8/13/2019 Molar2

1/6

1Journal of Paleolimnology 28:16, 2002. 2002Kluwer Academic Publishers. Printed in the Netherlands.

Mountain Lake Research

MOLARClimate variability and ecosystem dynamics of remotealpine and arctic lakes: the MOLAR project

Richard W. Battarbee1, Roy Thompson2, Jordi Catalan3, John-Arvid Grytnes4& H.J.B. Birks,1,41Environmental Change Research Centre, University College London, 26 Bedford Way, London WC1H 0AP, UK

(E-mail: rbattarb@geography.ucl.ac.uk)2Department of Geology and Geophysics, University of Edinburgh, West Mains Road Edinburgh, EH9 3JW, UK3Department dEcologia, Universitat de Barcelona, Diagonal 645, E- 08028 Barcelona, Spain4Botanical Institute, University of Bergen, Allegt. 14, N-5007 Bergen, Norway

Received 5 June 2001; accepted 9 January 2002

Key words: remote mountain lakes, palaeolimnology, climate change, instrumental records

Abstract

This paper introduces the results of an integrated project designed to compare high resolution analysis of proxyrecords of climate change in the sediments of seven mountain lakes across Europe with reconstructed instrumentalrecords of climate change over the last 200 years. Palaeolimnological methods used include radiometric dating (210Pb,137Cs), mineral magnetics, dry weight, loss-on-ignition, carbon, nitrogen, sulphur, pigments, diatoms, chrysophytecysts, cladocera and chironomids. Changes in fossil assemblages were summarised using principal components analysis.The stratigraphic data were compared with the instrumental record using linear regression techniques. The dated sedi-ment records for each proxy from each site were treated as the response variables and the various attributes of the

instrumental climate record as the predictor variables. The predictor variables were generated for each site for theperiod 1781 to 1997 using temperature reconstructions based on meteorological records. To harmonise the climaticpredictors and the response variables, the climatic variables were smoothed along time with a LOESS regression.The results of the various analyses at the seven sites are presented in the following papers. A synthesis of the projectand the relative performance of the different proxy methods are discussed in the final paper.

Introduction

There is increasing recognition that arctic and alpineaquatic ecosystems are being influenced by climatechange (Douglas et al., 1994; Sorvari & Korhola, 1998).As surface water temperatures tend to vary closely withair temperature, especially during spring and summermonths (Livingstone & Lotter, 1998), one of the effectsof global warming is an increase in lake-water tempera-ture. Changes in wind speed, wind direction, rain andsnowfall are also associated with climate change and,together with temperature, these changes affect the tim-

ing and intensity of thermal stratification, ice-cover,turbidity and light penetration (e.g., Catalan & Camero,1990). Chemical and biological changes ensue. Ifchanges in weather patterns are sustained over longerperiods, changes in water column productivity and inthe species composition of plant and animal pop-ulations can be expected, both by direct (e.g., tempera-ture) and indirect (e.g., pH, ice-cover) effects.

Equally, changes in weather patterns will also causecatchment changes to occur that will additionally influ-ence lake-water conditions. In mountain lakes increasedtemperatures lead to a reduction in catchment snow andice cover that cause changes in hydrology, speed upweathering processes and may accelerate soil erosion.

These effects of climate change are potentially re-corded in lake sediments and lake sediments can thusbe used as a means of reconstructing past climate overlong time-scales. However, the usefulness of this ap-

This is the first of 11 papers published in this special issue on thepalaeolimnology of remote mountain lakes in Europe resulting fromthe MOLAR project funded by the European Union. The guest edi-tor was Richard W. Battarbee.

8/13/2019 Molar2

2/6

2

proach depends on the sensitivity and accuracy of thevarious proxy methods that are used for climate recon-

struction and the extent to which the climate signal inthe sediment record is obscured by noise from otherinfluences (e.g., associated with human disturbance).

To maximise the climate signal we have selected sitesfor study that are as pristine as possible, situated abovethe timber-line, without apparent catchment disturbanceand largely unaffected by air pollution. To avoid prob-lems of acidification the sites are either located in areasof low acid deposition or in areas with calcareous bed-rock. At each site we have data from automatic weatherstations, modelled climate data spanning the last 200years, water column data and detailed multi-proxy analy-ses of 210Pb-dated sediment cores. We selected 200 years

as a key period for data comparison as this is the periodof time for which reliable climate records based on in-strumental data are available and also the time periodover which a reasonably reliable sediment chronologycan be established from 210Pb dating.

In this volume we describe the results of this projectand evaluate the sensitivity and usefulness of the differ-ent proxy-climate methods used. The work was carriedout under the auspices of the EU-funded project MO-LAR: Measuring and modelling the dynamic responseof remote mountain lake ecosystems to environmentalchange:A programme of Mountain Lake Research

(Battarbee et al., 2001). The papers are based on theresults from Work Package 3 of the MOLAR project thatwas specifically concerned with climate change. Thefollowing papers deal with the methods used to estab-lish the instrumental climate time-series for each site,with the limnological characteristics of the lakes andwith a comparison between the instrumental record ofclimate change spanning the last 200 years and the proxyrecords of climate change, on a site by site basis. Addi-tional data from these sites and from the wider MOLARproject has already been published (Strakrabov et al.,1999; Lami et al., 2000).

