.USGS science for a changing world EFFECTS OF SNOWMOBILE USE ON SNO\VPACK CHEMISTRY IN YELLOWSTONE NATIONAL PARK, 1998 Water-Resources Investigations Report 99-4148
u.s. Department of the Interior U.S. Geological Survey
Prepared in cooperation with the NATIONAL PARK SERVICE
I u.s. DEPARTMENT OF THE INTERIOR BRUCE BABBITT, Secretary
U.S. GEOLOGICAL SURVEY Charles G. Groat, Director
The use of firm, trade, and brand names in this report is for identification purposes only and does not constitute endorsement by the U.S. Geological Survey.
For additional information write to: Copies of this report can be purchased from:
District Chief U.S. Geological Survey U.S. Geological Survey Information Services Box 25046, Mail Stop 415 Box 25286 Denver Federal Center Federal Center Denver, CO 80225-0046 Denver, CO 80225 CONTENTS
Abstract I Introduction I Background 2 Purpose and Scope 3 Description of Study Area 4 Acknow ledgments 4 Sample-Collection Locations and Methods 4 Analytical Procedures 7 Snowpack Chemistry 8 Major Ions 8 Selected Hydrocarbons 10 Snowmel t Runoff 16 Patterns of Chemistry Relative to Snowmobile Use 21 Concl usions 22 References 22
FIGURES
I. Map showing location of snow-sampling sites in the Rocky Mountains in Western United States 2 2-3. Graphs showing: 2. Relation between ammonium and sulfate in mountain snow packs throughout the Rocky Mountain region, 1998 9 3. Ammonium and sulfate levels in snow at in-road and off-road sites in Yellowstone National Park. 1998 II 4-6. Maps showing: 4. Ammonium ion concentrations in snow in Teton- Yellowstone area, 1998 13 5. Nitrate ion concentrations in snow in Teton-Yellowstone area, 1998 14 6. Sulfate ion concentrations in snow in Teton- Yellowstone area, 1998 15 7-8. Graphs showing: 7. Ammonium, sulfate, and benzene levels in snow at selected sites in Yellowstone National Park, 1998 17 8. Hydrocarbons in snow on snowmobile-packed roads and olT-road sites 18
TABLES
I. Snow-sampling site locations and estimated snowmobile use levels in the Greater Yellowstone area and in northern Colorado 5 2. Snow chemistry of selected inorganic compounds at protocol-development sites 10 3. Snow chemistry of selected inorganic compounds in the Greater Yellowstone area 12 4. Snow chemistry of selected volatile organic compounds at protocol-development sites 16 5. Snow chemistry of selected volatile organic compounds in the Greater Yellowstone area 19 6. Squared Pearson correlation coefficients between selected organic and inorganic gasoline-engine- emission by-products 19 7. Snowmelt runoff chemistry of selected inorganic compounds in Yellowstone National Park 20 8. Snowmelt runoff chemistry of selected volatile organic compounds in Yellowstone National Park 20
CONTENTS III
I CONVERSION FACTORS, VERTICAL DATUM, AND ABBREVIATIONS
MultiplyquartfootinchTomile1.057obtainBy3.2810.62140.39370.03937 millimetercentimeterkilometermeter(mm)(km)(cm)liter(m)
Temperature in degrees Celsius (OC) may be converted to degrees Fahrenheit (OF) as follows:
OF = 9/5 (0C) + 32
Sea level: In this report "sea level" refers to the National Geodetic Vertical Datum of 1929-a geodetic datum derived from a general adjustment of the first-order level nets of the United States and Canada, formerly called Sea Level Datum of 1929.
ADDITIONAL ABBREVIATIONS
ng/L nanograms per liter J..leq/L microequivalents per liter mph miles per hour mL milliliter
IV CONTENTS Effects of Snowmobile Use on Snowpack Chemistry in Yellowstone National Park, 1998
By George P. Ingersoll
Abstract 1995). The majority of winter visitors tour Yellow• stone by motorized snowmobiles because unplowed Snowmobile use in Yellowstone National snow-covered roadways preclude highway access by Park has increased substantially in the past three automobiles into much of Yellowstone throughout the decades. In areas of greatest snowmobile use, snowfall season. ~he park is design~ed as Class I elevatedi~vels of by-products of ga~oline • under the Clean Air Act of I97"7,and degradation of combustion---- such as ammoniumand benzene have- jii;guality isprohibited. The quality of air is a growi~ been detected in snowpack samples. Annual concern because m-otorized winter visitation has snowpacks and snow-covered roadways trap increased nearly tenfold since 1968 (National Park deposition from local and regional atmospheric Service, 1990; Craig McClure, National Park Service, emISSIons. written commun., 1996). Air-quality monitoring at the West Entrance during the winter of 1995 detected S~owpack samples representing most of the carbon monoxide (CO) levels potentially exceeding winter precipitatIon were collected at about the time of maximum annual snow accumulation at a I:.ederal air-.9.~ality standards, which raised concerns about employee and visitor exposure to snowmobile variety of locations in and near the park to emissions (National ParKServfce, Air Quality Divi~ observe the effects of a raI}ge of_snowmobile slon, 1996)~frowever, at tl1is busiest entrance t;(he traffic levels. Concentrations of organic and inor• park, where -over 1,000 snowmobiles have entered on ganic compounds in s!!Qwsamples from pairs oJ peak traffic days, COc-oncentrations diminished atair• sites located directly in and off sllOw-paclsed . ---- ~~ 9uality sampling, sites 20 to 100 m away from the roadways used by snowmobiles were compared. t;,ntry point. Growing popularity of snowmobile use for times higher for the in-road snow compared to winter recreation in the Yellowstone National Park Concentrations of ammonium were up to three I off-road snow for each pair of sites. Thus, concen•J area has resulted in increasing numbers of snowmo• trations decreased rapidly with distance from biles operating in the park in the past decade (National roadways. In addition, concentrations of ammo• Park Service, 1990; Craig McClure, written commun., nium, nitrate, sulfate, benzene, and toluene in 1996). Local economies in towns adjacent to the National Park and National Forest lands benefit from snow were positively correlated with snowmobile use. the growing recreational attraction offered by snow• mobile touring and are dependent upon revenues generated by accommodating snowmobile recreation• alists. However, snowmobile use in Yellowstone has INTRODUCTION become a controversial issue in recent years because Because wintertime visitors to Yellowstone of the noise, traffic congestion, and air pollution National Park (fig. I) consider clean air to be an (Woodbury, 1997). important aspect of their experience in the park Increases in fossil-fuel combustion in the area (Littlejohn, 1996) and because vehicle emissions £..ouldcontribute to greater levels of emissisms enteJi.!1g~ degrade the air quality in Yellowstone, the park staff sensitive watersheds and animal habitats. Until have become increasingly concerned (Wilkinson, recently, concentrations and the extent of dispersion of
Abstract
I 116° 104° -J------I: ·······.. 0 ------i \ •.- 0":I 1I
\ , - ....'.:~:J MONTANA \ ...• /I •• '. '. - Yellowstone • ' .Q) \
V · Big Sky Park I ',.' '.' ~ National I ° _ \~o~h-e~~le. ~ Daisy !'!I~ \
45 I West Yellowston~(? _~i~ . @ ~:~:;~IS . Lewis OldLakeFai!h!UiDIvide (7(2 si.tes)sites) ~:" •F,ourSylvanMileLakesMeadow(2 sites) \ Garnet Canxoo •• Togwotee Pass I Rendezvous Mountain JacR~on Hole Airport I IDAHOTeton Pa:s •"','WYOMING \: --- I· \ ,------1I ..:'" 0;".I,.,. I I I •••••••••,:~ I
I L__ ". -- T Rab~it- Ear~: - - ~h~~S J--Creek I Pass •' •• vy.:0 Loch Vale
I I Supply Forks. :
I UTAH I COi:~RADO I . I I ... 37°'- I I
/------~- :.L· - -1- ("., ! ..:. NEW : ARIZONA i.······ MEXICO \ \ \
o 100 Miles I I I o 100 Kilometers
