Purpose and Scope

In autumn 2019, the U.S. Geological Survey (USGS) characterized the elemental composition of sediment retained by Enloe Dam. This study, completed in cooperation with the Priest Rapids Coordinating Committee, Habitat Subcommittee of the Grant County Public Utility District, focused on recently and historically deposited sediment within the Enloe Reservoir, emphasizing sediment sampling from lower sediment strata. Elemental concentrations, grain size, carbon content, radioisotope activities, and stable isotope ratios were measured in sampled sediment. The elemental analysis included a broad spectrum of rock-forming elements, elements associated with base and precious metal mining, and trace elements of environmental concern. Additionally, the abundance ratio of stable isotopes of nitrogen and carbon, were measured to assist future assessments of the historical use of the Similkameen River Basin by anadromous fish. These data and associated analyses provide resource managers information to evaluate sediment-management alternatives, feasibility, and potential costs that may be incurred if sediment retained by the dam contains high concentrations of trace elements. The purpose of this report is to present the methodology used and the results of this study. Data collected as part of this study are presented in Cox and others (2021).

Study Area

The 197-km long Similkameen River originates in Canada along the eastern flanks of the Cascade Range, with the final 43.5 km of the river in Washington State. The catchment above Enloe Dam drains about 9,190 square kilometers (3,550 square miles), of which 18 percent is within Washington State, and 82 percent is within British Columbia, Canada (Johnson, 2002). This study was conducted on sediment from the 2.6 km impoundment behind Enloe Dam. Enloe Dam (fig. 1) is a 16.2-meter (m) -tall (53-foot-tall), run-of-the-river style concrete dam. The dam is situated in a narrow, steep-walled canyon of the Similkameen River above Similkameen Falls 8.8 km upstream from the confluence of the Similkameen River with the Okanogan River (fig. 1). Sediment accumulation has filled much of the available storage capacity of Enloe Reservoir, resulting in water depths that are generally less than 6 m (20 feet). A previous USGS study estimated that as much as 1.37 million m³ (1.79 million yd³) of sediment is trapped behind Enloe Dam (Nelson, 1972). Sand-sized sediment appears to dominate reservoir substrate as Johnson and Plotnikoff (2000) reported from 0.0 to 3.5 percent fines in their eight samples. Nelson (1972) reported the predominance of sand and coarser-sized sediment, including gravel and cobbles.

Most of the Similkameen River Basin has a semi-arid climate, although the western part of the catchment drains the eastern slope of Cascade Range and is less arid. Snowmelt in the mountains provides the primary source of water to the Similkameen River, resulting in strong seasonal fluctuations in river discharge. Approximately 60 percent of the annual river discharge results during the spring freshet months of May and June when flows typically average about 227 cubic meters per second (m³/s) (8,000 cubic feet per second [ft³/s]). Discharge is appreciably lower during the remaining months, especially from August through October, when flows average below 28.3 m³/s (1,000 ft³/s) (U.S. Geological Survey, 2020). Above the dam, river flow is unregulated and on four occasions since 1928, flows greater than 850 m³/s (30,000 ft³/s) were recorded at the USGS gage, Similkameen River near Nighthawk (12442500; fig. 1). Enloe Dam lacks fish ladders and completely prevents the passage of fish.

The bedrock of the Similkameen River Basin varies, but most of the basin is underlain by marine sedimentary and coarse clastic sedimentary rock. In some areas, intrusion of granitic plutonic rock has generated low-grade metamorphism and hydrothermal alteration, leading to development of vein and porphyry type ore deposits (McMillan and Panteleyev, 1992; Ray and others, 1992; Stanley and others, 1996). Gold has been reported to occur as individual metallic fragments, often referred to as bleb or flecks, smaller than 25-micrometer (µm) that are associated with arsenopyrite in the garnet-pyroxene skarn of the Nickel Plate deposit (Sibbick and Fletcher, 1993). The entire basin was repeatedly glaciated during the Pleistocene, which sculpted bedrock and drainage features into a predominantly north-south orientation. Patches of unconsolidated till and silts, from Pleistocene glaciation, are common as a veneer throughout the basin (Cui and others, 2019). The north-south movement of continental glacier ice caused the transport of sediment and associated trace elements, such that analysis of the trace-element distribution in soils provides clues for the location of economically valuable elements (Sibbick and Fletcher, 1993; Shewchuk and others, 2020). Deposition of recessional glacial-fluvial sediment in the lower pre-glacial channel of the Similkameen River was extensive, resulting in the diversion of river flow near the location of Palmer Lake (fig. 1) eastward into the adjacent drainage of the Okanogan River (Willis, 1887). During seasonal spring freshet, flow from the Similkameen River enters Palmer Lake, which typically drains into the Similkameen River during the remaining seasons of the year.

