v - ^ .< f /->-» ' ' -" ' 7
(<^.fc ,AW*
EPA-600/3-77-112
1977 Ecological Research Series
f ESTIGiOE RUHOFF
WATERSHEDS IN GREAT LAKES BASIN
Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Athens, Georgia
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1 Environmental Health Effects Research
2. Environmental Protection Technology
3 Ecological Research
4 Environmental Monitoring
5 Socioeconomic Environmental Studies
6 Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials. Problems are assessed for their long- and short-term influ-
ences. Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects. This work provides the technical basis
for setting standards to minimize undesirable changes in living organisms in the
aquatic, terrestrial, and atmospheric environments
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161
-------
Eavirorr:-.'.-" "'
Lakes Ha'..'.
GLSPO L
EPA-600/3-77-112
October 1977
PESTICIDE RUNOFF LOSSES FROM SMALL WATERSHEDS
IN GREAT LAKES BASIN
by
B. G. Ellis, A. E. Erickson, A. R. Wolcott
M. Zabik, and R. Leavitt
Departments of Crop and Soil Sciences
and Entomology
Michigan State University
East Lansing, Michigan 48824
Grant No. R-800483
Project Officer
George W. Bailey
Environmental Research Laboratory
Athens, Georgia 30605
ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
ATHENS, GEORGIA 30605
-------
DISCLAIMER
This report has been reviewed by the Environmental Research Laboratory,
U. S. Environmental Protection Agency, Athens, Georgia, and approved for
publication. Approval does not signify that the contents necessarily reflect
the views and policies of the U. S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
11
-------
FOREWORD
Environmental protection efforts are increasingly directed towards
preventing adverse health and ecological effects associated with specific
compounds of natural or human origin. As part of this Laboratory's research
on the occurrence, movement, transformation, impact, and control of environ-
mental contaminants, management or engineering tools are developed for
assessing and controlling adverse environmental effects of non-irrigated
agriculture and of silviculture.
To meet the nation's food production needs, modern agriculture will
continue to rely on the application of fertilizers and pesticides. In
recognition of this, research continues to be directed toward delineating
the fate of these materials after their initial beneficial application.
The project described in this report evaluates the movement of several
herbicides in the water and sediment being lost from two agricultural
watersheds. These analytical data may also serve as a basis for develop-
ing and evaluating models of pesticide transport in soils and agricultural
landscapes.
David W. Duttweiler
Director
Environmental Research Laboratory
Athens, Georgia
in
-------
ABSTRACT
Patterns of runoff and sedimentation observed on two watersheds in
Michigan are described in relation to weather conditions at different seasons
of the year. An assessment is made of sources of variation in pesticide anal-
ysis for soil cores taken during the period May 1973 through September 1974
from the two watersheds. A number of relationships to methodology, chemical
species, topography, soil conditions, and weather are identified. Criteria
are given for assessing down-slope movement within and between sampling seg-
ments and movement within the profile. A detailed description is given of
weather and watershed conditions associated with wintertime runoff events on
the larger watershed and with major spring and summer events on both water-
sheds in 1975. Emphasis is placed on characterizing boundary conditions at
the beginning of each event in relation to weather sequences that preceded it.
Only portions of the data set, stored at the Environmental Research Laboratory,
Athens, GA, were used in these evaluations. However, important features of
soil, topography, management, and weather are identified in relation to use-
ful variation in the data. The described relationships should be helpful in
interpreting and modeling data from these watersheds for both pesticides and
nutrients.
The detailed descriptions of soils, topography, instrumentation, opera-
tional procedures, and management given in this report as background for in-
terpreting the pesticide data also apply to the nutrient data obtained under
EPA Grant No. R-802974-01-0. These nutrient data will be summarized and eval-
uated in a separate report.
This report was submitted in fulfillment of Grant No. R-800483 by Michi-
gan State University under the sponsorship of the U.S. Environmental Protec-
tion Agency. This report covers the period September 23, 1972, to October 31,
1975, and work was completed as of June 30, 1976.
IV
-------
CONTENTS
Abstract j_v
List of Figures vi
List of Tables vii
Acknowledgment viii
Sections
I Summary 1
Conclusions 3
II Recommendations 5
III Introduction 6
IV Experimental Methods 8
Description of Watersheds 8
Details of Construction and Operation 8
Soil Sample Collection 15
Runoff Samples 20
Residue Analysis 21
V Results and Discussion 23
Introduction 23
Runoff and Sedimentation Patterns 29
Soil Core Data, May 1973 to September 1974 37
Sources of variation 38
Evidence for movement down slope and in the profile 44
Paraquat in soils 45
Trifluralin in soils 48
Diphenamid in soils 49
Atrazine in soils 52
Summation 55
Selected Runoff Data, 1975 57
Wintertime runoff sequences 57
Unique spring and summer events 72
Summation 73
VI Literature Cited 76
VII Appendix 77
-------
FIGURES
Number
1 Principal soil types on the two watersheds (06 and 07)
at Michigan State University Farms ...... .
2 Contours and slope classifications. MSU watersheds 06 and
07. (SCS 1942) ...................... 10
3a Stainless steel linings of catchment basins (watershed 06) .. 11
3b Stainless steel Coshocton wheel (watershed 06) ........ 11
4a Stainless steel lead pipe and optional samples splitter
(watershed 06) ...................... 13
4b Stainless steel collection vessels in operation
(watershed 06) ...................... 13
5 Sampling segments 1973-74 .................. 16
6 Sampling segments 1973-74 (with soils overlay) ........ 17
7 Sampling segments 1974-75 (with topographic overlay) ..... 18
8 Sampling segments 1974-75 (with soil and topographic overlays 19
VI
-------
TABLES
Number Page
1 Recovery of Chemicals from Spiked Samples 22
2 List of Field Operations for Watersheds 06 and 07 24
3 Watershed 06 Paraquat May 1973 to May 1974 38
4 Watershed 06 Paraquat Summer 1974 39
5 Watershed 07 Paraquat May 1973 to May 1974 40
6 Watershed 07 Paraquat Summer 1974 41
7 Watershed 06 Trifluralin May 1973 to May 1974 42
8 Watershed 07 Trifluralin May 1973 to May 1974 43
9 Watershed 06 Diphenamid May 1973 to May 1974 50
10 Watershed 07 Diphenamid May 1973 to May 1974 51
11 Watershed 07 Diphenamid Summer 1974 53
12 Watershed 06 Atrazine Summer 1974 54
13 Weather and Watershed Conditions for the Period
03-12-75 to 03-11-75 58
14 Weather and Watershed Conditions for the Period
03-18-75 to 03-24-75 59
15 Weather and Watershed Conditions for the Period
03-25-75 to 03-31-75 60
16 Weather and Watershed Conditions for the Period
04-01-75 to 04-13-75 61
17 Weather and Watershed Conditions for the Period
04-14-75 to 04-18-75 62
18 Weather and Watershed Conditions on 08-21-75 and 08-22-75 ... 63
19 Rainfall, Runoff and Pesticides in Runoff for the Period
03-12-75 to 03-17-75 64
vii
-------
20 Rainfall, Runoff and Pesticides in Runoff for the Period
03-18-75 to 03-24-75 65
21 Rainfall, Runoff and Pesticides in Runoff for the Period
03-25-75 to 03-31-75 66
22 Rainfall, Runoff and Pesticides in Runoff for the Period
04-01-75 to 04-13-75 67
23 Rainfall, Runoff and Pesticides in Runoff from Watershed 06
for Unique Rainfall Events on Bare Soil (04-18-75) and on
Soil Under Corn Cover (08-21,22-75) 68
24 Rainfall, Runoff and Pesticides in Runoff from Watershed 07
for Unique Rainfall Events on Bare Soil (04-18-75) and on
Soil Under Soybean Cover (08-21,22-75 69
A-l Properties of Pesticides Used in This Study 77
Vlll
-------
ACKNOWLEDGEMENTS
We gratefully acknowledge the major contributions of Mr. Robert Hubbard
and Mr. A. B. Filonow. Mr. Hubbard assumed responsibility for much of the
installation and maintenance of runoff collection facilities and instrumenta-
tion at the site and for collecting and preparing meteorological data and
waterstage records for transmittal. Mr. Filonow had primary responsibility
for liaison with the Pesticide Analysis Laboratory and for preparing quar-
terly and annual reports and summaries under this project. He shared with
Mr. Hubbard responsibilities for collection of soil and runoff samples and
for maintaining field logs under this and the nutrient project. In the
Pesticide Analysis Laboratory, Mrs. Irene Amnock had responsibilities for
preparing samples for analysis and for coordinating analytical and data re-
cording activities. Her services are acknowledged with appreciation.
IX
-------
SECTION I
SUMMARY
Soils, topography, instrumentation, operational procedures and manage-
ment are described for two watersheds used in a 2-1/2 year study of movement
and losses of pesticides and nutrients from non-point sources in the Great
Lakes Basin. Observed patterns of runoff and sedimentation are described in
relation to weather conditions at different seasons of the year.
An assessment is made of sources of variation in pesticide analyses for
soil cores taken during the period May 1973 through September 1974. A number
of relationships to methodology, chemical species, topography, soil conditions
and weather are identified:
a) Random variation is high and derives from variable distribution
of chemicals during application, variability associated with
sampling and analysis, and random patterns of pickup and inter-
ception from runoff.
b) Variation is present and can be interpreted in terms of
movement downslope and within the plow layer, and in terms of
loss by degradation and/or volatilization or by leaching out
of the sampling zone into the subsoil.
c) Due to sampling bias, downslope movement within steeply sloping
segments is reflected by increasing recovery, rather than by
decreasing mass balances as might be expected. This relation-
ship may change if the capacitances represented by depositional
slopes and by retention in the plow layer are satisfied or
exceeded by heavier rainfall in 1975.
d) Because of its resistance to degradation and its susceptibility
to adsorption, paraquat was the most useful tracer for following
movement and retention in the landscape. Losses by degradation
or volatilization tended to mask the evidence for movement of
other herbicides in this study.
e) Mobility (both downward in the profile and in surface runoff) was
related directly to solubility: Paraquat > diphenamid > atrazine
> trifluralin. The great mobility of paraquat was not expected and
suggests interactions with fulvic acid components which inhibit its
biological activity without immobilizing it.
f) Redistribution patterns reflected the effects of a number of
factors on interception, infiltration and deep percolation of
-------
rainfall and runoff. East-west (E-W) planter rows determined the
direction of runoff from upper slopes, promoted cascading cross-row
runoff in the central drainageway, and provided micro-relief for
impoundment of runoff (with sorted sedimentation) at lower eleva-
tions. Granular structure of the plow layer on eroded slopes
promoted infiltration and deep percolation of rainfall and runoff.
Loosened structure and irregular surface of freshly tilled soil
promoted interception and infiltration during earlier portions of
each growing season. Length and pitch of slope were important, as
well as exposure of slope to drifting snow and winter sun. The
presence of a frost layer precluded deep percolation, promoted
pickup of saturated soil materials on upper slopes and ponding at
lower elevations. Lingering snow and ice promoted interception on
intermediate slopes.
g) Differences between the two watersheds in distribution and retention
of chemicals reflected mainly differences in length of E-W slopes
and the proportion of gentle slopes where runoff flows tend to slow
down or where ponding can occur. Soil differences were minor except
in relation to topography and the degree of erosion and incorporation
of granular subsoil materials into the plow layer.
h) Movement over the watersheds and through the soil was more rapid
in summer 1974 than in 1973, reflecting rains of greater magnitude.
In 1975, deeply penetrating rains and rains of erosive intensity
were more frequent than in 1974. Greater downslope movement of
chemicals may be expected because visible sediments were transported
further down the drainageway in both watersheds. Corn and soybeans
were much more vigorous in 1975, and differences between watersheds
may reflect differences in nature of the canopy and its rate of
development.
Weather and watershed conditions associated with runoff sequences during
winter 1974-75 are described. Emphasis is placed on characterizing boundary
conditions at the beginning of each event in relation to weather sequences
which preceded it:
a) The depth and distribution of snow cover reflected effects
of wind direction and local relief on patterns of drifting.
b) The depth, age and distribution of snow cover over the water-
sheds served to moderate the rates of penetration and dis-
appearance of frost in the profile. Direction of slope, in
relation to angle of insolation during the warmer part of the
day, was an important factor affecting depth and quality of snow
cover and disappearance of surface ice and of frost within
the profile.
c) Snow, particularly new snow, can intercept rain and runoff
and can delay net runoff from a watershed.
d) Frost in the profile prevents deep percolation of melt water
-------
or rain and permits runoff from minor amounts of precipitation.
e) When surface soil thaws on warm or sunny days, it tends to remain
saturated and soil particles are readily suspended in runoff.
Upper slopes can drain quickly by seepage as soon as sources of
melt water disappear.
f) Nightly freezing forms ice in saturated soil on slopes fed by
seepage from above and in areas where ponding occurs during
the day. This ice layer is slow to thaw, protects soil on slopes
from erosion, but serves to intercept seepage on slopes and to
seal the surface in areas of ponding. Under these conditions,
seepage from thawed and saturated soil over frost in sloping
areas can extend runoff for a time after rain stops or after
sources of melt water have disappeared.
g) Sediment pickup from frozen soil is low, but sediment loads
in runoff increase as the area exposed to thawing and satura-
tion of surface layers increases.
h) The recharge capacity of thawed soil increases as deep frost
layers recede. The quantity of precipitation needed to pro-
duce runoff increases accordingly.
i) Sediment yields in runoff collected at catchments provide a basis
for identifying events and watershed conditions which produced
major redistributions of sediments and chemicals within the water-
sheds.
Runoff data for major historic events in April and August 1975 are used
to illustrate effects of crop cover per se and of crop characteristics (brace
roots of corn, lodging of soybeans) on runoff volumes, sediment yields and
losses of chemicals.
CONCLUSIONS
1. Weather sequences which precede a runoff event can influence, some-
times determine, the ratio of runoff to infiltration and percolation. During
the early growing season, even light rains can modify surface structure and
micro-relief, left by tillage, in ways which favor runoff. During colder
seasons of the year, precipitation over a period of days or weeks can accumu-
late as snow which can be cumulatively effective as current precipitation
when it thaws. The distribution of snow over the landscape can vary dramati-
cally with drifting which, in turn, is influenced by wind velocity and by wind
direction in relation to local relief.
2. When winter thaws occur in the Great Lakes Basin, with or without
rain, infiltration and deep percolation are frequently controlled by the dis-
tribution of surface ice or of frost within the profile. The distribution
of surface ice and frost are influenced by topographical features which de-
termine insolation angle and patterns of down-slope seepage during periods
-------
of daily thawing and nightly freezing.
3. Patterns of sedimentation which can be observed visually are not
reliable for inferring patterns of redistribution of chemicals carried in
runoff. Visible sediments represent coarser, heavier, less adsorptive soil
fractions. Chemicals in solution or associated with very fine suspended
matter may be intercepted at any point where infiltration of runoff occurs.
Infiltration of runoff is promoted where runoff is slowed down on gentling
slopes or by lingering snow or ice and, in particular, where ponding occurs.
4. The micro-relief afforded by planter rows and the impedance afforded
by plant roots and stubble tend to control the direction of runoff on short:
slopes and contribute to the impoundment capacity in level basin areas. They
also provide the conditions for cascading runoff and severe erosion on
sloping areas which transect crop rows. Contour planting, no-till and good
soil conservation practices would appear to be most important for controlling
pollution from non-point sources in agricultural landscapes where cultivated
crops must be grown.
5. The mobility of pesticides in soil and over the landscape appears
to be primarily a function of their solubility and the degree to which they
are adsorbed on suspended sediments. On the permeable soils of these water-
sheds, net losses from surface applications appeared to be minimal in relation
to quantities retained by displacement into the plow layer and subsoil. In
the case of less persistent chemicals, net losses in runoff were further re-
duced by volatilization and/or degradation.
6. In spite of its cationic character, paraquat was the most mobile of
the chemicals in this study. It is likely that its displacement into the
profile would have been less rapid on finer textured soils with higher ex-
change capacity. However, it is possible that its high mobility (which does
correspond to its high solubility) was promoted by interaction with soluble,
macromolecular components of the fulvic acid fraction of soil organic matter.
Certainly, its phytotoxicity is destroyed by contact with the soil. If it
does move in complexes that inhibit its biological activity, its appearance
in runoff should pose no threat to natural systems in receiving waters.
-------
SECTION II
RECOMMENDATIONS
1. Sampling segments as delineated for these watersheds failed to dif-
ferentiate intermediate slopes where significant interception of runoff and
dissolved or suspended chemicals occurred with little or no visible sedimen-
tation. Each of the segments with major E-W slopes (01, 03, 04, and 06 in
Fig. 8) should have been divided along contours into at least two smaller seg-
ments. This should be considered in any future studies on these watersheds.
The principle applies to other watersheds as well.
2. Weather sequences which precede a runoff event can influence boundary
conditions dramatically, notably during winter months. Important boundary
conditions not characterized in this study, or for which only qualitative ob-
servations were made, include soil moisture content and distribution at the
beginning of summer events, and the distribution of snow cover, surface ice
and frost within the profile in relation to wind velocity, wind direction and
angle of insolation during winter months as influenced by local relief within
and around the watersheds. Costs for obtaining appropriately objective and
detailed data on these parameters will be high but should be considered in
any projections for developing second generation models for non-point sources
of pollution under conditions in the Great Lakes Basin.
3. Losses of pesticide could have been reduced substantially by sound
conservation practices such as contour planting, grassed waterways and winter
cover crops. But losses in runoff from these watersheds were very low except
for a major historic event in April 1975 in which 30 to 60% of the total loss
occurred. Much of the loss during this event can be ascribed to sediments
picked up from deep rills and gulleys cut along central drainageways by cas-
cading runoff across rows of crop stubble.
-------
SECTION III
INTRODUCTION
The challenge for the modern agriculturalist is to optimize food pro-
duction generating the highest quality of food and fiber with respect to pro-
ductivity, environmental quality and other national goals. This is necessary
to feed an ever increasing population at an economical cost and without de-
grading the environment. The use of pesticides and fertilizers, i.e., agri-
cultural chemicals, is a part of modern agriculture to meet the need for food
production.
The necessity of using both fertilizers and pesticides cannot be ques-
tioned, but the need for careful evaluation of the fate of these materials
once applied to farm land cannot be underestimated. Once applied to land,
a pesticide may (1) be adsorbed by soil particles and either retained in the
soil or move with the soil particles in the event of erosion, (2) remain in
the water fraction of the soil and be subjected to leaching downward or to
loss in surface runoff, (3) be vaporized into the air, (4) be absorbed by a
crop and thus added to our food chain, or (5) be degraded by microbial, chem-
ical, and photochemical processes in soils. The ultimate fate of an applied
pesticide is then dependent upon (1) properties of the pesticide, (2) soil
properties that affect the distribution of the pesticide between solid and
liquid phases in the soil, (3) climate within a particular region which in-
fluences the quantity of runoff from a watershed, (4) management practices,
and (5) watershed characteristics.
Many pesticides have been marketed in the past; undoubtedly, many more
will be cleared by regulatory agencies for use in the future. The chemical
properties of these pesticides may vary widely. To evaluate the runoff loss
potential of each new pesticide by field experimentation would not be feasible
from the view of either time or money. And many valuable years' use of a po-
tentially useful pesticide would be lost by holding the material from the
market until field experiments could be conducted. It is, therefore, impor-
tant to seek ways of evaluating the erosion and leaching hazard of new pesti-
cides by methods that do not require costly field experimentation for each
individual chemical. The development of models for pesticide transport in
field situations appears to offer a viable alternative for predicting pesti-
cide loss. These models can then be validated with field data for carefully
selected " model" pesticides and field locations.
