Scott E. Ishman

Stability of the Larsen B ice shelf on the Antarctic Peninsula during the Holocene epoch

The stability of the Antarctic ice shelves in a warming climate has long been discussed, and the recent collapse of a significant part, over 12,500 km2 in area, of the Larsen ice shelf off the Antarctic Peninsula has led to a refocus toward the implications of ice shelf decay for the stability of Antarcticas grounded ice. Some smaller Antarctic ice shelves have undergone periodic growth and decay over the past 11,000 yr (refs 7–11), but these ice shelves are at the climatic limit of ice shelf viability and are therefore expected to respond rapidly to natural climate variability at century to millennial scales. Here we use records of diatoms, detrital material and geochemical parameters from six marine sediment cores in the vicinity of the Larsen ice shelf to demonstrate that the recent collapse of the Larsen B ice shelf is unprecedented during the Holocene. We infer from our oxygen isotope measurements in planktonic foraminifera that the Larsen B ice shelf has been thinning throughout the Holocene, and we suggest that the recent prolonged period of warming in the Antarctic Peninsula region, in combination with the long-term thinning, has led to collapse of the ice shelf.

Productivity cycles of 200–300 years in the Antarctic Peninsula region: Understanding linkages among the sun, atmosphere, oceans, sea ice, and biota

Compared to the rest of the world9s oceans, high-resolution late Holocene paleoclimatic data from the Southern Ocean are still rare. We present a multiproxy record from a sediment core retrieved from a deep basin on the western side of the Antarctic Peninsula that reveals a dramatic perspective on paleoclimatic changes over the past 3700 yr. Analyses completed include measurement of magnetic susceptibility and granulometry, bed thickness, particle size, percent organic carbon, bulk density, and microscopic evaluation of diatom and benthic foraminiferal assemblages and abundances. Downcore variability of these parameters demonstrates the significance of both short-term cycles, which recur approximately every 200 yr, and longer term events (≈2500 yr cycles) that are most likely related to global climatic fluctuations. In the upper 600 cm of the core, lower values of magnetic susceptibility (MS) are correlated with lower bulk density, the presence of thinly laminated units, specific diatom assemblages, and generally higher total organic carbon content. Below 600 cm, magnetic susceptibility is uniformly low, though variability in other parameters continues. The magnetic susceptibility signal is controlled primarily by dilution of ferromagnetic phases with biosiliceous material. This signal may be enhanced further by dissolution of magnetite in the magnetic susceptibility lows (high total organic carbon). The role of variable primary productivity and its relationship to paleoclimate is assessed through the diatom data. In particular, magnetic susceptibility lows are characterized by higher than normal abundances of Chaetoceros resting spores. Corethron criophilum and/or Rhizosolenia spp. also are found, as is a higher ratio of the most common species of Fragilariopsis versus species of Thalassiosira . These assemblages are indicative of periods of high primary productivity driven by the presence of a meltwater stabilized water column. The 200 yr cyclicity noted in other paleoclimatic records around the world suggests a global forcing mechanism, possibly solar variability. In addition to the cyclic changes in productivity, overall elevated productivity is noted below 600 cm, or prior to ca. 2500 yr B.P. This increased productivity may represent the tail end of a Holocene climatic optimum, which is widely recognized in other parts of the world, but as yet is poorly documented in Antarctica.

Joint investigations of the Middle Pliocene climate I: PRISM paleoenvironmental reconstructions

Abstract The Pliocene epoch represents an important transition from a climate regime with high-frequency, low-amplitude oscillations when the Northern Hemisphere lacked substantial ice sheets, to the typical high-frequency, high-amplitude Middle to Late Pleistocene regime characterized by glacial—interglacial cycles that involve waxing and waning of major Northern Hemisphere ice sheets. Analysis of middle Pliocene (∼3 Ma) marine and terrestrial records throughout the Northern Hemisphere forms the basis of an integrated synoptic Pliocene paleoclimate reconstruction of the last significantly warmer than present interval in Earth history. This reconstruction, developed primarily from paleontological data, includes middle Pliocene sea level, vegetation, land—ice distribution, sea—ice distribution, and sea-surface temperature (SST), all of which contribute to our conceptual understanding of this climate system. These data indicate middle Pliocene sea level was at least 25 m higher than present, presumably due in large part to a reduction in the size of the East Antarctic Ice Sheet. Sea surface temperatures were essentially equivalent to modern temperatures in tropical regions but were significantly warmer at higher latitudes. Due to increased heat flux to high latitudes, both the Arctic and Antarctic appear to have been seasonally ice free during the middle Pliocene with greatly reduced sea ice extent relative to today during winter. Vegetation changes, while more complex, are generally consistent with marine SST changes and show increased warmth and moisture at higher latitudes during the middle Pliocene.

