U.S. Geological Survey
Water-Resources Investigations Report wrir.97-4235

Analysis of Bottom Sediment to Estimate Nonpoint-Source Phosphorus Loads for 1981-96 in Hillsdale Lake, Northeast Kansas

CONTENTS

Abstract
Introduction

Description of Hillsdale Lake Basin
Acknowledgments

Methodology

Sediment-Core Collection and Analysis
Bathymetric Surveys
Estimation of Bottom-Sediment Thickness, Volume, and Mass
Estimation of Phosphorus Loads

Sediment Deposition in Hillsdale Lake
Phosphorus in Hillsdale Lake
Estimated Annual Phosphorus Loads from Nonpoint Sources
Summary
References
Supplemental Information

FIGURES

1. Map showing location of Hillsdale Lake Basin in northeast Kansas and point sources of phosphorus discharge within the basin
2. Map showing location of sediment-coring sites, bathymetric range lines, and boundaries of out-of-channel lake partitions at Hillsdale Lake
3. Map showing estimated bottom-sediment thickness in Hillsdale Lake, 1996
4. Graph showing estimated bottom-sediment thickness in the Big Bull Creek channel of Hillsdale Lake, 1996
5. Graph showing estimated bottom-sediment thickness in the Little Bull Creek channel of Hillsdale Lake, 1996
6. Map showing estimated bulk density of bottom sediment in Hillsdale Lake, 1996
7. Map showing total phosphorus concentrations in bottom sediment of Hillsdale Lake, 1996
8. Graphs showing comparison of 1996 U.S. Geological Survey bathymetric data, 1978 U.S. Army Corps of Engineers topographic data, and 1993 and 1996 U.S. Army Corps of Engineers bathymetric data for Hillsdale Lake:
   8. Range line 6
   9. Range line 5
   10. Range line 4
   11. Range line 15
   12. Range line 13
   13. Range line 2
14. Graphs showing comparison of 1978 U.S. Army Corps of Engineers topographic data and 1996 U.S. Army Corps of Engineers bathymetric data for Hillsdale Lake:
    14. Range line 1
    15. Range line 3
    16. Range line 7
    17. Range line 8
    18. Range line 9
    19. Range line 10
    20. Range line 12
    21. Range line 14
    22. Range line 16
    23. Range line 18
24. Map showing additional bathymetric survey lines completed by the U.S. Geological Survey at Hillsdale Lake in 1996
25. Graphs showing 1996 U.S. Geological Survey bathymetric data for Hillsdale Lake:
    25. Range line 19
    26. Range line 20
    27. Range line 21
    28. Range line 22
    29. Range line 23
    30. Range line 24
    31. Range line 25
    32. Range line 26
    33. Range line 27
    34. Range line 28
35. Graph showing relation of sediment bulk density to percent moisture content for 19 sediment-core samples from Hillsdale Lake

TABLES

1. In-channel bottom-sediment thickness estimated from 1996 bathymetry and 1978 topographic information at bathymetric range-line sites in Hillsdale Lake
2. Estimated length, mean width, mean bottom-sediment thickness, and bottom-sediment volume in submerged stream-channel segments in Hillsdale Lake, 1996
3. Bottom-sediment thickness measured in core samples collected at out-of-channel sites in Hillsdale Lake, 1996
4. Estimated area, mean bottom-sediment thickness, and bottom-sediment volume in out-of-channel lake partitions of Hillsdale Lake, 1996
5. Estimated bottom-sediment volume for Hillsdale Lake, 1996
6. Estimated bulk density of bottom sediment at in-channel sites in Hillsdale Lake, 1996
7. Estimated bulk density of bottom sediment at out-of-channel sites in Hillsdale Lake, 1996
8. Estimated bottom-sediment volume, mean bulk density, and mass in submerged channels of Hillsdale Lake, 1996
9. Estimated bottom-sediment volume, mean bulk density, and mass in out-of-channel lake partitions of Hillsdale Lake, 1996
10. Estimated bottom-sediment mass in Hillsdale Lake, 1996
11. Total phosphorus concentrations and percent moisture content of bottom sediment at in-channel sites in Hillsdale Lake, 1996
12. Total phosphorus concentrations and percent moisture content of bottom sediment at out-of-channel sites in Hillsdale Lake, 1996
13. Latitude and longitude coordinates for the end points of the U.S. Army Corps of Engineers range lines used in the bathymetric surveys of Hillsdale Lake

