United States
             Environmental Protection
             Agency
Office of Water
Nonpoint Source Control Branch
(WH 585)
Publication 440/5-89-001
August 1989
$EPA     Off-site  Assessment
                 a national workshop
                                                Printed on Recycled Paper

-------
Off-site Assessment Workshop
   Proceedings of a National Workshop
      November 15,1988 • St. Louis, Missouri
  Cosponsored by the North American Lake Management Society and
     the U.S. Department of Agriculture Soil Conservation Service
        U.S. Environmental Protection Agency
                 Washington, D.C.
                   August 1989
              U.S. Environmental Protection Agency
              Region 5. library (Pt.ttJ)
              77 Wast Jackson Boulevard, |2tfi ftaf
              Chicago, ft 60604-3590

-------
                              Judith Taggart, Publications Editor
                              Lura Taggart, Production Manager
                                 Cover sketch by Patricia Perry
Points  of view expressed in this proceedings do not necessarily reflect the views or policies of the
U.S. Environmental Protection Agency nor of any of the contributors to this publication. Mention of trade
names and commercial products does not constitute endorsement of their use.
                         Copies of the Proceedings may be ordered from the

                                   Nonpoint Source Branch
                          Assessment and Watershed Protection Division
                             U.S. Environmental Protection Agency
                                      401M Street, S.W.
                                    Washington, DC 20460
                                           <»?f"v,:;r?.u";3 -C.
                                      il-J'-t  ,'iatdiJiiJ •.."A-v.
                              £I .to- vfiu ^  i'.-- •> '•'•»' **'

-------
Contents
                                                                       	V
Foreword	
James Meek

Quantification of Lake Sedimentation Rates UtUizmg Radioisotopes Present in the Environment   	1
S.C. Mclntyre, J.W.Naney, andJ.R. McHenry
                                                                                              7
Airphoto Inventories for Pinpointing Nonpoint Sources	
Frank R. Perchalski

 Contaminated Sediment Assessments	
 Anthony G. Kizlauskas and Bruce Kitsuse

 Physical Fish Habitat Components as Measures
 Donald M. Martin
of Beneficial Use Health	21
                                                                                             23
 EPA Remote Sensing Resource for Lake Management  ........................
 Mason J. Hewitt, III, Thomas H. Mace, and Ross S. Lunetta
 Modeling Linked Watershed and Lake Processes for Water Quality Management Decisions   ........ 29
 R.M. Summer, C.V.Alonso, andR^.. Young
                                                                                             39
 USDA Water Quality Program  ...................................
 Jam es Krider an d Bruce Kirschner
  The SCS Water Quality Indicators Guide: Surface Waters - A Tool to Assess Surface Water
  Quality Problems
  Charles R. Terrell and Patricia Perfetti
                                                                                              45
  Water Quality of the Missouri River   	
  John R. Rowland and John C. Ford

  Closing Remarks	
  Douglas A. Ehom

-------
Foreword
.^^^— ^-^— ^^^— "^^^•^^•^™^^^—

   The mandate is clear: as a nation, we must solve the problem of nonpoint source pollution
of our water bodies. And we must do it together: Federal and State programs working in tan-
dem  States are taking the lead with their nonpoint source pollution management programs,
but they are seeking information and assistance from the Federal level on funding and re-
    nsponse, pursuant to section 319(e) of the Clean Water Act, EPA is making available
 through publications, workshops, and other means, information regarding management
 practices and implementation methods.  Specifically, this information involves the follow-
 ing:
                 • technical training;

                 • educational and informational materials; and

                 • technology transfer materials.

    This Proceedings resulted from the first of a series of technical workshops EPA will
 cosponsor. As with the other workshops EPA has planned for the future, its objective was to
 facilitate technology transfer and the exchange of information.
    This Off-site Assessment Workshop, designed for local nonpoint source control profes-
 sionals, helped participants recognize and quantify nonpoint source problems and connect
 those problems with contributing sources.
    Future technical transfer efforts will include a newsletter, a clearinghouse, a nonpoint
 source modeling workshop, and a symposium on the applications of remote sensing for non-
 point source identification and tracing of best management practices.
    Solving nonpoint source pollution is truly a cooperative effort -among citizens, their
 communities, their States, and their Federal government.  As States develop their manage-
 ment programs, the sharing of information - among them and with the national program -
 becomes integral to the ultimate solution.
     At the Federal level, State efforts are coordinated by the Nonpoint Source Coordinator in
  each EPA Region. Contact your regional coordinator for information on nonpoint source
  management activities in your State, and for publications such as this proceedings.

                                        James W. Meek
                                        Chief, Nonpoint Source Control Branch, Assessment and
                                        Watershed Protection Division, U.S. Environmental
                                        Protection Agency, 401 M St., S.W., Washington, DC 20460

-------
                                                          OFF-SITE ASSESSMENT WORKSHOP, 1989: 1-5
Quantification of Lake  Sedimentation Rates
Utilizing Radioisotopes  Present  in the
Environment
S.C. Mclntyre
J.W. Naney
J.R. McHenry
U.S. Department of Agriculture, Agricultural Research Service,
Water Quality and Watershed Research Laboratory, Ditrant,
Oklahoma 74702
                                         ABSTRACT

            Sediment is the greatest pollutant by quantity to streams and lakes. Knowledge of sedimentation rates is
            frnpcTn to undJstandin'g aquatic processes, and for naturai resource planning^
            taining measurements of sedimentation rates, however, is not easily accomplished. Dir
            o  edimentation rates are possibie only if sediment deposition surveys have been made>^,
            °s often not the case. In the absence of periodic surveys sedimentat,on rates can  be eslma ted using
            radioactive elements present in the environment. Cesium-137, a fallout product of nuc ear testmg that t.ght-
            y adsorbs onto fine soil particles, can be used to trace soil as it is redistributed by eros,on and deposited as
            sediment. Measurement of cesium-137 activity permits estimates of sedimentation rates «nce 1954.
            Mother radioisotope, Lead-210, a decay product of radon, can be used to estimate sedimentation rates
            ™XT.pasMOO years. Results of radioisotope studies have proven useful in lake restoration planning at
            Reelfoot Lake  Tennessee. Estimates of sedimentation rates using radioactive elements can be obtamed
            relatively quickly and can provide information about lakes where historical data are lacking.
                 Introduction

  Sediment is the greatest pollutant in lakes and reser-
  voirs (Robinson, 1971). Its accumulation  in lakes
  reduces storage capacity for hydroelectric genera-
  tion, irrigation, and domestic water supplies. Sedi-
  ment also alters the aquatic environment, thereby
  reducing usefulness for wildlife habitat and recrea-
  tion. Management  of lakes and reservoirs requires a
  knowledge of sediment deposition rates. Historically,
  the only way to measure sediment deposition rates
  was to use standard survey procedures and survey
  the bottom of the  water body at least twice several
  years apart  (Task  Comm.  1970). In 1971, Krishnas-
  wamy and coworkers investigated four radionuclides
  to  determine sediment  deposition.  Of  the  four
   radionuclides, 137Cs and 210Pb have been the ones
most often used (Bobbins and Edgington, 1975; Pen-
ningtonetal. 1976; Von Guntenetal. 1987).
  137Cs, a product of atmospheric nuclear weapons
testing, is to a great extent removed from the air by
precipitation and deposited on the earth's surface
(Mitchell et al.  1983). 137Cs is tightly adsorbed on to
soil particles, and as the particles become deposited
as sediment changes in 137Cs  radioactivity makes it
possible to estimate time of sediment deposition.
  210Pb  is a  naturally occurring  isotope  of  the
uranium  series and  a  decay  product of radon
(Turekian et al. 1977). Trace amounts of radon are
found in most soils, but it also escapes into the atmos-
phere. As radon decays in the atmosphere, its decay
product 210Pb is deposited on the earth's surface. The
deposited 210Pb, known as unsupported 210Pb, is ad-
sorbed onto particles suspended in the water of lakes.
The particles subsequently settle to the bottom and

-------
S. C. MCINTYRE, J. W. NANEY, AND J. R. MCHENRY
become buried as sediment deposition accumulates.
The decay of 210Pb can be used to estimate time of
sediment deposition from about the last 100 years;
while 137Cs can only be used to date sediments for
about the past 30 years, since it did not appear in
measurable quantities until 1954.
  These two  radionuclides  provide  independent
methods of determining sediment deposition rates.
Dating of sediment deposits using radionuclides re-
quires that the  sediment once deposited not be dis-
turbed. After the radionuclides contact the surface of
the earth, there is not a long  delay before they are
deposited in sediment, and the radionuclides do not
move appreciably after deposition. The purpose of
this report is to describe how  137Cs and 210Pb were
used to quantify sediment deposition in a lake basin
where no historical survey data were available.
                Study Area

The lake basin investigated was Buck Basin (Mel ntyre
and McHenry, 1986). Buck Basin is the middle basin
of three that make up Reelfoot Lake (Fig. 1). Reelfoot
Lake began as an abandoned meander of the Missis-
sippi River. It is located about 5 km east of the present
river course, in the northwestern corner of Tennessee.
The lake increased  in size and depth due to ground
subsidence during the New Madrid  earthquakes of
1811 and 1812. The water level was controlled after a
levee and spillway were built  across the  natural
drainage way in 1917 (Smith and Pitts, 1982). The lake
has an open water surface area of about 4,100 ha, of
which 774 ha are located in Buck Basin.
   Buck Basin is surrounded by cypress swamps,
and a large number of cypress trees are standing in
the open water areas  of the basin. Many trees have
either died and fallen, or were  cut down,  leaving
numerous stumps and submerged logs in the basin.
Beyond  the swamps to the west  and north are
floodplain fields; to the east are bluff hills.
   Water flows into Buck Basin from floodplain fields
to the north through the Bayou du Chien. The Bayou
du Chien flows through Upper Blue Basin before flow-
ing into Buck Basin. Water flows into Buck Basin from
the steep hills to the east primarily by way of Reelfoot
Creek and 18 intermittent streams.
   Runoff water has brought noticeable amounts of
sediment to the  basin.  Contributing factors to the
sediment load are the  high average annual precipita-
tion (124 cm), easily eroded loess soils of the hills,
and the cultivation  of about half the hilly area and
most of the floodplain area of the watershed.
                             TENNESSEE
Figure 1.—Map of Reelfoot Lake, Tennessee showing loca-
tion of Buck Basin.
        Methods and Materials

Nine survey lines (ranges), running west to east, were
established across Buck Basin during May 1983. The
ranges were spaced approximately 300 m apart, with
sampling sites located about 400 m apart along each
range (Fig. 2). The gridlike pattern formed by the 33
sampling sites made it possible to measure differen-
ces in sediment accumulation throughout the basin.
   Sediment core samples were collected from a pon-
toon boat utilizing a plastic pipe 1.2 m in length and 76
mm in diameter. The pipe was pushed manually into
the sediment and retrieved with the assistance of a
winch. Water  depth  at  each  sample  site  was
measured  when core samples were collected. Five
cores were collected at 15  of the sample sites. The
five cores  from each site were sectioned by 10-cm
depth increments, and the sections were composited
by depth into plastic bags. Eight cores were collected
from the other 18 sample sites, and the cores were
sectioned  and composited by 5-cm  depth incre-
ments. Cores averaged 92 cm in length, ranging from
55 to 120 cm. In all, 406 composited samples  were
collected.
   In the laboratory, the samples were dried at 50°C
and passed through a 6 mm mesh screen in prepara-
tion  for 137Cs  analysis.  Approximately 1000  g of
screened sample were used to determine 137Cs ac-

-------
                                                            OFF-SITE ASSESSMENT WORKSHOP, 1989: 1-5
Figure 2.—Sediment sampling sites in Buck Basin, Reelfoot
Lake, Tennessee.

tivity levels (McHenry et al. 1980). Analysis for 137Cs
was conducted using a multi-channel analyzer with a
germanium lithium-drifted, solid state crystal detec-
tor.
   A 100  g subsample was taken from selected
samples at  site  103 and  ground with  mortar and
pestle. The ground samples were placed in 100-cm
aluminum cans, and the cans  were  hermetically
sealed. Analysis for 210Pb was  conducted using a
multi-channel analyzer with a low energy germanium
detector (Cutshall et al. 1983). Calculations of sedi-
ment deposition rates  from 210Pb and 137Cs were
made using the midpoint of each sediment layer.
   Sediment particle-size distribution was determined
by sedigraph analysis (Schiebe et al, 1983). Selected
subsamples were  pulverized by mortar and pestle
 before analysis. The sedigraph method measures the
 suspended sediment concentration in a settling tube
 using attenuation of a low energy x-ray beam.
                                                 as indicated by the increase in 137Cs activity, which
                                                 corresponded to the increase in fallout that occurred
                                                 in 1959 (Larsen, 1985). A median date of 1954 was
                                                 determined for the 25-30 cm layer using the    Pb
                                                 method  The five-year difference in dates between
                                                 using 137Cs and 21°Pb for the 25-30 cm layer is small,
                                                 considering the amount of possible variability.
                                                    Deeper sediment deposits at site 103  were dated
                                                  using the 210Pb method. 210Pb indicated a date of
                                                  1940 at a depth of 45 cm into the sediment core and a
                                                  date of 1930 at a depth of 70 cm. A date of 1880 was
                                                  found at a depth of 85 cm.
                                                    Sedimentation rates for the dated time periods at
                                                  site 103 were determined (Table 1). The dated time
                                                  periods correspond fairly closely to the major chan-
                                                  ges in agriculture of the region.

                                                   Table  1.—Time periods and sediment deposition at
                                                   site 103 in lower Buck Basin, Reelfoot Lake, Tennes-
                                                   see determined from 137Cs and 210Pb.
         Results and Discussion

 Sedimentation Rate
 From the 137Cs and 210Pb date, we found that 1963
 was the median date of sediment contained within the
 15-20 cm layer at site 103 (Fig. 3). The 25-30 cm layer
 at site 103 included sediments deposited during 1959,


TIME
PERIOD
1963 to 1983
1957* to 1963
1940 to 1957
1930(0 1940
1880 to 1930

SEDIMENT
DEPOSITION
(cm)
175
10.0
17.5
25.0
15.0
AVERAGE
ANNUAL
DEPOSITION
(cm/yr)
086
1.67
1.03
2.50
0.30
•1957 is average of 137Cs and 210Pb

  The 1963 to 1983 period, with an average annual
sedimentation rate of 0.86 cm/yr, was a time when
soybeans became the major cash crop (Smith and
Pitts, 1982). A rapid increase in the amount of land
planted to soybeans occurred with the conversion of
pasture land on the hillsides to soybean fields.
  The time period 1940 to 1963 included two dated
periods, but represented only one from an agricultural
standpoint. The average annual sedimentation rate
was 1.2 cm/yr. Three factors that contributed to the
sedimentation from 1940 to 1963 were channelization
of streams (including  Reelfoot Creek), a steady in-
crease in the amount of land planted in soybeans after
World War II, and a general increase in farm produc-
tion during the 1940s, related to World War II.
   The time period 1930 to 1940  had the highest
sedimentation rate, with an average annual rate of 2.5
cm/yr. This  period  had the largest amount of land
being farmed up until the late 1970s.
   The period 1880 to 1930 had a sedimentation rate
of 0.3 cm/yr, and extended from the time timber was
first cut through agricultural expansion.
    137Cs data from the 33  sample sites were analyzed
to  determine sedimentation rates since  1959. Sedi-
 ment layers containing 1959 deposition ranged from
 the 5-10 cm layer at locations 204 and 305, on the east

-------
S. C. MCINTYRE, J. W. NANEY, AND J. R. MCHENRY
              I37r   (mBq/g)
              10      20     30
 2IOpb   (unsupported)   (mBq/g)
10     20     30    40     50     60
    90 L

 Figure 3.—Distribution of 137Cs and 210Pb in the sediment profile from site 103 in lower Buck Basin, Reelfoot Lake, Tennessee.
 side of the basin, to the 40-50 cm layer at locations
 401 and 601, on the west side. Sediment deposition
 was greatest on the west side of Buck Basin, where
 the Bayou du Chien flows into the basin (Fig. 4). The
 overall average sediment deposition from  1959 to
 1983 was 21.8 cm, which was an average annual
 deposition rate of 0.9 cm/yr.
      Sediment Type and Quantity

  Sediment deposits since 1959 were analyzed for par-
  ticle-size distribution and bulk density. The sediment
  was found to be predominantly a clay sediment, with
  an average particle-size distribution of 5 percent
  sand,  31  percent silt, and 64 percent clay. The
  average wet bulk density was 1.18  g/cm , and the
  average dry bulk density was 0.35 g/cm3 . Using the
  average dry bulk density, the amount  of sediment
deposited in the basin since 1959 was 590,562 MT,
with an  average  of  24,607  MT/yr. The  sediment
delivered to the basin was 0.5 MT/ha of watershed per
year, and 0.85 MT/ha/yr if only fields were considered.
A watershed area of 50,703 ha was used to determine
the first delivery rate; this included the upland hills
and flood plains, but did not include the 7,330 ha of
wetlands immediately surrounding much of the lake.
   Buck Basin is a secondary receiver of sediment
from  its  two primary tributaries, so the sediment
delivery  rate does not generally  reflect  the large
amounts of sediment reaching the lake and its wet-
lands. Noticeable amounts of erosion have occurred
from  plowed hillside fields, and further conservation
 measures are needed to reduce soil erosion.
   Combining data on sediment deposition with data
 on water depths, an estimate can made of how much
 longer the basin will remain usable for fishing and
 other recreational  activities. Water depths  at sedi-

-------
                                                                       OFF-SITE ASSESSMENT WORKSHOP, 1989: 1-5
SAMPLING POINT
 Figure  4.—Bars  represent the centimeters of  sediment
 deposited trom 1959 to 1983 at each sample location In Buck
 Basin, Reelfoot Lake, Tennessee.


 ment sampling sites ranged from 1.4 m at sites 201,
 503, and 504, to 2.4 m at site 203. The average water
 depth of the basin was 1.8 m. If sediment continues to
 be deposited at the rate of 0.9 cm/yr, the basin will be-
 come too shallow, at 60 cm, for most uses in  about
 130 years. The basin will fill with sediment in about 200
 years.
                   Conclusions

  137Cs and  210Pb  made  it  possible to  estimate
  sedimentation  rates not easily obtained otherwise.
  We predict that Buck Basin will fill with sediment in
  about 200 years if the sedimentation rate remains the
  same.
                                                           ACKNOWLEDGMENT: The authors wish to thank the Ten-
                                                           nessee Department of Health and Environment, Division of Water
                                                           Management, and Division of Laboratory Services for assisting
                                                           with sample collection. We wish to thank Mr. Greg Denton, Ten-
                                                           nessee Division of Water Management, for his assistance in coor-
                                                           dinating sample collection.
                  References

Cutshall, N.H., I. L Larsen, and C.R. Olsen. 1983. Direct analysis of
   Pb-210  in sediment samples: self-absorption corrections.
   Nuclear Instr. Methods 206:309-12.
Krishnaswamy, S.D. Lai, J.M.  Martin, and M. Meybeck.  1971.
   Geochronology of lake  sediment. Earth Planet. Sci. Letter
   11'407-14.
Larsen R.J. 1985. Worldwide deposition of Sr-90 through 1983.
   Rep. EML-444. Environ. Measurements Lab., U.S. Dep. Energy,
   New York.                                    _      .
McHenry,  J.R., J.C. Ritchie, and C.M. Cooper.  1980. Rates of
   recent sedimentation in Lake Pepin. Water Resour. Bull. 6:
   1049-56
Mclntyre S C. and J.R. McHenry. 1986. Sedimentation in Buck
   Basin of Reelfoot Lake. Pages 51-62 in Conf. XVII, Int. Erosion
   Control Ass. Dallas, Texas.
 Mitchell, J.K., S. Mostaghimi, D.C.  Freeny, and J.R.  McHenry.
   1983. Sediment deposition estimation from cesium-137 meas-
   urements. Water Resour. Bull. 19(4):549-55.
 Pennington, W., R.S. Cambray, J.D. Eakins, and D.D.  Harkness.
    1976. Radionuclide dating of the recent sediments of Blelham
   Tarn. Freshw. Biology 6:317-31.
 Bobbins, J.A. and D.N. Edgington. 1975. Determination of recent
    sedimentation rates in  Lake Michigan using Pb-210 and Cs-
    137. Geochimi. Cosmo. Acta 39:285-304.
 Robinson, A.R. 1971. Sediment. J. Soil Water Cons. 26:61 -62.
 Schiebe,F.R., N.H.Welch, and L.R. Cooper. 1983. Measurement of
    fine silt and clay size distributions. Trans. Am. Soc. Agric.  Eng.
    26:491-94.
 Smith, W.L and T.D. Pitts.  1982. Reelfoot Lake: A summary report.
    Dep. Biol. Sci. Univ. Tennessee, Martin.
 Task Committee.  1970.  Sediment measurement techniques:
    reservoir deposits J.Hydrol. Div. Am. Soc. Civil. Eng. 96:2,417-2,
    446.
 Turekian, K.K., Y. Nozaki, and  L.K. Benninger. 1977. Geochemistry
    of atmospheric radon  and radon products. Ann.  Rev. Earth
    Planet. Sci. 5:227-55.
  Von Gunten, H.R., M. Sturm,  H.N.  Erten, E. Rossler, and F. Weg-
    muller. 1987. Sedimentation rates in the central Lake  Con-
    stance determined with Pb-210 and Cs-137. Schweiz 2. Hydrol.
    49(3): 275-83.

