Control of

WATER POLLUTION


from  cropland
mam^^mmmmmmam

Volumel— An overview
               ^^  Office of Research and Development

              ^ALfr-. Environmental Protection Agency

-------
                               REVIEW NOTICE

This report has been reviewed by the Office of Research and Development and the
Agricultural Research Service and approved for publication. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
                         DOCUMENT AVAI LABI LITY

While supply lasts, single copies may be requested from:

        (1) U.S. Environmental Protection Agency
            Agriculture and Non-Point Source
            Management Division (RD-682)
            Washington, D.C. 20460

            To order please cite

            REPORT NO. EPA-600/2-75-026b

        (2) U.S. Department of Agriculture
            ARS Information — Room 343A
            Federal Center Building - No. 1
            Hyattsville, Maryland  20782

            To order please cite

            REPORT NO. ARS-H-5-2

The public may also purchase this document from the National Technical Information
Service, 5285 Port Royal Road, Springfield, Virginia 22151 and from the Superintendent
of Documents, U.S. Government Printing Office, Washington, D.C. 20402
     For sale by the Superintendent of Documents, U.S. Government Printing Office. Washington. D.C. 20402

-------
                        Control  of
              WATER POLLUTION
                    from cropland
                   Volume I —An overview
    Authored by a Committee of Scientists of Agricultural Research Service, USDA.

  B. A. Stewart, Bushland, Texas	Coordinator
    D. A. Woolhiser, Fort Collins, Colorado	Hydrology
    W. H. Wischmeier, Lafayette, Indiana	Erosion
    J. H. Caro, Beltsville, Maryland	Pesticides
    M. H. Frere, Chickasha, Oklahoma	Nutrients

  Appendix C prepared by K. F. Alt, Economic Research Service, USDA.

  Prepared under an Interagency Agreement with the Office of Research and Development,
EPA. Project Officers were L. A. Mulkey, ORD, EPA,  and C. W. Carlson, ARS, USDA.
                            JUNE 1976
       Agricultural Research Service      / f* \ Office of Research and Development
       U.S. Department of Agriculture     \53ZZ/ Environmental Protection Agency

-------
                                          CONTENTS
Chapter
   I.   INTRODUCTION
          B. A. Stewart and D. A. Woolhiser
   2.   HYDROLOGIC ASPECTS OF NONPOINT POLLUTION
          D. A. Woolhiser [[[     7
            Fundamentals of Hydrology ............................................     7
            Components of the Hydrologic Cycle  ......................................    10
            Agricultural Chemical and Sediment Transport Models ...........................    17
            Agricultural Practices to Control Direct Runoff ...............................    18
            Research Needs [[[    23
            Literature Cited [[[    24

   3.   CROPLAND EROSION AND SEDIMENTATION
          W. a Wischmeier [[[    31
            Sediment Sources and Quantities .........................................    31
            Cropland Erosion  [[[    33
            Erosion Factors  [[[    36
            Erosion Control Methods  ..............................................    41
            Sediment Delivery Ratios  ..............................................    47
            Tolerance Limits [[[    48
            Research Needs [[[    50
            Literature Cited [[[    53

   4.   NUTRIENT ASPECTS OF POLLUTION FROM CROPLAND
          M. a Frere [[[    59
            The Problems [[[    59
            Sources of Nutrients  .................................................    61
            Transport from Cropland  ..............................................    68
            Effect of Control Practices .............................................    73
            Research Needs  .........  ..........................................    81
            Literature Cited [[[    82

   5.   PESTICIDES IN AGRICULTURAL RUNOFF
          J. a Caro  [[[    91
            Extent and Trends in Use of Agricultural Pesticides .............................    92

-------
  6.  INTERDISCIPLINARY RESEARCH NEEDS
         B. A. Stewart	   121
Appendix
A.   SIMULATION OF DAILY POTENTIAL DIRECT RUNOFF	   123
B.   SIMULATION OF POTENTIAL PERCOLATION AND NITRATE LEACHING	   149
C.   ECONOMIC ANALYSIS METHODOLOGY	   177
IV

-------
                                       Control  of

                            WATER  POLLUTION
                                  from  cropland

                              Volume Il--An  overview
                                            CHAPTER 1

                                         INTRODUCTION

                                    B. A. Stewart and D. A. Woolhiser
   Agricultural technology is one of the real strengths of
the United States. Although  the population has in-
creased steadily, food and fiber production has met the
domestic  needs  and  has  also  provided substantial
amounts for export, which is so important to the U.S.
trade balance. Fertilizers and pesticides have played  a
major role in this accomplishment because the acreage of
cropland has changed little in the last 45 years—agricul-
tural chemicals and other technological inputs have been
substituted for land.
   The marvels of agricultural technology have not gone
unchallenged.  Much of the blame  for polluted streams
and lakes is often placed on agricultural activities. Some
groups and  individuals have even called for a total ban
on  the  use of agricultural chemicals. At  the  other
extreme,  there  are those who claim that the use of
chemicals has not had  any  adverse effect on  the
environment and that there should be no restrictions on
or control of their use.
   The  ultimate decision  as to whether  agriculture is
contributing to pollution of particular water  bodies to
such an extent that active control measures are required
rests with  State or local  authorities. To assist these
officials   in reaching  this decision  and  in  choosing
appropriate  controls, the Federal Water Pollution Con-
trol Act Amendments of 1972, Public Law No. 92-500,
specify that the Administrator of the Environmental
Protection  Agency shall, in cooperation with  other
agencies,  provide guidelines for identifying and evaluat-
ing the nature and  extent  of nonpoint  sources of
pollutants. This two-volume  document on  control of
potential water pollutants from cropland was written by
scientists of  the  U.S.  Department of Agriculture in
response to this provision of the Act and at  the request
of the Environmental  Protection Agency. Volume 1 is a
User's Manual for guideline development. Here in Vol-
ume II we will review some of the basic  principles on
which control of specific pollutants is founded, provide
supplementary information, and present  some of the
documentation used in Volume I.
   Management  decisions relating  to the  control of
pollution from cropland involve a careful weighing of
potential costs and  benefits.  Some of the factors
affecting these decisions can be visualized by considering
the  schematic drawing  of  an  agricultural  system in
Figure  1. The system  itself is arbitrary and could consist
of a field, a state, or a river basin. Inputs to and outputs
from the system  can be identified and  inputs can be
classified as  controlled  or uncontrolled. Precipitation
and  solar radiation are  uncontrolled and contribute to
the stochastic nature of the outputs. The farmer has the
ultimate control over  the controllable inputs, subject to
physical  and legal constraints.  The  outputs can be
changed by varying the inputs or the system itself within
certain constraints imposed by physical laws.
   Most people would agree that the system should be so
modified and that the inputs to the system should be
controlled at a level that maximizes the net benefits to
                                                                                                     1

-------
society attributable to the system. Obviously the modifi-
cations  and controls chosen  depend  strongly on the
concept of social welfare and must include many costs
and  benefits not normally accounted  for  by a land
manager.
   The costs and benefits of the agricultural use of land
depend on the weather and other elements that may be
considered  as stochastic processes; therefore, the costs
and benefits associated  with the agricultural use of the
land  may   themselves   be considered  as  stochastic
processes. Consider the  four sample functions shown in
Figure 2. Xi(t) represents the amount of daily precipi-
tation; Ci (t) represents the amount of a chemical applied
to a field and is a stochastic process because the time of
application  depends  on  precipitation, stage  of crop
growth and  other factors associated with the particular
chemical; Yt(t) symbolizes daily surface  runoff which
may transport the chemical to a stream or lake; Y2(t)
represents  the  amount  of  chemical  transported  to
surface water; and B(t) represents the benefit process
(costs are negative benefits). Social costs  include those
incurred  when  surface  runoff  occurs shortly after a
chemical is  applied and those due to sediment. Other
costs include those normally borne by the farmer. The
benefit from the sale of the crop will vary annually as a
result of yield variability and the demand of society for
the particular crop, expressed as the price. Conceptually
the management decision problem is not difficult, but
practically it is formidable. First there is the question of
uncertainty—we do not know the long-term effects of
low, intermittent concentrations of many chemicals on
living organisms, including man. Therefore, we cannot
estimate the cost attributable to the specific transport of
a given  chemical. Since this, among many other uncer-
tainties, prevents the selection  of control practices and
institutional mechanisms that maximize net social bene-
fits, we may wish  to state the  objective  in  physical
terms. As an example, one could select those control
practices which maintained the average chemical concen-
tration below some threshold value for a given percent
of the time. If social costs associated with the  presence
of  this  chemical in  a stream exceeded  the  benefits
attributable  to it, this procedure would at least lead to
an  improved situation. However, as will be shown in
subsequent chapters,  our technology in predicting the
effects of changes in inputs and in the system  itself on
              Uncontrolled  Inputs

                 Precipitation
                 Solar  Radiation
                                       System
                                 (Field, Region)
             Controlled  Inputs

                Seed
                Fertilizer
                Energy
                Pesticides
                Management
                Capital
                Labor
                                                                  Outputs
                                                                     Crop  Production
                                                                     Water
                                                                     Sediment
                                                                     Nutrients
                                                                     Pesticides
                                    Figure 1.-Agricultural production system.

-------
X
                                                       Precipitation
o
                                                           Chemical
                                                           Application
                                                           Surface
                                                           Runoff
                                                                            -»-t
                                           Chemical Transport
                                                to  Stream
                                        Benefits  and Costs
GO
   1
                                                                     e
I      I
b      d
  Figure 2.-Sample functions of hydrologic processes and social costs and returns for agricultural system.
     a: planting cost to farmer; b: cost of applying chemical; c: social cost when chemical is transported to
     stream; d: harvest cost; e: return from sale of atop.

-------
concentrations of potential pollutants In surface waters
Is not developed well enough to  moke this approach
feasible.  As u  last resort, we  can  use direct  runoff,
percolation und  erosion as surrogate variables with the
assumption that u change  in any of these variables will
a Hoc I  water quality.  It must  bo  recognized that reduc-
tion of direct runoff or deep percolation may adversely
ulfcct  wutei quality  in Home Instances and, therefore,
may  create  water  quality und  quantity problems for
downstream water  users who depend  on  runoff  from
agricultural lands as u water supply.
   Before dealing with specific potential pollutants, it Is
important u> know something about the land resources
ol the U.S. because this has u significant bearing on the
use  ol  agricultural  chemicals,  Only the  land  in the
contiguous  48 slates  will hi- discussed, since there Is so
little cropland In Alaska and Hawaii.
   The contiguous 48 states contain  1,899,322,000 acres
of  land.  The  nonfederal   rural   land  comprises
 1,431 ,()30,000 acres, or 75 percent of the total. The use
 >l  this  rural land  is nearly  equally divided between
Cropland, pasture and range, and forest land (Table  1). A
land capability classification system1 has been developed
hy Hie U.S. Department of Agriculture  and u summary
of the amounts of various classes of soils and their use Is
given in  Figure 3. Class i soils are nearly level, have a low
erosion hazard, and arc suited to a wide range of plants.
They are deep, have high permeability and water-holding
capacity, are well drained, and are  fairly well supplied
with plant  nutrients or are highly  responsive to fertili-
sers, Soils in Class 11  have  some limitation* that reduce
the  choice  of plants  or require moderate conservation
practices.  They  often require   special   soil-conserving
cropping systems,  soil conservation practices, water-
control  devices,  or  tillage  methods  when used for
cultivated crops. Class U soils usually have gentle slopes
and  arc  moderately  susceptible  to  wind  and  water
erosion.
   ('lass  III soils are usually found on moderately steep
slopes  and  are  more  susceptible  to water and  wind
erosion  than  soils in  Clash II. They can  be used for
cultivated crops but  require highly effective conservation
practices that may be difficult to apply und maintain  if
erosion IK controlled.  Class IV soils ure  also suited for
cropland, but they require  careful management  and ure
often well suited for only two or throe common crops.
They ure usually found on steop slopes und are highly
susceptible to wind and water erosion.
Table I. Predominant land  uxc for nonfmloral rural lond In Hie
   contiguous 48 unto*  (USDA Sutlitlcal llullulln No. 461)
                                              Acre*
CROPLAND:
   Row crop* 	   160,041,000
   Cloie-ntruwn cropland liillow	,  , . . ,   132,620,000
   I'orunucropi	    77,629,000
   Cunwrvullon 11*0 .,,,,,	, ,    39,026,000
   Temporary Idle  	    II ,235,000
   Orchard*, vlncyurdi, and built I'nilti  	     5,060,000
   Open linul formerly cropped  	    11,592,000
                                           437,203,000
                                           101,061,000
                                           379,929,000
I'ASTURI AND RANCJI-
   Piiiturolund  	
   Ranuoliind   .......
 I'ORI'ST LAND
   Commercial  	  396,078,000
   Noimiiniiu'rriul	  ^62.860.000
                                           458,938,000
OTIII'K I  AND:
   In I'limi*  , , ,
   Nut In liirini ,
                                            27,779,000
                                            27,020,000
   1 National Inventory of Soil  and Water Conwrvutlon Need*,
1967. I'.S.  Department of Agriculture Stutlitlcul Hullolln No,
461, January 1971.
   Soils In Classes  V, VI, VII,  and VIII are limited in
their  use  and  ure  generally considered  unsuitable for
cultivation. Class V soils have little or no erosion ha/.ard
but  have  other  limitations that  are   impractical  lo
remove, lixamples arc bottom lands subject  to frequent
overflow,  stony soils, and ponded soils where drainage is
unfeasible, Class VI soils are usually limited lo pasture,
range, forest, or wildlife habitat. However, some Class VI
soils  can  be  used   for  common  crops with  careful
management.  Some  of  the soils  are also  adapted to
special crops such  us  sodded orchards, blueberries, und
similar crops. Class VII soils ure not suited for cropland,
and Class VIII soils are not only unsulted for cropland,
but have  limitations so severe  that  they are restricted
primarily  to recreation, wildlife habitat, water supply,
and esthetic uses.
   The  erosion ha/.ard  of cropland  increases sharply
from  Class I  through Class IV soils. Therefore, the larger
the cropland acreage on Class III and IV soils, the greater
the hu/.urd of erosion. Also, since sediment Is u principal
transport  mechanism for  agricultural   chemicals, the
potential for their loss is  much greater on these soils, For
example, It  is estimated  that from one-third to  one-half
of America's agricultural production  depends on fertil-
izer  use.  Therefore, If  fertilizer use were  eliminated,
cropland acreage would have  to be  greatly expanded.
Figure 3 shows that any  large increase in cropland would
have to come from Class III and IV soils, These soils ore

-------
   300
    250
   200
~ 150
    100
     50
                                                                                m
                                                                                &W;
                                                                                !•:•'•
                                                                                m
                                          CROPLAND

                                          PASTURE
                                          &  RANGE

                                          FOREST


                                          OTHER
                                             assa-a
   \y         v
LAND    CLASS
                                                                      v\
V\\
V\\\
       Figure 3.-UM of varloui clauoi of land In the 4ftcontlguoui itatoi (baud on data from USDA Statlitlcal Bulletin No. 461).
 leu desirable, not only because they are more erocltble,
 but  because  they   are  lower  In  fertility  and  yield
 considerably  less than Class I and II soils, particularly
 when fertilizers are not used.
    Unless sediment is controlled at a given level, the loss
 of  agricultural chemicals from equal treatments will
 usually  Increase as the soil class number increases. This
 suggests that one approach to control water pollution
 from  cropland is  to concentrate crops  to  the fullest
 extent possible on Class I and II  soils. These soils are
 naturally more productive, more responsive to fertilizers
 because of higher water-holding capacities, and easier to
 control  with  respect  to sediment losses. In'all likelihood,
 therefore, a high level of food and fiber production with
 the least Impact on the environment would result from
 using fertilizers and pesticides on the better lands whore
 their effectiveness Is  high and their loss is small. The use
 of  chemicals on  the  more  erosive soils  presents  a
 substantially  greater throat  to  the environment.  How-
          ever, it is possible to use them safely on these soils if a
          higher  level  of management  Is  practiced  to  control
          sediment and associated chemical losses, The treatments
          necessary to reduce losses are given In Volume I.
            How much agricultural  chemicals are affecting the
          environment is  certainly not clear. However, it appears
          that  sediment,  nutrient, and  pesticide losses urn  be
          controlled  at an  acceptable level  by the selection  of
          proper management systems. The challenge, therefore, is
          to develop appropriate assessment techniques and insti-
          tutional mechanisms so that controls are used only when
          needed. Also, recommending control practices for a large
          area is extremely difficult because the practices are often
          site-specific. The  concepts  presented in Volume I and
          the  material presented in the following chapters must,
          therefore, be considered only  us general  aids  to the
          decision-making process. Control  recommendations for
          specific sites must be developed by specialists within the
          area.
                                                                                                           5

-------
                                              CHAPTER 2

                       HYDROLOGIC ASPECTS OF NONPOINT POLLUTION

                                             D. A. Woolhiser
   Water,  running over the land surface or percolating
through the soil mantle to eventually appear as ground-
water runoff, is a potential carrier of  pesticides, nutri-
ents and sediment to streams and lakes. Any discussion
of nonpoint water pollution  from agricultural sources
necessarily involves  hydrology  because water  is the
primary transport medium.
   In this  chapter we will consider some hydrologic
fundamentals, including basic  physical principles and a
brief discussion of the stochastic nature of hydrologic
processes.  An understanding of the stochastic nature of
hydrologic processes  is important because it affects the
interpretation of experimental data. Components of the
hydrologic cycle  will  be described to illustrate the
physical basis  for  modifying surface  runoff  by agro-
nomic and engineering practices. These components have
been aggregated into  fairly general mathematical models
with the  objective of describing agricultural  chemical
transport.  Finally, documentation is provided  for most
of the 18  direct runoff control practices presented in
Section 4.2 of Volume I.
   Only those aspects of the hydrologic cycle that are
important in nonpoint pollution wfll be emphasized in
this report. Readers interested in a more comprehensive
discussion of hydrology are referred to several texts (23,
32, 64,  115}. Although results of experimental investi-
ptions  of the  effects of land use and treatment on
runoff from agricultural lands in the United States were
reported as early as 1927 (84), only recent experimental
work  wfll be considered here because dramatic changes
in agricultural practices have introduced time trends in
the amount of direct runoff from cropland (114).
   To understand how nonpoint pollutants move from
fields  to  surface  waters,  we must  first consider  the
physical form and placement of agricultural chemicals,
including nutrients and manures. Then we must consider
the various paths they must follow and the conditions
(such  as temperature, oxygen status, biological activity)
they may  encounter from field to stream or lake. Form
and placement of the potential pollutants are considered
in subsequent chapters; in this chapter, we wfll concen-
trate on the pathways.
                                FUNDAMENTALS OF HYDROLOGY
     Basic Physical Principles of Hydrology

   Two basic physical principles governing the amount
and distribution of water on the earth are those of mass
conservation and energy conservation. These principles,
along with several empirical relationships, form the basis
for most mathematical  descriptions of hydrologic phe-
nomena.
   The principle  of mass  conservation is frequently
illustrated by  the hydrologic  cycle or  by the  water
budget for an arbitrary volume of sofl. Morton's (52)
qualitative representation of the hydrologic cycle, Figure
1, is useful for introducing some hydrologic terms and
expressing the concept that the mass of water on earth is
assumed to be constant.
   If we consider the sector labeled "surface disposition
of precipitation-all forms" in Figure 1, applied to an
arbitrary volume of sofl with surface area, A, and depth,
d, as shown in Figure 2, we can write the conservation
equation for some arbitrary period of time, At:
                                                (D
where:
     P
    W
        precipitation received on the area, A
        water imported as a result of man's activities
Qs  =  net surface runoff (surface runoff leaving A
        less surface runoff entering A)
Op  =  net  lateral outflow (may  include ground
        water flow or unsaturated flow)

-------
                        Figuie l.-The hydrologic cycle-a qualitative representation [Horton (52)].
    AD  = increase  in surface storage (depression stor-
           age and detention storage)
    AS  = increase in soil water storage
     U  = net vertical outflow through soil or rock
     E  = evaporation   including  evaporation   from
           plants (transpiration).

   All  dimensions are in appropriate depth units. The
total  water  yield   for  this area,  both surface and
subsurface, is the difference between the total input of
precipitation  and  imported water, and evaporation,
assuming  changes in storage are insignificant. Each of
these components will be discussed in more detail in the
next section.
   The  amount  of evaporation is  controlled  by  the
amount of energy available at the layer of soil and air in
which plants grow. A conservation  of energy equation
may be written at this interface, expressing the relation:
   Net rate of incoming energy per unit area = net rate
of outgoing energy per unit area
       where:
          Rs
           P   =
         RL   -
           G   =
           H   =
           L   =
           E   =
           flux density of total short-wave radiation at
           the ground surface
           albedo  of the ground  surface (fraction  of
           incoming short-wave radiation that is re-
           flected)
           net flux density of long-wave radiation
           heat flux density into the ground
           sensible heat transfer into the atmosphere
           latent heat of vaporization of water
           evaporation rate
Rs(l-p)=RL
(2)
   Changes in heat storage in the vegetation and the heat
used  in photosynthesis have been ignored in Equation
(2).  They  would be  about 1% of RS.  The terms  in
Equation (2) are in units of heat energy per unit area per
unit time. The magnitude  of the terms in Equation (2)
may  vary substantially. If the soil surface is  wet  or
covered by actively  transpiring vegetation, most of the
available solar energy may be used to evaporate water. If
the soil surface is dry, most of the incoming energy may
be used to heat the air.

-------
   Equations (1) and (2) are linked by the evaporation
term,  E.  The  magnitude of  E  in  Equation  (1) is
effectively  limited   by the  amount of heat  energy
delivered to the surface A.
   Stochastic Nature of Hydrologic Processes

   A set of daily precipitation amounts on a particular
field, arranged chronologically, is an example of a time
series.  Other examples include  the daily direct  runoff
from a field, the daily amount of water percolating
below the depth d, or any of the terms in Equations (1)
or (2)  for  an  arbitrary period  of time. An essential
feature of these time series or processes is that they are
unpredictable  in  a  deterministic  sense. That  is,  we
cannot  predict with certainty how much rain will fall
tomorrow. These series can be viewed as sample  func-
tions of stochastic processes. A stochastic process may
be informally defined as a process developing in time in
a manner controlled  by probabilistic laws (81). Many
chance mechanisms are important in agriculture. Precip-
itation  is perhaps  the  most important,  but plowing,
planting and harvesting dates, and fertilizer and pesticide
application dates are certainly not deterministic.
   To analyze a  time series, one must  first  assume a
mathematical model for the stochastic process which is
completely specified except  for  parameter values that
can be  estimated on the  basis of an observed sample.
When the parameter values have been estimated, one can
obtain  certain  probability  expressions  that  may be
valuable in decision making.  For example, what is the
                                                 AS
                           •Q
                                                                                     B
                                   Figure 2.-Control volume for water balance.

-------
probability that the concentration of some substance in
runoff from  a field will exceed  a certain level  for 24
hours  or  more?  In an ideal  situation, our  stochastic
models would be constructed in accordance with Equa-
tions (1) and (2)  so that one might see how a change in
land management could affect the probability statement.
   We  cannot construct such ideal models. However, the
stochastic nature  of  hydrologic  phenomena  must be
appreciated  because,  if a  substance  is applied on  the
land, we cannot  guarantee  that it will never be trans-
ported to a stream. The probability of such  transport
happening in  a particular year may become infinitesi-
mally small, however.
   Another important concept is that of stationarity. A
stationary process is one where the chance mechanism
does not change with time. If we consider the process of
surface runoff from  a field in continuous corn, we can
see that it is not stationary. Not only are there periodic
changes within a year caused by seasonal phenomena but
also there are long-term  trends introduced by changes in
agricultural technology such as new tillage implements,
new crop varieties and increased fertilizer use. Therefore,
one cannot use long time series to estimate parameters
because the parameters are changing with time.
                           COMPONENTS OF THE HYDROLOGIC CYCLE
   In this section we will describe individual components
of  the  hydrologic  cycle  and  review  some  of  the
mathematical models that have been proposed or used to
describe  these  elements.  The discussions  will not  be
comprehensive  but will consider those aspects deemed
most significant for chemical  transport or for reducing
surface runoff.

                   Interception

   When rain begins, drops strike plant leaves and stems
and  are  retained  on these surfaces  by the forces of
adhesion and cohesion until a sufficiently thick film of
water  accumulates that  gravitation  overcomes  these
forces. If rain continues, the storage on an individual leaf
will become nearly constant, with as much water falling
from the leaf as falls upon it. Water will also be lost from
the film on vegetation by evaporation. There is some
disagreement as to whether this evaporation is a net loss
insofar  as  the  water balance of  a volume of soil is
concerned  (119).  If transpiration is limited  by  the
energy available, evaporation  from  the water stored  on
leaves is essentially equal to the amount of water that
would  be lost by transpiration unless the albedo of wet
vegetation  is  less than  that for  dry vegetation.  If
transpiration were  limited by  soil water content, how-
ever, the evaporation from a water film would be greater
and part of it could be considered as a net loss. Water
evaporated from mulch, dead leaves,  stems or  trunks
could  be considered a  net  loss if  energy  were not
limiting.  Rain intercepted by the canopy may  subse-
quently reach the ground by dripping from the leaves or
flowing down the stem. If stemflow is significant, it can
produce  substantial differences in soil-water content
over rather small distances (65).
   Although several have attempted to develop a mathe-
matical description of interception based  on physical
reasoning (51, 63), the  models have been rather crude.
Many mathematical watershed models do not include an
explicit component for interception (27, 50). Crawford
and  Linsley (26) combine interception and depression
storage into a single lumped storage with depletion by
evaporation and  transfer to  a lower  zone storage.
Boughton's model (10) and  the Tennessee Valley Au-
thority model (110)  assume  that  precipitation  will
accumulate in interception storage until  a threshold or
capacity value  is reached. The TVA model uses capa-
cities for forested watersheds of 0.05 inch in winter and
0.25 inch in summer. Saxton et al. (95) used a storage
amount of 0.10 inch for agricultural crops  and showed
that evaporation from  this source can be several inches
per year in a semi-humid climate.
   Zinke (119) concluded:
   "A survey  of  the data  in the literature indicates
interception storage amounts for rain of from 0.25 mm
to 9.14 mm (0.01 to 0.36 in.) and a similar range for
snow, 0.25 mm to 7.62 mm (0.01 to 0.30 in.).
   The storages indicate that one would not be greatly in
error  to  estimate  about  1.3  mm (0.05  in.) storage
capacities for rain for most grasses, shrubs and trees; and
3.8 mm (0.15 in.) for snow for trees."
   Jones (58) concluded that "a consistent difference in
storage capacity for trees, crops,  and grass of various
heights was not evident."
   From this brief review, interception does not appear
to  have  an  important  influence  on  runoff or deep
percolation from fields. However,  an increase in inter-
ception is partly responsible for the reduction in runoff
caused by conversion from clean-tilled crops to pasture
or meadow.
10

-------
                Depression Storage

   After interception storage has been  filled and the
infiltration capacity of the soil is exceeded so that all or
part of the soil surface is saturated, water will accumu-
late in surface depressions. Water  stored in depressions
either evaporates or infiltrates into the soil—none of it
runs off the surface.
   Depression storage can be  increased by agronomic or
engineering practices and, therefore, can be important in
reducing direct runoff from fields. For example, under
ideal circumstances, as much as 2.5 inches may be stored
in contour furrows constructed  with  a range furrowing
machine commonly used  in the  west  (77). Level bench
terraces  with  a capacity of over 2  inches have  been
installed in the  deep loess soils of western  Iowa (96).
Doty and Wiersma  (28) found  that  the  maximum
potential  depression storage  capacity for conventional
contouring and for bedding and listing practices ranged
from approximately  1 inch for contouring to as much as
3 inches for listing and bedding. The potential  surface
water storage  decreases as land slope increases and is
approximately half as great for a 7 percent slope as for a
1 percent slope.
   Agronomic   and  engineering practices  to increase
depression storage have a transient effect. With annual
cropping systems, storage capacity usually is maximum
in  the  planting  to  first  cultivation  period, which is
frequently the  most  important  for reducing losses of
agricultural chemicals by surface runoff. The  storage
capacity then decreases and reaches a minimum during
the harvest to plowing period (28). Contour furrows in
range and  pasture have maximum storage immediately
after installation.  Erosion and  trampling by livestock
gradually  reduce this storage capacity. For example,
contour furrows in eastern Montana had only half their
original  storage  capacity after 6-10 years,  and  the
average  effective life (storage > .05 inch) was about 25
years (77).
   Mathematical  descriptions  of  depression   storage
usually  represent it as a volumetric threshold that must
be exceeded before surface runoff occurs (10, 50). The
Stanford  model  (26) lumps depression  storage  with
interception but does not  assume a fixed  threshold
value. This approach can partially account for the spatial
variability of surface detention over the watershed. The
parameters in  these models are usually found by  trial
and  error  or  by  optimization techniques.  Very  little
information is  available that could serve as a guide in
choosing values for depression storage based on physical
measurements  in  the field.  The  work of Doty  and
Wiersma (28)  is  one exception for fairly simple  geo-
metric  shapes. Boughton (JO) suggested that the  "ran-
dom  roughness"of  soil surface microtopography  de-
scribed by  Burwell  and others (15, 16) might be an
adaptable measure of the depression storage.
   Although  manipulation  of depression storage  is an
obvious method of affecting surface runoff, the amounts
of change can be deduced only indirectly by analysis of
rainfall and  runoff.  The curve numbers for contoured
and contoured and terraced areas for the Soil Conserva-
tion Service method of estimating direct runoff shown in
Appendix  A  reflect  some  empirical data mixed with
judgment. Mathematical models that include depression
storage explicitly could be used  to  predict  changes.
However,  transciency  of  depression  storage  and  its
dependence on  precipitation, runoff, and erosion  make
prediction difficult.

                     Infiltration

   As  snow melts or rain falls on the soil surface or drips
from   the vegetation,  the  phenomenon of  infiltration
governs the amount  of water that will enter the soil and
thereby  greatly affects the  amount  of surface runoff.
Some  of the physical, chemical and biological charac-
teristics  of soil that affect infiltration can be manipu-
lated  by man through  agronomic and engineering prac-
tices. Therefore, changing the infiltration characteristics
of  soils  can profoundly affect the amount of surface
runoff as well as the amount of water stored in the soil
for plant use.
   Characteristics of both  the  porous medium and  the
fluid affect infiltration. The  porosity, pore-size distribu-
tion and  tortuosity  of soil pores all substantially  affect
infiltration  rates. Sands have higher infiltration rates
than silts or clays,  which have  a higher  porosity but
much smaller pores. Soil compaction by the trampling of
livestock  reduces  infiltration capacity and increases
surface runoff  (85).  From  this  evidence  it  can be
inferred  that compaction  by  machinery  would also
decrease infiltration  rates and that practices that reduce
machine traffic  on a field should reduce surface runoff.
   Raindrop impact on bare soil breaks up soil  aggre-
gates into their component particles or much smaller
aggregates. These particles  or small  aggregates can be
carried into larger  pores  by  water and  form  a thin
surface layer that has low hydraulic conductivity. This
surface layer may then control the infiltration rate (33,
47). Vegetation or mulches protect the soil surface from
raindrop impact and can prevent crust formation.
   Dense vegetation with  massive root  systems and
farming systems that leave substantial amounts of plant
residues  near the  surface  maintain high soil  organic
matter content  and promote aggregate stability, thus
maintaining high infiltration rates. Vegetation also has a
                                                                                                          11

-------
higher evapotranspiration rate between rains than evapo-
ration from  bare soil,  thus the soil water content is
reduced at the beginning of the next rain which increases
the rate and amount of infiltration.
   Tillage can increase the volume of large pores near the
soil surface  and thereby increase infiltration rates. The
effect is transient, however.
   Frozen soil usually has a lower infiltration rate than
unfrozen soil. If the soil is frozen  while wet, a dense,
nearly impermeable mass may result. However, if frozen
while dry, some soils will show little change in infiltra-
tion rate (76). The effect of increased viscosity of the
water  is apparently compensated for by the structural
change  caused  by freezing. Frost  usually penetrates
deeper if the soil is bare  than if  it is snow covered.
Therefore, practices that prevent snow from blowing
away tend to lessen frost penetration but the additional
snow deposited may increase runoff.
   Modern infiltration  theory based  on the  theory  of
unsaturated  flow or two-phase flow in porous media has
provided a basis for understanding infiltration behavior.
This theory has been presented in several recent texts or
reviews dealing with theoretical aspects of infiltration (8,
38, 46, 75, 83).
   Although  infiltration theory is  useful  in explaining
observed infiltration phenomena, it has just begun to be
used  in  quantitatively  estimating the effects of agro-
nomic  or engineering  practices  on infiltration and
surface runoff. The partial differential equations describ-
ing infiltration must be  solved by numerical methods—a
time  consuming and costly task if one wishes to find
long-term average effects or distribution functions  of
surface runoff.  Also,   this approach, with its  strong
physical  basis, requires costly  and difficult  measure-
ments of soil conductivity and diffusivity (109).
   Because of these difficulties, several infiltration equa-
tions,  either entirely empirical or  based  on simplifi-
cations of the  more general formulations, have been
used. Equations presented  by Norton (55) and Holtan
(48) are examples of the former. Gre.en and Ampt (39),
Philip (82, 83), Smith (104), Mein and Larsen (69), and
Brustkern and Morel-Seytoux (14)  used either simplifi-
cations of the  basic equations  or algebraic approxi-
mations of numerical solutions of the basic equations.
The first three of these  apply only to infiltration from a
ponded  surface  rather  than  to  rainfall conditions.
Although  solution of these equations is simpler and less
costly than  solution  of  the   more  rigorous  partial
differential equations, the  question of parameter esti-
mation remains. Usually they are estimated for different
soil and cover conditions from infiltrometer experiments
on small plots or data from small watersheds. Musgrave
and Holtan  (76) reviewed  much of the data available
before 1964.  Holtan  and his associates attempted to
develop  techniques for estimating parameters  in the
Holtan equation by using information available in soil
surveys (34) or by estimating parameters for  various
land-use or cover factors (49).
   Of the hydrologic models considered, none includes
an  infiltration  component  based  on  the numerical
solution  of unsaturated flow  or  two-phase  flow in
porous  media.  The   USDAHL  model (56)  utilizes
Hoi tan's  equation. The Bough ton  model, the  TVA
model and the Stanford model utilize empirical lumped
storage infiltration components, although the Stanford
model attempts to account for spatial variability by
assuming an invariant statistical  distribution of infiltra-
tion capacity. The USGS model (27) utilizes an  adapta-
tion of the Green and Ampt equation.

          Soil Water and Ground water

   Water  stored  in  the soil  and  rock is frequently
separated into  two   components:  the saturated or
groundwater zone and the unsaturated zone between the
groundwater and the surface. Water moves within the
unsaturated  zone in response to gravitational and  capil-
lary  potential  gradients. It may move generally down-
ward  during rainfall or snowmelt and generally upward
after a long, dry period, or it may move upward near the
surface and  downward in the lower part of the profile
simultaneously.  In general,  water movement  in the
unsaturated  zone will be predominantly vertical.
   In some soils, a rather permeable topsoil is underlain
by a slowly  permeable  clay layer. If infiltration is  rapid
enough, the  surface soil may become saturated, resulting
in flow which is predominantly in a  lateral downslope
direction and  is known  as  interflow. This water may
reappear on the surface some distance downslope or at
the foot  of the slope. Hydrologists generally agree that a
flow mechanism such as interflow exists; however, there
is some  argument about its importance.  Dunne  (31)
concluded from his measurements of subsurface storm
flow in Vermont that interflow (subsurface storm flow
in his terminology) did not contribute  significantly to
flood hydrographs. This does not mean, however, that
interflow is unimportant  to  water  quality in  some
regions.  Minshall  and  Jamison  (71)  presented   data
suggesting that interflow can exist on Midwest claypan
soils.
   An interflow runoff component  is included in the
Stanford model, the TVA model and the USDAH1/70
model. However, the volume of interflow runoff has not
been compared with field measurements because of the
difficulty in making such measurements. Therefore, it is
12

-------
difficult to ascertain if computed volumes are realistic or
are merely a result of curve-fitting procedures.
   As  Amerman (3) has  pointed  out, the separation
between the saturated zone and the unsaturated zone is
unnecessary from the  physical  point of view  and is
possibly misleading. Figure 3(a) shows a hypothetical
transverse  cross section  through   a valley  during  a
relatively dry  period and  Figure 3(b) shows a similar
section during a wet period. Under steady-state condi-
tions, the streamlines would represent the path lines of
water molecules or dissolved materials. However, hydrol-
ogic systems are usually unsteady so the streamlines are
continually shifting. The medium shown in this sketch is
isotropic  so the  streamlines are perpendicular to the
equipotential  lines. In an  anisotropic porous medium
that contained  a  relatively impervious  layer,  for ex-
ample, this would not be true.
   Figure 3 illustrates some  important  points about
transport  of dissolved  chemicals. Suppose  that in Figure
3(b) the soil surface from point A to point B was within
a single field. If we assume a steady state, the path line
from A to the stream  is much  shorter than that from B
to the stream. Therefore, a soluble chemical that leached
below  the root zone  on a particular day  would take
                                                 BEDROCK
                 Figure 3(a).-Cross section of hypothetical hydrological system during a relatively dry period.
             Figure 3(b).-Section of hypothetical hydrologic system during a wet period. [From Amerman, (3)]
                                                                                                           13

-------
much longer to reach the stream from point B than from
point A. Hydrodynamic dispersion will also affect the
arrival time at a stream  but has a relatively small effect
compared with  that  of macroscopic flow (78). How
much delay time might  be involved between the arrival
times of chemical constituents at the stream? This time,
of course, depends  on the path length and  velocity.
Some numerical models can answer this question if the
system  geometry and hydraulic characteristics of the
medium are known (12, 86,  100). As a crude approxi-
mation we might use the  results of Carlston (20), who
found that  the mean  residence  time  of groundwater
recharge  in  a Wisconsin drainage basin was  about 45
days. If we assume that  the time of travel for a particle
following the streamline originating between A and B is
equal to the  mean residence time and that the path
length of A is half the mean and the  path length of B is
1.5 times the mean,  the  arrival of a slug of chemical
distributed uniformly over the field extending from A to
B might appear at the stream over an interval of 45 days.
   Of course,  the  physical,  chemical  and  biological
processes that affect the particular constituent during its
travels  through  the  porous  medium  must   also be
considered.  For  example, if we  are  concerned with
nitrate transport, some  zones along the flow path may
be anaerobic and contain  carbon.  Under these circum-
stances, bacteria may convert the nitrate to harmless N
gas.  Such conditions  might well exist in the seep area
shown in Figure 3(b).
   Legrand (62) discussed  the  patterns of contaminated
zones of water in the ground. Although he considered
contamination  sources  of small  areal extent  (point
sources), his concepts can be readily applied to nonpoint
sources.  He   noted  that  when  contaminants  move
through the unsaturated zone  and reach the water table,
"enclaves" of contaminated  water  extend  from the
source in the direction of, groundwater movement,  as
shown in Figure 4. If the contaminant is not adsorbed or
chemically or biologically transformed, the enclave will
terminate at  a stream (Field  A, Fig. 4) and may cause
pollution.  Because  of  additional  water  entering the
stream,  the contaminant may be diluted to a harmless
level at a point C downstream.
   The boundary  of the  enclave  shown in Figure 4
assumes a constant inflow of the chemical uniformly
distributed over  the  field.  As a  rule, inputs will be
intermittent;  therefore,  the pattern  may consist of a
series of smaller enclaves moving toward the stream
completely surrounded  by uncontaminated water. The
dashed line emanating from the lower boundary of field
A terminates before reaching  the stream, illustrating the
situation in  which  some of the chemical may  pass
through a zone where chemical or biological reactions
may reduce its concentration to harmless levels before it
reaches  a stream. The same  situation  holds for the
contaminant moving from field B. Enclaves will change
in areal extent and in shape as the water table changes its
configuration naturally or by pumping of wells.
   Robbins and Kriz (92) presented  a  comprehensive
review of groundwater  pollution  caused by point and
nonpoint  agricultural sources. Their concern was pri-
marily with measurements of water quality within the
enclaves  of groundwater contamination,  not with the
effects on water  quality in streams and lakes.  For an
excellent  review  of  mathematical  models  describing
movement of chemicals in soils, see Boast (9).
   The subsurface transport  of  agricultural  chemicals
from a field to water bodies  is obviously very compli-
cated. Although we have a qualitative  understanding of
such transport, much uncertainty is involved in predict-
ing when and how much of a chemical  may reach a
stream or lake, or how much the amount can be reduced
by control practices.
   We can, however, identify certain goals of subsurface
water management  on agricultural  land. Maintaining
adequate  water in  the root zone  and encouraging a
vigorous  crop  are advantageous  from both the crop
production and water quality standpoints. Deep percola-
tion will occur in most humid and sub-humid climates.
Variations in soil characteristics will lead to substantial
differences in annual percolation, as shown in Figure 12,
Vol. I, and in Appendix B of this volume. In those areas
with  substantial  deep percolation, soluble agricultural
chemicals must be applied with more care.

                Evapotranspiration

   The sum of evaporation from the soil surface and
transpiration  from plants is called "evapotranspiration"
and represents the transport of water from the earth to
the atmosphere. It is important in agriculture because it
is required for crop growth. It is important in the loss of
potential pollutants from cropland because it affects the
volume of direct runoff and  the amount  of  soil water
that percolates to the saturated zone. Evapotranspiration
is obviously  a  major  component  in the  hydrologic
cycle—it  transports  about 70 percent  of the water that
falls  on  the  conterminous United States back to  the
atmosphere. This percentage can vary from 100 in arid
regions to about 50 or less in some mountainous areas of
the U. S.
   Three physical requirements must be met for evapora-
tion  from a  surface  to continue: 1)  There must be a
supply of heat  to convert liquid water to  vapor, 2) the
14

-------
                                                                 Direction of Ground
                                                                 Water  Concentration
         Pumped
            Well
                         C(Downstream  Limit of Harmful Concentration)
Figure 4.-Plan view of water-table aquifer showing enclaves of groundwater with high concentrations of a soluble material added to
                                              fields A and B.
vapor pressure of the  air must be less than that of the
evaporating surface, and 3) water must be continually
available. Evapotranspiration, through the latent  heat
term, is a component of the energy balance, Equation
(2), as well as the hydrologic water balance, Equation
(1).
   When  water is not limiting, the evapotranspiration
rate is limited by the radiant energy and advected energy
available.  Therefore,  a lower  limit  exists  for  total
runoff-the difference between  precipitation and poten-
tial evapotranspiration. Potential  evapotranspiration is
defined as  the hypothetical rate  of water loss from a
large, homogeneous  area  of  continuous  green crop,
under the given meteorological conditions, when there is
no resistance to water supply at the evaporating surface
(112).
   At most locations in the United States, soil water is
limiting some  time  during the  year,  so  the  actual
evapotranspiration will be less than the potential even if
the ground is fully covered by a crop. With annual row
crops, the ground will  not be covered by a transpiring
crop  canopy  for  a  substantial  period of  time,  so
evapotranspiration will be less than from grasses and the
total  runoff  will be greater. This is one  reason why
conversion from row crops to meadow or pasture usually
reduces runoff.
   The physics of evaporation and evapotranspiration is
discussed in several texts  (79, 93, 108). Here, we will
briefly outline the approaches  that have been used to
estimate  actual evapotranspiration  from cropland. In
general, the models used consist of a continuity relation-
ship, a means of computing potential evapotranspiration,
                                                                                                       15

-------
E0,  which  serves  as the  upper  limit of  the  actual
evapotranspiration rate,  a method  of computing actual
evapotranspiration, E, as a function of soil water content
when  it is  below some critical level, and a means to
modify E if the ground is  not  fully covered by  a crop
canopy.
   Potential evapotranspiration can be computed  by the
energy  budget method,  the aerodynamic method, or a
combination of the two (113). It can also be estimated
empirically  from evaporation pan data.  The equation
frequently used is:
where K is a "pan coefficient" and Ep is the evaporation
from a standard pan. Saxton et al. (94) found that daily
evapotranspiration computed by adjusted pan evapora-
tion was  highly  correlated (R2 = 0.87) with Eo calcu-
lated by the combination method.
   It now seems to be accepted  that actual evapotrans-
piration can  be less  than potential at soil water contents
above  the wilting point. Baier (4) presented a compre-
hensive review of this subject. The procedure used in the
simulations  of potential percolation in Volume I and
documented   in  Appendix B  of  this volume uses a
relationship  between E/EO and available water that is
similar to  those  presented  in the  literature. When
evapotranspiration estimates are  needed for  different
stages  of crop development, the evaporation rate can be
corrected by using  a  crop coefficient, Kp, that varies
according to the  stage of growth of the crop (57), or the
ratio E/EO may  be related  to the leaf area index,  LAI
(41, 90). The methods used for the  simulations  pre-
sented  in Volume  I  are  described in more  detail  in
Appendix B. For the extensive simulation study  of
percolation  and  nitrate leaching in Volume 1, a phys-
ically  more  realistic but more complex model such  as
presented by Richardson and Ritchie (89) or Saxton et
al. (95) would have been  difficult to  use with existing
time constraints.

                  Surface Runoff

   As the transport medium for dissolved chemicals and
for  sediments with their  adsorbed chemicals, surface
runoff is an important link  between fields and streams or
lakes.
   Surface runoff begins when the rainfall (or snowmelt)
rate  exceeds  the  infiltration  rate of  the  soil  and
depression  storage is filled.  Surface runoff is  classified
somewhat arbitrarily as either overland flow or channel
flow. Overland flow is  sometimes considered to be thin
sheet flow  over a relatively smooth surface. However, a
more general and realistic definition would be the  flow
that  is outside of the well-defined channel system. The
mean velocity of overland flow  is directly related to the
slope (laminar flow) or  the square root of the  slope
(turbulent flow) and is inversely related to the hydraulic
resistance of the surface. The hydraulic resistance varies
widely, depending on the surface characteristics, from a
Mannings resistance coefficient  of 0.02 for bare soil  to
0.4 for a dense turf (118). Such differences in hydraulic
resistance would result in water being about  six times
deeper on the turf than  on the bare soil for  the same
discharge.  The  velocity, of course,  would  be   only
one-sixth of that on the bare soil. The greater  depth on
the  dense   sod  would  allow  much  more   time  for
infiltration after the rainfall stopped,  resulting in less
runoff even if the infiltration characteristics of the soils
were the same. The decreased  shear stress on the soil
with a sod  cover would  also  result in  a much lower
erosion potential.
   Bailey, Swank and Nicholson (5) described the modes
of pesticide transport into and within the moving liquid
boundary during rainfall. The same processes would also
apply  to nutrient  transport by surface runoff.  This
transport process consists of four mechanisms, as shown
in Figure 5:1) diffusion  and turbulent transport of the
dissolved chemical from the soil water into the overland
flow  film,  2) desorption  of  the chemical  from soil
particles  into or toward the moving film, 3) dissolution
of stationary paniculate  matter trapped  at the bound-
ary,  and  4)  scouring of paniculate  matter  and its
subsequent dissolution.
   From  a consideration of these processes, one could
infer  that  practices that reduce  runoff  velocities and
prevent scour of particulate matter might reduce chem-
ical transport even if the total volume of runoff were not
reduced.  However, the  most effective practices  would be
those  that  increased  infiltration  rates  so  that  more
chemicals could be carried  into  the soil by bulk-flow
transport. An increase in depression storage would  have
a similar effect.
16

-------
                                                         Moving Liquid Boundary
                                                         (Film  Thickness  Increasing
                                                         Down  Slope )
                                                                                         Soil Surface
( D)  Moving
      Dissolution
PP = Pesticide  Particulate
in Motion
(C)  Stationary
     Dissolution
ps - Pesticide in
Solution in Motion
                                                                                     Water
                                                                                     Movement
(B) Pesticide     (A)  Liquid-Liquid
Desorption into or  Diffusion  Interchange
towards Moving    (Mass  Transfer of
Liquid  Boundary   Pesticide)
        D  Pesticide  Particle
        O  Soil Particle
         OPesticide  Adsorbed  on Soil  Particle
         /// Soil  Solution Containing   Pesticide
 Figure 5.-Modes of pesticide transport into and within the moving liquid boundary during a rainfall event. [From Bailey, Swank and
                                              Nicholson (5)]
              AGRICULTURAL CHEMICAL AND SEDIMENT TRANSPORT MODELS
   Models of agricultural chemical and sediment trans-
 port (which may be interpreted to include predictions
 made by them) represent our descriptions of how water,
 sediment, and chemicals move on fields or watersheds
 under  existing or  proposed  conditions. The  models
 should  not violate  the basic  physical principles of
 hydrology  and should  incorporate principles of chem-
 istry  and  biochemistry  needed  to describe chemical
 behavior in a biological system. They will also include a
 number of empirical relationships.
   Comprehensive  models of the  transport  of water
 (hydrologic models) have been used for about 10 years.
 Several of them have been discussed in previous sections
 of this  paper. Although special purpose water quality
 models were developed  as early as 1925 (JOT), general
 transport models were not  developed  until 1967  (54,
 55).
   Development   of agricultural   chemical  transport
 models  started around  1970. There have been several
 reviews  of the "state of the art" and the philosophy of
 modeling chemical transport (1,  24, 37, 59, 61, 116,
 117).
   Bailey et al. (5) have developed a conceptual model
 of pesticide runoff from agricultural lands, and several
                         quite general models are in the development and testing
                         stage (13, 25, 37).
                            The model  developers usually started with a hydrol-
                         ogic model that was developed for some other purpose
                         and  added  components  for  chemical  transport.  The
                         structure of the hydrologic model used thus served as a
                         constraint on  the  transport model, imposing all of its
                         constraints and shortcomings. When  these models have
                         been tested more thoroughly, many of the shortcomings
                         may be shown to be in the structure of the hydrologic
                         model.
                            These detailed  models may be quite useful in an
                         intensive study of a  particular field or watershed, but
                         they are far too complex for the extensive scope of this
                         report.  Field  data available for calibrating or testing
                         these models is also limited. Because of the present lack
                         of  knowledge  in modeling movement of agricultural
                         chemicals, we  used potential direct runoff and potential
                         percolation  as surrogate variables  in  Volume  I. We
                         implicitly  assumed that if surface runoff were reduced,
                         the  transport of chemicals would also be reduced. The
                         models  used to estimate potential  direct runoff and
                         potential percolation  are described in Appendices A and
                         B, respectively.
                                                                                                     17

-------
                  AGRICULTURAL PRACTICES TO CONTROL DIRECT RUNOFF
   Eighteen practices for controlling direct runoff, desig-
nated  as  Rl  though  R18, are presented in Volume I
(Table 14 and Section 4.2). Practices that reduce erosion
will  usually  reduce direct runoff, although to a lesser
extent. Therefore, the first 16 runoff control measures
have been assigned the same reference numbers as the
identical erosion control measures; only the alphabetical
prefixes differ. These agronomic or engineering practices
constitute means whereby direct  runoff may be reduced
as compared to  direct runoff from an index crop-
summer  row  crop (corn)  with  straight rows.  These
practices  are discussed in Volume I without supporting
documentation.  Table  1  contains citations of  articles
supporting statements made in Volume I and of articles
containing closely  related  information that may be
helpful in evaluating individual practices.
                                                     Most of the research cited in Table 1  was completed
                                                  within  the  last  15  years.  Earlier work  is not cited
                                                  because of possible nonstationarity caused by changes in
                                                  agricultural  practices.  The  percentage reductions  in
                                                  runoff are shown without any indication of statistical
                                                  significance. However, the decreases reported are consist-
                                                  ent with  the  physical  basis  of hydrology  discussed in
                                                  previous  sections.  Ranges  in  response for individual
                                                  practices  are  usually attributable to soil  and  climatic
                                                  differences and to sampling variability.
                                                     Additional documentation of land use and treatment
                                                  is given in the reports  of the Cooperative  Water Yield
                                                  Procedures Study Project (101, 102) and  in several re-
                                                  cent reviews (11,  61, 74).
                     Table 1. Bibliography on practices to control direct runoff (Volume I, Section 4.2)
  No.
                Runoff Control Practice
Page No.
   in
 Vol. I.
       Description
                                                       Citations
                                                               Significant Subjects
 Rl
           71
            No-till Plant in Residues
            of Previous Crop
 R2
   71
Conservation Tillage
                            Harrold, Triplett and
                            Youker (42)
                                                 Harrold and Edwards (45)
                                                 Harrold, Triplett and
                                                 Youker (43)
                                                 Harrold, Triplett and
                                                 Youker (44)
                                                 Smith and Whitaker (103)
Allis (2)
                                                 Free and Bay (36)
                                                 Manner ing and Burwell
                                                 (66)
Comparison of runoff and soil loss from
no-till and conventional tillage corn at
Coshocton, Ohio.

Single-storm runoff from a no-till field of
corn on a 21% slope was less than that from
straight-row corn field on a 6.6% slope but
slightly greater than that from contoured
corn on the 6.6% slope.

Comparison of 3 years of runoff and soil
loss data from no-till and conventional
tillage corn at Coshocton, Ohio.

Five-year average May-Sept, runoff was 0.44
inch for conventional and 0.04 inch for
no-till corn at Coshocton, Ohio.

In a 3-year period at McCredie, Mo., runoff
from corn with conventional tillage
averaged 5 in; runoff from  no-till fields was
6.7 in.

Over a 9-year period direct runoff from a
subtitled field in a corn-oats-wheat
rotation was 19% less than from straight-row
fields in the same rotation.

Runoff from a field of corn with mulch
tillage was greater than runoff from
conventional tillage.

Review runoff and erosion data from
various mulch tillage practices.
18

-------
                                                 Table 1. (continued)
                 Runoff Control Practice
 No.
Page No.
  in
 Vol. I.
R3
   72
R4
  72
R5
  73
R6
  73
        Description
 Sod-Based Rotations
            Meadowless Rotations
Winter Cover Crop
Improved Soil Fertility
                                                          Citations
                                                                                         Significant Subjects
                                                   Moldcnhauer et al (72)
                                                   Onstad (79)
Jamison, Smith and
Thornton (56)

Epstein and Grant (35)
                                                   Barnett (7)

                                                   Burwell and Holt (]7)
                                                   Carter, Doty and Carroll
                                                   (22)
                                                   Mannering, Meyer and
                                                   Johnson (67)
                                                   Moldenhauer, Wischmeier
                                                   and Parker (73)

                                                   Saxton and Whitaker (98)
                                                   Soil Conservation Service
                                                   (105)
                                                   Jamison, Smith and
                                                   Thornton (56)

                                                   Richardson (88)
                                                   Mannering and Burwell
                                                   (66)

                                                   Smith and Whitaker (103)
                                                   Jamison, Smith and
                                                   Thornton (56)
 Runoff from a till-plant field was slightly
 less than from a conventionally-tilled field
 for rainfall applied with a rainfall simulator
 in early June.

 Till-plant tillage up and down the slopes
 reduced runoff 42% over a 6-year period as
 compared to conventional tillage.

 Review of experiments at McCredie, Mo..
 Comparison of runoff and erosion from
 continuous potatoes and potatoes-sod-oats
 rotation at Presque Isle, Me.

 Rotation studies at Watkinsville, Ga.

 Compares runoff from corn-oats-hay
 rotation with runoff from continuous corn
 in west-central Minnesota.

 Runoff from Bermudagrass-corn rotation
 compared with continuous corn at Holly
 Springs, Miss.

 Evaluated effect of sod-based rotation on
 soil loss and  infiltration using a rainfall
 simulator.

 Runoff measured from corn-oats-meadow
 rotation and from continuous corn.

 Runoff measured from corn, small grain,
 meadow rotation and from continuous
 row crops.

 Runoff curve numbers established for
rotation meadow and row crops in
 rotation.

 Review of crop rotation experiments at
 McCredie, Mo.

 Measurements of runoff from cotton, corn,
oats rotation and oats, clover, cotton and
grain sorghum rotation at Riescl, Tex.

 No runoff reduction for corn interseeded
with legumes at LaCrosse, Wis.

 A small grain cover crop planted after
corn was removed for silage reduced runoff
 substantially (5.5 in. vs. 11.5 in.)

 Average annual runoff from plots at
 McCredie Mo. ranged from 6.8 to 11.7
inches during 1941-50. Lowest runoff was
 for well fertilized pasture; highest runoff
was for unfertilized corn-oats rotation.
                                                                                                                     19

-------
                                                  Table I.  (continued)
  No.
                  Runoff Control Practice
Page No.
  in
Vol.1.
       Description
                                                           Citation*
                                                                    Significant Subject»
 R7

 RX
   73

   73
 R9
   73
Timing of I idd Operation*

Plow-Plant Systems
Contouring
                                                     Carter, Dendy and Doty
                                                  !   Mofdenhauer. Wuchmeier
                                                     and Parker (73j
                                                     Saxton and Whitaker (98)
                                                     Wuchmeier (114)
None

Free and Bay (36.)
                                                     Mannering and BurweO
                                                     (66)
                                                     Wi*chmeier(H4)
Harrold and Edward* (45)
                                                     Alfc* (2;
                                                     Carter, Doty and
                                                  J   Carroll (22)

                                                     Onstad (79)
                                                     Wi*chmeier(iJ4>
                                                     SiilUfetal (9l>
                                                                       Average runoff wa* 9.01 inches from
                                                                       improved fertilized pasture* and 17,9 from
                                                                       unfertilized pasture* for a 6-year period.
                                                                       fcxperiment wa* at Holly Spring*, Mm.

                                                                       For a 10-year period at Oarinda, Iowa.
                                                                       runoff from continuous corn receiving
                                                                       nitrogen fertilizer wa* 2S'/' let* than runoff
                                                                       from unfertilized corn.

                                                                       Average annual runoff from row crop* with
                                                                       fall fertilization wa* 1.10 inche* at
                                                                       compared with 2.16 inche* from row crop*
                                                                       receiving only ttarter fertilizer.

                                                                       The ratio of runoff from corn land to
                                                                       runoff from adjacent fallow decrease* with
                                                                       increatci in corn yield. Much of the
                                                                       increase in corn yield » cauted fy higher
                                                                       fertilizer u*e.
   Average growing *ea*on runoff for plow-
   plant com wa* lew than runoff from con-
   ventional corn at MarceNu*, NY,

   Runoff from nnrulated rainfall on pkm -
   plant corn wa* tew than that from con-
   ventional com.

   Report* three *tudiet that 
-------
                                                  Table I. (continued)

No.

Runoff Control Practice
Page No.
in
Vol. I.
Description


Citation*


Significant Subject*

 RIO
 RH
 R12
 73
 73
R13


RI4
R15
RI6
74
74
           Graded Row*
 Contour Strip Cropping
 Terraces
Grassed Outlets
                     Ridge Planting
74
76
Contour Listing
                     Change in Land Use
                                                 |   OnstadandObonrap)
                                                    Pfest(I06)

                                                    Soil Con«ervation Service
:   Moldenhauer et al (72)

I

|   None



i   Baird and Richardson (6)



   Richardion (f§)
                                                    Spomer,, Hetnentann and
                                                    PiesHUM)
                                                   Saxton and Spomer (96)
                                                   Saxton, Spomer and
                                                   Kramer (97)

                                                   Soil Conservation Service
                                        None
                                        Maimeringand Burwell
                                        (66)

                                        HoUtntuuer ft aJ 
  Hitter ft al (91)

  Manneringand BurweU
  (66)
                                        Jamison, Smith and
                                        Thornton (56)
 Runoff from contoured and conservation
 tillage field* in corn.

 Compare* runoff from contoured corn with
 level terraced com and with meadow.

 Runoff curve number* established for con-
 toured row crop*.

 Graded row* on dope* of 3 A to 9% did not
 reduce runoff.

 Effects inferred from runoff reduction by
 meadow.

 Terracing alone on heavy day toil* of
 Texas Blacklandt had little effect on runoff
 volume.

 Effect* of conservation practice* including
 terracing on runoff.

 Level terrace* drastically reduced turface
 runoff in we*tem Iowa but groundwater flow
 increased.

 Level terrace* reduced discharge of water,
 sediment, nitrogen and phosphorus.

 Fourteen percent of water yield from level
 terraced watershed wa* surface runoff.
 Sixty-four percent of water yield from
 contour watershed  wa* surface runoff.

 Effects of level terracing on runoff and
 erosion.

 Runoff curve numbers established for
 graded terrace*.

 No data available on effect* of grassed
 outlets on surface runoff.

 Ridge planting on contour reduced direct
 runoff.

 Ridge planting on graded rows did not
 reduce direct runoff from a simulated rain.

 Ridge planting reduced pesticide runoff.

Cite Iowa study where annual direct
runoff from contour-listed com wa* 55% less
than that from straight-row planting up and
down the slope.

Direct runoff from pasture and meadow wa*
lower than that from com in central
Missouri.
                                                                                                                    21

-------
                                                   Table 1. (continued)
                 Runoff Control Practice
 No.
Page No.
   in
Vol. I.
       Description
                                                          Citations
                                                                   Significant Subjects
 R17
   76
Other Practices
 R18
   76
Construction of Ponds
                                                    Dragoun (29)
                                                    McGuinness and Harrold
                                                    (68)

                                                    Rice and Dragoun (87)
                                                    Saxton and Whitaker (98)
                                                    Spomer, Heinemann and
                                                    Piest (106)

                                                    Thomas, Carter,
                                                    and Carreker (111)

                                                    Wischmeier (114)
                                                    Hanson et al (40)
Soil Conservation Service
(105)

Dragoun and Kuhlman
(30)

Mickelson (70)
                                                    Neff (77)
                                                    Schwab and Fouss (99)
Langbein, Hains and
Culler (60)
                                                                       At Hastings, Nebr. average annual direct
                                                                       runoff was 0.20 inch from watershed in
                                                                       grass and 5 .24 inches from cultivated fields
                                                                       in row crops.

                                                                       Water yield decreased when watershed was
                                                                       reforested.

                                                                       Reseeding cropland with perennial prairie
                                                                       grasses reduced runoff by 94% in a 2-year
                                                                       period.

                                                                       Comparison of direct runoff from pasture
                                                                       and  meadow.

                                                                       Comparison of direct runoff from perennial
                                                                       grass, contoured corn and level-terraced corn.

                                                                       Bermudagrass meadow reduced runoff.
For nearly 5000 plot-years of data analyzed,
runoff from row crops averaged 12% of
total rainfall, while that from meadow
averaged 7%.

Effects of grazing intensity on direct
runoff from rangeland.

Runoff curve numbers established for
pasture and meadow.

Contour furrowing reduced runoff from
pastures.

Storing runoff in leveled areas for crop
production.

Storage capacity of contour furrows in
rangeland.

Surface runoff and tile flow from fields
with com and grass cover.

Hydrology of ponds and stock-water
reservoirs.
22

-------
                                          RESEARCH NEEDS
   Research on the effect of land management practices
 on  hydrology has  usually involved three  steps:  1)
 intensive  experimental  measurements  on  plots  and
 watersheds, 2) analysis of the data using some type of a
 mathematical model, and 3) generalization of results for
 more extensive application.
   Measurements  made in the first step are frequently
 governed by the model that is to be used in the second
 step. For example, in most of the experimental work
 examined, only rainfall and runoff were measured. This
 was adequate  when the only question was "Will treat-
 ment A reduce  surface runoff?" and when  the study
 could be maintained for enough time to obtain statis-
 tically  significant results.  Unfortunately,  we can no
 longer afford this luxury  of time. Policy decisions must
 often  be made  quickly  and by  the  time we  have
 statistical significance, the practice may be obsolete.
   The alternative is to develop more detailed models to
 use  as a framework in  analyzing the data and to obtain
 more intensive measurements in a  shorter time. The
 experimental  data  can  be used  to estimate model
 parameters  and  techniques  must  be  developed  for
 predicting parameters from readily obtainable physical
 measurements. Simulation can then be used to evaluate
 the  stochastic properties  of the system and to examine
 long-term effects.
   Stochastic  models of  point and areal  precipitation
 must be developed  and the parameters regionalized by
 mapping or other techniques.  As plant growth models
 and  other biological processes are included in hydrologic
 models,  the stochastic inputs must be expanded  to
 include temperature and  radiation. Obviously, the joint
 probability  structure   of precipitation,  radiation and
 temperature must be maintained.
   Prediction of  runoff  from complex  areas is still
difficult  and  needs  a great deal  of work from the
standpoint of  water quality. If concentrations of the
chemical are important  we must estimate the joint
probability  structure of discharge  of chemicals to a
stream and the quantity of water in the stream.
   A second generation of agricultural chemical trans-
port models should be developed after the first gene-
ration models have been tested and their strengths and
weaknesses  identified.  Material models, systems which
retain  many of the important  characteristics of real
watersheds  but  are  easier to  manipulate  and control,
may play an important part  in model testing and in
understanding the  significance  of parameters.  These
models, which  would  be less  than an  acre in size but
much larger and more  complex than a soil column  or a
lysimeter, would allow deliberate departures from homo-
geneity. The sensitivity of model parameters to such
variations could then  be  established under controlled
conditions.
   The third aspect of past research, generalization of
results for more extensive application, needs much more
emphasis.  The SCS curve number procedure for estimat-
ing direct runoff and the  Universal Soil Loss Equation
are examples. These techniques were developed before
the  advent  of,  or during the infancy of  high-speed
computers and the models used were accordingly simple.
This  constraint  has been  relaxed  considerably so  it
appears that significant improvements  could be made.
For example, the  direct  runoff  estimation procedure
could  be  improved  by  incorporating a  simple  soil
moisture accounting instead of the antecedent rainfall
index. The  functional  form of the equation could be
changed to more closely approximate results predicted
by modern  infiltration theory. Results from  complex
hydrologic models  should be  used along  with experi-
mental results to develop a new procedure for estimating
direct runoff.
   It should  be  emphasized that no single model  will
meet  all needs. We need a  set  of models, involving
increasing  abstraction,  and an objective procedure for
selecting the appropriate one for the job at hand.
                                                                                                      23

-------
                                         LITERATURE CITED
 1.  Aleti, A.,  Chiu,  S. Y., and McElroy, A. D.  1974.
    Methods for identifying and evaluating the nature
    and extent of nonpoint sources of pollutants from
    agriculture. Proc.  1974 Cornell Agr. Waste Manag.
    Conf., 10-23.

 2.  Allis, J. A. 1952. The story of two watersheds. Jour.
    Soil and Water Conserv. 7(5): 243.
12. Bredehoeft,  J.  D. and Finder, G. F. 1973. Mass
    transport  in  flowing  groundwater. Water  Resour.
    Res. 9(1): 194-210.

13. Bruce, R. R., Harper, L. A., Leonard, R. A., Sriyder,
    W. M., and Thomas, A. W. (Unpublished). A model
    for runoff of pesticides from  small upland water-
    sheds.
 3.  Amerman, C. R. 1973. Hydrology and soil science
    in field soil water regime. Soil Sci. Soc. Amer. Spec.
    Pub. No. 5, Madison, Wis.: 167-180.

 4.  Baier, W. 1967. Relationship between soil moisture,
    actual  and potential  evapotranspiration. In  Soil
    Moisture, Proc. Hydrology Symp. No. 6, Nat. Res.
    Counc. Canada: 151-191.
 14. Brustkern,  R. L., and Morel-Seytoux, H. J.  1970.
    Analytical treatment of two-phase infiltration. Jour.
    Hydr. Div. Proc. ASCE 96(HY12):2535-2548.