Abstract

Introduction

Hillsdale Lake is a Federal impoundment on Big Bull Creek in northwest Miami County, northeast Kansas (fig. 1). Completed in 1981, the lake is used for flood control, water supply, fish and wildlife habitat, and recreation. With the rapid expansion of urbanization in southern Johnson County, the use of the lake as a water-supply source is expected to increase substantially. As early as 1985, concerns existed that water quality in Hillsdale Lake may be affected by urbanization and increased point- and nonpoint-source discharges of phosphorus within the basin. Hillsdale Lake is reportedly the third-most, point-source-affected Federal lake in Kansas (Carney, 1994).

Phosphorus is an important nutrient because it is the principal limiting factor for primary production in most freshwater environments (Hakanson and Jansson, 1983). If phosphorus concentrations are too large, algal growth may become excessive and cause taste and odor problems for water suppliers. Additionally, excessive algal growth may be detrimental to aquatic life in, as well as discourage recreational use of, the lake. On the basis of widely accepted chlorophyll-a criteria, Hillsdale Lake has been classified as eutrophic by the Kansas Department of Health and Environment (KDHE) (Carney, 1994). A eutrophic lake contains nutrient-rich water and supports high biotic productivity (Cole, 1983).

Sediment deposition in a lake is important because about 95 percent of the phosphorus in streams tends to adhere to sediment particles (Hem, 1985). The sediment transported by streams and deposited in a lake acts as a sink where phosphorus may be stored and as a source of phosphorus to the overlying water and biota (Baudo and others, 1990). Although the release of phosphorus from lake-bottom sediment may occur under aerobic or anoxic conditions, the release rate is typically much greater during anoxic conditions (Bostrom and others, 1982; Wetzel, 1983). However, Welch and Cooke (1995) state that several mechanisms can combine to produce relatively high phosphorus release rates in shallow, aerobic lakes. Hillsdale Lake is typified by aerobic conditions with water depths ranging from shallow to relatively deep (50 ft or more).

Aside from resuspension, the release of phosphorus from lake-bottom sediment involves a mobilization from particulate to dissolved form followed by transport into the water column. Important environmental factors in the mobilization of phosphorus include redox potential, pH, and temperature. Transport processes include diffusion, turbulence, and bioturbation (Hakanson and Jansson, 1983).

In 1991, the Hillsdale Water-Quality Protection Project was initiated by the Hillsdale Lake Region Resource Conservation and Development Council (RC&D;) to establish long-term protection of the lake and drainage area. The goal of the project is to implement a nonpoint-source pollution-control program to manage further nutrient enrichment of Hillsdale Lake. Specifically, the objectives are to maintain a mean annual total phosphorus concentration of 0.06 mg/L or less in Hillsdale Lake water and a mean annual total phosphorus concentration of 0.10 mg/L, a low-flow mean concentration of 0.05 mg/L, and a runoff mean concentration of 0.40 mg/L in lake tributary water. RC&D; began water-quality sampling of the streams flowing into Hillsdale Lake in 1993. The sampling was performed during storm-runoff and base-flow conditions.

In a recent study, the U.S. Geological Survey (USGS) focused on determining both point and nonpoint sources of phosphorus during low-flow conditions (Putnam, 1997). Results of the study indicate that the point sources in the Hillsdale Lake Basin (fig. 1) are significant contributors to the phosphorus loads in Big Bull Creek and Little Bull Creek during low-flow conditions. The mean concentrations of total phosphorus ranged from 0.05 to 4.9 mg/L in water from 44 sites sampled in the Hillsdale Basin during low-flow conditions from May 1994 through May 1995. Because concentrations in water from these sites equaled or exceeded the RC&D; proposed low-flow total phosphorus concentration of 0.05 mg/L, it was concluded that nonpoint-source contamination is significant during low-flow conditions in the areas of the basin with no point-source discharges (Putnam, 1997).

