9 on """""~‘0HNW ....... THE EFFECTS OF MOIST AAU DRY HEAT UUUIUIIU I UPON MERCURY IIIALS III BAUASI TISSUE AAU THE , ’ ; :3 in; -- ; Q =: DISTRIBUTION OF MERCURY III UAUAIIS 0F UUUAAw “ ' ’ UALLAIIU UUCIIs AUAIIAISIEAEU IIIEIIIII MERCURY __ . . z CHLGRIDE Thesis for the DUAIUU Uf M S MICHIGAN STATE UNIVERSITY ELIZABETH JOY ROUGH . r' ' - 1972 ' LTP" 'QY Micki; State 3 Univcx .ity ABSTRACT THE EFFECTS OF MOIST AND DRY HEAT COOKING UPON MERCURY LEVELS IN BREAST TISSUE AND THE DISTRIBUTION OF MERCURY IN ORGANS 0F MCGRAN-MALLARD DUCKS ADMINISTERED METHYL MERCURY CHLORIDE BY Elizabeth Joy Hough Fifty McGraw Mallard ducks. (Anas platyrhynghos). divided into groups of 6 males and 4 females each. were administered methyl mercury chloride by gelatin capsule at dosage levels of 0.0. 2.0. 5.0, 6.0, and 10.0 mgMeHg/kg body weight. Birds given the 5.0 mg/kg dose were fed an additional 14 mgMeHg/bird via treated feed. The left side of breasts from half of the birds were browned for 6 min at 200°C and then braised at 90°C for 30 min and those from the other half were roasted at 177°C to an internal temperature of 90°C. Right sides of breasts from all birds provided uncooked samples which were analyzed for total mercury by flameless atomic absorption techniques as were the cooked breasts and the cooking drip. The concentration of mercury in the raw liver, kidney. thigh, heart. and brain was also analyzed for three birds from each dosage group to examine organ distribution. Mercury contents were calcu-- lated as ppm on a dry weight basis. The amount of mercury recovered after cooking in the combined ' cooked breast tissue and drip as compared with the uncooked breast tissue averaged 87.6% after braising and 96.6% after roasting. However. the difference in mercury concentration between uncooked tissue and that cooked by either method was not statistically significant. Roasted drip Elizabeth Joy Hough accounted for about 0,4% of the mercury recovered after cooking and braised drip accounted for about 0.8% recovery. The concentration of mercury in both cooked and uncooked tissue increased directly with methyl mercury dosage level administered. The concentration of mercury in drip showed a similar increase with the exception of roasted drip from the group given 10.0 mgMeHg/kg which had an average ppm of mercury lower than that for the roasted drip from the 6.0 mg/kg dose level. Comparison among heart, brain. liver, kidney, breast and thigh tissues from birds in the different dosage groups also revealed a direct relationship between the dose of methyl mercury given and tissue mercury level. Kidney accumulated the highest level of mercury and liver accumulated the second highest level of mercury. The rate of increase with increasing levels of methyl mercury administered was higher in kidney and liver than the levels in the other tissues. Average concentra- tions of mercury in breast, thigh, brain and heart were not significantly different although breast was consistently higher than thigh. THE EFFECTS OF MOIST AND DRY HEAT COOKING UPON MERCURY LEVELS IN BREAST TISSUE AND THE DISTRIBUTION OF MERCURY IN ORGANS 0F MCGRAW-MALLARD DUCKS ADMINISTERED METHYL MERCURY CHLORIDE BY Elizabeth Joy Hough A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science and Human Nutrition 1972 ,I I, ‘1 :‘I l '. .‘ ‘ ’ 1 a ‘1‘ fi‘ \ . I \' (I ACKNOWLEDGMENTS This study is dedicated to our environment with deepest gratitude to all those whose contributions have made completion of this project possible. ii TABLE OF CONTENTS INTRODUCTION ............. . ............ LITERATURE REVIEW ....................... Environmental Mercury Contamination . . .I. ..... . ..... . Sources . ...................... Hildlife ....................... Mercury Distribution and Metabolism in Animals ...... Experimental Administration ....... . ..... Avian Studies . . . . . ....... . .. . . . . . Metabolism ...................... Human Toxicity ...................... Clinical Aspects ...... . ............ Toxicity Levels .................. Human Exposure to Methyl Mercury from Foods ........ Level in Food .................... Cooking Studies . . . ........ . . Mercury Analysis Using Atomic Absorption ......... Digestion ...................... Analysis ...................... Interference ..................... EXPERIMENTAL PROCEDURE . . . . ................. The Sample . . .................... Mercury Administration ................ Processing ..... . ................ Cooking Procedures .................... Braisihg . . . .................... Roasting . . . . ................... Cooking Losses . . . . . . ........ . ..... Analysis Procedures . . .................. Moisture Determination ................ Hot Digestion .................... Analysis . . . . . . ................. Analysis of the Data . . ............... Time-Temperature Relationships ............ RESULTS AND DISCUSSION . . . .................. Physical Parameters of Cooking . . . . . ...... . . . Time-Temperature Relationships . ........... Cooking Losses . . . . . ........... Effects of Cooking Upon Mercury Concentration ....... Organ Distribution of Mercury . . . . . . . . . . . . . . . SUMMARY AND CONCLUSIONS . . . . . . . . . .......... RECOMMENDATIONS FOR FURTHER RESEARCH . ............. LITERATURE CITED ................... . . . . . iii TABLE 10 ' 11 12 LIST OF TABLES Page Content of mercury in organs of hens fed various 10 levels of methyl mercury dicyandiamide treated grain (Tejning, 1967). Averagea percentages and standard deviations of total, 33 drip and volatile cooking losses of duck breasts cooked by two methods. Analyses of variance for total. drip and volatile 34 cooking losses among breast samples due to cooking method, mercury dosage level and sex of bird. Averagea ppm of mercury on a dry weight basis. 36 before and after cooking, in duck breast tissue and in cooking drip. Analyses of variance of ppm of mercury in raw and 38 cooked duck breast tissue and in drip attributable :0 cooking method, mercury dosage level and sex of ird. Combined analyses of variance of ppm of mercury and 40 of total pg of mercury in raw and cooked duck breast tissue and in drip attributable to cooking method. mercury dosage level, sex of bird and state (i.e. cooked. uncooked or drip). Total micrograms of mercury in duck breast samples before 42 and after cooking and in cooking drip. Content of mercury with standard deviations and per- 46 centage solids in tissues of McGraw-Mallard ducks administered methyl mercury at five different levels. Two-way analysis of variance for the concentration of 48 mercury in selected tissues of ducks given methyl mercury at dosage levels of 0.0. 6.0. 10.0. 19.0 mgMeHg/kg of body weight. Significant differencesa among the mean concentrations 48 of mercury in different organs.of McGraw-Mallard ducks given MeHg at levels of 0 - 19 mg/kg. Significant differencesa among the mean concentrations 49 of mercury in each organ of McGraw-Mallard ducks given MeHg at levels of 0 - 19 mg/kg. The relationship between the level of methyl mercury 51 administered and accumulation in tissues. iv FIGURE 1 LIST OF FIGURES Page Time-temperature relationships for roasted and 31 braised duck breast samples averaged at 3 minute intervals. Percentage mercury recovered in duck breast tissue 43 and drip after roasting or braising as compared to raw tissue. Least squares regression lines for the relationship . 52 between the level of methyl mercury administered and accumulation in tissues. INTRODUCTION Mercury contamination of the environment has recently been recognized as a serious potential threat to man. Agricultural. industrial, and other human use of mercury compounds has resulted in measurably elevated levels of mercury in water and air (Hershaw, 1970; Hilliston, 1968). Environmental mercury becomes increasingly concen- trated as it progresses through the natural food chain (D'Itri, 1972; Hannerz. 1968). The result is that animals, particularly those who are predators (Borg gt_al,. 1969 and 1970; Fimreite. at 91,, 1970) or have long life-spans (Johnels. gt_al,. 1967; Bache §t_al,. 1971). have been found to have high tissue levels of mercury. Mono-methyl mercury is the form of mercury most toxic to man (Berlin gt_al,. 1969; Nelson gt_§l,, 1971) and to animals (Grolleau and Giban, 1966; Swensson and Ulfvarson. 1969). Varying degrees of irre- versible neurological damage were observed in persons having ingested toxic quantities of this compound from fish with high tissue levels of methyl mercury (Tokuomi. 1969; Takeuchi. 1970). Moreover. many forms of mercury have the potential to be methylated in nature (Dunlap. 1971; Jenson and Jerneldv. 1969; Mood §t_al,. 1968). Nearly all of the mercury in animal tissue is methyl mercury (Hestbb and Rydalv, 1969; Kiwimae eta}... 1969). Animal tissues. which provide man's most highly concentrated food source of methyl mercury. are rarely consumed without being cooked or processed in some manner. Thus. there is the possibility that the mercury content of food may be altered by processing or preparation prior to consumption. Only limited and conflicting reports are available on the effects of cooking on the concentration of mercury in muscle tissue. Norén and Hestbd (1967) reported increased mercury levels while Jergrelius (1971) reported decreased mercury levels in pike after frying and boiling. Results from both studies seem to be based upon a single sample for each cooking method. Further examination of the fate of methyl tissue as influenced by various cooking treatments is justified. In this study the mercury content of McGraw-Mallard duck (Ag§§_ platyrhynchos) breast tissue and of drip in samples cooked by different methods was compared with mercury content of uncooked breast tissue, from the same bird. Ducks were given different doses of methyl mercury ranging 0-19 mg/kg body weight and interrelationships among quantity of methyl mercury given. magnitude of mercury concentration in muscle. and influence of cooking treatment were examined. Although there are limited data on.the distribution of mercury in the tissues of other avian species, the distribution patterns in ducks have not been elucidated. Thus. the concentration of mercury in selected tissue and organs of these experimental ducks was also determined. LITERATURE REVIEW Contamination of the environment with mercury is a matter of concern due to its potentially harmful effects on wildlife and humans. Consequently, the literature was examined to determine the extent of known contamination. (Studies on distribution and metabolism of methyl mercury in animals and the mechanics of protein binding were reviewed. The problems of human toxicity were considered as well as degree of human exposure to methyl mercury through foods. Laboratory analysis for mercury by atomic absorption wasdiscussedI 3 Environmental Mercury Contamination The distribution of mercury in the environment results from natural occurrence and human activities. The latter source has aroused great concern as it has been implicated in causing elevated levels of mercury in the tissues of many wildlife species and in causing fatalities in many cases. W In 1968 people of the United States used more than 5.7 x 106 pounds of mercury or 30% of the world consumption and the trend is toward increased consumption (West. 1969). Undetermined portions of this mercury do escape into the environment. Mercury consuming industries in the United States. listed in descending order of amount of mercury used are: electrical apparatus. industrial control instruments. general laboratory use. chloralkali. paint. agriculture. dental. catalyst. paper and pulp. pharmaceutical and cosmetic. amalgamation. and miscellaneous uses. Large quantities of mercury may also be released into the environment by burning fossil fuels and smelting commercial ores ((D'Itri, 1972). Wildlife 'Using feathers of museum specimens as an index of body burden of mercury. Berg gt_al, (1966) found 10-20 fold elevations in the mercury content of various avian species which coincide with the onset of wide- spread industrial and agricultural use of mercurials in Sweden. This presents a strong case for considering mercury pollution to be the ' causative factor in wildlife contamination. I I Borg gt_al, (1969) conducted an extensive 8 year investigation of ‘mercury levels in game found dead in Sweden as well as animals shot for examination. Of 253 seed-eating birds found dead. 48% had mercury levels in liver of above 2 mg/kg, 30% were above 5 mg/kg, 20% were above 10 mg/kg, and 13% had more than 20 mg/kg. Of 298 seed-eating birds shot for investi- gation, 41% carried liver mercury levels of above 2 mg/kg, 19% were above 5 mg/kg. 12% were above 10 mg/kg, and 4% had more than 20 mg/kg. In the seed-eating birds examined, the incidence of birds at high mercury residue levels was significantly increased in late Spring and autumn as compared with other seasons. thus indicating a correlation with spring and autumn showing of seed treated with organo-mercury fungicides. 