Spatial-temporal reconstruction of Amazon flood pulse and dry season length over the past century using tree rings and isotopes of floodplain tree species Macrolobium acaciifolium

Climate observations for the Amazon Basin indicate the dry season is becoming drier and longer over recent decades. However, long-term climate data to put these trends in perspective are scarce for most of the Basin, thus hindering our understanding of spatial and temporal variability of the dry season in the Amazon region. I explored the potential of using tree ring-widths and oxygen isotopes in tree rings (d18OTR) of the floodplain tree species Macrolobium acaciifolium as a proxy for past climate conditions during the dry and transitional seasons, the growth period for floodplain trees. I sampled and analyzed wood cores of M. acaciifolium trees from floodplain forests located in western, central and southwest Amazon. M. acaciifolium presented clear ring definitions, but also frequent false rings requiring careful wood anatomic analysis and cross-dating. Despite these limitations, ring-formation dates estimated by cross-dating of ring-widths closely matched bomb-peak radiocarbon dating of tree-rings. Analysis of climate signals in ring-widths and d18OTR revealed potential for reconstruction of past dry season climate conditions, but the strength and nature of climate relationships differed between sites. Ring-width chronologies showed stronger climatic signals at sites located in smaller river catchments from western and southwestern Amazon, where the terrestrial phase of the trees coincided with the peak of the dry season. In contrast, d18OTR chronologies showed stronger climatic signals at the sites in western and central Amazon where growing conditions/seasons are wetter. Long-term increases in d18OTR since 1970 suggest there has been a drying trend during the dry season of the Amazon Basin, consistent with climate observations for this period. This Thesis shows that floodplain trees growing at sites in smaller catchments with mild dry seasons have the greatest potential as proxies for dry season climate. More studies exploring potentials of different floodplain species, including slow-growing old trees, are recommended.