VU Research Portal

Exploring early- and mid-Weichselian climate variability in Europe by applying chironomids as a proxy Engels, S.

2008

document version Publisher's PDF, also known as Version of record

Link to publication in VU Research Portal

citation for published version (APA) Engels, S. (2008). Exploring early- and mid-Weichselian climate variability in Europe by applying chironomids as a proxy.

General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal ?

Take down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

E-mail address: [email protected]

Download date: 05. Oct. 2021 Chapter 1

VRIJE UNIVERSITEIT

Exploring early- and mid-Weichselian climate variability in Europe by applying chironomids as a proxy

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad Doctor aan de Vrije Universiteit Amsterdam, op gezag van de rector magnificus prof.dr. L.M. Bouter, in het openbaar te verdedigen ten overstaan van de promotiecommissie van de faculteit der Aard- en Levenswetenschappen op vrijdag 7 maart 2008 om 15.45 uur in de aula van de universiteit, De Boelelaan 1105

door

Stefan Engels

geboren te Mill en St. Hubert

1 Chapter 1 promotor: prof.dr. J.F. Vandenberghe copromotoren: dr. S.J.P. Bohncke dr. O.M. Heiri

2 Chapter 1

Reading committee: dr. J.A.A. Bos dr. S.J.Brooks dr. K.F. Helmens dr. C. Kasse prof.dr. A.F. Lotter dr. B. van Geel

3 Chapter 1

This research was carried out at:

Vrije Universiteit Faculty of Earth and Life Sciences Department of Palaeoclimatology and Geomorphology De Boelelaan 1085 1081 HV Amsterdam The Netherlands

ISBN-13: 978-90-9022717-7

Exploring early- and mid-Weichselian climate variability in Europe by applying chironomids as a proxy

In Dutch: Analyse van klimaatsvariabiliteit tijdens het vroeg- en midden Weichselien, gebruik makende van chironomiden als klimaat-indicator

Cover photo: Channel connecting the Muddusjarvi river with one of the lakes sampled during the fieldwork in Kaamanen, 2006

This project was funded by the Council of Earth and Life Sciences of the Netherlands Organization for Scientific Research (grant-no 813.02.004).

4 Chapter 1

Exploring early- and mid-Weichselian climate variability in Europe by applying chironomids as a proxy

Stefan Engels

5 Chapter 1

Contents

Chapter 1: 9 Introduction / Summary Engels S

Chapter 2: 27 Chironomid-based palaeotemperature estimates for northeast Finland during Oxygen Isotope Stage 3 Engels S, Bohncke SJP, Bos JAA, Brooks SJ, Heiri O, Helmens KF Published online first for Journal of Paleolimnology DOI: 10.1007/s10933-007-9133-y

Chapter 3: 45 Present-day temperatures in northern Scandinavia during the Last Glaciation Helmens KF, Bos JAA, Engels S, Van Meerbeeck C, Bohncke SJP, Renssen H, Heiri O, Brooks SJ, Seppä H, Birks HJB, Wohlfarth B Geology 35 (11): 987-990

Chapter 4: 55 Rapid climatic events as recorded in Middle Weichselian thermokarst lake sediments Bohncke SJP, Bos JAA, Engels S, Heiri O, Kasse C In Press for Quaternary Science Reviews

6 Chapter 1

Chapter 5: 77 Intraregional variability in chironomid-inferred temperature estimates and the influence of river inundations on lacustrine chironomid assemblages Engels S, Bohncke SJP, Heiri O, Nyman M Published online first for Journal of Paleolimnology DOI 10.1007/s10933-007-9147-5

Chapter 6: 97 Environmental inferences and chironomid-based temperature reconstructions from fragmentary records of the Weichselian Early Glacial and Pleniglacial periods in the Niederlausitz area (eastern Germany) Engels S, Bohncke SJP, Bos JAA, Heiri O, Vandenberghe J, Wallinga J Accepted by Palaeogeography, Palaeoclimatology, Palaeoecology

Chapter 7: 119 The lacustrine archive of Oberwinkler Maar (Eifel, Germany): chironomid-based inferences of environmental changes during Oxygen Isotope Stage 3 Engels S, Bohncke SJP, Heiri O, Schaber K, Sirocko F Submitted to Boreas

Chapter 8: 135 Synthesis / Epilogue Engels S

Samenvatting 151 Acknowledgements 157

7 Chapter 1

8 Chapter 1

Chapter 1: Introduction / Summary

1.1 Climate change 1.1.1 Introduction During the last decade, there has been an increase in both the attention for and the awareness of global climate change. Predictions of future climate change as the result of increased human-induced levels of carbon dioxide and methane are alarming (IPCC 2007). However, even without the influence of mankind, earth’s climate has always been dynamic and the processes driving these changes are still acting, thus complicating the analysis of the effects of anthropogenic forcing on the climate system. To disentangle the natural variability in the climate system and the human- induced effects on the global climate, a critical analysis of climate change in the past may offer an understanding of the processes acting on the earth’s surface and driving the global climate system. However, the period for which there are historical records of changes in, for instance, precipitation or temperature is (in geological terms) relatively short. In most regions instrumental records do not start earlier than the 19th century. In order to reconstruct processes driving climate change on a longer timescale, we have to use indirect measurements of relevant parameters of the climate system, so called climate proxy-indicators. A wide range of proxies and techniques is available to study past changes in the climate system, and sediment, ice or even trees provide natural archives in which these proxies are preserved. The knowledge of both pattern and timing of climatic changes in the past is a prerequisite in order to understand the causes of changing climate at various time scales (Vandenberghe et al. 1998a). This study focuses on the last ice age (known in northwestern Europe as the Weichselian) during which abrupt climate changes have been reconstructed (see section 1.1.2) and applies chironomid analysis, a relatively new method, to infer past July air temperatures from fossil insect (chironomid) remains (section 1.2.3).

1.1.2 Palaeoclimatology The Quaternary (covering the last 2.6 million years of earth’s history (Gibbard et al. 2005)) is characterised by numerous repetitive oscillations in climate (e.g. Crowhurst 2002)). A periodicity of ~41 ka dominates in palaeoclimate records between 2.6 and 1 Ma ago, whereas a 100 ka cycle strongly characterizes climate change over the last 600 ka (Figure 1.1a). Stable oxygen isotope records from deep sea sediments (e.g. Crowhurst 2002), deuterium records from Antarctic ice cores (e.g. Epica community members 2004) and evidence on global sea level changes from e.g. fossil remains of

9

Chapter 1

LATE MIDDLE EARLY PLENIGLACIAL

LATE EARLY Chrono WEICHSELIAN

EEMIAN

stratigraphy HOLOCENE Age (ka BP) (ka Age .5 28 59 73 111 129 11 14.7 1 2 3 4 6 5e OIS 5a-d ø SMOW) Interstadial (‰ 23 Br rup 24 ø ø 2 4 3 5 6 7 8 Denekamp 9 -35 11 10 12 Hengelo 13 Moershoofd 14 Glinde 15 16 Oerel 17 18 19 20 21 Odderade 1 B lling-Aller d 22 -39 -43 oxygen isotopes isotopes oxygen P Stadial Hasselo Ebersdorf Schalkholz Rederstall Herning Younger Dryas Younger GRI The GRIP oxygen isotope record, showing the (numbered) abrupt climate oscillation known as b) 00 -4 -420 D (normalized) 0 -380 44 - EPICA Dome C deuterium record, spanning the last 700 ka (after community members 2004) and illustrating 100 cyclicity in 0

00 00 00 00 00 00 00

1 2 3 4 5 6 7 Age (ka BP) (ka Age Figure 1.1: a) temperature over Antarctica the past 600 ka; Dansgaard/ Oeschger events (Johnsen et al. 1992).

10 Chapter 1 coral reefs (e.g. Lambeck and Chappell 2001) show that the Quaternary has been a period characterized by numerous changes between ice-ages (glacials) and interglacials. All these changes are primarily attributed to variations in incoming solar radiation, as the result of variations in the eccentricity, axial tilt, and precession of the Earth’s movement (Milankovitch 1941).

Abrupt climate oscillations can be discerned superimposed on this long-scale Milankovitch-dominated signal, as for instance evident in the δ18O record of marine and ice-core records covering the last 110 ka (e.g. NGRIP members 2004). Within a matter of a few years, air temperatures could shift by more than 12 ºC in the high latitudes (Huber et al. 2006). During the glacial periods, these warm phases were however generally short (see Figure 1.1b), and followed by a step-wise return to colder conditions (e.g. Ganopolski and Rahmstorf 2001; Rasmussen and Thomsen 2004). These abrupt climate oscillations are known as Dansgaard/ Oeschger (D/O) events (e.g. Johnsen et al. 1992; Dansgaard et al. 1993). Although the exact origin of these millennial-scale cycles is still under debate, there is a relationship between the climate events witnessed in the Greenland ice-cores and past ocean circulation (e.g. Bond et al. 1993). D/O events seem to be centred on the North Atlantic region (Rasmussen and Thomsen 2004 and references therein), and changes in the North Atlantic thermohaline circulation most likely played a key-role in changing the amount of energy being delivered to the North Atlantic region (e.g. Ganopolski and Rahmstorf 2001).

Stream' 'Gulf

Warm surface water Cold/ deep water

Figure 1.2: Schematic outline of the thermohaline circulation in the present interglacial mode (after Rahmstorf 2006), illustrating the energy brought to mid and high northern latitudes around the Atlantic Ocean by the so-called ‘warm gulfstream’.

11 Chapter 1

1.1.3 European climate during Oxygen Isotope Stage-3 (59-28 ka BP) The modern climate of the European continent is strongly influenced by the energy brought to mid and high northern latitudes by the so-called ‘gulfstream’, which forms part of the larger thermohaline circulation (Figure 1.2). Former variations in the strength of this circulation, which are proposed as a driving mechanism for D/O- events, should therefore be evident on the European continent. Considering the generally westerly atmospheric circulation over western Europe this makes this area a key region for studying the D/O climate variability on land (Helmens et al. 2007).

Early palaeoclimatological studies focussing on the Weichselian climatic evolution on the European continent produced records in which the Early Glacial was characterized by the Amersfoort, Brørup and Odderade interstadials (e.g. Zagwijn 1961). The following Pleniglacial was considered as a generally cold period, consisting of a gradual expansion towards the maximum of the glacial (Van der Hammen 1952). The Late-Glacial period was characterized by the two warm periods Bølling and Allerød (Iversen 1942, Van der Hammen 1957). This general subdivision was used up to the end of the 1980s (Vandenberghe 1992). Subsequently, the Pleniglacial was subdivided into three periods: the “cold” Lower and Upper Pleniglacial and the “cool” Middle Pleniglacial, with the latter period characterized by several warm interstadials (Florschütz 1957; Van der Hammen et al. 1967; Zagwijn 1974; Vandenberghe 1992). This general subdivision of the Weichselian was correlated to the deep sea stratigraphical zonation (i.e. Oxygen Isotope Stages 2-5) by Woillard and Mook (1982), Vandenberghe (1985) and Guiot et al. (1989). The publications by Johnsen et al. (1992) and Dansgaard et al. (1993) showed that numerous abrupt short-term climate oscillations occurred over Greenland during OIS-3, and higher-resolution studies were performed in Europe, with the objective of determining whether the abrupt climate changes during OIS-3 as witnessed in the marine and ice-core records were also evident on the European continent.

To be able to reconstruct the short and abrupt climate oscillations as known from the marine and ice-core records in a terrestrial setting, long, continuous sediment records are essential. The number of terrestrial records in Europe registering climate change over OIS-3 is, however, limited. Continuous pollen records from France, including records from the Velay Region, Les Echets and La Grande Pile (Figure 1.3), show oscillations in the relative abundance of arboreal pollen which are assumed to be linked to D/O-like climate variability (e.g. Reille and De Beaulieu 1990; Reille et al. 2000). Furthermore, recent high-resolution studies by Veres (2007) on new cores taken from the Les Echets lake basin show new multi-proxy based evidence for the impact of D/O- events on the lake system. Spötl and Magnini (2002) published a speleothem record from the Central Alps (Austria), showing fluctuations in δ18O that are remarkably similar to Greenland interstadials 15-12 as recorded in the Greenland ice cores. A second speleothem record was recovered from the same cave, and the results published by Spötl et al. (2006) again showed a strong resemblance to the Greenland records.

12 Chapter 1

The loess sequence of Nussloch (Germany) spans the time interval between 19 and 31 ka BP (Rousseau et al. 2002). It shows 8 successive tundra-gley/loess units which have a periodicity of 1487.5 years, corresponding to the duration of Bond cycles or D/O-events, and the grain size variations in the Nussloch loess sequence show a strong similarity to the atmospheric dust content over Greenland (Rousseau et al. 2002). Vandenberghe et al. (1998b) recorded five cycles of cold phases (characterised by high sedimentation rates of coarse loess) and warmer intervals (during which gleysols were formed) in an exposure near Kesselt (Belgium). Using radiocarbon dates, these five cycles are placed in the time-interval between 41 and 27 ka.

20° 0° 20°

60° I

50° J B A C III II

E D

F H G

40°

0° 10° 20°

Figure 1.3: Study sites in northwestern and central Europe (stars; I = Sokli, II = Reichwalde/ Nochten, III = Oberwinkler Maar), covering (parts of) Oxygen Isotope Stage 3. Closed circles indicate the location of other records discussed in this chapter (A= Oerel, B= Denekamp/ Hengelo, C = Kesselt, D = Nussloch, E = La Grande Pile, F = Les Echets, G = Lac du Bouchet, H= Gossau, J = Upton Warron).

13 Chapter 1

Many sites have registered only part (-s) of OIS-3, and several interstadial periods have been reconstructed in northwestern and central Europe using fragmentary records from different sites. Locations from Great-Britain show coleopteran assemblages that indicate a temperate oceanic climate during OIS-3 with July air temperatures reaching values at least as warm as those of the present day (Coope 2002). Investigations of lacustrine and peat deposits intercalated in clastic fluvial and aeolian deposits in the Netherlands and northern Germany allowed for the identification of 5 different interstadials (e.g. Van der Hammen et al. 1967; Van der Hammen 1971; Zagwijn 1974; Behre 1989; Van Huissteden 1990; Behre and van der Plicht 1992). The opencast brown coal mines in eastern Germany and Poland provide additional sites where fragmentary records covering parts of OIS-3 are identified (e.g. Mol 1997; Bos et al. 2001; Kasse et al. 2003; Hiller et al. 2004). In the lignite mines at Gossau (Switzerland), Preusser et al. (2003) reconstructed several distinct interstadial periods in the middle Weichselian, some of which were characterized by a steppe vegetation with pine, and others by an open coniferous (Picea) vegetation. Beetle analysis suggests mean July air temperatures around 10 ºC for the steppe-dominated phases, and around 12-13 ºC for the coniferous-phases (Jost- Stauffer et al. 2001). However, quantitative data on both interstadial and stadial climate conditions during OIS-3 are still scarce, and the analysis of chironomid-remains provides for a new tool to obtain palaeotemperature estimates.

1.2 Chironomids 1.2.1 Introduction Chironomids are a diverse group of insects (Arthropoda: Insecta: Diptera: Chironomidae), including more than 5000 species worldwide (Cranston and Martin 1989). As other Diptera, chironomids have 4 life stages: egg, larva, pupa and imago (or adult). Most people will be familiar with chironomids as the black swarms of midges that occur near the rim of a lake on a warm day. The larvae of these insects mostly live in water and are typically among the most abundant invertebrates found in lakes and rivers (e.g. Cranston 1995). In the past 15 years, a new method has been developed through which the remains of the larvae of these animals can be used to reconstruct past temperatures. Below, information about the chironomid life cycle and ecology is presented, mostly based on the work by Pinder (1986), Armitage et al. (1995), and Porinchu and MacDonald (2003; see section 1.2.2). The latter part of this section focuses on the development of chironomids as a proxy for July air temperature (section 1.2.3).

1.2.2 Chironomid life cycle Adult male chironomids (Figure 1.4) form swarms, most often at dusk or dawn, and females enter this swarm to find a mate. After mating, the female deposits a gelatinous egg mass on or near the water surface of lakes, rivers or streams, often attaching it to

14 Chapter 1

Figure 1.4: Picture of an adult male specimen of Chironomus anthracinus, picture taken by K.P. Brodersen at Lake Esrum, Denmark (source: http://www.zi.ku.dk/ personal/kpbrodersen/Chiropics.htm). Used with kind permission of Klaus Brodersen. hard substrates such as rocks or water plants. A single batch of eggs can contain from 20 to 2000+ eggs. Hatching of the eggs occurs after a number of days, although the range of time involved can be between a few hours and several weeks. Timing of hatching is species-dependent, as is the preferential temperature at which hatching occurs. When hatching is successful, the newly emerged larvae swim vigorously (living planktonic in lakes) until they sink to the bottom of lakes, streams and rivers to find a suitable habitat for larval development. During the larval stage, chironomids moult (shed their cuticle) when they have outgrown their exoskeleton. The four different larval stages between the successive moults are called “instars”. The sclerotized parts of these molted cuticles (especially the head capsules of the exoskeleton) are deposited and preserved in lake sediment.

The larval stage is the most important life-stage of a chironomid, and can last between a number of days in tropical rock-pool dwellers to 7 years for certain arctic taxa. Most chironomid species that occur in subarctic and temperate environments are uni- to bivoltine, i.e. having 1 respectively 2 generations per year. The larvae inhabit a large suite of habitats, including lotic and lentic freshwater bodies, but they also occur in brackish to saline aquatic habitats, wet soils, soft substrates such as lake sediments, thermal springs, ephemeral pools, plant-held waters and in connection with aquatic macrophytes, hard substrata and submerged wood.

15 Chapter 1

(Days) (Days)

(Months) (Days)

Figure 1.5: Schematic representation of the life cycle of chironomids. Modified after Porinchu and MacDonald (2003), with kind permission of Dave Porinchu and SAGE publications Feeding strategies of the individual species are also diverse, and can be roughly classified into six groups, although most chironomids are not restricted to one mode of feeding: • Collector-gatherers: Feed on accumulated organic material, mostly fine particulate organic matter and algal remains. The larvae can be both tube- dwelling and free-living chironomids. This is the most common feeding pattern. • Collector-filterers: Suspension-feeders filter food from the water column with the use of silk nets. • Scrapers: Many Diamesinae, Orthocladinae, and a few Chironominae belong to this group, which use their mandibles to scrape food from rocks, plants, submerged wood or sediments. • Shredders: Chironomids of this group mine, chew, gouge or rasp their food, which mostly consists of coarse particulate organic matter. • Engulfers: Attack and ingest all or parts of their prey. Many Tanypodinae feed in this manner. • Piercers: Pierce their prey and suck the fluids out of it. This feeding strategy is mostly applied by some Orthocladinae.

16 Chapter 1

When an optimal temperature is reached (possibly in combination with a critical day- length or preferred light-conditions), the larvae pupate. Free pupae occur for maximally 72 hours, and most pupae are mobile, being able to swim. Some stay at the water surface, some only rise from time to time for oxygen and some only rise for emergence. After successful emergence, the male chironomids again form monospecific swarms. Most species prefer a particular time and location for swarming, although the preferred place varies depending on local factors. Swarming occurs in calm weather and at low wind speeds.

1.2.3 Chironomids as a temperature proxy Initial palaeolimnological studies in the 1960s applying chironomids as a proxy focussed on the reconstruction of the trophic state of the lake system, based on a lake classification system introduced by Thienemann in 1922 (Brooks (2006) and references therein). Walker (1987) and Walker and Matthewes (1987) were the first authors to suggest that midge (chironomid) fossils might be useful palaeoclimatic indicators. The authors proposed this relationship as they recognised the decline of Heterotrissocladius, a chironomid genus often associated with cold-oligotrophic lakes, following deglaciation in North America and Europe (Walker and Cwynar 2006). Furthermore, they realized that the best analogues for the late-glacial assemblages dominated by Heterotrissocladius are found in present-day Arctic and Alpine settings (e.g. Walker and Matthewes 1987; Walker et al. 1991; Walker and Cwynar 2006). The idea that a strong relationship exists between temperature and chironomid abundances was tested through surface sample collections of midge communities and multi-variate statistical analysis (Walker et al. 1991; Walker and Cwynar 2006). In these initial studies, the relationship between summer temperature and chironomids was shown to be strong. As a consequence, a number of modern datasets were developed, containing environmental information on lakes distributed over altitudinal or latitudinal gradients, and their associated chironomid faunas (the so-called training sets). In Europe, available training sets include those developed for the Swiss Alps (Heiri and Lotter 2005), Norway and Svalbard (Brooks and Birks 2001; unpublished data), northern Sweden (Larocque et al. 2001), and northwest Finland (Nyman et al. 2005). Application of these training sets initially focussed on the Late Glacial period (approximately 15-11.7 ka BP) and on the Holocene (11.7 ka BP – present). Battarbee et al. (2002) already indicated the considerable potential of chironomids as a proxy for reconstructing past climate, and chironomids are now seen as an essential component in many multi-proxy palaeoenvironmental studies based on lake sediment (Brooks 2006). Chironomid-based temperature reconstructions on Weichselian lake sediment sequences predating the late-glacial seemed a promising way to obtain reliable palaeotemperature estimates for a time-interval where such data is scarce (section 1.1.3).

17 Chapter 1

a) b)

Figure 1.6: Typical chironomid head capsules as encountered in lake sediments; a) Endochironomus albipennis-type, a taxon typically encountered in meso- to eutrohic water bodies (Brodin 1986), and b) Paracladius, a genus typical of cold, oligotrophic conditions (Walker et al. 1991). Source: O. Heiri Chironomids are a useful proxy because they are present in high concentrations in most freshwater bodies in the world. They often show diverse assemblages and have a good preservation potential. Furthermore, it is possible to identify preserved head capsules to species/ genus- level (Figure 1.6). As the adult chironomids are able to fly, and the chironomid life cycle is short, they are able to readily move from site to site, thus being able to rapidly respond to changes in environment or climate (e.g. Brooks et al. 2007). A large number of chironomids strongly depend on particular environmental conditions for their development, and using numerical techniques, quantitative estimates of environmental parameters that are important for chironomid survival (including oxygen availability, trophic state, salinity or July air temperature) may be derived from fossil chironomid assemblages (e.g. Brodersen and Quinlan 2006; Brooks 2006; Eggermont et al. 2006; Langdon et al. 2006).

1.3 Research objectives and results 1.3.1 Aim and research questions The processes driving the abrupt climate changes during OIS-3 are still poorly understood, and knowledge of the pattern of climate change over Europe during this period might improve our understanding of the climate dynamics and possible forcing mechanisms. The generally discontinuous nature of continental sedimentation and repeated erosion combined with poor dating control and a scarcity of high-resolution records presently hamper a detailed study of the Last Glacial climate in Europe (Helmens et al. 2007). The number of sites with a continuous registration of climate change during OIS-3 is limited (see section 1.1.3) and inferences of past changes in environment and climate are often only available for one or several interstadial

18 Chapter 1 intervals. As such, the magnitude and extent of climate change on the European continent during OIS-3 remains uncertain. This study aims to provide quantitative data on climate change during OIS-3 from key locations in Europe by employing chironomids as a new proxy for past summer temperatures during the Weichselian glaciation. Since many of the sediment records available for this time-window originate from floodplain lakes, an important aspect of this work was to explore the potential of sediments from floodplain lakes for chironomid-based quantitative climate inferences.

At the onset of this project, the general aim was to study the impact of D/O-like climate variability on the fluvial system of the Niederlausitz area in eastern Germany and on similar locations in northwestern and central Europe. During the project, opportunities to study D/O-climate variability on other key-locations in Europe arose (i.e. Eifel (Germany), Sokli (Finland)). The analyses of the lacustrine sediments derived from these localities provided information on past climate conditions for regions where such information was previously not available, as is the case in Sokli, or where no earlier studies using chironomids as a proxy were performed. Furthermore, a non- analogue situation between the former lakes from eastern Germany and the available modern training sets was studied in Finnish Lapland, and provided essential information on the influence of flooding by rivers on the chironomid fauna of floodplain lakes.

Specific questions addressed in this study include: • What is the potential of using chironomid remains that are preserved in lake sediments from OIS-3 for quantitative reconstructions of July air temperatures? • Can the abrupt climate oscillations (D/O-events) as documented in the marine and ice-core records, be recognized on the European continent by applying fast- migrating proxies such as chironomids? To what extent is the chironomid fauna of deep lakes influenced by D/O-like climate variability? • Are shallow floodplain lakes suitable archives for quantitative climate reconstructions based on chironomids? • Was the formation of thaw lakes during the Weichselian Pleniglacial in eastern Germany climate driven?

1.3.2 Thesis outline Apart from this introduction (Chapter 1), this thesis contains 6 papers which have either been published or which have been submitted for publication in peer-reviewed international scientific journals (Chapter 2-7). Due to the multi-proxy nature of palaeoenvironmental studies of lake sediments, all the papers presented here have a number of co-authors. I am the principal author of Chapters 2, 5, 6 and 7. Co-authors contributed to these papers in the form of data, ideas or contributions to the text. Two papers to which I contributed as a co-author provide a broader overview, summary or interpretation of the research projects my PhD-work provided a significant

19 Chapter 1 contribution to. For these two papers, I produced the chironomid data and their interpretation, and contributed to the writing of the text. They are therefore also included in this thesis (Chapter 3 and 4).

Chapter 2: Chironomid-based palaeotemperature estimates for northeast Finland during Oxygen Isotope Stage 3 The long sediment record from Sokli (northeast Finland) spans multiple glaciation- cycles, and includes a lacustrine sediment body covering part of OIS-3. The chironomid fauna encountered in these lacustrine sediments indicates that a shallow lake was present at the study site throughout the analyzed period. Using a Norwegian calibration data set (Brooks and Birks 2001, unpublished data), mean July air temperatures were reconstructed based on the chironomid assemblages. The palaeotemperature estimates are in the order of 10.5 – 14 ºC, which is similar to the current temperature at the study site of 13.1 ºC. As these reconstructed temperatures were unexpectedly high, the results were critically reviewed. Numerical analyses were performed to test the representation of the fossil Sokli-samples in the modern training set, various possible mechanisms that could have influenced the chironomid-inferred temperatures are discussed in this chapter, and the chironomid-based results were compared to other proxy-based reconstructions from the Sokli sequence and to temperatures derived from climate model simulations.

Chapter 3: Present-day temperatures in northern Scandinavia during the Last Glaciation Chapter 3 presents an overview of all proxy-data that was developed for the Sokli site by different project participants, including the work presented in Chapter 2, new pollen data, and quantitative pollen and macrofossil-based inferences of past climate change. Climate model-results are discussed in detail, and a possible mechanism to explain the high reconstructed July air temperatures is proposed.

Chapter 4: Rapid climatic events as recorded in Middle Weichselian thermokarst lake sediments A ~40 cm thick thermokarst deposit, recovered from the opencast lignite mine of Reichwalde (Germany), was analysed for pollen, macro-remains and chironomids and the results are presented in this chapter. Cryogenic features underlying the horizontally laminated lake sediments suggest cold conditions prior to the formation of the lake. During the initial infilling of the lake, July air temperatures were high, as is suggested by both the chironomid assemblages and the macro-remains. Based on botanical data, a minimum mean July air temperature of 12-14 ºC is reconstructed. The sharp decrease to lower temperatures (as reconstructed through semi-quantitative reconstruction of July air temperature based on the chironomid assemblages), together with the sudden drop in organic content of the sediment and the return to permafrost conditions (inferred from sedimentological features) all suggest a return to cold climate conditions during the deposition of the youngest part of the sediment record.

20 Chapter 1

The combined evidence suggests a D/O-like climate evolution forcing the formation and termination of the lake system.

Chapter 5: Intraregional variability in chironomid-inferred temperature estimates and the influence of river inundations on lacustrine chironomid assemblages Lakes on river floodplains are strongly affected by the regular inundations that occur in a natural environment. Flooding might have a distinct impact on the chironomid fauna living in a lake, as it might affect nutrient availability and water transparency, introduce new species, and influence habitat availability for chironomid larvae. For this reason, floodplain lakes are usually not included in modern training sets. However, the lake sediments presented in Chapter 4 were formed on a river floodplain, thus creating a non-analogue situation between our fossil samples and the modern lakes used in the training sets used to examine the relationship between chironomids and temperature. In order to assess the possible influence of river inundations on lacustrine chironomid-assemblages, 33 lakes were sampled during a 3- week fieldwork in Finnish Lapland. Of these lakes 13 were situated on a floodplain and thus prone to regular flooding, whereas the other 20 lakes were located outside the reach of the river, and thus did not experience any riverine influence. Only minor differences in the physical and chemical conditions in these two groups of lakes (inundated lakes versus lakes isolated from riverine influence) were detected although the environmental conditions were more variable in the group of isolated lakes. The chironomid fauna of the two groups did show differences, both with respect to taxon richness and chironomid concentration and in relative abundances of the different taxa. Using a Finnish calibration data set (Nyman et al. 2005), quantitative temperature estimates were derived from the 33 lakes. The results show surprisingly low variation in inferred temperatures, and imply that chironomid-assemblages derived from floodplain sediments can be used to quantitatively reconstruct July air temperatures, even when the training set that is used to calibrate the inference model is based on lakes that are isolated from riverine influence.

Chapter 6:Environmental inferences and chironomid-based temperature reconstructions from fragmentary records of the Weichselian Early Glacial and Pleniglacial periods in the Niederlausitz area (eastern Germany) In this chapter the sedimentary history of the opencast lignite mines of Nochten and Reichwalde, East Germany, is discussed. A chronology is obtained through optically stimulated luminescence (OSL)- and radiocarbon dating, and compared to previously existing chronologies published in other studies. A discrepancy between our chronology and earlier published records, concerning deposits formed in either OIS-4 or early OIS-3, is discussed. Two fragmentary lacustrine records, dated back to the Weichselian Early Glacial and to the Early Pleniglacial, are analysed for their chironomid content. The chironomid fauna of both records indicates that the former lakes were probably very shallow and meso-

21 Chapter 1 to eutrophic, and were situated on a floodplain. Using a Central European calibration data set (Heiri and Lotter 2005), quantitative palaeotemperature estimates were derived from the chironomid samples. The results are first compared to other proxy- records from the same cores, including temperature estimates based on macro-remains of botanical aquatic taxa. The chironomid-based temperature estimates are thereafter compared to other climate reconstructions from the Niederlausitz region, and finally placed in the larger framework of northwest and central European climatic history.

Chapter 7: The lacustrine archive of Oberwinkler Maar (Eifel, Germany): chironomid-based inferences of environmental changes during Oxygen Isotope Stage 3 There are only several lacustrine records available in Europe that continuously registered local and regional environmental and climatological change during the last ice-age, the northernmost of which is the Oberwinkler Maar record. The sediment record of the Oberwinkler Maar covers the entire OIS-3, and shows an alternation of organic-rich and clastic sediment intervals. The lower part of this record has been analysed for chironomids, and in this paper the results of these analyses and methodological problems associated with analyzing these sediments are discussed. The chironomid fauna of Oberwinkler Maar indicates that during the stadials, the former lake was relatively deep and oligotrophic. Cold-stenothermic taxa were abundant and show a surprisingly diverse assemblage. During the interstadial intervals, the number of chironomid remains encountered in the sediments decreases to very low numbers. The presence of different taxa belonging to the tribe Chironomini might indicate higher summer temperatures during these time-windows, but might also be the result of changes in oxygen-availability, trophic state or a combination of these factors. As count sums were generally low in this record, no attempt was made to quantitatively infer past changes in July air temperatures.

Chapter 8: Synthesis / Epilogue Chapter 8 concludes this thesis and the first sections focus on the potentials and problems associated with the application of chironomids as a proxy for past climate change. First, the relationship between chironomids and air temperature is discussed, and several factors that might influence this relationship are considered. The chironomid-based temperature inferences obtained in this study are compared with the results of the other proxies discussed in this thesis. Second, the palaeoclimatological data obtained in this study are placed in the broader framework of northwestern and central Europe, and the pattern that emerges from this comparison is briefly discussed. Finally, some concluding remarks concerning possible ways forward in the study of chironomids and their applications in palaeoclimatology, as well as some points concerning the progress in the study of D/O-events, are provided.

22 Chapter 1

References Armitage PD, Cranston PS, Pinder LC (eds): The Chironomidae: the biology and ecology of non-biting midges. Chapman and Hall, London, 194-224 Battarbee RW, Grytnes J-A, Thompson R, Appleby PG, Catalan J, Korhola A, Birks HJB, Heegaard E, Lami A (2002) Comparing palaeolimnological and instrumental evidence of climate change for remote mountain lakes over the last 200 years. J Paleolimnol 28: 161-179 Behre K-E (1989) Biostratigraphy of the last glacial period in Europe. Quatern Sci Rev 8: 25-44 Behre K-E, van der Plicht J (1992) Towards an absolute chronology for the last glacial period in Europe: radiocarbon dates from Oerel, northern Germany. Veget Hist Archaeobot 1: 111-117 Bond G, Broecker WS, Johnsen SJ, McManus J, Labeyrie L, Jouzel J, Bonani G (1993) Correlations between climate records from North Atlantic sediments and Greenland ice. Nature 365: 143–147 Bos JAA, Bohncke SJP, Kasse C, Vandenberghe J (2001) Vegetation and Climate during the Weichselian Early Glacial and Pleniglacial in the Niederlausitz, eastern Germany - macrofossil and pollen evidence. J Quatern Sci 16: 269-289 Brodersen KP, Quinlan R (2006) Midges as palaeoindicators of lake productivity, eutrophication and hypolimnetic oxygen. Quatern Sci Rev 25: 1995-2012 Brodin YW (1986) The postglacial history of Lake Flarken, southern Sweden, interpreted from subfossil insect remains. Internat Rev Gesamt Hydrobiol 71: 371-432 Brooks SJ (2006) Fossil midges (Diptera: Chironomidae) as palaeoclimatic indicators for the Eurasian region. Quatern Sci Rev 25: 1894-1910 Brooks SJ, Birks HJB (2001) Chironomid-inferred air temperatures from Lateglacial and Holocene sites in north- west Europe: progress and problems. Quatern Sci Rev 20: 1723-1741 Brooks SJ, Langdon PG, Heiri O (2007) The identification and use of palaearctic Chironomidae larvae in palaeoecology. Quatern Res Assoc Technical Guide 10, 276 pp Coope GR (2002) Changes in the Thermal Climate in Northwestern Europe during Marine Oxygen Isotope Stage 3, Estimated from Fossil Insect assemblages. Quatern Res 57: 401-408 Cranston PS (1995) Systematics. In: Armitage PD, Cranston PS, Pinder LC (eds): The Chironomidae: the biology and ecology of non-biting midges. Chapman and Hall, London: 31-52 Cranston PS, Martin J (1989) Family Chironomidae. In: Evenhuis NL (ed) Catalogue of the Diptera of the Australasian and Oceanic regions. Leiden and Honolulu: E.J. Brill and Bishop Museum Press, pp 252- 274 Crowhurst SJ (2002) Composite isotope sequence. The Delphi project. http://www.esc.cam.ac.uk/new/v10/ research/institutes/godwin/body.html. Dansgaard W, Johnsen SJ, Clausen HB, Dahl-Jensen D, Gundestrup NS, Hammer CU, Hvidberg CS, Steffensen JP, Sveinbjörndottir AE, Jouzel J, Bond G (1993) Evidence for general instability of past climate from a 250-kyr ice-core record. Nature 364: 218-220 Eggermont H, Heiri O, Verschuren D (2006) Fossil Chironomidae (Insecta: Diptera) as quantitative indicators of past salinity in African lakes. Quatern Sci Rev 25: 1966–1994 EPICA community members (2004) Eight glacial cycles from an Antarctic ice core. Nature 429: 623-628 Florschütz F (1957) The subdivisions of the Middle and Young Pleistocene up to the Late-glacial in The Netherlands, England and Germany, mainly based on the results of paleobotanical investigations. Geol Mijnb NS 19: 245-249 Ganopolski A, Rahmstorf S (2001) Rapid changes of glacial climate simulated in a coupled climate model. Nature 409: 153-158 Gibbard PL, Boreham S, Cohen KM, Moscariello A (2005) Global chronostratigraphical correlation table for the last 2.7 million years. Boreas 34: unpaginated. Modified 2007. Guiot J, Pons A, de Beaulieu J-L, Reille M (1989) A 140,000-year continental climate reconstruction from two European pollen records. Nature 338: 309-313 Heiri O, Lotter AF (2005) Holocene and Lateglacial summer temperature reconstruction in the Swiss Alps based on fossil assemblages of aquatic organisms: a review. Boreas 34: 506-516 Helmens KF, Bos JAA, Engels S, Van Meerbeeck C, Bohncke SJP, Renssen H, Heiri O, Brooks SJ, Seppä H, Birks HJB, Wohlfarth B (2007) Ice-free conditions and present-day temperatures during the last glacial at 50ka in the central area of the Scandinavian glaciations. Geology 35: 987-990 Hiller A, Junge FW, Geyh MA, Krbetschek M, Kremenetski C (2004) Characterising and dating Weichselian organogenic sediments: a case study from the Lusatian ice marginal valley (Scheibe opencast mine, eastern Germany) Palaeogeog Palaeoclimatol Palaeoecol 205: 273-294 Huber C, Leuenberger M, Spahni R, Flückiger J, Schwander J, Stocker TF, Johnsen S, Landais A, Jouzel J (2006)

Isotope calibrated Greenland temperature record over Marine Isotope Stage 3 and its relation to CH4. Earth Planet Sci Lett 243: 504-519

23 Chapter 1

IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the IPCC (ISBN 978 0521 88009-1) Iversen J (1942) En pollenanalytisk tidsfuestelse of Ferksvandslagene ved Nørre Lyndby. Med Dansk Geol Foren 10: 130-151 Johnsen SJ, Clausen HB, Dansgaard W, Fuhrer K, Gundestrup N, Hammer CU, Iversen P, Jouzel J, Stauffer B, Steffensen JP (1992) Irregular glacial interstadials recorded in a new Greenland ice core. Nature 359: 311-313 Jost-Stauffer M, Coope GR, Schlüchter C (2001) A coleopteran fauna from the middle Würm (Weichselian) of Switzerland and its bearing on palaeobiogeography, palaeoclimate and palaeoecology. J Quatern Sci 16: 257-268 Kasse C, Vandenberghe J, Van Huissteden J, Bohncke SJP, Bos JAA (2003) Sensitivity of Weichselian fluvial systems to climate change (Nochten mine, eastern Germany). Quatern Sci Rev 22: 2141-2156 Lambeck K, Chappell J (2001) Sea Level Change Through the Last Glacial Cycle. Science 292: 679-686 Langdon PG, Ruiz Z, Brodersen KP, Foster IDL (2006) Assessing lake eutrophication using chironomids: understanding the nature of community response in different lake types. Freshw Biol 51: 562-577 Larocque I, Hall RI, Grahn E (2001) Chironomids as indicators of climate change: a 100-lake training set from a subarctic region of northern Sweden (Lapland). J Paleolimnol 26: 307-322 Milankovitch M (1941) Kanon der Erdbestrahlung und seine Anwendung auf das Eiszeitenproblem. R Acad Spec Publ 133 Mol J (1997) Fluvial response to Weichselian climate changes in Niederlausitz (Germany). J Quatern Sci 12: 43- 60 NGRIP members (2004) High-resolution record of Northern Hemisphere climate extending into the last interglacial period. Nature 431: 147-151 Nyman M, Korhola A, Brooks SJ (2005) The distribution and diversity of Chironomidae (Insecta: Diptera) in western Finnish Lapland, with special emphasis on shallow lakes. Global Ecol Biogeogr 14: 137-153 Pinder LCV (1986) Biology of freshwater Chironomidae. Ann Rev Entomol 31: 1-23 Porinchu DF, MacDonald GM (2003) The use and application of freshwater midges (Chironomidae: insecta: diptera) in geographical research. Progr Phys Geogr 27: 378-422 Preusser F, Geyh MA, Schlüchter C (2003) Timing of Late Pleistocene climate change in lowland Switzerland. Quatern Sci Rev 22: 1435-1445 Rahmstorf S (2006) Thermohaline Ocean Circulation. In Elias SA (Ed): Encyclopedia of Quaternary Sciences, 10 pp Rasmussen TL, Thomsen E (2004) The role of the North Atlantic Drift in the millennial timescale glacial climate fluctuations. Palaeogeog Palaeoclimatol Palaeoecol 210: 101-116 Reille M, de Beaulieu J-L (1990) Pollen analysis of a long upper Pleistocene continental sequence in a Velay Maar (Massif Central, France). Palaeogeogr Palaeoclimatol Palaeoecol 80: 35-48 Reille M, de Beaulieu J-L, Svobodova H, Andrieu-Ponel V, Goeury C (2000) Pollenanalytical biostratigraphy of the last five climatic cycles from a long continental sequence from the Velay region (Massif Central, France). J Quatern Sci 15: 665-685 Rousseau DD, Antoine P, Hatté C, Lang A, Zöller L, Fontugne M, Ben Othman D, Luck JM, Moine O, Labonne M, Bentaleb I, Jolly D (2002) Abrupt millennial climatic changes from Nussloch (Germany) Upper Weichselian eolian records during the Last Glaciation. Quatern Sci Rev 21: 1577-1582 Ruddiman WF (2001) Earth’s climate: past and future. WH Freeman and Company, New York, 2001 Spötl C, Mangini A (2002) Stalagmite from the Austrian Alps reveal Dansgaard-Oeschger events during isotope stage 3: Implications for the absolute chronology of Greenland ice cores. Earth Planet Sci Lett 203: 507-518 Spötl C, Mangini A, Richards DA (2006) Chronology and paleoenvironment of Marine Isotope Stage 3 from two high-elevation speleothems, Austrian Alps. Quatern Sci Rev 25: 1127-1136 Thienemann A (1922) Die beiden Chironomusarten der Tiefenfauna der norddeutschen Seen. Ein hydrobiologisches Problem. Arch Hydrobiol 13: 609-646 Van der Hammen Th (1952) Dating and correlation of periglacial deposits in Middle and Western Europe. Geol Mijnb NS 14: 328-336 Van der Hammen Th (1957) A new interpretation of the Pleniglacial stratigraphical sequence in Middle and Western Europe. Geol Mijnb NS 19: 493-498 Van der Hammen Th, Maarleveld GC, Vogel JC, Zagwijn W (1967) Stratigraphy, climatic succession and radiocarbon dating of the last glacial in the Netherlands. Geol Mijnb 46: 79-95 Van der Hammen Th (1971) The Denekamp, Hengelo and Moershoofd Interstadials. Meded Rijks Geol Dienst 22: 81-85 Van Huissteden J (1990) Tundra rivers of the Last Glacial: sedimentation and geomorphological processes during the Middle Pleniglacial in the Dinkel valley (eastern Netherlands). Meded Rijks Geol Dienst 44: 3-138

24 Chapter 1

Vandenberghe J (1985) Paleoenvironment and stratigraphy during the Last Glacial in the Belgian-Dutch border region. Quatern Res 24: 23-38 Vandenberghe J (1992) Geomorphology and climate of the cool oxygen isotope stage 3 in comparison with the cold stages 2 and 4 in The Netherlands. Z Geomorphol Suppl 86: 65-75 Vandenberghe J, Coope R, Kasse C (1998a) Quantitative reconstructions of palaeoclimates during the last interglacial-glacial in western and central Europe: as introduction. J Quatern Sci 13: 361-366 Vandenberghe J, Huijzer BS, Mücher H, Laan W (1998b) Short climatic oscillations in a western European loess sequence (Kesselt, Belgium). J Quatern Sci 13: 471-485 Veres, DS (2007) Terrestrial response to Dansgaard-Oeschger cycles and Heinrich events: the lacustrine record of Les Echets, south-eastern France. Dissertations from the department of Physical Geography and Quaternary Geology No 6, Stockholm University, 110 pp Walker IR (1987) Chironomidae (Diptera) in paleoecology. Quatern Sci Rev 6, 29-40 Walker IR, Cwynar LC (2006) Midges and palaeotemperature reconstruction – the North American experience. Quatern Sci Rev 25: 1911-1925 Walker IR, Matthewes RW (1987) Chironomidae (Diptera) and postglacial climate at Marion Lake, British Columbia, Canada. Quatern Res 27: 89-102 Walker IR, Smol JP, Engstrom DR, Birks HJB (1991) An assessment of Chironomidae as Quantitative Indicators of Past Climatic Change. Can J Fish Aquat Sci 48: 975-987 Woillard GM, Mook WG (1982) C-24 dates at Grande Pile – Correlation of land and sea chronologies. Science 215: 159-161 Zagwijn W (1961) Vegetation, climate and radiocarbon datings in the Late Pleistocene of the Netherlands, Part I: Eemian and Early Weichselian. Meded Rijks Geol Dienst NS 14:15-45 Zagwijn W (1974) Vegetation, climate and radiocarbon datings in the Late Pleistocene of the Netherlands, Part II: Middle Weichselian. Meded Rijks Geol Dienst NS 25: 101-111

25 Chapter 1

26 Chapter 2

Chapter 2

Chironomid-based palaeotemperature estimates for northeast Finland during Oxygen Isotope Stage 3

S Engels, SJP Bohncke, JAA Bos, SJ Brooks, O Heiri and KF Helmens

Abstract Quantitative palaeotemperature estimates for the earlier part of Oxygen Isotope Stage (OIS-) 3 are inferred from subfossil chironomid remains. The high-latitudinal study site of Sokli, northeast Finland, provides a unique lacustrine deposit covering the earlier part of OIS-3, and the chironomid remains found in the sediments show that a shallow lake with a diverse fauna was present at the study site throughout the record. Using a Norwegian calibration data set as a modern analogue, mean July air temperatures are reconstructed. The chironomid-inferred July air temperatures are surprisingly high, reaching values similar to the current temperature at the study site. Other proxies that were applied to the sediments included the analysis of botanical and zoological macro-remains, and our results concur with temperature estimates derived from climate indicator taxa. Summer temperatures for interstadial conditions, reconstructed with climate models, are as high as our proxy-based palaeotemperatures.

This manuscript was published on OnlineFirst for Journal of Paleolimnology DOI: 10.1007/s10933-007-9133-y

27 Chapter 2

2.1 Introduction There is ample evidence from marine and ice core-records for the existence of abrupt climate oscillations, the so-called Dansgaard/ Oeschger (D/O) events, during Oxygen Isotope Stage (OIS)-3 (Johnsen et al. 1992; Dansgaard et al. 1993; Voelker et al. 2002). The mechanisms driving these abrupt and dramatic climate shifts are still poorly understood. An important step towards understanding these mechanisms is to analyze the global or hemispheric signature of the climate shifts associated with D/O events, for which quantitative information from a range of archives and for different regions including the ice-sheets, the marine realm, and the terrestrial realm is needed. However, the scarcity of terrestrial records dating back to OIS-3, which document palaeoenvironmental change on a local or regional scale, presently prevents correlation between the marine and ice core records and the terrestrial archives. Continuous high-resolution terrestrial records from the European continent, which show climatic trends similar to the ice core and marine records, are sparse and are found only in southerly locations where stadial and interstadial periods are registered in pollen-records (e.g. De Beaulieu and Reille 1992). In central and northwestern Europe, only a limited number of sequential interstadial records have been recognized (e.g. Van der Hammen et al. 1967; Huijzer and Vandenberghe 1998). Terrestrial climate records from high-latitudinal regions of Europe and dating back to the last glacial are, because of the past existence of an extensive ice-sheet, virtually non-existent. This absence is especially unfortunate since high-latitudinal regions are considered to be most affected by climate change (e.g. Cubasch et al. 2001). The Sokli study- site, situated in northeast Finland, consists of a small basin with atypical bedrock conditions. The exceptional geological setting provides for a sedimentary basin where the deposits show a sequence of several fossil-rich interstadial sediments (Helmens et al. 2000; 2007). Finely laminated lacustrine deposits dated to the early part of OIS-3 provide a unique opportunity to study the local environmental changes through the analysis of a range of different proxies. Chironomids (or non-biting midges) are recognized as a powerful proxy for inferring past climatic changes (e.g. Walker et al. 1991; Brooks and Birks 2001; Barley et al. 2006) and have been widely used to infer quantitative mean July air temperatures on both late-glacial and Holocene timescales (e.g. Ilyashuk et al. 2005; Brooks 2006; Walker and Cwynar 2006). Other studies used fossil chironomid assemblages in order to quantitatively reconstruct lake-water salinity (e.g. Eggermont et al. 2006), hypolimnetic oxygenation (Quinland and Smol 2001) or trophic conditions in lakes (Woodward and Shulmeister 2006). Quantitative palaeoclimatic inferences can contribute to our understanding of climate dynamics and forcing mechanisms (e.g. Vandenberghe et al 1998), as well as serving to validate climate models (e.g. Cane et al. 2006; Kageyama et al. 2006). Here we present the results of a detailed chironomid analysis of high-latitudinal lacustrine sediments, covering the earlier part of OIS-3, and palaeotemperature estimates for a time-interval where such quantitative data are sparse.

