Northern Prairie Wildlife Research Center
|TABLE 5.--Reported acute toxicities of fire-control chemicals to salmonoids in soft water; LC50 = concentration lethal to 50% of test organisms.|
|Formulation||Species||Life stage or
|Fire-Trol LCG-R||Rainbow trout||Swim-up fry--1.50||42||910-1,413||Gaikowski et al. (1996b)|
|Chinook salmon||Swim-up fry--2.86||40||1,080-1,195||Buhl and Hamilton (1998)|
|Fire-Trol 931||Rainbow trout||Swim-up fry--0.8||40||790-1,000||Johnson and Sanders (1977)|
|Coho salmon||Swim-up fry--1.0||40||930-1,000||Johnson and Sanders (1977)|
|Phos-Chek 259||Rainbow trout||Swim-up fry--0.8||40||94-165||Johnson and Sanders (1977)|
|Coho salmon||Swim-up fry--1.0||40||170-250||Johnson and Sanders (1977)|
|FireFoam 103B||Rainbow trout||0.5b||5.0b||41.1||Chemonics Industries (1993a)|
|FireFoam 104||Rainbow trout||0.5c||4.0c||34.6||Chemonics Industries (1993b)|
|ForExpan||Rainbow trout||N.R.d||N.R.||10.9||Norecol Environmental Consultants (1989)|
|Pyrocap||Mummichoge||0.15||N.R.||45.2||EA Engineering, Science, and Technology (1989)|
bC. Chang, Chemonics Industries (personal communication).
cR. Crouch, Chemonics Industries (personal communication).
dN.R. = not reported.
eFundulus heteroclitus: no data for salmonids; test was conducted in artificial seawater (19-22 g/L salinity) at 19-22oC.
The 96-h LC50s of the two FireFoams reported by the manufacturer (Table 5) are 2.7-3.4 times higher than our values for rainbow trout (Table 1). Conversely, our 96-h LC50s for ForExpan S and Pyrocap B-136 are 2.0-3.5 times higher than those reported by the manufacturer's contract laboratory. These differences in LC50 values are within the expected interlaboratory variation of fourfold for a given toxicant-species combination tested under similar conditions (Schimmel 1981).
Acutely toxic concentrations of NH3-N reported for Pacific salmon range from 0.08 to 1.1 mg/L (Russo 1985). However, because ammonia toxicity to aquatic biota is strongly influenced by pH and, to a lesser extent, by temperature (Russo 1985; USEPA 1986), comparisons of ammonia toxicity values should be limited to studies conducted at similar pHs. The 96-h LC50s we obtained for ammonia expressed as both TA-N (112 mg/L) and NH3-N (0.125 mg/L) are close to those reported by Thurston et al. (1981) for juvenile rainbow trout (100 mg/L as TA-N, 0.152 mg/L as NH3-N) tested at a pH of 6.80 in hard water (hardness, 200 mg/L as CaCO3). The time course of toxicity that we observed for ammonia (all mortality occurred within 24 h) is similar to that observed by Ball (1967) for rainbow trout tested under flow-through conditions at a pH range of 7.86-8.22. Inspection of his toxicity curve indicated that the apparent lethal threshold for ammonia to rainbow trout occurred within 24 h.
It is well established that nitrite is considerably more toxic to fish than is nitrate (Russo 1985), as was observed in this study (Table 2). The 96-h LC50 (0.79 mg/L as NO2-N) that we obtained for nitrite is 2.0-4.2 times higher than those reported by Russo et al. (1974) for juvenile rainbow trout (0.19-0.39 mg/L as NO2-N), but our 96-h LC50 is similar to that reported by Westin (1974) for juvenile chinook salmon (calculated value of 0.88 mg/L as NO2-N). Our 96-h LC50 of nitrate (1,658 mg/L as NO3-N) is close to those reported by Westin (1974) for juveniles of chinook salmon and rainbow trout (calculated values of 1,310 and 1,355 mg/L as NO3-N, respectively) tested in freshwater.
The toxicity of LAS to aquatic organisms generally increases with increasing mean alkyl chain length (Kimerle and Swisher 1977; Macek and Sleight 1977; Lewis 1991), which may account for the wide range in acute toxicity values reported for aquatic organisms (Kimerle 1995). The 96-h LC50 of 5.0 mg/L that we obtained for C11.4 LAS and rainbow trout is very similar to those obtained by Pickering and Thatcher (1970) for a commercial LAS blend with a mean alkyl chain length of C11.3 (Macek and Sleight 1977) and juvenile fathead minnow in static exposures (4.6 and 5.0 mg/L) and flow-through exposures (4.2 and 4.5 mg/L). These researchers also reported that most of the mortality occurred during the first 24 h of exposure. Calamari and Marchetti (1973) obtained a 96-h LC50 of 1.68 mg/L for LAS (65% C11 and C12) tested with rainbow trout, which is about one-third of our value. The difference in toxicity values for LAS between these studies with rainbow trout may be partly due to differences in composition of the LAS mixtures tested.
