USGS - science for a changing world

Northern Prairie Wildlife Research Center

  Home About NPWRC Our Science Staff Employment Contacts Common Questions About the Site

A Review of the Problem of Lead Poisoning in Waterfowl

Mortality from Lead Poisoning


Overlooked Losses

Dead ducks are seldom noticed in the marsh, and most hunters are unaware of the extensive losses of waterfowl caused by lead poisoning. Nevertheless, banding data indicate that approximately one-fourth of all ducks alive in September die from natural causes within the year - slightly more than are killed by hunters. The fall population of game ducks is usually around 90,000,000, although it declined to about 62,000,000 in 1985. With a natural mortality rate of 22.2 percent, a minimum of 14,000,000 to 20,000,000 of these ducks can be expected to die from natural causes each year.

The most convincing data on this aspect of waterfowl population dynamics stem from a study of 134,000 bands recovered from mallards by D.R. Anderson (1975). His analysis showed that adult male mallards suffered an annual mortality of 37 percent; adult females, 44 percent; and immatures, 50 percent. Natural losses accounted for 45 percent of the annual mortality of adult males, for 58 percent of that of adult females, and for 52 percent of that of immatures; hunting was responsible for the remainder of the losses. Most other species of ducks have slightly higher annual mortality rates than those of mallards, and, with the exception of the wood duck, natural losses account for an even higher percentage of the total mortality in these other species (Bellrose 1976).

One might wonder why the death of so many ducks goes largely unnoticed. Our observations indicate that when a duck becomes seriously ill, it leaves its flock and seeks dense cover out of water in marshes and along the shores of lakes. There it becomes a potential meal for a mink, raccoon, fox, coyote, eagle, hawk, owl, crow, gull, or any of a host of other predators.

A rapid disappearance of mallard and Canada and snow goose carcasses was reported by Humburg et al. (1983). At the Squaw Creek National Wildlife Refuge (NWR) in northwestern Missouri, 90 carcasses disappeared at the following rates: 9.4 percent after 1 day, 12.3 percent after 2 days, 36.7 percent after 3 days, and 62.2 percent at the end of 4 days. In central Missouri at the Swan Lake NWR, 62 carcasses were depredated at the following daily cumulative rates: 43.5, 67.7, 79.0, and 82.3 percent. In Texas coastal marshes, 47 carcasses disappeared at the following cumulative rates: 32 percent in less than 1 day, 47 in 2 days, 62 in 3 days, and 89 by the eighth day (Stutzenbaker et al. 1983).

Zwank et al. (1985) removed 1,072 sick, dead, and dying lead-poisoned waterfowl from Catahoula Lake, Louisiana, from 13 October 1980 through 31 January 1981. No reports of waterfowl die-offs were received during the time these collections were made, but removal of the waterfowl might have made reports of die-offs less probable. No die-off, however, had been reported during the 1979-1980 season on this same area when Smith (1980) found levels of ingested lead similar to those reported by Zwank and his colleagues. They concluded, "This magnitude of mortality without a corresponding reported die-off supports Bellrose's (1976) contention that the most important aspect of lead toxicosis mortality may not be the recorded massive die-offs, but the day-to-day losses" (p. 23).

As long as the numbers of ill and dead ducks do not exceed the ability of predators and scavengers to consume them, few carcasses are left as evidence. When mortality reaches greater proportions, however, carcasses become evident. Because lead poisoning generally results in the wasting away of flight muscles, victims are often unable to fly and are sometimes immobilized for a week or two before death, circumstances that make them easy prey. In diseases such as botulism and duck virus enteritis, birds succumb more quickly than they do to lead poisoning. Diseased carcasses therefore are more likely to be seen than lead-poisoned ones because predators and scavengers fail to keep pace with the death rate.

Unless waterfowl losses are so extensive that they focus attention on a particular area, they are often overlooked, especially by the public. Humburg et al. (1983:254) found that one-fourth of the waterfowl carcasses "planted" in quadrats on Swan Lake Refuge, Missouri, were not located when the areas were searched. Texas biologists (Stutzenbaker et al. 1983) experienced an even greater surprise. They planted 100 waterfowl carcasses on a 40.5-ha (100-acre) area, 10 carcasses in each of 5 cover types and 50 carcasses randomly tossed atop vegetation. Within 30 minutes after placement, 8 searchers were able to locate only 6 birds, all of which had been placed on top of vegetation. None of the carcasses placed in cover, where a sick or crippled duck might be expected to hide, was found. Because scavengers and predators probably did not remove all of the 94 unfound ducks within 30 minutes after they had been planted, these results also demonstrate the difficulty of finding waterfowl carcasses hidden in vegetation.

