Relationship between meteoric 10 Be and NO − 3 concentrations in soils along Shackleton Glacier , Antarctica

Outlet glaciers that flow through the Transantarctic Mountains (TAM) experienced changes in ice thickness greater than other coastal regions of Antarctica during glacial maxima. As a result, ice-free areas that are currently exposed may have been covered by ice at various points during the Cenozoic, complicating our understanding of ecological succession in TAM soils. Our knowledge of glacial extent on small spatial scales is limited for the TAM, and studies of soil exposure duration and disturbance, in particular, are rare. We collected surface soil samples and, in some places, depth profiles every 5 cm to refusal (up to 30 cm) from 11 ice-free areas along Shackleton Glacier, a major outlet glacier of the East Antarctic Ice Sheet. We explored the relationship between meteoric 10Be and NO−3 in these soils as a tool for understanding landscape disturbance and wetting history and as exposure proxies. Concentrations of meteoric 10Be spanned more than an order of magnitude across the region (2.9× 108 to 73× 108 atoms g−1) and are among the highest measured in polar regions. The concentrations of NO−3 were similarly variable and ranged from ∼ 1 μg g −1 to 15 mg g−1. In examining differences and similarities in the concentrations of 10Be and NO−3 with depth, we suggest that much of the southern portion of the Shackleton Glacier region has likely developed under a hyper-arid climate regime with minimal disturbance. Finally, we inferred exposure time using 10Be concentrations. This analysis indicates that the soils we analyzed likely range from recent exposure (following the Last Glacial Maximum) to possibly > 6 Myr. We suggest that further testing and interrogation of meteoric 10Be and NO−3 concentrations and relationships in soils can provide important information regarding landscape development, soil evolution processes, and inferred exposure durations of surfaces in the TAM. Published by Copernicus Publications on behalf of the European Geosciences Union. 1364 M. A. Diaz et al.: Relationship between meteoric 10Be and NO 3 concentrations in soils


Introduction
One of the most intriguing questions in biogeography concerns the relationship between the evolution of terrestrial organisms and landscape disturbance (e.g., glacial overriding, soil wetting), particularly in Antarctica. Current data indicate that organism lineages have survived in some Antarctic soils for possibly millions of years, despite multiple glaciations throughout the Pleistocene (Convey et al., 2008;Fraser et al., 2012;Stevens and Hogg, 2003). It is still unclear how and where these organisms found suitable glacial refugia given the high salt concentrations in high-elevation soils (Lyons et al., 2016). The most biodiverse soils in the Ross Sea sector are at low elevations near the coast, where the Ross Ice Shelf or sea ice meets the Transantarctic Mountains (TAM) (Collins et al., 2020). These soils are also those which are most susceptible to glacial overriding during glacial maxima, though the timing of retreat and glacial extent is still unknown on local scales (Golledge et al., 2012;Mackintosh et al., 2011).
Outlet glaciers are among the most responsive cryospheric components in Antarctica, and changes in their extents over time are recorded in nearby sedimentary deposits (Golledge et al., 2013;Jones et al., 2015;Scherer et al., 2016;Spector et al., 2017). However, only scattered information exists on TAM soil processes, ages, chronosequences, and their implications for terrestrial and ecosystem history (Bockheim, 2002;Dickinson et al., 2012;Graham et al., 1997Graham et al., , 2002Lyons et al., 2016;Scarrow et al., 2014;Schiller et al., 2009). Shackleton Glacier, an outlet glacier of the East Antarctic Ice Sheet (EAIS), flows between several exposed peaks of the central Transantarctic Mountains (CTAM), and ice-free areas are present at both low and high elevations. We report concentrations of meteoric 10 Be and nitrate (NO − 3 ) in soils from 11 ice-free areas and investigate their distributions at depth to explore 10 Be and NO − 3 relationships. The sampling methodology was designed to capture a range of soils which have low salt concentrations due to recent exposure from glacial retreat following the Last Glacial Maximum (LGM) and soils that were likely exposed since at least the last glacial period. These data include some of the only meteoric 10 Be and NO − 3 concentration data from the CTAM Campbell, 1968b, 1977;Graham et al., 1997;Lyons et al., 2016), inform knowledge of landscape disturbance and wetting history, may potentially be used to infer soil exposure duration, and are useful in understanding Antarctic terrestrial biogeography.