28 Chapter 2

2.2 Site description and materials 2.2.1 Sokli study-site The Sokli study-site (67º48’N, 29º18’E) is situated in the northeastern part of Finnish Lapland (Figure 2.1). The current mire system stretches along the Sokli rivulet for a length of approximately 15 km (Helmens et al. 2000), and is situated at an altitude of approximately 220 m a.s.l. Gently sloping hills, which reach an elevation of about 300-350 m, surround this mire and rivulet system (Helmens et al. 2000). The area is currently situated in the northern boreal forest dominated by birch (Betula), spruce (Picea) and pine (Pinus), and the coring site itself is characterized by a vegetation consisting mainly of sedges (mostly Carex). The present climate in the region is “cold temperate”, having a mean July air temperature of 13 to14 ºC and a mean February temperature of -14 ºC. Annual precipitation averages between 500 and 550 mm (Drebs et al. 2002). Systematic coring of the Sokli site started in the early 1990’s and, in 1996, two long cores (boreholes 900 and 901) were taken from the central part of the Sokli basin. The cores (2 m length, 4 cm diameter) were taken using a hydraulic piston corer designed by the Geological Survey of Finland, which was driven into the sediment by vibration. Lithological analysis of these cores showed major gaps in the sequence, for instance between the OIS- 5e (or Eemian) interglacial sediments and the OIS-3

N 70º N 20º E 25º E 30º E

Sokli

Arctic

65º N

km 60º N 0 100 200

Figure 2.1: Location map of Finland with major rivers and lakes, and the location of the Sokli site (asterisk)

29 Chapter 2 interstadial deposits (Helmens et al. 2000). In 2002, the Sokli A- and B-series were taken at a distance of 1-2 m from boreholes 900/901 in an attempt to close the major gaps in the former cores.

2.2.2 Borehole- stratigraphy of the Sokli B-series The Sokli B-series covers a depth of 26 m and shows only minor gaps in the sequence corresponding to the coarsest sediment layers, where drilling through coarse sands and gravels was difficult. Using both radiocarbon and optically stimulated luminescence (OSL) dating techniques, and using the pollen assemblages of the lowermost sediments to identify the OIS-5e (or Eemian) deposits, a chronology was established for the Sokli B-series (Helmens et al. 2007). A simplified lithology for Borehole Sokli B-series is shown in Figure 2.2. Nine Late-Quaternary stratigraphic units (SU) are distinguished (Helmens et al. 2007) the lowermost 5 m of sediment, consisting of diatom gyttja, and was formed during OIS- 5e. Above 21 m core depth, a gravelly deposit (lower part of SU-2) with a fluvial origin is present. The upper part of SU-2 consists of a 2 m thick sand/silt sequence, showing parallel laminae that could have formed either in a low-energetic fluvial system or in a shallow lake environment. The minerogenic sediments of SU-2 grade into a thick laminated gyttja deposit, which, in its upper parts, is interlayered with sand and gravel (SU-3). The OSL-dates, as well as the stratigraphy, indicate an age of OIS-5d and OIS-5c for SU-2 and 3 respectively. At 13.4 m core depth, a 2 m thick diamict deposit was recovered (SU-4). These glacially deposited materials correlate with OIS-5b. The overlying ice-marginal gravels and sands, grading into laminated sands and silts and then into laminated sandy gyttja (SU-5), and are correlated to OIS-5a. The till deposits of SU-6 (8.1-9.5 m core depth) represent the second phase of glaciation at the Sokli site. The till is overlain by glacio-fluvial sediments and thereafter by a 2 m thick minerogenic lacustrine deposit (SU-7). The sediments of this limnic deposit first become finer up the sequence, followed by a coarsening upwards, and are finally overlain by another till deposit (SU- 8). The OSL-dates, the (uncalibrated) radiocarbon date and the stratigraphy indicate an age of OIS-4, early OIS-3 and OIS-3/ OIS-2 for SU-6, 7 and 8 respectively. The top of the sedimentological record consists of glacio-fluvial sediments grading into a Holocene peat deposit (SU- 9).

2.2.3 Sedimentary record of SU-7 In this study, a detailed analysis of the sediments of stratigraphic unit 7 (early OIS-3), which has been defined as the Tulppio Interstadial (Helmens et al. 2007), is presented. The lowermost sediments of SU-7 are relatively coarse, low in organic content, and show no sedimentary structures. Gradually, the sediment transforms into clayey sediments with a slightly higher organic content. Between 6.85 and 6.75 m core depth, millimeter-scale alterations between clayey and silty layers are observed in the sediments. These laminations increase in thickness upward, and continue to 5.9 m

30 Chapter 2

Figure 2.2: Simplified lithological column of Borehole B-series at Sokli, with absolute age determinations (OSL-dates and uncorrected 14C-date (in italic)) and stratigraphical units (SU). The black bar Dates to the right of the sediment-column StratigraphicalLithology (ka BP) represents the Tulppio interstadial (see text for description) 0

SU 9

SU 8 5 54 ± 7/4 SU 7

48 ± 16 SU 6 74 ± 26 10 SU 5 80 ± 26

SU 4

94 ± 16 Depth (m) Depth 15 SU 3 94 ± 19

SU 2 Legend 20 Peat deposit Lacustrine deposit Lacustrine and fluvial deposits SU 1 Fluvial, glacio-fluvial and glacial deposits 25 Coring gap

Silt Clay Sand Gravel

31 Chapter 2 core depth. The sediments become coarser, containing a matrix of silts and, above 5.9 m core depth, of sands. Around 5.2 m core depth wood fragments are encountered in the core and, at 5.14 m core depth, a large pebble is present within the much finer matrix. The presence of this stone is probably the result of ice rafting, which brought coarse material into the low-energetic lake system and deposited so-called dropstones.

2.3 Methods 2.3.1 Chironomid analysis The 2 m thick deposit (between 7.1-5.1 m core depth) was subsampled in 5 cm thick slices. A total of 39 wet sediment samples (weight range: 3.65 – 38.54 g) was used for chironomid analysis. To remove fine particles from the samples, the sediments were deflocculated in cold 10% KOH for at least 4 hours and subsequently rinsed through a 100 µm sieve. Using a dissecting microscope, chironomid head capsules were hand- picked at 35x magnification using fine forceps and permanently mounted on glass slides using Euparal© mounting medium. The chironomid head capsules were identified following Wiederholm (1983), Heiri et al. (2004), Makarchenko and Makarchenko (1999), Moller Pillot (1984), Schmid (1993), Rieradevall and Brooks (2001) and Oliver and Roussel (1983). A chironomid percentage diagram was constructed using the computer programs TILIA and TG.VIEW (Grimm 1991-2004). Zonation of the chironomid diagram was carried out using the optimal sum-of- squares partitioning method as implemented in the program ZONE (Lotter and Juggens 1991), the significant number of zones was assessed by a broken stick model (Bennett 1996).

2.3.2 Temperature inference A Norwegian chironomid-temperature calibration data set was used as a modern analogue for our fossil chironomid assemblages. This training set contains 153 Norwegian lakes and spans a mean July air temperature range of 3.5-15.6 ºC (unpublished data; Brooks and Birks 2001). A 3-component weighted-averaging partial least squares (WA-PLS) model was selected as the inference model with the best predictive power, using mean July air temperatures as the response variable and chironomid taxa as the predictor variables. Using leave-one-out cross-validation (jack-knifing), the predictive powers of the model were estimated (Birks 1995). Our inference model had a RMSEP of 1.01 ºC, an r2 of 0.91 and a maximum bias of 1.05 ºC. A bootstrap cross-validation with 499 cycles was performed to calculate sample- specific error estimates. Using the modern analogue technique described by Birks et al. (1990), the occurrence of non-analogues in the fossil samples was calculated, where “no close” or “no good” analogues with the modern data were defined as the cut- level of the 2nd and 5th percentile of all chi-square distances in the modern calibration data. A cut-level of the 95th and the 90th percentile of the modern residual chi-square distances was used as an estimate of the fit of the fossil samples to temperature (“very poor” and “poor” respectively; Heiri et al. 2003a). The percentage of rare taxa in the

32 Chapter 2

fossil samples was calculated, where a rare taxon has a Hill’s N2 (Hill 1973) of 5 or less in the modern data set. Brooks and Birks (2001) state that the temperature optima of taxa with a N2 > 5 in the modern data are likely to be reliably estimated, whereas taxa with N2 < 5 are rare in the modern data, and the optima for these taxa are probably poorly estimated. The temperature inference model, the analogues and the percentage of rare taxa were calculated using C2 version 1.4.3 (Juggins 2003), and the modern residual chi-square distances using CANOCO version 4.0 (Ter Braak and Šmilauer 1998). The lowermost samples (below 6.79 m core depth) show low numbers of chironomid head capsules, even after all available material was examined for chironomid remains. These samples were therefore lumped and used as a single sample in the numerical analyses.

2.4 Results and ecological interpretation 2.4.1 Chironomid record A total of 65 chironomid taxa was identified in the Sokli sediments and a summarized chironomid sequence is presented in Figure 2.3. The average count sum in the samples above 6.75 m core depth was 89 head capsules (hc) per sample (range: 54 – 199 hc), although fewer counts were present in the lowermost part of the profile (2-24 hc). The high number of lacustrine taxa with a preference for littoral to sub-littoral habitats suggests that throughout the record a shallow lake was present at the Sokli site. Four zones have been distinguished in the chironomid record:

Zone S-Ch1 (7.10 - 6.71 m): Tanytarsus lugens-type, Microtendipes and Polypedilum nubeculosum-type were among the taxa that first appeared at Sokli, and might be considered the early colonisers of the newly formed lake at Sokli. The chironomid concentration does not rise above 1 hc/g, possibly due to high sedimentation rates. In the upper part of zone S-Ch1, count sums increase to values above 50, and Chironomus anthracinus-type (20%), T. lugens-type (15-40%) and Procladius (10%) are the most abundant taxa. Both C. anthracinus-type and Procladius have a broad distribution along the temperature gradient in the Norwegian training set: Procladius has a large temperature tolerance, as it was found in lakes with mean July air temperatures ranging from 4.5 ºC to 16.0 ºC. The abundances of C. anthracinus•-type show a bimodal distribution in the modern calibration set, with a small optimum in abundances at 8.5 ºC, and a large optimum in abundances around 14.6 ºC. T. lugens- type, the third taxon to show high numbers, also has a large temperature-tolerance, but has been considered to be indicative for cool and oligotrophic conditions (e.g. Brooks 2006). All these taxa might be indicative of a relatively deep lake.

Zone S-Ch2 (6.71 – 6.40 m): The onset of zone S-Ch2 is indicated by a sharp increase in the concentration of chironomid remains, reaching values of 50 hc/g wet sediment. The assemblages are

33

Chapter 2

Chironomid zone Chironomid

S-Ch4 S-Ch3 S-Ch2 S-Ch1

Prodiamesa

Eukiefferiella-type

ynocera ambigua ynocera

r 20%

o

opsectra insignilobus-type opsectra C

us Micr

20% Paracladi

20%

Stictochironomus

nytarsus mendax-type nytarsus 20% ancus-type 2 ancus-type

Ta

osum-type l

20 40% adotanytarsus m adotanytarsus

Cl

um nubecu um s-type

20 40%

Polypedil

es

endip ot

20 40% icr M

20% Endochironomus albipenni Endochironomus

20%

Ablabesmyia

nytarsus lugens-type nytarsus 20 40%

Ta ladius

20 40%

Proc

onomus anthracinus-type onomus

r 20 40% Single occurence Single

Chi Parakiefferiella nigra-type Parakiefferiella

Fine

Coarse

d concentration (hc/g) concentration d i

Laminations

onom

20 40 60

sum Chir

Count 2 3 3 3 2

logy 53 15 12 24 71 71 82 81 79 98 64 58 86 77 82 76 87 49 80 68 81 47 58 65 67 70 83 67

Sandy silt Sandy 152 193 109 104 125 131 Silt Clay Litho Lithology

: Summary diagram showing selected chironomid taxa recovered from the Sokli sediments. Abundances for individual are given as

5,10 5,30 5,50 5,70 5,90 6,10 6,30 6,50 6,70 6,90 7,10 Depth (m) Depth Figure 2.3 percentages, the head capsule count-sum in numbers, and chironomid concentration number of capsules per gram wet sediment. The samples below 6.79 m core depth yielded limited numbers of chironomid head capsules, and individual findings are indicated with a (+)

34 Chapter 2 initially dominated by T. lugens-type, with abundances of 40% at the onset of this zone. Ablabesmyia shows its highest occurrences in the middle part of this zone (10%). Toward the top of zone S-Ch2, the abundance of T. lugens-type declines and Tanytarsus mendax-type, Polypedilum nubeculosum-type and Cladotanytarsus mancus- type 2 become the dominant taxa. The latter two taxa (values above 20%) are indicative of higher July air temperatures and the transition to a chironomid- assemblage that is dominated by these two taxa could be the result of a steadily increasing summer temperature at Sokli. At the end of zone S-Ch2 there is a peak in the occurrence of Endochironomus albipennis-type, which temporarily becomes the dominant taxon with an abundance of 20%. This might indicate rising temperatures too, but could also be related to an increased nutrient availability in the lake.

Zone S-Ch3 (6.40 – 6.00 m): Endochironomus albipennis-type and T. lugens-type show lower abundances at the onset of S-Ch3, while Microtendipes returns to the lake and reaches abundances of 5-15%. P. nubeculosum-type and Cladotanytarsus mancus-type 2 remain the dominant taxa with an abundance of 20% and 20-30% respectively. Ablabesmyia shows declining abundances and temporarily disappears from the sequence during zone S-Ch3, while Corynocera ambigua appears in the middle of the zone. The concentration of chironomid head capsules is low throughout zone S-Ch3, between 5 and 10 hc/g. The occurrence of Microtendipes could indicate temperatures that are best classified as intermediate to warm (Bedford et al. 2004; Brooks and Birks 2000) and the gradual warming trend that was observed in zone S-Ch2 was probably interrupted.

Zone S-Ch4 (6.00 – 5.10 m): At the onset of S-Ch4, Microtendipes disappears from the sequence, whereas C. ambigua shows a constant abundance of 15%. This latter species was formerly considered to be a true cold-water stenotherm, but those ideas have recently been questioned by Brodersen and Lindegaard (1999), as they found C. ambigua in temperature, eutrophic Danish lakes. In the Norwegian training set, this species has a narrow temperature tolerance (± 1-2 ºC) with a temperature optimum of 9.9 ºC. Paracladius shows its first phase of high abundance, reaching values of 10%. At 5.83 m core depth, values of Paracladius decline and the head capsule density decreases to values below 10 hc/g. In the uppermost part of the core, between 5.10-5.50 m, this taxon reappears in the chironomid assemblage, but in much lower numbers. This genus is considered to be a true cold-water stenotherm, and its presence is possibly the result of a colder temperature at the study site. At 5.60 m core depth, the head capsule density increases again to approximately 50 hc/g. Stictochironomus reaches abundances of up to 10% in the upper part of the record. However, P. nubeculosum-type and Cladotanytarsus mancus- type 2 are still the dominant taxa in the assemblage. An interesting feature of the upper part of zone S-Ch4 is the occurrence of stream-inhabiting taxa such as Eukiefferiella and Prodiamesa.

35 Chapter 2

xa Inferred July Air Temperature (°C) are ta Fit-to-TR 10 11 12 13 14 Non-analogues

5.0

5.5

6.0 Depth (m) Depth

6.5

7.0

Figure 2.4: Chironomid-inferred mean July air temperatures, and sample spe- cific error-bars. The open circles indicate samples with no good analogues, and samples with more than 5% abundance of that are taxa rare in the modern training set. The solid circles indicate samples with no close analogues, with more than 10% abundance of taxa that are rare in the modern training set, and samples with a poor fit to temperature

The occurrence of taxa with a preference for lotic habitats, together with the coarser-grained sediment in this interval, suggests an increasing influence on the lake-catchment of proglacial streams from the nearby ice-sheet. The close proximity of the advancing ice sheet is further demonstrated by the uppermost part of the lacustrine deposit, which changes into a glacial till deposit, probably without a hiatus in the sedimentary sequence (Helmens et al. 2007).

2.4.2 Quantitative climatic reconstruction Figure 2.4 shows the reconstructed mean July air temperatures with the sample specific error bars. The reconstructed mean July air temperatures are relatively low (12.0 ºC) for the lowermost part of the record but they increase steadily to values around 13.5 ºC. At 6.30 m core depth, temperatures decline to values around 10.5-11.0 ºC after which they stabilise around 12.0-13.0 ºC up to 5.30 m core depth. The uppermost three samples show a return to colder conditions with an inferred temperature of 11.5-12.0 ºC. Since the reconstructed July air temperatures are not near

36 Chapter 2 the upper limit of the temperature range of the modern training set (15.6 ºC), the influence of so-called edge effects (Ter Braak and Juggins 1993; Birks 1998) can be excluded. Furthermore, where under colder climatic conditions the biodiversity tends to be reduced and the same cold-adapted assemblage of chironomids might exist over a wide temperature range (Birks and Birks 2006), our reconstructed temperatures are in the intermediate to warm range and so are not likely to suffer from this problem The cumulative abundance of the six taxa included in the fossil assemblages but absent from the modern calibration set is between 0.0-6.9 % per sample, whereas on average 98.7% of the identified fossil chironomids were used to obtain palaeotemperature estimates. Most of the identified fossil chironomids were well- represented in the modern training set, as there are only 3 fossil samples that have an abundance of rare taxa higher than 10%. These samples all have a high abundance of Paracladius, a taxon that in the modern training set has an effective number of occurrences (Hills N2) of 4.6, just below our threshold of 5. Only 1 sample shows a poor fit to temperature and not a single sample has a residual chi-square distance higher than the extreme 5 percent of the modern samples and therefore none is classified as having a “very poor” fit to temperature. The fossil sample that is classified as having a “poor fit” to temperature is situated in a part of the record where both the chironomid-assemblages as well as the reconstructed temperatures are stable. No fossil sample has a close analogue in the modern data, and 75% of the fossil samples have no good analogues. This result is surprising, as all the dominant taxa of the fossil record are all well-represented in the Norwegian training set and the training set contains many shallow, macrophyte rich lakes that could potentially provide good analogues for the fossil chironomid-assemblages. Probably, the relative abundances of taxa in the fossil samples are not (well) reflected in the individual training set lakes. Although there are no close modern analogues in the training set, WA-PLS can perform relatively well in poor analogue situations (Birks 1998). Since there is a good fit to temperature and a high number of fossil taxa that are well-represented in the modern training set, the inferred values are considered to be reliable given the properties of the modern calibration data.

2.5 Discussion The results presented in this study show the first chironomid-derived palaeotemperature estimates for early OIS-3 in northwestern Europe. Our reconstructed temperatures are surprisingly high, considering that the deposits were formed during the last glacial. The highest reconstructed temperatures are as high as the current mean July air temperature (13 ºC) at the study site, and even the lower reconstructed values of 10.5 ºC are temperatures that are currently found in northern parts of Finland and Norway (Drebs et al. 2002). Chironomid studies on both late-glacial and Holocene sediments have produced inferred temperature reconstructions that were in concurrence with other

37 Chapter 2 proxy-based temperature estimates (e.g. Heiri and Millet 2005; Magny et al. 2006), but also reconstructions that showed discrepancies between different sites (e.g. Velle et al. 2005) or with reconstructions based on different proxies (e.g. Birks and Ammann 2000). Larocque and Hall (2003) elegantly show that in their study there was a close similarity between measured and chironomid-inferred summer temperatures, but they also state that on longer temporal timescales, this relationship might not remain constant. Therefore, each record showing midge-palaeotemperature reconstructions must be interpreted cautiously, and in the context of all palaeoenvironmental data available (Heiri and Lotter 2005; Walker and Cwynar 2006). Below, we discuss mechanisms that could potentially have influenced our chironomid-inferred temperatures. Second, the chironomid-based results are compared to other proxy-based results from Sokli, and to model-based July air temperature reconstructions for north Finland. Finally, our results will be compared to other temperature records inferred from terrestrial sites in northwestern Europe.

2.5.1 Sampling design The fossil chironomid remains recovered in the sediments from the Sokli site are surprisingly well preserved, considering that an ice-sheet has overridden these deposits. Identification of the fossil material was possible to a degree of taxonomic resolution similar to the calibration set. Great care was taken to retrieve the core from the central part of the Sokli basin (Helmens et al. 2000; 2007) as was done during the sampling of the lakes in our modern calibration dataset, and to prevent any influence of within-lake variability in chironomid-inferred temperatures as presented in Heiri et al. (2003b).

2.5.2 Lake depth Walker and Cwynar (2006) state that the influence of lake depth on palaeotemperature inferences poses a problem that deserves more attention. Deep, thermally stratified lakes may provide habitats suitable for cold-stenothermous midges, whereas shallow lakes from the same region will not support these species. Palaeotemperature inferences from deep lakes might therefore result in lower temperatures than those inferred from the shallower sites. Both the chironomid- assemblages as well as the botanical macro-remains from the lacustrine sediments of SU7 suggest that a shallow lake was present at the study site throughout the period considered, although the lake might have been relatively deeper during zone S-Ch1. This is also suggested by preliminary diatom results (unpublished data). As no drastic increase in littoral macrophyte taxa such as Carex spp is reconstructed (unpublished data) we exclude a major influence of lake infilling on the lacustrine habitats available for chironomids.

2.5.3 Dispersion Velle et al. (2005) conclude that during the Holocene, mobility may not have been a limiting factor for the distribution of chironomids on the Scandinavian mainland.

38 Chapter 2

However, during OIS-3 there was a limited-sized ice-sheet situated close to our study site, possibly forming a major barrier for dispersal from the west or south. The lower sediment samples from our record (below 6.79 m core depth) show a low Hill’s N2- diversity. However, the samples above 6.79 m core depth show relatively high N2- diversity, suggesting that after 6.79 m core depth the colonisation of this lake has not been a problem for a wide range of taxa. Even a species with a limited mobility such as C. ambigua (Brodersen and Lindegaard 1999) reached the Sokli basin quickly with respect to the introduction of other chironomid taxa. Several taxa that reach high abundances only at a later stage in the record already have single occurrences in the lower parts of the record, before finally establishing themselves during a later stage (see for instance Paracladius). Therefore, local climate or habitat conditions rather than a restricted dispersion have most likely been the controlling factor for the composition of the chironomid assemblages.

2.5.4 Decoupling of water temperature and air temperature Many modern training sets have been designed with the specific goal of providing a tool for reconstructing mean July air temperatures. This however does not mean that the authors assume that the chironomid fauna responds exclusively to air temperature (Birks 1998). In fact, it is likely that chironomids respond to both air and water temperature, as water temperature influences the development of the midges during the relatively long larval stage, whereas air temperature has a direct influence only on the survival and distribution of the winged, short-lived adult stage (Brooks and Birks 2001). Livingstone and Lotter (1998) found a strong correlation between mean July lake water and air temperatures in Switzerland. However, there are several potential decoupling mechanisms between July air temperature and water temperature, for instance, an increased influence of winter precipitation in the form of snow on the temperature of the lake water (Birks and Birks 2006), or the influence of glacier-fed streams on a lake (Brooks and Birks 2001). In order to investigate whether these potential mechanisms played a role in the former lake, or whether past changes in nutrient availability have influenced the chironomid-based temperature inferences, the application of a single proxy will not suffice. Multi-proxy studies providing multiple lines of evidence for possible changes in climate or environment will help to identify factors influencing the composition of fossil chironomid assemblages.

2.5.5 Other proxy-records from the Sokli study-site and comparison with model results In the Sokli- project, botanical and zoological macro-remain analyses were carried out as an independent assessment of local climate conditions. Palaeotemperature estimates were made based on the botanical taxa by using the climate indicator plant species method (sensu Iversen 1954; Kolstrup 1980). Certain plants require certain minimum mean summer temperatures to flower and reproduce and the relationship

39 Chapter 2

Chironomid-Inferred Botanical macro-remain Organic mean July T (°C) minimum mean July T (°C) content (%)

10 11 12 13 14 681012 14 2.0 4.0

5.0 5.0 5.0

5.5 5.5 5.5

6.0 6.0 6.0 Depth (m) 6.5 6.5 6.5

7.0 7.0 7.0

Figure 2.5: Chironomid-inferred mean July air temperature, macro-remain inferred minimum mean July air temperature and organic matter content of the sediment between the geographical limit of plant distribution and temperature can be used to reconstruct past minimum temperatures based on fossil records of plant remains. Macroremains of Potamogeton mucronatus- type are found in Zone S-Ch2, and oogonia of Characeae are abundant in this zone as well. In the remaining part of the sediment sequence, macro-remains of aquatic plants are absent (unpublished data). Preliminary pollen results reveal that throughout the record an aquatic macrophyte vegetation was present consisting of taxa such as Isoetes, Myriophyllum spicatum and Potamogeton. No major trends are visible in the abundances of aquatic plant pollen, and we therefore do not assume major changes in available micro-habitats for chironomids as a result of changing aquatic vegetation. The quantitative temperature record inferred from the botanical macro- remains shows a temperature trend similar to the chironomid-based temperature- reconstruction. The aquatic macro-remains reconstruct the warmest interval to be between 6.7 and 6.4 m core depth (Figure 2.5). The organic content of the sediment is also highest during this part of the record, although variations in organic matter are comparatively low in the entire profile. As the reconstructed temperatures based on botanical macro-remains represent minimum mean July air temperatures instead of the mean July air temperature-estimates that are obtained through the analysis of chironomid remains, the macro-remain based palaeotemperatures are slightly lower throughout the record. A high number of statoblasts of the bryozoan species Fredericella indica is recorded in the interval between 6.7 and 6.4 m core depth as well. Instead of being indicative of a minimum mean July air temperature, it is known from modern

40 Chapter 2 observations that this species is most common in lakes with water temperatures between 11 and 15 ºC (Økland and Økland 2001), which corresponds well with the temperatures inferred from the chironomid remains. Increasing wetland vegetation fringing the lake and deposition of laminated clay and silts may have hampered a further expansion of Fredericella indica after 6.4 m core depth, as it prefers stony shores and avoids lakes with soft sediments and vegetation rich in Sphagnum (Økland and Økland 2001). Because of the presence of a reduced ice sheet over Scandinavia during OIS-3, the atmospheric circulation over Northeast Finland must have been different from the current circulation pattern. Using the ECBilt-CLIO-VECODE coupled atmosphere- ocean-vegetation model, circulation patterns for interstadial conditions during OIS-3 have been reconstructed (Helmens et al. accepted). These climate model runs show a stable circulation pattern in which dry, northerly winds advect dry air to northeast Finland. As summer-insolation was relatively high (even slightly higher than today), and reconstructed soil moisture in the model is low, there is a relatively high amount of solar energy available to heat up the surface. This results in model-based temperature reconstructions in the order of 13 ºC, similar to our reconstructed range of 10.5 – 14 ºC. 2.5.6 Comparison with other NW European sites The climate model runs performed by Helmens et al. (accepted) suggest that atmospheric circulation patterns over the UK were similar to those at the Sokli site, i.e. dry air that was advected from the ice-sheet to the adjacent land. Several terrestrial records registering the early part of OIS-3 are available from central and southern England (Coope 2002). Reconstructed summer temperatures, based on fossil coleopteran (beetle) assemblages, show that the climate during OIS-3 was probably temperate and oceanic in England with mean monthly July temperatures reaching levels at least as warm as those of the present day (Coope 2002). In contrast to the locations southwest (England) and northeast (Sokli) of the former ice-sheet, the model results by Helmens et al. (accepted) indicated that temperatures in locations south of the ice sheet (i.e. Germany) should be lower than those of the present day. Analysis of fossil coleopteran assemblages from the Oerel- site (northern Germany) suggested a mean July temperature of 12°C at this more southerly location (Behre et al. 2005). This cooler temperature is indeed in contrast with the high temperatures reconstructed at Sokli and in the British Isles. In eastern Germany, a location also south of the fossil ice-sheet but located further away from the ice-front than the Oerel-site, lacustrine sediments from the opencast lignite mines of the Niederlausitz area were used to obtain pollen- and macro-remain based minimum mean July temperature estimates (Bos et al. 2001; Kasse et al. 2003). Minimum mean summer temperatures were at least 12-13 °C at this location. This is cooler than the present day temperature at this location (17-18 °C), which is in line with the results from the Oerel site. In conclusion, the same temperature pattern can be derived from the few available records for OIS-3 as has been suggested based on the climate model simulations of Helmens et al. (accepted).

41 Chapter 2

2.6 Conclusions A unique fossil-rich lacustrine deposit, dated to the earlier part of OIS-3, with little post-depositional disturbance, was retrieved from the high-latitudinal Sokli-site. Analysis of this sequence showed changing concentrations of chironomid remains and a dynamic composition of the chironomid assemblages throughout the record: 1) The high number of lacustrine taxa with a preference for littoral to sublittoral habitats suggests that a shallow lake must have existed at the Sokli study site. 2) Using a Norwegian chironomid-based temperature inference model, palaeotemperature estimates were derived from the fossil chironomid- assemblages. The reconstructed mean July air temperatures are in the range of 10.5– 14 ºC, surprisingly high considering the current mean July air temperature of 13 ºC at the Sokli study site. 3) Botanical and zoological macro-remains of aquatic taxa, encountered in the Sokli sediments, provide independent evidence for a period of high (minimum) July air temperatures, in agreement with the chironomid-based inferences. 4) The warm July air temperatures reconstructed for Sokli during OIS-3 are in agreement with modelled results for interstadial conditions in northeast Finland. To our knowledge there exist no similar studies providing quantitative temperature reconstructions for high-latitudinal continental sites in Europe for the earlier part of OIS-3. At the Sokli study site, chironomid-remains provided independent evidence for high summer temperatures in northeast Finland during early OIS-3.

References Barley EM, Walker IR, Kurek J, Cwynar LC, Mathewes RW, Gajewski K, Finney BP (2006) A northwest North American training set: distribution of freshwater midges in relation to air temperature and lake depth. J Paleolimnol 36: 295-314 Bedford A, Jones RT, Lang B, Brooks SJ, Marshall JD (2004) A late-glacial chironomid record from Hawes Water, northwest England. J Quatern Sci 19: 281-290 Behre KE, Hölzer A, Lemdahl J (2005) Botanical macro-remains and insects from the Eemian and Weichselian site of Oerel (northwest Germany) and their evidence for the history of climate. Veg Hist Archaeobot 14: 31–53 Bennett KD (1996) Determination of the number of zones in a biostratigraphical sequence. New Phytol 132: 155- 170 Birks HH, Ammann B (2000) Two terrestrial records of rapid climatic change during the glacial-Holocene transition (14,000 – 9,000 calendar years B.P.) from Europe. Proc Natl Acad Sci 97: 1390-1394 Birks HH, Birks HJB (2006) Multi-proxy studies in palaeolimnology. Veg Hist Archaeobot 15: 235-251 Birks HJB (1995) Quantitative palaeoenvironmental reconstructions. In Maddy D, Brew JJ (Eds) Statistical modelling of Quaternary science data. Quaternary Research Association, Cambridge, pp 161-254 Birks HJB (1998) Numerical tools in palaeolimnology – progress, potentialities, and problems. J Paleolimnol 20: 307-332 Birks HJB, Line JM, Juggins S, Stevenson AC, Ter Braak CJF (1990) Diatoms and pH reconstruction. Philos Trans R Soc Lond B 327, pp 263-278 Bos JAA, Bohncke SJP, Kasse C, Vandenberghe J (2001) Vegetation and Climate during the Weichselian Early Glacial and Pleniglacial in the Niederlausitz, eastern Germany - macrofossil and pollen evidence. J Quatern Sci 16: 269-289

42 Chapter 2

Brodersen KP, Lindegaard C (1999) Mass occurrence and sporadic distribution of Corynocera ambigua Zetterstedt (Diptera, Chironomidae) in Danish lakes. Neo- and palaeolimnological records. J Paleolimnol 22: 41-52 Brooks SJ (2006) Fossil midges (Diptera: Chironomidae) as palaeoclimatic indicators for the Eurasian region. Quatern Sci Rev 25: 1894-1910 Brooks SJ, Birks HJB (2000) Chironomid-inferred Late-glacial air temperatures at Whitrig Bog, south-east Scotland. J Quatern Sci 15: 759-764 Brooks SJ, Birks HJB (2001) Chironomid-inferred air temperatures from Lateglacial and Holocene sites in north- west Europe: progress and problems. Quatern Sci Rev 20: 1723-1741 Cane MA, Braconnot P, Clement A, Gildor H, Joussaume S, Kageyama M, Khodri M, Paillard D, Tett S, Zorita E (2006) Progress in paleoclimate modeling. J Climate 19: 5031-5057 Coope GR (2002) Changes in the Thermal Climate in Northwestern Europe during Marine Oxygen Isotope Stage 3, Estimated from Fossil Insect assemblages. Quatern Res 57: 401-408 Cubasch U, Meehl G, Boer GJ, Stouffer RJ, Dix M, Noda A, Senior CA, Raper S, Yap KS (2001) Projections of Future Climate Change. In: Houghton JT, Ding Y, Griggs DJ, Noguer M, Van der Linden PJ, Dai X, Maskell K, Johnson CA (Eds) Climate Change 2001: The scientific basis. Contribution of working group I to the third assessment report of the intergovernmental panel on climate change. Cambridge University Press, Cambridge, pp 525 – 582 Dansgaard W, Johnsen SJ, Clausen HB, Dahl-Jensen D, Gundestrup NS, Hammer CU, Hvidberg CS, Steffensen JP, Sveinbjörndottir AE, Jouzel J, Bond G (1993) Evidence for general instability of past climate from a 250-kyr ice-core record. Nature 364: 218-220 De Beaulieu J-L, Reille M (1992). The last Climatic Cycle at La Grande Pile (Vosges, France) a new pollen profile. Quatern Sci Rev 11: 431-438 Drebs A, Nordlund A, Karlsson P, Helminen J, Rissanen, P (2002) Climatological statistics of Finland 1971-2000. Finnish Meteorological Institute, Helsinki, pp 99 Eggermont H, Heiri O, Verschuren D (2006) Fossil Chironomidae (Insecta:Diptera) as quantitative indicators of past salinity in African Lakes. Quatern Sci Rev 25: 1966-1994 Grimm EC (1991–2004) TILIA, TILA.GRAPH, and TGView. Illinois State Museum, Research and Collections Center, Springfield, USA http://demeter.museum.state.il.us/pub /grimm/ Heiri O, Lotter AF (2005) Holocene and Lateglacial summer temperature reconstruction in the Swiss Alps based on fossil assemblages of aquatic organisms: a review. Boreas 34: 506-516 Heiri O, Millet L (2005) Reconstruction of Late Glacial summer temperatures from chironomid assemblages in Lac Lautrey (Jura, France). J Quatern Sci 20: 33–44 Heiri O, Lotter AF, Hausmann S, Kienast F (2003a) A chironomid-based Holocene summer air temperature reconstruction from the Swiss Alps. Holocene 13: 477-484 Heiri O, Birks HJB, Brooks SJ, Velle G, Willassen E (2003b) Effects of within-lake variability on fossil assemblages on quantitative chironomid-inferred temperature reconstructions. Palaeogeogr Palaeoclimatol Palaeoecol 199: 95-106 Heiri O, Ekrem T, Willassen E (2004) Larval head capsules of European Micropsectra, Paratanytarsus and Tanytarsus (Diptera: Chironomidae: Tanytarsini). Version 1.0. http://www.bio.uu.nl/~palaeo/ Chironomids/Tanytarsini/intro.htm Helmens KF, Räsänen ME, Johansson PW, Jungner H, Korjonen K (2000) The Last Interglacial-Glacial cycle in NE Fennoscandia: a nearly continuous record from Sokli (Finnish Lapland). Quatern Sci Rev 19: 1605-1623 Helmens KF, Johansson PW, Räsänen ME, Alexanderson H, Eskola KO (2007) A series of ice-free intervals continuing into MIS 3 recorded in the central area of Fennoscandian glaciation: late Quaternary climate- stratigraphy at Sokli. Bull Geol Soc Finl 79: 17-39 Helmens KF, Bos JAA, Engels S, Van Meerbeeck C, Bohncke SJP, Renssen H, Heiri O, Brooks SJ, Seppä H, Birks HJB, Wohlfarth B (accepted) Ice-free conditions and present-day temperatures during the last glacial at 50ka in the central area of the Scandinavian glaciations. Accepted for publication in Geology Hill MO (1973) Diversity and evenness: a unifying notation and its consequences. Ecology 54: 427-432 Huijzer B, Vandenberghe J (1998) Climatic reconstruction of the Weichselian Pleniglacial in northwestern and central Europe. J Quatern Sci 13: 391 - 417 Ilyashuk EA, Ilyashuk BP, Hammarlund D, Larocque I (2005) Holocene climatic and environmental changes inferred from midge records (Diptera: Chironomidae, Chaoboridae, Ceratopogonidae) at Lake Berkut, southern Kola Peninsula, Russia. Holocene 15: 897-914 Iversen J (1954) The Late Glacial flora of Denmark and its relation to climate and soil. Danm Geol Undersøg 2: 88 119 Johnsen SJ, Clausen HB, Dansgaard W, Fuhrer K, Gundestrup N, Hammer CU, Iversen P, Jouzel J, Stauffer B, Steffensen JP (1992) Irregular glacial interstadials recorded in a new Greenland ice core. Nature 359: 311-313 Juggins S (2003) C2 User guide. Software for ecological and palaeoecological data analysis and visualisation. University of Newcastle, Newcastle upon Tyne, UK Kageyama M, Laine A, Abe-Ouchi A, Braconnot P, Cortijo E, Crucifix M, De Vernal A, Guiot J, Hewitt CD, Kitoh A, Kucera M, Marti O, Ohgaito R, Otto-Bliesner B, Peltier WR, Rosell-Mele A, Vettoretti G, Weber SL, Yu

43 Chapter 2

Y, MARGO Project Members (2006) Last Glacial Maximum temperatures over the North Atlantic, Europe and western Siberia: a comparison between PMIP models, MARGO sea-surface temperatures and pollen- based reconstructions. Quatern Sci Rev 25: 2082-2102 Kasse C, Vandenberghe J, van Huissteden J, Bohncke SJP, Bos JAA (2003) Sensitivity of Weichselian fluvial systems to climate change (Nochten mine, eastern Germany). Quatern Sci Rev 22: 2141-2156 Kolstrup E (1980) Climate and stratigraphy in Northwestern Europe between 30,000 BP and 13,000 BP, with special reference to The Netherlands. Meded Rijks Geol Dienst 32: 181-253 Larocque I, Hall RI (2003) Chironomids as quantitative indicators of mean July air temperature: validation by comparison with century-long meteorological records from northern Sweden. J Paleolimnol 29: 475-493 Livingstone DM, Lotter AF (1998) The relationship between air and water temperatures in lakes of the Swiss Plateau: a case study with palaeolimnological implications. J Paleolimnol 19: 181-198 Lotter AF, Juggins S (1991) POLPROF, TRAN and ZONE: programs for plotting, editing and zoning pollen and diatom data. Inqua-Subcommission for the study of the Holocene, Working Group on Data-Handling Methods, Newsletter 6: 4-6 Magny M, Aalbersberg G, Bégeot C, Benoit-Ruffaldi P, Bossuet G, Disnar JR, Heiri O, Laggoun-Defarge G, Mazier F, Millet L, Peuron O, Vannière B, Walter-Simmonet AV (2006) Environmental and climatic changes in the Jura mountains (eastern France) during the Lateglacial–Holocene transition: a multi-proxy record from Lake Lautrey. Quatern Sci Rev 25: 414-445 Makarchenko EA, Makarchenko MA (1999) Chironomidae. In: Key to Freshwater Invertebrates of Russia and Adjacent Lands. In: Tsalolikhin SJ (Ed.) Zoological Institute RAS, St. Petersburg, pp 670-857 Moller Pillot HKM (1984) De larven der nederlandse Chironomidae (Diptera) (Inleiding, Tanypodinae & Chironomini). Nederlandse Faunistische Mededelingen 1A Økland KA, Økland J (2001) Freshwater bryozoans (Bryozoa) of Norway II: distribution and ecology of two species of Fredericella. Hydrobiologia 459: 103-123 Oliver DR, Roussel ME (1983) The Insects and Arachnids of Canada, Part 11: The Genera of Larval Midges of Canada-Diptera: Chironomidae. Agriculture Canada Publication 1746, 263 pp Quinlan R, Smol JP (2001) Chironomid-based inference models for estimating end-of-summer hypolimnetic oxygen from south-central Ontario shield lakes. Freshw Biol 46: 1529-1551 Rieradevall M, Brooks SJ (2001) An identification guide to subfossil Tanypodinae larvae (Insecta: Diptera: Chironomidae) based on cephalic setation. J Paleolimnol 25: 81-99 Schmid PE (1993) A key to the larval Chironomidae and their instars from Austrian danube region streams and rivers with particular reference to a numerical taxonomic approach. Part I. Diamesinae, Prodiamesinae and Orthocladiinae. Wasser und Abwasser Supplement 3/93, 514 pp Ter Braak CJF, Juggins S (1993) Weighted averaging partial least squares regression (WA-PLS): an improved method for reconstructing environmental variables from species assemblages. Hydrobiologia 269/270: 485-502 Ter Braak CJF, Šmilauer P (1998) CANOCO reference manual and user’s guide to Canoco for windows. Wageningen, Centre for Biometry Wageningen, 352 pp Van der Hammen Th, Maarleveld GC, Vogel JC, Zagwijn W (1967) Stratigraphy, climatic succession and radiocarbon dating of the last glacial in the Netherlands. Geol Mijnb 46: 79-95 Vandenberghe J, Coope R, Kasse C (1998) Quantitative reconstructions of palaeoclimates during the last interglacial-glacial in western and central Europe: as introduction. J Quatern Sci 13: 361-366 Velle G, Brooks SJ, Birks HJB, Willassen E (2005) Chironomids as a tool for inferring Holocene climate: an assessment based on six sites in southern Scandinavia. Quatern Sci Rev 24: 1429-1462 Voelker AHL, workshop participants (2002) Global distribution of centennial-scale records for Marine Isotope Stage (MIS) 3: a database. Quatern Sci Rev 21: 1185-1212 Walker IR, Cwynar LC (2006) Midges and palaeotemperature reconstruction – the North American experience. Quatern Sci Rev 25: 1911-1925 Walker IR, Smol JP, Engstrom DR, Birks HJB (1991) An assessment of Chironomidae as Quantitative Indicators of Past Climatic Change. Can J Fish Aquat Sci 48: 975- 987 Wiederholm T (1983) Chironomidae of the Holarctic region. Keys and diagnoses. Part I. Larvae. Entomol Scand 19 Woodward CA, Shulmeister J (2006) New Zealand chironomids as proxies for human-induced and natural environmental change: Transfer functions for temperature and lake production (chlorophyll a). J Paleolimnol 36: 407-429

44 Chapter 3

Chapter 3

Present-day temperatures in northern Scandinavia during the Last Glaciation

Helmens KF, Bos JAA, Engels S, Van Meerbeeck CJ, Bohncke SJP, Renssen H, Heiri O, Brooks SJ, Seppä H, Birks HJB, Wohlfarth B

Abstract Scandinavia is generally considered to have been covered extensively with ice throughout Marine Isotope Stages (MIS) 4–2 between 75 and 10 ka (103 years). Here we present evidence for ice-free and warm conditions in the central area of the Scandinavian glaciations during MIS 3. Our multi-proxy data obtained from a lacustrine sequence in northern Finland reveal not only significant response in the northeastern sector of the Scandinavian Ice Sheet to warming during the early part of MIS 3, but also indicate rapid climate warming to present-day temperatures in this ice-free period. New climate-model simulations for MIS 3 interstadial conditions confirm the high mean July temperatures northeast of the Scandinavian Ice Sheet in response to the presence of the ice sheet and high insolation values during MIS 3.

This manuscript was published as Geology 35 (11): 987-990 DOI: 10.1130/G23995A.1

45 Chapter 3

3.1 Introduction High climate variability during the last glacial cycle shown in oxygen-isotope records in Greenland ice cores (Johnsen et al. 2001) and foraminifera records in North Atlantic deep-sea sediments (Bond et al. 1993) is characterized by frequent rapid shifts from cold stadial to warm interstadial conditions referred to as Dansgaard-Oeschger (D/O) oscillations (Dansgaard et al. 1993). High-frequency climate variability on the adjacent European continent has been inferred from long pollen records in the Mediterranean region (Allen et al. 1999), high-resolution sediment studies of maar lake sequences in Germany (Sirocko et al. 2005), and speleothem records in the Austrian Alps (Holzkämper et al. 2004). The Mediterranean records mostly reflect significant changes in precipitation, whereas the absence of permafrost as indicated by stalagmite formation at a high altitude-site in Austria suggests near present-day temperature conditions during the early part of MIS 3. The generally discontinuous nature of sedimentation and repeated erosion combined with poor dating control and a scarcity of high-resolution records, however, presently hamper a detailed study of the Last Glacial climate in Europe. As such, the magnitude of climate change on the continent through the last glacial cycle remains mostly unknown. An extensive database of geological sections and absolute age determinations along the mountainous Norwegian coast has recently indicated rapid phases of ice retreat, reaching to inland areas, along the western margin of the Scandinavian Ice Sheet during the late part of MIS 3 and MIS 2 (Olsen et al. 2001). In this paper, we present evidence for significant ice front retreat along the northeastern (continental) sector of the Scandinavian Ice Sheet during the early part of MIS 3. Quantification of the associated climate warming is made through a multi-proxy study on lacustrine sediments formed in northern Finland during deglaciation. A climate model experiment with a MIS-3 interstadial set-up is performed in order to test our proxy- based climate reconstructions.

3.2 Material and Methods The sediments from the Sokli B-series borehole (Figure 3.1), northern Finland (lat. 67º48’N, lon. 29º18’E, elev. 220 m a.s.l; Figure 3.3), form part of an unusually long and continuous sedimentary sequence of tills, glacio-fluvial and fluvial beds, inter- layered with a series of fossil-rich lacustrine sediments that reach from the present into the Late Saalian (MIS1–6, i.e., representing the last ca. 130 ka; Helmens et al. 2007). The Sokli sediments have been protected from erosion by Late Quaternary Scandinavian Ice Sheets due to their sheltered position in a steep depression. This depression occurs in relatively soft, highly fractured and deeply weathered rocks of a carbonate-rich magma intrusion in the crystalline Precambrian Shield. A first study of the entire Sokli sequence has indicated that the non-glacial sediment intercalations each have a characteristic lithological and palynological content, and represent individual and successive developments both in terms of depositional and vegetational change (Helmens et al. 2000; 2007). The Sokli B-series borehole is dated by independent AMS 14C and optically stimulated luminescence

46 Chapter 3

(OSL) dating (Helmens et al. 2007) and stratigraphic dating based on correlation with the deep-sea record (Figure 3.1). In the present study, a high-resolution analysis of a large variety of fossil remains was carried out on a lacustrine interval of laminated silts/clays at 5–7 m depth dated to the early part of MIS 3. This ice-free interval at Sokli has been defined as the Tulppio Interstadial (Helmens et al. 2007). A detailed lithological column and a selection of proxy-indicators that reflect local environmental changes and changes in regional vegetation are given in Figure 3.2. Different proxies were analyzed from the same samples. The frequencies of microfossils were calculated as percentages of the sum of tree, shrub, dwarf shrub and herb pollen (POLLEN SUM between ca. 275–400). The count sum of head capsules of chironomids ranged from 54 to 199. Counts were lower in the lowermost part of the profile (below 6.75 m core depth) and these samples were therefore lumped. Counts of macrofossils of plants and other aquatic animals were made in 10 and 20 cc samples. Mean July air temperatures were reconstructed by applying transfer functions based on modern-day animal/plant-climate calibration sets from northern Europe to the pollen (Seppä and Birks 2001) and chironomid records (Brooks and Birks 2001; right side of Figure 3.2). The sample specific prediction errors for the chironomid- based temperature inferences, as estimated through bootstrap cross-validation, range between ± 1.1–1.2 °C. The average error for the pollen-based inferences is ± 1 °C. Minimum mean July air temperatures were furthermore estimated based on aquatic plant fossils and detailed information of the present distribution of these species in northern Europe (Kolstrup 1980). Pollen of Myriophyllum spicatum and macrobotanical remains of Potamogeton mucronatus were used to infer minimum mean July temperatures of 10ºC (Kolstrup, 1980) respectively 13ºC (Brinkkemper et al. 1987) in Figure 3.2. We applied the LOVECLIM 3D climate model (Driesschaert 2006) to simulate the average MIS 3 interstadial climate. The extent and elevation of the continental ice sheets and other forcings (orbital paramaters, atmospheric greenhouse gases and dust content) are modified from full glacial conditions (Roche et al. 2006). The Sokli site is presently situated in the northern boreal forest (Helmens et al. 2000) and the climate is cold temperate with a mean July temperature of ca. 13 °C and mean February temperature of ca. ~14 °C. Mean annual precipitation is ca. 500 mm.

