Establishing a chronological framework for a late Quaternary seasonal swamp in the Australian 'Top End'

摘要

Swamps in the seasonal tropics have good potential for the reconstruction of late Quaternary monsoonal dynamics. Their successful use, however, has often been compromised by chronological limitations introduced by a variety of depositional and post-depositional processes actively modifying the swamp deposits. We here present and discuss the results of a multiple dating approach at Table Top Swamp (TTS) in northern Australia (the 'Top End'). Single-grain luminescence dating of quartz was successfully used to provide chronology in the lowermost core where insufficient organic material prevents the application of radiocarbon dating. In the uppermost, fine-grained and peaty section of the core, two different organic fractions (pollen concentrate and humins) were dated with AMS radiocarbon yielding significantly different chronologies. While this could point to the incorporation of younger pollen into the profile along seasonal dry cracks, older humins may also move up in the profile due to vertical mixing. Additional, spatially highly resolved measurements of the bulk OSL signal (Ln and Ln/Tn) combined with data on down-core variation in K, Th, and U concentration, grain size and moisture content were used to (i) guide the development of an age-depth relationship (i.e. age model) for the entire core based on three different data input scenarios, and (ii) test the applicability of novel luminescence screening techniques in seasonal swamp settings. Results suggest only minor differences among the applied models and scenarios, providing an overall reliable representation of the depositional history in the swamp. Even though all resulting age-depth models have relatively large uncertainties in the lower part of the core, there are significant changes in sedimentation rate over time, providing a chronological basis for a more detailed palaeoenvironmental analysis at TTS. The approach used may also be useful in developing age models in other complex environments, and has show