Several studies have provided estimates of annual point- and nonpoint-source loads of phosphorus to Hillsdale Lake. Montgomery (1991) used a modeling approach and the limited water-quality data available to estimate a mean annual phosphorus load of about 24,000 kg, with respective contributions from point and nonpoint sources of 20 and 80 percent. Carney (1994) also used a modeling approach and the limited water-quality data available to estimate a mean annual phosphorus load of about 19,000 kg, with respective contributions from point and nonpoint sources of 26 and 74 percent. Carney's study also indicated the need for a reduction of annual phosphorus loads from both point and nonpoint sources by 28 to 45 percent (Carney, 1994). Using water-quality data collected in 1994, a study by the Johnson County Environmental Department (JCED) and the RC&D; estimated the annual total phosphorus load contributed to Hillsdale Lake to be about 65,000 kg, of which 92 percent was from nonpoint sources (Holt, 1996). Because of the limited historical water-quality data available and the large hydrologic variability over time, an alternative approach was needed to determine the annual phosphorus loads to Hillsdale Lake from point and nonpoint sources.

A 2-year study by USGS, in cooperation with JCED and the Johnson County Unified Wastewater Districts, was begun in 1996 to estimate the historical phosphorus loads to the lake as well as the historical contributions of point and nonpoint sources. The historical perspective is important because regulatory and remediation strategies may be different depending on the total and relative contributions of phosphorus from the various point and nonpoint sources within the Hillsdale Lake Basin. The specific study objectives were to:

1. estimate the volume and mass of bottom sediment in the lake as well as the mean annual deposition since closure of Hillsdale Dam in 1981;
2. estimate the total mass of phosphorus in the lake-bottom sediment as well as the mean annual load; and
3. estimate the historical mean annual phosphorus loads to the lake from nonpoint sources.

The purpose of this report is to present the results of the USGS study to estimate the historical loads of phosphorus from nonpoint sources to Hillsdale Lake. From a national perspective, the methods and results presented in this report provide guidance and perspective for future lake studies concerned with the issues of sedimentation and water quality.

Description of Hillsdale Lake Basin

The Hillsdale Lake Basin is a 144-mi² area in Douglas, Franklin, Johnson, and Miami Counties, northeast Kansas. The lake has a surface area of 4,580 acres and a water-storage capacity of 68,000 acre-ft at the conservation-pool elevation of 917 ft above sea level. Water storage in Hillsdale Lake began on September 19, 1981, and the lake reached conservation-pool elevation on February 23, 1985. Principal tributaries that contribute flow directly to Hillsdale Lake include Big Bull Creek, Little Bull Creek, Rock Creek, Wade Branch, and Scott Branch (fig. 1). Bedrock in the basin is mostly Pennsylvanian-age limestone and shale. The predominate land use in the basin is agriculture, with about 40 percent of the land used for cultivated crops (Putnam, 1997). About 20 percent of the basin is wooded (Carney, 1994). Land use in the remainder of the basin is dominated by additional agricultural uses (for example, feedlots and pasture) as well as urban and residential uses.

Several point sources of phosphorus discharge are located within the Hillsdale Lake Basin and include wastewater-treatment facilities at Gardner, Edgerton, and the Johnson County New Century Air Center and wastewater lagoons at Connestoga Mobile Home Park and Lone Elm Estates (Carney, 1994) (fig. 1). Nonpoint sources of phosphorus in the Hillsdale Lake Basin include soils, bedrock, septic systems, feedlots, and cropland.

Long-term (1961-90) mean annual precipitation, as computed for Paola, Kansas, located 5 mi south of Hillsdale Lake, is 40.8 in. About 76 percent of the annual precipitation is received during the growing season (April through October) (National Oceanic and Atmospheric Administration, 1993).

Acknowledgments

Methodology

Sediment-Core Collection and Analysis

Bottom-sediment cores were collected during the summer and fall of 1996 at 53 sites (fig. 2) within the lake using a gravity corer. The liner used in the corer was cellulose acetate butyrate transparent tubing with a 2.875-in. outside diameter and a 2.625-in. inside diameter. The coring sites were located to provide a spatially representative sample of lake-bottom sediment both in and outside of submerged stream channels. A total of 22 in-channel and 31 out-of-channel sites were cored. Several of the cores were collected on or near the range lines used in the bathymetric surveys (fig. 2). The latitude and longitude for each coring site was obtained using global-positioning-system (GPS) technology with a horizontal accuracy ranging from 15.1 to 32.8 ft and a mean horizontal accuracy of 22.3 ft.