Sites

The lakes in this study share the same basic limno-logical characteristics. They are oligotrophic, typicallydimictic and have a winter ice-cover and a summerthermocline. They are all situated above or beyond theregional tree-line, and they have poorly vegetated catch-ments.

Most of the sites were included in the AL:PE.2 study(Wathne et al., 1997), and belong to the group of sites

identified in that study to be least influenced by aciddeposition. In the MOLAR project a number of new

sites were added to provide the study with a widergeographical and chemical range. These are locatedin Finland, Austria, Switzerland and Slovenia.

The sites are shown in Figure 1 and their main char-acteristics are listed in Table 1. They include sites inthe Arctic and northern Latitudes (vre Nedalsvatnin Norway, Saanajrvi in Finland), sites in the Alps(Gossenkllesee in Austria, Hagelseewli in Switzer-land, Jezero v Ledvici in Slovenia), sites in Spain(Estany Red in the Pyrenees, Cimera in the GredosMountains), and a site in the Slovakian Tatra Mountains(Nin Terianske).

Approach

In order to assess how well remote mountain lake sedi-ments record climate variability it is necessary not onlyto carry out a high resolution multi-proxy study of well-dated sediments from a range of sites, but also to es-tablish, if possible, the mechanisms that link climate

aanajrviSaanajrvi

vre Neadlsvatnvre Neadlsvatn

Nizn Terianske PlesoJezero v Ledvici

GossenklleseeHagelseewli

Estany Red

Nizn Terianske Pleso

Jezero v Ledvici

Gossenkllesee

Hagelseewli

Estany Red

Figure 1. Map of Europe showing the location of the seven moun-tain lales studied in this project.

8/13/2019 Molar2

3/6

3

change to the sediment record. Although there aremajor logistical difficulties working in remote moun-

tain lakes, we have approached this challenge by at-tempting to (i) generate 200 year-long meteorologicalrecords for each site; (ii) verify these records by com-parison with data from on-site automatic weather sta-tions; (iii) relate climate records to water columncharacteristics, (iv) relate sediment records to climaterecords using regression analysis, and (v) interpret thesediment records in terms of water column and mud-water interface processes.

Establishing weather and climate records on site

Establishing a time series of climate data that can be usedfor comparisons with sediment cores over the last 200years is a three-step process: constructing homogenousair-temperature series from lowland historical instru-mental records; transferring these records to the moun-tain sites taking into account differences in elevation andlocal microclimate; and testing the accuracy of the mod-elled data for mountain stations using established highelevation meteorological stations and data from on-siteautomatic weather stations (Agust-Panareda et al., 2000,Agust-Panareda & Thompson, this issue).

Europe is particularly fortunate in having the long-est and densest network of historical climate recordsof any part of the world. A unique body of climate datahas been assembled over the last five centuries andmuch of the data, the earliest reaching back to Janu-ary 1525 AD, have been converted into machine-read-able form. In many cases these data have been unified,compared, corrected and standardised in order to con-struct long-term homogenous air temperature time se-ries (e.g., Jones et al.,1986), and extensive historicalclimate databases have been created e.g., the GlobalHistorical Climate Network (Peterson et al., 1998) andthe North Atlantic Climate Data set (Frich et al., 1995).

Using these data and data from a more limited range

of upland meteorological stations we have attemptedto reconstruct air temperatures from 1781 to the presentday for our remote alpine and arctic lake study sites inEurope. The results are presented by Agust-Panareda& Thompson (this issue).

Relating meteorological data to water column

characteristics

Although some proxy-methods are designed to recon-struct temperature directly from the sediment record

Table1.Lakelocation,morphologyandcatchmentfeatures

Gossenkllesee

JezerovLedvici

Hagelseewli

vreNedalsvatn

Saanajrvi

N.Terianskepleso

Estany

Red

Latitude

4713

N

4620

N

4640

N

6246

N

6903

N

4910

N

4238

N

Longitude

110E

1347

E

802

E

900

E

2052

E

2000

E

046

E

Altitude(ma.s.l)

2417

1830

2339

728

679

1941

2240

Mountainrange

Tyrolean

Julian

CentralSwiss

Caledonian

Northe

rn

Tatra

Pyrenees

Alps

Alps

Alps

Finland

Maximumdepth(m)

9.9

15

18

18

24

44.4

73

Meandepth(m)

4.7

5.7

8.3

4.0

5.1

18.4

32.3

Lakearea(ha)

1.7

2.37

2.37

50

69.9

4.83

24

Lakevolume(106m

3)

0.08

0.135

0.197

1.996

3.6*

0.891

7.750

Watershedarea(ha)

20

notdefined-

36

1600

461

114

155

Renewaltime(years)

0.2*

0.1

0.3*

0.07

1*

0.8

4

Watershedtolakearearatio

11.8

15.2

32.0

6.6

22.8

6.5

Mainlithology

Granite,gneiss,

Limestone

Limestone

Gneiss

Schist,

gneiss

Granite

Granod

iorite

amphibolite

andlim

estone

Soilcover(%)

20