3.3 Results Macrofossils are scarce in the lowermost part of the silt/clay sequence but microfossils of algae indicate local aquatic conditions (local zone C1a in Figure 3.2). A more distinct lake environment fringed by wetland vegetation is recorded in the finely laminated sediments of zone C1b. However, it is not until zone C2 that a large diversification of aquatic life can be observed, reflected by, among others, the abundance of chironomids and the moss animal (Bryozoa) Fredericella indica. Chironomid-based temperature inferences indicate for zone C2 warming to present-

47 Chapter 3 day mean July temperatures (ca. 13–14 ± 1 °C). Elevated minimum mean July temperatures of 13 °C are also indicated by findings of macroremains of the aquatic plant Potamogeton mucronatus (Brinkkemper et al. 1987) in zone C2. Fredericella indica is presently reported to prefer mean summer water temperatures from 11 to 15 °C (Økland and Økland 2001). The chironomid record subsequently indicates decreasing mean July temperatures to ca. 11 ± 1 °C, which coincides with a fall in the abundance and diversity of aquatic life during zone C3. Although temperatures stabilized during zone C4, it is only at this latter stage that we register a distinct response to cooling in the lake catchment. Records of the soil fungi Glomus and Cenococcum geophilum indicate an increase in soil erosion (Van Geel et al. 1989; Walker et al. 2003), whereas an increased influence of the nearby ice-sheet through proglacial streams is suggested by the presence of various stream-inhabiting chironomids (e.g., Eukiefferiella-type, Prodiamesa and Rheotanytarsus). As a result, sandy silts accumulated in the lake. Eventually, a re-advancing ice-margin overriding the vegetated land is suggested by the findings of a dropstone and pieces of wood in the uppermost part of zone C4, and the final deposition of basal till (litho-facies unit D in Figure 3.2). The pollen assemblage in this lacustrine interval and the relatively warm pollen-inferred mean July temperature(s) of ~11 ± 1 °C (Figure 3.1) are similar to those presently recorded in the shrub tundra region close to the sub-arctic birch forest limit in coastal northern Norway. However, the distinct warming to present-day mean July temperatures, shown in our chironomid record for zone C2 only, did not result in the local immigration of pine or spruce, i.e., trees that presently form important components of the boreal forest at Sokli. This indicates that the latter warming must have been rapid, only allowing time for fast migrating aquatic fauna and flora to respond. In agreement with our proxy-based reconstructions, our climate model simulation suggests high mean July temperatures for the sector northeast of the Scandinavian Ice Sheet (Figure 3.3). In the model, warm summer conditions are the combined result of enhanced July insolation compared to present (+19 W/m2, Berger and Loutre 1991) and northwesterly winds advecting cool, but very dry air from the ice sheet. The winds are produced by a thermal high pressure cell over the ice sheet which is an extension of the anticyclonal cell situated over the cool northeastern Atlantic Ocean. The combination of high insolation and dry air leads to a strong sensible heat flux and relatively warm conditions near the surface.

3.4 Discussion Although absolute and stratigraphic age determinations indicate that this lacustrine sequence has an early MIS 3 age (Figure 3.1), the large error limits on both the optical (OSL) and 14C dates (Figure 3.2) preclude a direct correlation of this warm interval with the interstadial sequence (IS) in the Greenland record. However, it is arguable that thinning and gradual retreat of the northeastern margin of the Scandinavian Ice Sheet from its MIS 4 limit in Russia (Svendsen et al. 2004) started during IS 16 in the

48 Chapter 3

Sokli B-series (MIS)

stages 3 yr) Istope

Age (x 10Depth (m)Lithology Main vegetationMarine types Depositional environments 0 Northern boreal mire 1 forest lacustrine coring gap (data from other coring --(12)-- Permanent 2/3 fluvial/ lacustrine ice-cover fluvial 5 54 ± 7/4 glacio-fluvial shrub glacial tundra 3 1 48 ± 16 Permanent --(60)-- 1 4 74 ± 26 ice-cover --(74)-- detail in 10 sub-arctic 1 birch forest 5a Fig. 3.2 80 ± 26 limit --(85)-- permanent 5b ice-cover 1 --(93)-- 94 ± 16 Dating techniques 1 94 ± 19 15 birch-pine 5c 54 ± 7/4 AMS 14C forest uncalibrated --(105)-- 48 ± 16 OSL (1 =quartz, open tundra 2 =feldspar) 5d 20 2 119 ± 15 --(115)--

northern boreal 5e forest 25 (age x 103 yr) y d el la v silt *based on pollen c san ra

g data from :parallel cores

Figure 3.1: Simplified lithology of the Sokli B-series borehole is plotted against depth. Changes in depositional environments and main regional vegetation types are shown back to the Last Interglacial. The independent AMS 14C/OSL chronology for the Sokli B-series borehole is indicated to the left, and correlation with the deep-sea record (Martinson et al. 1987) to the right.

49 Chapter 3 earliest part of MIS 3, but that only the prominent IS 14 around 53 ka lasted long enough, and was warm enough, for the Sokli area to become deglaciated and the dramatic D/O warming event to be registered in the fossil record. Our warm early MIS 3 eastern Scandinavian summer conditions are in strong contrast with cool summer temperatures shown in proxies at Oerel (northern Germany) directly south of the ice sheet. A ca. 12 °C mean July temperature was reconstructed at this more southerly location (Behre et al. 2005). In our model simulation, the latter region is relatively cold in agreement with the palaeotemperature data from Oerel. In the model, these low temperatures are due to the prescribed land-ice configuration at this location, in combination with cold northerly winds descending from the center of the Scandinavian Ice Sheet. However, in early MIS 3, Oerel was ice-free, implying that the model exaggerates the cool conditions here. Interestingly, near present-day conditions are recorded for MIS 3 by proxy data from the UK (Coope 1987). As for our Sokli site, we ascribe this to higher insolation in combination with the altered atmospheric circulation.

3.5 Conclusion Ice-free conditions in northern Finland facilitated the deposition of laminated lacustrine sediments during early MIS 3. A multi-proxy based reconstruction of climate and environmental parameters shows warming to present-day temperatures. Our climate model simulation for MIS 3 interstadial conditions confirms the warm summer conditions and indicates that they result from high insolation in combination with dry winds blowing from the Scandinavian Ice Sheet. Our study not only reveals a highly dynamic Scandinavian Ice Sheet during the Last Glacial, but also demonstrates a clear response by biotic proxies to rapid climate warming.

References Allen JRM, Brandt U, Brauer A, Hubberten H-W, Huntley B, Keller J, Kraml M, Mackensen A, Mingram J, Negendank JWF, Nowaczyk NR, Oberhänsli H, Watts WA, Wulf S, Zolitschka B (1999) Rapid environmental changes in southern Europe during the last glacial period: Nature: 740–743 Behre KE, Hölzer A, Lemdahl J (2005) Botanical macro-remains and insects from the Eemian and Weichselian site of Oerel (northwest Germany) and their evidence for the history of climate. Veg Hist Archaeobot 14: 31–53 Berger A, Loutre MF (1991) Insolation values for the climate of the last 10 million years. Quatern Sci Rev: 297–317 Bond G, Broecker W, Johnsen SJ, McManus J, Labeyrie L, Jouzel J, Bonani G (1993) Correlations between records from North Atlantic sediments and Greenland ice, Nature 365: 143–147 Brinkkemper O, Van Geel B, Wiegers J (1987) Palaeoecological study of a Middle-Pleniglacial deposit from Tilligte, The Netherlands. Rev Palaeobot Palynol 51: 235–269 Brooks SJ, Birks HJB (2001) Chironomid-inferred air temperatures from Lateglacial and Holocene sites in north- west Europe: progress and problems. Quatern Sci Rev 20: 1723-1741 Coope GR (1987) Fossil beetle assemblages as evidence for sudden and intense climatic changes in the British Isles during the last 45,000 years. In Berger WH, Labeyrie LD (eds) Abrupt Climatic Changes. Reidel, Dordrecht, pp147–150 Dansgaard W, Johnsen SJ, Clausen HB, Dahl-Jensen D, Gundestrup NS, Hammer CU, Hvidberg CS, Steffensen JP, Sveinbjörndottir AE, Jouzel J, Bond G (1993) Evidence for general instability of past climate from a 250-kyr ice-core record. Nature 364: 218-220

50 Chapter 3

Driesschaert E (2006) Climate change over the next millennia using LOVECLIM, a new Earth system model including the polar ice sheets. PhD thesis, Louvain-la-Neuve, Université Catholique de Louvain-la-Neuve, 214 pp Helmens KF, Räsänen ME, Johansson PW, Jungner H, Korjonen K (2000) The Last Interglacial-Glacial cycle in NE Fennoscandia: a nearly continuous record from Sokli (Finnish Lapland). Quatern Sci Rev 19: 1605-1623 Helmens KF, Johansson PW, Räsänen ME, Alexanderson H, Eskola KO (2007) A series of ice-free intervals continuing into MIS 3 recorded in the central area of Fennoscandian glaciation: late Quaternary climate- stratigraphy at Sokli. Bull Geol Soc Finl 79: 17-39 Holzkämper S, Mangini A, Spötl C, Mudelsee M (2004) Timing and progression of the Last Interglacial derived from a high alpine stalagmite. Geophys Res Lett 31: L07201 Johnsen SJ, Dahl-Jensen D, Gundestrup N, Steffensen JP, Clausen HB, Miller H, Masson-Delmotte V, Sveinbjörnsdottir AE, White J (2001) Oxygen isotope and palaeotemperature records from six Greenland ice-core stations: Camp Century, Dye-3, GRIP, GISP2: Renland and NorthGRIP. J Quatern Sci 16: 299–307 Kolstrup E (1980) Climate and stratigraphy in Northwestern Europe between 30,000 BP and 13,000 BP, with special reference to The Netherlands. Meded Rijks Geol Dienst 32: 181-253 Martinson DG, Pisias WG, Hays JD, Imbrie J, Moore TC Jr, Shackleton NJ (1987) Age dating of the orbital theory of the Ice Ages: development of a high-resolution 0 to 300,000-year chronostratigraphy. Quatern Res 27: 1–29 Økland KA, Økland J (2001) Freshwater bryozoans (Bryozoa) of Norway II: distribution and ecology of two species of Fredericella. Hydrobiologia 459: 103-123 Olsen L, Sveian H, Bergstrøm B (2001) Rapid adjustments of the western part of the Scandinavian Ice Sheet during the Mid and Late Weichselian - a new model. Norsk Geol Tidsskr 81: 93–118 Roche DM, Dokken TM, Goosse H, Renssen H, Weber SL (2006) Climate of the last glacial maximum: sensitivity studies and model-data comparison with the LOVECLIM coupled model. Clim Past Discuss 2: 105–1153 Seppä H, Birks HJB (2001) mean temperature and annual precipitation trends during the Holocene in the Fennoscandian tree-line area: pollen-based climate reconstructions. Holocene 11: 527–539 Sirocko F, Seelos K, Schaber K, Rein B, Dreher F, Diehl M, Lehne R, Jäger K, Krbetschek M, Degering D (2005) A late Eemian aridity pulse in central Europe during the last glacial inception. Nature 436: 833-836 Svendsen JI, Alexanderson H, Astakhov VI, Demidov I, Dowdeswell JA, Funder S, Gataullin V, Henriksen M, Hjort C, Houmark-Nielsen M, Hubberten HW, Ingólfsson Ó, Jakobsson M, Kjær KH, Larsen E, Lokrantz H, Pekka Lunkka J, Lyså A, Mangerud J, Matiouchkov A, Murray A, Möller P, Niessen F, Nikolskaya O, Polyak L, Saarnisto M, Siegert C, Siegert MJ, Spielhagen RF, Stein R (2004) Late Quaternary ice sheet history of northern Eurasia. Quatern Sci Rev 23: 1229–1271 Van Geel B, Coope GR, Van der Hammen T (1989) Palaeoecology and stratigraphy of the Lateglacial type section at Usselo (The Netherlands). Rev Palaeobot Palynol 60: 25–38 Walker MJC, Coope GR, Sheldrick C, Turney CSM, Lowe JJ, Blockley SPE, Harkness DD (2003) Devensian Lateglacial environmental changes in Britain: a multi-proxy record from Llanilid, South Wales, UK. Quatern Sci Rev 22: 475–520

51 Chapter 3

C

14

Multi-proxy records from the

Figure 3.2: Sokli lacustrine sediments of the early part of MIS 3 at ~50 ka. A) Lithological column for depth interval 4.90–8.40 m in borehole Sokli B-series and absolute age determinations. Litho-facies units A and D represent glacial till, unit B proglacial gravels and sands fining upward into sands inter-layered with silts, and C laminated lacustrine silts. The AMS dating (UtC nr. 14712) was carried out on terrestrial plant macrofossils. B) Selected proxy-indicators from the multi- proxy analysis that reflect local environmental changes and changes in regional vegetation during local zones C1 to C4. C) Curves of pollen- and chironomid (aquatic insect)-inferred mean July air temperatures, with a Loess smoother (span = 0.2) superimposed on the inferred values. The chironomid- based temperature curve additionally shows minimum mean July temperature values as indicated by specific aquatic plant taxa (thin blue line). The present ca. 13 °C mean July temperature at the Sokli site is indicated for reference.

°C 13 13 (present) (present) mean July mean 11 temperature Minimum mean mean Minimum July temperature Chironomid-inferred 9 (based on aquatic plants) on aquatic (based

°C

13 (present) (present) 11 mean July mean temperature temperature 9 finely

coarsely laminated Pollen-inferred

c) s

herbs dwarf shrubs 50

(%)

Main ting chironomid ting (pine) Pinus sylvestris Betula pubescens/pendula (tree birch) (tree pubescens/pendula Betula

shrubs

geophilum Diagram Pollen 0 trees 100 orange orange coloring green coloring green

Stream-inhabi

coccum )

Ceno

0-2 (% 0-2

d.)

omus se

lla indica lla

Gl

cc

(%) 10

/

0-2

+ + nts + + Frederice + + + + + + + + +

+ + + + + + + + + + + + + + +

0-2 (cou 0-2 wood fragments clast gyttja re-deposited re-deposited debris organic

plants .)

Chironomid concentration Chironomid abundant 40

d

frequent se

0

present /gr

20 nts Wetland

(%) 0 (cou

gae Al 20 silt layering silt sand component 0 gravel gravel

5

•

content Organic (%) •

(%) b) ne zo cal Lo 0

b a a

C4 C3 C2

gravel

C1

Sand Silt •

•

• •

• • Clay

Lithology • • •

• •

unit es thofaci Li

D D C B A

ial nterstad I io pp Tul atographic unit atographic str ial

ial

1 mtae mtae li C 2 ad ad

C St

St

• • 16 ± 48 14

4 7/ ± 54

yr) a) 3

OSL Age AMS AMS (10 uncalibrated

5.00 5.20 5.60 5.80 6.00 6.20 6.40 6.60 6.80 7.00 7.50 8.00 5.40

Sokli B-series borehole (northern Finland) (northern borehole B-series Sokli Depth (m) Depth

52 Chapter 3

°

°

°

°

°

°

°

°

°

°C

13 13 ° (present) (present) mean July mean 11

temperature ° Minimum mean mean Minimum July temperature Chironomid-inferred 9 (based on aquatic plants) on aquatic (based ° ° ° ° ° ° ° °° ° ° ° ° ° °

°C

13

(present) (present) °C 11 mean July mean temperature temperature 9 finely coarsely laminated Pollen-inferred

c) s herbs Figure 3.3: Simulated mean July surface air temperature over the northern and dwarf shrubs 50

(%) middle part of Europe. Contour lines represent the geopotential height at 800

Main ting chironomid ting (pine) Pinus sylvestris

Betula pubescens/pendula (tree birch) (tree pubescens/pendula Betula hPa, while arrows give a schematic view of wind direction in the lower atmosphere. shrubs

geophilum Diagram Pollen

0 trees 100 The extent of the prescribed MIS3 Scandinavian Ice Sheet is indicated in gray and orange orange coloring green coloring green

Stream-inhabi the actual location of Sokli is represented by a black star. It should be noted that coccum

) the model is not suitable for analyzing results at the grid-cell scale because of the

Ceno

0-2 (% 0-2

d.)

omus se lla indica lla

low spatial resolution. Rather the model is used for (sub)continental-scale studies. Gl

cc

(%) 10

/

0-2

+ + nts + + Frederice + + + + + + + + +

+ + + + + + + + + + + + + + +

0-2 (cou 0-2 wood fragments clast gyttja re-deposited re-deposited debris organic

plants .)

Chironomid concentration Chironomid abundant 40 d

frequent se

0 present /gr

20 nts Wetland

(%) 0 (cou

gae Al 20 silt layering silt sand component 0 gravel gravel

5

•

content Organic (%) •

(%) b) ne zo cal Lo 0

b a a

C4 C3 C2 gravel

C1

Sand Silt •

•

• •

• • Clay

Lithology • • •

• •

unit es thofaci Li

D D C B A

ial nterstad I io pp Tul atographic unit atographic str ial ial

1 mtae mtae li C 2 ad ad

C St

St

• • 16 ± 48 14

4 7/ ± 54

yr) a) 3

OSL Age AMS AMS (10 uncalibrated

5.00 5.20 5.60 5.80 6.00 6.20 6.40 6.60 6.80 7.00 7.50 8.00 5.40

Sokli B-series borehole (northern Finland) (northern borehole B-series Sokli Depth (m) Depth

53 Chapter 3

54 Chapter 4

Chapter 4

Rapid climatic events as recorded in Middle Weichselian thermokarst lake sediments

SJP Bohncke, JAA Bos, S Engels, O Heiri and C Kasse

Abstract From a Middle Weichselian sediment sequence in the opencast brown coal mine of Reichwalde (eastern Germany), a ~40 cm thick thermokarst gyttja deposit has been sampled. The AMS 14C dates, although at the limit of detection, indicate an early Middle Weichselian age of the gyttja. Pollen, botanical and zoological (e.g. chironomids) macroremains have been analyzed. Botanical and chironomid taxa indicate warm climatic conditions in the bottom part of the sequence. For this lower part the botanical data suggest a minimum mean July temperature of 12-14ºC. Following this, a cooling is indicated, coinciding with an increased clastic deposition in the lake. A return to permafrost conditions is reconstructed for the upper part of the sequence. The combined evidence strongly suggests a degradation of permafrost due to increased warming in response to a D/O event as a forcing factor for the thermokarst lake formation.

This manuscript is In Press for Quaternary Science Reviews

55 Chapter 4

4.1 Introduction The Weichselian Early, Middle and Late Pleniglacial have been correlated respectively with oxygen isotope stages 4, 3 and 2 of the marine δ18O record (Martinson et al. 1987; Behre and van der Plicht 1992; Table 4.1). Investigations of the Greenland ice cores over this interval revealed many rapid climate oscillations in the oxygen isotope record (Johnsen et al. 1992; Dansgaard et al. 1993), the so-called Dansgaard/Oeschger (D/O) events. A D/O event typically starts with an abrupt warming of Greenland by 5-10ºC over a few decades or less, followed by a gradual cooling over several hundred to more than a thousand years and often ends with an abrupt final reduction of temperature back to cold (stadial) conditions (Ganopolski and Rahmstorf 2001). Climate studies based on marine cores from the Atlantic Ocean (Bond et al. 1993) also have shown these short-term climatic fluctuations. The period between the successive D/O events is most often around 1,500-years or a multiple thereof (Bond et al. 1997; 1999; Grootes and Stuiver 1997). However, the majority of these climate oscillations are not reflected in the NW European continental palynological record, except at more southerly locations, where interstadial periods seem to have been registered better in pollen records, e.g. Les Echets, La Grand Pile and the Velay region (e.g. Guiot et al. 1989; Reille and De Beaulieu 1990; De Beaulieu and Reille 1984; 1992; Reille et al. 2000). In the northwestern European terrestrial record only three to five interstadials have been recognized during the Weichselian Pleniglacial (e.g. van der Hammen et al. 1967; van der Hammen 1971; Zagwijn 1974; Kolstrup and Wijmstra 1977; Vandenberghe 1985; Ran 1990; Behre and van der Plicht 1992). Correlation between the ice core record and the terrestrial botanical record therefore remains unclear. Based on the ice core-records it is likely that there were intervals during the Pleniglacial when climate in NW Europe was suitable for the development of boreal forests. However, pollen and macroremain analyses have shown that such forests were not present (e.g. Kolstrup 1990; Bos et al. 2001). This absence seems to conflict with the temperature regime as reconstructed from pollen, macroremains and beetles and often has been the subject of speculations (e.g. Coope 1975; 2000; Kolstrup and Wijmstra 1977; Kolstrup 1990; Van Geel 1996; Ran 1990; Bos et al. 2001; 2004). A variety of factors has been suggested to explain the virtual absence of trees in NW Europe during the Pleniglacial: 1) too dry and continental climate, 2) wind stress, 3) heavy grazing pressure from large mammals, 4) highly dynamic soil conditions due to periodic permafrost, 5) the suddenness and intensity of the climatic warming and 6) the short duration of the warming intervals that left too little time for trees to migrate from their refugia in southern Europe. The latter two arguments suggest that vegetation response was not rapid enough to react to the short-lasting warming events. However, fast migrating biotic proxies such as Chironomidae and aquatic botanical taxa could be able to register even very abrupt warming events. An abrupt climate warming following a prolonged cold (stadial) period could lead to permafrost degradation and the formation of thermokarst lakes (so-called thaw lakes). This suggests that the response to a sudden and relatively short climate

56 Chapter 4 warming could be documented in the infill of such a thaw-lake. However, not every thermokarst lake is the result of external forcing mechanisms like a climatic warming. The formation of thermokarst lakes can also be due to internal forcing mechanisms, such as natural fires, erosion by river cut banks and inundations (Hopkins and Kidd 1988). To test whether thermokarst formation is internally or externally driven, the thermokarst infilling can be analysed for temperature-dependent proxies. If thermokarst initiation was triggered by climate warming, then evidence of such a thermal spike should be present in especially the lower parts of the thermokarst

GRIP oxygen isotopes OIS Chrono stratigraphy

Stadial -43 -39 -35 Interstadial Age (ka)

1 HOLOCENE Younger Dryas 11.5 1 Bø lling-Aller ø d LATE 14.7

2 2 LATE 3 28 4 5 6 7 8 Denekamp 9 10 11 3

12 Hengelo MIDDLE Hasselo 13 Moershoofd 14 Glinde Ebersdorf 15

16 Oerel PLENIGLACIAL 17 59 18 WEICHSELIAN 19 4

Schalkholz 20 EARLY 73 21 Odderade 22 Rederstall 5a-d 23 Brø rup EARLY 24 Herning 111 5e EEMIAN 129 6 SAALIAN

Table 4.1: Chronostratigraphy of the Weichselian and comparison with the d18O record of the GRIP ice core (Johnsen et al., 1992), the oxygen isotope stages (Martinson et al., 1987) and the terrestrial Interstadials and Stadials (e.g. Behre and van der Plicht, 1992; Dansgaard et al., 1993; Ran and van Huissteden, 1990). Ages follow Martinson et al. (1987) and Walker et al. (1999).

57 Chapter 4 infilling. Given the short duration of the Pleniglacial warm spikes in the ice-core record, one may assume that the bulk of the thermokarst infilling might already represent the waning stage of the thermal spike and the development towards a succeeding cold phase. Our hypothesis is that, analogous and synchronous to the Dansgaard/Oeschger events registered in the ice and marine cores, rapid climate warming occurred over northwestern Europe and that these rapid climate oscillations could have been recorded as thermokarst features.

4.2 Geological setting The opencast lignite mines in the Niederlausitz, eastern Germany, reveal extensive Weichselian sediment series deposited in a fluvial and partly aeolian context (Figure 4.1). These sediments are subject of an ongoing research program into the fluvial, palaeoenvironmental and climatological history of the region (Mol 1997a; 1997b; Bos et al. 2001; Kasse et al. 2003). The sequence in these quarries consists of local deposits from the rivers Spree and Neiβe. The latter flowed in a western direction through the ice marginal valley during part of the Middle Pleniglacial (Figure 4.1). Within the Weichselian fluvial and aeolian deposits frequent organic intercalations occur, which can be used for the reconstruction of the Pleniglacial vegetation and climate (Bos et al. 2001). In the opencast lignite mine of Reichwalde some of the organic sediments were interpreted as thermokarst lake deposits. The identification of former thermokarst situations and thermokarst sediments is complicated. Over the past decade we developed several diagnostic criteria by which thermokarst phenomena could be distinguished. These criteria are: 1) strong

WELZOW MINE Germany N WELZOW Poland SPREMBERG BAD MUSKAU Neiße Spree WEIßWASSER

NOCHTEN MINE HOYERSWERDA

SCHEIBE MINE REICHWALDE

MINE ¬

5 km REICHWALDE

older deposits ice-marginal valley cities lakes mines rivers sample location (mostly till) (former Neiße river) towns ¬

Figure 4.1: Location map

58 Chapter 4 deformations and ice-wedge casts below the thermokarst deposit ascribed to the degradation of the permafrost; 2) a sharp contact between the deformations and the overlying thermokarst infilling; 3) an abrupt and undeformed organic infilling of the thermokarst depression with gyttja reflecting an immediate presence of a lake after the permafrost degradation. A gradual deepening of the lake and a concomitant drowning of the vegetation in such case should be lacking; 4) the infill is mostly horizontally laminated.

4.3 Material and methods During a 2-month field work campaign in the Reichwalde opencast lignite mine (Figure 4.1) in June 1999, the sequence in the eastern wall of the mine was logged. Five stacked fluvial and aeolian units were identified, and the intercalated organic horizons were described. One of the organic layers, i.e. LM8, was sampled in a plastic box (20x25x80 cm) in order to secure enough material for pollen, macroremain, chironomid and loss-on-ignition analyses and for 14C dating on selected seeds/fruits (Figure 4.2). The lithostratigraphical position of the gyttja deposits could be tentatively correlated with the Middle Pleniglacial. The lithology of the deposit is displayed in Figures 4.2- 4.4.

4.3.1 Thermographic analysis A LECO TGA-601 was used to determine the loss-on-ignition of 18 samples of approximately 2 g (dry material). During the first 13 minutes of the treatment, samples were heated to a maximum of 105ºC. Moisture evaporates from the sample, and the dry weight of the sample is measured (Wdr (g)). Subsequently, the temperature is raised to 335ºC for a period of 35 minutes. The atmosphere in the oven consists of 100% oxygen, and all organic carbon is burned during this time interval. In order to determine the “classic” loss-on-ignition, the temperature in the oven is further raised to 550ºC for an additional period of 23 minutes, the atmosphere consisting of air again. Afterwards, the residue is weighed (Wgl (g)) and W550 is calculated using: W550 = ((Wdr – Wgl)/ Wdr) *100%. W550 is considered equivalent to the

“classic” LOI550 (Konert and Beets, in prep.).

4.3.2 Botanical analysis The plastic box was sampled at 1 cm intervals in the laboratory. Macroremain samples, varying between 20-60 g wet weights were treated with cold 5% KOH for 5 minutes (for organic samples) or Sodiumpyrophosphate (for clayey samples) and washed over a 200 µm sieve. Macroremains were picked out from the sieve residue. A Wild stereomicroscope (magnification up to 50x) was used during analysis. All material, including the left over sediment, has been stored in case the samples need to be re-examined in the future. Pollen samples were prepared following Faegri and Iversen (1975) with additional Sodiumpolytungstate heavy liquid separation to remove clastic material.

59 Chapter 4

Subsequently, to remove fine particles the material was sieved over a 7-8 µm nylon mesh, mounted in glycerine jelly and sealed with paraffin wax. A Zeiss light microscope with phase-contrast (magnification 400 and 630x) was used during analysis. To calculate the percentage curves of the individual taxa a pollen sum consisting of trees and shrubs, Ericales and dry herb taxa (including Poaceae) was employed. A combined pollen percentage and macroremain diagram (based on absolute values per sample) was constructed using TILIA.GRAPH (Grimm 1991-2004) and CORELDRAW computer programs. Plant macroremain and pollen identifications were made by comparison with modern reference material from the collection of the Vrije Universiteit Amsterdam. The nomenclature of pollen types in general refers to the Northwest European Pollen Flora (Punt et al. 1976-2003) or otherwise to Moore et al. (1991), plant macroremains to Berggren (1969; 1981), Anderberg (1994), Körber-Grohne (1964; 1991) and for Potamogeton to Cappers (1993). Palaeotemperature estimates were made based on the botanical taxa by using the climate indicator plant species method (sensu Iversen 1954; Kolstrup 1980). This method uses the relationship between the geographical limit of plant distribution and temperature, i.e. plants require a minimum mean summer temperature to flower and reproduce. Table 2 shows selected plant taxa and their required summer temperatures.

4.3.3 Chironomid analysis 18 samples of 4-8 cm3 were taken at approximately 2 cm intervals for chironomid analysis. The samples were soaked in 10% KOH for 4 hours and subsequently washed through a 100 µm sieve. Remaining fine sand grains were removed through decanting. Head capsules were handpicked at 35x magnification under a Wild stereomicroscope using a Bogorov sorting tray. After dehydration, head capsules were mounted on microscope slides using Euparal© mounting medium. Identifications were made at 100x and 400x magnification, and head capsules were identified mainly according to Hofmann (1971), Wiederholm (1983), Schmid (1993) and Moller Pillot (1984a; 1984b). To explore the possible relationships between the fossil samples and different environmental variables we used a subset of the 40 shallowest lakes of the Swiss 81- lakes training set developed by Heiri et al. (2003) as a modern analogue. Because of the differences in the general setting between the modern training set (the Swiss Alps) and the fossil environment (a braided-river floodplain) we decided to use a qualitative rather than a quantitative method to reconstruct environmental variables.

To test the relationship between mean July air temperature (Tjul) and the fossil chironomid assemblages, we performed a canonical correspondence analysis (CCA; ter Braak, 1986) using July air temperature as the sole constraining variable, the 40 shallow lakes as active variables and the 18 fossil samples as passive variables. Sample scores on the first ordination axes are plotted against core depth to visualize

60 Chapter 4

LM8 sand

thin sand bands clay

gyttja

¬ loading structure loam

sand

¬ frost fissure

Figure 4.2: Thawlake deposit LM8 and position of the sampled box

relative differences in fossil Tjul. CCA analyses are calculated using the program CANOCO (ter Braak and Šmilauer; 2002) and down-weighted percentage data.

4.3.4 Chronology Macroremains were selected for AMS 14C dating at two levels: one at the base of the gyttja (core depth 34-36 cm) and one higher up at core depth 20-22 cm. Because organic material reflecting atmospheric 14C concentrations such as seeds and fruits from terrestrial plants was absent, fruits of aquatic plants had to be dated. The sample at the base of the organic gyttja (34-36 cm core depth) consisted predominantly of Potamogeton mucronatus Schrad. (van der Meijden 2005 or P. friesii Rupr. Oberdorfer 1994) and P. praelongus. This sample provided an age of 45,850 ± 3750/2550 14C yrs BP. The second sample retrieved from core depth 20-22 cm consisted predominantly of P. praelongus. This sample yielded an infinite age of >47,000 14C yrs BP. The dates indicate that they are at the limit of the detection by the AMS 14C technique and although an ageing effect due to the uptake of old carbon can be expected in the aquatic macroremains, this effect can be ignored in view of the age indicated and the large standard error on the dates. Both the 14C ages, however, suggest an early Middle Weichselian age (early stage 3) for the infill of the thaw lake.

4.4 Results 4.4.1 Lithology and thermographic analysis Below the sampled horizon, floodplain sediments occur in which the influence of the river clearly declines. The grain size diminishes and eventually a light grey loam was

61 Chapter 4

Chironomid-based relative T Organic content (%) jul (CCA sample scores)

Depth (cm) Lithology 0 10 20 0.50 -0.5 -1.0 0 0 0 sand 150-210 µm, coarse up to 800 µm 5 5 5 continuous thin sand bands 10 10 10 dark grey brown clay

15 grey clay band 15 15

20 20 20

25 25 25 reddish brown gyttja 30 30 30

sand lenses, 35 intercalated with loam 35 35 light grey loam, undulating 40 base, i.e. loading 40 40 sand 3a. 3b. 3c.

Figure 4.3: LM 8 sequence: 3a. Lithology, 3b. TGA results, 3c. Chironomid-based relative TJul. deposited on top, i.e. core depth 40-38 cm (Figure 4.3a). These deposits have been interpreted as a distal facies of the then active river. The presence of a frost fissure suggests that from this moment onwards permafrost established on the site (Figure 4.2). Subsequently, cryoturbation or load-cast structures developed in a water- saturated environment. Under these circumstances high density loamy deposits sink down and due to permafrost degradation the low-density sandy sediments get pushed up. Moreover, the lithology demonstrates that the melting of the permafrost was not due to increased inundations of the river floodplain, because this would then have given horizontally bedded clayey overbank deposits (deriving from the inundations) on top of cryoturbation structures. Instead the loamy sediments have been incorporated into the cryogenic structures. Due to a progressive melting of segregation ice, water is further released and what was once the surface, sinks and forms the bottom of a small lake. Wave-erosion can potentially have enlarged the initial pool. Following the formation of the lake, horizontally bedded red-brown gyttja deposits were formed (38-16.5 cm core depth). This phase shows the highest organic content as demonstrated by the TGA results. Up to core depth 16 cm, the organic content steadily increases (Figure 4.3b). Between core depth 31 and 34 cm, a small perturbation disturbs the horizontally laminated gyttja and sand. Small clay pebbles are present in this interval. At core depth 16 cm, a sharp decline in organic matter is initiated and organic matter values drop to ~5 % (Figure

62 Chapter 4

4.3b). Between core depth 16.5 and 6 cm, a layer of dark grey-brown clay was deposited, with an increasing number of silty and sandy layers near the top of the sequence. These inputs of coarse clastic sediment are interpreted as moments of increased fluvial input to the basin. Above 6 cm the basin is filled with sand and the sample location is again within the reach of the floodplain. In these sandy sediments frost fissures are present, which suggest a return to permafrost conditions and mean annual air temperatures between -4 and -8°C (Huijzer and Vandenberghe 1998). The total thickness of the whole fluvial cycle is ca. 2 m thick (Bontebal and Smit, 2001). Five of these cycles have been registered during the logging of the profile faces. These sequences are interpreted as deposited by low-energetic anatomosing river systems.

4.4.2 The palaeobotanical record Consecutive samples were analyzed through the basal part of the gyttja, because if the thermokarst lake was formed due to external forcing by climate warming, then evidence for this Pleniglacial thermal spike should be present in the bottom layers of the infill. The overlying sediment was analyzed every 4 cm. The botanical record of sequence LM8 is divided into four zones (Figures 4.4a and 4.4b).

Zone RWB-1 (40.5-38 cm core depth) No pollen samples are available for this interval. However, macroremains of Carex spp., Characeae, Potamogeton praelongus, Bryozoa, Daphnia and Cenococcum geophilum are recorded. These indicate temporary, shallow water conditions after inundation of the floodplain. Deposition of loam during these flooding events represents the terminal phase of river activity. Finds of the soil fungus Cenococcum geophilum may be connected with surface erosion and reworking.

Zone RWB-2 (38-31 cm core depth) Zone RWB-2 is characterized by tree and shrub (Betula, Salix, Pinus, Alnus etc.) values of ca. 50-55%, high non-arboreal pollen (NAP) and Poaceae percentages (together up to 50%), and an increasing Juniperus curve with a maximum of 8% in the central part of this zone. Willow pollen may have been derived from dwarf willows such as Salix herbacea and S. polaris. Birch pollen was recorded with high values up to 35% and was probably derived from Betula nana, a cold-adapted dwarf-birch, which is also indicated by finds of its fruits (Figure 4.4a). Modern surface sample studies from the (sub) Arctic, however, suggest that B. nana has the tendency to be overrepresented in the pollen values (e.g. Iversen, 1954; Rymer, 1973). Some of the Betula pollen also may have been derived from tree birches through long-distance transport. Pinus pollen shows percentages in general below 20%, i.e. the rational limit of pine (Lotter et al. 1992; Lang 1994), and was interpreted as long distance deposition. Alder pollen may have originated from Alnus incana, a tall, thicket-forming shrub. A. incana is an early succession species and may have been a pioneer of recently deglaciated areas in Europe (compare Œrodoñ 1980; Litt 1994; Caspers and Freund 2001; Dambeck and Bos 2002; Bos et al. 2005). However, no A. incana macroremains were recorded that may suggest local presence. 63

Chapter 4

> 47,000 > 3750/2550 ± 45,850

C dates C

4 1 Biozone

RWB-1 RWB-4 RWB-2 RWB-3 Cenococcum geophilum Cenococcum 40%

20 20 40

Total of reworked palynomorphs reworked of Total

Pollen sum Pollen

215 206 217 233 235 224 229 179 180 125 163 144 127 158

Saxifraga androsacea Saxifraga

Chenopodiaceae

type Gentianaceae

Papaver radicatum Papaver

iflorae

Caryophyllaceae tharticum tharticum gul

Exaggeration of microfossil curves 5x. lantago major lantago

viviparum P

cf. cf.

Linum ca Linum

Compositae li Compositae icrofossils are displayed by curves and given in

type

Polygonum

Geum Geum

mex acetosa mex

Ru

teraceae tubuliflorae teraceae

As

Brassicaceae

Helianthemum

Fabaceae ca

Dryas or abundance (+++). Thalictrum

Tundra/steppe taxa and Poaceae Cornus sueci Cornus

type Artemisia

Lotus Apiaceae

20%

Poaceae type

type

20% Saxifagaceae

Potentilla

type

type lantago maritima maritima lantago

type P

tum tum

Oxyria

type

Ericales Plantago media Plantago

Aconitum napellus Aconitum

Bupleurum falca Bupleurum

Empetrum nigrum Empetrum

Bruckenthalia

Ericaceae indet. Ericaceae Alnus

40% , fruits ,

20 Trees and shrubs Pinus

5 Betula nana Betula 40%

20

Betula

NAP

Salix

Juniperus

es Frangula

ra/steppe taxa ra/steppe Erical

100% Tund

Poaceae 80 Shrubs Betula Pinus

Microfossil and macroremain diagram of LM8 sequence with a selection regional taxa. M

20 40 60 Trees

Lithology Depth (cm) Depth 0 5 percentages, macroremains are displayed by amounts as histograms or presence (+) Figure 4.4a: Microfossil and macroremain diagram with a selection of regional taxa 10 15 20 25 30 35 40

64 Chapter 4

The relatively high values of long-distance transported pollen (Pinus and possibly some Betula and Alnus), the low amounts of B. nana fruits and low pollen values of Salix therefore indicate that the landscape was rather open and treeless. The absence of trees in north and central Europe during this period was also suggested by other authors (Behre & Lade 1986; Behre 1989; Behre et al. 2005; Caspers and Freund 2001). The combined pollen and macroremain data indicate regional vegetation characterized by heliophilous herbs, grasses, and dwarf shrubs of Betula nana and Salix, which can be interpreted as low shrub tundra (sensu Bliss and Richards 1982). At this time, the lake was shallow and gyttja was being deposited. Aquatic communities developed with both submerged and floating-leaved taxa, e.g. Potamogeton spp., Myriophyllum spp., Nymphaea alba, Ranunculus Subgen. Batrachium, Characeae and algae. The botanical taxa suggest that the water was transparent, carbonate-rich and meso- to eutrophic. At the sample site the water depth was at most 2-3 m (compare

Minimum mean July Species Reference Temperature (oC)

Botanical taxa:

Polygonum viviparum 5 Kolstrup 1980 Betula nana 7 Brinkkemper et al. 1987; Ran 1990 Papaver 7.4 Vorren 1978

Empetrum 7.7 Vorren 1978 Filipendula ulmaria 8 Brinkkemper et al. 1987 Caltha palustris 8 Kolstrup 1980

Carex rostrata 8 Brinkkemper et al. 1987 Cornus suecica 8 Kolstrup 1980

Potentilla palustris 8 Brinkkemper et al. 1987

Juniperus communis 8 Isarin and Bohncke 1999 10 (taller plants) Kolstrup 1980 Cryptogramma crispa 9.9 Vorren 1978 Eleocharis palustris 10 Kolstrup 1980 Myriophyllum alterniflorum 10 Kolstrup 1980; Isarin and Bohncke 1999

Ranunculus subgen. Batrachium 10 Brinkkemper et al. 1987 Myriophyllum spicatum >10 Kolstrup 1979; 1980

Myriophyllum verticillatum >10 Kolstrup 1979; 1980 Nymphaea alba 12 Kolstrup 1979; 1980 Frangula alnus 13 Kolstrup 1980 Potamogeton mucronatus 13 Brinkkemper et al. 1987 Typha angustifolia 14 Kolstrup 1979; 1980

Zoological taxa:

Cristatella mucedo 10 Lacourt 1968

Table 4.2: Temperature indicator botanical and zoological taxa present in the deposits.

65

Chapter 4 Biozone

RWB-1 RWB-4 RWB-3 RWB-2

Batrachium

Daphnia

Subg.

Zygnemataceae

anunculus anunculus

R

p

grou s aquatilis s 40

20 5 10

Ranunculu

us braunii us ococc

20% 20% Botry

40 60% 60% Exaggeration of microfossil curves 5x. curves microfossil of Exaggeration , statoblasts ,

um s

cedo

20

Pediastr

icrofossils are displayed by curves and given in given and curves by displayed are icrofossils Aquatic taxa Aquatic

btusifolius

Cristatella mu Cristatella 10

ucronatus

pus Potamogeton praelongu Potamogeton

otamogeton o otamogeton

P 10

otamogeton m otamogeton

P 10

otamogeton cris otamogeton

P 10

um verticillatum um

otamogeton

um alterniflorum um icatum P

p (+++). abundance or riophyll

hyll

p My

10

yllum s yllum

Myrio

rioph

My 10

m m llu 40%

20 10% Myriophy

20 40

palustris Characeae

atica atica

Nymphaea alba Nymphaea qu

Eleocharis Eleocharis

type

Mentha a Mentha

alustris Galium

Filipendula

Potentilla p Potentilla 10

Carex nigra Carex 10

Carex aquatilis Carex

Carex elata Carex 10 10 10 10 10 10 Riparian herbs

eraceae

50%

Cyp m

isetu m

stifolium 20% Equ Sphagnu

10

Sparganium minimum Sparganium 10

arganium angu arganium

Sp 10

silate type

pe

ty

Sparganium 10%

Monolete p Monolete

Caltha palustris Caltha

Viola palustris Viola sp.

Typha angustifolia Typha

spp.

Calliergon Calliergon

+ + + + + + + + + Scorpidium scorpioides Scorpidium

+ + + + +

: Microfossil and macroremain diagram of the LM8 sequence with a selection of local taxa. M taxa. local of selection a with sequence LM8 the of diagram macroremain and Microfossil : Drepanocladus Drepanocladus

+ + + + + + + + + + + ++ + + + + + + + + +

Lithology Depth (cm) Depth 0 5 Microfossil and macroremain diagram with a selection of local taxa diagram with a selection Microfossil and macroremain 10 20 25 30 35 40

15

percentages, macroremains are displayed by amounts as histograms or as presence (+) (+) presence as or histograms as amounts by displayed are macroremains percentages, Figure 4.4b Figure

66 Chapter 4

Hannon and Gaillard 1997). Swamp vegetation inclusive Carex spp., Sparganium spp., Equisetum, Viola palustris, Caltha palustris, and Drepanocladus spp. fringed the shores, while Scorpidium scorpioides probably formed floating mats in these shallow, base-rich waters (cf. Dickson 1973). Pollen of thermophilous species present in this zone, i.e. Nymphaea alba (12°C), Frangula alnus (13°C, Table 4.2) and Typha angustifolia (14°C, Table 4.2), indicate minimum mean summer temperatures between 12 and 14°C. However, the most solid evidence for relatively high summer temperatures is provided by the presence of Potamogeton mucronatus fruits, a taxon that requires a minimum Tjul of 13°C (Table 4.2) and appears later in this zone. Furthermore, thriving juniper shrubs need a minimum mean July temperature that exceeds the 10°C (Iversen 1954), while the abundance of both pollen and seeds of Myriophyllum spicatum and M. verticillatum also suggests a minimum Tjul of >10°C (Table 4.2). The botanical taxa thus suggest a minimum mean Tjul of 13°C during this zone. For references to temperature indications see Table 4.2.

Zone RWB-3 (31-12.5 cm core depth) At 31 cm Sparganium spp. and Potamogeton crispus disappear, and later also P. obtusifolius and P. mucronatus, and the values of Equisetum, Myriophyllum and Characeae decline. In the lake, aquatic vegetation was dominated by taxa such as P. praelongus, Ranunculus Subgen. Batrachium and algae (Pediastrum, Botryococcus braunii and Zygnemataceae), and animal taxa such as Bryozoa and Daphnia. The robust plants with large submerged leaves of P. praelongus (Weeda et al. 1991) probably reduced the penetration of light and hence caused the decline in Characeae at ca. 30 cm. Swamp vegetation including Equisetum, Carex spp., Eleocharis palustris, Mentha aquatica, Filipendula, Potentilla palustris, Calliergon and Drepanocladus spp. fringed the shores. The local taxa indicate, besides a decrease in the lake water-depth, that the water became less carbonate-rich and less transparent. The latter was probably caused by the hydroseral succession process within the lake as well as by a higher abundance of algae, such as Pediastrum, Botryococcus and Zygnemataceae. These algae to a large extent caused the increase in organic content (Figure 4.3b). The increasing accumulation of organic matter probably also influenced the disappearance of M. alterniflorum (compare Weeda et al. 1987). The presence of Carex aquatilis and higher abundance of C. elata indicates seasonally fluctuating water-levels (Weeda et al. 1994). The abundance of Zygnemataceae supports this interpretation, since they pre- eminently occur in shallow waters and the spores can survive in unfavourable conditions, e.g. drying out of temporary pools (van Geel and Grenfell 1996). In this zone many thermophilious taxa (Juniperus, Frangula, Myriophyllum spp., P. mucronatus, Typha angustifolia and Nymphaea alba) decline or disappear. The remaining taxa, i.e. Ranunculus Subgen. Batrachium, M. alterniflorum, Eleocharis palustris and Cristatella mucedo are less warmth requiring and only need a minimum mean Tjul of >10°C.

67 Chapter 4

Chironomid diagram with a selection of taxa

pe ty

pe inus ty )

/g c um Depth (cm)Lithology anthrac lugens 0 es us a era ambigua mid s tion (h nemus c oc tars ladius pedilum noc one du ronomus c rotendipes rotendipes y rono ny a ioz 4 C Dates 5 Pro T Chi LimnophyPro Dic SergentiTribelosMic Poly CladopelmaMetriCor Chi B 1

14 61 10 57 RWC-2 56 58 15 97 89 90 20 85 66 > 47,000 76 RWC-1 25 68 161 139 30 70 70

76 35

81 ±

40 45,850 3750/2550

20 40 60 10% 20% 10% 20 40% 10% 10%10% 20 40% 10%10%10% 20%

Figure 4.5: Chironomid diagram of the LM8 sequence. Chironomid abundances are given in percentages of the total Chironomid sum.

Zone RWB-4 (12.5-6 cm core depth) In this zone, the percentages of Pinus, Alnus, Cornus suecica, reworked palynomorphs and numbers of Cenoccocum geophilum increase, while the percentages of Betula decrease. This suggests that in the surrounding region rather patchy tundra vegetation developed. The landscape became more open and surface erosion (increase Cenococcum) and long-distance transport (e.g. increase in Pinus and Alnus values) were important processes. At the sample location, sedge swamp vegetation including Carex elata, C. aquatilis, Ranunculus Subgen. Batrachium and Characeae was present, indicating fluctuating water levels and a water depth of ca. 0.25-1 m (compare Hannon and Gaillard 1997). Minimum mean July temperatures were probably around 10°C (indicated by Ranunculus Subgen. Batrachium, Myriophyllum alterniflorum and Cristatella mucedo).

4.4.3 The chironomid record The chironomid-assemblages recorded in sequence LM8 confirm that a shallow, meso- to eutrophic pond or lake existed at the sample location during the early

68 Chapter 4

Middle Weichselian (Figure 4.6). A total of 54 genera, species or morpho-types have been identified and the most abundant taxa are plotted in Figure 4.5. The genus Microtendipes is often interpreted as thermophilous with a limited distribution in arctic and alpine lakes (Porinchu et al. 2003; Laroque et al. 2001; Olander et al. 1999) and with a preference for shallow or littoral habitats. For instance, it currently occurs in temperate to subalpine lakes in the Swiss Alps (Lotter et al. 1997; Bigler et al. 2006). Microtendipes has a high abundance throughout the entire record (Figure 4.5). Other genera that occur throughout the fossil record are e.g. Limnophyes, Dicrotendipes, and Polypedilum, all suggesting a shallow lake environment. At core depth 12.5 cm, there is a shift in dominant taxa and the chironomid record is therefore divided in two zones.

