B23C-01 INVITED 13:45h
New Directions in the Study of Bacteria Inhabiting Very Cold Sea-Ice Formations
The development of new techniques for evaluating physical, chemical and microbial aspects of unmelted ice formations has led to a series of revelations about very cold saline ice (Arctic winter sea ice) as a habitat for life. Some of the remarkable features observed microscopically in unmelted ice at temperatures down to at least minus 15 degrees C include the physical connectivity of liquid brine inclusions on a micrometer scale, the extensive presence of complex exopolymers within the fluid inclusions, and the apparent attachment of dividing bacteria to the ice wall of a brine pore. When coupled with studies that involve melting ice but into salt solutions to minimize thermal and osmotic shocks to bacteria located within the brine inclusions, the hypothesis that bacterial activity continues even at the coldest temperature yet tested (minus 20 degrees C) has been supported. I recently worked to extend the study of very cold unmelted and brine-melted sea ice to include an assessment of extracellular enzyme activity (EEA). In other porous but less extreme environments, the detection of EEA (dominated by proteolytic activity) has reflected an important foraging strategy in use by heterotrophic bacteria dependent on acquiring small-sized, nitrogen-rich organic compounds for their livelihood. Taking advantage of facilities aboard the Canadian icebreaker Amundsen, newly renovated for science and frozen into Franklin Bay (Canadian Arctic) during winter 2004, I explored means to test for EEA (particularly leucine amino-peptidase activity), using fluorescently labeled substrate analogs, under in situ ice-brine conditions and over a range of temperatures and salt concentrations. Although methods remain to be fully optimized, preliminary results indicated detectable EEA down to minus 12 degrees C, with thermal optima much lower than previously detected in warmer (summer) sea ice and examples of possible activity at minus 18 degrees C, the lowest temperature tested. Graphical analyses of EEA in temperature-salt space suggested that high salt concentrations limit enzyme performance at the coldest temperatures, while other experiments pointed to the presence of particulate matter or exopolymers in the ice as possibly compensating for negative salt effects. The detection of enzyme activity under the extreme conditions of winter sea ice bears upon the nature of organic compounds found concentrated within the ice matrix (and available for release during the melting season), bacterial success in extremely cold ice formations, and the discussion of possible life processes in extraterrestrial saline ices.
B23C-02 INVITED 14:00h
Evidence for in-situ methane production in ice based on anomalous isotope analyses
Studying microbial ecology at low temperatures is important for understanding the limits of life processes as well the search for extraterrestrial life. Glacial ice sheets are special habitats where microbes have been preserved for geologically significant periods of time. Glaciers provide three distinct environments for microbial ecosystems. Subglacial lakes beneath the East Antarctic ice sheet provide one of the most intriguing environments that have yet to be explored. The upper portion of a glacier is formed from eolian derived (wind blown) materials (snow, impurities and microbes). Bulk impurity levels tend to be less than a few ppm, cell densities generally below 100 cells/ml and surface temperatures are generally below -15$^{o}$C. Subglacial environments (lowest 20m), on the other hand, tend to have (by comparison with the overlying glacier ) extremely high impurity concentrations, cell densities on the order of 10$^{6}$ cells/ml, and temperatures close to the pressure melting point (~ 0$^{o}$C). Microbial communities in the subglacial environments are comprised of eolian derived organisms that have traveled vertically through the ice sheet as well as organisms that inhabited the soil/rock environment before the glacier formed. Cell density measurements in glacier ice are fairly straightforward given proper cleaning techniques. Whether or not the cells in a glacier are able to grow (or at least maintain their metabolic functionality) while immured in the glacier has yet to be determined. This question remains unanswered largely because the metabolic rates of microbial communities in ice have not been measured in the lab. One way to infer in-situ microbial activity in ice is to analyze the elemental and isotopic composition of gaseous metabolic byproducts that are retained in the ice matrix. We present two case studies in which the measured methane (CH$_{4}$) concentration and isotope values in ice result from in-situ production. Methane measurements spanning the last 25kyr