Search

A 1990/1991 shoreline ecology program to assess the fate and effects of the Exxon Valdez oil spill in Prince William Sound was updated in 1998 and 1999. This update included a sediment sampling program for organisms at “worst case” sites and at randomly chosen reference sites. Correspondence analysis (CA), a statistical method that examines animal communities in terms of their similarity, was used to define community structure Statistical analysis of the degree of similarity between communities was used to assess effects of site-specific variables (sediment grain size, total organic carbon (TOC) and wave energy), interannual variation, and degree of oiling. Interannual variability had a significant effect on community structure, whereas site specific variables and degree of oiling did not. Differences in communities between 1998 and other years were particularly dramatic. The importance of interannual change demonstrates the importance of multi-year sampling and of appropriate study designs for separating impact effects from the natural occurring environmental factors which affect biological communities.

Beginning in 1989, scientists supported by ExxonMobil conducted a number of scientific studies to assess the fate and effects of the Exxon Valdez oil spill on shorelines in Prince William Sound (PWS), Alaska. The 1990, 1991, and 1993 field programs included concurrent sediment sampling for hydrocarbon chemistry and sediment toxicity. This sediment quality assessment found that spill residues on the oiled shorelines rapidly lost toxicity through weathering. The relative amounts of naphthalenes and chrysenes in the polycyclic aromatic hydrocarbons (PAH) in the sediments were found to be good indicators of weathering and toxicity. Using a standard sediment amphipod bioassay on the field samples increased mortality above background was found at exposure levels above approximately 2,600 ng/g total PAH (TPAH) for oil that had weathered a year or more. For samples with TPAH > 2,600 ng/g, fractions of naphthalenes (R = +0.76) and chrysenes (R = −0.63) significantly correlated with amphipod mortality where samples with high mortalities (> 80%) were dominated by relatively high fractions of naphthalenes (median = 0.26), and the low mortality category (< 30%) was dominated by relatively high fractions of chrysenes (median = 0.24). The amphipod mortality data fit significantly to a logistic model. Estimated LC10 and LC50 values were approximately 4,100 and 10,750 ng/g TPAH, respectively. Sediment grain size and total organic carbon were also found to contribute to increased amphipod mortality and were covariates in the data analysis. As petroleum weathers through exposure to the environment and lighter hydrocarbon fractions are lost, the results of this study show that the Exxon Valdez spill oil rapidly lost toxicity as the fractions of chrysenes increased and fractions of naphthalenes decreased.

A shoreline ecology program was performed in Prince William Sound (PWS), Alaska in 1990 and 1991 (1 and 2 years after the Exxon Valdez oil spill, EVOS) to assess the fate and effects of the oil in the intertidal zone. Major components of the study were repeated in 1998 and 1999. This update included a sediment-sampling program at formerly oiled “worst-case” boulder/cobble (B/C) sites and randomly chosen unoiled B/C reference sites. The samples were analyzed for petroleum hydrocarbons and benthic infaunal community characteristics. This paper focuses on the results of the benthic infaunal community analysis. Analysis of Covariance (ANCOVA) was used to analyze the 1990–1999 infaunal species composition data. Very little effect of oiling was detected in either the analysis of community structure parameters or in individual species abundances. Oiling effects were detected at some sites in 1990 and 1991, but not in 1998 and 1999. Nearly all the change in intertidal community parameters between 1990 and 1999 was attributed to natural interannual variation. The composition of the intertidal community of B/C shores changed over time because of natural factors not related to the spill. A core group of species was found in each of the 4 years. This group of species represented between 9 to 30% of all species identified. Two other groups of species did not co-occur. One group was present in 1990 and 1991, but not in 1998 and 1999; the other group was present in 1998 and 1999, but not in 1990 and 1991. The progressive change in the animal community observed between 1990 and 1999 is very likely related to long-term climatic changes occurring in the study area and not the oil spill. This long-term study demonstrates the importance of study designs that allow separation of oiling effects from natural factors that can affect biological communities.

