Journal of Quaternary Science最新文献

Closed basins within the Great Basin of the western United States were home to numerous lakes during the Pleistocene. One of these paleolakes along the western edge of the Great Basin, Lake Wellington, once filled a 10 × 25-km expanse of Smith Valley to depths approaching 90 m. This and other lakes that existed during the Pleistocene are generally considered to be pluvial, indicating contemporaneity with either or all a period of cooler climate, increased rainfall and snowmelt, and relatively reduced rates of evaporation as compared to today. Here we combine the results of ³⁶Cl terrestrial cosmogenic nuclide surface exposure dating with soils and geomorphic observations to suggest Lake Wellington was not a pluvial lake but instead the result of a large landslide prior to ~43 ± 15 ka along the West Walker River where it exited Smith Valley. The observations collected also reveal an ancestral course of the West Walker River ~85 m above the current river grade. Attributing the elevation difference to incision caused by active 0.05 ± 0.01 mm a⁻¹ uplift of the underlying Singatse and assuming the ancestral course followed the same path as today places the age of the paleoriver course at ~1.7 Ma.

Lago Viedma is a proglacial lake into which the Viedma Glacier flows from the Southern Patagonian Ice Field. This glacier has rapidly lost mass in recent times, and its deglacial history is reflected in the lake's subsurface. New high-resolution multi-channel seismic profiles acquired in the northwestern sector of the lake have allowed us to reconstruct the bathymetry of the area and identify several small sub-basins, which have a maximum depth of 240 m in this sector. Four seismic facies have been recognized, separated by erosional unconformities, reflecting the depositional conditions in this sector of the basin during different Quaternary sedimentation phases. There is a transitional phase from ice-contacted deposits to subglacial deposits, probably associated with a subglacial fan, and finally a phase of lacustrine sedimentation. In addition, three depositional stages were identified within the lacustrine deposits, indicating a different sedimentary input, and the morphologies observed at the lake bottom suggest the existence of at least ten stagnations of the glacial margin. This study helps to improve the history of the retreat of the Viedma Glacier in this area during the Pleistocene/Holocene and provides a basis for further geophysical measurements aimed at mapping this remote lacustrine environment.

During the Late Holocene, the majority of lowland river systems in temperate Europe transformed from low-energy multi-channel rivers in strongly vegetated marshy floodplains to more open floodplains characterised by single-channel meandering rivers with overbank deposits. While the general framework of this transformation in floodplain geomorphology, ecology and hydrology (i.e. geoecohydrology) is widely recognised many uncertainties remain as its timing varies significantly, both among different river catchments and within them. To unravel whether the observed differences in floodplain response can be attributed to differences in the timing and nature of the driving forces or to a difference in sensitivity towards them, we compare long-term and large-scale reconstructions of the geoecohydrological floodplain dynamics and of the (climatically and anthropogenically driven) land cover change for two contrasting regions: the central Belgian loess belt and the sandy Campine region. By using a combination of cluster analysis, ordination and Ellenberg indicator scores on a large multi-proxy and multi-site dataset, we revealed the major trends in the past geoecohydrological evolution of northeastern Belgian floodplains. These trends are probably determined by changes in floodplain wetness, which can in turn be linked to variations in upland forest cover. The Early and Late Holocene floodplain transformations appear synchronous with the respective increases and decreases in upland forest cover in the vicinity of the sites, largely determining the water availability in the river catchments and thereby their local geoecohydrological conditions. Initially, these evolutions were determined by climate, but during the Middle and especially Late Holocene anthropogenic influence became a far more important factor, causing the evolutions in the two studied regions to increasingly diverge. While marshy floodplains with forested margins can still be found in the sandy Campine region today, these have become rare in the central Belgian loess belt due to the combination of a higher level of human impact and greater erodibility of the soils in this area. Despite the strong spatiotemporal variability of the floodplain transformations, we observe a trend towards increasingly rapid floodplain responses to upland land cover changes over the course of the Holocene, probably related to the growing hillslope–floodplain connectivity. We conclude that the (dis)similarities in Holocene geoecohydrological floodplain change can be largely attributed to the (dis)similarities in (climatically or anthropogenically driven) land cover change in the uplands, although the differences in inherent sensitivity of the locations – linked to factors such as soil type and topography – further complicate the already non-linear impact–response relationships.

South America is a pivotal paleontological setting for Late Quaternary megafaunal research. Nonetheless, its paleobiogeography and paleoecology, and the causes for the recent extinction of megafauna in the Late Pleistocene/Early Holocene are not yet fully understood. This study aimed to investigate the potential geographical distribution and historically intertropical stable areas (HISAs) of South American megafauna. We generated maps using Paleo-Species Distribution Models for 12 megafaunal taxa (11 herbivores and one carnivore) during interglacial and glacial periods in the Late Quaternary: 120 and 21 ka. Our models, together with isotopic data, show that the HSIAs occurred mainly in low-altitude plains (<1000 m) and were superimposed on seasonal dry forest biomes. We propose the occurrence of two HSIAs: the West Intertropical Region and the Brazilian Intertropical Region, the latter being redefined. We suggest that the Brazilian Intertropical Region is much smaller than previously proposed, but was still a key zoogeographical region for megafauna in South America during the Late Pleistocene.

The quarry of the Sverdlov (Etalon) Factory stands out as a unique site in the northwest of the East European Plain since it exposes an almost complete Upper Pleistocene stratigraphic succession. Previous investigations have revealed a sequence of marine sediments deposited during the whole Mikulino Interglacial [Eemian, Marine Isotope Stage (MIS)5e], but the chronology of the Late Pleistocene deposits remains incomplete and requires further update. We present a study of the Middle–Upper Pleistocene units exposed in the Sverdlov Factory quarry by using geological methods, optically stimulated luminescence (OSL) dating and pollen analysis. The lower part of the Sverdlov Factory section comprises glaciomarine/glaciolacustrine varves, which accumulated during degradation of the Late Moscow (Late Saalian, Late MIS6) glaciation and correspond to pollen zone M₁ (lower peak of Picea). They are covered by interglacial marine clayey silt that includes all Mikulino regional pollen zones M₂–M₈. According to a Bayesian model based on K-rich feldspar luminescence ages, the interglacial marine deposition started 133 ± 8 ka ago and ended 109 ± 7 ka ago. The marine sediments are overlain by lacustrine silt with a K-rich feldspar age of 119 ± 7 ka, lacustrine/alluvial sand with plant detritus formed 47–41 ka ago, and lacustrine sand and silt without organic matter accumulated ~39 ka ago. The upper part of the section is composed of subglacial till formed during the Late Valdai (Late Weichselian, MIS2) under the Scandinavian Ice Sheet and glaciolacustrine varved deposits of the Baltic Ice Lake. Two ages were acquired using quartz OSL from these varves, 16.9 ± 1.6 and 15.7 ± 1.0 ka, which are assumed to slightly overestimate the true age.