-------
                                                         OFF-SITE ASSESSMENT WORKSHOP, 1989: 7-12
Airphoto  Inventories  for Pinpointing
Nonpoint Sources
 Frank R. Perchalski
 Maps and Surveys Department, Tennessee Valley Authority,
 200 Haney Building, Chattanoga, Tennessee 37402-2801
                                        ABSTRACT










            drainage to sinkholes and other limestone subsidence features.
                Introduction

  Nonpoint water pollution sources present new chal-
  lenges to the water resource manager. Until recently,
  environmental problems were largely site specific.
  Management efforts concentrated on individual sour-
  ces of pollution, making it appropriate to rely on tradi-
  tional approaches for collecting detailed field data
  over limited areas. From a water quality standpoint,
  the data situation was acceptable, as long as con-
  cerns remained limited to point sources of pollution.
     One hallmark of watershed water quality concerns,
  however, is complexity. This stems from the transport
  and interaction mechanisms possible between scat-
  tered sources and off-site effects.  These present a
   confusing array of data collection possibilities, each
   with its own combination of accuracy, reliability, and
   cost.  For many of these complex circumstances,
   aerial pohotographs are a valuable supplementary
   data source.
     Comprehensive  watershed  management intro-
   duces new spatial and temporal dimensions, with as-
   sociated  complexities and costs.  Water resource
   managers not only  face difficulties  of  pollution
transport, but now they must also deal with intermit-
tent sources spread all over a watershed.
               Background

In the search for alternatives to traditional data, a use-
ful source appeared to be county-level agricultural
summaries. Unfortunately, county boundaries usual-
ly bear little resemblance to watershed boundaries.
County summaries also indicate very little about the
distribution of potential sources, or about drainage
connections to the receiving waters. Methods were
needed to synthesize and integrate watershed data
about animal wastes, soil erosion, and chemical sour-
ces to their off-site impacts.
   These spatial and temporal aspects require new
and innovative approaches, beginning with data col-
 lection.   Data  from  point source sampling and
 monitoring efforts are not easily extrapolated to the
 uppermost reaches of a watershed.   Knowing that
 there is a water quality problem at a particular location
 leaves  the  investigator  with  nearly  everything
 upstream as a possible source. Today's problems re-
 quire more spatially complete watershed information.

-------
 F. R. PERCHALSKI
   By 1984, prototype nonpoint inventories in Ten-
nessee and Alabama  suggested that full watershed
inventories might be possible at low unit costs, using
aerial photographic methods (Tenn. Valley Author.
1987). Continued applications of the approach indi-
cate that the increased spatial and temporal com-
plexities of nonpoint source  problems are more
manageable using the information potential of aerial
photographs (Perchalskietal. 1988). Thus far, tens of
thousands of  square  kilometers have been inven-
toried in nine states.
   Aerial photographic methods do not replace site-
specific conventional  methods of data  collection.
There is always the need for soil, water and biological
samples collected by direct contact in the field.
Nevertheless,  aerial photographs can supplement
such data with information needed over large areas.
This supplementary information can include topo-
graphic or geographic features, drainage patterns
and other hydrologic  features, land uses and land
covers, and historic or seasonal variations.
   Airphotos can be used to inventory and  charac-
terize watershed nonpoint sources such  as  eroding
fields,  livestock  operations, failing septic systems,
construction sites, industrial and urban runoff, mined
lands,  and areas of agricultural chemical applica-
tions.  This watershedwide information is useful  for
ranking subwatersheds for remedial attention, and for
targeting individual sources for direct action. The
spatially complete picture of entire watersheds, made
possible by the use of aerial photographs,  also allows
the integration of dispersed sources and potential off-
site effects through detailed surface drainage map-
ping.  This   integrated   watershed   information
facilitates the planning of sampling and  monitoring
networks for documenting off-site impacts.
   The Airphoto Inventory Method

Although projects may have unique requirements, six
basic steps are common to most airphoto projects.
They include:  (1) project planning, (2) material ac-
quisition, (3) information extraction, (4) field checks,
(5) data transfer, and  (6) reporting.  If properly fol-
lowed, these steps can yield economical data for
watershed management that cannot be obtained by
other methods.
  Costs for  this six-step aerial nonpoint  source
method range from 20 to 70 cents per hectare (about
10 to 30 cents per acre).  Variables affecting costs in-
clude watershed size, intensity  of agricultural land
use,  drainage pattern  complexity, topography, and
final  products.  Watersheds of 400,000 hectares (1
million acres) can be inventoried for about 50 cents
per hectare (20 cents per acre).  This cost includes
acquisition of newairphotos, production of subwater-
shed  summaries on computer diskettes, and  a
hardcopy atlas.
   Levels of effort for each step in the method vary
with project characteristics. Table 1 summarizes typi-
cal ranges for each  step in a watershed nonpoint
source inventory.

Table 1.—Typical levels  of effort in the airphoto in-
ventory method for watershed nonpoint source
inventories.
METHOD STEP
1.
2.
3.
4.
5.
6.
Planning
Material Acquisition
Information Extraction
Field Checks
Data Transfer
Report and Graphics
EFFORT RANGE
5 to 1 0 percent
1 0 to 15 percent
30 to 40 percent
5 to 1 0 percent
30 to 40 percent
5 to 10 percent
       Step 1 -Project Planning

Success is  measured  by how well the airphoto-
derived information meets watershed management
needs. This starts with a careful definition of objec-
tives and time and cost constraints. These establish
the boundaries for defining the information require-
ments, and ultimately, the feasibility of the airphoto
approach.
   The airphoto analyst incorporates the user's infor-
mation needs into  an inventory  plan.  This plan
defines the classification scheme (i.e., both number
of  categories  and  sizes  of features), airphoto
specifications (e.g.,  date,  scale,  and film  type),
deliverable products, schedule, and costs.
   The user  defines important  land features and
physical  characteristics.  The analyst then  selects
subsets from appropriate, usually standardized, clas-
sification schemes.  For example, land use activities
might consist of selected categories from a detailed
land use and  land cover classification scheme, such
as that proposed  by the U.S.  Geological Survey for
use with  remote sensing imagery (Anderson et al.
1976).
   Classification schemes  define  the classification
resolution or  the number of categories  that  can  be
consistently identified at each level of detail.  They
also define the mapping unit resolution  or the mini-
mum  polygon  size  for various  scales  of aerial
photographs. The classification resolution of the land
use and land  cover classification scheme published
by the U.S.  Geological Survey  has nine Level  I
categories for use with satellite  imagery.  Level  III,
however, for  use  with  medium-scale aerial  photo-

-------
                                                               OFF-SITE ASSESSMENT WORKSHOP, 1989: 7-12
graphs  has several hundred categories.   Mapping
unit resolutions using satellite data are measured in
hectares.  For medium-scale aerial photographs, they
are measured in fractions of an hectare or tens of
square meters (hundreds of square feet).
   Nonpoint source inventories use a classification
resolution of at least Level III. User requirements may
even necessitate subdivision of Level III categories
resulting in selective identification and delineation of
features at Levels IV and V. Mapping unit resolutions
range from minimums of 1 /4 to 1 hectare (about 112 to
2 acres).
   Nonpoint source inventories typically have tem-
poral restraints that require the acquisition of new
aerial photographs very early in the spring of the in-
ventory year.  It would be unusual to find existing air-
 photos of a study watershed that had been acquired
 during the prior spring that met the proper specifica-
 tions.  The use of most existing airphotos would
 necessitate a significant tradeoff in the accuracy of
 the resulting data.   When documenting historical
 changes in a watershed, the limitations imposed by
 older airphotos must be clearly understood.
    Requirements for new airphoto acquisition will typi-
  cally specify stereoscopic coverage at  a scale of
  1:24,000, using infrared color film.
Table 2.—General guidelines for environmental air-
photo data collection. 	.	
      Step 2 -  Material Acquisition

  If a project requires  existing coverage, it may be
  necessary to check a number of possible sources.
  These will include federal, state, and local agencies.
  Private aerial survey firms in the project vicinity should
  also be queried for information they might have about
  past projects  in the area of  interest. For coverage
  before the 1950s, the National Archives is also a good
  source. Appendix II of ASTM  Special Technical Publi-
  cation 967  (Johnson  et al.  1988) contains a com-
  prehensive list of airphoto information sources.
     Airphoto sources often provide reproduction ser-
  vices. Standard products are  black-and-white or color
  contact prints, or transparencies and enlargements.
  Contact prints are one-to-one reproductions, usually
  from original negatives, made on film or paper that is
   10 inches square, with an image area 9 inches square.
   Black-and-white contact prints cost approximately $5
   each. Positive transparencies may also be obtained,
   this being most common for infrared color and some
   natural color airphotos.   Costs will be noticeably
   higher for color reproductions, ranging from $10 to
   $20 each.
     Acquisition of new aerial  photographs will require
   contracting with an aerial survey firm. Table 2 shows
   some general guidelines that should serve as a start-
CONSIDERATION
                     GUIDELINE
Flight-line orientation ..
Flight Height.
Camera	
Camera Focal Length.
Camera Calibration...
 Image Format
 Film Type	
 Forward Overlap..
 Sidelap   	
 Tip, Tilt, and Crab.

 Acquisition Time ..
 Clouds and Haze
 Other	
 Centered along plotted flight lines
 (compass east-west or compass
 north-south) for stereoscopic
 coverage of entire watershed
 12,000 feet above average ground
 elevation
 Wild RC-10 or equivalent
 Nominal 6 inches
 To U.S. Geological Survey
 specifications within the past 4
 years
 9 inches by 9 inches
,  Infrared color processed to
  positive transparencies
.  60 percent
.  30 percent
.  Maximum tip and tilt 3 degrees;
  maximum crab  10 degrees
  Between 10 am and 2 pm local
  time
.  None
.  Exposures to be numbered by
  contractor; contractor to prepare
  index by marking exposure centers
  on flight maps	
 ing point for discussing new photo acquisition with a
 contractor. Variations from standard mapping quality
 aerial photographs should be avoided,  since they
 may result in  excessive analysis and data  transfer
 costs  because of  inconvenient  image  formats.
 Several cost estimates should  be obtained,  since
 costs will vary depending upon proximity of the con-
 tractor to the site, type of photo aircraft, contractor
 workload, and other factors.
    There are fixed costs, which apply to just a  few
 photographs or to many, and variable costs, which
 decrease significantly for large projects.  New air-
 photo acquisition  costs are typically based  on crew
 and aircraft mobilization, transit time to and from a
 project site, and time online (i.e., actual time acquiring
 airphotos). These will normally cost several  hundred
 dollars for mobilization, about $100 per hour in transit,
 and several hundred dollars per hour while actually
  making the photographs. Processing of film materials
  into usable positive  images is usually included in the
  online costs.
     Step 3 - Information Extraction

  Nearly all airphoto applications  use subsets of two
  primary classification systems:  (1) land uses or land
  covers, and (2) landforms.

-------
F. R. PERCHALSKI
   Land covers are natural-appearing areas, such as
forestlands, lakes, and deserts.  Land uses, on the
other hand, are areas and features with which a use
can be  associated, such as agricultural fields, golf
courses, and schools.  With airphotos alone, some
land uses are difficult to distinguish from surrounding
land covers.   For example, forestlands within park
boundaries usually cannot be distinguished from sur-
rounding forestland.  Those areas are classified as a
land cover (e.g.,  forestland), unless supplementary
information establishes a park boundary.  Then the
area  is  classified as  a land  use (e.g,  forested
parkland).
   Most other visible  characteristics are classified as
landforms or associated features. Classification sub-
sets result from  specialties, such as photogeology
and photohydrology.  Thus, flat-lying massive lime-
stones are a primary landform class, and sinkholes,
interrupted surface drainage, and clay-rich residual
soils are associated features.
   Land uses, land covers, and  landforms  are iden-
tified by their particular characteristics, Their features
exhibit distinctive combinations of the following char-
acteristics: shape, size, shadow, tone or color, tex-
ture, site or surroundings, and topographic location.
Landforms  are  identified  by  combinations  of
topographic expression, drainage pattern and tex-
ture, erosion features, tone or color,  and land use or
land cover.
   Three levels of expertise are distinguished in the
process  of   extracting  information   from  aerial
photographs (Lueder, 1969). They are photo reading,
photo analysis, and photo interpretation.
   The simplest is photo reading. At this level, a per-
son uses basic feature characteristics to distinguish
between major land uses and land covers.  As an ex-
ample, the novice can usually distinguish forests from
plowed  fields, and farms from residential neighbor-
hoods.  This level of expertise can be compared with
the ability to  read a detailed map.  For inventorying
watershed features,  however,  it is  only marginally
useful.
   Airphoto analysis  is the next, more difficult infor-
mation  extraction  level.   Complex  classification
schemes are used, usually with the  U.S. Geological
Survey Level III as a minimum. Identification and clas-
sification of combinations of features begin at this
level.  To make effective use of subtle feature charac-
teristics in airphoto analysis, however, requires more
than the monoscopic viewing of airphotos, which is
sufficient at the airphoto reading level.  Stereoscopic
viewing of the airphotos is required, both for the clues
offered by topographic characteristics, and for the in-
creased accuracy and reliability provided  by simul-
taneously viewing subtle features from two different
perspective points.
   This   necessitates  using  stereoscopic  aerial
photographs and a convenient means for stereos-
copically viewing the airphotos.   Specifications for
new airphoto acquisition require a  minimum of 50
percent overlap of consecutive airphotos along  a
flightline, thus ensuring the two perspective views of
the same scene for stereoscopic viewing. Viewing in-
struments  can  range from  simple  pocket  stereo-
scopes costing less than $100 to mirror stereoscopes
with binocular magnifiers, costing more than $1,000.
   All photo analysis is performed  in basically the
same way, but important distinctions are made at the
photo interpretation level.  At that level, the training
and  background of the person come into play for
judging  the significance of what is identified at the
analysis level.
   Interpretation differences can be illustrated by the
manner  in which various  disciplines  draw  con-
clusions from the same analysis  result. For example,
a  geotechnical  engineer,  a soil scientist,  and  a
geologist  might each analyze  the  same  landform
characteristics  and  arrive  at  similar conclusions
about the  location,  extent,  and physical charac-
teristics of a  sandy soil on  a floodplain;  but  each
would infer different performance characteristics at
the interpretation stage. The geotechnical engineer
might infer a good borrow material source for road-
way construction because of the dry, sandy soil; the
soil scientist might infer poor crop productivity be-
cause of the droughty nature of the sandy soil; and
the geologist might infer high silica content of the
sand caused by upstream source materials.  Each
discipline sees the same sandy soil, but each assigns
different significance to the material.
   At each information extraction level, experience in
using airphotos is a crucial  factor,  and a poor ap-
preciation of this leads to problems.  As an example,
an expert in a particular discipline may read an ac-
count of the successful use  of airphotos, and  if the
person  had  some limited  past exposure  to the
method through formal instruction or other career ac-
tivities, he or she might attempt an application.  With
no analysis experience,  however, the results may be
only marginally useful.  As a  result, this "casual user"
may dismiss the entire approach as unworkable.
   Nonpoint  source  inventories  are  conducted
primarily at the airphoto analysis level. Animal waste
sites are located and identified. Agricultural fields are
located,  cropping and  conservation practices are
determined, active gullies are noted, and slopes and
slope lengths are estimated.  The area of each field is
also measured.  This airphoto-derived information,
                                                 10

-------
                                                              OFF-SITE ASSESSMENT WORKSHOP, 1989: 7-12
together with soils information and other factors from
Soil Conservation Service district conservationists, is
used to calculate the Universal Soil Loss  Equation
(USLE) for each field. Recommended chemical ap-
plication rates  are also calculated for  each field.
Detailed drainage is mapped from each animal waste
site and from each field.
   Obviously, not everyone will possess the same de-
gree of competence in  airphoto analysis and inter-
pretation.  This is a critical factor in determining the
levels of accuracy and  reliability to be attached to
results (Lueder, 1959).
          Step 4-Field Checks

 Field work associated with the airphoto approach falls
 into two categories. The first provides inexperienced
 analysts on-site familiarization with categories  of a
 classification scheme.  The second serves to verify
 the accuracy and reliability of the analyses and inter-
 pretations.
    Excessive field time should not be used in develop-
 ing classification schemes.  Overviews of the  area
 should be obtained from supplemental sources in the
 project planning stage. Inexperienced analysts, how-
 ever, may require more pre-analysis field time to build
 confidence  with the study areas and classification
 schemes.
    Verification field checks are selective, in that entire
 delineated features do not require field examination.
 Only enough of an individual feature need be ex-
 amined to enable the analyst to feel confident that the
  remainder  of  the feature's  area falls within the
  delineated category.  Likewise, not every similar-ap-
  pearing feature needs to be  examined.  Field  effort
  should only be used to confirm that the classifications
  are correct  or to resolve questionable decisions.
    Based on the  selective field checks, it may be
  necessary to correct or revise some of the inventory
  results.  The airphoto analyst takes  part in the field
  verifications and performs the necessary corrections
  and revisions.  This increases confidence and enhan-
  ces classification skills.
     Field verification should not be a substitute for poor
  airphoto analysis and interpretation. Used inthatway,
  field work can become very costly and time consum-
  ing.
     From the Glossary of Geology (Am. Geolog. Inst,
  1972), "ground truth... is a misleading term, since it im-
  plies that the truth may be found only on the ground;
  the whole  truth is  preferred."  The combination of
  ground and aerial data, supplemented  with existing
  information, should result in the  best available ap-
  proach to the "whole truth."
         Step 5 - Data Transfer

This step begins as soon as there are annotated air-
photos covering a significantly large, continuous area
for a  dedicated transfer effort.  This often occurs
before the information extraction step is completed.
Care must be taken, however, that data transferred to
the base maps will not require excessive corrections
or revisions as a result of subsequent field checks.
   Final map graphics consist of annotated clear plas-
tic overlays attached to (1) the original airphotos used
in the analysis step,  (2) 7.5 minute topographic quad-
rangles, or (3)  enlarged high-altitude aerial photo-
graphs. For small  sites in relatively flat or gently rolling
terrain, the analysis airphotos may serve as adequate
base maps. For projects covering the area of several
airphotos, however, this becomes an inconvenient ar-
rangement.
   Annotations from multiple airphotos must be trans-
ferred to a common base map. This is usually the 7.5
 minute topographic series of the National Mapping
 Program. I n many situations, these provide adequate
 planimetric and topographic references for the trans-
 fer of annotations from the airphotos.  Transference
 works best when airphoto and map scales are similar.
    The area of a single,  7.5  minute quadrangle is
 covered by approximately 10 airphoto stereopairs, at
 1-.24,000 scale.  The quadrangle format provides a
 good map base for multiple airphotos. Multiple quad-
 rangles can also  be composited for larger areas.
    The value of the airphoto information depends on
 the accuracy of the transfer of annotations to the base
 maps. Several manual techniques may be used.  The
 simplest is a visual match of  airphoto and map fea-
 tures. This works where accuracy is not the primary
 consideration, and where there is sufficient common
 airphoto and line map detail. This is the case for many
  nonpoint source  inventories  and  some  photo-
  hydrologic features.
    When  locational accuracy  is  required, optical
  projection devices are used to transfer data from the
  airphotos to the base map. The additional accuracy,
  however, can  more than double data transfer costs.
    A faster  manual technique is to  use high altitude
  aerial photographs.   Coverage  from the National
  High-Altitude  Photography (NHAP) Program can be
  enlarged to closely match the  scale of 7.5 minute
  quadrangles. Alternate exposures from the 1 -.80,000-
  scale black-and-white NHAP negatives are centered
   on 7.5 minute quadrangles and cover the entire area
   of these 1:24,000- scale sheets. These  provide con-
  venient image bases for the analysis results.
     The NHAP Program has been replaced by the Na-
   tional Aerial Photography Program (NAPP).  This still
   provides both  black-and-white and infrared  color
                                                   11

-------
F. R. PERCHALSKI
coverage, but now at a scale of  1:40,000.  Photo-
graphs cover one quarter of a 7.5 minute quadrangle,
with alternate airphotos centered over quarter-quad-
rangle areas.
   For many nonpoint source and hydrologic applica-
tions, visual data transfer is adequate. For most ap-
plications, high spatial accuracies at the analysis and
interpretation stages can result in unnecessary ex-
penditures of time and  money.  Some  sacrifice of
locational accuracy may  be justified for  nonpoint
source inventories,  due  to the large number of fea-
tures and the intricate drainage networks.
   For some applications, it is desirable to carry the
aerial data beyond  the map product. This may in-
volve computer storage of tabular data or digitizing
atlas  materials for use in a geographic  information
analysis system. Speed, accuracy, and cost need to
be evaluated closely if data are to be transferred and
stored in these forms.  Digitizing detailed data for
large areas is  prohibitively  expensive,  unless the
value of the resulting management decisions are
comparably high. Digitizing costs  can range from 60
to  100  percent  of the  airphoto acquisition and
analysis  costs.  The complexity  of some airphoto
products often makes digitization for entire water-
sheds impractical. For those situations,  a  multilevel
approach may be appropriate, with only critical data
or targeted areas being digitized.
    Step 6 — Report and Graphics
                Preparation

