Acute toxicity of ingested bismuth alloy shot in game-farm mallards. Toxicity of ingested bismuth alloy shot in game-farm mallards : chronic health effects and effects on reproduction Digitized by the Internet Archive in 2011 with funding from University of Illinois Urbana-Champaign http://www.archive.org/details/acutetoxicityofi35sand I I J 55:3-4 3 £ ILLINOIS ^^^J NATURAL HISTORY SURVEY Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards Glen C. Sanderson, William L. Anderson, George L. Foley, Loretta M. Skowron, Jeffrey D. Brawn, James W. Seets, and Karen L. Duncan Illinois Natural History Survey Bulletin Volume 35, Articles 3 and 4 April 1997 Natural History .Survey . Library Illinois Natural History Survey, Edward J, Arrhbrust, Acting;Chief A Division of the Illinois Department of Natural Resources A catalog of the publications of the Illinois Natural .History Survey is available without charge from the address below. A price list and an order blank are included with the catalog. Illinois Natural History Survey Distribution Center Natural Resources Building 607 East Peabody Drive _..'." Champaign, Illinois 61820 Citations: Sanderson, G.C, W.L. Anderson, GJL. Foley, L.M. Skowron, J.D. Brawn, and J.W. Seets. 1997. Acute toxicity of ingested bismuth alloy shot in game-farm mallards. Illinois Natural History Survey Bulletin 35(3): 185-2 16. Sanderson, G.C, W.L. Anderson, G.L. Foley, K.L.. Duncan, L.M. Skowron, J.D. Brawn, and J.W. Seets. 1997. Toxicity of ingested bismuth alloy shot in game-farm mallards: chronic health effects and effects on reproduction. Illinois Natural History Survey Bulletin 35(4):217-252. Editors: Thomas- E. Rice and Charles Warwick US ISSN 0073-4918 Printed by Authority of the State of Illinois (MJ412759-1M-4-97) Printed with soy ink on recycled and recyclable paper. ILUNOIS LIBRARY Contents Article 3: Acute Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards Acknowledgments ii Abstract 185 Introduction 185 Literature Review 185 Methods 187 Toxicity Study 187 Chemical Analyses 189 Storage of Samples 189 Digestions of Samples 189 Digestions for ICP Analysis 1 89 Digestions for GFAA Analysis 189 Analytical Methods 189 ICP 189 GFAA 190 Quality Control 190 Calculations 190 Statistical Analyses 190 Results 191 Survival 191 Retention and Dissolution of Shot 191 Body Weight 192 Organ Weights 192 Gizzard 192 Liver 192 Kidneys 192 Gonads 192 Hematocrit (Hct) 192 Heavy Metals and Essential Elements in Organs and Blood 192 Kidneys 195 Liver 195 Gonads 202 Plasma and Blood Cells 204 Feces 208 HlSTOPATHOLOGY 208 Gonadal Lesions 208 Female 208 Male 208 Liver 211 Kidneys 211 Gizzard 211 Discussion 212 Copper 212 Phosphorous 212 Iron 212 Calcium 212 Feces 212 Conclusions 213 Literature Cited 214 Continued on next page Article 4: Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards: Chronic Health Effects and Effects on Reproduction Acknowledgments ii Abstract 217 Introduction 217 Methods 217 Toxicity Sudy 217 Chemical Analyses 220 Storage of Samples 220 Digestions of Samples 220 Digestions for ICP Analysis 220 Digestions for GFAA Analysis 221 Analytical Methods 221 ICP 221 GFAA 221 Quality Control 221 Calculations 221 Statistical Analyses 222 Results 222 Chronic Toxicity Test 222 Survival 222 Hematocrit 222 Body Weight 222 Dissolution of Shot 222 Shot Retention 226 Organ Weights 228 Analyses of Tissues and Other Materials 231 Kidneys 231 Liver 231 Gonads 234 Blood 236 Reproduction 236 Eggs 236 Ducklings 240 Egg Weights 244 Egg Shell Thickness 244 Fertility Rates 244 Hatchability Rates 244 Egg Shell Analysis 244 Egg Content Analysis 244 Age of Embryo at Time of Death 247 HlSTOPATHOLOGY 247 Adults 247 Kidneys 247 Liver 247 Gonads 248 Heart 248 Lungs 248 Ducklings 248 Liver 248 Kidneys 248 Heart 248 Discussion 248 Conclusions 250 Literature Cited 251 ILLINOIS NATURAL HISTORY SURVEY Acute Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards fe^j Glen C. Sanderson Illinois Natural History Survey William L. Anderson Illinois Department of Natural Resources and Illinois Natural History Survey George L. Foley University of Illinois and Illinois Natural History Survey Loretta M. Skowron Illinois State Water Survey Jeffrey D. Brawn Illinois Natural History Survey James W. Seets Illinois Natural History Survey Illinois Natural History Survey Bulletin Volume 35, Article 3 April 1997 Illinois Natural History Survey Bulletin Vol. 35 Art. 3 Acknowledgments James W. Sergent and Wendy L. Grethen, Illinois Natural History Survey, fed the ducks, cleaned the pens, examined the feces for voided shot, saved fecal samples for chemical analysis, and assisted with other phases of the study. The following people assisted with weighing and dosing ducks and collecting and processing blood: Stephen P. Havera, Michelle M. Georgi, Aaron P. Yetter, and Christopher S. Hine, all with the Waterfowl Re- search Laboratory, Forbes Biological Station, Illi- nois Natural History Survey, Havana, Illinois; Ed- ward J. Heske, Linda K. Campbell, and Anne E. Zielske, all with the Illinois Natural History Sur- vey; and volunteers Beverley C. Sanderson and J. William Sanderson. Beverley C. Sanderson also assisted with many tasks in the preparation of this manuscript. William R. Manuel, retired, College of Veterinary Medicine, University of Illinois, pro- vided his expertise in the collection of blood. We thank Jerry L. Longcore, Leader, Patuxent Wildlife Research Center, Orono, Maine; Lawrence J. Blus, Wildlife Research Biologist, Biological Resources Division, U.S. Geological Survey; and Louis N. Locke, Wildlife Pathologist, and Milton Smith, Chemist, National Wildlife Health Center, for their reviews of the manuscript. Petersen Publishing Company, Los Angeles, California, provided fi- nancial support for the research and costs of this publication. 5 f 'iiU^a^u^ Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards 185 Abstract In a 30-day study involving penned game-farm mallards (Anas platyrhynchos), no harmful health effects were detected from dosing with either six, No. 4, bismuth/tin (Bi/Sn) alloy shot or six, No. 4, steel (Fe) shot, as compared with sham (0 shot) dosing. Survival, hematocrit (Hct) values, body weight, and mean weights of kidneys, livers, gonads, and gizzards were not affected. Mean concentrations of nutritionally essential elements (calcium [Ca], phosphorous [P], magnesium [Mg], zinc [Zn], copper [Cu], Fe, and Sn) were different among doses and between sexes in kidneys, livers, and gonads. However, concentrations of these elements in these organs and tissues in Bi-dosed ducks were not different from both 0- and Fe-dosed ducks. Bi/Sn alloy shot, as tested in this study, elicited no indications of toxicity in game-farm mallard ducks. Introduction To protect waterfowl from poisoning caused by ingested lead (Pb) shot, nontoxic shot regulations were implemented for waterfowl hunting on areas with severe problems with Pb poisoning ("hotspots") in the United States beginning in the early 1970s (Anderson 1992). Federal regulations became nationwide in 1991. Several European countries have converted or are planning to convert to nontoxic shot for waterfowl hunting (Moser 1992) and Canada will implement a nationwide ban on lead shot for all migratory bird hunting in 1997 (Canadian Wildlife Service [CWS] 1995). From the 1970s to the early 1990s, Fe was the only shot material approved as nontoxic by the U.S. Fish and Wildlife Service (USFWS) (Longcore et al. 1974). Although hunters have generally adapted to using Fe shot, some have urged that a search for alternative shot with greater ballistic capability be continued. Specifically, they wanted a shot that is non- toxic, inexpensive, and ballistically similar to lead shot. There have been numerous evalua- tions of potential substitute shot. Irby et al. (1967) evaluated three types of plastic-coated Pb, two Pb/Mg alloys, Fe, Cu, Zn-coated Fe, and molybdenum-coated Fe. Longcore et al. (1974) evaluated Pb shot with nickel coatings, Pb/phosphor Sn alloy shot, Pb shot with Sn/ nickel alloy coatings, steel shot with Pb coatings, Pb/Sn alloys, two types of disintegrable Pb shot, and Pb shot with biochemical additives. Haseltine and Sileo (1983) evaluated uranium. No satisfactory alternative was found until 1990, when John E. Brown, St. Catherines, Ontario, was awarded a U.S. patent for Bi shot. Our primary objective was to determine if Bi/Sn alloy shot caused toxic effects in captive game-farm mallards. Secondly, if toxic effects were manifest, we wanted to associate toxic ef- fects with amounts of Bi and other elements in the tissues. Our study complied with the "Acute Toxicity Test" guidelines of the USFWS and the CWS. Dr. Simon Nadeau, CWS, and Dr. Keith A. Morehouse, USFWS, reviewed our protocol be- fore we initiated our study. Literature Review The first known report of metallic Bi dosed in birds was by Hanzlik and Presho (1923), who administered metallic Bi, Pb, and other heavy metals to pigeons. The fatal dose of metallic Pb in their studies ranged from 0.6 to 2.28 g/kg. By contrast, none of the four Bi-dosed pigeons died after receiving doses that averaged 1 .39 g/kg, and the researchers concluded that Pb is more toxic — and mortality is higher with smaller doses—than other heavy metals, including Bi. Sanderson et al. (1992) conducted the first comprehensive study to determine the toxicity of ingested Bi shot (100% Bi) in birds (mallards). They followed with tests on ingested Bi/Sn alloy shot (Sanderson et al. 1997b) and the present study. Sanderson et al. (1997b) reported that reproduction of game-farm mallards was not affected after chronic dosing with Bi/Sn alloy shot. Sanderson et al. (1997a) reported no toxic effects of Bi shot embedded in the breast muscles of game-farm mallards. The International Commission on Radiologi- cal Protection (ICRP) (1960) reported that Bi is rapidly excreted by the kidneys except for small amounts retained in these organs. Some Bi is lost in bile. The estimated half-time for elimination in humans was about 5 days. Hamilton et al. (1972/ 1973) reported 0.4 Hg/g Bi (wet weight) in kid- neys and 0.004 |ig/g Bi in livers of autopsy cases (humans) with no known exposure to Bi. Kidneys of 22 individuals who had been given Bi salicylate had 33 Hg/g Bi and livers had 6.8 |ig/g Bi. 186 Illinois Natural History Survey Bulletin Vol. 35 Art. 3 Lee (1981) reported that after treatment was terminated, Bi declined about 2.6% in the urine daily, with half purged from the body in about 20 days. Fowler and Vouk (1979:348) stated, "In- gested bismuth is largely eliminated unabsorbed in feces. Model values for the daily balance of bismuth in reference