Palaeoenvironmental changes in the northernboreal zone of Finland: local versus regional

drivers

Shyhrete Shala

Department of Physical Geography and Quaternary GeologyStockholm University

Stockholm 2014

© Shyhrete ShalaISSN: 1653-7211ISBN: 978-91-7447-838-9Cover: Shyhrete Shala. Back cover portrait: Alistair Auffret.Layout: Summary and Papers II & III: Shyhrete Shala based on LATEX templates by Rickard Petterson.Published papers typeset by respective publishers, reprinted with permission. Graphics created using AdobeIllustrator, Grapher, R and Tilia.Printed by Universitetsservice US-AB, 2014.

Per familjen time.Per nenen dhe babain, vellezerit, motrat dhe Seadin e vogel.

Per gjyshen, te cilen e kam ne qiell, dhe gjithe ata qe ende me gezojne ne jete.„

Doctoral dissertation 2014Shyhrete ShalaDepartment of Physical Geography and Quaternary GeologyStockholm University

Abstract. In this work, a variety of geochemical and biotic proxies derived from theLake Loitsana sediment sequence (NE Finland) are employed to i) determine the timing ofdeglaciation, ii) characterise an early Holocene proglacial lake stage, iii) reconstruct Holocenelake development, iv) identify local-scale/site-specific processes that could cause shifts in bi-ological assemblages and v) assess the most probable Holocene mean July air temperature(T jul) development. The use of multiple proxies limits the uncertainties related to single-proxy approaches and facilitates the discrimination between local and regional driving fac-tors. The results are then used to assess possible implications for interstadial environmentaland climate reconstructions derived from older deposits in the adjacent Sokli basin. LakeLoitsana and its surroundings were deglaciated prior to 10 700 cal. a BP. The sedimentrecord reflects a shallowing lake with four local events; i) the presence of a deep glacial lake,ii) glacial lake drainage and formation of Lake Loitsana, iii) changes in fluvial input due toprogressive wetland expansion and iv) the gradual infilling of the lake. The results suggestthat local events have driven changes in fossil assemblages through a variety of processes,and that the biological assemblages reflect changes in different environmental parameters ina highly individual manner. This further emphasises the importance of using multiple prox-ies in palaeoenvironmental studies. Biological assemblages can themselves act as importantdriving agents promoting changes in the composition of other assemblages. It is thereforesuggested that future studies should consider for instance macrophyte abundance as wellas food-web interactions as equally important physical parameters/factors, when assess-ing changes in biological assemblages. This study also assesses the impact of locally drivenchanges on quantitative T jul reconstructions derived from different biological assemblages.While pollen-inferred temperatures follow the classical trend of gradually increasing earlyHolocene T jul with a mid-Holocene maximum July warming, the plant macrofossil and chi-ronomid assemblages reconstruct highest T jul during the early Holocene, i.e. at the peakof summer insolation. The relatively low early Holocene July temperatures recorded by theterrestrial pollen are the result of site-specific factors possibly combined with a delayed re-sponse of the terrestrial ecosystem compared to the aquatic ecosystem. The diatom-basedT jul reconstruction seems to be particularly affected by secondary variables and display mi-nor variability throughout the Holocene. Local scale events have a potential to drive changesin both biological assemblages and water chemistry, which can be reflected in the transfer-function based regional temperature reconstructions. The minimum T jul reconstructed onthe basis of indicator plant taxa seem to be reliable, however one should keep in mind thatabsence of a certain taxa during any stage of lake development cannot be interpreted directlyas cooling climate.

Keywords: Palaeolimnology, Palaeoecology, Biological assemblages, Sediment geochem-istry, Regional drivers, Local drivers, Quantitative temperature reconstructions, Transfer-functions, Holocene, MIS-5d-c, Northern boreal forest zone, NE Finland.

Sammanfattning. Eftersom instrumentella matningar sallan stracker sig langre tillbaka itiden an 200 ar behovs data som rekonstruerar miljoforandringar pa langre sikt for att vi skakunna forsta miljo- och klimatforandringar samt sarskilja den antropogena effekten. Exem-pel pa anvandningsomraden for resultaten av denna studie ar utvarderingen av framtida kli-matforandringsscenarier samt hur langtidsforvaringen av avfall fran karnkraftsreaktorer kanpaverkas av forandringar under glaciationscykler. I denna studie har pollen, vaxtmakrofossil,chironomider (mygglarver) och diatomeer (kiselalger), i kombination med sedimentgeokemi,analyserats fran en sedimentkarna tagen i sjon Loitsana (i Sokli, nordostra Finland). Syftetvar att i) bestamma tidpunkten for isavsmaltningen, ii) karakterisera en tidig holocenproglacial sjo iii) rekonstruera den holocena sjons utveckling, iv) identifiera lokala processersom kan orsaka forandringar i den biologiska artsammansattningen och v) bedoma den mesttroliga utvecklingen for julimedeltemperaturer under holocen. Genom att kombinera flerametoder begransas de osakerheter som uppstar vid anvandningen av en proxy och det un-derlattar differentieringen mellan de lokala och regionala processer som driver forandringari den biologiska artsammansattningen. Resultaten fran denna studie anvands aven for attbedoma eventuella konsekvenser for tolkningen av sediment fran MIS 5d-c (Brørup).

Deglaciationen vid sjon Loitsana och dess omgivningar agde rum strax innan 10 700 kali-brerade ar BP (kal. ar BP, d.v.s. fore 1950). Den holocena miljorekonstruktionen aterspeglaren komplex sjoutveckling med fyra lokala handelser; i) forekomsten av en djup glacialsjo, ii)draneringen av sjon och uppkomsten av sjon Loitsana iii) forandringar i fluvialt inflodepa grund av utbredning av narliggande vatmarker och iv) en successiv igenfyllning avsjobassangen. Resultaten tyder pa att dessa lokala handelser och deras associerade pro-cesser har drivit forandringar i biologiska artsammansattningar. Det pavisades aven attvarje biologisk proxy aterspeglar miljoforandringar pa ett individuellt satt vilket ytterli-gare betonar vikten av att anvanda en kombination av dessa vid rekonstruktioner av aldremiljoer. Studien visar ocksa att akvatiska vaxter sjalva kan verka som drivande faktoreroch framja forandringar i sammansattningen av andra biologiska grupperingar. Av den an-ledningen foreslas att framtida studier aven bor inkludera akvatiska vaxtmakrofossil vidbedomningen av forandringar i den biologiska artsammansattningen.

Denna studie granskar aven effekterna av ovannamnda lokalt drivna forandringar forkvantitativa rekonstruktioner av julimedeltemperaturer. Den pollenbaserade rekonstruktio-nen foljer den klassiska trenden med relativt laga temperaturer innan 8200 kal. ar BP ochen maximal uppvarmning under 8200-4200 kal. ar BP. Temperaturrekonstruktioner baser-ade pa makrofossil och chironomider visar emellertid en nastintill motsatt trend med hogatemperaturer (ądagens varde) aven under den tidiga delen av holocen. De relativt lagatemperaturerna som rekonstrueras av pollendata kan vara resultatet av lokala faktorer,som paverkar den fossila datauppsattningen, eventuellt kombinerat med en fordrojd re-spons av det terrestra ekosystemet jamfort med det akvatiska ekosystemet. Diatomeer arspeciellt paverkade av lokala faktorer och uppvisar minimala temperaturforandringar genomhela holocen. Lokala processer har en potential att driva forandringar i biologiska artsam-mansattningar, vilket kan reflekteras i regionala temperaturrekonstruktioner baserade pa sakallade transfer-funktioner. Minimum julitemperaturer baserade pa vaxtmakrofossil av in-dikationsarter fran akvatiska och vatmarksmiljoer, verkar vara tillforlitliga. Dock bor dethallas i minnet att franvaro av en art under nagot stadium av sjoutvecklingen inte direktkan tolkas som en indikation pa kallare klimat.

Palaeoenvironmental changes in the northern boreal zoneof Finland: local versus regional drivers

Shyhrete Shala

List of papers ˚

This doctoral dissertation consists of this summary and the following papers, whichare referred to by their Roman numerals in the text.

I Shala S., Helmens K.F., Jansson K.N., Kylander M.E., Risberg J. andLowemark L. in press. Palaeoenvironmental record of glacial lake evo-lution during the early Holocene at Sokli, NE Finland.Boreas (2013), 10.1111/bor.12043. ISSN 0300-9483.

II Shala S., Helmens K.F., Valiranta M., Luoto T.P., Weckstrom J., Salo-nen J.S. and Kuhry P. Evaluating environmental drivers of Holocenechanges in water chemistry and aquatic biota composition at Lake Loit-sana, NE Finland.Journal of Paleolimnology, accepted pending revisions.

III Shala S., Helmens K.F., Luoto T.P., Salonen J.S., Valiranta M., and Weck-strom J. Assessment of quantitative Holocene temperature reconstruc-tions using multiple proxies from the Lake Loitsana sediment record.Manuscript to be submitted to The Holocene.

IV Helmens K.F., Valiranta M., Engels S. and Shala S. 2012. Large shifts invegetation and climate during the Early Weichselian (MIS 5d-c) in-ferred from multi-proxy evidence at Sokli (northern Finland).Quaternary Science Reviews 41, 22-38.

˚Author contributions. The following people have contributed to this thesis: Jan Risberg (JR), J. Sakari Salonen (JSS), JanWeckstrom (JW), Karin F. Helmens (KFH), Krister N. Jansson (KNJ), Ludvig Lowemark (LL), Malin E. Kylander (MEK),Minna Valiranta (MV), Peter Kuhry (PK), Stefan Engels (SE) and Tomi P. Luoto (TPL). SS = Shyhrete Shala. All co-authorsare acknowledged for reading and commenting on the manuscripts.

I Conceived and designed by SS, KFH and JR. Written by SS and edited by KFH. KNJ isacknowledged for the DEM-based glacial lake reconstruction. MEK is acknowledgedfor assessing the quality of raw data obtained through XRF core scanning and theintroduction to principal component analysis. JR is acknowledged for supervisionand assistance during the diatom analysis. LL is acknowledged for assistance duringthe XRF core-scanning.

II Conceived and designed by SS and PK after discussions with KFH and MV. Written bySS under the supervision of PK and edited by KFH, MV and PK. TPL, JSS andJW are acknowledged for pollen analysis, chironomid analysis and producing thediatom-inferred pH reconstruction, respectively.

III Conceived, designed and written by SS after discussions with KFH and MV. TPL, JSSand JW are acknowledged for the transfer-function based temperature reconstruc-tions and compositional distance estimations using chironomids, pollen and diatoms,respectively.

IV Conceived and designed by KFH. Written by KFH, MV and SE. SS responsible for diatomanalysis and interpretation of results as well as preparation of the location figure.

Abbrevations

AMS accelerator mass spectrometry

C/N carbon/nitrogen ratio

cal. a BP calibrated years before present (BP = 1950)

CONISS constrained incremental sum-of-squares

Corg organic carbon

DEM digital elevation model

DOC dissolved organic carbon

FIS Fennoscandian Ice Sheet

IPCC intergovernmental panel on climate change

ka BP kilo years before present

LOESS locally weighted scatterplot smoothing

LOI loss-on-ignition

LWWA locally weighted weighted averaging

MIS marine isotope stage

OSL optically stimulated luminescence

PAR pollen accumulation rates

PCA principal component analysis

RMSEP root-mean-square error of prediction

SCM Sokli carbonatite massif

SKB Swedish nuclear fuel and waste management company

T jul mean July air temperature

TOC total organic carbon

WA weighted averaging

WA-PLS weighted averaging-partial least squares

XRF X-ray fluorescence

PALAEOENVIRONMENTAL CHANGES IN THE NORTHERN BOREAL ZONE

Introduction

In the light of recent debate regarding global warm-ing, the demand for data that reflect long-term nat-ural environmental variability has increased. As in-strumental records are relatively short and rarelystretch back in time more than 200 years, our under-standing of the nature and rate of climate change,and thus our ability to differentiate anthropogeniceffects is highly dependent on palaeoenvironmentalstudies (Bradley 1999, IPCC 2007). Recent advance-ments in method development and dating techniqueshas led to a rapid accumulation of studies using vari-ous archives (e.g. ice cores, deep-sea cores, lake sedi-ments, peat cores, tree-rings, speleothems) to recon-struct palaeoenvironmental and climate change. Thepronounced sensitivity of the high latitude/altitudetree-line ecotone to global climate change, has pro-moted research focused on these areas in particularduring the last decades (e.g Hyvarinen and Alhonen1994, Karlen et al. 1995, Kullman 1995, Karlen andKuylenstierna 1996, Barnekow 1999, Eronen et al.1999a, b, Korhola et al. 2000, 2002, 2005, Seppa andHammarlund 2000, Rosen et al. 2001, 2003, Seppaand Birks 2001, 2002, Bigler et al. 2002, 2003, Seppaet al. 2002, Kultti et al. 2003, Larocque and Bigler2004, Larocque and Hall 2004, Bjune et al. 2004,Rosqvist et al. 2004, 2007, Bergman et al. 2005,Bjune 2005, Nyman et al. 2005, Valiranta et al.2005, Velle et al. 2005a, Larsen et al. 2006, Sarmaja-Korjonen 2006, Allen et al. 2007, Sundqvist et al.2007, Jonsson et al. 2009, Grudd et al. 2002, Paus2010, Reuss et al. 2010, Weckstrom et al. 2010, Joneset al. 2011, Paus et al. 2011, Salonen et al. 2011,Self et al. 2011). Consequently, palaeoenvironmentalstudies from the northern boreal zone of Fennoscan-dia, i.e. from sites that have not been affected by tree-line fluctuations, are few (e.g. Sarmaja-Korjonenand Hyvarinen 1999, Barnekow and Sandgren 2001,Bjune et al. 2004, Siitonen et al. 2011, Valiranta etal. 2011, Birks et al. 2012).

The northern boreal forest is a wide vegetationzone that covers large parts of Alaska, Canada, thesouthern tip of Greenland, most of Iceland, north-ern Fennoscandia and large parts of northern Russia(Fig. 1). An extensive circum-arctic tree-line ecotone(13 500 km long) is situated between the boreal for-est and the tundra to the north. Species diversity ishigher in the boreal zone compared to the tundrabut lower compared to the more southerly vegeta-tion zones (Waide et al. 1999).In Fennoscandia, thenorthern boreal forest has relatively low tree speciesdiversity and is dominated by pine, spruce and birch.The number of herbs and ferns is also low and the

ground vegetation is usually richer in bryophytes andlichens than vascular plants (Esseen et al. 1997). Awarming in climate is predicted to affect high lat-itudes by causing northwards shifts in the tree-lineecotone and migration of species from south to north,which would lead to higher species diversity in thenorthern boreal zone (Root et al. 2003). Such shiftsare easily studied in Finland due to its relatively flattopography compared to the rest of Fennoscandia.Even small changes in altitude in such areas, wouldcause a more uniform latitudinal shift in climate andsubsequent regional vegetational response (Esseen etal. 1997).

Long-term natural environmental variability inFennoscandia has been the subject of numerouspalaeoenvironmental studies in the 20th century(e.g. von Post 1916, Cleve-Euler 1922, Renberg andHellerberg 1982, Robertsson and Garcıa Ambrosiani1988, Garcıa Ambrosiani 1991). While this has im-proved our understanding of e.g. vegetation suc-cessions through glacial-interglacial cycles (Iversen1957, Andersen 1994) and related forcing factors,i.e. astronomical cycles (Imbrie et al. 1984), ourunderstanding of driving mechanisms operating atsmaller scales (centennial/millennial) remains rela-tively limited. Although the use of lake sediments,which has been the main source of palaeoenviron-mental information, stretches far back in time (e.g.Agassiz 1850), most of the advancements in palae-olimnology (i.e. the study of lake deposits to in-fer past limnological changes) have taken place inthe recent decades (e.g. Smol 1980, Hodgson et al.1997, Battarbee 2000). Palaeoenvironmental studieson lake deposits have provided records of sedimentgeochemistry (LOI, TOC, C/N, XRF), biological as-semblages (pigments, pollen, macrofossils, chirono-mids, cladocera, diatoms), stable isotopes (18O, 13C,15N) and lithological characteristics (grain-size, lam-inations). These records have in turn been used to re-construct complex lake histories and to make conclu-sions of Holocene environmental and climate changeand the effects of in-lake processes on biological as-semblages (Birks et al. 2000, Velle et al. 2005, Ander-son et al. 2008). Similar studies and especially studiesincluding plant macrofossils are scarce in northernFennoscandia, and notably few macrofossil recordsare available from both the northern boreal zone aswell as the tree-line ecotone (Barnekow and Sand-gren 2001, Bjune et al. 2004, Valiranta et al. 2005,Valiranta et al. 2011, Siitonen et al. 2011, Birks et al.2012). Most studies in this region have focused on theHolocene climate development using primarily bioticproxies (e.g. Rosen et al. 2001, 2003), but often onlyone or two biological assemblages have been studied(e.g. Korhola et al. 2000, 2002, Larocque and Bigler

1

S. SHALA

Arctic Circle

Study area

B 500 kmB

20°E 40°E

70°N

60°N

Finland

Sweden

Swed

en Finlan

d

A

4**1

*3

*5

*6

*8*

2 *9*9

*16

*17

*15

*18*

19

* 10-13* **

*21

*22

*7

*14

*20

Boreal fore

st zo

ne

No

rthe

rn b

orea

l for

est z

one

Tree

-line

Figure 1. Map showing the position of present tree-line in the northern hemisphere and Fennoscandia (modified after

Rekacewicz 2005). The different green shadings depict the boreal forest zone and northern boreal forest zone (modified

after Hamet-Ahti 1981 and Moen 1998). Location of study site is indicated in red whereas the location of sites mentioned

in the discussion are numbered in the order they appears in the text.

2004, Larocque and Hall 2004, Seppa et al. 2009). Itshould also be noted that only few of these recordsdate back to the earliest part of the Holocene, i.e.prior to 10 000 cal. a BP (Kujansuu et al. 1998, Sze-roczynska et al. 2007).

Research objectives

The overall aim of this thesis was to reconstructpalaeoenvironmental variations based on the LakeLoitsana sediment sequence (NE Finland), whichis located in the northern boreal forest zone ofFennoscandia (Fig. 1), and covers most of theHolocene. Main focus is on the Holocene, particu-larly the early and mid-Holocene (prior to 4200 cal. aBP), but also include data derived from the adjacentancient sediment basin where the sediment sequencerepresents MIS 5d-c (c. 110–94 BP; Martinson et al.1987).

The present study is part of a larger projecton the late Weichselian-Holocene climate and en-vironment changes at nearby Sokli, funded by theSwedish Nuclear Fuel and Waste Management Com-

pany (SKB). The aim of this project is to reconstruct

depositional environments, vegetation development

and climate changes in northeastern Finland during

a series of ice-free periods. Results from this study

will contribute to assessments of long-term safety of

a planned repository for used nuclear fuel in Sweden

(SKB 2006), since these safety assessments focus on

a time period of 100 000 years and more, thus cov-

ering the timescale upon which Quaternary ice-sheet

cycles have operated. The effect of glacial conditions

upon repository safety needs to be analysed and eval-

uated. In this context, characterising environmental

conditions following deglaciation is of particular in-

terest. Previous palaeoenvironmental studies at Sokli

have focused on MIS 3 (c. 59–24 ka BP) and MIS 5d-

c (c. 110–94 ka BP) interstadial stages (e.g. Helmens

et al. 2007a, b, 2009, Valiranta et al. 2009). This

work concentrates on a detailed multi-proxy envi-

ronmental record of the Holocene epoch, which will

provide important background information to un-

derstand the much longer-term, glacial-interstadial-

interglacial, environmental changes. Questions that

were addressed in this work included:

2

PALAEOENVIRONMENTAL CHANGES IN THE NORTHERN BOREAL ZONE

a) Lake development

ą What was the timing of early Holocenedeglaciation?

ą What were the characteristic biologicaland sedimentological features of the earlyHolocene glacial lake stage?

ą What local processes can be identified inthe historical development of the subsequentLake Loitsana basin?

b) Regional climate

ą Are locally driven changes in biologicalassemblages reflected in transfer-functionbased regional temperature reconstructions?

ą In light of proxy evidence, what is the mostprobable regional climate development dur-ing the Holocene?

c) Implications for glacial-scale environmental de-velopment at Sokli

ą How does the Holocene record at LakeLoitsana support the interpretation ofdeposits from MIS 5d-c and other stagesfrom the last glacial-interglacial cycle atnearby Sokli?

To achieve these goals high-resolution records ofsediment geochemistry (LOI, C/N, XRF), siliceousmicrofossils (diatoms, phytoliths and chrysophytecysts) and macrofossils were compiled from an un-usually thick (9m) lake deposit accumulated in LakeLoitsana, NE Finland, that covers most of theHolocene. The Holocene chronology is based on 14CAMS dated terrestrial macrofossils from Lake Loit-sana.