Surface Sample Collection

Surface samples were collected at the SWI with a petite Ponar sampler at 10 transect locations, spaced approximately 50–400 m apart along the length of the reservoir (fig. 2). Surface samples were collected on October 17, 18, and November 11, 2019. At each transect, the channel cross-section was divided into thirds—left, right, and center—with samples being collected from near the middle of each third section. Sample identification was based on the location of the sample transects and position within the transect with respect to the channel center; for example, “BG8La blk” indicates the sample is a surficial grab sample (BG) from the left part of transect 8 (8L), a primary (a) sample as opposed to a reserve (b) sample, and unsieved or bulk (blk) sample. Samples were collected by carefully lowering the sampler to the riverbed where it triggered on contact, scooping up roughly 1–2 kg of material to a depth of about 2–5 centimeters (cm). Upon retrieval, the sediment sample was placed in a 23-cm by 30-cm polymer tray, and the sample was inspected to remove fragments of organic debris larger than about 1 cm. The sample was then transferred to a Ziploc bag using a polymer scoop and stored on ice. All samples were transported to the water-quality laboratory at the USGS Washington Water Science Center (WAWSC) in Tacoma, Washington, where the samples were stored frozen. Subsamples were drawn from original sediment samples and packaged for shipping to other laboratories for analysis.

Sediment Core and Slurry Sample Collection

Sediment cores were collected at six locations during November 14–17 and December 10–14, 2019, using a Vibecore-D coring device provided by Specialty Devices Incorporated. Core locations focused on the deeper deposits closer to the dam. The Vibecore-D was suspended from a 3.3-m tripod frame mounted on the deck of a 2.4 by 4.6 m floating platform equipped with an electric hoist. Sediment was collected in thin-wall (1.6 millimeters [mm]) aluminum (Al) core tubes with 2.5-, 5-, and 10-cm diameters. The platform was anchored in position from all four corners during coring operations. The target depth for the collection of cored sediment was the interface between accumulating sediment and the pre-reservoir surface of cobble or bedrock. At the time of coring, the maximum thickness of accumulated sediment in the reservoir was expected to be from 9 to 12 m, as reported by Nelson (1972) and Public Utility District No. 1 of Okanogan County (2012). A telescoping-coring technique was used to collect sediment at depths greater than 5 m below the SWI where a combination of 2.5-, 5-, and 10-cm diameter Al tubes were used. With this technique, a 5-cm diameter core tube, generally 6.1-m long, was driven from the SWI to a maximum depth and then extracted from the riverbed. After recovery of the 5-cm tube, a 10-cm tube was inserted to the same depth; sediment within the 10-cm tube was removed using a sediment-slurry pump. A 5-cm tube, generally 6-m long, was then inserted inside the 10-cm tube and, using additional tubes with sleeve connectors, driven to depths below the extent of the 10-cm tube. This telescoping process was repeated at some sites by cleaning out the 5-cm tube, inserting a 2.5-cm tube, and driving the tubes to depths greater than 9 m below the SWI. Sediment core tubes were generally driven until refusal, the point where no further downward vertical advancement of the Vibecore system was possible. A stainless-steel core-catcher retained sediment in the core tube during sample collection. Upon recovery, the core tube was drained of overlying water and capped. Summary data of core sections using multiple telescoping sections are listed in table 1.