 15. Burwell, R. £., Allmaras, R. R., and Amemiya, M.
    1963. A field measurement of total porosity  and
    surface  micro-relief of soils. Soil Sci. Soc. Amer.
    Proc. 20:697.
 5. Bailey, G. W., Swank, R.  R., and Nicholson, H. P.
    1974. Predicting pesticide  runoff from agricultural
    land:  A  conceptual  model. Jour. Environ. Qual.
    3(2):95-102.

 6. Baird, R. W., and Richardson, C. W.  1969.  Effects
    of conservation treatments on water yield. In W. L.
    Moore and C. W. Morgan, eds.  Effects of watershed
    changes  on streamflow: 69-78. Univ. Texas Press,
    Austin, Texas.
 7. Barnett,  A. P.  1965. Using perennial grasses  and
    legumes to control runoff and erosion. Jour.  Soil
    and Water Conserv. 20(5):212-215.

 8. Bear, J. 1972. Dynamics of fluids in porous media.
    American Elsevier Pub. Co., New York, 764 p.

 9. Boast,  C. W.  1973.  Modeling the movement of
    chemicals in soils by water. Soil Sci.  115:224-230.

10. Bough ton, W. C. 1968. A mathematical catchment
    model  for estimating  runoff. Jour. Hydrol. (N.Z.)
    7(2):75-100.

11. Boughton, W. C. 1970. Effects  of land management
    on quantity and quality of available water. Report
    No. 120, Water  Res. Lab.,  Univ. New South Wales,
    Manly Vale, N.S.W., Australia, 330 p.
 16. Burwell, R. E., Allmaras, R. R., and Sloneker, L. L
    1966. Structural alteration of soil surfaces by tillage
    and rainfall. Jour. Soil and Water Conserv. 21:61.

 17. Burwell, R. E., and Holt, R. F. 1967. Soil and water
    losses from  Barnes  soil in west-central Minnesota.
    Minnesota Science, Univ. Minnesota Agr. Expt. Sta.
    23(4):10-12.

 18. Burwell, R.  E.,  Schuman,  G. E., Piest,  R.  F.,
    Spomer, R. G., and McCalla, T. M. 1974. Quality of
    water discharged from  two agricultural watersheds
    in  southwestern   Iowa.  Water  Resour.  Res.
    10(2):359-365.

 19. Byers,  H. R.  1959. General meteorology. Third ed.
    McGraw-Hill, New York, 540 p.

20. Carlston, C. W. 1964. Tritium-hydrologic research:
    Some results of the U. S. Geological  Survey Re-
    search Program. Science 143(3608):804-806.

21. Carter, C. E., Dendy, F. E., and Doty, D. W. 1966.
    Runoff and soil loss from pastured Loess soils in
    north Mississippi. Proc. Mississippi Water  Resour.
    Conf., Jackson, Miss., April: 104-118.

22. Carter, C. E., Doty,  C. W., and Carroll,  B. R. 1968.
    Runoff and  erosion characteristics of the brown
    loam soils. Agr. Engin. 49(5):296.
24

-------
23. Chow, V.  T.  ed.  1964.  Handbook of  applied
    hydrology. McGraw-Hill, New York.


24. Cooper, C. F. 1973. Hydrologic modeling-a vehicle
    for quantifying man's  impact  on the environment.
    Trans. ASAE 16(3):578, 579, 581.

25. Crawford, N. H., and Donigian, A. S., Jr. 1973.
    Pesticide transport and  runoff model for agricultural
    lands. EPA-660/2-74-013, 211  p. U. S. Govt. Print-
    ing Off., Washington, D.C.

26. Crawford, N.  H.,  and Linsley, R. K.  1963. A
    conceptual model of the hydrologic cycle. I.A.S.H.
    Publication No. 63. Symp. Surface Waters: 573-587.

27. Dawdy, D. R., Lichty, R. W., and  Bergman, J. M.
    1972. A rainfall-runoff simulation model for estima-
    tion of flood peaks for small drainage basins. USGS
    Prof. Paper 506-B, 28 p.

28. Doty, C.  W., and Wiersma, J.  L. 1969. Geometric
    shaping  and  contouring  of  land  as related  to
    potential  for  surface-water  storage. Trans. ASAE
    12(3):322-325,328.

29. Dragoun, F. J.  1969. Effects of cultivation and grass
    on  surface    runoff.   Water  Resour.  Res.
    5(5):1078-1083.
35. Epstein, E., and  Grant W. J. 1966. Rock and crop
    management effects  on runoff and  erosion in a
    potato-producing area. Trans. ASAE 9(6):832-833.


36. Free, G. R., and  Bay, C. E. 1969. Tillage and slope
    effects  on  runoff  and  erosion.  Trans.  ASAE
    12(2):209-219.

37. Frere, M. H.,Onstad,  C. A., and Holtan, H. N. 1975.
    ACTMO, an agricultural chemical transport model.
    ARS H-3, U.S. Dept. Agr., Washington, D.C., 56 p.

38. Gardner, W.  R.  1967. Development of  modem
    infiltration  theory and  application  in hydrology.
    Trans. ASAE 10:379-381, 390.

39. Green, W. H., and Ampt, G. A. 1911. Studies on soil
    physics: I. Flow of air and water through soils. Jour.
    Agr. Sci.4:l-24.

40. Hanson, C. L., Kuhlman, A. R., Erickson, C. J., and
    Lewis, J. K.  1970. Grazing  effects on runoff and
    vegetation  on western  South Dakota  rangeland.
    Jour. Range Manag. 23(6) :418-420.

41. Hanson,  C. L. 1973. Model for predicting evapo-
    transpiration from native rangeland in the northern
    Great Plains. Ph.D. Thesis, Utah State Univ., Logan,
    Utah,116 p.
30. Dragoun, F. J. and Kuhlman, A. R. 1968. Effect of
    pasture management practices on runoff. Jour. Soil
    and Water Conserv. 23(2):55-57.


31. Dunne,  T.  1970.  Runoff production in a humid
    area.  ARS  41-160, U.S.  Dept. Agr.,  Washington,
    D.C,  107 p.


32. Eagleson, P. A. 1970. Dynamic hydrology. McGraw-
    Hill, New York, 462 p.


33. Edwards, W. M., and Larson, W. E. 1969. Infiltra-
    tion of water into soils as influenced by surface seal
    development. Trans. ASAE 12(4):463,465,470.


34. England, C. B. 1970. Land capability: A hydrologic
    response  unit  in agricultural watersheds.  ARS
    41-172, U.S. Dept. Agr., Washington, D.C., 12 p.
42. Harrold, L. L., Triplett, G. B., Jr., and Youker, R.
    E. 1967a. Less soil and water loss from  no-tillage
    corn. Ohio  Agr. Res. Develop.  Ctr., Ohio  Rep.
    52(2):22-23.

43. Harrold, L. L., Triplett, G. B., Jr., and Youker, R.
    E. 1967b. Watershed test of no-tillage corn. Jour.
    Soil and Water Conserv. 22(3):98-100.

44. Harrold, L. L., Triplett, G. B., Jr., and Youker, R.
    E. 1970. No-tillage-characteristics of the system.
    Agr.Engin.51(3):128-131.

45. Harrold, L. L. and Edwards, W. M. 1972. A severe
    rainstorm test of no-till corn. Jour. Soil and Water
    Conserv. 27(l):30-35.

46. Hillel, D. 1971.  Soil and water. Physical principles
    and processes. Academic Press, New York,  288 p.
                                                                                                      25

-------
 47.  Hillel,  D.,  and Gardner, W. R.  1970. Transient
     infiltration  into  crust-topped profiles. Soil Sci.
     109:69-76.

 48.  Holtan,  H.  N.  1961. A concept for  infiltration
     estimates in watershed engineering. ARS 41-51, U.S.
     Dept. Agr., Washington, D.C., 25 p.
58. Jones, J. R. 1969. An objective method for estimat-
    ing runoff from small rural catchments. M.E. Thesis,
    Univ. of N.S.W., Sch. Civil Eng, 138 p. (cited by
    Boughton, W. C. 1970.) Effects of land management
    on quantity and quality of available water. Rep. No.
    120,  Water Res. Lab., Univ. New  South Wales,
    Manly Vale, N.S.W., Australia, 330 p.
 49.  Holtan, H. N., and Creitz, N. R. 1967. Influence of
     soils, vegetation and geomorphology on elements of
     the  flood  hydrograph.  I.A.S.H.  Publ. No.  85,
     2(3): 756-767.

 50.  Holtan, H. N. and Lopez, N. C. 1971. USDAHL-70
     model of watershed hydrology, ARS Tech. Bui. No.
     1435, U.S. Dept. Agr., Washington, D.C., 84 p.

 51.  Horton,  R.  E.  1919.  Rainfall interception.  Mon.
     Weather Rev. 47: 603-623.
 52.  Horton,  R.  E. 1931. The field, scope and status of
     the  science  of hydrology.  Trans. Amer. Geophys.
     Union 12:189-202.


 53.  Horton,  R.  E. 1940. An approach toward physical
     interpretation of infiltration capacity. Soil Sci. Soc.
     Amer. Proc. 5:399-417.

 54.  Huff, D. D.  and Kruger, P. 1967a. The chemical and
     physical  parameters in a hydrologic transport model
     for radioactive aerosols. Proc. Internatl. Hydrology
     Symp., Ft. Collins, Col., 1:128-135.


 55.  Huff, D. D., and  Kruger, P.  1967b. A numerical
     model for the hydrologic transport  of radioactive
     aerosol from precipitation to water supplies. Isotope
     Techniques in the Hydrologic Cycle, Am. Geophys.
     Union, Geophys. Monogr. Ser. No. 11: 85-96.


 56.  Jamison, V. C.,  Smith,  D. D., and Thornton, J. F.
     1968. Soil and  water  research on a claypan soil.
     ARS Tech. Bui. No. 1379, U.S. Dept. Agr., Washing-
     ton, D.C., 111 p.
57. Jensen, M.  E. 1966. Empirical methods of estimat-
    ing or predicting evapotranspiration using radiation.
    Evapotranspiration and its role  in water  resources
    management. Conf. Proc. Am. Soc.  Agr.  Engr. St.
    Joseph, Mich., December 5-6:49-53,64.
59. Konrad, J. C., and Cain, J. M. 1973. Hydrologic and
    watershed  modeling for  regulating  water quality.
    Trans. ASAE 16(3):580-581.

60. Langbein, W. B., Hains,  C. H.,  and Culler, R. C.
    1951. Hydrology  of stock-water   reservoirs  in
    Arizona. U. S. Geol. Surv. Circ. 110,18 p.

61. Larson, C. 1973. Hydrologic effects of modifying
    small watersheds-   Is prediction by  hydrologic
    modeling  possible?   Trans. ASAE l6(3):560-564,
    568.

62. Legrand,  H.  D.  1965. Patterns  of  contaminated
    zones of water in the  ground. Water Resour. Res.
    1(1): 83-95.

63. Leonard  R. E. 1967. Mathematical theory of inter-
    ception. In Sopper, W. E., and Lull, H.  R., eds.,
    Forest hydrology.   Pergamon Press, New York:
    131-136.

64. Linsley, R. K., Jr., Kohler, M. A., and Paulhus, J. L
    H.  1958.  Hydrology for  engineers.  McGraw-Hill,
    New York, 340 p.

65. Lull,  H.  W.  1964.  Ecological and silvicultural as-
    pects. In  V. T. Chow ed., Handbook of applied
    hydrology: 6-1,6-30. McGraw-Hill, New York.

66. Mannering, J. V., and  Burwell, R. E. 1968. Tillage
    methods  to reduce  runoff and erosion  in  the corn
    belt.  ARS Agr. Inf.  Bui. No. 330, U.S. Dept. Agr.,
    Washington, D.C., 14 p.

67. Mannering, J. V.  Meyer, L. D., and Johnson, C. B.
    1968. Effect  of cropping  intensity on erosion and
    infiltration. Agron. Jour. 60: 206-209.

68. McGuinness,  J. L., and  Harrold,   L.  L. 1971.
    Reforestation influences on small watershed stream-
    flow. Water Resour. Res. 7(4): 845-852.
26

-------
69. Mein, R.  G.,  and Larson, C. L 1973. Modeling
    infiltration during a steady rain. Water Resour. Res.
    9(2):384-394.

70. Mickelson, R.  H.  1966.  Level  pan system  for
    spreading  and  storing watershed runoff. Soil Sci.
    Soc. Amer. Proc. 30(3): 388-392.

71. Minshall, N. E., and Jamison, V. C. 1965. Interflow
    in claypan soils. Water Resour. Res. 1(3): 381-390.

72. Moldenhauer, W. C., Lovely, W. G., Swanson, N. P.,
    and Currence. H. D. 1971.  Effect  of row grades and
    tillage systems  on soil and water losses. Jour. Soil
    and Water  Conserv. 26(5): 193-195.
81. Parzen,  E. 1962. Stochastic processes. Holden-Day,
    Inc., San Francisco, 324 p.

82. Philip, J.  R.  1964. An  infiltration equation with
    physical significance. Soil Sci. 77: 153-157.

83. Philip, J.  R.  1969. Theory of infiltration. Advan.
    Hydrosci.  5. Academic Press, New York: 215-305.

84. Ramser, C. E.  1927. Runoff from small agricultural
    areas. Jour. Agr. Res. 34(9): 797-823.

85. Rauzi, F., and Hanson, C. L. 1966. Water intake and
    runoff as  affected  by intensity of grazing. Jour.
    Range Manag. 19: 351-356.
73. Moldenhauer, W. C., Wischmeier, W. H., and Parker,
    D. T. 1967. The influence of crop management on
    runoff, erosion,  and  soil properties of a Marshall
    silty clay  loam.  Soil  Sci. Soc.  Amer. Proc.  31(4):
    541-546.

74. Moore, W. L., and Morgan, C. W. eds.  1969. Effects
    of watershed changes on streamflow. Univ. Texas
    Press, Austin, 289 p.

75. Morel-Seytoux,  H. J.  1973. Two-phase  flows  in
    porous media. Advan. Hydrosci. 9, Academic Press,
    New York: 119-202.

76. Musgrave,  G. W., and Holtan, H. N. 1964. Infiltra-
    tion,  Chap.  12  in  V. T. Chow, ed. Handbook  of
    applied hydrology: 12-1, 12-30. McGraw-Hill,  New
    York.

77. Neff, E. L. 1973. Water storage capacity of contour
    furrows in Montana.  Jour.  Range Manag.  26(4):
    298-301.

78. Nelson,  R. W., and Eliason, J.  R.  1966. Prediction
    of water movement  through soils—a first step  in
    waste transport analysis. Proc.  21st Industr. Waste
    Conf., Purdue Univ.  Engr.  Ext. Ser. 121, Part  2:
    744-758.

79. Onstad, C. A. 1972. Soil and water losses as affected
    by tillage practices. Trans. ASAE 15(2): 287-289.

80. Onstad, C. A., and Olson, T. C.  1970. Water budget
    accounting on two corn cropped watersheds. Jour.
    Soil and Water Conserv. 25(4): 150-152.
86. Reddell, D. L., and Sunada, D. K. 1970. Numerical
    simulation  of dispersion in  ground water aquifers.
    Hydrol. Paper 41, Colorado State Univ., 79 p.

87. Rice, W. L., and Dragoun, F. J.  1965.  Effects on
    runoff volume from perennial prairie grass seeded
    on cultivated land. Jour. Soil and Water Conserv. 20
    (2): 63-64.

88. Richardson, C. W. 1972. Changes in water yield of
    small watersheds by  agricultural  practices. Trans.
    ASAE 15(3): 591-592.

89. Richardson, C.  W., and  Ritchie,  J. T.  1973. Soil
    water balance for small  watersheds. Trans. ASAE
    16(1): 72-77.

90. Ritchie, J.  T.,  and  Burnett,  E. 1971.  Dryland
    evaporative flux in  a subhumid climate:  II.  Plant
    influences. Agron. Jour. 63: 56-62.
91. Ritter, W.  F.  Johnson, H. P., Lovely, W. G., and
    Molnau,  M. 1974.  Atrazine, propachlor and dia-
    zinon  residues  in  small  agricultural  watersheds.
    Runoff losses, persistence and movement. Environ.
    Sci. Technol. 8(1): 38-42.


92. Robbins, J. W. D., and Kriz, G. J. 1969. Relation of
    agriculture   to  groundwater  pollution, a  review.
    Trans. ASAE 12(3): 397-403.

93. Rose, C. W. 1966.  Agricultural physics. Pergamon
    Press, New York, 230 p.
                                                                                                       27

-------
94. Saxton, K.  E., Johnson, H.  P., and Shaw, R. H.
     1974a.  Watershed evapotranspiration estimated by
    the  combination  method.  Trans.  ASAE  17(4):
    668-672.

95. Saxton, K.  E., Johnson, H.  P., and Shaw, R. H.
     1974b.  Modeling evapotranspiration and soil mois-
    ture. Trans. ASAE 17(4): 673-677.

96. Saxton, K. E., and Spomer, R. G. 1968. Effects of
    conservation  on  the  hydrology of loessial  water-
    sheds. Trans. ASAE 11(6): 848-849,853.

97. Saxton, K. E.,  Spomer, R. G.,  and  Kramer, L. A.
     1971. Hydrology and erosion  of loessial watersheds.
    ASCE Proc. Hydr. Div., 97 (HY11):1835-1851.

98. Saxton, K. E., and Whitaker, F. D. 1970. Hydrology
    of  a  clay  pan  watershed.  Univ.  of  Missouri-
    Columbia, College Agr. Expt. Sta. Res. Bui. 974,47
    P-

99. Schwab, G. O., and Fouss, J. L. 1967. Tile flow and
    surface  runoff from drainage systems with corn and
    grass cover. Trans. ASAE 10(4): 492-496.

100. Schwartz,  F.  W., and Domenico, P. A.  1973.
     Simulation of hydrochemical patterns in  regional
     groundwater   flow.  Water  Resour.   Res. 9(3):
     707-720.

101. Sharp, A. L., Gibbs, A. E., and  Owen, W. J. 1966.
     Development of a procedure  for  estimating the
     effects of land and watershed treatment on stream-
     now.  Tech. Bui.  1352, U.S. Dept. Agr., Washing-
     ton, D.C. 57 p.

102. Sharp, A. L., Gibbs, A. E., and Owen, W. J. 1968.
     A  history of the cooperative  water  yield proce-
     dures  study project. U.S. Dept. Agr. and  U. S.
     Dept. Int. Res. Rep. No. 403, 1077 p.

103. Smith, G. E., and Whitaker, F. D.  1974. Losses of
     fertilizers and pesticides from claypan soils. Report
     on Project R-801-666 prepared for Office Res. and
     Monitoring, U.S. Environ. Protec. Agcy. 75 p.

104. Smith, R.  E.  1972.  The  infiltration envelope:
     Results  from  a  theoretical  infiltrometer.  Jour.
     Hydrol. 17(1/2): 1-21.
105. Soil Conservation Service. 1972. National engineer-
     ing handbook. Section 4. Hydrology. U.  S. Govt.
     Printing Off., Washington, D. C.

106. Spomer, R. G., Heinemann, H. G., and Piest, R. F.
     1971. Consequences of historic rainfall on western
     Iowa farmland. Water Resour. Res. 7(3): 524-535.

107. Streeter, H. W., and Phelps, E. B. 1925. A study of
     the pollution and natural purification of the Ohio
     River. U. S. Pub. Health Bui. No. 146.

108. Sutton, 0.  G.  1953. Micrometeorology. McGraw-
     Hill, New York, 333 p.

109. Swartzendruber, D., Skaggs,  R. W., and Wiersma,
     D. 1968.  Characterization of the  rate  of water
     infiltration  into soil. Tech. Rep.  5, Purdue Univ.
     Water Resour. Ctr., 120 p.

110. Tennessee Valley  Authority. 1972. Upper Bear
     Creek  Experimental Project. A continuous daily-
     streamflow model. Res. Paper No.  8. Knoxville,
     Tenn., 99 p.

111. Thomas, A. W., Carter, R. L, and Carreker, J. R.
     1968. Soil, water  and nutrient losses from Tifton
     loamy sand. Trans. ASAE 11(5): 677-679, 682.

112. Thornthwaite, C. E. 1948. An approach towards a
     rational classification of climate.  Geogr. Rev. 38:
     55-94.

113. Thornthwaite,  C.  E., and Hare,  F.  K. 1965.  The
     loss of water to the air. In  Agricultural meteor-
     ology.  Meteorological Monographs,  Amer. Mete-
     orol. Soc., Boston, Mass.6(28): 163-180.

114. Wischmeier, W. H.  1966. Relation of  field-plot
     runoff to  management  and physical factors.  Soil
     Sci. Soc. Amer. Proc. 30(2): 272-277.

115. Wisler, C. O., and Brater, E.  F. 1959. Hydrology.
     John Wiley and Sons, New York, 408 p.

116. Woolhiser, D. A. 1973.  Hydrologic and watershed
     modeling   state of the  art.  Trans. ASAE  16(3):
     553-559.

117. Woolhiser, D. A. 1975. The watershed approach to
     understanding  our environment. Jour. Environ.
     Qual.4(l):  17-21.
28

-------
118. Woolhiser, D.  A. 1975. Simulation of unsteady    119. Zinke,  P. J. 1967. Forest interception studies in
     overland flow. Chap. 12 in K. Mahmood and V.         the United States. In W. E. Sopper and H. R. Lull,
     Yevjevich, eds. Unsteady flow hi open channels,         eds., Forest  hydrology.  Pergamon  Press, New
     Vol. II. Water  Resour. Publications, Fort Collins,         York: 137-161.
     Col.: 485-508.
                                                                                                     29

-------
                                             CHAPTER 3

                          CROPLAND EROSION AND SEDIMENTATION

                                            W. H. Wischmeier
   Erosion is the wearing away of the land surface by
water, wind, ice, or other geological agents. Sediment is
defined as solid material, both mineral and organic, that
has been moved from its original source by these agents
and is  being transported or has come  to rest on the
earth's surface (66). Sediment impairs the quality of the
water resources in  which it  is  entrained and  often
degrades the location where it is deposited. It may carry
pesticides, toxic metals, and plant nutrients absorbed on
the soil particles (25, 69).
   This  chapter  documents technical background and
methodology for estimating and controlling  cropland
sediment production. It supplements the material given
in Volume I, Sections 3.3 and 4.1. Only  sediment from
cropland erosion by water was  considered pertinent to
the purposes of this  manual,  but literature  on wind
erosion control is cited (12, 60, 61, 99,  100). Observed
quantities of sediment from geological erosion and from
nonagricultural sources are cited to help portray crop-
land  sediment in  its  proper perspective,  and land
classifications  pertinent  to large-area appraisals of crop-
land sediment  potential are reviewed. Brief overviews of
(a) existing erosion research data, (b) the  mechanics of
the soil-erosion process,  and (c) progressive improve-
ments in prediction equations, provide pertinent back-
ground information for erosion-control technology. The
Universal  Soil  Loss Equation, soil loss tolerances, and
sediment delivery ratios are reviewed  as potential tools
for pollution-control planning. The major emphasis is on
discussions of erosion factors and important features of
erosion-control practices.
                             SEDIMENT SOURCES AND QUANTITIES
   Sediment concentrations  in rivers of  the United
States range from  200 to 50,000 ppm, with an  occa-
sional concentration as high  as 600,000 ppm (21). The
amount  of sediment moved by flowing water has been
reported to average at least 4  billion tons  a year, with
about one billion  tons reaching  major streams  (19).
Estimates  ascribe about  30%  of this country's  total
sediment to geological erosion  and about half of it  to
erqsion of agricultural lands (77).

                Geological Erosion

   The erosion that occurs under natural environmental
conditions  of  climate and vegetation, undisturbed by
man, is called geological, natural, or normal erosion (66).
Estimates  of annual rates of geologic deposition in the
United States range from less than 0.30 to  0.74 ton per
acre  (38, 65).  Even at such relatively low rates, a large
drainage area will produce large quantities of sediment.
The Missouri River's name attests to the turbidity of its
waters before it was discovered by Europeans. The rate
of erosion under natural vegetation reaches a maximum
where  the  mean annual rainfall is between 10  and 15
inches. Under  higher rainfall rates, improved vegetation
inhibits erosion;  under rates  of less than  10 inches
sediment-entraining runoff becomes  more  rare  (29).
Natural erosion over long geologic periods can be  quite
dramatic, as evidenced by the wearing away  of moun-
tains and building up of flood plains.
   The more rapid erosion  that is primarily a result of
activities of man  is called accelerated  erosion  (66).
Sediment produced by accelerated erosion comes from
many sources.

             Nonagricultural Sources

   Some major nonagricultural sources of sediment are:
erosion from construction  activities,  roadside erosion,
stream channel and streambank erosion, scouring of
flood-plain land by floodflow,  mining and industrial
wastes dumped into streams or left in positions suscepti-
ble to erosion, and mass wasting from landslides.
   In  some watersheds,  the  sediment  that  originates
from these sources may far exceed that from cropland.
                                                                                                       31

-------
A  1969 report by the Secretary of Agriculture and the
Office of Science and Technology  (19) gave the follow-
ing statistics:  During road construction in Scott Run
Watershed,  Fairfax County,  Virginia, sediment at the
rate  of about  140 tons per  acre  was produced at the
source, and  about half of this amount was measured at a
downstream gaging station. Erosion losses at rates of 42
to 289 tons per acre per year were  measured on bare
roadside  cuts  near Carterville, Georgia, and comparable
rates were measured on 35 road cuts in  the Baltimore
area. As much  as 2,000 cubic  yards of sediment per
square  mile  of access road  has been  measured in
mountainous country. Sediment  from construction  ac-
tivities in urbanizing areas  near Lake  Bareroft,  Virginia,
was  reported  equivalent to 39 tons  per acre annually.
Studies in southeastern Kentucky showed that sediment
yields  from strip-mined coal  land can be 1,000 times
that from forested land; there are about 2.3 million acres
of strip-mined lands in the United States. Erosion is a
serious  problem  on  at least 300,000 miles of stream-
bank.


                Cropland Sediment

   Cropland  does not produce the greatest amount of
sediment per unit of area, but because of the large area
involved,  our  437 million  acres of cropland as a whole
produce  more sediment  than  any other source. Annual
soil  loss from  cropland ranges from  about one ton to
more than  100 tons per acre, depending on  the crop
system, management  practices, rainfall, soil characteris-
tics, and  topographic features.  A  1967 Conservation
Needs  Inventory by  the USDA  (75, 77) showed that
about half of our country's cropland averages between 3
and  8  tons  of soil loss per acre per year, 30% averages
less  than  3  tons, and 20% averages more than 8 tons.
Individual states have published the  adjusted inventory
data, and the reports are available from state offices of
the Soil Conservation  Service, USDA (75).
         Large-Area Estimates of Cropland
                Sediment Hazards

   In 1940, Baver (7) listed the major erosion factors as
climate,  topography, vegetation, soils,  and the human
factor. The principal influence of climate is the  type,
amount,  and temporal distribution of the rainfall. The
human factor includes such items as crop sequence, soil
and crop management, and conservation practices. Each
of  these  factors  often  varies widely  within  a single
watershed or land resource area. All except climate often
vary appreciably even among different fields on a single
farm. Therefore,  soil-loss estimation  and control plan-
ning are  most effective on a local basis, by procedures
given in Volume I.
   On  a  large-area  basis, the cropland contribution to
sediment in streamflow is influenced by: the amount of
sediment produced on the cropland (gross erosion), the
density of cropland in the drainage area, and the portion
of the eroded  soil that actually reaches a continuous
stream system (sediment delivery ratio).

Gross Erosion

   The erosion potential on a relatively homogeneous
drainage  area can be estimated  by using representative
soil, cover, and  topographic features to evaluate  the
factors in the Universal Soil Loss Equation. Published
maps  and  standard  land classifications also provide
helpful information  for appraisals of cropland sediment
hazard on a large-area basis.
      The  map given  in Volume I as Figure 9 shows
relative  potential  contributions  of   cropland  in  the
conterminous United  States by major land resource
areas.
   Soil survey maps are  the best  sources of information
on soil characteristics and associated land features. These
maps generally include classifications of erosion and land
slope.  The mapped  erosion class is primarily  an indica-
tion of the  extent of prior erosion; quantitative  erosion
rates are not mapped because of their local nature.  The
slope  class  indicates whether the land is nearly level,
gently sloping,  moderately  sloping,  strongly sloping,
steep,  or very steep, but it does not provide information
on the slope shapes and lengths.
   Land  resource units are  geographic areas of land,
usually  several  thousand acres  in   extent,  that  are
characterized by particular patterns  of soil (including
slope  and erosion), climate, water resources,  land  use,
and  type of farming (73).
   Major  land  resource  areas consist  of geographically
associated  land  resource units.  The  156  major land
resource   areas of the  48   conterminous states  were
selected  as  the  basis  for  mapping  hydrologic  and
erosion-potential data in Volume I. Major characteristics
of the 156  individual  areas are given in Agriculture
Handbook No. 296 (73).
   Capability classes (27) are interpretive soil  groupings
made primarily for agricultural purposes. The  classifica-
tion  begins with  the individual  soil   mapping unit.  A
capability unit  is a grouping of soils  that are suited to
32

-------
the same  kinds of cultivated crops  and pasture  plants
and  have  about the  same  responses  to systems  of
management.  A  capability  subclass  is a  grouping  of
capability units having similar kinds of limitations or
hazards.  Four  kinds  of  limitations or  hazards are
recognized:  erosion, wetness, root-zone limitation, and
climate. In the broadest category of capability classifica-
tion all the soils are grouped in eight classes.
   The eight capability classes were briefly described in
Chapter  1.  A more detailed description is given  in
Appendix II of USDA Statistical Bulletin No. 461  (77).
Generally, erosion hazards increase as the capability-class
number increases (except Class V), but it is important to
recognize that  for some  areas the higher classifications
are due to  wetness, root-zone limitations, or  climatic
limitations  rather than erosion. The  class and  subclass
designations, together,  provide information about  both
the degree and the kind of limitation.
Cropland Density
   Acreage data for land in each subclass of each of the
eight  capability classes,  by states  and several land-use
classifications can be  obtained  from  the Conservation
Needs Inventory (75). These  data were used in  the
development of Figures 6  through 9,  Volume I. Wind-
erosion limitations  were  included in  the  capability
subclass data used  for Figure 7. The  other  tables and
charts in the erosion sections of Volume I are for water
erosion only.

Sediment Delivery Ratio

   This is  the  factor that  adjusts the  gross sediment
estimate to compensate for deposition  along the path
traveled by  the runoff as it moves from a field slope to a
continuous  stream  system. The  delivery ratio  will be
discussed in more detail at the last of this chapter.
                                         CROPLAND EROSION
              Erosion Research Data
   Measurements of runoff and soil loss from field plots
in the United States began about 1917, in Missouri (64).
Between  1929  and   1933  the U.  S.  Department  of
Agriculture  established ten  Federal-State  erosion  re-
search   stations, in  regions  where  the problem had
become most critical. In the next 25 years, erosion plot
studies  were established  at 32 more locations. Precise
measurements of  precipitation, runoff, soil  loss, and
related  field conditions at  the  42  stations in  23 states
were continuous for periods of 5  to 30 years (85). In
1960, studies were underway on 18 soils. Fundamental
studies  of erosion mechanics were conducted concur-
rently and have  received increased  emphasis since about
1960.
   In 1954, the Agricultural Research Service established
a  national runoff and soil-loss data  center at  Purdue
University.  The basic data  from more than  10,000
plot-years of erosion studies at 42 research stations were
assembled,  standarized in  units,   and  transferred  to
punched cards for summarization  and overall  statistical
analyses (79). Data from  continuing studies were added
annually for analysis with the previously assembled data.
   The  plot studies  and   fundamental  investigations
identified the  major erosion  factors and  provided a
wealth of information on erosion mechanics and control.
Inherent limitations of the  plot data will be pointed out
in the discussion of soil loss equations.
   Field-plot rainfall simulators are now used to expe-
dite filling voids in existing plot data and field  testing of
new erosion control concepts and practices. This equip-
ment can simulate the drop sizes and terminal velocities
of natural rain at common intensities, apply simulated
rainfall on  several 75-foot  plots simultaneously, and
apply  identical storms to plots on physically separated
soils and topographies (43).

                The Erosion Process

   Soil erosion is a process of detachment and transpor-
tation  of soil materials  by erosive agents (16).  It is  a
mechanical process that requires energy. Much of this
energy is supplied by falling  raindrops. The dead weight
of  the  water falling in  30  minutes  of a  Midwest
thunderstorm  may  exceed  100  tons  per  acre.  The
billions of drops which comprise this  100-ton volume of
water  strike  the  soil,  if unprotected, at an average
velocity of nearly 20 miles an  hour. The impact energy
during the 30 minutes may exceed 1,000 foot-tons per
acre (93).
   When raindrops strike bare  soil at  a high  velocity,
they shatter soil granules and clods and  detach particles
from the  soil  mass.  Splash action and shallow overland
flow  transport some  of the  detached particles directly
down  the slope and others  to implement  marks and
other  small  channels,  where the more concentrated
runoff  provides  transportation  for  them.  This soil
movement is called sheet erosion (6), or interrill erosion
(40). This type of erosion occurs rather uniformly over
the  slope  and may  go unnoticed until much of the
productive topsoil has been  removed. In sheet erosion.
                                                                                                          33

-------
nearly all of the soil-particle detachment is by raindrop
impact (32, 36,  101).
   The erosive potential of flowing water depends on its
velocity, depth,  turbulence, and  type and  amount of
material  it transports (77). Water moving down the slope
follows the path of least resistance and concentrates in
tillage marks, eroded flow channels, and depressions in
the  natural land surface,  where  it gains in depth  and
velocity. Erosion in these flow concentrations is directly
related to the hydraulics of the concentrated flow (40).
The concentrated  runoff may remove enough  soil to
form small  but  well defined channels, or rills. Rills are
often the first readily apparent evidence of erosion, but
tillage usually obliterates them.
   Rill erosion has  been defined as an erosion process in
which numerous channels only several inches deep are
formed  (66),  and  as   the erosion  occurring  in flow
channels (40). In rill erosion, soil  particles are detached
by the shearing action of water flowing over the soil
surface and by slumping of undercut side walls and small
headcuts. The detached particles are transported by a
combination of rolling, saltation,  and suspension. Parti-
cles transported by suspension may travel long distances
before being deposited  on the  land surface. The capabil-
ity of runoff to detach soil material is proportional to
the sheer stress  raised to a power  of approximately two
(17). Consequently, rill  erosion  increases rapidly as
steeper or  longer  slopes  increase  runoff flow depth.
Under continued rainfall, sheet erosion  continues be-
tween the rills.  Field soil losses are usually a combina-
tion of sheet and rill erosion, and  their relative contribu-
tions to  total  soil loss differ with soils and  surface
conditions.
   When water accumulates in narrow channels and,  over
short periods, removes  the soil from this narrow area to
depths of 1 to 2 feet, or more, the process is called gully
erosion (66).  Cully erosion produces large amounts of
sediment but can usually be prevented on cropland.
   A soil's inherent ability to resist erosion by  rainfall
and  runoff depends   on  its  physical  and  chemical
properties. Erosion control is accomplished by reducing
the  mechanical  forces  of the water acting  on the soil
particles or by increasing the soil's resistivity to erosion,
or both.

                Soil Loss Equations
soil  properties,  topographic  features,  and  numerous
management  details  occurred  at different  levels and in
different combinations in the various studies.
   Plot  data  predict specific-field  soil losses only if the
influence of  each of the major contributing parameters
can be  isolated and evaluated relative  to  the level at
which the parameter was present  in  the study, so that
the various  influences can  be  combined  in different
proportions  to simulate other situations.  However, ef-
fects of rainfall characteristics and  soil properties cannot
be isolated in a one-location  study, where rainfall and
soil are  either constant for the  plot series or  vary in
unison.  Also, many  relevant secondary variables cannot
be  controlled  in plot studies.  Some  of  these  vary
randomly  over time.  Some  differ with  seasons, and
others', such  as rainfall distribution and storm character-
istics, show  long-term trends at  a given  location  but
fluctuate  unpredictably for  short  time  periods. The
uncontrolled variables  interact with controlled variables,
and these interactions  can substantially bias brief-period
research  results.  Assembling  all   the  available  erosion
research   data  at  one location  for  overall statistical
analyses  (79) counteracted many of these limitations. It
enabled combining basic data from various locations in
analysis designs capable of providing in formation on the
major factor effects individually  and on  some  of the
most important interaction effects. It also helped mini-
mize bias of results by  random variables.
   Mathematical  relationships were derived whose basic
and  theoretical  validity   has been   substantiated by
subsequent  fundamental  research. When  these  factor
relationships  are  combined in a general soil  loss equa-
tion, planners  can determine what the average annual
soil loss  rate  and  the potential soil loss reductions from
various  alternative crop and  management  systems are
likely to be at specific  locations other than  that of a plot
study.
   The  most accurate soil loss equation that  is  now
field-operational  is  the  Universal  Soil  Loss Equation.
This equation has  been  used as  an  erosion-control
planning tool for more than  a decade in  the 37 states
east of the Rocky Mountains  and  is now used to a more
limited  extent also in the Western States,  Hawaii, and
several  foreign countries. However, the following  brief
overviews of four soil  loss equations are pertinent to the
subsequent discussion of erosion factors.
   The  literature of  the  past  40 years includes many
reports  of  local erosion  studies. These  reports may
appear to a  casual reader as inconsistent, and sometimes
incompatible, because  of wide  differences in  the  re-
ported results.  However, most of these differences can
be accounted for by  the  fact  that the rainfall pattern,
The Slope-Practices Equation

   This initial soil loss equation was developed gradually
in the early  I940's. Zingg (102) developed factors for
the effects of length and steepness of slope. Smith (62)
added crop and  conservation  practice factors and the
34

-------
concept  of a  limiting annual soil  loss. Browning and
coworkers  (10) proposed  soil-erodibility and  manage-
ment factors for Iowa, but their work was not published
until 1947.  With the cooperation of program leaders in
the  North  Central  Region  of the  Soil Conservation
Service, these initial developments were combined in the
Slope-Practice   Equation  for  use  throughout the  Corn
Belt.
   This equation used several dimensionless factors to
adjust  an initial  basic soil loss to  specific  field condi-
tions. Its basic  soil loss was the average annual loss from
corn-oats-meadow rotations  on research plots in  the
North  Central  States.  Factors  for  other crop systems
were estimated relative to this rotation. The equation
had no rainfall factor, and its soil factor was expressed
relative to 1.0 for Marshall silty clay loam. Zingg's slope
length  and steepness exponents (0.6 and 1.4) were used
to  adjust  the  soil-loss  computations  to  field  slope
dimensions.

The Musgrave Equation

   In 1946, a national committee, with G. W. Musgrave
as chairman, was assembled in Ohio  to reappraise  the
factors in the Slope-Practice  Equation and add a rainfall
factor.  The  modified model  became  known as  the
Musgrave Equation  (48).  A graphical solution of  the
equation was  published  in 1952 for the Northeastern
States  (35).
   The 1.75 power of the 2-year, 30-minute rainfall was
adopted as the rainfall factor,  and Zingg's  slope-length
and  percent-slope exponents were lowered  to 0.35 and
1.35, respectively. Annual cover factors were estimated
relative to a value of 1.0 for either continuous fallow or
continuous  rowcrop.  A  quantitative  soil  factor was
derived by  adjusting  annual soil  losses for effects  of
rainfall,  slope  and cover.  Subsequent research did  not
confirm the adequacy  of 2-year, 30-minute rainfall as an
index  of local differences  in  rainfall erosivity. The
lowered  slope-length  factor was compatible with  some
early sets of data but too low for others. Numerous plot
studies showed that continuous fallow and continuous
rowcrop are not interchangeable  and  that the  cover
effect of continuous rowcrops is highly variable.
   The Musgrave Equation  has been  widely used  for
estimating gross erosion from large heterogeneous water-
sheds.  Its highly generalized factor values are more  easily
assigned to  broad areas than are  factors based  on  more
specific descriptions of the erosion-influencing parame-
ters. However,  erosion hazards are highly localized. For
resource-conservation  and  pollution-control  planning,
soil  loss equations  need  to  reflect local conditions as
accurately as possible.
The Universal Soil Loss Equation (USLE)

   The  Universal  Soil  Loss  Equation  (80,  94,  95),
developed in 1958, overcame many of the deficiencies of
its  predecessors.   Its form is similar to  that  of the
Musgrave  Equation,  but the concepts, relationships and
procedures underlying the definitions and evaluations of
the erosion factors are  distinctly different (see section
on  Erosion  Factors).  The major improvements  (84)
included:
   1.  More complete separation of factor effects so that
      results of a change  in the level of one or several
      factors can be more accurately  predicted.
   2.  An erosion index that provides a good  estimate of
      the erosive potential of rainfall and  its associated
      runoff.
   3.  A quantitative soil-erodibility factor that is evalu-
      ated  directly  from  research data without refer-
      ence  to any common benchmark.
   4.  An equation and nomograph capable  of comput-
      ing the credibility factor for numerous soils  from
      soil-survey data.
   5.  A  method  of including effects  of interactions
      between cropping and  management parameters.
   6.  A method of incorporating effects of local rainfall
      pattern and specific crop cultural  conditions in
      the cover and management factor.
   The  Universal  Soil Loss Equation  computes average
annual  soil loss  as the product of two quantitative
factors  (soil-erodibility  and rainfall-erosivity) and four
qualitative factors (96).  The equation  is:

                   A=RKLSCP

where A is the average soil loss, in tons per acre, for the
time period used  for factor R (usually average annual).
R    is the rainfall and  runoff erosivity index.
K,    the soil erodibility factor, is the average soil loss in
     tons per  acre per unit of R, for a given soil on a
     "unit  plot"  which is defined  as  72.6 feet  long,
     with 9% slope, continuously fallowed, and  tilled
     parallel to the land slope.
L,   the  slope-length factor,  is the  ratio of soil loss
     from  a given length of slope to  that from  a
     72.6-foot length with all other conditions identi-
     cal.
S,   the  slope-steepness factor, is the ratio  of soil loss
     from a given percent-slope to that from a 9%  slope
     with all other conditions identical.  (In practice,
     factors L and  S are  usually combined in a single
     topographic  factor denoted by LS.)
C,   the  cover and management  factor, is the ratio of
     the  soil loss with specified cover and  agronomic
                                                                                                           35

-------
      practices to that from the fallow condition on
      which factor K is evaluated.
P,    the practice factor, is the  ratio of soil loss with
      supporting  practices  such as contouring or strip-
      cropping to that with straight-row  farming up and
      down the slope.
   The concepts and relationships underlying the evalua-
tions of these factors are reviewed in the discussion of
Erosion Factors.

Basic Erosion Models

   Basic mathematical models are being developed  that
combine fundamental principles, concepts and relation-
ships of erosion  mechanics, hydrology, hydraulics, soil
science,  and meteorology  to simulate the erosion  and
sedimentation processes. Substantial progress has been
made  in developing  static  and dynamic models capable
of predicting  spatial and temporal variations in erosion
and  sedimentation (14, 17, 40, 52). To the extent  that
these  simulation  models reflect  direct and  interacting
effects of  more  of the uncontrolled  and  secondary
variables, they will enhance analyses of erosion systems
and  control practices. These models have not become
field operational because additional research is needed to
bridge certain information gaps.  However, they have
already improved the understanding of erosion proc-
esses, helped explain some of the seeming inconsistencies
in the field-plot data, and improved the  accuracy  of
some of the factor evaluations for the USLE.
   The initial  basic models have added several important
new concepts. One is the treatment of soil detachment
by rainfall, detachment by runoff,  and transport by
runoff, as individual subprocesses that bear substantially
different relationships to the erosion factors and that
occur in widely differing combinations (39, 44). Either
detachment capacity  or transport capacity  can limit
erosion at a  given site.  Another new concept is  the
separation of rill erosion from interrill erosion (17, 40).
This distinction  will help clarify unexplained differences
in the credibilities of soils and effectiveness  of crop
canopies. Some soils allow very substantial sheet erosion
without  rilling; others  are much more susceptible  to
rilling.
                                          EROSION FACTORS
   The  climatic, soil, topographic, and management
parameters that largely  determine erosion rates have
wide ranges of possible values, or levels, that can occur
in  any  of an extremely  large  number of  possible
combinations. The six major erosion factors discussed in
this section estimate the effects of different levels of
these parameters on soil erosion by water. In a soil loss
equation, each factor must be represented by a number
that reflects  the specific local conditions, and all the
numbers must  be relative to the same, clearly defined,
benchmarks.  The  benchmark  conditions for the Uni-
versal Soil Loss Equation are free of geographic bounds
and are defined as follows.
   The benchmark management condition is continuous
fallow  that receives  primary and secondary tillage each
spring  and is  periodically tilled during the summer to
prevent vegetation   and  serious  crusting. The tillage
operations are  up and down the slope. This condition
was selected because: (a) continuous fallow is the only
condition under which soil effect  could be evaluated
independently  of cover,  management, and residual ef-
fects, and (b) it is a more constant condition than would
exist with any type of cropping. The fact  that this
condition rarely exists in practice is immaterial because
the soil loss computed by the  equation as a whole does
reflect existing field conditions.
   The  slope  length of 72.6 feet was  selected as  a
benchmark because most of the erosion research plots
since  1930 were of this length.  It is sufficient  for
measurement of runoff effect as well as raindrop-impact
effect. Slope steepness of 9% was the most representa-
tive  for the existing plot data. Straight-row fanning up
and  down the  slope represents  complete  absence of
support practices. The "unit plot" on which the quanti-
tative  soil factor is measured has  these  benchmark
conditions and factors L, S, C, and P have values of  1.0.
   The values of factors R, K, L, and S are essentially
firm for a particular location and, together, determine
the  location's  characteristic  erosion  potential.  The
farmer or planner has no control over rainfall pattern or
steepness  of the slope. The effective slope length can be
reduced  by  use of  terraces or  diversions,  but   this
reduction can be classified as a practice effect. Manage-
ment systems that gradually improve  soil structure  and
increase its organic-matter content can  affect  its credi-
bility,  but an appreciable change in the soil factor would
require many years. Factors C and P, on the other hand,
are highly responsive to executed management decisions.
Good management and erosion control practices reduce
sediment  production primarily through  their effects on
these two factors. The following discussions of the six
major erosion factors include the concepts and relation-
ships underlying their definition and evaluation for  the
USLE.
36

-------
    Rainfall and Runoff Erosivity (Factor R)

   Most cropland erosion by water is directly associated
with  rain events and is  influenced both by the  rain
intensities and  by the amount and rate  of runoff. The
function of factor  R  is to quantify these interrelated
erosive forces. The parameter used to evaluate R must be
predictable on  a probability  basis from  meteorological
data.  It  must be definable for specific storms and for
specific time periods other than annual, and its seasonal
or annual evaluation must be influenced by all signifi-
cant rains rather than only by annual maxima.


The Rainfall-Erosion Index, El

   The assembled plot data showed that when all factors
other than rainfall are  constant, storm  soil losses from a
cultivated field are  directly  proportional to an inter-
action term, which is the product of the rainfall energy
and the maximum 30-minute  intensity. This product is
the El parameter (80, 93). The relation of soil loss to El
is linear; therefore, individual-storm values of El can be
summed  to  obtain seasonal   or annual  values  of the
parameter. Frequency  distributions of annual, seasonal,
or annual-maximum-storm El values  follow the  log-
normal type of curve  that is typical of many hydrologic
data(S0).
   Median  raindrop  size  increases as  rain  intensity
increases, to  about 3  in/hr,  and terminal velocities of
free-falling waterdrops  increase with increased drop size
(22,  33). Since the  kinetic energy of a given mass in
motion is  proportional  to velocity  squared, rainfall
energy is directly related  to  rain intensity.  Analyzing
published   dropsize   and   terminal-velocity   data,
Wischmeier and Smith (93) derived the equation E = 916
+ 331 logioi, where E is the kinetic energy in foot-tons
per acre-inch of rain,  and i  is intensity in inches per
hour.  The energy of a  rainstorm can be computed from
recording-raingage data.  The  storm  is  divided into
successive increments  of essentially uniform intensity,
and a rainfall energy-intensity table (93) derived from
the above formula is used to compute the energy of each
increment. Thus, the energy of a rainstorm is a function
of all  its component intensities and rain amount.
   In  exploratory analyses of data from bare fallow
plots, rainfall energy  was  the best  single predictor of
associated runoff, but  was not a good predictor of soil
loss. For sheet  erosion, soil detachment is primarily by
raindrop impact on the surface, but the capacity of the
associated runoff to detach and transport soil material is
directly related to  its depth  and  velocity.  These are
directly related  to the  maximum prolonged intensity of
the storm.  Therefore, the erosive potential  of a rain-
storm is  a function of  its kinetic  energy,  maximum
prolonged  intensity,  and their interaction, all three of
which are reflected in the El parameter.
   The published rainfall energy-intensity table (84, 93)
applied the equation  given above to intensities up to 10
in/hr. Two recent studies showed that median drop size
does  not continue to increase when intensities  exceed
about 3  in/hr (11, 26).  Therefore, the energy given in
the table for a 3 in/hr intensity should be used for all
higher intensities as well. This change does not signifi-
cantly affect El computation  in  the United States
because prolonged intensities greater than 3 in/hr are too
rare to have much effect on average annual El values.
   For computation of average annual El values, contin-
uous  records of from 20 to 22 years are desirable in
order to  avoid  bias by  cyclical variations  in rainfall
pattern (49). Erosion index values were computed for
about 2,000 locations fairly uniformly distributed over
the 37 states east of the Rocky Mountains. By interpo-
lating between the computed point values, lines of equal
value (iso-erodents) were plotted on a map that included
county lines as references (82, 96). The mapped values
represent 22-year rainfall records (1937-1958). At sta-
tions  where 40-year records were available, the 40-year
average annual rain amounts generally coincided very
closely with the corresponding averages for the 22 years
used in development of the iso-erodent map.
   The computed annual  El values are reasonably well
correlated  with  the  2-yr,  6-hr  rainfall  probabilities
published by the Weather Bureau (74). The relationship
is  expressed by  El = 27.38P2-17, where P = the  2-yr,
6-hr rainfall (87). The El values given in Figure lOa,
Volume I, for the 11 Western States  were estimated by
this equation. Those for the  other 37 states were taken
from the original iso-erodent map (#2).
   Factor R in the USLE usually equals the pertinent El
value. For prediction of average annual soil loss, it is the
annual-El value  available from  Figure  lOa; for short
specific time periods, it  is the actual local El for that
period. However, there are two conditions for which the
computed El must be modified to evaluate factor R.
   1.  Where snowmelt  runoff on  moderate to steep
      slopes is significant, the El value must be adjusted
      upward to add the erosive effects of this runoff to
      the R value. The Palouse Region of the Northwest
      exemplifies this condition. Numerical  evaluation
      of the erosivity of runoff that is not an immediate
      consequence  of rainfall  is an  area of  needed
       research.  Only tentative estimates of the adjust-
       ment factor for the Palouse Region are presently
      available.
                                                                                                          37

-------
  2.  Experience has shown that on the Coastal Plains
      of the Southeast, factor  R  is less  than  the El
      values computed by  the standard procedure. This
      discrepancy may be due  to  the  combination of
      hurricane-associated  storms  and  flat slopes.  The
      hurricane  storms compute very  high  El  values,
      but the gentle slopes are soon largely covered by
      very slowly moving runoff that shields the soil
      surface from raindrop impact. In  a study on a
      similar soil  and slope in the Maumee Basin of
      Northeastern Indiana, using  a rainfall  simulator
      and inflow  at  the  upper end of the plot, drop
      impact on the  soil  surface was needed  to obtain
      significant soil  loss from  a 35-foot  plot (36). A
      maximum of 350 for El  values in the Southeast
      was  recently adopted as a  temporary measure
      until  research can provide "effective El"  values
      for these conditions.


 Runoff

   Surface  runoff is  not  a   separate factor in  the
 Universal Soil Loss Equation or its predecessors because:
 (a) no satisfactory prediction  equation  for cropland
 runoff existed,  and (b) the respective roles  of rainfall
 and runoff in the erosion process had not been separated
 in erosion research. Runoff data  alone do not predict
 soil loss. The sediment content of an acre-inch of runoff
 can range from  a mere trace to  many tons. For soil loss
 prediction, the  factors in  the USLE would need to be
 combined with the runoff factor, and the  runoff would
 first need to be  predicted as a function of essentially the
 same parameters. Therefore, it was advantageous to
 relate the factors directly to soil loss in an equation for
 widespread  field use. The El parameter combines esti-
 mates of runoff amount and rate with the potential of
 the rainfall  to detach soil material by drop impact and
 splash action.
   Researchers   have  recently  made good progress in
 separating rainfall-induced (interrill) erosion  from run-
 off-induced (rill) erosion (17, 40). With this separation, a
 runoff factor added to the soil loss equation should have
 substantial potential  for improved  accuracy. An equa-
 tion that predicts the two types of erosion as separate
 components of the total soil loss could largely solve the
 aforementioned  problems  with  factor R in the  North-
 west and Southeast. Also, some erosion-control practices
greatly reduce  soil loss  without appreciable  effect on
 runoff. Onstad  and  Foster (52) obtained good  results
from adding a runoff factor to the USLE when using the
equation to route sediment through a watershed. How-
ever, more research is  needed  to  make  this approach
 field operational.
            Soil Erodibility (Factor K)

   The susceptibility of a given land area to erosion is a
function of all the factors in the soil loss equation, but
some soils erode more readily  than others even when
rainfall, topography, cover and  management are identi-
cal. Soil credibility refers to a  soil's inherent suscepti-
bility to erosion by rainfall and runoff. This is a function
of complex interactions of soil physical and chemical
properties. Numerous researchers have measured differ-
ences in the credibilities of a few soils, and some have
related  credibility to specific soil properties (5, 10, 34,
46, 50,  51, 55,  76,  91). Water infiltration into soils was
reviewed by Pan and Bertrand (54).
   The  relation of soil  loss to El  is linear, and the
average  increase in soil loss for each additional unit of El
differs for different soils (93). The average  soil loss per
unit of  El, measured under the previously defined "unit
plot" conditions, is the numerical soil-erodibility factor
of the USLE. For 23 benchmark soils for which K was
measured in long-term plot studies under natural rain, its
value ranged from 0.03 to 0.69 (50, 96).
   Rainfall simulators were  used in  the Corn Belt, the
Southeastern States, and Hawaii to  evaluate other soils
and obtain soil loss data for study of the relationships of
various  soil properties to credibility (5, 89, 91). In a
Corn Belt study of about 60 soils selected to include a
broad range in soil properties,  24  primary and inter-
action terms  were  statistically   significant  in multiple
regression analysis of the data (91). This illustrates the
complexity  of the problem, but for practical purposes
many of these  terms can be neglected either because of
relatively small effect or because they are closely related
to particle-size distribution, organic-matter content, soil
structure or permeability.
   Two  recent  findings were  particularly  helpful for
simplifying  the prediction of inherent soil credibility:
(a) that from the viewpoint of erodibility, very fine sand
(0.05 -  0.10 mm) would be more properly classified  as
silt than as sand (91), and (b) that percentages of sand,
silt and clay  must be  considered in relation to each
other, because  of  strong interaction between particle
sizes. The most informative particle-size parameter in the
Corn Belt study was M = % silt(100 - % clay), where the
very fine sand is included in the silt fraction. When this
parameter was included with organic-matter content, a
soil structure  index, and the profile permeability class,
prediction of  the erodibility factor was well within the
accuracy needed for field use (89). The equation is:
 K = (2.1 x 10-*) (12 -On^M1 •'4 + 0.0325(8 - 2)
    + 0.025(P-3),
38

-------
where Om = percent organic matter, M = the particle-size
parameter presented  above, S = structure index, and P =
permeability  class (92).  Permeability is a profile  para-
meter; the other three pertain to the upper few inches of
soil.
   The soil-erodibility nomograph presented in Volume
I,  Figure 11, provides a quick graphic  solution to this
equation.  However,  the  relationship changes when the
silt fraction exceeds 70%. This change is reflected in the
nomograph by the bend  in the percent-sand curves, but
is  not  reflected in the  above equation. The structure-
index and permeability-class codings are defined on the
nomograph (89).
   For a few special conditions, the nomograph solution
may be  modified  to  improve  K-value accuracy: (1)
Fragipans   and  claypans  reduce   permeability in  wet
seasons, but do not greatly reduce it for thunderstorms
that occur when soil is relatively dry. Separate credi-
bilities can  be  computed for dry and wet seasons by
using different  permeability  ratings in the  nomograph
formula.  (2) The  mulching effects of stone, gravel, or
shale on  the surface  are not accounted  for in the
nomograph equation. If used on such soils, it would be
applied to mechanical-analysis data for the soil exclusive
of the  large fragments, and the indicated K value would
then be reduced by treating the large fragments as partial
mulch  cover.  (3) The nomograph lacks  sensitivity to
differences in erodibilities of  desurfaced high-clay sub-
soils, because other chemical properties become impor-
tant under those conditions. Recent studies showed free
iron and aluminum oxides were important for high-clay
subsoils but not for most topsoils (58).
   Standard texture classes are too broad to be accurate
indicators of erodibility. Therefore, the K values listed in
Table 2a of Volume I  are only first approximations.
Nomograph solutions  will  show  a  broad  range  of
erodibilities within a  texture class.


     Topographic Features (Factors L and S)

   Soil loss per  unit area increases  as slopes become
longer or  steeper. The  USLE denotes effects  of slope
length by  L and  effects of steepness  by S.  Both are
dimensionless and expressed  relative to the  "unit plot"
dimensions defined for factor K. In practice, the two are
combined in a single topographic factor denoted by LS.
   Slope length is the distance from the point of origin
of overland flow  to the point where  either  the  slope
decreases enough that deposition  begins,  or the runoff
water enters a well-defined channel (63). The effect of
slope length on  runoff per unit area is generally not of
practical significance, although there have been instances
of statistically significant direct and inverse relationships
(83).  Neither  is  soil detachment  per  unit  area  by
raindrop impact greater  on  long slopes.  The effect  of
slope  length  is,  therefore,  primarily  due  to greater
accumulation and more channelization of runoff on the
longer slopes.  This increases the capability of the runoff
to detach and  transport soil material.
   Factor L in the USLE is dimensionless. For slopes
steeper  than 4% it is generally computed by the formula
L = (A/72.6)0-5, where A = slope length in feet and 72.6
feet is the benchmark length. The exponent of 0.5 is the
average  of values obtained in 10 independent studies in
which the observed  values ranged from 0.3  to 0.9 (97).
Field  observations  indicate that the exponent  is prob-
ably about 0.3 for slopes of less than 3%, and 0.4 for 4%
slopes.  Increasing  the exponent to 0,6 when slopes
exceed  10%, as suggested in  Agriculture Handbook No.
282,  is  of questionable validity. The higher exponents
observed in the  length-effect studies  were  associated
with plowed-out bluegrass sod  or abnormally severe rain
events, on slopes that did not exceed 10%. Both L and S
are believed to be  influenced  by density of cover, soil
erodibility,  and rainstorm characteristics, but  existing
data  are inadequate for mathematical  evaluations  of
these  interaction effects.
   There have been field indications  that the slope-
length exponent becomes smaller  for extremely  long
slopes.  This  is logical  because  slopes  approaching  a
thousand feet in length  would  rarely have a constant
slope  steepness along their  entire length, and upslope
depositional areas would be likely.
   Slope steepness  affects both runoff and  soil loss. In
the assembled plot data, runoff from small grain tended
to increase  linearly with increases in slope.  For row
crops the   increase  was  curvilinear,  increasing at  an
increasing rate (83). Soil loss increases more  rapidly than
runoff as slopes steepen.
   The  combined effects of length and steepness for
uniform slopes were shown  in Table 3, Volume I. The
table  was derived by the formula
      LS =
             72.6
430 sin2 6 + 30 sin 6
         6.574
                                           + 0.431
where m = 0.5 if the slope is steeper than 4%, 0.4 for 4%
slopes, and 0.3 for slopes of 3% or less; and 6 = the angle
of slope.
   The last quantity in  this equation is an unpublished
conversion of an earlier formula (96) to an expression in
terms of the sine of the angle  of slope. Within the range
of the research data, the two forms are equally accurate,
but  an expression in terms of sin  0 is more logical and
                                                                                                          39

-------
computes  more  realistic values  when  extrapolated  to
steeper slopes.
   The research data used to derive relationships of slope
length and  steepness to soil  loss were from plots not
longer than  270 feet  and slopes not steeper than  18
percent. The extrapolated values shown  in Table 3  of
Volume  I   for  slopes  that  exceed  these  dimensions,
although  speculative, are the  best estimates presently
available.  Soil  loss  estimates  for  slopes steeper than
about  30% are potentially subject to  considerable error.
Research on steep slopes is a major need.
   Shape of the  slope is also important.  \vhen a slope
steepens or flattens significantly toward the lower end,
or  is  composed of a  scries  of convex and concave
segments, its overall average gradient  and length  do not
correctly indicate the topographic effect on  soil loss. An
irregular slope can be viewed as a series of segments such
that the gradient within  each segment can, for practical
purposes, be considered uniform. The segments  cannot
be  evaluated as independent slopes when runoff flows
from one segment to the next. However,  the amount of
soil  detached on each segment can be computed by a
recently published  formula (18) and summed for the
entire  slope length. For each segment, the  effective slope
length is the distance from the top of the overall slope to
the foot of the particular segment.
   If the segments are selected so that they are also of
equal length,  the slope-effect table for uniform  slopes
can  be used with  appropriate adjustment  factors for
position of the segment  on the overall slope. For most
field slopes,  three segments  should be sufficient. The
procedure is as follows (87):
   Ascertain  the  percent slope  for each segment. Enter
the slope-effect chart or table with the total slope length
and read the LS value corresponding to the steepness of
each of the  three segments. Multiply the chart LS value
for the upper segment by 0.58, the middle-segment value
by  1.06,  and the  lower-segment value  by  1.37. The
average of  the three products is a good estimate  of the
effective LS value for that slope. The three products also
indicate the relative  magnitudes of soil loss on the three
slope segments. (If two segments are  sufficient, use the
multiples: 0.71 and 1.29. For four segments: 0.50,0.91,
1.18, and  1.40.  For five segments:  0.45,  0.82, 1.06,
1.25, and \A2.)(87).

        Cover and Management (Factor C)
   The ability of a  soil to  resist the erosive forces of
rainfall  and  runoff is  profoundly influenced by the
direct and residual effects of vegetation, crop sequence,
management, and agronomic erosion-control  practices.
The  effects  of   cropping and  management  must  be
estimated in combination, because of many interrelated
variables.  Nearly any crop can be grown continuously or
in any one of numerous rotations. The sequence within a
system can be varied. Crop productivity can be low, or it
can be high. Crop residues can be removed, left on the
surface, incorporated near the surface, or plowed under.
The amount of residues  can vary  from scattered  pieces
to complete surface cover. The crop can be planted in a
pulverized and smoothed seedbed, in a rough and cloddy
seedbed, or  with extremely little soil disturbance. It can
be  intertilled after emergence, or  the weeds can be
controlled with chemicals.  The effectiveness of crop-
residue  management  will  depend  on the  amount of
available residue. This, in turn, depends on the rainfall
distribution, the fertility level, and various management
decisions  made by the farmer. Also, the residual effect
of  meadow sod depends on the  type and  quality of
meadow,  on how the succeeding seedbed  was prepared,
and on the length of time  elapsed since the sod  was
turned under. The erosion-reducing effectiveness of a
crop  system  depends  on  how the levels of all these
variables,  and others, are combined on the field.
    Factor C in  the Universal Soil Loss Equation is the
ratio  of  soil loss  from  land  cropped under specified
conditions  to the corresponding loss from clean-tilled,
continuous  fallow (96) and therefore includes the effects
of  all these variables.  If the actual  soil loss equals the
potential loss predicted by the product of factors R, K,
L,  and   S,  factor  C=l.  This would be  clean-tilled
continuous  fallow or land where mechanical desurfacing
has removed all of the surface vegetation and most of
the root zone. Where there is any vegetative cover, where
the upper layer of soil contains  significant amounts of
roots  or plant residues,  or where cultural practices
increase  infiltration and reduce velocity, soil loss is less
than  the  product RKLS.  Factor C brings  this reduction
into  the  soil loss computation.  On  cropped land,  C
ranges from about 0.60  downward  to less than 0.01.
This  great  flexibility  in  the value of C is  extremely
important to erosion-control planners.  If C  is reduced,
soil loss is reduced by the  same percentage.
    The  canopy protection  of crops varies  widely for
different  weeks or months in the crop year. The overall
erosion-reducing effectiveness of a crop depends on how
much of the erosive rain falls while the crop provides the
least  protection.  The correspondence of periods of
highly erosive rainfall with  periods of  good or poor
vegetative cover differs appreciably  between geographic
regions.  Therefore the C value  for a particular  crop
system will not be the  same in all parts of the country. A
field-tested   routine  is  available   for  computing site
C-values  that reflect  the  net effect of the interrelated
crop and  management  variables in whatever combination
40

-------
they occur at the site and in relation to the local rainfall
pattern.
   The entire rotation cycle  is divided into a series of
cropstage periods so defined that cover and management
may be considered approximately constant within each
period. The five cropstage periods are defined as follows,
for each crop-year in the system (57, 96):
   Period  F  -  Rough fallow. Turn  plowing  to  final
     seedbeed preparation. (No-plow systems  omit this
     period.)
   Period  1 - Seedling. Seedbed preparation to  1 month
     after seeding.
   Period  2 -  Establishment.  The second  month  after
     spring or summer  seeding.  For fall-seeded grain,
     this   period  extends  to about May  1  in  the
     Northern  States, April  15  in the Central States,
     and April  1 in the Southern States.
   Period  3 -  Developing and maturing crop. End of
     period 2 to harvest.
   Period  4 -  Residue  or  stubble.  Crop harvest to
     plowing or new  seeding. (When meadow  is seeded
     with  small grain, period 4  ends about 2 months
     after grain harvest. The vegetation is then classified
     as established meadow.)
   Probable calendar dates for the events that  begin the
successive periods  are selected  on  the  basis  of  local
climate and farm practice. The fraction of the annual El
that  normally  occurs  in that  locality  during  each
cropstage  for  each of  the  crops in the rotation  is
determined from the applicable El-Distribution Curve in
Agriculture  Handbook  No.  282. These  fractions are
multiplied  by the  corresponding  soil loss ratios  from
Table 2 in the same handbook. The sum of the products
obtained for the cropstage periods in any one year is the
C-value  for that  particular  crop  in that  system.  The
crop-system C-value  is   the  sum of  all  the  partial
products, divided by the  number  of years in the system.
This procedure  has been used by the Soil Conservation
Service to develop local C-value tables that are available
from their state offices. The illustrative C values given in
Table 4 of Volume I were also derived by this procedure,
but the seeding and harvest dates and the  El-distribution
data were  generalized  and  are  not  precise  for  any
particular location.
   The 33 regional El-distribution curves in Agriculture
Handbook  No.  282  were derived  from  the  22-year
rainfall records used to develop the iso-erodent (R-value)
map (82.) Corresponding data for the 11 Western States
and  Hawaii  are  presently  available  only as  tentative
estimates.
   The soil-loss-ratio table (96) was derived from analy-
sis of more than  10,000 plot-years of erosion data. The
data in this  table are percentages of soil loss from the
indicated combinations  of cover and  management  to
corresponding losses from continuous fallow. The table
has limitations that need to be recognized. The "mini-
mum tillage" classification  applies only to plow-based
systems in which disking and smoothing are omitted. A
partial list of ratios for no-plow systems that retain some
or all  of the residues on the surface was published in
1973 (86). The ratios for  corn in cropstage 4, residues
left, are for stalk cover  as left by the picker. Shredding
the stalks provides  more complete cover and reduces the
soil-loss ratio. Some of the crop systems in  the Western
States  and Hawaii are not represented, but approximate
values  for these systems  are now available from the Soil
Conservation Service, Western Technical Center, Port-
land, Oregon. Approximate C-values for range, wood-
land, and idle land were  published in  1974 (87).
   Practices that  depend on small rates of residue and/or
tillage-induced  surface   roughness for erosion-control
effectiveness  will be ineffective  if slopes are excessively
long. The precise length limits  for various mulch rates,
slope steepnesses, kinds of soil, and rainfall patterns have
not been determined. Investigations by the Agricultural
Research Service  are underway to improve or verify the
approximate  limits given  in Table 13, Volume I.
         Supporting Practices (Factor P)

   This factor is similar to C except that P accounts for
additional effects of practices that are superimposed on
the  cultural practices, such  as  contouring,  terracing,
diversion,  and  contour   stripcropping.  Approximate
values of P, related only to slope steepness, were listed in
Table  5, Volume  I. These values are  based  on rather
limited field data, but factor P has a narrower range of
possible values than the other five factors.  Influences of
type  of  vegetation,   residue  management,  rainstorm
characteristics, and soil properties on the value of P have
not been evaluated to the point of predictability.
                                   EROSION CONTROL METHODS
   Specific types of erosion-control practices were dis-
cussed in  Section 4.1  of Volume I.  These discussions
included general information on the advantages, limita-
tions,  and  variability of  each type  of practice. The
indicated percentages of reduction in soil loss were based
on C values estimated from all the available data rather
                                                                                                          41

-------
than on results of any specific local  experiment.  This
section  discusses principles and relationships that deter-
mine the effectiveness of erosion-control practices.
   If  surface  runoff can  be eliminated,  movement  of
sediment from a field will be insignificant. Breakdown of
soil aggregates by raindrop impact, rearrangement of soil
particles,  and surface  sealing can occur  before runoff
begins, but very little sediment will leave the field unless
surface  runoff is available to  transport it. Land treat-
ments that result in a deep, fertile topsoil, a high level of
organic matter, good  tilth, and  good vegetative cover
increase infiltration and reduce runoff. These conditions
may completely eliminate surface runoff from moderate
rainstorms on some areas. Generally, however, where the
rainfall  is adequate for crop production some of it falls
at intensities  greater than the soil  can infiltrate  even
when well managed, and runoff occurs.
   Land  treatments that increase infiltration and the
capacity  of the soil  to  store  water will  reduce small-
watershed flooding that results from short, intense  rains
during  the growing  season.  However,  when the  soil
becomes  saturated  to  a considerable depth, as is often
the case in major flood periods, cultural practices  have
much  less effect  on  runoff. Erosion-control  practices
must  also reduce the shear stress and transport capacity
of the  runoff.  This  means reducing runoff  amounts,
velocities, and depths  and dissipating the flow energies
on plant residues rather than on the soil surface.
   Erosion-control practices rely primarily on five means
of reducing erosion: 1) vegetation, 2) plant residues, 3)
improved tillage methods, 4) residual effects of crops in
rotation,  particularly  systems that include  grass  and
legume meadow, and 5) mechanical supporting practices.
A sixth potential approach would be use of chemical soil
stabilizers, but they have not yet  become economically
feasible for field use.