from the Sajama ice core from central Bolivia (18$^{o}$S, 69$^{o}$W, 6542masl), for example, were 1X-5X higher than contemporaneous values recorded in polar ice cores [Campen et al., 2003]. \delta$^{13}$CH$_{4}$ values from five discrete depths were compared to corresponding measurements made on the Taylor Dome ice core and suggest the additional (in-situ) $CH$_{4}$ in the Sajama samples has an average isotopic composition of -63.2\pm2.8\permil. For reference, atmospheric \delta$^{13}$CH$_{4}$ values range from -42 to -45/pm over this period. The Sajama isotope values are characteristic of methanogenic CH$_{4}$ emitted from most terrestrial ecosystems. The second case study revolves around ice that was recovered from a perennially ice covered lake in the McMurdo Dry Valleys, Antarctica. Previous work on ice from Lake Bonney demonstrated a rich microbial consortium located ~2m below the surface [Priscu et al., 1998]. Methane isotope analyses were made on ice from this depth interval to identify the presence of microbially produced CH$_{4}$. \delta$^{13}$CH$_{4}$ and \delta D$CH$_{4}$ results suggest the CH$_{4}$ arises from acetogenic CH$_{4}$ production as opposed to CO$_{2}$ reduction. Campen, R.K., T. Sowers, and R.B. Alley, Evidence of Microbial Consortia Metabolizing Within a Low-Latitude Mountain Glacier, Geology, 31 (No. 3), 231-234, 2003. Priscu, J.C., et al., Perennial Antarctic Lake Ice: An oasis for life in a polar desert, Science, 280, 2095-2098, 1998.
B23C-03 14:15h
Macronutrients in the Summer Sea ice of the Ross Sea.
The development and production of sea ice microbial communities is often linked to limitations imposed by light, temperature and salinity. Nutrients also impose a limitation upon biomass development and production. Nutrient constraints on biomass development was readily evident in surface habitats of the pack ice of the Ross Sea during the summer of 1999. The spatial extent of nutrient depletions, primary production and biomass accumulation- investigated at both large (100s of kilometers) and small (centimeter to meter) scales- showed the importance of the tight coupling between the ice cover's gross morphological features (e.g. brine tubes, cracks and floes size) in determining the extent to which extremely productive communities developed. Specifically, extreme nutrient depletions were often observed in areas where brine tubes or cracks were not readily apparent, whereas, relatively enhanced nutrients and large biomass accumulations (exceeding 50 to 1000+ ug Chla per liter) were often found in association with the floe edges, cracks and brine tubes. These features and links were documented over the entire latitudinal extent of the summer pack ice (January of 1999)- yet, were especially apparent in the southern region. Whether or not these features and the tight coupling observed are a ubiquitous and common feature of the late spring and summer pack ice of the entire Southern Ocean remains to be determined.
B23C-04 14:30h
A Microbial Community in Sediments Beneath the Western Antarctic Ice Sheet, Ice Stream C (Kamb)
In 2000, an ice-drilling project focusing on the "sticky spot" of Ice Stream C recovered cores of sub-glacial sediments from beneath the Western Antarctic Ice Sheet. We have characterized several chemical and microbiological parameters of the sole intact sediment core. Pore waters extracted from these sediments were brackish and some were supersaturated with respect to calcite. Ion chromatography demonstrated the presence of several organic acids at low, but detectable, levels in the pore water. DAPI direct cell counts were approximately 10$^{7}$ cells g$^{-1}$. Aerobic viable plate counts were much lower than direct cell counts; however, they were two orders of magnitude higher on plates incubated at low temperature (4 $\deg$C; 3.63 x 10$^{5}$ CFU ml$^{-1}$) than at higher temperatures (ca. 22$\deg$C; 1.5 x 10$^{3}$ CFU ml$^{-1}$); no colonies were detected on plates incubated anaerobically at either temperature. 16S rDNA clone library analysis indicates extremely limited bacterial diversity in these samples: six phylogenetic clades were detected. The three dominant bacterial phylogenetic clades in the clone libraries (252 clones total) were most closely related to Thiobacillus thioparus (180 clones), Polaromonas vacuolata (34 clones), and Gallionella ferruginea (35 clones) and their relatives; one clone each represented the other three phylogenetic clades (most closely related to Ralstonia pickettii, Lysobacter antibioticus, and Xylella fastidiosa, respectively). These sequences match closely with sequences previously obtained from other subglacial environments in Alaska, Ellesmere Island, Canada and New Zealand. Implications of this microbial community to subglacial chemistry and microbial biogeography will be discussed.
