EFFECT OF MINING AND RELATED ACTIVITIES ON THE SEDIMENT TRACE ELEMENT GEOCHEMISTRY OF THE SPOKANE RIVER BASIN, WASHINGTON, USA

Cecile A. Grosbois¹, Arthur J. Horowitz², James J. Smith² and Kent A. Elrick²

¹BRGM, Direction de la Recherche, Avenue C. Guillemin, BP 6009, 45065 Orleans Cedex 02, France (grosbois@usgs.gov)

² U.S. Geological Survey, Peachtree Business Center, 3039 Amwiler Road, Atlanta, GA 30360, USA

Abstract

Surface sediments in the Spokane River Basin are enriched in Pb, Zn, As, Cd, Sb and Hg relative to local background levels. Maximum enrichment occurs in the Upper Spokane River in close proximity to Lake Coeur d'Alene. On average, enrichment decreases downstream. Subsurface sediments also are enriched in Pb, Zn, As, Cd, Sb and Hg relative to background levels. Enrichment began between 1900 and 1920 in the middle of the basin; this is contemporaneous with similar findings in Lake Coeur d'Alene (the upstream source of the Spokane River), as well as the completion of Long Lake Dam (1913). In the most downstream part of the basin, enrichment began between 1930 and 1940. This temporal shift may reflect the latter's greater distance from the Coeur d'Alene Basin, but is more likely the result of the completion of Grand Coulee Dam (1934-1941) which backed up the Spokane River and elevated water levels by about 30 m in the most downstream section of the basin.

Introduction

The Spokane River Basin (SRB)is located in eastern Washington and begins at the northern outlet of Lake Coeur d'Alene (CDA). It is about 180 km long and is 110 km downstream from the CDA mining district. Water flow is mostly from east to west, beginning at Lake CDA (Idaho) and extending to the Columbia River (Lake Roosevelt). Land use in the basin is mainly agricultural, residential and recreational; it is not known to contain significant ore deposits like those found in the CDA Mining District. The Spokane River is heavily regulated by dams to control floods and to produce hydropower. For purposes of this study, the basin can be divided into 5 units based on hydrologic similarities and on dam locations (1) Upper Spokane River (60 km) - begins just after Post Falls Dam, is the major outflow of Lake CDA, and extends to Upriver Dam; (2) Nine Mile Reservoir (10 km) - begins downstream from the city of Spokane and extends to Nine Mile dam; (3) Long Lake (40 km) - begins northwest of the city of Spokane, downstream from Nine Mile Dam and extends to Long Lake Dam; (4) Post Long Lake (12 km) begins at Long Lake Dam and extends downstream to Little Falls Dam; and (5) Spokane River Arm (27 km) - begins at Little Falls Dam, and extends to Lake Roosevelt, which was created on the Columbia River between 1934 and 1941 as a result of the completion of Grand Coulee Dam.

In 1998, the U.S. Environmental Protection Agency began a remedial investigation/feasibility study (RI/FS) in the CDA and Spokane River Basins to determine if the Bunker Hill Superfund site should be extended further downstream. One of the major goals of this study was to evaluate downstream sediment-associated trace element concentrations to determine if the enrichments observed in the CDA Basin extended into the SRB. The results of that study are described herein.

Methods

Ninety-three surface grab samples were collected throughout the SRB in 1998/1999. Sampling sites were located using a portable global positioning system (GPS). The majority of the bed sediment samples were collected using stainless-steel grab samplers. However, in major sections of the Upper Spokane River, surface samples had to be collected by hand using non-metallic scoops because in this section of the basin (1) the water was not deep enough; (2) the current speed was too fast and boat access was severely limited; and/or (3) the river bottom was obstructed by large cobbles and bolders. In addition to the surface samples, 6 gravity cores were collected using a stainless-steel device with a clear polycarbonate liner and a non-metallic core catcher. All analytical procedures were similar to those employed in prior studies in and around Lake CDA (Horowitz, et al., 1993; 1995).

Results and Discussion

Although there is a fair degree of scatter in the data, based on the chemical ranges (minimum and maximum values), it is apparent that the surface sediments in the SRB are enriched in Ag, Pb, Zn, Cd, Hg, As, Sb, and Zn (Table 1). Not surprisingly, these are the same elements enriched throughout the CDA River and Lake CDA (Horowitz, et al., 1993; 1995). By far, the most elevated trace element levels occur in the Upper Spokane River, and are sufficiently high to represent both a potential aquatic as well as a human health problem (Shawn Sheldrake, U.S. EPA, written comm., 1999). The <63-µm fractions from the Upper Spokane, which are similar in grain-size distribution to the surface sediments in Lake CDA, also contain chemical concentrations on a par with those found in the lake (Table 1). On the other hand, as in Lake CDA, concentrations of Ba, V, Li, Be, P, Sr, Se, Cr, Ni, Ti, TOC, TC and TS do not appear enriched in Spokane River Basin surface sediments.

