Inferring Past Environmental Changes in High Arctic Lake Ecosystems Using Ancient DNA

Biology Department Research Group Protistology and Aquatic Ecology ______

Inferring past environmental changes in High Arctic lake ecosystems using ancient DNA

Laura Decorte Studentnumber: 01402099

Supervisors: Prof. Dr. Elie Verleyen, Prof. Dr. Wim Vyverman, Dr. Bjorn Tytgat Scientific tutor: Lotte De Maeyer

Master’s dissertation submitted to obtain the degree of Master of Science in Biology

Academic year: 2019 - 2020

© Faculty of Sciences – research group Protistology and Aquatic Ecology All rights reserved. This thesis contains confidential information and confidential research results that are property to the UGent. The contents of this master thesis may under no circumstances be made public, nor complete or partial, without the explicit and preceding permission of the UGent representative, i.e. the supervisor. The thesis may under no circumstances be copied or duplicated in any form, unless permission granted in written form. Any violation of the confidential nature of this thesis may impose irreparable damage to the UGent. In case of a dispute that may arise within the context of this declaration, the Judicial Court of Gent only is competent to be notified.

Content 1 Introduction ...... 2 2 Objectives ...... 4 3 Material and methods ...... 5 3.1 Description of study site ...... 5 3.2 Sample collection ...... 7 3.2.1 Samples of present-day community ...... 7 3.2.2 Lake sediment cores ...... 7 3.3 Analysis of present-day lake and surface sediment DNA samples ...... 7 3.3.1 Bioinformatic analyses, taxonomical assignment and downstream analysis ...... 7 3.4 Core analysis ...... 8 3.4.1 Radiocarbon dating ...... 8 3.4.2 Fossil pigment analysis ...... 8 3.4.3 Ancient DNA analysis ...... 8 4 Results ...... 12 4.1 Present-day lake and surface sediment samples ...... 12 4.1.1 Alpha diversity ...... 12 4.1.2 Taxonomic composition of the samples ...... 12 4.2 Lake sediment core ...... 14 4.2.1 Chronology ...... 14 4.2.2 Core description ...... 15 4.2.3 Fossil pigments ...... 15 4.2.4 Ancient DNA ...... 17 5 Discussion ...... 22 5.1 Representation of aquatic and catchment community in surface lake sediments ...... 22 5.2 Protocol for aDNA processing ...... 23 5.3 Lake evolution and environmental reconstruction ...... 23 5.3.1 Deglaciation and colonization after the Last Glacial Maximum ...... 23 5.3.2 Late Glacial and early-mid Holocene ...... 24 5.4 Perspectives ...... 25 6 Conclusion ...... 26 7 Summary...... 27 7.1 English summary ...... 27 7.2 Nederlandse samenvatting ...... 29 8 Acknowledgments ...... 32 9 Reference list ...... 33 10 Appendix ...... 41

1 Introduction The Arctic is the most sensitive region in terms of climate change, with warming being two to three times higher than the global annual average (IPCC 2018). Since the end of the Little Ice Age (LIA) in the late nineteenth century, most Arctic glaciers have been retreating (Dowdeswell et al. 1997). Glacier retreat exposes new terrestrial habitats that can be colonized (Pessi et al. 2019), which promotes the greening of the Arctic. This happens in a few successional steps where cryptogamic covers (plants that reproduce by spores) play an important role (Screen & Simmonds 2010, Kern et al. 2019). The first colonizers of deglaciated, exposed soils generally are biological soil crusts (BSCs), consisting of algae, cyanobacteria, fungi, lichens and mosses (Yoshitake et al. 2018). They act as ecosystem engineers as they contribute to soil stabilisation, water content of the soil and nutrient cycling (Pointing & Belnap 2012). This way, BSCs are very important for further colonization by higher plants. They have a direct positive influence on vegetation density, species richness and cover and as such contribute to the greening of the Arctic (Breen & Lévesque 2006). The dominant primary producers in High Arctic tundra biomes are dwarf shrubs, forbs, graminoids and cryptogamic covers (Kern et al. 2019). The tundra biome is impacted by climate warming in such a way that it is showing an increased shrub dominance (Myers-Smith et al. 2015). Pearson et al. (2013) also predict a widespread redistribution of Arctic vegetation. They claim that in general, low-lying vegetation with sparse plant cover will decrease, while larger shrubs and trees will increase their range.

Climate warming also induces changes in Arctic freshwater ecosystems through lengthening of the summer growing season. More in particular, increased spring and autumn temperatures are resulting in a decrease in the duration of lake ice cover. Moreover in deep lakes, warming temperatures and less ice cover may induce longer periods of thermal stratification. These changes in ice phenology (timing of ice break-up and freeze-up) and stratification lead to shifts in the life strategy of diatoms, generally marked by increasing abundances of small-celled cyclotelloid taxa and decreasing abundances of heavier diatoms like Aulacoseira taxa (Rühland et al. 2015). Furthermore, the decreased duration of lake ice cover may lead to increased development of mosses in the littoral zone, which in turn favours epiphytic diatom taxa associated with mossy substrates (Rühland et al. 2015). Interestingly, lake sediments, that accumulate over time, contain plenty of information about the response of the biological communities in the lake and their surrounding catchment to environmental changes (Smol & Douglas 2007). As sediments accumulate continuously, often anoxic conditions are created, which offers excellent preservation conditions (Parducci et al. 2017). An Arctic wide evaluation of diatoms in lake sediments has revealed significant and widespread changes in diatom community structure and ecological reorganisations in many High Arctic lakes within the last ~150 years (Smol et al. 2005).

In addition to diatoms, many studies on sediment cores in the Arctic up to now were focussed on pollen and macrofossils to assess the effect of climate changes on the vegetation in their catchments (e.g. Peros & Gajewski 2009, Zhang et al. 2018). Pollen data, however, give a relatively low taxonomic resolution (Sønstebø et al. 2010). Moreover, local pollen production in a treeless area like the Arctic is usually low and pollen that are wind-transported over long distances could dominate the pollen assemblages (Birks 2008, Birks & Birks 2000). This can lead to paleoenvironmental reconstructions being biased by the influx of exotic pollen (Crump et al. 2019). Plant macrofossils are generally deposited close to their origin and therefore better represent the local plant community. However, the representation is very species specific, as not all species get deposited or preserved as macrofossils

(Birks 2008). Moreover, macrofossil analysis requires expertise in botanical taxonomy, morphology and vegetation ecology and is therefore not straightforward (Birks & Birks 2000).

A more recent and very promising approach to reconstruct past environmental changes is the use of ancient DNA (aDNA) that accumulated in sediments (Willerslev et al. 2003, 2014). High-throughput sequencing techniques have drastically facilitated the development of DNA-based research in paleoecological studies during the last decade (Parducci et al. 2017). Sedimentary aDNA has proven to be efficient for reconstructing past vegetation and species diversity in the Arctic (Alsos et al. 2016, Parducci et al. 2017, Sønstebø et al. 2010). Although rare taxa are not always detected and DNA of some species rarely reaches the lake sediment (Alsos et al. 2018), aDNA analyses offer new perspectives for studying past vegetation changes. More species can be identified and at a higher taxonomic resolution than generally obtained using pollen or macrofossil analyses. aDNA in lake sediments also largely reflects the local flora, in contrast with regionally dispersed pollen (Parducci et al. 2017). Although aDNA is a powerful tool, it also has its drawbacks. DNA suffers from degradation and is thus often highly fragmented and subject to a variety of chemical modifications (Shapiro et al. 2012). Preservation conditions, however, play an important role. When the DNA is preserved in cold environments, like the Arctic, some of these damages are reduced (Willerslev et al. 2004). Another aspect that makes working with aDNA challenging, is the risk of contamination. Even a very low amount of contamination can have an effect, as modern DNA will show preferential amplification in a PCR reaction over damaged aDNA. Contamination can happen at multiple stages during the processing of the aDNA, from core retrieval in the field up to core processing in the laboratory. It is therefore essential to follow a detailed protocol which tackles all procedures and equipment that is needed during core processing in order to avoid contamination (Willerslev & Cooper 2005).

Another useful tool in paleolimnological reconstructions is the analysis of fossil pigments. Fossil pigment analysis is a good addition to aDNA analysis (Zhu et al. 2005) and can be useful to gain a deep understanding of changes in photoautotrophic communities in the lake and its catchment. This is because pigments reach the lake sediments both from autochthonous planktonic and littoral organisms, as well as from allochthonous plant matter from the catchment area (Guilizzoni et al. 2002). Fossil pigment compositions can tell a lot about previous changes in lake primary productivity (Swain 1985, Tavernier et al. 2014). Additionally, phototrophic phytoplankton communities can be identified based on fossil marker pigments. Chlorophyll derivatives and total carotenoids are indicators for algal abundance, while individual carotenoids are signatures for specific algal taxa (Jiang et al. 2011).

Svalbard is a very suitable region for paleolimnological studies. First of all, it has a wide range of lakes throughout the archipelago. These High Arctic lakes are believed to be one of the most sensitive ecosystems to climate change, due to a brief growing season that causes a low annual production (Birks et al. 2004a). Despite the presence of local mining industries in Svalbard and coal-fired power stations in the Isfjord area, levels of contamination in sediments are low and mainly observable very locally in Isfjord (Rose et al. 2004). Svalbard is also physically isolated from large industrial regions, although some long-range pollution has been reported (Birks et al. 2004a). Furthermore, it is one of the most accessible regions in the High Arctic (Birks et al. 2004a). Knowledge about the glacial history in Svalbard is essential for understanding paleolimnological changes. Previous studies suggest that the start of deglaciation over the western Svalbard shelf after the Last Glacial Maximum (LGM) occurred around 19,500-20,500 calibrated (cal) years BP (Husum et al. 2019, Rasmussen et al. 2007), which is earlier than previously thought (Jessen et al. 2010). The north-western part of the fjord mouth of 3

Kongsfjorden was deglaciated before 14,400 cal years BP (Husum et al. 2019, Jessen et al. 2010). Several glacier re-advancing events occurred across Svalbard during the Late Glacial and early Holocene (van der Bilt et al. 2015, Farnsworth et al. 2018). However, early-mid Holocene (9,000 – 5,000 years ago) climate is believed to have been warmer than today, which is already supported by several studies (Alsos et al. 2016, van der Bilt et al. 2019, Farnsworth et al. 2018, Voldstad et al. 2020). These glacier re-advances could have been controlled by glacier dynamics (surges) rather than climate (Farnsworth et al. 2018). Also during the LIA, glaciers in Svalbard re-advanced (Farnsworth et al. 2018, Larsen et al. 2018).

However, knowledge about the Late Glacial and Holocene climate and deglaciation history in Svalbard is sparse (Larsen et al. 2018) and long-term records of past environmental conditions using proxy indicators in lake ecosystems in relation to past climate changes in the Arctic are generally lacking (Jiang et al. 2011). This information is highly needed and also essential to better predict future environmental changes and the responses of biological communities.

2 Objectives In this thesis, a first objective was to investigate to what extent contemporary planktonic communities in a lake and the vegetation in its catchment are reflected in the sedimentary record. In order to achieve this, DNA found in littoral and pelagic samples was compared with DNA in sediment samples to assess the degree of overlap in species composition between the samples.

The second objective was to develop a protocol for aDNA analysis with a focus on how to avoid contamination during sample preparation.

The third objective was to study the evolution of lake Sarsvatnet in north-western Svalbard, and to reconstruct past changes in lake primary production and the eukaryotic community structure in and around the lake. In order to achieve this, aDNA and fossil pigments were analysed in two lake sediment cores.

3 Material and methods 3.1 Description of study site The study site is located near Ny-Ålesund (78°55′30″N 11°55′20″E), a former mining town on the southern shore of Kongsfjord (King’s bay) which has permanent research institutes from different countries. The town is situated on the Brøgger Peninsula in western Svalbard on the largest island of the archipelago, Spitsbergen, about 80 km north of the capital Longyearbyen (Fig. 1a). Western Svalbard has a milder climate than expected based on its northern location. The reason for this is the Norwegian current which flows northward along the western coast and transports warm Atlantic water masses into the Arctic Ocean, which results in an arctic-oceanic climate (Birks et al. 2004a, Kern et al. 2019). According to a weather station established close to Ny-Ålesund by the Norwegian Meteorological Institute (www.met.no), mean summer and winter temperature over the last two decades was 8 °C and -14 °C, respectively. Longer cold periods between -20 and -35 °C can occur during winter. Mean annual precipitation over the same period was 470 mm.

The cores and samples were taken in and around lake Sarsvatnet (78°57’02.6″N 12°29’54.6″E) in Ossian-Sarsfjellet, a nature reserve northeast of Ny-Ålesund (Fig. 1b). The deep lake is surrounded by steep cliffs on the northern and eastern side of the shore. There are several fresh glacial moraine ridges on the southern side of the lake, which is less steep (Birks et al. 2004b). The Kongsbreen glacier is situated east of lake Sarsvatnet and at the present day, glacial meltwater flows north and south of the lake. Whenever the glacier is large enough and reaches a certain threshold, glacial derived sediments can reach the lake (Støren et al. 2020). The vegetation in the catchment of the lake is rich and different compared to other lake catchments along the west coast of Svalbard. This is probably a result of the inner-fjord setting of the lake and hence warmer growing-season temperatures (Birks et al. 2004b). The vegetation belongs to the middle arctic-tundra zone (MATZ) and includes species such as Carex rupestris, Carex nardina, Cassiope tetragona and Dryas octopetala (Birks et al. 2004b, Kern et al. 2019).

(b) (a)

(c)

Figure 1 (a) Map of Svalbard, (b) a more detailed areal picture of the area around lake Sarsvatnet (adjusted from Støren et al. 2020) and (c) bathymetric map of lake Sarsvatnet (adjusted from Støren et al. 2020). Circles indicate location of the sampling campaign in 2017 for the recent samples (red: littoral, white: pelagic and surface sediment), star indicates location of the sampling campaign in 2018 for the two sediment cores.

3.2 Sample collection 3.2.1 Samples of present-day community Samples needed for analysis of the present-day community were taken in July/August 2017 in lake Sarsvatnet (Fig. 1c). Three samples were taken in littoral zones of the lake (OSLAC17A, OSLAC17B and OSLAC17C) and three sample were taken in the pelagic zone (OSPEL17A, OSPELDEEPA and OSPELDEEPB) using a filter. In addition, three cores were taken using a UWITEC corer, and subsequently the upper 0-0.5 cm (SA17G10.0_5, SA17G20.0_5 and SA17G30.0_5) and 0.5-1 cm (SA17G10.5.1, SA17G20.5.1 and SA17G30.5.1) from each core were subsampled and used for further analysis. Samples were stored immediately in liquid nitrogen in the field and at -80 °C in the laboratory until further processing.

3.2.2 Lake sediment cores Two cores were retrieved in the summer of 2018 (core 1: SV18L1_S1_LV1 and core 2: SV18L1_S1_LV2) using a Livingston piston corer from the centre of lake Sarsvatnet approximately at its deepest point. The sample names that were given afterwards (C1_01 up to C1_21 and C2_01 up to C2_37) referred to the core (C1 or C2) and the depth of the subsample within the core (e.g. 01 referred to 0-1 cm). The cores were frozen while they were transported to the laboratory in Ghent, where they were stored at -80 °C. 3.3 Analysis of present-day lake and surface sediment DNA samples 3.3.1 Bioinformatic analyses, taxonomical assignment and downstream analysis Quality filtering and clustering into Operational Taxonomic Units (OTUs) was performed using a customized bioinformatics pipeline (Tytgat et al. 2019). Overall raw sequence data quality was checked using FastQC reports. Forward and reverse reads were paired using PEAR (Zhang et al. 2014) with a minimum merge length of 50 bp, a maximum merge length of 600 bp, a minimum overlap of 10 bp, a minimum length of 200 bp after quality trimming (with a minimum Phred score of 30) and no reads with unknown bases allowed. Quality filtering was performed using USEARCH with a minimum Phred score of 30, a minimum length of 300 bp and maximum expected error of 0.5. Sequences were dereplicated and clustered into Operational Taxonomic Units (OTUs) with the UPARSE algorithm with a 97% identity threshold, while simultaneously chimeras were detected and removed de novo using UCHIME (Edgar et al. 2011). Sequences were subsequently classified using the PR2 reference database in Mothur v.1.39.5 (Schloss et al. 2009). Singletons, doubletons and contamination originating from mock communities (OTUs that had a higher abundance in mock samples than in lake and sediment samples) were removed.