0f the 412 predatory birds which were shot or found dead. 62% had liver mercury levels exceeding 2 mg/kg. 36% exceeded 5 mg/kg. 19% exceeded 10 mg/kg. and 11% had more than 20 mg/kg. Other birds found dead included a Mallard duck with a level of 62 mgHg/kg in mixed liver and kidney tissues and a Sheld duck with 40 mgHg/kg in the liver. Mercury level in livers of birds collected in Alberta and Saskatchewan were reported by Fimreite gt'al, (1970). Average levels for seed-eating birds from Alberta were 1.00-2.84 ppm and for those from Saskatchewan were 0.10-0.55 ppm. Average levels for their avian predators were 0.76-6.84 ppm for birds from Alberta and 0.45-0.73 ppm for birds from Saskatchewan. It was noted that in Alberta the usage of organo- mercury seed-dressing is more common than in Saskatchewan. * Dustman gt_al, (1970) found liver levels of 0.18-5.60 ppm of mercury for several species of ducks collected on Lake St. Clair in 1970. Birds from Michigan were also examined for mercury levels in breast tissue by Youatt and Zabik (1971). Levels ranging 0.01-0.64 ppm were found in pheasant samples collected from 21 counties in northern Michigan. The average mercury level for pheasants was 0.10 ppm and only 2 showed levels above 0.30 ppm. waterfowl from the Lake St. Clair - Detroit River area were analyzed for mercury in breast tissue. Levels for the 39 birds ranged 0.01-1.76 ppm and averaged 0.65 ppm.‘ Fourteen had levels lower than 0.50 ppm. Included in this sample were 3 Mallard ducks with a mean mercury concentration of 0.11 ppm and 25 Lesser Scaup with a mean concentration of 0.83 ppm. Differences in these two duck species may be due to feeding habits since Mallards are surface feeders while Scaup are divers. These studies support the hypothesis that in areas of higher mercury use. wildlife species have higher tissue levels of mercury than in areas of lower use. Since toxic levels have been found in animals. it is imperative that examination be made of the full extent of mercury contamination. its human implications and methods for reducing human intake. Mercury Distribution and Metabolism in Animals There are still many questions unanswered on the action of methyl mercury after it enters the body although several reasearch studies have recently been done in this area. Various methyl mercury compounds have been administered orally and intravenously in single doses or over time periods. Metabolic responses and distribution of the compound in various body tissues have been examined. Experimental Administration Méthyl mercury has been administered by capsule. in treated feed. and intravenously by different investigators and all found that it was absorbed and distributed similarly in body tissues. Grolleau and Giban (1966) administered methyl mercury dicyandiamide in gelatin capsules to various avian species. When given in this manner. it was found to be toxic proving its absorption in the gastrointestinal tract. Backstrom (1969) studied distribution in quail given methyl mercuric nitrate via a stomach tube or intravenously in a radial vein. Distribution was essentially the same for both methods of mercury administration. These findings of methyl mercury absorption from the digestive tract may be related to its acid stability and fat solubility. The methyl mercury compound selected for administration to animals varies among studies; however. Ulfvarson (1962) demonstrated that the anion does not appreci- ably influence the toxicological properties of the organo-mercury complex. Avian Studies 203Hg labeled) BAckstrom (1969) administered methyl mercuric nitrate ( in a single dose intravenously to quail. ‘He found differences in the concentration of mercury in organs by sex. Female blood. liver. kidney. *and brain reached maximum concentration sooner and greatly reduced their mercury load between the 10th and 30th day after exposure while males only slightly reduced their concentration of mercury during that time. These differences were probably due to excretion of methyl mercury in eggs. One day after exposure. the concentration of mercury in kidney and liver (c. 1000% of the g dose/g tissue) was much greater than in muscle and 'brain (c. 100% Of the g dose/g tissue). Some muscles (e.g. superficial pectoral) concentrated more than others. Small mounts of mercury appeared in the subcutaneous fat depots one hour after injection but no accumulation was seen over a longer survival time. 7 Borg gt_al. (1970) fed chicken tissues containing physiologically incorporated methyl mercury to 4 goshawks. After death due to alkyl mercury poisoning. goshawk organs were analyzed for mercury. A total intake of 18-20 mgMeHg/bird resulted in the following average organ concentrations of mercury (mgHg/kg wet wt): liver - 128. kidney - 120. skeletal muscle - 42. and brain - 41. Control birds had very low levels of 0.2-4.7 mgHg/kg in all tissues analyzed. Methyl mercury (Panogen 15) treated feed was given to chicks so that total intake per bird over a three week period was about 0.0. 1.7. 3.4. 5.1 mg yielding liver levels of 0.03. 3.94. 7.25. 10.00 ppm mercury respectively (Fimreite. 1970). Growth retardation occurred at all levels and 5.1 mgHg/bird was considered toxic. Methyl mercury dicyandiamide dressed seed was incorporated into the diet of hens by Smart and Lloyd (1963). Each of the 5 hens in group A & B consumed about 2.8 & 3.5 mgHg/week respectively over 4-8 week periods. The mean level of mercury found in breast and thigh muscle was 7.2 and 4.1 ppm. respectively. in group B and 4.5 and 3.2 ppm. respectively. in Group A. The mean mercury level in liver was 11.5 ppm for group A and 17.2 ppm for group B. Kidneys from only one hen in each group were analyzed and an average of 7.5 ppm of mercury was found. Mean levels in group B were higher than group A. There were no significant differences attributable to duration of feeding. The difference between the amount of mercury taken in by a hen and that contained in the eggs did bear a direct relation to the mercury found in tissues (particularly breast and liver). General health of the birds remained normal throughout the feeding period. 8 White Leghorn cocks were fed for three weeks with wheat dressed with methyl mercury hydroxide at'a concentration of 16mgHg/kg by . Swensson and Ulfvarson (1969). Higher mercury concentrations were produced in the liver than in the kidneys. A steady state in the blood and muscles was reached after 2 weeks of feeding and there was no significant decrease in mercury level 2 weeks after cessation of mercury feeding. A similar steady state was reached for brain at 3 weeks. Average mercury levels of organs in this steady state were about 9.4 ppm for blood. 7.1 ppm for muscle. and 6.1 ppm for brain. Mercury concentration in liver (30.5 ppm) and kidney (26.8 ppm) was highest after 2 weeks of feeding. After 5 weeks the levels were 21.1 and 12.2 ppm mercury for liver and kidney. respectively. No deaths or symptoms of disease were observed. Swensson and Ulfvarson also reported in this paper that the approximate lethal dose after single intravenous injections in cocks was 30 mgMeHg/kg of body weight. Hens were given feed treated with methyl mercury hydroxide to yield daily doses of 0.4 or 1.6 mg over a period of 140 days (Kiwimde at 31, 1969). The concentration_of mercury in muscle for one bird given 1.6 mgMeHg was 1.25 mg/kg. 94% of which was methyl mercury. Muscle of one bird fed 0.4 mgMeHg/day had 3.90 mgHg/kg. 100% of which was methyl mercury. _ Borg gt_a1, (1969) fed pheasants wheat dressed with methyl mercury dicyandiamide which had a mercury content of about 20 mg/kg. Death resulted after 29-61_days and the mercury residues at this time were 30-130 mg/kg in mixed liver and kidneys and 20-45 mg/kg in muscle tissue. Jackdaws fed a similar ration died after 26-38 days. Mercury residues amounting to 70-115 mg/kg were found in mixed liver and kidney. Four out of six magpies fed this dressed wheat every third day died within 35-71 days. Mercury levels of 50-200 mg/kg in mixed liver and kidney tissues were found at death. Swensson and Ulfvarson (1968) reported a fairly even distribu- tion of mercury in all organs at intervals after a single intra- venous dose of methyl mercury hydroxide. A dose of 6.0 mg/kg (estimated as 1/5 of the L050) yielded. after 10 days. average levels of 4.6. 7.2. 8.1. 4.3. 3.8 ppm mercury in blood. liver. kidneys. muscle. brain. respectively. Delayed increase in brain concentration of methyl mercury was noted and brain concentration continued to increase over time. Tejning (1967) fed 21 hens mixtures of Panogen (methyl mercury dicyandiamide) treated grain plus feed. Duration of feeding and mercury content of ration were varied. Average mercury concentrations found in organs are summarized in Table 1. With only 2 exceptions breast muscle contained more mercury than did thigh muscle. Liver and kidney had higher levels of mercury than did other organs. 10 Table 1. Content of mercury in organs of hens fed various levels of methyl .mercury-dicyandiamide treated grain (Tejning. 1967). Mean Ration Duration Mean Mercury Content (p919) in Tissues Group (ngg/g) (days) Breast Thigh Heart Brain Kidney Liver A 4.4 39.7 2.9 2.7 - 4.3 13.1 16.8 B + E 9.0 42.5 7.8 5.0 3.2 6.7 18.2 21.3 C 17.6 15.5 8.9 7.2 - 8.2 29.5 35.0 D + G 18.0 39.5 13.9 10.8 15.1 12.8 48.0 59.1 F 9.2 87.5 8.8 7.7 8.0 9.2 30.8 27.2 H 18.4 85.5 15.8 12.5 18.2 13.3 47.8 55.2 There was little accumulation of mercury in subcutaneous fat. Mercury content in muscle and organs generally reflected the increasing amounts of methyl mercury administered. AccumuIation rate does. however. vary with the particular organ or tissue as well as with the experimental animal. The concentration of mercury found in breast. brain. heart. and thigh are similar although thigh levels were reported slightly lower than breast. Liver and kidney were higher in mercury than other tissues. No consistent differences between the two organs were elucidated. Metabolism Very little is actually known about the metabolism of methyl mercury at the biochemical level. Methyl mercury behaves differently than other mercurials in that it is distributed rather uniformly throughout the body of most animals studied and is more slowly excreted. There is a very strong covalent bond between the methyl carbon and the mercury atom which largely remains intact in the body (Backstrom. 1969). In fact. there have been indications that this stable compound is formed in the body from other mercury compounds (Kiwimde gt_al.. 1969; Borg gt_al,. 1960; Gage. 1964). 11 Although methyl mercury is fat soluble (Hughes. 1957). only minute amounts are found in animal fat (Backstrom. 