5
Toxicological Effects of Dispersants and Dispersed Oil

One of the most difficult decisions that oil spill responders and natural resources managers face during a spill is evaluating the environmental trade-offs associated with dispersant use. The objective of dispersant use is to transfer oil from the water surface into the water column. When applied before spills reach the coastline, dispersants will potentially decrease exposure for surface dwelling organisms (e.g., seabirds) and intertidal species (e.g., mangroves, salt marshes), while increasing it for water-column (e.g., fish) and benthic species (e.g., corals, oysters). Decisions should be made regarding the impact to the ecosystem as a whole, and this often represents a trade-off among different habitats and species that will be dictated by a full range of ecological, social, and economic values associated with the potentially affected resources. Comparing the possible ecological consequences and toxicological impacts of these trade-offs is difficult. First, each oil spill represents a unique situation and second, it is often difficult to extrapolate from published research data into field predictions, especially regarding the possibility of long-term, sublethal toxicological impacts to resident species (Box 5-1 provides definitions for most the common terms used in discussions of toxicological effects).

Historically, the use of dispersants in the United States has been restricted primarily to deepwater (>10 m), offshore spills. In addition, the focus and the recommendations of the 1989 NRC report on oil dispersants were based on expected impacts of dispersants and dispersed oil during open ocean spills (NRC, 1989). As the potential use of dispersants is expanded into nearshore, estuarine, and perhaps even freshwater systems,



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects 5 Toxicological Effects of Dispersants and Dispersed Oil One of the most difficult decisions that oil spill responders and natural resources managers face during a spill is evaluating the environmental trade-offs associated with dispersant use. The objective of dispersant use is to transfer oil from the water surface into the water column. When applied before spills reach the coastline, dispersants will potentially decrease exposure for surface dwelling organisms (e.g., seabirds) and intertidal species (e.g., mangroves, salt marshes), while increasing it for water-column (e.g., fish) and benthic species (e.g., corals, oysters). Decisions should be made regarding the impact to the ecosystem as a whole, and this often represents a trade-off among different habitats and species that will be dictated by a full range of ecological, social, and economic values associated with the potentially affected resources. Comparing the possible ecological consequences and toxicological impacts of these trade-offs is difficult. First, each oil spill represents a unique situation and second, it is often difficult to extrapolate from published research data into field predictions, especially regarding the possibility of long-term, sublethal toxicological impacts to resident species (Box 5-1 provides definitions for most the common terms used in discussions of toxicological effects). Historically, the use of dispersants in the United States has been restricted primarily to deepwater (>10 m), offshore spills. In addition, the focus and the recommendations of the 1989 NRC report on oil dispersants were based on expected impacts of dispersants and dispersed oil during open ocean spills (NRC, 1989). As the potential use of dispersants is expanded into nearshore, estuarine, and perhaps even freshwater systems,

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects BOX 5-1 Common Toxicological Terms Related to Dispersant Toxicity Testing Exposure—Contact with a chemical by swallowing, breathing, or direct contact (such as through the skin or eyes). Exposure may be either acute or chronic. Acute—An intense event occurring over a short time, usually a few minutes or hours. An acute exposure can result in short-term or long-term health effects. An acute effect happens within a short time after exposure. Acute toxicity to aquatic organisms can be estimated from relatively short exposures (i.e., 24, 48, or 96 hr) with death as the typical endpoint. Chronic—Occurring over a long period of time, generally several weeks, months or years. Chronic exposures occur over an extended period of time or over a significant fraction of a lifetime. Chronic toxicity to aquatic organisms can be estimated from partial life-cycle tests of relatively short duration (i.e., 7 days). Sublethal—Below the concentration that directly causes death. Exposure to sublethal concentrations of a material may produce less obvious effects on behavior, biochemical and/or physiological function (i.e., growth and reproduction), and histology of organisms. Delayed Effects—Effects or responses that occur some extended time after exposure. Static Exposures—Exposures for aquatic toxicity tests in which the test organisms are exposed to the same test solution for the duration of the test (static non-renewal) or to a fresh solution of the same concentration or sample at prescribed intervals such as every 24 hr (static renewal). The concentration of the test material may change during the test due to bio- the trade-offs become even more complex. For example, the protection of sensitive habitats, such as tropical coral reefs and mangroves, is a priority in oil spill response decisions. Many studies have shown that oil, floating above subtidal reefs, has no adverse effects on the coral; however, if allowed to reach the shoreline, the oil may have long-term impacts to a nearby mangrove system. In addition, oil may persist in the mangrove system creating a chronic source of oil pollution in the adjacent coral reefs. The trade-off would be to consider the use of dispersants. Application of