Brief overview of Antarctic glacial and wetting history
Antarctica is believed to have maintained a persistent ice sheet since at least the Eocene epoch; paleorecords indicate that the East and West Antarctic Ice Sheets (EAIS and WAIS, respectively) have waxed and waned since at least the Miocene (Gasson et al., 2016;Gulick et al., 2017). Sediment core records collected from the Ross Sea and ice cores from the Antarctic interior indicate that the EAIS and WAIS underwent dozens of glacial and interglacial cycles throughout the Cenozoic (Augustin et al., 2004;Talarico et al., 2012). The WAIS is a marine-terminating ice sheet defined by a grounding line below sea level, which decreases the stability of the ice sheet and results in more rapid advance and retreat compared to the EAIS (Pollard and DeConto, 2009). The EAIS is grounded above sea level and is therefore generally more stable. The EAIS and WAIS were at their most recent greatest extent about 14 ka during the LGM (Clark et al., 2009). During the LGM, the EAIS expanded along its margins, and some of the greatest increases in height occurred at outlet glaciers which flow through exposed peaks of the TAM and drain into the Ross and Weddell seas (Anderson et al., 2002;Golledge et al., 2012;Mackintosh et al., 2014). As a result, many of the currently exposed TAM soils were overrun by ice during the LGM, and some may have only recently been exposed.
Much of the Antarctic continent is a polar desert, and geomorphological data from ice-free soils in the McMurdo Dry Valleys indicate that some regions have likely been hyperarid for as long as 15 Myr (Marchant et al., 1996;Valletta et al., 2015). As such, atmospherically derived constituents, including salts and metals, can accumulate in exposed Antarctic soils at concentrations similar to those from the Atacama and Namib deserts (Diaz et al., 2020;Lyons et al., 2016;Reich and Bao, 2018). Using soil NO − 3 concentrations from the Meyer Desert in the Beardmore Glacier region and NO − 3 fluxes calculated from a Dominion Range ice core, Lyons et al. (2016) estimated that at least 750 000 years have passed since the Meyer Desert had widespread soil wetting. It is likely that other high-elevation and inland locations in the TAM also have high concentrations of salts and similarly old "wetting ages", though this has not been thoroughly investigated.

Meteoric 10 Be systematics in Antarctic soils
10 Be is a cosmogenic radionuclide with a half-life of 1.39 Myr (Korschinek et al., 2010) that is produced in both the atmosphere (meteoric) and in situ in mineral grains. In the atmosphere, N and O gases are bombarded by high-energy cosmic radiation to produce meteoric 10 Be. Particle-reactive 10 BeO or 10 Be(OH) 2 is produced and removed from the atmosphere by wet and dry deposition (McHargue and Damon, 1991). At the Earth's surface, meteoric 10 Be sorbs onto clay particles and is insoluble in most natural waters of pH greater than 4 (Brown et al., 1992;You et al., 1989). The clay particles can be redistributed to lower depths in soils due to particle migration or can be transported by wind. As such, the total number of 10 Be atoms in a soil profile, its inventory, is a function of surface exposure duration, erosion, clay particle translocation, solubility, and sedimentation. If delivery rates can be determined, meteoric 10 Be can be used as a tool to determine exposure ages, erosion rates, and soil residence times (see Willenbring and Von Blanckenburg, 2010, and references within). Regarding terrestrial exposure ages, there are scattered exposure age studies from across the CTAM using a variety of in situ produced cosmogenic nuclides (Ackert and Kurz, 2004;Balter-Kennedy et al., 2020;Bromley et al., 2010;Kaplan et al., 2017;Spector et al., 2017), and previously reported exposure ages of CTAM moraines and boulders from these studies range from < 10 kyr to > 14 Myr.
The measurement of meteoric 10 Be in soil has enabled researchers to date surfaces (soils) and features in Antarctica. Previous studies have measured meteoric 10 Be in the McMurdo Dry Valleys (MDV) and Victoria Land soils and sediments to calculate exposure ages and to determine the onset of the current polar desert regime (Dickinson et al., 2012;Graham et al., 2002;Schiller et al., 2009;Valletta et al., 2015). In general, these previous studies found that highelevation, northern fringe regions along the Ross Embayment have been ice-free and possibly hyper-arid since at least the Pliocene. Few meteoric 10 Be data have been previously published from the CTAM (Graham et al., 1997), which represent ice sheet dynamics and climatic conditions closer to the Polar Plateau.

Nitrate systematics in Antarctic soils
The nitrogen cycle in Antarctica differs greatly from the nitrogen cycle in temperate regions, primarily due to scarce biomass and few vascular plants (Cary et al., 2010;Michalski et al., 2005). Nitrogen in CTAM soils primarily exists as NO − 3 and is sourced from the atmosphere, with varying contributions from the troposphere and stratosphere (Diaz et al., 2020;Lyons et al., 2016;Michalski et al., 2005). Similar to meteoric 10 Be, NO − 3 is deposited on exposed soils; however, nitrate salts are highly water-soluble. Once deposited on the surface, nitrate salts can be dissolved and transported down-gradient or eluted to depth when wetted (i.e., during ice-snow melt events). However, the hyper-arid climate of the CTAM can allow NO − 3 to accumulate at high concentrations in soils (Claridge and Campbell, 1968a;Diaz et al., 2020;Lyons et al., 2016). Soil NO − 3 concentrations have the potential to inform our knowledge of wetting history and possibly glacial history in the CTAM due to the relatively high solubility of nitrate salts, though uncertainties regarding het-erogeneous deposition and post-depositional alteration (such as re-volatilization and photolysis) require further investigation (Diaz et al., 2020;Frey et al., 2009;Graham et al., 2002).

