Dominique Genty

Precise dating of Dansgaard-Oeschger climate oscillations in western Europe from stalagmite data

The signature of Dansgaard–Oeschger events—millennial-scale abrupt climate oscillations during the last glacial period—is well established in ice cores and marine records. But the effects of such events in continental settings are not as clear, and their absolute chronology is uncertain beyond the limit of 14C dating and annual layer counting for marine records and ice cores, respectively. Here we present carbon and oxygen isotope records from a stalagmite collected in southwest France which have been precisely dated using 234U/230Th ratios. We find rapid climate oscillations coincident with the established Dansgaard–Oeschger events between 83,000 and 32,000 years ago in both isotope records. The oxygen isotope signature is similar to a record from Soreq cave, Israel, and deep-sea records, indicating the large spatial scale of the climate oscillations. The signal in the carbon isotopes gives evidence of drastic and rapid vegetation changes in western Europe, an important site in human cultural evolution. We also find evidence for a long phase of extremely cold climate in southwest France between 61.2 ± 0.6 and 67.4 ± 0.9 kyr ago.

Testing Theoretically Predicted Stalagmite Growth Rate with Recent Annually Laminated Samples: Implications for Past Stalagmite Deposition

Annually laminated stalagmites deposited over the last 30–160 years are analysed to determine their growth rate. Three natural and artificial cave sites in England, France, and Belgium were chosen for their wide range of variability in growth rate determining variables, and multiple samples were taken from each site. The annual nature of laminae deposition within the stalagmite calcite was confirmed by comparison to the date of cave/void opening, 14C analyses, or by using dated event horizons. Measured stalagmite growth rate was determined from annual laminae thickness measurements and compared to that theoretically predicted from the chemical kinetics of the calcite precipitation reaction. A good agreement is observed between empirical observations and theoretical predictions, although two complicating factors, variations in calcite porosity, and seasonal cessation of the water supply to the samples, both affect the growth rate. Implications for the extraction of palaeoclimate information from stalagmite growth rate are discussed.

Annually Laminated Sequences in the Internal Structure of Some Belgian Stalagmites--Importance for Paleoclimatology

ABSTRACT Fifteen stalagmites from four caves and one sealed tunnel in southern Belgium are composed of alternations of annually deposited white-porous and dark-compact laminae. This is demonstrated by comparing the number of laminae with the local history of the site for modern stalagmites and with radioisotopic ages for Late Glacial and Holocene stalagmites. Annual cyclicity in the internal structure of these speleothems is explained by the highly seasonal variations of the water excess, which influences underground water flow. Comparison between climatic data and modern stalagmites of a closed tunnel shows that growth laminae can record climatic variations: (1) there is a good correlation (R = 0.84) between lamina thickness in a stalagmite and water excess; (2) during years with a high water excess, dark-compact laminae are more developed, which makes the speleothem darker. Vertical successions of several laminae represent microsequences that may have recorded climatic variations with a time resolution of 1/2 year. In a Late Glacial stalagmite, successive laminae microsequences form very regular cycles of 11 years separated by a thick darkcompact lamina. It is supposed that, as for modern stalagmites, the thick dark-compact lamina corresponds to a period of high water excess. Hence, this 11-year cycle may reflect a climatic cycle.

Intra- and inter-annual growth rate of modern stalagmites

We measure the factors that determine growth rate (temperature, drip rate, calcium ion concentration) for 31 waters that feed stalagmites within six cave systems throughout Europe. Water samples were collected at a frequency of at least 1 month, to permit the modelling of both inter- and intra-annual growth rate variations, utilising the theory of Wolfgang Dreybrodt (Chem. Geol. 29 (1980) 89–105; Chem. Geol. 32 (1981) 237–245; Dreybrodt, W., 1988. Processes in Karst Systems. Springer-Verlag, Berlin 288 pp.). Inter-annual growth rates were measured using the stalagmites that were associated with the analysed water samples; growth rate was determined from annual lamina counting, specific time markers within the stalagmites, and location of bomb 14C. When compared to theoretically predicted values, a good agreement between theoretical and measured stalagmite growth rates is observed (R2=0.69). When compared to site climate and geochemical parameters, a good correlation is observed between measured growth rate and mean annual temperature for five sites (R2=0.63) and dripwater calcium content (R2=0.61), but not drip rate (R2=0.09). The good correlation with both calcium and temperature is due to soil CO2 production being primarily determined by surface temperature and soil moisture. However, when we compare our data to that in the Grotte de Clamouse, a site that has little soil cover, we observe that the growth rate–temperature relationship breaks down due to either the lack of soil CO2 production or prior calcite precipitation. Intra-annual data demonstrates that maximum growth rate occurs when calcium concentrations are high, and that this occurs under different seasons depending on the hydrology of each site. Our results demonstrate a stronger dependence of intra-annual stalagmite growth rate on dissolved calcium ion concentrations than drip rate for the range of drip rates investigated here (0.01<t<2drip s−1), but for lower drip rates, this factor becomes important in controlling growth rate. We suggest that for well-monitored and -understood sites, stalagmite growth rate variations can provide useful information for palaeoclimate reconstruction.