man are: dietary intake 20 Hg / g, fecal elimination 1 8 |!g / g, urinary excretion 1.6 Ug/g. • • • Absorbed bismuth is mainly ex- creted in the urine." Oehme (1979) reported that soluble and in- soluble Bi salts, suspended in oil to maintain levels in blood, were injected to treat syphilis. Other Bi compounds were used to treat malaria and amebiasis. Medicinal use of Bi decreased with the advent of newer treatments. Most hu- man exposure to Bi is in compounds that are insoluble and not readily absorbed whether in- gested or applied to the skin. Apparently tissue binding is slight, even when Bi is absorbed. An equilibrium is established among tissues, blood, and urine. Kidneys have the highest amounts of Bi, with the liver generally a poor second. With few exceptions, Bi compounds present no prob- lems whether by ingestion, inhalation, or dermal application. Poisoning from industrial exposure is rare (Oehme 1979). There are no federal standards for Bi or its compounds and no evidence linking Bi or Bi compounds with industrial poisoning. Also, all episodes of Bi poisoning were from soluble com- pounds used in medicine, and fatalities and near fatalities were mainly from intravenous or intra- muscular injection of soluble salts (Key et al. 1977). Venugopal and Lukey (1978) reported that low solubility limits the toxicity of Bi compounds, which are highly toxic. Locke et al. (1987) re- ported neurotoxic effects at Bi concentrations of < 0.1 Ug/g in blood. Thomas et al. (1988:124) reported, "Since tox- icity resulting from environmental or industrial exposure to bismuth or any of its compounds is not a problem, levels of tolerance have not been identified . . . ." Bi telluride, which is used as a semiconductor in the electronics industry (Oehme 1979), is an exception. Thomas et al. (1988) re- ported, however, that the French Ministry of Health banned the sale and use of all Bi com- pounds, and that Australia restricted the use of Bi subgallate. These authors suggested that in hu- mans amounts of Bi in blood > 0.48 umol/L (0.1 ug/mL) are potentially dangerous, that amounts > 0.05 and < 0.1 ug/mL call for careful monitoring of patients, and that amounts < 0.05 ug/mL are considered safe. Krigman et al. (1985) reported that levels of Bi in blood of humans differ between individuals who show side effects from chronic use of Bi and those who do not. Those who show no effects usually have < 0.05 ug/g Bi in their blood whereas those who show symptoms have > 0.05 ug/g Bi in their blood. These authors conclude that levels of Bi in blood > 0.05 Ug/g indicate a high risk and amounts < 0.05 Ug/g indicate a low risk. Other investigators (Hillemond et al. 1977; Serfontein and Mekel 1979) concluded that 0.05 ug/g Bi in blood is potentially neurotoxic. Dipalma (1988) stated that levels of Bi in blood should not be > 0.02 ug/g. Dipalma (1988) reported that exposure of humans to Bi is not considered a serious indus- trial hazard. According to him, there are few data on Bi concentrations in blood from either oral or topical applications because of the assumption that absorption of Bi is low. Dipalma (1988) indicated that bacteria in the intestine might me- thylate Bi to form a soluble compound. He re- ported (p. 244), "In animals, trimethyl bismuth is highly toxic and causes an encephalopathic syn- drome similar to that seen in man. Blood levels of bismuth should not exceed 20 ug per L (20 ppb)." In their review, Slikkerveer and de Wolff (1989) summarized the effects of Bi in mammals and reported a peak of Bi in blood 45 minutes after oral dosing with colloidal Bi in humans. Others had reported peaks between 4.7-21 Ug/g 15-60 minutes after dosing. With continued dosing, 3- 4 weeks were necessary to reach a steady-state of Bi in plasma. Persons who had not received Bi therapy had between 1 and 15 ug/L of Bi in their blood. Although the site of Bi absorption in the gastrointestinal tract is unknown, Slikkerveerand de Wolff believed that absorption after oral dos- ing is dependent on solubility and that cysteine, sorbitol, and lactic acid may promote absorption of Bi. They suggested that colloidal Bi is absorbed in the small bowel and stomach. Meaningful reference values for Bi levels in tissues are not available because of large varia- tions in experimental and analytical techniques, and the chemical form of Bi in blood is unknown. The highest concentrations of Bi were always in the kidney. After 14 months of dosing with colloi- dal Bi subcitrate in rats, Bi concentrations ranked from high to low in kidney, lung, spleen, liver, brain, and muscle tissues. When bone concentra- tions were measured, they were usually 10-20 times lower than in the kidney. Slikkerveerand de Wolff (1989) reported that Bi is found in both urine and feces. The Bi in feces April 1997 Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards 187 comes from Bi excreted in bile, which concen- trates plasma Bi by a factor of 10, and from intes- tinal secretion. In humans showing symptoms of Bi toxicity after exposure, concentrations in bone were 1.5-6.7 Ug/g wet weight compared with < 1 Ug/g wet weight in nonexposed individuals. Bi encephalopathy is mainly supported by elevated blood Bi. A steady-state Bi concentration of > 100 Ug/L (ppb) of blood in humans was arbitrarily suggested as an "alarm" level and 50 Ug/L was considered a "safety" level, but no proof supports these choices. Concentrations of blood Bi from 10 to 4,600 ug/L were found in 618 Bi encephalopa- thy patients. Abbracchio et al. (1985) administered tri-po- tassium-dicitrato bismuthate intraperitoneally and by gavage in laboratory rats. After intraperi- toneal injection, Bi reached peak concentrations in blood within 30 minutes and declined rapidly. When a dose 10 times higher was given by gastric intubation,much lowerblood concentrations were detected. They found no Bi in the brain after oral administration and concluded that there was ap- parently little risk of neurotoxicity after dosing with this derivative. They stated (p. 143), "This could also be the reason why no appreciable side effects have ever been described after use of this drug in humans." Gregus and Klaassen (1986) reported that feces and urine were equally important for the excretion of injected Bi compounds in rats. Biliary excretion apparently determined the fecal excre- tion of Bi, but the percentage of dosed Bi excreted in the bile was independent of the amount dosed. Woods and Fowler (1987) reported that little information was available on the effects of Bi in mammals in general, but noted that toxic effects in the liver, kidneys, and blood have been found in humans and laboratory animals after exposure to Bi compounds. In their studies with rats (P. 276), they found that "...bismuth significantly im- pairs the activities ofboth hepaticALA synthetase and heme synthetase at all dose levels." Ross et al. (1988) injected 2,500 ug/g of Bi subnitrate intraperitoneally in laboratory mice. Although Bi concentrations in blood and brain tissues of mice that showed signs of neurotoxicity were significantly higher than in dosed mice that showed no signs, they concluded that the concen- tration of Bi in blood did not predict neurologic signs. They suggested that 6 Ug/g of Bi in the brain show neurologic symptoms and that a con- centration of > 0.5-2.0 Ug/g of Bi in blood had to be maintained for several weeks to accumulate enough Bi in the brain to cause neurotoxicity. They also concluded that 5-1 ug / g Bi in the brain was associated with motor dysfunction in hu- mans and mice and that concentrations above 50 Ug/L (ppb) are necessary to produce frank en- cephalopathy in humans. Slikkerveer and de Wolff (1989) reported, however, that following oral dosing of trimethyl Bi to dogs, the level of Bi was higher in the liver than in the kidney, probably because of the or- ganic character of the molecule. They reported that early toxic effects of Bi may be related to effects on enzymes of the haem synthesis but that anemia has never been associated with ingestion of Bi. Methods The Bi/Sn shot used in this study contained 0.0040% to 0.0186% Pb fc= 0.0094%,SD = 0.0054%). Because Pb made up < 0.1% of the test shot, Environment Canada (1992) guidelines did not require that tissues be analyzed for Pb. We recog- nized, however, that researchers are interested in Pb, so we included this metal, albeit at a some- what high detection limit, in the analyses for residues. Seventy-five female and 75 male wild-type game-farm mallards 6 to 8 months of age were purchased from Whistling Wings, Hanover, Illi- nois. The ducks, reared on a 60-acre lake, were transported from Hanover to Champaign, Illi- nois, by truck in crates on 22 March 1994. Toxicity Study The ducks were weighed and one duck was ran- domly assigned to each pen. Forty ducks (20 females and 20 males) were randomly assigned to one of the three treatments—dosed with Bi shot, dosed with Fe shot, or sham dosed with shot (controls). Five male and five female ducks were randomly selected from each dosing group for collection of feces to be analyzed for excreted Bi, Fe, and Sn. Sanderson et al. (1997b) describe the methods we used to randomize the doses, ducks, and pen assignments for the present study. The pens were consecutively numbered, el- evated, outdoor, 1-m2 structures. They were cov- ered with vinyl-coated, 25.4-mm mesh, 14-gauge wire. A 9.1-m x 36.6-m pavilion (roofbut no sides) covered the pens (see Sanderson et al. 1992 for more details). Facilities forholding the ducks were inspected and approved by several members of the Labora- tory Animal Care Committee, University of Illi- nois, after the ducks were placed in the pens. The Illinois Natural History Survey Bulletin Vol. 35 Art. 3 committee also inspected the facilities once dur- ing the study. Commercial duck pellets (Heinhold 17% Duck Finisher Pellet™, Heinhold Feeds, Inc., Kouts, Indiana) were provided ad libitum during the 3-week acclimatization period. The duck pellets contained a minimum of 17.0% protein. On the date of dosing, the pellets were replaced with whole shelled corn ad libitum for the duration of the study. Protocols of theCWS and theUSFWS specified these diets (Environment Canada 1992). The three groups of ducks were each dosed as follows: sham dosed (controls); six, No. 4 (3.30 mm diameter), Fe shot; or six, No. 4, Bi shot. Ducks in each group are hereafter referred to as 0- dosed (controls), Fe-dosed, and Bi-dosed. We began the study on 12 April 1994 (Day 0) when we weighed, collected blood samples, and dosed the ducks. A small plastic funnel fitted with a plastic tube (9.5mm outside diameter, 22.9 cm long) was inserted