After the March 24, 1989 grounding of the Exxon Valdez and the release of 258,000 barrels of Alaska North Slope (ANS) crude oil into the marine environment of Prince William Sound (PWS), Alaska, a number of scientific studies were conducted from 1989 to 1998 to assess the fate and effects of the spill. These included the 1990–1991 shoreline ecology program (SEP), which detected little measurable impact of the spill, except at certain heavily oiled sites. In 1998, these 1990 and 1991 studies were updated. The 1998 study found spill small remnants from originally heavily oiled sites represent a minute fraction of the total PWS shoreline area. Any isolated deposits of remaining oil residues from the spill generally were found at the top of the tidal zone and highly weathered and therefore in a form and location unavailable and non-toxic to biota. Between 1991 and 1998, the concentrations of petroleum hydrocarbons at the “worst case” sites studied decreased dramatically. The polycyclic aromatic hydrocarbon (PAH) concentrations measured at these sites in 1998 were at least three times lower than the effects range-low (ER-L) sediment toxicity threshold values of 4,022 ng/g for total PAH. Mussels sampled at heavily oiled “worst case” sites in 1998 also had dramatically lower tissue PAH concentrations compared with samples taken in 1993, indicating low bioavailability of any hydrocarbons present and thus low risk to biota. Natural interannual variability in the structure of the biological infaunal communities is the largest, most consistent signal observed in this study, not any residual oil spill effects. The results of statistical analyses of the data (ANCOVA) indicated no continuing oiling effects in 1998. Consistent differences were observed between oiled and unoiled reference sites, but they were significant in only one analysis. These differences were unrelated to differences in the total PAH present among sites. Any observed differences are more likely related to the fact these “worst case” oiled sites were not selected in an unbiased, random manner.

During 1978 and 1979, sediment samples were collected in the mangrove areas on the west side of Bahia Sucia, Puerto Rico, to examine the effects of oil discharged from the tanker Zoe Colocotroni (March 18, 1973) on the infaunal community. Samples for benthic community analysis and hydrocarbon analysis were taken in representative types of mangrove habitats and compared with habitats in a reference area some distance from the spill site. Infaunal organisms larger than 0.5 mm were included in this study. Correlation of biological and chemical data indicates that for the high salinity habitats there were more infaunal organisms larger than 0.5 millimeters in the oiled area than in the unoiled area. Analyses of the red mangrove fringe areas showed that the distribution of infaunal organisms was not related to the distribution of oil in an obvious way. Large numbers of organisms were found in some heavily oiled areas. Even the most diverse communities seen in the impact area were not as diverse as the community seen in the unoiled reference area. We conclude that only in the red mangrove environment is there remaining damage to infaunal communities from the Zoe Colocotroni oil spill.

Samples of sediment were taken at intervals between 1979 and 1986 from soft sediment locations in the Aber Benoit, an estuary heavily impacted by the Amoco Cadiz oil spill of 1978, and from a comparable reference site outside the spill zone. Each sample was analyzed for aliphatic hydrocarbons by capillary gas chromatography. The data demonstrate the progressive weathering of the Amoco Cadiz oil resulting in an assemblage of biogenic hydrocarbons similar to the reference site by 1986 at all but the most heavily impacted locations. The data indicate that any residues of weathered Amoco Cadiz oil present in 1985-86 were in the final stages of degradation.

The fate and effects of two nearshore discharges of Murban crude oil at Long Cove, Searsport, Maine in August 1981 were studied following a one-year, pre-spill baseline study of the test areas. An upper and a lower intertidal sampling area within a 60 × 100 meter test plot were exposed to dispersed oil in water resulting from the discharge of 250 gallons of oil pre-mixed with 25 gallons of Corexit 9527 dispersant. Release of treated oil was around high-water slack tide on the surface of the water, with added mixing energy provided by mixing gates deployed by small boats. The maximum water depth over the test areas was 3.5 meters. Untreated crude oil (250 gallons) was released on an ebbing tide within a separate, boomed-off 60 × 100 meter test plot. A third test plot served as an oil-free reference plot. Water samples taken near the surface and near the bottom during and after discharge showed that chemically dispersed oil loses lower boiling hydrocarbons in both the aliphatic and aromatic fractions below n-C17as the droplets diffuse downward. Data are given for sediment samples taken from the test plots 11 months pre-spill and 10 months post-spill. Hydrocarbon analyses of the sediment samples show little incorporation of dispersed oil into the sediments of the treated oil plot relative to the sediments exposed to undispersed oil.