In order to comprehensively reconstruct lake on-togeny and to assess the potential local-scale/site-specific processes that drive changes in water chem-istry and biological assemblages, the above men-tioned data were incorporated with chironomid data(Engels et al.submitted) and previously publishedrecords of pollen (Salonen et al. 2013) derived fromthe same sediment sequence. Quantitative T jul re-constructions, based on pollen, chironomid and di-atom records, were made using the transfer-functionapproach whereas minimum T jul was inferred fromindicator plant species identified in the macrofossilrecord (Salonen et al. 2013, this thesis). Biological as-semblages are often used to infer past changes in tem-perature, however, the complex nature of biotic re-sponses to environmental change has made transfer-

function based reconstructions challenging to inter-pret. In this work, particularly, the sensitivity of bi-ological assemblages to different environmental vari-ables is assessed and the most probable HoloceneT jul development at and around Lake Loitsana isdiscussed and evaluated against the evidence of localdevelopment of the lake and the regional vegetationdevelopment.

Study area

The study area, i.e. Lake Loitsana and the surround-ing wetlands and the Sokli basin, is situated in thevicinity of the deserted mining hamlet of Sokli in NEFinland (around 67˝491N, 29˝171E; Fig. 2). Regionalpresent-day vegetation is northern boreal forest withbirch (Betula pubescens and B. pendula), pine (Pi-nus sylvestris) and spruce (Picea abies) as the dom-inant tree species. The mean annual temperature inthe area is approximately -1 ˝C; mean January tem-perature is -13.8 ˝C, mean July temperature is 13.4˝C, and mean annual precipitation is approximately550 mm (Fig. 3). The Sokli area has in the last 40years attracted numerous geology researchers due toits rare local bedrock type (e.g. Paarma 1970, Pert-tunen and Vartiainen 1992). The underlying bedrockconsists of a Palaeozoic carbonate-rich magma in-trusion, the Sokli Carbonatite Massif (SCM; Fig. 2),which is enriched in phosphorous, niobium, iron, sul-phur and calcium and surrounded by bedrock con-sisting of crystalline rocks of the Precambrian Shield(Vartiainen 1980).

Previous research carried out on sediments recov-ered from the Sokli basin have focused on geomor-phology and stratigraphy (Johansson 1995, Helmenset al. 2000), optically stimulated dating (Alexander-son et al. 2008) and palaeoenvironmental reconstruc-tions on various time scales (Helmens et al. 2007a, b,2009, Engels et al. 2008, Bos et al. 2009, Valirantaet al. 2009). Extensive glacial lakes formed duringthe latest deglaciation, in front of the retreatingice margin of the Fennoscandian Ice Sheet (FIS),has been reconstructed by Johansson (1995) usingraised shorelines, outlet channels, the distributionof coarse-grained outwash and finegrained glacio-lacustrine sediments, and topographic data. The re-construction shows that glacial lakes covered vast ar-eas in northern Finland. Most of the lakes, however,were rather small and restricted to the deepest partsof river valleys. The Sokli Ice Lake in NE Finland isone of the glacial lakes reconstructed by Johansson(1995). The main morphological features used in thereconstruction, i.e. dry channels and deep canyonscarved out in the Precambrian crystalline rock, are

3

S. SHALA

interpreted to represent glacial lake drainage sites or

spillways. Minerogenic sediments underlying gyttja

or peat are considered to be of glacio-lacustrine ori-

gin due to their relatively fine-grained lithology (Jo-

hansson 1995). The Sokli Ice Lake is the only ‘type

4’ glacial lake, i.e. a shallow lake with several meters

thick bottom sediments and indistinct shore marks,

reported from NE Finland (Johansson 1995). Thistype of glacial lake was, however, widespread in Swe-den (Lundqvist 1972).

Lake Loitsana (67˝481182, 29˝161562E; 214 ma.s.l.) is situated in a former meltwater channel asso-ciated with northwest-southeast oriented eskers (Fig.2C). Eskers situated northeast, northwest and southof the lake consist of sorted sands that were assum-

LAKE

FENITE

CARBONATITE

ESKER

RIVER

CORING SITE

SOKLI CARBONATITE MASSIF

CONTOUR LINE (10m)

WETLAND

L E G E N D

Sokli

N u o r tt i

20 km

RU

SS

I AV ä r r i o t u n t u r i

5 5 2

29°17'E

67°4

9'N

B

C

FI N

LA

ND

Finland

AArctic circle

70°N20°E10°E

60°N

B

240

220

220

240

240

220

67°4

8'N

67°5

0'N

67°4

7'N

Sokli

oja

Sokli

C

Loitsana

Sokl

iaap

ax1

MIS 5d-c

67°4

9'N

Yli-Nuortti

x2

x3

Saunavuotso

Maskaselkä314

1 00.5 km1Kyörtesselkä

282

29°18'E29°16'E29°14'E 29°20'E

Figure 2. Orientation map showing A. the location of the study area; B. the regional setting; and C. the location of

Lake Loitsana, coring site and geomorphology of the area (modified from Johansson (1995)). The light grey shading

marks areas below 220 m a.s.l. whereas the darker grey shading marks higher elevations (10 m interval). X1 and X2

mark the location of sampling for basal peat dating from the Saunavuotso and Loitsana wetland, respectively. X3 marks

the location of borehole 902 from which a basal peat age was inferred for the Sokli aapa.

4

PALAEOENVIRONMENTAL CHANGES IN THE NORTHERN BOREAL ZONE

Mea

n m

onth

ly te

mpe

ratu

re (°

C)

Mea

n m

onth

ly p

reci

pita

tion

(mm

)1971-2000

J F M A M J J A S O N D 0

20

40

60

80

-20

-10

0

10

20

Figure 3. Climatogram showing mean monthly precipi-

tation and temperature for the time period of 1971-2000

for Savukoski, (67˝171N, 28˝101E, 193 m a.s.l.). Data were

obtained from the Finnish meteorological institute.

ingly deposited during the last deglaciation (Pert-tunen and Vartiainen 1992). The lake is at presentmostly fed by groundwater inflow from the esker(Risberg J, personal communication). An extensivemire complex of the aapa-type (i.e. patterned fen)borders the lake to the south. Lake Loitsana drainsinto the Sokli rivulet (Soklioja), which is part of theNuortti River drainage basin; the latter drains intothe White Sea (Fig. 2B). The size of Lake Loitsana is„0.08 km2, water depth is between 1 and 2 m and thewater pH is „7.4 with an alkalinity of 1.90 mmol l´1.Ptot and Ntot are 120 and 520 µg l´1, respectively.Water colour is ă10 mg Pt l´1 and turbidity 2.7FNU (Finnish Environmental Institute 2010). Waterchemistry was measured in April 23rd 2008 when thewater temperature was 0.3 ˝C.

The highly erosion sensitive nature of the magmaintrusion upon which Sokli is situated, has favouredthe formation of cavities in the bedrock. Sedimentsaccumulated within these depressions have thus beensheltered from ice-sheet erosion. The latest ice-sheetis also considered to have been cold based duringextensive periods (Kleman et al. 1999), minimisingbasal erosion. This has enabled the formation of acontinuous sediment succession, which is unusuallylong for Fennoscandia, representing the last c. 135000 years. On the contrary to earlier conclusions, ice-free conditions governed at Sokli during MIS 5d-c c.112–94 ka BP, MIS 5a c. 80–74 ka BP and MIS 3c. 59–24 ka BP (e.g. Helmens et al. 2000, 2007a, b,

2009, Alexanderson et al. 2008). Additionally, nu-merous studies (e.g. Olsen 1988, Hattestrand 2008,Lunkka et al. 2008, Salonen et al. 2008, Pitkaranta2009, Hattestrand and Robertsson 2010, Valiranta etal. 2012) have provided evidence indicating that theFIS of the last glacial was considerably more dynamicthan what is commonly accepted (e.g. Mangerud1991, Lundqvist 1992, Donner 1995).

Methods and material

The studies presented in this thesis are mainly basedon down-core fossil data extracted from the LakeLoitsana sediment core (Papers I-III). Paper IV isbased on the sediment record of the B-series bore-hole in the Sokli aapa. A summary of the methodsthat were used in each study (Papers I-IV) as wellas the sample interval and the time period that eachstudy covers are presented in Fig. 4. The methodsand analyses employed to reconstruct palaeoenviron-mental and palaeoclimatological variations from theLake Loitsana sediment record are presented below.

Sediment and sediment coring

Nine meters of lacustrine sediment were collectedwith a Russian peat corer (Jowsey 1966) from the ice-covered Lake Loitsana in spring 2008 (67˝481182N,29˝161562E; 214 m a.s.l.). These sediments wereused in the palaeoenvironmental reconstructions pre-sented in papers I-III. The sediment succession con-sists of two major parts: a lower minerogenic unit(900–735 cm below sediment surface) and an upperorganic unit (735–0 cm). The upwards-fining minero-genic unit consists mainly of fine sandy sediment(900–795 cm) and silt (795–735 cm) with increasedclay content at 875–843 cm. The minerogenic unit at900–795 cm contains nine sand layers (1–2 cm thick)and a 5 cm sand layer between 800 and 795 cm. Thetransition from the minerogenic to the organic-richunit at 735–700 cm indicates a gradual change fromsilt with organic components to silty gyttja. The up-per organic unit consists of 700 cm gyttja with alter-nating dark and light-brown layers and intervals withincreased sand content at 700–635, 538–535, 377–340and 70–35 cm below sediment surface. Diffuse lami-nations occur between 900 and 770 cm. Continuouslaminations occur from 770 to 538 cm but are moredistinct at 770–735 and 575–538 cm.

Six and a half meters of lacustrine sediment, sit-uated 13.5–20.0 m below the wetland surface, werecollected from the B-series borehole in the Sokli aapain winter, 2002 (67˝481322N, 29˝181282E; „220 ma.s.l.). These sediments were used in the environmen-

5

S. SHALA

LOI

OS

L D

ATIN

G

1

2

3

4

5A

5B

5C

5D

5E

6

MARINE ISOTOPE

STAGES

LATE

HO

LOC

EN

EE

AR

LY H

OLO

CE

NE

MID

-HO

LOC

EN

E

0

2000

4000

6000

8000

10 000

AG

E (c

al. a

BP

)

A GE (

cal.

a B

P)

PAPER I

LOI

XR

F

10 0

0010

500

9500

LOI

C/N

PO

LLE

N

SIL

ICE

OU

S M

ICR

OFO

SS

ILS

MA

CR

OFO

SS

ILS

CH

IRO

NO

MID

S

AM

S 1

4 C D

ATIN

G

DIA

TOM

TJU

L

CH

IRO

NO

MID

TJU

L

PO

LLE

N T

JUL

M

AC

RO

FOS

SIL

TJU

L

SE

DIM

EN

T D

EP

TH (m

b.s

.)

14

15

16

17

18

19

XX

106

112

94

MIS

5C

MIS

5D

AM

S 1

4 C D

ATIN

G

SIL

ICE

OU

S M

ICR

OFO

SS

ILS

SIL

ICE

OU

S M

ICR

OFO

SS

ILS

MA

CR

OFO

SS

ILS

CH

IRO

NO

MID

S

AG

E (k

a B

P)

PAPER II PAPER III

PAPER IV

Figure 4. The time span of each paper, methods and sample resolution used in respective study. MIS stages are taken

from Martinson et al. (1987).

tal reconstruction presented in paper IV. The succes-sion consists of „2 m silty-sandy sediments, whichare overlain by a 2-m densely compacted laminatedsandy gyttja. The gyttja unit is disrupted by a hia-tus at 15.8–15.6 m depth. Layers of sand and gravelare found above and below this sediment succession.

Sediment geochemistry

LOI measurements. - In order to determine the or-ganic and inorganic carbon content of the sediment,LOI was performed at 550 ˝C for four hours and950 ˝C for five hours (Heiri et al. 2001) on samplesof similar size („1 cm3). The LOI (expressed as %dry weight) was calculated according to the followingequations:

LOI550 “ ppDW105 ´DW550q{DW105q ˚ 100 (1)

LOI950 “ ppDW550 ´DW950q{DW105q ˚ 100 (2)

The dry weight (DW) of the sample, after drying at105 ˝C, was measured before combustion (DW105),after heating to 550 ˝C (DW550) and after heatingto 950 ˝C (DW950).

C/N analysis. – Carbon and nitrogen concentrationswere measured through combustion in a Carlo ErbaNC2500 Elemental Analyzer at the Stable IsotopeLab at the Department of Geological Sciences, Stock-holm University. Freeze-dried sediment samples wereground and the material weighed into tin capsulesfor C/N-analysis. To account for carbonates in thesediment, material was also taken from all samplesfor Corg-analysis. The material was weighed into sil-ver capsules, treated with HCl (10%) and dried over-night before analysis. C/N ratios have proven effec-tive in revealing the source of the organic matter.As the aquatic flora has a high nitrogen composi-tion, samples with low C/N (ă10) ratios are gener-ally considered to indicate higher lake productivitywhile those with a high C/N ratio (ą20) could be in-dicative of low lake productivity. However, this ratiois also affected by the amount of terrestrial remainstransported into the lake and should therefore be in-terpreted with caution (Wetzel 2001).

XRF core scanning. – Recent advancements in theXRF technique have enabled nondestructive, in-situ XRF measurements at sub-millimetre scales

6

PALAEOENVIRONMENTAL CHANGES IN THE NORTHERN BOREAL ZONE

(Croudace et al. 2006). This has made XRF corescanning increasingly popular during the last decadefor the analysis of both marine and lacustrine sedi-ment cores (Francus et al. 2009, Croudace and Roth-well 2010, Lowemark et al. 2011). The increasingnumber of studies focusing on XRF core scanningdata derived from lacustrine sediment has linked sed-iment geochemistry to a wide array of variables suchas minerogenic input, grain-size changes, lake leveland productivity, and redox processes (Moreno et al.2007, Kylander et al. 2011, 2012, 2013, Vasskog etal. 2012).

The ITRAX XRF core scanner, which was usedin this study, produces ‘relative’ elemental data ex-pressed in peak area rather than ‘absolute’ measure-ments (i.e. concentrations) like those acquired usingconventional XRF analysis of distinct samples. In apalaeoenvironmental context, however, it is the rela-tive changes in the elemental profiles, which are of in-terest. Nevertheless, in order to avoid the closed sumeffect, normalisation of elemental data is required(Calvert 1983, Rollinson 1993, Weltje and Tjallingii2008). When working with homogenous sediments,normalisation by the incoherent + coherent scatter-ing has been used to correct for various instrumentalparameters such as dead times and tube aging. Whendealing with sequences with highly variable matriceshowever, dilution effects from e.g. carbonates and or-ganic matter (Lowemark et al. 2011, Kylander et al.2013) require data normalisation with an elementthat is conservative and biologically unimportant.One such element is Ti, which has been utilised toreveal patterns that are undetectable in single ele-mental profiles (Moreno et al. 2007, Kylander et al.2011).

In this study, XRF-data are used to detect en-vironmental and catchment changes recorded in theLake Loitsana sediments during the early Holocene.The cores were scanned using a molybdenum-tubeat 2 mm resolution (voltage 30 kV, current 45 mA)and 30 s measuring time (cf. Croudace et al. 2006).Focus lies on the minerogenic unit and the transitionto more organic sediments.

Biotic proxies

Fossil assemblages of pollen, macrofossils, diatomsand chironomids were extracted from the Lake Loit-sana record and used to reconstruct the lake on-togeny (paper II) and make transfer-function basedtemperature reconstructions (paper III). The re-sponse of diatoms to environmental change wasfurther assessed in paper I and IV. The analyseswere performed following standard procedures as de-scribed in Fægri and Iversen (1989) for pollen, Birks

(2007) for macrofossils, Battarbee (1986) for diatomsand Brooks et al. (2007) for chironomids. A combi-nation of these proxy assemblages was used in or-der to avoid uncertainties related to single-proxy ap-proaches and to facilitate the discrimination betweenlocal and regional drivers of changes in species dis-tribution and abundance.

Pollen records are commonly employed to recon-struct vegetational history and to infer past climatechanges. The interpretation of pollen rain can bechallenging, however, as it is composed of differ-ent proportions of local vegetation and long-distancetransported pollen (Birks and Birks 2000), and localpresence of certain species is difficult to deduce. Thisrelationship is particularly important in areas wherepollen production is low (e.g. the artic or alpine ar-eas) and where the long distant transported pollenmay dominate the pollen record.

Using a combination of macrofossils with pollenfacilitates not only the interpretation of local veg-etation development but often also yields a highertaxonomic resolution as plant macrofossils can oftenbe identified to genera or even species level (Birksand Birks 2000; Birks 2007). Macrofossil records areapart from being useful to reconstruct local vegeta-tion, also a valuable tool when addressing lake de-velopment, particularly when aquatic macrophytes,chironomids, diatoms and cladocera data are com-bined. These organisms are often found interactingwith each other (van Donk and van de Bund 2002,Velle et al. 2005, Siitonen et al. 2011, Valiranta etal. 2011). In this study, macrofossils are also usedto reconstruct minimum T jul based on presence ofwetland and aquatic indicator taxa and their cur-rent northern limit (Iversen 1954, Kolstrup 1979,Valiranta et al. 2009). Minimum T jul values of thesetaxa are inferred from current species distributionmaps (Lampinen and Lahti 2013) and the mean Julyair temperatures (1961–2000) calculated from dailymeasurements by Finnish Meteorological Institute(Venalainen et al. 2005).

Diatoms have a long track record in palaeoen-vironmental studies (e.g. Holst 1899, Sundelin 1917,Cleve-Euler 1922) and have been used to infer a widerange of environmental variables (pH, DOC, TOC,minerogenic turbidity, temperature) as they are par-ticularly sensitive to water body conditions and re-spond rapidly to environmental change (e.g. Renbergand Hellerberg 1982, Fritz et al. 1991, Wunsam andSchmidt 1995, Smol and Cumming 2000).

While it has become evident that diatom-basedHolocene temperature reconstructions can be rela-tively ambiguous (Bigler 2001), chironomids haveshown a particular prominence in this aspect and arecurrently widely used to reconstruct Holocene July

7

S. SHALA

air temperatures (e.g. Bigler et al. 2002, Korhola etal. 2002, Seppa et al. 2002). However, chironomid as-semblages are also affected by water body conditionssuch as oxygen concentration, trophic state and or-ganic carbon content (e.g. Velle et al. 2005, Nymanet al. 2008, Medeiros and Quinlan 2011). This hascaused discussions regarding the suitability of chi-ronomids in T jul reconstructions (Velle et al. 2010,2012, Brooks et al. 2012).

Data analysis

Principal Component Analysis (PCA). – Detectingtrends in compositional data is often not particu-larly straightforward. In such cases, using ordina-tion techniques can be useful as they reveal relation-ships among the data. There are several ordinationtechniques available, with or without stratigraphicalconstraint of the data. Stratigraphically constrainedordination techniques such as PCA, work by iden-tifying the first few axes, which capture the ma-jor variability of the data (Birks 2012 and refer-ences therein). This can be particularly useful whenworking with large data-sets such as the geochemicalrecord obtained through highresolution XRF core-scanning. To handle this data-set and detect poten-tial correlations between different elemental profiles,a PCA was made using JMP 9.0.0 software (correla-tion mode, varimax rotation). Since the original datawere initially normalised, the data-set became subse-quently ‘closed’. All elemental profiles were thereforelog transformed prior to analysis in order to avoidthe closure effect of compositional data (e.g. Aitchi-son 1982).

Transfer-functions. – The most commonly usedmethod to infer past changes in various environmen-tal variables (e.g. temperature) is the multivariatecalibration-function (Imbrie and Kipp 1971, Birks etal. 1990, ter Braak and Juggins 1993) more com-monly referred to as the ‘transfer-function’ approach(Birks et al. 2010). Shortly, this method involvestwo steps: the regression and the calibration. Theformer involves the establishment of modern rela-tionships between taxa and the parameter to be re-constructed. This requires a modern calibration set(c-set) with a number of calibration sites (for in-stance lakes) for which various environmental char-acteristics (e.g. depth, temperature, pH, conductiv-ity, alkalinity, number of taxa) have been measured.These sites are chosen along a gradient that coverschanges in the parameter of interest (pH, tempera-ture, TOC). During the calibration step, the estab-lished relationship is used in combination with thefossil data to infer past changes in e.g. temperature.

The calibration site where the modern assemblageresembles the fossil assemblage most yields the in-ferred temperature. The temperature estimate canbe based on assemblages from one calibration siteor it can be a weighted average estimate of severalcalibration sites. The transfer-function approach has5-6 basic assumptions. One of these assumptions isthat variables other than the one to be reconstructed,have insignificant influence (Birks et al. 2010 and ref-erences therein, Juggins 2013). This is, however, notalways the case when working with biological assem-blages, which can complicate transfer-function basedreconstructions of e.g. T jul (Anderson 2000, Rosenet al. 2003, Juggins 2013).

There are a number of various regression calibra-tion models available for quantitative reconstructions(Birks et al. 2010). Transfer-function based T jul re-constructions presented in paper III are prepared ac-cording to Salonen et al. 2013 (pollen) and Luotoet al. in press (chironomids). The diatom inferredT jul make use of a combined c-set (Weckstrom 2001,Weckstrom et al. 2003, Solovieva et al. 2005, 2008)with 178 calibration sites. A LWWA regression modelwith a selection of the 50 calibration sites that weremost similar to the fossil data in terms of speciescomposition is used in this reconstruction.

