J Paleolimnol (2011) 45:339–352 DOI 10.1007/s10933-011-9501-5

ORIGINAL PAPER

Holocene aquatic ecosystem change in the boreal vegetation zone of northern Finland Minna Va¨liranta • Jan Weckstro¨m • Susanna Siitonen • Heikki Seppa¨ • Jyri Alkio Sari Juutinen • Eeva-Stiina Tuittila

•

Received: 5 July 2010 / Accepted: 14 January 2011 / Published online: 29 January 2011 Ó Springer Science+Business Media B.V. 2011

Abstract We studied multiple variables in a sediment core from Lake Kipoja¨rvi, northern Finland, to investigate Holocene ecosystem changes in relation to catchment characteristics and known climate

M. Va¨liranta (&)  J. Weckstro¨m  S. Siitonen Department of Environmental Sciences, University of Helsinki, P.O. Box 65, 00014 Helsinki, Finland e-mail: [email protected] J. Weckstro¨m e-mail: [email protected] S. Siitonen e-mail: [email protected] S. Siitonen Arctic Centre, University of Lapland, P.O. Box 122, 96101 Rovaniemi, Finland H. Seppa¨  J. Alkio Department of Geosciences and Geography, University of Helsinki, P.O. Box 64, 00014 Helsinki, Finland e-mail: [email protected] J. Alkio e-mail: [email protected] S. Juutinen  E.-S. Tuittila Peatland Ecology Group, Department of Forest Sciences, University of Helsinki, P.O. Box 27, Helsinki, Finland e-mail: [email protected] E.-S. Tuittila e-mail: [email protected]

variations. We focused on a forested catchment because previous paleolimnological studies conducted in Fennoscandia focused mainly on subarctic lakes within a range of shifting treeline(s). Data on aquatic macrophytes, diatoms, Cladocera, C:N ratio, organic matter (LOI) and regional vegetation (pollen), revealed a three-phase limnological development. The early Holocene, species-rich, mesotrophic lake was transformed into an oligotrophic, speciespoor aquatic ecosystem by the early middle Holocene, ca. 7,500 cal years BP, earlier than has generally been reported. The transition involved considerable changes in aquatic macrophytes. Changes in the Cladocera and diatom communities appear to have been linked to aquatic macrophyte development, which in turn, was probably regulated by catchment development and hydrology, and a consequent decrease in nutrient input from the catchment. During the more humid late Holocene, surface flow from the catchment probably increased, but the lake’s nutrient status remained oligotrophic. Possible reasons for low nutrient concentration in the late Holocene include: 1) slower biogeochemical cycling due to cooler climate, 2) a new hydrologic outlet and associated shorter water-retention times, and 3) accelerated peatland development in the catchment that affected water flow patterns and nutrient cycling. Keywords Holocene  Aquatic macrophytes  Cladocera  Diatoms  Sediment proxies  Northern boreal lake  Aquatic ecosystem change

123

340

Introduction In Fennoscandia, northern boreal lakes, i.e. those that have always been located within a coniferous forest, are under-represented among Holocene paleoecological and paleolimnological studies, while subarctic lakes, located at or beyond shifting treeline(s), have been studied intensively (Seppa¨ and Weckstro¨m 1999; Barnekow 2000; Rose´n et al. 2001; Va¨liranta et al. 2005). Similarly, while knowledge of Holocene diatom and Cladocera community dynamics has increased, data on Holocene development of aquatic macrophyte communities are scarce and available for only a few Fennoscandian northern boreal lakes (Va¨liranta et al. 2005; Va¨liranta 2006). This bias is unfortunate because aquatic macrophytes are essential components of the ecosystem. They provide habitat, substrate and shelter for key food web organisms and play a vital role in nutrient cycling (Carpenter and Lodge 1986; van Donk and van de Bund 2002). In small, shallow lakes, the littoral flora contributes significantly to primary productivity and may dominate and regulate the metabolism of the entire lake ecosystem (Wetzel 2001). We used a multi-proxy paleolimnological approach (macrofossils, Cladocera, diatoms, pollen, organic matter [loss-on-ignition LOI] and C:N ratio) to infer the Holocene development of Lake Kipoja¨rvi, Finnish Lapland, in the context of changes in the surrounding terrestrial ecosystems, including the adjacent peatland, associated with large-scale Holocene climate variations. Various aquatic and terrestrial components were included to understand 1) how different biotic groups respond to long-term environmental changes, (2) how these responses are related, and 3) whether they are regulated mainly by internal and/or external (climate) drivers (Lotter and Birks 2003). Such studies are scarce in Fennoscandia. The best examples are the Kra˚kenes and Ra˚ta˚sjøen studies in Norway (Birks et al. 2000; Velle et al. 2005) and the Njargaja¨vri study in eastern Finnish Lapland (Va¨liranta et al. 2005; SarmajaKorjonen et al. 2006). Therefore, the principal aim of this study was to increase our knowledge of long-term aquatic ecosystem evolution in the northern boreal vegetation zone. Study site Lake Kipoja¨rvi is situated in the northern boreal vegetation zone in Finnish Lapland (69°180 N,

123

J Paleolimnol (2011) 45:339–352

27°320 E), 159.4 m above sea level (Fig. 1). Mean annual and July air temperatures are -1 and 13.1°C, respectively, and annual precipitation is 395 mm. The lake is covered by ice from the end of October to the end of May. The catchment-lake ratio is *7:1. Lake Kipoja¨rvi lies in glacio-fluvial terrain and is bordered by an esker to the east. Mineral soils that cover *50% of the catchment are dominated by pine-birch forest with abundant dwarf shrubs on the field layer. Kiposuo, a large, open aapa mire that was named by the authors, bounds the lake to the north. The lake area is *11 ha and the maximum water depth is 1.5 m. The lake has a small inlet that flows through the adjacent aapa mire and a narrow (*1 m wide), relatively shallow (*0.5 m deep) paludified outlet in the south. The lake is oligotrophic and mesohumic with a mean total organic carbon (TOC) content in summer 2006 of 7.5 mg l-1 and a pH of 7.3. Aquatic macrophyte species in the lake include Nuphar pumila, Myriophyllum alterniflorum, Potamogeton alpinus and Sparganium cf. angustifolium. Their modern occurrence is patchy. Shoreline species are Utricularia intermedia, Carex rostrata, Menyanthes trifoliata, Equisetum fluviatile and Potentilla palustris.