Zone RWCh-1 (36-12.5 cm core depth) In the lower part of this zone Chironomus anthracinus-type shows abundances over 15%. After this initial peak, the values of C. anthracinus-type decline in favor of Microtendipes. The presence of genera such as Dicrotendipes and Glyptotendipes suggests that macrophytes were present in the lake (e.g. Moller Pillot 1990), which was confirmed by the botanical data. The lake was meso- to eutrophic, as indicated by several species such as Ablabesmyia or Chironomus plumosus-type (e.g. Brodersen and Quinlan 2006; Lotter et al. 1998). Most of the abundant fossil taxa currently occur in lowland or temperate regions in NW Europe.

Zone RWCh-2 (12.5-8 cm core depth) At 12.5 cm core depth Microtendipes, Cladopelma, Metriocnemus terrester-type and Polypedilum values decrease while taxa such as Procladius, Tanytarsus lugens-type and C. anthracinus type show higher abundances. Chironomid productivity rates drop considerably, a trend that is often interpreted as a decline in either temperature or in trophic state. In view of the return of fluvial activity to the site, a decline in trophic state can be excluded as a forcing factor. T. lugens-type is one of the taxa that dominates in this part of the record, and is often considered an indicator of cool, oligotrophic conditions (e.g. Porinchu & MacDonald 2003, and references therein). The high values of C. anthracinus on the other hand also could point to an increase in nutrient availability, as they are often reported to be characteristic for high productivity shallow lakes (Brodersen & Lindegaard 1997). The botanical aquatic taxa however, suggest no increase in the nutrient availability within the lake. The genus Chironomus is also known to be common in shallow, relatively warm, arctic (or high elevation) ponds (Walker 1990).

Chironomid-based temperature reconstructions Figure 4.3c shows the scores of the fossil assemblages on the first ordination axis after the CCA-run with July air temperature as the sole constraining variable. The scores on this ordination axis represent relative temperature values, where a higher (or less negative) value indicates lower temperatures. Temperatures are high and rather stable in the lower parts of the record. At core depth 12.5 cm a decline in temperature

69 Chapter 4 temperature is inferred, as already suggested by the transition in the chironomid assemblages (Figure 4.5).

4.5 Discussion Not every thermokarst lake is formed as a result of external forcing mechanisms like a climatic warming, as they can also be formed as the result of an internal forcing mechanism. If thermokarst initiation was triggered by climate warming, then the evidence of such a thermal spike should be present especially in the basal part of the thermokarst infilling. Up to now, no explicit evidence for climatic control governing the process of thermokarst formation has been shown in the sedimentological record of eastern Germany. In the thermokarst infilling of LM8, aquatic plant remains suggest a high minimum mean Tjul of ca. 12-14ºC shortly after the formation of the lake and during the initial period of deposition. Aquatic plant species can spread relatively quickly by water bird transportation, are independent of soil formation and thus can react relatively quickly to climate warming (e.g. Iversen 1954; 1973; Figuerola and Green 2002). Chironomids are another fast-migrating proxy, as the life cycle of these animals is short and the adults are able to fly and thus potentially move a large distance within one generation. During the last cold stage, the temperate European flora and fauna became restricted to the mountains of southern Europe (i.e. Balkan, Italy and Iberian Peninsula) and in some cryptic refugia in central and northern Europe (e.g. Bennett et al. 1991; Willis 1996; Willis et al. 2000; Stewart and Lister 2001; Svenning 2003). With respect to the rapid spread of botanical and zoological taxa after deglaciation, especially these northern cryptic refugia, located in areas of sheltered topography, are important (Stewart and Lister 2001; Hampe et al. 2003).

The relatively high Tjul during initial deposition suggests that permafrost degradation and subsequent thermokarst lake formation was climatically induced. Apparently, the lower part of the record was formed during a relatively warm interval. At core depth 12.5 cm several events are witnessed: 1. the organic content of the sediment drops considerably to values of ~5%, which can either be due to a decline in primary production or to an increase in clastic material, or a combination of both, 2. the pollen record demonstrates an increase in the pine pollen values and Pinus becomes dominant over Betula, which suggests an increase in long-distance transport as a result of a more patchy shrub tundra vegetation, 3. both the larger abundance of Cenococcum geophilum sclerotia and the increased presence of reworked palynomorphs suggest a larger role for surface erosion. 4. chironomid production rates decline, there is a shift in dominant chironomid taxa, and CCA with Tjul as the sole constraining variable suggests a sharp decline in summer temperatures. Together with the return of permafrost conditions as witnessed in the sedimentary sequence (i.e. frost fissures), this suggests a recurrence of stadial conditions. The succeeding cycle starts again with interstadial conditions, the melting of the permafrost and the formation of load-cast structures or cryoturbations. Five of these

70 Chapter 4 cycles have been recorded in the Reichwalde mine during the early Middle Weichselian, which can be tentatively correlated with D/O cycles 17 to 13. Many of these changes, however, also could be a reflection of a lowering in the water level of the lake. The botanical taxa suggest a constant decrease in lake depth, initiated around 31 cm core depth. Since this is a gradual process, it does not explain the abrupt change in temperature at core depth 12 cm. This excludes lake depth as the forcing factor for the major change in the fossil chironomid-assemblages at core depth 12 cm. In contrast, the reconstructed temperatures do show a sharp transition at core depth 12.5 cm (Figure 4.3c), suggesting palaeotemperature as the forcing factor for the concurrent changes in the lake catchment. The high initial temperatures together with the sharp transition to a colder environment in the upper part of the sequence suggest a climatic evolution that resembles a D/O-event. In the absence of a reliable time control on the core, the rapidity of the temperature change at core depth 12.5 cm can only be a rough estimate based on sediment accumulation rates derived from the literature (e.g. Berglund 1986; Korhola et al. 2002). Sediment accumulation rates of gyttja in lakes come to 0.5-1 mm/year (Berglund 1986). In subarctic and arctic regions, however, sediment accumulation rates may be lower, 0.1-0.5 mm/year (e.g. Berglund 1986; Korhola et al. 2002). Taking into account a compaction of 50% of the original thickness this would than amount to 0.05-0.5 mm/year. The registered temperature change occurred within 3 cm, which would represent a time slot of 120-600 years. Climate oscillations such as the D/O events provide us with a framework in which many more warming and cooling events could occur during which permafrost could develop and subsequently degrade in West and Central Europe. The short duration of the warming events explains why there was little time for arboreal vegetation to react, e.g. by a shift in the vegetation zones over NW Europe. If the deduction that we are dealing with a warming event is correct, several of the numbered interstadials that have been distinguished in the Greenland ice cores (Johnsen et al. 1992) may be correlated with this event. These are interstadials 14, 15 and 16 (Table 1). The radiocarbon age of the deposit may indicate correlation with interstadial 14, i.e. the Glinde interstadial (Behre and van der Plicht 1992; Dansgaard et al. 1993).

4.6 Conclusions The combined evidence demonstrates that formation of the thermokarst lake was initiated by warm climate conditions. Pollen, botanical macroremains and chironomid analyses show that the basal infilling of the thaw lake occurred during a period with high mean July air temperatures (12-14ÚC). This warm interval was probably too short for boreal forest to react. The return to permafrost conditions higher up in the sequence meets the perception that we are dealing with an early Middle Weichselian D/O event. The rapid warming initiated the degradation of the permafrost and the formation of the thaw lake. The infilling of the thaw lake

71 Chapter 4 represents the waning phase of the warming and the subsequent return to cold climate conditions. The available AMS 14C dates, although at the limit of detection, indicate an early Middle Weichselian age of the gyttja and the D/O event registered possibly represent D/O 14, 15 or 16.

References Anderberg AL (1994) Atlas of Seeds and small fruits of Northwest-European plant species with morphological descriptions. Part 4, Resedaceae-Umbelliferae. Swedish Museum of Natural History. Risbergs Tryckeri AB, Uddevalla. Behre K-E (1989) Biostratigraphy of the last glacial period in Europe. Quatern Sci Rev 8: 25-44 Behre K-E, Lade U (1986) Eine Folge von Eem und Weichsel-Interstadialen in Oerel/Nieder-sachsen und ihr Vegetationsablauf. Eiszeitalt Gegenw 36: 11-36 Behre K-E, Hölzer A, Lemdahl G (2005) Botanical macro-remains and insects from the Eemian and Weichselian site of Oerel (northwest Germany) and their evidence for the history of climate. Veget Hist Archaeobot 14: 31-53 Behre K-E, van der Plicht J (1992) Towards an absolute chronology for the last glacial period in Europe: radiocarbon dates from Oerel, northern Germany. Veget Hist Archaeobot 1: 111-117 Bennett KD, Tzedakis PC, Willis KJ (1991) Quaternary refugia of north European trees. J Biogeogr 18: 103-115 Berggren G (1969) Atlas of Seeds and small fruits of Northwest-European plant species with morphological descriptions. Part 2, Cyperaceae. Swedish Natural Science Research Council. Berlingska Botryckeriet, Lund Berggren G (1981) Atlas of Seeds and small fruits of Northwest-European plant species with morphological descriptions. Part 3, Salicaeae-Cruciferae. Swedish Museum of Natural History. Berlings, Arlöv Berglund BE (1986) Palaeoecological reference areas and reference sites. In: Berglund BE (Ed) Handbook of Holocene Palaeoecology and Palaeohydrology. John Wiley. Chichester and New York pp 111-126 Bigler C, Heiri O, Krskova R, Lotter AF, Sturm M (2006) Distribution of diatoms, chironomids and cladocera in surface sediments of thirty mountain lakes in south-eastern Switzerland. Aquat Sci 68: 154-171 Bliss LC, Richards JH (1982) Present-day arctic vegetation and ecosystems as a predictive tool for the artic-steppe mammoth biome. In: Hopkins DM, Matthews JV, Schweger CE, Young SB (Eds.) Paleoecology of Beringia. Academic Press. Inc., New York, pp. 241-257 Bond G, Broecker W, Johnsen S, Mc Manus J, Labeyrie L, Jouzel J, Bonani G (1993) Correlations between climate records from North Atlantic sediments and Greenland ice. Nature 365: 143-147 Bond GC, Showers W, Cheseby M, Lotti R, Almasi P, de Menocal P, Priore P, Cullen H, Hajdas I, Bonani G (1997) A parvasive millennial cycle in the North Atlantic Holocene and Glacial times. Science 278: 1257-1266 Bond GC, Showers W, Elliot M, Evans M, Lotti R, Hajdas I, Bonani G, Johnson S (1999) The North Atlantic’s 1-2 Kyr Climate Rhythm: Relation to Heinrich Events, Dansgaard/Oeschger Cycles and the Little Ice Age. In: Clark PU, Webb RS, Keigwin LD (Eds) Mechanisms of Global Climate Change at Millennial Time Scales. American Geophysical Union, Washington DC, pp. 35-58 Bontebal M, Smit L (2001) Rivieractiviteit tijdens het Boven-Pleistoceen in relatie tot klimaatsveranderingen in het Lausitzer Urstromtal, Duitsland. MSc. Thesis Vrije Universiteit Amsterdam, 88 pp. Bos JAA, Bohncke SJP, Kasse C, Vandenberghe J (2001) Vegetation and Climate during the Weichselian Early Glacial and Pleniglacial in the Niederlausitz, eastern Germany - macrofossil and pollen evidence. J Quatern Sci 16: 269-289 Bos JAA, Dickson JH, Coope GR, Jardine WG (2004) Flora, fauna and climate of Scotland during the Weichselian Middle Pleniglacial - palynological, macrofossil and coleopteran investigations. Palaeogeogr Palaeoclim Palaeoecol 204: 65-100 Bos JAA, Huisman J, Kiden P, Hoek WZ, van Geel B (2005) Early Holocene environmental change in the Kreekrak area (Zeeland, S-W Netherlands): a multi-proxy analysis. Palaeogeogr Palaeoclim Palaeoecol 227: 259- 289 Brinkkemper O, van Geel B, Wiegers J (1987) Palaeoecological study of a Middle Pleniglacial deposit from Tilligte, The Netherlands. Rev Palaeobot Palynol 51: 235-269 Brodersen KP, Lindegaard C (1997) Significance of subfossile chironomid remains in classification of shallow lakes. Hydrobiologia 342/343: 125-132 Brodersen KP, Quilan R (2006) Midges as palaeoindicators of lake productivity, eutrophication and hypolimnetic oxygen. Quatern Sci Rev 25: 1995-2012 Cappers RTJ (1993) The identification of Potamogetonaceae fruits found in The Netherlands. Acta Bot Neerl 42: 447-460 Caspers G, Freund H (1997) Die Vegetations- und Klimaentwicklung des Weichsel-Früh- und – Hochglazials im nördlichen Mitteleuropa. Schriftenr Deutsch Geolog Gesellsch 4: 201-249 Caspers G, Freund H (2001) Vegetation and climate in the Early- and Pleni-Weichselian in northern central Europe. J Quatern Sci 16: 31-48

72 Chapter 4

Coope GR (1975) Climatic fluctuations in northwest Europe since the Last Interglacial, indicated by fossil assemblages of Coleoptera. In: Wright AE, Moseley F (Eds) Ice Ages: Ancient and Modern. Geol. J. special Issue 6, Seel House Press, Liverpool, pp. 153-168 Coope GR (2000) Middle Devensian (Weichselian) coleopteran assemblages from Earith, Cambridgeshire (UK) and their bearing on the interpretation of ‘full glacial’ floras and faunas. J Quatern Sci 15: 779-788 Dansgaard W, Johnsen SJ, Clausen HB, Dahl-Jensen D, Gundestrup NS, Hammer CU, Hvidberg CS, Steffensen JP, Sveinbjörndottir AE, Jouzel J, Bond G (1993). Evidence for general instability of past climate from a 250-kyr ice-core record. Nature 364: 218-220 Dambeck R, Bos JAA (2002) Lateglacial and Holocene landscape evolution of the northern upper Rhine river valley, southwestern Germany. Zeitschr Geomorph Supp. Bd 128: 101-127 De Beaulieu J-L, Reille M (1984) A long Upper Pleistocene pollen record from Les Echets, near Lyon, France. Boreas 13: 111-132 De Beaulieu J-L, Reille M (1992) The last Climatic Cycle at La Grande Pile (Vosges, France) a new pollen profile. Quatern Sci Rev 11: 431-438 Dickson JH (1973) Bryophytes of the Pleistocene. Cambridge University Press, Cambridge Faegri K, Iversen J (1975) Textbook of Pollen Analysis, third edition. Munksgaard, Copenhagen Figuerola J, Green AJ (2002) Dispersal of aquatic organisms by waterbirds: a review of past research and priorities for future studies. Freshw Biol 47: 483-494 Ganopolski A, Rahmstorf S (2001) Rapid changes of glacial climate simulated in a coupled climate model. Nature 409: 153-158 Grimm EC (1991–2004) TILIA, TILA.GRAPH, and TGView. Illinois State Museum, Research and Collections Center, Springfield, USA (http://demeter.museum.state.il.us/pub /grimm/) Grootes PM, Stuiver M (1997) Oxygen 18/16 variability in Greenland snow and ice with 10-3 to 105- year time resolution. J Geophys Res 102: 26455–26470 Guiot J, Pons A, de Beaulieu J-L, Reille M (1989) A 140,000-year continental climate reconstruction from two European pollen records. Nature 338: 309-313 Hampe A, Arroyo J, Jordano P, Petit RJ (2003) Rangewide phylogeography of a bird-dispersed Eurasian shrub: contrasting Mediterranean and temperate glacial refugia. Molec Ecol 12: 3415-3426 Hannon GE, Gaillard M-J (1997) The plant macrofossil record of past lake-level changes. J Paleolim 18: 15-28 Heiri O, Lotter AF, Hausmann S, Kienast F (2003) A chironomid-based Holocene summer air temperature reconstruction from the Swiss Alps. Holocene 13: 477-484 Hofmann W (1971) Zur Taxonomie und Palökologie subfossiler Chironomiden (Dipt.) in Seesedimenten. Arch Hydrobiolog 6: 1-50 Hopkins DM, Kidd JG (1988) Thaw lake sediments and sedimentary environments. Proceedings Vth International Conference on Permafrost, Trondheim, 790-795 Isarin RFB, Bohncke SJP (1999) Summer temperatures during the Younger Dryas in North-western and central Europe inferred from climate indicator plant species. Quatern Res 51: 158-173 Iversen J (1954) The Late Glacial flora of Denmark and its relation to climate and soil. Danm Geol Undersøg. 2: 88 119 Iversen J (1973) The development of Denmark’s Nature since the Last Glacial. Danm Geol Undersøg. V, 7-C, 126 pp Johnsen SJ, Clausen HB, Dansgaard W, Fuhrer K, Gundestrup N, Hammer CU, Iversen P, Jouzel J, Stauffer B, Steffensen JP (1992) Irregular glacial interstadials recorded in a new Greenland ice core. Nature 359: 311- 313 Kasse C, Vandenberghe J, van Huissteden J, Bohncke SJP, Bos JAA (2003) Sensitivity of Weichselian fluvial systems to climate change (Nochten mine, eastern Germany). Quatern Sci Rev 22: 2141-2156 Kolstrup E (1979) Herbs as July temperature indicators for parts of the Pleniglacial and the Late-glacial in The Netherlands. Geol Mijnb 59: 377-380 Kolstrup E (1980) Climate and stratigraphy in Northwestern Europe between 30,000 BP and 13,000 BP, with special reference to The Netherlands. Meded Rijks Geol Dienst 32: 181-253 Kolstrup E, Wijmstra TA (1977) A palynological investigation of the Moershoofd, Hengelo, and Denekamp interstadials in the Netherlands. Geol Mijnb 58: 377-380 Kolstrup E (1990) The puzzle of Weichselian vegetation types poor in trees. Geol Mijnb 69: 253-262 Konert M, Beets K (in prep) Simultaneously analysis of organic matter, structural water and carbonates on sediments with a thermo gravimetric analyzer: a comparison with classical analysis Körber-Grohne U (1964) Bestimmungsschlüssel für subfossile Juncus-Samen und Gramineen-Früchte. Probl Küstenforsch Südl Nordseegeb 7 Körber-Grohne U (1991) Bestimmungsschlüssel für subfossile Gramineen-Früchte. Probl Küstenforsch Südl Nordseegeb 18: 169-234 Korhola A, Sorvari S, Rautio M, Appleby PG, Dearing JA, Hu Y, Lami A, Cameron NG (2002) A multi-proxy analysis of climate impacts on the recent development of subarctic Lake Saanajärvi in Finnish Lapland. J Paleolimnol 28: 59-77 Lacourt AW (1968) A monograph of the freshwater Bryozoa - Phylactolaemata. Zool Verh 93: 3-159 Lang GH (1994) Quartaire Vegetationsgeschichte Europas. Gustav Fischer Verlag, Jena

73 Chapter 4

Larocque I, Hall RI, Grahn E (2001) Chironomids as indicators of climate change: a 100-lake training set from a subarctic region of northern Sweden (Lapland). J Paleolimnol 26: 307-322 Litt T (1994) Paläoökologie, Paläobotanik und Stratigraphie des Jungquartärs im nordmittel-europäischen Tiefland. Diss Botan 227: 1-185 Lotter AF, Eicher U, Birks HJB, Siegenthaler U (1992) Late glacial climatic oscillations as recorded in Swiss lake sediments. J Quatern Sci 7: 187 204 Lotter AF, Birks HJB, Hofmann W, Marchetto A (1997) Modern diatom, cladocera, chironomid, and chrysophyte cyst assemblages as quantitative indicators for the reconstruction of past environmental conditions in the Alps. I. Climate. J Paleolimnol 18: 395-420 Lotter AF, Birks HJB, Hofmann W, Marchetto A (1998) Modern diatom, cladocera, chironomid, and chrysophyte cyst assemblages as quantitative indicators for the reconstruction of past environmental conditions in the Alps. II. Nutrients. J Paleolimnol 19: 443-463 Martinson DG Pisias NG, Hays JD, Imbrie J, Moore TC, Shackleton NJ (1987) Age dating and the orbital theory of the ice ages: development of a high-resolution 0 to 300,000-year chronostratigraphy. Quatern Res 27: 1-29 Mol J (1997a). Fluvial response to climate variations. The Last Glaciation in eastern Germany. PhD Thesis, Vrije Universiteit, Amsterdam, Amsterdam Mol J (1997b) Fluvial response to Weichselian climate changes in the Niederlausitz (Germany). J Quatern Sci 12: 43-60 Moller Pillot HKM (1984a) De larven der Nederlandse Chironomidae (Diptera). Inleiding, Tanypodinae en Chironomini. Ned Faun Meded 1a Moller Pillot HKM (1984b) De larven der Nederlandse Chironomidae (Diptera). Orthocladiinae. Ned Faun Meded 1b Moller Pillot HKM (1990) De larven der Nederlandse Chironomidae. Autoekologie en verspreiding. Ned Faun Meded 1c Moore PD, Webb JA, Collinson ME (1991) Pollen analysis, second edition. Blackwell Scientific Publications, Oxford NEPF The Northwest European Pollen Flora 1-7 (1976, 1980, 1981, 1984, 1988, 1991, 1996, 2003) Punt W et al (eds.), Elsevier, Amsterdam Oberdorfer E (1994) Pflanzensoziologische Exkursionsflora. Ulmer, Stuttgart Olander H, Birks HJB, Korhola A, Blom T (1999) An expanded calibration model for inferring lakewater and air temperatures from fossil chironomid assemblages in northern Fennoscandia. Holocene 9: 279-294 Porinchu DF, MacDonald GM (2003) The use and application of freshwater midges (Chironomidae: insecta: diptera) in geographical research. Progr Phys Geogr 27: 378-422 Porinchu DF, MacDonald GM, Bloom AM, Moser KA (2003) Late Pleistocene and early Holocene climate and limnological changes in the Sierra Nevada, California, USA inferred from midges (Insecta: Diptera: Chironomidae). Palaeogeogr Palaeoclim Palaeoecol 198: 403-422 Ran ETH (1990) Dynamics of vegetation and environment during the Middle Pleniglacial in the Dinkel Valley (The Netherlands). Meded Rijks Geol Dienst 44-3: 141-205 Ran ETH, van Huissteden J (1990) The Dinkel Valley in the Middle Pleniglacial; dynamics of a tundra river system. Meded Rijks Geol Dienst 44: 209-220 Reille M, de Beaulieu J-L (1990) Pollen analysis of a long upper Pleistocene continental sequence in a Velay Maar (Massif Central, France). Palaeogeogr Palaeoclim Palaeoecol 80: 35-48 Reille M, de Beaulieu J-L, Svobodova H, Andrieu-Ponel V, Goeury C (2000) Pollenanalytical biostratigraphy of the last five climatic cycles from a long continental sequence from the Velay region (Massif Central, France). J Quatern Sci 15: 665-685 Rymer L (1973) Modern pollen rain studies in Iceland. New Phytol 72: 1367-1373 Schmid PE (1993) A key to the larval Chironomidae and their instars from Austrian Danube region streams and rivers. Part 1. Diamesinae, Prodiamesinae and Orthocladiinae. Federal Institute for Water Quality of the Ministry of Agriculture and Forestry, Wien, 524 pp Srodon A (1980) Paleohistoria olszy w Polsce (Palaeohistory of alder in Poland) (in Polish with English summary). In: Bia³obok S (ed) Olsze – Alnus Mill. Nasze drzewa leœne, PWN, Warszawa-Poznañ: 7-33 Stewart JR, Lister AM (2001) Cryptic northern refugia and the origins of the modern biota. Trends Ecol Evol 16: 608-613 Svenning J-C (2003) Deterministic Plio-Pleistocene extinctions in the European cool-temperate tree flora. Ecol Lett 6: 646-653 ter Braak CJF (1986) Canonical Correspondence Analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology 67: 1167-1179 ter Braak CJF, Smilauer P (2002) CANOCO reference manual and CanoDraw for windows user’s guide. Wageningen: Biometris Wageningen, 500 pp Vandenberghe J (1985) Palaeoenvironment and stratigraphy during the last glacial in the Belgian-Dutch border region. Quatern Res 24: 23-38 van der Hammen Th, Maarleveld GC, Vogel JC, Zagwijn W (1967) Stratigraphy, climatic succession and radiocarbon dating of the last glacial in the Netherlands. Geol Mijnb 46: 79-95

74 Chapter 4

van der Hammen Th (1971) The Denekamp, Hengelo and Moershoofd Interstadials. Meded Rijks Geol Dienst 22: 81-85 van der Meijden R (2005) Heukels’ flora van Nederland. Wolters-Noordhoff, Groningen van Geel B (1996) Factors influencing changing AP/NAP ratios in NW-Europe during the Late-Glacial period. Il Quaternario 9: 599-604 van Geel B, Grenfell HR (1996) Spores of Zygnamataceae (Chapter 7A). In: Jansoniu J, McGregor DC (Eds) Palynology: principles and applications. American Association of Stratigraphic Palynologists Foundation 1: 173-179 Vorren K-D (1978) Late and Middle Weichselian stratigraphy of Andøya, north Norway. Boreas 7: 19-38 Walker IR (1990) Modern assemblages of arctic and alpine Chironomidae as analogues for late-glacial communities. Hydrobiologia 214: 223 – 227 Walker MJC, Björck S, Lowe JJ, Cwynar LC, Johnsen S, Knudsen K-L, Wohlfarth B, INTIMATE group (1999) Isotopic ‘events’ in the GRIP ice core: a stratotype for the Late Pleistocene. Quatern Sci Rev 18: 1143-1150 Weeda EJ, Westra J, Westra Ch, Westra T (1987) Nederlandse oecologische flora, wilde planten en hun relaties 2. IVN, de Lange/Van Leer, Deventer. 304 pp Weeda EJ, Westra J, Westra Ch, Westra T (1991) Nederlandse oecologische flora, wilde planten en hun relaties 4. IVN, Salland/de Lange, Deventer. 317 pp Weeda EJ, Westra J, Westra Ch, Westra T (1994) Nederlandse oecologische flora, wilde planten en hun relaties 5. IVN, Lecturis BV, Eindhoven. 400 pp Wiederholm T (Ed) (1983) Chironomidae of the Holarctic region. Keys and diagnoses. Part I – Larvae. Ent Scan Supp 19 Willis KJ (1996) Where did all the flowers go? The fate of temperate European flora during glacial periods. Endeavour 20: 110-114 Willis KJ, Rudner E, Sümegi P (2001) The full glacial forest of Central and Southeast Europe. Quatern Res 53: 203- 213 Zagwijn W (1974) Vegetation, climate and radiocarbon datings in the Late Pleistocene of the Netherlands, Part II: Middle Weichselian. Meded Rijks Geol Dienst NS 25: 101-111

75 Chapter 4

76 Chapter 5

Chapter 5

Intraregional variability in chironomid-inferred temperature estimates and the influence of river inundations on lacustrine chironomid assemblages

S Engels, SJP Bohncke, O Heiri, M Nyman

Abstract Floodplain lakes are rarely analysed for fossil chironomids and usually not incorporated in modern chironomid-climate calibration datasets because of the potential complex hydrological processes that could result from flooding of the lakes. In order to investigate this potential influence of river inundations on fossil chironomid assemblages, 13 regularly inundated lakes and 20 lakes isolated from riverine influence were sampled and their surface sediments analysed for subfossil chironomid assemblages. The physical and chemical settings of all lakes were similar, although the variation in the environmental variables was higher in the lakes isolated from riverine influence. Chironomid concentration and taxon richness show significant differences between the two classes of lakes, and the variation in these variables is best explained by loss-on-ignition of the sediments (LOI). Relative chironomid abundances show some differences between the two groups of lakes, with several chironomid taxa occurring preferentially in one of the two lake-types. The variability in chironomid assemblages is also best explained by LOI. Application of a chironomid-temperature inference model shows that both types of lakes reconstruct July air temperatures that are equal to, or slightly underestimating, the measured temperature of the region. We conclude that, although there are some differences between the chironomid assemblages of floodplain lakes and of isolated lakes, these differences do not have a major effect on chironomid-based temperature reconstruction.

This manuscript was published OnlineFirst for Journal of Paleolimnology DOI 10.1007/s10933-007-9147-5

77 Chapter 5

5.1 Introduction The Chironomidae (Insecta: Diptera) is a very diverse family of aquatic insects, occurring in a multitude of habitats, and chironomids are frequently among the most abundant invertebrates found in lakes and rivers (e.g. Armitage 1995; Cranston 1995). Chironomids have been used in palaeoecological studies as a proxy for a range of different environmental variables. For instance, Gandouin et al. (2006; 2007) explored the ratio between lotic and lentic chironomid taxa as a tool to qualitatively 2 reconstruct flow-regimes in dead branches of two French rivers (Rhône and Garonne). Other research projects used fossil chironomid assemblages in lake sediments to quantitatively infer past changes in July air temperature (e.g. Heiri and Lotter 2005; Brooks 2006; Walker and Cwynar 2006), salinity (e.g. Heinrichs and Walker 2006; Eggermont et al. 2006), or trophic conditions in lakes (Lotter et al. 1998; Brooks et al. 2001). Chironomid-based temperature reconstruction has received increasing attention in recent years and a large number of chironomid-based July air temperature records are now available from formerly glaciated regions in Europe and North America (e.g. Walker et al. 1991; Brooks and Birks 2001). However, lacustrine sediments deposited in floodplain environments such as palaeochannels and oxbow lakes have to our knowledge not previously been used for quantitative temperature reconstruction based on fossil chironomids. Floodplain sediments have great potential for chironomid-based temperature reconstructions. They can be found in landscapes beyond the maximum extent of late Quaternary glaciations (Gandouin et al. 2005; 2006) and in time windows for which other lacustrine sediment records are rare. For example, in the opencast lignite mines in eastern Germany, lacustrine sediments dating back to Oxygen Isotope Stage (OIS) - 5a and early OIS-3 (ca. 80 and 55 ka BP, respectively) have been retrieved (e.g. Mol 1997; Bos et al. 2001; Kasse et al. 2003). The fossil chironomid and botanical assemblages, together with the sedimentological record, all suggest that these palaeolakes were situated on a river floodplain. Modern lakes, where high-discharge events of nearby rivers or streams can reach the lake, are usually not included in the modern calibration sets used to develop chironomid-temperature inference models because of the complex ecological processes that can be the result of such floods. As a consequence, lacustrine floodplain sediments potentially present non-analogue situations in respect to most modern chironomid-temperature transfer functions and it is presently unclear to what extent this influences quantitative temperature reconstructions. Studies on modern lakes have shown that certain lakes, and their associated chironomid faunas, are sensitive recorders of small-scale climate changes whereas other lakes are less likely to respond to climate variability, depending on their (relative) location (e.g. Velle et al. 2005; Heegaard et al. 2006). Within-region discrepancies are also recorded for down-core analysis of chironomid-assemblages and their associated inferred temperatures (e.g. Bigler et al. 2002; Korhola et al. 2002; Velle et al. 2005). Presently, it is unclear how large intraregional variability of

78 Chapter 5 chironomid-inferred temperature estimates can be, or whether this variability changes between floodplain lakes and lakes that are unaffected by inundations. In this study, we investigate the differences between the physical and chemical properties of lakes on a river floodplain that are annually inundated, and lakes that are isolated from such riverine influence. The subfossil chironomid assemblages of these lakes are compared and the influence of river inundations on the chironomid concentrations, taxon richness and relative chironomid abundances is explored. The intraregional variability of chironomid-inferred temperature estimates is determined and compared for the two classes of lakes with the aim of assessing whether chironomid-temperature inference models, developed for lakes unaffected by riverine influence, can be applied to floodplain sediment records.

5.2 Study area The 33 lakes studied in this project are all situated within a 6 km radius of the small town of Kaamanen, Finland (69º06’ N, 27º10’ E; 146 m a.s.l.). The landscape of the region is characterised by outcrops of metamorphic rocks (granite, gneisses), interspersed with lakes, wetlands and mires. Today, the area around Kaamanen is situated in boreal forest, in which birch (Betula pubescens) and pine (Pinus sylvestris) are the most important tree species. Mean July air temperature at Kaamanen is 13.1 ºC and the annual precipitation averages 441 mm (over the period of 1971 – 2000; Drebs et al. 2002). The floodplain of the Muddusjoki river (Figure 5.1), the largest river of the area, is covered in sedges and other wetland vegetation. The Muddusjoki is known to have fluctuating discharges (and water levels), the highest discharge being in May or June during the snow melt season. The average difference between the highest and the lowest monthly water level during the period of 1971 – 2005 was 137 cm (unpublished data, P. Räinä). As 13 studied lakes are situated on the same elevation as the river, they are inundated by the river during every flood. This was also evident from flood marks found during the fieldwork (e.g. evidence of erosion or debris deposits at a uniform height throughout the study area), and confirmed by local inhabitants. The other 20 lakes lie either higher or sheltered behind rock outcrops, and therefore remain isolated from riverine influence. Lake 33 differs from the other lakes in the data set as it has probably originated from the construction of a road close to the lake, and is fed by a drainage system that directs rain water away from the road and into the lake.

5.3 Methods 5.3.1 Fieldwork and laboratory analyses During a 3-week fieldwork campaign in August 2006, 33 ponds and small lakes were sampled for water chemistry and surficial sediments. Vegetation, lake perimeter, number of in- and outflows and possible anthropogenic influence were recorded. Lake bathymetry was established for every lake using a GPS and a water-depth gauge. At the deepest point of the lakes, water temperature, conductivity, pH and O2 profiles were recorded before sampling the lake sediments. Using a HON/ Kajak -

79 Chapter 5

27º05' 27º10' 27º15'

L30 L24 L29

2 L32 L31 L33

L23 L06

L05 L12

L21 L25 L04 L22

69º05'00"

L15 L14 L27 L20 L03 L19 L18

L02 L01 L11 L17 L10 L07

L08 L09 L13

L16

L28

69º02'30" L26

m 0500 1000

Figure 5.1: Location of the 33 studied lakes in northern Finland. Inundated lakes have a shaded text box, isolated lakes a white text box. The shaded areas (light gray) represent lakes and rivers/ streams.

80 Chapter 5 corer with a diameter of 6.2 cm (Renberg 1991), at least two cores were retrieved from every lake, and the sediment column was subsampled in 1-cm increments in the field. The topmost one centimetre of sediment was used in this study for loss-on-ignition (LOI was determined following standard protocols of the Dutch Centre for Normalization at 550 ºC using a CEM MAS-300) and chironomid analysis. Water samples were collected at 50 cm below the water-surface for every lake, and in the case of lakes deeper than 2 m or with thermal stratification, an additional water sample was retrieved from the lower part of the water column. Alkalinity of the water was measured in the field, whereas other water chemistry variables were measured in laboratories using a Dionex DX 120 Chromatograph and a Varian ICP- AES. δ18O of the water samples was measured using a ThermoFinnigan Delta+ mass spectrometer with a GASBENCH II preparation device at a 1σ precision of ~0.1‰.

5.3.2 Subfossil chironomid remains Prior to chironomid analysis, the sediments were freeze-dried in order to obtain a reliable estimate of chironomid concentrations per weight. Freeze-dried samples (0.019 -1.818 g) were treated with cold KOH (10%) for a period of at least 4 hours, before sieving over a 100 µm mesh. Chironomid remains were hand-picked from the residue, and mounted on permanent slides using Euparal© mounting medium. After omitting chironomid head capsules (hc) that were only identified to subfamily or tribal levels, minimum, maximum and mean count numbers per sample were 149, 415, and 184 hc respectively. Taxonomy follows Sæther (1976), Wiederholm (1983), Schmid (1993), Heiri et al. (2004) and Brooks et al. (2007). Using TILIA and TG.VIEW (Grimm 1991-2004), a chironomid abundance diagram was constructed.

5.3.3 Numerical analyses In order to compare chironomid assemblages from lakes that are regularly inundated and lakes isolated from riverine influence, a range of statistical analyses was performed, and the results are compared for the two classes of lakes. Rarefaction analysis was performed to assess the taxon richness (TR) of the different samples. This method simulates a random selection without replacement, estimating for all samples the taxon richness for the smallest count size recorded in the entire sequence of samples. In our case, this smallest sample contained 149 identified hcs, and the TR was determined for all 33 lakes for this count size using RAREPOLL (Birks and Line 1992). A t-test was performed to assess whether TR significantly differed between the two classes of lakes. In this study, 24 environmental, chemical and biological variables were determined for all 33 lakes (all variables on interval/ ratio scale from Table 1), and using redundancy analysis (RDA; Ter Braak and Looman 1994) with Monte Carlo permutation tests (unrestricted model, 999 permutations), the environmental variables that best explained the variance in TR were selected. Chironomid concentrations were compared for the two classes of lakes using a t-test. Similar to the TR-analyses, RDA was performed in order to select the environmental variables that best explained the variance in chironomid

81 Chapter 5

concentrations. As the chironomid concentrations showed a skewed distribution, they were log-transformed prior to analysis. The dataset containing the 24 environmental variables was subjected to a series of Principal Component Analyses (PCAs) in order to detect co-variation between the different variables, and to assess differences of the chemical and physical variables between our 2 classes of lakes. Environmental variables that show a skewed distribution (i.e. all water chemistry data except for [Na], [K], [Cl] and [Ca]) were log- 2 transformed prior to the analyses. Inspection of initial results showed a high co- variability between a number of variables (especially for the water chemistry data), and for subsequent PCAs redundant variables were omitted in order to create a biplot that was more readily interpretable. The omitted variables included all δ18 variables presented in Table 1b, except for O, TP, [NH4], [SO4] and [PO4]. Using detrended correspondence analysis (DCA; Hill and Gauch 1980), with detrending by segments, non-linear rescaling of the axes and down-weighting of rare taxa, the amount of compositional turnover in the chironomid data was determined. As the gradient length of the first axis was 2.08 standard deviation units, the application of both unimodal and linear methods was possible in subsequent analyses (Birks, 1998). Square-root transformed percentage abundances of the chironomid taxa, with down-weighting of rare taxa, were used in a correspondence analysis (CA) to explore the distribution of the different chironomid taxa in the two lake classes. A series of direct gradient analyses (canonical correspondence analyses (CCAs)) was performed to determine the unique and marginal effects of the individual environmental variables on the variability of the chironomid abundances. The marginal effect of a given environmental variable is defined as the amount of variability in the chironomid data that would be explained by a constrained ordination model that uses this environmental variable as the only explanatory variable (Lepš and Šmilauer 2003). The unique (or conditional) effects are the partial effects of the selected variables (tested through partial Monte Carlo permutation tests), which depend on the variables already selected (Lepš and Šmilauer 2003). A 2-component weighted averaging-partial least squares (WA-PLS) chironomid-temperature inference model was applied to the chironomid assemblages of our 33 lakes, in order to test the intraregional variability in chironomid-inferred temperature estimates, and to assess the influence of river inundation on quantitative temperature inferences. This model was based on an expanded and re-analysed training set by Olander et al. (1999) including 63 lakes in western Finnish Lapland (Vasko et al. 2000; Nyman 2007). The warmest lake in this dataset was identified as an outlier, and excluded from further analyses. The calibration data covers a mean July air temperature range of 7.9 to 14.9 ºC (1961-1990 Climate Normals data). All chironomid percentage values were square root transformed to stabilize the variance, and the model has a coefficient of determination (r2) of 0.75, a root mean square error of prediction (RMSEP) of 0.76 ºC, a mean bias of 0.023 ºC and a maximum bias of 1.19 ºC (all as assessed through leave-

82 Chapter 5 on-out cross validation). Sample specific errors were estimated through bootstrapping (999 cycles; Birks et al. 1990). The WA-PLS model was developed and implemented using C2 version 1.4.3 (Juggins 2003).

5.4 Results 5.4.1 Taxon richness and chironomid concentrations The taxon richness (TR) estimated for a count sum of 149 hcs (equivalent to our minimum count sum), ranges between 14 and 43 for the 33 lakes. Using RDA with forward selection, LOI was selected as the environmental variable explaining the λ most variance in TR ( =0.59 of total inertia of 0.90, p = 0.001). O2 was the second forward selected variable with p = 0.022, explaining an additional 10% of the total variance. The relatively strong (linear) correlation (r2 = 0.58) between the LOI and TR is plotted in Figure 5.2a. Most inundated lakes have a lower LOI than the isolated lakes, but there also seems to be a tendency for the isolated lakes to have a lower TR. The hypothesis that both groups of lakes have statistically different TR values was tested using a t-test. The resulting t = 3.56 and p = 0.0012 show that the chance of a type I error (false rejection of H0) is lower than 5%, and H0 (= both groups have the same means) is rejected.

a) r2 = 0.579 40.0

30.0 on richness

Tax 20.0

0 10 20 30 40 50 60 70 80 90 Loss-on-ignition (%) b) r2 = 0.574

10000

1000 (hc/g)

100 Chironomid concentrations

0 10 20 30 40 50 60 70 80 90 Loss-on-ignition (%)

Figure 5.2: a) Taxon richness and b) chironomid concentrations plotted against loss-on-ignition. Open circles indicate inundated lakes, solid circles isolated lakes; regression lines are linear models.

83 Chapter 5 85 85 type A indicates mperature of the 800 816 384 436 111 915 708 940 317 726 594 880 1313 1289 1993 1989 1030 9840 1454 hc/g 1701 1818 8857 9840 4538 1847 4873 1523 2960 1895 3243 1141 4109 1505 1120 the detection limit. Chironomid Chironomid 2 concentration % 75 79 83 77 59 Sediment content Alkalinity LOI 2 9.0 0.2 7.5 0.0 5.4 0.0 8.7 0.0 8.7 0.3 8 7 8 8 9 µS/cm ppm mg/l

T water T water pH EC O depth Water ha m ºC ha m area Open water Vegetation

155 152 163 152 149 m a.s.l. Elevation Physical parameters Water column Physical parameters Water nr Lake Class L16 Isolated B 1.6 2.3 16.7 2.9 16.7 2.3 L04 1.6 152 Isolated B B L15 6.6 L16 Isolated 6.4 16.9 4.9 10 Inundated 9.8 146 A 0.6 60 2.9 2.6 18.0 5.5 19 6.9 0.6 36 L055.0 19.9 1.0 B Isolated 0.3 L12 Inundated 148 B 1.9 L23 0.9 18.7 6.2 37 8.6 158 Isolated 1.4 18 B 15.8 6.9 14.4 5.4 18 8.1 0.6 40 L31 Isolated B 14.3 3.3 15.8 3.9 15.8 3.3 14.3 L24 B 152 Isolated L31 Isolated B 0.9 3.1 14.0 5.9 47 7.6 0.6 60 L02 L03 Inundated 146 Inundated 146 A A 0.2 3.0 1.6 18.7 5.2 29 3.4 16.7 5.3 20 9.6 8.6 2.4 44 0.9 32 Minimum 146 0.1 0.8 9.7 2.9 7 3.1 0.0 9 Mean 151 3.3 3.0 16.1 5.1 24 8.2 0.8 48 L06 146 Isolated L14 B L17 2.0 Inundated 146 1.8 19.0 4.4 10 A 9.1 156 Isolated 2.7 0.3 69 B 1.6 18.0 5.4 19 1.6 L29 7.1 3.9 16.8 3.5 10 0.9 34 8.0 153 Isolated 0.0 70 B 3.2 1.5 14.4 6.1 41 9.7 1.5 52 Maximum 163 15.8 15.9 19.9 6.3 63 10.2 2.4 82 L22 Inundated 147 A 3.3 3.1 15.7 5.0 21 8.9 0.4 30 L09 Inundated 146 A 5.9 2.5 17.5 5.7 24 7.8 1.7 31 L28 Isolated B 0.5 1.8 15.8 3.5 15.8 1.8 0.5 B L26 L28 Isolated 153 Isolated L33 B 2.5 156 Isolated 2.1 15.7 4.8 11 B 8.8 0.4 0.7 70 2.3 13.0 5.9 51 10.1 1.3 27 L11 Inundated 146 A 0.9 0.9 17.9 5.0 10 6.7 1.0 36 L13 Isolated B 2.3 5.0 18.0 4.1 18.0 5.0 2.3 B L10 L13 Isolated Inundated 146 A 0.1 0.8 17.9 5.6 22 8.9 0.9 10 L08 Inundated 146 A/B 2.9 1.2 18.9 6.3 31 9.2 1.0 37 L01 Inundated 146 A 2.3 2.3 16.0 4.4 17 9.3 1.6 53 L19 158 Isolated B 0.5 3.0 15.5 5.5 29 9.3 0.6 56 L21 147 Inundated A 10.7 1.6 15.6 5.2 17 10.0 0.6 39 0.6 10.0 17 5.2 15.6 1.6 L07 10.7 Inundated A 146 A/B Inundated 6.3 L18 2.0 18.1 5.8 23 L20 L21 147 8.9 158 Isolated 0.8 44 L25 157 Isolated B L27 B 4.5 L30 151 Isolated 15.9 1.8 L32 157 Isolated 9.7 5.3 28 B 4.1 6.8 13.9 5.8 35 154 Isolated B 5.6 0.9 1.2 41 156 Isolated B 3.3 0.4 0.8 30 10.6 4.4 33 B 5.4 3.1 11.3 5.4 63 6.7 5.2 1.7 0.1 0.9 26 15.3 5.9 21 10.2 1.0 1.3 47 17.0 4.4 44 9.3 0.5 55 0.5 74

Physical and chemical parameters of the 33 lakes in data set. Lake numbers refer to indicated Fig. 1. Vegetation

Table 5.1a a vegetation dominated by sedges, type B pine and birch. A value of 0 refers to concentrations below Minimum, mean and maximum values are presented for each variable (where applicable). Abbreviations include: T water: average te water column; LOI: loss-on-ignition; hc: head capsules; EC: electric conductivity

84 Chapter 5

A t-test performed on the chironomid concentrations (hcs encountered per gram of freeze-dried sediment) of the two groups resulted in a t = -2.49, where t- critical (α = 5%) = 2.2. This indicates that the mean chironomid concentration in the group of isolated lakes is significantly different than that of the inundated lakes. The relationship between the chironomid concentrations of the 33 lakes and the environmental variables was examined using RDA. LOI was selected as the variable with the strongest relationship with chironomid concentrations (explaining λ=0.57 of total inertia of 0.84, p = 0.001). Forward selection showed that [Cl] had a significant relationship (p < 0.05) with the chironomid concentrations, explaining an additional 10% of the variance in the data. Figure 5.2b shows the log-transformed chironomid concentration of the individual lakes, plotted against LOI.

5.4.2 Physical and chemical properties of the lakes PCA was performed on the physical variables, water column chemistry and LOI data presented in Table 5.1a, and the chemical data presented in Table 5.1b. Variables that were excluded from the analysis were chemical variables that showed a large co- variance with other variables retained in the analysis, and the data on nominal/ ordinal scale. The variance explained by the first three constructed PCA axes is: 26.3%, 18.0% and 14.5%. Figure 5.3a shows a biplot of the environmental variables and the sites on the first two axes. Water chemistry variables are strongly correlated with the first PCA-axis, whereas physical variables of the lake and its catchment dominate the second axis.