Figure 1. Location of snow-sampling sites in the Rocky Mountains in Western United States. many emissions from wintertime snowmobile traffic in corridors where snowmobiles operate are they Yellowstone were unknown. High CO levels measured detected? at the West Entrance, combined with increasing numbers of snowmobiles entering the park, raised the question of whether snowmobiles are causing Background increased levels of airborne emissions elsewhere in the park. If airborne snowmobile emissions are exten• Early studies of the effects from snowmobile sively distributed, at what distance from highway use in Minnesota (Wanek, 1971) and south-central 2 Effects of Snowmobile Use on Snowpack Chemistry in Yellowstone National Park, 1998 Canada (Neumann and Merriam, 1972) focused on nearby emission sources has been successfully demon• ecological effects and indicated that physical damage strated in northwestern Colorado (Turk and others, to ecosystems occurred as a result of over-snow vehic• 1992; Turk and Campbell, 1997). ular activity. As a result of the increased use of snow• I..he USGS, in c.ooperation with the National mobiles in the 1970's and 1980's, concerns about Park Service Air Resources Division, Denver, Colo• chemical emissions from a variety of mechanized rado, and the National Park Service staff, Yellowstone equipment used outdoors, including snowmobiles, NationarPark, Wyoming, has been monitoring atmos• prompted extensive research of engine emissions pheric deposition to snowpacks in and around Yellow• (Hare and Springer, 1974; U.S. Environmental Protec• stone National Park annually since 1993. During tion Agency, 1991). Collins and Sell (1982) detected 1993-98, chemicals in the snowpack, particularly elevated concentrations of lead along a snowmobile acidic compounds, were monitored annually trail in Wisconsin. Little work has been done, throughout a network of 50 to 60 sites in the Rocky however, to evaluate chemical deposition in snowpack Mountain region of the Western United States. The due to snowmobile use in the Western United States. network includes several sites in and near Yellowstone Numerous studies exist in the literature National Park. Patterns emerging from preliminary, concerning the occurrence of hydrocarbons such as the unpublished ch;-mical analyses of the seasonal snow- .• benzene, toluene, ethylbenzene, xylenes (BTEX), and packs in the Yellowstone National Park area during• methyl tert-butyl ether (MTBE) compounds in precipi• 1993-95 indicated that traffic vQlume otsnowmobiles tation, surface water, and ground water in urban might positively correlate with chemical concentra• hydrologic settings (Ayers and others, 1997; Bruce, tions of ammonium, nitrate, or sulfate. Thus, a pilot 1995; Delzer and others, 1996) and in less densely study to test the hypothesis that snowmobile use populated areas, especially in the Central and Eastern results in increased emission deposition relative to United States (Fenelon and Moore, 1996; Terracciano background levels was begun in 1996. Results from and O'Brien, 1997; Reiser and 0' Brien, 1998). the 1996 study of a small set of sites (three) demon• However, comparatively few studies concerning strated a positive correlation between the level of hydrocarbons in precipitation and water have been snowmobile traffic and chemical concentrations of done in the Rocky Mountain region (Dennehy and ammonium and sulfate (Ingersoll and others, 1997). In others, 1998), and even less is known about the occur• a separate study, snowmobile engines were found to rence of hydrocarbons in snowfall (Bruce and emit ammonia and sulfur dioxide in laboratory tests McMahon, 1996). (White and Carroll, 1998). The U.S. Geological Survey (USGS) has been monitoring the chemical composition of annual snow• packs in Colorado since the mid-1980's. Elevated Purpose and Scope levels of emissions from atmospheric deposition held in seasonal snowpacks have been indicated by chem• Building on the results of the 1996 pilot study, ical concentrations of species associated with water• the USGS, in cooperation with the National Park shed acidification (including nitrate and sulfate) at ~ervice, conducted a further investigation of the rela• alpine and subalpine sites in Colorado (Turk and tion between snowmobile use and snowpack chemistJY_ Campbell, 1987; Campbell and others, 1991). in Yellowstone National Park in 1998. The purpose of Comparisons between chemical concentrations down• t~e 1998 study was to determine if emissions from_ wind of or nearer to possible emissions sources and snowmobile traffic are detectable in a larger sampling those located crosswind or farther away from the network of seasonal snowpack and to investigate. source have revealed trends in snowpack chemistries whether emission levels tend to diminish rapidly -.yith. in Colorado (Ingersoll, 1996; Turk and Campbell, distance from the snowmobile thoroughfares as the _ 1997). At sites where atmospheric deposition was emissions disperse into the surrounding watersheds:JL affected by local emission sources, concentrations of dispersion is limited to a short distance from local acidic species were higher than at locations more sources such as snowmobiles, then two key results distant from the source (Ingersoll, 1995). This tech• emerge. Watershed-scale-- effects--from ----winter traffic nique of using annual snowpack chemistry to identify are unlikely, and further identification of regionally
INTRODUCTION 3
I influenced deposition of emissions may be possible National Park and nearby in Montana and Wyoming WIthout interference from locally-generated snow• (fig. I, table I). Two additional sites in Colorado mobile emissions. (Loch Vale and Supply Forks), where preliminary Knowledge of the extent of the deposition to sampling and protocol development was done, also annual snowpacks of acids and other compounds asso• were selected to complement the study. Conifers ciated with fossil-fuel combustion is important to park dominate the mostly undeveloped mountainous land• preservation; it also is important to establish chemical scapes at snow-sampling sites, facilitating sample baselines that can be used in evaluations that affect collection at locations where snowfall tends to accu• policy decisions controlling vehicle trafflc in protected mulate uniformly. Elevations of sampling sites range areas. Thus, snowpacks that represent a spectrum ot' from 2,035 m above sea level at West Yellowstone, .:-nowmobile use at locations throughout-~tone Montana, to about 2,570 m near Sylvan Lake in were sampled during March 1998 to measure wintet• Wyoming. At these elevations, the seasonal snowpack time deposition of selected org~inic and-inorganic by• is maintained throughout the winter into spring, and products of gasoline combustion. substantial snowmelt usually does not occur until In this report, a method using snowpack chem• spring runoff. istry as an indicator of snowmobile emissions is described. No permanent constructs, instrumentation, or maintenance were needed; data collection required Acknowledgments minimal environmental impacts and required only one annual visit per site. A single snow sample from the This research was made possible by the efforts annual snowpack provided data on the chemistry of many from the National Park Service and Yellow• stone National Park. The author also would like to representing most of the yearly precipitation. With this method, a cost-effective sampling network was readily express gratitude for the assistance with this project established to meet the objectives of the study, and the from Mary Hektner, Kristin Legg, Craig McClure, and only constraints were that sampling sites were located John Sacklin of Yellowstone National Park; and Kristi where seasonal snowpacks persist and safe accessi• Heuer and Kathy Tonnessen of the National Park bility (low avalanche hazard) was likely. Service, Air Resources Division, Denver, Colo.; and the District Ranger staff at Yellowstone National Park.
Description of Study Area SAMPLE-COLLECTION LOCATIONS AND The Yellowstone Plateau in southwestern METHODS Montana and northwestern Wyoming has a dry, mid• continental climate but frequently is in the path of the Snow-sampling sites were carefully chosen at winter jet stream that brings abundant moisture, locations where the annual snowpack typically persists mostly from the Pacific, to the mountains (Martner, from October through March so that chemical solutes 1986). The area has strong winds and frequent storms held in the snowpack structure that represent seasonal that deliver precipitation preferentially to elevated atmospheric deposition could be captured. Seasonal areas; surrounding plains areas are dry by contrast. snowpacks provide a composite record of atmospheric Annual snowpacks range in depth from 0.5 to about deposition of airborne emissions throughout winter if 4 meters. The high average elevation (more than 2,000 no substantial melt occurs before sample collection. m) and latitude of the area maintain cool temperatures Obtaining snow samples before melt begins is crucial and cause most annual precipitation to fall as snow to preserving the chemical record of the snowpack (Paulson and others, 1991). As in many subalpine because a concentrated elution, or ionic pulse, begins continental settings in the Rocky Mountain region, early in the annual snowmelt process (Campbell and snowcover is present about 6 months each year, gener• others, 1995). Sampling within 2-3 weeks before melt ally from October until March or April. begins permits data for most yearly snowfall to be The study area includes 28 snow-sampling sites collected in a single sample, and seasonal chemical at selected high-elevation locations in Yellowstone deposition remains intact.