Sediment in the first two cores, 6_2 and 7_2, was highly compacted during the coring process and the sediment could not be extruded from the core tubes in the field using the continuous screw-driven sediment extruder. Sediment was removed from the core tubes in 7.5–15 cm sections using polyvinyl chloride pipe split lengthwise for these cores. Sediment samples removed from the core tube were placed in Ziploc bags and stored on ice. For coring sites, 8_5, 8_6, 8_7, and 8_8, sediment core tubes were cut to remove the unfilled core tube, packed to remove void space at the upper end of the tube, and then capped at both ends so that sediment remained in place. Cores were returned to the USGS WAWSC laboratory in Tacoma, Washington, where the Al core tube was frozen with dry ice, and then split lengthwise. This method allowed the sediment cores to be split, showing grain size and sedimentary differences. Samples for laboratory analysis were then collected from the central part of the exposed sediment.

The core catcher provided a common reference point on each sample tube for sample identification and depth of penetration below the SWI. A unique core-sample identification code was developed based on the transect nearest the core location, the core segment, the distance of the sample above the core catcher, and the size fraction of the sample. For example, sample 8_6b.12-16<63 indicates the sample is from core site 8_6, from the second (b) telescoping segment, and was taken from material 12–16 cm above the core catcher and is the less than 63-µm fraction. An estimated location of the sediment-sample depth can be made from the depth of penetration of core segment 8.6b, which would be the depth location of the core catcher (table 1). While labeling and recording cores from sites 6_2 and 7_2, the distance above the core- catcher units was measured in inches and converted to International System of Units (metric) units; all other units were measured in centimeters.

At coring sites, 8_6, 8_7, and 8_8 (fig. 2), where penetration depths reached refusal depths of 8–9 m, samples of water and sediment, referred to as slurry samples, were pumped from the coring tube into 20-liter plastic buckets to ensure a minimum of 10 grams (g) of less than 63-µm sediment fines was available for analysis of ¹³⁷Cs and ²¹⁰Pb_ex radioisotopes.

Analytical Methods

Sediment grain size fractions were characterized using the Wentworth scale of less than 63 µm, very-fine sand (greater than 63 to less than 125 µm), fine sand (greater than 125 to less than 250 µm), medium sand (greater than 250 to less than 500 µm), coarse sand (greater than 500 µm; Horowitz, 1991). The grain size was determined gravimetrically by sieving and reported as percent of total sample by mass. Surficial samples were sieved into three size fractions: gravel (greater than 2 mm), sand (greater than 63 µm and less than 2 mm), and fines (less than 63 µm). Core sediment samples were sieved into five size fractions: less than 63 µm, greater than 63 to less than 125 µm, greater than 125 to less than 250 µm, greater than 250 to less than 500 µm, and greater than 500 µm.

The carbon content of sediment was determined for both core and surficial sediment. TC was determined using a coulometric titration following combustion at 1,000 °C in a stream of oxygen; the carbon dioxide (CO₂) liberated following combustion was titrated in a coulometer cell. Likewise, for inorganic carbon (IC), the untreated crushed sample was reacted with a perchloric acid solution to liberate CO₂ from inorganic carbonates and then titrated in a coulometer cell to measure percent IC. The difference between TC and IC determined total organic carbon (TOC) concentrations.

The activities of radioisotopes ²¹⁰Pb_, and ¹³⁷Cs were measured on five sediment samples, obtained from the deepest sediment cores sampled to confirm that deep core samples contained sediment deposited before the mid-1950s. Caesium-137 would not be expected in sediment deposited before thermonuclear weapons testing began in the mid-1950s (Olsson, 1986; Ritchie and McHenry, 1990). Excess lead-210 (²¹⁰Pb_ex)_, as defined later in this paragraph, would not be expected in sediment deposited more than approximately 100 years ago. To increase the likelihood of detecting ¹³⁷Cs and ²¹⁰Pb in the Enloe Reservoir sediment, bulk samples were sieved and only the less than 63-µm fraction was analyzed for ¹³⁷Cs and ²¹⁰Pb (He and Walling, 1996; Singleton and others, 2017). As much as 10 milliliters of fine sediment were sealed in gamma vials for 3 weeks to develop radioactive equilibrium between ²²⁶Rn and ²¹⁴Pb. Sediment vials were counted on an ultra-low background germanium well detector at the USGS Pacific Coastal and Marine Science Center Radiochemistry Lab in Santa Cruz, California, for 24 hours. Absolute activities of ¹³⁷Cs and ²¹⁰Pb were calculated relative to measured activities in certified reference materials from the International Atomic Energy Agency with the same geometries and counting times. Total lead-210 (²¹⁰Pb_total) was measured by its gamma peak at 46.5 keV, ¹³⁷Cs was measured by its peak at 661 keV, and ²²⁶Ra was measured by its indirect peaks at 351.87 keV (²¹⁴Pb) and 609.31 keV (²¹⁴Bi); data are reported as disintegrations per minute per gram of material (dpm/g). Unsupported, or excess ²¹⁰Pb (²¹⁰Pb_ex) was computed as the difference between ²¹⁰Pb_total and ²⁶⁶Ra (assumed to equal supported ²¹⁰Pb). Reported uncertainties include detector efficiency, background corrections, and one standard deviation of the counting error.