The sediment cores were refrigerated and typically processed within 48 hours of collection at the USGS laboratory in Lawrence, Kansas. Initially, each core was extracted from the liner and placed in a tray. Maintained at a slight incline to preserve the vertical integrity of the sediment, the core was split longitudinally (that is, lengthwise) to expose the relatively undisturbed inner part of the core for examination and sampling. On the basis of differences in moisture content, texture, and organic matter content (for example, root hairs, sticks, seed pods, leaves), the boundary between the lake-bottom sediment and the underlying pre-impoundment land-surface (or channel-bed) material was determined. The lake-bottom sediment was characterized by higher moisture content, finer texture, and little if any visible organic matter as compared to the underlying material. The thickness of the lake-bottom sediment in each core was measured to the nearest 0.01 ft.

The number of samples extracted from each core was dependent on the thickness of the lake-bottom sediment. Typically, a minimum of 0.2 ft of sediment was required for sample processing. Cores collected at the submerged in-channel coring sites typically had sediment thicknesses of several feet. For these in-channel cores, there was sufficient sediment for the extraction of five samples to assess the distribution of phosphorus concentrations with depth. Cores collected at the out-of-channel sites typically had sediment thicknesses of less than 0.5 ft. For these out-of-channel cores, either one or two samples were extracted depending on the sediment thickness. Each core was divided into segments of equal length. Then, an equal volume of sediment (defined as the space occupied by the sediment particles, water, and gases as measured in cubic units) was extracted longitudinally from both halves of each segment and combined.

The combined sediment volume was homogenized and quartered into subsamples of at least 10 g each. Two of the subsamples were placed in separate sample bottles, labeled, and refrigerated for subsequent phosphorus analysis. One subsample was analyzed for percent moisture content. This subsample was weighed to the nearest 0.10 g, oven dried at about 45 °C for 48 hours, and reweighed. Oven drying of the subsample continued as it was reweighed on a daily basis until no additional moisture loss was observed. The percent moisture content then was computed as follows:

Equation 1

Analyses of sediment samples for percent moisture content were performed at the USGS laboratory in Lawrence, Kansas.

Sediment density as bulk density also was estimated for most of the in-channel cores as well as a few out-of-channel cores for which sufficient sediment volume was available. Bulk density was computed as follows:

Equation 2

The volume for a cylindrical core sample was computed as:

Equation 3

The bulk densities then were converted to pounds per cubic foot for use in subsequent computations. Analyses of sediment samples for bulk density were performed at the USGS laboratory in Lawrence, Kansas.

For the in-channel cores that were divided into five segments of equal length, the second and fourth segments were sampled for bulk-density analysis. The two results then were averaged to estimate a representative bulk density for the site. For the out-of-channel cores, the limited amount of bottom sediment dictated that a single segment, either the middle half or the entire thickness of the core, be sampled. For each segment, one-half of the sediment volume, split longitudinally, was extracted for bulk-density analysis. The other half of the segment was required for phosphorus analysis. The bulk-density samples then were weighed, oven dried, and reweighed as previously described to estimate the mass of the samples.

Analyses of bottom-sediment samples for total phosphorus concentration were performed at the USGS National Water-Quality Laboratory in Arvada, Colorado, using the method described by Fishman and Friedman (1989). In this report, results are reported as mass (dry weight) of total phosphorus. As part of the study, 27 quality-assurance samples in the form of duplicate samples were analyzed. A total of 140 analyses were performed. Data for all bottom-sediment samples analyzed for total phosphorus are stored in the USGS WATSTORE data storage and retrieval system and the U.S. Environmental Protection Agency's STORET system.

