Here, we provide the raw plant DNA metabarcoding data archived in Lake Naleng sediment core spanning the late-glacial to the present (17,700 to 0 cal yr BP).

Lake Naleng (31.10° N, 99.7°5 E, 4200 above sea level) is situated in a glacially-formed basin on the southeastern Tibetan Plateau. It is an alpine freshwater lake with a maximum water depth of 36.7 m, an open surface area of 1.7 km²,  a specific conductivity of 0.045 mS/cm, a Secchi depth of 2.9 m, a pH of 8.1, a dissolved oxygen content of 6.9 mg/L (measured 19.09.2003), and a catchment area of 128 km². The catchment is used for seasonal grazing by yaks and sheep. The lake is located at the upper treeline with main vegetation of alpine meadow. The detailed geographic, biogeochemical and vegetation information are described in Kramer et al., 2010a, b, c (doi: 10.1007/s00334-009-0219-5, 10.1016/j.yqres.2009.12.003, and 10.1016/j.palaeo.2009.12.001) and Opitz et al., 2015 (doi: http://dx.doi.org/10.1016/j.quascirev.2015.03.011).

The lake sediment cores (17.81 m) was recovered from water depth of 32 m using an UWITEC (Niederreiter 60) piston corer system on the frozen lake surface in February 2004 (Kramer et al., 2010a).  Due to lack of macrofossil remains, ten bulk organic carbon samples were selected for AMS (accelerator mass spectrometry) radiocarbon dating at the Leibniz Laboratory, Kiel. The age-depth model was built after subtracting the lake reservoir effect of 1500 cal yr BP. The original ¹⁴C dating and chronology is given by Kramer et al., 2010a, b, c.

In this project, we inferred the plant richness based on DNA data and simulated glacial dynamics and habitat area within lake catchment. We synthesized these data with published palaeoenvironmental data  (temperature, soil development, and treeline location) to find the direct factor of plant richness variations over the past ~18,000 years. According to the inferred relationship, we predicted the plant richness change in the next 250 years (until 2300 C.E.).

1. Sedimentary ancient DNA isolation, polymerase chain reactions (PCR), and purification and pooling

Sedimentary ancient DNA (sedaDNA) was extracted from about 3–10 g of sediment using the PowerMax^® Soil DNA Isolation Kit (MoBio Laboratories, Inc., USA) with a modifed instructions in a dedicated ancient DNA laboratory.  The plant sedaDNA was amplified using general plant gh primer targeting the P6 loop (10-143 bp) of chloroplast trnL (Taberlet et al., 2007).  Each PCR reaction (25 µL) were prepared in ancient DNA laboratory and amplified in PCR thermocycler in post-PCR laboratory. Then, each PCR batch was run in agarose gel electrophoresis (2%) to check if it can be purified. Each PCR batch was replicated until obtaining two positive PCR products for each sediment sample with associated negative controls.  Positive means: the brightest staning is in a range of 100-200 bp and the gene band is evidently longer than one of control. The gene band with length < 50 bp is considered as primer dimer.  All qualified PCR replicates were purified with the MinElute PCR Purification Kit (Qiagen, Germany) in post-PCR lab.  Afterwards, 1 μL of PCR product was measured with the ds-DNA BR Assay and the Qubit^® 2.0 fluorometer (Invitrogen, USA). Finally, all purified PCR products were equimolarly pooled.  It should be note that the ancient DNA lab and the post-PCR lab are located in different buildings.  The genetic lab work was done in Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research, Potsdam. 

2. High throughput sequencing

The pooled PCR products were sequenced on the Illumina HiSeq 2500 platform (2 x 125 bp, paired-end reads) at Fasteris SA sequencing service, Switzerland. The mode is HiSeq High Output Version 4 with the HiSeq SBS Kit v4. Our project resulted in 9.5 Gb with 37,922,797 generated clusters >= Q30.

3. Bioinformatics

We used OBITools (Boyer et al., 2016) to do the upstream analysis of  the sequencing data. The raw sequecning data (two fastq files), scripts for creating the taxonomic database and analyzing the raw sequencing data are enclosed as well as the tag-sample-matirx.

1. There are no missing values.

2. Metabarcoding data: the raw NGS data

    Forward: 170421_SND393_B_L004_HUA-7_R1.fastq.gz

    Reverse: 170421_SND393_B_L004_HUA-7_R2.fastq.gz

3. NGSfilter file: tag-sample-matrix for demultiplexing the NGS data in ObiTools.

    Filename: 1-tagfile_04NC.txt

4. ObiTools script for building a reference database (v. embl127) for taxonomic assignment.

    Filename: 2-embl127creation_OBITools.txt

5. ObiTools script for analyzing the metabarcoding data

    Filename: 3-MetabarcodingData_analysis_OBITools.txt

6. Two reference databases for taxonomic assignment in ObiTools

    Filename: 4-arctborbryo_gh.fasta; 5-g_h_embl127_new_taxon_good.fasta

7. Taxonomy dump files of Arctic and Boreal vascular plant and bryophyte reference libraries

    Filename: 6-ecochange.zip.gz

8. Scientists who are working on palaeoecology, especially in using environmental DNA from lake sediment core, may be interested in this data.