Electronic Supplementary Material
Ancient DNA complements microfossil record in deep-sea subsurface sediments
Franck Lejzerowicz1, Philippe Esling1, Wojciech Majewski2, Witold Szczuciński3, Johan Decelle4, Cyril Obadia1, Pedro Martinez Arbizu5, Jan Pawlowski1
1Department of Genetics and Evolution, University of Geneva, CH 1211 Genève 4, Switzerland
2 Institute of Paleobiology, Polish Academy of Sciences, 00-818 Warszawa, Poland
3 Institute of Geology, Adam Mickiewicz University in Poznań, 61-606 Poznań, Poland
4 CNRS UMR7144 EPPO & UPMC Université Paris 6, Station Biologique de Roscoff, France
5 DZMB, Senckenberg am Meer, Südstrand 44, Wilhelmshaven, Germany
Table S1. Deep-sea sediment sampling date, coordinates and water depth. Water depths are given in meters below the sea level.
Station / Core / Date / Latitude / Longitude / Water depthME791/550-1 / 550 / 07/22/2009 / 26° 34.08' S / 35° 13.19' W / 4475.9
ME791/552-1 / 552 / 07/22/2009 / 26° 34.11' S / 35° 13.24' W / 4463
ME791/600-1 / 600 / 08/05/2009 / 3° 56.97' S / 28° 5.18' W / 5181.3
ME791/601-1 / 601 / 08/05/2009 / 3° 57.03' S / 28° 5.18' W / 5188.8
Table S2. PCR primer sets, amplification conditions and massive sequencing. For the Foraminifera, only the PCR products obtained with the primer combinations in bold were cloned and Sanger sequenced. The average PCR product lengths were calculated from in silico amplified fragments based on a foraminiferal SSU rDNA reference database containing 1752 sequences. The average lengths include the primer sequences. For next-generation sequencing (NGS), unique combinations of primers tagged at the 5'-terminal ends with tag sequences of 5 nucleotides (denoted by Xs) were used for PCRs. These primers include the sequence of the foraminiferal-specific primer 14F0 (F0-X) and a reverse primer (15-X). The reverse primer was designed in the conserved region adjacent to the 37f hypervariable region under the conditions that i) all known foraminiferans match its 3'-terminal end, ii) the sequence of each tag is at 5 differences (Levenshtein distance) from the conserved positions in the aignment and iii) the sequence of each tag is at 5 differences (Levenshtein distance) from the 5 conserved positions of the forward primer. Each PCR was done in a total volume of 40 µl and contained 1X AmpliTaq Gold buffer, 2.25 mM MgCl2, 1.5 U of AmpliTaq Gold DNA polymerase (Applied Biosystems), 0.2 mM of each dNTP, 0.2 μM of each tagged primer, 0.4 g.l-1 of BSA and 5 µl of metagenomic DNA extract. The amplicons were pooled in equimolar quantities prior to library preparation and NGS. Both the library preparation and the sequencing performed on a paired-end sequencing run of 150 cycles on the MiSeq instrument were realized at Fasteris (Plan-les-Ouates, Switzerland). Raw sequencing data was treated through a basecalling pipeline including MiSeq Control Software 2.1.13, RTA 1.17.22.0 and CASAVA-1.8.2.
Massive sequencing data analysis
The library containing the amplified deep-sea metagenomic DNA extracts was sequenced on a run shared with another library. We filtered the entire run at once only allowing reads with i) a mean quality score above 25, ii) a maximum of two 2 bases displaying a score lower or equal than 25 and iii) no ambiguous base. The paired-end reads were assembled allowing a minimum of 20 aligned bases and after sample demultiplexing based on perfect tagged primers matching and trimming of the primers, we removed the amplifed sequences shorter than 15 bases (primer dimers). A total of 253,356 sequences were further analyzed after removal of unique sequences composed of only one single read in a sample. A total of 1,737 unique sequences were dereplicated across all samples and individually assigned as in [1]. Unique sequences were pre-clustered based on the taxonomic signature represented by the 20 first nucleotides of the 37f hypervariable region and allowing up to 4 differences (edit distance). This signature will at least gather sequences of identical foramininferal genus as used in [1]. Based on pairwise distances calculated from Needleman–Wunsch global alignments (terminal gaps were counted as differences), we delineated operational taxonomic units (OTUs) within each pre-cluster as in [1]. Diversity indices were calculated from matrices the number of reads per OTU and per sample using the package vegan in R [2].