Bathymetric Surveys

Bathymetric surveys were performed independently by USGS and COE during the summer of 1996 using GPS technology to record the geographic location of the boat on the lake and a fathometer system to determine the depth to the sediment/water interface. The GPS and fathometer data were recorded digitally using a data-logging unit. The bathymetric surveys were conducted along 16 range lines that were established by the COE in 1978 (fig. 2); (see also figs. 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, and 23 in the "Supplemental Information" section at the end of this report). USGS collected additional bathymetric data along lines running between and across the 16 COE range lines (see figs. 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, and 34 in the "Supplemental Information" section). The additional bathymetric data were not used for the estimates provided in this report as there were no preceding data to provide a basis for comparison. However, the additional bathymetric data may provide supplemental baseline data for future studies of Hillsdale Lake. The latitude and longitude coordinates for the end points of COE range lines used in this study are provided in table 13 in the "Supplemental Information" section. The reliability of the fathometer was verified twice each day by suspending a metal pipe at known depths directly below the transducer. The accuracy of both USGS and COE bathymetric data is ±0.25 ft (Steve Sando, U.S. Geological Survey, oral commun., 1996; Harry Hartwell, U.S. Army Corps of Engineers, oral commun., 1997). A graphic comparison of the 1996 USGS bathymetric data, the 1978 COE topographic survey data, and the 1993 and 1996 COE bathymetric data for several range lines indicated consistency among the data sets (see figs. 8, 9, 10, 11, 12, and 13, in the "Supplemental Information" section). Therefore, it is believed that both the USGS and COE 1996 bathymetric data sets provide representative lake-bottom elevation information.

Estimation of Bottom-Sediment Thickness, Volume, and Mass

Total bottom-sediment volume (sediment plus water and gases) in Hillsdale Lake was estimated using a partitioning approach in which the conservation-pool surface area of the lake was divided into in-channel and out-of-channel components. The channels were divided further into segments as determined by the locations where the bathymetric range lines intersect the channels. The range-line intersections served as the end points for the channel segments (fig. 2). Likewise, the out-of-channel component was divided further into three lake partitions as guided by bottom-sediment thickness information. The out-of-channel lake partitions consist of the Big Bull Creek arm (including Rock Creek), the Little Bull Creek arm, and the main body of the lake near the dam (including Wade Branch and Scott Branch). Respectively, the out-of-channel lake partitions were identified as B, L, and D (fig. 2). Bottom-sediment volume was computed for all components the total surface area multiplied by the mean thickness of the lake-bottom sediment. The in- and out-of-channel results then were combined to provide an estimate of the total volume of bottom sediment in the lake.

The mean thickness of the lake-bottom sediment was estimated using the coring data for the out-of- channel sites and the bathymetric data for the in-channel sites. Because the bathymetric data have an accuracy of about ±0.25 ft, the bathymetric methodology was not considered sensitive enough to reliably measure out-of-channel sediment deposits that may only be a few hundredths of a foot thick. Conversely, the bathymetric data were considered to be the best source of information available for the in-channel thickness estimates given the limitation (described in the following paragraph) associated with the use of a gravity corer to sample soft lake-bottom sediment.

With a gravity corer, a phenomenon referred to as "core shortening" results in a recovered sediment-core sample that may be only about one-half of the actual thickness of sediment penetrated (Emery and Hulsemann, 1964). Core shortening is caused by the friction of the sediment against the inner wall of the sample tube as the corer penetrates the sediment (Emery and Hulsemann, 1964; Hongve and Erlandsen, 1979; Blomqvist, 1985; Blomqvist and Bostrom, 1987). In "normal" lake-bottom sediment at Hillsdale Lake, which is characterized by uniform texture with decreasing water content at depth, core shortening results in a core sample that provides a thinned but complete representation of all of the sediment layers that were penetrated (Emery and Hulsemann, 1964; Hongve and Erlandsen, 1979). Because, as the corer enters the sediment, there is a lag phase before initiation of core shortening (Blomqvist, 1985), the phenomenon is not considered to be a significant factor for the typically thin sediment layer encountered at the out-of-channel coring sites. However, at the in-channel sites where the sediment layer is thicker, substantial core shortening was observed. Therefore, the bathymetric data provided the best estimate of actual bottom-sediment thickness in the submerged channels.

Bottom-sediment thickness in the submerged channels was estimated at the sites where COE bathymetric range lines intersect the channels (fig. 2). At each range-line site (that is, a channel-segment end point), bottom-sediment thickness was computed as the difference between the mean 1996 and the mean 1978 (pre-impoundment) channel-bed elevation. The mean 1996 channel-bed elevation was determined using either USGS or COE bathymetric data depending on which data set provided the most complete information for the channel. Each channel segment was assigned a bottom-sediment thickness computed as the mean of the bathymetry-derived sediment thicknesses for the two range-line sites that defined the segment. Channel segments with the edge of the conservation pool as their upstream limit were assumed to have no sediment accumulation at these locations. This assumption is based on bathymetric evidence that indicates little or no sediment deposition in the upstream-most reaches of the tributaries. For example, the Big Bull Creek, Little Bull Creek, and Rock Creek channels had little if any sediment deposition at their upstream-most range-line sites (fig. 2). The most probable explanation for this condition is that the stream velocity at these sites is sufficiently high to transport incoming sediment farther downstream where eventually it is deposited.