through the pharynx into the proventriculus. To reduce friction, the tube was kept in a pail of water between dosings. Each dose of shot was poured into the funnel and flushed into the proventriculus with approxi- mately 5 mL of water. Controls were treated the same except that no shot were included. Before dosing, the shot were counted, weighed, and placed in individual vials in the laboratory. The type, number, and weight of each dose of shot were recorded on the top of each vial and on a computer printout for each duck. At dosing, the shot dose was matched with the corresponding duck. Blood was collected from the wing vein in heparinized microhematocrit capillary tubes for hematocrit determination and in 2.5-mL syringes for separation into cells and plasma. The plasma samples were analyzed for major elements (> 1% by wt in shot) and for major nutritionally essential elements (Ca, P, Mg, Zn, and Cu). Twenty-gauge, 25.4-mm needles were used (Baxter Healthcare Corporation, Scientific Products Division,McGaw Park, Illinois). The whole blood was injected into 10-mL lithium heparinized Vacutainer™ tubes and centrifuged to separate cells and plasma. Body weights were recorded and blood samples collected on Days 0, 15, and 30. As each group of 24 hematocrit samples was collected, the samples were centrifuged at the site in a mobile laboratory (house trailer). The tubes were spun for 5 minutes at 11,500 PvPM at 13,000- g force, after which the values were read and recorded. The whole blood samples also were centri- fuged at the site, in groups of 12 samples (capacity of the centrifuge) at a time. The tubes were spun for 5 minutes at 3,000 RPM. The plasma was removed with micropipettes and placed in 5-mL nonheparinized Vacutainer tubes. The cells were retained in the 10-mL lithium heparinized tubes. As the plasma and cells were separated, the tubes were placed in racks and put on ice in a styrofoam cooler. All samples were stored in a freezer (-10°C) until analyzed. The Bi shot were provided by William S. Montgomery, Jr., Bismuth Cartridge Co., Dallas, Texas. Seven shot were chemically analyzed in the laboratory of the Illinois State Water Survey, Champaign, Illinois, before the ducks were dosed. Mean (+ SD) percentages of elements in these shot were as follows: Bi = 98.35%, ± 0.86%; and Sn = 1.90%, ± 0.10%. Other elements averaged < 0.1% each; Pb ranged from 0.0040% to 0.0186% (x=0.0094%, + 0.0054%). Fe shot were removed from commercial 12-gauge shotgun shells and were not analyzed. The 120 ducks were weighed and blood was collected from the wing veins, as scheduled, on Day 30 (12 May 1994). Following these proce- dures, the ducks were killed by decapitation and necropsied on the same day (with the exception noted below). The gizzard, liver, kidneys, and gonads were excised from each duck. Two changes were made in the methods as originally approved by the CWS. First, because voided shot were not found in the feces, 20 dosed ducks were radiographed to obtain a positive record of shot retention in the gizzards. The 20 ducks for which daily fecal samples were being collected were chosen so that fecal material could be re-examined if the radiographs indicated dosed shot were missing from the gizzards. A dorsal- ventral and a right or left lateral view radiograph were made for each duck on Day 23 (5 May 1994) by the College of Veterinary Medicine, University of Illinois. The other change in the methods involved killing the ducks and performing the necropsies over 2 days (instead of 1) to ensure that tissue samples were obtained from freshly killed birds. The pathologist necropsied 30 ducks on 12 May and the other 30 ducks on 13 May 1994. After the pathologist had examined, weighed, and fixed representative samples of kidneys, liver, and gonads in 10% formalin for histopathology, the remaining residual tissue from these organs was placed in separate, numbered, plastic bags and stored in a freezer as backup samples. Organs from the remaining 60 ducks, which were not necropsied, were removed and weighed, placed April 1997 Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards 189 in individual, numbered, plastic bags, and stored in the freezer as additional backup samples. We took representative samples, after fixing in 10% formalin, from each of the 60 necropsied ducks and examined the samples histopathologi- cally. Sections of gonad (testis or ovary), liver, kidney, and gizzard were embedded, trimmed, and sectioned at 4 microns. Tissues on glass slides were stained with hematoxylin and eosin by stan- dard methods. All ducks were examined by a veterinary pathologist, who did not know the dose history of the ducks. Later, we associated group assignment and weight data with histo- logic findings to aid in interpretation. Chemical Analyses Storage of Samples Samples were inventoried when received, stored at -10°C, and monitored daily. Samples were allowed to thaw to room temperature, then pre- pared for metal analysis by labelling by tissue type and a number for identification. The sex of the duck and the shot dose it received were not disclosed to the individuals who analyzed the samples. Digestions of Samples Blood cells, blood plasma, livers, kidneys, go- nads, and feces were acid digested before analysis for metals with inductively coupled, argon-plasma emission spectroscopy (ICP) and graphite-fur- nace, atomic-absorption spectroscopy (GFAA). Because wet weight concentrations of the blood and organs were desired, these samples were not dried before digestion. Feces were dried at 104°C to determine percent moisture. The concentra- tions of metals measured in fecal samples are on a dry-weight basis. We analyzed for Bi, Sn, Fe, Pb, Ca, Mg, P, Zn, and Cu. ICP was used to measure these metals; beryllium (Be) was used as an inter- nal standard. GFAA was used to measure Pb and Bi when they were at low concentrations. Digestions for ICP Analysis We used samples of 0.5 to 1.0 g. A mixed portion of the sample was weighed to 1.0 mg with an electronic, top-loading balance and placed into a tared 50-mL conically tipped polypropylene, cen- trifuge tube. The tubes were precleaned for 24-hr with a 10% nitric acid (HN03 ) soak then rinsed in deionized water. Samples of feces were weighed to 0.1 mg. Approximately 30 to 50 mL of an acid and internal standard solution were added to the sample after taring. The final acid concentrations were 2% HNQ3 and 10% hydrochloric acid (HC1). The Be concentration was targeted at 2.00 mg/L. The samples were then homogenized into a slurry using a saw-toothed generator made of titanium and TFE-fluorocarbon (Pro Scientific, Monroe, Connecticut). The internal standard solution was used to rinse excess materials from the generator and the amount was accounted for in the total weight. Samples were prepared with the SpectrPrep System™, an automated microwave-digestion- system(CEM Corporation, Matthews, NorthCaro- lina). A 15-mL sample loop was used. After heating, cooling, and filtering, about 12.5 mL of the sample were collected and deposited by autosampler into 1 5 mL polypropylene test tubes. This digestate without further treatment was then used for ICP analysis. The automated microwave digestion system was a relatively new technique to prepare samples. A few problems arose in adjusting to the system; most were associated with clogging of the small-diameter tubing. A thorough homogenation followed by a few hours in a warm, ultrasonic bath usually improved the operation. Digestions for GFAA Analysis We used samples of 0.5 to 1.0 g. A mixed portion of the sample was weighed to 1.0 mg with an electronic, top-loading balance and placed into a tared TFE-fluorocarbon beaker. Approximately 20 mL of deionized water (DI H20), 0.250 mL concentrated HNO3, and 1 mL of hydrogen per- oxide (H2 2) were added. The mixturewas heated at approximately 95°C until the solution started to clear (about 0.5 hr). Approximately 20 mL of DI H2 and 2 mL H2 2 were added. Upon further heating the mixture cleared and "foamed up." DI H2 was used to rinse contents from the sides of the beaker. The beakers were then covered with TFE-fluorocarbon watch glasses and allowed to reflux for approximately 1 hr. The resulting solu- tions were usually clear to yellow. The samples were increased to 50 mL in a volumetric flask, filtered through 0.45-^m nitrocellulose filters, and stored in acid-washed, linear, polyethylene bottles. The ultimate acid concentration was 0.5% HNO3. High-purity acids and hydrogen peroxide (Baker Ultrex™ and Fisher Optima™) were used for all digestions. Analytical Methods ICP We used a Thermo Jarrell Ash (TJA) AtomComp™, Model 61, vacuum spectrometer. The instrument has a polychromator configured with 44 fixed 190 Illinois Natural History Survey Bulletin Vol. 35 Art. 3 channels, including analytical lines for high and low concentrations of Ca and Mg. Although we reported results for only a few elements, we mea- sured 30 analytes to monitor for spectral interfer- ences, which we did not detect, with blank sub- traction and background correction. We used USEPA Method 200.7, Revision 4.4, Determination of Metals and Trace Elements in Waterand Wastesby InductivelyCoupled Plasma- Atomic Emission Spectroscopy for our work. We modified the method and used a different diges- tion process, and we measured Bi, not a listed analyte. We chose Be as an internal standard because it was not present in the samples, it does not cause spectral or background interferences, and it is precisely detectable. GFAA We used a Thermo Jarrell Ash, Model 957, Atomic Absorption Spectrophotometer coupled with a Model 188, Furnace Atomizer and FASTAC™ autosampler. Samples were introduced as a spray and deposited directly into a carbon cuvette at 100°C so that samples dried on contact. We used method 3113 of Greenberg et al. (1992). We ana- lyzed samples in triplicate and reported the mean. Quality Control We calibrated instruments daily cind we verified the standard curve using National Institute of Standards and Technology (NIST) traceable, qual- ity control samples (QCS). Samples (usually 10) were bracketed by calibration blanks, laboratory fortified blanks, and instrument-performance, check solutions during analysis as well as peri- odic checks on the internal-standard solution. The ICP instrument was programmed to compen- sate for drift by recalculating the slopes of the calibration curves if any analyte was more than +5% of the true value while measuring the ICP check standard. If an analyte measured greater than +10% of the true value for this sample, the instrument was recalibrated and the affected samples reanalyzed. The ICP check standard was formulated