Speleothem δ 13 C record suggests enhanced spring/summer drought in south-eastern Spain between 9.7 and 7.8 ka

Abstract

South-eastern Spain is one of the driest regions in Europe and thus, prone to drought. Terrestrial climate records covering the late Glacial and Holocene from this area are sparse. Here, we present a flowstone record from Cueva Victoria, south-eastern Spain, which covers the late Glacial (15 ka) to the mid-Holocene (7 ka) including the Younger Dryas (YD). Between the onset of the Bølling/Allerød (B/A) and the early Holocene, flowstone δ¹⁸O values progressively decrease in accordance with sea-surface temperatures in the Alboran Sea, indicating an increase in precipitation in south-eastern Spain and a supra-regional signal of North Atlantic temperature change. At the same time, decreasing δ¹³C values suggest progressively increasing precipitation and vegetation density. This trend is interrupted by both colder and drier conditions during the YD. Between 9.7 ± 0.3 and 7.8 ± 0.2 ka, a large positive excursion of the δ¹³C values indicates a strong reduction in vegetation density, probably as a consequence of very dry spring/summer conditions. In combination with the continuously low speleothem δ¹⁸O values and a nearly unchanged growth rate, this suggests increased seasonality (i.e. drier spring/summer conditions, but not a strong reduction in annual precipitation). This is consistent with several other climate records from the Western Mediterranean region, showing that the Western Mediterranean realm (Spain, Italy) experienced pronounced spring/summer drought during this time interval. Interestingly, the timing of this dry period coincided with the African Humid Period. This may be part of a teleconnection with the North African Monsoon via the Hadley cell circulation.

Keywords

drought Holocene stable carbon isotopes stable oxygen isotopes south-eastern Spain U-series dating

Introduction

In the last decades, speleothems (secondary carbonates in caves) have been established as important terrestrial paleoclimate archives (Fairchild and Baker, 2012; Henderson, 2006). Their greatest advantage compared to other terrestrial archives, such as lake sediments, is that they can be very precisely dated by the ²³⁰Th/U-dating method, even beyond the limit of ¹⁴C dating (>50 ka) (Richards and Dorale, 2003; Scholz and Hoffmann, 2008). In addition, speleothems occur worldwide and are usually well preserved in the sheltered cave environment. Several climate proxies, such as stable oxygen (δ¹⁸O, e.g. Lachniet, 2009; McDermott, 2004) and carbon isotope values (δ¹³C, e.g. Genty et al., 2005; Ridley et al., 2015; Rudzka et al., 2011) as well as trace element concentrations (e.g. Fairchild and Treble, 2009), can be measured at high temporal resolution.

The δ¹⁸O values in speleothems from the Mediterranean region are commonly interpreted as a proxy for changes in precipitation (e.g. Ayalon et al., 1998, 2002; Bard et al., 2002), while at higher latitudes and altitudes, they are assumed to be more sensitive to temperature changes (e.g. Boch et al., 2009; Fohlmeister et al., 2012; McDermott et al., 2011). However, speleothem δ¹⁸O values can also be affected by several other processes (Lachniet, 2009; McDermott, 2004). Speleothem δ¹³C values can be interpreted as a proxy for changes in the composition (Cerling et al., 1993; Denniston et al., 2000; Dorale et al., 1998) and density (Fohlmeister et al., 2011) of the vegetation above the cave and microbiological activity in the soil (Breecker et al., 2012; Genty et al., 2003, 2006; Meyer et al., 2014). Since effective meteoric precipitation directly influences vegetation density and soil microbiological activity, it has an effect on soil pCO₂ and the δ¹³C value of the drip water (Meyer et al., 2014; Ridley et al., 2015). In general, lower δ¹³C values correspond to higher vegetation density and microbiological activity in the soil (Fohlmeister et al., 2011). Thus, δ¹³C values should increase in case of decreasing precipitation and vegetation density and vice versa.

Currently, only a few terrestrial climate reconstructions from the Mediterranean region are available, and these are mainly based on pollen records (Allen et al., 1999; Brauer et al., 2007; Tzedakis et al., 2004, 2006), near-shore sediments (Bardají et al., 2009; Mauz et al., 2012; Zazo et al., 2013), and speleothems covering glacial-interglacial timescales (Bard et al., 2002; Bar-Matthews et al., 2003; Hodge et al., 2008). In the Western Mediterranean, climate archives covering the onset of the Holocene are even more sparse and mainly limited to pollen sequences (Carrión et al., 2010; Pérez-Sanz et al., 2013; Tinner et al., 2009), lake-level reconstructions (Magny et al., 2011, 2012; Peyron et al., 2013), and speleothem records from Sicily (Frisia et al., 2006), central and northern Italy (Scholz et al., 2012; Zanchetta et al., 2007). The pollen records have a relatively low temporal resolution (e.g. Carrión, 2002) and often do not cover the onset of the Holocene. The speleothem data, in contrast, have a higher temporal resolution (e.g. Genty et al., 2006; Moreno et al., 2017). However, paleoclimate information from speleothems covering the Pleistocene/Holocene transition is not available yet for south-eastern Spain, which is one of the driest regions in southern Europe with a strong seasonality in precipitation. Thus, this is a key area to study past climate variability and changes in seasonality during the Holocene in southern Europe, which are poorly understood.

Here, we present a high-resolution flowstone stable isotope record from Cueva Victoria (CV), south-eastern Spain, covering the period from 15 to 7 ka. A previous investigation of flowstones from CV showed that they cover at least the last 1 Ma (Gibert et al., 2016) and mainly grew during warm and humid interglacials (Budsky et al., 2015). This demonstrates that the flowstones from CV represent a sensitive hydroclimate archive in this currently semiarid region.

Regional setting

CV is located in south-eastern Spain between Cartagena and Mar Menor (37.6°N, 0.82°W, 40 m asl, Figure 1). The area is one of the driest regions in Europe with an annual precipitation between 200 and 300 mm (Agencia Estatal de Meteorología, 2011). The climate is characterised by a strong seasonality with moderate precipitation during winter and spring (<30 mm/month), followed by a hot and dry summer (<10 mm/month) and a more humid autumn (≈50 mm/month, Figure S2, available online). It is classified as BSk climate according to the Köppen-Geiger classification, which reflects arid cold steppe climate with a mean annual temperature (MAT) <18°C (Kottek et al., 2006). September to November rainfall is transported to the area by cold air masses entering the warm Mediterranean Sea from the north-west (Araguas-Araguas and Diaz Teijeiro, 2005). During the rainy season, most of the rainfall occurs within 20–30 days, and 10% to 15% of the annual precipitation falls during flash flood events (Araguas-Araguas and Diaz Teijeiro, 2005). This rainfall does not penetrate the thin soil cover, but runs off superficially and is, therefore, not available for the vegetation. The annual precipitation pattern is negatively correlated with the Western Mediterranean Oscillation Index (WeMO), the pressure difference between the Gulf of Cadiz (Spain, San Fernando) and Padua (Italy, Po basin; Martin-Vide and Lopez-Bustins, 2006). This is the result of low pressure entering the Western Mediterranean through the Strait of Gibraltar and moisture transported to eastern Spain, whereas Italy is under the influence of high pressure. A positive WeMO index leads to enhanced precipitation in the eastern Western Mediterranean and reduced precipitation in eastern Spain. There is no significant influence on precipitation by the North Atlantic Oscillation (NAO, Comas-Bru and McDermott, 2014), which has a strong influence in many other parts of Europe (Deininger et al., 2016; Hurrell and Loon, 1997). The annual temperature pattern in south-eastern Spain shows a stronger relationship with the East Atlantic pattern (EA, Ríos-Cornejo et al., 2015).

Figure 1.

Modern precipitation (CRU TS 4.01 1901–2016; Harris et al., 2014) and mean sea-level pressure (contour lines; HadSLP 2r, 1850–2018; Allan and Ansell, 2006) and the estimated situation for the interval between 9.7 ± 0.3 and 7.8 ± 0.2 ka (dashed lines, see text). For (a) winter, (b) spring, (c) summer, and (d) autumn, the estimated northernmost position of the ITCZ (blue lines) is shown. In addition, several records discussed in the text are shown: 1. Ernesto Cave (Scholz et al., 2012); 2. Chauvet Cave (not covering the time interval, Genty et al., 2006); 3. Corchia Cave (Zanchetta et al., 2007); 4. Lake Accesa (Drescher-Schneider et al., 2007; Finsinger et al., 2010; Peyron et al., 2011); 5. Speleothem records from multiple cave sites from northern Spain (Stoll et al., 2013); 6. Kaite Cave (Domínguez-Villar et al., 2017); 7. Basa de la Mora (Pérez-Sanz et al., 2013); 8. Marcelino tufa deposits (Pellicer et al., 2016); 9. Ebro Basin sediments (Bastida et al., 2013); 10. Pollen record, La Garrotxa (Piqué et al., 2018); 11. Lake Estanya (González-Sampériz et al., 2017; Morellón et al., 2009); 12. Molinos Cave (Moreno et al., 2017), 13. Lake Villarquemado (Aranbarri et al., 2014); 14. Lake Salines (Burjachs et al., 2016) and Villena Lake (Jones et al., 2018); 15. Lake Siles (Carrión, 2002); 16. Nerja Cave (McMillan, 2006); 17. Refugio Cave (Walczak et al., 2015); 18. Sediments, San Rafael (Pantaléon-Cano et al., 2003); 19. Grotta di Carburangeli (Frisia et al., 2006); 20. Gorgo Basso (Tinner et al., 2009); 21. Alboran Sea sediment cores MD95-2043 (Cacho et al., 1999; Fletcher et al., 2010; Fletcher and Sánchez Goñi, 2008); ODP161-976 (Combourieu Nebout et al., 2009; Martrat et al., 2014); 22. Grotte de Piste (Wassenburg et al., 2016); 23. Lake Sidi Ali (Zielhofer et al., 2017); 24. GC27 (Tierney et al., 2017); and 25. GeoB790-2 (Tjallingii et al., 2008) (For interpretation of the colour-codes, the reader is referred to the online version of this article.)

CV formed in Triassic limestones and dolostones of the Inner Betic Cordillera. These rocks belong to the Alpujarride metamorphic complex (San Ginés unit) and are partly karstified (Manteca Martínez and Pina, 2015). In these zones, manganese ore occurs associated with reddish clay, silt and sand linked to south-west striking faults (Manteca Martínez and Pina, 2015). The cave system extends over 3 km laterally and 155 m vertically (Ros and Llamusí, 2015) and consists of several chambers. The original cave system was widened during the last century due to mining for the underlying manganese ore (Manteca Martínez and Pina, 2015). Ventilation of the original cave system was probably lower compared to the very well ventilated situation today (pCO₂ < 500 ppmV, data provided by the local caving group, CENM-naturaleza), which results from several artificial openings related to mining. The part, where the flowstone core was recovered, is less well ventilated, as is reflected by the high present-day relative humidity (>90%). Mean annual cave air temperature at the sampling site of the flowstone is 17°C (CENM-naturaleza).

CV is well known as an early Pleistocene fossil site with excellently preserved bones of a diverse fauna (Ferràndez-Cañadell et al., 2014; Gibert et al., 2016; Gibert and Ferràndez-Canadell, 2015), including hominin bones (Gibert et al., 2008; Ribot et al., 2015). The fossil-bearing red breccia is covered by a flowstone unit, which occurs throughout the cave and preserves a change in magnetic polarity (Gibert et al., 2016), presumably related to the Brunhes/Matuyama reversal (0.78 Ma). Flowstone thickness is typically between 5 and 15 cm, but may partly exceed 50 cm in the deeper parts of the cave (Gibert et al., 2016).

Material and methods

Sample collection and preparation

Several flowstone cores were collected from chamber Victoria III (Ros and Llamusí, 2015) using a mobile core drilling device with a diameter of 5 cm. Vic-III-4 has a length of 31 cm, whereas Vic-III-1 and Vic-III-3 are 41.5 and 40.5 cm long, respectively. All cores were embedded in gypsum and cut into slabs (thickness ≈1 cm). One half of the slab was used for polished thin sections (70 µm) to study the crystal fabric (Figure 2). Here, we focus on the upper 12 cm of Vic-III-4 (Figure 2), which correspond to the Holocene (section ‘²³⁰Th/U dating’). The other two cores mainly grew during older interglacials, and the Holocene is only contained in the upper few millimetres (Figure S1, available online).

Figure 2.

(Left) Scan of sample Vic-III-4 with red arrows indicating the positions of ²³⁰Th/U-dating samples. The black line on the right side of the slab marks the stable isotope traverse, and the red line represents the profile for trace-element analysis. (Right) The corresponding thin sections in cross-polarised light prepared from the opposite slab.

²³⁰Th/U dating

Twenty-two samples (100 – 300 mg) were cut from the upper 12-cm section of sample Vic-III-4 with a diamond-coated bandsaw following visible growth bands (Figure 2). In addition, one sample was obtained from the top of Vic-III-1 (Figure S1, available online). The samples were dissolved in 7 N HNO₃, and a mixed ²³³U–²³⁶U–²²⁹Th spike (Gibert et al., 2016) was added to the solution. Chemical separation of U and Th was carried out by ion exchange chemistry as described by Yang et al. (2015). U and Th isotope ratios were determined with a Nu instruments multi-collector inductively coupled plasma mass spectrometer (MC-ICP-MS) at the Max Planck Institute for Chemistry, Mainz. Samples were measured by a standard-sample bracketing procedure as described by Obert et al. (2016). Activity ratios and ages were calculated using the half-lives of Cheng et al. (2000) for ²³⁰Th and ²³⁴U, Le Roux and Glendenin (1963) for ²³²Th, and Jaffey et al. (1971) for ²³⁸U.