4 Effects of Snowmobile Use on Snowpack Chemistry in Yellowstone National Park, 1998
I Table 1. Snow-sampling site locations and estimated snowmobile use levels in the Greater Yellowstone area and in northern Colorado
[Estimated use levels based on numbers of snowmobiles passing sites per day (ranging from 0 to over 1.000 snowmobiles per day). local observations. and restrictions. Categories from very low to high are estimated as less than 10. 10 to 99. 100 to 499. 500 or more snowmobiles per day for very low. low. moderate. and high levels. respectively: m. meters: km. kilometersl
Location (50- and 1,OOO-meter distances measured unless Estimated level of Site name otherwise indicated; all other distances approximated) snowmobile use Big Sky Big Sky ski area. Montana. mid-mountain. off-road low Biscuit Basin 3 kilometers north-northwest of Old Faithful geyser. 50 meters ofT-road moderate-to-high Biscuit Basin (in-road) 3 kilometers north-northwest of Old Faithful geyser. in snowpacked moderate-to-high roadway paired with off-road site Canyon I kilometer west of snowpacked roadway. and 3 kilometers south-south• very low (if any) west of Canyon. Wyoming. and ahout 50 meters ofT utilities service road Daisy Pass 200 meters south of Daisy Pass. Montana. and 100 meters west of snow• low packed roadway Four Mile Meadow I kilometer north of plowed US Highway 287 and 10 kilometers west of low Togwotee Pass. Wyoming Garnet Canyon 3 kilometers west of plowed Jenny Lake road and 25 kilometers north of very low (if any) Jackson. Wyoming Jackson Hole Airport About 50 meters north of north edge of visitor parking lot and very low (if any) 100 meters east of active runway Lewis Lake Divide 5 kilometers south of Lewis Lake and 50 meters west of snowpacked moderate roadway Lewis Lake Divide (in-road) 5 kilometers south of Lewis Lake in snowpacked roadway paired with moderate ofT-road site Lionhead 2 kilometers northwest of Targhee Pass. Montana-Idaho and about 50 low-to-moderate meters from snowmobiling area Loch Vale 15 kilometers southwest of Estes Park. Colorado. and ahout 3 kilometers very low (if any) ofT-road
Old Faithful (1.000 01) 1.000 meters south of Old Faithful. Wyoming. ofT-road along Fern very low (if any) Cascade trail Old Faithful Corrals I kilometer west of Old Faithful. Wyoming. and 50 meters ofT-road low Old Faithful East Lot 500 meters southeast Old Faithful geyser and 50 meters off-road low Old Faithful 300 meters southwest Old Faithful geyser and 50 meters off-road high Old Faithful (in-road) 300 meters southwest of Old Faithful geyser in snowpacked roadway high paired with off-road site Rendezvous Mountain 500 meters northwest of summit of Jackson Hole aerial tram. out of very low (if any) hounds Supply Forks 5 kilometers northwest of Grand Lake. Colorado. and 50 meters off-road moderate-to-high Sylvan Lake I kilometer west of Sylvan Pass. Wyoming. and 50 meters ofT-road low Sylvan Lake (in-road) I kilometer west of Sylvan Pass. Wyoming. in snow packed roadway low paired with ofT-road site Teton Pass 2 kilometers west of summit of Teton Pass and about 50 meters ofT-road very low (if any) Togwotee Pass 25 kilometers west of Moran. Wyoming. and about 100 meters northeast low-to-moderate of US Highway 287 Tower Falls I kilometer south of Tower Falls junction and about 50 meters cast of very low snowpacked roadway 21-Mile About 21 miles (33 kilometers) north of West Yellowstone. Montana. very low (if any) and 150 meters west of plowed US Highway 191 West Yellowstone 100 meters east of West Entrance. Yellowstone National Park. and high 50 meters north of snowpacked roadway West Yellowstone (in-road) 100 meters east of West Entrance. Yellowstone National Park in snow• high packed roadway paired with off-road site
SAMPLE-COLLECTION LOCATIONS AND METHODS 5 I Table 1. Snow-sampling site locations and estimated snowmobile use levels in the Greater Yellowstone area and in northern Colorado-Continued
[Estimated use levels based on numbers of snowmobiles passing sites per day (ranging from 0 to over 1,000 snowmobiles per day), local observations, and restrictions. Categories from very low to high are estimated as less than 10. 10 to 99. 100 to 499, 500 or more snowmobiles per day for very low, low, moderate. and high levels. respectively; m. meters; km, kilometers I
Location (50- and 1,ODD-meterdistances measured unless Estimated level of Site name otherwise indicated; all other distances approximated) snowmobile use West Yellowstone ( I,000 m) 100 meters east of West Entrance, Yellowstone, and 1,000 meters north very low (if any) of snowpacked roadway West Yellowstone, 8 km east 8 kilometers east of West Yellowstone, Montana, and 50 meters ofT high snowpacked roadway West Yellowstone, 8 km east 8 kilometers east of West Yellowstone, Montana, and in snowpacked high (in-road) roadway paired with off-road site
Snowpack sampling sites along or near snow• closure dates for most segments of highways groomed packed routes were included in this study to represent for snowmobile traffic in the park. various levels of over-snow traffic (fig, I), Sites at During the 3 months before sampling the West Yellowstone, Old Faithful, and Sylvan Lake were Yellowstone snowpack, sampling methods used in this selected to repeat the pilot study of 1996. Additional study at Yellowstone were tested in Colorado in condi• sites were selected along the highway over Lewis Lake tions similar to those at Yellowstone to identify poten• Divide where snowmobiles are typically operating at tial sources of contamination during collection and cruising speeds; along the roadway 3 km from Old transportation of snow samples. Two sites were Faithful and 8 km east of West Yellowstone where selected: (I) Loch Vale in Rocky Mountain National cruising speeds also are typical; and at a location about Park, Colorado, representing background conditions 20 km from snowmobile routes near Tower Falls. At where motorized travel is not permitted within 3 km, snowpit-sampling locations a sample was taken at and snowmobile use is restricted to areas at least least 50 m off the nearest roadway in undisturbed 10 km away, and (2) Supply Forks near Grand Lake, forest, and at several sites the off-road sampling loca• Colorado, where regular snowmobile use offers the tion was paired with a location in the active snow• opportunity to sample both the snow-covered roadway covered roadway (West Yellowstone, 8 km east of and directly adjacent off-road snowpacks using the West Yellowstone, Biscuit Basin, Old Faithful, Lewis methodology described in the pilot study of Ingersoll Lake Divide, and Sylvan Lake). In-road snowpits were and others (1997). Certified blank waters (with docu• positioned directly in snow packed by snowmobile mented low ionic strength) for both inorganics and traffic. At about 1,000 m from snowmobile roadways, organics analyses were used in field blanks, laboratory two additional off-road snowpits were established blanks, and trip blanks collected to measure contami• adjacent to snowpits that were 50 m off-road at West nation during collection of snow samples. Laboratory Yellowstone and Old Faithful. blanks are used to test for contamination during Snowpacks in and near Yellowstone National analytical processing; trip blanks are kept sealed in Park were sampled during the period March 4-24, transit with samples to see if contamination of sealed 1998, when snow-water equivalent values at snow• samples might be occurring during transport. At both telemetry sites at sampling locations were 87 to sites, replicate snow samples were collected to test 88 percent of the 30-year (1961-90) averages (Natural reproducibility of results, and field blanks were Resources Conservation Service, 1998), Snow depths collected by pouring certified blank water directly over at sampling sites ranged from 12 to 132 cm at in-road locally cleaned sampling tools and into Teflon sample sites and 63 to 221 cm at off-road sites. Snow depths bags to check for contamination from collection in packed roadways were substantially less than snow equipment. Precautions were taken to prevent contam• depths at unpacked off-road sites because of frequent ination because of the extremely dilute chemistry of compression from grooming machines and daily snow. Further testing was done to detect possible snowmobile traffic. Field sample collection was corruption of samples due to nearby snowmobile use accomplished within just a few days of the season during snow sampling. Also, after investigators trav-
6 Effects of Snowmobile Use on Snowpack Chemistry in Yellowstone National Park, 1998 eled to the sampling sites by snowmobile, they donned presented in this report were collected before snow• Tyvek suits over their outer garments as an additional melt began. precaution against contamination. Workers wore latex Snowmelt runoff was sampled at five locations gloves when collecting samples. The bottom few during the first half of May 1998. Runoff from the centimeters of the snowpack were not sampled to melting snowpacks near snow-sampling sites was avoid inclusion of forest litter and soils in the samples. collected at West Yellowstone, Old Faithful, Lewis The top few centimeters of snowpack also were Lake Divide, Sylvan Lake, and Tower Falls. Snowmelt discarded to exclude snow contaminated by workers' runoff samples for inorganics were collected in activities resulting from transport to and preparation of Dl-rinsed, 2-L polyethylene bottles, refrigerated, and the snowpit. Clean plastic shovels and scoops (scoured shipped overnight to the USGS laboratory in Denver, and rinsed with local snow) were used to collect a Colorado. Detection limits for inorganics in snowmelt• vertical snow column for each sample that represented runoff samples were the same as mentioned above for the entire snowpack to be analyzed. Workers in the snow samples except for ammonium (2.0 Ileq/L). snowpits took precautions to avoid inclusion of Snowmelt runoff samples for organics were collected foreign substances such as soils, tree litter, clothing, in precleaned and sealed, amber, 40-mL glass vials fit saliva, or perspiration when filling sample containers. with caps designed for sampling YOC's. The snow samples were cut and placed in 8-L Teflon bags that were prerinsed in pure (electrical resistance of 18 megohm/cm) deionized water (Dl). Samples for ANALYTICAL PROCEDURES analysis of organic constituents were collected in precleaned, 40-mL amber glass vials designed for Snow samples (8 L each) were allowed to melt collection of volatile organic carbon (YOC). These in the Teflon collection bags and were processed Teflon bags and vials were sealed against contamina• within 12 hours by using a series of analytical proce• tion in plastic bags and cases and transported frozen to dures. Samples were kept cool by processing them soon after melting or refrigerating snowmelt aliquots USGS laboratories. Inorganic analyses were done at USGS Regional Research Laboratory in Boulder, reserved for subsequent analyses. Major cation concentrations were determined on filtered (0.45 mm), Colorado, and organic analyses were done at the acidified aliquots by using two emission-spectroscopy USGS National Water Quality Laboratory in Arvada, techniques-inductively coupled plasma [for calcium Colorado. Snow samples were kept frozen at -10°C to (Ca2+), magnesium (Mg2+), and sodium (Na+)] and prevent chemical reactivity until the actual dates of atomic absorption [for potassium (K+)]. Chloride (Cn laboratory analyses. and sulfate (S042-) concentrations were determined Snow samples were collected from snowpits on filtered (0.45 mm) aliquots by ion chromatography. prepared with a smooth, freshly cut, vertical face Nitrate (N03 -) and ammonium (NH4+) ion concentra• extending from the ground surface upward throughout tions were analyzed on filtered (0.45 mm), frozen the entire depth of the snowpack. Before snow subsamples by air-segmented, continuous-flow color• samples were collected, physical measurements of the imetry. Detection limits were as follows, in snowpack were made to ensure solute loss had not microequivalents per liter (Ileq/L): Ca2+, 0.5; Mg2+, occurred. Full-snowpack temperature profiles were 0.5; Na+, 1.0; K+, 0.3; Cl-, 0.5; S042-, 0.4; N03-, 0.2; recorded at 10- or 20-cm intervals to ensure tempera• and NH4+,0.5. Specific conductance was measured in tures below zero degrees Celsius predominated among microsiemens per centimeter at 25°C with a platinum snow strata. Snow-crystal size, type, and hardness of electrode; pH was determined in standard units with all homogeneous strata were measured to document a combination glass electrode designed for low-ionic• the history of the metamorphism of the snowpack strength waters. Alkalinity was determined using a during the winter. Further observations of a lack of Gran titration method. evidence of melt, saturated wet snow, and soil mois• Quality control involved systematically ture beneath the snowpack ensured snowmelt elution analyzing deionized-water blanks, an internal had not yet begun and that the snow to be collected reference sample, and USGS standard reference water maintained the seasonal atmospheric deposition in an samples. Water from Chaos Creek in Rocky Mountain ice phase. Snow samples from all sampling sites National Park, Colorado, was selected as the internal
