Search

Marine Isotope Stage 11 (MIS 11) from ∼424,000 to 374,000 yrs ago included one of the longest and warmest interglacials of the last 800,000 yrs, and is a potential analogue for the Holocene due to the similarity of Earth's orbital configuration at this time. The question of how the El Nino–Southern Oscillation (ENSO) responds to warmer background climates remains unanswered and is critical to understand how the ENSO system will evolve under the influence of anthropogenic warming. In this study, we present a 35 yr-long, high-resolution record of MIS 11 climate variability in the Western Pacific Warm Pool (WPWP) based on coupled measurements of skeletal Mg/Ca and δ18O in giant Tridacna gigas clams from Huon Peninsula, Papua New Guinea. The δ18O of modern T. gigas from Huon Peninsula faithfully records sea surface temperature, salinity/rainfall and regional ENSO variability. The geochemical integrity of the MIS 11 T. gigas for recording paleo-ENSO events was established through trace element screening, detailed petrography and SEM analysis. The fossil T. gigas δ18O record indicates that ENSO was operating during a 35-yr window in MIS 11, but with fewer events of shorter duration compared to those experienced during the last 100 yrs. The suppressed ENSO variability in the MIS 11 T. gigas record corresponds with a reduction in the amplitude of the average annual cycle in δ18O values. Distinctive changes in local insolation seasonality, and T. gigas δ18O, brought about by changes in Earth's orbit, provide an additional geochronological constraint on the timing of reef growth at Huon Peninsula to around 402 ka during the MIS 11.3 sub-stage (∼424–395 ka).

We report erosion rates determined from in situ produced cosmogenic 10 Be across a spectrum of Australian climatic zones, from the soil-mantled SE Australian escarpment through semi-arid bedrock ranges of southern and central Australia, to soil-mantled ridges at a monsoonal tropical site near the Arnhem escarpment. Climate has a major effect on the balance between erosion and transport and also on erosion rate: the highest rates, averaging 35 m Ma 21 , were from soil-mantled, transport-limited spurs in the humid temperate region around the base of the SE escarpment; the lowest, averaging about 1.5 m Ma 21 , were from the steep, weathering- limited, rocky slopes of Kings Canyon and Mt Sonder in semi-arid central Australia. Between these extremes, other factors come into play including rock-type, slope, and recruitment of vegetation. We measured intermediate average erosion rates from rocky slopes in the semi-arid Flinders and MacDonnell ranges, and from soil-mantled sites at both semi-arid Tyler Pass in central Australia and the tropical monsoonal site. At soil-mantled sites in both the SE and tropical north, soil production generally declines exponentially with increasing soil thickness, although at the tropical site this relationship does not persist under thin soil thicknesses and the relationship here is 'humped'. Results from Tyler Pass show uniform soil thicknesses and soil production rates of about 6.5 m Ma 21 , supporting a longstanding hypothesis that equilibrium, soil-mantled hillslopes erode in concert with stream incision and form convex-up spurs of constant curvature. Moreover, weathering-limited slopes and spurs also occur in the same region: the average erosion rate for rocky sandstone spurs at Glen Helen is 7 m Ma 21 , similar to the Tyler Pass soil-mantled slopes, whereas the average rate for high, quartzite spurs at Mount Sonder is 1.8 m Ma 21 . The extremely low rates measured across bedrock-dominated landscapes suggest that the ridge- valley topography observed today is likely to have been shaped as long ago as the Late Miocene. These rates and processes quantified across different, undisturbed landscapes provide critical data for landscape evolution models.

Differences in chemical weathering extent and character are expected to exist across topographic escarpments due to spatial gradients of climatic and/or tectonic forcing. The passive margin escarpment of south-eastern Australia has a debated but generally accepted model of propagation in which it retreated (within 40 Ma) to near its current position following rifting between Australia and New Zealand 85–100 Ma before present. We focus on this escarpment to quantify chemical weathering rates and processes and how they may provide insight into scarp evolution and retreat. We compare chemical weathering extents and rates above and below the escarpment using a mass balance approach coupling major and trace element analyses with previous measurements of denudation rates using cosmogenic nuclides (10Be and 26Al). We find a slight gradient in saprolite chemical weathering rate as a percentage of total weathering rate across the escarpment. The lowlands area, encompassing the region extending from the base of the escarpment to the coast, experiences a greater extent of chemical weathering than the highland region above the escarpment. Percents of denudation attributable to saprolite weathering average 57 ± 6% and 47 ± 7% at low and high sites respectively. Furthermore, the chemical index of alteration (CIA), a ratio of immobile to mobile oxides in granitic material that increases with weathering extent, have corresponding average values of 73·7 ± 3·9 and 65·5 ± 3·4, indicating lower extents of weathering above the escarpment. Finally, we quantify variations in the rates and extent of chemical weathering at the hillslope scale across the escarpment to suggest new insight into how climate differences and hillslope topography help drive landscape evolution, potentially overprinting longer term tectonic forcing. Copyright © 2009 John Wiley & Sons, Ltd.

Development of continental aridity has been linked to late Cenozoic global cooling, but the evidence is indirect, based on terrestrial loess deposits and eolian silt in marine sediments, whereas direct dating of the inception of arid landforms has been frustrated by a lack of suitable methods. Here we report the first age determination of a major arid-zone dune field, based on cosmogenic 10 Be and 26 Al measurements of drill cores from dunes in the Simpson Desert, central Australia. Results show that the dune field began to form ca. 1 Ma, whereas dating using quartz optically stimulated luminescence indicates episodic dune building during late Quaternary ice ages. Less intense desertification began earlier; the previous cosmogenic exposure dating showed that neighboring stony deserts began to form at the onset of Quaternary ice ages 2–4 Ma. Aridity deepened and the dune field formed when ice age cycles increased their amplitude and switched their periods from 40 k.y. to 100 k.y. ca. 1 Ma.

The sustainability of soil is a major issue for society. In principle, the evolution of soil resources can be constrained by comparing the rates of soil erosion and production. Cosmogenic isotopes provide one measurement of soil erosion rates. They can also be used to estimate soil production rates but only if erosion is assumed to be balanced by production. This implies that the evolution of soil resources (thinning, thickening or constant) since soil thickness is assumed to be constant with time. Here we utilise an independent method to estimate soil production rates, using uranium-series (U-series) isotopes. The study of a site in temperate Australia undisturbed by human activity shows that soil production rates inferred from U-series isotopes are similar to erosion rates derived from beryllium-10 (10Be) measurements, implying that at this site there is no net accumulation or loss of soil. Saprolite production rates (the migration rate of the weathering front into the bedrock) are also similar to erosion rates so the thickness of the entire weathering profile is effectively in steady-state. This study demonstrates that the combination of U-series and cosmogenic isotopes can be used to quantitatively assess soil evolution and the development of weathering profiles. Preliminary observations suggest that the rate of bedrock weathering (i.e. saprolite production) in temperate Australia is of the same order of magnitude as that inferred for laterites in tropical climates. This may suggest that, for thick weathering profiles, although the extent of weathering strongly differs between temperate and tropical climates, the migration of the weathering front into the bedrock occurs at a relatively uniform rate regardless of present-day climatic conditions.

