National Conference on
IRRIGATION RETURN FLOW
QUALITY MANAGEMENT
PROCEEDINGS
-------
PROCEEDINGS OF NATIONAL CONFERENCE
IRRIGATION RETURN FLOW
QUALITY MANAGEMENT
Edited by:
James P. Law, Jr., Chief
Irrigated Agriculture Section
Robert S. Kerr Environmental Research Lab.
U.S. Environmental Protection Agency
Ada, Oklahoma 74820
and
Gaylord V. Skogerboe, Professor
Dept. of Agricultural and Chemical Engrg.
Colorado State University
Fort Collins, Colorado 80523
Sponsored by:
U.S. Environmental Protection Agency
and
Colorado State University
May 16-19, 1977
Colorado State University
Fort Collins, Colorado
-------
The opinions expressed herein are solely that of the
authors and do not necessarily represent official policies
of the representative organizations.
Printed in the United States of America
Library of Congress Catalog Card No. 77-78996
Copies may be obtained at a cost of $20 each from:
Department of Agricultural and Chemical Engineering
Colorado State University
Fort Collins, Colorado 80523
This proceeding published by:
Office of University Communications
Colorado State University
Fort Collins, Colorado 80523
-------
PREFACE
Previously, the U.S. Environmental Protec-
tion Agency (EPA) and Colorado State Univer-
sity cooperatively hosted a National Conference
on Managing Irrigated Agriculture to Improve
Water Quality. This conference was held in
Grand Junction, Colorado from May 16-18,
1972. A few years prior to this time, EPA had
formulated its irrigation return flow research
and development program. Then, a number of
research grants had been initiated throughout
the western states to launch this program. The
conference in Grand Junction focused primarily
upon: (1) identifying and describing irrigation
return flow quality problems; and (2) describing
the research being undertaken to develop
solutions to problems of irrigation return flow
quality.
The EPA irrigation return flow research
and development program has made substan-
tial progress in the past five or six years.
Numerous research projects and investigations
have been recently completed (or are nearing
completion) which focused upon defining ap-
propriate technologies for alleviating water
quality problems from irrigated agriculture;
most of these technologies involve improved
water management practices. In addition, case
studies have provided necessary experiences as
to how such technologies might be im-
plemented. These field experiences, combined
with studies of legal approaches, as well as
studies concerned with defining the processes of
implementation and the socio-economic con-
siderations that must be taken into account
prior to and during implementation, have
provided valuable insights as to the available
alternatives for implementing programs of
irrigation return flow quality management.
Thus, holding this National Conference on
Irrigation Return Flow Quality Management
provides for the timely dissemination of the
results from EPA's program.
The primary objectives for this conference
to be held May 16-19, 1977 will be to : (1) present
the results of the irrigation return flow research
and development program of the U.S. En-
vironmental Protection Agency; (2) integrate
the research results into an interdisciplinary
approach for solving problems of irrigation
return flow quality management; and (3)
provide a forum for presenting and discussing
the alternatives in implementing irrigation
return flow quality control.
We are now at that point in time where
sufficient research and investigation has been
completed so that more intelligent decisions can
be made to solve the water quality problems of
irrigated agriculture. These problems need to be
addressed. From a national and local stand-
point, now is the time for everyone concerned to
provide their input — to weigh the alternatives
and voice their opinions — in order that ap-
propriate and viable programs will be im-
plemented that are sensitive to both local and
national needs and priorities.
The editors appreciate very much the
cooperation of the authors in meeting the
necessary time schedule for printing these
proceedings so that they would be available at
the time of the conference, May 16, 1977. We are
deeply indebted to Stephen W. Smith and
Richard L. Aust of the Department of
Agricultural and Chemical Engineering, and
Ms. Patricia Tietz of the Printing and
Publications Office, all at Colorado State Uni-
versity, for their many hours of effort to insure
that this proceedings would be available for the
conference registrants.
James P. Law, Jr.
Gaylord V. Skogerboe
April, 1977
-------
Contents
THE ROLE OF EPA'S OFFICE OF RESEARCH AND DEVELOPMENT IN IRRIGATED CROP
PRODUCTION RESEARCH — James P. Law, Jr. and Arthur G. Hornsby, Robert S. Kerr
Environmental Research Laboratory, EPA 1
NITROGEN IN RETURN FLOWS
SIMULATION OF NITROGEN MOVEMENT, TRANSFORMATIONS, AND PLANT UPTAKE
IN THE ROOT ZONE — J. M. Davidson, P. S. C. Rao, and H. M. Selim, University of Florida 9
NITRATE MOVEMENT IN CLAY SOILS AND METHODS OF POLLUTION CONTROL — Allen R.
Swoboda, Texas Agricultural Experiment Station 17
EFFECT OF THREE IRRIGATION SYSTEMS ON DISTRIBUTION OF FERTILIZER NITRATE
NITROGEN IN SOIL — A. B. Onken, C. W. Wendt, 0. C. Wilke, R. S. Hargrove, Walter Bausch,
and Larry Barnes, Texas Agricultural Experiment Station 27
NITROGEN AND WATER MANAGEMENT TO MINIMIZE RETURN FLOW POLLUTION FROM
POTATO FIELDS OF THE COLUMBIA BASIN — B. L. McNeal, B. L. Carlile, and R. Kunkel,
Washington State University 33
VARIABILITY OF NITRATE LEACHING WITHIN DEFINED MANAGEMENT UNITS — L. J. Lund
and P. F. Pratt, University of California 45
FIELD MEASURED FLUX OF VOLATILE DENITRIFICATION PRODUCTS AS INFLUENCED BY
SOIL-WATER CONTENT AND ORGANIC CARBON SOURCE — D. E. Rolston, D. A. Goldhamer,
D. L. Hoffman, and D. W. Toy, University of California 55
SOIL NITRATE CONCENTRATIONS IN CORN PLOTS TREATED WITH ISOTOPICALLY
LABELED NITROGEN FERTILIZER — F. E. Broadbent and A. B. Carlton, University of
California 63
THEORETICAL AND EXPERIMENTAL OBSERVATIONS OF WATER AND NITRATE
MOVEMENT BELOW A CROP ROOT ZONE — J. W. Biggar, K. K. Tanji, C. S. Simmons,
S. K. Gupta, J. L. Maclntyre, and D. R. Nielsen, University of California 71
WATER MANAGEMENT
MINIMIZING SALT IN RETURN FLOW BY IMPROVING IRRIGATION EFFICIENCY - Jan van
Schilfgaarde, U.S. Salinity Laboratory 81
MODELING SALINITY OF IRRIGATION RETURN FLOW WHERE SOURCES AND SINKS ARE
PRESENT — R. J. Hanks, L. S. Willardson, and D. Melamed, Utah State University 99
FIELD EVALUATION OF SPRINKLER IRRIGATION FOR MANAGEMENT OF IRRIGATION
RETURN FLOW — L. S. Willardson, R. J. Hanks, and R. D. Bliesner, Utah State University 109
EFFECTS OF IRRIGATION MANAGEMENT ON SOIL SALINITY AND RETURN FLOW QUALITY -
P. J. Wierenga and J. B. Sisson, New Mexico State University 115
EFFECTS OF IRRIGATION SYSTEMS ON WATER USE EFFICIENCY AND SOIL WATER SOLUTE
CONCENTRATIONS — C. W. Wendt, A. B. Onken, O. C. Wilke, Raford Hargrove, Walter Bausch,
and Larry Barnes, Texas Agricultural Experiment Station 123
SCIENTIFIC IRRIGATION SCHEDULING FOR SALINITY CONTROL OF IRRIGATION RETURN
FLOW — Marvin E. Jensen, Snake River Conservation Research Center, ARS 133
RETURN FLOW MANAGEMENT
MANAGEMENT GUIDELINES FOR CONTROLLING SEDIMENTS, NUTRIENTS AND
ADSORBED BIOCIDES IN SURFACE IRRIGATION RETURN FLOWS - D. L. Carter and
J. A. Bondurant, Snake River Conservation Research Center, ARS 143
-------
QUALITY OF IRRIGATION RETURN FLOW FROM FLOODED RICE PADDIES - K. W. Brown,
L. E. Deuel, F. C. Turner, and J. D. Price, Texas A & M University, Texas Agricultural Experiment
Station at Beaumont and Texas Agricultural Extension Service 153
EVALUATION OF SURFACE IRRIGATION RETURN FLOWS IN THE CENTRAL VALLEY OF
CALIFORNIA — Kenneth K. Tanji, James W. Biggar, Robert J. Miller, William O. Pruitt, and
Gerald C. Homer, University of California, and USD A Economic Research Service 167
AN ECONOMIC ANALYSIS OF IRRIGATION RETURN FLOW RECYCLE SYSTEMS IN THE
CENTRAL VALLEY OF CALIFORNIA — William Kinney, Gerald L. Homer, and Kenneth K. Tanji,
Department of Agricultural Economics, USD A Economic Research Service, University of
California 175
ON-FARM METHODS FOR CONTROLLING SEDIMENT AND NUTRIENT LOSSES —
D. W. Fitzsimmons, C. E. Brockway, J. R. Busch, G. C. Lewis, G. M. McMaster, and C. W. Berg,
University of Idaho 183
ECONOMIC ANALYSIS OF ON-FARM METHODS FOR CONTROLLING SEDIMENT AND
NUTRIENT LOSSES — Karl H. Lindeborg, Larry Conklin, Roger Long, and Edgar Michalson,
University of Idaho 193
COMBINING AGRICULTURAL IMPROVEMENTS AND DESALINATION OF RETURN
FLOWS TO OPTIMIZE LOCAL SALINITY CONTROL POLICIES — Wynn R. Walker,
Colorado State University 203
IRRIGATION RETURN FLOW MODELS
PRACTICAL APPLICATIONS OF IRRIGATION RETURN FLOW QUALITY MODELS TO
LARGE ACREAGES — Marvin J. Shaffer and Richard W. Ribbens, U.S. Bureau of
Reclamation 217
AREAL PREDICTIONS OF SOIL WATER FLUX IN THE UNSATURATED ZONE —
A. W. Warrick, University of Arizona 225
WATER DISTRIBUTION PATTERNS FOR SPRINKLER AND SURFACE IRRIGATION
SYSTEMS — David Karmeli, Colorado State University 233
HYDRO-SALINITY MODELS: SENSITIVITY TO INPUT VARIABLES — J. D. Oster and
J. D. Wood, U.S. Salinity Laboratory 253
MODELING THE IRRIGATION RETURN FLOW SYSTEM - CURRENT CAPABILITIES
AND FUTURE NEEDS — Wynn R. Walker, Colorado State University 261
CASE STUDIES
MESILLA VALLEY
APPLICATION OF MODERN IRRIGATION TECHNOLOGY IN THE MESILLA VALLEY,
NEW MEXICO — T. W. Sammis and C. M. Hohn, New Mexico State University 269
ECONOMICS OF CONTROLLING IRRIGATION RETURN FLOW IN THE MESILLA
VALLEY, NEW MEXICO — Robert R. Lansford, Lynn W. Gelhar, and Bobby J. Creel,
New Mexico State University 277
SAN JOAQUIN VALLEY
AGRICULTURAL DRAINAGE PROBLEMS IN THE SAN JOAQUIN VALLEY — Edgar P. Price,
U.S. Bureau of Reclamation 283
HOW THE NPDES PROGRAM WILL DEFINE PRESENT WATER QUALITY CONDITIONS -
Gene Merrill, Regional Water Quality Control Board 289
LOCAL SOLUTIONS TO DRAINAGE PROBLEMS — William R. Johnstone,
Westlands Water District 293
A VALLEYWIDE SOLUTION — THE INTERAGENCY DRAINAGE PROGRAM —
Louis A. Beck, Interagency Drainage Program 297
-------
YAKIMA VALLEY
IRRIGATION RETURN FLOW PROBLEMS IN YAKIMA VALLEY — John Spencer and
Marc Norton, Department of Ecology, State of Washington 299
THE SULPHUR CREEK PILOT PROJECT: A PRACTICAL APPROACH TO CONTROL OF
POLLUTANTS LEAVING IRRIGATED FARMLANDS —John Spencer, Marc Norton, and
J. M. Gleaton, Department of Ecology, State of Washington, and
Soil Conservation Service 307
THE "208" PLANNING EFFORT FOR IRRIGATED AGRICULTURE IN THE
STATE OF WASHINGTON — Marc Norton and John Spencer, Department of Ecology, State of
Washington 321
WELLTON-MOHAWK IRRIGATION AND DRAINAGE DISTRICT
THE 1973 AGREEMENT ON COLORADO RIVER SALINITY BETWEEN THE UNITED
STATES AND MEXICO — Myron B. Holburt, Colorado River Board of California 325
AN ASSESSMENT OF IRRIGATION EFFICIENCIES AND DRAINAGE RETURN
FLOWS FROM THE WELLTON-MOHAWK DIVISION OF THE GILA PROJECT —
D. L. Krull and D. L. Clark, Bureau of Reclamation 335
WELLTON-MOHAWK ON-FARM SYSTEMS IMPROVEMENT PROGRAM — Richard S. Swenson,
Soil Conservation Service 349
GRAND VALLEY
RESEARCH AND DEMONSTRATION APPROACH TO DEVELOPMENT OF
APPROPRIATE SALINITY CONTROL TECHNOLOGIES FOR GRAND VALLEY —
Gaylord V. Skogerboe and Wynn R. Walker, Colorado State University 353
THE HYDRO-SALINITY SYSTEM IN THE GRAND VALLEY — Wynn R. Walker,
Gaylord V. Skogerboe, Robert G. Evans, and Stephen W. Smith,
Colorado State University 361
MODELING SALT TRANSPORT IN THE IRRIGATED SOILS OF GRAND VALLEY —
James E. Ayars, David B. McWhorter, and Gaylord V. Skogerboe, University of Maryland and
Colorado State University 369
EVALUATING APPROPRIATE TECHNOLOGIES FOR SALINITY CONTROL IN
GRAND VALLEY — Robert G. Evans, Wynn R. Walker, Stephen W. Smith, and
Gaylord V. Skogerboe, Colorado State University 375
DEVELOPMENT OF BEST MANAGEMENT PRACTICES FOR SALINITY CONTROL
IN GRAND VALLEY — Wynn R. Walker, Gaylord V. Skogerboe, and Robert G. Evans,
Colorado State University 385
IMPLEMENTATION
THE EPA GENERAL PERMIT PROGRAM — Kathe Anderson, EPA 397
INTERFACE OF WATER QUANTITY AND QUALITY LAWS IN THE WEST -
George E. Radosevich, Colorado State University 405
AN INFLUENT CONTROL APPROACH TO IRRIGATION RETURN FLOW QUALITY
MANAGEMENT — George E. Radosevich and Gaylord V. Skogerboe, Colorado State
University 423
A PROCESS FOR IDENTIFYING, EVALUATING AND IMPLEMENTING SOLUTIONS
FOR IRRIGATION RETURN FLOW PROBLEMS - E. C. Vlachos, J. W. H. Barrett
P. Huszar, J. Layton, G. Radosevich, M. Sabey, G. V. Skogerboe, and W. L. Track,
Colorado State University 435
vii
-------
APPENDIX
PUBLISHED REPORTS FROM EPA'S RESEARCH PROGRAM ON IRRIGATION
RETURN FLOW QUALITY
V1I1
-------
IRRIGATION RETURN FLOW
QUALITY MANAGEMENT
-------
Introduction
-------
The Role of EPA's
Office of Research
and Development in Irrigated
Crop Production Research
JAMES P. LAW, JR. and ARTHUR G. HORNSBY
Source Management Branch,
Robert S. Kerr, Environmental Research Laboratory,
Office of Research and Development,
U.S. Environmental Protection Agency, Ada, Oklahoma.
ABSTRACT
In setting the stage for the conference, the
role of the Irrigated Crop Production research
program will be described. The legislative man-
dates of PL 92-500 to EPA will be reviewed in
relation to the charge to control water quality
degradation resulting from agricultural ac-
tivities. The overall objectives of the program,
established prior to the passage of PL 92-500,
appear to be valid and supportive of the efforts
of the program thus far. This conference
represents a compendium of the results of the
program's efforts as well as a hard look at the
future direction the program should take.
Supported research has included
technology development and evaluation of its
effectiveness in water pollution abatement
andJor control. The technology investigated has
included both structural and nonstructural im-
provements within the water delivery, farm,
and water removal subsystems of irrigated
agriculture. Additional studies have in-
vestigated the legal, socio-economic, and/or
other institutional constraints to water
management reform. More effort will be re-
quired in these areas.
Current program emphasis is evolving into
a phase which we are referring to as implemen-
tation research. Included in this is the prepara-
tion of research-based guidance documents aim-
ed at providing much needed information for
those state agencies responsible for planning
and implementing areawide wastewater
management alternatives involving nonpoint
source agricultural activities. The discussions
of this conference should provide great impetus
toward that goal.
INTRODUCTION
On October 18, 1972, the 92nd Congress
signed into law the "Federal Water Pollution
Control Act Amendments of 1972". This act sets
forth an ambitious goal to "restore and main-
tain the chemical, physical, and biological in-
tegrity of the Nation's waters". In order to
achieve this objective, Congress declared the
following national goals and policies:
Section 101. (a)
"(1) It is the national goal that the discharge of
pollutants into navigable waters be
eliminated by 1985;
"(2) It is the national goal that wherever at-
tainable, an interim goal of water quality
which provides for the protection and
propagation of fish, shellfish, and wildlife
and provides for recreation in and on the
water be achieved by July 1, 1983;
"(3) It is the national policy that the discharge
of toxic pollutants in toxic amounts be
prohibited;
"(4) It is the national policy that Federal finan-
cial assistance be provided to construct
publicly owned waste treatment works;
"(5) It is the national policy that areawide
waste treatment management planning
processes be developed and implemented to
assure adequate control of sources of
pollutants in each state; and
-------
INTRODUCTION
"(6) It is the national policy that a major
research and demonstration effort be made
to develop technology necessary to
eliminate the discharge of pollutants into
navigable waters, waters of the contiguous
zone, and the oceans."
The above statements come from Public
Law 92-500, 92nd Congress, S.2770, October 18,
1972, which documents the amendments to
Section 2 of the Federal Water Pollution Control
Act. The 1972 amendments replace the previous
languages of the Act entirely including the
Water Quality Act of 1965, the Clean Water
Restoration Act of 1966, and the Water Quality
Improvement Act of 1970, all of which have
been amendments of the Federal Water Pollu-
tion Control Act first passed in 1948. Subse-
quently, P. L. 92-500 has been amended by P. L.
93-207, December 8, 1973; P. L. 93-243, January
2, 1974; P. L. 93-592, January 2, 1975; P. L. 94-
238, March 23, 1976; and P. L. 94-558, October
19, 1976. These amendments have not substan-
tially changed the intent of P. L. 92-500 but
rather are meant to clarify the wording of
various sections and subparts thereof.
The Congress addressed the responsibility
of the Nation in achieving these goals and
policies in Section 101 (b). There is is stated "It is
the policy of the Congress to recognize, preserve,
and protect the primary responsibilities and
rights of States to prevent, reduce, and
eliminate pollution, to plan the development
and use (including restoration, preservation,
and enhancement) of land and water resources,
and to consult with the Administrator in the
exercise of his authority under this Act. It is
further the policy of the Congress to support and
aid research relating to the prevention, reduc-
tion, and elimination of pollution, and to
provide Federal technical services and financial
aid to State and interstate agencies and
municipalities in connection with the preven-
tion, reduction, and elimination of pollution."
The Congress also addressed the ad-
ministration of the Act in Section 101 (d):
"Except as otherwise expressly provided in this
Act, the Administrator of the Environmental
Protection Agency (herein after in the Act called
'Administrator') shall administer this Act."
Thus, we can see that the Environmental
Protection Agency has the responsibility of
administering the congressional intent of P. L.
92-500. The primary responsibilities of the
Agency are spelled out in Title I — Research and
Related Programs, Title II — Grants and Con-
struction of Treatment Works, Title III — Stand-
ards and Enforcement, Title IV — Permits and
Licenses, and Title V — General Provisions.
The intent of this paper is to provide some
insight into the Agency's charge given by
Congress and how research funded by the
Agency fits into the overall operation of a
regulatory Agency. The Agency was created by
Presidential Reorganization Plan No. 3 in
December 1970. This plan brought together 15
programs scattered among several Federal
Government Agencies to mount a coordinated
attack on environmental problems. These
problems include air and water pollution, solid
waste management, pesticides, water supply,
radiation, noise, and toxic substances.
Organization and Functions of EPA
EPA must maintain and enhance en-
vironmental quality in a way that is consistent
with other national goals. Functions performed
by EPA include: setting and enforcing en-
vironmental standards; researching the causes,
effects, and control of environmental problems;
assisting states and local governments through
a variety of planning and waste treatment
facility construction grants; disseminating in-
formation on environmental problems and
solutions; educating the public; demonstrations
of how to protect and enhance the environment;
and providing technical assistance in solution
of environmental problems. To better under-
stand this relationship it is appropriate to look
at the manner in which EPA is organization-
ally structured and the relative role of the
various key elements. This structure is given in
Figure 1. The functions of the key elements and
their interrelationships are briefly as follows:
Office of the Administrator. This office is
responsible for the supervision and direction of
the programs and activities of the Agency.
Office of the Assistant Administrator for
Planning and Management. This office is
responsible for the overall management ac-
tivities of the Agency. . . coordinates planning,
evaluation, and standard setting efforts . . .
directs the Agency's resource processes. . . and
provides administrative support.
Office of the Assistant Administrator for
Enforcement. This office serves as the principal
advisor to the administrator in matters per-
taining to the enforcement of standards for
-------
ROLE OF EPA IN RESEARCH
environmental quality and is responsible for the
conduct of enforcement activities on an Agency-
wide basis.
Office of the Assistant Administrator for
Water and Hazardous Materials. This office is
responsible for the Agency's pesticide, toxic
substances, and water programs . . . develops
standards, criteria, and national policy (except
for the area of enforcement and research).
Office of the Assistant Administrator for
Air and Waste Management. This office is
responsible for administration and operations
of the Agency's air, solid waste, noise, and
radiation programs . . . policy evaluation
assistance and selected demonstration projects.
Office of the Assistant Administrator for
Research and Development. This office is
responsible for the development, direction, and
conduct of a national research, development,
and demonstration program in pollution
sources, fate, and health and welfare effects;
pollution prevention and control, and waste
management and utilization technology; en-
vironmental sciences; and monitoring systems.
The AA for R&D serves as principal science
advisor to the Administrator and coordinator
for the Agency's policies and programs con-
cerning carcinogenesis and related problems.
Regional Offices. The country is divided
into ten regions, each region being administered
through a regional office. The primary respon-
sibility of the regional administrator in each
region is to accomplish the national program
objectives. The regional administrator and his
staff serve as the Agency's principal represent-
atives in that region in contacts and
relationships with Federal, State, and local
agencies; industry; academic institutions; and
other public and private groups.
With this overview of the Agency structure
and functions we can look more intensively at
the role of the Office of Research and Develop-
ment (ORD). In support of the Agency's mis-
sion, the ORD conducts a comprehensive and
U. S. ENVIRONMENTAL PROTECTION AGENCY
ASST ADMINISTRATOR
FOR PLANNING
-
of net o
AOMtNSTRA
OFFICE O
**C> CVAIUA
o..«o
AE SOURCE
Figure 1. Organizational Structure of the Environmental Protection Agency (June 1976).
3
-------
INTRODUCTION
integrated research and development program
to provide:
— The scientific and technical base for
reasonable standards and regulations.
— Standardized methods to measure and
assure quality control in programs to
assess environmental quality, imple-
ment regulations and enforce standards.
— Cost-effective pollution control
technology and incentives for accep-
tance of environmentally sound options.
— Scientific, technical, socio-economic,
and institutional methodologies needed
to judge environmental management op-
tions against competing national needs.
ORD's research is supplemented by general
scientific and technical research in other federal
agencies, colleges and universities, and
elsewhere. Governmental agencies function in a
manner different than academia in that the
scope of interest and objectives of endeavor are
closely tied to legislative mandates and they
must provide outputs applicable on a national
scale. ORD establishes its objectives and
priorities in response to the overall mission and
priorities of EPA and is highly mission oriented,
concerned with solving specific priority
problems rather than only advancing scientific
knowledge. Although the scope of ORD projects
may vary from quite fundamental research to
the full-scale engineering demonstration of new
pollution control processes, all projects are
directed at meeting specified objectives.
Within ORD, research related to the agri-
cultural industry is handled administratively
under the Office of Air, Land and Water Use
(Figure 1) in the technical program area of
Industrial Processes. The Industrial Processes
Program Area contains the Renewable
Resources Subprogram which includes
Irrigated Crop Production, Non-Irrigated Crop
Production, Forest Management, Animal
Production, and Alternate Pest Management.
The Renewable Resources Subprogram encom-
passes the development of total management
systems, including predictive methodology to
control air, water and land pollution from the
production and harvesting of food and fiber and
their related residual wastes; and assessment of
probable trends in the productiion of renewable
resources and their resulting environmental
impact.
Problem Definition and Program
Objectives
Irrigated Crop Production aspects of the
Renewable Resources Subprogram are ad-
ministered by the Irrigated Agriculture Section,
Source Management Branch, Robert S. Kerr
Environmental Research Laboratory, Ada
Oklahoma. The control of environmental
degradation caused by irrigated crop produc-
tion is a multifaceted problem involving
technical, legal, economic, and institutional
considerations. The objective of this program is
to develop and demonstrate the fundamental
technology needed for full-scale pollution con-
trol programs in irrigated areas. This
technology includes: canal and lateral lining
and other structural controls for water delivery
systems; methods to minimize water use and
increase water use efficiency; control of nutrient
losses; salinity control; sediment control; con-
trolled leaching losses; pesticide transport in
irrigated systems; and treatment processes.
Evaluation of the legal, economic and in-
stitutional constraints to water management
reform and technology changes is required.
Development and verification of mathematical
simulation and predictive techniques based on
physical-chemical-biological processes oc-
curring in irrigated soil systems is required to
assess the effects of on-farm water management
practices on the water quality of receiving
streams. These models can be used to develop
technically sound alternative pollution control
management schemes for irrigated systems.
The alternatives will include waste stream
treatment processes. The output would be used
by Federal, state, and local planning and pollu-
tion control agencies for the assessment and
control of pollutants resulting from irrigated
crop production activities. Water quality
problems arising from irrigation activities are
influenced by water, crop and land manage-
ment practices and by the agricultural
chemicals (fertilizers and pesticides) applied to
the land. The major water pollutants from
irrigated land include salinity (mineral salts),
sediments (soil erosion), fertilizer nutrients
(nitrogen and phosphorus), and bioactive
materials (toxic and crop residue).
-------
ROLE OF EPA IN RESEARCH
Program Accomplishments
Historically, the Irrigated Crop Production
program (formerly the Irrigation Return Flow
program) has looked extensively into treatment
processes and structural control measures to
alleviate pollution arising from irrigated crop
production. An intensive study involving the U.
S. Bureau of Reclamation, California Depart-
ment of Water Resources, and the EPA was
conducted on the Bio-Engineering Aspects of
Agricultural Drainage, San Joaquin Valley,
California. This project examined and
evaluated treatment methods for the removal of
nutrients and salinity from agricultural
drainage waters. The evaluation of structural
control measures has been performed primarily
in the Grand Valley of Colorado. Canal lining,
lateral lining, and drainage systems have been
studied to evaluate the relationship between
excessive seepage and the mass emission of
salts to the receiving stream — in this case — the
Colorado River.
Following the initial thrust of examining
treatment processes and structural control
measures, the program emphasis has evolved to
reflect the primary importance of resource
management as the key to reducing or
eliminating environmental degradation caused
by irrigated agricultural activities. This ap-
proach is aimed at solving the problem at the
point where it is created rather than treating the
effluent. Studies have been funded to evaluate
the effectiveness of water, salinity and nutrient
management practices on the reduction of mass
emission of pollutants from irrigated lands as
well as the effect on crop yields.
Aside from the obvious benefits to improve
water quality, the information derived from
these studies will provide a basis for develop-
ment of "Best Management Practices" required
in support of Section 208 of P. L. 92-500 where
the states are required to submit a plan which
"shall include ... (1) a process to identify . . .
agriculturally-related nonpoint sources of pollu-
tion . . . and, (2) a process to set forth procedures
and methods ... to control to the extent feasible
such sources." These studies will also provide
information to develop feasible effluent
limitations as required for point sources in
Section 301 "Effluent Limitations". Several
projects have been funded to address the need
for more definitive information to support the
development of the NPDES irrigation return
flow permits. These permits are required under
Section 402 "National Pollutant Discharge
Elimination System" where it is stated "the
Administrator may, after opportunity for public
hearing, issue a permit for the discharge of any
pollutant, or combination of pollutants,
upon condition that such discharge will meet
either all applicable requirements under Sec-
tions 301, 302, 306, 307, 308, and 403 of this Act,
or prior to the taking of necessary implementing
actions relating to such requirements, such
conditions as the Administrator determines are
necessary to carry out the provisions of this
Act."
Future Program Emphasis
During the past several years, many
research and demonstration projects have been
completed which have demonstrated and
evaluated the effectiveness of technologies for
the control of pollutants arising from irrigated
agriculture. These technologies have included
both structural and nonstructural measures to
reduce salinity, nutrient and sediment loadings
on receiving streams and/or groundwater
aquifers. Economic factors and cost-
effectiveness have been investigated as well.
The results clearly indicate the feasibility of
pollutant control at the source through im-
proved water management and control on-the-
farm. These observations lead to the question of
how such controls can best be implemented in
valley-wide areas and they also provide the
emphasis for future directions the program
should take.
Long-standing methods and practices tend
to be followed year after year unchanged until it
can be demonstrated that changes will result in
economic gain. It follows that implementation
of pollutant control measures will not be
accepted and adopted until it can be shown that
benefits will accrue to the user. Demonstration
type projects will be required in many more of
the problem areas to show the farming com-
munity what can be done in better water
management, fertilizer management, etc., for
irrigation return flow quality control. These
must be shown to be economically feasible as
well as cost-effective to reduce pollutant dis-
charges. Projects of this type are expensive and
require seceral years for successful evaluation
to be completed. Educational methods, informa-
tion exchange and good public relations are
required to encourage acceptance of improved
management practices.
-------
INTRODUCTION
Additional effort will be required to com-
plete the needed manuals of recommended prac-
tices for the use and guidance of the state
agencies responsible for planning and im-
plementing controls over major nonpoint
sources in the agricultural areas. These will be
needed in regional areas where water quality
problems and effective controls may differ
greatly. Further analysis and methodology will
be required to assess the socio-economic, legal
and institutional problems facing implementa-
tion of control technologies. These must include
alternative solutions and decision-making logic
to arrive at equitable choices of pollutant con-
trol programs to be implemented.
As more technologies are developed,
demonstrated and evaluated, the program
emphasis must shift more to implementation
research. Outputs of educational material,
manuals of recommended practices, total
system management models, and information
dissemination activities must be planned into
future years. The present conference is a signifi-
cant effort toward passing information along to
the user community. A research needs
workshop will be convened immediately follow-
ing this conference to identify needs and
priorities for emphasis by the Irrigated Crop
Production program. EPA's continued support
of its ORD programs is required if we are to
successfully follow through with a strong
program support of the implementation of
technologies which are required to show water
quality improvement in our western river
basins.
-------
Nitrogen in Return Flows
-------
Simulation of
Nitrogen Movement,
Transformations, and Plant
Uptake in the Root Zone
J. M. DAVIDSON, P. S. C. RAO, and H. M. SELIM
Soil Science Department, University of Florida
Gainesville, Florida
ABSTRACT
Two simulation models, a detailed
research-type and a conceptual management-
type, for describing the fate of nitrogen in the
plant rootzone are discussed. Processes con-
sidered in both models were: one-dimensional
transport of water and water-soluble N-species
as a result of irrigation/rainfall events,
microbiological N-transformations, and uptake
of water and nitrogen species by a growing crop.
The research-type model involves a
finite-difference approximations (explicit-
implicit) of the partial differential equations
describing one-dimensional water flow and
convective-dispersiue NH4 and NO3 transport,
along with simultaneous plant uptake and
microbiological N-transformations. Ion-
exchange (adsorption-desorption) of NH^ was
also considered. The microbiological transfor-
mations incorporated into the model describe
nitrification, denitrification, mineralization
and immobilization. All transformation
processes were assumed to be first-order kinetic
processes. The numerical solution was flexible
in its soil surface boundary conditions, as well
as initial conditions for soil water content and
nitrogen concentration distributions in the soil
profile. The solution can also be used for non-
homogeneous or multilayered soil systems.
The research - type model involves a
detailed description of the individual process
and requires a large number of input
parameters, most of which are frequently un-
available. Because of this a more simple
management-type model was developed.
Several simplifying assumptions were in-
troduced into the management model. This
model requires a minimal amount of input data
by the user, and provides a gross description of
the behavior of various nitrogen species in the
plant root zone.
INTRODUCTION
Large quantities of nitrogen are applied
annually to land surfaces in the form of com-
mercial fertilizers, animal manure, municipal
and industrial waste and irrigation water. The
ultimate fate of this nitrogen is of public interest
because of its potential for surface water and
groundwater pollution when in the nitrate form.
Also, an understanding of the dynamic
behavior of nitrogen in various soil-water-plant
systems is essential for increasing the efficiency
of nitrogen use by crops. This greater efficiency
is responsible for reducing the potential for
groundwater pollution by nitrates.
Various inorganic (NH4, NO 3, NO 2, N 2O
and N2) and organic nitrogen forms exist
simultaneously in the soil. These and other
nitrogen substrates undergo reversible and/or
irreversible transformations owing to chemical
and microbiological processes. The water-
soluble nitrogen species (NH4, NO%, and NO 2)
are also transported through the soil in response
to soil-water movement. The NH4 and NO 3
distribution is complexed further by absorption
by plant roots. The extent of water and nitrogen
uptake is determined, in part, by the transpira-
tion demand, which in turn is dependent upon
crop growth stage and meteorological variables.
The complexity of the soil-water-plant
system is further enhanced by the fact that all of
-------
NITROGEN IN RETURN FLOWS
the above processes occur simultaneously. The
relative importance of any or a combination of
these processes in determining nitrogen
behavior is dependent not only on several
physical, chemical and biological soil prperties,
but also on the growth stage of the crop.
Therefore, a prerequisite to modeling the fate of
nitrogen in soil-water-plant systems is an un-
derstanding of the processes discussed above. A
considerable amount of qualitative information
is available regarding nitrogen and its
agronomic aspects and individual processes in
soils (Bartholomew and Clark, 1965). However,
due to the nature and conditions under which
much of this research was conducted, it is
difficult to separate this information into a form
that can be used to develop relationships useful
for simulation and/or prediction purposes.
Intensive research efforts by several
researchers during the past decade have yielded
a multitude of models for simulation of nitrogen
behavior in soil-water-plant systems. However,
due to our limited understanding of the major
processes and the interrelationships among
them, considerable divergence exists among the
modeling approaches undertaken to date. A
review of the state-of-the-art of nitrogen simula-
tion models is presented by Tanji and Gupta
(1977).
In this manuscript we will describe and
evaluate two nitrogen simulation models for
describing the fate of nitrogen in a plant-root
zone. The first is a detailed research-type model
and the second a conceptual management-type
model.
SIMULATION OF WATER AND
NITROGEN
Three major processes that need to be
considered in all nitrogen simulation models
are: transport of water-soluble nitrogen species
by the soil water, chemical and/or
microbiological transformations, and plant up-
take of water and nitrogen. The following
discussion is centered around these processes.
Research-Type Model
One-dimensional water flow through satur-
ated and unsaturated soils is described by the
following nonlinear partial differential equa-
tion:
at
= J_fK,.. -*)
He)
-W
dz
dz
(z,t)
[1]
where d is the volumetric soil-water fraction
(cm ^/cm 3), t is time (day), z is vertical distance
(cm), K (0) is hydraulic conductivity (cm/day), h
is soil-water potential at a given location z (cm)
and W (z,t) represents plant root extraction
(hr-1). A numerical solution (finite differences
by and implicit-explicit procedure) to equation
[1] is presented by Selim et al. (1977a), among
others.
Plant root extraction of water, W(Z)t), was
described using the Molz-Remson (1970) model:
(z,t)
= T(t)
D (8) R (z,t)
[2]
where, T(t) is evapotranspiration rate
(cm/day), L is depth of root zone (cm), R(z,t) is
"effective" plant root distribution which is
proportional to root density (cm-2), and D(g)is
the soil-water diffusivity (cm ?/day). The total
available water (TAW) for plants is defined as
that held between "field capacity" and 15-bar
volumetric soil-water contents. The evapo-
transpiration rate is equal to the potential evap-
otranspiration rate (PET) when the available
water (AW) in the profile is greater than 20% of
TAW (Ritchie, 1973). PET is obtained from the
Penman equation. The value of PET was ad-
justed by multiplying it by a "crop factor" to
account for plant growth and age during the
growing season.
The movement of water-soluble nitrogen
species through soils occurs as a result of
molecular diffusion and mass transport in the
soil-water phase. Because of the general accep-
tance of chromatography theory and its
applicability to a soil-water system, this ap-
proach was used to describe the vertical move-
ment and distribution of water-soluble nitrogen
species. The partial differential equation for
one-dimensional solute transport is:
[3]
dt
= — D ,
dz *
± i describes the biological transformations
influencing the itn nitrogen species.
10
-------
NITROGEN IN THE ROOT ZONE
The mobility of the ammonium (NH , +) ion
in a soil-water system is directly influenced and
controlled by the adsorption-desorption charac-
teristics of the NH 4 + with the soil matrix.
Numerous equations have been used to describe
adsorption-desorption, but the most common
are the Freundlich, first-order kinetic, and
Langmuir equations (Davidson et al., 1976).
Other types of cation exchange equations that
could be used to describe the adsorption-
desorption of NH 4 "*" are described by Duttet al.
(1972). Thermodynamically based adsorption-
desorption equations require more information
about the composition of the soil solution than is
generally available. It is believed that simpler
adsorption models can be used as reasonable
approximations for the adsorption-desorption
of NH /*" in many soil-water systems.
Equations [1], [2] and [3] have been solved
numerically by Selim et al. (1977b) using a finite
difference (explicit-implicit) approximation
technique. The solution calculates soil profile
distributions of nitrogen species (NH^, NO 3
and organic-N) as well as soil water content (or
tension) at selected time intervals. The model,
written in FORTRAN IV programming
language, allows for water infiltration and
redistribution with either a water flux or water
head boundary condition at the soil surface. The
nitrogen transformation processes were
described by first order kinetic equations with
the rate coefficients made a function of soil
depth and soil water potential (h).
Assuming a linear Freundlich adsorption
for NH 4 ~*~ and first-order rate processes for the
nitrogen transformations shown in Figure 1,
equation [3] can be rewritten in the following
form for NH4 + and NO 3" in the soil solution
(Selim et al., 1977b):
dz-
8 dz
— k;i 0
(i
dt dz- o £.L
where A = concentration of NH , + in soil solution
(pg/cm:)),
B = concentration of NO 3 " in soil solution
(/jg/cm3),
O = amount of N in organic phase (pg/g),
k1? k2, k:), k4, k=, = kinetic rate coefficients,
respectively, for NH 4 +
nitrification, NO 3 "
immobilization, NH4 +
mineralization, immobili-
zation of organic-N, and
NO g " denitrification
(day -1).
V = q(z) -6 -^-- D-ff- , where q(z) is the
Darcy water flux (cm/day),
R = 1 + pK/e, retardation factor for NH4 +
exchange,
K = distribution coefficient of NH 4 in
exchangeable phase (cmVg).
The transformation processes for organic
N are described by:
P = k2 0 B + k4 0 A - k3 P 0 [6]
dt
and the gaseous loss of N due to denitrification is cal-
culated from:
=k5eB
at
[7]
where G is the sum total of NoO, NO, and/or
N2 gas (,ug/g). All transformation rate coef-
ficients (k j) were made a function of the soil-
water potential (h) or soil-water content (6),
and/or soil organic matter content (0) at any
given depth within the soil profile. These em-
pirical relationships, similar to those used by
Hagin and Amberger (1974), were devised from
published data and are reported by Selim et al.
(1977b).
Nitrogen uptake by plants involves the
movement of water-soluble nitrogen species
(NH 4 and NO 3) to the roots followed by their
absorption across the root surfaces. Mass flow
and diffusion are the two major processes by
which NO 3 and NH4 are transported to the
roots (Nye and Spiers, 1964; Passioura and
Frere, 1967; Marriott and Nye, 1968; Olsen and
Kemper, 1968; Phillips et al., 1976). Convective
flow of water towards roots in response to
transpiration results in the transport of NH4
and NO 3 to the roots along with the water. The
concentration of these ions at the root surface
decreases when the rate of root uptake exceeds
the rate of supply of these ions by mass flow.
Diffusion of NH 4 and NOs towards the roots
then occurs due to the concentration gradient.
The nitrogen species taken up by plant roots
are NH4 and NO 3 (Dibb and Welch, 1976).
However, due to the relatively rapid transforma-
tion of NH 4 to NO 3 and the mobility of the
latter ion, most researchers have considered
only the uptake of NO 3. Rao et al. (1977a) have
11
-------
NITROGEN IN RETURN FLOWS
considered the uptake of both NH 4 and NO 3 in
proportion to their presence in the root zone.
Data are unavailable to determine the frac-
tional uptake of NH 4 and NO 3 by plants when
both species are equally available. The max-
imum N-uptake demand (jug N/day/cm2 soil
surface) by the plant at any time during the
growing season, denoted QNmax> was deter-
mined in a similar manner to that used by Watts
(1975). The values of QNnaax were determined
by analyzing the cumulative N-uptake by the
crop grown under non-limiting conditions for
either water or nitrogen. The N-uptake rate,
q v}max, was calculated as follows:
max
N
= Qmax//" R(z,t)dz
N
[8]
where, Q Nm&X is defined above, the integral of
R(z,t) over the rooting depth L represents the
total root length in the profile. Note that q jvjmax
has the dimensions of MgN/day/cm root
length/cm2 soil surface/cm depth. The root
capacity for N-uptake is assumed to be constant
over the entire root length.
A Michaelis-Menton type relationship
determined the actual N-uptake rate (q jq) on the
basis of the soil solution concentrations of
nitrate (CNQ ) and ammonia
O
max
QN =qN
[9]
where, K is the sum of solution concentrations of
NO 3 and NH 4 at which q j^ equals 0.5 q Nmax •
It should be noted that the processes (diffusion
and mass flow) involved in transporting NO 3
and NH4 to the plant roots are implicitly
ignored in the approach described above. The
value of qj$, calculated as in equation [9],
multiplied by the root length density R(z,t) at
any soil depth, yields the value for the nitrogen
uptake sink term to be included in equation [3].
In the Molz-Remson (1970) plant root ex-
traction model, the transpiration demand is
expressed in units of cm water/cm root
length/day. This requires a knowledge of the
exact nature of root distribution in the soil
profile at all times during the growing season.
The water absorptive capacity of roots is assum-
ed to be uniform over the entire root length in the
model. Unfortunately, the water absorptive
capacity is known to be neither uniform or
constant (Newman, 1974). Models that incorpo-
rated the processes of die-off, regeneration,
proliferation and extension of roots in response
to soil-water stress are available in the literature
(Hillel and Talpaz, 1976). For a detailed discus-
sion of root-soil water relations and root
research problems, the reader is referred to
excellent articles by Barley (1970), Pearson
(1974) and Newman (1974).
Empirical root growth models were devised
(Rao et al., 1977a) based on measured corn (Zea
mays) root length distributions in the soil
profile at different times (NaNagara et al.,
1976). These distributions were used as inputs in
the modified Molz-Remson water uptake model
(equation [1], [2])and N-uptake model (equation
[8]). The root capacity for absorption of water
and nitrogen was assumed uniform over the
entire root length. Reduction in absorption
capacity due to age or suberization was
neglected. Die-off and regeneration of roots due
to water stress at any given soil depth, as
modeled by Hillel and Talpaz (1976), was also
neglected.
In summary, the research-type model
describing simultaneous one-dimensional tran-
sient water and solute transport, ion-exchange,
microbiological transformations, and plant up-
take of water and nitrogen is based on
numerical solutions of equations [1], [3], and [9]
with appropriate modifications as discussed
above. Further details of this complete model
and some typical simulations are reported
elsewhere by Selim et al. (1977b).
Management-Type Model
The research-type model, described in the
preceeding section, is conceptually pleasing in
that it represents a mechanistic description of
the soil-water-plant system. This model, and
others like it, require extensive amounts of input
data that are commonly unavailable. Also,
application of these models to describe the
behavior of nitrogen on large-scale watersheds
does not seem feasible due to the spatial
variability of the soil/plant properties. At the
present time, however, the research-models are
useful in performing sensitivity analyses to
identify the most significant processes and/or
parameters, thereby allowing for simplifica-
tion. The use of simpler models becomes more
desirable when only gross descriptions are
required. Such a simple conceptual
management-type model for describing the fate
of nitrogen in the plant root zone has been
12
-------
NITROGEN IN THE ROOT ZONE
described in detail by Rao et al. (1977a). Some
prominent features of this model are discussed
in the following paragraphs.
Several simplified forms of the transient,
one-dimensional water flow model, i.e. equation
[1] without the plant uptake sink term W(z,t),
have been used (Dutt et al., 1972; Beek and
Frissell, 1973; Duffy et al., 1976). Perhaps the
most simplified concept is that "piston-
displacement" used by Frere et al. (1975) and
Rao et al. (1976a). This concept is based on two
major assumptions: (i) all soil pore sequences
participated in solute and water transport, and
(ii) the soil water initially present in the profile
is displaced ahead of the water entering at the
soil surface. Based on these assumptions, Rao et
al. (1976a) showed that the depths of wetting
front (d wf ) and nonreactive solute front (d sf)
resulting from infiltration of I cm of water into
the soil profile were:
dwf = I/(0f -
dsf=I/flf
[10a]
[lOb]
and, dividing equation [lOb] by equation [lOa],
dsf
dwf
A,
HOc]
Published data for nitrate and chloride
movement in several soils were analyzed accor-
ding to equation [lOc] to test its validity. These
results, shown in Fig. 2, indicated that assump-
tions (i) and (ii) are indeed valid for a wide range
of soils examined. Equation [lOb] may further
be extended to calculate the depth of solute front
(d? sf) after the soil profile had drained to a "field
capacity" water content (0FC) as:
d sf =
FC
[11]
Finally, equation [11] was modified to consider
reactive solute and the case when solute front
was initially at some depth d j in the profile,
I
, R > 1 [12]
FC
where, R is the retardation factor due to adsorp-
tion.
Rao et al. (1976a) have also considered plant
water extraction using the Molz-Remson model
(equation [2]) in the above approach. The agree-
ment between prediction and data obtained
from a field study, Fig. 3, led to further im-
provements to include nitrogen transfor-
mations, and plant N-uptake and are discussed
in detail by Rao et al. (1977a). The
microbiological transformations included in the
modified model were: nitrification, mineraliza-
tion, and immobilization. Ion exchange of NH 4
was described by a linear isotherm. Nitrogen
species resulting from transformations were
estimated with analytical solutions (Rao et al.,
1977b) to the differential equations describing
first-order kinetic reactions. These solutions are
based on the total amounts of a given nitrogen
species within the soil profile and are valid
under the following conditions: (1) within the
region of interest, the net solute flux (input-
output) in the soil profile is zero, and (ii) the soil
profile is homogeneous with regard to adsorp-
tion partition coefficient for NH 4 and kinetic
transformation rates. Assumption (i) can be met
for deep profiles. Rao et al. (1976b) have shown
that when all rate coefficients were assumed
constant rather than depth-dependent,only
small errors were introduced in estimating N-
transformations during steady-state water
flow; thus, assumption (ii) may not be limiting.
Root growth and nitrogen uptake were describ-
ed in a similar manner to that for the research
model. However, the nitrogen uptake rate (q N
in equation 9) was determined by the total
amounts of nitrate and ammonia within the
root zone, rather than their concentration dis-
tributions as in equation 9. The q ^ values were
also adjusted when the plant was under water
stress as calculated by the Molz-Remson model.
MODEL SIMULATIONS
Research Model
It is not possible to validate the research
model in its entirity, as the values for all the
model input parameters are not provided for in
the limited data base currently available. Thus,
only simulated results are presented here to
demonstrate the research-model capabilities in
describing the fate of nitrogen in the crop root
zone. The soil parameters used in the
simulations represent a well-drained loam soil,
while the plant parameters are for corn crop.
The simulations presented here commenced 34
days after planting and proceeded until 83 days
into the crop growth season. This seven-week
time period was chosen since it was the most
active in terms of crop demand for nitrogen.
13
-------
NITROGEN IN RETURN FLOWS
The initial soil-water content (0f) on the
34th day was assumed uniform at 0.10
cm 3/ cm 3 over the entire crop root zone (0-90
cm). The soil profile was also assumed devoid of
any mineral-N (NH4 + NO3), while the
organic-N distribution was described by:
0(z) = 50.0 [exp (-0.025z)]
[13]
which amounted to a total of 2863 pg organic-
NX cm 2 in the root zone. A solution containing
200 Mg N/ml of NH 4 NO 3 was applied at the
surface for a period of 4 hours, followed by water
application for another 8 hours (i.e., infiltration
for a total of 12 hours). The soil surface was
maintained saturated (h = 0) during infiltra-
tion. The total amounts of nitrogen and water
applied in this manner were equivalent to 83 kg
N/ha and 9 cm of water. Two additional
irrigations of 3.2 and 3.3 cm of water were
applied, respectively, on the 48th and 62nd day
of the growing season. Water losses due to
evaporation at the soil surface were ignored,
while crop transpiration demand of 0.3 cm/day
was assumed throughout the simulation period.
Nitrate losses due to denitrification were ig-
nored in the simulations presented here.
(NH
(N0)5
Figure 1. Soil nitrogen transformations considered
in the research model. The subscripted symbol k is a
first-order rate coefficient, while the subscripts e, s, i,
and g refer to exchangeable, solution, immobilized,
and gaseous phases, respectively. K is theFreundlich
distribution coefficient.
Solution concentrations of NO3-N and
NH4-N in the soil profile at selected times
following each irrigation are shown in Figures 4
and 5. Changes in concentrations of these
nitrogen species below the 60-cm depth were
very small; hence are not shown. The organic-N
distributions are also not presented here. The
position of the NO 3-N front immediately at the
end of the first irrigation (curve labeled 0.5 days
in Figure 4) is at 25 cm depth and can be
calculated by equation [lOb] given I = 9 cm and
a saturated water content (0f) of 0.36 cm3/cm-
in the wetted zone. The NH4-N front was
calculated at about 17 cm depth; this retarda-
tion is due to ion-exchange. During the two-day
period following the first irrigation, redistribu-
tion of soil water had caused additional move-
ment of both NO3-N and NH4-N pulses
(Figures 4 and 5), respectively, to a depth of 29
and 19 cm. The total amount of NH4-N in the
soil solution had decreased during the 14-day
period following irrigation (Figure 5) principal-
ly due to nitrification and plant uptake.
O Balasubramanian, 1974
D Cassel, 197!
O Ghuman et al., 1975
• Kirdo et al., 1973
A Warnck et al., 197!
e,/ef
Figure 2. The relationship between dsf/dwf and
6* i/' 0 f as calculated by Equation [ lOc] compared with
experimental data for various sources (Reproduced
from Rao et al., 1976, where references to the
literature cited in this figure can be found).
A total of 43.2 cm of water was extracted
from the root zone during a two-week period.
This extraction resulted in an average water
content of 0.21 cm3 /cm3 in the region where
NO3-N and NH 4 -N pulses resided. The second
and third irrigations of 3.2 cm and 3.3 cm,
respectively, were not sufficiently large to cause
significant movement of these solute pulses.
The total amounts of NH4-N and NO3-N
present in the root zone continued to decrease as
a result of transformations and plant uptake.
NH4-N concentrations in soil solution had
diminished to less than 4 ng N/ml and that of
NO 3-N were less than 10 ^g N/ml by the 83rd
day of the growing season (i.e., 21 days after the
third irrigation).
The total amounts of NO 3-N, NH4-N (sum
of solution and exchangeable phases), and
organic-N remaining in the root zone, as a
percent of that at the initiation of the
14
-------
simulations on the 34th day, are presented in
Figure 6. The losses of nitrogen shown here are
due only to transformations and plant uptake,
as there was no movement of soil water beyond
the root zone. Rapid loss of NH 4 -N is evident in
Figure 6. The production of NO3-N from
nitrification was greater than that absorbed by
the roots during the first week, giving rise to the
plateau in the early portion of the NO 3 -N curve
in Figure 6. However, the amount of
decreased rapidly after this time as
became the sole source of N for plant uptake.
There was also a small net loss of organic-N as
2100
O 50
/"
3 I-
^
4 Q.
Z
5 -
20
30
TIME
40 50 60
days
Figure 3. Comparison between measured (solid
circles) and predicted (solid lines) position of the
nonreactive solute front in a sandy soil at various
times. The soil was planted to millet crop (Reproduced
from Rao et al., 1976a).
NITROGEN IN THE ROOT ZONE
the amount of N mineralized exceeded that
immobilized during the 7-week simulation
period.
Cumulative amount of nitrogen removed by
the crop during 34-83 day growing period is
shown in Figure 7. The curve marked "demand"
represents the amount of N taken up by the
plants if maximum N-demand (QNmax) is
satisfied at all times (i.e., ideal growth). The
amount of nitrogen present in the crop root zone
was insufficient during the later part of the
simulation period (times greater than 8 days in
Figure 7) to meet the maximum demand, result-
ing in a significant deviation of the simulated
curve from the "ideal" curve. Such a nitrogen
deficit, when it occurs under real conditions,
would lead to decreased dry matter accumula-
tion and reduced yields.
Management Model
NaNagara et al. (1976) have performed field
experiments to measure nitrogen uptake by corn
during an entire crop growing season. In addi-
tion to measuring accumulation of N in the
plant, these authors also obtained data on root
length distributions, nitrate concentration and
water content profiles throughout the season.
These experimental data will be utilized in
validation of the simple management model
described earlier. NaNagara et al. (1976) have
compared their data with predictions from two
conceptual microscopic models of N-uptake
described by Phillips et al. (1976). The first
model (model I) considers the mass flow of
NO3-N Solution Concentration (/jgN/ml)
4O 6O 6O 10O O 2O 4O 6O 80 100 0
20
40
60
E
u
20
CL
UJ
Q
4O
O
00
601-
2 days
1slirngation
2ndirrigation
21 days
3rdirrigation
Figure 4. NO3-N solution concentrations in the soil profile, simulated by the research model at selected
times following each of the three irrigations.
15
-------
NITROGEN IN RETURN FLOWS
0
NH4-N Solution Cone. (jugN/ml)
0 2O 40 6O 8O 10O O 20 4O 6O 0 20 40
E
u
20
CL
LJ
Q
40
O
60L
0.5 days
2 days
1st irrigation
days
2 days
2nd irrigation
3rd irrigation
FigureS. NH 4-N solution concentrations in the soil profile, simulated by the research model, at selected
times following each of the three irrigations.
nitrate into roots with water (i.e., passive up-
take) as a result of water extraction by roots in
response to the transpiration demand. By know-
ing the amount of water transpired in a given
time period and the average nitrate-N concen-
tration in the soil solution in a given region of
the soil profile, the cumulative N-uptake was
estimated. Model II considers the microscopic
processes of nitrate transport to root surfaces by
diffusion and mass flow. Furthermore, rate of N-
uptake by roots was assumed to be directly
proportional to the nitrate concentration. Note
that neither model I or model II consider uptake
of the NH4-N species.
Nitrogen uptake values calculated from the
simple management model are compared in
Table 1 with measured values, as well as those
predicted from models I and II (by NaNagaraet
al., 1976). Reasonable agreement between data
and all three models (with widely different
conceptualizations of the processes) makes
acceptance or rejection of any of these models a
difficult task. Considering the simplicity of
model III, its close agreement with data is
encouraging. However, further testing of this
model with additional data is needed. When
only a gross description of the nitrogen
TABLE 1
A comparison of measured nitrogen uptake by corn
grown under field conditions and that predicted by
three simulation models. Measured data and
calculations using Models I and II are taken from
NaNagra et al. (1976), while calculations using
model III were presented by Rao et al. (1977a).
Growth Measured
Period N-uptake Calculated N-uptake (mg N/plant)
(days) mg N/plant Model I Model II Model III
Figure 6. Percent of applied nitrogen remaining
within the root zone during the simulated growth
season. The curves were based on the data presented
in Figures 4 and 5.
34-49
49-76
76-97
Total
34-97
% Error
1435
1593
974
4002
1097
1101
1496
3693
-7.7
1254
2000
1278
4533
+13.3
1928
1948
683
4559
+13.9
16
-------
NITROGEN IN THE ROOT ZONE
behavior in the crop root zone is required, this
simple model seems to hold promise.
2ooor
Simulation
( research model )
0 8 16 24 32 40 48
Days After Application
56
Figure 7. Comparison of the cumulative nitrogen
uptake curve simulated using the research model
with that when maximum uptake demand is satisfied
at all times during the growth season.
ACKNOWLEDGMENT
This research was supported in part by
funds from the U.S. Environmental Protection
Agency (Grant No. R-803607) and in part by
special funds from the Center for Environmen-
tal Programs of the Institute of Food and
Agricultural Sciences, University of Florida,
Gainesville, Fla.
REFERENCES
1. Barley, K. P. 1970. The configuration of
the root system in relation to nutrient uptake.
Adv. in Agron. 32:159-201.
2. Bartholomew, W. V. and Clark, F. E., ed.
1965. Soil Nitrogen. Agronomy Monograph No.
10, Am. Soc. Agron., Inc. Madison, Wisconsin,
615 pp.
3. Beek, J. and Frissell, M. M. 1973. Simula-
tion of nitrogen behavior in soils. Pudoc.
Wageningen, the Netherlands, 67 pp.
4. Davidson, J. M., Ou, L. T., and Rao,
P. S. C. 1976. Behavior of high pesticide concen-
trations in soil water systems, in Residual
Management by Land Disposal. Proc. of the
Hazardous Waste Research Symp. Tucson,
Arizona. EPA-60079-76-015, p. 235-242.
5. Dibb, D. W. and Welch, L. F. 1976. Corn
growth as affected by ammonium vs. nitrate
absorbed from soil. Agron. J. 68:89-94.
6. Duffy, J., Chung, C., Boast, C., and
Franklin, M. 1976. A simulation model of bio-
physiochemical transformations of nitrogen in
tile-drained corn belt soils. J. Environ. Qual.
4:477-486.
7. Dutt, G. R., Shaffer, M. J., and Moore, W.
J. 1972. Computer simulation model of dynamic
bio-physiochemical processes in soils. Univ. of
Arizona Tech. Bull. No. 196, 101 pp.
8. Frere, M. H., Onstad, C. A., and
Holtan, H. N. 1975. ACTMO, an agricultural
chemical transport model. U.S. Dept.
AgrL, ARS-H-3, 54 pp.
9. Hagin, J. and Amberger, A. 1974. Con-
tribution of fertilizers and manures to the N-
and P- load of waters: A computer simulation.
Final Report to the Deutsche Forschungs Ge-
meinschaft from Technion., Israel. 123 pp.
10. Hillel, D. and Talpaz, H. 1976. Simula-
tion of root growth and its effects on pattern of
soil water uptake by nonuniform root system.
Soil Sci. 121:307-312.
11. Marriott, F. H. C. and Nye, P. H. 1968.
The importance of mass flow in uptake of ions
by roots from soil. Trans. 9th Int. Cong. Soil Sci.,
Adelaide, Australia, 1:127-134.
12. Molz, F. J. and Remson, I. 1970. Ex-
traction-term models of soil moisture use by
transpiring plants. Water Resour. Res. 6:1346-
1356.
13. NaNagara, T., Phillips, R. E., and
Leggett, J. E. 1976. Diffusion and mass flow of
nitrate-nitrogen into corn roots grown under
field conditions. Agron. J. 68:67-72.
14. Newman, E. 1.1974. Root and soil water
relations, in The Plant Root and Its Environ-
ment, ed. E. W. Carson., Univ. Press of Virginia,
p. 363-440.
15. Nye, P. H. and Spiers, J. A. 1964. Si-
multaneous diffusion and mass flow to plant
roots. Trans. 8th Int. Cong. Soil Sci., Bucharest,
Rumania, 3:535-542.
17
-------
NITROGEN IN RETURN FLOWS
16. Olsen, S. R. and Kemper, W. D. 1968.
Movement of nutrients to plant roots. Adv. in
Agron. 20:91-151.
17. Passioura, J. B. and Frere, M. H. 1967.
Numerical analysis of the convection and diffu-
sion of solutes to roots. Aust. J. Soil Res. 5:149-
159.
18. Pearson, R. W. 1974. Significance of
rooting pattern to crop production and some
problems of root research, in The Plant Root and
Its Environment, ed. E. W. Carson, Univ. Press
of Virginia, Charlottesville, Virginia, p 247-270.
19. Phillips, R. E., NaNagara, T., Zartman,
R. E., and Leggett, J. E. 1976. Diffusion and
mass flow of nitrate-nitrogen to plant roots.
Agron. J. 68:63-66.
20. Rao, P. S. C., Davidson, J. M.£ and
Hammond, L. C. 1976a. Estimation of nonreac-
tive and reactive solute front locations in soils.
Residual Management by Land Disposal. Proc.
Hazardous Waste Research Symp. Tucson,
Arizona. EPA-600/9-76-015. p. 235-242.
21. Rao, P. S. C.,Selim,H. M., Davidson,
J. M., and Graetz, D. A. 1976b. Simulation of
transformations, ion-exchange, and transport
of selected nitrogen species in soils. Soil Crop
Sci. Soc. of Florida Proc. 35:161-164.
22. Rao, P. S. C., Davidson, J. M., and
Jessup, R. E. 1977a. A simple model for descrip-
tion of the fate on nitrogen in crop root zone.
Manuscript prepared for Agronomy Journal.
23. Rao, P. S. C., Jessup, R. E., and
Davidson, J. M. 1977b. A model for kinetics of
nitrogen transformations during leaching in
soils: Analytical Solutions. Unpublished
manuscript.
24. Ritchie, J. T. 1973. Influence of soil
water status and meteorological conditions on
evaportion from a corn canopy. Agron. J.
65:893-897.
25. Selim, H. M., Hammond, L. C., and
Mansell, R. S. 1977a. Soil water movement and
uptake by plants during water infiltration and
redistribution. Vol. 36. Soil Crop Sci. Soc. of
Florida Proc.
26. Selim, H. M., Davidson, J. M., Rao,
P. S. C.,and Graetz, D. A. 1977b. Nitrogen
transformations and transport during transient
unsaturated water flow in soils. Water Resour.
Res. (submitted).
27. Tanji, K. K., and Gupta, S. K. 1977.
Computer simulation modeling for nitrogen in
irrigated crop lands.
28. Watts, D. G. 1975. A soil-water-nitrogen-
plant model for irrigated corn on coarse textured
soils. Ph.D. Dissertation, Utah State Univ.
187p Xerox Univ. Microfilms no. 76-6253; Ann
Arbor, Michigan.
18
-------
Nitrate Movement in Clay
Soils and Methods
of Pollution Control
ALAN R. SWOBODA
Soil and Crop Sciences Department,
Texas Agricultural Experiment Station,
Texas A&M University, College Station, Texas
ABSTRACT
Movement of nitrates were measured in
soils to determine quantitative amounts of
losses. Leaching losses in lysimeters ranged
from 0.04 to 6% depending on nitrogen source
and climatic conditions. Nitrate concentrations
in shallow wells within watersheds indicated
that nitrates from applied fertilizer was
leaching into the shallow groundwaters. Con-
centrations as high as 60 ppm NO3 -N were
found in some wells after the initial rains
following fertilization. Runoff from the
watershed contained very little nitrate and
amounted to less than 4% of the fertilizer
applied.
A descriptive model of nitrate movement in
clay soils is presented. Methods of reducing
nitrate leaching in soils are discussed. Field
studies indicated that inhibiting nitrification
was an effective means of reducing nitrate
movement from ammonium fertilizers. Slow re-
lease sulfur coated urea also reduced nitrate
movement when applied in the fall or winter.
Delaying fertilization until planting was also
an effective means of reducing nitrate losses
without reducing yields.
INTRODUCTION
Movement of nitrate is generally considered
to be more of a problem in light textured sandy
soils. However, it should not be misunderstood
that nitrate movement is not a serious problem
in clay soils. Nitrates are very soluble in water
and whenever water moves through a soil,
nitrate leaching can occur. In fact, Thomas and
Swoboda (1969) have reported anion movement
in clay soils as much as 5 times faster than
would be predicted if the water moved through
the soil as "piston-type" flow.
Ritchie et al. (1972) have shown that the
movement of water through a clay soil which
contains approximately 60% clay can be very
significant. Under field conditions, they mea-
sured hydraulic conductivity values of 2.5
cm/day. Using fluorescein tagged water they
illustrated that the majority of water movement
was not through the entire soil matrix but rather
around structural units and along cleavage
planes.
In connection with this study it was also
shown (Kissel et al., 1973) that Cl moved with
the water and by-passed the majority of the soil
structural units. They reported Cl movement to
be 9 times faster than what would be anticipated
under "piston-type" water movement. It is evi-
dent from these studies that water and anions
do move through clay soils in significant quan-
tities.
The importance of nitrate leaching as it
relates to fertilizer losses and ground water
pollution and the development of methods of
reducing nitrate movement were objectives of
this study.
Water Movement
The assumption is sometimes made that the
water content of a soil must increase as water
moves through the soil in much the same way as
a water front moving through a dry soil.
However, since water in clay soils in under a
high capillary tension, water movement is
usually rather fast. Following an irrigation or
rainfall the water content of the soil surface will
increase considerably since initial infiltration is
rapid. However, within a few hours after in-
filtration is complete the moisture content of the
soil surface will decrease again as the water
19
-------
NITROGEN IN RETURN FLOWS
moves through the soil. If the soil was at or near
field capacity prior to the addition of the water,
the moisture content of the soil below about
45 cm will change only slightly as water move-
ment occurs. Even under ponded conditions the
water content of the soil below 45 cm will
seldom attain a saturated condition. Figure 1
illustrates the change in water content of Miller
clay prior to and after two irrigations. The
initial moisture content of the soil was a little
over 5 cm of water per 15 cm of soil depth. Two
days after irrigation with 9.7 cm of water,
moisture readings indicated an increase in
moisture content of about 0.5 cm H 2 O in the top
15 cm and less than 0.2 cm 715 cm in the next
60 cm. This moisture data indicates that over
8 cm of water had moved through the 165 cm
soil profile in only two days. The second irriga-
tion of 8.9 cm increased the moisture content
slightly after 2 days, but again 8 cm of water
appears to have moved through the entire
profile without changing the moisture content
of the soil below 60 cm.
It is evident from this data that con-
siderable quantities of water do move through
some clay soils. Nitrate leaching can be of
significance in these soils when excess nitrates
are present.
MOISTURE CONTENT
(cm HgO/15 cm)
3.5 4.0 4.5 5.0 5.5 6.0
E
o
Q.
UJ
O
•II-
0
15
30
45
60
75
90
105
120
135
150
165
Figure 1. Moisture content of a Miller clay prior to
and 2 days following 2 irrigations of 9.7 cm and 8.9
cm.
Nitrate Leaching in Clay Soils
The results of a study to illustrate nitrate
leaching as a result of improper fertilization of a
coastal bermuda grass pasture are shown in
Figure 2.This Houston Black clay field was
fertilized with 150 kg-N/ha as NH4NO3 on
October 22, after growth had essentially
stopped. During the first 2 weeks of November
the field received 3 small rain storms totaling
5.7 cm. On November 21, one month after
fertilization, the soil profile was sampled for
nitrates. Only limited leaching had occurred
(Figure 2) due to the low moisture content of the
soil surface prior to the rains. Eighty percent of
the applied nitrogen could be accounted for as
nitrates (NO 3) in the top 45 cm of soil. In late
November and early December the field re-
ceived 3 additional rains, each over 2.5 cm and
totaling 9.5 cm. The soil profile was sampled
again on December 7, 16 days after the last
sampling, and considerable leaching had oc-
curred. Only 51% of the applied nitrogen could
be accounted for in the 180 cm soil profile. The
concentration of NO sin the top 45 cm of the soil
had been reduced considerably by the ad-
ppm N03-N
0 5 10 15 20 25 30
4-2-69
(CHECK)
Figure 2. Concentration of NO 3 -N in soil profile of
Houston Black clay at various times after fertilizing
with 150 kg-N/ha as NH 4 NO 3 on October 22,1968.
20
-------
NITRATE MOVEMENT
ditional rainfall. In the spring very little
available nitrogen as NO 3 for plant growth was
detected in the entire profile sampled. That is
not to imply that the unaccounted for NO 3 had
leached through the entire 180 cm soil profile. It
is believed that a large part of the unaccounted
nitrogen was immobilized in the soil by micro-
organisms. Recent unpublished work (Kissel,
1976) has shown considerable immobilization
occurs in these soils.
The movement of NO 3 through soils causes
considerable concern with respect to possible
contamination of shallow water supplies. Water
which has a NO 3 -N concentration as high as
10 ppm is generally not considered acceptable
for domestic use.
Nitrates in Underground Aquifers
George and Hastings (1951) reported that
about 3000 of a total of 20,000 water wells
TABLE 1
Concentration of NO 3 -N in shallow wells in
grassland watershed at Riesel, Texas in
1974 and 1975.
Date
10-16-73
3- 5-74
4- 1-74*
5- 7-74
6- 1-74
6-20-74
9- 6-74
9-13-74
11- 8-74*
11-20-74
11-27-74
12-13-74
1- 2-75
2- 5-75
2-25-75
2-27-75
3-16-75*
4-14-75
5- 1-75
5- 7-75*
5-27-75
6-11-75
7- 3-75
7-28-75
1
0.1
d
0.5
9.7
0.0
0.1
2.3
0.4
14.4
6.0
4.5
2.8
1.7
0.3
0.0
0.2
0.6
0.8
6.0
5.3
6.0
33.0
20.5
4.0
Well
2 3
1.0
2.1
61.0
20.0
d
d
16.5
d
18.8
d
5.4
6.2
8.0
2.7
d
2.8
3.8
d
11.2
d
10.2
d
d
d
0.2
0.3
d
0.2
d
d
d
d
22.6
d
d
7.1
12.0
3.4
d
d
d
d
9.9
d
11.2
d
d
d
4
0.2
d
d
d
d
d
d
d
2.8
d
d
1.6
1.3
0.1
1.2
1.8
1.0
d
5.7
d
3.7
0.4
0.2
d
*90 kh/ha of nitrogen as NH 4 NO 3 was applied on
March 5 and September 27,1974, and on March 16,
and May 7, 1975.
d-well was dry
checked in Texas prior to 1948 contained over
4.5 ppm NOs-N with many being over
400 ppm. These samples were taken prior to the
widespread use of commerical fertilizers. They
indicated that although most of the high nitrate
wells were less than 200 feet deep, their oc-
curence appears to be unrelated to rainfall,
geography, or cultivation. Although it appears
that commercial fertilizer nitrogen is not a
major cause of nitrate pollution of shallow
aquifers, there are some instances where
leaching can contaminate shallow water
supplies.
A study was initiated to measure the move-
ment of nitrates with runoff water from a
watershed and leaching within the watershed.
Water samples were collected and runoff mea-
sured from a 7.7 ha grassland watershed. Water
samples were periodically obtained from 4
shallow wells installed in the watershed. The
wells were placed at a depth to intercept water
moving along the semi-impermeable calcareous
marl layer underlying the Houston Black clay.
The wells were generally between 1.8 and 2.4
meters deep. Nitrogen as NH 4 NO 3 was applied
twice each year at the rate of 90 kg-N/ha.
Runoff from the watershed which
amounted to 31 cm in 1974 and 15.6 in 1975,
contained very little nitrates. Generally, the
concentrations in the runoff during the 2 year
study were less than 1 ppm NO3-N, although
one storm following a fertilizer application of 90
kg-N/ha produced runoff which had concentra-
tions as high as 7.2 ppm NO 3-N. The total
amount of NO 3-N which was carried off the wa-
tershed in 1974 was less than 1 kg/ha, and in
1975 the amount lost was 6.5 kg/ha.
Of the 4 wells installed in the watershed,
only one contained water at all times during the
two year study. The other 3 wells only contained
water intermittently following rainy periods.
The concentration of nitrates were generally
very low in all wells. The only exception was
after the initial leaching rain following each
application of fertilizer nitrogen. Table 1 shows
the concentrations in the wells at various times
during the 2-year study. It is readily evident
that fertilization of the pasture did increase the
NO 3 concentration in the wells. The highest
concentration observed was 61 ppm NO 3-N in
Well 2 on April 1, 1974. However, all 4 wells did
show indications of nitrates leaching.
Well 1, which always contained water and
appeared to intercept a dependable water sup-
21
-------
NITROGEN IN RETURN FLOWS
ply for domestic use, contained NO 3 -N levels
exceeding 10 ppm during 2 periods oi the study;
these were in November of 1974 and June and
July of 1975. Both of these periods closely fol-
lowed applications of fertilizer nitrogen. It is
important to note that the concentrations
decreased rapidly after each of these periods
indicating dilution within the aquifer was oc-
curring.
It appears that leaching of fertilizer
nitrogen into these shallow aquifers can and
does occur. However, the quantity of nitrate
which is leached appears to be small and
dilution within the aquifer can reduce the con-
centrations within a short period of time.
Methods of Controlling Nitrate Leaching
A study was initiated in the fall of 1973 to
determine the effectiveness of several systems
for reducing nitrate movement in clay soils.
Since it is well documented that nitrates can
readily move through soils which do not have a
significant anion adsorptive capacity, the basic
premise of this study was to devise methods of
maintaining fertilizer nitrogen in the am-
monium form (NH 4) as long as possible. The
ammonium form of nitrogen can be adsorbed by
the negatively charged sites on the clay and
restrict its movement through the soil.
TABLE 2
Total NO3-N and NH4-N in 120 cm profiles of
Houston Black clay fertilized with different sources
of nitrogen and treated with N-Serve and sampled
January 21, 1975.
Treatment*
Soluble N
SCU
Soluble N -
N-Serve
SCU - N-Serve
Oct
XO3-N
433
221
127
160
. 78
A7/4-.V
0
26
88
66
Date of Applicati
Not: 19
MM -A* AT/4 -A'
362
316
117
257
kg-N ha
18
12
52
98
Dec. 20
NO3-X
179
80
77
112
NH4-N
154
78
327
62
Re-
duc-
tion
01 NO3
%
—
37
67
46
*Soluble N are average of (NH 4) 2 SO 4 and urea
treatments, CSU are average of SCU-20 and
SCU-30 treatments.
Three basic techniques were investigated
for maintaining fertilizer nitrogen in the am-
monium form until plant utilization. The first
technique consisted of treating ammonium
forms of fertilizers with a specific microbicide
which inhibits the nitrification process in soils.
This method would delay the conversion of
NH 4 to NO 3. The second technique was the use
of slow release or slowly soluble nitrogen
sources which would limit the amount of NH 4
and NO 3 which would be present in the soil at
any time. The third system investigated was the
timing of nitrogen application with regard to
plant utilization.
These three systems were investigated at
two locations on two different soils and for a
period of two years. One soil was a silt loam
(Norwood, Typic Udifluvent) and the other a
clay (Houston Black, Udic Pellustert). The field
plots on Norwood silt loam were located in the
flood plain of the Brazos River near College
Station, Texas and the field plots on Houston
Black clay were located at the Blackland
Research Center near Temple, Texas.
TABLE 3
Total NO3-N and NH4-N in 120 cm profiles of
Houston Black clay on June 6, 1975 after fertilizing
with soluble sources and slowly soluble sources of
nitrogen with and without treatment with N-Serve.
Star 2s
AW .\'H4
Soluble N
SCU
Soluble N -
N-Serve
SCU - N-Serve
*Soluble N are average of (NH4)2SC>4 and urea
treatments, SCU are average of SCU-20 and
SCU-30 treatments.
The treatments at each location consisted of
two readily soluble nitrogen sources (am-
monium sulfate and urea) and two slowly solu-
ble nitrogen sources, both sulfur coated ureas
(SCU-10 and SCU-30). The numerical values
after each refer to the percentage of nitrogen
soluble in hot water (38° C) during 7 days of
incubation. Each nitrogen treatment was also
applied with a nitrification inhibitor, N-Serve
[2-chloro-6 (trichloromethyl-pyridine)]. The in-
hibitor was applied directly to the fertilizer at a
rate equivalent to 1% of the total nitrogen. Each
fertilizer was applied at a rate of 134 kg-N/ha
(120 Ibs-N/ac) in a band 12-18 cm below the
furrow. Fertilizers were applied each month
starting in October and continuing until plant-
ing time in March. Each treatment was repli-
cated three times. Grain Sorghum (Sorghum
-------
NITRATE MOVEMENT
vulgare) was planted each year on both soils as
the indicator crop. Soil core samples were taken
through the placement band to a depth of 120 or
150 cm twice each year to monitor nitrate move-
ment in the soils.
There was no significant difference
between the ammonium sulfate and urea
sources of nitrogen as to their effect on NO 3
movement and therefore, will be discussed
together and referred to as Soluble N. There was
a little difference between the two sulfur-coated
ureas with the more soluble form (SCU-30)
resulting in more NO 3 movement. However,
both sources were similar enough so that they
can be discussed as one source.
The rate of nitrification in the Houston
Black clay soil as affected by source of nitrogen,
treatment with N-Serve, and time of applying
nitrogen is shown in Tables 2 and 3. Soil
samples taken during the Winter (Jan. 21,1975)
had very little NH4 remaining in the soil from
the soluble sources of nitrogen except for the
December application. Since the December
application was made only one month prior to
sampling, it would be expected that con-
siderable NH 4would remain in the soil. Even
though the nitrogen was in the soil only a short
time, and with the cool winter temperatures,
over 50% of the nitrogen had been nitrified
(Table 2). It is evident that nitrification is very
active in the Houston Black soil.
The use of sulfur coated urea (SCU) reduced
the amount of NO 3 present in the soil at any
time during the 3 months by 37%. The lower
total amount of nitrogen reported for the SCU
sources was due to unhydrolyzed sulfur-coated
urea prills remaining in the soil.
The nitrification inhibitor was very effec-
tive during the winter months in reducing the
amount of NO 3 present in the soil. Even when
the treated soluble sources were applied 3
months prior to sampling 40% of the nitrogen
present remained in the less leachable NH4
form. The amount of NO 3 present in the soil at
any time during the 3 months was reduced 67%
by the N-Serve treatment (Table 2).
The data in Table 2 depict a significant
decrease in the total amount of inorganic
nitrogen present in soil from the N-Serve treat-
ment. This effect was evident at both research
sites and during each year of the study.
Jansson (1958) has reported that NH4 is
much more readily utilized by microorganisms
than is NO 3. The use of a nitrification inhibitor
with an NH4 fertilizer maintains a higher
concentration of NH4 in the soil for a longer
period of time, thereby enabling more nitrogen
to be immobilized. Thus, the decrease in in-
organic nitrogen shown in Table 2 for the N-
Serve treatments.
Treatment of the SCU's with N-Serve also
reduced the amount of NO 3 present in the soil at
any one time, although it was not as effective as
the treatment of the Soluble N source.
The summer sampling on June 6,1975 of the
Houston Black clay typifies the effects of fer-
tilizer timing and nitrification. Table 3 il-
lustrates that nitrification of the soluble NH 4
sources was very rapid and complete. No NH 4
remained in the soil in June from any of the
dates of application of the soluble N sources. As
in the winter samples the slowly soluble sources
(SCU) reduced the amount of leachable NO 3 in
the soil at any given time. There was an average
of 41% less NO 3 in the soil from the SCU
treatments as compared to the ammonium sul-
fate and urea treatments. Undissolved sulfur
coated prills and some NH 4 still remained in
the soil on June 6 even when the SCU was
applied the previous October.
TABLE 4
Amount of NO 3-N leached below 60 cm but above
120 cm in Houston Black clay on January 21, 1975
as influenced by nitrogen source, nitrification
inhibitor and time of application.
Date of Application
Treatment* Oct. 18 Nov. 19 Dec. 20 Reduced
Leaching
kg-N/ha %
Soluble N
SCU
Soluble N
+ N-Serve
SCU +
N-Serve
128f
88
60
68
92
64
68
72
48
36
40
44
—
30
38
32
*Soluble N are average of (NH4)2SC>4 and urea
treatments, SCU are average of SCU-20 and SCU-30
treatments.
tNC>3 content of control plots have been subtracted
from these values.
Treatment of the nitrogen sources with the
nitrification inhibitor also reduced the amount
of NO3 present in the soil in June for all dates of
application. There was a considerable amount
23
-------
NITROGEN IN RETURN FLOWS
of NH4 present in the soil even when the
nitrogen was applied in October, indicating the
N-Serve still had some inhibitory effect 7
months after application. The effect of N-Serve
in reducing nitrification was reduced somewhat
during the warmer part of the year as indicated
by the 22% average reduction in NO 3 content in
June as compared to the 67% average reduction
found for the soluble source in January (Table
2). There was very little difference between the
two soils at the different locations in relation to
the different treatments.
It is apparent that the use of slowly soluble
nitrogen sources and the treatment of fertilizer
with N-Serve were effective in reducing the
amount of leachable NO 3 present in the soil at
any time. Therefore it would seem that these
practices would also be effective in reducing the
quantity of nitrate leaching through soils.
The amount of NO 3 found below 60 cm in
the sampled profiles was found to be a good
indicator of nitrate leaching. Time of applica-
tion had a great influence on the amount of NO 3
found below 60 cm in both the silt loam and clay
soil. The winter samples on January 21, 1975
indicated that considerable movement of NO 3
had already occurred into the zone below 60 cm
for all treatments (Table 4). Since the mineraliz-
ed NO 3 found in the control plots was taken into
account to obtain the values in Table 4, the
values shown are a result of the fertilizer
applied. The earlier applications of nitrogen
caused considerable more NO 3 to move below
60 cm. Even the December application, which
was made only 30 days prior to sampling,
resulted in NO 3 movement below 60 cm.
The slow release sources and treatment
with N-Serve reduced the amount of NO 3 move-
ment, but they did not prevent leaching even
when applied only 30 days prior to sampling.
The NO 3 found below 60 cm represents a
significant portion of the total NO 3 in the
profile. Generally, about 35% of the NO 3 in the
profile was found below 60 cm at the January
sampling.
In June of 1975 the effect of fertilizer timing
on NO 3 movement was more evident than in
January (Table 5). Fertilization with the soluble
sources of nitrogen in the fall resulted in much
more NO 3 being present below 60 cm than
when fertilizer was applied in the spring.
However, this effect was not near as evident for
the other treatments, although the amounts of
NO 3 found below 60 cm was reduced by 42-53%
by the use of SCU and N-Serve treatment of
TABLE 5
Amount of NO 3-N leached below 60 cm but above
120 cm in Houston Black clay on June 6, 1975 as
influenced by nitrogen source, nitrification
inhibitor, and time of application.
Date of Application
Reduced
Leach-
Treatment? Oct. 18 Nov. 19 Dec. 20 Feb. 21 Mar. 25 ing
Soluble N
SCU
Soluble N +
N-Serve
SCU + N-Serve
171t
48
58
32
78
51
57
61
62 60
29
48
24
34
30
28
48
34
49
30
53
42
58
•Soluble N are average of (NH4)2SO4 and urea treatments, SCU are
average of SCU-20 and SCU-30 treatments.
tNO 3 content of control plots have been subtracted from these values.
ammonium sulfate and urea. The percentage of
total NO 3 in the profile found below 60 cm in
June ranged from as high as 63% from the
October application to as low as 16% in the
March application. It is very likely that some
NO 3 had already moved below the depth of
sampling in June since samples taken between
120-150 cm contained considerable amounts of
nitrates.
The use of sulfur coated urea and treatment
of urea and ammonium sulfate with a nitrifica-
tion inhibitor appear to significantly reduce the
amount of NO 3 leaching in both the clay and
silt loam soils. However, these treatments did
not have any affect on the yield of grain
sorghum. Although, it appeared that con-
siderable leaching did occur, probably less than
15% of the applied nitrogen actually moved
below the rooting zone of the grain sorghum.
The 134 kg-N/ha applied appears to be more
than required for maximum yields even when
considering 15% loss by leaching.
SUMMARY
Nitrate movement in clay soils can be of
significant magnitude. Pollution of un-
derground water supplies is possible when
nitrogen fertilizers are improperly used and
excessive rates are applied.
The use of slow release nitrogen sources,
and the treatment of readily soluble nitrogen
sources with an effective nitrification inhibitor,
can reduce the amount of nitrate movement
through soils. Applying nitrogen fertilizers to
the soil as near to the time when plants can
utilize it will also reduce the chances for pollu-
tion of underground water supplies.
24
-------
NITRATE MOVEMENT
REFERENCES
1. George, W. O., and Hasting, W. W. 1951.
Nitrate in the ground water of Texas. Am.
Geophys. Union 32:450.
2. Jansson, S. L. 1958. Tracer studies on
nitrogen transformations in soil with special
attention to mineralization-immobilization re-
lationships. Kgl. Lanthbruks-Hogskol Ann.
24:101-361.
3. Kissel, D. E. 1976. Immobilization of in-
organic nitrogen in clay soils. Unpublished
data.
4. Kissel, D. E., Ritchie, J. T., and
Burnett, E. 1973. Chloride movement in un-
disturbed swelling clay soil. Soil Sci. Soc. Pm.
Proc. 37:21-24.
5. Thomas, G. W., and Swoboda, A. R.
1969. Anion exclusion effects on chloride move-
ment in soils. Soil Sci. 110:163-166.
6. Ritchie, J. T., Kissel, D. E., and Burnett,
E. 1972. Water movement in undisturbed swell-
ing clay soil. Soil Sci. Soc. Am. Proc. 36:874-879.
25
-------
Effect of Three
Irrigation Systems
on Distribution of Fertilizer
Nitrate Nitrogen in
A. B. ONKEN, C. W. WENDT, O. C. WILKE, R. S. HARGROVE
WALTER BAUSCH and LARRY BARNES
Texas Agricultural Experiment Station, The Texas A&M
University System, Lubbock, Texas
ABSTRACT
Sweet corn (Zea Mays L.) was grown two
years on a Miles loamy fine sand (Udic Pal-
eustalfs) fertilized with band applied I5N
enriched sodium nitrate. Two plots, 6x6m, were
established under each of the three irrigation
systems -sprinkler, furrow and subirrigation. A
starter band of fertilizer was applied 7.6 x 7.6 cm
from the seed with the rest sidedressed 25 cm
either side of the center of the seedbed.Total
nitrogen applied was 124 kg/ha in 1973 and 105
kg/ha in 1974. In 1974, each of the six plots that
were established in 1973 were divided in half,
one half receiving 15N enriched fertilizer as
sodium nitrate, and the other half receiving the
same amount ofunenriched sodium nitrate. Soil
samples were taken periodically from furrow to
furrow in lateral distance increments of 25 cm
and through the starter fertilizer band to a
depth of 5.2 m in 30 cm increments. Plant
samples were obtained at the end of each
growing season. Soil and plant samples were
analyzed using standard procedures.
Fertilizer nitrogen moved differently under
the three irrigation systems. When sprinkler-
irrigated, fertilizer bands tended to maintain
their integrity during downward movement.
Under furrow irrigation, fertilizer bands tended
to merge in the center of the bed and move
downward. Under subirrigation, fertilizer tend-
ed to move down under the furrows. At the end of
two crop years, greatest depth movement of
fertilizer nitrogen was under sprinkler irriga-
tion and least under subirrigation. Additional-
ly, a greater amount of fertilizer nitrogen was
found in the organic nitrogen fraction under the
sprinkler system than under the furrow system
with the subirrigation system having the least.
The amount of nitrate-nitrogen from fer-
tilizer to a depth of 5.2 m was found to be in the
order: sprinkler > furrow > subirrigation. At the
end of two crop years, it was possible to account
for 92.6, 86.1, and 50.5 percent of the fertilizer
nitrogen applied to the sprinkler, furrow, and
subirrigation systems, respectively. The largest
differences between irrigation systems, in ac-
counting for applied fertilizer nitrogen during
the two years, occurred in the soil.
The amount of fertilizer nitrate remaining
in the upper portion of the profile for the longest
period of time was subirrigation > furrow >
sprinkler. The retention of nitrate in the upper
portion of the profile in conjunction with high
water potentials resulted in denitrification. The
amount of denitrification was apparently
greatest for subirrigation and least for sprinkler
irrigation. The quality of irrigation return flow,
relative to nitrates, to underground water was in
the order: subirrigation > furrow > sprinkler.
INTRODUCTION
Application of fertilizer nitrogen to sandy
soils low in organic matter is necessary for
sustained high levels of production. However,
movement of fertilizer nitrate-nitrogen through
porous soils into shallow underground water
strata has been considered to be a potential
pollution hazard, particularly under irrigation.
The purpose of the study reported here was to
determine the effects of three irrigation
27
-------
NITROGEN IN RETURN FLOWS
systems — sprinkler, furrow and subirrigation
— on the movement of fertilizer nitrate-nitrogen
in the presence of a growing crop.
METHODS AND MATERIALS
Sweet corn (Zea Mays L.) was grown two
years on a Miles loamy fine sand (Udic
Paleustalfs) fertilized with band applied 15N
enriched sodium nitrate. Two plots, 6 x 6 m, were
established under each of the three irrigation
systems-sprinkler, furrow and subirrigation. A
starter band of fertilizer was applied 7.6 x 7.6 cm
from the seed with the rest sidedressed 25 cm
either side of the center of the seedbed. Total
nitrogen applied was 124 kg/ha in 1973 and 105
kg/ha in 1974, Table 1. In 1974, each of the six
plots that were established in 1973 were divided
in half, one half receiving ^N enriched fer-
tilizer as sodium nitrate and the other half
receiving the same amount of unenriched
sodium nitrate. The bands were flagged and soil
samples were taken periodically from furrow to
furrow in lateral distance increments of 25 cm
and through the starter fertilizer band to a
depth of 5.2 m in 30 cm increments. The plant
samples were obtained at the end of each
growing season. Soil and plant samples were
analyzed using standard procedures (1, 2).
TABLE 1
15,
Dates and amounts of nitrogen as * N enriched
sodium nitrate applied to experimental plots.
Date
Starter
Sidedress
May 5, 1973
June 4, 1973
March 24, 1974
June 6, 1974
Nitrogen (kg/ha)
32.3
91.3
22.4
82.4
Irrigation water amounts applied for the
three systems for both years are shown in Table
2. The sprinkler and subirrigation systems
received approximately the same amount of
irrigation water both years. Somewhat more
water was applied to the furrow system due to
the fact that in the loamy fine sand soil, 75 mm
of water had to be applied in order to obtain
uniform distribution. Rainfall date are shown in
Table 3. There was a large variation in the
amount of rainfall received during the crop
growing seasons. A substantial amount of rain
was received between harvest of the 1973 crop
and planting of the 1974 crop.
TABLE 2
Dates and amounts of irrigation water applied to
sprinkler, furrow and sub-irrigated plots.
Irrigation
System
Year
1973
1974
Sprinkler
Furrow
Subirrigation
waiei
256
381
256
: Viuiii;
249
310
255
TABLE 3
Dates and amounts of rainfall received from
planting in 1973 to last sampling in 1974.
Time of Year
Year
1973
1974
Before Planting
Growing Season 104
After Harvest 275
-Rainfall (mm) •
15
193
RESULTS AND DISCUSSION
Distribution of fertilizer nitrate for 1973 in
the sprinkler-irrigated plots is shown in Table 4.
On the first sampling date of May 31, the starter
band of fertilizer was detected in the top 30 cm
sampling increment. Subsequent samplings
throughout the growing season show that the
fertilizer nitrate tended to move down through
the soil in distinct bands. Also, it will be noted
that after the first sampling date that little or no
fertilizer nitrate was detected in the top 30 cm.
Maximum depths to which fertilizer nitrate was
detected were in the 90 to 120 cm sampling
increment. Fertilizer nitrate concentration
decreased with time and tended to move down in
the profile. The reduction in the amounts
detected was probably due to leaching, plant
uptake, immobilization and denitrification.
A somewhat different distribution pattern
of fertilizer nitrate can be seen for the furrow-
irrigated plots, Table 5. Again, the starter band
was detected at the first sampling date in the top
30 cm of soil. Concentrations tended to decrease
with time and the fertilizer nitrate moved down
in the profile with the detectable amounts being
found in the 120 to 150 cm sampling increment
which was somewhat deeper than for the
sprinkler-irrigated plots. Also, the fertilizer
28
-------
FERTILIZER NITROGEN IN SOIL
TABLE 4
Average fertilizer nitrate-N concentrations for three
sampling dates by depth and lateral sampling
location for two spinkler irrigated plots fertilized with
1 ^ ~ ~ enriched sodium nitrate, 1973.
Samp. Depth
(cm)
0-30
30-60
60-90
0-30
30-60
60-90
90-120
120-150
0-30
30-60
60-90
90-120
120-150
150-180
180-210
Lateral Sampling Distance
0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
25
XT;J
iNll
M<
51
58
;rate-N (ppn
ly 31, 1973
0.0 0.0
0.0 0.0
0.0 0.4
June 22
2.7
59.8
24.7
0.0
3.7
6.2
3.3 0.4
0.1 1.1
July 26
0.0
0.0
1.1
0.2
0.0
0.0
0.0
0.0
0.0
1.7
0.2
0.0
0.0
0.0
29.4
6.1
0.8
,1973
76
n\
1)
0.0
0.0
0.0
0.0 0.0
3.8 21.0
8.5
2J
0.4
, 1973
0.0
0.0
8.4
0.0
0.1
0.0
0.0
7.8
0.2
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
(cm)
102
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.2
0.0
0.0
0.0
0.0
0.0
TABLE 5
Average fertilizer nitrate-N concentrations for three
sampling dates by depth and lateral sampling
location for two furrow irrigated plots fertilized with
enriched sodium nitrate, 1973.
Samp. Depth
(cm)
Lateral Sampling Distance
0
25
— Witr
51
58
ate-N (ppm
May 31 :
0-30
30-60
60-90
0.0
0.0
0.0
0.0
0.0
0.0
3.4
0.0
0.0
June 22
0-30
30-60
60-90
90-120
120-150
0.0
2.0
2.0
0.0
0.0
0.3
5.0
2.4
1.4
0.0
3.8
7.5
3.8
3.0
0.0
,1973
28.3
L8
LO
, 1973
16.0
19.0
2.0
3.8
0.0
76
\
0.0
0.0
0.0
2.2
1.3
0.8
0.1
0.0
(cm)
102
0.0
0.0
0.0
1.7
1.8
2.0
0.8
0.0
July, 28 1973
0-30
30-60
60-90
90-120
120-150
150-180
180-210
210-240
0.0
0.1
1.2
4£
0.2
0.0
0.0
0.0
0.0
0.0
0.2
2.2
3.3
0.5
0.0
0.0
0.2
0.0
0.3
LS
L6
0.4
0.0
0.0
0.0
0.0
6.4
9.5
4.2
0.5
0.0
0.0
0.2
0.0
2,2
2.6
0.1
0.0
0.1
0.0
0.2
5.8
3.0
0.4
0.0
0.2
0.0
0.0
nitrate tended to move toward the center of the
bed and to move downward at this point. It will
be noted that substantial quantities of fertilizer
nitrate were detected in the top 30 cm of the soil
through the third sampling date.
Fertilizer nitrate distribution for the sub-
irrigated plots for 1973 is shown in Table 6. This
distribution pattern is considerably different
from the other two systems in that the fertilizer
nitrate tended to move outward from the center
of the bed and toward the furrows. This was a
consequence of the fertilizer bands being placed
above and 25 cm either side of the subirrigation
lateral. Thus, movement of irrigation water was
always upward and toward the furrows with
respect to the fertilizer bands. It will be noted
that detectable quantities of fertilizer nitrate
were found in the top 30 cm of the soil through
the third sampling date and that it did not move
in detectable quantities below 90 cm.
Distribution of fertilizer nitrate for the
sprinkler-irrigated plots for 1974 is shown in
Table 7. At the first sampling date on May 22,
the fertilizer nitrate from the starter band was
detected to a depth of 90 cm. It was probably
located deeper in the soil in 1974 than in 1973
due to the greater amount of rainfall that fell
between its application in 1974 and the first
sampling date than for the same period in 1973.
Additionally, some quantities of fertilizer
nitrate were found in a depth range of 180 to 300
cm and was apparently that applied in 1973. At
this sampling date, the sidedressed bands were
detected in the upper 30 cm of soil along with
some additional downward movement of the
starter band. Also, fertilizer applied in 1973 had
moved downward an additional 30 cm. By the
last sampling date on July 16, further move-
ment downward of the 1974 sidedressing and
starter fertilizer bands can be seen in the upper
portion of the soil profile and of the 1973
fertilizer nitrate in the lower portions of the
profile. It is interesting to note the zone of
extremely low concentrations from 150 to 240
cm that resulted in a distinct separation
between fertilizer nitrate applied in 1973 and
that applied in 1974.
29
-------
NITROGEN IN RETURN FLOWS
Concentrations of fertilizer nitrate for the
samples taken from the furrow-irrigated plots in
1974 are shown in Table 8. Again, the starter
band of fertilizer was detected at the first samp-
ling date to the depth increment of 60 to 90 cm
probably due to additional amounts of rainfall
received between application and first sampling
in 1974 as opposed to 1973. Additional amounts
of fertilizer nitrate were detected at lower depths
(90 to 210 cm); somewhat higher in the profile
than for the sprinkler plots. At the second
sampling date on June 20, the two sidedressed
band applications were detected in the top 30 cm
of soil with some evidence of movement
downward and depletion of the starter band
fertilizer in the upper portions of the profile. The
distinction between fertilizer applied in 1973
and 1974 was not as clear for the furrow-
irrigated plots as for the sprinkler-irrigated
plots. There was additional movement
downward of fertilizer nitrate as determined by
the soil sampling on July 17. There was also
depletion of fertilizer nitrogen probably due, in
addition to leaching, to plant uptake, im-
TABLE 6
Average fertilizer nitrate-N concentrations for three
sampling dates by depth and lateral sampling
location for two subirrigated plots fertilized with
N enriched sodium nitrate, 1973.
Samp. Depth
(cm)
Lateral Sampling Distance
0
25
Nit
51
58
rate-N (ppm
May 31,
0-30
30-60
60-90
1.3
L8
0.0
0.8
0.0
0.0
4.5
0.1
0.0
June 22
0-30
30-60
60-90
90-120
120-150
7.4
0.4
0.0
0.0
0.0
6.8
2,0
0.1
0.0
0.0
17.1
0.2
22
0.0
0.0
July 28,
0-30
30-60
60-90
90-120
120-150
150-180
180-210
210-240
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.1
0.5
0.5
0.2
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1973
32.0
0.2
0.0
, 1973
19.4
0.5
0.2
0.0
0.0
, 1973
0.1
0.0
0.0
0.0
0.0
0.0
0.0
0.0
76
—
0.0
0.0
0.0
L6
0.0
0.4
0.2
0.0
0.2
0.2
0.0
0.0
0.2
0.2
0.0
0.0
(cm)
102
0.2
0.0
0.0
10.5
0.1
1.0
0.0
0.0
0.0
0.5
0.0
0.0
0.0
0.0
0.0
0.0
mobilization and denitrification. The more dif-
fuse pattern with a lack of distinct separation
between the fertilizer applied in 1973 and 1974
was probably due to the fact that the fertilizer
nitrate was held in the upper portions of the
profile longer under a furrow system than under
a sprinkler system. The fertilizer was applied in
a band in the bed at a vertical placement above
the bottom of the water furrow. Water moving
into and upward in the bed carried the fertilizer
TABLE 7
Average fertilizer nitrate-N concentrations for three
sampling dates by depth and lateral sampling
location for two sprinkler irrigated plots fertilized with
enriched sodium nitrate, 1974.
Samp. Depth
(cm)
Lateral Sampling
0
25
Nitr
51
•ate-N
May 22,
0-30
30-60
60-90
90-120
120-150
150-180
180-210
210-240
240-270
270-300
2.0
0.6
0.4
0.5
0.8
0.4
1.0
2.9
0.3
1.3
0.2
0.2
0.2
0.2
0.3
0.2
0.3
0.4
0.6
0.4
1.6
4.4
1.8
0.2
0.1
0.3
0.2
0.2
0.2
0.4
June 18,
0-30
30-60
60-90
90-120
120-150
150-180
180-210
210-240
240-270
270-300
330-360
0.3
0.5
0.5
0.4
0.6
0.5
0.3
0.7
2.7
2.1
0.1
7.2
0.1
0.2
0.5
0.6
0.7
0.4
0.2
1.0
2.5
0.4
Distance
58
(ppn
1973
2.0
2.6
0.2
0.2
0.4
0.0
0.2
0.2
0.0
0.0
1974
0.0 12.7
LQ.
3.8
0.6
0.5
0.5
0.4
0.3
0.4
2.3
1.1
July 15,
0-30
30-60
60-90
90-120
120-150
150-180
180-210
210-240
240-270
270-300
300-330
0.2
0.2
0.2
0.2
0.2
0.3
0.4
0.2
0.6
3.0
1.0
0.5
7.6
3.3
1.2
0.5
0.5
0.5
0.3
L3
1.4
0.8
0.5
1.3
2.2
2.1
0.6
0.4
0.6
0.5
1.8
1.5
0.6
0.8
2.5
1.3
0.3
0.5
0.2
0.2
1.0
2.9
0.7
1974
3.4
7.4
4.6
4.0
0.7
0.3
0.6
0.7
2.6
1.8
0.5
76
*\
i)
0.2
0.6
0.3
0.2
0.2
0.2
0.2
0.2
0.5
0.5
24.4
0.2
0.3
0.3
0.2
0.4
0.3
0.1
0.0
0.1
0.1
6.4
4.8
1.4
0.7
0.3
0.2
0.5
0.6
0.6
1.1
1.0
(cm)
102
0.2
0.2
0.2
0.1
0.2
0.1
0.2
0.2
1.0
0.3
0.1
0.1
0.2
0.3
0.2
0.1
0.2
1.1
L5
1.4
0.1
0.3
0.5
0.6
0.6
0.4
0.2
0.6
0.8
2.0
1.0
0.6
30
-------
FERTILIZER NITROGEN IN SOIL
along. Thus, fertilizer movement in the furrow
system was affected by irrigation water moving
upward as well as downward and rainfall
moving downward. Where all the water move-
ment under the sprinkler system was straight
down, the fertilizer nitrate tended to move the
same way under both irrigation and rainfall
resulting in the distinct separation of 1973 and
1974 applied fertilizer.
TABLE 8
Average fertilizer nitrate-N concentrations for three
sampling dates by depth and lateral sampling
location for two furrow irrigated plots fertilized with
enriched sodium nitrate, 1974.
Samp. Depth
(cm)
Lateral Sampling Distance
0
25
51
rate-N
May 20,
0-30
30-60
60-90
90-120
120-150
150-180
180-210
210-240
0.3
0.2
1.0
1.6
1.5
0.2
0.2
0.2
0.3
0.6
2J
3.4
2.0
0.5
0.5
0.6
0.8
15.4
2.0
0.4
1.0
LJ
1.2
0.5
June 20
0-30
30-60
60-90
90-120
120-150
150-180
180-210
210-240
240-270
270-300
300-330
330-360
360-390
0.2
0.6
0.6
1.0
4.5
4.8
3.1
0.8
0.1
0.0
0.0
0.0
0.0
44.0
20.3
2.2
3.4
6.2
W.
5.7
2.4
0.4
0.2
0.0
0.2
0.0
5.6
4.2
3JJ
4.5
7.6
4.3
jx4
L6
0.4
0.2
0.0
0.0
0.0
July 17
0-30
30-60
30-60
90-120
120-150
150-180
180-210
210-240
240-270
270-300
1.5
2.7
0.9
L2
L9
1.2
3J
2J,
0.4
0.4
5.6
3.6
L2
8.1
5J3
3.0
4.3
5£
Li
0.6
0.7
0.7
L§
4.Q
2.7
2.0
3.2
3^5
M
0.6
58
(ppm)
1974
0.2
5.0
0.4
0.9
0.8
1.5
1.8
0.0
, 1974
16.1
5.8
LS
3.0
5J5
4.4
3.7
1.4
0.3
2.2
2.7
1.0
0.0
, 1974
3.2
6.1
3.0
0.8
1.1
0.6
0.9
L2
0.8
0.2
76
0.2
L4
0.2
0.3
0.4
0.5
0.2
0.2
0.8
0.4
0.4
0.2
0.5
0.9
2.6
0.8
0.2
0.0
0.0
0.0
0.0
0.3
2.1
L2
0.5
0.2
0.8
0.8
0.4
0.2
0.0
(cm)
102
0.3
0.1
0.0
0.0
0.2
0.2
0.2
0.0
0.3
0.3
0.5
0.6
0.8
3.5
3.8
2J3
0.4
0.2
0.0
0.0
0.0
0.5
0.5
0.8
0.8
0.7
0.3
0.7
0.6
0.1
0.2
Data from the subirrigated plots for 1974
are shown in Table 9. The movement pattern is
different than for the two previously discussed
plots and somewhat similar to that in 1973
wherein the fertilizer nitrate was generally held
up in the profile. At the first sampling date on
May 16, the fertilizer nitrate from the starter
band was detected with only small quantities of
fertilizer nitrate other than that found in the
profile. The sidedressing bands were detected in
the upper 30 cm of the profile at the second
sampling date and by the third sampling date it
TABLE 9
Average fertilizer nitrate-N concentrations for three
sampling dates by depth and lateral sampling _
location for two subirrigated plots fertilized with 15 N
enriched sodium nitrate, 1974.
Samp. Depth
(cm)
Lateral Sampling
0
25
51
ntti.TsJ
May 17,
0-30
30-60
60-90
90-120
120-150
150-180
0.5
0.7
0.6
0.7
0.4
0.4
0.5
0.5
0.2
0.6
L2
0.5
0.3
1.0
7.9
0.0
0.1
0.2
June 14,
0-30
30-60
60-90
90-120
120-150
150-180
180-210
210-240
240-270
270-300
2.4
0.6
0.5
6.0
1.6
3.3
go
0.7
0.3
0.0
17.2
0.6
1.9
0.4
0.9
0.6
2.0
0.2
0.1
0.3
0.0
0.0
0.0
L4
0.2
0.4
0.0
0.0
0.0
0.0
July 18,
0-30
30-60
60-90
90-120
120-150
150-180
180-210
210-240
240-270
270-300
300-330
330-360
360-390
390-410
as
L5
0.2
0.4
0.2
0.4
0.4
0.4
0.1
0.6
2.0
L8
1.0
0.4
0.8
0.1
1.2
0.4
0.3
0.6
LS
L2
1.2
1.3
0.9
0.3
0.2
0.2
0.2
0.0
0.0
0.5
0.0
0.0
0.0
0.0
0.1
0.1
0.1
0.1
0.1
0.1
Distance (cm)
58
/TTTiYYl
V,ppiii
1974
22
0.6
9.5
0.0
0.0
0.7
1974
0.0
0.0
0.6
0.4
0.1
0.4
1.4
1.6
LO
0.9
1974
0.2
0.0
0.0
0.0
0.0
0.0
0.1
0.2
0.3
0.4
0.2
0.2
0.2
0.2
76
\
0.0
0.4
0.7
0.2
0.0
0.2
2O5
0.2
0.1
0.2
0.6
0.6
0.3
0.3
0.2
0.2
0.2
0.1
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.1
0.1
0.2
102
0.2
0.2
0.2
0.0
0.0
0.0
0.2
0.4
L2
2.1
2J)
1.8
0.6
0.4
0.4
0.5
2.8
0.2
0.0
0.1
0.0
0.2
0.2
0.2
0.2
0.4
0.5
0.7
0.8
LQ
31
-------
NITROGEN IN RETURN FLOWS
had moved outward under the furrows. There
was some small amount of leaching of fertilizer
nitrate in the profile along with decreases in
concentration due to the same reasons given
previously. The indications for the subirriga-
tion system are that fertilizer nitrate was held
higher in the profile for longer periods of time
than for either the furrow or sprinkler systems
due to the upper movement of the irrigation
water.
Utilizing soil analyses data and plant
analyses data from 1974, it was possible to
account for 92.6, 86.1 and 50.5 percent of the
fertilizer nitrate applied over the two-year
period for the sprinkler, furrow and subirriga-
tion systems, respectively. Thus the indications
are that denitrification was greatest under
subirrigation, next under furrow irrigation and
least under sprinkler irrigation. The distribu-
tion of fertilizer nitrate in the soil profile would
tend to confirm this in that it was held higher in
the profile in a position where the soil
microorganisms responsible for denitrification
could have a readily available energy source in
the order: subirrigation > furrow irrigation >
sprinkler irrigation. Tensiometer data indicated
that the upper portions of the soil profile were
maintained at water potentials of -20 cb or
greater for substantial periods of time. Field
capacity for this particular soil has been
determined to be -10 cb.
In summary, some general statements can
be made about the movement of fertilizer nitrate
under these irrigation systems. The banded fer-
tilizer under the sprinkler system tended to
move straight down in the profile, apparently
somewhat more rapildy than for the two other
irrigation systems. Thus, the sharp distinction
between the 1973 and 1974 applications. For the
furrow irrigation system, with the band applied
above the bottom of the water furrow, the
direction of water flow and subsequent move-
ment of fertilizer nitrate apparently was such
that it held the fertilizer nitrate up in the profile
somewhat longer than for the sprinkler system
even with greater water application rates. The
fertilizer tended to merge in the center of the bed
and then to move downward, either with the
movement of irrigation water or rainfall. Thus,
the more diffuse pattern of vertical distribution
and less distinct break between 1973 and 1974
applied fertilizer. For subirrigation, the fer-
tilizer nitrate apparently moved upward and
outward from the bands to a position next to the
furrow or under the furrow due to the movement
of the irrigation water that came from beneath
and to the side of the fertilizer bands. Movement
downward apparently occuurred due to some
irrigation water and rainfall.
While apparently a number of factors af-
fected fertilizer nitrate movement, including
rainfall between the crop years, crop utilization,
denitrification and immobilization, there is no
doubt that the method of irrigation played a
major role in the movement of fertilizer nitrate-
nitrogen in this study.
REFERENCES
1. Bremner, J.M. 1965. Isotope-ratio anal-
ysis of nitrogen in nitrogen-15 tracer in-
vestigations. In C. A Black (ed.) Methods of soil
analysis. Agronomy 9:1256-1286. Am. Soc. of
Agron., Madison, Wis.
2. Environmental Protection Agen-
cy. 1971. Methods of chemical analysis of
water and wastes. National Environmental Re-
search Center. Analytical Quality Control Lab-
oratory, Cincinnati, Ohio.
32
-------
Nitrogen and Water Management
to Minimize Return-Flow Pollution
from Potato Fields of the
Columbia Basin
B. L. McNEAL, B. L. CARLILE and R. KUNKEL
Department of Agronomy and Soils,
Washington State University, Pullman, Washington
ABSTRACT
Cooperative field studies have evaluated
levels of dissolved soil N, and corresponding
crop yields, for selected potato production prac-
tices in the Columbia Basin area of Washing-
ton. High dissolved-N levels were found
throughout the growing season in well-
managed potato fields. Such levels were de-
creased by decreased fertilization rate, use of
slow-release N fertilizers, or sprinkler-applica-
tion of N fertilizers.
Careful water management during
sprinkler irrigation proved capable of main-
taining dissolved-N within the root zone of
subsequent crops by season's end, even on very
sandy sites. Alternate-furrow irrigation also
proved effective in "trapping" much banded
fertilizer N within the plant root zone on
heavier-textured furrow-irrigated soils. Periodic
recovery of residual N by other crops in the
rotation is necessary to prevent eventual return-
flow contamination, however.
Site-to-site sampling variability necessi-
tated the use of composited soil samples, rather
than fixed-position soil solution extraction
cups, for adequate monitoring of dissolved-N
levels in soils of the area. Neither dissolved soil
N nor plant petiole NO 3-Nproved to be reliable
predictors of crop N needs at the high residual
soil N levels commonly found in recropped
potato fields of the Columbia Basin.
INTRODUCTION
In addition to return-flow contamination of
surface streams, agricultural practices in
irrigated areas can lead to significant return-
flow contamination of area groundwaters. Of
particular concern is the rapid leaching of
nitrates from soils whenever water in excess of
crop needs is allowed to percolate. A test crop of
particular sensitivity to fertilizer and water
management techniques is the potato, because
of its concurrent requirements of high fertility
and high soil moisture levels. Field studies of
nitrate leaching in conjunction with various
soil, crop and water management practices were
carried out in the Columbia Basin area of
central Washington, largely with Environmen-
tal Protection Agency funding from 1970
through 1974. This 200,000 hectare irrigation
project is largely responsible for the fact that the
state of Washington is second in the nation in
total potato production, and leads the nation in
potato production per acre. Hence, the potato
serves as a significant and sensitive indicator
crop for the area.
METHODS AND MATERIALS
Initial sampling of soil solutions from
potato-producing fields was carried out through
the use of ceramic extraction cups placed at 8 to
12 locations in each field. These cups generally
were placed at the 30, 60, 120 and 240 cm soil
depths, as well as immediately above the basalt
or caliche layer underlying the field. A portable
vacuum pump was used to produce a vacuum of
approximately 0.4 bar at each sampling site,
with a small butane tank being used as a
vacuum buffer and attached through a
manifold system to each set of cups. Soil solu-
tion extraction was terminated for each depth
after 50 to 100 ml of solution had been obtained
from the appropriate extraction cup, or after 24
33
-------
NITROGEN IN RETURN FLOWS
hrs had elapsed since irrigation. Samples were
stored in a refrigerator at approximately 4°C
until analysis.
As the large field variability from sampling
site to sampling site became evident, decreasing
reliance was placed upon fixed-position sets of
extraction cups, and greater reliance was placed
upon variable-position soil sampling for ob-
taining estimates of soil solution nitrate concen-
trations and amounts of nitrate leaching within
irrigated potato fields. Soil samples were
collected by hand with a King tube sampler
during the cropping season, and either with a
King tube or with an hydraulic Giddings
sampler at season's end. Samples were placed in
250- or 500-ml plastic cartons for transport the
same day to the laboratory in Pullman. Samples
were then frozen until they could be dried prior
to analysis. Drying was in a convective drying
oven at 60° C for 24 hrs. Samples were then
ground to pass a 2mm sieve and stored until
analyses could be performed on 2:1 watensoil
extracts. Extracts were clarified by centrifuga-
tion, after a 10-minute extraction period on a
reciprocating shaker.
Total dissolved inorganic nitrogen in ex-
traction cup extracts or soil sample extracts was
measured with a steam distilliation method
employing MgO and Devarda's alloy (Bremner,
1965). Most of this nitrogen was in the nitrate
form, as identified by combining the above
steam distillation procedure with one employ-
ing MgO alone during the initial phases of the
study. In addition, sample pH and electrical
conductivity were measured with a standard
laboratory pH meter and conductivity cell,
respectively, and chloride was measured poten-
tiometrically with a commercial chloridimeter.
Initial studies during 1970 and 1971 of the
project period were carried out at an extremely
sandy site in Block 21 of the Columbia Basin
project. Many soils of this unit had initially
been passed over for irrigation development
because their sandy nature was incompatible
with the surface irrigation systems employed on
most initially-developed lands of the Columbia
Basin Project. They are subsequently being
developed using solid-set or center-pivot
sprinkler irrigation systems. Field studies at
this location were carried out in connection with
workers from the WSU Irrigated Agriculture
Research and Extension Center at Prosser,
Washington, with joint funding from the U.S.
Bureau of Reclamation (Middleton et al, 1975).
The stratified sands and sandy loams at this
location were underlain by basalt at a depth of
2.4 to 3.1 meters. A water table developed at the
surface of the basalt beneath plots receiving
excessive water applications.
Remaining studies were carried out in
cooperation with Dr. R. Kunkel of the Depart-
ment of Horticulture of Washington State Uni-
versity. Many of the studies were carried out at
the WSU Othello Field Station on a silt loam
soil typifying many of the first lands developed
for the Columbia Basin Project. Such soils are
commonly furrow-irrigated, although sprinkler
irrigation is being practiced increasingly on
these soils as well. The soil at the Station is
commonly 1 to 1.5 meters in depth, and un-
derlain by caliche hard pan.
Studies were also carried out during the
1973 season at four locations in grower's fields.
The fields were arrayed in a 150-kilometer arc
extending from a location in the Horse Heaven
Hills area south of Pasco and Kennewick (just
north of the Oregon border) to a location east of
Moses Lake in the northern portion of the
Columbia Basin. All locations were managed by
the growers themselves, although
differentially-fertilized, planted and harvested
by WSU workers. They thus provided an es-
timate of potential ground water and drainage
water contamination under "real world"
operating conditions.
Most experimental plots throughout the
studies were ten to fifteen meters long and 2 to 4
rows (1.6 to 3.2 meters) wide. Four to six
replicates normally were included at each loca-
tion for each set of management conditions
studied.
RESULTS AND DISCUSSION
Sandy Site
Typical results from the extraction cup
samplings at the sandy experimental site are
presented in Table 1. Except for cases where
excessive leaching of fertilizer nitrogen oc-
curred (e.g., during late season for the high-rate
furrow-irrigated plots), soil solution dissolved N
values were commonly in the range of several
tens to a few hundreds of mg per liter. This
contrasts with the maximum drinking water
standard of 10 mg/1 for dissolved NOs-N.
Shallow ground waters in extensive potato
producing areas thus might commonly be ex-
pected to develop NO 3 -N values in excess of the
34
-------
NITROGEN MANAGEMENT — COLUMBIA BASIN
TABLE 1
Average soil dissolved inorganic N values, and tuber yields, sandy site, 1969-1971.
Average Dissolved N Values
Average Tuber Yields
Treatment
Low-rate Sprinkler
(QDb
High-rate Sprinkler
(Q3 = 1.5 x Ql)
Low-rate Furrow
(Wl)
High-rate Furrow
(W3 = 3 x Wl)
Depth
(cm)
0-120
120+
0-120
120+
0-120
120+
0-120
120+
Apr 20
-May 5
June
4-11
July
3-8
Aug
5-12
Sept
4-9
Mea- 110
sure kg N/ha
(mg /liter)
40
68
139
45
56
206
83
16
76
33
105
48
195
20
310
16
73
70
296
36
512
162
68
48
234
113
132
160
85
123
9
5
298
78
54
33
26
12
2
3
Total
Nol's
Total
Nol's
Total
Nol's
Total
Nol's
446
379
450
383
—
—
—
—
330
kg N/ha
560
kg N/ha
(quintals/ha)
554
414
553
475
389
221
368
205
587
413
617
481
387
155
379
192
a Ceramic extraction cup data, 5 to 8 data per average.
k Ql = replacement of ET, using evaporation pan estimates. Wl = time to wet to mid-furrow at
mid-run (122-meter run).
drinking water standard. Alternate water
sources may be required in such areas for young
infants and nursing mothers. The data of Table
1, as well as the data of most subsequent tables
of this report, demonstrate that the potato is not
an efficient nitrogen-utilizer. Soil solutions at
the end of the growing season in potato-
producing fields still commonly contain several
tens of mg/1 of NO 3-N (see also Kirkham et al,
1973). This is in contrast to results for efficient
nitrogen utilizers, such as wheat (Fanning et al,
1969). This is probably one reason why high
empirical fertilizer requirements have been
developed for this crop.
The low-rate sprinkler application (ap-
proximately 75 surface-cm of water) was suf-
ficient to produce near maximal yields at this
site (Jensen et al, 1961). Slight increases in total
tuber yield, and somewhat larger increases in
yield of U.S. #1 tubers at the site with higher
sprinkler application rates, probably result
from compensation for wind distortion of solid-
set sprinkler patterns at the site. Some localized
water stress undoubtedly developed under the
windy late-spring and early-summer conditions
for the low-rate sprinkler treatment.
The low-rate sprinkler treatment main-
tained most fertilizer nitrogen in the surface 120
cm of soil until season's end. This nitrogen thus
would be available for "scavenging" by subse-
quent deep-rooted crops in the rotation. The
high-rate sprinkler treatment led to decreased
dissolved-N values for the surface 120 cm of soil
near season's end. As yields were not increased
substantially, and as soils at the site appear to
be too sandy for much denitrification at this
water application rate, the decrease probably
reflects nitrogen leaching to the greater depths.
Such leaching may account for the increased
nitrogen concentrations at the greater depths
for this application rate on the August sampling
date. The data are far from unequivocal,
however, and reflect a common problem with
the use of ceramic extraction cups for field
studies involving realistic depth increments
and sampling periods. Substantial amounts of
solute can be present in the soil between extrac-
tion cup depths at any one extraction time, and
yet go essentially undetected. Leaching of
nitrogen to depths between the 240-cm and 300-
cm positions, or dilution of leached nitrogen
with low-N waters in the water table at the 300-
cm depth, could easily produce the
anomalously-low subsoil dissolved-N levels of
the early September sampling period.
As expected, the studies demonstrate that
use of furrow irrigation on such extremely
sandy sites would lead both to substantial
leaching of nitrogen and also to substantially
lower crop yields. Sprinkler irrigation had been
35
-------
NITROGEN IN RETURN FLOWS
at 1 to 3 day intervals during the study period,
and furrow irrigation was at 2 to 4 day intervals,
with the selected interval being shortened for
each successive year of the study. The pattern of
nitrogen leaching to greater soil depths is
clearly evident from the data for the low-rate
furrow-irrigated plots, where dissolved-N
values decreased progressively after early July
for samples from the surface 120 cm of depth.
Substantial amounts of nitrogen were present
at the greater soil depths (well beyond the
rooting depth of potatoes) on both the July and
August sampling dates. This water application
rate (averaging 350 surface cm of water) had
virtually depleted the soil profile of dissolved-N
by the end of the irrigation season. At the high
rate of furrow irrigation (averaging 920 surface
cm of water), root-zone dissolved-N values
decreased rapidly after early June. The entire
profile was leached essentially free of nitrogen
by early August at this water application rate.
In addition to problems with site-to-site
variability and solute placement between ex-
traction cup depths, an additional problem at
this sandy site was the fact that all extraction
cups did not yield solution on every sampling
date. As a result, variable sample populations
were included on many dates. This was not as
great a problem for finer-textured sites, but
subsequent sampling placed increasing
reliance on soil sampling in place of extraction
cup usage.
Silt Loam Site
In 1972, all sampling was carried out at the
Othello Field Station of Washington State Uni-
versity, on a silt-loam soil (McNeal and Kunkel,
1973). As evident from the data for the four
experiments summarized in Table 2, root zone
dissolved-N values at season's end were com-
monly on the order of several tens of mg/1. As
root zone dissolved-N values generally decreas-
ed during the growing season, because most
nitrogen in these experiments had been banded
or broadcast at planting time, this indicates
that even higher nitrogen concentrations are
probably present in most waters percolating
from these fields. For example, the subsoil under
a long-term fertilizer factorial experiment (Ta-
ble 2-d) increased 3-fold in dissolved-N levels
over a five year period at the highest nitrogen
application rate. Soils at this site commonly de-
velop a tillage or traffic pan at the 20 to 40 cm
depth during potato production (Gardner et al,
1975). Such pans are typical of many potato-
producing areas of the Columbia Basin. Water
flux estimates based on tensiometer data fell in
the range of 0.02 to 0.2 cm per day for this loca-
tion during 1972. These values are quite low in
relation to soil texture and existing irrigation
practices, and probably reflect the influence of
the compacted layer. Such water flux estimates
can be combined with values for dissolved
nutrient concentrations to provide estimates of
the quantities of nutrients being leached under
various farm management regimes. For exam-
ple, a flux of 0.1 cm per day for a 120-day
growing season corresponds to 12 surface cm of
water, or to 59 kg of N per hectare at an average
dissolved-N level of 50 mg/1.
The data of Table 2 generally substantiate
the trends of increased root zone dissolved-N
levels at higher nitrogen application rates
and/or lower water application rates. The lower
subsoil dissolved-N values at the head of furrow-
irrigated fields (Table 2-a) may reflect either
greater leaching of nitrogen from the subsoil
with greater amounts of water passage through
the profile, or increased tendency for denitrifica-
tion at these relatively shallow depths at higher
average water contents. As a basis of com-
parison for data of the table, commercial
nitrogen applications of 300 to 400 kg of N per
hectare are common for potatoes in the Colum-
bia Basin area at present, and values of 500 to
600 kg of N per hectare are not uncommon for
sandy soils irrigated with center-pivot systems,
where nitrogen can be applied through the
irrigation line throughout the irrigation season.
Although levels of dissolved-N were higher at
the higher nitrogen fertilization rates, there was
not a linear relation between nitrogen applica-
tion rate and nitrogen concentration of the soil
solution. Substantial amounts of nitrogen
remained in the soil solution at season's end
even at low nitrogen application rates, and
levels were increased only two to three fold even
when the nitrogen fertilization rate was in-
creased five to six fold.
An interesting observation from Table 2-c is
the maintenance of high levels of dissolved-N in
the root zone even after the over-winter period
following the 1971 suspension experiments. By
the fall of 1972, however, with the ground left
fallow during the summer of 1972, root zone
dissolved-N levels had decreased markedly.
This probably reflects incorporation of
dissolved-N into organic forms, for little irriga-
36
-------
NITROGEN MANAGEMENT — COLUMBIA BASIN
tion water had been applied prior to the fall 1972
sampling, and the ryegrass cover crop then
being grown on the plots had not developed
sufficiently to utilize much of the nitrogen
which was present. As indicated earlier, subsoil
dissolved-N levels under long-term fertilizer
plots were in the neighborhood 150 mg/1 for a
450 kg N/ha fertilization rate (Table 2-d).
Because of the poor utilization by potatoes of
applied nitrogen, however, subsoil dissolved-N
values were on the order to 50 mg/1 even at
extremely low nitrogen application rates. This
substantiates the potential for ground water
and drainage water contamination by return-
flows from potato-producing areas.
Grower's Fields
Data of Tables 3 and 4 demonstrate results
from the sampling of grower's fields during the
1973 season. Once more, dissolved nitrogen
levels of several tens to a few hundred mg/1 were
common in all cases, except for the furrow-
irrigated sandy site near Moses Lake where
leaching of nitrogen from the subsoil was ap-
parent once it "escaped" entrapment in root
zone soil during alternate-furrow irrigation
TABLE 2
Average soil dissolved inorganic N values, silt loam site, 1972.
Experiment
a) Furrow-rate
b) 1972 Suspension
Fertilizer
c) 1971 Suspension
Fertilizer
d) Long-term Fertilizer
Factorial
Depth
(cm)
0-60
60+
0-60
60+
0-60
60+
0-60
60+
0-60
60+
Average Dissolved N Values a
(mg/liter)
Head of Field Tail of Field
High Low
Waterb Water
30 38
6 13
Low High
Nc N
21 47
9 10
Broadcast
110 670
kg N/ha kg N/ha
37 89
21 43
Spring of 1972
110 670
kg N/ha kg N/ha
80 174
32 105
Nitrogen Plots
0-100 450
kg/ha d kg/ha
63 44
49 147
High Low
Water Water
49 104
31 93
Low High
N. N
64 88
36 88
Banded
110 670
kg N/ha kg N/ha
38 54
23 42
Fall of 1972
110 670
kg N/ha kg N/ha
15 24
39 67
P and K Plots
0-100 450
kg/ha kg/ha
44 63
106 102
aSoil sampling data, generally post-harvest.
k 122-meter run, wetting alternate sides of a given crop row every 6 or 2 days.
c 220 or 450 kg N/ha.
dAnnually as N, P2O5 or K2O for 1965-1967 and 1970-1971. 480 kg N/ha to all plots in
1972.
37
-------
NITROGEN IN RETURN FLOWS
TABLE 3
Averages and variations in soil dissolved inorganic N values, and tuber yields, suspension
fertilizer trials, grower's fields, 1973.
Location
Soil
Irrigation System Depth
Average Dissolved N Values a Average Tuber Yields
Early Season Late Season Total No 1's
Fib
F4
Fl F4
Fl F4 Fl F4
Tri-Cities Area
Horse-Heaven Hills
Moses Lake
Othello
Silt Loam
Sand
Sandy Loam
Silt Loam
Center-pivot
Center-pivot
Alternate-furrow
Solid-set
(cm)
0-60
60+
0-60
60+
0-60
60+
0-60
60+
(mg/liter)
193
52
177
56
124
12
155
95
461
48
204
68
293
26
367
121
121
39
48
61
52
8
43
69
178
68
154
127
120
13
101
111
(quintals/ha)
— — — —
451 454 409 411
466 539 278 296
731 743 431 381
Coeff. of Variation, dissolved N Values ('.
Late Season
Tri-Cities Area
Horse-Heaven Hills
Moses Lake
Othello
Silt Loam
Sand
Sandy Loam
Silt Loam
Center-pivot
Center-pivot
Alternate-furrow
Solid-set
0-60
60+
0-60
60+
0-60
60+
0-60
60+
Fl
98
66
65
72
98
112
96
111
F4
103
101
111
41
102
147
95
124
Early Season
Ba
72
65
74
67
92
107
104
66
Br
70
69
88
53
104
%
106
101
Ph)
Late Season
Ba
91
69
—
92
112
91
115
Br
110
98
88
56
107
148
99
121
a Soil sampling data.
bFlandF4=110and450kg/haN,P2O5 andK2OattheTri-Citiessite,250and520kg/haattheHorse-Haven Hills site,
and 110 and 450 kg/ha at the Moses Lake and Othello sites, respectively.
(where adjacent furrows are never wet
simultaneously during the growing season).
The use of alternate-furrow irrigation still main-
tained reasonably high dissolved-N levels in the
plant root zone at this site until late season,
however. Once more, dissolved-N values in-
creased with increasing fertilizer application
rate, and decreased during the growing season,
at each location. Nitrogen was applied through
the irrigation line, in addition to pre-plant
applications to all experimental plots, for each
of the center-pivot locations. Use of slow-release
nitrogen sources (Table 4) led to substantial
reductions in root-zone and subsoil dissolved-N
values early in the season at each location. By
late season, however, there commonly remained
more dissolved-N in root-zone soils from the
slow-release nitrogen plots than from those
fertilized with traditional nitrogen sources.
Thus, subsequent crop management would be
extremely important where slow-release
nitrogen sources were used to minimize ground
water or drainage water contamination in
potato-producing areas.
As was commonly observed throughout the
study period, only slight responses to nitrogen
fertilization in excess of 200 to 300 kg of N per
hectare were evident in most cases. In fact, yield
of U.S. No. 1 tubers was commonly decreased by
increasing nitrogen application rate. Such data,
when coupled with the above-mentioned com-
mercial fertilization rates, suggest that excess
nitrogen is being applied for potato production
in much of the Columbia Basin area at the
present time. One reason for this may be that
nitrogen recommendations developed for new
potato lands are being maintained as lands
revert to nearly-continuous potato production,
despite the facts that somewhat lower yields are
obtained (even with annual fumigation) under
continuous production and that nitrogen
release during subsequent growing seasons
from prior crop residues may constitute an
appreciable source of nitrogen. One exception to
this trend of decreasing yields with increasing
fertilization rate occurred at the Moses Lake
site, where furrow irrigation of a sandy site was
apparently leading to enough nitrogen leach-
38
-------
NITROGEN MANAGEMENT — COLUMBIA BASIN
TABLE 4
Average soil dissolved inorganic N values, and tuber yields, slow-release N experiments, grower's fields, 197:1
Location
Soil
Irrigation System Depth
Average Dissolved N Values a
Early Season Late Season
Tradb SR Trad SR
Average Tuber Yields
Early Season Late Season
Trad SR Trad SR
Horse-Heaven Hills
Moses Lake
Othello
Othello Station
Sand
Sandy Loam
Silt Loam
Silt Loam
Center-pivot
Altern ate-f urro w
Solid-set
Alternate-furrow
(cm)
0-60
60+
0-60
60+
0-60
60^
0-60
60-
469
83
256
18
296
84
354
81
Crag,
352
51
128
9
162
50
265
23
liter)
150
105
95
14
85
71
236
100
(quintals/ha)
226
64
94
12
146
57
154
32
385
484
719
606
412
434
712
655
318
263
503
361
325
231
502
392
a Soil sampling data.
b Traditional N sources = NH4NO3, (NH4 I2 and SO4 and urea. Slow-release sources = sulfur-coated ureas SC 100, SC 45 and
SC25.
ing losses to maintain a yield response at higher
nitrogen application rates.
Average tuber yields were commonly as
high or even slightly higher for slow-release
nitrogen sources, reflecting the somewhat
greater maintenance of nitrogen levels
throughout most of the growing season for plots
to which such fertilizer materials had been
applied. An exception was the alternate-furrow
site at Moses Lake, where slow-release nitrogen
sources apparently did not release nitrogen fast
enough to meet the nitrogen needs of the crop
under conditions where appreciable leaching of
fertilizer nitrogen occurred.
The data at the bottom of Table 3
demonstrate the high degree of variation that is
evident in field dissolved-N values. Coefficients
of variation were commonly in the range of 75 to
100% of the means for all field studies. Contrary
to earlier expectations, coefficients of variation
were not substantially lower for sprinkler-
irrigated than for alternate-furrow-irrigated
fields, for broadcast fertilizer applications than
for banded fertilizer applications, or for subsoil
than for root zone samples. The amounts of
variation in such field studies are simply large,
and have to be compensated by large numbers
of composited samples if one is to obtain reliable
estimates of nitrate leaching patterns on a field-
wide basis.
Additional Studies
Table 5 presents data obtained from studies
at the Othello Field Station of Washington State
University during 1973. In the nitrogation ex-
periment, nitrogen was applied through the
irrigation line at various fertilization rates and
application intervals. The data of the table
represent nitrogen application at two rates and
in equal amounts either every two weeks or
every six weeks of the growing season. Highest
yields were obtained at the lower nitrogen
fertilization rate and at the most frequent
nitrogen-application interval. Minimum soil
dissolved-N nitrogen concentrations remained
at about the same levels as observed in the other
fertilization studies, as would be anticipated
from the above-noted inefficiency of potatoes
with respect to nitrogen uptake. The maxima in
soil dissolved-N levels were considerably less in
these nitrogated plots, however. This approach
appears promising as a method for lowering
leaching losses of nitrogen from sprinkler-
irrigated potato fields. Fortunately, it is already
common usage in the Columbia Basin area.
The second study summarized in Table 5
consisted of an evaluation of several sprinkler-
irrigation rates at the silt loam site. A
fertilization-rate study was superimposed. Both
total yields and yields of U.S. No. 1 tubers were
nearly constant whether water was applied at
only 75% of estimated ET needs (Jensen et al,
1961) or at rates as high as 150% of estimated
needs. Somewhat higher yields were obtained at
the higher fertilization rates in this particular
set of studies. This is contrary to common trends
on recropped potato lands in the Columbia
Basin, and may reflect the growth of a crop
39
-------
NITROGEN IN RETURN FLOWS
TABLE 5
Average soil dissolved inorganic N values, and tuber yields, silt loam site, 1973.
Average Dissolved N Values a
Experiment Depth Measure
Early Season
Late Season
F2b F4 2 week 6 week
F2 F4 2 week 6 week
Nitrogation
(cm)
0-60
60+
(mg/liter)
102
71
33
17
49
39
59
20
Average
76
40
Tuber Yields
89
22
74
29
111
33
(quintals/ha)
Total
Nol's
F2
709
513
F4
643
461
2 week 6 week
710
493
646
471
Average Dissolved-NJ Values (mg/liter)
Irrigation Rate
Early Season
F1&2C F3&4 75% 100%
Late Season
150% F1&2 F3&4 75% 100% 150%
0-60
60+
280
30
Total
Nol's
729
129
290
96
583 640 26 108 153 22
39 103 15 139 112 91
Average Tuber Yields (quintals/ha)
26
30
F1&2 F3&4 75% 100% 150%
646
491
680
516
664
493
653
500
673
518
a Soil sampling data
b F2 and F4 = 340 and 670 Kg N/ha, respectively. Applied in equal increments every 2 or 6 weeks.
c Fl and F2 = 110 & 220 kg N/ha. F3 & F4 = 450 & 670 kg N/ha. Sprinkler irrigation at 75, 100 or 150%
of estimated ET, based on microclimatological estimates.
having a high nitrogen requirement on this
particular field of the Othello Station during the
preceeding season. Early season root-zone
dissolved-N values were extremely high in this
study, but late season root-zone, and entire sea-
son subsoil, dissolved-N levels were typical of
values obtained from our other studies as well.
As several studies had indicated that excess
nitrogen was being applied to many potato
fields of the Columbia Basin area, studies were
carried out in 1973 and 1974 to evaluate the
amounts of soil and plant nitrogen required for
maximum crop yields, and to evaluate various
parameters which might be used to estimate
crop yield from soil or plant nitrogen analyses.
Results of these studies are summaried in Table
6. In the 1973 studies, crop yields were correlated
with root zone dissolved-N levels or with plant
petiole N(>3 -N values at various times during
the growing season. Except for the Moses Lake
site, where nitrogen stress developed because of
some nitrogen leaching, no correlation was
obtained between yields and either soil nitrogen
or petiole NO 3 -N values on any of four dates
throughout the growing season. Correlations
were commonly negative, indicating that lower
yields were obtained at higher soil or plant
nitrogen levels.
For the 1974 studies, a series of plots were
established at the WSU Othello Field Station in
which an attempt was made to maintain plants
at various petiole NO 3 -N levels throughout the
entire growing season. As petiole NO 3 -N levels
tended to decrease with increasing
physiological age of the crop, it was only
possible to maintain the different levels on a
40
-------
NITROGEN MANAGEMENT - COLUMBIA BASIN
relative basis. However, the results of the study
suggest that attempts to maintain 10,000 ppm
petiole NOs-N until mid-August may be one
local practice accounting for excessive nitrogen
fertilization in much of the Columbia Basin
area. Of the various petiole NO 3 -N levels and
petiole sampling dates examined, only the in-
termediate NO 3 -N levels for the 90-day samp-
ling date proved well related to total tuber yield
or yield of U.S. #1 tubers. Maintenance of petiole
NO3-N levels of 10,000 to 12,000 ppm until
approximately mid-July would have produced
maximum yields in this particular case.
However, maintenance of such high levels
further into the growing season provided no
demonstrable correlation with either total yield
or yield of U.S. #1 tubers.
When the same data were evaluated in a
different manner, as outlined in the bottom
portion of Table 6, no "critical" petiole NO 3-N
levels were apparent. If such a level were
present, then yields should have been depressed
as the amount of time increased during which
the crop was below this critical level. No such
trends were evident, either for entire-season
petiole data or for petiole data to August 15th.
This was true even when petiole NO 3 -N levels
had dropped as low as 5,000 ppm. Plants grown
for substantial periods at petiole NO 3 -N levels
less than 5,000 ppm produced yields as high or
higher than plants which had never been allow-
ed to decrease below this petiole level. Thus, the
use of a critical petiole NO 3 -N value of 10,000
ppm for potatoes being grown in the Columbia
Basin does not appear to be warranted, except
possibly as an early-season indicator of ade-
quate soil nitrogen levels when most nitrogen is
being applied as pre-plant fertilizer.
The final data from the overall study to be
summarized in this report deal with lateral
variations in soil dissolved inorganic N values.
The first two sets of data in Table 7 involved
experiments at the Othello station site, with
alternate-furrow irrigation employed. The data
compare values beneath crop rows with those
from beneath irrigation furrows when using
this irrigation system. Little difference between
soil dissolved-N levels beneath rows and
furrows were evident when this irrigation tech-
nique was practiced. The data also
demonstrated that differences between the
heads and tails of alternate-furrow-irrigated
fields are normally more significant than
differences between rows and furrows. They
also show that the differences between rows and
furrows were largely eliminated when subsoil
values were dealt with under this type of irriga-
tion practice.
A third set of data from the table deal with a
comparison of extraction cup values for two sets
of extraction cups located in the same set of
TABLE 6
Correlation of tuber yields with soil dissolved inorganic N or petiole nitrate-N values, 1973 and 1974.
Parameter
Average Linear Correlation
Coefficients (1973)
Tuber Yields (1974)
Location or
Measure
Moses Lake
Othello
Othello Station
Total
No 1's
Total 1's. soil N at:
90
.68
. •)•)
-.66
vs. pet
LI"
584
424
101
.77
-.07
-.53
iole NO
L2
695
509
118
.14
-.52
•j at 90
L3
71S
507
vs. davs petit
Tuber Yields (1974) (related
to entire season's petiole
data)
Tuber Yields (1974) (related
to petiole data to
August 15)
Total
No 1's
Total
No 1's
0-20
727
480
690
497
NO'j v
21-40
707
523
773
533
132 davs
.06
-.53
davs:
L4
642
476
)le
A'o / 's i-s petiole NO;i at:
90
-.02
.11
-.41
vs.
LI
775
550
10.000 ppm
41-60
723
513
691
504
61-80
757
537
—
0-20
718
524
696
501
104
-.02
-.07
-.16
petiole
L2
761
494
vs.
NO'
21-40
644
457
728
546
118 132 days
— —
-.07 -.09
-.39 -.42
NO a at 150 davs:
L3 L4
843 797
611 481
days petiole
j •- 5.000 ppm
41-BO 61-80
771 899
547 656
732 —
447 —
Total vs petiole ArO.q at:
90
.01
.28
-.47
vs.
u
700
503
104 118
-.06 -
.16 -.16
-.52 -.46
petiole NO 3 at
L2 U
682 655
480 496
132 days
—
.16
-.47
120 davs:
IA
725
516
vs. days petiole
0-20
722
498
699
496
NO 3 <7.500
21-40 41-60
7.->l 671
556 483
676 748
521 479
ppm
61-8(1
791
570
—
—
aLl = < 8.000. < 4.000 and < 2,000 ppm at 90. 120 and l.iO days, respectively:
L2 = 8.000 - 11.000. 4.000 - 8.000. and 2.000 - 4.000 ppm on the respective dates:
L3 = 11.000 • 14,000. 8.000 - 12.000. or 4.000 - 6,000 ppm: and
L4 = <. 14.000. <. 12.000 or ^ 6.000 ppm.
41
-------
NITROGEN IN RETURN FLOWS
sprinkler-irrigated experimental plots at the
Block 21 (sandy) experimental site. Each pair of
extraction cups at a given depth were only a few
feet apart, with both located beneath the crop
row. As is evident from the data, large
variations in dissolved-N levels at a given depth
can be found within a few feet of lateral distance
in the field. Such variations make reliance on
data from a few sets of extraction cups highly
questionable in terms of estimating field-wide
leaching losses of nitrogen.
The final data of Table 7 provide a com-
parison of extraction cup data and soil samp-
ling data from adjacent soils during three of the
experimental study periods at the Block 21 and
Othello Station experimental sites. Variations
from one point in the field to another are evident
from the data for the July or September samp-
ling periods, when values for cups and for
nearby soil were compared with one another.
When the entire set of data for a given season
were averaged, however, both soil sampling and
extraction cups provided similar results. Thus,
either method could be used to obtain reliable
estimates of the amount of nitrogen being
leached at a specific point in the field, and each
technique serves as an independent confirma-
tion of the experimental approach. Because of
the greater versatility and increased numbers of
areas that can be sampled with the soil samp-
ling procedure, it is preferred over the extraction
cup approach.
CONCLUSIONS
High levels of dissolved inorganic nitrogen
commonly exist in soil solutions from within
and beneath potato fields of the Columbia Basin
area of Washington. A considerable potential
for ground water and drainage water con-
tamination by soil solutions displaced during
excessive irrigation thus exists for this area.
Careful sprinkler irrigation at quantities dic-
tated by micrometeorological or pan evapora-
tion data can produce near maximum yields of
high quality tubers even on extremely sandy
soils, while still maintaining residual dissolved-
N from the current crop season in the top 60 to
120 cm of soil, where it can be removed by
subsequent deep-rooted crops in the rotation.
However, a common tendency to over-irrigate
TABLE 7
Average soil dissolved inorganic N values, including lateral variations.
Location Irrigation System Depth
(cmi
Silt Loam Sit* Alternate-furrow
(1972)
0-60
60-
Silt Loam Site Alternate-furrow
(1972> 0-60
60-
Sandy site Sprinkler (high-rat*)
(1972)
180
240
Both Sites
Sprinkler (197H —
Furrow (1972) —
Furrow (1973) —
Average Dissolved N Values a
Head
38
8
Overall6
63
52
First
Rowsb
Tail
81
73
Rows
(mg liter)
Furrows
Sprikler
42
48
Furrows
83 44
61 50
Extraction Cupp
July 8 Aug 12
30 106
41 70
Average for
July
122
71
242
Sept
94
42
Sept 9
10
94
Cups
Overall
103
56
225
Head
32
11
Second
Tail Sprinkler
70 59
63 40
Extraction Cup
July 8 Aug 12 Sept 9
8
42
Average
July
148
74
336
323 56
41 46
for Nearby Soil
Sept Overall
74 106
48 57
— 209
a Soil sampling data, except where otherwise specified.
" Furrow-rate experiment, table 2.
c Long-term fertilizer factorial, table 2.
" 1972 suspension fertilizer experiment, table 2.
e Cups located a few feet apart, beneath the crop row. in the high-rate sprinkler plots of table 1.
42
-------
NITROGEN MANAGEMENT — COLUMBIA BASIN
late in the season, when plants have matured
and crop water needs have diminished, must be
curbed if the nitrogen is to be subsequently
"scavenged" by future crops.
Increases in dissolved-N levels consistently
were observed with increased fertilization rates,
although the increases were not strictly propor-
tional to fertilization rate. High levels of
residual inorganic nitrogen persisted from
fall until spring unless leached by over-winter
rains. However, these levels were reduced to
more acceptable values during a subsequent
fallow period or following subsequent growth of
a deep-rooted crop. Deep-leached nitrogen
appeared to be neither assimilated nor
denitrified to an appreciable extent, as evident
from subsoil dissolved-N values for a long-term
fertilizer experiment on a silt loam soil of the
area.
Use of slow-release nitrogen fertilizers sub-
stantially lowered soil dissolved-N levels. The
lower dissolved-N levels should have slightly
lowered leaching losses during the growing
season, although higher residual nitrogen
levels often remained at season's end in plots
fertilized with slow-release fertilizers. "Nitroga-
tion" (sprinkler application of N fertilizers
throughout the growing season) produced little
evidence of substantial nitrogen leaching even
at high fertilization rates and infrequent
application intervals. However, soil dissolved-N
levels of several tens of mg/1 persisted even in
nitrogated plots, because of the apparent in-
ability of potatoes to efficiently utilize all
nitrogen which may be present in the soil
solution.
Fertilization rates of 250 to 350 kg of N per
hectare were generally sufficient for maximum
yield of high level tubers on recropped potato
lands at the experimental locations sampled
throughout the Columbia Basin. Over-
fertilization appears to be common in many
portions of the Basin, and is poor practice from
both an environmental and an economic view-
point. No evidence of a "critical" plant petiole
NC>3-N level was found for petiole NC>3-N
levels above 5,000 ppm, even though a critical
level of 10,000 ppm is commonly recommended
in this area. This may account for some of the
apparent excessive usage of nitrogen on
potatoes in the Columbia Basin. Total yield or
tuber quality was predictable from dissolved
soil nitrogen or from petiole NO3-N only for
nitrogen-stressed (e.g. highly-leached) plots in
our studies.
Ceramic extraction cups constituted a poor
sampling procedure for dissolved soil nitrogen
in these studies, due to extremes in concen-
trations which were encountered over lateral
distances of only a few meters. The cups also
failed to extract consistently from all depths at
all times (producing variable sampling pop-
ulations from one period to the next). There was
also a high probability that solute peaks would
be located between extraction cup depths on any
given sampling date, and it proved hard to
predict the exact effects of cup location with
respect to fertilizer band location in band-
fertilized fields. Extrapolation from a few ex-
traction cup sites to an entire irrigated field
appears to be an unreliable approach for
monitoring or verifying fertilizer leaching
patterns in most areas of the Columbia Basin.
