United States
Environmental Protection
Agency
Office Of Water
(WH-553)
EPA841-F-92-012
December 1992
Number 7
&EPA TM.DL Case Study
Sycamore Creek. Michigan
Key Feature:
Project Name:
• Location:
Scope/Size:
Land Type:
Type of Activity:
Pollutant(s):
TMDL Development:
Data Sources:
Data Mechanisms:
Monitoring Plan:
Control Measures:
A watershed analysis that links
dissolved oxygen problems to
sediment loads and establishes NFS
load allocations
Sycamore Creek
USEPA Region V/Ingham County,
Michigan
Watershed area 274 km2;
subwatershed area 96 km2
Irregular plains
Agriculture
Sediment
NFS
State and local
DO model, NPS loading model
Yes
BMPs
Sycamore Creek
FIGURE 1. Location of the Sycamore Creek Watershed
in Michigan
Summary: Sycamore Creek (Figure 1) was targeted
for intensive watershed analysis because its water quality
problems are representative of many streams that drain primarily agricultural land in southern Michigan. It is listed on
Michigan's §303(d) list Sediment is the pollutant most responsible for impairment of Sycamore Creek. It has destroyed
aquatic habitat, and dissolved oxygen (DO) modeling results have indicated that sediment oxygen demand (SOD) is the
most significant oxygen sink under drought conditions. Model simulations have also shown that respiration by aquatic
plants contributes significantly to the DO deficit at some locations in the Creek. Aquatic plants depend on available
nutrients to grow, and since most nutrients are transported to Sycamore Creek while adsorbed to suspended sediment,
reducing sediment loadings will address this problem. Instream monitoring supported these conclusions by revealing State
water quality standard violations for dissolved oxygen (DO) at seven of eight locations in Sycamore Creek.
Michigan's Department of Natural Resources (MDNR) believes that reducing suspended solids loadings to Sycamore Creek
is the best overall strategy for increasing DO concentrations in the creek to meet the DO standard and improve aquatic
habitat Less sediment in the Creek will improve fish and macroinvertebrate habitat; provide a firmer stream bottom that is
more appealing for recreation; deepen the channel, thereby improving navigation potential; and increase oxygen
concentrations by reducing SOD and aquatic plant respiration. The first step in this direction was to estimate annual
average sediment loading to the stream from urban runoff, stream bank erosion, agricultural fields, septic tanks, and point
sources using modeling, channel surveying, and monitoring. Stream bank erosion, agricultural erosion, and urban runoff
were all significant sediment sources. Analyses indicated that suspended solids loading would have to be reduced by 52
percent in order to reduce DO levels sufficiently to meet the standard at all locations (except downstream of the marsh)
during drought flow. MDNR has no final plan on how to achieve the necessary reductions; however, one possible
allocation scheme that is reasonable and can achieve the 52 percent reduction is presented.
Contact: John D.' Suppnick, Michigan Department of Natural Resources, Surface Water Quality Division, RO. -Box
30028, Lansing, MI 48909, phone (517)335-4192 ' ""
Printed on Re cycled Paper
-------
; I,:;!; !;:;•:;;; ^.B^KgRjpU^
drou
•camore Creek is a small warm-water stream that
LjUIIIIIIIIIIBI''1™!!1!!''''!!*!!!!!!1!1! K< !Vl;;T!!iin^
,274 square ^°?
,_t flow is only 1.3 cubic feet per second at the
pFwastewater 'dl^criarge" so""that downstream1 from
t|^!!lliilli»!MK^ '.Jlin-rfU'iJ*! "Mmi". f'l-ihil*!**'», »' I1*!"1 at «• li * / Ml! "Ma? 11» 'it»'W W f" if?' 1*1
s point the stream is considered effluent-dominated.
f if! fifgffll ffet, Jaj«yn. jty£ .rniddle
(Figure ,1).