The inclusion of watersheds from the Great Lakes Region in such a model-
ing program is important because the soil types are considerably different
from Central, Southern, or Western United States; the climate includes snow-
fall leading to runoff and erosion from snowmelt during winter and spring.
Two small watersheds existed on the campus farms at Michigan State University
-------
which had been studied for many years prior to establishing the experiments
described in this report. Their inclusion in this study was desirable because
their history was known, weather records for about 35 years were available
for the site, a small weather station was located adjacent to the watersheds
for collection of weather data, and a portion of the physical structure neces-
sary for these experiments was already on site.
The objectives of the project were to
1. evaluate the movement of several herbicides (differing widely
in volatility, solubility, and susceptibility to adsorption)
in the water and sediment being lost from agricultural water-
sheds in the Great Lakes Basin;
2. furnish Athens Environmental Research Laboratory, EPA, with
analytical data and historical hydrologic and climatic data
which will allow for evaluation, by systems analysis, of
movement of pesticides in soils and agricultural landscapes;
3. describe modes of pesticide transport quantitatively in an
agricultural landscape.
4. determine the effect of soil factors and micro relief on
distribution and transport of pesticides from an agricultural
watershed.
-------
SECTION IV
EXPERIMENTAL METHODS
DESCRIPTION OF WATERSHEDS
Soil types typical of the Great Lakes Basin are found in two watersheds
on the soil science farm at Michigan State University. The east watershed
(hereafter referred to as 06) contains 0.80 hectare and the west watershed
(hereafter referred to as 07) contains 0.55 hectare. The principal soil type
of the watersheds is a Spinks loamy fine sand (figure 1). Watershed 07 con-
tains some Tuscola fine sandy loam in the northwest corner. The slopes in
the eastern 1/3 of watershed 06 are classified as Hillsdale fine sandy loam.
Depositional materials in the central confluence area of 06 are classified
as Traverse fine sandy loam. Official descriptions for these soil series
are attached in Appendix I.
A detailed topographical survey made in 1942 is presented in figure 2.
Slopes vary from 2 to 4 percent in front of the catchments to 10 to 12 per-
cent at certain points in the watersheds. Areas with slopes of 6 to 12 per-
cent in figure 2 correspond to areas of class 2 erosion in figure 1. In areas
of 8 to 12 percent slope, subsoil materials have been exposed. Due to their
coarsely granular structure, these materials form drouthy seedbeds so that.
germination is frequently reduced or delayed and early growth of crops is
delayed. On upper slopes, near the watershed perimeters, soil moisture re-
serves are quickly depleted, and crops develop symptoms of water stress more
quickly at any time during the season than at contour elevations only a few
feet lower downslope. A protective canopy develops more slowly on upper
slopes and is less dense at maturity. As a result, soils are exposed for
longer periods to the erosive action of heavy rains.
The runoff catchments were initially concrete, with water flow being
measured by means of standard waterstage devices at the weirs.
The watersheds were in continuous corn for over ten years prior to ini-
tiation of this project. During this period they received, on the average,
3.37 Kg/ha of atrazine and heavy applications of complete fertilizer each
year. Livestock manure was applied in 1970 at a rate of 40 metric tons per
hectare.
DETAILS OF CONSTRUCTION AND OPERATION
The precatchments, catchment basins and weirs were lined with stainless
steel (see figure 3a). Stainless steel, 1:100, Coshocton wheels were con-
structed and connected by a belt drive to two DC auto heater fan motors for
-------
o
M
CD
O
o
CO
^
X
0
in
o
2
LU
0
Id
_;
o
CO
o
o
0
(A
1-
-
in
o
CO
it-
CD
tn
o
o
-
in
m
o
M
ID
C.
Q.
CO
o>
CO
/
0^
5s m o> »
« N « °
CM | ^ v
1 *> = '
O CM CO
z
o
w - +
0
cc
UJ ^^^
x^
^S"
s
en
/
X
/
1
/
< -
Q CD
U O
»- ^
< CO
>
CO
in
01
a.
Q.
O
E
co
h
0)
0)
4J
CO
a
CO
00
H
CJ
CO
vO
O
CO
T3
01
^
co
M
01
1
f
CO
3
o.
E
o
CJ
>«
^
CM
O)
"
CO
o
CO
M
Ctt
C
o
*
.
o
CO
r
^
CL
UJ
>
0
s
4J
0)
4-J
C
o
co
0)
Cb
o
ca
o
a
H
0)
J^
M
H
-------
DETE/fM/HAT/ON OFJLOFf CLASS
Fig. 2. Contours and slope classifications. MSU watersheds
06 and 07. (SCS 1942)
10
-------
Figure 3a. Stainless steel linings of catchment basins (watershed 06)
Figure 3b. Stainless steel Coshocton wheel (watershed 06)
11
-------
each wheel. One is shown in operation in figure 3b. Current was supplied
by heavy duty storage batteries equipped with trickle chargers. A stainless
steel pipe from the Coshocton wheel terminated in an optional (1:1 or 1:10)
sample splitter prior to directing the samples to receiving containers on
a rotating carrousel (see figure 4a). As shown in figure 4b, samples were
collected in stainless steel pots. A float, attached to the delivery spout,
was equipped with a mercury switch adjusted to close the circuit to the car-
rousel drive motor when one pot was full (about 10 liters) and open it again
when an empty pot had advanced into place.
Heating tapes were installed below the stainless steel catchment liners
and around the waterstage wells, delivery pipes and sample splitters to pre-
vent those portions of the system from freezing during runoff periods in late
fall to early spring. These were not always effective. In particular, the
water in the waterstage well froze on several occasions, and two or three
major winter runoffs were not registered. Runoff volumes for these events
were calculated initially from collected sample volumes and nominal sample
reduction settings. The originally reported waterstage records for these
events have been adjusted for probably effective reductions calculated from
computer output for other events when satisfactory waterstage records were
obtained.
The Coshocton wheels and catchments were protected from direct precipi-
tation by a corrugated steel roof. Plywood sides were installed around the
front and sides of the catchments to reduce drifting of snow. A 5 cm gap
was left at the lip of the catchment for runoff to enter, and on several oc-
casions, southerly winds caused much more drifting inside the catchment
through this gap than if the sides had not been present.
The Coshocton wheels were not protected around the sides. Frequently
winter runoff under snow cover would continue for several hours after air
temperatures had fallen below freezing with the result that Coshocton wheels
and motors would be encased in ice, belts would break or pins would shear,
and motors burn out. On several other occasions, snow drifted to depths of
1.8 to 2.4 meters over the Coshocton wheels and in the after-flume runoff
channels. The snow was removed promptly from the Coshoctons and sample split-
ters, but on two occasions heavy runoffs occurred before the channels had
been cleared to the outlet tile with the result that runoff backed up and
overflowed through the sample splitter into the collection house. Fortunate-
ly, project personnel were on hand to open the channels before excessive
overflow into the house had occurred. (A similar overflow situation devel-
oped during a torrential downpour on August 21, 1975.).
Buildings were constructed to enclose the instruments and carrousels.
The collection houses were insulated and refrigeration units installed for
keeping samples cool during the summer. Small space heaters had to be in-
stalled in winter to prevent water which entered by seepage or overflow from
freezing and immobilizing the carrousels.
The paired fan motors used to drive the Coshoctons were normally ade-
quate. During very heavy runoff events, they lacked power to drive the slot
extension up into the flow of water. This contributed to erratic time inter-
vals between samples and variable runoff flows for individual samples as
calculated by the computer from recorded stage heights. Belt slippage may
12
-------
, q*.
Figure 4a. Stainless steel lead pipe and optional sample splitter (watershed 06)
Figure 4b. Stainless steel collection vessels in operation (watershed 06)
13
-------
have contributed to this, but motors also heated excessively and, on several
occasions, burned out during the event.
The motors for the Coshoctons were inexpensive, easily replaced, and
spares were always kept on hand. To avoid excessive attrition, a rain pot
with a water-contact switch was installed initially at each watershed to turn
on the Coshocton wheel after 0.00039 cm of rainfall. This switch was re-
placed in April 1974 with a relay to the carrousel drive circuit. This relay
activated the Coshocton motors after the first pot had filled.
The normal standby position for the Coshocton was with the slot centered
under the lip of the flume (1/5 reduction). The sample splitter was usually
set for a 1/10 reduction. Thus, the nominal reduction for the first sample
of an event was usually 1/50. With the Coshocton turning (1/100 reduction),
the nominal reduction for pots after the first one was usually 1/1000. On
occasion the Coshocton cover was removed, giving only the 1/10 reduction at
the splitter, or the sample splitter was sometimes opened so the nominal re-
duction was that for the Coshocton (1/100).
These reduction settings were recorded in the log for each event, but
it was observed that effective reductions could vary widely from the nominal
setting. During light runoff, the stream coming from the lip of the flume
could meander from one side to the other so that all of it or none of it
might enter the slot of the Coshocton. At any time, debris could lodge
against the sample splitter and change the split materially. In the case
of major events, pots were sometimes composited or some proportion discarded
in sequence in the field, or the collected samples were composited in the
lab or alternate samples discarded to reduce the numbers of samples for anal-
ysis. The disposition of each pot was recorded, and this has helped resolve
many of the apparent discrepancies in computer output between runoff flows
calculated from stage heights and flows expected from the numbers of samples
reported.
During the first few events of 1973, sample collection times were re-
corded by project personnel. Rainfall times were also based on observation
plus U. S. Weather (US WB) Bureau instruments and records for the official
USWB reporting station which is located at the watershed site. Beginning
August, 1973, rainfall times and sample collection times were recorded auto-
matically on the same time scale with a 10-pen event recorder connected by
relays to a tipping bucket rain gauge, to the Coshocton wheel motor circuits,
and to the carrousel drive motor circuits.
Attempts were made to record waterstage heights on the event recorder
also, but the waterstage recorders were disabled by the various hook-ups that
were tried. Thus, our reported waterstage records for each watershed are
based on clock times that are independent of each other and of the times re-
ported for samples and rainfall.
Every effort was made to keep the three clocks synchronized and to log
discrepancies when noted. Nevertheless, time discrepancies for numerous
14
-------
events became apparent in computer output received from AERL. These have
been reconciled by detailed examination of recorder charts and field logs
to give what we consider to be a realistic record of these events as we ex-
perienced them. The time discrepancies would have been virtually impossible
to detect without the parallel time frames supplied by the computer.
In spite of the operational difficulties noted, with a few noted excep-
ions, the data reported under this and the nutrient project is valid. Many
of our difficulties probably are not unique. Some of those encountered dur-
ing winter operations may be useful in design of similar facilities for
winter runoff studies in northerly areas.
SOIL SAMPLE COLLECTION
Soil residue data reported for the summer 1973 and the winter 1973-74
runoff seasons represent four sampling segments for each watershed. These
are shown in figure 5. Their relation to soil types and erosion classes is
shown in figure 6.
Questions arose regarding the possibility that watershed perimeters may
have been altered by tillage practices and that slopes may have changed due
to erosion since the original topographical survey in 1942. Also, it was
apparent that significant slope and soil differences could not be adequately
represented by only four sampling segments.
The watersheds were surveyed again in May 1974. The contours obtained
(figure 7) correspond well with those in the original survey (figure 2) ex-
cept near the perimeters where tillage and erosion had softened the sharp
ridges indicated in the earlier survey. By plowing and discing, a sharp berm
was formed along the original perimeters and seeded to bromegrass just before
the 1974 crops were planted. Breaks in the berm, resulting from harvest op-
erations in the fall, were repaired again before planting in 1975.
At the time of the 1974 survey six sampling segments were delineated
and their areas determined. These are shown in relation to contours in
figure 7 and in relation to soils and erosion classes in figure 8. These
segments did correspond well with observed patterns of wash-off, rill for-
mation, and sedimentation.
Except for pre-treatment and mass balance samplings in the spring of
1973, soils were sampled to a depth of 30 cm (or to the depth of water pene-
tration in dry soil). Seven depth increments were normally taken: 0 to 1 cm,
1 to 2.5 cm, 2.5 to 5 cm, 5 to 7.5 cm, 7.5 to 15 cm, 15 to 22.5 cm, and 22.5
to 30 cm. Ten to 15 cores were composited for each sampling segment (a
larger number of cores was needed for the 1 cm and 1.5 cm increments to pro-
vide sufficient sample for both herbicide and nutrient analyses).
Numerous sampling devices were tried during the summer of 1973. A
shielded, stainless steel sampling probe (approximately 4 cm I.D.) was de-
veloped which worked well after freshly tilled soil had settled. It was not
satisfactory for sampling loose or dry soil, and the volume obtained from
15
-------
East Watershed (006)
West Watershed (007)
Segment
I
2
3
4
Total
Acres
.14
.70
.61
.54
1.98
Segment
I
2
3
4
Total
Acres
.18
.36
.27
.54
1.35
MSU WATERSHEDS
Figure 5. Sampling Segments 1973 - 74
16
-------
SOIL LEGEND 510 Hlllidolt f 10 I
3 I I Tutcolo f 10 I
313 Travlrn f in I
S I 9 Spinki l.f ia.
EROSION
I 0 - 2 3 /.
2 23-75%
H Surf oc
deposition
MSU WATERSHEDS
Figure 6. Sampling Segments 1973 - 74 (With soils overlay)
17
-------
t-i
01
o
o
H
&
td
s-l
&0
o
tx
o
4J
H
S
0)
0)
CO
ex
s
CO
0)
S-l
00
H
18
-------
o o
01
c?
a.
DC
O
O
4-1
H
O
CO
J3
4J
H
C
cu
e
CO
CO
0)
M
00
H
Pn
CA a.
19
-------
10 cores for the 1 cm and 1.5 cm increments did not supply the sample needed
for all analyses.
A series of stainless steel "cookie cutters" was designed for sampling
the 1, 1.5, and 2.5 cm increments. Large stainless steel spatulas (7.6 by
30 cm) were inserted into the soil to form a 3-sided frame from which the
"cookie sections" could be taken with minimal contamination, even in dry,
loose soil.
Several special samplings were made with these sampling tools to derive
standard bulk density values for mass calculations.
RUNOFF SAMPLES
Each sample (approximately 3.6 liters) was passed through a stainless
steel three-way splitter. One fraction was used for pesticide analysis,one
for sediment analysis, and the remaining one for nutrient analysis.
The water and sediment phases were separated by passing each sample
through a 0.45 u millipore filter for nutrients or a Whatman 42 filter for
pesticidesthe sample being retained on the filter is hereafter referred
to as sediment, and that passing the filter is hereafter called water.
RESIDUE ANALYSIS
The analyses for paraquat, diphenamid, atrazine, and treflan (See appen-
dix table 1) were done in accordance with methods published by EPA workers
(Payne et^ _al., 1974, Pope and Benner, 1974). Special instrumentation,
methods, and parameters are noted below.
1. Paraquat was analyzed by scanning between 370-470 nm and
measuring the maximum peak height (above baseline) at ~394 nm
on a Beckman DBG spectrometer using 4 cm cells. When sediment
samples were less than 1 gram, sufficient adjacent samples
in time sequence were combined, where possible, before the acid
digestion step and average values reported. Blind spike samples
were run periodically to ensure uniformity among workers. The
minimum detectable amount was 100 ng.
2. Atrazine, diphenamid and treflan were analyzed on a Beckman
GC-4 adapted with a Coulson electrolytic conductivity detector
operating in the nitrogen mode. The column oven temperature
was set at 205° C for atrazine or trifluralin and 220° for diphena-
mid. The injection port was set to 270° C and the furnace was held
at 850° C.
Minimum detectable amounts (signal/noise =5) were 5 ng for
20
-------
atrazine and treflan and 30 ng for diphenamid. It should be
noted that soil samples were deactivated by equilibrating with
water before extracting for these three chemicals.
When sediment samples were less than 1 gram, sufficient ad-
jacent samples were combined, where possible, before G.C. analysis
and average values reported. Blind spike samples were run periodi-
cally to ensure uniformity among workers. Mean recoveries from
spiked samples are given in Table 1.
G.C. signals were fed to PDP-8/e computer via an AFC-8 A/D
interface. PAMILAR software permitted the computation of the data
to peak area, height and retention times numbers which in turn were
fed to a PDP-11/40-RSTS computer for storage and subsequent mani-
pulation for report generation.
21
-------
TABLE 1. RECOVERY OF CHEMICALS FROM SPIKED SAMPLES.
Chemical Added
A. Spiked soil samples
Atrazine
Diphenamid
Paraquat
Trifluralin 1
B. Spiked water samples
Atrazine
Diphenamid
Paraquat
Trifluralin 1
C. Spiked sediment samples
Atrazine
Diphenamid
Paraquat
Trifluralin
ppb
20
100
300
,000
1
1
100
,000
20
100
300
t
Recovered
%
85± 5
82± 6
90± 5
104± 3
85± 5
80± 5
84± 6
98+ 10
80± 7
80± 8
90± 3
Added
ppb
100
500
1,500
2,000
5
5
500
2,000
100
500
1,500
t
Recovered
%
86± 5
89± 5
91± 4
89± 4
90± 4
88± 5
85± 6
97± 10
90± 5
92± 5
92± 2
t No estimates of recovery from spiked sediment samples were made for trifluralin.
22
-------
SECTION V
RESULTS AND DISCUSSION
INTRODUCTION
The principal objective of this project was to furnish Athens Environ-
mental Research Laboratory, EPA, with data which will allow the evaluation
by systems analysis of movement of pesticides in soils and agricultural land-
scapes. In that the data from Michigan State University is only a portion
of the data input, this section will deal principally with observations that
are necessary to relate our data to soils and topography on the two water
sheds and to weather conditions during the period of the study. Detailed
discussion and conclusions that may ultimately be drawn from the data will
not be covered in this final report.
A list of the field operations for each watershed is presented in
table 2. Data from the study has been submitted at various times. A key
to reports where the data from this study can be found is given below.
First year Summary, June 1973 to May 1974
Rainfall records 06-16-73 to 05-16-74
Runoff records 06-16-73 to 05-16-74
Pesticide residue data 06-16-73 to 05-16-74
Soil core data 05-24-73 to 05-02-74
Quarterly Report, May 1974 to September 1974
Rainfall records 05-17-74 to 08-31-74
Runoff records 05-17-74 to 08-27-74
Quarterly Report, October 1974 to December 1974
Rainfall records 09-02-74 to 11-28-74
Runoff records 09-12-74 to 11-05-74
Pesticide residue data 07-02-74 to 09-12-74
Soil core data 05-22-74 to 09-03-74
Quarterly Report, January 1975 to March 1975
Monthly USWB reports May 1973 to March 1975
23
-------
4-1
p!
CU
f|3
o
u
r>^
O
e
^
y^Q
O
CO
o
H
EC
CO
P4
H
H
Pd
O
Pn
cr.
53
0
M
H
Pd
fj3
P-i
O
Q
"] pj
W 0
PK 4J
c«
PM M
o cu
ex
H O
1 1 T3
J rH
CU
H
FT i
CN
cu
rH
,0
ca
H
01
4-1
cO
Q
O
rH
c"j
4J
rl
'
rrj
Q)
CO
3
CO
cfl
&
cu
0
M CU
CXrH
ex
60 S
C3 cfl
rl CO
rH
f 1 1 ^
B cu
co ex
CO
CO
rH CU
H M
0 O
co o
g
o
M
HH
13
CU
4-1
O
01
rH
rH
O
O
CO
cu
i-H
ex
cO
Cfl
t-H
H
O
CO
4J
C
Ol
B B
rl O
01 m
ex
cu
1 0
CU 4-1
M
PH O
01
i-*.
1
CN
1
m
O
a
rl
o
o
CO
cfl
ex
0
o
CN
f^.