Bellrose's (1959: Table 31) data suggest that 58.5 percent of the male mallards in the Mississippi Flyway with one or more lead pellets in their gizzards at a given time will die of lead poisoning during that year. Of individual male mallards with one or more lead pellets in their gizzards at a given time, only 7.8 percent (58.5 percent ÷ 7.5, the number of 20-day intervals in the 150 days ducks typically spend in migration and on wintering grounds) will die of lead poisoning at that time. The remaining 50.7 percent of the deaths from lead poisoning will come from ducks that have not yet ingested shot or from the 92.2 percent that had shot in their gizzards but will not die until they ingest shot a second, third, fourth, fifth, sixth, or seventh time during the year.

As an approximation, however, and until better data are available, we can calculate the estimated mortality from lead poisoning of an assumed population of 50,000 male mallards on a given wintering or migration area in the following manner:

50,000 male mallards × 6.80 percent that have one or more lead pellets in their gizzards (Bellrose 1959) = 3,400 × 58.5 percent = 1,989 male mallards that will die of lead poisoning on the area.

If the area were searched each day for 150 days (the calculated time ducks spend in migration and on wintering grounds) and if 10 percent of the male mallards that died of lead poisoning were found - in a Texas study (Stutzenbaker et al. 1983), 8 searchers found only 6 percent of 100 dead ducks planted in a 40.5-ha area - searchers should find 1.3 male mallards per day dead of lead poisoning (1,989 male mallards dead of lead poisoning × 10.0 percent found = 199÷150 days = 1.3). Thus, we should not be surprised that "routine" losses from lead poisoning go largely unnoticed.

In contrast to lead poisoning, the self-perpetuating nature of duck virus enteritis, other contagious diseases, and botulism causes victims to spread the disease or toxin among healthy waterfowl. Local epizootics are then prone to develop. Even so, outbreaks are often far advanced before they are noticed. During the past several years, the highest known losses from diseases in the United States other than lead poisoning are botulism, 150,000 (1970); fowl cholera, 70,000 (1956-66) up to 100,000 (1974-75); and duck virus enteritis, 40,000 (1973) (U.S. Fish and Wildlife Service 1976). More recently, 122,555 waterfowl died from disease in California and Nevada. Of these, 78 percent died of botulism, 21 percent of avian cholera, and 1 percent from other diseases (memo from C.T. Osuzi, disease biologist, U.S. Fish and Wildlife Service, Klamath Basin Refuges, Tule Lake, California, to refuge biologists, U.S. Fish and Wildlife Service, 31 May 1984). Such epizootics among waterfowl have been sporadic both in time and place. Obviously, known losses from disease are but a tiny fraction of the nonhunting mortality of ducks.

Lead poisoning, however, is not as confined to particular times or places as are bacterial and viral diseases. Because of the widespread distribution of spent lead shot on the bottoms of lakes and marshes and on upland feeding grounds, the potential for lead poisoning is everywhere that ducks and geese are hunted.

Because lead-poisoned ducks are easier for hunters to bag than healthy ducks, man is frequently the agent that removes lead-poisoned ducks from the population. In a sense, the hunter fills the same role as other predators. The role of wild predators increases in importance after the hunting season when man is no longer removing a portion of the afflicted ducks.

In experiments in central Illinois, 1949-1951, trapped wild mallards were banded and released either as controls (not dosed with lead) or as experimental birds dosed with one, two, or four No. 6 lead shot (Bellrose 1959). During the following 25 days, hunters returned bands in the following proportions: 1.5 bands from birds dosed with one pellet, 1.9 bands from birds dosed with two pellets, and 2.1 bands from birds dosed with four pellets were returned for every band returned from non-dosed control birds. This evidence indicates that mallards ill from lead poisoning were more readily taken by hunters than were healthy ducks. The temporal effect of increasing levels of shot on the availability of dosed ducks to hunters is shown in Figure 4. Many of the ducks shot were undoubtedly suitable for human consumption, but others in the late stages of lead poisoning were emaciated and probably were discarded. Many thin ducks that we have examined in hunters' bags showed the effects of lead poisoning. Relationships between number of lead pellets in the gizzard and body weight in male and female mallards found dead or moribund in Minnesota, IIlinois, Arkansas, and Louisiana are shown in Table 6.

