Document 3ezbd9jJvd3bQGrvMj6xQQzk6

) p Joxu. Rf*c*rch VS. G*oL Surv*r VoL 1, No. I, Sfpt.*Oct. 1MI, P. !* -Ml A STUDY OF THE DISTRIBUTION OF POLYCHLORINATED BIPHENYLS IN THE AQUATIC ENVIRONMENT By HANS J. CRUMP-WtESNER, HERMAN R. FELTZ. and MARVIN L. YATES. Washington, D.C., Arlington, Va., Menlo Park, Calif. y4bifrr/.-Dta gathered from monitoring ctivilie* and project fttudica indicate the ubiqiiilou* ommcncc and diKiribution of pufychlorinated iHplicnyU in Uic acjunlir environment. ItcMclurp have b*rn delected in Minplca from 19 Stales, representing nearly every region of llie eounlry. Comcnlfatinns runsril from 0.1 to 1.0 m>/l in unfiltcnd water umple* and 5.0 to 3,20(1 np/kg in bottom sediments. I'CII residue* were atro found In fi1i and aqiulie jilaula. Simple* were prepared by lire ume leehniqne* used for general chlorinated inseclicldc detection, with aftcrinl atlrntion to cleanup and reparation of PCB't from other compounds, Uasic identification and quantification woe made by dual-column clrclron-capture gas chromatography and confirmed by gat chromatography-inat* spectrometry whenever pomjHc. The sampling program in r\pcc-tcd to be broadened geographically In 1973 and increased in successive year* in nrdcr to more adequately definr the distribution of FCB residues in the major drainage basins of the United States. AcknowledgmentThe authors appreciate the assistance of personnel in the U.S. Ceotogiral Survey laboratories at Austin, Tex., and Washington, D.C., in preparing data compilations. ANALYTICAL TECHNIQUES PCI) residues were analyzed hy the multiple-pesticide residue methods for water, suspended sediment, and bottom material, aa descrilied by Goerlitz and Drown (1*J72). The analytical procedures include not only rhlorinulcd pesticides, but also the general class of orgauochlorine compounds. Special atten tion was given to cleanup and separation of PCD a from coextractives. Extraction of water, rodlment, and biota Within the past three years, polychlorinated biphenyls (PCU's) have been recognized as a major environmental contaminant. Their detection has caused widespread concern and has genera led intense interest ill data relating to the presence and effects of these compounds, pnrln ul.irly in the aquatic environment. First produced ahnul dO years ago. I'Ll! com|>oiinds have become increasingly nselul in such industrial applications as components in transformers ami capacitors, heat exchangers, paints, inks, dyes, And dust control agents. While much attention Inis been focused on the. estimation of levels and potential hazards of PCU's in aquatic organisms, few data are available on the oceurrcncr of PCU's in water and bottom sediments. The U.S. Geological Survey, thruugh its water-quality-monitoring activities and walcr-resources-asscssmenl projects, has been alert to the PCU problem since it was first retried by Widmark (l%7). Although file Geological Survey lias no nationwide PCH assessment program, sufficient data hove ar.fimiulalcd from pesticide-residue programs to permit a preliminary assessment of PCU contamination of the Nation's hydrologic environment. The purpose of this paper is to present data gathered from these activities, showing tin; widespread occurrence of PCU'} in significant concentrations in unfiltered surface water and ground water, bottom sedimrnts, flora, and fauna. One-liter unfiltcrcd water samples were collected in precteancd gloss hollies and were extracted three times with hexane. The hexane portions were combined, dried with anhydrous Na2SO. , und concentrated to 1 ml before cleanup and analysis hy electron-capture gas chromatography (F.CCC). Fifty-grain sediment samples (dry-weight basis) were ex tracted with on acclone-heximc solvent. The sediment is dispersed First in acetone, mid hexane is added to recover the ocetonc together with the desorbed material. The extract is washed with distilled water, dried over Na,50, , anil eintrvntraleil to 5 ml for cleanup before 1'COC analysis, The two extraction procedures used for biota arc described in iui analytical marmiil issued by the Food and Drug Administration (I*.171). Fish samples were extracted with petroleum ether in a blender, whereas aquatic plants were extracted with acetonitrile. The chlorinated hydrocarbon fraction was partitioned between petroleum ether and aeetonitrile, dried, and