Sample Collection and Storage
In 1996 the USGS employed the R/V Seaward Explorer (cruise
ID # SEAX96024) to collect 99 sediment cores (20-65 cm in length) from
58 stations along 13 north-south transects in the Long Island Sound (LIS). The SEAX96024 cores were collected
with the USGS hydraulically damped gravity corer (HDGC; Bothner
and others, 1997; Buchholtz ten Brink and others,
2000a). The HDGC is a modified gravity corer that collects 11-cm diameter
cores up to 65 cm in length in clear polycarbonate core barrels. The HDGC provides for excellent preservation of
the sediment-water interface. A video camera mounted on the corer frame
allowed the investigators to verify that the core was taken in a representative
area of the sea floor and that there was not excessive disturbance of
the interface or leakage from the bottom of the core upon recovery. Between
one and six replicate cores were taken at 29 of the stations. On deck
the cores were capped and described. The overlying water was retained
in the core barrel to reduce disturbance of the sediment during transportation
and storage. Cores were stored vertically in a refrigerator at 45 deg
F or in a freezer at -10 to -25 deg F for long-term storage.
X-radiographs of the sediment cores were used to deduce the frequency
and persistence of sediment transport events, layering, lithology and
relative density. Each whole core was x-rayed vertically (see an example of a digitized x-radiograph here) alongside
a centimeter-scale ruler using a Picker x-radiograph machine at Falmouth
Hospital in Falmouth, Massachusetts*. All cores were x-rayed from two directions
(the second direction rotated 90 degrees from the first) to identify sloped
features. Some cores were x-rayed frozen with ice crystals appearing as
fine white lines. The size of the x-radiograph film limited the length
of the core that could be x-rayed to 30 cm per shot. The x-radiographs
were digitized with a flatbed scanner, and composite images for cores
longer than 30 cm were made using commercially available image processing
software. A computer-generated ruler was overlaid on the digitized x-radiographs
using the depths indicated by the ruler x-rayed with the core. X-radiographs
are presented as x-ray negatives; areas of high density (sand, shells,
stones) are bright and areas of low density (mud, voids) are dark.
*Note: X-radiographs were collected using the wall bucky in the thoracic,
spine, and lat mode with density and energy settings of 90 kvP x 2 seconds
x 300 mA x 1.5 master density.
Sixty-six of the 99 cores were selected for sectioning. At least one core from each sample location was sectioned with the exception of sample location A6. When multiple cores had been collected at a sample location, the core selected for sectioning was chosen based on length of core (and presumably length of time represented by the core sediments) and the best preservation of the sediment-water interface and sedimentary features. Cores were sectioned into depth intervals that
ranged between 0.25 and 2 cm in size and processed for further analysis.
Uncut cores were archived in cold-storage facilities.
Selected cores were sectioned in intervals ranging
from 0.25 to 2 cm. Most cores were sectioned within a few months of collection. The sectioning method involved placing the core barrel
in a ring stand and then using a piston jack to extrude sediment from
the barrel (see picture here). An acid-washed titanium knife was used to remove the outer
portion of the sediment (to preclude the effects of smearing along the
barrel during the coring and the extrusion processes). The knife was wiped
clean and then used to put the remaining sediment into a sample jar with
an aliquot of wet sediment removed for Clostridium perfringens (C.
perf.) analysis. C. perf. is an environmentally innocuous
bacterial sewage indicator. As the sectioning was done, the sediment
color (based on Munsell® Soil Color Charts) and other pertinent information
was recorded by the person(s) doing the sectioning.
The water content
of samples was determined for use in chemical calculations and porosity calculations.
Water content was calculated using the equation:
wet sample weight (g) - dry sample weight (g)
wet sample weight (g)
The numbers for this calculation were determined by taking the
weights of each sample before and after freeze-drying. The water content data
were also used to calculate in-situ porosity when no draining occurred.
Water contents measured during the C. perf. analysis were used
to verify the water content values and to correct water content values for selected intervals as necessary. Changed water content values were noted in the comments section of the data sheets.
Sediment Grain Size
Sediment grain size distributions
were measured in order to determine the historical distribution of sediment
texture in the region and to provide data needed for modeling sediment
transport, sediment accumulation rates, and contaminant distributions.