Table 6 - Average weights (g) in relationship to number of ingested lead shot of male and female mallards found dead or moribund during massive die-offs in Minnesota, Illinois, Arkansas, and Louisiana, 1938-1955 (unpublished data from Illinois Natural History Survey files.)

Number of Shot
Males
Females
No.
Wt.
Sx
No.
Wt.
Sx
0
48
900
154.0
43
767
83.7
1
106
838
60.1
87
748
69.1
2
71
860
51.1
48
767
101.3
3
60
841
117.0
41
785
87.0
4
43
868
108.2
21
767
45.1
5
30
889
92.7
14
842
82.9
6
28
852
37.8
13
757
46.8
7
9
847
87.8
6
771
70.2
8
11
866
82.3
3
771
90.7
9
10
816
122.8
4
794
45.3
10
9
892
130.2
1
726
-
>10
30
885
121.3
12
771
146.0
Total and Average
455
860a
89.3
293
768b
79.8
aAverage normal weight = 1,247g = 31.0 percent loss.
bAverage normal weight = 1,106g = 30.6 percent loss.

These differing rates may be accounted for by differing food habits. Mallard diets in Arkansas-Louisiana were more beneficial than those in the upper Midwest, and a higher level of shot ingestion was consequently required to incur mortality in mallards in Arkansas-Louisiana than in the upper Midwest. Stated another way, lead-poisoned birds dropped out of the population more quickly in the upper Midwest than they did in Arkansas-Louisiana.

Weight loss as a result of lead poisoning has been demonstrated in other studies of experimental dosing. The daily percentage weight loss in captive game-farm mallards dosed with from zero to six No. 6 lead pellets is shown in Figure 5, based on unpublished data in the Illinois Natural History Survey (INHS) files. The average daily percentage loss of body weight among mallards dosed with one No. 4 or one or two No. 6 lead pellets on various diets is shown in Figure 3 (from unpublished data in the INHS files). All were adults, but differences between males and females were shown in some cases.

GIF- Net percentage of daily body weight loss(pic)
Figure 5-Net percentage of daily body weight loss in game-farm mallards with increasing doses of No. 6 lead shot. Data from Illinois Natural History survey files.

Losses attributed to crippling often conceal losses due to lead poisoning. We found that two-thirds of the ducks we initially believed crippled proved instead to be dead or incapacitated from lead poisoning (Bellrose 1953:348). We also determined that lead-poisoned mallards weighed from 0.2 to 0.7 lb less than mallards that succumbed to shot wounds (Bellrose 1953). Most moribund ducks of light weight are lead-poisoned rather than victims of debilitating wounds. Similar findings were made in Missouri by Humburg et al. (1983:252), who reported that 83.5 percent of necropsied mallards died from lead poisoning; only 15.6 percent had died from gunshot wounds. In Louisiana, Zwank et al. (1985:25) calculated that ">7 times as many northern pintails, >5 times as many mallards, >7 times as many snow geese, and 3 times as many greater white-fronted geese died, or would have died, from lead toxicosis than died, or would have died, from wounding by hunters." Results were different in California, however, where 4,991 waterfowl carcasses were necropsied from 1977 through 1980 (Moore and King 1980). Of the 4,092 specimens for which cause of death was determined, 55 percent had died from avian cholera, 24 percent from crippling wounds, 13 percent from botulism, 6 percent from lead poisoning, and 2 percent from miscellaneous diseases.

Data from Duck Die-offs

Large-scale losses of waterfowl from lead poisoning are most evident after the hunting season, a time when few hunters or other people are in marshes and swamps. Lists of die-offs known to have been caused by lead poisoning were compiled by Bellrose (1959:240-241) and the Mississippi Flyway Council Planning Committee (Hawkins 1965). Most of these were noted in winter and early spring; only two were observed during the hunting season, both late in the season. As determined from carcasses collected periodically, Humburg et al. (1983:255) reported that mallard losses from lead poisoning in Missouri also appeared most frequently late in the hunting season and after the season had closed.