Materials and methods Field sampling A 329-cm sediment core was obtained from the deepest part of the lake (Fig. 1) in winter 2006 using a Livingstone piston corer. Bottommost sediment was obtained with a Russian peat corer. After coring, sediment was immediately sealed and stored intact in a cold room. The core was cut into 2-cm-thick slices and subsamples from selected intervals were analyzed for plant macrofossils, Cladocera, diatoms, pollen, LOI and C:N. Laboratory analyses Terrestrial plant material from the core was dated by Accelerator Mass Spectrometry (AMS). Dating was carried out at the Dating Laboratory of University of ˚ ngstro¨m Laboratory, Helsinki, Finland and the A University of Uppsala, Sweden. Radiocarbon dates (BP) were calibrated using the program Calib501,

J Paleolimnol (2011) 45:339–352

Fig. 1 Lake Kipoja¨rvi in eastern Finnish Lapland. The study site is located within the northern boreal vegetation zone, in subarctic climate conditions. The white dot indicates the

INTCAL04 calibration curve (Stuiver and Reimer 1993). An age-depth model was created by fitting a second-order polynomial curve to the calibrated dates (Fig. 2). One date, Hela-1606, was out of sequence and was omitted from the age-depth model. All other dates were in stratigraphic order (Table 1). Three additional bottom peat samples from the adjacent mire, Kiposuo, were dated using bryophyte remains. These analyses were carried out at the Poznan Radiocarbon Laboratory, Poland. For organic matter (LOI) and C:N ratio analyses, two consecutive 2-cm-thick slices were combined. Subsamples of known volume were analyzed for organic content, and carbon and nitrogen concentrations. Subsamples were dried at 70°C to constant mass, after which they were milled for determination of LOI at 550°C. Carbon and nitrogen concentrations of dried subsamples were analyzed using an elemental analyzer (Vario Max CN, Elementar Analysensysteme GmbH) and are presented as mass ratios (C:N). Plant remains were analysed from alternate 2-cm subsamples, with the exception of the bottommost 40 cm, where analysis was carried out contiguously. A total of 93 subsamples were analysed. No chemical treatment was necessary. The volume of subsamples varied between 10 and 25 cm3, with 20 cm3 being the

341

sampling site. The arrows show the position of the inlet and the outlet. Copyright for the aerial photo Maanmittauslaitos

Fig. 2 Age-depth model of the Lake Kipoja¨rvi sediment core

most common volume. Sediment volume was determined by displacement of water in a measuring cylinder. The sediment was sieved using a 124-lm

123

342

J Paleolimnol (2011) 45:339–352

Table 1 Results of radiocarbon analyses of Lake Kipoja¨rvi sediment core and the adjacent mire Kiposuo. O = Omitted from the age-depth model Sample code

Depth (cm)

Age

14

C

1 sigma range

Cal 14C median 1 sigma

Lab no

Kipoja¨rvi III Kipoja¨rvi III Kipoja¨rvi III

28–30

1,065 ± 35 BP

932–1048

970

Hela-1361

76–78

2,940 ± 35 BP

3,008–3,204

3,100

Hela-1362

96–98

3,435 ± 35 BP

3,637–3,814

3,690

Hela-1363

Kipoja¨rvi III Kipoja¨rvi III Kipoja¨rvi III

140–142

5,625 ± 65 BP

6,317–6,463

6,410

Hela-1364

184–186

6,765 ± 50 BP

7,586–7,656

7,620

Hela-1365

236–246

6,135 ± 45 BP

6,949–7,156

7,040

Hela-1606 O

Kipoja¨rvi III

326–328

8,090 ± 60 BP

8,798–9,128

9,020

Hela-1368

Kiposuo A

320–321

7,840 ± 40 BP

8,555–8,681

8,620

Poz-20664

Kiposuo B

165–166

8,830 ± 50 BP

8,743–10,120

9,900

Poz-18417

Kiposuo C

155–156

8,350 ± 50 BP

9,304–9,440

9,370

Poz-18416

mesh and the residue was systematically examined using a stereo-microscope and a high-magnification light microscope. Identification of vegetative remains of Potamogeton species was based on leaf tips. Narrow and pointed/blunt leaf tips with three veins are hereafter referred to as P. berchtoldii/pusillus, as reliable differentiation between these species was not possible. Only selected taxa are presented and discussed. Pollen analysis was carried out at 4-cm intervals on 1-cm3 subsamples, using standard KOH, HF and acetolysis treatments (Fægri and Iversen 1989). Each subsample was spiked with two Lycopodium tablets for pollen concentration and pollen accumulation rate calculations (Stockmarr 1971). A minimum of 500 terrestrial pollen and spores were identified from each sample. Percentages of terrestrial pollen and spore taxa were calculated on the basis of their total sum and percentages of Sphagnum on the basis of the total sum of terrestrial taxa plus Sphagnum. Only selected taxa are presented and discussed. Cladocera were analysed at 8-cm intervals using 1-cm3 subsamples. Sample preparation and identification were mainly based on Szeroczyn´ska and Sarmaja-Korjonen (2007). The sediment was heated for 30 min in 10% KOH and rinsed using a 38-lm sieve. Cladoceran remains were identified from permanent slides with a light microscope at 2009 magnification. A minimum of 200 remains from each subsample was counted, except for four samples below 8,500 cal years BP, which contained very low numbers of remains. The number of individuals was reconstructed using the most common fragment of

123

each species, and the results are expressed as relative abundances. Unidentified small and medium-sized Alona-type carapaces and headshields are grouped as small Alona spp. In addition to the obvious Bosmina (Eubosmina) coregoni and B. (E.) longispina-type headpores, intermediate forms were also common, precluding reliable identification. These are grouped under a collective name Eubosmina. Species diversity was calculated for each sample with Simpson’s Index of Diversity (Simpson 1949). It takes into account species richness (number of species) and relative abundance. Diatom samples were prepared using H2O2 digestion and HCl treatment, and the cleaned diatoms were mounted on microscope slides with Naphrax. At least 300 diatom valves per subsample were counted at 8-cm core intervals. Diatoms were enumerated on random slide transects using a light microscope with phase-contrast at 1,0009 magnification. Taxonomic identification was mainly based on Krammer and Lange-Bertalot (1986–1991). The pH history of Lake Kipoja¨rvi was reconstructed using a modern diatom-water pH calibration data set that consisted of 98 surface-sediment diatom assemblages and corresponding pH measurements (Seppa¨ and Weckstro¨m 1999; Weckstro¨m 2001). The 1-component weighted average partial least squares (WA-PLS) model is robust with a coefficient of determination (r2) of 0.68 between observed and diatom-inferred pH values, and a root mean square error of prediction (RMSEP) of 0.31 pH units. The quantitative diatom-based TOC reconstruction is based on a subset of 200 of the 388 TOC data-set

