A
METHOD FOR DETERMINING SUSPENDED-SEDIMENT AND TRACE-METALS TRANSPORT IN THE
CLARK FORK RIVER,
WESTERN MONTANA
J.
R. Knapton
U.S. Geological Survey
Helena, MT 59626
Abstract--Suspended
sediments are thought to be an important transport mechanism for toxic trace
metals that have been identified in the Clark Fork. Hydrologic data from a
network of water-quality stations combined with data simulation will be used to
calculate suspended-sediment and trace-metals transport. An ultimate goal of
such an investigation might be the development of a predictive model.
INTRODUCTION
Previous and current studies
within the Clark Fork River drainage of western Montana (fig. I) have documented
the existence of toxic trace metals that have been released to the riverine
system from mines and associated metals-extraction processes in headwaters
areas. Investigations to date have focused on such problems as source areas,
chemical sinks and effects on biota and ultimate fate of the metals. Although
commonly studied independently, these problems are linked by the movement of the
constituents through the Clark Fork system. Until conclusive studies have
documented the transport mechanisms and quantified trace-metals movement,
solutions to many of the trace-metals problems may be elusive. The methods
described herein provide a means for determining the transport of suspended
sediment and trace metals in the Clark Fork.
THE ROLE OF SEDIMENT IN TRACE-METALS TRANSPORT
Historically, many of the analyses for trace metals were performed on filtered
samples and, thus, only the concentration of the dissolved phase was determined.
Studies on such rivers as the Amazon and Yukon, as described by Gibbs (1), have
indicated that the dissolved phase commonly accounts for a small fraction of all
the trace metals in transport. In most aquatic systems, the suspended sediment
has a trace-metals concentration that is much greater than the concentration
dissolved in the water column. Therefore, an assessment of trace-metals
transport in a riverine system needs to consider the fraction carried in
suspension as well as the dissolved fraction.
Although suspended sediment as a whole can be examined as a
transport mechanism for trace metals, additional insight can be gained from a
knowledge of the particle-size distribution. The most significant factor
controlling sediment capacity for retaining trace metals is grain size,
according to Jenne
A recent view of the role of clays as metals concentrators
implies that clays function as substrates for the precipitation and flocculation
of organic matter and secondary minerals such as hydrous iron and manganese
oxides. Those secondary minerals, in turn, become substrates for the
accumulation of trace metals rather than the clays themselves, according to
Jenne (3). There is some debate whether accumulation takes place by organics and
secondary minerals or whether there is direct physical and chemical interaction
with the clays. However, there is general agreement that the largest
concentrations of metals are present with the clays and that the following clay
minerals have a decreasing order of association with trace metals: (I)
montmorillonite; (2) vermiculite; and (3) illite, chlorite and kaolinite.
Investigators recognize that sediments in suspension and fine
sediments present in bed material readily revert from one medium to the other in
response to stream flow dynamics. Accordingly, analytical techniques that
measure trace-metals concentrations from sediment, either in suspension or in
bed material need to be the same. The use of similar analytical techniques would
lend compatibility to a variety of independent studies being undertaken
throughout the Clark Fork basin. Recent analytical techniques developed by Horowitz
(2) provide for direct measurement of trace metals from suspended sediment after
extraction of sediment from the water-sediment mixture. The technique uses a
traditional approach in analyses of trace metals from bottom sediments.
MEASUREMENT
OF SEDIMENT AND TRACE-METALS TRANSPORT
A strategy has been devised to quantify the transport of trace metals in the
Clark Fork, based on the association between sediment and trace metals in stream
environments. The strategy involves establishing a monitoring program in which
the actual transport of suspended sediment will be measured. The transport of
trace metals will be accomplished through a combination of direct measurements
and indirect estimates based on correlations between selected measurements.
Hydrologic data will be collected from a network of fixed stations on the Clark
Fork mainstem and at the mouths of major tributaries. Cost constraints require
that the Clark Fork study be divided into specific stream reaches and that data
be collected one reach at a time from the headwaters to the mouth. The logical
first reach is from the confluence of Silver Bow and Warm Springs Creeks to the
Turah Bridge, just upstream from Milltown Reservoir.
Suspended-sediment stations have been established on the
Clark Fork at Deer Lodge and Turah (fig. 1). Sampling at these two sites will
enable the direct measurement of suspended-sediment loads into and out of the
most up- stream reach of the Clark Fork. Continuous-record stream flow data from
existing gages will allow quantification of sediment on an annual basis as well
as for short-term hydrologic events.
