since the late 1970s or earlier. This 

 point needs to be substantiated with 

 more data from sediment cores. The 

 data in Figure 3 were chosen because 

 the chronologies extend to at least 1 980. 

 There are many reports where the chro- 

 nologies end in the 1 970s. In these latter 

 cases, contaminant concentrations that 

 had been increasing since the industrial 

 revolution appear to have stopped in- 

 creasing, but evidence of decreases is 

 lacking. The reason for studying new 

 cores is to verify or refute the hypothesis 

 that contamination has been decreas- 

 ing over the last decade. 



Temporal Trends of 

 Contaminants in 

 Molluscan Tissue 



Except underspecial circumstances, one 

 cannot expect samples of surface sedi- 

 ment collected at a single site to differ 

 on an annual basis even if contaminant 

 inputs change. Mussels and oysters, on 

 the other hand, can change their con- 

 taminant levels in response to changes 

 in their surroundings (Roesijadi et al., 

 1984; Pruell et al., 1987). This and the 

 fact that they are immobile makes them 

 ideal for monitoring changes in chemi- 

 cal concentrations in the coastal envi- 

 ronment. 



Figure 4 is an 1 8-year record of tPCB in 

 mussels at the NS&T site in Royal Palms 

 Park on the Palos Verdes coast of Los 

 Angeles. It shows a dramatic decrease 

 that began in 1 971 when the U.S. began 

 to phase out PCB use. The record has 

 been constructed by combining three 

 sets of data from three separate pro- 

 grams. Sericano et al. (1990) have 

 combined data from diverse sources to 

 show similarly dramatic historical 

 changes in the average tDDT concen- 



tration in oysters in the Gulf of Mexico. 

 In that case as well as in Figure 4, the 

 major decreases occurred in the early 

 1 970s. Unlike trends found in sediment 

 cores, where all sections would have 

 been analyzed by a single laboratory, 

 trends from annual collections of mol- 

 lusks have had to be based on data from 

 several sources. As explained by Stout 

 (1986), analytical artifacts must be sus- 

 pected because chemical techniques 

 for quantifying organic compounds in 

 environmental samples have improved 

 dramatically overthe same period of the 

 apparent decreases. Nevertheless, the 

 timing and the large magnitude of the 

 decreases lends credibility to the argu- 

 ment that major decreases have oc- 

 curred in concentrations of now-banned 

 chlorinated compounds. 



Decadal trends in trace metal contami- 

 nation have been sought by comparing 

 NS&T data of 1986 through 1988 with 

 data from analyses of mussels and 

 oysters collected in 1 976 through 1 978 

 by a previous "mussel watch" program 

 (Goldberg et al., 1983) sponsored by 

 the U.S. Environmental Protection 

 Agency (EPA). Statistically, since the 

 earlierprogramcollected a single sample 

 each year, it was necessary to aggre- 

 gate the three years of data for each 

 decade. With that aggregated data it 

 was possible to estimate differences in 

 trace metal concentrations in mollusks 

 at the 50 sites that were common to both 

 programs. Figure 5 shows the 50 site 

 locations and demonstrates the over- 

 whelming (39 out of 50) dominance of 

 decadal decreases in lead. That excess 

 of decreases over increases or lack of 

 difference is itself statistically sufficient 

 to declare a national decrease in lead 

 concentrations since the late 1970s, a 

 result consistent with the phase-out of 



14 