Downstream analyses were performed in R v3.6.0 (R Core Team 2019) using the phyloseq package 1.30.0 (McMurdie & Holmes 2013). The DNA data was standardised to the smallest sample size (1,899 reads) using the rarefy_even_depth function, by randomly subsampling with replacement to account for differences in sequencing depth between the samples. This enabled investigating alpha diversity and differences between samples. Alpha diversity was assessed using the plot_richness function and differences between samples were evaluated using non-metric Multidimensional Scaling (nMDS) of a Bray-Curtis matrix of presence-absence data with the plot_ordination function.

3.4 Core analysis 3.4.1 Radiocarbon dating Two sediment samples from core 1 (C1_10 and C1_19) and four from core 2 (C2_04, C2_12, C2_23 and C2_34) were used for 14C dating (see appendix 1 for pictures). Plant material was collected from sample C2_04 and also used for radiocarbon dating. Carbon dating was performed by Beta Analytic (USA) using the Accelerator Mass Spectrometry (AMS) technique. Calibration of the radiocarbon ages was performed via the IntCal-13 database.

3.4.2 Fossil pigment analysis As pigments are light sensitive, work was always done in a dark room with red light. Samples were freeze-dried prior to extraction. 90% acetone filtered over a 2 µm filter was added to the freeze-dried sediment in an aluminium cup resistible to acetone. The samples were sonicated for 30 seconds by an amplitude of 40 and stored overnight by 4 °C. Extracts were filtered over a 0.2 µm PFTE filter.

Pigments were isolated and analysed following the method of Van Heukelem (Van Heukelem & Thomas 2001) and using an Agilent technologies 1100 series HPLC (high pressure liquid chromatography). The HPLC instrument consisted of a high pressure pump, injector (injection volume 100 µl), an Agilent Eclipse XDB-C8 column and a diode array detector (DAD) set to monitor 450 and 665 nm. A gradient of two solvents, namely methanol and buffered methanol (methanol/TBAA 28 mM 70/30) were used. The HPLC system was calibrated using the one-point calibration method with 30 certificated pigment standards provided by the Danish company DHI. The identification of the pigments was based on the retention time and specified maximum absorption spectrum of each pigment (Jeffrey et al. 1997). Pigments that only met one out of the two conditions were assigned as pigment-likes. The taxonomic assignment of the pigments was based on Jeffrey et al. (1997) and Leavitt and Hodgson (2001). Calculation of pigment concentrations in the samples was done using the specific response factor (Rf) of the pigment standards as follows:

Rf pigment standard x area x extraction volume Pigment concentration in sample = dry mass sediment

The pigment stratigraphy was divided into zones using constrained cluster analysis (CONISS; Grimm 1987) of a Bray-Curtis distance matrix and the significance of the zones was assessed using the broken stick model (Bennett 1996) in the rioja package 0.9-21 for R (Juggins 2017).

3.4.3 Ancient DNA analysis 3.4.3.1 Setting up an ancient DNA laboratory Since there was no dedicated ancient DNA laboratory present at Ghent University, a room had to be prepared for this purpose. This new ancient DNA room was isolated from all other rooms in the laboratory where for example PCR is performed routinely. During the whole period of processing the ancient DNA, cleaning staff and other lab users were not allowed to enter the room. In this room, subsampling, RNA and DNA extraction and PCR took place.

Preparations and procedures are described in detail in the protocol that can be found in the appendix (appendix 2). In short, before lab material was brought in, the room was completely cleaned with

bleach 0.1%. Working surfaces were additionally disinfected1 with ethanol (70%), DNA away (Thermo Scientific) and RNase away (Thermo Scientific).

The ancient DNA room was equipped with a Biohazard cabinet in which all sample preparation was done. The outside and inside were completely disinfected. A freezer, centrifuge, PCR machine, vortex adapter and pipettes were brought in from other lab rooms after being thoroughly disinfected. Milli- Q water was autoclaved and filter sterilised through a 0.22 µm filter.

3.4.3.2 Preparation and procedures before starting Every time the ancient DNA room was entered, a procedure was followed to be sure a sterile environment could be maintained (appendix 2). In short, hands were first cleaned with soap and two pairs of gloves were put on, both cleaned with alcohol gel. A hairnet, face mask and Tyvek suit (DuPont) were put on. The Biohazard and all surfaces that were used that day were disinfected. Small jars with filter-sterilised water were put as negative controls on strategic places in the ancient DNA room. One was just outside the lab where hands were washed, one where the Tyvek suit was put on, two on frequently used working benches and finally one on the window sill where disinfection products were placed. Those 5 negative controls were frequently refilled to be sure the water did not completely evaporate. Three negative controls were placed in the Biohazard and were replaced every day. In case there was not enough filter sterilised water, autoclaved milli-Q water was filter sterilised through a 0.22 µm filter.

Small equipment was sterilised with heat at 140 °C for 4h, or disinfected in the conventional way when sterilising with heat was not possible. Tweezers, spoons and spatulas were additionally sterilised in the autoclave.

3.4.3.3 Subsampling core Core subsampling was performed according to the protocol attached in the appendix (appendix 2). In short, the outer 0.5 cm of the cores was scraped off with a sterile sharp knife. The cores were subsequently sterilised with bleach, ethanol, DNA away (Thermo Scientific) and RNase away (Thermo Scientific) and rinsed of with autoclaved and filter-sterilised water after each solution. Each core was subsampled at 1 cm resolution. In total 58 samples were retrieved, 21 from core 1 and 37 from core 2. Next, the samples were subsampled for RNA and DNA extraction, pigment analysis and radiocarbon dating. Approximately 2 g of sediment was sampled for RNA and DNA extraction, 4 g for pigment analysis and 2 g of 6 samples was sampled for radiocarbon dating (appendix 3).

3.4.3.4 RNA and DNA extraction and RNA conversion to cDNA RNA of the core samples was extracted in order to find the active part of organisms in the samples, which enables recognition of present-day contamination as only aDNA was of interest. RNA extraction was done using the RNeasy PowerSoil Total RNA Kit (Qiagen, appendix 4). In short, 2 grams of sediment was homogenized in a tube containing silica carbide beads, lysis buffers, phenol/chloroform/isoamyl alcohol (pH 6.5-8) and solution IRS to ensure complete lysis of all cells and neutralization of RNases. Nucleic acids in the lysates were then precipitated and resuspended in a buffer optimized for binding to anion-exchange gravity flow columns. RNA was eluted using a high-salt buffer, pure RNA was precipitated and resuspended in RNase-free water. Extracted RNA was stored at -80 °C before

1 Disinfection means applying bleach 0.1%, ethanol 70%, DNA away (Thermo Scientific) and RNase away (Thermo Scientific), unless described otherwise. 9

conversion to cDNA. RNA quality and quantity of a selection of samples was checked on an Agilent Bioanalyzer 2100 with a Eukaryote Total RNA Pico chip.

RNA to cDNA conversion was performed using the iScript cDNA Synthesis Kit (Bio-Rad). First, iScript Reverse Transcriptase, iScript Reaction Mix and RNA template were mixed. The reaction mix was incubated in a thermal cycler for 5 min at 25 °C, 20 min at 46 °C and 1 min at 95 °C. cDNA was stored at -20 °C.

DNA of the core samples was co-extracted with the RNeasy PowerSoil DNA Elution Kit (Qiagen, appendix 5). An alternative high-salt buffer was added to the column to elute the DNA. The DNA was precipitated and resuspended in RNase-free water. Extracted DNA was stored at -20 °C. DNA quality and quantity was checked on an Agilent Bioanalyzer 2100 with a High Sensitivity DNA chip.

DNA of negative controls (jars with water and swabs of a falcon, Whirlpack and Ziploc) was extracted using the DNeasy PowerLyzer Microbial Kit (Qiagen, appendix 6). Water and swabs were mixed with a bead solution and added to a bead beating tube containing 0.1 mm glass beads. Lysis solution was added and cells were lysed by a combination of heat, detergent and mechanical force using a TissueLyser II (Qiagen). The DNA from the lysed cells was captured on a silica membrane in a spin column format. DNA was then washed, eluted and stored at -20 °C.

3.4.3.5 Amplification PCRs of the core samples and negative controls were performed in duplicate using the 18S rRNA primers TAReuk454FWD1 and TAReukREV3 (Stoeck et al. 2010), and the FastStart High Fidelity PCR System kit (Sigma-Aldrich). The final 25 µl volumes for amplification contained 2 µl of DNA sample, 0.25 µl FastStart High Fidelity Enzyme Blend, 2.5 µl FastStart High Fidelity Reaction Buffer, 2.5 µl dNTP’s, 1 µl of forward and reverse 18S rRNA primer, 1 µl BSA and 14.75 µl water. The PCR mixtures underwent an activation step of 5 min at 94 °C, followed by 40 cycles of 1 min at 94 °C, a touchdown annealing temperature of 57-52 °C for 1 min and 3 min at 72 °C, finishing with a final elongation step of 20 min at 70 °C.

3.4.3.6 Library preparation and Illumina sequencing Prior to library preparation, PCR amplicons were purified using paramagnetic beads with the Agencourt AMPure XP PCR purification system. Sequencing libraries were prepared by adding indexes and primer adapters. This was followed by sequencing on an Illumina sequencer. These steps were performed by NXTGNT (UGent).

3.4.3.7 Bioinformatic analyses, taxonomical assignment and downstream analysis Quality filtering and clustering into Operational Taxonomic Units (OTUs) was performed using a customized bioinformatics pipeline (Tytgat et al. 2019) as already described in ‘3.3.1 Bioinformatic analyses, taxonomical assignment and downstream analysis’. Unclassified OTUs with a high abundance (the 35 most abundant OTUs for each core) and unclassified Embryophyceae and Craniata were additionally compared to the NCBI nucleotide database using BLAST (http://www.ncbi.nlm.nih.gov/blast/; Accessed July 2020). OTUs that (1) had a higher frequency in negative controls than in samples and (2) were obviously non-native species, were removed.

Downstream analyses were performed in R v3.6.0 (R Core Team 2019) using the phyloseq package 1.30.0 (McMurdie & Holmes 2013). The aDNA data was standardised to the smallest sample size (5,399

reads) using the rarefy_even_depth function, by randomly subsampling with replacement to account for differences in sequencing depth between the samples. Alpha diversity was assessed using the plot_richness function. The aDNA stratigraphy was divided into zones using constrained cluster analysis (CONISS; Grimm 1987) of a Bray-Curtis distance matrix and the significance of the zones was assessed using the broken stick model (Bennett 1996) in the rioja package 0.9-21 for R (Juggins 2017). Only taxa that had a minimum frequency of 5% in the samples were used in the cluster analysis to ensure a clear division of zones and including as many taxa as possible (Bennett 1996).

4 Results 4.1 Present-day lake and surface sediment samples After quality filtering using a customized bioinformatics pipeline (Tytgat et al. 2019), 442,909 sequences were left, clustering into 921 OTUs. Two samples, OSPEL17A and SA17G20-5-1, were left out of the analysis due to very bad quality according to the FastQC files. After removal of singletons, doubletons and of contamination originating from mock communities, 441,995 sequences clustering into 884 OTUs remained.

4.1.1 Alpha diversity Because the rarefaction curve of the sample with the smallest sample size, OSPELDEEP A (1,899 reads), was already flattening, data was rarefied to this sample size. After random subsampling, 303 OTUs were removed because they were no longer present in any sample. The observed alpha diversity (= number of OTUs) was highest for the littoral samples, slightly lower for the top sediment samples and lowest for the pelagic samples (Fig. 2). Sediment second layer samples only consisted of 2 samples, one with a very low and one with a very high observed alpha diversity (Fig. 2). The Shannon index, accounting for both abundance and evenness of the OTUs present, showed similar patterns when comparing the different sample types, but differences between them were slightly less (Fig. 2).

Figure 2 Boxplots showing rarefied alpha diversity measures (observed alpha diversity and Shannon index) in present-day lake and sediment samples from lake Sarsvatnet.

4.1.2 Taxonomic composition of the samples All sediment and littoral samples were dominated by Metazoa (43-94% of all reads), while this phylum was nearly absent in the pelagic samples (Fig. 3). Metazoa in the sediment samples mainly consisted of Ostracoda, more specifically the species Candona candida, which is the OTU with the most sequences overall (134,937 reads pre-rarefaction). Metazoa in the littoral samples differed in their composition depending on the sample. OSLAC17A and OSLAC17C were also dominated by Ostracoda, but OSLAC17B was dominated by Rhabdocoela, more specifically the flatworm Castrada luteola. 12

Another relatively abundant organism in the littoral and some sediment samples within the Metazoa was Chaetonotus sp. (Gastrotricha).

The phylum Ochrophyta, which contains i.a. diatoms and chrysophytes, was present in most of the samples (2-20%) (Fig. 3). Chrysophytes dominated in the littoral and pelagic samples, while diatoms were the most abundant Ochrophyta in the sediment samples.

The phylum with the highest abundance in the pelagic samples was Dinoflagellata (92-93%) (Fig. 3). They mainly consisted of unclassified Dinophyceae and one identified subgroup, Gymnodiniales. Two dinoflagellates that also appeared in all samples were Borghiella tenuissima and Polarella glacialis.

Chlorophyta and Streptophyta, belonging to the superphylum Archaeplastida, were not found in the pelagic samples, but were present with a relatively low abundance in the sediment samples (1-3% and 0-1% for Chlorophyta and Streptophyta respectively) and with a slightly higher abundance in the littoral samples (4-12% and 0.3-18% respectively) (Fig. 3). Sample OSLAC17C had a very high abundance of Embryophyceae (Streptophyta) compared to the other samples.

Figure 3 Barplot of phyla present in present-day lake and sediment samples from lake Sarsvatnet. OSLAC17A-C were littoral samples, OSPELDEEPA-B were pelagic samples, SA17G10.0_5, SA17G20.0_5 & SA17G30.0_5 were the top 0.5 cm of the sediment and SA17G10.5.1 & SA17G20.5.1 were the second layer (0.5-1 cm) of the sediment. Phyla were grouped per Superphylum: Alveolata (greens), Amoebozoa (yellow), Archaeplastida (shades of pink), unclassified Eukaryota (cyan), Hacrobia (blues), Opisthokonta (oranges), Rhizaria (red) and Stramenopiles (purples).

According to the composition of different phyla in the samples, the littoral and sediment samples looked very similar, whereas the pelagic samples were very dissimilar. However, out of all OTUs found in the littoral samples, 40% was also found in the sediment samples, which was the same percentage for the pelagic samples. Non-metric Multidimensional Scaling (nMDS) showed that the pelagic, littoral and sediment samples clustered in three different clusters (with one outlying sediment sample), suggesting that that the three sample types had different species communities (Fig. 4).

Figure 4 Non-metric Multidimensional Scaling (nMDS) of a Bray-Curtis matrix of presence-absence data from lake and sediment samples from lake Sarsvatnet. 4.2 Lake sediment core 4.2.1 Chronology Ages were reported in 14C ages and calibrated years before present (cal. yr BP, BP = AD 1950). Calibrations of radiocarbon ages were applied to convert the Conventional Radiocarbon Age to calendar years. The average age range was 12,746 – 10,053 cal. years BP for core 1 and 18,837 – 9,286 cal. years BP for core 2 (Table 1). Organic sediment and plant material from the same sample (C2_04) differed approximately 500 cal. years in their age.

Table 1 Radiocarbon dates and calibrated ages for core 1: SV18L1_S1_LV1 and core 2: SV18L1_S1_LV2 from lake Sarsvatnet

Sample nr Depth (cm) Analysed material 14C age (yr BP) Age (cal. yr BP) Average age (cal. yr BP) C1_10 9-10 Organic sediment 8920 +/- 30 10188 - 9917 10053 C1_19 18-19 Organic sediment 10870 +/- 30 12795 - 12696 12746 C2_04 3-4 Organic sediment 8770 +/- 30 9907 - 9659 9783 C2_04 3-4 Plant material 8300 +/- 40 9434 - 9138 9286 C2_12 11-12 Organic sediment 9250 +/- 30 10519 - 10282 10401 C2_23 22-23 Organic sediment 10200 +/- 30 12049 - 11765 11907 C2_34 33-34 Organic sediment 15590 +/- 50 18954 - 18720 18837 14

4.2.2 Core description The deepest zone (21-3 cm) of core 1 had an olive grey colour and its appearance remained more or less constant (Fig. 5). The upper zone (3-0 cm) consisted of brown organic sediment. The sediment in this zone looked slightly disturbed which could be due to recent activity in the sediment or disturbance during coring.

