Document RaeMDG61j9oD216mqO790629V

Reprinted from BioScience Polychlorinated Biphenyls Another Long-Life Widespread Chemical in the Environment David B. Peakall and Jeffrey L. Lincer The recent finding that pelagic birds dying on the coasts of Great Britian had polychlorinated biphenyls (PCBs) in their livers in concentrations of several hun dred parts per million (Bourne and Mead, 1969) shows that these compounds are present in the ecosystem in large amounts. Thus, it seems worthwhile to summarize and evaluate our current knowledge of these compounds. Structural and Physical Properties The picture is complicated by the fact that we are not dealing with a single com pound. The basic structure of PCBs is shown in Figure I. Any of the positions marked with an x can be substituted by chlorine. Widmark (1968) has calculated that of the 210 possible combinations 102 are probable. His criteria for these lim itations are those compounds containing five to eight chlorine atoms per molecule and the number of chlorine atoms per ring differing by not more than one. The commercially available Aroclors (Monsanto Company Trademark) are des ignated by numbers (Monsanto Technical Bulletins). The first two digits represent the molecular type: 12 - chlorinated bi phenyls: 25 and 44 - blends of chlorinated biphenyls and chlorinated terphenyls (75% biphenyl and 60% biphenyl, re spectively): 54 - chlorinated terphenyls. The last two digits give the weignt per cent of chlorine. Thus, Aroclor 1242 is a chlorinated biphenyl containing 42% chlo rine. The biphenyls commercially avail able from Monsanto range from 21% to 68% chlorine. Mass spectrographic studies of Aroclor 1260 (Kocman et al., 1969a) show the presence of 11 isomers: five con taining six chlorine atoms, five containing seven chlorine atoms, and one containing eight. Bagley et al. (1970), studying Aro clor 1254, found 18 compounds: one con taining three chlorine atoms, four contain ing four chlorines, four containing five chlorines, five containing six chlorines, and four containing seven chlorines. Thus, The airihon arc currently at (He LancmMr Laborator.'. Corarfl Uftivcmty. Ithaca. N Y. 958 the number of compounds present is, for tunately, much smaller than is theoretical ly possible. PCBs are chemically inert, are not hy drolyzed by water, and resist alkalies, acids, and corrosive chemicals. They have low volatility, their boiling points ranging from 278 C for Aroclor 1221 to 415 C for Aroclor 1268 (Penning, 1930). All are stable to prolonged heating at 150 C, and the lower Aroclors can be distilled at atmospheric pressure without appreciable decomposition. PCBs are described as in soluble in water and very soluble in hydro carbon solvents, although no exact figures appear to be available. Nothing is known about the biological decomposition of PCBs, but it is likely that they are more stable than DDT and its metabolites since they lack the ethane component between the aromatic rings, which is the site of action of most of the transformations of DDT. Thus, PCBs have the necessary physical and chemical characteristics for persistence and accumulation up the food chain. Use of PCBs Polychlorinated biphenyls were first de scribed in the literature in 1881 (Schmidt and Shultz, 1881), and the successful com mercial production was achieved by the Swann Company in 1930. In that year the physical characteristics and commcrical possibilities of these materials was de scribed by Penning (1930). PCBs are now manufactured by Monsanto in the United States (Trade name Aroclor), Prodel6e in France (Phenochlo.r), and Bayer in Ger many (Colphen). Other manufacturers are located in Japan and the Soviet Union. No figures of the amount of these mate rials produced annually are available, but to judge from the uses listed in Table I they may be very large. The uses listed in the table are only suggested applications taken from Technical Bulletin 306 of the Monsanto Chemical Company, and again there is no information on the extent to which these recommendations are acted upon by the manufacturers of plastics and resins. The use of highly chlorinated Aroclors for extending the kill-life of formulations containing chlordane. aldrin, and dieldrin are mentioned under miscellaneous appli cations of Aroclors. This possibility has been considered in a number of papers. Sullivan and Hornstein (1953), Homstein and Sullivan (1953), and Tsao et al. (1953) found that the chlorinated terphenyl Aro clor 5460 increased the residual persis tence of lindane. Tsao et al. (1953) consid ered that there was some indication that chlorinated polyphenyl may have a syner gistic effect with lindane. Duda (1957) PC* XXX % oo x%x A O-l-O DOT X POW& P0MT3 FOR SUBSTITUTION OF CM.0MNE (Al r Iff ' U6HT DSW 332338 p- *OOMBN*nON -K a-c-ci "0-0" OKHLOROORCiett. Fig. I. Basic stractarc ol PCBs (A l sad possible reactions of PCBs (B). FREE RADICALS BioScience Vol. 20 No. 17 STLCOPCB4078914 sprayed dm saplings with a lindane Aroclor 5460 mixture (1:4) and then tested the residual effect on the leaves to infes tation with elm leaf beetle (Galervceila xanthomeiaena). He found that .eaves treated with lindane plus Arodor were protected longer and that the insecticidal effect was more rapid than with lindane alone. Peaks corresponding to Aroclor 5460 have not been detected in the envi