1969; Tejning. 1967). The protein binding characteristics of methyl mercury appear to play a significant role in its metabolism. The methyl mercuric ion as well as other mercury compounds are known to have an extremely high affinity to animal protein and especially to the sulfhydryl groups (Hughes. 1957; Passow at 31, 1961; Benesch and Benesch. 1962). This is well illustrated by the results of a study by Kiwimae gt_al, (1969). Various mercury compounds including methyl mercury were fed to laying hens at levels of 0.4 or 1.6 pg/day. Whites and yolks of the eggs produced were analyzed separately for mercury content. When methyl mercury was fed the mercury concentra- tion was very high in the white (10 or 400 mgHg/kg) compared to yolk (2 or 10 mgHg/kg). The feeding of other mercury compounds resulted in higher mercury concentrations in yolk than in albumen. Yolk contains virtually all of the lipid material in the egg and the albumen is known to be high in sulfhydryl groups. The results of Kiwimae's study probably reflect these compositional differences. Ostlund (1969) found that the excretion rate was significantly slower when large non-toxic doses of methyl mercury were given than when trace doses were given. He offered two hypotheses. First. he postulated that the excretion mechanism may have become so highly saturated as to be damaged or non-functional. The second hypothesis was that there are two types of binding sites for mercury each with different affinities for methyl mercury. One type of binding site could be a labile site which is first saturated and rapidly cleared. 12 possibly by the kidney. and the second type with greater affinity for methyl mercury would be slowly cleared, probably by liver detoxification systems. This would be consistent with his finding of initial excre- tion via kidney and later excretion at a lower rate in the feces. Other factors consistent with Ostlund's binding hypothesis are his finding of the concentration of mercury in plasma proteins proportional to the body level while that in blood cells greatly increases with higher body levels. One explanation of this is second site binding in the cell membranes. Arrhenius (1967) has shown. also. that the liver detoxification enzymes are constituents of endoplasmic reticulum which would be consistent with second site binding in membrane structures. Bdckstrom (1969) administered methyl mercury to several species of freshwater fish. The skeletal musculature showed relatively low concentration in white muscles whereas there was a pronounced uptake of mercury in the red muscles. White muscles are characterized by anaerobic. glycolytic enzymes which are found in the cytoplasm while red muscles are characterized by oxidative. mitochondrial enzyme systems which are membrane bound (Lawrie. 1966). It is conceivable that these histochemical differences influence the differential distribution of methyl mercury. Another factor may be that lipid content is higher for white than for red muscles (Beecher gt 31,. 1968). Ganther gt_al, (1972) has reported that selenium may decrease the toxicity of methyl mercury to animals. Both quail and rats survived longer on diets containing methyl mercury plus selenium than on diets with equivalent levels of methyl mercury only. 13 Human Toxicity Between 1953-1960 there were 121 cases of methyl mercury poison- ing in Minamata. Japan. 46 of which were fatal. In Nigata. Japan. . between 1965-1970. 47 cases. including 5 deaths. of methyl mercury poisoning were reported. Stricken persons had conSumed locally caught fish and shellfish containing up to 50 ppm of mercury. 'Such high tissue loads of mercury were accumulated by the fish and shellfish from waters contaminated by the effluent of vinylchloride and acetaldehyde factories (Takeuchi. 1970; Tokuomi. 1969). These events proved that mercury pollution can be a serious threat to humans. Clinical Aspects _ The biological half-life of methyl mercury in man is 78 :_5 days. Symptoms do not immediately follow exposure since there is a latency period of 1-2 months. The central nervous system is the critical organ for exposure to methyl mercury (Berglund at al.. 1971). Of the body burden of mercury. the brain accumulates 15-20% and the remainder is rather evenly distributed throughout the body. Severe poisoning caUses brain lesions and cerebellar atrophy of the granule cells with pre- ferential injury to the calcerine and also to other cortical regions (Tokuomi. 1969). Methyl mercury penetrates the placental barrier. The concentra- tion of mercury in bloodcorpuscles of the newborn averages 28% greater than in the mother's corpuscles (Tejning. 1968). Toxicity Levels Mercury concentration in blood increases directly with methyl "mercury.intake providing a good diagnostic index of the body burden of mercury. In Sweden a maximum allowable concentration of 0.1 ppm mercury l4 in whole blood is the legal limit for persons of non-fertile age (LUfroth. 1969). No recommendation was made as to the allowable mercury level in the blood of fertile age persons. Berglund and Berlin (l969) have presented two estimates of "allowable daily intake" (ADI) for methyl mercury. They calculated the methyl mercury intake in the average Swedish diet and the mercury content of erythorocytes of a healthy subject (l0 ng Hg/g) to both be equivalent to 0.l mg Hg/day (0.7 mg Hg/week). The second ADI was calculated from the biological half-life. the bodily distribution and the lowest toxic level of methyl mercury to yield an ADI equivalent to 0.06 mg Hg/day (0.42 mg Hg/week). The authors caution that no assessment of fetus sensitivity has been included in setting these ADI. Another aspect which has not been fully examined is the possi- bility of chromosome changes. Methyl mercury has been shown to have considerable potency as a mitosis disrupting agent causing pOlyploidy and chromosome disjunction in a concentrated substrate of 50 ng MeHg/g. However. no chromosome abberation has been reported for human cases (Ramel. l967). One of the major problems involved in establishing toxicity levels for methyl mercury is that there is no real threshold at which mercury is tolerated. There is only a threshold in the number of brain cells which must be damaged before symptoms become visible. Little is known about the long term effects of exposure to low levels of mercury such as those currently present in the environment. 15 Human Exposure to Methyl Mercury from Foods Traces of mercury have been found in all plant and animal materials. In general. plant derived food products contain much lower residues than do animal derived protein foods. The effect of cooking upon mercury levels in foods is unclear. Level in Food Smart (1968) reported mercury levels in the following foods: apples 0.001-0.12 ppm. pears 0.04-0.26 ppm. tomatoes 0.01-0.50 ppm. potatoes 0.01-0.17 ppm. grain 0.005-1.0 ppm. The higher levels of mercury residues were due to excessive applications of mercury containing chemicals at some stage during plant growth or due to poor soil drainage as in the case of grain (rice). Westh (1969) found the following average mercury concentrations (ppm) in animal foods from Sweden: Pork chop 0.03. pork liver 0.06. beef filet 0.012. hen's egg 0.029. Pike 0.04-8.40. crayfish 0.21-0.50. The levels of mercury found in wildlife (discussed previously) generally exceed those found in foods from domestic sources. Cooking.Studies When meat is heated there is a loss of water holding capacity due to denaturation of the sarc0plasmic and myofibrillar proteins. Solubility of these proteins decreases with increased heating time and temperature and is greatest between 40-600C. The pH increases during heating. probably dUe to charge change and/or hydrogen bonding within the myofibrillar proteins. Heating causes inactivation of most muscle enzymes. Unfolding of the peptide chains caused by heat denaturation lowers cation binding power of muscle proteins. (Paul and Palmer. 1972; Hamm. 1966). Hamm and Hofmann (1965) found that heating up to 70°C for 30 minutes causes an unfolding of actomyosin molecules to expose sulfhydryl groups. l6 At higher temperatures (l20°C for 30 min.). an oxidation of sulfhydryl to disulfide groups occurs. During heating at 120° for longer periods of time. hydrogen sulfide is formed and this originates from the free or easily reacting sulfhydryl groups of actomyosin. It is conceivable that the quantity of mercury present in foods can be altered by changes which occur in the food during cooking or processing. Very few attempts have been made to examine the effect of cooking upon the mercury content of foods and the few findings that have been reported are conflicting. Furutani and Osajima (as quoted in Smart and Hill. 1968) found no reduction in the mercury content of rice due to cooking. Noren and “WestUU (l967) submitted what appear to be single samples of pike to boiling and frying treatments. The methyl mercury content of both samples before treatment was 0.52 mg Hg/kg of fish muscle. After cooking treatments. the methyl mercury content was 0.54 mg/kg. when corrections were made for weight loss of sample during cooking. It was concluded that cooking does not remove methyl mercury. Another study in which pike were boiled or fried was conducted by Jegrelius (l97l). The uncooked fish sample for the frying experiment contained 7.3 ng Hg/mg of dry fish. Mercury concentration calculated tn~ dry weight was based on 86-88% water content. determined by Karl-Fischer analysis. After frying in margarine for l5 minutes. the sample contained 6.8 ng Hg/mg of dry fish based on a 45-50% water content. The uncooked sample for the boiling experiment contained 10.8 ng Hg/mg of dry fish based on a water content of 86-88%. After simmering for l5 minutes. this sample contained 4.3 ng Hg/mg of dry fish based on 75-77% water content. According to these results. the mercury level in pike samples was reduced 17 by 7% after frying and by 60% after boiling. The author postulated that mercury may have been soluted in the boiling water and then partly or entirely vaporized. During preparation of a fish meal for experimental feeding. Tokuomi (1969) boiled shell-fish at a temperature of 60-700C while in the shell. The shell-fish were then removed. dried and ground into meal. This product was fed to cats and mice. Prior to cooking it can be fairly certain that the shell-fish contained extremely high levels of methyl mercury as they were collected from Minamata Bay. Japan. which was highly contaminated with methyl mercury. A diet with 30 g of the meal per day induced symptoms of methyl mercury poisoning in cats after 22-40 days while a diet with 15 g/day produced symptoms after 75-150 days. Mice began to show visible symptoms of poisoning after consuming daily 9.5 g of the meal for about 30 days. These results indicate that the cooking treatment did not reduce mercury in the shell-fish to non-toxic levels. Mercury Analysis Using Atomic Absorption One of the most popular and widely used methods of determining mercury in biological materials is by atomic absorption spectrophoto- metry preceded by a wet digestion of the material to be analyzed. This method is claimed to be sensitive. rapid and reproducible for determina- tion of mercury down to 1.0 ppb (Uthe at 11.. 