GEOLOGICAL SURVEY CIRCULAR 692 Selenium, Fluorine, and Arsenic in Surficial Materials of the Conterminous United States Selenium, Fluorine, and Arsenic in Surfic"ial Materials of the Conterminous United States By Hansford T. Shacklette, Josephine G. Boerngen, and John R. Keith GEOLOGICAL SURVEY CIRCULAR 692 Washingto(l 1974 United States Department of the Interior ROGERS C. B. MORTON, Secretary Geological Survey V. E. McKelvey, Director Free on application to the U.S. Geological Survey, National Center, Reston, Va. 22092 CONTENTS Page Abstract .......................................... 1 Fluorine-Continued Introduction ...................................... 1 Results of analyses ............................ 6 Selenium ......................................... 2 Discussion .................................... 6 Analytical method ............................. 2 Arsenic ........................................... 10 Results of analyses ............................ 2 Analytical method ............................. 10 Discussion .................................... 2 Results of analyses ............................ 10 Fluorine .......................................... 3 Discussion .................................... 10 Analytical method ............................. 3 References cited ................................... 11 ILLUSTRATIONS Page FIGURE 1. Map showing selenium concentrations in surficial materials of the conterminous United States . 4 2. Map showing physiographic regions of the conterminous United States . 6 3. Map showing fluorine concentrations in surficial materials of the conterminous United States . 8 4. Map showing arsenic concentrations in surficial materials of the conterminous United States ............. 12 TABLES TABLE~ 1-3 · Concentrations of elements in samples of soils and other surficial materials from the conterminous United States: Page 1. Selenium..................................................................................... 2 2. Fluorine . 6 3. Arsenic . 10 III Selenium, Fluorine, and Arsenic in Surficial Materials of the Conterminous United States By Hansford T. Shacklette, Josephine G. Boerngen, and John R. Keith ABSTRACT trations were determined by J. W. Budinsky, B. A. Concentrations of selenium, fluorine, and arsenic in 912, 911, McCall, and Roosevelt Moore. and 910 samples, respectively, of soils and other regoliths from We measured the precision of the analytical sites approximately 50 miles (80 km) apart throughout the methods used for these elements by analyzing 48 United States are represented on maps by symbols showing five randomly selected samples in duplicate. The 48 ranges of values. Histograms of the concentrations of these duplicates were randomly interspersed among the elements are also given. The geometric-mean concentrations (ppm) in the samples, grouped by area, are as follows: other 912 samples and were unknown to the analysts. The precision of each method was es­ Selenium- Entire United States, 0.31; Western United States, timated by 0.25; and Eastern United States, 0.39. 48 ~(log x ·-logy ·) 2 Fluorine- Entire United States, 180; Western United States, . 1 £ t 250; and Eastern United States, 115. sJ= z= 0.00040, Arsenic- Entire United States, 5.8; Western United States, 96 6.1; and Eastern United States, 5.4. where s& is the precisiOn, and xi and Yi are, respectively, the determinations of each element INTRODUCTION for the ith sample and its corresponding duplicate. The concentrations of 38 elements in samples of The analytical reproducibility, as well as the soils and other regoliths from sites about 50 miles logarithmic variance, for each of the three (80 km) apart on travel routes throughout the con­ elements is given later in this report. terminous United States were given by Shacklette, Analytical values for the three elements were Hamilton, Boerngen, and Bowles (1971), transformed to a logarithmic form because the Shacklette, Boerngen, and Turner (1971), and frequency distributions are more nearly sym­ Shacklette, Boerngen, Cahill, and Rahill (1973). metrical on a logarithmic scale than on an arith­ After these reports were prepared, analytical metic scale. The best measure of central tendency methods became available for determining in a lognormal distribution is the geometric mean, selenium, fluorine, and arsenic in surficial which is the antilogarithm of the mean logarithm. materials in concentrations as low as 0.1 ppm (part The most convenient measure of variation is the per million), 10 ppm, and 1 ppm, respectively. geometric deviation, which is the antilogarithm of The samples were collected and prepared for the standard deviation of the logarithms. analysis in the same manner as reported earlier Estimates of the arithmetic mean (tables 1-3) were (Shacklette, Hamilton, and others, 1971) and were derived by the use of Sichel's (1952) technique. analyzed in sequence that was completely random These methods of statistical evaluation are the with respect to sampling locality. Selenium con­ same as those used for evaluating other elements centrations were determined by J. S. Wahlberg in the samples, as reported by Shacklette, and M. W. Solt, and fluorine and arsenic concen- Hamilton, Boerngen, and Bowles (1971). 