Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
3 Ecological Effects: Summary and Evaluation of Available Information As pointed out earlier, most spill assessments are directed toward spills of floating oil. Thus, as with aspects of spill behavior and spill response discussed in the previous chapters, this chapter often compares and contrasts the potential ecological effects of spills of emulsified fuels with those of floating oils. The ecological effects associated with spills of floating oils (including crude oil and distilled products) can generally be categorized as: (1) physical effects associated with smothering or coating on birds, mammals, or other organisms; (2) acute lethal and sublethal toxicological effects of component compounds (such as PAH) on exposed organisms; (3) long-term effects from persistent oil residues or recalcitrant component compounds in sheltered environments or permeable sub- strates; or (4) acute impacts that lead to long-term adverse impacts on population dynamics (e.g., a small spill at a sensitive breeding location). In addition to the acute and long-term effects associated with spills, there has been growing awareness in the ecological community that the multitude of small but frequent releases of hydrocarbons (e.g., urban runoff, permitted discharges from treatment plants, vessel operational discharges) to the environment results in a significant load of potentially toxic compounds (albeit at sublethal concentra- tions; National Research Council, 1985). Although spills may result in spatially and temporally restricted areas of high concentration of potentially toxic com- pounds, such releases generally account for a very small percentage of the total load delivered to the environment. Conversely, if the background concentration of toxic compounds attributable to chronic loads is high in a given area, small spills that would normally dilute rapidly to below levels of concern may raise the ambient background concentration above a concentration of concern for a signifi- 44
ECOLOGICAL EFFECTS 45 cant length of time. Thus, the physical, acute, long-term, and/or cumulative effects of a release of floating oil are complex to evaluate. Ecological effects associated with spills of emulsified fuels can be discussed in a similar fashion, and their effects will differ based on the physical, chemical, and biological be- havior of the component compounds and the setting in which the spill occurs. UNDERSTANDING TOXICITY The physical effects associated with spills of petroleum products (e.g., smoth- ering or coating of floating oil on birds, mammals, or other organisms) are well documented (Boesch and Rabalais, 1987; National Research Council, 1985; Wells et al., 1995). However, toxicology studies conducted to date on the enormous number of individual hydrocarbon compounds and mixtures (such as crude oils, distillate products, or emulsified fuels) are more complex. Thus, a basic intro- duction to the approaches used to understand the chemistry of these complex mixtures as well as the acute (lethal) and sublethal toxicological effects seems appropriate. The primary chemicals of initial concern from a spill of Orimulsion or other fossil fuel will be the toxic PAH. Most PAH will be released rapidly because these compounds are already dissolved in the water phase (30 percent) of the product. Although no direct measurement of the concentration of soluble PAH in Orimulsion has been made, estimates are that only about 15 to 30 µg/L (parts per billion [ppb]) would be present in the water phase of the product (French, 2000). Analyses by Ostazeski et al. (1998a, 1998b) of a nominal concentration of 3,550 mg/L Orimulsion in seawater found only 1 ppb of PAH in the filtered (soluble) phase and 16 ppb when this same amount was added to fresh water. The esti- mates and available data discussed in Chapter 2 suggest that the concentrations of PAH in the water during bioassays and in the field during a spill could never exceed 30 ppb. It is true that additional slow dissolution of PAH from the bitumen particles will contribute to potential biological effects, but the extent of dissolution is a function of the dilution with water, limiting the maximum con- centration to 15-30 ppb. These chemical and physical factors will result in continuous exposure, since the PAH in the water are replenished by PAH from the bitumen, but the total concentration to which organisms would be exposed should never be higher than about 30 ppb. Standard procedures with reference oils were developed to extract the water- soluble components from crude and fuel oils and to conduct tests on a range of tolerant and sensitive species. These tests were designed to measure both acute and chronic effects, as well as bioaccumulation of petroleum hydrocarbons. From these investigations in the 1970s, it became apparent that the toxic chemicals from oils were the monoaromatics (BTEX), the two-ring aromatics (e.g., naph- thalenes) and the three-ring compounds (e.g., phenanthrenes). Laboratory and field studies also showed that naphthalenes, phenanthrenes, fluorenes, and
46 SPILLS OF EMULSIFIED FUELS dibenzothiophenes accumulated in tissues when animals were exposed to water extracts of oils. Toxicity testing with pure compounds showed that mortality did occur when animals were exposed to BETX compounds. However, these com- pounds are volatile and evaporate rapidly, so a constant supply would be needed for mortality to occur. The two- and three-ring compounds were more toxic and were retained in solution and tissues longer. High-molecular-weight compounds were generally low in solubility and would not reach concentrations in the water that could produce acute toxicity during the normal four days (96 hours) of exposure. However, several of these four- to six-ring compounds, such as the six-ring benzo[a]pyrene, are carcinogenic and can produce chronic toxicity from long-term exposure in sediments. Long-term effects may involve genetic dam- age, resulting in reproductive failure, or histological damage as observed in some fish species. More recently Ankley et al. (1995) found photoenhancement of toxicity when some of the PAH are present in the tissues of organisms exposed to ultraviolet light. In addition to PAH that are measured routinely in water, sediment, and tissue samples and that have been investigated for their toxic effects on organisms, it should be recognized that any weathered fossil fuel may contain other compo- nents capable of producing toxicity. Recently Roland et al. (2001) described effects on marine mussels from what is generally referred to as an unresolved complex mixture (UCM). The monoaromatic UCM produced sublethal effects on the mussels, and mussels collected from a spill site contained levels of mono-, di-, and tri-aromatic UCM hydrocarbons. Toxicity studies with oils and petroleum compounds begin with acute toxic- ity tests. In these tests, organisms are exposed to high concentrations of the compound of interest for 96 hours. The results are expressed as the concentration that produces 50 percent mortality in 96 hours (96-h LC50). Less toxic com- pounds will have higher LC50 values, since it will take higher levels to produce 50 percent mortality. The LC50 values for the water-soluble fractions of reference oils and specific hydrocarbons, using a range of animals, provided the early evidence about which specific components of oils were responsible for the major- ity of the toxicity observed. Once data were obtained on mortality, sublethal and chronic toxicity testing was conducted to observe the magnitude of difference between the results of acute tests and chronic exposure parameters (growth, reproduction, physiology). These tests serve as a guide for toxicity associated with specific compounds. However, aquatic toxicology has been criticized in recent decades for embracing the intellectual framework associated with toxicity bioassays to evaluate the safety of potentially harmful chemicals. Bioassays were not originally developed to determine safe exposure levels. Toxicant concentrations that fail to cause sub- stantial mortality within a few days of exposure might nonetheless be sufficient to seriously harm populations. These problems are extremely difficult to assess with bioassays for mixtures of compounds (Box 3.1). Further studies employ-