Trace element measurements

Trace element concentrations were determined by laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) in line scan mode at the Institute for Geosciences, University of Mainz. Line scans extended from the bottom of the flowstone drill core parallel to the growth axis (Figure 2). Analyses were performed using the ESI NWR193 ArF Excimer laser ablation system equipped with a TwoVol2 sample chamber coupled to an Agilent 7500ce quadrupole ICP-MS. The laser was operated at a repetition rate of 10 Hz with a laser energy at the sample site of ≈3 J/cm². The line scans were carried out after pre-ablation of the sample surface with a beam diameter of 100 μm and a scan speed of 10 μm/s. Backgrounds were measured for 20 s prior to each ablation. NIST SRM 612 was used for calibration using the preferred values reported in the GeoReM database (http://georem.mpch-mainz.gwdg.de/, Application Version 21, January 2017, Jochum et al., 2005, 2011) to calculate the element concentrations in the samples. During each run, basaltic USGS BCR-2G (n = 6) and synthetic carbonate USGS MACS-3 (n = 6) were analysed as a quality control material (QCM) to monitor accuracy and reproducibility of the analyses. For all materials, ⁴³Ca was used as an internal standard: For the reference materials, the Ca concentration reported in the GeoReM database, and for the samples, a Ca concentration of 39 wt% (Mertz-Kraus et al., 2009) was used. Data processing was performed using Microsoft Excel following the data reduction scheme of Longerich et al. (1996) and Jochum et al. (2007). Details of the calculations are given in Mischel et al. (2017a). The limit of detection (LOD) was calculated according to Kaiser and Specker (1956) for each element as

{LOD = I}_{background} {+ 3 * SD}_{background}

(1)

where I_background is the mean intensity of the signal recorded during the background measurement, and SD_background is the standard deviation of the corresponding signal interval (Table S1, available online). We report element concentrations for Mg, Sr and Ba. The measured Mg concentrations of the QCMs agree within 10% with the reference values, that is, the preferred values of the GeoReM database for USGS BCR-2G and the preliminary reference values for USGS MACS-3 (personal communication S. Wilson, USGS, in Jochum et al., 2012, Table S1, available online). For Sr and Ba, the measured values of both QCMs deviate less than 5% from the reference values.

Stable isotope measurements

Stable isotope samples of samples Vic-III-1 and -4 were milled at a spatial resolution of 0.5 mm using a NWR MicroMill. For Vic-III-3, the sampling resolution was 0.25 mm. Stable isotope measurements were performed at the Institute of Geology, University of Innsbruck, using a Thermo Fisher Delta^plusXL isotope ratio mass spectrometer linked to a Gasbench II as described in Spötl and Vennemann (2003) and Spötl (2011). Long-term analytical precision (1σ) of the δ¹⁸O and δ¹³C measurements is 0.08‰ and 0.06‰, respectively. Stable isotope values are reported relative to the VPDB standard.

Moisture source modelling

Speleothem δ¹⁸O values often reflect the δ¹⁸O values of precipitation at the cave site, which depend on various parameters, such as latitude, temperature, rainfall amount, and moisture source (Lachniet, 2009; McDermott, 2004). In order to determine the major moisture sources for the precipitation at CV, we performed trajectory analysis using the HYSPLIT (Hybrid Single-Particle Lagrangian Integrated Trajectories) trajectory model of the Air Resources Laboratory of the National Oceanic and Atmospheric Administration (Stein et al., 2015) for three altitudes (1000, 1500, and 2000 m).

We used data from a meteorological station located 20 km NNW (San Javier, 37°46’48 N, 0°48’36 W) of the cave, which provides an almost complete daily-resolution dataset since the late 1940s (KNMI Climate Explorer; Klein Tank et al., 2002). Following the method of Krklec and Domínguez-Villar (2014), we analysed all days since 1950 AD with precipitation >0.5 mm at the San Javier meteorological station, which is considered to be sufficient to contribute to karst aquifer recharge. For each of these days, we calculated the trajectories 5 days backwards (Gimeno et al., 2010). Effective moisture uptake is defined by a change of specific humidity >0.5 g/kg within 6 h (Baldini et al., 2010; Rozanski et al., 1993). In order to account for the boundary layer elevation (Sodemann et al., 2008), we assumed a constant level of 900 hPa (Baldini et al., 2010) and only considered values of the trajectories >900 hPa. It is important to consider that along the pathway of a trajectory, several locations of moisture uptake and loss (precipitation) at different elevations can be defined for each day, month and year. To calculate the contribution (i.e. the importance) of the individual moisture sources to total rainfall at CV for specific time periods (e.g. season, year, decade), the corresponding rainfall events were weighted according to their amount in the trajectory analysis.

In order to explore the relationship between common atmospheric patterns in Europe and precipitation at the study site, we calculated the correlation between precipitation and the indices of the NAO (Iceland-Gibraltar, Jones et al., 1997), the EA (Barnston and Livezey, 1987; Rodriguez-Puebla et al., 1998), the Scandinavian pattern (SCA, Barnston and Livezey, 1987) and the WeMO (Martin-Vide and Lopez-Bustins, 2006, Table S3, available online).

Results

Petrography

The fabrics of flowstone Vic-III-4 (Figure 4) were classified according to Frisia (2015). The flowstone exhibits columnar calcite crystals with increasing crystal size towards the top of the sample. In detail, the flowstone is composed of an elongated columnar fabric (Frisia, 2015) with small crystals at the bottom, which become progressively larger up to 6.5 cm distance from top (dft). The section between 6.5 and 3 cm dft is characterised by open columnar fabric. However, there is no evidence of dissolution and/or recrystallization in this section. At 3 cm dft, the open columnar fabric changes to columnar fabric. There is no petrographic evidence of a hiatus within the upper 1 cm dft as revealed by ²³⁰Th/U-dating (next section, Figure 3).

Figure 3.

Uncorrected (black) and conventionally corrected (i.e. (²³²Th/²³⁸U) = 1.25; orange) ²³⁰Th/U-ages vs distance from the top of the flowstone core Vic-III-4.

The crystal fabric of flowstones Vic-III-1 and -3 is columnar and shows no indication of dissolution or recrystallization. Vic-III-1 shows clear evidence for a hiatus at 0.3 cm dft (Figure S1, available online). In Vic-III-3, a hiatus is present at ca. 0.25 cm dft, indicated by a detrital layer.

²³⁰Th/U dating

Twenty-two ²³⁰Th/U-ages were determined for Vic-III-4. The ²³⁸U concentration ranges from 0.1 to 0.3 µg/g (Table S2, available online) and shows decreasing values from the bottom to the top of the sample. The uncorrected ²³⁰Th/U-ages range from 13.70 ka at the bottom (10.3 cm dft) to 2.78 ka at the top (Table S2, available online and Figure 3). At ca. 9 cm dft, the ²³⁰Th/U-ages suggest a large change in growth rate. Between the two uppermost ages of Vic-III-4, we detected a 4 ka-long hiatus (Figure 3). Not all ages are in stratigraphic order (Figure 3). The samples contain moderate amounts of ²³²Th (0.03 to 2.3 ng/g, mean 0.9 ng/g). The corresponding (²³⁰Th/²³²Th) activity ratios range from 18.8 to 185.7 (Table S2, available online). The effect of detrital contamination on ²³⁰Th/U-ages is considered to be significant if (²³⁰Th/²³²Th) < 200 (Richards and Dorale, 2003). Thus, detrital contamination should have a significant impact on all ²³⁰Th/U-ages of sample Vic-III-4. We applied the conventional correction for detrital Th assuming a ²³²Th/²³⁸U weight ratio of 3.8 (i.e. a (²³²Th/²³⁸U) activity ratio of 1.25) for the detritus (Wedepohl, 1995) and ²³⁰Th, ²³⁴U and ²³⁸U in secular equilibrium. This results in a shift towards younger ages by 0.04 ka (sample Vic-III-4-02) to 0.33 ka (sample Vic-III-4-22; Figure 3). The largest correction is observed for samples Vic-III-4-01, -22, and -27, which have (²³⁰Th/²³²Th) ratios lower than 30 (Table S2, available online). The larger uncertainties for the corrected ages are due to the assumed 50% uncertainty of the (²³²Th/²³⁸U) activity of the detrital material. Obviously, the conventional detrital correction does not eliminate all age inversions (Figure 3).

The uncorrected ²³⁰Th/U-age at the top of Vic-III-1 is 14.37 ± 0.14 ka (Table S2, available online). This sample has a low (²³⁰Th/²³²Th) activity ratio of 2.9 resulting in a relatively large conventional correction of 3.7 ka (10.6 ± 1.8 ka). Below the hiatus at 0.3 cm dft, the ²³⁰Th/U-ages are > 45 ka (not shown in this paper). Thus, only the section above the hiatus corresponds to the Holocene. Similarly, in Vic-III-3, the ages below 0.25 cm dft are > 50 ka. Since the flowstones from this cave mainly grew during interglacials and interstadials (Budsky et al., 2015), we assume that the upper 0.25 cm correspond to the Holocene, even if direct dating is not possible.

Proxy data

Stable oxygen and carbon isotope values of Vic-III-4 show a decreasing trend from the bottom to the middle of the sample (6 cm dft; δ¹⁸O: −4‰ to −6‰, δ¹³C: −7‰ to −11‰; Figure 4). Between 6 and 1.7 cm dft, the δ¹³C values show a large positive excursion from ca. −11‰ to ca. −4‰, while the δ¹⁸O values remain at a constant level of about −6‰. At 1.7 cm dft, the δ¹³C values return to lower values of around −10‰. The position of the δ¹³C excursion is in broad agreement with the observed change from columnar to open columnar fabric (Figure 4). In the lower part of the flowstone, the δ¹⁸O and δ¹³C values show a strong positive correlation (R² = 0.93; Figure 4). Samples Vic-III-1 and -3 show a similar increase in δ¹³C values from ca. −9‰ to ca. −4‰ at 0.25 and 0.2 cm dft, respectively (Figure S1, available online). Subsequently, the δ¹³C values in both cores return to ca. −10‰ (Figure S1, available online). The δ¹⁸O values show a decrease from ca. −4‰ to ca. −6‰ in both cores (Figure S1, available online). Due to the large amount of water with very negative δ¹⁸O values stored in ice sheets, the δ¹⁸O values of the ocean were higher during the late Glacial compared to the Holocene. Therefore, we corrected the δ¹⁸O values of the speleothem record for the corresponding changes in ice volume (Grant et al., 2012; Lambeck et al., 2014). Sea level is directly linked to ice volume and can be used to estimate the ice volume correction (Waelbroeck et al., 2002). To calculate past sea level, we follow the approach of Waelbroeck et al. (2002) using the δ¹⁸O values of benthic foraminifera from the Red Sea (Siddall et al., 2003). The ice volume correction shifts the δ¹⁸O values in the late Glacial section of our speleothem record by ca. −1‰ (Figure 7c). For the youngest part of the flowstone, the correction is negligible (Figure 7c).

Figure 4.

Proxy data and petrographic log (according to Frisia, 2015, C: columnar, Co: open columnar, Ce: elongated columnar) vs distance from the top of Vic-III-4. Y-axes are inverted for δ¹³C and δ¹⁸O values.

Magnesium concentration in Vic-III-4 displays a trend from high values (≈6000 µg/g) at the bottom (10.5 cm dft) to lower values (2000 µg/g) at around 6 cm dft and then remains at a constant level (Figure 4). Strontium concentration shows the same decreasing long-term trend, although it is not correlated with the Mg concentration on shorter time scales. Barium concentration varies between 20 and 160 µg/g, displays no long-term trend and shows peaks at similar dft as Sr.

Trajectory analysis

Figure S4 (available online) shows back-calculated trajectories (120 h, Gimeno et al., 2010) for selected days with precipitation events at the San Javier meteorological station. These examples indicate various pathways of trajectories for CV. The results of the trajectory analysis for the complete rainfall data set from San Javier show that the major sources of precipitation in south-eastern Spain are the Alboran Sea and the surrounding landmasses (N Morocco, Spain) as well as the Iberian Margin (Figure 5). Moisture uptake over the more distant Atlantic Ocean also plays a role. In general, the contribution of the Atlantic Ocean is more pronounced during winter and spring. During summer months, regional moisture sources dominate.

Figure 5.

Moisture uptake for rainy days at the San Javier meteorological station for the period AD 1950 – 2010. Winter (DJF), spring (MAM), summer (JJA), autumn (SON) as well as the rainy season (October to March) are shown in the first five maps. The bar plots on the x- and y-axes show the summarised moisture uptake on a 0.5° × 0.5° grid with respect of the total amount of precipitation. The bottom right panel shows the wind direction (1500 m a.s.l.) transporting moisture to the cave site during the rainy season for the last 6 h before reaching the cave site.

In order to assess whether recent rainfall amount or air temperature has a significant effect on the δ¹⁸O values of precipitation, we examined δ¹⁸O values of modern precipitation from nearby meteorological stations. The weighted mean annual δ¹⁸O values of modern precipitation (GNIP data, Rozanski et al., 1992, http://www-naweb.iaea.org/napc/ih/index.html) in south-eastern Spain (Almeria, Murcia and Valencia) do not show a significant correlation with the amount of annual precipitation (Figure S2, available online). Very high δ¹⁸O values during summer correspond to dry conditions at all stations, whereas lower δ¹⁸O values are observed during the rainy season from October to March (Figure S1, available online). Similarly, no significant correlation with MAT is observed (Figure S3, available online). Thus, modern δ¹⁸O values of precipitation reflect neither precipitation amount nor temperature on the inter-annual scale.

Correlation analysis of winter precipitation (October to March) at San Javier (precip) with the WeMO, the SCA, the NAO and the EA (Table S3, available online) shows that the WeMO (Martin-Vide and Lopez-Bustins, 2006) has the largest impact on the amount of winter precipitation in the region of CV, followed by the EA pattern (Barnston and Livezey, 1987). These two patterns are not correlated with each other. The NAO (Hurrell and Loon, 1997) and the SCA (Barnston and Livezey, 1987) show no significant correlation with winter rainfall at San Javier. During summertime, no significant correlation with any circulation pattern is observed (Table S3, available online).