1970; Hatch and Ott. 1968). Digestion Determination of mercury in biological materials by atomic absorp- tion requires that the sample be oxidized by a wet digestion procedure 18 prior to analysis. A wide variety of oxidizing agents have been used by different experimenters for digestion of biological samples. No systematic evaluation of the relative merits of the various agents was found. The most commonly used oxidizing agents include: HN03. H2504. KMn04. HCl04. and H202. Munns and Holland (1971) used HN03. H2504. and HClO4 to digest fish tissue. .Uthe $3.31, (1970) used H2504 and ano4. Jeff s and Elkins (1970) used HN03. H2S04. and ch4 while Malaiyandi and Barrette (1970) suggest st04 and HN03 with a vanadium pentoxide catalyst. Smart and Hill (1969) commented that HClO4 yields a more complete oxidation than HNO3 and H2504 alone. Hydrogen per- oxide was not used because of long digestion time. Ward and McHugh (1964) used HN03. H 504. HClD4 and H 0 A Report by the Joint Mercury 2 2 2 Residues Panel (1961) recommended use of HN03.H2504 and H202 for tissue digestion. Most procedures recommend the use of heat to accelerate digestion and recoveries of 83.5-100% have been reported with this method (Munns and Holland. 1971; Jeffus and Elkins. 1970; Uthe g§_al,. 1970). However. due to its high volatility1 . the possibility of methyl mercury loss from heating is a problem inherent in this method (Johnson. 1965). Cold digestion avoids the risk of mercury loss from volatilization but a much longer digestion time is required. This is an advantage if storage of samples for several weeks is desired since they are stable. In contrast. hot digestions must be analyzed within hours of completion. Fatty material may not be completely digested without heat but only a negligible portiOn of the mercury is present in the fat (Johnson. 1965; Smart and Lloyd. 1963). 1The saturated vapor concentration of methyl mercury chloride at 20° C is 94 mg per cubic meter of air compared with 14 mg per cubic meter of air for metallic mercury (Klein and Herman. 1971). 19 Analysis The classical procedure for flameless atomic absorption analysis of mercury is that of Match and 0tt (1968). The sample to be analyzed must be completely oxidized so that the mercury is present in the mercuric form. Sodium chloride-hydroxylamine sulfate solution is then added to reduce excess oxidizing agents. Mercury is aerated from the solution by reduction to metallic mercury by stannous sulfate. This vapor is circulated in a closed system through the absorption cell of the atomic absorption spectrophotometer. Absorption of the 253.7 nm radiation is quantitatively detected. Interference Hatch and 0tt (1968) stated that the presence of large amounts of elements which are easily reduced can prevent the complete aeration of mercury. Substances which may interfere with the evaporation of mercury include noble metals (i.e. gold and platinum). tellurium. bromides. thiosulphate. and perhaps selenium (Lindstedt. 1970). Iron and sulphur compounds may interfere in some methods of analysis (Norén and WestUU. 1967; Fishman. 1970). Munns and Holland (1971) reported that testing involving aliquot size of digested fish showed that a component in the fish was responsible for lowering recovery. Aliquots of 25. 50 and 100 ml of fish digest were adjusted to 100 ml volume and identically spiked with methyl mercury chloride. Results of analysis yielded standard curves with progressively lower slopes from aliquots of 25. 50 and 100 ml respectively of the fish digest. EXPERIMENTAL PROCEDURE This study investigated the effects of moist and dry heat cooking upon the methyl mercury content of duck breast tissue. McGraw-Mallard breasts were either braised or roasted and mercury contents were compared before and after cooking. In addition. selected raw organs were examined for relative mercury concentration. The Sample The Department of Fisheries and Wildlife. Michigan State University. provided the 50 adult (age approx. 1.year) McGraw-Mallard ducks (Ang§_ platyrhynchos) used in this experiment from their standard flock. These birds had been raised on the floor and had been fed a commercial game bird chowa of the following composition: 319.0% crude protein. 32.0% crude fat. 512.0% crude fibre. Use of this feed was continued throughout the mercury study. Ducks were randomly divided into five groups with 4 females and 6 males in each. Group I constituted the control group which received no mercury. Birds in this group were taken directly from the floor and sacrificed immediately. Groups II. III. IV and V each received a particu- lar level of methyl mercury chloride (MeHg). At the onset of the experiment. ducks were placed in 31 x 48 x 14 inch wire cages which were divided in half allowing 2 birds per half cage. Free access to water and feed was provided and cages were cleaned every other day. Weight of each bird was.recorded approximately weekly. All birds experienced an initial weight loss following caging. To avoid a“Purina Game Bird Chow" Ralston Purina Co.. St. Louis. Missouri 20 21 abnormal metabolism of mercury which might be characteristic of a fasting animal. caged ducks were allowed to acclimate for a period of approximately 4 weeks. Mercury was not administered until the average group weight stabilized or showed an increase over a 7 day period. Mercury Administration At this time single doses of methyl mercury chloride (ACS Reagent Grade) were administered orally at levels of 2.0. 5.0. 6.0. 10.0 mg/kg to Groups II. III. IV. V. respectively. Calculated doses of methyl mercury were weighed to the nearest 0.1 mg into #3 gelatin capsules with a lactose filler and refrigerated overnight. Capsules were then placed into the crop and swallowing was assisted by water when needed. Oral administration in capsules was used to simulate normal entrance to body tissue via the gastrointestinal tract and to assure total dose ingestion. Swensson and Fulfvarson (1969) found that. following a single dose of methyl mercury. maximum equilibrium between blood and muscle occurred after two weeks. Therefore. ducks were sacrificed 14 days after capsule administration with the exception of Group III. A preliminary examination of blood levels indicated a low concen- tration of mercury. To assure adequate levels for subsequent studies Group III was given an additional 14 mg MeHg/bird via treated feed. Game bird chow pellets were treated with methyl mercury as follows: Heigh 1000 g of pellets into a wide mouth jar. Place 0.1334 9 methyl mercury chloride dissolved in 20 ml acetone in an atomizer. Spray feed pellets with this solution. shaking and turning bottle to attain even distribution of the mercury. When atomizer is depleted. add 5 ml acetone and spray on feed. Allow all acetone to evaporate from feed. 22 Then bag 15 g quantities in polyethylene so that each bag contains 2 mg methyl mercury. Store in freezer. Administration of treated feed to Group III was begun 14 days after the 5 mg MeHg/kg capsule treatment and feeding was continued for 7 days. The schedule for feeding ducks treated feed was as follows: At 9:00 a.m. water pans were removed and feed pans emptied of untreated feed. Then 15 g of treated feed containing 2 mg MeHg was placed in the pan. In the early afternoon. ducks that had consumed all of the treated feed were given water and untreated feed. If some treated feed remained. water only was given until all treated feed was consumed at which time untreated feed was provided. At 9:00 p.m. feed pans were removed leaving water only. Methyl mercury consumed in the form of treated feed was approxi- 'mately 14 mg per bird and the total ingested by each Group 111 bird was estimated to be 19 mg MeHg. Ducks in this group were sacrificed seven days after treated feed was discontinued. Processing_ Ducks were transported in poultry crates for less than 30 minutes immediately prior to slaughter. Following decapitation the skinless breast was removed and separated into left and right portions. looking from the anterior. The sternum remained in the left portion. Breasts were rinsed in tap water. blotted dry with paper towels and then heat sealed in polyester film pouchesa. Pouches were packed in crushed ice and then frozen at -23°C until needed for subsequent preparation and analysis. Also. packaged and stored according to the above procedures were the heart. liver. kidneys. entire head (brain) and a sample of a "Scotch pak". Kapak Industries. Inc.. St. Louis. Minnesota 23 thigh tissue. No kidney was obtained from ducks in Group 11. Cooking Procedures The ducks of similar weight at slaughter within each group were paired. One bird from each pair was selected for dry heat treatment and the other received the moist heat treatment. Gender of the birds was taken into consideration in this pairing process so that each treatment subgroup contained 2 females and 3 males. Frozen breasts were thawed at 2°C overnight (approximately 16 hrs) before cooking. Tissue from the right side of the breast. which had been designated as the raw sample. was removed from the bone and macerated in a Waring Blender for 20 seconds at high speed. The sample was then wrapped in aluminum foil and refrigerated at 2°C (24 hrs maximum) until mercury analyses. Braising A Temco hot plate. model HP-25158. was preheated at 200 :_5°C (450 setting) for 15 minutes. All weights in cooking procedures were taken on a Mettler top loading balance to the nearest 0.1 g. Weights were determined for the uncooked left breast with bone intact and for the 6" teflon II coated. heavy aluminum frying pan and lid. One gram from a common lot of hydrogenated vegetable shorteninga was placed in the frying pan which was then preheated for 5 minutes. The lid was removed throughout the browning process. The breast was browned at 200 :_5°C for 3 min on each side. the meat side being browned first. Then the temperature of the hotplate was reduced to a "Crisco" - Procter and Gamble. Cinncinati. Ohio. 24 90 :_5°C (250 setting) for braising. The meat side of the breast was left facing upward and the pan was covered throughout braising. Braising time for all samples was 30 minutes. the time determined as necessary to attain an internal temperature of not less than 90°C. To record the time- temperature relationship during cooking. an iron constantan thermocouple was inserted horizontally into the center of the pectoralis major muscle through a specially designed thermocouple holder in the side of the frying pan. The cooked sample was then weighed as was the pan with lid and drip. Drip was extracted from the pan with 20 ml of distilled H20 (with the exception of drip from Groups I and II which were extracted in ethyl ether which was then evaporated under a hood for a minimum of 2 hours). Drip samples were frozen at -23°C in sample jars until used in analysis procedures. The cooked tissue samples were cut from the bone and macerated in a Waring Blender for 30 seconds on low speed. Macerated samples were then wrapped in aluminum foil and refrigerated (2°C for maximum of 24 hours) until.used in mercury analysis procedures. Weight of bone only was determined for each sample. Roasting A Hotpoint deck oven. model HJ225. equipped with a Honeywell Versatronic controller. was preheated to and maintained at 177 :_5°C with grids set on medium and the damper half closed. A 9" x 9" teflon cake pan with an aluminum rack was used as the roasting pan. Weights of the pan plus rack and of the duck breast were determined prior to cooking. The breast was positioned with the meat side up on the rack in the pan. Each sample was roasted to an internal temperature of 90°C. as measured 'by a potentiometer lead which had been inserted into the center of the 25 pectoralis major muscle. After cooking. weighing and preparation procedures were followed-as for braised sample. Cooking_Losses Total. drip. and volatile cooking losses were calculated for each cooked sample and converted to percentage based on the raw sample weight according to the method of Funk‘gt;al, (1966). Analysis’Procedures Moisture Determination Moisture content of duplicate. homogenized cooked tissue. uncooked tissue. and organ samples was determined by drying in vacuo at 90°C for 6 hoUrs and under 27 inches of mercury (AOAC. 1965). Percentage solids was caltulated. Het Digestion ’ A cold. acid digestion procedure was used because it allowed digested samples to be stored