Sample cent stic agoni of total PCB oc or OR. Paj P see a concentrations hor jar ed h zed 1 ios les and ole f ater th ubstantially re liver S The poss thatt int vel he 1 on o. souss rage is t in ed mono, di, and t pre ls mthe for the five species ta slight inverse s fo nrease els PCB'S ont moved u his es r wer the dc from Ell Sloug and DD reater ve significantly C DUES IN I PCB and DDT ISHES OF ELKHORN SLOUGH AND THE PAJARO RIVER, CALIFORNIA By W. D. Koenig Hopkins Marine Station of Stanford University Pacific Grove, California Since 1967, when polychlorinated biphenyls (POB's were first identified in animal residues (1,2), there increasing concern about PCB concentrations in has bee the environment. Efforts to measure these concentrations have been considerably complicated by the complex nature of POB's, which are not a single group of analogs but a series of isomers with varying degrees of chlorination Until recently, quantification of POB's using gas liquid chromatography has been made by assuming identical response of the electron-capture detector for all isomers and using either p,p'-DDE (3) or commercial PCB mixtures (Aroclors) as standards (4,5). Rote and Murphy (6), showed that the detector response varies according to the degree of chlorination of the individual isomer. By semi-logarithmically plotting the total ine/ng from the chromatograms of six different Aroclors against their average chlorine content they formulated a response curve from which the absolute values of specific POB isomers can be estimated. Using this technique, POB 214 levels were determined in fishes from Elkhorn Slough and the Pajaro River, California. DDE, DDD, and DDT residues were also quantified and their combined values compared to the total POB content. METHODS Elkhorn Slough and the Pajaro River both drain agricultural lands in the Monterey Bay area. They empty into Monterey Bay near Moss Landing (Fig. 1) and are both susceptable to considerable salt water intrusion from the bay. Staghorn sculpin (Leptocottus armatus), shiner perch (Cymatogaster aggregata), and leopard sharks (Tr akis semifasciata) were seined and gill netted in Elkhorn Slough at Kirby Park (A). A starry flounder (Platichthys stellatus) was obtained from near the Moss Landing marina (B). All fishes from the Pajaro River including top smelt (Atherinop: finis), and striped bass (Roccus saxatilis) were gill netted about one mile from the mough (C). Samples were either dissected soon after capture or frozen and stored until ready for analysis. Three liver and two flesh samples were analyzed from the leopard sharks; all other samples were single preparations of flesh or whole fish. Tissues were digested in a perchloric-glacial acetic acid mixture, extracted with hexane, and evaporated to dryness. The fat residue was weighed and then cleaned by acetonitrile partitioning followed by elution with 6% diethyl ether through a florisil column (7). The resulting mixture of pesticides and PCB's was evaluated by gas liquid chromatography for the approximate DDE content and an aliquot containing no more than 20 ug was concentrated down to 5 ml. for separation. This involved elution by petroleum ether through a column packed with specially treated silicic acid following the method of Armour and Burke (8). POB residues are eluted with the first 250 ml. of runoff DDE and other pesticide residues remain in the column and can later be recovered using a methylene chloride-hexane¬ acetonitrile mixture (80-19-1). As some difficulty in separation was frequently encountered, the last 30 ml. of petroleum ether eluate was collected in 10 ml. fractions and tested for early DDE runoff before it was added back to the original 220 ml. All eluates were concentrated to appropriate levels for analysis by gas liquid chromatography employing identical equipment and operating conditions used by Rote and Murphy (6). Peak area was measured by an automatic disc integrator and quantitated according to detector response. In order to adjust for variations in machine sensitivity, standards