Discussion

Chronology

The ²³⁰Th/U-ages display several age inversions despite of the correction for detrital Th (Figure 3). This suggests that detrital contamination is not completely accounted for by the conventional correction. Several studies have shown that the conventionally used ²³²Th/²³⁸U ratio is not uniformly appropriate for speleothems and that substantially different values have to be used (Fensterer et al., 2010; Hellstrom, 2006; Hoffmann et al., 2010; Rivera-Collazo et al., 2015; Roy-Barman and Pons-Branchu, 2016). In order to constrain the ²³²Th/²³⁸U ratio of the detrital material, the stratigraphic order of the samples can be used (Drysdale et al., 2006; Hellstrom, 2006; Roy-Barman and Pons-Branchu, 2016). Here, we apply a similar approach.

We used several (²³²Th/²³⁸U)_d activity ratios for the detritus (i.e. 0.08 to 4, separated by an interval of 0.01) and calculated corrected activity ratios, (²³⁴U/²³⁸U)_corr and (²³⁰Th/²³⁸U)_corr, assuming secular equilibrium between detrital ²³⁰Th, ²³⁴U and ²³⁸U. Based on the corrected activity ratios, we then solved the age equation to obtain corrected ²³⁰Th/U-ages. To determine the appropriate correction factor for the CV flowstone, we calculated the sum of all inversions (in years) for each correction factor. For this purpose, each age was compared with all other ages, and age inversions (without taking into account the uncertainties) were summed up.

Figure 6b shows an example for three data points. The conventional correction factor ((²³²Th/²³⁸U)_d = 1.25, orange) leads to two inversions, whereas the red one only leads to one inversion. Finally, the blue dots indicate the correction factor leading to no age inversions. Unfortunately, for the whole data set, no correction factor removing all age inversions was found (Figure 6a), and the lowest value that can be achieved leads to six age inversions in total. For example, the age at 7.5 cm dft generates two inversions, one with the next age and another inversion with the subsequent one (Figure 6b). Due to its relatively large amount of detritus, this sample is very sensitive to changes in the correction factor and is strongly shifted towards younger ages with a decreasing correction factor (Figure 6b). Using the correction factor leading to the minimum sum of all inversions ((²³²Th/²³⁸U)_d = 0.3) results in a total amount of about 1200 a of inversions over the complete sample (Figure 6c). This is only half of the sum of inversions observed using the conventional correction factor. In addition, it is 500 a less than for the approach minimising the total number of inversions (Figure 6c). Therefore, (²³²Th/²³⁸U)_d = 0.3 was used for all further calculations. The effect of the detrital correction is generally stronger in the upper part of the flowstone (<9.5 cm dft, Figure 6a). To account for the potential uncertainties of our approach, we assumed a conservative uncertainty of ± 50 % for the determined (²³²Th/²³⁸U)_d activity ratio (i.e. (²³²Th/²³⁸U)_d = 0.3 ± 0.15), which was propagated to the corrected ²³⁰Th/U-ages. This results in relatively large uncertainties of the individual corrected ages (Table S3, available online). Taking into account these uncertainties, the final data set does not show any significant age inversions (Figure 6d).

Figure 6.

(a) ²³⁰Th/U-ages calculated assuming different detrital correction factors for the 22 analysed samples. (²³²Th/²³⁸U)_d ratios were varied from 0.08 to 4 with an increment of 0.01. Blue dots correspond to the correction factor resulting in the smallest number of age inversions, whereas red dots indicate the correction factor yielding the minimum sum of age inversions in years. For comparison, the orange dots display the ages calculated using the conventional correction factor ((²³²Th/²³⁸U)_d = 1.25). (b) Example for three ages. The conventional correction (orange) leads to three age inversions. The first age is 525 a older than the second age, and the second age is 135 a older than the third age, although both should be younger than the third age according to the stratigraphy. The third age is 660 a younger than the first age. The red model minimises the sum of age inversions for the whole data set leading to only one inversion of 226 a for the three ages shown. The blue model results in no inversions between the three data points. (c) Relation between the sum of age inversions and the (²³²Th/²³⁸U)_d ratio assumed for the detritus. The red dot highlights the value resulting in the lowest value for the sum of inversions, which was used for all further calculations. The orange dot indicates the conventionally used (²³²Th/²³⁸U) correction factor of 1.25. (d) Age model calculated with StalAge (Scholz and Hoffmann, 2011). The youngest age was excluded due to the long hiatus. The individual age uncertainties include the effect of the detrital correction with (²³²Th/²³⁸U)_d = 0.3 ± 0.15 and are substantially larger than for the uncorrected and conventionally corrected ages (Figure 3; Table S2, available online).

Some studies showed that diagenesis and post-depositional mobilisation of U can also occur in speleothems both with (Scholz et al., 2014) and without (Bajo et al., 2016) visible changes in the crystal fabric. Since our approach to account for detrital Th eliminates all age inversions within uncertainty and the crystal fabrics show no evidence of dissolution or recrystallization, we consider post-depositional remobilisation of U as very unlikely.

The StalAge algorithm (Scholz and Hoffmann, 2011) was then used to construct an age–depth model (Figure 6d). We excluded the youngest age from the age model because of the hiatus of approximately 4 ka at the top of the sample (Figures 3 and 7) and calculated the age model only to 1 cm dft (uppermost used age). In addition, we calculated two separate age models (from the bottom to 9.1 cm dft and from 9.1 cm dft to 1 cm dft) because of the large change in growth rate. This change is neither represented in the crystal fabric nor in the stable isotope and trace element data (Figure 4). All age uncertainties reported in the following sections are based on this age model and are below 500 a.

Figure 7.

Comparison of the δ¹⁸O (d, lower curve is not corrected for ice volume) and δ¹³C (h) values (both axes are inverted) of the CV flowstone with other records: δ¹⁸O values from the NGRIP ice core (a, Rasmussen et al., 2006); speleothem δ¹⁸O record from Corchia Cave (e, Zanchetta et al., 2007); SST reconstructions from the Alboran Sea (c, violet Cacho et al., 1999, black Martrat et al., 2014); winter precipitation reconstruction from Lake Accesa (f, Peyron et al., 2011); northern hemisphere temperature between 30°N and 90°N (b, Marcott et al., 2013) and the hematite-stained grain (HSG, g, Bond et al., 2001) record, which is an indicator for iceberg discharge in the North Atlantic.

For flowstones Vic-III-1 and -3, it was not possible to construct a chronology for the thin Holocene growth sections based on the ²³⁰Th/U-data. The decrease in δ¹⁸O values from −4‰ to −6‰, which is observed in both cores and similar in magnitude with the δ¹⁸O signal of Vic-III-4, suggests that these sections of the flowstones correspond to the transition from the late Glacial to the Holocene. Similarly, the positive excursion in the δ¹³C values observed in both cores, which is comparable in magnitude and shape to the large excursion recorded in Vic-III-4, can reasonably be interpreted to be related to this phase. Thus, although we cannot independently establish the timing of the two stable isotope records, comparison with the Vic-III-4 record allows assigning them to the Holocene.

Speleothem δ¹³C values and trace elements

Speleothem δ¹³C values depend on various parameters, such as the concentration and δ¹³C value of soil CO₂, the δ¹³C value of the host rock, prior calcite precipitation (PCP), drip rate and cave ventilation. For flowstones, which are fed by water flowing long distances inside the cave, geochemical effects occurring inside the cave, such as variable degassing of CO₂ induced by strong cave ventilation (Johnson et al., 2006; Spötl et al., 2005), progressive precipitation of CaCO₃ along the flow path (Hansen et al., 2017; Mattey et al., 2010) and disequilibrium stable isotope fractionation (Mühlinghaus et al., 2009), may be of particular importance.

In the epikarst, dissolved CO₂ may degas and trigger PCP. Similarly, PCP may occur inside the cave en route to the flowstone. This process may strongly influence the δ¹³C values of the solution and result in increasing δ¹³C values (Dreybrodt and Scholz, 2011; Fairchild and Treble, 2009; Scholz et al., 2009). Thus, PCP and processes inside the cave may further amplify the general relationship of higher δ¹³C values during spring/summer drought conditions. Increased PCP also leads to increased Mg/Ca, Sr/Ca, and sometimes Ba/Ca ratios in the drip water because Mg, Sr and Ba incorporation into calcite is suppressed in comparison to Ca (Fairchild et al., 2000; Stoll et al., 2012). This may lead to positive correlations between speleothem δ¹³C and Mg/Ca, Sr/Ca and Ba/Ca ratios (Stoll et al., 2012). Except for the millennial-scale trend in Mg and Sr from the late Glacial to the early Holocene, we do not observe a positive correlation between Mg, Sr, and δ¹³C (Figure 4). We thus exclude that PCP had a strong effect on our flowstone δ¹³C record (Sinclair et al., 2012; Treble et al., 2015).

Most recharge at CV occurs during the rainy season from October to March. The vegetation growth period in eastern Spain depends on the plant species, but generally ranges from early spring to early summer and may be longer in case of summer rainfall (Camarero et al., 2015; Pasho et al., 2011). In case of long and dry summers and reduced precipitation in spring, vegetation density is negatively affected (Linares et al., 2011). This process is partly associated with long-term effects (Gazol et al., 2017). A prolonged reduction in spring and summer rainfall over several decades will thus result in a decrease in soil pCO₂ and soil microbial activity and be reflected in higher δ¹³C values. Thus, the δ¹³C values of the CV flowstone should be a sensitive proxy for past spring and summer drought even if aquifer recharge and flowstone growth predominantly occur during the autumn and winter season (Carrasco et al., 2006). Therefore, we interpret the δ¹³C values as a proxy for soil microbial activity and vegetation density reflecting changes in the length of the vegetation period and, thus, spring/summer drought.

The long-term trend in Mg and Sr from the late Glacial to the early Holocene (Figure 4) was observed in several speleothem records from Central Europe and attributed to glacial aeolian deposits, which are then progressively dissolved and washed into the cave (Fohlmeister et al., 2012; Mischel et al., 2017b). This interpretation may also be valid for south-eastern Spain and CV because loess deposits are present in the region, for instance, in the Granada basin, Andalusia, and on littoral plains of the eastern side of the country (Calvo et al., 2016; Coudé-Gaussen, 1990; Günster et al., 2001). In particular, Pleistocene dunes occur ca. 7 km standard error (SE) of CV (geological map, Llano del Beal, map sheet 978, Espinosa-Godoy et al., 1972). However, the long-term trend in Mg and Sr may also be related to incongruent dolomite dissolution of the host rock (Fairchild et al., 2000) during the drier glacial phase, when water residence times were longer. Due to increasing precipitation during the Bølling/Allerød (B/A, Figure 7), the karst aquifer was reactivated and the Mg and Sr concentration of the recharge water was progressively diluted. Unfortunately, we are not able to distinguish between the two processes (leaching of loess deposits vs incongruent dissolution of dolomite), and it cannot be excluded that the observed signal represents a mixture of both processes.

The δ¹³C values of the CV flowstone decrease from the bottom of the flowstone (B/A) until the onset of the YD (Figures 7 and 8). As for the δ¹⁸O values, the YD is indicated by a peak of 1.5‰ more positive δ¹³C values and more negative δ¹⁸O values, presumably reflecting both drier and colder climate conditions on the Iberian Peninsula (Baldini et al., 2015; Moreno et al., 2010) with reduced vegetation density. Subsequent to the YD, the δ¹³C values further decrease until 11 ka and then remain on a constant level of −10.5‰ indicating a well-developed C₃ vegetation above the cave and enough precipitation to maintain a permanent vegetation cover (Figure 8).

Figure 8.

δ¹⁸O (c) and δ¹³C (g) records of the CV flowstone compared with speleothem δ¹³C records from Ernesto Cave (d, Scholz et al., 2012), Chauvet Cave (e, Genty et al., 2006) and Grotta di Carburangeli (f, Frisia et al., 2006). Mean annual precipitation of N Africa at 31°N displays the timing of the northernmost extent of the African monsoon (h, Tierney et al., 2017). Furthermore, pollen-inferred reconstructions of summer precipitation (i, Peyron et al., 2011) and lake levels (j, Finsinger et al., 2010) and a pollen record from Lake Accesa, Italy (k, Drescher-Schneider et al., 2007) are shown. Finally, three pollen records (l, herbs and shrubs from Gorgo Basso, Sicily (Tinner et al., 2009); (m) Pinus (dark green) and deciduous forest (light green) pollen from Basa de la Mora, NE Spain (Pérez-Sanz et al., 2013), and (n) the amount of xerophytes in Siles Lake, S Spain (Carrión, 2002), are plotted. Ages with adjacent uncertainties are plotted on top (a) and June insolation gradient between 60° and 30°N is plotted below (b, Berger, 1978). For the locations of the individual records, see Figure 1. The y-axes for records c to g are inverted.

At 9.7 ± 0.3 ka, the δ¹³C values show a sharp shift towards higher values (Figure 8g) suggesting a strong decrease in precipitation during the vegetation period, which resulted in a decrease in vegetation density above the cave, lower soil pCO₂ and reduced microbiological activity. Since the growth rate of the flowstone remained at a similar level as before and after the δ¹³C excursion (Figure 6), a strong reduction in total annual recharge is unlikely. This is also supported by the two additional flowstone cores, which record calcite deposition – albeit very slowly – just during this δ¹³C excursion indicating no significant reduction in annual precipitation. Thus, we interpret the large excursion in δ¹³C values as a result of very dry conditions during the growth period of vegetation (spring to summer). Autumn to winter precipitation, in contrast, probably increased as reflected by the high speleothem growth rates preventing PCP in the aquifer. This interpretation of the δ¹³C data is supported by the crystal fabric. Elongated columnar fabric, as observed at the bottom of the flowstone (Figure 4), forms in case of constant drip rates and high Mg/Ca ratios (Frisia, 2015). Open columnar fabric, in contrast, forms as a result of lower drip rates and lower Mg/Ca ratios (Frisia, 2015). The synchronous occurrence of an open columnar fabric and the large peak in the δ¹³C values (Figure 4), thus, suggests drier conditions in the catchment of the cave in agreement with our interpretation of the δ¹³C record. The formation of the open columnar fabric may further be related to the progressive decrease of the Mg/Ca ratio from the late Glacial to the early Holocene (Figure 4). We emphasise that the change in crystal fabrics and crystal growth effects are unlikely to be the cause of the large δ¹³C excursion. Frisia (2015) compared δ¹³C and δ¹⁸O values for different speleothem crystal fabrics, and columnar, open columnar and elongated columnar fabrics show identical δ¹³C and δ¹⁸O values within uncertainty. Therefore, we interpret the large excursion in δ¹³C values to result from a reduction in meteoric spring/summer precipitation, which also invoked the change in crystal fabrics.