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects logical uptake, volatilization, adherence to the test vessel, chemical degradation, etc. Flow-Through Exposures—Sample to be tested is pumped continuously into a dilutor system and then to the test vessels. This method is used to control sample concentration throughout the duration of the test. Spiked Exposures—Spiked Declining (SD) Exposures: Concentration of dispersant sample is highest at start and then declines to non-detectable levels after 6–8 hr using a flow-through exposures protocol developed by Chemical Response to Oil Spills Environmental Research Forum (CROSERF) participants. LCp—Lethal Concentration: The toxicant concentration that would cause death in a given percent (p) of the test population. For example, the LC50 is the concentration that would cause death in 50 percent of the population. The lower the LC, the greater the toxicity. ECp—Effective Concentration: A point estimate of the toxicant concentration that would cause an observable adverse effect on a quantal (“all or nothing”) response in a given percent (p) of the population. ICp—Inhibition Concentration: A point estimate of the toxicant concentration that would cause a given percent (p) reduction in a non-quantal biological measurement such as reproduction or growth. NOEC—No-Observed-Effect-Concentration: The highest concentration of toxicant to which organisms are exposed in a full or partial (short-term) life-cycle test that causes no observable adverse effects on the test organisms (i.e., the highest concentration of toxicant at which the values for the observed responses are not statistically different from the control). SOURCES: Singer et al., 1991; Rand, 1995; Grothe et al., 1996; EPA, 2002a,b, 2005; New York Department of Health, 2005. dispersant would result in dispersion of the oil in the water column and so provide some degree of protection to the mangroves; however, the reef system would now have to endure the consequences of an increase in dispersed oil in the water column (see section on coral reefs later in this chapter). Therefore, for oil spill responders to decide upon appropriate response strategies, it is important that decisions are based on sound scientific data. Ecological factors that go into this decision include: expected sensitivity of exposed resources, proportion of the resource that would be

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects affected, and recovery rates (Pond et al., 2000). There is a tremendous need to reduce the uncertainty associated with each of these decision criteria. This chapter reviews recent laboratory, mesocosm, and field studies on the toxicological effects of dispersants and dispersed oil, particularly those published since the 1989 NRC report on oil dispersants (NRC, 1989). The intention is first to summarize the current state of understanding regarding the biological effects of dispersants and dispersed oil, and second to make recommendations for additional studies that will help fill critical data gaps in the knowledge and understanding of the behavior and interaction of dispersed oil and the biotic components of ecosystems. The following discussion is limited primarily to studies of the toxicological effects on individual organisms, as opposed to populations or communities. This narrower scope reflects the current state of science in ecotoxicology (see Box 5-2). Although the research and management communities recognize the importance of considering higher order ecological effects, not enough is known to extrapolate from toxicity tests to population or community-level impacts—an issue that concerns all applications of ecotoxicology. Consequently, the explicit consideration of these impacts, and formulation of research to address them, is beyond the scope of this report on the application of ecotoxicological principles to oil spill research. Due to implementation of several of the recommendations made in 1989 (NRC, 1989), particularly the standardization of toxicity testing methods and information garnered from long-term monitoring of field studies, some general conclusions about the toxicity of dispersants and dispersed oil can be reached. However, there are still areas of uncertainty that will take on greater importance as the use of dispersants is considered in shallow water systems. Specifically, there is insufficient understanding of the fate of dispersed oil in aquatic systems, particularly interactions with sediment particles and subsequent effects on biotic components of exposed ecosystems. In addition, the relative importance of different routes of exposure, that is, the uptake and associated toxicity of oil in the dissolved phase versus dispersed oil droplets versus particulate-associated phase, is poorly understood and not explicitly considered in exposure models. Photoenhanced toxicity has the potential to increase the impact “footprint” of dispersed oil in aquatic organisms, but has only recently received consideration in the assessment of risk associated with spilled oil. One of the widely held assumptions is that chemical dispersion of oil will dramatically reduce the impact to seabirds and aquatic mammals. However, few studies have been conducted since 1989 to validate this assumption. Finally, more work is needed to assess the long-term environmental effects of dispersed oil through monitoring and analysis of spills on which

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects BOX 5-2 Assessing Population and Community-Level Impacts: A Central Issue in Ecotoxicology The decision of whether or not to use chemical dispersants in aquatic systems involves evaluation of the trade-offs between potential impacts on various natural resources. Toxicity tests are one of the primary tools that are used to predict these impacts. Much of the toxicological literature focuses on the effects of dispersed oil on individual organisms, because this is the level of biological organization that is most readily studied. Of far greater significance—and of far greater complexity as well—are the effects of dispersed oil on populations and communities of organisms. How to make meaningful predictions about toxicological effects on populations or communities is a problem that is not unique to the assessment of the impacts of an oil spill, but rather is a central question in the field of ecotoxicology. How does the loss or impairment of one or more individual organisms impact a population? How does damage to single or multiple populations impact a community? In the case of dispersed oil, numerous ecological factors may affect the impacts to, and recovery of, these higher levels of biological organization, including the proportion of the resource affected (which in turn involves an understanding of the toxicological sensitivity of organisms as well as the behavior, habits, and habitats that will affect the probability of a species being exposed to oil), birth and death rates of the affected species, the current status of the population (e.g., endangered or common species), life stages that are present, and time of year (e.g., nesting or spawning season, seasonal migration). Population and community models are tools that show promise in enhancing our understanding of the toxicological impacts to these higher levels of biological organization. Despite recent efforts to advance these approaches (SETAC, 2003), there is no scientific consensus on this issue. Consequently, the majority of ecological risk assessments of environmental chemicals are still based on species-specific tests of toxicological effects on individual organisms. Until population and community-level approaches are more widely accepted and utilized in ecotoxicology, evaluations regarding the impacts of oil spills will remain largely based on qualitative assessments and best professional judgment. However, progress has been made in our understanding of the long-term effects of oil spills on biological communities. The NRC (2003) report on Oil in the Sea III: Inputs, Fates and Effects provides a good summary of some of the long-term studies that have been conducted after oil spills, especially those assessing effects on benthic communities and seabirds. For the moment, these types of studies represent the best chance of improving our understanding of the effects of spilled and dispersed oil on biological populations and communities. SOURCE: SETAC, 2003.