of Aroclor 1232 and 1260 were injected soon before or after the samples. The total peak area in in/ng from these two standards was plotted semi-logarithmically against 2/6 O the average number of chlorine atoms per molecule in the mixture. This average is known from the Aroclor number: 1232 contains 32% chlorine by weight or an average of two chlorines per molecule; 1260 contains 60% chlorine by weight or an average of six chlorines per molecule. A straight line was drawn through these two points from which detector response levels for any given isomer with a known number of chlorine atoms could be determined. Chlorine content of specific isomers was determined whenever possible by comparison with the mass spectometry work of other investigators (4,9,10). Peaks not previously identified were assumed to contain the same number of chlorines as peaks having approximately the same retention time. Each isomer found in a chromatogram was quantified individually and summed to obtain a value for total PCB content. RESULTS A total of thirty-nine different peaks were resolved in ifteen samples. Thirty-two of these correspond to those identified by Rote and Murphy in their analysis of Aroclor standards and were numbered identically 1 to 32 (6). Seven additional peaks were detected and designated by the number of the peak preceding it plus a letter. Five of these peaks (7A, 9A, 19A, 27A, 27B) had retention times between two POB compounds with equal numbers of chlorine atoms and were evaluated by using the response curve value appropriate for the two surrounding isomers. Peak 31A was located 2/ between a nonachlorobiphenyl and a decachlorobiphenyl. Since only one decachlorobiphenyl isomer is possible it was evaluated as having nine chlorine atoms. The degree of chlorination of the compound corresponding to peak 32A could not be inferred from its retention time and was not included in total POB content. Retention times relative to p,p'-DDE and the known or approximate degree of chlorination for each isomer are given in Table I. Results calculated in total PCB and total DDT residue content in both fat and whole tissue are listed in Table II. Fig. 2 compares the residue levels (ppm whole tissue weight) found in the three species of fish that were sampled from both rivers. Both POB and DDT levels strongly suggest that fishes in the Pajaro are concentrating more polychlorinated residues than the same fishes in Elkhorn Slough, though this data become somewhat confused depending on whether values referring to ppm of lipid or total tissue weight are examined. A rough correlation of residues in whole tissue weight to the approximate trophic level on which the fishes from the Pajaro River feed (11) is shown in Fig. 3. Except for the high values found in the top smelt, a primary carnivore the total PCB content shows a distinct upward trend while the DDT values exhibit only minimal correlation. After individual quantification, the POB isomers were divided into ten groups according to their degree of /8 chlorination. In order to judge the relative amounts of each group accounting for the total POB's, each is listed in Table III in percentage of total POB content. Further analysis, made by adding up the percentages of mono, di, and trichlorobiphenyls present in the sample fishes from the Pajaro River show the slight inverse relationship to the idealized trophic levels of the fishes shown in Fig. 4. Samples run on shiner perch roe from Elkhorn Slough and on top smelt eggs from the Pajaro River show opposite trends for the initial residue levels passed on by these species to their young. Both POB's and DDT were concentrated over 2-fold in the shiner perch roe compared to the levels found in the adult flesh while both were present in only one-half the concentration of the adult flesh in top smelt eggs. Particularly interesting are the results on the three leopard shark livers and two flesh samples. All livers show higher levels of DDT than POB's with a ratio averaging 2.91. In contrast, flesh samples from the same sharks