IODP

doi:10.2204/iodp.pr.347.2014

Principal site results

Lithostratigraphy

At Site M0060 in the Anholt Loch, in the Kattegat area, we recovered the thickest stratigraphic succession to 229.5 meters below seafloor (mbsf). The lowermost stratigraphic units are dominated by clast-poor sandy diamicton with charcoal clasts deposited through mass-transport processes, overlain by fluvial or deltaic sands and a poorly recovered marine mollusk–bearing silty clay with dispersed clasts. The upper part of the succession comprises interbedded sand, silt, clay, and diamicton, as well as laminated clay with dispersed clasts, deposited in an ice-influenced depositional environment. These strata are covered by a thin sand unit containing marine mollusk beds in fining-upward sequences deposited near wave base.

Sites M0059, M0067, M0063, M0064, M0065, and M0066 in the Baltic Sea recovered different thicknesses of a deglacial sequence. This sequence comprises a diamicton or sand unit at the base, overlain by up to 45 m of both massive and laminated glacial lake clays with rhythmic laminations, convolute bedding, and intraclasts. The clays are unaffected by bioturbation, and the rhythmic laminae are interpreted as varves. Between 2 and 35 m of greenish black clay was deposited at the top at each site. Some intervals of greenish black clay have characteristic “mat-like” laminations interpreted to represent accumulations of algae through lake-dump events. Other intervals contain marine mollusk fragments and probably represent a brackish-marine depositional environment. Black, presumably iron sulfide intervals are common in this upper part of the succession, and variable intensities of bioturbation were observed.

Sites M0061 and M0062 were cored in the Ångermanälven estuary. Each site terminated in a sand unit that gradually changes uphole from interbedded sand and silt to rhythmically laminated silty clays with upward-decreasing lamina thickness. Lamina couplet characteristics suggest that rhythmites could be interpreted as varves. At the top of the succession at each site is 8–11 m of greenish black laminated organic clay with meter-scale beds of black iron sulfide concentrations. The stratigraphy at these sites represents a change from a fluvially or glaciofluvially dominated setting to a glacial lake and finally to an estuarine environment.

The intervals of iron sulfide concentrations in the greenish black clays of the Baltic Sea and the Ångermanälven estuary are laminated and lack evidence of bioturbation. These sedimentary properties are consistent with sediments deposited in a protected basin with periods of low oxygen to account for the lack of evidence for macrobenthic life and high organic matter deposition as a precursor to iron sulfide formation.

Biostratigraphy

Siliceous microfossils

Site M0059

Diatoms, silicoflagellates, ebridians, and chrysophycean cysts were identified in Hole M0059A. The lowermost samples from Core 347-M0059-26H (~82.7–83.2 mbsf) contain a mixed freshwater/brackish diatom assemblage. This assemblage is overlain by a series of barren samples, until Core 16H (~50–53 mbsf). Cores 16H and 15H contain a freshwater assemblage, dominated by Aulacoseira and Stephanodiscus taxa. Cores 14H through 1H (~46–0 mbsf) contain a brackish-marine assemblage containing Chaetoceros, Thalassionema, and Thalassiosira. Preservation of opal at Site M0059 is considered to be moderate to good, pending quantitative preservational analysis, based on the presence of lightly silicified taxa and fine structures (i.e., vella) in many intervals.

Site M0060

Diatoms and chrysophycean cysts were identified in Hole M0060A. Cores 347-M0060A-82H through 9H (~203–12 mbsf) contain only rare fragmentary diatom valves and rare chrysophycean cysts. Many of these cores are completely barren. Cores 8H through 6H (~12–6 mbsf) contain a low abundance mixed freshwater/brackish diatom assemblage. Chrysophyte cysts are found only in Core 6H. All samples analyzed from Core 6H were barren. Opal preservation is considered to be moderate, pending further analysis. Diatoms are generally fragmentary, and gracile species are rare.

Site M0061

Diatoms, ebridians, and chrysophycean cysts were identified in Hole M0061A. Cores 347-M0061A-10H through 5H (~25–11 mbsf) are barren of siliceous microfossils, with the exception of Core 6H. Core 6H contains rare diatom valves in one sample. The upper parts of Core 4H and lower sections of Core 3H (~9–7 mbsf) contain a freshwater diatom assemblage, primarily containing Aulacoseira and Stephanodiscus, as well as epiphytic Gomphonema species. This is overlain by a brackish diatom assemblage in Core 3H containing Chaetoceros and Thalassiosira species. Cores 2H and 1H contain an estuarine diatom assemblage. This assemblage is diverse, containing Aulacoseira, Cyclotella, and Thalassiosira. Influence of the Ångermanälven River is noted in the diatom assemblage present at Site M0061, especially in Cores 1H and 2H (~4.8–1.5 mbsf), by the presence of periphytic taxa. Sea ice diatoms are noted throughout the diatom-bearing section. Chrysophycean cysts are found below the diatom-bearing cores, in Cores 7H and 6H (~17.6–11 mbsf). Additional samples were analyzed from Cores 347-M0061B-2H and 3H. A very similar brackish/estuarine assemblage is found in Hole M0061B. Opal preservation is generally considered to be good, pending quantitative analysis, based on the presence of lightly silicified taxa and fine structures.

Site M0062

The siliceous microfossil record from Site M0062 strongly resembles that of Site M0061. One key difference is the lack of ebridians at Site M0062. Hole M0062A, however, shows a stronger influence of the Ångermanälven River. Specifically, Core 13H through the lowermost sections of Core 4H (~35.9–9 mbsf) are barren of siliceous microfossils. Diatoms and chrysophycean cysts are present for the remainder of the analyzed material, from Core 4H to the top of Core 2H (~10–0.5 mbsf). The upper portions of Core 4H and the lower portions of Core 3H contain a freshwater assemblage. This is overlain by a brackish assemblage in Cores 3H and 2H (~7–0.5 mbsf). The upper portions of Core 2H contain a freshwater assemblage, strongly influenced by the Ångermanälven River. The diatom taxa seen at Site M0062 are very similar to those recorded at Site M0061.