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects dispersants have been used. Interestingly, several of these data gaps were also identified in 1989 (NRC, 1989). TESTING PROCEDURES FOR DISPERSANT AND DISPERSED OIL TOXICITY Toxicity Tests Much that is currently known about the toxicity and biological effects of dispersants and dispersed oil has been derived from bench-scale acute toxicity tests. These tests typically consist of exposing a single species to varying dilutions of dispersant or dispersed oil preparations under carefully controlled laboratory conditions. Factors that influence such tests include: choice of test organism and life stage condition of oil (fresh versus weathered) method of preparing test solutions exposure conditions choice of response parameters Commonly used test organisms include fish, mollusks, arthropods, annelids, and algae. The choice of test organism is dictated by a combination of factors including potential risk, comparative sensitivity, suitability of the species to the testing conditions, and relative ecological and economic significance. An additional consideration is the specific life stage to be tested, because larvae and adults may respond to exposure in significantly different ways. The method of preparing test solutions is particularly important in the case of dispersed oil testing. Water and oil are not easily miscible, so factors such as mixing energy and loading method can readily affect the relative concentrations of oil components to which test organisms are exposed. Dispersants can also separate and form films on water unless test solutions are properly prepared and mixed. Exposure conditions in toxicity tests for dispersants and dispersed oil vary with the choice of test chamber (e.g., open or closed), the exposure model (e.g., static or flow-through, spiked or continuous), route of exposure (e.g., water or food), test duration, and other factors such as temperature, salinity, and buffering capacity. The choice of test duration alone can significantly overestimate or underestimate toxicity depending on the actual oil spill situation being simulated. The choice of response parameters measured in a test can be significant as well. Current generation dispersants appear to cause toxicity

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects through disruptive effects on membrane integrity and a generalized narcosis mechanism (NRC, 1989). Dispersed oil, on the other hand, exerts a toxic effect through multiple pathways including narcosis, more specific receptor-mediated pathways associated with elevated dissolved phase exposures, and possibly by additional pathways such as smothering by dispersed oil droplets. The presence of receptor-mediated pathways suggests that relatively short-term toxicity tests with death as the primary or sole endpoint may not be sufficient to adequately assess the potential risks of dispersed oil. Short-term tests are also incapable of addressing potential delayed effects due to metabolism of oil constituents, bioaccumulation, or possible photoenhanced toxicity. Although much of the literature on the toxicity of dispersants and dispersed oil is based on typical static exposures of 48–96 hr duration, such tests have been criticized as potentially overestimating the toxicity of oil and dispersed oil in actual spill scenarios (NRC, 1989; George-Ares, et al., 1999). In response to these concerns, a university-industry-government working group, the Chemical Response to Oil Spills Environmental Research Forum (CROSERF), was organized to coordinate and disseminate research on oil spill dispersant use. CROSERF developed toxicity test protocols involving spiked exposures of shorter durations and standardized preparations of water accommodated fractions (WAF) of oil and chemically enhanced water accommodated fractions of dispersed oil (CEWAF) (Singer et al., 1991, 1993, 1994a,b, 1995, 2000, 2001a,b; Clark et al., 2001; Rhoton et al., 2001). For clarity, the term “CEWAF” will only be used in this chapter when referring to a dispersed oil water accommodated fraction that is prepared using the CROSERF protocols. “Chemically dispersed oil” will be used to describe non-CROSERF preparation methods. The CROSERF test methods are summarized in Table 5-1. The main focus of CROSERF was to standardize methods (i.e., preparation and quantification of fractions and exposure protocols) to allow for greater comparability of toxicological data. In this regard, CROSERF was quite successful. Significant toxicological information was generated using these protocols that successfully addressed the relative toxicity of different dispersants and oil, as well as the relative sensitivity of test organisms. Refinements to the CROSERF protocols may be warranted for future toxicity testing of dispersants and dispersed oil, either to address specific concerns with the current test procedures (as highlighted below) or to provide greater site-specificity for risk assessment purposes (e.g., dispersant use in nearshore areas). For example, several refinements to the CROSERF procedures have been proposed to adapt the test to subarctic conditions, including changes in WAF preparation, exposure and light regimes, analytical chemistry, and use of subarctic test organisms (Barron

