Assessing sediment toxicity and arsenite concentration with bacterial and traditional methods.

Three sediment samples LP (pool where logs are stored), LF (brook through landfill area), KN (Kaskesniemi) which is in Lake Pyhäselkä downstream from the mill, were taken from an old sawmill area and one from the unpolluted Lake Höytiäinen. The arsenite concentration was measured by GFAAS and two arsenite biosensing bacterial strains Pseudomonas fluorescens OS8 (pTPT31) and Escherichia coli MC1061 (pTOO31). The toxicity of sediment and pore water samples was determined by using luminescent bacteria (Flash test) and, further, whole sediment toxicity was measured using 10 days growth test and 50 days emergency test with midges (Chironomus riparius). With the flash test a lowered EC50 value was found only in sediment LF (EC50=0.17 v/v%). The Flash test indicated that all sediment samples taken from the sawmill area were highly toxic to bacteria, whereas growth and the emergence of chironomids showed no effects in other samples than LF. The midges tolerate well the contaminated environment. In contrast, bioavailability of arsenite of sediment samples KN and LF was quite high determined using the biosensor-strains in a direct contact assay. The bioavailable fraction of sediment LP was 6-10% out of the total arsenite concentration obtained with GFAAS (0.46-0.77 microg g-1 dw). The results show that the choice of analysis method grossly affects the outcome without any of the method giving an incorrect result. Different methods measure different parameters of a toxic sample and can thus be used to complement each other.