                    Vegetation

   Vegetation (a) intercepts rainfall  and thereby reduces
runoff and soil-particle detachment  by drop impact, (b)
increases  the  soil's water-storage capacity through  tran-
spiration,  (c) retards erosion by  decreasing  runoff
velocity,  (d)  physically  restrains soil  movement, (e)
improves aggregation  and porosity  of the  soil, and (0
in  eases biological activity in the soil (59).
 Crop Canopies

   Leaves and branches that are not in contact with the
 soil  reduce runoff from small rains but have  relatively
 little influence  on the amount and velocity of runoff
from  prolonged rains.  In  542 plot-years of convention-
ally  planted corn,  the  average runoff per  thousand
foot-tons of computed rainfall energy was only about 15
percent less in cropstage  3  than  before  canopy  had
developed (90). But soil loss per El unit from a field of
clean-tilled  90-bushel corn  is about 60 percent less in
cropstage  3  than  in cropstage 1 (96,  Soil Loss Ratio
Table), primarily as a result of raindrop interception by
the canopy. Water drops  from  canopy  may regain
appreciable  velocity, but usually not the terminal veloci-
ties of free-falling  raindrops. Therefore, canopy reduces
rainfall erosivity by reducing its impact energy at the soil
surface. The amount of reduction depends  on its height
and density. Canopy effect can be viewed as a reduction
in the "effective" El of the rainstorms and as such can
be directly computed for specific situations.
   Figure  1  shows the  ratios of effective El's computed
for several drop fall heights to the El of unintercepted
rainfall (88). Percent cover was defined as the percentage
of the total ground area  that could not be  hit  by
vertically  falling raindrops because  of the  canopy. Soil
loss reductions due to canopy over  a bare soil should be
approximately  proportional to the  reductions in effec-
tive El. Figure 1 assumes a  median dropsize of 2.5 mm
for both  the  rain  and the droplets  formed  on the
canopy. Where rainfall is characteristically of low inten-
sity and small drops, canopy effect would be less.
   All crops develop some canopy, but this may require
several months, and most or all of it  may be lost with
the crop  harvest.  Good soil-fertility management and
narrow row  spacing  hasten the  development  of a
protective canopy. Crop sequences  can be  selected that
substantially reduce the length of time between succes-
sive plant covers,  and early seeded winter cover crops
can provide  interim cover.
 Vegetation At The Soil Surface

   Stands of grass or small grain are much more effective
 than a  raised  canopy.  Much of the  rain that  such
 vegetation intercepts moves down the blades and stems
 to a point so near the ground that the droplets regain no
 appreciable energy. The dense  vegetation  at the  soil
 surface also reduces runoff amount and velocity and
 physically restrains soil  movement. For about  5,000
 plot-years of data, runoff from small grain averaged 9
 percent of total precipitation, in contrast to 12 percent
 for  row crops (83).  Meadow  averaged  7 percent. The
 soil-loss-ratio table shows that soil loss from established
 small grain averages less than half of that under a canopy
 of conventionally  planted corn. Soil loss from a good
 quality grass and legume  meadow is generally negligible.
42

-------
        1.00
          ,80
  Lul
  U.
  U-
  Ul
  o
  2:
  <£
  O
  a:
  o
  o
  6
          ,60
         .40
          ,20
            0
                 *Average  fall   height  of
                  drops  from canopy  (feet)
                ±L
                            T
  ill
              0
20            40             60            80

PERCENT GROUND  COVER BY  CANOPY
   Figure 1.-Effect of crop canopy on effective EL
100
                 Plant Residues

  With one-crop systems, crop residues can supply very
effective cover during the approximately 8 months from
harvest until the next crop develops full cover. Stalky
residues such as corn and grain sorghum provide more
effective cover  when shredded than when left partially
standing.  Complete  residue  cover at  the  soil surface
virtually eliminates raindrop impact on the soil, greatly
reduces the detachment capability and transport capac-
ity of runoff, and usually increases infiltration. Runoff
through  or over a complete cover of residue  mulch is
very low in sediment content.
  For rowcrop production, high residue rates are most
fully utilized in the no-till systems. The seeds are planted
                       in narrow slots opened through the residue by a fluted
                       coulter or other device, without  tillage. No-till planting
                       for corn  has been very highly effective  in chemically
                       killed meadow or small grain,  in  grain  stubble, in
                       chopped  rowcrop residues, and  in winter-cover crops.
                       However, the practice is not adaptable to all fields (see
                       Section 4.1, Volume I).
                          Reductions in  soil loss by various rates of straw
                       mulch tested under simulated rain have varied with soil
                       type and surface conditions (1, 4, 9, 28, 37, 42, 45, 70).
                       Figure  2 shows the average  relation of  soil  loss  with
                       various rates  of mulch to corresponding losses with no
                       mulch, as observed on 35-foot cropland slopes subjected
                       to 5 inches of simulated rain in two 1-hour storms (86).
                       How much  the  slope length or steepness  could  be
                                                                                                    43

-------
increased before  unanchored mulch would be undercut
or transported by the flowing water has, however, not
been fully  determined. Mulch  applied on plowed and
disked  surfaces of silt loam soils with  slopes of 3 to 5
percent substantially reduced runoff, but mulch  on a
15'' slope of untilled loam from which oat stubble had
been removed with a scraper had no significant effect on
amount of  runoff. Under both conditions, however, one
ton of straw per  acre reduced the velocity of the runoff
by about 60 percent (45).
   Partial  incorporation of the residues by shallow
tillage, such as disking, reduces the percentage of surface
cover and loosens some of the soil for easy detachment.
The residues are  then less effective than equal quantities
left undisturbed on the surface.  In a  test under 5 inches
of simulated rain, no-till planting without prior disking
reduced soil loss  83 percent in contrast to a  73 percent
reduction with similar planting after disking (86). The
amount  that soil-loss is increased by shallow tillage will
vary in  relation  to initial amount of residue and how
much is covered by the tillage.
   Moldboard plowing inverts the upper 6 to 8 inches of
soil and usually covers virtually all  of the  residue. The
surface is  then quite  susceptible  to erosion. However,
even with annual turnplowing, leaving  the residues on
the  field  is far  better than  removing  them.  Regular
incorporation of crop residues by plowing gradually
increases the amount  of organic materials in the soil and
improves  water  intake  and  soil  structure.  For  82
plot-years  of continuous corn with residues removed
each fall, runoff during the seedling and establishment
months  averaged 83%  of  corresponding  losses from
fallow;  for 50  plot-years in  which the residues were
plowed  down each year, runoff in  those months aver-
                 1.0
                as
          o:
          o
          2
                a6
                0.4
                0.2
20
40
60
                                                                            80
                                     100
                        % OF  SURFACE   COVERED   BY MULCH
                               Figure 2.-Effect of plant-residue mulch on soil loss.
44

-------
aged 51% of that from fallow (90). Annual soil loss was
about 20% less where residues  were incorporated  than
where they were removed.

            Improved Tillage Methods

   That influence of a tillage practice on the disposition
of crop residues is extremely important for erosion and
sediment control is evident  from the preceding discus-
sion. The surface micro topography and condition of the
soil after tillage also strongly influence the amount of
soil erosion. Roughness of the soil surface and porosity
of the tilled layer are important parameters in describing
the structure of a tilled soil  (31). Rough surfaces detain
considerable quantities of  water in microdepressions
until they can enter the soil, and the porous soil layer
offers a channel  system to funnel water throughout the
tilled layer.  Random roughness reduces runoff velocity,
and  the  water  that  is  temporarily  ponded in  the
depressional areas shields portions of  the soil surface
from  particle  detachment  by  raindrop  impact.  Also,
some of the sediment  detached by raindrop impact on
the exposed  surfaces  is  deposited on  the  ponded
surfaces. Porosity and  roughness  are influenced by the
type of tillage and by the water content  of the soil at the
time of tillage. Pulverizing the soil increases erosion by
increasing the soil's detachability  and increasing  the
amount and rates of runoff.
   When soil dries after a rain that has  broken  down its
surface structure and washed fines into soil voids, crusts
develop that  are  strong  enough to  reduce  seedling
emergence (20,  47).  Surface seals  and  crusts reduce
water intake  by the  soil  and  substantially  increase
erosion (41, 78), Rough soil surfaces tend to concentrate
the dispersed material in the microdepressions and leave
the peaks  more porous, but mulches are more  effective
for preventing surface seal.
   Soil  compaction by heavy  equipment can  hamper
root and plant development and thereby increase  soil
erosion. Conservation  tillage practices generally require
less use  of heavy equipment  on  the field. Aspects of soil
compaction were recently summarized by the American
Society of Agricultural Engineers (2).
   Larson  (30)  points out  that the secondary  soil
aggregates around the seed and seedling roots must be
small enough to prevent undue  drying of the soil, must
provide  sufficient soil-seed  or soil-root contact  for
moisture transfer, must provide adequate  aeration, and
must not  be so  finely divided as to encourage surface
crusting or mechanical impedance when dry. However,
the area between the  rows,  which he designates as the
water management zone, may be rough and cloddy or
may be left untilled under a residue mulch.
   Conventional tillage includes primary and secondary
tillage  operations  normally performed in preparing a
seedbed  for a  given crop grown in a given geographic
area. Where the term is used in this manual as a basis for
comparisons, it includes moldboard plowing and several
disking and smoothing operations.
   Minimum tillage is  the minimum soil manipulation
necessary for  crop production  under existing soil and
climatic conditions. The term  is often loosely applied to
any system  with  fewer  operations  than a conventional
system,  but  it  is  most  accurate  when  applied to
plow-plant and wheeltrack-plant systems, in  which the
field is plowed but secondary  tillage is omitted. These
systems are  most  effective for erosion control when the
rowcrop  follows  one or  more  years of meadow, and
before the clods disintegrate.
   Conservation tillage includes  tillage  systems  that
create  as good an  environment   as  possible for the
growing crop and that optimize conservation  of the soil
and water  resources,  consistent with  sound economic
practices. Conservation tillage  includes maximum or
optimum retention of residues on  the soil surface and
use of herbicides to control weeds (98)
   No-till is a system whereby a crop is planted directly
into  a seedbed untilled since  harvest of the  previous
crop.

        Residual  Effects of Previous Crops

   The benefits of crop rotations for minimizing periods
of little or no vegetative cover  were pointed out earlier,
but crops and  management practices also have residual
effects that influence soil credibility under succeeding
crops.


Sod-based Rotations

   The greatest residual effects are derived from grass
and legume  meadows. In  data  assembled  from conven-
tional seeding and tillage practices,  soil losses from corn
following meadow ranged from 14  to 68  percent of
corresponding losses from corn on  adjacent  plots in
meadowless systems. Grass and legume mixtures were
more  effective  than legumes alone. The erosion-control
effectiveness of rotation  meadows  turnplowed before
corn planting was, in general,  directly  proportional to
the quality  of  the meadow, as measured by hay yields.
Erosion  reduction was greatest during the  fallow and
corn-seeding periods and decreased gradually for about 2
years. The effects of well-managed long-term meadows
were still apparent in the third year. When second-year
hay yield exceeded the first, 2-year meadows were more
                                                                                                        45

-------
effective than one-year meadows, but when allowed to
deteriorate in the second year they were less effective
(81).

Meadowless Systems

   Crops that are not sod forming also have beneficial
residual effects  on soil credibility, but they are  much
less  pronounced  than  those  of grass  and legume mead-
ows.  Corn generally  leaves the soil  less credible than
soybeans  but more erodible than good quality  small
grain.  All crop systems have beneficial residual effects
relative to continuous fallow; brief periods of fallow in a
rotation are  not  as  erodible  as continuous fallow.
Removal of the crop  residues year after year gradually
reduces soil  organic  matter and  adversely  affects soil
tilth. One-time incorporation of residues by moldboard
plowing had  little effect  on infiltration or erosion, but
repeated incorporation year after year had very substan-
tial effects (90).
          Mechanical Support Practices

   These are mechanical erosion-control practices used
when slopes are  too long or too steep for agronomic
practices alone to control erosion.

Contouring

   Furrows  made by plowing, planting, and cultivating
form natural channels in  which runoff accumulates. If
the tillage is up and down slope, the shear stress of the
runoff increases as the slope of the furrows increases,
and erosion may be serious.
   In contouring,  tillage  operations  are carried out as
nearly as practical on the contour. The general rule is to
lay out guidelines which assure that all tillage is within a
gradient limit of 1  to 2 percent  (59). On gently sloping
land, contouring will  reduce the velocity  of overland
flow by channeling it around the slope. Contoured ridge
or  lister  planting  substantially increases  the storage
capacity of the furrows  and permits  storage of large
volumes of water.  When contouring is used  alone on
steep slopes or under high  rainfall intensities and  soil
credibility,  the hazard of gullying is increased because
row breakovers may release the stored water. Breakovers
cause  cumulative  damage   as  the  volume of water
increases with each succeeding row.  If the  contour lines
are not carefully laid out and rows are  allowed to cross
natural depressions at gradients  much  greater than  2
percent, adverse  results of breakovers  may  completely
offset the beneficial effects of contouring. The effective-
ness of contouring is also impaired by decreased infiltra-
tion capacity due to surface sealing, and by reduction in
depression storage after tillage operations cease and the
soil settles (59).

Graded Rows

   Graded rows are land-formed to a precise gradient.
This improves  surface  drainage and decreases the likeli-
hood of row breakovers.
Contour Stripcropping

   Alternating contoured strips of sod with strips of row
crops is more effective than contouring alone. The sod
strips serve as filters when rows break, and much of the
soil washed  from a cultivated strip is filtered out of the
runoff as  it spreads within  the  first several feet  of the
sod strip  (64).  In  the  Mormon Coulee near  LaCrosse,
Wisconsin,  some  fields are  reported  to have  been
cropped in  strips for  more  than 70 years.  Where the
strips were on the contour, or nearly so, erosion control
was good. Where the strips were sufficiently off-contour
to give row  slopes of 5 percent or more, soil losses from
flow of runoff down the rows were high  (8).
   Systems  with alternate contoured strips in meadow
reduce soil  loss to about  half  of  that from the same
rotation with contouring alone. Three-year rotations of
sod,  small  grain,  and  row crop  were  slightly less
effective.  Alternate  strips of fall-seeded grain and row
crop  have effected some  reduction in soil  loss, but
alternate  strips of  spring-seeded grain and  corn  on
moderate  to steep slopes have not proved more effective
than contouring alone (64).
   Buffer  Stripcropping is a practice in which strips of
grass are laid out between contour strips of crops in the
regular rotations.  The  grass  strips  may be irregular in
width and may be placed on critical slope areas in the
field (59).


Terracing

   Terracing with contour farming is more effective foi
erosion control than Stripcropping, because it divides the
slope into segments with lengths equal to the terrace
spacing. With  Stripcropping or contouring,  the  entire
field slope length is the effective length. With Stripcrop-
ping,  the  saved  soil is largely that deposited in the sod
strips; with terracing  the  deposition is in the terrace
channels and may be as much as 80 percent (96) of the
soil moved  to  the  channel.  Erosion  control between
46

-------
terraces depends on the crop system and other manage-
ment practices; with stripcropping, an  effective sod-
based rotation is built into the system.
   If a  control level is desired  that will  maintain soil
movement  between  the  terraces  within the soil-loss
tolerance limit, the P factor for terracing should equal
the P factor for contouring. However, if the soil loss
equation is used to compute gross erosion for watershed-
sediment estimation, a  terracing P factor equal  to 20
percent  of the  contour  factor is warranted (96). Since
terracing shortens  the slope length, it also reduces the
bet ween-terrace soil loss by decreasing the topographic
factor. Dividing a  slope that is steeper than 4% into n
equal segments divides the value of factor L by \fn. In
the table of P values given  in Volume I, this reduction
was included in the Pf factor for convenience.
   The two major types of terraces are the bench terrace
and the broadbase terrace  (59). Broadbase terraces are
broad-surface  channels  or  embankments  constructed
across  the  slope  of rolling land.  They may  be  either
channel type or ridge type. The primary purpose  of the
graded or channel-type terrace is to remove excess water
in such  a  way as to minimize erosion. The primary
purpose  of the level or ridge-type terrace is moisture
conservation; erosion control is a secondary  objective.
The channel is level and is sometimes closed at both ends
to assure maximum water retention. The Zingg conserva-
tion bench  terrace is designed for use in semiarid regions
for moisture  conservation.  It  consists of an earthen
embankment and a very  broad  flat  channel that re-
sembles a land bench.
   The  steep-backslope   terrace is constructed with a
backslope of 50% or steeper, which is kept in  grass (8).
It may  be  either a graded or a  level  terrace. Parallel
grass-backslope terraces  with subsurface drains are now
gaining  popularity.  They  release the excessive  water
slowly and are also better adapted to use of large farm
implements than graded or level terraces.
                                   SEDIMENT DELIVERY RATIOS
   The sediment  delivery  ratio is the  parameter that
bridges  the gap  between  upslope  erosion  data and
drainage-area sediment yield. The sum  of the  soil-loss
estimates for the individual  tracts constituting a drainage
area approximates the  quantity of soil  moved from its  '
original  general  position.  To  compute  drainage-area
sediment yield, this estimate must be  adjusted down-
ward to compensate for deposition in terrace channels,
in sod waterways, in field boundaries, at the toe of field
slopes, in depressional areas, and along the path traveled
by the runoff as it moves from the field  to a continuous
stream system  or lake  (96). Sediment  additions from
sources along this path must also be taken into account.
Further changes in sediment content of runoff water will
occur during the  stream transport phase. The Universal
Soil Loss Equation computes gross sheet  and rill erosion,
but  it  does  not compute  deposition.  Nor  does  it
compute sediment from gully, streambank, and channel
erosion. The sediment  delivery  ratio  provides a method
of accounting  for the  sediment  losses  and gains that
occur below the areas where the USLE is applied.
   The delivery ratio is usually estimated from natural
drainage-area parameters and therefore does not account
for deposition  in terrace  channels  or  in constructed
settling basins or  traps.  The  amount  of sediment
deposited in these man-made devices near the sediment
source is subtracted  from the computed field erosion to
obtain the gross-erosion  estimate to  which the  delivery
ratio is applied. Two methods  of defining the  delivery
ratio will be discussed.
  Delivery Ratios for Dealing with Downstream
               Sediment Problems

   For this purpose, the delivery ratio is defined as the
ratio of sediment delivered at a given point in the stream
system  to  the gross erosion from all sources in the
watershed above  that point.  Guides for estimating this
ratio were given in Volume I, section 3.3c. The source of
most  of  the  information   presented  there was the
Sedimentation  Section  of   the  National  Engineering
Handbook  developed by the Soil Conservation  Service
(72).  The approximate delivery ratios that were listed
relative  to  watershed  size   were  obtained  from the
relationship curve derived from  published and  unpub-
lished data assembled by L. C. Gottschalk, G. M. Brune,
J. W. Roehl, R. Woodburn, S. B. Maner, L.  H. Barnes,
and L.  M. Glymph and presented in the  Engineering
Handbook. This curve relates the delivery  ratio to the
negative  0.2  power of drainage-area size. There have
been  indications  that the 0.1 power would be  more
accurate for large  drainage areas (3).
   Analyzing data  from  14  Texas  Blackland  Prairie
drainage areas  that  ranged  from  0.42  to  97.4  square
miles, Renfro  (57)  computed delivery  ratios ranging
from 0.62 for a drainage area of 0.5 square mile to 0.28
for an area of 100 square miles. These are  significantly
larger than would  have  been estimated from the SCS
general  relationship  curve, and emphasize  the need to
consider the  other  factors listed in Volume I as  well as
watershed size. Several other relevant publications are
                                                                                                         47

-------
listed in  the  literature citations (3,  13, 23, 24, 53, 56,
67).
   Delivery ratios derived on this basis are more appro-
priate for dealing with downstream sedimentation prob-
lems than for estimating the amount and composition of
cropland  sediment that reaches  a  continuous  stream
system.  However,  they  arc  presently more  directly
available than those discussed below and can be helpful
also for the latter purpose.

   Delivery  Ratios for Purposes of this Manual
   For evaluation of cropland contributions to sediment
in stream systems, the delivery ratio should be defined as
the ratio of  sediment  delivered at the place where the
runoff enters a continuous stream system or lake to the
gross  erosion in  the drainage  area above  that  point. It
will then not  be biased by sediment-content changes that
occur during  the stream transport phase. Where this ratio
is known or  can  be closely approximated from drainage-
area parameters,  multiplying it  by the computed gross
erosion will  estimate  the amount of sediment delivered
to the stream system.
   No general equation for sediment delivery ratios as a
function of drainage-area parameters is presently avail-
able.  A generally applicable upslope-deposition equation
is a major research need. However, guides for approxi-
mating  the  average   delivery  ratio  for  a  particular
drainage  area arc available. The  ratio can approach a
value of  1.0 for a particular field if  the runoff drains
directly into  a lake or stream system, with no obstruc-
tions  and no flattening of the land slope. On the other
hand, a wide expanse of forest duff or dense vegetation
below the eroding area may filter out essentially all of
the sediment except  some  of the colloidal  material.
These are the extremes.
    Anything that reduces runoff velocity (reduction in
 slope steepness, physical obstructions such  as ridges or
 living or dormant  vegetation,  ponded  water, etc.)  re-
 duces  its capacity  to  transport  sediment.  When the
 sediment load  exceeds  the  transport capacity of the
 runoff, deposition occurs. The observed sediment reduc-
 tions by  terracing or contour stripcropping are examples
 of the potential magnitude of upslope deposition.  More
 than 80% of the soil eroded between terraces may be
 deposited in  the terrace channels because of the  large
 reduction in runoff velocity due to the terraces. Most of
 the soil eroded from cultivated strips has been observed
 to be deposited in the sod strips when contoured  strips
 of  sod  were  alternated  with  equal-width  strips in
 cultivated crops.
    Relative to  the sediment-source area,  the delivery
 ratio  will generally be directly  related to  amount of
 runoff and  inversely related to soil particle size. Relative
. to the land between the  source area  and  the stream
 system,  the  ratio  will  be  directly  related  to  slope
 steepness  and  amount  of channel-type erosion, and
 inversely related to: distance of the source area from the
 stream system  or lake; density of vegetation at ground
 level; and number of flow obstructions  such as  field
 boundaries, culverts, etc.
    The delivery ratio for a given drainage area will not be
 constant  for  all  runoff events, because the depth and
 velocity  of runoff will  differ  with  storm  size  and
 antecedent  surface conditions. These differences will not
 only affect  transport efficiency; a major runoff event
 may  also pick  up  some  of  the  sediment deposited
 en route  to  the stream  or lake  in  prior  events. The
 average  delivery  ratio  for  a  drainage  area can  be
 estimated more closely and should suffice for estimates
 of long-term average sediment yields.
                                          TOLERANCE  LIMITS
   This section discusses merits and limitations of several
alternative methods  of defining soil  loss or sediment
limits, as background information for those who may be
involved  in developing state sediment control standards.
Soil loss  limits used to  illustrate points are not intended
as specific recommendations.
   Optimum  soil-loss limits for preservation of cropland
productivity  may  differ  substantially from optimum
sediment standards for  control of runoff pollution from
nonpoint  sources. The underlying considerations  are
quite different, and specific differences must be recog-
nized. Standards  will  be  most  beneficial when they
achieve both objectives with  the  least possible  adverse
effect on production of food and fiber.
      Tolerances for Preservation of Cropland
                     Productivity

    Tolerance limits on average annual soil loss have been
 used in this country for a quarter century to guide soil
 conservation  planning. Limits ranging from 2 to 5 tons
 per acre are applied to individual field slopes. Experience
 has shown  these  limits  to  be feasible and generally
 adequate  for preservation of high  productivity levels.
 The 2  to  5 ton  tolerances  represent the  collective
 judgment  of soil  scientists  in the  Soil  Conservation
 Service, Agricultural Research Service, and State agricul-
 tural experiment stations in the 1950's. Factors consid-
 ered in defining these limits were  published by the Soil
48

-------
Conservation Service in reports of five regional soil loss
prediction workshops held from 1960 to 1962.
   One of the major considerations was longtime mainte-
nance of adequate soil depth for good plant growth. The
rate  of natural soil  renewal  for  mature soils has been
hypothesized  to  balance  the  rate of  erosion  under
natural  conditions,  without  influences  of man  (65).
Since erosion  in excess of renewal rates reduces soil
depth, shallow soils were assigned lower tolerance limits
than  those  for  deep soils with  subsoil characteristics
favorable for plant growth. Prior erosion was a  factor
because of  its effect  on the soil profile. Other considera-
tions included the prevention of field gullying, sedimen-
tation  problems,  seeding  losses, soil  organic matter
reduction,  and  plant nutrient  losses.  Research directed
to precise definition of soil loss tolerances (65, 68) has
been extremely limited.


         Tolerances for Sediment Control

   Sediment-control  standards  that coincide with the
toleranfes established for purposes of soil conservation
have the distinct advantage that  a farmer is in compli-
ance if he  follows a conservation plan approved by the
SWCD. These  plans include a safety factor in that  they
are generally  designed to protect  the  most  erodible
portion of  the  field.  Since  field  slope  gradients are
seldom uniform, the average soil loss for the entire field
is usually less than that on  the slope the plan is designed
to protect.
   Uniformly applied sediment-control standards based
on  average  annual  soil losses are  perhaps   the  most
feasible starting point,  because of their  simplicity and
because knowledge of  precisely how much upslope soil
movement  can  be tolerated is inadequate. But  if the
initial standards fail  to attain the desired level of water
quality  control,  the next  step should  be a range in
standards to suit the  requirements  of various  local
conditions  rather than  successive lowerings of uniform
limits.  Uniformly lowering soil  loss  limits   to  attain
higher  water-quality  goals  would unnecessarily remove
substantial acreages from grain production.
   Before quantifying gross-erosion limits for  cropland,
specific objectives  of the limits  should  be  defined.
Uniform  soil loss  tolerances reduce the total quantity of
sediment produced. This is important  for reduction of
direct damage by deposition on upslope areas, on flood
plains, and  in lakes or drainage ditches. But for control
of water  pollution from nonpoint sources, other aspects
of the problem may be  more important than the amount
of soil  eroded  from  a particular  field slope.  These
include:  upslope deposition,  composition of  the  sedi-
ment, the protection needs, and fluctuations in  sediment
loads.

Upslope Deposition

   For control of water pollution from nonpoint sources,
soil material eroded from a field slope but deposited in
terrace channels, field boundaries, or elsewhere  along the
path  followed  by  the runoff en route  to the  stream
system is irrelevant. The  fractions  of sediments  eroded
from  upslope areas that  are delivered to a continuous
stream system or lake range from less than 10% to nearly
100%. Uniform  limits on erosion rates will allow a wide
range in quantities  of delivered sediments. Estimating
sediment delivery ratios was discussed in the preceding
section.  Low  sediment  delivery  ratios are  of  little
relevance to preservation of the eroding cropland, but
they  are  highly important for water quality control.
Basing sediment standards on gross erosion minus the
estimated upslope deposition would  achieve more uni-
form  control of sediment quantity  and allow  greater
cropping flexibility. This would be a great improvement,
but  sediment   quantity  is  not  the only important
criterion.
Composition Of The Sediment

   Sediment  traps or settling basins trap primarily  the
coarse  material. Clay,  fine silt, and light soil aggregates
remain  in  suspension  much longer  than the  coarse
material  and  are the  greatest concern as a source  of
turbidity and carrier of chemical  compounds.  There is
some particle-size selectivity in erosion, but generally the
composition  of washoff material as it leaves the field is
closely  related to  that  of the soil  from which it is
derived.  There  is  substantial size  selectivity in  the
transport and deposition phases, but the composition of
the sediment  as  it leaves  the field  will determine  the
proportion of fine material available  for transport  in
suspension.  Thus, for pollution control, variability  in
soil-loss limits should be related to  soil texture.

Protection Needs

   Sediment standards could also be selective in relation
to the  needs of the  particular  body of water  being
protected.  For example,  controls  need  to   be  more
intensive for land draining into recreational waters and
urban water  supplies  than for land draining into major
river channels.
                                                                                                          49

-------
 Fluctuations In Sediment Loads

   Short-time  high  sediment  yields  are  much  more
 relevant  for  pollution control  than for preservation of
 the land resource. The average annual soil loss from a
 particular crop system  on a given field is the  mean of
 yearly losses that may differ tenfold, or even a hundred-
 fold,  due to differences in the cover  and management
 effects of the crops in the system, fluctuations in rainfall
 erosivity, and intermittent crop failures.
   The  soil loss equation  shows  that  under  conditions
 where  a  6-year   rotation  of  corn-com-corn-wheat-
 meadow-meadow would average 5 tons of soil loss per
 acre per year with conventional planting and tillage, the
 first-year corn would average about 4  tons, second-year
 com 9 tons, third-year corn 14 tons, wheat 2.7 tons, and
 meadow 0.2  ton. On the average, at least half of the soil
 loss from the corn would occur during the first month
 after  preparation of the clean-tilled seedbed. Appendix
 tables in Agricultural Handbook No. 282 (96) show that
 about one year  in  ten the rainfall-erosivity  factor  is
 likely to exceed its  local average value by  50 percent,
 and one year  in twenty by 75  percent. If a 20-year
 rainfall occurred in the third corn year, the predicted
 soil loss for that year on this field would be  1.75  times
 14 tons, or  nearly 25  tons, even though the  longtime
 crop-system average would not exceed the 5-ton limit.
   Soil loss variability due to fluctuations in rainfall or
 occasional seeding failures cannot be prevented, and
 yearly or seasonal rainfall differences  can be predicted
 only  on a probability  basis. Because  these differences
 interact  with other  erosion factors, specific-storm soil
 losses can  presently not be accurately predicted.  How-
 ever,  for  each crop  in  the  system, the  effects of
 fluctuations  in rainfall tend  to  average-out over long
 time periods, and  the differences in cover and manage-
 ment effects of the crops in a  particular system are
 reasonably well known. Therefore, the  average annual
 soil loss for each  year in a crop  sequence can  be
 predicted by use  of the Universal Soil Loss Equation
 with about the same accuracy as crop-system  averages.
 This is done by deriving factor C on a yearly basis by the
 method illustrated in Agriculture Handbook No. 282.
   Limits prescribed on a crop-year basis would reduce
the frequency of very high single-year or single-event soil
losses. In the preceding example, a 5-ton limit on the
design loss in any year of the cropping system would
require the use of good  residue  management for the
second-year corn and no-till  planting in shredded-corn-
stalk  mulch  for the third corn  year.  If the soil  and
climate were  not compatible  with  no-till planting in
residue cover,  the  third-year corn would need to be
omitted from the cropping system.
   However, crop-year soil loss limits would need to be
higher or  more flexible than the present rotation-average
tolerances. If not, they  would prevent  production of
corn, soybeans, or other rowcrops on numerous fields
where these  crops  can be  grown  in  rotation  with
meadow and small grain, and they could also eliminate
the acceptability of periodic clean plowing for weed and
pest control  on  fields that are usually no-till  planted.
The reason for this is that crop-year limits would allow
no credit for  much  more drastic reductions in soil loss
during other years in the crop system.

          Modified Sediment Standards   •

   A  possible  alternative  would be to continue the
present crop-system-average tolerances and superimpose
limits on the  maximum design loss for  any one year in
the  system.  The first limit would allow  credit  for
meadows and other  low-erosion crops in  a  system
designed to preserve the productive capacity of the land.
The second limit could be sufficiently  higher to avoid
unnecessary  restrictions on  land use  and  yet guard
against frequent occurrence  of   very high  single-year
sediment yields.
   The second  limit  would  take into account  such
factors as the intermittent more erodible conditions that
cannot be avoided, upslope deposition, soil texture, and
specific control needs for the location.  Upslope deposi-
tion could be  increased by  requiring use of sediment
traps or filter strips of grass or  small grain  across the
lower end of a field in the years when it is plowed. The
same requirement could  apply to the second and  third
corn years in sod-based rotations.
                                           RESEARCH NEEDS
   The  past  40 years have brought great  progress in
erosion  control, but serious erosion  and sediment dam-
ages are still far  too frequent. Population pressures,
increased export demands for agricultural products, and
more substitution  of large  machines  for manpower
changed the erosion problems and intensified the hazard
on many millions of acres of productive cropland. Larger
fields generally mean longer continuous slopes. Exten-
sive monoculture sacrifices the potential residual effects
of sod-based systems. Large equipment greatly increases
production per man-hour  but is not compatible  with
following the field contours on  much of the cropland.
50

-------
Furthermore, soil conservation and sediment control are
two individual goals  and do not have the same require-
ments. Personnel and  resources  available  for  erosion
research in recent years have been insufficient to keep
pace with the changing needs.  Research  is particularly
needed in the following general areas. This research will
involve many preliminary and secondary investigations
that are not listed.

      Sediment Delivery to Stream Systems

   Average annual erosion  rates on cropland  can be
predicted  with reasonable accuracy, but the percentage
of this eroded soil  that reaches a continuous stream
system  cannot.   Sediment  delivery  ratios as  usually
defined by  geologists  for  dealing  with downstream
sediment problems are influenced  too much by stream
transport efficiency and sediment accretions from non-
agricultural sources to  provide  the information needed
for control of water pollution from nonpoint cropland
sources. If the sediment delivery ratio is used for this
purpose,  it should be defined as the ratio  of sediment
delivered at  the place where the runoff water enters a
continuous stream system to the gross erosion from the
drainage area above that point.
   The sediment  delivery  ratio,  by either definition,
represents an attempt to reflect deposition and sediment
accretions in a single factor. The net effect of the two
processes  would be  difficult to relate  to drainage-area
parameters  in  a  regression  equation   because  large
amounts of deposition and large sediment accretions can
occur in the same drainage area and balance each other.
Prediction equations for deposition  and for sediment
accretions from runoff-induced erosion below the field
areas need to be separately derived, each as a function of
the drainage-area  parameters pertinent  to that process.
Such  equations will  facilitate estimating  the effects of
cropland erosion  control  not  only on the amount of
sediment delivered to the stream system but also on the
composition  of sediment  yields  farther downstream.
Development of  a  better  understanding of the  basic
sedimentation  and erosion  processes involved between
the time  when runoff leaves  a field area and when it
reaches a continuous stream system is one  of the greatest
erosion and sediment research needs.

             Mathematical Modeling

   Recent progress  in  development  of mathematical
models for  simulating  the  erosion and  sedimentation
processes  on  field-size areas and  on watersheds  has
demonstrated the potential of such models to predict
temporal  and spatial distribution  of erosion and sedi-
mentation  and  to  predict  specific events  more accu-
rately. However, some of the basic relationships assumed
for these  initial models need  research testing, and the
parameters need to be defined for a wide  range of field
conditions. Relationships describing erosion  and  deposi-
tion in channels and gullies also need to be  derived.
   These models are generally complex and difficult to
use in  the field. A relatively simple model that computes
individual-storm soil  losses  more accurately  than the
Universal  Soil Loss Equation  is needed. Such a model
can use the basic format of the USLE, but  it  will need
separate  erosivity  factors for rill erosion and interrill
erosion, and  their  relationships to the other factors in
the equation  will need to be determined. Use of volume
and peak rate  of  runoff to predict rill erosion  shows
promise, but  it requires derivation of a cropland-runoff
prediction equation.
   Improvement of the basic models, and  research to
determine the needed parameter relationships, should be
emphasized. Such models can  provide more  dependable
interpretation and  extrapolation of field-plot data, and
the predictions  of spatial  and  temporal  variations in
erosion and deposition are needed for both conservation
and pollution-control planning.

               Residue Management

   Residue management is  one of the major tools for
erosion control. In  the densely populated  countries, few
residues  are  usually  available  for erosion-control use
because they are needed for other purposes. We  may
soon  have  similar problems  in  this  country if crop
residues  become  economically  profitable  sources of
energy, concentrated feeds, or  building  materials. Re-
search must determine  the  optimum treatment  and
placement  of very limited  residues and  the  optimum
amount  and  type  of  associated  tillage required to
minimize  erosion   in the  absence  of what  we now
consider adequate cover.
   Neither has  the  optimum amount and placement of
residues  where they  are  abundantly available  been
determined.  Optimum placement of a portion of the
residue may permit incorporation of the remainder into
the  topsoil.  This  may  reduce  soil-temperature  and
wetness problems  without  decreasing the erosion con-
trol.

    Critical Slope-Length Limits for  Practice
                    Effectiveness

   Critical slope-length limits for effectiveness of partial
mulch covers and favorable  microtopographies provided
by conservation tillage practices need to be  determined.
                                                                                                         51

-------
If limits can be defined in terms of depth and velocity of
runoff, they can then be related to soil, topography, and
rainfall characteristics for guidance in field application.
Clear definition of critical slope-length and drainage-area
limits is needed for improved terrace  spacing design and
to  prevent unexpected  failures  of some  agronomic
practices. The investigations should include evaluation of
anchored versus loose mulches and of different types of
residues.
   Slope-length limits for effective contouring need to
be more accurately defined in relation  to  permeability,
soil stability,  crop cover, and other factors. Successive
row breakovers can result in rill erosion that more than
offsets the  reduction in sheet erosion  effected by  the
contouring.

      Erosion Index for Special Conditions

   The El  parameter  is a good indicator  of the erosive
potential of the rainfall and  runoff in   most  of this
country, but  there  are  a few  conditions for which
further investigation  of this  factor is urgently needed.
The  erosivity of  surface  runoff that  is  not directly
associated with drop impact needs to be evaluated, such
as  runoff  from thaw  and snowmelt.  This item  is
particularly important  in  the  Palouse  Region of the
Northwest.  The  effects of soil-surface   shielding  by
ponded or very slowly  moving runoff  also need to be
identified and evaluated. These effects may account for
the difficulties experienced with the El parameter on the
Coastal Plains  of the Southeast.

               Topographic  Factor

   The topographic factor needs further research, both
with reference to factor interactions and with reference
to long or  steep slopes. The effects of slope length and
steepness on soil erosion are known to be more variable
than indicated by existing formulas.  There is evidence
that they are significantly influenced by  mutual inter-
action and by interactions with cover, soil texture, and
rainstorm characteristics (or runoff rate). These inter-
action effects  need to be quantified so that variations in
the topographic factor can be predicted. This is impor-
tant both for soil conservation planning and for pollu-
tion control guides.
   Topographic effect also needs to be determined for
steep roadbank and construction slopes and for long
watershed  slopes. Existing  slope-length and steepness
formulas were derived from data on slopes not steeper
than 18% and, with only one exception, not longer than
270 feet. Extrapolation of the formulas to slopes that
far exceed these dimensions is quite speculative.

                  Soil Erodibility

   The  soil-erodibility  nomograph needs  to be aug-
mented for greater accuracy on high-clay subsoils and on
sandy  loams. Effects of soil  chemistry on credibility
need further investigation and quantification. Suscepti-
bilities of soils to sheet erosion and to rill erosion should
be  evaluated separately, and influences of montmoril-
linitic  clays  need further  study. Stripmine  areas  and
spoilbanks need specific research attention.
   Particle  size sorting in erosion and  sedimentation is
important  for pollution control  and  has  not  been.
adequately investigated.

                 Runoff Equation

   A cropland runoff equation designed for general field
use  would  be a valuable  asset in pollution control
planning. It could also  provide an additional factor for
the Universal Soil Loss Equation  that would improve its
accuracy, particularly for moderate storms.

                  Sediment Traps

   Vegetated filter strips, settling basins, and sediment
traps can be used to cause sediments to deposit near the
point of origin, but more needs to be learned regarding
their design for optimum trapping efficiency and their
particle size selectivity. Also, extreme runoff events may
pick up substantial amounts of sediment from the traps.
The probabilities of such occurrences and methods of
minimizing them need to be investigated.
52

-------
                                        LITERATURE CITED
 I.  Adams, J. E. 1966. Influence of mulches on runoff,
    erosion, and soil moisture depletion. Soil Sci. Soc.
    Amer. Proc. 30:110-114.

 2.  American Society of Agricultural Engineers. 1971.
    Compaction of agricultural soils. St. Joseph, Mich.,
    471  p.

 3.  American Society of  Civil  Engineers, Task Com-
 •  mittee  on  Preparation  of Sedimentation  Manual.
    1970.  Chapter IV. Sediment sources and sediment
    yields. Jour. Hydraulics Div. 96:1283-1327.

 4.  Barnett, A.  P., Diseker, E. G., and Richardson, E. C.
    1967.  Evaluation of mulching methods for erosion
    control on  newly  prepared and seeded  highway
    backslopes. Agron. Jour. 59:83-85.

 5.  Barnett, A.  P., and Rogers, J. S. 1966. Soil physical
    properties  related  to   runoff  and erosion  from
    artificial rainfall. Trans. Amer.  Soc.  Agr.  Engin.
    9:123-125.

 6.  Baur, A. J.  (dim). 1952. Soil and water conserva-
    tion   glossary.  Jour.  Soil  and  Water  Conserv.
    7:41-52,93-104, 144-156.

 7.  Baver, L. D. 1940. Soil physics. John Wiley & Sons,
    New York, 370 p.

 8.  Beasley, R. P. 1972. Erosion and sediment pollution
    control, Iowa State Univ. Press, Ames, 320 p.


 9.  Borst, H. L., and Woodburn, R. 1942.  The  effect of
    mulches and surface conditions on  the water rela-
    tions and erosion of Muskingham soils. Tech. Bui.
    825, U.S. Dept. Agr., Washington, D.C., 16 p.

10.  Browning,  G. M., Parish, C. L., and  Glass, J.  A.
    1947.   A method  for  determining the  use and
    limitation of rotation and conservation practices in
    control of soil-erosion practices in  Iowa. Soil Sci.
    Soc. Amer. Proc. 23:246-249.
11.  Carter, C. E., Greer, J. D., Braud, H. J., and Floyd,
    J.  M.  1974.  Raindrop characteristics  in  South
    Central  United States.  Trans.  Amer.  Soc. Agr.
    Engin. 17:1033-1037.
12. Chepil,  W.  S., and  Woodruff,  N.  P. 1963.  The
    physics  of wind erosion and its  control. Advan. in
    Agron. XV:211-302. Academic Press, New York.

13. Corinth, R.  L.  1970.  Effects of watershed charac-
    teristics on reservoir sediment deposition. Misc. Pub.
    No. 970, U.S. Dept. Agr., Washington, D.C.

14. David, W. P., and  Beer, C. E.  1975. Simulation of
    soil  erosion.   Trans.  Amer.   Soc.   Agr.  Engin.
    18:126-133.

15. Dideriksen, R. I., and Grunewald. A. R. 1974. What
    is  happening  to  our  soil  resources:  the  current
    inventory of rural  land. Proc. 29th Ann. Mtg., Soil
    Conserv. Soc. Amer.: 15-17.

16. Ellison, W.  D.  1947.  Soil erosion studies   part  1.
    Agr. Engin. 28:145-146.

17. Foster, G. R., and Meyer, L. D. 1975. Mathematical
    simulation  of  upland  erosion  using  fundamental
    erosion mechanics. In  Present and prospective tech-
    nology for  predicting  sediment yields  and sources:
    190-207. ARS-S40, U.S. Dept.  Agr., Washington,
    D.C.

18. Foster, G. R., and Wischmeier,  W. H. 1974. Evaluat-
    ing irregular slopes for soil loss  prediction. Trans.
    Amer. Soc. Agr. Engin. 17:305-309.

19. Freeman, 0. L., and Bennett, I. L. 1969. Control of
    agriculture-related pollution. A report  to the Presi-
    dent, by  U.  S.  Dept. Agr.  and Off.  Sci.  and
    Techno!., 102 p.

20. Gamer,  T.  H., and   Bowen,  H. D.  1966. Plant
    mechanics in seedling emergence. Trans. Amer. Soc.
    Agr. Engin. 9:650-653.

21. Glymph, L.  M., and Carlson, C. W. 1968. Cleaning
    up our rivers and lakes. Agr. Engin. 49:590,607.

22. Gunn,  R., and  Kinzer, G.  D.  1949. The terminal
    velocity  of fall for  water droplets.  Jour.  Met.
    6:243-248.

23. Happ, S. C. 1971. Discussion  of sediment sources
    and  sediment yield. Amer.  Soc.  Civ. Engin., Jour.
    Hydraulics Div. 96:597-599.
                                                                                                       53

-------
24. Happ, S. C. 1972. Valley sedimentation as a factor
    in environmental impact on rivers:  27.1-27.25. H.
    W. Shen Pub., Ft. Collins, Col.

25. Holt, R. F., Johnson, H. P., and McDowell, L. L.
    1973. Surface water quality. Proc. Natl. Conserv.
    Tillage Conf., Soil Conserv. Soc. Amer.: 141-156.


26. Hudson, N. W. 1971. Raindrop size. In Soil conser-
    vation:50-56.  Cornell  Univ.  Press, Ithaca,  New
    York.

27. Klingebiel, A. A., and  Montgomery, P. H. 1961.
    Land capability  classification. Agr.  Handbook No.
    210. U. S. Govt. Printing Off., Washington, D. C.,
    21 p.

28. Kramer,  L. A.,  and  Meyer,  L. D. 1969. Small
    amounts of surface  mulch reduce soil erosion and
    runoff velocity.  Trans.  Amer.  Soc.  Agr. Engin.
    12:638-641,645.

29. Langbein, W. B., and Schumm, S. A. 1958. Yield of
    sediment in relation to  mean annual precipitation.
    Trans. Amer. Geophys. Union 39:1076-1084.

30. Larson,  W. E.  1962. Advances in minimum tillage
    for corn. 17th  Ann. Hybrid  Corn Industry Res.
    Conf., Amer. Seed Trade Assoc. 17:44-51.

31. Larson,  W. E., and  GUI, W. R.  1973. Soil physical
    parameters for designing new tillage systems. Proc.
    Natl. Conserv.  Tillage  Conf., Soil Conserv.  Soc.
    Amer.: 13-21.

32. Lattanzi, A. R., Meyer, L. D., and Baumgardner, M.
    F.  1974.  Influences  of mulch rate   and slope
    steepness on interrill erosion. Soil Sci.  Soc. Amer.
    Proc. 38:946-950.

33. Laws, J. O. 1941. Measurements of fall velocity of
    water drops and  rain drops. Trans. Amer. Geophys.
    Union 22:709-721.

34. Lelland, J. H., Rogers, H. T., and Elson, J. 1941.
    Effect  of  slope, character  of  soil, rainfall,  and
    cropping treatments on erosion  losses  from  Dun-
    more silt loam. Va. Polytech. Inst.  Tech. Bui. No.
    72,31 p.
35. Lloyd, C. H., and Eley,  G. W.  1952. Graphical
    solution of probable soil  loss formula for North-
    eastern  Region.  Jour.  Soil  and  Water  Conserv.
    7:189-191.

36. Mannering, J. V., Johnson, C. B., and Wischmeier,
    W.  H. 1974.  Erodibilities  of selected soils in the
    Maumee  Basin.   1974  Agron.  Abs.,  Amer.  Soc.
    Agron.:167.

37. Mannering, J. V. and Meyer, L.  D. 1963. Effects of
    various rates  of surface mulch  on infiltration  and
    erosion. Soil Sci. Soc. Amer. Proc. 27:84-86.

38. Menard, H. W. 1961. Some rates of regional erosion.
    Jour. Geol. 69:154-161.

39. Meyer, L. D.  1971. Soil erosion by water on upland
    areas. In River mechanics II, Chapter 27. H. W. Shen
    Pub., Ft. Collins, Col., 25 p.

40. Meyer, L. D., Foster, G. R., and Romkens, M. J. M.
    1975. Source of soil eroded by water from upland
    slopes. In Present and prospective technology for
    predicting sediment yields and sources:  177-189.
    ARS-S40, U.S. Dept. Agr., Washington, D.C.

41. Meyer, L. D., and Mannering, J. V. 1961. Minimum
    tillage for com:  its effect on  infiltration   and
    erosion. Agr. Engin. 42:72-75.
42. Meyer,  L. D., and Mannering, J. V. 1963. Crop
    residue  mulches for controlling erosion on sloping
    land under intensive cropping. Trans. Amer. Soc.
    Agr. Engin. 6:322,333, 337.

43. Meyer,  L. D.,  and  McCune,  D. L. 1958. Rainfall
    simulator for runoff plots. Agr. Engin. 39:644-648.

44. Meyer,  L. D., and Wischmeier, W. H. 1969. Mathe-
    matical  simulation of the process of soil erosion by
    water. Trans. Amer. Soc. Agr. Engin. 12:754-758,
    762.

45. Meyer, L. D., Wischmeier, W. H., and Foster, G. B.
    1970. Mulch rates required for erosion control on
    steep slopes. Soil Sci. Soc. Amer. Proc. 34:928-931.
54

-------
46. Middleton,  H. E.,  Slater, C. S., and Byers, H. G.
    1932-1934. Physical and chemical characteristics of
    the soils from the  erosion experiment stations.
    Tech.  Bui. No. 316 and No. 430, U. S. Dept. Agr.,
    Washington, D.C.

47. Morton, C.  T., and Buchele,W. F. 1960. Emergence
    energy of plant seedlings. Agr. Engin. 41:428431,
    453-455.

48. Musgrave, G.  W. 1947. The quantitative evaluation
    of factors in  water erosion, a first  approximation.
    Jour. Soil and Water Conserv. 2:133-138.

49. Newman, J. E. 1970. Climate in the 1970's. Crops
    andSoils22(4):9-12.

50. Olson, T.  C., and Wischmeier, W. H. 1963. Soil
    erodibility  evaluations for soils on the runoff and
    erosion   stations.  Soil   Sci.  Soc.  Amer.  Proc.
    27:590-592.

51. O'Neal, A.  M.  1952. A  key  for  evaluating soil
    permeability  by means of certain  field clues. Soil
    Sci. Soc. Amer. Proc. 16:312-315.

52. Onstad, C. A., and  Foster, G. R. 1975. Erosion
    modeling on  a watershed. Trans. Amer. Soc. Agr.
    Engin. 18:288-292.

53. Pacific Southwest  Inter-Agency Committee.  1974.
    Report  of  water  management  subcommittee, ero-
    sion and sediment yield methods, 45 p.

54. Parr,  J.  F.,  and  Bertrand, A. R.  1960. Water
    infiltration  into soils. In  Advan.  in  Agron.  XII:
    311-363. Academic Press, Inc., New York.

55. Peele, T. C., Latham, E.  E., and Beale, 0. W. 1945.
    Relation of the physical properties  of different soil
    types  to erodibility. South Carolina Agr. Expt. Sta.
    Bui. No. 357.

56. Piest,  R. F.  and Bowie,  A. J. 1974. Gully and
    streambank  erosion.  Proc. 29th Ann. Mtg., Soil
    Conserv. Soc. Amer.: 189-196.

57. Renfro, G. W. 1972.  Use of erosion equations and
    sediment delivery ratios for predicting  sediment
    yield.  Proc.  Sediment Yield  Workshop, Oxford,
    Miss. In press, U. S. Govt. Printing Off., Washington,
    D.C.
58. Roth, C. B., Nelson, D. W., and Romkens, M. J. M.
    1974. Prediction of subsoil erodibility using chem-
    ical,  mineralogical and  physical parameters. Rep.
    EPA-660/2-74-043, 111 p. U.S. Govt. Printing Off.,
    Washington, D. C.


59. Schwab, G. O.,  Frevert, R. K., Edminster,  T. W.,
    and Barnes, K. K. 1966. Soil and water conservation
    engineering, Second Ed.  John Wiley & Sons, Inc.,
    New York.

60. Skidmore, E. L., Fisher, P. S., and Woodruff, N. P.
    1970.  Wind  erosion equation: computer solution
    and  application.  Soil   Sci.  Soc.  Amer.   Proc.
    34:931-935.

61. Skidmore,  E. L., and  Woodruff, N. P. 1968. Wind
    erosion  forces  in  the  U. S.  and their  use  in
    predicting soil loss. U.S. Dept. Agr., Agr. Handbook
    No.  346. U. S. Govt.  Printing Off.,  Washington, D.
    C.,42p.

62. Smith,  D.  D. 1941. Interpretation of soil conserva-
    tion  data for field use. Agr. Engin. 22:173-175.

63. Smith, D. D., and Wischmeier, W. H. 1957. Factors
    affecting sheet and  rill erosion.  Trans. Amer. Geo-
    phys. Union 38:889-896.

64. Smith, D. D., and Wischmeier, W. H. 1962. Rainfall
    erosion.  In  Advan. in Agron. XIV: 109-148. Aca-
    demic Press, Inc., New York.

65. Smith, R. M.,and Stamey.W. L. 1965. Determining
    the   range   of  tolerable   erosion.  Soil  Sci.
    100:414424.

66. Soil  Conservation Society  of  America.  1970. Re-
    source  conservation glossary. 52 p.

67. Stall, J. B., and  Bartelli,  L. J.  1959. Correlation  of
    reservoir sedimentation  and watershed factors. 111.
    State Water Survey Rep. No. 37.

68. Stamey, W. L., and Smith, R. M. 1964. A conserva-
    tion  definition  of erosion  tolerance.   Soil  Sci.
    97:183-186.

69. Stanford, G., England, C. B., and  Taylor,  A. W.
    1970. Fertilizer use and water quality. ARS 41-168,
    U.S.  Dept. Agr., Washington, D. C., 19 p.
                                                                                                       55

-------
70. Swanson, N. P., Dedrick, A. R., Weakly, H. E.,and
    Haisc. H.  R.  1965. Evaluation  of  mulches  for
    water-erosion  control.  Trans.  Amer.  Soc.  Agr.
    Engin. 8:438-440.

71. U.  S.  Department  of  Agriculture.  1971.  Basic
    statistics-national inventory of soil and water con-
    servation  needs. Statis.  Bui. No.  461. Washington,
    D.C.

72. U. S. Department of Agriculture. Soil Conservation
    Service. 1971. Sediment sources, yields and delivery
    ratios. In National  engineering handbook. Sect. 3.
    Washington, D.C.

73. U. S. Department of Agriculture. Soil Conservation
    Service. 1972. Land resource regions and major land
    resource areas.  Agr.  Handbook No. 296. U. S. Govt.
    Printing Off., Washington. D. C.

74. U.  S. Weather Bureau.  1958.  Rainfall  intensity-
    frequency regime. Tech. Paper No. 29, 5 parts.

75. Van  Doren.  C.  A., and  Bartelli,  L. J.-  1956. A
    method  of  forecasting  soil  losses. Agr.  Engin.
    37:335-341.


76. Voxnesensky,  A. S., and  Artsruui, A. B. 1940. A
    laboratory method for determining the anti-erosion
    stability of  soils. Soils  and Pert., Commonwealth
    Bur. Soils 10:289.


77. Wadleigh, C. H. 1968. Wastes  in relation to agricul-
    ture and forestry. Misc. Pub. No. 1065, U. S. Dept.
    Agr., Washington, D. C., 112 p.

78. Whitaker, F. D., Heinemann, H. G., and Wischmeier,
    W. H. 1973. Chemical weed controls affect runoff,
    erosion,  and  corn   yields. Jour.  Soil  and  Water
    Conserv.73:174-176.

79. Wischmeier, W.  H. 1955. Punched  cards  record
    runoff and soil-loss data. Agr. Engin. 36:664-666.

80. Wischmeier, W.  H. 1959. A rainfall erosion index for
    a universal soil-loss  equation.  Soil Sci. Soc.  Amer.
    Proc.  23:246-249.

81. Wischmeier,  W.  H. I960.  Cropping-management
    factor evaluations for a  universal soil-loss equation.
    Soil Sci. Soc. Amer. Proc. 24:322-326.
82. Wischmeier, W. H. 1962. Rainfall erosion potential -
    geographic and  locational  differences of distribu-
    tion. Agr. Engin. 43:212-215.

83. Wischmeier.  W.  H.  1966.  Relation  of field-plot
    runoff to management and physical factors. Soil Sci.
    Soc. Amer. Proc. 30:272-277.

84. Wischmeier, W. H. 1972.  Upslope erosion analysis.
    Chap.  15 I'M Environmental impact on rivers. Water
    Resour. Publications, Fort Collins, Col.

85. Wischmeier,  W.  H.  1972.  Erosion  and sediment
    research in the United States. Proc. Second Allerton
    Conf.  on Environmental  Quality  and Agriculture.
    Spec. Pub. No. 25, College Agr., Univ.  lll.:21-23.

86. Wischmeier,  W.  H.  1973. Conservation tillage to
    control water erosion. Proc. Nad. Conserv. Tillage
    Conf., Soil Conserv. Soc. Amer.: 133-141.

87. Wischmeier,  W.  H.   1974.  New developments  in
    estimating water  erosion. Proc. 29th Ann. Mtg., Soil
    Conserv. Soc. Amer.: 179-186.

88. Wischmeier,  W. H. 1975.  Estimating the soil loss
    equation's cover  and management factor for undis-
    turbed areas. In Present and prospective technology
    for predicting sediment yields and sources: 118-124.
    ARS-S40, U.S. Dept. Agr., Washington, D.C.

89. Wischmeier, W. H., Johnson, C. B., and Cross, B. V.
     1971.  A  soil erodibility  nomograph  for  farmland
    and construction  sites. Jour. Soil and Water Con-
    serv. 26:189-193.

90. Wischmeier,  W. H.,  and  Mannering, J. V. 1975.
    Effect  of organic matter  content of the soil on
    infiltration.  Jour.   Soil   and  Water   Conserv.
    20:150-152.

91. Wischmeier,  W. H.,  and  Mannering, J. V. 1969.
    Relation of soil properties to its erodibility. Soil Sci.
    Soc. Amer. Proc. 33:131-137.

92. Wischmeier, W. H., and Meyer, L. D.  1973.  Soil
    erodibility  on construction  areas. Highway Res.
    Board, Nat. Acad. Sci. Spec. Rep. 135:20-29.


93. Wischmeier, W. H., and Smith, D. D. 1958. Rainfall
    energy and its relationship to soil loss. Trans. Amer.
    Geophys. Union 39:285-291.
56

-------
94. Wischmeier, W.  H.,  and Smith,  D. D.  1961.  A
    universal soil-loss  estimating  equation  to  guide
    conservation farm planning. Trans. 7th Cong. Inter-
    natl. Soil Sci. Soc. 1:418425.

95. Wischmeier, W. H., and Smith, D. D. 1962. Soil-loss
    estimation as a  tool in soil and water management
    planning. Pub. No. 59, Internati. Assoc. Sci. Hy-
    drol.: 148-159.

96. Wischmeier, W. H., and Smith, D. D. 1965. Predict-
    ing rainfall-erosion  losses from cropland east of the
    Rocky Mountains-Guide for selection of practices
    for soil and water conservation. Agr. Handbook No.
    282, U. S.  Govt. Printing Off.,  Washington, D. C.,
    47 p.

97. Wischmeier, W. H., Smith, D.  D., and Uhland, R. E.
    1958. Evaluation of factors in  the soil loss equation.
    Agr. Engin. 39:458462,474.
98. Witmuss, H. D., Triplett, G. B. Jr., and Greb, B. W.
    1973. Concepts of conservation tillage systems using
    surface mulches. Proc. Natl. Conserv. Tillage Conf.,
    Soil Conserv. Soc. Amer.:5-12.

99. Woodruff, N. P.  1966. Wind erosion mechanics and
    control. Proc.  First Pan Amer. Soil Conserv. Cong.
    Brazil:253-262.

100. Woodruff,  N. P., and  Siddoway, F. H. 1965. A
     wind erosion equation. Soil Sci. Soc. Amer. Proc.
     29:602-608.

101. Young, R. A., and Wiersma, J. L. 1973. The role of
     rainfall impact  in soil detachment and transport.
     Water Resour. Res. 9:1629-1636.

102. Zingg, A. W. 1960. Degree and length of land slope
     as it  affects soil  loss in  runoff.  Agr.  Engin.
     21:59-64.
                                                                                                      57

-------
                                              CHAPTER 4

                    NUTRIENT ASPECTS OF POLLUTION FROM CROPLAND

                                               M. H. Frere
   Nutrients are naturally occurring chemicals essential
for plant growth. Sixteen elements are essential for the
growth and reproduction of most plants. Most soils are
lacking in adequate amounts of nitrogen, phosphorus,
and potassium. Hence, fertilizers containing these nutri-
ents are  essential to maintain the current  level of
agricultural production.  The  other nutrient elements
may be added as impurities in the fertilizer or to treat
specific nutritional problems. Present evidence indicates
that nitrogen and phosphorus are the principal nutrient
pollutants and,  therefore,  only  these  nutrients are
considered in this chapter.
   A major source of nutrients reaching water bodies in
this country is sewage, both from municipal treatment
plants  and  nonsewered  residences.  These are point
sources of pollution and extensive  efforts are underway
to limit their contributions. Runoff from rural land is
another major  source.  Unlike point sources, runoff
integrates the contribution of nutrients and water from a
wide variety of landscapes that are continuously chang-
ing with time. It must be recognized that nutrients leak
from the  system even when fertilizer is not applied and
while we  cannot eliminate nutrient losses, it is desirable
to minimize them.
   While a number of review papers have been written
about nutrient losses  (47,  76, 89, 98, 145, 167) the
emphasis of this  chapter is on the practices that can
control nutrient losses and the background necessary to
use these practices effectively. This chapter reviews the
literature and  summarizes  the data for nitrogen  and
phosphorus leaving cropland by  runoff, erosion,  and
leaching. The  material covered is limited to precipita-
tion-induced transport of the nutrients from cropland
and  improved  pastures.  Beyond  the  scope  of  this
overview are the very important problems of irrigation
return flow, wind erosion, and losses from waste disposal
areas.
   It  must be emphasized at the beginning that the
dynamic system under consideration is very complex.
The  wide variety of climates and landscapes provides
such a wide range of results that there is no typical case.
Complications  are introduced by difficulties in chemical
analysis   for  nutrients  in  water samples. Numerous
procedures  have been  followed for chemical analysis.
Changes in nutrient form can occur between the time
the sample is taken and when it is analyzed. Some of the
nutrient reported  as soluble could have been associated
with  colloidal  material  not  removed. The  practical
significance of these complications is unknown, but they
are noted  at  this  time to  warn  the reader  of the
limitations of  the data associated with nutrient pollu-
tion.
                                           THE PROBLEMS
   Two problems are associated with nutrients in the
aquatic environment: the water may be toxic to humans,
animals,  or fish  when the  concentration of certain
nutrient forms exceeds a critical level; and eutrophica-
tion may be accelerated.
                     Toxicity

   Case (28), Lee (92), and Winton (181) reviewed the
problems associated with nitrates and nitrites in drinking
water. The nitrite form of nitrogen, which is the most
toxic,  interacts  with  components  in  the blood  to
interfere with oxygen transport. Methemoglobinemia,
the technical name  given to this illness, is often called
"the  blue  baby syndrome" because  infants are  very
susceptible. Most of the problems  with  drinking water
have been  associated  with farm  wells with faulty well
casings and located close to manure concentrations such
as barnyards.
   Nitrate is  5 to  10 times less  toxic than nitrite and
healthy mature animals with single  stomachs are able to
                                                                                                      59

-------
excrete  nitrate  in the urine. Cattle, young animals, and
children convert some of the nitrate to nitrite in their
stomachs  and can develop  methemoglobinemia. Since
food  also contains nitrite and nitrate, the response  to
nitrate in  drinking water could be quite variable. The  U.
S. Public  Health Service  Drinking  Water Standards  of
1962 set the limit for nitrate at 10 mg N per liter (45
ppm nitrate). Armitage (5) reported the Recommended
Drinking Water Standards of the World Health Organi-
zation to be: 0-50 ppm nitrate = recommended, 50-100
ppm  nitrate  = acceptable. There is  some  concern,
however,  that  even  nontoxic   nitrate  levels (chronic
conditions)  may  lower  resistance  to  environmental
stresses and interfere with normal metabolism.
   Dissolved ammonia is another form of nitrogen that
can  occur at levels  toxic to fish.  Microorganisms can
generate  free  ammonia from organic  matter in  lake
bottoms during summer stagnation periods (164). Trout
are sensitive  to 1-2  ppm ammonia (35) while goldfish
appear to be less sensitive (48).