Different levels of detail and types of products can be
used to document aerial data. These should reflect
user requirements and  needs.  Several  basic com-
binations  of  materials  and formats are  possible.
These include maps and atlases, printed tabular data,
digital tabular summaries, digital graphics files, and
reports.
   Nonpoint  source inventories  require the most
complex assembly of materials. Atlases may consist
of quadrangle-centered airphoto  enlargements at
1:24,000 scale, with clear plastic overlays showing in-
dividual sources, drainage connections, and labels.
Multiple overlays are necessary for complex annota-
tions.  The overlays show the location of each field
and animal waste site, drainage  connections from
each animal waste site, subwatershed unit  boun-
daries, stream access points for livestock, eroding
streambanks, unpaved roads, and eroding road cuts.
Photohydrologic supplements can also be added,
such as springs, seeps, and infiltration areas.
  Atlases are supplemented with tabular data sum-
maries in the form of computer printouts, computer
diskettes,  or both.  The tabular summaries can in-
clude soil erosion potential by field, animal waste site
locations,  and  agricultural chemical  estimates  by
field. These data can be summarized by field and site
categories, by subwatershed,  by 7.5  minute quad-
rangle, and by county.
  For spatial analysis, digital graphic and attribute
files can be produced. Graphic products help ensure
the utility of the resulting data.  Massive amounts of
data may  need to be summarized and presented in
formats suitable for standard-size  reports.  Where
graphics are to be used as courtroom exhibits, they
must be quickly understandable and unambiguous.
                 Summary

The use of aerial photographs for locating and iden-
tifying nonpoint sources and their detailed drainage
connections is now practical forfull watersheds. Sub-
watershed summaries provide valuable  planning
tools for ranking manageable-size areas to establish
control and remedial action priorities.  Within sub-
watersheds individual nonpoint sources can be easily
targeted through the combined use of atlas sheets
and summary data.
   This innovative data collection approach is a sig-
nificant step toward realistically quantifying nonpoint
source problems and  off-site effects at costs  and
levels of detail not obtainable by other methods.
                References

American Geological Institute.   1972.   Glossary of Geology.
  Washington, DC.
Anderson, J. R., E.E. Hardy, J.T. Roach, and W.E. Witmer. 1976. A
  Land Use and Land Cover Classification System for Use with
  Remote Sensor Data. U.S. Geolog. Surv. Prof. Pap. 964.
Johnson, A.I. and C.B. Pettersson, eds. 1988.  Geotechnical Ap-
  plications of Remote Sensing and Remote Data Transmission.
  ASTM Spec. Tech. Publ. 967. Am. Soc. Test. Mater. Philadel-
  phia.
Lueder, D. R. 1959. Aerial Photographic Interpretation. McGraw-
  Hill Book Co. Inc., New York.
Perchaliski, F.  R. and J. M. Higgins. 1988. Pinpointing nonpoint
  pollution. Civil Eng. February. Am. Civil Eng., New York.
Tennessee Valley Authority. 1987. Pinpointing Nonpoint Sources.
  Impact March. TVA/ONRED/SFO-87/1.  Chattanooga, TN.
                                                  12

-------
                                                             OFF-SITE ASSESSMENT WORKSHOP, 1989: 13-19
Contaminated  Sediment Assessments
Anthony G. Kizlauskas*
Bruce Kitsuse
U.S. Environmental Protection Agency, Great Lakes National
Program Office, 230 South Dearborn Street, Chicago, Illinois
60604
                                           ABSTRACT

            The primary concerns with contaminated sediments are the effects on aquatic life and,,possibly  human
            health through food chain contamination. The International Joint Commission, which is a binatonal (United
            S±-Cana9dian) organization established under the Great Lakes Water Qua jv' g^TS'.^.^
            lished a Sediment Subcommittee to address contaminated sediment ,ssues ,n the Great Ukes. The bed.
            ment Subcommittee is currently developing  guidance for assessing the significance of *°n~^d
            sediments In the initial step, a reconnaissance survey is conducted to collect a HmitecI number of M£pU»
            on Sich to conduct basic tests to confirm whether there may be a contaminated segment problem ma
            paSar ocation. These tests include surficial sediment chemistry, analysis of the body-burdens, of chemi-
            cals in he indigenous aquatic life, and a qualitative analysis of the community structure of the benth.c mac-
             oin  rteb ate oeonmln?ty. If a po'tentia, prob.em is indicated by these tests, then a deta,,ed assessment o
            the site is made. The objectives of the detailed assessment are to more precisely determine the na ure of
             he problem, and its spatial and temporal nature by developing sediment contamination maps; and tpro-
            ™de information required to define the most appropriate remedial option should  action  be required. To
            achlevTZse objectives, a three-phased  approach is recommended: (1) determ.ne the distribution of
            physical and chemica! characterises of the sediments; (2) assess the toxicity of the segments usmg
            laboratory bioassay tests; and (3) quantitatively assess the health of the ind.genous b,olog,cal community.
                  Introduction

  This paper will present a brief perspective on how
  contaminated  sediment assessments are performed
  in the Great Lakes. The next section will present some
  background on the kinds of sediment issues that have
  arisen, and the assessment methods derived to ad-
  dress those issues.
                  Background

   The contaminated sediment issue first surfaced in the
   Great Lakes in the middle to late 1960s, when concern
   arose over the practice  of disposal of sediments
   dredged from  harbors into the  open waters of the
   Great Lakes.
   Over four million cubic meters  of sediment  are
dredged annually to maintain these harbors. Back in
the early 1960s, virtually all of the dredged sediments
were disposed of in the open waters of the lakes. The
public saw these oily, mucky sediments discoloring
the water around the disposal site.  They saw  the
resulting oil slicks,  and  became increasingly con-
cerned that this practice was not desirable. They were
concerned that their use of the Great Lakes for drink-
ing water, as well as for recreation, might be adversely
affected by the practice of open water disposal.
   As a result of this growing awareness and public
and  political  pressure,  the Chicago office  of the
Federal Water Pollution Control Authority produced a
set of  sediment pollution  guidelines to determine
when dredged material  was suitable for open water
disposal. These first Great Lakes sediment guidelines
were published in 1968 (Table 1).  They were based
                                                                        r, Suite 1400, Chicago, Illinois 60602
                                                    13

-------
A. G. KIZLAUSKAS AND B. KITSUSE
Table 1.—FWPCA Chicago August 1968	

DEGREE OF POLLUTION OF HARBOR SEDIMENTS (mg/kg DRY WT.)
                                                      Table 2.—EPA 1971.
PARAMETER
Ammonia N
COD
Total Iron
Lead
Oil & Grease
Phenol
Total Phosphorus
Sulfide
% Volatile Solids
Zinc
LIGHT
0-25
0-40,000
0-8,000
0-40
0-1,000
0-0.26
0-100
0-20
0-5%
0-90
MODERATE
25-75
40,000-120,000
8,000-13,000
40-60
1,000-2,000
0 26-0.60
100-300
20-30
5-8%
90-200
HEAVY
over 75
over 120,000
over 13,000
over 60
over 2,000
over 0.60
over 300
over 60
over 8%
over 200
upon the bulk, or total sediment concentration, of
chemicals of concern. The 1968 guidelines contained
10 parameters, including two metals, iron and lead.
   The Great Lakes guidelines formed the basis for
the national guidelines issued by the EPA Office of
Water Programs in 1971 for freshwater and saltwater
sediments (Table 2). These were popularly referred to
as the "Jensen Criteria."
   The Great  Lakes sediment guidelines  were up-
dated in 1977, based upon additional sampling data
(Table 3). The  guidelines now  included  19 para-
meters, including metals and PCBs.
   The 1977 edition of the Great Lakes guidelines has
served well to date when used as a screening tool to
When sediment concentrations of one or more of the following
pollution parameters exceed the limits expressed below, the sed-
iment will be considered polluted in all cases, and therefore, un-
acceptable for open water disposal.
SEDIMENTS IN FRESH &
MARINE WATERS
•Volatile Solids
Chemical Oxygen Demand (COD)
Total Kjeldahl Nitrogen
Oil & Grease
Mercury
Lead
Zinc
CONCENTRATION IN &
DRY WT. BASIS
6.0
5.0
010
0.15
0.0001
0.005
0.005
determine which sediments can be disposed of in the
open waters of the Great Lakes in an unconfined man-
ner, which should be confined, and which require a
more careful  examination to make a decision.  The
Great Lakes  guidelines have  always been applied
using best professional judgment to take into account
site-specific differences between dredging  projects,
including such factors as the size of the project, the
dredging frequency, and the occurrence of naturally
high background concentrations of pollutants.
   At about this same time, the Canadian environ-
mental agency, the Ontario Ministry of the Environ-
Table 3.—U.S. Environmental Protection Agency, Region V—Chicago, Illinois, April 1977

                           GUIDELINES FOR THE POLLUTION CLASSIFICATION
                                 OF GREAT LAKES HARBOR SEDIMENTS


Volatile Solids (%)
COD (mg/kg dry weight)
TKN (mg/kg dry weight)
Oil & Grease
(Hexane Solubles)
(mg/kg dry weight)
Lead (mg/kg dry weight)
Zinc (mg/kg dry weight)




Ammonia
Cyanide
Phosphorus
Iron
Nickel
Manganese
Arsenic
Cadmium
Chromium
Barium
Copper

NONPOLLUTED
<5
<40,000
< 1,000


< 1,000
<40
<90



NONPOLLUTED
<75
<0.10
<420
< 17,000
<20
<300
<3
*
<25
<20
<25
MODERATELY
POLLUTED
5-8
40,000-80,000
1 ,000-2,000


1 ,000-2,000
40-60
90-200
MODERATELY
POLLUTED
(mg/kg dry
weight)
75-200
0.10-0.25
420-650
17,000-25,000
20-50
300-500
3-8
*
25-75
20-60
25-50
HEAVILY
POLLUTED
>8
>80,000
>2,000


>2,000
>60
>200


HEAVILY
POLLUTED
>200
>0.25
>650
>25,000
>50
>500
>8
>6
>75
>60
>50
'Lower limits not established
                    POLLUTED
      Mercury         &1  mg/kg dry weight
      Total PCBs      a 10 mg/kg dry weight
 The pollutional classification of sediments with total PCB concentrations between 1 0 and 10 mg/kg dry weight will be determined on a case-by-case basis.
                                                   14

-------
                                                               OFF-SITE ASSESSMENT WORKSHOP, 1989: 13-19
ment, published its own set of bulk sediment chem-
istry guidelines for dredging project evaluation.
   When decisions needed to be made on dredging
projects along the international  border that affected
both countries, the issue was raised that the sediment
guidelines of the two countries were not compatible.
   In  1978, the United States and Canada signed a
Great Lakes Water Quality Agreement. One of the
provisions of that agreement, entitled Annex 7, called
for establishing a Dredging Subcommittee to develop
compatible guidelines between  the two countries for
making decisions on dredging projects.
    The Dredging Subcommittee published this set of
 compatible Great Lakes sediment guidelines in 1982
 (Table 4). These international guidelines were based
 on a "non-degradation" principle, whereby concentra-
 tions of pollutants in sediments were compared to the
 average surficial  sediment concentrations in the
 depositional  basins of the lake where they would be
 disposed. The guideline parameters included  PCBs,
 nine metals, and phosphorus.
    Since the  early 1980s, the contaminated sediment
 issue has taken on a broader dimension. Increasing
 attention has  focused  on the  concern that con-
 taminated  sediments may pose a hazard to the  en-
 vironment,  whether or  not  they are  slated  for
  Table 4.-
-Average concentrations (dry wt.) of surficial
                 LAKE              LAKE
               ONTARIO
  dredging. They may be disrupting the benthic ecol-
  ogy by being outright toxic; they may be causing an
  imbalance in the benthic species diversity stemming
  from nutrient over-enrichment; they may be causing
  bioaccumulation or inducing tumors in the resident
  fish.  In this scenario, one has to determine whether
  the risks posed by the contaminated sediments are
   sufficient to require that some kind of remedial action
   betaken.
      This situation has received more attention recently
   in the Great Lakes because of a program called the
   Remedial Action Plan process.
      The International Joint Commission is a  binational
   organization  responsible  for overseeing the im-
   plementation of the Great Lakes Water Quality Agree-
   ment between the United  States and Canada. The
   commission has designated 42 "Areas of Concern" in
   the Great Lakes Basin: 28 in the United States and 14
   in Canada.
      Areas of Concern are locations where pollution im-
    pacts are evident, as measured by violations of stand-
    ards, guidelines, or criteria, and by the impairment of
    such  beneficial uses as recreation, drinking water,
    biological productivity, and so forth. Based upon ex-
    isting data, virtually all of  the Areas of Concern are
    believed to be affected by contaminated sediments.
sediment constituents in the Great Lakes.
                                              ERIE
  CATEGORY I
PCBs
Mercury
Lead

CATEGORY II
                        0.077-0.0891
                        0.653 (0.07)
                        1063 (30)
                                0.074-0.252'
                                0.584 (0.08)
                                1124 (28)
                                                               LAKE
                                                              HURON
         0 009-0.0331
         0.224 (0.08)
         494  (22)
                             LAKE
                            MICHIGAN
0.00972
0.107 (0.06)
406 (19)
                    LAKE
                  SUPERIOR
0.030'
0.084 (0.07)
44*  (21)
                        3 33
                        2.53(1.3)
                        1.06
                                3.24
                                2.57 (1 1)
                                0.796
         1.14
         1.44 (0.7)
         0.96
10.55(5.3)
 0.95
 1.28(1 8)e
174
1,24 (0.6)
0.66
                          503 (44)
                         1927(105)
                          483
                          523
                                  397 (29)
                                 1777(98)
                                  537
                                  494
         324 (41)
         62" (83)
         324 (36)
         394 (47)
 225(21)
 975 (74)
 465 (62)
 245 (36)
 824 (62)
 974 (106)
 1634(51)
 954 (57)
Arsenic
Cadmium
Selenium

CATEGORY III
  M-g/g
Copper
Zinc
Chromium
Nickel

CATEGORY IV
  mg/g
Total Phosphorus* _


                                      ^^^^
        '^Thomas, H. E Braun, D L Gross and T T Dav.es "Organo Chlorine Insectiades and PCB ,n Surfiaal Sediments o. Lake Michigan (1975)"
  j'ofG L Research, VII (1) 42-50 Int'l Assoc of G L Res 1981
 'implementation Committee Report to the Water Quality Board  977 .
 "implementation Committee Report to the Water Quality Board (1978).
 scahill (1981)
 6Traversy et a/. (1975).
 'International Working Group Report (1975)
 8Cahill (personal communication)                                                                     _
 sprye and Shimp (1973) _ _ _ . __ _ - -- - ---
 1               ~                                   15

-------
A. G. KIZLAUSKAS AND B. KITSUSE
  To restore these degraded Areas of Concern, the
IJC has called for the preparation and implementation
of Remedial  Action Plans, or raps. These raps will
identify very specific steps to restore the area, and will
have a detailed timetable for tracking  progress
toward that goal.
   In 1986, as a result of the growing recognition that
contaminated sediments were an issue that ought to
be addressed  in a context broader than that of
navigational  dredging,  the  Water  Quality  Board
restructured  the  Dredging   Subcommittee  and
renamed it the Sediment Subcommittee. The Sedi-
ment Subcommittee was given a broad  mandate to
address contaminated sediment issues of all kinds.
   The  IJC felt  an  urgent need to provide those
responsible for preparing the Remedial Action Plans
for Areas of  Concern  with guidance on how to ad-
dress contaminated sediment problems. Therefore, it
charged the Sediment Subcommittee with preparing
a guidance document on the assessment of  con-
taminated sediment  problems. A draft  guidance
document was presented at a meeting of  Remedial
Action Plan Coordinators in Toledo, Ohio, November
19, 1987 (Report of Sediment Subcommittee and its
Assessment Work Group to the Water Quality Board,
"Procedures  for the Assessment  of Contaminated
Sediment Problems in the Great Lakes"). The purpose
of the guidance document is to assist the  Remedial
Action Plan authors in designing the proper studies to
assess  contaminated  sediment   problems.  This
guidance is the  most comprehensive to date on the
assessment  of contaminated  sediment problems in
freshwater systems.
   The  remainder of this paper will summarize the
Sediment Subcommittee's guidance.
     IJC Sediment Subcommittee
  Sediment Assessment Guidance

The  Sediment  Subcommittee's  guidance  sum-
marizes current thinking and practices relating to the
assessment  of  the degree  and nature of con-
taminated sediment, and its potential effects on water
quality and aquatic biota. The guidance is meant to
provide the best set of tools currently available for ad-
dressing the contaminated sediment issue. However,
decisions will be based on the circumstances unique
to each location being investigated. The guidance
urges that best professional judgment be used in ar-
riving at the kinds and numbers of tests to use in prob-
lem assessment.
   Much work remains to be done in assessing the ef-
fects of contaminated sediment and its significance in
individual Areas of Concern. Many techniques that
were laboratory  oriented have been refined and
adapted to situations common to Areas of Concern.
Othertechniques show great promise, but require fur-
ther validation and field testing. This guidance iden-
tifies a set of the best tools currently available for
assessing the significance of impacts from sediment-
bound contaminants.
  To more systematically and adequately document
the  relative significance of  contaminants  in sedi-
ments, the IJC  Sediment Subcommittee guidance
recommends the following process.
   • confirmation of problem, and

   • detailed assessment.

The evidence, if any,  that exists to document con-
taminated sediment problems is in many cases very
weak: the data are often dated; or just as often, con-
sist of one or two samples, the results of which were
never substantiated. Under these circumstances, the
guidance recommends a confirmation of the problem
before remedial actions are contemplated.
  When the confirmation of contaminated sediment
problems is  required, the guidance recommends
conducting a basic set of tests, consisting of
   • surficial sediment chemistry,

   • analysis of body  burdens of indigenous aquatic
     life, and

   • a qualitative analysis of the community structure
     of the benthic community.

This is the minimum  set of tests recommended to
determine whether or not there is a contaminated
sediment problem in a particular Area of Concern. If
any of these tests indicates a possible problem, then
the investigation is escalated to the next level of the
assessment process,  the detailed  assessment. The
following describes the tests used for confirmation of
the problem.
  First, a basic examination of sediment chemistry of
the area is performed:

   • Based on the bathymetry of the area, about 5 to
     10 samples of bottom sediments are collected
     from either the depositional zone or the area pre-
     viously thought to have been contaminated.

   • The superficial layer, about the top 5 cm of the
     sediments, is  analyzed for nutrients, metals,
     chlorinated organics, and oxygen-consuming
     contaminants.

  Because definitive, scientifically based sediment
criteria do not exist at the present time, a conserva-
                                               16

-------
                                                            OFF-SITE ASSESSMENT WORKSHOP, 1989: 13-19
tive  non-degradation type of  approach is recom-
mended. Using this approach, if concentrations of
contaminants  are  found to be higher than back-
ground concentrations, there  is  confirmation  that
there may be a  contaminated sediment  problem.
Here background is defined as the concentrations of
contaminants that exist belowthe ambrosia horizon in
the closest open-lake depositional basin. What this
essentially means is that background concentrations
are those that existed before the area was settled by
the  Europeans. These background concentrations
 are found in Table 2 in the draft guidance.
   The second set of tests to confirm a problem are
 those that evaluate  bioavailability of sediment con-
 taminants.
   Bioavailability is a concern because of potential
 concentration of  a  toxic  contaminant in the food
 chain. It is  recommended that  an assessment  of
 bioavailability be done by measuring tissue con-
 centrations of contaminants in biota at the same sam-
 pling sites used to determine surficial  chemistry.
 Either indigenous benthic invertebrates, or carp, or
 both should be sampled.
    Advantages to using these organisms for confirma-
 tion of bioaccumulation are that:
     • Both carp and benthic invertebrates ingest large
       amounts of sediment.

     • Benthic invertebrates are in  direct contact with
       the sediment and are spatially representative.

     • Carp are also  benthivores; they have the  poten-
       tial  for accumulating contaminants from sedi-
       ments, and they provide  sufficient tissue  for
       analysis.

   One disadvantage of using benthic invertebrates is
   that it may be difficult to obtain  sufficient tissue for
   analysis. Disadvantages to using  carp as an indicator
   are that carp move  around, so their contamination
   may not come from  the sediments in the immediate
   vicinity of where the fish were caught; also carp may
   be responding to contamination of the water, rather
   than of the bottom  sediments.
      Nevertheless, sampling these two organisms will
   give the best indication of the bioavailability  of sedi-
   ment-bound contaminants.
      The criteria recommended for assessing the biota
   data are the levels of persistent  organics. If they are
   above detection limits in the tissues of either the ben-
   thos or the carp, there  is reason to think that a con-
   taminated  sediment problem   exists.  Metals con-
   centrations in the biota are more difficult to assess,
   because some metals are required in certain amounts
   for healthy functioning  of the organism. Metals must
therefore be assessed  on a  case-by-case  basis,
taking into account the physiological requirements of
the species under consideration for individual metals.
   The third component of the tests for confirmation of
a sediment problem is a preliminary estimate of com-
munity  structure impairment  at the  same sites
sampled for sediment chemistry and bioaccumula-
tion To keep costs and effort in perspective during
this preliminary stage of assessment, this community
structure estimate should be only qualitative, without
detailed sampling  and  replication,  or taxonomic
detailing.
   The advantages to sampling benthic invertebrates
 are that benthic invertebrates are present in almost all
 aquatic habitats, they are relatively easy to  sample
 quantitatively,  and  they  have  comparatively  well
 known responses to pollutants and a well established
 taxonomy. Populations are relatively stable over time,
 requiring less frequent sampling. They are compara-
 tively non-mobile, and therefore, are representative of
 the area being sampled.
    The  absence of a  healthy  benthic community
 would indicate a sediment contamination problem. An
  unhealthy benthic community would be defined by

     • either the absence of clean water  organisms
      such as amphipods or mayflies, or

     • a community dominated by oligochaetes, or

     • the absence of invertebrates where habitat would
      normally be suitable for them.