Site M0063

Sediment samples from Site M0063 contain diatoms, chrysophytes, ebridians, and silicoflagellates. The lower samples analyzed in Hole M0063A, from Core 347-M0063A-36H to the lower sections of Core 15H (~102–42.5 mbsf), are barren of siliceous microfossils. The upper sections of Core 15H contain a low-diversity brackish water assemblage, dominated by Thalassiosira. This is overlain by another barren interval from Core 14H through Core 10H (~42.5–26 mbsf). Core 10H contains a freshwater diatom (Aulacoseira and Stephanodiscus) and chrysophyte-bearing assemblage. This freshwater assemblage is overlain by diverse siliceous microfossil flora containing silicoflagellates, chrysophycean cysts, ebridians, and brackish water diatoms, principally from Chaetoceros, Cyclotella, and Thalassiosira. Sea ice–related diatom taxa are common throughout this uppermost brackish interval. Additional samples were analyzed from Holes M0063C and M0063E. In Hole M0063C, samples were analyzed from Core 347-M0063C-22H (42–44 mbsf), and both the lowermost barren interval and the lower brackish Thalassiosira interval were identified. In Hole M0063E, samples were analyzed from Cores 347-M0063E-22H through 13H (~44–24 mbsf). The lower barren interval and freshwater interval were identified. Opal preservation is considered to be excellent, pending quantitative analysis, based on the presence of lightly silicified taxa and fine structures.

Site M0064

Only one analyzed sample from Site M0064 contained siliceous microfossils. The uppermost sample, from Core 347-M0064A-1H (0–3 mbsf), contained a typical Baltic brackish water diatom assemblage, consisting of diatom taxa primarily from the genera Chaetoceros, Cocconeis, Cyclotella, and Thalassiosira. Chrysophycean cysts were also present. The remaining cores were barren of siliceous microfossils.

Site M0065

The siliceous microfossil record from Site M0065 is very similar to that of Site M0063. The lowermost samples analyzed from Hole M0065A were barren of siliceous microfossils, from Cores 347-M0065A-15H through 5H (~46–12 mbsf). The uppermost intervals of Core 5H contain a brackish-freshwater assemblage composed of Aulacoseira, Fragilaria, Navicula, Stephanodiscus, and Thalassiosira. This is overlain immediately by a low-diversity freshwater flora composed of diatoms from Aulacosiera and Stephanodiscus and chrysophytes in Core 4H (~12–8.5 mbsf). Cores 3H and 2H (~8.5–2 mbsf) contain a subrecent diatom, silicoflagellate, ebridian, and chrysophyte assemblage typical of the Baltic Sea. This assemblage is composed primarily of Chaetoceros, Dimeregramma, Grammatophora, Paralia, and Thalassiosira species. Diatom preservation is considered to be generally poor in the brackish sections, based on the level of visible corrosion and fragmentation of diatom valves. Preservation is of higher quality in the freshwater sections. Quantitative analysis will follow.

Site M0066

No siliceous microfossils were found in any samples from Site M0066.

Site M0067

Diatoms and silicoflagellates were identified in Hole M0067B. Chrysophytes were also present in Core 347-M0067B-2H only (3–5 mbsf). A typical subrecent Baltic Sea diatom assemblage was recorded from the lower portions of Core 2H to the top of Core 1H. This assemblage contained primarily diatoms from the genera Actinoptychus, Chaetoceros, Dimeregramma, Diploneis, Grammatophora, Opephora, and Thalassiosira. Cores 347-M0067A-1H and 2H (4–0 mbsf) were also analyzed, yielding a nearly identical assemblage, which is lacking in chrysophytes. Opal preservation is considered to be good, pending quantitative analysis, based on the preservation of gracile diatom species and fine structures.

Palynology

Sites M0059 and M0067

At Site M0059, the uppermost samples contain palynomorphs in good preservation, whereas samples from greater depths show signs of degradation/oxygenation and contain reworked (partly Tertiary) palynomorphs. Pollen spectra from the uppermost interval reveal a broad-leaf tree pollen succession, known from terrestrial pollen records from the southern Baltic region. High percentages of Fagus in the uppermost samples indicate an age younger than ~2.5 ka BP. Mass occurrences of Gymnodinium cysts in the uppermost 38 mbsf also imply sub-Atlantic age. High abundances of dinoflagellate cysts point to a marine environment. Between ~35 and 50 mbsf, high percentages of Ulmus pollen indicate ages between 9.5 and 6.0 ka BP. A decline in freshwater algae at 49.16 mbsf implies the transition from the Ancylus Lake stage to marine conditions. Samples from lower intervals contain limited in situ pollen and high amounts of reworked Tertiary pollen. Numerous freshwater algae and aquatic insect larvae remains are present at greater depths. Combined, these findings indicate meltwater inflow with redeposition of Tertiary sediments. Results for Site M0067 are similar to those for Site M0059.

Site M0060

Twenty-five sediment samples from Site M0060 were analyzed, but except for the uppermost sample (0.75 mbsf), all analyzed samples indicate a high degree of oxygenation and are characterized by high percentages of bisaccate pollen grains and high amounts of reworked Tertiary pollen. The uppermost sample revealed high pollen concentration and no reworked pollen. The pollen association indicates an early Holocene age for this sample.