In summary, our δ¹³C record indicates dry spring/summer conditions and probably increased seasonality (drier spring/summer conditions and more humid autumn/winter conditions) in south-eastern Spain between 9.7 ± 0.3 and 7.8 ± 0.2 ka. Subsequently, the δ¹³C values decrease to approximately −10‰ indicating a return to a dense vegetation cover above the cave.

Speleothem δ¹⁸O values

Speleothem δ¹⁸O values are affected by cave air temperature, drip rate and the δ¹⁸O value of the drip water. The δ¹⁸O values of the drip water are influenced by the δ¹⁸O values of meteoric precipitation, which are strongly related to the source of precipitation, the temperature and humidity during evaporation in the moisture source region (Lachniet, 2009), the δ¹⁸O value of the ocean (LeGrande and Schmidt, 2006; Rozanski et al., 1993), the duration of moisture transport, the pathway of the storm tracks, rainfall amount and air temperature (McDermott, 2004).

In the Mediterranean, modern δ¹⁸O values can vary by ≈10‰ for a single precipitation event depending on the air masses and the source region of the moisture (Baldini et al., 2010; Celle-Jeanton et al., 2001; Moreno et al., 2014a; Sodemann and Stohl, 2009). Whereas moisture uptake for rainfall in the area of CV during the dry summer season almost exclusively occurs over the Western Mediterranean Sea, a substantial fraction of precipitation is derived from the Atlantic Ocean during the more humid winter season (Figure 5). Speleothem δ¹⁸O values at CV should thus reflect both local and more distant changes in the North Atlantic. This leads to strong variability of the δ¹⁸O values over the year, which is visible in the GNIP data (Figure S1, available online).

The local climate is characterised by a strong seasonality (Figure S2, available online). Considering that the vegetation period in the area lasts from spring to summer (see the section ‘Speleothem δ13C values and trace elements’), a substantial amount of spring and summer rainfall will be used up by the vegetation. Thus, recharge of the aquifer is most effective during winter times, whereas the contribution of summer rainfall is insignificant (Carrasco et al., 2006), and the δ¹⁸O values recorded by the flowstone should mainly reflect October to March rainfall.

Ice-volume corrected flowstone δ¹⁸O values progressively decrease from the onset of the Bølling (15 ka) towards the early Holocene (10 ka), interrupted by higher values during the Younger Dryas (YD; Figure 7). The timing of the YD in our record is in good agreement with the NGRIP δ¹⁸O record and SST reconstructions from the Alboran Sea (Cacho et al., 1999; Martrat et al., 2014; Figure 7). The general decreasing trend in the δ¹⁸O values of the CV flowstone from the late Glacial to the mid-Holocene shows a similar evolution as the NGRIP δ¹⁸O values, which are interpreted to reflect temperature changes in the North Atlantic realm (North Greenland Ice Core Project Members, 2004), and SST in the Alboran Sea (Cacho et al., 1999; Martrat et al., 2014; Figure 7a, c and d). A potential explanation for the observed negative relationship may be the effect of cave air temperature on the oxygen isotope fractionation between water and calcite of between −0.18‰/°C (Tremaine et al., 2011) and −0.24‰/°C (Kim and O’Neil, 1997), which is even imprinted on speleothem calcite if precipitation occurs under conditions of disequilibrium isotope fractionation (Mühlinghaus et al., 2009). The increase in SST from the late glacial to the early Holocene is approximately 7°C (Figure 7c). The corresponding decrease in speleothem δ¹⁸O values related to SST changes from the YD to 10.5 ± 0.35 ka is −1.5‰ is about −0.21‰/°C, which is in good agreement with the theoretical (equilibrium) values. This suggests that changes in the δ¹⁸O values of the CV flowstone on millennial and orbital timescales could be explained by temperature changes. However, if the speleothem δ¹⁸O signal was only related to temperature, the δ¹⁸O values of the drip water would have remained constant from the late Glacial to the Holocene, which is very unlikely. Thus, we cannot exclude that changes in atmospheric circulation patterns or moisture source, in particular from the late Glacial to the mid-Holocene, played a role as well. The general complexity of processes controlling speleothem δ¹⁸O values is also highlighted by detailed cave monitoring studies performed in northern Spain, where individual processes affecting the δ¹⁸O signal could not be disentangled although these sites mainly receive moisture from the Atlantic Ocean (Bartolomé et al., 2015).

High flowstone growth rates between 10.5 ± 0.35 and 7.0 ± 0.5 ka (compared to the older section of the flowstone) may be related to increased rainfall and recharge, which can cause increasing drip rates and – as a result – decreasing speleothem δ¹⁸O values (Mühlinghaus et al., 2009; Scholz et al., 2009; Stoll et al., 2015). In addition, rainfall δ¹⁸O values in the Mediterranean may also be related to the amount effect (Ayalon et al., 1998; Bar-Matthews et al., 2003; Lachniet, 2009). Thus, periods of reduced rainfall could be reflected by higher δ¹⁸O values and vice versa. Although modern precipitation δ¹⁸O data do not show a significant relation to rainfall amount on inter-annual timescales, we cannot exclude that rainfall amount may have affected our speleothem δ¹⁸O record on decadal and longer timescales.

In summary, although we cannot disentangle the processes affecting the flowstone δ¹⁸O values and the interpretation of the δ¹⁸O record of the CV flowstone remains challenging, we assume that the δ¹⁸O values of the CV flowstone on orbital timescales and during the late Glacial might be related to temperature changes on the North Atlantic, which is supported by the good agreement with the SST records. However, the influence of other parameters, such as seasonality of rainfall, rainfall amount and changes in moisture source(s), cannot be excluded. For the period from 10.5 ± 0.35 to 7.8 ± 0.2 ka, higher growth rates coincide with lower δ¹⁸O values, which suggest an additional influence of changes in the amount of autumn/winter precipitation on millennial timescales.

Implications for Western Mediterranean climate between 15 and 7 ka

15–11.7 ka

The warming trend from the late Glacial to the early Holocene is reflected in the evolution towards lower δ¹⁸O values in the CV flowstone, which parallels the evolution in the NGRIP ice core and SST in the Alboran Sea (Figure 7). This trend is also observed in several other terrestrial records (speleothems, pollen) from Europe and the Mediterranean south of 45°N (García-Alix et al., 2014; Moreno et al., 2014b). In addition, other speleothem data from the Western Mediterranean region and the Iberian Peninsula exhibit hiatuses during the cold interval prior to the B/A (Constantin et al., 2007; Frisia et al., 2005; Genty et al., 2006; Moreno et al., 2010, 2017) suggesting limited recharge as a result of reduced rainfall. The onset of growth of the CV flowstone during the B/A coincides with the recovery of vegetation in the Mediterranean (Allen et al., 2002; Fletcher et al., 2010; Fletcher and Sánchez Goñi, 2008; Tzedakis, 2005), as a consequence of rising temperatures and humidity levels. This is also reflected in the δ¹³C values of the CV flowstone, which are approximately −10‰ prior to the YD, suggesting relatively humid conditions and a well-developed vegetation above the cave during the B/A. Decreasing aridity towards the early Holocene is also recorded in the sediments of Lake Salines (Burjachs et al., 2016).

The YD is clearly visible in the CV flowstone record by an increase in both the δ¹³C and the δ¹⁸O values, forming a trough with less negative values (Figures 7 and 8). This suggests both colder and drier conditions in south-eastern Spain during the YD, which is confirmed by the decrease in SST by approximately 4°C and the decline in pollen of woody taxa in Italian lake sediments by about 20% (Allen et al., 2002). Drier conditions during the YD were also observed in other speleothem δ¹³C records from southern Europe (e.g. Genty et al., 2006; Moreno et al., 2010; Figure 8e). Due to the relatively low temporal resolution of our record, we do not observe two phases of the YD, as has been reported for northern Spain (Baldini et al., 2015).

11.7–9.7 ka

The δ¹⁸O values of the CV flowstone progressively decrease from the YD to the maximum June insolation gradient (Figure 8a). This trend towards more negative δ¹⁸O values coincides with a positive trend in the NGRIP ice core δ¹⁸O record (Rasmussen et al., 2006) and SST in the Alboran Sea (Cacho et al., 1999; Martrat et al., 2014; Figure 7a and c). This decreasing trend in δ¹⁸O values is accompanied by a trend towards lower δ¹³C values, suggesting a further increase in vegetation density and spring/summer precipitation until the Holocene climate optimum. These trends can be observed in several records from the Mediterranean. For instance, the substantial decrease of xerophytes (Figure 8n) and the shift to 70% of Pinus pollen in south-eastern Spain during this period indicates warmer and wetter conditions (Carrión, 2002; Carrión et al., 2010). A similar trend in pollen assemblages was found in the Alboran Sea with increasing woody pollen taxa from the YD to 9.7 ± 0.3 ka (Combourieu Nebout et al., 2009) and total pollen amount (Fletcher et al., 2013).

9.7–7 ka

Many climate records from the Iberian Peninsula are available for the time period between 9.7 and 7 ka. We first present a comparison with records from the cave region, followed by a comparison with records across the Iberian Peninsula and finally discuss the circum-Western Mediterranean.

The CV speleothem δ¹³C values show a large positive excursion between 9.7 ± 0.3 and 7.8 ± 0.2 ka, while the δ¹⁸O record reaches a plateau at −6.6 ‰ around 10.5 ± 0.35 ka. As discussed above, we interpret this excursion in the δ¹³C values as a result of reduced vegetation density due to very dry conditions during the growth period of vegetation in spring and summer. Autumn/winter precipitation, in contrast, must have remained on a similar level or even increased, preventing PCP in the aquifer and resulting in a similar speleothem growth rate as for the previous section.

This interval of strong spring and summer droughts was also observed in several other terrestrial records from the eastern and south-eastern Iberian Peninsula. For instance, a pollen record from Lake Siles, in the mountains to the west near CV, shows an increase of the proportion of xerophytes (Carrión, 2002; Jalut et al., 2000) during the δ¹³C excursion (Figure 8n), suggesting drier conditions during the vegetation season (spring/summer). Similarly, pollen records from fluvial deposits in the region of Almeria indicate a decrease of vegetation and microbiological activity in the soil due to less precipitation (Pantaléon-Cano et al., 2003). Further west, a speleothem record from Refugio Cave (Walczak et al., 2015) indicates temperate climate with rainfall throughout the year until 7 ka. However, it cannot be ruled out that this record predominantly represents winter precipitation. In addition, the speleothem records from Refugio Cave and CV both show a growth interruption subsequent to 7 ka, which is evidence for a dry regional climate. However, high speleothem growth rates between 6.9 and 6.4 ka observed at Nerja Cave (McMillan, 2006) demonstrate the complexity of the climate system in this region and may – as the Refugio cave site – also have been under the predominant influence of the NAO pattern (Comas-Bru and McDermott, 2014).

Further north of CV and more inland, the eastern part of the Iberian Peninsula is still under the same precipitation regime as CV, which is strongly influenced by the WeMO in the present-day climate (Martin-Vide and Lopez-Bustins, 2006). Salines Lake documents sharp aridity events at 10 and 9.5 ka (Burjachs et al., 2016), although the occurrence of ostracods indicates a permanent lake level between 10 and 8 ka. The adjacent Villena Lake shows long-term aridity events at around 10 ka and from approx. 9 to 7 ka (Jones et al., 2018), which coincide very well with the large δ¹³C excursion in our speleothem record. Although the behaviour of Salines lake level appears contradictive to our speleothem δ¹³C record, this can be explained by the constant feed of groundwater to Lake Salines, as observed during the present-day situation (Burjachs et al., 2016). Lake Villarquemado shows pronounced abundance of xerophyte taxa from 10 to 9.3 ka, indicating a dry period (Aranbarri et al., 2014). This dry period might be shorter due to the high elevation and the less efficient evapotranspiration at this altitude (≈ 1000 m a.s.l.) (Aranbarri et al., 2014). Close to this lake, a late winter/early spring aridity has been observed in a speleothem record from central Spain (Molinos Cave, Moreno et al., 2017) showing increased δ¹³C values (9.5–7.1 ka) and even an interruption of speleothem growth (between 9.4 and 8.4 ka). In summary, several records suggest a consistently dry period (with enhanced seasonality) during the δ¹³C excursion of the CV record in south-eastern Spain.

In northern Spain, a short and pronounced stalagmite growth phase between 10 and ca. 7 ka ending no later than at 4 ka suggests unfavourable conditions for stalagmite growth due to a regional drought (Stoll et al., 2013). This is in agreement with our record, since the positive WeMO index correlates positively with precipitation in this region, but negatively in eastern Spain and at CV, especially at the end of winter/early spring (Martin-Vide and Lopez-Bustins, 2006). In the eastern range of the Cantabrian Mountains, high δ¹³C values in the Kaite Cave record (9.5–8.2 ka, Domínguez-Villar et al., 2017) might reflect low vegetation density. This period coincides with high δ¹³C values and a hiatus in the speleothem record from Molinos Cave between 9.4 and 8.4 ka, which indicates the end of winter/early spring aridity (Moreno et al., 2017), although this record indicates dry conditions lasting until 4.8 ka. In the Pyrenees, a sharp increase of Pinus and deciduous forest pollen in the lake record at Basa de la Mora (Pérez-Sanz et al., 2013; Figure 8m) also indicates strong seasonality with dry summers and humid winters. This is supported by sediments in the Ebro basin suggesting a thin vegetation cover and semiarid conditions with extensive erosion (Bastida et al., 2013). Arid conditions with pronounced rainfall seasonality are indicated by tufa deposits from Marcelino (South Pyrenees, Pellicer et al., 2016) and are also reflected by a high percentage of non-arboreal pollen until 8 ka at La Garrotxa (NE Spain, Piqué et al., 2018). Although lake-level reconstructions from this region suggest high rainfall, these may again be biased by seasonal precipitation and enhanced melting of ice and snow in the Pyrenees, which also supply the lowlands, leading to the formation of riparian environments (González-Sampériz et al., 2017).

