Amphibole asbestos minerals (riebeckite, grunerite-cummingtonite, anthophyllite,

Transcription

1 Soil Mineralogy Naturally Occurring Asbestos: Potential or Human Exposure, Southern Nevada, USA Brenda J. Buck* Dep. o Geoscience Univ. o Nevada-Las Vegas 4505 S. Maryland Pkwy. Las Vegas, NV Dirk Goossens Dep. o Geoscience 4505 S. Maryland Pkwy. Univ. o Nevada-Las Vegas Las Vegas, NV and Geography Research Group Dep. o Earth and Environmental Sciences Geo-Institute KU Leuven Celestijnenlaan 00E 3001 Leuven, Belgium Rodney V. Metcal Dep. o Geoscience 4505 S. Maryland Pkwy. Univ. o Nevada-Las Vegas Las Vegas, NV Brett McLaurin Dep. o Environmental, Geographical and Geological Sciences Bloomsburg Univ. o Pennsylvania 400 E. Second St. Bloomsburg, PA Minghua Ren Frederick Freudenberger Dep. o Geoscience 4505 S. Maryland Pkwy. Univ. o Nevada-Las Vegas Las Vegas, NV Amphibole asbestos minerals are known human carcinogens, and many regulations have been developed to limit occupational exposure. These minerals can also occur in the natural environment, where they may be more diicult to control. We applied a diverse set o analytical methods including scanning electron microscopy/energy dispersive spectroscopy, electron probe analysis, x-ray diraction, and ield-emission scanning electron microscopy to rock, soil, and dust samples and to particles attached to clothing samples and cars. We ound naturally occurring ibrous actinolite, a regulated amphibole asbestos mineral, in rock, soil, and dust that can be transported by wind, water, cars, or on clothing ater outdoor recreational activities. Sources o these ibrous amphiboles are several plutons in southern Nevada and Arizona and alluvial ans emanating rom asbestos-containing bedrock. The morphology o the amphibole ibers is similar to amphibole ibers ound in the USEPA Superund site at Libby, MT. We ound that the morphometry o the ibrous particles in the study area did not substantially change when the original bedrock weathered into soil, and particles were eroded and transported through wind and/or water and inally settled and accumulated on natural or other suraces. Because large populations in Boulder City, Henderson, and Las Vegas are located only a ew kilometers, sometimes even only a ew tens o meters, downwind rom the sources, and because most o the particles are transported in suspension ater they are emitted, potentially large populations in Boulder City, Henderson, and perhaps Las Vegas could be exposed. This study demonstrates a potential public health risk to several large population areas. Abbreviations: EDS, energy dispersive spectroscopy; FE-SEM, ield-emission scanning electron microscopy; IMA, International Mineralogical Association; SEM, scanning electron microscopy; XRD, x-ray diraction. Amphibole asbestos minerals (riebeckite, grunerite-cummingtonite, anthophyllite, tremolite, actinolite) are known human carcinogens (Agency or Toxic Substances and Disease Registry, 001; International Agency or Research on Cancer, 01). Exposure to ibrous amphiboles can cause asbestosis; lung, ovarian, and larynx cancer; mesothelioma; pleural ibrosis; and possibly other health eects including depressed immune unction, cardiovascular disease, and gastrointestinal cancer (Agency or Toxic Substances and Disease Registry, 001; Camargo et al., 011; International Agency or Research on Cancer, 01; Shannahan et al., 01). Although amphibole minerals are common, ibrous and asbestiorm amphiboles are less so and require speciic geologic processes that promote the growth o ibers, in particular rock deormation during or subsequent to amphibole growth (Ahn and Buseck, 1991; Virta, 00). The complexity o amphibole iber ormation is relected in their widely varying chemical composition, which Soil Sci. Soc. Am. J. 77:19 04 doi:10.136/sssaj Open Access Article Received 16 May 013. *Corresponding author (buckb@unlv.nevada.edu). Soil Science Society o America, 5585 Guilord Rd., Madison WI USA All rights reserved. No part o this periodical may be reproduced or transmitted in any orm or by any means, electronic or mechanical, including photocopying, recording, or any inormation storage and retrieval system, without permission in writing rom the publisher. Permission or printing and or reprinting the material contained herein has been obtained by the publisher. Soil Science Society o America Journal

2 determines their mineralogical classiication (Leake et al., 1997; Hawthorne et al., 01). Many amphibole minerals orm solid solution series, which are variations in chemical composition among two or more end members. Mineral names, thereore, represent a range o chemical composition, and speciic mineral names cannot be accurately assigned without chemical analysis using an electron microprobe (Meeker et al., 003). Additionally, amphibole asbestos classiication has been complicated by the use o industrial names (Lowers and Meeker, 00; Meeker et al., 003): or example, crocidolite and blue asbestos or riebeckite, and amosite or cummingtonite-grunerite (Virta, 00). Complications also arise because o evolving amphibole classiication schemes used by mineralogists (Leake, 1978; Leake et al., 1997; Hawthorne et al., 01). Further complications arise rom the use o terms to describe the crystal morphology o ibrous amphiboles (Lowers and Meeker, 00), in particular the term asbestiorm as applied to the regulation o commercial asbestos (Meeker, 009). The deinition o asbestiorm or regulatory purposes generally sets limits on size (diameter <0.5 mm) and aspect ratio (>0:1), with additional morphological characteristics (Lowers and Meeker, 00). It has long been known, however, that inhaling such minerals when they occur as long, thin ibers can cause disease, including those minerals that can be described as ibrous but all outside the narrower deinition o asbestiorm (Agency or Toxic Fig. 1. Predicted potential occurrences o ibrous amphiboles in Clark County, Nevada, and an area o adjacent