Sites M0061 and M0062

Twenty-five samples in total have been analyzed for the northernmost Sites M0061 and M0062. The uppermost samples contained enough palynomorphs to generate statistically relevant results, whereas samples from greater depths (from silty/sandy sediments) were almost barren of palynomorphs. Pine and birch pollen dominate the assemblages in all samples from the upper parts of the sites. The youngest samples contain spruce pollen in significant amounts, indicating an age of ~3.0 ka BP. The rarity of marine taxa in the upper parts points to a strong terrestrial influence. Nonpollen palynomorphs (algae and insect remains) indicate strong freshwater input or lacustrine conditions for the samples from the upper part of Site M0062. Some samples also contain Thecamoeba remains. An older phase is indicated by a pollen spectrum from Site M0062, which may be provisionally ascribed to the onset of the Holocene.

Site M0063

The pollen record from the interval from 0.06 to 39.45 mbsf at Site M0063 indicates a late to mid-Holocene age. Palynomorph frequency and preservation diminishes toward the bottom part of this interval. Analyses of two samples from the lower part resulted in reliable pollen spectra. One sample (56.80 mbsf) may reflect a late glacial interstadial. A pollen assemblage at 66.18 mbsf may point to a late glacial age and suggests a transition from colder, steppe-like landscape to a landscape with pine-birch type of forest.

Site M0064

Most of the palynomorphs in all samples analyzed for Site M0064 have been degraded/oxygenated to a high degree. No sample contained enough palynomorphs in situ to yield statistically relevant data. Reworked dinoflagellate cysts in samples from the lower part are of Paleogene origin.

Sites M0065 and M0066

Among the four examined sediment samples from Sites M0065 and M0066, only the two shallowest, at 2.17 and 8.82 mbsf, are characterized by high pollen concentrations. Relatively high Quercus, together with Picea percentages, imply a late Atlantic/Subboreal age for these spectra. There may, however, be pollen-transportation bias due to the offshore position of the site.

Foraminifers

Foraminifers in the BSB almost exclusively occur as benthic foraminifers. Planktonic foraminifers are only rarely found in the Kattegat because of a combination of very shallow water depths (average Baltic Sea depth = 50 m; Andren et al., 2012) and low salinity, which declines along a transect from the marine Kattegat (32) to the brackish Landsort Deep (12) and the fresher Ångermanälven estuary in the Bothnian Gulf (5) (Swedish Meteorological and Hydrological Institute [SMHI]; www.smhi.se/en). Additionally, low oxygen concentrations in the bottom waters of many of the Baltic basins, especially Landsort Deep, may limit the survival of benthic organisms. In general, it seems that higher diversity of benthic foraminiferal species occurring in the sediments corresponds to higher bottom water salinity.

Site M0059

Foraminifers commonly occur at Site M0059 in the Little Belt from the seafloor to 50 mbsf. The assemblage diversity is low with Elphidum excavatum dominant and other Elphidium species occurring occasionally. This low faunal diversity suggests salinities lower than in the Kattegat but higher than in the Bornholm Basin and the Landsort Deep. A second interval with abundant foraminifers occurs from 80 to 85 mbsf. In addition to the dominant Elphidium spp., other species like Bucella frigida and Ammonia beccari are present, possibly suggesting colder conditions.

Site M0060

Site M0060, located near the island of Anholt in the Kattegat, is the most marine of the sites cored during this expedition, with modern bottom water salinity as high as 32 (SMHI; www.smhi.se/en). Very diverse fauna were present in some deeper intervals of Site M0060. The upper 21 m of Site M0060, however, contains a moderate diversity faunal assemblage often dominated by the genus Elphidium (particularly Elphidium excavatum f. clavata), with Cassidulina sp., Bulimina marginata, and Quinqueloculina sp. occurring in low abundances. The limited diversity and lack of warmer boreal species compared to modern Kattegat fauna (Nordberg and Bergsten, 1988) in all but the uppermost core suggests that only a thin Holocene deposit is present. Between 21 and 75 mbsf, low diversity and the dominance of E. excavatum f. clavata together with Cassidulina sp. suggests that a potentially continuous deglacial sequence is present. Below 75 mbsf, two intervals between 2 and 10 m thick occur with diverse faunal assemblages, though they are generally poorly recovered. These assemblages are dominated by B. marginata and Hyalinea balthica, indicating warmer and more marine conditions possibly similar to today. Although absolute age determinations have not been established to date, these intervals might be linked to previous interstadials or interglacials.

Site M0061

The occurrence of foraminifers at Site M0061 in the Ångermanälven estuary is the first time foraminifers have been recorded in the Bothnian Bay. Foraminifers were found in an interval between 1.5 and 6 mbsf, and all belonged to the species Elphidium albiumbilicatum, which is known for its tolerance of very low salinity conditions (<10; Lutze, 1965). The absence of foraminifers in the most recent part of the record, which has a present-day bottom water salinity of just 5 (SMHI; www.smhi.se/en), suggests that conditions were slightly more saline at the time when E. albiumbilicatum occurred and can possibly be linked to the middle Holocene (Widerland and Andersson, 2011).

Site M0062

Foraminifers do not occur at Site M0062, presumably because of the low salinity in this upper estuarine site today and in the past.

Site M0063

Site M0063 in the Landsort Deep is today characterized by low bottom water oxygen concentrations, and these low-oxygen conditions have occurred to varying degrees throughout the Littorina stage of the Holocene. Foraminifers occur continuously downcore to nearly 30 mbsf and are often common to abundant to 20 mbsf, suggesting that oxygen levels were generally high enough for subsistence. The harsh conditions are nevertheless reflected in the low diversity of the fauna, which is dominated by opportunistic and stress-tolerant Elphidium spp., especially E. excavatum f. clavata and sometimes Elphidium excavatum f. selseyensis. The absence of more marine species during the middle Holocene suggests that bottom water salinity likely did not increase above 20 (Miettinen et al., 2014).

