Search

There are still many uncertainties about the climatic forcing that drove the glacier fluctuations of the Patagonian Ice Sheet (PIS, 38-55 degrees S) during the last glacial period. A key source of uncertainty is the asynchrony of ice lobe fluctuations between the northern, central, and southern PIS. To fully understand the regional trends requires careful mapping and extensive geochronological studies. This paper presents geomorphological and geochronological reconstructions of the glacial and deglacial landforms formed during the last glacial period at the Seno Skyring lobe, southernmost Patagonia (52 degrees S, 71 degrees W). We present a detailed geomorphological map, where we identify two moraine systems. The outer and older is named Laguna Blanca (LB) and the inner Rio Verde (RV). The LB moraines were built subaerially, whereas parts of the RV were deposited subaqueously under the palaeo lake Laguna Blanca, which developed during deglaciation. We conducted surface exposure Be-10 dating methods on boulder samples collected from LB and RV glacial margins. The moraine LB III and LB IV formed at 26.3 +/- 2.3 ka (n = 5) and 24.3 +/- 0.9 ka (n = 3), respectively. For the inner RV moraine, we obtained an age of 18.7 +/- 1.5 ka (n = 6). For the palaeo Laguna Blanca evolution, we performed Be-10 exposure ages on shoreline berms and optically stimulated luminesce dating to constrain the lake levels, and Be-10 depth profile dating on an outwash deposit formed by a partial lake drainage event, which occurred at 22 +/- 3 ka. For the RV moraine deglaciation, we performed radiocarbon dating of basal sediments in a peat bog, which indicates that the glacier retreated from the terminal RV moraine by at least c. 16.4 cal kyr BP. Our moraine geochronology shows an asynchrony in the maximum extents and a different pattern of ice advances between neighbouring lobes in southern Patagonia. We speculate that this may be due, at least in part, to the interaction between topography and

Improving our knowledge of soil formation is critical so that we can better understand the first-order controls on soil thickness and more effectively inform land-management decisions. Cosmogenic radionuclide analysis has allowed soil scientists to more accurately constrain the rates at which soils form from bedrock. In such analysis, the concentration of an isotope, such as Beryllium-10, is measured from a sample of bedrock. Because this concentration is partly governed by the lowering of the bedrock-soil interface, a cosmogenic depth-profile model can be fitted to infer the bedrock and surface lowering rates compatible with the measured concentrations. Given that the bedrock-soil interface is shielded by soil, the cosmic rays responsible for the in-situ production of the radionuclide are attenuated, with attenuation rates dependent on the density profile of this soil. Many studies have assumed that soil bulk density is either equal to that of the bedrock or constant with depth. The failure to acknowledge the variations in soil bulk density means that cosmogenically derived soil formation rates previously published may be under- or overestimates. Here, we deploy a new model called "CoSOILcal" to a global compilation of cosmogenic analyses of soil formation and, by making use of estimated bulk density profiles, recalculate rates of soil formation to assess the sensitivity to this important parameter. We found that where a soil mantle >0.25 m overlies the soil-bedrock interface, accounting for the soil bulk density profile brings about a significantly slower rate of soil formation than that previously published. Moreover, the impact of using bulk density profiles on cosmogenically derived soil formation rates increases as soil thickens. These findings call into question the accuracy of our existing soil formation knowledge and we suggest that future cosmogenic radionuclide analysis must consider the bulk density profile of the overlying soil. Highlights The effect of

Unanswered questions about the glacier and climate history preceding the global Last Glacial Maximum (LGM) in the southern temperate latitudes remain. The Marine Isotope Stage (MIS) 3 is normally understood as a global interstadial period; nonetheless its climate was punctuated by conspicuous variability, and its signature has not been resolved beyond the polar realms. In this paper, we compile a Be-10 depth profile, single grain infrared (IR) stimulated luminescence dating and C-14 samples to derive a new glacier record for the principal outwash plain complex, deposited by the western Patagonian Ice Sheet (PIS) during the last glacial period (Llanquihue Glaciation) on the Isla de Chiloe, southern Chile (42 degrees S). In this region, the Golfo de Corcovado Ice Lobe left a distinct geomorphic and stratigraphic imprint, suitable for reconstructing former ice dynamics and timing of past climate change. Our data indicate that maximum glaciation occurred by 57.8 +/- 4.7 ka without reaching the Pacific Ocean coast. Ice readavanced and buttressed against the eastern side of the Cordillera de la Costa again by 26.0 +/- 2.9 ka. Our data further support the notion of a large ice extent during parts of the MIS 3 in Patagonia and New Zealand but appear to contradict near contemporaneous interstadial evidence in the southern midlatitudes, including Chiloe. We propose that the PIS expanded to its full-glacial Llanquihue moraines, recording a rapid response of southern mountain glaciers to the millennial-scale climate stadials that punctuated the MIS 3 at the poles and elsewhere.

Soils deliver multiple ecosystem services and their long-term sustainability is fundamentally controlled by the rates at which they form and erode. Our knowledge and understanding of soil formation is not commensurate with that of soil erosion, in part due to the difficulty of measuring the former. However, developments in cosmogenic radionuclide accumulation models have enabled soil scientists to more accurately constrain the rates at which soils form from bedrock. To date, all three major rock types – igneous, sedimentary and metamorphic lithologies – have been examined in such work. Soil formation rates have been measured and compared between these rock types, but the impact of rock characteristics on soil formation rates, such as rock matrices and mineralogy, have seldom been explored. In this UK-based study, we used cosmogenic radionuclide analysis to investigate whether the lithological variability of sandstone governs pedogenesis. Soil formation rates were measured on two arable hillslopes at Woburn and Hilton, which are underlain by different types of arenite sandstone. Rates were faster at Woburn, and we suggest that this is due to the fact that the Woburn sandstone formation is less cemented that that at Hilton. Similarly, rates at Woburn and Hilton were found to be faster than those measured at two other sandstone-based sites in the UK, and faster than those compiled in a global inventory of cosmogenic studies on sandstone-based soils. We suggest that the cementing agents present in matrix-abundant wackes studied previously may afford these sandstones greater structural integrity and resistance to weathering. This work points to the importance of factoring bedrock matrices into our understanding of soil formation rates, and the biogeochemical cycles these underpin.

