Ulrich Radtke

A revised morpho- and chronostratigraphy of the Late and Middle Pleistocene coral reef terraces on Southern Barbados (West Indies)

The Barbados coral reef terraces are one of the few type localities worldwide that provide insights into interglacial sea level change during the Late and Middle Pleistocene. Several sea level estimates have been established since the late 1960s and each has contributed to the “Barbados Model” of sea level change. This paper presents new morpho- and chonostratigraphic data from the Barbados coral reef terraces developed over the last 12 years. The work is based on significant advances in Electron Spin Resonance (ESR)-dating of fossil coral, air photo interpretation, and a greatly improved geomorphic map of preserved fossil beach formations and reef terraces above present sea level. The need for a revision of past published morpho- and chronostratigraphies is best demonstrated on the southern part of the island. The morpho- and chronostratigraphic sequence in this region is both more complex and diverse than has been assumed so far. The revised morphostratigraphy presented here includes a differentiation of coral reef terraces, wave-cut platforms and other erosive features, such as notches and cliffs. Our study of these geomorphic features, combined with new numeric dating results (ESR, U/Th), enable a revised estimate of the spatial and temporal variation in tectonic uplift rate within south Barbados. These new rates are an essential requirement for more precise glacio-eustatic sea level reconstructions during the Late and Middle Pleistocene from this region.

A one million-year-long sequence of marine terraces on Sumba Island, Indonesia

Abstract An exceptional sequence of raised coral reef terraces is visible at Cape Laundi, on the north coast of the island of Sumba, in the Savu Sea, in which at least eleven terraces wider than 100 m, six of them even being over half a kilometer wide, are geomorphologically distinguishable between an ancient patch reef now at +475 m and the present sea level. ESR date estimations of almost unrecrystallized corals, most of them in situ, make possible a tentative identification of the terraces corresponding to isotope Stages 15 (ca. 600 ka ago) and 9 (ca. 330 ka ago). When the uplift trend deduced from the present altitude of dated terraces (0.49 ± 0.01 mm/yr) is extrapolated to the whole raised section, most geomorphological features appear to correspond to interglacial stages of the astronomically calibrated oxygen-isotope record, up to stage 27 (ca. 0.99 Ma), beyond the Matuyama/Jaramillo geomagnetic boundary. The upper terraces would in this case be twice as old as previously estimated (ca. 0.5 Ma). Main terraces seem to correspond to high sea levels at certain interglacial stages and substages, but in some cases the action of low sea levels at glacial stages cannot be ruled out. For terraces between 30 and 70 m in altitude, Th/U and ESR dates reveal that aragonitic corals at similar elevations may differ in age as much as 100 ka; this can be explained by taking into account a polycyclic origin for several terraces, reconstructions available of changes in the eustatic sea level during the last 250 kyr, and the fact that coral reef development may have been scattered at some periods only, without totally covering subjacent, older formations. The Sumba sequence of marine terraces at Cape Laundi is among the longest and most complete for the past million years reported in the literature. Its relatively good state of preservation will probably give impetus to a number of new studies on the chemical, physical, isotopic, morphological, palaeoclimatic, pedological, ecological and diagenetic changes which have occurred during the past one million years.

Correlation and dating of Quaternary littoral zones along the Patagonian coast, Argentina

Six areas of the Patagonian coast, Argentina, were investigated in order to determine the number, characteristics, geomorphology and development of Quaternary littoral zones. It became apparent that utilizing dl ratios of aspartic acid and leucine of various molluscan species would be the most useful in correlation, relative age dating and estimating the ages of these zones. The oldest littoral zone is found at elevations between 24 and 41 m above mean sea level and is judged to be older than Oxygen Isotope Substage 5e in age based on relatively high amino acid ratios and extrapolated from non-linear kinetic models. An intermediate aged littoral zone may be present at some locations based upon beach ridges or platforms varying in elevation between 16 and 28 m above mean sea level. dl ratios are generally lower than those for the oldest zone but show a greater variation. This zone may represent the Substage 5e sea level stand. Well defined young beach ridges 8–12 m above mean sea level are found in most locations and have been 14C dated, and verified by amino acid ratios, as being Holocene. The presence of Quaternary aged emerged littoral zones at roughly the same elevation suggest that the glacio-eustatic contribution is the primary cause of the high sea level stands whereas secondary variations are attributed to other factors.

