Dana F.C. Riechelmann

Bunker Cave stalagmites: an archive for central European Holocene climate variability

Holocene climate was characterised by variability on multi-centennial to multi-decadal time scales. In central Europe, these fluctuations were most pronounced during winter. Here we present a record of past winter climate variability for the last 10.8 ka based on four speleothems from Bunker Cave, western Germany. Due to its central European location, the cave site is particularly well suited to record changes in precipitation and temperature in response to changes in the North Atlantic realm. We present highresolution records of δ18O, δ13C values and Mg/Ca ratios. Changes in the Mg/Ca ratio are attributed to past meteoric precipitation variability. The stable C isotope composition of the speleothems most likely reflects changes in vegetation and precipitation, and variations in the δ18O signal are interpreted as variations in meteoric precipitation and temperature. We found cold and dry periods between 8 and 7 ka, 6.5 and 5.5 ka, 4 and 3 ka as well as between 0.7 and 0.2 ka. The proxy signals in the Bunker Cave stalagmites compare well with other isotope records and, thus, seem representative for central European Holocene climate variability. The prominent 8.2 ka event and the Little Ice Age cold events are both recorded in the Bunker Cave record. However, these events show a contrasting relationship between climate and δ18O, which is explained by different causes underlying the two climate anomalies. Whereas the Little Ice Age is attributed to a pronounced negative phase of the North Atlantic Oscillation, the 8.2 ka event was triggered by cooler conditions in the North Atlantic due to a slowdown of the thermohaline circulation.

Sensitivity of whole wood stable carbon and oxygen isotope values to milling procedures

RATIONALE Milling of wood samples is a widely applied preparation method for pooling tree-rings from different trees or periods of several years for determination of δ(13)C and δ(18)O values. In this study, whole wood samples were milled using different procedures in order to evaluate potential effects of this preparation method on δ(13)C and δ(18)O values. METHODS Subsamples of a 5 cm(3) wood piece of a single tree-ring from a lowland white fir were used. The samples were milled with different setups: (i) two and three stainless-steel balls, (ii) 3, 5 and 8 min milling time, and (iii) discontinuous and continuous milling. The δ(13)C values were measured using an elemental analyser connected to an IsoPrime mass spectrometer and δ(18)O values using a Thermo Scientific MAT 253 mass spectrometer and a TC/EA connected by a ConFlo IV. RESULTS The results show that varying the milling procedure does not alter the δ(13)C and δ(18)O values in comparison to non-milled blank samples. For shorter milling times, an increased variance of δ(18)O values is recorded, probably caused by isotopic gradient between early- and latewood portions of the tree-ring and thereby biasing the insufficiently homogenised samples. No overheating effects on the δ(13)C and δ(18)O values were detected. CONCLUSIONS Milling of wood samples for carbon and oxygen isotope analyses is an appropriate preparation method.

Site-specific climatic signals in stable isotope records from Swedish pine forests

Key messagePinus sylvestris tree-ring δ13C and δ18O records from locally moist sites in central and northern Sweden contain consistently stronger climate signals than their dry site counterparts.AbstractWe produced twentieth century stable isotope data from Pinus sylvestris trees near lakeshores and inland sites in northern Sweden (near Kiruna) and central Sweden (near Stockholm) to evaluate the influence of changing microsite conditions on the climate sensitivity of tree-ring δ13C and δ18O. The data reveal a latitudinal trend towards lower C and O isotope values near the Arctic tree line (− 0.8‰ for δ13C and − 2.4‰ for δ18O relative to central Sweden) reflecting widely recognized atmospheric changes. At the microsite scale, δ13C decreases from the dry inland to the moist lakeshore sites (− 0.7‰ in Kiruna and − 1.2‰ in Stockholm), evidence of the importance of groundwater access to this proxy. While all isotope records from northern and central Sweden correlate significantly against temperature, precipitation, cloud cover and/or drought data, climate signals in the records from moist microsites are consistently stronger, which emphasizes the importance of site selection when producing stable isotope chronologies. Overall strongest correlations are found with summer temperature, except for δ18O from Stockholm correlating best with instrumental drought indices. These findings are complemented by significant positive correlations with temperature-sensitive ring width data in Kiruna, and inverse (or absent) correlations with precipitation-sensitive ring width data in Stockholm. A conclusive differentiation between leading and co-varying forcings is challenging based on only the calibration against often defective instrumental climate data, and would require an improved understanding of the physiological processes that control isotope fractionation at varying microsites and joined application of forward modelling.

Comparison of δ13C and δ18O from cellulose, whole wood, and resin-free whole wood from an old high elevation Pinus uncinata in the Spanish central Pyrenees

ABSTRACT δ13C and δ18O values from sapwood of a single Pinus uncinata tree, from a high elevation site in the Spanish Pyrenees, were determined to evaluate the differences between whole wood and resin-free whole wood. This issue is addressed for the first time with P. uncinata over a 38-year long period. Results are also compared with published isotope values of α-cellulose samples from the same tree. The differences in δ13C and δ18O between whole wood and resin-free whole wood vary within the analytical uncertainty of 0.3 and 0.5 ‰, respectively, indicating that resin extraction is not necessary for sapwood of P. uncinata. Mean differences between cellulose and whole wood are 0.9 ‰ (δ13C) and 5.0 ‰ (δ18O), respectively. However, further analyses of different species and other sites are needed to evaluate whether the findings reported here are coherent more generally.

Warm season precipitation signal in δ 2H values of wood lignin methoxyl groups from high elevation larch trees in Switzerland

RATIONALE In this study, we tested stable hydrogen isotope ratios of wood lignin methoxyl groups (δ2 Hmethoxyl values) as a palaeoclimate proxy in dendrochronology. This is a quite new method in the field of dendrochronology and the sample preparation is much simpler than the methods used before to measure δ2 H values from wood. METHODS We measured δ2 Hmethoxyl values in high elevation larch trees (Larix decidua Mill.) from Simplon Valley (southern Switzerland). Thirty-seven larch trees were sampled and five individuals analysed for their δ2 Hmethoxyl values at annual (1971-2009) and pentadal resolution (1746-2009). The δ2 Hmethoxyl values were measured as CH3 I released upon treatment of the dried wood samples with hydroiodic acid. 10-90 μL from the head-space were injected into the gas chromatography/high-temperature conversion/isotope ratio mass spectrometry (GC/HTC-IRMS) system. RESULTS Testing the climate response of the δ2 Hmethoxyl values, the annually resolved series show a positive correlation of r = 0.60 with June/July precipitation. The pentadally resolved δ2 Hmethoxyl series do not show any significant correlation to climate parameters. CONCLUSIONS Increased precipitation during June and July, which are on average warm and relatively dry months, results in higher δ2 H values of the xylem water and, therefore, higher δ2 H values in the lignin methoxyl groups. Therefore, we suggest that δ2 Hmethoxyl values of high elevation larch trees might serve as a summer precipitation proxy.