CTs
rH
(-*
4-1
ex
cu
T3
B
a
in
CM
o
4-1
rrj
CU
[3
o
rH
ex
CO
rrj
cu
J3
CO
M
cu
4-1
CO
CO
f^
1
rH
0
1
^O
o
f{
4-1
O
O
B
CO
o
4J
fj
4J
0
0
4-1
60
C
H
jj
f-X
w
r^H
-U
*H
Jj
T3
cu
O
M
i-l
cfl
K
00
r^
1
0
1
O
CO
cfl
cu
4-1
cfl
ftf
rH
u
00
(3
rl
4J
CO
CO
o
*O
cfl
o
i-l
0
4-1
J^
O
H
j-i
ex
T3
cu
S
o
rl
cd
PS
OO
r^
1
0
1
vO
O
B
o
co co |
J3 X > 00 T3 T) TJ
HO O CU (3 O
Ml S rH MOOCUCO CUB
cu cu co cn
ex .^ o rC B ro in o co
4J 4J OCUCUrH OCfl
W "S 60 'S CJ^Cr>. rH>
CU -H C O rH Til 4-1
4-lrH -H T3 rHCXjSC CM
H a. >-, cu o cu cci
i-nex cfl co ccNrHcu
CO M 3 -HrHCUjS >-,
m ex i j2 4J j-i ^-N
rHOI CO CO T3CN CU T3O
CVJM 'Cfl CUrHBrQ rH
3 60 rH & CO | O >-,
CCO-C'H 3CN CU M
H CO in -H O CU rH OO T3 CU +1
CU ^3° ^ Cfl t"> CU cfl ^s> CD
0) M O O O M<4HOOB rH
>purHrHj:M Old CO-
H rH4JCX'&T3CO COrH
4-ICMT3O-H 4J CflCcfl &
O B C 4H ^ 60 O *^ rj cfl 3 ^ ^,
Cfl O Cfl CJ i 1 CU ^^. rH 4-1 ^Cj
^ t^a-HCXCUMCU (3 -H-HC
CflOOm rH'H rH C04iiT3 OCO OCO-H
r^ K/t ^~ cu i i ex M »H *H M co c [5
^^ CJ\ 4-1 CO B CU Cfl ^" CO 4-1 3 CU
OOCNO COM cflCXrCCN cflO CUT3>>
fcrfOO -H3 Cfl ~~,CNCU CJJ3 OrH
rj rrj f ) CO 60 rj »H cfl ^ CU
CMCNCU 0)l*H rHCU WJS4J rHO l4-|rH4J
rHCUg'H'HM 4-1 CXin M3CO
4J|S BM OO r>-'HO CXCX3 3^3M
rHCtJ4J H4-1 COO VO |$ 4-> <3O CO ' CU
1
^
/-s CU 1
M J3 "d -H
O CJ CU 13 4-1
4-1 r£3 V*-/ K^» Cfl
CO 4-1 Cfl T3 Cfl
4-1 B O 4-IMCUJ3 S
co -H o ja coexx-^ o
T3 60 3 CO -H 60
CU Cfl VO C CJ1 B r^l iQ
rC O Cfl CU
Cfl rH r^« M M CO (^- ^^
M cfl T3 cfl CU J2 00
Q) 4-1 O CU CX5~^'4JO
4-1 C4J4J COC-OC04-1
CO O C TJ M M 3
& N o cfl ccucucucro
H rH Cfl 4-1 4-1 > CO
J2 M S CX -H -H -H M B
4-1 O O /-srHrH60COO
O 42 M CU PH"~- CXM
-O ^ M-l M £2 60 O O 4H
CU ^ 00 4-1 CO
C M T3 J3 S^^fJlT)'
O O 0) CO ^- CU O)
H rH'-OrH CO 43r.t^-60cNrH'O
rH rH'rH ffl OT30^rHrH
CX rH r^ O !-> -H O MH
CX -H O CX3T3CNrHOO
cfl H1*-! M tscr'CoO
O Cfl CO S-''HCO''0 COCO
C /*s rt Q) ^ ^^ O^J Cj CJ 4^
rH r^» J ' r^ f ^ N^X yj nQ \^ f^ ^^ ^
rHO -rl4-l CX T3 -HB^O4J CXOI
CO J?CXS CflCU BrHCdTJBB
MT3 CUCfl C JS cO.jS-HCflCU
3C rdra co coco ecyi4-»>>B WVJ
rHcfl o) CUM OICNOMcO O
14-1 |-H cfl rH rt fU r] o ^) rj , | pj
H VO rH -H >. 4J CX fi 3 CU -H -H
MO -rlO O Ocfl -HOCrHX! O
H ' H4J c/i co|s oMOcnex co-*
00 00
r^1^ r*"*
1 1
oo o>
o o
1 1
<^O ^-O
o o
24
-------
4J
».
fi
0
u
a
CS|
0)
rH
3
H
4-1
£3
O)
3
0
a
0
H
4->
CO
0)
CX
O
-o
rH
0)
H
p£4
O)
CO
Q
-^
C
O
H
CO
3
MH
M-l
H
*rj
J-I
O
M-l
x)
0)
4-1
o
01 co
rH 0)
rH vH
0 13
0 3
4-1
CO CO
01
CX 4-1
6 -H
CO rH
CO iH
4J
rH Cfl
H rH
O O
CO >
CO CO
r^» r^
I 1
00 O rH
O 4J rH
VO ^D
O O
LO
O
? 1 r^-
4-1 4-1
H
4-)
0) CU
cn B
j_!
co o
cfl £3
[3 *r-|
QJ **}"
o
O *>
J-< 0)
CX rH
CX
t>o B
e cfl
H cn
rH
0 0)
CO CX
CO
CO
rH 0)
H J-i
DOS
CO O O
CO
4-1
£3
0)
^
0)
M-l
MH
O
£3
3
JH
O)
4J
MH
cfl
13
0)
rH
CX
Ctj
CO
CO
rH
H
O
CO
CO
r^
1
oo
rH
^D
O
in
o
rH r-.
4-1 4->
H
4J
0) O)
co B
3 CU
CO O
cfl £3
3 -H
0) ^f"
43
O *
J-i 0)
CX rH
CX
M B
(3 cfl
H cn
Q 0)
cfl CX
CO
CO
rH 01
H J-i
oos
CO O O
CO
4-1
£3
0)
^
0)
MH
MH
o
£3
3
J-I
Vl
0)
4J
MH
cfl
13
01
rH
CX
B
cfl
CO
CO
rH
H
o
CO
CO
f-x
1
-d-
O
1
O
o o
rH CO
4-> 4-1
H
4-1
<1) 0)
co 6
Jj
co o
cfl C3
S -H
O) r^»
o
o >
J-I 0>
CXrH
M B
C cfl
H CD
rH
B o>
CO CX
CO
CO
rH CU
H J-i
oos
CO O O
cn
4J
C
cu
^
cu
MH
MH
o
£3
3
J-I
^
CU
4J
M-l
cfl
TJ
0)
rH
CX
B
Cfl
CO
CO
rH
H
o
CO
CO
r^
1
rH
rH
oo
o
0)
cfl
1 1
o
CO
cfl
[5
cfl
jf*
*-*.
W)
*y{
o
p**
(N
r>
rH
MH
o
n
rH
cu
H
T3
0)
>
o
s
0)
J^
s
cfl
J^
4J
CO
T3
C
cfl
s
13
0)
4-1
CO
0)
>
J-I
cfl
43
01
J-I
0)
s
cn
£3
cfl
0)
43
o
CO
CO
r~»*
1
rH
ci
rH
13
§
O
0)
o co
£3
O
0)
cfl
O
CO £3
cfl cfl
O
C O
O W
H O
4-1
cfl C
o cu
rl 0)
CX 4J
CX 0)
< 43
O
O O
is cn
cu cu
co S
3 §
CO £3
!» -H
43
O «
J-i CU
CX
60 B
cx j^
S cu
cfl cx
CO
CO
rH 0)
H I-l
o o
co o
4-J
CO
cr
cfl
cfl
cx
13
cfl
/^
^s
LO
£3
43
O
r- >
CX
:=>
13
H
s
£3
0)
43
d, 0.24 Kg/liter active)
*H
3
cr
H
rH
6-S
rH
<^
CM
ti
0
J-I
3
CX43
H
Q
1
LO
0
I
rH
rH
U
*~s
06 & 07 watersheds in
7 1/ha at 2.82 Kg/cm2.
(~i
4-1
O
43
£3
O
13
0)
H
rH
CX
CX
cfl
Q)
J-I
0)
S
o>
CO
4J
Cfl
?*>
J-I
J-I
3
rH
cn
13
cu
X
B
cfl
£3
0)
CO
4-J
0)
M
CO
0)
cx
cfl
cn
rH
0
CO
0)
o
rj
Cfl
rH
Cfl
43
CO
CO
cfl
s
rH
Cfl
H
4-1
H
a
M
as thawed .
43
! 1
H
O
CO
J-I
CU
4J
MH
cfl
S
Cn
£3
H
13
0)
4J
0
CU
rH
rH
0
O
CO
CU
rH
CX
E
cfl
CO
rH
H
0
CO
1
m
o
1
CO
o
e soil samples were
o
rt
rH
Cfl
43
03
cn
cfl
s
rH
S
-H
j^i
0)
H
1
CM
O
1
un
o
-a
Q)
4-1
o
0)
rH
rH
O
O
25
-------
,
4J
"(3
0
CJ
CM
01
rH
43
ffl
H
4-J
c
01
£2
3
O
O
C3
O
H
4-J
cfl
5-1
CU
CX
O
TJ
rH
Ol
H
En
0)
4J
tfl
Q
w
cfl
43
W)
}s^
sj"
*3"
rH
TJ
a
cd
p_i
d
^
t>0
^
r^-.
r*.*
13
* CU
IS *H
rH
cfl CX
43 CX
60
**St Q)
5-1
r~- a)
cfl
O
4-1
C
H
13
CU
,-_4
rH
H
4J
13
a
CO
13
0)
H
rH
CX
CX
cfl
5-1
01
N
H
rH 43
H 4-1
4-1 CX
5-1 CU
0) 13
4-1
e
a o
H
co in
cfl
pq 1*^
r-~
1
O
CM
1
m
o
CO
T3
O)
CO
0
o
o
*
CT*
in
o
4-1
CX
0 0
O 5-1
o a
«\
-* 43
m o
cfl
. 01
co
IS 5-1
0 0
£ 03
a 43
r^ 5-1
O 0)
rH CX
/"N s~~\
O .V
oo i-<
i^* n3
CO pg
^J
(IJ CO
CU (3
p3 cfl
O CU
H 43
P-c >%
--' o
CO
a
M O
O 4J
O
13
O 0)
4J 4-1
fi
13 cfl
01 rH
4J pi
a
tfl r^
rH O
CX
rrj
^O CU
O 43
CO
*O 5-1
0> 0)
43 4-1
CO tfl
5-1 IS
CU
4-1 13
cfl C3
S cfl
;*
i
rH
CM
1
m
o
4-1
CO
3
cr1
tfl
5-1 O
cfl CO
CXO
rH
i*1
4-1 Tj
H £3
|S cfl
13 in
0) -
(3
H
N O
cfl 4-1
^i
4-1 13
cfl 0)
H
-a to
H 01
4J &
O
Cfl x-^
CU
cfl >
43 'H
-» 4J
60 0
!^ tfl
CM CO
rH 43
rH 60 O
13
4-1 O"\ 01
cfl 0- 43
3 CD
CJ4 vf 5-1
cfl 0)
5-1 4J 4-1
Cfl Cfl Cfl
PM ^ S
£
1
CM
CM
1
O
O
CO
O
rH
13
a
cfl
m
^J-
f"^.
o
13
CU
Ol
&
4-1
0)
T3
CU
H
rH
CX
CX
4J 13
Cfl 0)
3 43
o- co
cfl 5-i
5-1 CU
CO 4J
CX cfl
13
13 0
n) 4-1
/*"N T3
cu cu
r *r">
H rH
4-1 CX
O CX
cfl CO
cfl CU
43 to
--. CU
00 &
, ^
r^ cu
co >
-H
CO 4-1
a
4-1 tfl
CO
cfl
T) 43
H -^.
6 00
cfl £4
CU CM
43 rH
CX
H rH r~-
n ^o
rH
H
o
Cfl
0)
43
4-1
e
5-1 0
0
M-I m
13 r->
o>
CO <4-l
3 0
CU 43
5-1 4-1
Ol CX
& CU
13
^
CO CO
^1
0) O
4-1 4-J
4J
3 CO
O 4-1
13
0) CU
*H &
K/l Ol
O S-i
o a
u c
r -H
13
CU
4-1
O
CU
rH
rH
0
O
Cfl
0)
rH
CX
Cfl
CO
rH
H
O
CU
p r~H
0 CX
5-1 S
«W Cfl
CO
CO
Ol 5-1
rH 01
CX CX
j~
ctf CO
CO O)
I-l
S o
o a
5-1
M-l C3
cu
13 H
CU
CO
3
S
cu a
to a
CU 0 Q)
S co &
Cfl
CO O 4-1
0) 4-J
43 Ol
o m 5-1
IJ 01
cx r^ S
cfl
J5
43
4-1
0
43
e
0
5-1
M-l
13
0)
4-1
O
Ol
rH
rH
O
O
Ol
5-1
01
J5
CO
Ol
rH
CX
e
cfl
CO
CO
rH 13
H 01
0 43
Cfl CO
;*
i
o
CO
1
in
o
s^-
1^-fc
1
CM
CM
I
m
13
0
CO
cfl
13
CU
4J
O
0)
rH
rH
O
1
5-1
CU
4-1
cfl
Jj
43
4J
0
43
3
0
5-1
"4-4
-rj
0)
4^
a
0)
rH
T 1
0
o
Ol
5-1
0)
J5
to
0)
rH
CX
0
tfl
CO
CO
rH T3
H 01
0 43
Cfl CO
£
1
CO
o
1
0
v^-
f^
1
CM
CM
1
m
o
13
0
CO
Cfl
13
CU
4J
O
01
rH
rH
O
O
1
5-1
01
4J
cfl
^
43
4J
O
43
g
0
5-1
14-4
13
O)
4_J
O
cu
rH
rH
O
CJ
0)
5-1
Ol
^
CO
CU
rH
CX
S
cfl
CO
CO
rH 13
H CU
0 43
Cfl CO
£
1
in
0
|
oo
o
>^-
r^.
I
CM
CM
1
m
o
13
0
CO
cfl
13
CU
4-1
O
cu
rH
rH
O
o
1
5-1
CU
4J
cfl
IS
43
4J
0
43
CJ
0
to
14-1
13
CU
4-1
O
CU
rH
rH
0
O
CU
5-1
Ol
IS
CO
Ol
rH
CX
CO
CO
CO
rH 13
H 0)
0 43
Cfl CO
pj
1
rH
1
OO
O
26
-------
J
"c3
O
O
CM
cu
rH
rQ
cfl
H
4-1
C3
CU
B
o
u
C3
O
H
4J
cfl
CU
CX
0
13
rH
0)
H
Pn
01
4J
cfl
«
cfl
43
CO
pj
0
4-1
O
rl
H
4J
01
e
00
r^*
CM
co
cfl
Jj
13
rH
CU
H
vO
O
a
0
MH
13
0>
4J
CO
CU
>
rJ
Cfl
43
01
60
CO
rH
H
co
C3
\-t
0
O
s^"
[**.
1
CM
1
0>
O
cfl
43
60
ON
rH
O
*
CM
CO
cfl
&
13
rH
cu
H
f^.
0
rJ
13
CU
4J
CO
0)
>
Cfl
43
CO
C3
Cfl
CU
>-,
0
C/3
>^-
p^,
1
OO
rH
1
O
rH
13
cs
ttJ
cfl
43
'Z
60
O
CO
rH
CO
cfl
13
C3
O
rH
4J
CO
a
H cfl
rH 43
CX ^.
CXrH
cfl O
rH 60
cO ^
3
4J rH
cj 0-
0)
H -rt
4-1 rH
O CX
Cfl CX
Cfl
Cfl
rC3 S~**
^ CU
60 >
W *H
4-1
in a
CO Cfl
rH Cfl
4-1 60
cfl W
3
cra\
cfl vO
M
CO CM
PM ^
~^f
pX.
1
oo
o
I
rH
rH
*^~
|^
1
CM
CM
1
£j
O
CO
cfl
T^
0)
4-1
0
01
rH
rH
O
o
1
cfl
cfl
PL, f-x
o
TJ
C3 0
Cfl 4-1
s-^ Tj
ai oi
> -H
H rH
4-> CX
O CX
cfl cfl
CO s~ i
43 01
^~ > 13
60 'iH 0)
^ 4J 4J
a a
r-- cfl cu
rH
CO Cfl i I
43 0
4-) -^ a
cO 60
^-' W CO
0)
Tj CO rH
H CM CX
a e
Cfl rH Cfl
C3 ^ CO
01
43 4-1 rH
CX CO -H
H 3 O
Q CT CO
^
4J
H
co
C3
CU
T3
^jyj
rH
3
M
O
<4H
13
0)
4-1
O
Q)
rH
rH
O
O
CO
cu
rH
CX
0 CO
cfl *H
co co
K*"»
rH rH
H Cfl
0 C3
W cfl
«^-
r*x.
i
CM
CM
1
rH
rH
13
CU
rH
CX
6
Cfl
CO
e
0
rH
O
4-1
O
13
cu
4-1
0
cu
rH
rH
O
o
CO
CU
rH
CX
e
cfl
CO
rH
H
O
CO
in
P*.
i
o
1
CM
O
*^-
r**«.
i
CM
CM
1
m
o
C3
O
CO
cfl
T)
CU
4-1
a
cu
rH
rH
0
u
rH
CO
C3
H
e
S-l
0)
4J
Jb4
o
14H
*rj
0)
4-1
a
cu
rH
rH
O
O
co cu
0) CJ
CX cfl
cfl cfl
CO 43
rH CO
H CO
O cfl
C/3 0
m
I
oo
i
I
m
o
CM
r--.
rH
*
Cfl
Cl
PM
60
>y^
cfl
rH 43
CO --»
rH rH
U
S
Cfl 60
_rt sx
^s.
£5 \O
LO
60 rH
^
'rJ
co a
VO Cfl
13 cfl
0) 43
rl ^
, 1 1 t^*
CX
CX 60
cfl 01 -H
r-H nH
4J M rH
CO O -rl
CO & 4-1
O
13 13 13 CO
01 CO C3
& O cfl 43
O M 4-1
rH 43 rH -H
cx u is
CU W
CU M 6
M CU 13 0
CU S C
S co m
CO
co u r-.
13 0) /~v
0) N J- HH
43 -H O 0
CO rH CL,
H -H CM 43
01 4-1 ffi 4-1
4-1 rJ ^ CX
CO O! cfl 0)
!3 PL, U 13
m m
r*^ r*"*
i i
v^ \^
rH rH
1 1
m m
o O
cu
cx
CO
*rj
CU
0)
CO
O
o
o
o\
m
o
^
0 0
O CO
0 0
4J 4-1
13 13
0) 0)
4-1 4J
c3 C3
cfl cfl
rH rH
CX CX
\o r-*
O O
*"O *r3
cu 01
43 43
co co
M f-i
O) O)
4-1 4-1
cfl cfl
& S
27
-------
-W
»
c
o
u
*
CN
0)
rH
s
H
4J
e
CU
3
o
u
e
o
H
J-l
Cfl
rJ
CU
P.
o
T3
r-H
CU
H
fe
0)
4-1
cfl
O
01
c
H
N
0)
t-i
4J
Cfl
vO
TJ O
C
CO O
4-1
XX
Q) T3
> Q)
H -H
U rH
0 P.
cd P.
cfl
CO
J= /->
^ cu
60 >
W -H
4-1
OO CJ
(N Cfl
rH Cfl
^ j ( i
^^
4-1 00
Cfl ^
3
crvo
CO m
M
ca ex
H CO
4J
O s^
cfl CU
>
cfl -H
rC 4J
^ 0
bO cfl
Ni
CO
CN J3
rH ~^
W}
CO W
4J ro
cd o
<*-s
rH
T3 '-'
H
e j-1
cfl cfl
C 3
O
O
CO
rC
25
tod
W
<)
^o
CO
O
Z
J-
JT]
^
J3
4J
H
S
CO
cu
CO
CO
u^
t^
1
in
CN
^o
o
-------
Quarterly Report, April 1975 to June 1975
Rainfall records 12-01-74 to 05-30-75
Runoff records 01-08-75 to 05-30-75
Monthly USWB reports April and May 1975
Soil core data 11-08-74 and 02-03-75
Quarterly Report, July 1975 to September 1975
Rainfall records 06-01-75 to 09-29-75
Runoff records 06-05-75 to 08-31-75
Soil core data 05-08-75 to 06-02-75
Monthly USWB reports June 1975 to September 1975
Special Report, December 22, 1975
Pesticide residue data 01-08-75 to 08-22-75
Soil core data 06-06-75 to 08-27-75
Special Memo, February 1976
Discrepancies in first computer printout for period 06-03-73
to 11-28-74.
