Project Title: "Developing and Applying a New In Situ Technology for the Investigation of Episodic Contaminant Transport Events Within Estuaries"

Project Coordinator:

John W. King, University of Rhode Island

Additional Principal Investigators:
Alfred K. Hanson, Jr., SubChem Systems, Inc.
Christopher R. Kincaid, University of Rhode Island
Elizabeth Lacey, University of Rhode Island
James G. Quinn, University of Rhode Island

Project Duration: 7/1999-6/2001

Project Location: Narragansett Bay NERR including regions adjacent to the NERR site.

Abstract:

Seasonal and episodic events such as tidal and wind driven currents, rainstorms, channel dredging, discharges from sewage treatment plants, and ship activities may lead to the resuspension of sediments and significant changes in the concentration of dissolved oxygen, chemical contaminants and nutrients in coastal marine waters. We are using a combination of new and traditional technologies to characterize the water and sediment quality in the Narragansett Bay National Estuarine Research Reserve (NB-NEER). New technology used for this project includes the combination of in situ chemical analysis (dissolved oxygen, iron, chlorophyll), hydrographic (salinity, temperature, density) and current velocity and direction measurements. Traditional technology used for this project includes the collection of surficial sediments and analysis for total metals, sediment grain size, nutrients and Simultaneously Extracted Metals-Acid Volatile Sulfide, and analysis of suspended particulate matter captured in sediment traps. Over the past six months we have collected and analyzed surficial sediments and sediment trap material from three sampling rounds to give an indication of natural seasonal variability associated with the NB-NEER. In addition, we have extended the analytical capabilities of the SubChemPak Analyzer TM, conducted laboratory tests to evaluate WET Labs new ChemStar detector, and conducted initial field tests with the XZ-Profiler TM. These new oceanographic instruments and platform will be further tested and evaluated during a series of field tests in Narragansett Bay during the late summer and early fall.

Description of Objectives:

The objectives of the study are two-fold. First, we will combine newly-developed and existing technologies in an innovative manner to document the effects of natural and anthropogenic episodic events on the input, transport and fate of nutrients, oxygen, and organic and inorganic contaminants in an urban estuary. Second we will apply our findings to improve existing risk-characterization schemes for urban estuaries. During the first year of funding, we are concentrating on the first objective. The findings in this progress report focus on the characterization of natural seasonal cycles.

Description of Field and Analytical Methods:

Three rounds of sampling have been performed at stations around the NB-NERR area (Figure 1, see fig1.jpg). During round 1, November-December 1999, surface sediment samples were collected and sediment traps deployed from Stations 1-8. In round 2, March-April 2000, sediment was collected from Stations 1-11 and sediment traps at Station 2, 4 and 5. Between sampling rounds 1 and 2, ice covered the upper portion of Narragansett Bay and several of the floats for the sediment traps were frozen in the ice and then transported with the ice as it moved. New sediment traps were deployed at all of the stations during round 2. In round 3, June 2000, surface sediments were collected and sediment traps deployed at Stations 1-11, and sediment traps were retrieved from Stations 2-11. The sediment trap at Station 1 was not present.

Surface samples were collected using a grab sampler. The top 2 cm were subsampled in the field using teflon coated titanium tools and stored in cleaned containers. During round 1, mini cores (up to 13 cm in length) were collected from a subset of stations and then subsampled at 2 cm increments down the core. All of the collected surficial and sediment trap sediments were analyzed for SEM-AVS, total trace metal concentration, grain size, and nutrient concentrations and the subsampled mini cores were analyzed for total trace metal concentrations, nutrients and sediment grain size. Selected sediment trap samples were analyzed for organic contaminants.

The flux of sediment to the sediment traps (the accumulation rate) was calculated by dividing the total dry weight of the sediment in the trap by the product of surface area of the trap and the length of time the trap was in the water (i.e., g/cm2/day).

The sediments were prepared for trace metal analysis using a total digestion technique (dried sediment treated with concentrated hydrochloric, hydrofluoric and nitric acids, placed in a heated sonicator and neutralized with boric acid), and were analyzed for trace metals (Cd, Cu, Pb, Ni, Zn) using graphite furnace atomic absorption spectroscopy (GFAAS).

Non-carbonate, non-organic grain size was determined by filtering the sediment through a 63 µm sieve and then analyzing the less than 63 µm sediment with an Elzone Model 180XY particle size analyzer.

Simultaneously extracted metal/acid volatile sulfide (SEM-AVS) analysis was performed using the purge and trap technique. AVS was measured using a sulfide selective electrode. Filtered SEMs were analyzed for Cd, Cu, Pb, Ni, Zn by GFAAS.