                  Eutrophication

   Eutrophication is the enrichment of waters by nutri-
ents and the ensuing luxuriant growth of plants. Much
has been written about this  subject  in the last few years
(100, 111,  153, 269).  Rapid  growth  of  algae is the
greatest and most widespread eutrophication problem in
most states (2). Algae can create obnoxious conditions
in ponded waters, increase water  treatment costs  by
clogging screens and requiring more chemicals, and cause
serious  taste and odor problems  (7 7). When a large mass
of algae dies and begins to decay, the oxygen dissolved
in the water decreases and certain toxins are produced,
both of which kill  fish (49). The  complexities of the
ecosystem are  illustrated by the observation that the
nutritional status of a species of algae can vary from lake
to lake or even from different areas and depths of the
same lake on the same day (39). Streams, however,  do
not age in the same sense as lakes, but their biological
productivity can be  increased by added nutrients (69).
For  example, phosphate from  farm land was a  very
beneficial and important factor  in the high production
of brook trout in Canadian streams (137).
   Aquatic plants require  a number of nutrients for
growth, but nitrogen and phosphorus appear to be the
ones  accounting  for  most of  the excessive  growth.
Sawyer  (128) concluded that eutrophication becomes a
problem when the concentration of inorganic nitrogen
exceeds about  0.3   ppm   and  inorganic  phosphorus
exceeds about 0.015 ppm. These concentrations of the
inorganic forms of nutrients are maintained by microbial
conversion  of  organic  forms so  the total  input  of
nitrogen  and  phosphorus  per  unit area of the lake
(loading rate) is important (57). Current international
guidelines for eutrophication control are 1.8 to 4.5 Ibs.
of P and 45 to 90 Ibs. of N per surface acre of lake per
year (] 69).
   The various roles of nitrogen in eutrophication have
been recently  reviewed (24,  53).  Aquatic  organisms
assimilate nitrate and ammonium. Ammonia and amino
acids are  excreted  by live organisms and released  by
decaying organisms. Fungi  and bacteria can convert the
organic nitrogen in  dead plant material and sediment to
ammonia  and  nitrate.  Whenever the environment  be-
comes anaerobic,  in  the  presence of  decomposable
organic matter, nitrates are denitrified to gaseous nitro-
gen  compounds.  Bouldin (20) estimates  that the  daily
loss  of the nitrate in the bottom sediments can be 7 to
15 percent by  microbiological denitrification and 2 to
28 percent of the ammonia by volatilization.
   Some additional inputs of nutrients are often  over-
looked, such as aquatic birds, leaves, dust, and pollen.
Another source is the  fixation of atmospheric nitrogen
into organic nitrogen by a number of organisms such as
the  blue-green algae. This process  is considered to be
adaptive in that it occurs when other sources of nitrogen
are depleted.
   Kramer et al. (86) and Lee (93) reviewed the role of
various  phosphorus compounds  in eutrophication. Sol-
uble orthophosphate is usually regarded  as completely
available for algal growth. Soluble organic phosphates
and  polyphosphates are probably not too available, but
are readily converted to orthophosphate. Finally, phos-
phate in  participate organic matter and adsorbed to
mineral sediments is usually only slowly released. The
adsorption  capacity of the  sediment  for phosphate
ranges from low for quartz sand to very high for certain
silicate clays.
   Sediment  low  in phosphate will usually  remove
phosphate from solution as it settles out (63, 64, 86, 93,
179). When the sediments have high P contents and the
environment around the  clay  is electrochemically re-
duced, then some of the phosphate can  be released to
the solution (86). This released phosphate may form the
mineral apatite, which is relatively insoluble (179), or if
there is a mixing process,  such  as caused by wind, the
phosphate is redistributed through the lake (63,178).
   Because a  lake's ecological system is so complex,
Shannon and Brezonik (132) devised an index of seven
parameters to  quantitatively characterize  the  trophic
60

-------
state of a lake. For  55 lakes  in  Florida, the  relation
between this index and the loading rates of nitrogen and
phosphorus showed that an increased phosphate loading
was statistically  the  more important. An additive form
of  the  loadings  accounted  for  60 percent  of the
variability in the index.
                                      SOURCES OF NUTRIENTS
   Fertilizers are well known as a source of nutrients on
cropland, but they  are not  the only source and some-
times are  not  the most  important source. Fertile
cropland soil is a pool of nutrients with different degrees
of availability to the crop and to the transport processes
of  runoff,  erosion, and leaching. Precipitation  and
animal wastes are  other sources. In  addition, legume
crops, with the assistance of microorganisms, can biolog-
ically convert atmospheric nitrogen into organic nitro-
gen.
   The relative importance  of each  of these sources
depends on a number of factors, such as the geographical
location with its relation to climate and  soils and the
crop management  practices previously and presently
used. Table  1  shows  the estimated  1969  national
nitrogen inputs (119) and estimates for several water-
sheds in Wisconsin  (13). The  national input for phos-
phorus in fertilizer is for  1973 (79), the manure input is
from Table  3, and  the inputs from plant residues and
precipitation were calculated from the N inputs and N/P
ratios of  7.5 (Table  17,  Vol.1) and  100 (Fig.  3),
respectively.
   On the national  level, fertilizer is a major input and
manure is a  relatively small input. The relative contribu-
tion of these sources  can  be reversed  for a specific
geographic location, as shown by the data for Wisconsin
watersheds.
                  Nutrient Cycles
Nitrogen
   Volumes have been written about nitrogen behavior.
Two comprehensive reviews are: "Soil Nitrogen" (10)
and "Soil Organic Matter and its Role in Crop Produc-
tion" (3). Figure  1,  adapted  from  Stevenson  (149),
illustrates the numerous compartments and pathways of
nitrogen.
   Most of the  reactions in the soil portion of the cycle
are microbial  and  thus  the  rates  are  sensitive  to
temperature  and  moisture.  Warm  (90° F) and moist
(water  in 80  percent  of  the  voids) are  optimum
conditions for cycling  within the soil. The conversion of
organic nitrogen  to nitrate (ammonification and nitrifi-
cation) is often called mineralization. A study of 39 soils
from across the United States  showed that the rate of
mineralization was proportional to the pool of mineral-
izable nitrogen. The  size of this  pool was not highly
correlated with the total organic matter or total nitrogen
(148).  Thus, some forms of organic matter are readily
converted to mineral forms whereas other organic forms
are not. Part of the stable  forms may exist as metal-
organic and organic-clay complexes.
   Soils contain 0.075 to 0.3 percent total nitrogen or
1,500 to 6,000 Ibs. per acre in the top 6 inches (5). Soils
                     Table 1.  Sources of nitrogen and phosphorus on a national and a watershed scale
Source
Fertilizer 	
Fixation 	
Manure 	
Kant residues 	
Precipitation 	
Total 	

National
Nitrogen
Million tons
6.8
3.0
1.0
2.5
1.5
14.8
Percent
45.9
20.3
6.8
16.9
10.1
Phosphorus
Million tons
2.2
0
0.4
0.3
0.01
2.9
Percent
76
0
14
10
0
Wisconsin watersheds
Nitrogen
Ibs/acre
10
12
42
45
8
117
Percent
8.5
10.3
35.9
38.5
6.8
Phosphorus
Ibs/acre Percent
8 32
0 0
12 48
5 20
0 0
25
                                                                                                         61

-------
o\
K>
                                     AIR
                                                  ADSORPTION
HARVEST

 PRODUCTS
                                    ANIMALS
                       PLANTS I
                       VOLATILIZATION
                                 RESIDUES
                                                                            EROSION
                  IMMOBILIZATION
           SOIL  ORGANIC

             MATTER
AMMONIFICATION
                                                                           CLAY AND

                                                                          ORGANIC COMPLEXES
                         IMMOBILIZATION
          LEACHING
                        NITRIFICATION
                                        Figure l.-The nitrogen cycle in agriculture.

-------
cultivated for 100 years can still release 30 to 60 pounds
of nitrogen per acre per  year and have drainage waters
with 5 to 10 ppm nitrate nitrogen.  Fertile soils in the
Corn Belt are estimated to release 120 Ibs. of N per acre
(7). In the semiarid west, some fields are  fallowed (kept
free  of  vegetation)  every  other  year to accumulate
moisture. The increased moisture also promotes mineral-
ization of 20 to 50 Ibs. of N/acre (37).
   When  ammonia (a gaseous form of nitrogen)  reacts
with the water it  forms  positively charged ammonium
ions. These ions are  held by the negative charge of the
clays as exchangeable cations. Some of the ammonium
ions can be trapped or fixed between clay  platelets.
   Ammonia  absorption  by  the soil has not been
considered in the past as a major path of nitrogen input.
However, in areas of high ammonia concentrations, such
as downwind of industries  or feedlots, the soil, lakes,
and  plants have  absorbed  from  20  to  70 pounds of
N/acre/yr (55, 67, 68,138).
   Immobilization, the reverse of mineralization, is the
part of the N cycle that converts nitrate and ammonium
into organic forms. It occurs under aerobic or anaerobic
conditions and  basically involves the  uptake  of the
mineral forms by microorganisms in the synthesis of cell
tissue. Whenever organic residues low in nitrogen are
being  decomposed,  mineral  nitrogen  must  be  used
because the carbon-to-nitrogen ratio  of microbial tissue
is on the order of 5-10:1.  Dead microbial  tissue then
becomes  part  of the organic matter pool that  can be
mineralized. A major problem in quantifying the immo-
bilization process is  that  it is  impossible, without
nitrogen  tracers, to  measure  the small amount of the
product in the large organic-matter pool.
   Denitrification is not well understood  but appears to
be a very important part  of the nitrogen  cycle affecting
environmental  quality.  Denitrification is the  use  of
nitrate by anaerobic microbes for oxygen and  results in
the production of nitrogen and nitrogen oxide gases. The
necessary anaerobic conditions are most prevalent when
the water content  of the soil is high, which is  the same
condition needed for leaching. Another requirement is a
supply of carbon for an energy source. Lack of useable
carbon may be  the factor that prevents all the  nitrate in
the leachate from being denitrified. Carbon is usually
not very mobile and, therefore, once the nitrate  passes
below the root  zone, the opportunity for denitrification
is limited (114). Complete waterlogging is not essential
for denitrification. Since the soil contains a wide range
of pore sizes, an unsaturated soil can have areas where
the water contents and microbial activity are  sufficient
to produce  an  anaerobic environment and denitrifica-
tion.  Quantification  of this process has  been limited.
Most of  the  estimates  have been  based on nitrogen
budgets where all unaccounted-for nitrogen  is assigned
to denitrification. The nitrogen gases produced are very
difficult to measure under field conditions.
   Average  losses have  been estimated  at   10  to 30
percent of the total  yearly mineral nitrogen input (26).
When excessive rates of nitrogen arc applied, as much as
50 percent can be lost (99).
   The inputs of fertilizer, biological fixation, animal
wastes, and precipitation; and the  losses by leaching,
runoff and erosion, and plant uptake (avoiding excessive
fertilizer use) will be covered in subsequent sections.

Phosphorus

   The phosphorus cycle shown in Figure 2 is a lot less
complicated  than the nitrogen cycle,  although  it has
some of the same paths. Before considering  the similar
paths, we will examine those reactions in  the  soil which
are  unique for  phosphorus. Black (7.5), Olscn  and
Flowerday (775) and Ryden et al. (725)  have prepared
comprehensive reviews of the subject.
   The phosphate concentration in  the soil  solution is
low, usually 0.01 to 0.1  ppm P, although  the total P in
the soil ranges from 100 to 1,300 ppm (73). The mineral
forms of phosphorus—calcium,  iron,  and  aluminum
phosphates-have very low solubilities and  the phosphate
is highly adsorbed to clay minerals. The organic forms of
phosphate have  not  been studied as extensively as have
the inorganic  forms. However, the organic part of the
total phosphorus can range from 3 to 75 percent.
   Because of the low solution concentrations and high
degree of adsorption, phosphate  tends  not  to  leach.
After 286 Ibs. of P  had been applied in  11  years, the
plant "available"  level had increased only 18 Ibs. and
very little had moved  below  12 inches (32). After 82
years of fertilization, the total P had been doubled, but
no  added P  was found below  54  inches  (76).  The
availability of phosphate in the soil  decreases exponen-
tially with  time.  However, soils vary greatly in  their
conversion of added P to insoluble  forms (90). Recent
work (52) indicates  that chemical reactions immobili/.e
more than 50 percent  of added soluble phosphate in a
few hours and an additional 10 percent in a month or so.
The  amount of biological immobilization into organic
phosphates that occurs simultaneously with the chemical
reactions  depends upon the amount of biological activ-
ity.

                   Precipitation

   The amounts  of nitrogen and phosphorus added in
precipitation are  generally low and, for cropland, they
                                                                                                          63

-------
                                 HARVEST
                                  PRODUCTS
                                              ANIMALS
X FERTILIZERS
                                    RESIDUES
                                                                            EROSION
               ADSORPTION
                                     IMMOBILIZATION
                                                     SOIL  ORGANIC
                                                       MATTER
   /
MINING
               DESORPTION
                                         MINERALIZATION
                              Figure 2.-The phosphorus cycle in agriculture.

-------
are negligible in comparison to other inputs. On forests,
unimproved pasture and rangeland, and lakes, the input
of nutrients added  by precipitation can be significant.
The concentrations of nitrogen and phosphorus in rain
and snow vary not only across the country, but within
short distances and during a storm.
   Uttormark  et  al.  (166)  list 40  factors  that  can
influence the concentration of nutrients in precipitation.
Feedlots, industrial urban centers, power plants, disposal
sites, etc., are all relatively local sources that can increase
the general regional level of nutrients. They prepared a
map of the United States delineating areas of different
nitrogen  contribution  in rainfall. The Lake States had
the highest contribution, about 2 to 3 Ibs. of N/acre/yr.,
whereas  the  Western  States had  less than  1 Ib. of
N/acre/yr. Dry fallout was not included.
   Seasonal  maps of  ammonium  and nitrate  in  rain
across  the  United  States  (73) show  that the  highest
concentrations are in the spring and summer. It appeared
that  the  rainfall concentrations might be  related to the
soil because  the  Southeast acidic soils are  the lowest.
Ammonium  concentrations  may  change as much as
10-fold during the year, but nitrate changes  much  less.
  When  rainfall  samples are  taken at different times
during  a  storm, the nitrogen concentrations  are often
slightly higher during the first part (24, 50). This could
be due to evaporation from the drops into dry air or the
washout of dust.
  The phosphate input in most places may be associ-
ated with dust, either  as dry fallout between  storms or
washed out of the atmosphere with  rain. Dust or dry
fallout is a very important  component associated with
the precipitation  input of nutrients.  It has been  esti-
mated  (166) that in  arid  regions 70 percent of the
nitrogen  in  uncovered rain  gages is from dust. Storms
blowing in from  oceans are  quite low in phosphorus.
Phosphorus can also be associated with ash and smoke,
as indicated in concentrations of 0.24 ppm phosphate in
rain at Cincinnati compared to threefold  less at a rural
location near Coshocton, Ohio (174). A yearly input of
0.2 to  0.6  Ib. P/acre/yr has been estimated (24, 110).
One  problem associated with evaluating the importance
of dust is the absence of a measure of the loss by dust
into the atmosphere. There is no evidence that snow has
any  different concentrations  of nutrients  than  rain
would at  the same time of year.

                     Fertilizers

  Fertilizer  use  to improve crop  yields dates from
antiquity. Sanskrit writings of 3,000 years ago note the
value of  dung for fertilizer (103). Guano and Chilean
nitrate  were available in Europe as a fertilizer over  100
years ago. Yields from field  tests started at England's
Rothamsted  Experiment  Station  in  1840  have been
maintained near  maximum by  the use of manure and
fertilizer while those from unfertilized plots have de-
creased to an uneconomic level.
   Organic forms of nitrogen  were the cheapest sources
of nitrogen until about  1900 (70). Before the 1950's,
sodium nitrate and ammonium sulfate were the principal
nitrogen  sources. Then ammonium nitrate became the
leading source,  only  to  be  surpassed by  anhydrous
ammonia  and urea in the 1960's. Phosphate fertilizers
also  changed in the 1950's from normal superphosphate
to concentrated  superphosphates.  These  are  trends to-
wards the use of more concentrated forms, thus reducing
shipping  charges  per pound of  nutrient.  The source of
the nutrient makes no difference to the plant because of
the extensive cycling in the soil. Some sources, such as
ammonium sulfate, can increase the soil acidity if used
for extended periods.
   Table 2 shows how fertilizer  use on four major crops
has changed over a 10-year period. The acreage of corn,
wheat, and  cotton  harvested has remained relatively
constant except  in  1974  when acreage  controls were
removed. The acreage of soybeans has steadily increased.
Except for cotton, the percentage of acres fertilized has
consistently increased over the years. Note also that the
yield per acre has tended  to increase. Fertilizer rate has
increased  except  for nitrogen  in 1974 when  short
supplies and increased costs because of the energy crisis
caused many farmers  to  reduce their application rate.
The  relative plateau of fertilizer use on cotton deserves
some comment.  Cotton is relatively sensitive to nitrogen
and  excess nitrogen  can reduce  yields. Cotton has been
fertilized intensively for a long  time and  the optimum
rates have evidently been found. Pests, such as insects,
weeds, and disease, are probably limiting yield  more
than fertility.
   Speculation on future fertilizer use is fraught with
uncertainties.  The problem is  basically economics. How
much fertilizer should the farmer apply to maximize his
profits? The  yield response to fertilizer is less for each
subsequent  increment of fertilizer  and  as  the  yield
increases some other costs also  increase. In the 1960's,
farmers  received  $2.50  for  each  dollar invested  in
fertilizer  and so  they substituted  an investment  in
fertilizer  for  additional land or  labor (62). Real estate
costs and wages increased over  250 percent since 1950
while  plant  nutrient costs decreased  (38). While  the
energy crisis  will tend to increase  fertilizer costs  faster
than other costs, it will probably be sometime before
they reach the level of other farm costs. Thus, farmers
will  probably continue to  fertilize,  but  will  carefully
appraise  the  size of the  applications  and eliminate
                                                                                                          65

-------
                        Table 2.  The change in fertilizer use on four crops in the past 10 years
Crop
Corn 	 ....
Wheat . . 	
Soybeans . 	
Cotton 	 ....

Year
1964
1966
1968
1970
1972
1974
1964
1966
1968
1970
1972
1974
1964
1966
1968
1970
1972
1974
1964
1966
1968
1970
1972
1974
Acres
harvested1
Millions
55.4
56.9
55.9
57.2
57.4
63.7
49.8
49.9
55.3
44.1
47.3
64.1
30.8
36.5
41.1
42.1
45.7
52.5
14.1
9.6
10.2
11.2
13.0
13.1
Area fertilized2
N
82
91
92
94
96
94
47
49
56
61
62
66
6
17
21
21
22
22
75
75
73
72
77
79
P
Percent
75
85
88
90
90
87
36
38
43
44
44
46
10
24
27
27
29
28
56
58
55
48
55
58
Fertilizer rate2
N
Ibs./ac.
45
83
102
112
115
103
28
32
37
39
46
46
13
14
12
14
14
15
69
77
71
75
75
78
P
18
24
28
31
29
27
12
14
14
13
16
17
11
15
17
16
18
18
21
23
23
24
24
23
Yield1
bu.lac.
63
73
80
72
97
26
26
28
31
33
23
25
27
27
28
Ibs.lac.
517
480
516
438
507
   1 United States Department of Agriculture, "Agricultural Statistics", Years 1964 to 1974.
   2 1964-1970 data from "Cropping Practices" SRS-17, Statistical Reporting Service, USDA.
    1972-1974 data from "Fertilizer Situation", FS-5, Economic Research Service, USDA.
excessive  use. A food crisis  that increases the price of
farm products relative to fertilizer costs would stimulate
fertilizer  use. Fertilizer use in the United  States  will
increase 5 percent per year because of rising populations,
improved diets, and  increased exports, according to a
recent estimate (78).

                   Animal Wastes

   Before commercial fertilizers came into common  use,
animal manure supplied most of the nutrients added to
the soil. In the 1960's it  was more expensive to load,
haul manure several miles, and spread it than it was to
purchase and apply commercial fertilizer. In addition to
the economics, the convenience of commercial fertilizer
caused animal wastes  to  become a disposal problem
rather than a nutrient source.
   The following discussion  is based  on  information
from several references (29, 30, 96,  170,180).
   Manure is the excrement of animals that contains the
undigested food and the urine. The nutrient content of.
manure is different for different animals, type of feed,
and amount of water consumed. In addition, the amount
of bedding used to  absorb the urine or presence of
superphosphate to react with the ammonia, the method
of storage, and the duration of storage influence the
nutrient content of manure being applied to the soil.
   Fresh manure contains 50 to 90 percent moisture, 0.2
to 6 percent total nitrogen, and 0.06 to 2.5 percent total
phosphorus, on a  dry weight  basis. The  old  rule of
thumb  was that the typical ton of moist cow  manure
contained about  10 Ibs. of total nitrogen and  1 Ib. of
total phosphorus  when it was applied to the  field.
Present methods of confinement, feeding, and  manure
handling vary  enough  that this rule of thumb is no
longer adequate.
   Yeck et al (183) recently estimated that 5.8 million
tons of nitrogen are  excreted annually by  livestock in
the United States. The percentage that can be collected
varies from zero for cattle on the range to nearly 100 for
caged  poultry. The work of a number of investigators
indicates that about half the nitrogen collected is lost
during storage, handling, and spreading before it can be
incorporated into the soil. Thus, of the 2.4 million tons
66

-------
of nitrogen that  they estimate can be collected, only 1.2
million tons are available as a substitute for fertilizer and
a large part of this is already being applied.
   The pollution potential isn't a result of animals per
se, but  of the manure they produce.  Since there are
maps showing the geographical distribution of animals
(Figs. 20-24, Vol. I), we needed a method for conversion
to manure and  its nitrogen and phosphorus contents.
From the  collectible  nitrogen previously described for
each  class  of  livestock,  a  50 percent  loss  before
incorporation was  assumed. This can then be geographi-
cally  distributed with  the  animals.  To calculate  the
amount of manure associated with nitrogen one must
assume  some percentage  of N in  the manure  being
spread.  The amount  of phosphorus  can  likewise be
calculated from  the percent phosphorus. .These percent-
ages of  N and P in the manure were estimated for the
different  kinds  of animals  from data  of various re-
searchers. These calculations are summarized in Table 3.
   Not  all of the nitrogen  in  manure  is immediately
available. Most of  the nitrogen left in manure when it is
incorporated into the soil is in organic compounds. Like
soil organic matter, microbes must mineralize the or-
ganic nitrogen into ammonium  and nitrate, which are
taken up by the plant. About 40 or 50 percent of the
organic nitrogen is mineralized during the first cropping
season (772). Additional amounts are released in subse-
quent years so that yearly applications can  build up the
nitrogen-supplying capacity  of  the soil. Of course this
"slow release" of nitrogen also means that some of it can
be released when plants are not rapidly growing (fall and
spring) and thus  be available for leaching.
          Biological Fixation of Nitrogen

   One of  the  most  recent  reviews of the biological
fixation  of nitrogen is  by Allison  (3).  Another  good
source of information is three chapters in "Soil Nitro-
gen"  (10).  Nonsymbiotic (free-living) organisms  prob-
ably-fix  about  10 Ibs.  of N per acre annually. This
amount is of little practical importance in the nitrogen
balance of a cultivated field,  but since it is comparable
to  the  precipitation contribution,  it  is  important  in
nonfertilized grass sods.
   Nitrogen-fixing bacteria in a symbiotic relation with
roots   of  certain  plants, principally  legumes,  can  fix
sufficient  nitrogen to support  a grass-legume  pasture.
With low soil nitrogen, the amount of nitrogen fixed in
effective legume nodules correlates closely with the dry
weight of  the legume  tissue  produced. Generally, the
fixed nitrogen is produced only as the plant needs it and,
therefore, plants with  poor  growth  will not fix much
nitrogen.  Similarly, high soil nitrogen levels,  such  as
from  fertilization,  will reduce the amount of nitrogen
fixed  because it is not needed for plant growth. There is
some  indication  (12) that fresh organic matter  from
previously  fertilized crops stimulates N fixation. Biggar
and Corey (13) report 200 Ibs of N per acre per year as
being fixed in legume systems. Allison (3) cites several
references  where  150 to 300 Ibs. are common and  as
much  as  600 Ibs.  of N/acre/year  was observed.  He
concludes that a grass-legume mixture may fix more N
per acre  than  legumes  alone  because the  grass will
continually  remove  any nitrogen mineralized from the
soil or plant material.
            Table 3. Estimates of nitrogen and phosphorus in manure that is available for application to cropland
Animals
Beef cattle 	
Dairy cattle 	
Swine 	
Laying hens 	
Broilers ....

Totals 	

Vol.1.
Figure
No.
21
22
23
24
25



Collectible N1
Thousand tons
650
670
600
250
240

2410

Available N2
Thousand tons
330
330
300
125
PO

1 205

N content'
in manure
Percent
2.5
2.0
2 8
4.5
38



Available
manure
Million tons
13
17
M
2.8
3.T

47.0

P content3
in manure
Percent
0.8
0.6
1.0
1.7
1 3



Available Ps
Tltoitsand tons
100
100
110
50
40

400

   2 Estimated amounts of nitrogen that can be collected (183).
   3 Available after application assuming 50% losses during handling, storage, and spreading.
   4 Estimated contents based on data from various authors.
   5 Calculated from available N and N content.
     Calculated from available manure and P content.
                                                                                                            67

-------
                                   TRANSPORT FROM CROPLAND
   Water pollution by nutrients from cropland involves
one  of three transport processes:  leaching, runoff, and
erosion. In nature the results of these processes are not
easily distinguishable. Water may infiltrate into the soil
and  thus  cause  leaching,  but a  few  feet  or yards
downslope the water with its dissolved nutrients may
come to  the surface  and  join the  overland flow or
runoff.  Similarly, the distinction between runoff and
erosion  may be  quite difficult when nutrients in the
runoff are  adsorbed on  or released from the eroded
sediment that the runoff water is carrying. The larger the
drainage area of the stream sampled, the more mixed are
the  three transport  processes.  Some  cases where  these
processes operated independently will be examined so
that the system can be better understood.
   The  usual range of values is presented in Figures 3
and  4.  Occasionally,  more  extreme  values  will  be
observed because the system is highly variable. Concen-
trations are not provided for sediment transport because
the  concentrations  of  sediment  vary so much.  Soils
contain 0.075 to 0.3 percent nitrogen and 0.01 to 0.13
percent phosphorus and  deposition of coarse material
during transport can increase  the concentrations in the
transported sediment by 2- to 6-fold.

                      Leaching

   Leaching is the process whereby soluble chemicals are
dissolved  and removed from  the soil in water that is
percolating through  the  soil.  Nitrate is the  principal
nutrient form  found in drainage waters because it  is
seldom adsorbed to the soil minerals. Some red subsoils
in the southeast are an exception (156). Organic forms
of nitrogen and phosphorus and the orthophosphate ion
are  seldom  found  to  any extent in  drainage water
because they are held by the soil.
   Three  principal ways of evaluating leaching are by
lysimeter studies, monitoring tile drains, and taking core
samples. Each  method  has some limitations and it  is
important to recognize these limitations when interpret-
ing the data.

Lysimeters

   Drainage lysimeters are columns of soil  in the field
isolated  by impermeable cylinders  with facilities for
collecting the water  that  drains out of them. Cylinder
walls at or above  the  surface can prevent or reduce
runoff and, therefore, increase the amount  of leaching.
Also, if suction is not applied, the bottom of the column
must be  saturated  before  drainage will occur.  This
saturated zone can provide an opportunity for denitrifi-
cation and  dilution. Some lysimeters do not contain
undisturbed profiles  while others are very shallow. Both
conditions limit  their usefulness for field interpretation.
   Lysimeters have provided some valuable information
about leaching. Kolenbrander (55) found that as the clay
content of the soil increased from  10 to 50 percent, the
nitrogen loss decreased  from 40 Ibs. N/acre/year at  18
ppm to 4 Ibs. N/acre/year at 2  ppm. Wild (177) found
that nitrogen mineralized in the soil didn't leach quickly
in fine soil with  cracks, but Kissel  et al. (80) found that
applied fertilizer, simulated  by  chloride, did move
quickly through the  cracks of a clay soil. Kolenbrander
(&5) reported phosphorus losses of 0.2 Ib. P/acre/year at
0.08 ppm for both  cropped and grass lysimeters,  with
and without fertilizer.

Tile Drains

   Tile  drains  have recently  provided   most of  the
measurements of nutrient leaching. While tile drains
sample  a  much larger  area and thus provide a more
integrated value, they may not accurately reflect the
nutrient content and water volume leaving the field. One
problem is to define the boundaries of the drainage  field
so that losses can be  calculated on an area basis. Another
is that the tiles short-circuit the drain paths with aerated
conditions  reducing  the  time and opportunity  for
denitrification (757). As an illustration of this condition,
Thomas and Barfleld (157) monitored an area where
one-third  of the water flow was  from  tile lines  with
nitrate levels of  15 ppm N while the rest of the seepage
had only  3 ppm   N.  At  lower  flows,  the  seepage
accounted for nearly 90 percent of the flow with nearly
zero nitrate while tile lines had 10 percent of the  flow
with concentrations of 9 ppm N.
   Many  studies have  been  reported concerning  the
concentration and the average annual loss per unit area
of nutrients in drainage waters (83, 96, 125,  166). The
first feature to be recognized is the  extreme variability in
the data, both the loss per unit area and  the concentra-
tion. Thus, an average value has little utility. One of the
major factors causing the variability in the loss data is
that the water  flow, and thus the load  of  nutrients
transported, varies greatly between  dry  years and wet
years. The type  of crop grown is another major factor.
Nutrient concentrations  are  consistently lower in the
drain water from grasslands and woodlands than in that
from cropland. Grasslands and woodlands have a longer
68

-------
                                CONCENTRATION-ppm






O.QI	QJ	LQ	10.0	IQO.O
                         PRECIPITATION         Ezzzzzzzzzzza N
       CROPLAND RUNOFF     Pi               *--• ^tu (11 tutui n«\H
                        ¥// ///////////// ////////////J
N     NON-CROPLAND RUNOFF
                     P'                      i       DRAINAGE
                           Figure 3.-Range of nitrogen and phosphorus concentrations in different waters.

-------
                              SPATIAL  RATE - Ibs./acre/year
0.01
O.I
1.0
10.0
100.0
           PRECIPITATION
                                               N
                          r/ / / / / /
                                      / / / / / / / // / / // / // / //// // // / / // / / / / / / 1
                                               N  CROPLAND RUNOFF
                                                                   NON-CROPLAND RUNOFF
                        PC
                      DRAINAGE
                i /////////////
                                                                                     N
       NON-CROPLAND
          SEDIMENT
                       CROPLAND SEDIMENT     [//////////////////////////////////TTT*
                                                P
                        PC
            i// / / //////// (/ / / // / / ////i i
                           figure 4.-Range of spatial rates of nitrogen and phosphorus in waters and sediments.

-------
growing season in which to remove nutrients and reduce
water flow. Also, even though they are located  on soils
of  lower  fertility,  these  lands are seldom  fertilized
because the economic  return is insufficient.  Generally
then, the fertile land  that  is  cropped will  have the
highest concentrations in the drain water and the most
drain water. Fertilizer  applied  to these lands, tends to
increase  the concentration  of nitrogen in  the drain
water, but seldom that of phosphorus.
Soil Cores

   Cores of soil can be taken from plots and fields, then
separated  into segments  and analyzed. This provides a
means  to  follow the movement of the  nutrients down
through the soil profile.  In most soils, the distance the
band of applied nitrate moves down through the soil is
proportional to the amount  of water in excess of that
required to raise the water  content to field capacity (84,
94). Water in excess of crop needs and soil properties
such as texture and structure are important factors in
determining the rate of leaching.  In Wisconsin, Olsen et
al.  (114)  estimated the  annual percolation below the
root zone to be 6 inches.  A silt loam soil has a water
capacity of 30 to 50 percent, thus nitrate could move 12
to  18  inches  per  year under  these conditions and  it
would  take at least  20 years  to reach a water table at 30
feet.
   In the semiarid Great Plains, leaching  is not a hazard
except when additional water is  added by irrigation or
fallow  land prevents normal transpiration by plants.
North  Dakota soils wetted  to 6 feet only occasionally in
40  years  (121), but in Colorado,  nitrate accumulated
below  the root zone in a wheat-fallow system (150). The
average nitrate content in a 20-foot profile was 260  Ibs
N/acre for cultivated dryland compared to 90 Ibs N/acre
in native  grass. Since very little  fertilizer is applied to
dryland crops in eastern Colorado, the difference be-
tween  these averages  is probably  a result  of nitrate
production and leaching under cultivation and  fallowing.

                      Runoff

   Runoff occurs when the rate of precipitation exceeds
the rate of infiltration. A heavy mulch apparently acts as
a sponge to hold water as do the numerous small surface
depressions. Surface-applied  fertilizers can dissolve into
this water held  in  the depressions and by the mulch.
When  this water becomes runoff  it carries a load of
nutrients.  This is  why  the  highest concentrations  in
runoff occur when there  is runoff soon after surface

-------
because there is more  plant cover more of  the time.
Overgrazed lands have higher nutrient losses because of
increased  runoff and erosion. Concentrations of nutri-
ents will also be high occasionally if animals have direct
access to a stream, if animals are fed  near streams, or if
fertilizers and manures are surface applied.  Phosphate
concentrations in runoff may be higher from grasslands
than  from  adjacent  cropland   after harvest because
freezing and drying cause a release of nutrients from the
vegetation (163, 176). Timber harvesting also causes a
sudden release  of nutrients to runoff waters from the
decay  of the trimmed foliage (18). The reduced water
consumption after  harvest also  increases  the  annual
nutrient loss. Uttormark et al. (166)  and Brezonik (24)
provide a summary  of losses from land in various uses.
Ryden et al.  (125) provide a  recent review of phos-
phorus losses.

                     Sediment

   Sediment  is  the  major transport vehicle  for phos-
phorus and  organic  nitrogen.  Raindrop  splash  and
overland flow of water detach soil particles containing
adsorbed phosphorus and associated organic matter. The
flowing water  transports the  particles off the field.
Transport capacity depends primarily  on the volume and
velocity of water flow. Whenever the velocity is reduced,
such  as  by a  flatter slope, the transport  capacity  is
reduced  and any sediment in  excess  of the reduced
capacity settles out. Since the larger and heavier particles
settle out first, the remaining sediment contains a larger
percentage of the finer particles. The finer particles have
a  higher  capacity   per  unit  of sediment  to  adsorb
phosphorus and, also, organic matter is lighter and tends
to  be associated  with  the  fine particles.  Thus, the
transported sediment is richer in phosphorus and nitro-
gen than the original  soil (59).
   Bedell  et  al.  (11) found  that 98  percent  of the
sediment samples from 2- to 4-acre watersheds contained
as  much  or more  organic  matter, phosphorus,  and
nitrogen  than the soil. The  degree  of enrichment has
often been described by an enrichment ratio, the ratio of
the  nutrient  concentration  in  the  sediment to  its
concentration  in the. soil  of the watershed.  Recently,
enrichment  ratios of 2  to 6   have been  found for
phosphorus (130,155). Barrows and Kilmer (9) reported
an average enrichment ratio of 2.7 for nitrogen and 3.4
for "available" phosphorus. Available means  that frac-
tion  by  chemical extraction  that is expected to  be
available to plants.
   Massey et al. (104, 105) found that the enrichment
ratio  was inversely   related  to  the  concentration  of
sediment in the runoff and the total amount of sediment
lost. Doty and Carter (34) found that when the sediment
concentration was highest  at  peak flow, the chemical
and physical composition of the sediment was similar to
that of the soil. At lower flows, the sediment concentra-
tion was lower, the  percentage of clay in the sediment
increased, and the enrichment ratio increased.
   The loss of total N and P in sediment from cropland
ranges from 1 to  50 or  100 Ibs. per acre per  year. Two
factors can contribute to the  high  loss  from cropland:
the soil is fertile  and contains a lot of nutrients per
pound of soil, and tillage operations often leave the soil
very susceptible in the  most  erosive  part of the year.
Noncropland areas often have a lower nutrient content
than the cropland and also lower erosion rates because
of  more  vegetative  cover. An exception  is the  large
amounts of sediment produced by gullies that are often
prevalent in noncropland.

     Nutrient Losses from  Large Watersheds

   As  indicated  at the  beginning of this  section, the
nutrient composition of a  stream reflects a mixture of
the three  basic transport mechanisms. The composition
of streams changes with amount  of flow, season of the
year,  and distance down the stream as  new material is
added from tributaries, seepage, and outfalls.
   The Environmental Protection Agency (165) is under-
taking a national  eutrophication  study, but  at present
only a preliminary analysis is available. The data indicate
less variability in  total nitrogen  loss   than in  total
phosphorus loss, and increases in the  concentrations of
nutrients  in  the  water  are correlated  with increased
density of animals in the watershed. A recent summary
of the spatial loss  for streams (166) reports a range of 1
to  13  Ibs. N/acre/year and 0.03  to 2 Ibs. P/acre/year,
with averages of 5 and  0.4. Both studies indicate that
forest  land yields less nutrients  than agricultural land
and that the Midwestern states have higher losses than
other  states.
   Most perennial  streams have a  number of outfalls for
municipal and industrial wastes, which complicates any
budgeting of sources. For  example, agricultural land is
estimated to  have  contributed all of the inorganic N, 49
percent of the total  N, and 13 percent of the total P to
the Potomac  River  in  1966 (71). The  figures for the
agricultural contribution to the Hudson River were 37
percent of the total N and 27 percent of the total P.
   Streams are dynamic systems with living organisms
assimilating the nutrients  and particulate  matter ad-
sorbing or releasing the nutrients. Some  of the nutrients
removed from solution may go undetected in the debris
moving along  the  bottom of the stream or floating on
72

-------
the surface. Phosphate-deficient  sediments will  remove
phosphate from the solution {154). Thus, the concentra-
tion of phosphate  in a stream changes as soil is added
from noncropland and stream banks (88, 173) or gullies
(130).  Keup (75) provides a good discussion  of the
behavior  of phosphorus  in flowing streams. He cites
studies on two rivers where the overall loss from solution
was logarithmically related to the distance of streamflow
below a point source of phosphorus. The coefficient was
0.01  to 0.02 per mile; that is, 14 to 30 miles for a 25
perceni reduction.
                                 EFFECT OF CONTROL PRACTICES
   There is  every  reason  to  believe that  nutrient loss
from cropland can be controlled at an acceptable level if
proper management practices are used.  As pointed out
by Klingebiel (82), soil surveys are an important basis
for planning the optimum use of each field. By intensive
use of fields with high crop production potential and
low hazards from runoff, erosion, and leaching, the use
of marginal  land  with  higher hazards can often  be
reduced.
   The effectiveness of management practices  for the
control  of nutrient losses has  not been quantitatively
evaluated  to the same  degree as  have  the impacts on
runoff and erosion. Only in  recent years have adequate
data on nutrients been collected.
   The  possibility  of  creating another  problem  by
solving one  problem should  be the concern of all who
make recommendations. The nitrate contamination of
ground water in Runnels County, Texas (87) can be used
as an example of this  possibility. This land, which had
been dryland farmed since 1900, had nitrate formed and
leached  below the root zone but not down to the  water
table. Extensive  terracing after the drought in the early
I950's increased water retention and leached the nitrate
on down  to the water table. Thus, a nitrate leaching
problem  was created by the terracing done  to solve a
problem  of limited moisture. While hindsight is  much
clearer than  foresight, we should  learn from previous
experiences  and examine  our recommendations  for
secondary effects.

      Erosion and Runoff Control Practices

   Sediment is a major pollutant in itself. That it also
carries nutrients  and pesticides means that the first goal
in controlling pollution from  cropland should  be to
-:>ntrol  erosion.  For example, more  than 97 percent of
'lie N and P lost from some watersheds was associated
with sediment  lost  primarily in  the 2 months after
1'lanttng (27).
   Soil conservation practices have been stressed  for over
40 years  and sufficient data  have been  collected to
permit fair predictions of average  annual  soil loss. Old
punciples are continually being  used  to create new
methods of control (140). Data on the effectiveness of
these practices for controlling nutrient losses  are ade-
quate  for  only qualitative  predictions.  Generally,  a
reduction in sediment loss provides less of a reduction in
nutrient loss.

               Residue Management

   The greater the amount of residue left on a field, the
greater the reduction in  erosion. Zwerman et al. (184)
reported that  leaving  the  crop residues  instead  of
removing them decreased runoff 50 percent and did not
change the nutrient content of the runoff water. Thus,
losses of nutrients in runoff were reduced 50  percent.
Losses of nutrients with sediment were  also reduced.
Romkens et al. (124) used simulated rain  and observed
that  several conservation tillage  systems  reduced sedi-
ment loss but increased  the loss of soluble nitrogen in
the runoff. Ketcheson and Onderdonk (74) found that
covering broadcast fertilizer  with a chopped cornstalk
mulch reduced soil phosphorus losses 65 percent and
fertilizer losses 97 percent.
   No-till or zero tillage (7) is one of the most effective
practices for  reducing erosion.  The  effect  of these
practices on the amount of runoff is variable; sometimes
runoff is increased and sometimes it is decreased. Smith
et al. (136)  reported  that nitrogen in  runoff  was not
greatly affected by  the  no-till practice, whereas phos-
phorus increased 5- to 8-fold, probably from leaching of
residues. Since runoff is sometimes decreased, nitrate
leaching  can be increased (158). Schwab et al. (131)
found little difference in the  nitrogen and phosphorus
content of tile drainage from conventional  and no-tillage
plots.

Cropping

   Sod  reduces  runoff and permits very  little  erosion.
Therefore,  on an average annual basis, the rotations with
sod  should show a reduced nutrient loss. Results from
corn-wheat-clover  plots  (102,  134) indicate that the
rotation  reduces the total N and P loss 3- to  6-fold
compared to continuous corn or wheat. Schuman et al
                                                                                                         73

-------
 (729, 130) reported that a sod pasture lost about 10-fold
 less nitrogen than  continuous' corn.  The  phosphorus
 losses from  the pasture were lower by a factor of two,
 even though the concentration  in the runoff and on the
 sediment was higher. A higher concentration  of phos-
 phorus  in   the  runoff,  particularly  snowmelt,  from
 pasture  or   hay lands  has been reported by several
 workers (27, 161,  176).
   When little or no residues are left on a field, as when
 corn is harvested  for  silage, then planting a cover crop
 such as a small grain will protect the soil during the
 winter.  Smith  et  al. (136)  recorded  a  50  percent
 reduction in runoff and a 40-fold reduction in  losses of
 sediment, total nitrogen,  and  total phosphorus when
 corn was planted into a ryegrass cover crop. They found
 little effect  on the soluble nutrients lost in runoff, which
 were already low.

 Supporting Practices

   Several erosion control  practices such  as contouring
 and terraces are physical rather than agronomic. They
 can be  used by  themselves with regular cultivation or in
 conjunction with reduced tillage systems to achieve even
 greater  reductions.  Bedell  et  al.  (11)  reported  that
 contouring  a corn-wheat-meadow rotation reduced sedi-
 ment, nitrogen,  and phosphorus losses  from 3- to 5-fold
 on all crops of  the  rotation. Schuman et al. (729,  130)
 found  terraces   reduced  water, sediment, and  total
 nitrogen  losses  a little  over  10-fold  and  phosphorus
 losses a little less than  10-fold compared with contour
 tillage.  The  enrichment  ratio doubled  from 2 to 4 and
 the phosphorus concentration in solution doubled.
   Farm ponds  are  constructed for  a variety of reasons
 such as stock watering, recreation, etc. They are also an
 effective trap for sediment and  nutrients (77 7). In some
 soils, the ponds are difficult to seal and a local seepage
 problem can be created.

          Nutrient Management Practices

   Erosion control practices will probably solve most of
 the  phosphate  pollution  problems and  many  of  the
 nitrogen pollution  problems. These practices will have
 less  effect on controlling nutrients dissolved in runoff
 than in  sediment. They  have no effect and may even
 aggravate a nitrate leaching problem. In these cases, it is
 necessary to use alternative or  additional practices to
 achieve the  desired  degree of control. These  practices
 involve  changing the use of nutrients. Table 4 contains a
 list of these practices and some of the references used in
 the following discussion.
Eliminating Excessive Fertilization

   For  preventing  nitrate leaching,  Olson (776)  sug-
gested that  only enough nitrogen be applied to satisfy
the crop  needs,  that the soil's capacity for producing
nitrate be accounted for, that the nitrate already present
in  the  root zone be  taken  into  account,  and  that
adequate  levels  of other  nutrients be supplied so  that
there is maximum  efficiency. These suggestions can be
reduced to a single concept of eliminating excessive use
of fertilizer.
   Nitrate builds up in the soil when excessive levels of
nitrogen  fertilizers  are  used  (750),  But  .when   only
adequate  amounts  are used at  the proper ti^ne, little of
the nitrogen is left after harvest (95, 99, 714,135).
   The greatest difficulty  in preventing excessive fertili-
zation is in  predicting what levels of fertilizer should be
applied so that the resulting level in the soil is adequate.
The first requirement is to predict the potential yield of
the crop and thus the nutrient requirements. Then, the
soil's  ability  to  meet  these  requirements  must be
evaluated. Finally, the efficiency of the applied fertilizer
in meeting the remaining nutrient requirements must be
considered.
   This difficulty in accurately predicting fertilizer needs
and  low  nitrogen  costs  has  led  some  growers to
overfertilize so  that lack of nutrients would not limit
yields.  Recommendations based  simply on  "mainte-
nance" or  "balance"  approaches to  replace nutrients
removed by the  crop should be discouraged (720). They
fail to account for either the nutrient supplied by the
soil or the losses of applied fertilizers.
   The  yield of  any crop and its response to applied
fertilizer  depends  upon  many different  soil, plant,
climatic,  and  cultural  factors  (759).  For  example,
experiment  station reports from Maryland and Michigan
show the  yield of corn can vary 2-fold across  a single
state. Both  climate and soil properties can be involved.
As the precipitation during the growing season decreases,
the water stored in the soil when the plant starts to grow
becomes the major yield determinant.
   Stanford (141,  143) has  published  extensively on
estimating nitrogen fertilizer requirements. He argues
persuasively that there is an  internal nitrogen  require-
ment of the crop for the expected yield. To adequately
estimate  this  requirement  requires  considerable   field
work. A first approximation can be made by considering
the expected yields and  nutrient contents (Table 17,
Vol.  I). Good  farmers in  fertile farming areas  will
probably produce higher yields but it is anticipated that
the nutrient content of the crops will be proportionately
74

-------
                      Table 4. Bibliography fur nutrient management practices (Volume I, Section 4.3).
Nutrient management practice
No.
Page No.
in
Vol.1.
Description '
Citations
Significant subjects
                                                      General
Nl
          76
       Kliminating Fxcessive Fertili-
          zation
                              Herronffo/(60)


                              Linville and Smith (95)

                              Petersen and Sander (120)


                              Stanford (143)

                              Stewart (150)

                              Thomas and Hanway (159)

                              Thomas and Peaslee (160)

                              Viet.s(168)
                           Nil rale already in soil


                           Little left with adequate amount*

                           Discourages maintenance and balanced
                              approaches

                           tstimating  N fcrtili/er requirement

                           Build-up with excessive use

                           Response to fertilizer

                           Phosphorus recommendations

                           Implication of banning all fertiliser
                                                  Leaching Com rol
N2
N3
N4
N:7
78
           79
79
80


83



83
Timing Fertilizer Application
       Using Crop Rotations
                  Using Animal Wastes for
                     Fertilizer
                  Plowing-under Green Legume
                     Crops

                  Using Winter Cover Crops
                  Controlling Fertilizer Release
                     or Transformation
Aldrich M)

Bouldin, Reid, and
   Lathwell (22)

Lathwell. Bouldin. and
   Reid<2J)

Bezel icek. Mulford. and
   Magee(F2)

Olsen (113)

Stewart, Viets, and
   Hutchinson (152)

Ashraf (6)

Uttormark, Chapin, and
   Green (166)

Zwerman et al (185)

Lyon. Buck man. and
   Brady (97)

Frink (46)

Tliomas (156)

Anonymous (4)

Boswel) and Anderson (]_9)

Broadbent (25)

Hauck and Koshino (58)
Conditions for fall fertilization

Argument for summer sided ressim:


Period of maximum use


Soybeans don't need fertiliser N


Profile N proportional to amount applied

Alfalfa removes deep N


Cost of storage

N and P lost in snowmelt


Manure increases infiltration

References for amount of fixation


Reduced leaching by cover crops

Recommended planting time

Cost estimate

Field experiment with inhibitors

Poor future prospects

Advantages of slow release
                                                                                                                     75

-------
                                              Table 4 (continued)
Nutrient management practice
No.

Page No.
in
Vol. I.
Description


Citations


Significant subjects

                                          Control of Nutrients in Runoff
N8

N9

N10

83

83

83

Incorporating Surface Appli-
cations
Controlling Surface Appli-
cations
Using Legumes in Haylands and
Pastures
Timmons, Burwell, and
Holt (162)
Wagner and Jones ( 171 )

Allison (3)

Plow-down reduces fertilizer loss

Less-frequent P and K applications needed on
fertile soils
Grass uses N from legumes

                                       Control of Nutrient Loss by Erosion
Nil
83
Timing Fertilizer Plow-down
None

higher.  More accurate data  for  the area under  con-
sideration  are usually available from the State experi-
ment station.
   Given a yield  estimate and a crop requirement, the
next step is to estimate the amount of nitrogen the soil
will supply without fertilizer. There are several factors to
be considered. One is the capacity of the soil to produce
nitrate by  mineralizing organic nitrogen in the soil. An
incubation  method would  appear  to  give  a reliable
estimate of the potential (144, 148) which is adjusted
for temperature and moisture effects (146, 147). How-
ever, the time required for the incubation prohibits soil
testing laboratories from using it (33). A hot water or
steam  extraction  of the soil sample may  provide an
adequate estimate of the potential  mineralizable nitro-
gen (139,142).
   Also to be considered is the amount of nitrate already
in the soil  (60,  61).  In the more  humid regions, any
nitrate remaining in the  soil after harvest will be leached
out of the root  zone before the crop  can  use it the
following season. But in the more arid areas such as the
Great  Rains, this  leaching doesn't occur regularly.  A
final  factor  that  needs to be  considered for the soil
supply is the amount of nitrogen that is available  from
residues and/or cover crops.
   The  final step  is  to estimate the fraction of the
fertilizer that the crop will  use. Many field experiments
show that the plant takes  up less than 70 percent and
often  less  than   50 percent  of the applied  nitrogen
fertilizer  (91,  96, 116).  Some  of the fertilizer  is
immobilized into organic matter, some is denitrified, and
some can be leached out of the root zone. These changes
can be reduced to some extent by applying the fertilizer
when the plant is growing.
   The behavior of nitrogen is quite complex and several
estimates are required to predict the amount of fertilizer
needed. A  simpler, but often less accurate approach, is
presently used in most cases. This approach relies on the
results  of   previous  experiments  in  the  area  where
different rates of nitrogen were applied to  the  crop.
Figure 5 is a summary of such an experiment (66). In
this particular case, the soil and residues supplied enough
nitrogen for a yield of  65 bu/acre. Applying nitrogen
fertilizer at  the rate  of  120 Ibs/acre  produced 141
bu/acre, which was close  to the maximum yield observed
(147 bu/acre). Such information  would be the basis for
recommending that 120 pounds  of nitrogen be applied
to corn  under similar conditions.
   The  phosphorus cycle is less complicated than the
nitrogen cycle  and the phosphorus  fertilizer recom-
mendations are  also much easier to make (760).  While
less than 20 percent of the applied phosphate is usually
taken up because of the reactions with the soil, there are
essentially  no  losses  by leaching or  volatilization. In
addition, soil tests have been extensively correlated with
yield  responses so that the fertilizer requirement is more
readily predicted from soil tests.
   Reducing fertilizer application to a less-than-adequate
level doesn't always decrease pollution and may in fact
increase it. Inadequate  fertilization decreases growth and
can increase runoff, erosion, and leaching. Smith  (134)
reports  a  9-fold increase  in nitrate  loss  from inade-
quately fertilized corn. Viets (167) discusses the implica-
tions  of banning the  use of all fertilizer. The  effect
76

-------
 would range from very little with soybeans in Iowa to
 over  a  90  percent  reduction  in  per  acre  yield  of
 grapefruit  in Florida.  Acreage  would  need  to  be  in-
 creased 20  to 30 percent for the major crops of corn,
 wheat, and cotton in the first year while there was still
 some residual fertility. In addition, the added acreage
 would be of lower fertility and more erodible, thereby
 creating additional problems.
   Mayer and Hargrove (706)used an economic model to
 examine the impact of restricting fertilization to certain
 levels nationally and only in Iowa. Reduced use through-
 out  the  country would eliminate foreign exports,  in-
 crease cropland  acreage, and  increase prices of farm
 products. If only a single state  such as Iowa  restricted
 fertilizer, the impact on  the farmer would be very great,
 since his yield per acre would be reduced but the price
 for his product  would not go  up because of supplies
 from adjacent states.
                           Timing Fertilizer Application

                              The  time of  the  fertilizer application  can be  an
                           important tool  for increasing the efficiency of fertilizer
                           use and reducing fertilizer loss. Fertilizer nitrogen use is
                           maximized when  fertilizer is applied near the time of
                           maximum vegetative growth (21, 91). Most  crops grow
                           the fastest several  weeks  after the plant emerges, as
                           illustrated by corn  in Figure 33 of Volume  I (56). The
                           application of nitrogen fertilizers several weeks after the
                           plant  has  started  to grow is commonly called  summer
                           sidedressing. Bouldin  et  al. (22) provide a  number of
                           arguments for the summer sidedressing of corn based on
                           experiments in  New York. Since  the fertilizer is  used
                           more  efficiently, less  fertilizer is needed and the lower
                           fertilizer cost offsets the added cost of application. They
                           argue  that if the  field is too wet for sidedressing,  then
                           previously applied  fertilizer  will probably  be lost  by
     175
     150
     125
     100
2
o
I    75
o
o
     50
      25
                  40
80
120
160
200
240
280
32O
360
                                APPLIED NITROGEN, Ibs./acre
                       Figure 5.-An example of the yield response of corn to applied nitrogen (66).
                                                                                                         77

-------
leaching or denitrification. Conversely, if it is too dry to
move the nitrate down to the roots, then the growth will
be retarded by lack of water anyway.
   If leaching is not  a problem, then applying fertilizers
preplan!  in  the spring  or even in the fall (except on
sandy soils) may be  acceptable. For fall fertilization it is
usually recommended that an ammonium type fertilizer
be  applied  after the soil  cools below 50° F (7). This
recommendation is based on the facts that while nitrate
is mobile,  ammonium  is  relatively  immobile and  is
converted to nitrate very slowly below 50° F (40, 126,
182).
   Nelson and Uhland (112) were among the first to
show  regional  variation  in  leaching.  Using  Thorn-
thwaite's  calculations  with  a constant  4  inches of
water-holding capacity and implicitly accounting for the
temperature effects  on  nitrification,  they divided the
area east of the Rocky  Mountains into four  regions of
different leaching potential (Fig. 20, Appendix B).
   We have attempted  to  provide  a more detailed
mapping by combining a  nitrification model  with  a
percolation  model. These models are described in detail
in Appendix B. The results  of the simulations are also
presented in Appendix B as maps of percentage loss for
various soil  groups.  These maps can be used to estimate
average leaching  losses  for  an area if the hydrologic
characteristics  of the soils  are  similar to one of the
groups modeled. Losses of fall-applied and spring-applied
ammonia (Figs. 34  and 35, Vol.  I) by nitrate leaching
were prepared from these maps by selecting the appro-
priate loss for  the predominant soil group in each Land
Resource Area.
   These results still represent coarse approximations in
spite of the numerous  factors that were incorporated
into  the  model. Only  a  few  combinations  of soil
characteristics were  modeled. Another limitation was the
relation  used  for  the  ammonium-nitrate conversion.
Other factors such as moisture and pH could modify the
actual conversion  rate for a particular soil. Also, immobi-
lization  and denitrification  could remove some of the
nitrate produced, and the leaching process may not be
exactly plug flow. The model assumes little transpiration
and no uptake during the winter,  which  would tend to
overestimate  leaching in  southern states where  there
could be plant growth. Thus, these maps can  serve only
as  first approximations and more detailed information
on locally important variations must  be  obtained from
the Soil  Conservation Service, State experiment stations,
and extension staffs  for each area.
    Split applications, applying part of the nitrogen in the
fall or spring and  the  rest as a summer sidedressing,
combines some of the features of each time of applica-
tion. The first application provides enough fertilizer for
a poor year. The last application is not large enough to
cause toxicity  problems that sometimes occur  and it
provides an opportunity to adjust for favorable weather,
increased plant population, and optimum planting dates.

 Using Crop Rotations

   Crop rotations can  be  used  to  reduce the average
 amount of  nitrogen fertilizer required. High  nitrogen-
 requiring crops such as corn, cotton, and sorghum can be
 rotated with crops requiring less nitrogen, such as small
 grains, or legumes which require only small amounts of
 starter fertilizer, such as soybeans or alfalfa. Olsen (113)
 found  the  amount  of nitrogen in the  profile   was
 proportional to the amount  of nitrogen applied during
 the  rotation.  Thus,  the  average nitrogen content in
 drainage from a watershed  with diversified crops should
 be lower than if the watershed were completely in crops
 like corn.
   Alfalfa is particularly useful  because  its deep  root
 system  can remove some  nitrate from deeper  depths
 than most crops can (752). Soybeans are high cash value
 legumes that  don't  require  nitrogen fertilization  but
 appear  to respond to high levels of soil nitrogen from
 previous crops (12, 175). The major limitations in crop
 rotation are the loss of cash  income and/or the cost of
 additional equipment.

 Using Animal Wastes for Fertilizer

   Animal wastes, or manure, have been used as a source
 of plant nutrients  for  thousands of years. A previous
 section discussed  many of the properties of manure.
 Zwerman et al. (184, 185) report the increased infiltra-
 tion and reduced runoff from long periods of manure
 use. This  section will be concerned with the problems
 associated with using manure as a substitute for fertil-
 izer.
   The  most   serious  problem  from a water quality
 standpoint is the  loss of  nutrients  in runoff. Animal
 manure produced during  the  winter  must  be either
 stored or applied when the crops are not growing  and
 chances of loss are greater. Since equipment can't enter
 fields that are wet with fall or spring rains, farmers with
 little  or no storage capacity are forced  to spread  the
 manure on frozen or snow-covered fields. The resulting
 runoff  from rains  or snowmelt  can carry 10  to 20
 percent of the nitrogen and  phosphorus in the manure
 (109,  166).  The  losses from  a manure  application
 containing  100 pounds of N per  acre are  10 to 20
 pounds  of N per acre  and 3 to 10 pounds of P per acre
 (107).
78

-------
   Flowing-down  manure  soon after application  is the
most  appropriate  method of  controlling losses  from
broadcast applications. This method also prevenls nitro-
gen loss as  volatile ammonia.  However, meadows and
haylands can't be plowed, nor can frozen croplands. For
these cases,  the State of Maine guidelines (101) recom-
mend that only upland fields with less than 3 percent
slope be used for manure spreading when frozen or snow
covered. They also recommend against spreading on any
fields with slopes  greater than 25 percent or within 100
feet of wells, springs, ponds,  or lakes, or when there is a
high possibility of runoff.
   While most manure is in a relatively solid form, some
stored wastes  are in a slurry form. Slurries, principally
from dairy and swine operations, can be injected directly
into the  soil, thus almost eliminating  runoff losses. A
large part of the nutrients are in solution and can move
into the soil  even if surface applied. Storage of manure is
a large added  expense with little economic return. The
investment needed for manure storage on  a dairy farm
has been estimated as  three to five times the cost of
daily spreading (6).
   A second problem of substituting manure for fertil-
izer is determining how much nitrogen  is being applied.
Not only does the nutrient content of manure change
with different animals, but  also  with  their  feed  and
environmental conditions. If  bedding is used, this pro-
vides added  bulk with no added nutrients, but it absorbs
liquids and prevents nutrient  loss. The largest losses are
probably by volatilization of nitrogen  during storage,
spreading, and before  incorporation.  As  much  as  40
percent  of the nutrient value  can be lost by delaying
incorporation  for 4 days (127). Ammonia gas is con-
tinually being lost, while  denitrification of nitrate and
nitrite occurs in anaerobic  storage.
   Other problems include the fact  that not all the
nitrogen  is  available  during  the first  growing season.
Organic nitrogen   compounds  are  similar to  the  soil
organic matter in  that some are more easily mineralized
by microbes than others.  This  is not too  serious  since
repeated  yearly  applications  will build  the organic
nitrogen up  so that the total mineralized is equivalent to
the amount  added. The nutrients may  not  be in the
proper ratio  fora  particular crop on that field. However,
this can easily be corrected by adding nutrients to the
'icld or to the manure in storage.

?lowing-Under Green Legume Crops

   This practice was frequently used to  supply nitrogen
before  the  development of commercial fertilizers, but
little research has been done  on it since commercial
fertilizers became available. Thus, the principal sources
of  information are older soil science  books (97, 108).
The practice is based on the symbiotic relation between
some types  of microorganisms  and  legume plums in
which nitrogen from the atmosphere  is converted into
plant protein. From 40 to  60 pounds of N can  be
supplied per ton  of dry forage.  Part of the nitrogen in
the plant comes from mineralized soil  nitrogen, but  this
is probably  balanced by not considering the nitrogen in
the plant roots. Obviously, the greatest limitation of this
practice is loss of any return from this crop. If the forage
is harvested, then the net gain in soil nitrogen is small.

Using Winter Cover Crops

   Although  winter  cover crops arc recommended for
control of soil erosion during the  fall and winter (see
Erosion  Control Practices),  they  can also reduce nitrate
leaching through  plant  uptake  of  nitrate  and  reduce
percolation by drying the soil out. An oat crop reduced
nitrate leaching 4-fold on one soil and eliminated it  on
another. Vetch, however, reduced leaching only slightly
and  because  it is a legume, added nitrogen (14). Cover
crops of oats, timothy, and rye reduced leaching 40 to
60  percent  (46). Thomas (156) recommends  that  the
cover crop be planted by October for the most effective
control of leaching.
   The  major expense  of this practice  is planting  the
crop. Some economic  return  can be obtained  by using
the crop for  winter grazing.  Also,  in some areas it is
possible  to double crop; that is, grow a winter grain such
as wheat and then  plant a short season crop  such as
soybeans. A serious limitation of cover crops is that they
can  remove so much of the soil water that  the main
summer crop suffers, particularly in a dry year.
Controlling Fertilizer Release or
Transformation

   Many  researchers  have explored  the  possibility  of
controlling fertilizer release or availability. Recently the
interest has been  very great  and over 50 papers were
presented at the 1974 Annual Meeting of the American
Society of  Agronomy dealing with  this subject. Two
basic approaches are being used:  a slow release fertilizer
and a nitrification  inhibitor.
   Slow  release  fertilizers offer three advantages: i)
reduction in nutrient loss by leaching and runoff, ii)
reduction in immobilization before plant uptake, and iii)
reduction in losses by denitrification  and volatilization
(58, 77, 119,  122). Three processes are used to slow the
release  of the fertilizer from  the granule:  i) controlling
dissolution by a physical  barrier, ii) using compounds of

-------
limited  water  solubility, and  iii)  using  a  barrier that
decomposes. Of the  13 different slow release fertilizers
developed,  sulfur-coated  urea seems to be  the most
promising. Most of  the commercial production is being
used on  turf. The  present cost is 25 to 40 percent more
than  uncoated  urea (4).  The  greatest   problem is  to
control the release so that the fertilizer is available when
the plant needs it. If the release doesn't occur in a short
period of  time  for  row  crops, then  the  remaining N  is
susceptible for leaching  after harvest. The  future looks
promising, but more research is needed before large scale
recommendations can be made.
   Nitrification inhibitors  are  chemicals that  prevent
microbes from converting  ammonium to  nitrate. Five
chemicals  olfer possibilities. The most widely tested are:
2-chloi o-6-( t richloromethyl)    pyridine,   sold   as
N  SERVE  by   Dow  Chemical;  2-amino-4-chIoro-6-
rnethy I pyridine, sold as AM by Mitsui-Toatsu Industries;
and sodium a/ide. Experiments with soil in plastic bags
in the field from  November to April show  that most of
the conversion was  prevented by N-SERVE (19). One of
 the greatest difficulties  has been to keep the inhibitor
near  the ammonium; usually percolating water separates
 them. Broadbent  (25) doesn't consider the  prospects of
developing a practical method to be very good.

 Incorporating Surface Applications

    Immediate incorporation of surface-applied  fertilizers
and manure can prevent significant losses of nutrients.  A
 number of studies have shown that the losses are greatest
when the runoff  occurs soon  after application (see the
 section  on runoff losses). Timmons et al.  (162) report
 that  deep incorporation  of  the  fertilizer by plowing
 clown and subsequent disking reduced the nutrient losses
 to levels  similar  to those  in runoff from unfertilized
 plots. Broadcasting  on a plowed surface is adequate if no
 additional  tillage  is performed because the infiltration
 is  very  high.  Disking instead of plowing  broadcasted
 fertilizer  was not  effective.  Up to 30 tons/acre  of
 manure  have been  incorporated into the soil with little
 increase in the nitrate and ammonium  contents of the
 tail water  from irrigation (151).
 Controlling Surface Applications

    The time of application and the type of fertilizer can
 be controlled  to some extent. Fertilizer should not be
 applied  during periods of expected runoff. Fall-seeded
 grains are often  top-dressed with  fertilizer in the spring.
 If leaching is not a problem, then fertilization at planting
would reduce runoff losses. Pasture and haylands usually
require surface application of fertilizers and thus runoff
losses could  be a  problem. Wagner and  Jones  (777)
report that if a high level of fertility is maintained, then
timing of phosphorus and potassium fertili/er applica-
tions  is not critical. In fact, on slightly deficient soils
enough P and K can be plowed under when the forage is
planted  to last 4 or 5 years. Since nitrogen is so mobile,
the greatest efficiency  is obtained by applying it shortly
before or during the growing season. Cool season forages
such as  the brome and blue grasses make their growth in
the spring and fall,  whereas warm season grasses such as
the Bermudagrasses  make their best growth  during  the
summer.
   Spraying a  fluid fertilizer on the surface might have
lower losses in runoff than broadcasting granular ferti-
lizers even though there is no published research on  the
subject. Fluid fertilizers are liquids or suspensions of
micro  crystals  and  therefore  should come  in quicker
contact with the soil than granules that must dissolve.
Ammonia  cannot  be used because it volatilizes too
easily. Urea is also subject to some losses because it is
converted to ammonium.