concentrated to a final volume of 5 ml. Separation Dollom-sedimcut, fish, and plant extract* were cleaned up with a technique devehqied hy Law and Gocrlitz (written commun., l`J7l). This technique requints less time und smaller 603 DSW 026628 STLCOPCB4010589 ' 604 POLYCIILOI! tNAl El> MtMIt.NYLS IN AQUATIC KN VlltONMKNT volumes of solvent* for elution than other widely used methods. Liquid-solid column rlirimnlopr:i]liy was employed, by use of Iwo different type:. of si-mimicro columns in sequence, ax shown in figure 1. Ilrxuue extract* were first passed llirougli an alamimi column, and one fraction was futllicr chromatographed on silira gel to separate PCB's from chlorinated insecticide*. Successive chromatography on alumina and silica gel results in a simultaneous cleanup and separation of PCR's from (lie common insecticides, except alilrin, and a slight overlap of pp DDE. To achieve a reduction in background interference, mercury was added to remove sulfur from bottom-sediment extracts before they were ap plied to the silica column. In order to insure reproducible chromatographic conditions, the activity of the adsorln-nls wus carefully controlled. Water extracts were cleaned up on u deactivated ulumiua micro column (Law and Gocrlilz, 1070). When I'GII's were detected in llie clctmcd-up water extracts, they were also separated on a semimicro silica gel column. Identification Basic identification was made by dual-column ECGC (DC-200 and QF-l/OV-17) and confirmed by gas chroma tography-mass spectrometry (GC-MS) when sample size and concentrations were sufficient. The amount of PCB'e was f 0-- 20 ml fraction (h*no) PCB PCN't AJdrln Chio'dcn* ODD DOC ODT Hrpichor Llrwii<s# Toiftpficrt* Hexane extract Alumina 20--30 ml friction <hK*n) Olaldrln Crtdnn HtutAchlPf tpoiid* If 3S--50 ml traction Cthtoft Methyl p*athion Pcrpthion MvUirt frithiM gel 0--25 ml (taction (hoiane) rco'* PCN'fc Aidftn . 20--45 ml fraction (benzene) DOD DOC ODT Heptftchlor l*Hl*n* Toktpntn* Figure | .-Scheme for M'p.iralina poll chluiiicilrit tiiplw-m lx (I'dll'k) and imlyrhloriiciletJ n.iplillulcucs (Pt'N's) fumi |H--licilcx. determined by mutching the iinkiiowii p-uks on tin- chromato gram l<> the nearest fitting commercial formulation and measuring the areas of four corresponding peaks. Retention lime uud peak-area measurements were made with a digital electronic integrator. The lower detection limit for PCU rrsiducs wus 0.1 ^tg/l in water and 5.0 /ig/kg in bottom sediment. Reported levels are subjrcl to considerable errur In-cause of the complexity of multiple peaks, some peak alteration, and the occasional presence of mixtures of PCB's in environmental samples. At best, reported values are estimates that may lie as much as 50 percent in error. DATA DESCRIPTION Occurrence of pesticide residues in the aquatic environment have Ikt.ii documented over a period of years through monitoring programs of several Federal agencies. As early us 1957, studies of chlorinated hydrocarbon pesticides in major river basins were made by the Federal Water Pollution Control Administration, now a purl of the Environmental Protection Agency, hy use of the cnrbon-adsorpliun technique (Brcidonbach and others, 1964). In 1964, interagency coop-ration in pesticide-monitoring programs culminated in a proposal to begin a national monitoring program. The original program for water was described in 1967 in the first issue of the Pesticides Monitoring Journal (Green and Love, 1967). The purpose of this program, revised in 1971 (Felix and others, 1971), is to provide continuing information on the IcvcIh of pesticide residues in (he water resources of the Nation and to identify possible problem areas. Because PCB's arc analyzed by the multiple-pesticide residue techniques, routine reporting of tliesc conqiounds hits been incorporated into current pesticide programs. At present, samples from 20 of the 101 proposed network sites arc collected and auulyzcd by the Geological Survey. All the samples are collected from sites located wrst of the Mississippi River und provide continuity with a network esluhlisls-d to evuluule the quality of water used for irrigation (Brown and Nishioka. 