ANALYTICAL PROCEDURES 7 I reference sample to monitor instrument precision from the snow-covered roadway at four sites along because of its similarity of ionic strength compared to snowmobile routes in Yellowstone form a distinct snowmelt from the sites reported here. Approximately cluster in the upper right corner of the plot in flgure 2 40 percent of sample batch run time for the analytical and demonstrate the elevated levels of these instrumentation was dedicated to analyzing fleld compounds in snow-packed roads in Yellowstone rela• blanks, laboratory blanks, trip blanks, duplicates, tive to 50 to 60 other snowpack-sampling sites in the internal reference samples, and USGS standards. Cali• Rocky Mountain region. Ammonium and sulfate were bration veriflcations were made with standards at the shown to be good tracers of snowmobile emissions in beginning and end of each batch of sample analyses. the pilot study of snowmobile effects on Yellowstone Processing blanks were analyzed to detect possible snowpack chemistry (Ingersoll and others, 1997). contamination from DI rinse water, filtering apparatus, Concentrations of ammonium and sulfate in snow• and the Teflon collection bags. Charge balance was covered roadways were well correlated in both high calculated between total anions and cations to check and low traffic levels in that study of the 1996 snow• the quality of the analyses. pack at three locations in the park. Analyses of the selected organics benzene, Snow samples from the sites in Colorado that toluene, ethylbenzene, and xylenes (BTEX), and represent locations popular for snowmobile use were methyl tert-butyl ether (MTBE) were accomplished analyzed and compared to ofT-road, untracked snow using purge and trap/selected ion monitoring and gas during December 1997 and January 1998, before chromatography/mass spectrometry techniques. Snow sampling began in Yellowstone in March 1998 samples were kept frozen in transit from sampling (table 2). The Supply Forks site in Colorado proved to sites then allowed to melt in a refrigerator at 4°C. be a good site to test sampling protocols for detecting Liquid samples were then composited and acidifled snowmobile emissions in snow because it is in a with I: I concentrated hydrochloric acid. Snow popular snowmobiling area, and the same pattern of samples were analyzed within 14 days of melting. emissions of ammonium and sulfate were observed Field and laboratory blanks of certifled, nitrogen• there as in Yellowstone. Comparisons of the in-road purged organics blank water, trip blanks, and field and samples at the Supply Forks site near Grand Lake, laboratory replicates were analyzed for quality control Colorado, to samples obtained 50 m off-road reveal a and evaluation of the precision and contamination of trend very similar to that seen in the 1996 Yellowstone sample-collection and analytical techniques. pilot study. Elevated concentrations of ammonium (4.7 Numerous analyses of spiked reagent waters were to 17.1 Ileq/L) and sulfate (16.0 to 74.9 Ileq/L) were done to evaluate instrument performance and establish detected in snow at the in-road site; consistently lower method detection limits. Detection limits for organics concentrations of ammonium (2.8 and 3.9 Ileq/L) and analyses, in nanograms per liter (ng/L or parts per tril• sulfate (7.1 and 9.6 Ileq/L) were detected in snow lion), were 1.7 (benzene), 5.0 (MTBE), 4.8 (toluene), 50 m off-road and directly adjacent to the in-road 3.8 (m- and p-xylene), and 1.6 (o-xylene). Minimum sampling site. At the backcountry Loch Vale site in reporting levels for these organic compounds in this Colorado, about 3,000 m from wheeled-vehicle traffic study are at least twice the method detection limits. and at least 10,000 m from snowmobile traffic, simi• larly low levels of ammonium (4.5 to 4.9 Ileq/L) and sulfate (9.2 to 9.4 Ileq/L) were detected. Nitrate SNOW PACK CHEMISTRY concentrations were fairly consistent (14.3 to 17.1 Ileq/L) in all in-road, off-road, and backcountry snow samples except for the "exhaust" sample (29.3 Ileq/L), which was collected from the snow directly beneath Major Ions (20 cm) the exhaust pipe of a snowmobile after its Ammonium and sulfate ion concentrations engine idled for 5 minutes. weakly correlate (r2 = 0.12) in the Rocky Mountain TJle results for the inorganic majo,! ion analyses region, but when snow samples from areas of high exhibited higher concentrations of most constituents, snowmobile use are included, the association becomes especially ammonium and sulfate, in areas of high stronger (r2 = 0.56) (flg. 2). Values for these samples snowmobile use (fig. 3). A selected set of sampling
8 Effects of Snowmobile Use on Snowpack Chemistry in Yellowstone National Park, 1998 25.0
••.. ~CII ~ 200 CII • C• -UI c:: IiiCII .~ 5- 15.0 CII •..o .!:! -S UI c:: ~ 10.0 •..III 'E uCII c:: uo CII 10 5.0 ::j (/)
0.0 0.0 5.0 10.0 15.0 20.0 25.0 Ammonium concentration (microequivalents per liter)
Figure 2. Relation between ammonium and sulfate in mountain snowpacks throughout the Rocky Mountain region, 1998.
sites where both inorganic and organic analyses were West Yellowstone, Biscuit Basin, Old Faithful, Lewis performed is shown in figures 3, 7, and 8. See table 3 Lake Divide, and Sylvan Lake). At the six paired sites for inorganic chemistry and table 5 for organic chem• where both in-road and off-road samples were taken, istry of sites not shown in figures 3, 7, and 8. Consis• the in-road value is shown just below the corrc• tent with the results of the pilot study in 1996, similar sponding off-road value. Traffic volumes were greatcst patterns for ammonium, nitrate, and sulfate concentra• at West Yellowstone, where 1,000 or more snow• tions in snow emerged with analyses of the 1998 mobiles commonly enter the park daily; substantial snowpacks. In-road snow samples were significantly numbers of those vehicles include Old Faithful as a hi~her in aminonium (4.0 to 24.5 )Jeq/L, p = 0.007) destination. Concentrations of ammonium and sulfate and sulfate (4.6 to 21.2 )Jeq/L, p = 0.002) than nearby in snowpacks generally were higher at the in-road sites (within 50 m) oft'-road concentrations of ammonium bctween West Yellowstone and Old Faithful than else• (13 t9...9.1 )Jeq/L) and sulfate (3.3 to 6.Q )J_eg/L) where (figs. 4 and 6). On thc West Entrance road, high (table 3). The p-values of 0.007 and 0.002 indicate t;;ffi~ ~olumes typically result in much slower speeds 99.3 percent and 99.8 percent probability of the values (20-30 mph) and considerable acceleration and decel• being significantly higher, respectively. The original eration. In addition, groups stop on the snow road edge sample at West Yellowstone (in-road) is shown in or in groon£d turnouts to obserYeWiICWfc or the figure 3; table 3 includes an additional replicate Madison River or to reorganize their partt On the sample at that site which does not appear in figure 3. South and East Entrance roads, lighter volumes mean Spatial patterns of concentrations of key major that traffic moves more smoothly at higher speeds ()S• ions are shown in figures 4-6. Within the park 45 mph). Visitors typically do not stop at Lewis Lake boundary, six sets of in-road and adjacent off-road Divide or Sylvan Lake (John Sacklin, National Park samples are shown (West Yellowstone, 8 km east of Service, written commun., 1999). Considerably""------fcwer
SNOWPACK CHEMISTRY 9 I Table 2. Snow chemistry of selected inorganic compounds at protocol-development sites
[units: pH. standard: ammoniulll. sulfate. and nitrate. in microequivalcnts per liter: nJa. not applicable]
14.3 Site Date pH Ammonium<0.225.126.274.915.014.316.415.7~4.974.958.229.37.115.716.09.6<0.2Nitrate17.217.117.112.14.95.00.62.88.63.90.94.77.3Sulfate ~ 4.84 9.2 ) ~:.4 Supply Forks6.345.614.074.805.563.914.78n/a4.764.72 12/17/97 4.5 ) Supply Forks3.91(in-road) 12/17/97 Supply Forks (field blank) 12/17/97 Loch Vale 12/16/97 Loch Vale (replicate) 12/16/97 Supply Forks (field blank) 01/15/98 Supply Forks 01/15/98 Supply Forks (in-road) 01/15/98 Supply Forks (in-road. replicate) 01/15/98 Supply Forks exhaust 01/15/98
Minimum' Maximum' Mean (median for pH)! Standard deviation ISUlllmary statistics do not include blanks
s,!1owmobiles enter the park from the South or East mobile traffic. To the south. Garnet Canyon and Eptrances and pass Lewis or Sylvan Lakes enroute to Rendezvous Mountain in the Tetons and Four Mile their destinations, which also is Old Faithful in Meadow and Togwotee Pass in the Wind River Range most cases. During the sampling period of March had background levels typical of off-road snowpacks 1998, snowmobiles during the day at Old Faithful where both local and regional emission deposition is commonly numbered in the hundreds, which was 5 to minimal. 10 times more than elsewhere in the park. 1/1winter of 1998,-4Q~9 snowmobiles came in the West Entrance; 15.209 came in the South En_trance; 2,563 Selected Hydrocarbons came-in the East Entrance; and 1,469 came in the North (Mammoth Terraces) Entrance, for a total of ~en~ene, toluene, and xylenes are VOC's and are by-products of gasoline combustion and have been 60.110 snowmobiles. Tot~ recreation visitors in ~nter 1998 were 119,274, of which 72,834 came __ detected in snowmobile emissions (White and Carroll. ionon snowmobiles, 9,897 were in snowcoaches, and 1998: Morris and others. 1999). MTBE is a gasoline 40, 101 were in automobiles (National Park Service, additive used in Colorado. but generally not found in Yellowstone National Park, written commun., 1999). gasoline used in the Yellowstone area. These hydro• carbons are common in urban areas (Lopez and Nitrate concentrations (fig. 5) are minimally affected Bender. 1998) but have been detected in this study far ~y snowmobile traffic and probably reflect regional from urban development.. Concentrations of these ?epo.sition from sources other than snowl!!0biles. VOC's in.seasonal snowpacks at the protocol-develop• Also worthy of note are the similarities in ine_ntsites in Colorado were higher for in-road snow concentrations at the low- or moderate-traffic sa.r:nplesthan for off-road samples by factors of 2 to sampling sites at Lewis and Sylvan Lakes relative to UlOO (table 4). The ~nge of in-road c0!.1centrations other sites in and near the park in areas of reduced or fo..!:..benzene (37.419 1,820 ng/Lf MTBE (\ 0.1 to restricted snowmobile use like Canyon, Tower Falls, 374 ng/L), to~e (577 to 9,880 ng/L},.Jn- and p• and Big Sky ski area, where snowmobile travel is xylenes (640 to 14.100 ng/L). and o-xylene (377 to minimal. West of the park near Targhee Pass, the area "'0Q~g(L)~as substantially greater in most cases around the Lionhead site has low-to-moderate snow- than the range of ofT-road sample concentrations of
10 Effects of Snowmobile Use on Snowpack Chemistry in Yellowstone National Park, 1998
I In-road. sampling sites Off-road and background sampling sites =I' 25 15205 = ~ rz=::. 10 .~ ~ - ~~-[J~ (.)CI) CI) 0a.c: I f= CI) ':5Uiijc:
I!!!!!!!!