As part of the American Petroleum Institute sponsored tidal area dispersant project involving two test spills of Murban crude oil in Long Cove, Searsport, Maine in August, 1981, water samples were collected. This paper deals with the analytical results for the analyses of water samples collected for analysis of non-volatile hydrocarbons by: infrared spectrophotometric quantitation of total CCl4 extractables, and gravimetric analysis of aliphatic and aromatic hydrocarbon fractions followed by capillary gas chromatography. In the dispersant-treated oil discharge area, there were two primary water sampling locations during the discharge phase of the experiment: an upper intertidal area (maximum depth = 2 meters) and a lower intertidal area (maximum depth = 3.5 meters). The gas chromatographic data for the water samples were treated numerically to obtain parameters whose values reflect the extent of dispersed oil weathering. For the aliphatics, the peak area ratio for n C14/n C18 was calculated for each sample. For the aromatics, the ratio for the peak area sum of the mono, di, and trimethyl naphthalenes to that for the mono, di, tri, and tetramethyl dibenzothiophenes was determined for each sample. At both sampling locations, dispersed oil in water sampled 10 cm off the bottom consistently had a smaller fraction of lower boiling aliphatic and aromatic hydrocarbons than water sampled at the same place and the same time ½ meter below the surface. In addition, the data show that there is a 12–50 fold decrease in hydrocarbon concentration on going from near surface to near bottom at any given time, even in water as shallow as 2 meters. The data indicate that the primary mechanism for hydrocarbon loss involves volatilization of hydrocarbon fractions. Analyses of water samples taken from submerged plumes of dispersed oil outside the sampling areas demonstrated slower loss of low boiling components consistent with the importance of atmospheric exchange in the weathering process. In the chemical dispersal of an oil spill, it may be most advantageous to use mixing methods that minimize vertical mixing in order to maintain a high concentration of emulsified oil in the upper ½ meter water layer. This will maximize the extent of loss of lower boiling hydrocarbon components into the atmosphere and thus minimize the toxicity of any dispersed oil fractions that diffuse downward and interact with benthic communities.

The effect of two nearshore discharges of Murban crude oil on community structure in intertidal benthic communities was studied. One discharge consisted of 250 gallons of Murban crude only. Following the discharge, no measurable amount of Murban crude could be found in sediments exposed to the cloud of dispersed oil. Significant amounts were found in the test plot exposed to untreated oil. In the area exposed to untreated oil, more oil was found in the upper intertidal zone than lower down. Effects on infaunal communities mirrored the analytical results. There was no evidence of adverse effects on infaunal community structure from exposure to dispersed oil. There is clear evidence that exposure to untreated oil did adversely affect community structure. Some indigenous species were reduced in number or eliminated; there were blooms of opportunistic polychaetes. The changes in community structure brought about by the untreated oil are consistent with results observed at real-world oil spill sites.

On August 19, 1981, two test spills of Murban crude oil were carried out in Long Cove, Searsport, Maine. One spill was chemically dispersed; the other was not. Measurements were made to make quantitative comparisons of the chemical fates and biological effects of the two spills. Hydrocarbon analyses were carried out on water samples, animal tissue samples, intertidal sediment samples, and subtidal sediment samples. Biological measurements were carried out at the community level, whole animal level, and biochemical level of organization. No significant biological effects attributable to the dispersed oil spill were observed. This lack of effects is, in part, a result of changes in the physical and chemical properties of dispersed oil which help limit its availability and toxicity. The chemical fate and biological effects of the undispersed oil spill were typical of those reported from actual oil spill sites. Oil was incorporated into sediments and animal tissue. On the community level of organization mortality, reduced diversity and evenness, increased population density, and increased dominance by opportunists were all observed immediately after the spill, and up to 1 year later. On the whole animal level of organization, no effects on scope-for-growth were observed in two filter feeding bivalves. On the biochemical level of organization, activities of two sensitive enzyme systems were elevated.