Mohave County, Arizona, south o western Lake Mead. Plutons are Miocene in age; alluvium includes deposits o Miocene to Quaternary age. Only a small portion o these mapped areas has been examined. Substances and Disease Registry, 001; Camargo et al., 011; Meeker, 009; International Agency or Research on Cancer, 01). Naturally occurring asbestos reers to asbestos ound as a natural component o rocks and soils but may include ibrous minerals that do not meet the regulatory deinition o asbestos (Harper, 008). In addition to the regulated asbestos minerals, many other naturally occurring ibrous minerals are also implicated in causing disease. Examples include the amphibole minerals winchite, magnesioriebeckite, and richterite rom studies in Libby, MT (e.g., Meeker et al., 003; Shannahan et al., 01), erionite (a zeolite mineral) in Turkey and North Dakota (Carbone et al., 011; Ryan et al., 011), and antigorite (a serpentine mineral) in Poland (Wozniak et al., 1988) and New Caledonia (Baumann et al., 010). Toxicity is related to the width, length, aspect ratio, surace area, and surace chemical composition o the mineral iber (Aust et al., 011). Separable ibers longer than 5 mm with aspect ratios ³3:1 and those that contain Fe are considered to be more toxic (Aust et al., 011; Case et al., 011). Morphology, along with amphibole classiication, is at the heart o controversies over the potential public health risks and the necessity or remediation represented by naturally occurring asbestos. An example includes the naturally occurring ibrous actinolite-magnesiohornblende in the northern Caliornia community o El Dorado Hills (Harper, 008; Meeker et al., 006). Additionally, during the Libby, MT, criminal trial, deense experts or the mining company argued on the basis o morphology and amphibole classiication that the Libby amphiboles ell outside the regulatory deinition o asbestos (Meeker, 009), despite clear evidence or high rates o asbestos-related disease in the Libby population (Agency or Toxic Substances and Disease Registry, 00). Previous to this study, there were no reported natural occurrences o asbestos in Clark County, Nevada (Van Gosen, 008). Fibrous blue amphiboles (magnesioriebeckite, winchite, and actinolite) similar to those at Libby, MT, have been reported rom Miocene granitoid plutons (igneous intrusions) in an area o Mohave County, Arizona (Fig. 1), immediately adjacent to Clark County, Nevada (Potts and Metcal, 1997; Potts, 000). These ibrous amphiboles were ormed by hydrothermal alteration o aulted Miocene plutonic rock and are present as racture-ill veins and replacement o primary magmatic amphibole (Potts and Metcal, 1997; Potts, 000). Subsequent erosion o the altered pluton redistributed clasts containing ibrous amphibole into Miocene conglomerates (Williams, 003) as well as more recent Quaternary alluvium. The ibrous amphibole occurrences in Mohave County are all within unpopulated areas o the Lake Mead National Recreation Area; however, similar aulted and altered Miocene plutonic rocks and related conglomerate alluvium deposits crop out in and near populated areas o Clark County (Fig. 1). Thereore, in this study, we initiated an exploration or ibrous amphiboles using existing geologic maps and our expertise to create a map showing areas that contain the types o rocks that could potentially contain known or suspected carcinogenic ibrous amphiboles (Fig. 1). Field exploration o 193

3 one o these areas in the McCullough Range south o Las Vegas ound ibrous blue and green amphibole minerals lining ractures within Miocene granitoid plutons. Based on these results, we decided to perorm an initial exploration o this immediate area and its surroundings and determine the presence, location, mineralogy, and morphology o the ibrous amphiboles in rock, soil, and dust to guide uture research on the potential human health risk. Materials and Methods Forty-three samples were collected (Table 1). Seventeen rock samples were collected rom exposed granitoid outcrops: seven rom Boulder City, our rom Black Hill, and six rom the McCullough Range (Fig. ). Rock samples were collected by using a rock hammer to break resh samples rom exposed outcrops. Seventeen soil samples were collected, among which 11 were rom pristine desert suraces and six rom dirt roads (Fig. ). Soil was collected using a plastic scoop and sampling the uppermost <1-cm loose soil rom the surace. Airborne sediment was collected using Big Spring Number Eight (BSNE) dust traps (Fryrear, 1986) set at our levels: 5, 50, 75, and 100 cm. Dust collected at heights o 5 and 100 cm was analyzed or ibrous amphiboles using scanning electron microscopy (SEM) with an attached energy dispersive spectroscopy (EDS) detector (see below). One trap was placed in the backyard o a private residence in Boulder City rom 3 to 18 Nov. 01. This private residence was chosen because the backyard was exposed to the undeveloped, open desert. Other traps were placed downwind adjacent to dirt roads while vehicles (cars, trucks, and/or motorcycles) traversed them or a period o a ew hours. Lastly, our samples were collected rom clothing (pants and boots) and the side o a car tire ater a single morning o outdoor recreation in Boulder City. Activities during this time included driving and walking on dirt roads. The clothes and the car were washed beore and sampled ater the recreation activities. Samples were analyzed at the University o Nevada-Las Vegas using a variety o analytical methods (see Table 1) including: (i) SEM/EDS, (ii) ield-emission scanning electron microscopy (FE- SEM), (iii) x-ray diraction (XRD) analysis, and (iv) wavelength dispersive electron probe microanalysis (EPMA). For initial SEM/EDS and FE-SEM analyses, green and blue minerals visually identiied as amphiboles in rock samples were mounted on aluminum stubs with carbon tape. Dust and soil samples were mounted by pressing aluminum stubs with carbon tape against the sides o the plastic bags (e.g., Meeker et al., 003). Later, to more accurately measure particle size and shape, all soil samples, two partially disintegrated rock samples (Adams and H8R), and most BSNE dust samples were sieved through a 50-mm sieve. A total o 0. g was then put into 40 ml o deionized water and stirred with a magnetic stirring device. A 100-mL aliquot o each sample was collected and placed on a separate 0.4-mm pore diameter polycarbonate ilter (see Table 1. Sample locations, type, and analyses perormed using scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS), ieldemission scanning electron microscopy (FE-SEM), x-ray diraction (XRD), and/or wavelength dispersive electron probe microanalysis (EPMA). Sample GPS coordinates Analyses perormed Sample type Easting Northing SEM/EDS/FESEM XRD EPMA B soil (road) x B soil (road) x B soil (road) x B soil (road) x H soil (road) x H soil (road) x B soil x B soil x B soil x H soil x H soil x H soil x B soil x B soil x B soil x H soil x H soil x x Adams rock x x Adams rock x x Adams rock x BCP rock x BCP rock x BCP rock x x BCP rock x x F rock x F rock x F rock x H8R rock x MRA rock x MRB rock x x RJ rock x x RBKT rock x RBKT rock x RBKT rock x BCF dust x D dust x D dust x H8D dust x H9D dust x P dust (pants) x LR dust (boots) x LL dust (boots) x C dust (car tire) x GPS coordinates are reerenced to NAD 83 UTM Zone 11 North. Collected rom dust ater recreational activities in Boulder City. 194 Soil Science Society o America Journal

4 Fig.. Landsat image showing the study area in southern Nevada. Granitic plutons are shown in red. Estimated areas o alluvium deposited rom erosion o those plutons are shown in yellow. All sampling locations are shown as white circles and contained ibrous amphiboles. Meeker et al., 006). For two BSNE samples (BCF and H9D), most rock samples, and the dust samples rom clothing and the car tire, the quantities o loose sediment were too small to collect 0. g o sediment and the initial method was used. Stubs and ilters were coated with C and analyzed using a JEOL SEM ( JSM5600) equipped with an Oxord EDS detector and/or a FE-SEM ( JSM6700F) (Table 1). Samples were scanned or amphibole ibers at magniications ranging rom 500 to 000. Particle dimensions were measured through a combination o using the JEOL PC-SEM sotware calibrated with a certiied JEOL mag-standard (or FE-SEM images) and individually measuring SEM images using the scales provided with each image. Aspect ratios were calculated using the maximum length and average width o each particle as was done in previous studies (Campbell et al., 1977; Meeker et al., 003, 006) and to make accurate calculations o particle volume necessary to calculate the terminal all velocity (see below). A PANalytical X Pert Pro x-ray diraction spectrometer was used to analyze three rock samples and one soil sample. The samples were ground to a ine powder with a mortar and pestle and analyzed at the ollowing settings: 0 ma at 40 kv, q values 5 to 75; data were analyzed using X Pert Pro sotware. Amphibole mineralogical classiication was based on quantitative chemical analyses o three polished thin-sections measured on a our-wavelength dispersive spectrometer, automated JEOL 8900 electron probe microanalyzer. Amphibole structural ormulae calculations were perormed using the EPMA data ollowing recommendations o the International Mineralogical Association (IMA) (Leake et al., 1997); amphibole names were evaluated using both the Leake et al. (1997) and Hawthorne et al. (01) IMA classiication schemes. The ibrous nature o the amphiboles can result in poor polishing in probe sections, which in turn can result in lower oxide totals in electron probe microanalyses than might normally be expected (Meeker et al., 006). Following the approach used by the USGS EPMA laboratory in Denver (Meeker et al., 006), we rejected any analyses that yielded totals <9% (w/w). The atmospheric transport mode was determined or 765 conirmed amphibole particles. The atmospheric transport mode o a particle can be derived rom the ratio u /u*, where u is the terminal all velocity o the particle and u* is the riction velocity (Pye and Tsoar, 1990). The ollowing transport modes apply: saltation (u /u* > 1.5): particle transport on the order o centimeters to a ew meters, up to a ew tens o meters when particles experience succeeding jumps in a windstorm; modiied saltation (0.70 < u /u* < 1.5): particle transport rom a ew meters to a ew hundreds o meters; short-term suspension (0.10 < u /u* < 0.70): particle transport rom a ew hundreds o meters to several tens o kilometers; and longterm suspension (u /u* < 0.10): particle transport up to hundreds o kilometers and more. For nonspherical particles such as ibers, 195

5 which tend to have a cylindrical shape, u can be calculated with the ormula o Dietrich (198), taking the volume equivalent diameter as the reerence particle diameter and setting the Powers Index (Folk, 1965) equal to 6. Dietrich s ormula was derived or particle Reynolds numbers (Re) larger than approximately 0.1. Comparisons with experimental data sets (Goossens, 1987a) have shown that Dietrich s ormula provides excellent results or Corey shape actors (Corey, 1949) >0.0. For Corey shape actors <0.0 and Re >1.0, the ormula o Vakil and Green (009) can be used. For Corey shape actors <0.0 and Re <1.0, the ormula u ( p ) r r pg 6LR = 1h psina cosa ( R + L ) (see Appendix A) can be used, where r p is the mass density o the particle, r is the mass density o air, g is gravitational acceleration, L is the length o the particle, R is the radius o the particle, h is the dynamic viscosity o air, and a is the inclination o the particle in the low (a = 90 or orientation parallel to the low and a = 0 or orientation perpendicular to the low). For Re <0.1 (laminar low), the orientation o particles is random (a = 45 ), whereas or Re >1 (turbulent low), particles tend to orient themselves perpendicular to the low (a = 0 ) (Goossens, 1987b). For