Marginal effects Conditional effects Variance Variance Variable P P explained explained LOI 15.1% 0.001 15.0% 0.001 Elevation 7.4% 0.002 7.0% 0.001 O2 6.2% 0.008 5.3% 0.001 TOC 5.5% 0.017 4.4% 0.003 Depth 5.7% 0.019 4.4% 0.004 EC 7.1% 0.002 4.4% 0.010 SO 4 4.9% 0.059 Ca 6.1% 0.008 Mg 8.3% 0.001 T water 6.4% 0.010 PO4 3.2% 0.348 Open water 4.5% 0.096 δ 18O 10.0% 0.001 Alk.Acid 7.4% 0.003 K 3.4% 0.351 NH4 4.1% 0.154 Na 3.8% 0.223

pH 5.6% 0.023 Cl 4.2% 0.196 Si 6.9% 0.001 Mn 3.2% 0.398 Fe 3.5% 0.280 Al 1.8% 0.970 TP 1.9% 0.904

Table 5.2 Marginal and conditional effects of explanatory variables. Statistically significant relationships are plotted in bold

85 Chapter 5

4 3.1 0.0 SO mg/l 24.6 4 2.4 PO mg/l 13.4 202.5

2 4 NH mg/l 0.041 0.000 0.197 Cl 1.4 0.5 8.5 mg/l 4 0 1.7 6 0.068 9 0.0 1 1.1 5.4 0.122 1.0 0.000 0.9 20.8 1.0 0.053 3.6 0.125 8 1.3 10.2 9 21.3 1.0 9 1.0 1.2 4 0.5 2.5 0.006 0 1.3 1 0.008 0.7 7.2 0.000 2 0.9 2.4 0.013 6 0.8 1.7 4.3 0.013 0.7 0.7 3.7 0.013 1.3 31.2 2.5 0.005 37.9 1.2 9 0.040 1.0 4.6 3 1.1 4.2 1.2 1.4 1.0 1.1 0.008 0.024 7 9.3 3.6 2 0.9 0.5 1 0.7 9 0.038 0.9 9 2.4 3.3 0.012 9 1.4 0.052 0.9 1.0 3.7 0.040 8.5 3.0 0.127 1.0 3.3 0.117 16.7 5.5 2.5 3.6 1.6 4.4 10 10 1.1 0.000 05 4.7 00 05 1.1 2.8 1.1 0.038 0.9 0.010 3.7 0.047 202.5 1.4 3.1 2.5 2.4 41 00 08 1.3 1.3 0.021 1.4 0.104 00 4.5 19 0.037 4.2 08 1.1 4.5 4.8 2.2 4.0 00 0.003 1.6 3.9 0.197 3.9 0.005 1.8 11.4 1.1 2.9 0.026 24.6 0.0 3.4 08 4.3 1.8 0.001 4.0 4.2 TP mg/l 0.012 0.000 0.080 Si 1.0 0.0 4.5 mg/l Mn mg/l 0.012 0.000 0.084 Al mg/l 0.07 0.02 0.19 Fe mg/l 0.19 0.00 0.59 1.9 0.0 5.6 Ca mg/l 0.7 0.0 2.0 Mg mg/l K mg/l 0.37 0.15 0.73 1.2 0.1 2.6 Na mg/l O 18 -9.4 -5.8 δ -14.4 SMOW 5.05 1.33 8.95 5.51 4.35 5.00 4.72 -8.3 -11.8 5.42 3.34 1.8 -9.6 1.6 4.23 -8.8 0.43 4.27 1.3 -7.4 0.41 4.21 2.6 -6.8 0.6 0.40 -10.6 5.64 0.8 1.0 0.57 -10.9 5.22 0.8 1.4 0.8 0.37 1.4 -10.8 5.30 2.8 0.6 0.33 0.1 1.65 0.11 -9.9 0.42 1.7 0.2 0.20 1.2 5.46 -8.3 0.17 2.0 0.04 0.2 5.70 0.16 0.11 1.4 1.1 -9.4 0.21 0.7 0.005 3.63 0.12 0.7 0.0 -6.3 0.012 0.6 0.04 0.36 6.47 0.06 1.9 1.7 1.0 -8.9 0.19 0.1 0.005 0.20 3.22 0.04 0.1 0.6 2.9 -8.7 0.10 1.0 0.20 0.003 0.55 0.00 3.84 1.5 1.1 -6.2 0.0 0.09 0.6 0.4 0.06 0.004 0.23 5.17 0.05 1.0 -6.7 1.9 0.3 1.2 0.13 0.006 0.33 0.08 -11.3 7.38 0.04 0.5 0.011 1.4 0.1 0.2 0.30 0.00 -11.3 7.01 0.20 0.04 0.8 0.003 4.0 0.1 0.7 0.31 0.00 0.4 -11.6 6.46 0.38 0.4 0.005 0.2 0.05 0.8 0.36 0.01 1.5 2.92 0.18 0.8 -9.5 0.26 0.0 1.7 0.03 0.2 0.006 1.6 8.95 0.59 0.7 -9.4 0.41 0.0 1.7 0.05 0.2 0.012 -10.5 7.64 0.20 1.2 0.6 0.41 0.0 0.4 0.11 0.3 0.009 -10.5 4.92 0.26 1.3 1.0 0.4 0.03 0.1 0.056 0.36 0.01 1.1 -14.0 6.41 0.04 0.0 1.2 0.13 1.8 0.010 0.35 0.00 0.2 5.68 0.07 2.7 -7.1 0.32 0.06 0.1 0.8 0.00 0.009 0.01 1.7 -13.1 5.94 2.9 0.16 0.03 0.2 0.8 0.04 0.015 0.03 3.75 0.07 1.0 0.6 -7.6 0.42 1.8 0.3 0.12 0.010 0.00 2.1 5.76 0.02 0.1 -9.8 0.000 1.8 0.0 0.15 0.02 1.33 0.54 1.7 0.00 0.7 1.6 -8.0 0.000 0.64 0.0 0.00 0.26 1.8 1.1 2.2 -6.8 0.003 0.13 0.5 0.25 0.32 0.00 3.1 1.5 0.5 2.0 -5.8 0.0 0.11 0.01 0.004 0.65 -14.4 0.05 0.7 2.8 0.0 0.6 0.3 0.56 0.005 0.44 4.5 0.04 1.5 0.001 0.0 0.3 1.2 0.26 1.9 0.12 0.08 0.001 0.4 0.3 0.6 0.09 0.45 0.01 0.9 0.004 0.73 3.7 0.05 0.2 0.01 0.41 0.12 1.3 0.0 1.7 1.0 0.003 0.08 4.5 1.3 0.041 0.0 0.4 0.10 0.05 0.0 4.4 0.05 0.1 0.043 0.05 5.6 3.9 0.03 0.004 0.01 0.58 0.0 0.04 0.2 0.01 0.001 0.11 0.9 0.004 0.01 0.05 1.7 0.084 0.01 0.000 0.0 0.00 0.2 0.00 4.5 0.01 0.0 mg/l TOC nr Lake L01 L01 L02 L03 L04 L05 L06 L07 L08 L09 L10 L11 L12 L13 L14 L15 L16 L17 L18 L19 L20 L21 L22 L23 L24 L25 L26 L27 L28 L29 L30 L31 L32 L33 Mean Minimum Maximum Chemical parameters of the 33 lakes in data set. Lake numbers refer to indicated Fig. 5.1. A value 0 refers concentrations below

Table 5.2b the detection limit. Minimum. mean and maximum values are presented for each variable

86 Chapter 5

Axis 1 2.0 a) Elevation Depth PCA

1.0 LOI SO4 Open wat EC TPNH4 0.0 TOC Axis 2 Axis 18O Alk.Acid

pH -1.0 O2 PO4

T water -2.0 -2.0 -1.0 0.0 1.0 2.0 b) 1.5 CA

23 1.0

18

0.5 32 33 24 10 19 26 12 28 5 25 4 29 30 6 Axis2 27 0.0 13 9 20 17 31 16 22 11 8 15 7 2 -0.5 3 14 21 1

-1.0 -1.0 -0.5 0.0 0.5 1.0 1.5 Axis 1 Figure 5.3: a) Principal components analysis (PCA) biplot of selected environmental variables and sites. The variance explained by the first and second axis is 26.3% and 18.0% respectively. b) Correspondence analysis (CA) scatter- plot of sites. The variance explained by the first and second axis is 18.8% and 11.2% respectively. Open circles indicate inundated lakes, solid circles isolated lakes.

The third axis (not shown here) shows a strong correlation with cation concentrations. As can be seen in Figure 5.3a, the regularly inundated lakes cluster together in the PCA, implying that lakes belonging to this class have a similar physical and chemical environment. The means of the two different classes are similar for the variables that correlate strongly with the first axis, but there is a difference in lake-scores on the second axis: the inundated lakes are on the negative side of the axis, the isolated lakes on the positive side. Further analyses show that when “elevation” is omitted from the dataset, this difference in lake-scores on the second axis disappears. Therefore, the lakes in both classes seem to have similar means for most chemical and physical variables, but the larger scatter of PCA axis scores (Figure 5.3a) indicates that the variability in values is larger for the isolated group.

87

Chapter 5

-type

-type

Stempellinella Micropsectra radialis Micropsectra

20 40% Corynocera ambigua Corynocera

20% spp.

2 zalutschicola Zalutschia

Cladopelma -type

Corynoneura

hyes -type

Cricotopus

nata

type B type Limnop

20%

-type Zalutschia Zalutschia

s i

Zaltuschia mucro Zaltuschia

-type I -type ipes lidicorn

20%

mancus

Microtend

Tanytarsus pal Tanytarsus

tanytarsus

a orophila a type A type

ll 20%

e

Clado

-type

ns

Pagasti Parakiefferiella

20%

Tanytarsus luge Tanytarsus -type

20% -type Procladius

20%

-type

Tanytarsus mendax Tanytarsus -type

20%

-type

Chironomus anthracinus Chironomus

bathophila

a

iell Orthocladius

Sergentia

Parakieffer

-type

20%

Paratanytarsus penicillatus Paratanytarsus

smyia

ntrionalis

20%

Ablabe -type

20 40% type Psectrocladius septe Psectrocladius

20% Psectrocladius sordidellus Psectrocladius

20%

Dicrotendipes nervosus- Dicrotendipes

type VI type

Pseudochironomus

-type type

20%

Tanytarsus Tanytarsus

Lauterborniella/ Zavreliella Lauterborniella/

psectrocladius

lo Al

20%

Tanytarsus glabrescens- Tanytarsus type A type

20 40%

Zalutschia Zalutschia

Chironomid abundance diagram of selected taxa. Sites are arranged according to lake-class and subsequently according to loss-on- to according subsequently and lake-class to according arranged are Sites taxa. selected of diagram abundance Chironomid Lake L10 L12 L22 L09 L03 L14 L15 L11 L08 L21 L02 L07 L01 L25 L33 L20 L23 L18 L27 L29 L30 L19 L31 L04 L24 L06 L26 L17 L32 L05 L28 L13 L16

low low high high LOI LOI

Inundated lakes Inundated lakes Isolated

ignition value of their sediments. their of value ignition Figure 5.4: 5.4: Figure

88 Chapter 5

5.4.3 Chironomid abundance data Correspondence analysis (CA) showed that the isolated lakes and the regularly inundated lakes both cluster together in separate groups when using the chironomid- data as input (Figure 5.3b), implying that the chironomid faunas of the two classes of lakes differ. The colder lakes (Lake 23 and Lake 33), and those showing strong thermal stratification and bottom-water anoxia during the fieldwork period (Lake 18, Lake 25 and Lake 27), plot outside of the cluster of isolated lakes. The variance explained on the first two axes plotted in Figure 5.3b is 18.8% and 11.2%. In an initial series of CCAs with 24 exploratory variables, 13 variables were manually selected as explaining a statistical significant amount of the total inertia (p<0.05). Of these variables, LOI was selected as the variable that is most strongly related to variation in the chironomid abundances (15.1%; Table 5.2). This was followed by δ18O (10%), Mg (8.3%), and elevation (7.4%). A series of CCAs with forward selection was performed to determine the marginal and unique effects of the different environmental variables (Table 5.2). Again, LOI was selected as the most important explanatory variable, now followed by elevation, O2, TOC, lake depth and electric conductivity (EC). The first axis of a CCA with these 6 variables now explains 16.9% of the total inertia (λ = 0.192 of total 1.135).

A total of 86 chironomid taxa were identified for the 33 lakes from northern Finland. Figure 5.4 shows the abundance of selected taxa per lake in inundated and isolated lakes. Cricotopus-type to enhance the readability of the Figure.

15 Inundated lakes Isolated lakes

14

13

12

11 Mean July air temperature

10

9 5 7 4 6 8 1 4 21 23 2 33 L01 L02 L03 L07 L08 L09 L10 L11 L12 L1 L15 L L22 L04 L0 L06 L13 L16 L1 L18 L19 L20 L L L25 L2 L27 L2 L29 L30 L3 L32 L Lake id (see Fig. 5.1)

Figure 5.5: Chironomid-inferred mean July air temperature estimates in relation to the observed temperature for the Kaamanen region (horizontal line; Drebs et al. 2002). The sample specific error of prediction was determined by bootstrap cross-validation (999 cycles; Birks et al. 1990).

89 Chapter 5

Although the most abundant taxa all occur in both the inundated and the isolated lakes, there are certain taxa that show a preference for one of the two groups. Cladotanytarsus mancus-type I, Cladopelma, Cricotopus-type, Microtendipes and Tanytarsus pallidicornis-type all show a preference for the regularly inundated lakes, whereas Dicrotendipes nervosus--type, Lauterborniella/ Zavreliella, Paratanytarsus penicillatus-type, Sergentia, Tanytarsus type VI and T. glabrescens-type show higher abundances in the isolated lakes. The differences within the genus Zalutschia are also striking: whereas 2 Zalutschia type A almost exclusively occurs in isolated lakes with a high organic matter content of the sediments, the other identified taxa (Zalutschia type B, Z. mucronata-type and Z. zalutschicola) all are restricted to the inundated sites or sites with a low LOI. The different types within the genus Psectrocladius all show a preference for the isolated lakes, but they also feature a strong correlation with LOI: in both classes of lakes we see increasing abundances of e.g. P. septentrionalis-type with increasing LOI (Figure 5.4).

5.4.4 Intraregional variability in temperature inferences The results of applying a chironomid-temperature inference model to our 33-lake dataset are shown in Figure 5.5. The inferred temperatures are all around or lower than the measured 30 yr average July air temperature for Kaamanen (Drebs et al. 2002), and sample specific errors are ± 0.79-0.88 ºC. The group of inundated lakes shows less diverse temperature estimates and values closer to the observed temperature than the group of isolated lakes. The standard deviation of the temperature-inferences within the group of inundated lakes is 0.32 ºC, and within the group of lakes isolated from riverine influence 0.54 ºC. This low variability in inferred temperatures was expected for the group of inundated lakes, as the variation in the chironomid-abundance data was also low in this group (Figure 5.3b). Lake 33 can be considered as an outlier, as it has a reconstructed temperature of 10.6 ºC, approximately 1 ºC lower than the second lowest reconstructed value.

5.5 Discussion 5.5.1 Inundated lakes versus isolated sites: environment Thomaz et al. (2007) state that river floods result in the homogenization of aquatic habitats on the floodplain. An increased connectivity between the river course and the floodplain will enhance the exchange of water, sediments, minerals and organisms. During periods of low water, the heterogeneity of habitats on a floodplain may be increased through (amongst others) local water inputs from tributaries and ground water seepage (Thomaz et al. 2007). In our dataset, it is clear that the chemistry of the different lakes within the group of floodplain lakes is very uniform. In the PCA analysis (Figure 5.3a), the lakes of the inundated class all cluster close together, and the variation in scores on both the first and the second axis is low, implying a low within-group variability in environmental conditions. The lakes that are isolated from riverine influence are plotted in a separate group in the PCA diagram (Figure 5.3a). The scores of the lakes on the first axis are similar to those of

90 Chapter 5 the inundated class, but the variability in scores is higher, suggesting a higher variation in water chemistry. This is consistent with the expected higher variability due to the more diverse local processes at the isolated sites. The scores on the second axis differ between the two groups, which can be mainly attributed to floodplain characteristics: first, all lakes in the inundated class have the same elevation, as they are situated closely together on the same floodplain. Second, most of these lakes are shallow, as deep lakes on a floodplain will be preferentially filled in. Thirdly, other environmental variables that are correlated to the second axis, e.g. O2, are for some lakes related to depth as well (e.g. bottom water anoxia in deep, stratified lakes with a small area/wind fetch). This implies that there is a tendency for the floodplain lakes to have similar habitats available for chironomids and other aquatic macrofauna, and that the variability in environmental conditions is higher in lakes isolated from riverine influence.

5.5.2 Inundated lakes versus isolated sites: chironomid fauna In western Finnish Lapland, the chironomid taxon richness is most strongly correlated with alkalinity (Nyman et al. 2005), which is usually directly connected to weathering intensity of catchment bedrock. In their study, Nyman et al. (2005) assessed chironomid assemblages over a wide range of alkalinity values, whereas the range in alkalinity is limited in our dataset. The variation in TR in our dataset is best explained by LOI and the TR is significantly different in the two lake classes. The floodplain lakes, that on average have a low LOI, are situated at an ecotonal boundary between lotic and lentic habitats, which could result in a more diverse fauna and also a higher TR in this class. The positive correlation between LOI and the concentration of chironomid remains might be the result of higher (inorganic) sedimentation rates in the regularly inundated lakes as a result of the regular flooding of the lakes. This would result in a lower chironomid concentration per weight, and the correlation between LOI and the concentration of chironomid remains would then be directly related to floodplain processes. It is also possible that the more organic-rich sediments that are present in the lakes that are isolated from riverine influence provide for more food or suitable habitats for a wider range of chironomid taxa, and are possibly characterized by a higher chironomid production.

In all Fennoscandian datasets that describe the distribution of chironomid assemblages in lake surface sediments (e.g. Larocque et al. 2001; Nyman et al. 2005), LOI has been reported as a strong explanatory variable for the differences in relative chironomid abundance. However, in these datasets, LOI is correlated with summer temperatures, and the datasets have been designed to cover a large summer temperature gradient. Therefore, LOI also covers a large range of values. A Canadian chironomid-temperature calibration dataset, where the LOI-range is smaller (Walker et al. 1991), shows that the explanatory power of LOI is much lower than in the

91 Chapter 5

Fennoscandian datasets. In the results presented here, the LOI-range is large despite the similar local climatic circumstances; the large range of LOI is probably the result of the riverine input of clastic material in the floodplain lakes, and diverse local settings in the isolated lakes. Many chironomid larvae live as infauna in the lake sediment. The organic matter content of these sediments may affect the ability of different chironomid taxa to burrow into the sediments (Lindegaard 1997; Larocque et al. 2001). Furthermore, a 2 number of chironomids are detritus feeders (Pinder 1986; Korhola et al. 2000) and substrate composition may affect the success of this feeding strategy. Thus, the organic matter content of lake sediments can have a direct impact on the chironomid populations (Bigler et al. 2006). We can also see this in Figure 5.4, where a number of species show a preference for either low or high organic content of the lake sediments. Therefore, it is likely that the variations in relative chironomid abundances are directly related to sedimentary organic matter content and thereby (indirectly) driven by river inundations.

Interestingly, chironomid assemblages in the studied lakes contained only very few truly rheophilous chironomid taxa (as described by e.g. Moller Pillot (1990) and Moog (1995)). A number of the chironomid taxa shown in Figure 5.4 are commonly found in streams and rivers, e.g. Orthocladius-type, Cricotopus-type, and Limnophyes (e.g. Wiederholm 1983). However, all of these taxa can also live in shallow lakes, and the remains of these taxa are at most slightly more common in inundated lakes than in isolated lake basins. This suggests that the majority of chironomid head capsules found even in the regularly inundated lakes were produced in the lakes rather than transported there from running water habitats. Gandouin et al. (2006) showed that chironomid-assemblages encountered in lotic and lentic habitats in southern France could be used to qualitatively infer past changes in palaeodischarge and floods. In our study, we did not encounter chironomids restricted to running water habitats, as most of the abundant taxa are known to occur in a range of different habitats. Therefore, no inferences of e.g. the connectivity of the floodplain lakes with the main river channel can be made based on the chironomid assemblages.

5.5.3 Implication of river inundations for chironomid-based temperature inference The chironomid-inferred July air temperatures are all similar to or slightly lower than the observed modern temperatures in the Kaamanen region. The variability in the chironomid-inferred temperatures is surprisingly low, considering the different environmental settings of the lakes incorporated in the dataset. Ground water influx in the inundated lakes probably buffered the water-temperature differences between the different lakes on the floodplain. Most weighted averaging-based transfer functions suffer from the “edge- effect” where inferred temperatures at the edge of the environmental gradient of

92 Chapter 5 interest tend to be biased towards mean values (Birks 1998; Brooks and Birks 2001). Since the temperatures reconstructed for our study lakes are towards the warmer side of the values in our calibration set, it is not unexpected that our reconstructed temperatures are slightly lower than the measured temperature. Other differences between our study sites and the lakes from northwestern Finland on which the applied transfer function is based, like the clearly more acidic lakes in the region of Kaamanen (average pH of 5.1 in our study versus pH = 7.0 in the western Finnish training set (Olander et al. 1999)) might also have induced differences in the reconstructed temperatures. The reconstructed temperature for Lake 33 is distinctly lower than the observed temperature, even when the sample specific errors are taken into account. Lake 33 had a water-temperature that was 4 ºC lower than a lake that was only situated 300 m away (Lake 32; sampled on the same day) and can therefore be considered as an outlier in our data set. The most important conclusion for this project, which can be drawn from Figure 5.5, is that there is little difference in the chironomid-inferred temperatures between the two lake classes. This implies that, according to our results, fossil chironomid records derived from floodplain sediments can be used to reconstruct summer temperatures, and are as reliable as records based on chironomid assemblages from lakes unaffected by riverine influence. This seems to be the case even if the applied chironomid-temperature inference model is developed based on sediments from lakes isolated from riverine influence and annual flooding.

5.6 Conclusions In order to investigate the potential influence of river inundations on fossil chironomid assemblages, sediments and water-samples from 13 regularly inundated lakes and 20 lakes isolated from riverine influence where analysed and compared. The main conclusions that can be drawn from this study are the following: 1) Taxon richness is significantly higher and the chironomid concentration significantly lower in inundated lakes, and there are several taxa that occur preferentially (although not exclusively) in one of the two groups. However, the chironomid fauna shows only comparatively minor differences between the two groups of lakes. 2) Organic matter content of the sediment (as LOI) is the environmental variable that best explains the variation in taxon richness, chironomid concentrations, and in the relative chironomid abundance data between the different studied lakes. 3) Although there are differences in the chironomid fauna of the individual lakes, chironomid-inferred temperatures based on a modern calibration set and transfer function from northwest Finland are relatively constant. All reconstructedtemperatures are similar to or slightly lower than the measured temperature for the Kaamanen region, possibly related to the edge effect in the applied transfer function. However, on average there is no difference in

93 Chapter 5

the temperature inferences for lakes situated on the floodplain versus those isolated from riverine influence. The results suggest that floodplain sediments are a useful alternative for chironomid- based temperature reconstruction in regions and for time-windows for which lacustrine records unaffected by riverine influence are rare.

2 References Armitage PD (1995) The behaviour and ecology of adults. In: Armitage PD, Cranston PS, Pinder LC (eds): The Chironomidae: the biology and ecology of nonbiting midges. Chapman and Hall, London, 194-224 Bigler C, Larocque I, Peglar SM, Birks HJB, Hall RI (2002) Quantitative multiproxy assessment of long-term patterns of Holocene environmental change from a small lake near Abisko, northern Sweden. The Holocene 12: 481-496 Bigler C, Heiri O, Krskova R, Lotter AF, Sturm M (2006) Distribution of diatoms, chironomids and cladocera in surface sediments of thirty mountain lakes in south-eastern Switzerland. Aquatic Sciences 68: 154-171 Birks HJB (1998) Numerical tools in palaeolimnology – progress, potentialities, and problems. J Paleolimnol 20: 307-332 Birks HJB, Line JM (1992) The use of rarefaction analysis for estimating palynologival richness from Quaternary pollen-analytical data. The Holocene 2: 1-10 Birks HJB, Line JM, Juggins S, Stevenson AC, Ter Braak CJF (1990) Diatoms and pH reconstruction. Philos Trans R Soc Lond B 327: 263-278 Bos JAA, Bohncke SJP, Kasse C, Vandenberghe J (2001) Vegetation and Climate during the Weichselian Early Glacial and Pleniglacial in the Niederlausitz, eastern Germany - macrofossil and pollen evidence. J Quat Sci 16: 269-289 Brooks SJ (2006) Fossil midges (Diptera: Chironomidae) as palaeoclimatic indicators for the Eurasian region. Quat Sci Rev 25: 1894-1910 Brooks SJ, Birks HJB (2001) Chironomid-inferred air temperatures from Lateglacial and Holocene sites in north- west Europe: progress and problems. Quat Sci Rev 20: 1723-1741 Brooks SJ, Bennion H, Birks HJB (2001) Tracing lake trophic history with a chironomid-total phosphorus inference model. Freshw Biol 46: 513-533 Brooks SJ, Langdon PG, Heiri O (2007) The identification and use of Palaearctic Chironomidae larvae in palaeoecology. Quaternary Research Association Technical Guide 10, 276 pp Cranston PS (1995) Systematics. In: Armitage PD, Cranston PS, Pinder LC (eds): The Chironomidae: the biology and ecology of nonbiting midges. Chapman and Hall, London: 31-52 Drebs A, Nordlund A, Karlsson P, Helminen J, Rissanen, P (2002) Climatological statistics of Finland 1971-2000. Finnish Meteorological Institute, Helsinki, pp 99 Eggermont H, Heiri O, Verschuren D (2006) Fossil Chironomidae (Insecta:Diptera) as quantitative indicators of past salinity in African Lakes. Quat Sci Rev 25: 1966-1994 Gandouin E, Franquet E, Van Vliet-Lanoë B (2005) Chironomids (Diptera) in river floodplains: their status and potential use for palaeoenvironmental reconstruction purposes. Archiv Hydrobiol 162: 511-534 Gandouin E, Maasri A, van Vliet-Lanoë B, Franquet F (2006) Chironomid (Insecta: Diptera) assemblages from a gradient of lotic and lentic waterbodies in river floodplains of France: a methodological tool for paleoecological applications. J Paleolimnol 35: 149-166 Gandouin E, Ponel P, Franquet E, Van Vliet-Lanoë B, Andrieu-Ponel V, Keen DH, Brulhet J, Brocandel M (2007) Chironomid responses (Insect: Diptera) to Younger Dryas and Holocene environmental changes in a river floodplain from northern France (St-Momelin, St-Omer basin). The Holocene 17: 1-18 Grimm EC (1991–2004) TILIA, TILA.GRAPH, and TGView. Illinois State Museum, Research and Collections Center, Springfield, USA http://demeter.museum.state.il.us/pub /grimm/ Heegaard E, Lotter AF, Birks HJB (2006) Aquatic biota and the detection of climate change: are there consistent aquatic ecotones? J Paleolimnol 35: 507-518 Heinrichs ML, Walker IR (2006) Fossil midges and palaeosalinity: potential as indicators of hydrological balance and sea-level change. Quat Sci Rev 25: 1948-1965 Heiri O, Ekrem T, Willassen E (2004) Larval head capsules of European Micropsectra, ParaTanytarsus and Tanytarsus (Diptera: Chironomidae: Tanytarsini). Version 1.0. http://www.bio.uu.nl/~palaeo/ Chironomids/Tanytarsini/intro.htm Heiri O, Lotter AF (2005) Holocene and Lateglacial summer temperature reconstruction in the Swiss Alps based on fossil assemblages of aquatic organisms: a review. Boreas 34: 506-516 Hill MO, Gauch HG (1980) Detrended correspondence analysis: an improved ordination technique. Vegetation 42: 47-58

94 Chapter 5

Juggins S (2003) C2 User guide. Software for ecological and palaeoecological data analysis and visualisation. University of Newcastle, Newcastle upon Tyne, UK Kasse C, Vandenberghe J, Van Huissteden J, Bohncke SJP, Bos JAA (2003) Sensitivity of Weichselian fluvial systems to climate change (Nochten mine, eastern Germany). Quat Sci Rev 22: 2141-2156 Korhola A, Olander H, Blom T (2000) Cladoceran and chironomid assemblages as quantitative indicators of water depth in subarctic Fennoscandian lakes. J Paleolimnol 24: 43-54 Korhola A, Vasko K, Toivonen HTT, Olander H (2002) Holocene temperature changes in northern Fennoscandia reconstructed from Chironomids using Bayesian modelling. Quat Sci Rev 21: 1841-1860 Larocque I, Hall RI, Grahn E (2001) Chironomids as indicators of climate change: a 100-lake training set from a subarctic region of northern Sweden (Lapland). J Paleolimnol 26: 307-322 Lepš J, Šmilauer P (2003) Multivariate Analysis of Ecological Data using CANOCO. University Press, Cambridge Lindegaard C (1997) Diptera Chironomidae, non-biting midges. In: Nilsson A Aquatic insects of North Europe vol 2 Odonata – Diptera. Apollo books, Stenstrup, 440 pp Lotter AF, Birks HJB, Hofmann W, Marchetto A (1998) Modern diatom, cladodera, chironomid, and chrysophyte cyst assemblages as quantitative indicators for the reconstruction of past environmental conditions in the Alps. II. Nutrients. J Paleolimnol 19: 443-463 Mol J (1997) Fluvial response to Weichselian climate change in the Niederlausitz (Germany). J Quat Sci 12: 43-60 Moller Pillot HKM (1990) De larven der nederlandse Chironomidae (Diptera) Deel C: Autoekologie en verspreiding. Nederlandse Faunistische Mededelingen 1C, 87 pp Moog O (1995) Fauna Aquatica austriaca. Abteilung für Hydrobiologie, Fischereiwirtschaft and Aquakultur der Universität für Bodenkultur, Wien Nyman M (2007) Distribution of non-biting midges (Diptera, Chironomidae) in subarctic lakes of Finnish Lapland – applications in lake classification and palaeolimnology. Kilpisjärvi Notes 20 Nyman M, Korhola A, Brooks SJ (2005) The distribution and diversity of Chironomidae (Insecta: Diptera) in western Finnish Lapland, with special emphasis on shallow lakes. Global Ecol Biogeogr 14: 137-153 Olander H, Birks HJB, Korhola A, Blom T (1999) An expanded calibration model for inferring lakewater and air temperatures from fossil chironomid assemblages in northern Fennoscandia. The Holocene 9: 279-294 Pinder LCV (1986) Biology of freshwater Chironomidae. Ann Rev Entomol 31: 1-23 Renberg I (1991) The HON-Kajak sediment corer. J Paleolimnol 6: 161-170 Sæther, OA (1976) Revision of Hydrobaenus Trissocladius Zalutschia Paratrissocladius and some related genera (Diptera: Chironomidae). J Fish Res Board Can 195: 1-287 Schmid PE (1993) A key to the larval Chironomidae and their instars from Austrian danube region streams and rivers with particular reference to a numerical taxonomic approach. Part I. Diamesinae, Prodiamesinae and Orthocladiinae. Wasser und Abwasser Supplement 3/93, 514 pp Ter Braak CJF, Looman CWN (1994) Biplots in reduced-rank regression. Biometr J 36: 983-1003 Thomaz SM, Bini LM, Bozelli RL (2007) Floods increase similarity among aquatic habitats in river-floodplain systems. Hydrobiologia 579: 1-13 Vasko K, Toivonen HTT, Korhola A (2000) A Bayesian multinomial Gaussian response model for organism- based environmental reconstruction. J Paleolimnol 24: 243–250 Velle G, Brooks SJ, Birks HJB, Willassen E (2005) Chironomids as a tool for inferring Holocene climate: an assessment based on six sites in southern Scandinavia. Quat Sci Rev 24: 1429-1462 Walker IR, Cwynar LC (2006) Midges and palaeotemperature reconstruction – the North American experience. Quat Sci Rev 25: 1911-1925 Walker IR, Smol JP, Engstrom DR, Birks HJB (1991) An assessment of Chironomidae as Quantitative Indicators of Past Climatic Change. Can J Fish Aquat Sci 48: 975-987 Wiederholm T (1983) Chironomidae of the Holarctic region. Keys and diagnoses. Part I. Larvae. Entomol Scand 19

95 Chapter 5

2

96 Chapter 6

Chapter 6

Environmental inferences and chironomid-based tem- perature reconstructions from fragmentary records of the Weichselian Early Glacial and Pleniglacial periods in the Niederlausitz area (eastern Germany)

S Engels, SJP Bohncke, JAA Bos, O Heiri, J Vandenberghe, J Wallinga

Abstract We inferred past climate conditions from lacustrine sediments intercalated in Weichselian Early Glacial and Early Pleniglacial fluvial and aeolian sediments, exposed in two opencast lignite mines from the Niederlausitz area (eastern Germany). A chronology was established using radiocarbon and luminescence dating methods. Both the lithology and the chironomid faunas indicate that the former shallow lakes were situated on a floodplain. Palaeotemperature estimates calculated from the fossil chironomid-assemblages of the Early Glacial lacustrine deposit indicate mean July air temperatures of 15 ºC, which is in line with results derived in earlier studies from the Niederlausitz area and from northwestern Europe. The Early Pleniglacial lacustrine deposits consist of an organic-rich gyttja, intercalated with sand and silt lenses. The chironomid assemblages show that a shallow meso- to eutrophic lake was present at the study site, and chironomid-inferred palaeotemperature estimates show an abrupt decline in July air temperatures from 15-16 ºC to 13 ºC. In combination with other proxies from the same record, this suggests a Dansgaard/ Oeschger like climate evolution.

This manuscript is accepted for publication in Palaeogeography, Palaeoclimatology, Palaeoecology.

97 Chapter 6

6.1 Introduction The Weichselian period has been a research topic of interest for several decades as it is a period of dynamic climate evolution that was not substantially influenced by human activity. The number of lacustrine records covering (parts of) the Weichselian on the European continent is limited, and high-resolution quantitative reconstructions of past climate change are scarce. With the development of the so- called transfer-function approach (e.g. Birks 1998), new tools have become available to quantitatively infer past changes in climate from fossil assemblages of different groups of organisms such as diatoms, pollen or chironomids. Chironomids are sensitive indicators of past changes in water depth, nutrient availability and summer temperatures (e.g. Walker and Cwynar 2006; Brooks 2006) and well-preserved head capsules of the larvae are usually abundant in lake sediments. In recent years the development of chironomid-based transfer functions has greatly improved the usefulness of chironomids for palaeoenvironmental reconstruction. Using transfer functions, we are able to provide quantitative estimates of past environmental conditions based on fossil chironomid assemblages (e.g. Walker et al. 1997; Brooks and Birks 2001; Heiri and Lotter 2005). Initial applications of these transfer-functions have focussed on the Late-Glacial period (e.g. Walker et al. 1991; Brooks and Birks 2000) and the Holocene (e.g. Heiri et al. 2003; Velle et al. 2005) and were aimed at reconstructing past July air temperatures. Chironomid-based studies on lacustrine sediments from Europe predating the Late- Glacial are scarce, and include Becker et al. (2006), Gandouin et al. (2007) and Engels et al. (2007a). The latter study provides (to our knowledge) the first chironomid- inferred palaeotemperature estimates for the Middle Weichselian.

The opencast lignite mines of eastern Germany provide large exposures of Weichselian sediments that are mostly of fluvial origin, and have been the subject of extensive studies in the past (e.g. Wolf and Alexowsky 1994; Mol 1997a; Bos et al. 2001; Hiller et al. 2004; Kasse et al. 2003; Bohncke et al. in press). Lacustrine deposits, 6 intercalated in these fluvial and aeolian sediments, are suitable archives for palaeoenvironmental reconstructions based on the multitude of proxies available to infer past changes in vegetation, environment and climate from lake sediments (e.g. Smol et al. 2001a; 2001b). Lacustrine sediments encountered in the opencast mines of eastern Germany all originate from former lakes situated on river floodplains. Hence, they potentially may have been affected by flooding by the nearby river. Modern training sets normally do not include lakes that are prone to flooding by rivers, which creates a non-analogue situation between our fossil floodplain lakes and the existing modern training sets. To our knowledge, no quantitative palaeoenvironmental reconstructions on chironomid remains from floodplain lakes has been attempted so far. In a recent study of 33 lakes in Finland, Engels et al. (2007b) have shown that the variability of chironomid-based reconstructions and prediction residuals from

98 Chapter 6

Welzow

Spremberg

Bad Muskau Weißwasser SR-X1

Hoyerswerda Nochten LM8 Bärwalde Boxberg

5 km Reichwalde

Older depositsMine Cities/ Towns River/ stream Box core

Figure 6.1: Location map of the Nochten and Reichwalde mines in the Niederlausitz area (after Mol, 1997a). The locations of the SR-X1 and LM8 sample boxes are indicated by arrows. floodplain lakes is similar to lakes unaffected by riverine influence. This suggests that floodplain lakes are suitable alternatives for reconstructing summer temperatures in situations where no lakes isolated from riverine influence are available, as is the case in the Niederlausitz area of eastern Germany. In this paper we show the results of a detailed analysis of the sedimentary succession of deposits, exposed in two opencast lignite mines in the Niederlaustiz area: Nochten and Reichwalde (Figure 6.1). Two lacustrine deposits, dated to the Early Glacial and the Early Pleniglacial, are analysed for a range of proxies and in this paper we aim to reconstruct mean July air temperatures by applying chironomids as a proxy for these two time-windows for which quantitative data on past climate conditions is scarce.

6.2 Geological setting The Lausitzer ice-marginal valley, situated in the southern part of eastern Germany, was formed during the Saalian glaciation and has an E-W orientation. The former valley is characterised by a number of large opencast lignite mines, including the Reichwalde and Nochten mines (Figure 6.1). Active recovery of Miocene browncoal, situated approximately 100 m below the current surface, exposes the overlying sediments that are mainly of Saalian (130-200 ka) and Weichselian (12-110 ka) age. Previous studies in the Nochten mine include studies of the fluvial succession (Mol 1997a; 1997b; Kasse et al. 2003) and of the vegetation and climate development (Bos et al. 2001). The Reichwalde mine was previously visited by Bohncke et al. (in press) who qualitatively reconstructed environmental and climatic conditions from a short lacustrine record and provided a first chironomid record for the sequence.

99 Chapter 6

6.3 Methodology 6.3.1 Data acquisition In June 2004 we studied a 1000 m long N - S oriented exposure in the Nochten mine. As scree was present on the slopes, no continuous sections could be analysed but instead, many vertical logs were investigated. Special attention was given to lithology, sedimentary structures and periglacial phenomena. Samples for optically stimulated luminescence (OSL) dating were taken from the aeolian and fluvial deposits, while the organic material was collected using sample boxes, and transported to the Vrije Universiteit Amsterdam (The Netherlands) for detailed botanical (pollen, macro-remains) and zoological (chironomids) analysis. To obtain a better understanding of the chronology, additional macro-remains were taken from the sample boxes for radiocarbon dating. During fieldwork in 1999, a large N - S oriented exposure in the Reichwalde mine was analysed and the aeolian and fluvial deposits were studied in a fashion similar to the Nochten sediments.

6.3.2 Chronology Using both radiocarbon and OSL dating techniques, a chronology was established for the Weichselian sediments of the Nochten mine (Table 6.1; Figure 6.2). We applied quartz OSL dating to determine the time of deposition of the fluvial sediments (Wallinga 2002). The OSL signal of sand-sized quartz grains is set to zero by light exposure prior to burial, and builds up after burial due to ionizing radiation from surrounding sediments (with a small contribution from cosmic rays). The age of a sample is obtained from measurements of the dose absorbed by the grains since the

last light exposure (equivalent dose; De, expressed in Gy) and measurement of the radiation flux the grains were exposed to since burial (dose rate; expressed in Gy/ka). Quartz OSL dating is usually applicable to sediments up to 100 – 150 ka of age. For equivalent dose determination we used the single-aliquot regenerative dose (SAR) procedure (Murray and Wintle 2003). Dose recovery tests (Roberts et al. 1999; 6 Wallinga et al. 2000) indicated that a laboratory dose could be accurately measured

OSL Sample Unit Material Equivalent dose Dose rate OSL Age Gy s.e. Gy/ka s.e. ka s.e. NCL-6605051 N4 Fluvio-aeolian sand 25.5 1.4 1.19 0.05 21.4 1.5 NCL-6605052 N4 Fluvio-aeolian sand 31.0 1.5 1.32 0.05 23.6 1.4 NCL-6605053 N2 Shallow fluvial sand 88.8 4.3 1.10 0.05 81 5 NCL-6605054 N2 Shallow fluvial sand 81.0 3.4 1.26 0.05 64 4 NCL-6605055 N1 Shallow fluvial sand 69.1 3.1 0.93 0.04 74 5 NCL-6605056 N1 Shallow fluvial sand 107.1 5.3 1.25 0.05 86 5 NCL-6605057 N1 Shallow fluvial sand 46.3 2.2 0.57 0.04 82 6

14C Sample Unit Material 14C age ka BP s.e. GrA-30631 N1 Carex fruits, Salix twig fragments >45 GrA-30632 N1 Carex fruits, Salix twig fragments 43.0 ±0.9/0.7 GrA-22168 RW 3 Potamogeton fruits >47 GrA-22169 RW 3 Potamogeton fruits 45.8 ±3.8/2.6

Table 6.1: OSL ages and uncalibrated radiocarbon dates from the Weichselian sedimentary records of Nochten and Reichwalde

100 Chapter 6 using this procedure (average dose recovery ratio of 1.00 ± 0.01; in excellent agreement with unity). For dose rate estimation we employed gamma-ray spectrometry.

6.3.3 Chironomid analysis The organic lacustrine deposits of SR-X1 (Nochten mine, Figure 6.2) and LM8 (Reichwalde mine, Figure 6.2) were sampled every other centimeter, and were analysed for chironomid remains and a range of palaeobotanical proxies. For SR-X1,

Nochten Depositional Reichwalde m Sedimentary environment m Sedimentary Chronology Chronology Units asl structures Units rel. structures 134 Aeolian dunes RW6 N5 Soil Deflation surface 21 131 21.4 Fluvial flood deposits, RW5 aeolian reworking 23.6 N4

14 Fluvial deposits, RW4 125 braided river system

12

N3 Fluvial flood RW3 122 deposits 81.0 *LM8* >47 45.8

64.5 N2

4 74.0 116 RW2 85.8 Overbank deposits >45 N1 2 81.6 *SR-X1* 114 43.0 RW1 0 Fluvio-glacial deposits Lithology Sedimentary structures Chronology Organic/ loam No visible structures Involution OSL-date (ka) Sand Low angle cross-bedding Frost cracks 14C-date (uncal ka BP) Sand & gravel Horizontal bedding Box cores Trough and large-scale Ice-wedge cast *code* planar cross-bedding

Figure 6.2: Composite sedimentological logs of the Nochten and Reichwalde mines, with cryogenic features, chronology and sedimentary environments. The hatched lines indicate a possible correlation between the different sedimentary units.

101 Chapter 6

a total of 10 wet sediment samples with a weight range of 3.8 – 11.1 g were used for chironomid analysis. The LM8 sequence includes 18 sediment samples with a weight range of 2.0 – 16.7 g. The samples were treated with cold KOH for at least 4 hours and rinsed on a 100 µm sieve before hand-sorting the chironomid head capsules (hcs) under a dissecting microscope (35x magnification). The chironomid hcs were identified following Wiederholm (1983), Moller Pillot (1984), Schmid (1993), Rieradevall and Brooks (2001) and Heiri et al. (2004). A chironomid percentage diagram was constructed using the computer programs TILIA and TG.VIEW (Grimm 1991-2004). Of the available chironomid-temperature transfer functions and training sets, the Swiss training set was selected as it is closest to our site and it includes the majority of the taxa encountered in our fossil sediments. Although our sampling site is not situated within the region where the training set was developed, Heiri et al. (2007) and Ilyashuk et al. (2005) showed that transfer functions can be applied outside their region of origin provided that the results are evaluated carefully. The Swiss chironomid-July air temperature transfer function is based on a 2- component weighted-averaging partial least squares regression (WA-PLS) calculated on square root transformed percentage abundances of chironomid taxa identified in surface sediment assemblages from 115 lakes in the Alpine region (Heiri and Lotter 2005; Bigler at al. 2006). The model was screened for outliers and a total of 14 lakes were deleted from the model based on ecological criteria. These outliers included lakes affected by unusual hydrological conditions, exceptionally deep lakes, lakes with a strong running-water influence on chironomid assemblages and lakes strongly affected by local topography (see Von Gunten et al. (2007) for details). The screened inference model covered a July air temperature range of 5-18.4 ºC, had a leave-one-out cross-validated Root Mean Square Error of Prediction (RMSEP) of 1.4 ºC, a coefficient of determination (r2) of 0.88 and a maximum bias of 1.28 ºC. Sample specific prediction errors were calculated using 999 bootstrap cycles. Following Birks et al. (1990) and Heiri et al. (2003), the chi-square distance 6 between the individual fossil samples and the most similar sample in the modern calibration dataset was determined, and distances larger than the 2nd and 5th percentile of all chi-squared distances in the modern data were identified as fossil samples with “no close” and “no good” analogues, respectively. A canonical correspondence analysis (CCA) with mean July air temperature as the sole constraining variable was performed, where the fossil samples were added as passive samples. Cut-off levels of the 95th and 90th percentiles of all the residual distances to the first canonical axis in the modern training set was used as an estimate of the fit of the fossil samples to temperature (“very poor” and “poor” respectively; Birks et al. 1990). The percentage of rare taxa was calculated for each fossil sample, where a rare taxon was defined as having a Hill’s N2 (Hill 1973) below 5 in the calibration dataset (e.g. Heiri et al. 2003). WA-PLS, Hill’s N2 and chi-square distances were calculated using computer program C2 version 1.4.2 (Juggins 2003), CCA was performed using CANOCO version 4.51 (ter Braak and Šmilauer 2002).

102 Chapter 6

6.3.4 Botanical analysis Plant macrofossils were hand-sorted from large sediment samples that were derived from the same depth intervals as the chironomid-samples. Samples with varying weight were treated with cold KOH (5%) for at least 4 hours and washed over a 100 ìm sieve, after which macrofossils were picked from the sieve residue. Palaeotemperature estimates were made based on the plant taxa by using the climate indicator plant species method (sensu Iversen 1954; Kolstrup 1980). The relationship between the modern geographical limit of plant distribution and temperature can be used to reconstruct past minimum temperatures based on fossil records of plant remains, as the individual plants require certain minimum mean summer temperatures to flower and reproduce.

6.4 Lithology and geochronology The independent OSL-based chronology derived from the Nochten sedimentary record suggests a rather constant accumulation of sediments during the Weichselian, and is compared to existing chronological sequences from the Nochten (Bos et al. 2001; Kasse et al. 2003) and Scheibe mine (Mol 1997a; 1997b). The Reichwalde mine does not have an independent chronology (except for the two radiocarbon dates that are at the range limit of the radiocarbon dating technique). The general sedimentological sequence is, however, similar to that of the Nochten mine, although the Weichselian sediment column is 2 m thicker. Combining the assumed simultaneous occurrence of cryogenic features, deflation surfaces and the formation of the Usselo soil, the sedimentological sequences were correlated as indicated in Figure 6.2 (hatched lines). Due to the lack of independent determination of the elevation (in m asl) for the Reichwalde record, the relative positioning of that column has an uncertainty of a few meters.

Several studies have shown that sediments from the previous interglacial (the Eemian) are absent from the Niederlausitz area (e.g. Wolf and Alexowsky 1994; Mol 1997a). Kasse et al. (2003) identified a complex of organogenic deposits (their Unit 1a) intercalated by fluvial sands as Early Glacial sediments in an E-W exposure in the Nochten mine. These sediments lay directly on top of the Saalian glaciofluvial sediments. In this study, we dated Unit N1 to approximately 80 ka (3 OSL dates, Figure 6.2), which indicates an age near the end of the Early Weichselian Glacial. That these Early Glacial deposits are probably formed over a large area in the Niederlaustiz region, is further demonstrated by Mol (1997a) and Hiller et al. (2004), who also recognised a complex of organic-rich Early Glacial deposits discordantly overlying Saalian fluvio-glacial sediments in the Scheibe mine, and by the complex of organic sediments observed in the Reichwalde mine that are correlated to the Early Weichselian Glacial as well (Units RW1 and RW2).

The 6 m thick clastic deposit of Unit N4, characterised by parallel laminations and frost cracks, overlies the Early Glacial sediments. The OSL dates estimate an age of 60-

103 Chapter 6

70 ka for this sediment interval, roughly equivalent to the Early Pleniglacial and thus indicate that there is no (large) hiatus between the sediments of Units N1 and N2. Bos et al. (2001) and Kasse et al. (2003) document a similar 4 m thick sediment body consisting of sheet-flood or crevasse splay deposits. They report three luminescence dates in his interval, indicating an age of 40-50 ka. In addition, they show three radiocarbon dates that are near the limit of the method (~ 40 14C ka BP). Based on this age control Kasse et al. (2003) reconstruct a major hiatus spanning at least the Early Pleniglacial and possibly also the earlier parts of the Middle Pleniglacial. This interpretation is in sharp contrast to our OSL chronology. The luminescence ages reported by Bos et al. (2001) and Kasse et al. (2003) were obtained through a multiple aliquot infrared stimulated luminescence (IRSL) procedure on fine-grain feldspar (M. Frechen, personal communication 2007). The feldspar IRSL signal has been shown to be unstable on geological timescales (e.g. Huntley and Lian, 2006), and will tend to underestimate the burial age unless the ages are corrected for anomalous fading. We note that an anomalous fading rate of approximately 6% per decade would explain the discrepancy between quartz OSL and feldspar IRSL ages.Such fading rates are in line with those reported by Huntley and Lian (2006). However, we cannot be certain that anomalous fading resulted in age underestimation for the Bos et al. (2001) and Kasse et al. (2003) samples, since fading measurements were not carried out. We also need to consider the possibility that the quartz OSL signal was not completely reset in all grains prior to deposition. This would result in a remnant dose, causing overestimation of the true burial dose and thus of the age (e.g. Wallinga 2002). However, the equivalent dose distributions do not indicate the occurrence of heterogeneous bleaching. Hence we regard it unlikely that our OSL ages grossly overestimate the burial age. In the light of the uncertainties with regard to the previously reported ages (both 14C and IRSL) and convincing demonstrations of the accuracy of quartz OSL ages obtained using the SAR procedure (Murray and Olley 2001; Wallinga 2002), we 6 conclude that our new ages are more robust than the previous age estimates and that they provide a more reliable chronology for the site. It is interesting to note that multiple-aliquot OSL dates on quartz grains from a fluvial flood deposit in the Scheibe mine (also located some 10 meters below the surface) by Mol (1997a) yielded ages of approximately 70 ka, in accordance with the dates we derived for Unit N2. The uncertainties on the dates by Mol (1997a) were large and the authors interpreted their OSL- ages as overestimations of the true age of the sediments based on a priori expectations of the age of the sediments based on palynological results from underlying sediments. Our results suggest however that the OSL ages of Mol (1997a) may have been correct.