Estimates of regional erosion and sediment mixing from different sources in the Yangtse River system are presented, based on sand samples collected from major tributaries and the trunk stream, at 23 sites between western Sichuan and the Yangtse Delta. Mixing is estimated from concentrations of Mg, Ca, Sr, Ti, Mn and Fe, which are substantially higher in sand from major tributaries in the western Yangtse River catchment than from tributaries in the eastern catchment. Intermediate concentrations occur in sand from the Yangtse Delta, both for modern samples from the surface and for early Holocene samples from drill holes. Mixing ratios indicate that 35 ± 5% of sand in the delta came from eastern sources. A similar result was obtained using cosmogenic 10Be in quartz grains as a tracer of mixing. Regional erosion rate estimated from 10Be in sand grains from high mountain catchments of the western Yangtse River are mostly similar to rates based on sediment gauging but are sometimes higher, and range to over 700 m Ma− 1, while 10Be measured at upper Yangtse River tributaries on the northeast Tibetan plateau gave rates of 20–30 m Ma− 1. For the eastern catchments, 10Be measurements from quartz sand and sediment gauging both gave rates of 30–70 m Ma− 1. Eroding at this rate, the eastern catchments could not supply more than 20% of the sediment in the delta, in contrast with 35% estimated from geochemical fingerprints. The relative input from eastern sources may have been higher in Late Pleistocene times, under a different climatic regime, and reworking of Pleistocene deposits may still be in progress.

Since the end of the post-glacial sea level rise 6800 years ago, progradation of river mouths into estuaries has been a global phenomenon. The responses of upstream alluvial river reaches to this progradation have received little attention. Here, the links between river mouth progradation and Holocene valley aggradation are examined for the Macdonald and Tuross Rivers in south-eastern Australia. Optical and radiocarbon dating of floodplain sediments indicates that since the mid-Holocene sea level highstand 6800 years ago vertical floodplain aggradation along the two valleys has generally been consistent with the rate at which each river prograded into its estuary. This link between river mouth progradation and alluvial aggradation drove floodplain aggradation for many tens of kilometres upstream of the estuarine limits. Both rivers have abandoned their main Holocene floodplains over the last 2000 years and their channels have contracted. A regional shift to smaller floods is inferred to be responsible for this change, though a greater relative sea level fall experienced by the Macdonald River since the mid-Holocene sea level highstand appears to have been an additional influence upon floodplain evolution in this valley. Copyright © 2006 John Wiley & Sons, Ltd.

Closely-spaced transects measured by excimer laser ablation inductively-coupled plasma mass spectrometry (ELA-ICPMS) at 5 and 32 μm spatial resolution are used to generate trace element composition maps (Ba, Sr, Mg, U, Na, P and Al) from MND-S1, a previously studied modern stalagmite from southwest Australia ( Treble et al., 2003 : EPSL 216: 141). Rainfall at the site is highly seasonal, and trace elements in MND-S1 show strong seasonal variation. Trace element maps show that Ba, Sr, U and Na concentrations coherently follow annual growth layers identified from Scanning Electron Microscopy (SEM) images. The SEM images also reveal that stalagmite growth did not proceed uniformly: growth layers vary in thickness and locally pinch out. Highly preferential crystal growth, determined by nucleation sites left by the previous year’s growth, may be responsible for this uneven growth layering. Differential crystal growth apparently causes variability of trace element concentrations along each annual layer, although additional disequilibrium processes affect Mg, which is less distinctly banded than Ba, Sr, U and Na. Uneven and discontinuous growth layers influence the number of annual cycles, their wavelengths and seasonal amplitudes measured in any one transect. This has clear implications for studies that use annual trace element cycles as chronological markers, growth rate or seasonality proxies.

We present a record of seasonal and inter-annual oxygen (δO) and carbon (δC) isotope ratios from an 81-year-old stalagmite from Moondyne Cave, southwest Australia. The growth history of stalagmite MND-S1 is known since it grew on a cave boardwalk that was installed in 1911 and removed in 1992. This stalagmite provides an excellent test of speleothem climate proxies because the regional climate is strongly seasonal (wet winter/dry summer) and has experienced a 200 mm (20%) reduction of mean rainfall since the mid-1960s, and a 0.8 °C temperature rise since ∼ 1953. Seasonal variations in calcite δO were measured in situ by high spatial resolution ion microprobe, whilst inter-annual variations of δO and δC were measured by conventional gas-source mass spectrometry. Comparison of the speleothem stable isotopes and instrumental temperature records reveals that δO variations are too large to be driven by temperature alone, and are in the opposite sense. However, daily rainfall δO measurements show that the mean seasonal range in δO of rainfall in southwest Australia is large (2‰) and inversely correlated with rainfall amount. A rainfall driver for the speleothem δO is confirmed by the detection of seasonal shifts of 0.7-1.5‰ in speleothem δ that track rainfall δO, smoothed by storage in the overlying limestone. The seasonal range in speleothem δO is larger than any inter-annual and decadal variation observed in the record. The prominent annual cycles in speleothem δO revealed by ion microprobe analysis indicate that subtle changes in the frequency of intense winter rainfall events, or possibly also moisture sources, could produce significant changes in mean speleothem δO. The ion microprobe results also raise the possibility that the masses of speleothem calcite deposited in winter and summer could vary as a function of the seasonal drip rate and carbonate saturation state of these waters. If this is the case, then small changes in the relative masses of calcite deposited in winter and summer could produce significant shifts in mean δO and δC that have a complex relation to climate. This finding should be generally applicable to the interpretation of long-term trends in speleothem geochemical records for shallow cave sites where seasonal variations in geochemical tracers are relatively large, including most of the sub-tropical monsoon belts and mid to high latitudes with distinct rainfall seasons.

Moraines on Schnells Ridge, southwest Tasmania, have been dated using in situ 10Be. An age of 19,400 ± 600 yr is indicated for the well-preserved innermost moraine from consistent measurements on four large quartzite boulders. This corresponds closely with exposure ages reported by T.T. Barrows et al. (2002, Quaternary Science Reviews 21, 159–173) for Last Glacial Maximum glacial features farther north in Tasmania and southeast Australia. In contrast, ages between 39,000 and 141,000 yr were obtained from a series of boulders on a more extensive outer moraine, indicating that this has had a more complex history.