Special Report, September 1976
Field Notes on Crop Development: 1973, 1974, 1975.
Special Memo, October 1976
Reconciliation of computer printout received June 1976 with
transmitted data and all available records and field logs.
In this section we have two objectives: (1) to describe soil conditions,
topographical features, and patterns of runoff and sedimentation which would
have influenced profile distributions and mass balances within sampling seg-
ments and lateral redistribution from one segment to another; (2) to describe
weather and watershed conditions associated with a sequence of winter runoff
events in 1975 and with two major historic events in spring and summer 1975
(one without crop cover and one with mature crop cover).
Rationale will be presented for resolving some of the apparent anomalies
in our soil core data for pesticides through September 1974. Corrected run-
off data will be used to show effects of freezing and thawing on winter run-
off losses and to show unique differences between the two watersheds in their
susceptibility to losses of sediments and chemicals.
Runoff and Sedimentation Patterns
The two watersheds differ in size and are uniquely different in topo-
graphy (Fig. 2). Soil differences (Fig. 1) are minor except as they relate
to topography. The Tuscola fine sandy loam in the NW corner of Watershed 06
29
-------
was developed in somewhat finer parent materials than the Spinks, as was the
large area of Hillsdale on the east side of Watershed 06. However, subsoil
materials exposed on 8-12% slopes of Hillsdale and Spinks are similar (cf.
soil descriptions).
The large area of Traverse fine sandy loam in the central basin of Water-
shed 06 has no parallel in Watershed 07. The Traverse is a depositional soil,
formed in sediments eroded from surrounding slopes. Sedimentation patterns
observed in this area during the three growing seasons and two winter periods
of this study will be described later in this subsection.
The plow layer in the more severely eroded areas of Spinks, Hillsdale
and Tuscola (8-12% slopes) has a coarsely granular, open structure. Where
freshly plowed, these slopes have a high infiltration capacity. Under rain
action, infiltration is reduced by slaking at the surface and by the forma-
tion of a weak, thin crust (2 to 3 mm).
As observed in our core samples, the slaking action extended to a depth
of about 2 cm. A single grain structure was found below the surface crust
in our 0-1 cm depth increment. In the 1-2 1/2 cm increment, there was a
gradation from fine to coarse aggregates. Below 2 1/2 cm, an open., porous
structure was retained through the plow layer (25 cm) from one season to the
next. Some consolidation did occur because, two to three weeks after plow-
ing, probe sampling below 7 1/2 cm was virtually impossible unless the soil-
was moist.
Substrata below the plow layer in all areas drain freely, although local
variations in deep drainage may occur due to discontinuous textural bands
of finer materials. A 5 to 20 cm layer of silty clay loam is encountered
at depths of 75 to 150 cm in and around the central ridge between the two
watersheds (E. P. Whiteside, personal communication). Also, in the spring
of the year, a perched water table approaches the surface at lower eleva-
tions near the catchmentsa fact that was brought to our attention on
0514-74 when the tractor nearly mired making the last two passes with the
plow in front of the catchments.
The porous internal structure of the plow layer on eroded slopes, com-
bined with freely draining substrata, would be conducive to downward dis-
placement by sifting and percolation of fine materials released by slaking
at the surface. We think that this explains, at least in part, the unexpec-
ted downward movement of paraquat in the profile, which is indicated by our
data for segments which include severely eroded slopes (segments 3 and 4 in
1973, Fig. 6, and segments 01 and 03 in both watersheds, 04 in Watershed 06
and segment 06 in Watershed 07 in 1974 and 1975, Fig. 8).
Several light rains or a single rain of moderate intensity would serve
to slake the soil surface and smooth irregularities left by tillage opera-
tions so that runoff from eroded slopes could occur readily. Even a light
rain (5-minute intensity of 0.02 cm/hr ) could produce runoff from upper
slopes if it continued for several hours.
Frequently, rains of moderate intensity and short duration would produce
30
-------
runoff from upper slopes, but sediments picked up on upper slopes would be
intercepted on intermediate slopes before reaching the central draw or the
gentle slopes of the central basin. During winter runoff events, lingering
patches of snow on intermediate slopes were most effective in intercepting
sediments from runoff water seeping through them or spreading laterally to
flow around them.
Evidence of sedimentation on intermediate slopes was quickly erased by
later events so we discounted it in laying out our sampling segments. How-
ever, examination of our soil core data through September 1974 leads us to
believe that substantial movement of sediment from upper to intermediate
slopes did occur within sampling segments 3 and 4 in 1973 (Fig. 6) and with-
in segments 01, 03, 04, and 06 in summer 1974 (Fig. 8). Random sampling
within these segments would have weighted our composites unduly in favor of
depositional intermediate slopes. As a result, our data show increasing
total recoveries for these segments over time instead of decreases as would
be expected. Depositional areas within these segments include eroded slopes
with open, porous structure in the plow layer so that our sampling bias is
reflected at depths greater than 7 1/2 cm as well as in the upper increments.
The apparent sampling bias in eroded segments was greater for paraquat
than for other chemicals, as would be expected because of its greater persis-
tence and its total affinity to the particulate phase. Total recoveries for
both watersheds in 1974 exceeded the total applied. A similar result would
be expected for phosphorus and should be looked for in our nutrient data.
To understand patterns of erosion and sedimentation on these watersheds,
it must be recognized that crops were planted east and west, at approximately
right angles to the major slope and central drainageway. As a result, even
during the winter, runoff from upper slopes followed row middles east or west
toward the central draw of each watershed.
Because of the short E-W slopes, erosion down crop rows was mainly sheet
erosion. Only occasionally were small rills (5 to 8 cm deep) cut in the
track left by the covering discs on the planter. The principal occasion when
this occurred was during the heavy rains of 04-18-75 when ponded water out-
side the berm broke through at several points along the east side of water-
shed 06.
When row middles at upper elevations in the central draw filled to over-
flowing, cross-row rills would form quickly and produce a rapid cascading
discharge onto more level areas down slope. Deep cuts (10 to 30 cm) could
be produced very quickly. These might extend across ten or more rows before
reaching a point where sufficient ponding in row middles could occur to slow
the flow of water.
On slopes of 4% or less (Fig. 2), impounded water might spread several
meters up and down the row before breaking through into the next row middle.
At this point, a new deep rill or gully might form, or simply a succession
of mid-row ponds connected by shallow rills cutting across the ridges left
by the planter.
31
-------
At points where discharge from a rill or gully entered an area of
ponding, the heavier sediments would be dropped quickly. Conspicuous depos-
its of light-colored fine sand would be left, extending up and down the row
and sometimes in two or three successive rows below the point of discharge.
Near the extremities of these deposits the light-colored sands graded abrupt-
ly to darker colored, very fine materials which blended quickly with the soil
so that the limits of their lateral or downslope distribution could not be
ascertained.
Another important feature of cross-row erosion in central areas of both
watersheds is that deep rills and gullies were quickly obstructed by debris
intercepted by plant roots or stubble. During the course of a major event
or during subsequent lesser events, even a deep gully extending through the
plow layer could fill with sediment.
Sediments deposited in rills and gullies were mainly the heavier sand
fractions. These were less susceptible to cutting than unsorted soil. Suc-
cessive episodes of cutting would start at different points along the E-W
axis and at different points along the S to N slope of the draw. The result
was a random, meandering pattern of alternate cutting and filling along the
central NW-SE axis of both watersheds.
Usually, rills would cut no deeper than 10 to 20 cm before filling again.
However, during the near-record rains of 04-18-75 and 08-21-75, cutting at
several points in both watersheds extended through the plow layer to depths
greater than our 30 cm sampling. Gullies left open at the end of these
events had not been filled by the time of our terminal mass balance samplings
of 05-08-75 and 08-27-75. Those left in April were covered by plowing in
May; those left in August had largely filled with sediments by harvest time
in October.
Cross-row rills were not wide, usually no more than 10 or 15 cm. How-
ever, their meandering pattern and random distribution would have influenced
depth distributions of herbicides and nutrients over significantly large cen-
tral areas in both watersheds.
The sediments deposited in rills and gullies were mainly light-colored
fine sandin other words, the least adsorptive soil fractions. However,
there were textural bands of dark-colored finer materials, varying in thick-
ness or frequency, as well as occasional slumps of soil from the sides of
the rill or gully.
The random distribution of cross-row rills and gullies and the strati-
fied variability of sediments deposited in them must be considered in inter-
preting changes in profile distribution of chemicals and nutrients from one
sampling to the next. In particular, some of the date-to-date variation be-
low 7-1/2 cm may represent a random weighting of our composite samples by
cores taken in sediments deposited in deep rills.
On the other hand, changes over time in the upper two or three sampling
levels (depth increments) will be influenced by the sorting out of heavier
sediments on the surface along lateral mid-row sedimentation fans at points
32
-------
where cross-row rills discharged into areas of ponding. These light colored
surface deposits were generally thin and limited in extent at the higher ele-
vations corresponding to segments 0602 and 0702 in Fig. 8. Only limited
ponding could occur in these areas because of the steep E-W slopes. In these
areas, visible sedimentation was associated also with down-row runoff from
east and west. Sediments from steep slopes in segments 01 and 03 contributed
to rapid filling of cross-row rills and gulleys at these higher elevations
in the central draw.
Opportunities for ponding increased markedly on slopes of 4% or less,
beginning at about the 894 ft contour in watershed 06 and the 890 ft contour
in 07 (Fig. 2). At lower elevations, patterns of cross-row cutting and
lateral sedimentation were uniquely different on each watershed and varied
from season to season.
In watershed 06, extensive ponding can occur in row middles on the large
area of 2-3% slopes. Ponding was infrequent during summer 1973 and the fol-
lowing winter. Cross-row cutting at higher elevations was light also, and
the heavier, light-colored sediments from these rills were deposited in a
limited area along the central draw in segment 2 of Fig. 6. Meandering
shallow rills were formed in Segment 1, but conspicuous sorting out of
heavier sediments occurred in only a few rows and the lateral surface depos-
its were thin.
The following summer (1974), runoff flows were somewhat heavier, cut-
ting was more extensive, and sorting of sediments was observed all the way
to the catchment. At the time of the 09-0374 mass balance sampling, sur-
face deposits of light-colored materials were generally thin and scattered,
but, at several points in segment 0605 (Fig. 8), they extended for several
meters up and down the row and were 1 to 2 cm thick.
During winter runoff events, soils in the Great Lakes Basin which are
not under vegetative cover are, nonetheless, very frequently protected
against deep cutting by surface ice or by the presence of deep frost in the
soil. On the other hand, pickup of sediments by running water is facilitated
when bare soil on the immediate surface thaws, since internal drainage cannot
occur into the frost layer below. The thawed soil tends to remain saturated
and readily suspended.
In permeable soils on sloping areas, saturated soil over frost can
drain quickly by lateral seepage as soon as snow to supply melt water disap-
pears. Seepage from upper slopes serves to keep thawed soil further down-
slope saturated. During winter thaws, night temperatures usually fall below
freezing, and thawed surface soil freezes each night. The ice which forms
in saturated soil thaws less readily the next day. Over a succession of
daily snow melts with nightly freezing, ice builds up on the soil surface.
Soil on intermediate and lower slopes is protected, thereby, from signifi-
cant erosion. Lower basin areas are effectively sealed so that any runoff
from surrounding areas will accumulate on the surface. Overflow will occur
as soon as the impoundment capacity of local micro-relief is exceeded.
Deep frost disappears more slowly in areas where surface icing has
33
-------
occurred. Disappearance of ice and frost is affected further by lingering
patches of snow. Disappearance of snow cover is influenced by patterns of
drifting (as determined by wind direction and local relief) and by the mean
angle of incidence of insolation (as determined by season and by direction
of slope).
On these watersheds, snow disappeared first on upper slopes, next in
lower basin areas, and last on the north-facing slopes. Surface ice and
deep frost disappeared in the same order.
During winter runoff periods, overflow from lower basin areas in both
watersheds was frequently impeded by drifted snow in front of the catchments,
and extensive ponding occurred during the day. Meandering flows and sedimen-
tation in these areas were influenced further by lingering patches of snow
and by random patterns of disappearance of surface ice. In watershed 06,
meandering was promoted also, as snow and ice disappeared, by the relative
resistance to cutting afforded by sandy deposits left from summer events.
An additional factor in sedimentation and leaching was the reestablishment
of internal drainage as deep frost disappeared, beginning around the peri-
phery and progressing in random fashion through areas subject to ponding.
By the time of the 04-18-75 event, sorted sedimentation patterns at
points of rill discharge were scattered randomly in segment 0605 over an area
approaching the extent and outline of the Traverse fine sandy loam. Visible
deposits in the east and west thirds of this area were thin and not extensive.
These peripheral deposits mingled frequently with similarly thin (2-3 mm)
and non-extensive light-colored sediments originating in down-row runoff from
segments 0604 and 0606.
The widely meandering rills cut during winter events were not deep.
Deep cutting did occur during the event of 04-18-75. Major cuts (20 to 30
cm) occurred along the central draw, transecting sedimentation patterns laid
down earlier. Nevertheless, considerable meandering occurred in areas of
2-3% slope. Rills formed in these level areas were of moderate depth (10-15
cm), and extensive sedimentation fans were formed. In several places, light-
colored sand deposits, up to 5 cm deep, extended several meters east or west
from points of rill discharge and across several rows of corn stubble.
After plowing and planting on May 16 and 17, 1975, new patterns of cut-
ting and sedimentation were initiated quickly by frequent moderate to heavy
rains beginning 05-21-75. Meandering along the central draw increased as
the corn crop developed, particularly as brace roots were extended to ob-
struct cross-row flows, beginning early in July. Some moderately deep rills
(10-15 cm) were cut during early events, mainly in segment 0602 (Fig. 8).
Cutting became shallower as meandering increased. Sorted surface sediments
in 0605 were generally thin, but by mid August their random distribution was
as extensive as at the end of the previous winter.
As in the case of the 04-18-75 event, the very heavy rain of 08-21-75
produced deep rills and gullies which transected earlier sedimentation
patterns. However, in the presence of a fully developed corn crop, meander-
ing in areas of 2-3% slope was more extensive than in April, and a larger
34
-------
proportion of the area was affected both by deep sedimentation in rills and
gullies and by lateral surface deposits.
In contrast to 06, areas in watershed 07 where runoff down the central
draw can spread laterally are limited to a rather narrow band of 3-4% slopes
below the 890 ft contour (Fig. 2). The opportunity for ponding in row
middles reaches its widest extent between the 888 and 885 ft contours. This
area includes the wide portion of segment 2 in Fig. 6 and the south one-third
of segment 1. Ponding below the 885 ft contour was variable because of the
tendency, even during events of only moderate magnitude, for runoff flows
to converge into one or two deeper rills which would drain the area quickly
to the catchment.
Because of the limited impoundment capacity in watershed 07, rills cut
in areas corresponding to segment 0705 in Fig. 8 were deeper than in 0605,
meandering and lateral sedimentation were less extensive, surface deposits
were thinner, and the heavier light-colored sediments were carried further
down the drainageway.
During summer 1973 and again during summer 1974, cutting and filling,
together with sorted lateral sedimentation in row middles, was observed all
the way to the catchment. During both winter runoff periods (1973-74 and
1974-75), sedimentation in the area below about the 884 ft contour (Fig. 2)
was promoted by ponding due to drifted snow in front of the catchment.
During summer 1975, a number of events, beginning early in the season,
were of sufficient magnitude that deep rills were cut which drained the areas
below the 885 ft contour quickly before much ponding or lateral sedimentation
could occur. Over a succession of events, rills would fill and new ones form,
but meandering was narrowly restricted. Some lateral sedimentation did occur
during lesser runoff events.
During the very heavy rains of 04-18 and 08-21-75, a central gully was
scoured through the plow layer, beginning at about the 883 ft contour. A
substratum of glacial outwash cobbles was exposed over areas up to a meter
wide, and washing of the plow layer extended over a wider area.
Soybeans lodged extensively in central areas of segments 0702 and 0705
where deep cutting occurred on 08-21-75. The fallen vegetation served to
slow runoff flows during later events and promote sedimentation. This is
reflected in our runoff volumes and sediment yields for the event of the
following day.
Some sedimentation in rills and gullies undoubtedly occurred during the
event of 08-22-75. However, major cuts were still open at the time of our
final mass balance sampling on 08-27-75. By harvest in October, most deep
rills in segment 0702 and the upper half of 0705 had filled. Due to inter-
ception of sediments at higher elevations, not much sedimentation had occur-
red in the central gully below the 883 ft contour.
Another feature of difference between the two watersheds is that E-W
slopes in segment 0706 and parts of 0704 (Fig. 8) were steeper than in the
35
-------
corresponding segments of the other watershed. Sediments in down-row runoff
from these areas were carried further into the central draw and contributed
to visible surface deposits and to filling of cross-row rills in 0705. Be-
cause the slopes were short, the surface deposits were thin, however.
It is difficult to anticipate how the unique differences in patterns
of erosion and sedimentation on the two watersheds during each runoff period
will affect our soil core data. However, expected differences do appear in
the runoff data.
Because of the very much larger area where ponding and sedimentation
could occur in the central basin of 06, runoff which could be measured at
the weir occurred less frequently, sediment yields were generally lower, and
total sediment losses during major events were less than on 07. The larger
sediment losses from 07 included a larger proportion of less adsorptive sand,
and this is reflected in lower concentrations of paraquat in the sediment
phas e.
A further comment should be made regarding effects of freezing and thaw-
ing on patterns of sediment pick-up and resedimentation. Our observations
in winter 1973-74 were rather superficial but consistent with more detailed
observations in winter 1974-75.
Depth of freezing was related to slope and snow cover. Depending on
wind direction, snow would drift on slopes facing NE or NW, leaving only 5
to 10 cm trapped by stubble on upper slopes and variable depths in central
basins. During freezing cycles, frost would penetrate quickly and to great
depths (45 cm) if the soil were bare. Under snow cover, the rate of frost
penetration would be related inversely to the depth of snow. Once frozen,
however, the soil would not begin to thaw until snow cover became thin and
granular so that the sun's rays could penetrate. At that point, the surface
centimeter or two would thaw quickly in bright sun even when air temperatures
were at or slightly below freezing.
Soil thawing under departing snow cover is saturated with water that
cannot be removed by percolation into frozen soil underneath. Soil materials
are, therefore, readily picked up by moving water if snow melt is rapid and,
in particular, if snow melt is accompanied by even a light rain. During
periods of thawing in winter months, the soil usually freezes again at night.
Alternate freezing and thawing serves to keep soil materials on the surface
loose and readily suspended in moving water.
Because of the normally thinner snow cover on upper slopes and their
more direct exposure to a southerly sun, the upper slopes experienced fre-
quent cycles of freezing and thawing and were frequently bare of snow at the
time of winter rains.