Zonation. – When working with stratigraphical se-quences, large data sets are common. In order tofacilitate the description and highlight correlations,fossil data are usually divided into zones. The tech-niques available today are divided in two categories;splitting or divisive and agglomerative techniques.The former works by splitting the data set intosmaller groups, while the latter clusters stratigraph-ically adjacent samples into larger groups (Birks2012). In this study, an agglomerative technique(constrained incremental sum-of-squares: CONISS)was used for pair-wise clustering of samples with thehighest compositional similarity. While zonation ofthe MIS 5c diatom record was performed in Tilia1.7.16 (Grimm 2011), numerical analyses related tothe zonation of all Holocene records were carried outusing R 2.12.2 (R Development Core Team 2011)with the additional packages Rioja (Juggins 2009)and Vegan (Oksanen et al. 2011). The broken-stickmodel was employed to estimate the number of po-tentially significant zones in the record (Bennett1996). Although numerical techniques provide resultsthat are repeatable and less subjective than the tra-ditional approach, i.e. visual examination of the data(Birks and Gordon 1985), the main patterns can becaptured by both approaches. The zones obtainedthrough numerical analyses were therefore used as

8

PALAEOENVIRONMENTAL CHANGES IN THE NORTHERN BOREAL ZONE

a guideline for the final zonation (i.e. lake stages),which was carried out visually in all studies.

Chronology

Dating control. – Age-depth estimations in lacustrinesediments are commonly established through radio-carbon dating. The method involves measurementof carbon isotopes from e.g. dead plant tissue, moreparticularly the procedure measures the relation be-tween unstable 14C, which decreases at a constantrate, and the stable isotopes 12C or 13C. These threeisotopes are naturally occurring in the atmosphereand are incorporated into plants through uptake ofCO2 during photosynthesis (Hajdas 2006). The ratioof 14C to 12C or 13C in the plant tissue is thereforein equilibrium with that of the atmosphere until thetime of death (Libby et al. 1949) after which 14C isreduced with time. As the half-life of 14C is known, itis possible to date the timing of death by measuringthe ratio of 14C against the stable forms of carbon.However, variations in the production of these iso-topes vary both geographically and over time andcalibration curves need to be used to convert radio-carbon ages to calendar years (Olsson 1991). Thisconversion is often challenging due to wiggles on the14C calibration curve. Moreover 14C age plateaus cancause converted ages that have a probability distri-bution spanning over several hundred years (Hajdas2006, Reimer et al. 2009). Other possible problemswith the radiocarbon dating method include contam-ination by old inorganic carbon or soil carbon fromthe catchment that is incorporated in the tissues ofaquatic macrophytes and brown mosses thereby re-sulting in too old ages (MacDonald et al. 1991).

The radiocarbon calibration curve for the north-ern hemisphere (IntCal09) that now extends back to50 000 cal. a BP, is based on empirical tree ring data,which provide a direct measure of atmospheric 14C,and reservoir-age-corrected marine data (Reimer etal. 2009). Age determination for the Holocene sed-iment succession was therefore established throughthe use of AMS radiocarbon dating. The age of theMIS 5c-d sediment record, however, was determinedusing OSL dating, as its age beyond the limit of ra-diocarbon dating (Alexanderson et al. 2008). OSLdating has developed rapidly over the last decadesand is now widely used to date various sediment de-posits (e.g. Murray and Wintle 2000, Alexanderson2002, Duller 2006, Preusser et al. 2009, Blomdin etal. 2012). Basically, OSL dating works by measur-ing a luminescence signal, in minerals such as quartzand feldspar, which is sensitive to both light andheat. When sediments are exposed to sunlight orheat this signal begins to deteriorate and is even-

tually erased (also known as sediment bleaching). Inorder for OSL dating to work, it is essential for thesediments to become completely bleached and thenburied. The OSL date is then derived from the lu-minescence signal, which begins to accumulate oncethe sediment is buried and yields the age of sedimentburial (Preusser et al. 2008). The dating of the MIS5d-c sediments was performed on quarts grains ex-tracted from samples consisting of fine medium sand(Alexanderson et al. 2008).

Calibration and age-depth modelling. – Radiocar-bon dating was performed by accelerator mass spec-trometry (AMS) on plant macrofossil remains inPoznan Radiocarbon Laboratory, Poland. Sampleswere washed in a fine jet of water through a 100-µmsieve. The residue was analysed in a Petri-dish forappropriate dating material, which was put in glassvials filled with distilled water and a small drop of10% HCl for preservation purposes. The radiocar-bon dates were calibrated into calendar years beforepresent (cal. a BP) using IntCal09 (Bronk Ramsey2009, Reimer et al. 2009) in R 2.12.2 software (R De-velopment Core Team 2011). A smooth spline modelwas applied to the dates using the CLAM 2.1 pack-age (Blaauw 2010) for R. Weighted average estimatesand sediment accumulation rates (cm a´1) are basedon both the entire calibrated distribution of all datesand the applied model. These point estimates are,according to Blaauw (2010), more likely to give arobust age estimate than e.g. the median as all dat-ing information is taken into consideration. The LakeLoitsana chronology is compared with two basal peatdates from the adjacent Saunavuotso and Loitsanawetlands (Fig. 2C: X1 and X2, respectively) and oneinferred date from borehole 902 (Helmens et al. 2000)in the Sokli aapa (Fig. 2C: X3).

Results

The next chapters provide summaries of four paperspresented in this thesis.

Paper I

Palaeoenvironmental record of glacial lakeevolution during the early Holocene at Sokli,NE Finland

Shyhrete Shala, Karin F. Helmens, Krister N.Jansson, Malin E. Kylander, Jan Risberg andLudvig Lowemark

A palaeoenvironmental record of glacial lake develop-ment, following the ice-marginal retreat of the FIS inNE Finland, is presented in this study. A 3-m-thick

9

S. SHALA

600

700

750

800

850

900DEPTH (cm)

650

LITHOLOGY

9410±50 BP8630±130 BP 9370±110 BP8450±70 BP14C AGES

10 5

00

10 6

00

10 7

00

10 4

0010

300

10 2

0010

100

10 0

0099

00

9800

9700

9600

9500AGE (cal. a BP)

ACCUMULATION RATE (cm a -1)

LAMINATIONSZO

NE

S

1a1b23

0

20

80

100LOI950 (% dw)

LOI550 (% dw)

RESIDUE

0

0.4

0.1

0.30.

3

1.6 1

3

0

0.02

1

6

Ca/

TiS

/Ti

Zr/T

iS

i/Ti

Sr/T

i

Coa

rse

/ med

ium

silt

Nb/

Ti

Loca

l bed

rock

inpu

tS

and

Not

ava

ilabl

e

Reworked / Tertiary

Planktonic

Tychoplanktonic

Benthic

Aerophilic

Rheophilic

Abundance (counts)

0

20

4

0

Species richness

2040Phytoliths

2040

6080

100

LAK

E D

EV

ELO

PM

EN

T

Acc

. rat

e (c

m a

-1)

Lam

inat

ions

less

dis

tinct

Abu

ndan

ceab

unda

ntoc

curin

gsc

arce

<10

or e

mpt

y0.

10.

2

Phy

tolit

hsab

unda

ntoc

casi

onal

rare

GE

OC

HE

MIC

AL

DA

TA (X

RF

CO

RE S

CA

NN

ER)

SIL

ICE

OU

S M

ICR

OFO

SS

ILS

Cor

ing

site

in a

rela

tivel

y sh

elte

red

emba

ymen

t of a

larg

e an

d de

ep g

laci

al

lake

. Sho

re e

rosi

on re

sults

in in

flux

of p

ooid

phy

tolit

hs a

nd re

wor

ked

diat

oms

from

till

as. T

urbu

lent

and

olig

o-eu

troph

ic la

ke w

ater

con

ditio

ns.

10 5

80 -

10 1

90 c

al. a

BP,

Sta

ge 3

10 1

90 -

9400

cal

. a B

P, S

tage

4In

crea

sing

sed

imen

tatio

n ra

te in

a s

mal

l, re

lativ

ely

deep

Lak

e Lo

itsan

a w

ith in

flux

of s

tream

ing

wat

er c

lose

to th

e co

ring

site

.S

tead

y in

crea

se in

sed

imen

t org

anic

mat

ter.

Hig

h ab

unda

nces

of d

iato

ms

and

mes

o-eu

troph

ic la

ke w

ater

con

ditio

ns.

c. 1

0 74

0 - 1

0 58

0 ca

l. a

BP,

Sta

ge 2

>10

740

cal.

a B

P, S

tage

1Ic

e-m

argi

n ne

arby

. H

igh

influ

x of

min

erog

enic

sed

imen

ts a

nd e

rosi

on o

f the

loca

l bed

rock

. C

ondi

tions

too

hars

h fo

r dia

tom

gro

wth

. Sh

allo

win

g an

d re

lativ

ely

smal

l gla

cial

lake

with

hig

h in

flux

of s

ilts

from

ex

pose

d la

ke s

hore

s an

d in

crea

sed

influ

ence

of t

he lo

cal b

edro

ck. R

isin

g or

gani

c m

atte

r in

the

sedi

men

t. C

ondi

tions

unf

avou

rabl

e fo

r dia

tom

gro

wth

.

% d

iato

ms

% s

ilice

ous

mic

rofo

ssils

San

d la

yers

Figure

5.

Sum

mary

of

pala

eoen

vir

onm

enta

ldata

and

stages

duri

ng

the

earl

yH

olo

cene

gla

cial

lake

dev

elopm

ent

and

init

iati

on

of

Lake

Loit

sana.

For

lith

olo

gy

legen

dse

eF

ig.

6.

10

PALAEOENVIRONMENTAL CHANGES IN THE NORTHERN BOREAL ZONE

sediment succession, of which 2 m are comprised ofglaciolacustrine sediment collected from Lake Loit-sana (a remnant of Sokli Ice Lake; cf. Johansson1995), was analysed using lithological characteristics,LOI measurements, XRF core-scanning and siliceousmicrofossils (diatoms, phytoliths). The aim was todetermine the timing of deglaciation, reconstruct thedevelopment of a relatively small glacial lake andstudy the environmental conditions and response ofthe aquatic ecosystem following deglaciation in theearly Holocene. The maximum possible extent of theglacial lake during different stages was explored andvisualised using a DEM. Although glacial lakes cov-ered extensive areas in north Finland (Johansson1995), few of these glaciolacustrine sediment succes-sions have been subjected to detailed analyses andenvironmental conditions following deglaciation re-main thus rather unstudied.

The results suggest that the sediment record, con-sisting of laminated silts grading into gyttja, reflectsthe different stages in glacial-lake evolution recon-structed on the basis of DEM and geomorphologi-cal data (Johansson 1995, this study). The sequenceis dated to the earliest part of the Holocene (Fig.5). Based on 14C dating of tree-type Betula seedsfound in the deep glacial lake sediments, which con-firms local presence of tree birch, it is concluded thatthe deglaciation of Lake Loitsana, NE Finland, oc-curred shortly prior to 10 700 cal. a BP. The com-bination of the retreating margin of the FIS and thesurrounding topography allowed for the formation ofa glacial lake that existed with various water depthsand surface extensions. The gradual retreat of theice margin from the coring site is reflected in thelithology, which shows a change from coarse-grainedsandy sediment to fine-grained, laminated silty sed-iment (Fig. 5). Deposition of gyttja initiated as theglacial lake drained and sedimentation continued inthe relatively small and shallow Loitsana depression.The proximity and subsequent retreat of the ice mar-gin is also clearly reflected in the XRF core scanningdata by changing contributions of elements typicalfor the local bedrock (Ca, S, Nb) to the geochemi-cal record (Fig. 5). After glacial lake drainage, thecatchment of Lake Loitsana became restricted to amuch smaller area overlapping with the SCM. Thisis reflected in an increase in e.g. trace element (Nb)concentrations in the lake sediment. Principal com-ponent analysis on the geochemical data further sug-gests that grain-size is an additional factor responsi-ble for the variability of the sediment geochemistryrecord. The initial deep-water glacial-lake stage wascharacterised by relatively high abundances of heav-ily silicified planktonic diatoms (Aulacoseira subarc-tica, Cyclotella meneghiniana) reflecting turbulent

water conditions (Lund 1954, Gibson et al. 2003).A shallow glacial lake stage with a restricted catch-ment was reconstructed following the deglaciation ofa major spillway, the Nuortti river canyon (Fig. 2C),which led to a lowering in lake level, possibly by sev-eral tens of meters. The subsequent final drainageof the glacial lake is marked in the diatom recordby distinct shifts related to morphometric eutroph-ication (e.g. dominance of eutrophic Stephanodiscusparvus among the planktonic diatoms). The resultsfurther suggest that the aquatic ecosystem respondedquickly to the changing environmental conditions.Diatom abundance and species richness initially in-creased with increasing distance between the ice mar-gin and the coring site. As silt eroded from exposedshorelines were deposited into the lake during theshallow water glacial-lake stage, diatom abundancessharply decreased and pioneering Staurosira venterdominated the turbid waters. A change to higherdiatom abundances and increased species richnessmarks the final drainage of the glacial lake and ini-tiation of gyttja deposition in Lake Loitsana.

Paper II

Evaluating environmental drivers ofHolocene changes in water chemistry andaquatic biota composition at Lake Loitsana,NE Finland

Shyhrete Shala, Karin F. Helmens, MinnaValiranta, Tomi P. Luoto, Jan Weckstrom, J.Sakari Salonen and Peter Kuhry

This comprehensive study based on sediment char-acteristics, geochemical data and biological assem-blages (pollen, macrofossils, chironomids and di-atoms) was conducted to reconstruct the develop-ment of Lake Loitsana and evaluate driving agentsto changes in water chemistry and biological assem-blages throughout the Holocene. Exhaustive recordsof macro- and microfossils of aquatic and wetlandplants, zoological taxa and diatoms are here com-pared to chironomid records (Engels et al. submit-ted) to facilitate the discrimination between localand regional factors and study potential interactionsbetween the aquatic biota. The presence of an earlyHolocene proglacial lake makes this study particu-larly interesting, as there are few studies from thenorthern boreal forest zone of Fennoscandia that in-clude such an initial development stage and deal withmultiple biological assemblages.

Ten samples from the Lake Loitsana sediment se-quence were dated (Fig. 6). Two samples were con-sidered to be too old and were excluded from the age-depth model. In contrast to the other dated samples,which were performed exclusively on terrestrial plant

11

S. SHALA

12 000 10 000 8000 6000 4000 2000 0

cal. yrs BP

0

100

200

300

400

500

600

700

800

900

DE

PTH

(cm

)16

30±3

514

C A

GE

S

LITH

OLO

GY

0 2.5

AC

CU

MU

LATI

ON

RAT

E (m

m a

-1)

AGE-DEPTH MODEL

Silty gyttja w.clay components

Gyttja

Gyttja w. sand components

Sandy gyttja

Silty gyttja/Gyttja silt

LEGEND

Algal gyttja

Sandy gyttja w. silt components

Gyttja w.silt components

Silty gyttja

Silty gyttja w.occasional sand

Silt

Sandy silt

Silt w. claycomponents

Sand

Silty sand

Sandy silt w. clay

9370

±110

9410

±50 8630

±13084

50±7

0

7960

±60

6020

±40

4530

±40

Figure 6. Age-depth model, lithology and accumulation rate for the Lake Loitsana sediment sequence.

remains, these two samples also contained unidenti-fied epidermal remains, most likely of aquatic originand were thus most probably contaminated by oldcarbon as Loitsana is a hard water lake with car-bonates in circulation (MacDonald et al. 1991). Thelowermost radiocarbon date (760–755 cm) of the re-maining eight samples provided an age of c. 10 730cal. a BP, thus confirming that the core representsmost of the Holocene epoch. No material suitablefor dating was obtained from the lowermost 140 cmof minerogenic sediment (900–760 cm). It can, how-ever, be assumed that these sandy sediments weredeposited at relatively high sedimentation rates andmost likely represent a short time period. The sedi-ment record suggests that the lake development wascomplex and included i) a proglacial lake stage ii)the subsequent glacial lake drainage and formationof Lake Loitsana, iii) a history of fluvial input af-fected by nearby wetland expansion, which causedthe redirection of a major streamlet and iv) lake in-

filling in an eventual groundwater-fed shallow lake(Fig. 7). These three local factors have in turn drivenchanges in the fossil assemblages by influencing wa-ter body conditions (pH, stratification, trophic state,turbidity, and turbulence), available substratum, flu-vial activity, rate of sediment influx, water depth andshore erosion.

The biological assemblages reflect changes in dif-ferent environmental parameters in a highly indi-vidual manner, which emphasises the importance ofusing multiple proxies in palaeoenvironmental stud-ies. The results further suggest that biological as-semblages can themselves act as important drivingagents in providing habitat and food thus promot-ing changes in the composition of other assemblages.This is particularly evident in Lake Loitsana whenconsidering aquatic macrophyte abundance, more ex-actly Myriophyllum, which in this study appears tobe an important driving factor for Corynocera am-bigua, a chironomid with a modern distribution in

12

PALAEOENVIRONMENTAL CHANGES IN THE NORTHERN BOREAL ZONE

1b23a3b4a4b56

ZON

ES

Species richness

1020

300

714% stream taxa

5010

0

Oligotrophic

MesotrophicEutrophicUndetermined

CH

IRO

NO

MID

S

Planktonic

Tychoplanktonic

AerophilicRheophilic

Reworked / Tertiary

DIA

TOM

S &

OTH

ER S

ILIC

EO

US M

ICR

OFO

SS

ILS

Benthic

MA

CR

OFO

SS

ILS

Phytoliths

Chrysophyte cysts

SE

DIM

EN

T G

EO

CH

EM

ISTR

Y

2010

1010

2050

100

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

1000

0

AGE (cal. a BP)

LITHOLOGY

4080

Organic carbon LOI550 (% dw)

0

LOI950 (% dw)

108

1216C/N

Diatom inferred pH7.

27.

610

3050

Diatom species richness

Potamogeton berchtoldiiPotamogeton filiformisPotamogeton pusillus

20

Sphagnum leaves

Myriophyllum remains

Equisetum

Potamogeton remains

BASAL PEAT AGESSaunavuotsoLoitsana wetland Sokli aapa

2040

20

Figure

7.

Sum

mary

of

mult

i-pro

xy

reco

rdof

the

Holo

cene

Lake

Loit

sana

sedim

ent

sequen

ce(m

odifi

edfr

om

pap

erII

).D

iato

ms

are

gro

up

edacc

ord

ing

to

livin

ghabit

at

and

rew

ork

ed/T

erti

ary

taxa;th

eabundance

ofphyto

lith

sand

chry

sophyte

cyst

sis

rela

tive

toth

esu

mofall

silice

ous

mic

rofo

ssils.

Chir

onom

ids

are

gro

up

edacc

ord

ing

toth

eir

trophic

state

pre

fere

nce

(Luoto

2011b);

oligotr

ophic

=5–15,

mes

otr

ophic

=15–25

and

eutr

ophic

=25–100

TP

µg/l.

The

per

centa

ge

of

stre

am

taxa

incl

udes

sem

i-te

rres

tria

lta

xa

(ă2%

abundance

).P

lant

macr

ofo

ssils

are

pre

sente

das

bars

show

ing

conce

ntr

ati

ons

per

10

cm3

or

as

rela

tive

abundance

of

rem

ain

s.F

or

adet

ailed

lith

olo

gy

see

Fig

.6

13

S. SHALA

Finnish lakes related to cold oligo-mesotrophic wa-ters (Luoto 2011b). It is concluded that even thoughlake and catchment development are linked to region-alscale factors, observed changes in lake biota alsoreflect local-scale processes within the lake (properstream re-direction and infilling) and the immediatesurrounding catchment area (wetland initiation andforest development). Taking local environmental fac-tors into account when conducting quantitative en-vironmental/climatic reconstructions based on bio-logical assemblages is therefore not only necessaryin order to assess their impact on these reconstruc-tions but might also provide an explanation to thedifferent responses of these proxies.

Paper III

Assessment of quantitative Holocenetemperature reconstructions using multipleproxies from the Lake Loitsana sedimentrecord

Shyhrete Shala, Karin F. Helmens, Tomi P. Luoto,J. Sakari Salonen, Minna Valiranta and JanWeckstrom

Four mean July air temperature (T jul) reconstruc-tions, based on biotic proxies extracted from theHolocene sediment sequence of Lake Loitsana, werein this study evaluated in relation to each other andlocal-scale/site-specific processes in the lake devel-opment. These latter factors have been shown tosignificantly influence the species distribution of thefossil assemblages. The aim was to assess the reli-ability of each reconstruction and the most likelyHolocene development of T jul, particularly the tim-ing of highest T jul. Similar approaches were usedwhere possible in order to make the transfer-functionbased reconstructions comparable with each other.Pollen and chironomid-based T jul was reconstructedusing a two-component WA-PLS regression calibra-tion model (ter Braak and Juggins 1993) and thediatom-based T jul was reconstructed using a LWWAregression model with classical deshrinking (see Jug-gins and Birks 2012). The reliability of the recon-structions was assessed using the compositional fit(squared chord distance) between fossil samples andtheir closest modern analogues in the c-set, thetemperature optima (WA-optima) of the dominat-ing taxa as well as the distribution of lakes alongthe temperature gradient of the c-set. MinimumT jul values were inferred based on the presence ofindicator plant taxa. The reconstructed T jul dis-play highly differentiating patterns throughout theHolocene (Fig. 8).