High spatial resolution measurements of Mg, P, U, Sr, Ba and Na using 193-nm excimer laser-ablation inductively coupled plasma mass spectrometry has revealed clearly resolvable annual cycles in a modern speleothem from southwestern Australia. The age of this stalagmite is established by the dates of emplacement and removal of the boardwalk where it grew (1911–1992). This chronological constraint allows for the first confident comparison between the instrumental climate record and speleothem trace element content. Eleven laser-ablation transects across a ∼0.5 mm wide section of the speleothem growth axis were measured. Cycles that could be matched between adjacent transects were stacked into a master record to minimise variability between tracks and permit greater confidence in comparing the trace elements to the instrumental climate record. P and U positively, and Mg inversely, mimic the sudden 20% decrease in annual rainfall experienced by this region since 1965. We argue that P from seasonal vegetation decay is transported to the speleothem before it is mineralised in the soil, owing to the low P retention capacity of southwest Australian soils. Both vegetation activity and the transport of vegetation-derived HPO42− are sensitive to rainfall. Groundwater P concentration may also influence the transport of U through the strong affinity between phosphate and uranyl ions. Mg appears to be sensitive to groundwater residence time as this affects drip-water Mg/Ca composition by preferential loss of Ca during drier episodes when calcite precipitates before reaching the stalagmite. The effects of groundwater residence time may also be important for Sr on inter-annual scales. However, the behaviour of Sr on the annual cycle is opposite to Mg but compatible with Ba and Na, which are shown to depend on speleothem growth rate.

Geochemical anomalies and growth discontinuities in Porites corals from western Sumatra, Indonesia, record unanticipated reef mortality during anomalous Indian Ocean Dipole upwelling and a giant red tide in 1997. Sea surface temperature reconstructions show that although some past upwelling events have been stronger, there were no analogous episodes of coral mortality during the past 7000 years, indicating that the 1997 red tide was highly unusual. We show that iron fertilization by the 1997 Indonesian wildfires was sufficient to produce the extraordinary red tide, leading to reef death by asphyxiation. These findings highlight tropical wildfires as an escalating threat to coastal marine ecosystems.

Sea level cycles recorded by coral terraces at Huon Peninsula (HP) in Papua New Guinea match rapid climate changes that occurred between 30,000 and 65,000 years ago, seen in Greenland ice cores and marine sediments. Each cycle of 6000–7000 years ended with a sea level rise of 10–15 m lasting 1000–2000 years, following a longer period of falling sea level. Precise dating shows that each rise corresponds to a “Heinrich” episode of ice-rafted detritus in north Atlantic, signalling massive ice outbreak from north America. Sea level may be the trigger that forced near-synchronous breakouts of ice from north America and eastern Greenland. The HP sea level changes also correspond to similar cycles in benthic oxygen isotopes reported from the north Atlantic, which are too large to be explained entirely in terms of ice volumes. Jointly, the sea level and isotope records suggest that the north Atlantic deep ocean cooled as sea level fell and warmed as sea level rose, in each 6000-year cycle.

Samples of plankton, surface sediments and Holocene deposits from the South Alligator River, an energetic macrotidal river in northern Australia, are examinedto evaluate the usefulness of estuarine foraminifers for palaeoenvironmental interpretation of sediments deposited in tropical macrotidal conditions. Although marine and estuarine foraminifers in the estuarine channel are mixed by strong tidal currents, there are significant trends along the 80-km length of the tidal river. The relative proportions of marine andporcellaneous taxa d ecrease upstream from the mouth, but hyaline taxa andthose with brackish affinities increase upstream. Foraminifers in intertidal sediments have distinctly different assemblages in the uppermost intertidal, mid-tidal (mangrove), lower intertidal and subtidal zones. In addition to the effects of tidal mixing, preserved thanatacoenoses are affected by selective dissolution and/or pyritisation within the sediment bodies. Despite the effects of adverse taphonomic factors, foraminiferal assemblages preserved in drill cores provide an accurate guide to original depositional environments (intertidal/subtidal; brackish estuarine/coastal marine), in mid-Holocene sediments in the South Alligator area. Foraminifers are found to improve the precision of previous palaeoenvironmental reconstructions made on the basis of sedimentary facies and pollen analyses. r 2001 Elsevier Science LtdandINQUA. All rights reserved .

Lying in a semi-arid region, the Flinders Ranges have rocky, weathering-limited hillslopes and streams within the ranges are active only during rare downpours, when they transport boulders, gravel and coarse sand. However, silt- and clay-rich valley-fill deposits, incised by present streams, occur as terraces and terrace remnants. Such deposits are not accumulating today. In and upstream of Brachina Gorge, in the central ranges, these remnant valley-fills are exposed in bank sections up to 18 m high. At some localities, exposures show horizontal, centimetre-scale beds of fine sand and clayey silt that can be traced for several hundred metres, and contain gastropods, diatoms and phytoliths consistent with sluggish, shallow water flow under fresh to brackish conditions. AMS 14 C and OSL dating shows that these valley-fill deposits accumulated between ∼33 and 17 ka, an interval that embraces the Last Glacial Maximum (LGM). Formerly thought to be lake beds (Aust. J. Earth Sci. 46 (1999) 61), the valley fills are shown here to have accumulated in a fluvial wetland that extended westward from the middle reaches of Brachina Creek, through Brachina Gorge and joined with aggraded fan deposits beyond the ranges. Aggradation of this wetland requires a substantial reduction of both rainfall variability and evaporation, which would be favoured by the low temperatures and reduced incursions of summer rainfall. This is consistent with climatic reconstructions by Miller et al. (Nature 385 (1997) 241) and Johnson et al. (Science 284 (1999) 1150) for Lake Eyre, north of the Flinders Ranges, at the LGM. Demise of the wetland was heralded by a major influx of coarse alluvium followed by channelling and dissection. Erosion was interrupted by an episode of aggradation and floodplain widening, represented by remnants of a wide terrace inset below the primary wetland surface. This episode, which is interpreted as a return to lower climatic variability, ceased with establishment of the present climatic regime, which has resulted in stripping of the Late Pleistocene deposits from much of the Brachina valley.

Late Quaternary rates of apparent soil production, bedrock incision, and average erosion are determined for the southeastern highlands of Australia using in situ produced cosmogenic nuclide concentrations of 10 Be and 26 Al. Apparent soil production rates define a steep, inverse exponential function of soil depth with a maximum of 143 m Ma � 1 under zero soil depth. There were no observed soil depths between about 25 cm and zero, however, such that the maximum observed rate is about 50 m Ma � 1 . The Bredbo River catchment average erosion rate is 157 1mM a � 1 , and is similar to the average hillslope erosion rate of 167 1mM a � 1 . Bedrock incision rates average 9 mMa � 1 and suggest that the higher rates of hillslope erosion may be in response to a pulse of incision, perhaps generated by knickpoint propagation. Bedrock erosion rates inferred froma tor profile average 3.8 mMa � 1 , with higher rates on other, more weathered tor tops. An aboveground tor profile of nuclide concentrations is consistent with a simple model of rapid stripping of the surrounding saprolite, supporting the view that at least one episodic period of increased denudation has affected the landscape evolution of the highlands. We test this hypothesis by using a simple landscape evolution model to reasonably predict the spatial variation of soil depth as well as the emergence of tors. r 2001 Elsevier Science Ltd and INQUA. All rights reserved.

Sea level change during the Quaternary is primarily a consequence of the cyclic growth and decay of ice sheets, resulting in a complex spatial and temporal pattern. Observations of this variability provide constraints on the timing, rates, and magnitudes of the changes in ice mass during a glacial cycle, as well as more limited information on the distribution of ice between the major ice sheets at any time. Observations of glacially induced sea level changes also provide information on the response of the mantle to surface loading on time scales of 10 3 to 10 5 years. Regional analyses indicate that the earth-response function is depth dependent as well as spatially variable. Comprehensive models of sea level change enable the migration of coastlines to be predicted during glacial cycles, including the anthropologically important period from about 60,000 to 20,000 years ago.