REFERENCES
1. Bremner, J. M. 1965. Inorganic forms of
nitrogen. In: C. A. Black, D. D. Evans, J. L. White,
L. E. Ensminger and F. E. Clark, eds. Methods of
Soil Analysis, American Society of Agronomy,
Madison, Wisconsin, pp 1179-1237.
2. Fanning, C. D., A. R. Halvorson and F. E.
Koehler. 1969. Residual nitrates following winter
wheat with high rates of fertilization. Proc. 20th
Annual Pacific Northwest Fertilizer Conf. pp
50-57.
3. Gardner, W. H., C. Calissendorff and R.
Kunkel. 1975. Potato root development. Proc.
14th Annual Washington Potato Conf. page 29.
4. Jensen, M. C., J. E. Middleton and W. O.
Pruitt. 1961. Scheduling irrigation from pan
evaporation. Circular 386, Washington Agric.
Expt. Sta. Pullman. 14 pp.
5. Kirkham, M. B., D. R. Keeney and W. R.
Gardner. 1973. Uptake of labelled nitrate and
water at different depths in the root zone of potato
plants grown on a sandy soil. Agronomy
Abstracts, American Society of Agronomy,
Madison, Wisconsin, p. 125.
6. McNeal, B. L. and R. Kunkel. 1973. Nitrate
leaching following potato fertilization. Proc. 12th
Annual Washington Potato Conf. pp 103-109.
7. Middleton, J. E., S. Roberts, D. W.
James,T. A. Cline, B. L. McNeal and B. L. Carlile.
1975. Irrigation and fertilizer management for
efficient crop production on a sandy soil.
Washington Agric. Expt. Sta. Bulletin 811.10pp.
43
-------
Variability of Nitrate Leaching
Within Defined Management Units
L. J. LUND and P. F. PRATT
University of California, Riverside, California
ABSTRACT
In attempting to relate nitrate concen-
trations below root zones to root zone soil
characteristics, the experimental approach has
consisted of a comparison of a number of sites
within a defined management unit or field. The
basic assumption in this approach has been
that the management of the field is uniform and
variations in nitrate leaching result from
variations in soil characteristics. However,
results reported herein show that even within a
field of limited size, other sources and sinks can
be significantly different at various sampling
locations. The variation in these factors for a
number of fields in the Santa Maria Valley,
California is discussed.
INTRODUCTION
Increased interest over the past few years in
groundwater degradation and stratospheric
ozone destruction has led many scientists,
politicians and control agency personnel to take
an intensive look at agricultural uses of
nitrogen fertilizers and their effects on the
environment. In studying the fate of nitrogen
applied to agricultural soils, a number of
sources, sinks, and pathways need to be con-
sidered. The approach used most frequently in
nitrogen studies has been to develop nitrogen
balances that consider various sources and
sinks. An example of this type of study was
recently reported by Fried, et al. (1976). In this
simplified model, a steady state is assumed in
which the inputs of nitrogen are equal to the
outputs. The sinks of nitrogen described by this
model that are of most interest at the present
time are nitrogen losses by leaching and den-
trification. Nitrates that leach below the root
zone are considered those that are beyond
recovery of plants. One of the products of
denitrification is nitrous oxide, which may have
far-reaching implications with regard to deple-
tion of stratospheric ozone.
In developing nitrogen balances to consider
the magnitude of various sinks, areas of various
size can be considered, including drainage
basins (Ayers and Branson, 1973) or individual
field plots (Adriano, 1972). However, defined
management units (farmers' fields) will
probably be the most important in determining
amounts of nitrogen leached or dentrified, or in
attempting to change the magnitude of these
sinks. In determining nitrate concentrations
below root zones and nitrogen balances for
defined management units, hereafter called
field, it is necessary that one understands the
variability likely to be encountered. By un-
derstanding the contributions of various factors
to this variability, sampling and analysis
programs can be designed that will increase the
reliability of data.
As described by numerous authors, in-
cluding Fried, et al. (1976), a steady state model
for nitrogen can be described by the equation
NM-C-L-DN-P =
[i]
where Np is the nitrogen added in fertilizers,
N j is the nitrogen added in the irrigation water,
NM is the nitrogen added by miscellaneous
ways including NH 3 adsorption, rainfall, etc.,
C is the nitrogen removed in the harvested crop,
L is the nitrogen leached below the root zone,
DN is the nitrogen lost of denitrification, and P
is miscellaneous losses due to erosion, runoff,
NH3 volatilization, etc. At steady state the
assumption is that there is no net change in the
soil nitrogen pool over the time studied. Since at
this time it is very difficult and/ or expensive to
measure DN directly in the field, this equation is
generally rewritten as
DN = NF + NI + NM-C-L-P [2]
In attempting to relate nitrate concen-
trations below root zones to soil characteristics
45
-------
NITROGEN IN RETURN FLOWS
of the root zones, the experimental approach
has consisted of a comparison of a number of
sites within a defined management unit or field.
The basic assumption in this approach has been
that the management of the field is uniform f.nd
variations in nitrate leaching (L) result from
variations in soil characteristics. However, we
have found that even within a field of limited
size, the other sources and sinks in equation [2]
can be significantly different at various samp-
ling locations. The variation in these factors for
a number of fields in the Santa Maria Valley,
California is discussed in this paper.
TABLE 1.
Particle Size Distribution and Organic Carbon
Content of Surface Horizon for Eight Sites
Within a Defined Management Unit
Site Sand Silt
Organic
Clay Carbon Color
343
344
345
346
347
348
349
350
61.0
60.7
59.9
61.2
53.9
58.4
67.6
57.4
32.1
32.3
33.0
31.9
33.1
34.5
26.3
33.6
6.9
7.1
7 1
6.9
13.0
7.0
6.1
9.0
.39
.49
.43
.25
.17
.44
.47
.16
10YR3/4
10YR3/4
10YR3/4
7.5YR4/4
7.5YR4/4
10YR3/3
10YR3/3
7.5YR5/6
MATERIALS AND METHODS
Fields that have a variety of soil profiles
were selected as study areas for determining
nitrogen balances, nitrate concentrations below
root zones for typical management systems,
and the magnitude of denitrification losses.
Eight or nine sites were sampled in each field
that is described here. The soil profile to 6 feet at
each site was described, sampled and
characterized. Soil samples were collected from
three holes at each site in 1-foot increments
between 6 and 20 feet. The lower depth was less
in cases where obstructions or perched water
tables were encountered while drilling with a
Giddings hydraulic auger. Chloride concen-
trations were determined for saturation extracts
by automatic titration and expressed on a soil
solution basis by using field moisture contents
of the soils. Soil solution chloride values
reported here represent means of 12 to 42
samples. Particle size analyses were done by the
pipette method (Day, 1965) and organic carbons
were determined by wet oxidation (Allison,
1965).
Yields were determined by sampling from a
10 x 10 m plot at each site at harvest time. Three
yield determinations were made within each
plot and crop samples from each yield test were
analyzed for total nitrogen (semimicro-kjeldahl)
and chloride (Adriano, et al. 1973).
RESULTS AND DISCUSSION
Field Characteristics
A number of characteristics of a field have
important effects on the variable in equation [2],
some of which will be discussed later. However,
one of these characteristcs that is very impor-
18
Figure 1. Field in Santa Maria Valley studied for
nitrogen balance. Mapping units are MaA-Marina
sand, 0 to 2"o slopes; MaC-Marina sand, 2 to 9% slopes;
PnA-Pleasanton sandy loam, 0 to 2% slopes; and PnC-
Pleasanton sandy loam, 2 to 9% slopes. Location of
sites sampled are indicated by dots.
tant and needs to be considered is the complex of
soils present in the field. Soils can be described
or considered in many ways. Examples would
be soil mapping units, or soil pedons present, or
the variation in surface or subsoil textures. In
few instances does one find fields that have only
one mapping unit. By examining a soil map of
an area, for example, western Riverside County,
California, it is evident that most fields have at
least several mapping units. In addition, map-
ping units on a detailed soil map have only a
certain degree of uniformity. Thus, the varia-
tion in soil characteristics present in most fields
necessitates sampling more than one site within
a field to develop a nitrogen balance. Pratt, etal.
(1977), discussed the number of cores needed per
field to adequately represent the field for a
variety of situations in California.
Three fields in the Santa Maria Valley.
California which are presently being studied to
46
-------
NITRATE LEACHING VARIABILITY
develop nitrogen balances are shown in Figures
1 and 2. Soil boundaries as taken from the
Northern Santa Barbara County Soil Survey
Report (Shipman, 1972) are shown on the
figures along with the locations of sites that
have been sampled. The varying shades of gray
on Figure 1 are due to soil variation. This is
supported by particle size and organic carbon
data for surface horizons of the various sites
(Table 1). No obvious color patterns due to soil
variation are present in Figure 2. However, data
given in Table 2 show distinct particle size and
organic carbon differences between sites. Also,
these variations are not reflected in the soil
mapping units and thus care must be exercised
even when using soil maps to ensure that all soil
bodies within the field are sampled.
Other field characteristics that need to be
considered in certain instances include surface
configuration, depth to water table, and irriga-
tion system layout, as discussed later.
Figure 2. Two fields in Santa Maria Valley studied
for nitrogen balances. Mapping units are StA-
Sorrento sandy loam, 0 to 2% slopes and SuA-
Sorrento sandy loam, sandy sub-stratum, 0 to 2%
slopes. Fields boundaries are shown by dark straight
lines.
Nitrogen Inputs
Fertilizers (N p): The amount of nitrogen
added per unit area is one of the input values
needed for a nitrogen balance for a field. When
considering the field as a whole, one needs only
to determine the total amount of fertilizer
applied and the total area to calculate N p.
However, the distribution of that total nitrogen
applied becomes important when a specific site
within a field is sampled. The uniformity of
yields and nitrogen available for leaching will
depend on the distribution.
One of the obvious factors that will affect
the uniformity of distribution is the method of
TABLE 2
Particle Size Distribution and
Organic Carbon Content of Surface
Horizon for Seventeen Sites Within
a Defined Management Unit.
Site
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
Sand
73.6
74.7
73.9
67.2
63.9
71.3
74.0
65.4
52.3
54.1
66.7
55.2
55.5
70.5
61.5
68.3
79.7
Silt
18.8
17.6
18.2
22.3
25.1
20.4
18.4
24.8
34.0
34.0
22.3
30.5
31.5
20.3
26.2
22.0
14.2
Organic
Clay Carbon Color
7.6
7.7
7.9
10.4
11.0
8.3
7.7
9.8
13.7
11.8
11.0
14.3
13.1
9.2
12.3
9.7
6.2
.31
.33
.35
.40
.43
.30
.34
.42
.58
.58
.38
.60
.64
-
.51
.44
.30
10YR4/4
10YR4/4
10YR4/4
10YR3/3
10YR3/3
10YR3/3
10YR3/3
10YR3/3
10YR3/3
10YR3/3
10YR4/2
10YR3/3
10YR3/3
10YR3/3
10YR3/3
10YR3/3
10YR3/3
application. Methods such as banding or side-
dressing will result in more uniform application
along rows than across rows. Uniformity of
application by broadcasting with a gravity
spreader will depend largely on the operator. A
bulk spreader equipped with a centrifugal
spreader will give nonuniformity (Brinsfield
and Hummel, 1975; Cunningham, 1963). When
nitrogen is added in irrigation water, the uni-
formity of water application in the field becomes
important as does runoff of irrigation water.
These points will be discussed later.
Irrigation Water (Nj): Many waters used
for irrigation contain significant amounts of
nitrogen, usually as nitrate. This source of
nitrogen needs to be considered in developing
nitrogen balances and again the uniformity of
water application will determine the uniformity
of nitrogen applied in the irrigation water. As
an example of the magnitude of nitrogen added
by the irrigation water, water used to irrigate
crops in the Santa Maria Valley typically
contains 10 ppm NO 3-N. If 60 ha-cm of water is
applied to a crop, 60 kg of nitrogen would be
added per hectare. If one is growing vegetables
where 224 kg of fertilizer nitrogen are added per
hectare, this 60 kg would account for only 20% of
the input; but, on lima beans where a typical
47
-------
NITROGEN IN RETURN FLOWS
fertilizer rate is 56 kg/ha, the irrigation water
would account for over 50% of the nitrogen
input.
Miscellaneous (N jy[): Some miscellaneous
inputs of nitrogen would be soil or plant adsorp-
tion of NH 3 from the atmosphere, nitrogen in
rainfall, and fixation of atmospheric nitrogen.
Adsorption of NH 3 by soils has been discussed
by Malo and Purvis (1964), Hanawalt (1969),
and Luebs, et al. (1973). If fields of interest are
near feedlots or dairies, then contributions of
atmospheric NH 3 need to be considered. Rain-
fall is generally considered to contribute 5 to 7
kg of N per hectare per year. Again for most
intensive farming, this amount is negligible.
Most nitrogen balance studies are conducted for
non-nitrogen-fixing crops. If nitrogen-fixing
crops are studied, nitrogen inputs from this
source need to be included.
Nitrogen Outputs
Harvested Crop (C): Nitrogen is applied to
agricultural lands to satisfy the requirements of
the growing vegetation. Many researchers have
shown that the use efficiency of applied
nitrogen is considerably less than 100"o. Thus,
some of the applied nitrogen is lost to other
sinks. In determining the amounts of nitrogen
removed in the harvested crop, it is necessary to
understand various production systems. For
example, in dry lima bean production, the only
material that leaves the field, unless the straw is
used for bedding or feed, is the dry beans. In the
TABLE 4
Yields, Nitrogen Contents of Peppers and Amounts
of Nitrogen and Chloride Removed in the Harvested
Crop for Nine Sites Within a Defined Management
Unit (Figure 2).
Site
360
361
362
363
364
365
366
367
368
Mean
Yield
metric T/ha
12.72
12.41
16.96
27.01
26.72
18.91
17.12
10.75
12.74
17.25
TV
content
%
2.45
2.42
2.39
2.25
2.09
1.96
2.14
2.47
2.44
2.29
W
removed +
kg/ha
91 be
88 be
118 abc
176 a
163 ab
108 abc
107 abc
77 c
91 be
113
Cl
removed +
kg/ha
18 be
17 c
25 be
53 a
42 ab
40 ab
26 be
8c
9c
27
+ Values followed by the same letter are not signifi-
cantly different at the 1 % level.
case of cauliflower, part of the vegetative por-
tion is used to protect the head, thus more
nitrogen is removed from the field than just that
in the head.
When determining the nitrogen removed by
the harvested crop, the variation at different
sites within the field must be considered. In
working with nitrogen balances for fields in the
Santa Maria Valley, California, we have ex-
perienced this variability. Data on yields and
nitrogen removals for three crops from the fields
shown in Figures 1 and 2 are given in Tables 3,
TABLE 3
Yields, Nitrogen Content of Sugar Beets and
Amounts of Nitrogen and Chloride Removed in the
Harvested Crop for Eight Sites Within a Defined
Management Unit (Figure 1).
Site
343
344
345
346
347
348
349
350
Mean
Yield
metric T/ha
56.74
94.57
92.93
87.10
88.38
91.47
94.01
86.09
86.40
N
content
%
1.13
1.00
1.39
1.12
.71
1.39
.89
.59
1.03
N
removed+
kg/ha
94 d
179 a
181 a
149 b
116cd
190 a
137 be
91 d
142
Cl
removed+
kg/ha
93 c
45 d
96c
98 c
39 d
145 b
157 b
201 a
109
-(-Values followed by the same letter are not signifi-
cantly different at the 1 % level.
TABLE 5
Yields, Nitrogen Contents of Dry Lima Beans and
Amounts of Nitrogen and Chloride Removed in the
Harvested Crop for Eight Sites Within a Defined
Management Unit (Figure 2).
Site
370
371
372
373
374
375
376
377
Mean
Yield
kg/ha
2332
2324
2180
2002
2747
2978
2573
1692
2354
N
content
%
4.13
4.46
4.53
3.76
3.55
3.65
3.39
3.15
3.83
N
removed+
kg/ha
91 ab
98 ab
93 ab
71 c
91 c
102 a
82 be
50 d
84
a
removed+
kg/ha
1.5 b
1.8 b
1.7b
1.5 b
2.4 a
2.2 a
2.2 a
1.0 c
1.8
+ Values followed by the same letter are not signifi-
cantly different at the 1% level.
48
-------
NITRATE LEACHING VARIABILITY
4, and 5. The yields of sugar beets at eight
locations within a 14-ha field varied from 57 to
94 metric T/ha (Table 3). The amounts of
nitrogen removed in the harvested crop ranged
from 91 to 190 kg/ha with an average of 142
kg/ha. Yields of peppers grown on a 24-ha field
ranged from 11 to 27 metric T ha for nine
locations (Table 4). Nitrogen removed from the
field varied from 88 to 176 kg, ha with an
average of 113 kg/ ha. The amount of nitrogen
removed by dry lima beans was lower than
either sugar beets or peppers (Table 5). Fifty to
102 kg/ha of nitrogen were removed from eight
locations within a 49-ha field. The average
nitrogen removed was 84 kg/ha for eight sites
where the yield from 1692 to 2978 kg/ ha. From
these data it is apparent that there are not only
marked differences in nitrogen removal by
various crops, but there can also be significant
differences between locations within a field
with "uniform management."
Leaching losses (L): In determining the
amounts of nitrogen leaching below root zones.
it is necessary to know the amount of water
leaching from the root zone (drainage volume)
and the concentration of nitrogen, generally
NO 3-N, in that water. Two ways of determining
the drainage volume are (1) by using a value for
evapotranspiration (ET) in relation to the
DISTANCE
DISTANCE
Figure 3. Typical water advance curve and fate of
applied water for furrow irrigation. The dashed line
indicates the root zone.
amount of water applied, and (2) by using an
estimation of the leaching fraction as a portion
of the water applied. In most cases in California,
adequate ET data are lacking so the second
method is more generally used. The drainage
volume by this method is calculated as
D = IWapPLF [3]
where D is the drainage volume expressed as
surface cm, IWapp is the volume of water
applied expressed as surface cm and LF is the
proportion of applied water moving below the
root zone. The LF can be approximated by
LF=C1I/C1D [4]
where Cl I is the choloride concentration in the
irrigation water corrected for fertilizer inputs
and crop removals and C\D is the chloride
concentration in the drainage water. This
technique for determining LF has been dis-
cussed previously by Pratt, etal.( 1972,1977). By
combining equations [3] and [4] the drainage
volume can be calculated as
D = IW
Cl
app
C1D
[5]
In using this equation one must have
reliable data on the water inputs and chloride
concentrations in applied water, fertilizer,
harvested crop, and drainage water.
First, let's consider the application of water
to a field of interest. If received as rainfall, the
distribution over most fields will be uniform
except where runoff occurs due either to slope or
to rainfall intensity. But for most situations,
especially in the arid west, the rainfall can be
ignored and the uniformity of irrigation water
appliction is much more important. Sprinkler
and furrow irrigation are used for most field and
vegetable crops in California and they will be
considered here. Drip irrigation by which water
is applied at specific points is being used
extensively on tree crops but will not be con-
sidered here.
The uniformity of application of water by
sprinklers has been discussed by numerous
authors (Christiansen, 1942; Criddle, 1965;
Chaudhry. 1976). Factors that affect the uni-
formity of application include system layout,
wind patterns, and surface configuration. In
field crops where a sprinkler system is not
permanent and surface gradients are very
small, it is very difficult to generalize as to
specific parts of a field that may receive more
water than another part. When considering a
49
-------
NITROGEN IN RETURN FLOWS
steady state system, where the management
has been relatively uniform over a period of
years, it may be reasonable to assume that the
variations in application will even out over time
(Kruse, et al. 1962).
A quite different situation exists in the case
of furrow irrigation. With furrow irrigation,
water is always applied in the same direction
and this has implications relative to uniformity
of application. Figure 3 shows a generalized
example of water applied during furrow irriga-
tion in relation to distance from the source of
water (Willardson and Bishop, 1967: Merriam,
1969). In this example, it is assumed that the
soils are uniform throughout the length of the
furrow. As shown, more water is applied at the
upper end of the field than at the lower end. This
amount is increased with decreased efficiency of
the irrigation system. Thus, with furrow irriga-
tion there is a built-in "nonuniformity" as it
applies to water application. This nonuniformi-
ty should be more predictable for a field with
uniform soils than for one with a wide variety of
soils. In comparing furrow and sprinkler irriga-
tion, Kruse, et al. (1962) found that 28"o less
water was added to the root zone of the
downstream 1/4 of sorghum plots than to the
upstream 1 4. Water distribution variations
occurred more at random on the sprinkled plots.
If one considers the effects of furrow irriga-
tion on drainage volumes at various sites within
a field, possibly some of the variation in D can
be anticipated. From equation [4] as the LF
decreases, the chloride concentration in the
drainage water will increase with the constant
chloride input. From Figure 3, it is evident that
IB T
8 •
CL = B.Ill IS D + 2 43
r = .ES „
F = IB.73
IBB ZBB
DlSTHNCE FRDH HRTER SOURCE - METERS
3BB
Figure 4. Relationship between distance from water
source and chloride concentrations in the soil solu-
tion below root zone of fields shown in Fig. 2. These
fields are furrow irrigated.
i m 2m
DI5TRNCE FROM HBTER 5DURCE - METERS
300
Figure 5. Relationship between distance from water
source and chloride concentrations in the soil solu-
tion below root zone of a sprinkler irrigated field.
Distances measured from point where water would be
applied if furrow irrigated.
for furrow irrigation the LF would decrease with
the distance from the water source. This has
been the case for several sites that we have
investigated in the Santa Maria Valley. The
relationship between Cl D and distance from the
water source is shown for three fields in Figures
4 and 5. The fields represented in Figure 4 are
furrow irrigated, while the field represented by
Figure 5 is sprinkler irrigated. For the furrow
irrigated fields where the soils are reasonably
uniform sandy loams, there is a significant
increase in Dlrj as the distance from water
source increases. For the sprinkler irrigated
field, a relation between Dlpj and distance
would not be expected and that is the case for the
field shown in Figure 5.
Another point to consider in relation to
calculating drainage volumes from equation [5]
is that Clj the choloride concentration in the
irrigation water, needs to be adjusted for other
inputs of chloride such as in fertilizers or
organic amendments and losses in the
harvested crop. For example, if KCl is added as
a fertilizer or manure is spread on the field these
inputs need to be considered. In our studies in
the Santa Maria Valley, we have also found that
some crops remove significant amounts of
chloride. Depending on the amount of salt
added in the irrigation water relative to that
removed by the crop, the error in calculating D
by equation [5] can be quite large if the crop
removal of chloride is ignored. The % error can
be expressed as
x 100
[6]
50
-------
NITRATE LEACHING VARIABILITY
where D0 is the calculated drainage volume
assuming no crop removal of chloride and D cis
the calculated drainage volume after making
the appropriate correction for crop removal of
chloride. By appropriate substitution this equa-
tion becomes
% Error =
x 100
[7]
CliVK-Cr
where C r is the amount of chloride removed per
unit area, Clj is the chloride concentration in
the irrigation water and V is amount of water
applied in ha-cm and K is the conversion factor
to convert chloride concentration to an amount
per unit area. Plots of this equation for typical
water application rates and chloride concen-
trations are shown in Figure 6. In considering
this figure in connection with typical crop
removal of chloride for various crops (Table 6),
one can see that in many cases the crop removal
cannot be ignored. For example, if 60 ha-cm of
water (1.5 meq Cl/1) were applied to sugar beets,
the drainage volume would be overestimated by
53% if the crop removal of 112 kg/ha was not
considered.
Like nitrogen removals that were discussed
previously, chloride removals can be quite
variable at different locations within a field.
Significant differences were found between
sites within the three fields discussed previously
(Tables 3,4, and 5). These data also demonstrate
the large differences between crops.
The points discussed here show that many
factors need to be considered when using equa-
TABLE 6
Amounts of Chloride Removed by Various Crops
in Santa Maria Valley.
Chloride
Crop Chloride Yield removed Clj
Artichokes
Broccoli
Cauliflower
Celery
Lettuce
Lima Beans
Peppers
Potatoes
Sugar Beets
%
1.10
.66
.34
3.20
2.48
.08
.49
.36
.81
metric T/ha-l-
7.4e
7.8 e
11. 2e
67.0 e
56.0m
2.4m
18.6m
40.7m
86.3m
kg/ha
13
7
4
129
34
2
28
26
110
mg/1
1.6
1.6
1.3
4.8
4.8
1.4
1.3
1.6
2.3
+e—yield was estimated
m—yield was measured
tion [5] to calculate drainage volumes. Some of
these are apparent but others are more obscure.
Physical losses (P): As with the mis-
cellaneous additions of nitrogen, the mis-
cellaneous losses are important only in specific
cases. If nitrogen is applied as NH3, then
volatilization needs to be considered. When
nitrogen is applied in the irrigation water,
nitrogen lost in runoff must be taken into
account. Losses by erosion may occur where
steep surface gradients exist but generally they
can be ignored.
Dentrification (DN): Data obtained directly
for loss of nitrogen by denitrification for field
situations are generally lacking. Most
denitrification values assigned to soils, man-
agement or crop factors are obtained by
difference (equation [2]). Thus, the variability of
other factors (L, C, etc.) will affect the
magnitude of DN. Presently, extensive research
is being conducted to determine the
relationships between denitrification levels and
soil and management factors. With the use of
isotopes of nitrogen and field gas
measurements, data should be available in the
near future on the variability of DN within
fields. These data will also enable one to check
the "balance" of nitrogen within a system and
none of the factors will have to be calculated by
difference.
Implications
What does this research indicate relative to
irrigation return flow quality management?
One of the main implications of the research
reported here, along with previous work com-
pleted on our nitrate project, is that the use of
soil sampling to monitor nitrate concentrations
leaching below root zones is probably cost-
prohibitive. To obtain reliable data for nitrates
leaching below root zones, the variability of the
factors controlling the pool of nitrogen
available for leaching must be taken into con-
sideration when designing a sampling pro-
gram. As pointed out in this paper, crop removal
of nitrogen will vary over a field. Sampling must
include sites where crops have removed large
amounts of nitrogen and where crops have
removed small amounts. Only by studies
preliminary to a monitoring program can these
locations be identified.
Type of irrigation systems will also in-
fluence the design of a sampling program. A
stratified design may be necessary for furrow or
51
-------
NITROGEN IN RETURN FLOWS
border irrigation systems while a random
design may be satisfactory for fields irrigated
by sprinklers. Analysis of our data along with
literature relative to irrigation systems would
indicate that when sampling fields irrigated by
furrow or border methods, sites in close proximi-
ty to the water source should be sampled along
with positions at greater distances from the
water source. This is because of the lesser
quantities of water applied at the lower end of
fields irrigated by these methods.
H - I £ HCB/L, EB Hfl-CM
E - 1.5 ' IZB '
C - H.I ' BB "
- H H • I2B
MB 6B BB IBB
CHLORIDE REMOVED - KB/HR
Figure 6. Relationship between chloride removed by
harvested crop and "n error in drainage volume if Cl j
is not corrected for crop removal of chloride.
Soil variation also needs to be taken into
account when sampling. The effects of soil
characteristics on plant growth, water trans-
mission and retention characteristics and
denitrifiction potential will all be important in
determining the magnitude of the nitrogen pool
available for leaching. Even within a single
mapping unit, considerable variation exists and
one sample from the center of a soil body will not
adequately represent that mapping unit. Also,
soil properties traditionally used for mapping
soils are only indirectly related to water
transmission characteristics. These character-
istics are very important in relation to water
application vs. field position, period of satura-
tion following irrigation, and leaching volumes.
Sampling programs must include these
variations.
After all of these variations are considered
when designing a soil sampling program, one
often finds that a very intensive sampling
program is necessary. Data presented by Pratt,
et al. (1977) for 4 fields (25-90 ha) in Ventura,
Kern, and Madera Counties, showed that 5 to 21
sites were needed (8 samples with depth per site)
to obtain measured mean nitrate concen-
trations that would fall within 30% of the true
mean at the 95% confidence level. To be within
10% of the true mean it would have been
necessary to sample 23 to 161 sites per field.
Only by preliminary studies can these numbers
be determined. However, they can be an-
ticipated to some extent by considering the
factors discussed in this paper.
Monitoring nitrate leaching by soil samp-
ling does not appear to be feasible for basin-wide
studies. As a research tool to study the com-
ponents of the nitrogen cycle it has proved
useful, but we would not recommend it as a
monitoring method. A more positive aspect of
this research has been the accumulation of data
showing the amounts of nitrate that have
leached from root zones and an improved un-
derstanding of the nitrogen cycle and factors
that affect the pathways between various
sources and sinks. For improved nitrogen
management and thereby improved return flow
quality, the magnitude of the nitrogen pool
available for leaching must be minimized. This
can be done by increasing crop uptake efficiency
of applied fertilizers, lower fertilizer application
rates and/or increased denitrification of excess
nitrogen. Through present and future research
these points will be more fully elucidated and
"ideally" better nitrogen management will
result.
ACKNOWLEDGMENTS
Financial support of the Research Applied
to National Needs Division of the National
Science Foundation through Grant AEN-76-
10283 is gratefully acknowledged. Gratitude is
expressed to Richard Elliott, Stephen Whaley,
Fain Sutherland, and Edward Betty for their
assistance on this project.
REFERENCES
1. Adriano, D. C., P. F. Pratt, and K. M.
Holtzclaw. 1973. Comparison of two simple
methods of chlorine analysis in plant materials.
Agronomy J. 65:133-134.
2. Adriano, D. D., F. H. Takatori, P. F.
Pratt, and O. A. Lorenz. 1972. Soil nitrogen
balance in selected row-crop sites in southern
California. J. Environ. Qual. 1:279-283.
3. Allison, L. E. 1965. Organic carbon. In
C. A. Black (ed.). Methods of soil analysis, Part
2. Agronomy 9:1367-1378.
52
-------
NITRATE LEACHING VARIABILITY
4. Ayers, R. S., and R. L. Branson. 1973.
Nitrates in the upper Santa Ana River Basin in
relation to groundwater pollution. Calif. Agric.
Exp. Sta. Bull. 861.
5. Brinsfield, R. B., and J. W. Hummel.
1975. Simulation of a new centrifugal dis-
tributor design. Trans, of ASAE. 18(2):213-216.
6. Chaudhry, F. H. 1976. Sprinkler uni-
formity measures and skewness. Journal of the
Irrigation and Drainage Division, ASCE
102(4):425-433.
7. Christiansen, J. E. 1942. Irrigation by
sprinkling. Calif. Agr. Exp. Sta. Bull. 670.
8. Griddle, W. D., S. Davis, C. H. Poir, and
D. C. Shockley. 1956. Methods of evaluating
irrigation systems. USDA Ag. Hndbd 82.
Washington, D. C.
9. Cunningham, F. M. 1963. Performance
characteristics of bulk spreaders for granular
fertilizer. Trans, of ASAE 6(2):108-114.
10. Day, P. R. 1965. Particle fractionation
and particle size analysis. In C. A. Black (ed.).
Methods of soil analysis, Part I. Agronomy
9:545-567.
11. Fried, M., K. K. Tanji, and R. M. Van
De Pol. 1976. Simplified long term concept for
evaluating leaching of nitrogen from agricul-
tural land. J. Environ. Qual. 5:197-200.
12. Hanawalt, R. B. 1969. Soil properties
affecting the sorption of atmospheric ammonia.
Soil Sci. Soc. Amer. Proc. 33:725-729.
13. Kruse, E. G., P. E. Schleusener, W. E.
Selby, and B. R. Somerhalder. 1962. Sprinkler
and furrow irrigation efficiencies. Ag. Engr.
43(ll):636-639, 647.
14. Luebs, R. E., K. R. Davis, and A. E.
Laag. 1973. Enrichment of the atmosphere with
nitrogen compounds volatilized from a large
dairy area. J. Environ. Qual. 2:137-141.
15. Malo, B. A., and E. R. Purvis. 1964. Soil
adsorption of atmospheric ammonia. Soil Sci.
97:243-247.
16. Merriam, J. L. 1969. Irrigation
methods: Adaptability and efficiencies of sur-
face methods. Paper presented at the
Agricultural Water Conservation Conference.
California Department of Water Resources.
Davis, CA. June 23-24, 1969.
17. Pratt, P. F., L. J. Lund, and J. M. Rible.
1977. An approach to measuring leaching of
nitrate from freely drained irrigated fields. In
Nielsen, D. R., and J. McDonald (eds.). Nitrogen
and the Environment. Academic press.
18. Pratt, P. F., W. W. Jones, and V. E.
Hunsaker. 1972. Nitrate in deep soil profiles in
relation to fertilizer rates and leaching volume.
J. Environ. Qual. 1:97-102.
19. Shipman, G. E. 1972. Soil Survey of the
Northern Santa Barbara Area. USDA-SCS.
Washington, D. C. 182 pp., plus maps.
20. Willardson, L. D., and Bishop, A. A.
1967. Analysis of surface irrigation application
efficiency. Journal of the Irrigation and
Drainage Division, ASCE. 93(2):21-36.
-------
Field Measured Flux of Volatile
Denitrification Products as
Influenced by Soil-Water Content
and Organic Carbon Source
D. E. ROLSTON, D. A. GOLDHAMER, D. L. HOFFMAN,
and D. W. TOY
Department of Land, Air, and Water Resources, Soils and Plant Nutrition Section,
University of California, Davis, California
ABSTRACT
The amount of NO3' in irrigation return
flow waters is dependent upon each of the
components of the N cycle in soils. One of those
components for which absolute amounts and
rates are not well known is denitrification.
Volatile denitrification products, primarily
WpO and N2, ore evolved whenever anoxic
sites develop within the soil and when sufficient
carbon is available. Absolute amounts and
rates of denitrification from a Yofo loam field
profile at Davis, California, were studied in
relation to the influence of soil-water content
and organic carbon source. Field plots were
intensely instrumented with soil atmosphere
samplers, soil solution samplers, and ten-
siometers. Soil-water pressure heads (h) in the
upper 15 cm of soil were maintained constant at
-15 and -70 cm of water. Three levels of soil
carbon were evaluated from plots cropped with
ryegrass, uncropped plots, and plots to which
manure was mixed in the top 10 cm of soil.
Fertilizer was applied at the rate of 300 kg N
ha-1 as KNOs enriched with 20 and 40% 15N
for the h= -15 and h= -70 cm treatments,
respectively. The flux of volatile gases at the soil
surface was measured from the accumulation of
N2O and 15N2 beneath an air-tight cover
placed over the soil surface for Ior2 hours per
day. Denitrification occurred in order of
decreasing magnitude in manure (h = -15 cm),
manure (h = -70 cm), uncropped (h = -15 cm),
and uncropped (h = -70 cm)plots. Approximate-
ly 70% of the fertilizer nitrogen was denitrified
for the manure (h - -15 cm) treatment. Ap-
proximately 1 % of the added fertilizer was
denitrified in the uncropped (h = -70 cm) treat-
ment.
INTRODUCTION
The amount of NO 3" reaching the ground
water of irrigated lands is dependent upon each
of the components of the N cycle in soils. The
amount of fertilizer N applied and crop uptake
of N are easily measured parameters of the total
balance. The other components of the N
balance, however, are not as easily measured or
determined. The other components of the N
balance has been a subject of much study and is
the primary concern in terms of NO 3 "in ground
water. The natural spatial variability of the
leaching component has been demonstrated to
be large (Biggar and Nielsen, 1976). The other
components such as residual soil N, denitrifica-
tion, and NHs volatilization losses are also not
easily measured. The residual soil N is difficult
to measure due to the large organic N pool of
most soils and complicated by the natural
spatial variablily of that pool (Rolston, 1977).
The transient processes of immobilization and
mineralization within the large organic N pool
also contribute to the complexity of the measure-
ment. Volatilization loss of fertilizer as NH 3
occurs to varying degrees in calcareous soils.
This component can generally be minimized by
incorporating NH4+ fertilizer below the soil
surface. Another loss of N from the soil system
for which absolute amounts and rates are not
well known is denitrification. Volatile
denitrification products, primarily N£O and
N 2, are evolved whenever anoxic sites develop
within the soil and when sufficient carbon as
55
-------
NITROGEN RETURN FLOWS
supplied by soil organic matter, plant materials,
and manure is available.
Simulation models of the N balance in soil
systems attempt to predict the amount and
concentration of NC>3" in irrigation return flow
water as a function of irrigation and cropping
practices (Mehran and Tanji, 1974; Donigan
and Crawford, 1976; Shaffer et al., 1976; Tanji
and Gupta, 1977; and van Veen, 1977). In
general, the denitrification component of the
various mathematical models has not had ade-
quate input data especially for the rates of
denitrification. Total denitrification of applied
fertilizer is used quite frequently such as 10-15%
of the fertilizer N applied (Fried et al., 1976).
Very few experiments have evaluated the
absolute amounts and rates of denitrification in
the field. Rolston et al. (1976) demonstrated that
the volatile gases from denitrification could be
measured in a field profile. Total denitrification
from gas fluxes compared reasonably with
denitrification determined by difference for a
small, intensely-instrumented, field plot. Total
denitrification was determined by integrating
with time the flux of the gaseous denitrification
products as determined from measured soil
gaseous diffusion coefficients and concentra-
tion gradients. However, these studies only
evaluated the amount of denitrification under
one cropping or carbon input system and one
soil-water content near saturation. The objec-
tives of this paper were to measure the absolute
amounts and rates of denitrification from a field
profile at Davis, California, as influenced by
soil-water content and organic carbon source.
MATERIALS AND METHODS
The field plots were located on Yolo loam
soil (Typic Xerorthents). The average soil temp-
erature at the 5-cm depth was 23°C during the
experimental period. Six l-m^ plots were es-
tablished with a 60-cm deep redwood barrier
around each undisturbed block of soil. Three
plots were constantly maintained at a soil-water
pressure head (h) of -15 cm and three plots were
maintained at -70 cm of water. In order to
establish different carbon treatments within
each of the two water regimes, two plots were
cropped with perennial ryegrass (Lolium
perenne) for approximately 4 months prior to
the experiment, two plots remained uncropped,
and manure was mixed in the top 10 cm of soil of
two plots approximately 2 weeks before fer-
tilizer was applied. Each of the six plots were
instrumented with tensiometers, soil solution
samplers, soil atmosphere samplers, and ther-
mocouples. Five soil atmosphere samplers were
installed at the 2-, 5-, 10-, and 15-cm depths.
Three soil atmosphere samplers were installed
at the 20-, 30-, 60-, 90- and 120-cm soil depths.
Triplicate samplers, designed to function as a
tensiometer or solution extractor, were installed
at the 60-, 90-, 120-, 150- and 180-cm depths of the
soil profile. Triplicate tensiometers were also
installed at the 5- and 15-cm depths. Duplicate
thermocouples were installed at the 5-cm depth.
Soil solution samplers consisted of porous cups
glued to polyvinyl chloride tubing. Soil solution
samples were obtained by evacuating bottles
connected to the samplers. Soil atmosphere
samplers consisted of 0.1-cm inside diameter
nylon tubing glued into a 5-cm long, perforated,
acrylic plastic tube. For the deeper soil depths,
the small diameter nylon tubing was placed
inside a 1.3-cm diameter polyvinyl chloride tube
and the nylon tubing glued into a milled plastic
tip. For all samplers, the volume of the sampling
tubes was very small. Soil atmosphere
samples were taken in 1 ml increments and
N2O, CO2, O2, and N2 analyzed by gas
chromatography. Another 0.5 to 1.0 ml of gas
was taken to determine 15 N 2 with a mass
spectrometer. All gas samples to be analyzed by
mass spectrometry were pulled through
ascarite, dehydrite (Mg (CIO 4) 2), and a com-
mercial 0 2 scrubber.
Nitrate solution equivalent to 300 kg N ha'1
was uniformly applied to the plots by applying
approximately 3 ml of solution to each of 400
points at the soil surface. The 15 N enrichment
was 20 and 40% excess for the h = -15 cm and
h = -70 cm treatments, respectively.
Within 6 hours after applying the fertilizer
solution, an air-tight cover was placed over the
plots. Samples of the atmosphere beneath the
cover were taken after 1 or 2 hours for 15 N 2 and
15N2O determinations. Soil atmosphere sam-
ples from within the soil profile were also taken
soon after applying the fertilizer. Both soil
profile samples and samples from beneath the
cover were taken daily for a few days after the
fertilizer was applied and then at less frequent
intervals unitl 15N2 and 15N2O could no
longer be detected above background.
Soil solution samples were taken at weekly
intervals. The grass of the cropped plots was cut
periodically and the total clippings dried for
analysis.
56
-------
FLUX OF DENITRIFICATION PRODUCTS
RESULTS
Concentrations of 15^2 gas from
denitrification of the added fertilizer as a func-
tion of soil depth at 2 days after fertilizer
application for the plots of the h = -15 cm
treatments are given by Fig. 1. The concentra-
tion profiles demonstrate the enormous
differences in N2 production among the
cropped, manure, and uncropped plots. The
differences in anoxic development were also
evident from differences in concentration of 0 2
within the plots. For the cropped and manure
plot, the minimum 0 2 concentration was 9.0 and
2.7% at the 20- and 10- cm depths, respectively.
Concentration of O 2 within the uncropped plots
became no smaller than approximately 15%. It
appears from Fig. 1 that the greatest denitrifica-
tion in the manure plot was occurring in the
upper 10 cm of soil, whereas denitrification in
the cropped plot appeared to be occurring slight-
ly deeper. Similar data for the plots of the h =
-70 cm treatment for 2 days after application of
fertilizer are shown by Fig. 2. There was much
less production of N 2 in the h = -70 cm
treatments than in the h = -15 cm treatments.
Oxygen concentrations in the cropped, manure,
and uncropped plots became no smaller than 17,
15, and 17%, respectively. Similar concentration
profiles were measured for N2O. Since N2O
concentrations did not become great enough to
measure ^N without concentrating the N2O
sample, N 2 O concentrations greater than in-
itial values were considered to be derived from
the fertilizer.
Total flux of the gaseous denitrification
products (N2 + N2O) derived from the fer-
tilizer as a function of time for the two manure
plots are given by Fig. 3. These fluxes were
determined from the accumulation of 1° N 2 and
N£O beneath a cover placed over the soil
surface for 1- or 2-hour intervals. The numbers
below each of the soil-water pressure head
treatments are total denitrification as deter-
mined from the area beneath each of the flux
versus time curves. The flux calculated from the
concentration increase beneath the cover was
corrected for the decrease in the concentration
gradient with time as the gases accumulated.
The decrease in the concentration gradient with
time would result in an underestimation of the
diffusive flux as measured from concentration
changes beneath the cover. A simple correction
for this underestimation of the flux can be made
based upon the steady-state diffusion equation,
f=-Dp(dC/dx) [1]
where f is the gaseous flux, Dp is the soil
gaseous diffusion coefficient, C is gas concen-
tration, and x is soil depth. If it is assumed that
the concentration at the soil surface is equal to
the concentration beneath the cover and that
the concentration at the shallowest sampling
depth (2 cm) does not change with time, the
following boundary conditions apply:
C = C0 , x = 0 cm [2a]
C = C2 , x - 2 cm [2b]
The solution to Eq. [1] and [2] rearranged to
solve for D p is
Dp=-(VL)/(At)ln[(C2-Co)/C2] [3]
where V is the volume of the chamber placed
over the soil surface, L is the depth of soil for
which measurements are taken (2 cm), A is the
cross-sectional area of the soil covered with the
chamber, t is the time after covering the soil at
which the concentration beneath the lid (Co)
was measured, and In is the natural logarithm.
The value of D p was determined from Eq. [3] for
the 0 to 2-cm depth interval from the change in
concentration (C o) of 15 N 2 and N 2 O over a 1-
or 2-hour time period and the measured concen-
tration at the 2-cm depth (C 2 )• The calculated
Dp and the measured concentration gradient
(dC/dx) at t = 0 (assumed to be linear for 0 to 2
cm) were used to calculate the corrected flux
from Eq. [1]. This corrected flux was the best
estimate of the gaseous flux of 15 N 2 and N 2 O
if the cover had not been placed over the soil
surface. Measurements of concentration at the
2-cm depth both before and after the cover had
been in place for 2 hours demonstrated that
Condition [2b] was valid for a 2-hour period. The
greatest correction in the calculated flux was for
the cropped (h = -15 cm) treatment with 6 kg
N ha '1. Corrections for all other plots were less
than 5% of the total denitrified.
The increase in the flux at 1.4 days of the
h = -15 cm treatment (Fig. 3) was from a sample
taken in the afternoon of a fairly hot day. Thus,
the fluxes may be slightly underestimated for a
few days after application due to increased
denitrification as surface soil temperature in-
creased in the afternoon.
Similar data for the flux of N2 and N2O
from the fertilizer as a function of time for the
cropped and uncropped plots are given by Fig. 4.
The scale of the ordinate is considerably
different for the cropped and uncropped plots.
Both Fig. 3 and 4 show that 15N2 and N2O
57
-------
NITROGEN RETURN FLOWS
could not be detected beneath the lid after
approximately 20 days. Denitrification for each
of the six plots occurring as N 2 O, N 2, and total
is given in Table 1. In all cases, the amount of
N2 produced was much greater than N 2 O.
Figure 5 gives soil solution NO 3" concen-
trations derived from the fertilizer as a function
of soil depth at approximately 125 days after
application of the fertilizer to the h = -15 cm
treatments. As expected, the smallest concen-
trations in the soil solution were in the manure
plot because of a high amount of denitrification.
Higher concentrations were measured in the
cropped plot than in the manure plot. The
greatest NO 3" concentrations were measured in
the uncropped plot. For the h = -70 cm
treatments, the NO 3- was just measureable at
the 60- cm depth after approximately 115 days.
Thus, most of the NO 3" for this treatment was
still in the upper 60 cm of the soil profile after
115 days.
DISCUSSION
Gas sampling within the soil profiles de-
monstrated that considerable 15 N 2 was pro-
duced from denitrification of ^ NO 3" fertilizer.
Similar profiles were also measured for N 2 0.
For all plots except the manure plot at h =
-15 cm, the 15 N^ concentrations were nearly
constant with depth or decreased gradually
near the surface. For the manure (h = -15 cm)
treatment, however, the greatest concentration
of 15 N 2 and N20 were measured at the
shallowest sampling depth of 2 cm. Thus, con-
siderable denitrification occurred very near the
soil surface. The largest denitrification rates
generally occurred within 6 to 24 hours after
applying NO3". For the relatively high soil
temperatures at which these experiments were
conducted, denitrification began very quickly
with large initial rates. The generally rapid
decrease in denitrification with time was due to
decreasing NO 3" concentrations in soil solution
and the displacemnt of NO 3" into less anoxic
zones. Considerable 0 2 could be measured from
samples taken within zones where denitrifica-
tion was occurring. Thus, there must have been
sites or "pockets" within the soil which were
anoxic due to high microbial activity and small
02 diffusion. In general, the greatest anoxic
development occurred in the upper 60 or 90 cm of
this profile (Rolston et al., 1976).
The addition of manure to the soil greatly
increased the rate and total denitrification. The
addition of crop residues would also increase the
denitrification potential. The presence of living
plants greatly increased the amount of denitri-
fication as shown by the differences in N 2 and
N 2 0 flux at the soil surface of the cropped and
uncropped plots. The consumption of 02
through respiration of grass roots and addition
of carbon to the system from sloughed roots
increased denitrification by approximately
three to four times over that of the uncropped
plots at equal soil-water pressures. For the
uncropped plot maintained at a soil-water
pressure head of -15 cm, only 8 kg of N ha "1
were lost as N 2 and N 2 0. Thus, it appears that
very little denitrification of added NO 3" will
occur from uncropped Yolo loam even under the
most adverse soil-water conditions. The max-
imum rate of denitrification in the cropped plot
maintained at h = -15 cm was considerably
less than that reported by Rolston et al. (1976).
The field plot used by Rolston et al. (1976) had
been cropped with perennial ryegrass for ap-
proximately two years before application of
NO3", and the soil may have been slightly
wetter than that of this study since water was
occasionally pulled into the soil atmosphere
samplers. It appears that long-term cropping
with grass may substantially increase the
available carbon of the soil and the denitrifica-
tion potential.
The soil-water pressure of the soil also has a
large effect on denitrification. By decreasing the
soil-water pressure head from -15 to -70 cm,
denitrification in the manure plots was decreas-
ed from 280 kg N ha'1 to 47 kg N ha'1.
Decreasing the soil-water pressure head from
-15 to -70 cm for the cropped plots decreasing the
soil-water total denitrification from 34 to
9 kg N ha "1. Decreasing soil-water pressure in
the uncropped plots decreased denitrification
from 8 to 4 kg N ha'l. The very narrow range
of soil-water pressure for which denitrification
occurs for this particular soil, indicates that
slight manipulation of soil-water content may
be achieved to either increase or decrease denit-
rification depending upon particular objectives.
The large differences in denitrification
among the cropped, manure, and uncropped
plots of the h = -15 cm treatment also resulted
in large differences in the concentration of
fertilizer NO 3" amount of denitrification
resulted in the smallest NO 3' concentrations in
the soil solution. Denitrification in the un-
cropped plot of only 3% resulted in the largest
58
-------
FLUX OF DENITRIFICATION PRODUCTS
NO 3 concentration in the soil solution. The
peak of the NO 3 - pulse for all three plots of the
h = -15 cm treatment was at approximately
140 cm for 120 days after fertilizer application.
On the other hand, for the h = -70 cm
treatments, most of the NO 3* was still in the
upper 60 cm of the soil profile. An estimate of
evapotranspiration was obtained from the
water application rate for the h = -15 cm
treatments, the volumetric water content of the
soil profile, and the rate of movement of NO 3"
through the soil profile. Inasmuch as the
profiles were kept constantly wet, it would be
expected that evapotranspiration would be
similar for all plots. Applying this estimated
evapotranspiration to the h = -70 cm
treatments gave a pore water velocity for the
h — -70 cm treatments of approximately 0.5 cm
day' 1 as compared to 1.1 cm dayl for the
h = -15cm treatments. From these estimates of
pore water velocity, the peak of the NO 3 - pulse
should have been at about 55 cm at 110 days for
the h = -70 cm treatment. Soil solution
samples taken at 110 days show a slight in-
crease in NO 3" concentration at the 60 cm depth
of the h = -70 cm treatments.
The implication of denitrification on irriga-
tion return flow water quality is that the concen-
tration of NO 3 leaching below the crop root
zone can be partially manipulated through
denitrification of fertilizer. The carbon input
and soil-water content may be manipulated to
either increase or decrease denitrification as
TABLE 1
desired. The overall objective of nitrogen
management, of course, is to maximize fertilizer
use efficiency and minimize leaching of NO 3" to
ground waters. It is obvious from the results
presented here that the addition of an organic
carbon source to the soil such as manure greatly
increased denitrification in the upper portion of
the soil profile. Addition of manure and fer-
tilizer to the soil would be undesirable in terms
of fertilizer use efficiency unless the irrigation
water could be managed to maintain soil-water
pressures in the top part of the profile smaller
than the value for which denitrification oc-
curs (-70 to -100 cm for Yolo loam). A soil with
FROM FERTILIZER (mg N liter'1 soil air)
10 20 30
Day 2
h= -15 cm
• CROPPED
o MANURE
AUNCROPPED
Figure 1. Representative concentration profiles of
N 2 derived from the fertilizer within the soil of the
cropped, manure, and uncropped plots at a constant
soil-water pressure head of -15 cm.
Denitrification as N2O, N£ , and total for the six
treatments as determined from the accumulation of
gas beneath a cover placed over the soil surface for
a 1- or 2-hour period. The application rate was
300kg N ha-1 asKNOa.
TOTAL
Treatment
N/2 0 N2
(kg (kg (kg (%of
Nha'1) Nha'1) N ha'1) fertilizer
Manure,
h — -15 cm
Manure,
h - -70 cm
Cropped,
h = -15 cm
Cropped,
h = -70 cm
Uncropped,
h = -15 cm
Uncropped,
h = -70 cm
9.9
5.4
4.3
1.8
2.1
0.6
198
42
30
7
5.7
3.6
208
47
34
9
8
4
69
16
11
3
3
1
10
20
30
40
5O
60
FROM FERTILIZER (mg N liter"' soil air)
0.2 0.4 O.6
0.8
Day 2
h = - 70 cm
• CROPPED
oMANURE
A UNCROPPED
Figure 2. Representative concentration profiles of
N 2 derived from the fertilizer within the soil of the
cropped, manure, and uncropped plots at a constant
soil-water pressure head of-70 cm.
59
-------
NITROGEN RETURN FLOWS
low organic matter would result in the smallest
denitrification and the largest NO 3" leaching
at times when a crop was not actively taking up
part of the N applied.
The results of this research shows that
denitrification occurs very rapidly after applica-
tion of NO 3" to a wet soil and that denitrifica-
tion rates can potentially be very large. The
growth rate of denitrifying organisms during
wetting of initially dry soil, the amount of
available carbon, soil physical and chemical
properties, and length of irrigation will affect
the amount of denitrification. The irrigation
frequency and soil-water contents before, dur-
ing, and after irrigation should have a large
influence on the absolute amounts and rates of
denitrification. The position of NO 3" pulse in
relation to the soluble or available carbon
source of the profile will also greatly influence
the amount of denitrification. The position of
NO 3" within the profile and the soil-water
content may be manipulated through manage-
ment of irrigation. Irrigation management
practices to minimize NO 3 ~ in irrigation return
flow waters may or may not be the same
practices which would minimize denitrification
to attain maximum fertilizer use efficiency. The
proper management practices to attain minimal
environmental change and maximal fertilizer N
10 20
TIME (days)
30
Figure 3. Total flux of gaseous denitrification
products as a function of time after fertilizer was
applied to the manure plots at two soil-water
pressures (h). The number under each pressure head
value is the total denitrification for that treatment.
efficiency can be best determined from a
thorough analysis of the fluxes of N in the soil
including the volatile products of denitrifica-
tion.
ACKNOWLEDGMENTS
This research was supported in part by the
United States Environmental Protection Agen-
cy (Grant No. R804259-01-1) and the National
Science Foundation-RANN (Grant No.
134733X).
REFERENCES
1. Biggar, J.W., and Nielsen, D.R. 1976.
Spatial variablity of the leaching characteris-
tics of a field soil. Water Resources Research 12:
79-84.
2. Donigan, Jr., A.S., and Crawford, N. H.
1976. Modeling pesticides and nutrients in
agricultural lands. Rept. to U.S. Environmen-
tal Protection Agency, Environmental Protec-
z
o>
tr
Id
10 20
TIME (days)
Figure 4. Total flux of gaseous denitrification
products as a function of time after fertilizer was
applied to cropped and uncropped plots at two soil-
water pressures (h). The number under each pressure
head value is the total denitrification for that treat-
ment.
60
-------
FLUX OF DENITRIFICATION PRODUCTS
tion Technology Series EPA-600/2-76-043, 318
P-
3. Fried, M., Tanji, K. K., and Van De Pol,
R.M. 1976. Simplified long term concept for
evaluating leaching of nitrogen from agricul-
tural land. J. Environ. Quality 5: 197-200.
4. Mehran, M., and Tanji, K. K. 1974.
Computer modeling of nitrogen transfor-
mations in soils. J. Environ. Quality 3:391-396.
5. Rolston, D. E. 1977. "Application of
diffusion theory to measurement of denitrifica-
tion" in Nitrogen and the Environment.
Academic Press, New York, (in press)
6. Rolston, D. E., Fired, M., and
Goldhamer, D. A. 1976. Denazification
measured directly from nitrogen and nitrous
oxide gas fluxes. Soil Sci. Soc. Am. J. 40:259-266.
7. Shaffer, M. J., Ribbens, R. W., and
Huntley, C. W. 1976. Detailed return flow salini-
ty and nutrient simulation model, Vol. V of
Prediction of mineral quality of irrigation
return flow. Rept. To U.S. Environmental
Protection Agency from U.S. Bureau of
Reclamation.
8. Tanji, K. K., and Gupta, S.K. 1977.
"Computer simulation modeling for nitrogen in
irrigated croplands" in Nitrogen and the En-
vironment. Academic Press, New York, (in
press)
9. Taylor, S. A. 1949. Oxygen diffusion in
porous media as a measure of soil aeration. Soil
Sci. Soc. Am. Proc. 14: 55-61.
10. van Veen, H. 1977. Behaviours of ni-
trogen in soil. A computer simulation model.
Ph.D. Thesis, Wageningen, The Netherlands.
SOIL SOLUTION FERTILIZER N (ppm)
20
40
60
80
100
h = -!5cm
CROPPED, day 120
MANURE, day 133
UNCROPPED, day 127
Figure 5. Concentration of fertilizer derived NO 3"
(ppmN) in the soil solution as a function of depth for
the cropped, manure, and uncropped plots at a soil-
water pressure head of-15 cm for approximately 125
days after fertilization application.
61
-------
Soil Nitrate
Concentrations in Corn Plots
Treated with Isotopically
Labeled Nitrogen Fertilizer
F. E. BROADBENT and A. B. CARLTON
Department of Land, Air and Water Resources,
Soils and Plant Nutrition Section,
University of California, Davis, California
ABSTRACT
Soil solution composition has been moni-
tored for 3 years at depths ranging from 30 to
300 cm at an experimental site on Yolo fine
sandy loam where l^N-depletedammonim sul-
fate has been applied at 0,90,180 and 360 kg N
per year. Samples have been taken at ap-
proximately 2-week intervals during the grow-
ing season and less frequently during the winter
and early spring. In unfertilized plots receiving
60 cm/yr irrigation water NO3-N concentra-
tions fluctuated widely near the surface, rang-
ing from 3-4 up to about 35 ppm over a 2-1/2 year
period. At lower depths fluctuations were not as
great, but at 300 cm NO$-N consistently re-
mained above 10 ppm. At 90 and 180 kg-
fertilizer N, the latter level being sufficient to
produce the maximum grain yield, very little
NO 3 -N derived from fertilizer was found below
120 cm. However, at the 360 kg level, fertilizer-
derived nitrate was found to constitute a signifi-
cant fraction of the total at all depths down to
300 cm.
At another site on Hanford sandy loam
where fertilizer rates ofO, 112,224,336,448 and
560 kg N/year were used, soil solution nitrate
concentrations were determined once a year by
analysis of soil cores taken just after harvest. In
this soil of low organic N content most of the
residual nitrate in the soil was fertilizer derived.
Below the 212 kg N level there was little residual
nitrate, but it rose sharply as a function of N
applied at levels above the 212 kg rate.
The data indicate that with good manage-
ment, optimum corn production and minimum
nitrate pollution are compatible goals, but it is
impractical to maintain NO 3 -Nconcentrations
in the soil solution below the 10 ppm Public
Health standard. Unfertilized plots which ex-
hibited severe N deficiency had soil solution
NQ3-N levels above 10 ppm much of the time.
INTRODUCTION
Since nitrate in groundwaters results from
a dymanic, complex and highly interrelated
series of nitrogen transformations and trans-
port processes which are modified by biologi-
cal, chemical and physical properties of soil,
relating concentrations of nitrate in the subsoil
to management practices is not a simple matter.
Not least among the complicating factors is the
gradual and continuous mineralization of soil
organic nitrogen. The size of the organic pool is
larger than is commonly realized. For example,
if it is assumed that a soil has only 0.05%
organic N in the surface meter, mineralization
of 2% per annum would contribute 150 kg N per
hectare in the form of nitrate to the soil system.
Much larger amounts would be produced in soils
of higher organic content, particularly in cases
where the soils are deep and freely drained.
A number of recent articles have dealt with
the relationships of fertilizer and other practices
to accumulations of nitrate in soils, plants and
waters, including some excellent reviews (1,2,3,
4, 5, 6). For the most part, previous work in this
area has had no means of discrimination be-
tween nitrates derived from fertilizer and that
from other sources, including mineralization of
organic N within the soil. The possibility of
63
-------
NITROGEN IN RETURN FLOWS
establishing quantitative relationships be-
tween nitrate concentrations in effluents from
irrigated lands and nitrogen fertilizer manage-
ment has been made more feasible by the
availability of fertilizer materials which are
labeled with stable isotopes. In particular, the
production of l^N-depleted materials in quan-
tities sufficient for full-scale field trials at costs
which are not unreasonable for field operations
has represented a major breakthrough.
The normal composition of nitrogenous
materials as they occur in nature is about 0.366
atom % 15N and 99.634 atom % 14 N. Any
nitrogen-containing substance which differs
significantly from this isotopic composition can
be used as a tracer. Compounds containing less
than 0.005 % 15 N are now available in ton
quantities. If the minimum acceptable
difference from the natural isotopic composition
is taken as 0.01 atom % in order to ensure two
significant figures in calculation, it can be
determined that essentially pure 14 N materials
can be diluted about 30-fold without losing their
isotopic identity. This acceptable range of dilu-
tion is satisfactory for evaluating the con-
tributions of labeled fertilizer to nitrates in soil
solutions, even when taken at considerable
depth.
This paper describes two field trials utiliz-
ing 15N-depleted fertilizer in which measure-
ments of soil nitrates were made and in which
determination of the isotopic composition of the
nitrate samples permitted the computation of
the proportion which was derived from the
applied fertilizer.
EXPERIMENTAL PROCEDURE
In 1973 field experiments were established
at two locations, one at the University of
California in Davis and the other at the
Kearney Horticultural Field Station near
Reedley in the San Joaquin Valley. At the Davis
site a 2.2 hectare tract of Yolo fine sandy loam
was laid out in four blocks with four fertilizer
rates and three irrigation levels. Fertilizer rates
were 0, 90, 180 and 360 kg N as 15N-depleted
ammonium sulfate, all applied at planting time
to an area 4.57 x 6.1 m within each plot. The
irrigation treatments were 20, 60 and 100 cm of
water applied over the growing season at 2-week
intervals in amounts intended to follow the
consumptive use curve. These amounts are
equivalent to 1/3, 3/3 and 5/3 of the evapo-
transpiration requirement for corn at this loca-
tion, based on weighing lysimeter data obtained
in previous years at this site. All plots received a
pre-plant irrigation to bring the plots to field
capacity. The field was instrumented with
porous ceramic suction probes for extracting
soil solution at depths of 30,60,120,180,240 and
300 cm. Two probes were located at each depth
in 24 of the 48 plots, and in the other 24 plots
duplicate probes were placed at 240 and 300 cm
depths. Only the data from the 300 cm probes
are presented in this paper, and the values given
generally represent the mean of 8 replicate
probe samples, although there are some excep-
tions since every probe did not yield a sample at
every sampling date. Samples were collected at
approximately biweekly intervals during the
growing season and less frequently during the
intervening months, when sampling frequency
was determined by rainfall pattern. The soil
solution samples were analyzed for nitrate and
ammonium and the isotopic composition of the
inorganic N determined with a mass spec-
trometer. For all practical purposes the in-
organic N was identical with nitrate-N, since
very little ammonium was found.
Tensiometers were located in the plots at
the same depths as the soil solution probes, and
in addition, wells for neutron moisture meters
were located in each moisture treatment area.
Frequent readings were made throughout the
year.
The plots were planted to corn, with all
above-ground portions of the crop removed at
harvest time. Plant uptake of labeled N was
determined by analysis of grain and stover.
After harvest, soil cores were taken to a depth of
300 cm, two cores in each plot. Increments of 30
cm were analyzed for inorganic N by extraction
with N KC1, and the combined NH| + NOj
in the extract analyzed in the mass spec-
trometer. Concentrations on a soil solution
basis were calculated by using the neutron
meter readings to determine soil moisture at the
time of soil sampling. These values as reported
reflect not only nitrate in the soil solution, but
also absorbed ammonium which is potentially
convertible to nitrate.
At the Kearney site 24 plots, each 0.27 hec-
tare in area, were established on Hanford sandy
loam. Annual fertilizer rates employed were 112,
224, 336, 448 and 560 kg N/ha, of which one
third was applied pre-plant and the other two-
thirds later as a side-dressing. Irrigation was
uniform for the entire area, which received 67
64
-------
SOIL NITRATE CONCENTRATIONS
cm of water during the growing season, This
amount was 10% in excess of the calculated
evapotranspiration. At this location no probes
for sampling soil solution were installed. Soil
samples were obtained after crop harvest to a
depth of 180 cm in 1973, to 240 cm in 1974 and to
420 cm in 1975. Three cores were taken from
each plot, with samples at 30 cm increments,
and composited by depth for analysis. Nitrate
was extracted from the soil samples with water
and determined by Devarda reduction and
steam distillation. Moisture contents were deter-
mined by oven-drying field samples.
As at the Davis site, corn was planted at the
Kearney location, but in the latter case only the
grain was removed, with all other plant parts
returned to the soil after harvest.
RESULTS AND DISCUSSION
The rainfall pattern over the three cropping
seasons at the Davis location is recorded in
Table 1. These amounts may be put in perspec-
tive by noting that the total stored water in the
top 3 meters of soil ranged from about 70 to 120
cm from the driest to the wettest conditions,
respectively.
TABLE 1
Monthly rainfall at the Davis site from July 1, 1973
to January 1,1976.
typical data. Even with 8 replicates the stand-
ard deviations ranged from 36 to 78% of the
mean for total soluble N, and were even higher
for tagged N. These observations emphasize the
difficulties attendant to monitoring soil solu-
tion composition in a field situation, even with a
large number of probes.
TABLE 2
Means and standard deviations of N concentrations
in soil solution samples collected at the 300 cm depth
in plots receiving 100 cm irrigation water.
Month
1973
Monthly total, cm
1974
1975
January
February
March
April
May
June
July
August
September
October
November
December
0
0
.43
3.99
14.73
10.69
8.91
1.47
13.00
2.31
0
0.74
1.65
0
0
3.05
2.26
9.47
0.61
17.91
11.38
1.37
0
0
0.28
0.13
0
6.17
0.68
0.97
Suction Probe Data
Before discussing the results it is ap-
propriate to comment on the reliability of data
obtained by sampling soil solution with suction
probes. The variability among replicates were
considerable, as is shown in Table 2,which
presents means and standard deviations for
samples obtained on two dates in 1975. The
dates were selected at random, simply to show
£_
N level,
kg/ha
0
90
180
360
0
90
180
360
Soluble
N, ppm
7.1
9.4
13.3
31.4
8.4
10.5
14.3
39.7
Std. Dev. Tagged Std. Dev.
ppm N, ppm ppm
July 15, 1975
2.8
4.4
5.9
13.6
August 12, 1975
3.4
3.8
6.3
31.0
—
0.58
1.17
13.8
—
0.81
1.80
19.0
—
0.34
0.65
8.9
—
0.88
2.96
14.1
Concentrations of inorganic N at the 300 cm
depth are presented in a series of graphs show-
ing temporal trends from July 1, 1973 to
January 1, 1976.Where fertilizer was applied,
concentrations of inorganic N derived from
fertilizer are also shown. Since fertilizer was
applied in April or May of each year, the values
shown represent cumulative effects, there being
no way to distinguish between applications in
successive years. The values at 300 cm are
unaffected by plant uptake, since few if any
roots penetrate to that depth, and are probably
representative of solution which eventually
reaches the water table. It may be noted as an
aid to interpretation of the curves presented that
the soil was sufficiently well supplied with
available nitrogen that no yield response to any
level of fertilizer addition was obtained in 1973.
However, in 1974, 90 kg of fertilizer N was
required to produce maximum yield, and in 1975
a yield response was obtained up to 180 kg
additional N. In all three years the yield was
significantly lower in plots receiving only one-
third of the evapotranspiration requirement
(1/3 ET). Reference to Figure 1 shows a general
downward trend in concentrations of inorganic
N in the unfertilized plots receiving 20 cm
irrigation after 1973, but values did not fall as
low as 10 ppm at any time. Where 90 kg N was
applied concentrations went nearly up to 50
ppm in the latter part of 1974, but decreased
65
-------
NITROGEN IN RETURN FLOWS
rather sharply thereafter, reaching a low of
about 16 ppm at the end of 1975. The amount of
fertilizer derived N reaching the 300 cm depth
was almost negligible. In Figure 2 it will be
noted that at 180 and 360 kg fertilizer levels
very sharp increases in soluble N were observed
during the summer of 1974, with values in
excess of 50 ppm N being attained. At the 180 kg
fertilizer rate concentrations declined progres-
sively through 1975, but at the 360 kg rate
concentrations remained high. In the latter
instance, a significant proportion of the total
soluble N was derived from the added fertilizer.
In these plots receiving inadequate water for
maximum crop production high concentrations
of soluble N are perhaps not surprising, but the
minimal contribution of fertilizer to these levels
at the 90 and 180 kg rates is noteworthy.
1/3 ET, NoN
Figure 1.
Similar data for the 1 ET irrigation treat-
ment are presented in Figures 3 and 4. With
more water applied to the soil, the concen-
trations attained were not quite so high as at 1/3
ET, but the same general observations are
applicable. Again, only at the 360 kg N rate did a
significant amount of fertilizer N reach the 300
Figure 2.
cm depth. Data for the 5/3 ET irrigation treat-
ment are presented in Figures 5 and 6. In these
curves the influence of the heavy rains which
fell in March, 1974, are more evident in a
temporary reduction in soluble N concen-
trations. With reference to Figure 6, it will be
noted that concentrations of tagged N in plots
receiving 360 kg N increased progressively
during 1974 and 1975. Figures 3 and 5 show that
in unfertilized plots where water was adequate
soluble N did not approach values as low as 10
ppm until the latter part of 1975, in spite of the
fact that yields were severely depressed by
defficiency of nitrogen. For example, in the 1 ET
treatment area grain yields in the check plots
were 59% of the optimum obtained in 1974 and
only 44% in 1975.
Data from Soil Cores
Concentrations of inorganic N obtained
from soil cores at the Davis site in October, 1975,
are presented in Figure 7. These are expressed
on a soil solution basis for the layer of soil
between 180 and 300 cm depths. As expected,
concentrations were higher in the plots receiv-
66
-------
SOIL NITRATE CONCENTRATIONS
ing least water. Mean values for this 120 cm
thick layer were consistent with probe sample
values in showing concentrations above 10 ppm
N in the unfertilized plots. Concentrations did
not increase sharply with increasing fertilizer
application rates up to the point of maximum
yield, 180 kg N/ha, but increased significantly
with the further addition of N at the 360 kg level.
I ET, NoN
July I
1973
Figure 3.
Jon I
1974
July I Jon I July I
1975
Jon)
1976
The same general conclusions are
applicable at the Kearney site, data for which
are presented in Figure 8. Here the fertilizer-
derived nitrate is also identified. In this ex-
tremely nitrogen-deficient soil nitrate-N was
about 9 ppm on the soil solution basis in the 180-
300 cm soil layer after crop harvest in 1975. In
these plots grain production in the unfertilized
plots was only 18% of that in plots receiving 224
kg N. There was no appreciable contribution of
fertilizer N to nitrate in the 180-300 cm layer at
fertilizer rates of 224 kg and below, but the
concentration increased sharply above this fer-
tilizer rate, with concentrations in excess of 100
ppm being attained at the 560 kg N rate.
It might be argued that the lack of fertilizer
N in the soil solution at fertilizer rates below the
optimum is due to preferential utilization of
fertilizer N by the crop, thereby leaving more of
the soil N available for leaching. While it is true
that biological exchange involving microbial
activity undoubtedly occurred to some degree.
resulting in some fertilizer N being immobilized
and some soil N mineralized, the plant uptake
data presented in Tables 3 and 4 show that
utilization of soil N was actually enhanced by
fertilizer additions up to the levels of optimum
yield, and even beyond in several instances. The
optimum yield at the Davis site in 1973 was
obtained with no fertilizer, with 90 kg N in 1974
and with 180 kg N in 1975, whereas the optimum
rate at the Kearney site was 212 kg in all three
years.
I ET, 180 kg N
Totol
Figure 4.
The data indicate that with careful manage-
ment of irrigation water and fertilizer rates,
maximum corn production on these soils is
compatible with minimum nitrate pollution
hazard. However, it is not realistic to attempt to
maintain nitrate-N concentrations in the soil
solution below 10 ppm. In any event, total mass
flow of nitrate below the root zone is obviously a
much more important consideration than con-
centration per se.
67
-------
NITROGEN IN RETURN FLOWS
5/3 ET, No N
10
Figure 5.
July!
1973
Figure 6.
Jonl
1974
July!
Jonl
1975
Julyl
Jonl
1976
Figure 7.
50 100 150 200 250 300 350 400
FERTILIZER N APPLIED, kg/ho
TABLE 3
Uptake of soil N by crops as affected by fertilizer rate,
Davis site.
Pert.
rate-
1/3 ET
1 ET
5/3 ET
1974 1975 1973 1974 1975 1973 1974 1975
kg/ha
0 155 83 64 155 120 90 145 105 77
90 139 99 102 148 134 108 136 133 100
180 110 93 105 108 115 124 103 117 101
360 95 97 107 84 87 82 75 82 83
0
Figure 8.
IOO 200 300 400 500
FERTILIZER APPLIED, kg/ho
TABLE 4
600
Uptake of soil N by crops as affected by fertilizer rate,
Kearney site.
Fertilizer rate
kg/ha
0
112
224
336
448
560
1973
34
63
80
66
57
53
1974
kg /ha
35
36
37
25
23
18
1975
40
59
69
60
43
44
68
-------
SOIL NITRATE CONCENTRATIONS
ACKNOWLEDGMENT
This work was supported by Grants GI
34733X, GI 43664, and A EN74-11136 A01 of the
National Science Foundation.
REFERENCES
1. Adriano, B.C., Pratt, P.F., and
Takatori, F.H. 1972. Nitrate in unsaturated zone
of an alluvial soil in relation to fertilizer
nitrogen rate and irrigation level. J. Environ.
Quality 1: 418-422.
2. Bower, C.A. and Wilcox, L.V. 1969.
Nitrate content of the upper Rio Grande as
influenced by nitrogen fertilization of adjacent
agricultural lands. Soil Sci. Soc. Amer. Proc. 33:
971-973.
3. 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 move-
ment of nitrate through soil profiles. Soil Sci.
Soc. Amer. Proc. 34: 448-452.
4. Pratt, P.F., Jones, W.W., and Hun-
saker, V.E. 1972. Nitrate in deep soil profiles in
relation to fertilizer rates and leaching volume.
J. Environ. Quality 1: 97-102.
5. Stanford, G., England, C.B., and
Taylor, A.W. 190. Fertilizer use and water
quality. USDA-ARS 41-168.
6. Viets, F.G., Jr., and Hageman, R.H.
1971. Factors affecting the accumulation of
nitrate in soil, water, and plants. Agriculture
Handbook No. 413, ARS-USDA.
69
-------
Theoretical and Experimental
Observations of Water and Nitrate
Movement Below a Crop Root Zone
J. W. BIGGAR, K. K. TANJI, C. S. SIMMONS, S. K. GUPTA,
J. L. MacINTYRE, and D. R. NIELSEN
Department of Land, Air and Water Resources,
University of California, Davis, California
ABSTRACT
A report of the progress of an experiment
which attempts to deal with the spatial
variability of a field soil is presented. The
experimental objective is to measure the flux of
water and nitrate leaching below the root zone
of a crop by examining the behavior of nitrogen
applied to a corn crop grown under irrigation.
A primary objective of this conference and
that of the Environmental Protection Agency is
to be able to interpret periodic measurements of
water and its dissolved constituents at specific
locations at and below the soil surface in
relation to irrigation return flow quality man-
agement. The objective is relatively new to all of
us, regardless of our scientific or engineering
discipline, owing to the fact that we and most
other agriculturists throughout the world have
used the yield of a crop or a sequence of crops as
the foremost measure of good irrigation
management practices. Those measurements
have occasionally been complemented by gross
estimates of water and salinity budgets, partial-
ly substantiated by infrequent monitoring of
the quality of drainage effluents in areas of poor
drainage or salinity hazard. By in large,
measurements of soil and water have generally
been restricted to those within the crop root zone
to ascertain capacities of available water and
plant nutrients and estimates of total salinity or
leaching fractions. Today, owing to a greater
concern regarding the quality and quantity of
our water resources brought about through both
private and governmental sectors, researchers
have quickened their pace to be able to under-
stand what happens below the soil surface, and
below the root zone. We are a long way from a
fully developed technology to measure and
interpret events below the soil surface to
manage irrigation return flow quality. We are
also a long way from a fully developed
technology to mathematically model and inter-
pret measurements made below the soil surface
to manage irrigation return flow quality. This
conference is a testimony to the latter two
statements with research presentations often
falling into one of two classes — (1) Intensive
measurements and/or calculations restricted to
a single location without regard to their inter-
pretation and simulation over larger land areas,
and (2) Extensive measurements and/or simu-
lations for large land areas without adequate
regard to their verification at specific locations.
The former are oftentimes not suited to relative-
ly small rates of change evaluated over long
time intervals — years or decades, while the
latter are oftentimes impractical for short-time
events. Both are potentially affected severely by
the spatially varying properties of the soil. That
is to say, the spatial variability of soils poses a
problem to the researcher in identifying a mean
value of a soil parameter even for a relatively
small experimental site, and because of this
problem, the researcher hesitates to extrapolate
results to larger land areas or give such values
to mathematical modelers with relatively little
appreciation of their experimental uncertainty.
On the other hand, soil-water-solute
measurements taken at a few specific times and
locations may lead rightfully or wrongfully,
owing to spatial variability of soils, to an
acceptance of a particular mathematical model.
The purpose of this paper is to report the
progress of an experiment currently being con-
ducted in an attempt to deal with the spatial
variability of a field soil with the objective to
71
-------
NITROGEN IN RETURN FLOWS
measure the flux of water and nitrate leaching
below the root zone of a crop. The experiment,
which is a part of a larger project partially
supported by NSF, examines the behavior of
nitrogen applied as l^N [n l^N-depleted fer-
tilizer to a crop of corn being grown under
irrigation. The experiment has thus far covered
four growing seasons. The experimental site is
located in Davis, California, on a recent alluvial
fan derived from mixed but dominantly
sedimentary rock sources and is classified Yolo
loam and Yolo silt loam.
There are three irrigation regimes
replicated four times. Hence, there are 12 large
plots which are managed as individually
irrigated units, and within each of these plots,
fertilizer is applied to four subplots as 14N in
15N-depleted NH4SO4 at O, 90, 180 and 360
kg per ha. The water treatments correspond to
1/3, 3/3 and 5/3 of the normal evapotranspira-
tion (ET) requirements of the corn crop as
determined by several years of experience with
corn in the same field. These correspond to
about 20, 60, and 100 cm of irrigation water
applied, respectively. Selection of these
treatments was based on the objective of
providing for three different soil moisture
regimes with corresponding differences in the
flux of water that would drain from the root
zone. Consequently, the 5/3 ET water treatment
would provide excessive water that will drain
out of the root zone during and after the growing
season and leach mobile nitrogen. The 1/3 ET
water treatment does not provide equivalent
water lost by ET and drainage will be minimiz-
ed. Since the Yolo soil is deep and stores a
significant quantity of water provided by winter
rainfall and pre-irrigations, the corn corp will
not suffer severe stress during the growing
season. Drainage from the 3/3 ET water treat-
ment is expected to be intermediate and might
represent normal irrigation practices.
Irrigations are applied at 14 day intervals
providing a quantity of water to each water
treatment commensurate with the estimated
requirements based on evapotranspiration
rates. It should be understood that the experi-
ment does not require applications correspon-
ding exactly to the water loss as calculated by
ET since the treatments hav been established
only as guidelines to achieve three different
moisture regimes.
Soil water pressure measurements are made
using tensiometers placed 30, 60, 120, 180, 240
and 300 cm below the soil surface. Two plots in
each replication of each water treatment have
two tensiometers at each depth and two other
plots in the same replication have two ten-
siometers each at 240 and 300 cm. Thus, there
are 48 tensiometers at each depth except at 240
and 300 cm where there are 96 at each depth for
a total of 432. Figure 1 is a diagram of the
configuration in each water treatment plot.
Measurements of soil water pressure are taken
usually immediately before and after an irriga-
tion and in the interval between irrigations.
There are an equal number of suction probes in
the same configuration for extracting the soil
solution. All instruments are buried at or below
the 30 cm depth so that none protrude above the
soil surface to interfere with cultural practices.
This required a modification in design since all
tubing also is buried within the plot area and
comes to the surface in the roadways between
plots. Two neutron access tubes are located in
each replication of the water treatments as
shown in Figure 1, for a total of 24 locations.
Neutron readings are made at 15 cm intervals to
a depth of 300 cm twice a week during the
growing season and through the fallow season
on a reduced schedule.
Hence, during the past four years we have
obtained thousands of simultaneous measure-
ments of soil water content and soil water
pressure head at the above-mentioned depths at
two locations within each irrigated plot. Before
the data can be used in any constructive way, it
is necessary to examine them in view of the
uncertainty of each value, as well as to
IRRIGATION LIME
O SUCTION PROBES
• TENSIOMETERS
® NEUTRON TUBES
Figure 1. Schematic diagram of an irrigated plot
showing the relative locations of suction probes, ten-
siometers, neutron meter access tubes, and sprinkler
heads.
recognize how to deal with their distribution to
obtain a mean value and/or other useful
measure of their distribution. The uncertainty
stems from instrument and observer error while
72
-------
WATER AND NITRATE MOVEMENT
the distribution of observed values stems, in
addition to that uncertainty, to differences in
soil water regime caused by (1) the rainfall and
irrigation patterns, (2) patterns of water extrac-
tion owing to the corn roots being non-
uniformly distributed within the soil profile
both with respect to space and time, and (3) the
spatially varying properties of the soil as well as
their time-dependent nature owing to hyster-
esis. For the purposes of this discussion, we
shall concentrate on points 2 and 3, without
neglecting point 1, or the error of the measure-
ments.
Let us first examine measurements of the
soil water characteristic (soil water content
versus soil water pressure) taken in four ar-
bitrarily selected locations that were not
cropped but covered with plastic sheets within
the experimental site. These measurements
taken in the absence of corn plants are shown in
Figures 2 and 3 for soil depths 60 cm and 120 cm,
respectively, for only a 30-day period following
the application of a large quantity of water.
Data for soil water content are mean values of
four measurements and those for soil water
SOIL WATER CONTENT S (cm5 cm'3)
SOIL WATER CONTENT 8 (cm5 cm"3)
o
F
I,
m <
33 '
T>
(SI
c
I
m
o
u
rT —
3.8
Figure 2. Soil water characteristic curves measured
at the 60-cm depth in four locations within the ex-
perimental site. The solid lines are given by equation
(1).
Figure 3. Soil water characteristic curves measured
at the 120-cm depth in four locations within the ex-
perimental site. The solid lines are given by equation
(1).
pressure head are mean values of three mea-
surements. Notice that the curves in Figure 2 for
plots 1 and 3 are nearly identical while that for
plot 2 and that for plot 4 show progressively less
water retained at foil water pressure heads of
-200 cm. Similarly, .'n Figure 3, for the 120-cm
depth, the curve for piot 1 is completely different
from those for plots 2, 3 and 4. The solid line in
each graph is
= a{exp[p(0-0s)-l]) [1]
where p equals bfl g "1 , b is a parameter common
to all locations, 0g denotes the saturated water
content, and a is a parameter the value of which
depends upon each location. These differences
in soil water characteristic curves shown in
Figures 2 and 3 also exist at the other soil
depths. Even greater differences exist within
and between plots illustrated in Figure 1. The
net result is that the natural spatial variability
of a soil precludes an assumption that soil water
regimes in plots treated identically will behave
similarly. It demands that a special effort be
made to account for point measurements being
biased away from the norm owing to localized
soil properties in order to integrate these mea-
73
-------
NITROGEN IN RETURN FLOWS
surements in such a manner as to represent the
mean over the total area under consideration —
an experimental plot, four plots all treated the
same, or an entire field.
We demonstrate in Figure 4 a scaling
technique similar to that suggested by Warrick,
Mullen and Nielsen (1977) to coalesce the data of
Figures 2 and 3 into a single curve. Full details
will be available in a publication of Simmons,
Biggar and Nielsen (1977). All data points for
each of the graphs of Figures 2 and 3 are given
in Figure 4 with the solid line described by
h(S) = am(exp[b(S -!)-!]) [2]
where S equals 6/6 s, am is the scale mean
coefficient which holds for all locations, and b is
that parameter given in Equation [1]. The value
of a given in equation (1) for a specific location is
that of am divided by the scale factor a
explained in Simmons et al. (1977). Figure 4
illustrates that the data from Figures 2 and 3
can be coalesced into a single curve satisfying
equation (2) that can be used in mathematical
analyses for a specific location within a plot (for
a particular value of a) or for a much larger area
if the mean value of a is known. In a similar
fashion, the hydraulic conductivity K can be
coalesced by the equation
K = a2Km exp[c(S-l)] [3]
where Km is the scale mean water-saturated
hydraulic conductivity which holds for all
locations and K o (equal to a 2 K m ) is the water-
saturated hydraulic conductivity which
.,
s
:
:
.:
i-
I
:
08
06-
h *-36.2 (exp [-5.64(5-1 )]-l}
• 60 cm DEPTH
o 120cm DEPTH
J
_
0 -100 -200
SOIL WATER PRESSURE HEAD h (cm)
Figure 4. Scaled soil water characteristic curve for
the data given in Figures 2 and 3. Solid circles and
open circles represent the 60 and 120 cm depths, re-
spectively.
E
-
oe
UJ
o O.I
C/5
T~
120 cm DEPTH
o
VK0= 2.24 cm/day
o
o o
10 20
TIME (days)
3C
Figure 5. Calculated and measured values of soil
water flux at the 120 cm depth of plots 1 and 2.
dependent upon location, and c is a parameter
common to all locations. Hence, it appears that
the soil water properties [0(or S) versus h and K
versus 0(or S)] may be scaled to provide the basis
of calculations of soil water transport and
retention for any area desired provided that
appropriate values of the scale factor are
available. Figure 5 shows measured values as
well as calculations of the soil water flux at the
120-cm depth in plots 1 and 2 with the value 35.3
for the parameter in the equation being derived
from the average value of a for the four plots.
Using the thousands of paired values for d
and h stemming from measurements with
neutron meters and tensiometers taken in the 12
plots during the past four years, we have
ascertained 72 values of a (6 soil depths times 12
plots) which represent the experimental site.
These values, shown in Figure 6 as a cumulative
probability graph, are not normally distributed,
but skewed with a few values exceeding 5 lying
to the right off of the graph. The a values are
better described as a log-normal distribution
having a mean of 0.91 and a mode of 0.38.
The impact of the spatial variability of the
soil water properties reflected through the
values and distribution of a is illustrated in
Figures 7 and 8 where the soil water flux at the
240 cm depth and the average water content
from the soil surface to that same depth are
given as a function of drainage time in the
absence of corn roots for values of a equal to 0.4
and 2.0. During early stages of drainage, values
7 !
-------
WATER AND NITRATE MOVEMENT
of the soil water flux differ by ten-fold and after
20 days of drainage, the soil location
characterized by an a equal to 0.4 contains 20
cm more water in the profile to the 240-cm depth
than the soil location characterized by an a
equal to 2.0. These values are typical of those
manifested within an experimental plot. It is
readily apparent that a few measures or es-
timates of K, 0, and/or h within field plots may
easily lead to unreliable conclusions, and that a
thorough analysis of the inherent homogeneity
(or heterogeniety) of the site is required before
meaningful mass balances or rates of water and
nitrate transport may be simulated (or verified)
by mathematical models.
99-
05
90-
.
701-
50-
30-
Figure 6. Cumulative probability graph of i'2 val-
ues of o (and In «) measured at six soil depths in each
of the twelve irrigated plots cropped to corn.
In the presence of a crop, the movement of
water and solutes in a soil profile reflect soil
properties dealt with above, as well as space and
time dependent nature of the crop properties. We
include both of these properties in a simulation
model (Gupta et a/., 1977) to illustrate the
importance of recognizing and quantifying the
spatial variability of soils on the magnitude of
the soil water flux within and below the root
zone of the corn crop in our experiment. It is not
our intent here to justify the salient features of
the computer program used to simulate the soil
water movement, but we do wish to make clear
the concepts included in the calculations.
Features above the soil surface included (a) the
soil water flux during sprinkler irrigation was
10 20
TIME (days)
i
Figure 7. Soil Water flux at the 240 cm depth versus
drainage time after a heavy irrigation in the absence
of plants calculated using two sets of values of K 0
and a.
0.45
AVERAGE 9 OVER 0 TO 240cm DEPTH
:
= 04-0.029 ln(! + 348 K0t/240)
0 = 06lcm/day
a =0.4
.
-
I
:
0.35-
025
10 20
TIME (days)
Figure 8. Average soil water content from the soil
surface to the 240-cm depth corresponding to the con-
ditions given in Figure 7.
0.5 cm per hour, (b) lysimetric evapotranspira-
tion measurements for corn in the same ex-
perimental area were appropriately propor-
tioned into soil surface evaporation and poten-
tial transpiration based upon leaf area indices,
and (c) actual transpiration was linked to poten-
tial transpiration as a function of the soil water
content. Features below the soil surface includ-
ed (a) root growth as a function of days as
measured previously for corn on the same site,
(b) water absorption by roots based on the
application of Darcy's equation and root densi-
75
-------
NITROGEN IN RETURN FLOWS
ty as a function of soil depth and time, and (c) a
water table sufficiently deep to assume a unit
hydraulic gradient far below the root zone.
Figure 9 shows simulations of the soil water
flux at three soil depths from planting time (t = 0
when the soil water content was assumed to be
0.32 cm 3 cm "3) to near harvest (t = 131 days)
for the 5/3 ET (100 cm) irrigation treatment
during the summer of 1975. Calculations are
based on measured values of hydraulic conduc-
tivity and the soil water characteristic for two
scale factors 0.634 and 2.36. The former is larger
than the mode but smaller than the mean of the
distribution given in Figure 6, while the latter
reflects about 10% of the observations expected
in the field.
The seven peaks (in each graph) manifested
in Figure 9 reflect the seven irrigations of
120
E
o
cr
Ld
-------
WATER AND NITRATE MOVEMENT
growing crop. We shall have to incorporate into
the analysis made by Biggar and Nielsen the
various transformations of the nitrogeneous
species as well as their uptake by the corn crop.
ACKNOWLEDGMENT
Financial support from the National Sci-
ence Foundation, the Kearney Foundation of
Soil Science, and the California Agricultural
Experiment Station is acknowledged.
REFERENCES
1. Biggar, J. W. and Nielsen, D. R. 1976.
Spatial variability of the leaching
characteristics of a field soil. Water Resources
Research 12: 78-84.
2. Gupta, S. K., Tanji, K. K., Nielsen, D. R.,
Biggar, J. W., Simmons, C. S., and Maclntyre, J.
L. 1977. Field simulation of soil-water move-
ment with crop water extraction. University of
California, Water Science and Engineering
Paper No. 4013 (in preparation).
3. Simmons, C. S., Biggar, J. W., and
Nielsen, D. R. 1977. Scaling field-measured soil
water properties. Hilgardia (in preparation).
4. Warrick, A. W., Mullen, G. J., and
Nielsen, D. R. 1977. Scaling field-measured soil
hydraulic properties using a similar media
concept. Water Resources Research 13:(in
press).
77
-------
Water Management
-------
Minimizing Salt in Return Flow
by Improving Irrigation Efficiency
JAN VAN SCHILFGAARDE
USDA, Agricultural Research Service,
U.S. Salinity Laboratory,
Riverside, California
ABSTRACT
Return flow from irrigated agriculture has
been identified as the major source of salinity in
the Colorado River that may be controlled.
Thus, if the trend of increasing salinity is to be
reversed, it appears that irrigated agriculture
must bear a large portion of the burden.
Research at the U.S. Salinity Laboratory in-
dicates irrigated agriculture can reduce its con-
tribution by efficient irrigation that provides
water of low salinity in the upper portion of the
crop root zone while the salinity of the water in
the lower portion can be permitted to concen-
trate considerably more than had previously
been suspected without decreased yields. If
these results hold true under field conditions,
the leaching requirement of most crops grown
with Colorado River water could be reduced
below 10%. Lower leaching requirements, if
achieved, would also reduce salt discharge to
the river due to precipitation of lime and gyp-
sum in the soil and because of decreased salt
pickup from saline underground sources.
To evaluate the minimum leaching concept
for alleviating the salinity problem of a major
river basin, two field studies have been initiated
in the Wellton-Mohawk Irrigation and
Drainage District of southwestern Arizona. The
first field experiment was installed in December
1973 in citrus on coarse-textured, mesa soil, and
the second was started in September 1974 in
alfalfa on fine-textured, valley soil.
The paper will describe the experimental
design of both experiments and initial data on
crop water use, soil salinity profiles, and
leaching fractions achieved will be presented.
INTRODUCTION
Irrigation return flow can contribute
significantly to the salt loading of rivers and
other water bodies. For the Colorado River, EPA
(1971) estimated that 37% of the salinity at
Hoover Dam could be attributed to irrigation.
But what can be done through irrigation
management to reduce the salt contribution
from this source?
Three distinct processes may be listed as the
principal causes for increased salinity in irriga-
tion return flows. Seepage water (and to a lesser
degree tail water) may dissolve salts from the
soil or underlying saline strata; seepage water
may displace groundwater of higher mineral
content; and evapotranspiration always
reduces the volume of water that carries the
salts, thereby increasing the concentration. To
the extent that seepage water dissolves salt or
displaces salt water, reductions in the seepage
from canals, laterals and drains will reduce
salinity downstream. Here we restrict the dis-
cussion to the effects of changing the deep
percolation on cropped fields.
Research has demonstrated how both the
amount and the composition of salt dissolved in
the drainage water leaving the rootzone of a
crop can be varied by changing the leaching
fraction (Rhoades et al., 1973). At high leaching
fractions, i.e., when a large partof the irrigation
water applied passes through the rootzone,
water tends to dissolve salts from the soil
minerals at low leaching fractions, some salt
species tend to precipitate. As the leaching
fraction is reduced, by improved irrigation
management, the situation favoring dissolution
gradually shifts to one favoring precipitation.
Consequently, the total salts carried by the
drainage water are decreased. The quantities
involved vary with the situation, but they can be
substantial.
Also, Bernstein and Francois (1973) showed
the alfalfa could be irrigated with a far lower
81
-------
WATER MANAGEMENT
leaching fraction than previously thought
without significant yield reductions. They
postulated that, as long as water of low salinity
was freely available to the plant roots at
shallow depths, the roots in the lower part of the
rootzone could concentrate the soil solution to a
maximum value that could be determined by
extrapolating a plot of relative crop yield vs. soil
water salinity to zero yield. This interpretation
results in required minimum leaching fractions
of about 1/3 to 1/4 those generally recommend-
ed.
Combining the above findings with con-
siderations from soil physics concerning water
flow in soils, we proposed (van Schilfgaarde et
al., 1974) that irrigation management practices
could be divised that would increase irrigation
efficiency and thereby reduce substantially the
amount of salt in the drainage water. Two field
experiments were initiated to test this
hypothesis. This paper gives a progress report
on these field studies. More detailed information
may be found in USSL Staff (1977) and Hoff-
man et al. (1977).
The Setting
The experiments were established in the
Wellton-Mohawk Irrigation and Drainage Dis-
trict of Arizona. The District contains about
26,000 ha of irrigated lands along the Gila
River. Roughly 4,000 ha are on a coarse-textured
mesa that rises about 20 m above the flood
plain; the remainder is on the flood plain. The
valley, roughly 6 km wide and 65 km long, is
underlain by a gravel aquifer. On the east, it is
bounded by the Mohawk Mountains. On the
west, the Gila Mountains severely restrict sub-
surface drainage. The irrigation water is
diverted from the Colorado River at Imperial
Dam, lifted 52 m, and distributed through lined
'canals. Some 108 drainage wells pump into a
lined drainage canal. In recent years, the
drainage rate has averaged about 8500 1/s (2.7
X 108 m3/yr or 210,000 AF/yr) and in 1972 the
salinity of these flows averaged 3700 mg/1. The
drainage water can be returned to the Colorado
River either above or below Morelos Dam, the
point of diversion for irrigation in Mexico. Since
this drainage volume (2.7 X 10 8 m 3) makes up
about 16% of the amount delivered annually to
Mexico by treaty (1.7 X 10^ m 3), mixing itwith
the water detained behind Morelos Dam
significantly affects the salt concentration of
irrigation water in the Mexicali Valley; dis-
charging it downstream from Morelos Dam
increases the volume of water that must be
released at Imperial Dam to satisfy treaty
obligations.
Objectives
The primary objective of the field studies is
to determine the feasibility of reducing the salt
output in drain water by reducing leaching
without reducing crop yields, through uniform
and frequent irrigations. Additional objectives
are to determine the components of the water
and salt balance quantitatively and to establish
requirements for irrigation systems to achieve
low leaching under field conditions.
EXPERIMENTAL PROCEDURE
Citrus
The first experiment was established in
December 1973. It utilized nearly 2 ha of Valen-
cia orange trees (Citrus senensis L.) centered
within a 4-ha block planted in the fall of 1973. It
is located on the mesa, 3 km east of Tacna,
Arizona, on the Desert Valencia Ranch. The
experiment is surrounded by trees of similar
age, except on the south, where it is bordered by
the Mohawk Irrigation Canal. The 4-ha block
was chosen because of its history of high yield,
and the experimental site was located within
the block on the basis of the uniformity of tree
trunk circumference. The tree spacing is 4.9 X
6.7 m and the dripline of a typical tree is 4.5 m in
diameter. The randomized block experiment
consists of 5, 10, and 20% leaching treatments
replicated nine times, using 3- by 3-tree plots
(Fig. 1). This design should detect 12% yield
differences at the 5 % confidence level (Jones,
Embleton, and Cree, 1957). The experiment is
separated from the remainder to the flood-
irrigated grove by border trees irrigated to
achieve 20% leaching. Frost protection and tree
pruning management is the same as for the
surrounding groves.
The soil, a Dateland fine-sandy loam soil
(Typic, Haplargid, couarse laomy, mixed hy-
perthermic), is calcareous throughout, well-
drained, moderately permeable, and represent-
ative of the soils where citrus is grown in the
District. It is underlain with sand beginning at
a depth of 1.5 to 2.0 m and continuing to at least
2m.
82
-------
MINIMIZING SALT
LEGEND
K Vacuum Extractor
Ml Plot Identification
L 5% Leaching
M 10% Leaching
H 2O% Leoching
Border Drip
'O Oi
i Border Bubbler
10 o!
Border Dike
•bj
10
o
lo
i'°
ho
1 °
•i°
10
1°
o
||0
10
'0
i°
o
jl°
1C
1°
IP
1°
10
: -
:
|0
1C
lo
i°
juwwunguuuHnnnurnnnnn
OOOOOOOOOOl
OOOOOOOOOO
OOOOOOOOOOj
°[
:
:•
,
0
0|
:••
c^
'•:•
Ol
°i
O|
c
.:••
0
-
0
0
0
0
|0
. o
°M°I°
ooo
ooo
oLo2o
ooo
ooo
°L°3°
ooo
ooo
°H°4°
ooo
ooo
ooo
oVo
ooo
ooo
0M°6°
ooo
ooo
ooo
M 7
ooo
°H°8°
000
ooo
°L°9°
ooo
ooo
000
ooo
ooo
0H°2°
ooo
ooo
°M°30
ooo
ooo
oLo4o
ooo
ooo
°M°5°
o"o5o
000
o^o o
°H°6°
ooo
OgO o
O 0 O
OLOTO
°M°8°
00 0
0*0 o
ooo
H 9
ooo
ooo
0H°,°
ooo
ooo
0M°2°
000
ooo
°H°3°
OOO
ooo
°M°40
OOO
ooo
ooo
oLo5o
ooo
ooo
°L°6°
OOO
ooo
ooo
oVo
0L°8°
OOO
ooo
°M°9°
OOO
ooo
locfooooooooi
.0000000 ooo'
loooooooooo
o o o_ o_q_ o _o_ o_o__Oj
1
1
ij
: . :
Figure. 1 Design of minimum leaching experiment
on mature Valencia orange trees in southwestern
Arizona.
Irrigation water is continuously available
in the Mohawk Canal. A typical concentration
of total dissolved solutes in the irrigation water
during 1975 was 944 mg/1; the major salt
constituents and their concentrations were (in
meq/liter) Ca, 4.5; Mg, 2.7; Na, 6.8; HCOs, 2.8;
SO 4 , 7.7, and Cl, 3.5. Water for the experiment
is pumped directly from the canal, passed
through commercial sand and screen filters,
and delivered in buried, plastic mains to each
plot. The filters remove foreign material larger
than 75 micrometers in diameter. The 11-kW
centrifugal irrigation pump can deliver 161/s at
a pressure of 350 kPa.
The frequency and duration of irrigation
are controlled by programmable time clocks,
which operate the pump, and by automatic
irrigation controls, which operate electric
valves at each plot. The time clock for the
experimental plots is programmed to operate
the pump for 30 min as many as three (winter) to
six (summer) times every day. A flow control
valve at each tree delivers 32 ml/ sec or 57 liters
to each tree in a plot receiving water that
irrigation period; this is equivalent to a uniform
surface application of 1.7 mm over the entire
area. The volume of water applied to each plot is
measured by two water meters placed in series
to insure accurate, fail-safe measurements.
Each of the 243 experimental trees is irrigated
with a 35-m spiral of dual-chamber drip irriga-
tion tubing as shown in Fig. 2. The tubing has
0.5-mm diameter outlets every 0.3 m along its
length. This design was chosen to provide
uniform water application under each tree. The
border trees are irrigated by a second program-
mable time clock that operates the pump and
automatic irrigation controls for 45 min as
many as two (winter) to four (summer) times
every day when the experimental plots are not
being irrigated. Some of the border trees are
irrigated bubblers filling small basins formed
under each tree at a controlled rate of 63
ml/sec. Other border trees are drip irrigated
from capillary tube emitters (1.7 mm ID) in-
serted into 25-mm-diameter polyethylene pipe
laid on the soil surface about 1 m from the trunk
on both sides of each tree row. The emitters are
spaced 0.7 m apart, giving 14 per tree. Flow
control valves in each pipeline deliver an
average of 32 ml/sec to each tree. The three
irrigation systems are illustrated in Fig. 3.
Tree Drip Lin*
Tensiometers
Salinity Sensors:
For Irrigation, 0.3 -m deep
Depths of 0.15,0.30, 0.45, 0.60, 0.90r
Depths of 0.30, O.45, 0.90, 1.50 m
Depths of 0.30, 0.45,0.60, 1.20m
Figure. 2 Illustration of instrumentation under
center tree of three of the nine replications for each
leaching treatment. The remaining replications have
two sets of salinity sensors and one set of ten-
siometers.
83
-------
WATER MANAGEMENT
Dual - Chamber Irrigation Tubing
Distribution Orifices
(Emits water 01 low pressure)
Supply Orifice
Feeds water to distribution tube)
Main Chamber
(Water supply tube)
Secondary Chamber
(Water distribution tube)
Drip Irrigation Tubing
-25 mm Polyethylene
^^.^ Tubing
Bubbler Irrigation System
Flow Control Valve
Figure. 3 Sketch of dual-chamber, bubbler, and drip
irrigation systems.
Each plot is irrigated automatically by
signals generated at four tensiometers installed
at the 0.3-m depth and located 60 degrees apart
and 1.5 m radially out from the trunk of the
center tree of each plot (Fig. 2). The signal is
generated by a light-emitting diode and a
phototransistor placed directly opposite each
other on the manometer columns of the ten-
siometers (Austin and Rawlins, 1977). As the
soil dries, rising mercury in the manometer
column interrupts the light beam from the
diode. When a signal is received from two or
more of the four tensiometers, a circuit is
triggered that opens a solenoid valve to irrigate
the plot for 30 min. The soil matric potential at
which irrigation occurs is controlled by setting
the elevation of the phototransistors on the
manometer column. Lowering the setpoint
raises the soil matric potential and increases
leaching; raising the setpont decreases
leaching. Soil salinity measurements made
with in situ sensors (Richards, 1966; Oster and
Willardson, 1971) at the 0.30- and 0.45-m depth
serve as feedback to the tensiometer-controlled
irrigation system. To achieve the desired
leaching, the tensiometer setpoints are adjusted
based upon the difference between measured
salinity and that predicted from the soil water
composition model of Oster and Rhoades (1975).
The predicted salinity depends on the chemical
composition of the irrigation water, the partial
pressure of CP2 in the soil, and the soil water
uptake distribution of the crop, as well as the
leaching fraction. For purposes of control, the
border trees are treated as two plots. Their
setpoints are adjusted so that the total amount
of water applied approximately equals that for
the 20% leaching plots.
Additional tensiometers and salinity sen-
sors, in two spatial distributions, were installed
beneath the center tree of each plot. For three of
the nine replications of each leaching treat-
ment, for sets of salinity sensors and two sets of
tensiometers were installed along one radial
(see Fig. 2). For the remaining six replications,
only two sets of salinity sensors and one set of
tensiometers were installed. The salinity sen-
sors and tensiometers are read twice weekly.
Four vacuum extractors (Duke and Haise,
1973) were installed in one replication of each of
the three leaching treatments to measure the
volume and chemical composition of the soil
solution leaching below the undisturbed root
zone. The location of the extractors is shown in
Fig. 1. Each extractor within a treatment is
located on a line directed toward the center of
four different trees from a common manhole at a
depth of 1.2 m. A diagram of the extractor for
one tree is given in Fig. 4. Each extractor
consists of three soil-filled sheet-metal troughs
(0.15 m wide by 0.20 m high by 0.61 m long),
each containing two lines of ceramic tubes
(12 mm in dia.) at the bottom. Soil solution is
drawn through each set of these tubes to a bottle
held under partial vacuum. The vacuum is
adjusted so that two tensiometers installed over
the center of the extractor read the same as two
tensiometers about 0.3 m away from the extrac-
tor at the same depth (see inset in Fig. 4). With
uniform soil matric potential near the top of the
extractors, they should intercept the flux
representative of their cross-sectional area
without causing convergence or divergence of
flow.
The citrus grove around the experimental
site is irrigated by ranch personnel by border
84
-------
MINIMIZING SALT
CROSS-SECTION
A —A
SHEET METAL
TROUGH
Figure. 4. Schematic of vacuum extractor installation.
flooding; a border consists of 6 rows of 35 trees
each, a total area of 0.7 ha, surrounded by
earthen dikes. A typical border is irrigated to a
depth of 150 mm in about 45 min through six
sliding gates in the wall of a concrete irrigation
lateral. One border in the 4-ha block of trees
immediately north of the experiment was
selected for comparing fruit yield and quality,
trunk growth, and irrigation volume with the
experimental plots. Soil matric potential is
monitored with tensiometers, as described
above, under three separate trees. Because of the
high leaching percentage in the border, the soil
salinity level is low and is monitored periodical-
ly from soil samples. Water applied to this
border is measured with a concrete, critical-flow
flume (Replogle, 1977) installed in the irrigation
lateral. The irrigation volume can be calculated
to within about 2% from the irrigation period
and a continuous record of the water elevation
at the entrance relative to the flume floor at its
throat.
The experimental trees are fertilized by
foliar applications of urea and microelement
chelates instead of through the irrigation as is
done in the surrounding groves. One border of
flood-irrigated trees east of the experiment and
within the same 4-ha block is fertilized similarly
to serve as a fertilizer check for fruit yield and
quality and trunk growth with the experimental
trees.
The fruit from all the experimental trees,
the two rows of border trees on the east and west,
and three rows of trees from both the flood- and
fertilizer-check plots are harvested by in-
dividual tree in April of each year and weighed.
Some fruit is selected for quality deter-
minations. Trunk circumference is measured
annually as a simple measure of tree growth.
Alfalfa
The second experiment is 13 km northeast
of Tacna, Arizona, on the Snyder Ranch in the
flood plain of the Gila River. The experimental
site is the northern quarter of an 8-ha field that
had been previously cropped to alfalfa and flood
irrigated. Within 400 m of the south end of the
field, the topography rises 20 m to the mesa
where citrus is grown. A drainage well for water-
-------
WATER MANAGEMENT
table control is located next to the southwest
corner of the field. The experimental site was
instrumented and the irrigation system in-
stalled during September 1974. After cultiva-
tion, alfalfa (Medicago satiua L., cv. Hayden)
was planted early in October.
The Indio fine-sandy loam soil (Typic
Torifluvent, coarse silty, mixed hyperthermic),
representative of the valley soils in the district,
is calcareous and will drained. The soil profile
texture grades from very fine sandy loam to silt
loam at a depth of 0.3 m and to silty clay loam at
a depth of 0.8 m.
The experimental design, shown in Fig. 5,
consists of three leaching treatments, 5,10, and
20%, each replicated five times. Each plot is
12.2 m wide and 104 m long. The remainder of
the 8-ha field is irrigated by the rancher using
level-basin flooding.
The irrigation water, pump, and filter
system are identical to those for the citrus.
Water is applied with an electrically driven,
lateral-move irrigation system 195 m long with
wheel towers 12.2 m apart. As in the citrus
experiments, water for each plot passes through
two water meters in series and then through 20
spray nozzles (Fig. 6). At a pressure of 70 kPa,
each nozzle delivers 30 ml/sec. At a ground
speed of 30 m/hr, the irrigation system applies
about 6 mm each pass. The number of passes
varies from one every few days in winter to a
maximum of two passes per day in summer. The
differential leaching treatments are achieved
by regulating the pressure for each plot with a
throttle valve.
Two replications of each leaching treat-
ment were instrumented. A concrete manhole
was installed at the border between adjacent
plots for the six replications instrumented (Fig.
7). After excavating for the manhole, two
vacuum extractors, each 1.5 m long, were in-
stalled at the 1.2-m depth in each of the six
instrumented plots. The vacuum and pressure
LEGEND
- Plot borders and location of wheel tracks of irrigation system
HI 10 Replication number and leaching percentage
€ Instrumentation Manhole 1.7m deep X 1.5m dia.
75mm I.D. Conduit
c*
c
2
^_ ^
I*
E
IO
» /
,- /
y
w
o>
TJ
0
CO
...
i
120
15
no
•••••
1
I i
|"~^1 12.l9n^*
i | | i
n2o'rj5 inioiiri2diii5
I
i
i
i
i
C v • • -C
1 1 1
1 1
1 1
1 1
1 1
1
1 1
1 1
ni 10
IV201
— r—
IV5
TV 10
t
1
1
V20JV5
I
.
1
1
1
1
i
1
1
1
I
1
VIO
0
-a
0
CO
Figure. 5 Design of minimum leaching experiment for alfalfa in southwestern Arizona.
86
-------
TOP VIEW
MINIMIZING SALT
Water
meters
Valve
6-mm I.D. PVC Pipe
Wheel
SIDE VIEW
dOO-mm I.D Aluminum Mainline
18-mm I.D. Polyethylene hose
'^25-mm I.D. PVC Sub
Main
Figure. 6 Spray system for one section of the lateral move irrigation system serving one replication.
9 x
,Plot Border and Wheel Track
of Irrigation Tower
Excavated Pit
r
Suction Lysimeter
Extractor
^-Conduit Trench
i jljLJ
| ^^^O
x • x
t 1
1
i !
. V •
\ *
J~ *
vC \
Manhole]
Scale:
•MB
^Conduit Tunnel
X % X
LEGEND
^™ * "^V
x Salt Sensors O.3 , 0.6, 0.9, ft 1.2 m deep
Tensiometers 0.3, 0.6, 0.9,1.2 , 8 1.5 m deep
Figure. 7 Location of instruments for two of the six replications instrumented in the alfalfa experiment.
SI
-------
WATER MANAGEMENT
lines to the extractors were installed through an
underground conduit to each manhole. Salinity
sensors and tensiometers were installed as
shown in Fig. 7. In addition to the vacuum
extractors, a 1.2-m-deep suction lysimeter was
installed in each of the six instrumented plots.
These were formed by lining a 1.5-m-square hole
with 250-/im-thick plastic film. A wooden frame,
0.3 m below the soil surface, serves as the top lip
of the lysimeter. After backfilling with about
0.1 m of soil, ceramic tubes, identical to those
used in the vacuum extractors, were installed in
each lysimeter to extract the leachate. The
extractors were then filled with disturbed soil
and compacted by saturating with water.
The portion of the 8-ha field not within the
experiment, which serves as the check plot, is
flood irrigated as a level-basin from a single
gate located in one corner of the field. For each
irrigation, a depth of about 150 mm is applied at
the rate of about 0.4 m 3 / s. Irrigation volume is
measured with a critical-flow flume identical to
that used in the citrus experiment. Salinity
sensors were installed at four depths in six
locations in the check plot.
All of the crop management practices, in-
cluding planting, harvesting, and pest and
weed control, are performed by personnel of the
Snyder Ranch. The alfalfa is harvested by
cutting with a windrower conditioner and cub-
ing, following customary procedures. Harvest
dates are chosen by the ranch when the alfalfa
is at approximately 50% bloom
The yield of each cutting is determined by
weighing three samples from each replicate.
Each sample is taken from a 26.8-m 2 area in the
first swath (4.4 m wide) made down the center of
each plot. Nine similar samples are taken from
the flood check.
RESULTS AND DISCUSSION
The results presented here are illustrative
rather than complete. Furthermore, they repre-
sent an interim rather than a final report. Since
the initiation of the citrus study, three harvests
have been made. The first harvest, in April 1974,
no doubt reflected the irrigation management
before the experiment began. Citrus yields tend
to vary greatly and often show an alternate year
pattern (Jones et al., 1957, Parker and
Batchelder, 1932). Also, Bingham et al. (1974)
found that the effects on yield due to a change in
irrigation management were not noted till the
fourth year; and Raats (1975) predicted that a
period of many months, depending on leaching
fraction, would be required before a new quasi-
steady-state would be reached. Thus it would be
unrealistic to expect definite results in 1976. For
similar reasons, the time since initiating the
alfalfa experiment has been too short for
definitive results. In addition, some difficulties
were encountered in that experiment.
Citrus
Soil Properties
At the beginning of the experiment, exten-
sive soil samples were collected for detailed
analysis. The soil in the grove was far from
uniform, with wide changes in texture from
place to place; similarly, the chemical proper-
ties showed considerable variation. Table 1
gives representative values for a number of
properties. The relation between hydraulic con-
ductivity and matric potential, determined on
three 0.9-m-long undisturbed cores, is given in
Fig. 8.
TABLE 1
Representative Soil Properties of Dateland Fine Sandy
Loam Soil
Property
Unit
Typical Values Range
Cation exchange capacity
Exchangeable sodium percentage
Sodium adsorption ratio
Saturation percentage
Field water content
Electrical conductivity
pH
Soluble* calcium
Soluble* magnesium
Soluble* sodium
Soluble* potassium
Soluble* bicarbonate
Soluble* sulfate
Soluble* chloride
Soluble* nitrate
meq/lOOg
%
—
g/gx 100
g/gx 100
S/m@2S
—
meq/1
9-13
3-4
25-30
11-12
0.13-0.15
7.4
3-4
2
5-7
0.2
2
5-6
2-3
0.6
2.5-23
3-13
19-50
4-22
0.1-0.6
6.4-8.3
1.6-23
0.7-11
3-37
tr-0.6
0.8-4.6
2.3-56
0.9-28
0.2-6.2
•Soluble in saturation extract.
Water Use
Annual totals of water applied are given in
Table 2. When analyzed on a month-to-month
basis, the data show that the application rates
by treatments were not always in the expected
order; the totals, however, were ordered as
expected. Monthly averages show a range from
1 mm/day (3.271/tree-day) to 11 mm/day (360
1 /tree-day), if an inadvertent over-irrigation in
one month (July 1976) on the 20% leaching plots
is not considered.
88
-------
MINIMIZING SALT
-30
Matric Potential , kPa
-20 -10
0
E
100 E
50 si
10 o
o
5 £
o
I J-
0.5
O.I
Figure. 8 Relationship between soil matric poten-
tial and hydraulic conductivity for Dateland fine
sandy loam soil.
TABLE 2
Annual Irrigation, Rainfall, and Pan Evaporation for Citrus
5000
1000
500
a
TJ
Depth of water applied*
Cat-
Year
Leaching
Treatment Borders Flood Rain- PanEvap- culated
5% 10% 20% Bubbler Drip Check fall oration ET
mm
1974
1975
1976t
Average
1401
1498
1675
1525
1450
1560
1725
1578
1685
1802
1975
1821
1651
1942
2049
1881
1818
1912
1704
1811
_
2582
2750
2666
105
84
90
93
1838
1732
1650
1740
1328
1423
1550
1434
•Irrigation plus rainfall.
TData estimated for October, November, and December.
The 3-year average water applications sup-
port, but by no means prove, the hypothesis that
the desired leaching fractions were closely ap-
proximated. Using that assumption, one can
calculate an average annual evapotranspira-
tion rate (ET) of 1440 mm. Application of that
estimate of ET to the flood check water use,
yields an estimated leaching percentage of 46%.
ET was 83% of pan evaporation, as measured
with a pan placed within a nearby citrus grove
in a space cleared by removing one tree.
Referred to a pan some 5 km away in an open
area, the pan factor was 0.71.
Clearly, the above calculations and
assumptions are somewhat tenuous. Reliable,
direct measurements of ET are not available,
and the direct measurements of leaching
volumes are also considered unreliable.
However, some additional information should
be considered. In 1974, water was purposely
withheld to increase the salinity to the projected
levels; thus, some of the water used was from
soil storage. As experience was gained, it
became evident that the imposed changes in
water application rates lagged the changes in
pan evaporation. In 1976, in an effort to compen-
sate for this lag, we overirrigated for a period.
These two sets of events qualitatively explain
some of the year-to-year differences shown in
Table 2.
Comparisons were also made with ET es-
timates calculated from climatological data
using the modified Penman (Doorenbos and
Pruitt, 1975) and Jensen-Haise (Jensen, 1973)
equations and with the data published by Erie et
al. (1965). These comparisons led to the follow-
ing values for yearly evapotranspiration in mm:
Erie
Jensen-Haise
Penman
Field estimate
1000
1342
1421
1440
The citrus data reported by Erie et al. were taken
from earlier work where the water use was
determined gravimetrically, but the water lost
from the top 0.15 m was ignored (Jensen, M. E.,
personal communication, 1975). Thus one would
expect them to give an underestimate.
Soil Water Potential
The tensiometer readings give an indica-
tion of the water potential distribution in the
rootzone of the trees. The average profiles for
the three leaching treatments during 1974 and
1975 (Fig. 9) show that the soil was indeed
substantially drier at depths below 0.6 m for the
plots with 5% leaching than at 20% leaching.
Near the surface, the potentials were essentially
the same. Average profiles, however, do not
reflect the details of the flow regimes, nor do
they bring out the heterogeneity in the system.
To better understand the detailed flow
patterns, we installed 84 additional ten-
siometers under the center tree in plot H4 (Fig. 1)
89
-------
WATER MANAGEMENT
-25
Soil Metric Potential, kPa
-20 -15 -10 -5
4
:
6 -
o
V.
Figure. 9 Average soil matric potential profiles for
the three leaching treatments in the citrus during
1974 and 1975.
Radial Distance from Tree , m
North Radial
r
-
:
.-
Radial
Diagonal Radial
1.80
Figure. 10 Distribution of total head under the
center citrus tree of plot H4 in the evening on August
29, 1976.
in three planes through the tree trunk and read
them frequently. Figures 10,11, and 12 illustrate
some of the trends. They represent total head
distributions referred to the 0.3-m depth as
datum for August 24, August 29, and September
Rodial Distance from Tree , m
2
North Radial
3
.-
Radial
Diagonal Radial
Figure. 11 Distribution of total head under the
center citrus tree of plot H4 in the evening on August
29, 1976.
Radial Distance from Tree , m
:
North Radial
-
East Radial
Diagonal Radial
Figure. 12 Distribution of total head under the
center citrus tree of plot H4 on September 4, 1976.
-------
MINIMIZING SALT
JFMAMJJASONDJFMAMJJASONDJFMAMJJASOND
1974 1975 1976
Figure. 13 Salinity trends with time for the three leaching treatments of the citrus experiment.
4, 1976. On August 24, irrigation had
reasonably met crop demand, but that morning
the irrigation water was shut off and left off
until 2040 hours on August 29. Thesetpoint(i.e.,
the control point for irrigation at 0.3 m) during
this period was -5.5 kPa.
The data illustrate that the flow patterns
clearly were not one-dimensional. The head at
the 0.3-m depth was controlled by the setpoint
when the irrigation system was operating. The
6-day drying period is reflected in the substan-
tially drier profile on August 29. By September
4, the distribution again resembled that of
August 24 very closely. The direction of the
gradients strongly suggests there is little root
activity below 0.6 m.
Soil Salinity
The time course of average soil salinity
measured with salinity sensors is given in
Fig. 13 for the three leaching treatments.
(Salinity data for the 0.6- and 1.2-m soil depths
are not given to simplify the figure.) Initial soil-
salinity levels were about equal to the salinity of
the irrigation water, 0.13 S/m, because of
overirrigation before the experiment. Begin-
ning in February 1974, soil salinity increased
with time until quasi-stable values were at-
tained at the 0.3- to 0.45-m depth by July (5
months) and at the 0.9-m depth by January 1975
(11 months). Since March 1975, changes in soil
salinity have undergone two complete cycles.
The cyclic pattern is evident to a soil depth of
0.9 m and the cycle takes about 1 year. During
the first half of the year, the trees were un-
derirrigated and soil salinity increased; during
the last half of the year, the trees were
overirrigated and salinity decreased. Although
cyclic, the differences in salinity among
leaching treatments were as expected, with 5%
leaching being the most saline and 20%
leaching the least.
Figure 14 compares time-averaged soil
salinity as a function of soil depth for the
intervals of January 1974 to July 1975 and July
1975 to July 1976, the initial soil-salinity dis-
tribution, and the projected soil-salinity dis-
tribution at equilibrium from the model of Oster
and Rhoades (1975). In general, soil salinity
increased with time at all soil depths and
salinity decreased with increasing leaching.
The soil-salinity distribution has reached the
projected salinity values for all three leaching
treatments to a depth of 1.5 m, except at the
deeper depths in the 5% leaching treatment.
Water Uptake Distribution
Of interest are the distribution of water
uptake with depth and the changes in salinity
with depth over time. From extensive soil samp-
ling, chloride distributions have been obtained
at various times under selected trees. Oster (this
conference) will discuss these in detail. Here we
only point out that, with some difficulty, such
data can be used to estimate the water uptake
distribution. In Table 3, we present the results of
such calculations for one tree irrigated with a
91
-------
WATER MANAGEMENT
Soil Woter Salinity , S/m
0.2 0.4 0.6 0.8 1.0 1.2 1.4
£
f
**
ex
0
_.
"o
CO
V
0.3
0.6
0.9
1.2
\ \ \Ni 5% Leaching
' \ ^"v^A
V* \
\ ^'\ \ .--PROJECTED
"••A S SOU. SALINITY
- \ \./
0 0.2 0.4 0.6 0.8 1.0 1.2 1.'
0
0.3
0.6
0.9
1.2
1 1
| | 1 I 1 |
. . .. 10% Leaching
. 1 X^x, JAN '74 (Initial)
\ X '•• \ JAN '74 to
\\ I JULY '75
\ N''-\ JULY '75 to
. \X. JULY '76
• '.-
0 0.2 0.4 0.6 0.8 1.0 1.2 1.
0.3
0.6
0.9
1.2
i *
i i i i ii
. . r. 20% Leaching
\ i'\
- \ v\
- \ y>
" * \
Figure. 14 Time-averaged soil salinity distribu-
tions with soil depth at initiation, for two in-
termediate time periods, and the projected final
conditions.
leaching fraction targeted at 0.20 and one, at
0.05. Prom weighted chloride data, the actual
LF's were estimated to have been 0.30 and
0.06-Irrigation water was taken as containing
3.3 meq/1 of chlorides
Raats (1975) has indicated how the travel
time for a parcel of water through the soil profile
can be calculated if the uptake distribution is
TABLES
Chloride and Water Uptake Distribution Under the Center
Trees of a High and a Low Leaching Plot
PlotH4
PlotL7
Soil CI Accumulated
Depth Concentration RWL'
Accumulated
Concentration RWL'
m
0.3
0.6
0.9
1.2
1-5
1.8
meq/1
6.0
8.3
9.4
10.2
10.6
-
0.64
0.86
0.93
0.97
0.98
-
meq/1
12.0
14.5
18.0
23.0
33.0
45.0
0.77
0.83
0.88
0.92
0.97
1.00
•Relative water loss (part of water consumed above depth indicated).
known. For purposes of generalization, he as-
sumed an uptake distribution of the form
/3 = 6~1 exp (-z/5),
where /3 is the fraction of the total water taken
up above depth z and 8 may be interpreted as a
characteristic rooting depth. When z = 8, 63% of
the water uptake has taken place. Figure 15
shows such travel times for three leaching
fractions, assuming 8 = 0.2m and 6 = 0.4m.
When the data in Table 3 are used, the paralled
calculations give results very close to those for 8
— 0.4. These curves may be interpreted in terms
of the time required to establish a new
equilibrium after irrigation management has
changed. We conclude that we should expect to
establish a new quasi-equilibrium no sooner
than 9 months after initiation of a 5% leaching
treatment at a depth of 1.8 m, and no sooner
than 3 months for 20% leaching. These es-
timates are of the right order of magnitude, but
somewhat shorter than suggested in Fig. 13.
Vacuum Extractor Volumes and
Concentrations
It was stated earlier that a reliable direct
measurement of leaching volumes was not
available. We had hoped that the vacuum
extractors would yield such data. However, the
volumes collected have been consistently too
low. This result can be explained in part by
mechanical failures. Restrictions due to plumb-
ing failures and poor contact between the
extractors and the undisturbed soil have recent-
ly been noted in some of the extractors and
corrected. However, a more basic cause relates
to nonhomogeneity of the soil profile. The
extractors depend, for proper operation, on the
continuity of vertical flow. When the soil texture
changes from fine to coarse, such continuity is
not maintained. Especially where pockets of
sandy soil are imbedded in finer textured soil,
theory and experiment predict that unsaturated
flow will tend to bypass the coarse-textured
pockets. Field observation has verified that this
situation indeed occurs for at least some of the
extractor segments.
The extractor leachates have been analyzed
chemically. The leaching fractions calculated
from the chloride levels were 0.06,0.15 and 0.27
for the low, medium and high plots, respective-
ly. These values are reasonable and ap-
proximately the same as determined by alter-
nate methods. From early data, it is estimated
that the LF prior to initiating the experiment
92
-------
MINIMIZING SALT
Q.
0)
Q
o
CO
40
-
80
Travel Time , Days
120 160 200 240 280
320 360
0.3
_ 0.6 -
8 =0.4 m
8 = 0. 2 m
0.9 -
Figure. 15 Chloride-derived and calculated travel time for a parcel of water to pass through the soil
profile as a function of leaching fraction and rooting depth. The curves are calculated from Raat's theory
with 6 = 0.15 and ET = 7mm/day. The plotted points are derived from the chloride data (Table 3) for the
leaching percentages indicated.
was at least 0.33. This compares with 0.46
determined from the water use data on the check
plot.
Fruit Yield and Quality
The average annual yield, regardless of
treatment, was 120,149 and 83 kg/tree for 1974,
1975 and 1976, respectively. There was no
statistical difference in yield between any of the
treatments, including the border and check
plots. Similarly, treatments did not differ in
number of fruit harvested, or in fruit quality.
The drastic difference from year-to-year
appears to be typical of citrus, as indicated
earlier. Measurements of trunk diameter
verified that the trees are continuing to grow,
again without any differences attributable to
treatment. In view of the findings by Bingham
et al. (1974) that fruit yield depressions were
noted the fourth year after salinization, the
results to date must be interpreted with caution.
Alfalfa
Soil Properties
As for the citrus, we present some typical
soil properties, Table 4, and a representative
relationship between the hydraulic conductivi-
ty and matric potential (Fig. 16). The soil was
nonsaline, indicating that it has been well
leached in the past.
TABLE 4
Representative Soil Properties for the Snyder Ranch
Alfalfa Field
Property
cation -exchange-capacity
exchangeable- sodium-
percentage
sodium -adsorption-ratio
saturation percentage
field water content
pH
electrical conductivity
soluble* calcium
soluble magnesium
soluble sodium
soluble potassium
soluble bicarbonate
soluble sulfate
soluble chloride
soluble nitrate
Depth
Unit fmetert
meq/lOOg 0-0.6
>0.6
% 0-0.6
>0.6
- 0-0.6
>0.6
g/g x 100 0-0.6
>0.6
g/g x 100 0-0.6
>0.6
0-0.6
>0.6
S/m 0-0.3
>0.3
meq/1 0-0.3
>0.3
0-0.3
>0.3
0-0.3
>0.3
0-0.3
>0.3
0-0.3
>0.3
0-0.3
>0.3
0-0.3
>0.3
0-0.3
>0.3
Typical
values
15
28
8
10
6
7
40
50
14
28
7.4
7.5
0.25
0.35
5
15
3
8
12
20
0.3
0.3
2.3
2.5
16
28
5
10
0.6
0.3
Range
10-19
25-35
5-12
5-30
5- 8
5-30
30-45
42-60
11-30
23-36
6.9-8.2
7.0-7.8
0.15-0.50
0.17-1.0
3.8-26
2-30
3-16
1-18
10-27
10^0
0.1-0.7
0.1-0.6
1.9-4.0
1.4-6.0
8-30
15-80
3-10
5-30
0.2-3
0.1-1.7
'Soluble in saturation extract.
93
-------
WATER MANAGEMENT
Matric Potential , kPa
-30 -20 -10
5000
1000
500
a
TJ
\
E
E
100
50 >>
>
10
5
J
3
a
w
•o
>,
I
0.5
O.I
Figure. 16 Relationship between soil matric poten-
tial and hydraulic conductivity for Indio fine sandy
loam.
Water Use
The total amounts of water applied from
January 1975 to July 1976 were 3152,3322, and
3613 mm for the 5, 10, and 20% leaching treat-
ments, respectively, and 3598 mm for the flood
check.
By subtracting the planned leaching depth,
the total water use data can be used to arrive at
an estimate of evapotranspiration. This es-
timate is 1813 mm/year. This compares with
other estimates as follows:
Erie et al.
Penman
Brawley, Ca.
1890
2201
2060
The Brawley data were obtained in a weighing
lysimeter (LeMert and Kaddah, 1977). Arbitrar-
ily taking the average of these four estimates
(1991 mm) as the true value, the leaching
percentages for the 5,10, and 20% leaching treat-
ments would be -3,2, and 10%. Other data tend to
confirm that the alfalfa was indeed underirri-
gated.
Soil Water Potential
Tensiometric measurements have been
helpful in explaining some of the anomalies
noted. As illustrated in Fig. 17, the total head
distributions indicated both upward and
downward gradients. These and similar data
may be used to draw the following conclusions.
The reversal in gradient early in 1975 on the 10
and 20% leaching plots can be explained by
differential water uptake with depth. Overtime,
these profiles developed into monotonically
decreasing heads with depth, indicating that
water was moving downward through the
profile into the water table. In some of the 5%
leaching plots, however, there was evidence of a
substantial upward gradient below 0.9 m
associated with water uptake from the water
table. More detailed data, not shown, support
the hypothesis that in some parts of the field,
part of the water was supplied from the water
table. Whether upward flow takes place depends
not only on the irrigation management, but on
localized subsoil conditions and water table
position.
Soil Salinity
Initial salinity sensor measurements taken
in November 1974 averaged 0.79,0.62,0.68, and
0.73 S/m at the 0.3-, 0.6-, 0.9- and 1.2-m depths,
respectively. Due to operational difficulties with
the irrigation system during the beginning of
1975, insufficient water was applied, and salini-
ty at the 0.3-m depth increased after the experi-
ment began, By March, salinity at the 0.3-m
depth started to decrease. This decrease has
continued, although short-term cycles have oc-
curred. The lowest levels of salinity (about 0.4
S/m) were reached in January 1976 (Fig. 18).
The salinity below 0.3 m was expected to in-
crease initially, then become stable with time.
Such increases did occur, with stable but high
levels reached by January 1976. As shown in
Fig. 19, the salinity profiles have changed from
ones decreasing with depth during the first half
of 1975 to ones increasing with depth during the
second half of 1975. They have changed little
through the first half of 1976, except for a slight
increase at 0.9 and 1.2 m. These increases at
greater depths show the beginning of the profile
shapes expected to develop, but to date salinity
has not difference significantly among
treatments. For example, in May 1976, soil
salinity at the 1.2-m depth was about 1.1 S/m for
all leaching treatments; for irrigation water
94
-------
MINIMIZING SALT
having a salinity of 0.13 S/m, the resultant
leaching percentage is 12.
0.3
0.6
0.9
1.2
1.5
e 0.3
- 0.6
F 0.9
1.2
1.5
0.3
0.6
0.9
1.2
1.5
0_
UJ
Q
4-1-75
I "05
n-05
I,It - 10
n,m-20
6-5-75
9-15-75
-10 -8 -6 -4 -2 0
TOTAL HEAD , m of H20
Figure. 17 Hydraulic potential distribution for the
three leaching treatments in the alfalfa at selected
times during 1975.
For comparison, the salinity profile for the
flood check is shown in Fig. 19, along with that
for the 20% leaching treatment. The salinity
increased uniformly with depth, was quite
stable, and was 0.2 to 0.3 S/m lower than in the
experimental treatments. Based on a soil salini-
ty of 0.77 S/m at 1.2-m depth, the apparent
leaching percentage obtained on the check plot
was 17%, indicating that 5% more leaching had
occurred on the flood-irrigated check than on
the sprinkled experimental plots.
Vacuum Extractor Volumes and
Concentrations
The extractor data confirm the conclusions
reached from the salt sensor, the tensiometer,
and the water-use data, i.e., that the alfalfa
experiment was underirrigated. The extractor
volumes obtained indicate that the alfalfa was
irrigated at about half the targeted leaching.
The extractors in 1-05 were affected by upward
flow. The composition of the extractor leachates
also indicated less-than-desired leaching. The
compositions have not yet come to equilibrium;
continuing exchange reactions account for the
difference between observed and predicted com-
positions.
Crop Yield
The alfalfa was cut 7 times in 1975 and 5
times in 1976 through July. The 1975 yields were
J L
c
o
0.2 H
0 80 160 240 320 40 120 2OO 280 36O
1975 1976
Julian Date
Figure. 18 Salinity trends with time for the 10% leaching treatment of the alfalfa experiment.
95
-------
WATER MANAGEMENT
0
0.3
0.6
0.9
1.2
1.5
0
E
0.3
f °6
£ 0.9
o
_ 1-2
'o
en 1.5
o
0.3
0.6
0.9
1.2
1.5
Soil Water Salinity , S/m
.2 .4 .6 .8 1.0 1.2
1.4
5% Leaching
.2
.4
.8
1.0
1.2
1.4
\
10 % Leaching
.2
.4
.6
1.0
1.2
1.4
flood check
2O% Leaching
Figure. 19 Time-averaged salinity distributions
with depth for three leaching treatments and the
flood-irrigation check of the alfalfa experiment.
17.9, 17.6 and 17.6 Mg/ha for the 5,10, and 20%
leaching treatments; for 1976, they were 16.4,
15.8 and 15.9 Mg/ha. These yields were not
statistically different with respect to treatment.
The flood check yielded 21.3 Mg/ha in 1975 and
18.6 Mg/ha in 1976, 20 and 16% more than the
average of the experimental plots.
These results cannot be fully explained;
however, the experimental plots were plagued
with excessive weeds and had a poor stand
compared to the flood check. This was partly
due to some problems encountered with her-
bicide applications and partly to the method of
water application combined with the
harvesting operation. The high frequency of
irrigation maintained a wet soil surface that in
turn resulted in a low infiltration capacity;
apparently, this problem was aggravated by
compaction from the frequent passage of heavy
equipment for harvesting.
Some operational problems were also en-
countered with the irrigation equipment. As a
result, during some periods more water needed
to be applied than normally to make up for past
deficits. Measurements of oxygen at a depth of
0.45 m indicated that, during these periods, the
oxygen level in the soil air was often at 10% or
below. The time-averaged salinity in the top
0.2 m of soil, for the period July 1975 to May
1976,was 0.8 S/m in the experimental plots
compared to 0.5 S/m in the flood check. They
yield depression expected from this 0.3 S/m
difference is 12% (Maas and Hoffman, 1977).
After the July 1976 cutting, the alfalfa was
removed from the field, the field was relevelled,
and a new seeding established. At the same
time, the irrigation system was modified to
permit larger applications per irrigation. It is
hoped that these changes will overcome the
earlier problems.
INTERPRETATION
Experience thus far suggests that the
technology is available to irrigate tree crops at
controlled efficiencies to maintain the leaching
percentage at 20% or less. We believe that such a
management practice will not adversely affect
yield or , for that matter, cost of production in
the Wellton-Mohawk.
The experience with alfalfa has been less
satisfactory. The irrigation method chosen to
carefully control water application rates was, to
some extent, incompatible with the other farm-
ing practices, and the high water table at the
experimental field has compounded the results.
The flood irrigated check plot, however, has
indicated that a leaching fraction of around 0.10
can be maintained with conventional practices.
The consequences of these tentative con-
clusions may be illustrated briefly for the
Wellton-Mohawk District. It should be stressed
that, in different physical settings, the con-
sequences would no doubt be different.
Let us assume the following conditions. The
area (A) planted in citrus is 3000 ha; the typical
water delivery using flood irrigation is 3200
mm/yr (I p); the evapotranspiration (ET) is 1400
mm/yr; the projected average drainage water
concentration in 1980 is 3000 mg/1; the irriga-
tion water contains 944 mg/1 (Ci); and the
expected leaching fraction (LF) is 0.20. Then the
reduction in annual volume of drainage water
due to a change in irrigation management
would be
96
-------
MINIMIZING SALT
AV = A(Ip -
= 3000 [3200 - (1400/0.80) ]
= 4.35 x 107 m3 /yr
= 35,000 AF/yr
If we assume mass balance, without precipita-
tion of salts in the rootzone, then the amount of
salt leaving the rootzone in the drainage water
would be reduced by
A salt = C i AV
= 41,000 Mg/yr
= 45,000 tons/yr
However, we know that gypsum and lime will
precipitate. Making that correction (Oster and
Rhoades, 1975), the salt load would be reduced
an additional 11%:
A salt - 45,100 Mg/yr
- 49,600 tons/yr
This reduction in salt input to the water table
would be reflected over time in the drainage
water pumped and, presumably, delivered to the
proposed desalting complex. In the short run,
the leaching water will displace groundwater.
In 1973, the groundwater contained 3700 mg/1
of salt; it has been projected to be at 3000 mg/1 in
1980. Using the latter figure, the short-term
effect of changing the irrigation management
in the citrus would be reduction in the exported
salt load of 3000 mg/1 X (4.35 X 10 7 m3/yr), or
130,000 Mg/yr. This compares with a current
annual volume of around 26 X 10 7 m3/yr
containing, at 3000 mg/1, 780,000 Mg/yr.
A similar projection could be made for the
approximately 8000 ha of alfalfa irrigated with
an estimated leaching fraction of 0.25 in the
period 1970-72
It should be emphasized that the above
projections are site-specific. Under different
circumstances, entirely different consequences
could result. Suarez and Rhoades (1977) and
Rhoades and Suarez (1976) considered a series
of such circumstances and demonstrated the
dependence of the results on the river water
chemistry. The fact remains that irrigation
management can substantially affect down-
stream water quality, not to mention water and
energy conservation. The understanding
gained from studies such as reported here, and
the practical experience obtained in evaluating
different irrigation systems, should be useful as
necessary but insufficient inputs to the site-
specific evaluation required.
ACKNOWLEDGMENT
The work reported here represents the ef-
forts of most members of the staff of the U.S.
Salinity Laboratory. The author simply
represents the interests of his staff and does not
take credit for the work. He does accept respon-
sibility for any errors in preparation or inter-
pretation.
The work was supported in part by EPA
under Interagency Project EPA-IAG-D4-0370.
REFERENCES
1. Austin, R. S., and Rawlins, S. L. 1977.
Photo-interrupter liquid-level detector circuit:
Applications for automatic irrigation with ten-
siometers. Agr. Eng. (Submitted.)
2. Bernstein, L., and Francois, L. E. 1973.
Leaching requirement studies: Sensitivity of
alfalfa to salinity of irrigation and drainage
waters. Soil Sci. Soc. Amer. Proc. 37:931-943.
3. Bingham, F. T., Mahler, R. J., Parra, J.,
and Stolzy, L. H. 1974. Long-term effects of
irrigation-salinity management on a Valencia
orange orchard. Soil Sci. 117:369-377.
4. Doorenbos, J., and Pruitt, W. O. 1975.
Crop water requirements. Irrigation and
Drainage Paper No. 24, Food and Agric.
Organization of the U.N., Rome, 179 p.
5. Duke, H. R., and Haise, H. R. 1973.
Vacuum extractors to assess deep percolation
losses and chemical constituents of soil water.
Soil Sci. Soc. Amer. Proc. 37:963-964.
6. Erie, L. J., French, O. F., and Harris,
K. 1965. Consumptive use of water by crops in
Arizona. Univ. of Arizona Agr. Exp. Sta. Tech.
Bull. 169, 44 p.
7. Hoffman, G. J., Dirksen, C. Ingvalson,
R. D., Maas, E. V., Oster, J. D., Rawlins, S. L.,
Rhoades, J. D., and van Schilfgaarde, J.
Minimizing salt in drain water by irrigation
management — Design and initial results of
Arizona field studies. Agric. Water Manage-
ment. (In preparation)
8. Jensen, M. E. (ed.) 1973. Consumptive
use of water and irrigation water requirements.
Report of the Tech. Comm. on Irrig. Water
Requirements of the Irrig. and Drainage Div.,
Amer. Soc. Civil Engr., 215 p.
97
-------
WATER MANAGEMENT
9. Jones, W. W., Embleton, T. W., and Cree,
C. B. 1957. Number of replications and plot sizes
required for reliable evaluation of nutritional
studies and yield relationships with citrus and
avocado. Amer. Soc. Hort. Sci. 69:208-216.
10. LeMert, R. D., and Kaddah, M. T. 1977.
Lysimeter-determined and estimated evapo-
transpiration of alfalfa in the arid subtropical
climate of Imperial Valley, California. Agron. J.
(In preparation.)
11. Maas, E. V., and Hoffman, G. J. 1977.
Crop salt tolerance — current assessment. J.
Irrig. and Drainage Div., ASCE. (In press.)
12. Oster, J. D., and Rhoades, J. D. 1975.
Calculated drainage water compositions and
salt burdens resulting from irrigation with river
waters in the Western United States. J. En-
viron. Qual. 4:73-79.
13. Oster, J. D., and Willardson, L. S.
1971. Reliablility of salinity sensors for the
management of soil salinity. Agron. J. 63:695-
698.
14. Parker, E. R., and Batehelder, L. D.
1932. Variation in the yield of fruit trees in
relation to the planning of future experiments.
Hilgardia 7(2):81-161.
15. Raats, P. A. C. 1975. Distribution of
salts in the root zone. J. of Hydrology 27:237-
248.
16. Replogle, J. A. 1977. Portable, adjust-
able flow-measuring flume for small canals.
Transactions of the ASEC. (In preparation.)
17. Rhoades, J. D., Ingvalson, R. D.,
Tucker, J. M., and Clark, M. 1973. Salts in
irrigation drainage waters. I. Effects of irriga-
tion water composition, leaching fraction, and
time of year on the salt compositions of irriga-
tion and drainage waters. Soil Sci. Soc. Amer.
Proc. 37:770-774.
18. Rhoades, J. D., Oster, J. D., Ingval-
son, R. D., Tucker, J. M., and Clark, M. 1974.
Minimizing the salt burdens of irrigation
drainage waters. J. Environ. Qual. 3:311-316.
19. Rhoades, J. D., and Suarez, D. L. 1976.
Benefits and limitations of reduced leaching.
Proc., Conf. on Salt and Salinity Management,
Santa Barbara, Calif. Water Resources Cr.
Report No. 38:93-110.
20. Richards, L. A. 1966. A soil salinity
sensor of improved design. Soil Sci. Soc. Amer.
Proc. 30:333-337.
21. Suarez, D. L., and Rhoades, J. D.
1977. Effect of leaching fraction on river salini-
ty. J. Irrig. and Drainage Div., ASCE. (In
press.)
22. United States Environmental Protec-
tion Agency. 1971. The mineral quality problem
in the Colorado River Basin, Summary Report,
U.S. Environmental Protection Agency
Regions 8 and 9, GPO 790485.
23. United States Salinity Laboratory
Staff. Minimizing salt in return flow through
irrigation management. EPA-IAG-D4-0370. (In
preparation.)
24. Van Schilfgaarde, J., Bernstein, L.,
Rhoades, J. D., and Rawlins, S. L. 1974. Irriga-
tion management for salt control. J. Irrig. and
Drainage Div., ASCE 100(IR3):321-338.
98
-------
Modeling Salinity
of Irrigation Return Flow
where Sources and Sinks
are Present
R. J. HANKS, L. S. WILLARDSON, and D. MELAMED
Department of Soil Science and Biometeorology,
College of Agriculture, Utah State University, Logan, Utah
ABSTRACT
Managing irrigation return flow when
salinity sources and sinks are present presents
many complex difficulties. Models developed to
include the source sink term in an empirical way
show that wide variations in irrigation quality
and quantity will have little influence on return
flow for many years where the source-sink term
is important. The quantity of salt in return flow
is primarily determined by water flow because
the soil solution concentration changes very
slowly. Where a water table is present,
successful irrigation might be practiced with no
return flow for some conditions. Even where no
source-sink term is considered, model predic-
tions indicate irrigation management with no
return flow for several years is possible with
little reduction in yield. The influence of irriga-
tion system uniformity is shown to be signifi-
cant over several years time. Thus, it is evident
that many possiblities for return flow manage-
ment exist but the long term effects should not
be overlooked.
INTRODUCTION
Prediction of the influence of irrigation
management on irrigation return flow quality
and salinity of the root zone must be possible if
future irrigation is to be successful over a long
period of time. Many presently successful irriga-
tion systems based on design criteria that
ignore the salinity of the irrigation return flow
may need to be modified or changed to meet new
restrictions on irrigation return flow quality.
Many models have been developed to
predict the influence of irrigation management
practices on the salinity of the soil and drainage
water where the soil has been assumed to be
inert (summarized by Childs and Hanks, 1975).
However, many soils are not inert. Soil chemical
reactions have strong source-sink properties
such as exchange, precipitation, and solution.
These processes are especially important in soil-
water systems having a predominance of
calcium and magnesium sulfates. Some authors
have used chemical models to account for the
source-sink processes and have obtained good
predictions of soil solution and drainage water
solution and drainage water salinity for some
limited laboratory and field situations (Tanji et
al., 1972; Dutt et al., 1972; and Oster and
Rhoades, 1975). The water flow portions of these
models have been quite restrictive so that many
field situations cannot be simulated.
Consequently, we have been interested in
the development of a predictive model that is
comprehensive enough to include most field
related processes of water and salinity variation
related to irrigation return flow. The current
model (Nimah and Hanks, 1973) involves a
comprehensive water flow portion that ac-
counts for irrigation, rainfall, drainage,
evaporation, and transpiration (and thus root
uptake). A salinity modification which accounts
for mass and diffusion flow was made by Childs
and Hanks (1975). This modification has also
been used to predict the effect of salinity and
irrigation management on crop production. The
Childs-Hanks modification could not account
for the strong source-sink tendencies of the soil
studied at Vernal, Utah as discussed by
Bliesner et al. (1977), so a further modification
has been made as described by Melamed et al.
(1977).
99
-------
WATER MANAGEMENT
MODEL DESCRIPTION
The flow of water in the soil during irriga-
tion must be predicted so soil profile salinity can
be assessed. The basic water flow equation for
one-dimensional flow is
3t 3z
3Z
A(z,t)
(1)
where 6 is volumetric water content, t is
time, z is depth, K(0) is hydraulic conductivity,
H is hydraulic potential and A(z,t) is the root
extraction term. The root extraction term is
given by
. [HR
—
(1.05-z) - h(z,t) - A(z,t)]-RDF(z)-K(e)
Ax-Az
where HR is the effective root water poten-
tial at the soil surface (z = 0), the quantity
"1.05-z" is a combined term to account for
friction loss in the root and gravitational head
loss within the root, h is the soil matric poten-
tial, A is the soil solution osmotic potential, RDG
(z) is a root density function, Ax us the distance
between the roots and the point in the soil where
h(z,t) and s(z,t) apply and As is the depth
increment. Nimah and Hanks (1973) have given
details for the solution of the above two
equations.