Table 1: Average concentrations of surface sediment-associated enriched trace elements in
various sections of the Spokane River Basin.
	Ag	Pb	Zn	Cd	As	Sb	Hg
Category	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg
Canadian Background Levels		23	65	1.1	4.2		0.10
U.S. Background Levels		23	88		7.0	0.6	0.05
Lake CDA Background Levels	<1	24	110	2.8	4.7	0.7	0.05
Lake CDA Surface Average	6.0	1900	3600	62	151	23	1.8
Spokane River Bulk Background	<0.5	34	89	0.5	5.2	0.9	0.04
Spokane River <63-µm Background	<0.5	34	85	0.2	5.6	0.7	0.05
Upper Spokane Bulk	0.9	320	1200	5.8	14	3.4	0.10
Upper Spokane <63-µm	3.4	1000	3500	22	33	7.7	0.35
Nine Mile Reservoir Bulk	<0.5	41	430	2.3	8.9	1.1	0.04
Nine Mile Reservoir <63-µm	0.9	65	710	6.0	12	1.5	0.05
Long Lake Bulk	0.5	110	960	8.2	13	1.6	0.08
Long Lake <63-µm	0.8	120	940	8.2	13	1.5	0.07
Post Long Lake Bulk	<0.5	22	310	1.9	8.0	1.1	0.02
Post Long Lake <63-mm	0.5	39	340	1.5	15	1.7	0.09
Spokane River Arm Bulk	0.5	63	580	2.8	14	1.8	0.05
Spokane River Arm <63-µm	<0.5	110	920	4.1	18	2.0	0.09

The longitudinal/downstream variability of enriched trace element concentrations throughout the basin can be quite marked (Table 1). On average, for the bulk sediments, it ranges from factors as low as about 1.25 for Hg to as high as about 10 for Pb, whereas for the <63-µm fraction, it ranges from factors as low as 1.2 for Ag and Sb to as high as about 10 for Pb, Zn, Cd, and Hg. Longitudinal variability is readily apparent from plots of calculated downstream enrichment factors for each section of the basin. Enrichment factors [EF (also termed change ratios)] were calculated for both the bulk and the <63-µm fractions for each section of the SRB by dividing the individual sample concentrations by the average background (either for the bulk or the <63-µm fraction) concentration (e.g., Horowitz, et al., 1999). Without altering the basic underlying spatial patterns, smoother plots were obtained by using the mean chemical concentrations for each section, after first normalizing these values to their respective Ti concentrations to limit the effects of differential dilution by locally derived material .

As noted previously, the concentrations of Ba, V, Li, Be, P, Sr, Se, Cr, Ni, Ti, TOC, TC and TS do not appear enriched in SRB surface sediments. Hence, it is not surprising that the EF's for these elements are close to 1 for both the bulk sediments and the <63-µm fractions. The order of enrichment for the elevated trace elements is, on average, Zn>Cd>Pb>As>Sb>Hg in the bulk sediments and Cd>Zn>Pb>As> Sb>Hg in the <63-µm fraction. Based on basinwide plots of the EF's for both the bulk and the <63-µm fractions, only sediment-associated Zn and Cd are enriched throughout the basin; whereas As, Sb, and Hg only appear to be enriched in the Upper Spokane River, and Pb only may be enriched as far downstream as Long Lake. It should be noted that between Long Lake and the Spokane River Arm, the EF's for all the bulk sediment-associated enriched elements, and the <63-µm enriched Zn, Cd, and Pb display small but significant increases. This may be the result of one or all of several physical/chemical factors including: (1) post-depositional chemical remobilization in Long Lake followed by subsequent physical remobilization of the precipitates; (2) higher water velocities in the relatively short Post Long Lake section; (3) greater dilution with locally derived unenriched material in the Post Long Lake section; and/or (4) a potential secondary source for these trace elements in the Spokane River Arm.

Trace element partitioning in selected bulk surface sediments was inferred using the same two-step sequential extraction procedure employed with bulk surface sediments from Lake CDA (Horowitz, et al., 1993). This procedure partitions trace elements among 3 phases: an operationally-defined iron oxide phase, an operationally-defined organic/sulfide phase, and a residual or matrix-bound phase. As there was little or no sulfur present, (nor heavy minerals, Grosbois, unpublished data, 2000) trace element concentrations ascribed to the operationally-defined organic/sulfide phase should be viewed as being predominantly associated with organic matter. In the surface sediments of the SRB, on average, the majority of the Zn (~80%), Pb (~70%), and Cd (~95%) are associated with the operationally-defined iron oxide phase. Further, the majority of the Sb (>90%) appears associated with a residual phase. Although the percentages are somewhat lower for the Zn, Pb, and Cd, and somewhat higher for the Sb, this is similar to the patterns observed in Lake CDA surface sediments and would further support the view that the origin of the trace element-rich SRB sediments is Lake CDA and/or the CDA basin (Horowitz, et al., 1993).