ECOLOGICAL EFFECTS 47 BOX 3.1 Weaknesses in the Use of Standard Bioassay Conditions in Assessing the Toxicity of Complex Mixtures, such as Orimulsion-400 and Fuel Oils Standard bioassay procedures as described in American Society for Testing and Materials (ASTM) or EPA documents are most appropriate for testing single water-soluble chemicals or effluents with low particulate content (about 30 ppm [parts per million] or less). Mixtures of petroleum hydrocarbons are particularly difficult to use in bioassays. This is because there are both soluble and relatively insoluble phases, with some of the components floating in suspension and some sinking. In addition, the composition of this mixture in solution is constantly chang- ing. Recent bioassay procedures for testing the toxicity of sediments were not designed to test sediments spiked with mixtures such as Orimulsion- 400 and fuel oils, because one portion of these mixtures will be soluble, and then volatile, whereas another portion will remain coating the substrate. Bioassay organisms that do not come in contact with or ingest floating or suspended oil droplets are exposed primarily to the soluble components. Anderson et al. (1974a,b) demon- strated that within 24 hours about 50 to 80 percent of the soluble and toxic aromat- ic hydrocarbons from crude and fuel oil in aerated bioassay containers had volatil- ized. They also showed that there was little or no change in the LC50 from 24 to 96 hours. Therefore, the use of 96-hour bioassays to compare the toxicity of different mixtures of petroleum hydrocarbons is not very appropriate. Probably the best method of assessing the toxicity of complex mixtures, includ- ing dispersed oils, is a constant flowing exposure system where the exposure concentration and time to 50 percent mortality are measured (Anderson et al., 1981, 1984, 1987). Capturing and treating the waste from such testing make the use of this system difficult and expensive. The next best alternative is to expose organisms for 24 hours to known concentrations of the soluble and toxic hydrocar- bons derived from the test products (water-soluble fractions of fuel oils and filtered fraction of Orimulsion-400). Bioaccumulation studies have shown that the toxic naphthalenes and phenanthrenes are accumulated by exposed organisms (Ander- son et al., 1974b), which helps to explain the observed toxicity. Insufficient data are available on the bioaccumulation of these compounds or other hydrocarbons from the soluble or particulate phase of Orimulsion-400. Given all of the problems associated with characterization of the conditions that result in significant toxicity (time and concentration of components), it is difficult to compare the results of toxicity tests with one product, such as Orimulsion-400, to those conducted on fuel oils presently transported and used in the United States. This comparison is further complicated by the fact that the majority of the PAH (including the toxic naphthalenes and phenanthrenes) contained in Orimulsion- 400 are in 1- to 80-µm particles, while these compounds are initially present in the floating fuel oils and then partition into the water phase during a bioassay. A weakness in many of the toxicity study comparisons used to produce data on both No. 6 fuel oil and Orimulsion-400 is that the concentrations of some of the most toxic components (naphthalenes and phenanthrenes) were not measured at the beginning and end of the exposures. It is possible to estimate these concentra- tions, based on other studies (Ostazeski et al., 1998a, 1998b), but the accuracy of these estimates and the duration of exposure to these compounds are in doubt.
48 SPILLS OF EMULSIFIED FUELS ing methods specifically designed to evaluate chronic or sub-lethal effects from exposure to Orimulsion, and its components, should be carried out. With the previous discussion in mind, it is still possible to draw some general conclusions about potential toxicological effects associated with spills of Orimulsion. Available published toxicological studies examine the effects of two formulations of Orimulsion: Orimulsion-100, discontinued in 1998, and Orimulsion-400. Conclusions drawn in this report are applicable to either formu- lation unless specifically stated otherwise. The soluble PAH present in Orimulsion occur at very low concentrations, compared to other heavy fuel oils and crude oils (Anderson et al., 1974a; Table 2.2). Even from spills of large amounts of crude oil, the water column seldom contains a high enough concentration for a sufficient period of time to produce acute toxicity. The concentration of toxic hydrocarbons entering the water dur- ing a spill of Fuel Oil No. 6 would be approximately 10 times greater than that introduced by a spill of Orimulsion-400. It can therefore be assumed, that under identical conditions, the concentration of soluble PAH with known acute and long-term toxicity (naphthalene; two-ring aromatics and phenanthrene; three- ring compounds) would also be lower than from a spill of an equal volume of Fuel Oil No. 6 (Anderson et al., 1974b; Neff and Anderson, 1982). Still, an examination of the potential concentration of compounds of concern from a spill of a given volume at a specific location and time is necessary to fully evaluate the environmental risks. Furthermore, many of the same chemicals that are toxic are also those that would be present in the water column. Bioaccumulation from contaminated sediments, or sediments mixed with bitumen droplets, would in- volve high-molecular-weight compounds (Meador et al., 1995). The very soluble alkylbenzenes (BTEX) are nearly absent from Orimulsion, so no effects from these chemicals are likely. Consideration should be given to those compounds added to bitumen to disperse and stabilize the emulsion. Intan 400 AE constitutes 0.13 percent of Orimulsion-400. Tank studies by Ostazeski et al. (1998a, 1998b) showed that the addition of 14.2 liters of Orimulsion-400 to 4,000 liters of water (nominal con- centration of 3,550 ppm) produced an AE concentration of about 3 ppm within 30 minutes and this level was maintained in solution in both the seawater and the freshwater tanks for the entire 168 hours of testing. The fact that AE is present in the water phase of Orimulsion at concentrations 100 times higher than soluble PAH added to water during bioassay studies strongly suggests that the observed toxicity in the many bioassays was primarily from AE, with some contribution from PAH. For example, if a nominal concentration of 3,550 ppm Orimulsion was found to be the LC50 in a given test, the water would contain about 3 ppm of AE but at the most 30 ppb of PAH. Based on studies with both classes of chemicals, the AE would be at lethal concentration (Table 3.1) while the PAH would be somewhat less than the lethal concentration.