While pollen-based temperatures follow the clas-sical trend of relatively low early Holocene T jul and

a mid-Holocene maximum in July warming, the plantmacrofossils and chironomids reconstruct high T jul

already during the early Holocene, i.e. at the peak ofsummer insolation. The diatom-based reconstructiondisplays least variability, (only 0.6 ˝C amplitude)throughout the entire Holocene. The evaluation inrelation to local-scale/site-specific processes revealedthat the reconstructions are influenced at least tosome extent by local factors, which further empha-sises the advantage of using multiple proxies whenreconstructing climate variables. The poor composi-tional fit and low pollen-based T jul, reconstructedduring the early Holocene, is attributed to an over-representation of locally produced pollen of Cyper-aceae, Equisetum and Gramineae (syn. Poaceae) inthe fossil samples. These taxa are commonly foundin tundra vegetation and have thus low tempera-ture optima in the c-set. Minimum T jul values of 15˝C are inferred from the presence of plant macrofos-sils (Typha and Glyceria lithuanica) during the earlyHolocene. These temperatures might have prevaileduntil 6500 cal. a BP if the presence of Typha pollenis interpreted as local. The chironomid-based T jul

reconstruction appears to have been mainly affectedby processes such as macrophyte abundance, DOCand the general infilling of the lake. While shallowwater conditions, resulting from the general lake in-filling, caused an overestimation of T jul values dur-ing the late Holocene, macrophyte abundance andDOC appears to have favoured the abundance of coldwater Corynocera ambigua and cold stenotherm C.oliveri thus causing underestimation of the recon-structed T jul during mid-Holocene. However, maxi-mum warming based on chironomids is found con-sistently in the early Holocene irrespective of lo-cal lake development. The diatom-based T jul valueswere found to be underestimated mainly due to twofactors; the mass-occurrence of Fragilariaceae, whichconstitute more than 90% of the total abundance andhave a low WA-optima („11.5–12.5 ˝C), and the un-even distribution of calibration sites along the tem-perature gradient of the c-set with the majority ofthe c-sites within the range of 11.1–13.0 ˝C (Fig. 9).

It is concluded that although site-specific pro-cesses have clearly had a large effect on the relativelylow early Holocene pollen T jul, this might furtherhave been influenced by a delay in response of ter-restrial vegetation to climate change at Lake Loit-sana, NE Finland. The early- Holocene warming assuggested by the aquatic/wetland ecosystem in theLake Loitsana sequence (chironomids, macrophytes)seems to be reliable, indicating a T jul about 2 ˝Chigher, or more, compared to present-day.

14

PALAEOENVIRONMENTAL CHANGES IN THE NORTHERN BOREAL ZONE

2024

480

500

520

(W/m

2 )0

2 m

in T

jul (

°C)

0 2000 4000 6000 8000 10000-2

02

Tju

l (°C

)

60° N

INS

OLA

TIO

N

M. EU

TR.

AGE (cal. a BP)

JuneDecember

Pollen

sdi monori h C

Plant macrofossils

TEM

PE

RAT

UR

E A

NO

MA

LY F

RO

M P

RE

SE

NT-

DAY

Typha latifolia pollen

DEEP

SHAL

LOW

GLA

CIA

L

LAK

E

DEEP LAKE W. RIVER INPUT

MAINLY GROUNDWATER FED

SHALLOWING LAKE RIVE

R DI

VERS

ION

LAK

E D

EV

ELO

PM

EN

T

LOC

AL

STA

GE

S

123456

AB AB

NEARBY WETLAND DEVELOPMENT

L A K E L O I T S A N A

TER

RE

STR

IAL

VE

GE

TATI

ON

DE

VE

LOP

ME

NT

2040

6080

100

Shr

ubs

Her

bsFe

rns

Tree

s

Betula

Pinus

Picea

Alnus0

200

2040

020

EMUTESI UQ SNI A MER

CXERA DEES

EMUTESI UQ

CEAECAREPY

GAI RECYL ACI NAUHTI L DEES

SE

LEC

TED L

OC

AL

ELE

ME

NTS

GEAENI MAR

Diatoms

Figure

8.

Sum

mary

ofth

elo

caldev

elopm

ent

ofL

ake

Loit

sana

(as

reco

nst

ruct

edin

pap

erI

and

II).

Holo

cene

pollen

and

chir

onom

id-b

ase

dT

jul

reco

nst

ruct

ion

usi

ng

atw

o-c

om

ponen

tw

eighte

dav

eragin

g-p

art

ial

least

square

s(W

A-P

LS)

regre

ssio

n.

LO

ESS

smooth

er(s

pan

0.2

5,

one

robust

ifyin

git

erati

on)

isadded

to

both

Tjul

reco

nst

ruct

ions.

The

dia

tom

-base

dre

const

ruct

ion

usi

ng

loca

lly

wei

ghte

dw

eighte

dav

erage

(LW

WA

)w

ith

class

ical

des

hri

nkin

gis

dis

pla

yed

wit

ha

LO

ESS

smoth

er.

Sum

mary

dia

gra

mof

Holo

cene

dev

elopm

ent

of

terr

estr

ial

veg

etati

on

exte

nded

aft

erSalo

nen

etal.

(2013).

Sel

ecte

dp

ollen

(gre

ysi

lhouet

te)

and

macr

ofo

ssil

rem

ain

s(d

ash

edline)

of

taxa

dis

pla

yin

glo

cal

pre

sence

are

als

oin

cluded

.T

he

60°N

sum

mer

and

win

ter

inso

lati

on

(Ber

ger

and

Loutr

e1991).

15

S. SHALA

0

10

20

30

40

50

60

70

DiatomsChironomidsPollen7.

9-10

10.1

-11

11.1

-12

12.1

-13

13.1

-14

14.1

-15

15.1

-16

16.1

-17

17.1

-17.

3

Temperature range (°C)

Num

ber o

f cal

ibra

tion

site

s

Figure 9. The distribution of lakes along the sampled

temperature gradients of the diatom calibration set (7.9-

15.7 ˝C), chironomid calibration set (7.9-17.0 ˝C) and

pollen calibration set (9.0-17.3 ˝C).

Paper IV

Large shifts in vegetation and climate duringthe Early Weichselian (MIS 5d-c) inferredfrom multi-proxy evidence at Sokli (northernFinland)

Karin F. Helmens, Minna Valiranta, Stefan Engelsand Shyhrete Shala

In this paper, an extensive record consisting of mul-tiple proxies (macrofossils, chironomids, siliceous mi-crofossils, loss-on-ignition and lithological character-istics) is integrated and combined with chironomid-and indicator plant species-based palaeotempera-tures. The record was extracted from sediments ofMIS 5d-c age preserved in the Sokli basin and usedto infer the vegetation, depositional and climate his-tory at Sokli (NE Finland). The chronology wasconstrained with two OSL dates. The samples weretaken at 14 m and 15 m depth from sediment surface,and yielded ages of 94˘16 and 94˘19 ka BP (Alexan-derson et al. 2008). The results suggest highly dy-namic landscape changes. During MIS 5d the land-scape was characterised by a braided-river environ-ment, steppe-tundra vegetation and strong continen-tal conditions. This was followed by the developmentof an oxbow lake during MIS 5c; the ongoing ter-restralisation as a result of the gradual infilling led toa return to stream-channel deposition during the latepart of MIS 5c. A mixed boreal forest was developedas birch, pine and spruce (in that order) establishedin the area. The vegetation gradually opened up dur-

ing the late part of MIS 5c, which was followed by theMIS 5b stadial. This study also suggests mean T jul

of at least 12–14 ˝C during MIS 5d and of several de-grees higher than present day (13.4 ˝C) during MIS5c. The diatom assemblage presented in this studywas extracted from the laminated unit of sandy gyt-tja deposited during early and mid-MIS 5c (Fig. 10).

The record is characterised by mass-occurrence(ą20%) of Staurosira venter and S. construensthroughout the record and high abundances of Stau-rosira binodis, Staurosirella lapponica, S. pinnata,Pseudostaurosira brevistriata, Fragilariforma neo-producta, Martyana martyi and Staurosira binodis(henceforth Fragilariaceae). Fragilariaceae speciesare generally considered to be opportunistic andpioneering due to their wide range of ecologicalpreferences (Smol 1983, Anderson 2000). They arefavoured by relatively high alkalinity (Battarbee,1986) and are often found in lakes that have somesort of disturbance such as proglacial environments(Bigler et al. 2002, Risberg et al. 1999). Relativelyhigh abundances of phytoliths, chrysophyte cysts andTertiary diatoms during the early stages of MIS 5care associated with low concentrations of diatoms,most likely due to erosion caused by flooding duringthe earliest phase of infilling of the oxbow lake. Theon-going infilling of the lake eventually initiated ter-restralisation and a more diverse diatom flora devel-oped. An increased nutrient availability, reflected byincreasing abundances of e.g. Stephanodiscus parvusand Cocconeis placentula v. euglypta, and continu-ous fluvial influx (Meridion circulare) is recordedthroughout MIS 5c. Supplementary information as-sociated to this paper is available online at doi:10.1016/j.quascirev.2012.02.008.

Discussion

Holocene development of the Sokli Ice Lakeand Lake Loitsana

As the margin of the FIS retreated towards thenorth-west during the early Holocene deglaciation,meltwater was trapped between the ice-sheet andthe higher terrain to the east, and a glacial lake wasformed. The time period during which this glaciallake existed has been estimated to less than 100 years(Johansson 1995). Although the timing of deglacia-tion around Lake Loitsana roughly corresponds withthe early Holocene regional deglaciation pattern re-constructed for northern Finland (Johansson et al.2011), due to the low accuracy of the available datesfrom Lake Loitsana, which hampers precise age es-timation, it is difficult to estimate the exact the life

16

PALAEOENVIRONMENTAL CHANGES IN THE NORTHERN BOREAL ZONE

Sca

rce

Occ

urrin

gA

bund

ant

Dia

tom

abu

ndan

ce

Pla

nkto

nic

1

Tych

opla

nkto

nic

2

Ben

thic

3

Aer

ophi

lic 4

Rhe

ophi

lic 5

Rew

orke

d/Te

rtiar

y

Phy

tolit

hs

Chr

ysop

hyte

cys

ts

0 20 40 60 80 100 0 20 40 0 20 40

15.0

15.5

16.0

16.5

DEP

TH(m

)

Ste

phan

odis

cus

parv

us 1

Frag

ilaria

cap

ucin

a v.

cap

ucin

a2

Mel

osira

var

ians

2

Am

phor

a lib

yc a3

Coc

cone

is p

lace

ntul

a 3

Gom

phon

ema

angu

stat

u

m3

Nav

icul

a re

inha

rdti

i3

Tetra

cycl

us e

mar

gina

tus

3

Rei

mer

ia s

inua

t a4

Mer

idio

n ci

rcul

are

v. c

ircul

ar e5

Frag

ilarif

orm

a ne

opro

duct

a1

Mar

tyan

a m

arty

i1

20

Pse

udos

taur

osira

bre

vist

riat

a1

Sta

uros

ira

bi

nodi

s 1

20

Sta

uros

ira c

onst

ruen

s 1

20 40

Sta

uros

ira v

ente

r 1

20

Sta

uros

irella

lapp

onic

a1

20

Sta

uros

irella

pin

nat

a1

FR

AG

IL

AR

IA

CE

AE

DEPOSITIONAL

ENVIRONMENT

STREAM CHANNEL

STREAM CHANNEL

TE

RR

ES

TS

TR

AL

ISA

TIO

N

O X B O W L A K E

STACKED DIATOM RECORD SELECTED DIATOM TAXAOTHER SILICOUS REMAINS

Figure 10. Summary of the diatom record extracted from the MIS 5d-c sediment deposit. Cocconeis placentula is an

aggreggate of all placentula varieties.

span of the glacial lake itself. In general, the LakeLoitsana sediment record suggests a shallowing lakewith four limnological phases that had major im-pacts on the Holocene sedimentation dynamics, wa-ter chemistry and composition of biotic assemblages.These stages are i) a deep glacial lake phase ii) thedrainage of this glacial lake, which led to the forma-tion of a smaller and shallowing glacial lake and even-tually the much smaller Lake Loitsana, iii) a phasewith changes in fluvial input affected by nearby wet-land expansion, which caused the redirection of a ma-jor streamlet and iv) a gradual infilling of an eventualprincipally groundwater-fed lake.

Deep glacial lake phase. – The presence of aproglacial lake makes this study interesting as fewstudies from the northern boreal forest zone of Fin-land include such an initial developmental stage (Ku-jansuu et al. 1998; Fig. 1B: 1, Szeroczynska et al.2007; Fig. 1B: 2). These sediments are of importanceas they allow us to conduct high-resolution studiesof environmental and climate conditions directly fol-lowing the latest deglaciation. At high northern lat-itude sites, such as Lake Loitsana, this event coin-cides with the period of high summer insolation (Fig.8). The aquatic flora (i.e. siliceous microfossils, bry-ozoa) was during this time, mainly affected by physi-cal factors such as rate of sediment influx, turbulenceand possibly reduced light penetration caused by sus-pended minerogenic material (i.e. turbidity). For ex-

ample, low diatom abundances, associated with highaccumulation rates, have been recorded in sedimentsof Lake Inarijarvi in NE Finland (Kujansuu et al.1998), the Baltic Ice Lake (Kabailiene 1995; Fig. 1B:3) and Lake Agassiz in North America (Risberg etal. 1999; Fig. 1A: 4). Cold and deep-water associ-ated chironomid taxa Heterotrissocladius maeaeri -type and Tanytarsus lugens-type and bryozoan Fred-ericella indica inhabited the glacial lake; the formeris also reported to have dominated the deep wa-ters of the late-glacial Baltic Ice Lake (Luoto et al.2010; Fig. 1B: 5). The surrounding vegetation of thenewly deglaciated land was initially relatively openand characterised by shrubs (Betula nana, Salix ) andwetland taxa (Carex, Equisetum, Glyceria lithuanica,Juncus and Sphagnum). Tree Betula was present al-ready during the early Holocene, corresponding withprevious finds from northern Finland (Valiranta etal. 2005; Fig. 1B: 6) and northern Sweden (Barnekowand Sandgren 2001; Fig. 1B: 7).

Glacial lake drainage. – The glacial lake drainage ini-tiated at c. 10 500 and was completed at c. 10 200cal. a BP, which corresponds roughly to that of LakeInarijarvi further north (Kujansuu et al. 1998). Asthe proglacial lake was progressively drained, LakeLoitsana was confined to a topographical depressionand was affected by shore erosion from the newlyexposed areas. The subsequent more abundant pres-ence of littoral/wetland (Callitriche spp, Potamoge-

17

S. SHALA

ton spp, Myriophyllum, Typha) and terrestrial (Be-tula, Salix ) taxa in the macrofossil record might re-flect the closer proximity of the coring site to theshore (Hannon and Gaillard 1997). In the case ofLake Loitsana, however, changes in distance of thecoring site to the shore, appear not have been impor-tant for the presence of different macrofossils in thesediment record, probably due to the steep-slopednearby esker facing the lake. The increased abun-dance in macrofossils might thus also be the effect ofshore erosion. Relatively rich early Holocene aquaticvegetation was also found in the alpine Lake Njar-gajavri further north (Valiranta et al. 2005). The re-duction in lake volume following the final drainageof the glacial lake, combined with an increased inputof sediment, caused a morphometric eutrophication(cf. Hofmann 1998). This is mainly reflected by thecontinuous occurrence of the hypereutrophic diatomStephanodiscus parvus. High relative contributions oftrace elements in the sediment support the interpre-tation of increased input from the local Sokli Car-bonatite Massif.

Changes in fluvial input affected by nearby wetlanddevelopment. – Changes in surface inflow patternsinto to Lake Loitsana, have clearly affected thespecies composition of macrofossils, chironomids anddiatoms directly (input of rheophilic species) and in-directly (input of minerogenic particles/DOC). Theinflux was particularly pronounced between 10 200and 6800 cal. a BP. Fluvial input to Lake Loit-sana diminished after 6800 cal. a BP and is re-flected as a marked reduction in stream indicatorsin the sediment record. Input reduction correspondswith a wetland expansion along the lake’s south-ern shores, which resulted in a diversion of the Sok-lioja. The high carbon content, lack of fluvial influx(i.e. rheophilic diatoms and chironomids) and highamount of chrysophyte cysts further suggests thatthe lake might have been temporarily isolated be-tween 6800 and 6300 cal. a BP. This time period cor-responds roughly with lowest lake levels recorded inLake Jierstivaara and Lake Njargajavri, Finnish La-pland (cf. Korhola et al. 2005; Fig. 1B: 8, Valirantaet al. 2005). In the Lake Loitsana record, however,there is no clear stratigraphical evidence indicatinga shallow lake phase. Therefore, it may be that thetemporarily isolated phase is caused by local hydro-logical changes at the surroundings. Lake Loitsanabecame a mainly groundwater fed lake from c. 6300cal. a BP and received almost no surface inflow afterc. 4300 cal. a BP.

Lake infilling. – The progressively infilling Lake Loit-sana initially derived its sediments from a large

catchment area stretching beyond the boundariesof the local bedrock. Following the proglacial lakedrainage, these sediments originated from a muchmore confined catchment with strong carbonatite im-print. The mid-Holocene stream diversion resulted ina further confinement of the catchment area; afterthis the sediment input was mostly of very local ori-gin and sedimentation rates decreased significantly.The long-term natural acidification that is typical formost tree-line lakes in northern Fennoscandia (Ko-rhola and Weckstrom 2004, Sarmaja-Korjonen et al.2006; Fig. 1B: 6) was not observed in Lake Loitsanawhere the pH remained at „7.2–7.7 throughout theHolocene. It is possible that the alkaline nature of thebedrock kept the lake well buffered; however, pH ą7is not unique for Lake Loitsana as this has also beenobserved in other high latitude lakes such as alpineLake Somaslampi (Szeroczynska et al. 2007) and bo-real Lake Kipojarvi (Valiranta et al. 2011; Fig. 1B:9).

Detected changes in biological assemblagesand potential driving factors

High abundances of Fragilariaceae characterisedboth the Holocene (papers I-II) and the MIS 5d-c sediment records (paper IV). Observed genera ofthis family are generally considered to be oppor-tunistic and pioneering due to their wide range ofecological preferences and often dominate alkalinelakes that have some sort of disturbance, e.g. alpineproglacial lakes or recently deglaciated lakes withshallow photic zones (e.g. Battarbee 1986, Ander-son 2000, Battarbee 2000, Bigler et al. 2002; Fig.1B: 7, Ampel et al. 2009). In this study the mas-soccurrence of Fragilariaceae appears to be favouredby the alkaline water and disturbances through flu-vial/minerogenic influx as well as groundwater in-flow from the esker. Martyana martyi in particular,has a preference for high ion content (Witkowski etal. 1995/96). As boreal conditions have prevailed al-most since Lake Loitsana’s initiation, the length ofthe growing season or deficiency of nutrients haveprobably not been restricting factors affecting theaquatic macrophyte prevalence (cf. Valiranta 2006a,b; Fig. 1A: 10-13). In general, factors such as distanceof the coring site to the shore and fluvial activityappear to be of greater importance to the macro-phyte assemblages in the sediment record. This fur-ther highlights that absence of certain species in thefossil record should be carefully interpreted as it doesnot necessarily reflect its actual absence from thelocal vegetation (Birks 1973, Hannon and Gaillard1997, Valiranta 2005). In Lake Loitsana, however,changes in water depth did not result in a marked

18

PALAEOENVIRONMENTAL CHANGES IN THE NORTHERN BOREAL ZONE

alteration of the distance to the nearest shore dueto the presence of an esker in the vicinity of thecoring site with a steep side facing the lake. Thus,presumably, any changes in water level had no sig-nificant effect on macrophyte abundance. This is fur-ther supported by the more or less consistent pres-ence of littoral zone species such as Myriophyllum,Potamogeton and large terrestrial remains such asBetula throughout the record. An important aspectof aquatic macrophytes is that they can themselvesact as environmental drivers by providing habitatand food for other organisms. This is particularlyevident in the case of Corynocera ambigua, whichis associated with higher abundances of Myriophyl-lum. Corynocera ambigua has a modern distributionin Finnish lakes related to cold oligo-mesotrophic wa-ters but is known to have a complex ecology (Broder-sen and Lindegaard 1999). Psectrocladius sordidel-lus-type is another example of taxa associated withlittoral vegetation and water depth (Tolonen et al.2001; Fig. 1B: 14, Luoto et al. in press; Fig. 1B: 15)and appear also in the Lake Loitsana record to havebeen affected by macrophyte abundance. Other localfactors affecting chironomids species distribution in-clude water body conditions (e.g. streaming water,DOC). The presence of cold stenotherm Corynoceraoliveri -type in the sediment record coincides with in-creased fluvial activity and might thus be driven byincreased DOC contents as allochtonous carbon con-sists mainly of DOC (Jansson and Broberg 1994).Corynocera oliveri -type has, though generally con-sidered a cold water indicator, also been found inwarmer lakes (Palmer et al. 2002) as well as associ-ated to higher DOC concentrations in Canadian Arc-tic lakes (Gajewski et al. 2005, Medeiros and Quinlan2011).