for extended periods of time before atomic absorption analysis. A cold digestion should also minimize the risk of volatiliZation of mercury during the commonly used hot digestion pro- cedures. ' ‘ Frozen organ samples were thawed overnight at 2°C. Approximately 1 g of tissue was weighed into a tared 50 ml flask and the weight was recorded to the nearest 0.0001 g. The flask was then packed in ice and '10 ml of concentrated H2304 was added after which the mixture was allowed to stand overnight. After repacking flasks in ice. 5 ml of 30% H202 was slowly added. The mixture was allowed to stand until the digestion mixture became visibly free of undigested particles. At this time the flask was covered with Parafilm and stored at room temperature until 26 analyzed. Through preliminary investigation it was found that digestion was complete after three weeks so samples were analyzed as soon as possible after that time. Although all reagents were from a common lot. a reagent blank was included with every group of diges- tions. Duplicate digestions were prepared from all cooked and uncooked breast samples and from organs when there was sufficient sample. It was necessary to use different oxidizing agents for drip samples due to extreme effervescence upon addition of H202 which resulted in sample loss. The cooking drip sample was dried over P205 to obtain the dry sample weight. Digestion then proceeded as for tissue with the substitution of 3-5 ml concentrated HNO3 in place of 5 ml H202. A storage period of more than 4 weeks was required for drip samples in order to reduce foaming of sample during analysis. Analysis All samples were analyzed by flameless atomic absorption spectro- photometry with a Coleman Mercury Analyzer. model MAS-50. having a sensitivity equal to or better than 0.01 pg of mercury. Analysis procedures were essentially those outlined by Hatch and Ott (1968) and operational directions accompanying the instrument. The reagents and standard mercury solutions used in analytical procedures were as follows: Reagents - (Zabik. 1970) - saturated aqueous solution of potassium permanganate - sodium chloride - hydroxylamine hydrochloride solution. (60 ml of 25% (w/v) hydroxylamine MCI and 50 ml of 30% (w/v) sodium chloride diluted to 500 ml with water). 27 - 10% stannous chloride solution (dissolve 20 g of stannous chloride. 2H 0 in 40 ml of warm concentrated hydrochloric acid. Hhen i6 solution add 160 ml of water). Prepare fresh daily. Standard mercury solutions - (Thorpe. 1971) - stock solution (1 mg Hg/ml or 100 ppm H ) 'Weigh 0.6768 g mercuric chloride IACS Reagent Grade) into a 500 ml volumetric flask and bring to volume. - standards Pipette 10 ml of stock solution into a 1000 ml volu- metric flask and bring to volume. Take 1. 5. 10. 20 ml aliquots of this solution and dilute to 100 ml to yield 0.1. 0.5. 1.0. 2.0 ppm Hg standards. respectively. A complete set of standards was run before and after every 10-15 samples. The procedure was identical to that for samples except that standards were acidified with 5 ml of 50% H2S04 prior to adding KMnO4. The values for the two standard curves thus derived were used to calcu- late a least squares regression line which was used to correct the sample values to pg mercury. Before analysis the total volume of the digestion was recorded to the nearest 0.1 ml. This procedure incidentally served to homogenize the sample before a 1-3 ml aliquot was transferred to a 800 flask containing less than 90 ml deionized H20. .This solution was titrated to a permanent pink endpoint with potassium permanganate. The solution was then reduced with 5 ml of sodium chloride-hydroxylamine hydro- chloride solution. The total volume was brought to 95 ml with deionized H20 and the closed aeration system was connected. Five ml of stannous chloride solution was injected into the BOD flask and maximum absurp- tion Was recorded. A second aliquot from each digestion was subsequently Nil. 28 Mercury in parts per million (ppm) was calculated for each sample on a wet weight basis by the following formula: total volume ml analyzed = ppm Hg Sample Weight Mercury content on a dry weight basis was calculated by dividing ppm corrected pg Hg X mercury in wet tissue by the percentage solids. The total micrograms of mercury in cooked and uncooked breast and in organs was calculated by multiplying the wet weight (g) of the entire sample by the ppm of mercury on a wet weight basis. For drip. total micrograms of mercury was calculated from the total dried weight times ppm of mercury on a dry weight basis. The total percentage of mercury recovered after cooking was calcu- lated as the difference between the total micrograms of mercury in the uncooked breast and the total micrograms of mercury in the correspond- ing roasted or braised breast plus its cooking drip divided by the value for the uncooked tissue. Percentage of mercury recovered in drip was determined by dividing the total micrograms of mercury in the drip by that in the corresponding uncooked breast tissue and percentage recovery in the cooked breast tissue is the difference between the total percentage of mercury recovered and the percentage recovered in drip. All glassware used for mercury procedures was carefully washed and then thoroughly rinsed first with tap water and then with distilled and deionized water. Acetone was not used due to its high absorbance in the ultraviolet range. Pipettes were rinsed with water. soaked overnight in 'acid-chromate solution and then rinsed for more than 3 hours in hot tap water. 29 Analysis of the Data The cooking data were statistically analyzed for variance due to cooking method. animal. and mercury dosage level. Duncan's multiple range test (Duncan. 1957) was used to point out sources of significant differences. Organ data analysis included one-way and two-way analyses of variance due to mercury feeding level and organ mercury content. Duncan's multiple range test (Duncan. 1957) was used. correlation coefficients were calculated and least squares regression lines were derived for each organ. Time-Temperature Relationships Internal temperature of breast was recorded throughout the cooking period and averaged at three minute intervals. These data were plotted and examined for differences due to cooking method. RESULTS AND DISCUSSION Fifty McGraw-Mallard ducks were given dosages of methyl mercury at five different levels. The left side of each breast of half of the birds was braised while that of the remaining birds was roasted. Raw right sides of all breasts. cooked breast tissue and cooking drip were analyzed for total mercury. The concentration of mercury in cooked tissue and in drip was compared to that in the raw sample from the same bird to examine the fate of mercury after cooking. The effects of cooking methods. dosage level and sex were also evaluated. Three ducks were selected from each of the five groups for analyses of the concentration of mercury in raw liver. kidney. breast. thigh. heart. and brain. Relative mercury concentrations among organs were examined as was the relationship between dosage level and organ concentration of mercury. Physical Parameters of Cooking Time-Temperature Relationships Breast samples which received moist heat treatment were browned in one gram of hydrogenated shortening at 200°C for 6 minutes and then braised in a covered pan at 90°C for 30 minutes. It can be seen from time-temperature averaged from all braised samples (Figure 1) that there was a rapid increase in internal temperature during browning and the first 9 minutes of braising. A maximum internal temperature of 95°C was reached after approximately 21 minutes of total cooking time; after which the internal temperature gradually decreased during the remainder of the cooking period. The final internal temperature averaged 82°C. 30 3i .m_m>eoucw opacHE m um commeo>m mo_aEmm ammocn xuav vomfimen tcm noummoL Lo; mafischAHm.oe ocsumeoaeounoefih ._ oeamau Amopscflev mafia mcfixooo mm mm on hm om HN ma ma NH 0 m m o I .. u 2.; cm a. ....09.9 m .2xcv e ..00. —_I_. / 0%. em a .m e .8 m 0 9mm w... n .8 a )o 633m .2: w 32 Dry heat treatment of breast samples consisted of oven roasting of the uncovered sample at 177°C until an internal temperature of 90°C was reached. Time-temperature relationships for all samples show a relatively linear increase in internal temperature throughout the cooking period (Figure 1). Total cooking time averaged 36 minutes but individual sample times varied considerably ranging from 24 minutes up to 57 minutes. Variation was probably mainly due to differences in sample size. which ranged from 70.39 to 127.89. although variation in the temperature of samples at the beginning of cooking was also. undoubtedly. a factor. Internal temperature of samples could have been equalized before cooking was begun to minimize this factor; however. a constant thawing time was used in this study. Cooking_Losses Weight of the breast meat was reduced in all samples during cooking. These cooking losses were quantitated as percentages in categories of total. volatile and drip losses which are reported in Table 2. The influence of cooking method. mercury dosage level and sex of bird upon total. drip and volatile cooking losses from duck breast tissue was analyzed statistically (Table 3). These analyses revealed that the two different cooking methods yielded significantly different percentage drip and total cooking losses (P 5_ 0.001). Average total cooking losses were higher for braised samples (25.42%) than for roasted samples (19.83%). This difference is accounted for by the considerably higher percentage of braised drip (7.39%) as compared with roasted drip (1.04%). Similar results have been reported in the literature indicating that methods which cook meat in a moist atmosphere usually increase the losses (Paul and Palmer. 1972). Volatile cooking losses were not significantly 33 Table 2. Averagea percentages and standard deviations of total. drip and volatile cooking losses of duck breasts cooked by two methods. Cooking {Sizing Percentage of Coeking Losses Method (mgMeHg/kg) Total Drip Volatile Roast 0,0 17.36 :_6.88 0.69 :_O.38 16.67 1.6.66 2.0 20.21 :_3.49 1.44 :_O.80 18.77 :_4.11 6.0 24.22 :_5.13 1.03 :_0.69 23.19 i 5.37 10.0 16.95 1,6.17 0.91 :_0.46 16.04 :_5.86 19.0 20.39 :_7.04 1.14 :_0.43 19.26 :_7.00 (Method Average) (19.83) (1.04) (18.79) Braise 0.0 25.18 :_1.61 7.64 i_3.20 17.54 :_4.53 2.0 27.07 :_2.28 6.55 :_3.83 20.52 :_5.88 6.0 24.90 :_5.40 7.43 :_3.11 17.48 :_5.79 10.0 24.26 :_3.05 7.68 :_3.41 16.68 :_2.23 19.0 25.61 :_2.34 7.67 :_1.83 17.95 i 3.17 (Method Average) (25.42) (7.39) (18.03) aAverage of 5 birds. 3 males and 2 females. for each cooking treatment in each group. 34 Table 3. Analyses of variance for total. drip and volatile cooking losses among breast samples due to cooking method. mercury dosage level and sex of bird. QA 4 fl Mean Square Source of Variance Degrees of Cooking Losses Freedom Total DFip Volatile Total 49 Cooking Method 1 391.66*** 504.03*** 7.05 Mercury Level 4 26.65 0.23 27.96 Sex 1 1.38 8.83 3.24 Cooking Method- 4 21.33 1.31 22.38 Mercury Level Cooking Method-Sex 1 30.61 5.99 63.67 Error 38 18.31 3.92 21.57 ***Significantly different at the 0.1% level of probability. 