J Paleolimnol (2011) 45:339–352

343

lakes from the European diatom database (http:// craticula.ncl.ac.uk/Eddi/jsp/index.jsp; Juggins 2001). The locally weighted weighted averaging (LWWA) classical deshrinking model yielded an r2 of 0.50 and a RMSEP of 2.3 mg l-l. Constrained optimal sum of squares partitioning (Birks and Gordon 1985) was implemented to identify the most significant shifts in fossil pollen, Cladocera and diatom assemblages. The number of statistically significant zones was calculated using a broken-stick model as described in Bennett (1996). Optimal partitioning was performed using the program ZONE 1.2 (Lotter and Juggins 1991). The zonation based on pollen percentage data was also applied to the pollen influx data. The macrofossil record was visually divided into macrofossil assemblage zones based on the variation in remains of aquatic taxa. The diatombased pH and TOC reconstructions were made using the program C2 (version 1.5.0, Juggins 2003).

Results

Fig. 3 Loss on ignition, carbon/nitrogen (C:N) ratio and plant macrofossil record, with only selected taxa presented. Bars indicate the actual number of finds and a symbol (?) is used to

describe the relative abundance of the remains (? rare, ?? occasional and ??? abundant)

Chronology and sediment composition The sediment consisted of gyttja. The high LOI value (33%) and age of 9,200 cal years BP for the bottommost sample (Fig. 3) suggest that the core does not comprise the entire Holocene epoch. This conclusion is supported by the fact that basal fossil pollen and diatom assemblages lack the first post-glacial components that are typical for the region. Basal pollen assemblages from the earliest Holocene are usually characterized by high ([20%) values of Ericaceae pollen (Hyva¨rinen 1975; Seppa¨ 1998), and diatom assemblages are dominated by Fragilaria species (Seppa¨ and Weckstro¨m 1999). The organic content shows an increasing trend from the bottom upwards, reaching *70% by 8,000 cal years BP. After that, LOI remains relatively stable (60–65%) until *4,500 cal years BP. The high

123

344

LOI probably reflects low minerogenic input from the catchment as well as abundant aquatic macrophytes. From *4,000 cal years BP onwards, LOI remains at *30%. The transition coincides with the major middle Holocene/late Holocene climate shift towards cooler and more humid conditions (Seppa¨ and Hammarlund 2000) that probably resulted in more effective surface runoff from the catchment and reduced productivity in the lake and in the catchment. Variation in C:N is subtle, from only 9 to 13 (Fig. 3), but the observed pattern of higher early and middle Holocene values and lower late Holocene values coincides with the LOI data and the development of the aquatic macrophyte community (Fig. 3). In general, high C:N ratios ([10) indicate input of organic matter from higher plants (littoral/wetland/ terrestrial), and low ratios (\10) reflect input from phytoplankton (Wetzel 2001). Regional vegetation history The pollen diagrams (Fig. 4a, b) can be divided into two pollen assemblage zones, PAZ I (9,200–8,700 cal years

Fig. 4 a A simplified percentage pollen diagram from Lake Kipoja¨rvi. Two pollen assemblage zones, PAZ 1 and PAZ 2, are delimited by the solid line. Sub-zones I and II within PAZ 2 are separated by the dotted line. b The pollen accumulation rate diagram for key taxa

123

J Paleolimnol (2011) 45:339–352

BP) and PAZ II (8,700 to present). PAZ II is divided into two sub-zones, SZ 1 and SZ 2. PAZ I is strongly dominated by Betula (60–80%). The pollen accumulation rate (PAR) of 5,000–10,000 grains cm-2 year-1 (Fig. 4b) is much higher than the typical value of 500–2,000 grains cm-2 year-1 for modern birch forests in northern Fennoscandia (Seppa¨ and Hicks 2006). The reason for the extraordinarily high values might be that the basal date is too young (9,200 cal years BP), which resulted in an inaccurate age-depth model and artificially high sediment accumulation rates for the bottom section. There is, however, little doubt that the early Holocene forest was dominated by birch until 8,700 cal years BP, as the plant macrofossil record also indicates the dominance of birch forest and absence of pine forest (Fig. 3). The rapid increase in Pinus pollen at 8,700 cal years BP defines the onset of the PAZ II SZ1 (Fig. 4a, b). The percentage value rises gradually up to 60% and the PAR value rises more rapidly, from 300 to 400 to over 2,000 grains cm-2 year-1, reflecting a gradual change in forest structure as a result of competitive exclusion of birch by pine (Hyva¨rinen

J Paleolimnol (2011) 45:339–352

1975). Furthermore, the first macroscopic Pinus remains date to 8,700 cal years BP (Fig. 3). The main feature of the terrestrial vegetation during the middle to late Holocene, as reflected by the pollen data, is the long-term decline of Pinus PAR values. A particularly distinct decline takes place at SZ 2 around 5,000 cal BP, when Pinus PAR decreases from 1,500 to about 1,000 grains cm-2 year-1 (Fig. 4b). This is probably related to the general retreat of the pine treeline and the thinning of the pine forest in northern Fennoscandia that started ca. 5,000 cal years BP (Hyva¨rinen 1975). An additional cause for the regional decrease of pine may have been the accelerated spread of the aapa mires at the expense of formerly pinedominated mineral soils in the Lake Kipoja¨rvi area. In particular, this is suggested by the coeval rising trend of Sphagnum spores that begins ca. 5,000 cal years BP (Fig. 4). In the adjacent Kiposuo mire, peat accumulation had already started during the early Holocene (9,900–8,600 cal years BP) (Table 1), but relatively low Lake Kipoja¨rvi Pinus PAR values (Hyva¨rinen 1975; Seppa¨ and Hicks 2006) probably also reflect a regional-scale expansion of aapa mires (Korhola et al. 2010; Weckstro¨m et al. 2010). Macrophytes and bryozoa The macrofossil diagram was visually divided into three zones MAZ I–III. MAZ I (ca. 9,200–7,250 cal years BP) is characterised by the presence of a relatively diverse aquatic species community (Fig. 3). Such a rich pioneer assemblage is in good agreement with previous studies by Birks (2000) and Va¨liranta (2006). Several narrow-leaved Potamogeton species are present: P. berchtoldii/pusillus, P. filiformis and P. compressus, which occur together with Callitriche hamulata and C. hermaphroditica, Myriophyllum sp. and the macroalga Nitella. Statoblasts of the bryozoan Cristatella mucedo are abundant and a few Plumatella statoblasts are detected. Remains of limnotelmatic species are scarce and only seeds of Menyanthes trifoliata, vegetative remains of Equisetum sp., and occasional Carex/Cyperaceae remains are present. The early Holocene vegetation community represents a relatively nutrient-rich, shallow, littoral environment (Rintanen 1976). Potamogeton assemblages, in particular, suggest a warmer-thanpresent July temperature. The modern northern range limits of P. compressus and P. pusillus indicate that