The 96-h LC50 of SDS that we obtained for rainbow trout (24.9 mg anionic surfactant/L) is about five times higher than that obtained for the same species (4.62 mg/L) by Fogels and Sprague (1977). The disparity in toxicity values between the two studies may be partly due to temperature effects on the rate of toxic activity of SDS in fish. As was observed in our test, Fogels and Sprague (1977) were not able to obtain a definitive lethal threshold concentration for SDS and rainbow trout within 96 h. Differences in LC50s obtained at periods with no apparent lethal threshold may only reflect differences in the rates of response rather than differences in sensitivity (Fogels and Sprague 1977). The rate of SDS toxicity may have been higher at 15°C in their study than at 12°C in our study, which may account for the lower LC50 value at 96 h in their study. Similarly, Hokanson and Smith (1971) reported that the time of lethal threshold for LAS to bluegill Lepomis macrochirus was reduced from 48 to 51 h at 15°C from 20 to 24 h at 25°C.
Comparisons of estimated ammonia concentrations at the 96-h LC50 of the fire retardants and those of NH4Cl indicated that ammonia, expressed as either TA-N or NH3-N, was the major toxic component in these formulations (Tables 2, 3). Estimated concentrations of NH3-N at the 96-h LC50s of Fire-Trol LCM-R and Phos-Chek 259F were not significantly different from that of NH4Cl, which indicated that NH3 was probably the primary toxic component in these formulations. For Fire-Trol LCA-F, only the estimated TA-N concentration at the 96-h LC50 was similar to that of NH4Cl. The lower estimated NH3-N concentration at the 96-h LC50 of Fire-Trol LCA-F (compared with that of NH4Cl) was probably due to differences in pH of the test solutions. The low pH of Fire-Trol LCA-F solutions (5.92-6.56) resulted in a lower percentage of total ammonia as NH3 than occurred in the NH4Cl solutions (pH 6.69-6.78). For example, the percent NH3 in concentrations that bracketed the 96-h LC50 for Fire-Trol LCA-F (0.044-0.054%) was less than one-half of those for NH4Cl (0.105-0.122%).
The conclusion that ammonia was the primary toxic component in these fire retardants is supported by the time course of toxicity. In tests with the three fire retardants and NH4Cl, all mortalities occurred during the first 24 h of exposure. Moreover, the average loss of ammonia in test solutions of the three fire retardants after 96 h (10-19%) was similar to that in NH4Cl solutions (16%). The reduction in TA-N concentrations of about 10-20% over the 4-d test period (average of 2.5-5% per day) probably did not have a major influence on the time course of toxicity in these tests.
Based on comparisons of measured NO3-N and NO2-N in test solutions of the three fire retardants with their lethal concentrations to rainbow trout (Table 2), it seems highly unlikely that these compounds made a significant contribution to the toxicity of the fire retardants tested here. Measured concentrations of nitrate in fire-retardant solutions (≤0.9 mg/L as NO3-N) were at least three orders of magnitude lower than its 96-h LC50. Similarly, measured nitrite concentrations in the same solutions (≤0.02 mg/L as NO2-N) were at least an order of magnitude lower than its 96-h LC50 for rainbow trout.
Although ammonia was probably the major toxic component in these formulations, other ingredients in the formulations may have contributed to the observed toxicity. Without comparative testing of these components, it is not possible to accurately assess their relative toxic contribution.
Because of the proprietary composition of the foams, the types of anionic surfactant(s) used in these formulations are not known. Moreover, the analytical method used for anionic surfactant analysis is neither specific for LAS nor does it indicate the LAS homologs that are present. Consequently, it is not possible to identify the anionic surfactant component(s) in these formulations. However, comparisons of estimated anionic surfactant concentrations at the 96-h LC50 of the foams, C11.4 LAS, and SDS may provide insight into the observed differences in toxicity of these foams to rainbow trout.
In a toxicological review of fire-fighting foams, Norecol Environmental Consultants (1989) concluded that the toxicity of these foams is attributable to anionic surfactants, because the nonsurfactant components have a low order of toxicity to aquatic organisms and they comprise about the same percentage of the total formulation as the anionic surfactants. Assuming that the nonsurfactant ingredients did not significantly influence the toxicity of the foams (Norecol Environmental Consultants 1989), the observed differences in anionic surfactant concentrations at the 96-h LC50s of the foams indicated that the type of anionic surfactants used varied among the foams. If the foams contained the same anionic surfactant(s) or surfactants with similar toxicity to rainbow trout, it seems likely that the 96-h LC50s based on anionic surfactant concentration would have been similar for each of the foams.
Based on estimated anionic surfactant concentrations at the 96-h LC50s of the foams, the toxicity to rainbow trout of the surfactants in the two FireFoams were similar and greater than those of the surfactants in the other three foams (Table 4). Comparisons of estimated anionic surfactant concentrations at the 96-h LC50s of the foams with those of the reference surfactants indicated that C11.4 LAS may serve as suitable model to characterize the acute toxicity of ForExpan S and Fire Quench to rainbow trout, whereas it may underestimate the toxicity of the FireFoams and overestimate the toxicity of Pyrocap B-136 (Tables 2, 4). Conversely, SDS was considerably less toxic to rainbow trout than were the foams (based on anionic surfactant concentrations), and thus SDS probably would not be useful in characterizing their toxicity.