Holocene climate development in NEFinland based on different proxy records

The rich and diverse fossil record extracted from theLake Loitsana sediment sequence allowed for T jul

reconstructions dating back to c. 10 800 cal. a BP. Acomparison of the reconstructed T jul deviation frompresent day mean July air temperature revealed thatboth the absolute values and the general trends differconsiderably between the different proxies (Fig. 8). Itis inferred that changes in biological assemblages, i.e.pollen, plant macrofossils, chironomids and diatoms,driven by local scale/site-specific processes have hadan influence on quantitative reconstructions (paperIII). The diatom-based T jul displays a stable tem-perature around 12 ˝C (i.e. below the present dayvalue of 13.4 ˝C). Only a minor dip occurs dur-ing the early stages of Lake Loitsana correspond-

ing to the morphometric eutrophication phase (Fig.8). Although diatoms have been shown to captureHolocene climate variability (e.g. Bigler 2001), theirsensitivity to other variables (e.g. pH, trophic state)complicates temperature reconstructions especiallyduring time periods when the other variables are notconstant (Battarbee 2000, Bigler and Hall 2003; Fig.1B: 7). Paper III shows that the diatom-based T jul

reconstruction is hampered throughout the Holoceneby the mass-occurrence of Fragilariaceae as well asby the design of the calibration set where the ma-jority of the calibration-sites are within the rangeof 11.1-13.0 ˝C. Diatom-inferred temperature recon-structions are therefore not discussed further. BelowI will evaluate how local-scale/site-specific processesmay have hampered the temperature signal. The re-construction outcomes are discussed against previousstudies from the region. Three focus time-windowsare used; the early Holocene; prior to 8200 cal. aBP, the mid-Holocene; 8200–4200 cal. a BP and thelate Holocene; 4200 cal. a BP–present (Walker et al.2012).

Early Holocene (prior to 8200 cal. a BP). – Thereconstructed early Holocene temperatures displaycontrasting patterns (Fig. 8). Pollen-based recon-structions displays low T jul values („2 ˝C be-low present-day) while chironomids reconstruct highT jul of at least 2 ˝C higher than present. The highearly Holocene T jul is further supported by indepen-dent data provided by plant macrofossils, which sug-gest minimum T jul of at least 2 ˝C above presentprior to 10 000 cal. a BP. The low pollen-basedT jul temperatures reconstructed by Seppa and Birks(2001; Fig. 1B, 9) for NW Finnish Lapland andSeppa et al. (2009), seem to be in accordance withthe Lake Loitsana pollen reconstructions. Both ofthese records are, however, shorter than the LakeLoitsana record and do not extend prior to 10 000cal. a BP. The pollen reconstruction from Lake Loit-sana, however, seems to be influenced by tapho-nomic differences between the fossil data in rela-tion to the calibration set (Salonen et al. 2013; pa-per III). As the c-set is designed to delimit theamount of local pollen in the modern samples, thehigh abundance of locally produced pollen (Cyper-aceae, Equisetum and Gramineae (syn. Poaceae))in the early Holocene fossil samples of Lake Loit-sana becomes a limiting factor for the pollen tem-perature reconstruction. These taxa in the c-set arecommonly found in tundra vegetation and thereforemay result in too low T jul. Warmer than presentchironomid-inferred T jul values are reconstructedduring the early Holocene. The taxa, which yieldthese high temperatures (i.e. Polypedilum nubeculo-

19

S. SHALA

sum-type, Chironomus anthracinus-type, Cricotopuscylindraceus-type and Procladius) have been associ-ated with both vegetation and trophic state in pre-vious studies (Raunio et al. 2010, Eggermont andHeiri 2012). As parts of this time period were charac-terised by littoral vegetation and morphometric eu-trophication, one might argue that the reconstructedtemperatures are driven mainly by these factors. Ifthese species were indeed driven by the local vege-tation and higher trophic state, their highest abun-dance would be expected to be restricted to the timeperiods when these factors prevailed. This, however,is clearly not the case in the Lake Loitsana sedi-ment record as higher abundances of these taxa oc-cur also outside these phases (paper II). It seems,thus, unlikely that these factors have had a major im-pact on the species distribution and thus the recon-structed high T jul values, which prevailed alreadyduring the deep glacial lake phase and start to de-cline halfway through the morphometric eutrophi-cation phase. Macrofossil inferred minimum T jul isbased on the local presence of wetland/aquatic planttaxa. As such, this reconstruction is mainly restrictedby taphonomic factors that affect the representa-tion of remains in the fossil sample. Temperatureshigher than present are inferred for the earliest partof the record. Macrofossil remains suggest presentday temperatures from c. 10 000 cal. a BP onwards.Moreover, pollen of aquatic/wetland plants is gen-erally said to have relatively local dispersal range(Krattinger 1975, Birks and Birks 2000, Ahee 2013).Accordingly the occurrence of Typha latifolia pollensupports the interpretation of warmer than presentearly Holocene July conditions. A similar patternwith high chironomid inferred T jul and low polleninferred T jul have also been recorded in tree-lineecotone lakes (Jones et al. 2011; Fig. 1A: 16, Pauset al. 2011; Fig. 1A: 17, Salonen et al. 2011; Fig. 1A:16, Luoto et al. in press, Paus 2013), while in north-ern boreal lakes on the Kola Peninsula high chirono-mid inferred T jul were reconstructed (Ilyashuk et al.2013; Fig 1B: 18-19). Minimum T jul warmer thanpresent have also been inferred from the boreal LakeKipojarvi, northern Finland (Valiranta et al. 2011).The suggested very early Holocene warming, recon-structed by macrofossils and chironomids, is consis-tent with the contemporary high summer insolation(Berger and Loutre 1991) and also increased sunspotactivity (Solanski et al. 2004) as well as high summersea-surface temperatures (SST) from the Barents Sea(Hald et al. 2007; Fig 1A: 20).

Mid-Holocene (8200–4200 cal. a BP). – Throughoutthis time period the pollen-based reconstruction dis-plays T jul values higher than present day (i.e. 13.4

˝C). The temperature gradually increases until c.7000 cal. a BP after which it remains around 14 °C(Fig. 8). This increase is in accordance with the plantmacrofossil reconstruction, which suggests at leastpresent day temperatures during mid-Holocene. Chi-ronomid inferred T jul, however, display values belowpresent day (13.4 ˝C) with the exception between6800 and 6000 cal. a BP when conditions slightlywarmer than today, are reconstructed. The increasein pollen-inferred temperatures coincides with themigration and establishment of pine forest in thearea, which is based on PAR, as well as occurrenceof Alnus, Filipendula and long-distance transportedCorylus. Warmer temperatures are generally drivenby the occurrence of tree pollen such as Pinus, Tilia,Ulmus, Corylus, Carpinus, Fraxinus). With the ex-ception of Pinus, these taxa are presently only foundin the southernmost part of the calibration set (Sa-lonen JS, personal communication). The low T jul

values (ą13 ˝C) inferred from the chironomid as-semblages between 8600 and 6800 cal. a BP as wellas 6000 and 5000 cal. a BP, appear to be driven byhigh occurrences of Corynocera ambigua and C. oliv-eri -type, whose abundance/presence in this recordappear to be driven by local-scale processes. Simi-larly, an influence of the river diversion on the chi-ronomid assemblage and thus reconstructed T jul

around 6500 cal. a BP cannot be ruled out as theobserved change in species distribution appears tobe related to locally driven changes in water bodyconditions (e.g. lack of fluvial input). To summarise,pollen and plant macrofossil-based T jul reconstruc-tion seems to be least influenced by local-scale fac-tors and thus they most probably reflect the regionalmid- Holocene climate most reliably. Temperatureshigher than present are in accordance with earlier re-gional temperature reconstructions (e.g. Renssen etal. 2009, Seppa et al. 2009).

Late Holocene (4200 cal. a BP to present). – Whileplant macrofossils suggest at least comparable T jul

throughout the late Holocene, pollen and chironomidreconstructions indicate higher than present T jul

temperatures (Fig. 8). Relatively low spruce pollenpercentages are encountered since c. 4000 cal. a BPin comparison to other sites in Finnish Lapland (Sa-lonen et al. 2013). The WA-optima of Picea (15.2˝C) is 4th highest among the dominating taxa in thec-set and this may to some extent have led to anoverestimation of the reconstructed T jul values dur-ing the late Holocene, despite its underrepresentationin the fossil samples. Increasing chironomid-inferredtemperatures from c. 4200 cal. a BP until presentare clearly related to declining abundances of cold-indicating Corynocera ambigua and Paratanytarsus.

20

PALAEOENVIRONMENTAL CHANGES IN THE NORTHERN BOREAL ZONE

The most recent rise in T jul values from c. 1800 cal.a BP onward seems to be driven by increased abun-dances of warm indicating Cladotanytarsus mancus-type and Tanytarsus mendax -type (Eggermont andHeiri 2012). These taxa, however, have also a prefer-ence for shallow waters and nutrient-rich conditions(Brooks et al. 2007, Luoto 2011a). The inferred tem-perature of 15.2 ˝C is therefore probably unrealisti-cally high and driven by the further shallowing of thelake.

Implications for interstadial environmentreconstructions

Palaeoenvironmental records extracted from olderdeposits at Sokli have provided evidence of a moredynamic Weichselian glaciation and warmer intersta-dial temperatures than previously thought (e.g. Hel-mens et al. 2007a, b, 2009, Valiranta et al. 2009).To date the interpretation of these older sedimentdeposits in the Sokli basin have lacked the refer-ence baseline environment development successiondescription that is provided by this Holocene study.The lacustrine sediment sequence of Lake Loitsanais particularly interesting since most of the older de-posits in Sokli also have a lacustrine facies. A par-ticular issue that should be brought up in the studyof Sokli deposits is the influence of the special typeof bedrock in the area (Vartiainen 1980). The car-bonatite deposit is richer in nutrients (e.g. phospho-rous) than the surrounding Precambrian bedrock.This could possibly result in a local ‘fertilisation’ ef-fect and anomalous ecosystem response. Accordingly,also the high temperatures reconstructed based onplant macrofossils and chironomids for the previousinterstadial(s) can be speculated to be a result of the“fertilisation” effect. In the Holocene record of LakeLoitsana the reduction in lake size and volume follow-ing glacial lake drainage and the subsequent morpho-metric eutrophication, could have led to such a ‘fer-tilisation effect’ and anomalous environmental condi-tions due to the relatively large input of local carbon-atite bedrock material. This appears, however, notto have been the case as the chironomid-based T jul

reconstruction in the early Holocene does not showany specific response to marked local changes in theglacial lake to Lake Loitsana transition. The same isvalid also for pollen and macrofossil reconstructions,which do not show anomalous values for this period.Lake Loitsana proxy evidence consistently shows nomajor deviations from the environmental and cli-mate reconstructions derived from other boreal oreven subarctic sites in Fennoscandia (Valiranta etal. 2005, Valiranta 2006a, b, Larsen et al. 2006; Fig.1B: 21, Jones et al. 2011, Valiranta et al. 2011, Birks

et al. 2012; Fig. 1B: 22). Furthermore, several of theplant taxa recorded in the early Holocene sedimentsas well as the MIS 5d-c are presently not found inthe surroundings of Lake Loitsana despite the localbedrock. It seems most plausible therefore, that thelong records derived from the Sokli area provide re-liable information about climate changes during thelast Interglacial-Glacial cycle in NE Finland.

Conclusions

ą Lake Loitsana and its surroundings weredeglaciated prior to 10 700 cal. a BP. This is inagreement with the wider deglaciation history ofNE Finland.

ą The glacial lake that formed following deglacia-tion was initially deep and characterised by highinput of minerogenic influx from a wider catch-ment area. The continuing ice-marginal retreatof the FIS, lowered the glacial lake level due tothe deglaciation of a major spillway, leaving thecoring site in a relatively sheltered embaymentdominated by shore erosion that created turbidwater conditions.

ą Four limnological/hydrological events were iden-tified based on the Lake Loitsana sedimentrecord; the presence of a proglacial lake, thedrainage of this lake and eventual formation ofLake Loitsana, adjacent wetland expansion andsubsequent river diversion, and the general in-filling of the lake.

ą In addition to regional climate and catchmentchanges, the Lake Loitsana development was af-fected by local-scale/site-specific processes in-cluding changes in fluvial input, influx rateof eroded material, turbidity, turbulence, wa-ter chemistry and water depth. These changesresulted in marked shifts in local aquatic andwetland assemblages. Aquatic macrophytes werealso identified as potential drivers to changes inchironomid species distribution.

ą The transfer-function based regional tempera-ture reconstructions derived from different bi-otic assemblages can be influenced by these lo-cally driven environmental changes. The differ-ent proxies, however, respond individualisticallyto various processes and inferred T jul appearsto be reliable in varying level during differenttime intervals.

ą Summer temperatures in the Lake Loitsana areaseem to have been warm already during the

21

S. SHALA

early Holocene but relatively warm conditionsprevailed until the mid-Holocene, i.e. until c.6000 cal. a BP. During the latter part of theHolocene, T jul were close to present day value.

ą Holocene environmental and climate reconstruc-tions at Lake Loitsana display no major devia-tions from other boreal, or even subarctic, sitesin Fennoscandia. The Lake Loitsana Holocenereconstructions lent credence to interstadials re-constructions derived from deposits in the adja-cent Sokli basin. The data suggest that temper-ature reconstructions are not driven by for in-stance the anomalous bedrock composition andrelated fertilising effect but that the proxy as-semblages largely reflect the changes climate.

Future prospects

The complex nature of biotic response to environ-mental change that is being increasingly emphasisedin palaeolimnological/ecological studies (Birks andBirks 2006, Smol 2010) is further verified by thisstudy. The Lake Loitsana record has shown thatthe early Holocene lake sediments are of impor-tance and should be further investigated if we areto reconstruct changes occurring in direct connec-tion to the deglaciation phase. In northern Finlandthese changes are often recorded in glacial lake sedi-ments and as demonstrated in this study, even shal-low glacial lakes have a potential for palaeoenviron-mental reconstructions. For regional climate recon-structions the use of lakes where gyttja accumula-tion started directly following deglaciation should bepreferred. This delimits the effects of marked localenvironmental changes. Possible future sites for earlyHolocene palaeoenvironmental reconstructions havebeen recognised in the nearby Varriotunturi area(Fig. 2B). Finely laminated lacustrine sediments de-rived from one of the lakes has been dated using ter-restrial macrofossils, and the initiation of depositiondates prior to 10 000 cal. a BP. As emphasised inthis study, macrofossil records from the boreal for-est zone of Fennoscandia dating back to the earliestpart of the Holocene are scarce. Aquatic macrophytesare predicted to react quickly to climate change (e.g.Alahuhta et al. 2011) and have been found to occur inboth northern boreal and tundra sites prior to 10 000cal. a BP (Valiranta 2005, 2006a, b). As the aquaticspecies disperse quickly (Barrat-Segretain 1996 andreferences therein, Sawada et al. 2003) and are not re-stricted by factors such as soil development, aquaticmacrophytes can be successfully used to reconstructpost-glacial environmental conditions (Valiranta et

al. in prep). In combination with pollen studies, thisproxy might also further improve our understandingof migration dynamics of terrestrial vegetation. Fur-thermore, as highlighted in this study, macrophyteabundance appears to be related to changes in chi-ronomid species distribution. Aquatic vegetation haslong been acknowledged to be an important compo-nent in the food-web chain as they e.g. provide refugefor zooplankton which graze upon phytoplankton, re-duce nutrient availability through uptake of nutrientsfrom the water and are themselves important as foodfor various zooplanktons (e.g. van Donk and van deBund 2002, Siitonen et al. 2011). The effects of foodweb interactions on the aquatic biota are being in-creasingly stressed (van Donk and van de Bund 2002,Anderson et al. 2008, Furey et al. 2012, Luoto et al.2012). Although not fully addressed in this study,future studies should thus consider also this as anessential tool in the assessment of changes in biolog-ical assemblages. The initiation and subsequent ex-pansion of wetlands surrounding Lake Loitsana hasclearly influenced the development of Lake Loitsana.Several cores were collected from these wetlands andthey bear evidence of changing environmental con-ditions throughout the Holocene. The Saunavuotsowetland situated north of Loitsana (X1 in Fig. 2C) isparticularly interesting since it was once connectedto Lake Loitsana. Peatland initiation occurred at c.8500 cal. a BP (Fig. 7) and was following only ashort period of gyttja accumulation after the glaciallake drainage. The peat record holds evidence ofa complex Holocene development with an atypicalsuccession from bog to fen (Valiranta M, personalcommunication). This is contrasting to the succes-sion of most wetlands, which commonly develop fromgroundwater fed fens into precipitation fed bogs (e.g.Gorham 1957, Hughes et al. 2000). Further studies onthe cores collected from Saunavuotso might thus pro-vide additional information on wetland developmentas well as how local hydrological changes occurringwithin the wetland might have affect Holocene lakedevelopment.

Financial support

This PhD-project was funded by the Swedish Nu-clear Fuel and Waste Management Company (SKB)and the Department of Physical Geography and Qua-ternary Geology, Stockholm University. Many thanksto the Bolin Centre for Climate Research (Stock-holm University), BioCold and NEPAL projects,Carl Mannerfelt Foundation, Gerard De Geer fund,INTIMATE COST Action, K&A Wallenberg foun-dation, Margit Althin’s foundation and Swedish So-

22

PALAEOENVIRONMENTAL CHANGES IN THE NORTHERN BOREAL ZONE

ciety for Anthropology and Geography for the finan-cial support I received throughout this project.

ReferencesAgassiz L. 1850. Lake Superior: its physical character, vegeta-

tion and animals compared to those of other and similarregions. Gould, Kendall, and Lincoln, Boston.

Ahee J. 2013. The spatial dynamics of wind pollination inbroadleaf cattail (Typha latifolia): A new method to in-fer spatial patterns of pollen dispersal. M.Sc. Thesis, En-vironmental and Life Sciences Graduate Program, TrentUniversity, Peterborough, ON, Canada

Aitchison J. 1982. The statistical analysis of compositionaldata (with discussion). Journal of the Royal StatisticalSociety, Series B (Statistical Methodology) 44, 139–177.

Alahuhta J, Heino J, Luoto M. 2011. Climate change and thefuture distributions of aquatic macrophytes across borealcatchments. Journal of Biogeography 38, 383–393.

Alexanderson H, Adrielsson L, Hjort C, Moller P, Antonov O,Eriksson S, Pavlov M. 2002. Depositional history of theNorth Taymyr ice-marginal zone, Siberia – a landsystemapproach. Journal of Quaternary Science 17, 361–382.

Alexanderson H, Eskola KO, Helmens KF. 2008. Optical dat-ing of Late Quaternary sediment sequence from northernFinland. Geochronometria 32, 51–59.

Allen JRM, Long AJ, Ottley CJ, Graham Pearson D, Hunt-ley B. 2007. Holocene climate variability in northernmostEurope. Quaternary Science Reviews 26, 1432–1453.

Ampel L, Wohlfarth B, Risberg J, Veres D, Leng MJ, KaislahtiTillman P. Diatom assemblage dynamics during abruptclimate change: the response of lacustrine diatoms toDansgaard–Oeschger cycles during the last glacial period.Journal of Paleolimnology 44, 397–404.

Andersen STH. 1994. History of the terrestrial environmentin the Quaternary of Denmark. Bulletin of the GeologicalSociety of Denmark 41, 219–228.

Anderson NJ. 2000. Diatoms, temperature and climaticchange. European Journal of Phycology 35, 307–314.

Anderson NJ, Brodersen KP, Ryves DB, McGowan S, Johans-son SL, Jeppesen E, Leng MJ. 2008. Climate versus in-lake processes as controls of community development inan oligotrophic lake in the low-Arctic of West Greenland.Ecosystems 11, 307–324.

Barnekow L. 1999. Holocene tree-line dynamics and inferredclimatic changes in the Abisko area, northern Sweden,based on macrofossil and pollen records. The Holocene9, 253–265.

Barnekow L, Sandgren P. 2001. Palaeoclimate and tree-linechanges during the Holocene based on pollen and plantmacrofossil records from six lakes at different altitudes innortheren Sweden. Review of Palaeobotany and Palynol-ogy 117, 109–118.

Barrat-Segretain M. H. 1996. Strategies of reproduction, dis-persion, and competition in river plants: A review. Vege-tation 123, 13–37.