345

the July temperature was at least 0.5–1°C higher than at present (Lampinen and Lahti 2007), which probably also resulted in a longer open-water season. The presence of Nitella is indicative of high water transparency (Coops 2002). MAZ II (ca. 7,250–4,750 cal years BP) represents a transition period during which the diversity and the amount of aquatic species decreased. Only two aquatic genera that were encountered in MAZ I occurred: unidentified remains of Potamogeton and Myriophyllum sp. Vegetative remains of Nymphaeaceae appear, accompanied by a Nuphar seed. Otherwise, only a small number of Cristatella mucedo statoblasts and Equisetum sp. remains are detected. Species assemblages indicate littoral conditions and, in particular, continuous presence of Equisetum remains is indicative of shallow water (\1 m) and/or more proximate location of the shoreline (Hannon and Gaillard 1997), compared to MAZ I. There is a decline in the amount of Cristatella mucedo statoblasts. This might be related to a general decrease in nutrient status and a decrease in the abundance of aquatic macrophytes (Økland and Økland 2000). The aquatic vegetation record shows a significant decrease in species richness and a gradual replacement of submersed Potamogeton communities by oligotrophic, floating-leaved Nymphaeaceae species. Such a development suggests a change in trophic status as mesotrophic lakes have been shown to support more diverse aquatic plant species communities and better-developed belts of submersed and floating-leaved plants than oligotrophic lakes (Rørslett 1991; Hannon and Gaillard 1997). By the end of MAZ II a complete aquatic species turnover has taken place. MAZ III (ca. 4,750 cal years BP to present) is characterized by a scarcity of macroscopic plant remains. With the exception of Nymphaeaceae, no other remains of aquatic plant species are present. The Cristatella mucedo statoblast record is almost continuous, even though the numbers remain low. After ca. 4,250 cal years BP Equisetum vegetative remains are no longer present and only individual seeds of Carex are encountered. Cladocera In total, 34 species of Cladocera were encountered, though many are only present in a few samples and in

123

346

J Paleolimnol (2011) 45:339–352

Ag e D cal ep y H th ear o c BP Bo lobe m s d Pl mi ium n e Pl uro a lo gib e x n b Al uro us giro eru on xu trig s m a s u on tris qu n e ad cin llu ra at s D ng us is pa ul D ra ar ap lo is Al hni na on a l ro C a e ong stra hy x is ta do iqu pi ru a na s sp Al on ha G a er ra gu ic us bt tt a Si oleb ta m C oc eris am e t Si pt pha est da oc lu ud Al cr erc s sp ina r ia o y u Al no sta s r p. o p ll e Eu na sis ina ctir os bo int elo tri sm erm ng s in e ata a di a Eu ry Al ce on rcu a s af sp fin p is . Pa ra U lon na a La per pig t r Acton ura a ro a s la p Al e eti ten on ru fer s el s h a la a na rpa na e

Cladoceran taxa

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000 8500 9000

By t Al hot o re O na ph p r e Al hry ecta s lo o o D na xus ngu ngi a r l m Al phn usti gra a anu on ia ca ci s lis a pu Al ex le on ci x Le a sa p c R tod ost hy o at Si nco ra k a m ta i n ps lo d on na ti 1- fa C D lc AZ at a

Lake Kipojärvi

III 60

120

II 180

240

I

300 0 5 0 10 0 5 0 10 0 10 20 30 0 10 0 10 0 10 0 10 20 0 5 0 10 0 5 0 10 0 10 0 5 0 5 0 10 20 30 0 10 0 10 20 300 10 0 10 0 5 0 10 0 10 20 30 40 0 5 0 10 0 5 0 5 0 5 0 10 0 5 0 5 0 10 0,76

0,84

Analyst Susanna Siitonen

Fig. 5 Percentage Cladoceran diagram and calculated diversity index from Lake Kipoja¨rvi. The hollow silhouettes show a 59 exaggeration of the percentage values

low numbers. The Cladocera record was divided into three assemblage zones, CAZ I–III (Fig. 5). CAZ I (ca. 9,300–7,300 cal years BP) has the highest number of species (30) and diversity. The diverse assemblage is characterised by abundant Alona quadrangularis, Chydorus sphaericus, and small Alona spp. together with the presence of Pleuroxus uncinatus and Graptoleberis testudinaria. Bosmina longirostris is present in this zone only, and the few occurrences of Alonella exigua are almost exclusively in CAZ I. This pioneering species composition resembles the one detected in Lake Njargajavri, eastern Finnish Lapland (SarmajaKorjonen et al. 2006). The early Holocene species assemblages indicate mesotrophic conditions and abundant available habitats, probably provided by aquatic vegetation (Brodersen et al. 1998; Dodson et al. 2000; Sarmaja-Korjonen et al. 2006; Bjerring et al. 2009). The onset of CAZ II at ca. 7,300 cal years BP is characterised by a decline in diversity and in the number of taxa, from 30 to 24. A. affinis and A. nana increase towards the present and eventually become the dominant species, together with Eubosmina spp. P. uncinatus, which thrives in relatively nutrient-rich waters, disappears completely and macrophyte-associated G. testudinaria occurs only occasionally, in

123

low numbers. These changes reflect a decline in the trophic state and loss of habitat. From ca. 4,500 cal years BP (CAZ III), the proportion of planktonic Eubosmina spp. decreases steadily towards the present day, and A. nana becomes the dominant species. The relative abundances of littoral chydorids Rhyncotalona falcata and A. excisa also increase. The species shift implies further oligotrophication, a decline in pH and temperature, and because the Chydoridae taxa become more dominant, possible expansion of the littoral area (Sandoy and Nilssen 1986; Brodersen et al. 1998; De Eyto et al. 2003; Bjerring et al. 2009). Although the number of species (25) is comparable to CAZ II, the late Holocene assemblages are the least diverse and are dominated by few species. In general, the subfossil cladocera assemblage in Lake Kipoja¨rvi is exceptionally species-rich (34 taxa) in comparison to other cladocera stratigraphies derived from Finnish Lapland, which contain *20 taxa (Sarmaja-Korjonen and Hyva¨rinen 1999; Sarmaja-Korjonen et al. 2006). Zooplankton species richness usually shows unimodal response to productivity (Dodson et al. 2000). In Lake Kipoja¨rvi, the species richness was highest during the early Holocene, which indicates the occurrence of mesotrophic conditions and oligotrophication thereafter.