In summary, despite a seemingly contrasting signal to the lake-level records from the Iberian Peninsula (Morellón et al., 2018), we conclude pronounced spring and summer drought and more humid autumn/winter conditions for Iberian Peninsula between approximately 10 and 7 ka. The apparent disagreement with some lake-level records may be related to a seasonal bias in the lake-level reconstructions (González-Sampériz et al., 2017), different altitudes of the catchment areas and/or continuous feeding groundwater (e.g. Lake Salines; Burjachs et al., 2016). We therefore suggest that the lake-level records presented by Morellón et al. (2018) predominantly reflect an autumn/winter rainfall signal (i.e. the main period of precipitation in south-eastern Spain), but did not record the dry spring/summer conditions reflected in our δ¹³C record.

In a larger circum-West Mediterranean context, in northern Italy, higher δ¹³C values at Ernesto Cave between 8 and 7.5 ka indicate relatively dry conditions (Scholz et al., 2012; Figure 8d). A coeval speleothem record from Corchia Cave (Zanchetta et al., 2007; Figure 7e) shows low δ¹⁸O values, reflecting increased seasonality with drier summers and wetter winters. Lake-level reconstructions (Finsinger et al., 2010) based on pollen assemblages from Lake Accesa (Drescher-Schneider et al., 2007) in central Italy indicate a low lake level (Figure 8j; Peyron et al., 2011) due to reduced summer precipitation, but enhanced winter precipitation during the period of high CV flowstone δ¹³C values. In addition, the increase of Mediterranean pollen taxa at the same site (Drescher-Schneider et al., 2007) indicates summer drought (Figure 8k). Thus, the dry period recorded in the CV flowstone is also observed in several archives from the north-eastern part of the Western Mediterranean (Ligurian Sea).

In Sicily, herb and shrub pollen are abundant (Gorgo Basso; Tinner et al., 2009, Figure 8l), and the adjacent Lago Preola shows a similar pattern from dominant shrub to tree pollen assemblages at around 7 ka (Calò et al., 2012; Curry et al., 2016; Magny et al., 2011). A strong seasonality with high winter precipitation has also been suggested for northern Sicily based on high δ¹³C values in a speleothem from Grotta di Carburangeli (Frisia et al., 2006; Figure 8f). These records document a humid winter and dry summer interval in the south-eastern part of the Western Mediterranean (Tyrrhenian Sea). However, at higher elevations in Sicily, Lake Pergusa sediments record similar dry climate conditions with a rapid shift at already 8.9 ka towards a wetter climate and the establishment of a forest vegetation (Sadori and Narcisi, 2001).

Whereas in the Western Mediterranean the spring/summer climate was drier and winters were wetter during 9.7 ± 0.3 − 7.8 ± 0.2 ka, climate records from Morocco (Wassenburg et al., 2016; Zielhofer et al., 2017) and Tunisia (Genty et al., 2006) show a different pattern. On multi-millennial timescales, the Lake Sidi Ali δ¹⁸O record shows a drying trend from the early to late-Holocene, lacking evidence of a dry phase between 9.7 ± 0.3 and 7.8 ± 0.2 ka. This difference to the climate at CV can be explained by the strong influence of the NAO in North Morocco, as opposed to south-eastern Spain, where the relation to NAO is insignificant (Table S3, available online). The Alboran Sea pollen records that are located between south to south-eastern Spain and Morocco (Cacho et al., 2001; Combourieu Nebout et al., 2009) likely reflect a mixture of both regions. Nevertheless, the increase in the abundance of pollen from plants adapted to dry conditions is strong evidence for several dry periods interrupted by short humid phases during the early to mid-Holocene.

At the same time, in the Eastern Mediterranean, the deposition of a sapropel below 1000 m water depth indicates increased precipitation, but mainly during winter (Emeis et al., 2000; Rohling et al., 2015). Evidence for anoxic conditions in the Western Mediterranean due to increased river runoff were reported as well (Jimenez-Espejo et al., 2007). This might be related to increased autumn/winter precipitation as shown by the compilation of climate records in this study.

The implications for large-scale atmospheric circulation

The dry spring/summer conditions in the Western Mediterranean realm between 9.7 ± 0.3 and 7.8 ± 0.2 ka occurred when the North African monsoon reached its northernmost position, i.e. the so-called African humid period (deMenocal et al., 2000). Pollen records from the Sahara and the Sahel zone show high abundances of grass and trees (Hély et al., 2014; Tjallingii et al., 2008) indicating enhanced precipitation in northern Africa (Tierney et al., 2017). Especially the northernmost oceanic core off the Moroccan coast (31°N, Figure 1) shows a well defined wet period between 10 and 7.5 ka (Figure 8h; Tierney et al., 2017). The North African Monsoon is strongly related to the position of the Intertropical Convergence Zone (ITCZ), which together with the Azores Subtropical High is part of the Northern Hemisphere Hadley cell circulation. A change in either the strength or the position of the ITCZ will thus also affect the Azores Subtropical High. However, western Mediterranean precipitation is not only controlled by the summer season, and we suggest that the seemingly contrasting precipitation patterns derived from the climate records discussed in this paper can only be explained by taking into account all seasons. The timing and evolution of the June insolation gradient between 60°N and 30°N is in good agreement with long-term variation of the δ¹⁸O values of the CV flowstone. Low winter insolation versus high summer insolation induced a high seasonality in precipitation, with enhanced winter precipitation (Kutzbach et al., 2014). Pollen-based climate reconstructions (Davis and Brewer, 2009; Davis et al., 2003; Fletcher et al., 2010; Mauri et al., 2015) suggest lower summer and lower winter temperatures (ca. 2°C) between 10 and 7 ka in the Western Mediterranean/SW Europe and a reduced positive temperature anomaly in the northern hemisphere (Marcott et al., 2013; Figure 7b). In contrast, SSTs in the Alboran Sea record peak values (≈ 20°C, Figure 7c; Cacho et al., 1999; Martrat et al., 2014). Lower summer temperatures may suppress convection and therefore precipitation during summer. However, annual precipitation at CV was higher from 9.7 ± 0.3 to 7.8 ± 0.2 ka (Mauri et al., 2015), which might be explained by enhanced seasonality. In combination with high SSTs, cold low pressure systems in autumn and winter (Figure 1a and d) enter the Western Mediterranean and force enhanced convection. This should be especially pronounced in the coastal region as modelled by Mauri et al. (2015).

Figure 1 shows sea-level pressure patterns for different seasons for today and for the time period between 9.7 ± 0.3 and 7.8 ± 0.2 ka, interpreted from the climate records discussed in the text. During the time when the North African Monsoon reached its northernmost position, the Azores High has been shown to be weaker during the winter season (Tierney et al., 2017) associated with a southward shift of the Westerlies. We propose that this not only results in more negative NAO-like conditions during the early Holocene as suggested by Deininger et al. (2016), but also in more frequent shifts of a weakened Icelandic Low towards the Iberian Peninsula extending into the Mediterranean (Kutzbach et al., 2014; Figure 1a). According to the southward shifted low, an intensified WeMO-type pattern (Martin-Vide and Lopez-Bustins, 2006) might become an increasingly important factor for precipitation in the eastern Iberian Peninsula due to elevated SSTs at this time. Precipitation patterns between northern Italy and the eastern Iberian Peninsula confirm the importance of the WeMO as a dominant pattern in the early Holocene. Due to the southerly displacement of the Westerlies, Icelandic Lows predominantly entered the Western Mediterranean, inducing a WeMO structure in autumn/winter with elevated precipitation.

During the boreal spring season, northern hemisphere insolation increases. We argue that during the spring season, the ITCZ and the Azores Subtropical High were (analogically to the summer situation) located north compared to the present day, which explains the shortened growth season for vegetation and dry spring conditions in the Western Mediterranean. The Subtropical High moved roughly to the present-day summer position. Riparian forest sites might benefit from high river discharge and high lake levels due to high winter precipitation and snowmelt during spring.

During the summer season, northern hemisphere insolation was higher than today, which led to a more northern position of the ITCZ (Chiang and Friedman, 2012; Tierney et al., 2017) and the adjacent Subtropical High (Schneider et al., 2014). A recent modelling study shows a stable low pressure over northwest Africa during the time of maximum northward extent of the North African Monsoon (Figure 1c, Pausata et al., 2016). Therefore, the Western Mediterranean is under a high pressure cell with dry north-easterly winds, which might suppress local convection (Walczak et al., 2015).

During autumn, insolation decreases, and the ITCZ and the adjacent Subtropical High move south. Hence, the Westerlies can enter the Western Mediterranean, and their relatively cold air in combination with high SSTs leads to strong convection and intense rainfall during autumn. This intense rainfall leads to aquifer recharge at CV.

Many models focus on winter and summer precipitation (Mauri et al., 2015), but spring and autumn are not discussed. We propose that changes in precipitation during autumn and spring are most important for the annual precipitation budget in the eastern part of the Iberian Peninsula between 9.7 ± 0.3 and 7.8 ± 0.2 ka, although we cannot exclude a slight decrease in winter precipitation (Jin et al., 2012) as reported from Molinos Cave (Moreno et al., 2017).

Conclusion

We present a stable isotope and trace element records from flowstones of CV, south-eastern Spain, and based on several Holocene climate records from the Western Mediterranean, we present an estimation on atmospheric conditions for the time between 9.7 ± 0.3 and 7.8 ± 0.2 ka.

On shorter timescale, we cannot disentangle the different processes influencing the speleothem δ¹⁸O signal. However, on multi-millennial to orbital timescales, δ¹⁸O values are mainly influenced by North Atlantic temperature changes and seasonality. In general, they represent autumn/winter precipitation δ¹⁸O values, whereas δ¹³C values reflect the vegetation density and microbiological activity in the soil and are therefore a spring to summer signal.

Between the onset of the B/A and the early Holocene, δ¹⁸O values progressively decrease, suggesting an influence by increasing temperature and precipitation in south-eastern Spain in accordance with SSTs in the Alboran Sea. At the same time, decreasing δ¹³C values indicate progressively increasing precipitation and vegetation density. This trend is interrupted by colder and drier conditions during the YD.

Between 9.7 ± 0.3 and 7.8 ± 0.2 ka, the δ¹³C values show a large positive excursion indicating a strong reduction in vegetation density and, thus, very dry spring/summer conditions. At the same time, autumn/winter precipitation probably increased indicating enhanced seasonality. This dry period with enhanced seasonality is also observed in several other climate records from the Western Mediterranean region, showing that the entire Western Mediterranean realm (Spain, Italy) experienced spring/summer droughts during this period. We suggest that this was related to a more northward positioned and/or stronger Azores Subtropical High. This occurred at the same time when the North African Monsoon reached its northernmost position, which suggests that this is a teleconnection pattern related to the Hadley cell circulation. The increase in autumn/winter precipitation was also observed in the Eastern Mediterranean, which has been suggested as a cause of anoxic conditions and the formation of sapropels.

Supplemental Material

Supplement_Budsky – Supplemental material for Speleothem δ13C record suggests enhanced spring/summer drought in south-eastern Spain between 9.7 and 7.8 ka – A circum-Western Mediterranean anomaly?

Supplemental material, Supplement_Budsky for Speleothem δ13C record suggests enhanced spring/summer drought in south-eastern Spain between 9.7 and 7.8 ka – A circum-Western Mediterranean anomaly? by Alexander Budsky, Denis Scholz, Jasper A Wassenburg, Regina Mertz-Kraus, Christoph Spötl, Dana FC Riechelmann, Luis Gibert, Klaus Peter Jochum and Meinrat O Andreae in The Holocene

Footnotes

Acknowledgements

We thank Andrés Ros, the team of CENM-naturaleza and the city of Cartagena for the support of the sampling and opportunity to work in Cueva Victoria. The assistance of Beate Schwager in the geochemistry lab (Max Planck Institute for Chemistry, Mainz) and Manuela Wimmer in the isotope laboratory (University of Innsbruck) is highly appreciated. The authors gratefully acknowledge the NOAA Air Resources Laboratory (ARL) for the provision of the HYSPLIT transport and dispersion model and/or READY website () used in this publication. Constructive and detailed comments by the editor, Giles Young, and two anonymous reviewers were very helpful to improve the manuscript.

Funding

This work was funded by the German Research Foundation (ME3761/2-1 to R. Mertz-Kraus and SCHO 1274/9-1 to D. Scholz).

Supplemental material

Supplemental material for this article is available online.

ORCID iD

Meinrat O Andreae

References

Agencia Estatal de Meteorología (2011) Atlas climático ibérico: Temperatura del aire y precipitación (1971–2000) = Atlas climático ibérico: Temperatura do ar e precipitação (1971–2000) [Iberian Climate Atlas: Air Temperature and Precipitation (1971–2000)]. Madrid: Instituto Nacional de Meteorología.

Allan

Ansell

(2006) A new globally complete monthly historical gridded mean sea level pressure dataset (HadSLP2): 1850–2004. Journal of Climate 19(22): 5816–5842.

Allen

JRM

Brandt

Brauer

et al . (1999) Rapid environmental changes in southern Europe during the last glacial period. Nature 400: 740–743.