Sites M0064–M0066

In the Bornholm Basin and Hanö Bay (Sites M0064–M0066), foraminifers were only present at Site M0065. Elphidium spp. is dominant in the upper 12 m of Site M0065 (primarily Elphidium excavatum clavatum), but the absolute number of specimens is generally low. As modern conditions in the Bornholm Basin and Hanö Bay are more saline than in the Landsort Deep, the similarity between fauna at Sites M0065 and M0063 suggests that a full Holocene sequence may not have been recovered. This lack of recovery is most likely related to the operational need to open hole the upper 2 m of sediment because of the risk of contamination from dumped chemical munitions.

Site M0067

Site M0067, like Site M0059, is located in the Little Belt. The two short holes drilled here contained an interesting foraminiferal assemblage in the upper 4.5 mbsf. Although the sites are located near each other, the assemblage at Site M0067 is significantly different from the one at Site M0059. The uppermost core at this site is dominated by the agglutinated foraminifer Eggerelloides scaber (Frenzel et al., 2005), whereas several Elphidium spp. and A. beccarii dominate the fauna of the lowermost samples.

Ostracods

Ostracods are poorly represented in records from the Baltic Sea, mainly because of significant dissolution of calcareous carapaces in the organic-rich sediments (Frenzel et al., 2010). At the same time, abundant data are available on ostracod assemblages living in the Baltic Sea and on the ecology of the ostracod species (e.g., Elofson, 194; Rosenfeld, 1977; Frenzel et al., 2010). Baltic Sea bottom water salinity varies greatly from 31 in the Kattegat to values of 7–8 in the southeastern part and 10–12 in the central Baltic (Gustafsson and Westman, 2002). The presence of ostracod taxa is dependent on a specific range of environmental parameters such as salinity, temperature, and oxygen, making ostracod assemblage analysis a good instrument for reconstructions of Baltic Sea paleoceanography (e.g., Viehberg et al., 2008; Kristensen and Knudsen, 2006).

Site M0059

Ostracods occur at Site M0059 and are most abundant and diverse in the upper 60 mbsf. Variations in composition of ostracod assemblages suggest bottom water salinity changes from freshwater (Candona spp.) to brackish-marine conditions with the following characteristic taxa: Leptocythere spp., Palmoconcha spp., Cytheropteron latissimum, and Sarsicytheridea bradii. Around 80 mbsf, a different assemblage was recorded, comprising different ecological groups from freshwater to shallow-water marine together with redeposited pre-Quaternary valves, suggesting a high-energy shallow environment.

Site M0060

Site M0060 contains an abundant ostracod assemblage to 20 mbsf. Below this level only scattered juvenile and redeposited valves were recorded. The interval 6–20 mbsf is characterized by a diverse marine assemblage with Elofsonella concinna, Sarsicytheridea punctillata, Roundstonia globulifera, and Heterocyprideis sorbyana, with some of the marine taxa typical of North Atlantic waters and high salinity ≥26–30 such as Bythocythere constricta, Cytheropteron biconvexa, and Cytheropteron arcuatum occurring in the lower part of this interval. At ~6 mbsf, shallow brackish-marine and freshwater ostracods (Ilyocypris sp.) were identified, suggesting a shallow brackish environment.

Site M0061

At Site M0061, ostracods occur in low abundance to 6 mbsf. The assemblage is dominated by Paracyprideis fennica. Low abundance and taxonomic diversity implies harsh environments for ostracods. Similarity between ostracod and foraminifer records implies that salinity decreases uphole.

Site M0063

Site M0063 contains a very limited ostracod assemblage. Ostracods occur in low abundance in the upper 18 mbsf and at ~26 and ~39 mbsf. Low abundance and poor preservation prevented paleoenvironmental interpretation of the ostracod data.

Site M0065

Site M0065 is characterized by a low-diversity ostracod assemblage from the surface to 9 mbsf. Predominance of brackish water taxa (Palmoconcha spp.) in the lower part of the interval and brackish-marine (Cytheropteron latissimum and Sarsicytheridea bradii) in the upper part suggests increasing salinity uphole.

Site M0067

Site M0067 contained ostracods in the interval between 0.2 and 4.2 mbsf. Significant variations in abundance were observed. The highest abundances were recorded at ~3.2–3.5 mbsf, reaching 50–130 valves per 20 cm3 sample. All taxa identified from this site are shallow-water marine, with the following species being the most abundant: E. concinna, S. punctillata, S. bradii, and Robertsonites tuberculatus. A relatively low ratio of juvenile valves suggests a high-energy environment where some redeposition took place.

Sites M0062, M0064, and M0066

Sites M0062, M0064, and M0066 were barren with respect to ostracods.

Geochemistry

For the analysis of interstitial waters, a total of 789 samples were collected offshore using either Rhizons or squeezers. An additional 876 sediment samples were collected for analysis of headspace gas, and in 161 of these samples, methane was analyzed for all four microbiology sites during the offshore phase. Pore water analyses onboard the ship included measurements of salinity, pH, sulfide, alkalinity, and ammonium. An additional 23 chemical species in the pore water were measured during the OSP. The measurements include analyses of chloride, bromide, and sulfate by ion chromatography and aluminum, barium, boron, calcium, iron, magnesium, manganese, potassium, phosphorus, silica, sodium, strontium, sulfur, titanium, lithium, molybdenum, rubidium, vanadium, zinc, and zirconium by inductively coupled plasma–optical emission spectrometry (ICP-OES). We also analyzed 657 sediment samples for total carbon, total organic carbon, and total sulfur using a carbon sulfur analyzer at the University of Bremen.