Battarbee RW. 1986. Diatom analysis. In: Berglund BE. (ed)Handbook of Holocene Palaeoecology and Palaeohydrol-ogy. John Wiley & Sons, Chichester, pp 527–570.

Battarbee RW. 2000. Palaeolimnology approaches to climatechange, with special regard to the biological record. Qua-ternary Science Reviews 19, 107–124.

Bennett KD. 1996. Determination of the number of zonesin a biostratigraphical sequence. New Phytologist 132,155–170.

Berger A, Loutre MF. 1991. Insolation values for the climateof the last 10 million years. Quaternary Science Reviews10, 297–317.

Bergman J. Hammarlund D, Hannon G, Barnekow L,Wohlfarth B. 2005. Deglacial vegetation succession andHolocene tree-limit dynamics in the Scandes Moun-tains, westcentral Sweden: stratigraphic data comparedto megafossil evidence. Review of Palaeobotany and Pa-lynology 134, 129–151.

Bigler C. 2001. Diatoms as indicators of Holocene climate andenvironmental change in northern Sweden. PhD Thesis,Umea University, Umea, Sweden.

Bigler C, Grahn E, Laroque I, Jeziorski A, Hall RI. 2003.Holocene environmental change at Lake Njulla (999 ma.s.l.), northern Sweden: a comparison with four smallnearby lakes along an altitudinal gradient. Journal of Pa-leolimnology 29, 13–29.

Bigler C, Hall RI. 2003. Diatoms as quantitative indica-tors of July temperature: a validation attempt at cen-tury–scale with meteorological data from northern Swe-den. Palaeogeography, Palaeoclimatology, Palaeoecology189, 147–160.

Bigler C, Larocque I, Peglar SM, Birks, HJB, Hall RI. 2002.Quantitative multiproxy assessment of long–term patternsof Holocene environmental change from a small lake nearAbisko, northern Sweden. The Holocene 12, 481–496.

Birks HH. 1973. Modern macrofossil assemblages in lake sed-iments in Minnesota. In: Birks, HJB and West RG (ed)Quaternary Plant Ecology, Blackwells Scientific Publica-tions, Oxford, pp 173–189.

Birks HH. 2007. Plant macrofossil introduction. In: Elias SA.(ed) Encyclopedia of Quaternary Science. Elsevier press,Oxford, pp 2266–2288.

Birks HH, Birks HJB. 2000. Future uses of pollen analysismust include plant macrofossils. Journal of Biogeography27, 31–35.

Birks HH, Birks HJB. 2006. Multi-proxy studies in palae-olimnology. Vegetation History and Archaeobotany 15,235–251.

Birks HH, Battarbee RW, Birks HJB, Bradshaw EG, BrooksSJ, Duigan CA, Jones VJ, Lemdahl G, Peglar SM, SolemJO, Solhøy IW, Solhøy T, Stalsberg MK. 2000. The devel-opment of the aquatic ecosystem at Krakenes Lake, west-ern Norway, during the late-glacial and early-Holocene –a synthesis. Journal of Palaeolimnology 23, 91–114.

Birks HH, Jones VJ, Brooks SJ, Birks HJB, Telford RJ, Jug-gins S, Peglar SM. 2012. From cold to cool in northernmostNorway: Lateglacial and early Holocene multi-proxy envi-ronmental and climate reconstructions from Jansvatnet,Hammerfest. Quaternary Science Reviews 33, 100–120.

Birks HJB. 2012. Analysis of Stratigraphical Data. In: BirksHJB, Lotter AF, Juggins S, Smol JP (eds) Tracking En-vironmental Change Using Lake Sediments: Data Han-dling and Numerical Techniques. Springer, Dordrecht, pp355–378.

Birks HJB, Gordon AD. 1985. Numerical methods in Quater-nary pollen analysis. 317 pp. Academic Press, London.

Birks HJB, Heiri O, Seppa H, Bjune AE. 2010. Strengths andweaknesses of quantitative climate reconstructions basedon late-Quaternary biological proxies. Open Ecology Jour-nal 3, 68–110.

Birks HJB, Line JM, Juggins S, Stevenson AC, ter BraakCJF. 1990. Diatoms and pH reconstruction. Philosophi-cal Transactions of the Royal Society B 327, 263–278.

Bjune AE. 2005. Holocene vegetation history and tree-linechanges on a north-south transect crossing major climategradients in southern Norway – evidence from pollen andplant macrofossils in lake sediments. Review of Palaeob-otany and Palynology 133, 249–275.

Bjune AE, Birks HJB, Seppa H. 2004. Holocene vegetationand climate history on a continental–oceanic transect innorthern Fennoscandia based on pollen and plant macro-fossils. Boreas 33, 211–223.

23

S. SHALA

Blaauw M. 2010. Methods and code for ’classical’ age-modelling of radiocarbon sequences. QuaternaryGeochronology 5, 512–518.

Blomdin R, Murray A, Thomsen KJ, Buylaert J-P, SohbatiR, Jansson KN, Alexanderson H. 2012. Timing of thedeglaciation in southern Patagonia: Testing the applica-bility of K-Feldspar IRSL. Quaternary Geochronology 10,264–272.

Bos JAA, Helmens KF, Bohncke SJP, Seppa H, Birks HJB.2009. Flora, vegetation and climate at Sokli, northeasternFennoscandia, during the Weichselian Middle Pleniglacial.Boreas 38, 335-348.

Bradley RS. 1999. Paleoclimatology: Reconstructing Climatesof the Quaternary. 613 pp. Academic Press, San Diego.

Brodersen KP, Lindegaard C. 1999. Classification, assessmentand trophic reconstruction of Danish lakes using chirono-mids. Freshwater Biology 42, 143–157.

Bronk Ramsey C. 2009. Bayesian analysis of radiocarbondates. Radiocarbon 51, 337–360.

Brooks SJ, Langdon PG, Heiri O. 2007. The Identificationand use of Palaeoarctic Chironomidae Larvae in Palaeoe-cology. QRA Technical Guide No. 10, 276 pp. QuaternaryResearch Association, London.

Brooks SJ, Axford Y, Heiri O, Langdon PG, Larocque-ToblerI. 2012. Chironomids can be reliable proxies for Holocenetemperatures. A comment on Velle et al. (2010). TheHolocene 22, 1495–1500.

Calvert SE. 1983. Geochemistry of Pleistocene sapropels andassociated sediments from the eastern Mediterranean.Oceanologica Acta 6, 255–267.

Cleve-Euler A. 1922. Om diatomacevegetationen och dessforandringar i Sabysjon, Uppland, samt nagra damda sjoari Salatrakten. Sveriges Geologiska Undersokning C309,49–77.

Croudace IW, Rindby A, Rothwell RG. 2006. ITRAX. De-scription and evaluation of a new multi-function X-raycore scanner. In: Rothwell RG (ed) New techniques in sedi-ment core analysis. Geological Society of London, London,pp 51–63.

Croudace I, Rothwell G. 2010. Micro-XRF sediment core scan-ners: important new tools for the environmental and earthsciences. Spectroscopy Europe 22, 6–13.

Donner J. 1995. Quaternary History of Scandinavia. 200 pp.Cambridge University Press, Cambridge.

Duller GAT. 2006. Single grain optical dating of glacigenicdeposits. Quaternary Geochronology 1, 296–304.

Eggermont H, Heiri O. 2012. The chironomid-temperature re-lationship: expression in nature and palaeoenvironmentalimplications. Biological Reviews 87, 430–456.

Engels S, Helmens K F, Valiranta M, Brooks SJ, BirksHJB. 2010. Early Weichselian (MIS 5d and 5c) temper-atures and environmental changes in northern Fennoscan-dia as recorded by chironomids and macroremains atSokli, northeast Finland. Boreas 39, 689–704.

Eronen M, Hyvarinen H, Zetterberg P. 1999. Holocene humid-ity changes in northern Finnish Lapland inferred from lakesediments and submerged Scots pines dated by tree-rings.The Holocene 9, 569–580.

Eronen M, Lindholm M, Saastamoinen S, Zetterberg P. 1999.Variable Holocene climate, treeline dynamics and changesin natural environments in northern Finnish Lapland.Chemosphere: Global Change Science 1, 377–387.

Esseen P-A, Ehnstrom B, Ericson L, Sjoberg K. 1997. Borealforests. Ecological Bulletins 46, 16–47.

Fægri K, Iversen J. 1989. Textbook of pollen analysis. 4thedn. by Fægri K, Kaland PE, Krzywinski K. John Wiley& Sons, Chichester.

Finnish Environmental Institute 2010. Environmental dataobtained from the Environmental Information Sys-tem (HERTTA) data base: Vattenkvalite, resultatblad8.2.2010.

Francus P, Lamb H, Nakasawa T, Marshall M, Brown E,Suigetsu 2006 Project Members 2009. The potential ofhigh-resolution X-ray fluorescence core scanning: Appli-cations in paleolimnology. PAGES News 17, 93–95.

Fritz SC, Juggins S, Battarbe RW, Engstrom DR. 1991. Re-construction of past changes in salinity and climate usinga diatom-based transfer function. Nature 352, 706–708.

Furey PC, Lowe RL, Power ME, Campbell–Craven AM. 2012.Midges, Cladophora, and epiphytes: shifting interactionsthrough succession. Freshwater Science 31, 93–107.

Gajewski K, Bouchard G, Wilson SE, Kurek J, Cwynar LC.2005. Distribution of Chironomidae (Insecta: Diptera)head capsules in recent sediments of Canadian Arcticlakes. Hydrobiologia 549, 131–143.

Garcıa Ambrosiani K.1991. Interstadial minerogenic sedi-ments at the Levenaiemi mine, Svappavaara, Swedish La-pland. GFF 113, 273–287.

Gibson CE, Andersen NJ, Haworth EY. 2003. Aulacoseira sub-arctica: taxonomy, physiology, ecology and palaeoecology.European Journal of Phycology 38, 83–101.

Gorham E. 1957. The Development of Peat Lands. The Quar-terly Review of Biology 32, 145– 166.

Grimm EC. 2011. Tilia software v. 1.7.16–Illinois State Museum, Springfield, IL.(http://museum.state.il.us/pub/grimm/tilia/)

Grudd H, Briffa KR, Karlen W, Bartholin TS, Jones PD,Kromer B. 2002. A 7400-year treering chronology in north-ern Swedish Lapland: natural climate variability expressedon annual to millennial timescales. The Holocene 12,657–665.

Hajdas I. 2006. Radiocarbon dating for Quaternary studies –New developments and perspectives. Quaternary Interna-tional 154, 160–163.

Hald M, Andersson C, Ebbesen H, Jansen E, Klitgaard-Kristensen D, Risebrobakken B, Salomonsen GR, Sarn-thein M, Sejrup HP, Telford RJ. 2007. Variations in tem-perature and extent of Atlantic Water in the northernNorth Atlantic during the Holocene. Quaternary ScienceReviews 26, 3423–3440.

Hamet-Ahti L.1981. The boreal zone and its biotic subdivi-sion. Fennia 159, 69–75.

Hattestrand M. 2008. Vegetation and climate during Weich-selian ice fee intervals in northern Sweden – interpreta-tions from fossil and modern pollen records. PhD thesis,Stockholm University, Stockholm, Sweden.

Hattestrand M, Robertson A-M. 2010. Weichselian interstadi-als at Riipiharju, northern Sweden – interpretation of veg-etation and climate from fossil and modern pollen records.Boreas 39, 296–311.

Hannon GE, Gaillard M-J. 1997. The plant macrofossil recordof past lake-level changes. Journal of Paleolimnology 18,15–28.

Heiri O, Lotter AF, Lemcke G. 2001. Loss on ignition as amethod for estimating organic and carbonate content insediments: reproducibility and comparability of results.Journal of Paleolimnology 25, 101–110.

Helmens KF, Bos JAA, Engels S, Bohncke SJP, Renssen H,Heiri O, Brooks SJ, Seppa H, Birks HJB, Wohlfarth B.2007a. Present-day temperatures in northern Scandinaviaduring the last glaciation. Geology 35, 987–990.

Helmens KF, Johansson PW, Rasanen ME, Alexanderson H,Eskola KO. 2007b. Ice-free intervals continuing into Ma-rine Isotope Stage 3 at Sokli in the central area of theFennoscandian glaciations. Bulletin of the Geological So-ciety of Finland 79, 17–39.

24

PALAEOENVIRONMENTAL CHANGES IN THE NORTHERN BOREAL ZONE

Helmens KF, Rasanen ME, Johansson PW, Jungner H,Korjonen K. 2000. The Last Interglacial-Glacial cyclein NE Fennoscandia: a nearly continuous record fromSokli (Finnish Lapland). Quaternary Science Reviews 19,1605–1623.

Helmens KF, Risberg J, Jansson KN, Weckstrom J, Bernts-son A, Kaislahti Tillman P, Johansson PW, WastegardS. 2009. Early MIS 3 glacial lake evolution, ice-marginal retreat pattern and climate at Sokli (north-eastern Fennoscandia). Quaternary Science Reviews 28,1880–1894.

Hofmann W. 1998. Cladocerans and chironomids as indicatorsof lake level changes in north temperate lakes. Journal ofPaleolimnology 19, 55–62.

Hodgson DA, Wright SW, Davies N. 1997. Mass spectrometryand reverse phase HPLC techniques for identification ofdegraded fossil pigments in lake sediments and their appli-cation in palaeolimnology. Journal of Paleolimnology 18,335–350.

Holst NO. 1899. Bidrag till kannedomen om Ostersjons ochBottniska vikens postglaciala geologi. Sverige GeologiskaUndersokning C180, pp 128.

Hughes PDM, Mauquoy D, Barber KE, Langdon PG. 2000.Mire-development pathways and palaeoclimatic recordsfrom a full Holocene peat archive at Walton Moss, Cum-bria, England. The Holocene 10, 465–479.

Hyvarinen H, Alhonen P. 1994. Holocene lake-level changes inthe Fennoscandian tree-line region, western Finnish Lap-land: diatom and cladoceran evidence. The Holocene 4,251–258.

Ilyashuk EA, Ilyashuk BP. Kolka VV, Hammarlund D. 2013.Holocene climate variability on the Kola Peninsula, Rus-sian Subarctic, based on aquatic invertebrate records fromlake sediments. Quaternary Research 79, 350–361.

Imbrie J, Hays JD, Martinson DG, McIntyre A, Mix AC, Mor-ley JJ, Pisias NG, Prell WL, Shackleton NJ. 1984. The or-bital theory of Pleistocene climate: support from a revisedchronology of the marine δ 18O record. In: Berger AL, Im-brie J, Hays J, Kukla G, Saltzman B (eds) Milankovitchand Climate, Part 1. D. Reidel Publishing Co., Dordrecht,pp 269–305.

Imbrie J, Kipp NG. 1971. A new micropaleontological methodfor quantitative paleoclimatology: application to a LatePleistocene Caribbean core. In: Turekian KK (ed) TheLate Cenozoic glacial ages. Yale University Press, NewHaven, pp 77–181.

IPCC (Meehl, GA, Stocker TF, Collins WD, Friedlingstein P,Gaye AT, Gregory JM, Kitoh A, Knutti R, Murphy JM,Noda A, Raper SCB, Watterson IG, Weaver AJ, Zhao Z-C) 2007. Global climate projections. In: Solomon S, Qin D,Manning M, Marquis M, Averyt K, Tignor MMB, MillerHL, Chen Z (eds) Climate Change 2007: The Physical Sci-ence Basis. Contribution of Working Group I to the FourthAssessment Report of the Intergovernmental Panel on Cli-mate Change. Cambridge University Press, Cambridge, pp747–845.

Iversen J. 1954. The late–glacial flora of Denmark and itsrelation to climate and soil. Danmarks Geologiske Un-dersøgelse II 80, 87–119.

Iversen J. 1957. The bearing of glacial and interglacial epochson the formation and extinction of plant taxa. UppsalaUniversitets Arsskrift 1958, 6, 210–215.

Jansson M, Broberg A. 1994. Abiotiska faktorerskaraktaristiska, omsattning och funktion i sotvatten.Limnologiska institutionen. Uppsala universitet.

Johansson P. 1995. The deglaciation in the eastern part ofthe Weichselian ice divide in Finnish Lapland. GeologicalSurvey of Finland, Espoo.

Johansson P, Lunkka PJ, Sarala P. 2011. The glaciation of Fin-land. In: Ehlers J, Gibbard PL, Hughes PD (eds) Quater-nary Glaciations – Extent and Chronology: a closer look.Developments in Quaternary Science 15. Elsevier, Oxford,pp 105–116.

Jones VJ, Solovieva N, Self AE, McGowan S, Rosen P, SalonenJS, Seppa H, Valiranta M,

Parrott E, Brooks SJ. 2011. The influence of Holocenetree–line advance and retreat on an arctic lake ecosystem;a multi-proxy study from Kharinei Lake, North EasternEuropean Russia. Journal of Paleolimnology 46, 123–137.

Jonsson CE, Leng MJ, Rosqvist GC, Seibert J, Arrowsmith A.2009. Stable oxygen and hydrogen isotopes in sub-Arcticlake waters from northern Sweden. Journal of Hydrology376, 143–151.

Jowsey PC. 1966. An improved peat sampler. New Phytologist65, 245–248.

Juggins S. 2009. Rioja: Analysis of Quaternary Sci-ence Data, R package version 0.5–6. (http://cran.r-project.org/packagerioja)

Juggins S. 2013. Quantitative reconstructions in palaeolimnol-ogy: new paradigm or sick science? Quaternary ScienceReviews 64, 20–32.

Juggins S, Birks HJB. 2012. Quantitative environmental re-constructions from biological data. In: Birks HJB, LotterAF, Juggins S, Smol JP (eds) Tracking EnvironmentalChange Using Lake Sediments: Data Handling and Nu-merical Techniques. Springer, Dordrecht, pp 431–494.

Kabailiene M. 1995. The Baltic Ice Lake and Yoldia sea stages,based on data from diatom analysis in the central, south-eastern and eastern Baltic. Quaternary International 27,69– 72.

Karlen W, Bodin A, Kuylenstierna J, Naslund J-O. 1995. Cli-mate of northern Sweden during the Holocene. Journal ofCoastal Research, Special Issue 17, 49–54.

Karlen W, Kuylenstierna J. 1996. On solar forcing of Holoceneclimate: evidence from Scandinavia. The Holocene 6,359–365.

Kleman J, Hattestrand C, Clarhall A. 1999. Zoomingin on frozen-bed patches: scaledependent controls onFennoscandian ice sheet basal thermal zonation. Annalsof Glaciology 28, 189–194.

Kolstrup E. 1979. Herbs as July temperature indicators forparts of the pleniglacial and lateglacial in the Netherlands.Geologie en Mijnbouw 58, 377–380.

Korhola A, Vasko K, Toivonen HTT, Olander H. 2002.Holocene temperature changes in northern Fennoscandiareconstructed from chironomids using Bayesian modelling.Quaternary Science Review 21, 1841–1860.

Korhola A, Weckstrom J. 2004. Paleolimnological studies inarctic Fennoscandia and the Kola Peninsula (Russia). In:Pienitz R, Douglas MSV, Smol JP (eds) Long-term envi-ronmental change in Arctic and Antarctic lakes, volume8. Kluwer, Dordrecht, pp 381–418.

Korhola A, Tikkanen M, Weckstrom J. 2005. Quantificationof Holocene lake-level changes in Finnish Lapland usinga cladocera – lake depth transfer model. Journal of Pale-olimnology 34, 175–190.

Korhola A, Weckstrom J, Holmstrom L, Erasto P. 2000. Quan-titative Holocene Climatic Record from Diatoms in North-ern Fennoscandia. Quaternary Research 54, 284–294.

Krattinger K. 1975. Genetic mobility of Typha. AquaticBotany 1, 57–70.

Kujansuu R, Eriksson B, Gronlund T. 1998. Lake Inarijarvi,northern Finland: Sedimentation and late Quaternaryevolution. Geological Survey of Finland, Espoo.

Kullman L. 1995. Holocene Tree-Limit and Climate His-tory from the Scandes Mountains, Sweden. Ecology 76,2490–2502.

25

S. SHALA

Kultti S, Valiranta M, Sarmaja-Korjonen K, Solovieva N, Vir-tanen T, Kauppila T, Eronen M. 2003. Palaeoecologicalevidence of changes in vegetation and climate during theHolocene in the pre–Polar Urals, northeast European Rus-sia. Journal of Quaternary Science 18, 503–520.

Kylander ME, Ampel L, Wohlfarth B, Veres D. 2011. High-resolution X-ray fluorescence core scanning analysis ofLes Echets (France) sedimentary sequence: new insightsfrom chemical proxies. Journal of Quaternary Science 26,109–117.

Kylander ME, Klaminder J, Wohlfarth B, Lowemark L. 2013.Geochemical responses to paleoclimatic changes in south-ern Sweden since the late glacial: the Hasseldala Port lakesediment record. Journal of Paleolimnology 50, 57–70.

Kylander ME, Lind EM, Wastegard S, Lowemark L. 2012.Recommendations for Using XRF Core Scanning as a Toolin Tephrochronology. The Holocene 22, 371–375.