Study sites and region
Shackleton Glacier (∼ 84.5 to 86.4 • S; ∼ 130 km long and ∼ 10 km wide) is a major outlet glacier of the EAIS that drains north into the Ross Embayment with other CTAM outlet glaciers to form the Ross Ice Shelf (RIS) (Fig. 1). The ice flows between exposed surfaces of the Queen Maud Mountains, which range from elevations of ∼ 150 m a.s.l. near the RIS to > 3500 m a.s.l. further inland. The basement geology of the Shackleton Glacier region is comprised of igneous and metamorphic rocks that formed from intruded and metamorphosed sedimentary and volcanic strata during the Ross Orogeny (450-520 Ma) (Elliot and Fanning, 2008). The southern portion of the region consists of the Devonian-Triassic Beacon Supergroup and the Jurassic Ferrar Group, while the northern portion consists of pre-Devonian granitoids and the early to mid-Cambrian Taylor Group (Elliot and Fanning, 2008;Paulsen et al., 2004). These rocks serve as primary parent material for soil formation (Claridge and Campbell, 1968b). Deposits of the Sirius Group, the center of the stable vs. dynamic EAIS debate (Barrett, 2013;Sugden et al., 1993;Webb et al., 1984;Wilson, 1995), have been previously identified in the southern portion of the Shackleton Glacier region, particularly at Roberts Massif (Fig. 2) and Bennett Platform, with a small exposure at Schroeder Hill (Hambrey et al., 2003).
The valleys and other ice-free areas within the region have been modified by the advance and retreat of Shackleton Glacier, smaller tributary glaciers, and alpine glaciers. Similar to the Beardmore Glacier region, the Shackleton Glacier region is a polar desert, which results in the high accumulation of salts in soils. The surface is comprised primarily of till, weathered primary bedrock, and scree, which ranges in size from small boulders and cobbles to sand and silt. Clay minerals have been previously identified in all samples from Roberts Massif and are likely ubiquitous throughout the region (Claridge and Campbell, 1968b). The clays are a mixture of those derived from sedimentary rocks and contemporaneous weathering (Claridge and Campbell, 1968b). Thin, boulder belt moraines, characteristic of cold-based glaciers, were deposited over bedrock and tills at Roberts Massif, while large moraines were deposited at Bennett Platform (Fig. 2;Balter-Kennedy et al., 2020;Claridge and Campbell, 1968b). Most soils appeared to be dry, though some small ponds and water tracks have been documented near Mt. Heekin and Thanksgiving Valley (Elliot et al., 1996). Additional information on the sample locations and surface features is provided in Tables 1 and 2. , and an overview map of Shackleton Glacier (c). The red box in (b) encapsulates the Shackleton Glacier region. The red circles in (c) represent our 11 sampling locations, with an emphasis on Roberts Massif (orange), Bennett Platform (green), and Thanksgiving Valley (blue), which have the most comprehensive dataset in this study. The bedrock serves as the primary weathering product for soil formation (Elliot and Fanning, 2008;Paulsen et al., 2004). For reference, the East Antarctic Ice Sheet (EAIS), West Antarctic Ice Sheet (WAIS), Ross Ice Shelf (RIS), and Dominion Range are labeled (a, b). Base maps were provided by the Polar Geospatial Center.

Sample collection
During the 2017-2018 austral summer, we visited 11 ice-free areas along Shackleton Glacier: Roberts Massif, Schroeder Hill, Bennett Platform, Mt. Augustana, Mt. Heekin, Thanksgiving Valley, Taylor Nunatak, Mt. Franke, Mt. Wasko, Nilsen Peak, and Mt. Speed (Fig. 1). These areas represent soils from near the head of the glacier to near the glacier terminus at the coast of the RIS. Two surface samples (Table 1) were collected at each location (except for Nilsen Peak and Mt. Wasko, represented by only one sample each) with a plastic scoop and stored in Whirl-Pak™ bags. One sample was collected furthest from Shackleton Glacier or other tributary glaciers (within ∼ 2000 m) to represent soils that were likely exposed during the LGM and previous recent glacial periods. A second sample was collected closer to the glacier (between ∼ 1500 and 200 m from the first sample) to represent soils likely to have been covered during the LGM and exposed by more recent ice margin retreat.
Soil pits were dug by hand at the sampling locations furthest from the glacier for Roberts Massif, Schroeder Hill, Mt. Augustana, Bennett Platform, Mt. Heekin, Thanksgiving Valley, and Mt. Franke (seven sites). Continuous samples were collected every 5 cm until refusal (up to 30 cm) and stored frozen in Whirl-Pak™ bags. All surface (21) and depth profile (25) samples were shipped frozen to The Ohio State University and kept frozen until analyzed. We selected Roberts Massif, Bennett Platform, and Thanksgiving Valley as locations for the most in-depth analysis for the depth profiles. These locations were chosen to maximize variability in landscape development: Roberts Massif represented an older, likely minimally disturbed landscape; Bennett Platform represented a landscape with evidence of recent glacial advance and retreat, as well as substantial topographic highs and lows; and Thanksgiving Valley represented a landscape with possible hydrologic activity, as evidenced by nearby ponds (Table 2).