The deepest part of the second core (37-30 cm) consisted of grey clay, which were possibly sediments originating from a deglaciation event (Fig. 5). The following zone (30-21 cm) had an olive grey colour. This was followed by a more greyish zone (21-4 cm). The upper part of the core (4-0 cm) was again olive grey and also looked disturbed.

Figure 5 Visual representation of the two cores from lake Sarsvatnet. Depths, calibrated radiocarbon ages and pigment and aDNA zones were indicated.

4.2.3 Fossil pigments 4.2.3.1 Core 1: SV18L1_S1_LV1 The first core was divided in two distinct pigment zones (Zone C1-P1 and Zone C1-P2) based on the broken-stick model and CONISS cluster analysis of a Bray-Curtis distance matrix (Fig. 6a). Only pigments representing at least 2% of the total chlorophylls or total carotenoids were plotted.

Zone C1-P1 (21-19 cm depth) was characterised by relatively low pigment concentrations, with a mean of 24 µg/g of total chlorophylls and 9 µg/g of total carotenoids. Pheophytin a and pheophytin a-likes, which are chlorophyll a derivatives, represented the most abundant groups in the chlorophylls, followed by chlorophyll a (present in all photosynthetic algae and higher plants). The ratio chlorophyll a/degradation products (pheophorbide a, pheophytin a and pheophytin a-likes) was low in this zone (0.12-0.20). Carotenoids mainly consisted of diatoxanthin (32-33%, produced by diatoms, dinophytes and chrysophytes), lutein & lutein-likes (7-8% and 16-18% respectively, produced by chlorophytes and higher plants) and beta-carotene (7-15%, produced by chlorophytes, higher plants and some phototrophic bacteria). Other carotenoids in this zone with a lower relative abundance (all less than 15

9%) were pigments that are produced by cyanobacteria (zeaxanthin, echinenone, echinenone-likes and canthaxanthin), cryptophytes (alloxanthin and alloxanthin-likes) and chlorophytes and higher plants (violaxanthin-likes). The ratio of chlorophyll derivatives to total carotenoids (CD/TC) was relatively high (1.5-16.7).

Zone C1-P2 (18-0 cm depth) was characterised by an increased pigment concentration, with a mean of 154 µg/g of total chlorophylls and 167 µg/g of total carotenoids. Pheophytin a and pheophytin a-like still represented a high percentage of total chlorophylls, together with chlorophyll a. The ratio chlorophyll a/degradation products increased (0.21-0.59), suggesting a better preservation of the pigments. Overall, the composition of carotenoid pigments remained similar. Some small differences were the relative abundance of beta-carotene, which was slightly higher (17-26%), and the relative abundance of alloxanthin(-likes) decreased. Lutein & lutein-likes decreased (2-6% and 8-16% respectively) and diatoxanthin increased (32-42%) slightly in relative abundance. CD/TC ratio was lower than in the previous zone (0.6-1.2)

4.2.3.2 Core 2: SV18L1_S1_LV2 The second core was divided in three distinct pigment zones (Zone C2-P1, Zone C2-P2 and Zone C2-P3) based on the broken stick model and CONISS cluster analysis of a Bray-Curtis distance matrix (Fig. 6b). Only pigments representing at least 2% of the total chlorophylls or total carotenoids were plotted.

Zone C2-P1 (37-33 cm depth) was characterised by very low pigment concentrations, with a mean of 0.21 µg/g of total chlorophylls and 0.16 µg/g of total carotenoids. Chlorophylls only consisted of pheophytin a and pheophytin a-likes. Carotenoids only consisted of beta-carotene, with an exception in the deepest sample, where diadinoxanthin (dinophytes, diatoms, chrysophytes and cryptophytes), diatoxanthin, zeaxanthin and lutein occurred with a relative abundance of respectively 14%, 14%, 5% and 4%. CD/TC ratios ranged from 1.5 to 2.9.

Zone C2-P2 (32-31 cm depth) was characterised by slightly increased pigment concentrations, with a mean of 20 µg/g of total chlorophylls and 2.3 µg/g of total carotenoids. Chlorophylls consisted of pheophytin a, pheophytin a-likes and chlorophyll a. Chlorophyll a/degradation products ratio was low (0.19-0.22). Carotenoids consisted of diatoxanthin (22-34%), lutein (7-12%) and lutein-likes (11-19%), beta-carotene (8-14%), alloxanthin-like (5-14%) and pigments produced in cyanobacteria (canthaxanthin, echinenone and zeaxanthin with 6-13%, 5-10% and 3-6% respectively). CD/TC ratio was high (6.3-50.2).

Zone C2-P3 (30-0 cm depth) was characterised by remarkable higher pigment concentrations, with total concentrations of chlorophylls and carotenoids ranging from 32 to 408 µg/g and 27 to 390 µg/g respectively. The first part of this zone (30-19 cm) was marked with some peaks and valleys (Fig. 6b). At first there was an increase of chlorophyll and carotenoid concentrations between 30 and 27 cm until it peaked at 246 and 350 µg/g respectively, after which it decreased until 77 and 122 µg/g at 24 cm. Then concentrations increased again reaching similar concentrations as the peak before. Chlorophyll a abundance was low at 26 cm depth, which is slightly before the lower total pigment concentrations, and was compensated by a higher abundance of pheophytin a(-likes). Overall, chlorophyll a/degradation products ratio increased compared to the previous zone (0.32-1.32). Diatoxanthin and lutein-like abundance decreased (20-26% and 2-12% respectively) at 26-25 cm depth and was compensated by a higher abundance of a pigments produced in cyanobacteria, echinenone

(16-18%). The second part of the zone was more or less constant with a pigment composition very similar to zone C1-P2 from core 1. The most abundant carotenoids were diatoxanthin (26-42%), beta- carotene (18-23%), lutein-likes (10-14%) and zeaxanthin (6-12%). CD/TC ratios ranged from 0.6 to 1.4.

(a)

(b)

Figure 6 Stratigraphic diagram of fossil pigments in core SV18L1_S1_LV1 (a) and core SV18L1_S1_LV2 (b) from lake Sarsvatnet with chlorophylls in green and carotenoids in orange. The plot shows the relative abundance of chlorophyll and carotenoid pigments separately, with concentration of total chlorophylls and carotenoids expressed in µg/g dry weight. CD/TC shows the ratio of chlorophyll derivatives to total carotenoids. Depths are shown on the y-axis with corresponding calibrated ages.

4.2.4 Ancient DNA The aDNA of ten samples for each core was sequenced. After quality filtering using a customized bioinformatics pipeline (Tytgat et al. 2019), 212,741 sequences were left, clustering into 514 OTUs. In total 233 out of the 514 OTUs had a higher frequency in negative controls than in the core samples, of which 203 OTUs only occurred in negative controls. After removing OTUs that (1) had a higher frequency in negative controls than in samples and (2) were non-native plant species (Cucumis sativus, Taxus sp., Platanus occidentalis and Pelargonium sidoides) and Homo sapiens, 157,971 sequences were left, clustering into 275 OTUs.

4.2.4.1 Alpha diversity and taxonomic composition Data was randomly subsampled to the smallest sample size (5,399 reads). Observed alpha diversity ranged from 18 to 98 OTUs per sample (Fig. 7). Shannon indices for samples from core 1 were between 2.6 and 3.3, with one sample having a very low index (1.3 for sample C1_20). Shannon indices for samples from core 2 ranged between 2 and 3.1, with a slightly decreasing trend from younger to older samples.

Figure 7 Rarefied alpha diversity measures (observed alpha diversity and Shannon index) in aDNA core samples from lake Sarsvatnet.

A lot of unclassified Eukaryotes were present in every sample (3-44%), suggesting that the reference database was not sufficient to classify all organisms. The most abundant groups that occurred in almost all samples were Dinoflagellates (4-41% and 0-51% for core 1 and 2 respectively), Chlorophyta (11-29% and 7-55%) and Ochrophyta (1-42% and 4-23%) (Fig. 8). The abundance of dinoflagellates stayed stable throughout core 1, but they were absent in the oldest sample of core 2, then increased until 51% and then decreased again until 5% in the most recent sample. The OTU with the most sequences (18,511 reads pre-rarefaction) was the dinoflagellate Polarella glacialis and occurred in all samples.

Sample C1_20 was different from the other samples as Cercozoa were very abundant (63%). The deepest sample in both cores (C1_21 and C2_32) and two other samples in core 2 (C2_05 and C2_10) had more Metazoa (14-26%) compared to the other samples (0-3% and 0-6% in core 1 and 2 respectively) (Fig. 8).

Streptophyta occurred sporadically in core 1; they were present in 5 samples (C1_01-03, C1_10 and C1_16) with an abundance of 0.5-10% and were absent in the other samples (Fig. 8). Streptophyta had a high abundance in the oldest sample (C2_32) of core 2 (17%), but a very low abundance in the other samples (1% or less). The Phylum of Streptophyta only consisted of 4 OTUs: one green alga Cosmarium granatum, 1 moss Sanionia uncinata and two vascular plants Mitella sp. and an unclassified Asteracea.

Figure 8 Barplot of phyla present in samples from two cores from lake Sarsvatnet. Phyla were grouped per Superphylum: Alveolata (greens), Amoebozoa (yellow), Archaeplastida (shades of pink), unclassified Eukaryota (cyan), Hacrobia (blues), Opisthokonta (oranges), Rhizaria (red) and Stramenopiles (purples).

4.2.4.2 Stratigraphy The first core was divided in two distinct zones (aDNA-zone C1-D1 and C1-D2) based on the broken stick model and CONISS cluster analysis of a Bray-Curtis distance matrix (Fig. 9a). aDNA-zone C1-D1 corresponded to pigment zone C1-P1 and aDNA-zone C1-D2 with pigment zone C1-P2. Only taxa that had a minimum frequency of 5% in the samples were used in the cluster analysis. The 30 most abundant taxa were shown in the stratigraphic plot and OTUs that couldn’t be classified on phylum level using the PR2 reference database were not plotted.

Alveolates in aDNA-zone C1-D1 mainly consisted of the dinoflagellates Polarella glacialis and Woloszynskia cincta and a sister group to dinoflagellates, Perkinsea. Chlorophyta were only represented by Choricystis sp. and Tetracystis sarcinalis. Streptophyta were not present in this zone. Other taxa that had a high abundance in this zone were Craniata uncl. (possibly polar bear, see 5.1 Representation of aquatic and catchment community in surface lake sediments) and Glissomonadida uncl. Only a few Ochrophyta were present: the diatom Navicula radiosa and an unclassified Eustigmatophycea.

All Alveolate taxa had a marked increase in aDNA-zone C1-D2. Polarella glacialis remained abundant, with its highest abundance in the top samples. Choricystis sp. decreased in this zone and eventually almost disappeared, whereas other Chlorophyta appeared. Streptophyta occurred for the first time in the record with 3 species: the vascular plant Mitella sp., a moss Sanionia uncinata and a green alga Cosmarium granatum. Craniata uncl. sporadically appeared, but with low abundance. Glissomonadida completely disappeared in this zone. Several Ochrophyta appeared sporadically, of which Pseudotetraedriella kamillae had a high abundance throughout the whole zone.

The second core was also divided in two distinct zones (aDNA-zone C2-D1 and C2-D2) based on the broken stick model and CONISS cluster analysis of a Bray-Curtis distance matrix (Fig. 9b). The deepest aDNA-zone C2-D1 corresponded to the pigment zone C2-P2 and the uppermost aDNA-zone C2-D2 corresponded to the pigment zone C2-P3. Only taxa that had a minimum frequency of 5% in the samples were used in the cluster analysis. The 28 most abundant taxa were shown in the stratigraphic 19

plot and OTUs that couldn’t be classified on phylum level using the PR2 reference database were not plotted. aDNA-zone C2-D1 only consisted of one sample (C2_32) and was marked with the absence of Alveolata and almost all Chlorophyta. Choricystis sp. appeared in this zone with a low abundance. Taxa that dominated this zone were the moss Sanionia uncinata, Craniata uncl., Paracercomonas sp. and Nannochloropsis sp.

Figure 9 Stratigraphic plots of aDNA in core SV18L1_S1_LV1 (a) and core SV18L1_S1_LV2 (b) from lake Sarsvatnet with Opisthokonta (light orange), Rhizaria (dark orange), Alveolata (greens), Stramenopiles (red) and Archaeplastida (blue). The plots show number of reads for each OTU after rarefaction. Pigment zones were indicated on the left for comparison. *Classification based on NCBI database, detailed BLAST results can be found in appendix 7. **Had poor results when comparing to PR2 and NCBI database, presumably polar bear based on alignments.

In aDNA-zone C2-D2, several dinoflagellates like Polarella glacialis and Warnowia sp. and other Alveolata occurred for the first time in the record. Also several Chlorophyta appeared in this zone. Sanionia uncinata completely disappeared in this zone, Craniata uncl. still was present but with a low abundance and Paracercomonas sp. also disappeared. A Rhabdocoela flatworm was sporadically abundant. The diatom Navicula radiosa only appeared in the most recent sample and some Eustigmatophyceae occurred, of which Pseudotetraedriella kamillae had a relatively high abundance, as was also the case in the upper aDNA-zone C1-D2 of core 1. Nannochloropsis sp. still was present in the deeper part of the zone but then disappeared.

4.2.4.3 Comparison to fossil pigments In the pigment analysis, diatoxanthin was the most abundant carotenoid. It is produced by diatoms, crysophytes and dinophytes. The first two groups belong to the Ochrophyta, which was also a relatively abundant group in the aDNA analysis (Fig. 8). Dinoflagellates also represented a relatively important group in the aDNA analysis (Fig. 8). Based on the total concentrations of both chlorophyll and carotenoids, algal abundance was low in zones C1-P1, C2-P1 and C2-P2 (the deepest zones in each core) and increased in zones C1-P2 and C2-P3. The deeper aDNA zones also showed lower algal abundance in aDNA samples, and increasing algal abundance in the younger zones of both cores (Fig. 9). Alloxanthin, a pigment produced by planktonic cryptophytes and therefore an indicator for this group, was present in the core, but with low abundance and slightly higher in zone C1-P1 and C2-P2 (Fig. 6). However, chryptophytes were very rare in the aDNA data and the slight increase in these zones was not detected.

4.2.4.4 Comparison to present-day communities The high abundance of Metazoa and more specifically the ostracod Candona candida in the recent top sediment samples was not found in the aDNA analysis, instead C. candida only appeared in one sample (C2_07) with a low abundance (44 reads pre-rarefaction). Ochrophyte abundances were comparable in the two analyses. Chlorophyta had lower abundances in the recent sediment and littoral samples compared to aDNA samples. Dinoflagellates in the aDNA analysis had a higher abundance than in the recent sediment and littoral samples, but still lower than in the pelagic samples.

5 Discussion 5.1 Representation of aquatic and catchment community in surface lake sediments According to the present-day communities found in the lake samples, there is more similarity between littoral and sediment samples than between pelagic and sediment samples. The main difference between pelagic samples and the other samples is the dominance of dinoflagellates in the former. In all other samples, this group only represents a very small part, while in pelagic samples they are the dominant group. However, the number of reads not always represents the abundance of taxa. There are a number of factors that influence the abundance of an OTU in a sample, and one of them is the varying gene copy number among eukaryotes (Zhu et al. 2005, Hirakawa & Ishida 2014). The number of 18S rRNA gene copies was estimated by Zhu et al. (2005) and their results showed that this copy number ranged from 1 in the picoplanktonic species Nannochloropsis salina to more than 12,000 for a large dinoflagellate Akashiwo sanguinea. Also nanoplanktonic dinoflagellates were estimated to have around 1,000 copy numbers. The high abundance of dinoflagellate sequences in the pelagic samples should thus be interpreted with caution, as it can be the result of a high gene copy number from a particular or some dinoflagellate species, rather than a high abundance in the communities.