the interval 0.1 < Re < 1.0, the low is transitional. Calculations or the ibers investigated in this study showed that, in the transitional regime, a = 45 leads to much better agreement with Dietrich s ormula than a = 0, suggesting that particles are still more or less randomly oriented. Thereore, or Re <1.0, a value o 45 was used when calculating u. Geospatial methods used data rom existing geologic maps, data sets, and GIS sotware. A map or initial exploration was created using data rom House et al. (010) and Felger and Beard (010) to delineate areas in Clark County and northwestern Arizona that could potentially contain ibrous amphiboles (Fig. 1). On the House et al. (010) map, these include Miocene granitoid plutons and large alluvial deposits eroded rom them. The estimated areas draining plutonic rocks in Fig. were created within a GIS database by digitizing and integrating topography, hydrology, and aerial photography data sets (National Elevation Dataset, National Hydrology Dataset, and the Geospatial Data Gateway, datagateway.nrcs.usda.gov). These data were superimposed on the geologic data compiled rom House et al. (010), Anderson (1977), and Smith (1984) to determine the intersection o these drainages with exposed granitic plutons in the study area. Results Occurrence All 43 samples, including rock, soil, dust, car tire, and clothing, contained ibrous amphiboles. The original sources o these ibrous amphiboles are the Miocene plutons in the McCullough Range, Black Hill, and Boulder City areas (Fig. ). Fibrous amphiboles in these granitoid plutons occur as racture-ill veins and as alteration products o primary magmatic amphibole and appear hydrothermal in origin. In the McCullough Range and Black Hill, sediment eroded rom these sources orms large alluvial ans that drain primarily eastward into the Eldorado Dry Lake, along the western side o Highway 95, and north northwest along the southeastern edge o the city o Henderson. In Boulder City, sediment eroded rom the pluton is deposited in many areas in and surrounding the city (Fig. ). I we assume that our samples are representative o the surrounding geomorphic suraces, the estimated total surace area either containing ibrous amphiboles or likely to contain them (plutons plus alluvial ans in Fig. ) is approximately 14 km, which corresponds to about 53,000 acres. This calculation does not include the ibrous amphibole locality in Mohave County, Arizona (Potts, 000; Williams, 003), just east o the Colorado River (Fig. 1; center right o Fig. ). Additionally, other Miocene plutons in the region have not yet been studied but may also have the potential to contain similar ibers (Fig. 1). Thus, the precise locations and ull extent o ibrous amphiboles in this region remains to be determined. Mineralogy Although XRD analyses cannot identiy speciic amphibole phases (Meeker et al., 003), our analyses indicate the presence o amphibole: whole rock and soil samples analyzed by XRD identiied albite, amphibole, quartz, calcite, chlorite, and biotite. Electron microprobe analysis o ibrous amphiboles (111 analyses) in rocks rom Boulder City and the McCullough Range identiied these amphiboles primarily as actinolite (85%), one o the six regulated asbestos minerals (Fig. 3; Table ). Some analyses rom Sample MRB were classiied as magnesiohornblende; these analyses are clearly part o a solid solution compositional continuum with actinolite in the same sample (Fig. 3; Table ). We Fig. 3. Mineralogy o ibrous amphiboles in southern Nevada. Structural ormula calculations, including Fe 3+ estimates, and amphibole classiications ollow the 1997 recommendations o the International Mineralogical Association (Leake et al., 1997). Most analyses classiied as actinolite, which is one o six regulated asbestos minerals. Sites BCP4 and Adams are rom the Boulder City pluton; MRB is rom the McCullough Range near Henderson. 196 Soil Science Society o America Journal

6 are not aware o reports o asbestiorm magnesiohornblende in the literature, but actinolite-magnesiohornblende at El Dorado Hills, CA, is described as having ibrous to prismatic morphology (Meeker et al. 006, p. 0). The 1997 IMA classiication scheme (Leake et al., 1997) was used in Fig. 3 to acilitate comparison with prior work on ibrous amphiboles (e.g., Meeker et al., 006). Using the more recent IMA 01 classiication (Hawthorne et al., 01), the majority o our analyses (98%) were classiied as actinolite (most analyses that classiied as magnesiohornblende using IMA 1997 were reclassiied as actinolite using IMA 01). Table. Mineral chemistry data or analyses o 16 selected points rom three representative samples. Results or other analyses were similar. Oxide concentrations rom wavelength dispersive electron probe microanalysis; structural ormula calculations according to Leake et al. (1997). Sample ID BCP4 Adams MRB Point rim1 rim4 rim5 am5-13 am1-18 am-4 am-4 am3-9 am3-11 am3-14 rim1-7 am1-4 vein1- vein1-4 vein1-14 am3-1 Mineralogy Act Act Act Act Act Act Act Act Act Act Act Act Mg-hbld Act Act Act Oxides, % (w/w) SiO Al O TiO MnO MgO FeO CaO Na O K O F Cl Total Calculated structural ormula per 3 oxygens T site Si Al(T) Sum C site Al(C) Ti Fe Mg Fe Mn Sum B site Mn Fe Ca Na Sum A site Na K Sum Total cations Mg Avg. Fe 3+, % Act, actinolite; Mg-hbld, magnesiohornblende. Total Fe. T, C, B, and A sites reer to speciic cation sites in the amphibole structural ormula A 0-1 C B 5 T 8 O (OH ); Al(T) and Al(C) are the Al that is calculated to occur in the T and C sites, respectively. Mg/(Mg + Fe + )