Losses from lead poisoning occur most frequently during the winter and spring for several reasons. Perhaps the most important is that hunting deters waterfowl from feeding in many areas until the close of the season. Hunters place their blinds on or near the best waterfowl feeding grounds, and spent lead, of course, is deposited most densely in the vicinity of these blinds. After the close of the hunting season and if freeze-up has not occurred, waterfowl are attracted to the abundant food still available near the blinds and consequently ingest spent lead shot while feeding, apparently mistaking shot for seeds, tubers, small mollusks, and other food items. (Waterfowl may ingest lead during the hunting season by feeding at night near blinds, but this ingestion probably results only in chronic losses until the birds are stressed by winter weather late in the season.) When large numbers of ducks gather to feed on grounds that were heavily hunted, the number ingesting shot may increase mortality to a level beyond the appetites of scavengers and predators, and the die-off then becomes noticeable. In addition, these predators had been helped during the hunting season by hunters who killed and removed a sizable proportion of poisoned birds.

Although observed die-offs of waterfowl from lead poisoning represent a tragic loss, they represent only a small proportion of the actual loss. Knowledge of the magnitude of these obscure, usually overlooked, day-to-day losses comes from several sources: (1) the presence of ingested lead in waterfowl gizzards obtained from hunters during the fall and early winter, (2) the occurrence of lead in wing bones, (3) the level of lead in blood, (4) the level of lead in livers, and (5) the numerous experiments with penned or released ducks dosed with lead shot.

The shot levels found among mass die-offs of ducks from lead toxicosis provide ancillary information on the lethality of lead. Most dead or moribund male and female mallards in three regions of the Mississippi Flyway were found with three or fewer pellets (Table 7). From 4.4 to 18.1 percent of these birds contained no ingested pellets. In Illinois and Minnesota-South Dakota, no statistical differences were found between the percentages of males and females ingesting lead or between the number of pellets ingested by them (X2 = 12.5, P = 0.67; X2 = 11.4, P = 0.59). In Arkansas-Louisiana, however, mallard females died with significantly fewer pellets in their gizzards than did males (X2 = 22.0, P = 0.98). A comparison of mallard males in the three flyway regions revealed no statistical difference in rates of ingested shot between birds found in Illinois and birds found in Minnesota-South Dakota (X2 = 15.4, P = 0.83) but a significant difference between rates of ingested shot in birds found in Illinois and those found in Arkansas-Louisiana (X2 = 69.5, P = 0.99). Rates of shot ingestion by male and female mallards in Arkansas-Louisiana were considerably higher than they were for male mallards in Illinois (data from Bellrose 1959: Table 2, Table 3).

Table 7 - Incidence of ingested lead shot pellets in male and female mallards found dead on moribund at die-offs from lead poisoning, in three regions of the Mississippi Flyway, 1938-1955 (from Bellrose 1959: Table 2,3).

Shot Level
Illinois
Minnesota - South Dakota
Arkansas - Louisiana
Males
Females
Males
Females
Males
Females
No.
%
No.
%
No.
%
No.
%
No.
%
No.
%
0
37
11.9
32
18.1
21
13.4
11
11.5
8
4.4
14
8.6
1
90
29.0
57
32.2
53
33.8
42
43.8
9
5.0
18
11.1
2
49
15.8
30
16.9
30
19.1
20
20.8
35
19.3
23
14.2
3
36
11.6
19
10.7
20
12.7
10
10.4
24
13.3
34
21.0
4
18
5.8
7
4.0
10
6.4
6
6.3
25
13.8
18
11.1
5
23
7.4
5
2.8
4
2.5
3
3.1
18
9.9
19
11.7
6
9
2.9
5
2.8
8
5.1
2
2.1
16
8.8
12
7.4
7
5
1.6
1
0.6
2
1.3
0
0.0
8
4.4
10
6.2
8
7
2.3
2
1.1
1
0.6
1
1.0
9
5.0
2
1.2
9
4
1.3
1
0.6
0
0.0
1
1.0
8
4.4
5
3.1
10
6
1.9
1
0.6
3
1.9
0
0.0
6
3.3
2
1.2
>10
26
8.4
17
9.6
5
3.2
0
0.0
15
8.3
5
3.1
Total
310
99.9
177
100.0
157
100.0
96
100.0
181
99.9
162
99.9

The weights of mallards that died in the Mississippi Flyway die-offs discussed above were remarkably similar by sex over the most common levels of pellet ingestion (one to four pellets) (Table 6). Many of these specimens were alive when picked up but incapable of flight. Both males and females were 31 percent below their average weights (Bellrose 1976:229).