A series of storm deposits at Curacoa Island, a steep bedrock island in the Palm Group, north Queensland, record the passage of cyclonic events. Storm ridge accretion occurs when cyclonic waves and surge combine to transport coarse sediment, mostly broken coral, onto the shore and deposit it above normal tidal levels. Narrow fringing reefs at Curacoa act as a rapidly replenished sediment source, augmented by lithic gravel transported longshore. Beachface, stormberm, ridge platform and washover deposits can accrete during a single event. A series of storm ridges were examined via stratigraphic pits and trenches on two traverses; deposits from 22 events were identified and their ages were determined using 68 radiocarbon dates from coral clasts. Most dates from each storm deposit formed statistically accordant groups. Only 15% of the dates were older than the group ages and were interpreted as reworked specimens. Accordant ages within each group were combined and converted to calibrated radiocarbon ages. Ridge ages decrease uniformly with distance from the oldest ridge, indicating uniform depositional processes over the last 5000 years. A time series of storm deposit ages indicates that cyclone frequency was statistically constant over the last 5000 years. This suggests that cyclone frequency may not have been affected by sea surface temperatures in the region, which appear to have been about 1°C warmer about 5000 years ago (Gagan, M.K., Ayliffe, L., Hopley, D., Cali, J., Mortimer, G., Chappell, J., McCulloch, M., Head, M.J., 1998. Temperature and surface water balance of the mid-Holocene tropical western Pacific. Science 279, 1014–1018).

Eighty-eight conventional radiocarbon age measurements of corals, collected from Holocene transgressive reefs at Huon Peninsula have been published previously, but, except for 29 dates from a 52 m drillhole, the data have not been systematically analysed. From these data, we show that post-glacial reef growth tracked rising sea level (‘keep-up’ growth) at Huon Peninsula, from −70 m at 11,200 yr BP. The Holocene reef emerged about 6300 yr BP in the southeast of the region, where uplift rate is highest (3.3 m ka-1), and at about 5950 yr BP in the northwest, where uplift rate is 1.9 m ka-1. The uplift process is dominated by metre-scale coseismic uplift events; the difference in timing of reef emergence between northwest and southeast, is related to differences in the stepwise uplift of the two regions. Although uplift is stepwise, the uplift-corrected age–height data from the dated corals fall within a tight envelope, which represents a well defined eustatic sea level curve for the calibrated age interval from 7000 to 13,000 yr BP.

The eustatic component of relative sea-level change provides a measure of the amount of ice transferred between the continents and oceans during glacial cycles. This has been quantified for the period since the Last Glacial Maximum by correcting observed sea-level change for the glacio-hydro-isostatic contributions using realistic ice distribution and earth models. During the Last Glacial Maximum (LGM) the eustatic sea level was 125±5 m lower than the present day, equivalent to a land-based ice volume of (4.6–4.9)×107 km3. Evidence for a non-uniform rise in eustatic sea level from the LGM to the end of the deglaciation is examined. The initial rate of rise from ca. 21 to 17 ka was relatively slow with an average rate of ca. 6 m ka−1, followed by an average rate of ca. 10 m ka−1 for the next 10 ka. Significant departures from these average rates may have occurred at the time of the Younger Dryas and possibly also around 14 ka. Most of the decay of the large ice sheets was completed by 7 ka, but 3–5 m of water has been added to the oceans since that time.

Many landslides disrupt the coral terraces at northeast Huon Peninsula, Papua New Guinea. Each landslide is younger than the youngest terrace which it intersects and is older than the oldest coral reef or shoreline that overlies it. The ages of the raised coral reefs and terraces were established by previous studies using radiocarbon and uranium-series dating. The times of occurrence of 26 late Quaternary landslides are reported here; 7 are based on new radiocarbon dates from in situ corals or molluscs associated with landslide deposits, and the remainder are based on terrace ages that were established previously. The landslides include slumps, block slides and block flows which mostly involve failure of coral limestones of the terraces, and debris flows which originate from large landslides that cut through the coral limestones into underlying fan-delta and tuffaceous sedimentary rocks. The proportion of terrace area affected by Late Quaternary landslides ranges from 50% per km2 and depends on average gradient of the landsurface, tectonic uplift rate, and the thickness and geometry of the raised reefs and terraces. Clusters of landslides, which occurred at about 800–1250 and 6300–7800 yr BP, may have been triggered by large earthquakes which also caused metre-scale coseismic uplift. However, we did not identify landslides that could have been associated with every large coseimic uplift that has been documented in this region.

Up to six regressive terraces occur on the Holocene raised reef tract and up to 15 occur on late Pleistocene raised reef tracts along 40 km of coastline at Huon Peninsula, Papua New Guinea. We suggest that the regressive terraces represent repeated episodic uplift caused by great earthquakes. Ages of Holocene coseismic uplift events are established by radiocarbon dating; the late Pleistocene events are bracketed by U series ages of the raised reef tracts on which they occur. The mean recurrence interval of great earthquakes that caused the uplift events is 970 to 1165 years in the Holocene and probably the same in the late Pleistocene; the interval ranged from about 200 to 1900 years. The uplift rate increases parallel to the coast from northwest to southeast, and the amplitude of coseismic uplifts generally increases similarly, although some events produced uplift with little shore-parallel tilting. The mean amplitude of coseismic uplifts throughout the study area is ∼3 m in for both Holocene and late Pleistocene sequences. Large, late Quaternary landslides are numerous, and some probably were triggered by the great earthquakes that caused coseismic uplift. There appears to be no continuum between historical large earthquakes at Huon Peninsula of magnitudes >7 that produced no or only minor uplift, and the great earthquakes represented by meter-scale coseismic uplifts and very large landslides. Two tectonic subregions are recognized, which were uplifted together by some Holocene events but not by others. There is no surface trace of Holocene faulting between the subregions, and a buried fault is thought to separate them.

High-precision measurements of SrCa ratios are reported for Porites corals from the uplifted Holocene coral terraces at Huon Peninsula, Papua New Guinea. The early Holocene Porites have UTh mass spectrometric ages of 8920 ± 60 yr and 7370 ± 50 yr, and δ234U(t) values of 145 ± 2, similar to modern seawater. The SrCa coral records provide 5–6 year high resolution (near weekly) time windows into early Holocene sea surface temperatures. Seasonal temperature fluctuations are generally in the range of ± 1°C, with occasional excursions of ± 2°C, which may indicate the more frequent recurrence of very strong ENSO (El Nino-Southern Oscillation) events. Mean annual SrCa temperatures of 24.2 ± 1.1°C and 22.9 ± 0.8°C have been obtained, which are ∼ 2–3°C cooler than that exhibited by a modern Porites. These results indicate that, during the early Holocene, the equatorial western Pacific ocean was at least several degrees cooler than present-day temperatures. This is consistent with late glacial coral records from the Caribbean that indicate lower (∼ 6°C) sea surface temperatures for the equatorial oceans. The Huon Peninsula corals also indicate that SSTs were several degrees cooler than those in the Caribbean during the early Holocene. Thus, although the northern hemisphere summer radiation maximum occurred at ∼ 10 ka, there appears to have been a significant lag in the response of the equatorial western Pacific ocean to this warming. Cooler early Holocene sea surface temperatures in the western Pacific may have been due to changing patterns of ocean-atmosphere circulation, resulting from the exposure of large areas of continental shelf in the southeast Asia region, a consequence of lower glacial sea levels. It is likely that ocean temperatures in the Huon Peninsula were influenced by the opening at ∼ 7 ka of the Torres Strait, that now separates New Guinea from the Australian mainland.