Grain size analysis
Methods
The sediment samples from all four cores taken in 10 cm intervals were processed for volumetric grain size distribution analysis. The samples were treated with hydrogen peroxide to remove organic matter and analyzed with a Malvern Mastersizer 2000 Particle Analyzer at the Faculty of Geographical and Geological Sciences, Adam Mickiewicz University (Poznań, Poland). Grain size statistics were calculated using the logarithmic method of moments with the software GRADISTAT [3], and the data is presented as volume % converted into phi grain size units. The conversion of micrometers into phi values is based on the following formula:
phi = -log2 D
where D is the size of grains in millimeters. Higher phi values correspond to smaller grains.
Results and discussion
The grain size distributions of the analyzed samples are very similar. In table S3 are given the basic grain size statistics. The mean grain size values are in the range of 7.14 to 8.06 phi, corresponding to the fine and very fine silt fraction. All the sediments are poorly sorted (the sorting is 1.46 to 1.78). The grain size distributions are symmetrical to coarse skewed (skewness is -0.65 to -0.07) and mesokurtic (kurtosis is 2.62 to 3.69). All the samples contain some admixture of sand fraction, however most of the sediment is in silt fraction. The sum of the very fine silt and clay fractions in some samples reach slightly above 50%.
The changes in grain size in particular cores are small, however, a slight decrease in mean grain size is observed in most of the cores (figures S1),.
Table S3. Descriptive statistics of the granulometric analysis for each sediment layer. The grain sizes are converted into phi values.
core 550 / mean / sorting / skewness / kurtosis0 cm / 7.37 / 1.68 / -0.37 / 3.61
10 cm / 7.43 / 1.75 / -0.34 / 3.21
20 cm / 7.79 / 1.63 / -0.31 / 3.28
30 cm / 7.97 / 1.58 / -0.15 / 2.62
40 cm / 8.06 / 1.64 / -0.53 / 3.59
core 552
0 cm / 7.14 / 1.58 / -0.17 / 3.54
10 cm / 7.60 / 1.66 / -0.21 / 3.10
20 cm / 7.67 / 1.78 / -0.39 / 3.16
30 cm / 7.55 / 1.57 / -0.07 / 2.94
40 cm / 7.87 / 1.65 / -0.34 / 3.28
core 600
0 cm / 7.69 / 1.47 / -0.20 / 3.45
10 cm / 7.75 / 1.79 / -0.58 / 3.29
20 cm / 7.95 / 1.51 / -0.29 / 3.23
30 cm / 7.75 / 1.77 / -0.42 / 2.94
core 601
0 cm / 7.15 / 1.70 / -0.41 / 3.62
10 cm / 7.85 / 1.55 / -0.28 / 3.31
20 cm / 7.99 / 1.60 / -0.40 / 3.23
30 cm / 7.84 / 1.79 / -0.65 / 3.69
Figure S1. Vertical changes in mean grain size in the studied cores. The grain sizes are converted into phi values.
14C dating
Methods
Two samples composed of carbonate foraminifera tests were selected for high precision AMS 14C measurements. They were performed on the 1.5 SDH-Pelletron Model "Compact Carbon AMS” at Poznań Radiocarbon Laboratory, with raw data presented in 14C year BP. The calculation of the radiocarbon age of the sample assumes that the specific activity of the 14C in the atmospheric CO2 has been constant. However, the atmospheric 14C is not constant and in marine environment a reservoir effect must be taken into account. Therefore, the AMS 14C dates were converted into calibrated ages using the CALIB Rev. 6.1.0 [4] and the Marine09 calibration curve [5]. The difference (ΔR) in reservoir age correction of the model ocean and the region of the Brazilian Basin was not reported so far. Consequently, the values provided by [6] from eastern margin of the Brazilian Basin (St. Helena) were used and averaged asΔR: 306 with average uncertainty 36.