1967'. Munigohl and Schulze, 1969). Budgetary restrictions have prevented further iiiiplemenlulioii of (he network. After development of the technique to separate PCIl's from pesticide residues, examination for the presence ol I'Cli's in water and suspended- and bottom-sediment samples eolh-eted for the national monitoring program began in January 1971. Funding lias lieen requested to increase the number of network stations to 50 in I9i1, and to 100 stations in !9i 1. allowing a better assessment of l`('.l{'s and pesticides in major drainage basins throughout the United Stales. Analysis of 194 water samples and .1.1 but tom-sediment samples revealed no positive identifications of I`4.71 i s; however, these data are not Indy representative of the entire Nation because of the limited numlicr of sites sampled. Ill 19511, the (ieologieal Survey began op-ration of a bench-mark network to provide basic bydmhsgie data on DSW 026629 STLCOPCB4010590 CIlUMP.WlESNni, KKI.TZ, AND YATES b05 M-Irrlrd stream basins itirtxij.'In>ut 11n* United Slates that nre rs| i teii li> rem.iin ill their present natural condition or are lint rv peeled to tie significantly altered hy man. I.ncalions of the 57 heneli marks established in 17 States arc shown in fipirr 2. To in-urc minimum interference hy man, many of the hydrologic lieneli marks arc in national parks, wilderness areas. Stale parks, national forests, and areas set aside for scientific (Indy. A detailed description of the network basins, including drainage, climate, topography, geology, legrtution, hydrology, witter quality, and manmade influences, can he found in a rrport hy Cobb und lliei-eekcr (1971). Data gathered from 4ft of these sites are presented in table 1. Despile the careful acrrriiine for pristine location of bench.mark sites, two hottom-sediinenl samples analyzed in the 1972 water year contained PCD residues. A value of 5.2pg/kg was measured in the sample from South Fork Rocky Creek near Briggs, Tex., and C.fl pg/kg in the sample from Upper Twin Creek at McGaw, Ohio. Table 1.-Summary of I'Cli residue data, national hydrologic bench mark nc.tu'ork, January 1071 -June 1072 Type of simple No. of tamptes Oeeurrr.ncei Concert* lml ion Witrr ................... . . 54 Bottom lediincnt . - - Pft/kg . .. .. 51 0 2 5.2, 8.8 The majority of the. Geological Survey's water resources studies are. condueled in cooperation with Slate water re sources and pollnlion-eonlrol agencies, or in rrs|voiiM- to requests from other Federal agencies. PCD data from these programs in 15 Slates nre presented in tables 2, 1, and 4. Water samples alone, because of the tow water solubility of I'CIj's, are not a good indicator of the widespread occurrence of the. cnmfKumds, and show an incidence of slightly over 5 percent. Unfillcred water samples from 12 of the .15 States had I'CB concentrations ranging from 0.1 to 4.0 pg/l. However, a significant numlier of suspected traces of l'CB's were not reporteil because of several analytical limitations. Generally, resampling of areas where PCll's were first detected revealed that the compounds were still present several months later. Bottom sediments were, collected concurrently with many of the water samples reported in table 2. These samples were token from lakes and streams that drain a variety of land-use areas generally located away from industrial renters. The data in table 3 show that the Iroltom-sedimenl samples mav be used as an indicator of PCB contamination in the Nation's hydrologic environment. Significantly, of samples collected at random from 10 States, 13 contained PCB's in the range 5.0-2,400 pg/kg. Across the Nation, one of every five lrollom-srdiinenl samples examined contained I'CB's. Data available front Florida (tabic 4) merit special attention, because they reveal the distribution of residues in several environmental components. Only 12 of 231 unfiltcred water Figure 2.- M.i|> hlnovina loralinn uf ti> iliutugic farueli-niatk stations. NumUc* refer to li.-t in Cobb and IlieM-rker (1971). DSM 026630 STLCOPCB4010591 w, J2- - F,v dmi. 30/ /_--. c /<). ooo / <3. I ~ f/o/LA rA^ s-o |) i /7i><" I Q& Q '^sr~~ /D / ;loO /< osw 026631 STLCOPCB4010592 . POLYCHLORINATED III PHENYL S IN AQUATIC I NVIUONMKN f Tablf- 2.-Siimtnory of PC/? rufiilur i/'ila for tut fate otid fround wofrr, Table 4.-.Summary ,,f IT/I residue <Inin from selected aumpling lilri in January 1971 -funn I9?2 A 7ortJat Jammry 1 '>7f - /unn I *) 7J State *1 ~r **} *"&* n~--.m Cotci-n' tratinn tcn"t (^/D Median concentration <,/) Typed ample No. f Occur. aaiuplcs rencca C`,n<*n' tealwui range Median , en- traium Alaska................... Arixona................ 