These findings suggest that chronic rather than acute conditions usually prevail in lead poisoning die-offs. W.L. Anderson (1975) reported on certain acute effects of lead in lesser scaups. He observed a positive correlation between the number of ingested pellets and body weight, but these birds had ingested large levels of lead shot (44 percent had ingested more than 10 shot each). The lack of relationship between low numbers of ingested shot and body weight implies that once ducks are affected by lead the physiological results are similar among individual ducks no matter how many shot are in their gizzards.

In chronic lead poisoning, gizzard activity is reduced and the afflicted bird literally starves to death, dying at a weight approaching that caused by starvation. Conversely, birds succumbing to acute lead poisoning have massive destruction of blood and other tissues and die at much higher weight levels.

Data from Laboratory and Field Studies

The incidence of lead pellets found in the gizzards of waterfowl and the high levels of lead present in their wing bones, blood, and livers provide ample evidence of the magnitude of exposure to lead. Other data are needed to tell us how many waterfowl die from this exposure. Initially, we thought that experiments with penned waterfowl would provide data on the relationship between amount of lead ingested and mortality; however, varying kinds and amounts of food consumed affected the toxicity of lead so significantly that results were tenuous. Furthermore, a wild duck's diet cannot be precisely duplicated in the laboratory, and ducks in the wild eat larger quantities of food to maintain their weight than do captive birds. In addition, experiments with penned birds cannot reveal at what point a predator might take a duck ill from lead poisoning or under what circumstances stress from lead poisoning might increase a bird's susceptibility to other mortal diseases. Finally, wild mallards in captivity and game-farm mallards in captivity react differently to diet and to dosed lead. Wild mallards, no doubt, are under greater stress in captivity than are game-farm mallards.

Field experiments, on the other hand, provide meaningful data on the relationships between ingested lead and mortality in wild ducks. Such an experiment involves trapping large numbers of wild ducks of one species, fluoroscoping and weighing them to select healthy birds, dosing a portion with lead shot, banding all of them, and returning the birds to the wild - all within a few hours.

Bellrose (1959) conducted such a study. All trap sites were located within 0.8 km (0.5 mi) of the Havana Field Station of the Illinois Natural History Survey on the Chautauqua National Wildlife Refuge so that trapped birds could be speedily handled. During three consecutive autumns, beginning in October 1949, a total of 6,099 mallards were trapped; 4,307 were used in the experiment. These birds were divided into four groups: undosed birds (controls), birds dosed with one No. 6 lead shot, birds dosed with two No. 6 shot, and birds dosed with four No. 6 shot. To increase the number of band recoveries in 1950 and 1951, a $2.00 Reward Band was attached in addition to the standard band. This tactic increased the recovery rate 2.2 times (Bellrose 1955).

In this experiment, band recoveries from hunters provided the data for assessing the mortality of the several groups of mallards. We found a higher rate of band recoveries from dosed wild mallards during the first 10 months after banding than from the undosed controls (Fig.6). These higher recoveries began to appear 8 days after dosing (Fig 4). As previously noted, lead-poisoned birds are more likely to be killed by hunters. Moreover, reduced rates of band recoveries among dosed ducks during the hunting season 1 year after banding suggest that additional nonhunting mortality may have occurred in the interim. Clearly, adult male mallards dosed with one, two, or four No. 6 lead shot suffered greater mortality from lead poisoning than did juvenile males during the year of banding (Bellrose 1959:274). Differences in band recovery rates between dosed and unclosed birds during the second season were highly significant for adult males (X2 = 18.72, P < 0.005) and for juveniles (X2 = 16.02, P < 0.005). As in the experiments with penned mallards, juveniles in the wild were not as susceptible to lead toxicity during the fall as were adults. However, this difference in susceptibility appears to cease late in December; there after juveniles ingesting lead suffered losses similar to those of adults.