Results and discussion
The raw and calibrated AMS 14C ages are reported in table S4. Since the calibration method returns an age distribution instead of a single value, for the two processed samples the 2-sigma range (2σ) are presented.
Since the cores were taken using a multicorer preserving the sediment – water interface, the Recent age was assumed for the top-most layers and sediment accumulation rates could be calculated. The sediment accumulation rates are in the order of 2.3 cm per thousand years in the case of core 550 and 0.9 cm per thousand years for core 601. These results are consistent with previous observations in the deep sea Atlantic [7].
Table S4. AMS 14C dates and calibrated dates.
Core / Depth (cm b.s.f.) / Material / Lab. no. / Raw AMS 14C / Calibrated years AD±2σ550 / 30 / about 10 mg foraminifera tests / Poz-37140 / 11870 ± 70 BP / 12790 - 13245
601 / 30 / about 50 mg foraminifera tests / Poz-37141 / 28950 ± 260 BP / 32050 - 33230
Micropaleontological analysis
Eighteen sediment samples were wet washed over 63 μm and 20 μm sieves. The >63 μm residua were dried and picked for radiolaria and foraminifera. From most samples all specimens were picked. Rich samples were divided with a dry microsplitter into fractions, from which at least 100 specimens were picked for each group. Specimens were arranged by taxa on micropaleontological slides. Radiolaria are classified on the generic level, according to the classification utilized by [8]. Foraminifera above the species level are classified according to [9]. In addition, for radiolaria, the 20-63 μm fractions were soaked in peroxide, then washed with tap water and propanol. For each sample, a few drops of suspended material was left on a glass to dry, covered with Durcupan ACM resin and a cover glass. The micropaleontological slides were analyzed under Olympus BX50 light microscope (electronic supplementary material, figures S2 and S3).
Figure S2. Radiolaria from analyzed cores. 1.Acrosphaera sp., 550-10 cm; 2.Acanthosphaera sp., 600-surface; 3.Hexacontium sp., 550-10 cm; 4-5.Stylatractus spp., both from 601-30 cm; 6-7.Cenosphaera spp., both from 601-10 cm; 8. ?Spongoliva sp., 600-surface; 9.Didimocyrtis sp., 601-10 cm; 10-11.Dictyocoryne sp., both from 600-surface; 12.Spongaster sp., 600-surface; 13.Spongocore sp., 601-surface; 14. Spongodiscus sp., 552-10 cm; 15.Spongotrochus sp., 550-10 cm; 16.Stylodictya sp., 552-10 cm; 17.Octopyle sp., 552-surface; 18-19. various Pyloniidae, both from 601-surface; 20.Anthocyrtidium sp., 601-10 cm; 21.Cornutella sp., 601-10 cm; 22.Botryostrobus sp., 601-10 cm.
Figure S3. Foraminifera from analyzed cores. Textulariida: 1.Rhabdammina sp., 550-surface; 2.Buzasina galeata (Brady, 1881), 601-surface; 3.Cribrostomoides wiesneri (Parr, 1950), 601-surface; 4.Cribrostomoides subglobosum (Cushman, 1910), 550-30 cm; 5.Paratrochammina sp., 601-surface; 6. ?Alterammina sp., 601-surface. Rotaliida: 7.Globocassidulina subglobosa (Brady, 1881), 550-30 cm; 8.Pullenia salisburyi Stewart et Stewart, 1930, 550-20 cm; 9. Pulleniasubcarinata d'Orbigny, 1839, 550-30 cm; 10.Stainforthia sp., 600-surface; 11.Fursenkoina sp., 601-30 cm; 12.Osangulariellaumbonifera (Cushman, 1933), 550-30 cm; 13.Oridorsalis umbonatus (Reuss, 1851), 550-30 cm; 14-15.Oridorsalis cf. O.umbonatus (Reuss, 1851), 601-30cm, 550-30 cm; 16.Cibicides sp., 600-30cm; 17.Epistominella sp., 601-30 cm; 18.Ioanella tumidula (Brady, 1884), 601-30 cm. Globigearinida: 19.Globorotalia tumida(Brady, 1877), 601-20 cm; 20.Globorotalia crassaformis (Galloway et Wissler, 1927), 601-20 cm; 21.Globorotalia inflata(d’Orbigny, 1839), 550-30 cm; 22.Turborotalita quinqeloba(Natland, 1938), 550-30 cm; 23.Globigerinoides sacculifer (Brady, 1877), 601-surface; 24-25. various globigerinids, 601-30 cm, 600-surface.