Channel widths were estimated using the 1978 topographic survey data and the 1996 bathymetric data. The channel width at each range-line site was estimated only for the sediment-filled part of the channel as opposed to the entire bankfull channel. The channel length for each channel segment was estimated by digitizing the channels from USGS 1:24,000-scale topographic quadrangles and intersecting the bathymetric range lines using geographic-information-system software.

The sediment-thickness, channel-width, and channel-length data were used to estimate the bottom-sediment volume for each channel segment. Then, the segment volumes were summed to estimate the volume of bottom sediment in each channel as well as the total in-channel bottom-sediment volume for the lake.

The total mass (dry weight) of bottom sediment in the lake was estimated using the same lake components as described previously. Using the representative bulk density for bottom sediment at each of the in-channel coring sites, a mean bulk density was computed for each channel segment and used to convert bottom-sediment volume to bottom-sediment mass. Total in-channel bottom-sediment mass then was estimated as the sum of the sediment mass estimated for the individual channel segments. Because few bulk-density values were estimated for the out-of-channel coring sites, additional bulk-density estimates were computed using a regression-derived relation between bulk density and percent moisture content. A mean bulk density was computed for bottom sediment in out-of-channel lake partitions B and L and used to convert bottom-sediment volume to bottom-sediment mass. Because no bulk-density information was available for out-of-channel sites in partition D, an estimated mean bulk density was derived using the relation of mean in-channel to out-of-channel bulk densities in partitions B and L. A comparison showed that out-of-channel bulk densities were an average of 7.5 percent greater than in-channel bulk densities in lake partitions B and L. Therefore, the mean bulk density for out-of-channel sites in partition D was computed as the mean of the in-channel sites plus 7.5 percent.

Because it was not feasible to accurately distinguish annual layers of sediment deposition in the core samples, mean annual sediment deposition was estimated by dividing the total mass of bottom sediment by the number of years of deposition. Because the reservoir began filling in September 1981, about 15 years of sediment deposition had occurred in the lake at the time the core and bathymetric data were collected during the summer and fall of 1996. Therefore, mean annual sediment deposition was estimated as the total mass of bottom sediment divided by 15.

Estimation of Phosphorus Loads

The mean annual load (mass) of phosphorus added to the lake from point-source discharges was estimated on the basis of water-quality and flow data obtained from JCED and KDHE for the five largest point sources within the Hillsdale Basin. Included were the wastewater-treatment facilities at Edgerton, Gardner, and the Johnson County New Century AirCenter as well as the wastewater lagoons at Connestoga Mobile Home Park and Lone Elm Estates (fig. 1).

The annual amount of total phosphorus exiting the lake was estimated using available COE water-quality data on total phosphorus concentrations in the lake outflow and available USGS data on the quantity of flow. The flow-weighted, mean total phosphorus concentration and the mean annual discharge were determined and then multiplied to provide an estimate of the total phosphorus leaving the lake each year.

The mean annual load of total phosphorus contributed from nonpoint sources was estimated using the following mass balance:

Equation 4

Lacking the data to reliably quantify the residence or storage time of phosphorus in the streams downstream from the point sources, all phosphorus from the point sources was assumed to be transported to Hillsdale Lake each year. The temporal variability of phosphorus concentrations and storage in the lake-water column was assumed to be neglible. Therefore, the water-column component in the mass balance was considered insignificant relative to the other components and was not included.

Sediment deposition in Hillsdale Lake

Estimation of the total and mean annual sediment deposition in Hillsdale Lake was a multistep process. Initially, sediment-core and bathymetric data were collected to quantify bottom-sediment thickness in the lake. Sediment cores were collected both in and outside of the submerged stream channels (fig. 2).