GIF- Percentage of total mortality(pic)
Figure 6 - Percentage of the total 4-year mortality after banding that occurred during the first 10 months after wild mallards dosed with zero, one, two, or four No. 6 lead shot were released. (Data from Bellrose 1959, Table 27; data for females adjusted by information in Bellrose 1955).

Band data on female mallards were not of the same quality as data for males. Nevertheless, the data suggest that females suffer greater mortality during the fall and early winter than do males. This pattern agrees with the data from our laboratory experiments. As previously pointed out, however, females become less susceptible than males to lead toxicosis late in the spring.

A similar dosing experiment was conducted in California between 22 January and 23 March 1979 by the U.S. Fish and Wildlife Service and the California Department of Fish and Game (Deuel 1985). Slightly over 12,000 pintails were trapped and banded; 6,109 were dosed with two No. 5 lead shot, and all ducks were released. Bands subsequently recovered revealed no significant differences in survival between dosed and nondosed birds. The presence of lead shot had no apparent effect upon subsequent survival.

To understand the difference in results between experiments with free-flying wild mallards in Illinois and free-flying wild pintails in California is to understand the difference in food habits between the two species in the respective regions. Numerous laboratory experiments (Jordan and Bellrose 1951; Longcore et al.1974; Irwin 1977; Korandaetal.1979) have shown that duck diets high in protein, calcium, and phosphorus help to alleviate lead toxicosis.

Diets of pintails in three regions of California (Pederson and Pederson 1983; Euliss 1984; Deuel 1985) have a high protein level, particularly in late winter and early spring, the period of the experimental dosing study. Pintails were found to consume appreciable quantities of invertebrates; midge (chironomid) larvae were especially important and most prevalent in late winter and early spring diets. The crude protein of the most important invertebrates ranged from 46 to 76 percent (Deuel 1985: Table 7). Swamp timothy (Helechloa schoenoides), an important food for pintails in the Central Valley of California, has a crude protein content of 13.9 percent, high for a plant species (Deuel 1985). We believe that the high protein diet of pintails in late winter and early spring in California reduced mortality from lead at the level tested (two No. 5 pellets). Laboratory studies indicate that as shot levels increase they increasingly overwhelm the beneficial effect of diet. According to Pederson and Pederson (1983), Euliss (1984), and Deuel (1985), pintails feed most extensively on seeds of moist soil plants during the fall, a period when invertebrate consumption is at a seasonal low.

The diet of the mallards in the Illinois experiment was composed of corn, seeds of moist soil plants, and coontail, a minor item high in protein (Anderson 1959). Overall, this diet was high in carbohydrates but low in protein. The difference between the protein intake of mallards in Illinois and pintails in California appears responsible for the difference in mortality rates from lead poisoning.

Had the California experiments been conducted during the fall when the pintails feed more extensively on rice, barley, and weed seeds, results might have been different. Stendell et al. (1979:6) reported that 12.5 percent of the wing bones of adult pintails in California and 8.9 percent of those of juveniles had >20.0 ppm lead. These findings establish that significant amounts of lead entered the bodies of pintails during the hunting season, the period in which wing bones were collected for analysis.

Several reports of waterfowl, including pintails, dying of lead poisoning in California have been made. Moore and King (1980) conducted intensive waterfowl mortality surveys during 1979-1980 on Delevan National Wildlife Refuge and Grizzly Island Wildlife Area, California. They collected 779 waterfowl of which 340 were necropsied; 36 (10.6 percent) were diagnosed as lead poisoning mortalities. Pintails were also involved in 21.4 percent of the lead poisoning episodes in the Pacific Flyway as determined by the National Wildlife Health Laboratory (Table 4). Stendell et al. (1979) reported that 9.2 percent of wing bones from pintails from the Pacific Flyway contained >20.0 ppm lead (Table 5).


Previous Section-- Susceptibility to Lead Poisoning
Return to Contents
Next Section -- Practices to Reduce Lead Toxicosis

Accessibility FOIA Privacy Policies and Notices

Take Pride in America logo USA.gov logo U.S. Department of the Interior | U.S. Geological Survey
URL: http://www.npwrc.usgs.gov/resource/birds/pbpoison/mortalit.htm
Page Contact Information: Webmaster
Page Last Modified: Friday, 01-Feb-2013 19:42:36 EST
Menlo Park, CA [caww55]