Meteoric 10 Be analysis
A total of 30 subsamples of surface soils from all locations and the depth profiles from Roberts Massif, Bennett Platform, and Thanksgiving Valley were sieved to determine the grain size at each location. For each sample, the percentages of gravel (> 2 mm), sand (63 µm −2 mm), and silt (< 63 µm) are reported in Table S1 in the Supplement. Since there is a strong grain size dependence of meteoric 10 Be (little 10 Be is carried on coarse (> 2 mm) grains; Pavich et al., 1986), the gravel portion of the sample was not included in the meteoric 10 Be analysis. The remaining soil (< 2 mm) was ground to fine powder using a shatterbox.
Meteoric 10 Be (Tables 1 and S2) was extracted and purified at the NSF/University of Vermont (UVM) Commu-nity Cosmogenic Facility following procedures adapted from Stone (1998). First, 0.5 g of powdered soil was weighed into platinum crucibles, and 0.4 g of SPEX 9 Be carrier (with a concentration of 1000 µg mL −1 ) was added to each sample. The samples were fluxed with a mixture of potassium hydrogen fluoride and sodium sulfate. Perchloric acid was then added to remove potassium by precipitation and later evaporated. Samples were dissolved in nitric acid and precipitated as beryllium hydroxide (Be(OH) 2 ) gel, then packed into stainless-steel cathodes for accelerator mass spectrometer isotopic analysis at the Purdue Rare Isotope Measurement (PRIME) Laboratory. Isotopic ratios were normalized to primary standard 07KNSTD with an assumed ratio of 2.85 × 10 −12 (Nishiizumi et al., 2007). We corrected sample ratios with a 10 Be/ 9 Be blank ratio of 8.2 ± 1.9 × 10 −15 , which is the average standard deviation of two blanks processed alongside the samples. We subtracted the blank ratio from the sample ratios and propagated uncertainties in quadrature. The average blank ratio was about 4 orders of magnitude less than the measured sample ratios.

analysis
Separate, un-sieved subsamples of soil from all locations and depth profiles were leached at a 1 : 5 ratio of soil to deionized water for 24 h, then filtered through a 0.4 µm Nuclepore membrane filter. The leachate was analyzed on a Skalar San++ Automated Wet Chemistry Analyzer with an SA 1050 Random Access Auto-sampler (Lyons et al., 2016;Welch et al., 2010). Concentrations are reported as NO −  (Table 1) with accuracy, as determined using a USGS 2015 "round-robin" standard, and precision better than 5 % (Lyons et al., 2016).

Meteoric 10 Be inventory
We developed a mass balance using the fluxes of meteoric 10 Be to and from Shackleton Glacier region soils to understand the accumulation of 10 Be in glaciated environments (Pavich et al., 1984(Pavich et al., , 1986. The model assumes that soils that were overlain by glacial ice in the past and are now exposed accumulated less 10 Be than soils that were exposed throughout the glacial periods (Fig. 3). The concentration of meteoric 10 Be at the surface (N, atoms g −1 ) per unit of time (dt) is expressed as a function wherein the addition of 10 Be is represented as the atmospheric flux to the surface (Q, atoms cm −2 yr −1 ), and removal is due to both radioactive decay, which is represented by a disintegration constant  . Conceptual diagram of meteoric 10 Be accumulation in soils during glacial advance and retreat. In "ideal" conditions, 10 Be accumulates in exposed soils by wet deposition (with snow) and dry deposition (by gravity as indicated by black arrows), and 10 Be concentrations beneath the glacier are negligible at background levels (a). As the glacier retreats, 10 Be can begin accumulating in the recently exposed soil (b) and an inventory can be measured to calculate exposure duration. In the case in which the glacier has waxed and waned numerous times and the soils already contain a non-negligible "inheritance" concentration of 10 Be, the inventories would need to be corrected for 10 Be inheritance (c-d) to accurately determine exposure duration. Q represents of the flux of 10 Be to the surface, λ represents radioactive decay of 10 Be, E represents erosion, and I represents the migration of 10 Be from the surface to depth.
We accounted for uncertainties regarding 10 Be migration in the soil column by calculating the inventory (I , atoms cm −2 ) of the soil (Eq. 2) (Pavich et al., 1986). We used a density (ρ) of 2 g cm −3 and assumed that Q had not changed systematically over the accumulation interval. The inventory is the total sum of meteoric 10 Be atoms in the soil profile, and the change in inventory due to deposition, decay, and surface erosion is related to surface exposure duration (Eq. 3).
Meteoric 10 Be concentrations typically decrease with depth until they reach a "background" level (Graly et al., 2010). The background is identified as the point at which the concentration of meteoric 10 Be is constant with depth ( dN dz = 0). Typically, the background values can be used to calculate an initial inventory (I i , atoms cm −2 ) using Eq. (4), where N z is the 10 Be concentration (atoms g −1 ) at the bottom of the profile (z, cm). In this case, we assume that the initial concentration of meteoric 10 Be is isotropic. However, an accurate initial inventory can only be determined for soil profiles that are deep enough to capture background concentrations. This may not be the case in areas of permafrost where 10 Be is restricted to the active layer (Bierman et al., 2014).
Additionally, the initial inventory can be influenced by repeated glacier advance and retreat during glacial-interglacial cycles. For this case, the soil has "inherited" 10 Be during each subsequent exposure to the atmosphere, some of which may have been removed with eroded soil (Fig. 3c and d). For constructional landforms, such as moraines, the inheritance is equal to the background and/or initial inventory. Without information on drift sequences, it is difficult to correct the measured inventory for inheritance by distinguishing meteoric 10 Be that was deposited after the most recent ice retreat from 10 Be that was deposited during previous interglacial periods.

