Data Descriptor | Open

A foraminiferal δ18O record covering the last 2,200 years

  • Scientific Data 3, Article number: 160042 (2016)
  • doi:10.1038/sdata.2016.42
  • Download Citation
Published online:


Thanks to the precise core dating and the high sedimentation rate of the drilling site (Gallipoli Terrace, Ionian Sea) we were able to measure a foraminiferal δ18O series covering the last 2,200 years with a time resolution shorter than 4 years. In order to support the quality of this data-set we link the δ18O values measured in the foraminifera shells to temperature and salinity measurements available for the last thirty years covered by the core. Moreover, we describe in detail the dating procedures based on the presence of volcanic markers along the core and on the measurement of 210Pb and 137Cs activity in the most recent sediment layers. The high time resolution allows for detecting a δ18O decennial-scale oscillation, together with centennial and multicentennial components. Due to the dependence of foraminiferal δ18O on environmental conditions, these oscillations can provide information about temperature and salinity variations in past millennia. The strategic location of the drilling area makes this record a unique tool for climate and oceanographic studies of the Central Mediterranean.

Design Type(s)
  • time sampled measurement data set
  • observation design
Measurement Type(s)
  • oxygen-18 atom
Technology Type(s)
  • isotope ratio mass spectrometer
Factor Type(s)
    Sample Characteristic(s)
    • Globigerinoides ruber
    • Gulf of Taranto
    • marine sediment

    Background & Summary

    Knowledge of natural climate variability is required to understand recently observed trends and to assess anthropogenic effects on climate. Instrumental series, however, cover only the last few centuries. Marine cores with very high sedimentation rates allow for investigating climate variations on scales of decades up to millennia.

    For many years the Torino Cosmogeophysics group has studied sediment cores drilled from the Gallipoli Terrace in the Gulf of Taranto (Ionian Sea) and deposited in the last millennia. This site is close to the volcanic Campanian area, and this allowed us to date the cores with great accuracy by tephroanalysis. It also has a high sedimentation rate, allowing for a high time resolution (3.87 years).

    We measured the oxygen isotope composition δ18O of planktonic foraminifera in one of the cores extracted from the Gallipoli Terrace. These measurements provided a high-resolution, 2,200-year-long record (see Data Citation 1: Pangaea which allowed us to demonstrate that the millennial trend and the bicentennial oscillation detected in this series are temperature driven1.

    In a recent paper2, using these isotopic data together with historical Po River discharge data and oceanographic measurements, we have shown that the Po River discharge undergoes robust decadal fluctuations that reach the Ionian Sea, ~1,000 km south of the Po River delta, by salinity anomalies propagation. These salinity variations have been registered in the foraminifera shells3,4 and allowed for the first time a reconstruction of North Italian hydrological variability on millennial-scale.

    The value of this isotopic δ18O record, lying in its uncommonly high time-resolution and length, makes it worth of a thorough description. Therefore, here we provide details about the sediment core and its dating, as well as about the procedures of sample preparation and measurement. Moreover, in the last section we give several analysis and arguments validating the data-set.


    Drilling area and characteristics of the core

    The gravity core GT90/3, in which the δ18O series was measured, was drilled from the Gallipoli Terrace in the Gulf of Taranto (Ionian Sea) at 39°45′53′′N, 17°53′33′′E (see Fig. 1). It was extracted at a depth of 178 m and its length is 3.57 m. The map in Fig. 1 shows the extraction site of the GT 90/3 core and of the other ones we retrieved from the same Gallipoli Terrace (cores GT14 and GT89/3 (refs 5,6) were drilled very close to GT90/3 and are represented by the same point on this map). The cores are stored at the underground Laboratory of Monte dei Cappuccini (OATo-INAF) in Torino (Italy).

    Figure 1: Location of the extraction site of core GT90/3.
    Figure 1

    (a) Core GT90/3 was drilled in the Gulf of Taranto, Ionian Sea. The map was made with the Language for technical computing MATLAB, version R2015b (MathWorks Inc.) (b) Bathymetric map showing the Gallipoli Terrace, from which our cores were extracted. We measured δ18O in the GT90/3 core (red point).

    Thanks to its geographical location, the Gallipoli Terrace is a favourable site for climatic studies based on marine sediments, because of its closeness to the volcanically active Campanian area, a region that is unique in the world for its detailed historical documentation of volcanic eruptions (see the Technical Validation section). Tephra layers corresponding to historical eruptions were identified along the cores, thus allowing for accurate dating and determination of the sedimentation rate. The measurements performed in different cores from the same area showed that the sedimentation rate is uniform across the whole Gallipoli Terrace5,6. Our group measured several profiles in the cores, including 210Pb and 137Cs activity, as well as pyroxene density (see the Technical Validation section), CaCO3 content5,​6,​7,​8 and thermoluminescence9,10, in addition to oxygen and carbon isotopic ratios1,2,7,11,12.