The pore water composition at most sites reflects the rise in salinity associated with the transition from a freshwater environment to the modern brackish-marine Baltic Sea. The length of the sediment column impacted by the salinity change in the bottom water strongly depends on the sediment accumulation rate and diffusion. At Site M0059 in the Little Belt, for example, the brackish-marine sediment deposit is exceptionally thick at ~47 m. Concentrations of chloride, a conservative element in seawater, are high and relatively constant at ~380 mM throughout the upper 15 m of the sediment, before declining to 100 mM at depth. In contrast, at Site M0066 in the Bornholm Basin, the brackish-marine deposit is <2 m thick. Here, chloride concentrations display a comparatively moderate decline from near 170 mM near the seafloor, largely reflecting diffusive transport into the sediment. Evidence for fresher water at depth at Site M0060 may reflect flow through high-porosity sand units.

Differences in pore water salinity also reflect the position of each site within the BSB. Sites located a greater distance from the Danish Straits and the connection with the North Sea are generally characterized by a lower salinity in the pore water in the upper part of the sediment column. Thus, although the surface sediment salinity is ~33 at Site M0060 near Anholt, salinities of ~6 were observed at Sites M0061 and M0062 in the Ångermanälven estuary.

The chemical composition of the interstitial waters also records large spatial and temporal differences in the input and degradation of organic matter in the sediment. Key indicators for high rates of organic matter degradation include high alkalinities, elevated concentrations of ammonium and phosphate, and lack of sulfate in the pore water. Exceptionally high alkalinities of ~200 and 55 mEq/L are observed at Sites M0059 and M0063, respectively. Alkalinities are at least an order of magnitude lower at most other sites, indicating lower rates of organic matter degradation.

Methane was detected in the sediment at all microbiology sites (i.e., Sites M0059, M0060, M0063, and M0065), indicating an important role for methanogenesis in the degradation of organic matter at depth in the sediments. However, because of extensive degassing during coring, most of the values indicate presence or absence of methane only and cannot be used as a quantitative measure. Sediment total organic carbon and total sulfur profiles support the spatial and temporal variations in organic matter input and degradation as deduced from pore water profiles.

Physical properties

Physical property data obtained during the offshore, pre-onshore, and onshore phases of Expedition 347 vary in quality at each drill site, but several parameters are of constantly high quality and generally reflect the different lithologic and in many cases seismic units identified within each hole (see “Lithostratigraphy”). Natural gamma ray (NGR) data obtained during the pre-onshore phase were of particular importance for preliminary interpretations and predominantly reflect variations in clay, water, and organic content at all sites. Noncontact resistivity, magnetic susceptibility (MS), and gamma ray density obtained with the shipboard multisensor core logger (MSCL) also commonly aided downhole interpretation. At a number of sites, bulk density (i.e., wet density) obtained during onshore analyses corresponded very well to the shipboard MSCL gamma ray density data. Shipboard fast track MS, together with “normal” MSCL MS, allowed construction of composite stratigraphic splices at most sites (see “Stratigraphic correlation”). These splices proved to be very useful, especially for microbiology holes/cores that were heavily whole-core subsampled onboard. Interpreted together, the above physical property parameters also reflect the disturbance common in the uppermost intervals of cores.

Shipboard MSCL P-wave velocity and onshore color reflectance data exhibit greater variation in quality and were of limited utility at most sites. Problems with these two parameters may result from coring and postcoring disturbance, such as cracking due to expansion and undersized cores. Additional processing of the color reflectance data, aided by examination of line scanner images, is likely necessary to identify and remove intervals characterized by spurious values.

Downhole logging

Downhole geophysical logs provide continuous information on physical, chemical, textural, and structural properties of geological formations penetrated by a borehole. Offshore downhole logging operations for Expedition 347 were provided by Weatherford Wireline Service and managed by the European Petrophysics Consortium (EPC). The set of downhole geophysical instruments utilized during Expedition 347 was constrained by the scientific objectives, the coring technique, and the hole conditions at the nine sites. The suite of downhole geophysical methods was chosen to obtain high-resolution resistivity images of the borehole wall, to measure borehole size, and to measure or derive petrophysical or geochemical properties of the formation such as porosity, electrical resistivity, acoustic velocities, and natural gamma radioactivity. No nuclear tools were deployed during Expedition 347.

The Weatherford Compact suite comprised the following tools:

  • The gamma ray tool (MCG) measures natural gamma radiation.
  • The spectral gamma ray tool (SGS) allows identification of individual elements that emit gamma rays (e.g., potassium, uranium, and thorium).
  • The array induction tool (MAI) measures electrical conductivity of the geological formation. The output of the tool comprises three logs: induction electrical conductivity of shallow, medium, and deep investigation depth.
  • The sonic sonde (MSS) measures formation compressional slowness (inverse velocity).
  • The Compact microimager (CMI) is a memory capable resistivity microimaging tool. The eight arms in two planes of the CMI also provide four independent radii measurements, which can be used to identify near-borehole stress regimes.

A total of nine boreholes (Holes M0059B, M0059E, M00060B, M0062D, M0063A, M0064A, M0064D, M0065A, and M0065C) were prepared for downhole logging measurements. Measurements were performed in open borehole conditions (no casing). Despite difficult borehole conditions (nonconsolidated formations, risk of collapse, etc.), the recovery and overall quality of the downhole logging data is good. The gamma ray tool was used in the top of every tool string as a communication tool and for correlation between the different runs. The choice of tool string was based on borehole conditions and drilled depth. Because of borehole conditions, it was not possible to log with all tools in every borehole or to reach total drilled depth in most of the holes. Natural gamma ray measurements performed on unsplit cores will enable core-log correlation. Comparison of downhole logging units with petrophysical properties boundaries and lithologic unit boundaries showed generally good correlation.

