USGS


Ecotoxicology of White Phosphorus in an Alaskan Tidal Marsh

by
Donald W. Sparling , USGS Patuxent Wildlife Research Center

Skip past Group 1 Links

Patuxent Home

ABSTRACT
INTRODUCTION
ACUTE TOXICITY
SUBACUTE TOXICITY
MODES OF ACTION
REPRODUCTIVE EFFECTS
SECONDARY POISONING
TERATOGENIC EFFECTS
HAZARD ASSESSMENT

ABSTRACT

SPARLING, DONALD W. Ecotoxicology of white phosphorus in an Alaskan tidal marsh. U.S. Geological Survey, Patuxent Wildlife Research Center, Laurel MD 20708-4017.

White phosphorus (WP) has been used extensively by the military for artillery practice and training maneuvers. When WP contacts air it forms a thick, white, obscuring smoke but when it enters wet sediment or water it may remain in pellet form indefinitely. Extensive waterfowl mortality over a 12 year period in Eagle River Flats, a tidal marsh being used for artillery practice, prompted the U.S. Fish and Wildlife Service and the U.S. Army to investigate the toxicity and possible remediation of WP contamination at the site. An associated survey determined that WP is also present in wetlands of several other army bases in the United States.

This paper reviews research conducted on the toxicity of WP to waterfowl and potential predators feeding on dead and dying ducks at Eagle River Flats. Data are presented on the acute and subchronic toxicity of WP to waterfowl, histopathological effects, possible biomarkers of exposure, and teratogenic and reproductive effects. A model hazard assessment summarizes the ecotoxicology of WP.

  INTRODUCTION

White phosphorus (P4) or elemental phosphorus is used by the U.S. military in munitions to trace artillery impacts, as obscurants, and as incendiaries. When it comes in contact with air, P4 forms thick, white clouds of smoke. When in cool, wet soil or water, however, P4 can lie inert for decades and be picked up by feeding waterfowl. Since 1982, extensive P4-caused waterfowl mortality has occurred at Eagle River Flats, a 865-ha tidal marsh within Fort Richardson, Alaska (Fig. 1). At least until 1993, an estimated 1,000 to > 2,000 ducks, swans, and an untold number of shorebirds have died due to ingesting pellets of P4. The substance has been found in 71 National Priority hazardous waste sites in 29 states and in wetlands of at least 8 other U.S. Army installations in the United States.

This poster summarizes research conducted on P4 and birds at Patuxent Wildlife Research Center. We thank the U.S. Fish and Wildlife Service Region 1; U.S. Army, Fort Richardson; Cold Regions Research and Engineering Laboratory, U.S. Army Corps of Engineers; and AEHA for the support and financing of this project.

Figure 1: Location of study area, Eagle River Flats, Alaska.

Fig. 1. Location of study area, Eagle River Flats, Alaska.

ACUTE TOXICITY

The waterfowl species most affected by P4 toxicity at Eagle River Flats include mallards, pintails, green-winged teal, trumpeter swans and tundra swans. We used game farm mallards and mute swans as models to determine the acute toxicity of P4 in waterfowl.

The LD50 for male mallards using pelletized P4 was 3.90 mg P4/kg body weight (95% CI = 3.24 to 4.69).

Probit Mortality = -4.023 + 6.725*Log(dose); P = 0.023.

For mute swans the LD50 = 3.56 mg P4/kg body weight (95% CI=1.37 to 4.24).

Probit Mortality = -3.856 + 6.998*Log(dose); P = 0.028. (Fig. 2).

The median size of P4 pellets at Eagle River Flats is around 1 mg, about the size of many waterfowl food organisms. Thus an average 1200 g mallard would die from ingesting < 5 pellets and a 10 kg swan would need approximately 35 pellets for a lethal dose.