Sediment from surficial and core samples of various size fractions were analyzed for the concentration of major and trace elements (listed in table 2) by AGAT Laboratories (Calgary, Alberta, Canada). Samples were prepared for elemental analysis at the USGS WAWSC by drying overnight at 60° C and prepared as both sieved and unsieved; that is, bulk samples. Sieve sizes corresponded to the Wentworth size fractions defined above. Laboratory analysis of the elemental composition of sediment samples was performed by inductively coupled plasma-optical emission spectroscopy (ICP-OES) and inductively coupled plasma-mass spectroscopy (ICP-MS), following microwave-assisted strong acid (mixture of nitric, hydrofluoric, and perchloric acids) digestion of 1.0 g of sediment (Taggart, 2002). Data were deemed acceptable if recovery of each element was plus or minus 15 percent at five times the lower reporting limit and the calculated relative standard deviation of duplicate samples varied by no more than 15 percent (see section, “Quality Assurance and Quality Control”). Data for rhenium, although reported in Cox and others (2021), were variable with respect to replicate analysis and lacked reference sample concentrations or other appropriate quality-assurance (QA) data to be included in this report. All total-element contents are reported as either parts per million (ppm, equivalent to mg/kg) or percentage of the dry weight of sediment.

Sediment samples for analysis of the stable isotope abundance ratios of nitrogen and carbon (¹⁵N/¹⁴N and ¹³C/¹²C) were prepared from the fine-grain fraction of sediment obtained from near the lowest sections of the cored sediment. The samples were prepared from either individual core sections or sediment-water slurry mixtures pumped from the maximum cased depth used for core retrieval. Two to five replicate sample pellets were prepared and analyzed for each sample submitted for analysis. The analysis of stable isotope ratios was conducted at the University of California-Davis Stable Isotope Facility. The resulting ratios are expressed relative to international standards VPDB (Vienna Pee Dee Belemnite) for carbon and air for nitrogen. Results are reported in standard delta notation (δ) in units of parts per mil (‰).

Quality Assurance and Quality Control

Environmental samples analyzed for elements, carbon, and stable isotopes were corroborated against a wide-ranging set of QA data, including standard reference material, quartz blank samples, and replicate samples. Samples were collected following the Quality-Assurance Plan for Water-Quality Activities in the USGS WAWSC (Conn and others, 2017). A review of QA data indicated that data quality, for all elements other than Rhenium (Re), is acceptable for this study (table 2 and in Cox and others, 2021). Radioisotope data were used qualitatively to confirm that sediment was deposited pre-1950s; no QA samples were submitted for analysis.

To evaluate precision, the relative percent difference (RPD) between replicate samples was calculated as:

Carbon QA samples included quartz sand blanks and duplicate or triplicate samples. The two quartz blank sample results indicated no carbon was detected. The average RPD for TC was 9.4 percent, for TOC it was 6.6 percent, and for IC it was 3.8 percent. In replicate samples with zero or near zero measured IC, RPD cannot be calculated correctly and were not included in averages.

Briefly, samples of standard reference material were included with environmental samples submitted to the laboratory for analysis of trace elements; element recoveries that exceeded 15 percent of the accepted values are noted. Element concentrations measured in quartz blank samples were often below the minimum laboratory reporting limits. Concentrations of 14 elements reported in the quartz blank samples were greater than 5 times the reporting limit. The average concentrations of As and Pb measured in the three quartz blank samples were1.2 mg/kg for As and 0.7 mg/kg for Pb. Only the concentrations for Ge, Hf, Li, and Zr in the quartz blank material exceeded 10 percent of the minimum concentration reported in the Enloe Reservoir sediment samples, indicating negligible analytical or processing contamination in the resulting data.