Bathymetry-derived, in-channel bottom-sediment thicknesses (table 1 and fig. 3) were used to estimate the mean bottom-sediment thickness for each channel segment. At each bathymetric range-line site, bottom-sediment thickness was computed as the difference between the 1996 and the 1978 (pre- impoundment) mean channel-bed elevations. A comparison of sediment thickness using the 1993 and 1996 bathymetry at 10 range-line sites indicated that about 13 percent of that thickness was deposited in the channels after 1993. This converts to a mean annual depositional rate of about 4.3 percent of the total thickness for 1994-96 as compared to 7.3 percent for 1981-93. However, the difference could be due simply to 1 or 2 years of exceptionally large or small deposition. A longitudinal view of sediment deposition in the Big Bull Creek and Little Bull Creek channels is provided in figures 4 and 5, respectively. Table 2 provides the estimated channel length, mean channel width, mean bottom-sediment thickness, and bottom- sediment volume for each channel segment.

Bottom-sediment thicknesses measured in sediment-core samples collected at 31 out-of-channel sites are provided in table 3 and figure 5. Because of the atypically thick sediment encountered (1.23 ft) at site D-1 and the proximity to a small tributary of the main Little Bull Creek channel, there is uncertainty as to whether site D-1 is more representative of in- or out-of-channel conditions. Therefore, site D-1 was excluded from the calculation of mean sediment thickness for the out-of-channel sites.

To calculate mean values for the out-of-channel bottom-sediment thicknesses, the six cores with a thickness listed as less than 0.02 ft were assigned a value of 0.01 ft. The out-of-channel lake-bottom sediment ranged in thickness from less than 0.02 to 0.89 ft, with a mean sediment thickness of 0.24 ft. In general, the out-of-channel sediment appeared to be thickest in the Big Bull Creek and Little Bull Creek arms of the lake and thinnest in the main body near the dam (fig. 3). Table 4 provides the estimated area, mean bottom-sediment thickness, and bottom-sediment volume for each out-of-channel lake partition.

The total volume of bottom sediment in the lake was estimated as the sum of the bottom-sediment volumes determined for the in-channel and out-of-channel components of the lake (table 5). The total in- and out-of-channel bottom-sediment volumes were 51,300,000 and 40,100,000 ft³, respectively. Therefore, the total estimated volume of bottom sediment in the lake was 91,400,000 ft³ or about 2,100 acre-ft. In comparison, the COE, using bathymetry only, estimated the total bottom-sediment volume in 1993 to be 1,928 acre-ft (Phil Snell, U.S. Army Corps of Engineers, oral commun., 1997). The 2,100 acre-ft of sediment occupies about 3 percent of the lake's original water-storage capacity of 68,000 acre-ft at conservation pool.

Respectively, the in- and out-of-channel components represent about 6 and 94 percent of the total conservation-pool surface area of the lake. However, the in-channel component represents about 56 percent of the total volume of bottom sediment. Thus, as noted by Thornton and others (1990), initial sediment deposition is greatest in the submerged channels.

Bottom-sediment mass was estimated as the bottom-sediment volume multiplied by the mean bulk density of the sediment. Bulk densities were estimated at 18 in-channel (table 6) and 12 out-of-channel (table 7) sites in the lake (fig. 6). Of the 12 out-of-channel sites, bulk-density values for 8 sites were estimated using a relation between bulk density and percent moisture content. A statistical analysis was performed using 19 sites for which bulk-density and percent moisture-content information was available. For sites having data for multiple depth intervals, the uppermost depth interval was used on the assumption that it would most closely approximate conditions at the out-of-channel sites for which bulk density was being estimated. A Pearson's correlation coefficient of -0.94 (R² = 0.89) was calculated, thus indicating a strong inverse relation between bulk density and percent moisture content (see fig. 35 in the "Supplemental Information" section). A regression analysis was used to determine the following relation:

Equation 5

Estimated bulk densities ranged from a mean of 28.0 lb/ft³ at site D-2 (in-channel site) to 59.9 lb/ft³ at site BB-4 (out-of-channel site), with an overall mean of 42.8 lb/ft³. The range of estimated bulk densities is consistent with bulk densities estimated at nine sites within the lake by COE in 1993 (Al Coop, U.S. Army Corps of Engineers, written commun., 1996). COE bulk densities ranged from 30.0 to 50.1 lb/ft³. In general, the bulk density of channel sediment was highest in the upstream parts of the lake and lowest downstream near the dam. For example, the bulk density of bottom sediment in the Big Bull Creek channel ranged from a mean of 49.4 lb/ft³ at upstream site BB-13 to a mean of 28.8 lb/ft³ at site D-4 near the dam (fig. 6). The out-of-channel sites did not exhibit the same upstream-to-downstream trend of decreasing bulk densities.

