7170 - Laserfiche WebLink

Home Browse Search

I tion of water samples are also commercially avail- able. Use of these ampules reduces acid leaks in sampling kits. The on-site circumstances should indicate where and how many samples should be taken. As a minimum, samples should always be taken outside and inside the kill area. The control or reference site (outside the kill area) should always be free from the influence of the suspected toxic water. In a stream, one sample should always be taken above the kill area or above any point source potentially associated with the kill. If involvement of an effluent discharge is suspected, a sample of the effluent should be collected, as well as water samples collected down- stream from the outfall (Figs. 4.1 and 4.2), to delineate the contaminated zone. For streams more than about 60 m wide, samples should be taken at two or more points along a transect across the stream. In large streams, it may also be necessary to take samples at various depths. Sampling devices that can be used to take water samples are outlined in Chapter 12; others are given by EPA (1982), Hill (1983), and APHA et al. (1985). To determine if benthic or- ganisms have been killed, samples of the bottom materials must be collected with equipment such as this Ponar dredge. TOXIC SUBSTANCES 37 Sediment Samples It may not be necessary to collect sediment in all fish kill investigations. However, samples should be consistently taken from the same sites where water samples were taken (above, within, and below the kill area). Special sampling sites below point source inputs may be desirable and should always be carefully documented. The method of handling the samples after collection and before analysis is deter- mined by the type of test to be run. Samples should always be kept cool (4° C) or frozen and stored at -20° C or lower (EPA 1982; Palmer 1984; Tetra Tech 1986). If samples are to be used in toxicity tests, they should always be kept cool (4° C), but never frozen (M. K. Nelson, National Fisheries Contaminant Research Center, Columbia, Missouri, personal communication). Sediments are usually taken with a corer or me- chanical grab dredge (EPA 1982; Palmer 1984; Tetra Tech 1986). The needed sample size is usually not less than 50 g (Table 4.6). One-quart wide- mouthed glass jars with screw cap lids are accept- I