= JL [D(0,q) 9C . 3(qC) + aKg (R-Q (3)
9t 3z 3z 9z
where C is the concentration of the soil
solution, q is the water flux, a is 1 or 0
depending on the presence of soluble salts or on
the value of R and C, R is the soil solution
concentration at equilibrium where no
precipitation or dissolution takes place and K s
is a transfer coefficient. If C> R, precipitation
will occure and if C < R dissolution will occur
provided there are soluble salts in the solid
phase. D(0,q) is the apparent diffusion coef-
ficient (the sum of the molecular diffusion and
the hydrodynamic dispersion). More details
involving the solution of the above equation are
given by Childs and Hanks (1975) and Melamed
et al. (1977). Another consideration involved in
the solution of equation (3) is the amount of solid
phase salts in the profile that can be dissolved.
The last term of equation (3) accounts for
the source-sink properties of the soil-water
system. As given, it is empirical and has many
limitiations not the least of which is how to
define the variables as discussed by Melamed et
al. (1977). The value of R and the source strength
must be determined by a calibration type
procedure from field data. The value of R, at
present, is taken as the measured soil solution
concentration provided there is a source of
soluble salts present. The source strength is
estimated from a comparison of the concentra-
tion of the soil solution and the concentration of
a 1:5 or 1:10 dilution. As now defined, the model
is primarily useful for leaching studies because
once the solid phase salts are all dissolved the
value of R is not defined. Further work is
presently being done to quantify this source-
sink term in more basic terms.
MODEL CAPABILITY
The model in its present form is quite
general and has been used for a wide variety of
predictions. Several varieties of input informa-
tion are needed. Soil hydraulic properties must
be known for prediction of water flow. Soil
chemical properties are needed for salinity
prediction. Initial soil water and salinity status
at the beginning of the prediction period are also
needed.
Plant properties are also needed as input
conditions to the model. These include root
depth and plant cover-time relations. The plant
cover-time relations are required to estimate the
portion of potential evapotranspiration that is
potential transpiration.
Climate factors needed as input conditions
are: potential evapotranspiration and rainfall
as related to time during the season. Irrigation
input data needed are: irrigation amount and
salinity, and time of occurrence of the
irrigations.
The model will produce a wide variety of
output information, some of which may not be
useful depending on the reason for the predic-
tion. Soil water and salinity profiles, amount
and salinity of drainage water are among the
most useful data available from the model. Of
particular interest is the prediction of relative
transpiration because this has been shown to be
related to dry matter yield by the following
equation (Hanks, 1974):
Y = T
(4)
100
-------
MODELING SALINITY OF RETURN FLOW
where Y is dry matter yield for transpiration, T,
and Yp is potential dry matter yield where
transpiration is not limited, T p, by soil water.
This yield prediction is very useful for many
management decisions related to irrigation
treatment, timing, etc.
MODEL PREDICTIONS
Many model predictions of irrigation
management effects on salinity of the root zone
have been made as outlined in detail by Childs
and Hanks (1975) and Melamed et al. (1977).
Where the source-sink term is not important and
for the specific field conditions simulated, the
predicted influence of amount of irrigation and
initial salinity of the soil solution are shown in
Table 1. The data indicate an increase in
relative transpiration (thus relative yield) as
irrigation amount increases and also as the
initial soil salinity decreases. The data also
show considerable upward flow from the water
table (negative drainage) where irrigation was
less than water use. Where water movement is
upward from the water table the final salt
concentration increased.
Figure 1 shows some predictions of relative
yield as related to time in years for two different
amounts of water added, both of which were less
than evapotranspiration so that salt was
building up with time. Where irrigation was
highest, relative yield did not decrease for about
6 years whereas with low irrigation, relative
yield decreased continuously. This demon-
strates that even when no leaching occurs, yield
might not be decreased for several years.
One other interesting simulation was done
to evaluate the influence of different values of
irrigation uniformity on relative yield where the
average leaching was the same. Under any
practical irrigation system.water is applied un-
evenly. While the average application rate may
indicate no salt buildup because drainage oc-
curs, there may be areas of the field receiving
sufficiently less irrigation so that no drainage
occurs and thus salt builds up over time. The
results of the predictions, shown in Table 2,
simulate a good sprinkler system (Cu = 0.88)
and a poor gravity system (Cu = 0.42). The
data show that yield varies because of non-
uniform water application and that the high
uniformity system causes higher average yields
than the low uniformity system. After five
years, there was a larger decrease in yields for
the low uniformity. The result is caused by
higher salt buildup, on the average, for the lower
uniformity system. In addition, more salt out-
flow into the drains occurred for the low uni-
formity than for the high uniformity. Many
other simulations were made for an inert soil
(Childs and Hanks, 1975).
However, when comparing the simulation
with the field measurements, it became ap-
parent that the soil was not acting like an inert
medium. When this result became apparent, a
field trial was set up with different leaching
fractions, different depths to the water table,
and different qualities of irrigation water as
described by Bliesner et al. (1977). Figure 2
shows these field results where regardless of the
depth to the water table, there was no ap-
preciable change in soil solution EC. If the soil
were inert the change in EC would have been
appreciable as shown. Similar results were
shown by Bliesner (1977) for different leaching
fractions and quality of the irrigation water.
This is evidence that there was a strong source-
sink potential for this soil. Thus efforts were
made, as discussed earlier, to incorporate the
source-sink effect in the model. The details of
this effort have been reported in detail by
Melamed et al. (1977) and Hanks et al. (1976).
Figure 3 shows a laboratory leaching trial,
using the soil from Vernal,with and without the
source-sink term included. This comparison
between measured and computed EC is good
provided the source-sink term is accounted for.
Figure 4 is the result of a laboratory leach trial
similar to that shown in Figure 3 except that it
was terminated much sooner. The EC of the soil
solution and the quantity of solid salts were
measured throughout the profile. The agree-
ment between measured and computed data is
good. The solid salts were estimated by taking
the EC of the soil solution extract and com-
paring it with the EC of a 1:5 extraction. The 1:5
extraction was assumed to be a measure of the
total salinity for the purposes of these studies.
More dilute extraction would be needed if the
amount of solid phase salts is very large.
The real test of a model must be made for
actual field conditions. This was done and
results are illustrated in Figures 5 and 6. The
field data came from plots with large quantities
of water. When the source-sink term was not
considered in the model, the simulation was
very poor. When a uniform source sink term,
R = 5 mmhos/cm, was used the simulation
101
-------
WATER MANAGEMENT
was also poor. Good simulation was obtained
only when the R value was taken as a variable
with depth. The R value was assumed to be the
initial concentration of the soil solution
measured before irrigation was applied. The
simulation, with a variable R, was not as good
at 26 hours, shortly after an irrigation with 25
cm of water, as at 49.5 hours. This was probably
caused by the value of the transfer coefficient,
K, which was too low.
A simulation using the chemical
equilibrium model of Dutt (1972) was also made.
This model not only did not give good results,
but also gave an almost uniform salinity profile.
This is probably a result of using a uniform ratio
of cations and anions in the computation. This
was done because only limited data were
available. It was assumed that there was gyp-
sum throughout and that the ratios of
Mg++ /Ca++ and Na++/Ca++ were 0.40
and 0.14, respectively, and the ratios of Cl"
/SO 4 and HCO 3 /So 4 were 0.24 and 0.01,
respectively. It may be that a modification of
different chemical species that change with
depth could be found that would give a better
explanation. However, it is difficult to envision
changes that could account for a profile that
was not uniform with depth and did not change
appreciably when irrigated with the large
amount of water used in this study.
It should be emphasized that the source-
sink part of the model reported herein is limited
to fields like those encountered. Having to use a
variable R (measured from field data) causes the
model to be useful only for leaching studies.
Once the excess solid salt in the profile is
leached, there is no reason to expect that the
same value of R would again apply at the same
depth for a situation where precipitation would
occur with different salt species. To be able to
predict the changing value of R will probably
involve complex chemical models. Thus, we
conclude the model is useful to predict changes
involving dissolution but will not be too useful
for predicting salt precipitation until some
logical method for predicting the value of R is
found.
With these limitations in mind we have
made model predictions simulating different
leaching situations using the field data from the
Ashley Valley (Vernal), Utah farm. The soils
and climatic data are the same as used by
Childs and Hanks (1975) except for the initial
salt concentration profiles. The crop studied
was alfalfa with a rooting depth pattern that did
not change with time. A typical growing season
was taken as 125 days and was assumed the
same for all years. Irrigation was applied at
about 15-day intervals at approximately equal
amounts. Potential ET was 81.5 cm of which E p
(evaporation) = 8.8 and Tp (transpiration) was
72.7 cm.
Figures 7 and 8 show the profiles of salt
concentrations in both the solution and solid
salt concentrations in the profile for a series of
model simulation runs totaling 7 years. The
total irrigation applied was 106 cm (30 percent
leaching) each year with a concentration of 2
meq/1 in the irrigation water. With this situa-
tion 25 cm of water leached out of the profile
each year. There was a water table at 160 cm.
The value of R was considered variable with
depth and was taken to be the same as the initial
solution concentration shown in Figure 7. The
amount of solid salt present at the beginning of
the first year was estimated from the difference
between the electrical conductivity of the
saturation extract and 1.5 extract as given by
Melamed et al. (1977).
The predictions of this model show (Figure
9) that it took 7 years to leach the stored salts out
of the soil profile. The amount of salt going into
the drainage water was simply the product of
the relatively constant concentration of the soil
solution just above the water table and of the
total water drained out. The model predicted
precipitation of salts near the bottom of the
profile and solid salt dissolution in the upper
part of the profile. There was a net equal
removal of salt from the profile every year.
These data indicate that a long time period
is needed to cause any significant changes in
the soil profile salt situation. The predictions
agree with the measurements of Bliesner et al.
(1977) that considerable variation in leaching
fractions and irrigation water quality over a
single growing season had very little effect on
the soil solution concentration. This was borne
out by preliminary computations made with a
leaching fraction of near zero which predicted
essentially no change in solution concentration
over several years.
Figures 9 and 10 show soil solution and
solid salt concentrations for a condition of ET
equal to 81.5 cm and with 106 cm of total
irrigation having a concentration if 10 meq/1 in
the irrigation water. In this situation the salts
added are just slightly more than those re-
102
-------
MODELING SALINITY OF RETURN FLOW
moved. After 3 years a stable soil solution con-
centration was predicted. However, there was a
redistribution of solid salts from the surface
toward the bottom of the profile.
IRRIGATION MANAGEMENT
The model results indicate that where the
source-sink phenomenon is important that
there are many management implications. The
first implication is that, over a period of time
such as a year or two, the soil solution and
drainage water concentrations will change very
little regardless of the irrigation management
imposed. Consequently the salt load of the
drainage water can be evaluated by measuring
the water flow to drainage and multiplying it by
the salt concentration of the drainage water.
Since the solution concentration of the drainage
water changes so slowly, yearly measurements
of the solution concentration should be suf-
ficient for purposes of monitoring. Thus, the
most critical measurement is the total water
flow to drainage. This also implies that any
experiment designed to measure influences of
irrigation management variables on drainage
water quality in these systems must be con-
ducted over many years.
A second important implication is that
irrigation management could be practiced with
no leaching at all. Model predictions indicate
this could go on for a long period with no
adverse effects, provided soil water storage
depletion plus irrigation and upward flow are
flow are equal to evapotranspiration.
Caution should be used in accepting this
conclusion. If salts were precipitated, there
would be a limit beyond which there would be no
void space left in the soil for solid salt storage.
This limit would not be reached for a long period
of time, however. An examination of the irriga-
tion management practices over the years in
Ashley Valley, Utah, have indicated that this
situation has actually occurred for many years.
Irrigation is with gravity systems which apply
water with a very low degree of uniformity.
Farmer practice has been to irrigate a few times
a year with high amounts of water. Good yields
have resulted. There is a water table near the
surface (EC of about 2 mmhos/cm) which
supplies water between the infrequent
irrigations. The soils in the area are
heterogeneously permeable which would con-
tribute to a highly variable amount of water
moving through the soil on a field. Yet the yields
are not variable over the fields. Water moving
up from the water table has apparently supplied
crop needs in th under-irrigated areas without
causing yield decreases over the years. Most of
the salts transported by the water moving up
have apparently precipitated out of the soil
solution. This process has been going on for
about 50 years with no problems with crop
production — in fact a very simple and efficient
irrigation management system has developed.
To illustrate further the above situation, the
results of a field experiment on alfalfa con-
ducted in Ashley Valley in 1975 are cited.
Salinity level and irrigation amount variables
were established on mature alfalfa. Six levels of
initial salinity were imposed by adding water
with different levels of salt concentration (using
calcium chloride) at the beginning of the season
with the highest level being 6 mmhos/cm.
Water applied was varied from 0 to more than
evapotranspiration. There was no influence on
alfalfa yields due to either salinity or water
application treatment. Water coming from the
water table apparently supplied the needs of the
plants where no irrigation was applied. There
was a difference in soil profile salinity due to
treatment near the surface but not below 45 cm.
CONCLUSIONS
Use of a model that contains a relatively
simple source-sink modification has provided
an in sight into the behavior of an irrigated soil
that did not respond to water management
variables. Refinement of the model will allow
prediction of a water management technique
that includes both precipitation and dissolution
of the salt in the profile. On a short-term basis,
control of the amount of return flow is the only
way to reduce salt contributions from irrigated
agriculture.
103
-------
WATER MANAGEMENT
TABLE 1
Comparison of irrigation water applied and initial
salt concentration on relative transpiration of the
medium deep root crop T/Tp, total water used,
drainage, salt concentration. Irrigation quality is
6.35 meq/liter. Note that (-) drainage indicated
net upward flow from the water table
Irri-
gation
and
rain
cm
5.6
5.6
5.6
10.3
10.3
10.3
15.0
15.0
15.0
24.4
24.4
24.4
43.2
43.2
43.2
56.4
56.4
56.4
63.5
63.5
Salt flow Initial Final Salt
to salt concen-
Drain ground concen- tration
T T/Tp age water tration average
25.0 0.51
24.0 0.49
15.8 0.33
28.2 0.59
27.2 0.57
19.1 0.40
32.4 0.68
31.4 0.65
23.1 0.49
40.5 0.87
39.6 0.86
32.3 0.70
46.3 1.00
46.3 1.00
44.6 0.96
47.9 1.00
47.9 1.00
47.9 1.00
47.9 1.00
47.9 1.00
cm kg /ha
-9.4 —
-9.1 —
-7.5 —
-9.1 —
-9.0 —
-7.4 —
-8.9 —
-8.8 -
-7.3 —
-8.5 —
-8.3 —
-6.9 -
-5.2 —
-4.9 —
-4.1 —
0.56 90
0.56 247
0.43 740
9.30 1614
9.40 4036
—meq/liter—
20
50
200
20
50
200
20
50
200
20
50
200
20
50
200
20
50
200
20
50
44
102
278
42
94
269
42
93
268
42
91
263
27
62
225
24
52
195
21
44
1.2
I.O
H '
a
ui
s
s!<
I *R=43
T/Tp
O
I-
_1
O
in
i
UJ
4 6
TIME (Years)
K>
Figure 1. Computed relative yield (T/Tp) and
relative soil solution concentration for two different
irrigation amounts as related to time. I is irrigation
and R is rainfall (After Childs and Hanks, 1975).
TABLE 2
crop
Irri-
gation
and
rain Area*
cm
40.4
46.7
53.1
59.5
65.8
%
10.4
24.8
29.6
24.8
10.4
Average
10.6
31.9
53.1
74.3
95.6
Average
20.0
20.0
20.0
20.0
20.0
Rel.
yield
0.85
0.92
0.96
0.99
1.00
0.95
0.50
0.75
0.96
0.96
1.00
0.83
Year 1
Salt
outflow
kg /ha
Cu
—
—
—
516
1233
247
Cu =
—
—
—
2556
8138
2175
Final
soil Rel.
salinity yield
meq/liter
Yeard
Salt
outflow
tons / acre
Final
soil
salinity
meq / liter
= 0.88 Parabolic Distribution
27.6 0.84
25.7 0.91
23.9 0.96
22.4 0.99
21.2 0.99
—
—
—
0.03
0.64
70.8
52.5
38.7
29.6
24.0
0.95 0.14
- 0.42 Rectangular Distribution
32.3 0.39
29.9 0.68
23.9 0.96
20.4 0.96
20.0 1.00
0.80
—
—
—
1.19
3.86
1.02
161.6
50.5
38.7
21.1
20.0
1O4
-------
MODELING SALINITY OF RETURN FLOW
• SHALLOW MATER TABLE
o MEDIUM WATER TABLE
A DEEP WATER TABLE
MEASURED
EXPECTED
ELECTRICAL CONDUCTIVITY (mm ho. /cm)
e/u 6/27 7/9 7/ie e/6 a/i* s/zs
DATE
Figure 2. Measured and expected (assuming an
inert soil) average soil water salinity with time for
three depths of water table (After Bliesner et al.,
1977).
2i 5.0 7.5
NUMBED OF PORE VOLUMES
Figure 3. Computed and measured electrical con-
ductivity of the effluent during a laboratory leaching
trial for Mesa sandy clay from Vernal, Utah (after
Melamed et al., 1977).
ec or THE SOLUTION
(mmhe/em)
456 8 ..345 6 7 8
Figure 5. Measured and simulated comparisons
using no source-sink term (K — 0.0) with a variable
source-sink term (R - Var) for the Vernal, Utah field
trial (after Melamed et al., 1977).
ELECTRICAL CONDUCTIVITY (mmhoi/cm)
3 4 5678 ..345 6
Figure 4. Computed and measured electrical conduc- Figure 6. Measured and simulated comparison us-
tivity of the soil solution, concentration of solid salts ing a constant source-sink term (R = 5.0 mmhos/cm)
and total salts for a laboratory leaching trial like that and a chemical equilibrium model for Vernal field
of Figure 3 (after Melamed et al., 1977). trial (after Melamed et al., 1977).
105
-------
WATER MANAGEMENT
Figure 7. Soil solution concentration profiles com-
puted for 7 years for a total irrigation of 106 cm/year,
with good water quality (2 meq/1) and evapotranspi-
ration of 81.5 cm.
Figure 8. Soil solid salt concentration profiles com-
puted for 7 years for the same conditions as Figure 7.
so ao too
DEPTH-CM
Figure 9. Soil solution concentration profiles com-
puted for 7 years for a total irrigation of 106 cm/year
with poor quality water (10 meq/1) and evapotranspi-
ration of 81.5 cm.
60 80
DEPTH -CM
Figure 10. Soil solid salt concentration profiles
computed for 7 years for the same conditions as
Figure 9.
REFERENCES
1. Bliesner, R. D., R. J. Hanks, L. G.
King, and L. S. Willardson. 1977. Effects of
irrigation management on the quality of irriga-
tion return flow. Soil Sci. Soc. Amer. Jour. (In
press)
2. Childs, S. W., and R. J. Hanks. 1975.
Model to predict the effect of soil salinity on crop
growth. Soil Sci. Soc. Amer. Proc. 39:617-622.
3. Dutt, G. R., M. J. Shaffer, and W. J.
Moore. 1972. Computer simulation of dynamic
bio-physiochemical processes in soils. Agr. Exp.
Sta., University of Arizona, Tucson. Technical
Bulletin 196.
4. Hanks, R. J. 1974. Model for predicting
plant growth as influenced by evapotranspira-
tion and soil water. Agron. Jour. 66:660-665.
5. Hanks, R. J., L. S. Willardson, J. J.
Jurinak, and J. D. Melamed. Managing salini-
ty with irrigation where salinity sources and
sinks are present. Proceedings of Int. Conf. on
Managing Saline Water for Irrigation: Plan-
ning for the Future. Lubbock, Texas, Aug. 16-20,
1976.
6. Melamed, D., R. J. Hanks, and L. S.
Willardson. 1977. Model of salt flow in soil with
a source-sink term. Soil Sci. Soc. Amer. Jour. (In
Press)
7. Nimah, M., and R. J. Hanks. 1973.
Model for estimating soil water, plant and
atmospheric interrelation. I. Description and
sensitivity. II. Field test of the model. Soil Sci.
Soc. Amer. Proc. 37:522-527, 528-532.
106
-------
MODELING SALINITY OF RETURN FLOW
8. Oster, J. D., and J. D. Rhoades. 1975. 9. Tanji, K. K., D. L. Redell, G. V. Terry,
Calculated drainage water compositions and and R. S. Ayeres. 1972. Computer simulation
salt burdens resulting from irrigation with river analysis on reclamation of salt-affected soils in
waters of the Western United States. Jour. Env. San Joaquin Valley, California. Soil Sci. Soc.
Qual. 4:73-79. Amer. Proc. 36:127-133.
107
-------
Field Evaluation
of Sprinkler Irrigation
for Management of
Irrigation Return Flow
L. S. WILLARDSON, R. J. HANKS
Agricultural and Irrigation Engineering Department,
Utah State University, Logan, Utah
and R. D. BLIESNER
Superior Farming Company, Bakerfield, California
ABSTRACT
Sprinkler irrigation offers one alternative
for control of irrigation water application that
influences quality of return flaw. A two-year
field scale study of soil solution response to
controlled water application using sprinklers
was conducted in the Ashley Valley near Ver-
nal, Utah. The design level of water control was
attained with high coefficients of uniformity
but the soil solution did not respond as expected.
The soil acted both as a source and a sink for
salt. Root zone salinity was only slightly
affected by different leaching fractions.
Minimum average leaching fractions ob-
tainable are influenced by both the hydraulic
design and operation of the system.
INTRODUCTION
Mechanical control of irrigation by sprinkle
or drip systems allows application of more
precise amounts of water than do surface irriga-
tion methods. With mechanical irrigation it is
possible to apply water continuously, intermit-
tently, or in any prescribed amount. The only
restraint on a mechanical irrigation system is
the water absorption rate. As long as the
application rate is lower than the infiltration
rate of the soil, water can be applied at any rate
and in any amount. With surface water applica-
tion methods, the amount applied and the
application rate is ultimately controlled by the
soil.
Since sprinkle irrigation allows manage-
ment control of the amount of water applied,
this irrigation method may offer a means of
controlling the amount of return flow. Reduc-
tion in the amount of return flow is usually
accomplished by reducing the leaching fraction
or reducing the amount of water applied relative
to potential evapotranspiration. Studies have
indicated that reducing the leaching fraction
will decrease the total salt loading from irriga-
tion return flow. Rhoades (1) has suggested that
minimizing the amount of drainage water max-
imizes precipitation of calcium carbonate and
gypsum in the soil, minimizes soil weathering,
minimizes the dissolution of salts previously
deposited in the soil and thereby maximizes the
amount of soluble salts retained in storage in
the soil profile.
Bernstein and Francois (2) of the U.S. Sa-
linity Laboratory found that leaching fractions
could be reduced to as little as one-fourth the
values recommended by the U.S. Salinity Lab-
oratory without reducing yields on alfalfa.
Their research indicated that the lower portion
of the root zone may become highly salinized as
long as sufficient good quality water is applied
in the upper root zone.
Leaching fractions as low as two to three
per cent have been recommended in recent
years. Leaching fractions of this low magnitude
are actually being attained due to soil con-
ditions in the Imperial Valley. Measurements of
electrical conductivity drain effluent discharge
indicate that these low leaching fractions are
common. In the Imperial Valley, however, the
infiltration characteristics of the soil and the
low internal permeability rather than method of
109
-------
WATER MANAGEMENT
irrigation controls the leaching fraction. In
other soils which are more permeable, mechani-
cal means of water application control will be
required if low leaching fractions are to be
attained.
FIELD STUDY PROCEDURE
The U.S. Environmental Protection Agen-
cy supported a study by the Agricultural and
Irrigation Engineering Department and the
Soils and Biometeorology Department of Utah
State University to determine the capability of a
commercially available irrigation system to
attain the level of control necessary to achieve
the low leaching fractions suggested in the
literature. The system chosen for evaluation
was a portable solid-set sprinkler system using
No. 30 WSTNT Rain Bird Sprinklers with
0.36 mm (9/64 in.) nozzles on a 9 m by 15 m (30
ft. x 50 ft.) spacing. The operating pressure was
approximately 4.1 atm (60 psi). The system was
used to irrigate 8.1 ha (20 acres) of alfalfa and
designed so that four adjacent lines were
operated simultaneously.
For the purpose of evaluation, a precipita-
tion catch system using cans on a 1.5 m (5 ft)
grid spacing was set up in the approximate
center of the field using the two inner lines of a
four line set. Plants in the area of the catch were
clipped prior to each irrigation to keep the
alfalfa below the tops of the catch cans. The
depths of catch, wind movement, line pressure,
sprinkler discharge, and pan evaporation were
measured for each irrigation throughout the
season. Water was collected for the full duration
of each irrigation.
The distribution uniformity of each irriga-
tion was estimated by using Christiansen's
Coefficient of Uniformity. In addition, several
other water distribution calculations were made
including potential irrigation efficiency, low
quarter distribution uniformity, and low catch
distribution uniformity.
The average operating pressure for all tests
was 4.1 atm (60 psi). The average wind velocity
for each catch ranged from 2.04 km/hr
(1.27 mph) to 8.77 km/hr. (5.45 mph) with an
overall average of 4.07 km/hr (2.53 mph). The
wind direction for each irrigation was west-
southwest. Wind speed in the range shown did
not seem to have a significant effect on unifor-
mity of water application. Full details of the
study are reported by Bliesner (3) and Bliesner,
et al (4).
RESULTS
Results of the can-catch data for seven
individual irrigations are shown in Table 1. The
table shows the depth applied, the average
depth caught, the average caught in the low
quarter of the area, the absolute low catch and
the Christiansen Coefficient of Uniformity.
TABLE 1
Sprinkler uniformity measurements for individual
irrigations.
Date
6/05/74
6/23/74
7/04/74
7/17/74
8/02/74
8/14/74
8/24/74
Depth
Applied
in*
5.17
3.19
3.18
3.28
4.10
3.19
4.93
Ave.
Depth
Caught
in*
4.70
3.07
2.84
3.64
3.93
2.99
4.75
Ave.
Depth
Low 1/4
in*
3.80
2.53
2.49
2.92
3.53
2.53
4.01
Low
Catch
in*
3.57
2.39
2.33
2.58
3.38
2.43
3.65
CU
86.
88.
92.
87.
94.
91.
89.
*multiply by 25.4 to get depth in mm.
Coefficients of Uniformity varied from a low of
86 to a high of 93. The Coefficients of Uniformi-
ty indicate that it is possible to get good
uniformity of water application with a commer-
cial sprinkler system.
The uniformity is further improved if the
water application is composited over the season.
Table 2 shows the same data for the individual
irrigations but combines the effects of in-
dividual uniformity measurements progressive-
ly over the season. Variations that occur in a
single irrigation are sometimes compensated by
subsequent irrigations. Table 2 shows a con-
tinued increase in the Coefficient of Uniformity
over the season. The Coefficient of Uniformity
increased from 86 per cent for the first irrigation
to 94 per cent for the composite of all seven
irrigations.
To determine whether an additional man-
agement practice would improve uniformity of
distribution, a calculated composite uniformity
for the system was developed, assuming that
every other irrigation was an alternate set. That
is, the sprinklers were moved laterally one-half
spacing every other irrigation. Using the data
collected during the season and compositing the
alternate sets over the season, the Coefficient of
Uniformity for the seven irrigations was raised
to 98 per cent as shown in Table 3.
no
-------
In irrigation management to control
leaching, the most important aspect of the
system is distribution uniformity. In order to
avoid salt build-up, the minimum desired leach-
ing fraction should be attained at every point in
the field. Unless the irrigation system is capable
of 100 per cent uniformity of application, all but
the least watered area will receive more than the
desired minimum leaching when the minimum
leaching requirement is applied to the least
watered area.
TABLE 2
Cumulative composite sprinkler uniformities
for the season.
EVALUATION OF SPRINKLER IRRIGATION
Compositing and using the alternate set
procedure resulted in a decrease of the minimum
catch to 91 per cent and the maximum catch to
109 per cent of the average. Further examina-
tion of Figure 2 illustrates the problem of using
very low leaching fractions when actual dis-
tribution uniformity isconsidered. For example:
20 per cent of the field received 97 per cent of the
average application or less. This means that 20
per cent of the field would be accummulating
salt since 20 per cent of the field received 97 per
cent or less than the average application. Or, in
other words, only the portions of the field
receiving 97 per cent of the average application
or more would receive any leaching at all. The
Ave. Ave.
Depth Depth Depth Low
Date Applied Caught Low 1/4 Catch CU
6/05/74 5.17 4.70 3.80 3.57 86.
6/23/74 8.36 7.77 6.48 6.19 87.
7/04/74 11.54 10.61 9.21 8.80 90.
7/17/74 15.30 14.25 12.50 11.93 91.
8/02/74 19.55 18.18 16.51 15.79 94.
8/14/74 22.74 21.17 19.22 18.39 93.
8/24/74 27.67 25.92 23.51 22.60 94.
TABLE 3
Computed alternate set composite sprinkler
uniformities for the season.
Ave. Ave.
Depth Depth Depth Low
Date Applied Caught Low 1/4 Catch CU
6/05/74 5.17 4.70 3.80 3.57 86.
6/23/74 8.36 7.77 7.11 6.90 95.
7/04/74 11.54 10.61 10.11 9.87 97.
7/17/74 15.39 14.25 13.24 12.83 95.
8/02/74 19.55 18.18 17.18 16.90 97.
8/24/74 22.74 21.17 20.07 19.73 97.
8/24/74 27.57 25.92 24.95 23.95 98.
Although Coefficient of Uniformity is a
qualitative measure of water distribution, it is
helpful to see the percentages of the field
involved in over- and under-irrigation. Figures 1
and 2 show the full season composite can-catch
water distribution expressed as a percent of area
receiving a given percentage of the average
application depth or more. Figure 2 shows the
effect of compositing alternate can-catch data to
improve the distribution. Figure 1 shows that
the minimum application was approximately 86
per cent of the average and the maximum
application was 118 per cent of the average.
salt. F
alternj
cent o1
a 3 pei
80
85
_c
£• 90
Q
& 95
o
| 100
^ 105
*c
I "0
O-
115
120 '
Figure
tributic
given p
80
85
f 'O
0
8, 95
O
; 100
<
"o 105
8 110
£
115
120
Figure
catch \
area r<
more.
"igure 2 shows the seasonal composite
ate set can-catch and indicates that 12 per
the field would be accumulating salt with
' cent leaching fraction.
111!
0 20 40 60 80 100
Percent of Catch Area
1. Seasonal composite can catch water dis-
n expressed as the percent of area receiving a
ercent of average depth or more.
i i i i
i i i i
0 20 40 60 80 100
Percent of Catch Area
2. Seasonal composite alternate set can
vater distribution expressed as the percent of
jceiving a given percent of average depth or
Ill
-------
WATER MANAGEMENT
The above data refers only to the uniformity
of water distribution within four sprinklers at a
given location in the field. Another factor ef-
fecting the water distribution of the sprinkler
system is the normal pressure variation design-
ed into the system. Under strict management or
automation, it is possible to operate a system so
that the supply ends of all the laterals are at the
same pressure. If the standard 20 per cent
friction loss in a lateral is allowed, then the
pressure at the first outlet is 1.25 times the
pressure at the last outlet. The discharge at the
head of the line, therefore, is 12 per cent higher
than the discharge at the end of the line, and the
average discharge is 2-1/2 per cent higher than
the end discharge. Combining this pressure
variation along the lateral with the results from
the can-catch analysis yields a minimum at-
tainable average leaching fraction for the total
field of 15 per cent for the seasonal composite
and 10 per cent for the alternate set composite
pattern assuming that the least watered area
receives zero leaching. The maximum leaching
fraction anywhere in the field would be 34 per
cent for the seasonal composite and 24 per cent
for the alternate set composite. Again, the
absolute minimum leaching fraction would be
zero.
Since the uniformity values of this field
system are high compared to most field
sprinkler systems, the minimum average
leaching fraction that could realistically be
expected from a solid set sprinkler system would
be 10 percent. To obtain average leaching
fractions lower than this minimum value will
cause areas of salt build-up in the field since part
of the area actually has a zero leaching fraction.
The resulting salt build-up could be detrimental
to crop production.
As part of the same experimental effort, a
study was made of the effect of leaching fraction
on soil salinity profiles using the sprinkler
system to apply water. The experiment was
designed to test the interaction of three leaching
fractions and the salinity of the applied water.
Leaching fractions were designed to be zero and
were otherwise varied to give a theoretical drain
discharge conductivity of 25 mmhos/cm and
8 mmhos/cm. The leaching fractions were
referred to as low, medium and high. The
average leaching fractions were approximately
0, 10 and 25 per cent. Figures 3, 4, and 5 show
profile salt contents on three dates during one of
the irrigation seasons for low, medium, and
high leaching fractions, respectively. The salt
0
i
2
£ 4
a
u
e
~
8
9
1C
• 6/11/74
O 7/18/14
A 8/28/74
I.O 2.O 3.O 4.O 5.O
ELECTRICAL CONDUCTIVITY mmho/fcm
Figure 3. Average soil solution salinity profile for
low leaching plots on three dates.
0
I
2
3
Ul
O
6
7
£
9
• 6/11/74
O 7/18/74
A 8/28/74
I.O 2.O 3.0 4.0 5.0
ELECTRICAL CONDUCTIVITY mmho/cm
Figure 4. Average soil solution salinity profile for
medium leaching plots on three dates.
112
-------
EVALUATION OF SPRINKLER IRRIGATION
content of the soil profile did not change
significantly under any of the treatments im-
posed during the irrigation season. The same
result was observed during the second year that
the study was conducted. Figure 6 shows the
measured and expected average soil salinity
changes with time for the three leaching
treatments. The expected values of average soil
water salinity rose to values above
7 mmhos/cm, but the average measured values
remained at approximately 4 mmhos/cm. This
particular soil acts as either a source or a sink
for salt depending on water management.
Figures 3, 4, and 5 indicate that short term
management of leaching fractions will not
significantly effect the salinity at the bottom of
the root zone. Management of the return flow
becomes a matter of managing the quantity of
return flow on a short term basis, since the
quality of return flow shows little change due to
management.
0
I
u. 4
5
I
CL
_
Q
• 6/11/74
O 7/18/74
A 8/28/74
I.O 2.O 3.O 4.O 5.O
ELECTRICAL CONDUCTIVITY mmho/cm
Figure 5. Average soil solution salinity profile for
high leaching plots on three dates.
CONCLUSIONS
Sprinkle irrigation offers one alternative
for irrigation water application control that will
influence the quantity of return flow. The field
studies conducted emphasized the importance
of obtaining uniform water application in order
to prevent salt build-up in parts of the field.
Compositing the water application over a full
season improves water distribution but there
are still areas within individual sprinkler
patterns that receive inadequate amounts of
water to accomplish minimum leaching.
Variations along sprinkler laterals will further
complicate the nonuniformity problem. In the
> 6.0
1 !
• Low Leaching
c Middle Leaching
A High Leaching
Measured
- — Expected
6/H 6/27 7/9 7/18 6/6 8/14 8/28
DATE
Figure 6. Measured and expected average soil water
salinity changes with time for three leaching
treatments.
field, the leaching fraction varied from 0 to 34
per cent with an average of 15 per cent assuming
that the least watered area was irrigated to
exactly meet evapotranspiration. The alternate
set composite showed a minimum application
amount of 92 per cent of the average applica-
tion. The average leaching fraction for com-
posite alternate sets varied from 0 to 24 per cent
with an average of 10 per cent. Uniformity of
application and control of the amount applied is
the key to successful management of irrigation
return flow by sprinklers.
REFERENCES
1. Rhoades.J. D.,R. D. Ingvalson,J. M.
Tucker and M. Clark. 1973. Effects of irrigation
water composition, leaching fraction and time
of year on the salt compositions of irrigation
and drainage waters. Soil Science Society of
America, Proceedings 37;770-773.
2. Bernstein, L. and L. E. Francois. 1973.
Leaching requirement studies; Sensitivity of
alfalfa to salinity of irrigation and drainage
waters. Soil Science Society of America, Pro-
ceedings 37;931-942.
113
-------
WATER MANAGEMENT
3. Bliesner, R. D. 1975. A sprinkler irriga- 4. Bliesner, R. D., R. J. Hanks, L. C. King
tion system evaluation and irrigation manage- and L. S. Willardson. 1977. Effects of irrigation
ment technique to maintain salt storage in the management on the quality of irrigation return
root zone. Unpublished thesis. Agricultural and flow. Soil Science Society of America, Proceed-
Irrigation Engineering Department, Utah State ings (in press).
University, Logan, Utah.
114
-------
Effects of Irrigation
Management on Soil Salinity
and Return Flow Quality
P. J. WIERENGA and J. B. SISSON
Department of Agronomy, College of Agriculture and Home Economics,
New Mexico State University, Las Cruces, New Mexico
ABSTRACT
A field plot study was conducted to deter-
mine the effects of controlled surface irrigation
and trickle irrigation on soil salinity, and on the
quality and quantity of irrigation return flow.
Changes in soil salinity were determined by
taking extensive soil samples at least once a
year. Return flow quality was measured on soil
solution removed through vacuum samplers in
the subsoil of each plot, and by analyzing the
water quality at various levels below the
groundwater and in a nearby drain. The quanti-
ty of return flow was estimated from crop
growth and weather data, and amounts of water
applied. This method was found more reliable
than several other methods used for deter-
mining percolation losses.
Results of the first three years of this on-
going study showed that a larger change in soil
salinity was produced by altering irrigation
frequency than by changing irrigation efficien-
cy. Irrigating when 50 percent of the soil-water
had been depleted was the irrigation frequency
most conducive to salt retention in the upper soil
profile. However, different irrigation efficien-
cies and frequencies had no significant effect on
cotton yield due in part to the large spatial
variability of the physical properties of the soil
at the site. The use of trickle irrigation was an
effective method for controlling the volume of
return flow, while maintaining relatively low
salinity levels in the soil around the trickle lines.
Accumulated salts were readily moved from the
trickle lines by preplant irrigations and by
intense rains. The mean salt concentration of
the irrigation return flow, as estimated on soil
solution samples removed by vacuum extrac-
tion from behw the root zone, agreed well with
the average salt concentration of the upper
groundwater, but not with the quality of he
deeper groundwater. The quality of the drain
water was similar to the average composition of
groundwater. The results of this study indicate
that more efficient irrigation can reduce the
volume of return flow, without immediate
detrimental effects on crop yields or soil salini-
ty.
INTRODUCTION
Maintaining the quality of river water is a
major problem in the Western United States. In
New Mexico, the salt concentration in the Rio
Grande increases progressively going down-
stream from Santa Fe, in the north, to El Paso
on the New Mexico-Texas border in the south.
The greatest increases in salt concentration
occur in reaches where irrigated agriculture is
practiced. The increase per mile in concentra-
tion of dissolved solids was found to be more
than twice as great in an irrigated area along
the Rio Grande as in the non-irrigated area
immediately to the north (Wierenga and Patter-
son, 1972). This greater increase in salt concen-
tration in irrigated reaches is related to irriga-
tion return flow. Irrigation return flow origi-
nates when water in excess of evapotranspira-
tion is applied to the soil-plant system. The
excess water is often applied intentionally to
leach soluble salts from the plant root zone and
insure perennial agriculture. That portion of the
excess water flowing back to the river is called
irrigation return flow.
The management of irrigation to minimize
the effects of irrigation return flow water on
downstream river water has received con-
siderable attention in the past few years. For
example, Rhoades et al. (1973, 1974) demon-
strated that the salt burdens of percolated
115
-------
WATER MANAGEMENT
irrigation drainage waters could be appreciably
reduced by minimizing the leaching fraction.
Reduced drainage volumes effectively increased
the precipitation of salts, reduced weathering of
soil minerals, and reduced the total salt in the
drainage waters. Suarez and Rhoades (1977)
studied the effects of irrigation leaching
management on river water quality. They con-
cluded that the reduced salt load in the return
flow may or may not reduce the salt load of river
water, depending on the successive mixing and
readjustment ot lower CO 2 levels of the return
flow and the type or irrigation water. However,
reduced leaching will result in reduced volumes
or return flow. This latter conclusion was also
reached by King and Hanks (1975), who found
that total seasonal salt discharge from a tile
drainage system was directly related to the
quantity of water discharged. Reducing the
volumes of return flow can be accomplished by
improved irrigation management, including
uniform application of water by controlled
flooding, furrow, sprinkler, or trickle irrigation.
Applying the optimum amounts of irrigation
water for given climatic and soil conditions
using irrigation scheduling programs such as
the one developed by Jensen (1972), will also
reduce the volume of return flow. The objectives
of this paper are to present data on the effects of
irrigation treatments on soil salinity in field
plots irrigated with 1.2 mmhos / cm water over a
three-year period.
METHODS AND MATERIALS
The experimental work was conducted on
the Plant Science Research Center of New
Mexico State University, about eight miles
south and west of Las Graces, New Mexico.
The soil at the site was heterogeneous
which was characteristic of the alluvial soils
along the Rio Grande Valley. The soils at the
site were classified "typic torrifluvents" and
"vertic torrifluvents". The soil profile consisted
in general of about 30 cm silty clay loam over 20
to 30 cm of clay, over a variable amount of silty
loam. The latter changed rather abruptly into a
medium sand. However, the fine textured
material in some plots was deeper, so that the
depth to the sand on the two-acre site varied
from 60 to 120 cm below the soil surface. Root
penetration into the medium to fine sandy
subsoil appeared to be almost negligible.
The experimental design consisted of thirty
7.3 x 7.3 meter plots. The plots were separated
from the surrounding soil with polyethylene
plastic to a depth of 75 cm below the soil surface.
Wooden boards extending 20 cm above the soil
surface and 10 cm below, covered with
polyethylene plastic prevented surface runoff
from the plots. Irrigation was by surface
flooding the area within plot boundaries. All
water was metered through a single flow meter
installed in the main line. Prior to construction
of the field plots, the area was leveled and flood
irrigated with approximately 20 cm water.
Further details on the experimental setup can be
found in Wierenga (1977).
The treatments on the surface irrigated
plots consisted of combinations of three fre-
quencies of irrigation, with three field water
application efficiencies, resulting in a total of 9
treatments. Each treatment was block ran-
domized with three replications per treatment.
The application efficiencies were 50,75 and 100
percent for the first year of study, and 80,90 and
100 percent efficiency for the last two years of
the study. Frequency of irrigation was based on
the amount of "available" water in the soil, with
irrigations scheduled when 25, 50 or 75 percent
of the available water was depleted from the
root zone of the crop. Consumptive use at the
experimental site was obtained during the first
year of study from Class A pan evaporation
data, corrected with a stage of growth depend-
ent crop coefficient. During the second and third
year of study, consumptive use was estimated
from climatic data, using the Irrigation
Management Scheduling Program (IMS),
developed by Jensen (1972).
All plots were planted with cotton (Acala
1517-70) on a 1 m row spacing. Fertilizer was
applied uniformly at a rate of 134 kg N / ha and
76kgP2O5/ha.
Efforts were made to determine the
amounts of return flow, e.g. deep percolation
water, from knowledge of the hydraulic conduc-
tivity and the hydraulic gradient as measured
in each plot. Triplicate tensiometers were in-
stalled at two levels in each plot, well below the
root zone in the sandy subsoil. The actual
depths varied between 120 and 180 cm, below
the surface, with a vertical distance of 30 cm
between the two sets of tensiometers. Soil water
tensions were determined with mercury man-
ometers located outside each plot. Hydraulic
conductivity as a function of water content was
determined for three plots following the
procedure described by Nielsen et al. (1964).
116
-------
EFFECTS OF IRRIGATION MANAGEMENT
This procedure consists of thoroughly wetting
the soil profile, covering the soil with
polyethylene plastic, and measuring the hy-
draulic gradients and water contents during the
drainage stage following infiltration. From the
changes in water content of the soil profile with
time and the gradients, the hydraulic conduc-
tivity could be calculated at various water
contents.
Water contents within each plot were deter-
mined with a neutron meter. A calibration curve
was established from neutron meter readings
and gravimetric sampling of the soil within the
plot area.
The quality of return flow was monitored
with soil solution samplers installed at depths
halfway between the two levels of tensiometers
in each plot. The samplers were made with
standard tensiometer cups, 5.8 cm long, at-
tached to 2.2 cm O.D. PVC tubing. Vacuum
was applied through a central vacuum system.
At the end of each cropping season soil
samples were taken within each plot at 20 cm
depth intervals to a depth of 160 cm below the
soil surface. Samples were taken at two
locations in each plot, and composited for
laboratory analysis. The electrical conduc-
tivities of saturation extracts of these samples
were determined.
At a weather station adjacent to the plot
area, measurements were made of the daily
incoming radiation, daily wind run, maximum
and minimum temperature, relative humidity,
rainfall, and pan evaportation.
The irrigation water used in the experiment
came from an 8-inch well just outside the plot
area, and had the following average concen-
trations of ions, expressed in meq per 1: Ca+2 >
7.73; Mg +2,1.68; Na+, 5.46; K +, 0.20; CF, 2.53;
HCO'3, 5.16; SO-4, 5.56; NO"3, 0.01. The
water had an average EC of 1.22 mmhos/cm.
RESULTS
Table 1 presents the potential gradients,
averaged for all plots, as measured in the sandy
subsoil of the plots during the 1972 and 1973
growing seasons.
At all times and for all plots the hydraulic
gradients in the subsoil were near minus one, as
expected, and directed downwards, indicating
some downward movement of water. With these
gradients and values of the hydraulic conduc-
TABLE 1
Hydraulic gradients (cm/cm), averaged for all plots
and for each month, during the 1972 and 1973
growing seasons. Each number is the average of 450
observations.
1972
1973
June
July
August
September
-0.86
-1.10
-0.84
-0.58
-0.61
-0.70
-0.90
-0.76
tivity (K0)), drainage rates may be calculated.
However, values of the hydraulic conductivity,
determined during transient drainage ex-
periments on three plots, showed large
variations with water content in the subsoil
(Wierenga, 1977). Furthermore, subsoil water
contents varied greatly between plots. For ex-
ample, the mean water content at the 120 to
150 cm depth for all plots during the 1973
irrigation season was 0.125 cm^/cm^ with a
standard deviation of 0.04 cm 3/cm 3. The large
variation in subsoil water contents, the
steepness of the hydraulic-conductivity versus
water content relationships, (K(0)), and the
errors in water content measurements, as well
as in the determination of the hydraulic conduc-
tivity values (the latter could be more than 100%
in the range of interest here (Fliihler, et al. 1976),
made it difficult to obtain reliable drainage
fluxes on the basis of K(0) relationships deter-
mined for three plots only. For these reasons,
the total volume of return flow during each
season was estimated as the difference between
irrigation water applied and estimated con-
sumptive use. Consumptive use was estimated
from weather data with empirical coefficients
for computing actual evapotranspiration from
potential evapotranspiration (Wierenga, 1977).
An analysis of variance of the estimated
drainage losses showed significant differences
due to efficiency and depletion treatments.
Differences among depletion treatments, al-
though very small, were not intended, but re-
sulted from timing of irrigations and the cut-off
date of irrigations near the end of August.
An important soil physical property in-
fluencing salt retention in a soil is field capaci-
ty. Field capacity was estimated for each plot as
the total water held in the soil profile immediate-
ly following preirrigation. The variation among
the field capacities were investigated with an
analysis of variance. It was established that the
117
-------
WATER MANAGEMENT
field capacity values were uniformly distributed
among the treatments, with a coefficient of
variation of ±16% from a mean value of
38.4 cm of water.
Analysis of variance of the field capacity
and drainage losses indicated that the experi-
ment was effectively carried out as designed
and that field capacity was distributed uniform-
ly among the irrigation treatments.
The quality of return flow was estimated
from soil solution samples obtained through
suction cups, and also from saturation extracts
of soil samples removed from the 140-160 cm
depth under each plot. The mean electrical
conductivities of soil solution samples extracted
during the 1972, 1974 and 1974 growing season
were 8.10, 7.89 and 7.63 mmhos/cm, respective-
ly. The mean electrical conductivity of the
irrigation water during these years was
1.22 mmhos/cm. Thus, on the average, the
ratio EC irrigation water/EC drainage water
was 6.5. An analysis of variance was made for
the electrical conductivities of the soil solution
samples collected during the three years of the
study. This analysis showed no significant
changes between 1972 and 1974 in the quality of
return flow, as measured on soil solution
samples from the 130-160 cm depth. The only
significant effects of irrigation treatments on
the concentration of soil solution were the
depletion treatments. The treatment means for
depletions are presented in Table 2. Table 2
shows that for the combined years the mean EC
of the soil solution below the 75 percent deple-
tion plots was significantly lower than the EC of
the soil solution below the 25 and 50 percent
depletion plots, indicating enhanced leaching
by less frequent irrigation, but using the same
total amount of leaching water. Care should be
taken in interpreting the results, since there was
much variation in the EC of the soil solutions
extracted from the subsoil of each plot, as well
as in the EC's of the saturation extracts.
Figures 1-3 present frequency distributions
of the electrical conductivity of the saturation
extracts multiplied by percent saturation for
samples taken in 20 cm depth intervals to
160 cm from each plot. The EC e x percent
saturation values were plotted versus their
relative cumulative frequency on normal
probability paper. The EC e values were mul-
tiplied by percent saturation to reduce errors in
the preparation of saturation pastes between
years.
TABLE 2
Treatment means of the electrical conductivities
(mmhos/cm) of the soil solution samples extracted
through suction cups in 1972, 1973 and 1974.
1972
1973
1974
Mean
9
25
9.73
8.70
6.84
8.42 a
fo Depletion
50
9.12
7.80
9.53
8.82 a
75
5.44
7.17
6.53
6.38 b
All
depletions
8.10
7.89
7.63
7.87
a,b — Numbers followed by common letter not signif-
icantly different
For a normal distribution of the ECe x
percent saturation values the symbols should
follow a straight line. The logarithms of the
ECe x percent saturation values were also
plotted. Fig. 1 presents distributions for all
depths, while Fig. 2 and Fig. 3 represent dis-
tributions for the 0-20 and 140-160 cm depths,
respectively. When all depths were considered
simultaneously (Fig. 1) the frequency dis-
tributions were not linear, but showed linear
sections, indicating the presence of two or more
soil populations, attributable to extensive soil
layering at the site (clay over sand with greatly
varying depths of clay). The frequency dis-
tributions of the log IQ (EC x % saturation)
values for the 0-20 cm depth was nearly linear
(Fig. 2), indicating a log normal distribution of
the ECe values at that depth. At 140-160 cm,
neither of the two frequency distributions
presented were linear. Since the variation in
depth to sand was particularly great in the third
replicate, frequency distributions were also con-
structed for the ECe values from the first two
replicates only (not presented here). By
eliminating replicate 3, the frequency distrib-
tion of the values of EC e x percent saturation
appeared to be closer to a straight line, in-
dicating that the bimodal distributions in Fig. 3
were caused to some degree by nonuniformity of
the soil in replicate 3. However, it is not clear
what causes the ECe x percent saturation
values at the 0-20 cm and 140-160 cm depths to
be other than normally distributed. The im-
plications of log normal distributions of soil
salinity values on the quality of return flow
have not yet been fully investigated.
An analysis of variance was also made on
the soluble salt content of the soil in each plot.
118
-------
EFFECTS OF IRRIGATION MANAGEMENT
The soluble salt content of the soil in each layer,
expressed in g/100 g soil or percent salt, was
computed from the ECe of each layer. The
coefficient for computing soluble salt content
from the EC e was obtained by analyzing one set
of samples for anions and cations, and using the
equivalent weights of these ions as well as the
saturation percentage of each sample. No
significant effects of treatments on the salt
contents in the soil were found. The only factors
which had a significant effect on soil salinity
were depth and the year x depletion interaction.
The significant variation by depth was ex-
pected, since the soil profiles were sharply
stratified. The large spatial variation in soil
salinity (Fig. 2 and 3) was not expected,
however. An index of variation is the standard
deviation of the mean soil salinity. For depths
greater than 100 cm, the standard deviation
was typically as large as the mean. In general,
spatial variation in ECe and salt content
appeared to be greater in the subsoil than in the
topsoil. This was also evident when Figures 2
and 3 were compared. Soil salinity was indepen-
dent of the efficiency treatments, which was not
too surprising considering the relatively small
differences in the amounts of water applied on
the three efficiency treatments, and the extreme
natural variability of the experimental site.
Table 3 summarizes the data on soluble salt
content in the soil, determined from saturation
extracts.
TABLE 3
Mean soluble salt contents (g/100 g of soil), estimated from
saturation extracts, by depth for each sampling period.
Depth (cm) All
0-20 20-40 40-60 60-80 80-100 100-120 120-140 140160dep.
Control
Fall '72
Fall '73
Fall '74
.20
.12
.13
.13
.25
.19
.17
.23
.33
.36
.35
.34
.20
.30
.33
.36
.15
.16
.17
.19
.10
.11
.12
.12
.07
.08
.09
.08
.04
.06
.07
.06
.17
.17
.18
.19
The data in Table 3 show a large increase in
soluble salt content to about 0.3 g/100 g at 80 cm
and then a decrease to .06 g/100 g at 160 cm. For
most of the plots, the clay/sand interface was
approximately at 70 to 80 cm, although there
was much deviation. Thus, it appears that
extensive accumulation of salt at or above the
clay/sand interface has taken place over the
years. During the three years of the experiment,
the salt concentration in the surface soil was
lower than in the control, while the salt concen-
trations at 60-80 cm showed an increase. The
salt content averaged over all depths showed a
slight increase between 1972 and 1974.
However, the increase is relatively small, e.g.
from 0.17 g/100 g to 0.19 g/100 g, or about 10%.
SUMMARY AND CONCLUSIONS
The results of this study showed that, for
our conditions, soil salinity and quality of
irrigation return flow were relatively insen-
sitive to irrigation efficiency and leaching
treatments, over a three-year period. The salt
concentration of the soil solution of the sandy
subsoil did not vary significantly during the
experimental period. The same was true for the
salt concentration in the saturation extracts of
soil samples removed from the sandy subsoil.
Observations similar to ours were made by
Bliesner et al. (1977), who found insignificant
increases in average salinity in soil profiles
irrigated with water having EC's of 0.90, 2.22
and 2.80 mmhos/cm, respectively, and
leaching fractions varying between .003 and
.343. In order to predict salinity profiles in such
soils, Hanks et al. (1976) developed a model with
an empirical source-sink term to account for the
behavior of precipitated salts in the soil system
and the lack of changes in profile salinity under
different leaching practices. Although the
chemical precipitation of salts was not studied
here, it was possible that precipitation and
dissolution of soil salts were important factors
in our soils. This was indicated by the apparent
accumulation of salt near the clay-sand inter-
face.
The spatial variability of soil salinity in the
soil of this study greatly complicated the assess-
0-160 cm
--EC "SAT %
•-I08,.IECI SAT.%)
1512 OBSERVATIONS
_J 1 , L_
Ql I 5 10 30 50 70 90 95 99 993
RELATIVE CUMULATIVE FREQUENCY (%)—»•
Figure 1. Relative cumulative frequency of ECe
(mmhos/cm) x saturation (%) for the 0-160 cm depth.
119
-------
WATER MANAGEMENT
ment of the effects of irrigation treatments on
soil salinity. Spatial variability in soil salinity
was caused by differences in soil profile
characteristics, but may also have been related
to local variations in leaching patterns, which
were observed at the site (van der Pol, et al.,
1977). Refined statistical techniques will have to
be used to determine the effects of treatments
under our field conditions.
r
210 -
t
f
S '»[-
a \-
169 OBSERVATIONS
. .0°
Cll I 5 10 30 50 70 909599
RELATIVE CUMULATIVE FREQUENCY (%)—••
Figure 2. Relative cumulative frequency of ECe
(mmhos/cm) x saturation (%) for the 0-20 cm depth.
. !
hooj-
-» W^tEOSAT.X)
18V OBS£«YAT»OHS
St
*
a 60 t-
OJ i 510 30507D 909599 999
RELATtVE CUMULATIVE FREQUENCY ttW—••
Figure 3. Relative cumulative frequency of ECe
(mmhos/cm) x saturation (%) for the 140-160 cm
depth.
REFERENCES
1. Bliesner, R. D., R. J. Hanks, L. G. King
andL. S. Willardson. 1977. Effects of irrigation
management on the quality of irrigation return
flow in Ashley Valley, Utah, Soil Sci. Soc. Amer.
J. (In press).
2. Fluhler, H., M. S. Arkadani and L. H.
Stolzy. 1976. Error propagation in determining
hydraulic conductivities from water content
and pressure head profiles. Soil Sci. Soc. Amer.
J. 40:830-836.
3. Hanks, J. J., L. S. Willardson, J. J.
Jurinak and J. D. Melamed. 1976. Irrigation
management where salinity sources and sinks
are present. In H. E. Dregne (ed.), Managing
saline water for irrigation. Proceedings of the
Intern. Conf. on Managing Saline Water for
Irrigation. Lubbock, Texas, 16-20 Aug., 618 p.
4. Jensen, M. E. 1972. Programming
irrigation for greater efficiency. In D. Hillel
(ed.), Optimizing the soil physical environment
toward greater crop yields. Academic Press,
New York. pp. 133-161.
5. King, L. G. and R. J. Hanks. 1975.
Management practices affecting quality and
quantity of irrigation return flow. Environmen-
tal Protection Technology Series, EPA-660/2-
75-005, 156 p.
6. Nielsen, D. R., J. M. Davidson, J. W.
Biggar and R. J. Miller. 1964. Water movement
through Panoche clay loam soil. Hilgardia
35:491-506.
7. Rhoades, J. D., R. D. Ingvalson, J.
M. Tucker, and M. Clark. 1973. Salts in irriga-
tion drainage waters. I. Effects of irrigation
water composition, leaching fraction, and time
of year on the salt composition of irrigation and
drainage waters. Soil Sci. Soc. Amer. Proc.
37:770-774.
8. Rhoades, J. D., J. D. Oster, R. D.
Ingvalson, J. M. Tucker, and M. Clark. 1974.
Minimizing the salt burdens resulting from
irrigation with river water in the western
United States. J. Environ. Qual. 3:311-316.
9. Suarez, D. L. and J. D. Rhoades. 1977.
Effects of irrigation leaching management on
river and soil water salinity. In H. E. Drenge
(ed.), Managing saline water for irrigation.
Proceedings of the Intern. Conf. on Managing
Saline Water for Irrigation. Lubbock, Texas, 16-
20 Aug. 1976, 618 p.
10. van der Pol, R. M., P. J. Wierenga and
D. R. Nielsen. 1977. Solute movement in a
layered field soil. Soil Sci. Soc. Amer. J.41: (In
press).
11. Wierenga, P. J. 1977. Influence of
trickle and surface irrigation on return flow
quality. Completion report EPA Project 13030
GLM. 179 p.
12. Wierenga, P. J. and T. C. Patterson.
1972. Irrigation return flow studies in the
120
-------
EFFECTS OF IRRIGATION MANAGEMENT
Mesilla Valley. In Managing irrigated
agriculture to improve water quality.
Proceedings National Conference on Managing
Irrigated Agriculture to Improve Water Quality.
May 16-19:173-179.
121
-------
Effect of Irrigation Systems
on Water Use Efficiency
and Soil-Water Solute
Concentrations
C. W. WENDT, A. B. ONKEN, O. C. WILKE, RAFORD HARGROVE,
WALTER BAUSCH and LARRY BARNES
Texas Agricultural Experiment Station,
Lubbock, Texas
ABSTRACT
The effect of sprinkler irrigation (SpJ,
furrow irrigation (F), subirrigation (Su), and
automated subirrigation (ASu) systems on
water use efficiency and soil-water solute con-
centration was evaluated at a field site in Knox
County, Texas. Significant differences existed
in the irrigation water requirement of sweet
corn irrigated by the different systems
(F > Sp = Su > ASu). However, little difference
in total water requirement existed between
systems — the soil water was utilized more
efficiently (ASu > Su s Sp > F). Automation of
irrigation systems offers the possibility of
significantly enhancing irrigation water use
efficiency of supple?nental irrigated areas.
Soil-water solute concentrations were too
low to be of concern relative to degrading the
quality of irrigation return flows due to the
dilution effect of rainfall. The relative ranking
of the different irrigation systems for the
various solutes was:
Chloride
Sulfate
Sodium
Calcium
Magnesium
Potassium
Ammonium
Conductivity
-F>Sp>Su> ASu
-F>Sp = Su> ASu
- Su > F > Sp > ASu
-F>Sp>Su> ASu
- Sp > F > Su > ASu
- Sp > Su > ASu > F
- ASu > Su > Sp > F
-Sp>F>Su> ASu
In most cases, at depths below 3.0 m, the
concentrations of the above solutes in the soil
were less than those of the irrigation water.
Although soil-water solutes were not a
problem in the irrigation return flows, the lower
concentrations of the ASu and Su systems
indicate that they may be superior to F and Sp
systems in maintaining the quality of irrigation
return flows where solutes are a problem.
INTRODUCTION
In 1970, a cooperative project was initiated
between the Environmental Protection Agency
and the Texas Agricultural Experiment Station.
The major emphasis of the study was to deter-
mine if current irrigation and fertilization prac-
tices were contributing to the increasing nitrate
of the ground water and to develop management
practices, if it was determined that fertilizers
were a major contributor to the nitrate in the
water table. The results of these studies are
reported in a paper in the Nitrogen in Return
Flows section of this conference.
Also included in the study was an evalua-
tion of the influence of sprinkler irrigation,
furrow irrigation, subirrigation, and automated
subirrigation on water use efficiency and soil-
water solute concentration. It is the purpose of
this paper to summarize the results of these
studies.
METHODS AND MATERIALS
The study was conducted on a Miles loamy
fine sand (Udic Paleustalfs) located in Knox
County, Texas. Four types of irrigation systems
were evaluated in the study: sprinkler irrigation
(Sp), furrow irrigation (F), manual subirriga-
123
-------
WATER MANAGEMENT
tion (Su) and automated subirrigation (ASu).
The pipe of the subirrigation systems was
located at a depth of 30 cm and spaced 102 cm
apart. A total of 45 plots were available for
various aspects of the study. The plots of each
system were 16 102-cm rows wide and 60.8 m
long. Details concerning the irrigation system
design have been previously described in other
publications (7) (8).
The meteorological equipment necessary to
estimate evapotranspiration potential (ETp)
according to the method of Jensen et al (4) was
located at the site. To determine leaf area index
(LAI) on the plots, a relationship between leaf
area and stem diameter of the corn was derived
so that leaf area measurements of a large
number of plants could be made.
The generalized water budget model used in
comparison was:
AW = M + Ir - N - F - (E+T) [Hillel (3)j
where A - change in water content, M ~ pre-
cipitation, Ir = irrigation water, N ~ runoff,
F - deep percolation, E+T = evapora-
tion + transpiration.
During the growing season, there was no
runoff from the plots. Also, the location of peak
concentrations of a bromide tracer in the root
zone remained nearly constant during the grow-
ing season indicating that there was negligible
deep percolation during the study period (5).
Since there was no runoff or deep percolation
from the plots, the above equation became
ET = M+ Ir - AW. Precipitation (M) was
measured with a weighing rain gauge, Ir with
meters and AW with calibrated neutron probe.
Water was applied when the potential at
30 cm decreased to -40 cb in the Sp, F, and Su
plots. Due to the porosity of the loamy fine sand
soils, it was not possible to apply less than
76 mm of irrigation water per application with
the F system. In the Sp and Su plots, it was
possible to apply a percentage of ETp at each
application varying according to the LAI-evap-
otranspiration relationship presented by
Ritchie and Burnett (6). Water was applied as
needed to the ASu plots when the potential
decreased to -40 cm until the potential increased
above -40 cb.
Sweet corn [(Zea mays L.) var. Bonanza)
was grown on the plots and treated with cultural
practices necessary to assure adequate growth
and yield. All plots were fertilizeed with 112 kg
of nitrogen/ha. Eight replications of yield data
were obtained from each irrigation system.
A soil-water extraction system was in-
stalled to obtain samples for solute analyses.
The extraction system was composed of vacuum
pumps connected by an underground line to
sampling bottles and soil-water extraction
tubes. One problem that occurred was the evap-
oration of water from the collection bottles. This
was eliminated by adding mineral oil to the
bottles.
Chemical analyses on the extracts included
nitrite, nitrate, chloride, ammonium, ortho-
phosphate, sulfate, calcium, sodium, potas-
sium, magnesium and conductivity. Proce-
dures used in the analyses for the ions using an
Auto-Analyzer have been previously described
(8).
RESULTS AND DISCUSSION
Water Use Efficiency
Water use and yield of sweet corn in the
study are shown in Table 1. If only the irrigation
water data are considered, it would appear that
a breakthrough in the amount of water required
by sweet corn occurred. Essentially the same
yield of sweet corn was obtained when 142 mm
of irrigation water were applied with ASu,
248 mm with Su and Sp, and 351 mm with F.
However, when the significant differences in
change in soil-water content in the soil profile
are considered, there was no significant
difference in the consumptive use (CU) of the
crop(F = 361 mm, Sp = 346 mm,Su = 34()mm,
ASu - 300 mm). These data thus support the
fact that irrigation systems do not change
greatly the CU of crops, previously pointed out
by Bucks et al(l) (2).
Also included in the table are the number of
irrigations applied with each system. Three
irrigations were applied with the Su system and
5 irrigations were applied with F and Sp
systems. In the plots with ASu, 25 applications
were applied. In general, the amounts applied
with the ASu system were in the range of the
amounts received in showers from rainfall.
Automation of irrigation systems does af-
ford the possibility of making more efficient use
of water stored in soil profiles. This factor
should not be discounted in those areas where
supplemental irrigation rather than full irriga-
tion is used. A zone of lower moisture content
124
-------
SYSTEM EFFECTS ON WATER USE EFFICIENCY
TABLE 1
Water applied* and yield of sweet corn irrigated by various irrigation systems comparing
the water use efficiency of irrigation systems in Knox County, Texas, 1973.
Irrigation
System**
F
SP
Su
ASu
Irrigation
Water
Applied
(mm)
351 a
248 b
248 b
142 c
Number of
Irrigations
5
5
3
25
Soi/ Water
Content Change
During the
Growing Season
(mm)
-10 a
-98 b
-92 b
-168 c
Total Water
Consumptive
Use (CU) of
the Crop
(mm)
361 a
346 a
340 a
300 a
Yield**
(ears /ha)
38,918 a
40,154 a
43,242 a
39,536 a
*Numbers not followed by the same letter are significantly different at the 5% level of probability.
**F = Furrow, Su = Subirrigation, ASu = Automated Subirrigation.
***A total of 83 mm of rainfall was received on 13 days during the growing season.
can be developed to store rainfall, yet the crop
can obtain adequate moisture and not become
stressed. Automation of F and Sp systems
deserves further investigation as a means of
increasing application efficiency. Irrigation
return flow can be reduced in those areas that do
not require a leaching requirement either by
automating systems or by using a measure of
ETp as a basis of irrigation. However, in most
irrigated areas, leaching requirements could be
provided by automated systems or included
with the ET p.
The concept of irrigating only a portion of
the root zone in supplemental irrigated areas
using current irrigation systems so as to leave a
portion for storage of rainfall to obtain more
efficient utilization of irrigation water needs to
be evaluated. Some approaches to this would be
to irrigate every other furrow with furrow irriga-
tion systems and apply smaller amounts more
frequently with sprinkler irrigation systems.
Producers are using these practices to a certain
extent but they could be further refined to
increase irrigation water efficiency.
Soil-Water Solutes
Nitrite
Analyses for nitrite from samples obtained
during the first year of the study showed that 0.1
mg/1 or less existed in the soil; therefore,
analyses were not continued.
Phosphate (Orthophosphate)
Concentrations of orthophosphate of the
extracts were generally less than 1 mg 1 rang-
ing primarily from 0 to 0.3 mg/1. Such concen-
trations of phosphate are no problem relative to
the water quality of irrigation return flows.
Phosphate analyses were, therefore, discon-
tinued after the first year of the study.
Chloride
Changes in chloride concentrations in soil-
water extracts over a two-year period (1971-
1973) are given in Table 2. There was a general
increase in chloride concentration in the Sp
system between 1971 and 1972 and between 1.2
and 3.0 m to concentrations greater than those
of the irrigation water. Between 1971 and 1972
growing seasons, 329 mm of rainfall were
received in 25 different rainfall periods. This
rainfall was apparently adequate to move
chloride down in the profile but not out of the
profile. In 1973, the chloride concentration was
much lower than in the second year of the study.
This was due to rainfall received between the
1972 and 1973 growing seasons. The 707 mm
received in 40 rainfall periods was adequate to
decrease the chloride content below that of the
irrigation water throughout the profile.
The same trend was followed in the F
system between the surface and 3.0 m in that
the chloride content of the extracts decreased
between 1971 and 1972-73 to a depth of 3.0 m.
Data below 3.0 m were sketchy due to the
previously mentioned problems of obtaining
soil-water extracts. There was a consistent
increase at 7.5 m. However, these samples were
from within the water table and it is difficult to
ascertain the separate contributions from the
125
-------
WATER MANAGEMENT
TABLE 2
Chloride concentrations (mg/1) of porous bulb
soil-water extracts from various depths 1971-73
from different irrigation systems located on a
Miles loamy fine sand, Knox County, Texas.
(Irrigation water concentration - 73 mg/1)
Depth
(m)
0.15
0.30
0.45
0.60
0.90
1.20
1.50
1.80
2.10
2.40
2.70
3.00
4.50
6.00
7.50
9.00
Sp
19/i
107
113
90
28
15
66
76
23
10
12
4
—
—
—
131
116
Tinkler
1972 1973
10
82
46
64
104
105
107
103
27
93
—
117
49
37
134
72
60
56
17
33
10
14
5
12
17
5
51
—
43
48
33
82
Furrow
1971
26
97
90
74
88
81
79
49
—
33
40
—
—
—
33
25
1972
60
93
46
51
53
27
25
39
—
—
74
—
—
—
68
22
1973
85
87
77
43
53
14
17
28
37
48
24
—
79
—
101
56
Manual
Subirrigation
1971 197! 1973
21
75
94
31
83
77
67
62
42
—
28
59
78
—
98
69
73
47
73
10
55
42
23
76
86
70
78
81
42
95
121
86
66
15
63
3
43
10
14
28
—
10
48
12
48
43
53
71
Automated
Subirrigation
1972
20
39
46
43
46
53
61
24
—
15
_
14
5
58
61
63
1973
93
12
21
10
15
19
7
12
—
—
26
56
10
19
38
43
water table and leaching through the soil
profile.
In the Su system, there was a general
increase in the chloride concentration im-
mediately at the surface with a major decrease
at 0.6 m where the Su system had a major
influence. This decrease was apparent down to
1.5 m. At some depths (1.8 m, 2.1 m, 2.7 m,
3.0 m, 7,5 m.9.0 m), there was an increase in the
chloride content during 1972 — but a decrease in
1973. The decreases were probably due to the
large amount of rainfall (707 mm) received
between the two growing seasons.
The ASu system was similar to the Su
system. There was an increase in the surface, a
general decrease down to about 2.7 m, a slight
increase from 3.0 to 4.5 m and then a general
decrease from 6.0 to 9.0 between the years 1972
and 1973.
In summary, there was a major decrease in
the chloride content of soil-water extracts
between 1971 and 1973. Rainfall received
(707 mm) between 1972 and 1973 was apparent-
ly a major factor in improving the quality of the
extracts. The ranking relative to chloride con-
centration of the soil profile was
F > Sp > Su > ASu. The lower chloride content
of the ASu system may have been due to the fact
that less chloride was added since less irrigation
water was applied. No major accumulations of
chloride were noted under any of the systems.
Chloride concentrations were generally less
than the irrigation water at the end of the three
years. It was, therefore, concluded that, under
the conditions of this study, chloride would not
be a major pollutant of the water table.
Sulfate
The changes that occurred in sulfate con-
centrations from 1971 to 1973 in the irrigation
systems are shown in Table 3. The sulfate
concentrations of extracts from plots irrigated
with the Sp system exceeded those of the
irrigation water in 1971 at 0.15, 0.30, 0.45, 2.7
and 9.0 m. During 1972 and 1973, the concen-
trations in the surface 0.6 m fluctuated but were
generally lower than the irrigation water and
values obtained in 1971. There was an increase
in sulfate concentration between 1.5 and 2.7 m
compared to the values obtained during 1971.
Between 3.0 and 9.0 m the sulfate concen-
trations were generally lower in 1973 than in
1972. At the end of the three-year period, the soil-
water extracts exceeded the concentrations of
the irrigation water concentration only at 0.3,
2.1 and 2.4 m. In general, the quality of irriga-
tion return flows was better in 1973 than 1971 in
the Sp plots.
Sulfate concentrations of soil-water ex-
tracts from the F system show definite changes
in peak sulfate concentrations within the sur-
face 3.0 m between years. The peak concentra-
tion was at 0.3 to 0.6 m in 1971, 1.2 in 1972 and
1.8 to 2.7 m in 1973. Another peak occurred in
1972 and 1973 at 0.3 m, probably from evapora-
tion of irrigation water. The location of the peak
at 2.7 m in the F plot was in the same location as
one in the Sp plot. There was little change in the
sulfate concentrations below 3.0 m in the F
system.
Soil-water extract sulfate concentrations
for a Su system were similar to chloride in that
there was a constant decrease in the sulfate
concentration of the extracts between 1971 and
1973 at 0.6 to 0.9 m. In 1972, there was high
concentration of sulfate at 1.8 m and 2.7 to 3.0 m
in 1973. Below 3.0 m, there was little change in
sulfate concentration between 1971 and 1973. At
the end of the three-year period, the sulfate
concentration exceeded that of the irrigation
water only at 1.8 m, indicating that the quality
of the irrigation return flows relative to sulfate
was high.
With the exception of the concentration at
0.15 m, the ASu system had by far the lowest
sulfate concentrations in the profile. There were
126
-------
SYSTEM EFFECTS ON WATER USE EFFICIENCY
TABLE 3
Sulfate concentration (mg/1) of porous bulb
soil-water extracts from various depths 1971-73
from different irrigation systems located on a
Miles loamy fine sand, Knox County, Texas.
(Irrigation water concentration -129 mg/1)
Depth
(ml
0.15
0.30
0.45
0.60
0.90
1.20
1 .50
1.80
2.10
2.40
2.70
3.00
4.50
6.00
7.50
9.00
Sprinkler
197!
284
270
241
44
97
65
63
63
108
67
221
_
—
—
36
183
197:;
23
136
60
118
90
86
127
129
149
1 54
_
55
114
110
15
114
,97.
127
137
54
54
57
17
64
53
146
142
35
_
1 5
82
82
123
furrow
1971
30
169
179
176
137
112
117
87
108
278
200
—
—
—
22
32
197J
69
154
75
78
105
199
99
107
—
_
156
—
_
—
13
20
197:(
187
204
177
106
94
86
98
204
239
255
209
—
1 56
—
30
39
Manual Automated
Subirrigation Subirrigation
1971
80
116
142
103
111
61
75
75
145
—
151
84
73
—
26
41
,97,
127
83
123
13
61
114
98
220
132
109
83
1 23
119
45
13
15
197:1
146
51
151
20
34
60
36
27
—
161
128
156
71
114
30
23
1971:
35
50
87
46
22
41
50
29
—
23
—
39
S7
49
29
-7
,97:,
372
20
30
20
23
51
57
18
—
—
48
45
64
67
51
36
decreases in sulfate concentration at 0.3 to
0.6 m, slight increases at 1.2 and 1.5 m, with
little change in concentration occurring below
3.0m.
In summary, the overall ranking of the
sulfate concentration of the soil-water extracts
from the profile of the various irrigation
systems was F > Sp = Su > ASu. In most cases
the sulfate concentrations of the extracts were
lower than that of the irrigation water in-
dicating the quality of the leachate reaching the
water table was high from all systems. It was
therefore concluded from this study that sulfate
was not a pollution hazard.
Sodium-
There was a general increase in sodium
concentration between 1971 and 1973 between 0
and 3.0 m in the Sp system (Table 4) except at
1.2 and 1.5 m. Below 3.0 m, there was a decrease
in the sodium concentration of the soil-water
extracts between 1971 and 1973. Only at 0.15
and 0.60 m did the sodium concentration of the
soil-water extracts exceed the concentration of
the irrigation water in 1973.
The concentration of sodium in the F
system was generally higher than the concen-
tration of the Sp system at the end of the three-
year period. The sodium concentration of the
extracts from 0.15 to 0.45 m increased between
1971 and 1973 apparently due to evaporation. A
peak located at 0.9 m in 1971 apparently moved
to 1.2 m in 1972 and 1.5 to 1.8 m in 1973. No
significant change occurred in the sodium con-
centration below 3.0 m during the three years of
the study.
Decreases and increases in the Su system
were noted in sodium concentration during the
TABLE 4
Sodium concentrations (mg/1) of porous bulb
soil-water extracts from various depths 1971-73
from different irrigation systems located on a
Miles loamy fine sand, Knox County, Texas.
(Irrigation water concentration - 92 mg/1)
Dfpth
tmi
Sprinkler
UtTI 1^17-J 197.1
Furron
Manual
Subirrigalion
Automated
Subirrigation
three-year period. Decreases were noted
between 0.3 and 0.6 m and increases were noted
at 0.15 and 0.90 m indicating that the sodium
moved above and below the zone of influence of
the Su pipe. Similar decreases were noted in
chloride (Table 2) and sulfate (Table 3) concen-
trations. Below 3.0 m there was a decrease in the
sodium concentration of the soil-water extracts
between 1971 and 1973.
A zone of low sodium concentration was
also noted in the ASu system at 0.3 to 0.6 m in
the zone of influence of the subirrigation pipe
with higher concentrations located above
(0.15 m) and below (0.90 m). Below 3.0 m there
was little change in sodium concentration.
In summary, concentrations of sodium in
the soil-water extracts from 0.15 m from all
irrigation systems were all higher than those of
the irrigation water in 1973. However, there
were no deleterious accumulations of sodium in
the soil profile during the three years of the
study. At the end of the three-year period, the
sodium concentration of extracts from below
3.0 m from Sp, Su, and ASu systems was less
than that of the irrigation water. Only at 7.5 and
9.0 m was the concentration of sodium from the
127
-------
WATER MANAGEMENT
F system higher than the irrigation water, and
these concentrations did not change during the
three years for which soil-water extracts were
obtained. Above 3.0 m the sodium concentra-
tion from the different systems was greatest in
the Su system followed by the F, Sp, and ASu
systems. The Su system had low concentrations
of sodium between 0.3 and 0.6 m where the
system exerted a major influence.
Calcium
Calcium data obtained are shown in Table
5. In the Sp system, increases in the calcium
concentration were noted between the surface
and 2.7 m between 1971 and 1973. No major
changes in calcium concentration were noted
below 3.0 m.
TABLE 5
Calcium concentrations (mg/1) of porous bulb
soil-water extracts from various depths 1971-73
from different irrigation systems located on a
Miles loamy fine sand, Knox County, Texas.
(Irrigation water concentration - 63 mg/1)
Ik-pth
tint
0.15
0.45
O.fiO
0.90
1.20
1.50
1.80
2.10
2.40
2.70
3.00
4. .VI
6.00
7.50
9.W1
Sprinkler
1971
31
45
86
37
27
37
41
50
26
29
60
—
24
22
,97,
30
90
43
50
77
56
66
59
71
93
—
71
87
77
82
28
,H7:,
79
59
50
190
55
48
31
32
59
70
118
_
93
85
57
31
Furrutt
1971
40
43
78
15
15
26
50
35
—
50
43
—
_
_
14
12
,97,
66
66
47
52
49
34
72
84
—
_
115
_
_
_
34
22
197,
93
59
59
38
64
31
55
117
101
101
70
—
85
—
40
'-'"
Manual Automated
Subirrigation Subirrigatwn
1971
35
74
71
63
77
49
56
59
71
—
79
82
68
—
28
23
80 40
53 53
75 67
38 27
51 54
34 43
53 38
60 76
99 -
176 62
99 77
86 57
89 67
67 77
63 64
31 :»
,«
26
39
61
56
25
30
72
51
—
44
—
66
77
76
93
44
197.,
186
31
36
38
81
40
3S
32
—
—
27
45
52
50
55
45
The primary changes in the F system
occurred between 1971 and the last two years of
the study. Increases in the calcium concentra-
tion of the extracts from 1971 to 1972 between
0.6 and 1.2 m remained stable through 1973.
The rains received between 1972 and 1973
apparently did not affect the calcium concen-
trations as they did the chloride concentrations
(Table 2).
Decreases in the calcium concentration in
the 0.15- to 0.9-m zone in the Su system were not
as great as those of other ions (Tables 2,3 and 4).
A major increase in calcium concentration to
170 mg/1 at 2.4 m was noted in 1972. However,
in 1973 the calcium concentration of the ex-
tracts was approximately equal to that of the
irrigation water (63 mg/1) throughout the soil
profile.
The ASu system had a low calcium concen-
tration in the zone of influence of the subirriga-
tion pipe (0.30 to 0.61 m). Calcium concen-
trations increased at 0.15 and 0.90 m in the
periphery of the zone of influence of the Su
system. Calcium concentrations of extracts
below 1.5m were generally lower in 1973 than in
1972.
In summary, the overall quality of the
irrigation return flow with respect to calcium at
the end of the three-year period was high.
Excluding the surface, the order of calcium
concentrations at the end of the three-year
period with respect to irrigation systems was
R > Sp > Su > ASu. The Sp system had the
highest concentrations above 1.5 m while the F
system had the highest concentration below
1.5 m. At the end of the three-year period, the
calcium concentration of the extracts was ap-
proximately equal to that of irrigation water in
the Sp, F, Su plots and less than irrigation water
in the ASu plots. Since less water was applied
through the ASu systems, it appears that the
amount of water applied as well as the irrigation
system may influence the ion concentration in
the profiles below irrigation systems.
Magnesium
Data on magnesium concentration of soil-
water extracts from the various irrigation
systems are shown in Table 6. These data were
obtained only during the 1972 and 1973 crop
years. In the Sp system, there was some increase
in concentration in the surface at the 0.3 and 0.6
depths. Otherwise, there was a general decrease
in the magnesium concentration between the
two years. Concentrations of the extracts ex-
ceeded those of the irrigation water (42 mg/1)
only below 2.7 m.
Concentrations of magnesium in the sur-
face 1.5 m were generally lower in the F system
than the Sp system. Increases in magnesium
occurred between the surface and 1.2 m in the F
system. There was a general overall increase
between 1972 and 1973, but the final concen-
trations in the profile were not as great as they
were in the Sp system.
In the Su system, the pattern of the mag-
nesium concentrations were similar to those
obtained for other ions (Tables 2, 3, 4 and 5) in
that there was a decrease at 0.6 m in the major
128
-------
SYSTEM EFFECTS ON WATER USE EFFICIENCY
zone of influence of the Su system. With the
exception of the 0.15-m depth, the magnesium
concentrations generally decreased throughout
the profile in the Su and ASu systems.
In summary, the Sp plots had the highest
magnesium concentrations followed by the F
plots with the Su and ASu plots being the
lowest. In general, the magnesium concen-
trations of the soil extracts from the root zone
tended to be lower than the concentrations of
the irrigation water while the calcium concen-
trations tended to be higher than those of the
irrigation water. This suggests that a portion of
the magnesium applied in the irrigation water
was absorbed by the clays or precipitated to a
less soluble form.
Potassium
Potassium concentrations obtained during
the course of the study are shown in Table 7. A
decrease in potassium concentrations above
0.9 m between 1971 and 1973 was noted in the
Sp system. There is some indication of a tenden-
cy toward slight accumulations in the lower
TABLE 6
Magnesium concentrations (mg/1) of porous bulb
soil-water extracts from various depths 1971-73
from different irrigation systems located on a
Miles loamy fine sand, Knox County, Texas.
(Irrigation water concentration - 42 mg/1)
Depth Sprinkler
(m)
0.15
0.30
0.45
0.60
0.90
1.20
1.50
1.80
2.10
2.40
2.70
3.00
4.50
6.00
7.50
9.00
1972
25
29
17
14
35
45
27
35
46
48
59
43
46
40
13
1973
39
43
15
41
13
29
7
7
27
25
58
—
27
48
32
10
Manual
Furrow Subirrigation
1972
24
18
13
14
11
10
35
44
—
—
43
—
—
—
11
8
7973
30
31
26
15
18
10
13
52
42
49
36
—
52
—
13
8
1972
19
18
25
12
15
12
11
22
36
27
29
30
45
48
34
19
1973
23
17
28
7
17
9
29
58
—
6
25
20
35
37
31
17
Automated
Subirrigation
1972
11
18
19
20
17
20
25
34
—
31
—
44
52
47
33
26
1973
133
8
9
8
18
12
9
19
—
—
18
32
42
40
29
22
part of the profile during 1972 between 1.5 and
1.8 m. However, these concentrations were not
noted in 1973. The concentrations were higher
at 0.6 m in 1973 than they were in 1971 in-
dicating the possibility of leaching. However,
below 18 m the concentrations were slightly less
or approximately equal to those of the irrigation
water following the first year of the study. No
massive amounts of leaching were occurring
compared to the amounts produced in the sur-
face, indicating that the excess potassium may
have reacted with the clay.
TABLE 7
Potassium concentrations (mg/1) of porous bulb
soil-water extracts from various depths 1971-73
from different irrigation systems located on a
Miles loamy fine sand, Knox County, Texas.
(Irrigation water concentration - 2.8 mg/1
Depth
0.15
0.30
0.45
0.60
0.90
1.20
1.50
1.80
2.10
2.40
2.70
3.00
4.50
6.00
7.50
9.00
Sprinkler
,,7,
29.2
,,.-,
IH7.I
12.021.0
32.9 23.7
29.8
9.7
1.6
1.1
2.2
1.6
1.6
1.1
2.2
6.5
8.1
18.5
6.3
16.5
9.2 22.7
5.1
4.2
3.0
6.0
6.0
3.0
—
0.0
2.0
1.0
1.0
3.0
5.2
5.5
2.1
2.1
2.1
1.1
0.5
1.1
1.1
1.1
2.1
Furrou'
1H71
11.4
6.4
1.6
2.1
2.1
1,1
0.0
0.0
_
1.5
0.0
_
_
_
0.0
0.0
,..,-,> 1
.0
3.0
2.0
1.0
1.0
1.0
4.9
3.3
—
_
2.9
_
—
_
1.0
1.0
••,7,-i
10.4
6.3
3.2
2.1
2.1
1.1
1.1
0.5
5.5
5.5
2.1
—
1.1
—
1.1
2.1
Manual Automated
Subirrigation Subirrigation
I*.',
28.2
28.5
25.4
6.1
6.6
3.0
3.1
3.1
3.0
—
1.5
1.0
2.0
—
2.5
5.1
,,r,
15.5
21.3
13.5
9.7
4.0
2.1
I.I
2.1
1.1
3.3
1.1
1.1
1.1
2.1
2.1
3.0
,,r.
15.2
16.3
11.4
4.2
9.5
9.2
10.6
10.9
—
1.1
1.1
2.1
2.1
2.1
3.2
3.2
,^
13.3
15.0
7.2
9.2
6.2
2.1
3.1
10.6
—
3.3
—
5.5
1.1
2.1
1.1
2.1
«<:,
23.7
12.4
7.3
6.3
12.5
4.2
3.2
3.2
—
—
2.1
4.2
1.1
2.1
1.1
2.1
There was little change in potassium con-
centration in extracts from the F system
between 1971 and 1973. This plot was land
leveled; consequently, part of the micaceous
minerals which were high in potassium were
probably removed from the surface and the
potassium in solution in the resulting top soil is
not as great as the top soil which was removed.
In general, the potassium concentrations below
1.5 m were approximately equal to that of the
irrigation water (2.8 mg/1).
The Su system yielded data very similar to
the Sp system in that there was a decrease in the
potassium concentration at the surface between
1971 and 1973. There was a slight increase at 1.5
to 1.8 m indicating some movement out of the
zone around the subirrigation pipe to this par-
ticular area. The decreases in the surface were
greater than that of the Sp plots indicating that
the surface may have been drier with less
moisture available to make the potassium
available.
The same trend holds for the ASu system in
that the surface remained higher with a zone of
129
-------
WATER MANAGEMENT
low concentration at 0.6 m. There was some
indication that there was an increase in potas-
sium concentrations at 0.9 to 1.8 m, indicating
again some movement out of the zone where the
subirrigation pipe was located.
In summary, the greatest potassium con-
centrations were in the plots with a Sp system
followed by the Su, ASu, and F systems. Soil
minerals are indicated to play a major role in the
production of potassium in that the land leveled
F system had very low potassium concen-
trations in the soil solution.
Ammonium
Typical ammonium-N concentrations
found during the course of the study are shown
in Table 8. Significant amounts of ammonium
were found periodically in extracts from all
systems, i.e., at 9.0 m in 1971 in the Sp system,
2.7 m in 1972 in the F system, 9.0 m in 1971 in
the Su system and 3.0 m in 1972 in the ASu
system. Overall, the ranking relative to oc-
currence of ammonium was ASu > Su >
Sp > F. While there was some evidence of
ammonium-N in the profile below the root zone,
the inconsistency with which it was found in
concentrations above 3.0 mg/1 makes assess-
ment of movement impossible. Generally,
ammonium-N concentrations were uniformly
low and would appear to be of little consequence
from the standpoint of irrigation return flow
degradation.
Conductivity
Between 1971 and 1973, the conductivity of
soil-water extracts from 0.6 and 0.9 m (Table 9)
increased in the Sp system. Otherwise, there
was a general decrease in the extracts. There
was little discernible change in conductivity of
soil-water extracts of the F system. In both the
Su and ASu systems an increase in conductivity
of extracts from 0.15 m and a decrease from 0.3
to 0.9 m occurred in the zone of influence of the
Su system. Otherwise the conductivity of the
extracts from the remainder of the profile
remained stable or decreased.
In summary, the conductivity was highest
in the Sp system followed by F, Su, and ASu
systems. Most values were lower than the
average irrigation water conductivity of 947
^mhos. This is not surprising since the area is a
supplemental irrigated area and rainfall is a
major factor influencing the conductivity of
soil-water extracts. The conductivity data thus
indicate that the salt load of the irrigation
return flow in this study was not a problem.
TABLE 8
Ammonium concentrations (mg/1) of porous bulb
soil-water extracts from various depths 1971-73
from different irrigation systems located on a
Miles loamy fine sand, Knox County, Texas.
(Irrigation water concentration < lmg/1)
Depth
(m)
0.15
0.30
0.45
0.60
0.90
1.20
1.50
1.80
2.10
2.40
2.70
3.00
4.50
6.00
7.50
9.00
Sprinkler
W71
2.1
0.7
0.2
0.1
1.7
0.4
0.2
0.2
1.3
0.5
0.7
—
_
—
6.0
10.5
1972
0.3
0.3
0.4
0.1
4.9
1.2
2.0
3.9
1.2
1.4
—
2.4
2.5
1.6
0.3
5.1
1973
1.1
0.5
0.5
0.5
1.5
1.1
0.2
0.2
0.2
0.2
9.7
—
0.2
0.0
0.2
0.5
Manual
Furrow
197!
0.3
0.3
0.6
0.3
0.3
0.2
1.5
1.2
—
3.0
3.0
—
—
—
0.8
1.7
1972
0.4
0.5
0.6
0.8
0.4
0.7
2.0
1.1
—
—
5.7
—
—
_
0.6
2.8
lH7:i
0.2
0.0
0.2
0.5
0.2
0.2
0.7
2.3
3.6
2.4
0.5
—
0.2
—
0.2
0.2
Automated
Subirrigation Subirrigation
1S71
0.2
1.5
1.6
0.6
1.2
1.8
1.2
1.6
3.3
—
0.2
0.4
2.6
—
3.0
10.0
;»/2
4.9
0.8
0.3
1.0
2.5
1.9
2.5
1.1
1.2
7.2
1.5
2.3
4.6
1.5
0.5
5.7
1973
0.5
0.5
0.2
0.5
0.7
1.1
3.6
2.3
—
0.0
0.5
0.5
0.2
0.2
0.2
0.5
1971
10.2
4.1
0.7
1.5
5.5
1.1
1.2
19.5
—
8.1
—
32.5
1.2
2.5
0.4
7.4
197:1
1.7
1.7
0.7
2.0
2.4
3.4
1.2
1.1
—
_
0.5
3.4
1.0
1.0
0.5
1.0
CONCLUSIONS
1. The ASu systems are superior to Su and
Sp systems which are superior to F systems in
the amount of irrigation water required to
produce corn even if the manual systems
applications are scheduled based on ET p.
2. No significant differences in total water
requirement of corn were noted between the
various irrigation systems. The soil water was
used more efficiently when less irrigation water
was applied.
3. Supplemental irrigated areas can in-
crease irrigation water-use efficiency
significantly by utilizing systems so that a
portion of the root zone remains dry for the
storage of rainfall.
4. The data indicate that the quality of
irrigation return flows of Su systems will be
superior to F and Sp systems because the
concentrations of all solutes and the electrical
conductivity (EC) were lower.
5. A zone low in all solutes is formed in the
path of water flow around the subirrigation
pipe.
6. Because salts are moved upward from
subirrigation emitters, periodic supplemental
leaching with another system may be needed in
arid areas.
130
-------
SYSTEM EFFECTS ON WATER USE EFFICIENCY
TABLE 9
Conductivity (^mhos) of porous bulb soil-water extracts from various depths 1971-73 from different irrigation
systems located on a Miles loamy fine sand, Knox County, Texas. (Conductivity of irrigation water - 947 jjmhos)
Depth
(m)
0.15
0.30
0.45
0.60
0.90
1.20
1.50
1.80
2.10
2.40
2.70
3.00
4.50
6.00
7.50
9.00
Sprinkler
1971
1335
1124
1069
700
630
860
904
740
660
—
555
—
—
—
1085
1595
1972
260
860
850
930
920
940
960
395
340
845
—
300
800
879
980
1180
1973
550
922
850
1575
825
175
500
600
255
760
290
—
670
795
710
—
1971
530
820
1100
940
970
840
960
770
—
975
865
—
—
—
580
715
Furrow
1972
760
890
720
640
670
760
320
340
—
—
270
—
—
—
630
670
Manual
Subirrigation
1973
990
980
895
660
850
560
840
130
220
240
705
—
750
—
650
670
1971
660
710
835
550
700
810
710
640
810
—
530
555
645
—
672
815
1972
800
600
840
360
660
650
550
950
970
350
700
670
670
760
700
630
1973
935
345
800
305
325
410
190
135
—
670
710
660
680
760
690
700
Automated
Subirrigation
1972 1973
450
560
590
650
580
860
740
1185
—
310
—
200
560
620
580
570
1225
330
360
510
555
420
550
630
—
—
730
—
550
570
560
560
REFERENCES
1. Bucks, D. A., Erie, L. J. and French, O.
F. 1973. Limiting Quantities and varying fre-
quencies of trickle irrigation on cotton. Prog.
Agric. in Ariz. XXV (4) : 13-16.
2. Bucks, A. D., Erie, L. J. and French, O.
F. 1974. Quantity and frequency of trickle and
furrow irrigation for efficient cabbage produc-
tion. Agron. J. 66(l):53-57.
3. Hillel, D. Soil and Water. 1971.
Academic Press, Inc. Ill Fifth Avenue, New
York, New York 10003. p. 227.
4. Jensen, M. E., Robb, C. W., and Fran-
zoy, C. E. 1969. Scheduling irrigation using
climate-crop-soil data. Jour, of the Irri. and
Drainage Div., ASCE 96:25-38.
5. Onken, A. B., Wendt, C. W., Hargrove,
R. S. and Wilke, O. C. Relative movement of
bromide and nitrate in soils under three irriga-
tion systems. (SSSA Journal, Jan, Feb 1977)
6. Ritchie, J. T., and Burnett, E. 1971. Dry-
land evaporative flux in a subhumid climate: II
plant influence. Agron. J. 63:56-62.
7. Wendt, C. W., Harbert III, H. P., Bausch,
W. and Wilke, O. C. 1973. Automation of drip
irrigation systems. Paper No. 73-2505.
Presented to the 1973 Winter Meeting of Amer.
Soc. Agri. Eng.
8. Wendt, C.W., Onken, A.B. and Wilke,
O. C. Effects of irrigation methods on ground
water pollution by nitrates and other solutes.
Final Report on EPA Grant No. S-802806
(Formerly 13030EZM) (in press).
131
-------
Scientific Irrigation Scheduling
for Salinity Control
of Irrigation Return Flow
MARVIN E. JENSEN
USDA,
Snake River Conservation Research Center,
Kimberly, Idaho
ABSTRACT
Basic principles of irrigation water man-
agement and irrigation scheduling are pre-
sented. Commercial and agency groups ex-
panded rapidly in the 1970's providing field-by-
field scheduling services to over 600,000 acres in
1976.
The leaching fraction used on projects can
effect return flow quality. Most leaching frac-
tion/return flow models hypothetically assume
uniform water applications of exact quantities
to attain targeted leaching fractions. The
average or effective leaching fraction for a field
is dependent on the irrigation uniformity coef-
ficient. The effects ofnonuniform water applica-
tion on average leaching fractions will be
presented, along with the probable effects of
expected improvements in irrigation efficiency
on return flow. Also, estimates of the accuracies
in estimating evapotranspiration and measur-
ing water will be presented. Substantial im-
provements can be made in irrigation efficien-
cies before minimum leaching fractions are
reached on most western irrigated projects.
INTRODUCTION
The annual return flow volume from an
irrigation project is dependent on the annual
irrigation water volume diverted to the project,
precipitation, and the project evapotranspira-
tion (ET). The average project return flow
quality is influenced by precipitation, the quali-
ty of the irrigation water, the proportion of
irrigation water in the return flow, the in-
tegrated project leaching fraction (LF), and salt
pick-up. The salt pick-up is influenced by the
volume of project deep percolation and seepage
from canals and laterals.
The title of this paper implies that the
variables affecting the quantity and quality of
return flow can be controlled by irrigation
scheduling, but the potential degree of control
has not been delineated. The purpose of a
detailed study conducted in 1975 (Jensen, 1975)
was to evaluate the probable effects of im-
plementing scientific irrigation scheduling on
return flow quality. The results of that study
and another presented at the California Con-
ference on Salt and Salinity Management
(Jensen, 1976) are summarized in this paper.
What is Irrigation Scheduling?
Irrigation scheduling is predicting the time
and amount of the next irrigation. This process
is dependent on the precipitation and ET since
the last irrigation, the allowable soil water
depletion, and the expected precipitation. Irri-
gation scheduling can significantly influence
the volume of water diverted to a project. Thus,
scheduling can potentially influence the LF,
and to a limited extent canal seepage. Irrigation
scheduling will have little effect on ET when
crops are irrigated for maximum, or for optimal
yields.
Reducing the salt load from an irrigation
project requires a minimum leaching fraction
(LF*) permissible for the crops and water quali-
ty involved. The average LF will be greatly
dependent on the attainable irrigation uniform-
ity. Thus, even though irrigation scheduling
technology can be refined and implemented,
attaining low average LF's will require both
irrigation scheduling and significant im-
provements in irrigation systems to uniformly
apply water.
133
-------
WATER MANAGEMENT
Need for Irrigation Scheduling
If the management objective is to minimize
the salt load in return flows, then minimum
leaching fractions must be achieved on each
irrigated field. It is difficult to manage the soil
water reservoir of complex soil-crop-climate
systems because many variables are involved,
and the soil water status is not readily apparent.
The quantity of water applied at each irrigation
is important because a single overirrigation
during a growing season can drastically
decrease the seasonal irrigation efficiency and
prevent achieving the targeted field LF.
Accurate water control is easier to achieve
with irrigation systems that have limited oppor-
tunities for overirrigation. With sprinklers, for
example, the amount of water applied is con-
trolled by the system and not the soil when the
application rate is less than the intake capacity.
Thus, a targeted application can be achieved by
regulating pressure and hours per set, especially
with systems that apply water while moving.
Moving systems usually apply water more un-
iformly than those with stationary heads.
Specific amounts of water can be applied
with surface irrigation systems if modern
technology is used, but most surface systems are
operated today as they have been for the last
two or three decades. Stream sizes normally are
not increased and set times reduced enough to
achieve the targeted uniform irrigation when
intake rates change after tillage. Facilities do
not permit these detailed changes at each irriga-
tion, especially with older systems that do not
have water measuring devices, adjustable water
control structures, and lined channels or en-
closed distribution systems. The duration of
each irrigation set is often based on some
convenient period such as 12 or 24 hours
because of the labor required to change sets.
When coupled with long length of runs, light
irrigations are difficult to achieve on most
existing surface systems. The time of irrigation
can easily be changed with these systems, but
controlled water applications will require im-
proved facilities. There is an emerging demand
for new innovations, instrumentation, and
equipment to modernize surface irrigation
systems to meet the 1983 goals of Public Law 92-
500 for best available technology to control the
quantity and quality of irrigation return flow.
Typically, observed irrigation efficiencies
on surface irrigated lands in the U.S. are low.
Early studies in the 1890's cited overirrigation
as the first and most serious mistake made by
early settlers in Wyoming (Buffum, 1892). This
situation has not changed much on many older
projects. Until recently, only limited progress
has been made in modernizing surface irriga-
tion systems because there have been few incen-
tives for change because water costs in most
areas are low and represent a relatively small
percentage of annual farming cost. Also, ex-
cessive water application effects on both crop
yield and quality normally are not as apparent
as the effects of water deficits and salinity
caused by inadequate irrigation.
Most older surface irrigation systems re-
quire much labor to be operated efficiently.
Increasing labor costs generally have reduced
labor input, thus offsetting the effects of im-
provements in irrigation facilities. Similarly,
with low cost nitrogen (N) fertilizer, it has been
easier and more economical to compensate for
poor water management by applying excess N.
New regulations on N in return flows and
increasing N costs are now beginning to in-
fluence irrigation and fertilizer practices. The
costs of correcting drainage problems that
emerge after several years of excessive water
use are often distributed uniformly to all water
users in a project and not just to those using
excessive water. Sometimes drainage costs are
cost-shared, which in essence subsidizes ex-
cessive water use. All of these practices have not
been conducive to improving irrigation prac-
tices and systems.
Today the ecomonics of irrigation are rapid-
ly changing, and new constraints are emerging.
Fertilizer costs are increasing. More important,
energy costs are spiraling and the certainty of
continuing energy supplies for irrigation is
diminishing. These and other emerging
economic incentives will have a major impact
on irrigation water management practices dur-
ing the next decade. Pending state and federal
return flow regulations, involving both water
quantity and quality, coupled with increasing
labor and energy costs will be changing irriga-
tion farming objectives. New management
technology will be needed for irrigation farming
to remain solvent and competive with rainfed
agriculture. Increased capital investments will
be needed to achieve better water control with
less labor, and increased technical skills will be
needed to service complex irrigation systems.
Farm management must place greater
emphasis on maximizing net returns and yield
per unit of water and less emphasis on maximiz-
ing yield per unit area.
134
-------
SCHEDULING FOR SALINITY CONTROL
As the amount of irrigation water diverted
to irrigation projects approaches the consump-
tive irrigation water requirement, more ac-
curate consumptive use or ET data will be
needed to optimize system operations. The de-
mand for irrigation management services will
increase which should stimulate the develop-
ment of more rapid and ecomonical means for
monitoring the soil water content and its dis-
tribution on individual fields.
Attainable Irrigation Efficiency and
Leaching Fractions
Optimum irrigation water management
will require more accurate water applications
for both consumptive and nonconsumptive
uses. Essential or minimum amounts needed for
nonconsumptive uses (frost protection, leaching
hydrating a root crop before harvest and seed
germination) are easy to specify, but may be
difficult to achieve because of nonuniform water
distribution inherent with many existing
systems. For example, if target minimum LF's
of 0.05 to 0.1 were acceptable and the manage-
ment objective was to maintain these LF's on
the 10% of each field that regularly receives the
least amount of water, the average LF for the
field may be three to five times the mimimum LF
when using existing sprinkler systems (Table 1,
Jensen, 1975). The areas normally receiving the
least amount of water are the lower ends of
uniformly graded surface irrigated fields and
the area between sprinkler laterals that are not
moved during the irrigation season.
The water distribution uniformity within
individual fields is an important variable that
must be considered in estimating the potential
effects of minimum LF's on the quality of return
flow. This is a very important variable that
cannot be neglected because the relationship
between the LF and the precipitation of salts
within the root zone is nonlinear. The prospects
of achieving cummulative seasonal uniformity
coefficients (U c) that exceed 90% by 1983 is very
remote for most projects even though the cum-
mulative Uc increases with successive
irrigations.
Another important variable that must be
considered, except with automatic systems
operated by sensors, is the probable accuracy of
applying targeted amounts of water. When
using values suggested by Jensen and Wright
(1976), the coefficient of variation (estimated
standard deviation/drainage, expressed in per-
cent) during a 30-day period is about 10 percent
with a LF of 0.5 and an average ET of 6 mm/day
(0.24 in./day). It increases to about 50 percent as
the LF approaches 0.1 (Table 2). The probable
error is even greater with shorter time periods
because the standard error in estimating ET
decreases proportional to 1 /(/T.
TABLE 1
Average leaching fraction (LF) for a field for various
targeted minimum LF (LF*) in relation to the
uniformity coefficient (Uc) of water applications
(From Jensen, 1975).
Average LF with
a LF* of:
Uc s ad 0.05 0.10 0.15
%
100
95
90
85
%
0
6.25
12.50
18.75
1.00
.89
.79
.69
0.05
.15
.25
.34
0.10
.20
.29
.38
0.15
.24
.33
.41
Assumptions:
Average depth of water applied in the 10% of a
field regularly receiving the lease amount of water is
(1 + LF*) ET; the application of water by a sprink-
ler system is normally distributed with a standard
deviation, s, estimated from the equation
Uc - [100(1 - 0.8s/100], and is independent of the
amount applied; the distribution coefficient, a a, at
5% of the area represents the relative depth of water
applied to 10% of the area that receives the least
amount of water; irrigations are timed exactly so that
only LF*(ET) drains through the soil; and ETis not
affected by soil salinity level.
TABLE 2
Estimated coefficient of variability in applying
water to achieve various targeted leaching fractions
for 10-, 20-, and 30-day periods (From Jensen and
Wright, 1976). _
Period, days
Target leaching fraction
20
30
0.1
.2
.3
.4
.5
94
42
25
17
12
%,
58
27
16
11
8
48
22
14
10
7
Assumptions:
Mean ET = 6 mm/day (0.24 in./day) estimated
from daily climatic data.
Surface runoff (15%) and applied water measured
with an accuracy of ± 5%, or s Q = 0.025Q.
135
-------
WATER MANAGEMENT
Traditional and Modern Irrigation
Scheduling Practices
Traditional irrigation scheduling methods
are based on tensiometers, electrical resistance
units, pan evaporation data, and general fixed
irrigation dates and amounts for given crops
within a local area. A few farmers now use the
neutron probe for measuring soil moisture and
scheduling irrigation. Traditional approaches
usually require the farmer to use some type of
instrument, take soil samples, or use evapora-
tive data. Thus, he must first understand ET
and soil moisture depletion. If a tool or instru-
ment is needed he also must understand how it
functions and its relationship to the soil water
depletion to use it correctly.
Traditionally irrigation scheduling meth-
ods have not been very effective in the past,
perhaps because there have been insufficient
incentives to warrant significant improvements
in irrigation management practices. Basically,
traditional methods essentially require a "do-it-
yourself" approach to irrigation scheduling.
Thus, promoting only traditional methods has
limited the farmers access to information need-
ed to improve irrigation scheduling decisions.
Alternative procedures for providing this infor-
mation on a real time basis have not been
seriously considered, developed and evaluated.
Modern irrigation scheduling services
provide farm managers with estimates of the
current soil water status on each field, and
predicted irrigation dates and amounts to be
applied on each field to avoid adverse effects on
plant growth. With this information farmers
can modify their irrigation practices and
schedule irrigations more accurately. The in-
creasing demand for commercial irrigation
scheduling services during the past five years is
indicative of a long standing need for such
information.
A modern irrigation scheduling service
(ISS), utilizes the latest irrigation science and
technology to provide current information on
the available water status in individual fields
and projected irrigation dates based on expected
climatic conditions. The ISS may provide the
daily rate that high frequency systems should
apply water to maintain the desired soil water
level in each field. When water supplies are
limited or another variable, like fertilizer, limits
production, the ISS should also recommend the
optimum times and irrigation amounts to
achieve these goals. An effective ISS recom-
mends needed improvements in irrigation
systems to achieve greater irrigation uniformi-
ty, reduce water losses, and maintain a
favorable salt balance in the soil. Services like
these increase the farmer's managerial skills
and should increase his net returns (Jensen,
1975).
The irrigation scheduling approach
developed by Jensen et al. (1969,1971) known as
the USDA-ARS Computer Program, has been
widely accepted because it does not require
farmers to obtain technical knowledge and
training to apply modern irrigation technology.
Periodic updating of the current soil water
status, ET rates and projected irrigation dates
provide information that greatly increases the
farmers' understanding of the soil-plant-
atmosphere system. Experience gained in
testing this concept in 1968 and 1969 indicated
that even though farmers had been irrigating
for many years, information provided by this
program increased their understanding of
processes influencing and controlling this com-
plex system.
Many commercial and agency service
groups have adopted the program without
change or have modified the USDA-ARS Com-
puter Program to suit their special needs. New
commercial firms have been established, and
many firms have purchased their own small
computers for routine calculations and record
keeping on hundreds of fields. The professional
staff of an ISS group must be trained in
irrigation science and technology for successful
application of the program. They must monitor
the soil water status and sometimes measure
precipitation on each farm to periodically tune
the computer results to field conditions. Most
companies recommend and provide updated
irrigation dates at least weekly. Some fields are
inspected twice weekly during the growing
season. Some service groups also design im-
proved irrigation systems and provide plant
nutrient, pest, control and other services. Ser-
vice companies must maintain active commu-
nications with the farmer, sometimes on a 24
hour basis, and they must have a crew of trained
and experienced field technicians. If ISS groups
are to provide unbiased recommendations to
maximize net returns to the farmer, they should
not sell products they recommend. This practice
represents a serious potential conflict of in-
terest.
Table 3 is a summary of the general irriga-
tion practices and management options along
136
-------
SCHEDULING FOR SALINITY CONTROL
with the information needed and provided by
service groups to improve irrigation manage-
ment. Such information is needed for all irriga-
tion methods and management options, except,
perhaps, fully automated systems that operate
with soil-water sensors or a combination of
sensors and timed controllers. Periodic monitor-
ing of saline control may still be needed with
automatic systems, unless salt sensors are used.
TABLE 3
General irrigation practices and management
options and the types of scheduling information
needed (from Jensen, 1976).
Types of irrigation
scheduling information
needed
Irrigation practices and
management options
A. High frequency
1. Maintain nearly con-
stant soil water level
and target leaching
fraction.
2. Planned gradual de-
pletion of available
soil water during the
crop season with tar-
get leaching fraction
provided during the
noncrop or some oth-
er crop season.
3. Combination of A.1
and A.2
B. Normal periodic
irrigations
1. Irrigate to bring the
soil to field capacity
at each irrigation and
provide target leach-
ing fraction.
a. Constant or fixed
application
amounts.
b. Fixed irrigation
intervals
2. Planned gradual de-
pletion of soil water
during the crop sea-
Daily evapotranspiration
(ET) and rate of water ap-
plication. Periodic soil wa-
ter monitoring for content,
distribution, and salinity.
Same as for A.1.
Same as for A.I.
Daily ET, earliest next irri-
gation date permitting ef-
ficient irrigation along with
the latest next irrigation
date to avoid significant ad-
verse effects on crop produc-
tion. Periodic soil water
monitoring for content dis-
tribution, and salinity.
Daily ET, irrigation amounts
for efficient irrigation, and
periodic soil water monitor-
ing for content, distribution,
and salinity.
son with target leach-
ing fraction provided
during the noncrop or
some other crop sea-
son.
a. Constant or fixed
application
amounts
b. Fixed irrigation
intervals
3. Combination of B.I
and B.2
C. Limited and supplemen-
tal irrigation
1. Limited irrigations
applied to optimize
production or net re-
turns per unit volume
of water.
2. Alternating shallow,
well-watered and
deep rooted, nonirri-
gated crops.
Same as B.I.a.
Same as B.l.b.
Daily ET, earliest next irri-
gation date and amount for
efficient irrigation along
with the latest date and cor-
responding amount to avoid
significant adverse effects
on crop production and per-
mit efficient irrigation. Peri-
odic soil water monitoring
for content, distribution and
salinity.
Expected ET, and optimum
times and amounts of irriga-
gation considering expected
rainfall to maximize pro-
duction per unit of irriga-
tion water. Periodic soil wa-
ter monitoring for content
and distribution.
Expected ET and produc-
tion from alternative se-
quences.
Monitoring the soil water status in each
field may not be required during the entire
growing season if excess water is applied at
each irrigation, or if the amount of water
applied and rainfall are known with reasonable
accuracy. However, soil water monitoring is
usually required to calibrate or tune the com-
puter calculations, since the error of measuring
soil water either gravimetrically or with a
neutron probe is generally less than the error in
estimating the amount of irrigation water
applied. Since the confidence limits of predicted
irrigation dates are dominated by the compo-
nent with the greatest uncertainty, the amount
of water applied by surface irrigation systems
usually causes the greatest uncertainty and
widens the limits until the field can be
monitored again.
When irrigation service groups first begin,
they provide irrigation schedules without
recommending changes in existing systems.
Later as they become acquainted with the
137
-------
WATER MANAGEMENT
characteristics and constraints of the systems,
and as they and the farmers gain experience,
components that need improvemnt can be iden-
tified and system improvements scheduled. As
labor costs increase, or skilled labor becomes
less available, ISS groups will play a more
important role in applying modern irrigation
science and technology in irrigation water
management.
ISS also must be economical, with sufficient
accuracy to be compatible with the system's
ability to apply specific amounts of irrigation
water. Irrigation scheduling information only
supplements, and does not replace, the farmers'
experience.
Adoption of Irrigation Scheduling
Services
Jensen (1975) evaluated ISS provided in
1974 and found that ten western U.S. commer-
cial firms had 1 to 10 years of experience and
seven of the ten had five years or less. ISS was
provided for a fee to about 4450 fields involving
over 100,000 ha (250,000 ac.) of summer crops in
eight western states. All 10 firms provided plant
nutrition services, seven provided plant pest
management services, and six provided
engineering services. Technicians, who moni-
tored soil water-depletion in each field once or
twice weekly, serviced an average of 5800 acres
and traveled an average daily distance of
195 km (120 mi.). Agency or project services
were similar to commercial services except the
customers paid only part of the direct cost. The
balance was distributed uniformly to all water
users in the projects. Of 22 agency service
groups, 21 had five or less years' experience.
These groups provided ISS for about 3500 fields
involving 54,000 ha (133,000 ac.) of summer
crops in 12 western states. Only about 25% of
these groups provided plant nutrition services,
20% provided pest management services, and
15% provided engineering services. Besides
field-by-field services, the U.S. Bureau of Rec-
lamation provided weekly estimates for major
crops based on early, medium and late planting
dates for different general soil types. General
irrigation guides for major crops were provided
for about 10,000 fields involving about
94,000 ha (233,000 ac.). The Alberta Depart-
ment of Agriculture in Canada provided ISS to
about 140 fields comprising 4,500 ha(l 1,000 ac.)
on a field-by-field basis and provided guides for
about 100 fields totaling about 3,200 ha
(8,000 ac.).
In 1976, a comparison of seven commercial
groups that provided field-by-field services in
1974 and 1976 added over 1,000 fields or
34,000 ha (85,000 ac.) per year. In 1976 ET
estimates based on current climatic data were
used on 95% of the area served. Most technicians
serviced 1,400 to 1,600 ha (3,500 to 4,000 ac.).
The capacity of these seven companies in 1976
was only about 6,200 fields or about 174,000 ha
(430,000 ac.). These groups could not expand
immediately, mainly because of a lack of
trained personnel.
Fees varied widely depending on the
method of charging. Most prices ranged from $6
to $11 per ha ($2.50 to $4.50 per ac.) for irrigation
scheduling, or a flat fee of $175 to $250 per field.
Commercial and agency irrigation schedul-
ing service expanded from less than 40,000 ha
(100,000 ac.) on a field-by-field basis in 1971 to
over 243,000 ha (600,000 ac.) in 1976. These
services are expanding as rapidly as new staff
members can be trained and new companies can
be established.
The Role Of Irrigation Scheduling and
Salinity Control
This assessment of irrigation scheduling
and salinity control indicated that substantial
improvements in irrigation water management
and efficiency can be made before minimal
leaching fractions needed to maintain a
favorable salt balance in the soil are reached on
most western irrigated projects. Only about 10
percentage points improvement in average
farm irrigation efficiencies can be expected by
1985 without significantly increasing energy
requirements for irrigated agriculture. This
change is not expected to significantly influence
salinity in return flows except where salt pick-
up is a major factor. New emerging ecomonic
incentives may bring about more rapid
changes.
Continued improvement in irrigation water
management is needed, but general implemen-
tation of new scheduling technology may re-
quire one or more decades. Irrigation system
improvements to permit uniform applications of
specific amounts of water also will be required.
With scientific irrigation scheduling and
systems that apply uniform known amounts of
water, significant reduction in salt loads in
return flows can be achieved. Both components
are needed because potential efficiencies of new
irrigation systems and salt control probably
138
-------
SCHEDULING FOR SALINITY CONTROL
cannot be achieved without scientific irrigation
scheduling.
REFERENCES
1. Buffum, B. C. 1892. Irrigation and Duty
of Water. Wyo. Agr. Expt. Sta. Bull. No. 8.
2. Jensen, M. E. 1969. Scheduling irriga-
tions using computers. J. Soil and Water Con-
serv. 24:193-195.
3. Jensen, M. E., J. L. Wright, and B. J.
Pratt. 1971. Estimating soil moisture depletion
from climate, crop and soil data. Trans. Am.
Soc. Agr. Eng. 15(5):954-959.
4. Jensen, M. E. 1975. Scientific Irrigation
Scheduling for Salinity Control of Irrigation
Return Flows. Environ. Protection Tech. Series
EPA-600/2-75-964, 92 pp.
5. Jensen, M. E. 1976. On-Farm Manage-
ment: Irrigation Scheduling for Optimal Use. In
press; Proc. of Conf. on Salt and Salinity
Manage. Sept. 23-24, Santa Barbara, CA.
6. Jensen, Marvin E., and James L.
Wright. 1976. The Role of Evapotranspiration
Models in Irrigation Scheduling. Am. Soc. Agr.
Eng., Scientific Paper No. 76-2061.
139
-------
Return Flow Management
-------
Management Guidelines for
Controlling Sediments, Nutrients,
and Adsorbed Biocides in Irrigation
Return Flows
D. L. CARTER and J. A. BONDURANT
USDA, Snake River Conservation Research Center,
Kimberly, Idaho
ABSTRACT
Sediments in irrigation return flows arise
mostly from furrow erosion, and nearly all
nutrients and biocides in surface irrigation
return flows, except those applied directly to the
water, are adsorbed to the sediments. Therefore,
controlling erosion and sediment loss in these
surface return flows also controls the nutrients
and biocides. There are three general manage-
ment approaches for controlling sediments in
return flows. The first is to eliminate surface
runoff by using irrigation methods that produce
no runoff. These methods include properly
designed and operated sprinkler sys terns, basin,
trickle, and some border and level furrow
methods. The second approach is to eliminate or
reduce erosion by controlling the slope in the
direction of irrigation, the furrow stream size,
the run length, the irrigation frequency and
duration, and tillage practices. The third is to
remove sediments from surface return flows by
controlling the tailwater and utilizing sediment
retention basins. All three approaches are
applicable and necessary for adequate control
in most irrigated areas. Available technology
needs to be integrated and applied to these
approaches. Research to develop improved
irrigation systems and methods, improved
irrigation water distribution systems, and
better field management practices, and
research on design and operational criteria for
sediment retention basins are needed.
INTRODUCTION
Surface irrigation return flow is that por-
tion of the irrigation water applied to soil which
passes over the soil surface and becomes runoff.
It usually includes direct spill from canals and
water that flows through farm ditches but is not
applied to the land. Typically, 10 to 30 per cent of
the water applied to furrow irrigated land
becomes surface runoff. Surface irrigation
return flow can also result from irrigation by
wild flooding, some border systems and where
sprinkle systems apply water more rapidly than
the infiltration rate on sloping soils. Only a
small portion of the total surface irrigation
return flow results from these latter three irriga-
tion methods. No surface runoff results when
the water application rate is equal to or less than
the infiltration rate. Such application rates can
be achieved with properly designed sprinkle
irrigation systems and with trickle systems, but
the energy requirements of sprinkle systems
and the expense of trickle systems limit their
use. In contrast to sprinkle irrigation where the
entire soil surface is the infiltrating area, only
the furrows are the infiltrating area for furrow
irrigated land. Furthermore, the furrows also
serve as conveyance channels to supply water to
the down slope portions of the field. Surface
irrigation return flows do not occur with subsur-
face irrigation or with certain border and furrow
methods that confine applied water to a given
area, including pumpback systems.
Water passing over the soil surface has
limited contact and exposure to the soil at the
soil surface, and flow at the interface is into the
soil. Therefore, the quantities of soluble salts,
fertilizer nutrients and pesticides dissolved
or washed off the soil into the water flowing over
the soil surface are expected to be extremely
small. This is particularly true where water is
confined to furrows and contacts only a portion
143
-------
RETURN FLOW MANAGEMENT
of the land surface. Such water does pick up
debris, crop residue, applied manure residue,
nematodes, plant pathogens and other foreign
matter that tends to be floated away by water.
When erosion occurs, the most important
material picked up is soil and material attached
to it. Soil picked up in the erosion process is
usually referred to as suspended sediment or
sediment.
Erosion of irrigated land has been recogniz-
ed as a serious problem for many years.
Isrealson, et al. (1946) stated that excessive
erosion of irrigated lands was adverse to the
perpetuation of permanent agriculture in arid
regions. Gardner and Lauritzen (1946) reported
that it was apparent to every farmer that serious
damage resulted when attempting to irrigate
steep slopes unless the stream was very small.
They recognized that little erosion occurred on
lands with gently slopes even with relatively
large stream sizes. These observations led them
to suggest the vital importance of finding a
means to estimate the rate at which soil would
erode with various stream sizes at various
slopes.
Today, 30 years later, furrow irrigation on
steep slopes with stream sizes that are too large,
resulting in serious erosion is still commonly
observed. Much technology has been developed
to control erosion of irrigated land and to reduce
sediment concentrations in surface irrigation
return flows, but it has not been applied.
There is a need to apply available
technology and to develop new technology for
reducing erosion and sediment loss from
irrigated lands. The purposes of this paper are to
provide an overview and an assessment of the
problems associated with sediment and ad-
sorbed nutrients and biocides in surface irriga-
tion return flows, to assess currently available
technology for implementing control measures,
and to suggest research and demonstration
needs to develop and apply improved control
technology.
Erosion on Irrigated Land
Whenever water flows over cultivated land,
erosion may occur. Factors influencing the
amount of erosion include: (1) the slope in the
direction of irrigation; (2) the stream size;
(3) the soil texture; (4) the condition of the soil
surface; (5) the duration of the irrigation; and
(6) the crop. Most erosion on irrigated land
results from furrow irrigation, and basically is
erosion of the furrows. Isrealson, et al. (1946)
reported that furrows near the head ditches
eroded 2.5 to 10 cm in sugarbeet fields. Mech
(1959) reported soil losses of 50 metric
tons / hectare during a 24-hour irrigation of corn
on a fine sandy loam soil on a 7 per cent slope.
He further stated that even on relatively flat
fields with short runs, 30 cm of surface soil have
sometimes been lost after about 10 years of
cultivation and irrigation. Similar results have
been reported in the 1970's on Portneuf silt loam
planted to dry beans, sugarbeets, potatoes and
corn.
Each furrow on furrow irrigated land func-
tions as the absorbing surface and as a channel
for conducting water to irrigate the remaining
length of run (Mech and Smith, 1967).
Therefore, the stream size at the head of the
furrow must be sufficient to meet the infiltration
requirements over the entire furrow length and
to propagate the stream to the end of the furrow
fast enough to give a reasonably uniform dis-
tribution throughout the length of run; ideally,
it should not exceed that size. Obviously, larger
streams are required to irrigate longer runs. But
larger streams have greater energy to erode
soils and transport sediment on sloping lands,
and thereby cause more erosion. More erosion is
expected near the heads of the furrows where
runs are long because that is where the stream
size is largest. Practically, short irrigation runs
have not been used because cross ditches in-
terfere with tillage, seeding, cultivating and
harvesting operations. Also, shorter irrigation
runs require more labor for irrigation. Also, it is
difficult to control stream size so that just
enough water is added to meet infiltration needs
because the infiltration rate usually changes
during irrigation. As a result of these practical
factors, irrigation runs are usually longer and
furrow stream sizes are larger than needed, and
erosion results, particularly at the heads of the
furrows.
Characteristics of flow and silt load along
irrigation furrows in two closely controlled tests
were reported by Smith and Mech (1967) (Table
1). The flow was carefully controlled into each
furrow and the runoff and sediment loss was
measured from the upper, middle and lower
third of each furrow. The run length was 274 m
and the slope was 2 per cent. The flow into each
furrow was about 15 per cent greater in test 2
than in test 1. Results of these studies clearly
illustrate that erosion was greatest where
stream size was largest. Soil eroded from the
144
-------
SEDIMENT, NUTRIENT, BIOCIDE CONTROL
upper third was deposited in the middle and
lower thirds as the stream size, and thereby the
energy to erode and the capacity to transport
sediment, decreased because of infiltration.
Results from these studies contrast to erosion
resulting from rainfall which is usually more
severe down slope where stream sizes are large
enough to erode and where slopes are greatest.
The common practice on many furrow
irrigated farms today is to place a large enough
stream in each furrow so that the water reaches
the lower end of the furrow in about 2 or 3 hours
for a 12- or 24-hour set. This usually allows
sufficient infiltration time to replenish water
depleted by the crop without reducing the
stream size or requiring labor during the set.
With this practice, stream sizes are often large
and 40 to 60 per cent of the applied water
becomes runoff, and erosion is extensive.
Another serious erosion problem is associ-
ated with the common practice in some irrigated
areas of keeping the drain ditch at the lower end
of the field 10 to 20 cm deeper than the furrows
and at a slope steep enough that the tailwater
flows rapidly away. With this practice, the ends
of the furrows erode rapidly, even with very
small streams. This erosion moves up the slope
because erosion increases the effective slope
near the end of the furrows. As the process
continues, the slope is increased on the lower 5
to 10 m of the field, making it difficult to control
erosion and soil loss from this portion of the
field, and to achieve adequate intake because of
smaller wetted perimeters. The lower ends of
fields may have to be reshaped every few years
because of this practice. This type of erosion is
easily controlled by proper tailwater manage-
ment.
Many fields with steep slopes are irrigated,
and usually in the direction of the steepest slope,
even though it has been recognized for decades
that serious erosion results from such practices.
Isrealson, et at. (1946) demonstrated over 30
years ago that increasing the slope from 1.15 to
6.07 per cent increased the erosion 16 times.
About that same time, Gardner, et al. (1946) and
Gardner and Lauritzen (1946) presented rela-
tionships among furrow slope, stream size and
erosion. Unfortunately, irrigation farmers gave
little attention to these results.
Following the early work in Utah, the
USDA-SCS Division of Irrigation conducted
many tests throughout the western U.S. and
developed the relationship:
Max. Non-Erosive Stream Size,
I/sec =
0.63
Slope, %
(1)
Evans and Jensen (1952) and Mech (1949)
studied the effects of stream size, slope, and soil
surface conditions on erosion. All of the work to
date suggests that erosion may be expected on
most row-cropped soils when slopes exceed 1 per
cent. Erosion may be controlled reasonably well
on slopes up to 2 per cent if the stream size is
carefully controlled.
Public Law 92-500 has increased the in-
terest among farmers and irrigation districts to
control erosion and sediment in surface return
flows. Many questions raised about erosion and
sediment loss indicate that few irrigators and
other personnel associated with irrigation have
a good concept for visual determination of
erosion in furrows. Carter and Bondurant (1976)
presented a simple equation to estimate soil
erosion:
Soil erosion, _L = 1.2 x eroded area, cm2 (2)
ha furrow spacing, m
Equation (2) assumes a soil bulk density of
1.2g/cm3 or t/m3. They also presented a
nomogram for estimating erosion losses in
English or metric units.
Sediment in Surface Irrigation Return
Flows
Sediment concentrations in surface irriga-
tion return flows vary widely. Brown, et al.
(1974) reported concentrations ranging from 20
to 15,000 ppm from studies of two large irrigated
tracts in southern Idaho (Table 2). Sediment
concentration in the canal waters are given for
comparison. The sediment concentrations in
most surface drains exceed those in the irriga-
tion water several fold, an exception is the W
drain, which functions as a sediment retention
basin with a long retention time. The sediment
loss from a field or an irrigation tract is deter-
mined by the volume of surface runoff and the
sediment concentration. Brown, et al. (1974)
reported a net sediment inflow for the 65,350-ha
Northside tract and a net sediment outflow for
the 80,030-ha Twin Falls tract. There was ero-
sion on both tracts, but most of the sediment
settled in drains on the Northside tract.
145
-------
RETURN FLOW MANAGEMENT
whereas, much of the sediment reaching drains
on the Twin Falls tract was carried to the river
because flow velocities in the drains were
greater and sediments did not settle and deposit
in the drains.
Nutrients in Surface Irrigation Return
Flows
Nutrients in surface irrigation return flows
are in dissolved forms, or they are attached to
sediments eroded from the land. Bondurant
(1971) showed mathematically that little soluble
nutrient pickup could be expected to result from
nutrient diffusion out of the soil into water
passing over the soil surface, and he presented
field data to verify his contention. Carter, et al.
(1971) found that soluble nutrient and salt
concentrations in surface irrigation return
flows were essentially the same as those in the
applied irrigation water. Edwards, et al. (1972)
stated that once nitrate enters the soil surface, it
does not re-enter surface runoff. Fitzsimmons, et
al. (1972), Naylor and Busch (1973) and Carlile
)1972) reported that nitrate and ammonium
nitrogen concentrations were about the same in
surface runoff as in the irrigation water. Naylor,
et al. (1972) illustrated that nitrogen concen-
trations in surface irrigation return flows from
fields can be markedly increased when liquid
nitrogen is added to the irrigation water for
fertilizing the crop. Fertilizer losses in the
surface runoff from this practice were propor-
tional to the fraction of the applied water that
became surface runoff during the fertilizer
application. In these studies, the soluble nitro-
gen was added directly to the water, increasing
the soluble nitrogen concentration in the irriga-
tion water. The concentration did not change as
the water passed over the soil surface.
Phosphorus is tightly held by soil, and
essentially all phosphorus in surface irrigation
return flow is associated with sediment. Carter,
et al. (1974) and Carter, et al. (1976) have
extensively studied phosphorus-sediment
relationships in irrigation return flows, and
their results show that total phosphorus and
sediment concentrations in surface runoff are
closely related, but that no such relationship
exists between soluble orthophosphate and sedi-
ment concentrations. They developed a regres-
sion equation relating total phosphorus con-
centration to sediment concentration over a
wide range of conditions. Fitzsimmons, et al.
(1972) and Naylor and Busch (1973) attributed
greater total phosphorus concentrations in sur-
face irrigation return flow than in irrigation
water to the greater sediment concentration in
the runoff water. Data reported by Carlile (1972)
also illustrate the close relationship between
sediment and total phosphorus concentration.
Results from many investigations show
conclusively that increases in nutrient concen-
trations from the irrigation to the surface runoff
water are closely associated to erosion and
subsequent increase in the sediment concentra-
tion in the surface runoff water. Therefore,
controlling the sediment in surface irrigation
return flows will also control most of the
nutrients.
Biocides in Surface Irrigation Return
Flows
There is little published information on
biocide concentrations in surface irrigation
return flows. There is considerable information
available on biocide concentrations in surface
runoff from nonirrigated lands. A review of the
literature indicates that except where biocides
are applied to the water, or where they are
washed off plant material in soluble forms by
rain or by sprinkle irrigation, the biocides in
surface runoff water are adsorbed to sediments
(Evans and Duseja, 1973). Unpublished data
from analyses of surface drainage waters and
sediments from the Northside and Twin Falls
tracts show that essentially all of the biocides
are adsorbed to sediments (Carter, 1975). The
available information indicates that con-
trolling sediments in surface irrigation return
flows will also control most of the biocides.
Controlling Sediments and Associated
Nutrients and Bioeides in Surface
Irrigation Return Flows
There are three broad general ways to
control sediments and associated nutrients and
biocides in surface irrigation return flows. One
is to eliminate or reduce surface irrigation
return flow. The second is to reduce or eliminate
soil erosion so that there will be little or no
sediment in surface runoff from irrigation. The
third is to remove sediments and associated
materials from surface irrigation runoff before
these waters enter natural streams. If runoff
can be eliminated, obviously there would be no
need for the second and third general ways of
control. Any farmer or irrigation district mak-
ing sufficient progress on the first two ways, so
that sediment and associated material concen-
146
-------
SEDIMENT, NUTRIENT, BIOCIDE CONTROL
trations are reduced below problem levels, will
no longer need to consider the third way. Such
progress should be the aim of irrigated
agriculture, with the recognition that many
years may be required to achieve this goal.
However, much immediate progress could be
made if presently available technology were
applied (Carter, 1972).
Eliminating or Reducing Surface
Irrigation Return Flows
There are irrigation methods that produce
no surface runoff. These include properly
designed and operated sprinkle sytstems, basin,
trickle, some border irrigation and level furrow
systems. These methods all have limitations.
Basin, border and level furrow methods are
limited to nearly level land. Capital investment
is high for center pivot, side roll and solid set
sprinkle systems and even higher for trickle
systems. Furthermore, energy requirements for
sprinkle system operation are high and energy
is limited. Batty, et al. (1975) compared energy
inputs involved in the installation and opera-
tion of various sprinkle and surface irrigation
systems and found that on a total annual
energy basis, surface systems required only 10
to 22 per cent as much energy as sprinkle or
trickle systems where some pumping was re-
quired for surface systems (Table 3). Energy
requirements for gravity surface systems would
be less than for those requiring pumping
energy.
Sprinkle irrigation is an efficient means of
applying water and can be used on lands too
steep for surface irrigation and lands with
undulating topography. The land area under
sprinkle irrigation is rapidly increasing, but
energy restraints may limit development in
some areas. Certainly, utilizing sprinkler
systems where practical can eliminate or reduce
surface return flows. However, larger center
pivot systems apply water at high rates and
may cause serious runoff problems (Pair, 1968)).
The recirculating or pump-back system
described by Bondurant (1969) and others
(Davis, 1964 and Pope and Barefoot, 1973) is a
useful method for eliminating, or greatly reduc-
ing, surface irrigation return flows from farms.
This method uses a basin or pond at the lower
end of the field to catch surface runoff. A pump
returns the water to the upper end of the field or
to another field for reuse as irrigation water.
Erosion is not eliminated and sediments
deposited in the basins must be removed
mechanically, but sediment is prevented from
leaving farms and entering natural streams.
Carter and Bondurant (1976) have sum-
marized and discussed irrigation methods with
little or no surface runoff in more detail. They
point out that eliminating or greatly reducing
surface irrigation return flows may cause other
problems in the irrigated West. Many farmers
depend wholly or in part upon surface return
flows from upstream irrigated farms or tracts
for their irrigation water supply. Many surface
irrigated tracts operate on a reuse principle so
that the only water entering streams is surface
runoff from the farms at lowest elevation in
these tracts.
Eliminating or greatly reducing surface
runoff is a means to control sediments and
associated materials in surface irrigation return
flows in some areas, but changing to irrigation
practices with no runoff would, in many in-
stances, cause other problems and require costly
changes in system design and operation. Where
changes are practical, they should be im-
plemented.
Reducing or Eliminating Erosion
Controlling Slope
Land slope greatly influences erosion.
Results of many investigations have shown
that erosion may be expected on most row
cropped soils where slopes are 1 per cent or more
(Mech, 1959; Mech and Smith, 1967; Swanson,
1960; Swanson and Dedrick, 1967; Harris and
Watson, 1971). Erosion may be controlled
reasonably well up to slopes of 2 per cent, but
fields with slopes greater than 2 per cent should
be examined carefully to see if they can be
irrigated by different methods. Changing the
direction of irrigation to one of lower slope,
contour irrigation and land grading to reduce
the slope near the lower ends of the fields to
decrease flow velocity are possible ways of
controlling slope. These changes are not
without problems. Farmers resist contour farm-
ing because usually short rows result, thereby
adding difficulty to farming operations with
large equipment.Grading to decrease the slope
and flow velocity usually causes furrows to fill
with sediment and flooding or lateral flow
between furrows results. Nevertheless, where
slopes can be reduced to 1 per cent or less, the
147
-------
RETURN FLOW MANAGEMENT
amount of erosion and sediment loss can be
reduced.
Controlling the Furrow Stream Size
Excessive stream sizes cause serious ero-
sion on sloping land (Mech, 1959; Mech and
Smith, 1967). Devices that positively control the
amount of water from the pipeline, flume or
ditch into each furrow are essential to effective
erosion control and efficient irrigation. Most
valves, gates, siphon tubes, and other flow
control devices permit small flow adjustments
that remain unchanged until reset. This equip-
ment is available, but is often not used or used
incorrectly.
A greater initial flow is often desired to get
the water to the end of the furrow and allow a
uniform intake time. Once the water reaches the
end, the flow should be reduced or cutback to
decrease erosion and runoff. However, when the
stream size is reduced for a given water set, the
excess water from the set after cutback must be
used elsewhere or wasted in most systems with
open ditches. Applying it to other sections of the
farm would require that irrigation sets be made
several times each day, and this conflicts with
other farming operations. Humpherys (1971)
developed several systems for reducing flows in
furrows after water has reached the ends. One
system splits the set, applying all the water
alternately to half the set until water reaches
the ends of the furrows, then applies the water to
the entire set so that flow in each furrow is one
half the amount initially.
Much can be done with present technology
to reduce erosion by controlling stream size.
Further development and application of
automated systems with proper stream size
control would bring about a marked reduction in
erosion and sediment loss from furrow irrigated
land.
The Run Length
The run length and furrow stream size are
closed related. Short runs can be irrigated with
small streams with very little erosion and
sediment loss, but cross ditches interfere with
farming practices. The multiset systems
developed by Rasmussen, et al. (1973) provide
an alternative to cross ditches for shortening
the run length. Aluminum or plastic pipe dis-
tributes water at several points along the
furrows so that small streams are used and
erosion and runoff is essentially eliminated.
Pipes are moved for tillage, seeding and
harvesting operations.
Worstell (1975) field tested an adaptation of
the multi-set systems with buried laterals so
farming operations could be carried out without
moving pipe. The system is fully automated and
can be programmed to apply water daily ac-
cording to ET depletion, or less often if desired.
Initial results are promising, but further testing
is needed. These kinds of systems have great
potential for controlling stream size or length of
run. With such systems, erosion and sediment
loss can be essentially eliminated and irrigation
efficiency can be greatly increased.
Controlling Irrigation Frequency and Duration
Erosion and sediment loss are highest dur-
ing the early part of an irrigation after soils
have been disturbed by cultivation. Mech (1959)
reported a soil loss of 39.9 t/ha from a recently
cultivated corn field during the first 32 minutes
runoff. The total soil loss for a 24-hour irrigation
was 50.9 t/ ha, and it occurred during the first 4
hours even though runoff increased after that
because of decreasing intake. Less erosion
would occur with less frequent irrigations, par-
ticularly when irrigations follow cultivations.
Alternate furrow irrigation is another prac-
tice to reduce erosion. Only half as much soil
surface is contacted by water and erosion is less.
However, the success of alternate furrow irriga-
tion depends upon soil conditions. Some soils do
not permit adequate lateral water movement, or
deep percolation losses may be too great during
the increased time required for lateral move-
ment.
Removing Sediment and Associated
Nutrients and Biocides From Surface
Irrigation Return Flows
Controlling tailwater
The most important factor in controlling
tailwater is to limit the amount of runoff. The
smallest stream that will irrigate to the end of
the furrow will add nearly as much water to the
soil as a larger stream, and the amount of runoff
water will be much less and more easily con-
trolled. Practices that will assure more uniform
intake rates of individual furrows need to be
developed and utilized for better runoff control.
The drain ditch at the lower end of the field
should be shallow and at a low slope, or checked
148
-------
SEDIMENT, NUTRIENT, BIOCIDE CONTROL
so that water moves slowly and sediments settle
out before the water leaves the field. Checking
the drain ditch forms miniature sediment
basins. Brockway, et al. (1976) found that mini-
basins receiving runoff from a few furrows each,
effectively controlled sediment losses.
Passing tailwater through grass or other
close growing crops efficiently filters sediments
from water. Grass buffer strips, heavy seeded
fall grain strips, or alfalfa at the ends of row
cropped fields, can greatly reduce sediment
losses. Another alternative is to utilize runoff
from row crops to irrigate alfalfa, pasture or
other close growing crops.
Utilizing Sediment Retention Basins
Much of the sediment in surface irrigation
return flows can be removed in sediment reten-
tion basins. The need to remove sediments from
surface irrigation return flow will continue for
many years even though much can be done to
reduce erosion and sediment loss from irrigated
fields. Basins are a partial cure to the sediment
problem, not a prevention. Their construction
and periodic cleaning are relatively expensive.
The effectiveness of simple sediment reten-
tion basins is illustrated by a 0.45-ha basin
removing 2390 t of sediment during two irriga-
tion seasons from part of the runoff water from a
117-ha tract. The erosion loss was 20.5 t/ha.
The sediment removal efficiency exceeded 80
per cent when the sediment concentration ex-
ceeded 0.1 per cent and it was never less than 65
per cent (Robbins and Carter, 1975).
The trapping efficiency of sediment basins
is directly related to the forward velocity, settl-
ing depth and particle size of the sediment.
Basins can be designed to remove given particle
sizes if the flow volume is known so that velocity
relationships can be established. The trapping
efficiency of one district basin designed to
remove at least 50 per cent of the incoming
sediment has not been less than 65 per cent over
5 years (Brown, 1977). More information on
design criteria is needed and some is being
developed and tested currently (Bondurant, et
al. 1976).
Particle size segregation takes place as
sediments settle in basins. Sediments remain-
ing in suspension are mostly in the clay size
fraction, although much clay settles in
aggregates because dispersion is not complete.
Dispersion is greater in waters with very low
salt concentration, and more clay remains
suspended. The clay size fraction is richer in
phosphorus, so passing surface runoff through
a sediment retention basin can give an apparent
phosphorus enrichment when phosphorus is
measured per unit of suspended material.
However, sediment retention basins conserve
phosphorus because most of the sediment is
removed by the basins (Carter, et al. 1974).
The use of sediment retention basins can be
discontinued for any field, farm or district
where implementation of erosion control prac-
tices have eliminated excessive sediment con-
centrations in the surface irrigation return flow.
Also, use of basins may not be needed every
season, depending upon the crops grown. Non
use during one or more seasons when close
growing crops are grown would allow the
collected sediment to dry and time for cleaning.
CONCLUSIONS
The quantity of sediment and associated
nutrients and biocides in surface irrigation
return flows could be reduced significantly by
applying presently available control
technology. There are restraints to direct
application of some practices such as the energy
limitation for converting to sprinkle irrigation,
and the dependence of some farmers on the
surface runoff from upstream farms or tracts for
their supply or irrigation water. The develop-
ment of irrigation methods with precise flow
controls that distribute water over the entire
field with little or no runoff and with low energy
requirements should receive top priority. The
buried lateral multiset system is an example of
systems that might be developed and improved.
The basic relationships among stream size, flow
velocity, erosion, sedimentation, run length and
TABLE 1
Water Flow and Soil Loss along Irrigation Furrows
(Mech and Smith, 1967).
Distance
from
upper end
m ft
0 0
91 300
183 600
274 900
0 0
91 300
183 600
274 900
Travel Time
Flow per furrow
per minute
liters
26.6
17.0
7.3
2.5
30.6
20.7
11.9
5.4
gal
7.03
4.49
1.94
0.67
8.08
5.46
3.14
1.42
Sort toss
per furrow
kg Ib
Test no. 1
0 0
43.3 116
4.8 13
0.4 1
Test no. 2
0 0
51.1 137
14.2 38
0.7 2
Runoff
%
61
21
2
66
35
8
Frompoint
of
application
min
0
48
211
682
0
24
98
436
For 91-m
(300-ft!
distance
min
48
163
471
24
74
338
149
-------
RETURN FLOW MANAGEMENT
sediment settling velocity need to be integrated are needed, and new and better water control
into new technology that will permit modifica- systems need to be developed.
tion of various control parameters. New ideas
TABLE 2
Sediment Concentrations in Irrigation and Drainage Waters For Two Large Tracts During the
1971 Irrigation Season.
Drain
K
N-32
J-8
S
W-26
W
Rock Creek
Cedar Draw
Filer Drain
Mud Creek
Deep Creek
Hansen Drain
Kimberly Drain
4/20
240
380
1,580 1
320
160
160
—
—
—
—
—
—
—
Sampling Date
Northside Canal Company, 65,350
5/3 5/17 5/28 6/7 6/15 6/29 7/13 7/26
Sediment Concentrations in parts per million
190 270 140 200 160 110 120 90
100 150 120 170 90 70 30 180
,430 2,610 510 660 660 300 80 170
350 110 140 100 200 440 110 130
80 100 60 100 130 100 60 160
50 60 30 30 40 20 20 30
Twin Falls Canal Company, 82,030 ha
5/25 6/2 6/15 6/29 7/13 7/26
Sediment Concentrations in parts per million
— — 540 300 140 190 310 320
— — 200 210 100 120 220 550
_ _ 710 400 210 710 2,250 2,120
— — 260 180 140 130 120 200
— — 200 110 70 80 60 70
4/20 5/14 5/26 6/23 7/6 7/20 8/3
— — 1,550 380 510 3,180 14,500 4,970
— 4,180 1,080 360 610 2,860 1,420 4,960
8/10
90
20
110
90
100
20
8/10
390
520
110
190
110
8/17
290
180
8/24
90
20
70
60
40
20
8/24
200
330
820
250
10
9/2
3,160
150
9/8
40
60
100
130
50
10
9/8
120
150
270
260
100
9/16
280
70
9/28
40
50
110
140
50
40
9/28
150
200
290
130
90
10/5
—
40
Canal Water, Monthly Average Concentrations
Northside
Twin Falls
April
63
74
May June July August Sept. Oct.
63 29 37 33 26 26
40 52 85 55 29 29
Nov.
26
29
Dec.
—
29
150
-------
SEDIMENT, NUTRIENT, BIOCIDE CONTROL
TABLE 3
Total Annual Energy Inputs, in Thousands of Kilocalories (or Gallons of Diesel Fuel) Per Acre Irrigated for
Nine Irrigation Systems, Based on 36-in. (915-mm) Net Irrigation Requirement and Zero Pumping Lift
(Batty, et al., 1975).
Irrigation
system
Surface without Irrigated
Runoff Recovery System
Surface with Irrigated
Runoff Recovery System
Solid-set sprinkle
Permanent sprinkle
Hand-moved sprinkle
Side-roll sprinkle
Center-pivot sprinkle
Traveler sprinkle
Trickle
Installation
energy
103.2
179.9
614.1
493.6
159.7
200.3
388.5
288.9
530.5
Pumping Installation per
energy pumping energy ratio Labor energy
35.2
48.0
770.0
770.0
804.0
804.0
864.0
1,569.0
468.0
2.93
3.75
0.80
0.64
0.20
0.25
0.45
0.18
1.13
0.50
0.30
0.40
0.10
4.80
2.40
0.10
0.40
0.10
Total energy
138.9
(15.0)
228.2
(24.6)
1,384.0
(149.5)
1,263.7
(136.5)
968.5
(104.6)
1,007.1
(108.8)
1,252.6
(135.3)
1,858.0
(200.7)
998.6
(107.8)
"These figures were obtained by dividing the installation energy by the system life and by the net acres irrigated
and multiplying by 1.03 to include annual maintenance energy for all systems except for solid set where 1.01
was used.