Oceanographic and physiographic controls on modern sedimentation within Antarctic fjords

Physical oceanographic data and modern surface sediments were collected from eleven fjords along the western side of the Antarctic Peninsula and South Shetland Islands. Surface sediment samples (62) were analyzed for texture and total organic carbon content. The distribution of biogenic and terrigenous facies within the fjords is controlled by bay geometry and oceanographic regime. Climate plays a secondary role but, along with ice drainage basin size, controls the rate of terrigenous supply to the glacial marine environment. Specifically, fjords along the Danco Coast and Palmer Archipelago with a high length to width ratio tend to have bottom sediments that are arenaceous where ice-rafted sediment is released preferentially at the head of the fjord. Biogenic facies are favored where the bay geometry is complex. Where such complexity exists, separate oceanographic regimes develop that lead to separation of terrigenous and biogenic sediments. Processes of interflow (mid- and deep-water turbid cold tongues) and Coriolis deflection produce terrigenous facies along the inner fjord and western edges of a fjord system. Warm outer bay waters tend to develop a stable eddy circulation pattern that favors the productivity of phytoplankton in the surface layers. Outer bays are therefore floored with organic-rich siliceous muds and ice-rafted material. Only in the South Shetland Islands is melt-water input significant enough to generate estuarine circulation within the fjord, but here strong bottom currents result in arenaceous bottom sediments with no biogenic facies. Ice-rafted diamictons are produced proximal to the edges of small tide-water glaciers in the South Shetlands. The facies relationships established in this study provide a strong reference for paleoclimatic studies that utilize downcore measurements of texture and organic carbon.

Holocene history of the Larsen-A Ice Shelf constrained by geomagnetic paleointensity dating

A sedimentary record collected from beneath the former Lar- sen-A Ice Shelf reveals the Holocene history of the Larsen-A re- gion. The record begins with the transition from grounded ice to a floating ice shelf, completed by 10.7 6 0.5 ka, and ends with the modern recession. The record contains several late Holocene dia- tomaceous ooze layers that suggest proximity to productive open- water events. Radiocarbon ages obtained from these sediments were complicated by the presence of detrital and reworked carbon. We have eliminated these complications and constructed a chro- nology for the Larsen-A Ice Shelf history via tuning of the geo- magnetic field paleointensity record with a reference curve. This approach provides chronological control to sediment sequences that lack appropriate material for radiocarbon dating. Geomag- netic paleointensity features with wavelengths of 2-3 k.y. can be recognized and interhemispherically correlated, illustrating the po- tential to use geomagnetic paleointensity variations as a global cor- relation and dating tool at sub-Milankovitch time scales.

Climatic variability in the eastern United States over the past millennium from Chesapeake Bay sediments

Salinity oscillations caused by multidecadal climatic variability had major impacts on the Chesapeake Bay estuarine ecosystem during the past 1000 yr. Microfossils from sediments dated by radiometry ( 14 C, 137 Cs, 210 Pb) and pollen stratigraphy indicate that salinity in mesohaline regions oscillated 10‐15 ppt during periods of extreme drought (low fresh-water discharge) and wet climate (high discharge). During the past 500 yr, 14 wet-dry cycles occurred, including sixteenth and early seventeenth century megadroughts that exceeded twentieth century droughts in their severity. These droughts correspond to extremely dry climate also recorded in North American tree-ring records and by early colonists. Wet periods occurred every ~60‐70 yr, began abruptly, lasted <20 yr, and had mean annual rainfall ~25%‐30% and fresh-water discharge ~40%‐50% greater than during droughts. A shift toward wetter regional climate occurred in the early nineteenth century, lowering salinity and compounding the effects of agricultural land clearance on bay ecosystems.