Knowledge of the distribution and movement of fine-grained sediments aids
in understanding the behavior of contaminants in LIS because locations
with high proportions of fine-grained sediments tend to have elevated
concentrations of particle-reactive contaminants (e.g., Ag, Cd, PCBs)
introduced to the environment by human activity (see Mecray
and others, 2000 for examples in surface sediments).
In the laboratory, freeze-dried samples
selected for grain-size analysis were disaggregated and wet sieved to
separate the coarse and fine fractions. The fine fraction (less than 62
µm) was analyzed for phi sizes 5 to 10 by Coulter Counter (McCave
and Syvitski, 1991); the coarse fraction was analyzed for phi sizes
4 and larger by sieving and by use of a rapid sediment analyzer (Schlee,
1966). Bivalve shells and other biogenic carbonate debris were manually
removed from the gravel fraction before analysis. Size classifications
for gravel, sand, silt, and clay were based on the method proposed by
Wentworth (1929), the inclusive graphics
statistical method of Folk (1974), and the
nomenclature proposed by Shepard (1954).
A detailed discussion of the computer processing of the raw textural data
are given in Poppe and others (1998). Salinity and water content data
were used to correct dry sediment mass for the contribution from salt dissolved in pore water. Salinity values were taken from measurements taken during sample collection at each site. Grain
size data are presented in figures as the weight percent of gravel, sand,
silt, and clay. Full phi size distribution are available in the data tables.
Radionuclides and sedimentation rate calculations
Radionuclide data for the calculation of sedimentation rates were collected on a PGT well-type gamma detector. Concentrations of the short-lived radionuclides Cs-137 (661.6 keV) and Pb-210 (46.5 keV) were collected to constrain the recent (30-100 year) history of LIS sediments. Gamma decay data was collected until the measurement error was below 10% or 4-7 days had passed. The data reported are corrected for the radioactive decay that took place between sample collection and sample analysis; for the efficiency of the detector as determined using a known standard; and for the yield, that is the percentage of the radionuclide concentration represented by gamma decay at the measured energy levels. Important years assumed for sediment rate calculations are 1954 (Cs-137 onset with advent of thermonuclear bomb testing), 1963 (Cs-137 peak), and 1970 (peak heavy metals concentrations before passage of Clean Air Act). Pb-210 background levels were calculated for each sample within a core from the average of the efficiency- and yield-corrected Bi-214 and Pb-214 concentrations. The excess Pb-210 was calculated as the difference of the decay-, efficiency-, and yield-corrected Pb-210 and the background Pb-210. No correction was made for gamma ray self-adsorption by the sediment. To calculate Pb-210 rates, all intervals were divided by the Pb-210 concentration at the deepest interval, next the natural log of each interval was calculated, and then a linear best fit (alpha) was calculated from the slope of the Pb-210 values from the sediment surface down to constant background values. The sedimentation rate in cm/yr was calculated by dividing the lambda value (the inverse of the half-life, 0.045 for Pb-210, which has a half-life of 22.5 years) by the alpha value. Sedimentation rates were calculated in units of centimeter per year and were not converted to mass fluxes (grams per square centimeter per year). All sedimentation rates were calculated assuming a constant sedimentation regime. Bioturbation and other sediment mixing was assumed to be minimal. Additionally, the sedimentation model assumes that all sediment was deposited vertically through the water column with no horizontal transportation of sediment though there is some evidence for horizontal transportation in some of the cores. These assumptions are simplistic but allow for a first-order look at sedimentation rates in LIS.
All sampling tools and sample containers were acid washed. The metals analyses were performed at Boston University using an inductively coupled plasma emission spectrometer for most elements and an inductively coupled mass spectrometer to measure Ag, Cd, and Pb. All metals data has been normalized to allow for comparisons between analyses performed on two different instruments over several years. Hg analyses were performed at Wesleyan University. For the purpose of this open-file report, Cu, Hg (where relevant), Pb, and Zn data are presented along with radionuclide data as part of the sediment dating process. More details on the methodology and chemistry data will be forthcoming in a subsequent Open-File Report.