The sediment samples were analyzed for carbon and nitrogen by placing dried, homogenized sediment (approximately 10 mg) into pre-cleaned tin boats. The samples were analyzed using a Carlo-Erba CHN analyzer and concentrations were calculated using an acetanilide standard curve.

The organic contaminants that were analyzed included: total petroleum hydrocarbons (TPHs), polycyclic aromatic hydrocarbons (PAHs), polychlorinated biphenyl congeners (PCBs), organochlorine pesticides (OCPs) and several other organic components (Quinn et al., 1992). For the current investigation, we analyzed most of the analytes on the list of organic contaminants recommended by the NOAA Status and Trends Program for estuarine and coastal monitoring. These contaminants include 24 PAHs, 24 PCB congeners and 13 OCPs . In addition, we analyzed samples for TPHs (m/z 55) and two major benzotriazoles (BZTs). The sum of the 24 PCB congeners times two is equal to the &Mac183;PCBs (equivalent to total Aroclors such as the &Mac183; of Ar1242, Ar1254, and Ar1260), and the sum of the 24 PAHs is the &Mac183;PAHs. In addition, the sum of the 5 DDT compounds reported in this study is the &Mac183;DDTs and the sum of the C10-BZT and chloro-BZT is the &Mac183;BZTs. All concentrations are reported on a dry weight basis. Detailed information on the sample storage, analytical methods, and quality control procedures used in this study have been published by Quinn et al., 1992; Latimer and Quinn, 1996; and Reddy and Quinn, 1998).

Summary of Findings:

Accumulation Rate

The accumulation rate between the first and second sampling round ranged from .00066 g/cm2/day (station 4) to 0.00800 g/cm2/day (station 5). The average accumulation rate (based on stations 2, 4 and 5) was 0.00451 g/cm2/day. For the second round, the average accumulation rate was much higher, 0.02856 g/cm2/day. As indicated earlier, there was ice covering the northern portion of the bay between sampling rounds 2 and 3, this likely led to the observed lower accumulation rate during that time period. A graphical representation of the accumulation rate at the various stations is presented in Figure 2 (see fig2.jpg).

Note the comparatively low accumulation rates in the northern stations and the comparatively high accumulation rates for the southern stations. For the three stations where there is two sets of data (stations 2, 4 and 5), the accumulation rate for stations 2 and 4 is roughly similar between rounds, however, station 5 had a significantly higher accumulation rate between rounds 2 and 3. Historical data from 1989 (King et al, 1992) for the Ohio ledge area (station 2) indicated an accumulation rate of 0.00414 g/cm2/day which is very similar to the average accumulation rate for this location during the current study (0.00478 g/cm2/day).

Grain Size

The majority of the surficial sediment was silt (Figure 3a, see fig3a.jpg). Coarser (sand-sized) sediment was present at stations 1, 7 and 10. Stations 1 and 10 likely have rapid bottom currents and station 7 had a significant component of shells. Figure 3a (see fig3a.jpg) shows a breakdown (% sand, % silt, % clay) of the surficial sediment for all three rounds. By comparison, the material in the sediment traps (the suspended material) shows all stations with a dominant silt component (Figure 3b, see fig3b.jpg).

Figure 3a & b

Trace Metals

In general, there are higher trace metal (Cd, Cu, Pb, Zn) readings at station 6, located in the Providence River, and lower trace metals down the Bay. However, station 4, located in Potter’s Cove at Prudence Island, also had comparatively high trace metal concentrations (Figure 4, see fig4.jpg).

The horizontal dashed lines on Figure 4 are the values for the Effects Range-Low (ER-L) (large dashed line) and Effects Range-Median (ER-M) (finer dashed line) for each trace metal. Effects range values have been calculated to determine the potential for impact to the biota. Sediment with values below the ER-L likely does not have an impact on the biota. Sediment with values above the ER-M likely does impact the biota. Sediment with trace metal values between the ER-L and the ER-M may impact the biota. With the exception of copper and zinc at station 6 during round 1, all of the values are below the ER-M. The sediment trap material (reflecting current inputs and resuspended material) had lower trace metal concentrations than the surface sediments at all the stations (Figure 5, see fig5.jpg).

A comparison of trace metal concentrations in sediment from sediment trap samples that were collected in 1989 (King et al., 1995) and the sediment from rounds 2 and 3 was performed. Two station locations were roughly comparable, station 7 was located near the 1989 Conimicut Point location, and station 2 was located near the 1989 Ohio Ledge location. The further north station (Conimicut Point/station 7) showed a decrease in trace metal concentrations between the 1989 and 2000 sampling events, whereas the Ohio Ledge/station 2 location showed an increase in trace metal concentrations (Figure 6, see fig6.jpg). Comparatively high AVS concentrations were present at stations 3, 4, 6 and 8. Figure 7 (see fig7.jpg) shows the relative AVS concentrations for the three sampling rounds.