Lampinen R, Lahti T. 2013. Kasviatlas 2012. Botanical Mu-seum, Finnish Museum of Natural History, Helsinki.

Larocque I, Bigler C. 2004. Similarities and discrepancies be-tween chironomid- and diatominferred temperature recon-structions through the Holocene at Lake 850, northernSweden. Quaternary International 122, 109–12.

Larocque I, Hall RI. 2004. Holocene temperature estimatesand chironomid community composition in the AbiskoValley, northern Sweden. Quaternary Science Reviews 23,2453–2465.

Larsen J, Bjune AE, de la Riva Caballero A. 2006. HoloceneEnvironmental and Climate History of Trettetjorn, a Low-alpine Lake in Western Norway, Based on Subfossil Pollen,Diatoms, Oribatid Mites, and Plant Macrofossils. Arctic,Antarctic, and Alpine Research 38, 571–583.

Libby W F, Anderson EC, Arnold JR. 1949. Age Determina-tion by Radiocarbon Content: World–Wide Assay of Nat-ural Radiocarbon. Science 109, 227–228.

Lund JWG. 1954. The Seasonal Cycle of the Plankton Diatom,Melosira Italica (EHR.) Kutz. Subsp. Subarctica O. Mull.Journal of Ecology 42, 151–179.

Lundqvist J. 1972. Ice-lake types and deglaciation patternalong the Scandinavian mountain range. Boreas 1, 27–54.

Lundqvist J. 1992. Glacial stratigraphy in Sweden. In: Kau-ranne K (ed) Glacial Stratigraphy, Engineering Geologyand Earth Construction. Geological Survey of Finland,Special Paper 15, pp 43–59.

Lunkka JP, Murray A, Korpela K. 2008. Weichselian sedi-ment succession at Ruunaa, Finland, indicating a Mid-Weichselian ice-free interval in eastern Fennoscandia.Boreas 37, 234– 244.

Luoto TP. 2011a. The relationship between water quality andchironomid distribution in Finland – A new assemblage-based tool for assessments of long–term nutrient dynam-ics. Ecological Indicators 11, 255–262.

Luoto TP. 2011b. Indicator value of midge larvae (Diptera: Ne-matocera) in shallow boreal lakes with a focus on habitat,water quality, and climate. Aquatic Insects 33, 351–370.

Luoto TP, Kaukolehto M, Weckstrom J, Korhola A,Valiranta M. in press. New evidence of warm early-Holocene summers in subarctic Finland based onan enhanced regional chironomid-based tempera-ture calibration model. Quaternary Research (2013)http://dx.doi.org/10.1016/j.yqres.2013.09.010.

Luoto T, Kultti S, Nevalainen L, Sarmaja-Korjonen K. 2010.Temperature and effective moisture variability in south-ern Finland during the Holocene quantified with midge-based calibration models. Journal of Quaternary Science25, 1317–1326.

Luoto T, Nevalainen L, Salonen V-P. 2012. Midge distribu-tion patterns at multiple spatial scales in shallow Finnishlakes revealed through sedimentary records. EntomologicaFennica 23, 83–94.

Lowemark L, Chen H-F, Yang T-N, Kylander M, Yu E-F,Hsu Y-W, Lee T-Q, Song S-R, Jarvis S. 2011. Normaliz-ing XRF-scanner data: A cautionary note on the interpre-tation of high-resolution records from organic-rich lakes.Journal of Asian Earth Sciences 40, 1250–1256.

MacDonald GM, Beukens RP, Kieser WE. 1991. Radiocarbondating of limnic sediments: A comparative analysis anddiscussion. Ecology 72, 1150–1155.

Mangerud J. 1991. The Scandinavian Ice Sheet through thelast interglacial/glacial cycle. In: Frenzel B (ed) Klim-ageschichtliche Probleme der Letzten 130 000 Jahre. G.Fischer, Stuttgart, pp. 307–330.

Martinson DG, Pisias NG, Hays JD, Imbrie J, Moore TC,Shackleton NJ. 1987. Age Dating and the Orbital The-ory of the Ice Ages: Development of a High–Resolution 0to 300,000 year Chronostratigraphy. Quaternary Research27, 1–29.

Medeiros A, Quinlan R. 2011. The distribution of the Chi-ronomidae (Insecta: Diptera) along multiple environmen-tal gradients in lakes and ponds of the eastern CanadianArctic. Canadian Journal of Fisheries and Aquatic Sci-ences 68, 1511–1527.

Moen A. 1998. Nasjonalatlas for Norge; vegetasjon. StatensKartverk, Hønefoss.

Moreno A, Giralt S, Valero–Garces B, Saez A, Bao R, PregoR, Pueyo JJ, Gonzales–Samperiz P, Taberner C. 2007. A14 kyr record of the tropical Andes: the Lago Chungarasequence (18˝S, northern Chilean Altiplano). QuaternaryInternational 161, 4–21.

Murray AS, Wintle AG. 2000. Luminescence dating of quartzusing an improved singlealiquot regenerative–dose proto-col. Radiation Measurements 32, 57–73.

Nyman M, Korhola A, Brooks SJ. 2005. The distribution anddiversity of Chironomidae (Insecta: Diptera) in westernFinnish Lapland, with special emphasis on shallow lakes.Global Ecology and Biogeography 14, 137–153.

Nyman M, Weckstrom J, Korhola A. 2008. Chironomid re-sponse to environmental drivers during the Holocene in ashallow treeline lake in northwestern Fennoscandia. TheHolocene 18, 215–227.

Oksanen J, Blanchet FG, Kindt R, Legendre P,Minchin PR, O’Hara RB, Simpson GL, SolymosP, Stevens MHH, Wagner H. 2011. Vegan: Com-munity Ecology Package. R package version 2.0–1.(http://CRAN.R–project.org/package=vegan)

Olsen L. 1988. Stadials and interstadials during the Weichselglaciation on Finnmarksvidda, northern Norway. Boreas17, 517–539.

Olsson IU. 1991. Accuracy and precision in sediment chronol-ogy. Hydrobiologia 214, 25–34. Paarma H. 1970. A newfind of carbonatite in North Finland, the Sokli plug inSavukoski. Lithos 3, 129–133.

Palmer S, Walker IR, Heinrichs M, Hebda R, Scudder G.2002. Postglacial midge community change and Holocenepalaeotemperature reconstructions near treeline, southernBritish Columbia (Canada). Journal of Paleolimnology 28,469–490.

Paus A, Velle G, Berge J. 2011. The Lateglacial andearly Holocene vegetation and environment in theDovre mountains, central Norway, as signalled in twoLateglacial nunatak lakes. Quaternary Science Reviews30, 1780–1796.

Paus A. 2013. Human impact, soil erosion, and vegeta-tion response lags to climate change: challenges for themid–Scandinavian pollen–based transfer–function tem-perature reconstructions. Vegetation History and Ar-chaeobotany 22, 269–284.

Paus A. 2010. Vegetation and environment of the Rødalenalpine area, Central Norway, with emphasis on the earlyHolocene. Vegetation History and Archaeobotany 19,29–51.

26

PALAEOENVIRONMENTAL CHANGES IN THE NORTHERN BOREAL ZONE

Perttunen M, Vartiainen H. 1992. Glaciofluvial transport ofclasts and heavy minerals from the Sokli Carbonatite com-plex, Finnish Lapland. Geological Survey of Finland, Es-poo.

Pitkaranta R. 2009. Pre-late Weichselian podzol soil, per-mafrost features and lithostratigraphy at Penttilankangas,western Finland. Bulletin of the Geological Society of Fin-land 81, 53– 74.

Preusser F, Chithambo ML, Gotte T, Martini M, RamseyerK, Sendezera,EJ, Susino GJ, Wintle AG. 2009. Quartz asa natural luminescence dosimeter. Earth-Science Reviews97, 184–214.

Preusser F, Degering D, Fuchs M, Hilgers A, Kadereit A,Klasen N, Krbetschek M, Richter D, Spencer JQG. 2008.Luminescent dating: basics, methods and applications.Eiszeitalter und Gegenwart 57, 95–149.

R Development Core Team 2011. R: a language and environ-ment for statistical computing. R Foudation for StatisticalComputing, Vienna.

Raunio J, Paasivirta L, Hamalainen H. 2010. Assessing laketrophic status using springemerging chironomid pupal ex-uviae. Fundamental and Applied Limnology 176, 61–73.

Reimer PJ, Baillie MGL, Bard E, Bayliss A, Beck JW, Black-well PG, Bronk-Ramsey C, Buck CE, Burr GS, EdwardsRL, Friedrich M, Grootes PM, Guilderson TP, Hajdas I,Heaton TJ, Hogg AG, Hughen KA, Kaiser KF, KromerB, McCormac FG, Manning SW, Reimer RW, RichardsDA, Southon JR, Talamo S, Turney CSM, van der PlichtJ, Weyhenmeyer CE. 2009. IntCal09 and Marine09 ra-diocarbon age calibration curves, 0–50,000 years cal. BP.Radiocarbon 51, 1111–1150.

Rekacewicz P. 2005. Definitions of the Arctic.UNEP/GRID-Arendal (http://www.grida.no/graphicslib /de-tail/definitions–of–the–arctic 12ba).

Renberg I, Hellerberg T. 1982. The pH history of lakes insouthwestern Sweden, as calculated from the subfossil di-atom flora of the sediments. Ambio 11, 30–33.

Renssen H, Seppa H, Crosta X, Goosse H, Roche DM. 2012.Global characterization of the Holocene Thermal Maxi-mum. Quaternary Science Reviews 48, 7–19.

Reuss NS, Hammarlund D, Rundgren M, Segerstrom U, Eriks-son L, Rosen P. 2010. Lake Ecosystem Responses toHolocene Climate Change at the Subarctic Tree–Line inNorthern Sweden. Ecosystems 13, 393–409.

Risberg J, Sandgren P, Teller JT, Last WM. 1999. Siliceousmicrofossils and mineral magnetic characteristics in a sed-iment core from Lake Manitoba, Canada: a remnant ofglacial Lake Agassiz. Canadian Journal of Earth Sciences36, 1299–1314.

Robertsson A-M, Garcıa Ambrosiani K. 1988. Late Pleis-tocene stratigraphy at Boliden, northern Sweden. Boreas17, 1–14.

Rollinson HR. 1993. Using Geochemical Data: Evaluation,Presentation, Interpretation. Pearson Education, UpperSaddle River, New Jersey.

Root TL, Price JT, Hall KR, Schneider SH, Rosenzweig C,Pounds JA. 2003. Fingerprints of global warming on wildanimals and plants. Nature 421, 57–60.

Rosen P, Segerstrom U, Eriksson L, Renberg I, BirksHJB. 2001. Holocene climatic change reconstructed fromdiatoms, chironomids, pollen and near–infrared spec-troscopy at an alpine lake (Sjuodjijaure) in northern Swe-den. The Holocene 11, 551–562.

Rosen P,Segerstrom U, Eriksson L, Renberg I. 2003. Do Di-atom, Chironomid, and Pollen Records Consistently In-fer Holocene July Air Temperature? A Comparison UsingSediment Cores from Four Alpine Lakes in Northern Swe-den. Arctic, Antarctic, and Alpine Research 35, 279–290.

Rosqvist G, Jonsson C, Yam R, Karlen W, Shemesh A. 2004.Diatom oxygen isotopes in proglacial lake sediments fromnorthern Sweden: a 5000 year record of atmospheric cir-culation. Quaternary Science Reviews 23, 851–859.

Rosqvist GC, Leng MJ, Jonsson C. 2007. North Atlantic re-gion atmospheric circulation dynamics inferred from a lateHolocene lacustrine carbonate isotope record, northernSwedish Lapland. The Holocene 17, 867–873.

Salonen JS, Seppa H, Valiranta M, Jones VJ, Self A, HeikkilaM, Kultti S, Yang H. 2011. The Holocene thermal maxi-mum and late-Holocene cooling in the tundra of NE Eu-ropean Russia. Quaternary Research 75, 501–511.

Salonen JS, Helmens KF, Seppa H, Birks HJB. 2013.Pollen-based palaeoclimate reconstructions over longglacial–interglacial timescales: methodological tests basedon the Holocene and MIS 5d–c deposits at Sokli, northernFinland. Journal of Quaternary Science 28, 271–282.

Salonen V-P, Kaakinen A, Kultti S, Miettinen A, Eskola KO,Lunkka JP. 2008. Middle Weichselian glacial event in thecentral part of the Scandinavian Ice Sheet recorded in theHitura pit, Ostrobothnia, Finland. Boreas 37, 38–54.

Sarmaja-Korjonen K, Nyman M, Kultti S, Valiranta M.2006. Palaeolimnological development of Lake Njarga-javri, northern Finnish Lapland, in a changing Holoceneclimate and environment. Journal of Paleolimnology 35,65–81.

Sarmaja-Korjonen K, Hyvarinen H. 1999. Cladoceran and di-atom stratigraphy of calcareous lake sediments from Ku-usamo, NE Finland. Indications of Holocene lake-levelchanges. Fennia 177, 55–70.

Sawada M, Viau AE, Gajewski K. 2003. The biogeographyof aquatic macrophytes in North America since the LastGlacial Maximum. Journal of Biogeography 30, 999–1017.

Self AE, Brooks SJ, Birks HJB, Nazarova L, Porinchu D, Od-land A, Yang H, Jones VJ. 2011. The distribution andabundance of chironomids in high–latitude Eurasian lakeswith respect to temperature and continentality: develop-ment and application of new chironomid-based climate-inference models in northern Russia. Quaternary ScienceReviews 30, 1122–1141.

Seppa H and Birks HJB. 2001. July mean temperature andannual precipitation trends during the Holocene in theFennoscandian tree-line area: pollen-based climate recon-structions. The Holocene 11, 527–539.

Seppa H, Birks HJB. 2002. Holocene Climate Reconstructionsfrom the Fennoscandian Tree- Line Area Based on PollenData from Toskaljavri. Quaternary Research 57, 191–199.

Seppa H, Hammarlund D. 2000. Pollen–stratigraphical ev-idence of Holocene hydrological change in northernFennoscandia supported by independent isotopic data.Journal of Paleolimnology 24, 69–79.

Seppa H, Bjune AE, Telford RJ, Birks HJB, Veski S. 2009.Last nine thousand years of temperature variability inNorthern Europe. Climate of the Past 5, 523–535.

Seppa H, Nyman M, Korhola A, Weckstrom J. 2002.Changes of treelines and alpine vegetation in relation topost–glacial climate dynamics in northern Fennoscandiabased on pollen and chironomid records. Journal of Qua-ternary Science 17, 287–301.

Siitonen S, Valiranta M, Weckstrom J, Juutinen S, KorholaA. 2011. Comparison of Cladocera-based water-depth re-construction against other types of proxy data in FinnishLapland. Hydrobiologia 676, 155–172.

SKB 2006. Geosphere process report for the safety assessmentSR-Can. Technical report TR- 06-19, Swedish NuclearFuel and Waste Management Co, Stockholm, Sweden.

Smol JP. 1980. Fossil synuracean (Chrysophyceae) scales inlake sediments: a new group of palaeoindicators. CanadianJournal of Botany 58, 458–465.

27

S. SHALA

Smol, JP. 1983. Paleophycology of a high arctic lake near CapeHerschel, Ellesmere Island. Canadian Journal of Botany61, 2195–2204.

Smol JP. 2010. The power of the past: using sediments totrack the effects of multiple stressors on lake ecosystems.Freshwater Biology 55, 43–59.

Smol JP, Cumming BF. 2000. Tracking long-term changes inclimate using algal indicators in lake sediments. Journalof Phycology 36, 986–1011.

Solanki SK, Usoskin IG, Kromer B, Schussler MM, Beer J.2004. An unusually active Sun during recent decades com-pared to the previous 11,000 years. Nature 431, 1084–1087.

Solovieva N, Jones,VJ, Birks HJB, Appleby PG, Nazarova L.2008. Diatom responses to 20th century climate warmingin lakes from the northern Urals, Russia. Palaeogeography,Palaeoclimatology, Palaeoecology 259, 96–106.

Solovieva N, Jones VJ, Nazarova L, Brooks SJ, Birks HJB,Grytnes J-A, Appleby PG, Kauppila T, Kondratenok B,Renberg I, Ponomarev V. 2005. Palaeolimnological evi-dence for recent climate change in lakes from the northernUrals, Russia. Journal of Paleolimnology 33, 463–482.

Sundelin U. 1917. Fornsjostudier inom Stangans och Svartansvattenomraden med hansyn till den sen– och postglacialaklimatutvecklingen. Sveriges Geologiska Undersokning Ca16, Stockholm, Sweden.

Sundqvist H, Holmgren K, Lauritzen S-E. 2007. Stable iso-tope variations in stalagmites from northwestern Swedendocument climate and environmental changes during theearly Holocene. The Holocene 17, 259–267.

Szeroczynska K, Tatur A, Weckstrom J, Gasiorowski M,Noryskiewicz A, Sienkiewicz E. 2007. A multi-proxy as-sessment of the late-Quaternary environmental historyfrom a small subarctic lake, northwest Finnish Lapland.Journal of Paleolimnology 38, 25–47.

ter Braak CJF, Juggins S. 1993. Weighted averaging partialleast squares regression (WA-PLS): an improved methodfor reconstructing environmental variables from species as-semblages. Hydrobiologia 269/270, 485–502.

Tolonen KT, Hamalainen H, Holopainen IJ, Karjalainen J.2001. Influences of habitat type and environmental vari-ables on littoral macroinvertebrate communities in a largelake system. Archiv fur Hydrobiologie 152, 39–67.

van Donk E, van de Bund WJ. 2002. Impact of submergedmacrophytes including charophytes on phyto- and zoo-plankton communities: allelopathy versus other mecha-nisms. Aquatic Botany 72, 261–274.

Valiranta M. 2005. Plant macrofossil evidence of changes inaquatic and terrestrial environments in north-eastern Eu-ropean Russia and Finnish Lapland since late Weichselian.PhD thesis, University of Helsinki, Helsinki, Finland.

Valiranta M. 2006a. Long-term changes in aquatic plantspecies composition in North-eastern European Russiaand Finnish Lapland, as evidenced by plant macrofossilanalysis. Aquatic Botany 85, 224–232.

Valiranta M. 2006b. Terrestrial plant macrofossil records; pos-sible indicators of past lake-level fluctuations in north-eastern European Russia and Finnish Lapland? ActaPalaeobotanica 46, 235–243.

Valiranta M, Birks HH, Helmens K, Engels S, Piirainen M.2009. Early Weichselian interstadial (MIS 5c) summertemperatures were higher than today in the northernFennoscandia. Quaternary Science Reviews 28, 777–782.

Valiranta M, Kultti S, Nyman M, Sarmaja-Korjonen K. 2005.Holocene development of aquatic vegetation in shallowLake Njargajarvi Finnish Lapland, with evidence of water-level fluctuations and drying. Journal of Paleolimnology34, 203–215.

Valiranta M, Sarala P, Eskola, T. 2012. Uusia todisteita bore-aalisista olosuhteista Veikselinterstadiaalin aikana (”Newevidence of boreal conditions during Weichselian intersta-dial”) (with English summary). Geologi 64, 9–14.

Valiranta M, Weckstrom J, Siitonen S, Seppa H, Alkio J, Juu-tinen S, Tuittila E-S. 2011. Holocene aquatic ecosystemchange in the boreal vegetation zone of northern Finland.Journal of Paleolimnology 45, 339–352.

Vartiainen H. 1980. The petrography, mineralogy and petro-chemistry of the Sokli carbonatite massif, northern Fin-land. Geological Survey of Finland, Espoo.

Vasskog K, Paasche Ø, Nesje A, Boyle JF, Birks HJB. 2012. Anew approach for reconstructing glacier variability basedon lake sediments recording input from more than oneglacier. Quaternary Research 77, 192–204.

Velle G, Brodersen KP, Birks HJB, Willassen E. 2010. Midgesas quantitative temperature indicator species: Lessons forpalaeoecology. The Holocene 20, 989–1002.

Velle G, Brodersen KP, Birks HJB, Willassen E. 2012. Incon-sistent results should not be overlooked: A reply to Brookset al. (2012). The Holocene 22, 1501–1508.

Velle G, Larsen J, Eide W, Peglar SM, Birks HJB. 2005.Holocene environmental history and climate of Ratasjøen,a low-alpine lake in south-central Norway. Journal of Pa-leolimnology 33, 129–153.

Venalainen A, Tuomenvirta H, Pirinen P, Drebs A. 2005. Abasic climate data set 1961–2000 – description and illus-trations. Finnish Meteorological Institute, Reports 5.

von Post L. 1916. Om skogstradpollen i sydsvenska torf-mosselagerfo1jder. Geologiska Foreningens i StockholmForhandlingar 38, 384–390.

Waide RB, Willig MR, Mittelbach G, Steiner C, Gough L,Didson SI, Juday GP, Parmenter R. 1999. The relationshipbetween productivity and species richness. Annual Reviewof Ecology and Systematics 30, 257–300.