1 Although most of the samples studied were west of the 97th meridian, are given in table 1. collected along roads, the specific sampling sites Figure 1 shows the distribution of the sample sites were selected to obtain surficial materials that throughout the conterminous United States and were as representative as possible of their natural the selenium concentrations of the samples, ex­ condition. Some samples, of necessity, were pressed in terms of five geometric ranges of con­ collected in cultivated fields; the degree of con­ centration. tamination, if any, of these samples or of a few samples collected near road shoulders cannot be TABLE !.-Concentrations of selenium, in pa-rts per million, in sa.mples ofsoils and other surficial materials from the conter­ evaluated from the data at hand. Most surficial minous United States materials analyzed were sampled at a depth of [Number of samples is given in parentheses after area] about 8 inches (20 em). We believe that soils and ... ... other regoliths from this depth are influenced very ·;:::: = ·E·~ e... -; Area ~ 8 -~ little by the surficial contamination associated ~ ~ 0 .., .s g e ..,., = C,!) ~8 with roadways. ~ C,!) Many geologists and other workers of the U.S. Entire conterminous Geological Survey assisted in this study by collect­ United States (912) 0.1-4.32 0.31 2.42 0.45 Western United States, ing samples along travel routes to their own field­ west of the 97th meridian (492) ............. .1-4.32 .25 2.53 .38 study areas. This assistance, and that of computer Eastern United States, east of the 97th specialists, was acknowledged in the earlier meridian (420) ............. .1-3.88 .39 2.17 .52 reports of this sampling program (Shacklette, Hamilton, and others, 1971; Shacklette, Boerngen, 'Estimated by method of Sichel (1952). and Turner, 1971.) SELENIUM DISCUSSION ANALYTICAL METHOD The concentrations of selenium in natural Selenium concentrations in the samples were de­ materials, including soils, have been extensively termined by a chemical-X-ray fluorescence method. investigated because of the long-known toxicity of A 2-g (gram) soil sample is fused with 15-g sodium this element to domestic animals and man and, carbonate and 5-g sodium peroxide. The fusion more recently, because the essentiality of selenium cake is then dissolved in 200 ml (milliliters) of in animal metabolism has been established. The water, and the solution is filtered. The filtrate is dual metabolic role of this element was discussed acidified, a tellurium carrier added, and the by Frost (1972). The relationships of selenium selenium along with the carrier is precipitated by chemistry and agricultural practices were dis­ an iodide sulfite reduction reaction. The cussed in an agriculture handbook by Anderson, precipitate of selenium and tellurium is collected Lakin, Beeson, Smith, and Thacker (1961), and on a filter disk and dried, and the quantity of more than 200 literature references to the subject selenium is then determined by X-ray were given. More recently, reports were published fluorescence. of .selenium accumulation in soils, plants, and The logarithmic variance of the analytical animals to levels that are toxic (Lakin, 1972) and cf method was measured as 0.042729. This means effects of selenium deficiency in soils on animal that the analyses are reproducible within a factor health (Muth and Allaway, 1963; Allaway, 1969; of 1.61, computed as the antilog of sa at the 68- and Oldfield, 1972). Selenium is generally con­ percent level of confidence, or within a factor of sidered to be nonessential in plant metabolism. 2.59, computed as the square of the antilog sa at Low concentrations of this element, however, have the 95-percent level. The logarithmic variance of been shown to stimulate plant growth (Ganje, selenium measured in all 912 samples is 0.1476, in­ 1966, p. 394). dicating that the analytical-error variance con­ Lakin (1961, p. 27) outlined the sources of tributed less than 29 percent to the total variance selenium in soils .as follows: "The selenium in soils in the data. may be derived (1) from parent material weathered from the underlying rock; (2) from RESULTS OF ANALYSES wind- or water-deposited seleniferous materials; (3) Statistics for the selenium concentration of all from ground or surface water,