On March 16, 1978, the tanker Amoco Cadiz ran aground off the coast of North Brittany. Her cargo of 221,000 tons of light crude oil was released into the sea. More than 126 miles of coastline were oiled, including a number of oyster (Crassostrea gigas) growing establishments. The North Brittany coastline already was stressed by earlier additions of oil and metals. In December 1979, 21 months after the oil spill, measurements of glucose-6-phosphate dehydrogenase activity, aspartate aminotransferase activity, and condition index were made in 14 populations of C. gigas. Five of the populations were outside the oil spill impact zone; nine populations were within it. At the same time the body burden of Cr, Ag, Pd, Zn, Cu, Cd, and Hg, as well as the body burden of aliphatic and aromatic hydrocarbons, was determined. Also at the same time, the numbers of total and fecal coliform bacteria in the ambient water were determined. Results of stepwise multiple regression have shown that 21 months after the Amoco Cadiz oil spill, aromatic hydrocarbons were only one of three factors adversely affecting C. gigas populations in North Brittany. Multiple regression techniques can be extremely useful in identifying those stressors associated with physiological effects in populations of animals.

A study was performed on the state of an oil spill site on the southwest coast of Puerto Rico. The location of the study was Bahia Sucia, the site of the Zoe Colocotroni spill of 17 March, 1973. Particular attention was given to the weathering characteristics of the stranded oil remaining at the sites and to the presence of biogenic hydrocarbons in the sediments. High resolution gas chromatography of the aliphatic and aromatic fractions of the sediment hydrocarbons was used to distinguish the types of hydrocarbons present and their distribution. It was observed that oil weathers much more rapidly in a tropical environment as compared with spills in temperate areas. Even in the most heavily impacted areas, the Zoe Colocotroni oil had weathered practically to the point of being tar. It was also observed that a tropical site has a significant chronic input of hydrocarbons from both natural and anthropogenic sources. Gas chromatograms of extracts of leaf waxes from live mangroves compared with biogenic hydrocarbons observed in the sediments. Tar-balls also account for a significant fraction of the hydrocarbon burden in the sites studied. The conclusion is that a tropical area has the potential for making a much more rapid recovery from an oil spill than a temperate one. Moreover, in assessing the effects of a tropical oil spill, care must be taken to distinguish the relative contribution to the total hydrocarbon burden in a spill area by oil, pelagic tar, and biogenic sources.

Effects of crude oil and salinity stress on the metabolism of two common filter feeding animals (Mytilus edulis and Modiolus demissus) have been investigated. Carbon budgets have been calculated for each species under a variety of combination of oil content and salinity. Both reduced salinity and crude oil tend to decrease the net carbon balance for each species; stresses from each source were additive in their effects on experimental animals. Although similar responses to oil were shown by each species, Mytilus appeared to be somewhat more resistant to oil than Modiolus.

As part of an ongoing study of the long-term effects of the Zoe Colocotroni oil spill of March 17, 1973, in Bahia Sucia, Puerto Rico, the effects of the remaining oil on the red mangrove trees in the impact area were investigated. This study involved four sampling trips to the spill site and a reference area outside the spill zone between April 1979 and April 1981. The present study was based on the observation that stressed mangrove trees in the heavily oil-affected areas had a similar appearance to trees exhibiting stress due to hypersalinity in unoiled areas. The working hypothesis was that petroleum hydrocarbons induce stress in salt-excluding plants such as red mangroves by disrupting the ability of the roots to exclude ions from seawater. Measurements of sodium (the principal seawater cation) and potassium (a major physiological ion) were made on leaf samples taken from trees from oil-affected areas and reference areas. Sediment core samples were taken from the root zone of the trees sampled at each site and analyzed for hydrocarbons by gravimetric and gas chromatographic methods. The results show a relationship between sediment hydrocarbon concentration and the ratio of sodium to potassium for mangrove leaves sampled at each site. The results show that for trees exposed at the root zone to the least weathered oil, the values of this ratio for the leaves had the largest values, reflecting an oil-induced impairment of the salt (Na) exclusion mechanism. This approach to measuring the physiological health of mangrove trees at an oil spill site offers a potentially useful means of documenting oil stress and recovery from oil stress in salt-excluding halophytes. Because oil stress in mangroves appears to be a root membrane-directed effect, there may be a “window in time” between initial oil impact and plant damage, as oil penetrates the sediments, during which mitigation measures could be taken.