Late Holocene advance of the Muller Ice Shelf, Antarctic Peninsula: sedimentological, geochemical and palaeontological evidence

Marine sediment cores were obtained from in front of the Muller Ice Shelfin Lallemand Fjord, Antarctic Peninsula in the austral summer of 199&91. Sedimentological and geochemical data from these cores document a warm period that preceded the advance of the Miiller Ice Shelf into Lallemand Fjord. The advance of the ice shelf is inferred from a reduction in the total organiccarboncontent and an increase in well-sorted, aeolian, sand in cores proximal to the present calving line. This sedimentological change is paralleled by a change in the foraminifera1 assemblages within the cores. Advance of the ice shelf is indicated by a shift from assemblages dominated by calcareous benthic and planktonic forms to those dominated by agglutinated forms. A I4C chronology for the cores indicates that the advance of the Miiller Ice Shelf took place c. 400 years ago, coincident with glacier advances in other high southern latitude sites during the onset of the Little Ice Age. Ice core evidence, however, documents this period as one of warmer temperatures for the Antarctic Peninsula. We suggest that the ice shelf advance was linked to the exclusion of circumpolar deep water from the fjord. This contributed to increased mass balance of the ice shelf system by preventing the rapid undermelt that is today associated with warm circumpolar deep water within the fjord. We also document the recent retreat of the calving line of the Miiller Ice Shelf that is apparently in response to a recent (four decade long) warming trend along the western side of the Antarctic Peninsula.

Historical trends in Chesapeake Bay dissolved oxygen based on benthic foraminifera from sediment cores

Environmentally sensitive benthic foraminifera (protists) from Chesapeake Bay were used as bioindicators to estimate the timing and degree of changes in dissolved oxygen (DO) over the past five centuries. Living foraminifers from 19 surface samples and fossil assemblages from 11 sediment cores dated by210Pb,137Cs,14C, and pollen stratigraphy were analyzed from the tidal portions of the Patuxent, Potomac, and Choptank Rivers and the main channel of the Chesapeake Bay.Ammonia parkinsoniana, a facultative anaerobe tolerant of periodic anoxic conditions, comprises an average of 74% of modern Chesapeake foraminiferal assemblages (DO-0.47 and 1.72 ml l−1) compared to 0% to 15% of assemblages collected in the 1960s. Paleoecological analyses show thatA. parkinsoniana was absent prior to the late 17th century, increased to 10–25% relative frequency between approximately 1670–1720 and 1810–1900, and became the dominant (60–90%) benthic formaniferal species in channel environments beginning in the early 1970s. Since the 1970s, deformed tests ofA. parkinsoniana occur in all cores (10–20% ofAmmonia), suggesting unprecedented stressful benthic conditions. These cores indicate that prior to the late 17th century, there was limited oxygen depletion. During the past 200 years, decadal scale variability in oxygen depletion has occurred, as dysoxic (DO=0.1–1.0 ml l−1), perhaps short-term anoxic (DO<0.1 ml l−1) conditions developed. The most extensive (spatially and temporally) anoxic conditions were reached during the 1970s. Over decadal timescales, DO variability seems to be linked closely to climatological factors influencing river discharge; the unprecedented anoxia since the early 1970s is attributed mainly to high freshwater flow and to an increase in nutrient concentrations from the watershed.

Oceanographic controls on benthic foraminifers from the Bellingshausen margin of the Antarctic Peninsula

Abstract Surface sediment samples collected from the Bellingshausen Sea continental margin of the Antarctic Peninsula were analyzed for benthic foraminifers. A total fauna including live and dead specimen counts was used to statistically identify discrete benthic foraminifer assemblages. Two major assemblages were defined, the Bulimina aculeata and Fursenkoina spp. Assemblages. The distribution of the Bulimina aculeata and Fursenkoina spp. Assemblages are closely associated with the distribution of the two dominant bottom water masses on the continental margin, the Circumpolar Warm Deep Water and Weddell Sea Transitional Water, respectively. Their distributions cannot be related to organic carbon content or sediment texture of the surface sediments. The faunal composition of the Bellingshausen Sea margin assemblages and their water mass associations are similar to faunal-water mass relationships observed in the Weddell Sea and South Atlantic Ocean where Bulimina aculeata is associated with Circumpolar Deep Water and Nonionella iridea, N. bradii and Trifarina earlandi are associated with Eastern Shelf and Modified Weddell Sea Water.