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects TABLE 5-1 CROSERF Toxicity Test Specificationsa Parameter CROSERF Procedure WAF and CEWAF Preparation Water Local seawater recommended; minimal 0.45 µm filtration Oil Fresh and artificially weatheredb Oil loading Variable loading (0.01–25 g of oil per liter of water); serial dilution not recommended Vessel 1–20 L carboys or aspirator bottles as appropriate for amount of solution required Head space 20–25% by volume Mixing energy/durationc Original: 18–24 h at low mixing energy (approximately 200 rpm with no vortex) and no settling time for WAF, and moderate mixing energy (20–25% vortex) with 3–6 h settling time for CEWAF; Modifiedd: WAF and CEWAF both prepared with moderate mixing energy and settling Mixing conditions Sealed in dark at test temperatures Analytical chemistrye TPH and <C10 volatile hydrocarbons required, other analyses optional; TPH, alkanes measured by GC/FID; VOCs and PAHs measured by GC/MS Dispersant (dispersant:oil) Primarily Corexit 9500 and/or 9527 (1:10); occasionally Corexit 9554 and others Dispersant concentration verification UV–spectroscopy Test Procedures Test design Five treatments plus control, each with three replicates Test concentrations Exposure concentrations derived from a series of geometrically progressing oil loading rates; for toxicity comparisons, total hydrocarbon content (THC: TPH plus <C10 volatile hydrocarbons) recommended as concentration endpoint Exposure regime 48 or 96 h tests in sealed vessels; static-renewal exposures for duration of test, aeration discouraged; flow-through “spiked exposures” with concentrations decreasing to nondetectable levels in <8 h Test maintenance Renew solutions at unspecified intervals for static renewal tests, removing dead organisms; dead organisms not removed in flow-through exposures; feeding as specified for test species, with food amount adjusted for loss of test organisms

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects Parameter CROSERF Procedure Species/life stage Temperate aquatic species/early life stages Temperature; salinity Temperatures appropriate to species; salinity full-strength seawater Light regime Laboratory lighting (fluorescent) Toxicity endpoint Lethality assessed daily for length of test; sublethal endpoints assessed as appropriate for test organism Bioaccumulation Not measured aSOURCE: Singer et al. (1991); Singer et al. (2000); Clark et al. (2001), Rhoton et al. (2001), Singer et al. (2001a). bModified ASTM Method D-86 (1990 modification); oil “topped” by distillation to 200 °C roughly simulating 1 day at sea (Daling et al. 1990; Singer et al., 2001b). cWAF=Water accommodated fraction; CEWAF=Chemically enhaced WAF, or chemically dispersed oil; stir bar size 1–2 in as appropriate. dClark et al. (2001) modification of standard CROSERF mixing energy protocol for physically dispersed oil (WAF) using 20–25% vortex, followed by 6 h settling time. eTPH: total petroleum hydrocarbons; alkanes: >10 carbon alkanes; VOC: volatile organic compounds (<10 carbon alkanes and MAHs); PAHs: polycyclic aromatic hydrocarbons; GC: gas chromatography; FID: flame ionization detection; MS: mass spectrometry and Ka’aihue, 2003). However, the potential benefits of altering test protocols from the CROSERF procedures should be carefully weighed against the implications for potential loss of data comparability and reproducibility. Some factors to consider in possible refinements to the current CROSERF test protocols for future testing efforts include: procedures for making dilutions to be tested exposure regimes, including test chambers methods for quantifying petroleum exposure chemical measurements response parameters potential photoenhanced toxicity Two alternate methods for preparing WAF and CEWAF fractions have been suggested, discussed at great length, and remain the subject of scientific debate (see Singer et al., 2000; 2001a; Barron and Ka’aihue, 2003) The CROSERF protocols recommend preparation of toxicity test solutions

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects by variable loading using a series of decreasing concentrations of applied oil and dispersant (Figure 5-1). Other researchers (for example see Barron and Ka’aihue, 2003) have proposed the use of a single oil:water loading rate and the preparation of test solutions using various dilutions of the stock preparation. The decision of which method to use may depend ultimately on the specific scientific question being addressed. Singer et al. (2001a) argue for the variable loading method because they believe it is more “field relevant” since spilled oil slicks tend to be dynamic, continu- FIGURE 5-1 Comparison of variable loading and variable dilution methods of preparing toxicity test solutions. SOURCE: Barron and Ka’aihue, 2003; courtesy of Elsevier.