Very little movement of sediments was observed on these upper slopes
resulting from snow melt alone. However, as snow cover disappeared from
areas on intermediate and lower slopes surrounding the central basins, con-
siderable pick-up and redeposition was observed, without rain, due to water
flows originating in snowmelt and seepage from higher elevations. Patterns
36
-------
of redeposition were influenced markedly by lingering patches of snow and/or
ice.
In the central basins, lingering patches of snow increased the meander-
ing of runoff flows. Thus, areas affected by alternate cutting and filling
and lateral sedimentation increased markedly. For this reason, sediment
yields in our data for winter events involving mainly snowmelt are low and
do not reflect the extensive pick-up and redeposition of sediments that oc-
curred within each watershed. On the other hand, winter events involving
both snowmelt and rain usually produced sediment yields substantially higher
than did rains of similar intensity or duration on unfrozen soil at other
seasons of the year.
SOIL CORE DATA, MAY 1973 TO SEPTEMBER 1974
In tables 3 to 12, available computer summaries have been condensed to
focus on changes in recovery of pesticides in surface soil layers (0 to 7-1/2
cm) and at greater depths (7-1/2 to 30 cm).
Sources of variation
Recoveries from spiked samples were reproducible (±5 to 12%) and ranged
from 82 to 89% for atrazine and diphenamid and from 89 to 104% for paraquat
and trifluralin (Table 1).
Recovery on filter paper monitoring discs in the field was more variable
and appeared to have been influenced by slope, direction of movement of the
sprayer, and by wind direction at the time of application. Variation among
10disc means for upslope and downslope traverses and traverses on the level
was ±18 - 19% of overall recovery for atrazine, ±22 - 35% for diphenamid,
and ±8 - 19% for paraquat. Overall disc recoveries in the last three appli-
cations ranged from 54 to 112% of calculated tank delivery for atrazine, 67
to 79% for diphenamid and 71 to 141% for paraquat.
A disturbing feature is the very low recovery of all chemicals immedi-
ately after the first application in 1973 (Tables 3, 5, 7, 8, 9, 10). These
low initial recoveries may reflect sampling difficulties. The weather was
windy, the soil very dry and loose, and we were still experimenting with
sampling devices.
After the initial sampling, recoveries of paraquat and trifluralin in-
creased and the sums for both watersheds remained rather stable for 2 or 3
samplings (Tables 3 to 8). Total recoveries of diphenamid did not change
materially over the first 3 or 4 samplings (Tables 9 and 10). Averaging the
sums for 06-18 and 07-04-73, recoveries expressed as percent of applied (for
06 and 07 respectively) were: paraquat (73 and 95%), trifluralin (97 and
66%), and diphenamid (31 and 28%). These recoveries may be compared with
mean recoveries from spiked samples: paraquat (90± 5%), trifluralin (97+
12%), diphenamid (86± 10%).
Soils are loose and rough at the surface after tillage operations, more
conducive to infiltration than runoff. Hardly any net runoff occurred at the
37
-------
TABLE 3. WATERSHED 06 PARAQUAT* MAY 1973 TO MAY 1974
Date
05-24-73
06-09-73
06-18-73
07-04-73
08-11-73
09-30-73
11-05-73
03-05-74
05-02-74
Inputs
Depth
cm
0-7%
0-7%
0-7%
0-7%
0-7%
7%-30
0-7%
7%-30
0-7%
7%-30
0-7%
0-7%
7%-30
06-09-73*
11-05-73*
TOTAL
Grams paraquat remaining
Seg 1
0
25
43
36
53
99
48
32
55
39
110
93
2
64
64
128
Seg 2
0
81
287
213
127
147
148
25
354
196
293
555
134
317
317
634
Seg 3
0
82
186
203
174
238
109
244
396
189
188
514
158
277
277
553
Seg 4
0
140
137
216
133
121
84
293
230
161
82
429
119
245
245
491
Sums
Increment 0- 30
0
328
652
668
486
605
389
594
1036
586
673
1591
412
0
1091
983
1621
2003
898
898
1796
* Applied to deliver 1.12 kg/ha (tank delivery was not determined).
38
-------
WAI iltlHEL Cif-
SUMMER 1974
Date
(05-02-74
Depth
cm
segments)
0-7 1/2
7 i,2-30
05-13, 14-74: Plowed
05-22-74
05-30-74
07-03-74
08-05-74
08-14-74
09-03-75
Inputs
0-7 1/2
7 1/2-30
0-7 1/2
7 1/2-30
0-7 J./2
7 1/2-30
0-7 1/2
0-7 1/2
7 1/2-30
0-7 1/2
7 1/2-30
05-22-74*
To date
Segments
Seg
(25
134
771
179
423
23i
806
438
248
607
267
391
179
532
1 Seg
(3)
(514)
(158)
cm)
122
322
124
391
144
339
118
130
509
211
521
110
327
Grains p,
araquat. remaining
sampled beginning 05-22-74 Suma
2 Seg 3
(4)
r»?9i
(119)
124
679
155
174
\95
J'»5
260
193
531
148
965
147
438
Seg
156
771
175
,)26
2j8
546
312
312
759
296
574
206
613
4 Se« '
d>
(93)
(2)
140
453
iC2
170
153
46'J
i89
204
691
239
837
137
408
"" Seg 6 Incre-
ment
(2)
(5i5) (159 i)
(134) (4'2)
83 759
329 3325
17'; 912
130 i8!6
131 ll',2
156 7,861
U* 1482
267 1354
512 3609
126 1288
532 3820
128
381
C-3u
(2J03)
4084
2728
3973
4963
5108
907
2702
* Based on tank delivery of 907 g ( - 1.13 kg/ha).
Recovered on filter paper discs: 979 g ( -1.22 kg/ha).
39
-------
TABLE 5. WATERSHED 07 PARAQUAT* MAY 1973 TO MAY 1974
Date
05-24-73
06-09-73
06-18-73
07-04-73
08-11-73
09-30-73
11-05-73
03-05-74
05-02-74
Inputs
Depth
cm
0-7%
0-7%
0-7%
0-7%
0-7%
7%-30
0-7%
7%-30
0-7%
7%-30
0-7%
0-7%
7%-30
06-09-73*
11-05-73*
TOTAL
Grams paraquat remaining
Seg 1
0
48
75
95
44
0
58
10
133
66
144
183
47
82
82
164
Seg 2
0
37
210
219
74
17
86
107
155
106
299
244
of
164
164
327
Seg 3
0
59
180
74
89
34
38
252
105
110
275
187
84
122
122
244
Seg 4
0
97
135
175
157
152
229
231
221
173
-4
431
293
245
245
491
Sums
Increment 0-30
0
240
600
563
364
204
410
560
614
454
717
1046
423
0
568
1010
1068
1469
613
613
1225
* Applied to deliver 1.12 kg/ha (tank delivery was not determined).
t Minimum quantifiable was 50 ppb, equivalent to 12 g for 7%-30 cm in
segment 1, and 24 g for segment 2.
+ Not sampled.
40
-------
TABLE 6. WATERSHED 07 PARAQUAT* SUMMER 1974
Grams paraquat
Date
Depth
cm
remaining
Segments sampled beginning 05-22-74
Seg
(05-02-74 segments)
7
0-7 1/2
1/2-30
05-13,14-74 Plowed
05-22-74
7
05-30-74
7
07-03-74
7
08-05-74
08-14-74
7
09-03-74
7
Inputs
0-7 1/2
1/2-30
0-7 1/2
1/2-30
0-7 1/2
1/2-30
0-7 1/2
0-7 1/2
1/2-30
0-7 1/2
1/2-30
05-22-74*
To date
(25 cm)
109
698
149
291
192
296
221
298
425
207
553
176
554
1 Seg
(3)
(184)
(47)
53
268
64
143
100
204
93
116
475
127
251
71
222
2 Seg 3
(4)
(244)
(0)
77
281
103
205
125
298
174
153
295
150
363
110
347
Seg
68
366
134
168
176
564
197
276
777
163
601
120
378
4 Seg
(1)
(187)
(84)
61
164
66
199
104
302
97
103
321
153
314
71
222
5 Seg 6
(2)
(431)
(293)
101
261
114
339
156
371
200
231
490
142
556
130
409
Sums
Incre-
ment
(1046)
C423)
469
2038
628
1345
852
2035
982
1176
2783
943
2638
0-3C
(1469)
2506
1973
28R7
3959
3581
678
2132
* Based on tank delivery of 678 g ( - 1.24 kg/ha).
Recovered on filter paper discs: 481 g ( - 0.88 kg/ha).
41
-------
TABLE 7. WATERSHED 06 TRIFLURALIN* MAY 1973 TO MAY 1974
Date Depth
cm
05-24-73 0-7%
06-08-73 0-7%
06-09-73 0-7%
06-18-73 0-7%
07-04-73 0-7%
08-11-73 0-7%
7%-30
09-30-73 0-7%
7%-30
j>05"//4 Q-7'f
05-0/; -//i 0-7%
7%-30
Input 06-08-73*
Grams
Seg 1
0
8
83
42
93
23
7
21
2
36
10
2
64
of Trifluralin remaining
Seg 2
0
34
179
265
424
133
22
55
9
51
61
15
317
Seg 3
0
71
501
195
300
181
29
17
8
27
32
29
277
Seg 4
0
30
355
242
178
95
7
46
7
21
56
11
245
Sums
Increment 0-30
0
143
1118
744
995
433
65
140
27
135
159
57
0
498
167
216
898
Applied to deliver 1.12 kg/ha (Tank delivery was not determined).
42
-------
TABLE 8. WATERSHED 07 TRIFLURALIN* MAY 1973 TO MAY 1974
Date
05-24-73
06-08-73
06-09-73
06-18-73
07-04-73
08-11-73
09-30-73
03-05-74
05-02-74
Input
Depth
cm
0-7%
0-7%
0-7*2
0-7%
0-7%
0-7%
7%-30
0-7%
7%-30
0-7%
0-7%
7%-30
06-08-73*
Grams
Seg 1
0
18
31
49
51
30
2
18
2
1
5
2
82
of Trifluralin
Seg 2
0
17
133
104
151
42
11
22
5
17
16
2
164
Seg 3
0
25
69
107
73
121
4
22
4
16
31
3
122
remaining
Seg 4
0
29
105
109
160
97
7
41
14
17
54
10
245
Sums
Increment
0
88
338
369
434
289
24
103
25
51
107
17
0-30
cm
0
314
128
124
613
* Applied to deliver 1.12 kg/ha (tank delivery was not determined).
43
-------
catchments before 08-09-73. Nevertheless, during rains on 06-16 (2.49 cm)
and on 06-25 to 06-29 (3.05 cm), runoff occurred from upper slopes and mid-
row ponding occurred in central basin areas. Chemicals were undoubtedly dis-
placed downslope and into the profile. Some loss of less persistent chemi-
cals undoubtedly occurred by volatilization and degradation. These probable
redistributions and losses are masked in segments 3 and 4 by a sampling bias
which favored depositional intermediate slopes. Thus, the apparent constancy
in total recoveries between 06-18 and 07-04 or 08-11-73 is an integrated ex-
pression of changes in distribution and variations in sampling recovery with-
in segments and between segments in each watershed.
It is not possible to assess sampling recovery in soils of chemicals
from later applications in this study because of uncertainties regarding
background, especially after spring plowing. Sampling variation was certain-
ly as great as that indicated by the filter paper discs. Random interception
of chemicals from runoff on intermediate slopes and random patterns of cut-
ting and sorted sedimentation along the central drainageway in each watershed
would have contributed to sampling variation.
In spite of the indicated great variability in application and recovery,
patterns of change over time appear which can be interpreted in terms of
movement down slope or within the profile. Mass balance changes for indivi-
dual chemicals are consistent with their known properties, and unique effects
of topography and of soil and weather conditions are expressed.
Evidence for movement down slope and in the profile
Our sampling segments were laid out initially to differentiate between
areas subject mainly to sheet erosion down the row and areas where cross-
row rilling or visible sedimentation occurred. However, very fine sediments
and dissolved chemicals can be intercepted, without visible evidence, by in-
filtration at any point down slope from areas of pickup. We had not antici-
pated the extent to which this could happen on the permeable sandy soils in
these watersheds. Our composite samples in steeply sloping segments were
based on 10 to 15 random cores but were, in effect, weighted in favor of de-
positional intermediate slopes.
Because of this sampling bias, increasing recovery in segments 3 and
4 of Fig. 6 and in segments 01, 03, 04 and 06 in Fig. 8 is evidence for down-
slope movement within these segments.
Ac upper elevations in the central drainageway, increasing recoveries
will reflect interception from down-row runoff originating on slopes to the
east and west. These areas would have been included in segments 2, 3 and
4 of Fig. 6 and in segment 02 of Fig. 8. Decreasing recoveries in these
areas will reflect pickup in cross-row runoff, with displacement further
downslope into segments 2 and 1 of Fig. 6 and into segment 05 of Fig. 8.
On areas of 4% slope or less at lower elevations along the central draw,
increasing recoveries will reflect interception from both down-row and cross-
row runoff. Major deposition would be expected in areas where mid-row
ponding occurs frequently. Areas available for ponding are very different
44
-------
in the two watersheds. Large differences can be expected between watersheds
and between runoff events in interception and retention of chemicals in these
areas. They are included in segment 1 and portions of segment 2 in Fig. 6
and in segment 05 of Fig. 8.
In all segments, losses by volatilization or degradation and by movement
out of the sampling zone into the subsoil will contribute to decreasing re-
covery. Downward movement through the soil is more readily apparent in the
original data for seven sampling depths, but it is reflected also in the data
reported here for two depth categories (0-7-1/2 and 7-1/2 to 30 cm).
Paraquat in soils
It must be recognized that movement and retention in soils will reflect
differences between chemicals in their partitioning between the solution
phase and solid surfaces. A little known factor is the extent to which ful-
vic acid-type products of humification may influence the partitioning of even
a strongly adsorbed chemical such as paraquat.
Paraquat was never detected in the water phase of runoff. However,
concentrations less than 100 ppb in water were not detected routinely.
Thus, it cannot be assumed that paraquat moves only in association with
particulate matter. Its retention and residual distribution in soils can,
nonetheless, be interpreted in terms of interaction with adsorptive surfaces.
Our data for paraquat through summer 1974 are summarized in Tables 3
to 6. There is great variation from date to date for depths, segments and
watershed sums. Included in this variation are variations associated with
application, sampling and analysis, as well as random variation due to random
patterns of pickup and interception which were not represented proportionate-
ly in 10 to 15 cores per segment.
Total recoveries on each watershed increased more rapidly than inputs.
In the 09-03-74 sampling (Tables 4 and 6), recovery to 30 cm exceeded the
total for three applications by 90% in watershed 06 and by 70% in 07.
Greater total recovery in 06 is related to the longer east-west slopes
and the greater opportunity for interception of sediments and chemicals from
down-row runoff. This relation to length of slope can be seen also within
watersheds. Thus, during summer 1973, paraquat accumulated to higher levels
in segment 3 of watershed 06 than in segment 4 (Table 3 and Fig. 6). In 07,
the largest accumulation occurred in segment 4 (Table 5).
In the early growing season, interception on slopes is favored by the
loose, uneven surface left by tillage operations. During winter and early
spring, visible sedimentation is associated with lingering patches of snow
or ice. On these watersheds, the last snow to leave was always on intermedi-
ate slopes where drifting occurred frequently and the insolation angle was
low. Patterns of drifting varied with wind direction and were further in-
fluenced by a woodlot 50 meters directly south of the site and extending
about 100 meters to the west of 07.
45
-------
Marked accumulations of paraquat in surface layers (0-7-1/2 cm) of seg-
ments 2, 3 and 4 in the 05-02-74 sampling in watershed 06 (Table 3) were, as-
seeiated with a sequence of light snows (2 to "5 cm) in March. Between snows,
the snow on upper and lower slopes would melt, but a rather continuous snow
>r ice cover was maintained on intermediate slopes protected from the direct
.ays of the afternoon sun. Sediments were readily picked up by snowmelt and
light rains on upper slopes because the surface soil remained saturated ever
frost. Frost was still present at 7 to 10 cm on 03-30-74. Visible sedimen-
tation was observed where lingering snow and ice served to divert or slow
down runoff part way down the slope.
In watershed 07, the greatest accumulation of paraquat on 05-02-74 was
in surface layers of segment 4 where intermediate slopes were most protected
from the direct rays of the sun during the warmer parts of the day. In seg-
ments 2 and 3, marked surface accumulations had occurred earlier in the win-
ter, as indicated by our data for 03-05-74 (Table 5). By contrast, in water--
shed 06 (Table 3), paraquat in surface layers on 03-05-74 was sharply reduced
from 11-05-73 in sloping segments and there was some evidence for net move-
ment into segment 1. Such divergent wintertime sequences in the two water-
sheds reflect differences in patterns of drifting and disappearance of srow
and frost as influenced by wind direction and mean Incidence of insolation
on major slopes.
In the absence of frost, pa"?r;ij«^t -nov_d quickly into the profile. Sii'«-
stantial movement to depths grear.ti than 7--1/2 cm had occurred already by
the time of our first full depth enroling 'in 08-11-73 (Tables 3 and 5). By
09-30-73, 60% of the total in both watersheds was in the 75% of sampled soil
volume below 7-1/2 cm.
Due to the fall application on 11-05-73 and surface accumulation or fro-
zen soil in depositional areas curing the winter, the proportion below 7-1/2
cm was reduced to 20 or 30% in thr- 05-02-74 sampling. The surface accumula-
tions were inverted by plowing so that, immediately after the spring appli-
cation on 05-22-74, 80% of the total in both watersheds was found in the
lower 75% of the sampling zone (Tables 4 and 6).
The first rains after the spring application occurred on 05 -28-74 arid
05-29-74 and totaled 1.96 cm. The soil was still moist from 1.17 era of rain
on 05-17 but loose and rough on the surface from discing and tillivator op-
erations on 05-20. Most of the rain infiltrated where it fell, except for
a brief period on 05-29 when ponding occurred in the 05 segments and a slight
amount of runoff (140 1) was measured at the flume of watershed 07.
Our data for 05-30-74 suggest that 35 to 45% of the paraquat found below
7-1/2 cm on 05-22 may have been displaced by percolating water to depths be-
low 30 cm. The greatest apparent losses into the subsoil occurred in seg-
ments with severely eroded slopes: 1, 3 and 4 in 06 (Table 4) and segment
1 in 07 (Table 06).
The indicated loss from deeper layers in segment 5 of watershed 06 was
also great (60%). By contrast, there was no evidence for loss to the sub-
soil in segment 5 of watershed 07. This difference relates to the very much
46
-------
larger areas in 06 where mid-row ponding can occur. Ponded water infiltrates
into these sandy soils very quickly (unless frost is present). The soil in
areas where ponding occurs is subject to more intense leaching of water-sol-
uble pesticides than are soils on slopes where surface drainage is rapid.
After 05-22-74, watershed sums increased in each successive sampling
through 08-14-74. The proportion found in the 75% of soil volume below 7-1/2
cm remained rather constant at 70 to 75% of the total for each watershed.
Within individual segments, however, there were sequences of increasing and
decreasing recovery in depth increments below 7-1/2 cm. These sequences oc-
curred at different times in different segments but do not appear to be due
simply to random variation. It appears that, within a given segment, there
were periods of net movement into the plow layer from the surface and periods
of net movement out of the lower plow layer into the subsoil.
The rapid movement of paraquat into the profile was not expected.
Sifting of fine materials in the dry condition and infiltration and percola-
tion of dissolved species and suspended matter would have been facilitated
during early portions of each growing season by the loose condition of the
plow layer, notably on eroded slopes. Cutting and filling in central areas
would have contributed to in-depth distribution also. However, the principal
factors affecting movement into and through the soil would appear to be topo-
graphical features, variations in soil structure, and the presence or absence
of frost or of a vigorously growing crop, which determine net infiltration
and deep percolation of water.