The total in-channel mass of bottom sediment in the lake, estimated as the sum of the sediment mass computed for the individual channel segments (fig. 2), was 2,136 million lb. Total out-of-channel bottom-sediment mass, estimated as the sum of the sediment mass computed for the three lake partitions (fig. 2), was 1,837 million lb. Therefore, the total estimated mass of bottom sediment in the lake is 3,973 million lb of which about 54 percent is in the submerged channels. Tables 8 and 9 detail the bottom-sediment mass estimated for the in- and out-of-channel lake components, respectively. Table 10 provides an overall summary of bottom-sediment mass estimated for the lake.

Annual sediment deposition was estimated by dividing the total mass of bottom sediment in the lake by the number of years of deposition (that is, 15). The mean annual sediment deposition was estimated to be 265 million lb or 120 million kg.

Phosphorus in Hillsdale Lake

Concentrations of total phosphorus in the lake-bottom sediment were determined for both in- and out-of-channel sites. To facilitate comparisons, a depth-integrated mean was calculated (excluding the duplicate samples) to provide a representative phosphorus concentration for each site sampled (fig. 7). The in-channel phosphorus concentrations (tables 11a, and 11b) showed no apparent trend with depth. Likewise, there was no apparent upstream-to-downstream trend observed for either the in- or out-of-channel (table 12) phosphorus concentrations. Overall, the depth-integrated, mean phosphorus concentrations for the in-channel bottom sediment ranged from 410 to 810 mg/kg, with a mean of 610 mg/kg. The depth-integrated, mean phosphorus concentrations for the out-of-channel bottom sediment ranged from 210 to 710 mg/kg, with a mean of 400 mg/kg (excluding site D-1 due to uncertainty as to whether site D-1 is more representative of in- or out-of-channel conditions). These results are in general agreement with COE phosphorus concentration data determined for seven sites within the lake in 1993 and(or) 1995 (Al Coop, U.S. Army Corps of Engineers, written commun., 1996). For COE sites having multiple-date phosphorus concentration data, the mean phosphorus concentration was calculated as a basis for comparison with the single-date USGS phosphorus concentration data. The adjusted COE phosphorus concentration data ranged from 470 to 900 mg/kg, with a mean of 680 mg/kg. Sedimentary rocks may contain phosphorus concentrations in the range of about 280 to 730 mg/kg (Hem, 1985). Thus, a substantial part of the observed phosphorus concentrations in the Hillsdale Lake bottom sediment may originate from soils and limestone and shale bedrock within the basin.

For quality-assurance purposes, 27 duplicate bottom-sediment samples were analyzed for total phosphorus concentration. In comparison, the duplicates differed from the original samples by an average of ±9 percent. This is considered to be within the range of acceptable variability for the analysis of total phosphorus concentrations in bottom sediment.

The total mass of phosphorus in the lake-bottom sediment was estimated by calculating the mass of phosphorus separately for the in- and out-of-channel components of the lake and then summing. The in-channel mass of phosphorus was estimated as the total in-channel mass of bottom sediment multiplied by the mean phosphorus concentration for all the in-channel sites (that is, 610 mg/kg). Likewise, the out-of-channel mass of phosphorus was estimated as the total out-of-channel mass of bottom sediment multiplied by the mean phosphorus concentration for all the out-of-channel sites (that is, 400 mg/kg). The mean phosphorus concentrations were used given that no upstream-to-downstream trend in phosphorus concentrations was observed in the lake-bottom sediment. The total in- and out-of-channel mass of phosphorus was estimated to be 591,000 and 333,000 kg, respectively. Therefore, the total mass of phosphorus in the bottom sediment was estimated to be 924,000 kg, of which 64 percent is in the submerged channels. The mean annual load of phosphorus in the bottom sediment was estimated to be 62,000 kg for 1981-96.

Estimated annual phosphorus loads from nonpoint sources

Summary

References

Supplemental Information

For additional information                Copies of this report can 
   write to:                                 be purchased from:
District Chief                            U.S. Geological Survey
U.S. Geological Survey                    Information Services
4821 Quail Crest Place                    Box 25286
Lawrence, Kansas 66049-3839               Federal Center
                                          Denver, CO  80225-0826