ronment. However, under conditions of temperature and flow-rate commonly used for gas chromatographic analysis these compounds would not be detected (Reyn olds. 1970). Nearly 25 years ago it was noted that Aroclor 1242 was one of 175 compounds out of 6000 tested that was effective against mosquito (Aegyps) larvae (Deonier et al., 1946). Lichtenstein et ai. (1969) tested the effect of adding a variety of biphenyls and terphenyls to DDT and dieldrin with respect to their toxicity to house and fruit flies. It was found that PCBs had very low toxicity to house flics when given alone but PCBs increased the toxicity of dieldrin and DDT, especially the latter. The effectiveness of these com pounds decreased as the chlorinated level increased. For example, Aroclor 1224 in creased the mortality of fruit flies from 59% to 93%, whereas Aroclor 1268 in creased it to only 77%. Analytical Methods I) Identification. Since various com pounds with electron-capturing properties have been tentatively identified in atmo spheric samples in the past (Abbott et al., 1966), and some pesticides are capable of hybridizing in the soil to form a new com pound (Bartha, 1969), it is important that the presence of PCBs in field samples be proven beyond doubt. Identification by means of a combination of high resolution gas chromatography and mass spectrom etry has been carried out by three inde pendent laboratories in Sweden (Widmark, 1967), Holland (Koeman et al.. 1969a), and the United States (Bagley el al.. 1970). Widmark's report states that all peaks were identified by mass spectrom etry (although no data were given), but Koeman and co-workers gave full experi mental details, and Bagley et ai. (1970) demonstrated that most chemicals in the eagle samples examined were components of Aroclor 1254. Thus, despite a recent statement by the Monsanto Chemical Company (in Risebrough, 1970) that the case for labeling the peaks in question as PCBs was not proven, there is enough evidence to convince an unbaised scienti fic jury. 2) Separation. The chemical techniques preliminary to quantitation of residues in samples containing both PCBs and chlo rinated hydrocarbon pesticides fall into two groups--those necessitating the de struction or alteration of one or more of the compounds, and those which do not. Included in the first group is nitration. Treatment with a 1:1 mixture of sulfuric acid nitric acid at 0 C for 5 min destroys or alters aldrin, p,p'-DDE, p,p'-DDD (TDE), p.p'-DDT, and dieldrin such that they can no longer be delected at the original position on the chromatogram. This procedure leaves unaffected PCB, lindane, and BHC (Jensen and Widmark, 1967). A more rigorous nitration with a 1:1 mixture of sulfuric acid-fuming nitric acid for 15 min at room temperature re moves, in addition to DDT and its related products, aldrin, heptachlor, Kelthane, Perthane, Tedion, Telodrin, and Trithion while lindane, heptachlor epoxide, toxaphene, and Strobane are not removed (Erroet al., 1967). Risebrough et al. (1969) reported that this nitration also removed the chromatographic peaks of PCBs. Rey nolds (1969) reported that his attempts to nitrate samples were not fully successful in that there appeared to be loss of some of the more volatile PCBs while peaks with longer retention times appeared. Armour and Burke (1969) reported that complex chromatograms resulted after nitration which could not be related to the unreacled DDT-PCB mixture, and nitra tion was not pursued as a practical means of separating DDT and PCB for further tests. Saponification with alcoholic NaOH or KOH will dehyrodrochlorinatc Perthane, Toxaphcne, DDD, and DDT to their re spective olefins (Archer and Crosby, 1966; Klein and Watts, 1964). Risebrough et al. (1969) reported that PCB peaks are not removed or displaced but gave no data. The second, and in some instances more desirable, group of analytical tech niques allows the special separation of many chlorinated hydrocarbon pesticides from PCBs. Reynolds (1969, 1970) re ported on an activated Florisil column technique which separated heptachlor, aldrin, DDE. and PCB with the first elu tion (60 ml n-hexane) from lindane: hep tachlor epoxide, DDD, and DDT with the second elution (40 ml 50% ethyl ether in hexane). Armour and Burke (1970) devel oped a method utilizing a silicic acidCelite column eluting aldrin and PCB with the first fraction (250 ml petroleum ether), and lindane, heptachlor, hepta chlor epoxide, dieldrin, endrin, p,p'DDE, o,p'-DDT, p.p'-DDT, and p,p'DDD with the second fraction (200 ml acetonitrile hexane methylene chloride-- 1:19:80). Koeman et al. (1969a), using an activated Florisil column, eluted the apolar compounds including DDE and PCB with hexane, and then dieldrin and endrin with 10% diethyl ether in hexane. Mulhern (1968) reported on a method which utilized silica gel-coated thin-layer plates Il Mi 1244. 1244 4 INI I7-4V >I/ Mk 1221. 1121. 