  If any of the tests indicates a  potential contaminated
  sediment  problem, then the  next phase of assess-
  ment, the detailed assessment, must be performed.
     The objectives of the detailed assessment are:
     • to determine the spatial and temporal  nature of
       the contamination by developing sediment con-
       tamination maps;

     • to more precisely determine the nature of the
       problem with the sediment; and

     • to provide information required  to define the
        most appropriate remedial option.

   Spatial trends in sediment quality must be determined
   to identify and delineate those areas that are exces-
   sively contaminated with toxic chemicals, and which
   therefore require remedial action. Sediment deposi-
   tional zones accumulate and integrate toxic chemical
   inputs from nearby sources over time. Chemical data
   from  analyses of samples from such zones can be
   very helpful  in determining the history of inputs and
   spatial trends in contaminant levels.
                                                    17

-------
A. G. KIZLAUSKAS AND B. KITSUSE
   Unfortunately, traditional assessment of sediment
quality has usually involved only chemistry data. But
chemistry data alone may not provide an indication of
the degree of biological damage that may be occur-
ring.
   A three-pronged approach to detailed assessment
of sediment contamination is recommended:

   • First, the distribution of physical and chemical
     characteristics of the sediments is determined;

   • Next, the toxicity of the sediment samples is as-
     sessed in laboratory bioassays; and

   • Finally, the health of the indigenous biological
     community is assessed.

The information obtained from each of these three
components, when taken together, significantly en-
hances  understanding of the role of contaminated
sediments in an Area of Concern.
   Bulk chemistry data are needed to determine the
degree and nature of the contamination. These data
can provide information about possible  sources of
the contamination. They may also provide clues as to
why the biota in the laboratory bioassays and the in-
digenous biota behave the way they do.
   The  second component, toxicity assessment
using bioassays, can establish the relative toxicologi-
cal significance of the sediment contamination. One
possible shortcoming is that bioassays may not ac-
curately represent the  conditions to which the in-
digenous  biota are exposed, or Lheir responses to
living in the sediment at the site.
   This is where the third component comes into play.
In situ assessment of the resident biota is  needed,
both to confirm the results of laboratory assays, and
to provide a relative indication of contaminant effects
versus the complex interactions of chemical, physi-
cal, and biological factors to which the community
responds.
   Thus, although interpretation of in situ data is dif-
ficult, in the final analysis it is the ultimate measure of
impacts. For this reason, the IJC Sediment Subcom-
mittee considers  In situ assessment to  be a man-
datory part of the investigation.
   A more detailed description of the recommenda-
tions for implementing the detailed assessment ap-
proach follows.
   To  perform a detailed assessment efficiently,  it is
recommended  that it be carried out in phases. A
tiered approach  should  be taken,  as  different
methods require a varying commitment of resources.
   In Phase 1, a three-dimensional physical map of
the sediments in the study area is constructed. This
provides definition of the horizontal and vertical dis-
tribution of the sediments. It defines those areas
where more detailed sampling will be conducted in
subsequent phases of the assessment.  Possible
methods for this physical mapping include the use of
grid sampling, side-scan sonar, echosounding, and
video imaging, or combinations of these. It is recom-
mended that the greatest effort be put into sampling
the depositional basins. This  is where most con-
taminants will end up, and where the remedial actions
will usually be applied. The detailed map of the sedi-
ment will define sediment zones based on grain-size
distribution, and  will allow the definition of ap-
propriate sampling strata for Phases 2 and 3.
   Phase 2 would consist of using the mapping infor-
mation to establish sampling locations for surficial
chemistry,  tissue  concentration,  and  community
structure of the indigenous fauna. The guidance sug-
gests suitable sampling devices to use for the inves-
tigations. Sample sites should be selected randomly
within the strata, and numbers of samples should be
based upon  statistical  requirements. Community
structure assessment in this phase will be quantitative
and taxonomically detailed.
   Other  in situ assessment tools are also recom-
mended for consideration. These tests typically re-
quire more specialized expertise. Among the  other
types of in situ measures that can be used are bottom
fish histopathology, invertebrate deformities, carbon
uptake efficiency, and thymidine incorporation.
   Phase 3 consists of suitable bioassays and sedi-
ment coring.  Because  these tests are  costly, it is
recommended that sample sites be selected using
the information gathered in the Phase 2 investiga-
tions.
   The collection and analysis of sediment cores in
vertical slices are recommended to provide informa-
tion  on the vertical distribution and history of con-
tamination. This  information  will  be  useful  for
developing remedial options. The vertical distribution
of contaminants can help to establish whether active
sources still remain, whether contaminant inputs are
declining or increasing, and whether resuspension is
significant.
   The conduct of both bioaccumulation potential
and  chronic  toxicity bioassays  is recommended.
Bioassays provide a direct test of toxicity of sediment
samples. However, bioassays have some limitations
that must be kept in mind when the results are used to
make management decisions. A laboratory bioassay
is  a simplified exposure of a biological system to a
contaminant, whose objective is to predict the effects
of that contaminant in the real world. Contaminants
usually exert their effects at the biochemical level,
often  by  inhibiting  enzymes  or  interrupting
physiological  processes. In the environment,  how-
ever, effects of a grosser nature are of concern: chan-
                                                18

-------
                                                             OFF-SITE ASSESSMENT WORKSHOP, 1989: 13-19
ges in populations, communities, and ecosystems.
While effects on communities and ecosystems can
best be studied by directly looking at changes in these
systems, these methods are not predictive. They can
only be used after the fact, after contamination has al-
ready occurred. And, because of the complexity of
natural systems, the changes found are difficult to
ascribe to specific  contaminants. Thus, it may be dif-
ficult to validate the results of laboratory bioasays with
real-world situations.
   Despite their limitations, bioassays do serve an im-
portant role in  helping to answer the biological "So
what?" question. Where possible, bioassays should
be performed with sensitive, indigenous species, so
that the results can be more directly related to the
fauna at that site. To ensure a comprehensive data set,
tests using two or  three species are preferable, espe-
 cially for areas with moderate contamination that may
 be at or near the toxicity threshold for some species,
 but not for others. Bioassays  of sublethal effects
 should also be used, placing particular emphasis on
 reproductive impairment.
    Chronic  assays that assess  reproductive or life-
 cycle effects are preferable, since they are likely to be
 more sensitive than acute tests. Of the test organisms
 that  have been suggested for chronic tests, the UC
 Sediment Subcommittee recommends using

    • daphnia magna, and

    • chironomus   tentans,   hyatlela   azteca,   or
      hexagenia limbata.

  For an acute, sublethal test, the subcommittee recom-
  mends using the Microtoxtest.
    The bioavailability of the contaminants in the sedi-
  ment is also a concern. It is therefore recommended
that  organisms from the  chronic  benthic test be
analyzed after exposure to estimate bioavailability.
   In addition, for a direct test of bioavailability of the
contaminants to fish, the use of the fathead minnow
test is recommended.
   Although  the  UC  Sediment  Subcommittee
guidance  emphasizes contaminated in-place sedi-
ments, a complete interpretation of the role of sedi-
ments and  contaminant fate  will  require data on
suspended  sediment quality of point sources and
tributaries.
   In summary, the UC Sediment Subcommittee-
 recommended strategy for assessment has two data
 collection stages:
   If the data are not recent or are limited, then confir-
 mation of the problem is needed. Data collection at
 this stage would include:
    • representative   sampling  of   the  surficial
      chemistry,

    • determination of contaminant body burdens of
      the indigenous biota, and

    • a qualitative assessment of the benthic  com-
      munity structure.

    If sediment contamination  is confirmed, then a
 detailed assessment is called for. The detailed assess-
 ment stage has three basic components:
    • collection of data  on the physical and chemical
      characteristics of the sediments,

    • laboratory bioassay tests for acute and chronic
      toxicity, sublethal effects, and bioavailability, and

     • detailed examinations for the resident biota.
                                                   19

-------
                                                             OFF-SITE ASSESSMENT WORKSHOP, 1989: 21-22
Physical  Fish  Habitat Components as
Measures  of  Beneficial  Use Health
 Donald M. Martin
 Nonpoint Source Coordinator, U.S. Environmental Protection
 Agency, Region 10, 422 West Washington Street, Boise, Idaho
 83702
                                           ABSTRACT

            EPA is currently investigating the development of criteria for a physical fisheries habitat component, sub-
            strate quality for the northern Rockies ecoregion. This is a departure from traditional criteria development
            which has been basically limited to the chemical components of the water column. Also, other physical
            habitat components, such as large woody debris, riparian habitat, and so forth, which should be candidates
            for criteria development in various ecoregions. Fisheries and water quality in Idaho have suffered severely
            from the impacts of sediment from  nonpoint sources such as agriculture, logging, and mining. Although
            spawning and rearing habitats of resident and anadromous salmonids have been extensively degraded,
            such impacts are not unique to any particular state in the Northwest. All states are required by the regula-
            tions of the Clean Water Act to have and implement water quality standards, which should provide ade-
            quate protection of designated beneficial uses (i.e., fisheries) from point and nonpoint sources of pollution.
            Recent activities in Idaho indicate a new trend toward quantifying acceptable levels of impact to fish habitat
            from nonpoint source activities. The level of degradation of fish habitat (spawning, summer or winter rear-
            ing) has been used as a measure of impact on fish populations, the protected beneficial use. This approach
            may well be a  viable way of defining full protection,  as provided for in high quality waters under the an-
            tidegradation provisions of state water quality standards.
  This paper discusses the concept of using physical
  fish habitat components to measure the health of the
  designated  beneficial uses,  such as fisheries. The
  U.S. Environmental Protection Agency is evaluating
  the possibility of developing criteria for fish habitat
  parameters, such as spawning and rearing habitat for
  salmonids. These physical habitat  components, as
  well as others such as riparian habitat, large organic
  debris, and  stream structure,  may be future can-
  didates for  criteria.  These criteria would be unique
  with each different ecoregion. Those species we are
  concerned  with  in the Northwest  whose physical
  habitats  are affected  by  sedimentation  included
  Chinook salmon, sockeye/kokanee salmon, cutthroat
  trout, bull trout, and rainbow/steel head trout.
     This consideration offish habitat  has developed in
  the West because of a deficiency  in Idaho's water
  quality standards for the last eight years: the lack of
  an antidegradation policy and implementation proce-
  dures. All the states in the nation except Idaho have
  antidegradation  provisions  in  their water  quality
standards.  Idaho deleted this provision in 1980 be-
cause the state found the system to be unworkable.
The Idaho Division of Environmental Quality has spent
considerable time developing new methods to ad-
minister antidegradation based on the principle that
some limited degradation could be allowed in high
quality waters, provided that the beneficial use does
not suffer any long-term impairment.
   The problem in Idaho, the Pacific Northwest, and
the Intermountain region is that such nonpoint source
activities as timber harvest, road building, mining, and
agriculture all produce sediment in excess of natural
conditions. This situation might possibly be managed
to  minimize the  delivery of  sediments to surface
waters, were it not for an extensive geologic formation
called the Idaho Batholith. This large granitic forma-
tion and its border geologies cover roughly 22,000
square miles of central Idaho and western Montana.
This formation is not unique to Idaho and Montana;
similar formations are found in northern California and
southern Oregon.
                                                   21

-------
D. M. MARTIN
  When the parent material is exposed to land-dis-
turbing activities, it rapidly decomposes into gruss.
Gruss is essentially sand, and is easily eroded into
stream courses and fish habitat as bedload sediment.
The sand bedload  is  deposited in the physical sub-
strate of the stream, which is the spawning and rear-
ing habitat for anadromous and  resident salmonids.
Asa result, streams and rivers that appear to be clean
and clear are actually killing fish or otherwise prevent-
ing  their successful  reproduction.  This  happens
through the following  process.
  The  additional  sediment  reduces  water flow
through the interstices of the gravels. This reduced
flow restricts the supply of oxygen to the eggs and the
removal of metabolic wastes from the egg pocket or
redd.  The eggs suffocate and die in their own wastes.
Those fish that manage to hatch are entombed, be-
cause  they are unable to struggle  to the  surface
through the excess sand.  Significant numbers offish
are lost due to this habitat impact.
  The preservation of a winter-rearing habitat is criti-
cal, because of the severity of winter conditions in the
region and  the resulting stress on fish populations.
Critical habitats for young fish can be  found at this
time in the cobble  and rubble areas  in the streams.
Fish seek shelter in the interstices of the rocks and
boulders to avoid the lethal winter conditions of the
water column. Excess bedload sediments tend to fill
in these areas, reducing the quantity and quality of the
habitat. Whenfishareunabletolocatesuitablewinter
habitat, they generally die.
  Sedimentation of summer-rearing habitat can also
have a significant negative impact, especially where
this type of habitat  is already limited. In these situa-
tions,  sediment fills pool habitats, reducing the areas
available for  rearing  of young and avoidance of
predation. Excess sedimentation also directly affects
the health of fish and their ability to feed.
  Summer-and-winter-rearing   habitat quality  is
monitored using the cobble-embeddedness variable.
Cobble embeddedness is the measure  of the degree
to which a cobble  is  surrounded by  sediment. This
habitat variable is measured using a hoop method,
and each rock within the hoop is measured according
to a standard technique. Hoop placement varies ac-
cording to criteria designed for both winter and sum-
mer habitat characteristics. The relationship between
cobble embeddedness and fish survival is well docu-
mented in the literature. As cobble  embeddedness
increases, fish survival decreases.
  The quality of spawning habitat is monitored by a
variable referred to as "percent fines."  Percent fines
are measured using a set of standard soil sieves. The
sieve  sizes vary, depending on the method  used to
analyze the results of the sample.  A set of sieves is
used to analyze the sample, using the three available
techniques.  An example of one analytical tool avail-
able is the  U.S. Forest Service  model, FISHSED,
which uses the relationship of all fines less than 6.34
mm. The relationship between percent fines less than
6.34 mm, and the expected number of young fish that
will be able to emerge from the substrate, is well docu-
mented. As  percent fines in the spawning substrate
increase, the expected  emergence decreases.   A
relatively small increase in percent fines can have a
large impact on the percentage of emerging  young
fish.
   Sediment is  cumulative  in these fish  habitat
parameters.  Too much sediment at one time, or with
no allowance for recovery, will severely  reduce the
habitat potential to sustain the beneficial use, the
fishery.   Hydrologists and  fisheries  biologists  in
Idaho,  Montana, and the Northwest are now working
closely together, and  it has been found  that short-
term sediment delivery eventually has long-term im-
pacts on fish  habitat.  That is, although the total
amount of sediment moving through the system is
very great, it is actually delivered in very small incre-
ments  from a number  of different  activities that
cumulatively have a very great impact.  For example,
salmonids have  been shown  for 60  years  not  to
reproduce and rear well in streams with sand sub-
strates.
   EPA is currently looking at developing criteria for
fish habitat,  thus stepping into a whole  new arena
dealing with more site-specific  conditions—such as
which habitats are limiting, rather than just one com-
ponent of an ecoregion or subecoregion.  The fact is,
we are trying to  establish  relationships between
habitat alterations or degradation and the beneficial
use. This process can be used in a number of dif-
ferent ways.
   Idaho has not had an antidegradation policy in
place for some time. The antidegradation Level Two,
referenced  in   the   antidegradation  regulations,
provides for full protection  in high-quality waters.
This kind of criteria development provides some pos-
sible definitions of full protection of the beneficial use.
Is full protection 100 percent, 90 percent, 80 percent
of habitat  capability?  Is full protection defined as a
healthy, viable, biologically strong population, or can
it be the physical habitat that can affect the popula-
tion?   Developing such  criteria may be one  way of
defining full  protection for fisheries; in this case, sal-
monids.   The  development of  numerical   criteria
based  on fish habitat  as a parameter for measuring
the health of the beneficial use may soon become a
reality.
                                                 22

-------
                                                        OFF-SITE ASSESSMENT WORKSHOP, 1989: 23-27
EPA Remote Sensing  Resource  for

Lake Management
                       	.^^^«^^_^_*^—^«<
 Mason J. Hewitt, III
 Thomas H. Mace
 Ross S. Lunetta
 U.S. Environmental Protection Agency, Environmental
 Monitoring Systems Laboratory, P.O. Box93478,
 Las Vegas, Nevada 89193-3478
                                        ABSTRACT

           The EPA has employed remote sensing instrumentation for the last 10 years to address lake management
           p ob^ms  TheTSnmental Monitoring Systems Laboratory-Las Vegas (EMSL-LV) is responsibft> for
           p oneelg active and passive sensing systems to address Agency needs. Passive remote sensing techno -
           ™""L of aerial Dhotoaraphy as well as airborne and satellite-borne *V^ S™™-*"**™'*™'
                                          ig system, has been developed at EMSL-LV. Passive systems
                                          il, macrophyte, wetland, thermal point and nonpoint runoff is-
                                           ' i laser fluorosensor can be used for interrogating the water
            and geographical information systems (GIS) data bases.
                Background

  Water quality management is a high priority within the
  US  Environmental Protection Agency. Legislative
  mandates such as the Clean Water Act, Water Quality
  Act, and the Clean Lakes Act specifically task the EPA
  with  assessing, protecting and/or  remediating the
  water quality of the nation's lakes and rivers. However,
  faced with an ever-increasing mandate and diminish-
  ing funds, the EPA has been faced with finding cost-ef-
  fective methods for the assessment of lake problems.
  Over a decade ago, EPA management determined
  that  remote sensing technology could play a benefi-
  cial  role by providing  vital information to  address
   problems  at a low cost, while providing a  synoptic
  view of the lake and its watershed. The Environmental
   Monitoring Systems Laboratory in Las Vegas, Nevada
   (EMSL-LV) took the lead in acquiring, developing and
   testing remote sensing technology to meet the need
   for information.
  This paper will address the remote sensing sys-
tems employed by the EPA for lake and watershed
characterization. The intent is to provide the lake
manager with an introduction to the technology by
demonstrating  its use with  several  practical  ex-
amples.
   Introduction to Remote Sensing

 Remote sensing is the science of acquiring informa-
 tion from afar, employing instrumentation that records
 electromagnetic energy reflected from a target. In its
 simplest form, your home camera is a remote sensing
 device that records, on film, energy reflected from the
 sun or flash unit.
   Generally, remote sensing instrumentation is clas-
 sified into active and passive categories, depending
 on the source of reflected energy. Passive systems
 record energy reflected from an external source such
                                                23

-------
 M. J. HEWITT III, T. H. MACE, AND R. S. LUNETTA
 as the sun. Hand-held or airborne cameras record
 reflected energy with a field of view dependent upon
 the focal length of the camera and distance from the
 target.  Different portions  of the  electromagnetic
 spectrum may be recorded by changing the film type
 in the camera. Infrared film is useful for vegetation and
 algae  assessment,  because  chlorophyll reflects
 higher in the infrared portion of the spectrum.
   Active remote sensing instruments are a class of
 devices that interrogate a target with an emitted pulse
 of energy. An easy-to-understand example is Radar,
 which pulses  microwaves  and  receives a  return
 reflection. EMSL-LV is pioneering a class of  active
 sensors constructed around laser technology. The
 most promising laser-based sensor for water quality
 assessment is the laser fluorosensor.
   Both active and  passive collectors may record
 reflected energy as a digital signal. Cameras record
 energy as an analog signal on film, while digital sys-
 tems convert the reflected analog signal to a numeric,
 and then record it on magnetic tape.
   Film and digital  products have their advantages
 and  disadvantages. Film-based products must be
 analyzed by an experienced photo interpreter  to
 derive maximum benefit from the product. While this
 is useful for certain projects where the  cognitive
 ability  of the  photo interpreter is needed to dis-
 criminate among features, the process is limited by
 the  sensitivity  and emulsion  of the film.  Digital
 products, on the other hand, may be analyzed by a
 computer system for the extraction of information,
 and  provide a wider bandwidth coverage.  The main
 disadvantage of digital systems is the ground resolu-
 tion limit. The best resolution with an airborne digital
 system is around 2 m, with the average for satellite im-
 agery ranging between 20 and 30 m.
   Many factors affect the success of a remote sens-
 ing mission. Cloud cover completely obscures all fea-
 tures  below   the   clouds,  while  atmospheric
 disturbances and airborne material can cause distor-
 tions in reflected energy. Distortions caused  by air
 disturbance can usually be corrected  during the
 analysis process.
   Several measures of  lake quality and character
 may be  acquired  using  remote sensing devices.
 Since light penetration in water is directly related to
 suspended  material in the water column, measures of
turbidity and clarity may be made using active and
 passive collectors.  Sediment concentrations may
 also be measured using remote collectors, because
 suspended  sediments affect the reflectivity of the
water. If a relationship between sediments and other
 parameters of concern can be established, then sedi-
ment concentration may be used  as a  surrogate
measure for  such  parameters  as conductivity,
nutrients, and chemicals.
   In waters with low turbidity, passive remote sens-
 ing can be  utilized to characterize subsurface fea-
 tures. Dense stands of submerged macrophytes and
 shallow  water (<10 m)  bottom types  can  be
 delineated under optimal conditions.
   Additionally, remote sensing devices may be used
 to provide a synoptic view of the watershed. This is
 particularly  useful, since  lake  problems seldom
 originate at the  shoreline or within the lake itself.
 Remote sensing may be a cost-effective method of
 acquiring information about point and nonpoint sour-
 ces within the lake basin.
               Applications

This section of the paper will highlight specific ap-
plications of photographic and digital imagery, the ra-
tionale for choosing the sensor, and the results.