for a concentration at the midpoint of the calibration curve. It was traceable to NIST Standard Reference Materials (SRMs). TheGFAA QCS initially were required to be within 10% of the true value. Subsequent measurement of the bracketed internals was required to be +15%. If these limits were exceeded, the instrument was recalibrated and the affected samples reanalyzed. Ten percent of the samples were digested and analyzed in duplicate, half of them spiked. Additional liver samples were treated as dupli- cates as part of the process of evaluating the automated, microwave-digestion equipment. Digestion blanks and spiked digestion blanks were prepared at a frequency of 10%. They were processed through the complete digestion and analytical system in the same manner as the samples. Calculations We saved the ICP data during analysis in data- base files with ThermoSpec (TJA) software utiliz- ing Enable OA. Data were then imported into Enable spreadsheets for tabulations and calcula- tions. The Enable spreadsheets were saved in a Lotus 1-2-3 format on diskette. TheGFAA results were recorded on an instrument printer as con- centrations in ug/L based upon peak-area mea- surements. These data were manually entered into spreadsheets for tabulations and calculations. The Method Detection Limit (MDL) (Glaser et al. 1981) was used to establish the detection limits for concentrations of elements in tissues and other materials. Glaser et al. (1981:1426) describe the MDL as "a new performance crite- rion for chemical analysis . . . defined as that concentration of the analyte that can be detected at a specific confidence level." Also, "The detec- tion limit should be related to the standard devia- tion of the measured value at or near zero concen- tration of the analyte . . . ." They further report (1427), "MDL is considered operationally mean- ingful only when the method is truly in the detec- tion mode, i.e., analyte must be present. The method detection limit is defined as the minimum concentration ofa substance that can be identified. " To be considered a meaningful difference, the MDL procedure is required to provide a value that averages > two times theMDL (Glaser et al. 1981). For statistical analysis, values < MDL were en- tered as one-half the MDL value. Most values for elements in the tissues were determined by ICP. Results of ICP analyses for Bi, Pb, and Sn were usually lower than the MDLs. Thus, selected samples of kidneys, livers, and gonads were analyzed for Bi and Pb by GFAA. The remaining amounts of plasma and blood cells after analysis by ICP were inadequate for further analysis by GFAA. Graphite-furnace atomic ab- sorption is not a satisfactory method to analyze for Sn. Statistical Analyses In this report, when two values are reported as "different" or that they "differ," it means that they differ in a statistical sense at an alpha of (P < 0.05). April 1997 Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards 191 Differences in concentrations of various ele- ments in livers, kidneys, and gonads; weights of organs (post-mortem); numbers of shot recov- ered; and dissolution rates of shot were tested by one- or two-way ANOVA using sex and dose (shot type) as grouping factors. Homogeneity of variances among groups was assessed with Levene's test. Brown-Forsythe or Welch statistics were used in instances where variances could not be assumed equal. In instances where the overall test of differences among groups was significant, pairwise comparisons were performed and sig- nificance evaluated based on the Bonferroni cor- rection. In instances where comparisons were made with controls, Dunnett's procedure was used. Variation inbody weights, hematocrit counts, and concentrations of elements in plasma and red blood cells (all measured at Days 0, 15, and 30) were evaluated using repeated-measures ANOVA. As above, sex or dose or both, were used as between-subject factors. Within-subject tests for variation over time were also performed as were tests for interactions between dose and time. When assumptions of compound symme- try were violated, Huynh-Feldt-adjusted signifi- cance probabilities were used. Results Survival All 120 ducks (controls, Bi-dosed, and Fe-dosed) survived to the end of the 30-day test period. Retention and Dissolution of Shot No voided shot were found in the feces from the 20 dosed ducks (5 female and 5 male Bi-dosed and 5 female and 5 male Fe-dosed ducks) for which feces were saved for chemical analysis. Radio- graphs on Day 23 readily identified all six shot in the gizzard of each of these ducks. Six pellets, which were sometimes dissolved to small disks, were recovered from 38 of the 40 Bi- dosed ducks. One male contained only five Bi disks in his gizzard. Because most of the shot were highly dissolved, it is probable that the sixth pellet had dissolved. A second male contained four tiny Bi particles in his gizzard. The combined particles of Bi weighed only 42.1 mg, and the fifth and sixth pellets probably had dissolved. Six pellets were recovered from 35 of the 40 Fe-dosed ducks. One female had five tiny pellets in her gizzard that weighed 77.7 mg. This duck had the second highest (2,339 Hg/g) concentra- tion of Fe in the liver. The mean concentration of Fe in livers of Fe-dosed ducks was 1086 |ig/g. Thus, the sixth pellet undoubtedly had dissolved. One female contained no shot in her gizzard, but she had 1,782 |ig/g Fe in the liver. This duck probably also had dissolved the shot. The re- maining three ducks, one male and two females, each contained five pellets in their gizzards. All of these pellets were small, collectively weighing from 1 47.4 to 282.9 mg for each duck. One of these females had the highest concentration (2,412 jig/ g) of Fe in her liver of any duck. The other two ducks contained 645 ug/g and 1,043 |ig/g Fe respectively, in their livers. The sixth pellet in each of these three ducks may have been voided, but they probably were dissolved. None of the dosed ducks with missing shot in their gizzards was among the ducks that were radiographed and for which feces were saved for analysis. The retained Bi and Fe shot differed in ap- pearance. The Fe shot were usually round, al- though many were pitted or had empty spaces on their surfaces, whereas the Bi shot were generally disk-shaped or flattened. In several instances, five Bi disks plus two, three, or four tiny pieces (not flakes) of Bi were recovered from the gizzard. Obviously, when a Bi disk became thin enough, it disintegrated into several pieces. A small number of flakes of Bi were found in a few gizzards. This finding for Bi/Sn alloy shot is in contrast to the abundance of tiny flakes of Bi found in the dosing study that used 100% Bi shot (Sanderson et al. 1992). The dissolution rates were variable in both Fe-dosed and Bi-dosed ducks. Based on the shot recovered from the gizzards on Day 30, females dosed with Bi shot dissolved a mean of 69.5% and males 72.5% (Table 1 ) of the metal's original weight in 30 days (dissolution in individual ducks ranged from 38.2% to 96.4%). No difference between the sexes was detected for Bi-dosed ducks. Fe-dosed females dissolved an average of 69.2% and Fe- dosed males 55.6% of the metal's original weight in 30 days (range for individual doses was from 38.0% to 89.6%). The different dissolution rates between sexes for Fe-dosed ducks was expected (Table 1). Females approaching the breeding season in spring eat more food than males and thus produce more acid in their gizzards. As a result, Fe shot, which dissolve readily in the acid (HC1) environment of the gizzard, dissolve more rapidly in females than in males during this sea- son. Males dissolved more of the weight (72.5%) of the Bi shot than of the Fe shot (55.6%) in 30 days. 192 Illinois Natural History Survey Bulletin Vol. 35 Art. 3 Females dissolved no more (69.5%) of the weight of the Bi shot than of the Fe shot (69.2% ) in 30 days. An interaction, which was caused by the lower rate of dissolution of Fe shot by males as com- pared with the dissolution rate of Fe shot by females and no difference in the dissolution rates of Fe shot and Bi shot by females, existed between sex and dose (Table 1). Body Weight All groups ofdosed ducks, except Bi-dosed males, lost from 1.8 to 5.0% of their body weight during the 30-day study. All groups lost from 4.5 to 9.6% of their body weight from Day to Day 15, prob- ably because of the switch from duck pellets to a whole corn diet. By Day 30, most of the birds had regained weight lost after the change in diet. Bi- dosed males gained only 1.6% in body weight from Day to Day 30 (Table 2). Males weighed more than females, and an interaction in weight between sex and time ex- isted, with females losing a larger percentage of their weight from Day to Day 30 than males. Although ducks lost weight over time, the aver- age weight losses for females from Day to Day 30 were only -3.8% for 0-dosed, -3.8% for Fe-dosed, and -5.0% for Bi-dosed females. The average weight changes for males from Day to Day 30 were -1.8% for 0-dosed, -3.2% for Fe-dosed, and +1.6% for Bi-dosed ducks. No difference existed in body weights among doses (Table 2). Organ Weights Gizzard Mean gizzard weights ranged from 29.3 g for Fe- dosed females to 32.2 g for Bi-dosed males (Table 3). No difference was detected in the weight of gizzards between sexes or among doses. As a percentage of total body weight, mean gizzard weights ranged from 2.5% for each of Fe- dosed and Bi-dosed males to 3.0% for 0-dosed females (Table 3). Gizzards of females contrib- uted a higher percentage of the total body weight than males. No difference was recorded among doses in the percentage that gizzards contributed to total body weight. Liver Mean weights of livers ranged from 19.3 g for Bi- dosed females to 21 .7 g for Fe-dosed females. No differences existed between sexes or among doses (Table 4). When considered as a percentage of total body weight, mean values for livers ranged from 1.6% for Bi-dosed and 0-dosed males to 2.0% for Fe-dosed and 0-dosed females. Livers of females comprised a higher percentage of the total body weight than the livers of males. No difference was detected among doses in the mean percentage that livers contributed to the total body weight. Kidneys Weights of kidneys, the organ most involved in excretion of Bi, differed least between sexes and varied least among doses of the organs weighed. Mean weights of kidneys ranged from 6.4 g for Bi- dosed females, Bi-dosed males, and Fe-dosed fe- males to 6.6 g for 0-dosed males (Table 5). No differences were found between sexes or among doses in the weights of kidneys. As with weights of