The evidence that extensive movement occurred in association with pon-
ding through the essentially structureless fine sands of segment 5 in water-
shed 06 suggests that paraquat may have moved to a large extent in solution,
perhaps in association with solubilizing organics in the fulvic acid fraction
of soil organic matter. Paraquat adsorbed on suspended clay or silt frac
tions would move readily into the granular plow layer on eroded slopes but
would tend to be intercepted at the plow sole, which is included in our deep-
est sampling increment (22-1/2 to 30 cm). Our data provide no clear evidence
of progressive build-up at this depth.
Due to our sampling bias which favored depositional slopes, increasing
recoveries of paraquat in our steeply sloping segments actually reflect down-
slope displacement. With this in mind, the data in Tables 4 and 6 present
a picture of progressive downslope displacement of paraquat in runoff, with
interception and infiltration into the soil on intermediate slopes in seg
ments 1, 2, 3, 4 and 6 and in areas of ponding in segment 5. Differences
between the two watersheds reflect mainly differences in length of E-W slope
and the proportion of gentle slopes where runoff flows tend to slow down or
where mid-row ponding can occur.
It is apparent that the sandy soils on these watesheds have a limited
capacity for retaining paraquat. Although our total recoveries through
09-03-74 still exceeded inputs, it appears that substantial movement into
the subsoil had occurred. Losses into the subsoil undoubtedly exceed, by
a large factor, losses in runoff at the catchments. The concentration of
dissolved and suspended paraquat in percolating water is probably not greatly
47
-------
different from that in normal runoff. During the more intense runoff events,
the heavier sediments picked up will tend to be less adsorptive and contri-
bute proportionately less to runoff losses of paraquat.
By the end of summer 1974, both watersheds appeared to be approaching
a steady state where, with twice-a-year inputs of paraquat, the total re-
tained in the plow layer over each watershed will reflect mainly seasonal
or annual differences in rainfall and net percolation. Within individual
segments, large fluctuations in total recovery and depth distribution will
reflect effects of topographical features, frost, tillage operations and crop
on interception, infiltration and deep percolation of precipitation and run-
off. Variation associated with these factors appears to be well in excess
of random variation, although anomalous data points do appear and will need
to be given special statistical treatment.
Trifluralin in soils
Because of the volatility of trifluralin, the first mass balance sam-
pling was taken on 06-08-73, immediately after the trifluralin was applied
and worked in to 3 inches with a tillivator (Tables 7 and 8). These samples
were taken as single depth increments to 7-1/2 cm. On the following day,
after application of paraquat and diphenamid, soils were sampled again in
four increments to the same depth.
Trifluralin was encountered at all four depths on 06-09. The sum for
the four increments was 3 to 4 fold greater than in the single increment
taken the previous day. The lower recoveries in the single in-rement of 06-
08 may reflect losses of very dry fine materials during sample manipulation,
since a strong wind came up shortly after application. It was windy also
on 0609, increasing the likelihood for contamination as deeper increments
were exposed, in sampling, to fine materials blowing across the soil surface.
The largest recoveries of trifluralin at lower depths on 06-09-73
were in segment 2 of both watersheds and in eroded segment 3 of 06 and 4 of
07. In the same samples, varying proportions of paraquat and diphenamid were
also encountered at depths below the surface centimeter. Since these two
chemicals were not worked in, recovery at deeper levels must be attributed
to contamination. There was a tendency to deeper recovery of all three chem-
icals in eroded segments where the plow layer is coarsely granular due to
incorporation of subsoil materials. As samples were being taken, fine, dry
materials sifted readily downward into this loose, open structure but were
intercepted whenever moist soil was encountered. In all segments, variation
in recovery with depth in this sampling likely reflects variation in rate
of drying and distribution of moisture in the freshly tilled upper plow layer.
Due, at least in part, to its high volatility, trifluralin is the least
persistent in soils of the chemicals in this study. Kearney et al. (1969)
give 6 months for 75 to 100% disappearance. A discussion of factors affect-
ing trifluralin volatization is given by Harper et al. (1976).
There is no evidence of net loss from either watershed through 07-04-73
(Tables 7 and 8). Losses due to volatilization and degradation undoubtedly
48
-------
occurred but were masked by downslope movement and our sampling bias in
steeply sloping segments. Net accumulation had occurred by 07-04 in segment
2 of watershed 06 and segments 1 and 4 of 07.
Total recoveries declined sharply during July through September and then
remained rather constant through the winter (no fall application was made).
As in the case of paraquat, deposition associated with snow and ice on inter-
mediate slopes served to maintain or increase recoveries in segments 3 and
4 of both watersheds and, to a lesser extent, in segment 2. Net losses oc-
curred between 09-30-73 and 05-02-74 in the segments nearest the catchments
(segment 1, Fig. 6).
The original data show major losses over the year from surface incre-
ments (0-1, 1 - 2-1/2, 2-1/2 - 5 cm). There is some evidence for progres-
sive downward displacement of residual trifluralin at lower depths. Movement
within the plow layer after 06-09-73 was, however, very slow as compared with
paraquat. Net movement into the subsoil would appear to have been negligible.
This relative immobility of residual trifluralin is consistent with its very
low solubility (<1 ppm at 27° C).
Diphenamid in soils
The solubility of diphenamid (260 ppm at 26° C) (Herbicide Handbook,
1974) is much less than for paraquat but substantially greater than for tri-
fluralin. It is essentially non-volatile and, perhaps for this reason, it
is considered to be somewhat more persistent in soils than trifluralin (8
months vs 6 months for 75 to 100% disappearance).
Only about 1/3 of the first application of diphenamid was recovered in
the 06-09-73 sampling (Tables 9 and 10). In the original data, most of the
recovery was in the upper three increments (0 - 1, 1 - 2-1/2, 2-1/2 - 5 cm).
As in the case of paraquat and trifluralin, a large proportion was recovered
below the surface centimeter or two in eroded segments 3 of 06 and 4 of 07.
After 06-09-73, total recoveries remained rather constant through 08-
11 on 06 and through 07-04 on 07. Increasing recoveries in segments 1, 2
and 3 of 06 through 08-11 are evidence for extensive down-slope displacement.
The evidence for watershed 07 is less clear, but deposition on intermediate
slopes would have served to counter degradation losses and maintain recover-
ies in segments 2, 3 and 4.
Diphenamid moved into the plow layer more rapidly than trifluralin.
By 09-30-73, there was evidence of movement to depths below 7-1/2 cm in seg-
ment 3 of 06 (Table 9). During the winter, recoveries below 7-1/2 cm in-
creased markedly in all segments of 06. The original data indicate some di-
phenamid may have been lost from the plow layer into the subsoil.
The indicated downward movement of diphenamid during the winter con-
trasts with the behavior of paraquat, which appeared to have moved very
little within the plow layer, while marked surface deposition occurred in
association with lingering snow and ice on intermediate slopes (cf. Table 3).
Diphenamid is much less strongly adsorbed than paraquat and could have moved
49
-------
TABLE 9. WATERSHED 06 DIPHENAMID* MAY 1973 to MAY 1974
Date
05-24-73
06-09-73
06-18-73
07-04-73
08-11-73
09-30-73
11-05-73
03-05-74
05-02-74
Inputs
Depth
cm
0-7%
0-7%
0-7%
0-7%
0-7%
7%-30
0-7%
7%-30
0-7%
7%-30
0-7%
0-7%
7%-30
06-09-73*
11-05-73*
TOTAL
Grams
Seg 1
0
95
46
47
92
2
11
2
38
2
24
11
35
192
192
383
of Diphenamid remaining
Seg 2
0
270
182
316
351
9
31
9
211
9
58
38
173
951
951
1902
Seg 3
0
475
190
411
326
8
122
165
66
8
95
25
125
830
830
1660
Seg 4
0
73
290
164
141
7
77
7
175
7
95
11
74
736
736
1472
Sums
Increment
0
913
707
938
911
27
240
184
490
27
272
85
406
0-30
938
424
517
491
2695
2695
5389
* Applied to deliver 3.36 kg/ha (tank delivery was not determined).
50
-------
TABLE 10. WATERSHED 07 DIPEHNAMID* MAY 1973 TO MAY 1974
Grams of
Date
05-24-73
06-09-73
06-18-73
07-04-73
08-11-73
09-30-73
11-05-73
03-05-74
05-02-74
Inputs
Depth
cm
0-7 1/2
0-7 1/2
0-7 1/2
0-7 1/2
0-7 1/2
7 1/2-30
0-7 1/2
7 1/2-30
0-7 1/2
7 1/2-30
0-7 1/2
0-7 1/2
7 1/2-30
06-09-73*
11-05-73*
Total
Seg 1
0
36
63
27
27
2
15
2
188
2
9
.0
.0
.8
.6
.1
.4
.8
.4
.3
.5
.3
9.8
10.6
245
245
491
Seg
0.
183.
125.
154.
105.
4.
41.
4.
516.
4.
18.
16.
38.
491
491
981
Diphenamid remaining
2
0
6
3
3
0
8
8
8
3
8
7
5
7
Seg 3
0.0
62.2
175.8
84.4
28.3
8.3
15.0
3.6
352.4
39.0
11.2
5.1
30.6
366
366
732
Seg
4
£
iun
IS
Increment 0-30
0.0
164.
222.
178.
108.
16.
76.
7.
543.
7.
104.
7.
22.
736
736
1472
0
,2
0
0
7
2
3
1
2
6
9
6
0.
445.
587.
444.
268.
32.
148.
18.
1600.
53.
143.
39.
102.
0
8
2
3
4
2
8
1
1
5
9
3
5
300.6
166.9
1653.6
141.8
1838
1838
3676
* Applied to deliver 3.36 kg/ha (tank delivery was not determined).
51
-------
rapidly with percolating water during winter thaws, when the frost layer re-
cedes or disappears altogether, as snow cover is removed exposing bare soil
to the direct rays of the sun.
During this first year, movement of residual diphenamid within the plow
layer was less in watershed 07 than in 06. The evidence in the original data
and in Table 10 indicates that movement into the subsoil was also less in
07.
After the 11-05-73 application, diphenamid inputs were discontinued
on watershed 06, and no analyses were made after 05-02-74. Data following
the 05-22-74 application on 07 are presented in Table 11.
Diphenamid disappeared much more rapidly in 1974 than the previous sum-
mer. It is likely that degradation losses were greater due to adaptations
in soil microbial populations. However, leaching and runoff were also great-
er. The original data indicate that substantial quantities of diphenamid
moved through the plow layer into the subsoil. As in the case of paraquat,
major movement out of the granular plow layer in segment 1 occurred between
05-22 and 05-30 (cf. Tables 6 and 11).
In other segments, diphenamid residues were retained within the
sampling zone for longer periods than in segment 1 (Table 11). Longer re-
tention in sloping segments 3, 4 and 6 reflects differences in infiltration
capability during early rains and differences in grade and length of E-W
slopes, allowing for interception and infiltration of runoff on intermediate
slopes over a succession of events (Figs. 2 and 8). Repeated inputs from
runoff from higher slopes served to maintain detectable quantities of diphen-
amid within the sampling zone in segments 4 and 5 through the final sampling
on 09-03.
Atrazine in soils
Atrazine is somewhat less soluble than diphenamid (30 to 70 ppm vs
260 ppm) (Herbicide Handbook, 1974). It is generally considered to be some-
what more persistent in soils also (10 months vs 8 months for 75 to 100%
disappearance).
Ten annual applications had been made for corn on these watersheds be-
fore this study was initiated in spring 1973. Carryover from these earlier
applications would explain the presence of atrazine in all depth increments
to 30 cm immediately after the 05-22-74 application on watershed 06 (Table
12). Sixty percent of the total was found in the 0 - 1 cm increment, and
this represents 64% recovery of the calculated tank delivery. The soils were
moist to the surface at the time of application. We did not have the sam-
pling difficulties encountered in very dry soils at the time of the initial
samplings in 1973. We are confident that very little contamination of in-
crements below the surface centimeter occurred.
It is unfortunate that analyses for background atrazine were not made
prior to the 05-22-74 application. It is difficult to reconcile the apparent
persistence of atrazine (from the last previous application in 1972) with
52
-------
TABLE 11. WATERSHED 07 DIPEHNAMID* SUMMER 1974
Grams diphenamid
Segments sampled beginning
Date
(05-02-74
Depth
cm
segments)
0-7 1/2
7 1/2-30
05-13,14-74 Plowed (25
05-22-74
05-30-74
07-03-74
08-05-74
08-14-74
09-03-74
Inputs
0-7 1/2
7 1/2-10
0-7 1/2
7 1/2-30
0-7 1/2
7 1/2-30
0-7 1/2
0-7 1/2
7 1/2-30
0-7 1/2
7 1/2-30
05-22-74*
to date
Seg
1 Seg
2 Seg
3 Seg
remaining
05-22-74
4 Seg
5 Seg
6 Si
ims
Incre- 0-30
cm)
392
463
370
178
32
52
9
of
18
0
0
530
1484
(3)
(5)
(31)
167
26
83
40
6
18
0
0
0
0
0
213
596
(4)
(8)
(23)
166
125
173
88
24
13
30
0
0
0
0
332
930
169
97
155
156
35
34
63
1
175
0
54
362
1014
(1)
(10)
(ID
66
12
60
10
18
8
26
14
92
11
8
213
596
(2)
(16)
(39)
213
94
210
75
42
86
33
34
129
0
0
392
1097
ment
(39)
(102)
1174
817
1052
547
157
211
160
49
414
11
62
(142)
1991
1599
368
463
73
2040
5716
* Based on tank delivery of 2040 g ( - 3.73 kg/ha).
Recovered on filter paper discs: 1619 g ( » 2,96 kg/ha).
t Minimum detectable (15 ppb) for 0-7>i / 7^-30 cm: Seg 1 (2 g / 7 g), Seg 2
(1 / 3), Seg 3 ( 1 / 4), Seg 4 (1/5), Seg 5 (1 / 3), Seg 6 (1 / 5).
53
-------
TABLE 12. WATERSHED 06 ATRAZINE* SUMMER 1974
Grams Atrazine
Date
05-22-74
05-30-74
07-03-74
08-05-74
08-14-74
09-03-74
Input
Depth
cm
0-7 1/2
7 1/2-30
0-7 1/2
7 1/2-30
0-7 1/2
7 1/2-30
0-7 1/2
0-7 1/2
7 1/2-30
0-7 1/2
7 1/2-30
05-22-74*
Seg 1
Seg 2
Seg 3
Seg
Segments sampled beginning
586
130
353
9
185
0*
106
20
0
64
7
714
399
114
324
3
57
4
69
24
0
52
24
438
370
232
270
11
76
8
69
47
7
39
0
588
634
302
433
16
140
0
226
70
0
61
0
823
remaining
4 Seg 5
05-22-74
506
134
256
12
84
0
97
50
0
58
71
547
Seg 6
235
125
262
12
75
4
84
81
0
48
0
511
Sums
Incre-
ment
2731
1037
1898
63
618
16
651
292
7
322
102
0-30
3768
1961
634
299
424
3625
* Based on tank delivery of 3625 g ( - 4.52 kg/ha).
Recovered on filter papers discs: 3160 g ( - 3.94 kg/ha).
t Minimum detectable (5 ppb) for 0-7*s / 7!j-30 cm:
Seg 1 (0.7 g / 2.6 g), Seg 2 (0.4 / 1.6), Seg 3 (0.5 / 2.1), Seg 4 (0.8 / 3.0),
Seg 5 (0.5 / 2.0), Seg 6 (0.5 / 1.9).
54
-------
its rapid disappearance from the lower plow layer between 05-22 and 07-03-74.
Nevertheless, the freshly tilled soil was favorable for infiltration of rain-
fall. Moreover, there was evidence that substantial quantities of paraquat
in both watersheds and of diphenamid in 07 were displaced into the subsoil
during this period also.
Carryover atrazine from earlier applications had virtually disappeared
from the lower plow layer by 07-03-74. Residual atrazine from the 05-22-
74 application moved progressively downward through the four increments above
71/2 cm and was found in the 71/2 - 15 cm increment of segments 1, 2 and
5 in the last sampling of 09-03-74. Direct comparison cannot be made between
atrazine and diphenamid since they were not applied together on the same
watershed. However, comparison of data in Tables 11 and 12 suggest that
atrazine was less mobile in soils than diphenamid, as might be expected from
its lower solubility. (The evidence on this point is more obvious in the
original data where movement through the 7 sampling levels can be followed).
As in the case of diphenamid, degradation losses served to obscure evi-
dence for movement downslope. Soil populations adapted to degrading atrazine
would have been present after ten annual applications. Interception and ac-
cumulation on intermediate slopes and in central basin areas failed to keep
pace with disappearance. Larger quantities of atrazine than of diphenamid
were retained in the 09-03 sampling, which is consistent with reported dif-
ferences in their persistence (cf. Tables 11 and 12).
Summation
This preliminary assessment of our soil core data through summer 1974
has revealed a number of relationships to methodology, chemical species,
topography, soil conditions and weather which should be helpful in interpret-
ing and modeling these and later data through summer 1975.
a) Random variation is high and derives from variable distri-
bution of chemicals during application, variability asso-
ciated with sampling and analysis, and random patterns of
pickup and interception from runoff.
b) Useful variation is present and can be interpreted in terms
of movement downslope and within the plow layer and in
terms of loss by degradation and/or volatilization or by
leaching out of the sampling zone into the subsoil.
c) Due to sampling bias, downslope movement within steeply
sloping segments is reflected by increasing recovery rather
than by decreasing mass balances as might be expected.
This relationship may change if the capacitances repre-
sented by depositional slopes and by retention in the
plow layer are satisfied or exceeded by heavier rainfall
in 1975.
d) Because of its resistance to degradation and its suscep-
tibility to adsorption, paraquat was the most useful tracer
55
-------
for following movement and retention in the landscape.
Losses by degradation or volatilization tended to mask the
evidence for movement of other herbicides in this study.
e) Mobility was related directly to solubility: paraquat >
diphenamid > atrazine > trifluralin. The great mobility
of paraquat was not expected and suggests interactions
with mineral and organic colloids and soluble fulvic acid
components which inhibit its biological activity without
immobilizing it.
f) Redistribution patterns reflected the effects of a number
of factors on interception, infiltration and deep percola-
tion of rainfall and runoff. East-west planter rows de-
termined the direction of runoff from upper slopes, pro-
moted cascading cross-row runoff in the central drainageway,
and provided micro-relief for impoundment of runoff (with
sorted sedimentation) at lower elevations. Granular struc-
ture of the plow layer on eroded slopes promoted infiltration
and deep percolation of rainfall and runoff. Loosened
structure and irregular surface of freshly tilled soil
promoted interception and infiltration during earlier
portions of each growing season. Length and pitch of
slope were important, as well as exposure of slope to
drifting snow and winter sun. The presence of a frost
layer precluded deep percolation, promoted pickup of
saturated soil materials on upper slopes and ponding at
lower elevations. Lingering snow and ice promoted inter-
ception on intermediate slopes.
g) Differences between the two watersheds in distribution
and retention of chemicals reflected mainly differences
in length of E-W slopes and the proportion of gentle
slopes where runoff flows tend to slow down or where
ponding can occur. Soil differences were minor except
in relation to topography and the degree of erosion and
incorporation of granular subsoil materials into the
plow layer.
h) Movement over the watersheds and through the soil was more
rapid in summer 1974 than in 1973, reflecting rains of
greater magnitude. In 1975, deeply penetrating rains and
rains of erosive intensity were more frequent than in 1974.
Greater downslope movement of chemicals may be expected
because visible sediments were transported further down
the drainageway in both watersheds. Corn and soybeans were
much more vigorous in 1975, and differences between water-
sheds may reflect differences in nature of the canopy and
its rate of development.