Aerial Photography

The EPA has acquired thousands of photographs of
water bodies within the U.S. in order to characterize
the health  of  lakes.  These  missions can provide
photography  at scales that  vary from 1:24,000 to
1:6,000, using natural or infrared film. Even though
the majority of the aerial missions for the EPA are re-
lated to hazardous waste sites, often the photo mis-
sion  will  include photographing  nearby  lakes,
because of the threat imposed by the sites. In addition
to current photography, EMSL-LV is able to search the
National Archives for older photography.  Historical
aerial photography (dating back to the late 1930s for
most parts of the country) is an excellent resource to
use when attempting  to reconstruct the land use ac-
tivity within a watershed.
  Atypical photo  mission for lake assessment might
involve acquiring  information to map algal blooms,
submerged aquatic vegetation, and contiguous wet-
lands; to locate failing septic systems; to characterize
watershed land use, and to identify point source pol-
luters. EMSL-LV maintains a highly competent staff of
photo  interpreters who  can turn out  annotated
photographs bound in hardcover binders for use by
the lake manager.

Airborne Digital  Imaging
EMSL-LV maintains a Las Vegas-based aircraft con-
figured with a 9 x 9 inch  mapping  camera and a
Daedalus  1260-11 channel  multispectral  scanner
(MSS). Table 1 lists the wavelength coverage for the
scanner. Channel 11  can be set to provide high and
low gain in  the thermal infrared for thermal imaging
                                                24

-------
                                                            OFF-SITE ASSESSMENT WORKSHOP, 1989: 23-27
Table 1.-Daedalus  1260  MSS  Channel  and wave-
                       0.80-0.89
                       092-1 10
                       80 -14.0
 capability. Also, a mld-IR (1.55-1.75//m) band can be
 substituted for the high and low gain thermal bands,
 corresponding to TM band 5.
   There has been considerable research on the utility
 of MSS imagery for the assessment of water quality
 parameters.   Witzig and  Whitehurst  (1981)  have
 catalogued the applications of airborne MSS, but  in
 general  parameters such as thermal plumes, tur-
 bidity,   Secchi depth,  chlorophyll  a,  and  total
 suspended solids may be correlated with a spectral
 response. Provided a significant relationship can be
 established  with suspended sediments, regression
 coefficients   may  be derived  for  non-reflecting
 parameters  such as  phosphorous, nitrate-nitrogen
 and conductivity.
    A typical application involved the  Flathead Lake,
  Montana. The objective of the study was to determine
  if airborne MSS could be used to  describe the dis-
  tribution of nutrients associated with spring runoff
  events (Mace, 1982). The specific parameters in ques-
  tion included surface temperature, transmittance,
  Secchi depth, conductivity, total suspended solids,
  dissolved orthophosphorus, total  phosphorus, and
  nitrogen. Three flight lines at 6,400  m (21,000 ft) were
  flown to acquire MSS data for the entire lake. Nominal
  pixel size at nadir was 13.8 x 13.8 m. Airborne imaging
  was simultaneous with in situ lake monitoring to assist
  with the image analysis.
     The image analysis showed strong correlations be-
   tween all in situ sampling parameters and the spectral
   responses recorded by the MSS sensor. This correlat-
   ing of spectral response to in situ sampling allows the
   extrapolation of parameters to the  entire lake.

   Satellite Digital  Imaging
   EMSL-LV is able to acquire data  and analyze digital
   imagery from the two major earth-orbiting satellites:
   LANDSAT and SPOT. Both of these systems orbit in a
near-polar orbit which provides coverage of every
place in North America every 14 to 16 days. LANDSAT
and SPOT work on the same principle as the airborne
MSS system: they record spectral response from the
earth's surface in a digital form for analysis. The sen-
sors vary in terms of spectral coverage and ground
resolution,  and an experienced analyst should be
consulted in orderto select the best sensorforthe job.
Scarpaceetal. (1979) have provided a detailed review
of LANDSAT applications for water quality assess-
ment.
   EMSL-LV has undertaken several major  projects
that illustrate the use of these sensors. EPA Region 5
asked EMSL-LV to assess nearshore water quality for
 Lake Ontario (Mace, 1983). The water parameters of
 concern were Secchi depth, turbidity, conductivity,
 dissolved  oxygen, alkalinity,  nitrate-nitrogen, total
 nitrogen, total phosphorus, and chlorophyll a.
    LANDSAT 4 MSS was selected as the sensor of
 choice because of its ground resolution (80 m) and
 spectral coverage. Two scenes (one scene = 185 x
 185 km) of LANDSAT MSS were large enough to in-
 corporate the areas of concern around Niagara Falls,
 Rochester and Oswego, New York. Simultaneous in
 situ sampling of the lake was accomplished during a
 normal  lake cruise, and data stored in the STORET
 data base was matched with imagery from the satellite
 fly-over. The results of the lab analysis for the above
 parameters were incorporated into the image analysis
  routine.
    This analysis  showed significant relationships be-
 tween several of the water quality parameters and the
  Landsat  data. Total  phosphorus,  nitrate-nitrogen,
  alkalinity, and  conductivity possessed high  coeffi-
  cients of determination, relatively low standard errors,
  and  statistical significance better than 0.001 (Mace,
  1983).  It was also shown that these non-reflecting
  parameters were closely tied to turbidity, and that a
  statistical relationship existed as demonstrated by in
  situ sampling. Therefore, in this case, reflectivity from
  suspended  solids could be used  as a  surrogate
  measure for non-reflecting properties. This type of
  program-built upon lake monitoring lab analysis to
  establish relationships and image analysis-enables
   extrapolation of point data to entire lake bodies. In the
   case of any large lake, this can be measured in terms
   of cost savings to lake monitoring programs.
     An additional benefit of satellite remote sensing is
   that it provides a synoptic view of the watershed.
   Since lake problems seldom begin in the lake itself,
   this can provide the lake manager with an all-encom-
   passing view of the watershed, to assist in detecting
   point and nonpoint source problems.

-------
 M. J. HEWITT 111, T. H. MACE, AND R. S. LUNETTA
           Laser Fluorosensor

 The laser fluorosensor is an active remote sensing in-
 strument that has shown great promise for  lake
 monitoring. Under development now at EMSL-LV, this
 system is based around laser technology that pulses
 downward from an aircraft, interrogating the water
 column and detecting reflected and emitted energy.
 The system,  which does not  depend on reflected
 energy from the sun, has shown promise in detecting
 such parameters as chlorophyll, dissolved organic
 material (DOM), and optical attenuation. Established
 correlations with fluorescence of the water column
 can lead to extrapolation to other parameters, such
 as dissolved organic carbon and turbidity.
   To  demonstrate  the  capability  of   the  laser
 fluorosensor,  a prototype version was mounted in an
 airborne platform and flown along a 734 km segment
 of the lower Snake and Columbia rivers (Bristowetal.
 1985). This system consists of eight major hydro-
 electric dams and impoundments. River  centerline
 profiles of three parameters were measured concur-
 rently for chlorophyll a,  optical attenuation and dis-
 solved organic carbon.
   Flying at an average of 250 m in altitude, 4,800
 water column interrogations were obtained, essen-
 tially providing a nearly continuous profile of the im-
 poundment  system.  The airborne measurements
 were correlated with in situ sampling at 12 monitoring
 stations. The results of the survey  demonstrated
 strong correlations between the laser fluorosensor
 signal and chlorophyll, optical  attenuation, and dis-
 solved organic carbon  (r  =  0.96, 0.96  and  0.88
 respectively) (Bristow et al. 1985). It is expected that
 further work on the laser fluorosensor will provide the
 EPA with a unique tool for lake monitoring in a near
 real-time sense.
 color paper plots may be produced that highlight the
 features or problems of concern. The degree of detail
 captured by the paper product is governed by the
 ground resolution of the sensor.
   The GIS is a new method of displaying and using
 remote sensing data. GIS is a computer-based sys-
 tem that provides for the storage, retrieval, and dis-
 play  of spatial  information-it  is an  automated
 cartography system linked to a relational data base.
 This linkage allows the lake manager to not only cap-
 ture the cartographic component of the remotely
 sensed data, but to build a data base describing the
 cartographic features. The GIS also allows for the
 flexibility of overlaying remote sensing data with other
 data  sources  such as census, transportation  net-
 works, point sources,  and other ancillary data that
 can be mapped. In its final form, the GIS allows inter-
 active query of the data base with graphic output to a
 video terminal and/or plotter.
    Requests for Remote Sensing
                  Support

Each of the 10 EPA Regions has a remote sensing
coordinator on the staff who is responsible for fielding
requests for remote sensing support from the states
and  EPA programs within the region. Requests for
remote sensing products are forwarded to the Ad-
vanced Monitoring Systems Division of EMSL-LV.
   Lake  managers are  encouraged to contact  a
remote sensing professional early in the monitoring
design stage to obtain advice on the correct sensor,
costs, and timeline for their projects. Only when a
remote sensing project is carefully designed can suc-
cess be assured.
      Remote Sensing Products

EMSL-LV maintains a trained  staff of photo inter-
preters with experience in  environmental assess-
ment. The typical product is a "picture book" with
features annotated on mylar overlays. These hard-
bound volumes are a ready-made atlas that can be
used either in the field or in discussions with manage-
ment or the public.
   Products  from  the   digital  sensors  can  be
presented in two ways, paper plot and geographic in-
formation system. Once the digital data are classified,
                 Summary

The EPA may draw on a wide range of remote sensing
instrumentation for quantifying lake parameters. The
EPA's remote sensing arm resides within the EMSL-
LV laboratory, and consists of active as well as pas-
sive collection systems. EMSL-LV has demonstrated
the capability of airborne and satellite-borne collec-
tion system with actual projects, and is pioneering the
development of active systems to support the needs
of the EPA and the states for information about the
nation's lakes.
                                               26

-------
                                                                       OFF-SITE ASSESSMENT WORKSHOP, 1989: 23-27
                   References

Bristow, M.P., D.H. Bundy, C.M. Edmonds, P.E. Ponto, B.E. Frey,
   and L F. Small. 1985. Airborne laser fluorosensor survey of the
   Columbia and Snake rivers: Simultaneous measurements of
   chlorophyll, dissolved organics and optical attenuation. Int. J.
   Remote Sens. (6) 11:1707-34.
Mace. T.H. 1982. Characterization of lake water quality parameters
   with airborne multispectral scanner data: Flathead Lake, Mon-
   tana. Proc. ACSM/ASPRS Spring Conf. March 14-20. Am. Soc.
   Photogram. Remote Sens., FallsChurch.VA.
	 1983. LANDSAT MSS classification of near-shore water
   quality in Lake Ontario.  EPA Rep. TS AMD-82095. Environ.
   Monitor. Sys. Lab., Las Vegas, NV.                 ..,„„..,
Scarpace, F. L, K.W. Holmquist, and LT. Fisher 1979. LANDSAT
   analysis of lake quality. Photo. Eng. Remote Sens. 45(5): 623-
   33.
Witzig, A.S. and C.A. Whitehurst 1981. Literature review of the cur-
   rent use of MSS digital data for lake trophic classification.
   Proce. ACSM-ASP 1981 Fall Conference. Am. Soc. Photogram.
   Remote Sens., Falls Church, VA.
                                                            27

-------
                                                         OFF-SITE ASSESSMENT WORKSHOP, 1989: 29-37
Modeling Linked  Watershed and Lake
Processes  for Water  Quality  Management
Decisions
 R.M. Summer*
 U.S. Department of Agriculture, Agricultural Research
 Service, Morris, Minnesota 56267

 C.V. Alonso
 U.S. Department of Agriculture, Agricultural Research
 Sen'ice, Fort Collins, Colorado 80522

 R.A.Young
 U.S. Department of Agriculture, Agricultural Research Service,
 Mortis, Minnesota 56267
                                         ABSTRACT

            A physically based modeling approach is used to link lake processes with watershed processes and to
            simulate lake and watershed responses to land management and weather conditions. AGNPS (Agricultural
            Nonpoint Source model) simulates hydrology, erosion, sediment transport, and transport of nitrogen and
            phosphorus within a watershed. Using a cellular structure, runoff, sediment, and chemical input variables
            from the watershed are linked to a lake model, FARMPOND. This one-dimensional lake model simulates
            temperature stratification mixing by wind, sedimentation, inflow density current, and algal growth. Un-
            steady advection-diffusion equations characterize the dynamics of suspended sediment, nitrogen and
            phosphorus, and chlorophyll. Random generation of weather conditions on a daily basis are used to drive
            the model  Resulting impacts of alternative management plans are simulated  by changing agricultural
            practices and land use, thereby modifying inflow characteristics to lakes. Modeling capabilities are being
            tested on eutrophic lakes in Missouri and Minnesota to simulate long-term trends and impacts of best
            management practices.
                 Introduction

  Lakes are prominent features of the landscape. The
  quality  of  their  water  reflects  physical/chemi-
  cal/biological interactions between the water column
  and sediments with the surrounding drainage basin
  and the atmosphere. This interaction is most obvious
  when the drainage basin is large relative to the lake.
  Such is the case in many agricultural  watersheds.
  Here, chemical and sediment transfer from fields and
  ditches to ponds and lakes accelerates lake aging
(eutrophication) by promoting infilling and poor water
quality.
  The magnitude and timing of lake degradation by
agricultural-related activity is difficult to predict be-
cause of the complexity of the land and water system.
For example, erosion and  chemical  contaminant
transport occur at variable rates during cyclic varia-
tions in crop rotations. These rates of sediment and
chemical transport to water bodies are confounded
by seasonal variations, the random nature of weather
events, and occasional major changes in land use.
  * presently at USDA-ARS, P.O. Box 1430, Durant, OK 74702-1430
                                                29

-------
R. M. SUMMER, C. V. ALONSO, AND R. A. YOUNG
  Considering the complexity of water quality im-
pacts, models are useful in supplementing insights
gained  from field  observations  and  laboratory
analysis. Growing concern among the agricultural
community  and lake  users  has  stimulated  the
development of a model of processes governing the
movement of sediment and nutrient-laden water in
and through lakes. Hence, a watershed—lake project
was conducted to link watershed and lake processes
and simulate water quality and quantity leaving the
lake. The objectives of this phase  of study were to
evaluate trends in water quality  within and leaving
lakes and to assess impacts of best management
practices and changing land use on water bodies in
agricultural watersheds. Only short-term trends are
reviewed, although the strength of the model lies in
simulating long-term trends.
          AGNPS-Lake Model

This study assumes that ponds or lakes are com-
ponents of agricultural watersheds. The model must
be capable of realistically simulating nonpoint source
pollution under varying land management practices,
simulating the randomness of weather conditions,
and  linking selected watershed and lake processes
governing water quality.
  The models selected to meet these requirements
are a watershed model, Agricultural Nonpoint Source
model (AGNPS) (Young et al. 1987), and a lake model
adapted  from FARMPOND (Gallegos et al. 1983),
which itself was adapted from RESQUALII (Stefan et
al. 1982a). The physically based computer simulation
model, AGNPS, was developed by the USDA Agricul-
tural  Research Service as a tool  for objectively
evaluating nonpoint source pollution from agricul-
tural watersheds. The model is  event-based but is
being adapted to an annual basis by using a daily time
step. Runoff, erosion,  sediment transport,  chemical
transport of sediment-attached and soluble nitrogen
and  phosphorus, and chemical oxygen demand are
simulated from primarily nonpoint sources. It is a dis-
tributed parameter model that works on a cell basis,
whereby the watershed is divided into uniform grids
for  input and analysis. Descriptors of each cell in-
clude land type, soil, vegetation type and maturity,
cultural practice, fertilizer application, SCS (Soil Con-
servation Service) curve number, slope, and aspect.
Feedlots, gullies, and other point sources, terrace im-
poundments, and site-specific deposition can also be
included in the analysis. A groundwater submodel is
being  linked to  the  model  (Nieber  and Lopez-
Bakovig, 1988). Further details about mathematical
formulations and routing sequences can be found in
Young et al. (1987) and application and verification
studies are described in Lee (1987) and Bingner et al.
(1987).
  The lake model  is a vertical  one-dimensional
model consisting of routines for water storage and
routing,   temperature   stratification,   sediment
transport and routing, plunging density current, light
penetration, and  wind  mixing.  Components  for
chemical and  biological  simulation represent  the
transfer and kinetics of nitrogen and phosphorus and
growth of phytoplankton  represented by chlorophyll
a. Driving forces include inflow and weather condi-
tions. Flow rating at the outlet is simulated with func-
tions  describing volume/stage  relationships.  The
model can generate rating curves for various designs
of Soil Conservation Service retention structures.
Boundary conditions include heat flux through the
surface water-air interface, no heat transfer through
the lake bed, wind shear at the surface water-air inter-
face, no water transfer through the lake bed except at
the phreatic surface, and no sediment resuspension
(after sediment settles it is lost to the system) (Fig. 1).
   Dynamics of all substances and intrinsic proper-
ties in the water column are described by the one-
dimensional advection-diffusion equation:
         ac    acA   a   ac  sources
       A — + v — = — kA — ± 	
         at    az    az   az   sinks
   where:
   A  = surface area of each horizontal layer (m2)
   c  = concentration (mg/L) or intrinsic property (e.g. water
      temperature)
   t  = time (day)
   v  = vertical velocity (m/day)
      v = 0 for soluble substances and temperature
   z  = depth (m)
   k  = turbulent diffusion coefficient (m2/day)

   This relationship  describes changes in tempera-
ture  and concentration as a function of depth and
time. The function operates iteratively within sequen-
tial horizontal layers  and assumes instantaneous
mixing and equilibrium conditions in the horizontal
direction.
   Additional details on the theory behind the mathe-
matical formulations of the lake model can be found in
Stefan and Ford (1975a, b), Dhamotharan et al.
(1978), Ford and Stefan  (1980), Dhamotharan et al.
(1981), Stefan et al. (1982b, 1983) and Riley and
Stefan (1987).
        Time and Space Scales

A daily time step was chosen to simulate selected
processes   representing   watershed  and  lake
dynamics on a seasonal, yearly, and long-term basis.
Shorter time  steps were not selected because the
                                                30

-------
                                                           OFF-SITE ASSESSMENT WORKSHOP, 1989: 29-37
Figure 1 .-Schematic diagram of lake model (modified from Riley and Stefan, 1987).
thermal energy budget  of  lakes is considered in
balance only on a time scale of one day (Ford and
Thornton, 1979). Since parameters within the advec-
tion-diffusion equation used to simulate transfer of all
substances within the water column depend on the
thermal structure,  a time step of  one day is  ap-
propriate. Other factors governing  selection of the
time step are time requirements and memory restric-
tions in running the model and availability of a data
base for time intervals less than a day. Larger time in-
tervals tend to mask the effect of in-lake biochemical
changes and hydrologic events within the watershed.
   The spatial  scale appropriate for applying this
 model is difficult to define until the linked simulations
 have been thoroughly tested. Ponds and lakes rang-
 ing in area from several ha to 1,000 ha are considered
 appropriate for one-dimensional models according to
 most criteria (Ford and Thornton, 1979). The water-
 shed model has been applied on 1 to 60,000 ha water-
 sheds. Theoretically, there  is no limit to the size of a
 watershed; rather, when channel processes become
 a dominant factor, this model should not be used.
    Further discussion on the physical basis for select-
 ing time and space scales and uncertainties that arise
 during  "lumping"  and  "splitting"  time  steps  and
 processes is given by Ford and Thornton (1979) and
 Ferreira and Smith (1988)
functions for the watershed and lake models. A daily
sequence of meteorological parameters including air
temperature,  relative  humidity,  solar  radiation,
precipitation, wind direction and speed, and cloudi-
ness are generated and  the distribution of values
reflects a  combination  of  the  periodicity  of  a
parameter distribution for a selected geographic area
and stochasticity. The code is  written  in  FORTRAN
and executable  on IBM-compatible personal com-
puters.
   A geographic grid structure and timing of AGNPS
provide a framework for routing  inflow to the  lake
model. First, water cells representing a lake within a
watershed grid are tagged in AGNPS. Output from all
 land cells bordering and delivering runoff, sediment,
 and nutrients to the lake is accumulated. This ac-
 cumulated output is routed as input to the lake model.
 Additional inflow is contributed by the groundwater
 submodel and by dry and wet atmospheric loading of
 nitrogen  and phosphorus defined  within the lake
 model. Hydrologic and biochemical simulation in the
 lake model is not based on the grid  structure; rather,
 simulation is one dimensional in the vertical direction.
 At the end of a day, output at the lake outlet becomes
 input to the receiving cell downstream of the lake.
 Simulation then continues in AGNPS and culminates
 at the outlet of the watershed.
           Structure and Linkage

  The lake model is linked to the watershed model to
  more fully characterize the system  and meet  the
  needs of a water quality management assessment. A
  weather generator (Nicks etal. 1987) providesforcing
              Test Simulations

  A realistic and credible representation of system be-
  havior  is the ultimate goal of most water quality
  models such as AGNPS-LAKE. A testing phase is
  proceeding to determine if important processes oc-
                                                  31

-------
R. M. SUMMER, C. V. ALONSO, AND R. A. YOUNG
curring in the natural system are being represented
by the model. Initial conditions are given below fol-
lowed by tests in wetland and watershed manage-
ment.

