LANDSCAPE RESPONSE TO LATE QUATERNARY CLIMATE CHANGES IN NE QUEENSLAND, AUSTRALIA

The humid tropics of NE Australia (c17degS) experienced major climate changes during the Last Glacial Cycle, as shown by pollen records at Lynch's Crater. At the Last Glacial Maximum rainfall may have been reduced by c60% and sclerophyll woodland replaced rain forest. Present-day rainfalls vary from 2000->5000mm according to altitude. A record of landscape activity is drawn from the E facing escarpment and rivers draining to the Coral Sea. Numerous, mainly fine-grained alluvial fans are developed along the escarpment, from minor catchments and on larger rivers, and these were dated (using TL and 14C AMS) to 25-14ka; they were subsequently entrenched. Extensive colluviation implies increased sediment yield from hillslopes, while the fans on larger rivers may signify loss of stream power. These are effects of reduced rainfall and plant cover, but sedimentation continued under increasing rainfall in the early postglacial climates.

Large landslides and debris flows occur on the escarpment slopes and boulder beds mark alluvial/debris flows in valleys and along the coast. OSL dating of sandy facies within the coarse deposits show good bleaching and provide preliminary dates for some of these events. Two large landslides date to 43+/-4ka and 18+/-3ka, with evidence of renewed Holocene activity. Intercalated sands and boulder beds indicate deposition c80 to c29ka from different sites, again with some Holocene sediments superimposed. Very coarse flow deposits record scattered high-energy events, and these have occurred throughout the LGC. Landscape response to changing Quaternary conditions probably reflected the operation of several major factors including: 1/ hillslope/catchment weathering and sensitivity to short-term climate fluctuations and individual events; 2/ spatio-temporal distribution of extreme events; 3/ short-term 10²y-10³y climate oscillations; 4/ longer-term 10⁴y-10⁵y shifts in rainfall and vegetation patterns. System responses to these factors also involve lags and delays that are poorly understood. Available data cannot resolve the complex timescales involved, but they indicate the importance of understanding these factors.