To understand the rates at which soils form from bedrock, it is important to know the rates at which the bedrock surface lowers (the apparent erosion rate, which is assumed to be constant). Previous models that calculate apparent erosion rates using measured concentrations of cosmogenic radionuclides rely on the assumption that the bulk density of the soil which forms as a product of bedrock erosion either equals that of the bedrock itself or is constant with depth down the soil profile. This assumption fails to recognise that soils have significantly lower densities that might not be constant with depth. The model presented here allows for the calculation of isotopically-derived soil production rates, considering the bulk density profile of the soil overlying the bedrock surface. This calculator, which can be run both in MATLAB® and GNU Octave©, represents a novel and significant contribution to the derivation of soil production rates.

Arable soils are critical resources that support multiple ecosystem services. They are frequently threatened, however, by accelerated erosion. Subsequently, policy to ensure their long-term security is an urgent societal priority. Although their long-term security relies upon a balance between the rates of soil loss and formation, there have been few investigations of the formation rates of soils supporting arable agriculture. This paper addresses this knowledge gap by presenting the first isotopically constrained soil formation rates for an arable (Nottinghamshire, UK) and coniferous woodland hillslope (Shropshire, UK). Rates ranged from 0.026 to 0.096 mm yr−1 across the two sites. These rates fall within the range of previously published rates for soils in temperate climates and on sandstone lithologies but significantly differed from those measured in the only other UK-based study. We suggest this is due to the parent material at our sites being more susceptible to weathering. Furthermore, soil formation rates were found to be greatest for aeolian-derived sandstone when compared with fluvially derived lithology raising questions about the extent to which the petrographic composition of the parent material governs rates of soil formation. On the hillslope currently supporting arable agriculture, we utilized cosmogenically derived rates of soil formation and erosion in a first-order lifespan model and found, in a worst-case scenario, that the backslope A horizon could be eroded in 138 years with bedrock exposure occurring in 212 years under the current management regime. These findings represent the first quantitative estimate of cultivated soil lifespans in the UK.

The St Kilda archipelago lies similar to 65 km west of the Outer Hebrides and similar to 60 km east of the Atlantic shelf break, and represents a key site for testing the assertion that during the Last Local Glacial Maximum (LLGM; c. 27 ka) the British-Irish Ice Sheet (BIIS) extended to near the shelf edge in all sectors. Two consistent cosmogenic Cl-36 exposure ages averaging (>=) 81.6 +/- 7.8 kafor perched bouldersat 290 maltitudedemonstrate that the last ice sheet failed tooverrun high ground on the largest island, Hirta. Cl-36 and Be-10 exposure ages for glacially emplaced boulders on low ground indicate deposition by small, locally nourished glaciers that last occupied a north-facing valley (Gleann Mor) at c. 30.9 +/- 3.2 ka, prior to extension of the last ice sheet to the outer shelf, and a south-facing valley (Village Bay) at c. 19.2 +/- 2.3 ka, several millennia after the LLGM. Our dating evidence is consistent with previous interpretations of lithostratigraphical, seismostratigraphical and geomorphological evidence and confirms that the last ice sheet failed to encroach on St Kilda. A simple ice-flow model demonstrates that even if thin, low-gradient ice lobes encircled the archipelago during theLLGM, the ice margin can only have reached the outermost moraine banks, similar to 40 kmwest of St Kilda, under extremely low (<2 kPa) driving stresses, implying either surge-like transient streaming behaviour at the ice-sheet margin or that the moraine banks relate to an earlier, more extensive ice sheet. The final glaciation of the Village Bay area at c. 19.2 +/- 2.3 kawas out of phase with the behaviour of the BIIS, which was undergoing net retreat during this period.

Southern South America contains a glacial geomorphological record that spans the past million years and has the potential to provide palaeoclimate information for several glacial periods in Earth's history. In central Patagonia, two major outlet glaciers of the former Patagonian Ice Sheet carved deep basins similar to 50 km wide and extending over 100 km into the Andean plain east of the mountain front. A succession of nested glacial moraines offers the possibility of determining when the ice lobes advanced and whether such advances occurred synchronously. The existing chronology, which was obtained using different methods in each valley, indicates the penultimate moraines differ in age by a full glacial cycle. Here, we test this hypothesis further using a uniform methodology that combines cosmogenic nuclide ages from moraine boulders, moraine cobbles and outwash cobbles. Be-10 concentrations in eighteen outwash cobbles from the Moreno outwash terrace in the Lago Buenos Aires valley yield surface exposure ages of 169-269 ka. We find Be-10 inheritance is low and therefore use the oldest surface cobbles to date the deposit at 260-270 ka, which is indistinguishable from the age obtained in the neighbouring Lago Pueyrredon valley. This suggests a regionally significant glaciation during Marine Isotope Stage 8, and broad interhemispheric synchrony of glacial maxima during the mid to late Pleistocene. Finally, we find the dated outwash terrace is 70-100 ka older than the associated moraines. On the basis of geomorphological observations, we suggest this difference can be explained by exhumation of moraine boulders. (C) 2017 Elsevier Ltd. All rights reserved.