Conversion factors: 1 kcal = 4.19 kJ; 1 kcal = 0.000108 gal of diesel.
REFERENCES
1. Batty, J. C., Hamad, S.N., and Keller, J.
1975. Energy inputs to irrigation. J. Irrig. and
Drain, Div., Proc. Amer. Soc. Civil Eng.
101(IR4):293-307.
2. Bondurant, J. A. 1969. Design of recir-
culating irrigation systems. Trans. Amer. Soc.
Agr, Eng. 12:195-201.
3. Bondurant, J. A. 1971. Quality of sur-
face irrigation runoff water. Trans. Amer. Soc.
Agr. Eng. 14:1001-1003.
4. Bondurant, J. A., Brown, M. J. and
Brockway, C. E. 1976. Design criteria for sedi-
ment ponds. Personal communication — un-
published data.
5. Brockway, C. E., McMaster, G. M., and
Bondurant, J. A. 1976. Sediment removal by
minibasins. Personal communication — un-
published data.
6. Brown, M. J. 1977. Sediment trapping
efficiency of a district sediment pond. Personal
communication — unpublished data.
7. Brown, M. J., Carter, D. L., and Bon-
durant, J. A. 1974. Sediment in irrigation and
drainage waters and sediment inputs and out-
puts for two large tracts in southern Idaho. J.
Environ. Qual. 3:347-351.
8. Carlile, B. L. 1972. Sediment control in
Yakima Valley. Proc. Natl. Conf. on Managing
Irrigated Agriculture to Improve Water Quality,
U.S. Environmental Protection Agency and
Colo. State Univ. p. 77-82.
9. Carter, D. L. 1975. Studies of pesticides
in surface drainage waters and sediments for
two large irrigated tracts. Unpublished file
data.
10. Carter, D. L. 1976. Guidelines for sedi-
ment control in irrigation return flow. J. En-
voron. Qual. 5:119-124.
11. Carter, D. L., and Bondurant, J. A.
1976. Control of sediments, nutrients, and ad-
sorbed biocides in surface irrigation return
flows. Environmental Protection Technology
Series, EPA-60072-76-237, U.S. Environmen-
tal Protection Agency, 44p.
12. Carter, L. L., Bondurant, J. A., and
Robbins, C. W. 1971. Watersoluble NO 3-
Nitrogen, PO4-phosphorus, and total salt bal-
151
-------
RETURN FLOW MANAGEMENT
ances on a large irrigation tract. Soil Sci. Soc.
Amer. Proc. 35:331-335.
13. Carter, D. L., Brown, M. J., and Bon-
durant, J. A. 1976. Sediment-phosphorus rela-
tions in surface runoff from irrigated lands.
Proc. Third Federal Inter-Agency Sedimenta-
tion Conf., p. 3-41 to 3-52.
14. Carter, D. L., Brown, M. J., Robbins,
C. W., and Bondurant, J. A. 1974. Phosphorus
associated with sediments in irrigation and
drainage waters for two large tracts in southern
Idaho. J. Environ. Qual. 3:287-291.
15. Davis, J. R. 1964. Design of irrigation
tailwater systems. Trans. Amer. Soc. Agr. Eng.
7:336-338.
16. Edwards, D. M., Fischbach, P. E., and
Young, L. L. 1972. Movement of nitrates under
irrigated agriculture. Trans. Amer. Soc. Agr.
Eng. 15:73-75.
17. Evans, J. O., and Duseja, D. R. 1973.
Herbicide contamination of surface runoff
waters. Environmental Protection Technology
Series, EPA-R2-73-266.
18. Evans, N. A., and Jensen,M. E. 1952.
Erosion under furrow irrigation. North Dakota
Agr. Expt. Sta. Bimonthly Bull., Vol. XV, No. 1,
p. 7-13.
19. Fitzsimmons, D. W., Lewis, G. C.,
Naylor, D. V., and Busch, J. R. 1972. Nitrogen,
phosphorus and other inorganic materials in
waters ina gravity irrigated area. Trans. Amer.
Soc. Agr. Eng. 15:292-295.
20. Gardner, W., and Lauritzen, C. W. 1946.
Erosion as a function of the size of the irrigation
stream in the slope of the eroding surface. Soil.
Sci. 62:233-242.
21. Gardner, W., Gardner, J. H., and
Lauritzen, C. W. 1946. Rainfall and irrigation in
relation to erosion. Utah Agr. Expt. Sta. Bull.
326.
22. Harris, W. S., and Watson, W. S., Jr.
1971. Graded rows for the control of rill erosion.
Trans. Amer. Soc. Agr. Eng. 14:577-581.
23. Humpherys, A. S. 1971. Automatic fur-
row irrigation systems. Trans. Amer. Soc. Agr.
Eng. 14:466-470.
24. Isrealsen, O. W., Clyde, G. D., and
Lauritzen, C. W. 1946. Soil erosion in small
irrigation furrows. Utah Agr. Exp. Sta. Bull. No.
320, p. 39.
25. Mech, S. J. 1949. Effect of slope and
length of run of erosion under irrigation. Agr.
Eng. 30:379-383, 389.
26. Mech, S. J. 1959. Soil erosion and its
control under furrow irrigation in the arid west.
USDA, ARS, Agr. Inf. Bull. No. 184.
27. Mech, S. J., and Smith, D. D. 1967.
Water erosion under irrigation. In Robert M.
Hagan, Howard R. Haise and Talcott W. Ed-
minister (ed.) Irrigation of Agricultural Lands.
Agronomy 11:951-963, Amer. Soc. Agron.,
Madison, Wis.
28. Naylor, D. V., and Busch, J. R. 1973.
Effects of irrigation, fertilization, and other
cultural practices on water quality. Res. Tech.
Comp. Rep., Idaho Water Resour. Res. Inst., p.
19.
29. Naylor, D. V., Lewis, G. C., Fitzsim-
mons, D. W., and Busch, J. R. 1972. Nitrogen in
surface runoff resulting from addition of fer-
tilizers to irrigation water. Proc. 23rd Ann. Pac.
NW Fert. Conf., p. 67-73.
30 Pair, C. H. 1968. Water distribution
under sprinkler irrigation. Trans. Amer. Soc.
Agr. Eng. 11:648-651.
31. Pope, D. L., and Barefoot, A. D. 1973.
Reuse of surface runoff from furrow irrigation.
Trans. Amer. Soc. Agr. Eng. 16:1088-1091.
32. Rasmussen, W. W., Bondurant, J. A.,
and Berg, R. D. 1973. Multiset surface irrigation
system. Int. Comm. on Irrig. and Drain. Bull.,
p. 48-52.
33. Robbins, C. W., and Carter, D. L. 1975.
Conservation of sediment in irrigation runoff.
J. Soil and Water Conser. 30:134-135.
34. Swanson, N. P. 1960. Hydraulic char-
acteristics of surface runoff from simulated
rainfall on irrigation furrows. ARS 41-43:90-102.
35. Swanson, N. P., and Dedrick, A. R.
1967. Soil particles and aggregates transported
in water runoff under various slope conditions
using simulated rainfall. Trans. Amer. Soc. Agr.
Engr. 10:246-247.
36. Worstell, R. V. 1976. An experimental
buried multiset irrigation system. Trans. Amer.
Soc. Agr. Eng. 19:1122-1128.
152
-------
Quality of Irrigation Return Flow
From Flooded Rice Paddies
K. W. BROWN, L. E. DEUEL, F. C. TURNER, and
J. D. PRICE
Department of Soil & Crop Sciences,
Texas A & M University, College Station, Texas;
Texas Agricultural Experiment Station at Beaumont, and
Texas Agricultural Extension Service, College Station, Texas
ABSTRACT
A three year field and laboratory study was
conducted to determine the quantity and quality
of irrigation return flow from flooded rice
culture. Both intermittent and continuous flow
irrigation techniques were evaluated. Selected
pesticides and nutrients were applied at
recommended and excessive rates. The concen-
trations of nitrate, phosphate and potassium in
the return flow were all within levels acceptable
for drinking water. Only small amounts of the
fertilizer applied to the soil before flood or in the
floodwater were lost in the irrigation return
flow. The large amounts of rainfall received
each year diluted the salts in the floodwater so
that the salt load of the irrigation return flow
did not differ greatly from that of the irrigation
water. Increases in the electrical conductivity of
the floodwater resulted from the release of
certain ions from the soil after fertilizer
applications. These increases lasted for periods
of 5-10 days.
The 96 hour median tolerance level of
catfish to three of the four pesticides tested were
lower in the water from the paddies than for tap
water. Excessive rates of the pesticides resulted
in concentrations in the paddy water which
were greater than the 96 hour TLM concentra-
tion. For the recommended rates, the concen-
trations ofpropanil exceeded 10% of the 96 hour
TLM until 24 hours after application. For
Molinate 10% of the TLM was not exceeded at
any time, for carbofuran the period was 20 days,
while for carbaryl the period was 6 days. It is
suggested that water should not be released
from the fields during the specified periods after
application. Evidence was found that some of
the pesticides which remain on the foliage are
later washed into the flood water by rain, thus
increasing the concentration in the return flow
rather than decreasing them by dilutions. Only
low concentrations of the toxic metabolites of
the pesticides were found in the field.
INTRODUCTION
As technology advanced, particularly the
methodology and instrumentation, the distribu-
tion and levels of hazardous chemicals in our
environment are revealed. The simplest and
most expedient solution to eliminate hazardous
materials would be to ban the use of all potential
pollutants. This would include virtually all soil
amendments and chemicals employed in
agricultural production.
We have reached a point where the fertility
status of many of our soils is dependent upon
"bagged" inputs. We have mutated the hardiest
of insects, such that natural control is virtually
impossible. More succinctly stated, many of the
chemicals deemed pollutants are presently es-
sential if we are to maintain our current socio-
economic standard of living.
The most logical approach is to determine
the longevity and mobility of the chemicals used
for agricultural production and to select those
chemicals and management practices which
minimize pollution hazards.
Rice is presently an important cash crop in
Arkansas, Louisiana, Texas and California
with approximately 2.3 million acres of
irrigated rice grown yearly. Fertilizer amend-
ments and pesticides are essential for continued
production at economically feasible levels.
However, fresh water supplies for urban use and
the estuaries along the coastal regions are
153
-------
RETURN FLOW MANAGEMENT
relatively unbuffered geographically from the
rice growing areas.
The primary objective of this project was to
assess the pollution potential of irrigation
return flow from flooded rice cultivation with
respect to fertilizer elements and pesticides.
EXPERIMENTAL DESIGN
Field studies were conducted for three years
on a group of 12 small rice paddies. Earthen
levees were constructed along the boundaries of
the plots and plastic barriers interred to a depth
of 90 cm within the dikes to retard water from
moving horizontally between plots. Flooded
surface area of the plots averaged 300 m 2.
Two irrigation schemes, continuous flow
and a more static intermittent flow system, were
evaluated. A water balance was ascertained for
each plot based on the quantity and frequency
of irrigation, precipitation, evaporation, trans-
piration, runoff, and leachate through the soil
profile. A system of weirs and reservoirs was
installed to manage water levels in accordance
with the irrigation schemes tested (Figure 1).
Water stage recorders continuously monitored
RRIGATKM CANAL
S/~> ff WATER STAGE
V' I "EJOBOESS g
1o o tepth of
lnwt*r •urroundng
' HUM
Figure 1. Schematic diagram of two of the research
plots showing water control devices.
water levels in each plot. Intermittent irrigation
was managed by metering water into the plots
as needed to maintain a flood of between 4 and
10 cm. Continuous flow plots were provided with
an outflow at 10 cm above the soil surface. An
equivalent of 0.9 cm of water for the entire plot
was metered in each day.
Several methods were used to measure the
movement of water into the profile. Infiltration
under flooded paddy conditions was determined
by comparing lysimeter data and water loss
from plots. Infiltration was also estimated on
small plots isolated with metal frames and
surrounded by water.
TABLE 1
Rate of fertilizers and pesticides applied given in kg/ha
active ingredients.
Recommended Excessive
Nitrogen as N
Phosphate as P2O6
Potassium as K2O
Propanil
Molinate
Carbofuran
Carbaryl
134.4
44.8
22.40
3.36
3.36
0.56
1.12
179.2
112.0
89.60
6.72
11.20
3.36
5.60
Plots were randomized with respect to
application rates of nutrients and pesticides.
Excessive rates of both were applied to the same
plots. Rates employed for the pesticides and
fertilizers are given in Table 1. Fertilizers and
pesticides were applied to rice plots in a fashion
which was as similar to normal cultural prac-
tices as possible.
Flood water in plots were sampled on a
schedule designed to provide detailed informa-
tion following events such as irrigation, heavy
rainfall, and applications. In addition, the
irrigation water was sampled at the canal
source for comparison. Water samples for
pesticide analyses were taken as soon after
application as possible and assigned a relative
time of zero hours. Subsequent samples were
generally taken 24, 48, 96, 192, 384, and 768
hours. Time zero for propanil was approximate-
ly 24 hours following the application since that
was when the plots were flooded. Plots were
already under permanent flood when the other
chemicals were applied.
The original plan was to sample the solu-
tion of the soil profile through porous filters at 5,
15.4, 30, and 61 cm depths. However, this
approach to sampling was not reliable because
very little or no water could be withdrawn by
suction from the tight, fine-textured, very slowly
permeable soil. Thus, the lack of sample, sample
volume, and excessive variation within replica-
tion called for another sampling method.
A dialysis tube method of sample collection
was used. Dialysis tubing was filled with
distilled-deionized water, placed in the plots and
covered with 1 cm of soil. After 24 hours of
contact with the soil solution, the dialysis bags
were removed and analyzed to estimate the
ionic constituency of the soil solution.
Soil profile samples were taken at the
beginning of the experiment to establish the
154
-------
RETURN FLOW FROM RICE PADDIES
background ionic constituency. Originally, the
depths sampled were 0 - 15, 15 - 30, 30 - 45, 45 - 60,
60 - 76, and 76 - 91 cm in each plot. During 1974
and 1975 sampling was restricted to the surface
and the 0 - 15 cm depth since it became evident
that water percolation was very slow and the
first 15 cm of the profile was most important.
Fish bioassays were employed to aid in
evaluating the water quality of the paddy water
released at the end of the field experiment.
RESULTS AND DISCUSSION
Water Balance
The water balance of a rice paddy over a
season may be written as:
Wbter Depth (cm)
Wot«r D«pth (cm)
,» q =
where P is the amount of precipitation, I is the
depth of irrigation, EV is the evaporation, TS is
the transpiration, R is the runoff, and L is the
percolation loss. The left side of the equation
summarizes the gains and the right side the
losses. The water balance subdivided into gains
and losses during the period of permanent flood
for 1974 and 1975 for both irrigation treatments
is given in Table 2.
TABLE 2
Water balance during the period of permanent flood
for!974 and 1975 for both irrigation treatments.
Intermittent
Irrigation
Gains
Rainfall
Irrigation
Total
Losses
Runoff
% of total app.
Leachate
% of total app.
EVTS
% of total app.
Total
% of total app.
74
27.6
43.6
71.2
26.2
36.8
12.4
17.3
45.6
64.0
84.0
118.1
75
48.4
24.7
73.1
31.2
42.6
12.0
16.4
39.6
54.0
82.9
113.1
Continuous
Irrigation
74
27.6
90.0
117.6
34.3
29.2
12.4
10.5
45.8
38.9
92.5
78.6
75
48.4
77.9
126.3
73.7
58.3
12.0
9.5
39.6
31.4
125.3
99.2
Input from precipitation was greater in
1975 than in 1974. Total input was approximate-
ly the same in the intermittent irrigated plots,
but was a little higher in the continuous flow
plots. However, the difference generally reflects
the increased depth employed in the continuous
management scheme in 1975.
Figure 2. Seasonal patterns of water depth in in-
termittently irrigated plots during 1974. The date
line represents the bottom of the 10° outflow weir.
The increased runoff in 1975 was due main-
ly to the storm events. Runoff is reflected in the
seasonal patterns of water depth with time
relative to the 10° outflow weir in Figures 2 and
3 for 1974 and in Figures 4 and 5 for 1975.
The water which moved into the soil profile
which was not used up by the plants was
considered to be leachate. It amounted to about
12 cm each year, however piezometer data
shows that the wetting front never penetrated
past 60 cm; thus, movement through the soil
was very slow.
Under both management practices, far
more water was applied to the field than would
have been needed to supply the evapotranspira-
tion. With the continuous flow management
system, the water applied by irrigation was 2
1/2 times that needed to satisfy the
evapotranspiration and was over 3 times
155
-------
RETURN FLOW MANAGEMENT
Ctaptti (on) WoHr D«pth (cm)
., « * B = B S r»w»»«r< •«0
Dqritl (cm)
Figure 3. Seasonal patterns of water depth in con-
tinuously irrigated plots during 1974. The date line
represents the bottom of the 10° outflow weir.
greater than the evapotranspiration when the
rainfall additions are considered.
These excessive applications typical of field
operations, resulted in large volumes of runoff.
Rainfall was little less than the evapotranspira-
tion during 1974 but exceeded it during 1975.
While these rains are scattered and do not come
at exactly the time they are needed, careful
management of irrigation water levels would
make it possible to increase the amount of
rainfall utilized and to decrease the volume of
runoff.
Salt Balance
Electrical conductivity (E.G.) values of the
irrigation supply and plot flood water, averaged
over the respective treatments, are given in
Figures 6 and 7 for intermittent and continuous
flow irrigation schemes, respectively. Analyses
Figure 4. Seasonal patterns of water depth in in-
termittently irrigated plots during 1975. The date
line represents the bottom of the 10° outflow weir.
of variance indicated that time of sample collec-
tion, fertilizer application rate, and irrigation
management had highly significant effects on
mean E.G. values in 1974 and 1975. Data from
1973 are too sparse to indicate significant trends
due to treatments, although means did vary
significantly with time.
Conductance is a measure of the current
carried by electrolytes, and as one would expect
the excessive fertilizer application rate resulted
in higher E.G. values. E.G. values were greater
under the impoundment irrigation manage-
ment.
It is evident from the data that the increase
in E.G. following fertilizer application was
primarily a temporary effect. Fertilizer incor-
poration into the soil and/or applied to dry soil
prior to flood resulted in lower salt levels in the
plot water sampled, as evidenced by the fact
that peak concentrations were about equal,
although the pre-plant and tillering application
rates were twice the panicle differentiation rate
(Table 3).
156
-------
RETURN FLOW FROM RICE PADDIES
WoNr DmHt (cm)
dF1-
*••
f-
d*
S*-
r
!f
sr
rf*
jp
f
f
f
f
N
f
f
f
f
f
f
f
f
f
f
f
f
t
st
~$
2
*
f
f
g
W*r D** text
f 9 S ? • » 8
Figure 5. Seasonal patterns of water depth in con-
tinuously irrigated plots during 1975. The date line
represents the bottom of the 10° outflow weir.
Irrigation and paddy water pH values for
continuous flow and intermittent irrigation
management schemes are given in Figures 8
and 9. The general trend was for the pH of the
paddy to increase toward that of the irrigation
water with time. Significant decreases in pH
correspond to the (NH4)2864 applications.
One could expect this since (NH4)2SO4 is an
acidic salt. While the excessive application rate
resulted in generally lower pH values, analyses
of variance indicated that rate was not signifi-
cant at a 5% level in either 1974 or 1975.
The intermittent irrigation scheme resulted
in a significantly lower pH in 1974, but imparted
little variation on treatment means in 1975. The
difference between the two years may be due to
the fact that a smaller percentage of the total
water volume was exchanged under the con-
tinuous flow management scheme in 1975. Con-
1973 EC. of the Paddy
Water (Impounded flow)
Sampling Dates
1974 EC. of the Paddy
Water (Impounded flew)
Recommended
E*c*ss-ve
Canal
• V
SCO
§ *»
i
i _
Sampling Doles
1975 EC of the Paddy
Water (Impounded flow)
Sampling Dates
Figure 6. Electrical conductivity in ^mhos/cm for
water in impounded plots and in the canal.
tinuous flow plots had been made deeper in 1975
to investigate the influence of plot depth. Thus,
the deeper plots resulted in a larger total water
volume, lessening the impact from 0.9 cm/day
outflow.
The salt load of each irrigation and all
runoff was calculated from plot water concen-
trations and the water balance data for both
continuous and intermittent irrigation
treatments (Table 4). More salt was applied to
the intermittent flow plots during 1974 than in
1975, but the greater concentration of salts in
the runoff during 1974 resulted in more salt
being removed. Continuous flow plots received
much more water than intermittent irrigated
157
-------
RETURN FLOW MANAGEMENT
plots, consequently, the amount of salt added to
the plots was greater. However, that lost to
runoff was far less than the applied indicating
that even under the continuous flow system, the
plots were absorbing much of the input from the
fertilizer applications and load of the irrigation
supply.
Nutrients
The concentrations of nitrates, ammonia,
phosphates, potassium, and the associated
cations and ions were monitored in the
floodwater throughout all seasons. For the
continuous irrigation plots, through which
water was flowing at all times, the data
collected is representative of the outflow water.
!
j*.
1973 EC of the Poddy
Woter (Continuous flow)
S S, S.
Sampling Dales
*» \ •*
I
1
1974 EC of the Patty
Water (Continues flow)
t
\\
r£&\ --- Eic
- Cowl
S> So So H
§
I
* «fr
Sampling Dolts
1975 EC of the Poddy
Water (Continuous flow)
*» S
So
S> So So S S» «••» *. 7* ?4» S %
Somptng Datu
Figure 7. Electrical conductivity in ^mhos/cm for
water in continuous flow plots and in the canal.
TABLE 3
Associated ions added with fertilizers during
the three years.
Associated ion added
Growth stage Fertilizer Associated Recommended Excessive
ofrice element onion/cation rate rate
Preplant
Preflood
Panicle
Differen-
tiation
NH;
K-
H,PO4
NH;
NH;
so.-
ci-
Ca'
sor
sor
sor
184
16
: -
:
184
92
246
33
-
246
121
TABLE 4
Salt balance during the rice growing season during
1974 and 1975.
Irrigation Salt applied in Salt lost Salt gained
Year technique Irrigation water in runoff by flood
1974
1974
1975
1975
Impounded
Continuous
Impounded
Continuous
kg/ha
528
993
428
712
kg/ha
559
575
433
587.9
kg/ha
-31
417
5
124
For the intermittent plots, the data represents
the concentration in the outflow when it oc-
curred due to excess rainfall and, in addition,
may be interpreted as the concentrations that
would have occurred in the outflow had the
water been deliberately released at any time
during the season. Typical nitrate concen-
trations in the impounded treatment are shown
in Figure 10. The highest concentrations are
associated with the period between planting
and permanent flood when nitrogen was liable
to oxidize between irrigations. A small peak
occurred shortly after the NH4SO4 surface
application at the time the permanent flood was
initiated, and an even smaller peak occurred as
a result of the water application at panicle
differentiation. All of the concentrations
measured were well below the 10 ppm NOs-N
considered to be the upper limit for drinking
water. Typical data on the concentration of
NHs-N are shown in Figure 11. The concen-
trations were low during the period when plots
were not flooded despite 40% of the N being
applied at seeding. Peaks which have half lives
of the order of 2 to 3 days occurred after the 40%
was surface applied just before flooding and
again after the 20% application at panicle
differentiation. Peak concentrations were
158
-------
RETURN FLOW FROM RICE PADDIES
typically 10 ppm NH 4-N, but occurred only on
one or two days each season. By approximately
five days after applications, the concentrations
decreased to below those in the incoming canal
water.
Concentrations of K were low and typically
of the same magnitude as the canal water. The
concentrations of ortho-phosphates followed
similar patterns.
Recommended
Canal
1973 pH of theKfqddy
*
1*
« Water (Impounded flow)
*~~* T» *
vso
3/20
6/SO
TABLE 5
Propanil recovered in water from treated rice plots
sampled 0 and 24 hours following the flood in 1973.
Water mgt.
Impounded
Impounded
Flowing
Flowing
Treatments
Hours following flood
kg/ha propanil
3.4
6.8
3.4
6.8
Experiment Ave'T SNK (p=2), 0.692
0
kg/hat
1.608
2.210
1.442
2.343
1.901a
24
0.001
0.001
0.002
0.002
0.002D
TValues represent mean of 3 replications.
"TAverages over entire experiment not followed by the same
letter are significantly different at the 5% level using a Stu-
dent-Newman-Keul's range test.
1973 pH of the Poddy
Water (Continuous flow)
— Recommended
Excessive
x • Canal
v
4/30 5/20
'14*
-_•
6/10
6/30
7/10
8/10
1974 pH of the Poddy
Water (Continuous flow)
— Recommended
Excenive
4/30 5/20
6/10
6/30
7/10
8/10
1975 pH of the Paddy
Water (Continuous flow)
"£.
4/30 5/20
6/K5 6/30 7/10
SompfenQ Dotes
8/K5
Figure 8. pH of water in continuous flow plots and
in the canal.
Excewive
—Recommended
Canol
4/30 5/20 STO «/30 7/20
1975
—Excesiive
X-e
Q.
4/30 6/20
•710 6/30 7/20
Sampling Dates
Figure 9. pH of water in impounded plots and in the
canal.
Pesticides
The pesticides employed in this study were
propanil (3', 4' dichlo-ropropionanilide),
molinate (S-ethyl-hexahydro-1-H-azepine-l-car-
bothioate), carbofuran (2, 3-dihydro-2, 2-di-
methyl-7-benzofuranyl-N-methyl carbamate),
and carbaryl (1-naphthyl-N-methyl carba-
mate).
Propanil was applied as a foliar spray 24
hours prior to the permanent flood application.
The 1973 experiment indicated that propanil
dissipated from the flood water within 24 hours
(Table 5). A more rigorous sampling schedule
was employed in 1974 and 1975 to determine the
rapidity with which propanil dissipated in the
plots (Tables 6 and 7). Although propanil con-
centration was about constant or increased over
the first 12 hours, it did not persist at significant
levels 24 hours following the flood application.
Plot water concentrations were generally found
159
-------
RETURN FLOW MANAGEMENT
1973 NH4-N in the Poddy
Water (impounded flow)
com
Excessive
— Recommended
5/C 9/20 9/30 6/9 6/19 6/29 7/9 7/19 7/29 »/B 8/l»
Figure 10. Concentration of NO3-N in ppm in im-
pounded plots and in the canal water.
?**"
lo
|
1-
1973 - NOs-N in the Poddy
Water (impounded flow)
i
:
0 °
:
4
o *i
j< .» ,
Excessive
V* «. canai „
Bi?^ ^T« U ii -rf ii n i i — ,
*, S •S. ^ S v.
Sampling Dote*
Figure 11. Concentration of NH3 + in ppm in im-
pounded plots and in the canal water.
to be a function of that washed from the foliage
(Figures 12 and 13).
Molinate was applied in granular formula-
tion by broadcast over the entire plot. Applica-
tion followed the permanent flood by 8 days in
1973 and by 18 days in 1974 and 1975. While
persistence was generally longer in the 1974 and
1975 field experiment, residue levels were more a
function of application rate, and time elapsed
following application (Table 8). Residues were
generally greater under the intermittent irriga-
tion management scheme all 3 years, but
statistically significant only in 1973.
Molinate residual levels were found to be a
logarithmic function of time (Table 9). Half-life
or time required for 50% reduction in the
molinate residual level was calculated for each
I
3-
s too-
mo 300
wo «oo .100
(kt/ha)
.100 .MO
Figure 12. Propanil recovered in the water immedi-
ately following the flood as affected by the absorbed
foliar concentration prior to the flood application
in 1974.
replication within treatment blocks (Table 10).
Although half-life values were longer in 1974
and 1975, the difference between years was not
significant at the 5% level. Unlike the trend
observed on persistence of molinate, the ex-
cessive application did not result in a signifi-
cant increase in half-life.
TABLE 6
Propanil in water from treated rice plots
sampled 0, 3, 6,12 and 24 hours following the
flood in 1974.
TVeafmenfs
Hours following flood
Water mgt. propanil 0
:
24
Impounded 3.4
Impounded 6.8
Flowing 3.4
Flowing 6.8
Experiment Ave't
SNK(p=5l 0.113
0.136
0.330
0.078
0.322
0.217a
0.070
0.242
0.090
0.249
0.163a
kg/hat
0.041
0.249
0.087
0.236
0.153a
0.167
0.105
0.091
0.241
O.lSla
0.008
0.008
0.011
0.005
O.OOSb
tValues represent mean of 3 Replications.
•tAverages over entire experiment not followed by the same letter
are significantly different at the 5% level using a Student-New-
man-Keul's range test.
Irrigation management scheme had a high-
ly significant influence on half-life values
within a given year, where intermittent flow
(the more static system) resulted in a longer
half-life. A first order interaction between year
and irrigation management was significant at a
5% level, and suggested that the influence of
intermittent flow was more pronounced in 1974
and 1975.
It was established in the laboratory, under
simulated flood conditions, that molinate was
dissipated biologically and apparently by
160
-------
RETURN FLOW FROM RICE PADDIES
aerobic organisms, since no significant
degradation was noted in the sterilized controls,
and molinate dissipation was greater in the
more oxidized samples (Table 11). These results
are consistent with those of the 3 year field
experiment. Field plots were flooded 8 days prior
to molinate application in 1973; wheras, the
corresponding time interval was 18 days in 1974
and 1975. Thus, the longer half-life and sus-
tained persistence in the latter two years ap-
parently was due to a more reduced environ-
ment.
Results of the laboratory experiment in
which redox potential was the only variable,
clearly showed that molinate dissipation rates
TABLE 7
Propanil recovered in water from treated rice plots
sampled 0, 3, 6,12, and 24 hours following the
flood in 1975.
Treatments
Hours following flood
Water mgt.
Impounded
Impounded
Flowing
Flowing
Experiment
Ave't
SNK(p=5)
kg/ha
propanil
3.4
6.8
3.4
6.8
0.560
0
0.817
1.036
0.440
1.108
0.850a
3
0.176
1.203
0.466
1.036
0.720a
6
kg/hat
1.267
1.671
1.327
2.525
1.607b
12
0.822
0.989
2.306
1.447b
24
0.061
0.056
0.075
0.208
O.lOOc
tValues represent mean of 3 replications.
•tAverages over the entire experiment not followed by the same
letter are significantly different at the 5% level using a Student-
Newman-KeuI's range test.
I-
.£
So*
Y.03O5'05SO(X)
r-OTJ
OB 12 IS
Propanil Rnscd From the Ftitogt (kg/ho)
Figure 13. Propanil recovered in the water immedi-
ately following the flood as affected by the absorbed
foliar concentration prior to the flood application
in 1975.
are affected by redox potential (Figure 14).
Dissipation rates were drastically reduced at
E values less than +72 mV.
120
CO
£
60
40
20
242 202 62 122 82 42 0
Eh fciM
Figure 14. Molinate dissipation rates as a function
ofEh.
Residual levels of carbofuran found for the
various treatments with respect to time follow-
ing application are plotted in Figures 15,16, and
17 for the 1973, 1974 and 1975 experiments,
respectively. Residual levels were highest in-
itially and decreased rapidly to less than 50% of
the initial concentration within 24 hours in
1973. Carbofuran residues in the water followed
a different pattern in 1974 and 1975. The
amounts in the water were low initially, and
highest in the 24-hour samples. Carbofuran was
applied in the plot water in a commercially
available granular form, and sufficient time
may not have elapsed for dissolution in 1974
and 1975. However, a time differential cannot
explain the anomaly in the 24-hour samples. As
much as 60% of that applied at the excessive rate
could be accounted for in 1974, 24 hours later.
Conversely only about 30% of that applied in
1973 was in the water and this maximum
occurred in the zero-hour samples.
During 1973, no rain fell until after the 192-
hour sampling period. This and subsequent
rains had little affect on the carbofuran residual
level in the flood water. During 1974, a 1.24 cm
rain fell just before the 192 hour sampling period
causing a second peak in concentration. Subse-
quent rains resulted in no increase in the
161
-------
RETURN FLOW MANAGEMENT
TABLE 8
Molinate recovered in water from treated rice plots in 1973,1974 and 1975.
Treatments
Year
1973
1974
1975
Water mgt.
Continuous
Continuous
Intermittent
Intermittent
Continuous
Continuous
Intermittent
Intermittent
Continuous
Continuous
Intermittent
Intermittent
kg/ha
Molinate
3.4
11.2
3.4
11.2
3.4
11.2
3.4
11.2
3.4
11.2
3.4
11.2
0
1.133a
3.504a
1.560a
6.028a
2.339a
5.928a
1.572a
6.908a
0.007d
0.019d
O.OOlc
O.OOlc
24
0.766ab
2.802b
1.396a
4.910b
2.064ab
5.225a
1.151a
5.546a
1.780a
5.265a
1.626a
4.827a
Molinate residue {kg/hat*
(hours following application)
48 96 192
0.374bc
1.417c
1.157ab
3.659c
—
1.321b
4.598a
1.536a
4.505a
0.231bc
0.572d
0.797bc
2.172d
1.609b
3.589b
0.321b
3.762c
0.794c
3.333b
1.287a
3.785a
0.197bc
0.072d
0.251c
0.547e
0.428c
0.967c
0.587b
1.941d
0.235d
1.469c
0.7 19b
1.923b
384
0.004c
0.004d
0.036c
0.068e
0.021c
0.069d
O.OGlc
0.531e
0.017d
0.291d
O.lOlc
0.867bc
768
O.OOlc
0.000
0.004c
O.OOSe
0.000
0.002d
0.000
0.053e
O.OOOd
0.002d
0.014c
0.083c
*Values represent average of three replications. Values within a given treatment block not followed by the same
letter are significantly different at the 5% level.
TABLE 9
Regression coefficients determined from best fit
analysis to first order decay equation.
Year
1973
1974
1975
Treatments
kg/ha
Water mgt. molinate
Continuous flow
Continuous flow
Intermittent
Intermittent
Continuous flow
Continuous flow
Intermittent
Intermittent
Continuous flow
Continuous flow
Intermittent
Intermittent
3.4
11.2
3.4
11.2
3.4
11.2
3.4
11.2
3.4
11.2
3.4
11.2
Regression
Coefficient (r1)
0.94
0.98
0.98
0.98
0.96
0.99
0.86
0.98
0.99
0.98
0.98
0.99
amounts in the flood water. In 1975, a 0.53 cm
rain fell just before the 24-hour samples were
collected, and correspondingly resulted in the
greatest plot water concentrations. The values
decreased markedly between the 24 and 48-hour
samples. Two rains totaling 0.79 cm were
recorded between the 48 and 96-hour, which
may have washed additional material into the
water resulting in a second peak in 3 of the 4
treatments at 96 hours. The influence of subse-
quent rainfall was not evident. The cor-
respondence of peak plot water values and
TABLE 10
Half-life of molinate in paddy H2O sampled in 1973,
1974 and 1975 as affected by irrigation
management and application rate.
Treatments Half-life in hours*
Molinate
Water mgt. kg/ha 1973 1974 1975
Continuous flow
Continuous flow
Intermittent
Intermittent
3.4
11.2
3.4
11.2
68
41
84
73
50
63
88
106
45
61
100
125
* Values averaged over 3 replications.
rainfall events, indicated a fraction of the
granular material may have lodged in the
sheath of the rice foliage.
Application rate and time were found to
have a highly significant influence on the
amounts of carbofuran in the paddy water in
each of the 3 years tested. Irrigation manage-
ment scheme had no affect on plot water values
at the 5% level of significance. The only signifi-
cant interaction found was between application
rate and time. The excessive rate resulted in
longer persistence of significant residue levels.
A commercially available formulation of
carbaryl was foliarly applied approximately 3
weeks prior to harvest in 1973, 1974, and 1975
(Tables 12, 13, and 14). Peak concentrations
correspond to rainfall events. The greatest
162
-------
RETURN FLOW FROM RICE PADDIES
TABLE 11
Effect of time, substrate level, and redox
potential on the dissipation of molinate in
flooded soil samples under laboratory conditions.
Incubation
period
days
1
8
16
Sucrose
added
g
none
0.25
1.00
sterilized
controlt
none
0.25
1.00
sterilized
controlt
none
0.25
1.00
sterilized
controlt
Rep
2
-
1
2
1
',
•
2
.
!
1
2
1
:
i
2
Molinate
recovered
%
89.1
98.4
99.5
97.8
96.7
91.5
94.3
92.9
85.2
77.2
72.9
88.3
92.8
89.7
30.8
69.2
85.7
52.9
90.0
58.9
89.1
Eh
mV
+ 167
+ 144
-183
-176
-208
-243
+ 252
+ 87
+ 122
+ 167
+ 174
- 88
- 43
+ 392
+ 97
+ 72
+ 52
+ 67
- 18
+ 67
+ 302
tSterilized controls had 1 g sugar added to them.
quantities found in the plot water sampled in
1973 occurred at the 48-hour sampling period,
which was preceeded by a 7.6 cm rain. Similar,
peak values noted at 40 and 96 hours in 1974
correspond to 8.6- and 0.5-cm rains, respectively.
(Jarbaryl was more evenly dispersed over the
first 4 sampling periods in 1974 due to cor-
responding rains incurred. The amount of car-
baryl in the plot water sampled in 1975 peaked
28 hours following application, then was rapid-
ly dissipated over the next 20 hours such that
the 48 hour samples did not differ significantly
from the 96 hour samples which were essential-
ly zero.
Analyses of variance indicated that time of
sample collection and application rate had a
highly significant influence on carbaryl levels
measured in the water for each of the years
tested. Residues of carbaryl were to be greater in
those under the intermittent irrigation scheme
at a 5% level of significance in 1973, and at a 1%
level of significance in 1974. Irrigation treat-
ment had no affect on carbaryl values observed
in 1975.
The 24,48, and 96-hour TLM concentrations
for each pesticide showed clearly that the insec-
ticides carbofuran and carbaryl were the most
toxic to fish (Table 15), molinate was the least.
Plot water concentrations never exceeded these
values in the permanent flood released at the
end of the growing season. Correspondingly, no
TABLE 12
§ JJsss
e 8?a ea
O O O O
II II
CmHnuom
-6-
192 384
Hours Following Application
Figure 15. Average concentrations of carbofuran
in rice paddy water sampled in 1973.
Concentrations of carbaryl in flood water
following its application in 1973, and
statistical significance with respect to time.
Treatment*
Block
Hours Following Application
0
24
4B
96
192
384
(Kg/ha!
I.Ri 1
2
•*ave
I,R, 1
2
-
ave
I,R, 1
I,R, 1
2
0.160
0.055
0.095
O.lOSa
0.119
0.013
0.171
0.121a
0.085
0.103
0.147
0.112a
0.119
0.161
0.513
0.264a
0.117
0.031
0.034
0.061a
0.073
0.120
QJJJ3
O.lOla
0.047
0.051
0.067
0.055a
0.121
0.115
0.109
O.llSa
0.270
0.124
0.321
0.238b
0.660
0.642
0.672
0.655b
0.563
0.3%
0.222
0.394b
0.815
0.694
1.004
0.851b
0.001
0.001
0.000
O.OOla
0.000
0.001
0.006
0.00 2a
0.001
0.002
0.000
O.OOla
0.573
0.000
0.000
0.191a
0.000
0.000
0.001
0.000
0.001
0.000
ojiSi
0.027a
O.OOO
0000
0.000
0000
0.000
0000
0.000
0.000
0.001
0.000
0.000
0.000
0.002
0.000
0.001
O.OOla
0.002
0.000
0.000
0.001
0.001
0.001
O.OOO
O.OOlc
•I, and I, indicate continuous and impoundment irrigation treatments, re-
spectively. R, and R, indicate recommended and excessive application rates,
respectively.
••Averages with different letter subscripts are significantly different at the
0.01 level.
163
-------
RETURN FLOW MANAGEMENT
fish died in the static bioassay test determined
on water collected just prior to its release.
However, plot water levels exceeded the
TLM value established for carbofuran up to 192
hours following the application at the
recommended rate and through the 384 hour
sampling period at the excessive rate in 1974
(Figure 16), and 1975 (Figure 17). Impact from
runoff within this time frame becomes a func-
TABLE 13
Concentrations of carbaryl in paddy water
following its application in 1974, and
statistical significance with respect to time.
Treatment*
I,R,
:
•
IA
.
3
••ave
-
3
ave
.
3
•ve
3
•ve
Hour* Following Application
•s —
0.196
0.090
.
0.113*
0.034
0.175
-
0333ab
0.104
0.146
0.090
0.113*
0.717
1.140
1.875
T244*
'
0.061
0.072
•
0.125
0.090
0.068
0094*b
0.039
0.015
0.020
0.025*
0.173
0.659
1.807
6~880tb
m
S
0.498
0.032
0.007
0.179*
1.291
0.255
0.206
0.584*
0.161
0.069
0.038
0089a
1.250
0.820
0.003
0.691b
~m
ha)
0.251
0.054
0.034^
0.113*
0.433
0.677
0.484
0.531*
0.496
0.272
0.082
0.288*
0.859
0.623
1.534
l.OOSab
£
0.000
0.001
0000
0.000.
0.000
0.015
0.000
O.OOSb
0.000
0.000
0.000
O.OOOa
0.002
0.000
0.473
0.158C
0.000
0.000
0.000
0.000
0.000
0.000
0.000
O.OOOb
0.000
0.000
0.000
0.000
0.000
0.000
0.001
O.OOOc
•I, and 1, indicate continuous and impoundment irrigation treatments, re-
spectively. R, and R, indicate recommended and excessive application rates.
respectively.
••Averages with different letter subscripts are significantly different at the
0.01 level.
TABLE 14
Concentrations of carbaryl in paddy water
following its application in 1975, and
statistical signficance with respect to time.
Treatment*
I R
-.
-.
IA
Houri Following Application
Rep
.
_JL
•••ve
.
1
•»«ve
1
.
3
•••ve
.
.
3
•••ve
.102
033
.029
055*b
079
.
.112
--
025
.'
.011
.02 1«
.220
.170
.066
.152b
21 28
490 038
.437
.056
328*
3.513
1.407
.937
1 950*b
.181
.249
.611
.347*
2.368
.871
4.476
2.572*
.232
127
132ab
7.72S
2.644
1.110
3832*
1.320
.024
.102
.4*2*
2.927
.927
.300
1.385*b
48
.065
. I
.090
.120*b
242
M
.201
.212b
.056
-
.134
.129*
.670
2.119
1.760
1.516*b
95
.000
.000
.000
OOOb
.000
•
.000
.OOOb
.000
>
.000
.OOOa
.006
.000
.625
.210b
tl, and I, indicate continuous and impoundment irrigation treatments, re-
specuvely R, and R, indicate recommended and excessive application rates.
••Average* with different Istfer subscripts are significantly different at the
0.01 level
tion of the quantity of diluent water at the point
of release. Although carbaryl exceeded the TLM
levels at various sampling periods, the rapidity
with which it was degraded certainly lessens its
impact as a potential pollutant. Data suggest
that propanil and molinate should be of little
concern at current recommended rates and
cultural practices employed.
The values of TOG, COD, and BOD
measured for plot water at the end of each
season did not vary significantly from that of
the canal water used to irrigate test paddies
(Tables 16, 17, and 18).
CONCLUSIONS
With normal flooded water management
practices, excessive amounts of water are lost to
runoff as a direct result of over-irrigation.
TABLE 15
The 24, 48 and 96 hour TLM concentrations
and their 95% confidence intervals in
filtered tap water in ppm.
1973
Pesticide
Prop*oil
Molina t*
C*rbofur*n
C*rbofur*D
C«rb«ryl
Flow
Static
Static
Static
Intermittent
Static
24-Hour
20.81
119.68-22.44)
33.25
(31.82-34.96)
>1.5
_
.56
I.50-.62)
6.71
(5.89-7.781
48-Hour
14.51
(13.33-15.65)
33.24
(31.82-34.961
1.42
(1.33-1.80)
.52
I.47-.58)
1.30
(1.24-1.40)
96-Hour
7.94
(6.99-8.851
33.24
(31.82-34.961
1.42
(1.29-1.70)
.51
I.46-.56)
1.30
(1.24-1.701
Hors Fofcwng Appkcotion
Figure 16. Average concentrations of carbofuran in
rice paddy water sampled in 1974.
164
-------
RETUKN FLOW FROM RICE PADDIES
Losses are greatest in continuous flow systems
and may be as great as 1 m during some seasons
in the area investigated.
Some pesticides degrade rapidly in the
paddy environment, while others persist for
several days or weeks after application. Rain-
fall tends to wash many of the foliar applied
chemicals off the plants into the water, causing
greater concentration in the runoff after rainfall
events than before. For each pesticide tested, the
half-life is given and recommendations have
been developed for retention times necessary to
reduce concentration below toxic limits.
SS 6
SX ?
o o o
11
oC; o o
X ContMxxMt RvcotMnandm}
O Continwoul Elc*MM«
R*co(nm«lcM
Hcxrs Following Application
Figure 17. Average concentrations of carbofuran in
rice paddy water sampled in 1975.
TABLE 16
Average TOC, COD and BOD of flood water and
canal water at the time of final drainage. No
significant differences between results were
found in any year.
1974
Treatment
TOC
mg/l
COD
mg/l
BOD
mg/l
Impounded irrigation
recommended rates of
pesticides and nutrients 28 61 2.2
Irrigation recommended
rates of pesticides and
fertilizers 25.3 57 1.6
Continuous flow irrigation
excessive rates of fertilizer
and pesticides 29 45.7 2.2
Continuous flow irrigation
excessive rates of fertilizer
and pesticides 26.7 45 1.6
Canal water 28 55 1.0
TABLE 17
Average TOC, COD and BOD of flood water and
canal water at the time of final drainage. No
significant differences between results were
found in any year.
1975
TOC COD BOD
Treatment mg/l mg/l mg/l
Impounded irrigation
recommended rates of
pesticides and nutrients 29 47 2.1
Irrigation recommended
rates of pesticides and
fertilizers 23 52 2.2
Continuous flow irrigation
excessive rates of fertilizer
and pesticides 27 57 1.7
Continuous flow irrigation
excessive rates of fertilizer
and pesticides 21 48 1.6
Canal water 21 47 2.3
Only a small fraction of the fertilizer
nutrients are lost in the runoff water even when
runoff volumes are large and excessive amounts
of fertilizer have been applied. The soils
researched appear to have a large capacity to
absorb the nutrients and peak concentrations
immediately after application decrease to
background levels with three to five days.
Irrigation water use efficiency could be
greatly improved with careful management
TABLE 18
Average TOC, COD and BOD of flood water and
canal water at the time of final drainage. No
significant differences between results were
found in any year.
Treatment
TOC
mg/l
COD
mg/l
BOD
mg/l
Irrigation recommended
rates of pesticides and
fertilizers 11 28 5.3
Continuous flow irrigation
excessive rates of fertilizer
and pesticides 9.3 25.4 2.0
Border plots intermittent
flow irrigation
recommended rates of
fertilizer and pesticides 13 29.5 6.4
Canal water 17 36 3.4
165
-------
RETURN FLOW MANAGEMENT
including the maintenance of small depths of shortly after a pesticide application would
flood and by withholding irrigation water at the overflow the levees.
end of the season. These practices would reduce
irrigation return flow and the decreased water nirxrr»WT irnr'lviF'MT
depthinthefieldsatanygiventimewouldallow ACKNOWLEDGMENT
more volume for storage and would decrease the This work was sponsored in part by fcFA
probability that a rainfall which may occur Grant # S802008.
166
-------
Evaluation of Surface
Irrigation Return Flows
in the Central Valley
of
KENNETH K. TANJI, JAMES W. BIGGAR, ROBERT J. MILLER
WILLIAM O. PRUITT and GERALD L. HORNER
Department of Land, Air and Water Resources, University of California,
Davis, California
Natural Resource Economics Division, Economic Research Service,
U.S. Department of Agriculture, Davis, California
ABSTRACT
The variability in the quantity and quality of
surface return flows from two typical irrigation
districts is reported for the 1975 and 1976
irrigation seasons. Factors contributing to such
variations are noted. Emphasis is placed upon
the concentration and mass emission of TDS
and SS. The implications of these findings are
discussed relative to PL 92-500 and best
management practices.
INTRODUCTION
The Great Central Valley (Sacramento and
San Joaquin Valleys) of California contains
about 6.86 million acres of irrigated croplands
out of the total 9.1 million acres statewide (3).
Field crops (irrigated pasture, alfalfa, cotton,
small grains, rice, sugar beets, corn) make up
about 70% of this acreage, tree and vine crops
(grapes, almond, walnut, peaches, prunes,
citrus) about 22%, and vegetable crops (tomato,
beans, potato) about 8% (3). On a statewide
basis, about 41% of the 9.1 million acres are
irrigated by the border method, 34% by furrow,
8% by sprinkler, and the remainder by drip and
subirrigation methods (3). The depths of applied
water in the Central Valley range from 3.2 to 4.9
ft for field crops, 2.8 to 3.7 ft for tree and vine
crops, and 2.4 to 2.7 ft for vegetable crops (3). It
has been estimated (2) that about 30% of the
applied water in the Sacramento River and San
Joaquin River Basins is discharged as surface
return flows.
Much attention is now being focused upon
irrigation return flows due to recent legislation
on water quality and pollution control, and the
concerns for water and energy conservation.
Figure 1 identifies the major water-flow
pathways in the crop root zone portion of
irrigated lands and the components of surface
return flows; namely, surface runoff (irrigation
tailwater, precipitation runoff, operational
spills from distribution systems) and collected
subsurface drainage (effluents from tile
drainage and drainage wells, and subsurface
waters intercepted by natural and man-made
open channels). With some exceptions, these
components of surface irrigation ruturn flows
are collected in the same drain regardless of
their origin: surface or subsurface, point or
nonpoint source. Some of these drain waters are
reused, either planned or incidental, at the site
of production or downstream, while the
remainder is discharged with no readily ap-
parent beneficial uses. It should be noted that
surface irrigation return flows in California are
nearly nonexistent where water is scarce
and/or expensive.
This report presents some data on the flow
and quality characteristics of surface irrigation
returns from two typical irrigation districts for
the 1975 and 1976 irrigation seasons.
Differences in the quantity and quality of return
flows are noted, with particular regards to total
dissolved solids (TDS) and suspended solids
(SS), and attempts are made to relate these
differences to site specific factors and con-
167
-------
RETURN FLOW MANAGEMENT
ditions. In addition, the information and data
are evaluated with regards to PL 92-500 and
potential best management practices.
OREGON
Applied Water
& Rainfall
Evapotranspiration
Surface
Runoff
Surface
Return Flow
Collected Subsurface
Drainage
Percolation
Figure 1. The major water-flow pathways in the
crop root zone portion of irrigated lands. Surface
return flow is comprised of surface runoffs and
collected subsurface drainage.
DESCRIPTION OF STUDY AREAS
Figure 2 gives a location map of the two
irrigation districts considered, Glenn-Colusa
Irrigation District (GCID) in the Sacramento
River Basin and Panoche Drainage District
(PDD) in the San Joaquin River Basin. GCID,
Figure 3, is a 163,700 acre district that obtains
its supply waters by diversions from the
Sacramento River and Stony Creek, and dis-
San
Francisco
GLENN-COLUSA
IRRIGATION DISTRICT
ANOCHE
DRAINAGE DISTRICT
PACIFIC
OCEAN
Figure 2. Location map pointing out two study
areas in the Great Central Valley of California.
charges its return flows via 11 lateral drains
into the 70-mile Colusa Basin Drain (CBD),
which has an outfall to the Sacramento River
above Knights Landing as well as to the Yolo
Bypass. PDD, Figure 4, is a 43,800 acre district
that receives imported water from the
Sacramento River Basin via the Delta-Mendota
Canal and the California Aqueduct. Its return
flows are discharged mainly into the
GLENN-COLUSA IRRIGATION DISTRICT
CIO 1
Figure 3. Map of the 163,700 acre Glenn-Colusa Irrigation District in the Sacramento River Basin o
California.
168
-------
SURFACE RETURN FLOWS - CALIFORNIA
PANOCHE DRAINAGE DISTRICT
Figure 4. Map of the 43,800 acre Panoche Drainage
District in the San Joaquin River Basin of Californi a.
Grasslands Water District (GWD) for reuse by
irrigated pastures and waterfowl habitats, and
ultimately discharged into the middle reaches of
the San Joaquin River. Other significant reuse
of return flows discharged into the Sacramento
and San Joaquin Rivers is also made by down-
stream users.
The major crop grown in GCID is rice (10%
of irrigated acreage in 1975 and 58% in 1976)
which is continuously flooded throughout the
growing season. Irrigation usually begins in
April and ends in October. The major crops
grown in FDD are cotton, tomatoes, and melons
which are furrow irrigated, and barley and
wheat which may receive some pre-irrigation or
irrigation depending upon winter rainfall. In
FDD irrigation is practiced year-round, except
in November and December when the usual
maintenance and repairs take place.
WATER QUANTITY AND QUALITY
Table 1 gives detailed analyses of the sup-
ply and drain waters based upon the average of
3 to 4 grab samples taken over each irrigation
season. The quality of the supply waters for
GCID is excellent while the drain waters have
not been degraded significantly. In contrast, for
FDD the supply waters are similar to the quality
of drain waters from GCID and its drainage
waters are substantially degraded with regards
to dissolved mineral salts, turbidity, nitrogen,
and boron. Table 2 describes differences in
quality characteristics for the several com-
ponents of collected surface irrigation return
flows relative to the supply water. The substan-
tial degradation of drain waters from FDD may
TABLE 1
Quality of supply and drain waters in Glenn-Colusa Irrigation District (GCID) and
Quality parameters
EC, micromhos/cm
Solutes in meq/liter
Na
Ca
Me
*•*&
HCOg
Cl
SO 4
Turbidity, JTU
BOD5,mg/liter
Nitrogen in mg N/liter
Total N
Org. N
NH4
NOq
J.-1W £
Boron, mg/liter
GCID, 1975
Supply Drain
117 418
0.22 1.92
0.50 1.22
0.40 1.39
0.99 3.00
0.07 0.48
0.08 0.85
9 25
3.0 2.6
0.92 1.31
0.87 1.03
0.03 0.07
0.14 0.19
p ;
GCID, 1976
Supply Drain
133 386
0.32 1.86
0.45 1.20
0.46 1.48
1.18 3.26
0.10 0.54
0.08 0.91
6 23
0.9 2.2
0.83 1.29
0.71 1.00
0.04 0.15
0.10 0.17
— —
FDD, 1975
Supply Drain
389 2.779
1.75 15.3
1.22 8.55
0.97 3.90
1.57 3.22
1.41 11.6
0.76 13.8
31 125
0.6 0.6
1.13 13.5
0.46 0.70
0.05 0.10
0.62 12.6
0.12 4.3
FDD, 1976
Supply Drain
473 2,891
2.09 16.7
1.16 8.58
1.05 4.75
1.71 3.03
2.01 12.1
0.91 14.7
24 180
1.1 1.2
1.43 17.1
0.75 1.44
0.04 0.19
0.64 13.4
0.11 4.4
169
-------
RETURN FLOW MANAGEMENT
TABLE 2
Collected surface irrigation return flow components
and their quality characteristics as related to
applied waters.
Quality Operational Irrigation Subsurface
parameters spills tailivater drainage
General quality
Salinity
Nitrogen
Oxygen
demanding
organics
Sediments
Pesticide
residues
Phosphorus
0 -
0 0, -1-
0 0. -, —
0 -. 0 0. -. --
0. ~. -
0 — 0, -, -
0 ~- 0, -, -
not expected to he much different than
some slight increase pickup or decrease
deposition may occur.
usually expected to be significantly higher
due to concentrating effects, application of
agricultural chemicals, erosional losses.
pickup of natural geochemical sources, etc.
= usually expected to be significantly lower
due to filtration, fixation, microbial de-
gradation etc.
be largely attributed to the collected subsurface
drainage which is affected by the presence of
native soil salts, boron and nitrogen, and the
production of sediments from the highly erodi-
ble Panoche soils (6).
Figures 5 and 6 show monthly flows and
concentrations of TDS and SS for supply and
drain waters in GCID and FDD, respectively
(6. 7). The 1975 irrigation season was in a
"normal" hydrologic year while the 1976 irriga-
tion season was in a "drought" hydrologic year.
The flows of supply and discharge waters in
GCID are dominated by flooded rice culture.
Rice is planted in late April to early May and
irrigated until late August to early September,
which is reflected by the flows in Figure 5. The
drain outflow is usually characterized by peak
flows in the early and late irrigation periods.
The former is due to lowering of water levels in
rice fields to prevent erosion of levees by gusty
winds that usually occur in May and the latter is
due to the draining of rice fields prior to
harvesting in mid-September and early Oc-
tober. The peak irrigation and drain outflow
months in FDD usually occur between March
and August for field and vegetable crops, but
water is delivered year-round because of less
precipitation in FDD than in GCID (7.73 and
10.65 inches in FDD during October through
September for 1974 / 75 and 1975/76 hydrologic
years vs. 19.56 and 6.58 inches in GCID, respec-
tively) and to some extent double cropping.
Figures 7 and 8 contain schematic
diagrams for supply and surface drain waters
for both the 1975 and 1976 irrigation seasons in
GCID and FDD, respectively. Included in these
diagrams are seasonal quantities of supply
water and their sources; gross quantities of
Cli«« - COIUSI »«ltlllO« DISIIICI
! ^
SOPPK •
•••r
1
4
i-
-•-f
h
—- *— I«SS
, '
n i -
LJ
El!)BI - COlllSl I««i6»»0» DISTtlCl
Dom otmm«
n n
j u
IDS
I
IBS
t
ss
»!/!
, [ « » 1 I I » S t « 1 | I ( *H * I ) I S 0 » 0
1975 W«
Figure 5. Monthly flows of supply and drain waters
in Glenn-Colusa Irrigation District over two irriga-
tion seasons and concentrations of total dissolved
solids and suspended solids.
; i » i • : . i s o « B I i F « " « i : '
Figure 6. Monthly flows of supply and drain waters
in Panoche Drainage District over two irrigation
seasons and concentrations of total dissolved solids
and suspended matter.
170
-------
SURFACE RETURN FLOWS - CALIFORNIA
drainage, part of which are captured and reused
within the district and remainder discharged;
irrigated acreages; unit application and unit
drainage rates (acre-feet/acre); and correspon-
ding masses (concentration x water volume) of
TDS and SS (tons); as well as unit mass
application and unit mass emission rates
(tons/acre). It is not possible herein to fully
discuss these data, but an attempt will be made
to point out the major findings.
In GCID, an increase of 7,800 acres of
irrigated lands occurred between 1975 and 1976
with a corresponding increase in diverted water,
but the unit application rate was slightly less
V
SAC RIVER STONY CREEK
711.300/135.300 1 V 124.100/30.700 1
SUPPLY
t
F
835,400/866.000 Af
131.793/114.243 1 TDS
27.267/14.133 T SS
631/61! AF/A
1.00/81 T/A TDS
0.21/0.10 T/» SS
243.500/183.800 AF
80.803/ 59.242 T TBS
11.922/12.498 T SS
C8D
184/1.31 AF/A
0-61/0.42 T/A TDS
0.09/0.09 T/A SS
RECAPTURE
*| 221.900/214.700
AC
DRAIN
465.400/398.500 AF
GLENN-COLUSA IRRIGATION DISTRICT 1975/1976
IRRIGATED ACRES 132.400/ 140.200
TOTAL ACRES 163100
Figure 7. A schematic diagram of seasonal surface
inflow and surface outflow of water, salts, and
sediments over two irrigation seasons in Glenn-
Colusa Irrigation District.
86.190/84.141 AF r-i 41.859/31634 AF
22.620/30.210 T IDS i 13.435/11.721 T TOS
12.053/6.637 T SS Si 1.082/870 T SS
SUPPLY
39627/37.061 AF
110.642/94.152 1 IDS
18.755/15.020 T SS CCID
6.904/7.971 »F
GWD <
PWD
I
1A55/4.3M Af
128.749/121.775 IF
38.369/41.931 T TDS
13.135/7.507 T SS
3.23/3.22 AF/A
0.96/1.11 T/A TDS
0.33/0.20 T/A SS
4VOB2/4U65 ftf
0.91/085 AF/A
2.53/2.54 T/A TDS
0-43/0-41 T/A SS
PANOCHE DRAINAGE DISTRICT 1975/1976
IRRIGATED ACRES 39.922/37.835
TOTAL ACRES 43.762
Figure 8. A schematic diagram of seasonal surface
inflow and surface outflow of water, salts, and
sediments over two irrigation seasons in Panoche
Drainage District.
(6.31 vs. 6.18 acre-feet/acre) due mainly to a
reduction of 12,136 acres in rice. In contrast,
there was a decrease of 2,870 acres of irrigated
lands in FDD from 1975 to 1976 and a correspon-
ding decrease in imported water, but the unit
application rates were nearly the same. Of the
465,400 and 398,500 acre-feet of drain water
produced in GCID in 1975 and 1976, respective-
ly, about half of it was captured and reused
within the district. The net discharge of surface
irrigation return flow was 29 and 21% of
diverted water in 1975 and 1976, respectively.
For comparison, FDD recaptured only 3.5 and
10.4% of their drain waters from the 1975 and
1976 irrigation season, respectively. Central
California Irrigation District (CCID) also used
16.8 and 19.3% of FDD's drain waters in 1975
and 1976, respectively. The remainder of the sur-
face return flows were discharged into the
Grasslands Water District (GWD). The unit
drainage rate was 0.91 and 0.85 acre-feet /'acre
for these two irrigation seasons in FDD as
compared to 1.84 and 1.31 acre-feet/acre in
GCID. These differences may be largely at-
tributed to irrigation application methods and
crop culture.
The monthly flow-weighted average con-
centration of TDS (mainly dissolved mineral
salts) in GCID's supply water was 116 and 97
mg/liter and drain water 244 and 237 mg/liter
for the 1975 and 1976 seasons, respectively,
about a 2.1- and 2.4-fold increase in concentra-
tion. In contrast, the TDS in FDD's supply
water was 219 and 254 mg/liter and drain water
2,053 and 1,868 mg/liter for the 1975 and 1976
seasons, respectively, and a 9.4- and 7.4-fold
increase. But, on a mass basis, Figure 7 for
GCID indicates only 61 and 52% of the TDS in
1975 and 1976 irrigation water, respectively,
were discharged into CBD while Figure 8 for
FDD indicates 288 and 225% for the correspon-
ding values. With regards to TDS, the increase
in concentration of TDS in drain water relative
to the supply water for GCID is in the usual
range of 2- to 4-fold increase from the
evapotranspiration concentrating effects by
crop plants. The remainder of salts brought in
by the irrigation water presumably is lost as
deep percolation since accumulation of salts in
soils is not observed. On the other hand, the 9.4-
and 7.4-fold increase in the concentration of
TDS in FDD's drain waters must be attributed
to the pickup of salts through chemical weather-
ing of native soil salts and minerals. Soil
analyses (6) from FDD show that EC e typically
171
-------
RETURN FLOW MANAGEMENT
ranges from 1.6 millimhos / cm in soils con-
taining large amounts of gypsum (greater than
10 T/AF). Results from a conceptual
hydrosalinity model (4) indicate gypsum con-
tributes about 25 meq/liter of salts (2,152
mg/ liter TDS) to the soil solution. Furthermore,
this computer model indicates if gypsum was
not present, the TDS of surface irrigation return
flow from FDD would be 458 mg / liter instead of
the monitored 2,053 mg/liter in 1975. For com-
parison, the computer predicted 1,818 mg/liter
in the presence of gypsum.
The monthly flow-weighted average con-
centration of SS (mainly sediments) in GCID's
supply water was 24 and 12 mg / liter and drain
water 36 and 50 mg liter for the 1975 and 1976
irrigation seasons, respectively. But the mass of
SS discharged into Colusa Basin Drain (CBD),
Figure 7, was 44 and 88% of that brought into
the district from the supply water. Likewise, SS
in FDD's supply water was 75 and 45 mg / liter
and drain water 348 and 298 mg/liter for the
1975 and 1976 irrigation seasons, respectively, a
4.6- and 6.6-fold increase in concentration.
However, on a mass basis, Figure 8, the SS
discharged into GWD and CBD was 1.4 and 2.0
times the amount brought into the district from
their supply water. The difference in unit mass
emission rate of SS from GCID as compared to
FDD may be attributed to the fact that flooded
rice fields in GCID more or less act as settling
basins for SS while the soils in FDD are highly
erodible (6).
DISCUSSION
The implications of the above evaluations
on concentration and mass emission of TDS
and SS are of considerable importance. While it
is recognized that impact of pollutants on the
beneficial uses of waters may be appraised in
terms of concentration, the effectiveness of
controlling discharge of pollutants should not
only be considered in terms of concentration but
also on a mass emission basis. If concentration
is overly stressed, this tends to promote inef-
ficient water management (5).
A survey (5,6) of typical ranges of
variations in the quality and quantity of surface
irrigation return flows along with the data
presented herein indicates the quantity and
quality of surface irrigation may be highly
variable and site specific. This particular con-
clusion should be considered for irrigation
return flows (point or nonpoint, collected or
uncollected) when contemplating future water
quality and pollution controls. Some of the
principal factors affecting how much irrigation
return flows are generated include: availability
and cost of supply water; irrigation application
methods and efficiencies; extent of reuse of
drain waters at the onfarm, district and basin
levels; soil physical conditions; and special
cultural practices such as the lowering of water
in flood rice fields during strong winds to
prevent erosion of levees to germination of seeds
in furrow beds. Some of the principal factors
affecting the quality of surface irrigation return
flows are: quality of supply water; presence of
native soil salts, boron and nitrogen; leaching
fraction and salt pickup-salt deposition
phenomena; use of agricultural chemicals such
as fertilizers and soil amendments; insecticides
and herbicides, animal manures and waste-
waters; crop and other plant residues; erodibili-
ty of surface soils and open drain channel
banks; sedimentation and resuspension of
suspended matter as affected by current veloci-
ty and size fraction; and discharges into irriga-
tion drainage canals by other sectors of society
including other agricultural activities,
municipal and industrial treatment plants,
silviculture, rangelands and watersheds.
In our opinion irrigated agriculture is a
unique institutional structure for which effluent
guidelines and "end of the pipe" controls
promulgated in PL 92-500 cannot be applied as
reasonably as in most other industries. It is our
contention that irrigated agriculture does not
have the "plumbing and valves" found in
industrial and sewage treatment plants.
Irrigated agriculture operates in an "open
system" in contrast to a "closed system" (as in
most other industries). Many of the natural
conditions and processes under which irrigated
agriculture operates are unmanageable or un-
controllable by man, or, at best, difficult to
control. Thus, if irrigated agriculture is to
comply with the intent of PL 92-500, this in-
dustry will require pollution abatement controls
and policies different from those of most other
industries.
It is suggested that due considerations
instead be given to "best management prac-
tices" rather than concentrating on waste dis-
charge requirements. A recent publication (1)
points out alternative control and management
practices for surface irrigation return flows.
Since irrigation tailwater is considered to be
easiest to control of the components of surface
172
-------
SURFACE RETURN FLOWS - CAIJFORNIA
irrigation return flows, tailwater management
appears to be a logical best management prac-
tice. However, the effects of tailwater discharge
into surface waters may be either beneficial,
detrimental, or both, depending on the quality
constituents of interest and water flow. Be that
as it may, for application to Central Valley
conditions, Figure 9 presents an array of alter-
native controls on tailwater production and
potential pollution of surface receiving water
bodies (5). Some of these tailwater management
practices may be broadly applicable, while
others are not because they require extensive
structural modifications and/or costly water
control. Due to the site variability of receiving
waters and of tailwater production and quality,
it is difficult to recommend any single universal-
ly applicable best management practice. A
practice that is effective in one location may not
be as effective in another.
FLIMINATION of TAILWATER DISCHARGE
LIMITED TAILHATER
DISCHARGE UNDER Low
STREAM FLOW CONDITIONS
(AUGME'IT/VAINTAIN FLOW,
REQUIRE HOLDING RESERVOIR)
INSTALL SEDIMENTATION
SUMPS AND DISCHARGE
IMPROVE IRRIGATION
APPLICATION EFFICIENCIES
TO ATTAINABLE LEVELS
LIMITED TAILWATER
DISCHARGE UNDER HIGH
STREAM FLOW CONDITIONS
(MINIMIZE IMPACT OF
POLLUTANTS^ REQUIRE
HOLDING RESERVOIR)
INSTALL TAILWATER
RECOVERY SYSTEM AND
REUSE AT FARM SITE
DISCHARGE TAILWATER AS
REQUIRED ?Y CULTURAL
PRACTICES (LOWERING
HATES IN RICE FIELDS,
DURING SEED GERMINATION
IN BEDS, ETC.)
CAPTURE USEASLE
IRRIGATION RETURN
FLOWS AT DISTRICT
AND BASIN LEVELS
o
Figure 9. An array of potential best management
practices for the control of irrigation tailwater and
potential pollution of surface receiving water bodies.
ACKNOWLEDGMENT
This investigation was supported mainly
by EPA Grant Nos. R 803603-01-1 and
R 803603-02-1. This particular phase of the
investigation was made possible through the
cooperation and support of Glenn-Colusa Irriga-
tion District and Panoche Drainage District.
REFERENCES
1. Carter, D. L., and Bondurant, J. A. 1976.
Control of sediments, nutrients, and adsorbed
biocides in surface irrigation return flows. EPA-
600/2-76-277, October, Environmental Protec-
tion Technology Series, US EPA, 53 p.
2. State Water Resources Control Board.
1975. Water Quality Control Plan Report,
Sacramento River Basin (5A), Sacramento-San
Joaquin Delta Basin (5B), and San Joaquin
River Basin (5C). Vol. II.
3. Stewart, J. I. 1975. Irrigation in
California. Report to the California State Water
Resources Control Board, University of Califor-
nia, 64 p.
4. Tanji, K. K. 1977. A conceptual hydro-
salinity model for predicting salt load in irriga-
tion return flows. In Proc. Intern. Conf. on
Managing Saline Water for Irrigation: Plan-
ning for the Future. 16-20 August 1976 (in
press).
5. Tanji, K. K., Biggar, J. W., Horner, G.
L., Miller, R. J., and Pruitt, W. O. 1976. Irriga-
tion tailwater management. Conclusions and
recommendations with regards to PL 92-500
and the NPDES Permit Program. Report to the
California State Water Resources Control
Board for Standard State Agreement No.
4091400, March, 58 p.
6. Tanji, K. K., Biggart, J. W., Horner,
G. L., Miller, R. J., and Pruitt, W. O. 1976.
Irrigation tailwater management. 1975-76 An-
nual Report to U.S. EPA on EPA Grant No.
R 803603-01-1, University of California at
Davis, Water Science and Engineering Paper
4011, April, 200 p.
7. Tanji, K. K., Biggar, J. W., Horner,
G. L., Miller, R. J., and Pruitt, W. O. 1977.
Irrigation tailwater management. 1976-77 An-
nual Report to U.S. EPA on EPA Grant No.
R 803603-02-1. (under preparation.)
173
-------
An Economic Analysis
of Irrigation Return Flow Recycle
Systems in the Central Valley
of California
WILLIAM KINNEY, GERALD L. HORNER, and KENNETH K. TANJI
Department of Agricultural Economics, University of California, Davis;
Economic Research Service, USDA;
and Department of Land, Air and Water Resources, University of California, Davis
ABSTRACT
Irrigation return flow recycle systems have
been used in California for a number of years
under a variety of crop rotations, soil
characteristics and qualities of irrigation water.
Recycle systems operating on the field, farm and
district level were compared on the basis of cost
in two areas of the San Joaquin Valley. The
effects of three scales of recycle systems on the
least-cost combination of irrigation water
delivery systems and water management
techniques were evaluated.
INTRODUCTION
In an annual progress report to the En-
vironmental Protection Agency, it was conclud-
ed that the production of tailwater occurred in
irrigated areas where water costs were low,
water supplies plentiful, high water use did not
affect yields and facilities for tailwater disposal
existed (Tanji, 1976, p. 18). The report also
recommended that the amount of tailwater
could be reduced by either adopting more ef-
ficient water application methods or managing
the present systems more carefully.
Public Law 92-500 has established water
quality goals that could require a substantial
reduction in the amount of disposed tailwater.
Therefore, determining the cost of alternative
methods of reducing tailwater is important in
selecting a least cost policy to achieve the goals
established in PL 92-500. This paper contains
an evaluation of farm and field level irrigation
methods and systems used in the San Joaquin
Valley that eliminate or substantially reduce
the amount of tailwater produced.
Irrigated agriculture uses about 85 percent
of the developed water in California and about
75 percent of the state's irrigation water is used
in the Central Valley. Field crops dominate the
Central Valley agriculture and about 90 percent
of them are irrigated by border and furrow
methods.
A reliable estimate of the amount of
tailwater produced in the Central Valley does
not exist. In 1975, the Central Valley Regional
Water Quality Control Board required many of
the irrigated areas to have a permit for the
disposal of irrigation return flows. The permit
requires the monitoring and reporting of the
flow, electrical conductivity and suspended
solids of supply and discharge waters. These
results will provide more insight into irrigation
efficiencies in the Central Valley and facilitate
a more comprehensive economic evaluation of
tailwater disposal.
CONTROLLING TAILWATER IN A
FIELD
Furrow irrigation typically produces runoff
in a field of 20 to 40 percent of the water applied
(Tanji, 1976, p. 16). Attempts to reduce the
stream of water to control runoff usually result
in poor water distribution and reduced crop
yields. The installation of a tailwater reuse
system can eliminate runoff by collecting the
tailwater in a sump at the lower end of the field
and pumping it back to the head of the field for
reuse. The reuse system allows maximum
stream flow, water penetration and high irriga-
tion efficiency.
175
-------
RETURN FLOW MANAGEMENT
Most field crops currently using border or
furrow irrigation methods could be irrigated
with some type of sprinkler system. Sprinkler
systems require more capital than non-
pressurized systems but are more efficient in
water application, require less labor and, with
good management, will reduce tailwater sub-
stantially.
The cost of restricting tailwater losses from
a 170 acre tomato field was estimated by com-
paring the costs of a conventional furrow irriga-
tion system with three systems that essentially
eliminate tailwater. They are a furrow irriga-
tion operation using gated pipe and a reuse
system, a handmove sprinkler system and a
side roll sprinkler system.
The conventional furrow irrigation system
utilized earth head ditches, canvas dams and
1-1/2 inch siphons. The field was leveled an-
nually and the cost per acre of irrigating the
tomato crop was $177.21 (Table 1). No reuse
system was employed and five acre-feet of water
per acre was applied to the crop.
The second system was also a furrow
system, but instead of earth head ditches, gated
pipe was utilized with a settling pond and a
reuse pump. The total amount of water used was
reduced from five acre-feet to 3.75 acre-feet per
year. This reduction in water use was measured
on a 170 acre tomato field located in the Panoche
Water District. The amount and quality of water
applied, tailwater and water reuse were
monitored in 1975. The tailwater, as a percent of
applied water, was 28.7 percent for the field and
24 percent for a specific 63 furrow section of the
field (Tanji, 1976, p. 84). Therefore, the water
recovery factor of 25 percent was used in
calculating costs for this system. Limited
resources and errors in other monitoring efforts
prevented replicating these measurements as
planned. However, these results seem to be
within the observed limits of other reuse
systems.
The cost of irrigating the tomato field with
this system was $150.25, or almost $27.00 less
than the conventional furrow system. Invest-
ment per acre increased from $7.74 to $249.00
per acre and labor decreased from $135 to $90
per acre.
The cost of irrigating the tomatoes by the
hand move sprinklers was $152.21. Although
the capital requirements were less than the
gated pipe-reuse system, operating costs were
slightly higher.
The side roll sprinkler system resulted in
costs of $127.79 per acre. Capital requirements
were high compared to the other systems but
only one hour of labor per acre per irrigation is
needed to operate the system. This is con-
siderably less than the other irrigation methods
considered.
TABLE 1
Summary of Irrigation Costs per acre for
170 acres of Tomatoes by Irrigation Method
Panoche Water District, 1976
Furrow with
Siphon* and no
Reuse System
Furrow with
Gated Pipes
and Reuse
System
Side
Roll
Sprinkler
Hand
Move
Sprinkler
Dollars Per Acre
Investment
Depreciation1
Interest
Taxes
Total Overhead
Operating Costs:
Irrigation Preparation
Water Requirement
Water Cost
Labor (J4.50/hr.)
Repairs
Power
Lateral Installation
and Removal
Total Operating Cost
Total Cost
7.14
.65
.31
.16
1.12
10.76
5.00AF
30.00
135.00'
.33
176.09
177.21
249.00
9.97
9.96
4.98
24.91
9.38
3.75AF
22.50
90.00
.34
3.12
125.34
150.25
291.37
19.39
11.67
5.83
36.89
3.75AF
22.50
45.00'
11.65
10.34
1.41
90.90
127.79
165.88
10.98
6.63
3.33
20.94
3.75AF
22.50
90.00'
6.64
9.35
2.78
131.27
152.21
'•Reed, A, D., Jewell L Meyer and Falih K. Aljibury, "Irrigation Costs," Cooper-
ative Extension University of California, Leaflet 2875, Davis, Aug. 1976.
*An 8 percent interest rate and straight line depreciation was used. Specific cost
budgets are presented in the appendix.
These data indicate that irrigation systems
designed to apply water efficiently and without
the production of significant tailwater is also
the least-cost alternative. These results
probably could be achieved in most other areas
of the Central Valley because the physical
conditions in the test area are fairly represent-
ative. The cost of water used in this analysis
was $6.00 per acre-foot which is relatively
cheap. The higher the cost of water the greater
the cost savings from using the more water
efficient systems. If water were free, a conven-
tional furrow system would compare more
favorably with the more efficient systems, but
the high labor requirement of operating siphons
and canvas dams result in high operating costs.
The internal rate of return was ap-
proximated for the three irrigation systems that
are capable of controlling tailwater. The
176
-------
ECONOMICS OF RECYCLE SYSTEMS
savings in irrigation costs as a result of using
the three systems was used as the annual
return. The internal rate of return for the gated
pipe and reuse system was approximately 18
percent; the side roll sprinkler system, 16 per-
cent; and the hand move sprinkler system, 16.5
percent. The investment in more efficient irriga-
tion systems appears to be profitable. However,
a relatively few California irrigators have
adopted field recycle systems or sprinkler irriga-
tion systems. The reasons farmers have not
widely adopted field recycle systems may be a
topic of further research.
MANAGEMENT OF DRAINAGE WATER
AT THE FARM LEVEL
The problem of managing irrigation return
flows in the San Joaquin Valley is complicated
by high water table conditions in many
locations (San Joaquin Valley Interagency
Drainage Program). In the Panoche Soil Con-
servation District, water table elevations have
been rising steadily for a number of years
(Stoddard & Kaner). The Enrico farm, located
within this district, has been recycling a mix-
ture of surface runoff and subsurface tile
drainage water since 1969. A normative linear
programming analysis was developed based on
the actual operations of the Enrico farm. The
objective of the LP problem is to derive cropping
patterns and water use that maximize returns to
land and management for the model farm,
subject to restraints on crop rotation, quantity
and quality of applied water, and resource
availability.
Model Formulation
The Enrico farm is 1,675 acres in size with
an entitlement to 5025 acre-feet of district water.
District water is taken from the supply canal
near the lower end of the farm where it may be
mixed with return-flows before the mixture is
pumped to the upper end for application. The
quality of applied water can be varied by using
different proportions of project and recycled
water.
The surface and subsurface return-flows are
collected in a District drainage ditch located on
the farm's boundary. In addition, due to uphill
discharges into this drain, the farm has access
to additional drainage water for recycling pur-
poses. In order to represent a variety of con-
ditions for available uphill runoff, the model
was run with levels ranging from 0 to 3000 acre-
feet.
Gross revenues for cropping activities were
determined using area average prices and yields
from the period 1970 to 1976. Net revenues were
then derived by subtracting typical crop produc-
tion costs (University of California Cooperative
Extension). Water activities were defined for the
application of project and recycled water. The
costs of running these activities were then
subtracted from crop revenues in the objective
function. The cost of pumping return-flows back
to the fields was assumed to be $3.50 per acre-
foot. Model solutions were derived using a range
of prices for project water, varying from $5.00 to
$25.00 per acre-foot.
Crop alternatives include tomatoes, wheat,
barley, cotton, alfalfa seed, sugar beets, and dry
beans. The crops were separated into three
classes according to their yield tolerance to the
boron concentration of applied water (Universi-
ty of California Committee of Consultants). For
each boron tolerance level, limits are set on the
maximum proportion of return-flows in applied
water that will avoid yield reductions. These
limits are based on flow-weighted average
boron concentrations as measured on the
Enrico farm during the 1975 and 1976 irrigation
seasons. District water had a flow weighted
average boron concentration of 0.1 mg/1, com-
pared to about 2.5 mg/1 for recycled water. Total
dissolved solids of these waters were low.
Therefore, considering the boron tolerance of
crops was sufficient in determing the yield
reduction from TDS (University of California
Committee of Consultants) solutions were also
derived using an assumed boron concentration
of 10 mg/1 in drainage water, to reflect con-
ditions where the subsurface component was a
higher proportion of total return-flows.
Water application and runoff rates are
based on values estimated by the California
Department of Water Resources. Maximum
acreage rectrictions were included to reflect
rotational and crop contract requirements.
Model Results
Optimal project and drainage water use,
cropping patterns and returns per acre for the
farm were derived for various water prices and
amounts of external drainage water supplied
(Table 2). Assuming a $5.00 District water cost
and 3,000 acre feet of district return flows,
drainage water use was 3,382 acre-feet, or 68
percent of total water use. If a $20.00 District
water cost is assumed, the amount of drainage
177
-------
RETURN FLOW MANAGEMENT
TABLE 2
Optimal Cropping Pattern and Drainage Water Use
with Zero, 1,000 and 3,000 Acre-Foot of Drainage
Water Available and Project Water Costing $5.00
and $20.00 per Acre-Foot. Enrico Brother's Farm,
Panoche Irrigation District.
District Water Costs/Acre-Foot
$5.00 $20.00 $5.00 $20.00 $5.00 $20.00
Drainage Water Available
(Acre-Foot)
Drainage Water Use
(Acre-Foot):
High Quality Blend
Medium Quality Blend
bow Quality Blend
Total
Percent of Total Water
Use
!>ramage Water
Discharged
Project Water Use
(Acre- Foot):
High Quality Blend
Medium Quality Blend
Total
Total Water Applied
1 Acre-Foot 1
Cropping Pattern (acres):
Tomatoes
Wheat
Cotton
Alfalfa Seed
Sugar Beets
Dry Beans
Total
Returns Per Acre
3,000
1,000
1,777
607
3,382
68
1.363
1,000
523
1.523
4,905
500
600
400
75
100
0
1,675
336
3,000
1,000
1,169
2.180
4.349
76
396
1.000
344
1.344
5,693
5OO
75
400
600
100
0
1,675
322
1,000
633
1,777
380
2,790
57
0
1,570
523
2.092
4.882
500
600
4OO
0
100
75
1,675
335
1,000
1.101
1.309
380
2,790
57
0
1,101
991
2.092
4,882
500
600
400
0
100
75
1,675
316
0
0
1,410
380
1,790
37
0
2,203
890
3,092
4.882
500
600
4OO
0
100
75
1,675
334
0
1.101
309
380
1.790
37
0
1.101
1.991
3.092
4,882
500
600
400
0
100
75
1.675
306
water used for irrigation increased to 4,349 acre-
feet, or 76 percent of total water use. Average
returns decreased from $336 to $322 per acre, as
a result of the higher cost of water and the
change in the cropping pattern from wheat to
alfalfa seed which has a higher tolerance to
boron. Total water application increased from
4,905 to 5,693 acre-feet.
If external drainage water supplies are
decreased below 3,000 acre-feet, changes in
project water costs do not affect the amount of
drainage water reused. With 1,000 and zero acre-
feet of extra drainage water available, respec-
tively, 57 percent and 37 percent of the total
applied water is composed of drainage water.
This represents total reuse of farm and off-farm
return-flows.
CONCLUSIONS
The amount of field tailwater can be re-
duced significantly and the annual costs of
irrigation slightly reduced by adopting reuse or
sprinkler application systems. However, the
high capital requirements of $165 to $300 per
acre may prevent the adoption of such tech-
niques. For example, installing a hand move
sprinkler system or a side roll system on 160
acres would require $26,540 and $46,620, respec-
tively. Although the annual irrigation costs are
lowered with these systems, the purchase of
such a system may not be the optimal invest-
ment if capital is limited. Greater returns may
be realized from investment in land or
machinery. More efficient use of irrigation
water would also result if water prices were
increased, or the quantity of water decreased.
The profitability of augmenting the farm
water supply with drainage water depends on
the costs of both District water and recycle
system. Generally, the higher the cost of irriga-
tion water the greater the value of drainage
water. Assumming a $20.00 cost of irrigation
water, returns could be increased by $16.00 per
acre by pumping an additional 2,604 acre-feet of
drainage water from the District's drainage
canal. The optimal water supply would consist
of 75 percent drainage water. At a $5.00 water
cost about 68 percent of the optimal irrigation
water supply would be drainage water.
178
-------
ECONOMICS OF RECYCLE SYSTEMS
APPENDIX
TABLE A-l
Per Acre Cost of Furrow Irrigating a 170 Acre Tomato Field with Siphons and No Reuse System.
Annual Cost
Cost
Cost
Acre
Life
Depreci-
ation
Interest Taxes
Total
Investment
Pitcher, 54"
Siphons, 1V2-300
Canvass, 12
Ditch Closer
Total Overhead
Dollars Per Acre
2,100
475
350
2,200
2.63
1.36
1.00
2.75
7.74
20
10
5
15
.13
.14
.20
.18
.65
.11
.05
.04
.11
.31
.05
.03
.02
.06
.16
.29
.22
.26
.35
1.12
Operating Costs:
Irrigation Preparation:
Fuel and Repairs
Labor (.757Acre)
Total Leveling
Ditch Open & Close:
Fuel & Repairs
Labor (.127Acre)
Total Ditch Costs
Total Irrigation Preparation
Irrigation: (10X)
Labor (30 hours/acre)
Water (5 Acre-Foot)
Repairs
Total Irrigation
Total Operating Costs
Total Irrigation Cost
6.02
3.36
.82
.54
135.00
30.00
.33
9.38
1.36
10.76
165.33
176.09
177.21
179
-------
RETURN FLOW MANAGEMENT
TABLE A-2
Annual Cost
Investment:
PVC Pipes:
15"
10" or 12"
Gated Pipes
Pump1
Settling Pond1
Total
Cost
10.560
3,300
18,744
4,000
3,250
Cost
Acre
66.00
21.00
117.00
25.00
20.00
249.00
Depreci-
Life
30
30
30
10
30
ation
Dollars Per
2.20
.70
3.90
2.50
.67
9.97
Interest
Acre
2.64
.84
4.68
1.00
.80
9.96
Taxes
1.32
.42
2.34
.50
.40
4.98
Total
6.16
1.96
10.92
4.00
1.87
24.91
Operational Costs:
Irrigation Preparation:
Leveling:
Fuel and Repairs
Labor (.757Acre)
Total Irrigation Preparation
Irrigation: (10X)
Labor (20.00 hours/acre)
Water (3.75 (Acre Foot)
Power for Rearculation
Fuel & Repairs
Total Irrigation
Total Operating
Total Irrigation Costs
6.02
3.36
90.00
22.50
3.12
.34
9.38
115.96
125.34
150.25
1 Source: Ray Ram, Manager, Panoche Irrigation District, Firbaugh CA.
180
-------
ECONOMICS OF RECYCLE SYSTEMS
TABLE A-3
Per Acre Cost of Irrigating a 170 Acre Tomato Field with a Side Roll Sprinkler System
Annual Cost
Cost Depreci-
Cost Acre Life ation Interest Taxes Total
Investment: Dollars Per Acre
Aluminum Pipe:
10" 4,642.70 27.31 15 1.82 1.09 .54 3.45
8" 3,478.20 20.46 15 1.36 .82 .41 2.59
Hydrant Valves 535.50 3.15 20 .16 .13 .06 .35
Pump 10,062.50 62.50 15 4.17 2.50 1.25 7.92
Laterals 30,395.00 178.13 15 11.88 7.13 3.57 22.58
291.37 19.39 11.67 5.83 36.89
Operating Costs:
Irrigation (10X)
Labor (10. hours/acre) 45.00
Water (3.75 AF) 22.50
Power 10.34
Repairs (4%) 11.65
Total Irrigation 89.49
Lateral Installation & Removal:
Fuel & Repair .71
Labor .70
Total Lateral Installation
& Removal 1.41
Total Operating Costs 90.90
Total Irrigation Costs 127.00
181
-------
RETURN FLOW MANAGEMENT
TABLE A-4
Per Acre Cost of Irrigating a 170 Acre Tomato Field with a Hand Move Sprinkler System
Annual Cost
Cost
Investment:
Aluminum Pipe:
10" 4,642.70
8" 3,478.20
Hydrant Valves 814.30
Pump 10,062.50
Laterals 8,639.40
Operating Costs:
Irrigation (10X)
Labor (20 hours/acre)
Water (3.75 AF)
Power
Repairs (4%)
Lateral Installation & Removal:
Fuel & Repairs
Labor
Cost
Acre
Life
Depreci-
ation
Interest Taxes
Total
Dollars Per Acre
27.31
20.46
4.79
62.50
50.82
165.88
15
15
20
15
15
90.00
22.50
9.35
6.64
1.40
1.38
1.82
1.36
.24
4.17
3.39
10.98
1.09 .55
.82 .41
.19 .10
2.50 1.25
2.03 1.02
6.63 3.33
128.49
3.46
2.59
.53
7.92
6.44
20.94
Total Lateral Installation
& Removal
Total Operating Costs
Total Irrigation Costa
2.78
131.27
152.21
REFERENCES
1. Tanji, K. K., J. W. Biggar, G. L. Homer,
R. J. Miller, and W. O. Pruitt, "Irrigation
Tailwater Management" First Annual Progress
Report for EPA Grant No. R 803603-01-1, U. C.,
Davis, Water Science and Engineering Paper
4011, April, 1976.
2. Reed, A. D., Jewell L. Meyer and Falik
K. AJjibury, "Irrigation Costs", University of
California Cooperative Extension, Leaflet 2875,
Davis, Aug. 1976.
3. California Department of Water Re-
sources, Vegetative Water Use in California,
1974, Bulletin No. 113-3, The Resources Agency,
State of California, April, 1975.
4. Ayers, Robert S., and Roy L. Branson,
"Water Quality Guidelines for Interpretation of
Water Quality for Agriculture", University of
California Cooperative Extension, January 15,
1975.
5. San Joaquin Valley Interagency Drain-
age Program, Program Review Report, U.S.
Bureau of Reclamation, California State Water
Resources Control Board, March, 1976.
6. Stoddard Howard and Henry Karrer,
"Drainage Investigation and Master Drainage
Plan Panoche Conservation District", Stoddard
and Karrer Civil Engineers, Los Banos, June,
1968.
182
-------
On-Farm Methods for
Controlling Sediment
and Nutrient Losses
D. W. FITZSIMMONS, C. E. BROCKWAY, J. R. BUSCH, G. C. LEWIS,
G. M. McMASTER and C. W. BERG
Agricultural Engineering Department, Agricultural & Civil Engineering Departments,
Agricultural Engineering Department, Plant & Soil Science Department and
Agricultural Engineering Department, respectively,
University of Idaho, Moscow, Idaho
ABSTRACT
Field experiments were conducted in the
Boise and Magic Valley areas of southern Idaho
during the 1975 and 1976 irrigation seasons to
evaluate on-farm methods for controlling sedi-
ment and nutrient losses from irrigated fields.
In one experiment, four 1.7 haplots of field corn
were continuously monitored during the two
irrigation seasons. Varying furrow stream sizes
and stream cutback methods were used to apply
the same total amounts of irrigation water to
three of the plots. No control measures were
used on the fourth plot. Runoff from two of the
first three plots was run through a vegetated
buffer strip at the lower endofthe field and then
a settling basin. Runoff from the other two plots
was run through a basin only. The effects of
each treatment and control measure on the
quality and quantity of return flow from these
plots are presented.
Sediment yield from bean field plots which
were subjected to three different levels of
cultivation was evaluated during the 1975 and
1976 seasons. The effects of zero, one and two
cultivations on seasonal sediment yield, return
flow quantities and phosphate loss were deter-
mined. During the 1976 irrigation season, sur-
face runoff from the bean field plots was passed
through three sizes of mini-basins at the lower
end of the field. The basins were six feet wide
and collected runoff from three, four and five
furrows. The efficiency of duplicate mini-basins
was evaluated with and without grassed
overflow sections on the drain ditch bank.
The results of sediment and nutrient yield
determinations on several other field sites in the
two study areas are also presented. The effects
of various combinations of soil types, crops,
irrigation practices and sediment retention
devices on sediment and nutrient losses from
these sites were evaluated.
INTRODUCTION
The importance of irrigation in producing
crops in arid and semi-arid regions of the world
is well documented. This practice is particularly
important to western areas of the United States
where irrigation is required for the production of
most crops. While beneficial in many respects,
irrigation can have serious detrimental effects
on water quality since irrigation return flows
may contain dissolved salts, sediment, plant
nutrients, and other materials. The effects of
these materials on the quality of receiving
waters are a major concern in the western states
since irrigation return flows constitute a large
portion of the flow in many streams in this
region.
The major water quality problem in most
western irrigated areas is increased salt concen-
trations in the return flow as a result of the
pickup of minerals from the soils and the loss of
water through evapotranspiration (6). As a
result, a number of studies have been conducted
to find ways to decrease the concentration of
salt in return flows in these areas. Salinity is not
a major problem, however, in many irrigated
areas. Instead, the major problem is sediment in
the surface runoff from irrigated fields; and, to a
lesser extent, nitrate and other nutrients in
ground and surface waters in these areas.
183
-------
RETURN FLOW MANAGEMENT
There are a number of ways in which
sediment and nutrient losses from irrigated
areas can be effectively controlled. These in-
clude the use of irrigation and/or other practices
which eliminate or reduce surface runoff and
resulting pollutant losses from individual fields
and on-the-farm removal of sediment and
nutrients from the surface runoff from in-
dividual fields. Field investigations to deter-
mine the effectiveness of several on-farm
methods for controlling sediment and nutrient
losses from fields were conducted in two
irrigated areas in southern Idaho during the
1975 and 1976 irrigation seasons. Data obtained
from these investigations are presented and
discussed in the following paragraphs.
STUDY AREAS
The field investigations were conducted in
two major irrigated areas in southern Idaho, the
Boise Valley in southwestern Idaho and the
Magic Valley in southcentral Idaho. Both of
these areas have a dry, temperate climate which
is characterized by cool, wet winters and warm,
dry summers. Broad expanses of nearly level or
gently sloping lands in these areas are suitable
for intensive cultivation. Most of these lands are
used for growing irrigated row crops, grain, and
forage crops.
The Boise Valley lies in an upland plain of
unconsolidated lacustrine and fluvial materials
deposited by the Snake and Boise Rivers.
Terraces of stream-laid and lacustrine deposits
rise stepwise above the Boise River. In several
areas, local basalt ridges rise as much as 122
meters above the bottom land. The soils in the
area are highly variable and complex. Those
over the alluvial materials are generally quite
variable in depth. Duripans in many of the soils
restrict the downward movement of water and
plant roots. High water tables are common in
many valley areas.
Most of the farms in the Boise Valley are
gravity irrigated, primarily through the use of
siphon tubes and furrow methods. Sprinkler
irrigation is practiced in the valley, although its
use is not very extensive at present. Irrigation
return flows are collected by a complex network
of open ditch drains. Many of these drains
return all or a portion of their flow to the
delivery system for reuse. A number of crops
including alfalfa and clover for seed and hay,
grain, field corn, sweet corn, beans, potatoes,
sugarbeets, and onions are grown in the valley.
Hops, mint, and several other specialty crops
are also grown and some of the area is used for
orchards.
Most of the irrigated land in the Magic
Valley lies in a region designated as the Snake
River Plain aquifer. Land north of the river is
underlain by the Snake Plain aquifer. Land
south of the river slopes northward from
foothills and is underlain by basalts and
rhyolites at varying depths. The major soils in
the area are of loessal origin with calcareous
loamy subsoils. The silt loam soils are generally
underlain by a calcareous hardpan at a depth of
60 to 90 cm. This cemented layer is permeable
but restricts root penetration.
Water for irrigation is obtained from the
Snake River and is conveyed to farms in the
valley through unlined canals. Water applica-
tion is predominantly by gravity systems
through the use of siphon tubes and furrows.
The use of gated pipe is increasing and sprinkler
irrigation from ground or surface water sources
is common in the area. The use of pumpback
systems is also increasing. The principal crops
grown in the area are sugarbeets, beans, grain,
alfalfa, and potatoes.
FIELD INVESTIGATIONS
Most of the field investigations in the Boise
and Magic Valley areas were carried out on
gravity-irrigated fields at University of Idaho
Research and Extension Centers at Caldwell
and Kimberly respectively. A few investiga-
tions were also conducted on cooperators' fields
in the two study areas.
Boise Valley Investigations
Four field plots were monitored during the
1975 and 1976 irrigation seasons to determine
the effects of different irrigation water manage-
ment practices and on-farm sediment retention
devices on the quantity and quality of the
surface runoff from these plots. Descriptions of
the plot sites are given in Table 1. Sites 2,3, and
4 were located adjacent to each other in a large
field at the Caldwell Center. Site 1 was located
in a cooperator's field on land adjacent to the
Caldwell Center. The soil at all four sites is a silt
loam which is underlain by a cemented hardpan
at a depth of 75 cm or less. The erosion hazard
for this soil is moderate to high. Field corn was
grown on all four plots, both in 1975 and 1976.
184
-------
CONTROL OF SEDIMENT LOSSES
TABLE 1
Boise Valley Field Plot Site Descriptions
Site Year
1
2
3
4
1975
1976
1975
1976
1975
1976
1975
1976
Run
Area Length
(ha) (m)
1
1
1
1
1
1
1
1
.76
.76
.89
.78
.84
.74
.76
.66
237
237
345
345
327
327
321
321
Average Sediment
Slope Retention
(%) Device(s)
2.6
2.6
2.7
2.7
2.9
2.9
2.9
2.9
Pond
T-Slot & Pond
Pond
Grain Strip & Pond
Grass Strip & Pond
Grain Strip & Pond
Grass Strip & Pond
Grain Strip & Pond
As indicated in Table 1, the surface runoff
from each plot was diverted through one or more
sediment retention devices. In 1976, for exam-
ple, the runoff from Site 1 flowed through a
series of small T-shaped Basins (T-Slots) before
entering a small sediment pond. These basins
were installed at 20-meter intervals upstream
from the pond. In 1975, the runoff from Sites 3
and 4 passed through a grass strip before it
entered a sediment pond. Densely seeded barley
was used as a buffer strip at Sites 2, 3, and 4 in
1976.
All flows entering each plot or sediment
retention device were continuously monitored
during the 1975 and 1976 irrigation seasons.
Weirs and flumes equipped with stage recorders
were used to obtain continuous records of these
flows. Water samples were taken from the
headwater and surface runoff from each plot
and from the flow leaving each sediment reten-
tion device at regular intervals during each
irrigation. Data obtained from this monitoring
program were used to establish water and mass
balances for each site. The components of the
mass balances included sediment, nitrate, and
total phosphorus. The balance results were
summarized for each irrigation, each set within
an irrigation, and the entire season.
Magic Valley Investigations
Field and plot studies were conducted at five
locations in the Kimberly area to evaluate
sediment and phosphate reduction for alter-
native management practices and sediment
removal systems. Included were studies of
(1) cultivation and water management prac-
tices (tillage study), (2) mini-basins,
(3) vegetated buffer strips, (4) sediment removal
by a dense growing crop, and (5) furrow cutback
irrigation. In making these studies, weirs and
flumes were used to make the necessary flow
measurements. Water samples taken at ap-
propriate stations were used to determine the
concentrations of sediment in the flows. In
general, the flow and concentration data were
used as input to a computer program for
determination of flow volumes and sediment
yields for each study site for each irrigation.
RESULTS AND DISCUSSION
Boise Valley Investigations
Water, sediment, and nutrient losses from
the study sites were affected by the water
management practices and sediment retention
devices used at these sites. Evaluations were
made to determine the effects of different furrow
stream sizes, preplant irrigations, irrigation
scheduling, vegetated buffer strips, and sedi-
ment retention ponds on these losses.
Furrow Stream Size
The plots at Sites 2, 3, and 4 were irrigated
five times during the 1976 irrigation season
with different sizes of furrow streams. Stream
size and pollutant loss data for each site are
given in Table 2. A stream of 0.767 liters per
second was used to irrigate Site 2 during the first
three irrigations. This stream was reduced to
approximately 0.6 liters per second during the
last two irrigations to limit runoff. At Site 3, the
furrow stream was cutback to one-half its initial
size after 12 hours during the first three
irrigations. During the last two irrigations, the
TABLE 2
Results of Boise Valley Furrow Stream Size
Study, 1975 and 1976
Site Irrigation
2
3
4
2
3
4
2
3
4
2
3
4
2
3
4
1
1
1
2
2
2
3
3
3
4
4
4
5
5
5
Set
Length
Ihr)
24
36
48
24
36
24
24
36
48
24
24
24
24
24
24
Stream Net Sediment
Size Loss
Us) fkghal
0.767
0.767.0.384'
0.396
0.767
0.767,0.384
0.407
0.767
0.767.0.384
0.407
0.599
0.599
0.609
0.599
0.600
0.643
2456
1647
-656'
395
744
-508
183
155
-140
200
142
69
24
105
123
Net Total
PLoss
(kg'hal
1.42
1.36
-0.17
0.04
0.60
-0.06
-0.06
0.29
-0.19
0.15
0.01
-0.07
-0.03
0.0
0.01
Net Nitrate
Loss
(kg/ha)
0.09
0.07
-0.18
-0.26
-0.15
-0.33
-0.37
-0.32
-0.35
-0.06
-0.20
-0.45
-0.50
-0.37
-0.40
'Stream was cutback to smaller flow rate after 12 hours.
'Negative numbers indicate a net gain.
185
-------
RETURN FLOW MANAGEMENT
furrow stream was reduced to approximately 0.6
liters per second but was not cut back. A stream
of approximately 0.4 liters per second was used
to irrigate Site 4. This stream had to be in-
creased to over 0.6 liters per second during the
last two irrigations to get the wetting front to
advance properly.
The sediment loss during the first irrigation
of the season is usually greater than it is during
subsequent irrigations (5). This was the case for
Sites 2 and 3. However, the sediment loss from
Site 3 (with cutback) was almost one third less
than the loss from Site 2 (without cutback)
during the first irrigation. The sediment losses
from both sites were much less during the last
four irrigations than they were during the first
irrigation. The seasonal totals (Table 3) show
that the net sediment loss from Site 3 (with
cutback) was over 16 percent less than the net
loss from Site 2 (without cutback). The use of a
small stream size on Site 4 resulted in a net gain
of sediment on this plot during the first three
irrigations. However, the sediment losses dur-
ing the last two irrigations were nearly the same
as those from the other two sites. The overall
effect of using a reduced stream size during the
first three irrigations is shown by the seasonal
total which indicates a net sediment retention of
over 1,100 kg ha on this plot.
TABLE 3
Seasonal Totals for the Boise Valley Furrow Stream
Size Study, 1975 and 1976.
Site
2
3
4
Net Sediment
Loss
(kg/hat
3258
2793
-1112
Net Total
PLoss
(kg/ha)
1.52
2.26
-0.48
Net Nitrate
Loss
(kg/ha)
-1.10'
-0.97
-1.71
'Negative numbers indicate a net gain.
The runoff from all three plots was sampled
after it had passed through a vegetated buffer
strip and had flowed to the sampling point in a
ditch which had a slope of 0.5 percent. As a
result, a portion of the sediment retention
shown for these sites occurred in the buffer strip
and in the drainage ditch.
Total phosphorus losses from Sites 2 and 3
were minimal. The net retention of total
phosphorus on Site 4 is related to the net sedi-
ment retention, since total phosphorus losses
are associated with sediment losses (2, 4). The
net retention of nitrate at all three sites was due
to the fact that nitrate is infiltrated into the soil
writh the irrigation water. Similar results have
been obtained at other sites in the Boise
Valley (1, 3).
Prepiant Irrigation
Many irrigators apply a preplant irrigation
regardless of the soil moisture status. In order to
evaluate the effects of this practice, preplant
irrigations were applied to Sites 2, 3, and 4 ten
days before the 1976 crop was planted. Water
and sediment loss data for these irrigations are
summarized in Table 4. The data show that a
large amount of sediment was lost from each
plot even though the runoff was not excessive.
Over one-third of the seasonal sediment loss
from Sites 2 and 3 occurred during the preplant
irrigation. The amount of sediment lost from
Site 4 during the preplant irrigation was nearly
equal to the net amount returned on that plot
during the remainder of the irrigation season
(Table 3).
TABLE 4
Effects of Preplant Irrigations on Water and
Sediment Losses, 1976
Water Runoff
Site
2
3
4
Prepiant
Irrigation
(mm}
41
52
69
Percent of
Seasonal
Runoff
16
23
36
Net Sediment Loss
Prepiant
Irrigation
(kg/ha)
1773
1657
1188
Seasonal1
Loss
fhg/ha)
5031
4450
76
Percent of
Seasonal
Loss
35
37
1563
'Seasonal total includes preplant irrigation loss.
Irrigation Scheduling
Irrigation scheduling can have an effect on
sediment and nutrient losses from a field as well
as an effect on crop yields (1). If a field is over-
irrigated or irrigated too frequently, the
resulting surface runoff will be excessive and
will carry excessive amounts of sediment and
other pollutants from the field. Reducing the
number of irrigations and the amount of water
applied will usually reduce the losses from a
particular field. This was the case for Site 1. The
number of irrigations at the site was reduced
from eight in 1975 to six in 1976, partially as a
result of a 15 percent decrease in the consump-
tive use requirements of the crop. As shown in
Table 5, reducing the number of irrigations
186
-------
CONTROL OP SEDIMENT LOSSES
resulted in a 21 percent reduction in the
amount of water applied, a 31 percent reduction
in the amount of surface runoff, and 25 percent
reduction in the net amount of sediment lost
from the plot.
TABLE 5
Effects of Number of Irrigations on Water
and Sediment Losses
Consumptive Water Water-Use'' Surface Net Sediment
Numberof Use Applied Efficiency Runoff Loss
Year Irrigations (mm! (mml {%) (mml (kg'ha)
1975 8
1976 6
510
432
1270
1000
40.1
43.2
376
261
26.565
19.923
'Percentage of water applied going to consumptive use.
The water-use efficiencies (percentage of
water applied used to meet the consumptive use
requirements of the crop) were quite low both
years. However, these efficiencies are about as
high as can be expected for a furrow system
operating under the soil, slope, run length, and
other conditions present at this site. They could
be increased by the installation and use of a
pumpback system.
Vegetated Buffer Strips
In 1975, a sodded grass strip which was
1.83 m wide and 61.0 m long was laid in the
drainage ditch at the lower end of Site 3. The
runoff from Site 3 entered the strip from the side
while that from Site 4 entered the strip from the
upper end and ran the entire length of the strip.
Samples of the runoff from both sites were
collected downstream from the strip. The runoff
from the control plot, Site 2, did not run through
the strip.
Dense bluegrass in the strip was well es-
tablished and approximately 2.4 cm high just
prior to the first irrigation. The strip did an
excellent job of reducing the velocity of flow in
the drain ditch during the first irrigation. As a
result, it was completely inundated with sedi-
ment and was not effective during the
remainder of the irrigation season. The effec-
tiveness of the strip during the first irrigation is
shown by the data in Table 6. About 47 percent
of the net seasonal sediment loss from Site 2
(without strip) occurred during the first irriga-
tion; whereas, only about 19 and 8 percent of the
losses from Sites 3 and 4 (with strip), respective-
ly, occurred during the first irrigation. Due to
the ineffectiveness of the grassed strip, the net
losses during the last four irrigations were
essentially the same for all three sites.
In 1976, a 2.44 m wide strip of barley was
seeded across the lower ends of the three plots at
a seeding rate of approximately 100 kg/ha. The
grain was well established when the irrigation
season was started and was effective in retard-
ing the furrow runoff and causing sediment to
settle out of the runoff before it entered the drain
ditch running parallel to the strip. The effec-
tiveness of this strip is shown by the data
presented in Table 6. Site 2 was irrigated nearly
the same in 1976 as it was in 1975. Most of the
difference in the total net sediment loss
(3258 kg/ha in 1976 as opposed to 7038 kg/ha in
1975) can be attributed to the effect of the buffer
strip used in 1976. Although the net loss from
Site 2 during the first irrigation in 1976 was
nearly as great as the loss during the first
irrigation in 1975, use of the grain strip resulted
in a marked decrease in the net sediment losses
during the last four irrigations. The totals in
Table 6 indicate that the grain strip used in 1976
was more effective than the grassed strips used
on Sites 3 and 4 in 1975.
TABLE 6
Effects of Vegetated Buffer Strips on
Sediment Losses
Site
2
3
4
2
Year
1975
1975
1975
1976
Strip
Type
none
grass
grass
grain
Irrigation
Numberis)
1
2-5
total
1
2-5
total
1
2-5
total
1
2-5
total
Net Sediment
Loss
(kg/ha)
3312
3726
7038
740
3149
3889
367
4196
4563
2456
802
3258
Sediment Retention Ponds
Sediment retention ponds were used to
remove a portion of the sediment and
phosphorus from the surface runoff from the
four plots before it flowed from the study sites.
In 1976, six small T-shaped basins (T-slots) were
also installed in the drain ditch upstream from
Site 1. Four of the T-slots had a capacity of
187
-------
RETURN FLOW MANAGEMENT
approximately 2 m^ while the other two had a
capacity of about 0.5 m-^. The six T-slots were
filled with sediment during the first irrigation
and were ineffective thereafter. The effec-
tiveness of the T-slots and pond during the first
irrigation is shown by the data presented in
Table 7.
TABLE?
Sediment and Phosphorus Removal by T-Slots and
Sediment Pond at Site 1 During the First
Irrigation, 1976
Constituent
Tot«l
Phosphorus
fwdimwit
Into
T-Slots
/kg Hal
0.73
1055
Into
Pond
Ikghol
033
504
Out of
Pond
(kg ho)
024
189
Rtmovol b\
T-Slots
t<*>
45
48
Removal bv
T-Slots
A Pond
<*•>
78
66
The sediment retention ponds were quite
effective in reducing the net sediment losses
from all four field sites. Seasonal sediment
balances for the four field plots and the ponds
are given in Table 8. The sediment removal
efficiency of the ponds ranged from 46 to 82
percent. One reason for the low removal efficien-
cy of the pond at Site 1 in 1975 was the small size
of this pond. It was essentially filled with
sediment by the first two irrigations in 1975.
With the added retention of the T-slots in 1976,
the pond did not fill with sediment until after the
third irrigation. Even though the pond was
ineffective for over half of the irrigation season
each year, it did greatly reduce the net sediment
loss from this site.
All of the runoff from Sites 2, 3, and 4 was
diverted through the same pond. The data in
Table 9 show that the sediment removal ef-
ficiency of this pond was quite high both years.
This pond was large enough that it did not
require cleaning after the 1975 season and was
only about half filled with sediment at the end of
the 1976 season. Of course, the irrigation man-
agement practices and vegetated strips de-
scribed previously helped reduce the amount of
sediment entering this pond.
Magic Valley Investigations
The results obtained from the Magic Valley
field investigations also show that water and
pollutant losses from gravity-irrigated field
sites were significantly affected by the manage-
ment practices and sediment retention devices
used at these sites. Like the results of the Boise
Valley investigations, they indicate the need for
integrating the use of management practices
with the use of sediment retention devices to
minimize losses.
Tillage Study
An evaluation of the effects of cultivation
and irrigation practices on water and sediment
losses from a dry bean field was made in 1975
and 1976. A 1.5-hectare field at the Kimberly
Center was used for this study. The soil at this
location is Portneuf silt loam with field slopes
varying from 0.8 to 1.2 percent. Two methods of
preplant irrigation were used to prepare the field
for planting. In one case, a "broadcast" pro-
cedure was used to wet the entire field surface.
TABLE 8
Seasonal Sediment Balances for the Field Plots and Sediment Ponds, 1975 and 1976
Net Loss
Site
1
2
3
4
Year
1975
1976
1975
1976
1975
1976
1975
1976
On Plots
(kg/haj
621
482
1,051
1,004
1,352
775
1,652
1,787
Off Plots1
(kg/ha)
27,146
20,405
8,088
4,262'
5,241
3,568s
6,215
675'
Out of Pond
(kg/ha)
14,710
6,273
1,425
770
934
888
1,232
261
Removal by
Pond
(%>
46
69
82
82
82
75
80
61
w/o Pond
(kg/ha)
26,525
19,923
7,038
3,258
3,889
2,793
4,563
-1,112
with Pond
(kg/ha)
14,089
5,791
374
-234s
-418
113
-420
-1,526
'Includes effects of T-slots and vegetated buffer strips.
'Does not include preplant irrigation.
'Negative numbers indicate a net gain.
188
-------
CONTROL OF SEDIMENT LOSSES
In the other, a "banded" procedure was used to
wet only the seedbed area. After planting, the
two preplant irrigation treatments were split
into three cultivation treatments which were
cultivated 1, 2, or 3 times in 1975 and 0, 1, or 2
times in 1976. Each of the 30 plots in the field
was irrigated at 8 to 12-day intervals. A total of
eight irrigations were applied in 1975. The 1976
experiment was terminated after a hailstorm
destroyed the crop during the sixth irrigation.
The results from the 1976 experiment show
that significant differences in sediment yield
occurred between the two preplant irrigations.
In fact, the yield from the broadcast treatment
was 1,050 kg/ha while that from the banded
treatment was only 260 kg/ha. This difference
is to be expected since the furrow spacing for the
broadcast treatment was much less than that
for the banded treatment (11.8 versus 18.9 cm).
As a result, the runoff and resulting sediment
losses per unit area from the broadcast treat-
ment were greater than those from the banded
treatment. The results also show that field slope
had a significant effect on sediment yield. This
is shown by the regression equations presented
in Figure 1 for the five 1976 irrigations. As can
be seen, the sediment yield from the steeper field
slopes apparently increased as the irrigation
season progressed.
Statistical analyses of the 1975 and 1976
data show that there was no significant correla-
tion between the number of cultivations per-
formed and total sediment yield. Also, the
irrigation application efficiency (percentage of
water applied which is retained on a field) was
not significantly correlated with the number of
cultivations. The data do show, however, that
there is an inverse relationship between the
seasonal gross sediment yield from the bean
field and irrigation application efficiency. A
plot of the gross sediment yield versus applica-
tion efficiency for the 13 irrigations performed
in 1975 and 1976 is shown in Figure 2. This
relationship illustrates one of the basic concepts
in reducing sediment losses from gravity-
irrigated fields. Generally, any practice which
increases irrigation efficiency results in a
decrease in sediment losses.
Mini-Basins
In 1976, a mini-basin study was conducted
on a 1.38 hectare bean field adjacent to the
tillage study field. A total of 12 basins were
installed by construction of berms perpen-
dicular to the bank of the drain ditch along the
lower end of this field to form small shallow
ponds. The basins were varied in size so as to
collect the flow from 3, 4, or 5 furrows and were
constructed with either plastic or grassed
overflow sections. The overflow sections were
0.5 wide and were leveled to provide uniform
flow over the section. Five irrigations were
monitored to obtain data for determining the
sediment removal efficiency of the basins.
The data obtained from this study are
summarized in Table 9. The seasonal sediment
loss from the field, as indicated by the average
of the check furrow losses, would have been
14,950 kg/ha. The weighted average loss for the
mini-basins was only 680 kg/ha. The average
seasonal sediment removal efficiency of the
basins was 94.9 percent.
There is apparently no significant differ-
ence in the sediment removal efficiency of the
Field Slope X
Figure 1. Regression equations for field slope
on gross sediment yield for five irrigations of a
bean field near Kimberly, Idaho, 1976.
Figure 2. Sediment yield versus application
efficiency for a bean field near Kimberly, Idaho,
1976.
189
-------
RETURN FLOW MANAGEMENT
different sizes of mini-basins used in this study
even though the average depth of the larger
basins was less than that of the smaller basins.
There is also no apparent benefit to using a
grassed overflow section instead of a plastic
section. It is doubtful that a farmer would spend
the time and money to install plastic sections.
However, grassed overflow sections should be
relatively easy to manage, and constuction of
simple grassed overflow mini-basins can
provide an economical means of retaining
sediments on a field.
One of the 3-furrow basins was filled with
sediment after five irrigations. This indicates
that the larger 4 or 5-furrow basins are more
likely to remain effective over a full season.
TABLE 9
Results of Magic Valley Mini-Basin Study, 1976
Sfdimtnt Yield (kg haj
Basin
}
2
3
4
5
6
Basin-
Silt
3
3
4
4
5
5
Oi-erflou
Type
grass
plastic
grass
plastic
grass
plastic
Basins
930
690
400
720
630
770
Check
Furrou-s
13.480
16.540
14,670
11,080
17,720
16,180
Removal
Efficiency
<°il
93.2
95.4
96.4
93.3
96.3
95.0
'Number of furrows entering basin.
Vegetated Buffer Strips
The vegetated buffer strip study was con-
ducted in 1976 on a 1.8-hectare spring wheat
field at the Kimberly Center. Flow and sediment
yield data for the three plots established at this
site are summarized in Table 10. Plot 1 was the
check plot. Plot 2 was banded with a single
planting of springwheat while Plot 3 was band-
ed with a double planting. These bands or strips
were 2.4 m wide. The plots were all the same size
(approximately 0.6 hectares). The slopes of the
three plots were 1.2, 1.4. and 1.7 percent respec-
tively. Each plot was irrigated six times using
6.5-hour sets.
The total amounts of water applied to the
three plots ranged from 501 mm on Plot 2 to
535 mm on Plot 1. The runoff ranged from
148 mm on Plot 3 to 162 mm on Plots 1 and 2.
The average net application to the three plots
was 359 mm. The seasonal water application
efficiencies for the three plots were essentially
all the same. The seasonal sediment yields, on
the other hand, varied considerably — from
1828 kg/ha for Plot 3 (double planted) to 8618
kg/ha for the check plot.
TABLE 10
Seasonal Totals for the Magic Valley
Buffer Strip Study, 1976
Plot
1
2
3
Stand Density
(sternum1/
519
739
815
Inflow
(mm)
535
501
514
Outflow
(mm)
162
162
148
Application1
Efficiency
<%>
70
68
71
Sediment
Yield
(kg'ha)
8,618
3,411
1.828
'Percentage of inflow retained on plot.
As can be seen from the data presented in
Table 10, increasing the stand density of the
wheat resulted in a large reduction in sediment
yield. In fact, increasing the stand density from
519 to 815 stems/m 2 (an increase of 57 percent)
resulted in over a four-fold reduction in the
sediment yield. Non-significant differences
between plots in the total amounts of water
applied and the application efficiencies indicate
that retardation of flow in the buffer strips was
sufficient to remove sediment but did not in-
crease infiltration appreciably. Planting the
buffer strips required very little effort and did
not disrupt normal irrigation and farming
operations.
Sediment Removal by Alfalfa
A cooperator's 1.5-hectare field of alfalfa
was monitored during 1975 to determine the
magnitude of sediment reduction obtainable
when surface runoff is reused for irrigation. The
field was 146 m long and had an average slope of
1.7 percent respectively. Each plot was irrigated
six times using 6.5-hour sets.
The results obtained from this monitoring
program are presented in Table 11. The net
amount of water applied during the four
irrigations was 657 mm. The application ef-
ficiencies ranged from 85 percent for the third
irrigation to 95 percent for the first irrigation.
The average for the season was 88 percent. The
average sediment concentration of the outflow
was 340 mg/1 while that of the inflow was
170 mg/1. The net sediment removal varied
from 42 percent for the first irrigation to 97
percent for the fourth irrigation, and averaged
79 percent for the season. These results show
that utilization of sediment laden runoff to
irrigate a dense growing crop such as alfalfa is a
viable method for improving water quality and
retaining sediments on farm fields.
190
-------
CONTROL OF SEDIMENT LOSSES
TABLE 11
Results of Sediment Removal by Alfalfa Study,
1975
Irrigation
1
2
3
4
Totals
Water
Inflow
(mmf
172
225
207
146
746
Sediment
Outflow
(mm)
8
30
31
20
89
Application1
Efficiency
1%)
95
87
85
86
88
Inflow
(kg/hat
125
263
269
509
1,166
Outflow
(kg/ha)
45
152
36
14
247
Removed
<"c)
64
42
58
97
79
'Percentage of inflow retained on field.
Furrow Cutback Irrigation
The effects of furrow cutback irrigation on
water and sediment losses from a cooperator's
potato field were evaluated in 1975. This study
was conducted on a 1.7-hectare field which was
146 m long and had a slope of 1.0 percent. The
field was divided into two equal plots and
irrigated so that each plot received 12 uncon-
trolled or non-cutback irrigations and 12 con-
trolled or cutback irrigations. Seasonal water
and sediment budgets for the two plots are
presented in Table 12.
The net application of water to the non-
cutback set was 634 mm while that to the
cutback set was 948 mm. The estimated
seasonal consumptive use by the crop was
546 mm. The average application efficiencies
for the non-cutback and cutback sets were 43
and 80 percent respectively. The non-cutback
irrigations resulted in 853 mm of runoff and a
net sediment loss of 92,540 kg/ha, while the
cutback irrigations resulted in 241 mm of runoff
TABLE 12
Seasonal Water and Sediment Budgets for Cutback
and Non-Cutback Irrigations of a Potato Field, 1975
Cutback Non-Cutback
Irrigations Irrigations
Inflow (mm)
Outflow (mm)
Appb'cation Efficiency (%)'
Sediment On (kg/ha)
Sediment Off (kg/ha)
Net Sediment Loss (kg/ha)
1,189
241
80
2,014
14,309
12,295
1,487
853
43
2,201
94,741
92,540
'Percentage of inflow retained on field.
and a net sediment loss of 12,294 kg/ha. These
results re-emphasize the direct relation between
irrigation efficiency and sediment loss. In this
study, a 20 percent decrease in the total amount
of water applied to a gravity-irrigated field
resulted in an 85 percent decrease in the amount
of sediment lost from the field.
REFERENCES
1. Busch, J. R., D. W. Fitzsimmons, G. C.
Lewis and D. V. Naylor. 1972. Cultural in-
fluences of irrigation drainage water.
Proceedings, ASCE Irrigation and Drainage
Specialty Conference, Spokane, Washington.
September, 12 p.
2. Fitzsimmons, D. W., G. C. Lewis, D. V.
Naylor and J. R. Busch. 1972. Nitrogen,
phosphorus and other inorganic materials in
waters in a gravity-irrigated area. Transactions
of the ASAE 15(2): 292-295.
3. Fitzsimmons, D. W., G. C. Lewis, K. H.
Lindeborg, J. R. Busch, D. V. Naylor and D. H.
Fortier. 1975. Effects of on-farm water manage-
ment practices on water quality in the Boise
Valley. Advance Report, Corps of Engineers
(Walla Walla, Washington) Boise Valley
Regional Water Management Study. December,
73 p.
4. Naylor, D. V., G. C. Lewis, J. R. Busch
and D. W. Fitzsimmons. 1976. Quality of irriga-
tion and drainage waters in the Boise Valley.
Idaho Agricultural Experiment Station
Research Bulletin No. 97. June, 22 p.
5. Naylor, D. V., G. C. Lewis, J. R. Busch
and D. W. Fitzsimmons. 1976. Quality of irriga-
tion and drainage waters in the Boise Valley.
Idaho Agricultural Experiment Station
Research Bulletin No. 97. June, 22 p.
5. Mech, Stephen J. and Dwight D. Smith.
1967. Water erosion under irrigation. In Irriga-
tion of Agricultural Lands, Ed. Robert M.
Hagan, Howard R. Raise and Talcott W. Ed-
minster. Agronomy Series 11, pp. 951-963.
6. Wells, Dan M. 1974. Agricultural waste
management, by the Committee on Agricultural
Waste Management of the Environmental
Engineering Division. J. of Environ. Engr. Div.,
ASCE 100 (No. EE1): 1-6.
191
-------
Economic Analysis of On-Farm
Methods for Controlling Sediment
and Nutrient Losses
K. H. LINDEBORG, L. CONKLIN, R. B. LONG,
and E. L. MICHALSON
University of Idaho, Moscow, Idaho
ABSTRACT
Information from 150 farmers in the Boise
and Magic Valley areas was collected relative to
current crop production, tillage and irrigation
practices, cost of production, and income from
crop production. The physical and economic
data were processed by a budget generating
method allowing standardized procedures of
estimating costs of operating machinery and
equipment. Partial budgets for each important
crop grown in the two areas were developed
from the computer program. Representative
farms were simulated for the two areas based on
the partial budgets. The simulated farm sizes
will be a measure of the profitability of the
present farming operation under existing water
management practices.
The effects on farm net income of imple-
menting erosion control practices on the farms,
as well as by district, are analyzed using linear
programming models. The analysis is based on
two farm sizes in each area, and measures the
effects on net farm income for a given set of
rotation with increasing sedimentation loss, or
alternately, a changing cropping pattern with a
given lower limit of sediment loss with the
resulting change in farm net income.
In measuring the external diseconomies
effected by irrigation, an attempt was made to
first determine the physical changes and then
measure the economic cost. The physical im-
pacts of irrigation were found to focus about
water quality (sedimentation, specific conduc-
tivity, phosphorous, and nitrogen) and higher
groundwater levels. Sedimentation resulted in
annual costs of about $1.4 million in 1975 to the
farm community to keep irrigation canals and
ditches clean. High groundwater levels
prevented the use of land for certain purposes
(basements in houses) and actually cost the city
of Meridian about $150 per month in pumping
costs. Water quality in the Boise River (Boise
compared to Notus) changed considerably due
to economic activity in the valley. During the
irrigation season, specific conductivity, nitro-
gen and phosphorous levels were five to ten
times greater below the irrigated area. Boise
River water also was found to contribute to
pollution problems further downstream in the
Snake River at Brownlee Reservoir.
INTRODUCTION
The full impact of erosion on surface
irrigated land is becoming more and more
widely perceived in society today. Not only is
productive topsoil washed away over time, it is
filling in downstream waterways and reservoirs
and degrading the water quality of downstream
flows. Many people are concerned over what can
be done to reduce the magnitude of this problem,
and what the costs would be.
This report is a first approximation of cost
effectiveness estimates for selected methods of
reducing the sediment loss from farms. It draws
on data from three field monitoring studies and
a survey of farmers in the Magic Valley and
Boise Valley of southern Idaho in order to
evaluate sediment loss control practices in
terms of the cost effectiveness one could
reasonably expect on an actual farm.
DATA SOURCES
A farm cost survey was conducted after the
harvest season in 1975. One hundred and fifty
randomly selected farm operators were inter-
viewed and asked to describe their present
farming practices, with special emphassis on
field operations and irrigation practices. The
193
-------
RETURN FLOW MANAGEMENT
study areas that were sampled were delineated
so that the farms within them had soils,
topography, and water supply conditions
similar to those at the monitored field sites.
The information from the farm survey was
processed so that representative farm models
could be constructed for each study area. In this
procedure, fifteen of the interviews were exclud-
ed because most of the farm was located outside
of the study area, or had such sandy soils that it
should not have been included in the study
areas. Typical tillage, cultural and irrigation
practices were then combined with input and
commodity prices, and cost-and-return budgets
were developed for representative farm models.
The physical parameters of irrigation water
supply, return flow, and sediment loss were
estimated for typical crop and field conditions
in the study areas. These estimates were based
on data from field monitoring projects con-
ducted by the University of Idaho over the past
five years (1, 2, 3). Data from sediment loss
control investigations associated with these
projects were summarized to estimate the
physical effectiveness that can be reasonably
expected from each practice.
ESTIMATION OF PHYSICAL
PARAMETERS
There are so many variables that influence
sediment loss from surface irrigated fields that
it has been essentially impossible to predict
how much soil will be lost from a particular field
during the irrigation season. The figures
presented in Table 1 are crude estimates of
sediment loss on a typical farm in the areas
studied. In this case, the concept of typical
includes the assumptions of silt loam soils with
slopes varying from one-half to four percent and
current management practices.
TABLE 1
"Typical" sediment loss levels for surface
irrigation crops on farms in the Magic Valley
and Boise Vallev.
Crops
Typical Sediment Loss
kg ha
Corn, beans, beets
Potatoes
Small grain, peas
Alfalfa
In order to estimate the cost effectiveness of
alternative methods of sediment loss reduction,
it was first necessary to estimate the physical
effectiveness of these practices. The effec-
tiveness of the methods analyzed in this paper
was based on field work that has already been
reported to this conference (2). Estimates were
made of the expected sedment loss reduction for
each practice under typical farm conditions.
These estimates are presented in Table 2, in
terms of the percentage of "typical" sediment
loss that would be retained on the farm by using
a particlar practice.
The sediment loss reduction practices are
described briefly below.
Flow cut-back involves running the usual
amount of water down the furrow or corrugate
for the first six to eight hours of a set, or until the
water is through to the lower end of the field,
then cutting back stream size for the remainder
of the set. This reduced flow results in less
erosion and less soil transport in the furrows,
however, more labor is required to perform the
cut-back operation. It was assumed that the
irrigation labor requirement was doubled by
this practice because, in essence, the water must
be set and then re-set.
TABLE 2
Expected sediment loss reduction for
selected control practices on typical
irrigated farms in the Magic Valley
and Boise Valley.
Percent of "Typical" Sediment
Control Practice Loss Retained on Farm
Flow cut-back
Grass or grain strip
Sediment pond
Mini-basin
Sprinklers
30
50
67
90
100
8,070
40,340
3,140
900
To illustrate the effectiveness of the prac-
tice, assume that the sediment loss from a bean
field is 8070 kg / ha under current management.
Referring to Table 2, if a cut-back procedure was
adopted, sediment loss from the field would be
reduced by 30 percent, to 5649 kg/ha.
A grass or grain strip can be drilled in
across the lower end of a row crop field to slow
the velocity of tailwater running off the field
during irrigation.which will cause it to deposit
some of its sediment. Crops that are normally
planted with a drill can be double (or even triple)
planted in a strip across the lower end.
194
-------
ECONOMICS OF CONTROLLING LOSSES
A sediment pond is a pond dug into a
waterway for the purpose of reducing flow
velocity and retaining sediment. The effec-
tiveness of such a pond depends on its shape
and size and on the volume of flow. In several
monitored ponds, about two-thirds of the sedi-
ment entering these ponds over the irrigation
season was retained by the pond.
The mini-basins are small shallow ponds
constructed on the lower end of a field by putting
in a low berm along the bank of the drain ditch.
Other berms are constructed perpendicular to
the drain ditch, so that each basin retains the
tailwater flow of just a few furrows. The berm
along the drain ditch also serves as a spillway
when necessary, so it should be seeded with
grass to minimize erosion into the ditch.
Sprinklers are in use in many areas of the
West, mainly in areas that do not have facilities
for surface irrigation or that have sandy soils
that are difficult to surface irrigate. Some land
is being converted to sprinkler irrigation using
power-move systems, largely for the sake of
labor saving and field consolidation. In addi-
tion, it is usually fairly easy to calibrate
sprinkler systems so that the water application
rate is equal to the soil intake rate. This makes
more efficient use of irrigation water than most
surface application methods and allows
tailwater runoff (and sediment loss) from fields
to be reduced to zero.
ECONOMIC ANALYSIS
The assumptions and calculations involved
in figuring the cost effectiveness of the selected
sediment loss control practices are described in
the following sections.
Cut Back Flow
The irrigation labor that is typically in-
volved in growing a crop was computed on the
basis of the farm survey data. For each crop the
median number of irrigations were multiplied
by 1.0 hour/ha/irrigation to arrive at a figure
for the hours of irrigation labor used per hectare
over the irrigation season. This time does not
include pre-season ditch work on the farm,
which was computed separately for the crop
budget.
The cut-back procedure was assumed to
double irrigation, labor time, which was valued
at $3.00 per hour. Table 3 presents the cost
figures. For purposes of comparison, the cost per
ton of sediment retained on the field is computed
on the basis of both a 30 percent and an 80
percent retention of sediment that would be lost
under typical present management. Field trials
have shown a large variability in this sediment
retention ratio — from 16 percent to slightly
more than 85 percent (2).
Grass or Grain Strip
For row crops, a grass or grain strip at the
end of a field would probably entail putting
some land out of production. In order to estimate
how much land would be out of production, it
was necessary to make some assumptions
regarding field size and shape. From the farm
survey data, and from air photos in the Canyon
Area Soil Survey for the Boise Valley, it was
noted that the majority of fields in the study
area were 4.05 ha or 8.10 ha tracts. The 8.1 ha
fields were usually rectangular, with dimen-
sions of 201 meters by 402 meters. Water may
run across either the long or short headland,
depending on field slope. For the sake of stand-
ardization, it was assumed that all fields are
8.1 ha and that the low headland averages 300
meters in length. If the seeded strip is 2.44
meters wide then 732m2, or 0.0732 ha of
headland would be out of production.
It was further assumed that, for row crops,
the headlands are less productive per unit area
than the entire field — that headlands out of
production would have yielded 80 percent of the
field average used in the crop budgets. For
corrugated crops, it was assumed that a normal
harvest would be taken from the overplanted
strip. No overplant procedure was used with
alfalfa since an alfalfa field is one big
vegetative strip.
Table 4 shows the costs of putting row crop
land out of production for the vegetative strip,
based on opportunity cost, or the value of
production given up. Calculations are not done
for potatoes. A vegetative strip would probably
be obliterated by the first irrigation on potatoes.
The operating costs consist of the costs
associated with drilling in the strip at the
beginning of the growing season and spreading
the deposited sediment back over the field after
harvest. For row crops, the "hassle factor"
involved in getting the drill set up to plant the
strip, driving to the field, planting the strip,
driving back to the farmstead and putting the
drill away is difficult to estimate. For calcula-
195
-------
RETURN FLOW MANAGEMENT
TABLE 3
Costs Associated With Flow Cut-Back to Reduce Sediment Loss on a "Typical" Farm in the
Magic Valley or Boise Valley
Irrigation Labor
Sediment Retained
Cost Per Ton of
Sediment Retained
Crop
Dry Beans
Bean Seed
Corn
Sugar Beets
Potatoes
Pea Seed
Win. Wheat
Spr. Grain
Alfalfa Hay
No. of
Irrigations
1
8
7
10
12
5
5
4
6
Normal
(hr./ha.)
1
8
7
10
12
5
5
4
6
Cut-Back
(hr./ha.)
14
16
14
20
24
10
10
8
12
Added
Labor
Cost
<&ha.)
42.00
48.00
42.00
60.00
72.00
30.00
30.00
24.00
36.00
30 percent
Retention
(tJha.)
2.42
2.42
2.42
2.42
12.10
0.94
0.94
0.94
0.27
80 percent
Retention
(t/ha.)
6.45
6.45
6.45
6.45
32.27
2.51
2.51
2.51
0.72
30 percent
Retention
($/t.)
17.36
19.83
17.36
24.79
5.95
31.91
31.91
25.53
133.33
80 percent
Retention
Wt.)
6.51
7.44
6.51
9.30
2.23
11.95
11.95
9.56
50.00
NOTE: "t." is the abbreviation for "metric ton" or "tonne" (1000 kg.).
TABLE 4
Net Returns and Opportunity Cost of Land Out of
Production For Vegetative Strips
(Average For Farm Models).
Net Return to Land
And Management
Headlands Opportunity Cost
Field (2O% Lower of 0.0732 ha
Average Yield) out of Production
Crop (&ha) (Vha) $
TABLE 5
Annual Operating Costs Associated with
Vegetative Strips on 8.1-Hectare Fields.
Dry Beans
Bean Seed
Corn Seed
Sugar Beets
529
988
806
875
334
688
477
504
24.45
50.36
34.92
36.90
Row Crops
Machine variable costs for drill and tractor.
S3.88/hr. X 0.86 hr./ha, X 0.0732 ha. X 4
Labor cost of seeding:
S3.00/hr. X 0.86 hr Jha. X 0.0732 ha. X 4
Wheat or barley seed:
100 kg./ha. X 0.0732 ha. X $0.24/kg.
Dirt Spreading:
*3.70/ha. X 8.1 ha.
Corrugated Crops
$
0.98
0.76
1.76
Machine variable costs for drill and tractor:
*3.88/hr. X 0.86 hr./ha. X 0.0732 ha.
Labor cost of seeding:
*3.00/hr. X 0.86 hr./ha. X 0.0732 ha.
Seed- Wheat or Barley
Peas
Dirt Spreading:
*3.701ia. X 8.1 ha. X .05
Total Annual Operating Costs
Grain
*
0.24
= 0.19
1.76
14.98
17.17
Peas
t
0.24
0.19
16.00
14.98
31.41
tion of labor cost and machinery variable costs,
the actual field time was multiplied by four in an
attempt to include all the steps in the operation.
TABLE 6
Total Annual Costs Associated with Vegetative Strips on 8.1-Hectare Fields.
Crop
Dry Beans
Bean Seed
Corn Seed
Sugar Beets
Pea Seed
Grain
Opportunity Cost
Of Land Out of
Production
$
24.45
50.36
34.92
36.90
0
0
Total Costs
Operating
Costs
$
33.47
33.47
33.47
33.47
31.41
17.17
Whole
Field
$
57.92
83.83
68.39
70.37
31.41
17.17
Per
Hectare
$
7.15
10.35
8.44
8.69
3.88
2.12
Sediment
Retained
(tJha.)
4.03
4.03
4.03
4.03
1.57
1.57
Cost per
Tonne of
Sediment Retained
<$/t.)
1.77
2.57
2.10
2.16
2.47
1.35
196
-------
For grain and pea fields, only the time involved
in making another pass over the lower end of the
field was used in computing costs.
Machine field time and variable costs per
hour of operation are from the model farm
budgets for grain enterprises. It was assumed
the farmer already has the machinery and that
machinery fixed costs had already been
allocated to crop enterprises. Therefore, no fixed
costs were charged to the strips. Table 5 shows
the calculation of operating costs for row crops,
grain and peas. Total annual costs for
vegetative strips are shown in Table 6.
Dirt spreading was accomplished by an
extra time over the field with a land plane. This
would probably have to be done every year for
row crops, every other year for grain and peas.
Based on numbers computed for the cost
budgets, this operation would cost $3.70 per
hectare.
Sediment Ponds
The cost effectiveness of sediment ponds
was analyzed for a situation where each 8.1
TABLE 7
Sediment Pond Size and Excavation Cost
Estimates for Surface Irrigated Crops on
8.1-Hectare Fields.
Pond Sue
Crop
Beans, Beets
Corn Seed
Potatoes
Grain, Peas
Alfalfa, Hay
Area
(m'i
54.9
54.9
168.4
36.6
36.6
Volume
fm'>
65.9
65.9
252.5
43.9
43.9
Excavation Cost
<@$0.75/m*)
<$)
49.41
49.41
189.40
32.94
32.94
Percent of
Pond Filled
in One Year
50
50
100
30
8
ECONOMICS OF CONTROLLING LOSSES
hectare field would have a pond in its tailwater
drains. Pond size and excavation costs depend
on the volume of sedimentto be handled. Table 7
shows the figures used for planning purposes in
this study.
Dirt spreading costs are usually about
equivalent to excavation cost, so excavation
TABLE 8
Annual Operating Costs for Sediment Ponds, with
Surface Irrigated Crops on 8.1-Hectare Fields.
Excavation and Annual
Spreading Costs- Percent of Operating
Pond Full Pond Filled Costs
Crop $ in One Year $
Beans, Beets
Corn Seed
Potatoes
Grain, Peas
Alfalfa Hay
98.82
98.82
378.80
65.88
65.88
50
50
100
30
8
49.41
49.41
378.80
19.76
5.27
TABLE 9
Opportunity Cost of Land Out of Production for
Sediment Ponds on 8.1-Hectare Fields.
Crop
Dry Beans
Bean Seed
Corn Seed
Sugar Beets
Potatoes
Pea Seed
Winter Wheat
Spring Wheat
Spring Barley
Alfalfa Hay
Opportunity Cost
Net Return Land of Land Out
Given Up Area of Production
(Who.) (ha.) $
334
688
477
504
714
592
381
331
111
257
0.016
0.016
0.016
0.016
0.051
0.011
0.011
0.011
0.011
0.011
5.34
11.01
7.63
8.06
36.41
7.61
4.19
3.64
1.22
2.83
TABLE 10
Total Annual Costs Associated With Sediment Ponds on 8.1-Hectare Fields.
Crop
Dry Beans
Bean Seed
Corn Seed
Sugar Beets
Potatoes
Pea Seed
Winter Wheat
Spring Wheat
Spring Barley
Alfalfa Hay
Opportunity Cost
of Land Out of
Production
$
5.34
11.01
7.63
8.06
36.41
7.61
4.19
3.64
1.22
2.83
Total Costs
Operating
Costs
$
49.41
49.41
49.41
49.41
378.80
19.76
19.76
19.76
19.76
5.27
Whole
Field
$
54.75
60.42
57.04
57.47
415.21
27.37
23.95
23.40
20.98
8.10
Per
Hectare
$
6.76
7.46
7.04
7.10
51.26
3.38
2.96
2.89
2.59
1.00
Sediment
Retained
(t./haj
5.41
5.41
5.41
5.41
27.03
2.10
2.10
2.10
2.10
0.60
Cost per Tonne
of Sediment
Retained
($/t.)
1.25
1.38
1.30
1.31
1.90
1.61
1.41
1.38
1.23
1.67
197
-------
RETURN FLOW MANAGEMENT
costs were doubled to arrive at the total
operating costs for sediment ponds. Table 8
presents these operating costs on an annual
basis.
To compute the costs associated with put-
ting land out of production by the installation of
a sediment pond, the surface area of the pond
was tripled to allow for margins around it. This
area was then multiplied by the value of output
given up on a headland area of that size. Once
again, row crop yields on headlands were
assumed to be 20 percent below field average.
Table 9 shows the opportunity costs of the land
out of production, and Table 10 presents total
annual costs for the sediment ponds.
Mini-Basins
The land area out of production for mini-
basins would be the same as for vegetative
strips, so the opportunity cost of land out of
production would also be the same. It was
assumed that the basin along the drain ditch
bank would have to be shaped and seeded every
five years. Costs for this operation on a
headland 300 meters long would be as follows:
Labor: 5 hr. x $3.00/hr. H- 5 years = $3.00/year
Machine variable costs for tractor,
"V" ditcher and blade:
2.5 hr. x $2.00 hr. H- 5 years = $1.00/year
Grass seed:
(area seeded - 300 m. x 0.8 m = 240 m2
240 m2 -=- 61.5 rnVkg. of seed = 4.0 kg.
4.0 kg. x $2.20/kg H- 5 years
$1.76/year
$5.76/year
Maintaining the mini-basins would involve
spreading dirt and rebuilding the field berms
every year on row crops, every other year on
grain and peas. For alfalfa, basins put in before
establishment would serve for the duration of
the stand. Construction costs are shown in
Table 11.
The amount of dirt to be spread, relative to a
vegetative strip, would be proportional to the
effectiveness of the two methods. The basins
retain 1.8 times as much sediment as would a
vegetative strip for a given field and crop, so
spreading costs were figured to be 1.8 times as
much. The details of dirt spreading costs are
TABLE 12
Dirt Spreading Costs for Mini-Basins on
8.1-Hectare Fields.
Crop
Beans, Beets, Corn
Potatoes
Grain, Peas
Alfalfa Hay
Cost
53.95
74.92
26.96
7.74
TABLE 11
Basin Construction Costs for Mini-Basins on 8.1-Hectare Fields.
Crop
Beans, Beets
Corn
Potatoes
Grain, Peas
Alfalfa Hay-
Beans. Beets
Corn
Potatoes
Grain, Peas
Alfalfa Hay
Furrow No. of No. of Basins Machine Labor
Spacing Furrows on 300m. Time @ 0.1
(m) Per Basin Headland hr./Basin
1.12
0.76
0.91
0.76
Machine Labor
Cost @ $3.00 hr.
$
20.10
29.70
24.60
19.80
19.80
4
4
4
6
Shovel Work
@ $3.00/hr.
S
40.20
59.40
49.20
39.60
39.60
67
99
82
66
Machine Variable
Costs® $2.00 hr.
$
13.40
19.80
16.40
13.20
13.20
(hr.)
6.7
9.9
8.2
6.6
Total Berm
Per Time
73.70
108.90
90.20
72.60
72.60
Shovel
Work@
0.2 hr./Basin
(hr.)
13.4
19.8
16.5
13.2
Const. Costs
Annual
73.70
108.90
90.20
36.30
(72.60 •*• yrs.
of stand)
198
-------
ECONOMICS OF CONTROLLING LOSSES
shown in Table 12. Table 13 shows the total
annual costs for the mini-basins. Operating
costs are the sum of 15.76 plus total basin
construction costs on an annual basis (from
Table 11), plus dirt spreading costs (from Table
12.)
Sprinkler Irrigation
Any large scale conversion of surface
irrigated lands to sprinklers would have many
impacts that are beyond the scope of this paper.
The availability of power for pumping would
probably be a major constraint. Even so, it
might be useful to look at the costs of installing
and operating a sprinkler system for com-
parison with other methods of reducing sedi-
ment loss. A sprinkler system is to be operated
so as to eliminate water runoff and sediment
loss from most fields.
Cost estimates were computed for a side-roll
sprinkler system, as shown in Table 14. The
system consists of a pump taking water out of a
pond, a mainline and six laterals for 56.7
hectares of irrigated crops on a quarter section
of land. Depreciation was calculated on a
straight-line basis, with a useful life of 15 years
for all components and a salvage value of 10
percent of new price. Annual interest was com-
puted at eight percent per year on average
investment, which is defined by the formula:
Average Investment =
(New Price + Salvage Value)-^ 2.
Costs for power and repairs are shown on a per
irrigation basis in Table 14, as are irrigation
labor savings. In Table 15, the total (net) costs of
owning and operating the sprinkler system is
presented by crop. The cost of sediment reten-
tion is also presented in Table 15.
SUMMARY OF RESULTS
The sediment loss reduction practices con-
sidered in this paper show a wide variation in
costs.
Relative cost effectiveness depends on the
crops under consideration, their present level of
sediment loss and the physical effectiveness of
the control practice. The costs per ton of sedi-
ment retained on the farm are shown for each
practice in Table 16.
Flow cut-back is the most expensive way to
retain a given volume of sediment if sediment
loss is only reduced by 30 percent. With an 80
percent reduction, flow cut-back is still a more
expensive way to retain soil than mini-basins
for all crops except potatoes.
For all crops, sediment ponds can retain
about two-thirds of normal sediment loss at a
cost of less than two dollars per tonne. On grain
crops, a vegetated strip, or lower end overplant,
compares favorably with sediment ponds on a
cost per unit basis. However, the strips retain
less total sediment. Costs per ton of sediment
retained range from $1.77 to $2.57 for the
vegetative strips under row crops.
TABLE 13
Total Annual Costs Associated With Mini-Basins on 8.1-Hectare Fields.
Crop
Dry Beans
Bean Seed
Corn Seed
Sugar Beets
Potatoes
Pea Seed
Win. Wheat
Spr. Wheat
Spr. Barley
Alfalfa Hay
(2 yr. stand)
Opportunity
of Land Out
of Production
$
24.45
50.36
34.92
36.90
52.26
50.65
27.89
24.23
8.13
18.81
Total Costs
Operating
Costs
$
133.46
133.41
168.61
133.41
170.88
69.02
69.02
69.02
69.02
37.70
Whole
Field
$
157.86
183.77
203.53
170.31
223.14
119.67
96.91
93.25
77.15
56.51
Per
Hectare
$
19.49
22.69
25.13
21.03
27.55
14.77
11.96
11.51
9.52
6.98
Sediment
Retained
(t./ha.)