The Ostracod Candona candida, that was omnipresent in the littoral and sediment samples, is a common and widespread species with a Holarctic distribution. It occurred in littoral and benthic zones, which corresponds to the locations where the species is found in this study. Its absence in pelagic samples is, apart from the dinoflagellates, an important reason for the species composition to look very different from the other samples.

Several diatoms that were found in the sediment samples were also present in littoral samples. The sediment cores, however, contained less chrysophytes compared to littoral and pelagic samples. In general, littoral and sediment samples look similar in phyla composition. Out of all OTUs found in littoral samples, 40% was also found in sediment samples. This was the same for pelagic samples, as 40% of OTUs in pelagic samples was found in sediment samples. Numerous studies have already shown that aquatic organisms can be identified from DNA extracted from lake sediments (e.g. D’Andrea et al. 2006, Coolen et al. 2004). Capo et al. (2015) indicated that 71% of planktonic, unicellular eukaryotes found in the water column were detected in sediment samples, which is quite a bit higher than the percentage (40%) found in this study. Only a few groups, like Cryptophyta and Haptophyta, seemed to be poorly preserved in sediments. Indeed, this study also showed a very low abundance of Cryptophyta in recent sediment samples (0-0.1%), in contrast with a slightly higher abundance in pelagic samples (1.5-1.6%), although the difference was small and Haptophyta were nearly absent in all samples.

DNA from lake sediments has also proven to be efficient in detecting aquatic vegetation, as it may be comparable with, or even superior to in-lake vegetation surveys (Alsos et al. 2018). Several studies also assessed catchment (terrestrial) vegetation communities through aDNA analysis in lake sediments (e.g. Alsos et al. 2016, 2018; Parducci et al. 2017; Sønstebø et al. 2010). In a study by Alsos et al. (2018), a mean of 31% of taxa growing within 2 m of several lake shores in northern Norway were detected with DNA from lake sediment. However, in this study, a much more limited amount of the catchment vegetation was detected in the sedimentary DNA. Only 3 vascular plant species and 6 mosses were found in the recent sediment samples collected in 2017, and 2 vascular plant species and 1 moss were found in the two cores from 2018. Field studies indicate though that there is a very diverse plant community present around the lake, with species like Dryas octopetala, Carex rupestris and Cassiope 22

tetragona (Birks et al. 2004b, Kern et al. 2019). More intensive sampling may have led to the detection of a higher proportion of rare taxa (Alsos et al. 2018), like the use of a larger amount of sediment as suggested by Capo et al. (2015). However, this limited detection was most probably caused by the 18S rRNA barcode marker used in this thesis, which is not plant-specific (see 5.4 Perspectives). This universal barcode marker however made it possible to detect a wide variety of organisms, including mammals. One OTU was identified as mammal and appeared repeatedly in the two sediment cores. Based on alignment of the sequence with other mammals that appear in Svalbard, some closely related mammals and possible contaminating mammals (reindeer, deer, Arctic fox, seal, grizzly bear, cat and human) and the fact that it occurs in both cores in several depths and not in negative controls (so contamination would be very unlikely), polar bear would be the most likely explanation. Moreover, the NCBI database does not contain 18S rRNA sequences of polar bear or grizzly bear. Mammal DNA has also been retrieved from sediments in previous studies in i.a. the Arctic (Willerslev et al. 2014), northern Sweden (van Woerkom 2016) and Alaska (Haile et al. 2009). 5.2 Protocol for aDNA processing Out of a total of 514 OTUs that were found in the core samples, 233 had a higher frequency in negative controls than in the core samples, of which 203 OTUs only occurred in negative controls. Furthermore, an important amount of the organisms found in the negative control samples were marine organisms. These observations suggest that there was at least some contamination present in the air in the laboratory that ended up in the jars with water. It also shows that very few of these organisms were also found in the core samples, indicating that the protocol for contamination prevention was needed and effective. The aDNA found in the core samples can thus (after removal of OTUs that occurred with a higher abundance in negative controls compared to core samples) assumed to be reliable data, with contamination being limited. These results also show that protocols that address preparing ancient samples for DNA extraction and the DNA extraction itself are essential, as is already stated in previous studies (Shapiro et al. 2012). 5.3 Lake evolution and environmental reconstruction 5.3.1 Deglaciation and colonization after the Last Glacial Maximum The deepest part of the longest core (core 2, 37-33 cm), coinciding with pigment zone C2-P1 around 18,837 cal. years BP, was very distinct from other zones in the core. It was marked by a very low concentration of carotenoid and chlorophyll pigments, suggesting low primary production. The only pigments present, at extremely low concentrations, were pheophytin, a derivative of chlorophyll a, and beta carotene, which is one of the most stable carotenoids. The lithology of this zone suggests inflow of glacial deposition, as the zone consists of grey, glacial clays. This glacial clay could be coming from glaciers present in the catchment that began to retreat after the Last Glacial Maximum. Deglaciation in western Svalbard is also estimated by several studies to have taken place during the Late Glacial between 20,500 and 14,400 cal. years BP (Jessen et al. 2010, Husum et al. 2019). The inflow of glacial clay causes lower light intensities in the lake (Laspoumaderes et al. 2013), which can be the reason for the observed low primary production.

Pigment concentrations began to rise slightly in the following pigment zone C2-P2 (33-30 cm). This, combined with the relatively high abundance of the moss species Saniona uncinata in this clay zone (aDNA-zone C2-D1) a little after 18,837 cal. years BP, could be due to rapid colonization of the lake catchment after deglaciation. Bryophyte communities generally develop almost immediately in 23

deglaciated areas (Stebel et al. 2018). Sanionia uncinata is a common, widespread moss species throughout the whole Svalbard archipelago. It is a dominant moss species in the later stages of succession, but also occurs in the early successional stage (Nakatsubo et al. 2005). Studies of lacustrine sediments in northern and north-western Svalbard also suggest glacier retreat and species colonization during the Late Glacial (Gjerde et al. 2018, Voldstad et al. 2020). The presence of the microalga Nannochloropsis sp. could indicate cold water conditions, as it is suggested that Nannochloropsis sp. may be better adapted to cold-water habitats (Fawley & Fawley 2007). Light intensities were probably still low during this period, as there was still inflow of glacial clays, probably causing the low pigment concentrations and thus primary productivity. The presence of Nannochloropsis sp. could be explained by its ability to continue growing in low light intensities (Palacios et al. 2018).

5.3.2 Late Glacial and early-mid Holocene The oldest material in the shortest core (core 1) did not consist of glacial clay like the second core. Instead the core started a bit earlier than 13,000 cal. years BP (exact dating of the deepest part is lacking), suggesting the deglaciation event was terminated by then. Pigment zone C1-P1, that lasted until around 13,000 cal. years BP, had low pigment concentrations (only a little bit higher than pigment zone C2-P2 in core 2). This could indicate that algal blooms had not started yet. Furthermore, pigment preservation was probably low, as the chlorophyll a/degradation products ratio was low in this period. This was also confirmed by the high chlorophyll derivatives/total carotenoids (CD/TC) ratio, as carotenoids degrade faster than chlorophylls under oxidizing conditions (Gorham & Sanger 1967, Swain et al. 1985).

Nannochloropsis sp. was still present just before the Holocene in the longest core (the onset of aDNA- zone C2-D2 in core 2) but then disappeared completely. This suggests warmer water conditions in the early Holocene as the microalga is adapted to cold-water conditions (Fawley & Fawley 2007). Fossil pigment analysis shows very high pigment concentrations during the early Holocene in both cores (pigment zones C1-P2 and C2-P3), which is an indicator for high algal abundance (Jiang et al. 2011). The low CD/TC ratios in these zones probably indicate better preservation conditions (which was also confirmed by the high chl a/degradation products ratio) and high primary production (Jiang et al. 2011). Results of the aDNA analysis also show higher algal abundance in aDNA-zone C2-D2 compared to aDNA-zone C2-D1 from the Late Glacial. Voldstad et al. (2020) also suggested a transition from 10,800 cal. years BP towards higher algal abundance and more thermophilous species. This is also in accordance with a study from Mangerud and Svendsen (2018) on marine molluscs, indicating that climate on Svalbard was at least as warm or warmer as present at around 11,000-9,000 cal. years BP. Alkenone-based lake-water temperature reconstructions from a lake in north-western Svalbard also showed higher temperatures at the onset of the Holocene, peaking at around 10,500 cal. years BP (van der Bilt et al. 2018).

Whereas the deepest part of the second core had a moss with a relatively high abundance, Embryophyceae were very sparse or even absent in all other zones of the two cores. This is, as stated before, probably mainly because the 18S rRNA barcode region is not plant-specific (see 5.4 Perspectives). However, an additional reason can be the dominance of algal DNA that can cause underestimation of other taxa like bryophytes and vascular plants (Alsos et al. 2018, Voldstad et al. 2020). Excessive algal abundance can cause PCR competition, which could potentially be avoided by

using primers that are designed to block amplification of algae (Alsos et al. 2018), as was also done for human DNA in a study of rare mammal species from sedimentary aDNA (Boessenkool et al. 2012). 5.4 Perspectives At least three main fields for further research can be identified.

First, the 18S rRNA maker used is not optimal to detect (higher) plants. In both the recent sediment samples and the older core samples, there was only a very small part of the OTUs that belonged to Embryophyta (9 and 3 in recent and core samples respectively). However, field studies indicate that there is a very diverse plant community present around the lake, with species like Dryas octopetala, Carex rupestris and Cassiope tetragona (Birks et al. 2004b, Kern et al. 2019) and several studies showed that DNA in lake sediment is a powerful tool to reconstruct catchment vegetation (Alsos et al. 2016, 2018; Parducci et al. 2017; Sønstebø et al. 2010). This low abundance of plants in sedimentary DNA can probably be explained by the 18S rRNA barcode marker used in this thesis. This gene can be found in all eukaryotes and is one of the most commonly used markers for protistan taxa (Pawlowski et al. 2012). It has been suggested as an effective barcode marker for some groups like diatoms (Guo et al. 2015) and foraminifera (Pawlowski & Lecroq 2010). However, this barcode marker is less effective to target other taxa like plants (Pawlowski et al. 2012). According to the CBOL Plant Working Group (2009), the core DNA barcode markers for land plants are regions of two plastid genes, rbcL and matK. Chen et al. (2010) proposed that ITS2 can be a universal barcode marker for a broad range of plants. The short P6 loop region of the plastid DNA trnL intron is also a widely applied marker for vascular plants in environmental samples (e.g. Alsos et al. 2016, 2018; Parducci et al. 2013; Taberlet et al. 2007, Voldstad et al. 2020). These plant-specific markers would most probably detect a larger amount of (higher) plants compared to the 18S rRNA marker that was used in this thesis.

Second, while a multiproxy approach was used (i.e., fossil pigment and aDNA analysis), a diatom analysis of the core samples would provide interesting additional information. Diatom remains in lake sediments provide reliable records of changes in water quality, habitat and catchment processes (Smol et al. 2005). They also allow quantification of lake primary production (Kaplan et al. 2002) and have been critical in detecting recent changes in Arctic lakes (Rühland et al. 2015).

Third, the core chronology should be improved. Radiocarbon dating in this study was performed on 6 samples, namely two from core 1 and four from core 2. This resulted in a preliminary chronology of the cores. However, to gain insight in the precise timing of environmental changes recorded in the lake sediment, additional radiocarbon dating would be required. Especially the upper parts of both cores should be dated in order to reveal the age of the most recent samples. Furthermore, it would be interesting to know when exactly the transition from glacial clay to the other sediment type took place in order to infer the timing of deglaciation (i.e., when the lake was not influenced anymore by the glacier).

6 Conclusion The comparison between DNA found in recent lake (littoral and pelagic) and surface sediment samples suggests that at least part of the aquatic communities are represented in the sedimentary DNA. A more intensive sampling may have improved detection of rare taxa. However, the results show the potential of sequencing sedimentary DNA to reconstruct past responses of aquatic communities to environmental changes. DNA from vegetation in the catchment was much less represented in the sediment, as only a few vascular plants and mosses were detected. More plants would probably have been detected when a plant-specific barcode marker was used, as the 18S rRNA marker used in this thesis was not optimal to detect plants.

The protocol developed in this thesis has shown to be effective in preventing contamination of core samples, which is essential in aDNA research. A lot of the contamination that was found in the negative controls, probably coming from DNA particles in the air of the laboratory, was absent in the core samples. We can therefore assume that the risk of contamination, which is one of the biggest challenges of working with aDNA, can be overcome using our newly established protocol.

The core lithostratigraphy suggests an inflow of glacial deposits during and after the Last Glacial Maximum around 18,837 cal. years BP, as the deepest part of core 2 consisted of glacial clays. Very low pigment concentrations suggest low primary production due to low light conditions in the lake during this deglaciation period. The abundant presence in the following zone of Sanionia uncinata, a dominant moss species in later successional stages, reflects a rapid colonization of the lake’s catchment a little after 18,837 cal. years BP. Pigment concentrations were rising slightly but were still low, indicating that the algal abundance remained low. The absence of glacial clays around 13,000 cal. years BP in core 1 suggests the absence of a glacier in the catchment of the lake and hence regional deglaciation. This was followed by a rise in pigment concentrations until they reached very high concentrations, likely as a result of algal blooms and higher temperatures during the early Holocene. This was also confirmed by the disappearance of Nannochloropsis sp., an alga adapted to cold-water conditions.

This thesis offered some clues regarding past environmental conditions in and around a lake in north- western Svalbard, though some suggestions for further research can be made. Additional barcode markers, more specifically plant-specific markers, could give more insights in past vegetation changes. We also propose that diatom analysis of the core samples is expected to provide interesting additional information. Lastly, an improved core chronology would allow more exact timing of environmental changes.

7 Summary 7.1 English summary Climate change is amplified in the Arctic, as warming in this region is two to three times higher than the global annual average. Warming has already resulted in the retreat of most Arctic glaciers, which exposes new terrestrial habitats that can be colonized. The first colonizers of deglaciated, exposed soils are generally biological soil crusts (BSCs), consisting of algae, cyanobacteria, fungi, lichens and mosses. They have a direct positive influence on vegetation density, species richness and cover and as such contribute to the greening of the Arctic. The High Arctic tundra biome (dominated by dwarf shrubs, forbs, graminoids and cryptogamic covers) is already showing an increased shrub dominance due to climate warming. A widespread redistribution of Arctic vegetation is predicted where in general, low- lying vegetation with sparse plant cover will decrease, while larger shrubs and trees will increase their range.

Climate warming also induces changes in freshwater ecosystems. A previous study based on diatoms in lake sediments in the Arctic revealed significant and widespread ecological reorganisations in many High Arctic lakes. Lake sediments, that accumulate over time, contain plenty of information about the response of the biological communities in the lake and their surrounding catchment to these environmental changes. Svalbard is a very suitable region for paleolimnological studies, as climate warming is amplified, there are many lakes on the archipelago, it is relatively pristine (although local mining industries are present) and it is one of the most accessible regions in the High Arctic.

However, long-term records of past environmental conditions using proxy indicators in lake ecosystems in relation to past climate changes in the Arctic are generally lacking. This information is highly needed and also essential to better predict the responses of biological communities to future environmental changes. The main objective of this thesis was to reconstruct past environmental changes in a High Arctic lake ecosystem using a multiproxy approach including fossil pigment and ancient DNA (aDNA) analyses in two cores from lake Sarsvatnet in north-western Svalbard. To achieve this, we first investigated to what extent contemporary planktonic communities in the lake and the vegetation in its catchment are reflected in the sedimentary record. DNA in littoral and pelagic samples was compared with DNA in top sediment samples to shed light on the degree of overlap in community structure between the samples. Second, a detailed protocol was developed which tackles all procedures and equipment that should be used in aDNA analysis, with a focus on how to avoid contamination during sample preparation. Third, two lake sediment cores were analysed to study the evolution of the lake and reconstruct past changes in lake primary production and the eukaryotic community structure in and around the lake. Fossil pigments were isolated and analysed using an Agilent technologies 1100 series HPLC (high pressure liquid chromatography), consisting of a diode array detector (DAD) set to monitor 450 and 665 nm. aDNA was extracted and amplified, and Illumina high-throughput sequencing of the 18S rRNA gene was performed. A customized bioinformatics pipeline was used for quality filtering and clustering the sequences into Operational Taxonomic Units (OTUs).

