of fats, sulfuric acid was used as the final step of cleanup just 

 prior to injection of the samples into the GC (Patton et al., 

 1989). The florisil and silica gel absorbent materials had to be 

 extensively cleaned prior to use in order to reduce their blank 

 contributions to acceptable levels. These cleanup steps involved 

 ignition of the florisil overnight in a muffle furnace at 600°C 

 and batch extraction by Soxhlet of the silica gel with pesticide 

 grade petroleum ether. 



Extracts were analyzed by electron capture detection 

 using a Hewlett-Packard model 5890 GC equipped with an 

 autosampler. All data were processed using Nelson 

 chromatography software. The GC separation was completed 

 on a J&W DB-1701 (J&W Scientific, Folsom, California) 

 megabore fused silica capillary column, 30 m x 0.53 mm ID. 

 The carrier gas was H, at a flow rate of 20 ml/min. The heated 

 zones were 250°C for the injector and 325°C for the detector. 

 The GC program was as follows: 120°C hold for 1 min; 

 increase to 160°C at 20°C/min; then programmed at 2°C/min 

 up to 225°C; and finally held for 5 min. Quantitation was 

 performed using the external standard method. Standard peak 

 area integration was used for all single component pesticides; 

 however, peak heights were used for PCB quantitation. The 

 heights were summed for all peaks that matched the retention 

 limes of Aroclor 1242, Aroclor 1254. or Aroclor 1260. 



Each extract collected from a silica gel fraction was 

 prepared for injection into the GC by volume reduction to about 

 1 ml via a Kudema Danish concentrator assembly, shaking 

 with 1 ml of concentrated sulfuric acid for 30 s, and reduction 

 of the acid-treated extract to 0.5 ml by blowdown with N, gas. 

 With the sediment extracts, it was necessary to add 2-3 drops 

 of elemental mercury to each 0.5 ml of final extract, and mix it 

 until all elemental sulfur was removed. 



Samples selected for toxaphene analysis were not treated 

 with sulfuric acid in order to preserve structural integrity of the 

 toxaphene components. Also, the fractionated silica gel extracts 

 were recombined to allow for maximum recovery of any 

 toxaphene that might have partially split into separate fractions. 

 In preparation for injection these extracts were reduced by N, 

 blowdown to 0.2 ml and injected into the Varian GC attached 

 to a Finnigan TSQ-70 mass spectrometer operated in the 

 negative chemical ionization mode. The mass spectrometer 

 was set up according to the procedures of Swackhammer and 

 associates (Swackhammer ef a/. , 1 987 ) except for modifications 

 to the choice of the GC column (DB-1, 30 x 0.32 mm ID— 

 1 |i film thickness), minor changes to the GC operating 

 conditions, and use of the external standard method for 

 quantitation. For qualitative verification of toxaphene in the 

 samples, the retention times of the peaks of each of the 

 characteristic mass chromatograms were compared to similar 

 chromatograms of the standards. The capillary column was 

 able to resolve 68 peaks in the standard. The quantitation 

 routine in this method allows for both retention-time-matched 

 and nonretention-time-matched peaks to be included if they are 

 present in the appropriate mass ranges; therefore, there was 

 always plenty of integrated area above the blank levels to be 

 used in the quantitations. Because of the selectivity of this 

 program for identifying only toxaphene peaks, the toxaphene 

 results should be considered very reliable. 



Procedural blanks were carried through each analytical 

 step to correct for background interterences from reagent 

 contamination and handling. Samples were processed in 

 batches of 10 to 20, with at least one matrix spike, duplicate, 

 and blank to monitor the accuracy and precision of the analysis 

 for each batch. To report a residue as detectable, the raw 

 number had to be at least twice the blank. 



Recoveries were monitored by carrying out matrix spikes 

 with mixtures of the expected pesticides and Aroclor mixtures. 

 The levels for spiking ranged from 2.5 to 12.5 ng/g for the 

 organochlorine pesticides and 12.5 to 125 ng/g for the PCB's. 

 For the biota the average recoveries were as follows; Total 

 PCB— 49%; p,p'-DDE [2,2-Bis(p-chlorophenyl)- 1, 

 1-dichloroethylene]— 86%; HCB— 73%; alpha-HCH— 95%; 

 gamma-HCH — 81%';oxychlordane — 77%;trans-chlordane — 

 88%; cis-chlordane— 88.6%; trans-nonachlor— 78%; 

 p,p'-DDD [2,2-Bis(p-chlorophenyl)-l,I-dichloroethane] — 

 88.4%; cis-nonachlor— 76%; and p,p'-DDT 

 [2,2-Bis(p-chlorophenyl)- 1,1,1 -trichloroethane] — 65%. For 

 the sediment, the recoveries were all low but consistent as 

 follows: Total PCB— 39%; p,p'-DDE— 57%; HCB— +6%; 

 alpha-HCH — 45%; gamma-HCH — 44%; oxychlordane— 

 44%; trans-chlordane — 40%; cis-chlordane — 42%; 

 trans-nonachlor — 41%; p,p'-DDD — 55%; cis-nonachlor — 

 42%; and p,p"-DDT— 36%. 



Duplicate results were collected for each batch of 20 or 

 less samples. Generally there was good agreement between the 

 two values, with the relative percent differences averaging less 

 than 25%. 



Both electron impact and negative chemical ionization 

 mass spectrometry were employed to confirm the residues 

 identified by electron capture GC. In some cases, residues 

 were identified even though they were below the detection 

 limits of the electron capture methods. 



Results and Discussion 



Organochlorine residues were present in all of the samples 

 analyzed (Tables 1,2,3). The highest single component OC 

 measured was the HCH class of compounds, especially 

 alpha-HCH at 8. 1 2 ng/g in one of the bivalve samples. In the 

 mixed OC component classes of compounds, PCB's and 

 toxaphene comprised the highest residues. For example, in 

 neuston there was 67.9 ng/g total PCB's (Station 22); in 

 zooplankton 23.9 ng/g total PCB's (Station 11 3, Table l);and 

 in pollack 10.8 ng/g toxaphene (Station 4, Table 3). Of the 

 single-component organochlorine pesticides other than HCH' s, 

 HCB was relatively high in some of the crabs and bivalves, and 

 trans-nonachlor and p,p'-DDD were generally high in fish, 

 zooplankton. and phytoplankton. The sediment was notably 

 devoid of most of the OC's found in biota (Table 2). 

 Alpha-HCH was found in a few sediment samples, detectable 

 levels of DDT and PCB's were measured in sediment surface 

 layer (Station 45), and some chlordane peaks were evident in 

 the deeperhomogenized sample (Station 13). Our data suggests 

 that sediments are not reservoirs supplying organochlorines to 

 the biota but rather acting as sinks. Further, it appears that the 

 atmosphere is the major loading factor in this system. 



286 