Sev
"effoi
1 in iihiiliillllliWiWIi'fcftBijriiiiMfriailii, i*',,1" .iiu:!Gi>"ii! il:
,',:,!|i"l!!!l,i:.!!!lllii/lli:, ,iiil'l'!,.. S ,.»'i'J!' 'iBUlM!: II! il",:. ii'Ml I1'"* !J1:
.
f^^p^i focuses .on the 96 ..... square kilometers upstream of
gj jl^r'T^. |St aiper i Rpad™(Figure'" 2). ^f^^^^w^^jsT'se^^S "
..... Ilfllsi'iioSy"*^^
expanded by__dredgmg. Most of the steam
Federal and State agencies are coordinating their
In the Sycamore Creek watershed to improve its
'protect designated uses. Sycamore
' Water Quality Standards for
other indigenous aquatic
ingl,^ witjim.ie;, watershed have been dredged at least
""" 'ifiiS ,?'>;;jr,«! ;i-i)l-: s SiJfflSllWHBWWySBlW'wSii,'; •. i m i
CjOgS^ stick _asi corn, soy_ beans, wheat, and .alfalfa.
iJhQugh t|ey do not ^pically plow in the fall,
^^j^^^t^!j^« |s.'i^^^Wj2E]^u^3^^5e^te
agricultural land is highly credible.
rid wildlife, total body contact recreation, and
'Michigan's" United "S'tates ' Department
naB««««imi«i«nwBi^h«>iwn>»«>n«>«A<»!niri<_r>w-ii
culture (USDA) water quality nonpomt source
U)UaWAIW)dMk^UUUk«-
'tf^^Tfoe watershed, is, dpminajed,^ loam and, sandy loam
I^T. soils, tjut some organic soils are scattered about.
'r'| ;"' i : ;.;The City of Mason is located near the downstream end_
i'.,| -.'J... ,j , 's of Sycamore Creek. The city_ has no major industries
J' j, ":'' ,;|' '"''-that (fis,cjharge process wastewater to the stream, but it
does have ^ municipal plant that provides advanced
'I1;1 i* ! „ ,! !'' "I:'1'1!, ^..ii' .l.'Ji,11?'.L r1'^'"1""1 "**'' ' ':". '''"ll '!" 'i!!1'1!."'"'^'11 "'' 'l|1'"*''' MNINI.IIIN™.I»IUJIM!'.II,,,III|I[I,II UhiBiKiPii^
hydrologic unit project, the creek benefits from an
intensive e^ucatiorial program, technical assistance, and
cosTs baring to Implement best management practices
(BMPp) within the" watershed In addition^ the Ingham
Soil Cfoiiservatfon DistrictHas"a"|255"(J) grant from
to provide technical assistance to farmers in the
Agricultural Stabilization and
$3QPiOOO .^ SpgCj^j
Sgriciiltiiral Conservation Program cost share money for
5rcanigre_Q-eelq and,the Ingham County Department of
rub"lK| jj^jjj ha^ a g'319 grant from USEPA to study
grouriawater in the watershed.
••I- S' ^asiesvaicr' teatment |pr a|"pl3pu!a5on"oF 6019""" The^
,S''fo'"fifS' 1"3'jJiji^11^^!^!^ per'day! '"Sycamore Creek's"
and Prioritizing
iljllf!' lil'ili!!1'!!!!!''''!'''!'''!'!!!!:!!!1''!!!!?''!!''"
••'-14' ;-:". ,^''• • • ''•• \'':-'3BIGURE| 2. i 'Sycamore^ Creek"and Vicinity'
,,'_"_.! ... ..... . ^ " ',. " .''"!!" "',"! U'"'i-,"~i'"'."','i T-*''!•'•'>"".',
Sycaniore Creek, which is listed oh Michigan's §303(d)
h'st, was targeted for intensive watershed analysis
because its water quality problems are representative of
many streams that drain primarily agricultural land in
southern Michigan.' Feedback on the success of NFS
management measures in this watershed can therefore be
applieii to similar streams throughout the region.
Monitoring
Biological surveys (Clark, 1990), conducted to help
characterize problems in the Sycamore Creek watershed
and to serve as a baseline for documentation of future
improvement, revealed that intolerant fish species were
absent; and that macroinvertebrate diversity and
abundance in Sycamore Creek were low. This evidence
that the creek's aquatic community is stressed and
unhealthy, that designated uses are unpaired, and that the
DO standard is being violated, was supported by channel
surveys, continuous DO monitoring, and DO modeling.