Holes M0059B and M0059E

Hole M0059B was drilled to 204.03 m drilling depth below seafloor (DSF). In preparation for logging, the hole was circulated and the drill string was pulled back in the hole to 20 m DSF. The MCG/MAI was the first tool sting to be run, and it reached 72.5 m wireline log depth below seafloor (WSF) from where an uplog was started. Then the drill pipe was run again to 204 m DSF and set to 88.5 m DSF, and the MCG/MAI tool string was deployed again in order to log the bottom part of the hole. The tool string reached 183.5 m WSF from where an uplog was started. Pulling out of the hole, an overpull was observed and the decision was taken to set the pipe again to ~20 m DSF and deploy the MCG/SGS/MSS tool string. This tool string was run to 73.5 m WSF, and an uplog was started.

Hole M0059E was drilled from seafloor to 100.8 m DSF. In preparation for logging, the hole was circulated and the drill string was pulled back in the hole to 15 m DSF. The first tool string was the MCG/MAI tool string, which reached ~70 m WSF from where an uplog was performed. The second tool string was the MCG/CMI tool string, which reached 60 m WSF. The last tool string, MCG/SGS/MSS, with total gamma, spectral gamma, and sonic, reached the same depth from where an uplog was started.

Hole M0060B

Hole M0060B was drilled from seafloor to 86.2 m DSF. In preparation for logging, the hole was circulated and the drill string was pulled back in the hole to 18.5 m DSF. The first tool string deployed included the gamma and induction tools, followed by the gamma, spectral gamma, and sonic tool string. Both tool strings reached ~65 mbsf from where an uplog was started. The last tool string deployed was the gamma and the Compact microimager. A maximum depth of 56 mbsf was reached.

Hole M0062D

Downhole logging measurements in Hole M0062D were made after completion of coring to a total depth of 21 m DSF. In preparation for logging, the hole was circulated with sea water and the pipe was pulled back to 2 m DSF. Two tool strings were deployed. The MCG/MAI was run from seafloor to 9.5 m WSF from where an uplog was started. The MCG/SGS tool string, measuring total gamma ray and spectral gamma ray, was run from seafloor to 9.5 m WSF, and an uplog was started.

Hole M0063A

Hole M0063A was drilled from seafloor to 115.8 m DSF. In preparation for logging, the hole was circulated with sea water and the drill string was pulled back in the hole to 18.65 m DSF. Three tool strings were deployed. The MCG/MAI was run from seafloor to 108.5 m WSF from where an upload was started. The MCG/SGS/MSS tool string was run from seafloor to 108.5 m WSF, and an uplog was started. The MCG/CMI tool string was run down to the same depth, and a high resolution uplog was started.

Holes M0064A and M0064D

Hole M0064A was drilled from seafloor to 41.5 m DSF. In preparation for logging, the hole was circulated with sea water and the drill string was pulled back in the hole to 2 m DSF. The MCG-MAI tool string was run into hole, and a downlog was started. The seafloor was picked up by the gamma ray, and the tool string came out of the pipe. Immediately, a sudden drop in tension indicated that the tool string set up at ~5 m WSF. Logging operation was abandoned after this unsuccessful attempt.

Downhole logging measurements in Hole M0064D were made after completion of coring to a total depth of 41.2 m DSF. In preparation for logging, the hole was circulated with sea water and the pipe was pulled back to 9 m DSF. Two tool strings were deployed. The MCG/MAI was run from seafloor to 31 m WSF, and the MCG/SGS tool string was run from seafloor to 23 m WSF.

Holes M0065A and M0065C

Hole M0065A was drilled from seafloor to 73.9 m DSF. In preparation for logging, the hole was circulated with sea water and the drill string was pulled back in the hole to 15.11 m DSF. Two tool strings were deployed. The MCG/MAI was run from seafloor to 42 m WSF, and the MCG/SGS tool string was run from seafloor to 16 m WSF.

Hole M0065C was drilled from seafloor to 47.9 m DSF. In preparation for logging, the hole was circulated with sea water and the drill string was pulled back in the hole to 13.7 m DSF. Two tool strings were deployed. The MCG/SGS/MSS was run from seafloor to 40 m WSF, and the MCG/CMI tool string was run from seafloor to the same depth.

Paleomagnetism

The magnetic susceptibility (χ) and natural remanent magnetization (NRM) of a total of 1779 discrete cubic (volume = 7.6 cm3 or 1 cm3) samples were measured during the OSP. Sixteen U-channels (~1.5 m long) were taken from cores recovered at Sites M0061 and M0062 to complement the discrete sample data. Forty-three minicubes (volume = 1 cm3) were included in the suite of paleomagnetic samples from Site M0066, where the restricted diameter of the split cores did not allow for standard paleomagnetic boxes to be taken in addition to the requested U-channels.

The range of χ, which was normalized to the wet mass of each discrete sample, spans across four orders of magnitude and included one negative sample from the Cretaceous limestone recovered from 204 meters composite depth (mcd) at Site M0060. The highest magnetic susceptibilities of 8 × 10–6 m3/kg (discounting samples that showed signs of contamination) were displayed by samples from a unit of well-sorted sand at 122 mcd at Site M0060. In general, χ was less than 1 × 10–6 m3/kg, with values less than 0.4 × 10–6 m3/kg restricted to the relatively organic rich postglacial brackish and marine sediments that characterize the Littorina phase of the Baltic Sea. Units of varved clays, including those with silty and sandy components characteristic of the BIL stage of the Baltic Sea, have intermediate to high χ values between 0.4 × 10–6 and 1 × 10–6 m3/kg. It was noted that enhanced χ values were frequently associated with lithologic boundaries considered to reflect the transition from the Ancylus Lake to the Littorina Sea and were frequently connected to observations of iron sulfide precipitates.