Allen

JRM

Watts

McGee

et al . (2002) Holocene environmental variability – The record from Lago Grande di Monticchio, Italy. Quaternary International 88(1): 69–80.

Araguas-Araguas

Diaz Teijeiro

(2005) Isotope composition of precipitation and water vapour in the Iberian Peninsula: First results of the Spanish Network of Isotopes in Precipitation. In: International Atomic Energy Agency (ed.) Isotopic Composition of Precipitation in the Mediterranean Basin in Relation to Air Circulation Patterns and Climate: Final Report of a Coordinated Research Project, 2000–2004. IAEA-TECDOC 1453. Vienna: International Atomic Energy Agency, pp. 173–190.

Aranbarri

González-Sampériz

Valero-Garcés

et al . (2014) Rapid climatic changes and resilient vegetation during the Lateglacial and Holocene in a continental region of south-western Europe. Global and Planetary Change 114: 50–65.

Ayalon

Bar-Matthews

Kaufman

(2002) Climatic conditions during marine oxygen isotope stage 6 in the eastern Mediterranean region from the isotopic composition of speleothems of Soreq Cave, Israel. Geology 30(4): 303–306.

Ayalon

Bar-Matthews

Sass

(1998) Rainfall-recharge relationships within a karstic terrain in the Eastern Mediterranean semi-arid region, Israel: δ¹⁸O and δD characteristics. Journal of Hydrology 207(1–2): 18–31.

Bajo

Hellstrom

Frisia

et al . (2016) ‘Cryptic’ diagenesis and its implications for speleothem geochronologies. Quaternary Science Reviews 148: 17–28.

10.

Baldini

McDermott

Baldini

JUL

et al . (2010) An investigation of the controls on Irish precipitation δ¹⁸O values on monthly and event timescales. Climate Dynamics 35(6): 977–993.

11.

Baldini

McDermott

Baldini

JUL

et al . (2015) Regional temperature, atmospheric circulation, and sea-ice variability within the Younger Dryas Event constrained using a speleothem from northern Iberia. Earth and Planetary Science Letters 419: 101–110.

12.

Bard

Delaygue

Rostek

et al . (2002) Hydrological conditions over the western Mediterranean basin during the deposition of the cold Sapropel 6 (ca. 175 kyr BP). Earth and Planetary Science Letters 202(2): 481–494.

13.

Bardají

Goy

Zazo

et al . (2009) Sea level and climate changes during OIS 5e in the Western Mediterranean. Geomorphology 104(1–2): 22–37.

14.

Bar-Matthews

Ayalon

Gilmour

et al . (2003) Sea–land oxygen isotopic relationships from planktonic foraminifera and speleothems in the Eastern Mediterranean region and their implication for paleorainfall during interglacial intervals. Geochimica et Cosmochimica Acta 67(17): 3181–3199.

15.

Barnston

Livezey

(1987) Classification, seasonality and persistence of low-frequency atmospheric circulation patterns. Monthly Weather Review 115(6): 1083–1126.

16.

Bartolomé

Moreno

Sancho

et al . (2015) Hydrological change in Southern Europe responding to increasing North Atlantic overturning during Greenland Stadial 1. Proceedings of the National Academy of Sciences 112(21): 6568–6572.

17.

Bastida

Osácar

Sancho

et al . (2013) Environmental changes during the Upper Pleistocene–Holocene in Mediterranean NE Spain as recorded by the mineralogy and geochemistry of alluvial records. Quaternary International 302: 3–19.

18.

Berger

(1978) Long-term variations of daily insolation and Quaternary climatic changes. Journal of the Atmospheric Sciences 35(12): 2362–2367.

19.

Boch

Spötl

Kramers

(2009) High-resolution isotope records of early Holocene rapid climate change from two coeval stalagmites of Katerloch Cave, Austria. Quaternary Science Reviews 28(23–24): 2527–2538.

20.

Bond

Kromer

Beer

et al . (2001) Persistent solar influence on North Atlantic climate during the Holocene. Science 294(5549): 2130–2136.

21.

Brauer

Allen

JRM

Mingram

et al . (2007) Evidence for last interglacial chronology and environmental change from Southern Europe. Proceedings of the National Academy of Sciences 104(2): 450–455.

22.

Breecker

Payne

Quade

et al . (2012) The sources and sinks of CO₂ in caves under mixed woodland and grassland vegetation. Geochimica et Cosmochimica Acta 96: 230–246.

23.

Budsky

Scholz

Gibert

et al . (2015) ²³⁰Th/U-dating of the Cueva Victoria flowstone sequence: Preliminary results and paleoclimate implications. In: Gibert

Ferràndez-Canadell

(eds) Geología y Paleontología de Cueva Victoria [Geology and Paleontology of Cueva Victoria] (Mastia 11–13). Cartagena: Ayuntamiento de Cartagena, pp. 101–109.

24.

Burjachs

Jones

Giralt

et al . (2016) Lateglacial to Early Holocene recursive aridity events in the SE Mediterranean Iberian Peninsula: The Salines playa lake case study. Quaternary International 403: 187–200.

25.

Cacho

Grimalt

Canals

et al . (2001) Variability of the western Mediterranean Sea surface temperature during the last 25,000 years and its connection with the Northern Hemisphere climatic changes. Paleoceanography 16(1): 40–52.

26.

Cacho

Grimalt

Pelejero

et al . (1999) Dansgaard-Oeschger and Heinrich event imprints in Alboran Sea paleotemperatures. Paleoceanography 14(6): 698–705.

27.

Calò

Henne

Curry

et al . (2012) Spatio-temporal patterns of Holocene environmental change in southern Sicily. Palaeogeography, Palaeoclimatology, Palaeoecology 323–325: 110–122.

28.

Calvo

Sánchez

Acosta

et al . (2016) Granulometrical, mineralogical and geochemical characterization of loess deposits in the Tajo Basin. Quaternary International 407: 14–28.

29.

Camarero

Gazol

Tardif

et al . (2015) Attributing forest responses to global-change drivers: Limited evidence of a CO₂ -fertilization effect in Iberian pine growth. Journal of Biogeography 42(11): 2220–2233.

30.

Carrasco

Andreo

Liñán

et al . (2006) Contribution of stable isotopes to the understanding of the unsaturated zone of a carbonate aquifer (Nerja Cave, southern Spain). Comptes Rendus Geoscience 338(16): 1203–1212.

31.

Carrión

(2002) Patterns and processes of Late Quaternary environmental change in a montane region of southwestern Europe. Quaternary Science Reviews 21(18-19): 2047–2066.

32.

Carrión

Fernández

González-Sampériz

et al . (2010) Expected trends and surprises in the Lateglacial and Holocene vegetation history of the Iberian Peninsula and Balearic Islands. Review of Palaeobotany and Palynology 162(3): 458–475.

33.

Celle-Jeanton

Travi

Blavoux

(2001) Isotopic typology of the precipitation in the Western Mediterranean Region at three different time scales. Geophysical Research Letters 28(7): 1215–1218.

34.

Cerling

Wang

Quade

(1993) Expansion of C4 ecosystems as an indicator of global ecological change in the late Miocene. Nature 361: 344–345.

35.

Cheng

Edwards

Hoff

et al . (2000) The half-lives of uranium-234 and thorium-230. Chemical Geology 169(1–2): 17–33.

36.

Chiang

JCH

Friedman

(2012) Extratropical cooling, interhemispheric thermal gradients, and tropical climate change. Annual Review of Earth and Planetary Sciences 40(1): 383–412.

37.

Comas-Bru

McDermott

(2014) Impacts of the EA and SCA patterns on the European twentieth century NAO-winter climate relationship. Quarterly Journal of the Royal Meteorological Society 140(679): 354–363.

38.

Combourieu Nebout

Peyron

Dormoy

et al . (2009) Rapid climatic variability in the west Mediterranean during the last 25 000 years from high resolution pollen data. Climate of the Past 5(3): 503–521.

39.

Constantin

Bojar

A-V

Lauritzen

S-E

et al . (2007) Holocene and Late Pleistocene climate in the sub-Mediterranean continental environment: A speleothem record from Poleva Cave (Southern Carpathians, Romania). Palaeogeography, Palaeoclimatology, Palaeoecology 243(3–4): 322–338.

40.

Coudé-Gaussen

(1990) The loess and loess-like deposits along the sides of the western Mediterranean Sea: Genetic and palaeoclimatic significance. Quaternary International 5: 1–8.

41.

Curry

Henne

Mesquita-Joanes

et al . (2016) Holocene paleoclimate inferred from salinity histories of adjacent lakes in southwestern Sicily (Italy). Quaternary Science Reviews 150: 67–83.

42.

Davis

BAS

Brewer

(2009) Orbital forcing and role of the latitudinal insolation/temperature gradient. Climate Dynamics 32(2): 143–165.

43.

Davis

BAS

Brewer

Stevenson

et al . (2003) The temperature of Europe during the Holocene reconstructed from pollen data. Quaternary Science Reviews 22(15–17): 1701–1716.

44.

Deininger

Werner

McDermott

(2016) North Atlantic Oscillation controls on oxygen and hydrogen isotope gradients in winter precipitation across Europe; implications for palaeoclimate studies. Climate of the Past 12(11): 2127–2143.

45.

deMenocal

Ortiz

Guilderson

et al . (2000) Abrupt onset and termination of the African Humid Period. Quaternary Science Reviews 19(1-5): 347–361.

46.

Denniston

González

Asmerom

et al . (2000) Speleothem carbon isotopic records of Holocene environments in the Ozark Highlands, USA. Quaternary International 67(1): 21–27.

47.

Domínguez-Villar

Wang

Krklec

et al . (2017) The control of the tropical North Atlantic on Holocene millennial climate oscillations. Geology 45(4): 303–306.

48.

Dorale

Edwards

Ito

et al . (1998) Climate and vegetation history of the midcontinent from 75 to 25 ka: A speleothem record from Crevice Cave, Missouri, USA. Science 282(5395): 1871–1874.

49.

Drescher-Schneider

Beaulieu

J-L

de Magny

et al . (2007) Vegetation history, climate and human impact over the last 15,000 years at Lago dell’Accesa (Tuscany, Central Italy). Vegetation History and Archaeobotany 16(4): 279–299.

50.

Dreybrodt

Scholz

(2011) Climatic dependence of stable carbon and oxygen isotope signals recorded in speleothems: From soil water to speleothem calcite. Geochimica et Cosmochimica Acta 75(3): 734–752.

51.

Drysdale

Zanchetta

Hellstrom

et al . (2006) Late Holocene drought responsible for the collapse of Old World civilizations is recorded in an Italian cave flowstone. Geology 34(2): 101–104.

52.

Emeis

K-C

Struck

Schulz

H-M

et al . (2000) Temperature and salinity variations of Mediterranean Sea surface waters over the last 16,000 years from records of planktonic stable oxygen isotopes and alkenone unsaturation ratios. Palaeogeography, Palaeoclimatology, Palaeoecology 158(3–4): 259–280.

53.

Espinosa-Godoy

Martin Vivaldi

Pérez Rojas

(1972) Mapa geologico de Espana: Llano del Beal, 978. Madrid: Instituto geologgico y minero de Espana.

54.

Fairchild

Baker

(2012) Speleothem Science: From Process to Past Environments. Oxford; Hoboken, NJ: Wiley.

55.

Fairchild

Treble

(2009) Trace elements in speleothems as recorders of environmental change. Quaternary Science Reviews 28(5–6): 449–468.

56.

Fairchild

Borsato

Tooth

et al . (2000) Controls on trace element (Sr–Mg) compositions of carbonate cave waters: Implications for speleothem climatic records. Chemical Geology 166(3–4): 255–269.

57.

Fensterer

Scholz

Hoffmann

et al . (2010) ²³⁰Th/U-dating of a Late-Holocene low uranium speleothem from Cuba. IOP Conference Series: Earth and Environmental Science 9: 12015.

58.

Ferràndez-Cañadell

Ribot

Gibert

(2014) New fossil teeth of Theropithecus oswaldi (Cercopithecoidea) from the Early Pleistocene at Cueva Victoria (SE Spain). Journal of Human Evolution 74: 55–66.

59.

Finsinger

Colombaroli

Beaulieu

J-L

et al . (2010) Early to mid-Holocene climate change at Lago dell’Accesa (central Italy): Climate signal or anthropogenic bias? Journal of Quaternary Science 25(8): 1239–1247.

60.

Fletcher

Sánchez Goñi

(2008) Orbital- and sub-orbital-scale climate impacts on vegetation of the western Mediterranean basin over the last 48,000 yr. Quaternary Research 70(3): 451–464.

61.

Fletcher

Debret

Sanchez Goñi

(2013) Mid-Holocene emergence of a low-frequency millennial oscillation in western Mediterranean climate: Implications for past dynamics of the North Atlantic atmospheric westerlies. The Holocene 23(2): 153–166.

62.

Fletcher

Sanchez Goñi

Peyron

et al . (2010) Abrupt climate changes of the last deglaciation detected in a Western Mediterranean forest record. Climate of the Past 6(2): 245–264.

63.

Fohlmeister

Scholz

Kromer

et al . (2011) Modelling carbon isotopes of carbonates in cave drip water. Geochimica et Cosmochimica Acta 75(18): 5219–5228.

64.

Fohlmeister

Schröder-Ritzrau

Scholz

et al . (2012) Bunker Cave stalagmites: An archive for central European Holocene climate variability. Climate of the Past 8(5): 1751–1764.

65.

Frisia

(2015) Microstratigraphic logging of calcite fabrics in speleothems as tool for palaeoclimate studies. International Journal of Speleology 44(1): 1–16.

66.

Frisia

Borsato

Mangini

et al . (2006) Holocene climate variability in Sicily from a discontinuous stalagmite record and the Mesolithic to Neolithic transition. Quaternary Research 66(3): 388–400.

67.

Frisia

Borsato

Spötl

et al . (2005) Climate variability in the SE Alps of Italy over the past 17 000 years reconstructed from a stalagmite record. Boreas 34(4): 445–455.