7 Morphology The Nevada amphibole particles occur in many dierent morphologies ranging rom blocky crystals to lexible iber bundles and individual ibers (Fig. 4 and 5). The nomenclature used to describe the mineral habit is not consistent and can be problematic (Meeker et al., 003, 006). Oten, morphological deinitions cannot be determined through SEM analyses, and most importantly, morphological deinitions may not provide an accurate assessment o actual health risks (Meeker et al., 003, 006). Because iber morphology is part o regulatory deinitions (Crane, 1995), however, morphologic dierentiation is an important issue. In this study, individual ibers were deined as individual, narrow, usually very elongated crystals with straight, even edges (Fig. 4a and 5c). They may contain visual ibrils. Fiber bundles were deined as elongated amphiboles showing evidence o being in the process o separating into smaller ibers (Fig. 4b 4e). Many ibers and iber bundles oten showed evidence o curling or bending (Fig. 4b, 4c, 5d, and 5e). Some bundles and curved ibers it the deinition o asbestiorm (Lowers and Meeker, 00). We classiied prismatic crystals as any amphibole crystal that occurred with irregular, blunt edges and did not have ibers peeling away rom the larger mass (Fig. 4 ). Some o the latter may be cleavage ragments; however, since it is not possible with individual crystals to know i they have broken away rom a larger crystal (Meeker et al., 006), we cannot distinguish between prismatic crystals and cleavage ragments. When not deining speciic morphologies, we use the generic term particle here. Length and width (diameter) were measured rom 704 amphibole particles collected rom rock, dust, and soils across the entire study area and rom 61 amphibole particles collected rom clothes and a car tire. Energy dispersive spectroscopy analysis was used to veriy that these particles were amphibole (Fig. 6). An additional 59 visually identiied ibrous particles were measured with FE-SEM, which did not have EDS capability. To distinguish between the identiied and unidentiied particles, we used color or the ormer and gray or the latter in Fig. 7. Conirmed amphibole iber length ranged rom 3.9 to 67.6 mm (average 7.9 mm); conirmed amphibole particle length ranged rom 3.6 to 308. mm (average 18.1 mm). Conirmed amphibole average particle width was 3.9 mm (particles) and 1. mm (ibers). Sixty-nine percent o the amphibole particles measured and 97% o amphibole ibers were <3 mm in diameter and are thus considered respirable (National Research Council, 1984, p. 5 Fig. 4. Scanning electron microscope (SEM)/ield-emission scanning electron microscope (FE- SEM) images: (a) amphibole iber rom soil in an alluvial an along the western edge o Black Hill (Sample H7); (b) amphibole iber bundles rom rock at Martha P. King Elementary School (Sample Adams), Boulder City; (c) FE-SEM image o curling ibers rom rock at Martha P. King Elementary School (Sample Adams); (d) amphibole iber bundle rom rock at Martha P. King Elementary School (Sample Adams ); (e) amphibole iber bundle rom dust collected in dust trap in private yard in Boulder City (Sample D4); and () prismatic amphibole crystal rom soil in an alluvial an, near Highway 95 (Sample B9). 47). Fibers with aspect ratios >3:1 and those that are longer than 5 mm are more toxic (Aust et al., 011; Case et al., 011). In this study, we ound that 100% o all amphibole ibers and 97% o all amphibole particles had aspect ratios >3:1 (Fig. 8). To ind out how these particles will be transported through air, the annual requency distribution o the wind s riction velocity in the area was determined (Fig. 9a). The meteorological stations in Henderson and Boulder City do not record riction velocities, but during a previous project a detailed data set was collected rom Nellis Dunes Recreational Area, which is located approximately 8 km northeast o Las Vegas and has similar terrain types (data are available on request). For the 765 EDS-conirmed ibrous amphibole particles investigated in this study, the our transport modes o saltation, modiied saltation, short-term suspension, and long-term suspension were calculated using the riction velocity classes shown in Fig. 9a. The results show that even at very low riction velocities, almost all o the amphiboles will be transported in suspension (Fig. 9b). 198 Soil Science Society o America Journal

8 number o higher aspect ratio particles than the El Dorado Hills ibrous amphiboles (Meeker et al., 006). In act, their higher aspect ratio requency is comparable to that o the Libby ibrous and asbestiorm amphiboles (Meeker et al., 003) and the anthophyllite and tremolite asbestos characterized by Campbell et al. (1977) (Fig. 8). Additionally, there is no sharp morphometric delineation between individual ibers, iber bundles, and prismatic crystals but rather a very gradual transition (Fig. 7). The morphometry o the ibrous particles does not substantially change when the original bedrock weathers into soil; particles then become eroded, are transported through wind and/ or water, and inally settle and accumulate on natural or other suraces (Fig. 7). Although some granulometrical dierentiation likely takes place during these processes, the data suggest that changes in shape remain small. Altogether, comparing their morphology with that in Libby, MT, and their identiication as actinolite, one o the six regulated asbestos minerals, these data indicate that these ibers in southern Nevada are highly respirable and possibly carcinogenic. The vast majority o the population o Clark County, million in 01 (U.S. Census Bureau 013), is ound in the conurbation o Las Vegas Henderson Fig. 5. Scanning electron microscope images: (a) amphibole crystal with rayed ends collected as dust generated rom car driving on dirt road, Boulder City (Sample BCF); (b) amphibole North Las Vegas (Fig. 1). Populations in bundle rom soil in an alluvial an, western side o Black Hill (Sample H5); (c) amphibole iber Boulder City, Henderson, and probably rom soil in an alluvial an, near Highway 95 (Sample B10); (d) amphibole bundle rom soil the easternmost portions o Las Vegas may in an alluvial an south o Boulder City (Sample B8); (e) asbestiorm amphiboles rom broken rock collected in McCullough Range (Sample RJ1); and () ibrous amphibole rom rock in be exposed to these actinolite asbestos ibers. The most