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects ally changing in size, shape, and thickness. Consequently, these tests address the question: “At what oil to water loading ratio is WAF (CEWAF) toxic?” Barron and Ka’aihue (2003) advocate a variable dilution method for preparing a WAF for testing dispersant that standardizes the oil:water ratio and provides a consistent chemical concentration in a test-series for each oil-dispersant combination (Figure 5-1). This approach answers the question: “At what dilution is a given oil:water ratio of WAF (CEWAF) toxic?” Because it has not been conclusively demonstrated that either method more accurately simulates the temporal dilution of dispersed oil under actual spill conditions, we do not endorse one method over the other. As noted below, there are drawbacks to both approaches. In the variable loading method, the dispersant:oil ratios do not change and, therefore, each test preparation will have different amounts of oil and dispersant relative to the volume of water in the test chamber. As a result mixing energies change as loading rate (Singer et al., 2000), potentially affecting droplet size or coalescence. The drawback of the variable dilution method has been described as the production of the equal ratio of each specific PAH across the dilution range (Barron and Ka’aihue, 2003). WAF and CEWAF produce significant proportions of oil in the droplet phase, such that increasing dilution may differentially affect the partitioning of the PAH into the aqueous phase. In addition, Barron and Ka’aihue (2003) have argued that the variable dilution approach provides economies in analytical costs by reducing the need to analyze the composition of every tested concentration. However, if chemical analyses were limited to stock solutions, inaccuracies may occur due to differential partitioning in the test dilutions, adsorption of compounds onto test chambers, or loss to the gaseous phase. The interpretation of the results of toxicity tests can be significantly affected by the method of WAF and CEWAF preparation because of the variable solubilities of the many components in oil. For example, the variable loading method yields different mixtures of petroleum hydrocarbons at different loading rates (see Figure 5-1). The problems that arise between the two methods are due to the fact that often both methods report their data in the same form (i.e., in ppm of some overall metric, such as TPH or tPAH). Therefore, the elimination of any fractional characteristics can lead to a misunderstanding of what that concentration actually represents. For example, LC50 data derived from tPAH or TPH alone may result in under-or overestimation of toxicity depending on test preparation method used. Hence, more complete characterizations of chemical analytes are needed. Another issue with the CROSERF protocols concerns the mixing energies involved in the process of preparing test solutions. The various CROSERF protocols employ equal mixing energies for the production of CEWAF, but differ in the approaches for the production of WAF. For ex-

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects that it is much more likely that adverse effects to coral species would be manifested at sublethal levels. One relatively unstudied hypothesis that could explain delayed effects is that most of the toxicity is derived from exposure to the UV radiation in sunlight (see earlier section on Phototoxicity in this chapter). This phenomenon may be of particular relevance in explaining the high toxicity of accumulated oil in corals, species that are slow to depurate PAH. BOX 5-4 Case Study: TROPICS, Panama Spilled Oil Type/Volume/Conditions. In 1984, a field oil experiment called the Tropical Oil Pollution Investigations in Coastal Systems (TROPICS) was conducted in Panama. The objective of the TROPICS experiment was to evaluate the relative impacts of oil and dispersed oil on mangroves, seagrasses, and corals. Exposure concentrations were targeted to be as high as 50 ppm, in a worst-case scenario, with dispersants applied to oil directly over corals. Physical and Biological Setting. Sheltered shallow area near Bocas del Toro, Panama (Figures 5-7 and 5-8). Mature mangroves with extensive seagrass beds (water depth average about 40 cm), and coral reefs (water depth average 60 cm). Oil and Dispersed Oil Application. The oil, or dispersed oil, was applied inside boomed areas 30 m wide and 30 m deep, extending across all three habitats. The pre-mixed dispersed oil (4.5 barrels) was released over a 24-hour period so that the dispersed oil concentrations would stay elevated over the exposure period. The untreated whole oil (6 barrels) was released in two periods over the 24 hr, at an application rate of 1 liter/m2. After one more day, the remaining floating oil was removed with sorbents. Monitoring Results. Water Column Monitoring: Oil concentrations at each treatment site (oil or dispersed oil) were monitored continuously for 24 hr using a field fluorometer that was calibrated to convert fluorescence into the concentration of physically and chemically dispersed oil. Discrete and unfiltered water samples were collected for chemical analysis by gas chromatography (GC). In comparing the oil concentrations in the water as measured by both approaches, the field fluorometer readings were 3 times higher that the GC concentration for samples from the dispersed oil site, and they were 17 times higher than the samples from the undispersed oil site. Therefore, the oil concentrations as measured in the discrete water samples by GC were used to calculate the oil exposures because these results are more quantitative.