A major discrepancy exists between the late Quaternary sea level changes derived from raised coral reef terraces at Huon Peninsula in Papua New Guinea and from oxygen isotopes in deep sea cores. The two methods agree closely from 120 ka to 80 ka and from 20 ka to the present. However, oxygen isotopes from benthic forams indicate lower sea levels by 20 to 40 m during the later stages of the last glacial cycle, from 70 to 30 ka. New, high precision U-series age measurements and revised stratigraphic data for Huon Peninsula terraces that formed during this interval give similar sea levels to those based on planktonic and benthic δ18O data. The record of ice volume changes through the last glacial cycle now appears to be reasonably complete.

A flight of coral terraces, up to ca. 1, 000 m above sea level, is well preserved on the northern coast of the Huon Peninsula, Papua New Guinea, and records the sea level and tectonic history since ca. 300 ka. The region is located near the boundary between the Western Pacific and Australian plates. The uplift rate since the last interglacial maximum (oxygen isotope stage 5e) ranges from 0.7 m/ka at the northwest end of the coast, and up to 3.5 m/ka towards the southeast end. Late Quaternary paleoseismicity is estimated from subdivided Holocene and late Pleistocene terraces, each of which records meter-scale coseismic uplift events.The study area covers about 40 km of coast. Holocene coral terraces, up to 25 m above sea level, are divided into seven levels at maximum, four to five at most sites. The highest Holocene terrace is a reef crest representing former sea level of the culmination of the sea level rise at ca. 6-6.6 ka BP. The lower terraces are regressive terraces, which record successive intermittent uplift events, probably caused by great earthquakes. Timing of the Holocene uplift events is ca. 2.5, 3.8, 4.4-5.1, and 5.4 ka BP at Kwambu-Kilasairo area on the northwestern coast, and at ca. 0.8, 1.4, 1.8, 2.5, 3.9 and 5.4 ka BP at Kanomi-Nanda area to Hubegong on the central and southern coast. Difference in the timing of the uplift events implies that there are at least two tectonic suregions on this coast.Similar small regressive terraces are also recognized from the detailed profiles of late Pleistocene terraces. For example, in the central and southern part of the study area, 15-7 extra steps occur between transgressive terrace IIIa (ca. 52 ka) and terrace II (ca. 38 ka). Thus, we infer that repeated meter-scale uplift events have occurred at least since ca. 52 ka. The apparent interval of uplift events, about 1-1.5 ka, resembles to that for Holocene.Such a meter-scale uplift with ka-scale recurrence intervals records a different earthquake deformation cycle from the centimeter-scale uplift associated with the 1992 May earthquake (M= 7.2 : Pandolfi et al., 1994). Such a small event with shorter recurrence interval is apparently not expressed geomorphologically as in the form of regressive terrace.

Late Quaternary stratigraphy of a 50 km2 catchment on the south-eastern highlands of Australia reveals processes and history of denudation, and helps resolve a long-standing debate about factors controlling episodic valley aggradation and degradation during Holocene times. Valley sedimentation occurred when swampy vegetation fully colonized valley floors and obliterated all channels, promoting aggradation for periods of several thousand years, with most incoming sediment being trapped in swampy meadows. Much of the sediment was reworked from late Pleistocene alluvial fan and valley fill deposits, and primary hillslope erosion was minor during the Holocene. Differing sedimentation patterns between the Late Pleistocene, Holocene and Post-European settlement periods reflect regional changes in sediment supply and transport capacity as a result of major environmental change. Within the Holocene, however, valley fill stratigraphy is controlled by massive, episodic gully erosion terminating aggradation. Gully initiation appears to be controlled more by thresholds of incision into vegetated valley floors than by changes to sediment supply. Whether the thresholds are exceeded because of climatic change, autonomous change or extreme events cannot yet be determined. Overall, the Holocene history represents continuing complex response to events of the Late Pleistocene, and does not support the K-cycle concept, which has strongly influenced late Quaternary geomorphology in Australia.

There are two widely used sea level curves for the last glacial cycle, one based on coral terraces at Huon Peninsula (HP) in Papua New Guinea, the other on oxygen isotope data from deep sea cores. Previous sea level estimates from HP are 20-40 m higher than isotopic sea levels for the interval 30-75 ka but new sea level data from HP support the isotopic sea levels. Computer simulations of coral terraces are more similar to observed HP terraces when the computer model is driven by the isotopic sea level curve than by the previous HP sea level curve. HP sea levels and benthic isotope records were used previously to calculate deep sea temperatures through the last glacial cycle the new HP results lead to a sharper indication that deep sea temperatures were up to 1.8°C cooler during the last glacial cycle than in the interglacials. The temperature estimates depend on the isotopic composition of Pleistocene ice sheets. However, if deep sea temperatures did not change, the HP sea levels require unrealistically low oxygen isotope values for the Pleistocene ice sheets. The model of cooler deep sea temperatures during glacial cycles is adopted here and was used to derive isotopic sea level changes using the V19-30 record over the last 330 ka. The isotopic sea levels before the last interglacial are consistent with the limited data available but need to be tested by high quality data from older coral terrace sequences.

The extensive flight of uplifted reef terraces which occurs along the Vitiaz strait on the northern flank of the Huon Peninsula in PNG (Papua New Guinea) contains a particularly good record of sea level changes in the last 250 ky. The Huon terraces were the target of an international expedition which took place in July–August 1988. In particular, we searched for suitable samples for U-series dating in a reef complex designated as VII, which is correlated with the last interglacial episode and high sea level stand. This complex is composed of a barrier reef (VIIb), a lagoon, and a fringing reef (VIIa). Twelve corals from these terraces and two corals from the older reef complex VIII were selected for analysis. The petrography, oxygen and carbon isotope compositions, and magnesium and strontium concentrations were determined along with the concentrations and isotopic compositions of uranium and thorium. ^(230)Th-^(234)U ages of the corals with > 99% aragonite, having primary textures, and which show U/Sr ratios around 0.4 × 10^(−3) and initial δ^(234)U values close to that of present seawater, appear to be reliable. The “most reliable” ages from complex VII corals fall in two tight groups centered at 118 ky and 134 ky. Corals with δ^(234)U(T) values higher than 160 and U/SR ratios substantially lower than seawater are assumed to have undergone diagenetic alteration, which appears to be common in this area. The simplest model for sea level height for terrace VII is a continuous rise between 134 and 118 ky. Alternatively, there may have been two periods of rapid sea level rise. In contrast, in the Bahamas, there is evidence that sea level remained rather constant over the time interval 132 to 120 ky. The absence of ages between 132 and 120 ky in PNG could be the result of changes in the local tectonic uplift rates during that time, or erosion that disrupted the continuous record. In any event, we find no basis for accepting a single brief time for the age of the last interglacial and applying this age as a precise chronometer for worldwide correlation, or as a test of climatic models. The older ages reported here precede the Milankovitch solar insolation peak at 128 ky, and the younger ages are ~ 10 ky after this peak. If the present high-precision data are correct, then it will be necessary to reassess the validity of the Milankovitch theory of climatic changes. The fundamental issue which must now be resolved is a means of identifying coral samples that have not been disturbed by diagenetic processes.

