Measurements of stable isotopes of cellulose from tree rings on Holocene samples – influence of plant species and cambial age and their impact on the interpretation of palaeoclimate

Tree rings are one of the most commonly used proxies for palaeoclimatic studies of the Holocene, in particular the width of the rings (TRW) has been found to be valuable for climate reconstructions. However, it has been shown that other tree-ring parameters can also serve as climate proxies providing additional climate information, such as the chemical composition of the wood, in particular the stable cellulose isotopes of the tree rings, which consist of the stable isotopes δ18O, δ13C and δD (or δ2H). The first objective of this work was to complete the measurements of the Eastern Alpine Conifer Chronology (EACC), which consists of more than 7000 samples from about 200 trees of two conifer species collected in the Swiss and Tyrolean Alps at high altitudes at the tree line. In this way, a continuous record of cellulose and isotope content was established, covering the last 9000 years. This large database was designed for palaeoclimate studies as well as for physiological studies. In the initial analysis of the database, we found that one of the two species, the larch, had uniquely low δD values compared to the other species in the database, the cembran pine. We then extended the analysis to several conifer species from the Bern Botanical Garden and confirmed the uniquely low value of the deciduous Larix genus. This new finding was published in the journal Frontiers in Earth Science. This unique signature of larch could be useful in archaeology, as it is important to determine the wood species and it is not always possible to distinguish the sample from spruce and larch using standard wood anatomy analysis. To test whether deuterium measurements can solve the plant genus assignment problem; we analysed about 20 archaeological wood samples that were uncertainly determined. The isotope data allowed us to clearly distinguish the genus in all but one of the samples, although they came from very different sites. The results are included in a manuscript that we submitted to the Journal of Archeological Science. Before using the database for palaeoclimatic studies, it was important to check how the different cambial ages and the different absolute ages of the trees that make up the EACC database affect the isotope records. These two features offer the possibility of investigating the influence of cambial age on tree rings’ isotopic composition to separate it from climatic influences. We found no age-related trend for the three isotopes in samples with a cambial age of more than 100 years. This suggests that the values of the tree isotopes after the 100 cambial years are suitable for palaeoclimatic purposes without detrending. In contrast, trends in the juvenile phase are consistent, with differences between species. We published these results in the Biogeoscience Journal. In a further step, we continued the analysis by investigating whether and how the relationships between isotopes and TRW change with the age of the cambium by examining the correlations at different ages and how different detrending methods affect these relationships. The results showed a changing relationship between carbon and hydrogen across cambial age, which could indicate differential use of reserves in the juvenile phase. We submitted the results of this study to the journal Forests. The EACC isotope database was designed for climate studies throughout the Holocene. Volcanic eruptions are known to produce short thermal climate trends. Studying short climate trends avoids the problem of varying absolute values between sample origins by examining climatic effects within the life span of trees. We use the last 2000 years of the database to examine the effects of volcanic eruptions and which separate the Icelandic eruptions from the tropical eruptions. In addition, we have also studied two eruptions, Tambora 1815 CE and Katla 750 – 764 CE, with an annual resolution of cellulose isotopes. We found that oxygen records the tropical eruption with negative values, in contrast, the hydrogen and carbon isotopes show no obvious variations. In conclusion, we show that the complete EACC isotope database has potential for future analysis or comparison with other databases. Therefore, we plan to make all data publicly available in a suitable archive