Figure 6

Figure 7

The SEM-AVS data shows some seasonality to the trace metal bioavailability. The values for SEM-AVS were positive (indicating potential bioavailability) for only the third round (spring collection) at stations 2, 5, 10 and 11 (Figure 8, see fig8.jpg). The value for SEM-AVS during the spring collection was 1.80 at station 2, 0.22 at station 5, 1.06 at station 10 and 0.86 at station 11. It is important to note that shell fishermen were present in the area of station 2 (Ohio Ledge).

Nutrients

With the exception of station 7 which had very high carbon values (likely due to the plethora of shells that were present at this station), the carbon and nitrogen values correlated well with each other. The carbon to nitrogen ratios (which may be used as an indicator of contamination) were roughly similar (Figure 9, see fig9.jpg).

Organic contaminants

Ten samples of trap material, collected in March/April 2000 (round 2) and June 2000 (round 3) were analyzed for organic contaminants. These results are shown in Table 1 (see table1.xls) along with results from the analyses of three frozen trap samples collected in March/April 1989 and analyzed with the current samples.

The stations are arranged with decreasing distance from the Providence River to the mouth of the bay; thus, station 6 is at Sabin Point and station 10 is at the southern end of Prudence Island (Station 5 is near Mount Hope Point in Mount Hope Bay). In all cases, there is a general trend of decreasing concentration from station 6 to station 10 (Figure 10, see fig10.jpg). Exceptions to this trend include station 4 where higher values are sometimes found at this location in Potter Cove, Prudence Island. The trace metal analytical results showed a similar trend with high values in Potter Cove. Also, for some unexplained reason, the benzotriazoles (&Mac183;BZTs) show maximum values at station 2 (Ohio Ledge). The agreement between sampling dates in rounds 2 and 3 is fairly good for the &Mac183;PCBs, &Mac183;DDTs, &Mac183;chlordanes, &Mac183;PAHs and TPHs (Figure 10, see fig10.jpg), suggesting that a common and constant source (probably resuspended surface sediment) is contributing to the trap samples. However, the agreement for the &Mac183;BZTs is rather poor with the round 2 BZTs being considerably lower than the round 3 samples.

For comparison, values from the 1989 samples are generally high compared to the recent samples (Table 1). (The Fox point station is at the head of the Providence River and north of station 6, the Pawtuxet River sample location is closest to station 6 and the Conimicut Point sample location is between stations 6 and 3). These latter two stations show somewhat lower concentrations in the samples from rounds 2 and 3 for the &Mac183;PCBs, &Mac183;DDTs, &Mac183;chlordanes, &Mac183;PAHs, and TPHs. Surprisingly, the &Mac183;BZT values in the 1989 samples are lower than several stations from the recent samples. Based on sediment concentrations, there has been a significant reduction in BZT values in Narragansett Bay surface sediments between 1989 and 2000. Thus, the low values in the 1989 trap samples may reflect degradation of these components in the frozen samples over the 11 year storage period. Values for these compounds in 1989 samples analyzed at that time were over 1000 ng/g.

Instrumentation Development Status

During the past six months SubChem Systems, Inc. R&D effort on this project primarily involved:

1. Continued development, improvement and extension of the analytical capabilities of the SubChemPak AnalyzerTM,
   
2. Laboratory testing and evaluation of WET Labs new ChemStarTM detector,
   
3. Component acquisition, set up and initial field testing of the XZ-ProfilerTM.
   

The SubChemPak AnalyzerTM is a new multi-channel in situ chemical analyzer that has been developed by SubChem Systems, Inc. for real-time determination and high-resolution mapping of dissolved chemicals in aquatic waters. The SubChemPak reagent delivery module (Figure 11, see fig11.jpg) transforms underwater optical instruments (WET Labs, Inc.) into sensitive chemical analyzers for instantaneous measurements of nutrients and other environmentally important chemicals. The instrument can be deployed with standard oceanographic electronic profiling packages (i.e. CTD’s) for vertical and/or horizontal profiling (Hanson and Donaghay, 1998; Hanson 1998; 2000). A family of dual-nutrient analyzers is being developed for selected pairs of nutrients: nitrate and nitrite, ammonia and urea, phosphate and silicate, and iron (II) and iron (III). Nutrients are determined by continuous flow analysis with spectrophotometric and fluorometric methodologies that have been optimized for rapid in situ measurements.