1242 II1V MM 1214 (41V 1221 I 1241 MM fin H4-IIV MM I2M Mi-NV i 1221 CIV i 12W iuru mm-jtm ikM 1244 MV 1244 (1 4V September I. 1970 DSW 332339 959 STLCOPCB4078915 aad a hexane/ethyl etbcr (98.2) rolvent system. The plate* were developed, sprayed with a silver nitrate solution, and exposed to UV light. The plates were then divided into five horizontal sections. Dieldrin, endrin, T-BHC, heptachlor epoxide, p,p'-DDD, p,p'-DDT. o,p'DDT, and p,p'-DDE (in that order) were found in the first four fractions, while most of the interfering compounds found in wildlife samples were found in the fifth zone. Baglcy et al. (1970) used this meth od but mentioned that zones three and four contained practically all unknown components as well as p,p'-DDT (zone three) and p,p'-DDE (zone four). Armour and Burke (1969) used precoated (alu minum oxide) sheets and n-heptane and 2% acetone/n-heptane for solvent sys tems. PCBs (Aroclors 1254 and 1260) and DDE were not separated, but p,p'-DDT and p,p'-DDD were completely sep arated from PCB by both solvent systems. Another possible means of separating in terfering substances is to use a series of differing polarity columns in the gas chro matograph at the time of determination. This is Iks time consuming and may be useful when operating conditions can be selected such that PCB peaks are absent in the region where sought pesticides emerge (Simmons and Tatton, 1967). 3) Quantitation. Koeman et al. (1969a) semiquantitatively measured the residues in Japanese quail fed phenochlor DP6 by using one of the peaks in a phenochlor DP6 mixture as a standard. Risebrough (1969) quantitated relative levels of PCBs by assuming that each PCB compound produced the same peak height w-th the electron capture detector as the same amount by weight of p,p'-DDE. After summing the heights of the individual peaks, the total was multiplied by a factor derived from measurements of standard solutions with electron capture and microcoulometric detectors. Jensen et al. (1969) reported PCB amounts as the sum of all PCB components and based the estimate on a combination of mass spectrometry and microcoulometric and electron cap ture detection. Even with this elaborate approach, these investigators suggest that the method is still rough and may be cor rect only within a factor of 2. Anderson et al. (1969) devised a method for obtaining a crude estimate of PCB residues as Aroclor 1254 from chromatograms where sep aration had not been attempted. On an empirical basis, it was found that Aroclor 1254 could be quantitated by considering peak 10 as p.p'-DDT and multiplying that value by 10. Since this peak had originally been quantitated as p.p'-DDT, and in fact many of the original samples con tained little or no p.p'DDT, it was pos sible to estimate relative PCB values. Anderson and his co-workers also saponi fied samples to remove interfering p.p'DDT and p,p'-TDE and then quantitated PCBs as Aroclor 1254 by relating sample peaks 9 and 10 to the corresponding peaks of an Aroclor 1254 standard. Reynolds (1970) employed a method similar to Koeman et al. (1969a) but basecfhis quan titation on an average of two or more peaks, fn addition, his results were re ported as Aroclor 1254 or 1260 depending on the overall pattern of the chroma tographic peak profile. It is clear that we are still relatively unsophisticated in our PCB quantitation methodology and will continue to estimate only relative amounts of PCBs in field samples until we synthe size the individual PCB components com monly found in the ecosystem and are able to speak in terms of these individual peaks as we do for most pesticide residues. However, it has been kindly pointed out (Risebrough, pers. comm.) that, with ref erence to biological significance, the cor rect order of magnitude and accurate rela tive amounts of PCB give the essential information. This is because biological effects, such as enzyme induction, are re lated to degree of chlorination, and the existing methods give some indication of this activity. Therefore, that information would be lost if stress were placed only upon quantitating individual peaks. 4) Magnitude of Error. Ever since PCB peaks were recognized for what they are, residue chemists and other researchers interested in pesticide residues have asked what magnitude of error is likely to result from ignoring the presence of PCBs. Only recently have studies either directly or indirectly resulted in an esti mate of this error. Anderson et al. (1969) quantitated p,p'-DDE, p,p'-DDD, and p.p'-DDT in five egg samples before and after saponification. There was no appre ciable change in the p,p'-DDE, but ap parent p,p'-DDD was reduced by ap proximately 58% and p.p'-DDT by 90%. Reynolds (1970) looked at a large number of samples before and after his PCBFlorisil separation and found that the actual p,p'-DDD residue (relative to the apparent residue) represented from 0 to 7% in California gull (Larus occidentals) fat, 0% in cormorant (Ehalacrocorax auritta) eggs, 80% in 10 pooled mallard (Anas platyrhynchos) duck eggs, and from 0 to 104% in great blue heron (Ardea cinerea) eggs. Respective values for p.p'-DDT were 0 to 53%, 13 to 41%, 100%, and 18 to 102%. Actual residue levels of heptachlor epoxide generally represented a majority of the apparent values. Before and after values for p,p'-DDE and diddrin were not significantly different. Obviously, the magnitude of error may vary depending on the trophic level sampled and certainly with the area from which a sample is col lected (Risebrough et al., 1968, Jensen et al., 1969). Since the publishing of ade quate separation techniques, there is no reason why there should be any error in pesticide quantitation contributed by PCBs. Taxicalagy Despite some recent studies, the toxi cology of PCBs remains rather