Depth profile composition and concentrations of meteoric 10 Be
Sediment grain size is similar among the three soil profiles collected from Roberts Massif, Bennett Platform, and Thanksgiving Valley; the soils are primarily comprised of sand-sized particles, with less silt-sized and smaller material (Fig. 4). The proportions of silt and gravel are similar at Roberts Massif, although the majority of the profile is sand-sized. Thanksgiving Valley has the coarsest material, while Bennett Platform has a more even grain size distribution. The deepest profile is from Thanksgiving Valley, while the Roberts Massif and Bennett Platform profiles are half the depth. All three profiles are ice-cemented at the bottom and are shallow compared those collected from the McMurdo Dry Valleys (Dickinson et al., 2012;Schiller et al., 2009;Valletta et al., 2015), though they are comparable to profiles collected at Roberts Massif by Graham et al. (1997). Concentrations of meteoric 10 Be for both surface and depth profile samples span more than an order of magnitude in the Shackleton Glacier region and range from 2.9 × 10 8 atoms g −1 at Mount Speed to 73 × 10 8 atoms g −1 at Roberts Massif ( Fig. 5; Table 1). At individual sites where samples were collected at two locations, concentrations are typically highest for the samples furthest from the glacier, with notable exceptions at Roberts Massif and Thanksgiving Valley (Fig. 5). This trend is expected since our sampling plan was designed to capture both recently exposed soils (near the glacier(s)) and soils which exposed throughout the LGM and possibly other glacial periods. The measured inventories (Eq. 2) vary from 0.57 × 10 11 atoms at Bennett Platform to 1.5 × 10 11 atoms at Roberts Massif (Table 3).
The meteoric 10 Be depth profiles differ between Roberts Massif, Bennett Platform, and Thanksgiving Valley. The profile from Roberts Massif has the highest overall concentrations (Fig. 6). Within the profile, the 5-10 cm sampling interval has the highest concentration, followed by the bottom of the profile, then the surface. The profile behavior for Thanksgiving Valley is similar, though the differences in concentrations within both profiles are relatively small. Bennett Plat- form is the only location where the surface concentration is the highest compared to the remainder of the profile and the concentration decreases with depth (Fig. 6). Although we sampled the entirety of the active layer where modern particle mobility throughout the soil column occurs, no depth profiles appear to decrease to the background levels needed to calculate an initial meteoric 10 Be inventory (Eq. 4). As a result, we are not able to correct the measured inventory for background 10 Be, nor are we able estimate the inherited 10 Be concentration in the soil.

Measured concentrations of NO −
3 span 4 orders of magnitude across the seven depth profiles we sampled ( Fig. 6; Table 1). The lowest concentration is from Mt. Franke of ∼ 1 µg g −1 ; the highest concentration is from Roberts Massif of 15 mg g −1 . In addition, similar to the meteoric 10 Be profiles, the NO − 3 concentrations are highest for the samples that were collected furthest from the coast and at the highest elevations (Table 1). In general, the profiles from Roberts Massif and Thanksgiving Valley are similar (Fig. 6b); 10 Be and NO − 3 concentrations are highest just below the surface 3 concentrations in soils Figure 5. Spatial distribution of surface meteoric 10 Be concentrations in the Shackleton Glacier region (a). Where possible, two samples were collected at each location to represent surfaces closest to the glacier, which might have been glaciated during recent glacial periods, and samples furthest from the glacier that are likely to have been exposed during recent glacial periods. Insets of Roberts Massif (b), Bennett Platform (c), and Thanksgiving Valley (d) are included, as these locations have both 10 Be and NO − 3 depth profile data. Base maps were provided by the Polar Geospatial Center.
in the 5-10 cm interval and are fairly consistent throughout the profile. The NO − 3 depth profile mirrors the 10 Be profile at Bennett Platform; while 10 Be concentration decreases with depth, the NO − 3 concentration increases with depth. Since we measured NO − 3 concentrations for all seven depth profiles we collected, we compare the profile concentrations and shapes from the four profiles without 10 Be depth measurements (Mt. Augustana, Schroeder Hill, Mt. Franke, and Mt. Heekin) to the Roberts Massif, Bennett Platform, and Thanksgiving Valley profiles with both measurements (Fig. 6). Most of the NO − 3 profiles do not significantly change with depth and are similar to the profile from Thanksgiving Valley, though Schroeder Hill is most similar to Roberts Massif (Fig. 6). This is unsurprising given the similar latitudes, surface features, and environmental conditions between the different locations (e.g., high-latitude hyper-arid vs. lowerlatitude with possible evidence of wetter conditions) ( Fig. 1; Table 2). No other location had large terminal moraines as observed at Bennett Platform.