Sampling frequencies for these two mainstem sediment stations
will range from two to three times a week during low flow to daily throughout
medium to seasonal high flows. During periods of storm runoff, sampling will be
intensified further to better characterize suspended-sediment concentrations
during rising, peak, and falling stages of stream flow. Methods of sampling are
de- scribed in the report of the u. S. Government (5).
The most accurate way to quantify trace-metals transport at
the mainstem stations would be to independently analyze each set of sediment
samples for the desired trace metals (arsenic, cadmium, copper, iron, lead,
manganese, and zinc) and to couple these concentrations with daily stream flow
to determine loads. Laboratory costs, however, make this approach prohibitive.
Therefore, samples for trace metals will be collected less frequently and
trace-metals loads will be determined by indirect methods.
Depth-integrated, cross-sectional samples analyzed for
dissolved and suspended trace metals will be collected at the two mainstem sites
and near the mouths of four major tributaries: Little Blackfoot River, Flint
Creek, Rock Creek, and Blackfoot River (fig.
1). Samples from the mainstem sites
will be collected approximately bimonthly during low flows (August through
March) and twice monthly during high flows (April through July). Samples from
the tributary stations will be collected primarily during high flow periods.
Attempts also will be made to collect additional samples at all mainstem and
tributary sites during periods of runoff from major storms. Information from the
tributaries can be used to determine if significant amounts of trace metals are
being contributed to the mainstem. Should preliminary sampling show that they
are, sampling at the tributaries can be increased for better quantification.
Based on the analyses of the suspended trace-metals samples,
regression equations will be developed to determine the relation between
concentrations of individual trace metals and sediment concentrations. Complete
particle- size distribution of the sediment from the selected samples also will
be determined to evaluate the size class of sediments to which the metals are
commonly associated. More meaningful regression equations may be established
using correlations with a specific size class of sediment. Metals concentrations
for periods between selected samples will be estimated by simulating
trace-metals concentrations using the regression equations and sediment
concentrations from the daily sediment samples. The simulated trace-metals
concentrations and stream flow data can be used to calculate trace-metals loads.
The dissolved fraction of trace-metals loads will also be
calculated indirectly by relating measured metals concentrations to
corresponding stream- flow discharges at the time of sampling and thus
developing regression equations. The gaps in concentration data can then be
simulated from the regression equations and mean daily stream flow discharges.
After simulating the concentrations, loads can be calculated in the same manner
as for the suspended trace metals.
Throughout many river systems and in selected reaches of most
others, the movement of bed material along the bottom may account for a
measurable amount of the total sediment load. These bottom sediments are not
accounted for by standard suspended-sediment sampling techniques. Therefore,
trace metals that may be transported with bed sediments are not included in load
calculations.
To limit as much as possible the bed load fraction of
sediment, cross-sections for sampling have been located at sites where stream
velocities are sufficient to retain fine sediments in suspension, thus keeping
the streambed relatively free of fine material.
PREDICTIVE
METHODS
Because
aquatic systems are dynamic, the foregoing program will only describe existing
conditions during the data-collection phase. An ultimate goal might be the
development of a predictive model that would be used to determine sediment and
metals transport in the Clark Fork in response to various hydrologic conditions
and plans of resource management.
A primary requirement for developing such a model would be
the acquisition of a database consisting of temporal and spatial input from the
network stations. Any viable model needs to address both physical and chemical
aspects of trace-metals transport. Several physical transport models presently
exist. Research efforts may be required to identify chemical transformations
that occur in the various stream environments of the Clark Fork. A general
approach, then, might be to modify an existing physical transport model for
application to the Clark Fork system and to incorporate the necessary chemical
elements that are identified.
LITERATURE
CITED
1.
Gibbs, R. 1977. Transport phases of transition metals in the Amazon and Yukon
Rivers. U.S. Geological Society of America Bulletin 88: 829-843.
2.
Horowitz, A. 1984. A primer on trace metal-sediment chemistry. Geological Survey
Open-File Report 84-709.
3. Jenne, E. 1976. Trace metals sorption by sediments and soils-sites and processes. In: Chappell, W.; and Peterson, K., eds., Symposium on molybdenum. Marcel-Dekker, New York. 2: 425-553.
4.
Jenne, E.; Kennedy, V.; Burchard, J.; and Ball, J. 1980. Sediment collection and
processing for selective extraction and for total metals analysis. In: Baker,
R., Contaminants and sediments. Ann Arbor Science Publishers, Ann Arbor, MI. 2:
169-189.
5.
U.S. Government (agencies of). 1978. National handbook of recommended methods
for water-data acquisition, Chapter 3- Sediment.