ECOLOGICAL EFFECTS 49 TABLE 3.1 Acute Effects of Linear and Branched Alcohol Ethoxylate Concentrations (mg/L) AE Algae Crustaceans Fish Selanastrum Pimephales cornutum Daphnia magna promelas (96-h Linear AE (96-h growth) (48-h survival) survival) n-C12-15AE7 1.3b n-C13-15AE7 0.9b 0.6b 1.5 b n-pri-C12-15AE9 0.7b 1.6c 1.2 c 4.0 (LOEC, in 2.0 (LOEC, in 7-d 7-d survival survival tests)b tests)b Branched AE C12AE 7 (Exxal 12-based) 38.7d 6.8d 6.8 d C13AE 7 (Exxal 14-based) 37.2d 5.9d 4.4 d C13AE 7 (two methyl branches) 7.5b 8.6b 4.5 b 4.0 (LOEC, in 4.0 (LOEC, in 7-d 7-d survival survival tests)b tests)b C13AE 7 (four methyl branches) 10.0b 9.4b 6.1 b 6.0 (LOEC, in 2.0 (LOEC, in 7-d 7-d survival survival tests) b tests)b Skeletonema Scophthalmus costatum Acartia tonsa maximus (48-h (48-h growth) (48-h survival) survival) Orimulsion-400 C12-13AE9-22 (GENAPOL X 159) 2â4 Orimulsion-400 (fuel, per se) 500 (LOEC)a 100 (LOEC)a 200 (LOEC)a SOURCE: aBjornestad et al., 1998; bDorn et al., 1993; cKravetz et al., 1991; dMarkarian et al., 1989. The second chemical used in stabilizing the emulsion, MEA, is far less toxic than AE surfactant. Most LC50 values for MEA (Table 3.2) are in the hundreds of parts per million of Orimulsion or two orders of magnitude higher than the soluble PAH concentration. Computer modeling (discussed in Chapter 2) supports the view that AE con- tained in the water phase of Orimulsion would be diluted rapidly in open water. The available data suggest that the LC50 for either 48- or 96-hour exposure may be about 1 ppm, and a seven-day chronic test on the reproduction of Ceriodaphnia dubia reported a No-Observable-Effect Concentration (NOEC) of 0.17 ppm (Harwell and Johnson, 2000). The AE concentration would be at the acute lethal
50 SPILLS OF EMULSIFIED FUELS TABLE 3.2 Toxicity of Monoethanolamine Species End Point Value (mg/L) Water flea (Daphnia magna) LC 50, 24 h 140 LOEL (acute toxicity) 100 Mosquitofish (Gambusia affinis) LC 50, 96 h 337.5 Bluegill (Lepomis macrochirus) LC 50, 96 h 329.2 Golden orfe (Leuciscus idus) LC 50 224, 525 Goldfish (Carassius auratus) LC 50, 96 h 170 LC 50, 96 h (pH 10.1) >5,000 LC 50, 24 h (pH 10.1) 190 Rainbow trout LC 50, 96 h 150 (Oncorhynchus mykiss) Fathead minnow LC 50, 96 h 2,100 (Pimephales promelas) Zebrafish embryos LC 50, 8-cell stage until hatching 3,600 (Brachydanio rerio) NOEL, 8-cell stage until hatching 1,200 Clawed toad (Xenopus laevis) LC 50, 48 h 220 Blue-green algae Population growth (threshold cell 7.5 (Anacystis aeruginosa) multiplication inhibition test) Blue-green algae Population growth (decrease in cell 2.1 (Microcystis aeruginosa) multiplication), 8 d Green algae Population growth (threshold cell 0.970 (Scenedesmus quadricauda) multiplication inhibition test) Green algae Population growth (3% 0.750 (Scenedesmus quadricauda) decrease in extinction coefficient), 7 d 300 Flagellate euglenoid Population growth (>5% (Entosiphon sulcatum) decrease in growth), 72 h Cryptomonad Population growth (5% decrease 733 (Chilomonas paramecium) in cell count), 48 h NOTE: NOEL = no-observed-effect level. SOURCE: Davis and Carpenter, 1997. concentration (1 ppm) after a thousandfold dilution and below the acute sublethal concentration after ten-thousandfold dilution. For most spills the water component of Orimulsion is expected to be diluted by these ratios within hours. Thus, acute lethal and sublethal effects from AE in even large spills of Orimulsion-400 are not likely to be significant (long-term effects are discussed later). COMPARING TOXICITY VALUES AMONG DIFFERENT FUEL TYPES In an effort to place the risk associated with spills of Orimulsion into some context, much of the existing literature makes comparisons among Orimulsion and other petroleum products such as Fuel Oil No. 6 (Table 3.3). Two factors must be accounted for in order to effectively compare available information on
ECOLOGICAL EFFECTS 51 TABLE 3.3 Concentrations of Compounds of Known Toxicity in Orimulsion-400 and No. 6 Fuel Oil in mg/L of Whole Product Compounds No. 6 fuel oil Orimulsion-400 Naphthalenes 46,649 474 Fluorenes 12,767 248 Phenanthrenes 48,927 854 Fluoranthrenes 35,272 191 SOURCE: Ostazeski et al., 1998a, 1998b; Stout, 1999; and American Petro- leum Institute, 1999. the toxicity of a spill of a multiphase fuel such as Orimulsion-400 to other fuels. First, unlike typical fuel oils where the distribution of various compounds of concern is essentially uniform, the distribution of compounds in multiphase fuels may vary dramatically among the components. For example, in any volume of Orimulsion-400, the bulk of the total PAH (3,000 ppm; Table 2.1) is tied up in the bitumen droplets, and the extent and rate of transfer of toxic PAH from these particles to the water and organisms are not well understood. The more dispersed the droplets are, the greater will be the exchange of PAH to the water, but this will also mean high dilution in the water, decreasing the possibility of effects. The second major difficulty in comparing toxicity data for multiphase fuels with those derived for typical fuel oils is due to the behavior of the additives in water. Each component of a multiphase fuel may behave differently. For example, the bitu- men phase of Orimulsion-400 demonstrates unique physical behavior that varies with salinity, etc., while the water phase simply disperses into the water column (see Chapter 2). Together, these two factors make it difficult to compare the LC50 values derived for Orimulsion-400 with available data derived for floating oils such as No. 6 fuel oil. For example, toxicity tests intended to evaluate the risk to organ- isms in the water column from spills of floating oils typically focus on the toxicity of the WSF or water- accommodated fraction (WAF) below the oil slick. Johnson et al. (1998), however, expressed the results of bioassay toxicity tests in terms of the nominal concentration (i.e., amount of total added product) of Orimulsion-400. Thus, it is difficult to directly compare LC50 values based on the hydrocarbons in solutions (WSF) of oils with those reported in terms of added concentration of the whole product of Orimulsion-400. Based on numerous evaluations (French, 2000) and tank tests (Ostazeski et al., 1998a, 1998b) it can be estimated that up to 0.001 percent (30 ppb) of the total PAH (3,000 ppm) in Orimulsion resides in the water phase. Because understand- ing initial concentrations is imperative to understanding the potential toxicologi- cal effects from a spill, these uncertainties greatly limit the usefulness of bioassay