J Paleolimnol (2011) 45:339–352

347

Fig. 6 A simplified diatom percentage diagram and the diatom-inferred quantitative pH reconstruction of Lake Kipoja¨rvi. Only species with a maximum abundance [5% are

shown. Diatom-inferred pH and TOC reconstructions are shown with a LOESS (locally weighted scatterplot smoothing) with a span of 0.2

Diatoms

Lange-Bertalot 1986–1991). Planktonic assemblages usually reflect periods of higher water levels. The aquatic plant community, however, clearly indicates littoral habitat conditions at the coring site. A drop in the abundance of Fragilaria taxa defines the start of SZ 3 ca. 8,200 cal years BP. The change in the diatom assemblage might suggest a shift towards a more stable environment, but also a decline in habitat diversity because Fragilaria also thrive in epiphytic habitats (Sayer et al. 1999). Evidence for a loss of habitat availability is supported by a decrease in aquatic macrophytes (Fig. 3). The beginning of DAZ II at ca. 7,500 cal years BP is characterised by a shift from the dominance of Fragilaria to the dominance of a more diverse, circumneutral, benthic diatom flora comprised of Achnanthes, Brachysira, Cymbella, Navicula, Pinnularia and Stauroneis. This switch may indicate a gradual shift to more stable environmental conditions, but may also indicate nutrient deficiency because benthic species are less nutrient-demanding. The second significant change in diatom assemblages occurs at the beginning of DAZ III, ca. 5,000 cal years BP. There is a shift in the diatom community, from species favoring moderate to rich electrolyte conditions (e.g. Amphora libyca, Denticula kuetzingii, Navicula viridula v. vulpina), to species thriving in oligotrophic, low-electrolyte,

In total, 155 diatom taxa were recorded. The diatom diagram (Fig. 6) was divided into three distinct assemblage zones, DAZ I–III, and into three additional sub-zones, SZ 1–3. DAZ I dates to ca. 9,200–7,500 cal years BP and is sub-divided into sub-zones SZ 1, 2 and 3. The most striking feature in the diatom stratigraphy occurs during SZ 1, in which diatoms are found only in the bottommost sample. The dominance of genera such as Brachysira, Cymbella, Frustulia, Navicula and Pinnularia, and the almost total absence of Fragilaria, including the revised genera, supports the idea that the sediment that accumulated immediately after deglaciation is missing from the stratigraphy (Seppa¨ and Weckstro¨m 1999). The diatom taxa in SZ 1 reflect a relatively stable habitat. The beginning of SZ 2 ca. 8,700 cal years BP is characterised by the dominance of benthic Fragilaria species that often form the pioneering assemblage in lakes. These species are often associated with high environmental instability and are known to tolerate broad environmental gradients (Smol 1988; Denys 1990). SZ 2 is also characterized by the maximum occurrence of the planktonic species Cyclotella radiosa and Cyclotella distinguenda. In particular, C. radiosa is also associated with relatively eutrophic waters (Krammer and

123

348

J Paleolimnol (2011) 45:339–352

and even dystrophic conditions (e.g. Brachysira sp., Frustulia rhomboides v. saxonica, Navicula subtilissima). This roughly coincides with the increase in the amount of Sphagnum spores (Fig. 4) and reflects a change in regional environmental conditions to cool moist climate with expansion of peatlands. The diatom-inferred pH stays around 7.4 between ca. 9,200 and 6,000 cal years BP, after which it gradually decreases to 7.1 around 1,000 cal years BP (Fig. 6). The high correlation (0.95) between DCA1axis scores (not shown) and pH reconstruction suggests that pH (or variables with linear correlation to pH) is the main environmental variable controlling the diatom species composition in Lake Kipoja¨rvi. Variation in the amount of TOC, which is an indicator of dissolved colour, was relatively subtle, as the reconstructed value varies between 5.5 at ca. 7,000 cal years BP, to 7.8 mg l-1 at ca. 3,300 cal years BP (Fig. 6). The amount is lowest during the early middle Holocene, between ca. 8,700 and 6,500 cal years BP, after which the amount gradually increases. After ca. 5,000 cal years BP the TOC value remains relatively stable. During the Holocene, variations in pH and TOC are probably linked. pH decreases steadily and the

amount of TOC continues to increase until the late Holocene. Natural, long-term acidification is a common feature in high-latitude lakes resulting from soilforming processes and vegetation development in the catchment, especially replacement of birch-dominated forest by pine forest. This leads to an accumulation of acid humus, which is the main source of organic carbon on mineral soils that is transported to aquatic ecosystems, Thus, in many cases, long-term acidification and an increase in TOC are parallel processes (Korsman and Segerstro¨m 1998).

Fig. 7 Summary chart, indicating the main aquatic ecosystem and regional environmental changes in and around Lake Kipoja¨rvi during the Holocene. Additionally, literature-based general northern Fennoscandian water-level and temperature

reconstructions are shown. The zone limits of pollen, diatoms and Cladocera are based on the program zone and associated broken stick model, whereas the zones for plant macrofossils, LOI and C:N data were determined visually

123

Discussion All aquatic biological variables in the sediment core from Lake Kipoja¨rvi reveal three local paleoecological phases (Fig. 7): 1) the early Holocene ca. 9,200–7,500 cal years BP, 2) the middle Holocene ca. 7,500–4,750 cal years BP, and 3) the late Holocene ca. 4,750 cal years BP to present. The main focus of the discussion is on the prominent early to middle Holocene community changes and the inferred limnological evolution as set against contemporary environmental and climate conditions.