common exposure route southern Boulder City (Sample BCP 1). would be through airborne dust. In this arid Discussion region, signiicant dust emissions are created through both (i) natural wind erosion (Fig. 10a) and (ii) any anthropogenic Although more particles need to be measured in uture activities that disrupt the surace (Goossens et al., 01) (Fig. studies, our data suggest that the morphology o the ibers in 10b). Based on soil texture, vegetation, and visual evidence, southern Nevada is comparable to those in the Libby, MT, USEPA the area most vulnerable to wind erosion is ound south and Superund site (Meeker et al., 003) and that the ibers are longer southwest o Boulder City (Eldorado Dry Lake and the sandy than those in El Dorado Hills, CA, a well-known area containing suraces east o Highway 95; see Fig. ). Numerous types o naturally occurring amphibole asbestos classiied as actinolite and anthropogenic activities occur or are planned throughout the magnesiohornblende (Meeker et al., 006) (Fig. 8). The average area, and many dirt roads are present in the McCullough Range aspect ratio or our measured amphibole ibers was 16:1, which and surrounding Boulder City. In addition, the wind regime in is less than the amphiboles measured at Libby, MT (Meeker et the area is bimodal (Fig. 11), with strong south and southwest al., 003), but much greater than those ound in El Dorado Hills, winds in spring and summer and primarily northeast winds CA (Meeker et al., 006). Particles that are >5 mm in length and in autumn and winter (Goossens and Buck, 011). Southeast have aspect ratios >3:1 are countable as asbestos (Crane, 1995). winds are less common but do occur. Thereore, especially in Ninety-two percent o the ibers and 95% o all particles counted the spring, the populations o eastern Henderson and eastern would meet this deinition. As discussed in Meeker et al. (003), Las Vegas are located within only a ew hundred meters to a ew some deinitions o true asbestos require lengths >5 mm with kilometers downwind rom the emission sources. In Boulder aspect ratios >0:1. In this study, 8% o ibers met this deinition. In general, the Nevada amphiboles have a signiicantly greater 199

9 City, the sources o amphibole ibers are even closer, with a potential or exposure year-round. The aeolian transport mode can be used as a tool to estimate how ar downwind the ibrous amphiboles can be transported once they are emitted and become airborne. Even at very low riction velocities, almost all o the measured amphibole particles will be transported in suspension (Fig. 9b). This means that, once airborne, they can easily reach the populated areas o Boulder City, eastern Henderson, and the eastern portions o Las Vegas, although or the latter, concentrations will be lower because o natural dilution. Especially or Boulder City, more research is needed because the populated zones are entirely situated within the source area o the ibrous amphiboles. Several outcrops o the pluton occur within the city itsel, and the remaining part o the city has been built on soils derived rom the pluton or on suraces enriched with amphiboles transported by water and wind (Fig. ). More outcrops o the pluton occur directly east and north o the city, and amphibole-enriched alluvial ans and sand suraces prone to wind erosion are adjacent to the south. Thereore, populations in Boulder City are likely to be exposed to ibrous amphiboles regardless o wind direction, and anthropogenic activities can cause dust production year-round. No inormation is currently Fig. 6. Comparison o energy dispersive spectra o ibrous amphiboles representative o those collected rom amphibole particles in soil (Samples H6 H5, H4, B4, B3, and B1) (upper let), collected rom amphibole particles in dust (Samples H9D, H8D, and BCF), on clothes (Samples P and LR), or the car tire (Sample C) (upper right), and collected rom polished thin sections o rocks collected rom Boulder City (bottom). The spectra indicate similar amphibole compositions in dust (and clothes), soil, and rock. Fig. 7. Morphology o the Nevada ibrous amphiboles: (a) aspect ratio and (b) iber length o all 104 measured particles as a unction o the particle diameter or the three morphological iber types deined in this study, and (c) aspect ratio and (d) iber length o the same 104 particles as a unction o the particle diameter or our dierent types o sources. Scanning electron microscope/energy dispersive spectroscopy (SEM/EDS)-conirmed amphiboles are shown in color; ield-emission scanning electron microscope (FE-SEM) non-eds-conirmed but visually identiied amphiboles are shown in gray. 00 Soil Science Society o America Journal

10 Fig. 8. Comparison with ibrous amphiboles rom other sites: (a) energy dispersive spectroscopy (EDS)-conirmed amphiboles rom this study vs. amphiboles (tremolite and actinolite + magnesiohornblende) rom El Dorado Hills, CA (Meeker et al., 006) and asbestos (anthophyllite and tremolite) analyzed in a U.S. Bureau o Mines study (Campbell et al. 1977); b. amphiboles rom our study site vs. amphiboles (richterite and winchite) rom Libby, MT (Meeker et al. 003). The data or Libby, MT, are or particle diameters <5 mm and the minimum diameter or reliable EDS analysis in the Nevada study was 0.3 mm; thereore, the curves were calculated or the interval 0.3 to 5.0 mm, which represents 84% o the EDS-conirmed Nevada amphiboles. All curves are or total samples (individual ibers, iber bundles, and/or prismatic crystals, in their original proportions). available, however, on the actual airborne concentrations at any location in the county. This inormation is greatly needed to better understand potential health risks. Conclusions Given the mineralogy and morphology o these ibers, it is imperative that this problem be urther studied in southern Nevada. In Libby, MT, exposure to ibrous amphiboles has resulted in asbestos-related-disease mortality rates 40 to 80 times higher than other areas in Montana