Discussion
The Shackleton Glacier region soil profiles and surface samples have some of the highest meteoric 10 Be concentrations (∼ 10 9 atoms g −1 ) measured to date in Earth's polar regions (Fig. 6a). Though our profiles are shallower than profiles from the MDV and Victoria Land in Antarctica (Dickinson et al., 2012;Schiller et al., 2009;Valletta et al., 2015) and Sweden and Alaska in the Arctic (Bierman et al., 2014;Ebert et al., 2012), the soils from these previous studies reached back- ground concentrations of 10 Be within the top 40 cm, which is close to our maximum depth of 30 cm at Thanksgiving Valley. For comparison, the deepest profile collected by Graham et al. (1997) at Roberts Massif was 36 cm. The Bennett Platform soil profile is most similar to the soil profiles from other regions in Antarctica, as they have decreasing 10 Be concentrations with depth, while Thanksgiving Valley and Roberts Massif are relatively homogenous and more similar to profiles from the Arctic. The inventories from this study are also among the highest calculated for Antarctic soils. The inventories from Bennett Platform and Thanksgiving Valley are most similar (∼ 10 10 ) to inventories of saprolites and tills from Sweden (Ebert et al., 2012) and the MDV (Schiller et al., 2009), though they are higher than those measured from other high-elevation inland locations in Victoria Land (Dickinson et al., 2012;Valletta et al., 2015). Our inventory from Roberts Massif is the same as the inventory reported for a nearby location by Graham et al. (1997), and all of our inventories are within the range of values from the Arctic (Bierman et al., 2014), despite shallower profiles.
6.1 Relationships between meteoric 10 Be and NO −

and governing processes
Previous studies have proposed that atmosphere-derived salt concentrations at the surface may correlate with exposure ages and wetting ages in Antarctica (Everett, 1971;Graham et al., 1997Graham et al., , 2002Graly et al., 2018;Lyons et al., 2016;Schiller et al., 2009). Graly et al. (2018) showed that, in par-ticular, water-soluble NO − 3 and boron exhibited the strongest relationships with exposure age (R 2 = 0.9 and 0.99, respectively). Lyons et al. (2016) used NO − 3 concentrations to estimate the amount of time since the soils were last wetted, and Graham et al. (2002) attempted to calculate exposure ages using the inventory of NO − 3 in the soil. Graly et al. (2018) argue that boron is the preferable exposure proxy due to concerns related to NO − 3 mobility under subarid conditions (e.g., Frey et al., 2009;Michalski et al., 2005) and given that uncertainties in local accumulation rates and ion transport can result in inaccurate ages when using NO − 3 alone (Graham et al., 2002;Schiller et al., 2009). Based on the results presented here for hyper-arid CTAM ice-free regions and the concerns with boron mobility depending on whether the B species present in the soils is BO 3− 3 (borate) or H 3 BO 3 (boric acid), we suggest that NO − 3 is suitable for interpreting wetting and disturbance histories.
Both meteoric 10 Be and NO − 3 are sourced from atmospheric deposition in the Shackleton Glacier region, and there appears to be a relationship between the two constituents in the soil profiles (Fig. 6b). A similar relationship between soluble salts and meteoric 10 Be was previously documented at Roberts Massif (Graham et al., 1997). NO − 3 is highly mobile in wetter systems, while 10 Be is less mobile under circumneutral pH. Given sustained hyper-arid conditions, minimal landscape disturbance, and negligible biologic activity, one can expect meteoric 10 Be and NO − 3 to be correlated throughout a depth profile given the similar accumulation mechanism (Everett, 1971;Graham et al., 1997). Further, their inventories (Eq. 2) should increase monotoni-3 concentrations in soils Table 3. Estimated exposure durations using the relationship between maximum 10 Be concentration and the calculated inventory; see Fig. 7 (Graly et al., 2010 cally with exposure duration. Deviations from this expected relationship could be due to (1) soil wetting, either in the present or past, (2) deposition of sediment with different 10 Be to NO − 3 ratios compared to the depositional environment, (3) changes in the flux of either 10 Be or NO − 3 with time, and (4) additional loss of NO − 3 due to denitrification or volatilization. The latter two mechanisms are likely minor processes; however, NO − 3 deposition fluxes are known to be spatially variable (Jackson et al., 2016;Lyons et al., 1990). As described above, Roberts Massif, Bennett Platform, and Thanksgiving Valley were selected for further investigation as locations which may represent different depositional environments: hypothesized hyper-aridity, recent glacial activity with large moraines, and active hydrology, respectively. By comparing differences in the expected and observed relationship between 10 Be and NO − 3 , we can infer the processes that have influenced their relationship.