J Paleolimnol (2011) 45:339–352

After deglaciation, the relatively warm oceanic climate (Seppa¨ and Hammarlund 2000) was associated with high groundwater levels and lake stages in Finnish Lapland (Korhola et al. 2005) and Lake Kipoja¨rvi was initiated. Peat accumulation commenced on the mineral soil in the adjacent Kiposuo mire, with a basal age between 9,900 and 8,600 cal years BP (Table 1). The early Holocene diverse aquatic ecosystem suggests that (a) temperature was *1°C warmer than present day July temperature, (b) mesotrophic conditions prevailed, and (c) there was a littoral environment at the coring site. A warming early Holocene climate and high effective moisture probably increased runoff and transport of dissolved organic matter and nutrients from terrestrial to aquatic ecosystems (Qualls and Richardson 2003; Mattsson 2010) where they provided a nutrient pool for aquatic organisms (Qualls and Richardson 2003; Va¨ha¨talo et al. 2003). The period that commenced ca. 7,500 cal years BP is characterized by major changes in all aquatic communities, especially aquatic macrophytes. A change in Cladoceran species richness conforms to this development. The early middle Holocene reduction in Cladocera species richness and diversity in Lake Kipoja¨rvi is in contradiction to earlier boreal studies that reported a species-rich middle Holocene (Korhola 1990; Sarmaja-Korjonen and Hyva¨rinen 1999) linked to higher mid-Holocene temperatures. We suggest that in Lake Kipoja¨rvi, changes in lake nutrient status and consequent changes in aquatic vegetation were the main environmental variables that controlled the structure of cladoceran assemblages. The importance of macrophytes in structuring modern and subfossil zooplankton communities and the overall function of shallow lakes is well established (Carpenter and Lodge 1986; Jeppesen et al. 1997; Davidson et al. 2007). Because the aquatic macrophyte community structure seems to have played an important role in regulating other aquatic ecosystem components, it is important to understand the driving mechanisms behind the vegetation change. The most important factors controlling aquatic plant species composition are summer temperature, water colour (i.e. transparency), pH, water depth and nutrient status (Wetzel 2001). The middle Holocene (ca. 8,000–5,000 cal years BP) represents the warmest phase of the Holocene in Fennoscandia (Seppa¨ and Birks 2001), and the length

349

of the growing season was at least as long then as during the early Holocene. Such climate conditions should have been suitable to support aquatic macrophyte communities comparable to those of the early Holocene. Consequently, a change in summer temperatures was probably not a significant factor in early middle Holocene aquatic plant species composition change. Establishment of conifer forest and paludification would have resulted in dystrophication, i.e. darker water colour, over the long term (Forsberg 1992; Seppa¨ and Weckstro¨m 1999) as these catchment components are the main source of humic substances (Birks et al. 2000; Lotter and Birks 2003; Mattsson 2010). Diatom-based TOC reconstruction showed relatively low values (5.5–7.8 mg l-1) throughout the lake history, which would suggest that variation in TOC had little impact on lake water transparency in the littoral zone. Thus, water colour probably did not have any effect on the early middle Holocene aquatic plant species composition. Similarly, the diatom assemblages suggest that pH did not reach such high or low values that it would have had an impact on aquatic macrophytes. Increasing summer temperatures probably caused local and regional hydrological changes that may have had a major influence on the aquatic ecosystems. In northern Fennoscandia, increased evapotranspiration and decreased precipitation (Seppa¨ and Hammarlund 2000) led to a general lowering of lake levels during the middle Holocene (SarmajaKorjonen and Hyva¨rinen 1999; Korhola et al. 2005; Va¨liranta et al. 2005). The TOC reconstruction indicates a low middle Holocene water depth in Lake Kipoja¨rvi also. The amount of TOC was probably at its lowest during the early middle Holocene as a consequence of decreased precipitation and increased residence time, which resulted in a decrease in the TOC concentration via sedimentation and bacterial oxidation (Hessen et al. 1997). Macrophyte and cladoceran records suggest that in Lake Kipoja¨rvi, littoral conditions prevailed at the coring site throughout the sedimentation history. Based on multiple lines of evidence, we conclude that changes in water level were not the reason for the change in aquatic vegetation species composition. All biological proxies indicate that trophic status had already begun to decrease in the early middle Holocene. Water chemistry and physical properties of

123

350

northern lakes are primarily linked to catchment processes, and the input of nutrients from the catchment is regulated by vegetation, soil type, climate and, especially, hydrology (Forsberg 1992; Pienitz et al. 1999; Tranvik and Jansson 2002). In general, decreased precipitation results in reduced input of nutrients and carbon to the aquatic environment. Accordingly, during the dry middle Holocene, transportation of these compounds from the catchment to the lake probably slowed. Reduced input from the catchment is supported by water chemistry reconstructions. During the early Holocene vegetation changed from birch to pine forest but the consequent accumulation of acidic humus seems to have had only a minarticle can be found in an online versionor effect on the lake water chemistry, as both pH and TOC show only subtle changes. The impact of low effective moisture was probably more pronounced due to permeable and nutrient-poor esker environment. Thus, we propose that hydrologydriven nutrient deficiency caused the detected changes in the aquatic ecosystem, i.e. the reduction in the amount of aquatic macrophytes and a change in species composition. Decreased habitat availability and structure, in turn affected diatom and cladoceran species composition. Fennoscandian late Holocene climate reconstructions have indicated a change towards colder and more humid conditions (Seppa¨ and Birks 2001; Seppa¨ and Hammarlund 2000). Higher precipitation probably resulted in increased allochthonous minerogenic and dissolved carbon input to Lake Kipoja¨rvi. This is reflected as a decrease in LOI values and an increase in TOC. Declining sediment C:N ratios imply a further reduction in the amount of aquatic vascular plants. A cool climate probably reduced in-lake primary production and slowed biogeochemical catchment processes. Consequently, the trophic status remained oligotrophic despite increased material transport from the catchment. Another factor that may have contributed to sustained oligotrophy was the shorter water retention time as a result of formation of an outlet ca. 5,900 cal years BP, when the water level in the lake rose. General late Holocene climate cooling resulted in a shorter growing (open water) season and over time probably began to favour more cool-tolerant macrophyte species (Va¨liranta 2006).