livers, when kidney weights were expressed as a percentage of total body weight, sex differences were detected. Mean percentages ranged from 0.5% for each group of males to 0.6% for each group of females. Kidneys of females comprised a larger percentage of the total body weight than males, but no differences existed among doses. Gonads No differences among doses in the mean weights of gonads were found (Table 6). As was expected, mean weights of gonads differed between the sexes: 6.4 g for 0-dosed females, versus 26.4 g for 0-dosed males; 10.1 g for Fe-dosed females, ver- sus 28.0 g for Fe-dosed males; and 4.3 g for Bi- dosed females versus 22.5 g for Bi-dosed males. These sex differences also were evident in gonad weights when expressed as a percentage of total body weight; the means ranged from 0.4% for Bi- dosed females to 2.4% for Fe-dosed males (Table 6). No differences appeared among doses, but male gonads contributed a larger percentage of the total body weight than did the female gonads. Hematocrit (Hct) Mean hematocrits were not different among doses for the three sample times: Days 0, 15, and 30 (Table 7). The mean percentage changes in Hct values from Day to Day 30 did not differ be- tween the sexes. With sexes combined the mean percentage change in Hct from Day to Day 30 increased (P< 0.00001 ) by 6.6% for controls, 1 1 .8% for Bi-dosed ducks, and 1 2.8% for Fe-dosed ducks. Heavy Metals and Essential Elements in Organs and Blood For consistency in presentation of the data, usu- ally the mean concentrations of elements in each organ or tissue for each sex and for sexes com- Continued on page 195 April 1997 Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards 193 Table 1. Percent of the dosed shot accounted for and mean percent of weight of dosed shot dissolved in 30 days in the gizzard six, No. 4, Fe shot or six, No. 4, Bi shot in female and male game-farm mallards (n = 20 females and 20 males in each dosed group). % of Dosed Shot Mean % Wt of Sex Dose Accounted for' Shot Dissolved 11 F Fe 933" 692 5.04h 2.84 M Fe 99.2 55.6 0.84 2.22 F Bi 100.0 69.5 0.00 4.01 M Bi 97.5 72.5 1.82 3.59 " Based on the shot recovered from the gizzards on Day 30, when the ducks were killed. h SE. Interaction between sex and dose: F = 4.53; P = 0.0374. Difference between sexes for percent of Fe shot dissolved in 30 days: F = 14.89; P = 0.0014. 1,37 Difference between doses for males: F = 17.42; P = 0.0002. Table 2. Mean body weight on Days a , 15, and 30b of female and male game-farm mallards each dosed with shot; six, No. 4, Fe shot; or six, No. 4, Bi shot, and mean percentage change in body weight from Day to Day 30 (n = 20 females and 20 males in each group). Mean Body Weight (Kg) Sex Dose Mean % change in body DayO Day 15 Day 30 wt-Day to Day 30c 1.11 1.06 1.06 -3.8 0.024d 0.027 0.030 1.12 2.24 1.19 1.22 -1.8 0.020 0.025 0.022 1.38 1.10 1.04 1.05 -3.8 0.020 0.018 0.021 0.97 1.24 1.18 1.20 -3.2 0.023 0.027 0.025 1.12 1.08 1.02 1.03 -5.0 0.026 0.025 0.026 1.11 1.23 1.21 1.24 + 1.6 0.019 0.025 0.039 2.31 F M F Fe M Fe F Bi M Bi a Ducks were dosed on Day 0. h Ducks were killed on Day 30. e Because of rounding error, these means are sometimes slightly different than if calculated by differences in the Day and Day 30 columns. d SE. Mean body weight: Difference between sexes: F = 66.42; P < 0.00001. 1,114 Interaction between sex and time: F =3.21; P= 0.0471. 2,228 Difference among doses: F =0.11; P= 0.8995. 2,114 Change over time: F =33.17; P< 0.00001. 2,228 194 Illinois Natural History Survey Bulletin Vol. 35 Art. 3 Table 3. Mean weight of gizzard and mean percentage it contributed to total body weight in game-farm mallards 30 days after dosing with shot; six, No. 4, Fe shot; or six, No. 4, Bi shot (n = 20 females and 20 males in each group). Sex Dose Mean Weight(g) Mean % of body wt F M F M F M Fe Fe Bi Bi 31.8 1.04a 31.6 0.75 29.3 1.07 30.7 0.98 30.2 1.00 32.2 1.08 3.0 0.07 2.6 0.08 2.8 0.08 2.5 0.08 2.8 0.13 2.5 0.13 SE. Difference among doses in weight of gizzard: F = 1.50; P = 0.2265. 2,14 Difference between sexes in percentage gizzard contributed to total body weight: F = 34.43; P< 0.00001. 1,114 Difference among doses in percentage gizzard contributed to total body weight: F =1.99; P = 0.1422. Table 4. Mean weight of liver and mean percentage it contributed to total body weight in game-farm mallards 30 days after dosing with shot; six, No. 4, Fe shot; or six, No. 4, Bi shot (n = 20 females and 20 males in each group). Sex Dose Mean Weight(g) Mean % of body wt F M F M F M Fe Fe Bi Bi 21.1 1.49a 20.0 0.94 21.7 1.54 20.5 0.94 19.3 0.92 19.5 0.97 2.0 0.10 1.6 0.06 2.0 0.12 1.7 0.08 1.9 0.08 1.6 0.04 SE. Difference among doses in weight of liver: F = 1.10;?= 0.3370. 2,14 Difference between sexes in percentage liver contributed to total body weight: F = 22.21; P< 0.00001. 1,84 Difference among doses in percentage liver contributed to total body weight: F = 1.80; P= 0.1706. April 1997 Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards 195 continued from page 192 bined are listed in the tables. When no statisti- cally significant differences existed between sexes, usually only P values for the combined sexes are provided. Kidneys The MDL for analysis by ICP for Bi in kidneys was 17.8 Ug/g (wet wt). The mean concentra- tions were < MDL in kidneys of all but 2 of 120 ducks. Because the MDL for Bi by ICP in the kid- neys was unacceptably high, kidneys of 10 0- dosed and 1 1 Bi-dosed ducks were selected for analysis by GFAA. No sex differences were detected in the mean concentration of Bi in the kidneys of Bi-dosed ducks (Table 8). The mean concentration (6.86 ug/g) of Bi in the kidneys of Bi-dosed ducks, with sexes combined, was higher than the mean concentration of Bi in 0-dosed ducks (0.334 Ug/g). The mean concentration of Bi in the kidneys of Bi-dosed ducks was much higher than the mean concentration (2.23 Ug/g) of Bi in the livers of Bi-dosed ducks (Table 10). TheMDL for Pb in the kidneys was 6.54 ug/ g (wet wt) by ICP. All mean values for Pb in the kidneys were < MDL by this method. By GFAA, no sex differences existed in the concentration of Pb in the kidneys of 0-dosed or Bi-dosed ducks. No difference was found in the concentration of Pb in the kidneys of 10 (sexes combined) 0-dosed ducks (0.440 ug/g) compared with 11 Bi-dosed ducks (0.313 ug/g) (Table 8). TheMDL for Sn in the kidneys was 9.47 ug/ g (wet wt). Only six ducks had concentrations of Sn >MDL. The mean concentration of Sn in the kidneys of these six ducks was 14.3 Ug/g and ranged from 10.5 to 19.7 Ug/g. Three of these ducks were Bi-dosed (x = 6.7 ug/g), two were 0- dosed (x = 1 .0 Ug/g), and one was Fe-dosed (13.7 Ug/g)- No difference existed between sexes in the mean concentrations of Cu in the kidneys (Table 9). With sexes combined, no differences were detected among doses in the mean concentration of Cu in the kidneys: 6.31 Ug/g in 0-dosed, 7.31 Ug/g in Fe-dosed, and 6.14 Ug/g in Bi-dosed ducks. No sex differences existed in the mean con- centrations of P in the kidneys (Table 9) and no differences existed among doses in the mean concentrations of P in the kidneys. However, an interaction was found between sex and dose. No difference existed between sexes in the mean concentration of Fe in the kidneys (Table 9), but mean concentrations of Fe in the kidneys differed among doses. Fe-dosed ducks, sexes combined, had higher mean concentrations (145 Ug/g) of Fe in their kidneys than 0-dosed ducks (123 ug/g) or Bi-dosed ducks (123 Ug/g). No difference was detected in the mean concentra- tions of Fe in the kidneys of 0-dosed and Bi-dosed females (Table 9). Females, with doses combined, had higher mean concentrations of Ca in the kidneys than males, but no difference was found in the mean concentrations of Ca in the kidneys among doses. Mean concentrations ofMg in the kidneys did not differ between sexes, and with sexes combined, no difference existed among doses (Table 9). Mean concentrations of Zn in the kidneys of 0-dosed, Fe-dosed, and Bi-dosed ducks did not differ between sexes within each dose, but with doses combined, males had higher mean concen- trations of Zn in the kidneys than females (Table 9). With sexes combined, mean concentrations of Zn in the kidneys varied among doses. Bi-dosed ducks had higher mean concentrations of Zn in their kidneys (28.2 Ug/g) than Fe-dosed ducks (25.2 Ug/g), but not higher mean concentrations than 0-dosed ducks (26.6 ug/g) (Table 9). No difference was detected in mean concentrations of Zn in the kidneys of 0-dosed and Fe-dosed ducks. Liver The MDL (by ICP) for Bi in livers was 18.45 ug/g (wet wt). No concentration of Bi exceeded the MDL in the liver of any duck. Analysis by GFAA produced values of Bi in the livers of 1 1 Bi-dosed ducks that averaged 2.23 Ug/g (0.63 to 5.63 Ug/g). Mean amounts of Bi in the liver (Table 10) were not different between sexes. With sexes com- bined, Bi-dosed ducks contained a higher (2.23 ug/g) mean amount of Bi in the liver than did 0- dosed ducks (0.193 Ug/g ). TheMDL (by ICP) for Pb in the liver was 7.51 Ug/g. The concentrations were all below the MDL. The concentration of Pb in the livers, as determined by GFAA, were from \ 15, and 30b of game-farm mallards each dosed with shot; six, No. 4, Fe shot; or six, No. 4, Bi shot and percentage change in Hct from Day to Day 30 (n = 20 females and 20 males in each group). Sex Dose DayO Mean Hct Day 15 Day 30 Mean % change in Hct—Day to Day 30 F+M 46.7 0.491 F+M Fe 45.8 0.90 F+M Bi 45.7 0.98 48.2 0.41 49.7 0.58 48.7 0.47 49.6 + 6.6 0.48 1.25 50.8 + 12.8 0.55 2.49 49.7 + 11.8 0.52 3.92 ' Ducks were dosed on Day 0. b Ducks were killed on Day 30. ( SE. Difference among doses: Change over time: F = 0.71; P = 0.4961. 2, 117 F = 50.10; P< 0.00001. Table 8. Mean concentrations (|ig/g wet wt) of Bi and Pb in kidneys of game-farm mallards 30 days after dosing with shot (controls) compared with ducks dosed with six, No. 4, Bi shot, as measured by GFAA. Element Sex Dose Bib Bi Pb F 0.140 0.000' M 0.528 0.058 F&M 0.334 0.070 F 0.138 0.070 M 0.742 0.640 F&M 0.440 0.320 8.05 1.14 4.77 1.45 6.86 0.99 0.427 0.179 0.112 0.033 0.313 0.121 MDL = Method Detection Limit (ug/g wet wt) by GFAA for Bi in kidneys = 0.10 ug/j for 10 ducks and 0.27 ug/g for 11 ducks and Pb = 0.27 ug/g for 10 ducks and 0.15 ug/g for 11 ducks. N = 10. N = ll. SE. Difference between sexes in Bi-dosed ducks: F = 3.09; P = 0.1127. 1,9 Difference between doses: F = 43.32; P = 0.0001. 1,10 Pb Difference between sexes in 0-dosed ducks: F = 0.88; P = 0.4011. Difference between sexes in Bi-dosed ducks: F = 1.68; P = 0.2274. 1.9 Difference between doses: F = 0.15; P = 0.7035. 