B23C-05 14:45h
A Habitat for Living Microbes on Silt Grains in Greenland Basal Ice at 3053m Depth
Two ice cores 30 km apart extended ~3050 m down to bedrock at Summit, Greenland. Most attention has been focused on their high-resolution climate record covering the last $\sim$110,000 years. In the bottom few meters, we measured up to $\sim$1 wt% of silt (mean diameter 2.5$\mu$m). In that basal ice of GRIP, R. Souchez et al. found a factor of 500 excess of CO$_{2}$ and a factor of $10^4$ excess of CH$_{4}$, and T. Sowers measured a factor 2 $\times 10^4$ excess of CH$_{4}$ at 3043 m in the nearby GISP2 basal ice. V. Miteva et al. discovered $\sim$7 $\times 10^7$ microbial cells/ml in Sowers\'sample. Souchez et al. inferred from the huge excesses that the ice formed by mixing with a peat deposit. It is now inferred from our work and the observations of Miteva et al. that the excesses were the direct products of in-situ microbial metabolism. The basal ice, with its silt and microbial population, likely formed before the ice sheet developed. Using various stains and epifluorescence, we analyzed the relationships of microbes to silt grains at six depths from 3044 to 3053 m in GISP2 ice as well as the climate correlation with microbes along throughout the entire ice core. In the silty ice, microbial concentrations ranged from $\sim$ 10^7$ to $<10^9$/cm$^{3}$ of ice and correlated with weight fraction of silt rather than with depth. Between 92% and 99% of the microbes were attached to silt particles at a typical concentration of $\sim10^7$/cm$^{2}$ of grain surface. This correlation extended down to the smallest silt grains: typically, up to 4 cells were attached to a 0.5$\mu$m grain. With an SEM we found that some cells were dividing, and we were able to culture microbes in low-nutrient liquid media. These two results demonstrated their viability after $10^5$ yr at -9$\deg$C.
B23C-06 15:00h
Microbial activity and nutrient concentrations from a North American rock glacier: An Antarctic analogue.
Here we provide the first report on microbial activity in a previously un-investigated icy system thought to be devoid of life: rock glaciers. Water draining from the Green Lake 5 rock glacier (RG5) in the Colorado Front Range has been sampled for water chemistry since 1998 as part of the Niwot Ridge LTER program. Maximum nitrate concentrations from the outflow of RG5 of 136 umoles/L are an order of magnitude greater than most alpine streams in the western US. Both lab and field incubations showed that microbial activity on the surface of RG5 was similar to that of well-developed tundra soils. Source waters and flowpaths of RG5 outflow were determined using end-member mixing analysis (EMMA) with isotopic and geochemical tracers. EMMA showed that nitrate flux was conservative with respect to flowpaths. Interestingly, the highest nitrate concentrations in RG outflow were from the baseflow component, which we believe was internal ice melt. DOC concentrations in RG5 outflow were similar to nearby streams at 1-2 mg/L. Fluoresence index values for the DOC increased from about 1.4 early in the season to 1.6 late in the season, suggesting a switch from terrestrial DOC production to aquatic microbial precurser DOC material similar to Anarctic streams. More sophisticated PARAFAC analysis of DOC showed that loadings of aquatic microbial components, which are similar to components found in Antarctic streams, increased by a factor of 2-4 from early to late season. Unexpectedly, nitrate concentrations were found to be significantly correlated with these aquatic microbial components (R2 = 0.70). These results indicate a strong microbial presence within the rock glacier itself, which is similar to aquatic microbial activity and not terrestrial microbial activity. A synoptic sampling of the outflow of 11 additional rock glaciers in the Rocky Mountains of Colorado and Wyoming showed similar results, suggesting that our results may be applicable to many other rock glaciers in the Rocky Mountain region.