During fire-control operations, fire retardants and fire-suppressant foams are applied using a variety of techniques that employ fixed-wing aircraft, helicopters, and ground engines. Although much effort has been expended on improving the delivery of these chemicals to the target areas, accidental inputs of these chemicals into streams occur when the mixing or application sites are near streams. Using simulation models of retardant drops on mountain streams, Norris and Webb (1989) concluded that fish mortality could occur as far as 10,000 m below the drop site, depending on the characteristics of the application and stream.
Accidental drops or misapplications of fire-control chemicals can result in large volumes of these chemicals being deposited into aquatic systems. In Oregon, fish kills have been documented in the South Fork of the John Day River (1995), in Hidaway Creek (1996), and in the North Fork of the John Day River (1997) after these water bodies, or tributaries of these water bodies, were accidentally treated with a fire retardant during fire-control operations (T. Unterwegner, Oregon Department of Fish and Wildlife, personal communication). Using field data and several assumptions about the application technique, Buhl and Hamilton (1998) calculated a peak concentration of 755 mg/L for Fire-Trol LCG-F in the South Fork of the John Day River following an accidental drop in one of its tributaries in 1995. This concentration is close to the 96-h LC50s that we obtained for two similar Fire-Trol formulations tested with rainbow trout (Table 1), and it was similar to that observed during a fish kill (estimated at 23,188 specimens, 718 rainbow trout/steelhead) in this reach of the river (T. Unterwegner, Oregon Department of Fish and Wildlife, personal communication).
In order to assess the potential hazard of these chemicals to salmonids, toxicity data must be related to expected or actual environmental exposure concentrations (Cairns et al. 1978). In the early phases of the hazard evaluation process, acute toxicity data for representative species are compared to estimated environmental concentrations to determine the margin of safety. This safety factor coupled with the use pattern and fate of the chemical are used as decision criteria to assess the need and direction of further testing (Maki 1979). Because of the lack of data on measured concentrations of these chemicals in aquatic systems, toxicity values were compared to their field application concentrations in tank mixtures (Table 6). The ratio of the field tank mixture concentration to its 96-h LC50 value indicates the amount of dilution required in a receiving water in order to obtain a concentration lethal to 50% of the fish within 96 h. Based on these ratios, Phos-Chek 259F and the two FireFoams require the largest dilutions (769-820-fold), and Pyrocap B-136 requires the smallest (64-fold) to achieve a concentration equal to its 96-h LC50.
|TABLE 6.--Concentrations of fire-control chemicals in field tank mixtures and the ratios of mixture concentration to its lethal and nonlethal concentrations to rainbow trout. LC50 = concentration lethal to 50% of test organisms.|
|Chemical||Field tank mixture||Ratio of field tank mix to:|
|Fire-Trol LCA-F||1 gal : 5 gal||243,673||259||518|
|Fire-Trol LCM-R||1 gal : 4.25 galc||273,919||240||351|
|Phos-Chek 259F||1.14 lb/gal||136,617||813||1,752|
|aWeight or volume of chemical
concentrate combined with water to produce a recommended tank mixture
(Lavin 1997). To convert gallons (gal) to liters, multiply by 3.79; 1.14
lb/gal = 137 g/L.
bNAEC = no acute effect concentration; highest concentration with 0% mortality at 96 h.
cC. Johnson, Intermountain Fire Sciences Laboratory (personal communication).
Considering that many populations of Pacific salmon are in decline (Nehlsen et al. 1991), margins of safety based on 96-h LC50s may not be adequate. One conservative approach would be to use the highest test concentration with no mortalities as the acute safe concentration for each chemical and then compare these concentrations to those in field tank mixtures in order to derive appropriate dilution factors (Table 6). Based on these comparisons, Phos-Chek 259F and the two FireFoams require the largest dilutions (1,000-1,752-fold) in receiving waters to achieve nonlethal concentrations to rainbow trout. Consequently, these chemicals pose the greatest acute hazard to rainbow trout (and other species with similar or greater sensitivities) during fire-control operations.
The chronic hazard posed by these chemicals to salmonids during fire-control operations is believed to be low, because long-term exposures are not likely as a result of the transient nature of the chemical plume in streams and biodegradation of the major toxic components in the formulations; ammonia to nitrate (Norris and Webb 1989) and parent anionic surfactants to less toxic intermediates (Kimerle 1995). Moreover, the bioavailability of LAS (and presumably other anionic surfactants) is reduced as a result of adsorption and binding with solids and dissolved organic matter (Kimerle 1995). However, the lethal and nonlethal concentrations of fire-control chemicals that we obtained in this study are based on observations over a 4-d period in the laboratory; consequently, it is not known if there would be any delayed effects in exposed fish after 4 d, nor is it known how environmental factors, such as temperature and solar radiation, influence the toxicity of these chemicals.