68.

García-Alix

Jiménez-Moreno

Jiménez-Espejo

et al . (2014) An environmental snapshot of the Bølling interstadial in Southern Iberia. Quaternary Research 81(2): 284–294.

69.

Gazol

Ribas

Gutiérrez

et al . (2017) Aleppo pine forests from across Spain show drought-induced growth decline and partial recovery. Agricultural and Forest Meteorology 232: 186–194.

70.

Genty

Blamart

Ghaleb

et al . (2006) Timing and dynamics of the last deglaciation from European and North African δ¹³C stalagmite profiles – Comparison with Chinese and South Hemisphere stalagmites. Quaternary Science Reviews 25(17–18): 2118–2142.

71.

Genty

Blamart

Ouahdi

et al . (2003) Precise dating of Dansgaard–Oeschger climate oscillations in western Europe from stalagmite data. Nature 421: 833–837.

72.

Genty

Combourieu Nebout

Hatté

et al . (2005) Rapid climatic changes of the last 90 kyr recorded on the European continent. Comptes Rendus Geoscience 337(10–11): 970–982.

73.

Gibert

Ribot

et al . (2008) CV-0, an early Pleistocene human phalanx from Cueva Victoria (Cartagena, Spain). Journal of Human Evolution 54(1): 150–156.

74.

Gibert

Ferràndez-Canadell

(eds) (2015) Geology and Paleontology of Cueva Victoria (Mastia 11–13). Cartagena: Ayuntamiento de Cartagena.

75.

Gibert

Scott

Scholz

et al . (2016) Chronology for the Cueva Victoria fossil site (SE Spain): Evidence for Early Pleistocene Afro-Iberian dispersals. Journal of Human Evolution 90: 183–197.

76.

Gimeno

Nieto

Trigo

et al . (2010) Where does the Iberian Peninsula moisture come from?: An answer based on a Lagrangian approach. Journal of Hydrometeorology 11(2): 421–436.

77.

González-Sampériz

Aranbarri

Pérez-Sanz

et al . (2017) Environmental and climate change in the southern Central Pyrenees since the Last Glacial Maximum: A view from the lake records. CATENA 149: 668–688.

78.

Grant

Rohling

Bar-Matthews

et al . (2012) Rapid coupling between ice volume and polar temperature over the past 150,000 years. Nature 491: 744–747.

79.

Günster

Eck

Skowronek

et al . (2001) Late Pleistocene loess and their paleosols in the Granada Basin, Southern Spain. Quaternary International 76-77(1): 241–245.

80.

Hansen

Scholz

Froeschmann

M-L

et al . (2017) Carbon isotope exchange between gaseous CO₂ and thin solution films: Artificial cave experiments and a complete diffusion-reaction model. Geochimica et Cosmochimica Acta 211: 28–47.

81.

Harris

Jones

Osborn

et al . (2014) Updated high-resolution grids of monthly climatic observations – The CRU TS3.10 Dataset. International Journal of Climatology 34(3): 623–642.

82.

Hellstrom

(2006) U–Th dating of speleothems with high initial 230Th using stratigraphical constraint. Quaternary Geochronology 1(4): 289–295.

83.

Hély

Lézine

A-M

, APD Contributors (2014) Holocene changes in African vegetation: Tradeoff between climate and water availability. Climate of the Past 10(2): 681–686.

84.

Henderson

(2006) Caving in to New Chronologies. Science 313(5787): 620–622.

85.

Hodge

Richards

Smart

et al . (2008) Effective precipitation in southern Spain (~266 to 46 ka) based on a speleothem stable carbon isotope record. Quaternary Research 69(3): 447–457.

86.

Hoffmann

Beck

Richards

et al . (2010) Towards radiocarbon calibration beyond 28ka using speleothems from the Bahamas. Earth and Planetary Science Letters 289(1–2): 1–10.

87.

Hurrell

Loon

(1997) Decadal Variations in climate associated with the North Atlantic Oscillation. Climatic Change 36(3): 301–326.

88.

Jaffey

Flynn

Glendenin

et al . (1971) Precision measurement of half-lives and specific activities of ²³⁵U and ²³⁸U. Physical Review C 4(5): 1889.

89.

Jalut

Esteban Amat

Bonnet

et al . (2000) Holocene climatic changes in the Western Mediterranean, from south-east France to south-east Spain. Palaeogeography, Palaeoclimatology, Palaeoecology 160(3–4): 255–290.

90.

Jimenez-Espejo

Martinez-Ruiz

Sakamoto

et al . (2007) Paleoenvironmental changes in the western Mediterranean since the last glacial maximum: High resolution multiproxy record from the Algero–Balearic basin. Palaeogeography, Palaeoclimatology, Palaeoecology 246(2–4): 292–306.

91.

Jin

Chen

Morrill

et al . (2012) Causes of early Holocene desertification in arid central Asia. Climate Dynamics 38(7): 1577–1591.

92.

Jochum

Nohl

Herwig

et al . (2005) GeoReM: A new geochemical database for reference materials and isotopic standards. Geostandards and Geoanalytical Research 29(3): 333–338.

93.

Jochum

Scholz

Stoll

et al . (2012) Accurate trace element analysis of speleothems and biogenic calcium carbonates by LA-ICP-MS. Chemical Geology 318–319: 31–44.

94.

Jochum

Stoll

Herwig

et al . (2007) Validation of LA-ICP-MS trace element analysis of geological glasses using a new solid-state 193 nm Nd:YAG laser and matrix-matched calibration. Journal of Analytical Atomic Spectrometry 22(2): 112–121.

95.

Jochum

Weis

Stoll

et al . (2011) Determination of reference values for NIST SRM 610-617 glasses following ISO guidelines. Geostandards and Geoanalytical Research 35(4): 397–429.

96.

Johnson

Belshaw

et al . (2006) Seasonal trace-element and stable-isotope variations in a Chinese speleothem: The potential for high-resolution paleomonsoon reconstruction. Earth and Planetary Science Letters 244(1–2): 394–407.

97.

Jones

Jonsson

Wheeler

(1997) Extension to the North Atlantic oscillation using early instrumental pressure observations from Gibraltar and south-west Iceland. International Journal of Climatology 17(13): 1433–1450.

98.

Jones

Burjachs

Ferrer-García

et al . (2018) A multi-proxy approach to understanding complex responses of salt-lake catchments to climate variability and human pressure: A Late Quaternary case study from south-eastern, Spain. Quaternary Science Reviews 184: 201–223.

99.

Kaiser

Specker

(1956) Bewertung und Vergleich von Analysenverfahren. Fresenius’ Zeitschrift für Analytische Chemie 149(1–2): 46–66.

100.

Kim

S-T

O’Neil

(1997) Equilibrium and nonequilibrium oxygen isotope effects in synthetic carbonates. Geochimica et Cosmochimica Acta 61(16): 3461–3475.

101.

Klein Tank

AMG

Wijngaard

Können

et al . (2002) Daily dataset of 20th-century surface air temperature and precipitation series for the European Climate Assessment. International Journal of Climatology 22(12): 1441–1453.

102.

Kottek

Grieser

Beck

et al . (2006) World Map of the Köppen-Geiger climate classification updated. Meteorologische Zeitschrift 15(3): 259–263.

103.

Krklec

Domínguez-Villar

(2014) Quantification of the impact of moisture source regions on the oxygen isotope composition of precipitation over Eagle Cave, central Spain. Geochimica et Cosmochimica Acta 134: 39–54.

104.

Kutzbach

Chen

Cheng

et al . (2014) Potential role of winter rainfall in explaining increased moisture in the Mediterranean and Middle East during periods of maximum orbitally-forced insolation seasonality. Climate Dynamics 42(3): 1079–1095.

105.

Lachniet

(2009) Climatic and environmental controls on speleothem oxygen-isotope values. Quaternary Science Reviews 28(5–6): 412–432.

106.

Lambeck

Rouby

Purcell

et al . (2014) Sea level and global ice volumes from the Last Glacial Maximum to the Holocene. Proceedings of the National Academy of Sciences of the United States of America 111(43): 15296–15303.

107.

Le Roux

Glendenin

(1963) Half-life of ²³²Th. In: South African Atomic Energy Board (ed.) National Conference on Nuclear Energy, Application of Isotopes and Radiation: Proceedings of the National Conference on Nuclear Energy held in Pretoria, April 5-8 1963. Pretoria: Pelindaba, pp. 83–94.

108.

LeGrande

Schmidt

(2006) Global gridded data set of the oxygen isotopic composition in seawater. Geophysical Research Letters 33(12): L12604.

109.

Linares

Delgado-Huertas

Carreira

(2011) Climatic trends and different drought adaptive capacity and vulnerability in a mixed Abies pinsapo–Pinus halepensis forest. Climatic Change 105(1): 67–90.

110.

Longerich

Jackson

Günther

(1996) Inter-laboratory note. Laser ablation inductively coupled plasma mass spectrometric transient signal data acquisition and analyte concentration calculation. Journal of Analytical Atomic Spectrometry 11(9): 899–904.

111.

McDermott

(2004) Palaeo-climate reconstruction from stable isotope variations in speleothems: A review. Quaternary Science Reviews 23(7–8): 901–918.

112.

McDermott

Atkinson

Fairchild

et al . (2011) A first evaluation of the spatial gradients in δ¹⁸O recorded by European Holocene speleothems. Global and Planetary Change 79(3–4): 275–287.

113.

McMillan

(2006) Tests for Palaeoaridity in Holocene Stalagmites from SW Europe. PhD Thesis, University of Keele.

114.

Magny

Joannin

Galop

et al . (2012) Holocene palaeohydrological changes in the northern Mediterranean borderlands as reflected by the lake-level record of lake ledro, northeastern Italy. Quaternary Research 77(3): 382–396.

115.

Magny

Vannière

Calo

et al . (2011) Holocene hydrological changes in south-western Mediterranean as recorded by lake-level fluctuations at Lago Preola, a coastal lake in southern Sicily, Italy. Quaternary Science Reviews 30(19–20): 2459–2475.

116.

Manteca Martínez

Pina

(2015) Las mineralizaciones ferro-manganesíferas de la mina-cueva Victoria y su contexto geológico: Fe-Mn mineralizations of the mine-cave Victoria and their geological context. In: Gibert

Ferràndez-Canadell

(eds) Geología y Paleontología de Cueva Victoria [Geology and Paleontology of Cueva Victoria] (Mastia 11–13). Cartagena: Ayuntamiento de Cartagena, pp. 59–74.

117.

Marcott

Shakun

Clark

et al . (2013) A reconstruction of regional and global temperature for the past 11,300 years. Science 339(6124): 1198–1201.

118.

Martin-Vide

Lopez-Bustins

J-A

(2006) The Western Mediterranean Oscillation and rainfall in the Iberian Peninsula. International Journal of Climatology 26(11): 1455–1475.

119.

Martrat

Jimenez-Amat

Zahn

et al . (2014) Similarities and dissimilarities between the last two deglaciations and interglaciations in the North Atlantic region. Quaternary Science Reviews 99: 122–134.

120.

Mattey

Fairchild

Atkinson

et al . (2010) Seasonal microclimate control of calcite fabrics, stable isotopes and trace elements in modern speleothem from St Michaels Cave, Gibraltar. In: Pedley

Rogerson

(eds) Tufas and Speleothems: Unravelling the Microbial and Physical Controls. London: Geological Society of London (Special Publications 336), pp. 323–344.

121.

Mauri

Davis

BAS

Collins

et al . (2015) The climate of Europe during the Holocene: A gridded pollen-based reconstruction and its multi-proxy evaluation. Quaternary Science Reviews 112: 109–127.

122.

Mauz

Fanelli

Elmejdoub

et al . (2012) Coastal response to climate change: Mediterranean shorelines during the Last Interglacial (MIS 5). Quaternary Science Reviews 54: 89–98.

123.

Mertz-Kraus

Brachert

Jochum

et al . (2009) LA-ICP-MS analyses on coral growth increments reveal heavy winter rain in the Eastern Mediterranean at 9 Ma. Palaeogeography, Palaeoclimatology, Palaeoecology 273(1–2): 25–40.

124.

Meyer

Feng

Breecker

et al . (2014) Interpretation of speleothem calcite δ¹³C variations: Evidence from monitoring soil CO2, drip water, and modern speleothem calcite in central Texas. Geochimica et Cosmochimica Acta 142: 281–298.

125.

Mischel

Mertz-Kraus

Jochum

et al . (2017a) TERMITE: An R script for fast reduction of laser ablation inductively coupled plasma mass spectrometry data and its application to trace element measurements. Rapid Communications in Mass Spectrometry 31(13): 1079–1087.

126.

Mischel

Scholz

Spötl

et al . (2017b) Holocene climate variability in Central Germany and a potential link to the polar North Atlantic: A replicated record from three coeval speleothems. The Holocene 27(4): 509–525.

127.

Morellón

Aranbarri

Moreno

et al . (2018) Early Holocene humidity patterns in the Iberian Peninsula reconstructed from lake, pollen and speleothem records. Quaternary Science Reviews 181: 1–18.

128.

Morellón

Valero-Garcés

Vegas-Vilarrúbia

et al . (2009) Lateglacial and Holocene palaeohydrology in the western Mediterranean region: The Lake Estanya record (NE Spain). Quaternary Science Reviews 28(25-26): 2582–2599.

129.

Moreno

Pérez-Mejías

Bartolomé

et al . (2017) New speleothem data from Molinos and Ejulve caves reveal Holocene hydrological variability in northeast Iberia. Quaternary Research 88(2): 223–233.

130.

Moreno

Sancho

Bartolomé

et al . (2014a) Climate controls on rainfall isotopes and their effects on cave drip water and speleothem growth: The case of Molinos cave (Teruel, NE Spain). Climate Dynamics 43(1–2): 221–241.

131.

Moreno

Stoll

Jiménez-Sánchez

et al . (2010) A speleothem record of glacial (25–11.6kyr BP) rapid climatic changes from northern Iberian Peninsula. Global and Planetary Change 71(3–4): 218–231.