123

J Paleolimnol (2011) 45:339–352

Conclusions All aquatic biological variables in the Lake Kipoja¨rvi sediment indicate a three-phase limnological evolution towards oligotrophy that commenced in the early middle Holocene, ca. 7,500 years ago. Ecosystem responses began with a regional-scale vegetation succession, i.e. a change from birch to pine forest ca. 8,700 cal years BP. This development, however, had only minor impact on pH and TOC of the water column. A change in aquatic macrophytes ca. 7,250 cal years BP was accompanied by a decrease in cladoceran diversity and changes in diatom species composition. This process was likely regulated by decreased nutrient input from the catchment due to dry middle Holocene climate conditions, rather than a change in temperature, water colour or lowering of the water level. Several processes may have led to decreased nutrient status by the early middle Holocene, which is earlier than reported by previous Fennoscandian studies. Dense arboreal vegetation and surrounding mires in the catchment effectively retained nutrients and reduced nutrient availability. Nutrient-poor esker terrain rapidly lost its initial nutrient stores and a permeable soil enhanced the impact of a dry climate. Drier climate also resulted in reduced nutrient leaching and there was a decline in the delivery of nutrients to the lake. By 4,750 cal years BP, Lake Kipoja¨rvi had changed from its initial relatively nutrient- and species-rich state to a typical oligotrophic, humic boreal lake ecosystem. Acknowledgments Funding was provided to MV and JW by the REBECCA-project, supported by the Helsinki University Environmental Research Centre (HERC) and the Academy of Finland. We gratefully acknowledge funding to SS from Arctic Doctoral Programme, Arctic Centre, University of Lapland. We warmly thank Mirjam Orvomaa and Virpi Kuutti for help with fieldwork and analyses. We are grateful to Ossi Aikio and many other local people for their helpful and cooperative attitude towards our fieldwork in Kaamanen.

References Barnekow L (2000) Holocene regional and local vegetation history and lake-level changes in the Tornetra¨sk area, northern Sweden. J Paleolimnol 23:399–420 Bennett K (1996) Determination of the number of zones in a biostratigraphical sequence. New Phytol 132:155–170

J Paleolimnol (2011) 45:339–352 Birks HH (2000) Aquatic macrophyte vegetation development in Kra˚kenes Lake, western Norway, during the late-glacial and early-holocene. J Paleolimnol 23:7–19 Birks HJB, Gordon AD (1985) Numerical methods in quaternary pollen analysis. Academic Press, London Birks HH, Battarbee RW, Birks HJB (2000) The development of the aquatic ecosystem at Kra˚kenes Lake, western Norway, during the late-glacial and early-holocene—a synthesis. J Paleolimnol 23:91–114 Bjerring R, Becares E, Declerck S, Gross EM, Hansson L-A, Kairesalo T, Nyka¨nen M, Halkiewicz A, Kornijow R, Conde-Porcuna JM, Seferlis M, No¨ges T, Moss B, Amsinck SL, Odgaard BV, Jeppesen E (2009) Subfossil Cladocera in relation to contemporary environmental variables in 54 Pan-European lakes. Freshwat Biol 54:2401–2417 Brodersen KP, Whiteside MC, Lindegaard C (1998) Reconstruction of trophic state in Danish lakes using subfossil chydorid (Cladocera) assemblages. Can J Fish Aquat Sci 55:1093–1103 Carpenter SR, Lodge DM (1986) Effects of submersed macrophytes on ecosystem processes. Aquat Bot 26:341–370 Coops H (2002) Ecology of charophytes: an introduction. Aquat Bot 72:205–208 Davidson T, Sayer C, Perrow M, Bramm M, Jeppesen E (2007) Are the controls of species composition similar for contemporary and sub-fossil cladoceran assemblages? A study of 39 shallow lakes of contrasting trophic status. J Paleolimnol 38:117–134 de Eyto E, Irvine K, Garcia-Criado F, Gyllstro¨m M, Jeppesen E, Kornijow R, Miracle MR, Nyka¨nen M, Bareiss C, Cerbin S, Salujo˜e J, Franken R, Stephens D, Moss B (2003) The distribution of chydorids (Branchiopoda, Anomopoda) in European shallow lakes and its application to ecological quality monitoring. Arch Hydrobiol 156:181–202 Denys L (1990) Fragilaria blooms in the holocene of the western coastal plain of Belgia. In: Simola H (ed) Proceedings of the tenth international diatom symposium, Joensuu, Finland, 28th August–2nd September 1988. Koeltz Scientific Books, Koenigstein, pp 397–406 Dodson SI, Arnott SE, Cottingham KL (2000) The relationship in lake communities between productivity and species richness. Ecology 81:2662–2679 Fægri K, Iversen J (1989) Textbook of pollen analysis. Wiley, Chichester Forsberg C (1992) Will an increased greenhouse impact in Fennoscandia give rise to more humic and colored lakes? Hydrobiologia 229:51–58 Hannon GE, Gaillard M-J (1997) The plant-macrofossil record of past lake-level changes. J Paleolimnol 18:15–28 Hessen DO, Gjessing EG, Knulst J, Fjeld E (1997) TOC fluctuations in a humic lake as related to catchment acidification, season and climate. Biogeochemistry 36:139–151 Hyva¨rinen H (1975) Absolute and relative pollen diagrams from northernmost Fennoscandia. Fennia 142:1–23 Jeppesen E, Jensen JP, Søndergaard M, Lauridsen T, Pedersen LJ, Jensen L (1997) Top-down control in freshwater lakes: the role of nutrient state, submerged macrophytes and water depth. Hydrobiologia 342–343:151–164