The RPD of replicate samples was typically less than 6 percent for elements; higher RPD resulted in duplicates of bulk samples and was more common among element concentrations approaching minimum laboratory reporting levels, such as for Au, Pb, and Hg. The average recovery of elements in reference materials samples was within the target range of recovery of 85–115 percent of expected values for all elements except for Ag, Au, beryllium (Be), and selenium (Se), which were higher, and Hf, Y, and Zr, which were lower (see table 2). Recovery for Zr, Hf, Cr, Sn, Mo, and Y may be low as dissolution of recalcitrant minerals during sample preparations is often poor (Hu and Qi, 2014). Average RPD, by element, among replicate reference-sample pairs are shown in table 2.

For stable isotopes, samples of reference material (National Institute of Standard and Technology-1547, peach leaves) were prepared and analyzed for QA purposes. Data resulting from the replicate analysis of NIST-1547 indicated analytical variability and precision were acceptable; the average RPD was 2.23 percent for δ¹⁵N and 0.1 percent for δ¹³C. The long-term standard deviation for analysis of stable isotope ratios of ¹⁵N/¹⁴N and ¹³C/¹²C reported by the University of California-Davis Stable Isotope Facility is 0.3‰ for ¹⁵N and 0.2‰ ¹³C, respectively (University of California at Davis, 2020a, 2020b65). No preferred isotope value has been established for peach leaves as reference material, but RPD among the five replicate peach leaves submitted were less than 5 percent and within ranges found in other studies (Cox and others, 2016; Gendaszek and others, 2016)

Statistical Analysis

Summary statistics were calculated with Microsoft Excel data-analysis package software (version MS365 enterprise) and included minimum, maximum, median, mean concentrations, and standard deviation statistics. Numerical comparisons of mean chemical concentrations were conducted using a t-test where unequal variance was assumed to correct for unequal sample size; acceptance of significant differences was p less than 0.05.

To assess if elements concentrations in Enloe Reservoir sediment were enriched or depleted with respect to average continental crustal abundance (ACCA) values, the ratio of median element concentrations of reservoir sediment (in the size fraction of less than 63 µm) was compared to ACCA values to generate an enrichment/depletion (E/D) ratio. If the E/D ratio was greater than 1.5, the element concentrations were considered enriched; if less than 0.75, element concentrations were considered depleted. The E/D ratio is used as an indicator of potential anthropogenic (human) influence.

A non-parametric Spearman’s rank-order correlation was performed in R (R Core Team, 2021) using the Hmisc package (Harrell, 2021) and visualized using the corrplot package (Wei and Simko, 2021). These packages were used to investigate the correlation among elements. Correlation significance was accepted at p less than 0.01.

Additional figures were created using the R-base package or ggplot2 (Wickham, 2016).

Grain-Size Characterization

Locations of 27 surficial sediment-sampling sites and 6 sediment core collection sites are shown in figure 2. Data associated with these samples were tabulated in Cox and others (2021). Sediment accumulated in the Enloe Reservoir consist primarily of sand and silt-sized detrital grains of the major rock-forming minerals, including quartz, feldspar, mafic minerals pyroxenes, and amphiboles, along with fragments of lithic rock material. Surficial sediment was typically medium to coarse grain sand with limited silt-sized material. Very coarse sand sediment was deposited in areas where the overlying surface water was shallow enough to allow wind-induced waves to resuspend and remove finer grain material from the sediment bed surface. Whitecap waves and turbid water were observed during high winds in the shallow areas within 10–15 m of the left bank between cross-sections 6 and 8 (fig. 2) during sampling on October 25, 2019.