and the United States (Agency or Toxic Substances and Disease Registry, 00). In the arid climate o Nevada, populations may be exposed to these ibers through dust emissions that can occur rom both natural and C F = W D wind erosion and anthropogenic activities. The evidence o ibrous amphiboles on car tires and on clothing ater recreational activities shows that they can be brought back to amily members and thus increase the risk o exposure or other populations besides those directly exposed through outdoor dust emissions. Because health eects may occur even at low levels o exposure to ibrous amphiboles (Alexander et al., 011; Hillerdal, 1999), our data indicate a potential public health threat in southern Nevada. Any potential uture land-use projects should careully determine the risks to both workers and the regional populations because disturbances to these natural desert suraces cause increased dust emissions (Goossens et al., 01). Social behaviors and land management practices that limit dust production in areas where ibers are present could be implemented to reduce human exposure. There is a compelling need or epidemiology studies, additional geologic and mineral studies, and signiicantly more research on their location, emission, airborne concentration, and pathways o human exposure. Appendix A When terminal all velocity is reached, low resistance has become equal to the submerged particle weight: G F = F [1] B W where G is the particle weight, F B is the buoyancy orce, and F W is the low resistance. Since G= B ru A r p Vg F = r Vg where r p is the particle s mass density, r is the luid s mass density, V is the particle volume, g is gravitational acceleration, C D is the drag coeicient, u is the particle velocity, and A is the particle s cross-sectional area perpendicular to the low, CDru A Vg = [] ( rp r ) 01

11 Fig. 9. Atmospheric transport modes: (a) annual requency distribution o the wind s riction velocity near the study area; and (b) proportion o our aeolian transport modes calculated or the 765 energy dispersive spectroscopy (EDS)-conirmed ibrous amphibole particles measured in this study. Even at very low riction velocities, almost all o the amphiboles will be transported in suspension. For a cylindrical particle o length L and radius R alling at an angle a (Fig. 1), the particle s volume is LR p and the particle s cross-sectional area perpendicular to the low is R psina +RLcosa. Thereore, r u = + [3] ( ) r ( p r LR pg CD R psina RLcosa) For a spherical particle in laminar low (Re < 0.1), the drag coeicient is equal to 4/Re, where Re is the particle Reynolds number (Prandtl and Tietjens, 1934). For Re between 0.1 and 1.0, this ormula still gives a very good approximation (see Giles, 196). For nonspherical particles and Re <1.0, the drag coeicient can be written as (4/Re)S, where S is a dimensionless shape actor. A test with particles used in this study (Fig. 13) showed that or the range < Re < 1 and a = 45 (see main text), very good results are obtained when S is set equal to unity. The ollowing ormula thus gives a very good approximation: 4 r u = + [4] Re ( r ) ( p r LR pg R psina RLcosa) Rearranging Eq. [4]: rp r r Writing Re as 1 p = p a+ a Re ( sin cos ) LR g u R L [5] r Eu h where E is the particle diameter and h is the luid s dynamic viscosity: or rp r 1 h LR g u p = Rp a+ L a r Eu r ( sin cos ) [6] Fig. 10. Natural and anthropogenically produced dust: (a) a dust storm entering Henderson Las Vegas rom the south; the McCullough Range is obscured by dust; (b) dust generated by a single o-road vehicle driving on a dirt road in the McCullough Range. 0 Soil Science Society o America Journal

12 Fig. 11. Wind rose or the Las Vegas area. Percentages are based on hourly data and reer to the period 19 Dec. 007 to 15 Dec. 008 (data rom Nellis Air Force Base meteorological station). 1 = + [7] E ( rp r ) LRpg hu( Rpsina Lcosa) Hence, ( p ) r r LRpgE u= 1h psina cosa ( R + L ) [8] Choosing the volume equivalent diameter D eq to represent the particle s diameter in Re 3 ( eq = 6 ) ( p ) 3 r r LRpg 6LR u= 1h psina cosa ( R + L ) and ater simpliication: ( p ) 3 r r pg 6LR 4 5 u= 1h psina cosa ( R + L ) D LR : [9] [10] Acknowledgments Thanks to R. Johnsen, G. Smith, O. Tschauner, P. Rees, S. Schaer, D. Merkler, and three anonymous reviewers. Funding provided by the University o Hawaii Cancer Center. Fig. 1. Cylindrical particle o length L and radius R alling at an angle a. Fig. 13. Terminal all velocities o particles in this study calculated with the equation o Dietrich (198) and with Eq. [10]. The covered particle Reynolds number interval is < R e < 1. Reerences Agency or Toxic Substances and Disease Registry Toxicological proile or asbestos. ATSDR, Atlanta, GA. tp61.pd (accessed 17 Nov. 01). Agency or Toxic Substances and Disease Registry. 00. Mortality in Libby, Montana ( ). ATSDR, Atlanta, GA. Alexander, B., K. Raleigh, J. Johnson, J. Mandel, J. Adgate, G. Ramachandran, et al Radiographic evidence o nonoccupational asbestos exposure rom processing Libby vermiculite in Minneapolis, Minnesota. Environ. Health Perspect. 10: doi:10.189/ehp Anderson, R.E Geologic map o the Boulder City 15-minute quadrangle, Clark County, Nevada. USGS Map GQ USGS, Reston, VA. Ahn, J.H., and P.R. Buseck Microstructures and iber-ormation mechanisms o crocidolite asbestos. Am. Mineral. 76: Aust, A., P. Cook, and R. Dodson Morphological and chemical mechanisms o elongated particle toxicities. J. Toxicol. Environ. Health, Part B 14: doi: / Baumann, F., P. Maurizot, M. Mangeas, J. Ambrosi, J. Douwes, and B. Robineau Pleural mesothelioma in New Caledonia: Associations with environmental risk actors. Environ. Health Perspect. 119: doi:10.189/ehp Camargo, M., L. Stayner, K. Strai, M. Reina, U. Al-Alem, P. Demers, and P. Landrigan Occupational exposure to asbestos and ovarian cancer. Environ. Health Perspect. 