A viscous component of the NRM in the vast majority of discrete paleomagnetic pilot samples was effectively randomized by the application of an alternating field of 5 mT. The subsequently determined inclinations and declinations of all the discrete samples are shown projected as polar (stereoplots) in Figure F32. Cores were not oriented with respect to an azimuth, and the relatively few number of sample points per core section limits the usefulness of the declination data. This data set shows a distinct cluster of data points that vary a few degrees from the predictions of a geocentric axial dipole (GAD) model. There are, however, many samples with shallow and reversed inclinations, and these were almost entirely restricted to coarse-grained units (including some varved clays with silt) that were deposited during early stages of Baltic Sea development (Baltic Ice Lake, Yoldia Sea, and Ancylus Lake). No convincing evidence of known late Quaternary geomagnetic field excursions are present in the data set, and the shallow and reversed directions are ascribed to deposition in high-energy environments. Paleomagnetic samples recovered from units that were formed during the Littorina phase of the Baltic Sea (approximately the last 6000–8000 y) show no sign of mechanically induced inclination shallowing. Core disturbance caused by gas (methane) expansion physically disturbed the between-sample fidelity of the paleomagnetic record in the Littorina sediments recovered from Landsort Deep (Site M0063), although it was noted that the NRM intensity of laminated units of Littorina age was higher than nonlaminated units. A biplot of χ versus NRM intensity is shown in Figure F33. Evidence of the most distinct regional feature in Holocene paleomagnetic secular variation, which occurred ~2600 y ago and is defined by a peak in inclination and a relative east to west swing in declination, was identified at 5 mbsf at Site M0062, in a varved sequence from Ångermanälven.

Microbiology

Site M0059

Two holes, M0059C and M0059E, were drilled for microbiology, interstitial water chemistry, and ephemeral geochemical parameters. Counts of microbial cells were made by fluorescence microscopy using the acridine orange DNA stain direct count (AODC) method and by flow cytometry (FCM) using SYBR Green DNA stain. Microbial cell abundances in the deep Holocene clay are very high and decrease gradually with depth, from ≥109 cells/cm3 near the sediment surface to ~108 cells/cm3 at 50 mbsf. The high cell counts are in accordance with the high rates of organic matter mineralization, as indicated by extraordinary high alkalinity and ammonium concentrations. Interestingly, there is no abrupt change in cell abundance between the Holocene sequence and the underlying glacial clay in spite of a strong shift in organic carbon content and, expectedly, in the availability of organic substrates and energy. Whether there is a shift in the phylogenetic or functional composition of the microbial community may be revealed by later DNA-based analyses.

Cell counts made by the two approaches of fluorescence DNA staining, using microscopy or FCM, yielded very similar data with a 1:1 relationship and no significant difference by a paired sample t-test. This is the first successful offshore comparison between cell counts by FCM and the standard AODC, and the result is therefore also important for future IODP deep biosphere expeditions. The results show that the techniques used for FCM quantitatively extracted and identified microbial cells.

Perfluorocarbon (PFC) tracer was used while drilling the microbiology holes in order to evaluate potential contamination of microbiology samples with cells from the drilling fluid. There were initial problems with the delivery of the PFC to Hole M0059C that were partly resolved when drilling the second microbiology hole toward the end of the expedition. Furthermore, the PFC contamination test was significantly improved between the first and second hole after the drilling fluid samples were taken directly from the core liner upon retrieval of each piston core. The average contamination level corresponds to the potential introduction of 10–100 cells/cm3 of sediment. In comparison to the in situ cell abundance of 108–109 cells/cm3, this is still less than a millionth of the indigenous community.

Site M0060

Hole M0060B was drilled for microbiology, interstitial water chemistry, and ephemeral geochemical parameters. Counts of microbial cells were made by fluorescence microscopy using the AODC method and by FCM using SYBR Green DNA stain. A paired sample t-test shows that the cell numbers estimated by the two methods do not differ significantly when applied to samples from the same core depths. Cell counts are relatively low in the upper 6 m of well-sorted sand. In the deep sequence of glacial clay below the sand, cell counts are high, 108–109 cells/cm3. Although a broad peak of alkalinity and ammonium indicates a higher rate of organic carbon mineralization in the upper 0–30 mbsf than below, cell counts do not change significantly with depth to ~85 mbsf, where the clay shifts to sand and where uncontaminated sediment could no longer be sampled for microbiology.

PFC contamination tests show a high level of potential contamination in the uppermost sandy cores (347-M0060B-1H through 3H), where PFC concentrations are as high in the interior of the core as on the exterior. In the glacial clay, the calculated potential contamination ranges from 102–103 cells/cm3 in some cores to <1 cell/cm3 in others. Compared to the in situ cell abundance of 108–109 cells/cm3, contamination is thus <1:105 of the indigenous cells.

Site M0063

Hole M0063E was drilled for microbiology, interstitial water chemistry, and ephemeral geochemical parameters. Microbial cells were counted by fluorescence microscopy using the AODC method and by FCM using SYBR Green DNA stain. In accordance with high rates of organic matter mineralization, as indicated by a broad peak of alkalinity and ammonium concentration, fluorescence microscopy yielded very high cell counts, approaching 1010 cells/cm3 near the sediment surface and decreasing steeply to ~108 cells/cm3 at ~40 mbsf. From this depth, which marks the transition from Holocene, organic-rich clay to glacial, low-organic clay, cell counts decrease with depth at a significantly lower rate.

In contrast to Sites M0059 and M0060, FCM counts are similar to AODC counts only in the lower, organic-poor clay below 41 mbsf at Site M0063, whereas FCM counts are significantly lower than the AODC profile in the overlying Holocene, organic-rich sediment, with samples taken near the sediment surface showing the largest deviation (>10-fold). This discrepancy may derive from the difficulty of counting individual cells in samples from this interval because of the presence of large clumps of cells and very small cells (~0.1 µm in diameter), which escape detection by FCM. An improved sample preparation technique will need to be developed to deal with these samples.