The coarse-grained sediments of Unit N3 were not dated as insufficient bleaching of the material was expected from fast sedimentation in the high-energetic braided river system. The sediments overlying the coarse-grained sediments of Unit N3 are well- dated using OSL-techniques to ~22 ka (Late Pleniglacial) and therefore the sediments

104 Chapter 6 of Unit N3 might have been deposited in the Late Pleniglacial, but possibly also much earlier (Middle to Early Pleniglacial). Kasse et al. (2003) dated the same deposits in the Nochten mine through radiocarbon dating to ~36 14C ka BP, and there the deposit is wedged between two sedimentary units that are IRSL dated to ~20 ka and ~45 ka. Caution is however needed with interpretation of this chronology, as radiocarbon ages in this range may underestimate the true age (e.g. Briant et al. 2005) and the IRSL ages may be affected by anomalous fading.

Unit N4 is dated to the Late Pleniglacial (2 OSL-dates: 20-23 ka), which is in good accordance with 2 radiocarbon and 3 IRSL dates from Kasse et al. (2003) from the Nochten mine, and also with age estimates of 18-20 ka (OSL-dates) of a large fluvio- aeolian deposit encountered in the Scheibe mine by Mol (1997a;1997b). The large syngenetic ice-wedge casts that were encountered in the Nochten mine could therefore have been formed during the Last Glacial Maximum (LGM), and are also present in Unit RW5, which is correlated to Unit N4. The sediments from Units N4 and RW5 probably belong to a sedimentary facies that is widespread in the region, which was expected due to its (partly) aeolian origin.

The gravel string encountered in the upper part of unit N2 is likely to be the eastern German equivalent of the Beuningen Gravel Bed from the Netherlands (Van der Hammen et al. 1967; Zagwijn and Paepe 1968). The soil horizon that is found 50 cm above the gravel string would than be the equivalent of the German Finow Soil or the Dutch Usselo Soil, which are both dated to the Allerød period (Van der Hammen et al. 1967; Van Geel et al. 1989; Kaiser et al. 2006). The aeolian deposits of unit N5 therefore were probably formed during the Younger Dryas interstadial.

6.4.1 Lithology of sample box SR-X1 Sample box SR-X1 (114.5 m asl) makes up the lower part of sedimentary unit N1 and is OSL dated to approximately 80 ka. The lowermost 14 cm of the core consist of a silty gyttja. A sandy layer (~ 4 cm thickness), completely barren of organic matter, is present on top of the silty gyttja, after which a second phase of deposition of organic- rich sediments occurred. As the lower part of Unit N1 was not heavily cryoturbated at the location of SR-X1, continuous horizontal laminae consisting of sand- and silt lenses were still visible in the organic-rich deposit. They suggest a regular input of clastic material from a nearby river system. The upwards decreasing organic content of the core (Figure 6.3) suggests that the influence of the river increased in time. The exposure showed that the gyttja deposits of SR-X1 extend laterally for several tens of meters, after which they wedge out and are replaced by soils and peat deposits.

6.4.2 Lithology of sample box LM8 The total thickness of the whole fluvial cycle from which LM8 is retrieved is ca. 2 m (unit RW3), and consists of low-energetic anastomosing-river deposits. Below sample box LM8, sandy and silty floodplain sediments are present that have been interpreted

105 Chapter 6 -type) 16 Chironomid-inferred Chironomid-inferred Limnophyes, Cricotopus July Air Temperature Temperature July(°C) Air 12 14

Lotic (hc/g)

40%

nic content (%) content nic a

20

Org

Paraphaenocladius 20

s/ ype

10 Chironomid concentration concentration Chironomid

heotanytarsus

R etriocnemu

m

ype

ra

oneura lobata-t oneura

a 20%

P yn

Lentic pus-t

nschoeldi-type

to Cor

rico 20%

C

us rose us

ype s t

s- nom elo

b

20% ri

racinus-type T

20% tictochiro S

20%

type

Tanytarsus lugen Tanytarsus

s-

Chironomus anth Chironomus cladius

Pro

e tendipes pedellus-type tendipes

pes-type

icro

20 40%

M

ctrocladius sordidellu ctrocladius

flavi

m nubeculosum-type m

osus-typ

se

P m

icrotendipes nervosus-type icrotendipes

(semi-) terrestrial (semi-) D

us plu us

enopsectra Polypedilu

nom Pha

sand hiro

C

6 spp. s

osmittia Stempellinella-type ud

20%

Pse

Metriocnemu phyes

20 40% imno

L rthocladius rthocladius

20 40% Geo

organic-rich silt

Lithology

nt sum nt

ou C 7 87 71 53 81 66 73 60 76 165 103

18 20 22 24 26 28 30 32 34 36 38 40 Depth (cm) Depth : Chironomid abundance diagram of selected taxa from the SR-X1 sample box, and organic content sediment (estimated through loss- Figure 6.3 on-ignition). Chironomid taxa are classified based on the habitats they predominantly found in, although some (e.g. can occur in a range of habitats types. Chironomid-inferred mean July air temperature estimates with sample specific error bars as determined by bootstrap cross-validation are plotted to the right.

106 Chapter 6 as a distal floodplain facies of the river that was active at that time. The diminishing grain size and the eventual deposition of a light grey loam (core depth 50-37 cm) show the decreasing influence of the river. Subsequently, cryoturbation structures developed in a water-saturated environment (Bohncke et al. in press). Water is finally released as a result of progressive melting of the segregation ice at the top of the permafrost. Following the formation of the lake, horizontally bedded red-brown gyttja was deposited (38-16.5 cm core depth). This deposit has the highest organic content, increasing upward from ~10% to ~30% (Figure 6.4). At 16 cm core depth a sharp decline in organic matter is initiated and organic matter values drop to ~5 %. Between 16.5 and 6 cm core depth, a layer of dark grey-brown clay was deposited, and towards the top of the sequence the number of silty and sandy layers increases. These inputs of coarse clastic sediment are interpreted as moments of fluvial input to the basin, possibly through flooding of the lake. There are no clayey or silty deposits above 6 cm core depth; the fluvial sands suggest that the basin was completely filled during a dramatic flood event.

6.5 Chironomid records and their palaeoclimatic interpretations 6.5.1 Chironomid record of SR-X1 (Weichselian Early Glacial) The concentration of chironomid hcs per gram of wet sediment is generally low in the SR-X1 record, and follows the trend of the percentage of organic matter content of the sediments (Figure 6.3). Count sums are between 53 and 165 hcs up to 21 cm core depth. Near the top of SR-X1, there is a sharp decrease in chironomid concentrations, and the uppermost sample only yielded 7 hcs. A total of 47 taxa were identified in the SR-X1 record, of which a selection is plotted in Figure 6.3. Chironomid taxa are classified according to ecological information presented by Moog (1995). The encountered morphotypes of Metriocnemus spp. and Cricotopus-type have been amalgamated in this Figure to enhance readability. The majority of taxa in the SR-X1 record indicate that the lake was probably very shallow. Although most taxa are typically found in standing water habitats, several of the most abundant taxa, with occurrences over 10% of the assemblage throughout the record (e.g. Georthocladius), are often associated with terrestrial or semi-terrestrial habitats such as wet soils. Limnophyes is usually associated with very shallow water in the littoral of lakes and with streams, but can also occur in (semi-) terrestrial habitats (Brooks et al. 2007). Other taxa (plotted to the right of Figure 6.3) are known to occur in flowing water habitats and are uncommon in lake sediments (Rheotanytarsus) or can be found in both flowing and standing water habitats (e.g. Cricotopus-type). Both the chironomid taxa associated with flowing water habitats, as well as those associated with (semi-) terrestrial habitats, could have been transported to the lake through frequent floodings of a nearby river. This implies that the former lake was most likely situated in a floodplain. Although there are some shifts, especially in the abundances of Microtendipes pedellus-type, no clear trend can be identified in the chironomid record of core SR-X1. The genus Microtendipes is reported to be an indicator of intermediate temperatures in northern

107 Chapter 6 14 16 18 Lotic Chironomid-inferred

12

July Air Temperature (°C) July Temperature Air

nt (%) nt (hc/g) n

20%

anic conte anic g

10 Or

mid concentratio mid

o

Chiron 30 60

Eukiefferiella-type

cotopus-type dius-type

Cri

Orthocla

s-type

Lentic

us i

oclad

r

anthracinu

P Silt

type

- us us

osus-type

Chironom

ua

pe

y omus plum omus

ambig Tanytarsus lugens Tanytarsus

-type

Chiron Sand

vosus-type

Corynocera Corynocera ype

Sergentia coracina-t Sergentia

icrotendipes ner icrotendipes

D

Polypedilum nubeculosum Polypedilum

6 (semi-)

terrestrial ochironomus

Silty clay Cladopelma lateralis-t Cladopelma

Crypt

ndipes pedellus-type ndipes

ote

Micr -type) can occur in a range of habitats types. Chironomid-inferred mean July air temperature estimates with sample

nophyes Gyttja

Lim 20 20 40 60 20 20 20 20 20 40 20 40

riocnemus spp. riocnemus

Met

Pseudosmittia

Lithology t sum t

oun C 14 61 57 56 58 97 89 90 85 66 76 68 70 70 76 81 161 139

10 15 20 25 30 35 : Chironomid abundance diagram of selected taxa from the LM8 sample box (after Bohncke et al., in press), and organic content Limnophyes, Cricotopus Depth (cm) Depth Figure 6.4 sediment (estimated through loss-on-ignition). Chironomid taxa are classified based on the habitats they predominantly found in, although some taxa (e.g. specific error bars as determined by bootstrap cross-validation are plotted to the right.

108 Chapter 6

Europe (Brooks and Birks, 2001), and it is most common in the littoral and sublittoral zones of lakes with coarse sediments (Brodersen and Lindegaard, 1999; Brooks et al., 2001).

6.5.2 Chironomid record of LM8 (Early Pleniglacial) A total of 54 chironomid taxa are identified in the 18 samples from LM8, and selected taxa are plotted in Figure 6.4 (after Bohncke et al. in press). The count sum of the individual samples is 56-161 hcs, with the exception of the uppermost sample that only yielded 14 hcs. The chironomid-assemblages shown in Figure 6.4 suggest that a shallow lake existed at the study site during the period of infilling and the lake was probably meso- to eutrophic, as indicated by several taxa such as Ablabesmyia or Chironomus plumosus-type (e.g. Brodersen and Quinlan, 2006). Most of the abundant fossil taxa currently occur in low altitude or temperate lakes in NW and Central Europe. In the lower part of the record Chironomus anthracinus-type shows abundances over 15%. After this initial peak, the values of C. anthracinus-type decline in favor of M. pedellus-type, which has a high abundance throughout the remainder of the record (Figure 4). Dicrotendipes and Glyptotendipes, present in the lower part of LM8, are often associated with the presence of macrophytes (Moller Pillot 1984) and are most common in meso- to eutrophic lakes (Brodin 1986). At 12.5 cm core depth M. pedellus-type, Cladopelma lateralis-type and Polypedilum nubeculosum-type values decrease in favor of Tanytarsus lugens-type, Procladius, and C. anthracinus-type. The chironomid concentration decreases from 50 to 10 hc/g. T. lugens-type is one of the taxa that dominates in the upper part of the record, and is often considered as an indicator of cool, oligotrophic conditions (e.g. Brodin 1986; Porinchu and MacDonald 2003). The high values of C. anthracinus•-type, Procladius and T. lugens-type could indicate an increase in lake-depth, although this is unlikely considering the geomorphological situation of the lake. These three taxa are also indicative for lower temperatures than the taxa that were more abundant in the lower part of the record.

6.5.3 Quantification of July air temperatures The chironomid-inferred July air temperature record of SR-X1 shows stable reconstructed temperatures around 14.5 ºC (Figure 6.3). All sample specific prediction errors are between 1.5 and 1.6 ºC. The sample just below the clastic interval that is barren of chironomids (26.5 cm core depth) shows a lower reconstructed temperature of 12.9 ºC. Above the clastic interval, the reconstructed temperature again is 14.8 ºC. The uppermost sample is not plotted in this Figure, as its count sum is much lower than 50, the minimum count sum recommended for numerical analysis (e.g. Heiri and Lotter 2001; Quinlan and Smol 2001). Although most chironomid taxa identified in the SR-X1 record are well represented in the Alpine calibration dataset, the absence of Georthocladius in the modern data set results in high cumulative abundances of taxa included in the fossil data but absent in the modern training set (average: 14,0 %;

109 Chapter 6

range: 4,6-35,0 %). When Georthocladius is excluded from the analyses, the sum of rare taxa (Hill’s N2 < 5) is 0,0-0,4 % for all fossil samples from SR-X1. Furthermore, even when excluding Georthocladius, the nearest-modern-analogue (as chi-squared distance) and goodness-of-fit (as squared residual distance) calculations resulted in ‘no good’ analogues for all fossil samples (except for sample 26.5 cm, which has ‘no close’ analogues) and in ‘poor fit’ to ‘very poor fit’ scores for all fossil samples. This result is unexpected, as all the dominant taxa (except for Georthocladius) are well represented in the Swiss calibration dataset. Probably, the combination of taxa encountered in the fossil assemblages of SR-X1 is not found in the calibration data set, perhaps as a result of the influence of river inundations on the fossil lake ecosystem of SR-X1. Previous studies by Heiri et al. (2003) and Engels et al. (2007a) have shown that chironomid- assemblages from early Holocene or even older deposits can have poor analogue statistics, even though the abundant taxa in their records were well represented in the modern calibration datasets. The reconstructed mean July air temperatures for the LM8 record sequence (Figure 6.4) are relatively high (15 to 16 ºC) for the lowermost part of the record, but steadily decline above 17 cm core depth to values around 13.0 ºC. The sample specific error estimates are between 1.4 and 1.5 ºC for all samples. On average 92.4 % of the identified fossil chironomids of the LM8 record were present in the modern training set and used to obtain palaeotemperature estimates (range: 85.6 – 100%). There are no rare taxa present in the fossil chironomid-assemblages, and only 1 sample (12.5 cm core depth) shows a ‘poor fit’ to temperature. Again, all samples have ‘no good’ analogues in the calibration data set (with the exception of the sample at 11.0 cm core depth which has ‘no close’ analogues). Although there are no good analogues for both the samples of the SR-X1 record and the LM8 record, Birks (1998) states that WA- PLS can perform relatively well in poor analogue situations. Heiri et al. (2007) for instance obtained reliable mean July air temperature reconstructions for Lake Hijkermeer, even during intervals were the fossil chironomid-assemblages showed 6 ‘no good’ analogue conditions. 6.6 Palaeoclimatic reconstructions from other proxy records 6.6.1 SR-X1 (Weichselian Early Glacial) Palaeobotanical analyses indicate that treeless vegetation existed at the Nochten-site during at least part of the Early Glacial (Figure 6.5). The lower part of the record shows values of Pinus of approximately 10%, increasing to ca. 25% in the upper part of the record. Lotter et al. (1992) suggest that abundances of Pinus pollen below the 20% cut-off most often reflect long distance transported pollen, and do not indicate the local presence of pine stands. However, tree pollen values may have been suppressed in our record due to the local presence of Cyperaceae and Poaceae at the study site. Other tree taxa that might indicate long-distance transported pollen (e.g. Alnus) also show increasing abundances towards the top of the record. There is however no macro-fossil evidence for the presence of trees at our study site. The local vegetation probably consisted of a low-shrub tundra dominated by Betula and Salix shrubs.

110 Chapter 6

Pollen Botanical macro-remains

ts as

hinae tobl , sta milis, budscalesles ragments nutlets hu nutlets e twig f , ostrata, ceae a nana/ trachium, nutlets Pinus etulaetula nana-typeetul a istatella mucedo Lithology Trees andUpland shrubsCyperacea herbsGrassesErica AlnusPicea B B CeratophyllumB Salix leafsp.,Salix ecbudsca sp. Carex aquatilis,Carex r B Cr 18

20 22

24

26 28

Depth (cm) 30

32

34 36

38

40 20 60 100% 10% 10 20 20 40 20 40 100 200 300 10 10

Figure 6.5: Selected pollen-taxa (expressed as relative abundances (%)), plant macrofossils (expressed as number of remains per sample), and the number of statoblasts of the bryozoan Cristatella mucedo in the SR-X1 record.

Macro-remains of aquatic plants suggest that the site was a shallow lake surrounded by a sedge swamp, which is consistent with the chironomid fauna. Leaf spines of Ceratophyllum demersum, found in low numbers throughout the record, indicate a minimum mean summer temperature (MMST) of 15 ºC (Litt 1994). Other abundant temperature-indicator species like Ranunculus subgen. Batrachium or Cristatella mucedo indicate a MMST of at least 10 ºC (respectively Brinkkemper et al. 1987; Lacourt 1968). Although the 15 ºC MMST reconstructed using plant-climate indicator species is unexpectedly high when compared to other climate reconstructions from the Niederlausitz region (see below), it does concur with our high chironomid-based palaeotemperature estimates of 15 ºC.

6.6.2 Other climate reconstructions from the Niederlausitz region for the Weichselian Early Glacial Bos et al. (2001) studied a sequence of Early Glacial deposits, and reconstructed vegetation that is initially dominated by pine (Pinus sylvestris). Many pine wood- fragments were found, as well as macro-fossils of other conifers (Abies, Picea, Larix). Pinus pollen also reached an abundance of 40-60% during the interval that the authors correlated with the Brørup Interstadial, whereas Betula only reached values between 8 to 18%. These pollen assemblages resemble pollen diagrams derived from other locations in the Lausitz area that date back to both the Odderade and Brørup

111 Chapter 6

Interstadial (Bos et al. 2001). Findings of macro-remains of Typha spp. and P. sylvestris suggest that the minimum mean summer temperature during the Early Glacial was >13 ºC. In the upper part of their Early Glacial sequence (correlated with the Rederstall stadial/ early Odderade Interstadial), Bos et al. (2001) record high values of NAP (reaching 60%) with much lower percentages of Pinus pollen (below 20%).. However, in the record of Bos et al. (2001), Betula reaches high values up to 30% whereas it only reaches abundances of 10% in our record. Bos et al. (2001) find local indicators of a reed swamp, including the massive presence of Typha seeds that indicate a MMST of 13 ºC. Within the youngest Early Glacial deposits, Bos et al. (2001) found poor microfossil and macrofossil assemblages. A minimum mean summer temperature around 8-10 ºC, is reconstructed, and there is no evidence for the presence of trees. A detailed comparison between our record and the different records of Bos et al. (2001) is hampered because of the highly variable character of the Early Glacial pollen-assemblages. We reconstruct a treeless vegetation with an increasing influence of long-distance transported pollen as well as reworked pollen (both including Pinus), and it might be possible that our record correlates best with the deposits described by Bos et al. (2001) as belonging to the earlier parts of the Odderade Interstadial. The MMSTs reconstructed by Bos et al. (2001) are in the order of 13 ºC, whereas our chironomid-based palaeotemperature estimates are 2 ºC higher. Using chironomid-remains, we reconstructed mean July air temperatures instead of minimum mean summer temperatures as derived from plant-indicator species. It was therefore also expected that the chironomid-based palaeotemperature estimates are slightly higher than the estimates derived from plant-climate indicator species. Furthermore, Bos et al. (2001) used a core taken from a deposit formed in a reed swamp, a habitat not suitable for many aquatic plants that are used in the climate- indicator species approach. This might have prevented the local occurrence of such 6 plants, and might potentially have caused an underestimation of the MMST. 6.6.3 LM8 (Early Pleniglacial) Bohncke et al. (in press) show that aquatic plant remains and pollen-assemblages from the LM8 record suggest a high minimum meanJuly air temperature of ca. 12-14 ºC shortly after the formation of the lake and during the initial period of deposition of sediments. They reconstruct treeless vegetation that is initially dominated by Betula shrubs and other tundra-elements. At 12.5 cm core depth the pollen record shows an increase in the pine pollen values, and Pinus becomes the dominant taxon. The authors interpret this change as an indication of a relative increase in long- distance transport as a result of a more patchy shrub tundra vegetation and a decreasing pollen production at the study site. Bohncke et al. (in press) furthermore demonstrate a larger abundance of Cenococcum geophilum sclerotia, as well as an increased presence of reworked palynomorphs, which they interpret as indicating a larger influence of surface erosion. We show in Figure 6.4 that at the same core depth,

112 Chapter 6 the chironomid concentration declines sharply, and that there is a shift in dominant chironomid taxa. The organic content of the sediment drops considerably to values of ~5% and the chironomid inferred palaeotemperature estimates (Figure 6.4) show a decrease by approximately 4 ºC. Bohncke et al. (in press) interpret their qualitative inference of cold climate conditions before the formation of the lake, high temperatures during the infilling of the lake and the abrupt transition towards cold- climate conditions near the top of the record as evidence for a Dansgaard/Oeschger (D/O)-like climate evolution phase (e.g. Dansgaard et al. 1993). Our quantitative chironomid-inferred temperature reconstruction supports the initial interpretation by Bohncke et al. (in press).

6.7 Comparison with palaeoclimatic reconstructions from NW Europe 6.7.1 Weichselian Early Glacial summer temperature reconstructions Aalbersberg and Litt (1998) suggested a minimum mean July air temperature around 14 ºC for northwestern Europe for the Odderade and Brørup Interstadials, with local temperatures in central and northern Germany that reach 15 ºC to 16 ºC. For the Niederlausitz area, the Odderade Interstadial showed climatic conditions that were similar to that of the Brørup Interstadial, with mean summer temperatures estimates reaching a maximum of 15 ºC. Both the Brørup and the Odderade Interstadial temperature estimates fit with our chironomid-inferred temperature estimates.

6.7.2 Weichselian Pleniglacial summer temperature reconstructions Although our chronology suggests that the sediments of LM8 were deposited during the Early Pleniglacial, other chronologies from the Niederlausitz area indicate an alternative early Middle Pleniglacial age (i.e. ~55 ka; see above). Very few terrestrial sites offer quantitative temperature estimates for the Early Pleniglacial in northwestern and central Europe. Using Coleopteran (beetle) based temperature estimates, Coope et al. (1997) reconstruct temperatures of the warmest month of 10 ºC to 13 ºC for central London and between 7 ºC and 11 ºC near Oxford, England. To the south, mean July temperatures of 15-16 ºC are reconstructed using coleopteran- assemblages from La Grande Pile, France (Ponel 1994). Fossil coleopteran (beetle) assemblages from several sites in England show that the climate during the Middle Weichselian was probably temperate and oceanic with mean monthly July temperatures reaching levels at least as warm as those of the present day (Coope 2002). Analysis of fossil coleopteran assemblages from the Oerel- site (northern Germany) suggested a mean July temperature of 11°C at this location (Behre et al. 2005). This cooler temperature is in contrast with the high temperatures reconstructed in England. Using 29 sites from north-western Europe, Huijzer and Vandenberghe (1998) estimate a minimum mean July temperature of ca 7 °C in Poland and 11 °C in western Europe using palaeobotanical data. However, the number of indicator

113 Chapter 6

species was low in the Polish records, which are therefore considered to be less reliable by the authors. The temperature estimates based on Coleoptera data from western Europe show that there have been intervals during the Early and Middle Pleniglacial during which the temperatures of the warmest month reached values that were almost equal to those of the present-day. The reconstructions based on palaeobotanical data represent minimum mean summer temperatures, and therefore are lower than the palaeotemperature estimates derived from coleopteran remains. Our data from eastern Germany, based on chironomid-assemblages, indicates that former temperatures during summer months were slightly lower than those of the present-day situation, which is in line with the coleopteran-based palaeotemperature estimates.

6.8 Conclusions Chironomid-based temperature reconstructions for the Weichselian Early Glacial average around 15 ºC, slightly lower than the present-day temperature at the study site. The pollen-assemblages from the same record indicate that treeless vegetation was present at the study site, which suggests a delayed vegetation response, as temperatures were sufficiently high for the development of a forest. The high chironomid-based palaeotemperature estimates are in accordance with aquatic plant- indicator macro-fossils found in the same sediment core. Climatic reconstructions for northwestern Europe based on Coleopteran remains also indicated relatively high summer temperatures for the Weichselian Early Glacial.

The chironomid-inferred July air temperatures for the Early Pleniglacial are around 15-16 ºC for a large part of the record, but a sharp decrease to 13 ºC is reconstructed near the top of the sediment sequence. The organic content of the sediment, the chironomid concentration and the local vegetation also show abrupt changes at this core depth, and suggest a D/O-like climate event. Although a direct comparison 6 between different sites is hampered because of the uncertain chronology, there are several sites in northwestern and central Europe that also yielded high palaeotemperature estimates for the Early Glacial and Pleniglacial, which are as high or nearly as high as at present.

This is the first study that uses lacustrine floodplain deposits in order to obtain quantitative, chironomid-based mean July air temperature reconstructions, and the results show that these sedimentary environments have a high potential for palaeoclimatic studies even considering the complex processes occurring on floodplains. The combination of chironomid-based temperature inference with other means for quantitative temperature reconstruction, such as the analysis macrofossils of aquatic macrophytes, seems a useful way forward in developing floodplain lake sediments as palaeoclimate archives.

114 Chapter 6

References Aalbersberg G, Litt T (1998) Multiproxy climate reconstructions for the Eemian and Early Weichselian. J Quatern Sci 13: 367-391 Becker A, Ammann B, Anselmetti FS, Hirt AM, Magny M, Millet L, Rachoud AM, Sampietro G, Wüthrich C (2006) Paleoenvironmental studies on Lake Bergsee, Black Forest, Germany. Neues Jahrbuch für Geologie und Paläontologie - Abhandlungen 240: 405-445 Behre KE, Hölzer A, Lemdahl J (2005) Botanical macro-remains and insects from the Eemian and Weichselian site of Oerel (northwest Germany) and their evidence for the history of climate. Veg Hist Archaeobot 14: 31– 53 Bigler C, Heiri O, Krskova R, Lotter AF, Sturm M (2006) Distribution of diatoms, chironomids and cladocera in surface sediments of thirty mountain lakes in south-eastern Switzerland. Aquat Sci 68: 154-171 Birks HJB (1998) Numerical tools in palaeolimnology – progress, potentialities, and problems. J Paleolimnol 20: 307-332 Birks HJB, Line JM, Juggins S, Stevenson AC, Ter Braak CJF (1990) Diatoms and pH reconstruction. Philos Trans R Soc Lond B 327: 263-278 Bohncke SJP, Bos JAA, Engels S, Heiri O, Kasse C (in press) Rapid climatic events as recorded in Middle Weichselian thermokarst lake sediments. Quatern Sci Rev Bos JAA, Bohncke SJP, Kasse C, Vandenberghe J (2001) Vegetation and Climate during the Weichselian Early Glacial and Pleniglacial in the Niederlausitz, eastern Germany - macrofossil and pollen evidence. J Quatern Sci 16: 269-289 Briant RM, Bateman MD, Coope GR, Gibbard PL (2005) Climatic control on Quaternary fluvial sedimentology of a Fenland Basin river, England. Sedimentology 52: 1397-1423 Brinkkemper O, Van Geel B, Wiegers J (1987) Palaeoecological study of a Middle Pleniglacial deposit from Tilligte, The Netherlands. Rev Palaeobot Palynol 51: 235-269 Brodersen KP, Lindegaard C (1999) Classification, assessment and trophic reconstruction of Danish lakes using Chironomids. Freshw Biol 42: 143-157 Brodersen KP, Quinlan R (2006) Midges as palaeoindicators of lake productivity, eutrophication and hypolimnetic oxygen. Quatern Sci Rev 25: 1995-2012 Brodin Y-W (1986) The postglacial history of Lake Flarken, southern Sweden, interpreted from subfossil insect remains. Intern Revue d ges Hydrobiol 71: 371-432 Brooks SJ (2006) Fossil midges (Diptera: Chironomidae) as palaeoclimatic indicators for the Eurasian region. Quatern Sci Rev 25: 1894-1910 Brooks SJ, Birks HJB (2000) Chironomid-inferred Late-glacial air temperatures at Whitrig Bog, south-east Scotland. J Quatern Sci 15: 759-764 Brooks SJ, Birks HJB (2001) Chironomid-inferred air temperatures from Lateglacial and Holocene sites in north- west Europe: progress and problems. Quatern Sci Rev 20: 1723-1741 Brooks SJ, Bennion H, Birks HJB (2001) Tracing lake trophic history with a chironomid-total phosphorus inference model. Freshw Biol 46: 513-533 Brooks SJ, Langdon PG, Heiri O (2007) The identification and use of Palaearctic Chironomidae larvae in palaeoecology. QRA Technical Guide No 10. Quaternary Research Association, London Coope GR (2002) Changes in the Thermal Climate in Northwestern Europe during Marine Oxygen Isotope Stage 3, Estimated from Fossil Insect assemblages. Quatern Res 57: 401-408 Coope GR, Gibbard PL, Hall AR, Preece RC, Robinson EJ., Sutcliffe AJ (1997) Climatic and environmental reconstructions based on fossil assemblages from Middle Devensian (Weichselian) deposits of the River Thames at South Kensington, Central London, UK. Quatern Sci Rev 16: 1163-1195 Dansgaard W, Johnsen SJ, Clausen HB, Dahl-Jensen D, Gundestrup NS, Hammer CU, Hvidberg CS, Steffensen JP, Sveinbjörndottir AE, Jouzel J, Bond G (1993) Evidence for general instability of past climate from a 250- kyr ice-core record. Nature 364: 218-220 Engels S, Bohncke SJP, Bos JAA, Brooks SJ, Heiri O, Helmens KF (2007a) Chironomid-based palaeotemperature estimates for northeast Finland during Oxygen Isotope Stage 3. J Paleolimnol: DOI 10.1007/s10933-007- 9133-y Engels S, Bohncke SJP, Heiri O, Nyman M (2007b) Intraregional variability in chironomid-inferred temperature estimates and the influence of river inundations on lacustrine chironomid assemblages. J Paleolimnol: DOI 10.1007/s10933-007-9147-5 Gandouin E, Ponel P, Andrieu-Ponel V, Franquet E, de Beaulieu J-L, Reille M, Guiter F, Brulhet J, Lallier-Vergès É, Keravis D, von Grafenstein U, Veres D (2007) Past environment and climate changes at the last Interglacial/Glacial transition (Les Échets, France) inferred from subfossil chironomids (Insecta). CR Geoscience 339: 337-346 Grimm EC (1991–2004) TILIA, TILA.GRAPH, and TGView. Illinois State Museum, Research and Collections Center, Springfield, USA http://demeter.museum.state.il.us/pub /grimm/ Heiri O, Lotter AF (2001) Effect of low count sums on quantitative environmental reconstructions: an example using subfossil chironomids. J Paleolimnol 26: 343-350

115 Chapter 6

Heiri O, Lotter AF (2005) Holocene and Lateglacial summer temperature reconstruction in the Swiss Alps based on fossil assemblages of aquatic organisms: a review. Boreas 34: 506-516 Heiri O, Lotter AF, Hausmann S, Kienast F (2003) A chironomid-based Holocene summer air temperature reconstruction from the Swiss Alps. Holocene 13: 477-484 Heiri O, Ekrem T, Willassen E (2004) Larval head capsules of European Micropsectra, Paratanytarsus and Tanytarsus (Diptera: Chironomidae: Tanytarsini). Version 1.0 http://www.bio.uu.nl/~palaeo/ Chironomids/Tanytarsini/intro.htm Heiri O, Cremer H, Engels S, Hoek W, Peeters W, Lotter AF (2007) Late-Glacial summer temperatures in the Northwest European lowlands: a new chironomid record from Hijkermeer, the Netherlands. Quatern Sci Rev: DOI 10.1016/j.quascirev.2007.06.017 Hill MO (1973) Diversity and evenness: a unifying notation and its consequences. Ecology 54: 427-432 Hiller A, Junge FW, Geyh MA, Krbetschek M, Kremenetski C (2004) Characterising and dating Weichselian organogenic sediments: a case study from the Lusatian ice marginal valley (Scheibe opencast mine, eastern Germany) Palaeogeog Palaeoclimatol Palaeoecol 205: 273-294 Huijzer B, Vandenberghe J (1998) Climatic reconstruction of the Weichselian Pleniglacial in northwestern and central Europe. J Quatern Sci 13: 391 - 417 Huntley DJ, Lian OB (2006) Some observations on tunnelling of trapped electrons in feldspars and their implications for optical dating. Quatern Sci Rev 25: 2503-2512 Ilyashuk EA, Ilyashuk BP, Hammarlund D, Larocque I (2005) Holocene climatic and environmental changes inferred from midge records (Diptera: Chironomidae, Chaoboridae, Ceratopogonidae) at Lake Berkut, southern Kola Peninsula, Russia. Holocene 15: 897-914 Iversen J (1954) The Late Glacial flora of Denmark and its relation to climate and soil. Danm Geol Unders 2: 88- 119 Juggins S (2003) C2 User guide. Software for ecological and palaeoecological data analysis and visualisation. University of Newcastle, Newcastle upon Tyne, UK Kaiser K, Barthelmes A, Czakó Pap S, Hilgers A, Janke W, Kühn P, Theuerkauf M (2006). A Lateglacial palaeosol cover in the Altdarss area, southern Baltic Sea coast (northeast Germany): investigations on pedology, geochronology and botany. Neth J Geosci 85: 197-220 Kasse C, Vandenberghe J, Van Huissteden J, Bohncke SJP, Bos JAA (2003) Sensitivity of Weichselian fluvial systems to climate change (Nochten mine, eastern Germany). Quatern Sci Rev 22: 2141-2156 Kolstrup E (1980) Climate and stratigraphy in Northwestern Europe between 30,000 BP and 13,000 BP, with special reference to The Netherlands. Meded Rijks Geol Dienst 32: 181-253 Lacourt AW (1968) A monograph of the freshwater Bryozoa - Phylactolaemata. Zool Verh 93: 3-159 Litt T (1994) Paläoökologie, Paläobotanik und Stratigraphie des Jungquartärs im nordmittel-europäischen Tiefland. Dissertationes Botanicae 227: 1-85 Lotter AF, Eichter U, Birks HJB, Siengenthaler U (1992) Late-glacial climatic oscillations as recorded in Swiss lake sediments. J Quatern Sci 7: 187-204 Mol J (1997a) Fluvial response to climate variations. The Last Glaciation in eastern Germany. PhD Thesis, Vrije Universiteit Amsterdam, Amsterdam Mol J (1997b) Fluvial response to Weichselian climate changes in the Niederlausitz (Germany). J Quatern Sci 12: 43-60 Moller Pillot HKM (1984). De larven der nederlandse Chironomidae (Diptera) (Inleiding, Tanypodinae & 6 Chironomini). Nederlandse Faunistische Mededelingen 1A, European Invertebrate Survey, Leiden, The Netherlands Moog O (1995) Fauna Aquatica austriaca. Abteilung für Hydrobiologie, Fischereiwirtschaft and Aquakultur der Universität für Bodenkultur, Wien Murray AS, Olley JM (2002) Precision and accuracy in the optically stimulated luminescence dating of sedimentary quartz: a status review. Geochronometria 21: 1-17 Murray AS, Wintle AG (2003) The single aliquot regenerative dose protocol: potential for improvements in reliability. Rad Meas 37: 377-381 Ponel P (1994) Climatic variations during the Würmian Pleniglacial inferred from fossil arthropod assemblages at La Grande Pile (Haute-Saone, France) C r Acad Sci 319: 845-852 Porinchu DF, MacDonald GM (2003) The use and application of freshwater midges (Chironomidae: insecta: diptera) in geographical research. Progr Phys Geogr 27: 378-422 Quinlan R, Smol JP (2001) Setting minimum head capsule abundance and taxa deletion criteria in chironomid- based inference models. J Paleolimnol 26: 327-342 Rieradevall M, Brooks SJ (2001) An identification guide to subfossil Tanypodinae larvae (Insecta: Diptera: Chironomidae) based on cephalic setation. J Paleolimnol 25: 81-99 Roberts RG, Galbraith RF, Olley JM, Yoshida H, Laslett GM (1999) Optical dating of single and multiple grains of quartz from Jinmium rock shelter, northern Australia: Part II, Results and Implications. Archaeometry 41: 365-395

116 Chapter 6

Schmid PE (1993) A key to the larval Chironomidae and their instars from Austrian danube region streams and rivers with particular reference to a numerical taxonomic approach. Part I. Diamesinae, Prodiamesinae and Orthocladiinae. Wasser und Abwasser Supplement 3/93: 1-524 Smol JP, Birks HJB, Last WM (2001a) Tracking Environmental Change Using Lake Sediments. Volume 3: Terrestrial, Algal,and Siliceous Indicators. Kluwer Academic Publishers, Dordrecht Smol JP, Birks HJB, Last WM (2001b) Tracking Environmental Change Using Lake Sediments. Volume 4: Zoological Indicators. Kluwer Academic Publishers, Dordrecht ter Braak CJF, Šmilauer P (1998) CANOCO reference manual and CanoDraw for Windows user’s guide. Biometris, Wageningen Van der Hammen Th, Maarleveld GC, Vogel JC, Zagwijn W (1967) Stratigraphy, climatic succession and radiocarbon dating of the last glacial in the Netherlands. Geol Mijnb 46: 79-95 Van Geel B, Coope GR, Van der Hammen T (1989) Palaeoecology and stratigraphy of the Lateglacial type section at Usselo (The Netherlands). Rev Palaeobot Palynol 60: 25–129 Velle G, Brooks SJ, Birks HJB, Willassen E (2005) Chironomids as a tool for inferring Holocene climate: an assessment based on six sites in southern Scandinavia. Quatern Sci Rev 24: 1429-1462 Von Gunten L, Heiri O, Bigler C, Van Leeuwen J, Casty C, Lotter AF, Sturm M (2007) Seasonal temperatures for the past ~400 years reconstructed from diatom and chironomid assemblages in a high-altitude lake (Lej da la Tscheppa, Switzerland). J Paleolimnol DOI 10.1007/s10933-007-9103-4 Walker IR, Cwynar LC (2006) Midges and palaeotemperature reconstruction – the North American experience. Quatern Sci Rev 25: 1911-1925 Walker IR, Smol JP, Engstrom DR, Birks HJB (1991) An assessment of Chironomidae as Quantitative Indicators of Past Climatic Change. Can J Fish Aquat Sci 48:975- 987 Walker IR, Levesque AJ, Cwynar LC, Lotter AF (1997) An expanded surface-water palaeotemperature inference model for use with midges from eastern Canada. J Paleolimnol 18: 165-178 Wallinga J (2002) Optically stimulated luminescence dating of fluvial deposits: a review. Boreas 31: 303-322 Wallinga J, Murray A, Duller G (2000) Underestimation of equivalent dose in 29 single-aliquot optical dating of feldspars caused by preheating. Radiation Meaurements 32: 691-695 Wiederholm T (1983) Chironomidae of the Holarctic region. Keys and diagnoses. Part I. Larvae. Entomol Scand 19: 1-457 Wolf L, Alexowsky W (1994) Fluviatile and glaziäre Ablagerungen am äusserten Rand der Elster- und Saalevereiserung; die spättertiäre and quartäire Geschichte des sächsischen Elbgebietes. Altenburger Naturwiss Forsch 7: 190-235 Zagwijn WH, Paepe R (1968) Die Stratigraphie der Weichselzeitlichen Ablagerungen der Niederlande und Belgien. Eiszeitalt u Gegenw 19: 129-146

117 Chapter 6

6

118 Chapter 7

Chapter 7

The lacustrine sediment record of Oberwinkler Maar (Eifel, Germany): chironomid-based inferences of environmental changes during Oxygen Isotope Stage-3

S Engels, SJP Bohncke, O Heiri, K Schaber, F Sirocko

Abstract The lacustrine record of Oberwinkler Maar (Eifel, Germany) is the northernmost continuous record that documents the Weichselian Pleniglacial in central Europe, a period that is characterized by multiple abrupt climate oscillations known as the Dansgaard/ Oeschger cycli. Here, the results of a high-resolution study of chironomid remains is presented, with focus on the earlier part of Oxygen Isotope Stage 3 (60 – 50 ka BP) covering 4 stadial/ interstadial cycles. During the stadials, the chironomid fauna of the former lake was dominated by many cold-stenothermic chironomid taxa, indicating a cold, oligotrophic lake. The concentrations of chironomid remains were lower during the interstadials, and featured a higher number of warm-indicating taxa. This could be the result of a higher summer temperature at the study site, but also of bottom-water anoxia, an increase in trophic state or a combination of these factors. The occurrence of a taxon restricted to (sub-) 7 arctic environments suggests a change in the temperature regime rather than in-lake processes as the driving mechanism for the changes in the chironomid record. Although there consistently was a response of the lake-ecosystem to climate changes, the amplitude of this response was not constant.

This manuscript was submitted for publication to Boreas

119 Chapter 7

7.1 Introduction In the North Atlantic region the period referred to as Oxygen Isotope Stage (OIS-) 3 is characterized by abrupt climatic changes (e.g. Johnsen et al. 1992; Dansgaard et al. 1993; Bond et al. 1993). These so-called Dansgaard/ Oeschger (D/O) events often start with an abrupt climate warming, followed by a period of several hundreds to a thousand years of gradual cooling, and end with a sudden return to cold-climate conditions (e.g. Dansgaard et al. 1993; Ganopolski & Rahmstorf 2001). It is expected that these dramatic climate changes as reconstructed in the Greenland ice core and the North Atlantic marine records also resulted in climatic shifts over the European continent. Continuous records of environmental and climatic change during the last glacial are limited to only a small number of continental sites in Europe. Several pollen records from southern and central Europe that cover OIS-3 suggest vegetation shifts that are linked to the climate evolution as reconstructed in the North Atlantic marine records (e.g. Reille et al. 2000). No records are available in northwestern Europe that continuously registered climate evolution during OIS-3. Fragmentary records dated back to OIS-3 suggest one or several interstadial periods at a single location (e.g. Van der Hammen et al., 1967; Zagwijn 1974; Coope 2002; Behre et al. 2005; Helmens et al. 2007; Engels et al. in press). However, the correlation of these records to the ice-core and marine records is hampered by geochronological problems (e.g. Shackleton 2006; Engels et al., submitted). The Oberwinkler Maar is a former lake that is now completely filled in by sediments. It is situated in western Germany, 300 km to the northeast of the classic terrestrial pollen site of La Grande Pile (de Beaulieu & Reille 1992), and on a N-S gradient with other long terrestrial climate-records like those of the Velay region (Reille & de Beaulieu 1988) and Les Echets (de Beaulieu & Reille 1984; Fig. 1). It therefore provides a unique archive for the reconstruction of climate change in a key- location in Europe. Greyscale measurements on the lake sediments from Oberwinkler Maar reveal abrupt changes from organic-rich to clastic sediment intervals that have been correlated with D/O-events (Sirocko et al. 2005). The analysis of fast-migrating biological proxies such as chironomids or other botanical and zoological micro and macro-fossils can help to reconstruct past environmental changes on a local scale (palaeolimnology) as well as on a regional scale (palaeoclimatology). 7 Chironomids are a diverse group of insects, whose larvae live in freshwater habitats such as rivers and lakes. Fossil chironomid remains derived from lake sediments have been used to reconstruct a range of different environmental parameters, including salinity (e.g. Eggermont et al. 2006), changes in trophic state of lakes (e.g. Brooks et al. 2001). Especially the use of chironomids to quantitatively infer past changes in mean July air temperature has received increasing attention over the past decade (e.g. Heiri & Lotter 2005; Brooks 2006; Walker & Cwynar 2006). Here we present the results of a detailed analysis of the chironomid-remains encountered in the Oberwinkler Maar sediments dated to early OIS-3 (spanning 60-50

120 Chapter 7 ka BP). The aim of this study is to assess how and to what extent the multiple abrupt climate changes (D/O-events) during early OIS-3 as revealed in the marine and ice- core records influenced the lacustrine ecosystem of Oberwinkler Maar.

7.2 Material The Oberwinkler Maar (50º09’N, 6º57’E) is a former maar lake in the Eifel, a volcanic field in Germany (Fig. 1). The West Eifel Volcanic Field shows 68 maar structures, and in the framework of the Eifel Laminated Sediment Archive (ELSA) project (Sirocko et al. 2005; Schaber & Sirocko 2005), 30 dry maars and maar lakes were cored in order to obtain an archive of climatic and environmental change spanning the last 140 ka BP. The lacustrine sediment record of Oberwinkler Maar is approximately 40 m long, and the longest diameter of the maar structure is 800 m. A chronology was established using accelerator mass spectrometry (AMS) radiocarbon dating and additional fine- tuning of the greyscale record of the core to the NorthGRIP δ18O record (see Sirocko et al. (2005) for details). It is part of the long ELSA greyscale stack 2005, which was constructed from 4 individual cores from the dry maar lakes of the west Eifel region and which was presented in Sirocko et al. (2005), where a study focussing on the last interglacial (or Eemian) was presented. The Oberwinkler Maar record covers the period from 65 to 22 ka BP, or D/O- events 18 to 2 and this is the first study focussing on the sediments from the earlier

58° 0° 8° 16°

52° 52°

Oberwinkler Maar

La Grande Pile 7 Les Echets 46° 46° Lac du Bouchet

8° 0° 8° 16°

Figure 7.1: Location of selected palaeoclimatological study-sites: the asterisk indicates the location of Oberwinkler Maar, the closed circles represents sites referred to in the introduction.

121 Chapter 7 S14 S15 S16 S17 D/O14 D/O15 D/O16 D/O17

50 100 Paracladopelma 50 100

cold water taxa water cold

gens-type

Micropsectra radialis-type Micropsectra

dius grimshawi-type dius 20 Stictochironomus rosenschoeldi-type Stictochironomus

20 Tanytarsus lu Tanytarsus

20 Heterotrissocla

Monodiamesa

bpilosus-type su

20 40 60 Protanypus eterotrissocladius eterotrissocladius

20 40 H eri-type

20 Paracladius

20 40 Corynocera oliv Corynocera

20 Sergentia coracina-type Sergentia

ype a orophila a

20

Paratanytarsus penicillatus-type Paratanytarsus s Pagastiell

20 Tanytarsus mendax-t Tanytarsus

seudochironomu 20 intermediate P otendipes pedellus-type otendipes

20 Micr

s-type

idellu s-type

50 100 150

sectrocladius sord sectrocladius P

20 Polypedilum nubeculosum-type Polypedilum Cricotopus cylindraceu Cricotopus

20

Chironomus anthracinus-type Chironomus

Pseudosmittia 20 40 Orthocladius-type

Warm waterWarm taxa

Phaenopsectra flavipes-type Phaenopsectra

nervosus-type

20 Procladius

ma lateralis-type ma

umosus-type 20 40 7 Dicrotendipes

20 Cladopel

20 Chironomus pl Chironomus myia

20 40 Ablabes Chironomid concentration Chironomid Chironomid concentration diagram showing selected taxa. Taxa are expressed as the number of remains per 100 g wet sediment. “D/O- 9 2 6 5 0 3 0 2 1

3 40 22 43 52 32 33 23 12 3 7

15 10 34 9 1 6 Count sum Count 3 26 81 27 3 4 2 3 3 4 1 1 2 3 6 4 5 4 52 26 28 58 51 38 38 13

35,50 36,00 36,50 37,00 37,50 38,00 38,50 39,00 Depth (m) Depth Figure 7.2: number” refers to the correlated Dansgaard/ Oeschger-event (see Schaber and Sirocko 2005; et al. 2005).

122 Chapter 7 part of this record. Especially this lowermost section of the sediment record is characterised by an alternation between organic-rich and clastic lake sediments, which are correlated with interstadials and stadials respectively (Sirocko et al. 2005; Schaber & Sirocko 2005). This paper presents the results of a high-resolution study of chironomid remains focussing on the depth interval between 39 and 35.5 m core depth, covering D/O16 to D/O13 (i.e. 60-50 ka BP).