poorly known as compared to that of the chlo rinated hydrocarbon pesticides. For ex ample, no definite work has been done to establish LD50 values for the various formulations of PCBs. 1) Acute, single dose experiments. Tucker and Crabtree (1970) found that a single dose of 100 mg/kg Aroclor 1254 (stomach tubed in oil) was fatal to two out of three rats, while 500 mg/kg was not fatal to three rats. Aroclor 1268 killed one rat out of three at 500 mg/kg, 1000 mg/ kg, 2000 mg/kg, and 4000 mg/kg. Smyth (1931) found that a 4 g/kg (degree of chlo rination not stated) was nontoxic to guinea pigs and rabbits, but this appears to be due to the fact that material which was given as a paste passed through the intes tine unabsorbed. Tucker and Crabtree (1970) found that Aroclor 1242, 1254, 1260, and 1268 at a dose of 2000 mg/kg was not fatal to mallard ducks (four groups of three birds each). 2) Acute, feeding experiments. Mon santo Bulletin 306 states that 100 ppm diet had no effect on rats, although no details of the studies were given. The ex periments of Bennett et al. (1938) in which 0.05 g/rat of 65% chlorine biphenyl was given orally every other day led to 50% mortality. If one assumes a body weight of 200 g, then the alternate day dose is roughly 250 mg/kg, which can be com pared to oral LD50 for DDT of 113 mg/kg (Frear, 1968). Miller (1944) found that two oral doses of 69 mg of 42% chlorine biphenyl a week apart were fatal to guinea pigs. At an estimated body weight of 400 g, this gives a dose of 170 mg/kg. Tucker and Crabtree (1970) fed rats diets contain ing 10 and 1000 ppm Aroclor 1254. One rat out of six on the low dose died; this mortality was considered to be due to other causes. Four out of four of the high group died within 53 days and the cal culated intake was 1330-1520 mg/kg. The DSW 332340 OAT\ BioScience Voi ?fl No. 17 STLCOPCB4078916 food intake of the 1000 ppm group was only 79% of the control group. Presst et al. (1970) have examined the toxicity of Aroclor 1254 to Bengalese finches (Lonckura striata). This is a diffi cult species for which to calculate the dietary intake. Due to their dependence upon unshelled food, it is only possible to present the PCB-laden food for a few hours a day (Jefferies. 1967). Loss by evaporation and loss by spillage must be allowed for, and increase of weight by def ecation must be kept to a minimum. Thus, the calculated dietary intake is sub ject to more error than is usually the case. Presst et al. (1970) also measured the con centration of PCBs in the liver; they found that the range was large (i.e., 70 697 ppm in birds that died compared to 3-634 ppm in those that survived). These authors conclude that Aroclor 1254 has only 1/13 the toxicity of DDT, although the different shape of the mortality curves steep with DDT, gradual with PCB -makes comparison difficult. Jef feries and Walker (1966) found good cor relation between calculated dietary intake and liver concentrations for pp'-DDT in the Bengalese finch. However, other workers (Dale et al., 1963; Stickel et al., 1966; Stickel and Stickel, 1969) have con sidered that levels in the brain are a more reliable index of toxic levels than those in the liver or whole carcass. De Vos and Kocman (1970) fed Phenoclor DP6, Clophen A60, and Aroclor 1260 to chickens at a dosage of 400 ppm. Mor tality was complete (20/20) for those birds on Phenoclor in 12-58 days and in 13-29 days for Clophen. The mortality for Aro clor was only 3/20 for a 60-day period. This differential effect is unexplained and is surprising in view of the fact that all three formulations contain 60% chlorine. These workers measured the residue lev els in the liver, and, for some birds, in the brain, for chickens dying during the ex periment. Although there was consider able variation, most of the brain levels were between 210 and 420 ppm. which can be compared to 50-80 ppm for DDT (Stickel et al., 1966). On the basis of this work, PCB is 1/4-1/5 as toxic as DDT. Little difference was noted between the three different PCB formulationr, al though the numbers of determinations involved was rather small. If this finding is borne out by subsequent work, it would suggest that differential absorption from the gut or differential penetration of the blood brain barrier is involved. The liver values were more variable. There was a considerable number in the 200-400 ppm range, but there were several values in excess of 2000 ppm. This finding is in agreement with the findings of Dale, Stickel, and co-workers that the brain levels are the best indication of acute toxic levels. McCune et al. (1962) found no mor tality with chickens fed 100 or 200 ppm Aroclor 1242 in their diet for a 4-week period. On diets of 400 ppm and 800 ppm, the mortalities over a 4-week period were, respectively, 50% and 90%. During the first 3 weeks, the mortality figures were 10% and 50%, respectively. Flick et al. (1965), using the same material at 400 ppm in the diet, had three birds out of 24 die in a 3-week period. Koeman et al. (1969a) found that a diet containing 2000 ppm Phenoclor DP6 caused complete mortality with Japanese quail (5/5) in 5 13 days and rats (8/8) in 1-56 days. Schoettger (unpublished) found that the 96-hr TLm for Aroclor 1221 using cut throat trout was 1.2 mg/1 and for Aroclor 1260 was 60.9 mg/I. In general, they found that the