52 SPILLS OF EMULSIFIED FUELS TABLE 3.4 Acute Toxicity of Oil-in-Water Dispersions of Orimulsion-400 to Aquatic Organisms Test Species Test Type (Reference) Orimulsion-400 (ppm nominal) Bacteria Microtox Salt water 2,579-2,800a (Vibrio fischeri) 15-min test (4,3) ALGAE Selenastrum capricornutum Fresh water, 72-h chronic (4) >1,000 (100% stock) Skeletonema costatum Salt water, 72-h chronic (1) 500 Invertebrates Crustacean Salt water, 650 (Acartia tonsa) 48-h acute (1) 10 (egg production) 22-d chronic (1) Sea urchin Salt water, (Lytechinus pictus) 20-min fertilization (3) 5,730 (Paracentrotus lividus) Fertilization (4) 296 Mysids Salt water, (Mysidopsis bahia) 96-h acute (5,2) 24.6-42.7a Rotifer Fresh water, (Brachionus plicatillis) 24-h acute (4) 633 Branchiopod Salt water, 24-h acute (4) (Artemia salina) >1,000 Amphipod Salt water, 601 (10-d) (Corophium orientale) 10- and 20-d chronic (4) 293 (20-d) Crustacean Fresh water, (Thamnocephalus platyurus) 24-h acute (4) >1,000 (100 % stock) Water flea (Daphnia magna) Fresh water, 48-h acute (4) 464 Fish Turbot Salt water, 96-h acute (1) <2,000 (Scophthalmus maximus) Inland silverside Salt water, 96-h acute (5,2) 114-200 a (Menidia beryllina) Rainbow trout Fresh water, 96-h acute (3) 301 (Onchorhynchus mykiss) Threespine stickleback Salt water, 96-h acute (3,2) 3,200 (Gasterosteus aculeatus) a Two values available for this test. SOURCE: (1) Bjornstead et al., 1998 (WAF used); (2) Calabrese et al., 1995; (3) Jokuty et al., 1995; (4) Golder Associates Geoanalysis, S.r.l, 1999; (5) Johnson, 1998. studies completed to date. However, by keeping these relationships in mind and by assuming a maximum concentration of bioavailable PAH of 30 ppb, it is possible to examine and make some general comparisons among previous toxic- ity data (Table 3.4) for Orimulsion-400 and floating oils such as No. 6 fuel oil.
ECOLOGICAL EFFECTS 53 For the reasons expressed above, the best comparisons from Table 3.4 would be the ppm concentrations measured in the water during exposures to both No. 6 fuel oil and Orimulsion-400 that produce either 50 percent mortality in a specific number of hours, or a reduction of 50 percent in a sublethal end point. Some of the most sensitive marine organisms used in toxicity testing are mysids (Mysidopsis) and silversides (Menidia), first used to assess petroleum hydrocar- bon toxicity by Anderson et al. (1974a). Johnson et al. (1998) reported the results of a 96-hour exposure of mysids to a filtered WSF of Orimulsion-400, producing a nominal LC50 of 7,070 ppm of Orimulsion-400. Using the factors for PAH and AE discussed above it can be estimated that the mysids were exposed to 7 ppm of AE and 30 ppb of PAH. Both of these concentrations are near the LC50 values for these chemical classes, so it is only possible to state that the combination pro- duced the observed effects. These investigators also reported an LC50 for Menidia exposed to the filtered WSF from a nominal concentration of Orimulsion of 12,800 ppm. Using the same factors, this exposure would be to 12 ppm of AE and 30 ppb of PAH, and there is certainly evidence from AE testing (Table 3.1) that 12 ppm of AE would be above the LC50 value. There are many other toxicity testing data points in various tables from multiple publications, but without spe- cific chemical analysis of PAH in the water, the findings cannot be readily com- pared to tests with other fuel types. It is clear that the maximum concentration of PAH that can be produced in the water during either a toxicity test or a spill of Orimulsion is less than that produced by floating oils. It is important, however, to consider the additional contribution to toxicity from the AE in this product, as initial concentrations can be toxic to exposed organisms. To measure the bioavailability and effects of PAH from Orimulsion formula- tions at low concentrations (0.01-1.0 mg/L), sea scallops were exposed for 60 days (Armsworthy et al., 1999). No impacts were found with respect to survivor- ship or somatic and reproductive tissue weight, but some increases in clearance rate and absorption efficiency were noted. Analysis of the tissues after the 60- day exposures found the highest concentrations in the digestive glands (30 ppm) and gonads (12 ppm). The accumulation in the gonads after 60 days represented a factor of 100, which is the same magnitude reported for PAH in other studies. The only other study to measure uptake of PAH from Orimulsion was with oysters, but the detection limits for the analyses were so high that the data are not useful. Therefore, the potential for uptake is known for only one test organism, but the findings agree with other studies of bivalves using specific hydrocarbons or extracts of oils. It should be recognized that many of these earlier studies also followed the duration of PAH contamination in the organisms after exposure was terminated, and found clearance of the tissues by metabolic activity, depuration, or both. In the only study that examined the effects of dispersed Orimulsion on wild- life, mallard ducks exposed to 100 ppm Orimulsion-400 in a freshwater holding pond exhibited body weight and histopathology similar to controls (Wolfe et al.,
54 SPILLS OF EMULSIFIED FUELS 2001). Coverage of waterproof plumage was noted to be less in treatment birds in that bitumen residue adhered to 40-45 percent of the birdâs plumage, most where the birdâs body contacted the water. After three days of preening and bathing in clean water, half of that amount remained, indicating that exposure of these birds is likely by ingestion. Although this study found the effects of Orimulsion exposure to be less injurious than those of a heavy fuel oil there was some loss of waterproofing in plumage at a low Orimulsion concentration and four hour exposure duration. Bivalves exposed to PAH from Orimulsion, or from any other type of fuel spill, have the potential to be ingested by mammalian or avian predators. This exposure route to higher trophic levels is not well studied but should be acknowl- edged (Duffy et al., 1994; Fry and Lowenstine, 1985; Leighton et al., 1983). Although vertebrates have a high capacity for metabolizing aromatic hydrocar- bons including PAH (Spies et al., 1996), there is a need for studies addressing the possible reproductive effects of ingestion on predators and other effects on higher trophic levels. Data on ecological effects of the dispersed form of Orimulsion on vascular plants comes from work with Orimulsion-100, which has a similar bitumen com- ponent but different additives (surfactant and magnesium) and viscosity than Orimulsion-400. Given the similarities of the bitumen component of the two formulations, the results can be used to assess the toxicity of exposure of these plants to the droplets and PAH. Data indicate that long-term exposure (21 days) to a water column concentration of 10,000 ppm Orimulsion did not result in mortality to the seagrass Thalassia testudinum (Ault et al., 1995). However, lower concentrations (1,000 ppm) resulted in increased leaf senescence (Ault et al., 1995). Because whole product Orimulsion (not just the bitumen component) was used in this study, it is possible that the surfactant and magnesium compo- nents contributed to this effect. Although Thalassia exhibited a negative response to dispersed Orimulsion, it continued to grow and produce new leaves. This resiliency may be due to shoot and root meristems that are buried in sediment and protected from toxic sub- stances in the water column (Ault et al., 1995). Continued growth and refoliation are also supported by nutrients stored in buried rhizomes. In addition to PAH in the water phase of Orimulsion-400, there is 0.1350 percent (1,350 ppm) AE in the neat fuel. Based on tank testing and simple dilution calculations, only the initial pulse of these chemicals into a closed system would produce concentrations in the sub-lethal effects range for algae (Table 3.3). Acute toxicities to fish of both branched and linear AE generally varies from <l to 10 ppm (Table 3.1). Although the data from linear AE are more extensive than those for branched (Talmage, 1994), it appears that the carp and minnow families are more tolerant than commercial, recreational, and saltwater fish, such as sunfish, trout, or cod. Algae showed little or no observable effects from