7.26
7.26
7.26
7.26
10.09'
2.82
2.82
2.82
2.82
0.81
Cost Per
Tonne of
Sediment Retained
($/t.)
2.68
3.13
3.46
2.90
2.73
5.24
4.24
4.08
3.38
8.62
'Because of the large volume of sediment, the mini-basins on a potato field would probably fill up before the season
ended. For purposes of comparison, it is assumed that 25% of incoming sediment for the season will be retained.
199
-------
RETURN FLOW MANAGEMENT
Mini-basins retain a greater proportion of
normal sediment loss than do strips or ponds,
but are also more expensive, both in total and on
a per unit of sediment basis.
Converting the surface irrigation system to
side-roll sprinkler could eliminate sediment loss
from the farm, but involves a large jump in costs
from those associated with surface tailwater
management practices. Again, potatoes are an
exception, because of the huge volume of sedi-
ment loss under the surface irrigation practices
that were assumed to be typical. For other crops,
the conversion to sprinklers would cost about 10
times as much per unit of sediment retained as
would sediment ponds and five times as much
as the mini-basins. On grain, sprinklers would
cost 18 times as much as ponds and six times as
much as mini-basins, per unit of sediment
retained.
TABLE 14
Annual Costs of Owning and Operating a Side-Roll
Sprinkler System on 56.7 Hectares of Irrigated Crops.
Investment Annual Annual
Cost Depreciation Interest
Item ($) ($/yr.) f$/yr.)
Laterals 27,000
Mainline 6,600
50 HP. Pump 3,500
Pond 500
34,600
Costs Per Hectare:
Depreciation and Interest:
Power:
Repairs:
Labor Cost Savings:
Ditch Maintenance:
Irrigation:
1,620 1,188
396 290
210 154
33 20
2,259 1,652
$69.00
3.70 per irrigation
2.50 per irrigation
9.25
0.75 per irrigation
TABLE 15
Costs of Owning and Operating a Side-Roll Sprinkler System, Relative to Sediment Retention.
Power
No. of Cost
Crop Irrigations <$/ha.)
Dry Beans
Bean Seed
Corn Seed
Sugar Beets
Potatoes
Peas
Win. Wheat
Spr. Grain
Alfalfa Hay
7
8
7
10
12
5
5
4
6
25.90
29.60
25.90
37.00
44.40
18.50
18.50
14.80
22.20
Repair
Cost
(&ha.)
17.50
20.00
17.50
25.00
30.00
12.50
12.50
10.00
15.00
Labor Cost Savings
Ditch
Maint. Irrigation
(Vha.) (Vha.)
-9.25
-9.25
-9.25
-9.25
-9.25
-9.25
-9.25
-9.25
-9.25
-5.25
-6.00
-5.25
-7.50
-9.00
-3.75
-3.75
-3.00
-4.50
Depr.
and
Interest
Who.)
69.00
69.00
69.00
69.00
69.00
69.00
69.00
69.00
69.00
Net
Annual
Cost
($faa.)
97.90
103.35
97.90
114.25
125.15
87.00
87.00
81.55
92.45
Sediment
Retained
(tJha.)
8.07
8.07
8.07
8.07
40.34
3.14
3.14
3.14
0.90
Cost of
Sediment
Retained
($/t.)
12.13
12.81
12.13
14.16
3.10
27.70
27.70
25.97
102.72
TABLE 16
Cost Effectiveness Summary for Selected Sediment Loss Control Practices.
Cost Per Ton of
Sediment Retained
Crop
Dry Beans
Bean Seed
Corn Seed
Sugar Beets
Potatoes
Pea Seed
Win. Wheat
Spr. Wheat
Spr. Barley
Alfalfa
Vegetative
Strip
$1.77
2.57
2.10
2.16
2.47
1.35
1.35
1.35
Sediment
Ponds
$1.25
1.38
1.30
1.31
1.90
1.61
1.41
1.38
1.23
1.67
Flow Cut-Back
Mini-Basins
$2.68
3.13
3.46
2.90
2.73
5.24
4.24
4.08
3.38
8.62
30% Retention
$ 17.36
19.83
17.36
24.79
5.95
31.91
31.91
25.53
25.53
133.33
80 Retention
$ 6.51
7.44
6.51
9.30
2.23
11.95
11.95
9.56
9.56
50.00
Side-Roll
Sprinkler
$ 12.13
12.81
12.13
14.16
3.10
27.70
27.70
25.97
25.97
102.72
2OO
-------
ECONOMICS OF CONTROLLING LOSSES
Work is presently going forward to incor-
porate these cost data into linear programming
models of representative, or model, farms for
each study area. The results of this work will
describe how a farm manager would meet a
specified limit on sediment loss from his farm. It
will also show how this adaptation will affect
farm income.
REFERENCES
1. Ballard, Floyd Leon. 1975. Analysis
and Design of Settling Basins for Irrigation
Return Flow. Unpublished M. S. thesis, Uni-
versity of Idaho, Moscow, Idaho. January.
2. Fitzimmons, D. W., C. E. Brockway, G.
C. Lewis, J. R. Busch, G. M. McMaster, C. W.
Berg. 1977. On Farm Methods for Controlling
Sediment and Nutrient Losses. Presentation at
the National Conference on Irrigation Return
Flow Quality Management, Fort Collins,
Colorado. May 16-19.
3. Watts, F. J., C. E. Brockway, A. E.
Oliver. 1974. Analysis and Design of Settling
Basins for Irrigation Return Flow. Research
Technical Completion Report, Project A-042-
IDA, Water Resources Research Institutute,
University of Idaho, Moscow, Idaho. Septem-
ber.
201
-------
Combining Agricultural
Improvements and Desalting of
Return Flows to Optimize
Local Salinity Control Policies
W. R. WALKER
Agricultural and Chemical Engineering Department,
Colorado State University, Fort Collins, Colorado
ABSTRACT
Mathematical simulations of the cost-
effectiveness relationship for various agricul-
tural and desalination alternatives for con-
trolling salinity in irrigation return flows are
being developed. The question of respective
feasibility for each type of control is determined
through minimizing the total costs, thereby
optimally selecting the best measures to imple-
ment. To demonstrate the concept, desalting
and canal linings were compared for the Grand
Valley in western Colorado. Results indicate
that desalting exhibits superior feasibility to
linings of canals where seepage rates are low. In
general, desalting will exhibit its best feasibility
when applied to large-scale applications and
when salt pickup rates stemming from irriga-
tion return flows are small.
INTRODUCTION
The development of desalination technolo-
gy in the United States has been guided by the
basic objective outlined by Congress to the U.S.
Department of the Interior's Office of Saline
Water (now combined with the Office of Water
Resources Research to form a single department
entitled, "Office of Water Research and
Technology"). This objective is:
"to provide for the development of prac-
ticable low-cost means for producing from
sea water or from other saline waters
(brackish and other mineralized or
chemically charged waters), water of a
quality suitable for agriculture, industrial,
municipal, and other beneficial consump-
tive uses."
To satisfy this objective, Congress has
provided funding for a massive research and
development effort, although the application to
large scale systems is only now beginning to
occur. The traditional scope of saline water
conversion programs has been to reclaim
otherwise unsuitable waters for specific needs.
However, the scope has dealt almost exclusively
with utilization of product water directly rather
than returning it to receiving waters in order to
improve the overall resource quality. Thus, with
mounting concerns for managing salinity on a
regional or basin-wide scale, the potential for
applying desalination within the framework of
an overall salinity control strategy is an in-
teresting one. In fact, the use of desalting
systems to resolve critical salinity problems is
already being planned as part of the Colorado
River International Salinity Control Project
agreement between the United States and the
Republic of Mexico (U.S. Department of the
Interior, 1973).
In the context of regional salinity control,
desalting costs can be expressed in dollars per
unit volume of salt extracted in the brine
discharge, rather than the conventional index
of costs per unit volume of reclaimed product
water. In this manner the respective feasibility
of desalination and other alternatives for salini-
ty management can be systematically com-
pared during the process of developing
strategies for actual implementation of salinity
controls. This paper examines the use of
desalination in controlling the quality of irriga-
tion return flows in conjunction with the more
traditional approaches associated with max-
imizing irrigation efficiency.
203
-------
RETURN FLOW MANAGEMENT
DESALINATION COSTS
In general, the costs associated with
desalting systems may be either classified as
those expended during construction or those
required annually to operate and maintain the
facilities. These costs are subject to inflational
pressures and must therefore be periodically
updated. Once costs are current, various
relationships between the costs and system
performance can be formulated. A detailed
description of desalting costing models was
given by Walker (1976) based upon work
reported by Prehn, et al. (1970) and U.S.
Department of the Interior (1972).
Capital and Annual Costs
It might be noted that the capital costs
delineated by Walker (1976) are based on es-
timating functions current during mid-1971 and
therefore must be updated to prevailing price
levels. The costs for construction, steam supply,
and general site development can be estimated
in the present time frame by employing the
Engineering News Record Construction Cost
Index (Engineering News Record is published
monthly by McGraw-Hill, Inc.). In July of 1971
the ENR index was 952, whereas in January
1976 it had risen to 1354. Consequently, an early
1976 cost estimate would be 1354/952 or 1.42
times the 1971 functional estimate. Other
capital costs for interest during construciton,
owners' general expense, start-up, and working
capital are functions of construction, steam,
and site development costs and are therefore
updated automatically. Land costs may be
estimated on a current basis by utilizing ex-
isting land prices.
Estimates of capital costs for feedwater and
brine facilities require several other inflational
factors. For example, conveyance pipeline costs
are modified by the Bureau of Reclamation
Concrete Pipeline Cost Index, CPI (1.17 for July
1971 and 1.88 for January 1976). Others are the
USBR Pumping Plant Cost Index, PPI (1.26 for
July 1971 and 1.98 for January 1976). USBR
Pumps and Prime Mover Cost Index, PMI (1.41
in July 1971 and 2.13 in January 1976), and the
USBR Canal and Earthwork Cost Index, El
(1.27 in July 1971 and 1.92 in January 1976).
The annualization of the construction costs
is divided according to whether or not the costs
represent depreciating capital. Depreciating
capital cost are multiplied by a "fixed charge
factor", FCR, which is the percentage of total
depreciating capital cost that is encompassed
by interest, amortization, insurance, and taxes.
Non-depreciating capital costs are multiplied by
the prevailing interest rate selected for project
evaluation or incurred in borrowing.
Annual costs must also be updated for price
increases due to inflation. Labor and materials
and steam generation O & M are updated using
the Bureau of Labor Statistics Labor Cost
Indices, SIC494-7, which was 3.76 in July 1971
and 4.93 in January 1976. Chemical costs are
multiplied by the present to 1971 ratio of the
Bureau of Labor Statistics Cost Index for
chemicals and allied products (181 for 1976,
#104.4 for 1971). Fuel and electricity costs are es-
timated using present prices for these inputs
and are therefore always estimated currently.
Replacement costs are expressed as functions of
plant capacity and are thus not updated by cost
indexing.
Water and Salt Costs
After describing the individual costs
associated with desalting systems, it is general-
ly necessary to express such cost in either
dollars per unit volume of product water (for
water supply feasibility) or dollars per unit of
salt extracted (for salinity control studies).
These cost bases are determined in this study by
dividing the total annual costs by the annual
volume of product water or brine salts.
Depreciating capital costs for the plant
itself and the feedwater — brine disposal sys-
tems are multiplied by the fixed charge factor.
Non-depreciating costs are next multiplied by
the interest rate, added to the annualized
depreciating costs, and finally summed with the
remaining annual costs; thus, using the eight
capital cost categories and six annual cost
elements described by Walker (1976) gives:
6 8 14
FCR 1 C' i + I r i C' i + 1 C' i
i=l i=7 i=9
Cnw = - (1)
pw
C p + U f x 3.65 x 10-
for product water costs, and for salt costs:
C 8 =
FCR £ C'i
i= 1
8 14
2 C'i+ I C'i
i=7 i=9
C b
bo
f 3.65 x 10-<
(2)
204
-------
OPTIMIZING SALINITY CONTROL
in which,
^ pw
Cs
= unit cost of product water, $/m3;
= unit cost of brine salts, $/metric ton;
= total annual cost for element i, $/year
(feedwater + plant + brine)
C' i = construction cost;
C' 2 = steam generation;
C' 3 = site development;
C' 4 — interest during construction;
C' 5 = start-up costs;
C' 6 = owner's general expense;
C' 7 = land;
C' 8 - working capital;
C' 9 = labor and materials (operational);
C' 10 = chemicals;
C' 11 = fuel;
C' 12 = electrical;
C' 13 = steam generation O & M; and
C' 14 = replacement.
C p — product water volume, m3 /day;
C b = brine volume, m3 /day;
C bo = TDS concentration in brine, mg/1;
FCR = fixed charge rate;
I r — interest rate; and
U f = use factor — fraction of total time in
actual operation.
Available Processes
Saline water conversion processes involve
the use of a semi-permeable barrier, which
excludes either water or salt flow. The barrier
may be a membrane which excludes salt such as
reverse osmosis (RO); one which excludes water
such as electrodialysis(ED); or one that ex-
changes salt for hydrogen and hydroxide ions
which unite to produce water (ion exchange, IX).
The barrier may also be a "phase boundary"
which excludes the salts. For example, va-
porization of water using multi-stage flash
(MSF), vertical tube evaporation-MSF (VTE-
MSF) and vapor compression-VTE-MSF (VC-
VTE-MSF) leaves the salts in the remaining
solution as does solidification of water using
vacuum-freezing-vapor compression (VF-VC)
processes (Probstein, 1973).The driving poten-
tial for each of these processes is either heat
(distillation and freezing), pressure (RO), elec-
trical (ED), or chemical (IX).
Each desalination process has specific ad-
vantages depending on such factors as
feedwater chemistry and desired product water.
A general review of these factors along with a
description of the costing model for each
technology is too lengthy for this paper, so the
interested reader is referred to the original
references.
Sensitivity in Cost Estimates
The application of desalting technology to
regional water quality management tends to be
a very site specific problem. As a result,
generalization of cost analyses is difficult.
However, it might be useful to point out the
model's sensitivity to various input parameters
relative to an arbitrary "base" so the relative
importance of the variables can be reviewed.
Table 1 summarizes the base values of some
typical input parameters that might be found in
an irrigation return flow problem.
Desalting System Capacity
Desalting costs expressed in terms of
dollars per ton of salt removed, or dollars per
cubic meter of product water, exhibit substan-
tial economies of scale. For the base condition,
the scale effects for each process are illustrated
in Figure 1 (ion exchange has not been included
because of high TDS in the feedwater). In nearly
every process, the unit costs at 950 m3/day are
2-4 times the cost at 121,000 m3/day. The CTE-
MSF, VC-VTE-MSF, and RO process costs at
the lower value are 3.5 to 3.6 times the upper
capacity indicating much larger importance of
scale that is associated with MSF (2.80), VF-VC
(2.78), or ED (2.13). Of these specific processes,
electrodialysis is more affected by input
parameters and therefore should be evaluated
more closely in the reconnaissance investiga-
tion.
The scale factor in desalination will gener-
ally preclude small installations for salinity
control since other measures for reducing salini-
ty will be cheaper. However, as the level of
implementation increases and the marginal
costs of desalting decrease, this technology may
become highly competetive with the various
other alternatives. Consequently, a major
parameter in a salinity control analysis that
can be expected to affect the potential use of
desalination is the level of implementation.
205
-------
RETURN FLOW MANAGEMENT
TABLE 1
Standardized desalting model input parameters
for variable sensitivity analyses.
Cp = 1.5 x 104 m-Vday (4mgd)
TDSi = 5000 mg 1
TDSp = 5000 mg/1
ENR = 1354
BLSi = 4.93
BLS2 = 181
CPI = 1.88
PPI - 1.98
PMI = 2.13
El = 1.92
Db = 1000 m (3280 feet)
Df - 100 m (328 feet)
Dib ; 100 m (328 feet)
I) if - 1000 m (3280 feet)
K 1.D7 13.5 feet)
Ec = $7.2 x 10*. M Joules ($20/100 kwh)
FCK = 0.0856
Fr = $1.20 x 10s M Joules ($1.14 MBTU)
Ir = 7%
Lp = $4,942, ha ($2000 acre)
T = 15.6°C(60°F)
l"t = 0.90
Na = 1260 mg 1
Mg = 123 mg 1
Ca = 393 me I
K = 8 mg 1
HCOs = 106 mg 1
Cl = 2035 mg 1
SO 4 = 1075 mg 1
NO 3 = Omg/1
Feedwater is supplied by wells and brine is disposed
of with injection wells.
Feedwater Salinity
Because of referencing desalting costs to
salinity control, the feedwater salinity is an
important parameter in the evaluation of the
alternative processes. The distillation and freez-
ing methods are not substantially limited by-
input salinity since the same measures are
necessary to desalt 1000 mg /I feedwater as for
10,000 mg 1 . Consequently, the higher the
feedwater salinity the lower the unit cost for
these processes. Reverse osmosis and elec-
trodialysis, on the other hand, use membranes
to effect a salt removal and therefore are directly
affected by feedwater salinity. Calculations
were made at the base input condition with
various levels of input salinity to evaluate this
factor. The results are plotted in Figure 2, where
the assumption is made that individual ionic
species do not create limiting conditions.
200
:
100
20 4O 60 80 100
Product Water Capacity, m'/doy x I0~s
120
Figure 1. Illustration of desalting costs as functions
of desalting capacity for the base condition.
In all the methods simulated in the costing
model (except electrodialysis) the costs at 2000
mg/1 would be double those at 5000 mg/1 and
five times that at 13,000 mg/1. However, the rate
of change of the salinity versus cost ratio
diminished toward increasing salinity values.
The electrodialysis process is significantly less
affected by feedwater salinity mainly because of
its modular construction and the direct rela-
tionship between power consumption and
salinity.
Operation and Maintenance
Unlike the fairly predictable construction
cost items, operation and maintenance costs are
subject to year to year inflational pressures
which cannot always be effectively predicted.
At the base condition, the O & M costs were
typically somewhat more than 50% of the total
annual costs, as shown below.
206
-------
OPTIMIZING SALINITY CONTROL
£ 1.5
1.0
EL'
MSF, VTE-MSF,
VC-VTE-MSF, RO
VF-VC-
0 5,000 10,000
Salinity Concentration in Feedwater, ma./I
Figure 2. Effect of feedwater salinity on desalting
costs.
Process
Percent of Total Annual Cost
Attributable to O & M
MSF
VTE-MSF
VC-VTE-MSF
ED
RO
VF-VC
60%
51
46
58
56
56
It is interesting to evaluate the effects of
interest rate on the relative importance of
operation and maintenance costs. For example,
if the interest rate was increased from 7% at the
base to 10.5%, O & M costs decline from more
than 50% to 39%, or about 11% in most cases.
In terms of total costs, this 50% increase in
interest rates increased unit costs by a low of
21.5% for MSF processes to 28.6% for VC-VTE-
MSF systems. In each case studied, the effects of
scale on feedwater salinity did not affect the
percentages given above with the exception of
electrodialysis. In the ED analysis, the effects of
plant capacity produced O & M precentages
ranging from 58% to 66% as the capacity in-
creased from the base to 121,000 m3/day.
Specific items in the operating expense
categories are also of interest to water quality
management planning. For instance, electrical
and fuel costs accounted for the following
percentages of total annual costs:
Process
Cost of
Electrical Power
Cost of Fuel
MSF
VTE-MSF
VC-VTE-MSF
ED
RO
VF-VC
9.8%
5.4%
2.1%
27.5%
20.4%
32.8%
32.0%
33.4%
24.0%
0 %
0 %
0 %
Rate increases for both electricity and fuel
produce proportional increases in annual costs.
For example, in a MSF system if electricity rates
increase 50%, annual costs will increase (0.50)
(0.98) = 0.049, or 4.9%. These data illustrate the
importance of electrical costs for membrane and
freezing processes and fuel for distillation
methods of desalination.
Land
Because land area for desalting systems is a
non-depreciating capital costs and can there-
fore be amortized indefinitely, land costs for the
base condition ($5,000/ha) account for only 1-8%
of total annual desalting costs as shown below.
Process
Land Costs as a Fraction of
Total Annual Costs
MSF
VTE-MSF
VC-VTE-MSF
ED
RO
VF-VC
3.8%
3.9%
4.8%
5.8%
7.7%
0.7%
Feedwater and Brine Facilities
Of the factors involved in evaluating
desalting feasibility, the feedwater and brine
disposal facilities may be the most site-specific
variables. In the reference situation using
feedwater wells 100 m deep and 1000 m from the
plant, and brine injection wells 100 m deep and
100 m from the plant, and brine injection wells
207
-------
RETURN FLOW MANAGEMENT
1000 m deep and 100 m from the plant, the costs
were as follows:
Percentage of Total Annual Costs
Attributable to Feedwater and
Process Brine Disposal Facilities
MSF
VTE-MSF
VC-VTE-MSF
ED
RO
VF-VC
21%
21%
25%
30%
40%
12%
Differences in the values reflect different
volumes of feedwater, brine, and overall costs.
The choice of a feedwater or brine disposal
system is very important in evaluating these
alternatives. Feedwater wells, for example, can
account for about 15% of the costs of feedwater
and brine disposal systems. Likewise, the choice
of brine injection wells over evaporation ponds
can be a significant decision. In the base
example, evaporation ponds would cost about
70% more than injection wells for each alter-
native except VF-VD in which case they would
cost approximately 7% less. Thus the solution
and location of this equipment should be con-
sidered and optimized for each potential loca-
tion.
Technical Advances
The desalting submodel does not include
compensation for technological improvements
in equipment or processes that are almost
certain to appear. The potential user of this
work should therefore be cognizant that sub-
stantial errors can be introduced if updating
with current literature is not part of using this
desalting submodel.
AGRICULTURAL SALINITY CONTROL
COSTS
Improved management and structural
rehabilitation are often regarded as the most
feasible treatments of an irrigation system to
improve the quality of return flow. Indirect
approaches such as limiting irrigation diver-
sions, effluent standards, land use regulations,
and economic incentives may also be considered
although they appear more difficult to imple-
ment. Whenever salt pickup is the objective of
salinity control, however, the specific control
measures should impact segments of the irriga-
tion system which contribute to the magnitude
of local groundwater flow. This may be ac-
complished by reducing seepage from various
elements of the conveyance system and
minimizing deep percolation from over-
irrigation. Return flow quality may also be
improved by relief and interceptor drainage to
collect subsurface flows before a chemical
equilibrium is reached with the ambient soil or
aquifer materials. The respective feasibility of
individual measures depends on developing and
evaluating cost-effectivenss functions for con-
trolling each alternative input to the
groundwater region where salt pickup is as-
sumed to occur. These are based on soil and
aquifer chemical behavior as interpreted by
prerequisite analyses. Thus, the relationship
between groundwater flow and salt loading can
be determined in such a manner that control
costs can be related to reductions in salt loading.
The mathematical development of the
agricultural salinity control cost-effectiveness
functions is too lengthy for this presentation.
Since canal and ditch linings are among the
more popular salinity control measures and
demonstrate the pertinent concepts, the follow-
ing discussion will be limited to this salinity
control alternative.
Water Conveyance Linings
The contribution to local or regional salini-
ty problems from irrigation water conveyance
networks may be the result of a number of
factors. First, unlined channels allow seepage
into underlying soils and aquifers where
naturally occurring salts might be dissolved
and transported into receiving waters. Second,
the structural and managerial condition of a
system may support large areas of open water
surfaces or phreatophytes which concentrate
salinity in return flows (through evaporation
and transpiration). Finally, the operation of the
system may preclude efficient water utilization
by the individual irrigator, especially if
deliveries are not made in accordance with crop
demands. The costs to remedy these problems
are generally limited to those associated with
lining and rehabilitation of the waterway.
However, so far as improved management may
be required, some costs associated with educa-
tional programs and legal/administrative ad-
justments may be incurred.
To test the feasibility of conveyance system
improvements relative to other salinity control
measures, the mathematical model developed in
the following paragraphs will center on one
principal channel improvement alternative;
208
-------
OPTIMIZING SALINITY CONTROL
concrete lined systems. It is assumed that this
alternative along with supportive structures
represent the generally applicable technology
in therms of both utilization and cost.
Seepage from canal, lateral, and ditch con-
veyance networks may be reduced or eliminated
by lining the perimeter with an impervious
material such as concrete, plastic, asphalt, or
compacted earth to note several of the more
common methods. Concrete is probably the
most commonly employed lining material
because of the combined advantages of cost,
ease of construction, availability, reduced
maintenance, and low permeability. The costs
of concrete linings (either slip-form or gunite)
vary with local economic and topographic con-
ditions, channel geometry and size, and re-
quirements for miscellaneous water manage-
ment, safety, and environmental structures. For
specific locations it is important to prepare cost
estimates on a case by case basis, although for
planning purposes it is useful to have general-
ized expressions.
A review of concrete lining costs by Walker
(1976) indicated that such costs could be
reasonably well estimated as a function of
wetted perimeter and updated to present and
future conditions with an appropriate cost in-
dex. A simpler methodology based on design
discharge can also be utilized, and will be
variable in this model.
Data presented by U. S. Department of the
Interior (1963, 1975) and Evans, et al. (1976)
were evaluated by the following general
relationship:
where,
U c = unit lining cost, $/m;
Q = design discharge, m3 /sec; and
K1, K2= regression coefficients.
After transforming the data with the Bureau of
Reclamation canal and earthwork cost index to
a base time of 1975, the K i and K2 values were
29.70 and 0.56, respectively. It should be noted,
however, that even in the same locale these unit
costs vary substantially. Equation 3, therefore,
is intended only as a general estimating for-
mula. The unit costs in Equation 3 include only
the earthwork, relocation, and lining costs and
do not include costs for fencing, diversion
structures, safety structures, etc. The latter
costs are also highly variable depending upon
the many site-specific conditions. An examina-
tion of such costs as given by the U.S. Depart-
ment of the Interior (1975) showed a range of
$12/m to $50/m with a average of $22/m.
Unless otherwise specified, the average figure
will be used in this analysis. Consequently, the
construction cost may be written in 1975 dollars:
C c = 29.70Q °-56 + 22
(4)
In addition to the construction costs, one
must consider service facilities, engineering,
investigations, and other adminstrative ex-
penses. The Bureau of Reclamation uses a 35%
factor for these costs, so Equation 4 can be
written as a total cal cost:
C c = 40.1Q °-56 + 29.70 (5)
where,
C c = 1975 value of concrete canal linings in $/m.
In order to calculate the total costs for a
given length of canal, ditch, or lateral, Equation
5 must be integrated over the applicable limits.
Because water is continually being withdrawn
from a conveyance channel, both wetted
perimeter and discharge decline along the
length of the channel. The distribution of these
parameters may be estimated in a number of
ways if measurements are not available. For
instance, a linear decline can be assumed, or the
decrease can be formulated in terms of acreage
distributions. Assuming a linear decrease, the
wetted perimeter at a specific location can be
determined by:
WP = WP m (1-bL/L t) (6)
in which,
WPm= the wetted perimeter at the channel inlet,
m;
L = length from inlet to specified point, m;
L t = total length of channel, m; and
b = empirical constant representing the frac-
tion of maximum wetted perimeter re-
maining at end of the channel.
and similarly for the design discharge:
Q = Q m (1-bL/L t ) (7)
where,
Q m - inlet channel capacity in m3 /sec.
Combining Equations 5 and 7, yields,
C c = 40.1 x Q m°-56 (1-bL/L t) °-56 + 29.70
(8)
209
-------
RETURN FLOW MANAGEMENT
Then, the total capital costs for lining L meters
of channel, Cc, are determined by integrating
Equation 8 as it varies from 0 to L meters:
C c = -JO.lxQ ,
0.56
_LL_
1.56b
1-il-bL Lt )
- 29.7L
(9)
Equation 9 assumed that the lining proceeds in
the downstream direction, however, the choice
of either upstream or downstream lining direc-
tion depends on their relative cost-effectiveness.
To determine this choice, it is first necessary to
define the salinity control effectiveness result-
ing from a particular lining project. Only the
salt loading effects wili be considered in the
model at this stage of development. If a salt
pickup condition exists, the equilibrium salinity
in return flow will differ from the salinity in the
seepage water so that after accounting for
phreatophyte withdrawals,
AS
= AS c V
Qe - Q
(10)
Q
g
AS
Qg
QP
v
where,
AS l = reduction in salt loading due to the lin-
ings, metric tons m annually;
= difference between the equilibrium salini-
ty concentration in the return flows and
the salt concentrations in the seepage
seepage waters, mg 1;
= total groundwater addition, m< year;
= phreatophyte use of groundwater, m*
year; and
= total volume of seepage, m'A year, as de-
termined by:
Vs = Nd ASR Wp< (11)
in which,
N d = number of days per year seepage occurs:
\VP' = wetter perimeter of original channel m;
and
ASR = change in seepage rate affected by lining,
ms m2 day.
It might be pointed out that the ASR value
might also be written as a length distributed
parameter in this model if measurements or
other data are available.
Substitution of Equation 6 into Equation 10
leads to an expression than can be integrated
over the length of the lining to determine the
total salinity control. This expression for the
downstream lining case is:
SP = AS,
10- N d ASR WP'm —- U-U-bL L t)-) (12)
2b
This question of lining direction can be ad-
dressed by approximating the marginal costs at
both ends and then comparing the results. First,
let K i, K2, and KS be defined respectively as:
K ! = 40.1 Q
0.56
m
(13)
Qc - QD
K2=ScNd ASR g P-WFm 10-* (14)
Q
g
K 3 = 29.7 (15)
Then the marginal cost estimate would be found by
dividing Equation 9 by Equation 8;
MC= KKl-bL/Lt)0*"
(16)
K c, (l-bL/Lt )
4& V
in which,
MC = marginal lining cost estimate, $/ton.
At the inlet where L=0, the marginal cost estimate,
MC \ is:
" 1 ^ K^ (17)
K2
whereas at the end where L=L t, the marginal cost
estimate MC u , is:
MC u =
(18)
K2(l-b)
Subtracting Equation 18 from 19 and simplifying
gives:
MCi-MCu= 2LL (1--
K3
K2
(l-b)
0.44
(1-
(19)
(1-b)
210
-------
OPTIMIZING SALINITY CONTROL
Under the assumptions of this model, Equation
19 is always negative, leading to the conclusion
that lining should always proceed downstream.
Consequently, the cost-effectiveness functions
for concrete linings is determined from
Equations 9 and 12 by solving for the independ-
ent variable L in Equation 12 and substituting
the resulting expression into Equation 9. The
resulting cost-effectiveness function is:
0.78
cc =
in which,
Ki =
Ko
0.5
-if, (20)
Ki Lt
1.56b
K2Lt
2b
(21)
(22)
COMPARISON OF TECHNOLOGIES
The preceeding paragraphs demonstrate
the complexity of determining optimal irriga-
tion return flow quality controls. These
problems of complexity can be significantly
alleviated by "staging" the calculations (also
known as multi-level optimization). This
procedure can take several forms, but its ad-
ditive format is the easiest to understand. From
the array of agricultural alternatives, the op-
timum set is determined as a function of control
level. The same is determined for the desalting
alternatives. And finally, agricultural and
desalting cost-effectiveness functions are com-
bined to produce the optimal overall strategy.
To demonstrate these concepts, desalting will be
compared to canal lining for the conditions in
Grand Valley. Since the Grand Valley situation
is a case study in these proceedings, the descrip-
tion of local hydro-salinity systems will be
omitted from this paper.
Optimal Canal Lining Strategies
There are fourteen major canal and ditch
systems in the Grand Valley ranging in length
from 74 kilometers for the Government Highline
Canal (17 m3/sec capacity) to 4 kilometers for
the Mesa County Ditch (1 m3/sec capacity).
The pertinent parameters for each canal, along
with the seepage contribution to salt loading,
were determined for use in Equation 20. The
resulting functions were then minimized using
the Jacobian Differential Algorithm developed
by Walker, et al. (1973) for a range of possible
salinity reductions desired from a canal lining
program. These results are given in Figures 3
and 4. Figure 3 shows the total capital construc-
tions costs as a function of the annual salt load
reduction to be realized. The upper curve is the
minimum cost associated with each value on the
abscissa. Underneath the upper curve are the
costs attributed to the various valley wide
canals. For example, if the contribution of canal
seepage to the salt loading problem was to be
reduced by 87,500 tons annually through
linings, the capital construction cost would be
approximately $27 million with $13.5 million on
the Government Highline canal, $8.7 million
on the Grand Valley system, $2.2 million on the
small ditches (e.g., Price, Stub, etc.), $1.6 million
on the Orchard Mesa System, and the
remainder on the Redlands system. Figure 4
illustrates how much of the respective canal
systems would be lined under this minimum
cost policy.
Redlands System
Q
D
3
0
40
50 60 70 80 90 100
Annual Salt Load Reduction, metric ton/xlO~
110
Figure 3. Optimal canal lining strategy in the
Grand Valley of western Colorado.
211
-------
RETURN FLOW MANAGEMENT
Grand Volley,
Grand Volley Mrtnlirw
and
Grand Valley Highline
Kiet Eit.,
Indp. Ranch,
Price. Stub
County
0 I02O3O4O9O6O7O8090
Percent of Canal Seepage Contribution to be Eliminated
Figure 4. Canal lining implementation strategy for
the Grand Valley example.
The results obtained in optimizing canal
lining policies is interesting in the sense that
they demonstrate the need to initiate linings on
more than one segment of the conveyance
system when full scale implementation begins.
The cost savings in implementing according to
a minimum cost strategy amount to about 30%
in this case (Walker, 1976).
7r
IOO 2OO JOO
Annual Solt Tonoo* Remo«ed, mtontxIO""
Figure 5. Cost-effectiveness of desalting in Grand
Valley.
Desalting Policy
Desalting evaluations involve, first, deter-
mining the most effective process and, second,
the best feedwater-brine facilities. In the Grand
Valley situation, reverse osmosis is clearly the
most advantageous. A summary of desalting
costs as a function of salt load removal is given
in Figure 5.
It might be noted that whereas agricultural
salinity control costs (at least for canal and
ditch linings) exhibit increasing marginal costs
with scale, the reverse is true for desalting
technologies. Consequently, early improve-
ments (or ones of small scale) favor agricultural
improvements while larger operations enhance
the feasibility of desalination.
Combined Agricultural-Desalting
Methodologies
The respective feasibility of agricultural
and desalting methods is determined by op-
timizing the individual contributions from each
aggregated analysis. In the example of canal
lining and desalting by reverse osmosis, this
would imply combining Figures 3 and 5 in an
optimal manner.
The optimal canal lining-desalting strategy
for the Grand Valley is given in Figure 6 for
seven values of interest rates. (During the
analysis, the parameter having the largest
single influence on the respective feasibility
was interest rates for project evaluation). In-
terest rates, of course, are not the only important
factor. For example, Figure 7 was prepared
using canal lining costs 50% higher than com-
puted. It is obvious the desalting would be
employed more in a plan utilizing these results,
but the reverse would be true if desalting cost
estimates turned out to be conservative.
SUMMARY AND CONCLUSIONS
An attempt has been made in this paper to
evaluate the possible role of desalination
technology in controlling the quality of irriga-
tion return flows. The Grand Valley in western
Colorado has been utilized to demonstrate the
concepts proposed herein, but the space
available for this paper prohibited an examina-
tion beyond simply evaluating the respective
feasibility of canal lining and desalting. The
results are indicative of the need to develop
optimizational modeling analyses as part of
salinity control planning on both a local and
regional scale.
212
-------
OPTIMIZING SALINITY CONTROL
Canal Lining Total Canol Lining Cost = $ 40 million
Desalting
cv
2 80
a 60
6% 7% 8% 9% 10% 12% 15%
Interest Rote
Figure 6. Comparison of canal lining and desalting
in reducing salt loading from the Grand Valley.
Canol Lining
Desalting
Total Canal Lining Cost = $60 million
BC
I 60
S 40
20
6% 7% 8% 9% 10% 12% 18%
Interest Rate
Figure 7. Comparison of desalting and canal lining
salinity control alternatives assuming 50% increase
in lining costs.
Desalting is a viable alternative to lining
canals, laterals, and head ditches. Its relative
feasibility cannot be ascertained generally, but
must be evaluated case by case. However, in the
areas where salt pickup rates are high,
desalting will enjoy its most favorable advan-
tage even though desalting is likely to be
implemented after a substantial investment in
agricultural improvements.
REFERENCES
1. Evans, R. G., S. W. Smith, W. R. Walker,
and G. V. Skogerboe, 1976. Irrigation Field
Days Report 1976. Agricultural Engineering
Department, Colorado State University, Fort
Collins, Colorado. August.
2. Prehn, W. L., J. L. McGaugh, C. Wong,
J. J. Strobel, and E. F. Miller. 1970. Desalting
Cost Calculating Procedures. Research and
Development Progress Report No. 555. Office of
Saline Water, U. S. Department of the Interior,
Washington, D. C. May.
3. Probstein, R. F. 1973. Desalination.
American Scientist. Volume 61, No. 3, May-
June, pp 280-293.
4. U.S. Department of the Interior, Bureau
of Reclamation. 1963. Linings for Irrigation
Canals. Denver Federal Center, Denver, Colo-
rado.
5. U.S. Department of the Interior, Bu-
reau of Reclamation and Office of Saline Water.
1972. Desalting Handbook for Planners.
Denver, Colorado. May.
6. U.S. Department of the Interior, Bu-
reau of Reclamation and Office of Saline Water.
1973. Colorado River International Salinity
Control Project, Executive Summary.
September.
7. U.S. Department of the Interior, Bu-
reau of Reclamation. 1975. Initial Cost Es-
timates for Grand Valley Canal and Lateral
Linings. Personal Communication with USER
Personnel in Grand Junction, Colorado.
8. Walker, W. R. and G. V. Skogerboe.
1973. Mathematical Modeling of Water
Management Strategies in Urbanizing River
Basins. OWRR Completion Report 45, En-
vironmental Resources Center, Colorado State
University, Fort Collins, Colorado. June.
9. Walker, W. R. 1976. Integrating Desali-
nation and Agricultural Salinity Control
Technologies. Paper presented at the Inter-
national Conference on Managing Saline Water
for Irrigation. Texas Tech University, Lubbock,
Texas. August.
213
-------
Irrigation Return Flow Models
-------
Practical Applications of Irrigation
Return Flow Quality Models to
Large Acreages
MARVIN J. SHAFFER and RICHARD W. RIBBENS
U.S. Bureau of Reclamation;
Engineering and Research Center; Denver, Colorado
ABSTRACT
Numerous return flow quality simulation
models have been developed in recent years.
Practical use of these tools on large irrigation
projects is discussed in terms of study objec-
tives, data requirements and availability, model
detail and sensitivity, inherent limitations, and
reliability and interpretation of results. The
modeling efforts of the Bureau of Reclamation
and other institutions are used to illustrate the
utility and practicability of these new tools.
Potential and present model users should gain
valuable insight into the benefits and pitfalls of
return flow quality modeling.
INTRODUCTION
There has been a proliferation of irrigation
return flow simulation models in recent years in
terms of both basic development (e.g., 5,6,8) and
practical application (e.g., 6,10,11) to real world
problems. These models range from simple
conceptual formulations to complex programs
requiring a large digital computer. Although a
wide variety of techniques and models are
generally available to water resource manage-
ment and development agencies, consultants,
environmental organizations and others, they
sometimes fail to utilize the most appropriate
procedures. A rapidly advancing technology
together with a general lack of any real
guidelines for approach/model selection and
useage are among the causes. The situation is
compounded by the lack of a centralized source
of return flow model codes and documentata-
tion. As a result, studies are often completed and
management decisions made with less than
optimal useage of available tools.
This paper attempts to present a realistic
approach to prediction studies involving irri-
gation return flow quality and quantity, par-
ticularly from large areas. Return flows include
surface and subsurface components such as
surface runoff, operational losses, seepage from
water conveyance systems, and subsurface
drain returns. Quality parameters can include
concentrations of major cations and anions,
total dissolved solids, pesticides, nitrates,
phosphates, trace elements, dissolved oxygen,
BOD, temperature, etc. Suggestions are includ-
ed for needed improvements in dissemination of
information on models and modeling techni-
ques. Following the suggested study approach
will better enable water resource agencies and
others to make decisions involving irrigation
return flows.
Basic Approach
Studies involving predictions of irrigation
return flow qualities and quantities should
follow an approach which allows efficient
utilization of available tools and techniques, but
still affords a suitable feedback mechanism for
making needed changes during the study. Feed-
back is a vital consideration due to the many
uncertainties present at the beginning of any
irrigation return flow (IRF) study.
From the outset, a specific objective should
be identified and a study plan established which
states what is to be accomplished and deter-
mines the resources necessary to complete the
study. Far too many IRF studies are started
without any clear idea of what is to be done,
what resources are needed, and how the infor-
mation will be used. Although the study plan
217
-------
IRRIGATION RETURN FLOW MODELS
may be altered as the work progresses, an initial
plan is required to provide a starting place.
The three essential or key elements of any
good return flow study plan are illustrated in
Figure 1. These include the available models
and approaches, the data requirements and
availability, and sensitivity analyses. To be
functional, the key elements must interact with
each other and with the original objective.
Many problems associated with IRF studies can
be traced to absent or unreasonable objectives,
failure to consider one or more of the key
elements, or failure to initiate and maintain an
effective feedback mechanism.
Avoiloblt
Models •' Approoches
Study Objectnes
(eipected results.
required resources,
occurocy)
Doto Rtqwrements /
Avoiiobility
PRACTICAL STUDY PLAN
Figure 1. Study Plan Analysis
A practical study plan should evolve from a
realistic objective which has considered the key
elements, which in turn have been interacted
with each other.
Available Approaches/Models
A wide range of techniques and models are
currently available for use in IRF studies. As
illustrated in Figure 2, the practitioner working
in the field tends to favor models or approaches
with the least amount of technical detail. Con-
versely, theorists generally prefer models which
contain many micro-details. The applied
modeler who is (or should be) looking for the
simplest model or approach that does the job,
often finds himself between these two extreme
viewpoints. This position, as shown by the solid
line in Figure 2, is the most likely to produce a
practical model which still considers the impor-
tant variables. Depending on the particular
application, the detail of the model or approach
used could fall within a wide range. The real
task lies in selecting the appropriate technique
or model from the available spectrum.
Actual model selection should begin with
the study objective and progress via the interac-
tive or feedback process shown in Figure 1. A
thorough knowledge of the general field con-
ditions and the types and scope of available
tools is essential. Field personnel with site
specific knowledge of the individual project area
should be involved in the IRF study. Otherwise,
it will be extremely difficult to know which
approach or model to apply.
Figure 2. Probability of having the best model for a
given application.
Available tools should be reviewed to deter-
mine the most appropriate model or technique.
Unfortunately, only a handful of modelers in
the United States and elsewhere are aware of
the latest techniques available in this rapidly
expanding field. These individuals have their
own specific versions of IRF models. Field
personnel (modelers) wanting to apply IRF
techniques are usually forced to select an ex-
isting model or to resort to outdated techniques.
It is extremely difficult for someone to readily
adapt parts of various IRF models to individual
needs due to the lack of a centralized location for
source codes and documentation. Consideration
should be given to establishing such a central
library of IRF models or submodels. Another
highly beneficial information source would be a
comprehensive manual describing the methods
and techniques for conducting irrigation return
flow quality and quantity studies. No such
manual currently exists, and the practitioner is
forced to improvise based on in-house
capabilities.
Various attempts have been made to group
and classify IRF models. Designations such as
stochastic and deterministic, dynamic and
steadystate, conservative and nonconservative,
analytic and numerical are all commonly used,
218
-------
APPLICATION OF MODELS
and IRF models often incorporate several of
them. All models have certain inherent
limitations which affect the way they can be
used. For example, a steady-state, conservative
model would be extremely difficult to apply to a
problem requiring dynamic projections of non-
conservative variables. Likewise, many IRF
models utilize representative sets of data defin-
ing initial soil and aquifer conditions. The
model user cannot input continuous data, and is
limited to and must interpret a few represen-
tative cases.
The study objective together with the par-
ticular field situation determine to a great
extent the approach or model which can be used.
If the objective is to get reconnaissance level
estimates from an area with limited data, then a
simple conceptual model or experience in
similar areas may be the best approach. Con-
versely, long term monthly concentrations of
nonconservative constituents in return flows
from a project could require a dynamic, non-
conservative numerical model.
In any event, selection of a suitable model or
approach should consider the basic study objec-
tive, the field situation, variable sensitivities,
and data requirements and availability. It could
very well be necessary to alter the model or
approach during the study as feedback becomes
available.
Sensitivity Analyses
The sensitivity analysis has been included
as an essential element of any IRF study plan.
Models or approaches are sometimes used
which fail to properly consider the significant
variables, while other variables are handled
with too much detail. Every aspect of a study
can benefit from a sensitivity analysis. The
model/approach used, the data requirements,
and the basic objective itself can hinge on the
results of such an analysis.
In some cases, the entire study may be
centered around the sensitivity aspects. A deter-
mination of output variability as a function of
variability in the inputs could provide sufficient
information to make management decisions.
It should be pointed out, however, that
sensitivity studies can be overdone and abused.
Since the true future variability and interac-
tions of the input parameters are seldom known,
any attempt to use sensitivity analyses to draw
boundaries or zones of uncertainty around
predicted results should be viewed with caution.
Nevertheless, sensitivity analyses should be an
important part of the study plan.
A relative error diagram similar to that
shown in Figure 3 can be used to compare the
sensitivities of several input variables. Many
variables will plot similar to variables 1 and 2 in
the figure. That is, predicted results are not
found to be highly sensitive to any one input
parameter. Variables with slopes much greater
than var. 4 would require special attention. In
an actual application, the effects of many
variables tend to cancel each other giving an
overall attenuated effect similar to the com-
posite.
Var 4
Var 3
Var. 2
„ Composite
Vor. I
K) 20 30 40 50 60 70 80
RELATIVE ERROR OF INPUT DATA
|PERCENT|
Figure 3. Variable Sensitivity
Data Requirements/Availability
Data collection programs on large projects
all suffer from the same problem; there are never
enough resources to do high density sampling
on the entire project. The data requirements of
various models/approaches must be considered
in light of the availability of data. Indeed, the
basic objective of the study must often be
tempered to reflect the available data.
The relationship between data re-
quirements and availability is shown in Figure
4. Generally, models must be selected which are
between the extremes.
Many IRF models utilize a number of in-
dividual data sets which represent conditions
found in the project area. An effort should be
made to minimize the number of data sets and
computer runs. One possibility is to establish
basic mapping units within the project and
219
-------
IRRIGATION RETURN FLOW MODELS
more intensely study a representative portion of
each unit. It is often possible to correlate in-
dividual parameters with certain key
parameters so that extensive sampling and
analytical procedures are unnecessary for each
sampling location.
In all cases, the costs of data collection
programs should be weighed against the
penalties (costs) of not collecting data (see
Figure 5). For example, the risk of downstream
pollution could outweigh the cost of collecting
more data from the project. In any event, data
monitoring programs should be extended
beyond the time of the initial IRF studies to
further verify the projections and to allow for
any remedial actions which might be necessary.
DATA
AV»I LABILITY
Arm erf Compromise
DATA REQUIREMENTS
Figure 4. Data Requirements and Availability
Interpretation of Results
The manner in which the results of an IRF
study are interpreted sometimes causes con-
siderable confusion. A hypothetical time plot of
total dissolved solids (TDS) concentrations in a
stream below an irrigation project is shown in
Figure 6. Depending on which parts of the plot
are examined, varying conclusions can be made
about the impace of the project on the stream.
The "with project" curve indicates that the
stream will never exceed a total dissolved solids
concentration of 500 mg/liter which is equal to
some state water quality standards for drinking
water. However, during the maximum impact
period, stream salinity is increased as much as
four or five fold. At steady state, the salinity
increase is only about 50 percent.
It should always be recognized that con-
fidence bands surround both the projected and
base curves in Figure 6. If these bands are too
wide, the results become meaningless. While it
is currently beyond the state-of-the-art to com-
COSTS
Cost (Penalty) of not
collecting data
VOLUME OF DATA
Figure 5. Cost of Data Collection (or lack thereof)
pute the location of say the 95 percent con-
fidence bands on the predicted curve, research
in this area could yield valuable results. The
primary problem is the inability to predict
future economic and social conditions which
control certain inputs to the modeling process.
800
700
_ 600
|500
g «0
3 *»•
E 200
•n
100
Moiimum impoct
period
Sttody-rrott
I 2 34 5 6789 10 II 12
TIME (TEARS)
Figure 6. Model results
Further complications to Figure 6 occur
when the receiving streams experience low
flows or go dry periodically. In the Western
United States, high total dissolved solids con-
centrations are associated with low flows.
Although the mean or median concentrations
may be increased by irrigation return flows,
maximum concentrations are often reduced.
The question of whether a project will cause
adverse impacts downstream is always difficult
to answer. All factors must be examined in a
rational manner.
Whenever long term results of an IRF study
are presented, the entire picture should be given
including the base conditions, the maximum
impact period, and steady state. Even with
220
-------
APPLICATION OF MODELS
short term projections, the base curve should be
presented along with the predicted values.
Common Problems in IRF Studies
Studies involving irrigation return flow
quantities and qualities are often complex and
involve several scientific and engineering dis-
ciplines. Most organizational difficulties with
IRF studies arise because the efforts are not
properly coordinated, and sufficient time and
resources are not committed to the complex
task. Table 1 lists some of the more common
mistakes often made in IRF studies. Most of
these could be avoided if an approach such as
the one suggested in this paper was followed.
The proper attitude towards IRF studies,
particularly those dealing with large areas, is
essential to a successful study. The underlying
purpose usually is (or should be) to obtain
additional information which can be used to
help make a management decision. Results with
very high accuracy levels generally are not
possible nor are they necessary in most cases.
Strong Points of Current IRF Study
Techniques
Computer simulation models are finding
increased useage in IRF studies. While
problems still exist, such as lack of extensive
verification and "oversell" of computer predic-
tions, the techniques represent a major ad-
vancement in IRF technology. It is now possible
to simultaneously consider many of the impor-
tant variables which affect return flow qualities
and quantities. Previous "desk top" techniques,
while valid within the context of thier assump-
tions and limitations, could not hope to handle
the complex interactions now being simulated.
In addition to pure computing power,
simulation modeling forces the practicing scien-
tist and engineer to look at the overall system
rather than extrapolate from his often narrow
area of expertise. This means that other dis-
ciplines are more likely to be included in the
study. An interdisciplinary approach to the IRF
study provides a forum for highly beneficial,
although sometimes frustrating, exchanges
between the respective fields.
A third benefit of simulation modeling
stems from the modular nature of most IRF
model computer codes. This allows (or should
allow), subroutines developed for a particular
subsystem to be used as part(s) of other models.
TABLE 1
Common Mistakes Made in IRF Studies
1. Setting objectives which are not ob-
tainable within the confines of available
resources (dollars, time, manpower, etc.)
2. Failure to allow sufficient time for the
necessary feedback mechanisms to operate.
(at least 3 times the single run through
estimate should be allotted).
3. Conducting studies using great detail on
insensitive parameters. Not knowing which
parameters are really sensitive.
4. Data requirements exceed data
availability; insufficient time allotted for
data collection.
5. Use of the local approach or model
without considering other possibilities.
6. Failure to consider or use feedback from
on - going activities.
7. Adopting the viewpoint that results from
computer simulation models are either all
powerful or completely useless.
8. Overplay of results; failure to identify the
weaknesses; over emphasis on data suppor-
ting a single viewpoint.
9. Empirical extrapolations based on years
of experience rather than applying a sound
scientific approach.
10. A modeling staff which is not qualified
in certain aspects of the study.
11. Not knowing whether to use a long term,
short term, or steady state approach. Over
reliance on any one of these.
12. Crash programs to study irrigation
return flow quality and quantity without
adequate time for data collection and other
study aspects.
Thus, for example, solutions to complex
problems in soil chemistry can be used by soil
physicists without need for detailed involve-
ment in a less familiar area.
IRF Model Applications
The use of irrigation return flow models to
predict the quantity and quality of return flows
from large areas or projects has become fairly
221
-------
IRRIGATION RETURN FLOW MODELS
common in recent years. The Bureau of
Reclamation has been studying the long term
impacts of its projects on the environment.
Rather sophisticated computer simulation
models (3, 4, 5, 9) are being used to assess the
quantity and quality of irrigation return flows
and the effects of these waters on receiving
streams. Specific applications have included
the Garrison Diversion Unit (11), the Oahe Unit
(10) the Dolores, Animas-La Plata and San
Miguel Projects, the San Juan Collector System,
the Wellton-Mohawk Project (9), the Navajo
Indian Irrigation Project, and the San Luis
Drain.
The University of California at Davis has
been involved in modeling irrigation return
flows from large areas. Predictions have been
made for the Glenn-Colusa Irrigation District
and the Panoche Drainage District near
Sacramento, California (6) and the Sutter
Basin, California (7).
New Mexico Institute of Mining and
Technology, Socorro, New Mexico, has studied
irrigation return flows in the Mesilla Valley,
New Mexico (2).
Utah State University has been involved in
several IRF modeling studies on large acreages
including the Bear River Basin (1,8) and other
areas (12).
SUMMARY AND CONCLUSIONS
Useful IRF models depend on past research,
present technology and knowledge, and con-
tinuing assessments based on comparisons of
predicted and observed responses of operational
projects. Wise model use depends on a thorough
understanding of the model capabilities and
limitations. Successful application in turn
depends on an adequate data base.
Data collection, basic research, model
development, and model application are in-
terdependent activities that require continuous
feedback for maximum success. Thus basic
research provides a conceptual framework for
developing the model, the model aids in the
design of a data collection program, and the
data are used to apply the model to a given area.
The results of this application may dictate
additional research, modification of the model,
and an expanded data collection program, Ear-
ly activation of this feedback process will assure
maximum usefulness of any IRF model.
IRF studies involve a broad spectrum of
disciplines including researchers, data collec-
tors, modelers, and decisionmakers represent-
ing such varied professions as geology, hydro-
logy, soil science, agronomy, and hydraulics. A
multidisciplinary approach is indicated which
requires substantial effort to establish effective
communications.
IRF studies often require large masses of
data. Even though simple models are used,
reliance on computers is natural. This in turn
introduces computer associated problems that
have nothing to do with irrigation return flows,
including documentation and maintenance of
source codes.
Development of a practical study plan is
essential to a successful IRF study. It is based
on the interaction of available models/ap-
proaches, data requirements/availability, and
sensitivity analyses with the study objective
and between themselves. The final product is a
result of tradeoffs and compromises. Proper
interpretation of results should then assure
improved decisions regarding irrigation return
flows.
REFERENCES
1. Hill, R. W., Israelsen, E. K., andRiley, J.
P., 1973. Computer Simulation of the
Hydrologic and Salinity Flow Systems in the
Bear River Basin. Report No. PRWG104-1, Utah
Water Research Laboratory, Utah State Univer-
sity, Logan.
2. McLin, S. G. and Gelhar, L. W., 1977.
Hydrosalinity Modeling of Irrigation Return
Flow in the Mesilla Valley, New Mexico.
Managing Saline Water for Irrigation.
Proceedings of the International Salinity Con-
ference, Lubbock, Texas, pp. 28-48.
3. Ribbens, R. W. and Shaffer, M. J., 1976.
Irrigation Return Flow Modeling for the Souris
Loop. Environmental Aspects of Irrigation and
Drainage. Proceedings of ASCE Specialty Con-
ference, Ottawa, Canada, pp. 545-557.
4. Shaffer, .M. J., 1977. Detailed Return
Flow Salinity and Nutrient Simulation Model.
Managing Saline Water for Irrigation.
Proceedings of the International Salinity Con-
ference, Lubbock, Texas, pp. 127-141.
5. Shaffer, M. J., Ribbens, R W., and
Huntley, C. W., In Press. Prediction of Mineral
Quality of Irrigation Return Flow. Volume V.
Detailed Return Flow Salinity and Nutrient
222
-------
APPLICATION OF MODELS
Simulation Model. U.S. Environmental Protec-
tion Agency, Ada, Oklahoma, 227 pp.
6. Tanji, K. K., 1977. A Conceptual
Hydrosalinity Model for Predicting Salt Load in
Irrigation Return Flows. Managing Saline
Water for Irrigation. Proceedings of the Inter-
national Salinity Conference, Lubbock, Texas,
pp. 49-70.
7. Tanji, K. K., Henderson, D. W., Gupta,
S., Iqbal, M., and Quek, A. F. 1974. Water and
Salt Transfers in Sutter Basin, California.
Paper No. 74-2029,1974 ASAE Annual Meeting,
18pp.
8. Thomas, J. L., Riley, J. P., and
Istaelsen, E. K, 1971. A computer Model of the
Quantity and Chemical Quality of Return Flow.
Utah Water Research Laboratory, Utah State
University. June. 94 pp.
9. U. S. Bureau of Reclamation, 1975.
Preliminary Sizing Study Yuma Desalting
Plant, Arizona. Special Report, Lower Colorado
Region. July, 33 pp.
10. U. S. Bureau of Reclamation, 1975.
Summary Report, Initial Stage Oahe Unit.
Water Quality and Return Flow Study.
September, 42 pp.
11. U.S. Bureau of Reclamation, 1976.
Water Quality Study. Garrison Diversion Unit,
North Dakota. June, 248 pp.
12. Utah State University, 1975. Colorado
River Regional Assessment Study, Part IV.
Prepared for National Commission on Water
Quality, Contract No. WQ5AC054. Utah Water
Research Laboratory. October.
223
-------
Areal Predictions of Soil Water
Flux in the Unsaturated Zone
A. W. WARRICK
Department of Soils, Water and Engineering,
College of Agriculture, The University of Arizona, Tuscon, Arizona
ABSTRACT
Irrigation, salinity and potential ground or
surface water pollutants are intimately related
to water and solute movement in the soilprofile.
Ideally, soil water fluxes and water composition
are managed in order to provide a desirable
plant growing environment and/or acceptable
quantities and qualities of drainage water. The
problem of predicting or averaging such fluxes
is very difficult due to the intrinsic variability of
the soil's physical and chemical properties.
In this study, predictions of soil water flux
are made accounting for the spatial variability
of the soil water parameters. Calculations are
made using field-measured distributions of con-
ductivity and soil water tension — water con-
tent relationships for Panache loam and Pima
clay loam soils. In the first set of simulations a
conductivity of the form K-K0 exp[a(6 -60)] is
utilized where 6 is the volumetric water content
and K0 and 00 the initial values of hydraulic
conductivity and water content for the wetted
profile. The a is an empirical constant for each
site and depth. Monte Carlo simulations are
used to find flux distributions with the
simplified drainage equation J £, = K0 (1 +aK0
t/l)-l. The flux JL is a random output depend-
ent upon the stochastic nature of K0 and a and
is found to be approximately log-normal for all
times studied.
In the second set of calculations, flux dis-
tributions are found using the more cumber-
some, finite difference solution to the nonlinear
moisture flow equation. This allows inclusion of
more realistic boundary conditions as well as
plant water uptake.
The results are a step towards finding the
most effective way of determining and ex-
pressing water flux as a function of time. Also,
addressed are reliability of areal predictions
and measurements.
Irrigation, salinity and the transport of
water-carried constituents are intimately
related to soil water flux in the unsaturated
zone. As soil is heterogeneous by nature, the
velocity varies from site to site within any
specified area. The nature of the areal distribu-
tion is basic in order to describe either mean
values or relative flow amounts occurring at
adjacent points. In this paper, we will first
review measured distribution patterns of water-
related, soil physical parameters. Then, we will
explore ramifications of such distributions with
regard to water flux in the unsaturated zone.
OBSERVED DISTRIBUTION
PATTERNS
Frequency distributions have been pre-
sented for a variety of water-related, soil par-
ameters by Nielsen, Biggar and Erh (1973).
Coelho (1974), Cassel and Bauer (1975), Biggar
and Nielsen (1976), and Warrick, Mullen and
Nielsen (1977a, 1977b). Other recently reported
studies closely related include that by Keisling
(1974), Carvallo et al. (1976), Baker and Bouma
(1976) and Peck, Luxmoore and Stolzy (1977).
Results for water contents measured under
ponded water conditions or for water retained at
given tensions have been found to be ap-
proximately normally distributed in several
cases. In Figure 1, results are shown from
Coelho (1974) for water retained at fifteen bars,
the value commonly chosen to correspond to the
permanent wilting percentage. The data is for
500 sampling points each at a 50 cm depth and
on a 4 m grid within a 76 x 96 m area of Pima
clay loam. The data points are for a class
interval of 0.6% water content gravimetrically.
The distribution is approximately normal with
a mean of 12.2 and a standard deviation of 2.0%
water gravimetrically. Values of Coelho taken
at 36 sites over the 87 hectares of the same Pima
225
-------
IRRIGATION RETURN FLOW MODELS
0' 15
000
6 9 12 15
15-BAR WATER RETENTION (%)
Figure 1. Frequency distribution for 500 measure-
ments of 15-bar water retention (After Coelho, 1974).
clay loam also were approximately normal as
was the bulk density and a porosity index
factor. Cassel and Bauer (1975) also found
approximately normal distributions for 15 bar
moisture retention over a 1.3 hectare area of
Maddock sand loam, but a tendency towards
skewness for Beardon silty clay. Nielsen,
Biggar and Erh (1973) observed volumetric
water content distribution for infiltration under
steady ponding to also be normal over 150
hectares of Panoche soil. Both of these last two
studies found a near normal distribution for
bulk density.
For several properties more directly related
to flow velocities, skewed distributions seem to
be more the rule than the exception. Figure 2 is a
cumulative frequency diagram of the logarithm
of the saturated conductivity K sat. by Coelho
(1974) for the 87 hectare area of Pima clay loam.
The straight line is for a log normal distribution
which, of course, is highly skewed.
Nielsen et al. (1973) determined the un-
saturated conductivity for each site of the form
K = K0 exp[a(0-00)] I1!
where Ko and 0O are values of the hydraulic
conductivity and water content under steady
ponded conditions (K o and 6 o are less than the
saturated values). The frequency distribution of
K o was found to be highly skewed over the 150
hectares studied. Estimates of the mean and
standard deviation, assuming a log normal
distribution, were 26.2 and 30.8 cm/day, respec-
tively, for the 180 cm depth. A combination
fractile diagram and cumulative probability
distribution for a is given as Figure 3 after
Warrick, Mullen and Nielsen (1977 a). Values
are plotted using both a and In a. If the plot
with a was a straight line, we would have a
normal distribution; if the plot of In a gave a
straight line, we would have a log normal
distribution. The general conclusion is that the
population was better approximated by a log
normal distribution. The estimate of the overall
mean and standard deviation was 75.2 and 67.8.
Figure 2. Cumulative frequency distribution of
logarithm of the saturated conductivity over 87
hectares of Pima clay loam (After Coelho, 1974).
In an effort to describe soil heterogeniety
with one stochastic parameter, Peck, Luxmoore
and Stolzy (1977) and Warrick, Mullen and
Nielsen (1977 b) utilized a similar media concept
after Miller and Miller (1955, 1956). Similar
media have pores and grains which differ in size
but have the same relative geometry including
equal porosity. Consequences are that at a given
water content the pressure head hr (cm) is
related to an average pressure head h m (cm) by
Ar hr = \ hm [2]
226
-------
PREDICTING SOIL WATER FLUX
2.0
C
i 0
4.
-1.0
-2.0
99
K
?
70 t
a
30 x
u
10
9
100
ZOO
300
Figure 3. Fractile diagram and a cumulative prob-
ability distribution for a and In a for 150 hectares of
Panoche soil (After Warrick et al., 1977a).
where Xr is a scaling factor and A is the
average scaling factor. Similarly, the un-
saturated hydraulic conductivity at a given
water content would be given by
Kr/Ar2=Km/X2 [3]
where K m is an "average" conductivity and A r
and A are the same as in [2]. The A r /A thus is
the only variable describing the soil
heterogeniety. For expediency, Peck et al. took
the distribution of Ar/A to be normal and
pursued a water balance analysis on that basis.
Warrick et al. determined the "best fitting" set
of values of Ar/A for three sets of field data
collected in three different states in the U.S.
These were data of Nielsen et al. (1973), Coelho
(1974) and Keisling (1974). Figure 4 shows the
results for the 180 depth of Nielsen et al. for the
water pressure (hr,i) vs. soil water content
expressed as degree of saturation S r>i. Figure
4A is for the raw data and 4B is for the "best-
fitted" data using [2]. The sum of squares is
reduced between the data points and the scaled
values by approximately 90%. A coalescence of
widely scattered points into a relatively narrow
band was found for all cases studied including
that for unsaturated hydraulic conductivity vs.
degree of saturation. However, the set of Ar/A
found from h vs. S data was not the same as that
found for reducing the K values, although there
was a strong correlation. The distribution for
Ar/A values in every case was found to be
skewed and approximately log normal. Es-
timates of the mean and standard deviation for
the ln(A r /A ) of Figure 4 were found to be -0.14
and 0.51.
RAMIFICATIONS OF VARIABILITY ON
WATER FLUX
We now examine distribution of water flux
in the soil profile resulting as a consequence of
spatial variability. For our illustrations, we
consider two different examples. In each case,
we assume the unsaturated hydraulic conduc-
tivity is given by [1], i. e. K — K o exp [a(6~60)].
The Q and K0 take on different values depend-
0.8
0.6
0.4
w" 0.2
z
0
t-
2 0.0
3
CO
o l.O
•-LJ
111
S 0.8
o
0.6
0.4
0.2
0.0
180cm SOIL DEPTH
140 DATA POINTS
-SS = 2.2xlO
M-688[(I-S)-0.749(I-S2)-2.I4(I-S3)
-100 -20O -300 -400 -500
HEADhr | (cm)
SS = 3.2x10
h=-9740[(l-S)-2.25(l-S2)+2.l6(l-SS)
-0.742(I-S*)]S-'
-100 -200 -300 -400 -500
SCALED HEAD ^rhr -(cm)
Figure 4. Unsealed (A) and scaled (B) soil water
characteristic using the similar media concept (After
Warrick et al., 1977b)
227
-------
IRRIGATION RETURN FLOW MODELS
ing upon the site location within the area
considered. The distribution of a and Ko are
each assumed to be log normal with values of
3.91, 0.636, 2.48, 1.40 for mean of In a, standard
deviation of In a, mean In Ko and standard
deviation of In Ko, respectively. These are all
estimates based on the 180 cm depth for the
Panoche site (Nielsen et al. 1973; Warrick et al.,
1977a).
Example 1. Water Initially Ponded
This is the example discussed in detail in
Warrick et al. (1977a). We assume surface pond-
ing sufficiently long that a steady flow rate J L,
= KQ exists throughout the profile at each
location. From site to site, J L will take on
different values and in fact will be log normally
distributed. Figure 5 is a histogram showing the
arbitrary classes for t = O and t = 10 days. For t
= O (Part A), the distribution of J L is Just that
of the K o itself. Even though the mean value is
26 cm/day, 45% of the time the flux is less than
10 cm/day and the remaining 35% greater.
Thus, the high velocity regions are contributing
a disproportionate share of the total areal flux.
At time zero, the ponded water is gone and
the profile drains with negligible surface
evaporation. A simplified drainage equation
based on a unit hydraulic gradient has been
shown by Nielsen et al. (1973) to bea reasonable
approximation for these conditions:
JL=K0/(l+aK0t/L) [4]
where a and K o are as in [1], t is time and L the
depth at which J L is evaulated.
The flux distribution is simulated using
Monte Carlo methods as described, for example,
by Hammersley and Handscomb, (1964). A total
of n random values of Ko and then of a are
chosen from the appropriate log normal dis-
tributions. The distribution of these "n" values
of J L may then be examined over a given time
and the mean and variance estimated.
Figure 5 B shows the resulting flux distribu-
tion of J L for such a Monte Carlo sample of n =
200 and for t =10 days. The mean value is 0.39
cm/day. The first and second classes make up
about 20 and 40% respectively, of the total
population. Thus, the distribution is skewed for
this time just as for t = 0. Roughly 60 percent of
the values are below the mean and 40 percent
greater.
The mean value of J L as determined from
Monte Carlo simulations with n = 2000 for 1,5,
10 and 20 days are shown as the "solid" curve of
Figure 6. The vertical bars show the range of the
middle 50% of the values. Therefore, 25% of the
flux values at each time are greater than the
value for the top of each bar and 25% are below
the base of the bar. The dashed line is the
"deterministic" values obtained by substituting
the mean value of a and K o into [4]. The mean
value and the deterministic curve start together
at t = 0 but diverge with time. The true mean is
consistently above the other curve. These
differences are not a result of an inaccurate
determination of the mean, but correlates sam-
ple size to expected accuracy for J L determined
at t =0, 1 and 10 days. Assumptions include a
sufficient number of samples such that the
50
tu
N
5 40
v>
(A
* 30
O
20
O 10
c
A
t-.O day
n
1-2 2-4 4-10 >10
JLX10 (cm/day)
9V
Ul
N
» 40
V)
V)
2 30
0
Z
~ 20
Z
IU
010
Ul
0.
f±
B
-
-
—
—
t:10 day
nn
<.2 .2-.4 .4-.B .6-.8 >-8
JL (cm/day)
Figure 5.
h» ~
Frequency distribution patterns for 180 cm depth of Panoche soil at t = 0 and t = 10 days.
228
-------
PREDICTING SOIL WATER FLUX
"central limit" theorem is applicable, that is, n
is sufficiently large that estimates of the mean
of J L are normally distributed for variance Var
(J L)/n even though J L itself is not normally
distributed. Examining the results for t = 0, we
find a sample of 1000 would give an estimate
within 5 cm/day (about 16% of the true mean),
95% of the time. Similarly, samples of 100 and 50
would be within 16 and 22 cm/day. At t = 1 and
t =10 days, a sample of 1000, 100, and 50 would
result in estimates within about 4,14 and 20% of
the true mean.
Example 2. Constant Areal Flux Initially
We now consider flux over the entire area at
t = 0, such as might exist under sprinkler
irrigation. At time zero, the irrigation is stopped
and drainage allowed. The flux as a function of
time is
JL = I0/(l + «Iot/L) [5]
where I o was the initial sprinkling rate. Eq. [5]
is simpler than [4] in that a is the only variable
parameter.
The mean value of J L was determined by a
Monte Carlo simulation, normalized by I o and
plotted as a function of time in Figure 8 with 10
= 2.6 cm/day. At t =0, all of the J L values are
EXPERIMENTAL
MEAN VALUE
DETERMINISTIC
a i
10
TIME (DAYS)
Figure 6. Flux values for Example 1 as a function of
time. Bars are ranges of the middle 50% of values
(After Warrick et al., 1977a).
simply the constant IQ- As time progresses, we
see effects of variability. The mean values are
again consistently above the deterministic
value found by substituting the mean a value
into [5]. Also shown as the shaded region is a
band of values corresponding to the middle 50%
of the flux values can be found analytically in
this case since the high 25% of JL values
correspond to the low 25% of a values and vice
versa.
DISCUSSION AND CONCLUSIONS
Spatial variation of soil water parameters
exist, and as a result, variation of fluxes occur
within the profile irregardless of how water is
added. Results presented here are for variation
80
60
40
20
t«0
100 206 1000
NUMBER OF SAMPLES
4.0
3.0
2.0
1.0
0 100 206 1000
NUMBER OF SAMPLES
0.8
0.6
E 0.4
o
->"* 0.2
t* 10
0 100 200 1000
NUMBER OF SAMPLES
Figure 7. Mean values of J L, expected 95% of the
time for t = 0,1 and 10 (After Warrick et al., 1977a).
229
-------
IRRIGATION RETURN FLOW MODELS
in lateral directions, but, of course, vertical
stratification also occurs. Our initial moisture
condition and subsequent drainage was simple
compared to a realistic seasonal field moisture
regime. In addition, the approximate drainage
equations [4] and [5] facilitated numerical ease
for the Monte Carlo simulations. Nevertheless,
the results should be indicative of variations
under more complex conditions.
In some applications, the estimate of the
mean values might well be the singular impor-
tant factor. In this case, Figure 6 serves as a
guideline, although the large number of
samples required seems overwhelming. Effec-
tive sample stratification criteria should help.
In other problems, the flux distribution pattern
itself might be as important as the mean value
itself, for example, if ponded irrigation tech-
niques were used, the high intake area could
dominate the return flow. Low intake areas
might be more subject to salinization.
As far as numerical techniques are con-
cerned, the Monte Carlo technique is powerful
and has a long history of successful
applications elsewhere. Variance reduction
techniques are available for Monte Carlo
studies and might be used advantageously
when solving the non-linear water flow equa-
tion when computer times would be significant
for repeated calculations.
ACKNOWLEDGMENT
Support for this project was provided by
Western Regional Project W-68 and Environ-
mental Protection Agency, Grant Ident. No.
R804751010.
REFERENCES
1. Baker, F. G. and J. Bouma. 1976.
Variability of hydraulic conductivity in two
subsurface horizons of two silt loams. Soil
Science Soc. Amer. J. 40:219-222.
2. Biggar, J. W. and D. R. Nielsen. 1976.
The spatial variability of the leaching
characteristics of a field soil. Water Resources
Res. 12:78-84.
3. Carvello, H. O., D. K. Cassel, J. Ham-
mond and A. Bauer. 1976. Spatial variability of
in situ unsaturated hydraulic conductivity of
Maddock sandy loam. Soil Science 121:1-8.
4. Cassel, D. K. and A. Bauer. 1975.
Spatial variability in soils below depth of
5
.05
I0=2.6 cm/day
MEAN VALUE
-- DETERMINISTIC
1C
TIME (days!
1C
Figure 8. Flux values as a function of time for
Example 2. Shaded area is for the middle 50% of
values.
tillage: Bulk density and fifteen atmosphere
percentage. Soil Science Soc. Amer. Proc.,
39:247-250.
5. Coelho, M. A. 1974. Spatial variability
of water related soil physical parameters. Un-
published Ph. D. dissertation, the University of
Arizona. (Xerox Univ. Microfilms, Ann Arbor,
Michigan 48106. Order No. 75-11061,110 pages).
6. Fliihler, H., M. S. Ardakani and L. H.
Stolzy. 1976. Error propagation in determining
hydraulic conductivities from successive water
content and pressure head profiles. Soil Science
Soc. Amer. J. 40:830-836.
7. Hammersley, J. M. and D. C.
Handscomb. 1964. Monte Carlo Methods.
Metheun Company, London.
8. Kiesling. T. C. 1974. Predictions with
which selected physical properties of similar
soils can be measured. Unpublished Ph. D.
dissertation, Oklahoma State University
(Xerox Univ. Microfilms. Ann Arbor, Michigan
48106. Order No. 75-8812, 108 pages).
9. Miller, E. E. and R. D. Miller. 1955.
Theory of capillary flow: I. Practical im-
plications. Soil Science Soc. Amer. Proc. 19:267-
271.
230
-------
PREDICTING SOIL WATER FLUX
10. Miller, E. E. and R. D. Miller, 1956.
Physical theory for capillary flow phenomena.
J. Appl. Phys. 27:324-332.
11. Nielsen, D. R., J. W. Biggar, and K.
Erh. 1973. Spatial variability of field-measured
soil-water properties. Hilgardia 42:215-260.
12. Peck, A. J., R. J. Luxmoore and J. L.
Stolzy. 1977. Effects of spatial variability on
soil-water properties in water budget modeling.
Water Resources Res. 13 (In press).
13. Warrick, A. W., G. J. Mullen and D. R.
Nielsen. 1977a. Predictions oft
Erh. 1973. Spatial variability of field-measured
soil-water properties. Hilgardia 42:215-260.
14. Warrick, A. W., G. J. Mullen and D. R.
Nielsen. I977b. Scaling field-measured soil
hydraulic properties using a similar media
concept. Water Resources Research 13 (In
press).
231
-------
Water Distribution Patterns
for Sprinkler and Surface
Irrigation Systems
DAVID KARMELI
Department of Agricultural and Chemical Engineering,
Colorado State University, Fort Collins, Colorado.
(On leave from Technion, Haifa, Israel.)
ABSTRACT
A linear fit model for sprinkler and power
curve fit model for suface irrigation systems, are
suggested to represent actual patterns of dis-
tributions in irrigated fields. Both models allow
for efficient calculation of surplus and deficient
volumes, as well as the various relevant efficien-
cies.
Both environmental and economical
aspects require desired efficiency combinations
to be followed and this may be done when the
irrigation systems are represented by the
suggested models. Also, the various functions
related to irrigation performance, may be in-
tegrated to reach optimal results.
INTRODUCTION
Most areas in the world are facing food
shortages as well as problems with water quali-
ty and supply. Overall irrigation efficiencies are
known to be low, and better design and evalua-
tion techniques are warranted. Improvement of
irrigation efficiency will not only increase the
total available water supply, but also have an
environmental impact, as well as an increase in
yields per unit of land or water.
Irrigation efficiency can be increased by
better water management in accordance with
proper adjustment of an irrigation system to the
given resources and its various parameters. It
may be argued that most of the major specific
parameters involved in the behavior of different
irrigation systems have been studied, thus
allowing for the design of better fit irrigation
systems. However, due to the newly added
environmental features involved in water use,
added taxation is due in respect to more efficient
uses of water. The use of available information
enables an insight into the system's possible
performance. However, existing design vari-
ables, as well as a lack of sensitivity for specific
parameters' effects, do not allow its use toward
achieving these goals. Integration of estab-
lished and available information is required in
order to reach an overall strategy allowing sys-
tem performance evaluation, and adjustment of
its parameters to reach higher efficiencies.
cies.
From the standpoints of environmental
concerns (return flow quality), economics of
crop irrigation (lower yields due to excessive soil
moisture stresses; fertilizer losses due to deep
percolation) and water availability (water
losses in the various forms), two main irrigation
features may be identified: the application ef-
ficiency (Ea) and the spatial water distribution
efficiency (Ej))-
The field application efficiency is the ratio
of the desired application amount to the amount
necessary to get this desired application at the
minimum point, or the ratio of the amount of
water required and the total amount of water
actually applied by the system. Field applica-
tion efficiencies are usually lower when using
surface irrigation systems (borders and fur-
rows) and higher with pressure irrigation
systems (sprinkler and trickle). The main fac-
tors involved in the lower application efficien-
cies are the runoff and percolating water below
the root zone for the surface irrigation system,
and the water lost to evaporation, in addition to
deep percolation, for the sprinkler and trickle
irrigation systems.
The extent of runoff water is primarily a
factor of the basic design parameters. It may be
233
-------
IRRIGATION RETURN FLOW MODELS
reused, thus increasing the cost of water, or it
may deep-percolate, while degrading the quality
of underground water flows.
Under surface (border and furrow) and
sprinkler irrigation conditions, the field
application efficiencies may be expressed as:
D x A
til
where D — depth of water made available to plants
during irrigation (mm)
A — total area under irrigation (m-»
y _ total amount of water applied.
In tricle irrigation conditions, the field ap-
plication efficiencies can be expressed as:
K .. N x (1 mm * T <2>
where N
q
mm
y
— total number of tricklers
— minimum emitter flow rate
— irrigation time
(q . i and (T) are designed to meet the es-
mln tablished water requirements.
The field spatial water distribution efficien-
cy is an evaluation of the uniformity of water
distribution in the field, and the factors in-
volved may be classified into those related to the
operational practices and those originating in
the design. Factors related to the irrigation
management and operation tend to average
through the irrigation season and a series of
irrigation applications. The effects may be
difficult to detect in the field and are well
recognized only in the later stages, during yield
performance. Factors related to design may be
considered permanent in their impression on
the field, and the extent of damage is obvious
and easily estimated.
The field spatial water distribution efficien-
cy (E D» can generally be expressed as:
E
D
where D — average depth of water applied by the
system
_^D _ average deviation from average
depth of water applied by emitters
(sprinklers, tricklers) or water stored
by furrows or borders.
The field spatial water distribution or
pattern reflects on two major parameters: defi-
cient water application, and deep percolating
water. Deficient water application is normally
associated with excessive moisture stress and
salt accumulation. Both may result in yield
decrease, but the relationship between soil
moisture stress and maximum yield is such that
net profit may be maximized at a moisture
stress level yielding below the maximum
(Stewart and Hagan, 1973). Also, salt accumula-
tion may be periodically leached, thus avoiding
yield damages. Deep percolating water con-
tributes to the cost of irrigation water per unit of
land as that water is pumped, diverted, and
conveyed, to be lost to deep percolation, while
leaching and causing the loss of valuable fer-
tilizer. Deep percolation may also contribute to
drainage problems associated with rising water
tables and water-logging of the root zone
stratum. Of basic concern is the degradation of
irrigation return flows due to the salt leaching of
the deep percolating water.
Both application and field distribution ef-
ficiencies determine the irrigation quality and
are greatly influenced by the pattern of water
distribution in the field. It is feasible to identify
irrigation systems by the patterns of water
distribution and, accordingly, the relationships
of the system's performance regarding moisture
deficiency, deep percolation and runoff. The
purpose of this work was to develop models that
may supply patterns of distribution and ef-
ficiency parameters for sprinkler and surface
irrigation systems.
SPRINKLER IRRIGATION PATTERNS
OF WATER DISTRIBUTION
A single nonoverlapping sprinkler has a
nonuniform pattern of water distribution, and
overlapping of individual patterns is required.
Overlapping and uniformity of patterns are
achieved by adapted sprinkler and lateral
spacings, as well as adjustments to wind
velocities and directions, nozzle characteristics,
and pressure-discharge relationships.
Sprinkler irrigation distribution patterns
have been characterized by various statistical
uniformity coefficients based on the character-
istics of the distribution. Christiansen (1942)
was first to introduce a uniformity coefficient
(UCC) to a sprinkler system.
UCC = 1 -
N x"y~
234
-------
WATER DISTRIBUTION PATTERNS
or UCC = 1 -
where ^y or - I y 1 -y I
N
the mean deviation
about the mean, y.
Christiansen's coefficient (UCC) is the most
popular one used by decision makers who also
regard UCC > 0.70 as satisfactory.
Wilcox and Swailes (1947) suggested an-
other coefficient of uniformity, where the stan-
dard deviation (s) replaces (Ay):
UCW = 1-4
~y
(5)
Hart (1961) and Hart and Reynolds (1965)
developed a uniformity coefficient (UCH) on the
basis of the assumption that precipitation from
commonly used sprinklers, under regular spac-
ing conditions, is normally distributed, and that
a normal (Gaussian) distribution expression
would well approximate the sprinkler precipita-
tion pattern.
The function for a normal distribution can
be written as:
Y = j* -q exp
(6)
where y — the frequency of the occurence of
value Y
"y" — mean of Y values
s — standard deviation
N — total number of samples
q — length of class interval.
This function (Eq. 6), for a normal distribu-
tion and its shape, can be defined once s and "y
are known, and various parameters related to
applied water ratios and fractions of wetted
areas may be determined.
It is pointed out, as follows, that, when the
samples have a normal distribution, the mean
of the absolute values of deviation equals 0.798 s
, and uniformity coefficient (UCH) is suggested:
UCH =1-.
0.798 s
(7)
Hart (1961) and Seniwongse et al. (1972)
found that distributions of many practiced
sprinkler systems are normal, and they es-
tablished the relationship between UCC and
UCH as:
UCC = 0.030 + 0.958 UCH (r2 = 0.888) (8)
It is also indicated that the high correlation
allows for reliable prediction equations. Assum-
ing normal distributions, the s / y ratio is
reported for each analysis, determining the
shape of the normal distribution curve. This
allows the determination of various irrigation
parameters, such as fractions of area adequate-
ly irrigated, and a generalized table based upon
the normal distribution concept may be con-
structed.
The model, the purpose of this work is to
suggest a model with a standard function whose
properties will enable the characterization of
precipitation patterns of sprinkler irrigation
systems, mainly in reference to efficiency and
other irrigation quality parameters.
The model requirements are the establish-
ment of dimensionless cumulative frequency
curves (Figs. 1 and 2), representing relation-
ships between precipitation distributions, and
the description of these frequency curves by a
mathematical model.
The linear regression functional form
(Eq. 9) was chosen to express the relationships
between the involved variables (Fig. 1):
Y = a + bX
where Y
X
a,b
(9)
— dimensions precipitation depth
— fraction of area
— linear regression coefficients
(constants).
The least squares method (minimization of
sum of squares of deviations of estimated values
from the observed ones) is used to fit the straight
line (linear regression) to the frequency curve of
the dimensionless, Y versus X values.
Various other exponential functional forms
were also studied, but it was established that the
linear regression model fit best to most sprinkler
irrigation patterns covering a wide range of s/y~
values.
The dimensionless sprinkler frequency
curve usually takes the "S" shape, as the
distribution pattern usually tends towards a
normal distribution. The s/y has a relatively
small value when the pattern is highly uniform
and most of the distribution is about the mean.
235
-------
IRRIGATION RETURN FLOW MODELS
0.5
Fraction of Arto (X)
Figure 1. Linear regression fit of a normalized non-
dimensional distribution curve in sprinkler irriga-
tion.
— Linear Regression Fit
— Normal Distribution Fit
—— Actual Data
I 0.5
M
-
0.6 08 l.O 1.2 1.4
Cfeptti ol Irr.gotea Ateot
Q 2 0.4 0.6 0.8
Fraction of Area ( X )
Figure 2a. Fits of actual data into normal and
linear regression distribution in sprinkler irrigation.
However, when the pattern tends to be less
uniform, s V would increase as the deviation
from the mean is larger, and the "S" shape of the
distribution curve would stretch out to behave
more like a straight line.
It may be hypothesized that the normal fit
would be most suitable for distributions where
:ends to be small. However, the linear fit
may be just as good, as most of the distribution
curve would tend to concentrate around the
mean, and errors at both extremes of the fre-
quency curve would be relatively limited. For
distributions where the s y is larger (less fitted
1.5
a
-
2 i.O
|
-
Linear Regression Fit
— Normol Distribution Fit
Actual Data
Y«0.5I4 -f 0.972X
r« -0.929
UCC» 74.8
UCH -73.6
4-- 0.330
y
0.4
H.ltoi
-
O.B
-------
WATER DISTRIBUTION PATTERNS
2.0r
.8
; *
Linear Regression Fit
— Normal Distribution Fit
— Actual Data
= 0.3085+ I
r2=0.973
UCC =63.6
UCH = 64. I
•=•• 0.45
y
383X
0-2 0.4 0.6 0.6 1.0 1.2 1-4 16 18
Histogram for Dimensionless Precipitation
Depth of Irrigated Areas
o :
0.4 0.6 0.8
Froction of Areo IX)
1.0
2
Figure 2c. Fits of actual data into normal and linear
regression distributions in sprinkler irrigation.
depth (Ymin) and also Ymin = l-0.5b as
[a + (a+b)] / 2 = 1.0.
4. The estimated maximal precipitation
depth (Ymax ) equals a + b and also
5. When b = 2.0 , Ymmwill equal 0. Ac-
cordingly, when b> 2.0, Ymjn will equal
parts of the field will not receive any
precipitation (Y = a+ bX will show
negative values). Also, both average
deficit and surplus precipitations can be
derived by substituting 0.25 and 0.75 for
X.
A distribution coefficient (UCL) is suggest-
ed to describe the uniformity of distribution
when the precipitation cumulative frequency
curve is fitted into a linear regression.
Substituting "I y ' ' y Un Equation 4, from
N
Figure 1:
UCL = 1 - — ['/2 x 0.5 (y -y) ]
— HldX
y
(10)
Linear Regression Fi
Normal Distribution Fit
Actual Data
0 0.2 0.4 06 O.S IO i.2 14 16 1.8 2 0
Figure 2d. Fits of actual data into normal and
linear regression distributions in sprinkler irrigation.
and UCL =1 1 -
(y max -jF).
Since y = 1.0 and y max - y = 0.5b
UCL = 1 - ('A x 0.5b)
or UCL = 1 - 0.25b
and b - 4.0 - 4.0 UCL
(11)
(12)
Relationships between Y max , Y min , b and
UCL are given in Figure 3.
Linear regression frequency curves are
given in Figure 4 for different uniformity coef-
ficients (UCL), as derived from the linear
regressions: 50 percent, 60 percent, 70 percent,
80 percent, 90 percent and 95 percent, and the
fraction of irrigated area, X, to given dimen-
sionless precipitation depth, Y. Figure 4 can be
used to determine application efficiencies once
magnitudes of under-irrigated areas have been
established as design criteria.
To study the suggested model and coef-
ficient of uniformity, 19 sets of single sprinkler
patterns were chosen, The sprinklers operated
under varying pressures (40, 50, 60 psi) and
varying wind velocities, and using a computer
237
-------
IRRIGATION RETURN FLOW MODELS
2.0r
- '
Ltneor Regression Fit
— Normol Distribution Fi1
t* -O 289 + 2.579X
r '=0.974
UCC • 32.7
UCH- 39 7
n i i i fTI
04 0.6 0.8
Froction of Areo (X)
•
l.O 1.5 2.0 2.5
Irr igot ion Qua 111 y (t>)
Figure 2e. Fits of actual data into normal and linear
regression distributions in sprinkler irrigation.
program, it was overlapped into various
spacings (20', 30', . . .70') x (20', 30', . . .70') as
well as various geometrical arrangements (rec-
tangular and triangular). The total number of
combinations studied was 798.
The actual data for each combination is
transformed into a dimensionless frequency
curve and fitted into both linear regression and
TABLE
Figure 3. Uniformity coefficient (UCL) as related to
irrigation quality in sprinkler irrigation.
normal fits. UCC and s/y"values are also given.
The linear regression fit was found to fit the
frequency curve very well. The determination
coefficient, r 2, was < 0.800 for only 5.12% of the
798 patterns; 0.800-0.899 for 14.15%; 0.900-0.949
for 28.90% and > 0.950 for 51.83% of the
patterns.
Table 1 and Figure 2a-e present a sample of
five sprinkler patterns with UCL values rang-
ing from 0.355 to 0.882. Results of the two
different fits were compared to the actual data,
yielding the overall differences between the
actual and the normal (IN ^) and linear (XL ) fit
values. The normal fits do not exhibit smaller
Differences between actual, linear (AL) and normal (AN) distributions for the various fraction of area, X.
Fraction of Are*. X
Sprinkler
Fatten* •
• 0.16 J7.5
0.3;
M.
0.45 63.6
d 0.59
II
e 0.76 3;."
M. - Difference betw
iS • Difference betw
OCX T~ .1
§7.2 88.2 0.470 0.932
0.03:
-0.001
:. 9:9
-0.047
-0.055
64.1 65.4 1.383
0.095
52.8 50.6 1.976
0.145
39.7 35.5 2.579 0.974
-0.065
-0.034
tea actual and normal distribution
.:
0.028
0.030
-0.054
-0.013
0.037
0.023
D.ue
0.095
0.234
fit values
-0.045
-0.089
-0.017
0.031
-0.045
0.000
-0.069
0.014
0.164
-
for X •
_
-0.034
-0.031
0.019
0.059
-O.073
-O.045
-0.160
-
0.124
1-0.5
0.1-0.9
.5
0.013
0.009
0.055
-0.057
-0.053
-0.011
-0.012
-0.175
-0.174
-0.045
-0. 056
0.072
0.086
-0.042
-0.062
-0.054
-0.103
-0.114
-0.179
.7
0.002
-0.012
0.060
0.065
-0.037
-O.072
0.143
0.058
-0.051
-0.168
.8
0.049
0.039
0.041
0.054
0.040
0.007
0.052
-0.044
0.011
-0.125
9 III-'
003 0.009
010
156 0.046
094
0.080 0.035
0.10B
-0.
-
039 0.058
072
.113 0.099
.052
EA1T
0.009
0.037
0.038
0.100
0.260
238
-------
WATER DISTRIBUTION PATTERNS
0 0-1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Fraction of Area IX)
Figure 4. Uniformity coefficients assuming linear
relationships between precipitation depth and area
irrigated in sprinkler irrigation.
Regression. Y*o»b; Y=b,X»Wind
Spacing Pressure ( psi)
-:
u
.1 3.0
' 2.0
c
I. 130'
I A. (50'
2. (50'
2A.150'
3. (70'
3A.170'
4 6 8 10
(Wl- Wind Velocity, m.p.h.
Figure 5. Irrigation quality (b) as a function of wind
velocity for varying spacings and pressures in
sprinkler irrigation.
deviations from actual data than the linear fit.
Where s/y is significantly larger (patterns d and
e) the linear fit is much better. However, there is
a tendency for smaller deviations at the ex-
tremes, when the normal fit is exercised, while
the rest and major portions of the field tend to
lesser errors, with the linear regression fits.
Various relationships
from this study:
were established
b = 3.956 - 3.940 UCC (r2 - 0.998) (13)
where b — coefficient of the linear regression for the
dimensionless frequency curve.
Equation 13 shows great similarity to the theo-
retical Equation 12.
b = 4.170 - 4.180 UCH
ra = 0.999)
(14)
The "b" values derived (representing irriga-
tion quality or patterns) are related in Figure 5
to wind velocities for various spacings and
pressures to allow their use in decision-making.
The linear regression model allows for the
evaluation of the distribution pattern (UCL), as
well as reaching valuable irrigation decisions
on the basis of results from the given function.
TABLE 2
Estimates of Water and Area Distribution for
Five Sprinkler Patterns (Fig. 2a-e).
Sprinkler Pattern
Efficiency and
System Quality
UCL
b (Y = a + bX)
Y^ (estimate)
Ynfl, (actual)
* max (estimate)
Ymax l"ctual>
Deficiently
watered area
(Y<0.0)
Average
watered area
(Y=0.9-l.ll
Surplus
watered area
(Y>1.1)
Increase to
bring 100%
area to YJJ
Increase to
bring 90%
area to YD
Percent
volume in
deficient zone
Percent
volume in
average zone
Percent
volume in
surplus zone
Percent
volume in
excess
88.2
0.470
0.775
0.717
1.198
1.252
0.316
0.426
0.258
•
.: ••;
0.261
0.440
0.299
0.018
b
75.7
0.972
0.536
0.551
1-411
1.715
0.397
0.206
0.397
,:•-.
47
0.281
0.206
0.513
0.099
c
65.4
1.383
0.378
0.233
1.622
1.677
0.427
0.146
0.427
2.38
2.01
0.258
0.146
0.596
0.152
d
50.6
1.976
0.111
0.048
1.889
1.975
0.449
0.102
0.449
8.1
4.29
0.205
0.102
0.693
0.308
e
35.5
2.579
0.000
0.026
2.161
2.153
0.461
0.078
0.461
-
0.141
0.078
0.781
0.274
239
-------
IRRIGATION RETURN FLOW MODELS
Deficient and surplus volumes, in relation to
area fractions, as well as changes required to
adjust to requirements, may be reached using
some of the following relationships:
YD'a
Area deficiently irrigated, A = -
D b
where Y p — maximal precipitation in the area
irrigated deficiently.
Area irrigated in surplus, A = 1 -
where Y <, — minimal pnx-ipitation in the area irri-
gated in surplus.
Area irrigated, as designed (± deviation).
Average application in the surplus area,
Y
S ~ 1 -c 2
S 2
Average application in the deficient area,
Y ±
Y = P • 1 + 2
^
Volume estimates may be obtained if total area
(A rp ) and average application rate (D ^ ) are
given. The excess volume applied (V Q ) would
be: b
V , =--
max - Y S
Some of these estimates are given in Table 2,
for the patterns shown in Figure 2a-e.
SURFACE IRRIGATION PATTERNS OF
WATER DISTRIBUTION
The two most common parameters for
describing the quality of surface irrigations
have been the water application efficiency (E a),
as shown in Equation 1, and the water distribu-
tion efficiency (Ejj), as shown in Equation 3.
Although the above efficiency parameters
provide an insight into irrigation system perfor-
mance, they are lacking in several respects,
namely:
1. The amount of deep percolation or defici-
ent volumes of infiltrated water in an irri-
gated field cannot be deduced.
2. The actual distribution in infiltrated
depth of water throughout the field can-
not be established.
Thus, a need exists for a technique which
will describe distribution patterns and efficien-
cies. The technique must be able to:
1. Aid in developing optimal design of irri-
gation systems based on computation of:
a. Amount and distribution of deep per-
colation water volumes
b. Amount and distribution of under-
irrigated root zone volumes
c. Volume of runoff losses.
2. Aid in evaluation of already-designed
systems and evaluation of operational
variable changes on the performance.
A perfect irrigation would be one in which
each of the efficiency parameters are a unity.
These parameters are: the water requirement
efficiency (EDV), the deep percolation efficiency
(EDP), the tail water efficiency (ETW), and the
overall application efficiency (EB). The efficien-
cies are given by the following equations.
EB = E „ 100
a.
EDV = 1 - VDF/VRZ
where VDF — Total deficient volume of water
in root zone after irrigation
VRZ — Total amount of water needed by
root zone to reach field capacity.
EDP = 1 - VDP/VLI
where VDP — Total volume of deep percolation
VLI — Total volume of infiltration.
ETW = 1 - QRF TVL
where QRF — Total runoff volume
TVL — Total volume applied to field.
The optimal combination of efficiencies to
provide maximum benefit in any irrigation
system is dependent on many factors. EDV is
dependent on the loss in productivity due to crop
stress and salt accumulation. EDP is a function
of water cost, possible increased drainage costs,
loss of fertilizers, and reduction in yield due to
lost fertilizers. ETW is dependent on cost of
water, cost of water application, cost of reuse
and availability of water.
The primary factors which affect the dis-
tribution of infiltrated depth of water
240
-------
WATER DISTRIBUTION PATTERNS
throughout irrigated fields in surface irrigation
are many:
1. Soil infiltration characteristics
2. Slope
3. Roughness
4. Geometric configuration of channel or
width of border
5. Desired depth of application
6. Length of run
7. Volume inflow
8. Total time of application (total volume
applied).
An example of the potential effects on the
distribution of infiltrated water with varying
operational characteristics in an irrigation
system is illustrated in Figure 6, where only one
parameter, volume inflow, varies.
The distribution of water in an irrigated
field may be computed without resorting to
actual field measurements if the following are
known:
1. Rate of advance of the irrigation front
2. Infiltration characteristics of soil
3. Rate of recession of water from the soil
surface.
Several authors have proposed methods for
estimating the rate of advance in furrows and
borders. Bassett (1972) has proposed a model
which accounts for both mass and energy of the
advancing water, and volume balance
procedures have been proposed by several
I- ( l,,5l,Dl.Ll)«t[ Y-O.810 + O.637X •-•••, r»-0.9e9J
Z. (I,.S,.D,.L,>«.[Y-0.«IO*0.407X*~. <* -0.994]
1 U.S.. 0,.L,),,[»-<>«™»° *««"• .f*-OJ«]
4. (1,.S,.01.L,)<.[T.0.9I5 +O.I73X"". r«.0.99e]
5. ( I, ,S,,D,. L() «,[ V-O.94a + O.IOS X ••••• , r«- 0-999 ]
0.3 0.4 0.5 C
FraclM* «1 *f«a (X)
Figure 6. Power curve fit expressions for varying
discharges in furrow irrigation (Y = C + aX^).
authors. Among these are Hall (1956), Wilke and
Smerdon (1965) and Bishop and Fok (1965).
Although these authors make a seemingly gross
assumption that the momentum can be ac-
counted for by assuming a constant average
depth of surface storage, their comparisons with
field data indicate that these methods yield a
good approximation of irrigation advance for
most cases.
In this study, the method proposed by Wilke
and Smerdon (1965) was used to simulate ad-
vance on irrigation furrow because of its ease of
application and close approximation to field
data. The equation established by Wilke and
Smerdon is a solution of the Lewis and Milne
volume balance equation as proposed by Philip
and Farrell (1964). The general equation propos-
ed by Philip and Farrell is:
kt
m=0
(15)
•>:
• (2+mn)
where q — inflow - ft-Vmin/furrow in furrow, or
inflow ftVmin/ft in border
c — average area of surface storage, ft2, for
furrow, or average depth of surface
storage, in feet, for borders
x — distance of water advance at time t
t — time of advance
k,a — constants of cumulative infiltration
equation
n — an integer
F — symbol for the gamma function.
Wilke and Smerdon (1965) solved the equa-
tion for several values of n and the resulting
equation is:
kt"
3 = 1.0 + A
ex c
(16)
where q, t, c, x, k, and n are as defined previously
A — coefficient dependent on the exponent
of the infiltration equation. The solution
for values of n other than specifically
given are obtained through interpola-
tion.
Several empirical equations have been used
to describe the infiltration characteristics of soil
[Green and Ampt (1911), Gardner and Widstoe
(1971), Philip (1957) and Kostiakov (1932)]. The
241
-------
IRRIGATION RETURN FLOW MODELS
Kostiakov and modified Kostiakov equations
have been the most commonly used equations
due to their simplicity and applicability to most
field situations. The integrated form of the
Kostiakov equation is used in this study.
D =1 kt
(17)
where k.n — empirically determined constants for
a given soil and soil condition
t — total infiltration time
D — cumulative infiltration depth.
Recession in furrows and border irrigation
was discussed by Wu (1972) and others. In
borders, an estimation of the rate of recession is
essential, as the surface storage is usually a
substantial portion of the total applied water
when the discharge and the head of the border is
terminated. In furrows, the volume stored in the
furrow is usually a small portion of the in-
filtrated water. Wilke and Smerdon (1965)
assumed a horizontal recession in furrows and
stated that the assumption that surface storage
is negligible ". . .would not be satisfactory for
small time increments, for soils having very
slow infiltration rates or for unusually large
furrows or stream sizes." Recession was as-
sumed negligible in this study, as its inclusion
was not considered to be essential to the pre-
sentation of stated objectives.
The model. As with the sprinkler analysis,
the stated purpose of this study was the develop-
ment of a model for describing distribution
patterns and efficiencies. The model re-
quirements are:
1. Establishment of a frequency curve of
the dimensionless infiltration depth, Y,
versus fractional areas (fractional
length of run), X (Fig. 6).
2. Description of the frequency curve by a
mathematical model which enables the
companson of systems based on the var-
ious efficiencies (EDV. EDP, ETW) and
other statistical parameters such as the
deviation from desired application
depth.
The models considered in this work for de-
scribing the frequency distribution are:
1. Y - aXb (18)
where Y — dimensionless infiltration depth
= infiltrated depth desired
depth of infiltration
X — fraction of field, wetted
a — coefficient of distribution curve
* max
b — exponent of distribution.
2. Y = c -r a X b (Fig. 7) (19)
where Y, X — as defined above
— minimum depth ratio (Y
min
1 max * mm
b — exponent of distribution.
Both models (Eqs. 18 and 19) were fitted to
the frequency curves computed when using the
Wilke and Smerdon advance equation (Eq. 16)
and the Kostiakov infiltration equation (Eq. 17).
The curves represented a wide range of com-
binations for furrow irrigation with varying
design and operation variables. (Five com-
parisons are illustrated in Fig. 8.)
The model Y = a XD was fitted to computed
depths with minimum r2 — 0.89, and mostly
r2 > 0.95. This equation is inaccurate at very
small X, and underestimates Y max. (It is very
accurate only when water first reaches the end
of the field.) the model of the form Y = c + aX°
is more desirable, as the model requirements are
better met and it allows for more information.
• o
; ° i.O
| o
o 2
I
•
•
*„,„.
." IX-
0.5
Froction of Area (X)
Y« C + OI*
Y Dimensionless Infiltration Ratio* Actual Infiltrated
Depth / Desired Infiltration Depth
a= Ymolllf,um- Yfflinjmmn
b= Exponent of Distribution
c = 'mi mmum
X Dimensionless Distance = Distance from End of Field/Total
Length of Field
Xp Dimensionless Distance at which Deep Percolation Begins
CDEC Volume of Deficient Moisture Application
EFGE Volume of Deep Percolation Water
ACBA Volume of Run-off or Toilwoter
Figure 7. Power curve fit of a normalized distribu-
tion curve in surface irrigation.
242
-------
WATER DISTRIBUTION PATTERNS
The minimum application ratio, Ymjn, in the
field is the coefficient, c. The maximum applica-
tion ratio, Ymax, is (c + a).
From the Y = c + a X model, the deficient
and deep percolated volumes may be obtained,
and in many cases the runoff volume and total
applied volume may be accurately estimated.
The under-irrigated or deficient volume, DFV,
in Figure 7, is represented by the area CDE. This
area is:
DFV= CDE = DEIB - CEIB = X p - CEIB
where X -
cX+-
CEIB =
(
b+1 |X
P (c + a X b ) dX =
b+1
b-t-1
T
Jo
= cXp+_a_Xp
and DFV = X - c X - _JL_ X
P P b+1 P
b+1
b+1
(20)
The volume of deep percolation, VDP,
Figure 7, is represented by the area EFGE and
can be computed as:
(21)
VDP = EFG = EFHI - EGHI
EFGHI=
c + a X ) dX =
1
cX JL X b+1 j
b+1 JX
b+1
b+1
EGHI = 1 - X
and VDP = c+
b+1
- _?_ - aX
P b+1 P
a x b+1
b+1 P P
The sum of deviations from the desired infiltrat-
ed depth, SDEV, is:
SDEV = DFV + VDP
The approximate runoff volume, QRF, is
represented by the area ACB in Figure 7. This
area may be approximated by extending Line
CEF to the point where Y = 0. More exactly, this
area may be calculated by taking three points
along curve CEF for any combination of a, b, c
and solving two equations, one of which is
solved by trial and error, to obtain a power curve
fit Yn = AI XnBl . The following equations
apply:
- Y
1/B\ B
where Y 0 - Y min = c
Y 5 — Y at X - 0.5
Y 1 ~ Y max - c + a
A i , B i — constants in Equation Y n — AXn.
The distance AB in Figure 7 can then be
computed as:
XI=(YO/AI)I/BI
and the runoff (QRF) may then be estimated as:
QRF
= I"
O7
AlXBldx=AL_XBl + 1
B i +1
1
(22)
B
The total volume applied to the irrigated
field, TVL, is represented by the area under the
curve ACEF:
TVL
.5"
Al X
B
1 dX =
-^ xB'+I
or
TVL =
B,
1 /
- — (X +
+ 1 !
.B +1
(23)
The volumes computed can be used to calcu-
late the various desired quality parameters, and
the following equations apply:
EDV = 1 - DFV/VRZ (24)
where VRZ — volume desired in root zone =1
243
-------
IRRIGATION RETURN FLOW MODELS
EDP =
where
ETW =
EB =
1 — VDP VLJ
VLI — total volume infiltrated
=Area EFGHI
1 — QRF TVL
j _ TVL - QRF - VDP • DFV
TVL
Some results from a series of trials in which
only a single design or operation variable at a
time was changed are given in Table 3 and
Figure 9. The results show that if the distribu-
tion of infiltrated depth can be established.
EDV and EDP can be closely determined from
the equation Y = c -t- a X D . ETW was found not
to correlate as well for some values of c, a, and b.
This is probably due mostly to the effect of
surface storage and recession which was
neglected in this study. However, in the many
-
1.3
.
0.9
0.8
-
•
furrow cases examined, ETW is predicted ac-
curately for cases in which the surface storage is
a small fraction of the infiltrated volume. The
coefficients c, a, and b in the distribution model
are merely an expression of distribution in the
irrigated field, and thus the extrapolation of the
equation to obtain ETW may not be justified in
border irrigation, as recession is an important
factor. As an example of when extrapolation
would yield totally erroneous results is the case
when the rate of recession is approximately
equal to the rate of advance. In this case the
extrapolation of the distribution in the irrigated
field would indicate an infinite volume of runoff.
Since the discharge into a furrow or border
is usually known or can be easily measured,
ETW can be calculated if the distribution equa-
tion is known by the following simple
relationship.
I; - Cumulotive Infiltration Depth (m'/m)
I, -0.0122 t0™* m*/m
Sr Slope S,'0.5%
D - Desired Infiltration Depth (mm)
D_* 100 mm
LJ- Length of Furrow (m) L,'400m
q. ' Furrow in Flow (Uteri/tec)
q,«0 6 I/sec
q>« I 01/sec
q4« 1.5 t /sec
Ht-2 01 /we
Q.
1 , rf-0.984
301 /sec
I. Y-I.367X"
Y-0.810+ 0.637X0"', rf-0.989
2. Y.I.I58X0-"* , r1 =0.965
Y'0.810+ 0.407X0'**, r* = 0.994
3. Y-I.082X0 OTt, r* = 0.940
Y-0 873 + 0.246XOT*°, r'=0.998
4. Y-I.057X00", r* = 0.930
-
I. ,S. .D,. L,,q.
0.2 0.4 0.6 0.8
Fraction of Aria ( X )
Computed Doto
Y-0.915 + O.I73X'
, T*' 99.8
5 Y-I.036X00", r* = 90.9
Y'l.036 + O.I06X"
, r*=99.9
-
0 0.2 0.4 0.6 0.8 1.0
Fraction of Area ( X )
Figure 8. Comparison between fits of computed data into two power curve fits in furrow irrigation.
244
-------
WATER DISTRIBUTION PATTERNS
+
ETW =
where
b+1
.) x VDA
QT (25)
VDA — total volume desired in root zone
QT — total volume applied to field.
The calculation of efficiencies through use
of the equations may be eliminated, as the
efficiencies obtained from any given distribu-
tion may be estimated for any combination of a,
b and c using graphical presentations (Fig. 9a-
e).
Valuable insight into the performance of an
irrigation system can be obtained by studying
the change in efficiencies with changing design
or operation variables (Tabel 3). From this table
we see that EDV decreases rapidly from the first
time setting to the second for discharges q ^ and
qo. However, to completely fill the root zone
requires a substantial increase in the total
volume (time) of application. Thus, if a small
percentage of deficiency can be tolerated, a
substantial decrease in water use is realized.
Figure 9a. Surface irrigation quality parameters
and distribution coefficients (a, b, c) for minimal
application ratio, c = 0.4.
Relationships of varying irrigation design
variables and distribution coefficients (a, b, c)
may be plotted (Fig. lOa, b) to enable the
designer to compare systems and select the
optimal design or operation based on his
limitations on EDV, EDP and ETW. In the
practical design ranges, the curves for any one
system may be established by a few points. The
system performance for many different
operating conditions may then be deduced.
Table 4, in conjunction with Figure 11, illus-
trates one such analysis. In this analysis
weighting factors were assigned to each of
EDV, EDP and ETW. The overall performance
of an irrigation system (OP) over a range of
variable changes can thus be examined. The
equation which was chosen as an indicator of
system performance in each case is given by:
C1ETW
C2EDV
2.0 1.6 1.2 0.8 0.4
8
0.4 0.8 1.2 1.6 2.0
.
Figure 9b. Surface irrigation quality parameters
and distribution coefficients (a, b, c) for minimal
application ratio, c = 0.5.
245
-------
IRRIGATION RETURN FLOW MODELS
6 : Z
1
1
0 2
5
o-j
i
-
Figure 9c. Surface irrigation quality parameters Figure 9d. Surface irrigation quality parameters
and distribution coefficients (a, b, cl for minimal and distribution coefficients (a. b. c) for minimal
application ratio, c = 0.7.
application ratio, c = 0.8.
TABLE 3
Comparison between computed and estimated efficiency values.
Total
39.8
46.8
27.8
3».5
15.1
43.6
46.8
-
70.6
Mstr lbut lor.
Coefficient
0.000
0.810
l.OOO
0.810
l.OOO
•
0.873
l.OOO
0.915
0.308
0.347
0.246
0.173
0.149
0.283
0.685
0.746
0.638
0.694
0.780
0.835
0.875
• ater Requlr.
Efficiency.
rSL*
0.»J2
0.987
l.OOO
0.836
0.976
1.000
0.984
l.OOO
Cc*fn,te
-------
WATER DISTRIBUTION PATTERNS
2.0 1.6 1.2 0.8 0.4
0.4 0.8 1.2 16 2.0
Entire Profile Adequately
Irrigated , Thus:
EDV 1.0
-
b
Figure 9e. Surface irrigation quality parameters
and distribution coefficients (a, b, c) for minimal
application ratio, c = 1.0.
The weighting factors Cj, C%, and €3, as
shown in Figure 11, were asssigned arbitrarily
for purposes of illustration. In practice,these
weighting factors would include the many and
varied costs and/or benefits associated with the
performance of a system at any given set of
conditions.
CONCLUSIONS
The models, based on the linear fit for
sprinklers and power curve fit for surface irriga-
tion systems, were found to be good represen-
tations of actual patterns of distribution in an
irrigated field. The models also allow for ef-
ficient calculation of surplus and deficient
volumes, as well as the various relevant efficien-
cies. The use of the distribution patterns, as
studied by these models, is important when
changes in operational or design variables are
due. The concept of managing or designing an
irrigation system to obtain the desired efficien-
TABLE 4
Overall furrow irrigation performance with
changing design and/or operation variables*.
Total
Volume
Applied
QTOT, m3
35.0
38.0
41.0
44.0
47.0
Total
Volume
Applied
QTOT, m3
35.0
38.0
41.0
44.0
47.0
50.0
Total
Volume
Applied
QTOT, n3
35.0
38.0
41.0
44.0
47.0
50.0
*
t
:
0.52
0.76
0.87
0.97
1.00
t
0.69
0.80
0.86
0.93
0.96
1.00
t
c
0.69
0.73
0.75
0.82
0.84
0.88
CIETW
i
0.90
0.70
0.55
0.53
0.48
t
0.59
0.47
0.38
0.35
0.33
0.32
a '
0.33
3.32
0.28
).2<
3.23
0.23
i + C2
'1*1
"'
0.50
0.66
0.77
0.75
0.75
'1°!
b+
0.60
0.73
0.80
0.82
0.83
0.84
hSl
bt
0.68
0.73
0.73
0.74
0.75
0.80
EDPj + C
D3 L3 "2
ETW
0.99
0.92
0.83
0.79
0.77
D3 L3 "3
ETW
0.940
0.830
0.720
0.690
0.650
0.620
D3 L3 "4
ETW
0.850
0.81
0.78
0.74
0.70
0.66
3EDVi
EDV
0.94
0.975
0.99
0.997
1.00
EDV
0.955
0.970
0.980
0.990
0.995
1.000
EDV
0.88
0.91
0.925
0.935
0.940
0.945
I . .
0.915
0.835
0.825
0.820
0.815
:.
0.90
0.91
0.89
. 1 -
0.865
0.86
EDP
0.99
0.98
0.97
0.96
0.96
0.96
OP/
0.940
0.939
0.921
0.913
0.908
OP/
0.943
0.921
0.893
0.887
0.878
0.671
OP/
0.887
0.891
0.690
0.883
0.874
0.665
OP/
0.947
0.952
0.946
0.943
0.941
OP/
0.947
0.939
0.925
0.925
0.920
0.918
OP/
0.8S5
0.896
0.903
0.902
0.898
0.895
ETW tall water efficiency
EDV water requirement efficiency
EDP deep percolation efficiency
C . C2, C3 assumed weighting factors
c, a, b - coefficients of distribution
cy combinations rather than designing to
always fill the root zone is forthcoming if short
water supplies are of any concern.
Based on this work, the following procedure
may be followed to describe optimal distribution
patterns and efficiencies for surface irrigation
systems.
1. Establishment of soil infiltration rate,
rate of advance and recession with
different inflow rates.
2. Derivation of power curve fit models, us-
ing the above infiltration, advance and
recession expression. These models
247
-------
\.n
0.8
0.6
0.4
0.2
I -1
1.2
1.0
0.8
0.6
0.4
0.2
0.8
0.6
0.4
0.2
\
(l),S1,DJ,LJ),q3
(IpS^Dj.Lj) ,q4
(I,.S,,D3,L3),q5
n,,Sl,D3,L3),q6
I,«0.01221° 4 ms/i
S,-0.5%
D3= 100 mm
L3»400m
qzs 0.6 I/sec
q » 1.0 t/sec
q4- 1.5 t/sec
q5* 2.0 f/sec
qg= 3.0 J/sec
55
•'
85
Total Volume Applied (QTOT), nv
Figure 10a. Variation in distribution coefficients (a, b, c) for varying discharges and total volume (time) applied
(QTOT) in surface irrigation.
-------
1.0
O.8
0.6
0.4
0.2
1.4
1.2
I 0
0.8
X-
(I,,S|,D3),L2,(q3)
(I,,S1,D3) L3, (q3)
(I,,Sl,D3),L4,(q9)
[,•0.01221
S, = 0.5%
D3= 100 mm
v,
=I.O I/sec
L3=400m
L4=600m
L9= 800m
\
95
Total Volume Applied (QTOT), m
Figure lOb. Variation in distribution coefficients (a, b, c) for varying discharges and total volume (time) applied
(QTOT) in surface irrigation.
-------
IRRIGATION RETURN FLOW MODELS
«O 45
Total volunt ApBIKd IOTOT). »•
Figure 11. Overall performance as determined by
weighted irrigation quality parameters in surface
irrigation.
describe the distribution pattern as well
as allow for computing efficiencies.
3. Derivation of optimal combinations of
distribution patterns and efficiencies
with varying design and operation
variables such as length of run, inflow
rate, desired application depth and total
applied volume.
For sprinkler irrigation a single sprinkler
pattern may be overlapped in a wide series of
feasible alternatives. The results are fitted into
a linear regression which allows for assessment
of irrigation quality under varying conditions.
Previous studies by Griddle et al. (1956), Wu
(1970). Hart (1961). Hansen (1960) and others
have suggested various analytical and graphi-
cal means to evaluate specific performance
parameters.
The results of this study offer a good,
efficient and overall means of presenting actual
pattern distribution for both sprinkler and sur-
face irrigation systems.
The models suggested allow for integra-
tion of the various functions related to irrigation
performance such as water-yield function. This
is feasible since deviations from optimal depths
are established using simple and accurate
techniques.
REFERENCES
1. Bassett, D. L. 1972. Mathematical
model of water advance in border irrigation.
Transactions of the ASAE 15(5):992-995.
2. Bishop, A. A., and Fok, Y. S. 1965.
Analysis of water advance in surface irrigation.
Journal of the Irrigation and Drainage Divi-
sion, ASCE, Vol. 91, No. IR1, Proc. Paper 4251,
pp. 99-116. March.
3. Christiansen, J. E. 1942. Irrigation by
sprinkling. California Agricultural Experiment
Station Bulletin No. 570.
4. Criddle, W. I)., Davis, S., Pair, C. H., and
Shockley, I). G. 1956. Methods for evaluating
irrigation systems. Agricultural Handbook No.
82, Soil Conservation Service, USDA.
5. Gardner. W., and Widstoe, J. A. 1971.
Movement of soil moisture. Soil Science, vol. 11,
pp. 215-232.
6. Green, W. H., and Ampt, G. A. 1911.
Studies on soil physics. I. The flow of air and
water through soils. Journal of Agricultural
Science, Vol. 4, pp. 1-24.
7. Hall, W. A. 1956. Estimating irrigation
border flow. Agricultural Engineering, Vol. 37,
pp. 263-265.
8. Hansen, V. E. 1960. New concepts in
irrigation efficiency. Transactions of the ASAE
3(l):55-57, 61-64.
9. Hart, W. E. 1961. Overhead irrigation
pattern parameters. Agricultural Engineering,
pp. 354-355. July.
10. Hart, W. E., and Reynolds, W. N. 1965.
Analytical design of sprinkler systems. Tran-
sactions of the American Society of Agricultural
Engineering 8(l):83-85, 89. January-Jebruary.
11. Kostiakov, A. N. 1932. On the
dynamics of the coefficient of water percolation
in soils and on the neceessity for studying it
from a dynamic point of view for purposes of
amelioration. Transactions of the 6th. Com.
Inter. Society of SOIL Science, Part A. Rus-
sian, pp. 17-21.
12. Philip, J. R. 1957. Theory of infiltra-
tion: 4. Soil Science 84(3):257-264.
13. Philip, J. R., and Farrell, D. A. 1964.
General solution of the infiltration-advance
problem in irrigation hydraulics. Journal of
Geophysical Research, Vol. 69, pp. 621-631.
February 15.
250
-------
WATER DISTRIBUTION PATTERNS
14. Seniwongse, D., Wu, I. P., and
Reynolds, W. N. 1972. Skewness and kurtosis
influence on uniformity coefficients and
sprinkler irrigation design. Transactions of the
ASAE 15(2):266-271. March-April.
15. Stewart, Ian J., and Hagan, Robert M.
1973. Functions to predict effects of crop water
deficits. Journal of the Irrigation and Drainage
Division, ASCE, No. IR4, pp. 421-439.
December.
16. Wilcox, J. C., and Swailes, G. E. 1947.
Uniformity of water distribution by some under-
tree orchard sprinklers. Scientific Agriculture
27(ll):565-583.
17. Wilke, Otto, and Smerdon, E. T. 1965.
A solution of the irrigation advance problem.
Journal of the Irrigation and Drainage Divi-
sion, ASCE, Vol. 91, No. IRS. September.
18. Wu, I-Pai. 1970. Graphic relationships
of intake, length of run, and time. Journal of
Irrigation and Drainage Division, ASCE, Vol.
96, No. IRS, Proc. Paper 7506, pp. 233-240.
September.
19. Wu, I-Pai. 1972. Recession flow in
surface irrigation. Journal of the Irrigation and
Drainage Division, ASCE, Vo. 98, No. IRl, Proc.
Paper 8764, pp. 223-240. March.
251
-------
Hydro-Salinity Models:
Sensitivity to Input Variables
J. D. OSTER and J. D. WOOD
USDA, Agricultural Research Service,
U.S. Salinity Laboratory, Riverside California.
ABSTRACT
Return flows estimated from salinity, or a
water budget analysis, exhibit opposite sen-
sitivities to increasing field irrigation efficien-
cies. However, the sensitivities of both
methods to assess return flows may be com-
plimentary if both are used to obtain an op-
timized estimate of return flows. The degree of
data uncertainly would undoubtedly be greater
in areas with multiple return flow paths, greater
sources of underflows and rainfall, and shorter
growing seasons.
Complex methods are being increasingly
used to assess the environmental impact of
irrigation projects. The development of models
that account for the hydrologic and salinity
balances of a large agricultural area is a major
undertaking. Once developed, the need for
answers to a specific set of conditions often
precludes a thorough sensitivity analysis of the
calculated values based on the expected
variability of input data. For example, results of
an optimization study by the Bureau of
Reclamation showed that shifts in cropping
patterns greatly affect the calculated return
flows from the Wellton-Mohawk Irrigation and
Drainage District in Arizona.
At our laboratory we have been interested
in the use of salinity measurements to deter-
mine leaching fractions and field irrigation
efficiencies in several field experiments. We
have also been asked recently to assist SCS in
evaluating how much improved irrigation
methods may be expected to reduce salt loads of
irrigation return flows. In both endeavors the
nature and extent of soil salinity variability are
important. In the field experiments (U.S. Salini-
ty Lab Staff, 1977), we hope to demonstrate that
improved irrigation will reduce salt loading. In
the assistance to SCS, we're confronted with the
need to assess the magnitude of salt pickup or
precipitation under current operating con-
ditions. This conference provided an opportuni-
ty to combine our experience with that of the
Bureau of Reclamation to assess the sensitivity
of volume and salt content of return flows to
shifts in cropping patterns and soil salinity
variability.
RESULTS AND DISCUSSION
Soil Salinity Variability
Measurements of soil salinity (Oster and
Rhoades, 1975; Rhoades, 1976) or in situ
chloride concentrations provide a means of
assessing leaching fractions or field efficiencies
of irrigation water use. Chloride is a conser-
vative salt species, since it does not enter into
exchange or precipitation reactions. Conse-
quently, the chloride concentration in the soil,
Cs, depends upon the concentration in the
irrigation water, Cw, and leaching fraction,
LF, according to the well-known relationship
C w / C Q = LF.
(1)
The approximation symbol was used purposely
to indicate variations can occur due to several
causes: rainfall, changes in Cw with time, or
seasonal variation in water management
associated with over- or under-irrigation. The
Cs can also vary as a result of nonuniform
water infiltration or crop water uptake. The
variability in C s, both in magnitude and dis-
tribution, is relatively unknown. However, it is
understood to be large, and by inference, so is
the variability in soil salinity.
Variability of soil chloride concentrations
was characterized, based on data obtained from
three field experiments that comprise the follow-
ing irrigation methods, location, soils, crop and
soil depths:
253
-------
IRRIGATION RETURN FLOW MODELS
1. Sprinkler and flood.
a. Grand Valley, Colorado; Ravola
loam; corn; 0.60 - 0.90m and 0.90 -
1.20m.
b. Wellton-Mohawk Valley, Arizona;
Indio fine sandy loam; alfalfa; 0.9 -
1.2m and 1.2 - 1.5m.
2. Trickle and flood.
a. Wellton-Mohawk Valley; Dateland
fine sandy loam; citrus; 0.6 - 0.9m
and 0.9 - 1.2m.
The arithmetic and geometric means of soil
chloride concentration — the latter from log-
transformed data — standard deviations,
skewness and kurtosis were calculated for each
soil depth, crop and irrigation method and, for
citrus, by leaching treatment. A separate array
was created for each sampling date. If the
geometric means for the two depth intervals
were not significantly different (P = 0.10), the
data were combined into a single array. The
means of these arrays were then tested against
others for the same location. Where possible,
they were combined to create the largest possi-
ble sets of data. Probability diagrams of the
largest arrays for each crop are shown in
Figures 1 and 2. For corn and citrus the dis-
tributions in Figure 1 are skewed; more than
70% of the chloride concentrations are less than
the mean. The curvature is typical of data that is
log normally distributed. The coefficients of
variation (standard deviation/mean) ranged
from 0.6 for alfalfa to 1.0 for corn, which is also
typical of data not normally distributed. The
log-transformed data, Figure 2, were more nor-
mally distributed: the cumulative distribution
approached a straight line. The number of
observations less than the mean were about 55%
and the coefficient of variation rangedfromO.il
to 0.28, which is within normal distributions.
The above distributions were typical of all
data arrays, as shown in Figures 3 and 4. These
probability diagrams are plotted as a function
of the variable (Csj - Csi) S^, where S^ and
Csi are the standard deviation and mean for
each array i. Consequently, Figures 3 and 4
represent the general characteristics of all the
data. They show that the log-transformed data
are more normally distributed than the un-
transformed data.
The sensitivity of a statistical analysis may
be improved by the use of log-transformed
chloride data. For example, in our analysis of
variance of chloride data obtained beneath a
tree canopy, the use of log-transformed data
increased the number of significant sources of
variation. The standard deviation for
transformed data was relatively large, but in-
dependent of the mean. For untransformed
data, it was approximately equal to six-tenths of
the mean. However, the confidence limits about
the mean would be about the same. To illustrate,
we calculated (Table 1) the confidence limits
about the mean for the C s expected for different
LF's when C w =3.3 meq/1 and the sample size
(N)is 10. The selected value for Cw corresponds
to the chloride concentration of Colorado River
water at Imperial Dam. For untransformed
data, the upper and lower limits were calculated
from the expression,
or
Cw / LF± t.i S(CS)
Cs ± 1.84 * 0.6 C.
(3a)
(3b)
where 1.84 is the value of tat a probability of 0.1
for 9 degrees of freedom. For transformed data,
S was assigned the value 0.25, the average
found for all arrays. The antitransformation
from log C s back to equivalent untransformed
values was accomplished according to the
following equations, (Hald, p. 164, 1952):
Upper limit; Cs (1 + a),
Lower limit; Cs / (1 + a),
10log a = t.i S/Vn~
(4)
(5)
(6)
Both log-transformed limits are slightly higher
than the untransformed limits. The upper limit
is increased because of the abnormally large
number of high concentrations in the distribu-
tion of concentrations about the mean; the lower
limit, because the mode of the distribution is
somewhat less than the mean. For practical
purposes, the two values for the upper and lower
limits are the same. Even though the limits
calculated from transformed and un-
transformed data differ little, their theoretical
interpretation differs significantly. No valid
statement can be made about the probability
levels of the confidence interval calculated in
the untransformed case. The transformed data
closely approximate a normal distribution, so
that valid probability statements are possible.
Although the difference between the upper
and lower limits increases with decreasing
254
-------
HYDRO-SALINITY MODELS
leaching fraction, the confidence intervals
about the leaching fractions determined from
average chloride concentrations decrease with
decreasing leaching fraction or increasing field
efficiency of water use. The data in Table 1 were
converted to irrigation efficiencies, E, according
to
E=1-LF = 1-CW/C
(7)
Upper and lower limits of E were calculated by
using the corresponding limit values for C „.
The results are illustrated in Figure 5. The
confidence interval becomes smaller with in-
creasing efficiency because of its inverse
dependence on Cs. Consequently, for low field
efficiencies of 0.4 to 0.6, the likely error in
irrigation efficiencies calculated from chloride
data would be relatively large and so would the
error in predicted return flows.
Improvements in irrigation efficiencies
from 0.45 to 0.6 are commonly projected in study
plans being prepared as a result of the Colorado
River Basin Salinity Control Act (PL 93-320). If
these are representative of field irrigation ef-
ficiencies, then chloride measurements ap-
parently would not be useful in evaluating the
consequences after implementing improved
practices. At efficiencies above 0.8 the level of
confidence would be greatly improved.
Projections of mass emissions of salt from
irrigated projects based on soil salinity
measurements would be subject to a similar
uncertainty. Mass emissions of chloride, Figure
6, were calculated from the log limits in Table 1
from the equation
MASSC1=CS Dp = Cs * A* Cu (1 E-l) = Cs • .
r cx i
C»
[_CS-CWJ
(8)
where Dp represents deep percolation. Area, A,
was assigned a value of lha and consumptive
use, Cu, was assigned a value of 1500 mm.
Obviously, at irrigation efficiencies of 0.5 the
degree of uncertainty would be very large;
however, at efficiencies of 0.8, soil salinity
measurements would be a sensitive measure of
mass emissions.
Return Flows — Crop and Water Data
The water budget for an area is the stan-
dard method used to estimate the volume of
return flows from an irrigated field or irrigation
district. Deep percolation from irrigated fields is
often a major component of irrigation return
flows. Water use efficiency varies from crop to
crop because of different cultural practices.
Consequently, the sensitivity of a water budget
analysis to different cropping patterns is of
interest.
The Yuma office of the Bureau of Reclama-
tion recently reported the results of a sensitivity
study! of the Wellton-Mohawk Irrigation Dis-
trict. This district is relatively simple
hydrologically; irrigation return flows are con-
fined to a closed aquifer, which is pumped to
maintain satisfactory water table levels; un-
derflows and rainfall are small, about 4% of the
diversions; and return flows from the lined
canals and consumptive use by phreatophytes
are each estimated to be about 10% of the
diversions.
A linear optimization procedure was used to
perform the sensitivity analysis. Crop input
data included a 6-year (1970-1975) compilation
of actual irrigated areas by crop. The cropping
patterns of the 15 crops grown in the district
were modeled with 17 equations, one for each
month and five additional equations to account
for midmonth cropping pattern changes. Each
equation represented the total area of irrigated
crops. In the optimization procedure, bounds
were placed on the total and individual crop
areas so that they remained between 6-year
maximum and average areas. For a given set of
efficiencies, individual crop area was optimized
to obtain estimates of maximum and minimum
return flows, D p, according to the relationship,
15
DD = v A * CU * [1/Ei - 1], (9)
where i represents an individual crop.
The results given in Table 2 for one set of
efficiencies show that deep percolation can vary
considerably depending on the farmer's choice
of crops to be grown. At program level 1, which
represents a minimum level of improvement, it
varied from 0.132 to 0.175km*. Deep percolation
was maximized as a result of increased areas of
3Measures for Reducing Return Flows from the Well-
ton-Mohawk Irrigation and Drainage District, 1976
Annual Report, Technical Field Committee, Yuma
Projects Office, Bin 5569, Yuma, Arizona, 85364.
255
-------
IRRIGATION RETURN FLOW MODELS
cotton, lettuce, melons, wheat and sorghum.
The area of these crops were least constrained
because they can be grown in rotation with
others: 1) cotton, wheat and sorghum in rotation
with lettuce and barley, and 2) wheat in rotation
with sorghum, cotton, and melons.
Cropping patterns significantly affect deep
percolation from fields for different sets of
overall crop efficiencies (Fig. 7). Under the
cropping pattern for 1974-1975, deep percolation
ranged from 0.192 to 0.249 km:t and for irriga-
tion management improvement level 3 efficien-
cies it changed from 0.102 to0.142km1. Figure?
also shows estimated return flows for the total
system, which include losses due to canal
seepage and consumptive use by phreatophytes
and gains due to rainfall and underflows. The
variation because of different cropping
patterns, A, is a major component of that for the
total system, B (.48 > A/B < .71). The relative
uncertainty in estimated return flows for the
total system increases from 18 to 32% [100 *
.5B/(.5B -i-D)] as efficiency increases from
current levels to those projected for level 3. In
other words, the more efficient the irrigation
system, the greater the relative error associated
with estimated return flows; the largest single
component of uncertainty is a shift in cropping
patterns.
CONCLUDING COMMENTS
Return flows estimated from salinity, or a
water budget analysis, exhibit opposite sen-
sitivities to increasing field irrigation efficien-
cies. The former depends upon a ratio where, at
lower efficiencies, a small variation in a small
number results in a large change. At high
efficiencies, the larger numbers assure a small
ratio with little dependence on variability. The
sensitivities, at low efficiencies, for water
budget data depend on relatively small changes
in large numbers. The situation reverses at high
efficiencies; the water input is decreased,
whereas the consumptive use and its uncertain-
ty remains fixed. The sensitivities of both
methods to assess return flows may be com-
plementary if both are used to obtain an op-
timized estimate of return flows.
Salinity or chloride variability would be
dependent upon the volume of soil sampled. The
data we presented was based on a soil volume of
about 9 x 10-5 m 3 (3 cm 2 x 30 crn). This
variability would not apply to drainage tile
effluents; the effective volume sampled could
approach 2 x 10 4 m 3 (400m x 30m x 3m). Such a
sample integrates the variability over space as
well as time. The variability would be smaller,
but the concentration measured would be more
difficult to interpret. The variability of soil
salinity measured using resistivity techniques
would be intermediate because the sample
volume is about 1m 3.
Time of sampling is another variable we
have not considered (Jury, 1975; Raats, 1977). It
certainly is one limitation with regard to the
usefulness of salinity or individual ion concen-
trations to estimate return flows or mass
emissions. The lag time between imposition of
management changes and resulting change in
soil solution or effluent concentrations would
increase with increasing efficiencies. Long-term
decisions based on short-term measurements
could result in serious errors.
The sensitivity data we have presented
probably are conservative. The chloride data
were obtained from soils irrigated for many
years. The variability in unirrigated soils in the
West could be even greater. The Wellton-
Mohawk area represents a relatively simple
hydrologic system. The degree of uncertainty
would undoubtedly be greater in areas with
multiple return flow paths, greater sources of
underflows and rainfall, and shorter growing
seasons.
One objective of the presentation is to raise
several questions about the uses of large-scale
models: 1) Is it realistic to attempt large-scale
verification? 2) What constitutes proper use of
models quantitatively or to predict trends? 3)
Should we increase the efforts in sensitivity
analyses on more systems? The last question
may be the key. If more were known about the
sensitivity of model output to the variability of
TABLE 1
Ninety percent confidence interval of mean soil
chloride concentrations where the chloride
concentration of the irrigation water is
3.3 meq/1 and n = 10.
Untrara formed
Leaching
Fraction
0.6
0.5
0.4
0.3
0.2
0.1
0.05
cs
ICu/LF)
5.5
6.6
8.2
10.0
16.B
33.3
66.0
Stand.
Dev.
3.0
4.0
5.0
6.0
10.0
20.0
40.0
Limits
Upper
7.2
8.9
11.1
13.5
22.3
44.9
89.1
Lower
3.8
4.3
5.3
6.5
10.7
21.7
42.9
Log transformed
Stand.
Deu.
.25
.25
.25
.25
.25
.25
.25
Limits
Upper
7.7
9.2
11.4
13.9
23.0
46.4
92.1
Lower
3.9
4.7
5.9
7.2
11.8
23.8
47.3
256
-------
HYDRO-SALINITY MODELS
input variables, it might be possible to for-
mulate more definitive answers.
TABLE 2
Deep Percolation from Farms — Reflecting
various levels of efficiencies, historical
crop data and Acreage reduction.
99.99
Program Level 1.
Minimum Irrigation
Efficiencies
Crop
Alfalfa
Pasture
Cotton
Safflower
Lettuce
Melons
Wheat
Barley
Sorghum
Leaching
Misc.
Bermuda G.
Grapefruit
Lemons
Oranges
TOTALS
CropCu.
(mm)
1830'
1040'
1070
1160
240
188
671
Ml
980'
1220
1280
1000
Area
thai
9206
870
4302
257
3532
1264
5868
60
3874
139
661
2673
75
324
858
33962
Eff.
78
•_
'
-"
21
'
70
70
•
5
44
,.
.:.
40
67
Deep Perc
(km*l
0.0475
0.0083
0.0122
0.0009
0.0258
0.0105
0.0169
0.0002
0.0152
0.0008
0.0049
0.0112
0.0014
0.0062
0.0129
0.1748
Program Level 1.
Maximum Irrigation
Efficiencies
Area
(ha)
9112
870
3694
257
1450
239
5205
11
1525
58
661
2673
75
:•
858
27141
Eff.
79
1
77
21
fi
<:-
'
14
'
.
«
,
70
Deep Perc
tkm'l
0.0443
0.0051
0.0105
0.0009
0.0106
0.0020
0.0150
0.0004
0.0060
0.0003
0.0049
0.0112
0.0014
0.0062
0.0129
0.1315
'Consumptive use reduced to estimate incomplete growth cycles for a portion of
these crops which are normally in rotation each year.
"Consumptive use estimated to reflect expected miscellaneous crops obtained
from historical crop data.
O-O citrus; n= 248 , s
x—x corn, n= 142 , s =
A—A alfalfa , n = 94 s
0.01
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Log|Q Chloride Concentration
8
Figure 2. Cumulative probability for transformed
chloride concentrations.
;
_
.,
V
|
'_
99.99
99.80
99.00
95 00 h
8000
60.00
40.00
20.00
500
1.00
0.20
0.01
A—A alfalfa; n= 94, s= 10.4
x—x corn , n= 142, s= 13.0
O—O citrus ; n = 248 ,s= 11.2
10 20 30 40 50 60 70 80
Chloride Concentration (meq/l)
Figure 1. Cumulative probability untransformed
chloride concentrations. S represents the standard
deviation.
....
I
:
-C
:
rx
:
!
c
.7
r
J J J J
99.8O
9900
95.00
8000
60.00
40.00
20.00
5.0 O
1.00
0.20
n m
0
x
a~
O* D
^o?DD°
^*O°
A_j£ O
tfP
$**
Qfi1
/O corn ; n = 456
x citrus, n = 68l
A alfalfa; n= 142
a citrus, n- 8IO
A
- x -
9B
AX
A"
X
O
- O
x
1 1 1 i 1 II
-4 -3
1
.
/Si
Figure 3. Cumulative probability for untrans-
formed, reduced chloride concentrations.
257
-------
IRRIGATION RETURN FLOW MODELS
99.99
99.80
99.OO
95.00
"o 8000
6000
4000
2000
o
500
I OO
0 20
001
•
Ax°0
AxO
_ '
O corn, n =456
A olfalfo , n= 142
o citrus ,n=8 10
x citrus, n=68l
T
OB
1 1 1
-4-3-2-10 I 2 34
(Cs, - Csi) /Si
Figure 4. Cumulative probability for transformed.
reduced chloride concentrations.
.
•
-
.-
c
.
1.0
0.8
0.6
0.4
.
^ 0.2
ant it ran s formed
untronsformed
0.2 0.4 0.6 0.8
Irrigation Field Efficiency
Figure 5. Confidence intervals for inferred field ef-
ficiencies. Calculated from transformed (P = 0.10)
and untransformed chloride data assuming a sample
size of 10.
ACKNOWLEDGMENT
The authors want to express their apprecia-
tion to Dr. Jan van Schilfgaarde for helpful
discussions and advice during the course of this
work and to J. O. Goertzen, D. F. Champion. R.
D. Ingvalson and M. Clark for their help in
collecting and analyzing the soil samples.
REFERENCES
1. Hald. A. 1952. Statistical theory with
engineering applications. John Wiley and Sons,
Inc., New York. 738p.
2. Jury, W. A. 1975. Solute travel-time
estimates for tile-drained fields: II. Application
to experimental studies. Soil Sci. Soc. Amer. J.
Proceedings 39: 1024-1028.
3. Oster, J. D. and Rhoades. J. 1). 1975.
Calculated drainage water compositions and
salt burdens resulting from irrigation with river
waters in the western United States. J. Environ.
Qual. 4:73-79.
4. Raats, P. A. C. 1976. Convective
transport of solutes by steady flows.
Agricultural Water Management. (In press).
5. Rhoades, J. D. 1976. Measuring, map-
ping and monitoring field salinity and water
table depths with soil resistance measurements.
FAO Soils Bulletin 31:159-186.
6. U.S. Salinity Lab Staff. 1977. Interim
Report-minimizing salt in return flow through
irrigation management. EPA Report. (In press).
—
--
:
3
1
_
0)
I
_
_£
^
_
0
mean
max
mm
0.5 0.7 0.8
Irrigation Efficiency
Figure6. Confidence intervals, (P = 0.10) for
chloride content of irrigation return flows of different
levels of field irrigation efficiencies.
258
-------
HYDRO-SALINITY MODELS
*: E
Rl = A/B
240
200
160
in
c
*-
- 1 20
Q_) '
rr
a>
o 80
c
'o
Q
40
Q
-.30
in °
2
-.25
R2=OE
-.20
-.1 5
-.10
-.05
u.
fl
-^
,
•
:
2
o
h
B
^
•
/
' /
:
X
n~ a5A
0.5 A + C
P, -0.5B
56% ' 'O.SB-l-b
Rl =0.71
R3 = O.I8
o
-------
Modeling the Irrigation
Return Flow System —
Current Capabilities
and Future Needs
WYNN R. WALKER
Agricultural and Chemical Engineering Department,
Colorado State University, Fort Collins, Colorado
ABSTRACT
A large number of mathematical models
have been developed and tested for simulation
of irrigation return flow systems. The strengths
and weaknesses of the technology have been
examined as part of two recent studies by the
author. This paper discusses what seems to be
some of the more critical problems in using
these models, the relative strengths of those
existing, and important research and develop-
ment needs for maximizing their utilization in
the future.
INTRODUCTION
Return flows from irrigated agriculture
generally contain at least some concentrations
of salts, nutrients, pesticide residues, and
sediments. Each of these "pollutants" can have
adverse effects on the future benefical uses of
receiving waters. Measures to diminish the
impact of irrigation return flows must be
evaluated in terms of their cost-effectiveness,
effects on crop production, and the magnitude of
the detriments they create.
Mathematical simulations, generally com-
puter based, have become the tools utilized to
evaluate both the irrigation return flow system
and alternative means for its control. The
existing body of these "models" is comparative-
ly large, which might be attributable to a
number of factors including: (1) the broad and
complex scope of the irrigation return flow
system; (2) the large number of both scientific
disciplines and researchers interested in one
phase or another of the problem; and (3) model
duplication.
In the last few years, the author has been
collecting the various irrigation return flow
models and recently evaluating their
characteristcs. Listings from about 1 / 2 to 2/ 3 of
the more than 50 available models have been
obtained and examined. This paper is intended
to briefly summarize some general thoughts
regarding this experience.
The actual beginnings of my effort occured
during March of 1975. A three-day workshop
was sponsored by the Environmental Protec-
tion Agency which involved ten participants,
the author, and EPA Project Officer, Dr. Arthur
G. Hornsby. Eight of the workshop group are
participating in the papers or case study presen-
tations at this conference. Those present by no
means represent an exclusive body of modelers,
but were selected to sample the various research
and administrative concerns. The workshop
involved three phases: (1) discussions concern-
ing modeling philosophy; (2) presentations of
various modeling activities; and (3) examina-
tion of the recent modeling efforts within the
Bureau of Reclamation. Two of the most in-
teresting results from the workshop were the
opinions generated in regard to modeling
philosophy and the pertinent research needs for
improving this technology. After the workshop
had been evaluated, another small project was
funded by the Environmental Protection Agen-
cy to collect and evaluate the various models
available.
Space does not permit an exhaustive review
of the available models nor an in-depth look at
their capabilities. Instead, the writer would like
261
-------
IRRIGATION RETURN FLOW MODELS
to briefly summarize a number of issues that
have been raised in order to place a perspective
on today's irrigation return flow models.
PROBLEMS IN MODEL UTILIZATION
A simulation model is generally a set of
mathematical relations applied to approximate
boundary conditions that attempts to simulate
the relationships between variables in the
natural system. It is useful in understanding the
natural system and in evaluating its behavior
under proposed management alternatives. Re-
searchers have been prolific in applying model-
ing concepts to the various segments of the
hydrology in an irrigated area.
A number of modeling efforts have been
reported with various degrees of modification in
the work of others, but the most prevalent
practice is to develop a model for the specific
purpose at hand. Consequently, irrigation
return flow models of all kinds are somewhat
individualistic in nature, each tending to ex-
press the modelers perception of a problem
solution, rather than a more generalized ap-
proach simplified to a particular need.
The use of existing irrigation return flow
models pose some severe problems. First, an
investigator or planner is almost always faced
with analysis of problems extending beyond the
limits of his own expertise. Thus, at least some
part of most modeling efforts involve com-
putations which must be "trusted." However,
cases occur when models are improperly utilized
because the assumptions governing the original
formulation are no longer valid. Furthermore,
even when the assumptions are explicit (which
is not the general case), the untrained user may
not know whether the assumptions are valid or
not. It is, of course, the responsibility of the
model users to insure the applicability of a
model, but the model developer must also be
charged with defining the limitations of the
model.
A second major problem in using existing
models is the variability in programmer skills
and computer languages. One of the ironic
concepts encountered in using computers is the
notion that efficient coding is cost-effective.
Modern computers are so rapid in their internal
operations and software manipulations that the
difference between the expert and novice in
programming for a particular need is orders of
magnitude less than personnel costs, especially
in the agricultural and engineering disciplines.
In fact, research budget categories allocated to
computer utilization probably never exceed 5-
10% of the total research costs. Consequently,
the disadvantages in simpler programming
with respect to cost are probably more than
offset by the added ease in adapting to other
computers and utilization by a wider audience.
For example, most agricultural and engineering
students only become familiar with the Fortran
language so that use of other language pack-
ages such as CSMP (for IBM computers) may
significantly limit the utilization of the model.
In examining the models of others, and
doing substantial modeling myself, the conclu-
sion appears to be that very few major computer
codes are verified, or in other words, checked to
determine if they actually make a calculation
according to the intended format. Even the most
expert programmers misplace parentheses or
improperly nest indexes and counters. If the
generated results appear reasonable, then the
code is considered correct; otherwise, the
specific problem is amended. Occasionally,
problem solutions involve a small fraction of the
loops and branches found in a program so that
some unused program routes are not always
checked. When another user tries to employ the
model under different conditions where other
program segments are used, a great deal of
difficulty may be experienced with the program.
Models should be verified against computations
made outside the computer code.
Irrigation return flow models available to-
day occasionally contain segments having
vastly different sensitivities in their simulation
of various elements of the system. Many of the
soil physics and chemistry models were
developed under concise laboratory conditions
where data quality is controlled within reliable
limits. When these models are coupled with
hydrologic models developed for field scale
applications, a significant disparity exists in
refinement. Data at the field scale are highly
variable so that in using detailed models the
necessary assumptions may render them no
better and probably worse than more simplistic
approaches.
And finally, there is often a tendency to
place too much credibility on the output from a
computer simulation. The irrigation return flow
system is very complex. In fact, many of those
complexities are only marginally understood.
As a result, simulation models are nearly
always empirical to some extent, and therefore
are at best, tools to assist the investigator. They
262
-------
CAPABILITIES AND NEEDS
are not, however, instruments for verifying
extended hypotheses as is the case when con-
clusions are drawn from extensive output based
on questionable or poor data.
