Toxicity of Santander Bay sediments to the euryhaline freshwater oligochaete Limnodrilus hoffmeisteri

Abstract

The freshwater euryhaline oligochaete Limnodrilus hoffmeisteri was selected for use in bioassays with polluted sediment from Santander Bay. It is easy to culture; is tolerant of low to moderate, up to 15%, salinity; and is common in oligohaline conditions in European and North American estuaries. Worms were collected from an estuarine population and kept in unpolluted sediment for between 2 and 4 weeks, under laboratory conditions, at 7–8.5% salinity and 22.5 °C. Sediment from different sites in Santander Bay were sieved through 250 Ŝsmm mesh and adjusted to a salinity of approx. 7% prior to the bioassays, either by adding sea water or dechlorinated tap water to the overlying water. High levels of ammonia in some sediment, which may confound results in ecotoxicity bioassays, were reduced by oxidation of the sediments in shallow trays. Sediment bioassays were performed with sexually mature Limnodrilus hoffmeisteri worms in 250-ml beakers, with a 1:3 ratio of sediment:water and 4 worms per baker. Endpoints in the 14-day bioassay were%mortality, adult final biomass,%adults that have shown resorption of the clitellum, number of cocoons, and burrowing behaviour. It was possible to rank the sites according to their toxicity using both mortality rates and sublethal effects. The control site had the following values for the endpoints: 5% mortality, ($\bar x \pm sd$) 2.40Ŝ±1.52 cocoons per beaker and 1.271Ŝ±0.470 mg dw adult final biomass. The most toxic sediment resulted in 65% mortality, resorption of the clitellum in 67% of the adults, no production of cocoons and a low final biomass ($\bar x \pm sd = 0.681 \pm 0.489$ mg dw per adult). A second site had high mortality (60%) and no reproduction, although resorption of the clitellum did not occur in surviving animals. The remaining sites showed similar mortality (35–42%), and at only one of them was low reproduction observed (0.8Ŝ±0.447 cocoons per beaker). Behavioural effects, measured as length of galleries in a fixed area of the test-vessel at the end of the bioassay, were significant compared with control at only one site. Multivariate analysis showed themortality gradient to be the strongest, with a second unassociated gradient representing clitellum resorption. The mortality gradient was associated with Cu and Zn concentration, and PAH and Pb possibly with resorption.