toxicity of Aroclors was inversely proportional to their percentage chlorination and directly proportional to their solubilities. These figures suggest that PCBs are two to three orders of mag nitude less toxic to fish than DDT. Work by Lichtenstein et al. (1969) on house and fruit flics found that PCBs were 40-300 times less toxic than DDT, the toxicity decreasing as the chlorine content in creases. Thus, it appears that the lower vertebrates and invertebrates are much less susceptible than mammals to direct toxicity from PCBs. 3) Sublethat effects. As with the chlo rinated hydrocarbon pesticides, the most important effects are long-range sublethal effects. The pathologic changes in various organs are summarized in Table 2. The table shows some interesting differences between mammals and birds. The most striking finds in mammals are alterations to the liver, whereas fluid in the peri cardial sac, kidney damage, and reduced spleen was found in birds. McLaughlin et al. (1963) found that 25 mg Aroclor 1242 injected into the yolk sac of chicken eggs caused complete mor tality, whereas 10 mg caused 95% failure and teratogenetic effects were noted among the young that hatched (beak deformity, edema, and growth retardant). Induction of hepatic hydroxylating en zymes has been demonstrated in the pi geon (Risebrough et al., 1968), rat (Street et al., 1969). and American kestrel (Falco sparverius) (Lincer and Peakall. 1970). Street and co-workers studied the effects of a diet of 50 ppm and 100 ppm on sleep ing time induced by a standard dose of hexobarbital, in vitro rates of aniline hy droxylat ion and demethylation of p-nitroanisolc, and the rate of excretion of dieldrin. These workers studied 10 com pounds ranging in chlorine content from 21% to 68% and found that all the effects increased with increasing chlorine con tent. For example, 50 ppm of Aroclor 1221 reduced hexobarbital sleeping time by 11%, whereas for Aroclor 1248 and 1268 the figures were, respectively, 35% and 48%. Thus, the sublethal effects have direct correlation with chlorine content, while the lethal effects appear to be in versely correlated (Tucker and Crabtree, in press). Lincer and Peakall (1970) noted an increase of the in vitro rate of metabo lism of estradiol in kestrels fed 0.5 and 5 ppm Aroclor 1254 in their diet and also demonstrated increased levels of cytoplas mic RNA with the higher dietary level using a cytophotometric technique. Tucker (unpublished) found that a sin gle oral dose of 500 mg/kg Aroclor 1254 caused regular egg laying of Japanese quail (Cotumix cotumix) to turn first to scattered egg production and then to stop completely for a week. The scattered eggs had shells 9% thinner than normal, but thickness returned to control values when regular laying was resumed. Mallard ducks dosed with 1000 mg/kg of Aroclor laid one or no eggs before stopping for I2 weeks. The few eggs laid after the sin gle large dose of Aroclor had shells 18% thinner, and again eggshell thickness was normal after resumption of regular laying. The period before egg laying has been noted to be increased in the ring dove (Peakall, unpublished). It is possible that the mechanism involved is increased rate of metabolism of circulating estradiol in the liver (Peakall, 1970). Anderson et al. (1969) had suggestive evidence that PCBs affected eggshell thickness, al though to a less extent than DDL. Pre liminary results with ring doves support this conclusion (Peakall, unpublished). Levels of PCBs Fold ia Natore Roburn (1965), comparing total chlorine (by concentration cell techniques) with results calculated from gas chromatog raphy, found that some unknown chlorine compounds were present in several tissue samples and eggs of wild birds in Great Britain. The first identification of these materials was by Jensen and it is stated (Anonymous, 1966) that residues in feath ers collected in Sweden go back to 1944. PCB residues have been reported in wild life from Canada (Holden and Marsden, 1967 Anderson et al., 1969; Reynolds, September I, 1970 DSW 332341 961 STLCOPCB4078917 Table 2. p*i>*h|k t>m Mni by pcb* Salami Am *1M m| (42% 4 Uni UnHi Am AMD t--A IM AiA*' Aw* (MImI atiapAy. faaal taAtai tow EaaaatiMy AAtmM*. Mm*. mA mimmiA^i 1 344 * M#v taf I Am* (HA ca H a|AMy Im MMlraNM HA GB Marfa* (33% a|kt iacfataal. latfa aaatfcar af kyaia* flakAM fywAli* ImiA AM 21, M. b IHpyai m Am far 11 Aaya (21-aA O Aiactara) Am IM ppala Act 2M pra AtM 4M prab A AM pra AM (Afacfar 1242) CkickaB 214 A4M*ra hi Am far 3 waaka (42%. Aiactef) Ctackaa tacrecaa ia watra. affect racraaatef aatk iaciaeaaf cfctariaa caataat. Aractar 1212-1 V 124112V 12*414V 1244-24% at *4 pra Da affact Naaffact Marfa* aaA MaltM 0aaM(** lAa rkimM ra*A Dhm^ Nmu at 209 mm. aitaMrat ha--nOf. Mtarjff--1 at 406 Kffct HyAraaaricaiAaai HyAraparicaiAra HyAnra'catAma. kyAtaparitaaaaac Marfa*. lactaacaA llaM ia aaricarAal tac at Ik* kifkat caacaatraoaa. Hmw of g--crut, aatarnaiaat af aOraaal mi smf bImr at law mmm tiatiaa*. Al ki0nr MBaa (1A44I MMtt m m. iiiMi Sura at A. (la araaal McCa* M al. (1442) Hick at A. (IM*) Variate Amm (S4A a Aractar) Warfkl >a 32.4% A kiara warftrt far caatiAaaaA *3.1% far th*M Ayiaf Sifkt warfkt iacrMM. * fata AwaraA fc*A4 m paricarAA cm. Preact M A. (la *taaa) AM era > Aha fat M A*y (HA ) CkickM CaatiAakalat aeiacit (cra*A 1 aaAI). liver watfkt iaciaaaaA liata 2.71 |/1H g ta 4.31 |/IH 1 (era** 3). Fatty *e*eaSfCtMIt. TtWar Akatatna. (era** 1 aaA 2) flat* witk taayA. 