Site Description and Initial Conditions
Simulations were conducted on Eagle Lake water-
shed in West Central Minnesota (Fig. 2). The entire
watershed, except for some internally drained areas,
drains into Eagle lake, whose outlet (cell 207) is also
the outlet of the watershed. A sharp-crested weir at
cell 207 controls the outflow from the lake and water-
shed.
  The watershed, 3,630 ha  in area,  is primarily
agricultural with  residential development bordering
the 360 ha lake. The average residence time of lake
water is 3.9 years. Studies in 1974 and 1978(Latterell
et at. 1979) indicated  the  lake  was seasonally
eutrophic, although a chlorophyll a concentration of
12 ppb during late summer algal blooms is relatively
low compared to other Minnesota lakes.
  The following  scenarios are based  on the same
weather and watershed conditions unless  noted.
First,  June weather  conditions were  chosen  for
simulation and comparison  with  measured  data.
Agricultural practices and maturity of crops repre-
senting growth conditions in June during the early
1980s were assigned to the watershed. Then a 19-day
"warm-up" period was conducted to bring initial con-
ditions within the lake to a level of steady state before
running   wetland  and  watershed  management
scenarios. As  results in Table 1 suggest,  a longer
period of simulation may be required to more realisti-
cally represent nitrogen  and phosphorus distribu-
tions in the  epilimnion  (upper lake layers)  and
hypolimnion (lower lake layers). Following the 19-day

Table 1.—Simulated versus observed in-lake condi-
tions in June 1978, Eagle Lake, WIN.

Clay ppm
NO3 ppm epilimnion
hypolimnion
PO4 ppm epilimnion
hypolimnion
CHLOR-a ppb
TEMP C epilimnion
hypolimnion
SIMULATED
19 DAYS
8-14
.01
02
06
06
5.0
20
12
OBSERVED*
10-20
.01 -.07
004- 01
.001-018
.1-.17
1.0-8.0
22
14
'June observations in Latterell et al 1979
 • EAGLE  LAKE
 SPOINT  LAKE
 0WETLAND
 BPASTURE
 DCROPLAND
Figure 2.—Cellular grid representing Eagle Lake watershed, M N. One cell = 0.16 ha.
                                               32

-------
                                                          OFF-SITE ASSESSMENT WORKSHOP, 1989: 29-37
run, a precipitation event equivalent to a 25-year, 24-
hour storm (123 cm) for that location was simulated,
followed by 20 days of average June surface inflow to
the lake.

Comparison of Model Outputs
Before conducting simulation  runs of  the  linked
model, the lake output from AGNPS without the lake
model  and AGNPS-LAKE were compared.  Runoff,
erosion, and pollutant transport within the watershed
and lake have been previously simulated using an
event-based version of AGNPS without a lake com-
ponent (U.S. Soil Conserv. Serv. and Kandiyohi Con-
serv. Distr, unpubl.) This simulation used a simple
routine for routing runoff through the lake and settling
various sediment sizes.
   Comparison of results simulating a 25-year storm
 event with and without the lake model is illustrated in
 Figure 3. Output of suspended sediment (clay) from
 the event-based version of AGNPS gives a single
 value of clay mass at the lake outlet, whereas output
 from AGNPS-LAKE gives daily outflow quantities.
 Large differences exist; these are expected given the
 simple routing of a flood event through Eagle Lake by
 AGNPS versus the dynamics simulated by the lake
 model on a daily time step. These results demonstrate
 the need for such a linkage on a  daily time step be-
 tween a watershed and lakes, and the possibility that
 lakes may provide an attenuating influence on the sys-
 tem.
             EAGLE LAKE WATERSHED, MN
         CLAY MASS OUTFLOW ~ 25 YR EVENT
     1500

  CLAY
 (1000 kg)
     1000-
                   AGNPS (Event based)
                    LAKE (Daily  output)
                             I
         .GNPS-25 Y LAKE-25 Y   DAY 1     DA 5     DA 10
  Figure 3.—Outlow water quality using AGNPS and AGNPS-
  LAKE.
  Wetland Impact
  The role of wetlands near the major lake inlet was
  evaluated by simulating wetland and nonwetland con-
ditions in the watershed and comparing outflow and
in-lake profiles generated by AGNPS-LAKE. Five cells
(80 ha) of wetlands at the inlet (Fig. 2) were refor-
matted and treated as part of the lake during the non-
wetland simulation.
   As expected, routing surface water flow through
wetland cells decreased sediment mass, particularly
silt, by about one third (Fig. 4). The peak clay mass
decreased on  the  first day  under  the wetland
scenario, but, unlike silt, clay remains high for several
days following the storm event. The physical explana-
tion for this response is related to the depth of plung-
ing density currents  (Fig. 5) and depth of withdrawal
from the lake. The depth of withdrawal at the outlet on
the first day is 12 meters but decreases to about 9.5
 meters  on the second day (because of decreasing
 head at the outlet). Only peak clay concentrations of
 the wetland inflow are withdrawn on the second day,
 while the nonwetland inflow, being more dense, plun-
 ges deeper in the profile. The silt density current (Fig.
 5)  is  reduced  substantially  under the  wetland
 scenario, so thatthe depth of withdrawal does not sig-
 nificantly change the outflow concentration.

             EAGLE LAKE, MN
    SILT MASS OUTFLOW ~ 25 YR STORM
     3500

     3000


 SILT"00'
 (kg) 2000'
 •  NON-WETLAND

— WETLAND
               EAGLE  LAKE, MN
      CLAY MASS OUTFLOW -- 25 YR STORM
   Figure 4.—Effect of wetland on sediment outflow.
                                                 33

-------
R. M. SUMMER, C. V. ALONSO, AND R. A. YOUNG
                    EAGLE  LAKE, MN
                          NON WETLAND DAY 1
                      *	* WETLAND, DAY 1
                      G	ONON WETLAND. DAY 2
                      A——& WETLAND, DAY 2
                                                                        EAGLE LAKE, MN
                                                                       IN-LAKE N fc P •	2S YR EVENT_
               100  150 200 250  300  350  400  450 500

                      CLAY (mg/L)



                    EAGLE LAKE,  MN
                   IN-IAKE SILT — 25 YR EVENT

                            • NON WETLAND, DAY 1
                            * WETLAND, DAY 1
                        O	ONON WETLAND, DAY 2
                            WETLAND. DAY 2
                  12  16   20   24   28
                     SILT (mg/L)
Figure 5.—Effect of wetland on In-lake sediment distribution.

   In addition to reducing sediments, wetlands may
also decrease nutrients. Routing runoff and erosion
through wetlands decreased total nitrogen and phos-
phorus loading to the lake by  15 percent. This
decrease is reflected in less organic nitrogen and
phosphorus within the lake (Fig. 6), together compris-
ing approximately 80 percent of the nitrogen and
phosphorus loading following the storm event. Con-
centrations of soluble nitrogen and phosphorus in the
lake did not change noticeably.
   Maximum concentrations of chlorophyll a within
the lake occurred in  15 days and 8 days under non-
wetlands and wetland conditions, respectively (Fig.
7). The wetland simulation agrees with reported peak
chlorophyll a measurements in Eagle Lake (Latterell
et al 1979), and the higher concentration, 22 ppb, for
nonwetland is well within the range of concentrations
measured in other lakes in Minnesota. The higher
concentration can be attributed to higher loading of
nitrogen and phosphorus and a deep  mixing zone
driven by high winds.
  The simulated filtering  capability of wetlands is
reasonable and in agreement with field observations
in a variety of ecosystems as reviewed by Windell et
al. (1986). Because wetlands comprise a natural com-
ponent of this particular watershed, it is not possible
                                                                                    ORGANIC P — NON WETLAND-
                                                                                    ORGANIC P — WETLAND
                                                                                    ORGANIC N	NON WETLAND
                                                                                    ORGANIC N	WETLAND
                    N &  P  (mg/L)
Figure 6.—Effect of wetland on the distribution of in-lake or-
ganic nitrogen and phosphorus.
to determine how realistic their impact is on  lake
dynamis. The results corroborate findings by Obert
and Osgood (1989) that sediment-attached nutrients
flowing into lakes can be significantly reduced by wet-
lands. However, the role of wetlands in reducing in-
lake nutrient loading is more complex because of
variability in  mixing  depths,  temperature patterns,
and nutrient release from  bottom sediment. The ef-
fects of wetlands on lake water quality are discussed
by Goldstein (1986) and Obertsand Osgood (1989).
                    EAGLE LAKE, MN
                 | IN-LAKE CHLOR-g -- 25 VR EVENT
DEPTH-,
 (m)
                                NON WETLAND. DAY 1
                                NON WETLAND. DAY 15 (MAX)
                                WETLAND. DAY 1
                              -fi WETLAND. DAY a (MAX)
                       10      15
                    CHLOR-a (ppb)
Figure 7.—Effect of wetland on growth of chlorophyll a.


Watershed Management

Projected impacts of land management alternatives
were examined by comparing crop and pasture land
before and after implementation of a Conservation
Resource Program (CRP). CRP is a federal program
that retires highly erodible cropland for a 10 year
period.  Fourteen percent of the watershed (27 per-
cent of the cropland) was seeded with permanent
grass cover and several terraces were built.
                                                 34

-------
                                                          OFF-SITE ASSESSMENT WORKSHOP, 1989: 29-37
  No monitoring has occurred since the CRP prac-
tices began during 1986 and 1987. Therefore, only
relative differences are projected in light of a 25-year
storm event during the June growing period.
   Improvements in water quality, both in-lake and
outflow, under post-CRP conditions are illustrated in
Figure 8.  Clay and silt mass in the  outflow are
projected to decl ine markedly d uring the first five days
following the storm event. A net reduction in total
phosphorus and total nitrogen is also indicated. An in-
crease in nutrients in the outflow beginning nine days
after the storm event is short-lived and is attributed to
errors  incurred  in  mechanical adjustment  and
redistribution of lake layer thicknesses by the model
as density and volume changed during inflow, outflow,
and temperature and wind fluctuations.
   The timing and peak  concentrations of chlorophyll
a (Fig. 9) suggest that CRP decreases both the time to
 peak algae bloom and the magnitude, relative to pre-
 CRP conditions. Values of chlorophyll a concentra-
 tions are within the observed range in June under
 pre-CRP  conditions (Latterell  et al.  1979). Clay,
 nitrogen, and phosphorus in the outflow on the first
 and tenth day following the storm event are compared
 in Table 2. Total nitrogen and phosphorus declined by
 about  one-fourth under CRP conditions. NO3 and
 available phosphorus increased from day 1 to day 10
 under both conditions, reflecting the high loading of
 organic nitrogen and phosphorus in  the inflow and
 high temperatures supporting  nutrient transforma-
 tions.
     Clearly,  longer simulations  are   necessary to
  evaluate land use changes against the pertubations of
  weather over several seasons and years. Post-audit
  analysis is required to verify the trends predicted by
  the model  (Donigian and Rao,  1988). Unfortunately,
  lake management modification such as copper sul-
  fate treatment and a highly improved sewage treat-
  ment  system complicate Eagle Lake's response to
  changes in watershed management. Further monitor-
  ing is required to adequately separate water quality
  response to changes within the watershed versus that
  of lake  management before the effectiveness of a
  management practice can be assessed.
         Summary and Discussion

   Selected watershed and lake processes are modeled
   within the AGNPS-LAKE model to assess trends in
   water quality and  resource sensitivity to changing
   land use. A physically based modeling approach, with
   weather functions acting as driving forces, is used to
   link  lake  processes  with watershed  processes.
             EAGLE  LAKE, MN
     CLAY MASS OUTFLOW ~ 25 YR STORM
      25 —
                                 LAKE--PRE CRP

                                 LAKE-POST CRP
 CLAY
(1000 kg)
                         10   12  14   16  18   20
                      TIME (DAY)
             EAGLE  LAKE, MN
    SILT MASS OUTFLOW --  25 YR STORM
    1000
               EAGLE  LAKE,  MN
        N & P OUTFLOW  -- 25 YR STORM
  N & P
                        8   10   12
                        TIME (DAY)
                                                     Figure 8.—Effect of watershed management practices (CRP)
                                                     on outflow water quality.
  Runoff, eroded  sediment, and chemicals from the
  watershed model provide inputtothelake model.This
  one-dimensional lake model simulates temperature
  stratification, mixing by wind, sedimentation, inflow
  density current, and algal growth. Outflow provides
  input to  downstream  segments in the  watershed
  model.
                                                  35

-------
 R. M. SUMMER, C. V. ALONSO, AND R. A. YOUNG
Table 2.—Daily outflow water quality under pre- and post-CRP watershed management
DAYS AFTER STORM EVENT
PRE-CRP

Q (m3/day)
Clay (kg/day)
Avail P (g/day)
Organic P (g/day)
Unavail. P (g/day)
Labil P (g/day)
NH4 (g/day)
N03 (g/day)
Organic N (g/day)
Adsorbed NH4 (g/day)
1ST DAY
1 56,000
21,600
42,000
130,000
25,200
530
62,000
67,000
400,000
62,000
10TH DAY
85,600
3,700
60,000
60,000
4,500
200
45,000
130,000
190,000
45,000
POST-CRP
1ST DAY
142,000
1 7,200
35,000
110,000
21,200
380
51,000
61,000
320,000
51 ,000
10TH DAY
79,500
3 400
49,000
45,000
4,400
170
34,000
110,000
140,000
34000
                      EAGLE LAKE,  WN
                           567891011
                     CHLOR-a  (ppb)

Figure 9.— Effect of watershed management practices (CRP)
on growth of chlorophyll a.
   Short-term trends in changes in water quality were
assessed by modifying  wetlands impact and land
management changes.  However, as the model is
used and improved, it will be possible to assess long-
term trends.
   Several subroutines are being added to the current
model to allow greater flexibility in  simulating  in-
dividual lake conditions, lake management practices,
and in-lake treatments.  A  bottom  sedimentation
routine is being incorporated to reflect changes in
morphometry and lake infilling over long periods of
time. Also a routine to simulate dissolved oxygen will
be added to facilitate aquatic habitat assessment.
   Future  work  on  watershed-lake interactions
should include  sediment resuspension, pesticide
pathways, and  model  uncertainty. Resuspension of
sediment into the water column of the lake is complex
and involves sediment digenesis and nutrient flux.
Havis (1 989) demonstrated the complexity of predict-
ing phosphorus limitation on algal growth assuming a
well-mixed   lake,  which  requires   knowledge  of
seasonal fluctuations in lacustrian transport  rates of
internal phosphorus. Similarly, transfers and transfor-
mations of pesticide and toxic substances require fur-
ther study. A pesticide subroutine is included in the
lake model, although parameter estimation, e.g., pes-
ticide  partition  coefficients in  the aquatic environ-
ment,   remains  a  major obstacle  in  simulating
pesticide pathways. Parameter and model uncertain-
ties need further research and investigation to better
understand the limits of simulation. Beck (1987) com-
prehensively discusses the effect of introducing un-
certainties from several sources, while Winter (1981)
describes hydrologic uncertainties.
                 References

Beck, M.B. 1987. Uncertainty, system  identification, and the
  prediction of water quality, Chapter  1 in Uncertainty and
  Forecasting of Water Quality, Springer-Verlag, Berlin.
Bingner,  R.L, C.E. Murphree, and C.K. Mutchler.  1987.  Com-
  parison of sedimentyield modelsonvariouswatershedsin Mis-
  sissippi. In Proc. Am. Soc. Agri. Eng. June 28-July 1.
Dhamotharan, S., H.G. Stefan, and F.R. Schiebe. 1978. Turbid
  reservoir stratification and modeling. |n_  Proc.  26th  Annu.
  Hydraulic Division Spec.  Conf. Am. Soc.  Civil  Eng., Univ.
  Maryland, College Park.
Dhamotharan, S., J. Gulliver, and H.G. Stefan. 1981. One-dimen-
  sional settling of suspended sediment. Water Resour. Res 17:
  1125-32.
Donigian, A.S. and P.S.C. Rao. 1988. Selection, application and
  validation of environmental models. Jnjnt. Symp. Water Quality
  Modeling of Agricultural Nonpoint Sources, Logan, Utah.
Ferreira, V.A. and R.E. Smith. 1988. The limited physical basis of
  physically based hydrologic models. Pages 10-18 in Modeling
  Agricultural,  Forest,  and  Rangeland  Hydrology. Proc. Int.
  Symp. Am. Soc. Agric. Eng., December.
Ford, D.E. and  H.G. Stefan. 1980. Thermal prediction using in-
  tegral energy model. J. Hydraulic Div. Am. Soc. Civil Eng 106
  HY1:39-55.
Ford, D.E. and K.W. Thornton. 1979. Time and length scales for the
  one-dimensional assumption  and its relation to ecological
  models. Water Resour. Res. 15:113-20.
Gallegos, C.L, F.R. Schiebe, H.G. Stefan. 1983. Modeling water
  movement  and thermal  structure in agricultural impound-
  ments. Natural Resources Modeling Symp., Pingre Park, Co.
  U.S. Dep. Agri. Agric. Res. Serv., Soil Conserv. Serv.
Goldstein, A. L. 1986. Utilization of wetlands as BMPS for the
  reduction of nitrogen and phosphorus  in  agricultural  runoff
                                                    36

-------
                                                                          OFF-SITE ASSESSMENT WORKSHOP, 1989:  29-37
  from south Florida watersheds. Pages 345-50 in Lake Reserv.
  Manage. II. N. Am. Lake Manage, Soc.
Havis RN 1988. Analytical models of lake phosphorus dynamics.
  Presented at Symp. Lake and Watershed Management, N. Am.
  Lake Manage. Soc., St. Louis, Mo., November.
Latterell,  J.J., R.S. Abbott, T.E. Straw, J.B. Van Alstine, and C.F.
   Myette 1979. Eagle Lake pollution control project: Assessment
   of lake improvement. Bull. 100. Water Resour. Res. Center,
   Univ. Minnesota.
Lee  MT  1987 Verif ications and applications of a nonpomt source
   pollution model. Pages 708-13 in. Proc. National Engineering
   Hydrology Symp. Am. Soc. Civil Eng. New York.
Nicks, A.D., D.A. Woolhiser, and C. W. Richardson. 1987. Generat-
   ing climate data for  a water erosion and prediction model. In
   Proc  Int Winter Meet. Soc. Agric. Eng. .December.
 Nieber, J.L. and I. L. Lopez-Bakovic. 1988. Model of large scale
   subsurface water flow and chemical transport In Proc.  Int.
   Winter Meet. Am. Soc. Agric. Engi. (Abstr.)
 Osgood, G.L. and R.A.  1988. The water quality effectiveness of a
   detention/wetland treatment system and its effect on an urban
   lake.  Presented at Symp. Lake and Watershed Management, N.
   Am. Lake Manage. Soc., November.
 Riley M J.  and H.G. Stefan. 1987. Dynamic lake water quality
   simulation model MINLAKE. Proj. Rep. No. 263. St. Anthony
    Falls Hydraulic Lab., Univ. Minnesota.
Stefan  H.G. and  Ford, D.E. 1975a.  Temperature dynamics in
   dim'ictic lakes. J. Hydraulic Div. Am. Soc. Civil Eng. 101, HY:97-
   114.                                               .   .
	 1975b. Mixed layer depth and temperature dynamics in
   temperate lakes. Verh. Int. Verein.Limnol. 19:149-57.
Stefan H  G., J.J.  Cardoni, and A. Fu. 1982a. A dynamic water
   quality simulation program for a stratified shallow lake or reser-
   voir: Application to Lake Chicot, Arkansas. Proj. No. 209. St. An-
   thony Falls Hydraulic Lab., Univ. Minnesota.
Stefan H G., S. Dhamotharan, and F.R. Schiebe. 1982b. Tempera-
   ture/sediment model for a shallow lake. Pages 750-65 in Proc.
   Am Soc. Civil. Eng. Environ. Eng. Div.
Stefan H  G., J.J. Cardoni, F.R. Schiebe, and C.M.  Cooper. 1983.
   Model  o1 light penetration in a turbid  lake. Water Resour. Res.
   19:109-120.                                     ,      .
Winter, T.C. 1981. Uncertainties in estimating the water balance ot
   lakes. Water Resour. Bull 17:82-115.
 Windell, J.T. et al. 1986. An ecological characterization of Rocky
   Mountain  montane and subalpine  wetlands. Biolog. Rep.
   86(11). U.S. Fish Wildl.Serv.
 Young, R.A., C.A. Onstad, D.D. Bosch, and W.P. Anderson. 1987.
   AGNPS, Agricultural Nonpoint Source Pollution  Model Con-
    serv. Res. Rep. 35. U.S. Dep. Agric. Agric. Res. Serv.
                                                               37

-------
                                                          OFF-SITE ASSESSMENT WORKSHOP, 1989: 39-40
USDA Water  Quality Program
James Krider
U.S Department of Agriculture, Soil Conservation
Service, Room 5131-S, Box 2890, Washington, DC
20013

 Bruce Kirschner
 U.S. Environmental Protection Agency, P.O. Box 367,
 Portage, Indiana 46368
                                         ABSTRACT








            are expected to greatly increase the water quality expertise of SCS personnel.
  The U.S. Department of Agriculture (USDA) policies
  on water quality protection call for integration of sur-
  face and  groundwater protection and improvement
  activities into all programs and activities. Additional
  direction for USDA comes from the Water Quality Act
  of 1987, which states, "It is the national policy that
  programs for the control of nonpoint sources of pollu-
  tion be developed and implemented in an expeditious
  manner." The USDA National Program  for Soil and
  Water Conservation for 1988 through 1997 lists water
  quality protection as the number two priority. As a
  result of this increased emphasis on water quality, the
  Extension Service and the Soil Conservation Service
   (SCS) are improving the technical skills of their per-
   sonnel to enable them to provide greater technical
  and  educational water quality assistance  to land
   owners.
     The Extension Service is the federal partner in the
   Cooperative Extension System and the educational
   arm of USDA. SCS is the technical assistance arm of
   USDA: In cooperation with the conservation districts,
   it assists  landowners with water quality and soil
   erosion problems. Both agencies have offices in most
   of the counties of the United States. The Environmen-
tal Protection Agency, the United States Geological
Survey, the Agricultural Research Service, and many
other federal, state, and local agencies and groups
will be involved in the USDA water quality protection
program.
   In response to the need for action in regard to
protecting water quality, the SCS has initiated a Water
Quality Action Plan that is to be carried out at the na-
tional level. The USDA-Extension Service is a working
partner in the plan, especially as it relates to develop-
ment of guidelines for nutrient and pesticide manage-
ment.
   The plan is composed of three elements. One ele-
ment provides for developing the necessary technical
support for the effort. The nationwide scale must ac-
count for the wide diversity of farming methods and
 crops and variations in soils and other natural resour-
 ces. Another element involves structuring the evalua-
tion methodology. It calls for positioning the technical
 products in the conservation planning process. The
 third element consists of training for the SCS and
 other USDA personnel who will integrate water quality
 technical  assistance  and  education into  their
 programs.