The Siirface Water Quality Division of the Michigan
Department of Natural Resources (MDNR) measured
channel dimensions and sediment depth at 49 sites in the
watershed using a survey rod and hand level. MDNR
also made observations of bank erosion and riparian
vegetation at most sites. Based on these observations,
WMSiPMi ^maiksmumia « -
.#1^ IT "S|l|iwl|ii; .„. „.„;,„;,
•,i$;fj
'"':f;|:i
mm
eaf!
^fc^'iira'iSii&.l'
irt;!i ivtj.
""iFill, if"]1"'!
;,!?'!
". !ii,f!!,;i';!>', ''•"'^ii:;'
''T.;,*1}*"!^^;
'^ssf^std
m
-------
active channel erosion at each site was classified as high,
moderate, or low.
MDNR conducted continuous DO monitoring at eight
locations using recording electrode style monitors.
Monitoring lasted from 6 to 103 days for each location.
Cause-and-effect synoptic DO surveys were conducted
twice with sampling at nine locations hi the creek during
summer low flows to provide data to calibrate a low-
flow DO model. These samples and a 24-hour
composite sample of effluent from the Mason WWTP
were analyzed for DO, biochemical oxygen demand
(BOD), ammonia, solids, and nutrients. Stream flow
measurements were made with a current meter, and an
ethylene gas tracer was used to determine a reaeration
rate coefficient for Sycamore Creek.
A special monitoring program was conducted during
1990 and 1991 to collect sediment and nutrient loading
data for this watershed analysis. Three agricultural
subwatersheds—Marshall Drain, Willow Creek, and
Haines Drain—were sampled from March, after the
snowmelt, until the appearance of a crop canopy in July.
Marshall Drain and Willow Creek are within the
Sycamore Creek drainage. Haines Drain is adjacent and
was monitored to provide a control watershed that would
allow a paired analysis for determining the effectiveness
of NFS control strategies. Its soil, slope, and land use
characteristics are similar to those of the Sycamore Creek
watershed. Water quality samples were collected by
hand two times each month during baseflow and by
using automatic samplers at 1- to 4-hour intervals during
runoff events . Flow was continuously measured during
the monitoring season.
Two urban subwatersheds (Rayner Creek and Columbia
Drain) were also monitored during two summer storms
using an automatic sampler at 1/2- to 4-hour intervals.
These watersheds were monitored to assist with the
identification of urban pollution sources.
Modeling Dissolved Oxygen
MDNR used a quasi steady state DO model (O'Connor
and DiToro, 1970) to predict DO concentrations in the
creek during drought conditions. They sought to
determine whether the DQ standard would be met under
the most severe circumstances and to determine the
relative importance of oxygen-consuming factors during a
drought. The model was calibrated by adjusting the
plant respiration and photosynthesis terms to obtain the
best match with the synoptic DO data. It was also
calibrated to match continuous DO data collected at one
location during the 1988 drought Plant respiration and
SOD were estimated in this second calibration as a single
term and then separated assuming no net oxygen
production by plants.
Preliminary Conclusions
The channel survey documented severe sedimentation
throughout the watershed. Average sediment depth was
0.3 meter of primarily fine sand and silt. The survey
also revealed that the most active stream bank erosion
Was occurring along wooded banks where herbaceous
plants were sparse. Ninety percent of the stations where
no active erosion was noted were nonwooded.
Nonwooded sites usually had thick sod stabilizing the
bank.
Seven of the DO monitoring sites were upstream of
Mason's WWTP, and all but one recorded DO
concentrations less than the minimum 5 mg/1 standard
The DO standard was violated at these stations 53 out of
153 days. Three sites upstream of the WWTP, but
downstream from a marsh, violated the DO standard on
every day they were monitored. The other three sites
that violated the standard, located upstream of both the
WWTP and the marsh, did so only on days of surface
runoff or during drought conditions. Downstream of the
WWTP, the measured DO was less than 5 mg/1 on only
1 out of 103 days and a large runoff event occurred on
that day. These data indicated that nonpoint sources
were contributing more to the oxygen demand in the
stream than point sources.