132.

Moreno

Svensson

Brooks

et al . (2014b) A compilation of Western European terrestrial records 60–8 ka BP: Towards an understanding of latitudinal climatic gradients. Quaternary Science Reviews 106: 167–185.

133.

Mühlinghaus

Scholz

Mangini

(2009) Modelling fractionation of stable isotopes in stalagmites. Geochimica et Cosmochimica Acta 73(24): 7275–7289.

134.

North Greenland Ice Core Project Members (2004) High-resolution record of Northern Hemisphere climate extending into the last interglacial period. Nature 431: 147–151.

135.

Obert

Scholz

Felis

et al . (2016) ²³⁰Th/U dating of Last Interglacial brain corals from Bonaire (southern Caribbean) using bulk and theca wall material. Geochimica et Cosmochimica Acta 178: 20–40.

136.

Pantaléon-Cano

Yll

E-I

Pérez-Obiol

et al . (2003) Palynological evidence for vegetational history in semi-arid areas of the western Mediterranean (Almería, Spain). The Holocene 13(1): 109–119.

137.

Pasho

Camarero

Luis

et al . (2011) Impacts of drought at different time scales on forest growth across a wide climatic gradient in north-eastern Spain. Agricultural and Forest Meteorology 151(12): 1800–1811.

138.

Pausata

FSR

Messori

Zhang

(2016) Impacts of dust reduction on the northward expansion of the African monsoon during the Green Sahara period. Earth and Planetary Science Letters 434: 298–307.

139.

Pellicer

Corella

Gutiérrez

et al . (2016) Sedimentological and palaeohydrological characterization of Late Pleistocene and Holocene tufa mound palaeolakes using trenching methods in the Spanish Pyrenees. Sedimentology 63(6): 1786–1819.

140.

Pérez-Sanz

González-Sampériz

Moreno

et al . (2013) Holocene climate variability, vegetation dynamics and fire regime in the central Pyrenees: The Basa de la Mora sequence (NE Spain). Quaternary Science Reviews 73: 149–169.

141.

Peyron

Goring

Dormoy

et al . (2011) Holocene seasonality changes in the central Mediterranean region reconstructed from the pollen sequences of Lake Accesa (Italy) and Tenaghi Philippon (Greece). The Holocene 21(1): 131–146.

142.

Peyron

Magny

Goring

et al . (2013) Contrasting patterns of climatic changes during the Holocene across the Italian Peninsula reconstructed from pollen data. Climate of the Past 9(3): 1233–1252.

143.

Piqué

Revelles

Burjachs

et al . (2018) Interdisciplinary approach to the landscape and firewood exploitation during the Holocene at La Garrotxa (Girona, NE Iberia). Quaternary International 463: 401–413.

144.

Rasmussen

Andersen

Svensson

et al . (2006) A new Greenland ice core chronology for the last glacial termination. Journal of Geophysical Research: Atmospheres 111: D06102.

145.

Ribot

Ferràndez-Canadell

Gibert

(2015) Primates from Cueva Victoria. In: Gibert

Ferràndez-Canadell

(eds) Geología y Paleontología de Cueva Victoria [Geology and Paleontology of Cueva Victoria] (Mastia 11–13). Cartagena: Ayuntamiento de Cartagena, pp. 433–452.

146.

Richards

Dorale

(2003) Uranium-series chronology and environmental applications of speleothems. Reviews in Mineralogy and Geochemistry 52(1): 407–460.

147.

Ridley

Asmerom

Baldini

JUL

et al . (2015) Aerosol forcing of the position of the intertropical convergence zone since ad 1550. Nature Geoscience 8(3): 195–200.

148.

Ríos-Cornejo

Penas

Álvarez-Esteban

et al . (2015) Links between teleconnection patterns and mean temperature in Spain: Theoretical and Applied Climatology. Theoretical and Applied Climatology 122(1–2): 1–18.

149.

Rivera-Collazo

Winter

Scholz

et al . (2015) Human adaptation strategies to abrupt climate change in Puerto Rico ca. 3.5 ka. The Holocene 25(4): 627–640.

150.

Rodriguez-Puebla

Encinas

Nieto

et al . (1998) Spatial and temporal patterns of annual precipitation variability over the Iberian Peninsula. International Journal of Climatology 18(3): 299–316.

151.

Rohling

Marino

Grant

(2015) Mediterranean climate and oceanography, and the periodic development of anoxic events (sapropels). Earth-Science Reviews 143: 62–97.

152.

Ros

Llamusí

(2015) Reconstruction and genesis of the Cueva Victoria karst. In: Gibert

Ferràndez-Canadell

(eds) Geología y Paleontología de Cueva Victoria [Geology and Paleontology of Cueva Victoria] (Mastia 11–13). Cartagena: Ayuntamiento de Cartagena, pp. 111–125.

153.

Roy-Barman

Pons-Branchu

(2016) Improved U–Th dating of carbonates with high initial ²³⁰Th using stratigraphical and coevality constraints. Quaternary Geochronology 32: 29–39.

154.

Rozanski

Araguás-Araguás

Gonfiantini

(1992) Relation between long-term trends of oxygen-18 isotope composition of precipitation and climate. Science 258(5084): 981–985.

155.

Rozanski

Araguás-Araguás

Gonfiantini

(1993) Isotopic patterns in modern global precipitation. In: Swart

Lohmann

McKenzie

et al . (eds) Climate Change in Continental Isotopic Records. Washington, DC: American Geophysical Union, pp. 1–36.

156.

Rudzka

McDermott

Baldini

et al . (2011) The coupled δ¹³C-radiocarbon systematics of three Late Glacial/early Holocene speleothems; insights into soil and cave processes at climatic transitions. Geochimica et Cosmochimica Acta 75(15): 4321–4339.

157.

Sadori

Narcisi

(2001) The Postglacial record of environmental history from Lago di Pergusa, Sicily. The Holocene 11(6): 655–671.

158.

Schneider

Bischoff

Haug

(2014) Migrations and dynamics of the intertropical convergence zone. Nature 513: 45.

159.

Scholz

Hoffmann

(2008) 230Th/U-dating fossil corals and speleothems. Eiszeitalter und Gegenwart 57(1–2): 52–76.

160.

Scholz

Hoffmann

(2011) StalAge – An algorithm designed for construction of speleothem age models. Quaternary Geochronology 6(3–4): 369–382.

161.

Scholz

Frisia

Borsato

et al . (2012) Holocene climate variability in north-eastern Italy: Potential influence of the NAO and solar activity recorded by speleothem data. Climate of the Past 8(4): 1367–1383.

162.

Scholz

Mühlinghaus

Mangini

(2009) Modelling δ¹³C and δ¹⁸O in the solution layer on stalagmite surfaces. Geochimica et Cosmochimica Acta 73(9): 2592–2602.

163.

Scholz

Tolzmann

Hoffmann

et al . (2014) Diagenesis of speleothems and its effect on the accuracy of 230Th/U-ages. Chemical Geology 387: 74–86.

164.

Siddall

Rohling

Almogi-Labin

et al . (2003) Sea-level fluctuations during the last glacial cycle. Nature 423: 853–858.

165.

Sinclair

Banner

Taylor

et al . (2012) Magnesium and strontium systematics in tropical speleothems from the Western Pacific. Chemical Geology 294–295: 1–17.

166.

Sodemann

Stohl

(2009) Asymmetries in the moisture origin of Antarctic precipitation. Geophysical Research Letters 36(22): L22803.

167.

Sodemann

Schwierz

Wernli

(2008) Interannual variability of Greenland winter precipitation sources: Lagrangian moisture diagnostic and North Atlantic Oscillation influence. Journal of Geophysical Research: Atmospheres 113(D3): D03107.

168.

Spötl

(2011) Long-term performance of the Gasbench isotope ratio mass spectrometry system for the stable isotope analysis of carbonate microsamples. Rapid Communications in Mass Spectrometry 25(11): 1683–1685.

169.

Spötl

Vennemann

(2003) Continuous-flow isotope ratio mass spectrometric analysis of carbonate minerals. Rapid Communications in Mass Spectrometry 17(9): 1004–1006.

170.

Spötl

Fairchild

Tooth

(2005) Cave air control on dripwater geochemistry, Obir Caves (Austria): Implications for speleothem deposition in dynamically ventilated caves. Geochimica et Cosmochimica Acta 69(10): 2451–2468.

171.

Stein

Draxler

Rolph

et al . (2015) NOAA’s HYSPLIT atmospheric transport and dispersion modeling system. Bulletin of the American Meteorological Society 96(12): 2059–2077.

172.

Stoll

Mendez-Vicente

Gonzalez-Lemos

et al . (2015) Interpretation of orbital scale variability in mid-latitude speleothem δ¹⁸O: Significance of growth rate controlled kinetic fractionation effects. Quaternary Science Reviews 127: 215–228.

173.

Stoll

Moreno

Mendez-Vicente

et al . (2013) Paleoclimate and growth rates of speleothems in the northwestern Iberian Peninsula over the last two glacial cycles. Quaternary Research 80(2): 284–290.

174.

Stoll

Müller

Prieto

(2012) I-STAL, a model for interpretation of Mg/Ca, Sr/Ca and Ba/Ca variations in speleothems and its forward and inverse application on seasonal to millennial scales. Geochemistry, Geophysics, Geosystems 13(9): Q09004.

175.

Tierney

Pausata

FSR

deMenocal

(2017) Rainfall regimes of the Green Sahara. Science Advances 3(1): e1601503.

176.

Tinner

van Leeuwen

JFN

Colombaroli

et al . (2009) Holocene environmental and climatic changes at Gorgo Basso, a coastal lake in southern Sicily, Italy. Quaternary Science Reviews 28(15–16): 1498–1510.

177.

Tjallingii

Claussen

Stuut

J-BW

et al . (2008) Coherent high- and low-latitude control of the northwest African hydrological balance. Nature Geoscience 1(10): 670–675.

178.

Treble

Fairchild

Griffiths

et al . (2015) Impacts of cave air ventilation and in-cave prior calcite precipitation on Golgotha Cave dripwater chemistry, southwest Australia: Novel approaches to and new insights from speleothem-based climate reconstructions. Quaternary Science Reviews 127: 61–72.

179.

Tremaine

Froelich

Wang

(2011) Speleothem calcite farmed in situ: Modern calibration of δ¹⁸O and δ¹³C paleoclimate proxies in a continuously-monitored natural cave system. Geochimica et Cosmochimica Acta 75(17): 4929–4950.

180.

Tzedakis

(2005) Towards an understanding of the response of southern European vegetation to orbital and suborbital climate variability. Quaternary Science Reviews 24(14–15): 1585–1599.

181.

Tzedakis

Hooghiemstra

Pälike

(2006) The last 1.35 million years at Tenaghi Philippon: Revised chronostratigraphy and long-term vegetation trends. Quaternary Science Reviews 25(23–24): 3416–3430.

182.

Tzedakis

Roucoux

Abreu

et al . (2004) The duration of forest stages in Southern Europe and interglacial climate variability. Science 306: 2231–2235.

183.

Waelbroeck

Labeyrie

Michel

et al . (2002) Sea-level and deep water temperature changes derived from benthic foraminifera isotopic records. Quaternary Science Reviews 21: 295–305.

184.

Walczak

Baldini

JUL

Baldini

et al . (2015) Reconstructing high-resolution climate using CT scanning of unsectioned stalagmites: A case study identifying the mid-Holocene onset of the Mediterranean climate in southern Iberia. Quaternary Science Reviews 127: 117–128.

185.

Wassenburg

Dietrich

Fietzke

et al . (2016) Reorganization of the North Atlantic Oscillation during early Holocene deglaciation. Nature Geoscience 9(8): 602–605.

186.

Wedepohl

(1995) The composition of the continental crust. Geochimica et Cosmochimica Acta 59(7): 1217–1232.

187.

Yang

Scholz

Jochum

et al . (2015) Lead isotope variability in speleothems – A promising new proxy for hydrological change? First results from a stalagmite from western Germany. Chemical Geology 396: 143–151.

188.

Zanchetta

Drysdale

Hellstrom

et al . (2007) Enhanced rainfall in the Western Mediterranean during deposition of sapropel S1: Stalagmite evidence from Corchia cave (Central Italy). Quaternary Science Reviews 26(3–4): 279–286.

189.

Zazo

Goy

Dabrio

et al . (2013) Retracing the Quaternary history of sea-level changes in the Spanish Mediterranean–Atlantic coasts: Geomorphological and sedimentological approach. Geomorphology 196: 36–49.

190.

Zielhofer

Fletcher

Mischke

et al . (2017) Atlantic forcing of Western Mediterranean winter rain minima during the last 12,000 years. Quaternary Science Reviews 157: 29–51.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

5.80 MB

Speleothem δ 13 C record suggests enhanced spring/summer drought in south-eastern Spain between 9.7 and 7.8 ka – A circum-Western Mediterranean anomaly?

Abstract

Keywords

Introduction

Regional setting

Material and methods

Sample collection and preparation

230Th/U dating

Trace element measurements

Stable isotope measurements

Moisture source modelling

Results

Petrography

230Th/U dating

Proxy data

Trajectory analysis

Discussion

Chronology

Speleothem δ13C values and trace elements

Speleothem δ18O values

Implications for Western Mediterranean climate between 15 and 7 ka

15–11.7 ka

11.7–9.7 ka

9.7–7 ka

The implications for large-scale atmospheric circulation

Conclusion

Supplemental Material

Supplement_Budsky – Supplemental material for Speleothem δ13C record suggests enhanced spring/summer drought in south-eastern Spain between 9.7 and 7.8 ka – A circum-Western Mediterranean anomaly?

Footnotes

Acknowledgements

Funding

Supplemental material

ORCID iD

References

Supplementary Material

²³⁰Th/U dating

²³⁰Th/U dating

Speleothem δ¹³C values and trace elements

Speleothem δ¹⁸O values