The McArthur River drains from a semiarid, sandstone catchment into a shallow embayment behind the Sir Edward Pellew Group of islands in the southwestern Gulf of Carpentaria. It has built a broad Holocene delta, presently with two active distributaries and several abandoned, mangrove-lined, former distributaries. Augering indicates that much of the delta is underlain by shelly sands which contain distinct shell beds in their position of growth. These are interpreted as delta front deposits, and the elevation of the landwardmost beds above high tide level implies emergence of 1–2 m over the last 4000 years. This relative sea-level fall appears to have been a major cause of rapid mid-Holocene delta progradation. The eastern margin of the delta has undergone little net progradation over the last 2000 years, though there has been accretion of small mangrove-covered islands to the northwest of the delta. Distributaries have migrated across the upper deltaic plain by lateral migration, leaving nested sequences of fluvial ridges. In the lower deltaic plain, channel migration appears to have occurred mainly by avulsion; former distributaries have been infilled with fluvial sands and are now tidally dominated.

There are several estuaries associated with the pronouncedly seasonal rivers which drain northwards from the Middle Proterozoic sandstone Arnhem Land plateau, and the Tertiary Koolpinyah land surface, into the macrotidal van Diemen Gulf, in the Northern Territory of Australia. The Holocene development of these, investigated in greatest detail for the South Alligator River with an upland catchment of > 10,000 km2. through drilling, palynology and radiocarbon dating, comprises both estuarine infill and coastal progradation. Three phases of estuarine infill can be recognised: (i) a transgressive phase (8000-6800 years B.P.) of marine incursion; (ii) a big swamp phase (6800-5300 years B.P.) of widespread mangrove forest development; and (iii) a sinuous/cuspate phase of floodplain development since 5300 years B.P., during which the tidal river has meandered and reworked earlier estuarine sediments. Since 6000 years B.P., the South Alligator coastal plain has prograded at a decelerating rate, with two phases of chenier ridge formation. A similar pattern of estuarine infill, and decelerating coastal plain progradation, is demonstrated for the Adelaide and Mary Rivers, both with catchments of > 6000 km2. The southern shore of van Diemem Gulf appears to have changed its overall position little during the last 2000 years. The major source for the clay, silt and fine sands which have infilled the estuary and coastal plain has been from seaward. Dispite the similarity of development, coastal sediment build up has had different effects on the morphology of each tidal river. The Adelaide has undergone a major diversion and no longer flows directly into van Diemen Gulf, but occupies a former fluvial course, and the Mary has been blocked entirely, and its former estuarine palaeochannels have been infilled with tide-transported sediment.

SUMMARY At present, the Ramu River in Papua New Guinea flows directly into the Bismarck Sea, while the adjacent Keram flows into the Sepik River. There is evidence to suggest that the Ramu previously occupied the existing Keram water course, and has been gradually diverting its discharge into its present channel, probably over the last 4,000 years. The diversion process is not complete yet, and during the wet season flood water still travels across these two rivers. The most direct evidence to support this proposal of river diversion is that there are much larger meander bends on which the existing Keram stream channel is superimposed. The diversion of the Ramu is believed to be a result of relative sea‐level rise in the late Quaternary which altered the gradients of the Keram and Ramu rivers, and may have been facilitated through neo‐tectonic movement of the floodplain.

The march of environmental changes in eastern Australia, from the Last Interglacial to the Holocene, is reviewed. Palynologic evidence suggests that forest composition fluctuated substantially during Oxygen Isotope (OI) Substages 5a to 5e, but physical evidence for this period is as yet poorly defined. During OI Stage 3 and possibly part of 4, inland lakes were larger and the hydrologic balance was wetter than at any time since. Conditions were coldest, and dune activity was maximal throughout the continent, during OI Stage 2, but driest conditions as judged from lake levels occurred at the termination of the Pleistocene, when temperatures were rising rapidly. Lags between changes of boundary conditions and responses of natural systems are clearly seen in Holocene behaviour of coasts and reefs, and the effects of extreme events rather than climatic changes are seen in some sedimentary sequences. Establishment of the chronology of deposits older than about 35 ka, and forging links between the terrestrial and deep-sea records, are major challenges for future research.

Passive margin escarpments are extensively studied around the world, and understanding their evolution continues to present one of the more compelling interdisciplinary challenges tackled by earth scientists today. Escarpments reflect the mor photectonic development of passive margins and can separate regions with different climatic histories, but the inferred rapid rates of escarpment retreat have been at odds with actual measurements of land surface denudation. In this paper we present results from extensive cosmogenic 10 Be and 26 Al analyses across the escarpment of southeastern Australia to quantify the erosional processes evolving the highland, lowland, and scarp face landscapes. We document new relationships between soil production rates and soil thicknesses for the highland and lowland landscapes and compare these soil production functions with those published in our earlier studies from the highlands and at the base of the escarpment. Both new functions define exponential declines of soil production rates with increasing soil depths, with inferred intercepts of 65 and 42 m/m.y. for the highland and lowland sites, respectively, and slopes of –0.02. Exposed bedrock at both of the new sites erodes more slowly than the maximum soil production rates, at 22 ± 3 and 9 ± 2 m/m.y., respectively, thus suggesting a “humped” soil production function. We suggest that instead of a humped function, lithologic variations set the emergence of bedrock, which evolves into the tors that are found extensively across the highlands and at the crest of the escarpment by eroding more slowly than the surrounding soil-mantled landscape. Compared to soil production rates from previous work using 10 Be and 26 Al measurements from two different sites, these results show remarkable agreement and specifically quantify a soil production function for the region where soil production rates decline exponentially with increasing soil thickness, with an intercept of 53 m/m.y. and a slope of –0.02. Erosion rates determined from 10 Be concentrations from outcropping tors, bedrock, and saprolite from a main spur ridge perpendicular to the escarpment, and sediments from first- and zero-order catchments draining the main ridges, show a clear

DATING of coral reef terraces can provide a record of changes in sea level, which should be pronounced during the transition between glacial and interglacial periods. Cores drilled from coral reefs at Barbados in the equatorial west Atlantic1have revealed the sea-level changes that occurred during the Younger Dry as event at the end of the Last Glacial Maximum (∼11,000 yr BP). It has not been known, however, whether Pacific coral reefs can grow at a rate sufficient to keep up with the rise in sea level during such a transition. Here we report results obtained from a 52-m drill core from the post-glacial reef at Huon Peninsula, Papua New Guinea, spanning the interval from 7,000 to 11,000 14C yr BP, which show that coral growth kept pace while the relative sea level rose by 50m. Although the tectonic environment is very different from that at Barbados, the two records compare well when corrections are made for local tectonic uplift, showing that sea-level rise was similar at both locations. The rate of rise was greatest between 9,000 and 10,00014C yr BP, corresponding to the time of the Younger Dryas.