DO modeling showed that most of Sycamore Creek is
not expected to meet the DO standard under drought
flow conditions. The daily minimum DO expected at
drought flow is 0.0 mg/1 at West Service Drive Creek,
4.4 mg/1 at Cemetery Bridge, 4.5 mg/1 at Howell Road,
and 3.9 mg/1 at Harper Road. The primary DO sink
under drought conditions was shown to be SOD followed
by aquatic plant respiration (Figure 3). The segment
downstream of the marsh (represented by the station at
West Service Drive Creek) is expected to have DO
concentrations less than the standard, even under average
summer flow, because of very high SOD in the marsh.
The habitat observed to be destroyed by sediment and the
DO monitoring and modeling results show that sediment
is the pollutant most responsible for impairment of
Sycamore Creek. As a result, MDNR decided that
reducing suspended solids loads to the creek would be
the most appropriate way to decrease SOD and the
nutrient loads that may be stimulating aquatic plant
growth.
THE WATERSHED ANALYSIS
The pollutant load associated with each monitored runoff
period in the urban and agricultural subwate. heds was
calculated from the interval method (Richards and
Holloway, 1987) according to the equation
-------
JV!" .-"'<. •*'< ^i:
'i ||ri" I'" ! ; "T! :„
.i..'.,.;!!-!! ! i, , »4i" !: jlv,
WWTP = Mason WWTP BOD and Ammonia
RESP = Aquatic Rant Respiration
SOD = Sediment Oxygen Demand
,,,,.
:'" |1
eak dailV average flow at the USGS gage. These
models _were used to ^alculate^suspended
'
....... , ............................... ..... .............. „,, ..... , .................. , ....................
;,,£,;;,; ........... :^i ..................... gj=' goHutant concentration (nii/1);
; ;* ":;,;: »"' '1'=' ..... tfistaniabeous flQ^"™^:-^1^ ..... wifi' thiB
' ' ' ' "
the ..... usG<
1! p A i ill ii = !;, <•:'! !^
"T" i, I '"'I!;;'":,™! T;'v'Ii: *:l O" i:, T-..I::: <,: *,. if::..: sii^t ri ',m.,'s as: ki«= ss f n HI
,, i L | Jl,; k = HOIJ conyersipn cpnstaiit (0,00006),
nod of record included water years 1976-
j^Sp^^gj^wate^year 1990. A i significant istormi was
iuced a peak daily average flow
Sediment Loads from Agricultural
p ,,,„,„,,„„ p..,M, , „„ . ,, nan,jniiimpii'im >nq 1U,SPT in «? «T up »,«I::M:IT«IIIO"'«»I::M']«i«
• • • • — - • — —
^ ;.£> srnf
!!!="The re;
storms recorded at
6-year period of
;,;:2rn^/sec, at trie, Holt gage, with a peak flow at
•cent greater than t&e flow prior
fji»f! ..... i ....... !'"''! ..... jIF ...... !lif4'r| ........ |'»'j .......... "•" 'II ............... inllllllll*! '» "!lll|lllhlnl|ll IjINIjilillil rt ........ ; ...... S ..... :|l|g|||||r ....... jllljjll ilgji 'l!n||l||lp||| ........ pip 'lljlll
ijft if- p-:;;1' i1;1*":"1!;;"! ......... "I . I1;;!; i ; .is Jf *Jl'if.r •:;<; ' f iij" : j; "I: { fsltfif !"'
i" '''^'nce iC)'sitentin was
. . . [[[
^'''^'iJllnce ..... |^iC)^'s,i|tentipn; iiw,as:,toii|eveJ2D_ load reduction
'
gres
storms in
sion models for storm runoff loads in the
for storms with
. of the 61
the 6-year record at the gage, 53 fell within
. ............ this range and only 8 were larger. The regression
Various sources. J
snails; riiisini m
*
.