7.3 Methods In an initial round of sampling, 1 cm thick samples were extruded from every organic-rich and every clastic sediment interval of the Oberwinkler Maar record. These samples were treated with cold KOH for 4 hours, and subsequently rinsed on a 100 ìm mesh to remove fine particles. Chironomid head capsules (hcs) were hand- picked from the sediments using a stereomicroscope (35 x magnification) and fine forceps. The process of hand-sorting the chironomids was prolonged due to the compact nature of the deposits. The lake sediments were extremely densely packed, probably as the result of desiccation of the Maar lake, and did not disintegrate completely during the laboratory process described above. A second batch of 1 cm thick samples was taken from the interval with the highest concentration of chironomid remains, i.e. the interval spanning D/O-cycles 13 – 16. Because of the nature of the sediment, a range of approaches was attempted in order to disintegrate the sediments with as little damage to the chironomid hcs as possible. Our final methodology included a treatment with warm KOH for several minutes, after which the samples were rinsed over a 100 µm sieve for the first time. The residues were then split into 4 or 5 subsamples, which were individually subjected to ultrasonic treatment using an ultrasonic cell disruptor (Branson Sonifier II, model 450) for 5 seconds. This ultrasonic cell disruptor uses mechanical vibration of a horn or probe, with a frequency up to 20 kHz, to induce cavitation or ‘cold boiling’. It differs from an ultrasonic bath (Lang et al. 2003) as the treatment is more violent and results in a faster and completer disintegration of the sediments. Again, the samples were rinsed over a 100 ìm sieve. When samples were still not completely disintegrated, the treatment with the cell disruptor was repeated until the sediments were entirely degraded. A comparison between samples that were treated with the ultrasonic cell disruptor and samples that underwent the standard treatment follows below. Chironomid remains were identified primarily following Wiederholm (1983), 7 Heiri et al. (2004) and Brooks et al. (2007). A chironomid concentration diagram was constructed using the TILIA and TILIA.GRAPH computer programs (Grimm 1991- 2004). Non-chironomid macro-fossils that were observed while hand-sorting the chironomid remains were also retrieved from the samples and an abundance diagram is presented below. As the samples had low count sums, several samples between 36.4 and 39.15 m core depth were amalgamated with adjacent samples in order to create higher count sums (23 - 81 hcs). Amalgamation of samples occurred only within a specific lithological unit. Square-root transformed percentage abundances of the chironomid

123 Chapter 7

zoological remains botanical remains

female catkininus fruitsscale

nid hc n ephippia statoblasts ceae oogonia atella eratopogo 35,50 Plumatella statoblast Cladocera Mites TrichopteranC Ephemeropteran mandibleCrist mandibleChara BetulaBetula treeBetula fruits pubescensPotamogeton nanaPinus fruitsSorbus sylvestris alpAlnus seedsRumex seedsglutinosa unripe fruits seed

36,00 S14

36,50 D/O14

S15 37,00 D/O15 554

37,50 S16 Depth (m) 831

38,00

D/O16 38,50

S17

39,00 D/O17

100 200 300 100 200 300 100 200 300 50 50 50 100 150 300 50 50 50 50 50 50 50 50

Figure 7.3: Zoological and botanical macro-remain record. Taxa are expressed as the number of remains per 100 g wet sediment. “D/O-number” refers to the correlated Dansgaard/ Oeschger-event (see Schaber and Sirocko 2005; Sirocko et al. 2005).

taxa of these amalgamated samples were used in a correspondence analysis (CA) to summarize the compositional change between adjacent samples (e.g. Lepš & Šmilauer 2003). Rare taxa were down-weighed in this analysis, which was performed using CANOCO v 4.5.

7.4 Results 7.4.1 Chironomid record (Fig. 7.2) A total of 55 samples with a weight range of 8.6 to 36.0 g were analysed and 48 chironomid taxa were identified in the Oberwinkler Maar sediments. As the abundance of chironomids is too low for reliable percentage calculation, we present our results in the form of concentrations of chironomid remains per 100 gram of wet sediment. The chironomid taxa have been classified according to their temperature 7 optima in modern training sets (e.g. Brooks & Birks 2001; Heiri 2001) or references in literature (Wiederholm 1983; Moller Pillot & Buskens 1990; Brooks et al. 2007).

The sediment interval that is correlated with D/O17 shows low chironomid concentrations and a chironomid fauna that is initially dominated by Pseudosmittia. Paracladopelma and Sergentia coracina-type are present in relatively high numbers during this interval. Zone S17 consists of 2 samples which include a total of only 5 chironomid head capsules: Paracladopelma and Phaenopsectra flavipes-type.

124 Chapter 7

D/O16 has the lowest chironomid concentrations of the record under consideration, and besides Paracladopelma includes taxa like Dicrotendipes nervosus- type, Polypedilum nubeculosum-type and Ablabesmyia. Zone S16 includes 12 samples with relatively high chironomid concentrations and count sums between 22 and 58 head capsules per sample. The first taxa that show high abundances in this part of the record are Chironomus anthracinus-type, S. coracina-type and Corynocera oliveri-type. The chironomid assemblages are dominated by a high number of cold-stenothermic taxa such as Heterotrissocladius subpilosus-type, Micropsectra radialis-type and Protanypus for the remainder of Zone S16. Although the D/O15 interstadial includes only 3 samples, a distinct shift in the chironomid fauna is observed. The assemblages include a number of taxa typical for intermediate temperatures like Psectrocladius sordidellus-type, Microtendipes pedellus- type and Pseudochironomus, but cold-water taxa like Paracladopelma and M. radialis-type are also present in this interval. Most of the abundant taxa were already present during S16, but 6 cold-stenothermic taxa that were abundant during S16 disappear from the chironomid record during D/O15. During S15 there is a mixed chironomid assemblage with very low chironomid-concentrations. Paracladopelma is the dominant taxon, but many of the taxa that were present during D/O15 occur in this interval as well. Chironomini such as C. anthracinus-type, Cladopelma lateralis-type and D. nervosus-type again make up an important part of the chironomid assemblage during D/O14. Count sums are very low in this part of the record and cold-stenothermic taxa are absent from this interval, except for Paracladopelma and single occurrences of H. grimshawi-type. Chironomid concentrations increase again during stadial S14, which can mostly be attributed to an increase in the abundance of Paracladopelma. Cold stenothermic taxa such as M. radialis-type and Protanypus return to the lake. A final shift back to a mixed chironomid-assemblage with low chironomid concentrations is witnessed in the uppermost sample, which is attributed to D/O13.

7.4.2 Zoological and botanical macro-remain record (Fig. 7.3) The presence of two types of Bryozoan statoblasts (Cristatella-type and Plumatella-type) shows a remarkable pattern: Cristatella-type is present in relatively high numbers during D/O17, whereas Plumatella-type is absent from the record during this interval. During D/O16, D/O15 and stadial S14, Plumatella-type is abundantly present whereas Cristatella-type is absent from the record. During the latest part of our record, 7 Cristatella-type reoccurs in the lake, and the concentrations of Plumatella-type decline again. Cristatella mucedo is considered to be a climate-indicator species, as it needs a minimum mean Tjul of >10 °C to develop (Lacourt 1968). Stadial S16 shows high abundances of cladoceran (water flea) ephippia. Oribatid mites are present throughout the record, and show a peak-occurrence at the onset of D/O15, when the cladoceran ephippia disappear from the record. Several other insect remains are encountered, such as head fragments and mandibles of Ephemeroptera (mayflies) and Ceratopogonidae (biting midges), but no clear trends can be discerned in their concentrations.

125 Chapter 7

The record of botanical macro-remains includes a total of 25 fruits of tree-birch (Betula pubescens and/ or B. pendula), indicating the presence of trees at the study site. Other tree-seeds or –fruits that were encountered include 1 pine (Pinus) fruit and 1 alder (Alnus) fruit. The presence of dwarf-birch is indicated by the finding of a single Betula nana fruit, and indicates a minimum mean summer temperature of 7 ºC (Brinkkemper et al. 1987). A clear pattern in the presence of Characeae oogonia is visible, as these are abundant during S14 and S16 and present at lower abundances during the rest of the record. This Characeae type includes oogonia from both the Nitella-type and the Chara-type, although the latter was much more abundant than the former.

7.5 Discussion 7.5.1 Ultrasonic cell disruptor Due to the limited availability of material and the large amount of time associated with the hand-sorting of the material, we limited ourselves to coarse estimates of the recovery-rates of chironomid hcs using both the ‘standard’ technique, which involves sample treatment with KOH, and a second technique adopted here, using an additional treatment with the Ultrasonic Cell Disruptor (UCD). The standard methodology that only involved treatment with KOH initially seemed to result in a good disintegration of the samples that belong to the stadial intervals, which consisted of silts and clays. 24 Samples were treated using only KOH and the recovery-rate was approximately 29 hcs per sample. The residues of these 24 stadial samples were later reanalysed using the UCD-technique, and an additional 14% of hcs was recovered from the sediments. The time involved in hand-sorting the samples after the treatment with the UCD was in the order of 2 hours per sample, whereas the sorting-time was approximately 7 hours per sample with the KOH- treatment only. When samples from the stadial intervals were first treated with the UCD, and the residues of this analysis were later used with the ‘standard’ procedure, no additional hcs were retrieved. The samples belonging to the interstadial intervals did not seem to be affected at all when treated with KOH. Although the samples do contain clays and other fine particles, the compact nature of the organic-rich samples meant that hand-sorting an entire sample (approximately 25 g wet sediment) took 24-32 hours. To assess how well the UCD worked on these samples, small aliquots (~ 4 g) were taken and first analysed using the standard treatment, after which the UCD treatment was again 7 applied to the residues. Picking time was reduced from 4 hours to 15 minutes per aliquot. When comparing the chironomid counts obtained from the 4 g aliquots (n=13), it became clear that samples that were treated only with KOH yielded more chironomid hcs than samples treated only with the UCD, with a maximum difference of approximately 50%. It is very well possible that, due to the violent treatment of the samples in the UCD, head capsules broke apart and were lost when rinsed over a 100 ìm sieve. Although the number of head capsules is very low, and a statistically reliable estimate can not be attempted, it did not seem to be the case that only the heavily sclerotized head capsules survived the UCD treatment, as several specimens

126 Chapter 7 of weaker-sclerotized taxa such as Ablabesmyia were encountered after the UCD treatment. Concluding, the treatment with the UCD shortened the time involved in hand- sorting the chironomid remains from the sediments with a factor 5-10, and yielded a relatively large amount of head capsules, even from samples that were previously treated using the standard technique of only applying KOH to the samples. However, results on separate aliquots of 13 samples showed that the recovery of hcs was much lower when applying the UCD technique compared to the KOH-treatment, which might be attributed to the loss of hcs during the violent sample-treatment. More extensive tests will be needed to assess whether the UCD technique introduces a bias to chironomid analysis, especially in cases where fossil samples are compared with untreated modern assemblages using a transfer function approach (e.g. Brooks & Birks 2001; Heiri et al. 2007).

7.5.2 Paleolimnology The lowermost sample that is analysed in this study shows a combination of high values of Cladoceran ephippia and a high concentration of Pseudosmittia. The latter taxon is considered to be indicative for (semi-) terrestrial habitats such as wet soils but can also occur in lacustrine habitats, where it is associated with aquatic macrophytes (Brooks et al. 2007; Brodin 1986). The high concentration of Pseudosmittia might indicate a low lake level during this stage. The next 16 samples (covering S17 and D/O16) show a low number of chironomid remains, which might be the result of anoxic bottom-water conditions. The macro-remain record shows a steady increase in Plumatella statoblasts, and a continuous presence of oribatid mites throughout this zone. The few chironomid taxa that are present include a number of Chironomini such as D. nervosus-type and C. lateralis-type. The presence of the different Chironomini taxa might indicate a limited availability of oxygen, as most Chironomini are tolerant of low oxygen concentrations as a result of their high haemoglobin levels. The lake was probably mesotrophic, as indicated by e.g. C. lateralis-type (Saether 1979), and had aquatic macrophytes in the littoral part of the lake. This is indicated by e.g. Dicrotendipes, a taxon that is often associated with macrophytes and occurs in meso- to eutrophic waters (Brodin 1986; Moller Pillot & Buskens 1990). One cold-stenothermic taxon, Paracladopelma, is present during this interval. This taxon occurs in the littoral of oligotrophic arctic lakes and is also known from deep oligotrophic lakes in central 7 Europe (Heiri & Lotter in press). It is most often associated with sandy substrates, and is intolerant of eutrophication (Wiederholm 1983).

During S16, C. anthracinus-type is one of the first taxa to appear in higher numbers. This taxon is known to colonise the sublittoral and profundal of stratified lakes (Heinis & Crummentuijn 1992; Hamburger et al. 1995; Beer et al. in press). At the same time, the number of Characeae oogonia also increases. Characeae are generally associated with clear, nutrient poor and alkaline hardwaters, and a change towards an oligotrophic lake system might explain the high abundances of Characaea oogonia

127 Chapter 7

Grayscale CA Scores 40 35 30 25 20 -10 1 2

36 S14 36 Cold, oligotrophic lake

D/O 14

S15 Warmer,oligo- 37 37 D/O 15 to mesotrophic

S16 Cold, oligotrophic lake Core depth (m)

38 38 D/O 16

S17 39 39 D/O 17

Figure 7.4: Pictures of the Oberwinkler Maar sediments covering the early part of Oxygen Isotope Stage 3, with greyscale measurements, a plot of the scores of amalgamated fossil chironomid samples on the first correspondence analysis (CA)-axis and interpretations of the changes in the chironomid fauna. Greyscale data and correlation to Dansgaard/ Oeschger-events after Sirocko et al. (2005).

(Hutchinson 1975; Simons & Nat 1996). Although several warm-water taxa with a preference for littoral habitats are present in low concentrations during this interval, the chironomid fauna is dominated by a large number of cold-stenothermic taxa like H. subpilosus-type, Corynocera oliveri-type, Paracladius and Protanypus. All these taxa are typical inhabitants of the profundal of cold oligotrophic lakes and require high oxygen levels to survive, indicating a distinct change in the lake environment from D/O16 to this stadial. The sediments of S16 are clastic, and consist of silt-sized quartz. Taxa like M. radialis-type and Paracladius, abundantly present in this interval, are most often 7 associated with clastic rather than organic sediments, although taxa that occur in organic-rich sediments like P. sordidellus-type are also present. The strong increase in the number of Cladoceran ephippia is remarkable. Cladocerans are known to produce more ephippia when environmental conditions change to less favourable conditions (stress), which might include a change in temperature, food resources or the introduction of a new predator (Sarmaja-Korjonen 2004). However, a major change in the cladoceran population, perhaps as a result in changes in the planktonic foodweb, might be a possible explanation for the sharp increase in the number of encountered ephippia as well.

128 Chapter 7

During D/O 15, we can discern a shift from a chironomid-assemblage dominated by cold-stenothermic taxa typical for deep oligotrophic lakes or cool littoral habitats to an assemblage that includes a number of taxa typical for intermediate temperatures and littoral habitats such as P. sordidellus-type, M. pedellus- type and Pseudochironomus. The abundance of mites also increases dramatically, whereas cladoceran remains disappear from the sediments. P. sordidellus-type and D. nervosus-type are often associated with aquatic macrophytes (Moller Pillot & Buskens 1990; Brodersen et al. 2001) and might indicate the local presence of these plants, and taxa such as M. pedellus-type are often associated with sandy sediment rather than organic-rich sediments (Hofmann 1984). Taxa typical of relatively warm habitats like Ablabesmyia are also present in higher numbers. Although all these changes could point to an increase in trophic state or temperature at the study-site, taxa typical of cold-water conditions and oligotrophic habitats such as Paracladopelma and M. radialis- type are still present in this interval. If the lake was still deep, these species might have retreated to the profundal parts of the lake, where they could have been present as “relics” of the preceding cooler periods.

Stadial S15 consists of only 2 samples, and shows a chironomid fauna that is similar to that of D/O15. The macro-remain diagram shows a decrease in the number of mites and Plumatella statoblasts. D/O14 is in many perspectives similar to D/O16: the number of chironomid remains is extremely low, and the chironomid-assemblages are a mix of taxa that are most often associated with warm- to intermediate temperatures and the cold-stenothermic taxa Heterotrissocladius grimshawi-type and Paracladopelma. Chironomini make up an important part of the chironomid-fauna, and again might indicate low-oxygen conditions and a slight increase in trophic state, perhaps to mesotrophic conditions. The macro-remain record is very poor during this interval, only registering the continued presence of oribatid mites and single occurrences of Plumatella statoblasts and cladoceran ephippia.

Stadial S14 shows a sharp transition towards different environmental conditions. An increase in chironomid concentrations can be seen at the onset of this stadial, which can mostly be attributed to an increase in the abundance of Paracladopelma and a return of M. radialis-type to the lake. This latter taxon is indicative of oligotrophic and cold lakes (Bretschko 1974; Gerstmeier 1989; Brooks et al. 2007). Protanypus also indicates cold climate or deep, oligotrophic conditions at the study site (e.g. Walker & 7 MacDonald 1995). C. oliveri-type is a taxon that has hitherto mainly been found as a fossil or subfossil and is restricted to arctic/subarctic condition. Characeae oogonia are again abundant and might indicate a return to oligotrophic, clear-water conditions.

7.5.3 Environmental changes recorded in the Oberwinkler Maar record In the Oberwinkler Maar record there are clear transitions between dark intervals, associated with a high organic carbon content, and light sections which consist of fine clastic sediments, mostly silt-sized quartz grains (Fig. 4). Millimetre-to-centimetre

129 Chapter 7 Maar Bretschko (1974); Saether (1979); Gerstmeier (1989); Heiri and Lotter (2003); Brooks Birks (2001) Saether (1979); Itkonen et al. (1998); Nyman and Krhola (2005) References Moog (1994); Heiri and Millet (2005) Saether (1979); Brooks and Birks (2001); Heiri Millet (2005) Bretschko (1974); Saether (1979); Heiri and Millet (2005) Moog (1994); Brooks and Birks (2001); Heiri (2001) Brodersen and Anderson (2002); Nyman Korhola (2005) Moog (1994); Brooks and Birks (2001); Heiri (2001) Cranston et al. (1993); Moller Pillot and Buskens (1990); Moog (1994) Moog (1994); Olander et al (1999); Heiri (2001) Ecology Cold alpine and arctic lakes, hypolimnion of deep temperate oligotrophic lakes Cold arctic lakes and hypolimnion of ultraoligotrophic arctic, subarctic and boreal lakes Cold alpine and arctic lakes, hypolimnion of deep temperate oligo/mesotrophic lakes Arctic and subarctic lakes Cold alpine and arctic lakes, hypolimnion of deep temperate oligotrophic lakes Littoral of relatively warm (subarctic/subalpine- temperate) lakes Littoral of relatively warm (subarctic/subalpine- temperate) lakes Littoral of lakes over wide temperature gradient Littoral and profundal of lakes Littoral of lakes, somtimes (semi-) terrestrial 7 0.695 1.154 1.287 1.294 1.397 -0.932 -0.832 -0.767 -0.635 -0.415 score CA axis 1 e I ype

-t -typ type type es -type Scores on the first Correspondence analysis (CA) axis and summarized ecology of 10 abundant chironomid taxa in Oberwinkler M. radialis- C. oliveri Taxon Paracladius H. subpilosus Ablabesmyia C. mancus- Procladius Limnophy D. nervosus Protanypus sediments, that are also characterized by extreme scores on the first CA axis. Table 7.1

130 Chapter 7 scale laminae are present in both the organic-rich and the clastic intervals, although they clearly differ from the varved sediments that were recovered from several Eifel Maar lakes for two interglacial periods (Schalkenmehren Maar (OIS-1) and Maar west of Hoher List (OIS-5e); Sirocko et al. 2005; Schaber & Sirocko 2005). The laminae are probably the result of periods of increased primary production versus periods of increased input of clastic sediments with a possible aeolian origin. The sharp transitions as witnessed in the greyscale of the record are accompanied by the changes we reconstructed by employing chironomids as a proxy. Correspondence analysis (CA) shows that the samples from the stadial periods all feature low values on the first CA-axis, whereas the samples from the interstadials are characterized by high values. This implies that the chironomid assemblages in the two classes of samples differ in a systematic way, but it does not provide an ecological explanation. This type of ordination method configures the first axis to correspond with the greatest variability in the chironomid assemblages, which is not necessarily related to a single environmental variable (Lepš and Šmilauer 2003).

Table 7.1 indicates the axis scores and summarizes the ecology of 10 abundant chironomid taxa in the Oberwinkler Maar sediments that are characterized by extreme scores on the first CA axis. Taxa that plot towards the negative extreme of the first CA axis include cold-stenothermic species such as Paracladius, Protanypus and H. subpilosus- type. Of these taxa also favour oligotrophic conditions. Paracladius requires relatively high levels of oxygen in order to survive (Moller Pillot & Buskens 1990). Most of these taxa can also persevere in deep stratified lakes under temperate climatic conditions provided that the cold hypolimnion provides sufficiently high oxygen conditions for these polyoxybiontic chironomids. The first abundant taxon that shows a high positive score on the first CA axis is Limnophyes. This taxon is a typical inhabitant of very shallow waters, and is often encountered in (semi-) terrestrial habitats as well, and is capable of surviving in habitats with a low oxygen availability (Moller Pillot & Buskens 1990). The tolerance of low oxygen availability is also known for other taxa that plot toward the positive side of CA axis 1 (e.g. Procladius (Heiri & Lotter 2003), C. plumosus-type (Brooks et al. 2007 and references therein)), which is in sharp contrast to the majority of the species that plot toward the negative side of CA axis 1. All taxa that plot towards the positive side of CA axis 1 prefer meso-to eutrophic environments (e.g. D. nervosus-type, Brodin 1986) and have a temperature optimum that classifies them as intermediate to warm taxa. 7

The taxa shown in Table 7.1 suggest that the type of sediment might not have been of importance in determining the composition of the chironomid assemblages. Taxa that plot towards the negative side of the first CA-axis as well as those that plot towards the positive side include chironomids that preferentially occur on either clastic sediments or on organic sediments. The ecology of chironomid taxa characterizing stadials and D/O events (Table 7.1) suggests that limnological conditions in the lake have shifted from oligotrophic conditions with a high oxygen availability to more eutrophic conditions with clear

131 Chapter 7

indications of low oxygen conditions in the lake, at least during parts of the seasonal cycle. Many of the cold stenothermic chironomid taxa found in the record can also colonize the hypolimnion of deep lakes in temperate climates. However, the sporadic presence of subarctic/arctic faunal elements during at least some of the stadials (e.g. C. oliveri-type; Table 7.1) suggests that the changes in oxygenation and trophic state recorded in the chironomid record were most probably associated with a change in temperature regime.

7.6 Conclusions The lacustrine record of Oberwinkler Maar provides the unique opportunity to study the Weichselian Pleniglacial climate evolution on the European continent in a continuous setting. In this paper, we presented the first results of a high-resolution analysis of chironomid remains from the earlier part of OIS-3, covering D/O16- D/ O13. The aim of this study was to assess the impact of the multiple abrupt climate changes (D/O-events) during early OIS-3 on the lacustrine ecosystem of Oberwinkler Maar. Using chironomids and other encountered macroremains as a proxy we conclude the following: 1) The former Oberwinkler Maar lake was a cool, oligotrophic lake during two prominent stadials (S14 & S16). Taxa that dominated the chironomid faunas during these intervals were cold-stenothermic taxa such as H. subpilosus-type, Protanypus, Paracladius and M. radialis-type. 2) A slight increase to mesotrophic conditions is suggested during the interstadials. The higher relative abundance of many Chironomini during the interstadials might indicate higher summer temperatures at the study site, but may also be related to lower oxygen availability or a higher nutrient availability. 3) The occurrence of C. oliveri-type, a taxon that has so far only been recorded in subarctic and arctic environments, suggests that the changes in oxygen availability and trophic state, as suggested by the chironomid record, were most probably associated with a change in temperature regime. 4) Although there consistently was a response of the lake-ecosystem to climate changes, the amplitude of this response was not constant. Climate changes during two of the stadial periods had a strong impact on the chironomid fauna, whereas in the 2 other stadials only minor changes in the relative 7 chironomid abundances were recorded. This is the first research project focusing on the Eifel Maar lake sediments from OIS-3, and these older sediments called for a new methodological approach in order to extract the chironomid remains from the sediments. Initial results suggest that the tested UCD approach might be too destructive for fossil chironomids. This method will need to be further tested and refined in order to assess and optimize its performance. Ongoing investigations on the Oberwinkler Maar record will provide for new lines of evidence with respect to local and regional environmental changes, and our data provides the first biological evidence for local and regional environmental changes during the earlier part of OIS-3 for this location.

132 Chapter 7

References Beer R, Heiri O, Tinner W (in press) Vegetation history, fire history and lake development recorded for 6300 years by pollen, charcoal, loss on ignition, and chironomids at a small lake in southern Kyrgyzstan (Alay range, Central Asia). The Holocene Behre KE, Hölzer A, Lemdahl G (2005) Botanical macro-remains and insects from the Eemian and Weichselian site of Oerel (northwest Germany) and their evidence for the history of climate. Veget Hist Archaeobot 14: 31-53 Bond G, Broecker WS, Johnsen SJ, McManus J, Labeyrie L, Jouzel J, Bonani G (1993) Correlations between climate records from North Atlantic sediments and Greenland ice. Nature 365: 143–147 Bretschko, G (1974) The chironomid fauna of a high-mountain lake (Vorderer Finstertaler See, Tyrol, Austria, 2237 m asl). Ent Tidskr Supp 95: 22-33 Brinkkemper O, Van Geel B, Wiegers J (1987) Palaeoecological study of a Middle-Pleniglacial deposit from Tilligte, The Netherlands. Rev Palaeobot Palynol 51: 235-269 Brodersen KP, Anderson NJ (2002) Distribution of chironomids (Diptera) in low arctic West Greenland lakes: trophic conditions, temperature and environmental reconstruction. Freshw Biol 47: 1137-1157 Brodersen KP, Odgaard BV, Vestergaard O, Anderson NJ (2001) Chironomid stratigraphy in the shallow and eutrophic Lake Søbygaards, Denmark: chironomid-macrophyte co-occurrence. Freshw Biol 46: 253-267 Brodin Y-W (1986) The postglacial history of Lake Flarken, southern Sweden, interpreted from subfossil insect remains. Internationale Revue der Gesamten Hydrobiologie 71: 371-432 Brooks SJ (2006) Fossil midges (Diptera: Chironomidae) as palaeoclimatic indicators for the Eurasian region. Quatern Sci Rev 25: 1894-1910 Brooks SJ, Birks HJB (2001) Chironomid-inferred air temperatures from Lateglacial and Holocene sites in north- west Europe: progress and problems. Quatern Sci Rev 20: 1723-1741 Brooks SJ, Langdon PG, Heiri O (2007) The identification and use of Palaearctic Chironomidae larvae in palaeoecology. Quaternary Research Association Technical Guide 10, 276 pp Coope, GR (2002) Changes in the Thermal Climate in Northwestern Europe during Marine Oxygen Isotope Stage 3, Estimated from Fossil Insect assemblages. Quatern Res 57: 401-408 Dansgaard W, Johnsen SJ, Clausen HB, Dahl-Jensen D, Gundestrup NS, Hammer CU, Hvidberg CS, Steffensen JP, Sveinbjörndottir AE, Jouzel J, Bond G (1993) Evidence for general instability of past climate from a 250-kyr ice- core record. Nature 364: 218-220 De Beaulieau JL, Reille M (1984) A long Upper Pleistocene pollen record from Les Echets, near Lyon, France. Boreas 13: 111-132 De Beaulieau JL, Reille M (1992) The last climatic cycle at La Grande Pile (Vosges, France) a new pollen profile. Quatern Sci Rev 11: 431-438 Eggermont H, Heiri O, Verschuren D (2006) Fossil Chironomidae (Insecta:Diptera) as quantitative indicators of past salinity in African Lakes. Quatern Sci Rev 25: 1966-1994 Engels S, Bohncke SJP, Bos JAA, Brooks SJ, Heiri O, Helmens K (in press) Chironomid-based palaeotemperature estimates for northeast Finland during Oxygen Isotope Stage 3. J Paleolimnol Engels S, Bohncke SJP, Bos JAA, Heiri O, Vandenberghe J, Wallinga J (submitted) Chironomid-based temperature reconstructions on fragmentary records from the Weichselian Early Glacial and Pleniglacial of the Niederlausitz area (eastern Germany). Palaeogeog Palaeoclimatol Palaeoecol Ganopolski A, Rahmstorf S (2001) Rapid changes of glacial climate simulated in a coupled climate model. Nature 409: 153-158 Gerstmeier R (1989) Phenology and bathymetric distribution of the profundal chironomid fauna in Starnberger See (F.R. Germany). Hydrobiologia 184: 29-42 Grimm EC (1991–2004) TILIA, TILA.GRAPH, and TGView. Illinois State Museum, Research and Collections Center, Springfield, USA http://demeter.museum.state.il.us/pub /grimm/ Hamburger K, Dall PC, Lindegaard C (1995) Effects of Oxygen Deficiency on Survival and Glycogen-Content of Chironomus-Anthracinus (Diptera, Chironomidae) under Laboratory and Field Conditions. Hydrobiologia 297: 187-200 7 Heinis F, Crummentuijn T (1992) Behavioural response to changing oxygen concentrations of deposit feeding chironomid larvae (Diptera) of littoral and profundal habitats. Arch Hydrobiol 124: 173-185 Heiri O (2001) Holocene palaeolimnology of Swiss mountain lakes reconstructed using subfossil chironomid remains: past climate and prehistoric human impact on lake ecosystems. PhD thesis, University of Bern, pp 113 Heiri O, Lotter AF (2003) 9000 years of chironomid assemblage dynamics in an Alpine lake: long-term trends, sensitivity to disturbance, and resilience of the fauna. J Paleolimnol 30: 273-289 Heiri O, AF Lotter (in press) Chironomidae (Diptera) in Alpine lakes in Switzerland: a study based on subfossil assemblages in lake surface sediments. Boletim do Museu Municipal do Funchal Heiri O, Millet L (2005) Reconstruction of Late Glacial summer temperatures from chironomid assemblages in Lac Lautrey (Jura, France). J Quatern Sci 20: 33–44 Heiri O, Ekrem T, Willassen E (2004) Larval head capsules of European Micropsectra, Paratanytarsus and Tanytarsus (Diptera: Chironomidae: Tanytarsini). Version 1.0. http://www.bio.uu.nl/~palaeo/Chironomids/Tanytarsini/ intro.htm

133 Chapter 7

Heiri O, Cremer H, Engels S, Hoek W, Peeters W, Lotter AF (2007) Late-Glacial summer temperatures in the Northwest European lowlands: a new lacustrine chironomid record from Hijkermeer, the Netherlands. Quatern Sci Rev DOI 10.1016/j.quascirev.2007.06.017 (Published online via ScienceDirect) Helmens KF, Bos JAA, Engels S, Van Meerbeeck C, Bohncke SJP, Renssen H, Heiri O, Brooks SJ, Seppä H, Birks HJB, Wohlfarth B (2007) Ice-free conditions and present-day temperatures during the last glacial at 50ka in the central area of the Scandinavian glaciations. Geology Hofmann W (1984) Stratigraphie subfossiler Cladocera (Crustacea) and Chironomidae (Diptera) in zwei Sedimentprofilen des Meerfelder Maares. Courier Forschungs Institut Senckenberg 65: 67-80 Hutchinson GE (1975) A treatise on limnology, 3. Limnological Botany. J Wiley & Sons, New York, pp 660 Itkonen A, Marttila V, Meriläinen JJ, Salonen V-P (1999) 8000-year history of palaeoproductivity in a large boreal lake. J Paleolimnol 21: 271-294 Johnsen SJ, Clausen HB, Dansgaard W, Fuhrer K, Gundestrup N, Hammer CU, Iversen P, Jouzel J, Stauffer B, Steffensen JP (1992) Irregular glacial interstadials recorded in a new Greenland ice core. Nature 359: 311-313 Lacourt AW (1968) A monograph of the freshwater Bryozoa – Phylactolaemata. Zool Verh 93: 3-159 Lang B, Bedford AP, Richardson N, Brooks SJ (2003) The use of ultra-sound in the preparation of carbonate and clay sediments for chironomid analysis. J Paleolimnol 30: 451-460 Lepš J, Šmilauer P (2003) Multivariate Analysis of Ecological Data using CANOCO. University Press, Cambridge Moller Pillot HKM, Buskens RFM (1990) De larven der Nederlandse Chironomidae. Autoecologie en verspreiding. Ned Faun Meded 1c: 1-87 Moog O (1995) Fauna aquatica austriaca. Abteilung für Hydrobiologie, Fischereiwirtschaft und Aquakultur der Universität für Bodenkultur, Wien Nyman MT, Korhola A (2005) Chironomid-based classification of lakes in western Finnish Lapland. Boreal Environ Res 10: 239-254 Olander H, Birks HJB, Korhola A, Blom T (1999) An expanded calibration model for inferring lakewater and air temperatures from fossil chironomid assemblages in northern Fennoscandia. Holocene 9: 279-294 Pinder LCV, Reiss F (1983) The larvae of Chironomidae (Diptera: Chironomidae) of the Holarctic region. In: Wiederholm T (1983) Chironomidae of the Holarctic region. Keys and diagnoses. Part I. Larvae. Entomol Scand 19 Reille M, de Beaulieu JL (1988) History of the Würm and Holocene vegetation in Western Velay (Massif Central, France): a comparison of pollen analysis from three corings at Lac du Bouchet. Review of Palaeobotany and Palynology 54: 233-248 Reille M, de Beaulieu J-L, Svobodova H, Andrieu-Ponel V, Goeury C (2000) Pollenanalytical biostratigraphy of the last five climatic cycles from a long continental sequence from the Velay region (Massif Central, France). J Quatern Sci 15: 665-685 Sæther, OA (1979) Chironomid communities as water quality indicators. Holarctic ecology 2: 65-74 Sarmaja-Korjonen, K (2004) Chydorid ephippia as indicators of past environmental changes – a new method. Hydrobiologia 526: 129-136 Schaber K, Sirocko F (2005) Lithologie und Stratigraphie der spätpleistozänen Trockenmaare der Eifel. Mainzer Geowiss Mitt 33: 295-340 Simons J, Nat E (1996) Past and present distribution of stoneworts (Characeae) in The Netherlands. Hydrobiologia 340: 127-135 Sirocko F, Seelos K, Schaber K, Rein B, Dreher F, Diehl M, Lehne R, Jäger K, Krbetschek M, Degering D (2005) A late Eemian aridity pulse in central Europe during the last glacial inception. Nature 436: 833-836 Van der Hammen Th (1971) The Denekamp, Hengelo and Moershoofd Interstadials. Meded Rijks Geol Dienst 22: 81-85 Walker IR, Cwynar LC (2006) Midges and palaeotemperature reconstruction – the North American experience. Quatern Sci Rev 25: 1911-1925 Walker IR, MacDonald GM (1995) Distribution of Chironomidae (Insecta: Diptera) and other freshwater midges with respect to treeline, Northwest Territories, Canada. Arctic Alpine Res 3: 256-263 Wiederholm T (1983) Chironomidae of the Holarctic region. Keys and diagnoses. Part I. Larvae. Entomol Scand 7 19

134 Chapter 8

Chapter 8

Synthesis / Epilogue

8.1 Introduction The first applications of chironomids as a proxy for past temperature changes focused on reconstructing the climate evolution during the Late-Glacial period (ca 11-15 ka BP). This is the most recent period in the geological record that shows large- amplitude climate oscillations. Not only the large-scale changes in mean July air temperatures such as the Younger Dryas/ Holocene transition were reconstructed, but also more subtle oscillations such as, for example, the so-called Gerzensee Oscillation were recorded by chironomid analysis. In Europe, chironomid-inferred palaeotemperature records covering (parts of) the Late-Glacial are available for Norway (Brooks and Birks 2000a), the UK (Brooks and Birks 2000b; Bedford et al. 2004), the Swiss Alps (Brooks 2000; Heiri and Lotter 2005), the Jura mountains (Heiri and Millet 2005), the Netherlands (Heiri et al. 2007a) and Italy (Heiri et al. in press). Recent studies have also attempted to reconstruct the smaller-scale climate variability during the Holocene (ca 11 ka BP – 0), and quantitative temperature- reconstructions are available for Norway (Velle et al. 2005), Sweden (Larocque et al. 2001; Bigler et al. 2002), Finland (Seppä et al. 2002; Korhola et al. 2002), Switzerland (Heiri et al. 2003; Heiri and Lotter 2005) and England (Langdon et al. 2004; Marshall et al. 2007).

However, before the onset of this project, palaeoclimatological records based on chironomids were not available for Europe for the periods predating the Late-Glacial period. The scarcity of lacustrine records, together with the suboptimal preservation of chironomid head capsules (with head capsules often missing diagnostic features) and the nature of the sediments (often compacted or desiccated) hindered

135 Chapter 8 chironomid-based reconstructions of July air temperatures on lacustrine sediments dated to the earlier parts of OIS-3. In the preceding chapters, newly developed chironomid records from three different sites covering parts of OIS-3 to -5 have been discussed in detail. From the lacustrine record of Sokli, North Finland, unique data were derived reconstructing environmental and climatic evolution in a region where no such data were previously available. The Oberwinkler Maar record is one of the few continental sites where continuous sedimentation was recorded throughout OIS-3, and the chironomid record revealed that the D/O-like climate variability as witnessed in the marine and ice-core records also affected central Europe. The fragmentary records from eastern Germany predated OIS-3. However, they suggest that the formation of at least one of the thawlake deposits was climate-driven, and chironomid-inferred temperatures indicate high (almost present-day) mean July air temperatures, in line with evidence from other proxies and from different sites.

In this final chapter of this thesis, several issues concerning the application of chironomids in palaeoclimatology and problems encountered during the analysis of chironomid results presented in the previous chapters are critically discussed (section 8.2), and the chironomid-based climate reconstructions are compared to other published records (section 8.3). A short section (8.4) with suggestions for future research will conclude this chapter.

8.2 Proxy evaluation In this thesis, several chironomid-based mean July air temperature reconstructions are presented and validated in different ways. Chironomid-based palaeotemperature estimates can be compared to other proxy-based palaeotemperature estimates, as has been done in chapters 2-4 and 6. Climate model simulations have produced patterns of atmospheric circulation for interstadial conditions during OIS-3, and the resulting pattern of mean July air temperature over Europe can be compared to the chironomid-based palaeotemperature estimates, as has been done in chapters 2 and 3. Finally, several numerical methods are available to test the representation of fossil chironomid samples in the applied modern training sets, and a fit-to-temperature can be calculated as for instance has been done in chapter 2 and 6. Although one or more of these methods is often available to test the reliability of chironomid-based palaeotemperature estimates, a critical attitude towards organism-based climate reconstructions is always required.

8.2.1 Chironomids as a proxy for temperature Temperature has a dominant role in every aspect of the chironomid life cycle (Brooks et al. 2007), as it governs the rate of the development of eggs, larval and pupal development and many aspects of adult behaviour (Brooks et al. 2007, and references therein). Direct effects of water temperature on the development of chironomids might include the postponing of pupation or the timing of eclosion (emergence of

136 Chapter 8 imagos from pupae), and indirect effects include changes in trophic state or changes in food and oxygen availability. It therefore makes sense that, on the broad scale, chironomids are useful indicators of past changes in temperature (e.g. Walker and Matthewes 1987). The relationship between chironomids and temperature has been the subject of extensive discussions in the literature, the most recent reviews including Porinchu and MacDonald (2003), Velle et al. (2005), Brooks (2006) and Walker and Cwynar (2006). However, as for instance stated by Brodersen and Anderson (2002), factors controlling chironomid distribution (in this case in West Greenland lakes) are highly multivariate, and the correlations between these different factors imply that it is not always possible to pinpoint a single variable as the main controlling factor responsible for chironomid distribution on the local scale. When the different factors controlling chironomid abundances act similarly in the training sets and in the fossil records to which the training set is applied, quantitative inferences of past temperatures at these fossil sites will be reliable. However, when these factors behaved independently and/ or differed from those represented in the modern training sets, chironomid-based climate reconstructions potentially become unreliable.

In shallow lakes, there is generally a strong relationship between air temperature and water temperature, as shown by Livingstone and Lotter (1998) for a training set developed in the Swiss Alps. In modern training sets, lakes that have strong input of glacial melt water or cold water from another source invariably show cooler inferred air temperatures than expected (Brooks and Birks 2001; Heiri unpublished data). The decoupling of air and water temperature in this case results in temperature inferences that are too low, which suggests that the relative abundances of different chironomid taxa in a lake are more directly influenced by water temperature than air temperature (Velle et al. 2005).

Decoupling of air temperature and water temperature might also happen in lakes with thermal stratification. Especially in deep lakes, the profundal is usually colder than the surface waters during the summer months, and temperature inferences based on chironomid-assemblages from lakes with thermal stratification might produce temperatures that are too low (Velle et al. 2005). In our dataset from Finnish Lapland this was for instance evident in Lake 18 (Chapter 5). This lake was 15.9 m deep and showed thermal stratification at the time of sampling, and produced chironomid- based temperature estimates that were distinctly lower than the temperature measured at the study site. The chironomid-assemblages encountered in the Oberwinkler Maar record (Chapter 7) indicate that the former lake was a relatively deep lake, and might therefore have experienced seasonal stratification of the water column, potentially resulting in a decoupling of the air and water temperatures at the site.

137 Chapter 8

8.2.2 Other factors than temperature affecting the composition of the chironomid fauna of lakes Productivity On a broad scale, temperature and primary productivity act in a similar way: a higher temperature generally results in a higher productivity. Trophic variables such as total phosphorous (TP) often show a strong correlation with temperature in modern training sets (Brodersen and Anderson 2002, and references therein), and it is therefore difficult to decouple the influence of these two parameters on the chironomid fauna of modern lakes incorporated in a training set developed over an altitudinal or latitudinal gradient. Heiri and Lotter (2005) studied the co-variation between TP concentrations and summer temperature in detail for their calibration data set from the Swiss Alps. Their results show that summer temperature can explain a highly significant proportion of the variance in the chironomid assemblages independent of TP concentrations, indicating that temperature rather than TP is the most important variable in determining the composition of the chironomid fauna of a lake. In our study of modern lakes in Finnish Lapland (Chapter 5) we have no direct measurements of productivity, but values of total phosphorous (TP) and total organic carbon (TOC) did vary considerably between the different lakes, whereas the chironomid-based temperature inferences were not significantly affected by these differences. The former lakes encountered in the Niederlausitz area (Chapters 4 and 6) may have experienced fluctuations in e.g. TP as a result of river floodings, but based on our results obtained from Finnish Lapland we expect the chironomid-based temperature inferences presented in chapters 4 and 6 not to be affected by these inundations.

Organic content of the sediment Similar to the relationship between productivity and temperature, a strong relationship between organic content of the lake sediment (usually estimated through loss-on-ignition; LOI) and the chironomid fauna of the lakes can be observed in some training sets. Chapter 5 lists several studies where LOI has been reported as a strong explanatory variable for differences in relative chironomid abundances (e.g. Larocque et al. 2001; Nyman et al. 2005). These training sets are designed to cover a large temperature gradient, and as LOI is correlated with summer temperatures in subarctic regions, the LOI-range covered in the training sets is similarly high. Our results presented in Chapter 5 indicate that even when air temperature at all lakes incorporated in a training set is similar, LOI is still a strong explanatory variable, not only for explaining variability in relative chironomid abundances, but also for the concentration of chironomid remains and of the taxon richness of the individual lakes. However, the large differences in LOI were associated with only minor variations in chironomid-inferred temperature estimates.

138 Chapter 8

Again, these results show that, although we know that sediment composition can influence relative abundances of chironomid taxa (Pinder 1986), this factor did not seem to have a strong effect on chironomid-inferred temperatures from the studied floodplain lakes in Finnish Lapland. The trends in the down-core temperature- reconstructions as shown in the previous chapters all resembled the trends in the organic content of the sediments, but our results presented in Chapter 5 suggest that the changes in LOI probably did not affect the chironomid-inferred temperatures to a large extent.

Oxygen availability A third environmental parameter that might co-vary with temperature and which is important for the development and survival of chironomids is oxygen availability. Oxygen availability is a variable that has major consequences for the chironomid fauna of deep lakes and has a direct influence on the distribution of bottom-living chironomid larvae (Heinis and Davids 1993). Heiri and Lotter (2003) provide an example of a record where changing oxygen concentrations in the bottom waters of the lake (Sägistalsee, Switzerland) resulted in major changes in the chironomid fauna. The chironomid-based temperature record shows abrupt changes to cooler temperatures during the periods of low oxygen availability, which are most likely artefacts due to the influence of oxygen availability on the chironomid-based temperature reconstructions (Heiri and Lotter 2003; 2005). However, as oxygen availability in the hypolimnion is itself influenced by multiple factors such as lake morphometry, depth, temperature and productivity, it might be difficult to partial out the effects of these other factors from the direct effects of oxygen availability (Brodersen et al. 2004). Changes seen in the chironomid fauna of Oberwinkler Maar (Chapter 7) might have been influenced by changes in oxygen availability, and quantitative reconstruction of temperatures was hampered by the current scarcity of modern analogues where deep lakes with different oxygen conditions are incorporated in calibration data sets.

Water depth The depth of a lake is an important parameter in determining the lake’s chironomid fauna. It has a strong influence on the distribution and abundance of chironomid larvae (Brooks et al. 2007 and references therein), and many species have preferences for certain water depth ranges. Heiri et al. (2003b) show a case study where the decrease of lake depth (as a result of infilling of the lake basin) influenced the composition of the chironomid fauna of the lake, favouring Tanytarsus lugens-type. Furthermore, complete mixing of the water column will occur during the entire open water season in shallow lakes, whereas in deep lakes, thermal stratification might exist during the summer. Thermal stratification will not only decouple the relationship between air and water temperature, but can also lead to oxygen depletion in the hypolimnion, thus favouring taxa with certain oxy-regulating capabilities (Brodersen et al. 2004).

139 Chapter 8

Heiri et al. (2003c) studied the within-lake variability of chironomid-inferred temperature estimates, and show that chironomid-based inferences of air temperatures were lower for samples derived from the deepest parts of 5 shallow Norwegian lakes than for samples from the littoral zones of these lakes. Surprisingly, samples from intermediate water-depths produced the lowest inferred water- temperatures, and so far, no satisfactory explanation is available. However, the results of Heiri et al. (2003c) imply that a possible water-depth related bias might result from applying chironomid temperature inference models based on deep-water samples to sediment cores taken from shallower parts of a lake basin. Statistical analyses have shown that lake depth often is an important variable explaining midge distributions in modern training sets (Korhola et al. 2000; Barley et al. 2006; Walker 2006). Although statistical models to infer lake levels from fossil chironomid assemblages have been developed (Korhola et al. 2000; Barley et al. 2006), so far there have been no applications where these models were applied to infer past changes in lake level. Model errors imply that past lake level fluctuations must be large in order to reliably reconstruct lake level changes (Korhola et al. 2000; Walker 2006). A further problem in interpreting fossil chironomid assemblages is presented by the habitat selection of cold stenothermous chironomid taxa often encountered in arctic and subarctic shallow lakes. Most of these taxa are also capable of persevering in deep lakes in temperate regions, provided that oxygen-levels in the hypolimnion are sufficiently high. Chapter 7 shows an example of a record where transitions between different chironomid assemblages might be attributed to several different changes in the past environment, including changes in trophic state, temperature and lake depth. In this instance, the occurrence of a taxon that is only known from (sub-) arctic environments and is not known from deep lakes in temperate regions suggests that lake level might not have been the driving factor behind the changes in the chironomid assemblages. However, it seems worthwhile to study the effects of lake level and changes therein on modern chironomid distribution in more detail and on different spatial scales.

Concluding, for palaeolimnological studies it is important to note that increases in temperature, productivity or anoxia may result in similar changes in the chironomid fauna. When these factors behave similarly in the training set as they did in the lake ecosystem analysed for fossil chironomids, then palaeotemperature estimates based on the chironomid abundances will be reliable; however, when local (in-lake) processes change, chironomid-inferred climate reconstructions will potentially be affected (Brodersen et al. 2004). Heiri and Lotter (2005) show an example where they state that the high nutrient concentrations of lakes on the warm end of the temperature gradient in their training set are partly the result of anthropogenic influence, and that downcore reconstructions of nutrient concentrations should

140 Chapter 8 therefore be treated with caution. To assess whether in-lake processes might have influenced the chironomid fauna of a lake in the past, application of multiple proxies will enhance the reliability of a chironomid-based temperature reconstruction (e.g. Birks et al. 2000; Smol 2002; Heiri and Lotter 2005; Birks and Birks 2006). First, assessing temperature evolution by employing different proxies (e.g. using pollen- assemblages as well as chironomids), could point to periods where either of the inferences might be influenced by factors other than climate alone. In-lake processes could for instance affect chironomid-based palaeotemperature estimates, whereas no such reaction should be evident in pollen-inferred temperatures. Second, independent assessment of past changes in the lake system (such as changes in nutrient loading), might reveal periods of increased productivity in the lake, increased erosion in the lake’s catchment or changes in land-use, which all might result in oxygen deficiencies or other changes near the lake bottom.