show DDT/POB ratios approximating those values found in all other samples where either whole fish or flesh was analyzed, Absolute values for DDT and POB residues (ppm in lipid) in the shark livers, shark flesh, and the average of the three other flesh samples from Elkhorn Slough are shown in Fig. 5. The breakdown of the total DDT residue levels into DDE, DDD, and DDT percentages is found in Table IV. 219 DISCUSSION Holmes (2), Risebrough ét al (12), Koeman et al (4), and Rote and Murphy (6) all found the higher chlorinated POB's to be predominant in organisms high up in food chains such as fish, sea birds, and a sea otter. Jensen et al (13) and Koeman et al (4) both present evidence which indicates that there is significant breakdown of the lower chlorinated PCB's as they pass through biological systems. Though lower chlorinated POB's were found in all samples the trend shown in Fig. 4 indicates breakdown among these compounds relative to the total POB content as one goes from the primary carnivore (top smelt) to the tertiary carnivores (striped bass and starry flounder). This finding substantiates the evidence of Jensen and of Koeman. DDT/POB ratios have been studied extensively by Risebrough et al (14). Their findings showed a regional fallout pattern in the DDT/PCB ratios with low values between one and two being found in birds around San Francisco Bay and increasingly higher ratios being found in birds ogressively further from industrialization. DDT/POB living pr ratios from.4 to .8 have been recorded by Presst and Jeffries in great crested grebes from the British Isles (15). In marine animals along the west coast of Sweden the DDT/POB ratios were found to be as low as .15 (13). From these data Peakall and Lincer (16) suggest that POB's 206 C are not as easily transported to remote areas as are DDI residues and thus the DDT/PCB ratio tends to be lower the closer one comes to industrial centers. Findings reported here do not contirm this hypothesis. Both the Pajaro River and Elkhorn Slough are near relatively little yet DDT/POB ratios were is general found to be industr as low or lower than values reported from industrialized areas. In contrast to these findings are the DDT/PCB ratios found in the liver samples of the leopard sharks. All three samples have ratios over 1.0 with a mean of 2.91. T/POB ratios in the flesh of the same specimens, however, were comparable to those found in the other tissue samples. As an explanation for this phenomenon, perhaps the exceptionally high levels of DDT residues concentrated in the liver tissue affect POB storage such that the DDT/PCB ratio becomes considerably higher than would otherwise be expected. This view is substantiated by the absolute residue concentrations found in the lipid of these samples which, because of the high fat content of liver tissue, gives a better relative measure of the residue levels. Using these values, both PCB and DDT levels in the shark flesh are seen to be similar to the levels found in the flesh samples from the other fishes taken from Elkhorn Slough (Fig. 5). In the liver tissue, however, DDT is far more concentrated than in any of the other fish samples O while POB content is slightly below those levels found in either the tissue samples from the same sharks or any of the other tissue samples from the Elkhorn Slough fishes. A similar effect on dieldrin storage in fish and rats has been shown to be caused by the presence of DDT (17,18). SUMMAR Samples of fishes from the Pajaro River and Elkhorn Slough were analyzed for chlorinated hydrocarbon residues. All DDT/POB ratios for flesh and whole fish samples were less than 1.2 ; ratios were substantially greater than 1 in leopard shark liver samples. The possibility that the high concentration of DDT stored in the liver tissue is ic to POB storage is discussed. The percentage antagonist: of total POB content present in combined mono, di, and trichlorobiphenyls for the five species taken from the Pajaro River was found to exhibit a slight inverse relationship to the idealized trophic levels on which these fishes feed in accordance with findings by Jensen et al (13) and Koeman et al (4) that breakdown occurs in the lower chlorinated PCB's as they pass through biological system POB's more than DDT levels tended to increase in these same five fishes as one moved up this simplified fraction of a food chain. Samples from the Pajaro River were compared to the same species captured in Elkhorn Slough and found to have significantly greater PCB and DDT concentrations. 