Out of all OTUs found in the littoral and the pelagic samples, 40% were also found in the surface sediment samples, suggesting that at least part of the aquatic communities are represented in sedimentary DNA. Dinoflagellates dominated the pelagic samples, which can be the result of a high gene copy number in this group, rather than a high abundance in the communities. Dinoflagellates 27

were much less abundant in the littoral and sediment samples, in which Metazoa dominated, and more specifically the ostracod Candona candida. A very limited amount of the catchment vegetation was detected in the sedimentary DNA, although field studies in previous studies indicated that there is a very diverse plant community present around the lake. This limited detection was most probably caused by the 18S rRNA barcode marker used in this thesis, which is not plant-specific.

Out of a total of 514 OTUs that were found in the core samples, 233 had a higher frequency in negative controls than in the core samples, of which 203 OTUs only occurred in negative controls. This indicates that the protocol for contamination prevention was needed and effective. The aDNA found in the core samples can thus (after removal of OTUs that occurred with a higher abundance in negative controls compared to core samples) assumed to be reliable data, with contamination being limited.

The lithology of the deepest part of the longest core suggests inflow of glacial deposition during and after the Last Glacial Maximum around 18,837 cal. years BP, as the zone consists of grey, glacial clays. The very low pigment concentrations also suggest low primary production due to low light conditions during this deglaciation event. In the following, younger zone of the core a little after 18,837 cal. years BP, Sanionia uncinata was an abundant moss species. It is a dominant moss species in the later stages of succession, but also occurs in the early successional stage. Its abundance could be due to a rapid colonization of the lake’s catchment after deglaciation. Previous studies on lacustrine sediments in Svalbard also suggest glacier retreat and species colonization during the Late Glacial. Pigment concentrations rose slightly, but still remained low, indicating a low algal abundance. Glacial clays were absent in the shorter core, which started around 13,000 cal. years BP, suggesting the absence of glaciers in the lake’s catchment and hence regional deglaciation. Algal blooms probably had not started yet, as pigment concentrations were still low. The following zones in both cores, around the early Holocene, were marked by relatively high pigment concentrations, which indicate high algal abundance. This was probably due to higher temperatures, as was also confirmed by the disappearance of Nannochloropsis sp., an alga adapted to cold-water conditions. This was also in accordance with several previous studies suggesting high temperatures at the onset of the Holocene.

At least three suggestions for further research can be made. First, another barcode marker which is more plant-specific would probably detect more (higher) plants from the catchment. The 18S rRNA gene used in this thesis can be found in all eukaryotes and is one of the most commonly used markers for protistan taxa, but is not optimal to detect plants. Several markers like rbcL, matK, ITS2 and trnL are proven to be more suitable for vegetation reconstructions. Second, a diatom analysis of the core samples is expected to provide interesting additional information about changes in water quality, habitat, catchment processes and primary productivity. Third, additional radiocarbon dating would be required, as only a limited number of samples were dated in this thesis to gain insight in the precise timing of environmental changes recorded in the lake sediment. Especially the upper parts of both cores should be dated, in order to reveal the age of the most recent samples, and the transition from glacial clay to the other sediment type, in order to infer the timing of deglaciation.

7.2 Nederlandse samenvatting De Arctische regio ondervindt een sterk uitgesproken klimaatverandering. De opwarming in deze regio is twee tot drie keer hoger dan het globale jaarlijkse gemiddelde. Deze opwarming veroorzaakte al het terugtrekken van de meeste gletsjers, wat nieuwe terrestrische habitats blootstelt die gekoloniseerd kunnen worden. Biologische bodem korsten, bestaande uit algen, cyanobacteriën, fungi, korstmossen en mossen koloniseren blootgestelde bodems meestal als eerste. Ze hebben een positieve invloed op de densiteit van de vegetatie en soortenrijkdom en dragen zo bij aan de vergroening van het arctische gebied. De arctische toendra, gedomineerd door dwergstruiken, kruidachtige planten en sporenplanten, wordt door de klimaatopwarming meer en meer gedomineerd door struiken. Er wordt voorspeld dat er een wijdverbreide herverdeling van arctische vegetatie zal plaatsvinden, waar over het algemeen gebieden met laag-liggende vegetatie en een schaarse plantenbedekking zullen inkrimpen en gebieden met grotere struiken en bomen zullen uitbreiden.

Klimaatopwarming veroorzaakt ook veranderingen in zoetwater ecosystemen. Een voorgaande studie gebaseerd op diatomeeën in het sediment van meren in het noordpoolgebied toonde aan dat veel van die meren significante en wijdverbreide ecologische reorganisaties ondervinden. Deze meer- sedimenten, die accumuleren doorheen de tijd, bevatten veel informatie over de reactie van biologische gemeenschappen in en rond het meer op omgevingsveranderingen. Svalbard is een zeer geschikte regio voor paleolimnologische studies; klimaatopwarming is sterk uitgesproken, er zijn veel meren in de regio, het is relatief ongerept (al zijn er een aantal mijnindustrieën aanwezig) en één van de meest bereikbare regio’s in het hoge Arctische gebied.

Lange-termijn overzichten van omgevingscondities in relatie met klimaatveranderingen in het verleden, gebruik makende van benaderende indicatoren in meer ecosystemen, zijn over het algemeen schaars. Deze informatie is nodig, en essentieel om betere voorspellingen te kunnen maken over de reacties van biologische gemeenschappen op toekomstige omgevingsveranderingen. Het hoofddoel in deze thesis was het reconstrueren van voorbije omgevingsveranderingen in een arctisch meer-ecosysteem, gebruik makende van een combinatie van indicatoren bestaande uit analyses van fossiele pigmenten en fossiel DNA in twee boorkernen uit het Sarsvatnet meer in het noordwesten van Svalbard. Hiervoor werd eerst onderzocht in hoeverre hedendaagse planktongemeenschappen in het meer en de vegetatie rond het meer teruggevonden kunnen worden in het sediment. DNA in litorale en pelagische stalen werd vergeleken met DNA in top sediment stalen om de overlap in gemeenschapsstructuur tussen de stalen te bekijken. Ten tweede werd een gedetailleerd protocol ontwikkeld dat alle procedures en materiaal nodig bij een fossiel DNA analyse beschrijft, met de focus op hoe contaminatie te vermijden tijdens de voorbereiding van de stalen. Ten derde werden twee sediment boorkernen geanalyseerd om de evolutie van het meer te bestuderen en voorbije veranderingen in primaire productie en de gemeenschapsstructuur van eukaryoten in en rond het meer te reconstrueren. Fossiele pigmenten werden geïsoleerd en geanalyseerd met behulp van een Agilent technologies 1100 series HPLC (high pressure liquid chromatography), bestaande uit een diode array detector (DAD) ingesteld om 450 en 665 nm te detecteren. Fossiel DNA werd geëxtraheerd en geamplificeerd, waarna Illumina high-throughput sequencing van het 18S rRNA gen werd uitgevoerd. Een serie van bio informatica processen werd gevolgd voor kwaliteitsfiltering en het clusteren van de sequenties in Operational Taxonomic Units (OTUs).

Van alle OTUs die gevonden werden in de litorale en pelagische stalen werd 40% ook in de top sediment stalen gevonden, wat erop kan wijzen dat zeker een deel van de aquatische gemeenschappen terug gevonden kan worden in het sedimentair DNA. De pelagische stalen werden gedomineerd door dinoflagellaten, wat eerder verklaard kan worden door het hoge gen copy number (het aantal kopies van een gen) in deze groep van organismen, in plaats van een hoge abundantie in de gemeenschappen. Litorale en sediment stalen, die een veel lagere abundantie van dinoflagellaten hadden, werden gedomineerd door Metazoa, meer bepaald het mosselkreeftje Candona candida. Slechts een heel beperkt deel van de vegetatie rond het meer werd teruggevonden in het sedimentair DNA, hoewel voorgaande studies een heel diverse plantengemeenschap terugvonden rond het meer. Deze lage detectie werd hoogstwaarschijnlijk veroorzaakt door de 18S rRNA merker die gebruikt werd in deze thesis en niet specifiek voor planten geschikt is.

Van de 514 OTUs die gevonden werden in de stalen van de boorkern hadden 233 OTUs een hogere frequentie in negatieve controles dan in de stalen, waarvan er 203 alleen in de negatieve controles voorkwamen. Dit wijst erop dat het protocol voor de preventie van contaminatie nodig en geschikt was. Het fossiel DNA in de boorkern stalen kan dus (na het verwijderen van de OTUs met een hogere abundantie in de negatieve controles dan in de stalen) geïnterpreteerd worden als betrouwbare data, waarbij contaminatie zoveel mogelijk werd beperkt.

Het diepste stuk van de langste boorkern bestaat uit grijze, glaciale klei, wat wijst op glaciale afzetting tijdens en na het Laatste Glaciale Maximum rond 18.837 cal. jaren BP. De heel lage pigmentenconcentraties wijzen ook op weinig primaire productie door lage lichtintensiteiten tijdens deze deglaciatie. In een volgende, recentere zone van de boorkern kort na 18.837 cal. jaren BP was het mos Sanionia uncinata een abundante soort. Het is een dominante mos soort tijdens de latere successiestadia, maar komt ook voor tijdens eerdere stadia. Zijn abundantie kan het gevolg zijn van een rappe kolonisatie rond het meer na de deglaciatie. Voorgaande studies op lacustriene sedimenten in Svalbard suggereren ook terugtrekking van gletsjers en snelle kolonisatie tijdens het late glaciaal. Pigmentenconcentraties stegen lichtjes, maar bleven toch laag, wat een lage abundantie aan algen aantoont. Glaciale klei werd niet gevonden in de korte boorkern, die rond 13.000 cal. jaren BP startte, en suggereert de afwezigheid van gletsjers rond het meer en dus ook van regionale deglaciatie. Algenbloei vond waarschijnlijk nog niet plaats aangezien pigmentenconcentraties nog laag bleven. De volgende zones in beide boorkernen, rond het begin van het Holoceen, werden gekenmerkt door relatief hoge pigmentenconcentraties, wat wijst op een hoge abundantie aan algen. Dit was waarschijnlijk het gevolg van hogere temperaturen, wat ook werd bevestigd door het verdwijnen van Nannochloropsis sp., een alg aangepast aan koud water. Dit stemt ook overeen met voorgaande studies die hoge temperaturen suggereren aan de start van het Holoceen.

Minstens drie suggesties voor verder onderzoek kunnen worden gemaakt. Als eerste zou een plant- specifieke merker waarschijnlijk meer (hogere) planten van rond het meer detecteren. Het 18S rRNA gen die gebruikt werd in deze thesis is aanwezig in alle eukaryoten en is één van de meest gebruikte merkers voor protisten, maar is niet optimaal om planten te detecteren. Er is aangetoond dat verschillende merkers zoals rbcL, matK, ITS2 en trnL beter geschikt zijn voor reconstructies van de vegetatie. Ten tweede wordt verwacht dat een diatomeeën analyse interessante bijkomende informatie zou bieden over veranderingen in waterkwaliteit, habitat, processen rond het meer en primaire productiviteit. Ten derde zouden bijkomende koolstofdateringen nodig zijn om inzicht te krijgen in de precieze timing van omgevingsveranderingen opgeslagen in het sediment, aangezien in 30

deze thesis slechts een beperkt aantal stalen gedateerd zijn. Vooral de bovenste delen van beide boorkernen zouden gedateerd moeten worden, om de ouderdom van de meest recente stalen te kennen, en de overgang van glaciale klei naar het andere type sediment, om de timing van deglaciatie te kunnen afleiden.

8 Acknowledgments The development of this thesis would not have been possible without a group of people who have given the time and energy to help me.

To begin with, I would like to thank my supervisor Prof. Dr. Elie Verleyen for giving me the chance to be a part of the Protistology and Aquatic Ecology Research group and for the given expertise, insights and constructive feedback.

I am also grateful for the guidance, aid in the lab, useful feedback and motivation given by Lotte De Maeyer and Dr. Bjorn Tytgat. Their contributions have helped me a lot to complete this thesis.

Also a big thanks to Ilse Daveloose who was a great help during the pigment analysis and processing of the pigment data.

I also want to thank Sofie D’hondt for helping me with all the practical work and analyses in the lab.

Finally, I want to thank my brother Jonas Decorte for reading through the thesis and giving me useful feedback.

9 Reference list Alsos, I. G., Lammers, Y., Yoccoz, N. G., Jørgensen, T., Sjögren, P., Gielly, L., & Edwards, M. E. (2018). Plant DNA metabarcoding of lake sediments: How does it represent the contemporary vegetation. PLoS ONE, 13(4), 1–23. https://doi.org/10.1371/journal.pone.0195403

Alsos, I. G., Sjögren, P., Edwards, M. E., Landvik, J. Y., Gielly, L., Forwick, M., Coissac, E., Brown, A. G., Jakobsen, L. V., Føreid, M. K., & Pedersen, M. W. (2016). Sedimentary ancient DNA from Lake Skartjørna, Svalbard: Assessing the resilience of arctic flora to Holocene climate change. Holocene, 26(4), 627–642. https://doi.org/10.1177/0959683615612563

Bennett, K. D. (1996). Determination of the number of zones in a biostratigraphical sequence. New Phytologist, 132(1), 155–170. https://doi.org/10.1111/j.1469-8137.1996.tb04521.x

Birks, H. H., & Birks, H. J. B. (2000). Future uses of pollen analysis must include plant macrofossils. Journal of Biogeography, 27(1), 31–35. https://doi.org/10.1046/j.1365-2699.2000.00375.x

Birks, H. J. B., Jones, V. J., & Rose, N. L. (2004a). Recent environmental change and atmospheric contamination on Svalbard as recorded in lake sediments - An introduction. Journal of Paleolimnology, 31(4), 403–410. https://doi.org/10.1023/B:JOPL.0000022542.87722.a6

Birks, H. J. B., Monteith, D. T., Rose, N. L., Jones, V. J. & Peglar, S. M. (2004b). Recent environmental change and atmospheric contamination on Svalbard as recorded in lake sediments – modern limnology, vegetation, and pollen deposition. Journal of Paleolimnology, 31(4), 411–431. https://doi.org/10.1023/B:JOPL.0000022543.03382.fa

Birks, H. H. (2008). The late-quaternary history of arctic and alpine plants. Plant Ecology and Diversity, 1(2), 135–146. https://doi.org/10.1080/17550870802328652

Boessenkool, S., Epp, L. S., Haile, J., Bellemain, E., Edwards, M., Coissac, E., Willerslev, E., & Brochmann, C. (2012). Blocking human contaminant DNA during PCR allows amplification of rare mammal species from sedimentary ancient DNA. Molecular Ecology, 21(8), 1806–1815. https://doi.org/10.1111/j.1365-294X.2011.05306.x

Breen, K., & Lévesque, E. (2006). Proglacial succession of biological soil crusts and vascular plants: Biotic interactions in the High Arctic. Canadian Journal of Botany, 84(11), 1714–1731. https://doi.org/10.1139/B06-131

Capo, E., Debroas, D., Arnaud, F., & Domaizon, I. (2015). Is Planktonic Diversity Well Recorded in Sedimentary DNA? Toward the Reconstruction of Past Protistan Diversity. Microbial Ecology, 70(4), 865–875. https://doi.org/10.1007/s00248-015-0627-2

CBOL Plant Working Group. (2009). A DNA Barcode for Land Plants. Proceedings of the National Academy of Sciences of the United States of America, 106, 12794-12797. http://dx.doi.org/10.1073/pnas.0905845106

Chen, S., Yao, H., Han, J., Liu, C., Song, J., Shi, L., Zhu, Y., Ma, X., Gao, T., Pang, X., Luo, K., Li, Y., Li, X., Jia, X., Lin, Y., & Leon, C. (2010). Validation of the ITS2 region as a novel DNA barcode for identifying medicinal plant species. PLoS ONE, 5(1), 1–8. https://doi.org/10.1371/journal.pone.0008613