Impacts, mega-tsunami, and other extraordinary claims

We welcome the comments of Abbott et al., Firestone and West, and Bunch et al.; in fact, we anticipated and hoped for the opportunity to further discuss the rash of recent impact claims and alternative interpretations. Abbott et al. reiterate claims for multiple large Holocene ocean impacts, resulting in “mega-tsunami” recorded by “chevrons.” Longterm meteorological records from stations near two chevron sites show the features to be precisely aligned with dominant winds (Pinter and Ishman, 2008), and their morphology is that of parabolic dunes. Abbott et al. argue that it is impossible to transport marine microfossils into these chevrons by wind. In actuality, coastal dune sand is commonly full of marine micro(and macro) fossils. Two samples of Pleistocene eolianite that we collected in California from 114 m and 230 m elevation contained numerous well-preserved foraminifera spanning 11 genera. Abbott et al. also argue that ironand chromium-stained foraminifera in chevrons and ocean cores represent splashes of impact melt. CaCO 3 melts are documented for the Chicxulub impact site, but these are a far cry from “well-preserved carbonate microfossils” encased in silicate melt. Salge (2007) describes silica and associated carbonate at Chicxulub as “carbonate melt particles,” recrystallized and decomposed calcite, and CO 2 back-reacting to calcite. Abbott et al. also state “[r]ecent experiments have replicated grasses intact within silicate impact melts.” However, the Harris and Schultz (2007) abstract merely suggests a mechanism for “grass-like remains in impact melt breccias.” Fully carbonized plant remains are documented in the geological record (e.g., Scott, 2008), but few researchers would accept unaltered plant material or foraminiferal tests in direct contact with silicate or metallic impact melt. The comments by Firestone and West and Bunch et al. reiterate the claim for a catastrophic North American impact event at 12.9 ka, as presented in Firestone et al. (2006, 2007a). Firestone and West try to distance themselves from earlier claims (e.g., “cosmicray jets,” “deadly nerve toxins” in Pleistocene algal mats, etc.), yet Firestone et al. (2007b) identify a new 33-ka airburst impact event (i.e., no crater) based on micrometeorite “bullets” lodged in mammoth tusks. Regarding the “12.9 ka event,” Firestone et al. (2007a) present an unusual assortment of evidence well outside criteria accepted by the impact community (e.g., Reimold, 2007). A common theme among the bevy of marginal impact claims is that they eschew accepted criteria in favor of new and untested markers. Our article offers an alternative hypothesis. Bunch et al. advocate that spikes in microspherule frequency require an impact event— we suggest that much of this material represents micrometeorite ablation fallout, which can and should be concentrated at any depositional hiatus or in any condensed section (see calculations at www.geology.siu.edu/GSATSupplement.pdf). Further concentration could result from sediment reworking or from terrestrial sources. It is suggestive that Firestone et al.’s highest spherule counts are from Michigan and Alberta, downwind of the ca. 13 ka eruption sequence at Glacier Peak, Washington (Mastin and Waitt, 2000). Bunch et al. and Firestone and West also highlight iridium at up to 117 ppb, higher than many Cretaceous-Tertiary sites. This comparison is disingenuous. The 117 ppb concentration and another of 51 ppb Ir were from spherule and magnetic grain separates, fully consistent with a noncatastrophic micrometeorite source (see www.geology.siu.edu/GSATSupplement.pdf). Firestone et al.’s bulk Ir concentrations peaked at just 2.3–3.8 ppb; maximum values were below detection at six of their 10 sites. Firestone et al. (2007a) note that these values are anomalously low, requiring them to invoke an Ir-depleted impactor. Our hypothesis of ubiquitous micrometeorite debris, concentrated by variation in clastic input and other terrestrial mechanisms, is eminently testable. We predict elevated concentrations of most or all of Firestone et al.’s markers at multiple time horizons in similar sequences. Our research group and several others are now separating microspherules collected from a range of sites and ages. The 12.9 ka impact hypothesis may be confirmed, or it may soon run into a wall of contrary data. The latest Pleistocene and early Holocene span a period of dramatic paleo-environmental changes. Against such a background, workers seeking grand unifying explanations should be aware of the danger of “selection bias.” Selection bias was invoked following claims for Paleolithic artifacts at the Calico site in California. At Calico, ~200 crude man-made objects were identified among countless fractured and unfractured clasts in a chert-rich fanglomerate. After years of acrimonious debate, scientific consensus recognized that the presumed artifacts represented “a biased sample of lithics from the total population of naturally fractured lithics at that site” (Duvall and Venner, 1979)—i.e., a few markers collected in good faith from an abundant background, combined with a good story and some wishful thinking. The real lesson of Calico may be that 30+ years later, a few hardcore proponents cling to the original story. We have no desire to squelch new scientific hypotheses, but the danger is these “extraordinary claims,” once disseminated in the popular imagination, can be almost impossible to dispel.