1,19 198 Illinois Natural History Survey Bulletin Vol. 35 Art. 3 Table 9. Mean concentrations (|ig/g wet wt) of Cu, P, Fe, Ca, Mg, and Zn (by ICP) in kidneys of game farm- mallards 30 days after dosing with shot (controls) compa-red with ducks dosed with six, No. 4, Fe shot or six, No. 4, Bi shot (n = 20 females and 20 males in each group). Dose Element Sex Fe Bi Cu F 5.50 0.38'1 M 7.13 0.45 F&M 6.31 0.34 P F 2758 53 M 3167 203 F&M 2962 112 Fe F 110 6.0 M 136 7.5 F&M 123 5.6 Ca F 84.0 7.1 M 79.9 7.3 F&M 81.9 5.0 Mg F 196 3.5 M 216 14.8 F&M 206 7.7 Zn F 25.3 0.8 M 27.9 1.5 F&M 26.6 0.9 1 SE. Cu Difference among doses: F 2,57 = 0.32; P = 0.7280. P Difference among doses: F 2,55 = 0.63; P = 0.5384. Interaction between sex iand dose: F = 3.48; P = 0.0540. Fe Difference among doses: F = 4.98; P = 0.0103. Interaction between sex ijnd dose: F = 3.01; 2,5! P = 0.0574. Difference between 0-dosed and Fe-dosed ducks: P < 0.05. Difference between 0-dosed and Bi-dosed ducks: P>0.10. Difference between Fe-dosed and Bi-dosed ducks: P < 0.05. 5.34 0.26 9.08 3.37 7.31 1.79 2937 80 2903 30 2919 40 152 8.5 139 8.2 145 6.0 84.4 5.3 70.2 1.8 76.9 3.1 199 4.2 197 2.4 198 2.3 24.7 0.6 25.6 0.8 25.2 0.5 5.74 0.49 6.60 0.65 6.14 0.40 3050 67 3006 48 3030 42 118 8.4 129 8.1 123 5.8 87.2 4.5 72.2 1.6 80.4 3.0 206 3.8 204 3.7 205 2.6 27.4 1.2 29.3 1.1 28.2 0.8 Ca Difference between sexes:F = 6.99; P = 0.0107. 1.55 Difference among doses: F = 0.41; P = 0.6680. 2,33 Mg Difference among doses: F = 0.80; P = 0.4536 Zn Difference between sexes:F = 4.45; P = 0.0395. 1,55 Difference among doses: F = 4.56; P = 0.0147. Difference between Fe-dosed and Bi-dosed ducks: P < 0.05. April 1997 Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards 199 Table 10. Mean concentrations (ug/g wet wt) of Bi and Pb in livers of game-farm mallards 30 days after dosing with shot (controls) compared with ducks dosed with six, No. 4 Bi, shot (analyses by GFAA). Dose Element Bi Pb Sex F 0.140- o.ooo 1 M 0.246' 0.033 F&M 0.193 0.024 F 0.068 0.018 M 0.552 0.376 F&M 0.310 0.195 Bi 2.79c 0.675 1.25* 0.347 2.23 0.492 0.184 0.053 0.110 0.030 0.157 0.036 MDL = Method Detection Limit (ug/g wet wt) by GFAA for Bi = 0.27 for 10 ducks and 0.10 for 11 ducks and for Pb = 0.10 for 10 ducks and 0.15 for 11 ducks. - N = 5. b SE. c -N = 7. d N = 5. e N = 4. Bi Difference between doses: F = 17.14; P = 0.0020. Pb Difference between doses: F = 0.65; P = 0.4294. continued from page 195 The mean concentration of Cu in the livers of 0-dosed females was 85.5 |ig/g versus 191 ug/g in males, 56.3 ug/g in the livers of Fe-dosed females versus 172 ug/g in males, and 78.3 ug/g in the livers of Bi-dosed females versus 149 ug/g in males (Table 11, Figure 1). Males had higher mean concentrations of Cu in the liver than fe- males, but we found no differences among doses in the mean concentration of Cu in the livers. Females consistently had more P in their livers than males (Table 11): 0-dosed, 3,164 Ug/g in females versus 2,998 P ug/g in males; Fe-dosed, 3,258 ug/g in females versus 2,958 ug/g in males; and Bi-dosed, 3,154 ug/g in females versus 2,897 ug/g in males. We found no differences among doses in the mean concentrations of P in the livers (Table 11). We detected no difference between sexes in the mean concentrations of Fe in the livers, but the mean concentrations of Fe differed among doses: 411 ug/g in 0-dosed ducks versus 1086 ug/g in Fe-dosed ducks versus 399 ug/ g in Bi-dosed ducks, sexes combined. Differences were detected in the mean concentrations of Fe in 0-dosed versus Fe- dosed ducks and in Fe-dosed versus Bi-dosed, but not in 0-dosed versus Bi-dosed ducks (Table 11). The mean concentrations of Ca in livers did not differ between sexes, but with sexes com- bined, the mean concentrations of Ca in the liver were different among doses. The mean concen- tration of Ca in the livers was higher in 0-dosed ducks (62.8 ug/g) than in Fe-dosed ducks (50.4 ug/g), but was not higher in 0-dosed than in Bi- dosed ducks (51.4 ug/g) (Table 11). We detected no difference in the mean concentrations of Ca in the livers of Fe-dosed and Bi-dosed ducks. The mean concentration of Mg in the livers ranged from 21 1 ug/g in 0-dosed males to 224 ug/ g in Fe-dosed females, but no difference existed between sexes. With sexes combined, no differ- ences were detected among doses in the mean concentrations ofMg in the livers: 0-dosed ducks, 213 ug/g; Fe-dosed ducks, 219 mg/g; and Bi- dosed ducks, 214 ug/g (Table 11). Continued on page 202 200 Illinois Natural History Survey Bulletin Vol. 35 Art. 3 Table 11. Mean concentrations (ug/g wet wt) of Cu, P, Fe, Ca, Mg, and Zn (by ICP) in livers of game-farm mallards 30 days after dosing with shot (controls) compared with ducks dosed with six, No. 4, Fe shot or six, No. 4, Bi shot (n = 20 for each sex). Dose Element Sex Fe Bi Cu Fe Ca Me Zn F 85.5 16.4 1 M 191 37.6 F&M 138 21.9 F 3164 126 M 2998 71 F&M 3081 72 F 416 37.8 M 406 58.8 F&M 411 34.1 F 66.4 7.0 M 59.2 7.9 F&M 62.8 5.3 F 215 8.4 M 211 4.9 F&M 213 4.8 F 53.3 3.5 M 48.9 2.8 F&M 51.1 2.2 56.3 12.1 172 31.8 114 19.2 3258 77 2958 88 3108 62 1158 91 1015 111 1086 72 54.0 2.5 46.8 3.8 50.4 2.3 224 4.6 214 6.5 219 4.0 48.4 2.6 48.1 2.8 48.2 1.9 78.3 15.6 149 32.6 114 18.7 3154 93 2897 63 3026 59 435 43.6 362 24.9 399 161.0 52.9 3.6 49.8 4.6 51.4 2.9 216 4.5 212 5.0 214 3.2 50.8 3.2 45.4 1.9 48.1 1.9 a SE. )ifference between sexes: )ifference among doses: Ca F =20.64; P< 0.00001. 1,76 F = 0.56; P = 0.5721. Difference among doses: = 3.43; P = 0.035b. Hfference between Fe-dosed females and Fe-dosed males: P < 0.05. Mfference between 0-dosed females and 0-dosed males: P < 0.05. )ifference between Bi-dosed females and 0-dosed males: P < 0.05. )ifference between sexes: )ifference among doses: Difference among doses: F =11.07; P = 0.0012. 1. 114 F = 0.45; P = 0.6380. 2.114 F =67.53; P< 0.00001. )ifference between 0-dosed and Fe-dosed; P < 0.01. Difference between 0-dosed and Bi-dosed; P > 0.10. Difference between Fe-dosed and Bi-dosed;P < 0.01. Difference between 0-dosed and Fe-dosed; P < 0.05. Difference between 0-dosed and Bi dosed; P > 0.05. Difference between Fe-dosed and Bi dosed; P > 0.10. Mg Difference among doses: F = 0.66; P = 0.5182. Zn Difference among doses: Sn The mean level was < MDL: = 12.8 ppm F = 0.70; P = 0.5005. April 1997 Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards 201 200 M I C R G R A M S P E R G R A M 100 Females Males Figure 1 . Mean concentrations ofcopper (ng/g wet weight) in liver of game-farm mallards 30 days after dosing with 0; 6, No. 4, steel (iron); or 6, No. 4, bismuth shot. 202 Illinois Natural History Survey Bulletin Vol. 35 Art. 3 Table 12. Mean concentrations (u/g wet wt) of Bi and Pb in gonads of game-farm mallards 30 days after dosing with shot (controls) compared with ducks dosed with six, No. 4, Bi shot (by GFAA Furnace). Dose Element Sex a Br' Bi Pb F 0.050 0.000' M 0.050 0.000 F&M 0.050 0.000 F 0.080 0.000 M 0.080 0.000 F&M 0.080 0.000 0.677 0.455 0.155 0.048 0.468 0.277 0.093 0.013 0.100 0.020 0.096 0.011 MDL = Method Detection Limit for gonads (ug/g wet wt) by GFAA = 0.15 for Pb and 0.10 for Bi. a N - 20. SE. Bi Difference between 0-dosed males and Bi-dosed females: F = 6.2821; P = 0.0406. continued from page 199 The mean concentration of Zn in the livers ranged from 45.4 ug/g for Bi-dosed males to 53.3 (ig/g for 0-dosed females, and values were not different between sexes. With sexes combined, no differences were found among the mean concen- trations of Zn in livers of 0-dosed ducks, 51 .1 ug/ g; Fe-dosed ducks, 48.2 Ug/g; and Bi-dosed ducks, 48.1 ug/g (Table 11). Gonads The MDL for Bi in gonads by ICP was 12.0 Ug/g for 33 ducks and 13.2 ug/g for 28 ducks. All but five values for Bi in gonads were MDL ranged from 15.2 to 27.8 ug/g and averaged 19.9 Ug/g. As determined by GFAA Furnace, the con- centrations of Bi in gonads of 0-dosed ducks were all MDL. The mean concentrations of Pb, as determined by GFAA furnace, in gonads of 0-dosed and Bi-dosed ducks were all MDL ( 1 7.8 ug / g) ofSn in the liver. Mean concentrations of Cu in the gonads differed by sex in 0-dosed ducks, but not in Fe-dosed and Bi-dosed ducks. With doses combined, mean concentrations ofCu differed by sex with males having lower concen- trations than females (Table 13). No differences were found for the mean concentrations of Cu in the gonads of mallards among doses. Mean concentrations of P in the gonads were not different by sex within doses. With sexes combined, no difference was detected in the con- centrations of P in the gonads among doses (Table 13). Differences in the mean concentrations of Fe in the gonads of males and females were substan- tial (Figure 2) in all dosed groups: 0-dosed, 56.4 Ug/g in females versus 12.5 Ug/g in males; Fe- dosed, 53.5 ug/g in females versus 10.7 Ug/g in males; and Bi-dosed, 40.3 Ug/g in females versus 16.8 Ug/g in males. Mean concentrations of Fe in the gonads of ducks within doses were not differ- ent (Table 13). Females contained up to 17 times more Ca in their gonads than males (Figure 3): 0-dosed Continued on page 204 April 1997 Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards 203 Table 13. Mean concentrations (|ig/g wet wt) of Cu, P, Fe, Ca, Mg, and Zn in gonads of game-farm mallards 30 days after dosing with shot (controls) compared with ducks dosed with six, No. 4, Fe shot or six, No. 4, Bi shot (N = 10, 11, or 12 for each sex). Element Sex Dose Fe Bi Cu Fe Ca Mg Zn F 1.76 0.215' M 0.985 0.149 F&M 1.37 0.155 F 3132 388 M 2662 41 F&M 2897 197 F 56.4 7.5 M 12.5 2.8 F&M 34.4 6.4 F 540.0 165.4 M 34.5 1.7 F&M 287.2 99.2 F 113 15.9 M 203 2.5 F&M 158 13.0 F 23.7 4.0 M 13.9 0.4 F&M 18.8 2.2 1.70 0.152 1.12 0.113 1.41 0.114 3102 395 2717 91 2910 202 53.5 8.0 10.7 1.2 32.1 6.3 590.1 179.2 34.4 1.6 312.2 108.0 127 14.6 206 6.3 166 11.9 24.6 3.6 14.3 0.6 19.4 2.1 1.48 0.183 1.34 0.174 1.41 0.125 2566 342 2917 128 2726 195 40.3 5.4 16.8 5.4 29.1 4.6 334.4 102.5 139.6 104.5 241.7 74.6 126 16.2 201 11.8 162 13.0 19.3 2.9 16.4 2.0 18.0 1.8 SE. Cu Difference between sexes: F =13.20; 135 Difference between sexes in 0-dosed ducks: Difference among doses: Difference among doses: F = 0.03; 235 P = 0.0006. P < 0.05. P = 0.9664. 0.23; P= 0.7965. Fe Difference between sexes: F =64.51; P = 0.00001. U5 Difference between sexes in 0-dosed ducks: P < 0.01. Difference between sexes in Fe-dosed ducks: P < 0.01. Difference between sexes in Bi-dosed ducks: P < 0.05. Difference among doses: F = 0.57; P = 0.5704. Ca Difference between sexes: Difference among doses: Mg Difference between sexes: Difference among doses: Zn Difference between sexes: Difference among doses: F =19.50; P< 0.00001. 