8.2.3 Comparison of chironomid-based results and other proxy-based evidence of environmental change Even though temperatures inferred through the analysis of Coleoptera or botanical macro-remains indicate that the climate in northwestern and central Europe was suitable for the growth of forests during several interstadials in OIS-3, pollen- assemblages encountered in northeast Finland (Chapters 2 and 3) and eastern Germany (Chapters 4 and 6) indicate that a treeless tundra was present at the study- sites. We interpret this apparent contradictory situation as an example of vegetation being out of equilibrium with its environment. The refugia of certain thermophilic tree taxa were probably situated too far away from our study sites to reach the lake catchments in the available time of “improved” climate conditions. Furthermore, whereas the chironomid fauna of the thaw lake encountered in Reichwalde (presented in Chapter 4 and 6) is dynamic, and suggests a sharp decrease in temperatures in the uppermost sediments, the pollen-assemblages remain constant over the record. The analysis of macro-remains of botanical and zoological climate-indicator taxa provided independent evidence of minimum temperatures for northeast Finland and eastern Germany (Chapters 2-4 and 6). Plants require certain minimum mean summer temperatures to flower and reproduce and the relationship between the modern geographical limit of plant distribution and temperature can be used to reconstruct past minimum mean summer temperatures based on fossil records of fast- migrating aquatic plant remains (e.g. Kolstrup 1980). This is a different approach to temperature reconstruction than the method used to infer mean July air temperatures from chironomid remains. That latter method provides a mean temperature estimate with a Gaussian error distribution, whereas the former provides an absolute minimum value, and a detailed comparison of the resulting temperature inferences is therefore complicated. However, the analysis of botanical and zoological macro- remains not only offers a way to independently reconstruct minimum mean July air temperatures, but also provides information on past lake levels, nutrient availability, or for instance the presence of aquatic macrophytes, which may provide habitats for

141 Chapter 8

OIS Chrono GRIP δ18O Age Chironomid-based inferences stratigraphy Sokli Niederlausitz Eifel o Stadial Interstadial ka (oC) ( C) (CA-scores)

-43 -39 -35 ‰ SMOW

3 28 4 5 6 7 8 Denekamp 9 3 10 MIDDLE 11 (cold) (warm) 12 Hengelo 10 14 -1 0 1 2 13 Moershoofd 14 Glinde 15 16 Oerel 12 18

PLENIGLACIAL 17 59

WEICHSELIAN 18 4 19 EARLY 20 10 18 73 21 Odderade 22 5a-d EARLY 23 Brø rup 24 111

Figure 8.1: Quantitative palaeotemperature estimates for northeast Scandinavia (Sokli (67º48’N, 29º18’E); Chapters 2 and 3) and eastern Germany (Niederlausitz (51º60’N, 14º50’E); Chapters 4 and 6) and qualitative inferences of past climate change in western Germany (Eifel (50º09’N, 6º57’E); Chapter 7) compared to the δ18O record of the GRIP ice core (Johnsen et al., 1992), the oxygen isotope stages (Martinson et al., 1987) and the terrestrial Interstadials and Stadials (e.g. Behre and van der Plicht 1992; Dansgaard et al. 1993; Ran and van Huissteden, 1990), with the chronostratigraphy of parts of the Weichselian. Ages follow Martinson et al. (1987). Modified after table 4.1 different chironomid species (e.g. Birks 2000; Bos et al. 2001). It can thus complement chironomid-based studies and provide information which cannot be derived from chironomids alone.

8.3 Palaeoclimatologic reconstructions for OIS-3 In this study, different lacustrine deposits covering parts of OIS-3 were analysed, and the results provided new and exciting data on the climate variability during OIS-3. Although the age of the thaw lake deposit derived from Reichwalde (Chapters 4 and 6) is ambiguous, there is a clear climatic signal evident in the sedimentological and palaeoecological data. Before the formation of the thaw lake, permafrost was present in the Niederlausitz area. Only after permafrost degradation a thaw lake was formed, as evidenced by the horizontally laminated sediments of the lake deposit. Using botanical climate-indicator taxa as well as chironomid remains, high July air temperatures (~14 ºC) could be reconstructed for the earlier infilling of the lake. A

142 Chapter 8 sharp decrease in July air temperatures, with an amplitude of approximately 3 ºC, provides a minimal estimate of climate variability in central Europe during the Early Pleniglacial. The presence of frost fissures in the sandy sediments overlying the lacustrine deposits indicates a return to cold conditions with mean annual air temperatures below -1 °C (Chapter 4; Huijzer and Vandenberghe 1998). The lacustrine record of Oberwinkler Maar (Chapter 7) contains multiple transitions between organic-rich and clastic-dominated sediment intervals. The chironomid-assemblages encountered in this sediment record indicate that, although the former lake of Oberwinkler Maar was a relatively deep lake, the chironomid fauna consistently responded to changes in the climate at the site, and that the amplitude of this response was not constant. Climate changes during two of the stadial periods had a strong impact on the chironomid fauna, whereas in the two remaining stadials covered by the sequence only minor changes in the relative chironomid abundances were recorded. The Sokli record (Chapters 2 and 3) is assumed to correspond to Greenland Interstadial 14. Although there are no major oscillations in the reconstructed temperatures, it does provide exciting new data as there exist no similar studies providing quantitative temperature reconstructions for high-latitudinal continental sites in Europe for the earlier part of OIS-3. Together, these three records provide quantitative or qualitative data on different time intervals (see Figure 8.1) which can be compared to temperature reconstructions from other sites over northwest and central Europe.

For most of the study sites indicated in Figure 1.3, summer temperatures or temperatures-of-the-warmest-month were reconstructed for one or more interstadial periods during the earlier part of OIS-3. Most palaeotemperature estimates are sub- equal or equal to modern-day temperatures (e.g. Huijzer and Vandenberghe 1998; Coope 2002). The only site that does not show such high inferred summer temperatures is the Oerel site (Behre and Van der Plicht 1992; Behre et al. 2005). Behre et al. (2005) analysed fossil coleopteran assemblages and inferred summer temperatures of 11 ºC at the site, where the modern-day summer temperature is 16-18 ºC.

The pattern of summer temperatures reconstructed by these proxy records over the European continent is comparable with the pattern proposed in Chapter 3 where the LOVECLIM 3D climate model was applied to simulate average OIS-3 interstadial climatic conditions. The model results suggest high mean July air temperatures for the sector northeast of the Scandinavian ice sheet as well as for the southeast and southwest of the former ice sheet. For the region directly to the south of the former ice-sheet, palaeotemperature estimates are lower than those of the present-day situation, although the model exaggerates the cool climate conditions for this area as a result of a too large Scandinavian ice-sheet used as a boundary condition in the model runs.

143 Chapter 8

Care must be taken, however, because the palaeoclimatic data under discussion have been obtained through the analysis of different proxies that use different approaches to infer past summer temperatures, resulting in data that are difficult to compare directly (see above). Error estimates can also only be calculated using certain proxies (e.g. Coleoptera, Chironomidae) while for other proxies (e.g. botanical macro- remains) such estimates are not available. Furthermore, the chronologies of all the sites show major differences in both the applied dating methodology and in the number of dates available per site, and it is often not clear during which time interval (i.e. which Greenland Interstadial) the deposit is formed. Combined, this hampers a detailed comparison of the available palaeoclimatological data for the Weichselian in northwestern and central Europe. Furthermore, the climate model runs were conducted using a climate model of intermediate complexity, which is suitable to simulate climate patterns at a continental scale, but might be misleading when only interpreting the model-output for one or several grid cells of the model. The coarse spatial resolution of the model for instance prohibited the simulation of a reduced Scandinavian ice-sheet, and the overestimation of the size of the ice sheet during OIS-3 influenced the simulated pattern of both wind-direction and air temperatures over central Europe.

8.4 Outlook 8.4.1 Chironomids as a proxy The application of chironomids as a proxy for temperature change was initiated after the first publications by Walker and co-workers (Walker 1987; Walker and Matthewes 1987; Walker et al. 1991). Since then, the relationship between chironomids and summer temperature has been tested through the development of climate inference models and discussed extensively (e.g. Porinchu and MacDonald 2003; Velle et al. 2005; Walker and Cwynar 2006; Brooks 2006; Brooks et al. 2007). Section 8.2 illustrates that, although significant progress has been made over the last decade, there are still questions remaining concerning the direct and indirect effects of temperature on the development and survival of chironomids. Trying to disentangle the effects of different environmental variables such as changes in lake depth, productivity and sediment composition in a modern environment and on a local scale might lead to a better understanding of the modern chironomid distributions, and the application of more suitable training sets to fossil samples. The taxonomy of subfossil chironomid head capsules has greatly improved over the last decade, as illustrated by Brooks et al. (2007). The authors show that since the publication of Walker et al. (1991) the taxonomic classification of the tribe Tanytarsini has significantly improved, increasing from 3 taxa in Walker et al. (1991) to as many as 28 taxa in Brooks et al. (2007). As multiple species are often contained within a fossil chironomid-type or taxon, the relationship with temperature is assumed to be less strong than the relationship of individual chironomid species with temperature. In the Swiss Alps, different head capsule types of the genus Corynoneura show a distinctly different relationship with summer temperature, and therefore

144 Chapter 8 have the possibility of being useful palaeoecological indicators (Brooks et al. 2007). This illustrates that a further improvement in taxonomic resolution might produce better inference models (Brooks and Birks 2001). Promising results have been obtained by Korhola et al. (2002), using a climate inference model based on Bayes’ theorem. As this method involves a direct instead of an inverse approach to modelling the relationship between chironomids and temperature, it differs from the other, more commonly used approaches in quantitative palaeoenvironmental reconstruction (the so-called frequentist approaches; Vasko et al. 2000; Toivonen et al. 2001; Korhola et al. 2002). In a Finnish calibration dataset, the Bayesian model outperforms other methods as it yields the smallest prediction errors and bias and the largest coefficient of determination (r2) of the applied numerical methods (Korhola et al. 2002). The results of applying the Bayesian model to a Holocene lake record from northern Finland were consistent with those derived from the now standard method of weighted averaging partial least squares (WA-PLS) regression, but especially the standard error of the inferences is much lower. Although the methods developed by Korhola and co-workers have not yet been applied to more fossil records, it does indicate that there are still improvements to be made with respect to the numerical techniques employed in palaeolimnological research.

This study confirms the applicability of chironomids to time-intervals predating the Weichselian Late-Glacial, and shows that chironomid-based inferences of past climate change during OIS-3 provide reliable results. However, as discussed in sections 8.2.2 and 8.2.3, there are intercorrelations between e.g. hypolimnetic anoxia, trophic conditions, lake depth and water temperature in modern calibration datasets. As shown in Chapter 7, the interpretation of chironomids as the only available proxy analyzed in a sequence might be difficult because changes in a range of different environmental and climatic variables might result in similar changes in the chironomid fauna. The number of records covering early OIS-3 is limited, and most (former) lakes that continuously registered climate change over (parts of) OIS-3 were deep lakes, whereas modern calibration datasets are often composed of shallow lakes to lakes of intermediate depths. Chironomid-based inferences of past climatic and environmental change on samples derived from deep lakes are hampered by this non-analogue situation, and the study of modern deep lakes with stable thermal stratification of the water column might provide better analogues. Most fragmentary lacustrine records covering parts of OIS-3 are derived from shallow lake systems, and changes in lake level might be more important for the chironomid fauna of such shallow lakes than changes in e.g. oxygen availability. Although in a modern environment the maximum lake level is often determined through the morphometry of the lake catchment and readily determined, this is often not clear in older lake systems that are covered under cover sands or fluvio-aeolian sediments, as is the case for many lake deposits of OIS-3 age in northwest and central Europe, or even under glacial tills, as was the case in Sokli. In order to be able to

145 Chapter 8 determine possible fluctuations in lake levels, and the influence of such fluctuations on the chironomid fauna, independent proxy-evidence is needed.

8.4.2 Palaeoclimatology and dating of D/O climate variability To understand the causes of changing climate on various time scales it is not only necessary to obtain a clear picture of past variations in the local environment and climate conditions at certain locations in Europe, but both the timing and spatial expression of climatic changes must be known as well (e.g. Vandenberghe et al. 1998).

Comparing climate reconstructions from different sites with each other requires a firm grip on chronology. The most widely applied dating method in Quaternary geology is radiocarbon dating (Geyh 2001a). This method, however, does not (yet) offer the potential to calibrate samples older than ca 26 ka BP, as results obtained from several high-resolution radiocarbon calibration data sets show large offsets when compared to each other (Van der Plicht et al. 2004). Furthermore, as OIS-3 spans the time interval between 28 and 59 ka, sediments from the earlier parts of OIS-3 will be near to or even outside of the dating range of the radiocarbon method (~50,000 yr BP, Van der Plicht et al. 2004). An example of the care that has to be taken when interpreting radiocarbon dates that are near the limit of this range is not only given in Chapter 6, but also in Briant et al. (2005), who show a distinct difference in the ages derived from radiocarbon dating and the use of OSL dates, which they contribute to the vulnerability to contamination of Accelerator Mass Spectrometry (AMS) radiocarbon samples of material older than ca 35,000 years.

Luminescence-based dating techniques offer a way to date the last exposure of quartz or feldspar grains to sunlight or heat (Lian and Roberts 2006), and using optically stimulated luminescence (OSL) dating, sediments deposited during the full length of the last glacial cycle can be dated (Wallinga et al. in press), depending on the natural background radiation. The age estimates resulting from OSL dating usually have error bars that are approximately as large as 7-10% of the age of the sample (Murray and Olley 2002). The dates therefore may constrain the age of a sample to e.g. “the earlier parts of OIS-3”, but precise dating or detailed correlation of short-scale climatic oscillations between sites (e.g. with the Greenland ice-core records) is hampered because of the relatively large error bars. These errors might, at an age of 60,000 years, be as large as ± 6 ka, thus incorporating multiple D/O-events. A third method to date deposits older than 40 ka might be through the analysis of 230Th/U dating. The method of dating so-called ‘dirty samples’ (Geyh 2001b), up till now only applied to peat sediments, assumes that uranium dissolved in groundwater is entirely absorbed by fen peat. Where the outer 10 cm of peat thus continuously accumulate uranium from the groundwater, the inner layer is assumed to behave as a ‘closed system’, not having post-depositional enrichment of uranium, and therefore being suitable for 230Th/U dating (Hiller et al. 2004). Although this assumption might not apply to lacustrine deposits, Hiller et al. (2004) show an

146 Chapter 8 example of how 230Th/U dating in a sedimentary sequence provided a means to determine the chronology of a lacustrine deposit intercalated in this sequence. Although the 230Th/U dating method has been shown to perform very well in speleothem research and in other studies using calcites, the applications of 230Th/U dating to palaeoclimatological research using lacustrine deposits is still at its starting point, and needs to be tested thoroughly. A promising way forward for dating OIS-3 deposits seems the analysis of cryptotephra, microscopic shards of eruptive volcanic material that are spread over large areas and may thus potentially link different records, providing ‘hinge-points’ for more detailed correlations. All records shown in this thesis have been analysed for tephra, but, with the exception of the Eifel Maar lake records, no volcanic particles were recovered.

References Barley EM, Walker IR, Kurek J, Cwynar LC, Mathewes RW, Gajewski K, Finney BP (2006) A northwest North American training set: distribution of freshwater midges in relation to air temperature and lake depth. J Paleolimnol 36: 295-314 Bedford A, Jones RT, Lang B, Brooks SJ, Marshall JD (2004) A late-glacial chironomid record from Hawes Water, northwest England. J Quatern Sci 19: 281-290 Behre K-E, Van der Plicht J (1992) Towards an absolute chronology for the last glacial period in Europe: radiocarbon dates from Oerel, northern Germany. Veget Hist Archaeobot 1: 111-117 Behre K-E, Hölzer A, Lemdahl G (2005) Botanical macro-remains and insects from the Eemian and Weichselian site of Oerel (northwest Germany) and their evidence for the history of climate. Veget Hist Archaeobot 14: 31-53 Bigler C, Larocque I, Peglar SM, Birks HJB, Hall RI (2002) Quantitative multiproxy assessment of long-term patterns of Holocene environmental change from a small lake near Abisko, northern Sweden. The Holocene 12: 481-496 Birks HH (2000) Aquatic macrophyte vegetation development in Kråkenes Lake, western Norway, during the late- glacial and early-Holocene. J Paleolimnol 23: 7-19 Birks HH, Birks HJB (2006) Multi-proxy studies in palaeolimnology. Veg Hist Archaeobot 15: 235-251 Birks HH, Battarbee RW, Birks HJB (2000) The development of the aquatic ecosystem at Kråkenes Lake, western Norway, during the late-glacial and early-Holocene – a synthesis. J Paleolimnol 23: 91-114 Bos JAA, Bohncke SJP, Kasse C, Vandenberghe J (2001) Vegetation and Climate during the Weichselian Early Glacial and Pleniglacial in the Niederlausitz, eastern Germany - macrofossil and pollen evidence. J Quatern Sci 16: 269-289 Briant RM, Bateman MD, Coope GR, Gibbard PL (2005) Climatic control on Quaternary fluvial sedimentology of a Fenland Basin river, England. Sedimentology 52: 1397-1423 Brodersen KP, Anderson NJ (2002) Distribution of chironomids (Diptera) in low arctic West Greenland lakes: trophic conditions, temperature and environmental reconstruction. Freshw Biol 47: 1137-1157 Brodersen KP, Pedersen O, Lindegaard C, Hamburger K (2004) Chironomids (Diptera) and oxy-regulatory capacity: An experimental approachS to paleolimnological interpretation. Limnol Oceangr 49: 1549-1559 Brooks SJ (2000) Lateglacial fossil midge (Insecta: Diptera: Chironomidae) stratigraphies from the Swiss Alps. Palaeogeog Palaeoclimatol Palaeoecol 159: 261-279 Brooks SJ (2006) Fossil midges (Diptera: Chironomidae) as palaeoclimatic indicators for the Eurasian region. Quatern Sci Rev 25: 1894-1910 Brooks SJ, Birks HJB (2000a) Chironomid-inferred Lateglacial- early Holocene mean July air temperatures for Kråkenes Lake, western Norway. J Paleolimnol 23: 77-89 Brooks SJ, Birks HJB (2000b) Chironomid-inferred Late-glacial air temperatures at Whitrig Bog, south-east Scotland. J Quatern Sci 15: 759-764 Brooks SJ, Birks HJB (2001) Chironomid-inferred air temperatures from Lateglacial and Holocene sites in north- west Europe: progress and problems. Quatern Sci Rev 20: 1723-1741 Brooks SJ, Langdon PG, Heiri O (2007) The identification and use of Palaearctic Chironomidae larvae in palaeoecology. Quaternary Research Association Technical Guide 10, 276 pp Coope GR (2002) Changes in the Thermal Climate in Northwestern Europe during Marine Oxygen Isotope Stage 3, Estimated from Fossil Insect assemblages. Quatern Res 57: 401-408

147 Chapter 8

Dansgaard W, Johnsen SJ, Clausen HB, Dahl-Jensen D, Gundestrup NS, Hammer CU, Hvidberg CS, Steffensen JP, Sveinbjörndottir AE, Jouzel J, Bond G (1993). Evidence for general instability of past climate from a 250-kyr ice-core record. Nature 364: 218-220 Geyh MA (2001a) Symbiosis between geochronologists and Quaternary geoscientists. Geochronometria 20: 1-8 Geyh MA (2001b) Reflections on the 230Th/U dating of dirty material. Geochronometria 20: 9-14 Heinis F, Davids C (1993) Factors governing the spatial and temporal distribution of chironomid larvae in Maarseveen Lakes with special emphasis on the role of oxygen conditions. Aquat Ecol 27: 21-34 Heiri O, Lotter AF (2003) 9000 years of chironomid assemblage dynamics in an Alpine lake: long-term trends, sensitivity to disturbance and resilience of the fauna. J Paleolimnol 30: 273-289 Heiri O, Lotter AF (2005) Holocene and Lateglacial summer temperature reconstruction in the Swiss Alps based on fossil assemblages of aquatic organisms: a review. Boreas 34: 506-516 Heiri O, Millet L (2005) Reconstruction of Late Glacial summer temperatures from chironomid assemblages in Lac Lautrey (Jura, France). J Quatern Sci 20: 33–44 Heiri O, Lotter AF, Hausmann S, Kienast F (2003a) A chironomid-based Holocene summer air temperature reconstruction from the Swiss Alps. Holocene 13: 477-484 Heiri O, Wick L, van Leeuwen JFN, van der Knaap WO, Lotter AF (2003b) Holocene tree migration and the chironomid fauna of a small Swiss subalpine lake (Hinterbergsee, 1515 m asl). Palaeogeog Palaeoclimatol Palaeoecol 189: 35-53 Heiri O, Birks HJB, Brooks SJ, Velle G, Willassen E (2003c) Effects of within-lake variability of fossil assemblages on quantitative chironomid-inferred temperature reconstruction. Palaeogeog Palaeoclimatol Palaeoecol 199: 95-106 Heiri O, Cremer H, Engels S, Hoek W, Peeters W, Lotter AF (2007) Late-Glacial summer temperatures in the Northwest European lowlands: a new chironomid record from Hijkermeer, the Netherlands. Quaternary Science Reviews DOI 10.1016/j.quascirev.2007.06.017 Heiri O, Filippi ML, Lotter AF (in press) Lateglacial summer temperature in the Trentino region (Northern Italy) as reconstructed by fossil chironomid assemblages in Lago di Lavarone (1100 m asl). Studi Trentini di Scienze Naturali - Acta Geologica. Hiller A, Junge FW, Geyh MA, Krbetschek M, Kremenetski C (2004) Characterising and dating Weichselian organogenic sediments: a case study from the Lusatian ice marginal valley (Scheibe opencast mine, eastern Germany). Palaeogeog Palaeoclimatol Palaeoecol 205: 273-294 Huijzer B, Vandenberghe J (1998) Climatic reconstruction of the Weichselian Pleniglacial in northwestern and central Europe. J Quatern Sci 13: 391 – 417 Johnsen SJ, Clausen HB, Dansgaard W, Fuhrer K, Gundestrup N, Hammer CU, Iversen P, Jouzel J, Stauffer B, Steffensen JP (1992) Irregular glacial interstadials recorded in a new Greenland ice core. Nature 359: 311- 313 Kolstrup E (1980) Climate and stratigraphy in Northwestern Europe between 30,000 BP and 13,000 BP, with special reference to The Netherlands. Meded Rijks Geol Dienst 32: 181-253 Korhola A, Vasko K, Toivonen HTT, Olander H (2002) Holocene temperature changes in northern Fennoscandia reconstructed from chironomids using Bayesian modelling. Quatern Sci Rev 21: 1841-1860 Langdon PG, Barber KE, Lomas-Clarke SH (2004) Reconstructing climate and environmental change in northern England through chironomid and pollen analyses: evidence from Talkin Tarn, Cumbria. J Paleolimnol 32: 197-213 Larocque I, Hall RI, Grahn E (2001) Chironomids as indicators of climate change: a 100-lake training set from a subarctic region of northern Sweden (Lapland). J Paleolimnol 26: 307-322 Lian OB, Roberts RG (2006) Dating the Quaternary: progress in luminescence dating of sediments. Quatern Sci Rev 25: 2449-2468 Livingstone DM, Lotter AF (1998) The relationship between air and water temperatures in lakes of the Swiss Plateau: a case study with palaeolimnological implications. J Paleolimnol 19: 181-198 Marshall JD, Lang B, Crowley SF, Weedon GP, van Calsteren P,Fisher EH, Holme R, Holmes JA, Jones RT, Bedford A, Brooks SJ, Bloemendal J, KiriakoulakisK, Ball JD (2007) Terrestrial impact of abrupt changes in the North Atlantic thermohaline circulation: Early Holocene, UK. Geology 35: 639-642 Martinson DG Pisias NG, Hays JD, Imbrie J, Moore TC, Shackleton NJ (1987) Age dating and the orbital theory of the ice ages: development of a high-resolution 0 to 300,000-year chronostratigraphy. Quatern Res 27: 1-29 Murray AS, Olley JM (2002) Precision and accuracy in the optically stimulated luminescence dating of sedimentary quartz: a status review. Geochronometria 21: 1-17 Nyman M, Korhola A, Brooks SJ (2005) The distribution and diversity of Chironomidae (Insecta: Diptera) in western Finnish Lapland, with special emphasis on shallow lakes. Global Ecol Biogeogr 14: 137-153 Pinder LCV (1986) Biology of freshwater Chironomidae. Ann Rev Entomol 31: 1-23 Porinchu DF, MacDonald GM (2003) The use and application of freshwater midges (Chironomidae: insecta: diptera) in geographical research. Progr Phys Geogr 27: 378-422 Ran ETH, van Huissteden J (1990) The Dinkel Valley in the Middle Pleniglacial; dynamics of a tundra river system. Meded Rijks Geol Dienst 44: 209-220

148 Chapter 8

Seppä H, Nyman M, Korhola A, Weckström J (2002) Changes of treelines and alpine vegetation in relation to post- glacial climate dynamics in northern Fennoscandia based on pollen and chironomid records. J Quatern Sci 17: 287-301 Smol JP (2002) Pollution of Lakes and Rivers: A Paleoenvironmental Perspective. Arnold Publishers, London; Co- published by Oxford University Press, New York Toivonen HTT, Mannila H, Korhola A, Olander H (2001) Applying Bayesian statistics to organism-based environmental reconstruction. Ecol Appl 11: 618-630 Van der Plicht J, Beck JW, Bard E, Baillie MGL, Blackwell PG, Buck CE, Friedrich M, Guilderson TP, Hughen KA, Kromer B, McCormac FG, Bronk Ramsey C, Reimer PJ, Reimer RW, Remmele S, Richards DA, Southon JR, Stuiver M, Weyhenmeyer CE (2004) NotCal04 – Comparison/ calibration 14C records 26-50 cal kyr BP. Radiocarbon 46: 1225-1238 Vandenberghe J, Coope R, Kasse C (1998) Quantitative reconstructions of palaeoclimates during the last interglacial-glacial in western and central Europe: an introduction. J Quatern Sci 13: 361-366 Vasko K, Toivonen HTT, Korhola A (2000) A Bayesian multinomial Gaussian response model for organism- based environmental reconstruction. J Paleolimnol 24: 243-250 Velle G, Brooks SJ, Birks HJB, Willassen E (2005) Chironomids as a tool for inferring Holocene climate: an assessment based on six sites in southern Scandinavia. Quatern Sci Rev 24: 1429-1462 Wallinga J, Bos AJJ, Dorenbos P, Murray AS, Schokker J (in press) A test case for anomalous fading correction in IRSL dating. Quatern Geochronol Walker IR (1987) Chironomidae (Diptera) in paleoecology. Quatern Sci Rev 6: 29-40 Walker IR (2006) Chironomid overview. In: Ellas SA (ed) Encyclopedia of Quaternary Science volume 1. Elsevier, Amsterdam, pp 360-366 Walker IR, Cwynar LC (2006) Midges and palaeotemperature reconstruction – the North American experience. Quatern Sci Rev 25: 1911-1925 Walker IR, Matthewes RW (1987) Chironomidae (Diptera) and postglacial climate at Marion Lake, British Columbia, Canada. Quatern Res 27: 89-102 Walker IR, Smol JP, Engstrom DR, Birks HJB (1991) An assessment of Chironomidae as Quantitative Indicators of Past Climatic Change. Can J Fish Aquat Sci 48: 975-987

149 Chapter 8

150 Samenvatting

Samenvatting

De laatste jaren is de aandacht voor wereldwijde klimaatsverandering sterk toegenomen. De voorspellingen van toekomstige klimaatsveranderingen ten gevolge van de menselijke invloed op het klimaat zijn alarmerend, maar ook zonder de invloed van de mensheid is het klimaat op aarde altijd dynamisch geweest. De processen die deze natuurlijke veranderingen veroorzaakten zijn nog steeds actief. Om de natuurlijke en de menselijke invloeden op het klimaatsysteem te kunnen onderscheiden moeten de processen die aan het aardoppervlak spelen goed begrepen worden. Metingen van recente klimaatsveranderingen kunnen een beeld opleveren van deze processen. Helaas is de tijdspanne waarop we metingen hebben van belangrijke klimaatsparameters (zoals bv. temperatuur) slechts kort (op geologische tijdschaal), vaak niet verder terugreikend dan de 19e eeuw. Om de natuurlijke processen die op langere tijdschaal spelen beter te kunnen begrijpen, moeten we daarom gebruik maken van indirecte metingen van verschillende parameters van het klimaatsysteem: de zogenaamde proxy-indicatoren. Een recent ontwikkelde proxy is de analyse van fossiele resten van larven van chironomiden (veder- of dansmuggen). Deze resten kunnen worden aangetroffen in sedimenten die zijn achtergebleven op de bodem van poeltjes of meren, en door statistische bewerkingen kan een absolute schatting van juli-temperatuur afgeleid worden uit de verschillende chironomiden-resten die binnen 1 monster worden aangetroffen.

Chironomiden zijn een diverse groep insecten (Arthropoda: Insecta: Diptera: Chironomidae), wereldwijd bestaand uit meer dan 5000 soorten. De meeste mensen zullen chironomiden het beste kennen als de zwarte wolken insecten die op mooie dagen langs open water te vinden zijn. Chironomide-larven leven vaak in open water zoals meertjes of rivieren, en zijn vaak de meest voorkomende ongewervelden in deze omgevingen. Net zoals bij andere Diptera bestaat de levenscyclus van een chironomide uit 4 fasen: ei, larve, pop en imago (of volwassene).

151 Samenvatting

Chironomiden zijn een bruikbare proxy omdat ze vaak in hoge aantallen aanwezig zijn in meertjes. Vaak zijn er vele verschillende soorten chironomiden aanwezig in 1 enkel meer, en omdat delen van de larve bewaard blijven (en ook geïdentificeerd kunnen worden) kan de fossiele chironomiden-fauna van een meertje goed gereconstrueerd worden. Veranderingen in deze samenstelling hangen samen met veranderingen in de leefomgeving van de muggen, zoals milieu- of klimaatsveranderingen. Veel chironomiden-soorten zijn sterk gekoppeld aan bepaalde milieu- of klimaats-omstandigheden, waardoor ze goed bruikbaar zijn als indicator. Omdat de volwassen chironomiden kunnen vliegen, en de levenscyclus van chironomiden slechts kort is (vaak 1 of 2 generaties per jaar), zijn chironomiden in staat snel te reageren op veranderingen in klimaat en milieu. Tot slot heeft temperatuur een dominante invloed op alle vier verschillende levensstadia, en al deze factoren samen betekenen dat chironomiden gebruikt kunnen worden als een middel om temperatuur te reconstrueren. Gedurende de laatste 15 jaar is er een numerieke methode ontwikkeld waardoor absolute temperaturen afgeleid kunnen worden van fossiele chironomide- resten. Deze methode is vooral toegepast om de overgang tussen de laatste ijstijd en de huidige warme periode te analyseren. Dit proefschrift laat de eerste resultaten zien van de toepassing van deze methode op een iets oudere tijdschaal: verschillende periodes binnen de laatste ijstijd zijn geanalyseerd om de variabiliteit van het klimaat tijdens deze recente geologische periode te kwantificeren.

Het Kwartair is het geologische tijdvak dat de meest recente 2.6 miljoen jaar van de aardse geschiedenis beslaat. Het Kwartair wordt gekenmerkt door grootschalige afwisselingen tussen ijstijden (glacialen) en warme periodes (interglacialen). De overgangen tussen glacialen en interglacialen worden geïnterpreteerd als het gevolg van grootschalige veranderingen in de ontvangen instraling van de zon (dit fenomeen staat bekend onder de naam Milankovitch-cycliciteit). De laatste ijstijd (het Weichselien) duurde grofweg van 110 tot 11 duizend jaar geleden. Behalve het grootschalige Milankovitch signaal, worden in zowel sedimentkernen van de Atlantische oceaanbodem als in ijskernen van Groenland vele abrupte klimaatschommelingen herkend, de zogenaamde Dansgaard/Oeschger (D/ O)-oscillaties. Deze oscillaties worden gekenmerkt door een scherpe stijging van temperatuur in enkele decennia tijd, gevolgd door een periode van enkele honderden tot duizenden jaren waarin de temperatuur langzaam weer daalt, besloten met een scherpe overgang terug naar koude klimaatscondities. Op verschillende locaties in Europa zijn aanwijzingen gevonden dat D/O-klimaatsvariabiliteit ook invloed moet hebben gehad op het klimaat in Europa, maar er is een gebrek aan kwantitatieve informatie over veranderingen in bijvoorbeeld temperatuur en neerslag tijdens het Weichselien. Fossiele chironomide-resten aangetroffen in kernen (bestaande uit sediment dat zich verzamelde op de bodem van meertjes) bieden de mogelijkheid om absolute temperaturen te reconstrueren.

152 Samenvatting

In deze thesis worden de resultaten van de toepassing van chironomide-analyse op 3 verschillende kernen gepresenteerd. Deze meertjes werden alle gevormd tijdens het vroege en middelste gedeelte van de laatste ijstijd, een tijdspanne waarin de invloed van de mens op het klimaat nog nihil was. De geregistreerde veranderingen in de chironomide-fauna zijn daardoor dan ook een directe uiting van het natuurlijke klimaatsregime. Ook werd de invloed van overstromingen (door een rivier) op de chironomide-soorten die voorkomen in meertjes in noord Finland getest tijdens een veldwerk.

Hoofdstukken 2 en 3 van dit proefschrift laten de resultaten zien van klimaatsreconstructies (gebaseerd op chironomiden) voor noordoost Finland. De omgeving rond Sokli (noordoost Finland) wordt gekenmerkt door een atypische opbouw van de ondergrond. Het carbonaat-rijke gesteente nabij Sokli is relatief gemakkelijk verweerbaar, en hierdoor is er op deze locatie in de loop van de tijd een depressie ontstaan. In deze depressie heeft minimaal 3 keer een meertje gelegen gedurende de laatste ijstijd. Dit is zeer verrassend, aangezien er altijd van uitgegaan werd dat de Scandinavische ijskap tijdens het Weichelien over Sokli gelegen moest hebben. De sedimenten van een meer, gedateerd tot het vroege deel van het Midden- Weichselien, zijn geanalyseerd voor chironomide-resten. Het resulterende diagram dat getoond wordt in Hoofdstuk 2 laat zien dat er een zeer diverse chironomide- fauna in het meer geleefd heeft, en dat dit meer relatief ondiep en voedselrijk moet zijn geweest. Verschillende klimaatsveranderingen hebben zich voorgedaan tijdens het bestaan van het meer, en deze hebben de chironomide-fauna sterk beïnvloed. Eén van de meest opvallende resultaten is dat de gemiddelde juli temperaturen, afgeleid van de fossiele chironomide-resten, net zo hoog waren tijdens een gedeelte van het Weichselien als dat ze heden ten dage zijn. In Hoofdstuk 3 worden de op chironomiden gebaseerde temperatuursreconstructies vergeleken met andere proxy’s voor klimaat en milieu, waarbij opvalt dat het beeld dat door de chironomiden geschetst werd ondersteund wordt door de aanwezigheid van verschillende andere indicatoren van hoge zomertemperaturen, zoals bijvoorbeeld de aanwezigheid van overblijfselen van mosdiertjes zoals Fredericella indica. Ook worden in dit hoofdstuk simulaties gepresenteerd die gemaakt zijn met behulp van een klimaatmodel. Dit model simuleert (onder andere) de atmosferische circulatie tijdens een warm interval in het Midden Weichselien. Uit deze modelresultaten valt af te leiden dat de aanwezigheid van de Scandinavische ijskap het circulatiepatroon over noord Europa sterk beïnvloed moet hebben, resulterend in de aanvoer van zeer droge lucht naar noordoost Finland. Samen met de hoge zonne- instraling levert dit een mogelijke verklaring voor de hoge geïnterfereerde juli temperaturen voor Sokli.

Tijdens een veldwerk in 1999 werd een meerkerntje aangetroffen in de bruinkoolmijn nabij Reichwalde (oost Duitsland), en op basis van correlatie met een sedimentrecord

153 Samenvatting uit een nabije mijn werd aangenomen dat deze kern ontstaan moest zijn tijdens het vroege deel van het Midden Weichselien (Hoofdstuk 4). Verschillende proxy’s werden gebruikt om de omgeving van het meertje te reconstrueren, en om ook iets te kunnen zeggen over het klimaat dat toen geheerst moet hebben. Gebruik makende van stuifmeelkorrels en fossiele resten van planten en dieren die zichtbaar zijn met het blote oog (zoals bv. zaden en bladeren) werd afgeleid dat het vroegere meertje op een overstromingsvlakte van een rivier gelegen moet hebben, en dat de vegetatie bestond uit een struiktoendra. Vervormingen in de sedimenten die werden aangetroffen onder de parallel gelamineerde meersedimenten duiden op koude klimaatscondities voordat het meertje gevormd werd. Tijdens de invulling van het meerbekken waren de gemiddelde juli temperaturen waarschijnlijk hoog, waarbij plantenresten een minimum temperatuur van 12-14 ºC suggereren. In het bovenste gedeelte van de kern is een scherpe afname van temperatuur te zien (gereconstrueerd met behulp van een semi-kwantitatieve methode gebaseerd op de chironomiden-resten die werden aangetroffen in deze kern). Deze afname in temperatuur valt samen met een sterke afname van het organische gehalte van het sediment, en met een terugkeer van permafrost condities. Dit alles samen suggereert een overgang naar koudere klimaatscondities tijdens het bovenste gedeelte van de kern.

Meertjes die gelegen zijn op een overstromingsvlakte van een rivier (zoals het voormalige meer dat gepresenteerd werd in Hoofdstuk 4) worden sterk beïnvloed door de regelmatige overstromingen die in een natuurlijke omgeving voorkomen. Deze overstromingen kunnen een sterke invloed hebben op de chironomide-fauna die in deze meertjes voorkomen, bijvoorbeeld als het gevolg van veranderingen in de aanwezigheid van voedsel of nutriënten, door beïnvloeding van de troebelheid van het water, of door de introductie van roofdieren. Hierdoor worden meertjes die gelegen zijn in een overstromingsvlakte niet gebruikt in de moderne training sets (dit zijn de groepen meertjes die gebruikt worden om de huidige relatie tussen chironomiden en hun omgeving te onderzoeken). Om te onderzoeken hoe goed onze fossiele chironomide-monsters (zoals bv. aangetroffen in oost Duitsland) en de moderne analogen (monsters genomen van de meertjes uit de training sets) met elkaar overeenkomen, hebben we 33 meertjes in noord Finland bemonsterd. Twintig van deze meertjes waren geïsoleerd van de invloed van een rivier, terwijl de andere 13 jaarlijks overstromen. Er werden slechts kleine verschillen aangetroffen in de fysische en chemische samenstelling van het meer en in de waterkwaliteit tussen de 2 verschillende groepen meertjes (jaarlijks overstroomde meertjes versus geïsoleerde meertjes), ondanks het grote verschil in de omgeving (bijvoorbeeld vegetatie) rond de verschillende meren. Wel werden er grote verschillen gevonden in de samenstelling van de chironomide-fauna tussen de verschillende groepen meren. Deze verschillen waren duidelijk in de concentratie van chironomiden-resten, het aantal soorten

154 Samenvatting chironomiden per meertje en ook in het relatieve voorkomen van de individuele chironomiden-soorten. De verschillen in de samenstelling van de chironomide-fauna per meertje blijkt niet te leiden tot grote verschillen in gereconstrueerde temperatuur. Dit betekent dat voormalige meertjes, die op een overstromingsvlakte gelegen moeten hebben, toch gebruikt kunnen worden voor temperatuursreconstructie, ondanks dat de moderne meertjes die gebruikt worden in de training sets niet gelegen zijn op een overstromingsvlakte.

In Hoofdstuk 6 word de sedimentatiegeschiedenis in de bruinkoolmijnen nabij Reichwalde en Nochten (beide oost Duitsland) besproken. Gebruik makende van de zogenaamde “optically stimulated luminescence” (OSL) en koolstof-14 dateringsmethoden wordt een onafhankelijke tijdschaal gepresenteerd voor de afzetting van de zanden en grinden die in de mijnen zijn aangetroffen. Deze tijdschaal wordt vergeleken met al bekende resultaten, en een verschil tussen onze resultaten en die uit eerdere onderzoeken wordt besproken Twee korte sedimentkerntjes, die gedateerd zijn tot het vroege en het midden van het Weichselien zijn onderzocht op chironomide-resten. De resulterende chironomide-diagrammen laten zien dat beide voormalige meertjes op een overstromingsvlakte gelegen moeten hebben, en dat ze waarschijnlijk redelijk to zeer voedselrijk waren. Kwantitatieve temperatuursreconstructies werden gemaakt met behulp van een training set uit de Zwitserse Alpen, en worden vergeleken met lokale en regionale klimaatsreconstructies.

Er zijn slechts enkele meerkernen bekend in Europa die over een langere periode (~10 duizend jaar of meer) bestaan hebben tijdens de laatste ijstijd. Eén van deze meerkernen is bekend uit de Eifel (Duitsland), en laat tenminste 4 abrupte klimaatschommelingen tussen koude periodes en warme periodes zien (Hoofdstuk 7). De chironomiden laten zien dat tijdens de koudere periodes, het meer diep en voedselarm was. Er waren veel soorten aanwezig die typisch zijn voor (extreem) koude klimaatsomstandigheden. Tijdens de warmere intervallen waren er verschillende soorten aanwezig die kunnen duiden op warmere klimaatsomstandigheden.

Deze thesis laat de resultaten zien van de eerste toepassing van chironomiden als een proxy voor temperatuur op een tijdschaal ouder dan 15 duizend jaar. Onze resultaten tonen aan dat ook op deze langere tijdschaal chironomiden een goede indicator zijn voor de omgeving van voormalige meertjes, en dat het afleiden van absolute juli temperaturen betrouwbare resultaten oplevert. Onze resultaten tonen aan dat het waarschijnlijk is dat verschillende abrupte klimaatschommelingen, zoals te zien in sedimentkernen van de oceaanbodem en in ijskernen uit Groenland, ook hun invloed hebben gehad op het Europese continent.

155 156 Dankwoord

Dankwoord / Acknowledgements

Vier jaar en 11.867 dode muggen verder zet ook ik me aan het schrijven van het meest-gelezen gedeelte van een proefschrift, het dankwoord. Net als velen die mij voorgegaan zijn, had ik in mijn eentje deze klus niet kunnen klaren, en dankzij de hulp van vele anderen is dit werk geworden tot het boekje dat nu voor je ligt.

Ten eerste wil ik graag Sjoerd bedanken, die mij überhaupt de kans gegeven heeft om met dit werk te beginnen. In de afgelopen 4 jaar heb ik veel van je geleerd Sjoerd, niet alleen op het wetenschappelijke gebied, maar ook op het persoonlijke vlak. Jij bezat vaak net de eigenschappen die ik miste, zoals de omgang met boze Finnen of de kwaliteit om op het juiste moment een stapje terug te doen. Ook Oliver wil ik heel graag bedanken. Je bent een beetje via een omweg in dit project verzeild geraakt, maar bedankt dat je er altijd was als ik weer een soort niet kende, een vraag over statistiek had, of op het laatste moment nog een kritische blik nodig had voor een abstract of manuscript. Jef: ik kan me voorstellen dat dit project in het begin een wat ver-van-mijn-bed show voor je was. Maar zeker tijdens de veldwerken en richting het einde van het project heb ik veel aan je expertise en brede overzicht gehad. Sjoerd, Oliver en Jef, ik ben blij dat ik met jullie samen heb mogen werken gedurende de laatste 4 jaar.

I would like to thank the members of the reading committee for their effort of reading this thesis and their kind feedback: Hanneke Bos, Steve Brooks, Karin Helmens, Kees Kasse, Andy Lotter, Bas van Geel.

Furthermore, I’d like to thank all my co-authors and colleagues for their help and enthusiasm. I’d especially like to thank Frank Sirocko and Katja for the opportunity to work with their unique sediments, and for the hospitality during my visits to Mainz; and Karin Helmens, for inviting me to work with the Sokli-sediments and for believing in me enough to offer me a postdoc contract, providing me with the opportunity to continue doing what I like best. The members of the RESOLuTION- program are all thanked for the stimulating discussions and feedback during our annual meetings.

157 Dankwoord

Ik had lang niet al mijn resultaten kunnen verkrijgen zonder de inzet van Diane, Gemma, Lukas en Roel in het veld, en ook de hulp van Lesley als vrijwilligster mag hierbij niet onvermeld blijven! Martin, Maurice, Tineke, Martine en Roel van het sediment-lab stonden altijd voor mij klaar wanneer er last minute weer eens iets moest gebeuren, of wanneer iets niet lukte. Hetzelfde geldt voor Bouk, Frans en Wynanda van het gesteente-lab, in de periode dat er zo’n 1000 monsters gevriesdroogd mochten worden. Geert-jan, Jos en (vooral) Ron, bedankt voor de software hulp, met name tijdens het maken van dit boekwerk.

Ik heb me de afgelopen jaren niet exclusief bezig gehouden met mijn eigen werk, maar ook mogen assisteren bij het begeleiden van veldwerken, excursies en practica, waarbij ik af en toe het idee had dat ik degene was die nog het meeste leerde. Kay, Vincent en de Brabant-crew, bedankt voor de 3 gezellige jaren - en onthoud: het is hier geen hotel! Ronald (en Freek), ik kijk met veel plezier terug op de Excursies Nederland waaraan ik mee mocht doen, en wat voor mij soms echt een eye-opener is geweest met betrekking tot tijd/ruimte- schalen in het ontstaan van het Nederlandse landschap. Ook wil ik graag Wim Hoek en Andy Lotter bedanken voor de begeleiding aan het einde van mijn (doctoraal) studie tijdens de eerste stapjes op de ‘weg tot de wetenschap’, en natuurlijk ook voor de welgemeende adviezen, suggesties en hulp tijdens latere momenten, tot vandaag aan toe.

Mark, jij was mijn eerste “echte” kamergenoot op de VU. Hoewel onze muzieksmaak in het begin nogal ver van elkaar af lag, zijn we (waarschijnlijk tot grote spijt van Hanneke) elkaar mooi halverwege tegengekomen bij de Nederlandstalige meezingers, waarmee we vele middagen opgesierd hebben. Hanneke, jij zult af en toe wat te stellen hebben gehad met mij, maar ik ben erg blij dat ik je als kamergenote heb mogen ervaren. Als ik weer eens een zaadje of blaadje tegenkwam was je altijd enthousiast en behulpzaam, en het samen werken aan verschillende projecten is zeer stimulerend geweest. Ik ga jullie missen, hopelijk drinken we snel samen weer een biertje (1-tje maar, hè ;))

Dank voor de vele discussies aan de koffietafel, de verhalen tijdens de lunch, en/of de ontspanning op de donderdagmiddag-borrel: Aafke, Alex, Anco, Ane, Bram, Bert, Cedric, Dick, Didier, Edith, Els, Emma, Frank, Freek, Geert-Jan, GeoVUsie, Gerald, Glenda, Hanneke, Hans, Hubert, Jef, Jens, Jochem, Jop, Jos, José, Karen, Kay, Kees, Klaas, Lia, Maarten, Margot, Mark, Martin, Mascha, Mirjam, Orson, Paul, Philip, Ron, Ronald, Sjoerd, Simon J, Simon T & Suzanne. Ook de studenten die met me een biertje gedronken hebben bedank ik uiteraard voor de gezelligheid, met name Ananda, Ingeborg, Jos W, Koen en Marijke, en natuurlijk ook mijn beide paranimfen: Charlotte & Karlien. Als ik iemand vergeet (en die kans acht ik vrij groot), vul hier je naam in: …………….. Bedankt!

158 Dankwoord

Zelfs als AIO kom je zo heel af en toe nog aan je ontspanning toe: Alette, Daniel, Denise, Derrick, Ivo, Dion, Jeroen, Nadine, Onno, Thijs, en ook Bef, David & AsIce, Jasper, Karin B, Marjan, Nita, Sjoerd en Tim, bedankt voor de afleiding, de fun etc. Rivierwijkers 3 (ik doe alle 31 ex- en huidige leden maar in 1 keer) en alles wat daar omheen hangt, jullie hebben een belangrijke rol gehad in mijn leven de laatste jaren, en behalve de dins- en zaterdagen ga ik zeker de weekendjes, toernooien en feesten nooit meer vergeten (of misschien juist wel, haha).

Tot slot wil ik Manon, Remko, Karin & Roy en natuurlijk mijn ouders bedanken voor alle steun en interesse over de afgelopen jaren. Pap, mam, zonder jullie vertrouwen en geloof in mij had ik dit nooit kunnen bereiken.

159 160