Acute Toxicity of Fire-Retardant and Foam-Suppressant Chemicals to Hyalella azteca (Saussure)
Discussion
Water quality
The pH variations observed in the tests probably did not produce biased results because amphipods are tolerant of pHs as low as 5.7 [19]. Similarly, dissolved oxygen concentrations in the amphipod test treatments exceeded the lower limit set by the ASTM [13].
Amphipods were significantly more sensitive to four of the five chemicals when exposed to them in soft water than when the amphipods were exposed to the chemicals in hard water. The increased sensitivity of amphipods in soft water may be related, in part, to the physiological state of the test animals. The ASTM [13] recommends culturing test animals in the dilution water that will later be used for the test solutions. Amphipods used in these tests were cultured in hard water. Animals used in these tests were allowed to acclimate to dilution waters over a period of 48 h and were fed as recommended by the ASTM [13]. This acclimation period may not have been long enough to alleviate some of the stress placed on the animals during testing. However, mortality was not problematic during acclimation, and control mortality averaged 10% in soft water tests, whereas control mortality never exceeded 10% in hard water tests. This control mortality did not exceed the criterion for an acceptable test with the amphipod, which is >/=80% control survival [20].
Acute toxicity
The toxicity of fire-retardant chemicals to the amphipod Gammarus pseudolimnaeus reported by Johnson and Sanders [7] did not increase over time, in contrast to what was observed in the present tests. They reported that toxic effects of fire-retardant chemicals were observed within the first 24 h of their tests and did not change significantly throughout the test duration. In the present study, toxicity of FT GTS-R, FT LCG-R, and PC D75-F doubled between the 24-h and 96-h observations, thus indicating a possible delayed effect.
The 96-h LC50s for G. pseudolimnaeus in their tests conducted in hard water (hardness 272 mg/L as CaCO3) was 62 mg/L for Fire-Trol 100 (FT 100), which was about six- fold lower than our value for FT GTS-R, a similar chemical. Both FT 100 and FT GTS-R contain ammonium sulfate, but FT 100 contains 35% clay and dichromate as a corrosion inhibitor, whereas FT GTS-R contains no dichromate and uses 3 to 4% guar gum in place of the clay to control viscosity. The dichromate may have been an important toxicant in the FT 100 tested by Johnson and Sanders [7]. Similarly, the 96-h LC50 for Fire-Trol 931 (FT 931) to G. pseudolimnaeus in hard water was 55 mg/L, whereas our value for FT LCG-R, which differs from FT 931 primarily in the manufacturer of the polyphosphates (C. Johnson, personal communication), was 10 times higher.
The two Phos-Chek compounds, PC 202 and PC 259, tested by Johnson and Sanders [7] with G. pseudolimnaeus had 96-h LC50s of 52 and 40 mg/L, respectively. Phos-Chek D75-E a combination of monoammonium phosphate and diammonium phosphate with guar gum thickener, is similar to these compounds, except that PC 202 contained only diammonium phosphate with carboxylmethyl cellulose as the thickener and PC 259 contained only diammonium phosphate with guar gum as the thickener. However, our LC50 value was eight to 10 times higher for PC D75-F. Interlaboratory comparison of these results produces a ratio of 10.3 for Fire-Trol chemicals and 9.8 for Phos-Chek chemicals. These ratios are more than two times higher than the four-fold variation typically found between laboratories [21]. This substantial difference in toxicity suggests that G. pseudolimnaeus is more sensitive than H. azteca, or that the compounds tested in the present study are substantially less toxic than the compounds tested by Johnson and Sanders [7], or a combination of these two possibilities.
In general, foam-suppressant chemicals were more toxic to amphipods than fire retardants. Foam-suppressant chemicals are typically composed of approximately 30 to 40% surfactant, which is the most likely toxic constituent. These surfactants lower the surface tension of water thereby interfering with the animals' ability to obtain oxygen from water [10].
The actual toxic threshold of surfactants is dependent, in part, on their carbon-chain length [22]. Both PC WD-881 and Silv-Ex contain anionic surfactants of unknown carbon-chain length. Sanchez Leal et al. [8] reported a linear increase in toxicity to Daphnia magna with increasing surfactant molecular weight. No information was found on the toxicity of surfactants to amphipods. Daphnia magna has been reported to be more sensitive to surfactants than other invertebrates [9,22].
Ammonia
The toxicity of ammonia is apparently species specific for invertebrates and fish. Macroinvertebrates are reportedly less sensitive to ammonia than fish species [23]. Flow-through tests determined that ammonia was acutely toxic to 19 freshwater macroinvertebrate species at concentrations ranging from 0.53 to 22.8 mg/L, whereas ammonia toxicity to 29 fish species ranges from 0.083 to 4.60 mg/L [23]. Studies conducted by Williams et al. [24] reported 96-h LC50s for ammonia ranging from 0.71 to 2.95 mg/L for 11 macroinvertebrate species. The crustacean species in their study were less sensitive to un-ionized ammonia than non-crustacean species. They reported a 96-h LC50 of 2.05 mg/L for G. pulex exposed to ammonia in moderately hard water (hardness 98-106 mg/L as CaCO3). In contrast, Monda [25] reported a 96-h LC50 for Chironomus riparius of 9.4 mg/L.