Quaternary Raised Coral-Reef Terraces on Sumba Island, Indonesia

A spectacular sequence of coral-reef terraces (six steps broader than 500 meters and many minor substeps) is developed near Cape Laundi, Sumba Island, between an ancient patch reef 475 meters high and sea level. Several raised reefs have been dated with the electron spin resonance and the uranium-series dating methods. The uplift trend deduced from these reefs is 0.5 millimeter per year; most terraces, although polycyclic in origin, appear to correspond to specific interglacial stages, with the oldest terrace formed 1 million years ago. This puts them among the longest and most complete mid-Quaternary terrace sequences.

Electron spin resonance dating of the Pleistocene coral reef tracts of Barbados

Abstract The reef tracts of Barbados have been investigated by electron spin resonance dating; all parameters necessary for ESR dating (including effective α-efficiency and thermal stability) were determined without cross checking with another independent dating method. In addition, some U-series analyses were carried out in order to test the reliability of ESR. In most cases, the results show a satisfactory agreement between ESR and both these and previously published U-series dates ( M. L. Bender, R. G. Fairbanks, F. W. Taylor, R. K. Matthews, J. G. Goddard, and W. S. Broecker (1979) . Geological Society of America. Bulletin 90, 577–594). For the oldest samples, ESR dates tend to exceed He/U dates, suggesting that there might have been He loss from aragonite. Raised reef tracts are assigned to high sea stands from successive interglacial stages 5, 7, 9, 11, 13, and 15. ESR and U-series (230Th/234U) dates for corals and molluses in North Barbados do not support the suggestion of N. P. James ((1971) . Unpublished Thesis, McGill University, Montreal) that this region has subsided during part of the past 125,000 yr. Whereas the experimentally determined mean life (τ) of trapped electrons is only 500,000 yr for the ambient temperature in Barbados, there is no evidence from the comparison between ESR and other dates for thermal fading. This emphasizes the difficulty of experimental measurement of τ.

ESR dating stratigraphically well-constrained marine terraces along the Patagonian Atlantic coast (Argentina)

Abstract ESR dating of marine mollusc shells is an important tool in coastal geomorphology as well as in palaeosea-level and other research. Some severe problems, however, remain as far as the accuracy of the dating results are concerned. Case studies of Patagonian Middle and Younger Quaternary marine terraces help to illustrate the potential of an improved ESR dating technique. It can be demonstrated that a discrimination of Holocene, Last Interglacial, Penultimate Interglacial and “older” samples is possible. A chronostratigraphic separation of the well-known Last Interglacial substages 5a, 5c, and 5e is still missing, partly due to the misleading accuracy of ESR dating results. Consequently, it cannot be decided yet whether the morphologically well-conserved Last and Penultimate Interglacial age substages result from a relative sea-level fall (glacioeustatic and/or decreasing wind intensities) during the one and the same interglacial highstand shortly after the 5e or 7 transgression maximum or whether they were deposited during the much younger sea-level sub-highstands, e.g. 5c, 5a or 7c, or 7a.

ESR dating of corals

Abstract The ages of 24 coral samples from Barbados and the New Hebrides were determined simultaneously by ESR and another independent dating method (α- and MS-U-series and 14 C). All parameters for ESR dating were determined, without cross-calibration, with another dating method (including thermal stability and effective α-efficiency). The results show a relatively good concordance. Furthermore, a systematic ESR dating study on the reef tracts of Barbados (83 samples) revealed satisfactory agreement between ESR and U-series data whereas previously published He U data seemed to show open system behaviour. The thermal stability of the ESR signal at g = 2.0007 (±0.0002) estimated to be in the range of 500 ka does not seem to have an apparent influence on the ESR results.