YellowstoneYellowstoneBiscuit(8Sylvan21.916.2(8(in·road)Faithful21.119.64.84West4.22.7Tower6.24.3DivideYellow·West(in-road)(in·road)Basinstone68West9.1Lake6(in·road)4.6Lake7.824.5kmOld17.5kmFallseasteast)YellowstoneFaithful(1,0002320.6(1,OOOm)Old6.44 Westm) Lewis stone (in·road) in·road) Yellow· o ammonium Figure• sulfate3. Ammonium and sulfate levels in snow at in-road and off-road sites in Yellowstone National Park, 1998.
benzene (less than 10 ng/L), MTBE (less 1l1illll.9 to concentrations of other VOC's including toluene 10.1 ng/L), tolu~nec88.9 to 265 ng/L), m- and (65.3 ng/L). This indicates minimal interference from p-xylenes (13.7 to 22.8 ng/L), and o-xylene (5.4 to snowmobiles passing in-road sampling snowpits while l2.9 ng/L). -- -- samples were collected. Another exhaust sample Initial protocol development, including the (January 15, 1998) was collected in snow at the processing of field blanks, laboratory blanks, and trip Supply Forks site directly beneath the exhaust pipe blanks, indicated a sensitivity to contamination near after a snowmobile engine had been running at about nondetect limits in three of six blanks, but inexplicably 2,000 revolutions per minute for 5 minutes. This direct high contamination from toluene was detected iii-the deposition of exhaust emissions resulted in high -fielClblank from 12/17/97 (1,570 ng/L) and a trip concentrations of toluene and xylenes relative to other I5HinKfrom 0 I/22/98 (1,400 ng/L). Many hydrocar• in-road samples shown in table 4. The Supply Forks bOi1S'aredetected all over the worid, including both in-road sample of January 15, 1998, also had relatively poles and mid-latitude alpine zones of North America high concentrations for benzene, toluene, and xylenes. I (Blais and others, 1998). Even though gasoline• When that sample was collected, yellow and brown , powered vehicles were used to transport the sampling discoloration was observed in the snow pack, possibly J ~ IJ' 1 .;1 personnel and equipment for this study, these two due to spilled oil or fuel in the snowpacked roadway. toluene contamination levels are very high for blank Thus, contamination of that sample was likely, and samples. Anomalously high concentrations of toluene concentrations of both inorganic (table 2) and organic (relative to other hydrocarbons detected) are shown for compounds (table 4) may not reflect atmospheric three Yellowstone sites located 50 to 1,000 m ofT-road deposition because of possible sampling of fuel (fig. 8). The source of toluene contamination is spillage. Other in-road-sample concentrations of currently under investigation. hydrocarbons at the protocol development site were A snow sample collected 5 m from an idling similar to each other. snowmobile at the Supply Forks site had a fairly In the Yellowstone area in March 1998, samples elevated concentration of MTBE but relatively low collected for VOC analyses at 12 selected sites had the
SNOWPACK CHEMISTRY 11 I Hydro-4.944.854.855.014.945.105.004.995.045.195.235.065.\95.185.03n/a5.475.065.205.455.315.525.5.095.035.175.325.085.155.245.525.155.\924.521.421.912.417.513.\13.919.6\4.410.510.9sium16.211.214.44.24.04.4nium0.10.80.04.80.93.321.121.224.520.63.63.02.34.63.53.7ride2.723.02.0Ammo-0.86.00.14.79.14.34.84.44.55.75.86.84.24.15.4Chlo-5.02.53.00.07.41.77.02.39.70.98.714.75.32.66.21.\4.94.012.43.33.22.79.29.47.310.63.\8.86.33.42.03.59.57.96.47.17.85.98.40.47.73.710.02.28.18.67.6Nitrate2.811.01.11.413.015.415.313.51.510.51.01.71.31.5\1.210.0\4.111.59.1\0.24.114.1Calci-8.06.22.53.32.3urnSod4.903.49.88.77.96.33.53.09.16.59.35.92.58.16.87.\4.88.37.15.86.6 i-22\212130147124179172118144132110178101128n/a469685206463738670298912 Date 03/06/9803/05/9803/03/9803/19/9803/07/9803/04/9803/08/9803/16/9803/18/9803/24/9803/09/9803/06/9803/03/17/98I\9/980/98 54.2 Table 3.SulfateMagne-:uSnowpHDepthchemistry of selected inorganic22110512 compounds in the Greater Yellowstone area 8 km east gen Twenty-oneWestTogwoteeMaximum'StandardOldJacksonCanyonTowerSylvanLewisGametDaisyFourBiscuitWestMeanOld(blank)(in-road)(in-road)(1.0008FaithfulkmFaithfulMileIYellowstoneSummaryYellowstonePassLakeFalls(medianCanyonLakeBasinHolecastdeviationPassm)MeadowDivideMileCorrals(replicate)East(in-road)(1.000AirportstatisticsforLotp~!)'m)BigdoSkynot include blank. TetonWestRendezvousLionheadBiscuitYellowstonePassBasinMountain Minimum'[units: depth.Sitecentimeters:name pH. standard units: hydrogen through nitrate. in microequivalents per liter: m. meters: km. kilometers: n/a. not applicahle]
12 Effects of Snowmobile Use on Snowpack Chemistry in Yellowstone National Park, 1998 oBig Sky MONTANA W. Yellowstone-8 km
W.in-roadYellowstone-8 km O~• \ .\... .