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects Not only are corals in high-light environments, they are translucent and seek high intensity light environments (by regulating pigments or altering their position with respect to the sun) to foster the symbiotic relationship with photosynthetic algae. An additional stress for corals may be attributed to the physical toxicity of oil droplets. It has been observed that oil droplets adhere to the surface of the coral, which results in a complete breakdown of the under- FIGURE 5-7 Case study: (TROPICS, Panama) Map of TROPICS study sites near Bocas del Toro, Panama. SOURCE: Ward et al., 2003; courtesy of the American Petroleum Institute. Effects: The sites were monitored five times in the first two years and once in 1994, ten years later. At the oil-only site, the corals were exposed to a 24-hour average of 0.14 ppm and a 48-hour average of 0.14 ppm. No significant impacts to corals were observed during any monitoring period. At the dispersed oil site, the corals were exposed to a 24-hour average of 5.1 ppm (with a 1 hr maximum of 14.8 ppm) and 1.6 ppm at 48 hr. The average exposure over the 48-hour period was 3.4 ppm. At these expo-

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects FIGURE 5-8 Case study: (TROPICS, Panama) Aerial view of whole oil and dispersed oil sites. SOURCE: Coastal Science Associates, Southern Affiliate, Incorporated. sures, there were significant impacts to the shallow coral reef communities. Impacts observed at two years post-exposure included: reduced coverage by the major categories of all organisms (30 percent), hard corals (10 percent), all animals (30 percent), and plants (10 percent); reduced growth of the two most important hard coral species (Agaricia tennuifolia and Porites porites) but not two others (Montastrea annularis and Acropora cervicornis); and mortality of binding sponges. Studies conducted ten years post-exposure showed full recovery of coral coverage to levels equal those present pre-spill at the dispersed site and equal to conditions at the non-oiled control site. Dispersed oil concentrations over the shallower seagrass (Thalassia testudinum) habitat were five times higher than over the coral habitat, av- lying tissues (Johannes, 1975). Again this phenomenon may be of direct relevance in interpreting physically versus chemically dispersed oil toxicities. NRC (1989) stated that the smaller droplets in chemically dispersed oil did not adhere to the corals, in contrast to the larger, physically dispersed oil droplets, some of which were found on coral a few weeks after

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects eraging 22 ppm over 24 hr with a maximum of 70 ppm as measured in discrete water samples analyzed by GC. Even at these high exposures (the maximum likely oil concentrations), no negative effects were observed for plant survival, growth rates, or leaf blade area at the dispersed oil treatment site compared to the non-oiled reference site. Untreated, whole oil caused significant impacts to mangrove habitats with high levels of defoliation and 17 percent mortality of adult mangroves after 2 years. After 10 years, mangrove mortality increased to 46 percent and some subsidence of the sediment surface was observed at the oiled site. After 18 years, the oiled site started to show some recovery as new trees replaced the dead trees (Figure 5-9; Ward et al., 2003). This field experiment clearly demonstrates the trade-offs associated with dispersant use in shallow tropical settings. FIGURE 5-9 Results of 18 years of monitoring impacts to mangroves in Panama as part of TROPICS. Histograms reflect mangrove tree or seedling population counts (1984–2001) from whole oil (Site O) and dispersed oil (Site D) compared to a reference site (Site R). SOURCE: Ward et al., 2003; courtesy of the American Petroleum Institute. SOURCE: Summary compiled from Ballou et al. (1987), Dodge et al. (1995), and Ward et al. (2003). exposure to oil. In addition, a common stress response to oil pollution that has repeatedly been observed in coral species is the excessive production of mucus (see Shigenaka, 2001). This protective response can reduce the bioaccumulation of chemical contaminants by binding them in this lipid-rich mucus matrix that is ultimately “sloughed off” (or eaten by grazing

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects fish) the surface of the coral, so protecting the underlying tissues. It is unclear whether chemically dispersed droplets or physically dispersed droplets or accumulation of dissolved components could alter this response. The excessive production of mucus takes energy away from normal cellular processes potentially reducing the overall health and fitness of the coral. In the case of chronic oil pollution events, such as continued leaching from mangrove sediments, excess mucus production could ultimately lead to coral death. In conclusion, recent studies of coral larvae clearly demonstrate impacts of dispersants and dispersed oil on corals and, because of their life history and habitat characteristics, these species may be especially susceptible (Table 5-11). Consequently, decisions concerning dispersant use should take coral toxicity studies into consideration. In addition, laboratory studies are needed to estimate the relative contribution of dissolved-and particulate-phase oil to toxicity in representative coral species. Because corals typically experience high levels of natural sunlight, these toxicity tests should include an evaluation of delayed effects and photoenhanced toxicity. Mangroves Few reports have been published that address the use of dispersants in treating oil spills close to mangroves. Early work by Getter and Ballou (1985) used an experimental spill at a site in Panama and concluded that dispersant use reduced the overall impact of oil on mangroves. This was a long-term project (10 years), but lacked replication of study sites (Dodge et al., 1995). In order to investigate the types of oil spill responses that might reduce the impact of oil spills and to address the need for more relevant information on the effects of oil spills on mangroves, Duke, Burns and co-workers carried out a number of field trials to assess the benefits of two remediation strategies for mangrove forests (see Burns et al., 1999; Duke and Burns, 1999; Duke et al., 1998a,b,c, 1999, 2000). These experiments were aimed at bridging the gap between surveys of real spill incidents (e.g., Volkman et al., 1994; Duke et al., 1997, 1998c) and those obtained from seedling laboratory experiments (Lai and Lim, 1984; Wardrup, 1987; Duke et al., 1998a). Field experiments, named the Gladstone trials, investigated the effects of different oils and remediation strategies on mangroves over both short and long-term time scales (1995–1998) utilizing a variety of replicated trials. One study compared the effects of dispersant (Corexit 9527) or bioremediation (aeration plus nutrients) strategies on a controlled spill using pre-weathered (24 hr) Gippsland light crude oil. It should be noted that the dispersant Corexit 9527 was premixed and weathered with the oil mixture before application. There were

