of mussel larvae in the present study exceeded 90 percent in both seawater and sediment 

 controls and was less than 75 percent in the OA site sediments. Therefore, in comparison 

 with some samples from highly contaminated waterways elsewhere in San Francisco Bay 

 and in parts of Puget Sound, the samples tested in the present study were not as toxic to 

 mussel larvae. 



Mean survival of oyster larvae was less than 50 percent in 8 of 10 (80%) samples from 

 Bellingham Bay, 6 of 6 (100%) samples from Everett Inner Harbor, 4 of 4 (100%) samples 

 from the Duwamish Waterway, and 8 of 9 (89%) samples from the Commencement Bay 

 waterways (Long, 1985). Mean survival of mussel larvae was less than 15 percent in samples 

 from the Islais Waterway, 24 to 50 percent in samples collected off the Alameda Naval Air 

 Station and 50 to 83 percent in samples from the SP site tested in 1985 (Chapman et al, 

 1987). By comparison, 76 percent of the mussel larvae survived exposure to sediment controls 

 in the present experiment. As few as 29 percent survived in tests of sediments from the OA 

 site. Again, the survival end-point of this test indicated a degree of response that was 

 somewhat less than that observed in other studies of very highly contaminated sediments. 



The adult mussels can be induced to spawn year-round as exemplified in this study. The 

 tests were performed in February. However, several investigators have experienced 

 problems, especially with oysters, in attempting to induce bivalves to spawn in the winter 

 (Joe Cummins, US EPA; Paul Dinnel, University of Washington; Peter Chapman, E.V.S. 

 Consultants, personal communications). 



The high sensitivity of the M. edulis larvae test observed in this evaluation was also 

 reported by Chapman et al. (1987) in a previous study in San Francisco Bay in which six 

 toxicity end-points were measured. In that study, the end-points of percent abnormal 

 development and percent survival of M. edulis indicated that the most samples, 5 of 9 and 8 

 of 9, respectively, were "toxic" relative to controls. However, Williams et al. (1986) 

 reported that a similar test performed with oyster larvae (Crassostrea gigas) was the least 

 sensitive of three that were evaluated. It indicated 35 percent of the samples from 

 Commencement Bay waterways were "toxic" relative to controls, compared to 39 percent for 

 R. abronius and 63 percent for a Microtox™ test. In the data reported by Chapman et al. 

 (1987), the analytical precision was somewhat lower (average CVs were 25% and 32.5%, 

 respectively, for percent abnormal and percent survival) than in the present evaluation 

 (average CVs of 3.9 and 22.6%, respectively). The positive correlations between the results 

 of this test and the texture and TOC content of the sediments have not been quantified 

 through empirical experimentation. However, similar to the results observed here, 

 Chapman et al. (1987) often observed highest toxicity in samples that had the highest 

 percent fines and TOC content, as well as the highest concentrations of toxicants. In this 

 evaluation, fine-grained, organically enriched samples that were apparently not 

 contaminated with the quantified analytes were also toxic to this organism. In conclusion, 

 although the end-point of abnormal development of this test was the most sensitive and had 

 the highest precision and discriminatory power of the five evaluated, it also may be 

 sensitive to "nuisance variables" (Bayne et al., 1988) such as sedimentological properties 

 and/or it may be a sensitive indicator of the toxicity of sediments contaminated with 

 chemicals that are not routinely quantified in chemical analyses. 



Similar to the results reported here, the data from the M. edulis test and the R. abronius 

 test previously have indicated very high concordance with each other (Williams et al., 

 1986; Chapman et al., 1987). 



Rhepoxynius abronius. This test has been developed and evaluated through extensive 

 research (e.g., Swartz et al., 1985). The life history and sensitivity of the animal to many 

 toxicants and sediment types has been described (e.g., Swartz et al., 1985). The species 

 normally burrows in surface sediments and, therefore, is appropriate for use in solid phase 

 tests. Populations in Yaquina Bay, Oregon and off Whidbey Island, Washington have been 

 sampled year-round for use in toxicity tests. The analytical precision of the test has been 

 quantified and compared among five laboratories (Mearns et ah, 1986). 



80 