3 8 Arkansas .............. 32 California.............. 161 Colorado.............. 32 Connecticut.......... 13 Hawaii................... 5 low* ..................... 24 Kansas................... 10 Kentucky.............. 7 Louisiana.............. 9 Maine..................... 2 Maryland........ 6 Massachusetts ... 5 Michigan .............. 2 Minnesota ............ Mississippi............ 3 8 Missouri................. 21 Montana .............. 47 Nebraska.............. 44 New Jersey .......... New Mexico.......... 11 36 New York ............ 325 North Dakota ... 40 Oklahoma ............ 19 Oregon................... 13 Pennsylvania .... 2 Puerto Rico.......... 7 South Dakota ... 18 Texas..................... 660 Virginia......... 4 Washington.......... 25 West Virginia .... 4 Wisconsin....... 3 Wyoming.............. 18 0 ............ 0 .... 0 .. . . 2 0.1,0.1 1 .3 6 .1-- -2 0 .... 0 .... 0 .... 0 -----0 .... 0 .... 1 .1 13 0 .... 2 .1, J 0 .... 0 .... 0 .... 0 .... 3 -1 0 .... 52 .1-4.0 0 .... 0 .... 0 .... 1 -2 1 .1 0 .. .... ' 12 J-3.0 1 -1 0 .... 0 .... 0 .... 0 .... .... .... `. . . . .... 0.1 .... .... .... -----.... .... .... .... .... .... .... .... .... .... ^ -- .... .... .... .... .... .4 .... .... .... Water..................... mfl ... Rottum sediment . udkg.. Flora..................... jig/kg . . Fauna ...................ng/kg . . ill 118 16 43 12 0.1-2.1 0.2 50 5 -3,200 30 5 10-50 20 36 6-1,000 40 of tlic DDT family in soulherji Florida (Frit/, und Culliertson, 1972). DISCUSSION AND EVALUATION Preliminary data presented in tliix report indicate that ignifiraiil concentrations of lTlt'sure widespread in the water resources of the Nation. However, there are some short comings of datu compiled front pesticide-residue programs. First, there is the problem of nourepresctilulive sampling within Slates mid some repetition of sampling in n given basin. Second, the lower limit of detection for PCll residues on the ,basi of * one-liter water sample is inadequate for critical evaluation. Trace amount** of Ic6 than 0.1 grg/l were detected in a significant number of bantples hut were excluded from the tabulation because they could not be confirmed. Third, because of the low solubility of PClIs in water, especially the higher chlorinated ones, the hulk of PCI) residues in streams are associated with suspended sediment and lioltom material. Therefore, PCI) concentrations may Ire expected to vary directly with the suspended-sediment concentration in the Xible 3. - ^Ufimicry of PCfi residue doto for b*>ltotn irdimenti, Jtmior)' cross scelion of a stream. Most surface-water samples were 1971-June 1972 collected by depth integration at Hie center of flow, which State No. of Occur* aatnplea rcnccs Alatka ................. . Arkansas ............ . California............ . Connecticut ... . . Hawaii................. . Cfcorpn .............. . Mjryliud............ MtuiAftippl......... . New Jersey . ... . Orcpon ................. . 1'emtxylvama .. . South Carolina . . . Texas ................... . Yirpmia.............. . IVisliinpInn . . . . . West Virginia . . . . 3 23 13 1 4 12 8 12 4 16 11 293 10 10 2 0 4 3 1 0 10 5 2 10 2 n 8 23 e 0 i Concen tration (ag/fcg) 20-2,400 20-190 4u 10-1,300 10-1.200 50; 170 8-250 15; 140 10-50 30-200 7.9-290 5-00 10 Median concentration (C/k*l 60 85 .... 300 30 20 20 SO 8(1 40 .... .... usually does not provide a representative sample of suspended sediment. Future investigations sltould he made with it depth-integrating sampler using the equal-lransil-rule proce dure (Felix and Culbertson, 1972). In spite of these short comings, evidence for the ubiquity of PCD's in the hydrologic environment is clearly established. As ollters have pointed out (Coerlilr. and Law, 1972), polychlorinated naphthalenes (PCN's) are compound* that have uses similar to ITU's and may jsossibly he present in environmental samples. They can be separated from rliluii- nated hydrocarbon insecticides and are eluted in the same fraction as PClls by alumina-silica gel column thrum.ilug raphy. Analysis of sediment samples collected Iroin a sn-U i Florida drainage ditch contained mixtures of ranging from 1,250 to 5,000 pg/kg, whereas the water samples overlying the sediments averaged 5., pg/l. lilentilii .'