Sediment cores were collected that extended 4.3–10.9 m below the SWI; cores where multiple telescoping cores segments were utilized are summarized in table 1. Core sections through sandy sediment were difficult to advance owing to friction of sand within the core barrel, particularly with coarse sandy material. All cores were advanced to the point of refusal, where further downward movement was impossible. Core catchers inserted at the bottom of each coring section before deployment were typically intact following core recovery with little or no sediment loss from the base of the core tube. Overall core recovery of sediment within the core tube was typically only about 45 percent of the total length of the core tube pushed into the sediment. The limited core recovery was attributed to sediment not entering the core tube as the core barrel was advanced because of greater friction within the core tube than was necessary to push the saturated sediment aside during vibration of the core assembly. Only core 8_7 (fig. 2) showed evidence of encountering the pre-reservoir streambed presumed to consist of boulders or hard rock capable of damaging the sharpened beveled end of the core tube. Additionally, there was a notable change in vibration noise of the Vibecore at the point in coring 8_7, where the core could not be advanced any further. Data from pre-reservoir stream elevation profiles indicate that the depth of cores 8_5, 8_6, and 8_8 approached the pre-reservoir streambed (Nelson, 1972). The extent of sediment accumulated beneath cores 6_2 and 7_2 is not well characterized but expected to be less than 2 m.

The cored sediment appeared to consist of generally well-sorted and rounded medium sands without appreciable horizontal bedding of sedimentary structure (fig. 3). Some sediment particle-size variations were observed, including occasional clasts of pebbles and gravel-size fragments, particularly in the upper 1.5 m of sediment below the SWI as shown in core segment 8_7a, 98–124 cm (fig. 3A). Distinct horizontal bedding features were generally lacking; however, occasional layers of charcoal particles were present in some sediment cores, such as near the distal end of core 8_6a, 3–20 cm (fig. 3B). The charcoal, presumably detritus from forest fires, mainly consisted of particles 1–2 mm in size in 1-15 mm thick bands. The charcoal banding often was distorted near the core wall indicating friction at the core wall; however, distortion was not apparent in the interior of the sediment core, indicating that the interior of the core was relatively undisturbed. From visual inspection of cores, the grain size of sediment near the bottom of cores was often finer, as there was less coarse sand that was common near the top of the accumulated sediment. Silt and smaller size particles were typically scant in much of the sediment cores, although an approximately 10–cm layer of very fine sand and silt was observed in several cores at a depth of about 3–4 m (examples from core segments 8_7a and 8_5c are shown in fig. 3C and 3D). In these sections, the transition from predominantly medium sand to very fine sand and silty sediment texture was abrupt, indicating a rapid change in sediment input or source area. The Vibracore could not be advanced beyond the silty material at about 4.2 m at core 6_2 nor below a depth of 2.5 m below the SWI of core 7_2 (fig. 2). Upon recovery of core 6_2, the sediment in the core catcher consisted of compact silt and very fine sand (sample 6_2a, 0–3 cm) transitioning to sandier material in the zone 5–15 cm above (samples 6_2, 3–6 cm; and 6_2, 9–12 cm) and returning to sediment more typical of subsurface reservoir sediment. Similar resistance to coring was encountered at about 3 m below the SWI at core sites 8_5 and 8_7, although coring was able to penetrate the fine sediment and collect an additional 1.8 and 4.6–6.1 m of sediment at those sites.

The distribution of sediment grain size varied among sediment samples. The distribution of sediment grain size between silt (less than 63 µm), sand (63–2,000 µm), and larger (greater than 2,000 µm) fractions of surficial and cored sediment samples are shown in figure 4. The general sediment size distribution in both surficial and subsurface sediment samples was similar, predominantly fine and medium sand (fig. 4). The minimal variation in sediment size fractions of the surficial sediment indicates a uniform fluvial environment within the reservoir at the time of deposition. Variation in percentages of the sediment-size fractions was larger for the subsurface sediment samples, as indicated by the distributions shown in figure 4. A more detailed analysis of multiple sand size fractions for 17 samples of subsurface sediment samples (table 3) shows that the 125–500-µm fractions (that is fine and medium sand) constituted, on average, 65.7 percent of the sample mass.