35 different. Neither mercury dosage level nor sex of birds significantly influenced any of these cooking losses. Effects of Cooking Upon Mercury Concentration The concentration of mercury in cooked and uncooked breast tissue samples and in dried cooking drip was determined by atomic absorption spectrophotometry. Parts per million of mercury on a dry weight basis was calculated and is reported by group average in Table 4. Percentage solids for breasts in braised groups averaged 26.99% for the uncooked samples and 35.38% for the cooked samples. Breasts in roasted groups had an average percentage solids of 26.69% for uncooked samples and 34.07% for cooked samples. These values reflect the higher cooking losses found after braising than after roasting which resulted in a . greater weight loss during cooking for braised breasts. During analyses of tissues from control ducks a phenomenon similar to that reported by Munns and Holland (1971) was observed. Often. an absorbance was recorded which was below that of the deionized water blank. Since it is not possible for the sample to contain less mercury than the blank. the interference of some component in the tissue digest seems to be indicated. For the purposes of this study. all samples which had an absorbance equal to or less than that of the water blank were designated as having a negligible quantity of mercury which was below the detectability threshold for the analysis techniques used. Mercury readings from all cooked and uncooked breast tissue and drip from control ducks fell in this category. 36 Table 4. Averagea ppm of mercury on a dry weight basis. before and after 'cooking. in duck breast tissue and in cooking drip. A A ‘ ppm of Mercury (Dry Weight) . Dosage Cooking Level _ Treatment. mgMeHg/kg Group No. Uncooked Cooked ‘ Drip....q Braise 0.0 I n.db n.d n.d Roast 0.0 (control) n.d n.d n.d Braise 2.0 II 6.00 :_O.65 5.08 :_1.22 0.86 :_0.21 Roast 6.35 :_1.89 7.68 1‘2.05 1.94 :_0.72 Braise 6.0 IV 11.31 :_1.88 8.91 1 1.70 1.73 1 0.32 Roast 11.42 :_2.07 9.21 :_1.67 5.81 :_2.62 Braise 10.0 V 16.34 :_6.34 15.59 1.6.39 2.52 :_0.80 Roast 15.71 :_4.29 14.47 :_4.55 3.93 :_2.35 Braise 19.0 III 31.96 :_7.51 31.77 :_7.71 4.71 :_0.89 Roast 33.66 :_9.70 35.23 :_10.98 11.93 :_3.41 aAverage plus standard deviation of 5 birds. 3 males and 2 females. for each cooking treatment in each group. b Mercury concentration below detectable level. 37 For both cooked and uncooked tissue from ducks given methyl mercury at dosage levels ranging from 2.0 to 19.0 mgMeHg/kg. the ppm of mercury increased directly with methyl mercury dosage level. The concentration of mercury in drip generally followed this trend of increasing mercury concentration with increased dosage level with one exception. The average mercury in drip from roasted breast of ducks given 6 mgMeHg/kg was 5.81 ppm and that for drip from roasted breast of ducks given 10 mgMeHg/kg was 3.93 ppm. However. one sample in the latter group had a very low mercury concentration of 0.5 ppm which lowered the average for that group. The trend of increasing concen- tration of mercury in samples corresponding to increasing dosage levels was significant at the 0.1% level of probability (Table 5). A signifi- cant interaction was found between cooking method and mercury level for drip which was probably due to the low average ppm of mercury in roast drip from dosage level 10 mgMeHg/kg. Average ppm of mercury were lower for cooked breast than for uncooked breast in all braised breast groups and for roasted breasts in groups IV and V. Roasted breast from group II and III birds averaged mercury concentrations which were slightly higher after cooking than before cooking (Table 4). However. these differences in mercury concen- tration before and after cooking were not statistically significant (Table 5). By comparing ppm of mercury in braised drip with that in roasted drip. it may be seen that the ppm of mercury in drip from roasting was consistently higher than that from braising (Table 4). These differ- ences in the concentration of mercury in drip from the different cooking treatments were significant at the 0.1% level of probability (Table 5). 38 Table 5. Analyses of variance of ppm of mercury in raw and cooked duck breast tissue and in drip attributable to cooking method. mercury dosage level and sex of bird. ‘_ A .A Mean Square Source of Deggees Sample State ~ Variance Freedom Uncooked Cooked Drip) Total 49 Cook Method 1 1.09 10.49 94.12*** Sex 1 4.37 ' 5.71 1.36 Mercury Level 4 1549.07*** 1641.96#** 100.79*** Cooking Method- 4 1.86 8.20 20.94*** Mercury Level Cooking Method- 1 17.18 13.15 0.42 Sex Error 38 23.12 26.20 2.70 ***significantly different at the 0.1% level of PTOPBPIIItY' 39 An analysis of variance combining cooked. uncooked and drip statistics (Table 6) indicated a significant difference among ppm of mercury for these different states of the samples (P §_0.001). Duncan's multiple range test (1957) revealed that there were no signifi- cant differences between the average concentration of mercury in cooked tissue but that the average concentration of mercury in drip was significantly less than in cooked and uncooked tissue (P §_0.001). Duncan's multiple range test also showed that the concentration of mercury in Group III tissues and drip was significantly greater than in Group V which was greater than in Group IV which was greater than in Group II which was greater than in Group I (P §_0.001). These differ- ences directly reflect the quantities of mercury administered to the ducks. No significant differences in tissue levels of mercury attribut- able to sex of the birds was found when analyses of variance were calculated from ppm of mercury in the tissues (Table 5). However. when the mercury content was calculated as total micrograms per sample (Table 7). the analyses of variance (Table 6) showed a significant difference in total mercury attributable to sex (P §_0.001). Since total mercury in tissues is calculated by multiplying ppm of mercury by the sample weight. the fact that female birds tended to be smaller than males was reflected in the statistical analysis. The average uncooked breast weight for females was 94.09 while the average for males was 105.29. All other results of the analysis of variance for total microgram data showed similar trends to that which has been discussed for the mercury contents expressed as ppm on a dry weight basis. 40 Table 6. Combined analyses of variance of ppm of mercury and of total ug of mercury in raw and cooked duck breast tissue and in drip attribut- able to cooking method. mercury dosage level. sex of bird and state (i.e. cooked. uncooked or drip). A Mean Squares Degrees Source of Variance , of ppm Hg Total‘ugHg . . Freedom Total 149' State 2 1555.98*** 2.107.637.86*** Cooking Method 1 34.79 296.80 Sex 1 10.63 110.891.22*** Mercury Level 4 2678.15*** 1.618.981.42*** State - Cooking Method 2 35.46 1,216.48 State - Sex 2 0.41 27,458.53 State - Mercury Level 8 613.68*** 399,085.69*** Cooking Method - Sex 1 54.04 10,888.93 Cooking Method - 4 29.59 3,013.65 Mercury Level Error 124 16.50 15,979.60 ***Significantly different at the 0.1% level of probability- 41 The total micrograms of mercury in the entire tissue or drip sample (Table 7) was used to calculate percentage of mercury recovered after cooking (Figure 2). Since none of the concentrations of mercury in tissues before and after cooking treatment were significantly different (Tables 5 and 6) the variations in mercury recovery may only be considered as trends. In all groups. except Group V. the percent- .age of mercury recovered was less when tissues were braised than when they were roasted. Jegrelius (1971) also found greater mercury loss with moist heat cooking than with dry heat cooking. He reported a 7% reduction in the mercury level of Pike after frying compared to a 60% reduction after boiling. A higher internal temperature was reached sooner and maintained longer for braised duck breasts than for roasted samples (Figure 1). Both of these conditions are known to contribute to greater cooking losses (Paul and Palmer. 1972) and they ‘may have contributed to greater mercury losses as well. With the exception of Group II samples. the percentage of mercury recovered after cooking by either method increased with increasing tissue load (Figure 2). Cooking seems to have reduced total mercury by a similarly small amount rather than by a proportion of the total Imercury content. Irregularities in the data from Group II. such as the recovery of 114% of the mercury after roasting. may be partially explained by the fact that this group had a very low tissue concentration of mercury and the mercury analysis technique employed in this study was least reliable when determining such low mercury concentrations. Norén and Hest60'(1967) also found that cooking fish with a low mercury concentra- tion (0.52 mgHg/kg) yielded over 100% recovery of the mercury (0.54 mgHg/kg)° Another factor which may have contributed to greater than 100% recovery of mercury in some samples is the possibility that digestion of 42 Table 7. Total micrograms of mercury in duck breast samples before and after cooking and in cooking drip. Dosage Total_ug Mercurya Cooking Level Treatment (mgMeHg/kg) Group No. Uncooked Cooked Drip Cooked + Drip Braise 0.0 I n.db n.d n.d n.d Roast (control) n.d n.d n.d n.d Braise 2.0 11 148.07 C 124.31 1.13 125.44 Roast , 161.97 185.37 0.61 185.98 Braise 6.0 IV 324.90 241.24 2.49 243.73 Roast 338.75 268.08 1.26 269.39 Braise 10.00 V 463.10 424.18 4.06 428.24 RoaSt 430.67 393.32 1.70 395.02 Braise 19.0 III 926.36 891.84 7.22 899.06 Roast 909.37 909.07 5.09 914.16 aTotal ug of mercury are averages of 5 birds. 3 males and 2 females, for each cooking treatment in each group. bMercury concentration below detectable level. 3 h. .oammau 30; cu woemaeou mm mcmean Lo mcgummoL Louwm aweu ucm oammau “women xusu ca noLo>0uoL >L3ucoe ommucouLom .N ocsmau Amx\mmmamav HHH muonw Amx\ommanv > mfiouw Amx\mmmEov >H Quouw Amx\mmmENv HH machm 0.0 a.mm ummom omamum Ammom mmflmum ummom omamum ummom omamnm m6 To m.o To m6 v.0 m6 m.em m.Hm e.Hm a.me m.ee e.eHH m.em 0.3 mime 5mm 53 m.~m m.ooa H em D 3.5 axe: I wanna: ..ON ..o¢ ..om .Iom ..OOH .IONH dtla pue atosnw ut AJaAooea quaoxad 44 uncooked samples was not as complete as digestion of cooked samples. Perhaps there was some structural change in cooking. such as the changes in sulfhydryl groups reported by Hamm and Hofmann (1965) which released more mercury for analysis. The percentage of mercury recovered in drip from the braised breasts averaged 0.8% of the original total mercury in all groups whereas that from roasted ones was 0.4% of the total mercury (Figure 2). It 'may be seen that drip accounts for only very small amounts of the total mercury and consequently the toxicological significance of cooking drip as a source of mercury is small. It was previously mentioned that roast drip had a higher mercury concentration in terms of ppm of mercury than did braised drip (Table 4). The quantity of drip from braising was much greater than drip from roasting so that. even though the concentration of mercury was lower in braised drip. the total micrograms of mercury in the braised drip was greater than that in roasted drip (Table 7). These results probably indicate compositional differences between drip from roasting and braising cooking methods. The types and proportions of extracted materials such as proteins and lipids appear to vary with the cooking method. It appears that more material with which mercury is associated is extracted by roasting. The one gram of hydrogenated shortening which was added for the braising method must also be considered as a component of the braise drip which probably acts as a diluent. Organ Distribution of Mercury Three ducks. with relatively similar breast mercury levels. were selected from each of the five different experimental groups for mercury 45 analysis of other raw organs including heart. brain. liver. thigh and kidney (no kidney sample was obtained from Group 11 birds). Mercury levels in uncooked breast tissue from these birds had been previously determined for the cooking study. Birds selected for organ analysis from Groups I. 11. IV. and V. which had been fed 0.0, 2.0, 6.0. and 10.0 mgMeHg/kg body weight. respectively. were all males. The organs of 2 females and 1 male were analyzed from Group III (the 19.0 mgMeHg/kg dosage level). Total mercury in the selected organs was determined by flame- less atomic absorption. Mean percentage solids and mean ppm of mercury. calculated on a dry weight basis. for these tissues from ducks given 0.0 to 19.0 mgMeHg/kg are presented in Table 8. In all but some very recent literature. the concentration of mercury in biological materials has been expressed as ppm on a wet weight basis. Since moisture content is a variable factor. mercury concentrations determined on a wet weight basis are difficult to compare accurately. For instance. if data in Table 8 were presented as ppm of mercury by wet weight. the magnitude of the difference between the mercury concentration in liver and kidney would be obscured by the much higher percentage solids in liver than in kidney. Due to its very low percentage solids. brain would appear to have a much lower relative concentration of mercury than other tissues on a wet weight basis. Also the difference in mercury concentration between breast and thigh muscle appears to be greater when wet weight values are used than from dry weight values. Statistical analysis included a one-way analysis of variance of all organ data and a two-way analysis of variance which did not include data for organs from dosage level 2.0 mgMeHg/kg due to the absence of 46 Table 8. Content of mercury with standard deviations and percentage, solids in tissues of McGraw-Mallard ducks administered methyl mercury vat five different levels. Meanb ppm Mercury in Tissues (dry weight) at Dose Level (mgMeHg/kg) Organ Averagea .' 