Coolen, M. J. L., Muyzer, G., Rijpstra, W. I. C., Schouten, S., Volkman, J. K., & Sinninghe Damsté, J. S. (2004). Combined DNA and lipid analyses of sediments reveal changes in Holocene haptophyte and diatom populations in an Antarctic lake. Earth and Planetary Science Letters, 223(1–2), 225–239. https://doi.org/10.1016/j.epsl.2004.04.014

Crump, S. E., Miller, G. H., Power, M., Sepúlveda, J., Dildar, N., Coghlan, M., & Bunce, M. (2019). Arctic shrub colonization lagged peak postglacial warmth: Molecular evidence in lake sediment from Arctic Canada. Global Change Biology, 25(12), 4244–4256. https://doi.org/10.1111/gcb.14836

D’Andrea, W. J., Lage, M., Martiny, J. B. H., Laatsch, A. D., Amaral-Zettler, L. A., Sogin, M. L., & Huang, Y. (2006). Alkenone producers inferred from well-preserved 18S rDNA in Greenland lake sediments. Journal of Geophysical Research: Biogeosciences, 111(3), 1–8. https://doi.org/10.1029/2005JG000121

Dowdeswell, J. A., Hagen, J. O., Björnsson, H., Glazovsky, A. F., Harrison, W. D., Holmlund, P., Jania, J., Koerner, R. M., Lefauconnier, B., Ommanney, C. S. L., & Thomas, R. H. (1997). The Mass Balance of Circum-Arctic Glaciers and Recent Climate Change. Quaternary Research, 48(1), 1–14. https://doi.org/10.1006/qres.1997.1900

Edgar, R. C., Haas, B. J., Clemente, J. C., Quince, C., & Knight, R. (2011). UCHIME improves sensitivity and speed of chimera detection. Bioinformatics, 27(16), 2194–2200. https://doi.org/10.1093/bioinformatics/btr381

Farnsworth, W. R., Ingólfsson, Ó., Retelle, M., Allaart, L., Håkansson, L. M., & Schomacker, A. (2018). Svalbard glaciers re-advanced during the Pleistocene–Holocene transition. Boreas, 47(4), 1022– 1032. https://doi.org/10.1111/bor.12326

Fawley, K. P., & Fawley, M. W. (2007). Observations on the Diversity and Ecology of Freshwater Nannochloropsis (Eustigmatophyceae), with Descriptions of New Taxa. Protist, 158(3), 325–336. https://doi.org/10.1016/j.protis.2007.03.003

Gjerde, M., Bakke, J., D’Andrea, W. J., Balascio, N. L., Bradley, R. S., Vasskog, K., Ólafsdóttir, S., Røthe, T. O., Perren, B. B., & Hormes, A. (2018). Holocene multi-proxy environmental reconstruction from lake Hakluytvatnet, Amsterdamøya Island, Svalbard (79.5°N). Quaternary Science Reviews, 183, 164–176. https://doi.org/10.1016/j.quascirev.2017.02.017

Gorham, E., & Sanger, J. (1967). Plant Pigments in Woodland Soils. Ecology, 48(2), 306–308. https://doi.org/10.2307/1933116

Grimm, E. C. (1987). CONISS, a FORTRAN-77 program for stratigraphically constrained cluster analysis by the method of incremental sum of squares. Computers and Geosciences, 13(I), 13–35.

Guilizzoni, P. & Lami, A. (2002). Paleolimnology: Use of Algal pigments as indicators. The Encyclopedia of Environmental Microbiology, 2306-2317.

Guo, L., Sui, Z., Zhang, S., Ren, Y., & Liu, Y. (2015). Comparison of potential diatom ‘barcode’ genes (The 18S rRNA gene and ITS, COI, rbcL) and their effectiveness in discriminating and determining species taxonomy in the Bacillariophyta. International Journal of Systematic and Evolutionary Microbiology, 65, 1369–1380. https://doi.org/10.1099/ijs.0.000076

Haile, J., Froese, D. G., MacPhee, R. D. E., Roberts, R. G., Arnold, L. J., Reyes, A. V., Rasmussen, M., Nielsen, R., Brook, B. W., Robinson, S., Demuro, M., Gilbert, M. T. P., Munch, K., Austin, J. J., Cooper, A., Barnes, I., Möller, P., & Willerslev, E. (2009). Ancient DNA reveals late survival of mammoth and horse in interior Alaska. Proceedings of the National Academy of Sciences of the United States of America, 106(52), 22352–22357. https://doi.org/10.1073/pnas.0912510106

Hirakawa, Y., & Ishida, K. I. (2014). Polyploidy of endosymbiotically derived genomes in complex algae. Genome Biology and Evolution, 6(4), 974–980. https://doi.org/10.1093/gbe/evu071

Husum, K., Howe, J. A., Baltzer, A., Forwick, M., Jensen, M., Jernas, P., Korsun, S., Miettinen, A., Mohan, R., Morigi, C., Myhre, P. I., Prins, M. A., Skirbekk, K., Sternal, B., Boos, M., Dijkstra, N., & Troelstra, S. (2019). The marine sedimentary environments of Kongsfjorden, Svalbard: an archive of polar environmental change. Polar Research, 38(0), 1–16. https://doi.org/10.33265/polar.v38.3380

IPCC (2018). Summary for Policymakers. In: Global warming of 1.5°C. An IPCC Special Report on the impacts of global warming of 1.5°C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change, sustainable development, and efforts to eradicate poverty [V. Masson- Delmotte, P. Zhai, H. O. Pörtner, D. Roberts, J. Skea, P. R. Shukla, A. Pirani, W. Moufouma-Okia, C. Péan, R. Pidcock, S. Connors, J. B. R. Matthews, Y. Chen, X. Zhou, M. I. Gomis, E. Lonnoy, T. Maycock, M. Tignor, T. Waterfield (eds.)]. World Meteorological Organization, Geneva, Switzerland, 32 pp.

Jeffrey, S. W., Mantoura, R. F. C. & Bjørnland, T. (1997). Data for the identification of 47 key phytoplankton pigments, in: Jeffrey, S.W. et al. (Ed.) Phytoplankton pigments in oceanography: guidelines to modern methods. Monographs on Oceanographic Methodology, 10, 449-559.

Jessen, S. P., Rasmussen, T. L., Nielsen, T., & Solheim, A. (2010). A new Late Weichselian and Holocene marine chronology for the western Svalbard slope 30,000-0 cal years BP. Quaternary Science Reviews, 29(9–10), 1301–1312. https://doi.org/10.1016/j.quascirev.2010.02.020

Jiang, S., Liu, X., Sun, J., Yuan, L., Sun, L., & Wang, Y. (2011). A multi-proxy sediment record of late Holocene and recent climate change from a lake near Ny-Ålesund, Svalbard. Boreas, 40(3), 468– 480. https://doi.org/10.1111/j.1502-3885.2010.00198.x

Juggins, S. (2017). rioja: Analysis of Quaternary Science Data, R package version (0.9-21). (http://cran.r-project.org/package=rioja).

Kaplan, M. R., Wolfe, A. P., & Miller, G. H. (2002). Holocene environmental variability in southern Greenland inferred from lake sediments. Quaternary Research, 58(2), 149–159. https://doi.org/10.1006/qres.2002.2352

Kern, R., Hotter, V., Frossard, A., Albrecht, M., Baum, C., Tytgat, B., De Maeyer, L., Velazquez, D., Seppey, C., Frey, B., Plötze, M., Verleyen, E., Quesada, A., Svenning, M. M., Glaser, K., & Karsten, U. (2019). Comparative vegetation survey with focus on cryptogamic covers in the high Arctic along two differing catenas. Polar Biology, 42(11), 2131–2145. https://doi.org/10.1007/s00300- 019-02588-z

Larsen, E., Lyså, A., Rubensdotter, L., Farnsworth, W. R., Jensen, M., Nadeau, M. J., & Ottesen, D. (2018). Lateglacial and Holocene glacier activity in the Van Mijenfjorden area, western Svalbard. Arktos, 4(1), 0. https://doi.org/10.1007/s41063-018-0042-2

Laspoumaderes, C., Modenutti, B., Souza, M. S., Navarro, M. B., Cuassolo, F. & Balseiro, E. (2013). Glacier melting and stoichiometric implications for lake community structure: zooplankton species distributions across a natural light gradient. Global Change Biology, 19, 316-326. https://doi.org/10.1111/gcb.12040

Leavitt, P. R. & Hodgson, D. A. (2001). Sedimentary pigments. In Tracking Environmental Change Using Lake Sediments. Vol. 3. Terrestrial, Algal and Siliceous Indicators (Smol, J. P., Birks, H. J. B. & Last, W. M., eds), 295-325. Kluwer Academic Publishers, Dordrecht. https://doi.org/10.1007/0-306-47668-1_15

Mangerud, J., & Svendsen, J. I. (2018). The Holocene Thermal Maximum around Svalbard, Arctic North Atlantic; molluscs show early and exceptional warmth. Holocene, 28(1), 65–83. https://doi.org/10.1177/0959683617715701

McMurdie, P. J., & Holmes, S. (2013). Phyloseq: An R Package for Reproducible Interactive Analysis and Graphics of Microbiome Census Data. PLoS ONE, 8(4). https://doi.org/10.1371/journal.pone.0061217

Myers-Smith, I. H., Elmendorf, S. C., Beck, P. S. A., Wilmking, M., Hallinger, M., Blok, D., Tape, K. D., Rayback, S. A., Macias-Fauria, M., Forbes, B. C., Speed, J. D. M., Boulanger-Lapointe, N., Rixen, C., Lévesque, E., Schmidt, N. M., Baittinger, C., Trant, A. J., Hermanutz, L., Collier, L. S., … Vellend, M. (2015). Climate sensitivity of shrub growth across the tundra biome. Nature Climate Change, 5(9), 887–891. https://doi.org/10.1038/nclimate2697

Nakatsubo, T., Bekku, Y. S., Uchida, M., Muraoka, H., Kume, A., Ohtsuka, T., Masuzawa, T., Kanda, H., & Koizumi, H. (2005). Ecosystem development and carbon cycle on a glacier foreland in the high Arctic, Ny-Ålesund, Svalbard. Journal of Plant Research, 118(3), 173–179. https://doi.org/10.1007/s10265-005-0211-9

Palacios, Y. M., Vonshak, A. & Beardall, J. (2018). Photosynthetic and growth responses of Nannochloropsis oculate (Eustigmatophyceae) during batch cultures in relation to light intensity. Phycologia, 492-502. https://doi.org/10.2216/17-100.1

Parducci, L., Bennett, K. D., Ficetola, G. F., Alsos, I. G., Suyama, Y., Wood, J. R., & Pedersen, M. W. (2017). Ancient plant DNA in lake sediments. New Phytologist, 214(3), 924–942. https://doi.org/10.1111/nph.14470

Parducci, L., Matetovici, I., Fontana, S. L., Bennett, K. D., Suyama, Y., Haile, J., Kjær, K. H., Larsen, N. K., Drouzas, A. D., & Willerslev, E. (2013). Molecular- and pollen-based vegetation analysis in lake sediments from central Scandinavia. Molecular Ecology, 22(13), 3511–3524. https://doi.org/10.1111/mec.12298

Pawlowski, J., Audic, S., Adl, S., Bass, D., Belbahri, L., Berney, C., Bowser, S. S., Cepicka, I., Decelle, J., Dunthorn, M., Fiore-Donno, A. M., Gile, G. H., Holzmann, M., Jahn, R., Jirků, M., Keeling, P. J., Kostka, M., Kudryavtsev, A., Lara, E., … de Vargas, C. (2012). CBOL Protist Working Group:

Barcoding Eukaryotic Richness beyond the Animal, Plant, and Fungal Kingdoms. PLoS Biology, 10(11). https://doi.org/10.1371/journal.pbio.1001419

Pawlowski, J., & Lecroq, B. (2010). Short rDNA barcodes for species identification in foraminifera. Journal of Eukaryotic Microbiology, 57(2), 197–205. https://doi.org/10.1111/j.1550- 7408.2009.00468.x

Pearson, R. G., Phillips, S. J., Loranty, M. M., Beck, P. S. A., Damoulas, T., Knight, S. J., & Goetz, S. J. (2013). Shifts in Arctic vegetation and associated feedbacks under climate change. Nature Climate Change, 3(7), 673–677. https://doi.org/10.1038/nclimate1858

Peros, M. C., & Gajewski, K. (2009). Pollen-based reconstructions of late Holocene climate from the central and western Canadian Arctic. Journal of Paleolimnology, 41(1), 161–175. https://doi.org/10.1007/s10933-008-9256-9

Pessi, I. S., Pushkareva, E., Lara, Y., Borderie, F., Wilmotte, A., & Elster, J. (2019). Marked Succession of Cyanobacterial Communities Following Glacier Retreat in the High Arctic. Microbial Ecology, 77(1), 136–147. https://doi.org/10.1007/s00248-018-1203-3

Pointing, S. B., & Belnap, J. (2012). Microbial colonization and controls in dryland systems. Nature Reviews Microbiology, 10(8), 551–562. https://doi.org/10.1038/nrmicro2831

R Core Team. (2019). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/.

Rasmussen, T. L., Thomsen, E., Ślubowska, M. A., Jessen, S., Solheim, A., & Koç, N. (2007). Paleoceanographic evolution of the SW Svalbard margin (76°N) since 20,000 14C yr BP. Quaternary Research, 67(1), 100–114. https://doi.org/10.1016/j.yqres.2006.07.002

Rose, N. L., Rose, C. L., Boyle, J. F., & Appleby, P. G. (2004). Lake-sediment evidence for local and remote sources of atmospherically deposited pollutants on Svalbard. Journal of Paleolimnology, 31(4), 499–513. https://doi.org/10.1023/B:JOPL.0000022548.97476.39

Rühland, K. M., Paterson, A. M., & Smol, J. P. (2015). Lake diatom responses to warming: reviewing the evidence. Journal of Paleolimnology, 54(1), 1–35. https://doi.org/10.1007/s10933-015- 9837-3

Schloss, P. D., Westcott, S. L., Ryabin, T., Hall, J. R., Hartmann, M., Hollister, E. B., Lesniewski, R. A., Oakley, B. B., Parks, D. H., Robinson, C. J., Sahl, J. W., Stres, B., Thallinger, G. G., Van Horn, D. J., & Weber, C. F. (2009). Introducing mothur: Open-source, platform-independent, community- supported software for describing and comparing microbial communities. Applied and Environmental Microbiology, 75(23), 7537–7541. https://doi.org/10.1128/AEM.01541-09

Screen, J. A., & Simmonds, I. (2010). The central role of diminishing sea ice in recent Arctic temperature amplification. Nature, 464(7293), 1334–1337. https://doi.org/10.1038/nature09051

Shapiro, B., Barlow, A., Heintzman, P. D., Hofreiter, M., Paijmans, J. L. A. & Soares, A. E. R. (2012). Ancient DNA: methods and protocols. https://doi.org/10.1007/978-1-4939-9176-1

Smol, J. P., & Douglas, M. S. V. (2007). From controversy to consensus: Making the case for recent climate change in the Arctic using lake sediments. Frontiers in Ecology and the Environment, 5(9), 466–474. https://doi.org/10.1890/060162

Smol, J. P., Wolfe, A. P., Birks, H. J. B., Douglas, M. S. V., Jones, V. J., Korhola, A., Pienitz, R., Rühland, K., Sorvari, S., Antoniades, D., Brooks, S. J., Fallu, M. A., Hughes, M., Keatley, B. E., Laing, T. E., Michelutti, N., Nazarova, L., Nyman, M., Paterson, A. M., … Weckström, J. (2005). Climate- driven regime shifts in the biological communities of arctic lakes. Proceedings of the National Academy of Sciences of the United States of America, 102(12), 4397–4402. https://doi.org/10.1073/pnas.0500245102

Sønstebø, J. H., Gielly, L., Brysting, A. K., Elven, R., Edwards, M., Haile, J., Willerslev, E., Coissac, E., Rioux, D., Sannier, J., Taberlet, P., & Brochmann, C. (2010). Using next-generation sequencing for molecular reconstruction of past Arctic vegetation and climate. Molecular Ecology Resources, 10(6), 1009–1018. https://doi.org/10.1111/j.1755-0998.2010.02855.x