Most of the subsurface sediments have similar physical characteristics. They appear homogeneously dark brown. No evidence of gas production (relict bubbles) or biological activity was observed in any of the cores. Unlike the cores collected in Lake CDA, there was no evidence of ash from the 1980 eruption of Mt. St. Helens (Horowitz, et al., 1995). None of the SRB subsurface sediments displayed any banding, which demarked the metal-rich subsurface sections in Lake CDA (Horowitz et al, 1995). Grain size distributions appear similar in all the cores, except SRC3 and SRC6, and are dominated by the presence of fines (<63-µm) in the upper 65 to 75% of their length. Coarse sand is present in the lower sections of cores SRC1, SRC2, SRC4, and SRC5; gravel also is present near the bases of SRC2 and SRC4. Core SRC6, collected in the Upper Spokane River, is quite short (27 cm) and markedly more organic-rich than the others; it contains an abundance of sticks and leaves. Unlike the other 4 cores, neither SRC3 nor SRC6 bottomed in coarser material.

As observed in their surface counterparts, the majority of the subsurface sediments also appear enriched in Pb, Zn, Cd, As, Sb and Hg relative to local background levels. On average, cores SRC2 and SRC3, collected at the downstream end of Long Lake, and SRC6, collected in the Upper Spokane River, contain the most elevated trace element concentrations. The ranges of enriched trace element concentrations associated with subsurface sediments vary from levels similar to those recorded in the surface sediments (Ag, Mn,) to substantially higher levels (Hg, Pb, Cd). This would seem to indicate varying input levels for the enriched constituents through time. On the other hand, although displaying some level of variability, the downcore distribution patterns for all the unenriched trace elements, in all the cores, remain fairly constant. All the unenriched concentrations tend to fall within the range of variability observed in the surface sediments (Table 1).

The downcore distribution patterns for the enriched trace elements, especially Pb, Zn, Cd, As, Sb, and Hg, observed in cores SRC1, SRC2, SRC4, and SRC 5 are all similar to each other (Fig. 1). This occurs despite the marked differences in their lengths. The concentrations of the enriched trace element group are at or near surface sediment background levels at the bases of these cores. About one third of the way upcore, concentrations rapidly rise, and then remain relatively constant until the last third of their lengths. At that point, there is a small but significant decline in the concentrations; however, the levels never return to the background levels observed in the basal sediments of these cores (Fig. 1). It could be argued that the background trace element concentrations encountered in the basal sediments from these cores resulted from the marked increase in mean grain size noted previously. To clarify this point, aliquots of <63-µm material were separated from the basal sections of each of these cores. Although the concentrations in the grain-size limited aliquots were higher than in the bulk aliquots, they still remained within the range of background levels noted in the surface sediments. Hence, the background levels recorded in the basal sediments from these cores did not primarily occur as the result of grain-size differences. It would appear that deposition of these basal sediments pre-dates the trace element enrichments noted in both the surface and subsurface sediments throughout the SRB.

One of the major reasons for obtaining cores in selected sections of the Spokane River was to try to reconstruct the recent geochemical history of the basin from just before the inception of upstream mining and ore-processing activities in the CDA Basin, through to the present (a period of about 120 years). To that end, both ¹³⁷Cs and ²¹⁰Pb activities were determined for all the sampled layers in cores SRC1 (Long Lake), SRC2 (Long Lake), and SRC 5 (Spokane River Arm). The downcore ages and the geochemical patterns for both Long Lake cores are remarkably consistent (Fig. 1). Trace element enrichment began in this section of the Spokane River Basin between 1900 and 1920. This is contemporaneous with the onset of trace element enrichment in Lake CDA (Horowitz, et al., 1995). Interestingly, the onset of trace element enrichment in Long Lake also is contemporaneous with the 1913 closing of Long Lake Dam (John Roland, Washington Department of Ecology, pers comm, 1999).

In the most downstream part of the basin (Spokane River Arm), trace element enrichment began between 1930 and 1940. The temporal shift between Long Lake and the River Arm may reflect the latter's greater distance from the CDA Basin, the proposed source of the enriched trace elements in the SRB. However, it is more likely the result of the completion of Grand Coulee Dam (1934-1941) which backed up the Spokane River, and elevated water levels in the River Arm by about 30 m (Brett Smith, USGS, pers. comm, 1999). It should be noted that core SRC5 was collected in 28 m of water. Hence, prior to the closing of Grand Coulee Dam, this section of the River Arm was barely under water, if at all. Hence, the later date for the inception of trace element-rich deposition in this section of the SRB does not preclude the transport of trace element rich sediments through this area contemporaneously with that found for both Long Lake and Lake CDA. The small but significant decline in trace element enrichment noted in the upper third of all the Spokane River cores appears to have begun around 1970 in the three dated cores (Fig. 1). This is contemporaneous with, and may be the result of the installation of tailings ponds in the CDA Basin, which were designed to limit the downstream dispersion of mine waste.

References

Horowitz A.J, Elrick K.A and Cook R.B. (1993), Hydrological Processes, 7: 403-423.

Horowitz A.J. Elrick K.A. Robbins J.A. and Cook R.B. (1995), Hydrological Processes, 9: 35-54.

Horowitz A.J., Meybeck M. Idlafkih and Biger E. (1999), Hydrological Processes, 13: 1329-1340.