W. Yellowstone •. ~ Mile~ • '"""" • ~ North Entrance Road w. Yellowstone .e owstone ational W. Yellowstone 0O~ .f~T~enty-~~;( - - T~:;;a·lls - - ~ _DaiSY_~a~ 1,000 m,--.1" , ; Park in-roadLionheadr q,. ~" "-- II 'N· I ( East / -'~ \, I .... Roa~ ~ 0BISCUIt BaSin r . r0 Sylvan Lake at Old - -L. . d West Entrance Road -r' see inset - - _ Entrance~\ 0 Sylvan Lake , ~.'V', .. / I ' "right Faithful (. - In-roa Lewis Lake Divide " ....___SouthEntrance Road • BaSin Faithful Bis~uitBiscuit tr GeyserOld Basin ------~ewis Lake Divide ~ 1__" in-road GOld Faithful , " CorralsCrew 0'-) ~ ------Ol~In-roadFaithful In-road 1-., ;r"'if. ~ ~. _." ® East " Old Faithful 0 Lot IDAHO ! " 1,000 m . Snow~it locations 3 Grand near Old Faithful l OFourMeadowMile 1 \ Teton . .i o Togwotee Pass Garnet ICanyo~, , ·ON.&(. . RendezvousMountain EXPLANA TION Ammonium ion concentration, in microequivalents per liter (minimum value 2.3; maximum value 24.5) o 15 30 MILES 2.3 - 4.9 I I I I J o o 15 30 KILOMETERS 0··". 5.0 - 9.9 o 10.0 -14.9 o 15.0 -19.9 • 20.0 - 24.5 Figure 4. Ammonium ion concentrations in snow in Teton-Yellowstone area, 1998. SNOWPACK CHEMISTRY 13 I oBig Sky MONTANA W.Yellowstone-8 km 0. ~if ------Daisy Pass in-~~d • f'\._'"' /'O""'""~'"~, W. Yellowstone O. Ye owstone National ,!,. Yellowstone •~ IYMlle r-\.:. 1,000 m 'V Park In-roadLionhead-1', q ~ k U Canyon• r '(. East / ~ '0 Sylvan Lake ~-~West Entrance Road~"-1-''V' /E~~a:C;r-'_~< 0~ylvan Lake right Faithful rr - - In.:!:.oad ;; "s B' 't ,- Lewis Lake Divide " ...-- outh Entrance Road' 1SC:U1 1 91d Lewis Lake Divide " Biscuit Geyser Basin Faithful •• O~ in-road OOld Fai!hful in-road 0" Basin ~ 1-I "Corrals ) ~ In-road '. ;r .....::::--.-" . Crew 0' -;::::Ol~0Faithful IDA H 0 ,I }\ " '!------?~O~~hfUI 0 't~~t I _ Grand l Snow~it locations I \ Teton . Four Mile near Old Faithful , J N P. • ...J 0Meadow Garnet Fanyo~O, • y' 0 Togwotee Pass Rendezv~usMountain /'r\xq''LJ - ~Jac kson I0 Teton Airport WYOMING : Pass EXPLANA TION Nitrate ion concentration, in microequivalents per liter (Minimum value 4.7; maximum value15.4) o 15 30 MILES 4.7 - 6.9 I I 1 I ) o o 15 30 KILOMETERS o 7.0 - 8.9 o 9.0 - 10.9 o 11.0 - 12.9 • 13.0-15.4 Figure 5. Nitrate ion concentrations in snow in Teton- Yellowstone area, 1998. 14 Effects of Snowmobile Use on Snowpack Chemistry in Yellowstone National Park, 1998 , oBig Sky MONTANA W.in-roadYellowstone-8 km • .~ • W. Yellowstone-8 km O~ 450 L ~ ~ .~ .----:::North Entrance Road • W. Yellowstone • - Mile w. Yellowstone 0 e ow in-road1,000 m· l"VtIi II Lioj~a~t:q, •~ Canyon ) lEast / "v \ I Roadr '~ . ,-0 Sylvan Lake at - -4- _ '- In-road West\Entrance Road _'' 'V'seeinset / Entrance~\ • 0 ~ylvan Lake • I right aithful rlr - - _ L . L k t)" 'd ._____SouthEntrance Road' ISCUI Old Lewis LakeI DivideI " --" B·Bisc!Jit't Ii'Geyser Basin ------.eWls a e IVI e 0O~ " Basin Faithful In-road 1-.. " " Corralsin-roadI~ OOld In-roadFa~thful 44°,_ : /[ • ~ • _" • Crew O\. :;;;:Ol~0Faithful I · " Old Faithful 0 East 'L..- 1,000m Lot Grand "SnOW~itloc~tions IDAHO i \. Teton . Four Mile near Old Faithful , J l Meadow N P. • ..J o Garnet Canyon'OI ,,.J • y. 0 Togwotee Pass Rendezv~usMountain "LJ/"r\ EXPLANATION Sulfate ion concentration, in microequivalents per liter (Minimum value 2.7; maximum value 21.2) o 15 30 MILES 2.7 - 4.9 ! J o f I I o 15 30 KILOMETERS o 5.0 - 8.9 o 9.0 - 12.9 o 13.0-16.9 • 17.0-21.2 Figure 6. Sulfate ion concentrations in snow in Teton-Yellowstone area, 1998. SNOWPACK CHEMISTRY 15 I Table 4. Snow chemistry of selected volatile organic compounds at protocol-development sites fMTBE, methyl tert-butyl ether; ng/L, nanograms per liter; <, less thanJ Benzene MTBE m-9.8809,8807,8605771,1<251881,4001,57032,10032.100Toluene14.100IIIand<588.9(ng/L))37720,90064011.31,390(ng/L)36.5<25<1022.726512.112.913.71613,7(ng/L)14.91035.49.87,3806.520.47.20065.3<522.815.617.991.5765126548.1p-xyleneo-xylene Site and(or) sample name Date 374200298<10<1030,9<1010.146(ng/L)10.1<10<10<1010.1 (ng/L) Field Blank <10 12/17/97 <10 Field Blank 01/15/98 <10 Field Blank 01/22/98 <10 Laboratory Blank 01/28/98 <10 Trip Blank 01/15/98 <10 Trip Blank 01/22/98 15.8 Loch Vale 12/16/97 <10 Loch Vale (replicate) 12/16/97 <10 Supply Forks 12/17/97 <10 Supply Forks (in-road) 12/17/97 37.4 Supply Forks 01/15/98 <10 Supply Forks (in-road) 01/15/98 1,820 Supply Forks (exhaust) 01/15/98 824 Supply Forks 01/22/98 <10 Supply Forks (5 meters-exhaust) 01/22/98 <10 Supply Forks (in-road) 01/22/98 182 Minimum' <10 Maximum I 1,820.0 Median' <10 ---'-S-ummary statistics do not include blanks. ~ame basic trends seen with major ions <.!lIS.: ?)J partic• (at the high-use, in-road sites), toluene and xylene ul~rly benzene. Other hydrocarbons such as MTBJ;:, concentrations also were high. Although benzene and toluene, and the xylenes also showed similar positive MTBE concentrations did not follow this pattern in correlations to snowmobile traffic patterns (fi~. 8), each case, benzene concentrations were the highest at Low concentrations of these VOC's ge~nerally the high-use, in-road sites at West Yellowstone and Old Faithful, and lowest at off-road sites. o~c_urredin the off-road and backcountry sites~J11u~h higher concentrations were noted at sampling sites nearer to snowmobile emissions (table 5). A replicate sample at the Old Faithful in-road site had concentra• Snowmelt Runoff tions similar to the original sample for all constituents analyzed, The VOC concentrations in snow samples at ~reliminary sampling of five snowmelt runoff in-road sites also generally were lower at the low• sites near West Yellowstone, Old Faithful, Lewis Lake traffic sites near Lewis and Sylvan Lakes. Squared Divide, Sylvan Lake, and Tower Falls was undertaken Pearson correlation coefficients (r2) for relations as a first step in identifying whether local surface• between VOC's shown in table 6 indicate that strong ·water quality might be affected by elevated levels of correlations (r2 = 0.44 to 0.99) exist between all emissions detected in snow near snowmobile traffic compounds except MTBE. Additionally, good correla• routes in the park... In this initial measure of stream tions (r2 = 0.49 to 0.80) between all hydrocarbons chemistry after the onset of seasonal snowpack (except MTBE) and ammonium and sulfate support melting, only single grab samples were collected assertions that these inorganic and organic constituents during May 1998. Those grab samples, a field blank, are emitted from the same source. In general, when and a replicate were analyzed for major ions (table 7) concentrations of ammonium and sulfate were highest and VOC's (table 8). 16 Effects of Snowmobile Use on Snowpack Chemistry in Yellowstone National Park, 1998 In-road sampling sites Off-road and background sampling sites ... 180 cu Co .!!! -g _ 160 c cu cu cu u c Iii (jj' ~ 140 ._>-cu C ::I - CU C" ~ .c 120 curn• o "C ••• U c ~ 100 'E CU :: .!: § ~ 80 rn 'c rn .2 ~ ~ 60 - ... DivideLake YellowstoneYellowstone19.621.1<10Biscuit(8(8Sylvan6.299.34.2WestB<10643.BstoneTowerWestYellow·4.39.116.2<1021.9(in·road)6Faithful2.7West16723.6Lake4.B<10Basin4(in·km24.517.54.67.BOldFallsroad)easteast)YellowstoneFaithful(1,0006.4<104Old(1,OOOm)m) Lewis stone (in·road) in· road) Yellow· ~ ~ g 40 c-c 23 20.664.3 cu ••• CU co .-~ c 20 0- o West U ~;?B;.,~ri9 o ammonium • sulfate D benzene Figure 7. Ammonium, sulfate, and benzene levels in snow at selected sites in Yellowstone National Park, 1998. Major ion chemistry generally indicated low levels are moderate to high relative to hundreds of concentrations of ammonium and nitrate (0 to other streams, ponds, lakes, and springs sampled in the 5.0 )leq/L), with much greater concentrations of Rockies. chemical concentrations in Yellowstone sulfate (15.3 to 160 )leq/L) and calcium (121 to surface water (with the exception of sodium) are not l 432 )leq/L). magnesium (28.8 to 235 )leq/L), and unusually high (Musselman and others, 1996). Further sodium (102 to 359 )leq/L). The field blank showed monitoring and more intensive surface-water sampling normal low detections, but the replicate grab sample at are needed to verify and expand on this cursory, West Yellowstone yielded a moderate sulfate concen• preliminary description of snowmelt-runoff chemistry. , • 4 tration (18.3 )leq/L) compared to the high concentra• , .