3. \ HyAtapancatAhaa caaaaa wrtk caapAa. t aaA 2. Rat* aitk caapA. 3. lactaacaA patpkyria. apiaaa aaal aitk laAactia* A taA pAp aaA atrapky A akit* palp lira**1 aaA 2). Ipia AanaaaaA bra 4.14* |/1M|ta 4.134 f/1*H caapA. 31. Vaa aaA KsaatM (la praaa) 'thMwtw PP I IcmM It. OcrCw AAO {nat 21 at Aiada Ittfl Inaat SI ai nt OiHwata rftccii atM atw errat waa. AC ctakat Pa a ai t at I atta M tan. ah f Sk muMm a iai|f 3 1970). Germany (Fiuczynski and 'Vendland, 1968: Koeman et al., 1967), Great Britain. (Holmes et al., 1967; Press! apd Jefferies, 1969: Presst et al., 1970; Holden and Marsden, 1967), Netherlands (Koe man et al., 1967; 1969), Sweden (Anony mous, 1966: Jensen et al., 1969), and the United Slates (Anderson et al., 1969; Risebrough et al., 1968; Risebrough, 1969).. Biological Magnification No detailed studies, such as those for DDD at Clear Lake (Hunt and Bis:hoff, I960) and DDT and its metabolites in Lake Michigan (Hickey el al., 1966), have yet been made for PCBs. The most detailed studies currently available are those of Jensen et al. (1969). The figures (mean, range of values, and sample size) given in the table below are taken from their paper. PCBk (ppm in exiractublc fal) Muud , Baltic Stockholm Archipelago 4.3 (l .9-3.6) (40) 5.2(34-7.0) (15) Herring 6.1(0.5-23) (18) 5 1(3.3-8.5) (4) Seal 34(16-44) ( 3) M) (>6-56) ( 3) Guillemot eg> \ White-tailed Egtc 250(140-360 M 9) Pectoral mmde -- 14.000 (MOO-17,000) ( 4) Brain --- 910(490-1500) ( 3) Eggi Heron -- 540(250-1100) ( 5) -- 9400 ( I) The levels in three species of fish in Clear Lake in 1968 were 0.03-0.005 ppm (wet weight) compared to 0.098 ppm in the breast muscle of a western grebe (Rise brough et al., 1969). Anderson et al. (1969) found that in most fish extracts the levels of PCBs were less than 0.1 ppm, whereas the levels in the eggs of cormorants (Phalacrocorax auritus) were 5-9 ppm. Presst, Jefferies, and Moore (unpublished) found that the livers of fish-eating birds in the British Isles ranged up to a maximum of 900 ppm (wet weight), bird feeders up to 70 ppm, mammal eaters to 50 ppm, and insectivorcs to I ppm. Unfortunately, no average values or prey items were in cluded. The physical properties of PCB and the available residue data clearly indicate that these materials are capable of bio logical magnification up the food chain. 962 BioSciencc Vol. 20 No. 17 DSW 332342 STLCOPCB4078918 T Rati* ( DDT to PCB Risebrough et al. (1968) and Risebrough (1970) have examined the ratio of total DDT, i.e., DDT and its metabolites to PCB. He has found that the ratio DDT/ PCB was 1-2 in San Francisco Bay and 5 10 for seabirds in the Pacific: in the Gulf of California, a region rdatively remote from contamination, the ratio was 9-10. Vermeer (quoted in Reynolds, 1970) in western Canada found a DDE/PCfl ratio of 7 for California gull tissues and 13 for great blue heron eggs. In the Baltic the ratio was 1-2 (Jensen et al., 1969), al though along the west coast of Sweden the ratio was as low as O.IS. In grebes in the British Isles the ratio was 0.4-0.8 (Prcsst and Jefferies, 1969). For sea-bird eggs. Presst et al. (unpublished) found ratios of 0.06 to 0.3. The overall impression is that (he amount of PCB in tissues tends to paral lel that of DDE, at least in local ecosys tems, and the DDE/PCB ratio is lowest near industrial areas suggesting that PCB is not carried quite so readily to remote areas. Nevertheless, the variation of the ratio is small enough to suggest tnat the routes of dispersal arc similar. Since the evidence points to aerial fallout as the route of dispersal of the chlorinated hy drocarbon pesticides (Risebrough et al., 1968; Frost, 1969), it is likely that this is also the main route for PCBs. The path ways by which PCBs escape into the eco system are poorly known, although the possibilities have been recently discussed at some length (Risebrough, 1970; Rey nolds, 1970). Since a large number of plastics and resins may contain PCBs (Table I), the most likely route is combus tion of these materials. This supposition remains to be tested. The possibility that PCBs could be derived from DDT should also be considered. The possibility that this conversion occurs in tissue is most unlikely for two reasons. First, it has never been detected despite the detailed work on the metabolism of DDT. Second, the only mechanism likely to give rise to a biphenyl is via free radicals and this is unlikely to occur in tissue. However, in the atmosphere under the influence of UV light, such a breakdown is more prob able. A possible reaction is shown in Figure I. Tautomeric shift could lead to a variety of isomers of dichlorophenyl, but it is difficult to envision the formation of more highly chlorinated biphenyls by this route. Since PCBs extracted from biological material match well with higher Aroclors (i.e., 1254), it seems unlikely that PCBs found in nature could be derived from other materials. Significaace ef Cams) Levels In view of the similarity of PCBs to DDT and its metabolites, the addition of PCB residues to the environment is rough ly equivalent to an increase of DDE resi dues. However, since a synergistic effect of PCBs on DDT has been demonstrated in insects (Tsao et al., 1953), this pos sibility