CURRENT CAPABILITIES
During the 1975 modeling workshop, a
number of the segments of the irrigation return
flow system were thought to be reasonably well
simulated with existing models. This opinion of
course is in the relative sense in considering the
irrigation return flow models as a whole. The
examination of the various available models
since the workshop has generally substantiated
the earlier expressions.
Of the irrigation return flow system models,
probably the best developed for field scale
applications are the evapotranspiration-irriga-
tion scheduling models. Evapotranspiration es-
timating procedures have been developed for a
wide variety of data availabilities, although
irrigation scheduling is generally limited to two
or three of the more inclusive methods. The
weakness of these models is that they are
primarily limited to predicting early growth
water demands and the effects of stress on use
rates.
The flow of water after entering the soil is
also well modeled with existing programs. Re-
cent developments in regards to two- and three-
dimensional flow predictions have broadened
the scope of this area of irrigation return flow
models to encompass all of the irrigation system
varieties. Unsaturated flow models of both
transient and steady-state conditions can be
found in numerous places throughout the
literature. Most of these models, however, do not
include terms to account for moisture extraction
by growing crops, and the boundary conditions
assumed for surface water applications are
occasionally unrealistic in the true physical
sense. The most severe constraints in using
these models lie in characterizing soil hydraulic
properties since such data are not generally
available from one location to another.
Chemistry predictions associated with un-
saturated flow are relatively well developed for
the basic components of salinity. Some pesti-
cides and volatile constituents like the nitrogen
species are much more complex and therefore
less accurately modeled. Again, respesentative
field data pose the more rigid restrictions on the
reliability of these models. A number of soils
contain naturally occurring salts that are dis-
solved by flows passing through the un-
saturated zones (salt pickup effects). Most
models do not handle this condition, although
some of the more recent modeling studies in
Utah have considered the problem by adding
sink and source terms to existing models.
Moisture and chemical flows in the
saturated regions can be simulated with
reasonable accuracy whenever data are
available and the physical nature of the hidden
system can be ascertained. Unfortunately, most
areas have very little groundwater data. In
irrigated fields underlain by drainage systems,
the model capabilities are reasonably good
since verification has been possible during
model development.
These areas of irrigation return flow model-
ing represent the best developed segments of the
modeling topic, but they consider only in-
dividual processes in the system. On a more
integrated scale, there exist a number of general
hydro-chemical models which include, to some
degree at least, the various processes present.
These models generally have expanded time
(one-week to one-month) and spatial (one-ha to
several thousand ha) reference frames and
therefore predict the distribution of the average
conditions. These models also considered very
simplistic simulations of the detailed hydro-
chemical processes because large scale data
resolution is very poor. However, in those cases
where verification has been possible, predictive
capability has often been comparable to more
detailed analyses.
A final group in the modeling technology is
the models which utilize simulation results in
management and optimizational contexts. Very
few if any such models exist with enough
generality to allow widespread utilization by
others, and in fact, this area of modeling is
almost exclusively a philosophy and "art"
oriented effort, primarily because of the totally
site-specific nature of problems at this level of
investigation. The science utilized in such
modeling efforts is extremely well developed
and proven even though structuring problems
in the "systems analysis" format is generally a
skill.
FUTURE NEEDS
The future needs in irrigation return flow
modeling can be broken down into two areas:
(1) research and development needs; and
263
-------
IRRIGATION RETURN FLOW MODELS
(2) needs in regards to disseminating existing
models to the user community.
Irrigation return flow models have devel-
oped over the last few years at three distinct
levels. First, what might be called process level
models which are limited in scope to a single
hydro-chemical process. These models have
been written at the most basic level, justified by
existing information. However, a specific
process might have been modeled at various
levels of sophistication depending on the re-
quirements of its use and available data.
Simulation of cropland evapotranspiration is a
good example of the typical process model. In its
most refined forms, the crop use rates are
determined by a relationship like the Penman
combination equation. Other simple versions
may use pan evaporation, Jensen-Haise,
Blaney-Criddle, etc., depending on data and
intent. Other process models include such
processes as unsaturated flow and associated
chemistry. Among the process level models,
there are several important research and
development needs. Those include: (1) improved
simulation of irrigation water application uni-
formities as influenced by system design and
operation; (2) evaluation of the impact of irriga-
tion frequency and timing on root zone water
and chemical flows; (3) improved simulation of
nitrogen transformations in irrigated soils; (4)
improved simulation of root extraction of
moisture and nutrients, particularly the depth
distribution of the extraction pattern; (5) in-
vestigation of crop growth-stage and stress
coefficient for estimating early season
evapotranspiration; and (6) evaluation of un-
saturated soil chemistry predictions when large
quantities of natural salts are present in
irrigated soils.
The second level of model development
might be termed the subsystem level represent-
ing a combination of various process models
capable of simulating a major part or all of the
irrigation return flow system. For instance,
subsystem models can be fabricated to examine
in various degrees of detail the region between
the soil surface and water table. Process models
would include irrigation uniformity (including
infiltration), evapotranspiration-root extrac-
tion, unsaturated flow and unsaturated chemis-
try. A sampling of the opinion of the earlier
workshop participants indicated that most
potential model users involved in research
would prefer to develop their own subsystem
level models to meet the specification of scope
and detail associated with their problem. One
must suspect, however, that the planner or
consultant would prefer the system level model
since they may not be comfortable synthesizing
process level models which may be only vaguely
understood. Consequently, an important fu-
ture need concerns the sensitivity of these
models to various kinds of input data. In other
words, for subsystem level models with varying
degrees of sophistication and varying time and
spatial resolutions, which data must be most
carefully considered? In addition, the con-
ditions under which these models can and
cannot be applied should be clearly delineated.
The third modeling level is the hydrologic
and irrigated system model which includes the
irrigation return flow system as a subprogram,
or as the entire program treated at a "macro"
level. These models tend to large area, large time
averaging systems, predicting on an input-
output basis. They are generally comprised of
the more simplistic process simulations and
subsystem level programs since their time and
spatial resolution usually do not justify more
sophisticated programs. And, large scale data
are generally poor and few. Further needs in
respect to these models may be more oriented to
evaluation of how well sampling data represent
field conditions, although parameter sensitivity
analysis and assumption validity remain
critical.
One of the interesting results of the
workshop alluded to earlier was that most of the
participants did not believe information re-
garding the various irrigation return flow
models was generally available, particularly
with regard to operational instructions. Most of
the participants felt that very little information
was being published relative to the sensitivity of
model parameters and results as affected by
various kinds and qualities of input data. Con-
sequently, the majority of models are not util-
ized by other than the authors and a few
associates, especially by those in other dis-
ciplines. Thus, an important future need is to
collect the various computer programs
available, simplify their codes in order to max-
imize their adapability to the wide range of
computers being utilized, and make the
programs available to potential users with
sufficient instructions to facilitate their use in
the proper manner. After this step, it might be
useful to develop a more consistent or uniform
program library by reprogramming those
already existing models and adding models,
264
-------
CAPABILITIES AND NEEDS
where research findings have added to model-
ing capabilities.
The irrigation return flow modeling system
is readily applicable to many other problems
associated with the irrigation of agricultural
lands. Many of the models are already being
utilized to increase irrigation efficiency,
evaluate organic waste disposal practices, max-
imize crop production, and conserve energy,
fertilizer, and water resources. It appears that
our current model capabilities are sufficient for
most needs although hard to apply and uncer-
tain to interpret. A few simple steps in our future
work should go a long way in improving these
capabilities.
265
-------
Case Studies
-------
Application of Modern
Irrigation Technology in the
Mesilla Valley, New Mexico
T. W. SAMMIS and C. M. HOHN
Agricultural Engineering Department, and Cooperative Extension Service, respectively,
New Mexico State University, Las Cruces, New Mexico
ABSTRACT
Current engineering technology is being
applied to a Demonstration Farm in the Mesilla
Valley, New Mexico to demonstrate the
feasibility of decreasing the large amount of salt
returned to the Rio Grande River System
through return-flow drainage. The reduction in
return flow is accomplished by irrigation
scheduling techniques and more-efficient
irrigation methods, such as drip irrigation. A
small pecan orchard has been converted from
flood irrigation to drip irrigation resulting in the
water application reduction greater than 50
percent with a marked visible increase in
growth rate over that of the previous year.
Irrigation scheduling increased the efficiency of
irrigation timing on the farm resulting in an
improved soil condition, in some cases one less
irrigation. Monitoring of the applied water on
the Demonstration Farm resulted in estimates
of irrigation efficiencies for the season ranging
from 54 percent for wheat to 97 percent for
alfalfa. The high efficiency for alfalfa was
probably due to the high water table which
contributed water to the crops in addition to the
applied water. Monitoring of the soil salinity is
enabling the investigators to follow the move-
ment and buildup of salt below the root zone
with increased irrigation efficiencies. Informa-
tion gained from the Demonstration Farm is
being used by the Cooperative Extension Ser-
vice in an educational program. The various
methods used to make the public aware of EPA's
efforts on the Demonstration Farm include
media releases, tour brochures, and field days.
INTRODUCTION
Current engineering technology is being
applied to a 182-hectare Demonstration Farm in
the Mesilla Valley, New Mexico to show the
feasibility of alternative water-management
practices on the quality of drainage return flow
and soil salinity in the Upper Rio Grande Basin.
The crops grown on the farm are: wheat,
tomatoes, cotton, lettuce, tobasco peppers, chile
peppers, alfalfa, and a pecan orchard. The
source of the irrigation water is Rio Grande
water from the Elephant Butte Irrigation Dis-
trict augmented by wells installed at various
locations throughout the farm.
Figure 1 presents a map of the farm with the
irrigation distribution system and the
associated hectares under production for each
crop. Because of the deep drainage problem that
exists in the valley, drains have been installed
to maintain the water table at a depth of 2-2.5
meters.
METHODS
To improve the farm irrigation efficiency
for the various crops, it was necessary to
measure the water by installing meters on all of
the irrigation pumps, and Parshall flumes in all
the irrigation ditches receiving surface water. A
complete record was kept of the water applied
from both surface and groundwater sources on
each of the fields.
Rain gauges were installed at the Demon-
stration Farm and the New Mexico State Uni-
versity Plant Science Farm to measure the
precipitation in the area; however, the mean
annual precipitation averages less than 25 cm
with most of the rain falling in the summer from
thunder showers caused by tropical air masses
coming in from the Gulf of Mexico. Essentially
all the consumptive-use requirements of the
crops are satisfied by irrigation.
269
-------
CASE STUDY: MESILLA VALLEY
Figure 1. A map of the crops grown on the
EPA Demonstration Farm in the
Mesilla Valley.
Flow rates and water qualities were moni-
tored in the La Mesa Drain passing through the
center of the farm in order to establish the
amount and quality of the return flow from the
surrounding area. Piezometer wells monitored
the groundwater fluctuations.
To implement the objective of the
demonstration project, 1.33 hectares of pecans
were converted from flood irrigation to drip
irrigation. In the process of converting the
orchard to drip irrigation, a well was installed to
provide clean, good-quality water for irrigation
and domestic uses. An automatic back-flushing
sand filter was installed to supply the necessary
filtration required to operate a drip irrigation
system.
An irrigation management schedule service
is being provided to the farm for scheduling
irrigations based upon the estimated consump-
tive use of the crops and the available moisture
within the root zone (2). The consumptive use is
estimated by a computer model using microme-
teorological data. Prediction is made on the
estimated consumptive use in the future. When
the available soil moisture has been depleted by
50%, irrigation is scheduled. The information is
checked weekly in the field and the model
updated continously to account for the changes
in the micrometeorology. Using this approach,
the timing of the irrigations are as close to
optimum as is practical. Information about the
daily consumptive-use rate on several crops has
never been determined for the conditions in this
valley. Under these circumstances, the
irrigations were scheduled based upon soil
moisture availability measurements made in
the field.
RESULTS AND DISCUSSION
Table 1 presents the irrigation date of
applied water, and the estimated consumptive
use for that time period for each field on the
Demonstration Farm. The table also presents
the information on the percentage of water
supplied by surface vs. groundwater ranging
from 36% surface water for the tomato fields to
87% for the cotton fields. The farm, on an
average, received 69% of its water from the
Elephant Butte Irrigation District and the rest
from pumped groundwater. The seasonal irriga-
tion efficiency, based upon the consumptive use
divided by applied water without considering
rainfall, ranged from 54% for one of the wheat
fields to 97% for the alfalfa fields. The alfalfa
evapotranspiration was measured ina lysimeter
nearby. The cotton and wheat evapotranspira-
tion rates were determined from the Jensen-
Haise method (1). Because of the shallow water
table, the alfalfa roots were able to extract
moisture from the groundwater system. If the
amount used from the water table was added to
the applied surface water, the computed efficien-
cy would be less. The average farm irrigation
efficiency, using the service of irrigation
scheduling, was 72%. The variability in the
seasonal irrigation efficiency from field to field
growing the same crop was due, in part, to how
closely the irrigations were supplied to the field
compared to the dates recommended by the
irrigation scheduling. Irrigation efficiencies
between fields growing similar crops were
similar for the tomato and wheat fields but
varied for the cotton and pepper fields.
Data on the amount of applied water to the
pecan orchard using the drip irrigation system
is given in Table 2. The seasonal amount of
applied water, spread over the entire 1.33 hec-
tares, for the growing season was 21.6 cm. This
is considerably less than would be applied by a
surface-irrigation method estimated to be 60 to
80 cm, based upon 6 to 8 irrigations applying
10 cm per irrigation. The low application rate
supplied by the drip system is made possible
because the emitters only wet a small area
around the trees.
270
-------
MODERN TECHNOLOGY APPLICATION - NEW MEXICO
TABLE 1
Irrigation water application, source of water, and
evapotranspiration for crops on the EPA
Demonstration Farm, 1976.
Irrigation
Date
Alfalfa-Fd. 10
3-15-76
5-18-76
6-3-76
6-22-76
7-12-76
8-7-76
8-20-76
9-19-76
10-8-76
Season Totj!
Percentage
Seasonal Ft
Cotton-Fd, 4 We
3-27-76
5-12-76
8-5-76
8-17-76
8-2S-76
10-27-76
Season Total
Recommended Total
Irrigation Water
Date Applied
3-17-76
5-13-76
6-8-76
6-24-76
7-10-76
8-5-76
8-21-76
9-18-76
10-7-76
Surface Water
eld Irrigation Efficiency
st
pre- irrigate
SIC
6-30-76
7-31-76
8-22-76
9-2-76
Percentage Surface Water
Seasonal Field Irrigation Efficiency
3-29-76
5-18-76
6-22-76
8-7-7b
8-27-76
10-5-76
10-27-76
Season Total
Percentage
Cotton-Fd. 12
3-19-76
i-13-76
6-29-76
7-25-76
8-21-76
8-26-76
9-29-76
10-27-76
Season Total
Seasonal F
Wheat-Fd. 3
1-13-76
3-14-76
4-7-76
4-29-76
5-13-76
7-12-76
pre- irrigate
NI
7-1-76
8-5-76
finished
Surface Water
NI
7-1-76
7-22-76
8-26-76
ield Irrigation Efficiency
3-17-76
4-8-76
4-27-76
5-12-76
harvested
Seasonal Field Irrigation Efficiency
Wheat-Fd. 5 k 6e
1-28-76
3-26-76
4-22-76
5-18-76
6-16-76
3-26-76
4-23-76
5-16-76
harvested
Percentage Surface Water
Seasonal Field Irrigation Efficiency
Lettuce-Fd. 3
8-8-76
8-26-76
9-27-76
10-13-76
none
46.34
25.02
72.34
65.86
57.31
32.50
37.63
37.63
41.91
416.54
60. Ti.
9i . 6i
39.93
15.4!
17.99
16.28
13.19
12.94
115.73
847>
24,65
9.8o
15.61
12.37
21.07
83.56
75V.
777
5.02
5.18
6.83
6.83
6.20
6.42
36.48
637.
1127.
40.87
41.90
42.52
42.89
26.25
194.43
79%
557.
19.23
15.41
35.00
22.17
91.81
10W
547.
41.04
32.66
31.79
14.67
120.16
Appl icat ion
Per Hectare
1.643
.883
2.556
2.343
2.039
1.156
1.339
1.339
1.491
14.789
2.465
.943
1.126
1.004
.882
.791
7.151
2.313
.913
1.461
1.156
1.948
7.791
.735
.761
1 . 004
1.004
.913
.943
5.360
1.463
1.491
1.522
1.534
.938
6.948
1.461
1.156
2.647
1.674
6.938
2.739
2.191
2.100
.974
8.004
Source of
Water
$
G
S
S
s
s
s
s
c
s
s
s
s
s
s
s
G
s
s
G
s
s
G
G
S
S
s
s
s
s
s
s
S(527.)
S(377.)
S(687.)
G
Evapotranspiration
(ET)
for time period
(kro-'/ha)
1.040
2.011
.999
1.265
1.420
1.750
2.790
3.398
.776
14.151
.139
.766
2.055
.652
.596
1.778
5.990
.185
.538
1.149
1.310
2.392
.423
5.997
.177
.766
1.347
1.544
.294
1.296
.573
5.997
.005
.477
.870
.766
.583
1.164
3.865
.005
.872
.941
1.144
.817
3.779
4.5141
Percentage Surface Water 46%
Seasonal Field Irrigation Efficiency 56%
271
-------
CASE STUDY: MESILLA VALLEY
TABLE 1 continued
Recoraaended
Date Date
lomatoes-Fd. 8
3-20-76
3-23-76
5-1-76 5-1-76
5-7-76 5-7-76
6-17-76 6-13--6
7-13-76 7-10-76
8-4-76 8-2-76
8-20-76 I'.nished
S«-.,son Total
Percent age Sun... Water
7o«jto( s-Fd. ll'1
5-18-76 5-ltt-7o
6-22-76 6-20-76
7-72-76 7-15-76
£-3-76 M
8-7-7fj .s-5-76
8-77-70 S-22-70
S-31-76 M
9-29-70 9-11-76
Season Totnll? U-10 8)
Peppers-Fd. 4 East
3-Ii-"6 pre-irrigate
0----0 6-13-76
t-22-?6 6-31-76
7-12-76 7-10-76
6-5-76= 7-31-76
6-18--6 8-17-76
8-28-76 8-27-76
10-5-T6 9-20-76
Season Total
Tabasco Peppers-Fd. 2
4-12-76 pre-irrigate
.-22- 76
5-12-76
5-15-76 5-15-76
6-7-^6 6-2-76
6-20-76 6-30-76
7-8-76 7-16-76
8-8-76 8-5-76
8-18-76 8-17-76
9-7-76 9-1-76
9-27-76k 9-21-76
10-5-76
Season Total
Percentage Surface Water
Tiba&co Peppcrs-FJ. 7
5-S4-76 5-1. -"6
5-18-76
5-20-76 6-3-76
6-17-76 6-20-76
6-76-'f 7-2-76
--1V-6 7-13-76
S-7-7c 8-3-76
S-711-^6 8-18-76
9-3-^6 9-2-76
9-21-"6 9-21-76
10-1-70
Season Total
Total
Water
Applied
23.41
27.73
19.72
6.5i
10.23
11.95
1..82
1.64
116.03
85. ST.
42.00
83.19
32.04
6.16
28.96
22.08
15. 30
229.73
35.97.
27.85
10.60
1 7.25
1-.29
11.79
13.06
20.33
11.71
126.88
75*
6.16
3.20
3.82
4.81
8.24
8.26
5.55
5.61
8.21
7.15
7.21
68.22
54X
9.21
11.21
11.03
16.54
18.98
19.72
12.54
35.37
18.49
6.14
13.19
172.42
Appl ication
Per Hectare
1.735
2.0-.9
.48.
.7M
.883
1 .095
.122
8.590
1.674
3.317
1.278
.243
1.156
.883
.609
9.160
2.374
.883
1.461
1.217
1.004
1.126
1.735
1.004
10.604
1.156
.609
.730
.913
1.557
1.565
1.065
1.065
1.552
1.339
1.369
12.920
.560
.680
.669
1.004
1.156
1.187
.761
2.156
1.136
.375
.801
10.485
Evapo trans pi rat ion
Source of (ET)
Water for time period
S
S
S
S
S
S
c
r,
5.782K
S
sw;
S
c
G
5.7828
S
S
S
S
S
G
G
7.8033
S
S
C
S
G
G
C
C
S
S
S(551)
7.803*
S
S
G
S
G
S
S
G
S
S
S
7.8038
Percentage Surface Water 62Tt
Seasonal Field Irrigation Efficiency 741
3 • Surface Water
C - Groundwater
No irrigation recomended
Received tail water froa Field 3
*Field #6 irrigated at the sane time
Gregory, E. J. and Eldon G. Hanson,
as Field 5, uith the
Predicting Constmptiv
Research Institute, Report No. 066, April 1976
Seasonal estimate based on ninimia value measured in San
Bulletin 113-3, Department of Water Resources, California
Harvesting started 9-18 and continued through 10-8, with
Irrigated Cotton Field 4-Uest at same time
JBlaney, Harry F. and Eldon C. Hanson, Consumptive Use and
same amount and ET
e Use with CliAstoloftical Data. Hew Mexico Water Resource
Joaquin Valley, California, Vegatative Water Use,
irrigation between first and second picking
Water Requirements in Hew Mexico. Hew Mexico State
Engineer, Santa Fe, New Mexico, Technical Report 32, 196S
kOther areas irrigated
272
-------
MODERN TECHNOLOGY APPLICATION - NEW MEXICO
The last column in Table 2 is the average
daily applied water m^/ha over the total
acreage. The applied water represented a low
application rate compared to the normal con-
sumtive use of a crop. However, if the daily rate
is divided by the ratio of the projected area of the
trees to the total area (48%), then the daily rate
approaches the daily consumptive rate which is
responsible for a small pecan tree (3). The daily
irrigation rate will be increased this next year to
adjust for the tree growth.
One of the problems associated with the
successful operation of a drip-irrigation system
is to have an adequate filter system. The
original screen filter system installed was in-
adequate, passing sand particles through it,
and was subsequently replaced by a sand filter
system. Another problem associated with the
drip-irrigation system is the possibility of an
accumulation of salts. As can be observed in
Table 3, there were more salts in the 5th and 6th
foot depths beneath the pecan orchard than
beneath the alfalfa and lettuce fields. Because
salinity analysis was not done before the initia-
tion of the drip-irrigation system, it i3 not
possible to determine that the salts are the result
of the irrigation system. However, salt buildup
will be followed carefully in the future. This
should be a standard procedure with the initia-
tion of a drip-irrigation system onto a field until
experience is acquired about the operation of the
system and its affect on salt buildup. As long as
the high salt level remains at the bottom of the
root system, no damage is expected to the trees.
With the high irrigation efficiencies made possi-
ble with the drip system, it is expected that the
salts will remain captured in the lower depths
and not return to the groundwater system
through return flow.
The flow rates in the La Mesa Drain are
presented in Figure 2 along with the measured
salinity and the groundwater level. As may be
observed, the flow rates in the drain are cyclic,
increasing when irrigation practices start in
March and decreasing at the end of the irriga-
tion season in October with a very short lag
response time (approximately 15 days). There is
a negative correlation between water quality in
the drain and flow rates. The flow rate is
presented only for the drain at the input to the
farm area, because it was impossible to detect,
within the degree of measurement accuracy,
any increase in the flow as the drain passed
through the farm. Consequently, the measured
values in the drain are representative of a
TABLE 2
Irrigation water applied on pecans by trickle
irrigation system by weekly irrigation periods,
EPA Demonstration Farm, 1976.
Irri-
gation Applied
Period* Water
(m
5-12/5-19
5-20/5-27
5-30/6-4
6-11/6-18
6-21/7-1
7-2/7-10
7-11/7-17
7-22/7-29
7-30/8-5
8-6/8-27
8-30/9-7
9-8/9-14
9-22/10-1
Total
Vtree)
.47
.48
.35
1.46
.28
.13
.43
.20
.82
.38
.28
.17
1.06
Average Daily
Applied Water
(mVday/tree)
.071
J369
.058
.209
.028
.053
.061
.030
.118
.018
.030
.025
.089
Total Total
Applied Average Daily
Water Applied Water
(mVha)
.115
.117
.087
.361
.069
.105
.105
.059
.202
.509
.143
.040
.261
2.170
(mVha)
1.66
1.47
1.44
4.47
0.63
1.16
1.49
.74
2.88
2.30
1.77
.61
2.17
*Divided into weekly irrigation periods; trickle irrigation sys-
tem operated on continuous basis.
Note: Seasonal rainfall from March 1 to October 31 recorded
at the New Mexico State University Plant Science
Farm was 22.19 cm.
326 pecan trees on 1.3 ha (Field #1). Pan evaporation
5-12 to 10-1, 1976 equal to 119 cm, applied water/pan
evaporation = 0.18.
regional response and not the local response of
the farm.
It is interesting to note that although the
flows in the drain increased substantially dur-
ing the summer months, the groundwater table,
as measured by the piezometer located 426
meters from the drain, remained almost con-
stant.
The demonstration project is concerned
with presenting the results on the Demonstra-
tion Farm to the general public.
The extension personnel have been in-
volved in conveying the information gained
from the farm study to interested groups
through media releases, tour brochures, and
field day presentations.
CONCLUSIONS
Return flow to the river system through
deep drainage can be decreased by increasing
irrigation efficiencies. One of the side effects
273
-------
CASE STUDY: MESILLA VALLEY
TABLE 3
Soil analyses for salt content for selected samples,
EPA Demonstration Farm, January 6, 1977.
Plant
Type
Alfalfa
Lettuce
Pecan
samples
located
at tree
trunk
Pecan
samples
located
.5 n
from
trunk
of tree
Pecan
samples
located
1 ffi from
trunk
ct tree
Pecan
samples
located
2.7 m
from
trunk
of tree
Concentration*
Sample
(cm)
0.0- 7.5
7.5- 15.0
30.0- 37.5
60.0- 67.5
90.0- 97.5
120.0-127.5
150.0-157.5
175.0-182.0
0.0- 7.5
7.5- 15.0
30.0- 37.5
60.0- 67.5
90.0- 97. 5
120.0-127.5
150.0-157.5
175.0-182.0
0.0- 7.5
7.5- 15.0
30.0- 37.5
60.0- 67.5
90.0- 97.5
120.0-127.5
150.0-157.5
175.0-182.0
0.0- 7.5
7.5- 15.0
30.0- 37.5
60.0- 67.5
90.0- 97.5
120.0-127.5
150.0-157.5
175.0-162.0
0.0- 7.5
7.5- 15.0
30.0- 37.5
60.0- 67.5
90.0- 97.5
120.0-127.5
150.0-157.5
175.0-162.0
0.0- 7.5
7.5- 15.0
30.0- 37.5
60.0- 67.5
90.0- 97.5
120.0-127.5
150.0-157.5
175.0-182.0
E.G.
(mmhos)
.92
.89
.82
1.22
2.31
1.03
.74
.54
1.58
1.61
1.19
.86
1.02
.61
. 74
.97
1.21
1.11
1.25
1.81
4.00
7.58
5.55
5.33
1.27
1.09
1.53
4.25
4.41
6.40
5.64
7.17
1.11
4.34
1.40
2.58
1.76
5.33
6.72
1.48
2.11
2.20
3.71
4.91
4.44
5.06
3.63
1.41
pH
Saturation
Ca
Mg
Ma
K
Cl
CO, HC03
SO,
(%) (meq/L)
7.73
7.77
7.85
8.03
8.03
8.20
8.35
8.54
8.08
7.94
8.15
8.29
8.48
8.35
8.32
8.28
7.88
7.85
8.01
8.01
7.83
7.99
8.26
8.26
8.02
8.08
8.06
7.88
7.85
7.95
8.36
8.21
8.08
7.91
8.12
7.99
8.24
7.98
8.21
8.58
8.03
8.14
8.00
7.84
7.97
8.16
8.32
8.45
32.5
29.8
31.1
28.9
25.6
23.8
22.0
25.8
61.3
56.2
40.4
31.9
30.1
33.3
28.8
24.2
50.0
50.1
66.2
50.4
54.8
50.5
34.5
31.3
46.1
50.8
71.7
85.0
52.3
46.4
35.1
33.0
49.0
80.1
53.7
85.0
30.5
35.1
31.9
23.1
49.4
53.6
54.5
80.9
29.8
36.8
28.0
25.5
4.24
4.13
3.16
3.95
3.76
2.18
1.59
.68
4.28
4.60
2.76
1.81
2.24
.53
1.91
1.22
1.95
2.01
1.31
4.13
21.51
14.43
7.69
5.43
2.94
3.20
1.42
22.14
18.38
15.38
6.63
5.93
3.88
17.77
1.57
5.07
4.92
16.50
11.89
.53
6.17
6.15
12.73
20.47
20.33
10.82
4.70
1.16
1.48
1.42
1.24
1.41
2.92
.99
.59
.38
2.31
2.85
1.98
1.47
1.54
.75
.72
1.22
1.68
1.64
1.26
1.46
8.20
15.14
7.39
5.87
1.56
1.13
1.17
9.12
9.91
12.52
6.93
7.97
1.27
9.81
1.19
3.19
1.77
8.28
10.40
.86
2.43
2.41
5.10
10.37
9.75
8.13
4.32
1.08
3.84
3.95
4.17
9.22
18.28
8.02
6.24
4.59
7.87
7.74
5.97
4.75
5.28
4.75
5.84
7.39
8.30
7.94
10.05
14.91
25.64
54.90
53.62
55.54
8.86
7.58
13.45
28.83
29.68
54.26
55.76
74.04
7.70
29.68
11.15
21.37
13.78
40.06
60.37
13.68
14.57
16.01
30.06
34.44
29.12
43.70
34.25
12.68
.51
.21
.21
-.26
.39
.19
.14
.14
.58
.65
.45
.32
.31
.20
.22
.20
1.07
.94
.88
.05
.60
.79
.45
.47
.91
.75
.36
.50
.45
.54
.43
.50
.72
.43
.63
.51
.24
.69
.59
.16
.87
.86
.83
.72
.45
.47
.30
.18
1.89
1.70
1.86
2.96
6.76
2.42
1.98
1.62
4.91
4.91
3.29
2.13
2.48
1.20
2.13
2.15
3.12
2.76
4.58
6.67
5.08
5.62
19.11
18.23
4.29
3.48
4.16
7.07
11.34
23.16
20.54
28.61
3.86
12.87
4.75
8.35
5.66
21.52
26.90
3.85
7.32
7.69
14.80
15.31
10.93
21.38
11.29
2.78
0 4.04
0 3.50
0 3.02
0 3.26
0 3.18
0 1.48
0 2.08
0 1.68
0 2.18
0 3.08
0 2.40
0 1.54
0 1.54
0 1.98
0 1.78
0 2.46
0 5.50
0 5.64
0 2.58
0 3.38
0 .84
0 1.46
0 1.46
0 1.62
0 4.54
0 3.34
0 2.16
0 1.20
0 1.32
0 1.32
0 1.48
0 1.50
0 3.48
0 1.26
0 3.18
0 1.72
0 2.06
0 13.6
0 1.36
0 2.22
0 2.86
0 2.26
0 1.46
0 1.28
0 1.22
0 1.18
0 1.80
0 1.50
4.16
4.44
4.36
8.00
15.40
7.20
3.96
2.52
7.60
7.70
5.10
3.60
5.00
2.76
4.70
5.08
5.00
4.70
6.10
10.60
51.60
73.60
48.80
44.40
5.40
5.40
9.90
51.60
48.80
60.80
47.60
60.00
5.80
46.00
7.10
20.20
13.20
44.00
57.20
9.30
14.20
15.60
37.60
48.00
46.00
38.00
30.80
10.50
*Based on analysis of saturation extract samples at indicated saturation percent.
from increased irrigation efficiency is a
decrease in the amount of fertilizer that is
needed to be applied to the fields because less of
the applied fertilizer will be leached out of the
root zone. In most cases during 1976, the
recommendations given to the farmer by the
irrigation scheduling service were followed,
resulting in a reasonable farm-irrigation ef-
ficiency and consequently a minimum return
flow rate. However, the farm efficiency can be
improved with closer cooperation between the
farmers' irrigation dates and the irrigation
dates recommended by the irrigation schedul-
ing service. The feasibility of using irrigation
scheduling techniques is supported by the
response on the Demonstration Farm and by
the fact that the scheduling service is commer-
cially available and is being used by other
farmers in the valley. The use of a trickle-
irrigation system to supply the consumptive use
of a pecan orchard resulted in a substantial
water saving and consequently a reduced return
274
-------
i:
1.0
E .5
10 20 30 10 20 30 10 20 30 10 20 10 10 20 30 10 20 30 10 20 30 10 20 30 10 20 10 10 20 30 10 20 30 10 20 30
lanu.iry Kehruarv March April May lime .Tuly August September November December
Depth
10 20 30 10 20 30 10 20 30 10 20 30 10 .'0 U) 10 20 30 ] o 20 30 10 20 30 in ,'n 10 10 20 30 10 20 30 10 20 30
•'•'""'""' February M.n-h April M;iy Juno "ily August September October November Derumbi-r
Figure 2. Flow rates and water quality in the La Mesa Drain along with the associated ground-water depth fluctuations.
1
:•:
'
;.:
'•'
i
(
:
• i
i
-
:
-------
CASE STUDY: MESILLA VALLEY
flow. However, the salts problem associated
with the pecan orchard needs further investiga-
tion and monitoring.
In the Mesilla Valley, the response time of
the drain system to changes in deep drainage is
very rapid. Consequently, any farm practices
over a large area that affects return flow quanti-
ty and quality will immediately be observed in
the response of the system.
REFERENCES
1. Jensen, M. E. and H. R. Haise. 1963.
Estimating evapotranspiration from solar
radiation. J. Irrig. and Drain. Div., Am. Soc.
Civ. Engr., 89:15-41.
2. Rio Grande Project, New Mexico-Texas,
Annual Report. 1973. Irrigation Management
Services. U. S. Department of the Interior,
Bureau of Reclamation, December 1974.
3. Rio Grande Project, New Mexico-Texas,
Annual Report. 1972. Irrigation Management
Services. U. S. Department of the Interior,
Bureau of Reclamation, December 1973.
276
-------
ECONOMICS
OF CONTROLLING IRRIGATION
RETURN FLOW IN
THE MESILLA VALLEY,
NEW MEXICO
ROBERT R. LANSFORD, LYNN W. GELHAR and BOBBY J. CREEL
New Mexico State University; New Mexico Institute of Mining and Technology;
New Mexico State University, Las Cruces, New Mexico
ABSTRACT
Preliminary budgeting indicates that net
returns of water are reduced by incorporation of
irrigation water management practices, but
that irrigation water requirements may be
reduced. Thus, incorporation of these improved
irrigation water management practices may be
feasible under restricted water availability con-
straints. If irrigation water use can be reduced
without affecting the local agricultural
economy by optimal crop selection incor-
porating the improved practices, the quantity of
irrigation return flows may be reduced and
possibly the quality may be improved. Further
evaluation in the current and last year of the
three-year project will provide opportunities for
refinements and sensitivity evaluation in the
linear programming model as well as more
extensive iterative analysis with the
hydrological model.
INTRODUCTION
The quality of irrigation return flow
represents a major problem in the western
United States because of the goal of restoring
and maintaining the quality of the nation's
waters as required by the Federal Water Pollu-
tion Control Act (PL 92-500). The water of the
Upper Rio Grande has been reported as a classic
example of water quality degradation.
A U.S. Senate Select Committee (1961)
report estimated that the Upper Rio Grande and
Pecos Basins were the shortest of water in
relation to projected demand to 1980 of any
basin in the continental United States. This
projection was documented by the U.S. Water
Resources Council study (1968) which studied
the complete Rio Grande drainage area in
Colorado, New Mexico, and Texas. The Water
Resources Council study identified the major
problems to be water deficiencies and
ground water storage depletion, poor quality
because of mineral pollution, heavy sediment
loads in many tributaries, excessive non-
beneficial consumptive use, and frequent floods
causing extensive damage.
The Council further stated that these water
problems must be solved since projected water
withdrawals for the year 2020 are estimated to
be about two and one-half times greater than the
present average runoff in the region.
The Upper Rio Grande from the headwaters
in Colorado to Fort Quitman, Texas (Figure 1)
probably comes closer to consumptively using
all of the surface water of the river than any
other major river basin in the United States.
The only surface waters that escape the region
at Fort Quitman are occasional flood waters
from thunderstorms. Therefore, concentration
of dissolved solids became a major problem in
the southern reach of the basin. Table 1 in-
dicates the variation in composition of the Rio
Grande water from Otowi Bridge near Santa Fe
to El Paso, Texas. It shows a progressive
increase in the concentration of total dissolved
solids and percent sodium from the upper to the
lower sampling stations. The relatively large
increase in dissolved solids in the river along
277
-------
CASE STUDY — MESILLA VALLEY
TABLE 1
Surface water quality of the Rio Grande at selected gaging
stations. 1967
Site-
Alter- trical
agt DU- Conduc-
Di>- tolued tivity
charge Co Ma Na Cl SO* HCO3 Solid* EC x 1
-------
ECONOMICS OF CONTROL - MESILLA VALLEY
The objective of this paper is to project the
changes in the quantity of irrigation return
flows as a result of alternative water manage-
ment practices that could be adopted by farmers
in the Mesilla Valley. They are (1) irrigation and
management scheduling, (2) trickle irrigation
on tree crops, and (3) sprinkle irrigation on
vegetable crop emergence. The effects of these
practices will be simulated for the approximate-
ly 100,000-acre Mesilla Valley.
THE ANALYTICAL SYSTEM
The analytical system consists of two
specific models sequentially linked to simulate
the agricultural production and hydrological
adjustments that would occur as a result of
implementation of the alternative water
management practices. The first is a linear
programming model to estimate the economic
impact and the irrigation water requirements.
The solution is constrained by the usual
physical, institutional, and market restrictions.
The results of the LP model serve as inputs to
the physical model. The physical model is
partitioned into interdependent submodels that
analyze the hydrology and salinity balances.
The submodels estimate the effects of irrigation
water use on the water table depths and the
quality and quantity of irrigation return flows.
Solutions from the models are iterated a suf-
ficient number of times to simulate the
hydrologic adjustments from a change in water-
use as a result of alternative water management
practices.
The Physical Model
The hydrosalinity model selected for this
study was developed by the U.S. Bureau of
Reclamation (1975) for the Environmental
Protection Agency. This discussion of the
hydrosalinity model is a summary of material
presented at the International Conference on
Managing Saline Water for Irrigation in Lub-
bock, Texas (McLin and Gelhar, 1976). This
model can represent a general water and mass
balance for a river basin, or portion thereof. The
model includes components which simulate
surface and ground water reservoirs, diversions
to meet irrigation needs and other uses, irriga-
tion return flows, and chemical transfor-
mations in the soil. Several nodes can be used to
simulate the conditions in different segments of
a basin. The model would be classified as a
multicell lumped parameter model. Input data
on water use, surface and groundwater quality,
and streamflows and diversions are required to
operate the model.
The USER model represents an irrigated
area by a series of nodes. The number of nodes
depends on the physical features of the study
area, the availability of data, and the number of
points at which the model response is desired
within the area. The model structure is based on
water balance for a given time interval which is
computed and maintained for each node. The
transfer of water between river and aquifer may
be essential to maintain this balance. A soil
column is included to simulate the chemical
exchange between soil and water as water
percolates downward through the soil column.
The model allows the mixing of one water wit!
another and computes the chemical quality oi
the mixed waters in proportion to the volumes
mixed. Thus, a chemical mass balance is
simultaneously maintained. All these com-
putations are performed for one node at a time
and progress from the upstream to the
downstream nodes.
A node includes simulation of several
features as shown in Figure 2. These are:
simulation of magnitude and quality of river
flow, diversion to meet the irrigation demand,
water transfer between river and aquifer, irriga-
tion return flows directly to the aquifer, through
the soil to the aquifer, and to the river. The river
flow at the start of nodal operation is a known
quantity, a portion of which is diverted to meet
the irrigation demand. The irrigation return
flow is then computed by subtracting the crop
consumptive use from the irrigation demand.
The chemical quality of mixed waters is com-
puted at each point. At the end of nodal opera-
tion, the balance of inflow to, and outflow from,
the river is determined. If the river flow is in
excess of that observed, the additional water is
transferred to the aquifer; if the river flow is in
deficit, water is withdrawn from the aquifer to
maintain the balance. In the model operation,
no consideration is given to spatial variability
of hydrologic variables within a node.
Hydraulic properties of soil and aquifer are not
considered.
The input data needed to operate the model
include irrigation demand, crop consumptive
use, diversion, river water quality, initial
aquifer water storage and chemical quality, soil
moisture content and its chemical quality, and
precipitation.
279
-------
CASE STUDY - MESILLA VALLEY
Originally, the computer program for the
model was written for a Control Data Corpora-
tion computing system. To make the program
compatible with an IBM 360-44 system,
numerous changes involving definition of
variables, and format and dimension
statements were required. The revised computer
program is listed in a recent report (Lansford, et
al., 1976) with all of the program changes
identified.
SCHEMATIC FLOW CHART
CONSUMPTIVE
USE
-
DIVERSION
•
IRRIGATED
AREA
RETUR I FLOW
AQUIFER
Figure 2. Schematic flow chart showing the features
simulated in each node of the model.
Simulation for the Mesilla Valley
The objectives of these simulations are to
determine what model structure is suitable for
the Mesilla Valley and to test the sensitivity of
the model results to observed and estimated
input data. The model was implemented with
existing field data for the period June 1967 to
June 1968. A complete set of ground water
quality data was available at the beginning of
this period.
The irrigated areas served by the Leasburg
and Mesilla diversion dams form natural nodes
for the model (Figure 3). The quantity and
chemical quality of flow in these diversions are
observed. Each node includes several
operational features as shown in Figure 4.
Economic Model
The linear programming model derives a
cropping pattern that maximizes returns to
water in each of the nodes, subject to the amount
of surface and groundwater available and the
crop rotation and marketing requirements of
the area. The locations of crops were specified in
the base year, with the location of additional
Figure 3. Location of existing surface data and
sampling points. Node 1 contains approximately
40,800 acres; node 2, approximately 67,800 acres.
280
-------
ECONOMICS OF CONTROL — MESILLA VALLEY
acreages of crops only being constrained by
market characteristics. Water use in the base
year approximated actual water use reported by
the area irrigation districts. Average commodi-
ty prices for 1967-76 are justified by constrain-
ing the LP crop production.
Alternative crop production activities and
coefficients will be developed by utilizing a
submodel budget generator to derive engineer-
ing cost approach crop enterprise budgets. The
base year budgets will be designed to simulate
the cost and returns and input requirements for
typical farming operations in the Mesilla
Valley. Alternative crop production budgets
incorporating irrigation water management
practices will be developed by modification of
the base year budgets, thus providing a series of
levels of irrigation water management practices
in the Mesilla Valley.
RESULTS
Experience with the USBR-EPA simulation
model has demonstrated the difficulty of routine
application of large and complex computer-
programs even when the model is designed to be
general and flexible. This model is actually
computer- and site-specific; several subroutines
must be modified for each application and thus
requires substantial knowledge of the logic of
the entire program.
The preliminary simulation results for the
Mesilla Valley indicate a need to improve the
input data and model structure (McLin and
Gelhar, 1976). Several sensitivity analyses con-
ducted with the model in its preliminary form
showed the effects of consumptive water use,
soil chemistry, and groundwater chemistry. The
scheme originally used in the model to transfer
water between the aquifer and the stream may
be physically unrealistic in that it is indepen-
dent of aquifer properties and water level.
Improvements which relate the stream-aquifer
interaction to aquifer properties and water level
(Gelhar and Wilson, 1974) will be incorporated
after the model has been tested over a long time
frame in its original form. Baseflow recession
and well drawdown data will be used to estimate
the required aquifer parameters for the propos-
ed model improvements.
NODE
NODE 2
N SEQUENCE NUMBERS
-N = DIVERSION XTN
+N - INFLOW V**/
PUMPAGE FROM C. W. TO X7\
SUPPLY DEMAND /*^-X
I
1 NODE 1 AQL'IFER
RIO GRANDE FLOW BE
L
**US
INDI
TION
0
i
2
i
EAS BITRC
E 'CONL
CATE TH
OF RFI
j) C
z
i
s
CANAL
SE* CAf
IS DIS1
URN FLC
) C
5
5
g
«
LOU
!D TO /*7\
W
>. z
11
r
RIO GRANDE AT HEAD OF SYSTEM
RIVER FLOW AFTER DIVERSION
TO ••'3 BELOW) *
IRR. DEMAND SUPPLIED FROM G. W.
IRRIGATION DEMAND IRRIGATED
QP
i IRRIGATION
\^ RETURN FLOW "
NODE 1 WASTEWAYS TO RIVER
«** RET
CS
s
CO
TRANSFER
RIVER TO
€
^?
INFLOW
!RNED TO RIVER /
•R' ON CONl'SE CARD) /
Ij'csv' ON CONI'SE CARDJ 5
DIL t
LUMN -
r ?
^
z:
NODE 1 AQL'IFER |
OF FLOW ^~\
AQUIFER ^Xi/
) INFLOW TO AQUIFER
FROM RIVER
' AQL'IFER TO RIVER
TO RIVER FROM AQUIFER
r\
PREDICTED OUTFLOW OF MODEL
<
(OBSERVED OUTFLOWS) "^ J
HYDROLOGIC BALANCE
0
PUMPAGE FROM G. ^_>-^
JTO SUPPLY DEMANDV^yxTV
IKODE 2 ^~~"^
AQL'IFER y^v
RIO GRANDE FLOW BELOW
MES
**CSE 'CC
INDICATE
TION OF E
C
z
-j-.
1 RIO GRANDE OUTFLOW A'l COURCI
ILLA DAM
N'USE' CARD TO f-*>
THIS DISTRIBU-CO
ETURf; FLOW ^"^
? I
X
1
ill
1^ s
1
1
1
RIO GRANDE
AT MESILLA DAM
HEAD" OF NODE 2 ^-^
RIVER FLOW AFTER DIVERSION ^p
(TO 03 BELOW) g
IRR. DEMAND SUPPLIED FROM G. W. ~
IRRIGATION
0(7\ .:
DEMAND J IRRIGATED | 5]
I AREA
IRRIGATION
\i/ RETURN FLOW ~~
NODE 2 WASTEWAYS TO RIVER l
***; RETURNED TO RIVER t
C'SUR' OS CONUSE CARD) /
**',: TO SOIL COLUMN f a:
^ '
SOIL
COLUMN
J
TRANSFER OF
GWV' ON CONUSE CARD) ±
|
i
u
«
NODE 2 AQLUFER j*^
FLOW ^
RIVER TO AOL'IFER \^J
O INFLOW TO AQUIFER
FROM RIVER
©TRANSFER OF FLOW
AQUIFER TO RIVER
INFLOW TO RIVER FROM AQUIFER — .
PREDICTED OUTFLOW OF MODEL
(OBSERVED OUTFLOW \^/
XCHANGE MECHA1JISH SETS
HYDROLOGIC BALANCE
Figure 4. Mesilla Valley system flow chart
281
-------
CASE STUDY - MESILLA VALLEY
Of the features examined in the sensitivity
analyses for the one-year simulation period, it is
evident that the average initial chemistry of
aquifer waters and the combined effects of crop
consumptive use and irrigation efficiency play
an important role in the predicted TDS output.
Preliminary budgeting indicates that net
returns to water are reduced by incorporation of
irrigation water management practices, but
that irrigation water requirements may be
reduced. Thus, incorporation of these improved
irrigation water management practices may be
feasible under restricted water availability con-
straints. If irrigation water use can be reduced
without affecting the local agricultural
economy by optimal crop selection incor-
porating the improved practices, the quantity of
irrigation return flows may be reduced and
possibly the quality may be improved. Further
evaluation in the current and last year of the
three-year project will provide opportunities for
refinements and sensitivity evaluation in the
linear programming model as well as more
extensive iterative analysis with the
hydrological model.
Additional results are expected and will be
reported during the conference.
ACKNOWLEDGMENT
The work upon which this report is based
was supported in part by funds provided by the
United States Environmental Protection Agen-
cy under Grant number S803565-01-0 through
the New Mexico Water Resources Research
Institute.
The report is published as Journal Article
621, Agricultural Experiment Station, New
Mexico State University, Las Cruces, New Mex-
ico 88003.
REFERENCES
1. Gelhar, L. W., and Wilson, J. L. 1974.
"Groundwater Quality Modeling,"
Groundwater, Vol. 12, No. 6.
2. Lansford, R. R., et al. 1976. Demonstra-
tion of Irrigation Return Salinity Control in the
Upper Rio Grande — Annual Report, Year 1.
New Mexico Water Resources Research In-
stitute, Las Cruces, New Mexico. March, 121 p.
3. McLin, S. G., and Gelhar, L. W. 1976.
"Modeling of Irrigation Return Flow in the
Mesilla Valley, New Mexico." Paper presented
to the International Conference on Managing
Saline Water for Irrigation. Lubbock, Texas.
4. U.S. Bureau of Reclamation. 1975.
Prediction of Mineral Quality of Irrigation
Return Flow, Vol. 1. Engineering and Research
Center. Denver, Colorado, 58 p.
5. U.S. Senate, Senate Committee on
National Water Resources. 1961. Water
Resource Activities in the United States.
Eighty-sixth Congress, Government Printing
Office, Washington.
6. U.S. Water Resources Council. 1968.
The Nation's Water Resources, The First
National Assessment. Government Printing Of-
fice, Washington.
282
-------
Agricultural Drainage Problems
of the San Joaquin Valley
E. P. PRICE
U.S. Bureau of Reclamation,
Sacramento, California
ABSTRACT
A brief description of the physical setting of
the San Joaquin Valley and the drainage
problems anticipated are presented. Estimates
are preliminary and subject to revision as
studies continue.
INTRODUCTION
This paper, in combination with the three
which follow it, describes the present and im-
pending agricultural drainage problems in the
San Joaquin Valley of California and the efforts
of three agencies to cope jointly and comprehen-
sively with these problems. The three agencies
are the U.S. Bureau of Reclamation, the Califor-
nia Department of Water Resources, and the
California State Water Resources Control
Board. They have undertaken what has been
named the San Joaquin Valley Interagency
Drainage Program — or the IDP, for short.
One purpose of the IDP is to coordinate
studies of agricultural drainage water and salt
management which the three agencies engage
in independently. This coordination should in-
sure that the funds which the three agencies
have for studying drainage problems will be
spread efficiently over a broad range of work
without duplication of effort or omission of
essential components. A second purpose is to
develop cooperatively an economically feasible,
environmentally sound, and politically accept-
able plan for managing the agricultural
drainage problems of the San Joaquin Valley
and for disposing of the salts which are, after
all, the fundamental problem.
The Physical Setting
The San Joaquin Valley is the southern half
of the great Central Valley of California (see
Figure 1). The San Joaquin Valley is bounded
on the east by the Sierra Nevada Mountains, on
the south by the Tehachapi Mountains, on the
east by the Coast Range Mountains, and on the
north by the Delta formed by the Sacramento
and San Joaquin Rivers. The floor of the valley
is approximately 200 miles long and about 40
miles wide on the average. The floor slopes
upward from sea level at the Sacramento-San
This paper presents a brief overview of the
problem. The three papers which follow will
give more details about the problem, the local
partial solutions, and some of the possible
comprehensive solutions under consideration. Figure 1. Location map
283
°<«- ~"*-SsSg'S T E H A C H A PI MOUNTAINS-^! »\
^P^^^V^fX
V
-------
CASE STUDY: SAN JOAQUIN VALLEY
Joaquin Delta to a rim elevation of from 500 to
700 feet above sea level on the three moun-
tainous sides. An imperceptible drainage divide
crosses the San Joaquin Valley from east to
west near its midpoint. North of this divide the
surface waters drain to the Pacific Ocean via the
San Joaquin River, the Sacramento-San Joa-
quin Delta, and San Francisco Bay. The
southern half, known as Tulare Lake Basin, is
an interior drainage; only in years of heavier
runoff does surface water escape northward to
the San Joaquin River and thence to the ocean.
For purposes of this discussion of drainage
problems, it is convenient to talk of the San
Joaquin Valley in six components rather than
to treat the valley as a whole.
One component is the eastern side of the
San Joaquin Valley north and east of the San
Joaquin River. This is called Eastside North
(see Figure 2). Here the soils are moderately or
highly permeable. Nearly all of this area has
been developed for irrigation using the high
quality water which comes from the snowpack
of the Sierra Nevada. The combination of
permeable soils and high quality water makes it
unlikely that this area will ever have drainage
problems. Should high water tables develop,
most of the water could be reused.
The second area has been named the East-
side South. It lies south of the San Joaquin River
bounded on the east by the Sierra Nevada
foothills, on the south by the northern boundary
of Kern County, and on the west by the Tulare
Lake area. As with the Eastside North, most of
the Eastside South soils are permeable or
moderately permeable and it has a supply of
high quality irrigation water from the Sierra
Nevada runoff. It does not have enough of this
high quality water, however, to irrigate all of the
irrigable land. Consequently, at some future
time, water will have to be imported to overcome
the ground water overdraft that now exists in
the area. The imported water may contain
moderate amounts (300-800 milligrams liter) of
salt and ultimately will cause a drainage
problem. The problem most likely will manifest
itself on the west side of the component area
near the valley trough.
The third component area is the service
area of the Delta-Mendota Canal. It lies west of
the San Joaquin River and extends from almost
the Sacramento-San Joaquin Delta to the San
Luis service area. These soils are appreciably
Figure 2. Designated drainage areas in San Joa-
quin Valley
tighter than those of the east side and a large
portion of the water supply is imported with a
salinity of 300-800 mg/1. The remainder of the
water supply is either ground water or local west
side runoff, both of which are as saline, or
perhaps more saline, than the imported supply.
Irrigation development in this area began more
than 100 years ago using water from the San
Joaquin River. With the completion of the Delta-
Mendota Canal 25 years ago, the San Joaquin
River water was replaced by the imported water
and the area of irrigation was expanded. A few
localized drainage problems have developed.
South of the Delta-Mendota Canal service
area lies the San Luis service area. Here again
the soils are tighter than those on the east side,
in some cases being only slowly permeable. The
area is interspersed with impervious strata
which result in numerous perched water tables.
The drainage picture is further complicated by
the fact that the water supply is either imported
water or water pumped from the underground.
In both cases there is an appreciable saline
284
-------
DRAINAGE PROBLEMS
content. Drainage problems have been slowly
increasing in this area with the growth of
irrigation. Even before irrigation development,
however, natural conditions produced pockets
of highly saline shallow ground water along the
trough line of the valley. In some cases, these
perched water bodies have salinities of 100,000
mg/1.
The fifth component area is the Tulare Lake
area. It is the bed of what under natural
conditions in recent geologic times was an
intermittent shallow lake or marsh. The instiga-
tion of upstream irrigation developments,
which reduced the inflow, and the construction
of a series of diked ponds in which to manage
the runoff which did reach the old lake, have
permitted the reclamation and agricultural
development of the area. The salts which have
accumulated in the soils over the centuries and
the tightness of the lacustrine soils have com-
bined to produce one of the most serious and
immediate drainage problems in the San Joa-
quin Valley.
The sixth component area is the Kern
County portion of the valley floor. In many
respects it is similar to the Tulare Lake area.
Unlike the Tulare Lake area, however, the Kern
County area had only a severely limited water
supply until the late 1960's. Now it receives
supplemental imported water from two sources.
It can be anticipated that drainage problems
will develop rapidly in the Kern County area.
The Ultimate Magnitude of the Drainage
Problem
The San Joaquin Valley Interagency
Drainage Program has estimated the annual
quantities which may be captured in a master
drainage facility. This of course, will not be all
the bad quality water that will result from the
irrigation and the concentration of salts by the
plants in the soil. Some of the bad quality water
will escape past tile drains into the ground
water and will show up elsewhere. How much is
captured for management and disposal depends
upon the efficiency of water application by the
irrigator and the efficiency of the tile drainage
system.
On the Delta-Mendota service area, we
anticipate that ultimately about a million and a
half acre-feet of water will be applied annually
for irrigation. We assumed a basin efficiency of
87 percent, and this means that 196,000 acre-feet
of waste water will be generated every year. If
we can capture 50 percent of this, we will have
about 100,000 acre-feet of recoverable waste to
our drain from the Delta-Mendota service area.
In the San Luis service area, we anticipate
an ultimate annual application of about 1,760,-
000 acre-feet. The efficiency of use within the
area will range from 86 percent to 94 percent and
about 216,000 acre-feet of poor quality water will
be generated each year. We anticipate recover-
ing approximately 70 percent of this, or 150,000
acre-feet annually.
In the Eastside South area, ultimately
about 4,900,000 acre-feet of water will be applied
at efficiencies ranging from 85 percent to 94
percent. This will produce 380,000 acre-feet of
waste water to be disposed of, and we expect
that from 50 percent to 70 percent of it will be
recovered in a master drainage facility, or about
220,000 acre-feet.
In the Kern County area, ultimately 2,980,-
000 acre-feet of water will be applied with
efficiencies ranging from 92 percent to 96 per-
cent. Thus, 205,000 acre-feet of waste water will
be produced untimately each year and we will be
able to capture somewhere between 65 percent
and 75 percent of it for a total of 140,000 acre-feet
of poor quality water discharged to the drain
each year.
For the Eastside North area, we anticipate
that 860,000 acre-feet a year will be applied with
an efficiency of 85 percent. This means 129,000
acre-feet of drainage water will be generated,
but only about 20 percent or 25 percent of it, or
about 30,000 acre-feet, will be of such poor
quality as to require disposal via a master drain.
We expect the Tulare Lake area to generate
ultimately about 200,000 acre-feet of poor quali-
ty drainage water per year. Three-fourths of this
will be disposed of in local drainage ponds, but
the remaining 50,000 acre-feet probably will
have to be disposed of via a master drainage
facility. These estimates are based on obser-
vations of existing practice in the area rather
than by estimating application and drainage
efficiencies. The volume of water applied varies
so much from year to year in this area (due to
lack of upstream regulatory storage) that any
implication of a fairly constant irrigation ef-
ficiency would be misleading.
The Buildup of the Drainage Quantities
Estimating the rate at which drainage
flows will build-up in a master drain for an area
285
-------
CASE STUDY: SAN JOAQUIN VALLEY
such as the San Joaquin Valley undoubtedly is
much more of an art than a science. The first
question to be answered is when will the water
supply development be complete. At the
present time the Eastside North area, the Delta-
Mendota service area, and the Tulare Lake
Basin area have adequate supplies. The San
Luis service area is in the process of building up
its annual use of supplemental water as is the
Kern County area. The Eastside South area,
however, is seriously deficient in water supply
and there still is great uncertainty about what
sort of water development project will be built to
supply this area.
After the question of the rate of buildup of
water supply comes a question of the action of
the shallow ground water table. For the San
Joaquin Valley Interagency Drainage
Program, we have assumed that areas which
develop a shallow ground water table within 5
feet or less of the surface will have to have
drainage provided. There are some areas in the
San Joaquin Valley, however, that have been
operating for years with full irrigation and a
shallow ground water table within 5 feet of the
surface. The shallow ground water shows no
sign of rising, and no drainage is required.
Consequently, we must recognize the fact that
very low rates of permeability which are dif-
ficult to measure accurately may nevertheless
be adequate to carry away the leaching water
from an efficient farm operation.
Finally, of course, in order to know how the
drainage quantities will build up with time we
must predict how the individual farmers will
install the tile drains. This is a matter of
individual economic judgment on the part of
each farmer and one which is very difficult to
foresee. The timing will be further complicated
by the related factors of financial assistance
which may be made available to farmers for in-
stalling tile drains, and the availability of the
drains themselves, and the people to install
them. Experience has shown that in the
Coachella Valley in southern California at
least, the latter factor may be the limiting one as
far as the rate of installation of the tile drain
systems is concerned. At any rate, we have done
our best in these three areas of uncertainty and
have come up with estimated rates of growth of
the drainage qualities, decade by decade,
through the year 2080 (see Table 1).
TABLE 1
Estimated annual quantities of subsurface agricultural drainage water in the San Joaquin Valley.
1980
2000
2020
Subarea
Delta-Mendota
Canal service
area
San Luis service
1000 1000
A.F. Acres
1000 1000
A.F. Acres
1000 1000
A.F. Acres
2040
1000 1000
A.F. Acres
2060
2080
1000 1000
A.F. Acres
1000 1000
A.F. Acres
35
40
50
60
60
75
80
100
100
125
100
125
area
Tulare Lake1
Tulare Lake-
Eastside South
Kern County
Eastside North1
10
5
(20)
0
0
0
20 70
9r-! 20
(30)^ (60)
0 50
0 40
0~ 5
140
27
(80)
125
55
6
110
30
(100)
120
80
10
220
40
(150
300
115
12
120 240 130 260 150 300
40 60 50 75 50 75
(135) (200) (150) (215) (150) (215)
180 450 200 500 220 560
120 175 130 190 140 205
20 24 30 35 30 35
Eastside North3 (100) (120)-' (100) (120) (100) (120) (100) (120) (100) (120) (100) (120)
Total 50 68 235 413 410 762 560 1,049 640 1,185 690 1,300
via a valleywide drain system.
^Retained drainage water to be disposed of in local evaporation ponds — not included in totals for valleywide
system.
3Leaves subarea via San Joaquin River, ground-water flow to west side, evaporation and consumptive use on non tilled
areat.
286
-------
DRAINAGE PROBLEMS
Quality of the Drainage Effluent
The quality of the drainage effluent is an
all-important factor in deciding what kind of
disposal management will be required. The
estimates of amounts of water applied, the
quality of the applied water, and the efficiency
of application lead us directly to the quality of
the effluent. There is, however, one somewhat
empirical factor to be taken into account. This is
the existence of the pockets of the highly saline
perched ground water along the trough of the
San Joaquin Valley. Because of the presence of
these perched water tables, the estimated
overall quality of the drainage effluent begins at
a poorer quality and decreases as time
progresses, finally leveling off.
The valley-wide average quality of the
agricultural drainage to be disposed of from the
San Joaquin Valley through a master drainage
facility is expected to vary as follows:
Quality
Electrical
Conductivity
micromhos/cm
7,500
5,500
5,000
5,000
4,500
4,300
Total
Dissolved
Solids
mg - 1
5,200
3,800
3,500
3,500
3,200
3,000
Year
1980
2000
2020
2040
2060
2080
CONCLUSION
This completes a brief description of the
physical setting of the San Joaquin Valley and
of the drainage problems which we anticipate
there. Needless to say, these estimates I have
given you are preliminary and subject to revi-
sion as our studies continue.
287
-------
How the NPDES Program
Will Define Present Water
Quality Conditions
GENE MERRILL
Central Valley Regional Water Quality
Control Board,
Sacramento, California
ABSTRACT
Monitoring of surface irrigation supplies
and surface return flows is being carried on
through the NPDES permit program under
Public Law 92-500. This program encompasses
approximately 1.3 million acres of land in the
SanJoaquin Valley of California alone. The EC
and suspended solids of the irrigation returns
frequently exceed basin water quality control
plan objectives (standards) adopted in 1971 by
the Central Valley Board for the receiving
water.
The NPDES irrigation return flow monitor-
ing program is administered in California
through the State Water Resources Control
Board and nine regional water quality control
boards. The largest regional board is the Cen-
tral Valley Regional Water Quality Control
Board which covers 6.8 million of the State's 8.8
million acres of irrigated land. The San Joaquin
Valley lies in the southern portion of the Central
Valley region and contains approximately
4,500,000 acres of irrigated land.
Water quality standards, called objectives,
have been set for some surface streams by the
Central Valley Board. An example is the objec-
tive of 500 ppm salinity (total dissolved solids)
at Vernalis on the San Joaquin River. Any
discharge to the river, including agricultural
returns, can be restricted by the Regional Board
to meet these objectives and keep river salinity
below 500 ppm. This course of action has not
been pursued, lacking concomitant alternative
actions such as development of a dilution water
supply and a saline water drainage facility.
Another water quality objective for the San
Joaquin River allows discharges to increase
turbidity a small amount over background
levels — usually a 10 percent increase. Enforce-
ment of turbidity objectives has been delayed
for most agricultural discharges pending
development of sufficient data and selection of
appropriate available technology.
When faced in 1974 with the issue of ad-
ministering a permit program for irrigation
returns, California sought advice from a State
Agricultural Water Quality Advisory Com-
mittee. The committee, composed of irrigation
and water leaders and technical persons,
recommended:
1. Regional Boards issue permits to water
purveying and draining entities, be they
districts or individual farms.
2. Allow grouping of entities into single
permits around hydrographic areas to
facilitate sampling and permit issuance.
3. Initially the permittees to monitor three
items — quantity of flow, both supply
and discharge; salinity (electrical con-
ductivity); and suspended solids
(primarily sediments).
At this time 22 NPDES permits covering 2.9
million acres of irrigated lands have been issued
in the Central Valley Region (Sacramento and
San Joaquin Valley). Monitoring has been
carried on for up to two years by some per-
mittees. Table 1 shows permits issued in the San
Joaquin Valley area. These do not include
3,100,000 acres of land in the south half of the
San Joaquin Valley (Tulare Lake Basin), as in
this more arid location return flows to surface
streams are very infrequent. A monitoring
program is being devised for this area and will
be carried out under section 208 of PL 92-500
rather than the permit section.
289
-------
CASE STUDY - SAN JOAQUIN VAT .LEY
San Joaquin River irrigation return flows
constitute, in normal water years, a significant
portion of the irrigation season flow of the San
Joaquin River. In the permit areas, surface
irrigation returns and subsurface irrigation
returns are usually co-mingled. Thus, the
monitoring sites, which are mostly located near
discharges to the river, represent a combination
of surface and subsurface runoff or drainage.
Table 2 shows preliminary estimates and
measurements of quantities of water supplied
through canal conveyances for four permittees.
Supply locations were selected by sites that
represented at least 80 percent of the total
surface water supply to a permit area.
Groundwater sources are not reported although
some existing data is available. If it is assumed
that typical applied irrigation water is 3-1/2
acre feet per irrigated acre in the permit areas
indicated, then the percentage of the supplies
monitored would be as indicated in the table.
The table also shows preliminary estimates
and measurements of quantities of return flows.
The percentage of total returns which these are
thought to represent are indicated in the table.
Representative supply and surface return flow
sites described above were selected with
assistance from irrigation district staff, con-
sultants, representatives of the Soil Conserva-
tion Service, Agricultural Extension Service
and others.
The Central Valley Board looks at the
salinity issue as a problem requiring a basin-
wide solution: i.e., the Board looks to the Inter-
Agency Drainage Program, the Basin Plan and
other sources for solutions. Typical electrical
conductivities (EC) of supply and discharge
water are shown and compared in Table 3. It can
be seen that a relatively high discharge EC of
"westside" permit areas indicates a need for
drainage facilities to separate and remove
saline drainage waters from the return flows.
The higher EC of "westside" permit areas is
considered to be largely due to relatively higher
EC of supply waters in that area.
Table 3 also shows typical suspended solids
concentrations in supply and discharge waters.
Field observations indicate that high sediment
loads in the discharges are mainly due to
irrigation conditions. More detailed studies
made by the University of California, the Soil
Conservation Service, and the Stanislaus Coun-
ty Public Works Department confirm that the
sediment loads are from irrigation surface
returns. These organizations are working with
the local Resource Conservation District, the
Central Valley Board staff and others to iden-
tify and implement solutions to the sediment
problem.
During the process of selecting monitoring
points, persons with local irrigation and
agricultural expertise were invited to informal
discussions and were asked to make
suggestions. These discussions were useful in
that monitoring and research activities by other
organizations and agencies were found to com-
plement the NPDES monitoring program. In-
itially it was thought that more internal
monitoring would be included in the NPDES
permits. However, as a result of these informal
discussions, some of the research and in-
vestigations of others were directed to meet
much of the initial internal monitoring needs.
In addition to the parameters of flow, salini-
ty, and suspended solids, the Central Valley
Board is cooperating with other organizations
such as the California Department of Fish and
Game to look into possible problems of
agricultural chemicals in irrigation flows.
The NPDES permit program has not been
accepted without reservations by a wary
agricultural community. There has been,
however, some recognition in the value of par-
ticipating in the program and some monitoring
findings have been made which were not an-
ticipated. For example, it was found that there is
a substantial daily variation in the EC of the
water supplied through the USER Delta-
Mendota Canal to the Mendota-Crows Landing
area.
SUMMARY
Monitoring of surface irrigation supplies
and surface return flows is being carried on
through the NPDES permit program under
Public Law 92-500. This program, administered
by the State of California, encompasses ap-
proximately 1.3 million acres of land in the San
Joaquin Valley of California alone. Sampling
locations are usually located near the point of
discharge to surface waters. The EC and
suspended solids of the irrigation returns fre-
quently exceeds basin water quality control
plan objectives (standards) adopted in 1971 by
the Central Valley Board for the receiving
water, in this case the San Joaquin River. The
solution to the high EC problem lies in basin-
wide management of saline waters including
290
-------
large scale drainage conveyances where ap-
propriate. The solution to the high suspended
solids is being sought through investigations
NPDES PROGRAM — SAN JOAQUIN
being conducted by local Resource Conserva-
tion Districts, Soil Conservation Service, and
Agricultural Extension Service personnel.
TABLE 1
NPDES Irrigation Return Flow Monitoring in the San Joaquin Valley
Monitoring
Permittees
Central Delta Water Agency
South Delta Water Agency
Patterson Resource Conservation Dist.
Mendota-Crows Landing Return
Flow Group
Stanislaus-Tuolumne Rivers Water
Quality Committee
Byron-Bethany Irrigation Dist.
Turlock Irrigation Dist.
Merced Irrigation Dist.
Madera County Flood & Water
Conservation Dist.
Columbia Canal & Newhall
Land & Farming Co.
El Nido Irrigation Dist.
TOTAL
Irrigated
Acres
120,000
102,000
50,000
400,000
200,000
11,000
170,000
110,000
150,000
21,000
9,000
1,343,000
Date
Permit
Issued
8/27/76
8/27/76
5/28/76
7/25/75
8/22/75
8/27/76
8/22/75
8/22/75
3/26/76
8/22/75
8/27/76
Date
Monitoring
Began
—
—
6/76
8/75
9/75
9/76
9/75
9/75
7/76
8/75
—
Number
Supply
Points
—
—
3
2
2
1
1
1
3
—
1
14
Number
Discharge
Points
—
—
3
3
8
3
2
3
6
2
—
30
TABLE 2
Quantities of Supply and Discharge Water, 1976 Season (Preliminary Data)
Permittee
Patterson RCD (W)1
Mendota-Crows Landing (W)
Turlock I.D. (E)
Merced I.D. (E)
Irrigated
Acres
50,000
400,000
170,000
110,000
Assumed
Irrigation
Demand
(3.5AF/ac.)
(acre feet)
175,000
1,400,000
595,000
385,000
Supply
Measured
(acre feet)
87.0002
930,000
622,800
400,000
Percent
of Assumed
Demand
66%
104%
103%
Discharge
Measured
(acre feet)
10,000
120,000
74,000
39,000
Percent
of Total
(estimate)
60%
90%
30%
70%
'(W) denotes "westside" location or west of the river, (E) east of the river.
2Data only for June-December, 1976.
TABLE 3
Typical Electrical Conductivities and Suspended Solids of Supply and Irrigation Return Water,
1976 Season (Preliminary Data)
Electrical Conductivity
Permittee
Patterson RCD (W)
Mendota-Crows Landing (W)
Turlock I.D. (E)
Merced I.D. (E)
Supply
Water,
(micromhos
per cm)
850
480
40
38
Irrigation
Re turn Flow,
(micromhos
per cm)
1230
1640
260
224
Ratio
ofEC,
(return •+• supply)
1.4
3.4
6.5
5.9
Suspended Solids (mg/l)
Supply
Water
119
68
1
7
Irrigation
Return Flow
333
218
7
33
291
-------
Local Solutions
to Drainage Problems
W. R. JOHNSTONE
Westlands Water District,
Fresno, California
ABSTRACT
Throughout the San Joaquin Valley the
installation of on-farm tile systems can in-
tercept and collect saline subsurface drainage
water and thereby reduce or eliminate the
damage from high water tables. Five subareas
in the San Joaquin and Tulare Lake Basins of
the San Joaquin Valley have individually
developed local solutions for disposing of saline
subsurface drainage water collected within the
subareas.
The West side of the San Joaquin Valley
has been divided into five different subareas for
evaluation of drainage problems depending on
the source of irrigation water, access to surface
drainage, and the source and quality of subsur-
face drainage effluent. Each area has an ex-
isting subsurface saline drainage problem and
some need to better control surface runoff or
return flows. The five general areas are located
in two distinct basins of the valley — the San
Joaquin and the Tulare Lake Basins. See the
map in the first paper of this case study. It is
anticipated that all tailwater or surface return
flows generated in the San Joaquin Valley will
eventually be reused within the water-short
valley.
The first two areas are in the San Joaquin
Basin where the San Joaquin River provides a
natural outlet to the Pacific Ocean through the
Delta.
The Northern Part of the Delta-Mendota
Service Area
The northern part of the Delta-Mendota
Service Area has a number of subsurface
drainage systems which were installed several
years ago prior to any attempts to place water
quality controls on agricultural discharges with
the collected drainage effluent going directly
into the San Joaquin River.
The drainage problem area has increased
from a few acres in 1955 to over 60,000 acres
with a perched water table less than 5 feet below
the soil surface. There are also over 100,000 and
90,000 acres affected by high water tables less
than 10 and 20 feet from the ground surface
respectively, in this area.
About 60,000 acres of the area now have
drainage systems installed which control high
water table conditions, and on-farm drainage
facilities are being installed continuously.
The quality of the subsurface drainage
effluent produced in this area has the lowest
TDS of any area in the valley ranging between
1,500 and 2,500 ppm TDS.
The local solution for disposing of the
subsurface drainage effluent is the construction
of collector systems by local districts with the
systems discharging the effluent into the San
Joaquin River. The discharges will be removed,
however, if and when a valley drain is con-
structed. Considerable surface runoff or
tailwater is also discharged into the river from
this area, but that water would not be accepted
in a valley drain because the water is of
reuseable quality for irrigation.
The Southern Part of the Delta-Mendota
Service Area
The southern part of the Delta-Mendota
Service Area is also partly drained with subsur-
face drainage systems. However, most of the
drainage effluent produced by these drainage
systems is reused by recirculation through open
drains and irrigation canals that traverse
through the area from the south to the north.
Surface runoff or tailwater is also reused with
the subsurface effluent several times before all
of the runoff finds its way into the Grasslands
Water District, a waterfowl habitat area. The
Grasslands area takes all of the drainage water
293
-------
CASE STUDY: SAN JOAQUIN VALLEY
produced in the southern part of the Delta-
Mendota Service Area for flooding waterfowl
hunting areas and native pasture. The quality
of the subsurface drainage effluent in this area
ranges between 2,500 and 9,000 ppm TDS. After
all of the reuse, all drainage discharges from the
Grasslands Water District eventually end up in
the San Joaquin River, which is the source of the
agricultural water supply for much of the
northern part of the Delta-Mendota Service
Area and the Delta areas.
Recently, since more stringent water quali-
ty controls for agricultural discharges have
been contemplated, one small district has bann-
ed the installation of drainage facilities until a
suitable outlet, other than the river, can be
obtained. The district directors have concluded
that no more saline drainage effluent should be
mixed with the irrigation water because signifi-
cant salt balance problems are being created
within their district.
The southern three areas are located in the
Tulare Lake Basin where very few on-farm
drainage facilities have been installed, when
compared to the size of the problem area,
because there is no natural outlet to the ocean.
Federal San Luis Service Area
The Federal San Luis Service Area has
about 300,000 acres that will ultimately need
subsurface agricultural drainage. Currently,
more than 200,000 acres of land within the area
has a perched water table less than 20 feet from
the ground surface at this time.
The portion of the area with less than a 10-
foot water table increased from about 10,000
acres to almost 120,000 acres between 1967 and
1976.
The quality of the perched water table in
this area generally ranges between 6,000 and
15.000 ppm TDS with small areas having salini-
ty concentrations of up to 100,000 ppm. The
local solution for disposal of subsurface
drainage effluent from this area will be through
the use of the United States Bureau of
Reclamation's San Luis Drain. This drain
which is currently under construction is a
concrete-lined open drain with a 300 cfs capaci-
ty, and a control reservoir to provide a drainage
outlet to the Delta and Ocean.
The Bureau is also presently constructing
the initial portion of a drainage collector system
for the 600,000-acre Westlands Water District in
the portion of the district where the water table
is less than 10 feet from the ground surface. The
collector system will convey on-farm drainage
water from the farms to the San Luis Drain.
Each landowner or farmer is responsible for the
installation of the on-farm drainage facilities.
Tulare Lake Area
The Tulare Lake Area has a drainage
problem area of about 165,000 acres. The only
time there is a natural outlet to the ocean from
this area is during peak flooding periods of the
Kings, Kaweah, and Tule Rivers which flow
into the lake area. This is another portion of the
valley where only a small portion of the
drainage problem area is drained with subsur-
face drainage facilities because again no collec-
tion disposal system has been developed. Where
drainage systems exist, all drainage water has
been recirculated for reuse as irrigation water
systems. Recently, the Tulare Lake Drainage
District has constructed a 300-acre pond in the
northern part of the district to evaporate subsur-
face drainage effluent collected by the district.
Planning is also underway for the construction
of a larger pond in the southern end of the
district where all of the district's drainage
effluent will be evaporated, and all the salt that
is contained in the drainage water will be
stockpiled.
The California Regional Water Quality
Control Board has established discharge re-
quirements for the northern and southern
basins. The local solution for the disposal of
subsurface drainage effluent for the Tulare
Lake Basin is ponding and evaporation.
Kern County Area
The Kern County area has developed
drainage problems from increased irrigated
agricultural activities during the last 20 years.
The area with the perched water table within 5
feet of the ground surface has increased from
about 180 acres to over 22,000 acres in the last 12
years. Crop yield will be reduced due to the
drainage problem with resulting substantial
income losses with some properties probably
being removed from agricultural production.
The Kern County Water Agency predicts that
unless the drainage problem is solved, the
annual damages are expected to increase from
$2,665,000 in 1975 to $19,735,000 in 1985 and
over $45,000,000 in the year 2005.
294
-------
LOCAL SOLUTIONS
The Kern County Water Agency has pro-
posed a local solution for their drainage prob-
lem through a project consisting of on-farm tile
drainage systems to remove excess water from
the crop ground root zone; district facilities to
collect effluent from various farms; a main
drain to convey the drainage from the problem
area; and the use of the collected drainage water
for cooling purposes at the proposed San Joa-
quin Nuclear Project Power Plant. It is projected
that under normal operation the cooling towers
at the San Joaquin Nuclear Project would put
the saline water to a beneficial use and also
reduce the volume of the drain water by about 50
percent, helping the disposal problem.
Kern County officials also are investigating
the possible use of the drainage water for
wildlife enhancement either before or after it is
used as cooling water at the power plant and as
a potential attraction for industry which can
economically utilize concentrated brines
developed through additional solar evapora-
tion.
The Wheeler Ridge-Maricopa Water
Storage District, which is a member agency of
the Kern County Water Agency, has about 6,600
acres of land with perched water table within 5
feet of the ground surface and an additional
17,600 acres with a potential drainage problem
that will require on-farrn drainage facilities
within 10 to 15 years. This district may not be
able to wait for the development of a valley-wide
or regional system and is considering several of
their own local alternatives for disposing of
drainage effluent to be collected within the
district. They are: 1) local evaporation, 2) dis-
posal through the Kern County Water Agency
master drainage program, or 3) a plan to be
developed by the San Joaquin Valley Interagen-
cy Drainage Program.
SUMMARY
Throughout the San Joaquin Valley the
installation of on-farm tile systems can in-
tercept and collect saline subsurface drainage
water and thereby reduce or eliminate the
damage from high water tables. However, the
five subareas in the San Joaquin and Tulare
Lake Basins of the San Joaquin Valley have
individually developed local solutions for dis-
posing of saline subsurface drainage water
collected within the subareas. The northern part
of the Delta-Mendota Service Area discharges
its drainage water into the San Joaquin River;
the southern part of the Delta-Mendota Service
Area disposes of its drainage water into the
Grasslands Water District where the water is
used for irrigating native pasture and wildlife
habitat.
The San Luis Drain is being developed to
convey drainage effluent from the Federal San
Luis Service Area to the western Delta; the
Tulare Lake Drainage District is developing
evaporation ponds to totally evaporate the
drainage water collected in that district; and the
Kern County Water Agency is contemplating
the use of saline drainage water for cooling the
reactors in a nuclear power plant.
295
-------
A Valley-Wide Solution
L. A. BECK
San Joaquin Valley Interagency Drainage Program,
Fresno, California
ABSTRACT
The San Joaquin Valley Interagency
Drainage Program (IDP) is an action-oriented
program with objectives including: 1) coordina-
tion of the on-going drainage water manage-
ment activities of the cooperating agencies
(USER, SWRCB, and DWR); 2) development of
alternative plans for managing the drainage
waters; 3) determination of potential uses for
drainage waters; 4) development of a
recommended plan for managing drainage
waters; and 5) recommending ways to finance
drainage facilities.
Drainage problems in localized areas of the
San Joaquin Valley are severe. Yields on in-
dividual fields may be reduced as much as thirty
percent. The third paper of this case study
described the rapid increase in acreage affected
by drainage problems. However, on a valley-
wide basis only 15 to 20 percent of the potential
or ultimate drainage problem area is presently
affected. These present drainage problem areas
represent only about 4 percent of.the valley's
irrigated land.
This small percentage on an overall basis
makes it difficult to provide a valley-wide or
ultimate solution now. Table 1 in the first paper
of this case study showed that it would be after
the year 2000 before drainage production was
half of the ultimate projected. It would be
difficult to justify operating facilities at partial
capacity for so many years. It would also be
difficult to finance the ultimate facilities with
this limited repayment capacity while the flows
are increasing. Those individuals without
problems now are often difficult to convince
how severely they will be affected in the future.
The relatively small number of people affected
now have not been able to convince the
legislature to authorize drainage facilities.
Local drainage solutions may not be the
long term or ultimate solution. Salts are the
problem component of the drainage waters. The
salts must be managed and ultimately removed
from the valley to provide a salt balance and
protect agricultural production.
Planning and implementation must be
done now to provide relief for those with ex-
isting problems and to integrate any present
actions into future and ultimate solutions.
The drainage problem has been recognized
for many years. A 1965 report, DWR Bulletin
No. 127, "San Joaquin Master Drain",
recommended an ultimate solution that was not
implemented for the aforementioned reasons. It
was decided in 1975 to develop a drainage
solution that could be implemented. The San
Joaquin Valley Interagency Drainage Program
(IDP) was formed by the U.S. Bureau of
Reclamation (USER), the California State
Water Resources Control Board (SWRCB), and
the California Department of Water Resources
(DWR). The IDP objectives include: 1) coordina-
tion of the on-going drainage water manage-
ment activities of the cooperating agencies; 2)
development of alternative plans for managing
the drainage waters of the San Joaquin Valley,
while maintaining agricultural productivity
and surface and ground water quality; 3) deter-
mination of potential uses for drainage waters;
4) development of a recommended plan for
managing drainage waters; and 5) recommend-
ing ways to finance drainage facilities.
IDP staff consists of four professionals and
two clerical. The staff works with members of
each agency to coordinate activities. The staff
also meets routinely with a public participation
group. The public group represents irrigators,
local government and areas under considera-
tion as disposal sites.
IDP is an action-oriented program that will
develop recommendations that can be im-
plemented in the near future, as well as an
ultimate program. The recommended program
may consist of local or regional projects in the
first phase and then integration into valley-
297
-------
CASE STUDY: SAN JOAQUIN VALLEY
wide facilities that accomplish salt manage-
ment and disposal. The recommended San Joa-
quin Valley Interagency Drainage Program
will be developed by the Fall of 1978. Additional
time will berequired to complete necessary work
to upgrade the report so that it can be presented
for legislative authorization.
Five basic alternatives are being in-
vestigated by IDP. They are: 1) no valley-wide
action; 2) evaporation ponds; 3) discharge to the
ocean; 4) discharge to the ocean via the San
Joaquin River and San Francisco Bay; and 5)
discharge to the ocean via the western Delta and
San Francisco Bay. Different disposal points
are subalternatives within each alternative.
USER is conducting an economic analysis
of the alternatives. SWRCB is conducting an
environmental assessment of the alternatives.
DWR is developing a financial program and
studying legal and institutional constraints.
When these studies of the alternatives are
completed, the recommended program will be
developed. The recommended program may
include components from more than one alter-
native. It will include staging of facilities and
reuse of drain water before disposal.
One major difference between IDP and
previous studies is that reuse of drainage waters
is a primary consideration. The San Joaquin
Valley is in an overdraft situation (about 1,500,-
000 acre-feet per year). Even though the
drainage waters are brackish, there are many
uses for any water in this water-short area.
Potential uses of drainage waters include power
plant cooling, marsh development for waterfowl
enhancement, recovery of salts, salinity repul-
sion, and aquaculture.
In addition to reuse, reclamation by
desalting is being studied. All the water supplies
in the valley are now allocated. A new water
development project to import water into the
valley would probably cost $150 per acre-foot
when the water was delivered. Desalting costs
are not much higher than this cost, and the
volume of water for disposal would be reduced.
One difficulty with desalting is that reverse
osmosis does not remove the 10 mg/lof boron in
the drainage waters.
Power plant cooling and marsh develop-
ment appear to be the most practical at this
time. Two nuclear power plants are under
consideration in the San Joaquin Valley. The
timing of construction of these and future power
plants may fit with the development of drainage
waters. Development of additional marshes
would use significant quantities of drainage
water. Marshes and power plants might be used
in series. The effluent from marshes and power
plants would be increased in salinity and would
need to be exported from the valley.
Salinity repulsion would take place in the
western Delta. Fresh water release is required to
prevent sea water intrusion into the Delta.
Discharge of drainage waters at the western
Delta would reduce the requirement for fresh
water to be used elsewhere.
Salt reclamation and aquaculture do not
seem too practical now, but are being in-
vestigated because they may be in the future.
The salt companies are not interested in recover-
ing salt from drainage water because there is
more sodium sulfate than sodium chloride.
Chemical companies may be interested in in-
dividual constituents in the future, but initially
the quantities would be too small. Additional
study needs to be done before the practicality of
aquaculture with drainage water can be
demonstrated.
It is important that the recommended
program be very flexible with the possibility of
changes in technology and the uncertainty of
drainage flow projections. At any decision point
in the future, the option to change or adjust the
drainage program must be available.
298
-------
Irrigation Return Flow Problems
in Yakima Valley
JOHNSPENCERand MARC HORTON
Department of Ecology;
State of Washington; Olympia, Washington
ABSTRACT
Studies completed in recent years show
that irrigated agriculture is a significant source
of the pollutants in the lower reaches of the
Yakima River. Sediment, phosphates and
nitrates are the principal pollutants resulting
from irrigation activities. Improved on-farm
water management practices should be under-
taken to alleviate this water quality degrada-
tion.
Since the passage of P.L. 92-500, attempts
to apply NPDES permits to irrigation return
flows have been stymied by legal questions and
a lack of knowledge of relationships between
farming practices, soils, weather conditions,
and water quality. In addition, there are legal
questions regarding the authority of irrigation
districts in the State of Washington to control
on-farm water management practices.
INTRODUCTION
The Yakima Valley has a total area of 6,062
square miles and occupies the north central part
of the State of Washington. About three-fourths
of the valley is in the Columbia Basin
physiographic province, with the remainder in
the Cascade Range. The Cascade Range
generally bounds the valley to the west, with the
Wenatchee Mountains forming the northern
boundary, and Rattlesnake Hills to the east and
Horse Heaven Hills to the south separating
from the Columbia Valley.
The Yakima River and its tributaries drain
the valley. The river heads near the crest of the
Cascade Range, northeast of Mount Rainier
and flows for 180 miles in a generally
southeasterly direction to its confluence with
the Columbia River near Richland. It is the
largest single river system located entirely
within the State of Washington. Major
tributaries include the Naches, Cle Elum,
Kachess and Teanaway Rivers. The total an-
nual flow leaving the valley under natural
conditions would be approximately 3,580,000
acre feet. The actual flow is 2,340,000 acre feet.
The average annual irrigation diversions
within the basin are 2,365,000 acre feet of which
1,140,000 acre feet are consumed and 100,000
acre feet diverted from the basin.
The climate varies from desert conditions in
the lower valley to a moist alpine type in the
higher mountains. The irrigated area of the
upper valley receives from 10 to 15 inches of
precipitation per year, while the lower valley
receives less than 10 inches.
Agriculture is the primary economic activi-
ty. Yakima is a center for processing of both
fruit and vegetable crops grown in the adjacent
valleys. There are a limited number of lumber
and wood product plants.
Population densities are high in the valleys
and low in the surrounding rangelands. The
present population is slightly in excess of the
1965 figure of 184,500, and is centered in the
towns along the Yakima River.
The first known attempt at irrigated
agriculture in the valley was made in 1853 at the
Ahtanum Mission near Tampico. In 1867,
farmers were diverting water from the Naches
River and by 1870 there were 1,000 acres under
irrigation in the Yakima Valley. The early
systems were privately constructed and served
easily accessible lands along the main stem of
the Yakima River and the lower portions of the
major tributaries.
During 1886-1888 the Northern Pacific
Railroad's transcontinental line reached the
area and extended over the Cascades to Puget
Sound. As it had been granted a considerable
acreage in the valley, the railroad and other
299
-------
CASE STUDY: YAKIMA VALLEY
sources of private capital undertook large in-
vestments in irrigation in order to attract
settlers. The Sunnyside Canal was started in
1890 to divert water from the Yakima River and
by 1892 forty-two miles of canal were completed.
In 1900, management was taken over by the
Washington Irrigation Company. In the mean-
time, several other canals were constructed by
private capital so that by 1900 the Yakima
Valley was the most extensively irrigated area
in Washington, with 67,000 acres. This had
expanded to 125,000 acres by 1905 when irriga-
tion by the Bureau of Reclamation and the
Office of Indian Affairs began.
The Yakima Valley is now one of the most
extensively irrigated areas in the United States,
having six storage dams, five diversion dams,
two hydroelectric plants, six major governmen-
tal irrigation projects plus numerous small
private irrigation systems and districts. The
total area served by irrigation facilities in the
valley is approximately 505,000 acres of which
416,000 are served by governmental projects.
Water for the remaining irrigable land is
supplied through Warren Act contracts and by
individuals diverting from small streams.
About 18,000 acres are irrigated by
groundwater.
Approximately 80 percent of the irrigated
area of the valley is watered by furrow or flood
irrigation methods. The remainder is sprinkler
irrigated except for a few hundred acres of drip
irrigation. There is currently a marked trend
towards sprinkler systems.
With the advent of irrigation in the valley
the quality of water in the Yakima River and its
tributaries has declined. Some of the degrada-
tion is directly due to irrigation return flows,
while further deterioration is associated with
fruit processing plants and other agricultural
and industrial enterprises attracted to the
valley by the strong irrigated agriculture base.
In the upper reaches of the valley, water
quality is virtually unimpaired. Although
mineralization increases more than four-fold in
going from the headwaters to the mouth, the
water still remains in the low salinity class and
is suitable for reuse as irrigation water. The
main effect of irrigation activities is high con-
centrations of suspended sediment, phosphates
and nitrates.
Suspended solids in return flows settle and
interfere with drain design flow capacity and
cause wear of pumps and sprinklers. Where
discharged to the river, they settle behind
diversion structures and require periodic
removal. Associated with the sediment is at-
tached phosphate which frequently reaches
levels in the lower river above the potential
algal bloom limiting concentration. Similarly,
nitrogen levels, which are a consequence of the
percolation of irrigation water applied in excess
of crop needs, greatly exceed the potential algal
bloom limiting concentrations in the lower
river. Heavy growths of plankton and higher
forms of aquatic life are consequently found in
the river from Wilson Creek, a few miles below
Ellensburg, to the mouth. The photosynthesis
and respirational activities of these organisms
cause a wide diurnal fluctuation in several
water quality parameters including dissolved
oxygen, pH and alkalinity.
Federal and State Irrigation Return Flow
Activities
Federal Water Pollution Control Act
Amendments (P.L. 92-500)
The Federal Water Pollution Control Act
Amendment of 1972 (P.L. 92-500) was the most
far-reaching water pollution legislation ever
undertaken by Congress. The ramifications of
its passage began to be felt immediately.
However, debates on its intent and applicability
were to continue as its various programs were
undertaken. Some debate has been resolved
through judicial processes. Much debate will
continue among and within administrative
agencies as they struggle with interpretation of
the legislation. In the agricultural pollution
area, the intent of the Act was less clear than
with other segments. After four years of inter-
pretation, both by the courts and involved
agencies, the struggle continues as to how to
carry out the intent of Congress in the
agricultural sector.
The potential for pollution from
agricultural lands was mentioned in the Act
under two specific headings. First, Section 304
required from EPA information and "guidelines
for evaluating the nature and extent of nonpoint
sources of pollutants" and "processes, proce-
dures and methods for control of pollution"
resulting from agricultural activities (among
others). Section 208 of the Act required, as an
element of area-wide planning processes, the
identification of agricultural related nonpoint
sources of pollutants. This section also required
identification of procedures and methods to
control these sources to the extent feasible.
30O
-------
RETURN FLOW PROBLEMS - YAKIMA VALLEY
The Act had one other connection to return
flows from irrigated lands. Section 502 —
general definitions — defined a point source as
"Any discernible confined and discrete con-
veyance, including, but not limited to, any pipe,
ditch, channel, tunnel, conduit, well, discrete
fissure, container, . . from which pollutants are
or may be discharged."
Point sources were to be treated quite
specifically and in their own particular manner.
Section 402 of the amendments spelled out the
details of a NPDES program which would place
dischargers of pollutants from point sources
under increasingly more stringent limitations.
The goals of the Act were rather ambitious —
such as the 1983 goal of "fishable and swim-
mable" waters. Large portions of the Act were
devoted to spelling out the details of the NPDES
program. Looking back, it is apparent that
discharges from industries and municipalities
via pipes and ditches were seen as a major
threat to the nation's water quality. From all
that has been seen of the NPDES program, it
has had the most significant effect with the
treatment of sources with discrete processes. It
is also apparent that, in addition to process
changes, end of the pipe effluent treatment was
seen as a major tool in improving water quality.
Along with funds for sewage treatment plant
construction and upgrading, research and plan-
ning, the NPDES program was one of the major
tools provided for the attainment of the ACT's
ambitious goals.
In spite of all the above, it remains a fact
that the only specific referral to agricultural
return flow was as a nonpoint source. These
sources of pollution were to be treated under the
results of long-term planning — specifically
that of Section 208. On the other hand, the
administrating agency, the Environmental
Protection Agency (EPA), could not ignore the
definition of point source and their mandatory
treatment under the NPDES permit program.
In December 1972, EPA began to solicit
information concerning the types of
agricultural point sources which would come
under the NPDES program. The result of these
activities were regulations pertaining to
agricultural point sources published July 5,
1973. These regulations as they apply to irriga-
tion return flows, attempted to exempt what
EPA considered to be the less significant por-
tion of dischargers.
NPDES requirements apply to discharges
of irrigation return flow (such as tailwater,
tile drainage, surfaced groundwater flow or
bypass water), operated by public or private
organizations or individuals, if: (1) there is a
point source of discharge (e.g., a pipe, ditch,
or other defined or discrete conveyance,
whether natural or artificial) and; (2) the
return flow is from land areas of more than
3,000 contiguous acres, or 3,000 non-
contiguous acres which use the same
drainage system.
It is the individual or organization who
actually has control of or responsibility for
the discharge of irrigation return flow who
must apply for a permit. For example, if
water is supplied by an organization but the
discharge of return flow to navigable
waters is controlled by an individual who
has more than 3,000 acres under irrigation,
it is the individual who must apply for a
permit. On the other hand, if an irrigation
organization supplies and controls the
irrigation return flow from a total of 3,000
acres to navigable waters, the organization
must apply for a permit, even though one
individual may be supplied with water for
3,000 acres or more.
The initial regulation did not specifically
state the individual or organization who was to
receive the permit. However, because of their
legal powers and size, it was obvious that
irrigation districts were the planned recipients.
The interpretation of the above paragraphs
was to determine the extent to which
Washington's irrigated agricultural lands
would be affected. For instance, if all irrigation
districts over 3,000 acres received permits, ap-
proximately 94 percent of Washington's
irrigated land would be affected. If, on the other
hand, those districts with 3,000 acres draining
to one discharge were permitted, then at the
most 70 percent of the irrigated acreage would
have been affected.
Three Major Issues
The details of which districts were to receive
permits was to be one of three primary issues. A
second of the major issues also pertained to the
districts — it was whether any irrigation district
had the authority over, and responsibility for,
irrigation return flows. The third issue per-
tained to the 3,000-acre limitation and EPA's
301
-------
CASE STUDY: YAKIMA VALLEY
authority for granting exemptions to certain
categories of point sources.
1. Irrigation districts in Washington
receive their authority to operate under Chapter
87 Revised Code of Washington. The purposes of
irrigation districts are rather clearly
enumerated in that chapter, and among those
purposes and powers are those of construction
and maintenance of delivery and drainage
"systems". Also, districts are granted the power
of assessment for costs of operation of such
systems. In relation to specific delivery systems,
the district's control terminates at "the point of
individual distribution". Because the point of
control and authority specifically stops at the
point of individual distribution, districts argued
that they had no control over the manner in
which an individual managed his water. This
has been (and is) a major obstacle to acceptance
of the permit concept by districts.
2. Also at issue is the question of whether
or not districts have the authority to operate
"systems" for pollution control. RCW 87.03.120
and 87.03.125 grant the authority for operation
of sanitary sewage or drainage systems by a
district when it "is of special benefit to the lands
of a district as a whole". It seems then that
districts have the power to operate the facilities
for pollution control and the maintenance of
"systems" is left for interpretation.
It appears that the powers of pollution
control (although not specifically termed as
such) and assessment are available. However,
three more subtle questions result from this
observation: 1) Is "end of the pipe" pollution
control cost-effective or socially acceptable? 2)
With a pollution control and assessment system
within the district, could the political structure
of the districts survive? 3) Can assessments be
made to those (the poorly managed farms)
within the district who receive district serv-
ices?
The former two questions will be dealt with
in a subsequent portion of this discussion;
however, to the latter question, it should be
pointed out that irrigation districts routinely
utilize their manpower and machinery to carry
out pollution control construction and
maintenance for individuals. They also routine-
ly assess those individuals for a portion of the
cost incurred (e.g., construction and cleaning of
sediment basins).
3. A third major issue was developing at
the time EPA formalized its regulations on
agricultural point sources. This issue was that
of EPA's authority for the 3,000-acre exclusion.
The 3,000-acre limitation apparently was placed
in the regulation for two reasons: 1) To minimize
the workload by State agencies and EPA by
minimizing the number of permits issued, and
at the same time, 2) To focus governmental
resources toward the major portion of the
problem. According to EPA estimates, 80 per-
cent of lands irrigated by irrigation districts or
similar irrigation organizations fell within the
3,000-acre criteria.
The Natural Resources Defense Council
(NRDC) took exception to the July 5, 1973
agricultural regulations by filing an action with
the District Court in Washington, D.C. Their
exception was based on their opinion that the
Law (P.L. 92-500) provided no authority for
exclusion of point source discharges from the
NPDES permit system. Secondly, the NRDC
felt that a significant source of pollutant
loading could come from operators under 3,000
acres.
Washington State's Involvement
It was not until November 1973 that the
Washington State Department of Ecology
(DOE) accepted the responsibility for ad-
ministration of the NPDES program. Transfer
of the agricultural portion of that program was
completed in July of 1974. At that time, none of
the major issues had been resolved. Further,
opposition was becoming formally organized
and battle lines were forming.
The DOE quickly developed a scheduled
program for obtaining NPDES applications
and issuance of permits. Although program
planning got off to a flying start, the program
quickly fell behind schedule because none of the
major issues had yet been resolved.
With pressure to get the program moving,
the DOE sought the guidance and advice of
entities concerned with water and skilled in the
area of irrigation. This took place with the
formulation of a State Technical Advisory Com-
mittee for Water Quality Improvement (TAG).
Formation of this group took place in September
1974, and agencies and organizations which
took part were: U.S. Geological Survey, U.S. Soil
Conservation Service, U.S. Bureau of Reclama-
tion, Washington State University, Washing-
ton Association of Conservation Districts,
Washington Reclamation Association, State
Association of Washington Irrigation Districts,
302
-------
RETURN FLOW PROBLEMS — YAKIMA VALLEY
and the Department of Ecology. The initial
tasks of this group were to review the problems
of application of the NPDES permits to irrigat-
ed agriculture and to provide input to the con-
tent of the proposed permits.
During the first eight months of DOE's
administration of the agricultural permits,
some applications were received and, with the
guidance of the TAG, these applications were
processed. The process was carried out to the
point of "legal notice," but the proposed permits
were not issued. Several districts refused to
submit an application. However, those who did
submit applications submitted them only for
those facilities for which they felt they had
control (operational wasteways).
In early 1975, the NRDC received a
favorable judgment regarding the 3,000-acre
limitation. The Court ordered EPA to draft and
promulgate a new set of regulations relating to
irrigated agriculture and gave them one year to
complete the task.
At this point, the TAG should have been
dissolved, but there remained the spirit of
cooperation, awareness of the agricultural
pollution problem, and the momentum toward a
positive approach to abatement of these
problems. The TAG was to remain intact and
active for over a year after the NRDC decision.
In fact, it was this group, with the attributes
mentioned above, that was responsible for
guiding initiation of the Sulphur Creek project.
Water Quality in the Yakima River Basin
As part of its statewide water quality plan-
ning efforts, the DOE, in 1973, undertook the
development of a water quality management
plan for the Yakima Basin. The basis of this
plan was to be developed through contracts to
the consultants CH2M-Hill (CH2M).*
A portion of the report from CH2M dealt
with a current characterization of water quality
in the Yakima Basin. This report evaluated
sources of pollutants entering the Yakima
River. CH2M conducted its own monitoring but
also relied extensively on existing data. A map
of the Yakima River system is shown in Figure 1
(CH2M). Figure 2 (CH2M) shows the results of
*CH2M, "Characterization of Present Water Quality
Conditions in the Yakima Basin," Department of
Ecology, February 1975, and CH2M "Agricultural
Return Flow Management in the State of Washing-
ton," Department of Ecology, April 1975.
its source load evaluation. From this overall
view, it can be seen that the Yakima River below
Sunnyside Dam (the diversion structure for the
Sunnyside Valley Irrigation District) is degrad-
ed, largely due to the impact of irrigation
activities.
The preface of the characterization report
states that information "clearly indicates
agriculture as a significant source of
pollutants," and further, "any attempt to im-
prove the quality of the waters in the Yakima
Basin must make considerable efforts to im-
prove the quality of return flows." The emphasis
on return flows follows from an estimate that in
an average year, 80 percent of the summer flow
below the Sunnyside Dam to the mouth of the
Yakima River is attributable to return flows.
Following completion of the water quality
management study, the DOE again contracted
with CH2M to look at agricultural return flow
management in the State. The objectives of the
study were as follows:
1. Determine the relative contribution of
point and nonpoint sources to water quality
degradation in a predominantly agricultural
area.
2. Determine alternative methods of
reducing or eliminating the addition of wastes
from these sources to water bodies.
3. Identify institutional constraints on
applying the NPDES permit program to return
flow management.
4. Recommend methods of applying the
NPDES permit program to feedlot and irriga-
tion return flows to result in improved instream
water quality.
CH2M conducted the study in three phases.
The first phase was a detailed study of the
Sulphur Creek Drainage, a subbasin within
Washington's Yakima Basin. The second phase
was a semidetailed study of the Sunny side-Roza
Area of the Bureau of Reclamation's Yakima
Project. In the third phase, the results of the first
two phases were applied to the entire Yakima
Basin.
Results of the study pointed to improvement
of on-farm management practices as the initial
step in improvement of water quality in the
Yakima. Sediment, phosphates, and nitrates
were the principal causes of degradation from
irrigation activities. For instance:
1) Phosphate losses in some areas were as
high as 40 Ibs/ac/yr,
303
-------
CASE STUDY: YAKIMA VALLEY
2) Nitrogen losses were as high as 30
Ibs/ac/yr - - up to 20 percent of that
applied — Nitrogen concentration in the
lower Yakima has increased five-fold
over the last 20 years, and
3) Cleaning sediments from the drains in
the Sulphur Creek basin alone cost
$65,000/year — up to 190 tons per day are
transported from the Sulphur Creek
drainage to the Yakima River.
In the institutional area, the report pointed
out that it is unfortunate that our current water
rights system "does not encourage better
management of our water resources" (CH2M,
page 10, Summary Report).
Nevertheless, living within our present
legal framework and with the need of ad-
dressing the NPDES permit system, the report
suggested the following guidelines for any
program developed to meet current needs:
1. Recognize that individual farm
operations are the significant cause of pollution
in irrigation return flows. The strategy should
be geared to ways of improving on-farm prac-
tices. The initial emphasis should be on the
small percentage of farmers responsible for the
majority of the problems.
2. The irrigation districts must play a
major role in seeking solution to water quality
problems by identifying problem areas, im-
x-. -"•" ^- p-~
»i< v'~- v- -7 —
:*£' ^ / '' v
V?- ";
YKIMA WVER
X^(BASW-
s xX
SCIU II Pi.!'
Figure 1. Map of the Yakima River system.
304
-------
«T CLE ELIM
•I ROZ* 0*1
«T SUIWYSIDE D*N
«T mod*
rim
TITIl
GISSOUEO SOLUS
urn
JUIMitL
TOItL
•timUML 0.02
IIIIUIIOH 4
IIHIICIPtL 0.3
ICIPU 0.2
IDIIItTION I
lt»tl 1.9
ICIML 0.9
lltlUIIM i
SEPTIC TIHI 0.3
SIPTIC TtM 1
SCPTIC TMI O.I
H
!
i
:
1
MPTIC TIM 1
•MICIHl II
HIIMTIM II
Figure 2. Sources of pollutants in the Yakima River. Values are expressed in percent.
:•
•
-•
•
-------
CASE STUDY: YAKIMA VALLEY
proving water deliveries, and encouraging
better water management.
3. Statewide and regional advisory com-
mittees could be an effective way to assist the
DOE in developing and implementing a pollu-
tion abatement program for irrigated
agriculture.
4. The basin approach to improving water
quality should be through an education
program designed to assist individual farm
operators. The Bureau of Reclamation, the Soil
Conservation Service, and Washington State
University have developed programs en-
couraging better on-farm water management.
5. Water quality monitoring should be
continued and should be designed to measure
long-term quality changes. Quantity
measurements should be an integral part of all
quality measurements. (CH2M-1975).
In all of the work that has been completed or
is underway concerning agricultural pollution
abatement, there seems to be three general
topics: 1) legal frameworks, 2) technical alter-
natives, and 3) the socio-economic realities
(institutional frameworks). The broad nature of
these topics in itself points to the complexity of
the problem. It is apparent that any long-range
solution to pollution from irrigated agriculture
must deal with all of these aspects. However, as
a beginning, the Sulphur Creek project was
developed to deal mainly with the technical and
socioeconomic factors.
306
-------
The Sulphur Creek Pilot Project
A Practical Approach to
Control of Pollutants Leaving
Irrigated Farmlands
JOHN SPENCER and MARC HORTON
Department of Ecology; State of Washington,
Olympia, Washington
JIM GLEATON
Soil Conservation Service; Sunnyside, Washington
ABSTRACT
The Sulphur Creek project grew out of the
momentum and confusion resulting from
attempts to apply a National Pollution Dis-
charge Elimination System (NPDES) permit to
irrigation return flows. The result has been a
project containing what appeared to be compati-
ble and, for the most part, admirable objectives
— soil and water conservation, improved water
quality, improved fertilizer and pesticide
application, improved crop production, and
local control. Along with education and
technical assistance for the individual farmer,
this unique program is structured* to provide
local control to decisions affecting the farmer.
The Best Management Practices concept
assumes that informed decisions and improved
management will generally provide improved
crop production, conservation of soil resources,
and cleaner water.
The project is currently just over one year
old. Details of the technical success of the
project will not be available for some time.
Unlike most types of industry covered by the
NPDES permits, irrigation return flow improve-
ment represents complex hurdles which have to
be overcome by facing the realities of
technology, politics, economics, and social
patterns.
Description Of Area
Sulphur Creek (Figure 1) is a perennial
tributary to the Yakima River, draining over
48,000 acres near Sunnyside, Washington. Two
major irrigation districts (Sunnyside and Roza)
cross this drainage.
Major crops grown in this area include
mint, sugar beets, corn, hops, tree fruit, and
small grain. Total crop value exceeds $46
million annually.
Within Sulphur Creek, there are about 550
farm operators. A 1974 study of this area found
that these farmers receive an average of 584
cubic feet/second (cfs) of water during the
irrigation season. Of this, approximately 208 cfs
is returned to the Yakima River, about 35
percent of the total delivery. These return flows
carry with them an average of 190 tons of soil
per day. Sulphur Creek is the most significant
drainage to the Yakima River in terms of
sediment load, as shown in Table 1.
Conceptual Framework
Soon after the NRDC versus EPA decision
was entered (March, 1975), the Department of
Ecology (DOE) and the State Technical Ad-
visory Committee for Water Quality Improve-
ment (TAG) were wrestling over how to deal
with the 12-month lull while new regulations
were developed. The momentum generated
through the TAG was positive and forward in
nature. It was not surprising that through this
group the idea of a state pilot project developed.
Sulphur Creek began with discussions
between the U.S. Soil Conservation Service
(SCS) and DOE concerning possibilities of in-
307
-------
CASE STUDY: YAKIMA VALLEY
•4-
• ^
LEGEND
WaftMvty Dischargt Points
Surface Dnin*
1*4''* -W-,
UP i _ / -' • t
A.C.R Referrols
Figure 1. Sulphur Creek drainage area. The abbreviation "DID" refers to "Drainage Improvement District.'
308
-------
TABLE 1
Comparison of Major Drains in Yakima Basin
(Average values for 1974 irrigation season)
CO
Measured Parameters
Flow (cfs)
Temperature (°C)
Dissolved Oxygen
(Mg/L)
pH (Units)
Conductivity
(Micromhos/cm)
Total Coliform
•H (Count/100 ml)
o Fecal Coliform
ffl (Count/100 ml)
Cod (Mg/L)
<| NO + NO (Mg/L)
O Z
•H Kjeldahl (Mg/L)
!<2
-------
CASE STUDY: YAKIMA VALLEY
teragency cooperation in a program of
agricultural pollution abatement. These dis-
cussions culminated in submission of an issue
paper to the TAC by the Washington Associa-
tion of Conservation Districts (WACD) and the
SCS. This paper specifically dealt with the
conservation districts' proposed role in the
NPDES irrigation return flow program. Calling
for Conservation District (CD) and SCS
assistance to both DOE and the farmer, the
paper outlined a proposal whereby farmers
could complete the traditional conservation
planning process and, in that way, meet the
requirements of the State from a water pollution
control standpoint. In the conservation plan-
ning process, a particular farm unit is evaluated
against current technology and recognized con-
servation practices. Following evaluation, the
individual farmer is provided a list of alter-
native pathways which, when carried out,
would get him from where he is today to
conservation farming at some point in the
future. This type of planning process contained
all of the flexibility and local considerations
thought necessary for a successful program in
pollution abatement for irrigated agriculture.
Conservation districts (CD) and the SCS also
offered their assistance in the implementation
of conservation plans and with the recording
and documentation processes that would follow.
The initial paper presented to the TAC recog-
nized the need for information education. It
called for full utilization of the Washington
State Cooperative Extension Service's capabili-
ties in informing the farmer of current
technology.
After considerable discussion, a final pro-
ject proposal was prepared by the DOE utilizing
the WACD SCS issue paper as a basis. The
following is a description of how that program
was to work:
1. The proposed program for the Yakima
Valley was not a study but an operational
program for decreasing erosion on the farm
and reduction of associated adverse water
quality effects. Results were to be ac-
complished by providing better manage-
ment practices while considering the in-
dividual farmer's goals. Also, this program
was to promote refinement of water delivery'
and return systems within irrigation dis-
tric*s which would complement the on-farm
effort.
2. The on-farm portion of the program was to
be a joint SCS, CD, irrigation district (ID),
farmer endeavor which would result in a
farm plan. This plan, when certified by the
Department of Ecology, would assure the
farmer that he would be in compliance with
water pollution laws and was using the best
reasonably available technology according
to his individual circumstances.
3. The program was to begin with a valley-
wide public information release which
would explain the reason for the program
and the elements of the program.
4. The Operations Committee (a group of local
irrigation conservation specialists) was to
consider problem areas in the valley from
information supplied by SCS, irrigation
districts, and DOE and was to designate a
priority list of problem areas. These areas
were then to receive a more intense public
information effort which would explain in
detail the program and alert farmers to the
arrival of SCS specialists.
5. The most important facet of the program
was that the farm plan was to be designed
by the farmer himself to fit a particular
farm. Technical help was to be provided by
the SCS specialist as he and the farmer
worked together developing each plan.
6. Following the intense information program.
the SCS specialists were to begin calling on
farmers to begin developing farm plans. The
resulting plan would be reviewed by the
local conservation district, and, if accep-
table, the local district would recommend
certification by the Department. The cer-
tified plan was to indicate that the farmer
would be using the best available
technology and would be in compliance with
Federal and State water pollution control
laws.
7. Implementation of the plans was to begin
prior to the beginning of an irrigation
season with technical help from the SCS,
CD, or irrigation district. These plans would
be written for periods of time ranging from
three to ten years, depending on each in-
dividual's desired goals and economic
capabilities. The SCS would assess all im-
plemented plans and would submit to DOE
an annual summary of progress. Each CD
was to maintain records of developed farm
plans. These records would be available to
DOE as necessary for review.
310
-------
8. The order or priority to be used by SCS
personnel when entering a "problem area"
was as follows:
a. Work with those individuals who were
interested, would cooperate voluntari-
ly, and could accomplish im-
provements without major investment
requirements.
b. Work with those individuals who
would cooperate voluntarily but would
require some cost sharing for struc-
tural or other improvements. These
could be identified initially and
bypassed until cost-sharing funds are
located.
c. Identify those individuals for ap-
propriate enforcement action who were
refusing to cooperate.
9. All farmers in a "problem area" would not
necessarily require a farm plan for correc-
tive action. Those farmers who were
managing their operations responsibly
were to be identified to the CD by SCS and
bypassed.
10. The scope of this program was, ideally, the
entire Yakima Valley. Initial funding
limitations precluded this broad geo-
graphic scope and indicated concentration
of the effort to, perhaps, the lower Yakima
Valley (Granger-Sunnyside vicinity),
where most indicators suggested the
SULPHUR CREEK PROJECT
greatest problem existed. If additional
funds became available, the program could
easily be expanded to include other "pro-
blem areas" in the valley.
11. The SCS and CD's have worked successful-
ly with this approach for 40 years with
those who would cooperate. The past ap-
proach had been strictly voluntary. It was
envisioned that eventually a regulation
might be developed which would specify
the need for management plans in those
areas where voluntary effort had not been
successful. Such a regulation would
provide for a certification procedure which
would assure farmers with certified plans
that they are in compliance with water
pollution control laws.
The basic concepts of this approach are il-
lustrated in Figure 2.
Project Initiation
The initial steps in getting a project such as
Sulphur Creek underway included organization
of the Operations Committee, orientation of the
technical staff, and the gathering of available
information.
The initial meeting of the Operations Com-
mittee was held in June 1975, and the actual
Committee membership included the following:
Conservation Districts — Four members
Irrigation District Boards — Three mem-
bers
Operation* Committee
Recommends Activity
in Problem Area
On
Farm
Assistance
Priority One
Priority Two
Priority Three
Problem Area.
Lands with Significant
Land and Wafer
Management Needs
Figure 2. Flow chart for implementation of the Sulfur Creek Pilot Project.
311
-------
CASE STUDY: YAKIMA VALLEY
U.S. Bureau of Reclamation (Yakima Pro-
ject) — One member
Bureau of Indian Affairs (Wapato Irriga-
tion District) — One member
Cooperative Extension Service (Yakima
County) — One member
U.S. Corps of Engineers (Seattle) — One
member
Department of Ecology (Yakima Office) —
One member
This management team was formed to
provide local control. The purpose of the
Operations Committee was to make the farm
plan concept work. The 12 members who made
up this Committee were, for the most part, local
to the Yakima Basin and intimately involved in
irrigation and farming activities.
The Committee was charged with the tasks
of 1) giving general guidance to the program,
2) determining areas of need for the program.
and 3) acting as a coordinating body between
the individual farmers and governmental agen-
cies. In this way, the Committee has provided
local representation with the best interest of the
individual farmer and the environment under
consideration.
Details of funding for the project also had to
be worked out in the early months. The DOE
and the SCS both agreed to contribute $100.000
for the initial two years of the project. The
majority of this $200,000 funding was to provide
technical assistance to the farmers. The actual
contract between the SCS and DOE took longer
than expected to be developed. It wasn't until
mid-1975 that the papers were signed and the
project was ready to go.
The Operations Committee went quickly to
work and identified Sulphur Creek as the initial
target area for the Yakima program. However, it
was August, and the irrigation season was
essentially completed in 1975 before the SCS
technical staff was identified, familiarized with
the area, and ready to begin development of
farm plans.
During this organizational period (June
through August, 1975), the Operations Com-
mittee was thoroughly updated as to the status
of regulations under P.L. 92-500, the mechanics
of the program, and Yakima River water quality
data (principally CH2M studies). After review-
ing the available water quality information and
gaining an awareness of the limitations in
manpower, the Operations Committee decided
to more narrowly focus its attention to sub-
drainages within the Sulphur Creek drain,
Drainage Improvement Districts 5 and 18 were
selected (Figure 1).
In the initial selection of an area, several
strategies were considered. For instance, the
Committee could have selected an area farther
upstream, such that improvements in sediment
load to the Yakima River would be more visibly
seen. This strategy was appealing, because it
was felt that if individuals could see visible
improvement in river water quality there would
be more of a tendency to gather behind and
support the program. However, as time pro-
gressed, it became apparent that the Operations
Committee made the right choice. With Federal
regulations pending, it was obvious that it
would eventually be necessary to document the
results of the project in terms of water quality
data. The previous study in the Sulphur Creek
area by CH2M-Hill provided most of the needed
background data. Monitoring programs were to
be developed to follow up previous work and so
that the results of this type of program could be
adequately assessed.
When Sulphur Creek was identified as the
target area, work by the SCS technical staff
began. There were no published soil surveys
available at that time, so that the staff went to
work developing a "mini" survey for the area.
Soil series were identified, descriptions ob-
tained, and aerial photographs accumulated.
These, along with other pertinent and up-dated
sections of technical handbooks containing
systems specifications, irrigation guidelines,
and minimum conservation practices, were
gathered to formulate a planner's handbook,
specific to Sulphur Creek.
The SCS technical staff consisted of four
people of varying experience. In spite of their
technical experience, there was a need to
become orientated with the irrigation techni-
ques of the area and the topography of Sulphur
Creek. A week-long training session was
organized by the SCS personnel for Sulphur
Creek staff and for any interested irrigation
district personnel. Representatives from both
major irrigation districts attended (Sunnyside
Valley Irrigation District and the Roza Irriga-
tion District). The session covered such items as
theory and computer capabilities for soil intake
values and irrigation efficiency. Field evalua-
tions were made of rill systems, including
analysis of soil moisture and intake rates.
312
-------
SULPHUR CREEK PROJECT
Information and Education Program
It was recognized that if a program aimed at
changing on-farm management practices was
to succeed, it would be necessary to accomplish
two tasks: 1) Raise the level of general aware-
ness in the farming community as to the
technical and financial opportunities available;
and 2) Inform the farming community of the
goals and framework of the project. Virtually all
types of media were utilized. Although a
planned and corrdinated approach was needed,
efforts to develop such a program never succeed-
ed.
Prior to SCS staff actually contacting any
farmers, a color brochure was developed. Input
into the contents of the brochure was received
from the State Technical Advisory Committee
(TAG) and the Operations Committee. Specific
effort was made to insure that the message to
the farmers was clear and to the point.
Initial intentions were to have a valley-wide
approach and to mail the brochure to all irriga-
tion district members in the Yakima Valley.
Valley-wide coverage of the project was in-
evitable, due to the major newspaper distribu-
tion and TV/radio transmission areas.
However, the attempt to cover the entire valley
with limited resources lead to the initial mistake
of not giving the project a consistent image and
name. For example, the initial project was titled
"The Yakima Basin Farm Plan Program". The
title of the brochure, "The Yakima Basin Farm
Plan," was slightly different. When emphasis
shifted to the Sulphur Creek area, the program
then became "The Sulphur Creek Pilot Project",
"The Sulphur Creek Project", etc. In this transi-
tion, the basic goal of name familiarity was lost.
By the time the color brochure was ready to
mailing, it was decided that the initial informa-
tion effort should be geared more intensively to
the target area — Sulphur Creek. Time con-
straints forced limitation of the initial mailing
to operators in Drainage Improvement Districts
(DID). The Roza and Sunnyside Valley irriga-
tion districts agreed to mail the brochure in a
special mailing with a conservation district
cover letter to these people. Controversy arose
over the wording of the cover letter and, conse-
quently, only partial cooperation with the the
irrigation district was achieved. Nevertheless,
the general mailing was made to DID 5 and 18
farmers, with agreement of the irrigations dis-
tricts that a future mailing of the brochure
would be made valley-wide with irrigation dis-
trict year-end water statements.
Prior to the time of the first mailing of the
brochure, it was apparent that farmers were not
going to "break the door down" to get involved
in the program. Consequently, the conservation
district cover letter was written to invite
operators to a public meeting at a local Grange
Hall in August of 1975. Members of the
Operations Committee spoke at the meeting and
reaction appeared favorable. Something above
20 percent of the 200 operators in DID 5 and 18
attended this session. No specific yes/no re-
quests were made during the session in terms of
cooperating with the project.
As the project was undertaken, several
news releases were made through the local news
media (newspapers, radio, and TV). Spinoff
coverage of the project developed through the
existing controversy over the NPDES program
and through visits to the area by EPA per-
sonnel.
Adding to this type of news coverage, news
articles were developed by the SCS personnel on
the project as one of their job requirements.
Timely and well-written articles were developed
and received favorable coverage by farm jour-
nals.
In preparation for public meetings and
presentations, a slide show was prepared. This
show, with minor modifications, is still being
utilized to publicize the project.
Another mechanism used for general infor-
mation was the development of a poster. The
theme presented was "Are You Losing More
Than You're Using." The title was accompanied
by a two-color sketch of an eroded field.
Over the first year of the project, there has
been an effort to promote tours in the Sulphur
Creek area so that public awareness might be
increased. Guided bus and car excursions
around the basin with conservation district
representatives, irrigation district people, ad-
ministrative (EPA) personnel, and farmers has
helped to show the effect of rather simple
conservation practices on water quality.
Unfortunately, the level of information/ed-
ucation never became anything more than a
general awareness program. In other words, the
more or less intensive information and educa-
tion effort was never a reality. It should be
noted, also, that almost the entire effort was
313
-------
CASE STUDY: YAKIMA VALLEY
informational in nature and not educational.
For a variety of reasons, educational produc-
tions were limited. It has been apparent that
awareness of the program does not alone com-
mit people to it.
The problems in the information education
effort were seen early in the project. Respon-
sibility for the program was not defined even
though efforts were made by DOE and SCS to
divide and more clearly define responsibilities.
At the onset of the program, an information and
education schedule was developed for discus-
sion, but this schedule contained no financial
commitments or constraints. Consequently, in
the first crucial months, information attempts
were following an unrefined schedule with un-
defined financial needs or abilities.
Six months after the Sulphur Creek crew-
was on the ground, several alternatives to the
information education levels were developed
with budgets. It was DOE's responsibility to
find the funds for the effort. One such proposal,
having a budget of $20,000 for the remaining 18
months of the project, was selected. Neverthe-
less, funds of this magnitude were never located.
In February' of 1976. $5,000 was located in
the SCS contract funds and a program was
developed around that figure. Posters, slide
shows, and educational brochures were listed as
major expenditures. Nevertheless, without the
information and educational responsibility
clearly located, this program has only been
partially carried out. Additionally, it should be
noted that the intensive effort which was
planned for the first crucial months of the
project was lost and not recoverable.
Farmer Contact and Farm Management
Agreements
Following formation of the Operations
Committee, orientation of the technical staff,
selection of Sulphur Creek, and initiation of the
information education effort, work began in
soliciting farmer cooperation in the program.
As an initial step, several farmers with fair to
good management practices were enlisted to
work out the details of the farm management
process. Specifics on the contents of the
agreements were developed, along with details
of a signature sheet. The signature sheet is
signed by all parties to the agreement and each
signature prefaced by a statement of intent.
As mentioned, each agreement follows
closely the framework of the traditional farm
plan, with the exception that consideration is
given to improvement in water quality as well
as soil and cropping improvements. A list of the
information contained in a typical farm
management agreement is presented below:
1. The conservation district cooperative
agreement, which sets forth, in general
terms, what the district will do or provide
and what the farmer will do.
2. A map showing delineation of the soils
within the entire farm unit.
3. A map showing field boundaries, land
uses, access routes, headquarters, struc-
tural facilities (mechanical and electrical
facilities used in the farming operation),
and identification of all locations within
each irrigation district unit of water
delivery and return flow discharge
point(s).
4. The soil description and interpretation
upon which land use and treatment
decisions will be made.
5. Tabulation of identified problems and ex-
isting soil and water management prac-
tices by irrigation unit.
6. Alternative vegetative, structural, or
management practices specific to appli-
cable cropping systems that will provide
acceptable improvement of return flows.
7. Tabulation of decisions by farmer or by
farm manager and SCS technician which
relate to the above alternatives of specific
identified practices (vegetative, manager-
ial, or structural).
8. A statement to the effect that "Farm
management agreements developed under
this cooperative program are for the sole
use of the landowner/operator, the conser-
vation district, and DOE for the exclusive
purpose of evaluating and implementing
irrigation return flow improvements."
9. Review and approval by the appropriate
conservation district as to knowledge and
practicability of the farm management
agreement and its contents.
10. A record showing contacts and supple-
mental data gathered during the planning
and implementation period.
The above elements were included in a set of
guidelines developed for the project staff.
Originally, these were to be developed for the
purposes of inclusion into regulations, since it
314
-------
SULPHUR CREEK PROJECT
was anticipated that the project would be ex-
panded on a statewide level. This concept was
premature, and the guidelines have remained as
a working document only.
Following the first public meeting with the
farmers of DID 5 and 18, a list of "significant
contributors" was developed. This list was
developed principally through the cooperation
of the irrigation districts and was intended to
focus attention on that minority of farmers who
apparently had land and water management
problems. Approximately 50 farmers were on
this list.
The Operation Committee recognized that a
significant problem existed on leased land. In
some cases, farmers utilizing land owned by
others were not concerned with the conserva-
tion of the soil resource or water quality. The
problem of leased land had to be addressed. It
was decided that attempts had to be made to
identify the landowner and leasee. In some
cases, this was possible through conservation
district and irrigation district records, but in
many cases the lessee could not be identified
readily.
An early attempt was made by DOE
regional personnel to contact some of the 50
farmers in person. The demand on time and
manpower made this attempt infeasible. In
addition, the level of cooperation achievable
through sole DOE contact was questionable.
The Operations Committee decided that the
most promising route to follow was to hold a
meeting specifically for the 50 listed farmers.
An invitation letter was mailed through the
local conservation district calling for a meeting
in late January 1976.
Response was good. There were 22 of the 50
who came to the session chaired by the South
Yakima Conservation District (on behalf of the
Operations Committee). The slide show was
presented and remarks explaining the program
were given by various members of the
Operations Committee. Following a request by
the Operations Committee, 100 percent of those
attending requested farm evaluations. This was
accomplished through distribution of request
forms requiring a definite yes-no commitment.
A second similar meeting, preceded with
invitational letters, was called the following
month. Ten of the remaining 28 farmers
appeared, and this session again resulted in 100
per ent agreement to have their farm units
evaluated. At this time, the point of diminishing
returns had been reached and no further
sessions were scheduled.
During these early sessions, it was recog-
nized that there were probably some "falsely
accused" on the list of 50. Following farm
evaluation, this proved to be true and several of
the 50 have been given assurances that they are
utilizing good land and water management
techniques.
Financial Programs
As an incentive to getting farmer involve-
ment, the most readily available source of
funding was explored — the USDA Agricultural
Stabilization and Conservation Service pro-
gram of cost sharing (ASCS). This program
provides cost sharing for up to 50 percent of the
cost of specified conservation practices. In this
program, cost sharing can be provided through
long-term agreements with ASCS for up to 10
years on any farm unit with a financial limita-
tion of $25,000 (ASCS funds) on any particular
farm.
Background and objectives of the Sulphur
Creek Project were discussed with the Yakima
and Benton County ASCS Committees. It was
necessary to bring the Benton County ASCS
Committee into the project because small por-
tions of DID 5 and 18 lie in that county. These
committees, made up of conservation district
board members, direct the use of funds. Alter-
native proposed measures and possible costs to
operators were explained to the committees. A
special project was recommended for funding
the eligible measures under the ASCS program.
Under the suggested proposal, funds would be
earmarked mainly for use in long-term
agreements (LTA). It was felt that LTA would
most ideally blend with the farm management
agreement program.
The ASCS county committees and Sulphur
Creek technical staff developed two special
project applications for funding from the state
ASCS funds. Again, it was necessary to have
one located in Yakima County and one in
Benton County. These proposals resulted in
funding totaling $80,000 — $10,000 in Benton
County and $70,000 in Yakima County.
The Yakima County committee set no
priorities for LTA funds. Thus, the option for
utilization of LTA as a long-range planning tool
was nonexistent for Yakima County land-
owners. The Benton County committee did give
315
-------
CASE STUDY: YAKIMA VALLEY
farmers the option of LTA and several requests
were received as a result.
ASCS cost-sharing requests resulted in the
following planned improvements in Sulphur
Creek area: (Because of the spinoff publicity
from the activity in DID 5 and 18, other requests
came in from the Sulphur Creek area in general:
the following list represents those planned
improvements in the total Sulphur Creek
drainage.)
Number of
Units
Planned Improvements
31 Piping open ditches for more efficient
water control
12 Excess water management system im-
provements
32 Conversions from rill to sprinkler
2 Livestock storage and disposal systems
11 Desilting and debris basin tail water re-
covery svstems
During the first year of operation, several
resource conservations and development pro-
jects (RC & D) have evolved. RC & D funds are
provided through the USDA, with the SCS
providing the leadership in the program.
Though the SCS provides leadership, RC & D
projects are developed at a local level, with
locally developed goals, and basically, locally
run. RC & D projects differ from projects under-
taken with ASCS funds in many ways, but
primarily in the area of size. RC & D projects are
community in nature and may help in supply-
ing funds for such things as flood prevention,
erosion and sediment control, agricultural
water management, agricultural related pollu-
tion control, or fish, wildlife, and recreation
development.
For example, in Sulphur Creek one applica-
tion has been made to improve drainage outlets,
and another is forthcoming on a group gravity
delivery' system to provide pressure delivery to
about 700 acres. Pressure will be sufficient for
sprinkler operation, doing away with numerous
pumps.
Manpower seems to be the major constraint
on the number of RC & D projects which are
undertaken. It takes considerable effort to make
all landowners aware of the program and to
gain their cooperation in these types of projects.
It is felt that if sufficient time were spent with
the involved landowners in the Sulphur Creek
area, there are a large number of RC & D
projects which might be generated.
Project Outputs
Following the first year of operation of the
Sulphur Creek Project, the following results
have been obtained. In DID 5 and 18 and of the
original 50 "problem" farm units, 5 agreements
have been signed and another 22 are in the
planning process. In the Sulphur Creek area,
and outside of the significant 50, another 6
agreements have been signed and 9 are in the
planning stage. Of the 50 farmers intensively
contacted in DID 5 and 18, only 5 have shown
some resistance to the Sulphur Creek approach.
During the 1977 irrigation season, it is
anticipated that 23 plans will be signed in the
Sulphur Creek area. Of these, 13 will come from
that original list of 50 farmers. In addition, one-
third (8) of these plans will cover farm units of
over 1,000 acres.
Technical Problems and Economic
Considerations
As mentioned, one of the first tasks of the
technical staff as they entered the Sulphur
Creek area was to develop some of the technical
tools necessary for soil and water conservation
planning. Along with conducting "mini" soil
surveys of the area and gathering aerial
photographs, the technical team set out to
update what is commonly referred to as the
"Technical Guide" for SCS planners. The
Technical Guide sets out minimum conserva-
tion treatments required in specific situations to
maintain the soil resource. The above, along
with applicable portions of the SCS publication,
"Irrigation Guide — Benton, Kittitas, Walla
Walla and Yakima Counties'" (1974), were in-
corporated into a planner's handbook specific
for Sulphur Creek.
As planning progressed during the first
year of the project, it became apparent that
some of the information had to be constantly
updated, based on experiences specific to Sul-
phur Creek. For instance, the "Irrigation
Guide" lists maximum criteria for irrigation
including that for nonerosive streams, intake
rates, row spacing by crop, peak and monthly
consumptive use, maximum length of run
(based on soil and slope), net moisture replace-
ment (based on soil), and gross application (-
based on slope and irrigation efficiency). In the
316
-------
SULPHUR CREEK PROJECT
Sulphur Creek area, it was found that the
"Irrigation Guide" contained some errors. Some
soil intake rates, for instance, varied ten-fold
from those in the Irrigation Guide. The effect of
errors of this magnitude are apparent — low
intake rates will affect length of run, initial
stream size, erosive stream size, advance rates,
irrigation time, etc. Consequently, new realistic
intake rates had to be found to reflect the
situation in Sulphur Creek.
Another significant problem found in the
Sulphur Creek area was erosion in tail water
systems. Many irrigators were found to be
properly managing water at the head end of the
fields. However, because in many cases 50-100
small flows would accumulate in collection
ditches, erosive streams would be developed.
This problem was solved on several units
through piping collection ditches and utiliza-
tion of dropped inlets. Dropped inlets are ver-
tical pipes which intersect the main collection
pipe and are generally spaced at about 50-60 foot
intervals. Sediment was measured leaving one
field of wheat and found to be reduced by eight-
fold with this type of system. The first irrigation
following tillage generally causes the largest
sediment load to return systems. This can be
reduced or, in many cases, eliminated through
the removal of a few inches of soil around the
dropped inlet pipes — thus creating small
sediment basins.
During the first year of the Sulphur Creek
Project, many of the solutions recommended to
the irrigators were simple and relatively inex-
pensive. For instance, the dropped inlet collec-
tion system described above would cost ap-
proximately $1,500 for a 40-acre field, with
dropped inlets spaced every 50 feet ($37.50 per
acre). The cost of creating small sediment
basins around the dropped inlets would also
need to be added, but would be a minimal yearly
cost, probably done by wheel tractor/blade or by
hand labor.
This type of collection system coupled with
a sediment basin and repump system would cost
approximately $5,500 per 40-acre block (or
$137.50 per acre). Conversion to sprinkler
systems on a particular unit would average $250
to $800 per acre, depending upon the complexity
of the system and the type of water delivery.
Any particular farm might require a com-
bination of practices to reduce sediment losses.
For example, a field that is currently well
irrigated, utilizing a long run (1500 to 1700 feet)
and has a sediment-loss problem, might need
one or more of the following:
Items
Cost Per Acre
$50 per acre
$37.50 per acre
Splitting the run in half
Tailwater collection system
(dropped inlet)
Sediment basin installation and $12.50 per acre
maintenance
The total cost for this particular unit would
run approximately $100 per acre. Addition of a
repump system to the sediment basin installa-
tion would raise the total capital investment to
$225 per acre. When the costs of return flow
clean-up under rill irrigation get this high, then
sprinkler systems, with lower labor cost, begin
to become more feasible.
Many farmers are already operating at a
good management level. However, 5 to 10 per-
cent are not, and the result has been a cost to
their downstream neighbors and, of course, in
terms of overall water quality in the basin.
Taking the average cost between the two
systems mentioned above ($100 with sediment
basin and $225 with sediment basin and
repump) — $160 per acre and projected over 10
percent (problem units) of the 247,000 acres of
rill irrigated land in Yakima County (24,700
acres), it would yield a total projected cost of
approximately $4 million to carry out these
practices. This figure, of course, is an estimate
not considering the size of those operations
which seem to have a problem. It has been the
experience in the Sulphur Creek area that those
farming larger operations tend to have more of a
problem. Additionally, annual operation and
maintenance costs have not been considered.
It has been found that utilizing a rather
simple evaluation process, the conservation
planner can pinpoint problems on an irrigated
unit and make meaningful recommendations to
aid in clean-up of return flows. From an overall
standpoint, and from the results of Sulphur
Creek, most recommendations include one or
more of the following: Improved water manage-
ment, shortening furrow lengths, reducing
steepness of slopes through leveling, contour
ditching, grassed waterways, filter strips, sedi-
ment basins, repump systems, or sprinklers on
steeper more erosive soils. It is felt that the
Sulphur Creek evaluation procedure will yield
the proper set of alternatives for a particular
317
-------
CASE STUDY: YAKIMA VALLEY
field. They allow the operator to make an
informed and intelligent decision on how to get
the most out of his farm without undue damage
to the resource base.
SUMMARY
Spawned by futile efforts to apply the
NPDES permit system to agriculture, the Sul-
phur Creek project has clearly pointed to alter-
natives as a result of its first year of operation.
Attitude changes, due to information and educa-
tion, local input and control, and realistic
regulations, are the key elements to minimizing
pollutant loadings from irrigated lands. Any
number of alternate programs would be de-
signed containing these elements.
The three controversial issues apparent in
1973 have not yet been resolved: 1) Who is to
receive the NPDES permit (should there be a
permit system at all)? 2) Do those entities have
the responsibility and or ability to control
waste from irrigated lands? and 3) Does EPA
have authority to exempt certain categories of
dischargers from the permit system?
From the administrative and technical
standpoints, many common problems have
developed during the past year. Some of these
problems and solutions include:
1. A clear definition as to who had authori-
ty, control, and responsibility of the project
should have been developed early in the
project. A good deal of confusion and futile
effort can develop in any project when the
above factors are not clearly given to a
project leader. In Sulphur Creek, almost six
months had elapsed before a clear assign-
ment of authority, control, and responsibili-
ty was obtained.
2. Information, and more importantly,
education, should be given a high priority.
Individuals cannot be expected to change
their manner of operation without being
influenced by something. It is felt that
negative incentives such as threat of
penalties should be reserved for extreme
abuses of land and water resources. In a
project such as Sulphur Creek, it is felt that
education should be emphasized as a
mechanism to bring about change and fund-
ing committed to this objective. Only
through education will the farming com-
munity as a whole recognize that the objec-
tives of the improved soil and water
management, improved crop production,
and improved water quality are compatible.
3. More emphasis should be given to the
current level of technology. That is, there is
some question of the ability of ad-
ministrative agencies to either predict the
specific results of any massive change in
management practices or predict the
economic impact on the farming communi-
ty. Is our technical knowledge at a level such
that accurate predictions can be made as to
the interrelationships between tillage prac-
tice, water management, crop rotations, and
economics? Can these predictions be made
for any area, let alone a relatively large area
such as a river basin? Experience in Sulphur
Creek has indicated that we have a lot to
learn. Also, it tells us that any requirements
applied to a hydrogeographic area will have
to be rather specific and aimed only at those
gross abusers of land and water resources.
At that general level, accurate predictions
as to water quality improvement cannot be
made at this time. Only through intensive
intergovernmental cooperation, combina-
tion of resources, research, and clearly
defined agency goals will predictions of this
nature ever by accurately made.
4. The objectives and scope of any program
should be clearly defined and remain un-
altered during the life of the project. In
Sulphur Creek, the information program
was initiated to bring out voluntary im-
provement in management practices.
However, the objectives and scope of the
project changed in its initial year, giving
rise to some unresolvable problems. An
example is that the general publicity given
to the project as it was originally designed
(headlines such as "Farmers Told by-
Ecology Engineer to Clean Up Water")
provided improvement on lands outside the
target area. However, when the
USGS DOE/WSU monitoring programs
began, the objectives became research in
nature (as opposed to general action
program). Control areas for a research pro-
ject were not not available due to this "spin-
off activity." Additionally, detailed outputs
for the project were never prepared (it was
designed as a general action program). It
seems, however, that the success or failure
of Sulphur Creek from an administrative
standpoint may be based on the number of
318
-------
SULPHUR CREEK PROJECT
agreements completed, the number of
farmer contacts, or a clear definition of a
"significant contributor." Perhaps these
should have been objectives from the begin-
ning.
In spite of these problems, Sulphur Creek
can be termed a success after one year of
operation. Although not documented in detail, it
is felt that the levels of concern and awareness
of agriculturally related water quality problems
has risen in the agricultural community. For a
voluntary program, cooperation has been
satisfactory. It is known that agricultural
management practices are improving because
of the program. However, the resulting level of
improvement of water quality may not be
known until the results of monitoring programs
have been thoroughly analyzed. This analysis
will necessarily include considerations of farm
economics and weather during the monitoring
period. Perhaps, a reliable correlation between
the project and water quality may never be
found. Probably the largest successes of the
program have been the recognition of the value
of local control and input into governmental
programs and the significant accomplishments
that can be made through the combination of
objectives of various agencies and organiza-
tions toward a single goal — a realistic ap-
proach toward a quality environment.
319
-------
The "208" Planning Effort
for Irrigated Agriculture
in the State of Washington
MARC HORTON and JOHN SPENCER
Department of Ecology;
State of Washington; Olympia, Washington
ABSTRACT
The Department of Ecology is reponsible for
developing a "208" Plan for irrigated agricul-
ture in the State of Washington. It is anticipat-
ed that Best Management Practices (BMP) will
be utilized by individuals to come into com-
pliance with a state-wide regulatory program
(possibly NPDES). Water quality committees
formed through local conservation districts,
with representatives from all interested groups,
will develop BMP based on the following
criteria: (1) economic feasibility; (2) local accep-
tability; and (3) water quality improvement.
INTRODUCTION
As directed by Section 208 of P.L. 92-500
and initial regulations, three urban drainage
areas of Washington were designated as area-
wide waste treatment planning areas. They are
portions of King, Snohomish, and Clark coun-
ties. Comprehensive planning has been un-
derway in these areas for some time.
Through Federal court direction (NRDC
filed suit based on the urban-industrial limita-
tion of initial regulations), DOE is active in
"208" planning for those areas of the State not
designated under initial regulations. Because of
funding limitations, the Department of Ecology
(DOE) has opted for a program which will
emphasize "non-point" source problems in
specific areas of the State. Hopefully, solutions
to these problems developed during the "208"
process can be extrapolated to other areas of the
State having similar problems. The Federal
courts have given states until November 1978 to
complete their 208 planning.
The major elements of Washington's ap-
proach are as follows:
Problem
Responsible Planning
Entity
Urban-industrial non-point
sources
Irrig ated-agricultur al
sources
Dryland agricultural
sources
Dairy waste
Groundwater disposal
Silviculture
Implementation
assurances
Designated areas
DOE and State
Conservation
Commission
State Conservation
Commission
State Conservation
Commission
DOE
DOE
Consultant
Irrigated Agriculture
Comments here will be restricted to the
irrigated-agricultural element. The final
product of the irrigated-agricultural element
will be a result of local planning, not by
agencies, but by local citizens. Basically, there
will be two objectives: (1) to develop Best
Management Practices (BMP) for on-farm
management improvements and (2) to develop a
program to address the NPDES regulations.
It is anticipated that following the "208"
process, Best Management Practices (BMP) will
be utilized by individuals to come into com-
pliance with a state-wide regulatory program
(possibly NPDES).
Major activity in the irrigated-agricultural
element will occur at the local level. Conserva-
tion districts in irrigated areas are being urged
to accept the responsibility for organization of
the local input. At this time, it appears that
water quality committees will be formed
321
-------
CASE STUDY: YAKIMA VALLEY
through the conservation districts with
representation from all interested groups. The
principal function of these water quality com-
mittees will be to develop a list of BMP. The
criteria for judging any particular management
practice will be three-fold:
1. It must be economically feasible.
2. It must be locally acceptable.
3. It must improve water quality.
Washington State Cooperative Extension
Service, the SCS, the Conservation Com-
mission's irrigated-agricultural specialists, and
DOE personnel will aid these local groups in
organizational and technical matters.
Basin-wide work groups will also be formed
in each of the two major irrigated areas in
Washington — the Yakima Basin and the
Columbia Basin (Figure 1). These groups will
have the principal function of assimilating
information provided to them from the local
committees. They will look for similarities in
BMP from various areas and refine the develop-
ing list of BMP. Membership in this group will
include representation from all of the local
groups, irrigation districts, grower groups, and
other interested entities.
Additionally, this group will have primary
responsibility for development of a program to
meet the objectives of existing NPDES
regulations. Current regulations call for
issuance of area-wide permits in 1978. Also, they
provide for development of a State program
which will meet or exceed the objectives of the
proposed area-wide permits.
Permit proposals and refined lists of BMP
will be passed on to a state-wide advisory
committee for irrigated agriculture. This group
has a similar but expanded membership from
that of the former State Technical Advisory
Committee for Water Quality Improvement.
The tasks of this particular group will be to:
(1) provide overall guidance to the process, (2)
review permit proposals and BMP, and (3) guide
the development of pilot projects, such as the
one in Sulphur Creek.
The above process is not stepwise in nature,
but rather is intended to be a dynamic in-
terchange of ideas and information from the
local to State level.
To assure local input from both major
basins and from other irrigated areas, several
additional items have been designed into the
Conservation District Activity —
*
Information to and from Local
People
Water Quality Committees
Workshops
Newsletters
BMP
Basin Work Groups- Yakima and Columbia
Basin Projects
'Assimilate Information, Determine
Similarities in BMPs.Aid in NPDES
or Equivalent Program Development
Refined BMP
Proposed NPDES /Equivalent
Program
State Advisory Committee for
208 Irrigation Element
* Pilot Project Development,
Overall Guidance, Planning
Review
State 206 Policy
Advisory Committee
*208 Plan
Figure 1. Washington's 208 planning process for
irrigated agriculture.
program. For instance: (1) creation of a large-
scale information/education program, (2) an
information dissemination and feedback
system utilizing periodic newsletters, etc., (3) a
series of workshops (1977) and hearings (1978)
on the refined BMP list and NPDES proposals,
respectively.
Information from the Sulphur Creek project
will be utilized in the 208 process to aid in the
development of BMP and permit proposals.
Because of the anticipated value of Sulphur
Creek information, the project may be extended
to November 1978, the proposed termination of
the 208 process. Additionally, other pilot pro-
jects are under consideration currently and may
be undertaken in the next irrigation season
(1977).
In the Yakima Basin, long-range planning
has concurrently been conducted by three
federal agencies — the Bureau of Reclamation,
the Corps of Engineers, and the Soil Conserva-
tion Service. The DOE has entered into an
agreement with the three agencies for the pur-
322
-------
"208" PLANNING EFFORT
pose of avoiding duplication and providing a should be noted that some interim outputs will
free exchange of information. Because of this have to be used, because planning by the Bureau
agreement, outputs from the planning of these of Reclamation and Corps of Engineers will
agencies will be incorporated, as appropriate, continue far beyond November 1978 and on into
into the final Washington State 208 Plan. It 1980 and 1981.
323
-------
The 1973 Agreement on
Colorado River Salinity
Between the United States
and Mexico
M. B. HOLBURT
Colorado River Board of California,
Los Angeles, California
ABSTRACT
The background of the conflict between the
United States and Mexico over the salinity of
the Colorado River water delivered to Mexico,
the early attempts to resolve the conflict, the
meetings and negotiations that led to the latest
agreements, a description of the agreement, the
legislation that implements that agreement,
and necessary future actions are presented.