    Sample preparation and δ18O measurement

    We sampled the core using a spacing of 2.5 mm. Each sample of sediment (5 g) was soaked in 5% calgon solution overnight, then treated in 10% H2O2 to remove any residual organic material. Subsequently it was washed with a distilled-water jet through a sieve with a 150 μm mesh. The fraction >150 μm was kept and oven-dried at 50 °C. The planktonic foraminifera Globigerinoides ruber were picked out of the samples under a microscope. For each sample, 20–30 specimens were selected from the fraction comprised between 150 and 300 μm.

    The use of a relatively large number of specimens for each sample reduces the isotopic variability of individual organisms, giving a more representative δ18O value. The stable isotope measurements were performed using a VG-PRISM mass spectrometer fitted with an automated ISO-CARB preparation device. Analytical precision based on internal standards was better than 0.1‰. Calibration of the mass spectrometer to VPDB scale was done using NBS19 and NBS18 carbonate standards.

    δ18O profile

    The δ18O profile we obtained is shown in Fig. 2. It consists of a continuous record of 560 points from 200 BC to 1979 AD, with a sampling interval of Δt=3.87 y.

    Figure 2: δ18O isotopic ratio profile measured in GT 90/3 core.
    Figure 2

    Mean value is 0.46‰ and standard deviation is 0.23‰. The sampling interval is Δt=3.87 years and the data are expressed in VPDB standard. Since δ18O is inversely correlated with temperature, the δ18O series is plotted on a reversed y-axis.

    Several profile features clearly correspond to distinct climatic periods: the low δ18O values near 1000 AD are associated with the Medieval Warm Period (MWP) and to the Industrial Era (IE); relatively high values correspond to the Little Ice Age (LIA, 1600–1800 AD). We also find a δ18O maximum at about 0 AD, suggesting low temperature at that time (this controversial period is deeply discussed in Sec. 5 of Taricco et al. 1).

    Data Records

    The δ18O record (Fig. 2), covering the last 2 millennia, is available on Pangaea (see Data Citation 1: Pangaea in machine-readable ASCII format. The file has four columns, the first one with the depth of the sample (in cm), the second and the third ones with the date of the sample (in years AD and BP respectively) and the fourth with the δ18O value (permil).

    Technical Validation

    Uncertainties of δ18O measurements

    Regarding the uncertainties of δ18O measurements, we did not replicate the isotopic analysis at a given depth in the core to estimate the uncertainty, because of the small number of Globigerinoides ruber in the sediment. In order to obtain a value of δ18O, which is representative of the sample, we need to collect about 20–30 specimens, which is about the total number of suitable foraminifera contained in each 2.5 mm-thick slice. For this reason, we prefer to choose high resolution and to extract later the signal from the noise, using statistical methods applied to the whole series. This is done by a Monte Carlo procedure that allows for assigning a confidence level to each variability mode detected in the series, as described in the next subsection.

    Extraction of signal from noise

    The δ18O series was analysed by different spectral methods1. Using Singular Spectrum Analysis13,​14,​15,​16(SSA) we demonstrated that the statistically significant part of the δ18O time series is given by the sum of components with periods of 770, 300, 180, 125 and 11.4 years, respectively (periods deduced by applying the Maximum Entropy Method (MEM)). The signal accounts for roughly 42% of the total variance in the series.

    Monte Carlo-SSA15 allows us to verify that the statistically significant part (98% c.l.) of the δ18O time series is given by the sum of these five components, with a residue of red noise. We obtained this result after rejecting a whole range of null hypotheses, including different combinations of components.

    Recently we focused on the decennial range of periods2 and we showed that the robust decadal component is related to the Po River discharge variations.

    High-frequency (decadal) variability in the δ18O series

    The highest-frequency variation detected in the δ18O series is a decadal oscillation. This component is central in Taricco et al. 2, where it allowed for reconstructing the Po river discharge over two millennia. For this reason, it deserves a more in-depth discussion.

    Despite its short period with respect to the time resolution of the series, this oscillation is visible even looking at the raw data. This can be appreciated comparing (Fig. 3) the decadal variation reconstructed by SSA (black curve) and the raw data (red curve) over a part of the series.

