Stable isotope dendroclimatology of New Zealand kauri (Agathis australis (D. Don) Lindl.) and cedar (Libocedrus bidwillii Hook. F.)

摘要

This thesis presents research that improves our understanding of the relationship between climate and tree-ring stable isotopes in New Zealand kauri (Agathis australis (D. Don) Lindl.) and cedar (Libocedrus bidwillii Hook. F). As a whole, this research creates a framework from which future investigations of stable isotope dendroclimatology can pursued at high temporal resolution across longer term temporal intervals. Such a framework is required to realize the full potential of New Zealand tree-ring records as archives of palaeoclimatic information. In the Southern Hemisphere, mid-to-high latitude terrestrial climate proxy records are sparse; the long-lived endemic conifers of New Zealand present a rare opportunity to gain valuable insight into Late-Quaternary and Holocene climatic change.A major barrier to long stable isotope dendroclimatic proxy records has been the time-and-resource intensive nature of associated sample preparation and analytical processes. Through experimental assessment of the SBrendel α-cellulose extraction method, this research demonstrates that this simple processing method can, with some caveats, be successfully applied to resinous conifers, facilitating rapid chemical preparation of samples. Further efficiency gains are documented in through use of dual element low temperature pyrolysis of cellulose, providing carbon and oxygen stable isotope determinations on a single sample, rather than the traditionally separate analyses. These preliminary investigations into methodological efficiency enabled the ~3000 stable isotope determinations on kauri and cedar, on which the following investigations into isotopic variability, chronology quality and climate-isotope relationships are based.Both kauri and cedar exhibit substantial intra and inter-tree isotopic variability, resulting in greater than normal sample numbers being required to isolate the common variance in composite time-series. Despite sampling 7 trees (18 cores) at one site and 8 trees (21 cores) at another, only oxygen isotope time-series reach an Expressed Population Signal score of 0.85, a common benchmark of chronology quality. Further investigations using high-resolution sequential sampling of single rings show that this variability extends to a sub-annual scale. Within single rings there are large ranges in δ13C and δ18O, generally corresponding with regular annual cycles of up to 4.6‰ for δ13C and 8.1‰ for δ18O. These data show high frequency (sub weekly) changes in δ18O, demonstrating sufficiently fine resolution for event-to-seasonal scale climate/weather reconstruction. However, that potential is complicated by the lack of strong climatic correlations with isotopic cycles.Due to their stronger common signal, δ18O time series are investigated for inter-annual stable isotope dendroclimatic reconstruction from kauri and cedar. Bootstrapped correlation and multiple regression models show strong relationships between kauri and cedar δ18O and relative humidity, accumulated rainfall and soil moisture deficit. The correlations are strongest for all variables during the period from Autumn prior to growth to the current growth summer. Tree-ring δ18O in kauri and cedar is identified as an effective recorder of regional hydroclimate leading up to and during the Austral growth season.While it is demonstrated that tree-ring δ18O holds significant promise for inter-annual palaeoclimate reconstruction, it is equally apparent from this research that stable isotope dendroclimatology in New Zealand faces significant challenges. Foremost is gaining an improved understanding of seasonal dendrochemical cycles and their relationship with tree-physiology. Increasing the spatial and temporal coverage of tree-ring isotope records is vital to realising New Zealand’s vast potential for stable isotope dendroclimatic reconstruction.