ECOLOGICAL EFFECTS 55 exposure to concentrations of one family of branched AE that were acutely toxic to fish and invertebrates (Markarian et al., 1989). Yet branched AE used in experiments by Dorn et al. (1993) were comparably acutely toxic (i.e., in the 4.5 to 10 mg/L range), with the freshwater fathead minnow being the most sensitive and a green alga the least sensitive. However the criterion for the algae was growth, while that for the fish was survival. Acute and chronic toxicities of nominal concentrations of specific surfactants provide only approximations of the risks to the specific type of organism used in the test, and not to the health of the waterway. Branched AE are generally less acutely toxic than linear AE (Table 3.1) and are more resistant to biodegradation so they can be more persistent in the envi- ronment. Talmage (1994), citing works of Dorn et al. (1993), Kravetz et al. (1991), and Markarian et al. (1989), found linear AE to be up to forty-fold more acutely toxic than branched AE to the growth of a common green planktonic alga, and from two- to four-fold more toxic to a freshwater crustacean zooplankter and a common minnow (Table 3.1). In seven-day toxicity studies, Dorn et al. (1993) found that the lowest-observed-effect concentration (LOEC) of branched AE on the survival of the water flea was the same as, or 50% less toxic (depending on the number of branches) than linear AE, and branched AE were equally or two- fold less toxic for the minnow. Bjornestad et al. (1998) studied GENAPOL X 159 (the branched AE in Orimulsion-400) in survival tests with a common ma- rine planktonic crustacean, Acartia tonsa, and reported an LC50 of 2 to 4 mg/L. These concentrations are comparable to those of linear AE and about two-fold more toxic than the branched AE used in two-day tests by Dorn et al. (1993) described above. No studies were performed on fish or algae with the surfactant, but tests with the Orimulsion-400 emulsified fuel, showed that it was up to fifty- fold less toxic to the planktonic crustacean than the surfactant and of low concern to a marine alga and a species of bottom fish. The data for MEA in the water phase of Orimulsion-400 (Table 3.2) show it to be relatively nontoxic, since even Daphnia had a 24-h LC50 of 120 ppm. Thus, because MEA has very low toxicity, evaluation of the potential ecological effects of a spill of Orimulsion-400 should be based on the sum of toxicity and long-term effects of exposure to the PAH and AE in solution from spills in both fresh and salt water. UNDERSTANDING ECOLOGICAL RISKS ASSOCIATED WITH SURFACTANT Although the impacts of petroleum hydrocarbons on marine and aquatic organisms have been studied for years, the long-term chronic effects of the sur- factants used in Orimulsion-400 and many other household products, are less well understood. Consequently, additional discussion of these compounds seems
56 SPILLS OF EMULSIFIED FUELS warranted. Surfactants in oil and bitumen emulsions are predominantly two types: (1) the AE and (2) the alkylphenyl ethoxylates (APE) and nonylphenol ethoxylates (NPE), together referred to here as APE (Karsa, 1998; Talmage, 1994). In 1998 an estimated 465 million pounds of AE and LAS were used predominantly as surfactants in household products in the United States (Stanford Research Institute, 1998). Given their widespread use, their possible toxicologi- cal interactions (i.e., additive effects), their presumed concentrations in some major U.S. rivers, and the risks they pose in aquatic systems, major synthetic surfactants (AE, APE, and LAS) are receiving greater attention from environ- mental protection agencies worldwide. This is especially true since background concentrations of surfactants and their intermediary decay products, together with their additive effects with indigenous pollutants, may already have impacted aquatic resources and human health, notwithstanding the effects from accidental spills of emulsified fuels. Criteria for allowable concentrations of the synthetic commercial blends of surfactants (AE, APE, and LAS) apparently do not exist in the United States. Yet their reportedly low concentrations in the 1980s may only reflect dilution by the very high discharges in U.S. rivers, compared to the smaller rivers in other industrialized nations (Ahel et al., 2000; Naylor et al., 1992). Currently, there is no monitoring program in the United States designed to track changes in loading rates of specific surfactants to rivers. Only one study in the late 1980s reported a concentration of 0.5 ppb above, and 1.1 ppb below, an effluent of the commonly used AE in a river (Talmage, 1994). In view of this, it is instructive to view environmental risks of spilled emulsified fuels from another perspective. A spill of the entire contents of one barge loaded with 3,000 barrels of Orimulsion-400, at 6.2 barrels per metric ton, would release about 1,380 pounds of AE (at 0.13 percent by volume in Orimulsion-400). Were this to occur in the lower Mississippi River at average discharge (about 500,000 cubic feet per second [cfs]), the concentration of surfactant would be about 80 ppb immediately downstream of the spill, if the background levels were zero. This concentration (80 ppb) is about one-fourth of the 0.28 ppm concentration derived for use in risk assessment of AE, esti- mated from stream mesocosm investigations (Dorn et al., 1997). However, it is near the predicted NOEC (110 ppb) for AE, estimated to be 50 to 100 times lower than the predicted environmental concentrations, derived from monitor- ing results of removal rates for AE and other surfactants by sewage treatment facilities (van de Plassche et al., 1999). SPILL SCENARIOS In an effort to understand the ecological risks associated with spills of Orimulsion in a variety of environments, the potential physical, acute lethal and
ECOLOGICAL EFFECTS 57 sublethal effects that may occur in the six spill scenarios are discussed below (based on a review of existing literature). For ecological risk assessment, the six scenarios can be broadly combined into three major groups: (1) marine and brackish (salinities greater than 5-7 psu); (2) fresh water (salinities less than 5 psu); and (3) on land, including wetlands. The discussion for these groupings begins with a general statement about factors contributing to overall effect, fol- lowed by a greater discussion of the potential effects on four major groups of organisms (i.e. wildlife, including birds and mammals; water column resources, including fish and invertebrates; benthic organisms, including epifauna and in- fauna; and nonplanktonic primary producers or vascular plants, including kelp in marine settings and plants and trees in freshwater settings). Marine and Brackish Water Environments As discussed in Chapter 2, some of the dispersed bitumen droplets will coalesce and surface as either tarballs or tar patties. The amount of re-floating bitumen will be a function of spill volume, water density, and current dynamics and is difficult to predict. It could range from zero to as much as one-third of the spilled volume. Floating bitumen tarballs will be very sticky and persist for a long time. Recovery of floating