The Gladstone field experiment, Queensland: Weathering of hydrocarbons in mangrove sediment: testing the effects of using dispersants to treat oil spills

Brief description

Experimental plots were established at three sites within mature stands of Rhizophora stylosa (FLNS, FLSN, FLIS), in an area approved for reclamation by the Gladstone Port Authority. The sites were matched visually for tidal elevation, mangrove tree composition and condition, sediment composition and benthic invertebrate populations. Three replicate 6 m² plots were constructed at each site. Around each plot, prop roots were cut in a path about 0.5 m wide to allow installation of plastic retaining walls. The retaining walls were dug into the mud to a depth of 20 cm and supported to a height of 1 m. A gate was installed to allow tidal waters to move in and out while retaining the oil. One of three treatments was applied to each plot: no oil (control); oil; and dispersed oil. Three undisturbed control sites (FLNN, FLSS, CISS), were chosen outside the reclamation area, and plots were delineated by marking tape stretched between the trees only. A litter trap was also suspended in the trees above each plot. Before treatment with oil, initial biological measurements were taken and each plot was sampled for sediment grain size, total organic carbon and background hydrocarbon content. The oil used in this experiment was a medium range crude oil from the Bass Strait Basin in southern Australia and the dispersant used was Corexit 9527. Prior to application to the plots, the oil and the oil plus dispersant mixture was preweathered to simulate an oil spill arriving from seaward. Dispersant was added to the oil to achieve a 1/20 dispersant to oil (v/v) ratio in accordance with industry recommended usage. Treatments were added to the appropriate plots at high tide at a rate of 5 l/m² and pumps were used to distribute the oil evenly throughout each plot as the tide fell.The average rate of dosing in each plot was calculated from surface sediment (0-2 cm) samples collected 40 hours after treatment. Four replicate samples, each comprised of at least 5 pooled surface samples were collected from each plot for analysis. At 1, 7, 13 and 22 months, 4 replicate 10 cm diameter cores were collected from each plot, sliced and the 0-2, 10-12 and 20-22 cm sections from each replicate were pooled. Observations, including the presence of roots, animals, burrows and oil were made before the samples were frozen for later analysis.The following analyses were conducted on the sediment samples: Total Extractable Organic Matter determined gravimetrically; Total oil determined by UV Fluorescence analysis (UVF); Total hydrocarbons determined by gas chromatography with flame ionisation detection (GC-FID); The percentage of unresolved hydrocarbons; Concentrations of individual isoprenoid and n-alkanes (µg/g dry weight); Sum of 218 aromatic and alkylaromatic isomers in the naphthalene/biphenyl, fluorene, phenanthrene/anthracene, benzanthracene/chrysene, fluoranthene/pyrene and benzopyrene to benzoperylene series as determined by Selected Ion Monitoring-Gas Chromatography/Mass Spectroscopy (SIM GC/MS); Ratios of specific alkyl phenanthrene isomers to illustrate selective biodegradation; Sum of triterpane biomarkers in the hopane series as determined by SIM GC/MS using m/z 191; Selected sterane and potential demethylated hopane biomarkers as determined by SIM-GC/MS using m/z 217 and m/z 177; Ratios of specific biomarkers useful as biodegradation indices. This research was a component of the project "Fate and effects of oil and dispersed oil on mangrove ecosystems in Australia" and was undertaken to provide an assessment of immediate and short term impacts of oil spills.The objectives of the hydrocarbon analyses were to determine whether dispersant usage to control an oil spill affects:1. the initial concentration of oil absorbed by sediments2. the depth of penetration of oil in the sediments3. the persistence of oil over time4. its rate of biodegradation5. the pattern of its internal triterpane and sterane biomarkers The study was designed to mimic a catastrophic oil spill approaching the mangroves from the sea. The study site was within several hundred hectares of Rhizophora stylosa forest designated for destruction in an extension of the Port of Gladstone.