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects no differences observed between oil alone and dispersed oil treatments on resident fauna. Death of mangrove trees, however, was significantly lower in the plots treated with dispersant, similar to data previously obtained from laboratory and field studies (Duke et al., 1998a,c; Duke and Burns, 1999). With oil alone, long-term impacts on the fauna and little sign of recovery of trees led the authors to conclude that dispersion of spilled oil before it reaches mangroves should be considered for reducing the long-term impact of oil on mangrove habitat. It was interesting to note that the use of Corexit 9527 resulted in no difference in the amount of oil absorbed by the sediments, the penetration of oil to depth, or the weathering patterns of the oil over time. IMPROVING THE USE OF INFORMATION ABOUT EFFECTS IN DECISIONMAKING As discussed in Chapter 2, the ultimate decision regarding the use of dispersants in spill response generally rests upon answering the question as to whether use of dispersants will reduce the overall impact (Figure 2-4 in Chapter 2) by reducing the effects on some specific and sensitive species or habitat, without causing unacceptable harm to another specific and sensitive species or habitat. This decision represents a trade-off that will be dictated by a range of ecological, social, and economic values associated with the potentially affected resources. When spills occur offshore, where the potential magnitude and duration of impacts on organisms in the water column or seafloor can be assumed to be minimal, a decision to use dispersant can be made with information that is generally available. As the capability to deploy dispersants offshore increases, however, the capability to use dispersants in nearshore and shallower water settings will also increase. At the present time, the current understanding of the risk of dispersant use to shallow water or benthic species during a given spill is typically not adequate to allow rapid and confident decision-making. Several factors contribute to this uncertainty. The rate of processes controlling the ultimate fate of dispersed oil is poorly understood. Of particular concern is the fate of dispersed oil in areas with high suspended solids and areas of low flushing rates. There is insufficient information to determine how chemically dispersed oil interacts with suspended sediments, as well as biotic components of aquatic systems, both short- and long-term, compared to naturally dispersed oil. Relevant state and federal agencies, industry, and appropriate international partners should develop and implement a focused series of experiments to quantify the weathering rates and final fate of chemically dispersed oil droplets compared to undispersed oil. Results from these experiments could be integrated with results from biological exposures

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects TABLE 5-11 Toxicity Studies of Chemically Dispersed Oil (or Dispersant Alone) to Coral Species in Laboratory and Field Studies (since 1988) Species Oil (D:O ratio) Dispersant Exposure Coral reef (primarily Porites porites and Agaricia tennuifolia) PBCO (1:20) Commercial nonionic glycol ether-based 24 hr continuous release Acropora spp. (growth), variety of corals visually assessed Arabian light crude (1:20) Corexit 9527 24 hr and 120 hr exposures plus 1 year recovery. Growth assessed. Acropora palmata, Montastrea annularis, Porites porites Oil (W) not detailed (1:10) 12 D including Corexit 9527, Corexit 9550, Finasol OSR7 DO and O, 6–10 hr, 2 week recovery and delayed assessments in clean SW. Larvae of Stylophora pistillata and Heteroxenia fuscescense Egyptian crude (1:10) Inipol IP-90, Petrotech PTI-25, Biosolve, Bioreico R-93, Emulgal C-100 WSF (of O), DO WAF and D (5–500 ppm). 2–96 hr, static Acropora millepora (eggs and larvae) Heavy crude oil (1:10/100) Corexit 9527 WAF, DO and D alone. Exposures; 4 hr fertilization assays (FA), 24 hr larval metamorphosis assay (LM); static Stylophora pistillata (adult) Egyptian crude (1:10) Emulgal C-100 WSF (of O), D and DO WAF. 24 hr, static with recovery in clean SW. NOTE: D, dispersant; DO, chemically dispersed oil; D:O, dispersant:oil ratio; HC, Hydrocarbon concentration (ppb); O, oil; PBCO, Prudhoe Bay Crude Oil; SW, seawater; TPH, total petroleum hydrocarbons; WAF, water-accommodated fraction; WSF, water soluble fraction. comparing uptake of dissolved, colloidal, and particulate oil to provide a comprehensive model of the fate of dispersed oil in aquatic systems. There is insufficient understanding of the actual concentrations and temporal/spatial distributions and behavior of chemically dispersed oil from field settings (from either controlled experiments or actual spills). Data from field studies (both with and without dispersants) are needed to validate models, provide real-world data to improve knowledge of oil fate and effects, and fulfill other information needs. Relevant state and federal agencies, industry, and appropriate international partners should develop and implement steps to ensure that future wave-tank or spill-of-opportunity studies (or during the Natural Resource Damage