Figure S4. Photomicrographs of 20-63μm fraction of selected samples from sites 550 and 600, all to the same scale (scale bar: 100 μm). Abundant and diverse radiolarian assemblages (A and C) contrats with mineral-rich material almost barren in radiolarian skeletons (B and D).
Table S5. Foraminiferal microfossil counts. Foraminiferal abundances from > 63 μm fraction in number of specimens per gram of a dry sediment throughout the four cores. Note different scales for different groups (see attached supplementary material).
Table S6. Radiolarian microfossil counts. Radiolarian abundances from > 63 μm fraction in number of specimens per gram of a dry sediment throughout the four cores. Note different scales for different groups (see attached supplementary material).
Figure S5. PCR amplifications of different size fragments in deep-sea subsurface sediment samples. A. Core 600. B. Core 552. For each layer of each core, three fragment sizes were amplified using highly specific primers targeting the foraminiferal SSU rRNA gene sequence: ~1000 bp (a), ~400 bp (b) and ~150 bp (c). The numbers on top correspond to the sediment layer depths in centimeters. The size of selected molecular weight marker (M) fragments are indicated on the right (in base pairs).
Figure S6. Changes of OTU diversity measures across the sub-samples of each core. Diversity indices were the Shannon diversity index, the Simpson index and Fisher's alpha index. The measures were realized on a matrix containing for each sample the number of reads obtained for each OTU.
Figure S7. OTU richness and sequence abundance of the species and genus shared in both the microfossil record and the DNA sequencing data. The left panel displays the distribution across samples of the OTUs assigned to species or genus found in the morphology. The right panels display the number of Sanger and NGS sequences found for each of the OTU of the left panel.
References
- Lecroq, B., Lejzerowicz, F., Bachar, D., Christen, R., Esling, P., Baerlocher, L., Østerås, M., Farinelli, L. & Pawlowski, J. 2011 Ultra-deep sequencing of foraminiferal microbarcodes unveils hidden richness of early monothalamous lineages in deep-sea sediments, Proc. Natl. Acad. Sci. USA.108, 13177–13182. (doi:10.1073/pnas.1018426108)
- R Development Core Team 2010 R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing.
- Blott, S. J. & Pye, K. 2001 GRADISTAT: a grain size distribution and statistics package for the analysis of unconsolidated sediments. Earth. Surf. Process. Landforms 26, 1237–1248. (doi:10.1002/esp.261)
- Stuiver, M., Reimer, P. J., and Reimer, R. W. 2005. CALIB 5.0. [WWW program and documentation].
- Reimer, P. J. et al. 2009 IntCal09 and Marine09 radiocarbon age calibration curves, 0–50,000 years cal BP. Radiocarbon51, 1111–1150.
- Lewis, C. A., Reimer, P. J. & Reimer, R. W. 2008 Marine reservoir corrections: St. Helena, South Atlantic Ocean, Radiocarbon50, 275–280.
- Mollenhauer, G., Schneider, R. R., Jennerjahn, T., Müller, P. J. & Wefer, G. 2004 Organic carbon accumulation in the South Atlantic Ocean: its modern, mid-Holocene and last glacial distribution. Glob. Planet. Change40, 249–266. (doi:10.1016/j.gloplacha.2003.08.002)
- Boltovskoy, D. 1998 Classification and distribution of South Atlantic recent polycystine Radiolaria. Palaeontologia Electronica1, 116 pp.
- Loeblich, A. R. & Tappan, H., 1987 Foraminiferal Genera and their Classification. Van Nostrand Reinhold Company, New York, 2 Vols, 970 pp. + 847 pls.