-------
J. KRIDER AND B. KIRSCHNER
   One  of the more important requirements in im-
plementing the Water Quality Action Plan has been to
develop the necessary tools and training for integrat-
ing water quality into the resource conservation plan-
ning process. Specifically, SCS is providing nationally
developed products for use in Field Office Technical
Guides, along with guidelines for SCS offices in each
state that will allow for:
   1. Adaptation of national products for local condi-
tions
   2. Production of local technical guide information
   3. Aid in training local field office staff in implement-
ing the Water Quality Policy
   Water Quality Action Plan products that may be of
the most interest to lake and watershed managers in-
clude:
   1. In the General Resource References-The Ef-
fects of Conservation Practices on Water Quality and
Quantity;
   2. From the Soil and Site Information - Soils Rating
Related to the Movement of Pesticides; and
   3. From the Resource Management Systems Sec-
tion -The Guide to Selecting Conservation Practices.
   Although this information is considered to be state-
of-the-art and is much advanced over the information
previously available to SCS field offices, it is only the
first generation of water quality evaluation and treat-
ment techniques. Much like the  first generation of
computers, it is expected the Water Quality Act Plan
products will  be extensively refined in future years.
Accordingly, a much more valuable product should
evolve.
                                                40

-------
                                                      OFF-SITE ASSESSMENT WORKSHOP, 1989: 41-43
The  SCS  Water  Quality  Indicators Guide:
Surface  Waters-A Tool to  Assess  Surface
Water Quality Problems	
 Charles R.Terrell
 National Water Quality Specialist, Soil Conservation
 Service, P.O. Box2890, Washington, DC20013

 Patricia Perfetti
 Geoscience and Environmental Studies Department, University
 of Tennessee, Chattanooga, Tennessee 37403
                                       ABSTRACT






                                                E=^
                 eSthefieldsheeton animal wastejustasif an SCS District Conservationist were assessing a nonpomt
                                  ^^
            through the use of conservation and best management practices.
  Nonpoint source pollution is pollution of the nation's
  waters from diffuse sources,  such as agricultural
  fields and urban areas. In 1984, the U.S. Environmen-
  tal Protection Agency reported nonpoint source pol-
  lution as the leading type of pollution in 39 percent of
  the nation's rivers, 52 percent of the lakes, and 48 per-
  cent of the estuaries. Six of the 10 EPA regions named
  nonpoint source pollution as the principal cause of in-
  adequate water quality. In addition to surface waters,
   nonpoint source pollution also threatens ground-
   water quality. Agricultural activities contribute non-
   point  source  pollutants through  sediment from
   eroding lands, nutrients from excessively applied fer-
   tilizers, bacteria from animal wastes, pesticides from
   pest control operations, and salts from irrigation ac-
   tivities.
      For more than 50 years, the Soil Conservation Ser-
   vice (SCS) has provided technical assistance to land-
owners and groups who manage and use the nation's
soil, water, plant, and animal resources. More recent-
ly SCS has become actively involved in water quality
issues, involving both surface and ground waters. In
1983, SCS published the Water Quality Field Guide to
illustrate types of agricultural nonpoint source pollu-
tion. Additionally, SCS, with other federal/state agen-
cies and private groups, published the Baybook in
1987.  The  Baybook was  designed to address
Chesapeake Bay issues, but it has wider applications,
including helpful ways to work with urban nonpoint
source pollution problems.
   Another recently published SCS book, the Water
 Quality Indicators Guide: Surface Waters, is a com-
 panion volume to the Water Quality Field Guide. While
 the Field Guide discusses what agricultural nonpoint
 source pollution is, the Indicators Guide tells how to
 assess agricultural nonpoint source problems, and

-------
 C. R. TERRELL AND P. PERFETTI
 which conservation and best management practices
 (BMPs) can be used to remedy or reduce an agricul-
 turally related water quality problem.
   The SCS  Water Quality  Indicators Guide  is
 designed as an easy-to-use tool to assist field person-
 nel in evaluating agricultural water quality problems
 through the use of field sheets. The field sheets can be
 used to assess sediment, nutrient, animal waste, pes-
 ticide, and salt problems without performing chemi-
 cal testing or elaborate species identification. The
 guide was developed over the past six years. The field
 sheets were tested by the authors and by SCS water
 quality and other technical specialists in five states
 across the nation: Tennessee, Pennsylvania, Kansas,
 Minnesota and Colorado. The  guide  is general be-
 cause it covers many types of pollutants and environ-
 ments, but it has been  shown to be a good tool for
 performing a "first-cut"  assessment of water quality,
 even though further testing may be necessary in
 some situations. The ability to examine and recom-
 mend conservation or best management practices is
 an added advantage of the guide.
   Over the past two years, we made presentations
 about the Water Quality Indicators Guide to the Na-
 tional Science Teachers and National Association of
 Biology Teachers Conventions. The guide was given
 an excellent reception  by these educators, and we
 have had several requests to use the guide for water
 quality curriculum development, from elementary
 through college levels.  Dr.  Patricia Perfetti, Head of
 the Geoscience and Environmental Studies Depart-
 ment of the University of Tennessee at Chattanooga,
 has been instrumental in developing the WaterQuality
 Indicators  Guide,  having  written  much  of the
 manuscript.
   At this  point we will do a "practice run" of the
 guide's field sheet on animal waste. The field sheets
 are in Appendix F, at the rear of the guide. The animal
 waste field sheet is number 2A. We will complete the
 animal waste field sheet just as if we were making an
 assessment on a farm or a ranch. To do this, I will
 show some slides that illustrate real nonpoint source
 conditions. They have been assembled from various
 parts of the country, but they will represent a single,
 hypothetical nonpoint source situation in the Ap-
 palachian region.
   We will complete the field sheet just as if an SCS
 District Conservationist were assessing a nonpoint
 source problem. The objective is to learn how to use
 the field sheets, and to see how they can be used to
 assess nonpoint source pollution. Your job is to as-
 sess the situation, determine to what extent a non-
 point source pollution  situation  exists,  and then
 complete the field sheet. After the slides, you will have
time to complete the field sheet; then I will guide you
through it indicating appropriate answers. I  recom-
 mend that you work in a group of two or three. In this
 way you will be able to interact on the concepts and
 get a better discussion of the situation. At the con-
 clusion we will discuss how any problems detected
 could be remedied through the use of conservation
 and best management practices.
 1.
4.
          Field Sheet Scenario

     In your first week on the job as a Soil Conserva-
     tion Service District Conservationist in the North-
     east, you are visiting farmers from a list of names
     left by your predecessor. You drive to Farmer
     Hensen's dairy operation and immediately are
     struck by the odor even at an eighth-of-a-mile
     distance.

 2.   Mr. Hensen greets you and tells you he's been in
     the dairy business for 30 years, and his herd now
     numbers 85 animals. He gives you permission to
     tourthe farm on your own, which you do.

 3.   Starting with the barn, you find where the manure
     odor is coming from.  The barn is located in a
     stream-bottom area, where the land behind the
     barn slopes down on three sides.

     The barn siting means that manure removal is al-
     ways uphill, so Mr. Hensen takes the easier route,
     directly into the stream. Even with that, Mr. Hen-
     sen has  not always been attentive to manure
     removal.  Consequently, as you enter the barn
     area you are walking in manure four inches deep.

     The medium-sized stream is doing most of the
     manure removal. Cows are pastured nearby and
     have easy  access to the stream, trampling the
     banks, leading to erosion and sedimentation of
     the steam.

 6.   At the bottom of one of the hills is a farm pond.
     You inspect the pond and  see a  green scum
     completely covering it. It is obvious that the farm
     pond  receives considerable drainage from  the
     hillsides.

     The cows have extensively trampled the banks,
     preventing  vegetation from growing. Erosion of
     the banks and sedimentation in the pond are also
     readily evident.

     One side of the pond is so eroded that the pond's
    water  capacity is only two thirds of the design
    capacity.  Excess flow leaves the pond and has
    cut a channel below the pond.

9.  When you leave the farm you go to a side road
    and stop at a bridge downstream of the Hensen
5.
7.
8.
                                               42

-------
                                                                  OFF-SITE ASSESSMENT WORKSHOP, 1989: 41 -43
           farm. You see the stream is filled with floating
           manure and the bottom is coated with organic
           matter.
                          10.  Further downstream, at a recreational lake where
                              the stream enters, green slime covers the water,
                              and  people no longer swim  or boat  there
                              anymore. On the way back to the office you try to
                              assess in your mind what you have just seen.
Animal Waste
             FIELD SHEET 2A: ANIMAL WASTE        „„_,....
INDICATORS FOR RECEIVING WATERCOURSES AND WATER BODIES
Evaluator 	 	 . 	
Water Body Evaluated .
Rating Item
. 	 	 — — — • 	
	 , 	 • 	
1 Evidence of
animal
waste
visual and
olfactory
__ 	 	 • 	 	 —
2. Turbidity &
color
(observe in
slow water)
_~ 	 • 	 - — 	 	
3. Amount of
aquatic
vegetation
4 Fish behavior
in hot weather;
fish kills,
especially before
dawn
. 	 • 	
5. Bottom
dwelling
aquatic
organisms
County/State _ 	 . 	 uulu 	 	 ~
	 . 	 	 	 — 	 	 	 ' Total Score/ Rank 	 	
— " \A/-itnr nnHv I nration . 	 . 	 — ~~ ^~~~~ ~~~~ n_._
Excellent
	 	 	 	 	 	
(Circle one number among th
-- No manure in or near water
body
-- No odor.
•- OTHER
9 	
:-- Clear or slightly greenish
; water in pond or along the
: whole reach of stream.
:-- No noticeable colored film on
: submerged objects or rocks.
-- OTHER
9
-- Little vegetation— uncluttered
look to stream or pond
-- What you would expect for a
pristine water body in area.
-- Usually fairly low amounts
of many different kinds of
plants.
- OTHER
6
:-- No fish piping or aberrant
: behavior
:-- No fish kills
-- OTHER
8
-- Intolerant species occur
mayflies, stoneflies,
caddisfhes, water penny,
riffle beetle and a mix
of tolerants
-- High diversity
':-- OTHER
9

e four choices in each row which
a condition has characteristics of
_ 	 — 	 • 	 '
Occasional manure '•••
droppings where cattle :
cross or are in water. :--
- Slight musk odor. :
- OTHER
6
_ 	 •
- Occasionally turbid or
cloudy. Water stirred up &
muddy & brownish at animal
crossings.
Pond water greenish.
Rocks or submerged objects
covered with thin coating
of green, olive, or brown
build-up less than 5 mm
thick.
- OTHER
6
	 _ — 	 	
- Moderate amounts of
vegetation; or
- What you would expect for
the naturally occurring
site-specific conditions.
- OTHER
6
- In hot climates, occas-
sional fish piping or
gulping for air in ponds
just before dawn.
- No fish kills in last
two years.
- OTHER
5
- A mix of tolerants:
shrimp, damselflies,
dragonflies, black flies.
-- Intolerants rare.
-- Moderate diversity.
- OTHER
5
SL2SS*jSS"S?«°»
Manure droppings in concen-
trated localized areas.
Strong manure or ammonia
odor.
OTHER
2
- Stream & pond water bubbly,
brownish and cloudy where
muddied by animal use.
- Pea green color in ponds
when not stirred up by
animals.
Bottom covered w/green or
olive film. Rocks or sub-
merged objects coated with
heavy or filamentous build-
up 5-75 mm thick or long.
- OTHER
3
	 	
. Cluttered weedy conditions.
Vegetation sometimes
luxurious and green.
- Seasonal algal blooms.
- OTHER
3
- Fish piping common just
before dawn.
- Occasional fish kills.
%
- OTHER
3
-- Many tolerants (snails,
shrimp, damselflies.
dragonflies, black flies).
-- Mainly tolerants and some
very tolerants.
-- Intolerants rare.
- Reduced diversity with
occasional upsurges of
tolerants, e.g tube worms,
and chironomkls.
- OTHER
3
_ 	 	 	 • 	
the watercourse or
core ) 	 	
• Dry and wet manure all
over banks or in water.
Strong manure & ammonia
odor.
OTHER
0
Stream & pond water
brown to black,
occasionally with a
manure crust along banks.
Sluggish & standing
water— murky and bubbly
(foaming).
Ponds often bright green
or with brown /black
decaying algal mats.
OTHER
0
- Choked weedy conditions
or heavy algal blooms
or no vegetation at all.
- Dense masses of slimy
white, greyish green,
rusty brown or black
water molds common on
bottom.
- OTHER
0
- Pronounced fish piping.
- Pond fish kills common.
- Frequent stream fish
kills during spring thaw.
- Very tolerant species
(e.g., bullhead, cattish).
- OTHER
0
- Only tolerants or very
tolerants: midges,
craneflies, horseflies,
rat-tailed maggots, or
none at all.
— Very reduced diversity.
upsurges of very
tolerants common.
-- OTHER
0
— _ 	 — 	 TOTAL (
1 . Add the circled Rating Item scores to get a total for J^^ ****^ ..exce|,ent» i( tne score totals at least 35. Check "good" if the score falls between 21 and 34,
Sheet 28, or 2B-. m^noi-un I Fair (7-20) [ ] Poor (6 or less) |
RANKING1 Excellent (35-43) [ ] Good (21 -34) [ ] rair (^u, i
43

-------
                                                          OFF-SITE ASSESSMENT WORKSHOP, 1989: 45-50
Water  Quality of the Missouri River
John R. Rowland
John C. Ford
Missouri Department of Natural Resources, P.O. Box 176,
Jefferson City, Missouri
           Historic Perspective

 One of the first historical accounts of water pollution
 in the Missouri River was a U.S. Public Health Service
 publication of 1913. That report presented data on
 typhoid death rates for many Missouri River towns,
 and concluded that sewage pollution was at least par-
 tially responsible for an increase in typhoid deaths.
 This report was instrumental in pointing out the need
 for effective disinfection of drinking water supplies.
    During the next 40 years,  emphasis continued to
 be placed on bacterial contamination.  A  second
 USPHS publication in 1952 documented that a sub-
 stantial increase in the average number of bacteria in
 the river since the 1913 report.
     This increase in bacteria  contamination was un-
  derstandable considering human population growth
  and the expansion of the meat-packing industry. The
  1952 report contained  pictures of municipal, meat-
  packing, and stockyard wastes flowing untreated into
  the Missouri and its tributaries.
     Samples collected in 1950 at many places on the
   river showed that, indeed, coliform bacterial levels in
  the Missouri were always highest immediately below
   large towns or  stockyards. These levels gradually
   dropped but were still high by the time the water
   flowed by the next large town that withdrew water for
   public consumption.
     A second water quality  problem, low dissolved
   oxygen in the river water, was also noted in this study.
   During the summer, the acceptable minimum of 5 mil-
   ligrams per liter was not being  achieved at locations
   40 and 150 miles below Kansas City.
     In addition to bacterial contamination, a  1958
   report by Kittrell mentioned the problem of grease,
   which during  periods of rapid water temperature
   decrease, tended to coat and clog intake screens at
   the Omaha water treatment plant.
     The  problem of nonpoint pollution sources was
   highlighted by a fish kill in the river in the Kansas City
area in May 1964. Widespread heavy thunderstorms
washed large amounts of organic materials into the
river or resuspended them from river bottom deposits
as river  stages and flow velocities increased. The
oxygen-demanding wastes in the river at the time
were estimated to equal the discharge from 50 million
people. Dissolved oxygen levels as low as 1.5 mg/L
were measured below Kansas City a few days after
the fish  kill, and there was some speculation that a
portion   of  the  river  had  been   completely
deoxygenated. Occurring at a time when local, state,
and federal governments were wrestling with the
problems of point source discharges,  this incident
revealed that serious water pollution control efforts
on the Missouri  would require more than the con-
 struction and operation of facilities to treat municipal
 and industrial wastes.
   The mid-1960s marked the completion of the main
 stem reservoir system as we know it today. The Fort
 Peck Dam in Montana was  completed in 1937 and
 was filled  by 1942. The other five dams, Garrison
 (North  Dakota), Oahe,  Big Bend, and Ft. Randall
 (South Dakota), and Gavins Point (South Dakota-
 Nebraska border), were completed between  1953
 and 1964.
    The most obvious and farthest-reaching impact of
 those six impoundments upon the lower Missouri is
 the reduction in suspended sediment concentrations.
 This reduction is greatest just below these impound-
  ments, but is still significant at the mouth of the Mis-
  souri,  over 800 miles  downstream.  For example,
  annual suspended sediment loads at Yankton, South
  Dakota, immediately below Gavins Point Dam, have
  been reduced to 1 percent of what they were prior to
  construction of the five lower reservoirs. At Omaha,
  approximately  200  miles  below  Gavins Point,
  suspended sediment loads have only 14 percent of
  the levels prior to impoundment, and at Hermann,
  Missouri, over  700  miles below Gavings Point, the
  present suspended  sediment load is about one third
  of what it was prior to impoundment.
                                                 45

-------
 J. R. HOWLAND AND J. C. FORD
  Water Quality Trends and Existing
               Water Quality

 Bacterial Contamination

 Figure 1 plots the increase in total coliform levels that
 have been  measured  since the 1920s. This trend
 reflects the earlier-mentioned increase in population
 and wastewater volumes. There is some indication of
 lower levels during recent years. Similar time trends
 can probably be seen for most sections of the river,
 with the exception of major metropolitan areas where
 disinfection is practiced.
   Total coliform  levels fluctuate greatly as  water
 flows through the  lower Missouri. Figure 2 shows that
 very high levels  occur immediately below  cities
 and/or large stockyards, followed by a general die-
 off, then another peak at the next major source.
   Fecal coliform  concentrations, an indication  of
 fecal contamination of water, followed a similar spatial
 trend in the river. The first extensive measurements of
 fecal coliform were  made  during surveys in  1968-
 1969; however, regular  monthly monitoring of fecal
 coliform levels did not begin until 1972. Figure 3 sum-
 marizes much of the available fecal coliform data on
 the river.
     50
       Figure 1 .—Total coliform trends in the Missouri River at the
       Kansas  Cities  and  St. Joseph  (from  USPHS, 1952;
       Burkholder, 1981; and Mines, 1980).
g-y
il
°i
= 2 20'
     12-
     4-
       Figure 2.—Annual median coliform bacterial levels in the
       lower Missouri River, 1950.
                    •	 USGS 1972-79
                          (27-60 samples/sta.)

                    •-  Fall 1968 (8 samples sta.)


                    ••••  Winter 1968-69
                          (ave. 13 samples/sta.)
                                     T
                          600
400
—I—
 200
                                                  River Mile

Figure 3.—Fecal coliform bacteria concentrations in the Missouri River.
                                                46

-------
                                                            OFF-SITE ASSESSMENT WORKSHOP, 1989: 45-50
  Seasonal trends in fecal coliform are not easy to in-
terpret. Monthly geometric means at St. Joseph are
shown in Figure 4. Late winter concentrations are low,
with sizable increases in spring and early summer cor-
responding to periods of maximum surface runoff and
erosion. Elevated levels of fecal coliform in late fall and
early  winter do  not correspond to periods  of  in-
creased flow, high suspended solids, high nitrogen
levels, or low dissolved  oxygen.  This suggests that
nonpoint runoff pollution is not totally responsible for
these high bacterial levels. Although Sioux City and
Council Bluff, Iowa, discontinues chlorination of was-
tewater effluents in October, these cities are believed
to be too far upstream to affect bacterial densities at
St. Joseph.  The increase may be due to increased
coliform survival rates in the river.
   It is difficult to discuss time trends for the fecal
 coliform data, since few exist. The only change that
 appears to  be significant between the 1968-69 data
 and the 1972-79 data  is the  lower levels of fecal
 coliform present within the Sioux City to Omaha area.
 These are probably caused by chlorination at Sioux
 City  and  improved wastewater  management  at
 Dakota City, Nebraska.
   9000-
 3
 g 6OOO-
 o
 3


   3000
               MAR.       JUNE      SEPT.
                          Monlti
  Figure 4.—-Seasonal mean coliform concentrations in Mis-
  souri River at St. Joseph (USGS1972-79).
tributaries such as the Kansas and Osage Rivers have
also helped reduce solids and turbidity on the Mis-
souri.
   Table  1 compares annual suspended sediment
load in the lower Missouri before and after impound-
ment (U.S. Army Corps Eng. 1948, 1957, 1965, 1970,
1972,1976). Prior to 1953, Fort Peck, in Montana, was
the only main stem reservoir on the river; but by 1955,
Fort Randall,  Garrison, and Gavins  Point were all in
operation.