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects Response Comments Reference DO decrease in coral cover—complete elimination of A. tennuifolia. Continuous field measurement of TPH and C1-C10 hydrocarbons aBallou [et al., 1989 Delayed sublethal impacts in all plots (bleaching); DO 120 hr exposure plots recovery less. No difference in growth rates. HC concentrations measured over time (to 120 hr) aLegore et al., 1989 Mortality was D dependent. Nominal exposures Thorhaug et al., 1989 Varied with exposure—from unsuccessful larval settlement to death. D toxic, DO WAF more toxic cf. WSF (and D alone). Nominal exposures (dilutions of stocks) Epstein et al., 2000 FA; WAF no effect. DO slight more toxic c.f. D alone. LM; DO more toxic cf. WAF, D toxic but at higher levels cf. [D] in DO. Measured THC mg/L in stocks. Nominal concentrations calculated for dilutions. Negri and Heyward, 2000 No death in WSF. D alone (1% or >) very toxic within 24 hr, delayed death (day 6) at 0.1%. DO WAF similar to D alone. Nominal exposures (dilutions of stocks) Shafir et al., 2003 aField study. Assessment investigations of oil spills that are not treated with dispersants) implement a field program to measure both dissolved-phase PAH and particulate/oil-droplet phase PAH concentrations for comparison to PAH thresholds measured in toxicity tests and predicted by computer models for oil spill fate and behavior. Accomplishing this will require the development and implementation of detailed plans (including preposition of sufficient equipment and human resources) for rapid deployment of a well-designed monitoring plan for actual dispersant applications in the United States. The RRT Region 6 Spill of Opportunity Monitoring Plan for dispersant application in the Gulf of Mexico should be finalized and implemented at the appropriate time. In addition, con-

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects sideration should be given to long-term monitoring of sensitive habitats and species (e.g., mangroves, corals, sea grasses) after dispersant application to assess chronic effects and long-term recovery. These data will be valuable in validating the assumptions associated with environmental trade-offs of using dispersants. One of the widely held assumptions concerning the use of dispersants is that chemical dispersion of oil will dramatically reduce the impacts of oil to seabirds and aquatic mammals, primarily by reducing their exposure to petroleum hydrocarbons. Evaluating the validity of this assumption is critical, because it is often a key factor in the decision on whether or not to use dispersants on a particular spill (e.g., in the ecological risk assessment workshop analyses). In addition, populations of waterfowl and some aquatic mammals may be higher in nearshore and estuarine areas; therefore, validating this assumption becomes even more important. Unfortunately, there is very little available information on the effects of dispersed oil on wildlife, especially aquatic mammals. Of additional concern is the effect of dispersed oil and dispersants on the waterproof properties of feathers and their role as thermal insulators. One of the recommendations of the NRC (1989) report was that studies be undertaken “to assess the ability of fur and feathers to maintain the water-repellency critical for thermal insulation under dispersed oil exposure conditions comparable to those expected in the field.” This recommendation is reaffirmed because of the importance of this assumption in evaluating the environmental trade-offs associated with the use of oil dispersants in nearshore and estuarine systems and because it has not been adequately addressed. The primary assumption for models predicting acute toxicity of physically and chemically dispersed oil is additive effects of dissolved-phase aromatic hydrocarbons. However, the possibility of photoenhanced toxicity and particulate/oil droplet phase exposure is generally not considered. A number of laboratory studies have indicated toxicity due to PAH increases significantly (from 12 to 50,000 times) for sensitive species in exposures conducted under ultraviolet light (representative of natural sunlight), compared to those conducted under the more traditional laboratory conditions of fluorescent lights. In addition, the toxicity tests typically do not consider delayed acute or sublethal effects. Consequently, current testing protocols may significantly underestimate toxicity for some species. For example, corals appear to be particularly sensitive to dispersants and dispersed oil due to the potential for photoenhanced toxicity and delayed effects. Similarly, toxicological effects due to increased exposure to oil from smothering, ingestion, or enhanced uptake are not explicitly considered in exposure models. Better understanding of these variables will decrease the uncertainty associated with predicting ecological effects of dispersed oil. Relevant state and federal agencies, industry,

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects and appropriate international partners should develop and implement a series of focused toxicity studies to: (1) provide data that can be used to parameterize models to predict photoenhanced toxicity; (2) estimate the relative contribution of dissolved and particulate oil phases to toxicity with representative species, including sensitive species and life stages; and (3) expand toxicity tests to include an evaluation of delayed effects. Detailed chemical analyses should accompany these tests, including characterization of dissolved and particulate oil composition and concentrations, as well as bioaccumulation. By refining our understanding of these variables, and incorporating them into decision-making tools, such as fate and effects models and risk rankings, the ability of decisionmakers to estimate the impacts of dispersants on aquatic organisms will be enhanced.

OCR for page 193
Oil Spill Dispersants: Efficacy and Effects This page intentionally left blank.