Based on PFC data, Cores 347-M0063E-3H, 4H, 30H, 31H, and 42H show the highest calculated contamination levels, with potentially as many as 104 drilling fluid-derived cells/cm3 in their interiors. On the other hand, Cores 1H, 2H, and 17H have no detectable PFC in the interior, and Cores 5H, 10H, 12H, 25H, and 41H have only moderate contamination (<100 cells/cm3) and are therefore suitable for microbiological analyses. No clear depth- or lithology-related trend in the level of contamination could be observed.

Site M0065

Hole M0065C was drilled for microbiology, interstitial water chemistry, and ephemeral geochemical parameters. Microbial cells were counted by fluorescence microscopy using the AODC method and by FCM using SYBR Green DNA stain.

Cell densities obtained by AODC are among the highest observed in all marine sediments examined by scientific drilling, with a sample at 3.53 mbsf yielding a cell count of 1.23 × 1010 cells/cm3. Similar to Site M0063, AODC counts decrease steeply through the upper 14 m of marine, high-organic sediment. Below this depth, which marks the transition to late glacial clay, cell counts decrease at a significantly lower rate. High cell densities in the organic-rich sediments above 14 mbsf coincide with the interval of high degradation rate of organic matter, as indicated by a broad maximum in alkalinity and ammonium concentrations at this depth interval. Salinity decreases linearly to ~30 mbsf and therefore does not appear to be directly related to the cell profile.

FCM counts are significantly lower than those of AODC in the upper, high-organic section of Hole M0065C. In this same interval, and similar to what was observed at Site M0063, the occurrence of large clumps of cells made individual cell counting problematic, likely resulting in underestimation of the real microbial densities. No significant deviation between FCM and AODC counts was observed in samples from the glacial clay below 14 mbsf.

PFC contamination tests indicate that most cores bear very limited contamination in their interiors (<100 contaminating cells/cm3), with Cores 347-M0065C-3H, 6H, and 7H potentially having <10 contaminating cells/cm3 in their inner part. Compared to the other sites, M0065 has the highest fraction of cores that are suitable for microbiological analyses.

Stratigraphic correlation

Site M0059

At Site M0059, it was possible to correlate five holes and create one continuous splice from 0 to ~87 meters composite depth (mcd; end of glacial clays). Comparison of seismic boundaries (calculated with a simple velocity-depth equation, using seismic velocity values measured onshore from each unit) with lithologic unit boundaries and physical property boundaries showed good results. Below 87 mcd, sediment recovery was sporadic, and splicing beyond that point was not possible.

Site M0060

At Site M0060, sandy intervals and noncontinuous recovery presented some difficulties in correlation. From seismic images, it was possible to find good matches for lithologic unit boundaries, using a simple velocity-depth equation and discrete velocity values measured for each unit. Lithologic unit boundaries were also easily identifiable from physical parameters data. As Hole M0060B was a microbiology hole, it was not possible to construct a splice on this site, as most of the material from the second hole was consumed by microbiological sampling and recovery was from one hole only below ~90 mcd.

Sites M0061 and M0062

Site M0061 provided material for good correlation to ~26 mcd, where rhythmically variable silt and clay changed into sand. Within the sand unit, reliability of correlation was difficult to check. Data integration suggests a possible unconformity/erosion approximately between a sulfide-rich clay unit and clay and silt rhythmite units at both sites. Good correspondence between two-way traveltime values was calculated for lithologic unit boundaries, physical properties, and features in seismic data. It was possible to construct an almost continuous splice to ~35 mcd. Site M0062 was very similar to Site M0061.

Site M0063

Site M0063 provided many challenges to correlation. Similar larger scale trends are visible in all physical properties in all holes (microbiology Hole M0063E was an exception, as only fast track magnetic susceptibility data were available from the entire length). Hole-to-hole correlation remains at a relatively approximate level (0.5–1.0 m) because of large nonlinear sediment expansion and coring disturbances. Sediment expansion (uppermost 40 mcd) made it difficult to precisely align prominent lithologic features between adjacent holes. Strong noise in the physical property data in each core top (coring disturbances) had to be cleaned. Based on lithologic information (see “Lithostratigraphy”) and seismic profiles, it is also possible that some type of gravity flows have occurred in the area, deforming sediment. A comparison of NGR data and downhole log gamma data suggests that despite sediment expansion and coring disturbances, combined sediment record recovery from Holes M0063A, M0063C, and M0063D was nearly continuous.

Site M0064

Site M0064 provided some variation in soft-sediment unit thickness. Correlation between holes was possible because lithologic changes (different types of clay sediments and diamicton) were clearly visible in the physical property data and slabbed core scan images. However, within the diamicton unit, correlation between holes could not be checked; therefore, splicing was restricted to the uppermost ~10 mcd.

Site M0065

At Site M0065, correlation and splicing was possible to ~48 mcd. Sediment recovery was nearly continuous, with some small gaps between 19 and 35 mcd. Seismic units could be connected well to major lithologic changes, although some finer unit boundary details were difficult to detect.

Site M0066

Correlation for Site M0066 was relatively good for the uppermost ~2–15 mcd, and it was possible to construct a splice. Below that point, sediment recovery was poor, and it was not possible to connect holes accurately with any physical parameters or a lithologic feature. Lithologic unit boundaries could be detected from the seismic image fairly well.

Site M0067

Two holes at Site M0067 were very shallow, with a thin organic-rich sediment blanket covering sandy and gravelly, and a possible diamicton-like material. Correspondence with seismic units was relatively good, although no measured velocity data were available from the diamicton unit. Correlation and splicing were possible only for the uppermost 3.10 mcd.