Lineage

Maintenance and Update Frequency: notPlanned

Statement: Statement: Analysis of Samples:Subsamples of the sediments collected before oil was applied were analyzed for grain size by shaking dried sediments on a rotary shaker in a set of stainless steel geological sieves and weighing the sediment retained on each sieve. Sieves were 63 µm, 125 µm and 250 µm mesh sizes. Other sub-samples were ground and analyzed for total organic carbon (TOC) content by a high temperature combustion process corrected for inorganic carbonate (Sandstrom et al., 1986).Sandstrom MW, Tirendi F and Nott ALJ (1986) Direct determination of organic carbon in modern reef sediments and calcareous organisms after dissolution of carbonate. Marine Geology 70: 321-329.Sediments for hydrocarbon analysis were defrosted, homogenized and subsampled for dry weight. Approximately 10 g wet sediment was weighed and mixed with 2 to 3 times its weight with anhydrous sodium sulfate (Na2SO4) to bind water. This mixture was placed in a Teflon centrifuge tube and 20 ml of methylene chloride (CH2Cl2) added for extraction. Surrogate standards (n-alkene, C22:1 and orthoterphenyl, OTP) were added to each extraction to track any losses during the analysis procedures. Docos-1-ene usually resolves from C22 in the GC-FID determinations, while OTP is an aromatic hydrocarbon that can be distinguished in GC-FID analysis and does not interfere with UVF analysis. The extraction tubes were placed in a beaker of water with the probe of a sonicator in the water set to maximum energy for 15 minutes. Extracts were filtered through glass wool into a 100 ml boiling flask. Samples were extracted three times in total. Combined extracts were reduced in volume using rotary evaporation. When reduced to a few ml, the extracts were again filtered through a pasteur pipette containing glass wool and one gram of Na2SO4 into graduated glass centrifuge tubes with glass stoppers. If further volume reductions were required, this was achieved with a gentle stream of ultra pure nitrogen. Samples were adjusted to between 5 and 1 ml for total extractable organic matter (EOM) determinations using 10 µl aliquots gently evaporated onto the pan of a microbalance. The EOM weight provided a gravimetric determination of total oil content and was used to control lipid loading for the normal-phase chromatography cleanup procedures. Extracts were screened for their content of aromatic hydrocarbons and derivatives using ultraviolet fluorescence (UVF) spectroscopy. The cleanup procedure used 8 g of Al2O3 in a 1 cm diameter burette. The alumina was precleaned with methylene chloride in a Soxhlet extractor, dried and then activated at 200°C for 4 hours. Then 2% by weight of Milli Q water was added to obtain consistent activation. EOM loading was limited to 40 to 80 mg per column. The total hydrocarbon fraction was eluted with 10 ml of hexane, followed by 10 ml of 20% CH2Cl2, then with 10 ml of 50% CH2Cl2 in hexane. Sulfur in extracts was removed by adsorption onto activated copper.Cleaned extracts were then reduced with rotary evaporation and/or nitrogen and transferred into vials for analysis by gas chromatography (GC-FID) for total hydrocarbons and GC-mass spectroscopy (SIM-GC/MS) for individual aromatic hydrocarbons and biomarkers. Internal standards, added immediately before instrument injection, were used for quantification (C20:1 for GC-FID, and deuterated aromatic hydrocarbons for SIM-GC/MS). A composite reference oil biomarker standard distributed by the Australian Geological Survey Organization was used to identify peaks and retention times for triterpane and sterane biomarkers by SIM-GC/MS.For GC/FID analysis, all samples were analysed on Carla Erba GC8000 series gas chromatograph using cold on column injection and a DB-5MS column (J&W Scientific). For GC/MS analysis, all samples were analyzed on a HP 6890 Series II gas chromatograph/mass spectrometer equipped with a DB-5MS column (J&W Scientific). Samples were analysed by full scan and selected ion monitoring (SIM). Similar oven temperature programs were used for both instruments as follows: 50°C for 1 min, ramp 6°C/min to 100°C, then 4°C/min to 300°C, and hold for 18 min. More specific analytical details for hydrocarbon analysis are outlined in UNEP/IOC/IAEA (1992) and are further discussed by Burns (1993). Because of the evaporation of the most volatile hydrocarbons in the benzene series and the light alkanes during extract concentration steps, these procedures are quantitative for detecting petroleum hydrocarbons in sediments over ~ C12 to C36 or higher elution range.Burns KA (1993) Analytical methods used in oil spills. Mar. Pollut. Bull. 26(2): 68-72.UNEP/IOC/IAEA (1992) Determination of petroleum hydrocarbons in marine sediments. (Prepared by Burns KA et al.) IOC Manual and Guides No. 20 Rev. 1, 97p.Quality Assurance/Quality Control (QA/QC):Quality Assurance/Quality Control (QA/QC) was built into the chemistry program at multiple levels. The procedures and the analyst were first tested by duplicate analysis of an international reference material (IAEA Marine Sediment 357) for EOM, UVF and GC-FID total oil content and individual aromatic hydrocarbons by SIM-GC/MS. Results were satisfactory with each component falling within one standard deviation of the reported mean from an international inter-calibration exercise. A method blank was included with each batch of 20 sediment samples. With the wide range of oil concentrations observed, it was impossible to estimate an appropriate amount of surrogate standards to add to initial extracts due to the need for extract dilution and/or co-elution on the GC. Recovery of surrogate standards in the blanks and controls was carefully monitored and ranged from 66 to 88% for OTP and C22:1. Each extract had several estimates of oil content. These were used to construct linear regressions against each other and any outliers were scrutinized for sources of error in calculations. If this did not resolve an error, then another subsample of the sediment was analysed. There was also an independent estimate of oil by THC-FID done by a commercial laboratory. These data were not shown to the analyst until after this dataset was complete. Agreement between the laboratories was good over 4.5 orders of magnitude in oil content. Detection limits for samples analysed by AIMS were better at low concentrations because the commercial laboratory did only one solvent extraction and did not reduce the volume of extracts before GC analysis. Linear regression between THC-FID for the two laboratories was highly significant (y = 0.789x + 1466, where 'x' was the AIMS determination and 'y' was the contract laboratory determination, n = 41, r² = 0.92, P Pratt, C (1997) Weathering of hydrocarbons in tropical mangrove sediments: using analytical techniques to test the effects of dispersants to treat oil spill. B.Sc. Honours Thesis. Chemistry Department, James Cook University, Townsville, Queensland.

Notes

Credit
Burns, Kathryn A, Dr (Principal Investigator)