Electron spin resonance (ESR) techniques applied to mollusc shells from South America (Chile, Argentina) and implications for palaeo sea-level curve

ESR dating of marine mollusc shells gives no evidence for a so-called Mid-Wisconsin sea-level high stand in Argentina and Chile. The idea of this high stand around 30,000–40,000 BP is still alive, especially in Argentina, but the evidence for it is, in our opinion and that of others, solely based on the misinterpretation of 14C dates. All ESR ages from southern South America belong either to the Holocene or the Last Interglacial or older Interglacial periods. Apart from the regional aspect, this paper deals with some methodological problems in ESR dating. Estimates have been improved by using only mollusc shells that are still closed and in living position and by introducing a plateau screening test for the determination of the accumulated dose (AD). The estimate of the environmental dose has been refined by analysing the infill sediments in the shells for their U, Th and K contents, separately from the embedding matrix. Special interest has been paid to the debated problem of early or continuous U uptake; the U contents of recent Holocene molluscs support the early uptake model. A probable solution to the problem of age overestimation of Holocene shells is also presented. Present-day shells are shown to possess frequently a fully developed signal and to give AD of 2–4, in one case up to 9 Gy. If the Holocene AD values of 14C-dated shells are corrected by subtracting these “recent”, and spurious, values, the resulting age estimates are more consistent with radiocarbon ages.

Geochronology and subsurface stratigraphy of Pukapuka and Rakahanga atolls, Cook Islands: Late Quaternary reef growth and sea level history

Eustatic sea-level cycles superposed on thermal subsidence of an atoll produce layers of high sea-level reefs separated by erosional unconformities. Coral samples from these reefs from cores drilled to 50 m beneath the lagoons of Pukapuka and Rakahanga atolls, northern Cook Islands give electron spin resonance (ESR) and U-series ages ranging from the Holocene to 600,000 yr B.P. Subgroups of these ages and the stratigraphic position of their bounding unconformities define at least 5 periods of reef growth and high sea-level (0–9000 yr B.P., 125,000–180,000 yr B.P., 180,000–230,000 yr B.P., 300,000–460,000 yr B.P., 460,000–650,000 yr B.P.). Only two ages fall within error of the last interglacial high sea-level stand (∼125,000–135,000 yr B.P.). This paucity of ages may result from extensive erosion of the last intergracial reef. In addition, post-depositional isotope exchange may have altered the time ages of three coral samples to apparent ages that fall within glacial stage 6. For the record to be preserved, vertical accretion during rising sea-level must compensate for surface lowering from erosion during sea-level lowstands and subsidence of the atoll; erosion rates (6–63 cm/1000 yr) can therefore be calculated from reef accretion rates (100–400 cm/1000 yr), subsidence rates (2–6 cm/1000 yr), and the duration of island submergence (8–15% of the last 600,000 yr). The stratigraphy of coral ages indicates island subsidence rates of 4.5 ± 2.8 cm/1000 yr for both islands. A model of reef growth and erosion based on the stratigraphy of the Cook Islands atolls suggests average subsidence and erosion rates of between 3–6 and 15–20 cm/1000 yr, respectively.

Quaternary uplifted coral reef terraces on Alor Island, East Indonesia

A flight of six major coral reef terraces, up to 700 m in altitude, occurs along the eastern and northern sides of Kabola Peninsula, Alor Island, Indonesia. Some radiometric dates have been obtained from unrecrystallized coral samples collected in growth position by three different methods (14C, 230Th/234U, ESR). This enabled the identification of the terraces corresponding to the Holocene and to oxygen-isotope stages 5c, 5e and 7. According to the present elevation of the dated terraces, a 1.0–1.2 mm/y mean rate of uplift can be discerned. Extrapolation of this trend to the whole sequence of terraces reveals a good correlation between the development of major terraces and interglacial or interstadial stages corresponding to astronomically calibrated oxygen isotope records, up to stage 13. The relatively rapid uplift rate in this region minimized the possibility of polycyclic sea-level stands at the same levels and contributed to the good preservation of some morphological reef features. Two superimposed marine notches are visible near the present shoreline, with retreat points at about 5.0 m and 8.6 m respectively above the present MLWST level. They can be interpreted as corresponding to a glacial interstadial (the upper notch) and to the Holocene sea-level peak (the lower one). As Holocene emergence has been less than what could be expected from a 1 mm/y rate of uplift, a major coseismic vertical displacement may occur in the future.