Dead carbon in stalagmites: carbonate bedrock paleodissolution vs. ageing of soil organic matter. Implications for 13C variations in speleothems

Twenty-one 14C accelerator mass spectrometric analyses were obtained for three Holocene stalagmites from the Uamh-an-Tartair cave (Sutherland, Scotland) in order to estimate the past dead carbon proportion (dcp). Results show that the dcp increases from 22 to 38% from 3780 years ago to the present. Because δ13C variation is small within each sample, it is concluded that this dcp increase is the product of the ageing of soil organic matter related to peat bog development above the cave that produced older soil CO2 and not from a more intense dissolution of the surrounding carbonates, which would have led to a δ13C increase. Comparison with samples from other sites in Europe shows no intersite correlation between dcp and δ13C, but a relatively good intersite correlation is observed between dcp and average site temperature. Thus, temperature may be a major factor controlling the production of old soil organic matter CO2 and, therefore, the dead carbon content of seepage water. In contrast to the Scotland stalagmites, two other Holocene samples from sites in southern France and Belgium exhibit a good correlation between δ13C and dcp, which can be explained in terms of variations in the intensity of limestone dissolution. Consequently, δ13C variations observed in stalagmites are not always due to changes in the vegetation type (C3/C4) as has been commonly assumed; 13C/12C variations in speleothem calcite may also be controlled by the soil organic matter age and, in some cases, by the intensity of the limestone dissolution. Conversely, a largely constant speleothem δ13C signal, as observed for the Scotland stalagmites, does not necessarily imply that surface climate and vegetation conditions were stable since the dcp variations, in this case, are clearly related to the peat bog development during stalagmite growth.

Carbon transfer dynamics from bomb-14C and δ13C time series of a laminated stalagmite from SW France—modelling and comparison with other stalagmite records

Twenty-two AMS 14C measurements have been made on a modern stalagmite from SW France in order to reconstruct the 14C activity history of the calcite deposit. Annual growth laminae provides a chronology up to 1919 A.D. Results show that the stalagmite 14C activity time series is sensitive to modern atmosphere 14C activity changes such as those produced by the nuclear weapon tests. The comparison between the two 14C time series shows that the stalagmite time series is damped: its amplitude variation between pre-bomb and post-bomb values is 75% less and the time delay between the two time series peaks is 16 years ±3. A model is developed using atmosphere 14C and 13C data, fractionation processes and three soil organic matter components whose mean turnover rates are different. The linear correlation coefficient between modeled and measured activities is 0.99. These results, combined with two other stalagmite 14C time series already published and compared with local vegetation and climate, demonstrate that most of the carbon transfer dynamics are controlled in the soil by soil organic matter degradation rates. Where vegetation produces debris whose degradation is slow, the fraction of old carbon injected in the system increases, the observed 14C time series is much more damped and lag time longer than that observed under grassland sites. The same mixing model applied on the 13C shows a good agreement (R2 = 0.78) between modeled and measured stalagmite δ13C and demonstrates that the Suess Effect due to fossil fuel combustion in the atmosphere is recorded in the stalagmite but with a damped effect due to SOM degradation rate. The different sources of dead carbon in the seepage water are calculated and discussed.

Annually Laminated Speleothems: a Review

This review of annually laminated speleothems firstly considers the four types of annual laminae found within speleothems: fluorescent laminae formed by annual variations in organic matter flux; visible or petrographic laminae, formed by annual variations in calcite texture or fabric; calcite-aragonite couplets; and finally trace element laminae. The methods available to confirm the annual nature, or otherwise, of lamina deposition are reviewed. We consider the use of annual laminae in chronology building, with particular relevance to palaeoclimate reconstructions. Finally, the use of annual lamina width as a palaeoclimate proxy is reviewed.

