|
<br />I
<br />I
<br />I
<br />I
<br />I
<br />I
<br />I
<br />I
<br />I
<br />I
<br />I
<br />I
<br />I
<br />I
<br />I
<br />I
<br />I
<br />I
<br />I
<br />
<br />Aquatic macroinvertebrates collected from the six sites and retained on the standard
<br />#40 stainless steel sieve (~0.425 mm) were separated by order. These samples and a
<br />portion of the zooplankton that passed through the sieve (< 0.425 mm), along with some
<br />tadpoles collected in light traps. were stored frozen and analyzed for selenium concentration
<br />by atomic absorption. Prior to analysis. samples were freeze dried and the results reported
<br />on a dry weight basis. The limit of detection was 0.15 j-tg/g, the procedure blanks had
<br />background equivalent concentrations less than the LOD, the percent relative standard
<br />deviation (triplicate sample preparation and analysis) was 3.7%, and recovery of selenium
<br />was within recommended ranges in NISI standard reference material 3149 (8.99-11.01 j-tg/L)
<br />and National Research Council of Canada (NRCC) reference material DORM-1 (dogfish
<br />muscle)(1.35-1.91 j-tg/g). The digested sample spike solutions had a mean percent recovery
<br />of 100%, and spiked samples analyzed for matrix suppression or enhancement had mean
<br />selenium recoveries of 98 % .
<br />
<br />A portion of the zooplankton that passed through the sieve ( < 0.425 mm) were
<br />collected, stored frozen, and analyzed for 31 elements by ICAP. Prior to analysis, samples
<br />were freeze dried and the results reported on a dry weight basis. The procedure blank had
<br />background equivalent concentrations less than the LOD for all elements except Ag, Ca, and
<br />Zn in one blank and Ca, Co, Fe, Mg, Sr, and Zn in a second blank (concentrations of these
<br />elements exceeded their LOD by very little), the mean percent relative standard deviation
<br />(duplicate sample preparation and analysis) was 5.3, the mean spike recovery was 90 %, and
<br />the recovery of trace elements in NBS reference material 1566A (oyster tissue) was within
<br />the recommended range. The estimated limit of detections [j-tg/g; mean (standard error)]
<br />were: Ag 0.9 (0.3), Al5 (1), As 10 (3), B 1.4 (0.4), Ba 0.13 (0.03), Be 0.07 (0.02), Bi 10
<br />(3), Ca 2.7 (0.7), Cd 0.07 (0.2), Co 1.4 (0.4), Cr 3.8 (1.0), Cu 0.7 (0.2), Fe 1.4 (0.4), K
<br />65 (17), Li 0.5 (0.1), Mg 0.6 (0.2), Mn 0.16 (0.04), Mo 1.1 (0.3), Na 5 (1), Ni 3.7 (1.0), P
<br />16 (4), Pb 7 (2), Sb 6 (2), Se 9 (2), Si 29 (7), Sn 5 (1), Sr 0.11 (0.03), Ti 0.30 (0.07), TI
<br />16 (4), V 0.9 (0.3), Zn 0.30 (0.07).
<br />
<br />Live fish at the end of studies 2. 3, and 4 were collected for analysis of selenium
<br />concentrations by neutron activation at the University of Missouri, Columbia. MO, and are
<br />reported on a dry weight basis. No fish were available from study 1. Prior to analysis,
<br />samples were freeze dried. The limit of detection was 15 ng/g, the percent relative standard
<br />deviation (triplicate sample preparation and analysis) was 3.1 %, and recovery of selenium in
<br />NIST standard reference material 1577 (bovine liver) was within the recommended range
<br />(3.5 i::0.4 j-tg/g).
<br />
<br />Statistical analyses
<br />
<br />Toxicant effects on percent survival (arcsine-transformed values), growth, and
<br />whole-body concentrations of inorganics in fish (logarithmically transformed values) were
<br />evaluated by one-way analysis of variance. Food density effects in the supplemental feeding
<br />study on percent survival (arcsine-transformed values) and growth in fish were evaluated by
<br />one-way analysis of variance. Treatment means were compared with the least-significant-
<br />
<br />16
<br />
|