Description: Interlaboratory comparisons involving nine European stable isotope laboratories have shown that the routine methods of cellulose preparation resulted in data that generally agreed within the precision of the isotope ratio mass spectrometry (IRMS) method used: +/- 0.2% for carbon and +/- 0.3% for oxygen. For carbon, the results suggest that holocellulose is enriched up to 0.39% in C-13 relative to the purified alpha-cellulose. The comparisons of IRMS measurements of carbon on cellulose, sugars, and starches showed low deviations from -0.23 to +0.23% between laboratories. For oxygen, IRMS measurements varied between means from -0.39 to 0.58%, -0.89 to 0.42%, and -1.30 to 1.16% for celluloses, sugars, and starches, respectively. This can be explained by different effects arising from the use of low- or high-temperature pyrolysis and by the variation between laboratories in the procedures used for drying and storage of samples. The results of analyses of nonexchangeable hydrogen are very similar in means with standard deviations between individual methods from +/- 2.7 to +/- 4.9%. The use of a one-point calibration (IAEA-CH7) gave significant positive offsets in delta H-2 values up to 6%. Detailed analysis of the results allows us to make the following recommendations in order to increase quality and compatibility of the common data bank: (1) removal of a pretreatment with organic solvents, (2) a purification step with 17% sodium hydroxide solution during cellulose preparation procedure, (3) measurements of oxygen isotopes under an argon hood, (4) use of calibration standard materials, which are of similar nature to that of the measured samples, and (5) using a two-point calibration method for reliable result calculation.

Description: We present the first European network of tree ring delta(13)C and delta(18)O, containing 23 sites from Finland to Morocco. Common climate signals are found over broad climatic-ecological ranges. In temperate regions we find positive correlations with summer maximum temperatures and negative correlations with summer precipitation and Palmer Drought Severity Indices (PDSI) with no obvious speciesspecific differences. Regional delta(13)C and delta(18)O chronologies share high common variance in year-to-year variations. Long-term variations, however, exhibit differences that may reflect spatial variability in environmental forcings, age trends and/or plant physiological responses to increasing atmospheric CO(2) concentration. Rotated principal component analysis (RPCA) and climate field correlations enable the identification of four sub-regions in the delta(18)O network - northern and eastern Central Europe, Scandinavia and the western Mediterranean. Regional patterns in the delta(13)C network are less clear and are timescale dependent. Our results indicate that future reconstruction efforts should concentrate on delta(18)O data in the identified European regions.