Bomb 14C recorded in laminated speleothems : Calculation of dead carbon proportion

We performed radiocarbon measurements using accelerator mass spectrometry (AMS) on 6 stalagmites, 3 stalactites and 7 seepage waters from four different caves in Southwest France and Belgium in order to calculate the dead carbon proportion (dcp). All the speleothems studied are modern and annually laminated, which offers the advantage of an accurate chronology, with better than one-year resolution. Coupled with the fact that very little calcite is necessary for an AMS measurement (between 1.5 and 7 yr of calcite deposit), we obtained dead carbon values within an uncertainty limit of ± 1.5%. Results show that the dead carbon proportion varies from 9.2% to 21.9% for calcite deposits and from 3.6% to 21.9% for water. In each sampling site, the dcp is homogeneous. Although the inter-site dcp varies by >11%, its average value of 15.5% ± 4.4 still lies within the uncertainty range of the accepted value of 15% ± 5 (dilution factor of 0.85 ± 03). We compare the average dcp of each site with the local geology, vegetation and climate. Given similar geology and temperature, the highest dcp values are found under forest cover; dcp difference is up to 9%. However, the Belgian site, which is also under a forest, shows a dcp very close to the dcp found under grassland sites of Southwest France, which proves that other unknown factors may play an important role in dissolution processes. Secondary calcite deposition and redissolution in the soil zone or more likely in the fracture system before reaching the cave itself could also explain the inter-site differences. The IAEA isotopic model (Pearson model adapted for open systems) is in good agreement with the measured activities.

Calculation of Past Dead Carbon Proportion and Variability by the Comparison of AMS 14C and Tims U/TH Ages on Two Holocene Stalagmites

Twenty-two radiocarbon activity measurements were made by accelerator mass spectrometry (AMS) on 2 Holocene stalagmites from Belgium (Han-stm lb) and from southwest France (Vil-stm lb). Sixteen thermal ionization mass spectrometric (TIMS) U/Th measurements were performed parallel to AMS analyses. The past dead carbon proportion (dcp) due to limestone dissolution and old soil organic matter (SOM) degradation is calculated with U/Th ages, measured calcite (super 14) C activity and atmospheric (super 14) C activity from the dendrochronological calibration curves. Results show that the dcp is different for the 2 stalagmites: between 10,800 and 4780 yr from present dcp = 17.5% (sigma = 2.4; n = 10) for Han-stm lb and dcp = 9.4% (sigma = 1.6; n = 6) between 3070 and 520 yr for Vil-stmlb. Despite a broad stability of the dcp during the time ranges covered by each sample, a slight dcp increase of about 5.0% is observed in the Han-stmlb sample between 8500 and 5200 yr. This change is synchronous with a calcite delta (super 13) C increase, which could be due to variation in limestone dissolution processes possibly linked with a vegetation change. The dcp and delta (super 13) C of the 2 studied samples are compared with 5 other modern stalagmites from Europe. Results show that several factors intervene, among them: the vegetation type, and the soil saturation leading to variable dissolution process systems (open/closed). The good correlation (R (super 2) = 0.98) between the U/Th ages and the calibrated (super 14) C ages corrected with a constant dcp validates the (super 14) C method. However, the dcp error leads to large (super 14) C age errors (i.e. 250-500 yr for the period studied), which is an obstacle for both a high-resolution chronology and the improvement of the (super 14) C calibration curves, at least for the Holocene.

Palaeoclimate research in Villars Cave (Dordogne, SW-France)

C to any vegetation and climatic changes. Drip rate monitoring, carried out under four stalactites from the lower and upper galleries, has shown a well marked seasonality of the seepage water with high flow rates during winter and spring. A time delay of about two months is observed between the water excess (estimated from outside meteorological stations) and the drip rate in the cave. A great heterogeneity in the flow rate amplitude variations and in the annual quantity of water between two nearby stalactites is observed, confirming the complexity of the micro-fissure network system in the unsaturated zone. At a daily scale, the air pressure and drip rates are anti-correlated probably because of pressure stress on the fissure network. Cave air CO 2 concentration follows soil CO 2 production and is correlated with its δ 13 C content. Since the beginning of the monitoring, the cave air temperature, in both lower and upper galleries, displays a warming trend of ~+0.4°C±0.1/10yrs. This might be the consequence of the outside temperature increase that reaches the Villars Cave galleries through thermal wave conduction. Chemistry monitoring over a few years has shown that the seepage water of the lower gallery stations is significantly more concentrated in trace and minor elements (i.e. Sr, Mg, Ba, U) than the upper stations, probably due to the 10-20 m depth difference between these galleries, which implies a different seepage pathway and different water/rock interaction durations. There is also, in the elemental concentration (i.e. [Ca]), a seasonal signal which causes variation in the speleothem growth rates. Modern calcite deposit experiments conducted for several years have permitted the calculation of vertical growth rates, which are extremely high in Villars (i.e. 1.0 to 1.75 mm/ yr). Pollen filter experiments in the cave have demonstrated that most of the pollen grain found in the cave comes from the air and not from the water. The specificity of the Villars Cave records is that the climatic variations were well recorded in the calcite δ 13 C whereas the δ 18 O is usually used in such studies. Overall, these results are helpful for the interpretation of speleothem records for palaeoclimatic reconstructions, but more work is needed, especially numerical modelling of the temperature, chemistry and hydrology.