56
-------
Selected Runoff Data, 1975
Our objective in this section is to illustrate how weather sequences
which precede a runoff event influence the capacity of soils to accept rain-
fall or to intercept runoff which contains potential pollutants.
During the growing season, even light rains can serve to slake surface
aggregates and level off irregularities left by tillage, thereby reducing
infiltration capacity and facilitating runoff during later events. Penetrat
ing rains, by satisfying the waterholding capacity of the soil, will increase
the likelihood of runoff when it rains again. This will be true in particu-
lar during the period when surplus gravitational water is retained in the
soil. On the permeable soils in these watersheds, the difference between
runoff and no runoff during Spring, Summer, and Fall was frequently deter-
mined by rain which occurred 24 to 36 hours before the rain which actually
produced runoff. During the wintertime, runoff volumes and sediment loads
were influenced by weather sequences, extending over periods of days or weeks,
which determined the depth and distribution of snow, ice, and frost.
An effort was made during winter 1974-75 to record snow, ice, and frost
conditions in greater detail than during the previous winter. These obser-
vations and associated weather conditions are summarized for important se-
quences of events in Tables 13 to 16. Precipitation summaries for these
events and runoff data for watershed 06 are presented in Tables 19 to 22.
Two major historic rainfall events were experienced in 1975, one on bare
soil in April and one in August when vigorous, fully developed crop cover
was present. Significant soil conditions and weather sequences associated
with these events are summarized in Tables 17 and 18. Precipitation sum-
maries and runoff data for both watersheds are summarized in Tables 23 and
24.
Runoff and residue data in Tables 19 to 24 have been calculated by in-
terpolating linearly between points on the waterstage calibration curve.
Runoff volumes integrated by the computer will be more accurate, but our cal-
culations are adequate for our objective in this section.
Wintertime runoff sequences
Beginning late in NOvember 1974, the surface soil would freeze overnight
or for periods of days but would thaw during periods of rain or melting snow.
Thus, the soil remained permeable, and no runoff occurred until 01-08 to 01-
10-75 when a combination of snow melt and rain totalling 2.74 cm did produce
runoff at the catchments. On 01-10-75, there was no evidence of frost to 30
cm.
Immediately following these first wintertime events, there was a two-
week period of very cold weather (night temperatures ranging down to -20 C).
Snow cover was light (0 to 2-1/2 cm), and the soil froze deeply. Major run-
off events from late January through February occurred on frozen soil and
were due to rain or melting snow or a combination of snow melt plus rain.
During this period bare soil areas or areas under thin and aging (granular)
57
-------
TABLE 13. WEATHER AND WATERSHED CONDITIONS FOR THE PERIOD 03-12-75 TO
03-11-75
Daily runoff due to snow melt on soil initially frozen to
46 cm (18 in).
02-13-75 Five to ten centimeters snow on soil frozen from surface to 46 cm.
Air temperature rose from -4 C (25 F) overnight to 3 C (38 F)
at about 1600. Rained in a.m. (.18 cm) from 0530 to 0630, but
runoff due to snow melt did not start until 1300.
03-13-75 Air temperature rose from overnight low of -8 C (18 F) to maxi-
mum of 2 C (35 F) at about 1600. Runoff due to snow melt began
at 1000 continued to 1930. Runoff flowed under crusted snow and
ice, over frozen soil beginning to thaw (1.25 cm) by late p.m.
03-14-75 Air temperature rose from overnight low of -7 C (20 F) to
maximum of 0 C at about 1600. Runoff due to snow melt (bright
sun) began at 1225 and continued to 1810. Runoff flowed under
crusted snow and ice (2.5 cm) over frozen soil beginning to
thaw (1.25 cm) again in p.m.
03-15-75 Air temperature rose from overnight low of -8 C (18 F) to a
maximum of 6 C (43 F) in p.m. Runoff due to snow melt began
at 1100 and continued to 1850. Areas of bare soil appearing and
thawing rapidly, but saturated to surface.
03-16-75 Air temperature rose from overnight low of -3 C (27 F) to
maximum of 7 C (45 F) at about 1400. Runoff due to snow melt
began at 0930 and continued to 2000. Only patches of snow left.
Soil thawed to depth of 5 to 7 cm.
03-17-75 Air temperature rose from overnight low of -3 C (27 F) to maxi-
mum of 11 C (52 F) in p.m. Runoff due to melting of patches
of snow remaining on intermediate slopes began at 0916. All
snow had disappeared by 1100, but runoff due to seepage from
saturated surface soil continued to 1345. Surface soil was
thawed to depth of 7 cm, but frozen below that to depths greater
than 30 cm.
58
-------
TABLE 14. WEATHER AND WATERSHED CONDITIONS FOR THE PERIOD 03-18-75 TO
03-24-75
Thawing sequence followed by snow and rain producing runoff over
thawed plow layer over frozen subsoil.
03-18 to Daily maximum air temperatures ranged from 10 to 13 C (50 to
03-21-75 55° F) and daily minimums from -2° C to 4° C (28 to 39° F).
Soil was bare and thawing at rate of 2 to 5 cm. per day. Plow
layer (0-25 cm.) had thawed by 03-22-75, but subsoil was still
frozen to 46 cm. There was .08 cm rain on 03-19 and 1.47 cm rain
equivalent as snow changing to rain in A.M. of 03-21. Snow depth
of 10 cm. was observed in A.M., 5 to 7 cm at 1400 and 2.5 cm
at 1700 on 03-21. Thawed plowed layer took up rain and snow
melt without runoff.
03-22-75 Air temperature rose from overnight low of 0 C to maximum of
7° c (44° F) in P.M. Snow from 03-21 disappeared during the
night. Runoff from snow melt began at 0144. Seepage from
saturated plow layer probably contributed also. Runoff in-
creased sharply following each of three showers between 0445
and 0750. Scattered light showers continued to 0950, for a
total of 1.12 cm. Runoff continued to 1058.
03-23 to Air temperature rose from overnight low of -2 C (29 F) to maxi-
03-24-75 mum of 9° C (48° F) in P.M. of 03-23, dropped to 1° C (33° F)
during the night, then rose to a maximum of 16 C (61 F) on
03-24. Scattered showers from 2055 on 03-23 to 0550 on 03-24
(total 0.81 cm) produced runoff beginning 2156 and continuing
to 0830. Plow layer was near saturation from the precipitation
of 03-22 and there was still frost in the subsoil.
59
-------
TABLE 15. WEATHER AND WATERSHED CONDITIONS FOR THE PERIOD 03-25-75 TO
03-31-75
Freezing sequence followed by rain and snow producing runoff over
frozen soil subject to diurnal thawing and freezing at the sur-
face.
03-25 to Low night temperatures (-3 to -8° C or 18 to 27° F) created new
03-27-75 frost layer which would thaw during the day to a depth of 2.5 cm
(daily maximum air temperature on 03-26 and 03-27 was 0 C).
There were traces of snow each day.
03-28 to Air temperature rose from overnight low of -3 C (26 F) to maxi-
03-29-75 mum of 2° C (36° F) in early P.M. of 03-28. Light drizzle
beginning at 1540 and continuing intermittently to 0320 on 03-29
(total 0.53 cm) produced runoff beginning at 1904 on 03-28 and
continuing to 1408 on 03-29. Runoff occurred over soil thawing
and saturated at the surface. Frost was encountered at depth
of 2.5 cm at end of event (1400 on 03-29).
03-30-75 Air temperature dropped to -7° C (20° F) during night of 03-29,
producing about 2 cm of snow (equivalent to 0.20 cm rain). Snow-
fall began about midnight and continued to 1330. Temperature
rose by noon on 03-30 to 0 C maximum for the day. Runoff due to
snow melt began at 1252 and continued to 1544. Runoff was ter-
minated by falling temperature.
03-31-75 Air temperature rose from overnight low of -9 C (15 F) to
maximum of 9 C (49 F) at about 1400. Runoff due to snow melt
began at 1334 and continued to 1750. Runoff was terminated by
falling temperature and by the fact that essentially all of the
snow had melted.
60
-------
TABLE 16. WEATHER AND WATERSHED CONDITIONS FOR THE PERIOD 04-01-75 TO
04-13-75
Heavy snow (33 cm), followed by high winds and drifting, 5 days
of freezing and 6 days of daily runoff due to snow melt on soil
initially frozen to 15 cm.
04-01 to Watershed was bare on 04-01. Snowfall beginning in early A.M.
04-07-75 of 04-02 and continuing through 0545 on 04-03, equivalent to
2.26 cm rain, deposited 33 cm of snow. Severe drifting on 04-04
left variable snow cover (5 to 20 cm) on watershed and buried
collection facilities (drifts of 2 to 2 % m.). Low night temper-
atures -7 to -10° C (13 to 19° F) from 04-03 to 04-07 froze soil
to 15 cm. Below 15 cm, earlier frost had disappeared (unfrozen
to 60 cm), except frozen soil was encountered at 30 cm in one
core on a steep northerly slope which receives winter sun at a
very narrow angle of incidence. Bright sun and maximum air
temperatures of 1° C on 04-05, 2° C on 04-06 and 5° C on 04-07
melted snow on upper slopes, producing runoff which was inter-
cepted by snow on lower slopes but beginning to pond in conflu-
ence area (segment 5) on 04-07. Under snow cover, soil remained
frozen to surface but began to thaw quickly wherever the snow
cover was reduced to a thin crust.
04-08 to This was a period of daily runoff due to melting of snow from
04-13-75 04-02 and 04-03. No precipitation, bright sunny days with daily
maxima of 3 to 9 C (37 to 48 F), overnight lows of -3 to
-5 C (22 to 27 F). Soil on lower slopes and over much of
confluence area remained frozen to surface under snow or in bare
soil areas where saturated soil was subject to nightly freezing.
On bare upper slopes, frost had disappeared through the plow
layer by 04-11. Daily runoff declined on 04-12 and 04-13 as snow
and frost disappeared from confluence areas, permitting soil to
drain and accept continuing snow melt from drifts on lower slopes.
61
-------
TABLE 17. WEATHER AND WATERSHED CONDITIONS FOR THE PERIOD 04-14-75 TO
04-18-75
Sunny days and rising temperatures, culminating in historic rain-
fall event on bare soil (corn stubble on Watershed 06, soybean
stubble on 07).
04-14 to This was a period of sunny days and rising temperatures (daily
04-17-75 maxima rose from 8° C (47 F) on 04-14 to 16° C (60° F) on
04-17, overnight lows did not go below freezing and went only to
10° C (50° F) during night of 04-17. A few small patches of
crusted snow and ice remained on north-facing steep slopes in
A.M. of 04-18. Elsewhere, frost had disappeared, soils had
drained and were dry or drying at the surface.
04-18-75 Air temperature rose from overnight low of 10 C (50 F) to
maximum of 19 C (66 F) at 1700. The temperature maximum was
preceded by light rains beginning at 1320. At 1700, heavy rain
(5-min. intensities up to 7.4 cm per hour) began and continued
without interruption to 2200, after which light rains continued
to 2400. Total for the event (10.95 cm = 4.31 inches in 10 hours)
represents a long term historical extreme. It produced major
flooding in the Lansing areathe worst since 1947. Severe
washing and gullying occurred on the watersheds. An estimated
1400 kg of coarse sediment was deposited in the catchment.
Similar quantities of intermediate particle size sediment were
deposited in the after flume drainage channel; these represented
back-up water tapped off at intervals by hand directly through
the sample splitter which was under water. Actual loss of
sediment from the watershed was likely 3-fold (or more) greater
than can be calculated from our runoff data.
62
-------
TABLE 18. WEATHER AND WATERSHED CONDITIONS ON 08-21-75 AND 08-22-75
Historic rainfall event on soil under fully developed corn
(Watershed 06) or soybeans (Watershed 07).
08-21-75 This was a major historical growing season event. The USWB report
shows 2.46 inches (6.25 cm) for the 24-hours ending at 1700 on
08-21. Of this total, 0.51 cm came as a light drizzle beginning
at 1720 on 08-20 and continuing to 0320 on 08-21. This rain
served to saturate the canopy and slake surface soil prior to
near-record rainfall beginning at 1300 and continuing to 1910
(5.95 cm = 2.35 inches in 6 hours, with 5-min intensities up to
16.15 cm/hour). In spite of the cover provided by vigorous,
fully developed crops of corn and soybeans (better than average
for these watersheds), extensive washing and gulleying occurred
and sediment losses were substantial.
08-22-75 This event occurred while soils were still well charged with
water from the very heavy rain of 08-21. Runoff data for the
two watersheds serve to illustrate unique features, related to
canopy and topography, which influenced runoff and losses of
sediments and chemicals.
63
-------
o
o
ai
o
a
M
u
H
CO
PH
O
03
<
E-l
S
H
M
S
*"
a
3
2
a
CL,
NO
0
o
01
0)
rl
01
W
CQ 4J
s c
01
a
0)
CO
«j
UH
O
1
5
3.
I
Q
S
o
CM
S
g
V
CO
o
U!
O
CM
33
"O
01
CO
s
X
CO
IW 0)
O rH
O
s
o S
Z 10
K g
a rH
S MH
1
2 §
M v
3 t-4
5
i ra
o 8
5 x
S a
A8
c
^ 1
CM
2 §
33
g.
0.
S
a
g
Cu
E
a.
***
00
«
^j
"
^
j^
a
o
.0
§
M
*s
o
h
X
NO 1-
r- o
r» r-
§ S
CO f
Cv
4 d
a N<
rH f
rH i-
0 C
V V
CO C<
-a" Nt
rH C-
CM S
O C
CM <
ON O
ST Nj
r- it
m a
rH
1
!
i
i
i
1
I
1
(
i
rH r-
§ i
g >
C (
CO O
1
1
1
m i/
i
CM C*
rH r
CO C
O C
4 rH CO
N Sf CO
* -a1 in
i co -a-
> 0 0
1 O CM
1 rH rH
i in NO
> ON *
4 CM rH
1 rH rH
5 O O
V V
> ON *
J CM rH
in ON
S O CM
r H -a-
% CM CM
5 rH ON
rH
t r- CM
^ m co
1
i
1
1
1
1
!
i
i i
i i
H rH rH
1 g §
I g g
3 a c
a co co
i i
i i
i i
t m in
k 1 J,
H iH i-l
-, A A
300
CO
i-l
in
rH
O
3
/-N
T3 O
a M
01 rH
.£ O
V
C SI-
01
S O
O. CM
H
3
S
00
O
m
o
0
in
1
NO
|
O
rH
CM
O
CM
rH
O
sr
CM
00
rH
NO
O
1-1
CO
CM
r-.
rH
0
0
in
1
f^
1
CO
o
C
S
a
O 0
01 CM
C C
tH iH
e e
e. a.
CM O.
n
0
:*!
*
64
-------
m
r-.
1
CN
1
f"l
O
O
H
in
r-.
1
00
rH
CO
O
o
o
PS
Pu
SB
H
O
fe
O
ai
w
w
Q
I-H
M
H
CM
a
23
fc
fa
g
§
*
Jj
r
S
o
CM
u
n
2
VO
0
o
01
2
4)
4J
B
rH
a
Bi
0)
4J
CO
a
4J
3
a
rl
id
P-
4J
B
a
u
M
U-I
o
1
S
0
CM
«
o
0}
en
IE
S
o
CM
01
CO
B
U-I 0)
O iH
o id
5
5 i-l
2 0
(5 4J
B
H
1 S
i-H
B
H
in e
a
H
1 eg
CM S
CO B
M 0
3 -H
O u
CM rH 0)
-» CO fi
H
N
ON OO fl]
CM CM H
B 0 0 4J
O. . . CO
a
cd
rH vO 4J
I tS
CU rH
J +
rH CM ctj
3
cr
B rl
O. O O co
a. v v a.
i-H
VO CO CO
a 4J
PI* s*. ov «cf O
D. UH rH CM H
UH
Q
B
3
^i
2
00 CO
rJ B
^ IH CM m
00 CO
h
o
oo r--
oo
/-. B
UH -H
U-I 00
B B S r. in
H 3 .C CM VO
e rl 0
^ rH
rJ 0 -
* %
H CM vo
t-t id r
fl Q -^
-------
r-.
1
*^
i
0
g
r-*
t
m
7
o
o
M
Pi
Ed
PH
S3
EH
pi
g
b
O
3
pa
Z
CO
u
o
u
6-i
en
PH
|
fc
Pt4
0
03
,-J
EZJ
3
S
*
r-H
rsl
TABLE
SO
O
o
0)
«
01
4J
1
B
T-
K
Otf
0
01
H
SI
a si
a
4J 13
0 CM CO O 0)
& m oo H en
o«
1 O CO
J 01 -3- O O
«- tJ B
DO rH CO i-
H
Ot ^M
>s 0
<0
H X
U
Iu B
a p:
o
^ CM -* u
M *H O
01
i1
(i
01
I-l U
js co m
^- 4J CO iH
g -a
H
l-i 3 O O
JZ O CO CO
"s rt '
o *
§
B 4
M CO r
00 SO SO CO
o) m co so
4J * * *
i-t eg co so
nJ CM
H
1
1 x-s
E
B
H
. -. i
J 4-1 OS X-* /*N
-t H U U
i MI m ^ >^
j£ IM |S rS a § B
8O -H
S S a
co o o u
0) C! C
U C.
JJ V
M CO eg
33 O ^" 0 H
4J 4J
n in .n in in
i i i I I
in r^ oo cr» o
eg eg eg eg co
O O O O O
Q CO H
-- Ci *
1 O
; H
18 01 Ol
^ B e
co o a s
e in eg 5 o
O C3 CO CO
I 0
1 1 «
in m m in m
co es cs en m
en en en
-------
f-^
1
co
1
O
0
H
m
I
o
i
o
Q
0
w
fcj
H
o
tn
o
1
uj
Q
M
<
C/2
|
o
Z
g
~
rJ
-T<
h- 1
CM
CM
rJ
CQ
^0
O
13
Q
g i i
c o i/
O rH r
0. tH r-
IS 0 C
a. x v s
Ql
in a
a. d
CXi CJ & rH r-
U
O
vo m
0 ^
u o
4J
4-J
CO VJ
*J O CU ^O I'
^- M r1^
60 UH « O C
rH
O
CO
O
rH i-t
H U-t ^D v
O
CO S
CU
cu a
C H
H C3 T3 O ->
B ,0 cu -,
rC O iH
. 4J
B rC
0 60
H
S
0 -HOC
<-> S (
2 s
n ^ -H B e
j: r^ act
^ *H CO U
O CU
rl
a.
0
X! rH
O CN O
m in m in m i/
1 1 1 1 1
CN fi -^ r~« co c
o o o o o C
4- ' ' ' vo vo TO
H H CN CM 3
cr
rt
H rH H «H rH H
t8
5 O O O O p.
rH
3 m in c?i m cfl
* *J
H rH CM CN H
^ CM o m vo
3 rH rH O O
r vo CM rH PH
-d" O -3" CM
j o O O 0 O
e
d
o
1
Tj
Q)
W
e
en
o
a
*o
u
T3
QJ
CO
rH
Cij
4J
O
H
iw
o
c
r3
rH
cfl
4J
O
H
H
CO
rH
, |
rtj
U
O
H
|cy\ in oo r*- r~t CNJ
-* m < rH o o
oo -^ m ^r rH o o
rH rH CO rH rH rH
o o o o o o
V V
O r^ CO OO CTi O
O rH P"* rH O CO
rH rH CM rH
*j co m r^- co >*o
00 CTi O CO O O CM
rH CM rH
\D CO CM O rH
-------
.
o
w
1
ca
z
o
en
H
O
05 '
*
CM
ca
S
>£>
0
TJ
j2
a
so cn
0 H
V4 ** O\ rH
J3 rH CM OO
BCJ\ CO CO
1 t
CO CM CO
,C co ^o cn
r-i
m in m
r** r*n r*»
i i i
GO rH CM
H CM CM
-* CO 00
000
1
cn
o
I
a
01
CO
cn
o
cs
W
T)
CU
a
Ol
m
rH
CO
4-1
O
H
OH
O
§
m
rH
CO
4J
O
H
e
^H
CO
2
5
o r- CM
r- m O
in
CM
CD Sf rH
r- co o
00 0
CM
CM CO i-H
a\ CM o
-»
rH O CO
1 O VO rH
1 co in
1
1 o co o
00 -a1 1-1 rH
so CO
I
| rH l-~ CO
so * O
rH CM
rH
in in rH
00 so rH CM
& t-- 0
It ft
CA rH
CO 0 0 H
i-i in \o r-*
a) ^o -* r--.