...... 2HE!252| [[[ the three Watersheds, including storm and base flow,
weremade
.-was 81
. ,- . .
models hi R values of 0.94, 0.75, and 0/70 for
. .......... ._ ......................... Marshall rain, Haines Drain, and Willow Creek,
respectivJfy. ^ total a ^^ load
for
began
tpbpc ,1989. Total annual flow for baseflow days was
t ..... ilEBuill ..... iiiii ......... Jliiiiiilill ..... UiiilLU^^^ ..... Ii ...... Sliilll ..... . .. ~^J . ,- .
by correlating a_ Um'ted _States_Geo;logical
'
- - three agricultural subwatersheds—Marshall Drain, Willow
Creek, and Haines Drain—was assumed to be
ive of the nonurban portion of the Sycamore
— - - Creek waljershed upstream of Harper Road. This
- — - assumption was valid because these subwatersheds have
- - soil and l.md use characteristics that are similar to those
• of the larger Sycamore Creek watershed upstream of
Harper Road. They are also subject to the same regional
-farming practices. Combined, these three subwatersheds
.
;0ds, a ..... linear ..... regression model was derived from the
-ivwHAlf'SI
-------
Estimating Sediment Loads from Eroding
Banks
Annual average channel erosion for actively eroding
banks was determined by multiplying the bank height at
the location of erosion, the length of the eroding portion,
the lateral recession rate, and the density of the soil in
the eroding banks. Length and height were estimated
from channel survey data collected by MDNR in 1989.
The lateral recession rate for a 3.6-kilome^r length of
the Willow Creek channel with organic soils was
determined by comparing 1989 channel cross section
measurements with design criteria for the channel when
it was last dredged in 1952. For other actively eroding
banks, the lateral bank recession rate was assumed to be
the same as that of Willow Creek if the soil was organic,
and half this rate for banks with loamy soil.
The fraction of the eroded soil that would travel as
bedload was subtracted from the channel erosion
estimates. This allowed direct comparison with the
agricultural and urban load estimates that were based on
measurements that did not include bedload. For organic
soils, the coarse sand fraction of the bank soil was
estimated by collecting a composite soil sample from the
stream bank and separating the components
gravimetrically by shaking the sample in a bottle with
water and observing the thickness of the sand layer after
settling. Particle size distributions in the Ingham County
Soil Survey (USDA, 1989) were used to estimate the
fraction of loamy soil that would travel as bedload.
In Willow Creek, erosion of organic soil on the
streambank contributed to the suspended solids loads
measured in that watershed. In addition to measured
loads, organic aggregates from stream banks were also
observed to be traveling semisuspended near the stream
bottom. No active bank erosion was occurring in either
Marshall Drain or Haines Drain, and therefore measured
loads originated only from upland areas in these
watersheds. The contribution of channel erosion to
measured suspended solids load in Willow Creek was
estimated by analyzing COD, turbidity, and suspended
solids measurements. Samples from Willow Creek that
contained primarily organic sediment from the stream
banks could be identified by turbidity measurements less
man 75 NTU and/or a ratio of COD to suspended solids
that was greater than 0.35. Figure 4 shows the
correlation between COD and suspended solids for low-
turbidity samples. There is a good correlation because
the source of organic solids in the stream bank is
homogeneous. These samples were usually collected
during the rise and peak of the storm hydrographs.
Estimating Sediment Loads from Urban Areas
An urban load estimation model (Driver and Tasker,
1988) was used to predict pollutant loads from the
Mason urban area. The model used rainfall, drainage
area, impervious area, population density, and mean
January temperature to predict pollutant loads from
individual storms. City sewer maps were used to
delineate drainage boundaries, and drainage areas were
then estimated by overlaying a grid and counting grid
squares in the watershed. Aerial photographs were used
to estimate impervious area, and the 1980 census
provided population density values. The model was used
to predict the suspended, solids load for each of the
storms monitored. While predicted loads agreed with
measured loads for Columbia Drain, agreement with
3OO
25O-
2OO-
Q .