~ tion (160 )leq/L) in the original sample. Free acidity in - VOC species tend to reach equilibrium- with- the all stream waters sampled was very low (0.02 to atmosphere when liquid-phase species are allowed to 0.09 )leq/L hydrogen ion concentration) relative to interact with atmospheric gases (Rathbun, 1998). As other streams in headwaters basins in the Rocky snowmelt begins, hydrocarbons in solution mix Mountain region (Campbell and others, 1991; 1995). readily with the atmosphere and tend to voJatilize into The corresponding pH values ranged from 7.17 to 7.82 the gaseous phase as they enter the atmosphere. All for these slightly basic surface waters. This low acidity VOC's studied apparently exhibited this tendency is likely due to generally high cation concentrations except toluene; all other constituents registered below that give the streams potential for neutralizing acidic reporting limits while toluene persisted in snowmelt snowmelt. Sodium concentrations are unusually high runoff waters (table 8). Additional monitoring and (102 to 359 )leq/L), especially at West Yellowstone analyses are needed to verify the persistence of and Old Faithful. Although pH and other major ion toluene in snowmelt runoff. SNOWPACK CHEMISTRY 17 I In-road sampling sites Off-road and background sampling sites 400 co •..s::: 500600 0III -IIIs:::C)•..Coco•..1,000800200100300 •..s:::111 U:;;~0(J 700900 =0 IIIc:E 0 YellowstoneYellowstoneBiscuitYellowstone12.923.6237<10127167Sylvan(in-road)290<10(8<5West19762.2DivideTower52.227.210108West43.8~726594617837973544<5<10434449Basin(in-road)1089.325355West<1048.213.8Lake<599.36.560.8kmFallsOldeast)YellowstoneFaithful(1,OOOm)<10<10<5(1.000m)16.4438Old Lewis ..", (in-road)..".." -Y Weststone (8 km(in-road)east I in-road) Yellow- 51562930712.2 64.3 ~ ..,- -LYLY{1~ I:] benzene • MTBE o toluene o m- & p-xylene • a-xylene Figure 8. Hydrocarbons in snow on snowmobile-packed roads and off-road sites. 18 Effects of Snowmobile Use on Snow pack Chemistry in Yellowstone National Park, 1998 I Table 5. SnowBenzenem-chemistry438515472<10237726594617290Tolueneand<10197<10434<1016744983762930727,297325354435622,876255,029012712,218.312,910813264.399,371.323,643,86.552,262,255,0<513,848.260.860,8MTBE89.316.4p-xyleneo-xyleneDateof selected03/06/9803/03/9803/05/9803/103/04/9803/19/980/98volatile organic<1016767.8 compounds in the Greater Yellowstone area WestLewisMaximum(in-road)Median(in-road)(1.000(in-road)(in-road,BiscuitYellowstoneYellowstone.Lake1Summary1m)I Dividereplicate)Basinstatistics(1.0008 km eastdom)notincludeblanks. TowerJacksonWestSylvanOld FaithfulYellowstoneFallsLakeHole Airport Site name MinimumInglL. nanogramsI perliter; m. meters,km. kilometers;<. lessthanl Table 6. Squared Pearson correlation coefficients between selected organic and inorganic gasoline-engine• emission by-products IMTBE. methyltert-butylm-Toluene(ng/L)0.060.910,520.150,610.780,060.490.76and(ng/L)MTBE0,770,680,030.440.80ether;p-xyleneo-xylene0.780.66nglL, nanogramsper0.45IiterJ 0,30 0.74 0.99 m-Tolueneo-xyleneSulfateMTBEBenzeneand p-xylene (ng/L) Ammonium Benzene SNOWPACK CHEMISTRY 19 I Hydro-Magne- Table 7.no150160250358359305Snowmelt0.50.011610215.3115nium10042.3Ammo-Chloride4.30.041.015.31022352610.000.10.40.30.535916068.80.767.5220Nitrate43216.65.01.421971.634.92.12.2432data1.318.310425515.840.3Sulfate28.812114226.813077.30.020.070.80.060.090.044.25.22.354.30.05.086.10.0pH0.020.03987.257.827.177.247.047.425.38 Daterunoffsium05/09/9805/11/9805/04/9805/05/9805/19/98chemistry n/aof7.047.257.82selected inorganic compounds in Yellowstone National Park gen CalciumSodium TowerWestOldMeanStandardMaximum(replicate)(fieldFaithfulYellowstoneI(medianSummaryFallsblank)deviation1 forstatisticspH)1do notincludeblank. Lewis Lake Divide Sylvan Lake Minimum![units: SitepH, namestandard; inorganic compounds, in microequivalents per liter; n/a, not applicable] Benzene Table 8. Snowmeltm-<10<10Tolueneand<10<25252.0(ng/L)<10252<546.633.3MTBE80.2Daterunoffp-xyleneo-xylene05/04/9805/19/9805/09/9805/11/98chemistry of selected<10 volatile organic compounds in Yellowstone National Park (ng/L) WestMaximumMedianTowerOld(replicate)blank)LewisFaithfulYellowstone1 MedianFallsI Lake(fieldvaluesDivideexcludefield blank. Sylvan Lake Minimum[MTBE.Sitemethylnametert-butylether:ng/L. nanogramsperliter: <. lessthanJ 20 Effects of Snowmobile Use on Snowpack Chemistry in Yellowstone National Park, 1998 I PATTERNS OF CHEMISTRY RELATIVE TO SNOWMOBILE USE VOC's in Rocky Mountain snowpacks. Bruce and McMahon (1996) reported concentrations in snowfall Although clear patterns have emerged to estab- UOluenecollected(tablein the8).DenverLittle ismetropolitanknown aboutarealevelsto beoflow. - lish ammonium and sulfate as reliable indicators of Toluene concentrations in snowmelt runoff in J snowmobile- emissions in nearby snowpacks, particu•~ Yellowstone (less than 25 to 252 ng/L; table 8) further larly along the corridor from West Yellowstone to Old indicate the potential sensitivity to contamination of Faithful, nitrate concentrations are not much influ- \ snow and surface-water samples. Even at Loch Vale (table 4), the backcountry site in Colorado several extreme exposure of the direct exhaust sample at kilometers from the nearest roadway, toluene concen• Supply Forks, snowpack concentrations of nitrate trations were similar to those detected in the snow• enced by these local effects. With the exception of tJ \ w~re relatively unaffected by snowmobile traffic. packed roadway at Sylvan Lake (108 ng/L; table 5). Additionally, toluene concentrations in the snow• \' Siting off-road sampling sites 50 m from snow• , mobile routes seems adequate to eliminate contamina• packed roadway at Old Faithful also were very similar I tion from snowmobiles and allow observation of to the concentration in snow I km off the highway , regional effects. Comparisons between chemistries at (table 5). In some cases, there was a more clearly the West Yellowstone sites 50 and 1,000 m off-road observable pattern, such as with comparisons between show similar values for all major ions and also are in-road and off-road sites at West Yellowstone and at similar to background levels elsewhere in the Rocky the site 8 km east of West Yellowstone (West Yellow• stone, 8 km east, table 5). The Tower Falls site, several mobiles is less likely 50 m from highway corridors, I kilometers from snowmobile traffic, had a low concen• I-especiallyMountain region;when comparedtherefore, tocontaminationin-road chemistry.from snow•-.J tration (89.3 ng/L) similar to that detected in both the Furthermore, two sites 50 m ofT-road and a third site original (91.5 ng/L) and replicate (III ng/L) snow 1,000 m off-road around Old Faithful also had good samples at Loch Vale, Colorado (table 4). Oddly, the agreement between major-ion concentrations and also snowmelt runoff grab sample from the area near Tower were unaffected by snowmobile traffic, as shown by Falls contained the highest concentration of toluene the in-road snow chemistry. Comparisons of these off• (252 ng/L). Clearly, more investigation is needed to road and in-road sites in the Old Faithful area located determine whether these anomalously high values for within 2 to 3 km of the geyser also indicate negligible toluene (relative to benzene, MTBE, and xylenes) in effects on sampling results from the geothermal snowmelt runoff are due to the sampling methodology, activitY., other sources of contamination, analytical techniques, Hydrocarbon levels in the snowpacks near or ambient conditions. In spite of these uncertainties, snowmobile use were elevated relative to background the toluene snow chemistry positively correlates with snowpack chemistry in the study but were lower, in other hydrocarbon and major-ion concentrations. Drinking-water standards for benzene - nationwide representing a full spectrum of watershed (5,000 ng/L), toluene (1,000,000 ng/L), and xylenes settings ranging from subalpine to urban (Dennehy (10,000,000 ng/L) published by the U.S. Environ• 1general,and others,than1998).concentrationsDetectable atconcentrationshundreds of locationsof mental Protection Agency (1996) far exceed any levels detected in either snow or snowmelt runoff at Yellow• VOC's in Yellowstone ranged from 12.2 to 973 ng/L I (table 5). VOC concentrations detected in urban itone in this study. A drinking-water standard for '---" storm water in the United States have been found to MTBE has not yet been determined, but future regula• range from 200 to 10,000 ng/L, with more concen• tion is planned. Even the highest detections of benzene trated levels observed less frequently (Lopez and in snow (167 ng/L at in-road site 8 km east of West Bender, 1998; Lopez and Dionne, 1998). In a variety Yellowstone) or snowmelt (less than 10 ng/L at all of urbanized, forested, and agricultural settings in sites), or toluene in snow (726 ng/L at in-road site New Jersey (Reiser and O'Brien, 1998), median 8 km east of West Yellowstone) or snowmelt concentrations of seven streams detected for benzene (252 ng/L near Tower Falls) at Yellowstone are far less (60 ng/L), MTBE (420 ng/L), toluene (60 ng/L), and than the established standards for water consumed by o-xylene (10 ng/L) were markedly higher than concen• humans (less than 4 percent and less than I percent, trations in snowmelt runoff at Yellowstone except for respectively). PATTERNS OF CHEMISTRY RELATIVE TO SNOWMOBILE USE 21 I CONCLUSIONS tion of persistent organochlorine compounds in moun• tains of western Canada: Nature, v. 395. p. 585-588. Snowpack-chemical analyses for ammonium Bruce. Breton. 