1,29 F = 0.22; P = 0.8021. 235 F =65.70; P< 0.00001. 1.39 F = 0.22; P = 0.8051. F =12.66; P = 0.0012. ui F = 0.17; P = 0.8402. 236 204 Illinois Natural History Survey Bulletin Vol. 35 Art. 3 continued from page 202 females (540.0 mg/g) versus 0-dosed males (34.5 mg/g); Fe-dosed females (590.1 mg/g) versus Fe- dosed males (34.4 mg/g); and Bi-dosed females (334.4 mg/g) versus Bi-dosed males (139.6 mg/ g). With doses combined, females also contained higher mean concentrations of Ca in their gonads than males. No differences existed, however, in the mean concentrations of Ca in the gonads among doses within each sex (Table 13). Males had higher mean concentrations ofMg in their gonads than females (Table 13), but no differences were detected among doses within each sex. With doses combined, females had higher mean concentrations of Zn in their gonads than males (Table 13). With sexes combined, Zn values in gonads varied little among doses. Plasma and Blood Cells TheMDLs (by ICP) for Bi in plasma were 7.38 ug/ g(wetwt) forDay 0,21.8 Ug/gforDay 15,and 11.8 Ug/g for Day 30. The MDLs for Bi in blood cells were 8.72 Ug/g for Day 0, 9.35 Ug/g for Day 15, and 16.3 Ug/g for Day 30. All mean levels were MDL only for Day 30, the only data included in this report. Females had higher mean concentra- tions of Zn in their plasma than males (Table 14). Continued on page 208 April 1997 Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards 205 60 50 M I C R O 40 G R A M S 30 P E R G R A M 20 10 Controls Fe Females Males Figure 2. Mean concentrations of iron (ug/g wet weight) in gonads of game-farm mallards 30 days after dosing with 0; 6, No. 4, steel (iron); or 6, No. 4, bismuth shot. 700 M I C R G R A M S P E R G R A M 600 500 400 300 200 100 Controls Fe Females Males Figure 3. Mean concentrations of calcium (ug/g wet weight) in gonads of game-farm mallards 30 days after Hosing wifh 0- 6 Nn 4. stppl (iron): or £>. Nn 4 hismnth ^hot- 206 Illinois Natural History Survey Bulletin Vol. 35 Art. 3 Table 14. Mean concentrations (|ag/g wet wt) of Cu, P, Fe, Ca, Mg, and Zn in plasma of game-farm mallards dosed with shot (controls) compared with ducks dosed with six, No. 4, Fe shot or six, No. 4, Bi shot (N = 18, 19, or 20 for each sex). Sex Day Elements Detected Dose Fe Bi F&M Cu F 15 30 P M 15 30 F&M 15 30 F 15 30 Fe M 15 30 F&M 15 30 F 15 30 Ca M 15 30 F&M 15 30 F 15 30 Mj 0.334 0.036" 179 18.0 245 14.1 291 20.4 220 19.4 257 15.2 259 9.6 199 13.5 251 10.2 275 11.5 8.40 1.2 6.31 0.8 7.47 0.9 15.4 4.5 5.94 0.8 7.71 1.1 11.8 2.3 6.14 0.5 7.48 0.7 88.3 6.1 144 12.2 176 17.3 83.0 6.3 110 5.4 107 1.7 85.7 3.9 127 7.2 141 10.2 17.4 1.2 23.3 1.0 26.6 0.304 0.031 204 20.9 268 12.5 303 14.3 202 11.8 262 6.3 252 7.3 203 11.8 265 6.9 277 9.0 14.5 3.5 7.55 0.6 9.07 1.0 7.62 1.0 5.67 0.5 6.08 0.6 11.1 1.9 6.61 0.4 7.58 0.7 87.9 5.7 140 7.8 169 12.6 80.1 5.8 117 1.7 111 1.3 84.0 4.0 129 4.3 140 7.8 19.2 1.4 24.5 0.8 27.0 0.370 0.048 196 13.9 225 8.2 270 13.3 215 17.0 282 22.3 251 14.0 206 11.0 252 12.4 260 9.8 10.7 1.7 5.08 0.8 6.85 0.7 13.8 3.4 8.19 1.2 5.17 0.5 12.2 1.9 6.59 0.7 6.03 0.4 87.0 5.8 120 3.8 168 12.4 80.4 6.2 115 1.5 109 0.5 84.1 4.2 118 2.1 139 8.2 17.8 1.2 21.7 0.5 26.2 April 1997 Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards 207 Table 14 continued Elements Detected Dose Sex Day Fe Bi M 15 30 F&M 15 30 F 30 M 30 F&M 30 Zn 18.0 1.5 22.5 1.2 24.4 0.4 17.7 0.8 22.8 0.8 25.5 0.6 4.56 0.39 2.83 0.13 3.69 0.25 16.3 1.1 22.7 0.5 24.3 0.4 17.8 0.9 23.6 0.5 25.7 0.6 4.27 0.34 2.77 0.08 3.52 0.21 17.1 1.3 23.8 1.0 24.4 0.6 17.5 0.9 22.6 0.6 25.3 0.5 4.63 0.30 2.72 0.17 3.64 0.23 SE. Cu Difference among doses; Day 0: 0.74: P = 0.4789. F = 0.39: P = 0.6747. 2,111 42.24: P < 0.00001 8.07; P-.= 0.0005. 0.03; P-.= 0.9747. Mean concentrations for Days 15 and 30 were SE. s N = 9. MDL = Method Detection Limit by ICP: MDL for Bi = 103 ug/g (dry wt) for Days 0, 1, 2, and 1-10 and 58.9 ug/g for Days 11-30. MDL for Sn = 19.5 ug/g (dry wt) for Days 0, 1, and 2 and 14.9 ug/g for Days 11-30. MDL for Fe = 31.8 ug/g (dry wt) for Days 0, 1, and 2 and 19.2 ug/g for Days 11-30. continued from page 208 and one Bi-dosed duck. In the Fe-dosed duck, a small-sperm granuloma was found, but the re- sidual parenchyma was normal. The Fe-dosed duck had evidence of a locally extensive tubular atrophy consisting of a decreased height in the seminiferous epithelium in one zone of the exam- ined testis. The adjacent tubules were within normal limits with no evidence of inflammation. The testis of the Bi-dosed duck contained scattered aggregates of lymphocytes and plasma cells, and multinucleated giant cells (presumably sloughed spermatocytes) were observed within scattered tubules. Evidence of normal produc- tion of spermatozoa was present on the slide. A small percentage ofducks in all three groups had mild vacuolar changes in the seminiferous epithelium. These ducks had spermatozoa within the genital ducts and in seminiferous epithelium. Liver Nearly all ducks had a variable number of lym- phocytes and plasma cells within the liver. The most common pattern was around the portal tri- ads. Occasionally, the inflammatory cells formed small nodules scattered in the parenchyma. One 0-dosed male had abscesses within the liver, which probably represented an acute bacterial infection. The hepatic lipidosis seen histologically seemed to correlate with livers that were heavier. Kidneys Nearly all ducks had a variable number of lym- phocytes and plasma cells in the wall of the ureter. Gizzard On Day 30, the contents of the gizzards from all 120 ducks were removed and saved, and those from Fe-dosed and Bi-dosed ducks were exam- 212 Illinois Natural History Survey Bulletin Vol. 35 Art. 3 ined for retained shot. The linings of all gizzards appeared to be unaffected, and there was no pattern of variation among doses. No lesions were detected in any of the gizzard sections exam- ined. Discussion Copper Approximately twice the concentration ofCu was detected in the livers of male ducks as in the livers of female ducks in all dosed groups (Table 11). Hanson and Jones (1974) found significantly higher concentrations of Cu in the feathers of female Ross' geese (Anser rossii) than in males. They presumed estrogen was responsible for the difference. Underwood (1971:61, 63), discussing Cu in the liver, stated, "There is no effect of sex, except in the Australian salmon (Arripis trutta) in which the female carried higher concentrations than the male." Van Campen (1971:214) reported that "Ad- ministration of estrogens induces large increases in serum copper in humans, rats, and swine." He also reported that androgens increased serum Cu concentrations in humans. Hill and Matrone (1961) found that when both Cu and Fe were low in the diet, an increase in one partly compensated for the deficiency of the other. Matrone (1960) concluded that Cu absorption is not directly af- fected by Fe. Thus, the Fe:Cu interaction is af- fected by something other than absorption. In the present study, the diet (corn) of the ducks (both females and males) was low in Fe, but dosing with Fe shot did not have a significant effect on the level of Cu in the livers. VanCampen (1971:221-222) stated, "The fac- tors that are most influential in determining the tissue levels of copper are age, hormones, disease and diet. . . . calcium apparently can either in- crease or decrease copper absorption, depending on the composition of the diet to which they are added." In our study, females had higher mean con- centrations of Cu in their gonads than males, which is in contrast to kidneys and livers where males always had higher mean concentrations of Cu. Phosphorous The lower concentrations of P in livers of Fe- dosed males, as compared with females, resulted from decreases of P in the livers of Fe-dosed males. Concentrations of P were only slightly higher in the livers of Fe-dosed females than in 0- dosed females. Dosing with Bi shot also caused a decrease in the concentration of P in livers of males, which resulted in a significant difference between the sexes. Concentrations of P were essentially the same in 0-dosed females and Bi- dosed females (Table 11). Iron Dosing with Fe shot resulted in large concentra- tions of Fe deposits in the livers, but dosing with Bi shot did not significantly affect the concentra- tions of Fe in the liver (Table 11). Although females dissolved a higher percentage of the dosed Fe shot than males (Table 1), the Fe-dosed females did not have significantly higher concentrations of Fe in their livers than the Fe-dosed males. Females had significantly higher concentra- tions of Fe in their gonads than males. These sex differences in the concentrations of Fe in the go- nads may be related to the preparation of the ovaries for egg laying. No differences were found in the mean concentrations of Fe in the gonads attributed to dosing with either Fe shot or Bi shot. Calcium It appears that dosing with Fe or Bi shot is associ- ated with lower concentrations of Ca in the livers and kidneys of ducks as compared with controls (Table 9). Forth and Rummel (1971:182) reported that increases in Fe or Ca mutually inhibited each other in their transfer through the small intestine of the rat. They concluded that it was possible there is ". . .a common transport mechanism for iron and calcium . . . ." Perhaps Bi induces a similar reduction in the transfer of Ca, although Bi has apparently not been studied in this context. Ca increased substantially in the plasma of both males and females for all dosed groups from Day to Day 15 to Day 30 (Table 14). The increase cannot be related to increase in Ca in the diet because after dosing all ducks were on a corn diet, and corn is low in Ca. Ca in the plasma among doses did not vary statistically. Feces Both Bi and Sn greatly increased in the feces of Bi- dosed ducks the day after dosing. Birds excreted Bi in the feces at high concentrations to the end of the 30-day study. Mean concentrations of Bi were not substantially different between 0-dosed and Fe-dosed ducks on Days 0, 1, and 2. Bi was much higher in feces of Bi-dosed ducks than in either 0- dosed or Fe-dosed ducks on Days 1 and 2. It appears that almost all of the Bi dissolved from Bi April 1997 Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards 213 shot in the gizzards is excreted in the feces of ducks. The mean concentration of Sn in the feces of Bi-dosed ducks was higher at the end of the 30- day study than the background level found the day prior to dosing. However, the mean concen- tration of Sn in the feces of Bi-dosed ducks de- clined substantially after Day 10. These findings for Sn in the feces seem to support