0
0
15O-
100-
5O-
Turbidity «c 75 ••
Turbidity s- 75-f-
O 10O 2OO 3OO 4OO 5OO QOO TOO 8OO 9OO 10OO
Suspended Solids (mg/l)
FIGURE 4. Relationship between chemical oxygen demand and suspended solids for Willow Creek runoff samples
-------
; ;; , ,;;;;;;;,,;;;; ,;;,;;;; ,;,;;;;; ;;;;; ; ;,;; j~i i' ? • ....................... T - • • ............... a ............. ..... >•• j T^ .......... ,. , 11117™"^^..^. ..... ,.1,™ ..... ..... w ••/•„ -^wKmrm rmnmmrmnmiMmn mm^Km^vtmsm^
The model was calibrated by adjusting the bias correction
•' ; ...... , • , :.!" , ,;J V»i t*'"J ....... ' ............... 'Sti'lS'l.. «.«* ...... Jilil.™ ....... 8 ......... —--—,— , .......... ,,— ...... — .,!—'—„ , .....
; ;;; ;£;• coefficient until model predictions matched measured
!!' ....... " !'"' '' ' I'" "I '"' HI1 fl'l"' ' '''" ll*lpili!!P' ill iii!lli!i||!!i'i{H'!|||!'i"'''ipl|ii|iii|! ...... |l' I'l'i'lljifilli'illilliliiliiif l|||l!|!IPi!!l|ifl|q|||i|llpil|||"i!!;|!!||||iT 11HB*[ ................... „!,„„,„ ....... „„„,.,„,„ ...... ipfliMpH ........... nig ...... » ...... 01™™ ........ •II«IH«L« ,»• ...... HUHnnmnnnnnuiitiiiiiinH.!
s, A +290 percent^""adjistment was made for Rayner
..... fi
t^^
..... '
Mmu.™ !•«•<».n. i™.'
water quality goals, the pollutant
""are achievable' with Best management
esfaHS the'Ievel of pollutant reduction' 'that' is
"Te'.'r'detecteble)' "response
in die stream.
.—
to be reasonable in
^ii~]i3S:i;The calibrated model predicted si
r'!#iii i,inl i,./" j"'. •",,', 'i iii'lii ,,"i,',"ii .'I, Hii'ij,,ail '';<": i^ui uiiil 1:3,1:LzijiHyt: ia i -,m BUI* Mmym
iii:!! sf-s : '•&•&*'< for rainfall events between April
i i i,;;,;,;,!:, i (1 ii ;Ji .' ,;,;„:,! Hit! fUlLAL.atMl I .viJdlf!ljJ!j«*llllH3|i«lhi 0=1,
suspended solids load
and October for 1976
I' '"'II I ' I |iiii|ii'i|«|i|i, 7*7 ,,, „ ,: , ,, : ,„,,,, , |f ,r,ji,,, , L| ,| „, h ,', ¥: i",],!',,!,,!,"!,,,!',"in,~ii,ii'mSiin«i„:,!HBiH|i!iiiii|i' ig||..iii i"iinnppmHUPiii'iiimiih'iiiiiqiiHiHii'niHi'uvnui ni
; jfeiSfr; amount of rainfall for each day within this period from
il, 11 in-' , • • • ,1, • j ., j,T "| n;' S'"' i nI! lAr ,j i.'I'-iik!,,,'' al' v± 1 'ti-1:»;!' "x •>*&I ',r «iUiBi!1 »•,..inwiiiRi'inii'Mimu1'! um\ 4i -"an
assi mption of a proportional response,
, -
3 shows
19''"km "soutSwS" of'Masoinl
'i^JSS,to, caculate
"
that the suspended sohds loading would have to be
,,,_, _.,_, .,^.,— _™._,,,_ ^™ '13O"Tevels sufficiently to
t downstream' of
^
' '
^during
rch), because of the possibility Sat
!i'" 'f', ii !,!'E I'liffii *''i*r\! iff ''5' iiiiiv 'i1' n.jf 'i lliBiKiilHinniiyiB^ii1 »IMI» i^immniiiiy' «PUIJ»W»«III»»N»IN
cent rSucHori'1 'in ..... S'O'D ..... would' improve DC) at
er liter. Since the'
, ,,!t. .„.,,, ,,,,,, r j.^QXeini
^^VjjKiftiiifeji'^i^lpitation 'during this" season
" ! '" " "! : "which' does' "not .run off "immediately!'' In'steadT'
t,run off ,im
1""1""!!*:1' .1'"' <""";::j i'i!" *::iTT;:,iii :T, '" mlrl11!"1:1."": rMSiilTjiifii.'''!!'!!!)1!! lii'lSiiarawiaiilsiHM^ M ™»ii?"pOMiau i!iiii»*i»'«B»i«iia
1 loads were estimated to be 6Q percent of summer loads.
