8181 - Laserfiche WebLink

Home Browse Search

SELENIUM IN RAZORBACK SeCKERS IN GREE;-..i RIVER. LT-\H 135 scan across the mass spectral range for 61 elements. Accu- racy of the semiquantitative scan typically has j: 30 to j: 50% lower detection limits than regular ICP analysis, depending on analyte mass. Elemental response factors were adjusted prior to analysis with a certified reference solution (High-Purity Trace Metals in Drinking Water; High Purity Standards, Charleston, sq. This solution was analyzed every 10 samples to provide an estimate of accuracy throughout the sample analysis. NIST 1643d (water) certifi- ed references solution was analyzed as a laboratory control sample. Beryllium (80 Ilg/liter), rhodium (10 Ilg/liter), and bismuth (10 Ilg/liter) were used as internal standards to correct for instrumental drift. Elemental concentration values obtained for dirty ethanol were blank-corrected by subtracting clean ethanol results. Similarly, the spiked ethanol was background-corrected by subtracting results obtained from two ethanol blanks. Background correction for the larvae digestion came from the four nitric acid digestion blanks. All tissue concentrations are reported on a dry weight basis. Statistics The weighted mean was calculated for selenium concen- trations in individual larvae and composites of larvae by dividing the concentration in the composite by the number of larvae in the composite. The Pearson correlation coeffic- ient for the relation between fish total length and whole- body selenium concentrations was determined using Statist- ical Analysis System programs (SAS, 1990). RESULTS Neutron Activation and ICP-MS Quality Control Results from analysis of four samples of NIST 1577 (bo- vine liver) standard reference material by neutron activation were all within the certified range, and method precision was 6.3% relative standard deviation. The limit of detection for the Se 17m method was 15 ng/g dry weight or 0.45 ng of selenium. The accuracy, precision, and limit of detection checks by MURRwere all based on 48-mg samples. Conse- quently, the accuracy and precision of selenium measure- ments in the larvae samples may have been lower because their sample weights ranged from 0.8 to 8.0 mg. For ICP-MS analysis of fish larvae, detection limits were categorized as follows: <lllg/g: Li, Sc, Ga, Ge, As, Rb, Y, Nb, Ag, Se, Sb, Cs, Th, U, and 27 rare-earth elements; <10 Ilg/g: V, Cr, Co, Ni, Zr, Mo, Cd, Sn, and Pb; <50Ilg/g: Ti, Mn, Cu, and Ba; and > 200 Ilg/g: Na, Mg, AI, K. Ca, Fe, and Zn. Recovery of 11 elements (Cr, Ni, Cu, Zn, As, Se, Cd, Sn, Sb, Hg, Pb) spiked into ethanol ranged from 70 to 135% for most elements. Severe losses were indicated for arsenic (48% recovery) and selenium (55% recovery), whereas con- tamination was indicated for lead (175% recovery). Recove- ries of these same 11 elements spikes into a blank solution prior to microwave digestion ranged from 94 to 108%. but contamination was apparent for tin (289% recovery). mer- cury (181 % recovery). and lead (174% recovery). Analysis of High-Purity Trace Elements in Drinking Water analyzed after every 10 samples indicated close agreement with certi- fied values. Analysis of 24 elements in NIST 1643d (water) reference material resulted in recoveries of 80 to III % for 20 elements, but were low for boron and iron and high for selenium and molybdenum. There seemed to be little loss of elements from larvae into the ethanol storage medium. Only calcium, sodium. and zinc concentrations were elevated in dirty ethanol compared to clean ethanol. There was scant selenium in either the clean or the dirty ethanol. Consequently, storage oflarvae in ethanol did not seem to alter inorganic concentrations in larvae. Selenium and Other Elements in Larvae Selenium concentrations in larvae from the five sites fell within a range of 2.24 Ilg/g at Cliff Creek to 7.42 Ilg/g at Stewart Lake Drain (Table 1). In general, fish collected at later dates were larger than those collected earlier (Table 1). For larvae collected at Cliff Creek and Stewart Lake Drain. selenium concentrations in larvae seemed to increase over time as fish grew, whereas they decreased at Sportsman's Drain and fluctuated at Greasewood Corral. Selenium con- centrations in larvae were positively correlated to fish total length at Cliff Creek (R = 0.65, P = 0.16), Stewart Lake Drain (R = 0.996, P = 0.05), and Greasewood Corral (R = 0.77, P = 0.13), but not at Sportsman's Drain (R = - 0.46, P = 0.54). For all sites the correlation coeffic- ient for the combined data was R = 0.57 (P = 0.01). At Cliff Creek, larvae collected on May 16 had the lowest selenium, those collected between May 19 and June 2 had intermediate concentrations, and those collected on June 13 had the highest concentration (Table 1). The same pattern of increasing selenium concentration in larvae over time oc- curred in collections at Stewart Lake Drain. At Greasewood Corral, larvae in the earliest collection on May 23 had lower selenium concentrations than those collected later. At Sportsman's Drain, which had only two collections, sel- enium concentrations were lower in larger larvae than in smaller larvae and were lowest at the last collection date. No pattern was apparent at Old Charlie Wash, which had one collection. These results suggest that larvae at Cliff Creek, Stewart Lake Drain, and possibly Grease- wood Corral were accumulating selenium from the environ- ment, but those at Sportsman's Drain were depurating selenium. The selenium concentration measured by ICP-MS in the 279-larvae composite sample from Stewart Lake Drain was 8.0 Ilg/g (Table 2). which was close to the weighted mean of