Using Legumes in Haylands and Pastures

   Legumes can be planted with grass in  pastures and
haylands to supply much  of  the nitrogen requirement
and thus reduce the need  to fertilize with nitrogen. The
legumes can be  very effective in this situation because
the grass receives nitrogen in leakage from the legumes as
well as the mineralized soil nitrogen (3). As pointed out
by Kilmer (76), leaching losses will be  higher  from
legumes than from grasses. However, if leaching is not a
problem and runoff losses are, then legumes can be used
effectively.  A major problem is that competition  for
other nutrients, water, and sunlight causes the grass to
crowd out the legume in a  few years.

Timing Fertilizer Plow-Down

   When nutrients are being lost by  sediment transport,
erosion  control practices are the obvious answer in most
cases.  An  additional  procedure  that  can  be recom-
mended is plowing during the least erosive  period and
leaving  the  field  in  the  least erosive condition. For
example, if erosion is less in the fall than in the spring,
phosphorus and potassium  fertilizer might  be plowed
under in the fall using stubble mulching techniques or
followed by a cover crop so that an erosive period in  the
spring can be avoided.
 80

-------
               Mathematical Models

   The soil-plant  system  is very  complex and  dynamic
and therefore  the impact  of various management prac-
tices can vary considerably. One of the most  efficient
ways  of  testing alternatives on  complex and  dynamic
systems is to describe  the system with a mathematical
model. If (he model adequately  represents the  behavior
of the system for  known responses and is based on
sound  fundamental  relations,  then its  responses  for
various alternatives  can be quickly  tested with modern
computer equipment.
   A number  of  programs have  been started in  recent
years  to  develop models for  sediment and chemical
transport from watersheds. Most  of these efforts are still
in  the development  stage  and have yet to be adequately
tested against field results.
   AC'TMO,  an agricultural  chemical transport  model
(44, 45}  linked  together  a hydrology  model  (6,5), an
erosion  model  (//
-------
                                         LITERATURE CITED
 1. Aldrich,  S.  R. 1970.  The  influence of cropping
    patterns, soil management and fertilizer on nitrates.
    Proc. Twelfth Sanitary Engin. Conf., Univ. Illinois,
    Urbana:152-169.

 2. Alexander, M., Moyle, J. B., Patrick, R., Weinberger,
    L., and O'Bryan, D. 1965. Aquatic blooms, Appen-
    dix Y9.  In  Restoring  the  quality of our environ-
    ment.  Environ. Poll.  Panel, President's Science
    Advisory Committee, The White House.

 3. Allison, F. E. 1973. Soil organic matter and its role
    in crop  production. Elsevier Scientific Pub.  Co.,
    Amsterdam, The Netherlands.

 4. Anonymous. 1974.  Enter: Sulfur-coated urea, a
    slow-release  fertilizer. Chem. Engin., 32 and 34.

 5. Armitage, E. R. 1974. The runoff of fertilizers from
    agricultural  land and  their  effects on the  natural
    environment. In D. E. G. Irvine and B. Knights, eds.,
    Pollution and the use of chemicals  in agriculture.
    Butterworths Pub., London.

 6. Ashraf,  M.  1974. An  analysis  of the  impact of
    manure  disposal regulations on dairy farms. Amer.
    Jour. Agr. Econ. 56(2): 331-336.

 7. Baeumer, K.,  and Bakermans,  W.  A.  P.  1973.
    Zero-tillage. Advan. in Agron. 25: 77-123.

 8. Barnett,  A. P.,  Carreker, J. R., Abruna, F., Jackson,
    W. A., Dooley, A. E., and Holladay, J. H. 1972. Soil
    and nutrient losses in runoff with selected cropping
    treatments  on  tropical soils.  Agron.  Jour. 64:
    391-395.

 9. Barrows,  H. L., and  Kilmer, V. J. 1963.  Plant
    nutrient  losses  from soils by water erosion. Advan.
    in Agron. 15:303-316.

10. Bartholomew, W. V., and Clark, F. E. (Eds). 1965.
    Soil nitrogen. Amer. Soc. Agron., Madison, Wis.

11. Bedell, G. D., Kohnke, H., and Hickok, R. B. 1946.
    The effects of two farming systems on erosion from
    cropland. Soil Sci. Soc.  Amer. Proc. 11: 522-526.
12. Bezdicek, D.  F., Mulford, R. F., and Magee, B. H.
    1974. Influence of organic nitrogen on soil nitrogen
    nodulation,  nitrogen  fixation, and yield  of soy-
    beans. Soil Sci. Soc. Amer. Proc. 38(2): 268-273.

13. Biggar, J. W., and  Corey, R. B. 1969. Agricultural
    drainage  and  eutrophication.  In   Eutrophication:
    causes, consequences,  correctives. Natl. Acad. Sci.,
    Washington, D. C.:  404-445.

14. Black, C. A. 1966. Crop yields in relation to water
    supply and soil fertility./« W. Pierre, D. Kirkham, J.
    Pesek, and  R. Shaw,  eds. Plant  environment  and
    efficient water  use. Amer. Soc. Agron., Madison,
    Wis.: 177-206.

15. Black, C. A. 1970. Behavior of soil  and fertilizer
    phosphorus in relation to water pollution. In T. L.
    Willrich and G. E. Smith, eds. Agricultural practices
    and water quality. Iowa State Univ. Press, Ames,
    la.: 72-93.


16. Blanchar, R. W., and Kao, C. W. 1971. Effects of
    recent  and  past  phosphate  fertilization  on  the
    amount  of  phosphorus  percolating  through  soil
    profiles  into  subsurface  waters.   Missouri Water
    Resour. Res. Ctr. Proj. No. A-031-Mo.

17. Borchardt, J. A. 1970. Secondary effects of nitro-
    gen in water. In Nitrate and water supply: source
    and control.  Proc. Twelfth Sanitary Engin. Conf.,
    Univ. Illinois, Urbana, 111.: 66-76.

18. Bormann,  F. H.,  Likens, G. E.,  Siccama, T. G.,
    Pierce, R. S., and Eaten, J. S. 1974. The export of
    nutrients and recovery of stable conditions follow-
    ing deforestation at Hubbard Brook. Ecol. Monogr.
    44(3): 255-277.

19. Boswell, F. C., and Anderson, 0. E. 1974. Nitrifica-
    tion inhibitor studies  of soil in field-buried poly-
    ethylene  bags.  Soil Sci.  Soc.  Amer.  Proc. 38(5):
    851-852.

20. Bouldin, D. R., Johnson, R. L., Burda, C., and Kao,
    C.  W.  1974.  Losses  of inorganic nitrogen  from
    aquatic  systems.  Jour.  Environ.  Qual.  3(2):
    107-114.
82

-------
21. Bouldin, D. R., and Lathwell, D. J. 1968. Timing of
    application is key to crop nitrogen  use. Food and
    Life Sci. 1:9-12.

22. Bouldin,  D. R., Reid,  W. S.,  and Lathwell, D. J.
    1971. Fertilizer practices which minimize nutrient
    loss.  In Agricultural  wastes: principles and guide-
    lines for  practical solutions: 21-35. Cornell Univ.,
    Ithaca, N. Y.

23. Bradford,  R. R.  1974. Nitrogen and phosphorus
    losses  from agronomy  plots  in  north Alabama.
    EPA-660/2-74-033. U. S. Govt. Printing Off., Wash-
    ington, D. C.

24. Brezonik, P. L. 1972. Nitrogen: Sources and trans-
    formations  in natural waters. In H. E. Allen and J.
    R. Kramer, eds. Nutrients in natural waters: 1-50.
    John Wiley and Sons, New York.

25. Broadbent, F.  E. 1971.  Nitrogen in  soil and water.
    In Proc. symp. on nitrogen in soil and water:56-77.
    Dept. Soil Sci., Univ. Guelph, Ontario.

26. Broadbent, F. E., and Clark, F. E. 1965. Denitrifica-
    tion. In W. V. Bartholomew and  F. E. Clark, eds.
    Soil nitrogen. Amer. Soc. Agron., Madison, Wis.

27. Burwell, R. E., Timmons, D.  R., and Holt, R.  F.
    1975. Nutrient  transport in surface runoff as influ-
    enced by soil cover and seasonal  periods. Soil Sci.
    Soc. Amer. Proc. (In press).

28. Case, A. A. 1970.  The health effects of nitrates in
    water. In  Nitrate  and  water  supply:  source and
    control. Proc. Twelfth Sanitary Engin. Conf., Univ.
    Illinois, Urbana, El: 40-46.

29. Committee on  Agriculture and  the Environment.
    1973. Productive agriculture and a quality environ-
    ment.  Agr. Board, Div. Biol. and Agr., Natl.  Res.
    Council.

30. Cornell University.  1969. Animal  waste manage-
    ment. Conf. on Agr. Waste Management, Syracuse,
    N.Y.

31. Crawford, N.  H.,  and   Donigian, A.  S.Jr. 1973.
    Pesticide transport and runoff model  for agricultural
    lands. EPA-660/2-74-013, 211  p. U.  S. Govt. Print-
    ing Off., Washington, D.  C.
32. Cywin,  A. and Ward, D. 1971. Agricultural pollu-
    tion of the Great Lakes basin. Combined report by
    Canada and the United States. EPA-13020-7/71.

33. Dahnke, W. C., and Vasey, E. H. 1973. Testing soils
    for nitrogen. In L. M. Walsh and J. D. Beaten, eds.
    Soil testing and plant analysis: 97-114. Soil Sci. Soc.
    Amer., Inc., Madison, Wis.

34. Doty, C.  W.,  and Carter, C.  E.  1965.  Rates and
    particle-size distributions of soil erosion from unit
    sources areas. Trans. ASAE 8(3): 309-311.

35. Downing,  K.  M., and  Merkens,  J.  C.  1955. The
    influence of dissolved oxygen concentration on the
    toxicity of unionized  ammonia to  rainbow trout.
    Ann. Appl. Biol. 43:  243-246.

36. Dunigan, E. P., Mondart, C. L., Jr., Phelan, R. A.,
    and Shamsuddin, Z.  H. 1974. Surface runoff losses
    of fertilizers. Louisiana Agr., Summer 17(4): 4-7.

37. Eckert,  R. E., Jr., Klomp, G. J., Young, J. A., and
    Evans,  R.  A.  1970.  Nitrate-nitrogen status of fal-
    lowed rangeland soils.  Jour.  Range  Manag. 23(6):
    445^47.

38. Engibous,  J. C. 1972. Economics of fertilizer use.In
    The fertilizer  handbook: 143-153.  The Fertilizer
    Institute, Washington, D. C.

39. Fitzgerald, G. P. 1972. Bioassay analysis of nutrient
    availability. In H. E. Allen and J. R. Kramer, eds.
    Nutrients  in natural  waters:  147-169. John Wiley
    and Sons, New York.

40. Frederick,  L.  R. 1956.  The  formation  of nitrate
    from ammonium nitrogen in  soils:  I.   Effect of
    temperature.   Soil  Sci.  Soc.  Amer.  Proc.  20:
    496-500.

41. Frere, M. H. 1971. Requisite sampling frequency for
    measuring nutrient  and  pesticide movement  with
    runoff  waters. Jour.  Agr.   Food  Chem. 19(5):
    837-839.

42. Frere, M. H. 1975. Integrating chemical factors with
    water and sediment transport from  a  watershed.
    Jour. Environ. Qual. 4(1): 12-17.
                                                                                                       83

-------
43. Frere, M. H., Jensen, M. E., and Carter, J. N. 1970.
    Modeling water and nitrogen behavior in the  soil-
    plant system. Proc. 1970 Summer Computer Simu-
    lation Conf.: 746-750.

44. Frere, M. H., Onstad, C. A., and Holtan, H. N. 1974.
    Modeling the  movement of agricultural chemicals.
    Proc. 1974 Summer Computer Simulation Conf.:
    271-274.

45. Frere, M. H., Onstad, C. A., and Holtan, H. N. 1975.
    ACTMO, an agricultural  chemical transport model.
    ARS-H-3, Agr. Res. Serv., U. S. Dept. Agr., Washing-
    ton, D. C.

46. Frink, C.  R.  1970. The nitrogen cycle of a dairy
    farm. In  Relationship of agriculture  to  soil  and
    water  pollution:  127-133.  Cornell  Univ., Ithaca,
    N.Y.
47.  Frink,  C.  R.  1971.  Plant nutrients  and  water
     quality. Agr. Sci. Rev. 9(2):  11-25.

48.  Fromm, P.  O.  1970. Toxic action of water soluble
     pollutants on fresh-water fish. U.  S. Water  Qual.
     Off., Environ. Protec. Agcy. Grant No. 18050 DST.

49.  Fry,  F. E. J.  1969. Some possible physiological
     stresses induced  by eutrophication.  In Eutrophi-
     cation: causes, consequences, correctives: 531-536.
     Natl. Acad. Sci., Washington, D. C.

50.  Gambell,  A. W., and Fisher, D. W. 1964. Occurrence
     of sulfate and nitrate in rainfall.  Jour. Geophys.
     Res. 69(20): 4203-4210.

51.  Gburek, W. J., and  Heald, W. R.  1970. Effects of
     direct runoff from  agricultural land  on the  water
     quality of small streams. In Relationship of agricul-
     ture  to soil  and  water pollution: 61-68. Cornell
     Univ., Ithaca, N. Y.

52.  Ghoshal,  S. 1974.  Fate of fertilizer phosphorus
     under aerobic decomposition. Plant and Soil 40(3):
     685-688.

53.  Goering,  J. J.  1972.   The role  of nitrogen  in
     eutropnic  processes. In Water  pollution micro-
     biology: 43-68. Wiley-Interscience, New York.
54. Hagin, J., and Amberger, A. 1974. Contribution of
    fertilizers  and manures to the N-  and P-load of
    waters.  Final  report  submitted to the Deutsche
    Forschungs Gemeinschaft.

55. Hanawalt, R. B. 1969. Environmental factors influ-
    encing the sorption of atmospheric ammonia by
    soils. Proc. Soil Sci. Soc. Amer. 33(2): 231-234.

56. Hanway, J. J.  1966. How a corn plant grows. Spec.
    Rep. No. 48. Iowa State Univ., Ames, la.

57. Harmeson, R. H., Larson, T.  E.,  Henly,  L. M.,
    Sinclair,  R. A., and Neill, J. C. 1973. Quality of
    surface water  in Illinois,  1966-71.  Bui. 56, Dept.
    Registration and Education, 111. State Water Surv.

58. Hauck, R. D.  and  Koshino, M. 1971. Slow-release
    and amended fertilizers. In R. A. Olson, T. J. Army,
    J. J.  Hanway, and V. J. Kilmer,  eds. Fertilizer
    technology and use: 455494. Soil Sci. Soc. Amer.,
    Inc., Madison, Wis.

59. Heinemann, H. G., Holt, R. F., and Rausch, D. L.
    1973. Sediment  and nutrient research on  selected
    corn  belt reservoirs. In W. C. Ackermann, G. F.
    White, and E. B. Worthington, eds. Man-made lakes:
    their  problems and environmental effects:381-386.
    Amer. Geophys. Union, Geophys. Monogr. 17.

60. Herron, G. M., Dreier, A. F., Flowerday, A. D.,
    Colville,  W. L., and Olson, R. A.  1971. Residual
    mineral N accumulation in soil and its utilization by
    irrigated com. Agron. Jour. 63: 322-327.

61. Herron, G. M.,.Terman, G. L., Dreier, A.  F., and
    Olson, R. A. 1968.  Residual  nitrate nitrogen in
    fertilized  deep loess-derived soils. Agron. Jour. 60:
    477-482.

62. Hildreth,  R. J., and Williams, M. S. 1968. Fertilizer
    use economics. In L. B. Nelson, M. H. McVickar, W.
    C. White, R.  D. Munson,  L. F. Seatz, and  S. L.
    Tisdale, eds. Changing patterns in fertilizer use.'433.
    Soil Sci. Soc. Amer., Inc., Madison, Wis.

63. Holt,  R. F. 1969. Runoff and sediment as nutrient
    sources. In  Water  pollution  by nutrients-sources,
    effects and control.  Bui.  13: 35-38.  Minnesota
    Water  Resour. Res. Ctr., Univ. Minn., Minneapolis.
84

-------
64. Holt,  R. F.,  Dowdy,  R. H., and Timmons, D. R.
    1970. Chemistry  of sediment in water. In T. L.
    Willrich  and G. E. Smith, eds. Agricultural practices
    and water  quality:  21-34. Iowa  State Univ. Press,
    Ames, la.

65. Holtan,  H. N., and Lopez, N. C. 1971. USDAHL-70
    model of watershed hydrology. ARS Tech. Bui. No.
    1435, U. S. Dept. Agr., Washington, D. C.

66. Hunter,  A.  S.,  and  Yungen, J. A.  1955. The
    influence of  variations in fertility levels upon the
    yield  and protein content of field corn in  eastern
    Oregon.  Soil Sci. Soc. Amer. Proc. 19: 214-218.

67. Hutchinson, G. L., Millington, R. J., and Peters, D.
    B.  1972.  Atmospheric  ammonia:  absorption  by
    plant leaves. Science 175: 771-772.

68. Hutchinson,  G. L., and Viets,  F. G., Jr.  1969.
    Nitrogen enrichment of surface water by absorption
    of ammonia  volatilized  from cattle feed lots. Sci-
    ence 166: 514-515.

69. Hynes, H. B. N. 1969. The enrichment of streams.
    In  Eutrophication: causes,  consequences,  correc-
    tives:  188-196. Natl. Acad. Sci.,  Washington, D. C.

70. Ibach, D. B., and Mahan, J. N. 1968. Fertilizer use
    patterns and potentials. In L.  B.  Nelson, M. H.
    McVickar, R. D. Munson, L. F. Seatz, S. L. Tisdale,
    and W. C. White, eds.  Changing patterns in fertilizer
    use: 1-31. Soil Sci. Soc. Amer., Inc., Madison, Wis.

71. Jaworski, N.  A., and  Hetling, L. J. 1970. Relative
    contributions of nutrients to the Potomac River
    basin  from  various sources. In Relationship  of
    agriculture  to soil and water pollution.  Cornell
    Univ., Ithaca, N.Y.

72. Johnson, J. D., and Straub, C. P. 1971.  Develop-
    ment  of a mathematical model to  predict the role of
    surfacp runoff and  ground water flow in over-fertili-
    zation of surface waters. Minnesota Water Resour.
    Res. Ctr., Univ. Minn., Minneapolis.

73. Junge, C. E. 1958. The distribution of ammonia and
    nitrate in rain water over the United States. Trans.
    Amer. Geophys. Union 39(2): 241-248.
74.  Ketcheson, J. W., and Onderdonk, J. J. 1973. Effect
     of  corn  stover  on  phosphorus  in  runoff from
     nontilled soil. Agron. Jour. 65(1): 69-71.

75.  Keup, L. E. 1968. Phosphorus in flowing  waters.
     Water Res. 2(5): 373-386.

76.  Kilmer, V. J. 1974. Nutrient losses from grasslands
     through leaching and  runoff. In D.  A.  Mays, ed.
     Forage  fertilization: 341-362.  Amer. Soc. Agron.,
     Madison, Wis.

77.  Kilmer, V. J., and Barber, J.  C. 1971. Fertilizer use
     and misuse. In Proc. symp. on nitrogen in soil and
     water. Dept. Soil Sci., Univ. Guelph, Ontario.

78.  Kilmer, V. J., Davis,  C.  H.,  Douglas, J., Harre, E.,
     Jones, U.  S., McVickar, M. H., Nelson, L. B., Reed,
     J. F., and White, W.  C.  1974.  The U. S. fertilizer
     situation and outlook. A report by a task force of
     the Council for Agr. Sci. and Technol., Ames, la.


79.  Kilmer, V. J., Gillian, J. W., Joyce, R. T., and Leitz,
     J. F.  1972. Loss of fertilizer  nutrients from soils to
     drainage waters. I. Studies on grassed watersheds in
     western North Carolina.  Water Resour.  Res.  Inst.
     North Carolina, Rep. No. 55.

80.  Kissel, D. E., Ritchie, J. T.,  and Burnett, E. 1973.
     Chloride  movement in  undisturbed  swelling  clay
     soil. Soil Sci. Soc. Amer. Proc. 37(1): 21-24.

81.  Kling, G. F. 1974.  A computer model  of diffuse
     sources of sediment and phosphorus moving into a
     lake.  PhD Thesis, Cornell Univ., New York.

82.  Klingebiel, A. A. 1972. Soil and water management
     to  control  plant  nutrients  in natural waters.  In
     Effects of intensive  fertilizer  use on the  human
     environment: 153-178.  Soils Bui. No.  16, FAO.
     Pub.  by Swedish Internatl. Devlpmt. Authority.

83.  Kolenbrander,  G. J.  1969. Nitrate  content and
     nitrogen loss in drain water.  Neth. Jour. Agric. Sci.
     17: 246-255.

84.  Kolenbrander, G. J. 1970. Calculation of parameters
     for the evaluation of the leaching of salts under field
     conditions, illustrated by nitrate. Plant and Soil 32:
     439-453.
                                                                                                       85

-------
85. Kolenbrander,  G. J.  1972.  Eutrophication  from
    agriculture with special reference to fertilizers and
    animal waste. In Effects of intensive fertilizer use on
    the human environment:  305-327. Soils  Bui. No.
    16,  FAO. Pub.  by  Swedish Internatl. Devlpmt.
    Authority.

86. Kramer, J. R., Herbes, S. E., and Allen, H. E. 1972.
    Phosphorus:  analysis of water, biomass, and sedi-
    ment.  In   H. E.  Allen  and  J.  R.  Kramer,  eds.
    Nutrients  in natural waters: 51-100. John Wiley and
    Sons, New York.
94. Levin, I. 1964. Movement of added nitrates through
    soil columns and undisturbed soil profiles. Eighth
    Internatl.  Congr. Soil Sci.,  Bucharest, Romania:
    1011-1022.

95. Linville, K. W., and  Smith, G.  E.  1971.  Nitrate
    content of soil cores from corn plots after repeated
    nitrogen fertilization. Soil Sci. 112(4): 249-255.

96. Loehr, R. C. 1974.  Agricultural waste management.
    Academic Press, Inc., New York.
87. Kreitler, G. W., and Jones, D. C. 1974. Natural soil
    nitrate: the cause of the  nitrate contamination  of
    ground water  in  Runnels  County,  Texas.  Proc.
    Second Natl. Ground Water Quality Symp., Denver,
    Col.:  179-188.

88. Kunishi,  H.  M.,  Taylor,  A. W.,  Heald,  W.  R.,
    Gburek, W. J., and Weaver, R. N.  1972. Phosphate
    movement from  an  agricultural watershed during
    two rainfall  periods. Jour.  Agr.  Food Chem. 20:
    900-905.

89. Kurtz,  L. T. 1970. The fate of applied nutrients in
    soils. Jour. Agr. Food Chem. 18: 773-780.
97. Lyon, T. L.,  Buckman,  H. 0., and Brady, N. C.
    1952. The  nature  and  properties of soil.  The
    MacMillan Co., New York.

98. McCarthy, P. L., Hem, J. J., Jenkins, D., Lee, G. F.,
    Morgan,  J. J., Robertson, R. S., Schmidt, R. W.,
    Symons, J. M., and Trexler, M. V. 1967. Sources of
    nitrogen  and  phosphorus in  water  supplies. Jour.
    Amer. Water Works Assoc. 59: 344-366.

99. MacGregor,  J. M., Blake, G. R., and  Evans, S. D.
    1974. Mineral nitrogen   movement into  subsoils
    following continued annual fertilization for corn.
    Soil Sci. Soc. Amer. Proc. 38(1): 110-112.
90.  Larsen, S., Gunary, D.,and Sutton, C. D. 1965. The
     rate  of immobilization  of applied phosphate  in
     relation  to  soil properties.  Jour.  Soil  Sci.  16:
     141-148.

91.  Lathwell, D. J.,  Bouldin, D.  R.,  and  Reid, W. S.
     1970. Effects of nitrogen  fertilizer  application in
     agriculture.  In  Relationship of agriculture to soil
     and water pollution: 192-206. Cornell Univ., Ithaca,
     N.Y.

92.  Lee, D. H. K. 1970. Nitrates, nitrites, and methemo-
     giobinemia.  Environ. Res. 3(5/6): 484-511.
 100.  Mackenthun,  K.  M.  1965. Nitrogen and  phos-
      phorus in water. U. S. Dept. Health, Education and
      Welfare, Washington, D. C.

 101.  Maine Experiment Station. 1972. Maine guidelines
      for manure and manure sludge  disposal  on land.
      Misc. Rep. No. 142, Agr. Expt. Sta., Orono, Me.

 102.  Martin, W. P., Fenster, W. E., and Hanson, L. D.
      1970. Fertilizer management for  pollution control.
      In T. L. Willrich and G. E. Smith, eds. Agricultural
      practices and water quality. Iowa State Univ. Press,
      Ames, la.
93.  Lee, G. F. 1973. Role of phosphorus in eutrophica-
     tion and diffuse source control. In S. H. Jenkins and
     K.  J. Ives, eds. Phosphorus in fresh water and the
     marine  environment:  111-128.  Pergamon  Press,
     Oxford, England.
 103.  Mason, H. C. 1974. The need for chemical control
      of pests and the use of fertilizer. In D. E. G. Irvine
      and  B. Knights, eds.  Pollution and the use of
      chemicals  in  agriculture.   Butterworths  Pub.,
      London.
86

-------
104.  Massey, H.  F and Jackson, M. L. 1952. Selective
     erosion of soil fertility constituents. Soil Sci. Soc.
     Amer. Proc. 16:  353-356.

105.  Massey, H.  F., Jackson. M. L., and  Hayes,  0.  E.
     1953. Fertility  erosion  on  two Wisconsin  soils.
     Agron. Jour. 45: 543-547.

106.  Mayer. L. V., and  Hargrove, S. H.  1971.  Food
     costs, farm  incomes, and crop yields with restric-
     tions on fertilizer use. Dept. Econ., Center for Agr.
     and Econ. Development, Iowa State  Univ., CAED
     Report No. 38, Ames, la.

107.  Midgley, A. R. and Dunklee, D.  E. 1945. Fertility
     runoff losses  from  manure  spread during the
     winter. Vermont Agr. Expt. Sta. B.ul. No. 523.

108.  Millar, C. E., and Turk. L. M. 1951. Fundamentals
     of soil science. John Wiley and Sons. New York.
115. Olsen, S. R.,and Flowerday. A. D. 1971. Fcnili/cr
     phosphorus interactions in alkaline soils. In  R. A.
     Olson, T. J. Army, J. J. Hanway. and V. J. Kilmer.
     eds.  Fertilizer technology  and use: 153-185. Soil
     Sci. Soc. Amer., Inc., Madison, Wis.

116. Olson,  R.  A. 1972.  Maximixing  the utilization
     efficiency of fertili/er N by soil and crop manage-
     ment. In Effects of intensive fertili/.er use on the
     human  environment:   34-52.  Soils Bui. No 16,
     FAO.  Pub.  by Swedish  Intcrnatl. Devlpmt. Au-
     thority.

117. Olson, R. A., Seim, E.  C.. Muir, J.,and Moshcr, P.
     N. 1972. Influence of fertili/.cr practices on water
     and   the  quality   of  the environment.   Univ.
     Nebraska,   Proj.   No.  B-004-NEB.   Rep.   No.
     PB211081, Natl. Tech. Inform. Scrv.. Springfield.
     Va.
109. Minshall, N.  E.  1970. Stream  enrichment from
     farm operations. Jour. Sanitary  Eng.  Div., Proc.
     ASCE96(SA2): 513-524.

110. Murphy, T.  J., and Cesarotti, D. J. 1974. Phos-
     phorus inputs from the atmosphere. W75-01199.
     OWRT-A-065-ILL(1).  Rep.  No. PB237-341/3WP.
     Natl. Tech. Inform. Serv., Springfield, Va.

111. National  Academy of Sciences.  1969. Eutrophi-
     cation:  causes, consequences, correctives. Natl.
     Acad. Sci., Washington, D. C.

112. Nelson,  L. B., and  Uhland, R. E. 1955. Factors
     that  influence loss of fal   applied fertilizers and
     their probable importance in different sections of
     the  United States.  Soil Sci. Soc. Amer. Proc.  19:
     492496.

113. Olsen, R.  J.  1970.  Effect  of various factors  on
     movement of nitrate nitrogen in soil profiles and
     on transformations of soil  nitrogen. PhD  Thesis,
     Univ. Wisconsin, Madison.

114. Olsen, R. J., Hensler, R. F.,  Attoe, O. J., Witzel, S.
     A., and  Peterson, L.  A. 1970. Fertilizer nitrogen
     and  crop  rotation  in  relation  to movement  of
     nitrate nitrogen through soil profiles. Soil Sci. Soc.
     Amer. Proc. 34(3): 448452.
118. Onstad, C. A., and  Foster, G. R. 1975. Erosion
     modeling  on  a  watershed.  Trans.  ASAE   18:
     288-292.

119. Parr, J. F. 1972. Chemical and biochemical con-
     siderations for maximizing the efficiency of ferti-
     lizer nitrogen. /// Effects of intensive fertilizer  use
     on the human environment: 53-86. Soils Bui. No.
     16, FAO.  Pub.  by  Swedish  Internal!. Devlpml.
     Authority.

120. Pctersen,  G.  A.,  and  Sander, D. H.  1974. The
     continuing need for  soil  testing. Soil  Sci. Soc.
     Amer. Proc. 38(5):  859-860.


121. Power, J. F. 1970.  Leaching of  nitrate-nitrogen
     unHer dryland agriculture in the  Northern Great
     Plains. In Relation  of agriculture to soil and water
     pollution: 111-122. Cornell Univ.,  Ithaca. N. Y.

122. Prasad,  R., Rajale.  G. B., and  Lakhdine, B. A.
     1971. Nitrification retarders and slow-release nitro-
     gen fertilizers. Advan. in Agron. 23: 337-383.


123. Rogers,  H. T.  1942. Losses of surface-applied
     phosphate and  limestone through  runoff from
     pasture land. Soil  Sci. Soc. Amer. Proc. 7: 69-76.
                                                                                                       87

-------
124. Romkens, M. J. M., Nelson, D. W., and Mannering,
     J. V. 1973. Nitrogen and phosphorus composition
     of surface  runoff  as affected by tillage  method.
     Jour. Environ. Qual. 2(2): 292-295.

125. Ryden, J. C., Syers, J.  K., and Harris, R. F. 1973.
     Phosphorus  in runoff and  streams.  Advan. in
     Agron. 25: 1-45.

126. Sabey, B. R.,  Bartholomew, W. V., Shaw, R., and
     Pesek, J.  1956.  Influence  of  temperature  on
     nitrification in soils. Soil Sci. Soc. Amer. Proc. 20:
     357-360.

127. Salter, R. M., and Schollenberger, C. J. 1939. Farm
     manure. Bui. No. 605, Ohio Agr. Expt. Sta., 69 pp.

128. Sawyer,  C. N.  1947.  Fertilization  of  lakes by
     agricultural  and  urban  drainage. New  England
     Waterworks Assoc. 61:  109-127.

129. Schuman, G.  E., Burwell, R. E.,  Piest, R. F., and
     Spomer, R. G.  1973.  Nitrogen losses in surface
     runoff from agricultural watersheds on Missouri
     Valley loess. Jour. Environ.  Qual. 2(2):  299-302.

130. Schuman, G.  E., Spomer, R. G., and Piest, R. F.
     1973.  Phosphorus losses from four agricultural
     watersheds on Missouri Valley loess.  Soil Sci. Soc.
     Amer. Proc. 37(3): 424427.

131. Schwab, G. O.. McLean, E. O., Waldren, A. C.,
     White, R. K., and Michener, D. W. 1973. Quality
     of  drainage  water from  a  heavy  textured  soil.
     Trans. ASAE 16: 1104-1107.

132. Shannon, E. E., and Brezonik, P. L. 1972. Rela-
     tionships between lake trophic state and nitrogen
     and phosphorus loading rates. Environ. Sci. Tech.
     6(8): 719-725.

133. Sievers,  D.  M.,  Lentz, G. L., and Beasley, R. P.
      1970. Movement  of  agricultural fertilizers  and
     organic insecticides in surface runoff. Trans. ASAE
      13(3): 323-325.

134. Smith, G.  E.  1967. Are fertilizers creating pollu-
     tion  problems? In  Soil  and  America's future:
      108-114. Soil Conservation Soc.  Amer., Ankeny,
     la.
135. Smith, G. E. 1969. Control of waterborne pollut-
     ants. Res. Com., Great Plains Agr. Council Publ.
     No.  34, Vol. 1:  147-155.  Kansas State Univ.,
     Manhattan.

136. Smith, G. E., Whitaker, F. 0., and Heinemann, H.
     G. 1974. Losses of fertilizers and pesticides from
     clay  pan  soils.  EPA-660/2-74-068, U. S. Govt.
     Printing Off., Washington, D. C.

137. Smith,  M.  W.  1959. Phosphorus enrichment of
     drainage waters  from farmlands. Jour.  Fisheries
     Res.  Board Canada, 16(6): 887-895.

138. Smith, P. F. 1974. Soil absorption of atmospheric
     nitrogen in Florida. Commun. Soil Sci. and Plant
     Anal. 5(4): 341-353.

139. Smith, S. J., and Stanford, G. 1971. Evaluation of
     a chemical index of soil nitrogen availability. Soil
     Sci. Ill: 228-232.

140. Soil Conservation Society of America.  1973. Con-
     servation tillage. A  handbook  for  farmers.  Soil
     Conservation Soc. Amer., Inc., Ankeny, la.
 141. Stanford, G.  1966. Nitrogen requirements of crops
     for maximum  yield. In  M. H. McVickar, W. P.
     Martin,  I.  E.  Miles,  and W. H.  Tucker,  eds.
     Agricultural anhydrous ammonia; technology and
     use. Soil Sci. Soc. Amer., Madison, Wis.

 142. Stanford, G. 1968. Extractable organic nitrogen
     and nitrogen  mineralization in  soils.  Soil Sci.  106:
     345-351.

 143. Stanford, G.  1973. Rationale  for optimum nitro-
     gen fertilization in corn production. Jour. Environ.
     Qual. 2:  159-166.

 144. Stanford, G., Carter, J. N., and Smith, S. J. 1974.
     Estimates of potentially mineralizable soil nitrogen
     based  on short-term incubations. Soil  Sci.  Soc.
     Amer. Proc. 38:99-102.

 145. Stanford, G., England, C.  B.,  and Taylor, A. W.
     1970.  Fertilizer  use   and water quality.  ARS
     41-168, Agr.  Res. Serv., U. S. Dept. Agr., Washing-
     ton, D. C.
 88

-------
146.  Stanford,  G.,  and  Epstein,  E.  1974.  Nitrogen
     minerali/ation-water  relations  in  soils.  Soil  Sci.
     Soc. Amer. Proc. 38(1): 103-107.

147.  Stanford, G., Frere, M. H., and Schwaninger, D. H.
     1973. Temperature  coefficient of soil nitrogen
     mineralization. Soil Sci. 115: 321-323.

148.  Stanford,  G.,  and Smith,  S.  J.  1972. Nitrogen
     mineralization  potentials of soils.  Soil  Sci. Soc.
     Amer. Proc. 36(3): 465472.

149.  Stevenson, F. J. 1965. Origin  and  distribution of
     nitrogen in soil. /// W. V. Bartholomew and F. E.
     Clark,  eds.  Soil nitrogen.  Amer.  Soc. Agron.,
     Madison, Wis.

150.  Stewart,  B. A.  1970.  A  look   at  agricultural
     practices in relation to nitrate accumulations. In O.
     P.  Engelstad, ed. Nutrient mobility in soils: accum-
     ulation and  losses: 47-60.  Soil Sci. Soc.  Amer.,
     Inc., Madison, Wis.

151.  Stewart, B. A.,  Mathers, A. C., and Thomas, J. D.
     1974. Manure  effects on rate  of advance, intake,
     and quality  of runoff from irrigated Pullman clay
     loam. Agron. Abstr., 170.

152.  Stewart, B. A., Viets, F. G., Jr., and Hutchinson,
     G. L. 1968. Agriculture's effect on nitrate pollu-
     tion of groundwater. Jour. Soil and  Water Conserv.
     23(1): 13-15.

153.  Stewart, K.  M., and  Rohlich, G. A. 1967. Eutro-
     phication—a review. Calif. State Water Qual. Con-
     trol Board, Sacramento, No. 34, 188 pp.

154.  Taylor, A. W., and Kunishi, H. M. 1971. Phosphate
     equilibria on stream sediment and  soil in a water-
     shed draining  an  agricultural region. Jour. Agr.
     FoodChem. 19:827-831.

155.  Thomas, A.  W., Carreker, J. R., and Carter, R. L.
     1969. Water, soil, and  nutrient losses on Tifton
     loamy sand. Univ. Georgia, College of Agr. Expt.
     Sta. Res. Bui. 64.

156.  Thomas, G. W.  1970.  Soil and climatic  factors
     which affect nutrient mobility. In  Orvis P. Engel-
     stad, ed. Nutrient mobility in soils: accumulation
     and  losses:  1-20.  Soil  Sci.  Soc.  Amer.,  Inc.,
     Madison, Wis.
157. Thomas, G.  W., and  Barfield,  B. J.  1974. The
     unreliability of tile effluent for monitoring subsur-
     face  nitrate-nitrogen  losses from soils. Jour. Envi-
     ron. Qual. 3(2): 183-185.

158. Thomas, G. W., Blevins, R. L., Phillips, R. E.,and
     McMahon,  M.  A.  1973.  Effect  of a killed sod
     mulch on nitrate movement and corn yield. Agron.
     Jour. 65(5): 736-739.

159. Thomas, G. W., and Hanway, J. 1968. Determining
     fertilizer needs. In L. B. Nelson, M.  H. McVickar,
     R. D. Munson, L. F. Seatz, S. L. Tisdale, and W. C.
     White, eds.  Changing  patterns in  fertilizer  use:
     119-140. Soil Sci. Soc. Amer., Inc., Madison, Wis.

160. Thomas, G. W., and  Peaslee, D. F. 1973. Testing
     soils  for phosphorus. In  L. M.  Walsh and  J. D.
     Beaton,  eds.  Soil  testing  and  plant  analysis:
     115-132. Soil Sci. Soc. Amer., Inc., Madison, Wis.

161. Timmons, D. R., Burwell, R.  E., and Holt. R. F
     1968. Loss of crop nutrients through runoff. Minn.
     Sci. 24(4): 16-18.

162. Timmons, D. R., Burwell, R.  E., and Holt, R. F.
     1973. Nitrogen  and  phosphorus losses in surface
     runoff  from  agricultural  land as influenced by
     placement of broadcast fertilizer. Water Resour.
     Res. 9(3): 658-667.

163. Timmons, D. R., Holt, R. F., and  Latterell, J. J.
     1970. Leaching  of crop  residues as a source of
     nutrients in surface  runoff water. Water Resour.
     Res.  6(5): 1367-1375.

164. U. S.  Environmental  Protection Agency.  1973.
     Measures for  the restoration and enhancement of
     quality of freshwater lakes. EPA-430/9-73-005. U.
     S. Govt. Printing Off., Washington, D. C.

165. U. S.  Environmental  Protection Agency.  1974.
     Relationships between drainage area characteristics
     and nonpoint source  nutrients in streams. Working
     Paper No. 25,  Pacific Northwest  Environ.  Res.
     Lab., Corvallis, Ore.

166. Uttormark, P. D., Chapin, J. D., and Green, K. M.
     1974. Estimating nutrient loadings of lakes from
     non-point sources.  EPA-660/3-74-020. U. S. Govt.
     Printing Off., Washington,  D. C.
                                                                                                        89

-------
167. Viets, F. G., Jr.  1971. Fertilizer use in relation to
     surface  and  ground  water pollution. In  R.  A.
     Olson, T. J. Army, J. J. Hanway, and V. J. Kilmer,
     eds.  Fertilizer technology and use: 517-532. Soil
     Sci. Soc. Amer.,  Inc., Madison, Wis.

168. Viets, F. G., Jr.  1971. Water quality in relation to
     farm use of fertilizer. BioScience 21: 460-467.

169. Vollenweider, R. A. 1971. Scientific fundamentals
     of the eutrophication of lakes and flowing waters,
     with  particular  reference to  nitrogen and  phos-
     phorus  as  factors in  eutrophication. Organ, for
     Econ. Coop, and Devlpmt., Paris.

170. Wadleigh,  C.  H. 1968. Wastes  in  relation  to
     agriculture and forestry. Misc.  Pub. No. 1065. U.S.
     Dept. Agr., Washington, D. C.

171. Wagner, R. E., and Jones, M. B. 1968. Fertilization
     of high yielding forage crops. In L. B. Nelson, M.
     H.  McVickar, R. D.  Munson, L. F. Seatz, S. L.
     Tisdale, and W.  C. White, eds.  Changing patterns in
     fertilizer  use:  297.  Soil Sci. Soc.  Amer.,  Inc.,
     Madison, Wis.

172. Walsh,  L. M. 1973. Soil and applied nitrogen. Fact
     Sheet No. A2519, Univ. Wisconsin - Extension.

173. Wang, Wun-Cheng. 1974. Adsorption of phosphate
     by  river particulate matter.  Water Res. Bui. 10(4):
     662-671.

174. Weibel, S  R., Weidner, R.  B., Cohen, J. M., and
     Christiansen,  A. G.  1966. Pesticides and  other
     contaminants in rainfall and  runoff.  Jour. Amer.
     Water Works Assoc. 58(8): 1075-1084.

 175. Welch,  L.  F.  1974.  Nitrogen  on soybeans.  Crops
     and Soils 26(9): 9-10.

 176. White,  E. M. 1973. Water-leachable nutrients from
      frozen  or  dried prairie vegetation. Jour. Environ.
      Qual. 2(1): 104-107.

 177.  Wild, A.  1972.  Nitrate leaching under bare fallow
      at a site in northern Nigeria. Jour. Soil Sci. 23(3):
      315-324.
178. Wildung, R. E., Schmidt, R. L., and Gahler, A. R.
     1974.  The  phosphorus status  of eutrophic lake
     sediments as related  to changes in limnological
     conditions    total  inorganic  and organic phos-
     phorus. Jour. Environ. Qual. 3(2): 133-138.

179. Williams, J.  D. H., and Mayer, T. 1972. Effects of
     sediment  diagenesis and  regeneration  of phos-
     phorus with special reference to Lakes Erie and
     Ontario.  In  H. E.  Allen and  J.  R. Kramer, eds.
     Nutrients in natural waters: 281-315. John Wiley
     and Sons, New York.

180. Willrich,  T.  L. and Smith, E. G. 1970. Agricultural
     practices and water  quality. Iowa State Univ. Press,
     Ames.

181. Winton, E. F. 1970. The health effects of nitrates
     in water. In Nitrate and water supply: source and
     control.  Proc.  Twelfth  Sanitary Engin. Conf.,
     Urbana, 111.

182. Yaalon,  D.  H. 1965.  Downward  movement and
     distribution of anions  in soil profiles with limited
     wetting.  In  Experimental pedology:  157-164.
     William Cloves and Sons, Ltd., London.

183. Yeck, R. G., Smith, L.  W., and Calvert.C. C. 1975.
     Recovery of nutrients from animal  wastes   an
     overview of existing options and potentials for use
     in  feed.  Proc.  Internatl.  Symp.  on  Livestock
     Wastes, Univ. Illinois, Urbana.

184. Zwerman, P. J., Bouldin, D. R., Grewelding, T. E.,
     KJausner, S. D., Lathwell, D. J., and Wilson, D. O.
     1971.  Management of nutrients  on agricultural
     land  for improved water  quality. Project  No.
     13020 DPB, Cornell University, New York. Rep.
     No. PB209-858, Natl. Tech. Inform. Serv., Spring-
     field, Va.


185. Zwerman, P. J., Klausner, S. D., Bouldin, D. R.,
     and  Ellis, D. 1972. Surface runoff nutrient  losses
     from  various  land disposal  systems  for  dairy
     manure.  In  Waste management research: 495-502.
     Cornell Univ., Ithaca, N. Y.
 90

-------
                                              CHAPTER 5

                             PESTICIDES IN AGRICULTURAL RUNOFF

                                                 J. H. Caro
   The  use  of "chemical  pesticides  has  resulted in
enormous benefits to mankind, chiefly in the areas of
public  health  and  agricultural production.  In  public
health, insect  control programs have saved millions of
lives by  combatting  such  diseases as malaria, yellow
fever, and typhus. In India, for example, the use of DDT
has  been credited  with  increasing  the average  life
expectancy from  32 to 47 years  of age. In Sri Lanka,
annual malaria cases dropped from 2 million in 1950 to
17 in 1963; when use of DDT was then discontinued,
the number of cases immediately rose and again reached
1 million  in 1968 (45). Agricultural benefits are  many:
pesticide chemicals have promoted higher  crop  yields
and  improved  quality of produce; aided the mechani-
zation of agricultural production, with substantial reduc-
tion  in labor requirements; and have helped to improve
the utilization  and management of land. Use of agricul-
tural pesticides also has resulted in important economic
benefits to both the farmer and the consumer of food
and fiber. Insecticides are widely estimated to return $5
to the farmer for  every $1 expended (127), which often
tips  the scales to economic  profit from a crop  rather
than economic loss. Agricultural products would prob-
ably cost the consumer two or three times more than at
present if the use of chemicals were eliminated (150).
   Despite  all  the  far-reaching  benefits,  the use of
pesticides  has brought about  a  conflict  of interest
because of the possibility of harmful impact on environ-
mental  quality.  Conservationists  have  often indicted
pesticide residues as being responsible  for a variety of
injurious effects,  including fish kills, reproductive  fail-
ures in  birds,  and acute illnesses in man and animals.
Agricultural applications have just as often been charged
with being primary sources from  which  the chemicals
dissipate into the environment. Although acute adverse
occurrences have indeed taken place, the  sources  of the
damaging pesticides have been a matter of some dispute.
With respect  to chronic effects, the true significance of
low  residue  levels  of most  pesticides  in  the general
environment  resulting from long-term use of the chem-
icals is still  not  well  understood.  In  any event, it  is
widely expected that the use of chemical pesticides will
remain an  integral  part  of agricultural technology for
many years and will in fact increase at least through the
next decade. Consequently,  information on the path-
ways by which pesticides  leave  the  site of application
and distribute throughout the environment will continue
to be actively sought so that appropriate controls can be
instituted.
   One  such  pathway is the movement of pesticides
away  from  treated  fields in  runoff  water  and  on
sediment carried  along in  the water. In Volume I, we
presented guidelines for identifying areas  of potential
pollution problems arising from this movement and also
described appropriate  pesticide  management practices
that would alleviate the problems.  In this  chapter, we
will  provide  documentation  to support  the recom-
mended practices. We will  also indicate  the size of the
potential problem by showing the extent of agricultural
use  of  pesticides, and  we will examine the  state of
knowledge  concerning  pesticide transport  in runoff.
Related areas to be covered include (1) information on
pesticide persistence in  soil, which  affects the  relation-
ship between amounts of residues moved in runoff and
the time elapsed since application of the pesticide to the
field; (2) characteristic levels of pesticides found in the
aquatic ecosystem;  and (3) the impact of pesticides on
aquatic organisms, which will permit some assessment by
the reader  of potential hazards of the reported levels. In
addition, we  will summarize information on methods for
removing pesticide  residues from  the  aquatic  environ-
ment,  and we  will show  the areas within  the  broad
subject of pesticides  in  runoff that  clearly require
additional research.
                                                                                                        91

-------
                EXTENT AND TRENDS IN USE OF AGRICULTURAL PESTICIDES
   Because of the demands of our increasing population
for space for cities, roadways, and recreational areas, and
because economic trends have  made  smaller and less
efficient  farm  units unprofitable, the amount of crop-
land that supports each of us has,  until very recent
times, declined steadily since the 1920's (Figure 1). The
same pressures have favored ever more intensive farming
of the  acreage  still cultivated. An important component
of this modem, high-efficiency agriculture is the use of
chemicals to combat the pests  and blights  that attack
our crops and  agricultural products. The insects, weeds,
and  plant diseases that cause  significant  agricultural
damage are many and varied. In  the United States alone,
for example, there are an estimated  10,000 species of
damaging insects and mites, about 600  of which are
serious pests that require control every year (114).
   The chemicals employed  to control these pests have
been highly effective. The general impact of pesticide
use on crop yields is indicated in Table 1, which shows
an  apparent relationship  between rates  of pesticide
application and  crop yields in  major geographic areas.
Although yields may be increased by any of a number of
improved agricultural practices,  the importance of pesti-
cide use is undeniable.
                                 The number of specific chemicals is impressive. The
                              current domestic market for pesticides includes  more
                              than  1800 biologically active compounds sold in  over
                              32,000 different formulations.  In 1971, 833 million
                              pounds of the  compounds  were used in the United
                              States, of which almost 60% were accounted for by use
                              on farms (Table 2). Of  all  pesticides used on farms,
                              herbicides comprised 52%, insecticides 39%, and chemi-
                              cals for control of plant diseases 9%.
                                 The extent and distribution of pesticide use on major
                              crops in  1971 is shown in Table  3. Three crops—corn,
                              soybeans, and cotton—accounted for almost 80%  of all
                              herbicide  use  on farms;  two crops—cotton and corn-
                              accounted  for nearly 70% of insecticide use; and  fruit,
                              nut, and vegetable crops accounted for 85% of fungicide
                              use. The  most  extensively  treated  crop was  peanuts,
                              whereas only small percentages  of alfalfa  acreage were
                              treated. The pests of most concern in specific crops are
                              shown clearly in the table.  Weeds, for example,  are a
                              severe problem in soybeans, but insect damage is limited.
                              Conversely, tobacco generally requires protection from
                              insects but not from weeds.
                                 The geographic distribution of cropland treated  with
                              pesticides is shown  in Table 4. Although the table shows
        400-
                                                               200
        250
                                                               50
           1910
1920
1930
1950
I960      1970
                          1940
                          Year
Figure l.-Land in crops and population in the United States. From Barrens (17).
92

-------
   Table 1. Rates of pesticide application and yields of major
           crops in countries and geographic areas'
   Table 3. Acres of crops grown and percentage treated with
             pesticides, United States, 1971
Country or Area


United States 	
Latin America 	
Oceania 	 ...
India 	
Africa 	

Pesticide
application
rates
Ounces/A
154 0
26 7
21.3
3.1
2.8
2.1
1.8

Yields of
major
crops
Ib/A
4 890
3 050
2 320
1 760
1 400
730
1,080

     From T.nnisetal (61).
   Table 2. Use of pesticides and percentage used by farmers,
                  United States, 19711
Type of pesticide
Herbicides 	
Insecticides 	
Fungicides 	
Total 	

Million pounds
used2
359
319
155
833

Percentage used
by farmers
63
53
27
59

   1 From Andrilenas (5).
   2 Active ingredients.
data for 1969, the relative distribution has probably not
changed  significantly in  succeeding  years. By far the
largest acreage  treated  is in  the two North Central
regions, which basically  comprise the Corn  Belt  and
much of the Wheat Belt. These 12 states contain 63% of
the  cropland  receiving  herbicides  and  47%  of that
receiving insecticides. The treated area is lowest  in the
Northeast (New England  and Middle Atlantic regions).
   In the  United  States, the  use  of pesticides  has
continued to increase. Total annual  sales for domestic
use  exceeded  one  billion  pounds  in  1973 (6), as
contrasted with  833 million  pounds in 1971  (Table 2),
and much of the  increase is undoubtedly attributable to
expansion in agricultural  demand. There are  two main
reasons for this. Within the past year or two, farm prices
have risen, so that the crop is worth more to the farmer
and he will  tend to apply pesticides  at a lower level of
infestation to protect it. Second, and also as a result of
changing  farm economics,  land  is  once  more  being
converted to cropland after a long period in the reverse
direction (Figure 1). The U.S. Department of Agricul-
ture has projected an annual  increase in cropland within
the near  future of approximately 4%. This corresponds
to over 3.5  million  additional acres  of corn,  of which
about 57% will require herbicides and 33% insecticides,
                                                                 Crop
                     M..|.     Percentage of total acres
                      acres   treated with Pesticides for
                      	'    control of

Alfalfa 	
Corn 	
Cotton 	
Fruit crops

Rice 	
Small grains1 . . . .
Sorghum 	
Soybeans 	
Sugarbeets . . . .
Tobacco . .
Vegetable crops . . .

27.5
74.0
12.4
4 0
1 5
1.8
91.7
20.8
43.5
1.4
0.8
4.8
Weeds
1
79
82
28
92
95
37
46
68
75
7
43
Insects
8
35
61
81
87
35
5
39
8
30
77
62
Diseases
<0.5
1
<0.5
53
85

<1
<0.5
2
13
7
27
                                                            1 Adapted from Andrilenas (5).
                                                            2 Includes wheat, oats, barley, rye, and mixed grains.
                                                              Table 4. Cropland acreage treated with pesticides, by
                                                                          geographic region, 19691
                                                                    Region
                             Cropland acreage treated for
                                      control of
                                Weeds
Insects'*
New England 	 ,
Middle Atlantic 	 ,
East North Central 	
West North Central . .
South Atlantic 	 ,
East South Central 	
West South Central . . . . ,
Mountain 	
Pacific 	 ,

United States 	

1000 acres
. . . . 260
. . . . 1645
. . . . 20060
. . . 33375
. . . 4030
. . . . 4265
. . . . 9990
. . . . 5655
. . . . 5635

84915

1000 acres
245
675
7890
10885
4^10
2670
7545
1985
3780

39880

   1 From U.S. Bureau of the Census (154).
     States in the regions are:
      NE:   ME,NH,VT,MA,RI,CT
      MA:   NY.NJ.PA
     ENC:   OH,IN,IL,MI,WI
    WNC:   MN.IA.MO.ND.SD.NB.KS
      SA:   DE,MD,VA,WV,NC,SC,GA,FL
     ESC:   KY,TN,AL,MS
     WSC:   AR, LA, OK, TX
      MT:   MT, ID, WY, CO, NM, AZ, UT, NV
      PA:   WA, OR, CA, AK, HI
     Not including land for hay crops.


and more than 8  million  additional acres of wheat, of
which  about  2.5 million  acres  will  need  herbicide
treatment.
   Some economists estimate that the use of chemical
pesticides will increase up to 15% annually over the next
few years (45) because of the economic situation and
                                                                                                            93

-------
potential fuel shortages. To save fuel, fanners may well
replace band treatment with broadcast applications, with
a resultant increase in pesticide applied per acre. Within
the growing market, the  patterns  of use  of specific
chemicals will differ from the present because of shifts
in  cropping patterns  and  pest  populations and  the
buildup of  resistance in  target species  of pests.  In the
North Central States, for  example, there  has  been a
marked movement away from  production of extensive
crops (wheat, oats, barley, rye,  flax) and toward row
crops (corn,  soybeans). Between 1961 and 1973, row
crops increased  from 58  to  80 million acres, whereas
extensive crops fell from 49 million to 34 million (141).
Weed populations  will change because the varieties that
are more easily  controlled will be selectively removed.
Despite the difficulties in  moving chemicals through the
registration process, new pesticides will continue to be
introduced into the marketplace to combat stubborn
weeds and  to counteract pest  resistance to older chem-
icals.
                       DISSIPATION OF PESTICIDES FROM TREATED LANDS
   During  application  to soil or foliage, pesticides may
 be lost in  spray drift or by volatilization; after applica-
 tion, they  disappear  from  the  site  of application by
 various pathways. The chemicals may undergo biological
 or chemical degradation; on foliage or the  soil surface,
 they may  degrade under the action of sunlight or they
 may evaporate; they may be taken up into the plant and
 removed in the harvested crop;  they may be  adsorbed
 onto soil particles and  moved off the treated area in
 eroded material; or they may dissolve  in rainwater (or
 irrigation water)  and  move  away  in surface runoff or
 down  through the soil in the soil solution, perhaps later
 to reappear in surface runoff or groundwater. The  rates
 of   disappearance and  the  fractions moving  by  each
 pathway depend primarily on the properties and formu-
 lation of the  pesticide;  the  type, microbial population,
 moisture level, and type of management of the soil; the
 extent  and intensity  of rainfall; and  the  soil and air
 temperatures.
   The movement in runoff water, eroded sediment, and
 subsurface water  is of direct concern here  and will be
 examined in detail. A number of excellent  reviews are
 available that  deal with the other pathways:  Kaufman
 (98) on microbial degradation of pesticides, Crosby (54)
 and Armstrong and Konrad  (9) on nonbiological degra-
 dation, Spencer  (145) and  Guenzi and Beard (76) on
 pesticide  volatilization,  Caro (33) and  Nash (120) on
 uptake of insecticides by plants, and Foy et al (71) on
 uptake of herbicides by plants.


 Factors Influencing Pathway of Movement into
                  Water Courses

 Adsorption and Solubility

   The  pathway that a pesticide takes in its movement
 away from the site of application through the action of
 water-that  is, whether  it  moves  in runoff water or
eroding sediment  or leaches down  into  the soil-is
governed by sorption and solution equilibria that depend
primarily  on the water solubility of the chemical,  the
degree and strength of its adsorption on soil, and on the
interaction of both soil  and pesticide with water (144).
Generally, compounds that  are more water-soluble will
move primarily in  runoff water and those more strongly
adsorbed  will move  mostly on sediment. An inverse
relationship exists between  solubility  and extent  of
adsorption, but only within families of  compounds.
Some pesticides, such as paraquat and diquat, are very
water-soluble  but   will  move  only on  the  sediment
because of strong, irreversible  adsorption; others have
low water solubility but will nevertheless move in  the
water except  when applied to a rather adsorptive  soil
(153).
   Soil  characteristics are  clearly  very important  in
determining the degree  of adsorption, as illustrated by
the fact that adsorptivity of a pesticide can vary  by as
much as  1 S-fold over a range of soil types. The  most
important  soil  property that  influences  the  way a
pesticide  partitions between soil and water, as deter-
mined by a number of investigators,  is  the  organic
matter  content, which usually  gives  a  good  direct
correlation with the degree  of adsorption.  Other  prop-
erties that  may be important are  the  acidity, cation
exchange  capacity, moisture content,  temperature, and
clay mineral content. With some pesticides such as  the
triazine and triazole herbicides,  adsorption  depends
primarily  on soil acidity: in acid soils, they associate
with free  hydrogen ions to  form cations  that adsorb
strongly  to the negatively  charged soil; in neutral  or
alkaline soils, they are in molecular form and are held
much more weakly by  the  soil  (163). The acid herbi-
cides, such as  2,4-D and picloram, also adsorb  more
strongly  in acid soils. Bailey and White (13) showed
correlation coefficients for  a number  of soil properties
and adsorption  as part of a  comprehensive review of the
adsorption and desorption of pesticides in soil.
94

-------
   Water  in  the  soil  competes  with  pesticides  for
adsorption sites on soil particles, so that as the moisture
level in  the soil  decreases,  the fraction of the chemical
adsorbed increases (82, p. 100). Rain falling on a dry soil
will therefore desorb  a  portion of the pesticide, which
would then move with the  water in any ensuing runoff.
Generally,  adsorption  decreases  as  soil temperature
increases,  and the response to  changes in temperature
becomes less  as  the adsorption  bonds weaken (82, p.
l>7).
   Pesticides adsorb  preferentially on  smaller soil par-
ticles because of the high surface area per unit weight of
these particles. When  runoff occurs,  the small particles
are transported greater distances than coarser material.
Since rill and sheet erosion primarily involve surface soil,
such  erosion will tend to favor movement of the more
strongly  adsorbed  pesticides  (707,  p.  431).  Higher
pesticide  concentrations in eroded  material  do not
necessarily mean that gross losses will be greater in the
sediment than in runoff water; the reverse sometimes is .
true because the amounts of water moved are so much
greater (134, 166).

teachability

   Pesticides in the soil or on its surface may move down
through  the soil profile dissolved in water. The principal
factors affecting the  movement are  the  same as those
controlling overland movement-adsorption and solubi-
lity-because a pesticide is partitioned between soil and
water in leaching as well as in runoff. Other  parameters
influencing leaching are water flow rate and amount, and
the formulation, concentration, and rate of degradation
of the pesticide (89). The  correlation of solubility and
adsorption with  pesticide movement  through soil sug-
gests that solubility  may  be  important in the  initial
movement  from  the  point of application,  whereas
adsorption may  be  the determining process  in  later
movement. Therefore, adsorption will, in general, be a
better indicator of overall potential movement than will
solubility  (21). As examples of  this,  prometryne (48
ppm  water solubility) moves less in soil than simazine (5
ppm) because it is more strongly adsorbed, and monuron
(230 ppm) moves about the same  as atrazine (33 ppm)
for the same reason (85).
   The pesticide moves downward through  the profile
either by mass flow of water from impacting rainfall or
by molecular diffusion  in  the soil solution.  Diffusion,
which is influenced by bulk density and temperature in
addition  to soil  moisture,  is slow in comparison with
mass flow and  is  important only over short  distances
(-/),  so  that  mass  flow  is  the primary  means  of
movement  under  most  conditions.  Water  does not,
however,  continuously  move downward, except in very
high rainfall areas.  Evaporation  at  the surface  causes
upward movement of subsurface water and its comple-
ment  of dissolved pesticides, which then concentrate at
the surface (107, p. 42°). Pesticides in the water can also
move  laterally  when they encounter  a /.one of water
saturation  or when  they  reach the  boundary  between
two areas  of different  soil moisture, since  water will
move  laterally  into  the drier soil (59). If the  laterally
moving water intercepts the sloping surface of the land,
the dissolved pesticide  will join the  overland tlow and
appear in surface runoff.
   Despite occasional reports of low-level groundwatcr
contamination by pesticides, measurements have  not in
general shown that groundwater pollution by leaching of
pesticides  through soil  is extensive or significant. Much
excess  water must be applied even to a relatively mobile
chemical  to move   it  deeply  into  the profile.  Many
reported  findings in  experiments  with  picloram, a
relatively  teachable herbicide, are in agreement: except
in sandy soils, picloram does not  leach below the  2-foot
depth  (107, p.  430).  For a  more strongly adsorbed
pesticide such as  dieldrin,  several hundred years  would
be  required  for  the  chemical  to  be transported  in
solution at a residual concentration of 20 ppb to a depth
of  1   foot in neutral  soils   (67).  The groundwatcr
contamination  that  does occur  may be  caused  by
pesticides  being carried on soil particles washed  down
into deep  cracks in the soil in drought-breaking rainfalls
(125. 169).

Formulation
   The formulation in  which a pesticide is applied also
may affect the pathway  of  movement, especially  if
runoff  occurs  shortly  after   application, before   the
chemical has equilibrated  with the  soil. For example,
ester formulations of the herbicide  2,4-D applied to a
sandy  loam soil in a set of experiments (16) were  tar
more susceptible to washoff than an  amine salt formula-
tion. The  amine formed a true  solution with water and
leached into the  soil,  whereas the  relatively insoluble
esters were adsorbed  and moved on the eroded sediment.

   Factors Influencing Amounts of Pesticides
            Moved  into Water Courses

   The  quantity  of a pesticide  moving  into  a  water
course from a  treated  area in  any given runoff occur-
rence  depends  on a number of  associated  factors. Its
relationship to topography, intensity  and duration of
rainfall,  soil crodibility,  and  land  management  and
cropping  practices   are discussed in  the  chapters  on
erosion and runoff.  Obviously, the amount moved will
                                                                                                         95

-------
increase with the amount of pesticide initially applied to
the area. It also depends on the following parameters.


Time After Application

   Characteristically,  pesticide  losses are  highest in the
first runoff occurring after application of the chemical,
and the magnitude of the loss generally decreases as time
between application and runoff increases. The effect of
elapsed time is particularly  noticeable with short-lived
pesticides and with pesticides that are not incorporated
into the soil. Concentrations of the chemical in subse-
quent runoff events decrease  at  a rate that  depends
largely  on the persistence of  the  pesticide  in the soil.
Field experiments with  the  carbamate insecticides car-
baryl and carbofuran showed  that pesticide concentra-
tions in both runoff water and sediments  in  the  third
runoff, which occurred  within 1  or 2 months  after
application, were less than  5% of the concentrations in
the first runoff (35, 36). By contrast, concentrations of
the persistent insecticide dieldrin in the third runoff, 3
or 4  months after application, were about  15% (water)
and over 30% (sediment) of  those in the  first runoff
(34). The pattern of relatively high concentrations in the
first runoff, decreasing with time to eventually negligible
concentrations in succeeding runoffs, has been noted in
experiments with  many  pesticides. Quantitative  ex-
amples are illustrated below in the section on pesticide
levels in runoff.

Persistence in Soil

   As mentioned, the persistence  of a pesticide in soil
affects the  change with time in amounts lost in runoff.
However, many  factors influence  the persistence  of an
individual pesticide and, consequently, it can be  quite
variable.  Picloram, for example, has been  reported in
specific instances to  effectively disappear from the soil
in as little as 50 days (110) or as long as 6 years (30),
but its persistence under moderate conditions is gen-
erally about  1.5 years.
   A pesticide applied to the soil is subject to a sequence
of overlapping loss processes—application losses, volati-
lization, sorption, leaching, and eventually chemical and
biological  degradation  (92). As a  result, the loss rate
changes  rapidly during the  early  period  when  the
chemical is distributing and equilibrating in the soil, then
becomes  nearly  constant over a  relatively  longer  time.
This  later rate is dictated by many conditions, including
among others the weather; cultural practices; and type,
temperature, moisture  level,  and acidity of the soil.
Pesticides that are subject to microbial degradation will
have reduced persistence when applied to an area that
had received an earlier application of the same chemical
because  of growth in  populations  of  active   micro-
organisms after the first treatment (99).  If, as is often
the case, more than one pesticide is applied, interactions
between the chemicals may  also  markedly alter the
persistence of the individual compounds in both soil and
aquatic environments (97).

Antecedent Soil Moisture

   Some pesticides will have greater losses  in runoff if
applied to wet soil than if applied to dry  soil,  partic-
ularly  if runoff occurs soon  after application. Experi-
ments  showing this have been reported for  2,4-D (16)
and for fluometuron (75). The effect is probably related
to  the competition of water with  the  pesticide for
adsorption sites on the soil particles.
Proximity to Water Course

   Sloping cropland  rarely abuts  continuous streams.
Consequently, pesticide-containing  runoff usually must
traverse some untreated land before reaching the water.
This intervening area  can trap  some  of the pesticide,
resulting in lowered contamination  of the stream. Large
decreases can be obtained. In one set of measurements,
flow over only 5 feet of untreated soil with slopes of 3%
or 8% reduced picloram losses in runoff from small field
plots more than 50% (151); in another case, dieldrin
applied at  10 to 20 times normal levels to strips of land
12  to  15 feet away from  the edges of ponds, with
shallow to steep slopes intervening, did not appear in the
pond water or bottom mud,  except for very low (0.3
ppm) contamination of the mud when the pesticide was
left on the surface of the soil (58).