In August 1973, the United States and
Mexico executed the latest in a series of
agreements that have attempted to settle the
dispute between the two countries concerning
the salinity of the Colorado River. The agree-
ment is called the "Permanent and Definitive
Solution to the International Problem of the
Salinity of the Colorado River". In June 1974,
the Colorado River Basin Salinity Control Act
was enacted which authorized the construction
of facilities considered necessary to implement
the agreement and, in addition, authorized a
major basin wide salinity control program. This
paper discusses the background of the conflict
between the United States and Mexico over the
salinity of Colorado River water delivered to
Mexico, the early attempts to resolve the con-
flict, the meetings and negotiations that led to
the latest agreements, a description of the
agreement, the legislation that implements the
agreement, and necessary future actions.
Mexico's Use of Colorado River Water
The Colorado River is extremely important
to northwestern Mexico. Colorado River water,
supplemented by groundwater supplies, has
irrigated between 425,000 to 500,000 acres in
Mexicali and San Luis Valleys. For many years,
the dominant crop was cotton. In 1959, 84% of
the irrigated acreage was planted to cotton. In
recent years, other crops have replaced cotton,
as indicated by the following table:
IRRIGATED ACREAGE IN MEXICALI
AND SAN LUIS VALLEYS IN ACRES
Crop
1970 1975
Wheat and Barley
Cotton
Miscellaneous
Alfalfa
Safflower
172,140 151,764
145,273 107,713
51,996 124,482
36,452 46,393
20,739 39,654
Total
426,600 470,006
The city of Mexicali, a city of about 400,000
people, obtains its basic supplies of 32,000 acre-
feet a year from the Colorado River. Since 1972,
the city of Tijuana, with a population of about
500,000 persons, has been receiving an average
of 8,000 acre-feet a year of Colorado River water
under a temporary agreement with several
California agencies while an aqueduct was
being planned and constructed. The aqueduct,
scheduled for initial operation in late 1977, will
enable Mexico to deliver 100,000 acre-feet a year
of Colorado River water to Tijuana.
The Mexican Water Treaty
In February 1944, the United States and
Mexico signed a treaty concerning utilization of
the Colorado River as well as the Tijuana River
and the Rio Grande. The most important provi-
sion of the treaty, with respect to the Colorado
River, is the allotment to Mexico of a guaran-
teed annual quantity of 1.5 million acre-feet a
325
-------
CASE STUDY: WELLTON-MOHAWK DISTRICT
year (maf yr). The treaty was a result of years
of negotiations and one of the contentious is-
sues was salinity.
California made a major issue of water
quality during the Senate hearings. The State
Department defended the large allotment to
Mexico by stating that up to 750,000 af yr would
come from irrigation return flows below Im-
perial Dam and would probably go to Mexico
irrespective of any treaty. Senator Downey of
California questioned the usability of the sup-
ply going to Mexico if the State Department
estimates of return irrigation flows were correct.
He made a remarkably accurate prediction in
1945 by stating that because of the ambiguity in
the Treaty concerning water quality, Mexico
would come back in 25 or 30 years and demand
better quality water. State Department
representatives, their consultants, and Senate
supporters denied that there was any ambiguity
in the Treaty. They stated that water quality
was extensively discussed, and that Mexico
fully understood that the Treaty required them
to take irrigation return flows irrespective of the
salinity of those return flows. In response to a
question at the hearings before the Senate
Foreign Relations Committee, the State Depart-
ment consultant said the United States could
deliver water to Mexico under the Treaty, as
much as 500,000 to 750.000 af yr, even if it
would not have any value for irrigation pur-
poses. The State Department pointed out that
the specific provisions that were included in the
Treaty to insure that Mexico must accept return
flow and drainage water are in Articles 10 and
11. Article 10 states that Mexico's allotment
included water from "any and all sources" and
would be "for any purpose whatsoever". Article
11 states that "waters shall be made up of the
waters of the said river whatever their origin".
In April, 1945. the Senate ratified the Treaty
by a vote of 76 to 10. Of the 14 Basin state
senators, only the two California senators and
one Nevada senator voted against the Treaty.
In testimony before the Mexican Senate, the
Mexican negotiators were telling a different
story from that told by their United States
counterparts. One of the Mexican negotiators
said it was understood between the two coun-
tries that the water delivered to Mexico must be
of good quality. He stated that Mexico could
demand water similar in quality to that which
she was currently using. Recognizing the dif-
ficulty of this demand, he said that Mexico
would not object to receiving water similar to
that used by the United States at Imperial Dam,
the last diversion point in the United States.
The Mexican Senate unanimously ratified the
Treaty in September 1944.
Drainage from Arizona's
Wellton-Mohawk Project
Between 1945 and 1961, there were no major
problems with respect to the river, as the
salinity of the water delivered to Mexico at the
Northerly International Boundary was general-
ly within 100 parts per million (ppm) of the
water at Imperial Dam, the last major diversion
for users in the United States. In 1961, the
Wellton-Mohawk Irrigation and Drainage Dis-
trict in Southwestern Arizona commenced
operation of a system of drainage wells which
discharged saline water into the Colorado River
below the last United States diversion, but
above the Mexican diversion. The drainage
water included a substantial proportion of high-
ly saline groundwater that had been concen-
trated through reuse during the previous half-
century. Initially, it had a salinity of around
6,000 ppm. As a sharp reduction in river flows to
Mexico occurred at the same time, the combined
impact of the Wellton-Mohawk drainage water
and reduction of dilution wrater was a sharp
increase in the salinity of the water delivered to
Mexico, from an average of around 800 ppm in
1960 to plus 1500 ppm in 1962. Mexico raised
strenuous objections to receiving the drainage
waters.
Although, as previously indicated, the Unit-
ed States intended that Mexico must receive
return flows below Imperial Dam under the
Treaty, it had not been anticipated that there
would be return flows as high in salinity as the
Wellton-Mohawk drainage or that there would
be such a precipitous rise in the salinity of the
waters delivered to Mexico. Consequently, after
the winter of 1961-62, the United States under-
took certain provisional measures to minimize
the impact of the high salinity drainage returns
from Wellton-Mohawk. The United States also
entered into negotiations with Mexico to arrive
at a practical solution. The State Department
asked each of the governors of the seven
Colorado River Basin states to appoint two
members to a reconstituted Committee of Four-
teen in order to advise the State Department in
connection with the salinity problem. The Direc-
tor of the California Department of Water
326
-------
SALINITY AGREEMENTS
Resources and myself are currently the two
California members of the Committee.
In the past several years, approximately
500,000 af/yr of the 1,500,000 af/yr guaranteed
to Mexico has come from drainage water below
Imperial Dam. Of this amount, the Wellton-
Mohawk Project contributes about 220,000 af.
Interim Salinity Agreements
Minute No. 218
Extensive negotiations were conducted
between 1962 and 1965 and, in November 1965,
a five-year agreement was incorporated in
Minute No. 218. (The "Minute" form is a record
of the International Boundary and Water Com-
mission, United States and Mexico.) Under
Minute No. 218, the United States undertook the
following actions at a cost of $12 million:
(1) Constructed an extension of the
Wellton-Mohawk Drain so that drainage water
could either be bypassed around Morelos Dam
(the Mexican Diversion Point) or mixed with
other Colorado River waters above Morelos
Dam, at the option of Mexico.
(2) Constructed additional drainage wells
in the Wellton-Mohawk Project which allowed
selective pumping of the most saline waters at
times when Mexico would be bypassing
Wellton-Mohawk drainage water, and allowed
the pumping of higher quality groundwater at
times when Mexico would be using Wellton-
Mohawk water.
(3) Replaced a portion of the bypassed
Wellton-Mohawk water which resulted in the
release of approximately 40,000 acre-feet of
mainstream water per year from Imperial Dam
in excess of the 1.5 million acre-feet per year
guaranteed by the Treaty.
Under the measures taken by the United
States, the quality of the water delivered to Mex-
ico was improved from average annual values
of about 1500 ppm in 1962 to 1240 ppm in 1971.
Minute No. 218 was entered into for a specific
period and was to expire in November 1970. Ac-
cordingly, the United States and Mexico com-
menced negotiations with the purposes of arriv-
ing at another five-year agreement. However,
Mexican officials did not want to enter into a
new long-term agreement in November 1970,
since a new administration was taking office in
December 1970.
Negotiations commenced in 1971 with the
new Echeverria administration. The United
States, supported by the Committee of Fourteen,
proposed a new Minute, which would have
provided to Mexico, Colorado River water hav-
ing the same salt concentration as would exist
were the Wellton-Mohawk Project and all other
projects in the United States below Imperial
Dam in salt balance. This means that the
tonnage of salt in drainage water originating
from lands below Imperial Dam in the United
States and delivered to Mexico would not exceed
the tonnage of salt in the water applied to these
lands.
Although the United States negotiators
thought they were near agreement with Mexico
in November 1971, Mexico finally rejected the
United States' proposals and negotiations were
discontinued, pending the results of a forthcom-
ing meeting between Presidents Nixon and
Echeverria. In the interim, the two countries
agreed to continue operations under Minute No.
218.
Minute No. 241
On June 15 and 16, 1972, Presidents Nixon
and Echeverria met and, following the
meetings, issued a joint communique dated
June 17, 1972. With respect to the Colorado
River, President Echeverria gave the essence of
the current Mexican position as wanting water
under the 1944 Treaty to be the same quality as
the water at Imperial Dam. President Nixon
replied that "this was a highly complex problem
and needed careful examination of all aspects".
He said that the United States was prepared to:
"(a) Undertake certain actions immediate-
ly to improve the quality of water going to
Mexico;
(b) designate a special representative to
begin work immediately to find a perme-
nent, definitive and just solution of the
problem;
(c) instruct the special representative to
submit a report to him by the end of the
year; and
(d) submit this proposal, once it has the
approval of this government, to President
Echeverria for his consideration and ap-
proval."
The immediate action referred to by the
President was formalized as Minute No. 241 of
the International Boundary and Water Com-
mission on July 14, 1972, and replaced Minute
No. 218. This agreement was based on the salt
balance concept previously advanced by the
327
-------
CASE STUDY: WELLTON-MOHAWK DISTRICT
United States and provided that the United
States discharge 118,000 af/yr of Wellton-
Mohawk drainage waters below Morelos Dam.
The United States agreed to substitute for this
water, on an interim basis, additional Colorado
River water releases and waters pumped from
wells on the Yuma Mesa. Thus, the total
deliveries to Mexico would exceed the 1,500,000
af/yr guaranteed by the Treaty. Actions by the
United States under Minute No. 241 included
the discharge of 119,490 af of Wellton-Mohawk
drainage water and the substitution by the
United States on an equal amount of water
during 1972-73. This resulted in reductions in
the average annual salinity of waters made
available to Mexico from 1242 ppm in 1971 to
1140 ppm for the year ending June 30, 1973.
Under the Minute, Mexico requested the
United States to discharge the balance of the
Wellton-Mohawk drainage waters (95,550 af),
below Morelos Dam which was charged to
Mexico's 1.5 maf/yr deliveries. As a result, the
water diverted at Morelos Dam for the year
ending June 30,1973, had an average salinity of
980 ppm, which was about 130 ppm higher than
the mean salinity of water arriving at Imperial
Dam for the same period.
"Permanent and Definitive Solution"
to International Salinity Problem
On August 16, 1972, President Nixon
designated former Attorney General Herbert
Brownell, Jr., as his special representative and
gave him the task of finding a permanent
solution to the Mexican salinity problem.
A federal task force consisting of policy and
working level representatives from a number of
the major departments of government, includ-
ing Interior, State, Agriculture, Environmental
Protection Agency and Office of Management
and Budget, was formed to assist Mr. Brownell.
The Task Force developed possible solutions,
evaluated them and presented the results to Mr.
Brownell. He also met with the Committee of
Fourteen periodically to seek their advice on
possible solutions. Mr. Brownell completed his
report on time and submitted it to the President
by December 31, 1972.
Several months after receipt of the Brownell
report, President Nixon appointed Mr. Brownell
as a special Ambassador and negotiations com-
menced between him and the Mexican represen-
tatives in the Spring of 1973. Ambassador
Brownell continued to meet and discuss the
negotiations with the Committee of Fourteen.
Although the Committee supported the thrust of
the proposed agreement with Mexico, they con-
tinued to state that the agreement would require
certain actions on the part of the Federal
Government in order to avoid damage to the
basin states. The Committee of Fourteen was
unable to receive firm assurances from responsi-
ble members of the government that actions
satisfactory to the states would be taken.
Concerns of Basin States
On July 5,1973, the Committee of Fourteen
wrote Mr. Brownell, summarizing the issues
previously raised in meetings that were still
unresolved. Senators and congressmen from the
seven basin states were also concerned that
terms of the agreement could be detrimental to
the water interests of the states. On July 20,
1973, all 14 senators and 36 congressmen from
the seven basin states signed a letter to Presi-
dent Nixon, asking that final negotiations with
Mexico not take place until there was substan-
tial agreement on the issues raised by the basin
states.
Ambassador Brownell met with the Com-
mittee of Fourteen on August 18, to discuss the
issues that had been raised by the states,
senators, and congressmen. Although agree-
ment was not reached with the states on the out-
standing issues, Brownell stated that he intend-
ed to try to reach an agreement with Mexico.
The actions requested by the basin states of
the United States, and their eventual disposi-
tion, are summarized as follows:
1. Federal government to assume respon-
sibility for permanent replacement of the reject
brine stream from desalting plant.
The Administration responded by propos-
ing that the replacement be conditional upon
augmentation of the Colorado River by 2.5
maf/yr. The legislation supported by the states,
and enacted as the Colorado River Basin Salini-
ty Control Act, Public Law 93-320 (P.L. 93-320),
declares it to be a national obligation and
directs studies on feasible methods to provide
replacement water are to be completed by June
30, 1980.
2. Simultaneous authorization of features
to implement the agreement with Mexico and
the domestic Colorado River Basin Salinity
Control Program.
The Administration disagreed with this
requested action and submitted legislation to
328
-------
SALINITY AGREEMENTS
Congress dealing only with the Mexican agree-
ment features. The States developed legislation
incorporating both features, and these were
enacted as Titles I and II of P.L. 93-320.
2. Any agreement with Mexico to not im-
pair or injure land owners in the United States.
The Administration agreed to negotiate on
this issue, and it was incorporated in P.L. 93-
320, particularly with respect to land owners
within the Wellton-Mohawk Irrigation and
Drainage District.
4. Federal government to continue research
program to increase conversion efficiency of the
Yuma Desalting Plant from 70 to 90 percent.
The Administration apparently concluded
that the need for such improved efficiency could
be delayed for several years and, in addition,
may be attainable through private industry re-
search; however, the basin states have sup-
ported appropriation of funds to continue
federal research on increasing efficiencies.
5. Administration to proceed with develop-
ment of Colorado River Basin water projects
authorized by Congress, recognizing rights of
the Basin states to develop their water
resources.
The Administration agreed to ease restric-
tions set by OMB on planning on water projects
impacting on salinity of the Colorado River.
This issue is complicated by other problems
affecting federal funding of water projects.
6. Federal government to obtain power
needed for Yuma Desalting Plant from new
sources so that power would not be withdrawn
from existing uses.
The Administration initially planned to
obtain its power needs by recapture of capacity
and energy through contract provisions of the
Parker-Davis Project System. However, P.L. 93-
320 negated this proposal by stating that
sources of power for the desalting complex shall
not diminish the supply of power to preference
customers from the Federal power systems
operated by the Secretary.
In addition to the issues as enumerated
above, the States were concerned by Mexico's
pumping of ground water along the Arizona-
Mexico border which originates within the U.S.
They urged that this issue be treated in any
agreement with Mexico so as to protect and
maintain the rights of the U.S. P.L. 93-320
included authorization of a well field in Arizona
to balance the groundwater pumping by Mexico
in the event the two countries are unable to
reach an agreement limiting groundwater
pumping.
Negotiations with Mexico
In his negotiations with Mexico, Mr.
Brownell had to consider a number of conflict-
ing views. The most significant are summarized
below:
(a) The United States agencies represented
on the federal task force had differing solutions
to the problem. For example, the State Depart-
ment desired a negotiated settlement in order to
avoid further conflict with Mexico and the
possibility of having to agree to a solution
decided by a third party, such as the World
Court. Also, there were a number of major out-
standing problems with Mexico such as drugs,
immigration, and economic policies that did not
lend themselves to satisfactory solutions. The
salinity problem was one that appeared to be
capable of solution by expenditure of money.
The Office of Management and Budget wanted
a solution that did not involve expenditure of
large sums of money.
(b) The seven basin states were anxious to
work with the Federal Government to achieve a
solution to this vexing problem. They did not
object to deliveries by the United States of water
in excess of 1,500,000 af/yr on a temporary basis
in order to have a practical solution to the
problem. However, the basin states opposed any
permanent commitment to Mexico of water
deliveries beyond that required by the Treaty.
(c) Mexico's basic position was that it was
entitled to the same quality water as is delivered
to Imperial Valley just across the border from
Mexicali Valley. Mexican representatives also
believed that they were entitled to substantial
compensation for damages caused by saline
Wellton-Mohawk waters since 1961.
Agreement with Mexico
After extensive negotiations, agreement
was reached between Ambassador Brownell
and Secretary of Foreign Relations of Mexico,
Emilio O. Rabasa in the latter part of August
and was approved by the two Presidents on
August 30,1973, and subsequently incorporated
in Minute No. 242 of the International Bound-
ary and Water Commission, which terminated
Minute No. 241.
Minute No. 242 contains the following ma-
jor provisions:
329
-------
CASE STUDY: WELLTON-MOHAWK DISTRICT
1. The United States shall adopt measures
to assure that by no later than July 1,1974, the
waters delivered to Mexico upstream from
Morelos Dam will have an average annual
salinity of not more than 115 ppm, plus or minus
30 ppm, over the annual average salinity at
Imperial Dam. This quality guarantee becomes
effective upon authorization by Congress of the
funds to construct the necessary works.
2. Until Congress authorizes the necessary
works to provide the quality guarantee, the
United States shall continue to bypass Wellton-
Mohawk drainage water at the annual rate of
118,000 acre-feet per year and substitute
therefore an equal volume of better quality
water.
3. The United States will continue to
deliver approximately 140,000 acre-feet to Mex-
ico on the land boundary at San Luis, Mexico,
with a salinity essentially the same as that of
the waters customarily delivered there (ap-
proximately 1550 ppm).
4. The concrete-lined Well ton-Mohawk
drain shall be extended approximately 53 miles
to Santa Clara Slough (on the Gulf of Califor-
nia) with a capacity of 353 cubic feet per second.
Construction and operation in Mexico would be
performed by the Mexican Government, but at
the expense of the United States.
5. Each country shall limit pumping of
groundwaters in its territory within five miles of
the Arizona-Sonora boundary near San Luis to
160,000 acre-feet annually.
6. The United States will support efforts by
Mexico to obtain appropriate financing for
improvement and rehabilitation in the Mexicali
Valley. The United States will provide non-
reimbursable assistance for those aspects of the
rehabilitation program relating to the salinity
.problem, including tile drainage. The extent of
this participation is to be negotiated in the
future.
7. The new Minute is to be recognized as a
permanent and definitive solution to the Colo-
rado River salinity problem.
Analysis of "Permanent" Solution
Mexico receives major new benefits from
the agreement. She receives a guaranteed quali-
ty at Morelos Dam related to the quality of water
at Imperial Dam. All of the costs in money or
water to achieve this quality guarantee is to be
borne by the United States. Mexico has the
promise of financial assistance with respect to
salinity problems, including tile drainage in
Mexicali Valley. In addition, the United States
will support efforts by Mexico to obtain
favorable financing for the improvement and
rehabilitation of Mexicali Valley. Also at the
United States expense, a concrete-lined canal is
to be constructed in Mexico to discharge saline
water.
The United States negotiators believe there
are considerable tangible benefits to the United
States. The agreement eliminates the possibili-
ty of long years of acrimonious controversy
between the two countries. The agreement does
not require any payments to Mexico for any past
damages. Since the agreement is described as a
permanent solution, presumably Mexico has
waived any future rights to press for monetary
damages. Mexico agreed to permanently accept
140,000 acre-feet per year of their treaty right at
the Arizona-Sonora boundary. This is largely
drainage water with a salinity considerably
higher than that of the Colorado River.
Although Mexico has accepted this water for
years and utilized it, there was apparently no
obligation on Mexico's part to accept the waters
at this location until this agreement was signed.
Although the United States will reduce the
salinity of the Wellton-Mohawk drainage
waters under the new agreement, Mexico has
agreed that other drainage waters below Im-
perial Dam will continue to be accepted as part
of the U.S. Treaty obligation. In summary,
negotiators believe that the concessions made
by the United States are in terms of money, not
water. However, this is not entirely true, as the
reject stream from the desalting plant currently
estimated at 43,000 af/ yr, will not be included as
part of the 1,500,000 af annual obligation and
will have to be replaced.
While the agreement is silent with regard to
salinity control upstream of Imperial Dam, both
the U.S. and Mexican negotiators agree that
upstream salinity control is vital to the long-
term success of the agreement. The Mexican
government apparently believes that the Unit-
ed States has committed itself to a program to
control the river's salinity upstream from Im-
perial Dam, for Foreign Minister Rabasa made
the following statement to the press in Mexico
City when the agreement was announced on
August 30, 1973:
". . . the final result will be that the
Mexicali farmers will have forever — they
330
-------
SALINITY AGREEMENTS
and their children and their children's
children — water whose average annual
salinity will never exceed 1010 ppm, which
is perfectly acceptable."
When correction is made for the method of
analysis for salinity used in Mexico, which is
different than that used in the United States,
and the 115 ppm differential between Imperial
Dam and Morelos Dam is added, the 1010 ppm
referred to above is equivalent to salinity at
Imperial Dam at the time the agreement was
signed.
Mr. Brownell has stated that unless the
United States controls the salinity at Imperial
Dam, in the future we will have a new salinity
problem with Mexico.
Colorado River Basin Salinity Control Act
The Administration desired that the legisla-
tion to authorize the necessary works for im-
plementation of Minute No. 242 cover just those
works. The Colorado River Basin states and
their Congressional representatives, however,
considered that there is one Colorado River
salinity problem and that any legislation
should encompass both the Mexican salinity
measures and a basin-wide salinity control
program. Further, there were a few features
relating to the Mexican salinity measures, such
as replacement of the reject stream from a de-
salting plant, that the states consider to be an
essential part of any legislation on the agree-
ment with Mexico. Accordingly, two bills were
introduced in Congress, the Administration's
bill and one by the basin states, which latter bill
included all the items deemed necessary to
implement Minute No. 242, plus the items
considered essential by the states implement-
ing the Minute, and, in addition, a basinwide
salinity control program.
Congress enacted the bill supported by the
states and the Colorado River Basin Salinity
Control Act, Public Law 93-320, became law on
June 24, 1974. It consists of two parts, Title I —
Programs Downstream From Imperial Dam,
and Title II — Measures Upstream From Im-
perial Dam. The Title I measures consist of
those features necessary to implement the
agreement with Mexico, including some that are
needed to protect the Colorado River Basin
states. Title II directs the Secretary of Interior to
implement a basinwide salinity control
program. It includes authorization for construc-
tion of four salinity control projects and directs
that planning reports be undertaken for twelve
other projects.
Title I Features
Under Title I, the Secretary of the Interior is
directed to proceed with a program of works to
enable the United States to comply with its
obligations under the agreement. Plate A shows
the location of the major features of Title I.
These authorized works and actions are:
1. A 129 million gallon per day desalting
plant and the necessary appurtenant works to
reduce the salinity of the Wellton-Mohawk
drainage water.
2. A lined brine disposal channel to convey
the brine from the desalting plant to Santa
Clara Slough in Mexico.
3. Replacement of the first 49 miles of the
Coachella Canal in California. This will reduce
seepage losses by about 132,000 acre-feet a year,
and the calculated water savings will be
credited against stored water used by the
government in complying with the agreement
with Mexico. When deliveries to California are
reduced in the future, the government will no
longer be able to use the water salvage credits
and, therefore, Coachella Valley County Water
District will make annual payments on the
canal replacement.
4. Various changes in the Wellton-Mohawk
Irrigation and Drainage District in order to
reduce the amount of annual drainage water
from that District to 175,000 acre-feet or less.
These changes would include a reduction in the
lands irrigated from the authorized level of
75,000 acres to a level of about 65,000 acres, an
improvement in irrigation efficiency, and other
necessary activities.
5. Direction to obtain a source of energy for
the desalting plant that does not diminish the
supply of federal power for preference customers
in the region.
6. A groundwater well field with a capacity
of 160,000 acre-feet a year on the Yuma Mesa
located within five miles of the Mexican border,
and to acquire rights in lands overlying the
groundwater areas to be pumped.
7. A study to be completed by June 30,1980,
of feasible measures for the replacement of the
brine from the desalting plant.
The Congress authorized for appropriation
the sum of $155.5 million for the construction of
331
-------
CASE STUDY: WELLTON-MOHAWK DISTRICT
the preceding features, and, through fiscal year
1977, the Congress appropriated a total of $93.6
million to enable work in most phases to
proceed rapidly. Extensive testing of the
Wellton-Mohawk drainage water for problems
to be overcome in desalting the water has been
underway for over two years. This testing
program has evaluated necessary pretreatment
processes and costs, and has enabled possible
desalting equipment and alternative desalting
membrane combinations to be performance-
tested in the actual field setting of the Yuma De-
salting Plant. The Plant is scheduled to be
completed by December 1981.
Funds have been made available to Mexico
for it to construct the portion of the brine
disposal channel within Mexico, and that work
is now completed. A contract has been awarded
for the construction of the United States section
of the channel, and completion of the concrete
lining is scheduled for March 1977, with the
final work to be completed shortly thereafter.
The 49-mile Coachella Canal reconstruction
design is about completed, a form of contract for
the repayment of allocated costs by the
Coachella Valley County Water District is near-
ing final approval and the canal construction is
scheduled for completion in March 1981.
Several thousand acres of land in the
Wellton-Mohawk district have been acquired
and retired from irrigation, and research is
continuing on techniques and facilities to
further reduce return flows from that district.
The Bureau of Reclamation has performed
preliminary studies of sources to replace the de-
salting plant brine, and has commenced an
investigation of alternative sources. The border
well pumping field is being acquired and pump-
ing wells and appurtenant facilities laid out.
As this progress report shows, the United
States is engaged in a major effort to observe its
agreement with Mexico. Pending completion of
the desalting works, the United States has been
meeting the 115 ppm ± 30 ppm salinity differen-
i
\
'
'
BYPASS DEAIN
SOUTHERN INT'L BOU N D A R Y
UNHID STATIS
DIPAITMINT Of IMI IMTIilOt
IUKIAU OF IICLAMATION
OMICI 0» SAIINI WATII
COIOBAOO RIVER MlEHIWIIMl
SALINITY CONlROt PBOJECI
PIOJICT HAN AtlA
MAP NO. »9i-300- J«
CLARA
SLOUGH
* '
Plate A.
332
-------
SALINITY AGREEMENTS
tial by discharging all Wellton-Mohawk
drainage water below Morelos Dam without
charge against the treaty and by substituting
therefore Colorado River water from storage.
The quantities bypassed were 111,000 acre-feet
in the last half of 1974,216,000 acre-feet in 1975,
and 206,000 acre-feet in 1976. This stored water
is to be replaced by the water salvaged by the
Coachella Canal lining until such salvaged
water is needed in California.
This program has effectively reduced the
salinity of the water delivered to Mexico. In
1971, the last full year prior to the combined
U.S.-Mexico programs to bypass Wellton-
Mohawk drainage waters around Morelos Dam,
The costs of the facilities authorized by
Congress has escalated sharply since 1974. The
latest estimated cost of Title I facilities is
$317,000,000, with a major portion of the in-
crease due to the desalting plant and related
facilities.
Future Actions
Although the title of the latest Colorado
River agreement with Mexico (Minute No. 242)
is "Permanent and Definitive Solution to the
annual salinities were 1,160 ppm at Morelos
Dam and 880 ppm at Imperial Dam, for a
differential salinity of 280 ppm. Under the
bypass program, annual salinities at Imperial
and Morelos Dams were as follows:
ANNUAL SALINITIES, ppm
Year Imperial Darn Morelos Dam Differential
1974
1975
1976
830
830
820
960
960
960
130
130
140
International Problem of the Salinity of the
Colorado River", it is apparent that for the
agreement to indeed become the hoped-for per-
manent solution, a number of actions now
underway must be carried to completion. These
actions include (a) construction of the world's
largest desalting plant and appurtenant
features, requiring appropriation and expen-
diture of hundreds of millions of dollars, (b)
construction of the first 49-miles of the
Coachella Canal, (c) continued development of
the authorized Colorado River Basin salinity
control program encompassing both natural
and man-made sources of salinity in the entire
Colorado River Basin, (d) appropriation and
expenditure of over a hundred million dollars
for the authorized salinity control units, (e)
authorization, over the next several years, of
additional feasible salinity control units and
the subsequent construction of such units, and
(f) better irrigation practices in both the United
States and Mexico.
333
-------
An Assessment of Irrigation
Efficiencies and Drainage Return
Flows from the Wellton-Mohawk
Division of the Gila Project
D. L. KRULL and D. L. CLARK
Yuma Projects Office, Bureau of Reclamation,
U.S. Department of the Interior, Yuma, Arizona
ABSTRACT
Public Law 93-320 authorized the construc-
tion of a large desalting plant to enable the
United States to comply with its obligation
under the agreement with Mexico of August 30,
1973 (Minute No. 242 of the International
Boundary and Water Commission, United
States and Mexico). The desalting plant will be
constructed to treat the drainage waters from
the Wellton-Mohawk Division of the Gila Pro-
ject so these waters will be available for delivery
to Mexico under the 1944 Mexican Water Treaty
and Minute No. 242. By reducing the volume of
drainage flows, the size of the desalting plant
can be reduced. An interagency committee was
established to determine and implement means
of increasing irrigation efficiencies so as to
reduce drainage return flows to the lowest
practical volume and to identify an optimal
economic solution relating desalting costs to
specified irrigation efficiency levels. An assess-
ment was made utilizing the water budget
process to determine the relationship between
increased irrigation efficiency levels and
drainage return flows.
INTRODUCTION
Wellton-Mohawk Division of the
Gila Project
The Reclamation Project, known as the Gila
Project, was authorized and established under
the provisions of the Reclamation Laws, the Act
of June 16, 1933, and various appropriations
acts for the purpose of reclaiming and irrigating
lands in the State of Arizona. The Gila
Reauthorization Act of July 30, 1947, provided
for a reduction in the acreage and established
the boundaries of the Gila Project, as shown on
Figure 1. The Wellton-Mohawk Division was
thereby authorized to include an area com-
prising approximately 75,000 irrigable acres of
land, or such number of irrigable acres as can be
adequately irrigated by the beneficial consump-
tive use of no more than 300,000 acre-feet of
water diverted annually from the Colorado
River. Wellton-Mohawk Division lands in
southwest Arizona were reclaimed for agricul-
tural purposes by importing Colorado River
waters, beginning in 1952.
The Gila Gravity Main Canal delivers
irrigation water to the Gila Project from Im-
perial Dam. The Wellton-Mohawk Check and
Turnout located about 15 miles downstream
from Imperial Dam diverts water to the Wellton-
Mohawk Canal that serves the Wellton-
Mohawk Division. The water delivery facilities
in the Division include 108 miles of main canals
and 227 miles of laterals and appurtenant
structures. Three large pumping plants on the
Wellton-Mohawk Canal lift the water a total of
170 feet. Small relift pumps are scattered
throughout the Division on laterals.
The Wellton-Mohawk Irrigation and Drain-
age District (District) was formed in 1951 and in
1952 assumed the repayment obligation of the
irrigation and drainage facilities. The Gila
Project delivered the first Colorado River water
to the Wellton-Mohawk Division in 1952 and the
entire distribution system was in operation by
the end of 1957. Care, operation, and mainten-
ance of these facilities were assumed over a per-
iod of time by the District upon completion of
construction.
335
-------
CASE STUDY: WELLTON-MOHAWK DISTRICT
N A
'
UNITE; STATES
DEPARTMENT CF THE INTERIOR
BUREAU OF ACCLAMATION
YUMA AREA PROJECTS
ARTZOKi
FEBRUARY 1977
Figure 1.
The Physical Setting
The District lands begin 8 miles east of the
confluence of the Gila and Colorado Rivers in
the vicinity of Yuma, Arizona. The Gila River
floodplain throughout this reach is referred to
as the Dome and Wellton-Mohawk Valleys, The
adjacent southern Gila River terrace is known
as the Wellton Mesa. Of the 75,000 irrigable
acres in the District, 61,270 acres are valley
lands and 13.730 acres are mesa lands. The area
has a 12-month growing season with potential
frost danger to tender crops between November
15 and February 15. Summers are dry and hot
with maximum temperatures near 120°F.
Minimum winter temperatures are about 25°F.
Average annual rainfall is about 3.5 inches.
The bottom land soils have been formed
from alluvial deposits of sands, silts, and clays.
The soils are often stratified and platy and have
had noticeable quantities of organic matter
added through heavy vegetative growth. The
arable soils have a relatively high water
holding capacity because of the predominance
of silt Textures range from coarse sand to clay
loam, but the most common textures are silt and
silt loam.
The textures of the mesa soils are generally
coarser than the valley soils, resulting in a lower
moisture holding capacity. The soil textures
range from sand to sandy loam, with a
dominance of good quality loamy sands.
The Salinity Problem
After irrigation with water from the
Colorado River was begun, the water table rose
and by 1957 crops were threatened by drowning.
Some old irrigation wells were reactivated for
drainage wells, and others were constructed to
remove excess ground water. This drainage
water was originally discharged to the Gila
River and was quite saline, initially averaging
6,000 ppm of total dissolved solids (TDS). From
early 1960 through late 1961, the concrete lined
Wellton-Mohawk Main Conveyance Channel
(WMMCC) was constructed throughout the
length of the District to carry drainage waters
from 108 wells that were ultimately put into
operation. The drainage water then emptied
into the Gila River near its confluence with the
Colorado and thence flowed to Morelos Dam
where it was delivered to Mexico. In addition,
336
-------
EFFICIENCIES AND RETURN FLOWS - GILA PROJECT
«s>*
UNITED STATES
DEPARTMENT OF THE INTERIOR
BUREAU OF RECLAMATION
COLORADO RIVER BASIN
SALINITY CONTROL PROJECTS
TITLE I
WELLTON-MOHAWK DRAINAGE
CONVEYANCE FACILITIES
MAP NO. I2S2-300-1
Figure 2.
excess flows of the Colorado River which Mex-
ico had received prior to 1961 came to a near end
in that year. The excess flows had diluted the
drainage water entering the river below Im-
perial Dam in previous years.
The combined effect of these developments
increased the salinity of Colorado River water
delivered to Mexico upstream of Morelos Dam
from an annual average of about 800 ppm TDS
to nearly 1,500 ppm TDS in 1962. The increase
in salinity created an international situation
and the President of the United States and the
President of Mexico agreed in 1962 that it was
urgent to find a mutually satisfactory solution.
In late 1965 the 12-mile long Main Outlet
Drain Extension (MODE) was completed that
allowed discharge of the Wellton-Mohawk Divi-
sion drainage waters either upstream or
downstream of Morelos Dam. A diagram of the
Wellton-Mohawk drainage facilities, including
the MODE are shown on Figure 2.
Operations of the Colorado River system in
accordance with interim agreements with Mex-
ico and coupled with improvement in the salini-
ty of Wellton-Mohawk Division drainage waters
reduced the average annual salinity of water
made available to Mexico to about 1,140 ppm
TDS for the year ending June 30, 1973. During
this period Wellton-Mohawk Division drainage
was in the range of 210,000 to 220,000 acre-feet
per year. Some of the drainage was bypassed to
the Colorado River below Morelos Dam and was
not utilized by Mexico.
However, the United States and Mexico
pursued a definitive, equitable and just solution
to the problem of salinity of the water delivered
to Mexico from the Colorado River.
Minute No. 242
Minute No. 242 of the International Bound-
ary and Water Commission, United States and
Mexico was approved by the two Governments
on August 30, 1973. The Minute is entitled
"Permanent and Definitive Solution to the
International Problem of the Salinity of the
Colorado River." Point 1 (a) of Minute No. 242 is
as follows:
"1. Referring to the annual volume of
Colorado River waters guaranteed to
337
-------
CASE STUDY: WELLTON-MOHAWK DISTRICT
Mexico under the Treaty of 1944, of 1,500,-
000 acre-feet (1,850,234,000 cubic meters):
(a) The United States" shall adopt
measures to assure that not earlier
than January 1, 1974, and no latter
than July 1, 1974, the approximately
1,360,000 acre-feet (1,677,545,000
cubic meters) delivered to Mexcio up-
stream of Morelos Dam, have an an-
nual average salinity of no more than
115 ppm+ 30 ppm U.S. count (121
ppm + 30 ppm Mexican count) over
the annual average salinity of
Colorado River waters which arrive at
Imperial Dam. with the understand-
ing that any waters that may be
delivered to Mexico under the Treaty
of 1944 by means of the All-American
Canal shall be considered as having
been delivered upstream of Morelos
Dam for the purpose of computing this
salinity."
As Imperial Dam releases proceed down-
stream to Morelos Dam, they are influenced by
return flows from irrigation projects which
increase the salinity. The 115 ppm differential
was negotiated with Mexico principally on the
basis of excluding return flows of the Wellton-
Mohawk Division of the Gila Project from the
deliveries to Mexico and including other return
flows to the river below Imperial Dam.
A large desalting plant has been authorized
by the Congress for construction to treat the
Wellton-Mohawk return flows, thereby reducing
the salinity of the water delivered to Mexico
upstream of Morelos Dam and preserving that
quantity of the treated water for use in the
United States.
An increase in irrigation efficiencies in the
District will reduce desalting plant cost by
reducing the quantity of water that will be
treated. Reducing the quantity of water treated
by the desalting plant will also decrease the
quantity of water that will be lost as brine in the
desalting process.
Public Law 93-320
Public Law 93-320 of June 24,1974, author-
izes construction, operation, and maintenance
of certain works in the Colorado River Basin to
control the salinity of water delivered to users in
the United States and Mexico. The Act (88 Stat.
266) is cited as the "Colorado River Basin
Salinity Control Act." Title I of the Act pertains
to works authorized for construction
downstream of Imperial Dam to meet the re-
quirements of Minute No. 242, and among other
things, provides for the following:
1. Construction of a desalting complex and
appurtenant works.
2. Acceleration of the cooperative program of
Irrigation Management Services with the Dis-
trict for the purpose of improving irrigation
efficiency.
3. Reduction of approximately 10,000 acres in
the authorized irrigable acreage in the District
and further reduction in the irrigable acreage
with the consent of the District as may be
deemed appropriate.
4. Authorization, either in conjunction with or
in lieu of land acquistion, to assist water users in
the District in installing systems im-
provements, such as ditch lining, change in
field layouts, automatic equipment, sprinkler
systems and bubbler systems, as a means of
increasing irrigation efficiencies; provided, that
all costs associated with the improvements
authorized and allocated to the water users on
the basis of benefits received shall be reim-
bursed to the United States.
5. Authorization for amendment of the con-
tract dated March 4, 1952, as amended, between
the United States and the District to provide for
reduction of the repayment obligation associ-
ated with the reduction of irrigable acreage in
the District and for credit against the outstand-
ing repayment obligation to offset any increase
in operation and maintenance assessments per
acre which may result from the District's de-
creased operation and maintenance base.
Advisory Committee on Irrigation
Efficiency in the Wellton-Mohawk and
Drainage District
Prior to proposing legislation to Congress *-
(concerning the United States requirements of
fulfilling Minute No. 242) President Nixon, in
mid-1973, directed that a working group be
established to determine and implement means
for increasing Wellton-Mohawk onfarm irriga-
tion efficiencies.
The group includes representatives of the
Department of the Interior, Environmental
Protection Agency, Department of Agriculture,
and the Office of Management and Budget. The
'The proposed legislation eventually became Public
Law 93-320.
338
-------
EFFICIENCIES AND RETURN FLOWS - GILA PROJECT
group became known as the Advisory Com-
mittee on Irrigation Efficiency, Wellton-
Mohawk Irrigation and Drainage District. In
August 1973, the Technical Field Committee
(TFC) was created by the Advisory Committee
for the following objective of assignment:
"The objective is to determine means of
reducing the pumped drainage effluent
from the District to the lowest practical
volume through improvement of onfarm
irrigation efficiencies and related mea-
sures. In addition, the TFC will identify an
optimal economic solution relating desalt-
ing costs to specific efficiency levels."
The Advisory Committee also charged the
TFC with specific tasks. A principal one was to
determine the best estimate of the quantity and
quality of Wellton-Mohawk return flows for use
in design of the desalting plant. The task was to
include (1) determining the measures required
for attaining higher irrigation efficiencies or
otherwise reducing the volume of drainage
effluent from the District; (2) estimating ranges
of costs and implementation times for at-
tainable levels of onfarm and system irrigation
efficiencies and estimating the volume and
quality of the drainage effluent that would be
associated with each cost, time, and efficiency
level; (3) developing alternatives and recom-
mending specific measures and a plan of imple-
mentation to achieve the stated objective; and
(4) recommending a followthrough program for
implementation, including the additional in-
puts or studies needed to refine or confirm
irrigation efficiencies findings and conclusions.
The TFC consists of interagency members
which include the Bureau of Reclamation, the
Agricultural Research Service, the Soil Conser-
vation Service, and the Environmental Protec-
tion Agency, with observers from the District
and the U.S. Section, International Boundary
and Water Commission.
THE ANALYSIS
Existing Practices
The analysis which associated onfarm
measures with efficiency improvements was
undertaken by the Technical Field Committee
(TFC) utilizing cropping and water use data for
calendar years 1970, 1971, and 1972.
The first step was an inventory and evalua-
tion of the existing onfarm irrigation systems in
the District. Soil surveys, aerial photographs,
field investigations by the Soil Conservation
Service (SCS), and records of the District and
Natural Resource District2 were used. All fields
serviced by operating irrigation turnouts (about
1,200) were considered for the following data:
1. Lengths of lined and unlined farm
ditches.
2. Field layouts including the length and
width of borders and sizes of irrigation sets.
3. The slope and levelness of the fields
4. Size, type, and number of field turnouts.
5. Types of soil.
The inventory showed that over 47,000
acres had been developed into level basins and
that about 633 miles of the 744 miles of onfarm
ditches were concrete lined. In the leveled
basins no tailwater occurs, and water in excess
of consumptive use percolates to the ground-
water table.
Of the approximately 65,000 acres in irriga-
tion rotation in the District in 1973,about31,000
acres of farm lands had been serviced by SCS
and had adequate physical systems installed
for efficient water management. Onfarm
systems on the remaining acreage were ana-
lyzed to determine measures needed to improve
irrigation efficiency and reduce deep percola-
tion. Basic data collected on each field were used
to determine present efficiencies and evaluated
to determine (1) those fields where significant
improvements are needed, and (2) what is
needed to do a complete job of systems improve-
ment.
Existing Assistance Programs
The Bureau of Reclamation and the District
had established a cooperative IMS program in
1973. Under this program the farmer is provided
information on when to irrigate and how much
water to apply to replenish depleted soil
moisture. The program involves determining
the soil-water-plant relationship and uses
various methods to determine the soil moisture
status and the soil moisture depletion rate.
The SCS provides technical assistance
through the Wellton-Mohawk Valley Natural
Resource Conservation District. Information on
2The Wellton-Mohawk Valley Natural Resource Con-
servation District is the local organization of the
Arizona State Land Department. It is responsible to
fulfill requirements of the Arizona conservation
laws. Assigned responsibilities includes restoration
and conservation of land and other natural
resources in Arizona.
339
-------
CASE STUDY: WELLTON-MOHAWK DISTRICT
irrigation management along with system im-
provement technical assistance is provided.
Limited financial assistance is available.
Improvement Potential
Soil Conservation Service criteria which
adapt the length and width of set to the soil
intake rate and available water holding capaci-
ty were used to determine potential onfarm
gravity system improvements. Further con-
sideration was given to established irrigation
practices in the District resulting in the follow-
ing design assumptions:
1. Minimum practical length of run is 330
feet.
2. Depth of water application in a field
would be a maximum of 6 inches.
3. Exactly 15 cfs would be available for
each irrigation.
4. Fields are planted to close growing crops
such as alfalfa.
Using these criteria, preliminary designs
were developed except for the citrus acreage.
The need for leveling was determined, as well as
the most appropriate size of set. It was not
believed practical to change the length of run or
the grade on the citrus fields. Most of the fields
in the District can be efficiently irrigated with
660-foot runs, although 400-foot runs were con-
sidered for some fields with coarse textured
soils. Concrete lined ditches were included for
all fields where they do not exist. Outlets and
measuring devices were included as needed.
Pressure system improvements for the citrus
acreage were considered.
Citrus is located on the mesa where soils are
coarse and highly permeable. Irrigation by
flooding is not efficient on these soils and there
is little prospect for improving irrigation ef-
ficiencies. Pressurized systems using tricklers
or bubblers, however, were thought to be most
promising and also offered advantages for
fertilizer applications and savings in labor
costs.
Many combinations and levels of assis-
tance programs that could improve efficiencies
were considered. They included improving
gravity systems, use of new technology such as
drip and bubblers on citrus and automation on
both gravity and pressure systems, technical
assistance and information service to design
new systems, research and demonstration of
new systems and how to apply the required
irrigation application, and better measurement
of water. Other factors that affect water use
include economic benefits to farmers, cost shar-
ing on assistance programs, water pricing
policy, and a willingness on the farmer's part to
improve his water management.
Program levels were developed to reflect (1)
management and system improvements assoc-
iated with various levels of irrigation efficien-
cies, and (2) a range of alternatives for selection
of a recommended program.
Potential Irrigation Efficiencies and
Costs
Four program levels and associated irriga-
tion efficiencies were developed to evaluate and
project reductions in return flows for 65,000
irrigable acres. A summary of the programs
utilized to evaluate the potential for reducing
return flows is presented in Table 1.
Program Level 1
Irrigation efficiencies for Program Level 1
were determined by reviewing the irrigation
efficiencies for each crop for 1970, 1971, and
1972. With the knowledge gained in gathering
the basic data and the evaluation of the existing
onfarm systems capabilities, judgment was
used to determine a probable increase in ef-
ficiencies that could be attained with effort
placed on water management only. No physical
improvements are included in Program Level 1
other than those available through ongoing
Department of Agriculture programs. A
research and demonstration effort with
emphasis on water use management would be
initiated with interested agencies. The IMS
program in the District would be accelerated.
It was estimated that average onfarm
irrigation efficiencies in the District would be
increased from the average 1970 through 1972
level of 56 percent to 64 percent under Program
Level 1 at an estimated total cost of $1,650,000
over 5 years.
Program Level 2
Program Level 2 includes IMS assistance
and research and demonstrations, the same as
for Program Level 1. In addition, a cost-share
program would be implemented to improve
onfarm systems with the lowest irrigation ef-
ficiency. Improvements on gravity sytems
would be on a total of 11,000 acres. In addition,
conversion to pressure systems on citrus would
340
-------
EFFICIENCIES AND RETURN FLOWS — GILA PROJECT
TABLE 1
Summitry of Programs to Reduce Return Flows
Wellton-Mohawk Irrigation and Drainage District
Price Base: 1973
Improved Management
Program Lewi
1
Farm Efficiency
64 percent
2
Farm Efficiency
69 percent
3
Farm Efficiency
72 percent
4
Farm Efficiency
82 percent
IMS
56,000 ac.
7 men
$200,000/yr
5 yean
56,000 ac.
7 men
$200,000/yr
5 years
56,000 ac.
7 men
$200,000/yr
5 years
56,000 ac.
7 men
$200,000/yr
5 years
Tech. A«st.
2 men
$50,000/yr
5 years
5 men
$150,000/yr
5 years
6 men
$200,000/yr
5 years
8 men
$250,000/yr
5 years
Research and Total
Demonstrations On/arm System Improvement in,tniiniinn
Total Cost Gravity
Reorientation
of ongoing
$400,000 program
11,000 ac. (in
low efficiency
$400,000 group)
($2.45 M)
20,000 ac.
$400,000
($4.07 M)
30,000 ac.
$400,000
($5.82 M)
»«""* <*mL>
1.65
3,000 ac. drip &
bubbler; 450 ac.
sprinklers 6.74
($2.14 M)
450 ac. sprinklers
4,000 ac. drip &
bubblers 9.30
($2.63 M)
8,000 ac. drip &
bubblers
3,000 ac. sprink-
lers 14.50 14.50
($6.03 M)
Federal Annual
Installation Equivalent
Cosl1 Federal Cost'
($ Million) (i Million)
1.65 .10
5.53 .33
7.51 .45
11.15 ,67
NOTE- Annual equivalent private cost2 including 25 percent cost sharing and operation, maintenance, and re-
pair amount to: P.L. No. 1 - going; P.L. No. 2 - $325,000; P.L. No. 3 - $502,000; and P.L. No. 4 - $827,000.
ifiased on 75 percent cost sharing for permanent onfarm irrigation improvement measures plus the value of pro-
gram expenditures for technical assistance, IMS and R and D.
interest at 5-5/8 percent, 50-year project life.
be undertaken on 3,000 acres. The total cost of
Program Level 2 was estimated to be $6,740,000
over 5 years. The total average onfarm irriga-
tion efficiencies were not estimated but believed
to be about midway between Program Levels 1
and 3.
Program Level 3
Program Level 3 includes total system im-
provements to gravity systems on 19,860 acres
and pressure system installations on 4,000 acres
of citrus in addition to IMS assistance and
research and demonstrations. The total es-
timated cost of the program is $9,300,000
resulting in onfarm irrigation efficiencies of 72
percent.
Program Level 4
Program Level 4 includes total gravity
system improvements on 29,600 acres and com-
plete conversion to pressure systems on 8,000
acres which is all of the citrus acreage in the
District. In addition, the IMS assistance and
research and demonstrations would be includ-
ed. The total estimated cost of the program is
$14,500,000 and the average onfarm irrigation
efficiency would be increased to 82 percent.
Program Level 4 efficiencies were based on
the design efficiencies of the fields in the
District assuming that all required im-
provements would be made. The design efficien-
cy on flood irrigated fields would be 80 percent
and pressure installations on the citrus acreage
would operate at 85 percent. The overall average
on farm irrigation efficiency would be 82 per-
cent. It was determined that the potential for
achieving Program Level 4 efficiencies was
theoretically possible but beyond practical
achievement on the entire District acreage in
the foreseeable future.
Table 2 presents the irrigation efficiencies
of the 1970, 1971, and 1972 average and for
Program Levels 1 and 3. The District was
divided into six segments according to
topographical characteristics of the District
and salinity characteristics of the aquifer to
better evaluate the water and salt mixing taking
place in the aquifer. There are five valley
segments and one mesa segment.
341
-------
CASE STUDY: WELLTON-MOHAWK DISTRICT
Water Budgets
Accounting for water entering and leaving
the aquifer underlying the District is the key to
understanding relationships between onfarm
irrigation efficiencies and drainage return flow.
Water budgets were used to assess historical
conditions, to predict future drainage returns,
and are being used to assess the effects of the
programs implemented to increase onfarm
irrigation efficiencies. The water budget
diagram depicted on Figure 3 illustrates the
many components of water considered for
balancing inflows and outflows with changes in
storage.
The quantity and quality of drainge return
flows from the District are influenced by (1) size
of area irrigated, (2) crop distribution and
related consumtive use, (3) onfarm irrigation
TABLE 2
Projected Irrigation Efficiencies for Program Level 1 and Program Level 3
Wellton-Mohawk Irrigation and Drainage District
Segment
Irrigation Efficiencies
Irrigation Efficiencies
Irrigation Efficiencies
Irrigation Efficiencies
3 Yr. Program Level
Aug.'- 1 3'
3 Yr. Program Level
Aug.' 1 3'
3 Yr. Program Level
Avg.' 1 3'
3 Yr. Program Level
Avg.' 1 3'
Dome
Roll
Tyson
Texas Hill
South Valley
Mesa
Wt, Avg. (Valley)
Wt. Avg. (Mesa)
Dome
Roll
Tyson
Texas Hill
South Valley
Mesa
Wt. Avg. (Valley)
Wt. Avg. (Mesa)
WHEAT
SORGHUM
ALFALFA
COTTON
38
62
77
57
59
_ —
64
__
60
67
82
63
66
__
70
__
70
71
85
69
70
__
74
—
53
59
76
45
52
__
56
__
58
63
79
50
57
__
63
—
60
67
81
60
62
—
68
—
83
91
82
55
71
55
82
55
84
85
83
58
76
60
81
60
85
85
85
63
82
65
83
65
72
84
98
66
70
__
79
—
77
85
85
71
73
__
79
—
82
85
85
76
76
__
81
—
LETTUCE
MELONS
BURMUDA SEED
TREE FRUITS3
20
18
21
24
20
25
25
25
25
25
27
27
27
27
27
32
—
—
—
—
37
—
—
—
—
40
—
—
—
—
75
68
71
50
57
77
73
74
55
62
80
75
76
60
66
—
40
—
—
--
65
—
--
20 25 27
32 37 40
66 70 72
28
28
40
40
40
65
65
65
PASTURE MISCELLANEOUS NOT HARVESTED
(Includes Other Hay) (Less Citrus Not Harvested)
Dome
Roll
Tyson
Texas Hill
South Valley
Mesa
Wt. Avg. (Valley)
Wt. Avg. Mesa)
39
70
43
48
67
36
60
36
44
72
60
53
72
41
64
41
51
75
75
57
75
43
70
43
34
31
—
46
45
—
41
39
40
—
51
50
—
44
41
50
—
51
53
—
49
Adjusted from District records. High irrigation efficiencies for alfalfa and cotton reflect contribution from
ground-water.
Includes pressure systems for orchards and limited sprinkler germination of vegetables.
Includes grapefruit, lemons, limes, oranges, tangerines, and pecans.
4For computer run pecans in Tyson segment remain in miscellaneous at 66 percent irrigation efficiency.
342
-------
EFFICIENCIES AND RETURN FLOWS — GILA PROJECT
WELLTON- MOHAWK IHRKUTKW AND DRAINAGf DISTRICT
WATER BUDGET ANALYSIS
I 8DD INFLOWS
f K- WXr«*«.
0»" HtCL^UATIOH
BUDGET DIAGRAM
WELLTON- MOHAWK IRRIGATION
AND
DRAINAGE DISTRICT
JUNE 1976 li32-30i-IISI
YUMA . ARIZONA
Figure 3.
efficiencies, (4) distribution system operating
efficiencies, (5) consumptive use by
phreatophytes and other noncrop vegetation,
(6) soil and drainage conditions, (7) aquifer
recharge attributable to flooding, (8) rainfall, (9)
salinity of the ground water, (10) salinity of the
applied irrigation water, and (11) pumping
operations.
Table 3 presents water budgets for the 1970,
1971, and 1972 base condition and for Program
Levels 1 through 4. The closure on the water
balance for 1970,1971, and 1972 was in error by
31,000 acre-feet. This inbalance was believed to
be related to many factors, such as (1) surface
and subsurface inflow to the aquifer from side
washes, (2) phyreathophyte use, (3) application
of crop consumptive use values, (4) effective
rainfall, and (5) subsurface outflow from the
mesa to the south. The evaluation for the base
year of 1970, 1971, and 1972 consistantly
reflected the outflow to be greater than the
inflow to the aquifer.
Because of the complexity of the many
factors influencing the aquifer and the level of
confidence that was placed on individual es-
timates, no attempt was made to force a balance
on inflow versus outflow. It was believed more
appropriate to project an inbalance along with
the other components to determine pumping
requirements when deep percolation is reduced
by increasing irrigation efficiencies. ^
Salinity Projections
The salinity of drainage returns was es-
timated by use of a digital simulation model.
The model considers the more important
chemical reactions such as solution or precipita-
tion of lime and gypsum, ion exchange, interac-
tions of calcium and bicarbonate ions with the
partial pressure of carbon dioxide and ion
pairing; other soluble salts are assumed to be in
3Since 1973, adjustments and refinements have been
mad i and the average inbalance has been reduced to
about 5,400 acre-feet per year for the base years of
1970, 1971, and 1972.
343
-------
CASE STUDY: WELLTON-MOHAWK DISTRICT
TABLE 3
Return Flows for 65,000 Irrigable Acres
Wellton-Mohawk Irrigation and Drainage District
Input Data
Irrigation Rotation
Irrigated
Crop Consumptive Use
Irrigation Efficiency
Derived Data
Delivered to Farms
District System Losses
Station 790, GGMC
GGMC Losses
Imperial Dam Diversion
Acauifer Inflow
Deep Percolation
District System Losses
Texas Hill
Effective Rainfall
Subtotal
Aquifer Outflow
Underflow South
Gila River at Dome
Phreatophyte Use
Subtotal
Flow at 8 + 30, WMMCC
Inflow - Outflow
Adjustment
Estimated Return Flow
(Ac)
(Ac)
(KAF)
ft)
(KAF)
(KAF)
(KAF)
(KAF)
(KAF)
(KAF)
(KAF)
(KAF)
(KAF)
(KAF)
(KAF)
(KAF)
(KAF)
(KAF)
(KAF)
Water Budget
for 1970,
1971, & 1972
65,000
61,400 -
252 V
56
447
51
498
16
514
195
51 „,
2 ^/
li~
25?
9
4 V
63 5^
"76"
183
31
214" I/
Improved
Management
Efficiency
of 64%
Level 1
65,000
61,800 °/
253 y
64
395
51
446
13
459
142
51
2
11
205
8 I/
4
58 I/
136
31
157
Improved Management and Systems
Efficiency
of 69%
Level 2
65,000
61,800 6/
253 I/
69
367
51
418
13
431
114
51
2
11
T75
3Z/
4
55
116
31
147
Efficiency
of 72%
Levels
65,000
61,800 6/
253 I/
72
351
51
402
12
414
98
51
2
11
152"
oZ/
4
53
~57
105
31
136"
Efficiency
of 82%
Level 4
65.000 , .
61,800 Sf
253 y
82
309
51
360
11
371
56
51
2
11
120"
oZ/
4
53
"57
63
31
"94"
Diversions - Returns
(KAF)
300
292
284
278
277
'Consumptive use on irrigated land estimated at 4.1 acre-feet per acre.
^Estimate of underflow into Wellton-Mohawk District.
includes surface flow and estimated underflow.
4Consumptive use of 3.9 acre-feet per acre on about 16,200 acres.
5Flows at 8 + 30, Wellton-Mohawk Main Conveyance Channel adjusted for estimated changes in aquifer
storage.
6Assumes 95 percent of irrigable land under irrigation.
7Underflow south will decrease as irrigation efficiencies on the mesa improve.
8Phreatophyte use will decrease with increased capability to provide better drainage in critical areas.
the solution phase. Simultaneously, the model
concentrates the input irrigation water in
proportion to the consumptive use and per-
colates the water through the soil profile. The
resultant deep percolation from the valley,
mesa, and other areas was thus mixed and
imputed to the aquifer system. The water and
major ionic constituents were then routed
through the aquifer by utilizing the principle of
conservation of mass. The chemical reactions
were presumed to occur in the upper 10 feet of the
soil. Soil samples were withdrawn and analyzed
from 80 borings to a depth of 10 feet made across
each of the delineated segments. This provided
the soil chemical and physical inputs. The
chemistry of the water was determined from
historical records. It is interesting to note that
from 8 to 14 percent of the salts applied in the
irrigation water are precipitated in the soil. The
amount of the precipitation depends primarily
344
-------
EFFICIENCIES AND RETURN FLOWS - GILA PROJECT
HE
LJ
C^
M
5
<
a
C£
\
AM
UK) .g/l ta »•« «000,
A8E . -- I.,.,l.l D«. ll«0 ••/! tofMr MOO. Onfclf* ISO K»F/V« Ir y.«r l» iO
DIB ••••MM
!»• KAF/Y« 1. y«or lt»O
ASE D — -
rl.l D*m OO.H..I rt •*» «t/l , Br«to.f. ISO KAF/YR In »«r 19.0
PREDICTED TREND OF WATER QUALITY
FOR WELLTON-MOHAWK RETURN FLOWS
Figure 4.
on the salinity of the input water, the partial
pressure of carbon dioxide in the soil, and the
irrigation efficiency.
Figure 4 shows return flows of 136,000 and
150,000 acre-feet with associated salinities pro-
jected for future conditions. The water quality of
the return flows and the salinity of the aquifer
are influenced primarily by quality and quanti-
ty of water applied for irrigation, crop and other
vegetative consumptive use, chemical reactions
within the system, and distribution system
losses. Other inflows to the aquifer are believed
to be of minor significance. During the initial
operation of the pump drainage system, average
annual salinity of the drainage was about
6,000 ppm total dissolved solids (TDS). As a
result of aquifer freshening due to irrigation
with Colorado River water and pump drainage
operations, there has been a gradual improve-
ment. The average annual salinity of the
drainage was about 3,700 ppm TDS in 1973,
which is about 100 ppm TDS higher than
shown in Figure 4. However, the long-term
change in the aquifer salinity should be ap-
proximately as shown provided that other con-
ditions affecting the aquifer do not change
significantly. Any assessment of the impact of
changing irrigation efficiencies on the quality
of the return flows must consider changes
occurring in the aquifer.
A salinity of 3,000 ppm TDS is projected for
1980 with 136,000 acre-feet of return flow and
about 2,490 ppm TDS is projected for about
2025 when the aquifer is in salt balance, assum-
ing the salinity of Colorado River water
stabilizes at about its present level. If no salinity
controls are implemented above Imperial Dam,
the salinity of the Colorado River is projected to
increase to 930 ppm TDS in 1980 and to about
1,160 ppm TDS by 2000 4 ; under these con-
ditions, the estimated salinities of the return
4Based on estimates included in "Status Report,
Colorado River Water Quality Improvement
Program," U.S. Department of the Interior, Bureau
of Reclamation, January 1974.
345
-------
CASE STUDY: WELLTON-MOHAWK DISTRICT
flow would be about 3,020 ppm TDS in 1980 and
about 3,230 ppm TDS in about 2025 when the
aquifer is in salt balance.
For 150,000 acre-feet of return flow, the
salinity of the aquifer is projected to be about
2,970 ppm TDS in 1980 and about 2,390 ppm
TDS in about 2025 at salt balance, assuming no
increase in the salinity of Colorado River water.
For corresponding time periods and assuming
an increase in salinity of Colorado River water,
as previously discussed, the projected salinities
are 2,990 ppm TDS and 3,050 ppm TDS, respec-
tively. These projections take into account es-
timated precipitation of harmless salts in the
soil profile as irrigation efficiencies increase,
but do not account for possible infiltration from
Gila River floodflows.
Acreage Reduction
Coupled with the proposal to construct
facilities to treat Wellton-Mohawk return flows
to meet requirements of Minute No. 242 is the
provision to hold irrigated acreage in the Dis-
trict to near the present level. This will preclude
increasing return flow from additional develop-
ment. Authorizing legislation (Public Law 93-
320) provides for reducing authorized irrigable
acreage in the District from 75,000 to 65,000
acres and provides for further acreage reduction
with the consent of the Wellton-Mohawk Irriga-
tion and Drainage District.
Another consideration in acreage reduction
is the District's allocation of Colorado River
water under the Gila Reauthorization Act as
previously discussed. Water budget analyses
reflect that about 65,000 acres of land in irriga-
tion rotation can be irrigated with the District's
allocation of Colorado River water based on the
definition of consumptive use contained in the
Supreme Court Decree (Arizona vs. California)
and the cropping pattern that prevailed in 1970,
1971, and 1972.
Irrigation Efficiency Improvement
Impacts
A parametric study of desalting costs was
undertaken to evaluate the relationship
between desalting costs and volumes of return
flow. In the study, the total range of probable
desalting costs was explored. Design level
parameters, such as plant recovery rates, TDS,
and ionic water composition changes were ex-
amined over a broad range of drain flows to
determine general cost sensitivities. Desalting
cost variability to select levels of irrigation
efficiency, river salinity levels, and long-term
aquifer changes was also examined. The signifi-
cant results of the cost study are presented in the
following tabulation:
TABLE 4
Desalting costs for selected program levels
Federal Cost
Return Annual
Level Efficiency Flow Capital1 Equivalent-
(1,000
(Percent) AF/Year) ($X106) ($X106)
1
2
3
64
69
72
167
147
136
91.1
94.2
80.9
14.5
13.1
12.4
'For a representative, reverse osmosis membrane desalting
process based on 1973 prices.
-Includes interest at 5-5/ 8 percent for 50 years on installa-
tion cost and all operation, maintenance, replacement, and
power costs.
Cost Sharing for Onfarm Systems
Improvements
The successful implementation of an on-
farm program to provide high offsite benefits
requires cost sharing incentives together with
technical assistance to assure proper installa-
tion. The extent of cost sharing required in-
volved three basic considerations:
1. The Federal benefits versus the owner
benefits from the improvements.
2. The ability of the farmer to pay for the re-
quired improvements.
3. The short time period to install the required
improvements and realize the associated reduc-
tion in return flows.
Crop budgets were used to evaluate
economic incentives to farmers and to deter-
mine the benefits farmers would receive from
implementation of an onfarm improvement
program. The crop budget analysis included
determination of gross returns from an average
type and size farm operation in the District for
various management levels. Production costs
were subtracted from the gross returns to obtain
net returns. It was assumed that installation of
structural measures would raise management
levels, resulting in increased net returns as
benefits accruing to the farmer. Composite or
average farms in the valley area and on the
adjacent mesa were used to construct a net
346
-------
EFFICIENCIES AND RETURN FLOWS — GILA PROJECT
return per acre for each management level
across the entire project area. The incremental
increase in net returns that the average farmer
would realize by moving from a low manage-
ment level, or a medium management level, to a
high management level were applied to the
acres in each category to be served by the
program to obtain the total increase in net farm
returns.
Federal benefits are the savings in
desalting costs. Incremental drainage reduc-
tions were computed for each management level
and multiplied by the incremental desalting
plant savings per acre-foot to obtain desalting
cost savings in terms of acres. These figures
were then multiplied by the acres to be served to
obtain the total Federal benefits. The procedure
utilized indicated that the ratio of farmer
benefits to total benefits would be about 30
percent. The crop budget process also indicated
that the increase in net income would be suf-
ficient to assist the farmer in paying for his
share of the improvements.
Several cost sharing arrangements were
considered for implementing the program. It
was determined that procedures similar to those
used previously by the Department of
Agriculture to solve problems associated with
high offsite benefits should be utilized. The
contract concept involving a conservation plan
developed with the farmer and an agreement
between the farmer and the Federal Govern-
ment to carry out the program over a period of
years appeared the most appropriate. It was
also believed that a successful program would
require 70 to 75 percent Federal cost sharing to
assure results on a timely basis. Program Level
3 appeared attainable with agressive manage-
ment improvement, technical assistance,
research and demonstrations, and an intensive
education program.
CONCLUSIONS AND
RECOMMENDATIONS
Conclusions
It was concluded by the Advisory Com-
mittee that the highest level of irrigation ef-
ficiency capable of achievement prior to the
desalting plant startup was a desirable goal.
This was identified as Program Level 3 and has
a cost-effective ratio of 6:1. Program Level 4,
although projected as effective in reducing
desalting cost was believed not practical of
achievement in the time period required. The
calculated return flow for Program Level 3 of
136,000 acre-feet per year is the objective. Irriga-
tion efficiencies of 64 percent on the mesa lands
and 74 percent on the valley lands are projected.
The average District wide irrigation efficiency
is 72 percent. The incremental reduction in
desalting costs compared to the estimated costs
of measures to reduce return flows are shown in
Figure 5.
a
i
-
o
REDUCTION IN
DESALTINGS PLANT
COSTS
IRRIGATION
EFFICIENCY
IN PERCENT-
9
1
COSTS FOR
REDUCING RETURN
FLOWS
Figure 5.
COST COMPARISON
INCREMENTAL REDUCTION IN DESALTING
PLANT COSTS COMPARED TO COSTS
FOR REDUCING RETURN FLOWS
If the measures under Program Level 3 are
fully implemented and if the effectiveness of
these improvements increase irrigation efficien-
cies as projected in these studies, return flows
should eventually reduce to 136,000 acre-feet per
year. However, there are uncertainties in
reaching this objective. They are (1) imprecision
in the calculations of return flow quantities, (2)
reservations that not all projected onfarm im-
provements will be fully implemented prior to
the plant startup date or that the measures will
be as effective in increasing irrigation efficien-
cies as estimated, and (3) changes in cropping
347
-------
CASE STUDY: WELLTON-MOHAWK DISTRICT
patterns. The Advisory Committee was highly
confident that return flows due to drainage
requirements would be reduced to 175,000 acre-
feet per year and considered it reasonable to
expect that return flows would be reduced to
150,000 acre-feet per year by the early 1980's.
It was further concluded that the irrigable
acreage in the District should be reduced by
approximately 10,000 acres to preclude ad-
ditional development that might increase
drainage return flows and to insure that water
use does not exceed the District's Colorado River
allocation.
Recommendations
The Advisory Committee in its Special
Report 5 made the following recommendations:
1. Acreage Reduction. The irrigable acreage
in the District should be reduced from the
authorized 75,000 acres by approximately
10,000 acres at an estimated cost of $9.5 million.
This is in accordance with provisions of Public
Law 93-320.
2. Program Level 3. It is recommended that
Program Level 3, which includes the following
items, be implemented:
a. Onfarm Measures. Install onfarm
structural features to improve gravity
irrigation systems on 19,800 acres, convert
to pressure systems on about 4,000 acres of
mesa land for use on citrus. Continued
limited use of portable sprinkler pressure
systems to germinate lettuce is anticipated.
Apply a cost sharing of 75 percent Federal
participation and 25 percent land owner
participation to all onfarm structural
measures, except portable sprinkler
systems. Consideration should be given to
continuing similar cost sharing
arrangements after the desalting plant
startup date to take advantage of oppor-
tunities that may exist to further reduce
return flows.
b. Improved Management. Expand the
IMS program to include insofar as prac-
ticable most of the irrigated acreage in the
District.
c. Technical Assistance. Provide assis-
tance by the SCS through cooperation with
the Wellton-Mohawk Natural Resource
Conservation District to achieve implemen-
tation of Program Level 3.
d. Research and Demonstra-
tion. Conduct research and demonstra-
tions to evaluate and demonstrate
measures to increase irrigation efficiency.
e. Education and Information. Develop
and conduct programs to disseminate infor-
mation to the water users so that they can
utilize the improved systems and the
technical and management assistance
provided under Program Level 3.
f. Cost. The Federal share of the cost of
Program Level 3 was estimated to be $7.5
million.
3. Distribution System Improvements.
Deficiencies in the District distribution system
that limit attaining higher irrigation efficien-
cies should be corrected and means found for
financing required improvements.
4. Water Pricing Policy. The District should
review the rate structure annually and set a rate
for supplemental water sufficiently high to act
as a deterrant against over-applications of
irrigation water.
5. Operation and Maintenance of the Irriga-
tion System. The District should improve its
quality of operation and maintenance to the
highest practical level for the efficient distribu-
tion and accurate accounting of water.
5Advisory Committee on Irrigation Efficiency,
Wellton-Mohawk Irrigation and Drainage District.
September 1974. Special Report Measures for Reduc-
ing Flows from the Wellton-Mohawk Irrigation and
Drainage District. U.S. Bureau of Reclamation.
348
-------
Wei Iton-Mohawk
On-Farm Systems
Improvement Program
R. S. SWENSON
Soil Conservation Service,
USDA, Phoenix, Arizona
ABSTRACT
The objective of the SCS in the Wellton-
Mohawk Irrigation Improvement Program is to
increase irrigation efficiency on farms in the
district and thereby reduce return flows from
the district. To accomplish this, the SCS enters
into contracts with eligible landowners and
operators to implement conservation practices
that will further the program goals. Cost shar-
ing is provided on a 75% federal, 25% cooperator
basis.
Authority for U.S. Department of
Agriculture and Soil Conservation Service par-
ticipation in the Wellton-Mohawk Irrigation
Improvement Program (Wellton-Mohawk On-
Farm Systems Improvement Program) is con-
tained in Title I of Public Law 93-320, the
Memorandum of Understanding between the
Departments of Interior and Agriculture, and
the Memorandum of Agreement between the
Bureau of Reclamation and the Soil Conserva-
tion Service.
The objectives of the program are to in-
crease irrigation efficiency on farms in the
Wellton-Mohawk Irrigation and Drainage Dis-
trict and reduce return flows from the District.
The Soil Conservation Service enters into
contracts with eligible landowners and
operators (cooperators) to install conservation
practices that will directly contribute to the
objectives of the program.
A cost-share rate of 75 percent Federal and
25 percent cooperator was established by the
Secretary of Interior. Funds to cover the costs of
contracts and technical services are transferred
from the Bureau of Reclamation to the Soil
Conservation Service under the Memorandum
of Agreement.
Lands eligible for the program must be
irrigated farmland within the Wellton-Mohawk
District. Land being considered for purchase by
the Bureau of Reclamation under the acreage
reduction program is excluded.
A priority system for assisting applicants
under the program was established by the Soil
Conservation Service. Those lands with low
efficiency are high priority.
As of January 1,1977, ninety-four of the one
hundred sixteen applications were on a high
priority list.
Irrigation efficiency in the district averaged
56 percent during the three-year period 1970
through 1972 (33 percent on the sandy mesa
lands and 65 percent on the fine-textured valley
lands). The goal of the program is to raise the
average efficiency in the District to 72
percent. (1)
Implementation of the program requires
input from several disciplines. The soil scientist
surveys the soils, and prepares a detailed soil
map. The soil conservationist develops treat-
ment alternatives and assists the cooperator
select those alternatives that will contribute to
program objectives. The engineer designs the
structural practices and checks for compliance
with contract specifications. Technicians make
planning and design surveys and inspect con-
struction. All staff members assist farmers with
irrigation water management. Practices are
installed by private contractors.
The contract specifies the practices to be
installed and the timing of the installation. The
contract extends two years beyond the year in
349
-------
CASE STUDY: WELLTON-MOHAWK DISTRICT
which .he last cost-shared practice is installed.
This if to permit application of a water manage-
ment system and to evaluate the results in terms
of program objectives.
The design delivery of water from the
irrigation district lateral to the farm headgate is
fifteen cubic feet per second (c.f .s.). Two types of
farm irrigation systems are being installed;
flood and pressure pipeline. All the flood
systems designed to date provide for applying
the entire head of available water to a single
level border. The borders are sized to enable a
minimum 3-inch water application to be made.
So far in our experience, the border sizes have
varied from 21/2 to 20 acres depending upon the
soil characteristics.
An integral part of the flood system is the
water measuring device. During an irrigation it
is important that the irrigator readily be able to
determine the volume of water so that he may
make adjustments in time of application. The
irrigator has available to him a "farm irrigation
guide" from which he can determine the time
needed to apply a specific amount of water on
each border.
A trapezoidal type critical flow depth flume
developed by the Agricultural Research Service
is the measuring device that was selected for use
in the program. A staff gauge installed on the
side of the flume can be read by the irrigator to
readily determine water volume.
Water is delivered to the borders either
through ports or turnout structures. Typical
ports are 16-inch concrete pipe with a slide gate
inlet. Three ports will normally handle a 15 c.f.s.
head of water.
Where feasible, the entire head of water is
delivered to a border through a single large
turnout structure. Water delivery is regulated
through a 46-inch wide jackgate. The structure
is flared from the gate to a width of 15 to 20 feet.
The field edge of the structure is usually beyond
the ditch berm and poses no obstacle to farming
operations.
Two types of pressure pipeline systems that
qualify for the program are sprinkler and drip
irrigation. A system is selected on an individual
basis depending upon the land capability, cost,
and farmer's choice. One drip system has been
installed to irrigate mature orange trees and
young lemon trees on sandy soils. The system
components of the 100-acre system include a
regulating reservoir, two miles of polyvinyl
chloride pipe and 109 miles of polyethylene tub-
ing. To date no sprinkler systems have been in-
stalled.
On flood systems, precision leveling of the
field surface is a must if we are to achieve the
degree of water control necessary to achieve
high irrigation efficiency. Fields are surveyed
on a 100-foot grid and designed with a flat
grade. Construction specifications call for a
tolerance of one-tenth foot from high to low on
80 percent of the field shots. Most of the earth is
moved with heavy construction equipment.
Finish leveling is accomplished by using a
laser controlled system. The control unit is a
laser beam that rotates five to ten times a
second. A receiving unit is mounted on a drag
scraper which activates the scraper's hydraulic
system to control the elevation of the blade.
Fields checked after laser leveling are within
one-tenth foot from high to low on all shots.
Some have checked to within five hundredths
foot from high to low.
Soils within the Wellton-Mohawk Irriga-
tion and Drainage District vary in texture from
fine sand to clay, often within the same field.
Where this condition exists water intake rates
obviously vary.
On fields where this is a problem, the areas
of problem soils are excavated to a depth of 18
inches minimum and back filled with suitable
soil. The excavated material is either wasted off
site or mixed with soils on the rest of the field to
achieve uniformity.
The final and perhaps most critical step in
achieving high irrigation efficiency is the preci-
sion handling of the water. Good irrigation
water management requires that the irrigator
have a practical knowledge of soils, consump-
tive use, flow rates, basin size, crop response,
gross and net water application needed and
frequency of irrigation. An irrigation water
management plan is developed and made a part
of each contract. The plan includes a table list-
ing the application time needed for each unit to
be irrigated depending upon the quantity of
water available (Table 1). Another chart
describes the normal irrigations required per
month and the net amount of each irrigation.
There is only limited data available on the
effects of the program because only a few
farmers have completed the on-farm in-
stallations and subsequently completed crop
seasons on the same land. As of July 30, 1976,
350
-------
ON-FARM SYSTEMS IMPROVEMENT
the following efficiencies are shown for farms
having completed their treatment.
CROP ACREAGE
PRE-
TREATMENT
1974-75
AVERAGE
EFFICIENCY
POST
TREATMENT
1976
EFFICIENCY
Wheat
Alfalfa
293
216
52%
65%
79%
92%
REFERENCES
1. The Advisory Committee on Irrigation
Efficiency, Wellton-Mohawk Irrigation and
Drainage District. 1974. Special Report on
Measures for Reducing Return Flows from the
Wellton-Mohawk Irrigation and Drainage Dis-
trict.
2. Soil Conservation Service National
Engineering Handbook, Section 15, Irrigation,
Chapter 4, Border Irrigation.
TABLE 1
Farm Irrigation Guide, October, 1975
Field: 1, 13.1 Acre Level Borders
NET APPLICA TION
Field: 2, 12.3 Acre Level Borders
NET APPLICATION
CFS hr
12 3
13 3
14 3
15 2
16 2
17 2
18 2
3"
min
40
20
05
55
45
35
25
0.3
4"
Inch
hr min
4
4
4
3
3
3
3
50
30
10
50
40
25
15
Intake Family
5"
hr min
6 05
5 35
5 10
4 50
4 30
4 15
4 00
hr
7
6
6
5
5
5
4
6"
min
15
45
15
50
25
10
SO
hr
4
3
3
3
3
2
2
3"
min
10
45
30
10
00
45
35
hr
5
4
4
3
3
3
3
4"
min
05
35
15
55
40
25
10
0.5 I)
hr
6
5
5
4
4
4
3
ich Into
5"
min
00
30
00
40
20
05
45
ikel
hr
6
6
5
5
5
4
4
family
6"
min
50
20
50
25
05
50
35
"Developed using Soil Conservation
Service National Engineering
Handbook, Section 15, Irrigation
Chapter 4, Border Irrigation" (2)
351
-------
Research and Demonstration
Approach to Development
of Appropriate Salinity Control
Technologies for Grand Valley
GAYLORD V. SKOGERBOE and WYNN R. WALKER
Department of Agricultural and Chemical Engineering,
Colorado State University, Fort Collins, Colorado
ABSTRACT
Introduction of channel seepage and irriga-
tion percolation losses into the underlying soils
and marine aquifer, and the eventual return of
these flows to the Colorado River with their
large salt loads, make the Grand Valley in
western Colorado one of the more significant
salinity sources in the Upper Colorado River
Basin. The Grand Valley Salinity Control
Demonstration Project was formulated to
delineate the magnitude of the water and salt
flow components from the irrigation system, to
evaluate the effectiveness of various water
management technologies in reducing the salt
load reaching the Colorado River, and to
demonstrate appropriate technologies on
farmers' fields.
Colorado River Salinity
The most serious water quality problem in
the Colorado River Basin (Fig. 1) is salinity.
Often referred to as mineral quality or mineral
pollution, salinity concentrations progressively
deteriorate towards the lower reaches of the
river system resulting in significant economic
penalties to Lower Basin users.
Analysis of many years of data indicate the
bulk of the Lower Basin salinity problem is
attributable to the salt loads acquired in the
Upper Basin. Therefore, salinity management
in the Upper Basin must deal directly with
controlling the salt loads in the river con-
tributed by municipal, industrial, and
agricultural water uses as well as some natural
sources. However, most of the future water
developments planned in the Upper Basin are
Figure 1. The Colorado River Basin.
directed toward meeting the rapidly expanding
urban demands which lie outside of the
Colorado River Basin. Thus, the export of this
high quality water from the basin will
significantly increase salinity concentrations
throughout the basin by reducing the dilution
capacity of the stream. These developments will
have to be accompanied by associated reduc-
tions in the total salt load passing downstream
resulting from salinity control programs to
avoid further downstream damages.
353
-------
CASE STUDY: GRAND VALLEY
Although such salinity control alternatives
as reducing reservoir evaporation and
phreatophyte consumption will eventually be
considered, the first priority is the reduction of
salinity associated with agricultural return
flows, presently constituting 37% (Fig. 2) of the
total salt load from the Upper Colorado River
Basin (EPA, 1977). The major aspects include
conveyance channel linings, improving the on-
farm water use efficiency, more stringent water
controls by irrigation companies, and effective
coordination of local objectives between the
various organizations. The definitions and the
interpretation of western water right laws pose
a difficult obstacle to salinity control incentives
for water users and thus must also be evaluated.
Among the most significant salt sources in
the Upper Basin region is the Grand Valley
(Fig. 3) in west-central Colorado. The area was
selected for a detailed study of important
parameters in the hydro-salinity flow system in
order to quantify the sources of the valley salt
contribution to the stream system. The detail of
the investigation also made it possible to quan-
tify the salinity control feasibility of several
alternatives such as water conveyance channel
linings and irrigation scheduling.
UPPER COLORADO RIVER
BASIN
AVERAGE SALT LOAD TONS/DAY
Juiw 1965 - May 1966
NATURAL POINT SOURCES
AND WELLS
•
S5T T/«J AND
NOUSTRIAL
UPPER MAIN STEM
SUBBASIN
LOWER COLORADO RIVER
BASIN
AVERAGE SALT LOAD TONS/DAY
Nowvfct' '963 — Octoow 1964
NET
GREEN RIVER
SUBBASIN
LOWER MAIN STEM
SUBBASIN
SAN JUAN RIVER
SUBBASIN
Figure 2. Major sources of salinity in the Colorado
River Basin (U.S. Environmental Protection Agen-
cy, 1971).
Figure 3. Relative magnitude of agricultural salt
sources in the Colorado River Basin (U.S. Environ-
mental Protection Agency, 1971).
The relative contribution from the valley to
the total salt load in the upper river system
attributable to agriculture is approximately 18%
(Fig. 3), or between 0.6 and 0.9 million metric
tons of salt annually (Hyatt, 1970 and EPA,
1971). Usually called "salt pickup," this con-
tribution results from water contact with the
naturally saline soils and aquifers overlying a
marine deposited Mancos Shale formation. The
shale, which is also the origin of much of the
soils, contains numerous lenses of salt and can
be readily dissolved. The introduction of water
into these soils and aquifers from irrigation
sources produces hydraulic gradients which
displace groundwater into the river. Since this
displaced water has usually had sufficient con-
tact time to reach chemical equilibrium with the
ambient salinity concentrations, large quan-
tities of salts are added to the river.
General Approach
The Colorado River enters the Grand Valley
from the east, is joined by the Gunnison River at
Grand Junction, Colorado, and then exits to the
west. The primary source of salinity is from the
extremely saline aquifers overlying the marine
deposited Mancos shale formation. The shale is
characterized by lenses of salt in the formation
which are dissolved by water from excessive
irrigation and conveyance seepage losses when
354
-------
SALINITY CONTROL — GRAND VALLEY
it comes in contact with the Mancos shale
formation. The introduction of water through
these surface sources percolates into the
shallow groundwater reservoir where the
hydraulic gradients it produced displace some
water into the river. This displaced water usual-
ly has sufficient time to reach chemical
equilibrium with the salt concentrations of the
soils and shale. These factors also make the
Grand Valley an important study ar^a, since
the conditions encountered in the valley are
common to many locations in the basin.
In order to adequately define the water and
salt flow system, both in terms of surface flows
and subsurface flows, and taking into con-
sideration available resources, it was necessary
to designate a demonstration area. This area
could be more intensely investigated than if the
entire valley were measured for water and salt
transport. Therefore, one of the requirements for
such a demonstration area would be that it is
representative of the irrigated lands in the
Grand Valley.
Since salinity is really a problem involving
subsurface irrigation return flows, then con-
siderable knowledge must be gained regarding
the groundwater hydrology. Also, one of the
more difficult aspects of evaluating the effec-
tiveness of improved irrigation water manage-
ment practices for reducing the salt pickup
returning to the Colorado River is being able to
define the salt transport occurring in conjunc-
tion with moisture movement through the soil
profile.
Basically, reducing the salt loading from
Grand Valley will require a reduction in the
volume of groundwater flows returning to the
Colorado River. Consequently, the investiga-
tion must establish the sources of groundwater
flow (e.g., canal seepage, lateral seepage, deep
percolation losses, phreatophyte consumptive
use, etc.). The next paper, "The Hydro-Salinity
System in the Grand Valley" presents the
results of these studies.
After developing a knowledge of the surface
hydrology and subsurface hydrology, including
the sources of groundwater flows, the next
crucial step is to establish the chemical changes
that occur as moisture moves through the soil
profile into the groundwater reservoir. In Grand
Valley, this step was accomplished by using a
number of research plots which served as large-
scale lysimeters. These plots are discussed
below and the results are reported in the case
study paper, "Modeling Salt Transport in the
Irrigated Soils of Grand Valley."
With the above information, it then
becomes possible to evaluate the effectiveness of
a variety of water management technologies in
reducing the salt load reaching the Colorado
River. For Grand Valley, most of the salt load is
the result of present irrigation practices.
Therefore, the field studies must evaluate each
of the appropriate technologies for improving
irrigation water management by establishing
the reduction in groundwater flow for each
technology and then predicting the salt load
reduction in the Colorado River. By evaluating
such technologies on farmer's fields, it becomes
possible to gain some measure of the acceptabil-
ity of each technology, or combination of
technologies, by the farmer. The final two
papers in this case study, "Evaluating Ap-
propriate Technologies for Salinity Control in
Grand Valley" and "Development of Best
Management practices for Salinity Control in
Grand Valley" present the results of this phase
of the research and demonstration program.
Figure 4. The Grand Valley of Colorado.
The Demonstration Area
The demonstration area, shown in Fig. 4,
was chosen as an intensive study area in which
the bulk of the investigation was to be con-
ducted and also includes most of the construc-
tion and demonstration efforts. This area was
designated for detailed investigations regard-
ing various salinity control measures on the
water and salt flow systems in an irrigated area.
The intensive study area was selected for its
accessibility in isolating most of the important
hydrologic parameters, but also has the impor-
tant advantage that it allowed five irrigation
companies to participate in one unit.
355
-------
CASE STUDY: GRAND VALLEY
In undertaking the Grand Valley Salinity
Control Demonstration Project, one of the first
tasks was to conceptualize a hydro-salinity
model of the intensive study area (Fig. 5). This
model has to have sufficient sensitivity to detect
the effects of various salinity control measures
upon the salt pickup reaching the Colorado
River. Then, the model could be used to design
the field data collection system. Finally, the
model could be used to extrapolate results from
the intensive study area to the entire Grand
Valley.
A difficulty often encountered while prepar-
ing water and salt budgets is the variability in
the accuracy and reliability with which the
hydrologic and salinity parameters are
measured. Usually, the mesurement precision
varies with the scope of the research and the
!«*_,
Cn
1
ter Ftows
t Flows
Evaporation
I
(—
t i
Canal
Seepage
1
i
1
1
1
1
J
— *'
i
i
i \
1
1
1
1
1
t-i
'
* "
River
Outflows
Figure 5. Schematic of hydro-salinity model for
Grand Valley.
area of the study. The intensive study area on
this project has been observed in great detail.
Since the hydrologic system is difficult to
monitor and predict, it is impractical to expect
models to operate without applying some ad-
justments in order that all components will be in
balance. In short, the budgeting procedure is
usually the adjustment of the segments in the
water and salt flows according to a weighting of
the most reliable data until all parameters
represent the closest approximation of the area
that can be achieved with the input data being
used. The vast and lengthy computation
procedure of calculating budgets is facilitated
by a mathematical model programmed for a
digital computer. A complete listing and ex-
planation of its operation has been previously
reported (Walker, 1970).
The discussion below is intended to give the
reader some insight into the types of field
investigations that were undertaken in order to
develop a hydro-salinity model calibrated for
the Grand Valley Salinity Control Demonstra-
tion Project area, yet capable of predicting
water and salt flows for the entire Grand Valley.
Field Investigations
The effects of poor water management are
significant in the study area as 50% of the
irrigable acreage has been reduced to marginal
production because of high water tables. The
hydrologic factors responsible for these con-
ditions were that they could be studied in
reasonable detail. In this investigation, a
balance between physical size of the area, level
of funding, and the detail with which the system
parameters, could be studied resulted in an
experimental design consisting of two phases:
(i) instrumentation and (ii) peripheral in-
vestigations.
The instrumentation in the area (Fig. 6)
provided valuable data describing much of the
water and salt movement. Some parameters
such as drainage discharges, lateral diversions,
water quality, and precipitation could be
measured directly. Other parameters such as
groundwater movements, deep percolation, and
evapotranspiration were computed from data
supplied by piezometers, wells, soil sample
analyses, and climatologic data.
Many of the water and salt subsystems do
not readily lend themselves to direct measure-
ment and must be quantified from an examina-
356
-------
SALINITY CONTROL - GRAND VALLEY
tion of a sample of the area, in which it is
assumed that the sample is representative of the
whole. Such studies included farm efficiency in
which the proportions of evapotranspiration,
deep percolation, tailwater runoff, and soil
moisture storage were evaluated, along with the
resulting changes in salinity.
Instrumentation
The hydrostatic pressures through the soil
profiles, hydraulic gradients, and conduc-
tivities were determined by numerous clusters of
0.95-cm diameter steel pipe piezometers. In each
cluster, three to seven piezometers were located
at varying depths depending on the textural
changes encountered in the placement of the
first piezometer to as great a depth as possible.
The technique used to extend the piezometers
employed a jetting rig which pumped water at
moderate pressure through the pipe to jet
material away from the end of the piezometer.
Because the small piezometers were limited
to the soil profiles, the examination of water and
salt movements in the underlying cobble
aquifers was facilitated by drilling 5-cm
diameter wells into these strata. The bottoms of
the pipes extended to the top of the Mancos
Shale formation in most cases, but several wells
were extended only midway through the aquifer
or just into the top in order to test the differences
that may exist in the conditions in the aquifer.
The function of these instrument points was
essentially the same as that of the small
piezometers. The large diameters facilitated
both depth measurements and water sampling.
Small Parshall and Cutthroat flumes were
used in the study to monitor discharges in the
drainage system. They were also extensively
used as part of the special studies on farm
efficiency, lateral ditch seepage, and farm water
deliveries. At the most important locations,
flumes with continuous stage recorders were
installed to minimize the error in evaluating the
discharges. Both types of flumes involve
primarily free flow operation, but continual
maintenance was required to maintain this
condition as mossing, sediment, and bank
vegatation created considerable submergence
effects as the study progressed.
An essential element of the analysis was
the measurement of the water diverted from the
canal system into the farm laterals for irriga-
tion. In the smaller capacity canals where the
discharges rarely exceed 1.4 m^/sec, the inlets
and exits to the study area were rated by
correlating current water measurements with
the stage in the canal. One exception to this was
at the inlet to the Mesa County Ditch in which a
1.5-m submerged orfice was calibrated. The
discharges in two of the main canals were of the
order of 17 m 3 / sec and it was felt that even if a 2
to 5% inlet and exit rating could be developed,
this error would mask most of the diversions to
the cropland. To remedy this situation, the farm
turnouts from each canal were individually
rated. Prior to the irrigation season, the relation
between thread rod height and gate opening for
each of the culvert-type turnouts was deter-
mined. The rating procedure involved the place-
ment of a portable Parshall flume downstream
from the gate to measure the flow, thereby
producing a rating curve with less than a man-
day of effort.
Peripheral Investigations
A number of segments within the water and
salt flow system could not be measured with the
available instrumentation and special studies
were required. The first such special study was
the determination of the seepage rates from the
canal and lateral systems. The ponding method,
which involves diking the canal or lateral into
several ponds and then correlating the falling
water surfaces to seepage rates, was selected to
assure accurate measurements. However, in
many of the smaller laterals the inflow-outflow
method was employed. The use of the two
methods depended on the discharge capacity of
the conveyance channel.
In order to evaluate the quantity of water
being evaporated and transpired from the soil
and crop surfaces, a detailed land use inventory
of the Grand Valley was conducted. The
procedure involved carrying aerial photos (1 cm
= 120 m) into the field and making note of the
various land uses on them. Then each crop area
was determined by measuring the respective
area on the photo (Walker and Skogerboe, 1971).
The water and salts being applied to the
cropland either result in tailwater, deep percola-
tion, or root zone storage. To delineate these
components, farm efficiency studies were con-
ducted on typical farms. Inflow-outflow
measurements and soil moisture samples were
taken to identify the water and salt movements.
Other smaller studies were made to deter-
mine the effectiveness of the existing drainage
system and the subsurface soil and aquifer
357
-------
CASE STUDY: GRAND VALLEY
characteristics. These investigations proved to
be helpful in understanding the hydrologic
network and added confidence to the results.
Soil Chemistry Research Plots
The hydro-salinity model describes the pre-
sent situation in the study area regarding water
and salt flows. However, the only method for
predicting the reduction due to salinity control
measure(s) is by assuming a one-to-one
relationship between water and salt. That is, if
the subsurface return flow is reduced by 50%, the
salt is also reduced by 50%. In order to overcome
this limitation, a project "Irrigation Practices,
Return Flow Salinity, and Crop Yields" was
initiated.
Three adjacent fields containing 23 acres
was leased for this study. The area was divided
into 54 plots which are 100 feet by 100 feet in
size, two plots which are 40 feet by 200 feet, two
plots which are 40 feet by 300 feet, and five plots
which are 40 feet by 500 feet. Each plot is used
for a different replication of the crop, fertilizer
and irrigation treatments. They were con-
structed so that each plot performs as a large
lysimeter. A trench was excavated slightly into
the shale along the lines dividing the plots. A
plastic curtain was then placed vertically in the
center of the trench to divide the individual
plots. The lower edge of the curtain is "sealed " to
the shale by back-filling to the original eleva-
tion of the shale with compacted clay. The
average depth to shale is approximately ten
feet.
The drainline encased in a gravel filter
material was then placed inside the curtain and
continued around the periphery of the plot.
Upon leaving the plot area, the water is
transported via solid pipeline to a measuring
station where water quality and quantity is
monitored.
The irrigation system is designed to deliver
water through a closed conduit to each plot and
allow measurement of the flows onto each plot.
Since furrow irrigation is used almost exclusive-
ly throughout the valley, this method has been
employed on the project area.
The crops being grown are corn, grass,
alfalfa, and winter wheat, since these are the
main crops grown commercially in the valley.
By varying irrigation timing and amounts,
crops, and nitrogen fertilizer levels on the
different plots, and by monitoring quality and
quantity of both inflow and outflow waters, the
effects of these parameters on return flow salini-
ty and crop yields can be evaluated.
The third paper in the case study, "Model-
ing Salt Transport in the Irrigated Soils of
Grand Valley" reports the results of this phase
of the research program.
Demonstration of Salinity Control
Measures
The principal study area in Grand Valley
(Fig. 4), was first used for evaluating the effec-
tiveness of canal and lateral lining, as well as
irrigation scheduling and tile drainage in reduc-
ing the salt load entering the Colorado River.
Later, the same area was used as a demonstra-
tion project during the 1974, 1975, and 1976
irrigation seasons. The advantage in con-
tinuing to utilize this study area is that the
hydrology is already known. The wealth of
available information provides a strong basis
for evaluating the effectiveness of salinity con-
trol measures.
With the available knowledge regarding the
study area, a lateral including the associated
land served by the lateral water supply was used
as a sub-system for evaluating the salinity
reduction in the Colorado River resulting from
the implementation of a salinity control
technology package. In order to facilitate con-
tinued participation by most irrigation interests
in Grand Valley, this demonstration project
utilized laterals under each of the five canals in
the study area. These particular laterals were
selected to represent a wide variety of irrigation
and drainage problems.
The selection of a lateral as a sub-system,
rather than an individual farm, has a tremen-
dous advantage in allowing control at the
lateral turnout. In this way, both the quantity of
flow and the time of water delivery can be
controlled, thereby facilitating improved water
management throughout the subsystem.
The most significant aspect of this par-
ticular demonstration project was the employ-
ment of a salinity control technology "pack-
age", rather than a single control measure.
Experience in Grand Valley has shown that the
most significant progress is made when the
gamut of questions can be answered regarding
the interrelationships between water manage-
ment and agricultural production. Thus, the
concept of a technology package, along with an
358
-------
SALINITY CONTROL — GRAND VALLEY
understanding of the "system" including other
agricultural inputs, provides the necessary base
for providing sound advice to the farmer, which
in turn facilitates the development of credibility
and consequently farmer acceptance.
This phase of the research and demonstra-
tion program is reported in the fourth paper of
this case study, "Evaluating Appropriate
Technologies for Salinity Control in Grand
Valley."
Best Management Practices
The results from the demonstration project
will be projected to valley-wide conditions in
developing best management practices for
salinity control in Grand Valley. This will
integrate several years of concentrated study in
the valley and will serve as a basis for an action
salinity control program. Such a program would
detail the optimal strategy for implementing
various levels of individual salinity control
measures into a comprehensive technology
package. To develop this kind of policy, cost-
effectiveness functions relating the reductions
in the system salt loadings resulting from a
specified investment would be individually
assessed in an optimizational format to arrive
at the least cost combination for achieving a
desired level of salinity control. Since salinity
control in Grand Valley must evolve with the
development of water resources in the Upper
Colorado River Basin, the time-varying
characteristics of salinity control strategies
must be described. The results of this optimiza-
tion are described in the final paper of this case
study, "Development of Best Management
Practices for Salinity Control in Grand Valley."
REFERENCES
1. Hyatt, M. Leon. 1970. Analog computer
model of the hydrologic and salinity flow of
systems within the Upper Colorado River
Basin. Ph.D. dissertation, Department of Civil
Engineering, College of Engineering, Utah
State University, Logan, Utah. July.
2. U.S. Environmental Protection Agen-
cy. 1971. The mineral quality problem in the
Colorado River Basin. Summary Report and
Appendices A, B, C, and D.
3. Walker, W. R. 1970. Hydro-salinity
model of the Grand Valley. M.S. Thesis CET-
71WRW8. Civil Engineering Department,
College of Engineering, Colorado State Univer-
sity, Fort Collins, Colorado. August.
4. Walker, W. R., and Gaylord V. Skoger-
boe. 1971. Agricultural land use in the Grand
Valley. Report AE71-746 WRW-GVSl, Agricul-
tural Engineering Department., College of En-
gineering, Colorado State University, Fort
Collins, Colorado.
359
-------
The Hydro-Salinity System
in the Grand Valley
WYNN R. WALKER, GAYLORD V. SKOGERBOE, ROBERT G. EVANS,
and STEPHEN W. SMITH
Department of Agricultural and Chemical Engineering;
Colorado State University; Fort Collins, Colorado.
ABSTRACT
The Grand Valley hydro-salinity system is
described by research data collected over an
eight-year period. A review of previous es-
timates indicates the great variability that can
be expected when data are few and of poor
quality. The Grand Valley system contributes
approximately 630,000 metric tons of salts to the
river annually. This figure represents 78 metric
tons of salt per hectare-meter of water reaching
the underlying groundwater aquifer from irriga-
tion return flows with canal and ditch seepage
contributing 23 percent of the salt pickup,
lateral seepage 32 percent, and on-farm head
ditch seepage and deep percolation losses con-
tributing the remaining 45 percent of the salt
pickup.
INTRODUCTION
In the mid-1960's, a concerted effort was
undertaken to identify the sources of salinity in
the Colorado River Basin. The Federal Water
Pollution Control Administration, then within
the Interior Department, utilized U.S.
Geological Survey stream gaging data as well
as an extensive water quality sampling
program to identify the major salt contributors
in the basin (U.S. Environmental Protection
Agency, 1971). The Grand Valley in western
Colorado was described as one of the largest
agricultural sources of salinity (about 18% of the
total Upper Basin's agriculturally related con-
tribution), and it subsequently became the site
for the first studies to evaluate field-scale salini-
ty control measures. The involvement by
Colorado State University has been sum-
marized by Skogerboe and Walker (1972) and
Skogerboe et al. (1974a).
The early studies identified the Grand Val-
ley as a major problem area by mass balanc-
ing water and salt flows into and out of the
valley region. Similar investigations by lorns et
al. (1965) and Hyatt et al. (1970) produced
supportive results, although a substantial
variability in the specific nature of the Grand
Valley salinity problem emerged. Since that
time, a great many individual calculations
describing the valley salt loading and the
respective components have been made, but
very little agreement existed until early 1977
when most studies were completed. Although
some variability still exists among the various
investigative groups, the differences are suf-
ficiently close that they become relatively unim-
portant in determing the overall Best Manage-
ment Practices for the Grand Valley as will be
shown in the last paper of this case study.
The purpose of this paper is to describe the
Grand Valley hydro-salinity system as existing
data defines it, and in so doing, hopefully
provide some insight into how various data
might be interpreted in agriculturally related
water quality studies.
HYDRO-SALINITY BUDGETING
The procedures for delineating the water-
salt flow system in an irrigated area is termed
hydro-salinity budgeting, or hydro-salinity
modeling, since computers are generally needed
to handle the large number of necessary
calculations. In the Grand Valley, the composi-
tion of this system has been extensively in-
vestigated at various levels of sophistication.
As the salinity investigations were continued,
refinements in the valley's basin-wide impact
have been made and verified. Interestingly, the
research evolution in the Grand Valley case
study suggests a fairly sound approach for other
areas as well.
361
-------
CASE STUDY. GRAND VALLEY
The problem of remedying an irrigation
return flow problem causing detrimental water
quality deterioration can be divided into four
logical steps. First, the magnitude of the
problem and the downstream consequences
must be identified in relation to the irrigated
area's individual contribution to the problem. In
this way, the most important areas can be
delineated for further consideration, thereby
making the most cost-effective use of available
personnel and funding resources. As noted
previously, this step led to the exhaustive efforts
in the Grand Valley that this case study reports.
Next, the components of the problem must be
segregated. In most large areas, the cost of
studying the entire system are prohibitive, so
smaller "sampling" studies are conducted from
which projections are made to predict the
behavior of the entire area. The next step is to
evaluate management alternatives on a
prototype scale in order to assess their cost-
effectiveness and develop a sensitivity about
the capability for implementing such
technologies. And finally, if the measures which
can be applied are effective in reducing salinity
and economically feasible, the final step is the
actual application of the technology to solving
the water quality problem. This paper sum-
marizes the first two of these steps as they
developed in the Grand Valley.
PROBLEM IDENTIFICATION
The Grand Valley was identified as an
important agricultural source of salinity in the
Colorado River Basin through a series of
analyses involving mass balance of the valley
inflows and outflows. loms et al. (1965)
evaluated stream gaging records for the 1914 to
1957 period, concluding that net salt loading
(salt pickup) from irrigation in the valley
ranged from about 450,000 to 800,000 metric
tons annually. This range of numbers has been
generated independently by Hyatt et al. (1970),
Skogerboe and Walker (1972), Westesen (1975),
and the U.S. Geological Survey (1976). More
recent consideration of data by the writers and
others indicates a long-term salt pickup rate
between 600,000 to 700,000 m tons/year. This
figure is now generally accepted by the various
research groups and action agencies involved
with Grand Valley salinity investigations.
The fact that the valley's salinity contribu-
tion has been such a disputed figure over the last
five years exemplifies the importance of es-
tablishing the total valley contribution. In
areas like the Grand Valley where the total
valley impact is only 5-8% of the river inflows or
outflows, the impact of irrigation must be estab-
lished using statistical analyses of the available
data. However, the natural variability can
cause serious errors in conclusions regarding
salt pickup if not tempered by other data. For
example, a major problem in early in-
vestigations was deciding how much of the
inflow-outflow differences were due to natural
runoff from the surrounding watershed.
Because of the meager precipitation locally, the
writers assumed the natural salt contribution
would be negligible. This conclusion was later
substantiated partially by Elkin (1976) who
estimated an upward limit for the natural
contribution of about 100% of agricultural
figures.
SEGREGATING THE IRRIGATION
RETURN FLOW SYSTEM
In the Grand Valley, as in numerous other
irrigated areas, water is supplied to the cropland
in a canal, ditch, and lateral conveyance
system. Water is diverted from the Colorado and
Gunnison Rivers into three major canals: (1) the
Government Highline Canal; (2) the Grand
Valley Canal; and (3) the Redlands Power
Canal. These large canals in turn supply the
smaller canals and ditches as follows:
Government
Highline
Canal
Stub
Price
Grand Valley
Canal
G. V. Mainline
G. V. Highline
Redlands
Power
Canal
Redlands #1
Redlands #2
Orchard Mesa Power Mesa County
Orchard Mesa #1 Kiefer Extension
Orchard Mesa #2 Independent Ranchmen's
A description of the hydraulic characteristics of
these canals and ditches is given in Table 1,
based on information provided by the Bureau of
Reclamation. From the canals and ditches,
water is diverted into the small, largely earth
ditches leading to the individual fields. This
lateral system involves approximately 600
kilometers of ditch carrying from 0.06 m 3 /sec
to 1 m3 /sec. A frequency distribution of the
lateral lengths based on data provided by the
Bureau of Reclamation is given in Figure 1
indicating that the average length is about 400
meters. The average capacity is about 0.10
m 3 /sec.
362
-------
HYDROSALINITY SYSTEM - GRAND VALLEY
Nearly all fanners in the valley apply water
using the furrow irrigation method. The Soil
Conservation Service extensively inventoried
the valley's irrigation system and were kind
enough to make that data available for use
herein. There are over 9000 individual fields in
the Grand Valley having widths, slopes and
lengths as distributed in Figures 2, 3, and 4.
Thus, the typical field is 140 meters wide, 160
meters long, with a slope (toward the south
generally) of 1.125%. A frequency distribution of
field acreages is given in Figure 5 showing the
typical field encompassing about 2 hectares.
Calculating the length of unlined field head
ditches based on the SCS data indicates a total
length of 1300 kilometers. Tail water ditch
length would approximate head ditch length.
TABLE 1
Hydraulic characteristics of the Grand Valley
canal and ditch system.
Name
Government Highline Canal
Grand Valley Canal
Giand Valley Mainline
Grand Valley Highline
Kiefer Extension of
Grand Valley
Mesa County Ditch
Independent Ranchmen's
Ditch
Price Ditch
Stub Ditch
Orchard Mesa Power Canal
Orchard Mesa til Canal
Orchard Mesa #2 Canal
Redlands Power Canal
Redlands #1 & VI Canals
Length
(km)
73.70
19.80
21.70
37.00
24.50
4.00
17.40
9.50
11.30
3.90
24.10
26.10
2.90
10.80
Initial
Capacity
(m'/sec)
16.99
18.41
7.08
8.50
3.96
1.13
1.98
2.83
0.85
24.07
3.12
1.98
24.07
1.70
Terminal
Capacity
fmVsec;
0.71
14.16
0.71
3.96
0.71
0.06
0.85
0.28
0.11
24.07
0.17
0.17
24.07
0.06
Initial
Perimenter
(m)
19.19
16.67
13.86
12.62
7.25
6.67
3.17
7.27
2.94
18.20
6.46
3.58
16.88
3.95
Irrigation water is applied to approximate-
ly 25,000 hectares during the course of a normal
irrigation season (Walker and Skogerboe, 1971).
This acreage has been substantiated by the
recent SCS inventory and generally accepted by
the other agencies. A graphical breakdown of
the acreage and miscellaneous land use in the
valley is given in Figure 6.
Based upon lysimeter data reported by
Walker et al. (1976) the weighted average con-
sumptive use demand by the irrigated portion of
the area equals about 0.745 meters per season.
These evapotranspiration estimates when
applied to all of the respective demands in the
Grand Valley are indicated in Table 2.
The irrigation return flow system in the
valley may be divided according to whether or
0 2 4 6 6 10 12 14 16 18 20 22
0 80O 1600 2400 1200 4000 4800 5600 6400 7200
Figure 1. Frequency distribution of lateral lengths
in the Grand Valley.
0 200 400 600 800 1000 1200 1400 1600
Figure 2. Frequency distribution of field widths in
the Grand Valley.
0 OS 1.0 1.5 2.0 2.0 30 i5 40 45
Figure 3. Frequency distribution of field slopes in
the Grand Valley.
not the return flows are surface or subsurface
flows. Surface flows occur as either field
tailwater or canal, ditch, and lateral spillage.
Subsurface flows include canal and ditch
seepage, lateral seepage, and deep percolation
from on-farm water applications (deep percola-
363
-------
CASE STUDY: GRAND VALLEY
TABLE 2
Figure 4. Frequency distribution of field lengths in
the Grand Valley.
Consumptive use estimated for the Grand Valley.
Consumptive Use
Volume Depth
in ha-m in Meters
Open water surface evapora-
tion and phreatophyte use1
Open water surface evapora-
tion and phreatophyte use2
Cropland
TOTAL
3,450 0.138
8,400 0.336
18.600 0.745
30.450 1.219
1 adjacent to river
2along canals and drains
0 5 10 15 20 25 SO 35 »0 «5
Figure 5. Frequency distribution of field areas in
the Grand Valley.
tion includes head ditch and tailwater ditch
seepage).
Canal and Ditch Seepage
Since the early 1950's, five major seepage
investigations on the valley's major canals and
ditches have been conducted (Skogerboe and
Walker, 1972 and Duke et al., 1976). Seepage
rates have been noted over a wide range, but
representative rates are presented for the four-
teen canal systems in Table 3. Walker (1977)
developed a mathematical relationship describ-
ing seepage:
IZO
100
20 -
Figure 6. Land use in the Grand Valley.
V8 =
• WPm «Sr *L
2b
[1-U-b)']
(1)
in which
V 8 = seepage volume of a canal in mVyear,
N d = number of days in operation per year,
WP m = inlet wetted perimeter, m;
S r — seepage rate, mVmVday;
L = canal length in m; and
b = empirical constant representing the frac-
tion of maximum wetted perimeter re-
maining at the end of the canal
If the values in Table 2 are substituted into Eq.
(1), the seepage volume for each canal and ditch
can be determined (Table 1). In the Grand
Valley, the writers estimate canal seepage to be
approximately 3,700 ha-m per year.
364
-------
HYDROSALINITY SYSTEM - GRAND VALLEY
TABLE 3
Seepage data for the fourteen major
canal systems in the Grand Valley.
Name of Canal or Ditch
Government Highline
Grand Valley
Grand Valley Mainline
Grand Valley Highline
Kiefer Extension
Mesa County
Independent Ranchmen's
Price
Stub
Orchard Mesa Power
Orchard Mesa #1
Orchard Mesa #2
Redlands Power
Redlandal&2
Seepage
Days in Seepage Rate Volume
Operation b mVmVdoy ha-m
214
214
214
214
214
214
214
214
214
365
214
214
365
214
0.80
0.17
0.69
0.29
0.60
0.64
0.31
0.65
0.44
0.001
0.72
0.62
0.001
0.67
0.091
0.045
0.061
0.061
0.061
0.061
0.061
0.061
0.061
0.076
0.076
0.076
0.065
0.137
1652.53
290.84
257.16
521.16
162.31
23.68
60.84
60.86
33.83
196.80
162.05
104.86
116.08
83.17
3726.17
Lateral Seepage
Tests reported by Skogerboe and Walker
(1972) and Duke et al. (1976) indicate seepage
losses from the small ditches comprising the
lateral system probably average about 8.79 ha-
m/km/year in the Grand Valley. Thus, for the
600 km of small laterals, the total seepage losses
are 5273 ha-m annually (or approximately 5300
ha-m annually). Combined lateral and canal
seepage is, therefore, approximately 9000 ha-m
annually.
On-farm Deep Percolation
Numerous studies in recent years have
attempted to quantify deep percolation from on-
farm water use. Skogerboe et al. (1974a, 1974b)
estimated these losses (including head ditch
and tailwater ditch seepage) to be about 0.246
ha-m/ha. Duke et al. (1976) estimated these
losses, independent of ditch seepage, to be 0.15
ha-m/ha. Although the earlier studies by the
writers assumed negligible head ditch seepage
(which is true for the test fields reported), it is
probably correct to allow for these losses. The
fact of the matter is, however, that many fields
fit both assumptions and given the large
number of fields tested by various investigators,
total on-farm losses are probably about 7500
ha-m.
Canal Spillage and Field Tailwater
The operational wastes and field tailwater
are difficult to define because, first, they do not
generally create problems associated with
salinity degradation, and second, data regard-
ing these flows are sparse. Skogerboe et al.
(1974b) listed field tailwater as 43% of field
application whereas Duke et al. (1976) listed this
percentage as 36%. On the other hand, Skoger-
boe et al. (1974b) and Duke et al. (1976) reported
estimates of canal spillage as administrative
wastes which were 18% and 35% of field applica-
tions, respectively. Estimates of spillage and
tailwater by the Bureau of Reclamation were
slightly smaller than the authors' estimate.
Using the 43% figure for field tailwater and the
18% figure for canal spillage yields about 37,000
ha-m per year field tailwater and spillage.
Aggregating the data presented previously
with inflow-outflow records in the vicinity of
Grand Valley gives a clear picture of how the
irrigation system relates to the overall hydrolo-
gy (Figure 7). The flow diagram is particularly
helpful in visualizing the relative magnitude of
the irrigation return flows from the agricultural
area.
( 42,100 ho-m)
Figure 7. Mean annual flow diagram of the Grand
Valley hydrology.
IDENTIFYING THE SALINITY
CONTRIBUTION
The salinity contribution of the Grand
Valley hydro-salinity system can be developed
in a number of ways. For example,
Net Salt Pickup 630,000 m tons/year
Net Groundwater Return Flow 8,100 ha-m/yr
Pickup =778 m tons
Return flow ha-m
= 7,800 mg/1
Data reported by Skogerboe and Walker (1972)
indicate an average groundwater salinity of
8,000 to 10,000 mg/1 (average of 8,700 mg/1)
which compares favorably with the salt pickup
rate of 7,800 mg/1 plus an irrigation water
salinity of 500-700 mg/1 (say 600 mg/1) which
365
-------
CASE STUDY: GRAND VALLEY
gives a total predicted salinity concentration of
8,400 mg/1.
The U.S. Geological Survey and others have
recently measured drainage return flows at
selected areas in the valley. These data indicate
an average salinity of 4,000 mg/1. Thus, as Duke
et al. (1976) pointed out, if all return flows were
through the drainage channels and
phreatophyte consumptive use was not con-
sidered (as earlier assumptions are inap-
propriate concerning these losses), the calcula-
tion would be:
i Deep Percolation - Seepage! 14.000 mg li i.Ol ; i -- Salt Pickup
or
(16,500 ha-m) (40) = 660,000 m tons
Consequently, the two salt loading figures, as
predicted by inflow-outflow mass balancing
and calculation using local data, check out.
Ignoring the effect of evapotranspiration (a
concentrating effect), the following con-
tributions to salinity loading (salt pickup) can
be delineated:
Canal and Ditch Seepage
Lateral Seepage
On-farm Losses
23%
32%
45%
REFERENCES
1. Duke, H. R., E. G. Kruse, S. R. Olsen,
D. F. Champion, D. C. Kincaid. 1976. Irrigation
return flow quality as affected by irrigation
water management in the Grand Valley of
Colorado. Agricultural Research Service, U.S.
Department of Agriculture, Fort Collins,
Colorado. October.
2. Elkin, A. D. 1976. Grand Valley salinity
study: Investigations of sediment and salt
yields in diffuse areas, Mesa County, Colorado.
Review draft submitted for the State Conserva-
tion Engineer, Soil Conservation Service,
Denver, Colorado.
3. Hyatt, M. L., J. P. Riley, M. L. McKee,
and E. K. Israelsen. 1970. Computer simulation
of the hydrologic salinity flow system within
the Upper Colorado River Basin. Utah Water
Research Laboratory, Report PRWG54-1, Utah
State University, Logan, Utah. July.
4. lorns, W. V., C. H. Hembree, and G. L.
Oakland. 1965. Water resources of the Upper
Colorado River Basin. Geological Survey
Professional Paper 441. U. S. Government
Printing Office, Washington, D.C.
5. Skogerboe, G. V., and W. R. Walker.
1972. Evaluation of canal lining for salinity
control in Grand Valley. Report EPA-R2-72-047,
Office of Research and Monitoring, En-
vironmental Protection Agency, Washington,
D.C. October.
6. Skogerboe, G. V., W. R. Walker, J. H.
Taylor, and R. S. Bennett. 1974a. Evaluation of
irrigation scheduling for salinity control in
Grand Valley. Report EPA-660/2-74-052, Office
of Research and Development, Environmental
Protection Agency, Washington, D.C. June.
7. Skogerboe, G. V., W. R. Walker, R. S.
Bennett, J. E. Ayars, and J. H. Taylor. 1974b.
Evaluation of drainage for salinity control in
Grand Valley. Report EPA-660/2-74-084, Office
of Research and Development Environmental
Protection Agency, Washington, D.C. August.
8. U.S. Environmental Protection Agen-
cy. 1971. The mineral quality problem in the
Colorado River Basin. Summary Report and
Appendices A, B, C, and D. Region 8, Denver,
Colorado.
9. U.S. Geological Survey. 1976. Salt-load
computations — Colorado River: Cameo,
Colorado to Cisco, Utah. Parts 1 and 2. Open
File Report. Denver, Colorado.
10. Walker, W. R. 1977. Combining
agricultural improvements and desalination of
return flows to optimize local salinity control
policies. Paper presented at the National Con-
ference on Irrigation Return Flow Quality
Management, Fort Collins, Colorado, May 16-
19.
11. Walker, W. R., and G. V. Skogerboe.
1971. Agricultural land use in the Grand Valley.
Agricultural Engineering Department,
Colorado State University, Fort Collins,
Colorado.
12. Walker, W. R., S. W. Smith, and L. D.
Geohring. 1976. Evapotranspiration potential
under trickle irrigation. American Society of
Agricultural Engineers Paper Number 76-2009.
June.
366
-------
HYDROSALINITY SYSTEM - GRAND VALLEY
13. Westesen, G. L. 1975. Salinity control
for western Colorado. Unpublished Ph.D. Dis-
sertation. Colorado State University, Fort
Collins, Colorado. February.
367
-------
Modeling Salt Transport in
the Irrigated Soils of Grand Valley
JAMES E, AYARS, DAVID B. McWHORTER,
and GAYLORD V. SKOGERBOE
Agricultural Engineering Department;
University of Maryland, College Park, Maryland; and
Department of Agricultural and Chemical Engineering;
Colorado State University; Fort Collins, Colorado
ABSTRACT
This study was undertaken to evaluate the
effects of the volume ofleachate leaving the soil
profile on the -quality of the leachate. A
numerical salt transport model was selected for
use in the study. Field data to calibrate the
model were collected on 63 research plots
located in the Grand Valley. The model was
tested and calibrated with the field data and
then used in a series of hypothetical simulations
designed to provide the required information.
The modeling results indicate that the salt
concentration of the leachate is independent of
the volume of leachate. The six-year simulation
showed that the concentration of salt below the
root zone was relatively constant. Therefore, the
salt loading due to subsurface irrigation return
flows can be calculated from a water budget
analysis, with salt load reductions being direct-
ly proportional to the reduced volume of subsur-
face return flow.
INTRODUCTION
Previous studies conducted on methods of
salinity control in the Grand Valley assumed
that the concentration of salt in the subsurface
return flow was independent of the volume of
the return flow. This implied that any method
which reduced the volume of return flow would
effect a similar reduction in the salt load. The
current study was designed to evaluate the
validity of this assumption. The first phase of
the study was to evaluate a salt transport model
using soil moisture and salinity data gathered
in the Grand Valley. Once the evaluation was
complete, the model was used to simulate a
series of hypothetical irrigation treatments.
Data from these simulations were used to
evaluate the effect of the volume of return flow
on the concentration of ionic species in the
solution, both in the soil profile and leaving the
soil profile. If the soil solution became saturated
with a particular ionic species, then further salt
pickup could be prevented as the return flow
moved over the shale bed.
The focus of this particular study was the
adaptation and evaluation of a numerical model
which could be used to characterize the salt
transport occurring in the soils of the Grand
Valley. A numerical model developed by Dutt et
al. (1972), which is currently being used by the
Bureau of Reclamation, USDI (see earlier con-
ference paper by Shaffer and Ribbens), was
selected for use in the study. The method of
calculating the value of hydraulic conductivity
and diffusivity used in the difference equation
of the soil water flow program was changed
from that found in Dutt's model. Also, the
functional relationships used to calculate
hydraulic conductivity and diffusivity were
changed. The soil water flow and soil chemistry
data used in the evaluation of the model were
collected as part of an on-going study in which
the effect of irrigation on crop yields and
salinity of deep percolation was investigated.
The research was conducted on 63 research
plots located on a 9.3 hectare site in the Grand
Valley. Eight irrigation treatments, four crops
and two fertilization treatments were used to
generate the moisture flow and salt transport
data required to calibrate the numerical model.
The irrigation schedules in the hypothetical
simulations used either 7 or 14 day irrigation
intervals and a depth of irrigation equal to the
evapotranspiration in the interval plus a
leaching increment which ranged from 1% to
369
-------
CASE STUDY: GRAND VALLEY
40% of estimated evapotranspiration. The
evapotranspiration for the simulations was
estimated using meteorological data collected in
the Grand Valley in conjunction with the irriga-
tion scheduling program of the Agricultural
Research Service (Kincaid and Heerman, 1974).
SOIL CHEMISTRY MODEL
The primary objective is to model the
transport of salts through the soils. (See Ayars,
1976, for a more thorough discussion.) The first
portion of the flow of water and consequent
transport of salts is through the root zone,
which is usually a zone of partial saturation. A
numerical model of the moisture flow and
chemical and biological reactions occurring in
the root zone has been developed by Dutt et al.
(1972). (This reference contains a complete
listing of the computer program.) This is the
basic model which will be used to describe salt
transport.
The model consists of three separate
programs. The first program describes the soil
moisture movement and distribution with time.
The second program interfaces the soil moisture
movements with the chemical-biological model.
This is needed because the horizons used in the
calculations of soil moisture and chemistry
differ. The third program computes the
chemical and biological activity occurring in
the soil profile. Figure 1 is a block diagram of the
over-all model. A brief description of the
moisture flow and chemical-biological models is
included to serve as a basis for understanding
the data collection requirements.
The flow is one-dimensional and was
developed using the Richards equation with a
sink term. Schematically, the model is given in
Fig. 2. Mathematically, the flow is described
using Richards equation in the form:
dt
where
9
t
z
D(0)
D(6)d6
dz
• K(0)j - S(E t
(1)
S(Et)
— volumetric moisture content,
= time,
= space coordinate in vertical direction,
= soil moisture diffusivity,
— hydraulic conductivity,
= evapotranspiration sink term (volume of
water consumed per unit volume of soil
per unit time).
INPUTS - WATER
APPLICATION a
CONSUMPTIVE USE
DATA
MOISTURE FLOW
PROGRAM
INPUTS = INITIAL
WATER CONTENT 8
PHYSICAL PROPERTIES
OF SOIL
OUTPUTS = SOIL
MOISTURE CONTENT 8
MOVEMENT WITH TIME
INPUTS = FERTILIZER
8 ORGANIC -N
APPLICATIONS,
TEMPERATURES,
CROP TYPES
BIOLOGICAL
CHEMICAL PROGRAM
INPUTS = INITIAL
CHEMICAL 8 PHYSICAL
PROPERTIES OF SOIL
OUTPUTS - WATER ,
NITROGEN 8 SALTS
ENTERING GROUND
WATER
Figure 1. Generalized block diagram of the model.
370
-------
SALT TRANSPORT - GRAND VALLEY
This is the diffusivity form of the equation
which means that only flow in the partially
saturated zone of the soil profile can be de-
scribed. The evapotranspiration sink term,
S(Et), is computed using the Blaney-Criddle
equations for evapotranspiration (other
equations could be adapted to the model such as
the root extraction term reported by Hanks,
Willardson and Melamed earlier in this con-
ference) with the loss due to evapotranspiration
being distributed through the soil profile by
assuming a specific root distribution for the
crop. The root distribution and coefficients for
the Blaney-Criddle equations are supplied by
the user. Actual values of evapotranspiration
can be used in the sink term when they are
known.
Salt transport is described by the following
equation in one dimension.
dec _ a
dz
(0D
dc
dz )
3(vc)
9z
+ S
(2)
(START MOISTURE "\
FLOW PROGRAM J
\ '
PROGRAM MOISTRE
READ CONTROL AND INPUT DATA
COMPUTE MOISTURE CONTENT AND
FLUX FOR EACH DEPTH NODE AND
TIME STEP
WRITE ON MAGNETIC TAPE OR
PRINT OUTPUT
,
SUBROUTINE THEDATE
COMPUTE CALENDAR DATE FROM
DAY NUMBER
1
SUBROUTINE CONUSE
COMPUTE VALUE OF MACROSCOPIC
SINK TERM
,
c
STOP MOISTURE
FLOW PROGRAM
Figure 2. Generalized block diagram of Moisture
Flow Program.
where
c = solute concentration,
D = diffusion-dispersion coefficient,
v = volumetric flux given by Darcy's Law,
S = source or sink term for the chemical species.
By assuming the term 3/3z (0D 3c/3z) is
negligible compared to 3(vc)/3z the equation
reduces to ddc/di = - 3(vc)/3z + S. This assump-
tion implies that transport due to dispersion in
partially saturated soils is negligible compared
to the convective transport which occurs. (The
earlier conference paper by Hanks, Willardson
and Melamed presents a method for handling
the source or sink term, S.) This is generally a
good assumption.
The model computes the moisture flow (v)
and couples the flow with the chemical changes
3c/3z computed in the biological-chemical
program to give the salt transport. This tech-
nique is the basis for the mixing cell concept.
The chemical exchange model computes the
equilibrium chemistry concentrations for calci-
um, magnesium, gypsum, sodium, bicarbon-
ates, carbonates, chlorides, and sulfates. The
nitrogen chemistry including ammonium,
nitrates, and urea-nitrogen uses a kinetic in-
stead of an equilibrium approach. The kinetic
approach is needed since microbial activity
involved in nitrogen transformation occurs over
a period of weeks and days instead of minutes
and seconds. The equilibrium chemistry for in-
organic salts is a good approximation since the
reactions describing their chemistry occur in a
matter of minutes or seconds in a flow regime
which is changing very slowly. A block dia-
gram of the biological-chemical model is given
in Fig. 3. A block diagram illustrating the
chemical reactions that are considered in one of
the subroutines is shown in Fig. 4.
SUMMARY OF MODEL RESULTS
From the calibration of the moisture flow
model using infiltration data, water content
profiles and storage change data, it was con-
cluded that the water flow could be adequately
modeled for the Grand Valley. Several
modifications were made to Dutt's original
program before the above conclusion could be
made.
The functions originally used to calculate
K(0) and D(0) in the model did not permit
accurate computation of soil-water flux at water
371
-------
CASE STUDY: GRAND VALLEY
contents close to full saturation for the con-
ditions of this study. A functional relationship
developed by Brooks and Corey (1964) was used
in the program to calculate hydraulic conduc-
tivity [K(f?)J. The function used in the model to
calculate soil-water diffusivity [D(f?)] was
developed using the relationship developed by
Brooks and Corey (1964) for K(0) and the
relationship for the soil-water characteristic
developed by Su and Brooks (J975).
The method used to compute the average
values of K(0) and D(0) required to solve the
difference form of Richards equation was also
changed. The average value of K(0) and D(0)
was originally computed using the average
water content of the two nodes being considered.
The averaging in the flow model was modified
so the conductivity is now calculated by using
/"START BIOLOGICAL-"\
V CHEMICAL PROGRAM^/
PROGRAM MAIN
READ CONTROL AND INPUT DATA
STORE INITIAL SOIL-CHEM DATA
PRINT CONTROL AND INPUT DATA
(OPTIONAL)
SUBROUTINE EXECUTE
MAKE ANY FERTILIZER AND/OR .
ORGANIC MATTER APPLICATIONS
INITIALIZE OR UPDATE SOIL
TEMPERATURES (WEEKLY)
READ MOISTURE FLOW DATA FROM
MAGNETIC TAPE
SUBROUTINE COMBINE
FOR EACH SEGMENT.
CALL EXCHANGE SUBROUTINE
CALL NITROGEN SUBROUTINE
CALL SOLUTE REDISTRIBUTION
SUBROUTINE
CALL PLANT -N UPTAKE SUBROUTINE
SUM CHEMISTRY CHANGES AND
UPDATE VALUES IN STORAGE
PRINT OR WRITE SPECIFIED VALUES]
X X
(STOP BIOLOGICAL-A
CHEMICAL PROGRAM/
Figure 3. Generalized block diagram of Biological-
Chemical Program.
Call EQEXCH (First Time)
Calculate CaCO, Solubility Constant at
Specified Moisture Content
I
Consider Solubility Reaction CaSQ, x2HzO:
Co*** SO, + 2HZ0
I
Consider Undissociated Ion Pair Reaction
CaSO, =Ca«*-
Consider the Exchange Reaction
2No*+Co-R = Co**+2Na-R
Consider the Exchange Reaction
Mg**+ Co-R rCo** + Mg-.R
Consider the Exchange Reaction
NH* + Na-R = No*+NH4-R
Consider Undissociated Ion Pair Reaction
MgS04 = Mg
1
Consider the Solubility Reaction
Return to Combine if Equilibrium
Constraints Satisfied
N
Figure 4. Generalized block diagram of subroutine
XCHANGE.
the moisture content at each node and then the
calculated conductivities are averaged. The dif-
fusivity is now calculated as an integrated
average diffusivity between the water contents
at adjacent nodes.
After these changes were made, it was
possible to predict infiltration, water content
distributions and changes in storage that
agreed satisfactorily with field measurements.
Since the model assumed a homogeneous
profile, it was necessary to calibrate the flow
model so as to incorporate the variability of field
properties into the simulations. The soil water
characteristic was calculated as an average
from water-content pressure head data gathered
through the entire depth of the soil profile in a
small area of the test site. This average
characteristic was then used to calculate K(0)
and D(0) in the calibration simulations.
The flow model was calibrated by matching
computed infiltration depths and times, water
content profiles and storage changes to field
measured parameters. The field measured water
372
-------
SALT TRANSPORT — GRAND VALLEY
content profiles were averages of profiles
measured at 4 separate locations in the test plot.
Only the value of K8 was changed during the
calibration, since all other parameters were
specified. The calibration was considered com-
plete when the computed values of infiltration
duration, storage change and water content
profiles were approximately equal to the field
measured values. The above procedure is essen-
tially an averaging process which incorporates,
to some degree, the field variability of the
hydraulic parameters into the flow model.
From comparisons of simulated and field
data used in evaluating the chemistry compo-
nent of Dutt's model, it was concluded that TDS
concentrations were adequately modeled but
that individual ionic species concentrations
were not. The simulations used to compare
computed and field chemistry data were made
using field data for initial and boundary con-
ditions in both the chemistry and flow models.
Field data on the chemical composition of the
soil solution extracted at a depth of 1.1 meters
for a 30 day period was used to compare with
calculated salt concentrations.
Comparisons of calculated and measured
data indicate that the CaSCU - CaCOs -
Ca(HCO 3) 2 system is not adequately modeled
for the soils in the Grand Valley. The model
computed Ca^~^ concentrations at a depth of
1.1 meters that were greater than theoretical
maximum values expected for this soil system.
A study of the CaSO4 - CaCO3 - Ca(HCO3>2
equilibrium equations indicated that the
solubility product of Ca(HCO3>2 calculated
using ion activities was incorrect. The solubility
of calcium bicarbonate, Ca(HCO 3)2, was then
calculated based on the partial pressure of
carbon dioxide. A reasonably good agreement
between computed and measured TDS concen-
tration was obtained using a value of 7 matin for
the partial pressure of carbon dioxide in the
simulations.
Data for single growing season simulations
using 7 and 14 day irrigation schedules and 2%,
5%, 20%, and 40% leaching increments (Figure
5), coupled with data from a 6 year simulation
using a 14 day irrigation schedule and 20%
leaching increment (Table 1, year 1 and 6)
indicate that the salt concentration of the
leachate is independent of the volume of
leachate. TDS profiles calculated at the begin-
ning and end of the six-year simulation (Figure
6) show the concentration of salt in the profile
below a depth of 122 cm, which is the bottom of
the root zone in the simulation, to be relatively
constant. Apparently, this zone acts as a buffer
and causes the salt concentration in the
leachate at the bottom of the profile to be
relatively constant.
In summary, the salt loading due to irriga-
tion return flow can be calculated from a
knowledge of the water balance in the system.
Reductions, therefore, in salt loading would be
directly proportional to changes which reduced
the volume of return flow. (Similar results were
obtained by Hanks, Willardson and Melamed as
reported earlier in these proceedings.) Es-
timating salt loading on a valley-wide scale
would require a knowledge of the volume of the
return flow and the concentration of salts in the
flows. The latter can be predicted from the
model used in this study if measured values are
not available.
3600
3500
3400
3300
3200
500
400
300
200
3200
3100
3000
2900
2800
27OO
• 2 %
D 20 % 0° °
0 40 % 0°
• &
_1_ _J 1 1 i t J | 1 1 1
-
' . a o • f'"***** ^
III 1 1 f ' ' ' !
?> °CP
- ^ D • °» **
-
1 1 1 1 1 1 1 1 1 1
Cumulative Leachat* (cm)
Figure 5. TDS and chloride concentrations as a
function of cumulative leachate at a depth of 2.1 m
calculated by hypothetical simulations using a 14-
day irrigation interval.
373
-------
CASE STUDY: GRAND VALLEY
TABLE 1
TDS concentrations and chloride concentrations
in cumulative leachate at 2.13 m for 6 year
hypothetical simulation using 14 day irrigation
schedule and 20% leaching increment.
Cumulative Cumulative
Julian Infiltration Leachate
Date (cm) (cm)
Cl TDS TDS-Cl
ppm ppm ppm
157
171
185
199
213
227
241
255
269
283
293
157
171
185
199
213
227
241
255
269
283
293
8.33
11.45
16.46
23.50
31.64
40.55
51.25
60.49
68.36
75.11
80.55
409
413
418
425
433
442
452
462
470
476
482
Year 1 of 6
5.02
6.70
7.48
8.05
8.92
10.11
11.93
12.86
13.66
14.31
14.75
Year 6 of 6
71.85
72.60
73.29
73.85
74.71
75.89
77.49
78.69
79.45
80.10
80.55
269
290
298
302
297
299
302
317
323
327
327
446
411
413
412
398
390
382
391
392
392
390
3276
3318
3336
3352
3352
3365
3391
3416
3432
3444
3444
3479
3435
3439
3438
3422
3414
3406
3415
3418
3421
3418
3007
3028
3038
3050
3055
3066
3089
3099
3109
3117
3117
3033
3024
3026
3026
3024
3024
3024
3024
3026
3029
3028
TDS (ppm)
1000 2000 3000 4000
15
3O
46
61
76
- 9.
2 107
.e
a 122
S .37
150
168
183
198
213
REFERENCES
1. Ayars, James E. 1976. Salt transport in
irrigated soils. Unpublished Ph.D. Dissertation,
Colorado State University, Fort Collins,
Colorado. August.
2. Brooks, R. H. and A. T. Corey. 1964.
Hydraulic properties of porous media.
Hydrology Paper No. 3. Colorado State Univer-
sity, Fort Collins, Colorado. March. 27 p.
3. Dutt, G. R., M. J. Shaffer, and W. J.
Moore. 1972. Computer simulation model of
dynamic bio-physio-chemical processes in soils.
Technical Bulletin 196, Department of Soils,
Water and Engineering, Agricultural Experi-
ment Station, University of Arizona, Tucson.
October.
4. Hanks, R. J., L. S. Willardson and
D. Melamed. 1977. Modeling salinity of irriga-
tion return flow where sources and sinks are
present. Paper presented at National Con-
ference on Irrigation Return Flow Quality
Management, Colorado State University, Fort
Collins, Colorado. May 16-19.
5. Kincaid, D. C. and D. F. Heermann.
1974. Scheduling irrigations using a program-
mable calculator. ARS-NC-12, ARS-USDA.
February.
6. Su, C. and R. H. Brooks. 1975. Soil
hydraulic properties from infiltration tests.
Proceedings, Watershed Management Sym-
posium, Irr. and Drainage Div., ASCE, Logan,
Utah. August 11-13. p. 516-542.
200
Cl ( ppm)
400 600 800 1000
Year I
• Year 3
A Year 6
Figure 6. TDS and chloride concentration profiles at day 293 calculated by a 6-year hypothetical simu-
lation using 20% baching increment and 14-day irrigation interval.
374
-------
Evaluating Appropriate
Technologies for Salinity
Control in Grand Valley
ROBERT G. EVANS, WYNN R. WALKER, STEPHEN W. SMITH and
GAYLORD V. SKOGERBOE
Agricultural and Chemical Engineering Department;
Colorado State University; Fort Collins, Colorado
ABSTRACT
A summary of the results of applied
research on salinity control of irrigation return
flows in the Grand Valley of Colorado is
presented for the period of 1969 to 1976. Salinity
and economic impacts are described for the
Grand Valley Salinity Control Demonstration
Area which contains approximately 1600 hec-
tares and involves most of the local irrigation
entities in the valley. During the eight years of
the demonstration project, 12.2 km of canals
were lined, 26.54 km of laterals were lined, 16.4
km of drainage tile was installed, a wide variety
ofon-farm improvements were constructed, and
an irrigation scheduling program was im-
plemented. The total value of the constructed
improvements in the demonstration area was
almost $750,000.
INTRODUCTION
In 1967, the irrigation companies of
Colorado's Grand Valley became concerned
with the potential financial burden which could
be placed upon the valley's water users by
salinity damages downstream, especially if
they were forced to comply with salinity control
measures at their expense. Consequently, they
began efforts to initiate action based on the
concept that abatement of the salinity problem
would have state, regional, national and inter-
national benefits.
The irrigation companies formed a
cooperative organization called the Grand
Valley Water Purification Project, Inc. (later
called Grand Valley Canal Systems, Inc.), and
petitioned the Federal Water Quality Ad-
ministration for matching funds for
demonstrating canal lining as a salinity control
measure. This money was forthcoming, and in
1968 the Agricultural Engineering Department
of Colorado State University contracted to do
the technical evaluation regarding the effec-
tiveness of canal lining for reducing the Grand
Valley's salt load contribution to the Colorado
River.
As time progressed, additional studies were
initiated to determine the potential benefits of
other technologies such as irrigation schedul-
ing, drainage, and lateral improvements for
controlling salinity in the Grand Valley. A
summary of the results of these various in-
vestigations is reported herein. A map il-
lustrating the study area and the various in-
vestigations is shown in Figure 1.
CHANNEL LINING
The results of the 1969 channel lining
studies indicated that canal and lateral lining
in the study area reduced salt inflows to the
Colorado River by about 4,000 metric tons
annually (Skogerboe and Walker, 1972). The
bulk of this reduction is attributable to the canal
linings, but clearly indicated is the greater
importance of lateral linings. The length of
laterals (600 km), plus the farm head ditches
(1640 km), is about eight times greater than the
length of canals (286 km). The economic
benefits to the Lower Basin water users alone
exceed the costs ($350,000 construction plus
$70,000 administration) of this project. Conse-
quently, it is justifiable to conclude that con-
veyance lining in areas such as the Grand
Valley, where salt loadings reach 18 metric tons
or more per hectare, are a feasible salinity
control measure. The local benefits accrued
from reduced maintenance, improved land
375
-------
CASE STUDY: GRAND VALLEY
Legend
Water Supply
".'.'.'.'.' V.'.'j Land Under Study Lateral
t::-X:X:X:X:XX| Previous Drainage Study
Irrigation Scheduling Protect
Hydrotogic Boundary
Canal or Ditch (No Improvements)
Drain or Wash
Trapezoidal Concrete
Slip-form Lining
Gunite Lining
Gumte ,Do»(ihill
Bank (My
Grand Valley Canal
Stub Ditch
overnment
i Highline
Canal
/ Price Ditch
Scale I Kilometer
Figure 1. Grand Valley Salinity Control Demonstration Area showing the locations of the various
investigations conducted from 1969 to 1976.
value, and other factors add to the feasibility of
conveyance linings as a salinity management
alternative.
The first and most important consideration
in improving farm water use is control. Implied
in this realization is the requirement of sound
water measurement at the farm turnout and
again at critical division points below the
turnout. This would necessitate a considerable
rehabilitation of both the canal and lateral
system and the implementation of a "call
period" to allow canal operators more time for
flexible water handling. In addition, it is impor-
tant that the canal companies extend their
control of the water below the canal turnout
structure to include key division points within
the lateral system to insure equitable allocation
of water among users.
IRRIGATION SCHEDULING
The irrigation of agricultural lands in the
Colorado River Basin is a significant cause of
the salinity concentrations encountered in the
Colorado River. Emphasis towards stemming
further salinity increases has logically centered
upon improving the quality of irrigation return
flows. This emphasis, especially in the high salt
contributing areas like the Grand Valley in
western Colorado, focuses upon reducing the
flows which pass through the saline soils and
aquifers, thereby reducing the salt pickup which
occurs by dissolution. Since a major fraction of
the water percolating through local soils in this
manner comes from over-irrigation, measures
aimed at improving irrigation efficiencies have
a high potential for controlling salinity. Among
the methods for achieving higher water use
376
-------
TECHNOLOGIES FOR SALINITY CONTROL — GRAND VALLEY
efficiencies on the farm, "scientific" irrigation
scheduling is one of the most important (Skoger-
boe, Walker, Taylor and Bennett, 1974).
Irrigation scheduling consists of two
primary components: crop evapotranspiration
and available root zone soil moisture.
Evapotranspiration is calculated by using
climatic data. The other major category of
required data pertains to soil characteristics.
First of all, field capacity and permanent
wilting point for the particular soils in any field
must be determined. More importantly, infiltra-
tion characteristics of the soils must be
measured. Only by knowing how soil intake
rates change with time during a single irriga-
tion, as well as throughout the irrigation
season, can meaningful predictions be made of:
(a) the proper quantity of water which should be
delivered at the farm inlet for each irrigation;
and (b) the effect of modifying deep percolation
losses. With good climatic data and meaningful
soils data, accurate predictions for the next
irrigation date and the quantity of irrigation
water to be applied can be made. In order to
enable the irrigator to apply the proper quantity
of water, a flow measurement structure is ab-
solutely required at the farm inlet.
The results of this demonstration project
indicated that irrigation scheduling programs
have a limited effectiveness for controlling
salinity in the Grand Valley under existing
conditions. Excessive water supplies, the
necessity for rehabilitating the, irrigation
system (particularly the laterals), and local
resistance to change preclude the proper
management of water applied during successive
irrigations. To overcome these limitations,
irrigation scheduling must be accompanied by
flow measurement at all the major lateral
division points and farm inlets. In addition, it is
necessary for the canal companies and irriga-
tion districts to assume an expanded role in
delivery of the water. Also, some problems have
been encountered involving poor communica-
tion between farmer and scheduler and certain
deficiencies in the USER computerized schedul-
ing program which was used in this study
dealing with evapotranspiration and soil
moisture predictions. Correcting these con-
ditions is easily rectified and will make irriga-
tion scheduling much more effective and ac-
ceptable locally.
Water budgets from which the study results
were generated were obtained from intensive
investigation on two local farms. The selection
of the two study farms was intended to be
representative of conditions valley-wide.
Analysis of the budgets reveal that ap-
proximately 50 percent of the water applied to
the fields came during the April and May period
when less than 20 percent of the field
evapotranspiration potential had been ex-
perienced. Salt pickup estimates during this
early part of the season amounted to about 60
percent of the annual total for each field.
Another indication of the importance of early
season water management is presented in an
analysis of irrigation efficiencies. As the season
progressed, the soils became less permeable and
the crop water use increased, causing marked
improvements in irrigation efficiency. Thus, if
irrigation scheduling is employed in its optimal
format, salt pickup from the two fields can be
reduced as much as 50 percent or more.
The results of this demonstration project
show that irrigation scheduling is a necessary,
but not sufficient tool for achieving improved
irrigation efficiencies. The real strides in reduc-
ing the salt pickup caused by over-irrigation will
come from the employment of scientific irriga-
tion scheduling in conjunction with improved
on-farm irrigation practices.
DRAINAGE
Drainage investigations in the Grand
Valley began shortly after the turn of this
century when local orchards began failing due
to high saline water tables. Studies showed the
soils to be not only saline but also of low
permeability. At the time, the future develop-
ment of the Bureau of Reclamation's "Grand
Valley Project" loomed as a severe threat to the
low lying lands between it and the Colorado
River. As an answer to these apparent drainage
needs, the solutions were clearly set forth but
never fully implemented. Rather, a local
drainage district was formed to construct open
ditch interceptor drains and some buried tile to
correct trouble spots. All of these efforts barely
contained the rise in water tables, and today
more than fifty years later, the local conditions
remain essentially unchanged.
This study was undertaken with the history
of local drainage well in mind, but for a different
purpose, the skimming of water from the top of
the water table before it reaches chemical
equilibrium with the highly saline soils and
aquifers. A farm owned by Mr. Bob Wareham
377
-------
CASE STUDY: GRAND VALLEY
was used in the study to demonstrate the
skimming effect by installing field relief drains
on 12.2-meter centers. The field had been under
poor irrigation management for several years.
so the results are not immediately discernable.
However, analysis of water quality throughout
the study area indicated that effective relief
drainage would intercept flows with a salinity
concentration as much as 3000 mg 1 lower than
existing groundwater concentrations.
In viewing the results of this study, it is
obvious that field drainage is a curative rather
than preventive measure. High costs of such
programs illustrate the need of first minimizing
the flows passing through the root zone or
seeping from canals and laterals. The small
amount of water entering the groundwater
could then be effectively removed by drainage
systems located at selected locations. Thus, field
drainage as it pertains to objectives of salinity
control is a remedy which must be considered
but will probably not be implemented until the
last stages of salinity control in the valley
(Skogerboe, Walker, Bennett, Avars and Taylor,
1974).
LATERAL IMPROVEMENTS
Since this study involved the selection of
several laterals in which the irrigators to be
served would participate by contributing to the
construction costs, the total cooperation of these
people was imperative. By first publicly adver-
tising the project, and then personally contact-
ing one or more parties who seemed most
interested, about eleven potential groups in the
project area were identified. After the people
had discussed the matter among themselves,
project personnel arranged meetings where
specific details of the project were outlined and
questions answered. Since the average farm size
in the area is about 2 to 6 hectares (5 to 15 acres),
the number of people involved could have been
as high as one hundred. In actuality, there were
89 persons involved. Interestingly, problems
anticipated in coordinating such a large group
did not materialize.
The first step in the lateral selection process
began when an announcement of the project,
along with a project location map. was placed in
the local newspaper (Figure 2). The article
stated the funding availability, its purposes,
conditions for qualifying, and the availability
of project representatives at an upcoming ques-
tion and answer session. Prior to this "open
house", the writers had envisioned the possibili-
ty of a long period of door-to-door field contact.