"expEctSi" O'O' at Harper"KoacT is" 3.9 'milligrams''per liter,
"~aT[! milligrams per liter iinprovemenTiri'DO would
con
®
, ™™ ,-
" of '"Se"' city ' servecl 'by" storaa'sewiers" "
, iilili ziiiiS! linii s:i Sliitii 111 It!
X presents one possible allocation scheme to
[eve 'th'e""§2 peVceiit'rectuction'.'1 '"Tfiis" scheme includes
erosion by 56"percent, reducing
.,. JTflii iarea,[''ipi,itfiei,,eri'tire,;
;| j^D^^mQuL^SCunKinled^oji^site treatment,were_
_stream £anFerosion tn'orgaluc^bils 6y 100 percent,
..^j,|-g.—.__.-g.j,g,^,..,,™™,,s,..l,^^™l,^|™ ,^_. ^Q
"penpentl an'3 reHuclng"urU'an" sources"'of'suspended solids
at
to be directly discharging to the stream or to
IHIigii'liii I l'i»i|'
jJ£^&^i^^;Sepdc tapks
i iii;j;1;|iii;.ii;ii; •:;U,^jtJ*jfS;H!Stl''JjS'i
•'' t1''!''1!;""?'''1;!!:*'!'' vi'ilsi!"
IBffllBWi'M'pPliiafi!BrPl»?Kl|ln'i''rflin"l!Kn*«l''iliir''aMl" ff'fliMi' ft! JIBRirWc* l*il:lli,,slirlJliS''RW',,F,*nlr''71l?t.r'W:3l'
reduction strategies are possible, but this strategy
achieve^'by'carefully targeting
.^_,^,,^^.™ ™^,__^_
areas inc'lucEe'agricultural fields
HiiNiii ill,: INN i .U «i«>.r "ligiii: !|ii i, i M#i in iiiliiii I i ^ininiiini iimiii • nhliiMijii ..Nil illhiii1 VHi hiiiiii liiii.ii.ili ii i I IU!iiiiiiiiiiiiiiiii|
:»!.: !!• I..: !! ':.,.:...: :,:.!:..! 1.1!!:..:.!... • '. .»'" 1 :! i: ,< ' ::, : I!....!: .''».! ....:.,!!ii,i ",.!.!.!!!:! ..::. : ..lllife '•: :l I!,: i II11C !» ill LJih I
to the stream, construction sites that are" 'adjacent" |'' (
^\lf ^'^^'^i^^^ystomswi^"in the""'
ay of Mason, "and'the'most severelyye'rod^g'stream '
|i«S1»
solids is most important if the
LEMENTING POLLUTION
XX)NTR6tS
provide
for recreain;"'d^^ien ithe" channeirthereby
^lll^^tions^y reducing SOD and aquatic plant
r, "|6|^ilatipn. Mo33 slmuIaSons revealed tEat respiration
laSr '-'f'^ri—T^^^
'I'M^JIUII^I .h.,'1,1 ,,,
:ficrt at
necessary load reductipris will be achieved, in part,
iirn*m«i«««ii« »i»»im'™:"™5i»a
to the DO
pjants
by jin ongping USD A prpgram that is intended to reduce
e^!iOTponragricuituraj land^by 50 percent by targeting
MgMy erpdible areas. .However, additional programs
j^'C;!?^'!?'1 f^?5°^"?n a^a^®^ °u*5SH^i 5?! JES^:":^1^^11*^!" "" " """ * BSiJ^ 12 KlSS6, SX9SH I9?^S? *2 5?, ,PeF£ent~
"l: etJMpspbrte^ to'^ycamore SSE'wEIIe """
anss Wl
tr\tc time no
1v nredict
"rpiufca,,,suspended solids toad on habita^ _aquatic life, or
SD." !n ISe absence
* !''":« :,! I? efforts to reduce, sus*
1 'i,;,,;,; -I ;,,/;;,, V;JT j;1-; ......Jt'ithj,,;!,^,!;!;, IK'II! i
ea soli3s'" loadings,"
"Ti / "i ; i : ', ,,",i;!ii|i.i:iT',i;:;i;1111 n [in
'"v;;i :Tl;t ^,r:;,/';^T,^1^f'lfifep
|||-jii<;: J, i,"'1^ I-"'' ,S:i '!i'i5^«'j|''':'ii i'jijiiili"! •,"^f;7^li/' MI3JJ; >^|?'i!iill''(SIIJ!!!|!i
Follow-up monitoring is necessary to .indicate whether
IB [TMDL adequately protects water quality and the
liflSfeilHil^^^^
-------
TABLE 1. Annual average load of suspended solids from various sources to Sycamore Creek upstream of Harper
Road and the load reductions necessary to meet water quality standards