    Figure 3: Comparison between the decadal component of the δ18O record (black curve) and the raw data (red curve).
    Figure 3

    The actual presence of this oscillation is supported by the following arguments. First, we detected it at high confidence level (99%); second, its period is not too close to the Nyquist period (7.7 years).

    Moreover, it must be underlined that the experimental procedure rules out the issue of frequency aliasing. Indeed, the discretization of our series is not the result of a punctual sampling of a continuous signal, but derives from the measurement of consecutive sediment slices performed after mixing the material contained in each of them. This mixing removes any possible frequencies higher than the Nyquist frequency, thus acting as a low-pass filter, which avoids the frequency aliasing.

    Theoretical validation of the modern experimental δ18O values

    As is common knowledge, planktonic foraminifera are strongly influenced by the environmental conditions of the near-surface sea in which they live, particularly due to their sensitivity to environmental temperature and salinity.

    The theoretical (expected) δ18Oc in the foraminiferal calcite was determined at different depths using the relationship with temperature (T) and δ18Ow of the water, given by Shackleton3, (1) δ 18 O c = 3 . 86 + δ 18 O w 0 . 23T and the relationship derived by Pierre4 for Mediterranean surface waters (S is the salinity of the considered layer), (2) δ 18 O w = 0 . 25S 8 . 2

    By combining these equations it was possible2 to calculate δ18O at different depths using observed salinity and temperature17,18 in the vicinity of the core location. Figure 4 compares the measured δ18O values with the calculated ones over the common period 1945–1979 covered by both the series of foraminiferal δ18O and the sea temperature/salinity ones.

    Figure 4: Comparison between measured Globigerinoides ruber of the core GT 90/3 δ18O (red points) and theoretical values of calcite δ18O (from Taricco et al., 20152).
    Figure 4

    The expected δ18O series and the corresponding 1-standard-deviation uncertainties were obtained using the Shackleton equation with salinity and temperature yearly-resolved data at different depths (from ref. 18), corresponding to the same Gallipoli site (plotted values are anomalies with respect to the average value of the 0–20 m layer).

    Here we recall the results obtained by this comparison and discussed in Taricco et al.2

    The five theoretical δ18O series of Fig. 4 show that: a) decadal variations are present in the upper layers; b) in spite of the different temporal resolution, these variations are in phase with the foraminiferal δ18O (red points), and the experimental amplitudes of the variations (0.2–0.3 permil) are in good agreement with the theoretical estimates for the layer in which these organisms dwell, i.e., 0 to 20 m depth.

    Core dating

    The δ18O series here described is particularly valuable due to the accurate dating of the core. It is thus worth recalling the tephroanalysis and radiometric measurements that led us to determine the sedimentation rate of the Gallipoli Terrace.


    The high accuracy of the dating of Ionian Sea cores was made possible by the closeness of the drilling site to the Campanian area, a region which is unique in the world for its detailed historical documentation of volcanic eruptions over the last two millennia19. The markers of these eruptions were identified along the cores as peaks of the number density of clinopyroxenes crystals. We found 22 sharp pyroxene peaks, corresponding to historical eruptions of the Campanian area, starting with the 79 AD eruption of Vesuvius that buried Pompei and ending with the last one in 1944 (see Table 1). Figure 5, panel a, shows the time-depth relation over the last two millennia.

    Table 1: Historical date and corresponding pyroxene-peak depth of the 22 eruptions identified along the core.
    Figure 5: Time-depth calibration of the core.
    Figure 5

    (a) The depth at which a volcanic peak is found in the sediment is plotted versus the historical date of the corresponding eruption, expressed in years counted backwards from 2000 AD. Each red dot in the figure represents a volcanic eruption, corresponding to a peak in the clinopyroxene time series: 22 historical volcanic eruptions occurred in the Campanian area during the last two millennia and have been identified along the core. The straight line resulting from a linear regression fit to the experimental data is also shown (black line). In the right part of panel a, the pyroxene number record is plotted. The arrows indicate the correspondence between the most prominent peaks of the record and the associated points on the calibration line. (bd) Peaks of the pyroxene number record corresponding, respectively, to the eruptions of Pompei, Pollena and Ischia. These peaks are those indicated by the arrows in panel (a).

    Each point represents a pyroxene peak found at a given depth and corresponding to a historical eruption. The linear regression (correlation coefficient r=0.99) gives a sedimentation rate of (0.0645±0.0002)cm y−1. This relatively high value allows for high time-resolution studies: the sampling interval of 2.5 mm corresponds to 3.87 years. The highly linear time-depth relation demonstrates that the sedimentation rate has remained constant, to a really good approximation, over the last two millennia. Moreover, the measurements performed in different cores retrieved from the same area showed that this rate is also uniform across the whole Gallipoli Terrace5,​6,​7. Taricco et al.20 confirmed this dating by applying advanced statistical procedures21,22. The tephroanalysis dating confirmed, improved and extended to the deeper part of the core the dating obtained in the upper 20 cm by the 210Pb method23,24.