B23C-07 15:15h
Organic Carbon Dynamics in Glacier Systems
The biogeochemical cycling of organic carbon (OC) has important implications for aquatic system ecology because the abundance and molecular characteristics of OC influence contaminant transport and bioavailability, and determine its suitability as a substrate for microbial metabolism. There have been few studies of OC cycling in glacier systems, and questions remain regarding the abundance, provenance, and biogeochemical transformations of OC in these environments. To address these questions, the abundance and molecular characteristics of OC is investigated in three glacier systems. These systems are characterized by different thermal and hydrological regimes and have different potential OC sources. John Evans Glacier is a polythermal glacier in arctic Canada. Outre Glacier is a temperate glacier in the Coast Mountains of British Columbia, Canada. Victoria Upper Glacier is a cold-based glacier in the McMurdo Dry Valleys of Antarctica. To provide an indication of the extent to which glacier system OC dynamics are microbially mediated, microbial culturing and identification is performed and organic acid abundance and speciation is determined. Where possible, samples of supraglacial runoff, glacier ice and basal ice and subglacial meltwater were collected. The dissolved organic carbon (DOC) concentration in each sample was measured by combustion/non-dispersive infrared gas analysis. Emission and synchronous fluorescence spectroscopy were used to characterize the molecular properties of the DOC from each environment. When possible, microbial culturing and identification was performed and organic acid identification and quantification was measured by ion chromatography. DOC exists in detectable quantities (0.06-46.6 ppm) in all of the glacier systems that were investigated. The molecular characteristics of DOC vary between glaciers, between environments at the same glacier, and over time within a single environment. Viable microbes are recoverable in significant (ca. 10$^{3}$ colony forming units/ml) levels from John Evans Glacier and Victoria Upper Glacier. Identification of these organisms has revealed similarities to bacteria isolated previously from glacier ice and arctic and antarctic terrestrial environments. Organic acid concentrations are higher in the basal ice than in the glacier ice at John Evans Glacier and Victoria Upper Glacier. These results suggest that environmental conditions, such as overridden soil and vegetation type, influence the characteristics of DOC and that microbial cycling of OC is active in glacier systems.
B23C-08 15:30h
Evidence for anoxic conditions in subglacial and proglacial environments
Glacial sediments may be host to a range of microbial communities, which drive oxic waters towards anoxia along hydrological flowpaths. Chemical and isotopic signatures in meltwaters from Finsterwalderbreen, a polythermal glacier on sedimentary bedrock in Svalbard, show clear evidence for anoxia at the glacier bed and in waterlogged proglacial sediments. Increases in 34S and 18O of sulphate in subglacial upwelling waters sampled in 1997 indicate the microbially-mediated reduction of sulphate. The 13C of the dissolved inorganic carbon is isotopically light \{13C = -8 %\}, which is consistent with the use of bedrock kerogen and/or the necromass of sulphide oxidizing bacteria as organic substrates for the sulphate-reducing bacteria. Meltwaters sampled from the same outflow two years subsequently do not show evidence of sulphate reduction. Instead they are depleted in 34S and 18O of sulphate, suggesting that anoxic sulphide oxidation is a significant process and that subglacial flowpaths have become more oxygenated. Proglacial groundwaters are isotopically light with respect to 34S and 18O, demonstrating that anoxic sulphide oxidation is an important process here also. Variations in the isotopic composition of sulphate between years in the subglacial case and within a single summer in the proglacial case are linked to shifts in local hydrological conditions and drainage reorganisation.