Aim High-latitude biodiversity distributions can preserve signals of the timing and geography of past glaciations, and as such ground truth ice-sheet models. Discrete polar archipelagos offer the fewest confounding factors for testing historic ice position records in extant biodiversity. At South Georgia, two competing geological hypotheses suggest that either the Last Glacial Maximum (LGM) ice was extensive, nearly covering the continental shelf (H 1 Big ice) or restricted to the inner fjords (H 2 Little ice). We examined the past configuration of the South Georgia ice cap using seabed biodiversity.Location South Georgia, Southern Ocean. MethodsWe used a bespoke camera lander (SUCS) and Agassiz trawl deployments across 'big ice' and 'little ice' hypothesized positions of LGM grounded ice around the South Georgia continental shelf. We investigated faunal assemblage structure and richness, especially of brooders, and modelled low dispersal taxa, for example, those with limited pelagic larvae (bryozoans and sponges). ResultsWe found a striking 'line' of major richness discontinuity, with significantly higher richness, especially of brooders and low dispersal model taxa, mainly conforming to the 'big ice' hypothesized position. What few bryozoans and sponges occurred inside this line were a subset of those outside.Main conclusions Benthic biodiversity is consistent with extensive LGM grounded ice advancing to near the shelf break in most, but not all locations around South Georgia's shelf, for example, the eastern shelf area. We suggest that most of the shelf is still undergoing recolonization from when grounded ice covered the shelf~20 kyr ago. Our alternative hypothesis of LGM ice position, H 3 'Limited-Extensive ice', best fitted our data and is easily further testable, but if verified, shows that shelf recolonization following glaciation is much slower than previously thought. This contrasts with surprisingly rapid colonization of continental shelves after ice-shelf collapses, but these are not grounded, which may be crucial to polar recolonization rates.

The Holocene and late glacial history of fluctuations of Stutfield Glacier are reconstructed using moraine stratigraphy, tephrochronology, and dendroglaciology. Stratigraphic sections in the lateral moraines contain tills from at least three glacier advances separated by volcanic tephras and paleosols. The oldest, pre-Mazama till is correlated with the Crowfoot Advance (dated elsewhere to be Younger Dryas equivalent). A Neoglacial till is found between the Mazama tephra and a paleosol developed on the Bridge River tephra. A log dating 2400 BP from the upper part of this till indicates that this glacier advance, correlated with the Peyto Advance, culminated shortly before deposition of the Bridge River tephra. Radiocarbon and tree-ring dates from overridden trees exposed in moraine sections indicate that the initial Cavell (Little Ice Age (LIA)) Advance overrode this paleosol and trees after A.D. 1271. Three subsequent phases of the Cavell Advance were dated by dendrochronology. The maximum glacier extent occurred in the mid-18th century, predating 1743 on the southern lateral, although ice still occupied and tilted a tree on the north lateral in 1758. Subsequent glacier advances occurred ca. 1800–1816 and in the late 19th century. The relative extent of the LIA advances at Stutfield differs from that of other major eastward flowing outlets of the Columbia Icefield, which have maxima in the mid–late 19th century. This is the first study from the Canadian Rockies to demonstrate that the large, morphologically simple, lateral moraines defining the LIA glacier limits are actually composite features, built up progressively (but discontinuously) over the Holocene and contain evidence of multiple Holocene- and Crowfoot-age glacier advances.

Climate-related disturbance regimes are changing rapidly with profound consequences for ecosystems. Disturbance is often perceived as detrimental to biodiversity; however, the literature is divided on how they influence each other. Disturbance events in nature are diverse, occurring across numerous interacting trophic levels and multiple spatial and temporal scales, leading to divergence between empirical and theoretical studies. The shallow Antarctic seafloor has one of the largest disturbance gradients on earth, due to iceberg scouring. Scour rates are changing rapidly along the Western Antarctic Peninsula because of climate change and with further changes predicted, the Antarctic benthos will likely undergo dramatic shifts in diversity. We investigated benthic macro and megafaunal richness across 10–100 m depth range, much of which, 40–100 m, has rarely been sampled. Macro and megafauna species richness peaked at 50–60 m depth, a depth dominated by a diverse range of sessile suspension feeders, with an intermediate level of iceberg disturbance. Our results show that a broad range of disturbance values are required to detect the predicted peak in biodiversity that is consistent with the Intermediate Disturbance Hypothesis, suggesting ice scour is key to maintaining high biodiversity in Antarctica’s shallows.