should not be overlooked in higher organisms. The enzyme induction effects of PCBs have been well documented, and carbonic anhydrase inhibition is likely. The effect of PCBs on photosynthesis is a critical experiment that has not been done. Wurster's (1968) experiments with phytoplankton should be repeated with PCBs. The most critical area for research on PCBs is to discover the major source(s) of escape into the environment. Legislative control of a material escaping as a side effect of its use may have a different set of problems than the control of pesticides which are broadcast as a function of their normal use. The evidence suggests that it is important to find the leak and stop it. Acknowl edgmeats We are grateful to Dr. Koeman (Uni versity of Utrecht), Drs. Presst and Jef feries (Nature Conservancy, Great Britian>; Dr. Reynolds (Ontario Research Foundation), Dr. Street (Utah State Uni versity), and Dr. Tucker (U.S. Fish and Wildlife Service) for advance copies of important material. Some of the work presented here was carried out under NIH Grant ES00306, Dr. T. J. Cade, Principal Investigator. The review was written while one of us (D.B.P.) was an Established Investigator, American Heart Association. Thanks go to Dr. T. J. Cade for reviewing the manuscript. References Abbott, D. C,, R. B. Harrison, J. O'G. Tilton, and J. Thomson. 1966. Organochlorine pesti cides in the atmosphere. Nature. 211: 259 261. Anderson, D. W,, J. J. Hickey, R. W. Ritfbrough, D. F. Hughes, and R. E. Christen sen. 1969. Significance of chlorinated hydro carbon residues to breeding pelicans and cormorats. Can. Field Natur., 13: 89-112. Anonymous, 1966. Report of a new chemical hazard. NewSci.. 32: 612. Archer, T. E., and D. G. Crosby. 1966. Gas chromatographic measurement of toxaphene in milk, fat. Mood, and alfalfa hay. Bull. Environ. Cotumm. Toxicol.. 1: 70-75. Armour, J., and J. Burke. 1969. Polychlorinated biphenyls as potential interference in pesti cide residue analysis. F.D.A. Laboratory In formation Bull. No. 918, II p. ---------- 1970. A method for separating poly chlorinated biphenyls from DDT and its analop. J. Assoc. OJfic. Anal. Chem. (in press). Bagley. G. E., W. L. Reichel, and E. Cromartie. 1970. Identification of polychlorinated bi phenyls in two bald eagles by combined gasliquid chomatography-mass spectrometry. J. Assoc. OJfic. Anal. Chem.. S3: 251-261. Bartha, R. 1969. Pesticide residue interaction creates hybrid residue. Science. 166: 1299 1300. Bennett, G. A., C. K. Drinker, and M. F. War ren. 1938. Morphological changes in the livers of rats resulting from exposure to cer tain chlorinated hydrocarbons. J. Ind. Hyg. Toxicol.. 28: 97-123. Bourne, W., and C. Mead. 1969. Seabird slaughter. B.T.O. News. No. 36, p. 1-2. Dale, W. E,, T. B. Gaines, W. J. Hayes. Jr., and G. W. Pearce. 1963. Poisoning by DDT: Re lation between clinical sips and concentra tion in rat brain. Science. 142: 1474-1476. Deonier, C. C., H. A. Jones, and H. H. Incha. 1946. Organic compounds effective apinst Anopheles quodrimaculatus. Laboratory test. J. Econ. Entomol., 34: 459-462. Duda, E. J. 1957. The use of chlorinated poly phenyls to increase the effective insecticidal life of lindane. J. Econ. Entomol.. 50: 218 219. Erro, F., A. Bcvenue, and H. Beckman. 1967. A method for (he determination of toxaphene in the presence of DDT. Bull. Environ. Conlam. Toxicol.,2: 372-380. Fiuczynski, V. D., and V. Wendland. 1968. The population of hfltyus migrant in Berlin 1952 1967. J. Omithoi.. 189: 462-471. Flick, D. F,, R. G. O'Dell, and V. A. Childs 1965. Studies of the chick edema disease. 3. Similarity of symptoms produced by feeding chlorinated biphenyl. Poultry Sci.. 44: 1460 1465. Frear, D. E. H. 1968. Pesticide HandbookEnloma. College Science Publishers. Frost, J. 1969. Earth, air, and water. Environ ment. II: 15-33. Hickey, J. J., J. A. Keith, and F. B. Coon. 1966. An exploration of pesticides in a Lake Michipn ecosystem. J. Appl. Ecol.. 3 (Suppl): 141-154. Holden, A. V,, and K. Marsdcn. 1967. Organo chlorine residues in seals and porpoises. Nature. 216: 1274-1276. Holmes, D. C , J. H. Simmons, and J. O. 'G. Tatton. 1967. Chlorinated hydrocarbons in British wildlife. Nature. 216: 227-229. Homstcin, I., and W. N. Sullivan. 1953. The role of chlorinated polyphenyls in improving lindane residues./. Econ. Entomol.. 46: 937 940. Hunt, E. G,, and A. I. Bischoff. I960. Inimical effects on wildlife of periodic DDD applica tions to Clear Lake. Calif. Fish Gamt. 44: 91-106. Jefferies, D. J. 1967. The delay in ovulation pro duced by pp'-DDT and its possible significancc in the field. Ibis. 189: 266-272. Jefferies, D J., and C. H. Walker 1966. Uptake of pp'-DDT and iti post-mortem breakdown in the avian liver. Nature. 212: 533-534. Jensen, S.. and G. Widmark.1967. Organochlorinc residues, OECD preliminary study 1966 67. Report given at the OECD Pesticide Conference. Jensen, S., A. G. Johnds. S. Olsson, and G. Otteriind. 1969. DDT and PCB in marine DSW 332343 I IQIO STLCOPCB4078919 animali from Swedish water*. Nature. 224: 247-250. Klein, A. K., and J. O. Walts. 1964 Separation and measurement of Pcrthanc, DDD (TDE), and DDT in leafy vegetables by electron capture gas chromatography. J. Assoc. OJfic. AmI. Chem.. 47: 311-316. Koeman, J. H., A. A. G. Oskamp, J. Veen, E. Brouwer, J. Rooth, P. Zwart, E. V. D. Brock, and H. Van Genderen. 