    Radiometric core dating

    The 210Pb activity evaluation23 was carried out at the low-level counting Laboratory of ‘Monte dei Cappuccini’ (OATo-INAF) in Torino (70 m.w.e.). The activity was determined through the detection of the β activity of the 210Bi (T1/2=5.013 d) in equilibrium with 210Pb, using a set of Geiger-Muller counters operating inside a shielded well. The experimental measurements are shown in Fig. 6.

    Figure 6: 210Pb activity as a function of sediment depth (from Bonino et al.23).
    Figure 6

    The sediment samples for the activity estimation were taken at 10 different depths, with a sampling thickness of 1 cm. Green and blue hyphens are the total and excess activity, respectively. The black line is the least-square fit of the excess activity. 137Cs activity is reported in the left-hand bottom corner of the figure (in red).

    The decreasing solid line, which represents the least-square exponential fit to the excess activity, allowed to deduce the sedimentation rate S=0.064 cm y−1, with an estimated error of 5% and a correlation coefficient r=0.99 (ref. 23). Moreover, in order to validate the results obtained by the 210Pb method and to test the presence of the core top, Bonino et al.23 measured 137Cs activity (T1/2=30 yr), which is primarily due to nuclear bomb testing and related to the radioisotope concentration maximum in the atmosphere that occurred during the 1960s. The γ-activity measurement of 137Cs (see Fig. 6) was performed using a HPGe detector with relative efficiency of 25%, located at the underground Laboratory of ‘Monte dei Cappuccini’ in Torino. The activity maximum was revealed in the first sample, as expected on the basis of 210Pb dating. This result ensured that the core top is present and has not been disturbed during drilling operations or by biological activity, at least within the sampling thickness.

    Limited bioturbation of the sediment and presence of the core top

    The very sharp pyroxene peaks (see, for example, Fig. 5—panels b,c and d showing the peaks corresponding to the Pompei, Pollena and Ischia eruptions) indicate that bioturbation by bottom-dwelling organisms is quite limited. Therefore, this climatic record is not significantly affected by sediment mixing. Moreover, the presence of the core top is demonstrated by the 137Cs activity, as described in the previous paragraph.

    Validity of the dating over the whole Gallipoli Terrace

    The profiles of CaCO3 content measured in different cores retrieved from the same area showed that the sedimentation rate is uniform across the whole Gallipoli Terrace5,​6,​7,​8. The CaCO3 content of each layer was determined by titration with EDTA, using the procedure explained by Barnes25.

    The profiles of the CaCO3 content of different cores display a remarkable layer-to-layer correspondence, within a fraction of a percent. This unusual feature guarantees an essential requisite for the reproducibility and generality of our results, i.e., the uniform stratigraphy of the Gallipoli Terrace.

    Additional Information

    How to cite this article: Taricco, C. et al. A foraminiferal δ18O record covering the last 2,200 years. Sci. Data 3:160042 doi: 10.1038/sdata.2016.42 (2016).


    1. 1.

      , , & Two millennia of climate variability in the Central Mediterranean. Climate of the Past 5, 171–181 (2009).

    2. 2.

      et al. Marine Sediments Remotely Unveil Long-Term Climatic Variability Over Northern Italy. Scientific reports 5, 1–9 (2015).

    3. 3.

      & Paleotemperature history of the Cenozoic and the initiation of Antarctic glaciation: oxygen and carbon isotope analyses in DSDP sites 277, 279, and 281 Initial Reports of the Deep Sea Drilling Project 29 (U.S. Government Printing Office, 743–755, 1975).

    4. 4.

      The oxygen and carbon isotope distribution in the Mediterranean water masses. Mar. Geol 153, 41–45 (1999).

    5. 5.

      , , , & The carbonate profile of two recent Ionian sea cores: evidence that the sedimentation rate is constant over the last millennia. Geophys. Res. Lett. 17, 1937–1940 (1990).

    6. 6.

      , , , & The CaCO3 profiles of deep and shallow Mediterranean sea core sas indicators of past solar-terrestrial relationship. Nuovo Cimento C 15, 547–563 (1992).

    7. 7.

      , & Long term solar-terrestrial records from sediments: carbon isotopes in planktonic foraminifera during the last millennium. Adv. Space Res. 29, 1537–1549 (2002).