Signs of P4 intoxication include ataxia, excessive drinking, loss of appetite, listlessness, rhythmic head movements and convulsions. The listlessness makes intoxicated waterfowl easy prey for several species of raptors inhabiting Eagle River Flats and the convulsions attract predators, thereby increasing the possiblity of secondary poisoning to eagles, hawks, ravens, and gulls.

Figure 2. Dose-response curves for mallards (solid line) and mute swans (dotted line) dosed with pelletized white phosphorus Fig. 2. Dose-response curves for mallards (solid line) and mute swans (dotted line) dosed with pelletized white phosphorus.

SUBACUTE TOXICITY

Repeated exposures to individual doses of P4, each of which would be sublethal by itself, can increase risk of mortality to mallards. Over a 10-day period 60% of the birds dosed with 3.4 mg/kg P4 and 40% of those dosed with 2.4 mg/kg died (Fig. 3). At the lower dose, no mortality was observed until the eighth day of dosing. The increased mortality was probably due to an accumulation of P4 in fat over the exposure period. Several of the birds dosed with pelletized P4 had intact pellets in their gizzards 6 days after the last dose and dead waterfowl have had `smoking gizzards' upon necropsy, suggesting that birds can carry lethal or sublethal doses of P4 with them when they leave the staging grounds at Eagle River.

Fig 3. Percent mortality of mallards given repeated doses of P4. Figure 3. Percent mortality of mallards given repeated doses of P4

MODES OF ACTION

White phosphorus toxicity appears to have two modes of action. One is fast-acting (1-6 hr) and is characterized by rapid onset of lethargy, followed by head bobbing or a `drunken' appearance and, in the case of mallards, intense convulsions lasting 0.25-1 hr. In swans the convulsions are not as severe and onset may be delayed by a longer retention of P4 in the gizzard. We believe that this mode is created by oxidation of hemoglobin into methemeglobin and by the destruction of red blood cells as both hemoglobin and hematocrit decrease dramatically in birds that have been given a lethal dose of P4 (Fig. 4).

The other mode of action is more protracted and focuses on hepatic and renal damage. Enlarged, discolored livers (Fig. 5) accompanied by hepatocellular vacuolations and necrosis (Fig. 6) occur frequently among birds that have survived for a day or more. Renal effects, although somewhat less frequent, involve damage to the tubules. Increases in plasma constituents such as LDH-L (Fig. 7) and AST (Fig. 8) are associated with this damage and might be used as bioindicators of exposure.

Figure 4. Change in hemoglobin concentration in whole blood following dosing with P4.

Fig. 4. Change in hemoglobin concentration in whole blood following dosing with P4 .

Fig. 5. Normal liver of mallards (A) produced by P4 Fig. 5. Fatty liver of mallards (B) produced by P4
Fig. 5. Normal liver of mallards (A) and a fatty liver (B) produced by P4.
Fig. 6. Photohistogram of a normal (A) liver in mallards. Fig. 6. Photohistogram of a fatty (B) liver in mallards.
Fig. 6. Photohistogram of a normal (A) and fatty (B) liver in mallards.

Fig. 7. Changes in plasma concentrations of LDH-L with P4 dosing.

Fig. 7. Changes in plasma concentrations of LDH-L with P4 dosing.

Fig. 8. Changes in plasma concentrations of AST with P4 dosing.

Fig. 8. Changes in plasma concentrations of AST with P4 dosing.

REPRODUCTIVE EFFECTS

White phosphorus has been found in the eggs of a herring gull and can be transferred to the eggs of chickens through dosing hens. Thus it has the potential of causing reproductive effects either by affecting hens or embryos.

We found that the most obvious affect of P4 on reproduction is a rapid cessation of egg laying which lasted for > 3 weeks when mallard hens were dosed with 2.0 mg/kg. This can be seen in decreased egg-laying rates over a 15-day period (Fig. 9). There were no significant effects on fertility or hatchability. Similarly, there were no observed effects on the breeding success of males that were dosed with 1.0 mg/kg P4.