1995, Denver's urban ground-water and sulfate have proven to be repeatable indicators of quality-Nutrients, pesticides, and volatile organic -snowmobile use in Yellowstone National Park and in compounds: U.S. Geological Survey Fact Sheet Colorado, and the hydrocarbons benzene, tolu~ne, and FS-I06-95. 2 p. xylenes correlate well with patterns observed in 1998 Bruce, W.B., and McMahon, P.B., 1996, Shallow ground• for ammonium and sulfate in the park. Concentrations water quality beneath a major urban center, Denver, of ammonium and sulfate at the sites i~ snowpacked Colorado: Journal of Hydrology, v. 186. p. 129-151. roadways between West Yellowstone and Old Faithful Campbell, D.H., Turk, J.T, and Spahr, N .E., 1991, were greater than those observed at any of 50 to 60 Response of Ned Wilson Lake watershed, Colorado. to other snowpack-sampling sites in the Rocky Mountain changes in atmospheric deposition of sulfate: Water Resources Division, v. 27, p. 2047-2060. region and clearly were linked to snowmobile opera• Campbell. D.H., Clow, D.W., Ingersoll. G.P., Mast. M.A" tipn. Concentrations of ammoniu!1~, sulfate, and Spahr, N.E., and Turk, J.T. 1995, Processes controlling hydrocarbon compounds found in gasoline correlate the chemistry of two snowmelt-dominated streams in 'with snowmobile use and traffic levels; where traffic the Rocky Mountains: Water Resources Research. volumes per day were greater, so were chemical v. 31. p. 2811-2821. concentrations. Thus, these combined analyses of Collins, B. J.. and Sell. N. J., 1982. Lead contamination chemistry of Yellowstone snowpacks are good indica• associated with snowmobile trails: Environment tors of the effects of high or low snowmobile traffic Reservoir, v. 27, p. 159-163. levels in the park. These chemical data establish Delzer, G.c.. Zogorski, J.S., Lopes. T.J .. and Bosshart. R.L.. important baselines for future evaluations. Further, 1996, Occurrence of the gasoline oxygenate MTBE these results indicate that snowmobile use along the and BTEX compounds in urban storm water in the routes originating at the South and East Entrances, and United States, 1991-95: U.S. Geological Survey n9t including the immediate area (within I km) Water-Resources Investigations Report 96-4145, 6 p. $.urrounding Old Faithful, may not be substantially Dennehy, K.F., Litke, D.W"Tate, C.M., Qi, S.L., McMahon. affecting atmospheric deposition of ammonium, P.B., Bruce, B.W., Kimbrough, R.A., and Heiny, lS., sulfate, and hydrocarbons related to gasoline combus• 1998, Water quality in the South Plalle River Basin, tion. Colorado, Nebraska, and Wyoming: U.S. Geological Survey Circular 1167,38 p. Preliminary analyses of snowmelt-runoff Fenelon, J.M .. and Moore, R.c., 1996, Occurrence of vola• chemistry from five of the snow-sampling sites indi• tile organic compounds in groundwater in the White cate that elevated emission levels in snow along River Basin, Indiana, 1994-95: U.S. Geological highway corridors generally are dispersed into Survey Fact Sheet FS-138-96, 2 p. surrounding watersheds at concentrations below l~y:els Hare, c.T., and Springer, KJ., 1974, Exhaust emissions likely to threaten human or ecosystem health. Local-. from uncontrolled vehicles and related equipment ized, episodic acidification of aquatic ecosystems in . using internal combustion engines, Part 7. Snow• these high snowmobile-traffic areas may be possible, mobiles: Southwest Research Institute, San Antonio, but verification will require more detailed chemi~al Texas, report no. SWRI-AR-946, 90 p. analyses of snowmelt runoff. Ingersoll, G.P., 1995, Maximum-accumulation snowpack chemistry at selected sites in northwestern Colorado during spring 1994: U.S. Geological Survey Open-File REFERENCES Report 95-139. 14 p. Ingersoll. G.P" 1996, Snowpack chemistry at selected sites Ayers, M.A., Baehr, A.L.. Baker. RJ., Hopple. lA .. in northwestern Colorado during spring 1995: U.S. Kauffman, and LJ., and Stackelberg, P.E., 1997, Geological Survey Open-File Report 96-411, 16 p. Design of a sampling network to determine the occur• Ingersoll, G.P" Turk, J.T. McClure, c., Lawlor, S., Clow, rence and movement of methyl tert-butyl ether and D.W., Mast, M.A., 1997, Snowpack chemistry as other organic compounds through the urban hydrologic an indicator of pollutant emission levels from cycle: American Chemical Society. v. 37. no. I. motorized winter vehicles in Yellowstone National p. 400-40 I. Park: Proceedings, 65th Annual Meeting. Western Blais, lM., Donald, D.B., Kimpe, L.E., Muir, D.C.G" Snow Conference, May 4-8, Banff, Alberta, Canada, Rosenberg, B., and Schindler, D.W., 1998. accumula- p.I03-113. 22 Effects of Snowmobile Use on Snowpack Chemistry in Yellowstone National Park, 1998 I Littlejohn, M., 1996, Yellowstone National Park visitors Reiser, R.G., and O'Brien, A.K., 1998, Occurrence and study, report 75: National Park Service Cooperative seasonal variability of volatile organic compounds Park Studies Unit, Moscow, University of Idaho, in seven New Jersey streams: U.S. Geological January 1996. Survey Water-Resources Investigations Report Lopez, T.J., and Bender, D.A., 1998, Nonpoint sources of 98-4074. IIp. volatile organic compounds in urban areas-Relative Terracciano, S.A., and O'Brien, A.K., 1997, Occurrence of importance of land surfaces and air: Environmental volatile organic compounds in streams on Long Island. Pollution, v. 100, p. 221-230 New York, and New Jersey: U.S. Geological Survey Lopez, TJ., and Dionne, S.G., 1998, A review of semivola• Fact Sheet FS-Q63-97, 4 p. tile and volatile organic compounds in highway runoff and urban storm water: U.S. Geological Survey Op~n• Turk. JT. and CampbelL D.H .. 1987. Estimates of acidifi• File Report 98-409, 67 p. cation of lakes in the Mt. Zirkel Wilderness Area, Martner, B.E., 1986, Wyoming climate atlas: Lincoln, Colorado: Water Resources Division, 23, 1757-1761. Nebr., University of Nebraska Press, 432 p. Turk, J T, and Campbell, D.H., 1997, Are aquatic resources Morris. J .A.. Bishop, G.A., and Stedman, D.H., 1999, Real• of the Mount Zirkel Wilderness Area in Colorado time remote sensing of snowmobile emissions at affected by acid deposition and what will emissions Yellowstone National Park-An oxygenated fuel reduction at the local power plants do?: U.S. Geolog• study, 1999: Prepared for Western Regional Biomass ical Survey, Fact Sheet FS-Q43-97, 4 p. Energy Program, PO. box 95085, IIII 0 Street, Turk, J.T., Campbell, D.H., Ingersoll, G.P., and Clow, D.W.. Suite 223, Lincoln, NB 68509,40 p. Musselman, R.C.. Hudnell, L., Williams, M.W., and 1992. Initial findings of synoptic snowpack sampling in the Colorado Rocky Mountains: U.S. Geological Sommerfeld, R.A .. 1996, Water chemistry of Rocky Survey Open-File Report 92-645. 6 p. Mountain Front Range ecosystems: USDA-Forest Service, Research Paper RM-RP-325, 13 p. U.S. Environmental Protection Agency, 1991, Non-road National Park Service, 1990, Winter Use Plan Environ• engine and vehicle emission study report: U.S. Envi• mental Assessment. Yellowstone and Grand ronmental Protection Agency, Ann Arbor, Mich .. Teton National Parks and John D. Rockefeller Jr., Office of Mobile Sources, report no. EPA/460/3• Memorial Parkway, Wyoming, Idaho, and Montana: 91/02, 566 p. U.S. Government Printing Office. U.S. Environmental Protection Agency, 1996, Drinking National Park Service, Air Quality Division, 1996, Results water regulations and health advisories: Office of of preliminary investigation of carbon monoxide and Water, U.S. Environmental Protection Agency, particulate matter levels at Yellowstone National Park Washington, D.C., October, 1996. 15 p. West Entrance Station during 1995: Denver, Colo., National Park Service review draft, March 21. 1996. Wanek, WJ., 197 I, Observations on snowmobile impact: Natural Resources Conservation Service, 1998, Wyoming The Minnesota Volunteer, Minnesota Department of Basin outlook report, April I, 1998: USDA-NRCS, Natural Resources. v. 34, p. 1-9. Casper, Wyoming, p. 9-11. White, lJ., and Carroll, J.N., 1998, Emissions from snow• Neumann, PW., and Merriam. H.G., 1972, Ecological mobile engines using bio-based fuels and lubricants: effects of snowmobiles: Canadian Field-Naturalist, Final report 08-7383, Southwest Research Institute, v. 86, p. 207-12. San Antonio, Texas, 184 p. Paulson, R.W., Chase, E.B., Roberts, R.S, and Moody, Wilkinson, Todd, 1995, Snowed under-The roar of snow• D.W., 1991, National water summary 1988-89• mobiles: National Parks, v. 69, Washington, D.C., p. Hydrologic events and floods and droughts: U.S. 32-37 Geological Survey Water-Supply Paper 2375. Rathbun, R.E., 1998, Transport, behavior, and fate of vola• Woodbury, Richard, 1997, Arctic cats and buffalo- Yellow• tile organic compounds in streams: U.S. Geological stone may not be big enough for both its growing herds Survey Professional Paper 1589, 151 p. of snowmobilers and its bison: Time, v. 149, p. 62-3. REFERENCES 23 -I< u.s. GOVERNMENT PRINTING OFFICE: 1999 - 773-084 / 20077 Region No. B I --,--. @ Printed on recycled paper I