bitumen will vary by spill volume, distance from shore, and oceanographic conditions that would either disperse or concen- trate the tarballs. Effects on Wildlife Floating bitumen is likely to behave much like pelagic tar. Therefore, the primary concern for wildlife will be contact with any sticky tarballs on the water surface or those that strand on shorelines. Even small spills of a persistent oil can impact hundreds to thousands of marine birds, such as the 600-barrel spill of a heavy crude oil off California that killed an estimated 9,000 marine birds (Page and Carter, 1986). Birds that spend most of their time on the water surface in dense flocks (e.g. seabirds, waterfowl) are at greatest risk. The effects of pelagic tar on sea turtles have also been well documented (Witham, 1983). Turtles feed on objects floating at the water surface, and therefore are susceptible to ingestion of tar balls that can block the oral cavity and digestive tract. Floating tar can coat the flippers and the mouth can become coated as the turtle attempts to clean its flippers. Tarballs stranding on the shoreline will be very sticky and adhere to the intertidal substrate. The degree of coating will depend on the volume spilled, the distance from shore, and the energy of the wind moving tarballs toward shore. Complete coating is unlikely, but patches of shoreline may contain a sufficient quantity of the emulsion to impact infauna and epifauna.
58 SPILLS OF EMULSIFIED FUELS Effects on Water Column Resources The bioavailability of PAH bound in the particles has not been determined. This is a factor to be considered in light of the filter feeding of either benthic or pelagic organisms that pass the particles over their membranes or ingest them. The most sensitive filter feeders in the water column would likely be small fish, whose gills could become clogged, or that could transfer PAH to the circulatory system and accumulate it in the liver. By âcoughing,â fish can dislodge particles from the gills. Uptake of PAH from suspended bitumen droplets in the water column by animals is likely to be limited because of low uptake efficiencies of PAH, reflecting slow kinetics and short residence time in the gut (Meador et al., 1995). Thus, the major toxicological threat is posed by PAH already dissolved in the water fraction of Orimulsion and subsequently dissolved from the droplets when diluted. Modeling results for a 10,000-barrel spill of Orimulsion in Dela- ware Bay (French McCay and Galagan, 2001) indicated that total PAH levels dropped below 25 ppb (acute toxicity level for the most sensitive species exposed over an extended period) within about 12 hours. In areas where the concentration of dissolved PAH is rapidly diluted, acute toxicity impacts on water column resources are expected to be low. Impacts would be higher for conditions in which dilution is slower, such as water bodies that are confined, have low flush- ing rates, or are very shallow. Effects on Benthic Organisms Spills of Orimulsion are not expected to affect benthic organisms in deep water because the droplets tend to suspend in the top 3 meters, and any eventual sinking due to adsorption onto natural particles will be slow and widely dis- persed. In shallow nearshore and estuarine settings, benthic organisms are poten- tially at risk from three pathways of exposure. First, the dissolved fraction could have acutely toxic impacts. However, because of rapid dilution with depth benthic organisms would typically be at low risk. Second, suspended bitumen droplets and bitumen-sediment agglomerates could be mixed throughout the water col- umn, where they could be ingested by benthic filter feeders. Sessile organisms will be the most at risk. PAH from Orimulsion can accumulate in shellfish tissues (Armsworthy et al., 1999) resulting in potential transfer to higher trophic levels. Food web magnification of PAH is unlikely however, because organisms at the higher trophic levels often have the greatest capacity for PAH metabolism (Meador et al., 1995). The third pathway of exposure is via tarballs that sink after mixing with sediment, either in the surf zone or after stranding onshore. This sticky oil can coat epifauna as it rolls around on the bottom, especially where the oil accumu- lates in a trough or pit. It is difficult to estimate how much of an Orimulsion spill
ECOLOGICAL EFFECTS 59 would behave this way, but similar behavior has been observed for spills of heavy oil (National Research Council, 1999). Effects on Nonplanktonic Primary Producers Macrophytic algae (e.g. kelps, âseaweedsâ) and vascular plants (seagrasses, mangroves, salt marsh grasses) are important components in coastal intertidal and subtidal communities. These organisms fix carbon and may contribute sig- nificantly to nearshore food webs. They provide vertical and horizontal structure and a substrate for attachment of other organisms, they function as breeding and nursery areas for many other ecologically important organisms, and they stabilize intertidal and subtidal sediment. Orimulsion may interact with coastal nonplanktonic primary producers ei- ther as a sticky floating bitumen residue that is predominantly influenced by wind or as a dispersed, nonsticky fluid moving with currents and tide. Clumps of floating Orimulsion residue reaching the shore may come in contact with emer- gent vegetation of the intertidal zone (e.g. salt marsh grasses, mangroves), adher- ing to plant tissue and forming patches of oiled vegetation. Associated epiphytic organisms within these patches will likely be smothered. There is no information in the literature regarding the potential effects of Orimulsion floating residue on vascular plants, but it is likely that they are similar to those of other weathered heavy fuels (Michel et al., 1995). The dispersed form of Orimulsion may interact with subtidal as well as intertidal primary producers. A study by Ault et al. (1995) suggests that, at least for the seagrass Thalassia, >1,000-fold dilution of Orimulsion (which would occur rapidly under most spill scenarios) would not result in long-term injury to the vegetation from water column exposure. The reported sublethal effects at lower concentrations, however, suggest a potential for short-term stress in af- fected Thalassia populations, which may in turn impact other components of the seagrass community. Potential effects on other species of seagrass are unknown. Chronic effects on the entire community from the accumulation of bitumen bound to the sediments are also uncertain. Because Thalassia has a large proportion of its living biomass buried in the sediment, it is likely to survive a coastal spill of Orimulsion. This suggests that other submerged or emergent vascular plants with high proportions of buried biomass (e.g. Spartina) could survive such a spill as well. However, sublethal impacts, such as leaf chlorosis or death, will affect plant production and may have repercussions throughout the food web. Another important finding from the Thalassia work was the decrease in water column oxygen concentrations in treatments with Orimulsion added (Ault et al., 1995). This effect may have contributed to the invertebrate mortalities