351 Juggins S (2001) The European diatom database, user guide, version 1.0 Juggins S (2003) C2 user guide. Software for ecological and paleoecological data analysis and visualization. University of Newcastle, Newcastle Upon Tyne Korhola A (1990) Paleolimnology and hydroseral development of the Kotasuo bog, Southern Finland, with special reference to the Cladocera. Ann Acad Sci Fenn A III 155:1–40 Korhola A, Tikkanen M, Weckstro¨m J (2005) Quantification of the holocene lake-level changes in Finnish Lapland by means of a cladocera-lake depth transfer model. J Paleolimnol 34:175–190 Korhola A, Seppa¨ H, Ruppel M, Va¨liranta M, Virtanen T, Weckstro¨m J (2010) The importance of northern peatland expansion to the late-holocene rise of atmospheric methane. Quat Sci Rev 29:611–617 Korsman T, Segerstro¨m U (1998) Forest fire and lake-water acidity in a northern Swedish boreal area: holocene changes in lake-water quality at Makkassjo¨n. J Ecol 86:113–124 Krammer K, Lange-Bertalot H (1986–1991) Bacillariophyceae. In: Ettl H, Gerloff J, Heynig H, Mollenhauer D (eds) Su¨ßwasserflora von Mitteleuropa, vol 2 (1–4). Gustav Fischer Verlag, Stuttgart/Jena Lampinen R, Lahti T (2007) Kasviatlas 2006. Botanical Museum, Finnish Museum of Natural History. Helsinki. [Digital atlas of vascular plants in Finland at: http:// www.luomus.fi/kasviatlas] Lotter AF, Birks HJB (2003) The holocene paleolimnology of Sa¨gistalsee and its environmental history—a synthesis. J Paleolimnol 30:333–342 Lotter AF, Juggins S (1991) PLOPROF, TRAN and ZONE. Programs for plotting, editing and zoning of pollen and diatom data. INQUA Commission for the study of the Holocene, Working Group on Data Handling Methods, Newsletter 6 Mattsson T (2010) Export of organic matter, sulfate and base cations from boreal headwater catchments downstream to the coast: impacts of land use and climate. Monograph Boreal Environ Res 36:1–45 Økland KA, Økland J (2000) Freshwater bryozoans (Bryozoa) of Norway: distribution and ecology of Cristatella mucedo and Paludicella articulate. Hydrobiologia 421:1–24 Pienitz R, Smol JP, MacDonald GM (1999) Paleolimnological reconstruction of holocene climatic trends from two boreal treeline lakes, Northwest territories Canada. Arct Antarc Alp Res 31:82–93 Qualls RG, Richardson CJ (2003) Factors controlling concentration, export, and decomposition of dissolved organic nutrients in the Everglades of Florida. Biogeochemistry 62:197–229 Rintanen T (1976) Lake studies in eastern Finnish Lapland. I. Aquatic flora: Phanerogams and Charales. Ann Bot Fennici 13:137–148 Rørslett B (1991) Principal determinants of aquatic macrophytes richness in northern European lakes. Aquat Bot 39:173–193 Rose´n P, Segerstro¨m U, Eriksson L, Renberg I, Birks HJB (2001) Holocene climatic change reconstructed from diatoms, chironomids, pollen and near-infrared spectroscopy at an

123

352 alpine lake (Sjuodjijaure) in northern Sweden. Holocene 11:551–562 Sandoy S, Nilssen JP (1986) A geographical survey of littoral crustacea in Norway and their use in paleolimnology. Hydrobiologia 143:277–286 Sarmaja-Korjonen K, Hyva¨rinen H (1999) Cladoceran and diatom stratigraphy of calcerous lake sediments from Kuusamo, NE Finland. Indications of holocene lake level changes. Fennia 177:55–70 Sarmaja-Korjonen K, Nyman M, Kultti S, Va¨liranta M (2006) Palaeolimnological development of Lake Njargajavri, Northern Finnish Lapland, in a changing holocene climate and environment. J Paleolimnol 34:203–215 Sayer C, Roberts N, Sadler J, David C, Wade PM (1999) Biodiversity changes in a shallow lake ecosystem: a multi-proxy palaeolimnological analysis. J Biogeogr 26:97–114 Seppa¨ H (1998) Postglacial trends in palynological richness in the northern Fennoscandian tree-line area and their ecological interpretation. Holocene 8:43–53 Seppa¨ H, Birks HJB (2001) July mean temperature and annual precipitation trends during the holocene in the Fennoscandian tree-line area: pollen-base climate reconstruction. Holocene 11:527–539 Seppa¨ H, Hammarlund D (2000) Pollen-stratigraphical evidence of holocene hydrological change in northern Fennoscandia supported by independent isotopic data. J Paleolimnol 24:69–79 Seppa¨ H, Hicks S (2006) Integration of modern and past pollen accumulation rate (PAR) records across the arctic treeline: a method for more precise vegetation reconstructions. Quat Sci Rev 25:1501–1516 Seppa¨ H, Weckstro¨m J (1999) Holocene vegetational and limnological changes in the Fennoscandian tree-line area as documented by pollen and diatom records from Lake Tsuolbmajavri, Finland. E´coscience 6:621–635 Simpson EH (1949) Measurement of diversity. Nature 163:688 Smol JP (1988) Paleoclimate proxy data from freshwater arctic diatoms. Verh Int Ver Limnol 23:837–844

123

J Paleolimnol (2011) 45:339–352 Stockmarr J (1971) Tablets with spores used in absolute pollen analysis. Pollen Spores 13:615–621 Stuiver M, Reimer PJ (1993) Extended 14C data base and revised CALIB 3.0 14C age calibration program. Radiocarbon 35:215–230 Szeroczyn´ska K, Sarmaja-Korjonen K (2007) Atlas of subfossil Cladocera from Central and Northern Europe. Friends of the Lower Vistula Society, S´wiecie Tranvik LJ, Jansson M (2002) Climate change (Communication arising): terrestrial export of organic carbon. Nature 415:861–862 Va¨ha¨talo AV, Salonen K, Munster U, Ja¨rvinen M, Wetzel RG (2003) Photochemical transformation of allochthonous organic matter provides bioavailable nutrient in a humic lake. Arch Hydrobiol 156:287–314 Va¨liranta M (2006) Long-term changes in aquatic plant species composition in North-eastern European Russia and Finnish Lapland, as evidenced by plant macrofossil analysis. Aquat Bot 85:224–232 Va¨liranta M, Kultti S, Nyman M, Sarmaja-Korjonen K (2005) Holocene development of aquatic vegetation in a shallow Lake Njargajavri, Finnish Lapland with evidence of water level fluctuations and drying. J Paleolim 34:203–215 van Donk E, van de Bund WJ (2002) Impact of submerged macrophytes including charophytes on phyto and zooplankton communities: allelopathy versus other mechanisms. Aquat Bot 72:261–274 Velle G, Larsen J, Eide W, Peglar S, Birks HJB (2005) Holocene environmental history and climate of Ra˚ta˚sjøen, a low-alpine lake in south-central Norway. J Paleolimnol 33:129–153 Weckstro¨m J, Seppa¨ H, Korhola A (2010) Climatic influence on peatland formation and lateral expansion dynamics in subarctic Fennoscandia. Boreas 39:761–769 Weckstro¨m J (2001) Assessment of diatoms as markers of environmental change in northern Fennoscandia. Unpublished Ph.D. dissertation, University of Helsinki, Helsinki Wetzel R (2001) Limnology. Academic press, San Diego