Source of Suspended Solids
Organic soil from stream banks
Loamy soil from stream banks
Nonurban (e.g., agricultural
fields)
Urban runoff
Point sources
Septic tanks
Annual Average
Load (metric
tons/year)
209
238
438
153
7.9
4.3
Load Reduction
to Meet Water
Quality
Standards
100%
20%
56%
30%
0%
0%
Method of Calculation
change in channel volume over time
field estimate
regression model and monitoring
data
calibrated regression model
self-monitoring data
worst-case estimate
aquatic community and to better quantify loads, verify
models, and evaluate the effectiveness of controls.
Monitoring of three agricultural subwatersheds using a
paired sampling approach (Spooner et al., 1985) is being
conducted to provide feedback on whether best
management practices reduce sediment loads to the
stream. Agricultural management practices are being
documented by periodic site visits during the sampling
season (approximately March-July). MDNR is storing
these land use data in the form of input files for the
Agricultural Nonpoint Source (AGNPS) model. The
AGNPS model results are being compared to actual
runoff data for each runoff event that is monitored. This
is possible since recording rain gages are being operated
in eaqh subwatershed during the monitoring season.
Three years of data collection have been completed.
Data from the first 2 years of monitoring were used to
estimate loads from agricultural areas as described above.
At this writing (December 1992), it appears promising
that funding for an additional 6 years is forthcoming
from USEPA under the §319 national monitoring
program. The future monitoring data can be used to
verify and refine the agricultural loading model described
above.
REFERENCES
Clark, K. 1990. A biological investigation of Sycamore
Creek and tributaries Ingham County, Michigan.
Michigan Department of Natural Resources, Lansing,
Michigan.
Driver, N.E., and G.E. Tasker. 1988. Techniques for
estimation of storm-runoff loads, volumes, and selected
constituent concentrations in urban watersheds in the
United States. U.S. Geological Survey Open Hie Report
88-191.
O'Connor, D.J., and D.M. DiToro. 1970.
Photosynthesis and oxygen balance in streams. Journal
of the Sanitary Engineering Division, ASCE, 96, SA2,
547.
Richards, R. P., and J. Holloway. 1987. Monte Carlo
strategies for estimating tributary loads. Water
Resources Research 23(10): 1939-1948.
Spooner, J., R.P. Maas, S.A. Dressing, M.D. Smolen, and
F.J. Humenik. 1985. Appropriate designs For
documenting water quality improvements from
agricultural NFS control programs. In Perspectives on
Nonpoint Source Pollution, EPA 440/5-85-001, pp. 30-
34.
USDA. 1989. Soil survey of Ingham County Michigan.
U.S. Department of Agriculture, Mason, Michigan.
This case study was prepared by John Suppniek, Michigan
Department of Natural Resources, in conjunction with
USHPA Office of Wetlands, Oceans and Watershed
Watershed Management Section. To obtain, copies, contact
your EPA Regional 303{
-------
------- |