1967. Insecticides as a factor in the mortality of the Sandwich tern (Sterna sandvicensis). A preliminary communication. Meded Riijksfac. Landbouwwetensch. Gent.. 32: Ml-853. Koeman, J. H., M. C. Ten Nocver de Brauw, and R. H. de Vos. 1969a. Chlorina'ed bi phenyls in fish, mussels and birds from the River Rhine and the Netherlands coastal area. Nature. 221: 1126-1128. Koeman, J. H., J. A. J. Vink, and J. J. M. de Gocij. 1969b. Causes of mortality in birds of prey and owls in the Netherlands in the win ter of 1968-1969. A rdea. 57: 67-76. Lichtenstein. E. P.. K. R. Schulz, T. W. Fuhremann, and T. T. Liang. 1969. Biological inter action between plasticizers and insecticides. J. Eton. Entomol., 62: 761-765. Lincer, J. L,, and D. B. Paakall. 1970. Induced hepatic steroid metabolism and increased cytoplasmic RNA by polychlorinated biphe nyls (PCB) in the American kestrel (Falco sparverius). Nature (in press). .. McCune, E. L., J. E. Savage, and B. L. O'Dell. 1962. Hydropcricardium and ascites in chicks fed a chlorinated hydrocarbon. Poultry Sci., 41: 295-299. McLaughlin, J., Jr., G. P. Marliac, M. J. Verrett, M. K Mutchler, and O. G. Fritzlaugh. 1963. The injection of chemicals into the yolk sac of fertile eggs prior to incubation as toxicity test. Toxicol. Appl. Pharmacol. 5: 760-771. Miller, J. W. 1944. Pathologic changes in ani mals exposed to a commercial chlorinated diphenyl. U.S. Pub. Health Rec.. 59: 1085 1093. Monsanto Technical Bulletins 0/PL-3IIA, O/PL-306, and O-FF/I. Mulhern, B. 1968. An improved method for the separation and removal of organochlorinc insecticides from thin-layer (dates. J Chromalogr.. 34: 556-558. PeakaH, D. B. 1970. Pesticides and the repro duction of birds. Sci. Amer. 222 : 73-78. Penning, C. H. 1930. Physical characteristics and commercial possibilities of chlorinated diphenyl. Ind. Eng. Chem.. 22: 1180-1182. Preset, L, and D. J. Jefferies. 1969. Winter num bers, breeding success, and organochlorine residues in the great crested grebe in Britain. Bird Study. 16: 168-185. Presst, I., D. J. Jefferies, and N. W. Moore. 1970. Polychlorinated biphenyls in wild birds in Britain and their avian toxicity. Environ. Pollut (in press). Reynolds, L. M. 1969. Polychlorobiphcnyls (PCBs) and their interference with pesticide residue analysis. Bull. Environ. Contam. Toxi col.. 4: 128-143. 1970. Pesticide residue analysis in the presence of polychlorobiphenyls (PCBs). Residue Rev. (in press). Risebrough, R. W. 1969. Chlorinated hydrocar bons in the global ecosystem. In: Chemical Fallout. G. G. Berg and M. W. Miller (Eds.), Charles C Thomas, Springfield, III., p. 5-23. 1970. More letters in the wind. Envi ronment. 12: 16-27. Risebrough, R. W., P. Reiche, D. B. Peakall, S. G. Herman, and M. N. Kirven. 1968. Poly chlorinated biphenyls in the global ecosystem. Nature. 220: 1098-1102. Risebrough. R. W,, R. Reiche, and H. S. Olcott. 1969. Current progress in the determination of polychlorinated biphenyls. Bull. Environ. Contam. Toxicol.,4: 192-201. Roburn, J. 1965. A simple concentration-cell technique for determining small amounts of halide ions and its use in the determination of residues of organochlorine pesticides. Analyst, 90: 467-475. Schmidt, H., and G. Shultz. 1881. Einwirkung von Fiinffach-Chlorphosphor auf das ydiphenol. Ann. Chem., 207: 338-344. Simmons, J. H., and J. O'G Tattoo. 1967. Im proved gas chromatographic systems for de termining organochlorine pesticide residues in wildlife. J. Chromatogr.. 27: 253-255, Smyth, H. F. 1931. The toxicity of certain ben zene derivatives and related compounds. J Ind Hyg. Toxicol.. 13: 87-96. Stickd, L. F,, W. H. Stickel, and R. Christensen. 1966. Residues of DDT in brains and bodies of birds that died on dosage and in survivors. Science. IS1: 1549-1551. Stickel, L,, and W. Stickel. 1969. Distribution of DDT residues in tissues of birds in relation to mortality, body conditions, and time. Ind Med. Surg . 3B: 44-53. Street, J. C,, F. M. Urry, D. J. Wagstaff, and A. D. Blau. 1969. Comparative effects of poly chlorinated biphenyls and organochlorine pesticides in induction of hepatic microsomal enzyme. Presented at ACS meeting Septem ber 8-12, New York. Sullivan, W. N,, and I. Homstcin. 1953. Chlo rinated polyphenyls to improve lindane resi dues. J. Econ. Entomol.. 46: 158-159. Tsao. Ching-Hsi, W. N. Sullivan, and I. Hornstein. 1953. A comparison of evaporation rates and toxicity to house flies of lindane and lindane-chloriated polyphenyl deposits. J. Econ. Entomol.. 46: 882-884 Tucker, R. K,, and D. G. Crabtree. 1970. Hand book of Toxicity of Pesticides to Wildlife. U.S. Dept, of Interior, Bureau of Sport Fish eries and Wildlife, Resources Publication No. M. de Vos, J. G., and J. H. Koeman. 1970. Com parative toxicological study with polychlori nated biphenyls in chickens with special ref erence to porphyria, edema formation, liver necrosis and tissue residues. Toxicol. Appl. Pharmacol, (in press). Widmark, G. 1967. Possible interference by chlorinated biphenyls. J. Assoc. Offic. Anal. Chem . 50: 1069. 1968. Determination of number of com pounds which can result from the chlorination of biphenyl, and development of a simple sys tem by which these may be codified. OECD report Sweden. Wursler, C. F. 1968. DDT reduces photosynthe sis by marine phytoplankton. Science, 159: 1474-1475. DSW 332344 STLCOPCB4078920