1% Solids 19.0 10.0 5.0 .. 2.0 0.0 Kidney 25.93 87.88 43.72' 30.74 - 3.49 ' :_22.55 19.82 :1.38 10.73 Liver 31.99 55.40 33.87 22.53 14.74 n.d.c 115.21 :5.74 14.13 12.57 Breast 27.09 25.95 13.74 11.21 7.93 n.d. :5.17 11.95 :0.25 :1.53 Thigh 25.19 22.52 13.87 8.92 8.19 n.d. 33.01 13.35 10.92 :3 54 Brain 20.81 21.01 11.17 8.19 4.95 n.d. :5.95 11.50 11.41 11.52 Heart 28.52 18.57 11.95 7.92 3.79 n.d. 13.05 $3.78 10.88 11.25 a15 ducks; 3 in each of 5 groups. bAverage content of three birds from each dosage level. cMercury concentration below detectable level. 47 mercury values for kidney in that group. Results of the two-way analysis of variance are reported in Table 9. There were significant differences among mercury levels and among tissues and the interaction between these two variables was significant. The one-way analysis of variance showed very highly significant differences among items and Duncan's multiple range test (1957) was used to further elucidate these differences (Tables 10 and 11). Duncan's multiple range test was based on the one-way analysis of variance since all of the data were included in this analysis. At all dose levels the concentration of mercury in tissue was found to be greatest in kidneys. second highest in liver. with no significant differences among other organs (Table 10). With methyl mercury administered at 19 mg/kg. the concentration of mercury in kidney was greater than the concentration of mercury in liver which was greater than that in other organs (P 5_0.01). At dose level 10 mg/kg. the concentration of mercury in kidney and liver was greater than that in other tissues (P 5_0.01). With methyl mercury given at 6 mg/kg. the concentration of mercury in kidney was greater than in all other tissues (P 5_0.01). These results agree closely with patterns of distribution in cocks after a single injection of methyl mercury hydroxide (Swensson and Ulfvarson. 1968). Other studies in which avians were fed methyl mercury compounds via treated feed reported the concentration of mercury in liver to be higher than that in kidney (Borg gt_al,, 1970; Smart and Lloyd, 1963; Swensson and Ulfvarson. 1969; Tejning. 1967). Smart and Lloyd (1963) reported higher levels of mercury in breast than in thigh tissue which agrees with results of this experiment although these differences were not statistically significant. 48 Table 9. Two-way analysis of variance for the concentration of mercury in selected tissues of ducks given methyl mercury at dosage levels of 0.0. 6.0.‘10.0. 19.0 mgMeHg/kg of body weight. (Source of Variance Degrees of Freedom Mean Square Total 71 Mercury Level 3 3981.46*** Tissue 5 2041.27*** Tissue - Mercury Level 15 471.43*** Error 48 43.49 *** Significant at the 0.1% level of probability. Table 10. Significant differencesa among the mean concentration of mercury in different organs of McGraw-Mallard ducks given MeHg at levels of '0 - 19Amg/kg. Mean ppm Mercury in Tissues (dry weight) Dose Level (mgMeHg/kg) Kidney Liver Breast Thigh Brain Heart 0.0 3.48 0.00 0.00 0.00 0.00 0.00 2.0 - 14.74 7.93 8.19 4.96 3.79 6.0 30.74 22.63 11.21 8.92 8.19 7.92 10.0 43.72 33.87 13.74 13.87 11.17 11.96 19.0 87.88 55.40 26.91 22.62 21.01 18.57 aAny two means not underscored by the same line are significantly different ,.at the 1% level of probability according to Duncan's multiple range test (1957) calculated from a one-way analysis of variance. 49 Table 11. Significant differencesa among the mean concentrations of mercury in each organ of McGraw-Mallard ducks given MeHg at levels of 0 - 19 mg/kg. Mean ppmMercury in Tissues (dry weight) at DoséTLevel Organ 19:0 10.0 (mgMgHg/kg) 0.0 Kidney 87.88 43.72 30.74 - 3.48 . Liver 55.40 33.87 22.63 14.74 0.00 Breast 26.96 13.74 11.21 7.93 0.00 Thigh 22.62 13.87 8.92 8.19 0.00 Brain 21.01 11.17 8.19 4.96 0.00 Heart 18.57 11.96 7.93 3.79 0.00 aAny two means not underscored by the same line are significantly differ- ent at the 1% level of probability according to Duncan's multiple range test (1957) calculated from a one-way analysis of variance. 50 In all organs the concentration of mercury was greatest when 19 mgMeHg/kg was given, followed in order by organs from birds given doses 10.0. 6.0, 2.0. and 0.0 mgMeHg/kg (Table 11). Kidneys from dose level 19.0 were significantly higher in mercury (P 5_0.01) than kidneys from levels 10.0 and 6.0 which were higher than kidneys of control ducks. In liver samples. those from dose level 19.0 were significantly higher (P 5_0.01) in mercury than all others. Mercury levels in breast. brain, heart and thigh were not significantly differ- ent (P 5_ 0.01). Correlation between the dose of methyl mercury given and tissue mercury levels for all tissues analyzed was very high. This relation- ship is illustrated by the regression coefficients and very highly significant correlation coefficients of 3_O.965 (Table 12), and by the least squares regression lines (Figure 3) derived from the regression coefficients. All tissues showed increasing concentration with increased dose. but the regression equation slope was greatest for the kidney data followed by that for liver. The slope of the equation relating breast tissue total mercury levels to methyl mercury adminis- tered was also slightly higher than that for thigh. brain or heart tissue. These ducks were clearly accumulating mercury in the kidneys at a high rate and although liver accumulation was also high, the level of mercury in kidney was consistently higher. Since these findings are not in agreement with the majority of the literature on avian distribution. none of which deals with ducks. perhaps the kidneys are more actively involved in the excretion of methyl mercury in ducks. This hypothesis is based on Ostlund's (1969) finding that mercury levels in liver and 51 Table 12. The relationship between the level of methyl mercury administered and accumulation in tissues. Correlation Regression Coefficients Organ Coeffgcients Slope Intercept Kidney 0.998 4.4065 2.8931 Liver 0.986 2.7278 5.1343 Breast 0.977 1.2785 2.4991 Thigh 0.985 1.0632 3.3523 Brain 0.990 1.0310 1.4506 Heart 0.989 0.9480 1.4448 ppm Mercury in Tissues (dry weight) 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 52 .1 J Kidney Breast /" Thigh Brain 0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 Methyl Mercury Administered gag/kg) Figure 3. Least squares regression lines for the relation- ship between the level of methyl mercury administered and accumulation in tissues. 53 kidney appeared related to mercury excretion activity of those organs. The fact that mercury was found in kidney of control ducks, which were only exposed to background levels of mercury, further suggests a natural detoxification function related to mercury concentration in the kidneys. Further investigations in this area are required before any conclusions can be drawn. SUMMARY AND CONCLUSIONS To investigate the effects of cooking treatment upon concentra- tion of mercury in muscle tissue, the breast tissue of McGraw-Mallard ducks which had been given doses of 0.0, 2.0, 6.0, 10.0 and 19.0 mgMeHg/kg body weight were either roasted or braised. These cooked tissues and their cooking drip were analyzed separately for total mercury by atomic absorption techniques as were comparable uncooked tissue samples. The concentration and location of mercury before and after cooking were determined. Accumulation patterns for mercury in tissues as related to dosage of methyl mercury administered were examined from data of the cooking study as well as for the selected raw organs: liver, kidney. thigh, heart, brain and breast. The braising cooking treatment yielded higher cooking losses and a greater quantity of drip than did the roasting treatment. Braised meat samples contained 86.8% of the mercury found prior to cooking whereas roasted meat samples contain 96.2%. Neither average parts per million of mercury nor total micrograms/ sample, determined for cooked and uncooked tissues in all groups, showed statistically significant differences. Therefore, it must be concluded from these data that braising and roasting as performed in this study are not satisfactory methods of reducing mercury residue levels in muscle tissue. Braising showed a tendency to remove slightly more mercury than roasting so perhaps a more severe moist heat treatment would magnify this difference. However, the magnitude of mercury residue reduction required for toxicological significance probably cannot be expected to result from normal cooking treatments. 54 55 The concentration of mercury in braised drip was lower than in roasted drip but since the quantity of braised drip was greater, more total mercury was found in braised drip than in roasted drip. These results suggest compositional differences between roasted and braised drip, particularly in relation to mercury associated components. There was a statistically significant. direct relationship between the quantity of methyl mercury administered to birds and the concentration of mercury found in uncooked breast tissue. in both braised and roasted breast tissue. and in braised drip. Roasted drip also exhibited this relationship with the exception of samples from the group of ducks given 10.0 mgMeHg/kg. All organs analyzed accumulated increasingly higher levels of mercury which were directly related to increasingly higher intakes of methyl mercury. Correlation coefficients reflecting this relationship were 3_0.965. The rate of increase in mercury concentration was higher for kidney and liver than for all other organs. At all dosage levels the concentration of mercury in kidney was greater than the concentration of mercury in liver and both kidney and liver concentrations were greater than those in other organs. There were no significant differences between mercury concentrations in breast. thigh. heart and brain within the same dosage level although breast was consistently higher in mercury than was thigh. _ It was hypothesized that the high levels of mercury found in kidney and the somewhat lower levels of mercury found in liver were related to detoxification function in ducks. The indications are that kidney is the most actively involved organ in excretion of mercury at the levels administered in this study. RECOMMENDATIONS FOR FURTHER RESEARCH During the course of this study, it became apparent that knowledge is incomplete concerning many aspects of the action of mercury in bio- logical and food systems. Further investigation in the following areas is recommended: 1. Metabolism mechanisms and distribution patterns of methyl mercury in ducks should be more thoroughly investigated to explain. for instance, the role of liver and kidney in detoxification as related to accumulation in these organs. 