Uplifted coral terraces at Huon Peninsula, Papua New Guinea, preserve a record of sea level, sea-surface temperature, and salinity from the penultimate deglaciation. Remnants have been found of a shallow-water reef that formed during a pause, similar to the Younger Dryas, in the penultimate deglaciation at 130,000 ± 2000 years ago, when sea level was 60 to 80 meters lower than it is today. Porites coral, which grew during this period, has oxygen isotopic values and strontium/calcium ratios that indicate that sea-surface temperatures were much cooler (22° ± 2°C) than either Last Interglacial or present-day tropical temperatures (29° ± 1°C). These observations provide further evidence for a major cooling of the equatorial western Pacific followed by an extremely rapid rise in sea level during the latter stages of Termination II.

Switches in the glacial-interglacial cycle are linked to climatic change, which is one reason for wanting to know more about them. The history of such switches can in part be traced by estimating sea level, which reflects the amount of water locked up in ice sheets. A study that uses the Red Sea as a gauge now takes estimates of lows in sea level back 500,000 years, well beyond the last glacial period.

The frequency and intensity of disturbance on living coral reefs have been accelerating for the past few decades, resulting in a changed seascape. What is unclear but vital for management is whether this acceleration is natural or coincident only with recent human impact. We surveyed nine uplifted early to mid-Holocene (11,000-3700 calendar [cal] yr B.P.) fringing and barrier reefs along similar to 27 km at the Huon Peninsula, Papua New Guinea. We found evidence for several episodes of coral mass mortality, but frequency was 16 km along the ancient coastline, occurred ca. 9100-9400 cal yr B.P., and is associated with a volcanic ash horizon. Recolonization of the reef surface and resumption of vertical reef accretion was rapid (< 100 yr), but the post-disturbance reef communities contrasted with their pre-disturbance counterparts. Assessing the frequency, nature, and long-term ecological consequences of mass-mortality events in fossil coral reefs may provide important insights to guide management of modern reefs in this time of environmental degradation and change.

Stony deserts are durable indicators of aridity but until now have not been directly dated. Using 21Ne and 10Be produced in surface rocks by cosmic rays, we show that Australian stony deserts formed 2–4 Ma, at the time when global cooling initiated the Quaternary ice ages and intensified aridity-induced major landscape changes in central Australia. This is the first direct determination of stony desert ages, using a new method for determining cosmogenic 21Ne in the presence of various neon components from other sources.

The comment by van Woesik ([ 1 ][1]) discusses other cases of reef death in the eastern Indian Ocean region during 1997 and concludes that while wildfires may have exacerbated phytoplankton blooms off the coast of Sumatra (Mentawai Islands), upwelling related to the Indian Ocean Dipole (IOD)—

The functional dependence of bedrock conversion to soil on the overlying soil depth (the soil production function) has been widely recognized as essential to understanding landscape evolution, but was quantified only recently. Here we report soil production rates for the first time at the base of a retreating escarpment, on the soil-mantled hilly slopes in the upper Bega Valley, southeastern Australia. Concentrations of 10 Be and 26 Al in bedrock from the base of the soil column show that soil production rates decline exponentially with increasing soil depth. These data define a soil production function with a maximum soil production rate of 53 m/m.y. under no soil mantle and a minimum of 7 m/m.y. under 100 cm of soil, thus constraining landscape evolution rates subsequent to escarpment retreat. The form of this function is supported by an inverse linear relationship between topographic curvature and soil depth that also suggests that simple creep does not adequately characterize the hillslope processes. Spatial variation of soil production shows a landscape out of dynamic equilibrium, possibly in response to the propagation of the escarpment through the field area within the past few million years. In addition, we present a method that tests the assumption of locally constant soil depth and lowering rates using concentrations of 10 Be and 26 Al on the surfaces of emergent tors.

Emerged coral reef terraces on the Huon Peninsula in New Guinea were reported in a reconnaissance dating study by Veeh and Chappell 1970. Age definition achieved was not good for several important terraces, and we report here a series of new 230Th/234U dates, which further clarify the history of late Quaternary eustatic sea level fluctuations. More than 20 reef complexes are present, ranging well beyond 250,000 yr old: we are concerned with the seven lowest complexes. Major reef-building episodes dated by 30Th/234U are reef complex I at 5–9 ka (kilo anno = 1000 yr), r.c. IIIb at 41 ka (four dates), r.c. IV at 61 ka (four dates), r.c. V at 85 ka (two dates), r.c. VI at 107 ka (two dates), and r.c. VII at 118–142 ka. Complex II was previously dated by 14C at 29 ka: this age has not yet been confirmed, and may be only a lower limit. The reef crests were built during or immediately before intervals of sea level maxima, when rates of rising sea level and tectonic uplift briefly coincided. The culmination of each reef-building episode was only a few thousand years in duration, and multiple dates from the same reef complex generally group within the statistical errors of the individual dates.Several methods can be used to estimate the altitude of each sea level maximum relative to present sea level. The least complicated is to calculate mean tectonic uplift rate for each profile of the terraces, and use the mean rate to calculate the tectonic displacement of each dated reef complex on that profile. The difference between the present altitude of a reef complex and its calculated tectonic uplift gives the paleosea level at the time the reef grew. We estimate uplift rates for six surveyed sections by calibrating against published paleosea level estimates from Barbados and elsewhere, viz 125 ka, paleosea at +6 m; 103 ka, −15 m; 82 ka, −13 m. For each section the individual uplift rates for reefs V, VI, and VIIb are within 5% of their section means. Using the mean rates. paleosea level estimates for reef crests II, IIIB, and IV are made for each section. Consistency of estimates between sections is good, giving −28 m for the 60 ka paleosea level, around −38 m for the 42 ka level and −41 m for the 28 ka level (if the age is older the paleosea level would be lower. Using the mean uplift rates, the 82 ka and 103 ka paleosea levels are also estimated for each section: all individual estimates are plotted graphically, and a sea level curve drawn. The reef stratigraphy indicates sea level lowerings between each dated reef crest: the crests probably represent the interstadials of the Wisconsin (Würm, Weichsel) Glaciation, and intervening lower levels correspond to stadials. Since the last time of eustatic sea level higher than the present (about 125 ka), five sea level maxima occurred at roughly 20-ka intervals, none being as high as the present.