The U.S. Environmental Protection Agency [23] has established an ammonia concentration criterion of 0.02 mg/L of unionized ammonia as the concentration below which all aquatic life may be protected. The un-ionized ammonia concentrations of the fire-retardant chemicals, but not the foam suppressants, exceeded this criterion as much as 120 times (Table 2). Un-ionized ammonia is believed to be more toxic to aquatic organisms than ammonium (ionized ammonia)[23]. The concentration of un-ionized ammonia is greater than ammonium when the pH is high and studies have indicated that ammonia toxicity increases with increasing pH. Ammonium was considered to be nontoxic or at least significantly less toxic [26]. Erickson [27] proposed a joint model in which un-ionized ammonia contributes more to ammonia toxicity at higher pHs and ammonium contributes more at lower pHs. Recent studies support Erickson's model, citing total ammonia as nitrogen toxicity to H. azteca [28-30] rather than the un-ionized form, which is toxic to Ceriodaphnia dubia [31]. Borgmann [28] tested the toxicity of ammonia to H. azteca and reported a 96-h LC50 of 28 mg/L total ammonia as nitrogen, which he extrapolated from a mortality curve for a chronic toxicity study. Likewise, Ankley et al. [30] reported 96-h LC50s of 20 to 23 mg/L total ammonia as nitrogen for H. azteca.
Ammonia toxicity is also dependent on water hardness [28,29]. Ankley et al. [29] reported that in soft water (hardness 42 mg/L as CaCO3), the toxicity of ammonia was similar at pHs 6.5, 7.5, and 8.5. In tests with harder water, toxicity of ammonia became more pH dependent. In moderately hard water (hardness 100 mg/L as CaCO3), the LC50 doubled between pH 6.5 and 8.5, and in hard water (hardness 240 mg/L as CaCO3) the LC50 increased six-fold between pH 6.5 and 8.5. In addition, the LC50s between hardnesses increased as much as 10-fold at pH 6.5. Their results indicated that as water hardness increased, a joint toxicity between ammonium and un-ionized ammonia occurred, which coincides with the joint model for ammonia toxicity that also considers the contribution of ionized species, as proposed by Erickson [27]. Our results are similar to those of Ankley et al. [29] because in the hard water tests with the three fire retardants, the amphipods were as much as eight times less sensitive to ammonia than in the soft water tests. These values indicate that the toxicity of the ammonia-based fire retardants was probably influenced by the concentration of total ammonia as nitrogen (the sum of ammonium and un-ionized ammonia) rather than unionized ammonia.
The ammonia constituent has been determined to be the toxic portion of fire-retardant chemicals to fish [32]. However, toxicity of the ammonia constituent of fire-retardant chemicals may be influenced by other constituents such as spoilage and corrosion inhibitors. Both FT LCG-R and PC D75-F, in soft water, contained less total ammonia as nitrogen than has been shown to be toxic to H. azteca (Table 2) [28,29]. Other toxins such as corrosion or spoilage inhibitors may be synergistic with the ammonia constituent thereby increasing the toxicity of ammonia. In contrast, all the fire retardants tested in hard water contained more total ammonia as nitrogen than has been reported to be toxic to H. azteca. Other constituents of the compounds may be antagonistic with the ammonia constituent, thereby ameliorating the toxic effect of the total ammonia as nitrogen when it exceeds reported LC50 values. For example, rainbow trout (Oncorhynchus mykiss) exposed to a mixture of nitrate and ammonia produced antagonistic results, except when the chemical ratio was low, whereas copper and ammonia were reported to be synergistic [33].
Trophic interactions
Amphipods may be the most abundant year-round macrobenthos [34] inhabiting lotic and lentic systems. They are a major link between top carnivores and the large energy store contained by detritus and its associated microflora and fauna. However, in feeding studies, amphipods have demonstrated a preference for algae over detritus [35].
In tests exposing Selenastrum capricornutum to the same fire-retardant and foam-suppressant chemicals as the amphipods of this study, this alga was more sensitive to three of the five chemicals tested [36]. A sublethal response of the alga to four of the five fire-retardant and foam-suppressant chemicals in these tests resulted in a stimulation of growth of 3 to 43% more algal biomass than the controls. An increase in algal biomass at concentrations of fire-retardant chemicals that were sublethal to amphipods would increase the animals' opportunity to graze and hide [35]. However, the fecundity and growth rate of the amphipods might be impaired when they are fed a poor-quality diet. Algae grown in the presence of fire-retardant and foarn-suppressant chemicals may not provide the nutrition required by amphipods because the ideal ratio of phosphorus: nitrogen required for algal growth will be altered by the addition of the chemicals. Sterner et al. [37] reported a decrease in fecundity and growth rate of daphnids when they were fed phosphorus-limited algae. The animals were also sluggish and easily caught, making them easy prey for predators.