23 C 10 ACKNOWLEDGEMENTS I would like to thank James Rote and Philip Murphy for providing invaluable aid throughout all phases of this project. REFERENCE 1. Widmark, G. 1967. Possible interference by chlorinated fic. Anal. Chem. 50: 1069. biphenyls. J. Assoc. C 2. Holmes, D. C., J. H. Simmons, and J. 0. Tatton. 1967. Chlorinated hydrocarbons in British wildlife. Nature 216: 227-229. Risebrough, R. W. 1969. Chlorinated hydrocarbons in the 3. lout, G. G. Berg and global ecosystem. In: Chemical E M. W. Miller (Eds.), Charles C. Thomas, Springfield, Ill., p. 5-23. 4. Koeman, J. H., M. C. Ten Noever de Brauw, and R. H. de Vos, 1969. Chlorinated biphenyls in fish, mussels and birds from the River Rhine and the Netherlands coastal : 1126-1128. area. Nature2 71. Pesticide residue analysis in the 5. Reynolds, L. M. 1 presence of Polychlorobiphenyls (POB's). Residue Reviews 34: 27-5 6. Rote, James W. and Philip G. Murphy. A Method for the quantitation of polychlorinated biphenyl (POB) isomers. In press. 7. Pesticide Analytical Manual Vol. III, Food and Drug Administration, Washington, D. C. 1970. 223 10. and J. A. Arm It from DDT Bipheny Separat Polyc 61- I m. Anal. Analog ssoc. Of ic. mar 197 , W. L two ion aphy-mass 3-1 age yo inedg 1. Chem. rom fic. 1. bery Wall ios, ancis J. C enyls in human adipose tissue. aed Poly hlor 201. 5: 317-3 . Cont ishes of Califor. Baxter, John L. 1966. In De t. of Fish and State of H. lott. e, and brou Reid O ati p. i the dete Tox 91. 4 nyls. m. T nviron. Con Ott Jensen, S., A. G. Johnels, S. om Swedi ine animal 1969. T and P : 247. Watu . Peakall, Reiche, iseb lorinated biphen; Kirven. 1968 an 1.. 29: 100 osy Natur inter nu ffries. 196 d D. esst, t lorine residu Sird St 168 rebe ritai 70. nat ate the at cresed eding - r, F Toxic 300-31 adipose tissue of Lince inated pread chemical i fe wides the 20: d. I nd J. w trout. Bull. Cont. drin long Tig Fig. GENL JURE llectio owing e Pajaro al who 1s PCB om both Elkhor sf d in sue iver. the Slough ole levels (ppm res and P DDT the Pajaro Ri ihes tro vels. hic wi combine idue O, o River in comparison with thei evel levels ( Comparis of) e of three se shark 1 opa sh erage of one sample each) shes( e oth Elkhorn Slough. of fror SANTACRUT A Fig. MONTEREY BAY MONTEREY MILES PAJARO RIVER ELKHORN SLOUGH M08 LANDING SALINA! RIVER V 22 ppm Flesh P( DO Sculpin DO Perch PCE DDI Flounder Pajaro Elkhorn Fale 228 ppm issue Pajaro River PCBIDOT PCBIDOT ABIDDT Perch Smelt Sculpin oCarn. 2° Carn. ABIDOT Bass PCBIDOT Flounder arn. Fe a 20 Figure ajaro River Smelt Sculpin Perch Bass Flounder 3°Carn. PCarn 2°Carn. PCBs in mono-trichlorobiphenyls % Total 230 90 70 ppm ipd 30 DDT PCB Shark liver N=3 DDT PCB Shark flesh N-2 Elkhorn Slough DDT PCB Sculpin Flounder Perch N=3 Fioure 3 able able I Table able aks ue) Per and DDT LEG levels ( DDT/PCB r Sloug atom who ver. nfishes f mElkhor Sontent referring to DDE. DDD idu 230 NO. CHLO 2(P 1.00 11 3 8 1.06 12 1.2 1.4 1.7 2.13 2.2 2.7 3.16 3.62 4. 8.0 E 9 8 aaaa- * * . . * — .. k- 8 a- 6 — — - akaaaa- oooooooo OO 1 2 ataaaaaaa- + 1 kaa- aavaaa- ataa- a- ooooaoo o — — . —-e taaaaaaaaaaa- oo- 8 atakaakakaakakaaaa- 3 5 0 0 + 8 7 a- 8 29 234 + 0 1 1. O + oo o o O - oo o o aka- tataa- o o ataataa- atatatataaa- Oo atatataataakaataoaa- — O O VO katatatakataaa- — — - O oo o o 10 0 ooo ataa- atakakaaa- tataaa- 23 IPT SAMLE Staghorn s lpin lounder lounder ner per perch iner per roe (egg Leo Livel esh Flesk Pajar Paj or Paja hor Elkhorr QDDE 0 76 e se Poor ser atio. atio atio orssue 18 639 188(DDE on 302 L.O E only 3(D 61. 2.1 18.56 081(1 091