Walker MJC, Berkelhammer M, Bjorck S, Cwynar LC, FisherDA, Long AJ, Lowe JJ, Newnham RM, Rasmussen SO,Weiss H. 2012. Formal subdivision of the Holocene Se-ries/Epoch: a Discussion Paper by a Working Group ofINTIMATE (Integration of ice74 core, marine and ter-restrial records) and the Subcommission on QuaternaryStratigraphy (International Commission on Stratigraphy).Journal of Quaternary Science 27, 649–659.

Weckstrom J. 2001. Assessment of diatoms as markers of envi-ronmental change in northern Fennoscandia. PhD thesis,University of Helsinki, Helsinki, Finland.

Weckstrom J, Snyder JA, Korhola A, Laing TE, MacDonaldGM. 2003. Diatom inferred acidity history of 32 lakes onthe Kola peninsula, Russia. Water, Air, and Soil Pollution149, 339–361.

Weckstrom J, Seppa H, Korhola A. 2010. Climatic influenceon peatland formation and lateral expansion in sub-arcticFennoscandia. Boreas 39, 761–769.

Weltje GJ, Tjallingii R. 2008. Calibration of XRF core scan-ners for quantitative geochemical logging of sedimentcores: Theory and application. Earth and Planetary Sci-ence Letters 274, 423–438.

Wetzel RG. 2001. Limnology: Lake and River Ecosystems.Third Edition. Academic Press, San Diego, CA.

Witkowski A, Lange–Bertalot H, Metzeltin D. 1995/96.The diatom species Fragilaria martyi (Heribaud) Lange-Bertalot, identity and ecology. Archiv fur protistenkunde146, 281–292.

Wunsam S, Schmidt R. 1995. A diatom-phosphorous transfer-function for alpine and prealpine lakes. Memorie dell’ Is-tituto Italiano di Idrobiologia 53, 85–99.

28

Appendix A. Identified diatom taxa in the Holocene and MIS 5d-c record

Diatom groups Holocene MIS 5c

Planktonic taxaAulacoseira ambigua (Grunow) Simonsen x xAulacoseira islandica (O. Muller) Simonsen x xAulacoseira subartica (O. Muller) Haworth x xCyclostephanos dubius (Fricke in A. Schmidt) Round xCyclotella meneghiniana Kutzing xCyclotella ocellata Pantocsek xCyclotella radiosa (Grunow) Lemmermann xCyclotella rossii Hakansson xCyclotella spp (Kutzing) Brebisson xStephanodiscus alpinus Hustedt xStephanodiscus cf. hantzschii Grunow (in Cleve & Grunow) xStephanodiscus medius Hakansson xStephanodiscus parvus Stoermer & Hakansson x xStephanodiscus spp Ehrenberg x

Tychoplanktonic taxaAulacoseira distans (Ehrenberg) Simonsen xAulacoseira lirata v. biseriata (Grunow) Haworth xAulacoseira valida (Grunow) Krammer x xFragilaria capucina Desmazieres x xFragilaria capucina v. mesolepta (Rabenhorst) Rabenhorst xFragilaria capucina v. rumpens (Kutzing) Lange-Bertalot xFragilaria capucina v. vaucheriae (Kutzing) Lange-Bertalot xMelosira varians Agardh x xNitzschia intermedia Hantzsch x xTabellaria flocculosa (Roth) Kutzing x xUlnaria ulna (Nitzsch) Compere x x

Benthic taxaAchnanthes bioretii Germain xAchnanthes clevei Grunow xAchnanthes conspicua Mayer xAchnanthes lanceolata (Brebisson) Grunow x xAchnanthes lanceolata ssp biporoma xAchnanthes lanceolata ssp frequentissima Lange-Bertalot x xAchnanthes laterostrata Hustedt xAchnanthes oblongella Oestrup xAchnanthes peragalli Brun & Heribaud x xAchnanthes ziegleri Lange-Bertalot xAchnanthidium minutissimum (Kutzing) Czarnecki xAmphora aequalis Krammer xAmphora inariensis Krammer x xAmphora libyca Ehrenberg x xAmphora ovalis (Kutzing) Kutzing x xAneumastus pseudotusculus (Hustedt) Cox & Williams xAneumastus tuscula (Ehrenberg) Mann & Stickle x xAnomoeoneis sphaerophora (Ehrenberg) Pfitzer x xAulacoseira alpigena (Grunow) Krammer xAulacoseira crassipunctata Krammer xAulacoseira italica (Ehrenberg) Simonsen x xAulacoseira italica v. tenuissima (Grunow) Simonsen xCaloneis bacillum (Grunow) Cleve xCaloneis schumanniana (Grunow) Cleve xCaloneis ventricosa (Ehrenberg) Meister x xCampylodiscus noricus v. hibernicus (Ehrenberg) xCocconeis disculus (Schumann) Cleve xCocconeis neodiminuta Krammer xCocconeis placentula Ehrenberg x xCocconeis placentula v. euglypta (Ehrenberg) Grunow x xCocconeis placentula v. klinoraphis Geitler xCocconeis placentula v. lineata (Ehrenberg) van Heurck x xCocconeis placentula v. pseudolineata Geitler x xCocconeis pseudothumensis Reichardt xCraticula cuspidata (Kutzing) Mann x xCymatopleura solea (Brebisson & Godey) W. Smith x xCymatopleura spp W. Smith xCymbella affinis (Kutzing) xCymbella amphicephala (Naegeli) xCymbella aspera (Ehrenberg) Cleve xCymbella cistula (Ehrenberg) Kirchner x xCymbella cuspidata Kutzing x xCymbella cymbiformis Agardh xCymbella ehrenbergii Kutzing x xCymbella helvetica Kutzing xCymbella heteropleura (Ehrenberg) Kutzing xCymbella minuta Hilse xCymbella paucistriata Cleve-Euler xCymbella proxima Reimer in Patrick and Reimer x

Cymbella reichardtii Krammer xCymbella simonsenii Krammer xCymbella subaequalis Grunow in Van Heurck x xCymbella subcuspidata Krammer xDidymosphenia geminata (Lyngbye) M. Schmidt xDiploneis elliptica v. elliptica (Kutzing) Cleve x xDiploneis oculata (Brebisson) Cleve xDiploneis parma Cleve xEllerbeckia arenaria f. teres (Brun) Crawford xEncyonema brehmii (Hustedt) Mann xEncyonema silesiacum (Bleisch in Rabenhorst) Mann x xEpithemia adnata (Kutzing) Brebisson x xEpithemia frickei Krammer xEpithemia sorex v. sorex Kutzing x xEpithemia spp Brebisson x xEpithemia turgida v. granulata (Ehrenberg) Brun x xEpithemia turgida v. turgida (Ehrenberg) Kutzing xEunotia curvata (Kutzing) Lagerstedt xEunotia implicata Norpel, Lange-Bertalot & Alles xEunotia praerupta Ehrenberg x xEunotia praerupta v. papilio (Grunow) Norpel xFallacia helensis (Schulz) Mann xFragilaria bidens (Ehrenberg) Grunow xFragilaria nitzschioides (Grunow) xFragilaria robusta (Fusey) Manguin x xFragilariforma neoproducta Lange-Bertalot x xFrustulia amphipleuroides (Grunow in Cleve & Grunow) Cleve-Euler xFrustulia vulgaris (Thwaites) De Toni x xGeissleria paludosa (Hustedt) Lange-Bertalot & Metzeltin xGomphonema acuminatum v. acuminatum Ehrenberg x xGomphonema angustatum (Kutzing) Rabenhorst x xGomphonema clavatum Ehrenberg x xGomphonema gracile Ehrenberg x xGomphonema minutum (Agardh) Agardh x xGomphonema olivaceum (Hornemann) Brebisson x xGomphonema parvulum v. parvulum (Kutzing) Kutzing x xGomphonema spp Agardh x xGomphonema truncatum Ehrenberg x xGyrosigma acuminatum (Kutzing) Rabenhorst xHippodonta capitata (Ehrenberg) Lange-Bertalot, Metzeltin & Witkowski x xHippodonta lueneburgensis (Grunow) Lange-Bertalot, Metzeltin & Witkowski x xMartyana martyi (Heribaud) Round x xNavicula abiskoensis Hustedt xNavicula absoluta Hustedt x xNavicula amphibola Cleve xNavicula cari Ehrenberg x xNavicula clementioides Hustedt xNavicula clementis Grunow xNavicula constans Hustedt x xNavicula cryptocephala Kutzing xNavicula cryptotenella Lange-Bertalot xNavicula decussis Østrup x xNavicula digitoradiata (Gregory) Ralfs x xNavicula eidrigeana (Carter) xNavicula exigua (Gregory) Grunow xNavicula explanata Hustedt xNavicula gastrum (Ehrenberg) Kutzing x xNavicula hambergii Hustedt in Hamberg xNavicula harderi Hustedt in Brendmuhl xNavicula joubaudii Germain xNavicula laevissima Kutzing x xNavicula lapidosa Krasske xNavicula laterostrata Hustedt xNavicula menisculus Schumann x xNavicula menisculus v. upsaliensis Grunow x xNavicula placentula (Ehrenberg) Kutzing x xNavicula protracta (Grunow) Cleve xNavicula pseudoventralis Hustedt xNavicula pusilla W. Smith xNavicula radiosa Kutzing x xNavicula reinhardtii Grunow x xNavicula rhynchocephala Kutzing x xNavicula rotunda Hustedt xNavicula schoenfeldii Hustedt xNavicula scutelloides W. Smith ex. Gregory xNavicula spp Bory x xNavicula submuralis Hustedt xNavicula subrotundata Hustedt xNavicula viridula (Kutzing) Ehrenberg x xNavicula vitabunda Hustedt xNavicula vulpina Kutzing xNeidium affine (Ehrenberg) Pfitzer xNeidium ampliatum (Ehrenberg) Krammer x xNeidium dubium (Ehrenberg) Cleve x xNeidium iridis (Ehrenberg) Cleve x

Neidium spp Pfitzer xNitzschia amphibia Grunow xNitzschia fossilis Grunow xNitzschia spp Hassall xOrthoseira dendroteres (Ehrenberg) Crawford xPinnularia abaujensis (Pantoscek) Ross x xPinnularia interrupta W. Smith x xPinnularia microstauron (Ehrenberg) Cleve x xPinnularia pulchra Østrup x xPinnularia spp Ehrenberg x xPinnularia streptoraphe Cleve x xPinnularia subcapitata Gregory x xPinnularia viridis (Nitzsch) Ehrenberg xPinnularia viridis f. cuneatiformis Cleve-Euler xPlaconeis elginensis (Gregory) Cox x xPseudostaurosira brevistriata (Grunow in Van Heurck) Williams & Round x xPseudostaurosira pseudoconstruens (Marciniak) Williams & Round xRhoicosphenia abbreviata (Agardh) Lange-Bertalot xRhopalodia gibba v. gibba (Ehrenberg) O. Muller x xSellaphora bacillum (Ehrenberg) Mann x xSellaphora pupula (Kutzing) Mereschkowsky x xStauroneis acuta W. Smith xStauroneis anceps v. anceps Ehrenberg x xStauroneis phoenicenteron (Nitzsch) Ehrenberg x xStauroneis producta Grunow xStauroneis spp Ehrenberg x xStaurosira binodis (Ehrenberg) Lange-Bertalot x xStaurosira construens Ehrenberg x xStaurosira venter (Ehrenberg) H.Kobayasi x xStaurosirella lapponica (Grunow in Van Heurck) Williams & Round x xStaurosirella leptostauron (Ehrenberg) Williams & Round x xStaurosirella pinnata v. pinnata (Ehrenberg) Williams & Round x xSurirella linearis v. constricta Hustedt xSurirella linearis v. linearis W. Smith xSynedra parasitica (W. Smith) Hustedt xSynedra parasitica v. subconstricta (Grunow in Van Heurck) Hustedt x xTetracyclus emarginatus (Ehrenberg) W. Smith x xTetracyclus glans (Ehrenberg) Mills xUlnaria capitata (Ehrenberg) Compere x

Aerophilic taxaAchnanthes exigua Grunow xDiploneis oblongella (Naegeli) Cleve-Euler x xHantzschia amphioxys (Ehrenberg) Grunow in Cleve & Grunow x xLuticola mutica (Kutzing) Mann xNavicula minima Grunow xNeidium bisulcatum v. subampliatum Krammer x xPinnularia borealis v. scalaris (Ehrenberg) Rabenhorst x xReimeria sinuata (Gregory) Kociolek & Stoermer x x

Rheophilic taxaAmphora pediculus (Kutzing) Grunow xAulacoseira italica f. crenulata (Ehrenberg) R. Ross in Hartley x xCymbella naviculiformis Auerswald x xDiatoma hyemalis (Roth) Heiberg xDiatoma mesodon (Ehrenberg) Kutzing x xEunotia pectinalis v. minor (Kutzing) Rabenhorst xFragilaria virescens Ralfs xMeridion circulare v. circulare (Greville) Agardh x xMeridion circulare v. constrictum (Ralfs) van Heurck x x

Reworked / Tertiary taxaCoscinodiscus spp Ehrenberg xHemiaulus spp Ehrenberg x xOpephora gemmata (Grunow) Hustedt xParalia sulcata (Ehrenberg) Cleve x xParalia sulcata v. biseriata Grunow xStephanopyxis spp Ehrenberg x

Unknown ecologyAchnanthes spp Bory xAmphora spp Ehrenberg x xAulacoseira spp Thwaites x xCaloneis spp Cleve xCymbella spp Agardh x xDiatoma spp Bory x xEunotia spp Ehrenberg x xFragilaria spp Lyngbye x xNavicula sect. Minusculae xNeidium spp Pfitzer x xNitzschia spp Hassall x xSurirella spp Turpin x xTabellaria spp Ehrenberg x x

Appendix B. List of synonyms

Taxon names used in this study Basionyms

Achnanthidium minutissimum (Kutzing) Czarnecki Achnanthes minutissima KutzingAneumastus pseudotusculus (Hustedt) Cox & Williams Navicula pseudotuscula HustedtAneumastus tuscula (Ehrenberg) Mann & Stickle Navicula tuscula (Ehrenberg) GrunowAulacoseira italica f. crenulata (Ehrenberg) Ross in Hartley Aulacoseira crenulata (Ehrenberg) ThwaitesCaloneis ventricosa (Ehrenberg) Meister Caloneis silicula (Ehrenberg) CleveCraticula cuspidata (Kutzing) Mann Navicula cuspidata (Kutzing) KutzingEncyonema brehmii (Hustedt) Mann Cymbella brehmii HustedtEncyonema silesiacum (Bleisch in Rabenhorst) Mann Cymbella silesiaca BleischEunotia curvata (Kutzing) Lagerstedt Eunotia bilunaris (Ehrenberg) MillsEunotia pectinalis v. minor (Kutzing) Rabenhorst Eunotia minor (Kutzing) Grunow in Van HeurckFallacia helensis (Schulz) Mann Navicula helensis SchulzFragilaria dilatata (Brebisson) Lange-Bertalot Ulnaria capitata (Ehrenberg) CompereFragilariforma neoproducta (Lange-Bertalot) Williams & Round Fragilaria neoproducta Lange-BertalotFrustulia amphipleuroides (Grunow in Cleve & Grunow) A. Cleve-Euler Frustulia rhomboides v. amphipleuroides

(Grunow) De ToniGeissleria paludosa (Hustedt) Lange-Bertalot & Metzeltin Navicula ignota var. palustris (Hustedt) LundHippodonta capitata (Ehrenberg) Lange-Bertalot, Metzeltin & Witkowski Navicula capitata EhrenbergHippodonta lueneburgensis (Grunow) Lange-Bertalot, Metzeltin & Witkowski Navicula capitata v. lueneburgensis (Grunow)

PatrickLuticola mutica (Kutzing) Mann Navicula mutica KutzingMartyana martyi (Heribaud) Round Fragilaia leptostauron v. martyi (Heribaud)

Lange-BertalotPinnularia abaujensis (Pantoscek) Ross Pinnularia gibba EhrenbergPlaconeis elginensis (Gregory) Cox Navicula elginensis (Gregory) RalfsPseudostaurosira brevistriata (Grunow in Van Heurck) Williams & Round Fragilaria brevistriata GrunowPseudostaurosira pseudoconstruens (Marciniak) Williams & Round Fragilaria pseudoconstruens MarciniakReimeria sinuata (Gregory) Kociolek & Stoermer Cymbella sinuata GregorySellaphora bacillum (Ehrenberg) Mann Navicula bacillum EhrenbergSellaphora pupula (Kutzing) Mereschkowsky Navicula pupula KutzingStaurosira binodis (Ehrenberg) Lange-Bertalot Fragilaria construens v. binodis (Ehrenberg)

HustedtStaurosira construens Ehrenberg Fragilaria construens v. construens (Ehrenberg)

HustedtStaurosira venter (Ehrenberg) H.Kobayasi Fragilaria construens v. venter (Ehrenberg)

HustedtStaurosirella lapponica (Grunow in Van Heurck) Williams & Round Fragilaria lapponica GrunowStaurosirella leptostauron (Ehrenberg) Williams & Round Fragilaria leptostauron v. leptostauron

(Ehrenberg) HustedtStaurosirella pinnata (Ehrenberg) Williams & Round Fragilaria pinnata EhrenbergSynedra parasitica (W. Smith) Hustedt Fragilaria parasitica (W. Smith) GrunowSynedra parasitica v. subconstricta (Grunow in Van Heurck) Hustedt Fragilaria parasitica v. subconstricta GrunowUlnaria ulna (Nitzsch) Compere Fragilaria ulna (Nitzsch) Lange-Bertalot

Thank you...I am very thankful for where I am today and the people I have met on my way here have played a large partin making it a wonderful experience.

To my supervisors Karin Helmens, Krister Jansson, Peter Kuhry and Jan Risberg (StockholmUniversity); thank you for giving me the opportunity to pursue this PhD project, for the support you havegiven me through the years and for always having your doors open for me when I needed advice. Pullingthis through without your help would have been difficult. Special thanks to my co-supervisors Peter Kuhryand Jan Risberg who have been instrumental for my continued studies at PhD level. Peter’s inspiringlectures first stirred my interest for palaeoenvironmental studies, while Jan’s outstanding supervision duringmy master’s thesis, motivated me to continue my studies in Quaternary Geology.

I am also very thankful to my external supervisor Minna Valiranta (University of Helsinki) for herinterest in my studies, the tremendous support she has given me during these past few years and for giving methe opportunity to spend three weeks at the Environmental Change Research Unit. My time at ECRU wasmade enjoyable to a large part by the wonderful ECRU members whose kindness, enthusiasm, team-work,excellence and impressive skills in Molkky will always be an inspiration for me. Thank you for making me feelat home from day one.

During my studies I’ve had the pleasure of working alongside several collaborators whose enthusiasmand expertise has helped broaden the value of this work. Jan Weckstrom has always been of great support.His analytical thinking made him just the right kind of person with whom I could discuss the curious world ofdiatoms and transfer-functions. Malin Kylander; working with XRF-data never made much sense until youcame into my life. Thank you for making everything easier and much more enjoyable, your are an inspiration.Sakari Salonen and Tomi Luoto, thank you for the countless discussions we’ve had this last year, forpatiently answering all my questions and quickly responding to my e-mails. Collaborating with you has trulybeen a pleasure. I have also had the priviledge of receiving support from Antonio Martınez Cortizas,Gavin Simpson, Liselott Wilin, Malin Johansson, Martin Margold and Paivi Kaislahti Tillman,which has undeniably made many of my days easier. Hilary Birks was kind enough to introduce me to plantmacrofossil analysis for which I am deeply grateful.

INK would not have been the great working place it is were it not for the wonderful people who workhere. I am particularly thankful to Susanna Blandman, Linus Richert, Bjorn Gunnarsson, CarinaHenriksson, Erik Hansson, Helle Skanes, Stefan Wastegard, Sven Karlsson and Yanduy Cabrerawho have always done a great job and been supportive in every way possible. To all my fellow PhDstudents both at INK and elsewhere in the Geobuilding; thank you for all the motivating chats, laughs, dis-cussions, after-works and relentless efforts to win the pub-quiz, which unfortunately only ever resulted in 2nd

place. Ewa, Alistair, Steve, Ping, Annika, Marika, Moo and Reto; you’ve made life a little lovelier for me.

To all my friends, especially Anna, Alba, Carolina and Nando; the good times and laughs thatwe shared, our long chats, wandering around, discovering new places and countless other little things havemade this last year bearable. My arctic friends; Svalbard would not have been as amazing were it not foryour company. Lewis, Ivan, Isabelle; thanks for being an inspiration, for always answering my questionsand for cheering me up.

I am beyond grateful to my family who has been with me on every step of this road. To mom, dad,Jeton, Enver, Edona, Qendresa, Qendrim, my brother-in-law, my lovely sisters-in-law and mybeautiful nephew who has reoriented my life; thank you for supporting me through all my years at theuniversity, for never doubting me, for always being there. Your love and support has pulled me through, thisis for you.

Last but not least I’d like to thank all the wonderful people I’ve met at various courses, workshopsand conferences, who have shared their inspiring work with me and taken interest in mine.