Placement of the Pesticide

   The most important effect  of  pesticide placement
with  respect  to environmental  contamination is that
soil-incorporated pesticides will  not be lost in runoff to
as great an  extent as those applied and  left on the
surface or  sprayed  on foliage. The subject is discussed in
more detail in section 4.4 of Volume I.
96

-------
                   PERSISTENCE AND FATE OF PESTICIDE RESIDUES IN THE
                                      AQUATIC ENVIRONMENT
      Distribution of Pesticides on Entering
                  Water Bodies

   A pesticide carried in agricultural runoff entering a
receiving water body-stream,  pond, or lake-will dis-
tribute within the aquatic system in a manner and at a
rate that depends primarily on whether the chemical is
initially dissolved in the water or adsorbed on particles
of eroded soil suspended in the water.
   A dissolved  pesticide will be  diluted  in  the larger
volume  of water and  will be subject to processes  that
dissipate it.  In  a flowing stream, it  will  simply be
transported away from the point  of  entry, later to
undergo degradation  or removal  from  the water. In a
pond or lake, it may  sorb or concentrate  in algae and
aquatic  vegetation  or  it  may  attach  to  suspended
sediment and other particulates in the  water such as
bacterial floes, diatoms, and general organic or inorganic
fragmentary  material.  In either case,  it is  eventually
deposited  on the bottom  of the lake  unless  it  is
chemically or biologically degraded before it reaches the
bottom or is taken up by living organisms. The sorption
processes appear to be generally quite rapid and  effi-
cient, as shown  in  measurements of organochlorine
insecticide sorption on algae (91) and  bacterial  floes
(106).  Highly soluble pesticides that are  only  weakly
adsorbed may be hydrolyzed or biologically degraded in
solution at a rate that depends on the types and numbers
of microorganisms in the water.
   The fate of pesticides entering water bodies adsorbed
on sediment has been  discussed in detail by Pionke and
Chesters (126). The pesticide will distribute first with
the carrying  sediment,  then  will equilibrate with the
remainder of the aquatic system. Sediments entering
water bodies  will segregate on a particle-size basis: in a
stream,  the fractionation will depend on stream velocity;
in a lake, the particles will settle on the bed in decreasing
order of particle size.  The finer particles containing the
highest  concentrations of pesticides will be transported
farthest and will be localized in a stream; in a lake, these
particles will settle last and remain at the water-sediment
interface. In  large,  thermally stratified lakes,  density
currents  may control the movement  and  mixing of
incoming sediments and settling may be very slow.
   Conditions in the water body may affect the  adsorp-
tion and desorption of the pesticide on  the sediment. If
the pH of the lake or stream is higher than that of the
inflow, desorption of acidic compounds (such as 2,4-D,
2,4,5-T, or picloram) or weakly basic compounds (such
as the triazine  or urea herbicides) will be favored, and
the reverse will be  true if the pH of the body is lower
than that of the inflow. Salinity in the lake will favor
adsorption of acidic pesticides  and desorption  of basic
pesticides, but  the  effect  is generally minor. A lower
temperature in the  water  body will increase pesticide
adsorptivity, but this too is a minor effect under field
conditions (126). If there is any oil pollution in the
water, adsorbed oil will  significantly concentrate  the
pesticides on the sediment (87).

           Post-Distribution Processes

   Pesticides  never  reach  true equilibrium in  water
bodies because the  systems are dynamic,  with  many
processes continually operating  to remove the chemicals
from  the system at  rates that change  as conditions
change. Pesticides  sorbed  on  bottom  muds may  be
churned  up and  carried along with sediment  during
periods of turbulent  flow  or they may remain where
originally   deposited.  Since  sorption  on  participate
matter is generally reversible, the bottom muds provide a
continuous supply of desorbed pesticides to the over-
lying  water.  In  the  water,  the  pesticides may  be
chemically or biologically degraded, they may reach the
surface and volatilize, or they may be decomposed near
the surface by the action of sunlight. (Sunlight energy is,
however, probably too weak to induce much photode-
gradation  in  natural waters.)  At the surface of  the
bottom muds,  pesticide  degradation is extensive.  Be-
cause  organic matter accumulates there, it is an area of
high microbial  activity. The microbial populations may
consume so much oxygen that the environment becomes
anaerobic,  a condition that favors the  degradation of
many pesticides. For example, most chlorinated hydro-
carbon insecticides,  although normally highly persistent,
will  degrade  at an  appreciable rate under anaerobic
conditions when the temperature is 20° C or higher (91).

  Pesticide Persistence in Aquatic Environments

   Measurements have been made  of the persistence in
aquatic systems of a large number of specific pesticides,
as summarized  by Pionke and Chesters (126), Paris  and
Lewis  (123),  and  Echelberger and  Lichtenberg (60),
                                                                                                        97

-------
among others. Information of a more general nature is
presented here.

Organophosphorus Insecticides

   As a class,  organophosphorus insecticides are among
the  less hazardous  pesticides  in the aquatic system
because they  hydrolyze  rapidly.  Within  the pH  range
6.0-8.5, which covers most systems, most organophos-
phorus compounds hydrolyze within 8 to 12 days, with
hydrolysis  occurring both in water  and  on sediments
(118, 126). Parathion is  an exception, being somewhat
resistant to chemical hydrolysis. It is, however, readily
susceptible to  microbial degradation in both aerobic and
anaerobic environments.  In the  absence  of an active
microbial population, parathion remains in  the aquatic
environment for several months; in the  presence  of an
active  population, it is degraded  in a matter of weeks
(75).
Organochlorine Insecticides

   As noted earlier, these compounds will degrade very
slowly, if at all, in aerobic  aquatic  systems, but will
decompose  more  rapidly  in anaerobic environments
through the action of microorganisms. The degradation,
which generally involves a simple dechlorination of the
molecule, may require up to several months for comple-
tion in natural systems. There is, however, some varia-
tion among the individual members of the class, persist-
ence  increasing  in  the   following  order:  lindane,
heptachlor,  endrin,  DDT,  DDD, aldrin,  heptachlor
epoxide, dieldrin (90, 91).

Other Pesticides

   The  carbamate  insecticides  have  been  shown  to
degrade in  slightly  alkaline  river water in  less than 4
weeks (60), but since the stability of these compounds is
pH-dependent, appreciably longer persistence would be
expected in a more acid environment.
   The decomposition of many  herbicides is primarily
biological.  In  aerobic lake water, for example, 2,4-D
persisted for up to  120 days, whereas in lake muds, it
was  substantially decomposed in  24 hours  once the
microbial populations had adapted to  the chemical (2).
The  herbicide  dicamba also  dissipates from water most
rapidly under nonsterile conditions. The rate of disap-
pearance depends greatly on the  temperature, especially
in the presence of sediments containing active microbial
populations (139).
          CHARACTERISTIC LEVELS OF PESTICIDES IN THE AQUATIC ECOSYSTEM
   When pesticide-containing runoff occurs from agricul-
 tural land, the chemicals are quickly diluted in the water
 bodies  receiving  the  runoff and  are  also partitioned
 among  the various components of the  environment-
 water, bottom  sediments, and living organisms—so that
 each component eventually bears a concentration of the
 pesticide. The magnitudes involved have been measured
 by many investigators under a variety of conditions and
 are summarized here, including levels in the runoff itself
 and in drainage streams, farm ponds, lakes, and oceans,
 so  that some appreciation may be  gained of the impact
 of agricultural activities on water quality.
   The  importance of  maintaining  a constant surveil-
 lance of the aquatic system in the United States has been
 recognized by  the Federal government. Comprehensive
 programs for continuous monitoring of pesticides in fish,
 estuarine shellfish, water, and bottom sediments, among
 other  components of the general environment,  are
 conducted by various agencies to establish baseline levels
 and to  signal significant trends.  The overall program is
 coordinated by an interagency committee (122). With
 respect to  the quality of our waters, standards have been
 set for  acceptable limits of certain pesticides in drinking
water  (Table  5). An obvious goal of pesticide control
programs is  to assure that the natural  waters of  the
country  are sufficiently pesticide-free to drink without
purification.

            Pesticide Levels in Runoff

   Concentrations of pesticides leaving treated fields in
runoff water  and entrained sediments during  the crop
season following application are almost always measur-
able, so  that there is little doubt that agriculture does
contribute to the pesticide residues found in the general
aquatic  ecosystem.   Both  concentrations  and  gross
amounts lost  depend on numerous factors, including
among others the  intensity of rainfall, the time after
pesticide application that  runoff occurs,  and the mode
of application. Concentrations of a given pesticide may
therefore differ substantially in runoff occurrences at
separate  locations  under  different  sets of conditions.
Nevertheless,  it is  useful to examine the results of
specific  measurements reported in the literature, not
only because  an understanding may be  gained of  the
orders of magnitude  involved, but  also  because  some
 98

-------
Table 5. Recommended limits for pesticides in drinking water1
             Pesticide
Recommended
   Limit
            Organochlorine Insecticides

Chlordane 	
DDT 	
Dieldrin 	
Endrin 	
Heptachloi 	
Heptachloi epoxide 	
Lindane 	
Methoxychlor 	
Toxaphene 	

ppb
	 1
	 3
.... 50
. . . 1
.... 0 5(0 2)2
. . . . 01
. . . . 01
	 "•' .
.... 5(4)
. . . 1000(1 OO)2
. . 5

                Phenoxy Herbicides
2, 4-D 	
Silvex 	
2,4,5-T 	

	 20(1 OO)2
	 30(10)
	 2

   1 From Environmental Protection Agency (62).
    Numbers in parentheses indicate limits proposed by the
Environmental Protection Agency to take effect in December
1976.
informative general conclusions may be drawn. Table 6
lists a  number of such measurements made on water-
sheds under normal agricultural conditions.
   The figures in the table show that pesticide concen-
trations in runoff are generally, but not always, highest
in the  first runoff occurrence following application of
the chemical. Presumably, small flows can occasionally
produce higher concentrations than more intensive flows
preceding  them.  Concentrations  are always  lower in
runoffs occurring  later in the season, irrespective of
which pesticide is applied, and the reduction is generally
greater for water-borne pesticides than for those carried
on sediment. The table also shows that the organochlo-
rine insecticides,  as  is well  known, adsorb to a great
extent  onto sediments, leaving only very low concentra-
tions in water. Control of erosion will therefore reduce
the movement  of these chemicals substantially. Atrazine,
on the  other hand, moves with both water and sediment.
In general, the concentrations of pesticides in runoff are
considerably above drinking water standards (Table 5)
and must be diluted substantially in drainage streams to
avoid acute harmful effects on aquatic organisms in the
streams. Finally,  the table  shows, particularly in  the
-ases of carbaryl and  picloram,  how widely runoff
concentration  of  a pesticide  may vary under different
conditions.
   Though amounts  of  a pesticide lost in runoff may
vary among specific treatments, it is almost always true
that the gross loss over the course of the year following
application will represent only a small percentage of the
amount  of chemical  that  had  been applied to  the
cropland. Results of  watershed  and  field-plot experi-
ments reported in the literature (Table 7) show the wide
applicability of the relationship. Except for one experi-
ment with atrazine in which a heavy rain occurred one
week after  application, total amounts lost  in  runoff
water and sediment were always 5 percent or less of the
application. This  appears to hold true irrespective of the
soil type or degree of incorporation of the pesticide into
the  soil,  and applies  even on  relatively steep slopes.
Simulated rainfall experiments, in which the amount and
intensity of water falling onto  a small sloping plot can be
controlled, show  that  rainfalls of the size and intensity
that might be expected to occur every year produce only
small percentage  losses of pesticides in runoff, but that
heavier  "10-year" rains may cause  larger losses (11,
166).
                        Pesticide Levels in Drainage Streams

                    Almost all the measurements that have been made of
                  pesticide  concentrations in  flowing streams draining
                  treated  areas  indicate  that, as might be  expected,
                  concentrations are  substantially lower  than  in  direct
                  runoff. With many herbicides, residues were  always
                  below detectable limits  in waters a few hundred yards
                  below sprayed  areas  (66,  78).  Even where  detected,
                  levels of such herbicides as 2,4-D, 2,4,5-T, dicamba, and
                  picloram  were well below  median  tolerance limits for
                  trout. In one instance, the herbicide  fenuron was applied
                  at a high rate of 23 Ib/A along stream channels. Only
                  2.4% of the application was lost in the stream water over
                  a 27-month period, with a maximum concentration of
                  430  ppb  following a  heavy rain  (56).  Results  for
                  insecticides were much the same as those for herbicides.
                  Phosphorus insecticides  were  just at detection levels in
                  drainage streams in California (14).  Chlorinated insecti-
                  cides were generally at  low  but measurable levels.  In
                  streams  draining sugarcane  fields,  waters  contained  a
                  maximum of 820 ppt (parts per trillion) of endrin over a
                  4-year period, generally decreasing to 30 to 40 ppt  3
                  months  after treatment. Streambed material  averaged
                  100 ppb, with levels decreasing as the season progressed.
                  Dieldrin,  BHC, and DDT were also found in the waters
                  in the parts-per-trillion range (104).  In drainage streams
                  from a commercial orchard that had received substantial
                  applications of Organochlorine insecticides, no residues
                  were found in the  water, but detectable levels of DDT
                  compounds,  dieldrin,  and endrin appeared in the silt,
                  organic debris, and  bottom organisms (117).
                                                                                                         99

-------
8
                                              Table 6. Characteristic concentrations of pesticides in runoff:  maxima and rates of decrease
Pesticide
Chlorinated
Insecticides
DDT 	
DDT 	
Dieldrin 	
Dieldrin 	
Dieldrin 	
Endosulfan . .
Endrin 	
Endrin 	
Endrin 	
Methoxychlor ....
Other
Insecticides
Carbaryl 	
Carbaryl 	
Phorate 	
Herbicides

Atra^ine .......
Fluometuron 	
Picloram 	
Picloram 	
Picloram 	
Picloram 	
2,4,5-T 	
2,4,5-T 	

Application
rate
Ib/A
15
065
1.5
5 0
5 0
0.31
0 3
0.3
025
200
45
15
0.67
8.0
1 0
4.0
30
2.0
1.0
025
3.0
1 2

Runoff occurrence giving Maximum concentration
maximum concentration
Number after
application
1st
1st
1st
2nd
1st
1st
1st
2nd
1st
1st
1st
1st
3rd
1st
3rd
1st
3rd
3rd
Days after
application
1
13
13
4
2
6
1
18
17
24
24
86
30
6
10
86
22
Runoff occurrence giving
reduced concentration
In water In sediment Num*er **" Dl»? ^ er
application application
ppb ppb
70.0 30,000
83.01
70.0 30,000
20.0
14,200
19.01
2.73
5.02
49.01
8.8
248 12,200
1220
19.0
4600 6,200
700 950
870
19.0
14.4
89.7
17.0
287
380
5th
3rd
10th
3rd
3rd
3rd
4th
3rd
2nd
2nd
4th
3rd
8th
2nd
4th
5th
46
18
82
14
27
23
33
29
35
35
197
98
30
20
197
38
Reduced concentration
In water In sediment
ppb ppb
1.0 10,000
7.01
1.0 30,000
6.7
, 5,000
2.01
0.53
2.88
8.01
1.0
8.4 80.0
980 3,300
180 550
9.0
LO
6^0
50
Citation

Haan(77)
Epstein and Grant (63)
Haan (77)
Caro era/ (34)
Caro et al (34)
Epstein and Grant (63)
Willis and Hamilton (169)
Willis and Hamilton (169)
Epstein and Grant (63)
Edwards and Glass (59)
Caro, Freeman, and Turner
(36)
Fahey (68)
Fahey (68)
Hall, Pawlus, and Higgins (81)
Hall, Pawlus, and Higgins (81 )
Wiese (167)
Bovey et al (26)
Baur, Bovey, and Merkle U9)
Baur, Bovey, and Merkle (19)
Baur, Bovey, and Merkle (19)
Bovey if a/ (26)
Edwards and Glass (59)
          1 Water-sediment mixture.

-------
                      Table 7. Percentages of applied pesticides lost in runoff1 in field experiments
Pesticide
Atrazine 	




Carbofuran . . .
DDT 	
DDT 	
Dieldrin 	
Dieldrin ... . . . .




Methyl parathion . . .
Methyl parathion . . .
Propachlor 	
Toxaphene . . . .
Trifluialin 	
Trifluralin 	

Incorpo-
rated depth
In.
... 0
	 0
	 0
	 2
... 3
	 2
	 0
	 0
.... 3
	 3
.... 0
	 0
.... 0
	 0
.... 0
	 0
	 0
.... 0
.... 6
.... 6

Soil texture
Silty clay loam
Silty clay loam
Silt loam
Silt loam
Silt loam
Silt loam
Loamy sand
Gravelly loam
Silt loam
Silt loam
Gravelly loam
Gravelly loam
Silty clay loam
Various
Loamy sand
Sandy loam
Silt loam
Loamy sand
Loamy sand
Sandy loam

Slope
%
14
14
10-15
10
9
10
2-4
8
14
10
8
8
02
0 1-4
4
2
10-15
2-4
4
2

Pesticide
in runoff
% ofappln.
4.8-5.0
2.6
2.5-15.9
0 1
0.9
1.9
1.0-2.8
0.7
2.3
0.02
0.25-0.35
0.01-1.0
0 1
<3.0
0.01-0.02
0.13-0.25
3.1
0.4-0.6
03-05
0.5-0.8

Citation
Hall (80)
Hall, Pawlus, and Higgins (81)
Ritterera/034)
Caro, Freeman, and Turner (36)
Caro et al (35)
Caro et al (35)
Bradley, Sheets, and Jackson (27)
Epstein and Grant (63)
Caro et al (34)
Caro et al (34)
Epstein and Grant (63)
Epstein and Grant (63)
Willis and Hamilton (169)
Wiese(167)
Sheets, Bradley, and Jackson (142)
Sheets, Bradley, and Jackson (142)
Ritterera/(134)
Bradley Sheets, and Jackson (27)
Sheets Bradley and Jackson (142)
Sheets Bradley, and Jackson (142)

   1
     Both water and sediment.
         Pesticide Levels in Farm Ponds

   Concentrations of pesticides in the waters of farm
ponds adjacent  to treated areas are clearly sensitive to
the amount of pesticide applied in the area and the
length of time between application and the first heavy
rain. In a pond near cotton plots, for example, DDT and
toxaphene  concentrations in  the water  were always
significant after application and were  especially high
when intense  rain  closely  followed  the  application.
Concentrations  ranged  from  0.4 to 13.4 ppb for DDT
and  from  2.9   to  65.2  ppb  for toxaphene  (142).
Similarly, in measurements of the herbicide picloram in
ponds at several locations in Texas, concentrations in the
water ranged from 55 to 184 ppb if the first rainfall
occurred within 2 weeks after application, but were only
2 to 29 ppb if the first rainfall was delayed for 6 weeks.
Concentrations  in all  cases dropped to  1  ppb or less
within 6 months (78).
   The few measurements that have been made indicate
that farm ponds are not always contaminated despite
proximity to treated areas. In Virginia in  1966, when use
of organochlorine insecticides was high, heptachlor was
1 >und in the water of only 10 of 35 ponds examined, at
levels up to 5 ppb. In the bottom muds, the conversion
product heptachlor epoxide  was  found in 14 of the
ponds at levels of 1  to 60 ppb. As  is evident, the
majority of the ponds  contained no detectable residues
(162).  In  another set of measurements, the waters of a
farm  pond near  a  5-lb/A treatment  with  carbaryl
contained no  detectable residues  throughout the crop
season, and a pond near a 0.67  Ib/A treatment with
phorate showed a maximum of only 4 ppb, becoming
undetectable later in  the season (68). Of course, ponds
or any water body can receive relatively heavy doses of
pesticides by drift or inadvertent direct spray from aerial
applications.

            Pesticide Levels in  Rivers

   Virtually no published  information is  available on
pesticide residues in  river waters  of the United  States
showing concentrations occurring later than about 1970.
Nevertheless, the pattern is clear: contamination  of the
streams peaked about  1966 (Figure 2) and then de-
creased steadily, probably right up to the present, with
decline in domestic  use of the most significant contami-
nants,  the organochlorine  insecticides. Concentrations
have always been low and are often at trace levels, which
are less than about  1 to 5 ppt for most of the chemicals.
   The national situation is perhaps best summarized by
examining the results of a few broad-scale monitoring
studies. Lichtenberg et  al (108) combined five annual
synoptic surveys of U.S. rivers for 1964 through 1968.
Except for dieldrin in 1964, no pesticide appeared in
more than 40% of the  samples, and the frequency of
                                                                                                        101

-------
                                 Dieldrin

 0
25

 0
25
 0
       r  no
       |  data
                                 Heptachlor
       i
                                  Lindane 8 BHC
                                          WTTTft
          no
r
l data
 no
data
Chlordane
         1964   1965    1966    1967   1968
Figure 2.-Percent positive occurrences of ten organochlorine
   insecticides in river waters of the United  States, 1964-68.
   From Lichtenberg era/ (108).
positive occurrences declined sharply after 1966 (Figure
2).  Maximum concentrations found during the 5-year
period for the 10 compounds shown in the figure ranged
from 0.84 ppb for ODD to 0.048 ppb for heptachlor. An
intensive examination of Mississippi and Missouri River
drinking water samples during the same  period (137)
showed frequencies of occurrence similar to those of the
broader surveys: over 40% of the samples were positive
for dieldrin,  over 30% for endrin and total DDT, and
20%  for  chlordane.  Little  or  no aldrin, heptachlor,
toxaphene, or methoxychlor was found. For  the later
period  from  1968  through  1971,  measurements  of
organochlorine insecticides and three herbicides (2,4-D,
2,4,5-T, silvex)  in  waters taken  from 20 stations in
Western  U.S. rivers  (138)  showed  the  characteristic
decline in positive insecticide occurrences. In  1967-68,
there were 165 insecticide occurrences and 70 herbicide
                                                occurrences; corresponding figures in  1970-71 were 53
                                                and 54. Maximum concentrations found were 0.46 ppb
                                                DDT and 0.99 ppb 2,4-D. In sum, contamination of river
                                                waters by pesticides is at low  levels,  sporadic, and
                                                decreasing. Even at worst, residues  were well below
                                                acceptable limits for drinking water (Table 5).
                                                   Pesticide  residues in  riverbed sediments  are  also
                                                relatively  low. In the Mississippi River, high concentra-
                                                tions  were found only in sediments located just below
                                                pesticide  manufacturing and formulating plants (18).
                                                Large amounts of organochlorine insecticides applied to
                                                crops in the Mississippi River Delta did not contaminate
                                                streambed sediments widely (112). Dieldrin and endrin
                                                occurred in only 18 to 20% of the sediments taken from
                                                a Louisiana estuary in 1968-69; maximum levels were 4
                                                or 5 ppb (136).  In  Rhode Island streams, measurements
                                                made about 1970 showed that DDT and its metabolites
                                                occurred  in  almost  all  sediments  at  concentrations
                                                generally  below  500 ppb. Chlordane and  dieldrin were
                                                also found, but not as often and usually at less than 50
                                                ppb (133). The reported results suggest  that riverbed
                                                surfaces of streams draining extensive row-crop farmland
                                                could be contaminated by locally used pesticides at low
                                                part-per-million levels.
             Pesticide Levels in Lakes

   Pesticide concentrations are generally lower in large
lakes than in rivers. Water samples from  Lake Erie in
1971-72 mostly showed no measurable residues. In the
few  positive samples  obtained, low levels of diazinon,
dieldrin, atrazine,  and simazine were  found. It  was
concluded  that the contribution by agricultural sources
to pesticide  pollution of the lake was negligible  and
insignificant  (156). In 1967-68 in Lake Poinsett,  the
largest  natural lake  in South Dakota, DDT  and its
metabolites  predominated,  averaging 80 ppt in   the
water,  with  lesser amounts  of  other  organochlorine
insecticides appearing in most of the samples (83). The
waters in Lake Michigan contained only 2 or  3 ppt of
organochlorine insecticides, chiefly the DDT family, in
1969-70. Upper sediment layers contained median con-
centrations of 18.5 ppb total DDT and 2.0 ppb dieldrin,
with traces of heptachlor epoxide and lindane (105).
   Being large,  diverse impoundments, lakes  provide
excellent environments for measurement of the process
of biological magnification of pesticides through food
chains.  Two  studies clearly illustrate the effect. In one,
DDT in Lake Michigan appeared at levels of about 2 ppt
in the  water, 14 ppb  in the  bottom muds, 410 ppb in
sand fleas,  3  to 6 ppm in fish, and as much as 99 ppm in
herring gulls, which are near the top of the chain (114).
102

-------
In the second study, in which organochlorine insecti-
cides  in Lake  Poinsett were  measured, the effect was
considerably smaller. Residues found were in the ratio:
water 1, bottom sediments 18, crayfish  18, zooplankton
37, algae 37, and fish 790 (83). Whatever the magnitude,
biological magnification must  be considered when evalu-
ating  the  significance of very low  concentrations of
persistent pesticides in waters and bottom sediments.

            Pesticide Levels in Oceans
   Pesticides in the  marine environment originate from
many sources and it is clearly not possible to define the
contribution from agricultural runoff.  Taken together,
however, all sources have contaminated the ocean to
only  extremely low  levels, and  only in coastal areas.
DDT, the  most ubiquitous pesticide, is undetectable in
Atlantic deep-sea water or sediments, but does appear in
water, sediments, shellfish, and  finned fish along the
coast. However, concentrations in shellfish have declined
to insignificant levels since the curtailment of DDT use
(88)  and,  although not reported, concentrations in the
 other  components  of  the  ocean  environment  have
 probably  also declined.  Where  they  occur,  concentra-
 tions are highest in surface slicks; in 1968, coastal slicks
 contained up  to  13 ppb of the organochlorine insecti-
 cides,  whereas the underlying waters were at low ppt
 levels  (140). Whole seawater off the Pacific Coast in
 1970 also contained DDT, but  only  at a maximum of
 less than 6 ppt (50).
   The pattern of decreasing DDT concentration with
 distance from  shore and with depth was confirmed in a
 more recent investigation of Pacific Ocean waters (168).
 DDT concentrations in surface film  samples taken in
 1971 and 1972 were 11-15 ppt in coastal waters, 0.4 ppt
 in the  offshore California current, and less than 0.02 ppt
 in the North  Central Pacific. In subsurface waters, by
 contrast,  DDT levels  were  less than  0.01  ppt in  the
 North Central Pacific and only  0.1 ppt in the offshore
 current. The measurements also  showed that concentra-
 tions of polychlorinated biphenyls were ubiquitous and
 were as much as two orders of magnitude higher than
 those of DDT.
                   IMPACT OF PESTICIDES ON THE AQUATIC ENVIRONMENT
   System planners  can  best evaluate  the  need for
instituting controls for pesticides in agricultural runoff if
they understand  the  effects that the chemicals may
produce in the downstream aquatic  environment. The
information presented here will emphasize some major
effects but does  not cover all that  is known  on the
subject.
   A  variety  of  pesticide-induced effects  on  aquatic
birds, fish, aquatic plants, invertebrates and microorgan-
isms has been documented. Acute responses have been
measured quantitatively and subtle effects produced by
long-term  exposure to  low, sublethal pesticide concen-
trations  have   been   recognized, even though  not
measured directly. Perhaps the most  prominent effects
have been the catastrophic fish kills during the 1960's in
which  pesticides  were implicated, though  never posi-
tively established  as the cause. One such took place in
the Mississippi River, presumably caused by endrin; a
second occurred  in the Rhine River and was attributed
to endosulfan. These and similar kills probably resulted
from  high pesticide  concentrations emanating  from a
point source rather  than  from  nonpoint  agricultural
runoff. Runoff was, however, probably partly responsi-
ble for contamination of Lake Michigan salmon by DDT
in 1969 that, though not  fatal to the fish, resulted in
widespread confiscation of the commercial  catch, with
an economic loss estimated at $3 to $4 million (95).
   Unless pesticide concentrations are very  high, re-
sponses of organisms  in a particular aquatic system are
extremely  difficult to  predict  because  of the great
variability and natural complexity of ecosystems and the
assortment of environmental insults that  man imposes
on the systems. Ponds,  lakes, and streams vary in their
content of water, dissolved salts, temperatures, acidity,
and nature and populations of plants and animals on the
bottom.  Pesticide  contamination in  small bodies  of
water, for  example, has  been judged  to  be  more
transitory and less serious than in large bodies because of
greater bottom surface area per unit  volume of water,
higher flushing rates, and greater biological activity (95).
Furthermore, different  forms of the active pesticidal
ingredient may differ in toxicity, and additives in  the
formulation, such as wetting agents or binders, may be
more  toxic to aquatic organisms than  the active ingredi-
ent itself (119).
   In  general,  herbicides  are   less   toxic  to aquatic
organisms than insecticides (Table 8), though there are a
number  of exceptions  (Vol. I, Tables  8a  and  9a).
Herbicides, especially those applied directly, also have
desirable as well as  undesirable effects on  water bodies.
They  may be responsible for opening and maintenance
of navigable waterways, saving  of  irrigation  water,
increase in aesthetic and monetary value of waterfront
property, control of mosquitoes and snails by removal of
                                                                                                       103

-------
     Table 8.  Relative toxicity of selected pesticides to
                  aquatic organisms1
     Pesticides
                               Organism
                  Plankton  Shrimp Crab   Oyster  Fish
Herbicides 	
Phosphorus ....
Insecticides
DDT. 	

1
0.5
2

1 i
1000 800
400 200

1
1
300

1
2
700

     From Butler (31). Based on the arbitrary assignment of
the value of 1 to herbicides.
aquatic weeds, restoration  of recreational waters, im-
provement in  fish  management,  and elimination of
flavors and odors from algal blooms. On the debit side,
many herbicides are acutely toxic to fish (Vol. I, Table
8a).  Serious losses of fish and other aquatic fauna may
also  occur when herbicides kill  aquatic weeds, which
gravitate  to  the  bottom  and  decompose, removing
necessary oxygen from the water. Moreover, phenols
resulting from the hydrolysis of phenoxy herbicides such
as 2,4-D may impart objectionable flavors  and odors to
water. In comparison  with insecticides, however, the
hazards  of herbicides in the aquatic  environment are
small.  Most  herbicides  have  little  or no toxicity to
humans, wildlife, or livestock; they may  reduce phy-
toplankton populations initially,  but recovery generally
occurs  within  2  or 3  weeks; they  do  not undergo
biological magnification in food chains, and shellfish are
tolerant  of them,  accumulating residues  only  tempo-
rarily after exposure (72).

                  Acute Toxicity

   The toxicity of a pesticide to fish is affected by
numerous parameters, including the size, age, and species
of the fish; water temperature and acidity; and physical
differences  at  the  aquatic  site.  Survival  time  after
exposure generally correlates directly with body weight,
probably because the smaller fish consume a  propor-
tionately greater  diet and have  less fat for  storage
detoxification. Higher water temperature increases toxi-
city of some pesticides and decreases it for others. DDT
and methoxychlor  are examples of insecticides that are
less toxic at higher temperature; toxaphene,  endrin,
malathion, and parathion are more toxic (32). The effect
can be very pronounced, as shown in Table 9. There are
differences in response within species as well as between
species.  Diquat,  for  example,  was  toxic  to  female
mosquitofish under conditions in which the males were
not affected (170).
   It is important to recognize that the high toxicity to
fish  of some  herbicides (trifluralin,  for example)  is
tempered  in nature by processes  that inactivate the
compound,  such  as  strong  adsorption  and  relative
immobility  on soil surfaces,  so that contamination of
natural water  bodies is minimal if the chemical is used
according to label directions.
   Another  aspect of toxicity to fish is the development
of resistance to pesticides, which has been documented
for several  species  of fish.  For  example,  fish in a
contaminated  lake in Mississippi had higher tolerance for
endrin, DDT,  and toxaphene than  those in a relatively
clean lake (22). Toxicity will  also result in the flourish-
ing of one group of organisms in an aquatic environment
while others are suppressed. Thus, elimination of algae-
eating species will produce increases  in  algae  and
anaerobic bacteria populations (130).

                 Chronic Toxicity

   Fish  mortalities  have been  observed  to occur in
nature by long-term, low-level exposure to pesticides.
Numerous pathological effects on the tissues and organs
of fish have also been  noted, including lesions of liver
and gills and  changes  in the  intestines, kidneys, brain,
and blood.  However,   some  chemicals—notably meth-
oxychlor and  carbamate insecticides such as carbofuran
and carbaryl—are  rapidly hydrolyzed on ingestion and
therefore do not have chronic effects.

         Fish Reproduction and Growth

   Persistent pesticides such  as DDT and dieldrin can
have strong adverse effects on reproduction in fish. In
one typical case, the  hatching success of landlocked
Atlantic  salmon from  a lake contaminated with DDT
was 36% lower than  in control fish (109). Mortality
usually  occurs in  salmon sac-fry during the period of
  Table 9. Effect of water temperature and exposure time on
           the toxicity of trifluralin to bluegills1
Water temperature
V
85
75
65
55
45
48-Hour
LC50
tig/liter
8.4
66
200
380
590
24-Hour
LC50
tig/liter
10
120
360
530
1300
   1 From Cope (48).
 104

-------
yolk-sac absorption. Some herbicides also may produce
reproduction  problems in  fish,  causing  atrophy of
spermatic  tubules  and  production of abnormal  sper-
matozoa (48). In addition to reproductive failures, loss
of appetite and restricted growth have been reported to
result from exposure of fish to pesticides.


                  Fish Behavior

   Cases have been reported  of changes in the condi-
tioned responses and  locomotor patterns  of fish as a
result of exposure to  pesticides. One  well-documented
effect is an increase in sensitivity to low water tempera-
tures, including active avoidance, in  fish exposed to DDT
(32).

            Effects on Aquatic  Plants

   Herbicides, by their very  nature, are more toxic to
aquatic plants than insecticides, but adverse effects do
not  always occur  where  they might be expected.
Phytoplankton are sometimes unaffected by pesticides,
sometimes multiply when predators are removed by the
chemicals,   and  sometimes are  seriously  inhibited in
growth, depending on  the particular conditions at hand.
One apparent and obviously serious effect is the disrup-
tion of photosynthesis in phytoplankton: in a compre-
hensive series  of tests,  carbon fixation  by  estuarine
phytoplankton  was  reduced  by 45 of 54 chemicals
tested, with reductions  of over 90% for several of the
compounds (158). Some  aquatic plants act as concen-
trating agents for the organochlorine insecticides, so that
when the  plants die, concentrations of the pesticides, as
well as of  plant  nutrients, are released into the water.

                  Odor and Taste

   Several of the organochlorine insecticides,  including
toxaphene, endrin, and heptachlor, impart objectionable
odors to water  at concentrations of only a few parts per
billion; a  number of herbicides generate  a strong odor;
and  the solvents  used in many formulations are highly
odorous in concentrations as low as 16 ppb (131). The
herbicide  2,4-D hydrolyzes in water to 4-chlorophenol
and  2,4-dichlorophenol,  both of  which  impart dis-
pleasing flavors and odors at low ppb levels. However,
the phenols are only rarely detected in natural waters
(72). Decomposing aquatic plants that have been killed
by herbicides are  a major source of foul odors in aquatic
environments.
               REMOVAL OF PESTICIDES FROM THE AQUATIC ENVIRONMENT
   Obviously,  no removal  of pesticide residues  from
bodies of water  would be needed if chemicals that are
rapidly degradable in the aquatic environment were the
only ones being used, but such is not the case. Pesticides
that are  relatively persistent  pose  a threat to water
quality that has prompted investigations of methods for
their removal,  chiefly in connection with the protection
of  drinking water supplies. Residues can be removed
directly from water bodies by such measures as dredging
sediments and  removing weeds, debris, and coarse fish.
However, these methods are generally not economically
feasible and the  method that is generally  followed for
nondrinking waters is to simply allow a period of time
for natural renovation to occur.
   With public water supplies, available data indicate
that present-day conventional water-treatment processes,
such  as  lime-alum  coagulation,  sedimentation,  sand
filtration, chlorination, and  pH adjustment, will reduce
high pesticide  levels substantially, but are inadequate for
removal of chronic contamination at low levels (39). The
degree of removal depends on the water solubility and
adsorptivity of  the  individual pesticides, with more
efficient  removal for compounds of low  solubility or
high adsorptivity. In one series of tests,  for example,
conventional treatment eliminated less than 10% of the
lindane and only 20% of the parathion in the water, but
removed 55% of the dieldrin, 63% of the 2,4,5-T ester,
and 98% of the DDT (135).
   Much research effort has been expended on adsorb-
ents to purify the water beyond the levels attainable by
conventional treatment. Activated carbon is clearly the
most effective adsorbent for pesticides, having a removal
efficiency about 4 orders of magnitude greater than that
of soil, 3 orders greater than that of algae, and 2 orders
greater  than that of coal  (100). The effectiveness of
removal depends on the contact time, the concentration
of activated carbon, the concentration of the pesticides,
and the presence of organic material  in the water that
may compete with the  pesticides for adsorption sites on
the carbon. If the water is clarified by other processes
before the activated carbon is introduced, the competi-
tive organic matter  can be at least partially controlled
(121). Removal efficiency decreases  at pesticide  con-
centrations  below 1 ppb. Although  it is possible to
remove  lower concentrations of organochlorine insecti-
cides, inordinately large amounts of carbon are required
(39).  At the 1-ppb  level, organochlorine insecticides in
drinking water account for only about 5% of the dietary
                                                                                                       105

-------
intake of these pesticides and would not pose a threat to
human health, at4east with respect to acute effects (94).
   Other techniques for pesticide removal from water
have been explored, with limited success. Chemical or
biological oxidation will degrade some  pesticides, but
toxic products are formed in many cases (39); ozone will
attack even the stable organochlorine insecticides, but
only at large  and impractical concentrations and with
unknown products  (135);  and  strongly basic  anion-
exchange resins will remove high concentrations of such
pesticides  as 2,4-D salts (3). Use of  reverse osmosis
membranes has shown promise for removal of a wide
variety of pesticides (40). Other effective processes may
yet be developed that will  take advantage of specific
characteristics  of  individual pesticides,  one possible
example being the introduction of microorganisms that
are specific  for rapid inactivation of certain classes of
chemicals.
                       PRACTICES FOR REDUCING ENTRY OF PESTICIDES
                                INTO THE AQUATIC ENVIRONMENT
   A total  of 15 pesticide management practices, desig-
nated as P 1 through P 15, are presented in Volume I
(Table  18  and  Section  4.4).  TTiese practices reduce
pesticide losses in runoff from treated fields by manipu-
lation  of the chemical itself and are meant to supple-
ment the basic control of runoff and erosion carrying
the pesticide, which is dealt with in other sections  of
Volume  I.  The various aspects of the practices are
discussed in  Volume I without supportive  documen-
tation.  Appropriate documentation  for each  of the
practices is, however, presented in Table 10.
   Table 10  contains citations of articles supporting
direct  statements made in Volume  I  and of articles
containing  closely related information that may  be of
benefit in evaluating the individual practices. Informa-
tion for both Volume I and this volume was obtained
not only from the published papers cited, but also from
discussions and meetings with agricultural and pesticide
specialists and  from internal  progress  reports of the
CRIS  (Current-Research-in-Science)  information  re-
trieval  system of the U.S. Department of Agriculture and
the State agricultural experiment stations.
                                           RESEARCH NEEDS
   The length of this review constitutes first-hand evi-
 dence that much is already known about pesticides in
 the aquatic environment, yet  numerous important ave-
 nues for future research remain. Further efforts should,
 of course, be directed to minimization of those agricul-
 tural  practices that contribute to  erosion and  runoff
 from  cropland and thereby produce excessive losses of
 applied chemicals, but our concern here is with aspects
 of the system that deal with the pesticides directly. Such
 needs  appear to fall into five general areas: (1) predic-
 tion of pesticide behavior in the aquatic ecosystem; (2)
 definition of significance of residues occurring in water
 bodies; (3) investigation of means for lowering rates and
 frequency  of  application  of  pesticides, so  that  the
 potential for contamination of waters would be lessened;
 (4)  development  of  new  pesticides  having environ-
 mentally favorable  properties; and (5) research on
 corrective measures to reduce  or remove contamination
 by applied pesticides. No order of priority among these
 is intended in the brief discussions that follow.

         Prediction of Pesticide  Behavior

   The development of mathematical models  to predict
 the behavior of pesticides after application is a relatively
new and important area of research. The ideal model is
one that is able to predict the consequences of any given
practice  or set  of  conditions;  applying  it, one  can
identify optimum modes of pesticide use with respect to
some particular attribute. However, a large amount of
work is  required  to construct useful models  of  the
typically complex agricultural systems. Moreover, data
collection could be  a limiting step; a complete model
may demand  such an  elaborate  input of hydrologic,
chemical, biologic, and management data that collection
of real-world numbers, including analyses, might take so
long that events would outrun predictions.
   Despite  these difficulties, development and refine-
ment of exploratory models is being actively pursued.
With respect to pesticide movement in runoff, a model
has been developed  with  the  objective of minimizing
water pollution (12, 53).  It takes into account condi-
tions both during and  between runoff events,  and has
given satisfactory predictions for  pesticides moved en-
tirely on sediment, but not yet for those moved in both
water and  sediment. Models in related areas have also
been proposed. The movement of  agricultural chemicals
through the soil profile has been  described mathemati-
cally, but existing models do not take into account the
ongoing  natural soil-forming processes, so that their
106

-------
                       Table 10. Bibliography on pesticide management practices (Volume I, Section 4.4)
          Pesticide management practice
 No.
Page No.
   in
 Vol. I.
Description
                                                          Citations
                                                              Significant subjects
PI
           85
             Using Alternative Pesticides
P2
           85
             Optimizing Pesticide
               Placement With Respect
               to Loss
P3
   85
             Using Crop Rotation
                     Craig et al (52)
                                                 Bailey (10)
                                                 Spencer (144)
                                                 Erbach and Lovely (64)
                     CaroeM/(35)
                                                 Ritterefa/(134)
                                                 Apple (1)
                                                 Constien ef a/(4J)
                                                 Moomaw and Robison (116)
                                                 Reid and Peacock (132)

                                                 Erbach, Lovely, and
                                                  Bockhop(65)

                                                 Bode and Gebhardt (25)
                     Wax (159)

                     Epstein and Grant (6j)

                     Wax (159)

                     Stockdale, DeWitt, and
                       Ryan (147)

                     Daniels (55)

                     Fleming (6J), p. 327
                                                 Blakely, Coyle, and Steele
                                                   (23), p. 305

                                                 Wade (155)
Typical examples of alternative pesticides
   effective against same pests in same crops

Discussion of pesticide properties pertinent
   to movement in runoff

Discussion of pesticide properties pertinent
   to movement in runoff

Desirability of rotating equally effective
   pesticides in succeeding years on same crop

Comparative pesticide loss in runoff:  in-furrow
   vs. broadcast applications
Comparative pesticide loss in runoff:  ridge
   planting vs. surface-contour planting

Comparative toxicity to crop seed of insecti-
   cides when placed in seed furrow

Necessity for placing insecticides in seed furrow
   in no-till management

Satisfactory performance of herbicides placed
   in narrow bands

Subsurface sweep applicators

High efficiency of precise spacing of herbicides


Advantage of disk over other implements in
   incorporating pesticides to minimize loss
   in runoff

Current trends toward broadcast application

Lower runoff loss of pesticides from rotation

Weed reduction by crop rotation

Improved insect control by crop rotation


Improved insect control by crop rotation

Typical examples of improved insect control
   by crop rotation

Reduction in erosional losses by rotation
                                                                    Intercropping of corn and peanuts to reduce
                                                                       attack by corn borers
                                                                                                                     107

-------
                                                 Table 10. (continued)
 No.
          Pesticide Management Practice
Page No.
   in
 Vol. I.
       Description
                                                          Citations
                                                                  Significant subjects
P4
P5
P6
P7
   86
           86
   86
   87
Using Resistant Crop
   Varieties
             Optimizing Crop Planting
                Time
Optimizing Pesticide
   Formulation
Using Mechanical Control
   Methods
                                                 Kuhlman, Cooley, and
                                                  Walt (103)

                                                 AUaway(l),p. 391-2
Sprague and Dahms (146)


Chant (38), p. 204

Hoffman (93)


Fleming (69), p. 328


Craig era/(52), p. 48


Wellhausen (164)


Waxefaf(160)


Bulkhead et al (28)

Wax (159)

           t

Foy and Bingham (70)


Mullison (119)


Barnett effl/(16)


Miles and Woehst (115)

Clack (42)

Depew (57)


Burnside and Colville (29)


Wax (159)
                                                 Whitaker, Heinemann, and
                                                   Wischmeier (165)

                                                 Crafts and Robbins (51),
                                                   p. 140-154

                                                 Behrens (20)
Comparative acreage receiving insecticides:
   continuous crop vs. rotations

Comparative weed control: continuous crop
   vs. rotations

Review of crop resistance to insects and
   reductions in use of insecticides

Wheat varieties resistant to Hessian fly

Examples of resistant crop varieties and insects
   resisted

Time of crop planting: summary of effects
   on insect infestations

Advantages of early plantings for combatting
   European corn borer

Break in sorghum plantings to combat sorghum
   midge

Advantages of late plantings of soybeans to
   combat weeds

Summary of planting dates of field crops

Addition of surfactants to increase penetration
   of herbicides; comparability of liquids and
   granules

Surfactants or oils to enhance herbicide
   penetration in plants

Toxicity of components  of formulations other
   than active ingredient

Comparative runoff potential: 2,4-D esters
   vs. amine salt

Controlled release formulations

Use of foam formulations for weed control

Superiority of granular formulations over
   liquids in seed-furrow applications

Superiority of tillage-herbicide combination
   for weed control

Discussion of tillage and  flame cultivation for
   weed control

Ability of cultivation to  reduce erosive soil
   loss

Discussion of tillage methods in weed control
                                                                    Disadvantages of cultivation and tillage
108

-------
                                                 Table 10. (continued)
          Pesticide management practice
 No.
Page No.
   in
 Vol. I.
      Description
                                                          Citations
                                                                  Significant subjects
P8
   87
P9
   87
P10
   87
Pll
   88
P12
   88
PI 3
Eliminating Excessive
   Treatment
Optimizing Time of Day for
   Pesticide Application
Optimizing Date of Pesticide
   Application
                     Using Integrated Control
                        Programs
                     Using Biological Control
                        Methods
                     Using Lower Pesticide
                        Application Rates
Turnipseedgfa/(152)


Chiang (41)


Shore (143)

Ware et al (157)
Cooperative Exten. Serv.,
  Illinois (47), p. 238

Apple, Walgenbach, and
  Knee (8)
                                                 Harrison and Press (86)

                                                 Anderson (4)
                                                Texas Agr. Exten. Serv.
                                                   (149)
                           Erbach and Lovely (64)


                           Summers, Byrne, and
                             Pimentel (148)

                           Hanson (84)

                           Council Environ. Qual. (49)

                           Chant (38)

                           Casey, Lacewell, and
                             Sterling (37)

                           Giese, Peart, and Huber (73)


                           Knipling (101)


                           Quraishi (129)

                           Putnam and Duke (128)

                           Patti and Camer (124)

                           Tumipseed et al (152)
                                                Casey, Lacewell, and Sterling
                                                   (37)
 Effectiveness of lower than recommended rate
    for insect control in soybeans

 Development of recommendations for deter-
    mining threshold pest damage

 Mathematical computation of optimum dosages

 Higher efficiency of early morning spraying


 Timing of spray to avoid harm to honeybees
Comparative effectiveness: planting-time vs.
   cultivation-time treatments for corn
   rootworm

Timing of sprays against corn borer

Optimization of time of foliar application of
   herbicides

Recommendations for long-interval preplan!
   applications of herbicides on cotton and
   peanuts

Critical period for applying herbicides on corn
   and soybeans

Advantages of early insecticide application
   for alfalfa weevil control

Aspects of integrated pest control programs

Aspects of integrated pest control programs

Aspects of integrated pest control programs

Example of reduction in insecticide use by
   introduction of pest management strategy

Reliability of computer-based pest manage-
   ment systems

Overview of biological control of insects
                                                                   Aspects of biological control of insects

                                                                   Example of biological control of weeds

                                                                   Example of usefulness of Bacillus thuringiensis

                                                                   Examples of effectiveness of rates less than
                                                                      recommended levels

                                                                   Example of lower dosage in an  integrated
                                                                      control program
                                                                                                                    109

-------
                                              Table 10. (continued)
          Pesticide Management Practice
  No.
Page No.
  in
Vol. I.
      Description
                                                      Citations
                                                              Significant subjects
 P14
   89
Managing Aerial Applications
 P15
   89
Planting Between Rows in
   Minimum Tillage
Johnstone (96)



Wax (159)


Marstonefg/(lll)

Cole, Barry, and Frear (43)

Glotfelty and Caro (74)

Coop. Exten. Serv. Ohio (46)


Way (161)
Insecticide application in ultra-low-volume
   sprays


Advantages and disadvantages of aerial
   application of herbicides

Pesticide in stream water from aerial spraying

DDT in environment after aerial application

Movement of airborne pesticides

Practice recommended for reduction of corn
   rootworm populations

Disadvantages of practice
ability  to  predict field  behavior  is limited (24). A
relatively  simple  model  has been developed  (79) in
which the quantities of pesticide to be applied are
optimized with respect to profits to the grower. Refine-
ment of this model could well lead to reduced use of
pesticides and lessened environmental contamination.

             Significance of Residues

   The  true  significance  of  pesticide residues  in the
environment  is perhaps the least  understood aspect of
the system, particularly with regard to chronic contami-
nation at very low concentrations. Changes in behavioral
patterns of aquatic organisms have been observed as a
result of chronic  exposure, but little is truly known of
possible long-term,  subtle effects (118).  To  aid in
assessment of the hazard, we need (1) in-depth studies of
declining species; (2) studies of the gain, loss, or change
in residues in both living and nonliving components of
the environment,  to  relate trends to observable effects;
and (3) lexicological measurements under conditions
that simulate the natural environment more closely than
the conditions used in such tests in the past.
              Reducing Pesticide Use

   Research directed to reduction in use of pesticides by
more efficient application of the chemical or by substi-
tution of nonchemical methods of pest control  offers
                                                the broadest opportunity for decreasing the potential for
                                                environmental contamination by pesticides. One impor-
                                                tant facet that will require considerable future effort is
                                                the  development of large-scale integrated control pro-
                                                grams in which minimum amounts of chemical pesticides
                                                are  used. Much more information  is needed for inte-
                                                grated control than is generally required to use pesticides
                                                alone. A  successful program  for  insect control, for
                                                example, requires knowledge of the dynamics of the pest
                                                population, life history of the pest, natural enemies,
                                                nutritional  requirements, host plants,  economic thres-
                                                hold of the insect population, and behavior of chemicals
                                                and organisms  used in the  program  with respect to
                                                effects on  nontarget environmental components (93,
                                                114).  Each of  these must be  examined in  detail and
                                                interrelated and there  are also  researchable associated
                                                matters, such as the selection of the most suitable of
                                                alternative methods of control for incorporation into a
                                                program, the  use of adverse natural phenomena to signal
                                                the  appropriate time for attacking insects by integrated
                                                control,  and  the  bringing  of  experimentally  proven
                                                integrated control methods up to practical application
                                                (38, 93).
                                                   Several approaches to more efficient application and
                                                utilization of pesticides are being actively investigated,
                                                but require additional effort. One such  is the develop-
                                                ment and testing  of foams, gels, and polymer-encapsu-
                                                lated  slow-release formulations that  can reduce drift,
                                                minimize movement of the pesticide in the environment,
                                                and perhaps decrease the number of applications needed
 110

-------
because the chemical will be used more efficiently. A
second is the use of electrostatically charged sprays to
decrease drift and optimize deposition of the chemicals
onto  plant surfaces. Optimization of  spray droplet
particle sizes for efficient  on-target deposition is  also
being investigated. A third avenue under investigation is
the reduction of pesticide volatilization from plant and
soil surfaces, and a fourth is the precision placement of
pesticides in the soil by devices such as subsurface sweep
applicators. Another aspect worthy of further study is
the  development  of computer  programs  to predict
occurrences of pest infestations. Such a program has
already been successful with potato late blight, saving
growers an average  of 4  sprays annually in comparison
with the normal practice  of spraying at 10-day intervals
(102).

         Development of New Pesticides

   Present pesticides, although effective, are  far from
perfect. Opportunity  is  still  great  to develop  new
chemicals that are highly specific for the pest, safe to
man and wildlife, and have little effect on the quality of
the environment. The ideal compound would have low
solubility in fats to  minimize the possibility of biomag-
nification; would  be biodegradable,  but only after its
intended function is completed; and would be nontoxic,
but convertible to a toxicant in the presence of the pest
(113).
                Corrective Measures

   Research should be conducted on means for shorten-
ing the persistence of relatively stable compounds at the
site  of application or in the aquatic environment by
deliberate manipulation  of their modes of dissipation.
Work of this type would be directed to such goals as
enhancement  of volatility  and photodecomposition or
adjustment of adsorptivity and teachability, perhaps by
use of adjuvants. Other possible corrective efforts could
involve  the use of aquatic plants  as  traps  to  remove
residues  from water  by  absorption, addition  of an
inoculum  of microorganisms to degrade  the  pesticides
after their  biocidal activity is no longer needed, direct
chemical inactivation in  soil by addition  of a reactant,
use of additives to control  the metabolic degradation of
herbicides within plant tissues, and use of new adsorp-
tive  media or ion exchange resins to remove residues
from water (70).
                                                                                                        Ill

-------
                                        LITERATURE CITED
 1. Allaway, W. H. 1957. Cropping systems and soil./w
   Soil,  the  yearbook of agriculture, 1957: 386-395.
   U.S. Dept. Agr., Washington, D.C.

 2. Aly, O. M., and Faust, S. D. 1964. Studies on the
   fate of 2,4-D and ester derivatives in natural surface
   waters. Jour. Agr. Food Chem. 12(6): 541-546.

 3. Aly,  O.  M., and Faust, S. D.  1965.  Removal of
   2,4-dichlorophenoxyacetic  acid  derivatives  from
   natural waters. Jour.  Amer. Water Works Assoc.
   57(2): 221-230.

 4. Anderson, L.  E.  1973. Factors  affecting herbicide
   performance.   Agron.   8-1.  Exten.   Div.,  Univ.
   Missouri, Columbia, 2p.

 5. Andrilenas, P. A. 1974. Farmers' use of pesticides in
    1971. Econ. Res. Serv., U.S. Dept. Agr., Agr. Econ.
   Rep. No. 252, 56 p.

 6. Anonymous. 1974. Pest. Chem. News,  3(17).  Food
   Chem. News, Inc., Washington, D.C.

 7. Apple, J. W.  1971. Response of com to granular
   insecticides applied  to  the row at planting.  Jour.
   Econ. Entomol. 64(5): 1208-1211.

 8. Apple, J. W., Walgenbach, E. T., and  Knee,  W. J.
    1969. Northern corn rootworm control by granular
   insecticide application at planting and cultivation
   time. Jour. Econ. Entomol. 62(5): 1033-1035.

 9. Armstrong, D.  E., and Konrad, J. G. 1974. Nonbio-
   logjcal  degradation  of  pesticides.  In  W.  D.
   Guenzi, ed. Pesticides in soil_and water: 123-131.
   Soil Sci. Soc. Amer., Inc., Madison, Wis.

10. Bailey, G. W.  1966. Entry of biocides into water
   courses. Univ. Calif. Water Resour. Center Rep. 10:
   94-103.

11. Bailey, G. W., Barnett, A. P., Payne, W. R., Jr., and
   Smith, C. N.  1974. Herbicide  runoff from  four
   coastal plain  soil types. EPA-660/2-74-017.  U.S.
   Govt. Printing Off., Washington, D.C., 98 p.

12. Bailey, G. W., Swank, R. R., Jr. and Nicholson, H.
   P.  1974.  Predicting pesticide runoff from agricul-
   tural land: a conceptual model. Jour. Environ.  Qual.
    3(2): 95-102.
13. Bailey, G. W., and White,  J. L. 1964. Review of
    adsorption and desorption of organic pesticides by
    soil colloids, with implications concerning pesticide
    bioactivity. Jour. Agr. Food Chem. 12(4): 324-332.

14. Bailey, T. E., and Hannum, J. R. 1967. Distribution
    of pesticides in  California.  Jour. San. Eng. Div.,
    Proc. Amer. Soc. Civil Eng. 93(SA5):  27.

15. Baldwin, F. L., Santelmann, P. W., and Davidson, J.
    M.  1975. Movement of fluometuron across  and
    through  the  soil.  Jour.  Environ.  Qual.  4(2):
    191-194.

16. Barnett,  A.  P., Hauser, E.  W.,  White, A. W.,  and
    Holladay, J. H. 1967. Loss of 2,4-D in washoff of
    cultivated fallow land. Weeds 15(2): 133-137.

17. Barrens,  K. C. 1971.  Environmental benefits of
    intensive crop production.  Agr.  Sci. Rev. 9(2):
    33-39.

18. Barthel,  W. F.,   Hawthorne, J. C.,  Ford, J.  H.,
    Bolton, G.  C., McDowell, L. L., Grissinger, E. H.,
    and  Parsons, D.  A.  1969.  Pesticide residues in
    sediments of the  lower  Mississippi River and its
    tributaries. Pestic. Monit. Jour. 3(1): 8-66.

19. Baur, J. R., Bovey, R. W., and Merkle, M. G. 1972.
    Concentration  of picloram  in runoff water. Weed
    Sci. 20(4): 309-313.

20. Behrens, R. 1975. Corn  and weeds. Weeds Today
    6(1):  15-18.

21. Biggar,  J. W. 1970. Pesticide  movement  in  soil
    water. In Pesticides  in the soil: ecology, degrada-
    tion, and movement. Internal!. Symp. on Pesticides
    in the Soil: 107-119. Michigan State Univ., East
    Lansing.

22. Bingham, C. R. 1969. Some effects of pesticides on
    Mississippi waters. Proc. Miss. Water Resour. Conf.,
    MSU Water Resour. Res. Inst., State College, Miss.

23. Blakely, B.  D., Coyle, J. J., and Steele, J. G. 1957.
    Erosion on  cultivated land. In Soil, the yearbook of
    agriculture, 1957:  290-307. U.S. Dept. Agr., Wash-
    ington, D.C.
112

-------
24. Boast,  C. W.  1973. Modeling  the movement  of
    chemicals in  soils  by  water.  Soil  Sci.  115(3):
    224-230.

25. Bode, L. E., and Gebhardt, M. R. 1969. Equipment
    for incorporation of herbicides. Weed  Sci. 17(4):
    551-555.

26. Bovey, R. W., Burnett, E., Richardson, C., Merkle,
    M.  G., Baur, J. R.,  and  Knisel, W.  G.  1974.
    Occurrence  of 2,4,5-T  and  picloram  in surface
    runoff water  in  the  Blacklands  of Texas.  Jour.
    Environ. Qual. 3(1): 61-64.

27. Bradley, J. R., Sheets, T. J., and Jackson, M.  D.
    1972.  DDT and  toxaphene movement in surface
    water from cotton plots. Jour. Environ. Qual. 1(1):
    102-105.

28. Burkhead, C.  E., Max, R. C.,  Karnes, R. B., and
    Reid, E. 1972. Usual planting and harvesting dates,
    field and seed crops. Agr. Handbook No. 283, 84 p.,
    Statis. Rep. Serv., U.S. Dept. Agr., Washington, D.C.
29. Burnside, 0. C., and Colville, W. 1964. Soybean and
    weed yields  as affected by  irrigation, row spacing,
    tillage, and amiben. Weeds 12(2): 109-112.

30. Burnside, 0. C., Wicks, G.  A., and Fenster, C. R.
    1971.  Dissipation  of  dicamba,  picloram,  and
    2,3,6-TBA  across  Nebraska.  Weed  Sci.  19(4):
    323-325.

31. Butler, P. A. 1965. Effects of herbicides on estu-
    arine fauna. Proc. So. Weed Conf. 18: 576-580.

32. Cairns, J., Jr., Heath, A. G., and Parker, B. C. 1975.
    Temperature  influence  on chemical  toxicity to
    aquatic  organisms. Jour. Water Poll. Contr.  Fed.
    47(2): 267-280.

33. Caro, J. H. 1969. Accumulation by plants of organo-
    chlorine insecticides from the soil. Phytopathology
    59(9): 1191-1197.

34. Caro, J.  H., Edwards, W. M., Glass, B. L., and Frere,
    M.  H.  1972.  Dieldrin  in  runoff from treated
    watersheds. Proc. 1970 Symp. on  Interdiscip. As-
    pects  of  Watershed  Mangt.,  Bozeman, Mont.:
    141-160. Amer. Soc. Civil Engin., New York.
 35.  Caro,  J.  H., Freeman,  H.  P., Glotfelty,  D. E.,
     Turner, B. C., and Edwards, W. M. 1973. Dissipation
     of soil-incorporated carbofuran in the field. Jour.
     Agr. Food Chem. 21(6): 1010-1015.

 36.  Caro, J. H., Freeman, H. P., and Turner, B. C. 1974.
     Persistence in soil and losses in  runoff of soil-incor-
     porated carbaryl in a small watershed. Jour. Agr.
     Food Chem. 22(5):  860-863.

 37.  Casey, J. E., Lacewell, R. D., and Sterling, W. 1975.
     An example of economically feasible opportunities
     for reducing pesticide use in commercial agriculture.
     Jour. Environ. Qual. 4(1): 60-64.

 38.  Chant, D. A.  1966. Integrated  control systems. In
     Scientific  aspects of pest control:  193-218. Natl.
    Acad. Sci.-Natl. Res. Council, Washington, D.C.

 39. Chesters, G., and Konrad, J. G. 1971. Effects of
     pesticide usage on water quality. BioScience 21(12):
     565-569.

40. Chian, E.  S. K., Bruce, W. N.,  and Fang, H. H. P.
     1975.  Removal  of  pesticides by reverse osmosis.
    Environ. Sci. Technol. 9(1): 52-59.

41. Chiang, H.  C. 1973. Ecological considerations in
    developing recommendations  for chemical control
    of pests: European  corn borer  as a model. FAO
    Plant Protec. Bui. 21(2): 30-39.

42. Clack, J. E.  1972. Application of herbicides in foam
    to reduce drift.  Proc. South. Weed Sci. Soc. 25:
    495^97.

43. Cole, H., Barry, D.,  and Frear, D. E. H. 1967. DDT
    levels in fish, streams, stream sediments, and soil
    before and  after DDT aerial spray  application for
    fall cankerworm in northern  Pennsylvania. Bui.
    Environ. Contam. Toxicol. 2(3):  127-155.

44. Constien,  E.  J., Anderson,  L., Murphy,  W.  J.,
    Woodruff,  C. M.,  Palm,  E.,  Thomas, G.,  and
    Workman, H. 1974.  Conservation  tillage-planting
    systems. Exten. Div., Univ. Missouri, Columbia, 4 p.


 45.  Cooperative  Extension Service, The  Ohio State
     University.  1973. Pesticides and the  environment.
     Bui. 536,20 p.
                                                                                                      113

-------
46. Cooperative  Extension  Service,  The  Ohio  State
    University. 1974. Insect control in no-tillage corn.
    Bui. L-208,4 p.

47. Cooperative Extension Service, Univ.  Illinois. 1974.
    Guide  for  the custom application of pesticides in
    Illinois. In  Summary of 26th Illinois custom spray
    operators  training school: 234-252.  Univ. Illinois,
    Urbana-Champaign.

48. Cope,  O.  B. 1965. Some  responses of fresh water
    fish to herbicides. Proc.  South. Weed Conf. 18:
    439-445.

49. Council on Environmental  Quality. 1972. Integrated
    pest management. Washington, D.C., 41 p.

50. Cox, J. L. 1971.  DDT residues in  seawater and
    particulate matter in the California current system.
    Fish. Bui., Natl. Mar. Fish.  Serv. 69(2): 443-450.

51. Crafts,  A.  S., and Robbins, W. W.  1962. Weed
    control. 3rd ed., McGraw-Hill, New York, 660 p.

52. Craig, W. S., Brown, K. E., Huggans, J. L., Thomas,
    G.  W., and Jones, F. G.  1974.  Insect control
    recommendations, Missouri,  1974. Coop.  Exten.
    Serv., Univ. Missouri, Columbia, 155 p.

53. Crawford,  N. H.,  and Donigian, A.  S., Jr.  1973.
    Pesticide transport and runoff model for agricultural
    lands.  EPA-660/2-74-013,  211 p. U.S. Govt. Print-
    ing Off., Washington, D.C.

54. Crosby, D. G. 1970. The nonbiological degradation
    of  pesticides  in  soils. In Pesticides  in  the soil:
    ecology, degradation,  and  movement.  Internatl.
    Symp.  on  Pesticides in the Soil: 86-94. Michigan
    State Univ., East Lansing.

55. Daniels, N.  1971.  Detection of insecticidal residue
    and control of soil  insects. Jour. Econ. Entomol.
    64(1):  175-177.

56. Davis,  E.  A., and Ingebo,  P.  A. 1970. Fenuron
    contamination of stream  water from  a chapparal
    watershed in Arizona. Res. Prog. Rep., West. Soc.
    Weed Sci.: 22.

57. Depew, L.  1971.  Evaluation  of foliar  and soil
    treatments for greenbug control on sorghum. Jour.
    Econ. Entomol. 64(1): 169-172.
58. Edwards, C. A.,  Thompson, A. R., Beynon, K. I.,
    and  Edwards, M. J. 1970. Movement  of dieldrin
    through soils. I. From arable soils into ponds. Pestic.
    Sci. 1:  169-173.

59. Edwards, W.  M., and Glass, B.  L. 1971. Methoxy-
    chlor and  2,4,5-T  in  lysimeter  percolation  and
    runoff water. Bui. Environ. Contam. Toxicol. 6(1):
    81-84.

60. Eichelberger, J.  W., and Lichtenberg, J. J.  1971.
    Persistence  of pesticides in river water. Environ. Sci.
    Technol. 5(6): 541-544.

61. Ennis, W.  B., Jr., Jansen,  L.  L., Ellis, I. T., and
    Newson, L. D. 1967. In The world food problem. A
    report of  the President's Science Advisory Com-
    mittee, Vol. Ill: 130-175. The White House.


62. Environmental  Protection  Agency.  1973.  Water
    quality criteria, 1972. EPA-R3-73-033,  594 p. U.S.
    Govt. Printing Off., Washington, D. C.

63. Epstein, E.,  and Grant, W. J. 1968.  Chlorinated
    insecticides  in runoff  water as affected by crop
    rotation. Soil Sci. Soc.  Amer. Proc. 32(3): 423426.

64. Erbach, D.  C.,  and Lovely,  W. G.  1974. Weed
    control with conservation tillage production of corn
    and soybeans. Jour. Soil and Water Conserv. 29(1):
    44-45.

65. Erbach, D. C., Lovely, W. G., and Bockhop, C. W.
    1975.  Area  of influence  of  herbicide granules.
    Submitted to Weed Sci.

66. Evans, J.  0., and Duseja,  D.  R. 1973. Herbicide
    contamination of surface runoff waters. EPA-R2-
    73-266, 99 p. U.S. Govt. Printing Off., Washington,
    D.C.

67. Eye, J. D. 1968. Aqueous transport  of dieldrin
    residues in  soils.  Jour. Water Poll. Contr. Fed. 40(8)
    part2:R316-R332.

68. Fahey, J.  E. 1970. Contamination of  farm ponds
    following treatment of watersheds for pest control.
    In Effect  of pesticide  residues and other organo-
    toxicants on the quality of  surface  and  ground
    water resources:  9-13a. Purdue Univ. Water Resour.
    Res. Center, Tech. Rep. No. 10.
114

-------
69. Fleming, W. E. 1957. Soil management and insect
    control. In Soil, the yearbook of agriculture, 1957:
    326-333. U.S. Dept. Agr., Washington, D.C.