Sediment Carbon Content

The concentration of sedimentary carbon was low in Enloe Reservoir sediment (fig. 5). TC in surficial sediment ranged from 0.07 to 0.78 percent, with a median 0.22 percent. Nearly all the carbon present in the surface sediment was TOC. IC was detected in only 10 of 27 surface-sediment samples, with a median concentration of 0.002 percent. However, higher carbon concentrations were present in the buried sediment with TC concentrations ranging from 0.36 percent to 5.78 percent, with a median of 2.07 percent. The share of TOC carbon in buried sediment ranged from 71.3 to 100 percent; IC had a median concentration of 0.06 percent. TOC concentrations are in the same range as found by Johnson and Plotnikoff (2000).

Radioisotopes Confirm Sampling of Pre-1955 Deposited Sediment

The sampling of sediment deposited in Enloe Reservoir prior to 1955 in the lower strata of the cored sediment was confirmed by undetectable ¹³⁷Cs activities (table 4). Although collected from similar stratigraphic strata, the results of ¹³⁷Cs and ²¹⁰Pb_ex from samples 8_6c and 8_6slurry were within expected measurement limits for duplicate analysis; one sample was collected by slurry method and the other was taken from the cored sediment. The 8_7slurry sample had a low but statistically significant ²¹⁰Pb_ex activity, possibly due to a small amount of contamination with sediment from above. Slurry samples were pumped from the open outer core tube after the deepest core segment had been retrieved. The slurry may have entrained a mixture of sediment from the level above the open casing. The ²¹⁰Pb_ex activity of 1.6 dpm/g in the 8_7slurry (fig. 2) was the same as the ²¹⁰Pb_total activity measured in an older sediment core from Palmer Lake (fig. 1) for sediment deposited before 1955 (Ford, 2010). The ²¹⁰Pb_ex and ¹³⁷Cs activity data in Enloe Reservoir sediment provide consistent indications that sediment samples obtained closest to the pre-reservoir streambed were deposited prior to 1955.

Element Concentrations in Enloe Sediment

Element concentrations were determined for a total of 112 samples of Enloe Reservoir sediment. The samples included 36 surficial sediment samples from 27 sample locations and 76 samples of cored sediment from the 6 core locations (fig. 2). Cored sediment samples were selected from 35 sediment horizons from the six sediment cores collected. Elemental analysis was performed on the less than 63-µm fraction of all 27 surficial sediment samples, with a subset of 9 samples also analyzed as bulk sediment. The preparation of cored sediment samples was more diverse. Analysis of cored sediment samples included (1) 31 bulk sediment samples; (2) 17 samples of the less than 63-µm size fraction; (3) samples from two sediment strata (8_5c, 0–69 cm; and 8_8 c, 30–39 cm), sieved into five grain size fractions; and (4) six samples of slurry sediment. These subsurface sediment samples included 12 samples selected from horizons where visual inspection of the cored sediment indicated a larger fraction of finer-grained sediment. These samples were selected to improve the potential of detecting the presence of mercury and other elements that have an affinity to sorb to fine-grain sediment particles.

Element concentrations in the deeper sediment (considered historical or previously deposited) were often higher when compared to surficial sediment considered modern (deposited in 2019). However, the concentrations of several elements including Hg were higher in modern sediment. A statistical comparison of element concentrations was made between the most recently deposited surficial sediment and the deepest sediment obtained in the slurry samples. The comparison was made using the less than 63-µm fraction of sediment, which is more homogeneous in composition than bulk sediment samples (table 5). The concentration of elements in the slurry samples from the base of cored sections shows significantly larger (p less than 0.05) concentrations for 21 of 41 trace elements analyzed, including Ag, As, Be, Bi, Co, Cs, Cu, Ga, Li, Mn, Mo, Ni, P, Pb, Sb, Sc, Ta, Te, Th, V, and Zn. Concentrations of six elements (Ge, Hf, Hg, Re, Sn, and Sr) were significantly larger (p less than 0.05) in surficial sediment than in cored sediment. The remaining 15 elements (Au, Ba, Cd, Ce, Cr, In, La, Nb, Rb, Se, Tl, U, W, Y, Zr) had no significant difference in concentrations between these two sediment types.