Underwood (1971), who reported that the available evidence for humans shows that Sn is poorly absorbed, poorly retained, and excreted primarily in the feces. In humans, the amount of Sn ingested with food was approximately the same as the amount excreted in the feces. Underwood's conclusion was that Sn shows little toxicity, probably because it is absorbed slowly and is excreted rapidly in feces. The mean concentration of Fe in feces de- clined sharply for both 0-dosed and Bi-dosed ducks starting on Day 1 . The decline continued to the end of the study in Bi-dosed ducks. Feces of 0- dosed and Fe-dosed ducks were not analyzed for the entire study. The decline of Fe in feces of 0- dosed and Bi-dosed ducks was probably a result of switching on Day from a diet of commercial duck food to corn, which is low in Fe. With two exceptions (kidneys of males, mean = 0.528 ug/g [range 0.36 to 0.72 Ug/g], and livers of males, mean = 0.246 |ig/g [range 0.18 to 0.37 Ug/g]), Bi was not found in the livers, kidneys, or gonads of 0-dosed ducks. The mean concentra- tions of Bi in kidneys of Bi-dosed ducks were 8.05 (ig/g for females (range 4.69 to 12.6 Ug/g) and 4.77 (ig/g (range 2.02 to 8.82 Ug/g) for males. The mean concentration of Bi in livers of Bi-dosed females was 2.79 Ug/g (range 1.19 to 5.63 Ug/g)- The differences for Bi-dosed versus 0-dosed ducks were significant for both kidneys and livers. Both macro and micro histological observations de- tected no toxic effects of Bi on the kidneys, liver, or gonads. Conclusions We detected no toxic effects in game-farm mal- lards dosed with six Bi/Sn alloy shot and ob- served for 30 days. Survival, body weight, Hct, and weights of organs were not affected. Gross and microscopic examination of the kidneys, liver, and gonads of the ducks also revealed only slight tissue changes. Our data support the conclusions of Sanderson et al. 1992, who reported no toxic effects in game-farm mallards dosed with 100% Bi shot. A number of differences in weights of organs and in mean concentrations of individual ele- ments were detected between females and males. These differences appear related to physiological changes associated with the onset of breeding, especially in egg-laying females. Although a few "anomalies" were linked to dosing with Fe shot or with Bi shot, no toxic effects were detected with either. For example, livers and kidneys ofboth Bi-dosed and Fe-dosed ducks had lower mean concentrations of Ca than livers and kidneys of 0-dosed ducks. The difference in the mean concentrations of Ca in the livers of Fe- dosed ducks versus Bi-dosed ducks was not sig- nificant (Table 11). 214 Illinois Natural History Survey Bulletin Vol. 35 Art. 3 Literature Cited Abbracchio, M.P., W. Balduini, A. Cavallaro, P. Adamoli, M. Fittipaldi, F. Muzio, S. Malandrino, and F. Cattabeni 1985. Brain and blood levels of bismuth after oral or parenteral administration of tripotassium-dicitrato bismuthate to rats. NeuroToxicology 6(3):139-144. Anderson, William L. 1992. Legislation and law- suits in the United States and their effects on nontoxic shot regulations. Pages 56-60 in D.J. Pain, ed. Proceedings of an international waterfowl and wetlands research bureau workshop, Brus- sels, Belgium. 1991. IWRB Special Publication 16, Slimbridge, UK. Canadian Wildlife Service. 1995. Minister Copps acts on two toxic substances. News release dated 11 July 1995. 6 pp. Dipalma, J.R. 1988. Bismuth toxicity. American Family Physician 78(5):244-246. Environment Canada. 1992. Guidelines regard- ing the toxicity tests required for the approval of candidate non-toxic shot (to be submitted to the meeting of the executive in January 1993). Envi- ronment Canada. 9 pp. Forth, W., and W. Rummel. 1971. Absorption of iron and chemically related metals in vitro and in vivo: specificity of the iron binding system in the mucosa of the jejunum. Pages 173-191 in S.C. Skorya and D. Waldron-Edward, eds. Intestinal absorption of metal ions, trace elements and ra- dionuclides. Pergamon Press, Oxford,New York, Toronto, Sydney, Braunschweig. 431 pp. Fowler, B.A., and V. Vouk. 1979. Bismuth. Pages 345-353 (Chapt. 20) in L. Friberg, G.F. Nordberg, and V. Vouk, eds. Handbook on the toxicity of metals. Elsevier/North-Holland Biomedical Medical Press, Amsterdam, New York. Glaser, J.A., D.L. Foerst, G.D. McKee, S.A. Quave, and W.L. Budde. 1981. Trace analyses for waste- waters. Environmental and Science Technology 15(12):1426-1435. Greenberg, A.E., L.S. Clesceri, and A.D. Eaton, eds. 1992. Standard method for the examination of water and wastewater. American Public Health Association, American Waste Water Works Asso- ciation, and Water Environment Federation. 18th ed. Section 31 13 Metals by electrothermal atomic absorption spectrometry: 3-20 through 3-28. Gregus, Z., and CD. Klaassen. 1986. Disposition of metals in rats: a comparative study of fecal, urinary, and biliary excretion and tissue distribu- tion of eighteen metals. Toxicology and Applied Pharmacology 85:24-38. Hamilton, E.J., M.J. Minski, and J.J. Cleary. 1972- 1973. The concentration and distribution of some stable elements in healthyhuman tissues from the United Kingdom. Science of the Total Environ- ment 1:341-374. Hanson, H.C., and R.L. Jones. 1974. An inferred sex differential in copper metabolism in Ross' geese (Anser rossii): biogeochemical and physi- ological considerations. Arctic 27(2):1 11-120. Hanzlik, P.J., and E. Presho 1923. Comparative toxicity of metallic lead and other heavy metals for pigeons. Journal ofPharmacologyand Experi- mental Therapeutics 21(2):145-150. Haseltine, S.D., and L. Sileo. 1983. Response of American black ducks to dietary uranium: a pro- posed substitute for lead shot. Journal of Wildlife Management 47:1124-1129. Hill, C.H., and G. Matrone 1961. Studies on copper and iron deficiencies in growing chickens. Journal of Nutrition 73:425. Hillemond, P., M. Palliere, B. Laquais, and P. Bauvet. 1977. Traitment bismuthique et bismuthemie. Semaine des Hopitaux de Paris 53:1663-1669. International Commission on Radiological Pro- tection. 1960. Report of Committee II on permis- sible dose for internal radiation. International Commission on Radiological Protection, Publica- tion No. 2. Pergamon Press, Oxford 218-219. Irby, H.D., L.N. Locke, and G.E. Bagley. 1967. Relative toxicity of lead and selected substitute shot types to game farm mallards. Journal of Wildlife Management 31:253-257. Key, M.M., A.F. Henschel, J. Butler, R.N. Ligo, and I.R. Tabershha, eds. Lorice Ede, manuscript ed. 1977. Occupational diseases. A guide to their rec- ognition. Bismuth and compounds. Page 338. United States Department of Health, Education, April 1997 Toxicity of Ingested Bismuth Alloy Shot in Game-farm Mallards 215 and Welfare, Public Health Service, Center for Disease Control, National Institutes of Occupa- tional Safety and Health. Revised Edition. Krigman, M.R., T.W. Bouldin, and P. Mushak. 1985. Metal toxicity in the nervous system. Mono- graphs in Pathology 58-100. Lee, S.P. 1981. Studies on the absorption and excretion of tripotassium dicitrate bismuthate in man. Research Communications in Chemical Pa- thology and Pharmacology 34:359-364. Locke, M., H. Nichol, and C. Kotola-Pirie. 1987. Binding of bismuth to cell components: clue to mode of action and side effects. Canadian Medi- cal Association Journal 137:991-992. Longcore, J.R., R. Andrews, L.N. Locke, G.E. Bagley, and L.T. Young. 1974. Toxicity of lead and proposed substitute shot to mallards. U.S. Department of the Interior, Fish and Wildlife Service, Special Scientific Report—Wildlife No. 183. 23 pp. Matrone, G. 1960. Studies on copper and iron deficiencies in growing chickens. Journal of Nu- trition 93:425. Moser, M. 1992. International Lead Poisoning Newsletter. International Waterfowl and Wet- lands Research Bureau, September. 1992. 17pp + appendices. Oehme, F.W., ed. 1979. Pages 603-605 in Toxicity of heavy metals in the environment. Part 2. Bis- muth. Marcel Dekker, Inc., New York and Basel. Ross, J.F., Z. Sahenk, C. Hyser, J. P. Mendell, and C.L. Alden 1988. Characterization of a murine model for human bismuth encephalopathy. NeuroToxicology 9(4):581-586. Sanderson, G.C., S.G. Wood, G.L. Foley, and J.D. Brawn 1992. Toxicity of bismuth shot compared with lead and steel shot in game-farm mallards. Transactions of the North American Wildlife and Natural Resources Conference 57:526-540. Sanderson, G.C., W.L. Anderson, G.L. Foley, K.L. Duncan, L.M. Skowron, J.D. Brawn, and J.W., Seets 1997b. Toxicity of ingested bismuth /tin alloy shot in game-farm mallards: chronic health effects and effects on reproduction. Illinois Natu- ral History Survey Bulletn 35(4):21 7-252. Serfontein, W.J., and R. Mekel. 1979. Review of bismuth blood and urine levels in patients after administration of therapeutic bismuth formula- tions in relation to the problem ofbismuth toxicity in man. Research Communications in Chemical Pathology and Pharmacology 26:391-411. Slikkerveer, A., and FA.de Wolff 1989. Pharma- cokinetics and toxicity of bismuth compounds. Medical Toxicology and Adverse Drug Experi- ence 4:303-323. Thomas, D.W., T.F. Hartley, P,. Coyle, and S. Soecki. 1988. Bismuth. Chapter 11, pages 115-127 in H.G. Seiler and H. Segil, eds. Handbook on toxicology of inorganic compounds. Marcel Dekker, Inc., New York and Basel. Underwood, E.J. 1971. Trace elements in human and animal nutrition. 3rd Ed. Academic Press, New York and London. 543 pp. Van Campen, D.R. 1971. Absorption of copper from the gastrointestinal tract. Pages 211-227 in S.C. Skoryon and D. Waldron-Edward, eds. In- testinal absorption of metal ions, trace elements and radionuclides. Pergamon Press, Oxford,New York, Toronto, Sydney, Braunschweig. 431 pp. Venugopal, B., and T.D. Luckey. 1978. Chemical toxicity of metals and metalloids. Pages 215-219 and 354-401 in Metal toxicity in mammals. 2. Plenum Press, New York and London. Woods, J.S., and BA. Fowler 1987. Alteration of mitochondrial structure and heme biosynthetic parameters in liver and kidney cells by bismuth. Toxicology and Applied Pharmacology 90:274- 283. Sanderson, G.C., W.L. Anderson , G.,L. Foley, S.P. Havera, L.M. Skowron, J.W. Brawn, G.D. Taylor, and J.W,. Seets. 1997a. Effects of lead, iron, and bismuth alloy shotembedded in the breast muscles of game-farm mallards. Journal of Wildlife Dis- eases. In press. 216 Illinois Natural History Survey Bulletin Vol. 35 Art. 3 UNIVER9ITY OF ILLIN0I9-URBANA 3 0112 044858188 Illinois Natural History Survey Natural Resources Building 607 East Peabody Drive Champaign, Illinois 61820 217,333-6880 A Division of the Illinois Department"of Natural Resources ILLINOIS NATURAL RESOURCES