A delay of three weeks in laying may affect mallard populations in the contiguous United States and southern Canada if there is differential nest success or duckling mortality through the breeding season. In the more compressed breeding season in Alaska, however, a 3-week delay could effectively prevent a female from breeding that year.

Figure 9. Rate of egg laying (2SEM) over a 15-day period for mallards given one of three levels of P4 per day and for controls. Figure 9. Rate of egg laying (2SEM) over a 15-day period for mallards given one of three levels of P4 per day and for controls.

SECONDARY POISONING

Several species of raptors including bald eagles, golden eagles, peregrine falcons, and harriers as well as herring gulls and ravens feed on dead and dying waterfowl. These birds, in turn, may be exposed to P4 either by ingesting intact pellets from gizzards or by eating P4-laden tissue. Although P4 is eliminated from living birds within a day or two either by excretion or metabolism, it may linger in the tissues of dead animals for substantially longer periods. We experimentally exposed kestrels to poultry chicks that had been dosed with P4 and then either dissected to remove the upper digestive tract and hence any pelletized P4 (NOPEL) or kept intact and had a 1.4 mg P4 pellet implanted into the crop (PELL). Kestrels fed PELL chicks decreased feeding (inappetance is characteristic of P4 toxicity) compared to controls and experienced a 55% mortality over 7 days of treatment (Fig. 10). These birds also experienced a significant drop in hemoglobin and increases in LDH-L and AST. Kestrels fed NOPEL chicks experienced a 22% mortality and showed intermediate changes in food consumption, LDH-L, and AST. P4 could be measured in the fat of both groups. The study indicates that raptors are at risk from eating contaminated prey, either by eating gizzards or simply by ingesting tissues. It corroborates the findings of dead eagles in the field and P4 in the eggs of herring gulls.

Fig 10. Mortality of kestrels fed dosed poultry chicks with (PELL) or without (NOPEL) pellets of P4. Fig 10. Mortality of kestrels fed dosed poultry chicks with (PELL) or without (NOPEL) pellets of P4

TERATOGENIC EFFECTS

As described above, P4 has some ability to enter eggs in that traces of P4 have been found in gull eggs and in the eggs of dosed chickens. In a pilot study several ducklings experienced teratogenic effects, the most serious involving scoliosis, spina bifida, micropthalmia, hydrocephalia, and lordosis (Fig. 11). In the full reproductive study we observed other ducklings with deformities. The percent of ducklings with abnormalities were: controls - 0.9%, 0.5 mg/kg - 7.2%, 1.0 mg/kg - 5.3% and 2.0 mg/kg - 21.0% (p for dose = 0.008). The most common deformities included scoliosis/lordosis (36% of those with deformities), edema (32%), hydrocephalia (20%) and bill curvature (12%).

Fig. 11. Deformed duckling from a hen dosed with P4. Fig. 11. Deformed duckling from a hen dosed with P4.

HAZARD ASSESSMENT

Obviously, trying to estimate the hazard P4 poses to waterfowl is a complex process (Fig. 12). First, one would have to estimate the probability that an average duck would be exposed to P4 based on feeding habits, feeding location, duration of stay at a contaminated site, and spatial distribution of P4. At Eagle River Flats American widgeon are the most common species of waterfowl but are seldom found dead from P4 because they tend to strip seeds from aerial shoots of aquatic plants and spend less time rooting in the sediments than the more vulnerable mallards and pintails. Once a duck is exposed, the amount consumed will determine if the effects will be lethal or sublethal. If its a female there may be reproductive consequences or its young may be impaired. Then, will the behavior induced by the P4 make the bird more vulnerable to predation? If the duck is eaten by a predator or scavenger, a similar set of questions have to be asked for the predator.

Fig. 12. Flow chart illustrating risk assessment for P4 in birds using Eagle River Flats. Fig. 12. Flow chart illustrating risk assessment for P4 in birds using Eagle River Flats.


Patuxent Home