Stebel, A., Ochyra, R., Konstantinova, N. A., Ziaja, W., Ostafin, K., & Maciejowski, W. (2018). A contribution to the knowledge of bryophytes in polar areas subjected to rapid deglaciation: A case study from southeastern Spitsbergen. Acta Societatis Botanicorum Poloniae, 87(4), 1–26. https://doi.org/10.5586/asbp.3603

Stoeck, T., Bass, D., Nebel, M., Christen, R., Jones, M. D. M., Breiner, H. W., & Richards, T. A. (2010). Multiple marker parallel tag environmental DNA sequencing reveals a highly complex eukaryotic community in marine anoxic water. Molecular Ecology, 19(SUPPL. 1), 21–31. https://doi.org/10.1111/j.1365-294X.2009.04480.x

Støren, E. W. N., Brun Bjerkås, A., Bakke, J., Linge, H., D`Andrea, W., van der Bilt, W., Røthe, T., Balascio, N. L., Bradley, R. S., Grant, O., Fabel, D., & Xu, S. (2020). Reconstruction of Holocene glacier fluctuations at Kongsbreen based on sediments deposited in lake Sarsvatnet, Ossian Sarsfjellet, Svalbard. EGU General Assembly 2020. https://doi.org/10.5194/egusphere-egu2020- 19216

Swain, E. B. (1985). Measurement and interpretation of sedimentary pigments. Freshwater Biology, 15(1), 53–75. https://doi.org/10.1111/j.1365-2427.1985.tb00696.x

Taberlet, P., Coissac, E., Pompanon, F., Gielly, L., Miquel, C., Valentini, A., Vermat, T., Corthier, G., Brochmann, C., & Willerslev, E. (2007). Power and limitations of the chloroplast trnL (UAA) intron for plant DNA barcoding. Nucleic Acids Research, 35(3). https://doi.org/10.1093/nar/gkl938

Tavernier, I., Verleyen, E., Hodgson, D. A., Heirman, K., Roberts, S. J., Imura, S., Kudoh, S., Sabbe, K., De Batist, M., & Vyverman, W. (2014). Absence of a medieval climate anomaly, little ice age and twentieth century warming in skarvsnes, lützow holm bay, east antarctica. Antarctic Science, 26(5), 585–598. https://doi.org/10.1017/S0954102014000029

Tytgat, B., Nguyen, D. T., Nguyen, T. X. P., Pham, T. M., Long, P. K., Vanreusel, A., & Derycke, S. (2019). Monitoring of marine nematode communities through 18S rRNA metabarcoding as a sensitive alternative to morphology. Ecological Indicators, 107(July), 105554. https://doi.org/10.1016/j.ecolind.2019.105554

van der Bilt, W. G. M., Bakke, J., Vasskog, K., D’Andrea, W. J., Bradley, R. S., & Ólafsdóttir, S. (2015). Reconstruction of glacier variability from lake sediments reveals dynamic Holocene climate in Svalbard. Quaternary Science Reviews, 126, 201–218. https://doi.org/10.1016/j.quascirev.2015.09.003 van der Bilt, W. G. M., D’Andrea, W. J., Bakke, J., Balascio, N. L., Werner, J. P., Gjerde, M., & Bradley, R. S. (2018). Alkenone-based reconstructions reveal four-phase Holocene temperature evolution for High Arctic Svalbard. Quaternary Science Reviews, 183, 204–213. https://doi.org/10.1016/j.quascirev.2016.10.006 van der Bilt, W. G. M., D’Andrea, W. J., Werner, J. P., & Bakke, J. (2019). Early Holocene Temperature Oscillations Exceed Amplitude of Observed and Projected Warming in Svalbard Lakes. Geophysical Research Letters, 46(24), 14732–14741. https://doi.org/10.1029/2019GL084384

Van Heukelem, L., & Thomas, C. S. (2001). Computer-assisted high-performance liquid chromatography method development with applications to the isolation and analysis of phytoplankton pigments. Journal of Chromatography A, 910(1), 31–49. https://doi.org/10.1016/S0378-4347(00)00603-4 van Woerkom, A. (2016). Ancient DNA from soils and sediments from the Lake Krigstjärn area , northern Sweden. November. https://www.diva- portal.org/smash/record.jsf?pid=diva2:1047299

Voldstad, L. H., Alsos, I. G., Farnsworth, W. R., Heintzman, P. D., Håkansson, L., Kjellman, S. E., Rouillard, A., Schomacker, A., & Eidesen, P. B. (2020). A complete Holocene lake sediment ancient DNA record reveals long-standing high Arctic plant diversity hotspot in northern Svalbard. Quaternary Science Reviews, 234. https://doi.org/10.1016/j.quascirev.2020.106207

Willerslev, E., & Cooper, A. (2005). Ancient DNA. Proceedings of the Royal Society B: Biological Sciences, 272(1558), 3–16. https://doi.org/10.1098/rspb.2004.2813

Willerslev, E., Davison, J., Moora, M., Zobel, M., Coissac, E., Edwards, M. E., Lorenzen, E. D., Vestergård, M., Gussarova, G., Haile, J., Craine, J., Gielly, L., Boessenkool, S., Epp, L. S., Pearman, P. B., Cheddadi, R., Murray, D., Bråthen, K. A., Yoccoz, N., … Taberlet, P. (2014). Fifty thousand years of Arctic vegetation and megafaunal diet. Nature, 506(7486), 47–51. https://doi.org/10.1038/nature12921

Willerslev, E., Hansen, A. J., Binladen, J., Brand, T. B., Gilbert, M. T. P., Shapiro, B., Bunce, M., Wiuf, C., Gilichinsky, D. A., & Cooper, A. (2003). Diverse plant and animal genetic records from holocene and pleistocene sediments. Science, 300(5620), 791–795. https://doi.org/10.1126/science.1084114

Willerslev, E., Hansen, A. J., & Poinar, H. N. (2004). Isolation of nucleic acids and cultures from fossil ice and permafrost. Trends in Ecology and Evolution, 19(3), 141–147. https://doi.org/10.1016/j.tree.2003.11.010

Yoshitake, S., Uchida, M., Iimura, Y., Ohtsuka, T., & Nakatsubo, T. (2018). Soil microbial succession along a chronosequence on a High Arctic glacier foreland, Ny-Ålesund, Svalbard: 10 years’ change. Polar Science, 16(March), 59–67. https://doi.org/10.1016/j.polar.2018.03.003

Zhang, H., Piilo, S. R., Amesbury, M. J., Charman, D. J., Gallego-Sala, A. V., & Väliranta, M. M. (2018). The role of climate change in regulating Arctic permafrost peatland hydrological and vegetation change over the last millennium. Quaternary Science Reviews, 182, 121–130. https://doi.org/10.1016/j.quascirev.2018.01.003

Zhang, J., Kobert, K., Flouri, T., & Stamatakis, A. (2014). PEAR: A fast and accurate Illumina Paired-End reAd mergeR. Bioinformatics, 30(5), 614–620. https://doi.org/10.1093/bioinformatics/btt593

Zhu, F., Massana, R., Not, F., Marie, D., & Vaulot, D. (2005). Mapping of picoeucaryotes in marine ecosystems with quantitative PCR of the 18S rRNA gene. FEMS Microbiology Ecology, 52(1), 79– 92. https://doi.org/10.1016/j.femsec.2004.10.006

10 Appendix

Appendix 1: Pictures of dried sediment and plant material used for radiocarbon dating. Pictures in column 1 are from core 1, pictures in column 2 from core 2.

C1_10 C2_04

C2_04

C1_19 C2_12

C2_23

C2_34

Appendix 2: Ancient DNA room preparation and core subsampling

Standard Operating Protocol for the setup of a (temporary) ancient DNA laboratory, subsampling of a sediment core and extraction of ancient DNA.

Authors: Decorte L., De Maeyer L., Tytgat B. & Verleyen E. A. Goal

The purpose of this protocol is describing how to set up an ancient DNA laboratory for subsampling a lake sediment core, DNA and RNA extractions and PCR, with procedures, materials and devices that can be used.

DNA that can be found in lake sediment that is several hundreds or thousands years old is generally degraded to some extent. The DNA strands accumulate damage over the years, which results in highly fragmented DNA containing a variety of chemical modifications. This makes this ancient DNA vulnerable for contamination with non-degraded DNA from the present-day environment. Even a very low amount of contamination can have an effect, as modern DNA will be preferred to be amplified in a PCR reaction over damaged ancient DNA. Contamination can happen at multiple stages during the processing of the ancient DNA. It can happen during the core retrieval, where modern environmental DNA can contaminate the bottom of the core while the corer is pushed into the sediment. Contaminating DNA can also be introduced by the researchers during core processing in the laboratory. Also DNA that has been previously amplified and that is present in the laboratory environment can be a source of contamination. It is therefore essential to follow a detailed protocol which tackles all procedures and equipment that should be used. B. Setup of a (temporary) ancient DNA laboratory

There is no permanent room present at Ghent University for working with ancient DNA. The processing of the sediment core and extracting of ancient DNA should take place in a room that is strictly separated from subsequent steps like PCR, preferably even in a separate building. During the whole period of processing the ancient DNA, cleaning staff and other lab users were not allowed to enter the room.

1. Preparation of the ancient DNA room a. Label the door (“aDNA room, forbidden to enter”) to make clear the ancient DNA room shouldn’t be entered. b. Make sure all devices that are needed are in the room. These devices are: i. Biohazard ii. Freezer iii. Centrifuge c. List of materials and reagents i. General supplies 1. Ziplocs 2. Whirlpacks 3. Aluminium foil 4. DNA away (Thermo scientific) 5. RNase away (Thermo scientific) 6. Ethanol 70% 7. Milli-Q water

8. Bleach 0.1% 9. Tyvek suits (DuPont) 10. Hairnets 11. Face masks 12. Gloves 13. Markers 14. Falcons 15. Little jars 16. Tork PaperCircle® ii. Specific for subsampling 1. Microtome blade 2. Aluminium/plastic bins 3. Racks 4. Core holder 5. Hacksaw with multiple sawblades 6. Tweezers/spoons/spatulas… 7. Scale d. Place a mop drenched in bleach 0.1% at the entrance of the room. 2. Disinfection2 of devices and the ancient DNA room a. Biohazard i. Disinfect the inside and outside of the biohazard. Make sure there are no sediment residuals left from previous subsampling procedures. ii. Put on the UV lamp for some time before starting or overnight. b. Freezer and centrifuge i. Disinfect both inside and outside. c. Clean all surfaces, working benches, window sills, chairs, door handle, light switch and floor with bleach 0.1%. d. Disinfect working surfaces. e. Thoroughly clean the room and Biohazard in the same way every week. C. Preparation and procedures before starting

Every time the ancient DNA room is entered, a procedure should be followed to be sure a sterile environment can be maintained. Negative controls should be placed and replaced regularly. When these preparations and procedures are done, the actual work can be started.

1. Procedure when entering the ancient DNA room a. Take off shoes. b. Wash hands with soap. c. Put on gloves, sterilise them with ethanol gel, put on second pair of gloves, sterilise them again with ethanol gel. d. Enter the ancient DNA room and put on slippers (that never leave the room), hairnet and face mask. e. Wipe your flipflops over the mop. f. Open the package with the Tyvek suit. g. Remove outer pair of gloves, put on new ones and disinfect them.

2 Disinfection means applying bleach 0.1%, ethanol 70%, DNA away (Thermo Scientific) and RNase away (Thermo Scientific), unless described otherwise. 43

h. Remove the Tyvek suit from its package and put on the suit without touching the outside of the suit with your clothes. i. Change and disinfect gloves frequently during the day. 2. Preparation a. Filter-sterilise autoclaved milli-Q water through a 0.22 µm filter. b. Put small jars with autoclaved, filter-sterilised water as negative controls at strategic places in the ancient DNA room: working benches, entrance of the room,… c. Place 3 negative controls in the Biohazard: one halfway each wall. Replace these negative controls every day and between different processes that happen on the same day. For example, if you remove the mantle of the core in the morning and start subsampling afterwards, it is advised to replace the negative controls in order to keep track during what stage the possible contamination has occurred. d. Disinfect inside of Biohazard and all surfaces that will be used that day. D. Subsampling core

1. Material preparation a. Sterilise aluminium foil, tweezers, spoons and spatulas with heat at 140 °C for 4h. b. Disinfect microtome blade, aluminium bins, racks, core holder, hacksaw and sawblades. 2. Removing the mantle of the core a. Place an aluminium/plastic bin with a rack on it in the Biohazard. b. Place the core on the rack at a 45° angle with the oldest part of the core (bottom) on top. c. A second person removes the outer ± 5 mm from top to bottom in zones of ± 10 cm. Disinfect the microtome blade each time before moving to the next zone. d. When the outer surface is removed, the second person removes the rack, puts on new gloves and disinfects them, and places a new rack. Place the core on this clean rack. e. Pour filter-sterilised 70% ethanol over the core and immediately rinse with autoclaved, filter-sterilised water. f. Spray the core with DNA away, rinse with autoclaved, filter-sterilised water, spray the core with RNase away and again rinse with autoclaved, filter-sterilised water. g. Wrap the core in aluminium foil, mark bottom and top of the core, place in a Ziploc and in the freezer. 3. Subsampling the core a. Place an aluminium bin with a rack on it and the core holder on it in the Biohazard. b. Take the core out of the freezer and place it in the core holder with the oldest past of the core (bottom) sticking out of the holder. c. Start sawing a slice of 1 cm thickness, make sure the sawdust lands in the aluminium bin. d. Let the slice fall on aluminium foil and wrap the slice in it. Label and put it in a whirlpack and in the freezer. e. Take another (clean) bin, rack, sawblade and core holder. f. Saw a second slice in the same way. g. Wrap the end of the core in aluminium foil, then the whole core in another aluminium foil and place it in the freezer.

h. Disinfect the inside of the Biohazard, aluminium bins, racks, sawblades and core holders. i. Repeat this procedure of sawing for every two slices to make sure the core doesn’t thaw completely during sawing. 4. Subsampling core slices Depending on what the purposes of the study are, the slices have to be divided in different pieces for every analysis. In this case, the core will be used for ancient DNA, GDGT and pigment analysis (samples need to be freeze dried for both GDGT and pigment analysis, so these can be sampled together) and radiocarbon dating. a. Take 10 slices out of the freezer and let them thaw (again start with the bottom of the core). b. Weigh 10 whirlpacks in the meantime. c. Take each sample out of the aluminium foil, put it in a new whirlpack and weigh it. d. Homogenize the samples by kneading it. e. Sampling for ancient DNA extraction i. Take approximately 2 g of sediment out of the whirlpack and put it in a falcon with beads (the falcon that is needed for DNA extraction). f. Sampling for GDGT and pigment analysis i. Take approximately 4 g of sediment out of the whirlpack and put it in a falcon. ii. Cover the falcon with aluminium foil. g. Sampling for radiocarbon dating i. Check which samples are needed for radiocarbon dating. ii. Take approximately 2 g of sediment out of the whirlpack and put it in a 2 ml Eppendorf tube. iii. If there are macrofossils present in the sediment core, take some and put them in another Eppendorf tube. h. Put the whirlpack with the remaining sediment in a Ziploc and in the freezer. i. Put the subsamples in the freezer. j. Repeat this procedure for every 10 samples. E. RNA and DNA extraction

During RNA and DNA extraction, the same preparations and procedures should be executed before starting. A protocol can be followed to extract the RNA and DNA. In this thesis, the RNeasy PowerSoil Total RNA Kit (Qiagen) and RNeasy PowerSoil DNA Elution Kit (Qiagen) were used.

Appendix 3: Table with subsampled weight used for each analysis. Weight for pigment analysis is indicated in dry weight.