The SubChemPak Analyzer has the unique capabilities to:

• Provide high-resolution vertical and horizontal profiles of nutrients in real-time and
  
• Determine nutrients accurately at trace levels (nanomolar to micromolar).
  
  The operation, in situ calibration and data acquisition for the instrument is computer-controlled and user-friendly. The concentration readings for nutrients are instantaneously displayed on the computer monitor. The SubChemPak Analyzer may be co-deployed with standard oceanographic electronic sensor packages (CTD) for vertical and/or horizontal (toyo) profiling.
  
  The SubChemPak Analyzer is comprised of three modular components,
  

1. The SubChemPak reagent delivery module
2. The real-time instrument control and data acquisition system
3. The WET Labs optical detectors.
   

The SubChemPak Analyzer was initially developed and utilized for the determination of nitrite and iron(II) in seawater (Hanson and Donaghay, 1998; Hanson, 2000). During the past few months continuous flow chemical reduction techniques have been optimized so that the oxidized forms of these chemicals, nitrate and iron(III), can also be determined simultaneously with the reduce forms, nitrite and iron(II). The nitrate determination uses heterogeneous reduction with a cadmium column. The iron(III) determination utilizes homogeneous reduction with ascorbic acid. Analytical technologies were also evaluated for ammonia and urea determinations with the SubChemPak.

The ChemStarTM is a new, miniaturized optical detector, developed jointly by WET Labs, Inc. (Philomath, OR) and SubChem Systems, Inc., for spectrophotometric analyzes with the SubChemPak Analyzer. The ChemStar optical cell has considerably smaller dimensions (2.5 cm x 2.5 cm x 18 cm) that the A-Star detectors used in the past. The flow cell has a small bore (1 mm) and a 15 cm path length. The LED light source and detector are physically incorporated into the ends of the detector body. A four-channel prototype ChemStar detector is presently being evaluated for determinations of nitrate, nitrite, iron(II) and iron(III), and urea. A second four-channel ChemStar prototype has been ordered, configured for spectrophotometric determinations of copper, phosphate and silicate with the SubChemPak Analyzer.

The XZ-ProfilerTM is a lightweight instrumentation package that can be towed behind a small boat and used to detect, track and map chemical plumes in coastal waters. The primary components are a four-channel SubChemPak Analyzer with ChemStar detectors, a CTD system (Sea Bird Electronics SBE25, with additional sensors for light irradiance and attenuation, oxygen, pH, and chlorophyll) and an AcrobatTM lightweight tow vehicle (SeaSciences, Inc.). During the past few months the required components were acquired (Acrobat, tow-cable, GPS, CTD, and instrumentation sub frame) and assembled into a functional system. An initial field test with the Acrobat tow-body (using a dummy payload) conducted in the Pettaquamscutt Estuary on August 10th was successful. As deployment experience is gained, additional instrumentation will be added to the payload and tested in the field. The improved SubChemPak Analyzer, new ChemStar detectors and the XZ-Profiler will be further tested and evaluated during a series of field tests in Narragansett Bay during the late summer and early fall.

References

Hanson, A.K. and P.L. Donaghay. 1998. Micro to fine-scale chemical gradients and layers in stratified coastal waters, Oceanography, Vol. 11, No. 1, 10-17.

Hanson, A.K. 1998. A dual-channel profiling chemical analyzer for nutrients and iron(II) in marine waters, AGU Ocean Sciences, San Diego, CA. EOS, Transactions, American Geophysical Union, 79(1): 2.

Hanson, A.K. 2000. A new in situ chemical analyzer for mapping coastal nutrient distributions in real time. OCEANS 2000 MTS/IEEE Conference Proceedings. In press.

King, J., J. Corbin, R. McMaster, J. Quinn, P. Gangemi, D. Cullen, J. Latimer, J. Peck, C. Gibson, J. Boucher, S. Pratt, L. LeBlanc, J. Ellis, M. Pilson. 1995. A Study of the Sediments of Narragansett Bay, Volume I: The Surface Sediments of Narragansett Bay. Final Report Submitted to the Narragansett Bay Project, 201 p.

Latimer, J.S., and Quinn, J.G. 1996. Environ. Sci. Technol. 30, 623.

Quinn, J.G., Latimer, J.S., LeBlanc, L.A., and Ellis, J.T. 1992. Narragansett Bay Project Report Number NBP-92-111, 205 p.

Reddy, C.M. and J.G. Quinn. 1998. GC-MS analysis of total petroleum hydrocarbons and polycyclic aromatic hydrocarbons after the North Cape oil spill, Mar. Poll. Bull.