60 SPILLS OF EMULSIFIED FUELS observed and should be considered as a potential risk to these organisms in estuaries where dissolved oxygen levels are already low. Freshwater Environments There are two important differences in the behavior of Orimulsion spills in fresh water versus seawater: (1) because Orimulsion is denser than fresh water, spills will result in dispersal of the bitumen droplets into the water column with no surfacing; and (2) the surfactant and stabilizer associated with Orimulsion are more stable in fresh water, thus there is less tendency to form clumps. Therefore, spills in flowing fresh water such as rivers and streams have a greater tendency to remain completely distributed within the water column, with no surfacing or settling. There will still be some attachment of droplets to natural material, but less than in estuarine or marine conditions, because of the stability of the dispers- ant. As the suspended droplets move down stream, high concentration pockets of Orimulsion may build up on the bottom where there are back eddies or depres- sions. If such an accumulation were to take place, there could be impacts on fish or invertebrates spending significant time at these sites. With an increase in flow, these pockets would be flushed downriver. For spills into quiescent water bodies, the bitumen droplets are expected to form a thick cloud that settles through the water column. Based on tank tests, the droplets would remain dissociated from one another and slowly settle to the bottom. Because the droplets are so small, they would readily be resuspended by any disturbance. The soluble components (PAH, surfactants, stabilizers) would diffuse into the water column above the settling droplets and subsequently move into the surrounding water, slowly diluting. Effects on Wildlife Risks to wildlife in the water column swimming through the dispersed form of Orimulsion-400 are unknown. There is no information on whether bitumen droplets suspended in the water column will stick to fur or feathers. Swimming at the surface of a water body containing dispersed Orimulsion does present a risk to wildlife, however, because coalesced floating bitumen has more of a tendency to stick to fur or feathers. Mallard ducks exposed to completely dispersed (100 ppm) Orimulsion in a pond had bitumen droplets stuck on their feathers, resulting in a reduction in water proofing (Wolfe et al., 2001). It is possible that the agitation of the water column by the swimming of the ducks in this confined water body promoted the coalescence and floating of the bitumen droplets.
ECOLOGICAL EFFECTS 61 Effects on Water Column Resources Because all of the product mixes into the water column, spills of Orimulsion into freshwater environments have a greater potential to impact water column resources than floating oil spills. The extent of impact is directly related to the spill volume and rate of dilution of the PAH and AE associated with the water phase and the PAH desorbed from the dispersed droplets. Dilution of AE by a factor of 1,000-10,000 would bring concentrations below levels of concern for short-term exposures (1,350 mg/L in neat Orimulsion, 1-2 mg/L as an acute LC50, and 0.1 mg/L as a NOEC). The difference between the potential impacts of PAH and AE is that the bitumen droplets contain about 3,000 ppm of PAH; thus, PAH desorption from bitumen droplets after the emulsion is broken provides addi- tional soluble PAH to the water column. The rate and extent of PAH exchange from the bitumen to the water is not well understood, but will be a function of dilution in the water. However, PAH concentrations will not exceed 30 ppb, the estimated initial concentration in the water phase of the fuel. Depending on flow rates, spills into rivers may have a slower rate of dilution than spills into bays and oceans. In rivers, motile species may be able to escape as the plume of droplets and dissolved PAH moves downstream, but nonmotile species could suffer acute toxicity from larger spills. Effects on Benthic Organisms Spills of Orimulsion in freshwater environments have a greater potential than floating oil spills to impact benthic organisms because bitumen droplets are denser than fresh water. Very small bitumen droplets will remain in suspension under all but the most quiescent settings, so filter-feeding benthic organisms could be exposed to the cloud of droplets as they move past. Based on limited data for sea scallops (Armsworthy et al., 1999), PAH uptake from solution will be approximately as described for oils and specific hydrocarbons (Meador et al., 1995), but accumulation from bitumen particles is not clearly understood. The risk of exposure to particles or oxygen depletion would be greatest for the most quiescent settings where bitumen could settle out of the water column and accu- mulate as a cloud of droplets suspended above the bottom. Benthic organisms would also be exposed to dissolved PAH and AE concentrations similar to those encountered by water column resources discussed above. Effects on Nonplanktonic Primary Producers Concentrations of bitumen associated with spills in rivers and harbors will present minimal risk to submerged macroalgae or vascular plants except in the most quiescent settings. Although macrophytes will be exposed to PAH and AE
62 SPILLS OF EMULSIFIED FUELS concentrations similar to those encountered by water column resources, they tend to be much less sensitive, which suggests that spills in freshwater settings pose minimal risk to these organisms. On Land, Including Wetlands Orimulsion spills on land are expected initially to flow downgradient like a heavy fuel oil. However, the emulsion is expected to break, allowing the water phase to drain away, leaving a surface layer of very sticky bitumen. On land and in wetlands, there may be some penetration of free product Orimulsion into coarse substrates. Effects on Wildlife Drying of the product on land and in wetlands would produce a very sticky substance, posing a hazard of coating for any wildlife entering the area. Small furbearers would be at risk of fouling, ingestion during cleaning, and habitat displacement. The risk to biota in the path of the spill would be primarily from smothering. Effects on Vascular Plants Plants may be impacted by spills of Orimulsion by smothering or through toxic effects resulting from exposure to roots or stems. Smothering occurs if large patches of Orimulsion are left in place on the ground surface, reducing oxygen diffusion into soils and interfering with root respiration. Although cleanup on land is possible (even a layer of the product in a marsh-type environment could be removed and recovered), spill cleanup may injure plants through physical dam- age to tissues, soil removal, or soil compaction. Summary of Ecological Effects The ecological effects associated with spills of petroleum can generally be categorized as (1) physical effects associated with smothering or coating (2) acute lethal and sublethal toxicological effects of component compounds (such as polycyclic aromatic compounds) on exposed organisms; (3) long-term effects from persistent oil residues in sheltered environments or permeable substrates or acute impacts that lead to long-term adverse effects on population dynamics (e.g., a small spill at a sensitive breeding location); and (4) effects associated with chronic releases. Thus, the more significant conclusions presented so far regard- ing ecological effects from spills of Orimulsion can be categorized in a similar fashion.