    8. 8.

      , , & Common spectral features in the 5500-year record of total carbonate in sea sediments and radiocarbon in tree rings. Radiocarbon 34, 798–805 (1992).

    9. 9.

      , , & Record of thermoluminescence in sea sediments in the last millennia. Nuovo Cimento C 20, 1–8 (1997).

    10. 10.

      , & The global and persistent millennial-scale variability in the thermoluminescence profiles of shallow and deep Mediterranean sea cores. Nuovo Cimento C 21, 453–460 (1998).

    11. 11.

      , , & Solar radiation variability in the last 1400 years recorded in the carbon isotope ratio of a mediterranean sea core. Adv. Space Res. 29, 1987–1994 (2002).

    12. 12.

      , , , & Solar activity in the last millennium recorded in the d18O profile of planktonic foraminifera of a shallow water Ionian Sea core. Solar Physics 188, 191–202 (1999).

    13. 13.

      , & Singular-spectrum analysis: A toolkit for short, noisy chaotic signals. Physica D 58, 95–126 (1992).

    14. 14.

      et al. Advanced spectral methods for climatic time series. Rev. Geophys. 40, 3.1–3.41 (2002).

    15. 15.

      & Monte Carlo SSA: detecting irregular oscillations in the presence of coloured noise. J. Clim 9, 3373–3404 (1996).

    16. 16.

      & Past and Present Variability of the Solar-terrestrial System: Measurement, Data Analysis and Theoretical Models (eds Cini Castagnoli G. & Provenzale A. 137–159 (IOS Press, 1997).

    17. 17.

      et al. The Western Mediterranean Deep Water: a proxy for climate change. Geophys. Res. Lett. 32, L12608, 1–4 (2005).

    18. 18.

      MEDAR Group Mediterranean and Black Sea database of temperature, salinity and biochemical parameters and climatological atlas. 4 CD-ROM and European Commission Marine Science and Technology Programme (MAST). Ed. IFREMER (2002).

    19. 19.

      et al. Quaderni de La Ricerca Scientifica vol. 114 (ed. Santacroce R. ) 53–103 (CNR, 1987).

    20. 20.

      , & Sequence of eruptive events in the Vesuvio area recorded in shallow-water Ionian Sea sediments. Nonlin. Processes Geophys 15, 25–32 (2008).

    21. 21.

      , & A signal extraction approach to modelling hormone time series with pulses and a changing baseline. J. Am. Stat. Assoc. 94, 746–756 (1999).

    22. 22.

      , , & An automatic statistical methodology to extract pulse like forcing factors in climatic time series: Application to volcanic events, Volcanism and the Earth’s Atmosphere. Geophysical Monograph 139, 177–186 (2003).

    23. 23.

      , , & Radiometric and tephroanalysis dating of recent Ionian sea cores. Nuovo Cimento C 16, 155–161 (1993).

    24. 24.

      , , & Geochronology of lake sediments. Earth Planet. Sci. Lett. 11, 407–414 (1971).

    25. 25.

      Apparatus and methods of oceanography (Allen & Unwin London, 1959).

    Download references

    Data Citations

    1. 1.

      Taricco, C., Alessio, S., Rubinetti, S., Vivaldo, G. & Mancuso, S. Pangaea (2016).


    We wish to thank S.M. Bernasconi (Geological Institute, ETH, Zuerich, Switzerland) for the mass spectrometer measurements and Mr A. Romero for his precious technical support.

    Author information


    1. Dipartimento di Fisica, Università di Torino, Torino 10125, Italy

      • Carla Taricco
      • , Silvia Alessio
      •  & Sara Rubinetti
    2. Osservatorio Astrofisico di Torino, INAF, Pino Torinese 10025, Italy

      • Carla Taricco
      • , Silvia Alessio
      •  & Salvatore Mancuso
    3. IMT School for Advanced Studies, Lucca 55100, Italy

      • Gianna Vivaldo


    1. Search for Carla Taricco in:

    2. Search for Silvia Alessio in:

    3. Search for Sara Rubinetti in:

    4. Search for Gianna Vivaldo in:

    5. Search for Salvatore Mancuso in:


    C.T. designed the study. C.T., S.R., S.A. wrote the paper. S.R., G.V. prepared the foraminifera samples. All the authors collaborated to data analysis and technical validation of the series.

    Competing interests

    The authors declare no competing financial interests.

    Corresponding author

    Correspondence to Carla Taricco.

    Creative CommonsThis work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit Metadata associated with this Data Descriptor is available at and is released under the CC0 waiver to maximize reuse.