119: doi:10.189/ehp Campbell, W., R. Blake, L. Brown, E. Cather, and J. Sjoberg Selected silicate minerals and their asbestiorm varieties: Mineralogical deinitions and identiication-characterization. In. Circ U.S. Bureau o Mines, Washington, DC. Carbone, M., Y. Baris, P. Bertino, B. Brass, S. Comertpay, A. Dogan, et al Erionite exposure in North Dakota and Turkish villages with mesothelioma. Proc. Natl. Acad. Sci. 108: doi: / pnas Case, B., J. Abraham, G. Meeker, F. Pooley, and K. Pinkerton Applying deinitions o asbestos to environmental and low-dose exposure levels and health eects, particularly malignant mesothelioma. J. Toxicol. Environ. Health, Part B 14:3 39. doi: / Corey, A.T Inluence o shape on the all velocity o sand grains. M.S. thesis. Colorado A&M College, Fort Collins. Crane, D.T Polarized light microscopy o asbestos. OSHA Method ID- 191 (revised 1995). Occup. Sa. Health Admin., Washington, DC. (accessed 4 Apr. 013). Dietrich, W.E Settling velocity o natural particles. Water Resour. Res. 18: doi:10.109/wr018i006p01615 Felger, T.J., and L.S. Beard Geologic map o Lake Mead and surrounding regions, southern Nevada, southwest Utah, and northwestern Arizona. In: P.J. Umhoeer et al., editors, Miocene tectonics o the Lake Mead region, Central Basin and Range. Spec. Pap Geol. Soc. Am., Boulder, CO. p

13 Folk, R.L Student operator error in determination o roundness, sphericity and grain size. J. Sediment. Petrol. 5: Fryrear, D A ield dust sampler. J. Soil Water Conserv. 41: Giles, R.V Fluid mechanics and hydraulics. nd ed. Schaum Publ. Co., New York. Goossens, D. 1987a. A drag coeicient equation or natural, irregularly shaped particles. Catena 14: doi: /s (87) Goossens, D. 1987b. Sedimentatiemechanismen bij Natuurlijke Stodeeltjes in Lucht. Ph.D. diss. Katholieke Univ. Leuven, Leuven, Belgium. Goossens, D., and B. Buck Eects o wind erosion, o-road vehicular activity, atmospheric conditions and the proximity o a metropolitan area on PM10 characteristics in a recreational site. Atmos. Environ. 45: doi: /j.atmosenv Goossens, D., B. Buck, and B. McLaurin. 01. Contributions to atmospheric dust production o natural and anthropogenic emissions in a recreational area designated or o-road vehicular activity (Nellis Dunes, Nevada, USA). J. Arid Environ. 78: doi: /j.jaridenv Harper, M th Anniversary critical review: Naturally occurring asbestos. J. Environ. Monit. 10: doi: /b810541n Hawthorne, F., R. Oberti, G. Harlow, W. Maresch, R. Martin, J. Schumacher, and M. Welch. 01. Nomenclature o the amphibole super group. Am. Mineral. 97: doi:10.138/am Hillerdal, G Cases associated with non-occupational and low dose exposures. Occup. Environ. Med. 56: doi: /oem House, P.K., H. Green, and A. Grimmer, and Nevada Digital Dirt Mapping Team Preliminary suricial geologic map o Clark County, Nevada. Open-File Rep Nevada Bureau o Mines and Geology, Reno. International Agency or Research on Cancer. 01. Asbestos. In: A review o human carcinogens. IARC Monogr. Eval. Carcinogenic Risks to Humans 100C. WHO Press, Lyon, France. p Leake, B.E Nomenclature o amphiboles. Can. Mineral. 16: Leake, B.E., A. Woolley, C. Arps, W. Birch, M. Gilbert, J. Grice, et al Nomenclature o the amphiboles: Report o the subcommittee on amphiboles o the International Mineralogical Association, Commission on New Minerals and Mineral Names. Am. Mineral. 8: Lowers, H., and G. Meeker. 00. Tabulation o asbestos-related terminology. USGS Open-File Rep USGS, Reston, VA. o/00/or-0-458/index.html (accessed 18 Apr. 013). Meeker, G Asbestos sans mineralogy? A view rom a dierent hilltop. Elements 5:69. Meeker, G., A. Bern, I. Brownield, H. Lowers, S. Sutley, T. Hoeen, and J. Vance The composition and morphology o amphiboles rom the Rainy Creek Complex near Libby, Montana. Am. Mineral. 88: Meeker, G., H. Lowers, G. Swayze, B. Van Gosen, S. Sutley, and I. Brownield Mineralogy and morphology o amphiboles observed in soils and rocks in El Dorado Hills, Caliornia. USGS Open-File Rep USGS, Reston, VA. National Research Council Asbestiorm ibers: Nonoccupational health risks. Natl. Acad. Press, Washington, DC. Potts, D.A Sodium metasomatism and magnesio-riebeckite mineralization in the Wilson Ridge pluton. M.S. thesis. Univ. o Nevada, Las Vegas. Potts, D.A., and R.V. Metcal Sodium metasomatism and riebeckite mineralization in an extensional terrane: Wilson Ridge pluton, northwest Arizona. Geol. Soc. Am. Abstr. 9(5):57. Prandtl, L., and O.G. Tietjens Applied hydro- and aeromechanics. Dover Publ., New York. Pye, K., and H. Tsoar Aeolian sand and sand dunes. Unwin Hyman, London. Ryan, P., M. Dihle, S. Griin, C. Partridge, T. Hilbert, R. Taylor, et al Erionite in road gravel associated with interstitial and pleural changes: An occupational hazard in western United States. J. Occup. Environ. Med. 53: Shannahan, J., M. Schladweiler, R. Thomas, W. Ward, A. Ghio, S. Gavett, and U. Kodavanti. 01. Vascular and thrombogenic eects o pulmonary exposure to Libby amphibole. J. Toxicol. Environ. Health A 75: doi: / Smith, E.I Geologic map o the Boulder Beach quadrangle, Nevada. Map 81. Nevada Bureau o Mines and Geology, Reno. U.S. Census Bureau State & county QuickFacts: Clark County, Nevada. U.S. Census Bureau, Washington, DC. states/3/3003.html (accessed 18 Apr. 013). Vakil, A., and S.I. Green Drag and lit coeicients o inclined inite circular cylinders at moderate Reynolds numbers. Comput. Fluids 38: doi: /j.compluid Van Gosen, B Reported historic asbestos mines, historic asbestos prospects, and natural asbestos occurrences in the southwestern United States (Arizona, Nevada, and Utah). USGS Open-File Rep USGS, Denver, CO. Virta, R. 00. Asbestos: Geology, mineralogy, mining, and uses. USGS Open- File Rep USGS, Reston, VA. Williams, M.M Depositional history o the Black Mountain conglomerate, Mohave County, Arizona: Sedimentary response to Miocene extension. M.S. thesis. Univ. o Nevada, Las Vegas. Wozniak, H., E. Wiecek, and J. Stetkiewicz Fibrogenic and carcinogenic eects o antigorite. Pol. J. Occup. Med. Environ. Health 1: Soil Science Society o America Journal