Concentration data for major and trace elements analyzed in surficial and buried sediment from core samples from Enloe Reservoir are summarized in tables 6 (surficial sediment) and 7 (buried cored sediment). The summarized data include minimum, maximum, median, mean element concentrations, along with standard deviation for element concentrations of the less than the 63-µm sediment fraction; the mean concentration in bulk sediment samples, and three reference values for comparison. Reference values include average elemental concentrations of all the elements analyzed in Enloe Reservoir sediment; these data were from reference samples of upper continental crustal material compiled by Rudnick and Gao (2014) and Haynes and others (2016) and are referred to as average continental crustal abundance (ACCA) in this document. The ACCA values were used to determine element enrichment or depletion. Background element concentrations from locations more local to the study were provided from soil obtained from the Spokane, Washington, region (San Juan, 1994) and average concentrations of selected elements from sediment from several small streams in the Similkameen River Basin (U.S. Geological Survey, 2004).

Major elements were typically not enriched with respect to ACCA values (E/D ratios less than 1.5), whereas many trace-element concentrations were more variable with E/D ratios often greatly exceeding 2.0. In surficial sediment, 14 trace elements were considered enriched (bolded in table 6), including Ag, As, Au, Bi, Cd, Cu, In, Mn, Sb, Se, Sn, Sr, Te, and Y. Depleted trace elements (E/D ratios less than 0.75) included Be, cesium (Cs), Hf, Nb, Ni, Pb, Ta, Tl, and Zr; for some elements such as Zr and Hf, this result may be affected by poor analytical recoveries of elements from recalcitrant minerals (Hu and Qi, 2014; table 2). In buried cored sediment concentrations of 17 elements were considered enriched (table 7), including 13 of the 14 enriched in surface sediment. The one exception to this result is Sr that was near the cut-off enrichment value of 1.5 in both surface and cored sediment. The additional enriched elements in cored sediment included Co, Mo, V, and Zn. Trace elements showing the greatest enrichment were Te, Mn, and Au, with E/D ratios in surficial sediment of 100, 85, and 38.7, and cored sediment of 430, 105, and 73, respectively. E/D ratios for As and Cu were 6.08 and 2.04 in surficial sediment and 56 and 6.3 in cored sediment, respectively. Seven of the trace elements with elevated concentrations (Ag, As, Au, Bi, Cu, Sb, and Te) were noted to be associated with hydrothermal mineralization and ore deposits that were mined near the Headley, British Columbia, throughout much of the 20th century (Ray and others, 1992; Ray and Dawson, 1994).

As expected, element concentrations measured in the less than 63-µm fraction of sediment were higher than concentrations measured in unsieved or bulk samples for most trace elements (37 of 41 elements analyzed). Trace elements with higher concentrations in all bulk-sediment samples included Ba, whereas trace elements Tl, Rb, and Re were sometimes higher in the bulk fraction. Ratios of elemental concentrations in the less than 63-µm fraction compared to the concentration measured in bulk sediment varied from 0.8 to 19, with an average of about 2.4. The average ratio of sieved to bulk sediment for As was 2.6 in surficial sediment but 5.2 in cored sediment. For Hg, the average ratio was 2.3 in surface sediment and 1.6 in cored sediment (table 8). Mercury (Hg) was not detected in 42 percent of bulk sediment samples from cored sediment but was detected in all samples from the less than 63-µm size fraction.

Element concentrations varied among the size fractions of the two samples that were sieved into five size fractions; 8_5c0-69 and 8_8c30-39 (fig. 2; tables 9 and 10). Gold and As concentrations in the less than 63-µm fractions were typically more than 10 times higher than concentrations in the sand fractions; exceptions to this result being the coarse sand (greater than the 500-µm fraction) of the 8_8c30-39 sample. The trend of increasing element concentrations with decreasing grain size across size fractions less than the 500 µm for many trace elements, including Ag, As, Au, Bi, Cd, Cs, Cu, La, Mn, Mo, Nb, Ni, P, Pb, Sb, Sn, Te, Th, U, and Zn. Concentrations in the very fine sand fractions were typically intermediate between the larger sand size fraction and the less than 63-µm fraction. Composite element concentrations were calculated from the individual size fractions of sample 8_8c30-39 and the composite concentration compared to concentrations measured in the bulk sample. Calculated elemental RPDs typically were less than 5 percent, although the RPD for As concentrations were 27 percent. Notably, the less than 63-µm fractions of both samples were less than 1 percent of the total sample mass.