ECOLOGICAL EFFECTS 63 Physical Effects One of the most damaging impacts during spills of floating oils is on birds coated by the oil when landing in or diving through the oil slick. In most instances, however, a spill of Orimulsion would not form a surface slick. In open-ocean or coastal environments (salinty >5psu) Orimulsion tends to coa- lesce, resurface, and form floating tarballs. These tarballs will be highly persis- tent and would contribute to pelagic tar in the ocean that affects both marine birds and turtles. Another key impact from oil slicks is the covering of the intertidal zone with a layer of spilled oil, some of which may move back into the subtidal zone. Complete coverage of the intertidal zone with a coating of Orimulsion is unlikely, because only a fraction of the spill is expected to re-float. It is there- fore likely that impacts from a spill of Orimulsion on birds, marine mam- mals, and intertidal organisms may be significantly less than spills of float- ing oils. Acute Lethal and Sublethal Effects Floating oils, including crude oil and refined products, presently transported and used in the United States contain up to 10 times higher concentrations of PAH than Orimulsion-400. The monoaromatics (BTEX) contribute to the toxic- ity of spilled fuel oils in the early stages of a spill, but these are present in such low concentrations in Orimulsion that they pose little risk to aquatic organisms. In addition, most of the PAH in Orimulsion occurs in the bitumen droplets, and the bioavailability of these compounds to non-filter-feeding organisms will likely be low, since the chemicals would first have to exchange into the water, at very dilute concentrations. The accumulation of PAH by filter feeders from the soluble phase and bitumen droplets is not well understood at present. Stud- ies should be undertaken to fully evaluate the bioavailability of PAH to filter feeders. There are, however, significant problems and complexity in attempting to compare the effects of Orimulsion to those of No. 2 fuel oil or No. 6 fuel oil. Previous research has shown that the soluble components of floating oils (BTEX, naphthalenes, phenanthrenes, etc.) are responsible for the observed toxicity to sensitive aquatic organisms. Bioaccumulation of naphthalenes and phenanthrenes during exposure provides additional evidence of their relationship to the toxic effects on animals. Some organisms are capable of metabolizing or at least depurating these compounds over time if a lethal dose is not received. However, there can be long-term effects from these exposures, such as histological damage in fish. In attempting to estimate the concentrations of toxic PAH in the water during a toxicity test with Orimulsion, it was necessary to use chemical data from tank
64 SPILLS OF EMULSIFIED FUELS tests without animals. Any additional testing with Orimulsion should provide the detailed chemical analyses necessary to make this type of comparison. Whole product testing with sensitive fish species (rainbow trout and silversides) showed that levels of about 100 to 300 ppm produced 96-h LC50 values. These fish were exposed to both the particulate and the soluble phases of the product, as would be the case during a spill, so these values are likely good estimates of the potential acute effects on fish from the dispersed product. Some additional key questions remaining on the impacts of Orimulsion re- late to the bioavailability of PAH from the whole Orimulsion product and from the bitumen droplets only. Short-term bioaccumulation studies for key commer- cial crustaceans and bivalves of the whole product dispersed in water would answer questions about potential uptake and seafood safety. To more fully address this question, standard bioaccumulation tests should be conducted, with exposure of organisms both to the whole product mixed in sediment and to the solid phase only (after filtration to remove the 30 percent water) mixed in sediment. These studies would have to demonstrate both solid-phase toxicity and bioavailability. Long-Term Effects The long-term effects of oil spills are a function of the bioavailability of the chronically toxic, high-molecular-weight hydrocarbons that remain in and on fine sediments. No spills of Orimulsion have occurred in which the potential effects of such coating or mixing with fine sediments could be observed. Studies are recommended that measure the bioavailability of PAH from mixtures of Orimulsion and sediments, and the acute as well as chronic effects of these exposures. Effects from Chronic Releases of Surfactants Surfactants are widely used, they are acutely and chronically toxic, they may interact with or add to the toxicity of other pollutants, and they are not monitored in U.S. waterways. Currently there are no criteria in the United States for allow- able concentrations of surfactants or their biodegraded intermediary products. Current background concentrations of surfactants and degraded products in U.S. waterways are unknown. Only an approximation of risks from spills of emulsi- fied fuels in U.S. waterways can be made, using models that are based mostly on toxic concentrations derived from acute toxicity studies of nominal amounts of the untreated toxicant. Better understanding of the significance of AE and its by-products in the environment would require the establishment of monitor- ing programs for specific types of synthetic, commercial blends of surfac- tants and their more persistent biodegraded intermediary products in U.S.
ECOLOGICAL EFFECTS 65 waterways. Furthermore, the establishment of criteria for allowable (low risk) concentrations of each commercial blend of surfactant and its biodeg- radation products would be needed to minimize environmental and health risks. Long, linearly arranged alkyl chains (up to C20) are less acutely toxic than short chains and branched chains, and AE with longer EOn moieties are less toxic as well. Biodegraded intermediary products (PEG and carboxylated AE) appar- ently persist longer than die-away tests (28-day) have shown; the period and products of ultimate degradation are unknown. Therefore, surfactants with longer, less complexly branched alkyl chains and long-chain EO moieties should be used in emulsified fuels (as was done with Orimulsion). Complexly branched hydro- phobic and hydrophilic moieties should be avoided in emulsified fuels because of their recalcitrance to degradation in sewage treatment facilities. The use of naturally derived surfactants in emulsified fuels should be investigated.