70. Foy, C. L., and Bingham, S. W. 1969. Some research
    approaches toward minimizing herbicidal residues in
    the environment. Residue Rev. 29: 105-135.

71. Foy, C. L., Coats, G. E., and Jones, D. W. 1971.
    Translocation of growth regulators  and herbicides;
    vascular plants. In P.  L. Altman  and D. S. Dittmer,
    eds.  Respiration  and circulation:  743-791.  Biol.
    Handbooks, Fed. Amer. Soc. Expt.  Biol., Bethesda,
    Md.

72. Frank, P. A. 1972. Herbicidal residues in  aquatic
    environments. In Fate of organic pesticides in the
    aquatic  environment:  135-148.  Advan.  in Chem.
    Ser. Ill, Amer. Chem. Soc., Washington, D.C.

73. Giese, R. L., Peart, R. M., and Huber, R. T. 1975.
    Pest  management. Science 187(4181): 1045-1052.

74. Glotfelty, D. E., and Caro, J. H. 1974. Introduction,
    transport, and fate of persistent pesticides in the
    atmosphere. In V.  R. Deitz, ed. Removal of trace
    contaminants from the air: 42-62. ACS Symp. Ser.
    17, Amer. Chem. Soc., Washington, D.C.

75. Graetz, D. A., Chesters, G., Daniel,  T. C., Newland,
    L. W., and Lee, G. B. 1970.  Parathion degradation
    in lake sediments. Jour. Water Poll.  Contr. Fed. 42:
    R76-R94.

76. Guenzi,  W. D., and  Beard, W.  E.  1974. Volatili-
    zation of pesticides. In W. D. Guenzi, ed. Pesticides
    in soil and  water: 107-122. Soil Sci. Soc. Amer.,
    Inc., Madison, Wis.

77. Haan, C. T.  1971. Movement of pesticides by runoff
    and erosion. Trans. ASAE 14(3):  445-447,449.

78. Haas, R. H., Scifres, C. J., Merkle, M. G., Hahn, R.
    R.,  and Hoffman, G. O.  1971. Occurrence  and
    persistence of picloram in grassland water sources.
    Weed Res. 11(1): 54-62.

79. Hall, D. C., and Norgaard,  R.  B.  1973. On the
    timing and  application of pesticides. Amer. Jour.
    Agr. Econ. 55(2): 198-201.
80. Hall,  J.  K.  1974. Erosional  losses  of s-triazine
    herbicides. Jour. Environ. Qual. 3(2): 174-180.

81. Hall, J. K., Pawlus,  M., and Higgins, E. R. 1972.
    Losses of atrazine in runoff water and soil sediment.
    Jour. Environ. Qual. 1(2): 172-176.

82. Hamaker, J. W., and  Thompson,  J.  M.  1972.
    Adsorption. In C. A. I. Goring and J. W. Hamaker,
    eds. Organic chemicals in the soil environment, Vol.
    I: 49-143. Marcel Dekker, Inc., New York.

83. Harmon,  M. R., Greichus,  Y. A., Applegate, R. L.,
    and Fox, A. C.  1970.  Ecological distribution  of
    pesticides in  Lake Poinsett, South Dakota. Trans.
    Amer. Fish. Soc. 99(3): 496-500.

84. Hanson, A. A. 1972. A directed ecosystem approach
    to pest control and environmental quality. Jour.
    Environ. Qual. 1(1): 45-54.

85. Harris, C.  I. 1966.  Adsorption, movement, and
    phytotoxicity of monuron and s-triazine herbicides
    in soU. Weeds 14(1): 6-10.

86. Harrison, F., and Press,  J.  1971. Timing of insecti-
    cide applications for European corn borer control in
    sweet  corn.   Jour.   Econ.   Entomol.   64(6):
    1496-1499.

87. Hartung,  R., and  Klingler, G.  W. 1970. Concen-
    tration of DDT by sedimented polluting oils. Envi-
    ron. Sci. Technol. 4(5): 407-410.

88. Harvey, G. R. 1974. DDT and PCB in  the Atlantic.
    Oceanus 18(1): 19-23.

89. Helling, C.  S. 1970. Movement of s-triazine herbi-
    cides in soils. Residue Rev. 32: 175-210.

90. Hill,  D. W.  1967. Degradation of selected chlori-
    nated  hydrocarbon  pesticides  in anaerobic and
    aerobic  water  environments.  Diss.  Abstr. B27:
    2399-2400.

91. Hill,  D. W., and  McCarty, P.  L.  1967.  Anaerobic
    degradation  of selected chlorinated hydrocarbon
    pesticides.  Jour.  Water Poll.  Contr.  Fed.  39(8):
    1259-1277.
                                                                                                     115

-------
92. Hiltbold, A.  E.  1974. Persistence of pesticides in
    soil. In W.  D. Guenzi, ed. Pesticides in  soil and
    water:  203-222. Soil Sci. Soc. Amer., Inc., Madison,
    Wis.

93. Hoffman, C.  H. 1970. Alternatives to  conventional
    insecticides—control  of insect  pests.  Agr. Chem.
    25(9):  14-19A.

94. Huang, J. C. 1972. Pesticides in water-effects on
    human  health.   Jour.  Environ.  Health  34(5):
    501-508.


95. Johnson,  H.  E.,  and Ball, R. C.  1972. Organic
    pesticide pollution  in an  aquatic environment. In
    Fate of organic pesticides in the aquatic environ-
    ment:  1-10.  Advan.  in  Chem.  Ser.  Ill,  Amer.
    Chem. Soc., Washington, D.C.


96. Johnstone, D. R.  1973. Insecticide concentration
    for ultra-low volume crop  spray application. Pestic.
    Sci. 4(1): 77-82.

97. Kaufman, D. D.  1972. Degradation  of pesticide
    combinations. In A. S. Tahori, ed.  Fate of pesticides
    in  environment.  Pesticide  Chemistry Vol.  VI:
    175-204. Gordon and Breach, Inc., New York.

98. Kaufman, D. D. 1974. Degradation of pesticides by
    soil microorganisms. In W. D. Guenzi, ed. Pesticides
    in soil and water:  133-202. Soil  Sci. Soc. Amer.,
    Inc., Madison, Wis.

99. Kearney, P.  C., Nash, R. G., and  Isensee, A. R.
    1969. Persistence of pesticide residues in soils. In M.
    W.  Miller  and G. G.  Berg,  eds. Chemical fallout.
    Current research  on  persistent pesticides: 54-67.
    Charles C. Thomas, Pub., Springfield, m.

100. King, P. H., Yeh, H. H., Warren, P. S., and Randall,
     C. W. 1969. Distribution of pesticides in surface
     waters. Jour. Amer. Water Works Assoc. 61(9):
     483486.

101. Knipling, E. 1972. Use of organisms  to control
     insect pests. Jour. Environ. Qual.  1(1): 34-40.
102. Krause, R. A., and Massie, L. B. 1973. Application
     and implementation of computerized forecasts of
     potato late blight. Phytopathology 63(2): 203.
 103. Kuhlman, D. E., Cooley, T. A., and Walt, J. 1974.
     Adult  com  rootworm  populations  in  Illinois,
     August 1973. In Summary of 26th Illinois custom
     spray  operators training school:  159-164. Coop.
     Exten. Serv., Univ. Illinois, Urbana-Champaign.

 104. Lauer, G.  J., Nicholson, H.  P., Cox, W. S., and
     Teasley, J. I.  1966. Pesticide contamination of
     surface waters  by sugarcane fanning in Louisiana.
     Trans. Amer. Fish. Soc. 95(3): 310-316.

 105. Leland, H. V., Bruce, W. N.,  and Shimp, N. F.
      1973.  Chlorinated  hydrocarbon  insecticides in
     sediments  of southern  Lake Michigan.  Environ.
     Sci. Technol. 7(9): 833-838.

 106. Leshniowsky, W. O., Dugan,  P. R., Pfister, R. M.,
     Frea,  J. I., and  Randies, C.  I.  1970.  Aldrin:
     removal from  lake water by flocculent bacteria.
     Science 169: 993-995.

 107. Letey, J.,  and  Oddson, J. K. 1972. Mass transfer.
     In C.A.I. Goring and J. W. Hamaker, eds. Organic
     chemicals in the soil environment, Vol. I: 399-440.
     Marcel Dekker, Inc., New York.

 108. Lichtenberg, J. J.,  Eichelberger, J.  W., Dressman,
     R. C., and Longbottom, J. E.  1970. Pesticides in
     surface  waters of the  United States—a  5-year
     summary,  1964-1968. Pestic. Monit. Jour.  4(2):
     71-86.

 109. Locke, D.  O., and Havey, K. 1972. Effects of DDT
     upon  salmon from Schoodic Lake, Maine. Trans.
     Amer. Fish. Soc. 101: 638.

.110. Lutz, J. F.,  Byers, G. E., and Sheets, T. J. 1973.
     The persistence and  movement of picloram and
     2,4,5-T  in. soils.  Jour.  Environ.  Qual.  2(4):
     485-488.

 111. Marston, R. B., Schults, D. W., Thiroyama, T., and
     Snyder, L. V. 1968. Amitrole  concentrations in
     creek waters downstream from an aerially sprayed
     watershed  sub-basin. Pestic. Monit.  Jour.  2(3):
      123-128.

 112. McDowell, L. L., Grissinger, H. H., Bolton, G. C.,
     and Parsons, D. A.  1971. Chlorinated hydrocarbon
     insecticide contamination of streambed sediments
     in the Mississippi River Delta. Proc. Miss. Water
     Resour. Conf., Water   Res.  Inst., MSU,  State
     College, Miss.
116

-------
113.  McNew, G.  L.  1972. Interrelationships  between
     agricultural chemicals and environmental quality in
     perspective. Jour. Environ. Qual. 1(1): 18-22.

114.  Metcalf, R.  L.  1972. Agricultural  chemicals in
     relation to environmental quality: insecticides to-
     day and  tomorrow.  Jour. Environ.  Qual.  1(1):
     10-14.
124. Patti, J.  H., and  Garner, G.  R.  1974. Bacillus
     thuringiensis investigations for control of Heliothis
     spp.  on  cotton.  Jour. Econ.  Entomol. 67(3):
     415-418.

125. Phillips, W. M., and Feltner, K. C.  1972. Persist-
     ence and movement of picloram in two Kansas
     soils. Weed Sci. 20(1): 110-116.
115. Miles, J. W., and Woehst, J. E. 1969. Formulations
     for  controlled  release  of  Abate  in water. In
     Pesticidal formulations research: 183-191. Advan.
     Chem. Ser. 86, Amer. Chem. Soc., Washington,
     D.C.

116. Moomaw, R., and Robison, L. 1973. Broadcast or
     banded atrazine plus propachlor with tillage vari-
     ables in corn. Weed Sci. 21(2): 106-109.

117. Moubry, R. J., Helm, J. M., and  Myrdal, G. R.
     1968. Chlorinated pesticide residues in an aquatic
     environment  located  adjacent to  a  commercial
     orchard. Pestic. Monit. Jour. 1(4): 27-29.

118. Muirhead-Thomson,  R. C.  1971. Pesticides and
     fresh water fauna. Academic Press, New York, 245
     P-

119. Mullison, W.  R. 1970. Effects of herbicides on
     water  and  its  inhabitants.  Weed  Sci.  18(6):
     738-750.

120. Nash, R. G.  1974.  Plant uptake of  insecticides,
     fungicides,  and fumigants from  soils. In  W. D.
     Guenzi,  ed. Pesticides in soil and water: 257-313.
     Soil Sci. Soc. Amer., Inc., Madison, Wis.

121. Nicholson,  H.  P.,  and  Thoman,  J. R.  1965.
     Pesticide persistence in public water,  their detec-
     tion and removal.//!  L. O. Chichester, ed. Research
     on  pesticides:  181-189. Academic  Press,  New
     York.

122. Panel on Pesticide Monitoring, Working Group on
     Pesticides.   1971.  National  pesticide monitoring
     program. Pestic. Monit. Jour. 5(1): 35-71.

123. Paris, D. P., and Lewis, D. L.  1973. Chemical and
     microbial degradation of ten selected pesticides in
     aquatic systems. Residue Rev. 45: 95-124.
126. Pionke, H. B.,  and Chesters, G. 1973. Pesticide-
     sediment-water  interactions. Jour. Environ. Qual.
     2(1): 29-45.

127. President's Science Advisory  Committee.  1965.
     Restoring  the quality  of our environment. The
     White House.

128. Putnam, A.  R., and Duke, W. B. 1974. Biological
     suppression of weeds: evidence for allelopathy in
     accessions of cucumber. Science 185: 370-372.

129. Quraishi,  M. S. 1971.  Nontoxic approaches  to
     insect control, N. Dak. Farm Res. 28(4): 33-35.

130. Raghu, K., and  MacRae, I. C. 1967.  The effect of
     the gamma-isomer of BHC upon the microflora of
     submerged rice soil. Can. J. Microbiol. 13:  173 and
     625.

131. Randall, C. W.  1967. Pollutional aspects of pesti-
     cides in  natural streams. Amer. Water  Resour.
     Assoc. Bui. 3(3): 47-53.

132. Reid, J. T.,  and Peacock, H. A. 1973. Subsurface
     sweep for applying herbicides. Agron. Jour. 65(6):
     996-998.

133. Rhode  Island  Water  Resources Center.  1972.
     Eighth annual report: 26-32.

134. Ritter, W. F., Johnson, H. P., Lovely, W. G., and
     Molnau,  M. 1974.  Atrazine,  propachlor, and
     diazinon residues in small agricultural watersheds.
     Runoff losses,  persistence, and movement. Envi-
     ron. Sci. Technol. 8(1): 38-42.

135. Robeck, G.  G., Dostal, K.  A., Cohen, J. M., and
     Kreissl, J. F. 1965. Effectiveness of water treat-
     ment processes  in pesticide removal. Jour. Amer.
     Water Works Assoc. 57(2): 181-199.
                                                                                                      117

-------
136. Rowe, D. R., Canter,  L. W., Snyder, P. J., and
     Mason, J. W. 1971. Dieldrin and endrin concentra-
     tions  in a Louisiana  estuary.  Pestic. Monit. Jour.
     4(4):  177-183.

137. Schafer, M.  L., Peeler, J. T., Gardner, W. S., and
     Campbell, J. E. 1969. Pesticides in drinking water.
     Waters from the Mississippi  and Missouri rivers.
     Environ. Sci. Technol. 3(12): 1261-1269.

138. Schulze, J. A., Manigold, D. B., and Andrews, F. L.
     1973.  Pesticides  in selected western  streams—
     1968-71. Pestic. Monit. Jour. 7(1): 73-84.

139. Scifres, C. J., Allen, T. J., Leinweber, C. L., and
     Pearson,  K.  H. 1973. Dissipation and  phytotoxi-
     city of dicamba residues in water. Jour. Environ.
     Qual.  2(2): 306-309.

140. Seba,  D. B., and  Corcoran, E. F. 1969. Surface
     slicks  as  concentrators of pesticides in  the marine
     environment. Pestic. Monit. Jour. 3(3): 190-193.

141. Sharpies, J. A., and Walker,  R.  1974. Shifts in
     cropland use in  the North Central  region. Agr.
     Econ. Res. 26(4): 106-110.

142. Sheets, T. J., Bradley, J. R.,  Jr., and Jackson, M.
     D. 1972. Contamination of  surface and ground
     water with pesticides applied to cotton. Univ. No.
     Car., Water Resour. Res.  Inst., Rep. No. 60, 63 p.

143. Shore, J. 1974. Accurate residue prediction. Pest
     Control 42(12): 18-21.

144. Spencer,  W.  F.  1970. Distribution of pesticides
     between  soil, water,  and air. In  Pesticides in the
     soil:  ecology, (degradation, and movement.  Inter-
     natl.  Symp. on  Pesticides  in  the Soil: 120-128.
     Michigan State Univ., East Lansing.

145. Spencer W.  F., Farmer, W. J., and Cliath, M. M.
     1973. Pesticide  volatilization. Residue Rev. 49:
     147.

146. Sprague, G.  F., and Dahms, R. G. 1972. Develop-
     ment  of crop resistance to insects. Jour. Environ.
     Qual.  1(1): 28-34.

147. Stockdale, H. J., DeWitt, J.,  and Ryan, S.  1974.
     Summary of Iowa insect pest control recommenda-
     tions  for 1974. Coop. Exten. Serv., Iowa St. Univ.,
     Ames. IC-328(Rev.), 21p.
148. Summers, C. G., Byrne, H. D., and Pimentel, D.
     1971. Spring timing applications for control of the
     alfalfa weevil in New York. Jour. Ecoh. Entomol.
     64(2): 478479.

149. Texas Agricultural Extension Service. 1974. Sug-
     gestions  for weed control with chemicals. Part
     I-MP 1059,  32p. Texas A&M Univ., College Sta-
     tion.

150. Texas  A&M University, College of Agriculture.
     1970. Special report:  impact of drastic reduction
     in the  use of agricultural  chemicals on food and
     fiber production and cost to consumer. 62p.

151. Trichell,  D. W., Morton, H. L., and Merkle,  M. G.
     1968. Loss of herbicides in runoff water. Weed Sci.
     16(4): 447-449.

152. Turnipseed,  S. G., Todd, J. W., Greene, G. L.,and
     Bass, M. H. 1974. Minimum rates of insecticides on
     soybeans: Mexican bean beetle, green cloverworm,
     corn earworm and velvetbean caterpillar.  Jour.
     Econ. Entomol. 67(2): 287-291.

153. Upchurch, R. 1966. Behavior of herbicides in soil.
     Residue Rev. 16: 47-85.

154. U.S. Bureau of  the  Census.  1973. Census of
     agriculture,  1969. U.S. Govt. Printing Off.,  Wash-
     ington, D.C.

155. Wade, N. 1975. International agricultural research.
     Science 188  (4188): 585-589.

156. Waldron,  A. C. 1974. Pesticide movement from
     cropland into Lake Erie. EPA-660/2-74-032, 95p.
     U.S. Govt. Printing Off., Washington, D.C.

157. Ware, G. W., Estesen,  B. J.,  Cahill, W. P., and
     Frost, K. R. 1972. Pesticide drift VI: target and
     drift deposits vs. time of applications. Jour. Econ.
     Entomol. 65(4): 1170-1171.

158. Ware, G. W., and Roan, C. C. 1970.  Introduction
     of  pesticides with  aquatic  microorganisms and
     plankton. Residue Rev. 33: 1545.

159. Wax, L. M. 1973. Weed control. In B. E. Caldwell,
     ed.  Soybeans: improvement, production, and uses:
     417457. Amer. Soc. Agron., Madison, Wis.
118

-------
160. Wax,  L.  M.,  Stoller, E. W.,  Slife,  F.  W., and
     Andersen, R. N. 1972. Yellow nutsedge control in
     soybeans. Weed Sci. 20(2): 194-201.

161. Way, W. A. 1968. Soil tillage for corn: maxi-mini-
     zero.  Farm  Facts No.  40, Exten. Serv.,  Univ.
     Vermont. 2p.

162. Weatherholtz, W. M., Cornwell, G. W., Young, R.
     W.,  and Webb, R. E.  1967. Distribution of hep-
     tachlor residues  in  pond  ecosystems in south-
     western Virginia. Jour. Agr. Food Chem. 15(4):
     667-670.

163. Weber, J. B., Weed, S. B., and Sheets, T. J. 1972.
     Pesticides-how  they  move and react  in  the soil.
     Crops and Soils 25(1): 14-17.

164. Wellhausen, H. W. 1972. Sorghum production  for
     grains  or  silage.  Coop.   Exten. Serv.,   Univ.
     Arkansas, Leaflet 427 (Rev.), 8p.

165. Whitaker, F. D., Heinemann, H. G., and Wisch-
     meier, W. H. 1973. Chemical weed controls affect
     runoff, erosion, and corn yields.  Jour. Soil and
     Water Conserv. 28(4): 174-176.
166. White, A. W., Barnett, A. P., Wright, B. G., and
     Holladay, J. H. 1967. Atrazine losses from fallow
     land caused  by runoff and erosion. Environ. Sci.
     Technol. 1(9): 740-744.

167. Wiese, A. F.  1975. Movement and persistence of
     fiuometuron in the South: II. Lateral movement in
     runoff water. Proc. South. Weed Sci. Soc., in press.

168. Williams,  P.  M., and Robertson, K. J.  1975.
     Chlorinated hydrocarbons in sea-surface films and
     subsurface waters at nearshore stations and in the
     North  Central  Pacific  Gyre. Fish.  Bui.,  Natl.
     Ocean. Atmos. Admin. 73(2): 445-446.

169. Willis, G. H., and Hamilton, R. A. 1973. Agricul-
     tural chemicals in surface runoff, ground water,
     and  soil.  I.  Endrin.  Jour.  Environ.  Qual.  2(4):
     463-466.


170. Yeo,  R.  R.  1967.  Dissipation  of  diquat  and
     paraquat and effects on  aquatic weeds  and fish.
     Weed Sci. 15(1): 4246.
                                                                                                     119

-------
                                               CHAPTER 6

                              INTERDISCIPLINARY RESEARCH NEEDS

                                                B. A. Stewart
   The preceding  chapters have listed research needs
relative  to particular disciplines.  This chapter briefly
discusses their interrelationships as they relate to non-
point pollution from cropland. Nonpoint pollution can
best  be  controlled by controlling  erosion. However,
knowledge is critically lacking concerning the degree of
erosion  and  sediment  control necessary  to control
nutrient and pesticide losses.
   The  Universal  Soil  Loss  Equation is  being  used
extensively in predicting  nonpoint pollution.  It is an
excellent  tool because  it quantifies  soil loss  and the
effect of  various practices on  reducing this loss.  The
Universal  Soil Loss Equation, however, estimates the
average annual soil loss from a  field area and does not
indicate the amount of sediment that is delivered to a
stream. The sediment  delivery ratio  is an attempt to
relate field losses to amounts reaching the stream. The
difficulty  with this concept for predicting water pollu-
tion is that it takes into account both the deposition of
sediment as  it moves toward the stream and the gains
from channel erosion. Pollutants, particularly pesticides
and nutrients from  added fertilizers, are  usually  not
associated with sediment from channel erosion.
   The enrichment ratio is a measure of the increase in
the concentration of a pollutant associated  with the
sediment that actually reaches a stream compared to the
concentration in the watershed  soil. The concentration
usually increases because more nutrients and pesticides
are adsorbed on fine-textured particles than on coarse
particles, and more coarse particles are deposited as the
sediment  moves  from  the  field  area  to  the stream.
Consequently, the  most pressing research need is to gain
a better understanding of how the Universal Soil Loss
Equation, the sediment delivery ratio, and the enrich-
ment  ratio can  be meshed.  These measurements  and
evaluations will not be  made easily, cheaply, or quickly.
Data obtained from small plots or field area can provide
only  crude  applications because the size  effect is so
significant. Since cropland is  so diffuse, monitoring of
agricultural areas  is  impractical. The most  likely ap-
proach, therefore, is to develop predictive  models. A
major effort should be directed toward obtaining neces-
sary data and developing such models. Results from such
an effort cannot be expected to be precise  but should
represent a statistical approach that establishes relation-
ships and relative levels.
   The Universal Soil Loss  Equation predicts average
annual soil  losses. Additional accuracy  is  needed  to
predict single events. These can be important in their
effect on water quality, especially if the time  of loss is
associated with applications of agricultural  chemicals.
Again, information is needed on large watershed areas. It
is fairly easy to instrument a field-sized area and measure
losses of sediment and  associated pollutants. The diffi-
culty is in determining how much of these actually reach
the stream. The effects and proper design of filter strips,
settling  basins, and sediment traps should   also  be
determined. Some evidence indicates that these can very
effectively reduce losses; however, there are also indica-
tions that this deposited material may  be  moved during
extreme runoff events.
   No-till  and  conservation  tillage systems  are highly
effective  for controlling soil  loss. Additional research is
needed, however, because it is not known how widely
these practices can be used. Insect and disease hazards
are  greater,  and if these  practices are  used   on vast
acreages, outbreaks could occur. Also, in some areas, it
may be necessary to occasionally plow to loosen the soil.
More data are also required concerning  pesticide and
nutrient  losses from no-till and  conservation tillage
systems as compared to conventional systems. Sediment
losses  are  drastically  reduced,  but  the  meager  data
available do not show proportionate losses in pollutants.
   Another  primary  research  need is to evaluate the
economic impact of control measures for  controlling
sediment and  chemical losses. Available technology is
adequate to  control these pollutants in most instances.
Implementation of these measures could produce major
changes  in  cultivation  practices,  timing of  chemical
applications, and location  of production  and general
                                                                                                         121

-------
farming  practices.  These  changes  would have direct
effects on the profitability of agricultural production,
supplies of food and fiber, and rural land use. An urgent
need is to appraise alternative procedures for controlling
pollutants so that economic hardships can be minimized.
   Since  our landscapes are not uniform and all areas are
not equidistant from a water body, numerous combina-
tions  of practices can  achieve the same water-quality
goal. The most desirable approach is to give each farmer
the opportunity to select the appropriate combination
of practices. To do this, he needs to  know how much
control each practice will provide. The costs and benefits
of implementing the various practices should also be
known. This  should include regional  and national im-
pacts. Thus, much research  is needed in a variety of
disciplines  and the  first  step  in  any of  them  is to
quantify the relationships.
   Last, and  perhaps most important, criteria must be
established as to what levels of sediment and chemicals
constitute  pollution. The criteria  might be absolute
limits (not to exceed X ppm) or conditional limits (not
to exceed  Y ppm in Z  years). Although the establish-
ment of these criteria will require  research outside of
traditional  agricultural areas—limnology, aquatic botany
and  zoology, and water treatment—it is of great impor-
tance  to  the agricultural community. Without  these
criteria it is impossible to  determine which practice will
be  adequate, much less  decide what  the  economic
impacts will be.
 122

-------
                                           APPENDIX A

                    SIMULATION OF DAILY POTENTIAL DIRECT RUNOFF

                                          INTRODUCTION
   The amount and seasonal distribution of direct runoff
was estimated to assess potential transport of pesticides
and nutrients. The effects  of some land management
practices on direct runoff were also estimated. Hydrol-
ogists have developed several rainfall-runoff models of
various degrees  of complexity for making  these esti-
mates. The  more  physically realistic models are quite
complicated and require a great deal of input informa-
                                                    tion  and computer  time. The national scope of this
                                                    report and the severe time constraints involved dictated
                                                    the use of a rather simple method of estimating runoff
                                                    from rainfall. Any input information required must also
                                                    be readily  available. After  considering several possi-
                                                    bilities, we decided to use the Soil Conservation Service
                                                    procedure  for estimating direct runoff from  storm
                                                    rainfall (4).
THE SOIL CONSERVATION SERVICE PROCEDURE FOR ESTIMATING DIRECT RUNOFF
                                 FROM STORM RAINFALL
  The Soil Conservation Service procedure for estimat-
ing direct runoff from storm rainfall (sometimes called
the SCS curve number method) was designed to use the
most generally available rainfall data: total daily rainfall.
For this reason rainfall intensity is largely ignored. The
basic relationship is the equation:
    Q =
            (P-Ia)2
                      Pj>L
(1)
where
    Q
    P
    L
          runoff in inches
          rainfall in inches
          initial abstraction in inches
     u
    S  =  potential maximum retention plus initial
          abstraction.
  The  initial  abstraction before runoff begins is con-
sidered to consist mainly of interception, infiltration and
surface  storage. Utilizing limited data from small experi-
mental  watersheds, the following empirical relationship
was developed:

   !a  =  (0.2)S.                              (2)

Substituting this relationship into equation (1) gives
   0=   (P-C.2S)2
   V     P+0.8S
                    ,P>(0.2)S
(3)
                                                                                                  SCS
       which is  the rainfall-runoff  relation used in the
       method.
         The parameter CN (runoff curve number of hydrol-
       ogic soil-cover complex number) is defined in terms of
       the parameter S as:
                                                      CN =
                                                            1000
                                                            S+10
                                                     (4)
Note that runoff equals rainfall when S = 0 and CN =
100.
   The potential maximum retention, S , and therefore
the runoff curve number are related to sofl surface and
profile properties,  the vegetative  cover, management
practices, and the soil water content on  the day of the
storm. Solutions of equation (3) are shown as a family
of curves in Fig. 1.
   Soil water  content  on the day of the storm  is
accounted for by  an  Antecedent  Moisture Condition
(AMC) determined by the total rainfall in the 5-day
period preceding the storm.
   Three  AMC groups have been established with the
boundaries between groups dependent upon the time of
year as shown in Table 1.
   The seasonal difference in the AMC  groupings is  an
attempt to account for the greater evapotranspiration
between storms during the growing season.
   The different infiltration characteristics of soils are
accounted for by classifying soils into four groups based
                                                                                                  123

-------
 HYDROLOGY:  SOLUTION  OF  RUNOFF  EQUATION   Q=(Clnil)2
                                                                    P+0.8S
                                                                                P= 0 to 12 inches
                                                                                Q = 0 to 8 inches
                          Rainfall  (P)
                          RUNOFF  (Q)
                   With  P* in:  S2ln+F;
                                                     . L  and F.P-I0-Q|;
                                                     • VTT i i  • : • i : : • • I • • 11
                                      .'Curves on this  sheet are  for
                                           case Ia« 0.2 S, so that
         ;lmtiah
         -abstraction  In iTTTT
Infiltration-
  curve1•H
                                              4567
                                                 RAINFALL  (P) IN  INCHES
REFERENCE
  Mockus, Victor;  Estimating  direct runoff amounts from  storm rainfall:
              Central  Technical Unit, October 1955
                                                             u. a MFAKnmrr or AOUCUWJKE
                                                            SOIL OONSEBVATION SEHV1CB
                                                            ftmoBnn omnt - wotouoaf UANCB
STANDARD DUG. NO.
ES- 1001
SHEET 1  Of  2
OKIE  6-29-56
                                                                                                                     REVBED 10-1-64
                                       Figure l.-Solutions of Eq. 3. [From SCS National Engineering Handbook (4)]

-------
      Table 1. Seasonal rainfall limits for antecedent
                moisture conditions1
        AMC group
 Total 5-day antecedent rainfall

Dormant season Growing season

I
II
III
inches
<0.5
0.5-1.1
>1.1
inches
. <1.4
1.4-2.1
>2.1
   1 From SCS National Engineering Handbook (4).
upon  the minimum rate of infiltration obtained for a
bare soil after prolonged wetting. The influences of both
the surface and the profile of a soil are included. The
hydrologic soil groups as defined by SCS soil scientists in
the National Engineering Handbook are:
A. (Low runoff potential). Soils having high infiltration
rates  even  when  thoroughly  wetted  and consisting
chiefly of deep, well to excessively drained sands or
gravels. These  soils have a high rate of water transmis-
sion.
B. Soils  having  moderate   infiltration  rates  when
thoroughly wetted and consisting chiefly of moderately
deep to deep, moderately well to well drained soils with
moderately fine to  moderately coarse textures. These
soils have a moderate rate of water transmission.
C. Soils having slow infiltration rates when thoroughly
wetted and consisting chiefly of soils with a layer  that
impedes downward  movement  of water, or soils with
moderately fine to fine  texture. These soils have a slow
rate of water transmission.
D. (High runoff  potential). Soils having very  slow
infiltration rates when thoroughly wetted and consisting
chiefly of clay soils  with a high swelling potential, soils
with a permanent high water table, soils with a claypan
or clay layer at or  near the surface, and  shallow soils
over nearly impervious material. These soils have a very
slow rate of water transmission.
   The SCS has classified over 9,000 soils in the United
States and  Puerto Rico according to the above scheme.
A sample  from the extensive table in the SCS National
Engineering Handbook is  shown in Table 2.  Rainfall-
runoff data from small watersheds or infiltrometer plots
were  used to make the classifications where such  data
were  available, but most are based on the judgement of
  'il scientists  and correlators who  used physical prop-
erties of the soils in making the assignments.
   The  interaction of hydrologic soil group (soil) and
land  use  and  treatment  (cover)  is accounted for by
assigning a runoff curve number for average soil moisture
condition (AMC II) to important soil cover complexes
for the  fallow  period and the growing season. Rainfall-
runoff data for single soil cover complex watersheds and
plots were analyzed to provide a basis for making these
assignments. Average runoff curve numbers for several
soil-cover complexes are  shown in Table  3.  Average
runoff curve numbers (AMC II) are for the average soil
moisture  conditions. AMC  I  has the  lowest runoff
potential. AMC III has  the highest runoff potential.
Under this condition the watershed  is practically satu-
rated from antecedent rains. Appropriate curve numbers
for AMC I and III based  upon the curve number for
AMC II  are shown in Table 4.
   Curve numbers for a  "good hydrologic  condition"
were  used in  the potential  direct runoff simulations.
"Hydrologic condition" refers to the runoff potential of
a particular cropping practice. A row  crop  in good
hydrologic condition will  have higher infiltration rates
and, consequently, less direct runoff than the same crop
in poor hydrologic condition. Good hydrologic condi-
tion seemed an appropriate  description  of corn  under
modern management practices.
   Seasonal variation not accounted for by the seasonal
dependency of the AMC classes is included by varying
the average moisture condition curve number according
to the stages  of growth of  a particular crop.  For the
simulations reported here, with straight row corn as the
index crop, the average (AMC II) curve number was set
equal to that  for  fallow for the  period  from March  1
until  the  average emergence date  for corn. Emergence
dates were assumed  to be  2 weeks  after the average
planting date  reported by the USDA (5).  During the
growing  season, AMC 41 curve numbers for each day
were calculated by the following equation:
                              where
                                 CNj  = the curve number for the ith day for AMC
                                         II.
                                 F    = fallow curve number.
                                 Cj   = crop coefficient for the ith day. Cj < 1.
                                 Cave = average  crop  coefficient  for the  growing
                                         season.
                                 CNave= average  growing  season  curve number for
                                         AMC II.
                                 The crop coefficients Cj are  defined as the ratio of
                              the crop evapotranspiration to potential evapotranspira-
                              tion for a given day when soil water is not limiting. Crop
                                                                                                       125

-------
                                Table 2.—Soil names and hydrologic classifications1 (Sample)
AA8ERG
AASTAD
ABAC
AdAjO
ABBOTT
AdBOTTSTUtfN
A8CAL
AifcGG
ABEL*
ABELL
AdtROctN
ABtS
ABUcNc
AalNbTON
ABIguA
ABO
ABOR
ABRA
ABRAHAM
AUSAnOKfct
A6SCOTA
Att&HER
ABSTcO
ACACIO
ACADEMY
ACAD1A
ACANA
ACASCO
A i. c IT UN AS
ACEL
ACKER
ACKMtN
ACMt
ACO
ACULITA
ACUHA
ACOVt
ACRtt
ACKELANc
Aw TON
A;UFF
ACWCIRTH
ACY
ADA
AOAJK
ADAMS
AOAMSUN
AOAMSTJMN
AdAMSVILLc
AOATUN
A3AVEN
ADOIELDU
ADD I SON
AODV
AUE
AOcL
AOELAIDfe
ADELANTU
AOEL INU
ADELPH1A
ADfcNA
AOCER
ADILIS
AOIRJNDACK
ADIV
ADJUNTAS
AOKINS
ADLtR
AOOLPH
ADRIAN
AENEAS
AETNA
AFTON
AGAi*
AGASSIZ
AGATe
A GAM AM
AiifcNCY
AutR
AGNcR
AuNEM
AGNUS
AGUA
A GU AD ILL A
AGUA OULtfc
AGUA FRIA
AGUALT
AGUEUA
AGU1L1TA
AGUIRRE
A GUST IN
AHATONE
C
B
0
C
D
C
0
B
B
B
0
0
^
0
C
B/C
D
C
B
C
B
0
u
C
c
0
b
0
B
0
B
B
C
B
B
C
C
c
c
B
B
B
C
a
0
A
B

C
0
D
C
D
C
A
A
D
B
a
c
c
D
A

B
C
B
C
D
A/0
B
B
0
B
0
0
B
C
u
b
d/C
a
B
A
C
B
B
8
B
U
B
0
NJTES A
AHL
AHLSTRON
AHMEtK
AHOLT
AHTANUM
AHMAHNbt
AIBtlNITQ
AIKEN
AIKMAN
AILEY
AINAKEA
AIRMONT
AIROTSA
AIRPORT
AITS
A JO
AKAKA
AKASKA
AKkLA
ALADDIN
ALAfc
ALAELOA
ALAGA
ALAKAI
ALA MA
ALAMANCE
ALAMO
ALAMOSA
ALAPAHA
ALAPAI
ALBAN
ALBANC
ALBANY
ALBATON
ALBEt
AL8EMARLE
ALBERWILLE
ALBIA
ALBION
ALBRIGHTS
ALCALDE
ALCtSTER
ALCOA
ALCONA
ALCOVA
ALOA
ALOAX
ALDEN
ALDER
ALDERDALE
ALDfcRMOOD
ALBINO
ALDMELL
ALEKNAGIK
ALCMEDA
ALEX
ALEXANDRIA
ALEXIS
ALFOKD
ALGANSEb
ALGERITA
ALGIERS
ALGOMA
ALHAMBRA
ALICE
ALICEL
ALICIA
ALIUA
ALIKCHI
ALINE
ALKO
ALLAGASH
ALLARD
ALLEGHENY
ALLEMANOS
ALLEN
ALLENUALE
ALLtNS PARK
ALLENSV1LLE
ALLENTINE
ALLENMOOD
ALLESSIO
ALLEY
ALLIANCE
ALLIGATOR
ALLIS
ALLISON
ALLOUEZ
ALLOMAV
ALMAC
ALMENA
ALMONT
BLANK HYDROLOGIC
TWO SOIL GROUPS SUCH AS
C
C
a
D
c
c
c
8/C
0
B
B
C
B
D
B
C
A
B
C
B
A
B
A
0
B
B
D
C
0
A
B
D
C
D
C
B
C
C
B
C
C
B
B
B
B
C
0
D
B
C
C
C
C
B
C
B
C
B
B
B
B
C/0
B/D
B
A
B
B
B
B
A
0
B
B
B
0
B
C
B
C
D
B
8
C
B
0
0
C
C

B
C
D
SOIL
ALMY
ALOHA
ALONSO
ALOVAA
ALPENA
ALPHA
ALPON
ALPOMA
ALPS
ALSfcA
ALSPAJGH
ALSTAD
ALSTON
ALTAMONT
ALTAVISTA
ALTO'ORF
ALTMAR
ALTO
ALTOGA
ALTON
ALTUS
ALTVAN
ALUM
ALUSA
ALVIN
ALVIRA
ALVISO
ALVOR
AHAOUR
AMAGON
AMALU
AMANA
AMARGOSA
AMARILLO
AMASA
AMBERSON
AMBOY
AMBRAM
ANEDEE
AMELIA
AMEN1A
AMERICUS
AMES
AMESHA
AMHERST
AMITY
AMMON
A MOLE
AMOR
AMOS
AMSDEN
AMSTERDAM
AMTOFT
AMY
ANACAPA
ANAHUAC
ANAMITE
ANAPRA
ANASAZ1
ANATONE
ANAVERDE
ANAWALT
ANCHO
ANCHORAGE
ANCHOR BAY
ANCHOR POINT
ANCLOTE
ANCO
ANDERLY
ANDERS
ANDERSON
ANDES
ANOORINIA
AND OVER
ANDREEN
ANDREESON
ANDRES
ANDRE* S
A NED
ANETH
ANGELICA
ANGELINA
ANGELO
ANGIE
ANGLE
ANGLEN
ANGOLA
ANGOSTURA
ANHALT
ANIAK
ANITA
ANKENY
GROUP INDICATES
B
C
B
C
B
C
B
B
C
B
C
B
B
D
C
D
B
C
C
B
B
B
8
D
B
C
D
C
0
D
D
B
D
B
B

C
C
A
B
B
A
C
B
C
C
B
C
B
C
B
B
0
D
B
0
D
8
B
D
8
D
B
A
D
D
D
C
C
C
B
C
C
0
B
C
8
C
0
A
D
B/D
C
C
A
B
C
B
0
0
0
A
ANLAUF
ANNABRLA
ANNANOALE
ANNISTON
ANOKA
ANONES
ANSARI
ANSEL
ANSELM3
ANSON
ANTELOPE SPRINGS
ANTERO
ANT FLAT
ANTHO
ANTHONY
ANT I GO
ANTILON
ANTIOCH
ANTLER
ANTOINE
ANTROBUS
ANTY
ANVIK
ANWAY
ANZA
ANZIANO
APACHE
APAKUIE
APISHAPA
APISON
APOPKA
APPIAN
APPLEGATE
APPLE TON
APPLING
APRON
APT
APTAKIS1C
ARABV
ARAOA
ARANSAS
ARAP I EN
ARAVE
ARAVETON
ARBELA
ARBONE
ARBOR
ARBUCKLE
ARCATA
ARCH
ARCHABAL
ARCHER
ARCH IN
ARCO
ARCOLA
ARD
AROEN
ARDENVOIR
ARCH LA
AREDALE
ARENA
ARENALES
AR.ENOTSVILLE
ARENOSA
ARENZVILLE
ARGONAUT
ARGUELLO
ARGVLE
ARIEL
ARIZO
ARKABUTLA
ARKPORT
ARLAND
A RLE
ARLING
ARLINGTON
AR10VAL
ARMAGH
ARMIJO
ARMINGTON
ARMO
ARMOUR
ARMSTER
ARMSTRONG
ARMUCHEE
ARNEGARO
ARNHART
ARNHE IM
ARNO
ARNOLD
ARNOT
ARNY
C
B
C
B
A
C
D
B
A
B
C
C
C
B
B
B
B
0
C
C
B
B
B
B
B
C
D
A
C
B
A
C
C
C
B
B
C
B

C
0
c
a
B
c
B
B
B
B
B
B
C
C
B
C
C
B
8
C
B
C
A
B
A
B
D
B
B
C
A
C
B
B
B
D
C
C
D
D
D
8
B
C
D
D
B
C
C
D
B
C/0
A
THE SOIL GROUP' HAS NOT BEEN
AROOSTOOK
AROSA
ARP
AURINGTON
AR(tIT3LA
ARROLIME
A!UON
ARROW
ARROMSMITH
ARROYO SECO
ARTA
ARTOIS
ARVADA
A«VANA
ARVESON
ARVILLA
ARZELL
ASA
ASBURV
AS;ALON
ASCHOFF
ASHBV
ASHCROFT
ASHDALE
ASHE
ASHKUN
ASHLAR
ASHLEY
ASH SPRINGS
ASH TON
ASXUE
ASNUELOT
ASHuaoo
ASKEW
ASO
AS3TIN
ASPEN
ASPERHONT
ASSINNIBOINE
ASSUMPTION
ASTATULA
ASTOR
ASTORIA
ATASCkDERO
ATASC3SA
ATCO
ATENCIO
ATEPIC
ATHELHOLD
ATHENA
ATHENS
ATHERLY
ATMERTON
ATHNAR
ATHOL
ATKINSON
ATLAS
ATLEt
ATHDRE
AT3KA
ATON
ATRVPA
ATSION
ATTERBERRV
ATTEM4N
ATTICA
ATTLEBORO
ATMATER
ATMELL
ATHOOD
AUBBEENAUBBEE
AUBERRV
AUBURN
AUBURNDALE
AUDI AN
AU GRES
AUGSBURG
AUGUSTA
AULO
AURA
AURORA
AUSTIN
AUSTKELL
AUXVASSE
AUZQUI
AVA
AVALANCHE
AVALON
AVERY
AVON
AVONBURG
AVONDALE
DETERMINED

C
C
B
0
C
0
B
B
B
C
C
D
C
D
B
C
B
B
B
B
C
B
B
B
C
B
A
C
B
B
C
C
C
C
C
B
8
8
B
A
A/D
B
C
D
B
B
D
B
B
B
B
B/D
C
B
B
D
C
B/D
C
B
C
C
B
A
B

B
C/0
B
B
B
C/0
0
B
C
B
C
D
B
C
C
D
D
B
C
B
B
B
C
D
£

B/C INDICATES THE OR AI NEO/UNDRAI N ED SITUATION
            1 From SCS National Engineering Handbook
126

-------
                      Table 3.-Runoff curve numbers for hydrologic soil-cover complexes1
                               (Antecedent moisture condition II, and I  = 0.2 S)
Cover

Land use Treatment or practice
Fallow Straight row
Row crops "
a
Contoured
"
" and terraced
	
Small grain Straight row
n
Contoured
n
" and terraced
n n n
Close-seeded legumes2 Straight row
or rotation meadow "
Contoured
n
" and terraced
n H n
Pasture or range


Contoured
"
"
Meadow
Woods


Farmsteads
Roads (dirt)3
(hard surface)
' From SCS National Engineering Handbook (4).
, Close-drilled or broadcast.






hydrologic sou group
Hydrologic condition
A
	
Poor
Good
Poor
Good
Poor
Good
Poor
Good
Poor
Good
Poor
Good
Poor
Good
Poor
Good
Poor
Good
Poor
Fair
Good
Poor
Fair
Good
Good
Poor
Fair
Good
	
....
	


77
72
67
70
65
66
62
65
63
63
61
61
59
66
58
64
55
63
51
68
49
39
47
25
6
30
45
36
25
59
72
74


B
86
81
78
79
75
74
71
76
75
74
73
72
70
77
72
75
69
73
67
79
69
61
67
59
35
58
66
60
55
74
82
84


C
91
88
85
84
82
80
78
84
83
82
81
79
78
85
81
83
78
80
76
86
79
74
81
75
70
71
77
73
70
82
87
90


D
94
91
89
88
86
82
81
88
87
85
84
82
81
89
85
85
83
83
80
89
84
80
88
83
79
78
83
79
77
86
89
92


Including right-of-way.
                                                                                                           127

-------
                          Table 4.-Curve numbers (CN) and constants for the case Ia = 0.2S1
CN for
condi- CN.for
... _ conditions
"°n i HI

100
99
98
97
96
95
94
93
92
91
90
89
88
87
86
85
84
83
82
81
80
79
78
77
76
75
74
73
72
71
70
69
68
67
66
65
64
63
62
61

100
97
94
91
89
87
85
83
81
80
78
76
75
73
72
70
68
67
66
64
63
62
60
59
58
57
55
54
53
52
51
50
48
47
46
45
44
43
42
41

100
100
99
99
99
98
98
98
97
97
96
96
95
95
94
94
93
93
92
92
91
91
90
89
89
88
88
87
86
86
85
84
84
83
82
82
81
80
79
78
s a
values
(inches)
0
.101
.204
.309
.417
.526
.638
.753
.870
.989
1.11
1.24
1.36
1.49
1.63
1.76
1.90
2.05
2.20
2.34
2.50
2.66
2.82
2.99
3.16
3.33
3.51
3.70
3.89
4.08
4.28
4.49
4.70
4.92
5.15
5.38
5.62
5.87
6.13
6.39
Curve2
starts
where
P =
(inches)
0
.02
.04
.06
.08
.11
.13
.15
.17
.20
.22
.25
.27
.30
.33
.35
.38
.41
.44
.47
.50
.53
.56
.60
.63
.67
.70
.74
.78
.82
.86
.90
.94
.98
.03
.08
.12
.17
.23
.28
CNfor
condi-
tion
II

60
59
58
57
56
55
54
53
52
51
50
49
48
47
46
45
44
43
42
41
40
39
38
37
36
35
34
33
32
31
30

25
20
15
10
5
0


CNfor
conditions
1 III

40
39
38
37
36
35
34
33
32
31
31
30
29
28
27
26
25
25
24
23
22
21
21
20
19
18
18
17
16
16
15

12
9
6
4
2
0



78
77
76
75
75
74
73
72
71
70
70
69
68
67
66
65
64
63
62
61
60
59
58
57
56
55
54
53
52
51
50

43
37
30
22
i3
0


s 2
values
(inches)
6.67
6.95
7.24
7.54
7.86
8.18
8.52
8.87
9.23
9.61
10.0
10.4
10.8
11.3
11.7
12.2
12.7
13.2
13.8
14.4
15.0
15.6
16.3
17.0
17.8
18.6
19.4
20.3
21.2
22.2
23.3

30.0
40.0
56.7
90.0
190.0
infinity


Curve2
starts
where
P =
(inches)
1.33
1.39
1.45
1.51
.57
.64
.70
.77
.85
1.92
2.00
2.08
2.16
2.26
2.34
2.44
2.54
2.64
2.76
2.88
3.00
3.12
3.26
3.40
3.56
3.72
3.88
4.06
4.24
4.44
4.66

6.00
8.00
11.34
18.00
38.00
infinity


             1 From SCS National Engineering Handbook (4).
             2 For CN in Column 1.
128

-------
coefficient curves for  corn  were obtained by fitting a
Fourier  Series to a curve  presented by Kincaid  and
Heermann (2).
   Curve  numbers for antecedent moisture conditions I
and HI were obtained from Table 4.
   At harvesting date or when CNj = CNave, whichever
came first, the curve number was set equal to CNave and
remained a constant until the next March 1. The SCS
recommends  that the after-harvest curve number be set
equal to the average growing season curve number if 1/3
of the soil surface is exposed. This simulation represents
a situation where residues  are left on  the field after
harvest.
   Any precipitation that  occurred when the mean air
temperature was less than  O°C was assumed to be snow
and was accumulated as snow storage until the temper-
ature  went above  0°C. Snowmelt was calculated by
using a degree-day factor:
    S   = KT                                  (6)
where
    S   = snowmelt in inches
    K  =  degree-day snowmelt factor (inches/day/0C)
    T  =  mean daily temperature, °C

   A  degree-day snowmelt  factor K  = 0.18 in/day/°C
(0.10 in/day/0 F) was used in all calculations. This is
approximately the mid-range  of the  values quoted by
Linsley, Kohler and Paulhus, (0.06 -0.15in/day/°F)(5).
The snowmelt calculated in this manner was used to
estimate the antecedent moisture condition and the SCS
curve  number procedure was used to estimate snowmelt
runoff. The SCS National Engineering Handbook does
not recommend the use of curve numbers in estimating
runoff from  snowmelt, because there is no way to
account for frozen ground. SCS  considers the entire
snowmelt as computed by equation  (5) to be runoff,
which is good  practice when one is concerned with
floods. Because  this  study was not concerned with
floods, it was deemed more appropriate to use the curve
number  procedure  to estimate snowmelt runoff despite
its  limitations.  Obviously, the  snowmelt runoff as
calculated may have significant errors.
                                                                                                    129

-------
                                     SIMULATION PROCEDURE
                       Data

   The daily precipitation data and temperature  data
required for the simulations were obtained on magnetic
tape from the National Climatic Center, Environmental
Data  Service,  NOAA,  U. S. Dept.  of Commerce, at
Asheville,  N. C. The data set obtained is termed Day
Deck 345.  The  normal period of record  was from
January 1948, through December 1973. A year begin-
ning on March 1 was used in all simulations. The stations
used are listed in Table 5.
   Simulations were performed only for stations east of
the Rocky Mountains for the following reasons:
   1. This report  is intended to cover only nonirrigated
cropland  and  much  of the  cropland  in the  West is
irrigated.
   2. Rainfall gradients tend to be  very steep in the
West because of orograpnic effects. Therefore, interpola-
tion  between widely  separated meteorologic  stations
would be misleading.

               Computer Program

   The  program  SCSRO (Soil Conservation  Service
Runoff) was written in  FORTRAN IV. A  generalized
flow chart is shown in Fig. 2.

       Assigning Hydrologic Soil Groups to
          Land  Resource Areas (LRAs)

   Land Resource Areas (LRAs) are shown in the map in
Fig. 2, Vol. I  and are  discussed in Section 3.1, Vol. I.
Although Land Resource Areas are defined as geographic
areas characterized by  a particular  pattern of soil type,
topography, climate,  water resources, land use and type
of farming (1) they are large enough that each  of these
factors varies significantly within the area. Therefore, it
is  impossible to characterize an entire LRA by a single
soil series. In many cases, however, the  major soil series
listed for  a  LRA  have similar hydrologic characteristics
in that they fall into one hydrologic soil group. Where
there is a  wide range of hydrologic characteristics within
a  LRA the  hydrologic soil group  of the predominant
agricultural soil was used.
   The simulation results  shown in Vol. I and in  this
Appendix should not be considered representative of the
entire  LRA. However, they  are representative of the
predominant agricultural soils of the LRA,  subject, of
Table S.-Meteorological records used in simulations
Location
Wichita, KA
Columbia, MO
Dodge City, K A
Kansas City, MO
Springfield, MO
Chicago, IL
Cleveland, OH
Columbus, OH
Lansing, MI
Sault Ste. Marie, MI
Green Bay, WI
Fargo, ND
LaCrosse.WI
Des Moines, IA
Grand Island, NB
Huron, SD
Omaha, NB
Sioux Falls, SD
Bismark, ND
Williston, ND
Scottsbluff, NB
Rapid City, SD
Cairo, IL
Indianapolis, IN
Lexington, KY
Springfield, IL
Savannah, GA
Miami, FL
Houston, TX
Brownsville, TX
Raleigh/Durham, NC
New Castle/Wilmington,
DE
Charleston, SC
Columbia, SC
Jacksonville, FL
Memphis, TN
Mobile, AL
Lake Charles, LA
Dallas, TX
Little Rock, AR
Oklahoma City, OK
Buffalo, NY
Newark, NJ
Boston, MA
Portland, ME
Syracuse, NY
Wilkes-Barre/Scranton,
PA
El Paso, TX
Amarillo, TX
Cape Hatteras, NC
Tallahassee, FL
Pittsburgh, PA
Period of
record
48-67
48-67
48-67
43-67
43-67
43-67
48-67
48-67
49-53,60-69
47-66
50-69
48-67
48-67
46-70
50-70
43-67
48-67
43-67
48-68
35-62
48-67
49-68
Missing
years
none
"
"
'*
"
"
H
"
"
II
'"
"
"
It
"
"
"
"
62
48,49,50
none
a
30-67 48-53,56-62
48-67
48-67
43-67
51-71
48-68
48-68
48-68
48-68
48-68

48-68
48-68
48-68
48-68
48-69
48-58
48-68
48-68
48-67
48-69
48-68
48-68
48-68
48-68
49-53,56-71

48-68
48-68
57-69
48-68
48-68
none
ii
"
52
49
49
49
49
49

49
49
49
49
49,62
49
49
49
none
49,65
49
49
49
49
50

49
49
58
49
49
Total
years
20
20
20
25
25
25
20
20
15
20
20
20
20
25
21
25
20
25
20
25
20
20
25
20
20
25
20
20
20
20
20
20

20
20
20
20
20
10
20
20
20
20
20
20
20
20
20

20
20
12
20
20
130

-------
          Read
          Field
          Character 1st ic s
          CALL SUBROUTINE
              CN
          CALL SUBROUTINE
              TPREAD
CO
oc
<
LLJ
LL.
O
LU
CO
LU
CC
53
LU
a
LU
CL.
 Subroutine CN computes
 an average curve number for
 each day of the year.
 Subroutine TPREAD reads in
.1 year of daily precipita-
 tion and temperature data.
          CALL SNOW
          CALL AMCF
 Subroutine SNOW determines
 which daily values of precipi-
 tation  are snow,  accumulates
 it  and converts it to snowmelt
 on  the appropriate days.
 Subroutine AMCF computes the
• antecedent precipitation
 index for  each day.
          CALL SRO
 Subroutine SRO computes daily
 surface runoff using the SCS
 curve number procedure.
          Calculate rainfall
          and runoff statistics
          for 14-day periods,
          seasons , and year.
                   I
                PRINT
                      Figure 2.-Program SCSRO flowchart
                                                                        131

-------
course, to the limitations of the SCS runoff estimation
procedure.  Predominant  agricultural soil series in  each
LRA were  obtained  from  Austin (1) and  hydrologic
classifications  for  the soil  series were  obtained from
Table  7.1 of the SCS National Engineering Handbook
(4). A list of the LRA's and the assigned hydrologic soil
groups is presented in Table 6. These assignments were
reviewed and modified  by personnel of the Technical
Service Centers of SCS and their assistance is gratefully
acknowledged.
Table 6.—Hydrologic soil group and available water holding
   capacities for predominant agricultural soils in land
   resource areas
Table 6.-Hydrologic soil group and available water holding
   capacities for predominant agricultural soils in land
   resource areas-Continued
Lantl r>~m;—« \,,,Ar~\s>^.^ Available water
rpcr,nr™> Dominant hydrologic
resource soil croup capacity in
area sou B1™? 4.ft. root zone

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43-51
52
53
54
55
56
57

Mountains
B
Mountains
Soil Information Lacking
Forest
Mountains
B
B
C
C
B
Mountains
B
B
D
D
D
D
D
Mountains
D
Mountains
Soil Information Lacking
" " "
D
D
D
Desert
a
"
Irrigated Desert
B
Mountain
B
B
C
B
Mountain
"
Soil Information Lacking
n a n
B
Mountains
B
B
B
B
D
B
(inches)

8




8
8
8
8
8

8
8
6
6
6
6
6

6



6
6
6




8

8
8
6
8




8

8
8
*6
8
4
8
Land
resource
area

58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
Dominant hydrologic
soil group

B
C
D
B
Mountains
D
B
A
B
B
B
B
C
B
B
B
B
B
D
C
C
A
C
D
C
D
B
D
D
D
Forest
Forest
B
A
Forest
Forest
Forest
B
A
B
B
D
B
B
B
B
C
B
B
Available water
capacity in
4-ft. root zone
(inches)
*6
8
4
8

4
8
2
8
8
8
8
6
8
8
8
8
8
*4
8
8
4
8
6
8
6
8
6
6
6


8
2



8
2
4
8
6
8
8
8
8
8
8
8
132

-------
Table 6--Hydrologic soil group and available water holding
    capacities for predominant agricultural soils in land
    resource areas-Continued
Land
resource
area

107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
Dominant hydrologic
soil group

B
B
C
C
C
D
D
D
B
C
Mountains
D
Mountains
C
C
B
C
C
Mountains
C
Mountains
B
B
Mountains
D
D
B
C
D
B
A
B
C
C
C
D
Mountains
A
B
C
B
C
C
D
Swamp
D
C
A
B
Swamp
Available water
capacity in
4-ft. root zone
(inches)
8
8
*6
*6
8
*4
*4
4
8
*4

6

*4
8
8
8
8

8

8
8

6
6
8
8
6
8
4
4-
*4
8
8
6

4
8
8
8
8
8
6

6
6
4
8

   * Available water-holding capacity reduced because root
zone is shallower than 4 feet.
                                                                                    133

-------
                                        SIMULATION RESULTS
   Program SCSRO output for each 14-day period for
the n years of record may include:
   1. A  listing of rainfall  amounts ordered by magni-
tude.
   2. A  listing of simulated  runoff ordered by magni-
tude.
   3. The mean and standard deviation of rainfall and
runoff events.
   The  statistical  summary  for the n-year simulation
included:
   1. A  table showing the  number of runoff events for
each 14-day period for each year.
   2. The  probability  that there would be no  runoff
events in any year for each 14-day period.
   3. The mean annual simulated runoff.
   4. The mean growing season simulated runoff.
   Maps  of the mean annual and  seasonal simulated
runoff (potential direct  runoff)  are  shown  for  each
hydrologic  soil group in Figs. 3 through  10. Because of
the relatively small area of soils classified in hydrologic
group  A, simulations with  this group  were not per-
formed for all rainfall stations. The growing season was
taken  as  the time  interval  between emergence and
harvest and varied with location. These maps can be used
to supplement the information presented in Figs. 3 and
4 in Vol. I.
   Top few rainfall stations were used in  this analysis to
depict  climatic  and orographic influences in the Appa-
lachian Mountains; therefore,  care must be used  in
interpreting the maps in these regions.
134

-------
                            Figure 3.-Mean annual potential direct runoff in inches. Straight-row corn in good hydrologic condition-Hydrologic Soil Group A.
U»

-------
OJ
                                                                                                                                                      0
                            Figure 4.-Mean annual potential direct runoff in inches. Straight-row corn in good hydrologic condition -Hydrologic Soil Group B.

-------
Figure 5.-Mean annual potential direct runoff in inches. Straight-row corn in good hydrologic condition-Hydrologic Soil Group C.

-------
OJ
oo
                            Figure 6.-Mean annual potential direct runoff in inches. Straight-row corn in good hydrologic condition -Hydrologic Soil Group D.

-------
                                                                                                                                         94
                          Figure 7.—Mean growing season potential direct runoff in inches. Straight-tow corn in good hydrologic condition-Hydrologic Soil Group A.
UJ

-------
Figure 8.-Mean growing season potential direct runoff in inches. Straight-row corn in good hydrologic condition-Hydrologic Soil Group B.

-------
Figure 9.-Mean growing season potential direct runoff in inches. Straight-row corn in good hydrologic condition-Hydrologic Soil Group C.

-------
Figure lO.-Mean growing season potential direct runoff in inches. Straight-row corn in good hydrologjc condition-Hydrologic Soil Group D.

-------
                     CONSISTENCY CHECK AND DISCUSSION OF ERRORS
  The  accuracy of the simulated runoff amounts was
checked by comparing  average annual potential direct
runoff with measured average annual runoff from several
small, single-crop watersheds in straight-row crops (pri-
marily corn and cotton) (7, 8). The watersheds used in
this comparison and a brief data summary are presented
in Table 7. The following linear regression equation was
obtained:
       Qs  =  1.365 +0.578 Qf
(7)
where

        Qs  =   simulated average annual direct runoff.
        Q0  =   observed average annual runoff.
   The coefficient of variation (r2) was 0.616. A scatter
diagram of computed versus  observed average annual
runoff is shown in Fig. 11. The vertical lines emanating
from the plotted points indicate the range in simulated
runoff that would occur if the soils were  assumed  to
belong to adjacent hydrologic soil groups. This indicates
                                  Table 7.-Surface runoff consistency check
Watershed
College Park, MDW-3

Americus, GA W-l

Lafayette, IN W-4

" W-5

" W-8

" W-10

" W-l 2

" W-l 3

" W-l 5

Clarinda, IA W-V

" W-W

" W-Y

Coshocton.OH W-l 15
" W-l 10
" W-l 18
" W-192
" W-l 06
Guthrie, OK W-2
Garland TX W-III

Spur,TX W-2

Riesel, TX Y-7
Hastings, NB 3-H
Oxford. MS WC-1
WC-3
C hickasha, OK C-l
Area
acres
6.06'

17.9

2.01

2.87

1.96

2.06

3.37

3.02

3.59

3.25

1.97

3.25

1.61
1.27
1.96
7.59
1.56
3.21
10.4

9.39

40.0
3.95
3.88
1.61
17.8
Crop
Soybeans
Sweet corn
Corn
Cotton
Corn
Soybeans
Corn
Soybeans
Corn
Soybeans
Corn
Soybeans
Corn
Soybeans
Corn
Soybeans
Corn
Soybeans
Corn

Corn

Corn

Corn
Corn
Corn
Corn
Corn
Cotton
Cotton
Corn
Cotton

Row crops
Corn
Corn
Corn
Corn
Hydrologic
soil group
B

B

B

84% B
16% C
B

66% B
34% C
89% B
11% C
80% B
20% C
B

90% B
10% D
90% B
10% D
72% B
28% D
C
C
C
C
C
B
D

50% B
50% C
D
B
C
C
81% C
Mean annual
Observed
2.47

1.27

2.26

4.14

4.89

5.88

3.98

3.93

4.33

1.41

3.37

1.06

2.85
2.41
1.97
3.13
3.23
7.67
11.67

2.70

7.00
4.85
14.99
13.77
1.24
runoff (in.)
Simulated
2.60

7.17

2.60

3.05

2.60

3.57

2.91

3.17

2.60

2.32

2.32

2.89

3.05
3.05
3.05
3.05
3.05
2.30
8.30

2.05

8.75
1.65
11.00
11.00
3.32
                                                                      19% B
                                                                                                   143

-------
    15
V)

•o
 0>
O
5


4
	,_  v^


I'*
     1
    0
      Qs= l.365-t-0.578 Q0

      r2=0.6!6

      s.e. = l.44"
                         I  Std.  Dev
                          of Mean
             8. o
             C CO
             o

            o ^
               0»
             u. O E
            ^ o O
               1_
              (O 1-)
                      a.
                      3
Regression
   Line
      0    I    2    3    4   5    6    7   8   9    10   II    12   13    14

          Q0 ,  Observed Mean  Annual Direct Surface Runoff, inches


                   Figure 11.-Simulated and observed mean annual direct surface runoff.
                                                                        15
144

-------
the inherent limitation in the SCS method caused  by
lumping all soils into four distinct groups. Horizontal
bars  emanating from plotted  points indicate estimated
standard  errors of  the  mean of observed data. This
illustrates the problem of short, fragmented records. The
SCS  method tended to underestimate runoff of more
than 3 inches.
   Meaningful  comparisons  are  difficult  because the
observations were made over  such a long time. Agricul-
tural practices have changed  drastically so the data are
not  stationary. For  example, hybrid corn and higher
fertility levels  have led to rapid canopy establishment
and more residues after harvest. The simulated condition
after harvest—approximately  67%  cover by residues—
probably is not consistent with the practices on water-
sheds where the data were obtained. One would antici-
pate that the simulated runoff would be less than the
observed in this case. Data from some of the watersheds
listed in Table 7 were undoubtedly used in developing
the  SCS curve number  procedure  so this is not  an
independent test of its predictive capabilities.
   Although the  relationship between simulated  and
observed direct runoff shown in Fig. 11 is not as good as
one would wish, it must be compared with the available
alternatives before  one can  judge  its usefulness. One
alternate that has been suggested is to use the map of
surface-water runoff prepared by  the  U.S. Geological
Survey (6) as an indicator of potential loss by direct
runoff.  To test this method, consider the  following
regression relationship between the average annual run-
off from the USGS map for the locations in Table 7 and
the observed average annual direct runoff:
      QG = 6.74 + 0.503 Q0 ;  r2 = 0.138
(8)
 where QQ is average  annual  surface-water runoff from
 the USGS map.
   The simulated results obviously are superior to those
 obtained from the runoff map as indicators of potential
 direct runoff.
   One would anticipate that the sum of the simulated
 average annual direct runoff and the average annual deep
 percolation  estimated by the procedures described  in
 Appendix  B of this  volume  should  be  rather  well
      correlated with the average annual streamflow from the
      USGS maps.
         The following regression equation was obtained be-
      tween  simulated direct  runoff plus percolation and
      runoff (streamflow) from the USGS map for 45 of the
      52 meteorological stations used:
        (Qs + Qp) = 0-409 + 0.979 QG ;  r2 = 0
                                        1.884
(9)
where Q_ is the simulated average annual deep percola-
tion. Seven  stations in the karst area of Florida and in
the coastal area of the Southeastern United States were
omitted because anomalies on the USGS map indicate
that much of the groundwater runoff flows directly into
the ocean. These crude checks indicate that the simula-
tions  provide  reasonable estimates of annual  direct
runoff and percolation.
   Records of runoff from continuous straight-row corn
were not  readily available for periods shorter  than one
month so it was not  possible to check the accuracy of
the time distribution of simulated potential direct runoff
within the year. However, 20 years of runoff data were
available for a small watershed in meadow at Coshocton,
Ohio. Direct runoff for 14-day periods simulated by the
SCS procedure  was  compared  with  observed  data.
Simulated and observed  mean runoff per event, mean
number of runoff events per period, and  mean  runoff
amount per period are shown in  Fig. 12. The  standard
deviation of the observed runoff per period is indicated
by a vertical line for each  period.
   The  simulated  runoff  per period  is  within  one
standard deviation  of the observed runoff for 14 of the
26  periods.  Assuming that the mean value is  normally
distributed for each period and that the 26 periods are
independent trials, the null  hypothesis cannot  be re-
jected at the 10 percent level.
   Sample distribution functions  of runoff amount per
event for two periods are shown in Fig. 13. A Kolmo-
gorov-Smirnov test comparing the distribution functions
indicates  that the  null hypothesis cannot be rejected at
the 10 percent level for both periods.
   Although it is impossible  to make strong inferences
on the  basis of the limited  tests performed, the SCS
method appears adequate for arriving at a first estima-
tion of direct surface runoff.
                                                                                                       145

-------
             1.2 -
         tn
         
c
      0
H-

o 0.35
c

ir 0.30


   0.25


   0.20


   0.15


   0.10


   0.05
               " I    3   5    7   9   II   13   15   17   19  21   23  25  27  29
               MAR  MAR APR  MAY   JUN   JUL  AUG  SEPT  OCT   NOV  DEC   JAN  JAN
                 I    28   25   23   20   18   15   12   10   7    5    2    3O

                                             Period


       Figure 12.-Comparison of simulated and observed runoff records by 14-day periods. Coshocton, Ohio meadow.
                       Mean  Runoff per  Period
146

-------
c
o

o
c
=j
U-

c.
o

"5
.a
CO

O


CD
 3

 £
 3
O
                                            Coshocton  Ohio,  Meadow
                                              o   Observed

                                              x   Simulated

                                             	Period I, Mar  1-14

                                             	Period 11, July 18- Aug 2
                        I    I    I    1   I    I
                                                           II   I
0      0.2     0.4     0.6    0.8     1.0      1.2     1.4     1.6     1.8

                         x   Runoff ,  inches
                                                                                   2.0
           Figure 13.-Sample distribution functions foi simulated and observed records. Ooshocton, Ohio meadow.