H VO ^ OV
iJ
-------
eJ
O
CO
«
ffl
Z
o
CO
H
rt
g
hJ
»-J
S
IH
tM
H
o-
Z
Bi
2
r~-
^ ^^^
O in
Cx] r~*
CO CN
«: JH
I §
M
O O
35 U
§ 5
Z Ed
M £Q
CO O
W CO
S &
CJ H
M O
H Z
PL, _]
( 1
Q O
Z 00
2 S
§ 5
cd
rJ Is-
< 00
fti rH
i s
^
CM
,,
1
CO
H
O
0
j?
K
4J
CO
S
B
H
CU
JJ
CO
O
TJ
H
g
CO
B
0)
,£
p.
H
Q
CO
3
o
B
CU
4J
g
3
H
1
4-1
4-1
O
I
S
*?J
EC
O
CU
CO
a
X
O
to
o
-------
snow cover would thaw at the surface during the day but would usually freeze
again at night. Soil under drifted snow on intermediate, northerly-facing
slopes or under ponded water in central basin areas would thaw less quickly.
On intermediate slopes and in areas of ponding, ice would form at night
in the saturated surface soil and then build up on top over a succession of
diurnal thawing and freezing cycles. On intermediate slopes, this surface
ice and lingering snow which was frequently associated with it served to in-
tercept or divert seepage and runoff from slopes higher up. In bottom areas,
surface ice effectively sealed the soil surface so that ponding would occur
with even light runoff from surrounding slopes.
The ice layer in lower basin areas persisted until a thaw in mid-March
when it disappeared briefly, then formed again. This new ice layer disap
peared quickly at the beginning of April as internal drainage was restored
by disappearance of a deep frost layer. On intermediate, northerly-facing
slopes, surface ice and crusted snow persisted to the middle of April.
Several sequences of runoff, beginning 03-12-75, permit a detailed ex-
amination of effects of snow cover, surface icing, and frost on runoff losses
of sediments and chemicals under a variety of wintertime weather conditions
which are characteristic for the Great Lakes Basin.
The first sequence of weather and watershed conditions is described
in Table 13. To start with, the soil was frozen to 46 cm. An earlier aging
snow cover (about 0.5 cm on upper slopes and lower basin areas) had been re-
plenished by 5 to 10 cm of fresh snow on 03-06-75 and 03-07-75. This snow
would have been equivalent to about 0.5 to 1.0 cm of rain.
Because of the impermeable frost layer over all of the watershed area
and the iced soil surface in areas of ponding, snow melt during daylight
hours produced net runoff at the catchments on six consecutive days before
the last snow disappeared (cf. Tables 13 and 19). On sunny days, melting
snow would produce runoff even when air temperature did not rise above freez-
ing (as on 03-14-75).
Air temperatures fell below freezing each night. Thawed surface soil
would freeze and ice would form on ponded water. Nevertheless, runoff would
continue under the pond ice well into the night, fed by seepage from sloping
areas. Even after all snow had disappeared on 03-17-75, runoff continued,
fed by seepage from saturated surface soil over frost on sloping areas.
Sediment yields in this sequence were initially low, but increased with
daily temperatures and runoff volumes, until snow to provide melt water was
nearly gone (Table 19). Increasing sediment pickup reflected decreasing snow
cover and increased exposure of surface soil to thawing during the day. Be-
cause of the impermeable frost layer, thawed soil remained saturated except
on upper slopes which drained quickly by seepage downslope as soon as sources
of melt water disappeared. Pickup and movement of sediments could be ob-
served readily wherever runoff occurred over saturated soil.
Chemical concentrations were much greater in the sediment phase than in
70
-------
the aqueous phase. Kd values were of the order of 103 for paraquat and
for atrazine. The values for paraquat are based on 99 ppb in the aqueous
phase, since it was never detected; so the Kd values are undoubtedly low for
paraquat. Total loss of chemicals is a function of sediment yield and run-
off volume. During these events, major loss of both chemicals was in the
aqueous phase.
The second runoff sequence followed a heavy snowfall changing to rain
on soil thawed through the plow layer but still frozen in the subsoil (Tables
14 and 20). Ten centimeters of snow in the A.M. of 03-21-75 melted quickly,
its disappearance hastened by rain. The total precipitation was 1.47 cm,
but no runoff occurred at the catchment until the snow had practically dis-
appeared in early A.M. of 03-22-75.
Three significant features served to delay runoff: (1) interception
by the snow cover, (2) the recharge capacity of the thawed plow layer which
had had several days to drain by lateral seepage over the impermeable frost
layer, and (3) the impoundment capacity of the lower basin where surface ice
and ponded water from earlier events had disappeared.
In early A.M. of 03-22-75, runoff due to seepage from the saturated
plow layer in sloping areas preceded the first of a series of showers. Since
thawed soil was saturated and impoundment areas were already full to over-
flowing, runoff flows responded promptly to each of the early morning showers
on 03-22-75 as well as to each of a series of showers during the night of
03-23 to 03-24-75.
Sediment yields increased progressively over this sequence (Table 20).
This is indirect evidence for the extensive redistribution of sediments which
occurred on the watershed. Visible sediments were deposited in areas of pon-
ding in segment 5 and in association with lingering snow and ice on northerly
and easterly slopes in segments 2 and 3 (Fig. 8).
During the event of 03-23 to 03-24-75, losses of paraquat in sediments
exceeded losses in the aqueous phase.
The next two sequences (Tables 15 and 16) occurred during a period of
low night temperatures during which a new frost layer formed at the surface
while the older deep frost layer continued to recede. By 04-07-75, the new
frost layer was 15 cm thick, whereas the old frost layer was encountered only
on northerly facing slopes.
During the first of these sequences (Tables 15 and 21), soil on the im-
mediate surface would thaw during the day and freeze again at night. The
presence of the new frost layer is reflected in the rather heavy runoff from
0.53 cm of rain on 03-29-75 and in the fact that 2 cm of snow (0.2 cm rain
equivalent) produced runoff as it melted over two consecutive days.
On 04-02 and 04-03, 33 cm of snow fell on bare and frozen soil. After
a day of high winds and drifting, the snow on upper slopes disappeared quick-
ly under sunny skies, producing runoff which was intercepted by snow on inter-
mediate and lower slopes and by ponding under snow in lower basin areas.
71
-------
No runoff occurred at catchments until 04-08 (Tables 16 and 22). Runoff
volumes and sediment yields increased to a maximum on 04-10, then declined
as frost disappeared at points around and within impoundment areas, per-
mitting these areas to drain.
Unique spring and summer events
Except for northerly facing slopes, all snow and surface ice had disap-
peared and soils had drained and were drying at the surface by 04-18 when
a major historic rainfall occurred (Table 17). Useful comparisons can be
made between this event and a later major event in August when a vigorous
crop canopy was present (Table 18). Due to the magnitude of these events,
collection facilities were swamped. Data for the two watersheds in Tables
23 and 24 have been adjusted to allow for some of the difficulties encoun-
tered. Sediment yields, in particular, are probably 30% or less of the ac-
tual loss. The discrepancy between calculated and real losses is probably
greater for watershed 07 than for 06 because drainage away from the collec-
tion facilities is more rapid and our visual estimate of after-flume sedi-
ments (grossly approximate) was not realistic.
In our Special Memo of October 1976, we stated that undetected sediment
losses may have been less at 07 than 06. However, this is inconsistent with
our observation that, within the watersheds, sediments were displaced further
down the drainageway in 07 than in 06. Also, extreme scouring occurred in
segment 0705, and our field log notes that the pipe leading from the
Coshocton to the collection house had clogged with sand during the 08-21 event.
Total rainfall and total runoff on 08-21 were about one-half as great
as on 04-18, but rainfall intensities and maximum runoff flows were about
two-fold greater (Tables 23 and 24). Deep rills and gulleys were cut down
central drainageways during both events, but much less sediment was carried
to the catchments in the presence of vigorously growing corn and soybeans
on 08-21.
Total runoff from the two watersheds during these two events was roughly
in proportion to the watershed areas. However, sediment concentrations in
the 04-18 runoff from watershed 07 was 2-1/2-fold greater than for 06. Total
loss of sediment from the smaller watershed was almost twice as great. This
is consistent with our observation that sediments were transported further
down the central drainageway in 07 because of the much smaller area where
ponding and mid-row sedimentation could occur.
The much lower concentration of paraquat and diphenamid in sediments
from 07 during the 04-18 event is consistent with our inference that the pro-
portion of heavier, less adsorptive sediments increases as runoff velocities
and sediment loads increase. Objective comparison of sediment yields during
the 08-21 event cannot be made because of sampling difficulties, but the
lower paraquat in sediment from 07 again reflects the larger proportion of
heavier sediments due to more rapid flow of runoff down the narrower central
drainageway in 07.
On 08-22, while soils were still well charged with water from the 08-21
72
-------
event, 1.8 cm of rain produced runoff from both watersheds (Table 18). Total
runoff and sediment loss from 07 was proportionately much less in relation
to 06 than would have been expected (cf. Tables 23 and 24). Soybeans along
the central draw in 07 had lodged severely when their roots were uncovered
by deep rills and gulleys cut during the 08-21 event. The fallen vegetation
served to slow down runoff and intercept sediments more effectively than the
standing crop might have. Rills and gulleys in segment 0702 and upper por-
tions of 0705 were, to a great extent, filled again with sediment during the
08-22 event (cf. Fig. 8).
Corn on 06 did not lodge where deep cross-row rills were cut on 08-21
because of its well established brace roots. Also, cross-row cutting was
not as deep on 06 mainly because of the greater opportunity for runoff flows
to meander along the E-W axis in segment 0605 and lower portions of 0602.
Meandering was promoted by interlacing brace roots which, by intercepting
debris and sediments, would effectively block an established rill, forcing
runoff to find another channel.
Many of the deep rills cut in watershed 06 on 08-21 were still open af-
ter the 08-22 event. Nevertheless, sedimentation in deep rills and along
meandering midrow ponds contributed to greatly reduced sediment losses on
08-22. The 1.8 cm of rain on 08-22 was approximately 1/3 of that on 08-21,
but it produced only 1/30 the volume of runoff and 1/50 the sediment loss.
Of course, it must be recognized that the intensity of the rain on 08-
22 was about 1/5 as great as on 08-21 (5-minute maxima: 3.05 vs_ 16.15 cm/hr).
Interception by the dense canopies of corn and soybeans and by infiltration
into permeable sandy soils would have retained a larger proportion of the
08-21 rainfall on both watersheds.
Summation
Weather sequences which precede a runoff event are important in deter-
mining boundary conditions for the event itself. This is particularly true
during the colder seasons of the year when precipitation over periods of days
or weeks can accumulate as snow on the landscape, and when infiltration and
deep percolation are controlled by the distribution of surface ice or of
frost within the profile.
A variety of weather and watershed conditions were associated with run-
off sequences during winter 1974-75. These have been described in preceding
sections. Salient observations should be helpful in interpreting and model-
ling our data for winter runoff events:
a) The depth and distribution of snow cover reflected effects
of wind direction and local relief on patterns of drifting.
b) The depth, age and distribution of snow cover over the water-
sheds served to moderate the rates of penetration and dis-
appearance of frost in the profile. Direction of slope, in
relation to angle of insolation during the warmer part of the
73
-------
day, was an important factor affecting depth and quality of snow
cover and disappearance of surface ice and of frost within
the profile.
c) Snow, particularly new snow, can intercept rain and runoff
and can delay net runoff from a watershed.
d) Frost in the profile prevents deep percolation of melt water
or rain and permits runoff from minor amounts of precipitation.
e) When surface soil thaws on warm or sunny days, it tends to
remain saturated and readily suspended in runoff. Upper slopes
can drain quickly by seepage as soon as sources of melt water
disappear.
f) Nightly freezing forms ice in saturated soil on slopes fed by
seepage from above and in areas where ponding occurs during
the day. This ice layer is slow to thaw, protects soil on slopes
from erosion, but serves to intercept seepage on slopes and to
seal the surface in areas of ponding. Under these conditions,
seepage from thawed and saturated soil over frost in sloping
areas can extend runoff for a time after rain stops or after
sources of melt water have disappeared.
g) Sediment pickup from frozen soil is low, but sediment loads
in runoff increase as the area exposed to thawing and satura-
tion of surface layers increases.
h) The recharge capacity of thawed soil increases as deep frost
layers recede. The quantity of precipitation needed to pro-
duce runoff increases accordingly.
i) Sediment yields in runoff collected at catchments provide a
basis for identifying events and watershed conditions which
produced major redistributions of sediments and chemicals
within the watersheds.
The effect of crop cover on runoff and on losses of sediments and chemi-
cals was illustrated by two major historic events in 1975. Runoff volumes
and losses were much less in the presence of corn or soybeans in August as
compared with an event of longer duration but lesser intensity in April.
During a second event of moderate intensity in August, differences in runoff
volumes and losses for the two watersheds were related to differences in
topography and differences in properties of the crop cover (corn vs soybeans).
At all times, losses of chemicals were a function of runoff volume and
sediment yield. The concentration of chemicals was much greater in the sedi-
ment phase than in the aqueous phase (apparent Kd values of the order of 10^
to 1C)3). Nevertheless, total loss in the aqueous phase exceeded that in sed-
iments, except for paraquat. The Kd for paraquat was higher than for other
chemicals, consistent with its known susceptibility to adsorption. When
sediment yields were high, more paraquat moved with sediments than in the
74
-------
water phase. Total paraquat in runoff was usually much greater than for
other chemicals. This is consistent with its much greater persistence in
soils. However, calculated losses of paraquat are artifactually high to the
extent that actual concentrations in the water phase were less than the 99
ppb input for undetected concentrations.
75
-------
SECTION VI
LITERATURE CITED
Harper, L.A., A.W. White, Jr., R.R. Bruce, A.W. Thomas, and R.A. Leonard.
1976. Soil and Microclimate Effects on Trifluralin Volatilization.
J. Environ. Qual. 5:236-242.
Herbicide Handbook of the Weed Science Society of America. 1974. Third ed.
Champaign, IL.
Kearney, P.C., R.G. Nash, and A.R. Isensee. 1969. Persistence of Pesticide
Residues in Soils. _In M.W. Miller and G.G. Berg, Eds. Chemical Fall-
out. Current Research on Pesticides. Charles C. Thomas, Publishers,
Springfield, IL.
Payne, W.R., Jr., J.D. Pope, Jr., and J.E. Benner. J. Agric. and Food Chem.,
22, 79 (1974).
Pope, J.D., Jr. and J.E. Benner. J. Assoc. Off. Anal. Chem., 57, 202 (1974).
76
-------
H
U
W
CO
X
M
Q
S3
W
PH
5!
S-)
o
H
CO
CO
rH
W
H
!3
M
Q
W
CO
CD
CO
w
o
M
O
M
H
CO
W
P-f
|JL|
0
CO
w
M
H
PS
W
P-(
O
Pi
P-i
rH
1
W
H
C
o co
H CU
4J -rl
CO 4-1
rH 5-1
3 0)
B 0.
rl O
O M
cu
B
cfl
C
rH
Cfl
U
H
B
CU
U
J_l
CU
3
4J
o
cfl
M-t
3
C
cfl
S
0)
B
cfl
G
CU
T3
CO
}_4
H
0)
B
CU CO
rO C
H
0 C
H O
4-1 B
co B
0) O
P-i O
6-S x~>
CTv rH
CN cO
M
4-J *^».
rH CO
Cfl 13
to C
3
CU O
*"O CX
H
!-i (N
o
0 -H
H
13 cfl
1
-*j-
- C
*d~ O
1 -H
rH
s^ B
43 3
4-J -rl
Q) (3
B -H
H T>
Q -rl
1 rl
rH pi
" *H
rH 43
C
0
M
^
0)
U
4-1
cfl
3
0*
cfl
}_(
cfl
P-I
4J
Cfl
3
CT
cfl
V-J
cfl
a
0)
T3
Jj
O
CX
CU
H
tri
cfl
4J
4-1
01
&
B-S
O
1
H
TJ
CN
A Ol
CM T)
1 -H
rH B
P>*» Cfl
43 4-J
4J Ol
cu a
B cfl
H rH
o >
1 G
** . ^
J2J O^
(3
O
r-)
CX
CJ
H
C
W
13
H
B
cfl
C
CU
CX
H
T3
CO
13
G
3
O
CX
*^"
H
tfl
6-^
rH
, 1
43 O
4-j e
CU -H
1 B
<) cfl
1 rH
O >, 0)
rl CX f3
O O -H
rH !-l N
43 CX CO
0 O -H
1 CO h
csj -H 4-1
CX
M
O
O
r*")
60
H
0)
O
1
CO
&
H
CJ
XI
0)
J-l
4-J
43*
^*
QJ
C
H
N
cfl
M
4-1
cfl
77
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO. 2.
EPA-600/3-77-112
4. TITLE AND SUBTITLE
PESTICIDE RUNOFF LOSSES FROM SMALL WATERSHEDS IN GREAT
LAKES BASIN
7. AUTHOR(S)
B. G. Ellis, A. E. Erickson, A. R. Wolcott, M. Zabik,
and R. Leavitt
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Departments of Crop and Soil Sciences and Entomology
Michigan State University
East Lansing, MI 48824
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Research Laboratory - Athens, GA
Office of Research and Development
U. S. Environmental Protection Agency
College Station Road r Athens RA 30605
3. RECIPIENT'S ACCESSI ON- NO.
5. REPORT DATE
October 1977 issuing date
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
1HB617
11. CONTRACT/GRANT NO.
R-800483
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/01
15. SUPPLEMENTARY NOTES
16. ABSTRACT
An assessment is made of sources of variation in pesticide analyses for soil
cores taken during the period May 1973 through September 1974 from two watersheds. A
number of relationships to methodology, chemical species, topography, soil conditions,
and weather are identified. Criteria are given for assessing down-slope movement
within and between sampling segments and movement within the profile. A detailed
description is given of weather and watershed conditions associated with wintertime
runoff events on the larger watershed and with major spring and summer events on
both watersheds in 1975. Emphasis is placed on characterizing boundary conditions
at the beginning of each event in relation to weather sequences that preceded it.
Only portions of the pesticide data set, stored at the Environmental Research
Laboratory, Athens, GA, were used in these evaluations. However, important features
of soil, topography, management and weather are identified in relation to useful
variation in the data. The described relationships should be helpful in interpreting
and modelling data from these watersheds for both pesticides and nutrients.
17 KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Pesticides
Runoff
SedJjnentation
13. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
b. IDENTIFIERS/OPEN ENDED TERMS
Watershed studies
19. SECURITY CLASS (This Report)
UNCLASSIFIED
20. SECURITY CLASS (This page)
UNCLASSIFIED
c. COSATI Field/Group
68E
91A
21. NO. OF PAGES
88
22. PRICE
EPA Form 2220-1 (9-73)
73
',£ll S^GOTNMENTjlNG OFFICE 1977-757-140/6578 Region No. 5-11
--" -*" -
U S. Env
Groat Lakes IT^
------- |