  Table 1.—Average annual suspended sediment load
  of the lower Missouri River (in million tons).	
  LOCATION
                                       SEDIMENT
                APPROXIMATE AVER. ANNUAL   LOAD
                 RIVER MILE  PRIOR TO 1953 AFTER 1955
Yankton, SD
Sioux City, IA
Omaha, NE
Nebraska City, NE
St. Joseph, MO
Kansas City, MO
Boonville, MO
Hermann, MO
811
732
626
562
452
360
180
100
1378

163.8

257.2
328.0
350.0
326.2
1.4
11.8
28.6
47.1
57.6
79.0

100.8
    Although reductions in suspended sediment are
 most spectacular immediately below the Gavins Point
 Dam, threefold reductions occur over 700 miles below
 Gavins Point, showing that the nature of the entire
 lower Missouri has changed. Not only has there been
 a long-term  and rather abrupt  trend  of  reduced
 suspended sediments, but  more short-term influen-
 ces can also significantly affect suspended sediment
 levels.
    It is therefore  apparent  that the  main stem im-
  poundments have  not only  been  instrumental  in
  reducing suspended sediments throughout the lower
  Missouri, but that in the first 100 miles below Gavins
  Point, they have also eliminated most of the normal
  variability in suspended sediment concentration.
    Turbidity of the river water was also reduced after
  impoundment. At the Kansas City, Kansas, water  in-
  take, turbidities ranged from 1300 to 3200 ppm prior
  to impoundment, and dropped to 470 to 800 ppm after
  impoundment, a decrease of 65 percent. At St. Louis,
  turbidities after impoundment decreased by more
  than 50 percent.
       Suspended Solids, Turbidity

   Sizable reductions in the amount of suspended solids
   and turbidity in the lower Missouri resulted from im-
   poundment of the main stem middle Missouri in North
   and South Dakota (U.S. Army Corps Eng. 1948,1957,
   1965,  1970,  1972, 1976).  Impoundments on major
     Total Dissolved Solids, Sulfates

   Because of its large area, there are considerable
   climatic differences within the Missouri basin. Of par-
   ticular importance are the lesser amounts of annual
   rainfall in the western part of the basin. This pattern of
   rainfall increases the relative importance of evapora-
                                                   47

-------
 J. R. HOWLAND AND J. C. FORD
 tion in the western and northern parts of the basin,
 and therefore, the total dissolved solids (IDS) con-
 centration. As  the Missouri flows east and  south,
 tributaries with lower IDS dilute its water. Thus, while
 the Missouri River in central Missouri has an average
 total dissolved  solids (TDS) concentration of about
 420 mg/L, the concentration in its major tributary in
 that location, the Grand River, is only 250 mg/L.
   A similar situation exists for sulfate  ion. Figure 5
 shows the magnitude  of downstream decreases in
 sulfates  (approximately 25 percent) and total  dis-
 solved solids (approximately 15 percent).
I
5
I
o
O
  200-
         800
                   600         400

                       Hlvtr Mil*
                                        200
Figure 5.—Average concentrations of total dissolved solids
and sulfate inthe lower Missouri River (USGS1974-79).
                                                     cent of the time) there is no clear relationship between
                                                     flow in the lower Missouri and sulfate concentration.
                                                     Since there is considerable variation in sulfate levels,
                                                     however, their variability must be due to tributary in-
                                                     flow. This assumption proves correct. Figure 6 shows
                                                     that sulfate in the Missouri has a much stronger
                                                     relationship to flow on the Nodaway River, a tributary
                                                     of the Missouri above St. Joe, than to flow in the Mis-
                                                     souri itself.
                                                    •io « (1000 cl.)
                                                     •I SI Jot.pt,
                                                    Nodftwcp ft (ell) r
                                                     nr Orvgon   0
                                                     Figure 6.—The relationship of sulfate in the Missouri River at
                                                     St. Joseph to flows in the Missouri and Nodaway Rivers.
   Trend analysis by linear regression and t-test for
significance of slope were performed on all flows at
St. Joseph. No significant trends over time were
found. Similar analysis was made on monthly average
TDS and sulfate levels from 1950 to 1979 on the Mis-
souri near St. Louis. No significant time trend was
detected for TDS on these data either. A slight but
statistically significant trend of sulfate concentrations
increasing with time was noted.
   Shorter-term trends are also apparent. The follow-
ing discussion of short-term variability  of sulfate
should also apply for total dissolved solids.
   Shorter-term variation in sulfate concentration in
rivers typically has an inverse relationship with dis-
charge, higher flows causing lower concentration of
sulfates, and vice versa. The lower Missouri, however,
presents a rather special case because of the main
stream impoundments. Most of the flow in the lower
Missouri comes from  Gavins Point Dam,  behind
which is a large body of water with  rather constant
concentration of sulfate. Therefore, increasing  the
released flow from Gavins Point Dam  by several
thousand cfs could make very significant changes in
the flow in the Missouri without changing the sulfate
concentration.
   Actual observation of the relationship between dis-
charge and sulfate bears out this point.  At flows of
less than 70,000 cfs at St. Joe (which occur 90 per-
                                                                       Nitrogen

                                                     Although not a major constituent of Missouri River
                                                     water, nitrogen is an important nutrient to both plants
                                                     and animals in the river system. Agricultural fertilizers
                                                     and discharges of organic wastes such as domestic
                                                     sewage, stockyard, and meat  packing wastes are
                                                     major contributors of nitrogen to the river.
                                                       Nitrogen is most commonly found in one of three
                                                     forms within the river: organic N, ammonia, or nitrate.
                                                     Organic nitrogen is nitrogen contained in plant and
                                                     animal  material  (living and dead) and  animal and
                                                     human wastes. Ammonia NH3 is produced by the
                                                     bacterial decomposition of organic material. A con-
                                                     ventional secondary wastewater treatment plant will
                                                     contain almost the entire process of decomposition
                                                     from organic nitrogen to ammonia within the plant. A
                                                     third form of  nitrogen commonly found in the river is
                                                     nitrate  nitrogen   (N03).  This  oxidized  form  is
                                                     produced from ammonia by bacteria in the presence
                                                     of oxygen. Unless a wastewater treatment plant has a
                                                     special unit process for nitrogen (oxidizing of am-
                                                     monia), the nitrogen discharge from the plant should
                                                     be mostly ammonia and a small amount of nitrite plus
                                                     nitrate  N. The remaining  nitrification takes  place
                                                     within the river, and some of the dissolved oxygen in
                                                     the river is used to form the nitrites and nitrates.
                                                 48

-------
                                                             OFF-SITE ASSESSMENT WORKSHOP, 1989: 45-50
  Figure 7 shows the distribution of nitrogen among
the three major forms, and how nitrogen concentra-
tions change in various segments of the river. Due to
settling of sediment-bound nitrogen and algal uptake,
releases of impounded water from Gavins Point Dam
are relatively low in nitrogen, but nitrogen increases
dramatically between Sioux City and St. Joseph. At St.
Joseph,  average total nitrogen (organic plus am-
monia plus nitrate) is five times higher than at Gavins
Point. Although point source discharges such as
municipal sewage  treatment plants,  meat  packing
plants, and waste treatment facilities  at stockyards
are obvious contributors of nitrogen to the  Missouri
between Sioux City and St. Joseph, the predominant
source is associated with agricultural activities in the
 Corn Belt.
  8-
  4-
                 50            100
                  Discharge (1000 cts)
Figure 8.-The relationship of total nitrogen and discharge in
the Missouri River at St. Joseph, MO.
  Figure 7 —Average concentration of the three major forms of
  nitrogen in the lower Missouri River (USGS1974-79).

     The influence of nonpoint sources becomes very
  apparent when nitrogen values are compared with in-
  creases in discharge. Figure 9 shows the relationship
  between total nitrogen and the number of days follow-
  ing a hydrographic event. The data points appearto fit
  a linear or an exponential model equally well. Figure 8
  shows the relationship between total nitrogen and dis-
  charge, and suggests an  exponential relationship.
  Both these trends indicate that nonpoint runoff does
  increase nitrogen concentrations in the Missouri, and
  the slope of the regression lines suggests nonpoint
   sources are the major source of nitrogen during the
  times the river is receiving runoff.
      Seasonal trends  in nitrogen concentrations are
   also evident, as shown in Figure 9. For the Missouri at
   St.  Joe, the four months of highest nitrate nitrogen
   and organic nitrogen are the same: March, April, May,
   and June. The same observations hold true over a 30-
   year-period on the Missouri near St. Louis. This period
   corresponds to the interval between spring thaw and
   the development of good  vegetative cover on tilled
   land, the period when the land is most susceptible to
   erosion.
                                                       4-
                                                     I
                                                     — 2*
                   NO3-N
           T—i	r
              MAR.
T	1—I	1	1	i
  JUNE      SEPT.
  Month
  Figure 9.—Seasonal trends in nitrogen In the Missouri River at
  St. Joseph (USGS 1972-79).
              Dissolved Oxygen

  All four states bordering the lower Missouri have es-
  tablished a minimum dissolved oxygen standard of 5
  mg/Lforthe protection offish and otheraquatic life re-
  quiring oxygen in the water. There is abundant dis-
  solved oxygen in the upper part of lower Missouri
  throughout the year, and in the entire lower river from
  late autumn to early spring. During the warm weather
  months,  the  increasing oxygen demand  of organic
  materials lowers the dissolved oxygen in the river as
  the water flows downstream. Figure 10 plots average
  saturation  values for oxygen for various sections of

-------
 J. R. HOWLAND AND J. C. FORD
100
fl
6 • *°'
| 40-
£
20-
0-
10-
8-
| ,
6
Q
2-

-------
Closing  Remarks
Douglas A. Ehorn
Deputy Chief, Water Quality Branch, U.S. Environmental
Protection Agency, Region V,230 South Dearborn Street,
Chicago, Illinois 60604
                                                          OFF-SITE ASSESSMENT WORKSHOP, 1989: 51-53
                                                          © 1989 North American Lake Management Society
 Today, more than ever, there is a need to address the
 nonpoint source pollution control program in an effec-
 tive manner. Our speakers have provided us with a
 picture of the world as it exists today in relation to the
 challenge of a very complex issue-measuring and
 controlling nonpoint sources of pollution. It would be
 too simple to say that a cause-and-effect relationship
 exists between human activities and the pollution of
 our waterways and the loss of the world's richest soils.
 Our speakers have presented dramatic evidence of
 that fact. They have clearly demonstrated that there is
 a relationship between human activity, not only at the
 original site, but off-site, as well. The costs and  the
 consequences of  off-site impacts largely have  not
 been anticipated or measured, so little information is
 available to provide us with the institutional (society)
 costs. As  I have reviewed the materials presented
 today, I  am inclined to make five observations that I
 believe  are fairly representative of these presenta-
 tions.
testified to that truth. A quick examination of their work
is in order.
   Four years ago, as I was doing a similar wrap-up of
a conference in Chicago, I noted a lot of confusion in
terminology between the agricultural community and
the water pollution control community. Today, by con-
trast,  no speakers had to spend time defining their
terms and excusing their vocabulary. The point is that
we have a broader perspective of each other's ac-
tivities and roles. We have the knowledge that it takes
to tackle the tough problems. For instance, we heard
discussions  on some  very  sophisticated topics of
sediment control, erosion potential, best manage-
ment practices  (or resource management systems),
riparian measures, and remote sensing techniques.
We have heard  that on the state level, there is broad
cooperation in dealing with very complex situations,
and general agreement by the local groups that work
 needs to be done for a variety of reasons; not the least
 of which are economic and human health-related.
   Jointly We Have The Knowledge To
                 Do The Work

  Nonpoint source control programs are not new to any
  of us. The ideas have been with us for a long time and
  have been advanced  by many clear-thinking  in-
  dividuals over the past three to four decades. Aldo
  Leopold has made one of the clearest presentations
  of the major themes in his writings, now called A Sand
  County Almanac. In 1940, Mr. Leopold dramatically
  provided an awareness of the problem of off-site im-
  pacts, as he told us that the gully is the enemy of the
  plover; and perhaps we will find it is our enemy, too.
  With that knowledge, many people have taken  on the
  work of conservation, using a variety of names and
   modes  of operation. Our knowledge of the world
   around  us has grown rapidly, and our  speakers have
  There Are Some Stumbling Blocks

 Our speakers have established that, although there is
 great progress on a number of fronts, certain stum-
 bling blocks continue to impede progress. Some of
 those are politics, lack of funding, need for authority to
 proceed, need for direction, need for a consensus,
 need to eliminate counterproductive laws, and a need
 for a whole solution.  In fact, there was even the al-
 lusion to the fact that we cannot make progress be-
 cause we do not know how to measure the current
 situation or the progress. All this may be true! What is
 important to recognize is what you and I intend to do
 about some of these things. We have a choice. And if
 we have a choice, then the real stumbling block may
  be only our own attitudes.
                                                 51

-------
D. A. EHORN
 The Work Is About Half Completed

I stated that we have come a long way in recognizing
the problems and issues that face us. And I believe
that is really true. But this is not a time to let up or slow
down. We have the knowledge that will allow us to be
successful if we desire to be successful. How do we
become successful, in light of the stumbling blocks
that I have just listed? I have two major themes here.
   First, every day you have the abilityto make a num-
ber of changes in attitudes and program approaches.
You, as managers, are constantly bombarded with
good ideas and concepts from your staffs. Please
keep in mind that if you analyze the situation, over 65
percent of the ideas and concepts that are brought to
your attention over the year can be implemented if
you alone will take an action. You do not need new
authority. You do not need new legislation. You do not
need new money. You do not need new staff. Well,
what does it take? It only takes two things. A little risk
on your behalf to say, "Let's take a little deviation from
the norm and create some excitement." And second-
ly, you must authorize or empower both you and your
staff to do  something.  That  may seem a little too
simple. But I strongly  urge you  to try a few items
where you perceive the risk to be low, just to prove to
yourself that it can be done; and by the way, you will
find both self-satisfaction and external gratitude for
your decisions.
   Second, enough money is currently available to in-
itiate the work. The task then is to hunt down the avail-
able resources and convince the people who control
the dollars and staff time that your project is a unique
opportunity,  and that results of the work will benefit
the geographic area where it is conducted and meet
the needs of the party providing the desired support.
By example, it will also help provide the foundation for
work elsewhere. There is a great deal  of evidence to
support this premise. In today's presentations, we
heard about the joint, cross media efforts of state and
local  governments  working  together  to  resolve
problems. In the North American Lake Management
Conference that follows today's workshop, there will
be more evidence. However, I could specifically men-
tion three or four projects that  occurred  because
either the local people worked  to get the federal
government's attention, or the federal government
worked to get local and state attention to solve issues
related to erosion, sedimentation, clean lakes, and
wetlands. The rule of thumb now tends to be that a
project requires the skills, knowledge,  and abilities of
a great number of programs to be fully successful. It
used to be an exception if we found people working
together; I, for one, am very glad that the situation has
turned around.
   Five years ago, the people in the agricultural com-
munity said that they could not get involved in water
quality aspects of on-ground programs. Today we
have seen,  in almost every presentation, that water
quality is factored into the projects. This is a very sig-
nificant turnaround. Now it is time to put the pressure
on the water pollution control agencies to ensure that
each project set in motion has a definitive erosion
control program, and measures that aid the agricul-
ture  community in implementing their mandates.
Both sides are recognizing the differences in the con-
trols necessary to address the rural and urban com-
ponents of nonpoint sources of pollution. Watershed
management approaches are beginning to appear at
all levels. Today we heard about environmental  cor-
ridors,  the role of wetlands in nonpoint source con-
trol, use of open space, terraces, and other resource
management systems. There seem to be fewer turf is-
sues.
   But  remember, even in this area, the work  is  only
half done. We need to further reduce the barriers by
eliminating  the conflicting and  contradictory  pro-
grams that each of us manage. We need to parley our
resources with others to provide solutions. We need
to eliminate the names that earlier created the turf bat-
tles to begin with. We can no longer be soil scientists,
environmental  engineers,  biologists, lawyers,  poli-
ticians; we must be CONSERVATIONISTS. We must
subscribe to the concept  of the Conservation Ethic
that was laid out for us by Aldo Leopold. Mr. Leopold
defined conservation as ".... the state of harmony
between man and land."
   Using  Mr. Leopold's  advice  and our current
knowledge  of  control systems,  we can recognize
where the potential for problems exists, and where we
should be using our precious resources. We  may
have to do this in two ways. The first is to address the
current worst case problems, because Congress ex-
pects that much. But also, in our daily efforts, we need
to find time and effort to prevent future problems, and
tell the  people in authority what is cost-effective. That
requires some innovation and demon-stration.
  We need to educate not only the public about the
associated costs of pollution control, but the cost dif-
ferential of not allowing the problem to occur. Make
sure that there is a matter of choice. We have all been
educating ourselves and that is significant. After all, it
brought us  to this  point of time. But the fact  is that,
since problems are still arising, we need to carefully
examine  a  new training agenda that includes the
rulemakers, the taxing authorities, and the voters in a
way that has not been accomplished to date. Only
after you put forward the issues in economic, human
health,  and aesthetic  terms,  will there  be  some
modification of behavior.
                                               52

-------
                                                          OFF-SITE ASSESSMENT WORKSHOP, 1989: 51-53
A Summary Of Behavioral Changes
   To Revolutionize The Situation

1.  The guard at the federal level is
    changing.
We have all been waiting for a number of years for that
big influx of dollars and other resources into the Non-
point Source Program. And now that we have legisla-
tion, our expectations have  grown even  higher.
Therefore, we could have a greater disappointment if
we do not get a grasp on the reality of the situation.
There will never be "enough money" to do all the work
that our imaginative minds will dream up. The non-
point source  control program  may  never be  a
regulatory program at the federal level. Therefore, it is
incumbent on the states and the local governments to
 put  together  programs  to  address  the  basic
 geographic issues facing them. There must be a firm
 will to force environmentally sound, economically
 feasible solutions.  For the local governments,  you
 must continue to discover innovative ways to con-
 vince the federal and state governments to support
 you. You are doing a good job, and don't let anyone
 tell you otherwise.

 2.  Be persistent in your efforts.
 Since there will never be enough money or staff, or
 data to support everything you  might want to ac-
 complish on a short-term basis, build your programs
 so that they will last through the economic droughts.
 You can find a cul-de-sac of sanity by finding a level of
  effort that is reproducible, sustainable, and measur-
  able, that will provide a level of satisfaction. Then have
  your contingency plans ready, just in case the miracle
  of funds ever occurs.

  3.   Do not count on the federal
      government to bail you out.
  Aldo Leopold said that there is tendency in American
  conservation to relegate to government all necessary
  jobs that private landowners fail to perform. A lot of
  people would rather wait for the federal or state
  government to come in and fix the problem. That is a
  very  short-sighted approach, and if you look at the
  history of the nonpoint  source efforts,  may not even
  be a solution at all. Some people are waiting for direc-
  tion on how to proceed. There are a couple of good
  reasons why they are waiting. First, they do not have
  money, or they do not want to perform activities that
  will preclude any future funding coming their  way.
  Others just simply do not knowwhattodo. There must
  be strategies to address these situations. Some solu-
  tions are easy. You, as responsible partners in conser-
  vation, can train local people to do the right thing. We
do that every day in the classrooms, and we need to
do more of the education right in the field. Further, we
have to keep the local efforts in mind, and ensure that
appropriate credit is given when other funds are avail-
able.
   Each of us has to be accountable for our hours of
work. We need to be responsibl e to the public who we
guaranteed we would assist when we took our posi-
tions. We  need to encourage the efforts of others
when we see something going in the right direction.
   We can no longer be agricultural agents, farmers,
consultants, lawyers, biologists, engineers. We must
be conservationists who are willing to put in the extra
time and effort to control pollution and improve the
environment.

 4.  We need a way to educate the
     legislatures and Congress.
 The truth is that we are working to balance the budget,
 and there will  be some reluctance  to fund  new
 programs. So what do we do? We need to ask the rule
 makers to make intelligent decisions, based upon
 good solid data that show what the cost-effective ap-
 proaches are to nonpoint source pollution control. We
 have to show them that with some money from the
 federal and state budgets, a lot of work will be done.
 We have to be able to prove that the locals cannot to-
 tally foot the biil, and that the cost should be spread
 nationally to people who benefit (in terms of food and
 fiber) from areas where pollution control is needed.
 We need  to demonstrate where there are counter-
  productive programs that should be eliminated.

  5.  Be prepared.
  I assist in a Cub Scout Den, and I have been reedu-
  cated and realigned to the simplicity of these words. I
  was told there would not be a Webelo Den if I did not
  do it. Well, I had a choice.
    I am telling you that the nonpoint source program is
  on the line every day of the year, and it needs YOU.
  You have made a difference so far, and you must con-
  tinue to do that every day.
    You must  be willing to travel to the field. You must
  be ready to make some innovative  decisions. You
  must empower staff. You must take a risk. I used to say
  that there are a lot of vehicles going in the right direc-
  tion; simply get one that you feel comfortable with and
  go get the work done.
     Aldo  Leopold expressed his opinion of the com-
  munity in the following thought: "..the individual is a
  member of a community of interdependent parts. His
  instincts prompt him to compete for his place  in that
   community,  but his ethics prompt  him  also  to
   cooperate."
     We all have choices every day. Make yours count.
                                                 53

-------

-------

-------