.lion wa- samples contained ITU's ranging from 0.1 to 2.1 pg/l, liul over 40 percent of tIn: associated bollorli sediments analyzed during tin-some period were conlnniinatcilwitli I'Cil's ranging confirmed by both niieroemthnnelry amt I his mjy possibly be tin- first evidence oi tin- occurrence of l'( ,N s in an environmental sample and illustrates the import.nice oi devel from 5.0 lo ,'1,20(1/tg kg. Tin- median I'l.l! conivnlrations of oping aiialylical capability for the surveillance of other 20 Jtg/kg mill TO pg/kg louml ill aquatic plants and fisli, organoi lilorine compounds that may behave like chlorinated rcBfHrclividy, follow llu* wlicitic of tin* biological accumulation hvdi'ix urhoii jieslicides. I 0SW 026632 STLCOPCB4010593 CIlUMp.WIEsNKIl. I KI.TZ. AND YATES 607 The entrance of I'CH's into till' aquatic riivirtiiiiiirut is probably related primarily to low-trmpiTalnro iuciucratiim of aolid wastes, industrial wade di.-jm.'al into waterways. ami sewage out rails. The highest levels are usually associated with industrial arcJs and nearby aquatic food chains. In contrast, we have noticed (significant concentrations in the hotloni sediments of drainage ditches, multipurpose canals, und suburban real eslute lakes remote from major industrial and metropolitan areas. The presence of I'Ofl's in real estate lakes was attributed to a variety of rniistrurlioti materials used ill tire housing industry. PCfl's in surface and ground water used for public water supplies were delected through cooperative programs. The highest concentration was d.O tg/l in an untreated source for a city in the Stale id New York. It is Hear that the presence of I'CR's and other organnchlorine compounds in the aquatic environment merits con tinuing observation because of the limited evaluation that ran be made from tlie meager data available. It is important l measure baseline levels of I't'.li's in streams mid lakes in order to determine trends. Long-term monitoring on u systematic basis will provide the data necessary to ossc.su the presence of PCD residues and concurrently reveal problem areas. Brown. Eugene. anil Nidiiuka. Y. A.. 1067. Pedirtdcs in srli rlnt wrstrrn *ttrams- rnulrilHilinn t the National Program: I'rsliridr* Monitoring Juur., v. I. no. 2. p. 211 --W. Cuiili, K. D- and lliesreki-r. J. E,, 1071. T1*e uatinunl hydrologic brnrli-mark nrlwork: II.S. Orul. Survey Grc. -UiO tt. 311 p. Krllz, It. tl., and Culbertson, J. K., 1972. Summing procedures and problems in the drtrrmniatiou id (s-sUeide rrsidues in the hydro logic rnvirunmriil: Pesticides Monitoring jour., v. 6, no. 3, p. 171 -178. Kelt/., Il.lt., Sayers. W. T.. and Nicholson, II. P., 1971, Nations! monitoring program fur die ussessinent of pesticide rrsidues in water: tYslicidrx Monitoring Jour., v, 5, no. l.p. 54--62. tiorrill/, I). K.. and Brown, Eugene, 1972. Methods tor ao.itV .3 - ,3 organic substances in water: II.S. (Teol. Survey Techniques Water Hesources luv, TWI 5-A3, 40 p. (Book 5, laboratory analysis.) Gorrtilz, D. F.,, and Law, L. M.. 1972, Gdoriiutrd naphlhalmc. in pesticide analysis: tluli. Environmental Containinjtion ami Ihmu.I. ogy, v. 7, p. 243-251. <Jrr.cn, It. S., and lavvc, S. K., 1967, Network tn monitor hydrologic environment covers major drainage rivers: Pesticides Monitoring Jour., v. 1, no. 1, {>. 13-16. Law, L. M.. and Gm-rlilz. I). K,, 197 . Micrucoliimn rliromalograpliic rlramip for the analysis of pextii-i r in water: Assoc. Official Anal. Oicmisls Jour., v. 53, p. 1276- I :Ii. Manigold, D, II., and Schulze, J. A.. I'-69. Pesticides in srleeted western streams-a progress rejuirt: Peslieides MoitiUiciug jour., v. 3, no. 2, p. 124-135. U.S. Food and Drug Administration. 1971, Peaticidea analytical REFERENCES CITED manual, v. 1: U.S. Food and Drug Adin. Wldtnsrk, C., 1967, Pesticide residues, possible Interference by chlori Breidenbach, A. \Y., and others, 1961, Tlic identification and measure ment of chlorinated hydrocarbon pesticides in Airfare waters: U.S. fuWir llrallh Service Puli. 1211, TO p. nated biphenyls, in Procrrdings of the IUPAC Commission nu the Development, Improvement, and Standardization o! Methods of Pesticide Itrsiduc Analysis: Assoc. Official Anal. Cheiniats Jour., v. SO, no. S, p. 1069. DSW 026633 STLCOPCB4010594