Implications for landscape disturbance and paleoclimate
Our work demonstrates that NO − 3 and 10 Be are correlated in much of the Shackleton Glacier region and that the soil profiles can inform our understanding of surficial processes and soil wetting for the region. Exposure age and cosmogenic nuclide data from across Antarctica show that a polar desert regime began in the mid-Miocene and has persisted into modern time (Lewis et al., 2008;Marchant et al., 1996;Spector and Balco, 2020;Valletta et al., 2015). Additionally, Barrett (2013) provides a detailed review of studies focused on Antarctic glacial history, particularly centered around the "stabilist vs. dynamicist" debate concerning the overall stability of the EAIS. Interpreting 40+ years of data from published literature, it is concluded that the EAIS is stable in the interior with retreat occurring along the margins, including at outlet glaciers (Golledge et al., 2012). Given these findings, we would expect NO − 3 and meteoric 10 Be concentrations to be correlated in hyper-arid Antarctic soils, such as those from the Shackleton Glacier region, as both constituents are derived from atmospheric deposition with minimal alteration at the surface. The major differences between the two concern transport mechanisms; meteoric 10 Be transport is limited by clay particle mobility, and NO − 3 is mobile upon soil wetting. If we assume an "ideal" situation in which an undisturbed hyper-arid soil has accumulated meteoric 10 Be ( Fig. 3a and b), 10 Be concentrations would be highest at the surface and eventually decrease to background levels at depth. None of the profiles we sampled and measured for meteoric 10 Be and NO − 3 reached background concentrations. All profiles had an active layer much shallower than those from the MDV (Graham et al., 2002;Schiller et al., 2009;Valletta et al., 2015). This indicates that the active layer may have deepened and shallowed throughout time, and modern 10 Be mobility is limited to the top ∼ 20 cm for most of the Shackleton Glacier region. Though clay particle translocation by percolating water can explain the correlated behavior of 10 Be and NO − 3 at Roberts Massif and Thanksgiving Valley, it is unlikely that the region had sufficient precipitation for significant percolation over the last 14 Myr given the high NO − 3 concentrations (Menzies et al., 2006). The concentrations of fine particles in the soil profiles also do not change significantly with depth, as would be expected if large precipitation or melt events were frequent (Fig. 4). Additionally, the soil horizons are moderately well defined (Fig. 4), suggesting minimal cryoturbation.
Similar to Arena Valley and Wright Valley in the MDV (Graham et al., 2002;Schiller et al., 2009), NO − 3 concentrations are highest just beneath the surface at Roberts Massif, indicating shallow salt migration under an arid climate. These data indicate that the samples furthest inland at Roberts Massif and Thanksgiving Valley have been fairly undisturbed since at least the middle to late Pleistocene given the estimates of exposure duration (see Sect. 6.2). Since the meteoric 10 Be and NO − 3 profiles at Bennett Platform are mirrored, we argue that the difference could be due to (1) additional 10 Be delivery or (2) enhanced NO − 3 transport. Bennett Platform was the only location we sampled on a large moraine (Fig. 2c), and as a constructional landform we would expect 10 Be to be highest at the surface and decrease to back- Figure 7. Relationship between the measured maximum (or surface) meteoric 10 Be concentration and the calculated inventory (Eq. 2). This relationship is used to infer 10 Be inventories given a maximum or surface concentration (Graly et al., 2010). The solid black line is the power relationship between concentration and inventory, while the dashed grey line is the regression from Graly et al. (2010). ground concentrations. This is generally the observed behavior. The NO − 3 profile behavior is similar to those throughout the Shackleton Glacier region, though the concentrations continue to increase with depth, possibly indicating some percolation of NO − 3 -rich brine. What may be considered the "anomalous" data point is the surface concentration of meteoric 10 Be. Even though we sampled a constructional landform, the sample was collected between two boulder lines in a small, local depression (∼ 1 m) ( Table 2). It is probably no coincidence that this location also has the greatest proportion of fine-grained material in the soil profile. The two boulder lines impede wind flow and act as a sediment and snow trap, possibly resulting in a higher concentration of meteoric 10 Be than expected simply from atmospheric deposition. The snow in the depression may also aid in NO − 3 transport when melted. In this case, additional sediment-laden 10 Be deposition (superseding any erosion) and/or possible salt transport need to be considered to accurately date the moraine.