Coral reef terraces consisting of an upward succession of fringing and barrier reef types are preserved on land along the raising coast of Huon Peninsula, New Guinea. Seven coral reef units, I–VII from the coast landwards spaced at about 20-kyr interval, provide a rather complete record of sea levels, oxygen isotopes and temperature of teh surface tropical ocean over the duration of the high sea level events of the last glacial cycle. The systematic sampling of the coral reefs was restricted to the massive Tridacna gigas species (the giant clam) which are about 0.7%0 enriched in 18O relative to calcite equilibrium value and the presence of symbionts seems to have no effect on the δ18O of their aragonitic skeleton. Sea levels rose rapidly to 5–6.5 m above present during the last interglacial which on stratigraphic grounds and δ18O compositions can be separated into an early phase whose δ18O is similar to modern (reef VIIa dated at 138 kyr) and a late phase that is 0.6%0 heavier (reef VIIb dated at 118 kyr). Fringing reefs VI, V, IV, III and II record short-lived high sea levels stands at 107, 85, 60, 45 to 40 and 31 kyr respectively that stopped short of the present sea level (−12, −19, −28, −32 to −42 m). The general decline of the high sea level stands corresponds to an average 18O enrichment of 1.7%0 measured between the early interstadials of isotope stage 5 (VI and V) and the late interstadials of isotope stage 3 (IV, IIIa, IIIb, and II). The maximum 18O enrichment of 1.7%0 relative to present was registered during a regressive phase of reef IIIa. Stratigraphic evidence indicates that sea level lows in the range of −37 to −55 m below present have occurred in between the high sea level stands but those are unlikely to represent the lowest sea levels of the last glacial cycle. Sea levels and δ18O ddta pair offer important clues regarding the partitioning of ice volume and temperature in δ18O records. The results suggest that the ice volume effect accounts for the observed δ18O changes during the early interstadial culminations but only for 29–46% during the late interstadial culminations. Sea water cooling beyond the threshold limit of coral reef growth (∼ 18°C) did not occur during the last glacial cycle. The temperature of the surface tropical ocean was similar to present during the early interstadials of isotope stage 5 and cooler by 3°C below present during the later interstadials of isotope stage 3.

Sealevel changes over the past 250,000 yr, determined from dated coral reefs on rising islands, are synchronous with precessional changes in solar radiation when orbital eccentricity is large. Oxygen isotope variations in deep-sea cores correspond in timing but not in amplitude, reflecting changing isotopic composition of Pleistocene icecaps.

The South Alligator River, Northern Territory of Australia, has a macrotidal estuary. Tidal influence (spring tidal range 5–6 m at the mouth) extends 105 km up the channel. It is dominated by freshwater in the wet season (December-April) with a salt wedge near the mouth, but is well mixed and becomes saline throughout the dry season. The tidal channel can be divided into four different channel types: an estuarine funnel, a sinuous meandering segment, a cuspate meandering segment (in which the inside of bends are pointed) and an upstream tidal channel. The distribution of morphologically defined land classes and morphological units within each land class on the floodplain flanking the estuary differs from one channel type to another. Several stratigraphic and morphostratigraphic units have been recognized from drill holes on the coastal and deltaic-estuarine plains, and a model of development is proposed on the basis of extensive radiocarbon chronology and palynology. The coastal plain has prograded with most rapid sedimentation between 5000 and 3000 yr BP. A similar pattern of progradation is identified in the estuarine funnel. In the sinuous segment of the estuary the channel has migrated laterally across the floodplain. Previous channel positions are indicated by palaeochannels and the meander tract is occupied by laminated channel sediments. Within the cuspate segment there are numerous sinuous palaeochannels on the plains. In the upstream segment, the channel and palaeochannels have long straight reaches with irregular bends and discontinuous levees, and channel avulsion is indicated. Mangrove mud is a widespread stratigraphic unit throughout the plains. The initial phase of development is a transgressive phase. 8000–6800 yr BP, when mangrove forests extended landwards into a pre-existing valley as sea-level rose. As sea-level stabilized, the transgressive phase was followed by a widespread mangrove phase, termed the ‘big swamp’ 6800–5300 yr BP. The mangrove forests disappeared from most of the plains as vertical accretion continued, and were replaced by grass and sedge-covered floodplains. During the sinuous phase, about 5300–2500 yr BP, the channel migrated laterally and eroded the deltaicestuarine plain and deposited lateral accretion deposits (laminated channel sediments). Part of the channel of the South Alligator River has then progressed from sinuous to cuspate in form, and erosion of river banks has occurred. Transgressive and big swamp phases occurred under rising and stabilizing sea-level, respectively. Later morphodynamic channel adjustments occurred under conditions of stable sea-level. The depositional model has direct application to other estuaries in northern Australia, and may be applied to other areas where sea-level change has been similar.

Swadling et al. (1988) suggest that an inland sea, open to the north, was formed in the Sepik-Ramu lowlands by rising sea level in the late Pleistocene. This sea was infilled during the later Holocene. They also imply that these geographical changes would have significantly affected the human and ecologie prehistory of the region, although without documenting any detail. We report here evidence for existence of the inland sea and for its history of recession towards the current Sepik River mouth. We also report evidence of unusually early use of pottery at sites on former shorelines east of the lower Ramu River and of a 5000-year-old Tridacna shell adze found in alluvium about 300 km from the Sepik mouth.

Previous studies of Holocene sea level in north Queensland have been based on a variety of sea-level related deposits, with interpretations of mid-Holocene levels ranging from near zero up to 4 m above Present. We report data from 8 carefully controlled new sites, located from the Palm Group northwards and extending into the Gulf of Carpentaria. Field evidence is from coral reef flats and prograded coastal deposits. The highest sea level on inner shelf islands and mainland coast occurred 5500 yrs ago. The results of this and of certain previous studies are reduced to a common reference surface, or isobase, which varies systematically in elevation across the region. The isobase correlates strongly with predicted hydro-isostatic deformation caused by postglacial sea-level rise.

Lady Elliot Island (Great Barrier Reef) is a coral cay constructed from near-concentric shingle ridges composed of coral, Tridacna shells and bioclastic sand bound by guano-derived phosphatic cement. Radiocarbon dating of individual Tridacna samples is used to investigate the evolution of the island and place constraints on the variations in storminess of the last 3200 years. A shallow platform reef existed at Lady Elliot since at least 6500 yrs B.P. Shortly before 3200 yrs B.P. island growth by shingle progradation commenced and proceeded to the present at a rather uniform lateral rate (60 m ka −1 to leeward and 90 m ka −1 windward). Phosphatization may have occurred continuously or episodically at any time throughout the last 3200 years. The youngest cemented ridge top is less than 770 years, whereas at the base of uncemented ridges, phosphatization is a modern (∼1950 A.D.) process. Similarly, low-level beach rock has continued to form within the last 25 years. A significant period of instability occurred less than 500 years ago during which marine erosion truncated the southwestern extremity of Lady Elliot Island, formed an eastern “spit” and stranded high-level beach rock. However, the radiocarbon ages of Tridacna samples from shingle ridges from Lady Elliot and Curacoa Islands are uniformly distributed (Kolmogorov-Smirnov test) throughout the last 4000 years. We interpret this to indicate no significant variation in ridge-forming processes or fluctuation in average storminess during this interval.