-------
                                        LITERATURE CITED
1.  Austin, M. E.  1965. Land resource regions and major
   land resource areas of the United States. SCS USDA
   Agriculture Handbook No. 296,82 p.

2. Kincaid, D. C. and Heermann, D. F. 1974. Scheduling
   irrigations using a programmable calculator.  ARS
   NC-12.5Sp.
3. Unsley, R. K., Jr., Kohler, M. A. and Paulhus, J. L.
   H. 1958. Hydrology for Engineers. McGraw-Hill, N.
   Y.

4. Soil  Conservation Service, USDA. 1971.  SCS Na-
   tional Engineering Handbook, Section 4, Hydrology.
   U.S. Govt. Printing Off., Washington, D.C.

5. USDA Statistical Reporting Service.  1972. Usual
   planting and  harvesting dates. Agr. Handbook No.
   283.
6. U. S. Geological Survey. 1970. The National Atlas of
   the United States of America. 417 p.

7. U. S. Dept.  of Agriculture. Hydrologic Data  for
   Agricultural Watersheds in the United States. USDA
   Misc. Publications  No. 945 (1956-59); 994(60-61);
   1070(1962); 1164(1963);  1194(1964);  1216(1965);
   1226(1966); 1262(1967).

8. U.S.  Dept. of Agriculture. 1957. Monthly precipita-
tion and runoff for small agricultural watersheds in the
United States. USDA ARS, Soil and Water Conservation
Research Branch, Washington, D.C. (June).
 148

-------
                                            APPENDIX B

           SIMULATION OF POTENTIAL PERCOLATION AND NITRATE LEACHING

                                          INTRODUCTION
  Potential  percolation was  defined in Section 3.4,
Volume I, as the annual amount of water that would
percolate below the root zone in a field of straight-row
corn.
  The relationship between  potential percolation and
other  hydrologic  variables  can be expressed by  the
vertical water balance equation for a column of soil as
     P  =
(1)
where P is the precipitation, Qs is the direct runoff, E is
the evapotranspiration, Qp is the percolation from the
bottom  of the root zone and AS is the change in soil
water storage in the root  zone during the time under
consideration. In Eq. (7),  it  is assumed that imported
water, lateral porous media flow, and change in surface
detention storage are negligible.
   Soluble agricultural chemicals that are not strongly
adsorbed by the soil may be carried below the root zone
by percolating water. After the  percolating  water has
reached the  ground water table, it will move laterally
and eventually reappear in a stream, lake or possibly the
ocean.  Some ground water may also flow into the root
zone in seepage areas and be transpired by vegetation.
   Because precipitation can be considered as a stochas-
tic process,  all of the other variables in Eq. (7) are
stochastic in nature. A simulation approach is necessary
to estimate each of the other terms because of the rather
complex  relationships between them. To carry out the
simulation we  need a soil water model, a direct runoff
model  and  an  evapotranspiration  model.  The SCS
procedure as described in Appendix A was used as the
direct  runoff  model. The  soil-water  model  and the
evapotranspiration model  are  described in this Appen-
dix.
                                     THE SOIL-WATER MODEL
   The soil-water model described in this section utilizes
the approximation that soil-water moves readily under
gravitational forces when its water content is above field
capacity. It is further assumed that water does not move
downward  when   the  water content is below field
capacity and that when the water content reaches the
wilting point it is no longer available to plants.
   The structure  of  the  three-compartment  soil-water
model  used  is shown in  Fig. 1. The  water content of
compartment i is  designated as Sj(t) and the maximum
capacity  of the i*h  compartment is  Kj,  which corre-
sponds to field capacity.  Compartments 1 and 2 repre-
sent the active root zone and compartment 3 represents
'he water storage below the current root zone and above
the maximum  depth  of rooting, d3- The  depth  of the
      surface layer is dj and d2(t) is the time varying depth of
      the root zone. The extraction of water from different
      depth zones varies with stage of crop development, so
      the capacities  of compartments 2 and 3 vary with time
      as the crop canopy expands but their sum is a constant.
      The maximum capacities of the compartments can be
      interpreted as the available water-holding capacity per
      unit area for the depths dj, d2(t)-di and d3-d2(t).
         Input to the system is Xj(t), which is the difference
      between daily rainfall plus snowmelt and direct runoff.
      System output is Yj(t), ¥2(1) and ¥3(1) where
          Y! (t) = daily evaporation from the soil surface
          Y2 (t) = daily transpiration
          Y3 (t) = dafly seepage below the maximum root zone
                                                                                                    149

-------
   The  evaporation  and  transpiration model will be
described in the next section. The rules for movement of
water between compartments are as follows:
        y,2(t+l) = S,(t) + X1(t+l)-Y1(t+l)-Y4(t+l)-K1; otherwise
                                                                                                      (2)
where yj2(t+l) is the flow from compartment 1 to com-    transpiration and flow to compartment 2. Flow from
partment 2.                                            compartment  1  to compartment 2 exists only if the
   Equation (2) states that the water content  of com-    available waterholding capacity  of compartment 1 is
partment 1 in period t+1 is equal to the water content in    exceeded. No provision is made for upward flow in this
period t plus the infiltration on day t+1 less evaporation,    model, although such flow is physically possible.
                         y12(t+l)-Ys(t+l)-y23(t+l)

                         t) + y12(t+l)-Y5(t+l)
-------
   Transpiration  Evaporation

         Y2(t)        Y,(t)
Infiltration

   X,(t)
1 •
*

Y4(t)
Y5(t)
(
i
1 S,(t)
\
2 Q
2
y,2(t)
t)
                                                           K,

                                                           K,(t)
                                                                                     d2(t)
                                                                                   r
3
S3(t)
                                                           K,(t)
                                                 Y3(t)
                                         Percolation

                                   Figure 1.-Soil-water model schematic.
crude concept of "field capacity." The approximations
and assumptions included in the model probably would
not lead to serious errors in deep, well-drained soils.
         Substantial errors may occur where water tables are
         shallow or in soils with shallow, relatively impermeable
         layers.
                             THE EVAPOTRANSPIRATION MODEL
  The evapotranspiration (ET) model is based on the
i -quently used assumption that ET will take place at
the "potential" rate if the soil has adequate water and a
complete crop canopy or if the surface  is wet. Actual
evapotranspiration rates will be less than potential as the
soil dries.
            The model can be described by the following equa-
         tions:
            Evaporation:
         Y1(t)=[l-C(t)]KpP(t)  SiW
(9)
                                                                                                 151

-------
                       O)
                       0)
                       o
                       c
                       0)
                       en
                       ai

                       LJ
       d
       o
       c
       o
      O
                                                             i
o
Q
to
0)
                             o
                             I
                             Time
                                   Figure 2.—Seasonal variation of root zone.
where  C(t) is  a time varying crop coefficient related to
the portion of the soil surface covered by vegetation, Kp
is a coefficient to convert pan evaporation to potential
ET, P(t) is the pan evaporation in inches per day and
Sj(t) and KI have been previously defined.
   Total transpiration is given by:
Y2(t) = C(t)KpP(t)f .                          (10)

where  f is the ratio of actual to potential evapotranspira-
tion and depends on the water  content of compartments
1 and 2 as shown in Fig. 3.
   The transpiration loss from compartment 1 is:
Y4(t)-Y2(t)
                  s,(t)
                   (ID
and the transpiration loss from compartment 2 is:
Y5(t) = Y2(t)-Y4(t)  .                          (12)

   An examination of Eqs. (9) through (12) reveals that
for bare fallow conditions, [C (t) = 0], only evaporation
occurs, i.e., Y2(t) = 0. For full canopy conditions C(t) =
l,Yj(t) = 0 and all loss is through transpiration. The
form of this evapotranspiration model is identical to that
used by Hanson (5).
                                    NITRATE LEACHING MODEL
   The purpose of this model is to gain a quantitative
insight  into  what percentage  of  nitrogen  applied  as
ammonium in the fall or spring would move below the
root  zone before the  roots had  reached their full
extension in the following crop year. Any nitrogen
present  in  the profile at  the  time of fertilization is
ignored  as are denitrification losses. It is also assumed
that  there is no nutrient uptake by  weeds or winter
152

-------
          0
                                                          0.6
.0
                                S2(t)]/[K,(t)-hK2(t)]
Figure 3.—Relationship between actual and potential evapotranspiration and root zone water content.
cover  crops. Water flow through the soil  during the
dormant period is assumed to be piston flow. The upper
compartment with capacity Kt was retained in the soil
model. At the  time ammonium was applied in the fall,
the water in compartments 2 and 3 was assumed  to be
uniformly distributed in the depth d3 - d!. Any time the
capacity of compartment 1 is exceeded, the volume of
water yu(t) moves as a piston flow, displacing some of
the soil water ahead of it as shown in Fig. 4.
   The depth increment of the i*h output from compart-
ment 1 to the lower zones is given by the expression



where  E is an exclusion factor to  account for the
fraction of the  soil water not containing nitrate and 0pis
the volumetric  field capacity (21,22), E as used in these
calculations is  the ratio  of the  volume  of soil water
excluding nitrate  to  the  total  soil  volume. Each time
                                         Xi2(t) is nonzero,  a  new  increment  is introduced,
                                         pushing all those ahead of it down a distance AZ.
                                            A  number of studies  (9, 12, 28) have shown that
                                         anions, like nitrate, move with the wetting front through
                                         dry soil. The depth to the peak concentration can often
                                         be estimated by the ratio of infiltration to field capacity.
                                         Laboratory leaching studies (21, 22) with water contents
                                         near  saturation show that these  anions are excluded
                                         from  some of the soil water and thus move faster than
                                         the total  soil  water does. Opinions differ  as  to the
                                         importance  of  this  factor for field  conditions. Some
                                         consider it important (23), others do not (1). We have
                                         chosen to incorporate an exclusion  factor because a large
                                         part of the nitrate movement being modeled will occur
                                         at moisture contents  above field capacity and it will also
                                         reflect some of the channelized flow in clay soils (8).

                                            When ammonium  fertilizer is applied, it is assumed to
                                         be concentrated at depth, di. Ammonium is converted
                                                                                        153

-------
         E(d3-d1)

                          0
           e
                                     Figure 4.—Piston flow soil water model.
to the nitrate according to the following temperature-
dependent relation after a lag of 5 days:
N(t+l)=N(t) + k(T)A(t)
(14)
where N(t) is the nitrogen in the nitrate form on day (t),
A(t) is the nitrogen in the ammonium form on day (t),
and k(T) is a temperature-dependent rate function given
by the following equations (4, 19):
  = 0.0032T-.012  ;  10°C
-------
        o
       O
March  I
Q)

O
Q



O
                                                                                     0)


                                                                                    Q
                                                                                      -
                                                                                     o
                                                                                    I
 a
LL!
                   O
                                                                                                            i   /
                                                                                                            K
                                                                                                                  o
                                                                                                                  0)
                                                                                                              Wo
                    Hfcu
                                                       Figure 5.-Nitrate leaching model operation in time.

-------
zone extension, all Wj from NR-1 that are below the root
zone are  summed  and  divided by  the  ammonium
nitrogen  present  at  spring  fertilizations  to give  the
percentage loss. Note that for the case portrayed in Fig.
5,  no nitrate  would be lost from  spring fertilization
because  the first increment  of recharge  after spring
 fertilization was not below the root zone. In the piston
 flow model for water movement and nitrate leaching it is
 assumed  that  the  nitrate  is completely mixed within
 each element  AZ,  but that there is no  diffusion  or
 dispersion allowing nitrate exchange between elements.
                                      SIMULATION PROCEDURE
                        Data
   Daily precipitation and temperature data used in the
simulations were obtained on magnetic  tape from the
National Climatic Center, Environmental Data  Service,
NOAA, U.S. Dept. of Commerce, Asheville, N. C. The
data set obtained is termed Day Deck 345. The normal
period  of  record  was  from January  1948  through
December  1973. A year beginning on March 1 was used
in all simulations. The stations used are listed in  Table 5,
Appendix A. Simulations were limited to stations east of
the Rockies because of the steep rainfall gradients in the
West and because most of the situations where  leaching
may be a  problem are in irrigated  areas and  thus are
excluded from this report.
   Mean monthly pan evaporation data  were obtained
for the stations used or for nearby stations from a U. S.
Weather Bureau publication (25). Fourier series were fit
to  these monthly values  and were converted to  mean
daily values. A single harmonic explained more than 97%
of the variance for most of the stations. The mean daily
evaporation and the amplitude and  phase angle of the
first harmonic were plotted on maps and isolines were
drawn. These parameters were then estimated by inter-
polation from the maps for stations where evaporation
pan data were not available.  The pan coefficient, Kp,
was obtained  from  the map  presented  by  Kohler,
Nordenson and Baker  (10).
            Estimation of Parameters

   The index  crop considered was  straight-row corn.
Planting  and harvesting  dates  for each locality  were
obtained from maps prepared  by the USDA Statistical
Reporting Service (24). Plowing and  spring fertilization
were  arbitrarily  assumed to have  been done  14  days
before planting.  The fall fertilization date was the day
the 5-day moving average temperature went below 50° F
(10° C) or December 15, whichever occurred first. Corn
was assumed to reach full canopy 80 days after planting.
Root zone depths, available soil water capacities, field
 capacities, and exclusion fractions most commonly used
 in the simulations are shown in Table 1.
  Table l.-Most commonly used soil-water model parameters
   Parameter
                        Hydro logic soil group
                           B
D
d, 	
d,1 	
Ki ....
KI +K2 "*" 1^3 . .
0F . ...
E 	
4 in.
4 ft.
.33 in.
4.0 in.
.123
.04
4 in.
4 ft
.67 in.
8.0 in.
.237
.07
4 in.
4ft.
.67 in.
8.0 in.
.327
.10
4 in.
4 ft
.50 in.
6.0 in.
.345
.15
    1 Root zone depths were reduced for shallow soils.
   It  was  assumed  that the  available  water-holding
capacity of a soil was the difference between the water
content at  0.3  bars and  15 bars tension. (Approximate
field capacity and wilting point). Typical  textures of
soils in each hydrologjc soil group were then selected
and the total available water  content was rounded to the
nearest inch. The storage capacities were assigned to land
resource areas on the basis of the characteristics of the
predominant  agricultural soils. The  assignments were
reviewed and corrected where necessary by soil scientists
of the  SCS Technical Service  Centers. The  assignments
by land resource areas are shown in Table 6, Appendix
A.  The  values of the  .exclusion fraction, E  , were
estimated for the assumed water-holding capacities from
published data on 15 soils (21).

               Computer  Program

   The  subroutine. ETRANS (Evapotranspiration) was
written  in  FORTRAN  IV.  It is called from program
SCSRO described  in Appendix A. A generalized flow
chart is shown in Fig. 6.
156

-------
           Read
           Field
           Char ac t er i s t ±c s
           Read evaporation
           and crop coeffi.
           series parameters
REPEAT FOR DESIRED NUMBER OF YEARS


Call CN


Call TPREAD
*
Call SNOW
i
Call AMCF
\
Call SRO
*
Call STEMP
                                For explanation of these sub
                              >  routines,  see Fig.  2,
                                Appendix A.
          Call  ETRANS
Subroutine  STEMP computes a
5^day moving average temperature
for use in  the nitrification
calculations.

Subroutine  ETRANS computes evapo-
ration, transpiration,  percolation
nitrification  and nitrate move-
ment on a daily basis  for 1 year.
          Compute rainfall,
          runoff,  percolation
          evapotranspiration
          and N  loss
          statistics.
         Write
         Output
         Summary
Figure 6.-Generalized flow chart Program SCSRO with percolation and nitrate leaching option.
                                                                    157

-------
                                        SIMULATION RESULTS

   The program output for each 14-day period included:       5. Distribution functions of loss of fall-applied nitro-
   1.  An ordered listing of daily rainfall.                  gen, spring-applied nitrogen and leaching.
   2.  An ordered listing of simulated runoff.                 6. Mean annual percolation.
   3.  An ordered listing of daily percolation.                 7. Mean annual evapotranspiration.
   4.  The mean and standard deviation of each of the       Maps of the mean annual percolation, fall-applied N
above.                                                 loss and spring-applied N loss for each of four available
   The  statistical summary for the n-year simulation    soil water-holding  capacities  are shown  in  Figs.  7
included:                                               through 18. Because of the limited area in which soils of
   1.  A table  showing the number of runoff events for    -Hydrologic Group A are predominant, simulations were
each 14-day period for each year.                         not performed for all stations. Therefore, isolines could
   2.  The probability that there would be no runoff in    not be drawn  in the east central portion of the United
any year for each 14-day period.                          States.  The mean annual precipitation for the period of
   3.  The mean annual simulated runoff.                  record used in the simulation is shown in Fig. 19.
   4.  The mean growing season simulated runoff.
 158

-------
               Figure 7.-Mean annual percolation below a 4-foot root zone in inches. Hydrologic Soil Group A. Four inches available water-holding capacity. Straight-row com.
ct
\o

-------
                                                                                                                                     12.0
                                                                                                                                    14.0
                                                                                                                                  O
Figure 8.-Mean annual percolation below a 4-foot root zone in inches. Hydrologic Soil Group B. Eight inches available water-holding capacity. Straight-row corn.

-------
Figure 9.—Mean annual percolation below a 4-foot root zone in inches. Hydrologic Soil Group Q Eight inches available water-holding capacity. Straight-row corn.

-------
to
                Figure lO.-Mean annual percolation below a 4-foot root zone in inches. Hydro-logic Soil Group D. Six inches available water-holding capacity. Straight-row corn.

-------
                                                                                                                                                                50
                                                                                                                                                        0,0
      Figure 11.-Mean percentage loss of fall-applied nitrogen. Straight-row com in good hydrologic condition. Hydrologic Soil Group A. Available water-holding capacity - 4 inches.
Ui

-------
Figure 12.-Mean percentage loss of fall-applied nitrogen. Straight-row corn in good hydrologic condition. Hydrologic Soil Group B. Available water-holding capacity - 8 inches.

-------
             Figure 13.- Mean percentage loss of fall-applied nitrogen. Straight-row com in good hydrologic condition. Hydiologic Soil Group C Available water-holding capacity - 8 inches.
o\

-------
                                                                                                                                                30
Figure 14.-Mean percentage loss of fall-applied nitrogen. Straight-row coin in good hydrologjc condition. Hydtologic Soil Group D. Available water-holding capacity - 6 inches.

-------
Figure 15.-Mean percentage loss of spring-applied nitrogen. Straight-tow com in good hydrologic condition. Hydrologic Soil Group A. Available water-holding capacity - 4 inches.

-------
Rgure 16.-Mean percentage loss of spring-applied nitrogen. Straight-row corn in good hydrologic condition. Hydrologic Soil Group B. Available water-holding capacity - 8 inches.

-------
         v_
--)
      ^
                         I 0*0
                                         I
                                           0*0
                                                  0-0
                         x—-<	
                                         	)
                           T
                         0«0
                        (-__.
                        !o«o
                                                o-o !
                                                     i
                                                  0.0	
                                                  ^
      0*0,

0«0-'—t—
^x
\
V
/ T """""T-J
/ ' i
\J—— / 	 i±!
{ / T— i
/ / io«o
-/ I i L-
\ /
D.O°^___.V^(
0'° °Co c
°'° 'i °'° ?
-- — -,{, ,'
i 	 1
O«O 1 \
1 1
^/-^
£^~ /'
)io 	 -,-- A
">V
                                                                                  0-0
                                                           /0«0

                                                      0«0.
                                                     I      i     °
                                                     6jfiuT*i.r26*
                                                                                   •^
                                                                                   '/
Figure 17.-Mean percentage loss of spring-applied nitrogen. Straight-row corn in good hydrologic condition. Hydrologjc Soil Group C Available water-holding capacity - 8 inches.

-------
                                                                                                                           •o   yv-oj
                                                                                                                      °n>^
                                                                                                                          °'°0-0
                                                                                                                                 /
                                                                                                                              0»0
Figure 18.-Mean percentage loss of spring-applied nitrogen. Straight-row corn in good hydrologic condition. Hydrologic Soil Group D. Available water-holding capacity - 6 inches.

-------
Figure 19.-Mean annual precipitation for period of record used in simulations.

-------
                                               DISCUSSION
   Simulated mean annual percolation and nitrate leach-
ing cannot be compared with observations because such
data are not generally available. However, the excellent
correlation between  the sum of the simulated potential
direct  runoff and percolation and the surface runoff
from USGS maps as presented in Appendix A suggests
that the simulated results are reasonable.
   Isolated  bits  of  data  are  available for additional
checks. Minshall  (14)  in a study of 25 years of runoff
data (1940-1964) on  the Platte  River  in southwestern
Wisconsin found  that the mean annual base flow was 5.7
inches.  The  mean  annual  potential  percolation  for
Hydrologic Group B is somewhat greater than 5 inches
(Fig. 8).
   Hanway  and Laflen (6)  reported 3-year averages of
1.11 and 4.64 inches of subsurface  drainage  for  tile
outlet  terraces in Creston, Iowa and Charles City,  Iowa,
respectively. From Fig.  8,  the mean annual potential
percolation  for these -sites is about 2 inches and 3 inches,
respectively. One would anticipate greater percolation
losses  from tile outlet terraces than from straight row
corn but the period  of record is too short to make valid
comparisons.  Again  it appears that the simulated per-
colation is reasonable.
   Saxton, Spomer and Kramer (20) reported on  meas-
urements of base flow for small watersheds with  contour
corn near Treynor,  Iowa.  Six-year average base flow
from two watersheds  was  2.52 and 2.47 inches.  From
Fig. 8, simulated deep percolation in this area is about 2
inches.  Rainfall  during  the 6-year period was above
average for 5 of the 6 years.
   Simulated mean annual percolation is compared with
lysimeter data in Table 2. Only one set of data is for
corn (Coshocton, Ohio), and the simulated percolation is
very close  to the observed. The other data sets agree
favorably with the simulated percolation when the crop
canopy differences are considered.  The shallow  lysime-
ters at Windsor, Conn, probably account for much of the
difference between observed and computed percolation.
   We  were  unable  to  find  any data  showing  the
percentage  of fall-applied nitrogen lost during the winter
and spring.
   Although  the comparisons  between simulated per-
colation and data cannot be considered as conclusive,
they do suggest that the simulations provide a reasonable
ordering of  Land  Resource  Areas  with  respect  to
percolation losses. The absolute amounts also appear to
be realistic.
   The only way a technique such as this can be judged
is  against  readily available  alternatives.  The leaching
hazard  map  prepared by  Nelson  and  Uhland  (18)  is
shown  in  Fig. 20.  The  material presented   in  this
Appendix and in Vol. I clearly  presents a more detailed
picture  of percolation and of the relative hazards of
nitrate leaching from fall fertilization.
   Care should  be  used  in interpreting the maps of
potential percolation and nitrate  leaching  where it  is
known that the model assumptions are seriously in error.
For example, in the Southern United States the assump-
tion of no nutrient uptake or  transpiration  during the
winter would be inaccurate if winter  cover crops are
planted. In  this case,  both the  percolation and the
nitrate loss would be overestimated.
                        Table 2.-Comparison of simulated mean annual percolation with lysimetei data
Location Citation Soil Sf&Sp
Ithaca, N.Y. Bizzell (2) Pet oskey gritty A
sandy loam B
Average percolation
Crop Simulated
Observed (com)
inches
Vegetables 17.76
17.76
inches
13
11
  Geneva, N.Y.         Collison et al. (3)   Ontario, Dunkirk
  Knoxville, Tenn.      Mooers et al. (15)   Cumberland
Windsor, Conn.
Windsor, Conn.
Coshocton, Ohio
Morgan &
Jacobson(16)
Morgan, et al.
(11)
Harrold &
Dreibelbis (7)
Merrimac,
20" depth
Merrimac,
30" depth
Muskingum
   B

   A


   A


   C
Barley-clover         12.7       11
  rotation

Fallow              22.5       15

Tobacco            13.63      20
                                                                      Tobacco with         12.45       20
                                                                        winter cover

                                                                      Corn years in          7.43        7
                                                                        CWMM rotation
 172

-------
Figure 20.—Relation of degree of leaching to geographic area. Leaching ranges from nil in Area I to very high in Area IV.
                                         From Nelson and Uhland (18).
                                                                                                               173

-------
                                        LITERATURE CITED
1.  Appelt, H., Holtzclaw,  K., and Pratt, P. F. 1975.
   Effect  of anion exclusion on movement of chloride
   through soils. Soil Sci. Soc. Amer. Proc. 39:264-267.

2.  Bizzell, J. A. 1943. Lysimeter experiments-VI:  The
   effects of cropping and  fertilization on the losses of
   nitrogen from the soil. Cornell Univ. Agr. Exp. Sta.
   Memo. No. 252: 1-24.

3.  Collison, R.  C., Beattie,  H.  C., and  Harlan, J. D.
   1933.  Lysimeter investigations:  III.  Mineral  and
   water relations and  final nitrogen balance in legume
   and  nonlegume crop  rotations for  a  period of 16
   years. New York (Geneva) Agr. Exp. Sta. Tech. Bull.
   212:1-81.

4.  Frederick,  L. R.  1956.  Formation  of nitrate from
   ammonium nitrogen in soils: 2. Effect of population
   of nitrifiers. Soil Sci. 83:481-485.

5.  Hanson, C. L. 1973. Model for predicting evapotrans-
   piration from native rangelands in the Northern Great
   Plains.   Ph.D.  dissertation, Utah State  University,
   Logan.

6.  Hanway, J. J. and Laflen, J. M. 1974.  Plant nutrient
   losses from tile-outlet terraces. Jour. Environ. Quality
   3(4): 351-356.

7.  Harrold, L. L. and Dreibelbis, F. R. 1958. Evaluation
   of agricultural  hydrology by  monolith lysimeters
   1944-55. U. S. Dept. Agr. Tech. Bui. 1179,166 p.

8.  Kissel,  D. E., Ritchie, J. T. and Burnett, E. 1973.
   Chloride movement in undisturbed swelling clay soils.
   Soil Sci. Soc. Amer. Proc. 37:21-24.

9.  Kolenbrander, G. J.  1970. Calculation  of parameters
   for the evaluation of the leaching of salts under field
   conditions, illustrated  by  nitrate.  Plant  and  Soil
   32:439-453.

10. Kohler, M. A., Nordenson, T. J., and Baker, D. R.
    1959. Evaporation maps for the United States. U. S.
    Weather Bureau Tech. Paper No. 37, 13 p.
11.  Kohler, M. A., Nordenson, T. J., and Baker, D. R.
    1963. Rainfall-runoff models. General Assembly of
    Int. Assoc. of Scientific Hydrology Surface Waters,
    Berkeley, Calif.: 479-491.

12.  Levin, I.  1964. Movement of added nitrates through
    soil columns and undisturbed soil profiles. Proc. 8th
    Intern.  Congress  of  Soil  Science,  Bucharest,
    Romania.

13.  Makkink,  G.  F.  and van Heemst, H.  D. J. 1966.
    Water  balance  and  water  bookkeeping  regions.
    Verslagen  en  mededlingen van de Commissie voor
    Hydrologisch  Orderzock T. N.  O. No. 12, 9-112.
    Cited by  F. E.  Schulze, Chapter 1, Rainfall and
    rainfall  excess  in recent  trends in  hydrograph
    synthesis.  Proc. of Tech.  Meeting 21. Versl. Medel
    Comm. Hydrol. Onderz. T.N.O. 12. The Hague. 103
    P-

14. Minshall,  N. E.  1967. Precipitation  and base flow
    variability. Proc. International Assoc.  of Scientific
    Hydrology, Bern: 137-145.

15. Mooers, C. A. Maclntire, W. H., and Young, J. B.
    1927. The recovery of soil nitrogen under various
    conditions as measured  by lysimeters of different
    depths. Tennessee Agr. Exp. Sta. Bui. 138: 1-30.

16. Morgan,  M. F. and Jacobson, H. G. M.  1942. Soil
    and crop interrelations of various nitrogenous fertil-
    izers: Windsor lysimeter Series B. Connecticut Agr.
    Exp. Sta. Bui. 458: 269-328.

17. Morgan, M. F., Jacobson, H. G. M.,  and  Lecompte,
    S. B. Jr.  1942. Drainage  water losses from  a sandy
    sofl  as  affected by  cropping and cover  crops:
    Windsor  lysimeter Series  C.  Connecticut Agr. Exp.
    Sta. Bui. 466: 729-759.

18. Nelson, L. B. and Uhland, R. E. 1955. Factors that
    influence  loss of fall applied  fertilizers and their
    probable  importance  in  different sections of the
    United   States.   Soil   Sci.  Soc.  Amer.   Proc.
    19(4):492-496.
174

-------
19. Sabey, B.  R., Bartholomew, W. V., Shaw, R., and
    Pesek, J. 1956. Influence of temperature on nitrifi-
    cation   in  soils.   Soil  Sci.   Soc.  Amer.  Proc.
    20:357-360.

20. Saxton, K. E., Spomer, R. G., and Kramer, L. A.
    1971. Hydrology and erosion of loessial watersheds.
    Proc. ASCE Jour. Hydr. Div. 97 (HY11): 1835-1852.

21. Smith, S.  J. 1972. Relative rate of chloride move-
    ment   in  leaching  of  surface soils.  Soil  Sci.
    114(4):259-263.

22. Smith,  S.  J., and Davis,  R.  J.  1974.  Relative
    movement  of bromide and nitrate through soils.
    Jour. Environ. Qual. 3(2): 152-155.

23. Thomas, G.  W.  1970. Soil and  climatic  factors
    which affect nutrient mobility. In Engelstad, O. P.
    ed., Nutrient mobility in soils: accumulation  and
    losses, Soil Sci. Soc. Amer., Inc., Madison, Wise.
24. USDA  Statistical  Reporting Service. 1972.  Usual
    planting and harvesting dates. Agr. Handbook No.
    283.

25. U. S. Weather Bureau. Climatography of the United
    States.  Climatic summary  of the United States-
    Supplement for 1951 through 1960.

26. Van Wijk, W. F., Larson, W. E. and Burrows, W. C.
    1959. Soil temperature and the early growth of corn
    from mulched  and unmulched soil. Soil Sci. Soc.
    Amer. Proc. 23(6): 428-434.

27. Wiser, E. H. and van Schilfgaarde, J. 1964. Predic-
    tion of irrigation water requirements using a  water
    balance. Proc. VI^1 International Congress of Agr.
    Engr., Lausanne: 121-129.

28. Yaalon,  D.  H.  1965. Downward movement and
    distribution of anions in soil profiles with limited
    wetting. In  Experimental Pedology. William Cloves
    and Sons, Ltd. London: 157-164.
                                                                                                     175

-------
                                              APPENDIX C
                               ECONOMIC ANALYSIS METHODOLOGY
   The following discussion details the application of the
method presented in Section 5, Volume I, to  evaluate
the decision-maker's optimal choice in the example given
in  Section  6.2,  Volume  I. This  example is clearly
site-specific and cannot hope to show the full gamut of
variables which may potentially be of significance in other
situations,  such as irrigation,  hired labor, other  crop
rotations and  the  like. The decision-maker will have to
adjust the budgeting system shown here to his particular
situation.
   Several important assumptions were made.  For cer-
tain computations,  the  size  of the  farm became  a
parameter,  and the  assumption was  made that the
example farm  had 250 tillable acres. Other assumptions
are detailed in the following tables. In addition, it was
assumed that  none  of the  macro effects  described in
Section 5.2 of Volume I influence any of the  decision
variables noted here. This assumption implied that any
machinery which may become obsolete due to a change
in cropping practices would be  sold at a cost close to its
depreciated value and that, consequently, there  was
no cost of disposing of obsolete equipment to be added
to the actual machinery costs.
   The determination of the relevant production alterna-
tives  resulted  in  five potential choices,  namely, (1)
continuous  corn  no-till planted in 70 percent residue
cover,  contoured, (2) a corn-corn-corn-wheat-meadow
rotation with  moldboard plowing on the first year corn
and no-till planting on the  second and third year corn,
contoured, (3) continuous corn with rotary strip tillage,
terraced,  (4)   continuous   corn  with  chisel  planting,
terraced, and  (5) a corn-soybean rotation with no-till
planting,  terraced. The  costs  and  returns for a  sixth
production  method  (i.e. continuous corn, residue  left,
with moldboard  plowing,  straight row) are shown for
comparison purposes only. This particular production
method does  not meet the soil erosion limitation and
can therefore  not be considered as an available alterna-
tive.
   Three of the five  viable alternatives required terrac-
ing, which  contributed an  additional production  cost,
summarized in Table 1. The example assumed a  farm
 Table 1. Broadbase terrace construction and maintenance costs
                Item
Amount
 Terrace spacing, feet   	     120
 Slope length, feet  	     350
 Number of terraces per slope ......       2
 Feet terrace/acre	         249
 Construction cost/foot terrace3, $	       0.60
 Construction cost/acre, $  	     149.40
 Prorated construction costb, $	      13.74
 Maintenance cost, foot3. $	       0.00023
 Maintenance cost, acre, $	       0.06
 Yearly terrace charge/acre, S	      13.80
 Total yearly terrace charge (250 acres), $.   3.450.00

   a Source: Sidney James (ed), Midwest Farm Planning
 Manual, 3rd Edition, 1SU Press, Ames, Iowa, 1973, p. 33.
   b Assume 20 year life of terrace. Interest at 8 percent.
located  on Monona silt loam with more than 3 percent
organic  matter, a land slope of 6 percent, and an average
slope  length of 350 feet.  According  to  the  technical
standards1 for terrace construction, the construction of
level broadbase terraces with a spacing of 120 feet would
be  appropriate in this situation. Table  1  shows the
assumptions  used  in  the  computation of the cost of
terracing the entire farm.
   Each of these production systems requires a specific
set of field operations and implements. Table 2 lists the
implements  considered in this study and  the computa-
tion of  the fixed costs for each machine. The computa-
tion of the depreciation  cost  used  the straight-line
method over the economic life of the implement. Not all
of  the  implements were  used  in any particular crop
production activity;  Table 3  shows which  implements
were used in each production alternative, the total hours
of machinery use, and the total implement cost. These
costs  did not  include the  cost of the tractor (listed
   1 U.S. Department of Agriculture, Soil Conservation Service,
Iowa, Technical Standards and Specifications for Conservation
Practices,  Section 4A-Cropland, Work  Unit Technical Guide,
Code No. 600 and 602, January 1973.
                                                                                                        177

-------
00
                                                                      Table 2. Machinery fixed costs
                   Machine
Size
Initial3
 cost
Salvageb
 value
Economic0
  life
   Yearly
depreciation
Taxes, insurance^
  and housing
Interest6
 Yearly
fixed cost
                                                             Dollars        Percent          Years         Dollars          Dollars          Dollars        Dollars
       Stalk shredder   	   12'flail                2,350            13.7            12            169.00           70.50         101.83         341.33
       Moldboard plow	   5-16"                 2,590            17.7            10            213.16           77.70         113.96         404.82
       Chisel plow	   15'                    1,700            13.7            12            122.26           51.00          73.67         246.93
       Disk, tandem	   20'                    4,385            17.7            10            360.89          131.55         192.94         685.38
       Harrow	   20'                      340            17.7            12             23.32           10.20          14.73          48.25
       Sprayer   	   tractor mounted          680            17.7            10             55.96           20.40          29.92         106.28
       Banter - conventional  	   4-38"                 1,430            17.7            10            117.69           42.90          62.92         223.51
       Rotary strip planter	   4-38"                 3,675            17.7            10            203.45          110.25         161.70         574.40
       No-till plant (fluted coulters)  ...   4-38"                 4,375            17.7            10            360.06          131.25         192.50         683.81
       Wheat drill (with grass seeding
          attachments)	   12'                    2,740             9.7            14            176.73           82.20         117.43         376.36
       Cultivator	   4-38"                 1,470            17.7            10            120.98           44.10          64.68         229.76
       Duster	   4-row                   400            13.7            12             28.77           12.00          17.33          58.10
       Combine, self-prop	   small 70-80 hp        16,100            18.9            10          1,305.71          483.00         708.40       2,497.11
       00111 ho"1	   2-38"                 2,800            18.9            10            227.08           84.00         123.20         434.28
       Hatform	   13'                    2,500            18.9            10            202.75           75.00         110.00         387.75
       Hay mowers	   7'                       960            12.5            i2             70.00           28.80          41.60         140.40
       Hay conditioners   . . .-.	   7'                     1,300            12.5            12             94.79           39.00          56.33         190.12
       Hay rake	   side delivery              980            12.5            12             71.46           29.40          42.47         143.33
       Hay baler	      PTO	3,500	21.1	8	345.19	105.00	157.50	607.69

          » Source:  Background Information for Use with CROP-OPT System. FM 1628, ISU Cooperative Extension Service, Ames, Iowa, November 1974.
          0 Source:  George E. Ayres, Estimating Used Machinery Costs, A.E. 1078, ISU Cooperative Extension Service, Ames, Iowa, January 1974.
          <• Source:   Sidney James (ed), Midwest Farm Planning Manual, 3rd Edition, ISU Press, Ames, Iowa, 1973, Table IV-7, p. 129.
           Taxes and insurance at 2 percent of initial cost; housing at 1 percent of initial cost. Source:  George E. Ayres, Estimating New Machinery Costs, A.E. 1077, ISU Cooperative
       Extension Service, Ames, January 1974.
          e Assumed at 8 percent per annum.

-------
Table 3. Machinery costs
H(
Implement
)urs per
Acres of
useb
Times
over0
Total Repair cost per
hours 100 hours"
Corn, residue left, spring turn-plow, conventional
Stalk shredder 	
Moldboard plow 	
Sprayer 	
Disk 	
Harrow 	
Planter 	
Cultivator 	
Combine 	
Corn head 	
Total 	
Com, fall shred stalks, chisel
Stalk shredder 	
Chisel plow 	
Sprayer 	
Harrow 	
Planter 	
Cultivator 	
Combine 	
Corn head 	
Total 	
.18
.36
.21
.10
.10
.22
.21
.63
.63

250
250
250
250
250
250
250
250
250

1
1
1
1
1
1
2
1
1

45.0
90.0
52.5
25.0
25.0
55.0
105.0
157.5
157.5

Dollars
94.00
129.50
34.00
219.25
10.20
114.40
73.50
322.00
56.00

Total repair
cost
Dollars
42.30
116.55
17.85
54.81
2.55
62.92
77.18
507.15
88.20

Yearly fixed
cost
Dollars
341.33
404.82
106.28
685.38
48.25
223.51
229.76
2,497.11
434.28

Total
cost
Dollars
383.63
521.37
124.13
740.19
50.80
286.43
306.94
3,004.26
522.48
5,940.23
plant, 30-40% residue cover
.18
.17
.21
.10
.22
.21
.63
.63

Corn, residue left, strip-till row zones
Stalk shredder 	
Sprayer 	
Rotary strip-till planter . .
Cultivator 	


Total 	
Corn, fall shred, no-till plant.
Stalk shredder 	
Sprayer 	
No-till planter 	
Duster 	
Combine 	
Corn head 	
Total 	
.18
.21
.22
.21
.63
.63

50-70%
.18
.21
.22
.21
.63
.63

250
250
250
250
250
250
250
250

1
1
1
1
1
2
1
1

45.0
42.5
52.5
25.0
55.0
105.0
157.5
157.5

94.00
85.00
34.00
10.20
114.40
73.50
322.00
56.00

42.30
36.13
17.85
2.55
62.92
77.18
507.15
88.20

341.33
246.93
106.28
48.25
223.51
229.76
2,497.11
434.28

383.63
283.06
124.13
50.80
286.43
306.94
3,004.26
522.48
4,961.73
,40-50% residue cover
250
250
250
250
250
250

residue cover
250
250
250
250
250
250

Corn-corn-com-wheat-meadow. residue left, no-till
Stalk shredder 	
Moldboard plow 	
Sprayer 	
Disk 	
Harrow 	
No-till planter 	
Wheat drill 	
Duster 	
Combine wheat 	
Corn head 	
Platform 	
Hay mower 	

Hay rake 	
Hay baler 	
Total 	
.18
.36
.21
.10
.10
.22
.25
.21
.63
.30
.63
.30
.31
.31
.30
.63

150
50
150
100
50
150
50
150
150
50
150
50
50
50
50
50

1
1
1
2
1
1


1
1
1
1
1
1

olant 2nd
1
1
1
1
1
1
1
1
1
1
1
1
3
3
3
3

45.0
52.5
55.0
105.0
157.5
157.5


45.0
52.5
55.0
52.5
157.5
157.5

and 3rd corn
27.0
18.0
31.5
10.0
5.0
33.0
12.5
31.5
109.5
94.5
15.0
46.5
46.5
45.0
94.5

94.00
34.00
294.00
73.50
322.00
56.00


94.00
34.00
350.00
8.00
322.00
56.00


94.00
129.50
34.00
219.25
10.20
350.00
219.20
8.00
322.00
56.00
50.00
96.00
52.00
58.80
210.00

42.30
17.85
161.70
77.18
507.15
88.20


42.30
17.85
192.50
4.20
507.15
88.20


25.38
23.31
10.71
21.93
0.51
115.50
27.40
2.52
352.59
52.92
7.50
44.64
24.18
26.46
198.45

341.33
106.28
574.40
229.76
2,497.11
434.28


341.33
106.28
683.81
58.10
2,497.11
434.28


341.33
404.82
106.28
685.38
48.25
683.81
376.36
58.10
2,497.11
434.28
387.75
140.40
190.09
143.33
607.69

383.63
124.13
736.10
306.94
3,004.26
522.48
5,077.54

383.63
124.13
876.31
62.30
3,004.26
522.48
4,973.11

366.71
428.13
116.99
707.31
48.76
799.31
403.76
60.62
2,849.70
487.20
395.25
185.04
214.27
169.79
806.14
8,038.98
                                                                 179

-------
Table 3. (continued)
u
Implement
Corn-soybeans, no-till plant,
Stalk shredder 	
Sprayer 	
No-till planter 	
Duster 	
Combine corn 	
Combine soybeans ....
Corn head 	
Platform 	
Total 	

r«r
fall shred
.18
.21
.22
.21
.63
.30
.63
.30
Acres of
useb
corn stalks
125
250
250
125
125
125
125
125
Times
over0
1
1
1
1
1
1
1
1
Total Repair cost per
hours 100 hound
22.5
52.5
55.0
26.25
116.25
78.75
37.5
Dollars
94.00
34.00
350.00
8.00
322.00
56.00
50.00
Total repair
cost
Dollars
21.15
17,85
192.50
2.10
374,32
44.10
18.75
Yearly fixed Total
cost cost
Dollars
341.33
106.28
683.81
58.10
2,497.11
434.28
387.75
Dollars
362.48
124.13
876.31
60.20
2,871.43
478.38
406.50
5,179.43
   Source: Background information for use with CROP-OPT system, FM 1628, ISU Cooperative Extension Service, Ames, Iowa,
November 1974.
   ° Acres on which implement is used each year.
   c Number of trips through field with implement.
   d Computed as percentage of list price. Used 2% for combine, platform, corn head and duster; 4% for stalk shredder and hay condi-
tioner; 5% for moldboard plow, chisel plow, cultivator, sprayer, and disk; 6% for hay rake and hay baler; 8% for planters and wheat
drill; 10% for hay mower. Source:  George Ayres, Estimating new machinery costs, AE  1077, ISU Cooperative Extension Service,
Ames, Iowa, January 1974.
 180

-------
separately in Table 4) or fuel and lubrication (Table 5).
The implement hours  per acre  (from Table 3) were
aggregated for each production alternative. The total was
augmented by a 10 percent figure for traveling to field,
idling, etc., to result in  the total tractor hours figure
listed  in  Table 4.  The  depreciation  cost  assumed  a
straight-line depreciation  over the economic life of the
tractor.
   The fuel costs for the  tractor and the combine were
computed as shown in Table 5. These fuel  costs were
presented separately from  the  other machinery  and
tractor variable costs for the purpose of emphasizing the
differences  in fuel  consumption  among production
alternatives.
   Table 6 shows the computations for the seed costs of
the five production  alternatives. The assumption was
made that the no-till alternatives would be subject to  a
higher seed mortality rate than the other alternatives,
due to the higher crop residue levels.
   Agricultural chemicals  were selected on the basis of
the recommended  nutrient and pesticide practices. It
was assumed that the nitrogen was applied in NH3 form
and the phosphate and potassium in granular bulk form
(Table 7). The restrictions on optimal timing (N2) are
assumed to be met by fertilizer application just prior to
planting, which in the case  of two alternatives (corn
chisel-plant and corn rotary-strip-till) also implies incor-
poration (N8 and 12).
   The pesticide costs (Table 8) were estimated on the
basis  of pesticide recommendations by the ISU Exten-
sion  Service for  control  of the major  pests.2' 3  The
pesticide  costs for the several  production alternatives
may vary since each rotation requires the use of a unique
mix of pesticides. For example, the herbicide cost for
the no-till alternatives was higher than for conventional
tillage because greater amounts of and more expensive
types of herbicides were  assumed to  be  used with this
alternative. The insecticide cost for  the rotation includ-
ing meadow was assumed greater than for continuous
corn  alternatives due to the expected incidence of the
first-year corn insect complex.  No insecticide cost was
assumed for  soybeans, since the acreage of soybeans
ordinarily treated with insecticides was quite small.
   Labor costs for the five production alternatives were
computed as shown in Table 9. The labor requirement
per acre was  estimated  as 130  percent of the tractor
hour  requirement  to account for overhead labor in
addition to the direct requirements. The labor cost per
hour was assumed equal to the present average wage rate
for Iowa.
   Table 10 presents two additional cost components,
namely the corn drying costs and interest charges. It was
assumed that  the costs  (variable and fixed) of drying
corn amounted to 12 cents/bushel, which is the current
charge for  custom drying in Iowa.4 It was assumed that
the out-of-pocket costs  involved  the  use of borrowed
capital. The interest costs on machinery  and the tractor
were included in their total  costs and are not repeated
here.
   The gross revenue for each of the production alterna-
tives was computed as shown in  Table  11. The no-till
alternatives were assumed to have a slightly lower yield
than the more conventional tillage alternatives due  to
increased production and harvesting complexities.
   The  final table,  Table 12, summarizes all  of  the
preceding  computations  and shows the gross  revenue
and net return  figures for each  of the  six production
methods. A land  cost was included based on an assumed
land value  of $974.00 per acre5 and a cash rent of $7.40
per $100 value.6 Since this land  charge  applied equally
to all  six  production  methods, any  error  in this  land
charge will change  only the absolute levels and not the
differences in net returns among the six alternatives.
   It  appears  that the (unavailable) alternative of con-
tinuous corn with conventional moldboard tillage has a
significantly higher  net revenue than any of the other
(available)  production  alternatives. There is only a small
variation  in  net  return  of the  top  three (available)
production alternatives with a major net return drop to
the  corn-soybeans  alternative.   The   corn-corn-corn-
wheat-meadow rotation has by far the lowest net return
among  these six  alternatives, indicating that the savings
in fertilizer cost generated  by  the  nitrogen  nutrient
credit  from the  legume meadow are  not sufficient  to
offset the increases in other costs.
   2Harold J. Stockdale, Insect Pest Control Recommendations
for 1975, IC-328 (Rev.),  ISU Cooperative Extension  Service,
Ames, Iowa, January 1975.
   3Vivan M. Jennings, Weed Control Guide for 1975.  Pm 601
(Rev.), ISU Cooperative Extension Service, Ames, Iowa, January
1975.
   ^Estimated  1975 Iowa  Custom Rates,  ISU Cooperative
Extension Service, FM 1698, Ames, Iowa, January 1975.
   s$974.00 is the November 1, 1974 average price for high
grade farmland in West Central Iowa, reported in William Murray
etal, Land Values Double in 5 years, FM 1681, ISU Cooperative
Extension Service, Ames, Iowa, January 1975.
   6 A  rent of $7.40 per $100 value is the average cash rental
rate for corn and soybean land reported in E. G. Stoneberg and
Ronald Winterboer, Cash Rental Rates from Iowa Farm Land,
FM  1626  (Rev.),  ISU Cooperative Extension Service,  Ames,
Iowa, August 1973.
                                                                                                          181

-------
                                                   Table 4. Tractor costs
               Item
                                      Straight-row
                                        C conv.
                       Contour
                C no-t.
            CCCWM no-t.    C chisel
                                        Terraced
                            C strip
                           CB no-t.
Tractor hours per acre3	
Total tractor hours''	
Tractor initial cost,0 dollars .  .  .
Economic life, years^   	
Salvage value, percent6    ....
Yearly depreciation, dollars .  .  .
Taxes, insurance and housing/
   dollars   	
Average annual interest,^ dollars
Total fixed costs, dollars   . .  .  .
Repair costs,11 dollars   	
Total tractor costs, dollars (excl.
   fuel)   	
     2.21
   607.75
18,230.00
    11
    27.5
 1,201.52

   546.90
   795.49
 2,543.91
   886.34

 3,430.25
     1.65
  453.75
18,230.00
    13
    23.5
 1,072.52

  546.90
  785.29
 2,404.95
  661.75

 3,066.46
     2.20
   605.00
18,230.00
    11
    27.5
 1,201.52

   546.90
   795.49
 2,543.91
   882.33
     1.92
   528.00
18,230.00
    12
    25.5
 1,131.78

   546.90
   789.97
 2,468.65
   770.04
     1.65
   453.75
18,230.00
    13
    23.5
 1,072.76

   546.90
   785.29
 2,404.95
   661.75
     1.34
   368.50
18,230.00
    14
    21.5
 1,022.18

   546.90
   781.29
 2,350.36
   537.42
 3,426.24     3,238.69
               3,066.70     2,887.79
   a Assume tractor is required for harvest hauling, in amount equivalent to time requirements for combine. Add 0.2 hours per acre
for application of fertilizer with rented implements.
   b Increased by 10 percent for idling, travel to field, etc.
   ciOOPTOhpdiesel.
   d From Sidney James (ed.) Midwest Farm Planning Manual. 3rd Edition, ISU Press, Ames, Iowa, 1973, Table IV-7, p. 129.
   e From George E. Ayres, Estimating Used Machinery Costs, A.E. 1078, ISU Cooperative Extension Service, Ames, Iowa, January
1974.
   f Taxes and insurance at 2 percent and housing at 1 percent of initial cost. Source: George E. Ayres, Estimating New Machinery
Costs. AE 1077, ISU Cooperative Extension Service, Ames, Iowa, January 1974.
   S Assume 8 percent interest.
   h 0.8 percent of list price per 100 hours of use. Source: Ibid.
Table 5. Fuel costs

Item
Total tractor hours 	
Fuel cost per tractor hour,3 dollars .
Tractor fuel cost, dollars 	
Total combine hours 	
Fuel cost per combine hour, dollars
Combine fuel cost, dollars 	
Total fuel cost dollars 	

Straight-row
C conv.
607.75
2.071
1,258.65
157.50
1.106
174.20
1,432.85
Contour
C no-t.
453.75
2.071
939.72
157.50
1.106
174.20
1,113.92
CCCWM no-t.
605.00
2.071
1,252.96
109.5
1.106
121.11
1,374.07
C chisel
528.00
2.071
1,093.49
157.50
1.106
174.20
1,267.69
Terraced
C strip
453.75
2.071
939.72
157.50
1.106
174.20
1,113.92

CB no-t.
368.50
2.071
763.16
116.25
1.106
128.57
891.73
   a Fuel consumption gallons per hour = 0.044xPTO hp. Lubrication costs at 15 percent of fuel cost. Source: Sidney James (ed.)
Midwest Farm Planning Manual. 3rd Edition, ISU Press, Ames, Iowa 1973, p. 125. Assume diesel fuel at $0.40/gal.
   b Gasoline consumption = 2.35 gal./acre. Source: George E. Ayres, Fuel Required for Field Operations, AE 1079, ISU Cooperative
Extension Service, Ames, Iowa, March 1974. Lubrication costs at 15 percent of fuel costs. Assume gasoline at $0.40/gal.
182

-------
                                                 Table 6. Seed costs
               Item
                                    Straight-row
                                      C conv.
Corn
   Seeding rate (seeds/acre)
   Assumed mortality, %  .
   Final stand	
   Seed amount8, bu.  ...
   Seed cost", dollars  .  . .
Wheat
   Seed amount, bu.
   Seed costc, dollars
Hay
   Seed amount**, Ibs
   Seed '-oste, dollars
Soybeans
   Seed amount', bu.  .  . .
   Seed costS, dollars  .  . .
Seed cost per acieh, dollars
Total seed cost, dollars .  . .
   23,000
       10
   20,700
    0.274
    6.85
    6.85
1.712.50
                      Contour
                                         Terraced
                 Cno-t.    CCCWMno-t,   C chisel
                                          C strip
   26,000
       20
   20,800
    0.310
    7.75
    7.75
1,937.50
   26.000
       20
   20,000
    OJ10
    7.75
                                 1-5
                                11.25
                                  15
                               24.45
   11,79
2,947.50
   244)00
       13
    0.286
    7,15
   24,000
       13
   20380
    0.286
    7.15
    7.15
1,787.50
    7.15
1,787.50
                                         CB no-t.
   26,000
       20
   20,800
    0.310
    7.75
    1
    9.50
    8.62
2,155.00
   a Based on 84,000 seeds per bushel.
   b Assuming price of $25.00 per bushel (Iowa price, U.S. Department of AgikiJtme. AgrinOtvnlPrices. Apr. 15,1974.
   c Price of S7.50 per bushel (U.S. Department of Agriculture, Agricultural Prim, Sept. 15.1974),
   d Source: Sidney James (ed.), Midwest Farm Planning Manual. 3rd Edition, ISU Vuas, Aae*. Iowa, 1975, p. 18.
   e Price of $163.00 per 100 Ibs. (U.S. Department of Agriculture, AgriadaniPrices, Sept. 15,1974).
   ' Source:  Sidney James, op. cit., p. 20.
   i Price of $9.50 per bushel (U.S. Department of Agriculture, Agricultural Prices. Sept. 15,1974).
   " Average seed cost.
                                                                                                                  183

-------
                                                Table?. Fertilizer costs
Item

Corn
N* 	
P20< 	
K20 	
Wheat
N 	
P20S 	
K20 	
Soybeans
P,0S 	
K20 	
Average amount0
N 	
p,0s 	
K20 	
Cost of fertilizer per acred, dollars. .
Total cost of fertilizer, dollars ....
Rental of application equipment6,
dollars 	
Total fertilizer cost, dollars 	

Straight-row
C conv.
Contour
C no-t. CCCWM no-t.
C chisel
Terraced
C strip

CB no-t.

170
30
20
170
30
20
30.90
7,725.00
187.50
7,912.50
170
30
20
170
30
20
30.90
7,725.00
187.50
7,912.50
113° 170
30 30
20 20
60
25
30
80 170
23 30
18 20
17.06 30.90
4,265.00 7,725.00
125.00 187.50
4,390.00 7,912.50
170
30
20
170
30
20
30.90
7,725.00
187.50
7,912.50
150b
30
20
30
30
75
30
25
18.38
4,595.00
125.00
4,720.00
   a Fertilizer recommendations based on: Regis D. Voss, General Guide to Fertilizer Recommendations in Iowa, AG-65 (rev.), ISU
Cooperative Extension Service, Ames, Iowa, August 1973.
   b Includes fertilizer credit from meadow or soybeans.
   c Amount per year of rotation if other than continuous com.
   d Assume N as NH3 and P20S as 46 percent P20S. Prices per pound are $0.136 for N, $0.206 for P20S, and $0.080 for K20.
Source: Iowa price in U.S. Department of Agriculture, Agricultural Prices, September 15,1974.
   e Assume 50^/acre for NHs knife and 25^/acre for 4-ton bulk spreader.
                                                 TableS. Pesticide costs
               Item
Straight-row

  C conv.
                                                            Contour
             Terraced
                                                       C no-t.    CCCWM no-t.
C chisel
C strip
CBno-t.
Com
   Herbicide, dollars
   Insecticide, dollars
   Acres 	
   Total cost, dollars
     11,00          16.00         16.00         11.00        11.00           18.00
      7.00            7.00          9.00          7.00         7.00            7.00
    250            250          150          250        250            125
  4,500.00       5,750.00      3,750.00     4,500.00    4,500.00        3,125.00
Soybeans
Herbicide, dollars 	
Acres 	
Total cost, dollars 	
Total pesticide cost, dollars ....
11.00
125
1,375.00
4,500.00 5,750.00 3,750.00 4,500.00 4,500.00 4,500.00
184

-------
                                   Table 9. Labor costs
Item
                       Straight-row
       Contour
                             Terraced
                        C conv.
C no-t.
CCCWM no-t.    C chisel
C strip
CB no-t.
Total direct labor, hours ....
Overhead (30%), hours 	
Total labor, hours 	
Cost per hour, dollars 	
Total labor cost, dollars 	

765.25
229.58
994.83
2.50
2,487.08

611.25
183.38
794.63
2.50
1,986.58

714.50
214.35
928.85
250
2,322.13

685.50
205.65
891.15
2.50
2,227.88

611 25
183.38
794 53
2 50
1 986 58

484 75
145 43
630 18
2 50
1 575 45

Table 10. Other costs
Item
Corn dry ing
Grain harvested 	
Cost per bushel 	
Total cost 	
Interest (8%) on operating capital
Fertilizer (8 mo.) 	
Seed (8 mo.) 	
Pesticide (6 mo.) 	
Fuel (3 mo.) 	
Labor (3 mo.) 	
Total interest 	
Total other costs 	

Straight-row
C conv.

27,500
0.12
3,300.00
558.98
91.28 '
180.00
28.66
36.08
895.00
4,195.00
Contour
C no-t.

26,250
0.12
3,150.00
558.98
103.27
230.00
22.28
23.56
938.09
4,088.09
CCCWM no-t.
	 jju . .
16,500
0.12
1,980.00
342.19
157.10
150.00
28.54
33.18
711.01
2,691.01
C chisel

27,500
0.12
3,300.00
558.98
95.27
180.00
25.35
31.36
890.96
4,190.96
Terraced
C strip

27,500
0.12
3,300.00
558.98
95.27
180.00
22.28
26.98
883.51
4,183.51

CB no-t.

13,125
0.12
1,575.00
340.32
114.86
180.00
18.75
17.71
671.64
2,246.64
                                                                                                      18S

-------
                                                    Table 11. Revenue
                Item
                                      Straight-row
                                       C conv.
                        Contour
                 C no-t.
             CCCWMno-t.    C chisel
                                          Terraced
                             C strip
                            CB no-t.
Corn
   Expected yield, bu./ac	        110.0            105.0
   Area cropped, acres	        250              250
   Total output, bu	     27,500           26,250
   Expected price, dollars/bu	           2.75             2.75
   Gross revenue, dollars  	     75,625.00        72,187.50

Wheat
   Expected yield, bu./ac. .
   Area cropped, acres	
   Total output, bu	
   Expected price, dollars/bu.. .
   Gross revenue, dollars  	

Meadow
   Expected yield, tons/ac	
   Area cropped, acres	
   Total output, tons  	
   Expected price, dollars/ton.   .  .
   Gross revenue, dollars  	

Soybeans
   Expected yield, bu./ac	
   Area cropped, acres  	
   Total output, bu	
   Expected price, dollars/bu	
   Gross revenue, dollars  	
Total gross revenue, dollars	      75,625.00       72,187.50
                                 105
                                 150
                              15,750
                                   2.75
                              43,312.50
                                  45.0
                                  50
                               2,250.0
                                   4.00
                               9,000.0
                                   4.0
                                  50
                                 200.0
                                  45.00
                               9,000.00
                               110.0
                               250
                            27,500
                                 2.75
                            75,625.00
                               110.0
                               250
                             27,500
                                  2.75
                             75,625.00
                                105.0
                                125
                              13,125
                                   2.75
                              36,093.75
                              61,312.50     75,625.00     75,625.00
                                                             40.0
                                                            125
                                                          5,000.00
                                                              6.00
                                                         30,000.00
                                                         66,093.75
                                                   Table 12.  Summary
               Item
Straight-row

  C conv.
                                                              Contour
                                          Terraced
                                                       C no-t.
             CCCWM no-t.    C chisel
                              C strip
                            CB no-t.
Gross revenue	
Costs
   Tractor (excl. fuel)  . .  .
   Implements (excl. fuel).
   Fuel  	
   Seed  	
   Fertilizer	
   Pesticides	
   Labor 	
   Terracing	
   Other	
   Land charge (see text)  .
   Total cost	
Net return	
75,625.00

  3,430.25
 •5,940.23
  1,432.85
  1,712.50
  7,912.50
  4,500.00
  2,487.08
    0
  4,195.00
18,020.00
49,630,41
25,994.59
72,187.50

 3,066.46
 4,973.11
 1,113.92
 1,937.50
 7,912.50
 5,750.00
 1,986.58
    0
 4,088.09
18,020.00
48,848.16
23,339.34
	dollars	•
 61,312.50    75,625.00
              75,625.00    66,093.75
  3,426.24
  8,038.98
  1,374.07
  2,947.50
  4,390.00
  3,750.00
  2,322.13
     0
  2,691.01
 18,020.00
 46,959.93
 14,352.57
 3,238.69
 4,961.73
 1,267.69
 1,787.50
 7,912.50
 4,500.00
 2,227.88
 3,450.00
 4,190.96
18.020.00
51,556.97
24,068.05
 3,066.70
 5,077.54
 1,113.92
 1,787.50
 7,912.50
 4,500.00
 1,986.58
 3,450.00
 4,183.51
18,020.00
51,098.25
24,526.75
 2,887.79
 5,179.43
   891.73
 2,155.00
 4,720.00
 4,500.00
 1,575.45
 3,450.00
 2,246.64
18,020.00
45,626.04
20,467.71
186

-------
   As the discussion in Section 5, Volume I, points out,
there are a number of intangible variables not accounted
for in  the net  return figures. One of these intangibles,
that of scheduling, may be of only minor significance in
this  example. The scheduling of alternatives  with the
highest  net  returns does  not  differ  sufficiently  to
influence the decision.
   A more important consideration is the variability of
yields. The variance of yield under no-till is higher than
for production alternatives utilizing more tillage. This
higher variance for no-till may be partly due to a lack of
familiarity  with this method on the part of growers. In
the present example, the no-till production alternatives
were assumed  to have a lower  yield  than the  other
production  alternatives to  account  for this  potential
yield impact. A farmer who is a  risk-averter or who is
utterly unfamiliar with no-till planting may be willing to
accept  a lower  net  return  with a higher  degree of
certainty if that alternative excludes no-till planting.
   One  additional consideration related to the cost of
terracing. The  present example assumes tacitly  that the
full cost of terracing is borne by the farmer. Historically,
society  has  reimbursed terracing  costs through various
government programs so that the farmer usually paid
half the  cost or even  less. Under any such cost-sharing
program,  the relative differences in  net revenue  will
change. In the present example, a cost-sharing program
with  a 50-50 split would  give  two  of the terraced
alternatives  a net revenue practically  identical  to the
(unavailable) conventional tillage continuous corn activ-
ity.
*US. GOVERNMENT PRINTING OFFICE:1977 722-669
                                                                                                          187

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
  EPA-600/2-75-026b
                              2.
4. TITLE AND SUBTITLE
  Control  of Water Pollution  from Cropland: Volume  II—
  An  Overview
             3. RECIPIENT'S ACCESSION-NO.
                                                           5. REPORT DATE
                                                              June 1976
             6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
 E. A.  Stewart, D. A. Woolhiser,  W.  H. Wischmeir,
 J. H.  Caro, and M. H. Frere
             8. PERFORMING ORGANIZATION REPORT NO
                ARS-M-5-2
9. PERFORMING ORGANIZATION NAME AND ADDRESS
 Agricultural Research Service
 U.S.  Department of Agriculture
 Washington, D.C.   20250
             10. PROGRAM ELEMENT NO.

                1HB617
             11. CONTRACT/GRANT NO.
                                                              IA6 D4-0485
12. SPONSORING AGENCY NAME AND ADDRESS
  Environmental Research  Laboratory - Athens
  U.S.  Environmental Protection Agency
  Athens,  Georgia   30601
             13. TY.PE OF REPORT AND PERIOD COVERED
                Final Jan.  '74  -  June '76
             14. SPONSORING AGENCY CODE
                EPA-ORD
15. SUPPLEMENTARY NOTES
 Prepared  as a joint publication  of Office of Research  and Development, EPA,  and
 Agricultural  Research Service, USDA.
16. ABSTRACT
       Engineering and agronomic techniques to control  sediment, nutrient, and
 pesticide losses from cropland are identified, described,  and evaluated.  Method-
 ology is  developed to enable  a user to identify the potential sources of pollutants,
 select a  list of appropriate  demonstrated controls, and  perform economic analyses
 for  final selection of controls.   The basic principles on  which control of  specific
 pollutants is founded are  reviewed, supplementary information is provided,  and
 some of the documentation  used in Volume I is presented.   Volume I (Report  Mo.
 EPA-600/2-75-026a) is available from NTIS as report no.  PB 249-517.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.lDENTIFIERS/OPEN ENDED TERMS
                           c. cos AT I Field/Group
 runoff
 pesticides
 nutrients
 non-point source pollution
 hydrology
 sediment control
 erosion
   agriculture
   cropland
  13 B
 1. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report)
21. NO. OF PAGES
 Unlimited
                                              20. SECURITY CLASS (Thispage)
                           22. PRICE
EPA Form 2220-1 (9-73)

-------