Bottom-dwelling invertebrates such as H. azteca would be susceptible to physical impairment when flocculent materials such as guar gum thickener from fire-retardant chemicals accumulated on the substrate. Such materials could also clog the respiratory systems of these animals. Hargrave [38] observed that the egestion rate of H. azteca decreased as the flocculent sediment material was consumed. Concern for secondary effects on aquatic ecosystems was expressed as early as 1977 by Johnson and Sanders [7].
Relationship to environmental considerations
Because environmental factors differ between and along streams and lakes, it is difficult to predict the impact that an accidental introduction of these chemicals would have on aquatic organisms. The impact of an accidental exposure of aquatic organisms to fire-retardant and foam-suppressant chemicals is dependent on a number of factors such as: the route of entry, behavior of the chemical, magnitude of spill, water velocity, shape of the streambed, and substrate composition [39]. A study was conducted to measure the chemical changes in stream water quality using five streams possessing different characteristics [40]. Norris and Webb [40] reported that the physical characteristics of streams and the characteristics of the chemical drop both affected the impact of the chemical in the stream and that water chemistry changes were detected as far downstream as 2,700 m.
A recent incident was reported on September 17, 1995, when a C-130 retardant bomber dropped a partial load of Fire- Trol LCG-F across about 67 m of Murderers Creek in the South Fork John Day River, Oregon (T. Unterweger, personal communication). An estimated 23,000 fish were killed along 2,740 m of stream, including about 718 rainbow and steelhead trout. Murderers Creek is the most significant steelhead trout production stream in the South Fork John Day River sub-basin, and the fish losses were considered biologically significant. Previous fish kills have been attributed to similar introductions of fire-control chemicals, but have not been well documented [39,40].
The application of fire-fighting chemicals is accomplished using a wide variety of aircraft equipped with different storage tank configurations, door sizes, and sequencing speed, which results in a wide range of drop patterns [1]. Drop patterns and the resulting ground patterns are also modified by drop height, retardant type, relative humidity, temperature, and wind speed and direction. Although there has been a substantial effort to improve the performance of fire-retardant delivery in fixed-wing aircraft, the large number of combinations of aircraft types and tank configurations, retardant types and tanks, and environmental conditions precludes a straightforward comparison of laboratory toxicity data to potential effects in aquatic ecosystems.
Nevertheless, the wide variety of aircraft delivery systems have been modified to deliver similar amounts of fire-fighting chemicals. Retardant use ranges from 0.41 L/m2 (1 gallon per 100 square feet; gpc) for fires in annual and perennial grasses or tundra to >2.44 L/m2 (>6 gpc) for fires in mixed chaparral or heavy slash [1]. Fire-retardants are prepared for field use by mixing 1.66 pounds of FT GTS-R per gallon of water to produce 1.1 gallons of slurry, which is equivalent to 198,930 mg/L. Field mixtures for other fire-fighting chemicals are given in Table 3. Comparing the concentrations of field mixtures to the acute toxicity values for the three fire retardants for arnphipods gives ratios ranging from 365 for PC D75-F to 3,704 for FT LCG-R tested in soft water. Thus, an accidental drop of PC D75-F in an aquatic environment would require a dilution of 365- to 2,713-fold to dilute it to a concentration equivalent to the 96- h LC50-a concentration that would cause a substantial amount of mortality. Similarly, application of FT GTS-R would have to be diluted to at least 548- to 1,566-fold to reach a concentration equal to the 96-h LC50; for FT LCG-R the dilution would have to be 505- to 3,704-fold.
A safety factor may be applied to toxicity data to estimate a safe concentration for aquatic organisms. The basis for a safety factor is the same rationale as an application factor, that is, the ratio of the highest observed no-effect-concentration (NOEC; lower limit of the maximum acceptable toxicant concentration) to the acute toxicity value (i.e., 96-h LC50) [26]. An application factor of 0.01 is typically used; the inverse gives 100 as a safety factor to use in estimating a possible NOEC. Similar approaches to extrapolating from toxicity data to environmental-concern concentrations using assessment factors of 100 and 1,000 have been used in European countries in the hazard assessment process [41,42]. Applying a safety factor of 100 to the toxicity information above would require a 54,800 to 156,600-fold dilution of FT GTS-R in soft water, a 50,500- to 370,400-fold dilution for FT LCG-R, and a 36,500- to 271,300-fold dilution for PC D75-F to approach safe concentrations (i.e., NOEC).
Foam suppressants are applied at about 0.1 to 1% foam, which is equivalent to 10,000 mg/L. Comparing the concentration of field mixtures to the acute toxicity values for the two foam suppressants on amphipods gives ratios for PC WD-881 of 1,000 in soft water and 454 in hard water, and 417 in soft water and 370 in hard water for Silv-Ex (Table 3). Applying a safety factor of 100 to these values would require a 45,400- to 100,000-fold dilution for PC WD- 881 and a 37,000- to 41,700-fold dilution for Silv-Ex to approach safe concentrations.
Previous Section -- Results
Return to Contents
Next Section -- Conclusions