Sample aDNA Pigments Radiocarbon Sample aDNA Pigments Radiocarbon number (g) (g dry weight) dating (g) number (g) (g dry weight) dating (g) C1_1 2.03 0.2647 C2_1 2.01 0.3414

C1_2 2.02 0.2602 C2_2 1.99 0.2344

C1_3 2.04 0.4759 C2_3 2 0.3684

C1_4 2.01 0.2378 C2_4 2 0.4971 2.28

C1_5 2 0.3659 C2_5 2.02 0.3312

C1_6 0.95 0.2952 C2_6 2 0.4612

C1_7 2 0.5341 C2_7 2 0.4463

C1_8 1.99 0.39 C2_8 2.02 0.5429

C1_9 2.04 0.4289 C2_9 2.05 0.393

C1_10 2 0.4228 2.3 C2_10 2.04 0.2856

C1_11 1.98 0.5696 C2_11 2 0.5

C1_12 2 0.6595 C2_12 2.02 0.5257 2.38

C1_13 2.02 0.5729 C2_13 2.09 0.5369

C1_14 2.01 0.5068 C2_14 2.05 0.5939

C1_15 2.04 0.304 C2_15 1.99 0.6031

C1_16 2 0.4006 C2_16 2.01 0.4936

C1_17 2 0.7186 C2_17 2.03 0.6157

C1_18 2.02 0.6682 C2_18 2.07 0.5959

C1_19 2 0.8597 2.03 C2_19 1.99 0.5714

C1_20 2.04 0.5904 C2_20 2.09 0.6404

C1_21 1.01 0.5721 C2_21 2.02 0.5245

C2_22 2.04 0.6135

C2_23 2.03 0.6299 2.49

C2_24 2.03 0.7832

C2_25 2.01 0.6628

C2_26 2 0.6036

C2_27 2.05 0.6379

C2_28 2.02 0.7492

C2_29 2.02 0.885

C2_30 2.05 1.1793

C2_31 2.01 1.5449

C2_32 2.06 2.1862

C2_33 2.15 2.2136

C2_34 2.04 2.1873 3.92

C2_35 2.04 2.405

C2_36 2 2.2149

C2_37 2.03 1.3313

Appendix 4: RNeasy PowerSoil Total RNA Kit Protocol: Detailed

Important points before starting

- If Solution IRS has precipitated, heat at 60°C until precipitate dissolves. - Perform all centrifugation steps at room temperature (15–25°C). - Wear RNase-free gloves at all times and remove RNase from the work area.

Procedure

1. Add up to 2 g of soil to the 15 ml PowerBead Tube (provided). Please refer to the Troubleshooting Guide for information regarding the amount of soil to process. 2. Add 2.5 ml of PowerBead Solution, 0.25 ml of Solution SR1 and 0.8 ml of Solution IRS. Note: The PowerBead Solution is a buffer that disperses cells and soil particles. Solution SR1 contains SDS and other disruption agents which aid in complete cell lysis. In addition to aiding in cell lysis, SDS is an anionic detergent that breaks down fatty acids and lipids associated with the cell membrane of several organisms. Solution IRS is a precipitation reagent that removes non-DNA organic and inorganic material including cell debris and proteins. Note: If it gets cold, Solution SR1 will form a white precipitate. Heating to 60oC will dissolve the SDS and will not harm the other disruption agents. 3. Add 3.5 ml of phenol/chloroform/isoamyl alcohol (pH 6.5–8.0, [User supplied]). Cap and vortex the PowerBead Tube to mix until the biphasic layer disappears. Note: Phenol/chloroform/isoamyl alcohol maximizes lysing efficiency and yield. Lysed cell components are trapped in the solvent and proteins are denatured leaving the nucleic acid in solution. 4. Place the PowerBead Tube on a Vortex Adapter (cat. no. 13000-V1-15) and vortex at maximum speed for 15 min. Note: Cells are lysed by a combination of chemical agents from steps 1-3 and mechanical shaking introduced by vortexing. Use of the vortex adapter will maximize homogenization, which can lead to higher DNA yields. Using tape to attach tubes is not recommended. 5. Remove the PowerBead Tube and centrifuge at 2500 x g for 10 min. Note: Centrifugation results in phase separation of the sample mixture. Three phases will be visible after centrifugation: the lower organic phase containing proteins and cellular debris, the interphase containing humics and other organic and non-organic material, and the upper aqueous phase containing total nucleic acid. Note: The thickness of the interphase will depend on the sample type. Samples high in organic content will have a thicker interphase. 6. Transfer the upper aqueous phase (avoid the interphase and lower phenol layer) to a clean 15 ml Collection Tube (provided). Discard the phenol/chloroform/isoamyl alcohol. Note: The biphasic layer will be thick and firm in soils high in organic matter and may need to be pierced to remove the bottom phenol layer. The upper aqueous phase containing total nucleic acids from the sample is transferred to a new tube. Cellular debris, proteins and organic material are left behind. Take care not to transfer material from the lower phase or the interphase. 7. Add 1.5 ml of Solution SR3 to the aqueous phase and vortex to mix. Incubate at 2–8°C for 10 min and then centrifuge at 2,500 x g for 10 min at room temperature. Note: Solution SR3 is a secondary precipitation step to further remove proteins and cellular debris. 8. Transfer the supernatant, without disturbing the pellet (if there is one), to a new 15 ml Collection Tube (provided). Note: The supernatant containing nucleic acids are transferred to a new 15 ml tube. Non-nucleic acid material is left behind. 47

9. Add 5 ml of Solution SR4 to the supernatant in the Collection Tube and invert or vortex to mix. Incubate at room temperature for 30 min. Note: Previous protocol instructions were to incubate at –20°C. If you have achieved good results for your soil type using the previous protocol, you may continue to follow it. 10. Centrifuge at 2500 x g for 30 min. 11. Decant the supernatant and invert the 15 ml Collection Tube on a paper towel for 5 min. Note: Solution SR4 is 100% isopropanol. Nucleic acid is precipitated and the isopropanol is discarded. 12. Shake Solution SR5 to mix and add 1 ml to the 15 ml Collection Tube. Resuspend the pellet completely by repeatedly pipetting or vortexing. Note: If the pellet is difficult to resuspend, place the tube in a heat block or water bath at 45°C for 10 min, followed by vortexing. Repeat until the pellet is resuspended. Solution SR5 is a proprietary salt solution used to resuspend the precipitated nucleic acids from Step 11. It is also used to equilibrate the JetStar Mini Column in Step 13 and to wash and prep the column for RNA elution in Step 15. 13. Prepare one JetStar Mini Column (provided) for each RNA isolation sample: a. Remove the cap of a 15 ml Collection Tube (provided) and place the JetStar Mini Column inside it. The column will hang in the Collection Tube. b. Add 2 ml of Solution SR5 to the JetStar Mini Column. Allow it to completely gravity flow through the column and collect in the 15 ml Collection Tube. Note: Do not allow the column to dry out before loading the RNA isolation sample. 14. Add the RNA isolation sample from Step 12 onto the JetStar Mini Column and allow it to gravity flow through the column into the 15 ml Collection Tube. 15. Add 1 ml of Solution SR5 to the JetStar Mini Column and allow it to completely gravity flow into the 15 ml Collection Tube. Note: The sample is added to the JetStar Mini Column and the nucleic acids are bound to the column matrix. The Capture Column is then washed with a second volume of Solution SR5 to ensure unbound contaminants are removed from the sample and column prior to the elution of RNA. 16. Transfer the JetStar Mini Column to a new 15 ml Collection Tube (provided). Shake Solution SR6 to mix and then add 1 ml to the JetStar Mini Column to elute the bound RNA. Allow Solution SR6 to gravity flow into the 15 ml Collection Tube. Note: The Solution SR6 is a proprietary salt solution that allows for the preferential release of RNA from the JetStar Mini Column, leaving DNA, residual debris and inhibiting substances in the column. 17. Transfer the eluted RNA to a 2.2 ml Collection Tube (provided). Add 1 ml of Solution SR4. Invert at least once to mix and incubate at –15°C to –30°C for a minimum of 10 min. 18. Centrifuge the 2.2 ml Collection Tube at 13,000 x g for 15 min to pellet the RNA. 19. Decant the supernatant and invert the 2.2 ml Collection Tube onto a paper towel for 10 min to air dry the pellet. Note: Solution SR4 is 100% isopropanol. Eluted RNA from the Capture Column is precipitated, centrifuged and allowed to air dry before it is resuspended and concentrated. 20. Resuspend the RNA pellet in 100 μl of Solution SR7. The RNA is now ready for downstream applications. Note: Solution SR7 is RNase/DNase-free water used to resuspend the pelleted RNA. Solution SR7 contains no EDTA. For long term storage of samples, 10 mM Tris (pH 8.0) or TE buffer may be used to resuspend the pelleted RNA.

Appendix 5: RNeasy PowerSoil DNA Elution Kit Protocol: Detailed

Notes before starting

- Wear RNase- and DNase-free gloves at all times. - Remove RNases and DNases from work surfaces before starting.

Procedure

1. Transfer the RNA Capture Column from step 16 of the RNeasy PowerSoil Total RNA Kit (cat. no. 12866-25) to a 15 ml Collection Tube (provided). 2. Add 1 ml of Solution SR8 to the RNA Capture Column to elute the bound DNA into the 15 ml Collection Tube. Allow Solution SR8 to gravity flow into the Collection Tube. Note: Solution SR8 is a salt solution that allows for the preferential release of DNA from the RNA Capture Column, leaving residual debris and inhibiting substances in the column. 3. Transfer the eluted DNA to a 2.2 ml Collection Tube (provided) and add 1 ml of Solution SR4. Invert at least once to mix and incubate at –15°C to –30°C for 10 min. 4. Centrifuge the 2.2 ml Collection Tube at 13,000 x g for 15 min at room temperature to pellet the DNA. 5. Decant the supernatant and invert the 2.2 ml Collection Tube onto a paper towel for 10 min to air dry the DNA pellet. Note: Solution SR4 is 100% Isopropanol. DNA eluted from the RNA Capture Column is precipitated, centrifuged and allowed to air dry before resuspending and concentrating. 6. Resuspend the DNA pellet in 100 μl of Solution SR7. Note: Solution SR7 is RNase/DNase-free water and is used to resuspend the pelleted DNA. Solution SR7 contains no EDTA. The eluted DNA is now ready for downstream applications. For long term storage of samples, 10 mM Tris (pH 8.0) or TE buffer may be used to resuspend the pelleted DNA. Note: Although RNA carryover does not occur with most soil types, certain soils high in organic matter may present unique carryover situations. When the absence of RNA contamination is critical, an RNase treatment of the isolated DNA is recommended; please refer to the Troubleshooting Guide for instructions.

Appendix 6: DNeasy PowerLyzer Microbial Kit

Important points before starting

• The PowerLyzer 24 Homogenizer may cause marring of labels on the tops of the PowerBead Tubes. To ensure proper sample identification, label sides and tops of the tubes. • If Solution SL has precipitated, heat at 60°C until the precipitate has dissolved. • Shake to mix Solution SB before use.

Procedure

1. Add 1.8 ml of microbial (bacteria, yeast) culture to a 2 ml Collection Tube (provided) and centrifuge at 10,000 x g for 30 s at room temperature. Decant the supernatant and spin the tubes at 10,000 x g for 30 s at room temperature and completely remove the media supernatant with a pipet tip. Note: Based on the type of microbial culture, it may be necessary to centrifuge longer than 30 seconds. 2. Resuspend the cell pellet in 300 μl of PowerBead Solution and gently vortex to mix. Transfer resuspended cells to a PowerBead Tube Glass, 0.1 mm. 3. Add 50 μl of Solution SL to the PowerBead Tube. Note: To increase yields, to minimize DNA shearing or for cells that are difficult to lyse, refer to the Troubleshooting Guide. 4. Homogenization options: a) PowerLyzer 24 Homogenizer: Balance PowerBead Tubes in the tube holder for the PowerLyzer 24. Homogenize for 5 min at 2000 RPM. Note: Depending on the sample, you can homogenize at a higher speed for less time. b) Vortex: Secure PowerBead Tube horizontally using the Vortex Adapter tube holder (cat. no. 13000-V1-24). Vortex at maximum speed for 10 min. Note: To minimize DNA shearing, refer to the Troubleshooting Guide. 12 DNeasy PowerLyzer Microbial Kit Handbook 06/2017 5. Make sure the PowerBead Tubes rotate freely in the centrifuge without rubbing. Centrifuge the tubes at a maximum of 10,000 x g for 30 s at room temperature. 6. Transfer the supernatant to a clean 2 ml Collection Tube (provided). Note: Expect 300 to 350 μl of supernatant. 7. Add 100 μl of Solution IRS and vortex for 5 s. Incubate at 4°C for 5 min. 8. Centrifuge 10,000 x g for 1 min at room temperature. 9. Avoiding the pellet, transfer all of the supernatant to a 2 ml Collection Tube (provided). Note: Expect approximately 450 μl of supernatant. A small carryover of glass beads is possible. This will not affect the results. 10. Add 900 μl of Solution SB to the supernatant and vortex for 5 s. 11. Load about 700 μl into a MB Spin Column and centrifuge at 10,000 x g for 30 s at room temperature. Discard the flow-through, add the remaining supernatant to the MB Spin Column, and centrifuge at 10,000 x g for 30 s at room temperature. Note: Each sample processed will require 2–3 loads. Discard all flow-through. 12. Add 300 μl of Solution CB and centrifuge at 10,000 x g for 30 s at room temperature. 13. Discard the flow-through and centrifuge at 10,000 x g for 1min at room temperature. 14. Being careful not to splash liquid on the spin filter basket, place MB Spin Column in a new 2 ml Collection Tube(provided). 15. Add 50 μl of Solution EB to the centre of the white filter membrane. 16. Centrifuge at 10,000 x g for 30 s at room temperature.

17. Discard the MB Spin Column. The DNA is now ready for downstream applications. Note: We recommend storing DNA frozen (–20°C to –80°C) as Solution EB does not contain EDTA. To concentrate DNA, see the Troubleshooting Guide.

Appendix 7: Classification of OTUs with a high abundance in the aDNA analysis based on BLAST results using the NCBI database. Query cover was 100% for each classification.

OTU Id Name on plot PR2 database result Blast result Score (per ident) OTU_37 Criconemoides sp.* Chromadorea_uncl.(100) Criconemoides annulatus/ C. 99,28%/ 98,8% myungsugae OTU_18 Gieysztoria sp.* Rhabdocoela_uncl.(99) Gieysztoria sp. 94,90% OTU_124 Dinophyceae uncl. Dinophyceae_uncl.(100) Warnowia/ Gymnodinium 95,94%/ 95,94% otu_124* OTU_20 Dinophyceae uncl. Dinophyceae_uncl.(100) Warnowia/ Gymnodinium 96,66%/ 96,66% otu_20* OTU_388 Dinophyceae uncl. Dinophyceae_uncl.(100) Heterocapsa rotundata/ 96,66%/ 96,66% otu_388* Gyrodinium OTU_293 Woloszynskia cincta* Dinophyceae_uncl.(100) Woloszynskia cincta 99,76% OTU_41 Syndedra Araphid-pennate_uncl.(100) Syndedra berolinensis 99,05% berolinensis* OTU_4 Pseudotetraedriella Eustigmatophyceae_uncl.(100) Pseudotetraedriella kamillae 96,94% kamillae* OTU_25 Chlorodendraceae Chlorodendraceae_uncl.(100) Scherffelia dubia/ Tetraselmis sp. 99,52%/ 98,33% uncl.* OTU_38 Chlamydomonadales Chlamydomonadales_uncl.(91) Tetracystis sarcinalis/ 97,35%/ 97,11% uncl.* Vitreochlamys nekrassovii OTU_16 Oophila Chlamydomonadales_uncl.(100) Oophila amblystomatis 98,56% amblystomatis* OTU_10 Pleurastrum sp.* Pleurastrum_sp.(84) Pleurastrum/ Macrochloris/ 99,52%/ 99,52%/ Chlorococcales 99,52% OTU_7 Carteria sp.* Chlorophyceae_uncl.(80) Carteria cerasiformis/ C. 97,36%/ 97,36% eugametos OTU_23 Chlorophyceae uncl.* Chlorophyceae_uncl.(96) Carteria sp./ Hormotilopsis 94,51%/ 94,48% gelatinosa OTU_32 Tetracystis sp.* Chlamydomonadales_uncl.(100) Tetracystis texensis/ T. 99,28%/ 99,28% excentrica OTU_21 Cosmarium Zygnemophyceae_uncl.(100) Cosmarium granatum 99,52% granatum* OTU_36 Mitella sp.* Embryophyceae_uncl.(100) Mitella sp. 97,61%