2. The actual binding mechanism and sites to which methyl mercury is bound should be determined and application of this information could facilitate the removal of mercury from meats for human con- sumption. a. The mechanism of differential binding between red and white muscle could be examined. b. The relationship between methyl mercury in muscle tissues and sulfhydryl groups requires further study. particularly regarding changes which occur at these sites during cooking. 3. If there is actually a tissue component present in the digestion which interferes with atomic absorption analysis of total mercury, this component could be identified and studied. 4. An effective means of removing excessive concentrations of methyl mercury from foods must be sought since environmental contamination cannot immediately be eliminated. a. Boiling Contaminated meat in large quantities of water containing various solutes might be tried as a means of reducing mercury ‘ concentration. 56 57 b. A means of reducing mercury concentration by ion exchange might be devised. The interaction between selenium and methyl mercury which produces reduced toxicity of the latter requires further investigation in view of the possibility of detoxifying mercury contaminated foods with selenium. LITERATURE CITED Arrhenius, E., 1967. Effects of organic mercury compounds on the detoxi- fication mechanism of liver cells in vitro. Oikos 9 (Suppl):32. Association of Official Analytical Chemists, 1965. Meat and meat products - moisture. "Official Methods of Analysis," 10th ed., p. 346. Washington. D. C. Bache. C. A., N. H. Gutenmann, and D. J. Lisk, 1971. Residues of total mercury and methyl-mercuric salts in lake trout as a function of age. Science 172:951. Backstrom. J.. 1969. Distribution studies of mercuric pesticides in quail and some fresh water fishes. Acta Pharmacol. et Toxicol. 27 (Suppl 3):1. Beecher. G. R., L. R. Kastenschmidt, R. G. Cassens, w. G. Hoekstra and E. J. Briskey. 1968. A comparison of the light and dark portions of a striated muscle. J. Food Sci. 33:84. Benesch, R. and R. E. Benesch. 1962. Determination of sulfhydryl groups in proteins. Meth. Biochem. Anal. 10:43. Berg, w.. A. Johnels, B. Sjostrand and T. Westermark. 1966. Mercury goztent in feathers of Swedish birds from the past 100 years. i 05 17:71. Berglund, F., and M. Berlin, 1969. Risks of methyl mercury accumulation in man and mammals and the relationship between body burden of methyl mercury and toxic effects. in "Chemical Fallout, Current Research on Persistent Pesticides." p. 258, M. v. Miller and G. G. Berg. Ed. Charles C. Thomas Publisher, Springfield, Illinois. Berglund. F.. and M. Berlin, G. Birke, R. Cederldf, V. Von Euler. L. Friberg. B. Holmstedt. E. Jonsson. K. tuning. C. Ramel. S. Skerfving. A. Swensson, and S. Tejning, 1971. Methyl mercury in fish. Nord. Hygienisk Tidskrift 4 (Suppl):1. Berlin, M. H.. T. w. Clarkson. L. T. Friberg. J. C. Gage. L. J. Goldwater. A. Jernelov, G. Kazantzis, L. Magos, G. F. Nordberg. E. P. Redford, C. Ramel, S. Skerfving, R. G. Smith, T. Suzuki, A. Swensson, S. Tejning, R. Truhaut. and J. Vestal, 1969. Maximum allowable concentrations of mercury compounds. Arch. Environ. Health 19:891. Borg, K.. H. Hanntorp, K. Erne. and E. Hanko, 1969. Alkyl mercury poisoning in terrestrial Swedish wildlife. Viltrevy 6: 301. Borg. K., K. Erne, E. Hanko. and H. Wanntorp. 1970. Experimental secondary methyl mercury poisoning in the goshawk (Accipiter. G. Gentilis L.). Environ. Pollution 1:91. 58 59 D'Itri, F. M., 1972. "The Environmental Mercury Problem." p. 25,110, Chemical Rubber Co. Press, Cleveland, Ohio. Duncan. 0. B.. 1957. Multiple range tests for correlated and hetero- scedastic means. Biometrics 13:164. Dunlap. L.. 1971. Mercury: anatomy of a pollution problem. Chem. & Eng. News 7/5:22. Dustman. E. H.. L. F. Stickel, and J. 8. Elder, 1970. Mercury in wild animals. Lake St. Clair. 1970 Int. Conf. on Environ. Mercury Contam., Ann Arbor, Michigan (in press). Fimreite. N.. 1970. Effects of methyl mercury treated feed on the mortality and growth of leghorn cockerels. Can. J. Anim. Sci. 50:387. Fimreite. N.. R. H. Fyfe. and J. A. Keith. 1970. Mercury contamination of Canadian Prairie seed-eaters and their avian predators. Can. Field-Natur. 84:269. Fishman. M. J., 1970. Determination of mercury in water. Anal. Chem. 42(12):1462. Funk. K.. P. J. Aldrich, and T. L. Irmiter, 1966. Delayed service cooking of loin cuts of beef. J. Am. Diet. Assoc. 48:210. Furutani, s. and v. Osajima as cited in N. A. Smart and A. Hill, 1968. Pesticide residues in foodstuffs in Great Britain. VI. Mercury residues in rice. J. Sci. Fd. Agric. 19:315. Gage. J. C., 1964. Distribution and excretion of methyl and phenyl mercury salts. Brit. J. Industr. Med. 21:197. Ganther. H. E., M. L. Sunde, C. Goudie, M. J. Kopecky, P. Hagner, Sang-Hwan Oh, and w. G. Hoekstra, 1972. Selenium: relation to decreased toxicity of methylmercury added to diets containing tuna. Science 175:1122. Grolleau, G. and J. Giban, 1966. Toxicity of seed dressings to game birds and theoretical risks of poisoning. J. Appl. Ecol. 3 (Suppl):199. Hamm. R.. 1966. Heating of muscle systems. .in "The Physiology and Biochemistry of Muscle as a Food.” p. 363. E. J. Briskey, R. G. Cassens, and J. C. Troutman. Ed. Univ. of Wisconsin Press, Madison, Nisc. Hamm. R. and K. Hofmann. 1965. Changes in sulphydryl and disulphide ‘ gggupgsgn beef moscle proteins during heating. Nature, London Hannerz, L.. 1968. Experimental investigations on the accumulation of ‘mgrcury in water organisms. Inst. Freshwater Res. Drottningholm :120. 60 Hatch, N. and W. Ott. 1968. Determination of sub-microgram quantities of mercury by atomic absorption spectrophotometry. J. of Biol. Chem. 40(14):2085. Hughes. N. L.. 1957. A Physiochemical rationale for the biological activity of mercury and its compounds. Ann. N.Y. Acad. Sci. 65:454. Jeffus. M. T. and J. 5. Elkins. 1970.. Determination of mercury in biological materials. J. of the AOAC. 53(6):1172. Jegrelius, G.. 1971. Mercury analysis of raw. fried, and boiled pike. Research Note. A B Glykocid Co. Vadstena, Sweden (8/25). Jenson, S. and A. Jerneldv. 1969. Biological methylation of mercury in aquatic organisms. Nature 223:753. » Johnels. A. G.. T. Nestermark. N. Berg, P. I. Persson and B. Sjostrand, 1967. Pike (Esox lucius L.) and some other aquatic organisms in Sweden as indicators of mercury contamination in the environ- ment; Oikos 18:323. . Johnson. N. C.. Chairman. Analytical Methods Committee, the Society for Analytical Chemistry. 1965. The determination of small amounts of mercury in organic matter. Analyst 90:515. Kiwimfle, A.. A. Swensson. u. Ulfvarson. and e. Westdd, 1969. Methylmercury compounds in eggs from hens after oral administra- tion of mercury compounds. J. Agr. Food Chem. 17(5):1014. Klein, I7Zand25. Herman. 1971. Precautions with alkyl mercury. Science :87 . Lawrie, R. A., 1966.’ Metabolic stresses which affect muscle. in "The Physiology and Biochemistry of Muscle as a Food." p. 137, E. J. Briskey, R. G. Cassens, and J. C. Trautman, Ed. Univ. of Wisconsin Press. Madison, Wisc. Lindstedt, G., 1970. ‘A rapid method for the determination of mercury in urine. Analyst 95:264. Lbfroth, G., 1969. Methylmercury. Swedish Natural Science Research Council. Stockholm, Sweden. ' Malaiyandi, M. and J. P. Barrette, 1970. Determination of submicro 3(133151e5 of mercury in biological materials. Anal. Letters : 79. Munns, R. K. and D. C. Holland. 1971. Determination of mercury in fish by flameless atomic absorption: a colloaborative study. J. of the AOAC. 54(1):202. Nelson, N., T. c. Byerly, A. c. Kolbye. L. T. Kurland, R. E. Shapiro, S. I. Shibko, H. H. Stickel, J. F. Thompson. L. A. Van Den Berg. A. Neissler, 1971. Hazards of mercury. Environ. Res. 4:1. 61 Norén, K. and G. Westdd. 1967. Metylkvicksilver i fisk. Var FUda. 2:13. . Ostlund. K., 1969. Studies on the metabolism of methyl mercury and dimethyl mercury in mice. Acta Pharmacol. et Toxicol.. 27 (Suppl 1):1. Passow, H.. A. Rothstein and T. W. Clarkson, 1961. The general pharmacology of the heavy metals. Pharmac. Rev. 13:185. Paul, P. c. and H. H. Palmer. Ed., 1972. "Food Theory and Applications." p.335, John Wiley and Sons, Inc.. New York. Ramel, C., 1967. Genetic effects of mercury compounds. in "The Mercury Problem." Oikos 9 (Suppl):35. Report by the Joint Mercury Residues Panel of the Advisory Committee on Poisonous Substances Used in Agriculture and Food Storage. the Analytical Methods Committee on Poisonous Substances of British Manufacturers of Agricultural Chemicals, 1961. Analyst 86:608. ‘ Smart, N. A., 1968. Use and residues of mercury compounds in agri- culture. Residue Rev. 23:1. Smart, N. A., and A. R. Hill, 1969. Determination of mercury residues in potatoes, grain, and animal tissues using perchloric acid digestion. Analyst 94:143. Smart, N. A., and M. K. Lloyd, 1963. Mercury residues in eggs flesh and livers of hens fed on wheat treated with methylmercury dicyandiamide. J. Sci. Fd. Agric. 14:734. Swensson. A. and U. Ulfvarson. 1968. Distribution and excretion of various mercury compounds after single injections in poultry. Acta Pharmacol. et Toxicol. 26:259. Swensson, A. and U. Ulfvarson, 1969. Investigations on the toxic effects of different mercur compounds on young. White Leghorn cocks. Poultry Sci. 48 (5 :1567. Takeuchi. T.. 1970. Biological reactions and pathological changes of human beings and animals under the condition of organic mercury contamination. 1970 Int. Conf. on Environ. Mercury Contam.. Ann Arbor. Mich. Tejning, 5.. 1967. Biological effects of methyl mercury dicyandiamide ' treated _rain in the domestic fowl (Gallus gallus L.). Oikos ‘ * 8 (Supplg:1. Tejning, 5.. 1968. Kvicksilverhalterna: blodroppor och-i blodplasma hos "normala" modrar och deros nyfodda barn. Report No. 680220. Clinic of Occup. Med. Univ. Hosp., Lund, Sweden. 62 Thorpe, V. A., 1971. Determination of mercury in food products and biological fluids by aeration and flameless atomic absorption spectrophotometry. J. of AOAC. 54(1):206. Tokuomi. H., 1969.. Medical aspects of Minamata Disease. Rev. Intern. Oceanogr. Med. XIII-XIV:5. . Ulfvarson, U.. 1962. Distribution and excretion of some mercury com- pounds after long-time exposure. Int. Arch. Gewerbepathol. 19:412. Uthe. J. F., F. A. Armstrong, and M. P. Stainton, 1970. Mercury determination in fish samples by wet digestion and flameless atomic absorption spectrophotometry. J. of Fisheries Res. Brd. of Can. .27(4):80. . Ward, F. N. and J. B. McHugh, 1964. Prof. Paper U.S. Geol. Surv. 501D:128. Wershaw, R. L.. 1970. Source and behavior of mercury in surface waters. U.S. Geol. Surv. Prof. Paper 173:29. West. J. M., 1969. Mercury. 1968 Minerals Yearbook, U.S. Dept. of the Interior, Bureau of Mines, U.S. Govt. Printing Office, Washington, D. C., L-11:639. Westdb, G.. 1969. Mercury and methylmercury levels in some animal food products. var Fdda 7:137. Westbd, G. and M. Rydalv, 1969. Kvicksilver och metylkvicksilver: fisk och kraftor. var Fdda 21:19. Williston. S. H., 1968. Mercury in the atmosphere. J. Geophys. Res. 73:7051. Wood, J. M., F. S. Kennedy, and C. G. Rosen, 1968. Synthesis of methyl mercury compounds by extracts of methanogenic bacterium. Nature. London 220:173. Youatt, W., and M. J. Zabik. 1971. Mercury in Michigan wildlife. Annual Meeting of Wildlife Disease. Colorado State University, Ft. Collins, Colo. ' Zabik, M. J., 1970. Mercury residue analysis procedure. (unpublished).