Attempt at inferring surface exposure duration approximation and thoughts on glacial history
We used the relationship between the maximum meteoric 10 Be concentration in the soil profile and the meteoric 10 Be inventory (Graly et al., 2010) to speculatively infer 10 Be inventories and estimate maximum exposure durations for all 11 locations with and without erosion using Eq. (5) (Fig. 7; Table 3). As is the case for Roberts Massif and Thanksgiving Valley, the highest 10 Be concentrations may not always be at the surface for all locations; however, the relationship is sufficiently strong to provide an estimate of the 10 Be inventory and thus an exposure duration estimate.
We did not measure erosion rates in this study. Balter-Kennedy et al. (2020) determined erosion rates for boulders at Roberts Massif, which were less than 2 cm Myr −1 . Considering we are investigating soils, we chose a conservative value of 5 cm Myr −1 for our calculations. We selected a 10 Be flux value (Q) of 1.3 × 10 5 atoms cm −2 per year from Taylor Dome (Steig et al., 1995) due to a climate similar to that of the CTAM and an absence of local meteoric 10 Be flux data. Compared to the measured inventories from Roberts Massif, Bennett Platform, and Thanksgiving Valley (from the 10 Be depth profiles; see Sect. 5.1), the inferred inventories differ by ∼ 16 %-130 %. The inferred exposure estimates with erosion range from 58 kyr to > 6.5 Myr, and the estimates without erosion range from 57 kyr to 1.94 Myr for Mt. Speed and Roberts Massif, respectively ( Fig. 8; Table 3). With the exception of Roberts Massif, Thanksgiving Valley, and Mt. Speed, the oldest surfaces are those which we sampled furthest from the glacier, which is consistent with our sampling methodology to capture younger and older soils. The sample from Roberts Massif collected closest to the glacier has an estimated exposure duration that is outside the model limits (> 6.5 Myr).
The youngest surfaces we sampled from the Shackleton Glacier region are those from the lowest elevations and closest to the Ross Ice Shelf (Fig. 8). This is generally consistent with previous glacial modeling studies, which show that the greatest fluctuations in glacier height during the LGM were along outlet glacier and ice shelf margins (Golledge et al., 2012;Mackintosh et al., 2011Mackintosh et al., , 2014. Given the low erosion rates throughout Antarctica (Balter-Kennedy et al., 2020;Ivy-Ochs et al., 1995;Morgan et al., 2010) and possibly low background concentrations of meteoric 10 Be (Dickinson et al., 2012;Schiller et al., 2009;Valletta et al., 2015), the Mt. Speed, Mt. Wasko, and Mt. Franke samples were all likely covered by Shackleton Glacier during the LGM, as were the lower-elevation, near-glacier samples from Mt. Heekin, Bennett Platform, and Mt. Augustana. The soils from Schroeder Hill and Roberts Massif have likely been exposed since the early Pleistocene (Fig. 8). We also attempted to estimate exposure durations using two additional methods: (1) the measured 10 Be inventories for Roberts Massif, Bennett Platform, and Thanksgiving Valley and (2) by calculating 10 Be concentrations using regressions of NO − 3 and 10 Be for all seven locations with depth profiles, as detailed in the Supplement. These exposure estimates are similar and range from ∼ 100 kyr at Bennet Platform to < 4.5 Myr at Roberts Massif ( Fig. S4; Table S3).
Sirius Group deposits were observed at Roberts Massif and were deposited as Shackleton Glacier retreated in this region (Fig. 2a). Evidence for a dynamic EAIS is derived primarily from the diamictite rocks (tills) of the Sirius Group, which are found throughout the TAM and include  well-documented outcrops in the Shackleton Glacier region, but their age is unknown (Hambrey et al., 2003). Some of the deposits contain pieces of shrubby vegetation, indicating that the Sirius Group formed under conditions warmer than present with woody plants occupying inland portions of Antarctica (Webb et al., 1984(Webb et al., , 1996Webb and Harwood, 1991). Sparse marine diatoms found in the sediments were initially interpreted as evidence for the formation of the Sirius Group via glacial overriding of the TAM during the warmer Pliocene (Barrett et al., 1992), though it is now argued that the marine diatoms are wind-derived contamination, indicating that the Sirius Group is older (Scherer et al., 2016;Stroeven et al., 1996). We document a large diamictite at site RM2-8 that is underlain by soils with an inferred exposure of at least 1.9 Myr, possibly greater than 6.5 Myr. These exposure duration estimates indicate that the loose Sirius Group diamict was deposited at Roberts Massif at some point after the Pliocene. While these data cannot constrain the age of the formation, we suggest that the diamict could have formed prior to the Pliocene and was transported during the Pleistocene glaciations.

Conclusions
We determined concentrations of meteoric 10 Be and NO − 3 in soils from 11 ice-free areas along Shackleton Glacier, Antarctica, which are among the highest measured meteoric 10 Be concentrations from the polar regions. Concentrations of meteoric 10 Be spanned from 1.9 × 10 8 atoms g −1 at Bennett Platform to 73 × 10 8 atoms g −1 at Roberts Massif. The concentrations of NO − 3 were similarly variable and ranged from ∼ 1 µg g −1 near the ice shelf to 15 mg g −1 near the Polar Plateau. In general, the lowest concentrations of 10 Be and NO − 3 we measured were at low elevations, near the ice shelf, and closest to the glacier.
Since NO − 3 and 10 Be are both derived from atmospheric deposition, we expect the shape of their accumulation profiles to be similar at depth in hyper-arid soils. In general, this was true for Roberts Massif and Thanksgiving Valley, while NO − 3 and 10 Be concentrations were mirrored at Bennett Platform. We conclude that much of the southern Shackleton Glacier region has maintained persistent arid conditions since at least the Pleistocene, though the region may have been warmer and wetter in the past, as evidenced by the presence of the Sirius Group diamict. The onset of aridity is particularly important in understanding refugia and ecological succession in TAM soils. Since parts of the region have remained hyper-arid and undisturbed for upwards of a few million years, prolonged exposure has resulted in the accumulation of salts at high concentrations in the soils. It is an enigma how soil organisms have persisted throughout glacial-interglacial cycles. However, it is possible that organisms have survived near the glacier at locations like Mt. Augustana, where glacial advance appears to have been minimal during the LGM, but seasonal summer melt has the potential to solubilize salts.
Overall, our data show that the relatively youngest soils we sampled were at lower elevations near the Shackleton Glacier terminus and lower elevations further inland (typically near the glacier). Inferred estimates range from 57 kyr (though likely post LGM when corrected) to 1.94 Myr, possibly > 6.5 Myr with erosion. Our sampling scheme was successful in capturing a range of surface exposure durations, which can contribute to growing archives for the CTAM. There are outstanding issues regarding inheritance dynamics of meteoric 10 Be in disturbed environments, particle erosion and deposition rates, and NO − 3 mobility. We hope that future studies will further evaluate the relationship between watersoluble salts (e.g., NO − 3 ) and meteoric 10 Be as a proxy for landscape disturbance and exposure age.
Data availability. The datasets generated for this study are included in the article or Supplement. The meteoric 10 Be data are stored with the US Antarctic Program Data Center (https://doi.org/10.15784/601421; Diaz et al., 2021).