However, forty individuals representing ten
laterals within the demonstration area attended
the session, and further contact was net
necessary. The overwhelming response at the
open house resulted in considerable time
savings and undoubtedly ranks as one of the
most important events leading to the project's
success.
At the open house, each inquirer was ad-
vised that the best action at that time would be
to contact others on the lateral, briefly explain
the project objectives, and enlist support. Con-
tact with the individuals who came to the open
house was re-established several weeks later,
and meetings with project personnel were
scheduled. With the exception of two cases, the
meetings were unqualified successes in gaining
the acceptance of the people involved. Lateral
groups desiring the project's help were told final
site selection would not be made until the fall or
winter of 1974-75 so that each lateral could be
evaluated for its usefulness in satisfying the
project objectives. Also, this time was used by
irrigators to finalize their own willingness to be
included in the project and to reach agreement
among themselves concerning such matters as
cost sharing and the desirable operating
characteristics of the improved irrigation
systems.
The project was established on the basis
that the project would pay 70 percent and the
participants 30 percent of the total construction
costs, not including engineering or administra-
tion. The 30 percent matching requirement
could be paid in cash or by equal value
arrangements such as direct labor, equipment
rentals, land leveling costs, or through the
voluntary assistance of local organizations
such as the Grand Junction Drainage District.
As more was learned about the various
lateral systems and the attitudes of the
irrigators, it was necessary to continually re-
evaluate each lateral in terms of project objec-
tives. In addition, throughout the first year,
project personnel received numerous requests
(at least 2 or 3 per week during the summer
months of 1974) from other interested land-
owners within the project area. In fact, requests
for assistance are still being received after
completion of the project.
378
-------
TECHNOLOGIES FOR SALINITY CONTROL — GRAND VALLEY
$230,000 EPA grant to fund
new seepage control project
The EPA has granted $230,000 for the
lining of laterals, construction of new
on-farm irrigation systems, and in-
stallation of tile drainage.
The funds can be used to pay 70 per
cent of the construction costs, with the
farmer paying the remaining 30 per
cent.
The demonstration project will use
two laterals under each of the five
canals in the study area. Laterals will
be selected to represent a wide variety
of conditions.
To participate, all of the irrigators
under a lateral must be willing to
share in the costs of lateral lining and
on-farm irrigation improvements. A
few of the laferals ha^e already been
extensively lined with concrete under
the previous demonstration project
CSU officials said the selection of a
lateral and all the crop land served by
a lateral, rather than an individual
farm, has a tremendous advantage in
allowing control at the lateral turnout.
Thus both the quantity of flow and the
time of water delivery can be con-
trolled, thereby providing improved
water management and higher crop
yields.
The new construction program will
be explained by CSU personnel at the
Holiday Inn from 9 a.m. to 4:30 p m.
Feb. 27. Any irrigator having lands in
the study area can inquire at that time
about possibilities for participating
Figure 2. Announcement of grant award in Daily Sentinel (Grand Junction, Colorado).
Funding has been received from the
U.S. Environmental Protection
Agency to construct irrigation im-
provements in the area between Grand
Junction and Clifton, according to the
Agricultural Engineering Dept at
Colorado State University.
The area is the same that received
funding five years ago for concrete and
gunnite lining of canals and laterals to
reduce seepage.
CSU officials said the advantage in
continuing work in the area is that
much is already known about the
underground water and the salt
flowing into the Colorado River from
the area. Additionally, they said,
considerable money has been spent on
both equipment and personnel for
instrumenting the particular
demonstration area.
The amount of information provides
a strong basis for evaluating the ef-
fectiveness of irrigation im-
provements in reducing river salinity
The study area was originally
selected because it is fairly
representative of the Grand Valley
Five canals traverse the area, thereby
allowing greater participation by the
majority of irrigation entities in the
valfey. '
The new study will use a variety of
irrigation methods, including "tuning
up" methods presently in use. CSU
said considerable experience has been
gained in improving the existing
irrigation methods while evaluating
irrigation scheduling as a salinity
control measure in the Grand Valley
However, more advanced irrigation
methods have not been evaluated in
the Grand Valley for salinity benefits.
Irrigation systems to be constructed
under the new project include
automated farm head ditches, border
irrigation, sprinkler irrigation, and
trickle irrigation. Tile drainage also
will be constructed on some farms.
In particular, some of the lands near
the Colorado River will require
drainage facilities to reclaim them for
high level productivity.
CSU officials said the most
significant aspect of the project is use
of a salinity control "package"' rather
than a single control measure.
Field days will be conducted in the
third year — 1976 — of the project,
probably during August.
The selection of a lateral as the subsystem,
rather than an individual farm, has the advan-
tage of maintaining control at the lateral turn-
out. In this way, both the quantity of flow and
the time of water delivery can be controlled,
thereby facilitating improved water manage-
ment throughout the system. The lands selected
demonstrated a wide variety of irrigation and
drainage problems which provided a represen-
tative cross section of the valley's irrigated
lands. With the available knowledge regarding
the study area, a lateral and its associated lands
could be used as a logical subsystem for
evaluating the salinity reduction in the
Colorado River resulting from the implementa-
tion of a salinity control technology package.
The laterals selected were evaluated on the
basis of four broad criterion:
1. Participation from all the water users on
lands served by the lateral;
2. The degree of participation in all three
phases of the project (pre-evaluation,
379
-------
CASE STUDY: GRAND VALLEY
construction, and post-evaluation) cover-
ing the project's three-year period;
3. The type and extent of irrigation and
drainage problems represented, and the
different solutions and alternatives
which were agreeable and economically
advantageous to the landowners; and
4. The analysis of the least cost expendi-
tures, demonstration value to other
farmers, and maximum production of re-
search results in order to realize project
objectives.
One hundred precent participation was re-
quired to accomplish one of the major goals of
the projecfcto demonstrate the effectiveness of a
"package" of technological improvements on a
broad scale for purposes of salinity control. This
was not a problem in any way during the
project.
Numerous lateral group meetings and in-
dividual discussions with irrigators were used
to evaluate, as objectively as possible, the
anticipated degree of voluntary participation in
the project's three phases, as well as their
willingness to change existing irrigation prac-
tices and methods. This was very critical since
many of the proposed lateral systems might be
designed in such a way that a return to old
methods and practices would be practically
impossible, and proposed management
methods might be mandatory for continued
operation of the system. The results and im-
plications were fully explained to all the par-
ticipants before any final decisions were
mutually agreed upon.
The type of physical problems and the
extent of these problems were carefully exam-
ined by project personnel to insure that un-
necessary duplication did not occur and that as
many different problems as possible could be
treated by a variety of technologies. The long
range objectives of agricultural salinity control
in the Grand Valley, and the Colorado River
Basin, were also taken into account in choosing
the types of problems to be studied.
The composition of various existing salini-
ty control technologies were carefully matched
to achieve a maximum effect on each lateral
subsystem. Planned treatments included
sprinkler irrigation, drip (trickle) irrigation,
concrete lateral linings, concrete head ditches,
gated pipe, automated cut-back irrigation, land
shaping and clearing, flow measurement,
tailwater removal systems, buried PVC plastic
irrigation pipelines, field drainage, irrigation
scheduling, and various improved water
management practices for each subsystem.
Once the selected laterals were identified by
their special problems, the alternative solutions
for alleviating these problems were presented to
the landowners and a proposed course of action
was planned in complete accordance with the
wishes of all parties. Then, project personnel
analyzed the costs of various alternatives and
prepared basic quantity take-offs and
preliminary cost estimates. Further meetings
were held with the water users under each
lateral and final plans were mutually adopted.
General Considerations in Design and
Construction
All of the improvements were organized
into a logical experimental design in order to
effectively evaluate the objectives of the project,
and an overall design philosophy was formu-
lated to govern the designs of improvements
within the lateral subsystems.
The first major consideration was the place-
ment of flow measurement devices in each
lateral subsystem. These measurement devices
•were placed immediately below all the lateral
headgates and measurement structures were
placed at all flow divisions and after each farm
delivery point on the main delivery system.
Also, all measurement devices can be read
directly without the use of tables or calculations.
To accomplish this, propeller meters read direct-
ly in cubic feet per second (cfs), and special
enameled metal staff gauges were designed and
manufactured for all the Cutthroat flumes
which read directly in cfs and Colorado miner's
inches. Two sizes of Cutthroat flumes were
standardized throughout the project: 1) a 20.3-
cm throat width by 91-cm length (8-inch throat
width by 3-foot length), and 2) 7.6-cm throat
width by 91-cm in length (3-inch throat width by
3-foot length).
Another consideration was the grade for all
pipelines and concrete ditches would be gov-
erned by the slope of land surfaces when possi-
ble. This reduced costs considerably because it
eliminated the necessity of many costly drop
structures and energy dissipation facilities. Im-
provements followed old channels where possi-
ble and an attempt was made to consolidate
ditches and laterals to minimize duplication of
facilities. Also, efforts were made to consolidate
land where possible under each lateral in order
380
-------
TECHNOLOGIES FOR SALINITY CONTROL — GRAND VALLEY
to maximize the usefulness of the im-
provements.
Anytime a lateral passed or went through a
subdivision or other urban area, the water was
conveyed in a closed conduit for health and
safety reasons, for the aesthetics of eliminating
an open ditch, and to minimize debris problems
caused by children playing in the ditch.
Under roadways and access routes, the
PVC plastic irrigation pipes were encased with
concrete pipe. Anytime a corrugated metal pipe
(CMP) culvert was replaced or relocated, it was
replaced with a sulfate resistant concrete pipe.
The concrete pipe was about one-half the
material cost of CMP, but the initial installation
costs were higher. However, in the highly saline
soil conditions of the Grand Valley, the concrete
would be expected to outlast the CMP by at least
20 years.
When water was taken from a lateral to
irrigate landscapes in a subdivision, where
possible the subdivision water was separated
from the agricultural water because the
methods of operation are so different as to be
incompatible with one another.
Construction of Lateral Improvements
Construction was completed in three stages
during the fall of 1974, the spring of 1975, and
the fall of 1975 with the majority of the construc-
tion completed by the start of the 1975 irrigation
season. Based upon the designs and a complete
list of materials, contractor and materials
specifications were prepared in the fall of 1974.
A summary of the construction by lateral and
the value of the improvements is presented in
Table 1. An illustrative summary of the mixing
of technologies utilized on a single lateral (MC
10) to maximize the research and demonstration
effectiveness is presented in Figure 3.
Irrigator Response to the Project
Several of the lateral groups elected to do
much of the construction work themselves and
were thus very involved in day-to-day opera-
tions. On other laterals, where the construction
TABLE 1
Summary of project improvements on lateral subsystems.
LATERAL
Types of Improvements*
Concrete Ditches (ml
Buried Plastic Pipelines
Gravity systems (m)
Pressurized systems (m)
Gated Pipe (m)
Drip Irrigation (h)
Overhead Sprinklers (h)
SideroU Sprinklers (h|
Drainage Works (h)
Plastic Drainage Tile (m)
Concrete Drainage Tile (m)
Flow Measurement (No.)
Cutthroat Flumes (No.)
90°V-NotchWeirs(No.)
Parshall Flumes' (No.)
12" Propeller Meters (No.)
10" Propellor Meters (No.)
8" Propellor Meters (No.)
Other Meters (No.)
Metering Headgates (No.)
Debris Removal Equipment (No.)
Land Shaping, etc. (h)
Irrigated Hectares (Possible)
Total Value
Value/Hectare
Value/Irrigated Hectare
HLC
113.1 hi
2.2'
244
(3)
1
2
11.6
$4,860.
370.
420.
HLE
!35.Sh>
,
792
3,274
5.2
8.0-
198
(4)
2
1
1
1
34.2
$30,760.
860.
900.
PD 177
127.8 h)
230'
2,051
207
2.2
(1-4)
3
1
3
1
1
3
2
1
22.7
$43,970.
1,580.
1,940.
GV92
124.3 h)
189
817
(3)
2
1
20.4
$13,600.
560.
670.
GV95
179.1 h)
2,789
2,312
378
583
4.0
6.5
2.667
(25)
18
1
1
1
2
1
1
1
24.3
69.8
$104,790.
1,320.
1,600.
GVISO
178.7 h)
1,898
2,573*
11.5
3,496
127)
26
1
53.5
$84,680.
1,080.
1,580.
MC3
13.7 h)
157
2.5
1,958
(3)
2
1
3.0
$18,440.
4,980.
6,150.
MC10
154.0 h)
2,723"
1,054
564
6.1
4.958
(18)
14
2
1
1
11.3
44.2
$68.540.
1,270.
1,550.
MC30
114.1 h)
1.040
195
122
'
(5)
3
2
13.8
$8,690.
620.
630.
TOTAL
(330.7 hi
9,026
9,794
3,652
1,476
2.2
5.2
4.0
36.8
13,079
442
( 102 TOTAL)'
70
3
10
3
4
6
4
2
3
35.6
275.3
$378,330.
1,140.
1,370.
"These laterals were part of the earlier canal and lateral lining study and contain approximately an additional 1390 meters of concrete ditches and 790 meters of concrete
pipe not included above,
"This lateral was part of a previous drainage study and contains an additional 3353 meters of plastic drainage tile on 4 hectares not, included above.
These flumes were removed at the end of the project since they measured field runoff.
'Interceptor drains, concrete tile. HL C tiled a targe open drain, HL E is a new drain.
'Includes 99 meters of 25 and 38 cm diameter concrete pipe.
*m = meters, h — hectares, No. = number.
This total flow measurement count does not include the flow measurement structures used in monitoring the hydrology for the whole demonstration area.
381
-------
CASE STUDY: GRAND VALLEY
was facilitated by contractors, some irrigators
were out every day asking questions and mak-
ing suggestions on construction procedures and
improvement of system performance. The
willingness of the irrigators to become involved
in the construction is desirable because they
develop a much better understanding of the
system design, operation, and maintenance.
However, on one lateral (GV 95) the fact that the
irrigators opted to pay for the construction
contributed to many problems encountered later
in the project. Due to the lack of daily involve-
ment, many of these irrigators lacked a com-
plete understanding of the system and its opera-
tion. Ultimately, this caused some conflicts
which should not have occurred. However, with
considerable time and effort, these conflicts
have been resolved.
D Rood
N
El'
rO il
§i
0«
ii
[;
0 50 100
Scole m meters
30cm, 374m
5 .. c -i . 3 8 -
v6.1 hectares, 10 cm Dia.
,'Plastic Tile ^j&SS&Sjij&J
M2m Centers .X\k-'''."•.'•.'•''••.'•'.'•'-
Mesa County
Legend
Drainage Ditch
Road
Canal
Field Boundary
Concrete Ditch
Buried Pipeline
Gated Pipe
Sprinkler Irrigation
Drainage
Figure 3. Map of lateral and on-farm improvements under MC 10 lateral system.
382
-------
TECHNOLOGIES FOR SALINITY CONTROL — GRAND VALLEY
The construction work was very personally
gratifying in many ways. Many people went out
of their way to help and assist others on the
laterals. For example, on the Price Ditch 177
lateral subsystem, several people from the sub-
division on the end of the lateral assisted in
laying pipeline for agricultural users, and two of
these people donated equipment for construc-
tion work on the project.
In another case, on the Grand Valley Canal
160 lateral subsystem, people without water
rights in the lateral assisted their neighbors in
pipeline construction. Also, on this lateral, an
8-inch diameter plastic pipeline replaced more
than a quarter mile of unlined ditches to one
farm. The pipeline was completely installed by
neighbors of of the family because of head of the
household had suffered a heart attack.
On the Mesa County 10 lateral subsystem,
people donated their own equipment for the
construction of the pipeline system. One elderly
gentleman with just a few acres had his sons
come and do his share of the work on the pipe
installation.
These examples illustrate occurances on
many of the laterals where community effort
was require to complete the work.
Cooperation Among Irrigators
In general, mutual cooperation between
irrigators on the project has improved quite
noticeably since construction began, particular-
ly where participants did much of the installa-
tion themselves. The new systems have reduced
cases of antagonism due to ditch maintenance
or inequitable water allocation. For instance,
under the new systems, yearly maintenance is
reduced because of concrete lined laterals or the
use of pipelines. Also, ineffective or outdated
division structures were replaced with new
structures containing flow measuring devices,
removing many areas of previous discontent.
Prior to the project some of the laterals had
previously developed water rotation systems.
The construction of these new systems greatly
facilitated the ease and speed of water delivery,
and contributed to the development of a new
awareness of water rights through water
measurement. Consequently, rotations have
become more widely accepted as a beneficial
practice.
The construction process also helped the
irrigators become much more aware of water
delivery problems. They now have more con-
sideration for their neighbors. There is more
communication between irrigators in deter-
mining the timing and amount of deliveries
because of increased emphasis upon improved
water management practices.
Participation by Local Organizations
There was a large amount of participation
by local organizations, which substantially
contributed to the project's success. The largest
degree of participation was by the Grand Junc-
tion Drainage District, closely followed by the
Mesa County Road Department and the local
irrigation companies.
The Grand Junction Drainage District,
through their standard drain installation agree-
ment, installed almost 13 kilometers (8 miles) of
drainage tile for this project. The Mesa County
Road Department installed and replaced
numerous road crossings and culverts and
provided some backfill for deeply eroded areas
near county roads. The Grand Valley Irrigation
Company and the Grand Valley Water Users
Association suggested ways of modifying
lateral operational procedures and replaced
worn out headgates, and they agreed to let
project personnel manage the operation of
headgates on the selected laterals. This was a
very important component of the project's
success in demonstrating improved water
management practices.
Meetings were also held with the Grand
Valley Rural Electric Association, the Mesa
County Tax Assessor, Mountain Bell Telephone
Company, local natural gas companies, bank
officials, attorneys, and water and sewer dis-
trict officials to obtain necessary information,
cooperation, and other assistance on construc-
tion easements, utility locations and
relocations, possible legal problems, and
various financial aspects of the project.
Almost 610 meters (2000 linear feet) of 30-
cm (12-inch) diameter plastic irrigation pipeline
was installed, free of charge, by a local construc-
tion company. They installed the pipe because
lateral GV 160 crossed about 122 meters (400
feet) of land belonging to the company. In order
to make the area more usable for their construc-
tion related activities, the contractor offered to
install the pipe at no charge. The project
supplied all the materials.
In general, the support given by local
business and organizations was overwhelming
383
-------
CASE STUDY: GRAND VALLEY
and undoubtedly a large reason for the attain-
ment of the project's objectives. The Grand
Junction Daily Sentinel, a local daily newspa-
per, was also of great assistance in promoting
project goals and reporting on project activities
and developments.
Summary of Improvements and Costs
In summary, the improvements completed
in the project area since 1969 as part of the
demonstration of salinity control include: 12.2
km (7.6 miles) of large canal linings. 26.59 km
(16.51 miles) of lateral linings, 16432 meters
(53,913 feet) of perforated field drainage tile,
construction of a wide variety of on-farm im-
provements, and an irrigation scheduling
program. The costs of the various im-
provements, which totaled almost $750,000, are
listed in Table 2.
The total value of the lateral improvements
(since 1974) for the project is $378,330, installed
on 330.7 hectares for an average cost of $1140
per hectare ($460 per acre).
REFERENCES
1. Skogerboe, G. V. and W. R. Walker.
1972. Evaluation of Canal Lining for Salinity
Control in Grand Valley. Environmental
Protection Technology Series, EPA-R2-72-047.
Office of Research and Monitoring, U.S. En-
vironmental Protection Agency, Washington,
D.C. October.
2. Skogerboe, G. V., W. R. Walker, J. H.
Taylor, and R. S. Bennett. 1974. Evaluation of
Irrigation Scheduling for Salinity Control in
Grand Valley. Environmental Protection
Technology Series, EPA-660/2-74-052. Office of
Research and Monitoring, U.S. Environmental
Protection Agency, Washington, D.C. June.
3. Skogerboe, G. V., W. R. Walker, R. S.
Bennett, J. E. Ayars, and J. H. Taylor. 1974.
Evaluation of Drainage for Salinity Control in
Grand Valley. Environmental Protection
Technology Series, EPA-660/2-74-084. Office of
Research and Development, U.S. Environmen-
tal Protection Agency, Washington, D.C.
August.
TABLE 2
Construction program of previous improvements in the
Grand Valley Salinity Control Demonstration Project.
Map
Designation
Company Same
Canal \'amt
Area I ^Demonstration Aral
A Grand Valley Irrigation Co
Mesa County Canal
B Palisades Irrigation Din.
Price Ditch
C Grand Valley Waters User*
Assn Gov't Highline Canal
D Mesa County Irrigation Co.
Stub Ditch
£ Grand Junction Drainage Co.
Open Drains
Closed Drains
Laterals
Tvpfof Length Perimeter Area L'nifCosr
Improvement Iml 1\d*j fm't (tvd't dm'!
Gunit* Lining 2.2 3.5 14 4.3 17.500 14.632 3.25 3.89
Slip Form Lining 1.9 3.1 15 4.6 16.T20 13.980 325 3.89
Gunite Lining 1.0 1.6 15-*- 4.6 S.8OO 7,358 3.50 4.19
Slip Form Lining 25 4.0 10 3.1 14.700 12.290 325 3.89
Slip Form Lining
Tile
Slip Form Lining 4.83 7.77
Miscellaneous Total
Costs Cost
Id <$l
2.100.00 58.975.00
2.900.00 56.240.00
5.800.00 36.6OO.OO
3.500.00 51.275.00
4.000.00
16,000.00
110.815.00
Are. 11
F Grand Valley- Irrigation Co. Gunite Lining 0.15 0.24
Grand Valley Canal
Area III
G Redlands Water and Power Slip Form Lining 05 0.8
SUBTOTAL
(ftl tm>
Drainage Costs 11.000 3.353
SUBTOTAL
Lateral Improvements
TOTAL VALUE of direct benefits to the Grand Valley
15+ 4.6 1.320
12 3.7 3.500
(in.! Icml lac!
4J 10.2J 10
1.104 3.50
2.926 3.25
lhai IS acl
4.1 1,694.00
4.19
3.89
IShai
4.185.82
4,000.00 8.620.00
1,600.00 11,475.00
354,000.00
0.00 16,940.00
*370,940.00
378.330.00
»749,270.00
•Costs of pre-constructiOD and post-construction ponding tests above amounts in CSU contract, plus costs of installing headgates. etc.
tTWumhill bank KB ing _ onlv.
tDiameter of tile.
384
-------
Development of Best
Management Practices
for Salinity Control in Grand Valley
WYNN R. WALKER, GAYLORD V. SKOGERBOE, and
ROBERT G. EVANS
Department of Agricultural and Chemical Engineering;
Colorado State University; Fort Collins, Colorado
ABSTRACT
Cost-effectiveness functions for various on-
farm, conveyance, and desalting salinity con-
trol alternatives were developed for the Grand
Valley of western Colorado. These functions
were then optimized to determine the best
management practices on a valley-wide scale.
The results indicate the relative importance of
lateral linings and on-farm improvements over
either canal lining or desalting. Generalized
curves are presented for application of this
analysis to other similar areas where less infor-
mation is available.
INTRODUCTION
The array of potential practices which can
be applied in an irrigated region to improve the
quality of irrigation return flows, or reduce the
impact of these discharges on receiving waters,
might include such structural changes as canal
and lateral linings, drainage, and conversion to
more site-suitable irrigation systems. In addi-
tion, improved irrigation practices could also
produce benefits. For instance, aiding the
irrigator with an irrigation scheduling service
to maximize his water application efficiency
would diminish the volume of irrigation return
flows. Various other improvements in designs
or practices might be noted, some of which
would be more applicable than others to the
multitude of conditions which exist in irrigated
areas in the western United States.
The costs associated with water quality
problems and the measures necessary for their
resolution are generally high, particularly when
the problem results from irrigated agriculture. It
is therefore important to implement only the
management alternatives which promise the
most cost-effective treatment of the problem,
subject of course, to environmental and political
constraints. The ultimate use of research and
demonstration data is in determining these best
management practices to be implemented for
the solution to the water quality problem.
The Grand Valley in western Colorado
contributes substantially to the serious salinity
problem in the Colorado River Basin, and has
therefore, been given considerable attention not
only to remedying its particular impact but also
in testing methodologies for application to
similar areas. A previous paper in this case
study described the efforts made to quantify the
components of the valley's complex hydro-
salinity system and the demonstrations of alter-
native management practices. These results
provided the data necessary to determine the
optimal policies for implementing a salinity
control program. The purpose of this paper is to
summarize the procedure and illustrate the
results for the Gand Valley case study.
GENERAL APPROACH
Optimizing the alternatives for managing
salinity implies that the relationships between
costs (or benefits) and salinity reductions are
known. These cost-effectiveness functions are
then fashioned into a format suitable for an
optimizational analysis (minimization of costs
or maximization of benefits, for example), and
the optimal is computed. In mathematical
terms:
n
Z = min 2 ff (X j
i=l
in which,
(1)
385
-------
CASE STUDY. GRAND VALLEY
Z = objective function; and
f i (X i ) = cost of implementing the ith
alternative at a level X j .
Equation 1 is of course restricted by a set of con-
straints such as:
I Xi>Xi
t=l
Xi
-------
MANAGEMENT FOR SALINITY CONTROL
4. Application Efficiency
— Cropland Evapotranspiration
Farm Deliveries-Tailwater + Precipitation
= 64%
On-farm water management improvements
may be divided into four general categories: (1)
improved irrigation practices implemented
through irrigation scheduling; (2) structural
rehabilitation; (3) conversion to more effective
irrigation methods; and (4) field drainage.
Irrigation Scheduling
Recent studies by the writers in Grand
Valley have indicated that irrigation schedul-
ing services, even when accompanied by flow
measurement structures, generally do not im-
prove farm and application efficiencies by more
than 5-10% (Skogerboe, et al., 1974a). A west-
wide review of irrigation scheduling by Jensen
(1975) indicated that a 10% improvement (from
40 to 50%) is realistically possible without
system conversions or more energy intensive
operations. In the Grand Valley, an irrigation
scheduling service which included water
measurement and farmer training would cost
an estimated $30/ha and would reduce return
flow salinity by about 20,000 metric tons an-
nually. Since it is not known how irrigation
efficiencies may be distributed it is assumed
that these figures may be linearly extrapolated
yielding a cost-effectiveness function for irriga-
tion scheduling of $37.50/ton with a limit of
20,000 metric tons amenable to this approach.
The overall impact of irrigation scheduling
being only 10% of the total estimated on-farm
potential improvement is insignificant by itself
when considering the sensitivity of these type of
costing estimates. Consequently, irrigation
scheduling is not considered a separate alter-
native salinity control measure, but rather
taken to be an integral part of each of the
remaining remedies. This assumption is par-
ticularly valid when viewing the costs of the
other alternatives (irrigation scheduling's low
cost would result in its inclusion in any im-
plementation policy generated by an op-
timizational analysis).
Structural Rehabilitation
Irrigation efficiency can often be substan-
tially improved by rebuilding and remodeling
existing systems. The most commonly
employed irrigation methods in the west are
those generally classified as surface irrigation
methods (furrow, border, or basin). Structural
improvements in this system may include con-
crete lined head ditches and gated pipe to reduce
seepage losses, land leveling for better water
application uniformity, adjusting field lengths
and water application rates to be more con-
gruent with soil and cropping conditions, and
automation to provide better water control.
Flow measurement and scheduling services
should accompany these types of improvements
in order to maximize their effectiveness.
In the Grand Valley, head ditch re-
quirements are generally less than the capacity
of the smallest standard ditch available
through local contractors (12-inch, 1:1 side
slope, slip form concrete). Consequently, lin-
ing costs can be expected to be linearly distri-
buted. In an earlier paper, an equation was
presented to estimate concrete lining costs:
C c = 40.1Q °'56 (4)
where,
C c = total lining cost, $/m; and
Q = channel capacity, m3/sec.
Assuming an average head ditch capacity of
0.05 m^/sec, Equation 4 yields an estimated
unit cost of $7.50/m. This figure is well within
the range encountered in the last two seasons in
the valley. As an alternative to concrete linings,
six-inch diameter aluminum pipe costs almost
exactly the same. Thus, both improvements
could be arbitrarily substituted with equal cost-
effectiveness depending on farmer preference.
There are approximately 1.3 million meters of
head ditches in the Grand Valley contributing
an estimated 30% of the total salt load from on-
farm sources. As a result, the cost-effectiveness
of head ditch improvements is $113.507 ton. The
limit of salt reductions through head ditch
linings is about 86,000 tons annually.
Adjusting field lengths and land leveling
probably has little or no salinity related benefits
in the valley because infiltration rates are so
low. In fact, advance rates are approximately
25% of the irrigation interval and uniformities
are already high.
Automatic cutback furrow irrigation
systems have demonstrated which when com-
bined with irrigation scheduling, may improve
irrigation efficiencies 15 to 20% (Evans, 1976).
387
-------
Cost
Cost of achieving desired
salinity control at
level 4
LEVEL 4 SALINITY CONTROL
COST-EFFECTIVENESS FUNCTION
: Optimal investment m
' alternative 2 at level 3
; Optimal investment in
alternative I at level 3
Cost
Optimal level 3
Costs from level 4
ALTERNATIVE I, LEVEL 3
COST-EFFCTIVENESS
FUNCTION
Desired salinity control
at level 4
x
alt. I, level 2 investments
alt. 2, level 2 investments
Cost
,'ALTERNATIVE 2,
LEVEL 3 COST-
EFFECTIVENESS
FUNCTION
O
fe
M
I
a
o
Effectiveness
Effectiveness
Cost
ALTERNATIVE 1, LEVEL 2
COST-EFFECTIVENESS
FUNCTION / ,
' costs
Optimal level 2
Effectiveness
Cost from level 3
Cost
ALTERNATIVE 2 LEVEL2
COST-EFFECTIVENESS
FUNCTION
level I
/ costs
Optimal level 2
Cost from level 3
Effectiveness
Cost
ALTERNATIVE 3 LEVEL !
COST- EFFECTIVENESS,
FUNCTION
/ level I
costs
Effectiveness
Cost
ALTERNATIVE 4 LEVEL
COST-EFFECTIVENESS
FUNCTION
' ^ level I costs
Effectiveness
Figure 1.
Illustration of the multi-level, multi- divisional nature of optimal management policies for regional salinity control.
-------
MANAGEMENT FOR SALINITY CONTROL
In 1975, the installed cost of the cutback
systems was $11.50/m. Thus, in addition to the
salt load reductions by lining the head ditch
(86,000 tons), an additional 60,000 m ton reduc-
tion could be expected due to increased irriga-
tion efficiency. The resulting cost effectiveness
is then computed as $102.50/ton, again assum-
ing a linear distribution. In the case of the
Grand Valley, it appears more cost-effective to
implement a more costly but efficient improve-
ment for the additional efficiency related salini-
ty reductions which are possible.
For other areas, where head ditch capacities
are large, concrete lining would generally be
more cost-effective than piped systems.
Whether or not automated cutback would enjoy
the advantage over regular linings would re-
quire evaluation under those specific con-
ditions.
System Conversion
In order to completely control the quality
and quantity of irrigation return flows in an
area like Grand Valley, irrigators would need to
convert their irrigation to either sprinkler or
trickle irrigation systems. Then with proper
management, irrigation efficiencies could be
improved to 75-85% with sprinklers and 90-95%]
with trickle systems. Portable or mobile
30r
_- 20
M
I
-
g
I
'Improvements of
Existing Systems
I. Heod Ditch Lining
2.Irrigation Sched.
3. Cutbock Systems
i. Side Roll Sprinklers
; 2 Drip Irrigation
! 3. Limiled Tile Drc
100 200
Annual Salt Load Reduction, m Ionx 10~
Figure 2. Optimal on-farm water management
strategies in the Grand Valley.
sprinkler systems (not including center-pivot
types) for the Grand Valley condition are
costing approximately $1250/ha with solid set
sprinklers and automated trickle systems ap-
proximately double this figure. At the time of
this writing, we have not evaluated these two
irrigation methods in sufficient detail to con-
firm their respective feasibilities. However, it
appears the added efficiency of trickle systems
is about equally offset by the higher costs.
Consequently, within the sensitivity of the
estimates, sprinkler and trickle systems enjoy
about the same advantage and could be applied
where individual circumstances dictate. The
Grand Valley cost-effectiveness for pressurized
irrigation systems is computed to be ap-
proximately $145/ton and could be expected to
essentially eliminate excessive deep percolation
(beyond leaching requirement) and field
tailwater. However, these systems are much
more energy intensive which was not given
major attention in this paper.
Field Drainage
The low permeability of Grand Valley soils
dictate relatively close drain spacings (12-24
meters). Although field drainage has been
proven partially effective in reducing salt
pickup (Skogerboe, et al., 1974b), the costs are so
high that drainage would not be competitive
with other salinity control measures. Therefore,
we will not offer further consideration to
drainage in this paper.
Optimal On-Farm Improvement Strategies
The cost-effectiveness relationships for the
previously described on-farm improvement
possibilities were examined in an optimization
context. The results shown in Figure 2 are
interesting. Total capital costs in millions of
dollars were plotted against the expected reduc-
tion in annual salt load from the valley for the
minimum cost array of practices. The curve
agreed with data reported recently by the Soil
Conservation Service in a public information
brochure.
Two major strategies evolved in the
analysis of on-farm improvements: (1) im-
provements to the existing system creating
salinity reductions up to about 150,000 tons; and
(2) system conversions to provide controls up to
approximately 250,000 tons. Of particular in-
terest here, is the fact that both alternatives are
mutually exclusive. In other words, in im-
389
-------
CASE STUDY: GRAND VALLEY
plementing an on-farm salinity management
plan, either one or the other is optimally chosen.
For instance, if planners selected on-farm im-
provements to reduce salinity by more than
150.000 tons, the alternatives would be limited
to changing to sprinkler and drip irrigation
methods. Below the 150.000 ton figure, im-
provements to the existing systems would be
optimal. This structure of the cost-effectiveness
functions is unique among the alternatives as
the reader will note in succeeding sections. This
uniqueness is based on the fact that on-farm
improvements themselves are mutually ex-
clusive and limited in their expected effec-
tiveness. For example, head ditch linings would
not be considered in the conversion to a
sprinkler system.
Lateral Lining and Piping
We have defined laterals as the small
capacity conveyance channels transmitting
irrigation water from the supply canals and
ditches to the individual fields. Most of these
laterals operate in a north-south direction and
can carry the flows in relatively small cross-
sections. Although the capacities of the laterals
may vary between 0.06 and 1.4 m^ sec, most
capacities would be within the range of 0.06 to
m3 sec. Utilizing a median value of 0.20
m •'* sec yields a concrete lining cost of ap-
proximately $16 m. Alternative use of PVC pipe
approximates concrete lining costs for this
capacity and a further distinction will not be
made. However by this assumption we are
neglecting the small seepage losses which
would still occur from concrete lined channels.
An earlier paper noted that Grand Valley
laterals extend approximately 600,000 meters,
less than one half the length of field head
ditches. Seepage under existing conditions con-
tributes about 202,000 metric tons, or slightly
less than the on-farm contribution. Although no
attempt was made to distribute the lateral
lining costs to account for variable capacity (at
least for the purposes of this paper), the cost-
effectiveness function for Grand Valley lateral
lining is about $49.50 per metric ton. Thus, the
estimated costs of lining the total lateral system
in the valley is about $10 million.
Canal Lining
The analysis of canal and ditch linings for
the Grand Valley was presented in an earlier
paper of this conference by the senior author
entitled, "Combining Agricultural Im-
provements and Desalination of Return Flows
to Optimize Local Salinity Control Policies."
The results are repeated in Figure 3, in which
the upper curve represents the minimal cost
curve. At any point along this curve, a vertical
cost distribution among the alternative canals
indicates the optimal investment in each
system.
Redlandi System
10
40 5O 60 70 80 90 IOO 110
Annual Salt Load Reduction, metric ton/ilO"'
Figure 3. Minimum cost canal and ditch lining
strategy for the Grand Valley.
Desalination
Application of desalting technology was
also discussed in the earlier paper mentioned
above. These results will not be reiterated here
other than to say a desalting cost-effectiveness
function for the Grand Valley assuming pump
drainage, reverse osmosis, and deep well injec-
tion of brines would be approximately linear at
$310 per metric ton of salt removed.
Determining Best Management Practices
The individual cost-effectiveness functions
for each previously discussed alternative can be
integrated to determine what might be called an
optimal overall policy for the Grand Valley. To
present this on a somewhat general basis, the
distribution of the cost effectiveness functions
has been computed and presented in Figure 4.
Lateral linings and desalting are represented as
linear distributions
390
-------
MANAGEMENT FOR SALINITY CONTROL
Y = X (5)
in which
Y = fraction of the total costs; and
X = fraction of the total potential salinity re-
duction associated with the linings.
i.o
0.9
o.e
0.7
0.6
and canal linings:
I 0.5
I 0.4
"o
20.3
0.2
O.I
0
L
yx
Desalting \
oteral Lining!
(On-Farm Improvements)
y- 0.525H + O.OI5** + 0.46xs
(Canal Lining)
O.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Fraction of Total Salinity Reduction Associated
with Respective Control Measure
1.0
Figure 4. Distribution of cost-effectiveness func-
tions for the primary salinity control alternatives
in the Grand Valley.
Equation 5 assumes that the total costs achieve the
potential salinity reduction attributable to that spe-
cific measure, with the potential being the measure of
maximum effectiveness. For on-farm improvements
and canal linings, respectively
Y = 0.645X + 0.295X2 + 0.08X3 (6)
Y = 0.525X + 0.015X2 + 0.46X3 (7)
The appropriate limits can be substituted
into Equations 5-7 to yield the problem's objec-
tive functions. For desalting, the liner structure
provides that:
Y d = 310 X d (8)
where,
Y d = capital cost of removing X d metric tons from
the irrigation return flow system.
similarly for lateral linings:
Y l = 49.50 X l
For the on-farm improvements:
Y f = 29.7 x 10«
(9)
0.645[ - - — )
L \2.5xlOs/
/ Xf \2 / Xf VH
0.295[ - - - + 0.08 - — )
\2.5 x 10s/ \2.5 x 105/ J
Y c = 40.xl06
0.525 f—
/ Xc \2 / Xc \3~]
0.015 S—) +0.461 —)
Vl.lxlO5/ Vl.lxlO5/ J(ll)
where the l.lxlO5 factor in Equation 11 is the
salt reduction possible from canal ditch linings
as reported in an earlier paper. The objective
function is therefore:
Z = minYd (Xd
Yf (Xf ) + Yc (Xc
(12)
in which the individual costs are represented as
functions of the expected salt reduction. Equa-
tion 12 must be constrainted to account for the
maximum potential of each approach. Thus,
X i
S f
X c
and,
< 202,000 tons
< 250,000 tons
< 110,000 tons
(13)
(14)
(15)
(16)
where,
(10)
— the total combined effectiveness to be
achieved, tons.
The result of this analysis for a wide range of
X T 's is shown in Figure 5.
DISCUSSION
In representing what we have chosen to call
the best management practices for the Grand
Valley, Figure 5 deserves several explanatory
notes. First, it must be realized that the four
major implementation alternatives (lateral lin-
ing, on-farm improvements, canal linings, and
desalting) only represent what might be
denoted as "structural measures." Consequent-
ly, non-structural alternatives such as land
retirement, influent and effluent standards,
taxation, and miscellaneous enforcement op-
tions are not included. The structural im-
provements would, however, consider irrigation
education (irrigation scheduling and extension)
type activities.
The second point which should be examined
is the value of this sort of analysis. In Grand
391
-------
CASE STUDY: GRAND VAT.LEY
200 SOO 400 500 600
Total Salinity Reduction, m Ion x 10*'
TOO
Figure 5. Minimum cost salinity control strategy
for the Grand Valley.
Valley, existing plans call for the lining of
canals, ditches, and laterals in combination
with some on-farm improvements. Desalting is
probably not being considered. Canal linings
would cost approximately 40 million dollars (in
1976 costs) and reduce salinity 110,000 tons.
Lateral lining is estimated to cost about $30
million by the Bureau of Reclamation, but
should be possible for at least one-third this
figure (10 million dollars) and would reduce
salinity by about 202,000 tons annually. Under
their most intensive plan, the Soil Conservation
Service estimates that $23 million could be
spent for on-farm improvements to produce a
209,000 ton improvement. (Our figures indicate
that $23 million would impact on-farm salt
sources by 205,000 tons annually.) Therefore, 73
million dollars would be invested in improving
irrigation return flow quality by an equivalent
517,000 tons ($141.20 per ton). Examination of
Figure 5 indicated that a salt load reduction of
517,000 tons could be achieved at a cost of only
$57 million by using less canal lining and
increasing on-farm improvements. This is a
savings of more than 25% or 16 million dollars.
A third point of interest is "How much
salinity control is feasible, given current es-
timates of downstream detrements?" Walker
(1975) reviewed much of the literature descrip-
tive of the California, Arizona, and Republic of
Mexico damages. At that time, Valantine (1974)
had proposed damages of $175,000 per mg/1 of
increase at Hoover Dam ($146 per ton in Grand
Valley assuming 8% interest). Other estimates
in terms of equivalent damages attributable to
Grand Valley range upward. A representative
figure is $190/ton as proposed by the Bureau of
Reclamation (Leathers and Young, 1976). If the
minimum cost curve in Figure 5 is differentiated
to approximate marginal costs and be con-
gruent with these damage figures, the $146 per
ton damage estimate of Valantine(1974) falls at
a 400,000 ton reduction, while the $190 per ton
occurs at 450,000 tons. Thus, not considering
secondary benefits in the Grand Valley, or
obviously all the consequences in the lower
basin, the level of investment in the Grand
Valley should not exceed $30-40 million.
Otherwise, the costs are apparently not justified
by the damages and another salt contributor
should be evaluated for possible control. Our
conclusion is, therefore, considering the
preliminary stage of analysis encompassed by
this paper, the best management practices for
the Grand Valley should be limited to on-farm
improvements and lateral linings. If
downstream damage can be shown to be higher,
then limited canal lining should proceed.
Furthermore, irrigation improvements should
consist of conversion to other methods rather
than treatments of the existing system. This
conclusion could of course be amended if local
farmers would adapt the measures necessary to
achieve 85-90% surface irrigation efficiencies
(which would imply mandatory compliance
with irrigation scheduling criteria and com-
prehensive automation).
And a final point to be made herein con-
cerns the sensitivity of these results to the
assumptions in the analysis. Our estimates
would need to be approximately 100% wrong to
affect the respective feasibility of lateral linings
versus on-farm improvements, and roughly the
same between the remaining alternatives. Such
magnitudes of error are improbable given the
years of experience and the level of research
effort applied to the Grand Valley by the
authors and many other federal and state
agencies.
REFERENCES
1. Evans, R. G. 1976. An Improved Cut-
Back Surface Irrigation System. Paper 76-2048,
Presented at 1967 Annual Meeting of ASAE.
Lincoln, Nebraska. June.
2. Jensen, M. E. 1975. Scientific Irrigation
Scheduling for Salinity Control of Irrigation
392
-------
MANAGEMENT FOR SALINITY CONTROL
Return Flow. EPA-600/2-74-064. Robert S. Ken-
Environmental Research Laboratory, Office of
Research and Development, Environmental
Protection Agency, Washington, D. C. August.
3. Leathers, K. L. and R. A. Young. 1976.
Evaluating Economic Impacts of Programs for
Control of Saline Irrigation Return Flows: A
case study of the Grand Valley, Colorado.
Report for Project 68-01-2660, Region VIII, En-
vironmental Protection Agency, Denver,
Colorado. June.
4. Skogerboe, G. V., W. R. Walker, J. H.
Taylor, and R. S. Bennett. 1974a. Evaluation of
Irrigation Scheduling for Salinity Control in the
Grand Valley. EPA-660/2-74-052. Office of
Research and Development, Environmental
Protection Agency, Washington, D. C. June.
5. Skogerboe, G. V., W. R. Walker, R. S.
Bennett, J. E. Ayars, J. H. Taylor. 1974b.
Evaluation of Drainage for Salinity Control in
Grand Valley. EPA-660/2-74-084. Office of
Research and Development, Environmental
Protection Agency, Washington, D. C. August.
6. Valantine, V. E. 1974. Impacts of
Colorado River Salinity. Journal of the Irriga-
tion and Drainage Division, American Society
of Civil Engineers, Vol. 100, No. IR4, pp. 495-
510. December.
7. Walker, W. R. 1975. A Systematic
Procedure for Taxing Agricultural Pollution
Sources. Grant NK-42122, Civil and En-
vironmental Technology Program, National
Science Foundation. Washington, D. C. Oc-
tober.
393
-------
Implementation
-------
The EPA General Permit Program
KATHE ANDERSON
U.S. Environmental Protection Agency;
Washington, D.C.
ABSTRACT
The Environmental Protection Agency
(EPA) has developed a General Permit Program
under the National Pollutant Discharge Elima-
tion System (NPDES) which applies to both
separate storm sewers and agricultural point
sources (irrigation return flow conveyances).
The proposed rules for this general permit
program were printed in the Federal Register,
Vol. 42, No. 24, February 4, 1977. This paper
provides a guide as to how the General Permit
Program is intended to function.
Purpose for the General Permit Program
The regulations are being developed to
comply with the court order following the case of
NRDC v. Train, better known as the Flannery
decision. That decision required EPA to develop
and administer a permit program for all point
sources in the feedlot, separate storm sewer,
agriculture and silviculture categories. The con-
ventional NPDES permit program was applied
to feedlots and silvicultural point sources, but
this general permit program will be applied to
separate storm sewers and agricultural point
sources (irrigation return flow conveyances).
The discharge sources affected by this general
permit program are defined for separate storm
sewers at page 11303, and for agricultural point
sources at page 28493, both in volume 41 of the
Federal Register.
Goals
This general permit program is being
proposed to meet the following goals:
A. To enable the permit-issuing agency to
issue an authorization to discharge to a number
of people with one simplified administrative
procedure. This will allow the agency to focus its
attention on the most severe pollution problems.
B. To enable the permit-issuing agency to
coordinate better with forthcoming 208 and
other local plans. Because universally appli-
cable techniques for controlling and abating
pollution from separate storm sewers and
agricultural point sources are currently un-
available, locally-oriented areawide plans will
better determine what pollution reduction prac-
tices are required.
C. To progress towards a comprehensive,
well-managed and effective program for im-
proving water quality. Much needs to be done to
gather information; develop pollution control
practices and technology; and enhance coopera-
tion among affected persons and groups,and
involved local, State and Federal agencies.
General permits will help define the problem,
establish priorities and direct resources toward
these ends.
D. To provide flexibility of approach. The
problems faced by owners and operators of
separate storm sewers and agricultural point
sources vary considerably throughout the coun-
try. This proposed program provides a number
of options for dealing with different problems
and thus allows for maximum flexibility within
the requirements of the court order.
Outline of the Program
If these regulations are finalized in the
same form as they are proposed, the following
outline shows the steps the Federal government
will take to administer the program.
Step 1: The Regional Administrator will
draw a map of the area in which a
general permit will be issued. This
area is called a general permit
program area (GPPA) and could
cover separate storm sewers or
agricultural point sources, or both.
Step 2: The Regional Administrator will
publish a notice in local papers and
post a notice in public places in the
397
-------
IMPLEMENTATION
GPPA saying that he is proposing
to issue a general permit affecting
all owners and operators of
separate storm sewers and/or
agricultural point sources in that
area. The notice will also indicate
what kinds of conditions the
affected owners and operators may
be subject to under the proposed
general permit.
Step 3: The Regional Administrator will
also send copies of the proposed
general permit to the District
Engineer of the Corps of Engineers,
the U.S. and State agencies for
agriculture and commerce, the ap-
propriate 208 planning agencies
and the State water pollution con-
trol agency in the State where the
discharge will occur. These agen-
cies and organizations may then
comment on the general permit as
proposed.
the general permit as proposed or to
modify it in some way. If the propos-
ed general permit remains intact, it
will probably become effective in a
month. If the general permit is
substantially modified from when
it was first proposed, the Regional
Administrator will publish and
post the changes as he did the
original notice.
Step 7: After the conditions of the general
permit have been finally deter-
mined, the permit will be published
in the Federal Register and in other
appropriate places in the area of the
GPPA. Any person wanting an in-
dividual copy of the permit may
request one from the Regional Ad-
ministrator. Once the general per-
mit becomes effective, all affected
owners and operators must comply
with the permit's terms and con-
ditions.
Step 4: After the notice is published and
posted, affected owners and
operators have at least 45 days to
comment on the proposed general
permit. If any person wants more
information or would like to request
a public hearing on the proposed
general permit, those opportunities
will be available.
Step 5: If several persons request a public
hearing, the Regional Ad-
ministrator may hold a hearing to
gather information concerning the
proposed general permit and to
hear the opinions of the persons
present. In holding a public hearing
the Regional Administrator will
make sure that all those persons
who would be interested in it will be
notified of the time, date and place,
and other pertinent information.
Step 6: After a public hearing, the Regional
Administrator will consider all the
comments received on the proposed
general permit and make final
determinations on whether to issue
Step 8: Should any affected persons dis-
agree with the terms and conditions
of the general permit, that person
may request an evidentiary hearing
(also known as an adjudicatory
hearing) within 10 days after the
general permit is published in the
Federal Register. Any request for
an evidentiary hearing will be
handled like those requests under
the conventional NPDES program.
It should also be noted that not every owner
or operator of any agricultural point source or
separate storm sewer may agree with the
general permit approach. Indeed, some such
owners or operators already have conventional
individual NPDES permits, and would prefer to
remain subject to those permits rather than to
be subject to a new general permit. Thus, it is
EPA's intent to allow those owners or operators
to keep their individual permits as they so
desire. In fact, even owners or operators of
agricultural point sources or separate storm
sewers who do not currently hold NPDES
permits have the option to apply for individual
permits rather than be subject to the general
permit. Procedures for obtaining such an in-
dividual NPDES permit are set forth in the
proposed regulations.
398
-------
EPA PERMIT PROGRAM
A Graphic Example of
the General Permit Program
\GpoPA
\5 '
The NPDES permit program is ad-
ministered by EPA in State A. The Regional
Administrator has designated four (4) GPP As
in State A as follows:
GPPA 1: This area, defined by the boun-
daries of a major river basin, is primarily
agricultural and is populated by hundreds of
owners and operators of agricultural point
sources (irrigation return flow ditches). There is
no substantial water quality problem in this
area, but the potential for water quality
problems is recognized. However, this GPPA is
not yet identified as a high priority area for
water quality improvement. Because of the
recognized potential for water quality problems,
however, the general permit for this GPPA
requires owners and operators of agricultural
point sources to submit reports of any sampling
they do. The general permit does not require
sampling at this time, but the term of the permit
is only 3 years so that the need for sampling
may be reassessed in the near future. The
general permit for this GPPA would resemble
the sample general permit included as Part VI of
this guide.
GPPA 2: This area, defined by the bound-
aries of a large Bureau of Reclamation project, is
primarily agricultural and is populated by hun-
dreds of owners or operators of agricultural
point sources (irrigation return flow ditches).
Unlike GPPA 1, this area has long been
recognized as plagued with water quality
problems, particularly siltation. For many
years the conservation districts have worked
with the farmers in the area to reduce erosion
and thereby decrease economic losses and im-
prove water quality. When the 208 agency
identified this area as having significant water
quality control problems, it began to coordinate
its water pollution control activities with ongo-
ing programs, and together with local
agricultural colleges and organizations devised
a 20-year plan for implementing proven tech-
niques to reduce siltation. The first 5 years
included plans to educate farmers to the
program, to train them in the techniques to be
implemented, to work with cost-sharing
programs for aid, to introduce appropriate
legislative changes, and to begin action in
increasing the width of grass filter strips and
planting appropriate protective vegetation. The
local and State governments of this area will
devise and enforce ordinances and regulations
which will ensure the implementation of this
plan. If regulatory programs are not imple-
mented satisfactorily then the general permit
may be strengthened with more substantial
requirements.
GPPA 3: This area, defined by city limits, is
primarily urban and contains several
publically- and privately-owned separate storm
sewers. There is a recognized water quality
problem, but techniques to solve the problem
have not been fully developed, primarily
because there is a lack of information about the
pollution and where it comes from. No 208 plan
requirements have been set forth as yet.
However, because of the need for information,
the general permit requires each owner or
operator to register with the permit-issuing
agency, reporting information on the number
and location of outfalls of their separate storm
sewer systems. Monitoring and reporting may
be general permit requirements at a later date.
GPPA 4: This area is both urban and rural
and encompasses all that area of the State not
specifically identified as GPP As 1, 2 and 3. No
serious water quality problems are currently
recognized by the 208 agency, and thus it has
been determined that no exceptional re-
quirements need be imposed through the
general permit. Thus, the permit simply
authorizes all owners and operators of separate
storm sewers and agricultural point sources to
discharge from those sources, subject only to the
standard language of the general permit. If
water quality problems are recognized at a later
date, regulatory measures to control these
problems must be devised. If these regulatory
measures prove ineffectual, then the general
permit may be revoked as to a portion of the area
(GPPA 5), and a new general permit, containing
more specific and stringent requirements, may
be imposed.
399
-------
IMPLEMENTATION
The variations in the form and re-
quirements of any general permit are dependent
on several factors:
1) The severity of the water quality prob-
lem.
-Does it affect health?
-Does it affect drinking water sources?
-Does it affect recreation facilities or
aesthetics?
2) The context in which the water quality
problem occurs.
-What will be the economic impact on the
area if the problem is corrected?
-What will be the economic impact on the
area if the problem remains unsolved?
-What are the political ramifications of
solving the problem?
3) The interest of the persons affected by
the water quality problem.
-Are farmers interested in solving the
problem?
-Are cities interested in solving the prob-
lem?
-What is the reaction of the community?
-What is the role of EPA in solving the
problem?
-What is the role of the State and local
governments in solving the problem?
-Is an environmental group particularly
interested in the problem?
4) The timetable in which other develop-
ments occur.
-When does the 208 agency complete its
plans?
-When do other plans and developments
impacting separate storm sewers and
agricultural point sources contemplate
implementation?
-When are research, development and
demonstration projects complete and
ready for functioning?
-When are effluent guidelines promul-
gated or appropriate technology identi-
fied?
Ten Most Asked Questions
1) What will a general permit look like?
A general permit could take many
forms depending for the most part on the
variations listed above. An example of a
possible general permit for southwest
Iowa is set forth later in this guide as a
sample for an area where it is assumed
there is no current or potential water
quality problem. As you can see, it
applies to all owners and operators of
separate storm sewers and agricultural
point source in an area of 8 counties. It
contains some standard prohibitions
and other language relating to the legal
responsibilities of the owners and
operators. It contains very few man-
datory provisions requiring action on the
part of most owners and operators. For
those owners and operators who do sam-
ple their discharge, reporting of that
sampling is required.
2) Who is the permittee?
Under this general permit program
there are few named permittees. In some
areas, it may be feasible and desirable to
name the permittees (city sewer districts,
for example). However, in most cases the
permittees will be referred to as the
owners and operators of separate storm
sewers and agricultural point sources.
Owners and operators are those persons
who own, operate or use the conveyances
of separate storm sewers and
agricultural point sources.
3) What about water quality standards?
EPA recognizes that State water
quality standards will not be met over-
night with the implementation of the
general permit program. Meeting water
quality standards is a long term project
that will require the cooperation of many
individuals, groups, organizations and
government agencies. In close coordina-
tion, the affected persons should be able
to work within the existing framework of
208 planning agencies, conservation dis-
trict education and support, and other
currently available and diverse
mechanisms for improving water quali-
ty. Improved water quality is one goal of
the general permit program, but it can
only be met after increased efforts to
coordinate this program with existing,
ongoing programs that have the same
outcome.
4) How does 208 planning fit into this
general permit program?
Although it is still somewhat
speculative exactly how the general per-
40O
-------
EPA PERMIT PROGRAM
mit program will coordinate with 208
plans, the theory is this: Plans developed
under section 208 of the Federal Water
Pollution Control Act are to identify
areas with water quality problems and
methods of solving such problems, in-
cluding management practices and
maintenance techniques and most im-
portantly, regulatory measures to imple-
ment these practices and techniques. If
these regulatory measures are insuf-
ficient, or unsuccessful, then an existing
general permit could be modified, or
individual permits might be issued in
lieu of a general permit. Either a
modified general permit or individual
permits could then incorporate the prac-
tices and techniques identified in the 208
plan. For example, suppose a 208 agency
identified the lower basin of a river as an
area with water quality problems. In
conjunction with the affected farmers of
the area, the 208 agency arranges a
water distribution and application
agreement which will help abate the
problem. Most of the farmers agree with
the arrangement but some are hesitant
about losing their water rights. Because
208 agencies are to develop institutional
and legal mechanisms for bringing
about the desired goals, legislation could
be proposed to preserve all water rights
so long as water is used in accordance
with the agreement.
This is a very simplistic example,
but it illustrates the flexibility of 208
plan as a management tool. It also
illustrates the importance of the role of
concerned citizens in the development of
208 plans.
5) Will farmers have to change the methods
of their operations?
This will depend on the conditions of
the general permit. EPA anticipates that
the vast majority of farmers affected by
the general permit program will not have
to change the methods of their
operations, particularly not in the near
future. Changes may be necessary where
water quality problems are identified.
However, farmers should begin now to
adopt management practices to protect
water quality. Involvement in and
cooperation with 208 planning agencies
can not be overemphasized as a function
of concerned citizens and organizations.
It is through such involvement and
cooperation that the most effective, well-
managed and best-accepted plans will
evolve.
6) Is the general permit program legal?
Yes. Not only does the Federal Water
Pollution Control Act not specifically
exclude the possibility of general per-
mits, Judge Flannery, as the interpreter
of the Act in the NRDC case, specifically
encouraged EPA to develop a flexibile
administrative approach to dealing with
separate storm sewers and agricultural
point sources.
7) What will happen over the next two
years in this area?
There are two anticipated de-
velopments.
A. This regulation will be
promulgated in final form, the
designation of GPP As will take
place, and general permits will begin
to be issued.
B. Since all 208 plans are re-
quired to be submitted by November
1978, water quality control problems
should be clearly identified and im-
plementation of the plans should
begin.
There is an additional possible
development:
Better information and data
could be developed which might
make effluent guidelines feasible for
agricultural point sources or
separate storm sewers, if not un-
iformly, at least in portions of the
Nation. This could pave the way for
use of the conventional NPDES per-
mits issued to individual owners and
operators.
8) What do I have to do if I own a separate
storm sewer or an agricultural point source?
Watch for notices in your area, most
probably in the Fall of 1977, that you are
in a GPPA for which a general permit
is proposed to be issued. Take an active
interest in its issuance. Attend the public
hearing if one is held. Before next Fall
find out about the 208 agency in your
401
-------
IMPLEMENTATION
area — what it is doing about water
quality problems you perceive in your
area, what its plans are, how it will affect
you, and how you can get involved.
9) What should I do if I'm affected by a
water quality problem that I want to help abate?
Contact your State water pollution
control agency and other local agencies
that may not be aware of the problem,
but that do have the authority to aid in
its resolution. Conservation districts,
land grant colleges, farmers'
organizations, environmental action
groups and other concerned persons are
often invaluable sources of information
and ideas. Again, the 208 and other
planning agencies should be utilized as
tools for solving such problems. Working
together with these government and
private organizations will help provide
the appropriate mechanisms for
cooperative effort to progress toward
meeting water quality goals.
10) What should I do if I don't like this
whole program?
Let your opinion be known. Com-
ment on the proposed regulations.
Vocalize your doubts at the forums
provided as part of the public informa-
tion and public participation aspects of
this regulatory process. As you know,
EPA originally did not require permits
for either separate storm sewers or
agricultural point sources. However, as a
result of a 1975 judicial interpretation of
the law, EPA is now required to develop a
permit program applicable to these point
sources. Given this requirement EPA
would appreciate any ideas for alter-
natives to this program. We value your
input.
Sample Permit
Permit No. IA-GP00001
AUTHORIZATION TO DISCHARGE
UNDER THE NATIONAL POLLUTANT
DISCHARGE ELIMINATION SYSTEM
In compliance with the provisions of the
Federal Water Pollution Control Act, as
amended, (33 U.S.C. 1251 et seq; the "Act"),
All owners and operators of separate
storm sewers as defined in 40 CFR
125.52 (a) (1) and all owners and op-
erators of agricultural point sources
as defined in 40 CFR 125.53 (a) (1)
are authorized to discharge from such sepa-
rate storm sewers and agricultural point
sources located in the following counties:
Pottawattamie Montgomery Page
Cass Adams Taylor
Mills Fremont
Such discharges are authorized into the fol-
lowing rivers:
Missouri Nishnabotna E. Nodaway
W. Nishnabotna Boyer Platte
E. Nishnabotna W. Nodaway and their
Middle tributaries,
Nodaway
in accordance with all conditions set forth
herein.
This permit shall become effective on
March 15, 1978.
This permit and the authorization to dis-
charge shall expire at midnight, March 15,
1981.
Signed this
. day of
402
-------
EPA PERMIT PROGRAM
A. MANAGEMENT REQUIREMENTS
1. Facilities Operation
The permittee shall at all times maintain in
good working order and operate as efficiently as
possible any treatment or control facilities or
systems installed or used by the permittee to
achieve compliance with the terms and con-
ditions of this permit.
2. Adverse Impact
The permittee shall take all reasonable
steps to minimize any adverse impact to
navigable waters resulting from non-
compliance with any conditions specified in this
permit.
3. Bypassing
Any diversion from or bypass of facilities
necessary to maintain compliance with the
terms and conditions of this permit is
prohibited, except (i) where unavoidable to
prevent loss of life or severe property damage, or
(ii) where excessive storm drainage or runoff
would damage any facilities necessary for com-
pliance with this permit.
4. Removed Substances
Solids, sludges, filter backwash, or other
pollutants removed in the course of treatment or
control of wastewaters shall be disposed of in a
manner such as to prevent any pollutant from
such materials from entering navigable waters.
B. LEGAL RESPONSIBILITIES
1. Violations of Permit
All discharges authorized herein shall be
consistent with the terms and conditions of this
permit. Any discharge from any separate storm
sewer or agricultural point source inconsistent
with the terms and conditions of this permit
shall constitute a violation of the permit.
2. Right of Entry
The permittee shall allow the head of the
State water pollution control agency, the
Regional Administrator, and/or their author-
ized representatives, upon the presentation of
credentials:
a. To enter upon the permittee's
premises where an effluent source
may be located or in which any
records are required to be kept under
the terms and conditions of this
permit; and
b. At reasonable times to have
access to and copy any records re-
quired to be kept under the terms and
conditions of this permit; to inspect
any monitoring equipment or
monitoring method required in this
permit; and to sample any dis-
charge.
3. Availability of Reports
Except for data determined to be confiden-
tial under section 308 of the Act, all reports
prepared in accordance with the terms of this
permit shall be available for public inspection at
the offices of the State water pollution control
agency and the Regional Administrator. As
required by the Act, effluent data shall not be
considered confidential. Knowingly making
any false statement on any such report may
result in the imposition of criminal penalties as
provided for in section 309 of the Act.
4. Civil and Criminal Liability
Except as provided in permit conditions on
"Bypassing" (A.3.), nothing in this permit shall
be construed to relieve the permittee from civil
or criminal penalties for noncompliance.
5. Oil and Hazardous Substance Liability
Nothing in this permit shall be construed to
preclude the institution of any legal action or
relieve the permittee from any responsibilities,
liabilities, or penalties to which the permittee is
or may be subject under section 311 of the Act.
6. State Laws
Nothing in this permit shall be construed to
preclude the institution of any legal action or
relieve the permittee from any responsibilities,
liabilities, or penalties established pursuant to
any applicable State law or regulation under
authority preserved by section 510 of the Act.
7. Property Rights
The issuance of this permit does not convey
any property rights in either real or personal
property, or any exclusive privileges, nor does it
authorize any injury to private property or any
invasion of personal rights, nor any infringe-
ment of Federal, State or local laws or
regulations.
8. Severability
The provisions of this permit are severable,
and if any provision of this permit, or the
application of any provision of this permit to
403
-------
IMPLEMENTATION
any circumstance, is held invalid, the applica-
tion of such provision to other circumstances,
and the remainder of this permit, shall not be
affected thereby.
C. PERMIT MODIFICATION
After notice and opportunity for a hearing,
this permit may be modified, suspended, or
revoked in whole or in part during its term for
cause including, but not limited to, the follow-
ing:
a. Violation by any owner or operator sub-
ject to this permit with any terms or con-
ditions of the permit;
b. A finding that circumstances
changed with respect to:
have
(i) The availability of demonstrated
technology or practices for the con-
trol or abatement of pollutants from
separate storm sewers or agricul-
tural point sources;
(ii) The availability of effluent guide-
lines for separate storm sewers or
agricultural point sources;
(iii) The availability or implementa-
tation of an approved 208 plan; or
(iv) The amounts, rates or concentra-
centrations of pollutants dis-
charged from point sources covered
by the general permit;
c. An act of God, flood, drought, materials
shortage, or other event over which the
owners or operators subject to the gen-
eral permit have no control;
d. Issuance of individual NPDES permits
for all point sources in a GPP A.
D. PERMIT REQUIREMENTS
1. Monitoring and Reporting
The results of any monitoring done by any
owner or operator of a separate storm sewer or
agricultural point source subject to this permit
shall be recorded by such owner or operator.
Any such reports shall be summarized bi-
annually and submitted to the Regional Admin-
istrator and the State at the following ad-
dresses:
Water Division
Environmental Protection Agency,
Region VII
1735 Baltimore Avenue
Kansas City, Missouri 64108
Iowa Department of Environmental
Quality
Lucas State Office Building
Des Moines, Iowa 50319
Such summary shall include:
a. The name and address of the person sub-
mitting the data;
b. The places and dates of monitoring,
c. The dates of analysis;
d. The persons who performed the analy-
sis;
e. The analytical techniques used; and
f. The results of the analysis.
2. Effluent Limitations
No owner or operator of a separate storm
sewer or agricultural point source subject to this
permit shall discharge into navigable waters:
1. Any radiological, chemical, or biological
warfare agent;
2. Any high-level radioactive waste;
3. Any pollutant in violation of State cer-
tification;
4. Any pollutant in violation of limitations
or prohibitions established under sec-
tion 307(a) of the Act; or
5. Any pollutant in violation of specific re-
quirements authorized under sections
208 and 209 of the Act.
404
-------
Interface of Water Quantity and
Quality Laws in the West
GEORGE E. RADOSEVICH
Colorado State University; Fort Collins, Colorado
ABSTRACT
Water pollution from irrigated agriculture
in the West has received major attention during
the past five years, primarily as a result of
federal and state endeavors to identify irriga-
tion return flow quality problems and to develop
a viable control strategy. The national goal of
"cleaner water" emerged as a result of the
deterioration of water quality by degraded dis-
charges from various sources.
The key to irrigated agricultural return flow
quality control is proper utilization and
management of the resource itself, and an
accepted tool in our society is the law. By legal
classification, it is divided into laws for quanti-
ty control and laws for quality control. The laws
on water quality control are recent, relatively
uniform between states and with little excep-
tion, constrain improvement of return flows
from irrigated agriculture. Unfortunately, in
most others, they cannot really be, said to
facilitate this consequence either. The laws
pertaining to water resources quantity control
and management are complex, voluminous,
inconsistent and lack uniformity among the
seventeen states of the West.
Directly affecting the management of water
in the West is the water right designed to
provide the water user with the same con-
stitutional guarantees extended to real proper-
ty. The resulting effect upon agricultural users
is that certain rigidities in the exercise and
protection of the right inhibit adaptation of
more efficient practices. Furthermore, the water
right holder is primarily concerned with his
immediate geographic area, and not with the
effects from exercising his right upon
downstream users who may be in another state
and themselves subject to different rules and
regulations.
It can only be anticipated that in the near
future greater uniformity in the water laws
among the states and in water use efficiency
criteria will occur. The recognition of the in-
terdependence of water quantity allocation and
water quality control for agricultural use is
imperative. In spite of several federal efforts to
introduce regulations to control this pollution
discharge, strong resistance and little success
summarizes the status. The state agencies,
particularly the water quality agencies, are very
much concerned with the implementability of
any program to control irrigation return flow
quality. Unless the nature of the problem is
incorporated into the law and its administra-
tion, water users cannot be expected to act in a
manner different from historical patterns.
WATER QUANTITY LAWS IN THE WEST
A System of Water Law
Water law in the United States is a
"federal" system with a delineation of jurisdic-
tion over water at the national and state govern-
ment levels. Federal government water law is
uniform and nation-wide with regional flexibili-
ty in the implementing agency regulations. But,
each of the fifty states adopted quantity control
surface and ground water laws with significant
variations. State water quality control laws are
more uniform, however, and follow a pattern set
by federal legislation.
It is commonly held that waters arising
within a state's boundaries are under the
jurisdiction of the state, unless subject to powers
reserved by the federal government. Conse-
quently, as local customs developed and states
were formed, each state adopted its own par-
ticular system of water law.
Surface water laws developed concerning
two distinct philosophies which were consistent
with the geo-climatic condition of the state. In
the humid eastern half of the country and along
the west coast, the riparian doctrine was
405
-------
IMPLEMENTATION
adopted. The more arid western half of the
country was faced with an immediate problem
of deciding how to allocate a scarce resource and
thus was compelled to develop a system of law
peculiar to arid lands. The result of trial, error
and compromise is the doctrine of prior ap-
propriation. Some states have a varied water
availability and concluded by adopting a mixed
riparian/prior appropriation system. Despite
the classification of state systems into three
groups, there is a wide variation between states
following the same doctrine as to the manner for
determining water rights, exercise of the right,
water use efficiency criteria, and system for
obtaining water rights and administering and
enforcing the law.
Ground water legislation occurred much
later in the states due in part to the lack of
knowledge of subsurface supplies and in part to
adequate surface sources. The basic principles
for use and control often followed the surface
doctrines, but again each state adopted and
modified the law to fit its perceived needs. Four
different systems of control emerged.
The system of water law and administra-
tive mechanism of these western states can be
classified as "use-oriented" — the dominant
objective being to utilize the water to produce an
economic gain, which to many meant a
livelihood and to others a profitable venture.
The following discusses the important features
TABLE 1
A Summary of Western Water Law
v f ^ I i
2 ! 3
XV^V/fiATEs uw DOCTRI nts
StatJ<& i ££•"
I-»RI2 : P.A.
2-CAL P.A.I R
3-COLO P.A.
4-10A P.A.
S-lA« P.A. .
R :
6-MMT P.A.
7-KfB P.A. •
R ;
9-«.w.. P.H.
10-H.D. P.»-
;;-OtLt P.A. •
R.:
I:-O«E P.*. «
U-S.D. P.A. •
' R.I
14-TEI : P.A. «
R :
•^•flK • P.A.
'6-MASH ; P.». .
17-WO ] PA.
Ground
Water
R.U.:
C.R.
Ownership
Puolic
People
P.A. (PuMic
P.A. jstate
P.A.
P A.
«*op
R.U.1 IPubliC
P.A.
PA.
....
P..
P.A.
A.O.
P.A.
P.A.
P.A.
! P.A.-Prio
Appropri tion
! R.-Ripari n
; A.O.-AtSO ute
; Ownershi
• R.U. -Reasonable
! C.R. -Corrective
! Rights
:A11 new water
i by P.A.
: -"Lack compre-
1 hensi ve groun
• Mter laws
i
i
Public
Publ 1C
Public
People
State
Public
Public
State
1
4 1
E.idencJ
of Hater!
Right i
Permit
Per.it
decree
&.».- ;
License |
.
Per.it
Per.1t
Per»1 1
^
6
* . , Criteria
it, t-nn- of
"'""""^Allocation
B.U. B.U.
B.I R.U. B.I R.U.
B.U. ;B.U.
B.U. |lcfs./
__j_50 a.
1 to 2
t.' /a.
B.U. 1 .iners"
/a.
B.U. 1 cfs/70a.
or 3 a; /a.
7
Preference
of Use
(Order)
1-2-3-4-5
1-2-
l-2o«er 5
8 ! 9
Oate of
Priority
0.0. A.
D.O.A.
1st ite«
G.K.-
0.0. A.
1-2" tD.O.A.
1-2-5-6-3 | D.O.A.
None D.O.A.
1-2 overS
B.U. Conditions None
1 Needs i
Per»1t B.U. B.U. I
! good agr.
practices
Pfr.it
Per.it1
Per«H
License
Per»it!
Per«1t
remit
Per»it
erent
types
'In Pin-
ing dis-
tricts.
4 over
2 1 5
-Cox>le«,
see text
D.O.A.
D.O.A.
Hone D.O.A.
Appurtency
Strict
Unlimited
None
Unlimited
B.U. 1 cfi/90». 1-2 1 D.B.U. '
5-6 ! i
B.U. 6.U.
5.U. 8.U.
B.U.
B.U.
B.U.
B.U.
IB.U.
t
B.U. -Ben-
eficial
use.
B.I R.U.-
Benefic-
ial t
Reason-
able Use
1 cfs/TOl.
or 3 a. '/a.
B.U.
Nature of
Use
Reasonably
Necessary
10
•ater
Rights
Registry
Original
Current
Original
teriied)
Current
(Lilted)
'
Original
(Li.lted)
7
Original
Original
11
Mater
Quality
In Right
Case
Case *
Statute
Case
Case
Use •
Statute
Use
Use
U»
Case
Use
Mine 0.0. A. fStrict [Current {Use
D.O.A.
Strict
1- D.O.A.
1-5-2-4-
3-7-6
1-2
None
1 cfs/70a. |l-5
'Dowstic
1 Mjn1-
2Agrlcul-
tural
(irriga-
tion)
'Mining
'Nfg. 1
.Industria
°tecrMtia
D.B.U.
D.O.A.
O.B.U.
0.0. A. iStrict
D.O.A.-
Oate of
applica-
tion
D.B.U.-
Date of
benefi-
cial
use
1
n
n
Strict-
can trans-
fer but
criteria
estab-
lished
Llnlted-
water
right for
specific
pa reel. but
transfer-
able
Originll
Original
Current
Current
Current
(CO.CU-
terlied)
Original
U,e
Use
Use
Use
Case •
Statute
Case
Original-
Initial
filing
recorded
Current-
User Hist
notify
Agency
of naM,
use. place.
etc..
transfers
unlimited
12
Forfeiture
of Rights
Syrs
3 yrs
— '
5 ,rs
3 yrs
-.'
3 yrs
5 yrs
« yrs * 1 yr
after notice
3 yrs
7 yrs
5 vrs
3 yrs
10 yrs
5 yrs
Syrs
5 yrs
'All state
recognize
loss by
abandon-
•ent
'10 yrs Is
evidence
of aban-
donment
13
Drainage
Rules
C.E. I
C.L.'
R.3.
C.L.
fled)
C.L.
C.L.
C.E.
C.E.
C.L.
C.L.
R.D.
K.O.
C.L.
C.L.
C.L.
C.E.
C.E.
Undecldei
C.E.-
C.L.-
Civil La
R.D.-
Reason-
able Di
charge
•C.E.-
flood
waters
C.L.-
natural
flows
14
Basin
of
Orialn
res
res
j^State
1-ARIZ
2-CAL
3-COLO
I-IDA
5-KAK
6-HOKT
res 7 -NEB
B-NEV
9-ri.M.
lei
10-N.D.
11 -OKU
12-CM
res
w
-
13-S.O.
14-TEX
1S-UTAH
16-MASH
17-KTO
406
-------
QUANTITY AND QUALITY LAWS
and status of the law of water allocation and use
in the western states, summarized by Table 1.
SURFACE WATER LAWS
As previously stated, the seventeen western
states have adopted one or both of the basic
water law systems found in the United States.
The system adopted by every western state is
the doctrine of prior appropriation, with those
states on the western seaboard and from North
Dakota to Texas also employing the riparian
doctrine to lands adjacent to watercourses.
There is a definite trend to eliminate the
riparian doctrine as demand on surface waters
increase. For all practical purposes, most of the
states with both doctrines have relegated the
riparian system of surface water control to an
insignificant role. However, a brief explanation
of the doctrine's salient points will enable the
reader to recognize the attitudes and drawbacks
in efforts to control the quality of irrigation
return flows.
Riparian Doctrine
Those states in the West applying the
riparian doctrine follow the American Rule of
Reasonable Use. Under this rule, riparian land-
owners can divert a reasonable amount of water
with respect to all other riparians on the stream;
and nonriparian lands may, under certain con-
ditions, make a reasonable use of remaining
waters.
Nature of the Riparian Water Right
Waters in states following the riparian
doctrines are a public resource, held in trust for
use by the people of the state. Thus, a land-
owner whose land borders a stream does not
have an ownership right to the waters of the
stream, but rather has a fundamental right by
virtue of his land location to a reasonable use of
the water and to be free from unreasonable uses
of others that cause him harm (Rancho Santa
Margarita v. Vail, 11 Cal. 2d 501, 81 p. 2d 533,
1938). He is essentially a correlative co-user with
all other riparians on the water source and as
between riparian uses, priority of use does not
establish priority of right in times of decreased
flows (Pabst v. Finland, 192 Cal. 124, 211 P.ll,
1922). Consequently, his right to the use of water
is not a right for a fixed quantity of flow or
volume, but rather is dependent largely upon
the extent of development that has taken place.
Manner of Allocation
Fundamental to the riparian law is the
location of land on a water source. Although
this requirement has been relaxed in many
eastern states to permit use of water on non-
riparian lands — as between riparians and
nonriparians — water will first be allocated to
the riparian landowner. Among the western
states, California remains the one state in
which the riparian right doctrine has major
significance. In California, a riparian right can
only be established upon riparian land. And, if a
portion of the land to which a riparian right
attached is severed from the original parcel, and
itself does not have access to the watercourse or
the riparian right is not specifically reserved for
the portion, then the right is lost and cannot be
restored.
More important to the allocation of water
under this doctrine is the requirement of
reasonableness in use, since the right does not
consist of a definite quantity of water.
The reasonable use requirement limits the
use of water to that quantity reasonably re-
quired for a beneficial use and prohibits waste
or unreasonable use, or unmeasurable methods
of use, or diversions (Const, of Cal., Art, XIV,
Section 3).
Other rules on allocation and exercise of
riparian rights include restrictions against
transfer of riparian rights to nonriparian lands,
water cannot be stored for later use under a
riparian right (Moore v. California-Oregon
Power Co., 22 Cal. 2d 725, 140 P.2d 798, 1943),
nor can foreign waters be claimed under a
riparian right (Crane v. Steuinson, 5 Cal. 2d 387,
54 P.2d 1100, 1936).
Preference of User
Under either the natural flow or reasonable
use theory, there is a preference for the "natural
wants" over all other uses. The "natural wants"
include household and limited livestock needs
and have generally placed domestic-called uses
in a preferred position. As between other uses —
agricultural, industrial, recreational, etc. —
there is no clear preference, but rather the courts
have looked upon the reasonableness of use to
determine conflicts between these uses.
Nonuse and Misuse of Rights
Riparian water law does not require a land-
owner to use the water in order to maintain the
right in good standing. Unless the right to use
407
-------
IMPLEMENTATION
water from an adjacent water source has been
sold or transferred to other lands or uses, the
right will continue as long as the land and water
is continuous. Abandonment of water right is
nonexistent under riparian law. There is a
possibility, however, that a riparian who does
not object to the open and notorious use by
another, through prescription, may have his
right reduced or lost.
Misuse of the right may result in a restric-
tion on use and/or judgment for damages to
those adversely affected. Parties injured
through the misuse must assert their claim in
court.
Statutory Modifications
There have been a number of significant
recent changes in the water law of the riparian
states primarily brought about by the inability
of existing water supplies to meet the expanding
demands on one side, and a recognition of public
interest in water resources on the other.
The changes can be summarized into two
major components: 1) establishment of a permit
system to allocate water among certain users;
and 2) creation of administrative machinery to
assess water resources through the permit
system.
Among the western states, the modifica-
tions to the original doctrine are strongly in-
fluenced by the simultaneous application of the
doctrine of prior appropriation, increased
demands on surface supplies for in-basin as well
as out-of-basin use stimulated in part by large-
scale reclamation projects, and heavy reliance
upon ground waters in some states, i.e., Califor-
nia, Nebraska, Oklahoma, and Texas. In all of
the dual-doctrine states except California, all
new claims to the use of surface waters must
comply with the statutory requirements of the
prior appropriation doctrine. Since the 1967
Water Rights Adjudication Act in Texas, all
surface water rights are now required to be filed,
approved and administered by the Texas Water
Rights Commission. All riparian right
claimants are to file their claim with the Com-
mission, or the right would be extinguished
(T.C.A., Section 5.301 to 5.341). If the Commis-
sion desires, it may also adjudicate these claims.
Similarly, in Washington, under the 1967
Water Right Claims Act, riparian right
claimants must have their claims adjudicated
in order to protect their interests against holders
of appropriation rights (W.R. C. §90.54.010). In
California, if a suit involving the determination
of water rights is brought in the superior court,
the court can refer the case to the State Water
Resources Control Board for a determination of
all water rights in the stream system (C.W.C.
§2000 and 2001; C. W. C. §2500 to 2900 sets out the
procedure).
Regarding Irrigation Return Flow
The riparian doctrine has as an inherent
component, the requirement that a riparian user
make a reasonable use of the water and his right
shall not be impaired in quantity or quality by
the unreasonable use of another riparian. Thus,
in theory, if the upstream riparian's return
flows were degrading the quality of the water
used by a downstream riparian, the latter has a
basis for judicial redress. It must be pointed out
that under the riparian system, an ad-
ministrative structure for allocation and regula-
tion of water does not exist (except in those
western states where riparian claimants are
required to file their claim or obtain a permit)
and, thus, the injured party must rely upon his
remedies in the courts through a private law-
suit.
The evolution of this doctrine was a for-
tunate event for it proved as useful for
agriculture as it was for mining. As mining
became more competitive, many miners and
newcomers to the area began farming. The
doctrine protected the first settler to use water
on his land. Later settlers had to respect the
prior ownership of land and the amount of water
which the prior settler was using. Hence, the
clique "first in time, first in right" symbolized
most everything the early water users were con-
cerned about.
The system that emerged was simple and
direct. Although there are many variations
between the appropriation doctrine states, a
number of key principles exist to establish
commonality, if not relative uniformity, among
the states. These principles are:
1. There had to be a division from a natural
stream or body of water. This has been
relaxed in most western states during the
last decade to allow in-stream use for recrea-
tion and fish and wildlife protection.
2. The water must be applied to a beneficial
use. Initially, this was defined in con-
stitutions and/or statutes to be domestic,
municipal, stock watering, irrigation, and
408
-------
QUANTITY AND QUALITY LAWS
certain industrial and power uses. Some
state laws, like Wyoming, reflect the
economic influence of one sector over
another, i.e., the railroad uses were
preferred to agricultural uses. In most of the
western states, however, the rural represen-
tation insured agriculture a high position as
a beneficial user. Beneficial use also
referred to the nature of use, but will be
discussed later.
3. When these two acts were completed, a
water right was created. This right entitled
the holder to continued use so long as the
use was beneficial. The attributes of this
right are discussed later.
4. Every water right acquires a priority
date such that priority of right and not
equality of right is the basis for distributing
water.
The Water Right
The entire system of prior appropriation is
based upon and evolves around the allocation of
water under the concept of the water right.
Simply put, this doctrine creates the right of
private use of a public resource under certain
conditions, which use has been declared to be a
public use. The right does not automatically
exist by virtue of the presence of water upon,
flowing through, or under land. In all western
states, these waters (some exceptions) are
declared to be the property of the public, people,
or state (see Column 3, Table 1). Regardless of
whether the state or the public (people) own the
water, the courts have held the state as a trustee
to the public for the proper allocation and
distribution of water and granting and protect-
ing the right to use the water so allocated.
Wyoming law states, for example:
A water right is a right to use the water of
the state, when such use has been acquired
by the beneficial application of water under
the laws of the state relating thereto, and in
conformity with the rules and regulations
dependent thereon (W.S.A. §41-2).
In Colorado, the Supreme Court very early
in the state's history announced a rule that can
be found in every statutory or judicial law of the
other appropriation doctrine states. The famous
Coffin v. Lefthand and Ditch Co. was decided in
1882 and held:
. . . water in the various streams thus ac-
quires a value unknown in moister climates.
Instead of being a mere incident in the soil,
it rises when appropriated to the dignity of a
distinct usufructory estate or right of
property . . . the right to property in this
country by priority of appropriation
thereto, we think it is and has always been
the duty of the national and state
governments to protect (6 Colo. 443).
The right so acquired has two legal
characteristics. First, the right itself is a real
property right. It is an exclusive right, which
like other property interests, can be defined, is
valuable and can be sold, transferred, mort-
gaged, or bequeathed. But the right differs from
the right that attaches to land or chattels, for it
is only a right to use the resource. Thus, it is
called a usufructory right (see Coffin case above
and O.S.A. 82 §105.2 as examples of judicial and
legislative holdings).
The second characteristic is that since it is
only a usufructory right and can only be exer-
cised when the water authorized for diversion
under the right is available and can be put to
beneficial use, there is no absolute ownership in
the corpus of the water prior to diversion. This
water is still a public resource, and if the right
holder cannot put it to beneficial use, he must
allow it to flow past his point of diversion to
other appropriations. However, if he can ap-
propriately use the water, that water which is
diverted into his delivery system is his personal
property until it returns back to the stream or
escapes his control.
The water right under the appropriation
doctrine consists of several defining elements
that give it value, dependability and security to
the holder. The right:
• exists to a definite source of supply, e.g.,
specific river, lake, or ground water ac-
quifer;
• is for a fixed and stated maximum quanti-
ty divertable;
• has a definite point of diversion to which
conditions are to be maintained as of the
time the appropriation took place;
• specifies the type of use for the diverted
water;
• identifies the place of use (for application
in the case of irrigation);
• implies the annual time of use based upon
type and place of use; and
• assures the holder of at least an implied
protection to the maintenance of water
quality necessary to carry out the pur-
409
-------
IMPLEMENTATION
poses for which the water was
propriated.
ap-
As was previously stated, one of the key
principles to the prior appropriation doctrine is
the "priority of right" that is granted a user over
subsequent appropriation. It is most often the
priority date, dependability of flow in stream
and location of point of diversion that gives a
water its value.
In most states, the priority date is the date
the application for a water right is received by
the state agency. Generally, an application
must be filed with pertinent information
relative to the user, use and source of supply. If
the application is approved, the water right will
normally have the priority of the date of applica-
tion (Column 8, Table 1). If the use requires
construction of diversion, storage and delivery
works over a period of years, the right, if the
application is approved and notice to proceed
given, will still retain the date of application
when the water actually is put to use, through
operation of the "doctrine of relation back." If,
however, the applicant does not construct the
works within the time period acceptable to both
parties, and the delay is unexcusable, the right
may have a priority beginning on the date the
water is put to use.
Several systems were developed by the
states to allocate water and provide evidence of
water rights, including posting a notice at the
point of diversion and filing a record with the
county clerk. The predominant approach now is
the permit system (Column 4, Table 1). An
application is filed with the appropriate state
agency who then takes the procedural steps of
evaluating and determining its disposition. If
approved, a permit is issued which may contain
conditions of use. If denied, the applicant may
appeal the administrative decision to the court.
In some cases, the finalized water right may be
called a license or certificate.
A few states have different classes of per-
mits which greatly enhance their ability to
allocate and regulate the use of water among
competing interests. In Texas, there are eight
classes of permits:
• regular permit — year-round perpetual
right;
• seasonal permit — portion of calendar
year (irrigation season and perpetual);
• temporary permit — short-lived specific
use, no longer than 3 years;
• term permit — fixed number of years and
expires;
• contractual permit — authorizes an ap-
propriator to contract the use of his water
to another for a term;
• permit under Section 5.141 — authorizes
impoundment on nonnavigable stream
on permittees' own property of less than
200 acre-feet and use for any specified
purpose;
• storage permit — storage of water for
project;
• emergency permit — allows emergency
appropriation for not more than 30 days
for public health, safety and welfare
(T.W.R.C. Rule 129.02.05.001-.008).
Oklahoma has two broad categories of
permanent and non-permanent. The former is
subdivided into regular and seasonal, while the
latter is divided into temporary and term
(O.W.R.B., Rule 350). All states grant direct flow
and storage permits. Colorado is the only state
that does not have public representation in the
water right allocation process. Water rights
applications are submitted to the appropriate
Water Court (one water court in each of the
seven water divisions). Through a statutorily
defined process, the Court and its referee act
upon the application by giving notice through
newspaper publication to water users in the
area and holding hearings so protests to the
application can be heard. If the application is
acceptable, the Water Court will issue a decree
as evidence of the water right. A conditional
decree may be issued if the work to complete the
diversion and put the water to beneficial use is
to take place over a period of time.
One of the frustrating problems to water
administrators and planners and often costly to
water users under the current high demand for
water and increase in sales and transfer is the
recordation of water rights. The majority of
states have a registry of the original issued
water rights (Column 10, Table 1) which iden-
tifies the original appropriation, point of diver-
sion, source of supply, amount divertable in
continuous flow, or total volume terms, and type
and place of use. In all states, any change or
transfer in place or type of use and point of
diversion must be approved by the state agency.
This is primarily to protect other appropriators
who may be adversely affected by the transfer if
conditions of the stream and return flow are not
accounted for. But, few states actually maintain
a current registry of water rights that reflect
410
-------
QUANTITY AND QUALITY LAWS
current ownership. These statutes, or adopted
regulations requiring notice to the agency of all
ownership changes. Oklahoma and Texas have
a computer card type report form that right
holders must submit annually with pertinent
data concerning ownership, amount of water
diverted, and what uses were made of the
resource. In some cases, failure to notify the
state agency of ownership and other changes is
prima facie evidence of non-use and could lead
to forfeiture or abandonment. The burden of
notice is usually placed upon the current owner.
Even in many of these states, however, the
current listing is not complete.
In the past, the purely engineering concerns
of source, diversion point and type of use may
have been sufficient to distribute water, that is,
water could only be diverted out of a fixed and
definite headgate. But, with problems of in-
creased demand on the available supplies and
the present need to resolve such water quality
problems as those stemming from irrigation
return flows, greater efficiencies in use must be
achieved. This can, and in many cases are,
being brought about by: 1) transfer of right to
use water to higher value uses; 2) some volun-
tary action by water users to improve the
diversion and delivery structures and locations;
and 3) tighter administration of the conditions
of water use granted a water right holder under
the law. To effectively administer these laws
(beneficial use and non-waste provisions), carry
out water planning and development, and even
take full advantage of transfer characteristics
inherent in the property right in water, accurate
records of current ownership and use are re-
quired.
One last feature of a water right for irriga-
tion use must be mentioned. In all states of the
West, when a water right is granted for irriga-
tion use, the right becomes appurtenant to the
land(s) described in the permit, that is, it
attaches to those lands and cannot be used
elsewhere without approval of the state agency
(or Water Court in Colorado). Many variations
as to the extent of attachment exist, however.
Most states use what, for lack of a better term,
we call limited attachment. This only requires
approval of the state agency when proper
measures and adjustments are made in the
transfer to prevent impairment of other users'
rights. In a few states like Colorado, the water
right does not necessarily transfer with the sale
of land (James v. Barker, 99 Colo. 551, 64 P2d
598, 1937).
In a minority of other states, the ap-
purtenancy rule is strict. That is, the transfer
can be approved if for uses other than irrigation
(A.R.S., §45-172), or if it becomes impracticable
to use the water economically or beneficially on
the original lands (in re Determination of
Relative Rights to Use of Waters of Pantona
Creek, 45 Ariz. 156, 41 p2d 288, 1935). Wyom-
ing applies the appurtenancy rule only to direct
flow rights, but as noted by Meyers and Tarlock
(1971, p. 528), some fourteen statutory excep-
tions riddle the principle. The other states with
strict provisions on transfer are Oklahoma,
Oregon, Nebraska, Nevada and South Dakota.
The reason for many states tying water rights to
land at the turn of the century was to prevent
some of the fraudulent land and water sale
practices that had gone on under earlier federal
settlement schemes in the West.
Basis and Criteria of Allocation and Use
Beneficial Use
The cornerstone of water allocation under
western water law as it has evolved is that
beneficial use is the basis and measure of the
right to use water. This is often the extent of
definition found in the majority of western
states water law (e.g., A.R.S. §45-101; N.R.S.
§533.035; S.D.L.§46-l-8; W.S.A. §41-2). Thus, in
order to use water, it must be taken for a
beneficial purpose. This has evolved into the
position that not only must water be used for a
beneficial purpose, but beneficial use is the limit
of the right (Farmers Highline Canal and
Reservoir Co. v. Golden, 129 Colo. 575, 272 P2d
629, 1954). Usually the term "beneficial use" is
not defined per se but is decided on a case-by-
case method. It has two aspects, referred to
above, that complicate the concept even more.
Water is allocated to a beneficial use, so for that
reason, many statutory provisions list types of
uses recognized as being beneficial. Among the
uses recognized as beneficial are: irrigation,
domestic, power production, municipal, in-
dustrial, recreation and minimum flows for
aquatic life. This short list is not meant to be
comprehensive but, rather only to illustrate the
spectrum recognized.
The other aspect is the use of the water itself
must be beneficial and carried out in a beneficial
manner. On this point, several states have
elaborated the definition to provide directly to
the administration of the laws. In California,
the use of water is subject to the constitutional
requirements that such use:
411
-------
IMPLEMENTATION
. .. shall be limited to such water as shall be
reasonably required for the beneficial use to
be served, and such right does not and shall
not extend to the waste, or unreasonable use
or unreasonable diversion, of water (Cal.
Const. Art. XIV, Sec. 3).
Texas applies this broad but more defined
approach to beneficial use also. It requires that
no more water be allocated and used than that
amount "economically necessary for the pur-
pose authorized when reasonable intelligence
and reasonable diligence are used in applying
the water to that purpose" (T.C.A., Sec. 5.002).
New Mexico law directs itself to irrigation
specifically by placing a limitation on all rights
by instructing the State Engineer not to allow
the diversion of more water for irrigation than
can be used consistent with good agricultural
practices to produce the most effective use of
water (N.M., Sec. 75-5-17).
Washington similarly addresses the use of
water by agriculture. Its laws provide that an
appropriator will be provided that quantity of
water reasonably necessary to irrigate his land,
but this irrigation is to be accomplished by the
most economical method of artificial irrigation
according to the methods employed in the
vicinity where the land is situated (R.C.W., Sec.
90.03.040). The most economical method is to be
determined by the court.
The concept of reasonableness is playing an
increasingly more important role in appropria-
tion states. For example, it may no longer be
reasonable to irrigate a crop by flooding when
another method, readily available, will produce
crops as well or better and simultaneously save
some of the water. Thus, even though the use —
irrigation — is beneficial, the method of applica-
tion is not reasonable.
Duty of Water
In addition to the requirement that water
will be allocated to a user for a beneficial use,
most states have adopted criteria to be followed
in allocating water to agriculture. This criteria
is commonly referred to as the statutory duty of
water. To quote from the Supreme Court of
Colorado:
It is that measure of water, which by
careful management and use, without
wasteage, is reasonably required to be
applied to any given tract of land for such
period of time as may be adequate to
produce therefrom a maximum amount of
such crops as ordinarily are grown
thereon. (Farmers Highline Canal and
Reservoir Co. u. Golden, 129 Colo. 575 at
584, 270 p2d 629, 1954).
The majority of states incorporate this into
their determination of the amount to be granted
the water right applicant. But several states
have quantified the duty. Little uniformity
exists indicating the different conditions found
in the states. Idaho, Wyoming and North
Dakota allow 1 cfs per 50, 70 and 80 acres,
respectively. South Dakota and Nebraska also
allow 1 cfs per 70 acres but no more than 3 acre-
feet per acre. Montana allows 1 miners inch per
acre and Kansas varies between 1 to 2 acre-feet
per acre, depending upon the circumstances
(Colume 6, Table 1).
Provisions in Nevada are particularly im-
portant to the subject of this report. In Nevada,
the State Engineer is to consider the duties of
water established by court decrees or by ex-
perimentation in the area where the water is to
be used (N.R.S., §533.070). He is also instructed
to consider the growing season, type of culture,
and reasonable transportation losses. This flex-
ibility allows the State Engineer to be precise in
allocating water.
Waste
The corollary to beneficial use of water is
the duty not to commit waste of water. This
requirement is expressed by statute or court
decision in all the western states. However, it
must be stated that as what constitutes
beneficial is difficult to define except on a case-
by-case approach, it is equally difficult to
categorically state what amounts to waste. In
the analysis of case law by Hutchins, Elles and
DeBraal, they note that an appropriator need
not take extraordinary precautions to prevent
waste if it is a reasonable use of the water
according to the customs of the community
(citing: Joerger v. Pacific Gas and Electric Co.,
207 Cal. 8, 273 P.1017, 1929), so long as the
custom does not involve unnecessary waste of
water (Hutchins, Vol. 1, p. 498).
Many states have statutory provisions like
that found in Arizona (A.R.S. §45-109) and
Nevada (N.R.S. §533.460) which prohibit waste
and charge the party so committing waste to the
detriment of another to be guilty of a mis-
demeanor.
412
-------
QUANTITY AND QUALITY LAWS
Preferences and Priority to Use
These are two concepts in the appropriation
doctrine that are often intermingled and con-
fused in use of the terms. Priority of right has
been described above as the date of a right that
distinguishes it from all other rights for pur-
poses of distribution of available water supplies
in the source from which the appropriation is
attached. This enables the senior right holder to
demand and receive his allocation at the time
the senior places a call for his water before the
junior is entitled to exercise his. During low
flows or scarcity, diversions are shut-off in
inverse order. Hence, the value of an early water
right.
Preferences, however, do not address a date
of appropriation, but rather the type of use that
receives preferential treatment by laws. In
many states, certain types of uses are placed in
an order establishing their preferred position
(see Colume 7, Table 1). This has two primary
purposes. The first, in the allocation of water
between competing, simultaneous applicants,
the allocating agency can use the criteria that
as between the applicants for supply of water
insufficient to meet the needs of all, the appli-
cant requesting water for a preferred use can be
granted his right over those requesting water
for a lower or non-preferred use.
The second aspect of the preference system
is that during periods of scarcity, the preferred
use has a right to exercise the condemnation
power to obtain water from non-preferred uses.
In all states but Texas, compensation must be
paid. Texas has an "absolute preference"
system which provides municipalities with the
right to take water without compensation. As a
practical matter, this is rarely if ever utilized.
GROUND WATER CONTROL SYSTEMS
Four Doctrines
Ground water resources are beginning to
play a major role in agricultural, municipal and
industrial water use. Approximately one-fifth of
the water withdrawn in the country comes from
this source.
Laws controlling the extraction and use of
ground water have become as complex as sur-
face water doctrines. Basically, however, the
states apply one of four doctrines — absolute
ownership, reasonable use, prior appropriation,
or correlative rights (see Corker, 1971). Colume 2
of Table 1 identifies the doctrines adopted by
each state. A thorough treatment of ground
water laws up to 1970 can be found in Hutchins
(1974).
Absolute Ownership
The doctrine of absolute ownership had its
origin in the United Kingdom with the 1843
decision of Acton v. Blundell (152 Eng. Rep.
1223.1843). Simply stated, the doctrine holds
that a landowner can withdraw any water from
beneath his land without liability to his
neighbors resulting from such action.
Among the western states, only Texas has
retained this rule. In Texas, the rule applies only
to percolating waters and not subterranean
streams or tributary stream underflows. But the
presumption is that all ground water is per-
colating, thus allowing a landowner to take and
use or sell all the water he can capture from
beneath his land (Texas Water Plan, 1968). In
areas where a defined acquifer exists, land-
owners can organize into a ground water conser-
vation district and establish location, depth,
discharge, and use rules (T.C.A., §7880).
Reasonable Use
Due to the extreme position of ground water
use without liability as proclaimed under the
absolute ownership doctrine, many states
began modifying the laws into what has become
known as the "American Rule of Reasonable
Use." This change is synonymous to the
modifications in the surface riparian doctrine.
The rule of this doctrine is: since the rights of
adjacent landowners is similar and their enjoy-
ment in the use of ground waters is dependent
upon the action of other overlying landowners,
each landowner is restricted to a reasonable
exercise of his own rights and reasonable use of
his own property, in view of the similar rights of
others (Meeker v. E. Orange, 11 N. J.L. 623, 74
A.379, 1909).
Among the two western states that have
retained the reasonable use doctrine (Arizona
and Nebraska), Arizona holds that one land-
owner can withdraw ground water, even though
some harm is dealt his neighbor, if heis making
a reasonable use of the water on land from under
which the water is taken (Bristor v. Cheatham,
77 Ariz. 227, 255, P.2d 173, 1953), while
Nebraska will allow out-of-basin diversions for
municipal use if no damage is done to overlying
landowners in the area where the water is
413
-------
IMPLEMENTATION
extracted (In re. Metropolitan Utilities District
of Omaha, 179 Neb. 783, 140 N.W.2d 626,1966).
It appeared that Arizona was leaning toward
the Nebraska rule in the Jarvis cases (Jaruis v.
State Land Department, 456 P.2d 385,1964 and
Jarvis v. State Land Department, 479 P.2d 269,
1970), with little limitation. Then, in early 1977,
the court held in Farmers Investment Co. v.
Bettery, et al. (558 P.2d 14) that water cannot be
transported out of a basin if other overlying
landowners are injured by the withdrawals. It is
also important to note that neither state has yet
adopted a comprehensive ground water code.
This doctrine leaves much speculation as to
what is "reasonable use," but on the other hand,
affords some measure of protection to property
now existing and greater justification for the
attempt to make new developments (Katz v.
Walkinshau; 141 Cal. 116, 74 P.766, 1903).
Correlative Rights
The doctrine of correlative rights in ground
water originated in California and is a further
refinement to the reasonable use concept.
Several states originally adopted this doctrine,
then changed to another rule (i.e., Utah), but
now only California, among the western states,
applies this rule. The doctrine holds that among
landowners overlying an underground water
supply, each landowner can make a reasonable
use of that supply so long as the source is
sufficient. But when the supply becomes insuf-
ficient due to the drought or draw-down effect,
each landowner is entitled to water in propor-
tion to the percent of his land in relation to all
other lands overlying the underground waters
(Katz v. Walkinshaw, supra). The net effect is to
provide great flexibility of ground water use in
an effort to maximize the resources, while
providing equitable allocation when shortages
occur.
Prior appropriation
Most of the western states found little
reason to differentiate their systems of law for
surface waters and ground waters. As a conse-
quence, they adopted ground water statutes of a
similar philosophy stating that this source
should be allowed maximum development with
recognition and protection given prior users.
This does not imply, however, that surface
water law was automatically applicable to
ground water. In fact, several states enacted
laws to control ground waters as late as mid-
1950's and 1960's.
The rule provides that ground water is
subject to appropriation for beneficial use
providing the intended user complies with the
statutory requirements, obtaining a permit or
license as the case may be. The administrative
official must determine if unappropirated
ground water exists and what adverse effects
would occur from approving the application.
In most states, the law allows the state
water official, upon a determination that a
particular ground water basin needs close
management of withdrawals, to classify the
area as a critical or designated ground water
basin (see Colorado or New Mexico laws). When
this occurs, the users are placed under direct
control for the protection of the aquifer and
vested rights.
DRAINAGE
The rights of landowners to protect their
property from diffused surface waters is only
incidental to the irrigation return flow quality
problem, because most agriculturalists are con-
cerned with how they can capture and use their
source of supply. However, a brief explanation
of the rules is considered useful in light of the
growing awareness and ability of meeting
different plant requirements for moisture and
the need to adopt conservation practices which
prevent erosion and lead to sediment control.
The three rules are: common enemy rule,
civil law rule and reasonable use rule. Under the
common enemy rule, diffused surface waters are
considered an "enemy" of the landowner and he
can construct dikes, drains or other necessary
steps to protect his land from the damaging
effects of the surface waters (Tillinger v. Frisbie,
138 Mont. 60,353 P.2d 645,1960; Gillespie^Land
and Irrigation Co. v. Gonzalez, 93 Ariz. 152,372
P.2d 135, 1963, as regards to flood waters).
The civil law rule holds just the opposite. It
is "essentially a rule of natural drainage hold-
ing that lower land is burdened with a natural
easement of drainage in favor of higher land"
(Colorado u. Brannon Land and Gravel Co., 534
P.2d 652, Colo. App., 1975). However, the higher
land cannot increase the burden of the lower
land, and the latter can, if necessary, take
protective measures to prevent damage to his
land from unreasonable discharges (Harper v.
Johannesen, 84 Ida. 278, 371 P.2d 842, 1962).
The third rule is in between the two
previously discussed rules and basically holds
there can be a reasonable interference with the
414
-------
QUANTITY AND QUALITY LAWS
natural flow of water by either party to protect
their property (Iven v. Roder, 431 P.2d 321,
Okla., 1967). See Colume 13, Table 6-1 for the
rules as adopted by the western states.
INCENTIVES AND DISINCENTIVES TO
EFFICIENT USE
Basin of Origin
Several states have adopted rules to protect
the water needs of landowners and populations
within a watershed from future shortages
caused by out-of-basin diversions and uses.
California adopted a county of origin rule in
1931 (C.W.C. §10505) and a Watershed Protec-
tion Act in 1933 (C.W.C. §11460 to 11463). These
provisions give a general protection to in-
habitants within the county and basin to
reclaim water in the future if needed from non-
county or basin uses. The impetus to the
Watershed Protection Act was the California
Water Plan which has proceeded in spite of the
reservations for future use within the basin.
Colorado and Nebraska also have limita-
tion on exportation of water from a natural
basin (C.R.S. §37-45-1 18(IV); and N.R.S. §46-
206 and 46-265). In the case of Colorado, the law
requires projects that use water out of the
natural basin shall not impair present and
prospective uses of water for irrigation and
other beneficial consumptive use purposes
within the natural basin, nor increase the cost
at the expense of users within the basin.
Texas and Oklahoma have taken a
different approach to out-of-basin diversions. In
Texas, water for transfer out of the basin is
restricted to those waters which will be surplus
to the reasonably foreseeable water supply
requirements within the basin of origin for the
next 50-year period (T.C.A. §8280-9). In 1972,
Oklahoma enacted the Stream Water Use Act
(O.S.A. 82§105). It protects the current water
users within a stream system from damaging
out-of-basin transfers by requiring the Texas
Water Resources Board to review the water
needs in the area of origin every five years
Rights and Duties to Return Flows
Return flows are an important source of
water in the arid western states, and as such are
considered by water users and administrators
as a commodity that should not be dealt with
lightly. As water rights to natural flows were
granted, the streams began to be augmented
from seepage, tailwater runoff and percolation.
Subsequently, other water rights were granted
based upon this source of water and junior in
time. As such, the courts have generally held
that junior appropriators can rely on these
return flows and protect their rights in this
source (Boulder v. Boulder and Left Hand Ditch
Company, 557 P.2d 1182, Colo., 1977; East
Bench Irrigation Co. v. Desert Irrigation Co., 2
Utah 2dl70, 271, P.2d 449, 1954).
Also as a general rule, irrigation districts
can recapture return flows before they leave
boundaries and reuse these waters (Ide v. Unit-
ed States, 263 U.S. 497, 1972). But, this rule
normally does not extend to individuals as re-
turn flows are considered by the courts to be the
non-consumptive uses of water that returns to
the stream from the proper and beneficial
application of water.
This leaves a fine line between waste water
and return flow. In Binning v. Miller, the
Wyoming Supreme Court stated a rule common
to many jurisdictions that one can recapture his
waste water on his property and reuse it thereon
(55 Wyo. 451, 102 P.2d 54,1940). Other jurisdic-
tions have gone on to say that a downstream
user can appropriate waste water, but cannot
compel the person committing the waste to
continue to discharge nor prevent him from
adopting improved practices that eliminate the
waste (Wedgworth v. Wedgworth, 20 Ariz. 518,
181 P.952, 1919).
Loss of Water Rights
Water rights under appropriation doctrine
can be lost through misuse. There are four ways
in which this may occur.
The first is abandonment. Should a right
holder not use his right for a statutory period of
time and intend not to use it, his water right may
be lost. The important element is intent, but this
may be difficult for the state or party claiming
abandonment to show. The state or another user
must bring action against the user and prove
both elements. Abandonment rules exist in all
the western states.
The second is forfeiture. This is a statutory
remedy to nonuse and only requires a showing
of nonuse of all or a part of the right for a
specified period of time. Automatically after the
statutory term (usually three or five years), the
water reverts to the public for appropriation by
another (see Colume 12, Table 1). There are
415
-------
IMPLEMENTATION
variations in state laws as to notice and
procedures for carrying out the forfeiture
provisions.
Adverse possession is the third method, and
this occurs when another openly and notorious-
ly uses the water right of a person, and that
person does nothing about it. If this continues
for a specific period, the former can claim the
right as his own. The practice is not looked upon
with great favor by the courts, however.
Condemnation is the fourth major type.
This occurs when a preferred user or public
entity exercises the right of condemnation.
Normally the only real issue is the amount of
compensation. Colorado, however, recently
passed a law requiring municipalities con-
demning agricultural water rights to show the
necessity for taking such action (C.R.S. §138-6-
201 to 216; see Radosevich and Sabey, 1976).
Water Quality in Water rights
There is at least an implied right of water
quality in the water right under the doctrine of
prior appropriation. In only two states (Califor-
nia and Washington) do the statutes make
specific provision of the element of water quali-
ty in a right, such that a user can make the same
demands on an agency to protect his interest in
water quality as he can his interest in the
quantity he is entitled to according to the
priority of right.
In all the western states, the courts have
recognized the common law doctrines of
nuisance and trepass as applying to the protec-
tion of property interests in water. Most of the
cases surround the discharge of mine tailings
into a stream and subsequent injury to
downstream agricultural users. An example of
the nature of cases, as they pertain to
agriculture, is the Idaho decision Ravndal v.
Northfork Placers, 60 Idaho 305, 91 P.2d 368
(1939). In that case, plaintiffs ditches and crops
were injured by the hydraulic mining process
employed by the defendant mining company.
The Idaho Supreme Court affirmed the damage
judgment awarded by the district court and held
that:
Numerous authorities announce the doc-
trine that while a prior use of the water of a
stream for mining purposes necessarily
contaminates it to some extent, such con-
tamination or deterioration of the quality
of the water cannot be carried to such a
degree as to inflict substantial injury upon
another user of the waters of said stream
(Ravndal v. Northfolk Placers, 60 Idaho
305, 91 P.2d 368, 1939).
In spite of this result, one law review article
has stated that private nuisance actions have
"provided virtually no incentive to the offenders
to reduce their harmful discharges into the
waterways" (Wood, 1971).
The conclusion from an analysis of all the
cases studied to date on the water quality
protection issue is that a water user must pursue
his own remedy in court if he wishes to protect
the quality of water he is receiving.
WATER QUALITY LAWS IN THE WEST
Water pollution control has actually had a
long history in the United States, dating back to
the local laws for sewage control in the colonial
times. By 1948, when the first significant piece
of federal legislation was enacted, most states
throughout the country had adopted some form
of pollution control legislation. State and local
health departments were responsible for carry-
ing out the law. There was little activity between
the federal and state governments up to that
point, for no federal program existed. The
federal attitude was that pollution control was
primarily a state responsibility. In fact, this
view prevailed until 1965 when the federal
government declared it would initiate control
where states failed to adopt standards and
enforcement procedures (Water Quality Control
Act of 1965, P.L. 89-234).
At present, all of the western states have
established water pollution control laws that
are highly uniform. The legislative enactments
and promulgations of rules that have taken
place in the past four years have been primarily
to qualify for grants under the Water Pollution
Control Act as amneded and to adopt a permit
discharge system comparable to the federal
NPDES permit program.
The pattern of present state water quality
control legislation is very similar to the com-
ponents and declarations found in P. L. 92-500.
As a general statement, the state acts consist of
five components:
1. Water quality policy
2. Criteria for pollution control.
a. classifications of waters
b. water quality standards
416
-------
QUANTITY AND QUALITY LAWS
c. effluent discharge standards
3. Control activities.
a. permit system
b. construction grants and programs
c. public participation in planning and
setting of standards
4. Sanctions and enforcement measures.
Water Quality Control Policy
The foundation of water quality control
within each state is found in policy declarations
introducing the legislation, or in pronounce-
ments by the agency charged with ad-
ministering the law. Fifteen of the states have
policy statements which are very similar.
Typical of the water policies found in the
other states is the statutory provision contained
in Oregon's law:
(1) To conserve the waters of the state;
(2) To protect, maintain and improve the
quality of the waters of the state for public
water supplies, for the propagation of
wildlife, fish and aquatic life and for
domestic, agricultural, industrial, munici-
pal, recreational and other legitimate
beneficial uses;
(3) To provide that no waste be discharged
into any waters of this state without first
receiving the necessary treatment or other
corrective action to protect the legitimate
beneficial uses of such waters;
(4) To provide for the prevention^ abate-
ment and control of new or existing water
pollution; and
(5) To cooperate with other agencies of the
state, agencies of other states and the
Federal Government in carrying out these
objectives (O.R.S. §468.710, emphasis
added).
As a general proposition, the policy declara-
tion also contains specific reference to the
protection of beneficial uses of water as provid-
ed in the water allocation laws, non-
degradation of waters that exceed current water
quality standards and target dates identical or
close to those found in federal legislation. The
specific policy of Idaho's Board of Health and
Welfare towards water quality is "to provide for
an orderly and economically feasible com-
prehensive water pollution control program,
which program shall be administered to con-
serve the waters of the State for all legitimate
beneficial uses, including uses for domestic
purposes, agriculture, industry, recreation and
fish and wildlife propagation" (Water Quality
Standards and Wastewater Treatment Re-
quirements, Idaho Dept. of Health and Welfare,
June 1973).
Other provisions found in state laws and
policy statements formulated by the agencies of
importance are the salinity policy of Arizona
(A.W.Q.S. Rules R9-21-103), emphasis upon
soil erosion control and programs of Kansas
(K.S.A. §2-1902) and Arizona (A.R.S.
§45.2001), and recognition by Colorado that the
problem of water pollution in Colorado is closely
related to the problem of water pollution in
adjoining states (C.R.S. §25-8-102(1)). Califor-
nia is concerned over ground water quality
(C.W.C. §12922) and waste water reclamation
(C. W. C. §13510). Many states have policies for
water quality control in their environmental
acts.
One of the most important declarations as
far as irrigation return flow is concerned is that
found in Montana's recent act. The law states
that it is the public policy of the state of
Montana to:
(a) conserve water by protecting, main-
taining and improving the quality and
potability of water for public water
supplies, wildlife, fish and aquatic life,
agriculture, industry, recreation and other
beneficial uses;
(b) provide a comprehensive program for
the prevention, abatement and control of
water pollution (R.C.M. §69-4801(1)).
The section further states that "it is not
necessary that wastes be treated to a purer
condition than the natural condition of the
receiving stream as long as the minimum treat-
ment requirements are met." "Natural" refers to
conditions or material present from runoff or
percolation over which man has no control "or
from developed land where all reasonable land,
soil and water conservation practices have been
applied. Conditions resulting from the reason-
able operation of dams at the effective date of
this Act are natural" (Ibid., §2, emphasis add-
ed). As a general rule, however, agricultural
return flows are included in the definition of
"pollutant" in the state laws.
Criteria for Water Quality Control
Three district criteria for control of water
quality are usually found in state laws. They are
classification of waters, water quality stand-
417
-------
IMPLEMENTATION
ards and effluent discharge standards. A few
states also have set pretreatment standards
(i.e., South Dakota) and performance and toxic
effluent standards (i.e. Nebraska and Wyom-
ing).
In most every state, the water quality
control agency is directed to classify the state
waters and develop and maintain a comprehen-
sive program for prevention, control and abate-
ment of water pollution. Except for New Mexico
and Oregon, all the states have developed a
classification system based upon the beneficial
uses of water. These systems range from 2 to 14
classifications, but in general, include domestic
water supply, full body contact recreation, fish,
wildlife and other aquatic life protection,
agricultural and industrial, and aesthetic uses.
Often, special provisions in the classifications
are made for perennial and non-perennial
streams, streams designated for no future dis-
charge, as return flow streams or streams
requiring advanced waste discharge.
In California, the regional water quality
boards are to develop a classification scheme
according to the particular regional needs. New
Mexico has not developed a classification
scheme but instead has chosen to regulate water
quality according to specific geographical areas
with water quality standards adopted for the
various river basins. Oregon has not adopted a
classification system for water quality control.
In conjunction with classification of state
waters, all states have developed water quality
standards according to bio-chemical composi-
tion that is to be protected or limits not to exceed.
The typical approach is to develop general and
specific standards with various parameters.
The parameters for general standards vary in
number from 6 to 15.
Indicative of the parameters for specific
standards are those found in Idaho:
A. Coliform Concentration
B. Dissolved Oxygen
C. Hydrogen Ion Concentration
D. Temperature
E. Turbidity
F. Total Dissolved Solids
(Water Quality Standards and Waste-
water Treatment Requirements, Idaho
Dept. of Health and Welfare, §VIII, June
1973).
Several states have specific reference to
irrigated agriculture in their water quality
standards. Arizona requires their Water Quality
Control Council to consider, among fifteen other
items:
16. In formulating any applicable stand-
ard pertaining to agricutural irrigation
and drainage waters, the Council shall be
guided by the principle that such waters
are put to beneficial use within the state for
the irrigtion of lands or become return
flows to the waters of the state and subse-
quently reused, and that such standards
shall not diminish the water available for
such uses nor deprive the state of such
water(A.R.S. §36. 1857 A).
The above statute further commends the Coun-
cil to:
1. Not require any present or future ap-
propriator or user of water to divert, cease
diverting, exchange, cease exchanging,
store, cease storing, or release any water
for the purpose of controlling pollution in
the waters of the state.
2. Exclude from water quality standards
wholly private waters closed to all public
uses and not discharging into or polluting
any other waters of the state (A.R.S.
§36.1857 B).
South Dakota has adopted specific water
quality parameters for irrigation waters. The
specific areas covered by these regulations
include specific levels of coliform and fecal
coliform organism, total dissolved solids, con-
ductivity and sodium absorption ratios (S.D.
Surface Water Quality Stats. Ch. 34:04:02:43,
1975). The criteria for irrigation waters is
applicable only from May 15 to September 30,
Also, the criteria for coliform and fecal
organisms is applicable only to water used to
irrigate root crops.
All western states have also adopted waste
discharge standards, but have very little
reference to irrigation return flows.
Control Activities
There are basically three control activities
employed by the western states water quality
agencies to carry out the policies and criteria
discussed above. They are: 1) permit system for
effluent discharges; 2) construction grants and
programs; and 3) planning and development
with public participation in setting various
standards for water quality control. These tools
are not to be confused with the legal tools for
418
-------
QUANTITY AND QUALITY LAWS
enforcing the law against violators, such as
cease and desist orders and judgments.
The only activity to be discussed here is the
permit system. The construction grants and
programs do not directly apply to irrigated
agriculture; at least there has been little if any
use at the state level. The planning activities are
in conjunction with the federally developed and
promoted 208 planning and other planning
activities required by federal law.
A permit is required from the state water
quality agency to lawfully discharge or emit
wastes into the surface waters of the state. Each
state has particular requirements and con-
ditions for obtaining a permit. The conditions
normally consist of four, but may vary from
state to state. They are:
1. Effluent limitations;
2. Schedule of compliance and interim
dates;
3. Special conditions; and
4. A monitoring program if appropriate.
The usual activities requiring a permit
include:
1. Discharge of any wastes into the waters
of the state from any industrial or commercial
establishment or activity or any disposal
system;
2. Construct, install, modify or operate any
disposal system or part thereof or any extension
or addition thereto;
3. Increase in volume or strength of any
wastes in excess of the permissive discharges
specified under an existing permit;
4. Construct, install, operate or conduct
any industrial, commercial, or other establish-
ment or activity or any extension or modifica-
tion thereof or addition thereto, the operation or
conduct of which would cause an increase in the
discharge of wastes into the waters of the state
or which would otherwise alter the physical,
chemical or biological properties of any waters
of the state in any manner not already lawfully
authorized;
5. Construct or use any new outlet for the
discharge of any wastes into the waters of the
state (O.R.S. §468.740).
Most permits are for a term of five years
with renewal if the permittee is still in com-
pliance with the conditions set out in the permit
and if other conditions have remained un-
changed. Permits can be modified or terminated
if: a) the permittee violates the conditions of the
permit; b) the permit was obtained by false or
misrepresentation or failure to fully disclose all
relevant facts; or c) physical conditions change
in the receiving waters that require either a
temporary or permanent reduction or elimina-
tion of the permitted discharge.
The permit programs of ten western states
have received EPA designation to administer
the entire NPDES process, that is, to issue new
permits, reissue expired permits, monitor permit
compliances, and enforce violations (Water
Quality, 1976, p. V-19). Table 2 identifies the
states accepted by EPA to implement the
NPDES program, and the reasons the remain-
ing states have not received designations.
Sanctions and Enforcement Measures
The intent of federal and state water pollu-
tion control laws is to prevent and abate dis-
charges which are harmful to the public health
and welfare and the beneficial uses to which
water is put. In so doing, the various laws have
described or granted administrative authority
to determine what activities must be stopped
because of immediate harm, what discharge
limits will be tolerated or excluded from control
under certain conditions, such as discharges
into non-perennial streams, and what dis-
charges require a permit and compliance to an
abatement and limitation schedule. Many of
these points have been discussed above.
However, beyond a doubt, the sanctions
which allow certain discharges under permitted
or excluded conditions do not constitute the bais
for a right to discharge. To dispel any notion
that a discharge requirement is in the nature of
a permit to discharge, the Porter-Cologne Act
specifically provides that discharges do not
receive the status of vested rights, but are in the
nature of a privilege (C.W.C. §13267(c) ).
One of the major problems and complaints
of pre-1972 state and federal water pollution
control legislation was the lack and inability of
enforcement of the law by the regulatory agen-
cies. The problem was two-fold. First, the legal
mechanism was lacking or too weak. A polluter
could "beat" the agency in court. Second, where
the legal procedure was adequate, the penalties
were often so small or minor that it "paid" to
pollute rather than incur the cost of pollution
control. Also, political and economic interests
often got in the way of effective enforcement.
419
-------
IMPLEMENTATION
TABLE 2
Status of NPDES Delegation
in the 17 Western States*
EPA Regions
Reason for EPA
Accepted Non-Acceptance
Region VI
New Mexico
Oklahoma
Texas
Region VII
Kansas
Nebraska
Region VIII
Colorado
Montana
North Dakota
South Dakota
Utah
Wyoming
Region IX
Arizona
California
Nevada
Region X
Idaho
Oregon
Washington
Legal
Legal
Legal
6-28-74
6-12-74
3-27-75
6-10-74
6-13-75
1-30-75
5-14-73
9-19-75
9-26-73
11-14-73
Resources
Legal
Legal
Legal
•Status as of February 1,1977, Update to Table V-5,
p. V-21, Staff Report to the National Commissionon
Water Quality, April 1976.
As a result of P.L. 92-500, the enforcement
provisions are mandatory, not discretionary,
and the legal tools are more precise. In all the
state laws and regulations examined, the state
water quality control agencies now have a host
of approaches with varying degrees of impact.
The range begins with notice to the polluter and
an opportunity to voluntarily comply with the
law. This may be done by submitting a detailed
plan and time schedule to come into compliance.
If the discharger fails to respond to this
administrative notification or directive, the
agencies have four other direct actions that can
be taken. They are:
1. Issuance of a regulating order, common-
ly referred to as a cease and desist order.
2. Initiation of court action to obtain a
restraining order and/or injunction to prevent
further discharges.
3. Levying of civil penalties.
4. Initiation of criminal proceedings by the
county or district prosecutor or state attorney
general.
Several states also permit the state agency
to expend monies to clean up the waste (Califor-
nia) or restock fish killed (Nebraska) and charge
the violator with reasonable costs.
REGULATIONS AFFECTING
IRRIGATION RETURN FLOW QUALITY
There has been very little legislative or
agency rule-making activity by the seventeen
western states in the specific area of irrigation
return flow quality control up to this time. Only
three states have adopted regulations impact-
ing the quality of these return flows. They are
California, Idaho and Nebraska, with Idaho's
Department of Health and Welfare adopting
regulations requiring a permit to discharge
from large irrigation projects only within the
past few months (Environment Reporter, Vol. 7,
No. 42, p. 1610, Feb. 18,1977). These regulations
have been recently repealed by the Idaho State
Legislature.
North Dakota has adopted regulations per-
taining specifically to salinity but not to irriga-
tion return flows generally (Standards of Sur-
face Water Quality State of North Dakota,
Section VII, North Dakota State Department of
Health, 1973). South Dakota has promultated
water quality standards that include reference
to irrigation return flows. Oklahoma has
adopted regulations for return flow streams and
mixing zones, but not specifically for the source
of pollutant from irrigated agriculture.
In most states, irrigation return flows are
ignored, exempted, or subject to qualifications
in the application of control regulations. The
Water Quality Control Council in Arizona may
make a determination that water quality stand-
ards do not apply when the entire flow in a
watercourse that would otherwise be dry is
effluent from agricultural irrigation return flow
(Arizona Water Quality Standards, R9-21-
210A). Furthermore, water quality standards do
not apply to the collection, return, or drainage of
agricultural irrigation return flows, excess or
tailwaters to canals, laterals or other manmade
irrigation water delivery facilities within an
irrigation system, or chemical maintenance of
irrigation facilities within an irrigation system
where agriculture is the only designated prim-
ary beneficial use, or physical or mechanical
maintenance of irrigation facilities within an
420
-------
QUANTITY AND QUALITY LAWS
irrigation system. But, the law of Arizona does
make it unlawful for any person ". . . to dis-
charge any agricultural, irrigation or drainage
waters into waters of the state and thereby
reduce the quality of such waters below the
water quality standards established therefore
by the Council in violation of an order issued
pursuant to Section 36.1854" (A.R.S., Section
36.1858).
California applies a unique approach to
handling the irrigation return flow problem.
First of all, irrigation return flow is not treated
as a point or non-point source. It is simply
looked at as a water quality degradation
problem. The state is divided into various
regions. Each region has the authority to con-
trol the water quantity and water quality within
its region. The State Water Resources Council
Board must adopt guidelines which set forth
minimum standards for the disposal of wastes
(Calif. Regs., Section 2555), and each Regional
Board develops its own system of regulations to
comply with the NPDES.
Monitoring programs must determine the
quality and quantity of supply waters entering
agricultural areas and return flows from
agricultural areas entering streams and/or
receiving waters. A monitoring program may
be carried out under permits which have been
issued to agricultural areas for either two or five-
year periods.
Planning programs are initiated by the
Regional Administrator or the State Board
upon the advice of the Agricultural Water
Quality Advisory Committee. Through the
planning process, agricultural discharges
which can be controlled by a permit are to be
identified and permits issued. These programs
must develop the best management practices for
agriculture which will reduce pollutant loads to
surface waters, and develop policies or
guidelines for individual farm operations and
basin-wide management problems. The Califor-
nia approach to meeting EPA's requirements of
issuing permits for irrigation return flows as a
point source is to identify all entities within an
irrigation system or subsystem from which
discharges can be identified and monitored, and
have these entities cosign the permit. This
satisfies EPA and enables the state to decen-
tralize monitoring and management to the local
entities while retaining control and enforce-
ment if necessary.
In Colorado, a permit is not required for any
flow or return flow of irrigation water unless a
federal act or regulation so requires (C.R.S.,
Section 25-8-506).
Idaho did accept the designation of irriga-
tion return flows as point source discharges, but
has not been delegated by EPA to carry out the
federal NPDES program.
Although Montana has no specific regula-
tions pertaining to irrigation return flow, the
policy of the state strongly implies that action
will be taken to protect receiving streams from
runoff or percolation where unreasonable land,
soil and water conservation practices have been
applied.
A number of states have retained regula-
tions adopted prior to the decision oiN.R.D. C. v.
Train (1975) which were based upon the original
regulations promulgated by EPA (Idaho,
Nebraska, Nevada). These regulations exempt
irrigation return flows except if the return flows
are from lands of more than 3,000 acres or 3,000
non-contiguous acres discharging into the same
drainage systems. This exemption was struck
down in the N.R.D.C. case and the states are
merely waiting for new federal regulations
before amending or trying to carry out the
existing ones.
Summary
As complex as the water law systems are in
the West, any situation taxing the ability of
current water supplies to meet increasing
demands will require changes in the laws. The
constraints to improved management lie
primarily in the water rights laws, and as a
general proposition, it is concluded that these
laws:
• lack specificity in definition, inhibiting
implementation of concepts such as
beneficial use and waste prevention
• provide an ad hoc approach to resolving
key water problems through the courts
• insure that the property right concept in
water will continue to be the most single
critical roadblock to curative solutions
• recognize customary standards of the
community as the criteria for beneficial
use
• fail to require the adoption of a recorda-
tion system to assist in implementation of
legally enforceable concepts and pro-
grams for improved water management
421
-------
IMPLEMENTATION
• fail to integrate water quantity and quali-
ty control, conceptually and administra-
tively
* provide disincentives for recapturing or
properly using return flows, or salvaged
and excess waters by prohibiting trans-
fers to other lands or users.
To systematically attack the water manage-
ment problem and particularly the return flow
quality problem, changes in state laws must
include:
• efforts to achieve uniformity within state
laws and administration
• efforts to achieve workable uniformity in
the laws between states
• development of criteria for efficiency in
water allocation and use according to key
principles of:
a) beneficial use
b) waste
c) duty of water
and apply this criteria to all uses, private
and public
• recognize and promote the "public trust"
of water agencies and the public duty in
the user for use of public resources
• incorporate, specifically, the element of
water quality in a water right
• shift to a term permit or periodic evalua-
tion of effective water use
• adopt a program of influent control which
includes outflow analysis with criteria to
determine the degree of improvement
relative to needs and opportunity costs.
422
-------
An Influent Control Approach
to Irrigation Return Flow
Quality Management
GEORGE E. RADOSEVICH and GAYLORD V. SKOGERBOE
Department of Economics;
and Agricultural and Chemical Engineering Department, respectively.
Colorado State University, Fort Collins Colorado
ABSTRACT
The Influent Control Approach is based
upon the assumption that improved water
management plus improved agricultural prac-
tices will significantly contribute to improved
water quality, and the conclusion that best
management practices plus best agricultural
practices will provide irrigation return flow
quality control, which in turn will contribute
significantly to the national goal of cleaner
water through improved water quality.
The Influent Control Approach consists of
eight specific components: (1) designate areas
for irrigation return flow quality management
and designate the responsible area entity;
(2) develop standards and criteria for beneficial
use in designated areas; (3) introduce incentives
to use water more efficiently; (4) include the
element of water quality in new or transferred
and changed water rights; (5) adopt and enforce
a reporting and recording system for water
rights; (6) recognize reasonable degradation
from agricultural water use; (7) adopt an
Agricultrual Practices Act; and (8) promote
close cooperation or integration of state water
agencies and related functions.
PHILOSOPHY AND CRITERIA FOR
EFFECTIVE CONTROL
The Need
Water quality control from irrigation return
flows has perhaps caused the greatest degree of
disenchantment among state and federal per-
sonnel charged with carrying out water quality
programs under P.L. 92-500 than any other
category of pollution sources. Since the time
that first regulations for irrigation return flows
were promulgated in 1973, there has been strong
and distinct differences of opinion among the
various agencies dealing with water at both
state and federal levels of government, and
within their ranks as well. Many western states
have called a stop to their programs until EPA
adopts what the states consider a workable
approach. The legal gyrations of the past four
years have caused them to undertake minimum
activity so as not to directly violate any par-
ticular law or regulation. Not one western state
has completely and enthusiastically embraced
a program of including irrigation return flows
as a "point source" and thus subjecting all
irrigation to the NPDES program.
Part of the problem for the disenchantment
stems from the physical difficulties in dealing
with the irrigation return flow quality problem
where it does exist. Equally important is the
lack of a philosophical foundation and thrust to
resolving a problem of this immense complexi-
ty, as well as inherent resistance to control.
Viewing the problem from both the position
of water users and agency personnel charged
with controlling the problem, this study
attempted to be both logical and pragmatic in
formulating an implementable and sustaining
approach to irrigation return flow quality con-
trol. The philosophy and criteria which follow
are building blocks to the proposed Influent
Control Approach (ICA) set out in this paper. It
is the authors' opinion that awareness, not con-
currence, is essential to an understanding and
acceptance of a program.
The complex situation faced by public of-
ficials and water users is illustrated in Figure 1.
To achieve the highly desirable goals of cleaner
water called for by the Federal Water Pollution
423
-------
IMPLEMENTATION
IRRIGATED AGRICULTURE
WATER MANAGEMENT PRACTICES
AGRICULTURAL PRACTICES
(Jtt of Agr. Chemtcols
"X T
CONVEYANCE
Dehvtry SyiHm
X. \CROPLANU
Apphcotton
linoolto/*
Procteet
SOURCES OF
RETURN fLOW
Seepage Loises
ISub&urtoce
Return flows)
Agronom*
Practice
A
/
Deep Percolation
[Subutrfoct Return
Flows)
1DCGRAD N3 EFFECTS 1
UPON*ATE« QUALITY,
Salmrty Sa
tftltv
1 Nitrates
Irrigation
Returr Flows
c 1 Application 1
t Methods and I
\
\
Tatlwoter Runoff
(Surface Return
snJL
Phosphates
Crop Residue
Bacteria
"T
Figure 1. Impact of water management and agricul-
tural practices upon irrigation return flow quality.
Control Act of 1972 (P. L. 92-500), it is necessary
to realistically assess the specific nature and
problems caused by surface and subsurface
return flows from irrigated agriculture. Present
water management and agricultural practices
create irrigation return flows from the con-
veyance system and cropland that may have a
degrading effect upon water quality in the
receiving waters.
Philosophy
The proposed philosophy upon which to
formulate a successful program for control of
irrigation return flow quality consists of four
interlocking propositions. First, the ultimate
goal achieved by the federal and state agencies
is improved water quality by way of improved
water management, with this particular study
focusing upon the return flow characteristics
and problems of irrigated agriculture. In this
context, improved water management means
water quality enhancement through reductions
in tailwater runoff, seepage losses, and deep
percolation losses (i.e., surface and subsurface
return flows, which are point and non-point
sources, respectively).
Second, the program should promote social
and economic well being through cooperative
action. Every effort should be made to prevent
polarization between state and federal agencies
and local water user groups. An approach must
be designated to stimulate a tripartite relation-
ship between the water users and the state and
federal agencies, while still maintaining the
identity of each.
Third, only attack the problem after it has
been realistically identified.
Fourth, voluntary compliance is more de-
sirable than forced or involuntary compliance
in implementing a management or control ap-
proach.
Criteria
Based upon this philosophy, the following
criteria must be met for an implementable and
sustaining program in irrigation return flow
quality control:
1. The approach should result in improved
water management practices specifically
and improved agricultural practices gener-
ally. In the context of this paper, improved
agricultural practices can be achieved
through joint progress in two areas:
(a) proper land use and (b) proper applica-
tion of agricultural chemicals. As stated
previously, water management includes
affects upon the quantity and quality of
surface and subsurface return flows.
2. The approach should prevent social
disruption and polarization of water users
(e.g., individuals, irrigation companies and
water districts) and state and federal agen-
cies. Maintaining separate identities is
necessary, maintaining opposition is not.
In most western states, water users have
gone on record opposing past federal and
state efforts to control irrigation return flow
quality.
3. The approach should be palatable to
water users. Some of the most often ex-
pressed concerns of the water users are:
What is the problem and how significant is
it; how am I involved; if I agree to a permit,
what does that mean to me now, as well as
potential control in the future (i.e., what
rights and liberties am I giving up by
agreeing to a nebulous program); what is it
going to cost; what benefits will be a-
chieved; and, who pays for nature's dis-
charges?
4. The approach must be feasible, flexible
and allow for state agency discretion in
working with local entities. State agency
concerns which must be met include: a
424
-------
INFLUENT CONTROL APPROACH
determination of the significance of the
problem; ability to implement the program
because of manpower limitations; iden-
tification of pollution sources; failure of
past programs to include subsuface flows;
credibility with water users; conflicts with
other state agencies (e.g., water quality
administration and agriculture agencies);
and, the ultimate impact of such a program
upon water quality, flow regime of streams
and water users.
5. The approach should improve the
credibility of state and federal agencies.
Presently, most agricultural water users
feel alienated against the federal and state
water quality control agencies.
6. The approach should utilize existing
institutions (e.g., laws and organizations)
and accepted concepts (e.g., designation of
problem areas such as critical ground water
basins, beneficial use, and duty of water) as
much as possible.
In order to implement any approach, both
the water users and the agencies must have
knowledge of the resources they are "manag-
ing." The irrigation system is too highly in-
tegrated and complex to be subjected to a frigid
unilateral control program.
AN INFLUENT CONTROL APPROACH
Irrigation Return Flow Characteristics
Irrigation return flows have few character-
istics which allow them to be viewed as a typical
point source pollution discharge problem. Most
pollution contained in irrigation return flow
occurs as a natural process of diverting and
using water for a legally appropriated beneficial
use. But, the pollution often occurs beyond the
boundaries of the control of the water user.
Thus, the first distinct feature is problem iden-
tification.
Because the degraded return flows may be
either surface or subsurface, and most often
diffused rather than collected into a discrete
conveyance system from the contribution
source, the second feature is contributor
identification. From a technological stand-
point, contributor identification requires an
evaluation of the sources of pollution; whereas,
from a legal viewpoint, contributor identifica-
tion requires a determination of who is
polluting.
Allocation of water under western states
laws requires that it be for a benefical use. Water
allocated for irrigation is generally allocated to
specific lands with the quantity based upon a
fixed state-wide duty of water standard (e.g.,
1 cfs/ 70 acres in Wyoming, 2 acre-feet per acre
in Nebraska).
The right to use water is a property right to
the holder, which is to be exercised according to
priority with other users and availability of
flow. Because there is no prorationing during
shortages as under the riparian doctrine, and
because this property right is one of rapidly
increasing value in the West, the inherent
incentive to the holder is to protect that right by
diverting the full entitlement without regard to
the fine line between beneficial use and waste.
Thus, a third factor emerges, i.e., law and
customary diversion preempt equal weight to
external diseconomies.
The fourth feature gives rise to the proposed
influent control approach described hereafter.
This is the correlation of input to output. In
addition to the possibility of controlling degrad-
ed return flows at their discharge, or effluent
control, an impact can be exerted upon the
quality of water discharged from irrigation uses
by changes at the input stages, i.e., delivery and
application of water. Because of the elusive
nature of irrigation return flows, the traditional
approach of effluent control is not considered
adequate nor feasible in light of a more
economic, simple and functional alternative
within the control of the water user and in-
fluence of state water officials. That alternative
is to control the influent. In the case of irrigation
return flow, this includes water user discretion
on the delivery and application of water (use of
well known technologies such as canal lining to
curb excessive seepage losses during con-
veyance, or improved irrigation methods and
practices to reduce deep percolation losses and
tailwater runoff), proper land use to retard or
prevent erosion of soil and subsequent sediment
pollution in tailwater, and proper application of
fertilizers and pesticides.
These four features plus: a) the system of
water quantity and quality administration at
the state level, and b) the peculiarities of our
judicial system, rules of evidence and burden of
proof, require an approach which is both
'In a few states, Nevada for example, discretion to de-
termine quantity is given the State Engineer.
425
-------
IMPLEMENTATION
preventative and curative but within the
parameters of a known demonstrated problem.
Because end-of-pipe treatment is neither
technically satisfactory, nor economically
justifiable, the Influent Control Approach (1C A)
is designated to get at the cause, not the
consequence, of the problem and promote alter-
native solutions within the control and capabili-
ty of irrigation water users generally. Where
voluntary action to alleviate the known problem
is not taken, existing laws for water quantity
use and control of discharges by permit can be
exercised.
Theme of Influent Control Approach
The underlying theme of the ICA can be
summarized as:
1. Proof before control
2. Proceed cautiously and positively.
3. Stimulate voluntary action based on
demonstrated need to change.
4. Maintain relationship between agen-
cies and water users.
5. Create or maintain credibility.
Assumptions
The Influent Control Approach is premised
upon ten assumptions. They are:
1. Achieving the goals of P. L. 92-500, the
Federal Water Pollution Control Act of
1972, and policies of federal and state laws
to improve the use of our national resources
is highly desirable;
2. The concept of property rights in water
and other constitutional guarantees will be
maintained;
3. The legal procedures of the judiciary
and agencies will be utilized;
4. Improved agricultural practices and im-
proved water management will result in
improved water quality;
5. Irrigation return flow problems and
appropriate solutions to these problems are
site specific;
6. Water users (farmers) will respond
when it has been demonstrated that there is
a problem to which they are contributing;
7. Technical and legal solutions to iden-
tify problems must be appropriate and
viable (technically sound, economically
feasible, legally implementable, and social-
ly acceptable);
8. Many irrigators will respond on a
voluntary compliance basis;
9. Those users who do not respond will feel
a local social pressure as a result of being
"out-of-tune" with the newly evolved
customs of the community; and
10. Regardless of approach, there will be
some users who will not respond or will
resist change, thereby requiring some
mechanism for enforcement.
Influent Control Approach Components
As was alluded to above, the distinction of
this approach to irrigation return flow quality
control is to indirectly correct the unreason-
ably degraded discharges caused by irrigated
agriculture by directly affecting the influent or
input to the system. This approach is based
upon the assumption that improved
agricultural practices (IAP) and improved
water management (IWM) will contribute to
improved water quality (IWQ). In specific con-
text of this paper, it is further conduced that best
management practices plus best agricultural
practices will provide improved irrigation
return flow quality control (IRFQC), which in
turn yields improved water quality.
BMP + BAP - IRFQC —>IWQ
Because the nature of agricultural pollution
from irrigation is too complex to rely upon end-
of-pipe treatment, the cause of the problem is
examined in its broader context, i.e., present
water management and agricultural practices,
with the emphasis upon only those elements of
agricultural practices relating to or having an
affect upon return flows. The concept of best
management practices is currently employed by
EPA and the states and refers here to im-
provements in local ^ water management. Best
agricultural practices is used here to include
proper land use and proper application of
agricultural chemicals.
To reiterate, every effort in formulating this
approach was made to decentralize the act of
control to the lowest common denominator —
the irrigator — because of his ability to volun-
tarily and directly impact the quality of return
-Local is used to distinguish water quality control
within the irrigation system or subsystem from
state and national control.
426
-------
INFLUENT CONTROL APPROACH
flows and because of a recognition that
agriculturalists traditionally are independent
people who prefer to be actors, not pawns. The
components of the ICA thus provide: (1) the
design and direction to irrigation return flow
quality control; (2) the opportunity for volun-
tary compliance by water users in problem
areas; and (3) the means to effectively assert
involuntary compliance upon those con-
tributing to the problem who refuse to adopt
better practices by the responsible government
agencies.
The Influent Control Approach is designed
to improve water quality by reducing excessive
seepage, tailwater runoff and deep percolation,
reducing sediment in return flows through ero-
sion control and reducing chemical concen-
trations in return flows through licensing
and/or control resulting from over-application
of pesticides and fertilizers. Since irrigation
return flow quality problems differ from one
irrigation system to another, the approach
provides the latitude to introduce change and
control according to the nature of the problem,
without requiring unnecessary compliance by
those irrigators outside problem areas.
Based upon this background, the Influent
Control Approach (Figure 2) is designed with
eight specific components. The first six com-
ponents pertain to improving local water
management, with components 1 and 2 having
application in the problem area only and com-
ponents 3 to 6 having state-wide jurisdiction.
Component 7 pertains to land use and chemical
applications affecting water quality and has
state-wide jurisdiction. Component 8 focuses
upon the functional ability of agencies to carry
out the program.
The Influent Control Approach consists of
the following components to be carried out by
the states:
1. Designate Areas for Irrigation Return Flow
Quality Management and the Responsible Area
Entity.
Action — Based upon monitoring and
analysis for indentifying significant irriga-
tion return flow problem areas within the
state, the state agency will: a) designate the
boundaries of the problem area, which may
be the boundaries of an irrigation system or
subsystem or watershed; b) designate an
entity, i.e., legally constituted body
representing water users within the area, to
undertake responsibility for working with
the water users, collecting data and dis-
seminating information. The area entity
may be a newly formed organization, an
existing organization, i.e., irrigation dis-
trict, that assumes the program or a federa-
tion of numerous existing organizations in
the designated areas; and c) insure that the
entity is carrying out the best management
practices developed for this area, as well as
best agricultural practices.
Rationale— Applying the designated
area approach to controlling unreasonable
degradation from irrigation return flow
enables the state to focus only upon those
areas within its boundaries where a pro-
blem has been identified. Thus, not all
irrigators are depicted as shifting exter-
nalities (i.e., costs from use of degraded
water) upon the public and downstream
users; all irrigators regardless of how well
they manage their water and land resources
are not subjected to the time consuming
procedures and implications of a permit
system. Consequently, water users are not
collectively polarized against the efforts of
state and federal agencies to reduce and
prevent water quality degradation. This
first component is the corner-stone of the
influent control program because it draws
attention only to problem areas without the
guilt insinuations or accusations that
farmers are so sensitive to.
From a practical point of view, the area
entity, which may be represented by an
existing irrigation or water-related
organization, would utilize a representative
board of commissioners that would be
responsible for carrying out monitoring,
discussing ways to alleviate unreasonable
degradation by irrigation return flows to
receiving waters with water users in the
designated area, and encouraging volun-
ary improvement of agricultural practices
jy those users identified as contributing to
the area's problem. For those users (or
entities representing those users within an
area) who refuse or fail to respond as
recommended, the area entity would notify
the state water quality control agency of the
specific non-compliance, and the state
would then proceed under existing federal
and state law to initiate control and enforce-
ment that is, under the general provisions of
427
-------
IMPLEMENTATION
A SOLUTION.
An Influent Control Approach (ICA)
ASSUMPTION:
Improved Agricultural Practices + Improved Water Managements Improved Water Quality
( IAP + IWM = IWQ )
CONCLUSION:
Best Management Practices + Best Agricultural Practices = Irrigation Return Flow Quality Control
( BMP+ BAP = IRFQC -IWQ )
DEFINITIONS:
BMP - Improved Local Water Management ( ILWM )
BAP = Proper Land Use (PLU)and Proper Application of Agricultural
Chemicals ( PA AC )
ACTION
I
INFLUENT CONTROL APPROACH (ICA)
BEST MANAGEMENT
PRACTICES (BMP)
Improved Local Water
Management (ILWM )
BEST AGRICULTURAL
PRACTICES (BAP).
Proper Land Proper Application of
Use (PLU) Agricultural Chemicals (PAAC)
i
[ PROBLEM AREA
COMPONENTS
STATEWIDE
STATEWIDE
I. Designate area
and area entity.
2. Develop standards
and criteria for
beneficial use.
3. Introduce incentives
for ILWM.
4. Add water quality to
water rights.
5. Add reporting and
recording for water
rights.
6. Recognize reasonable
degradation from irri-
gation return flows.
7. AdoptAgricultural Practices
Act. ,
Sediment
and erosion
Licensing and control
over application of
agricultural chemicals:
fertilizers and pesticides
J
8. Promote close cooperation or integration of state water
agencies anJ other related functions.
(RodOMvich a SKogerboe, 1977)
Figure 2. Influent Control Approach to irrigation return flow quality management.
428
-------
INFLUENT CONTROL APPROACH
the water pollution laws, prohibiting dis-
charges of pollutants and violation of
stream standards. In these cases, the iden-
tified and non-complying irrigator dis-
charge can be required to obtain a permit
under the regular NPDES program. The
area entity is thus responsible for assisting
in managing the agricultural practices
within the designated area, but control and
necessary enforcement are appropriately
left to the state.
Precedent — The concept of designated
areas for resources control and manage-
ment is well recognized and applied in
many western states for ground water and
municipal water supply. In most instances,
water users participate on commissions or
boards having jurisdiction over the man-
agement area.
2. Develop Standards and Criteria for Benefi-
cial Use in Designated Areas.
Action — For each designated area
within a state, the water quantity and
quality agencies will collaborate to arrive at
standards and criteria for beneficial use of
water. Such standards and criteria will not
constitute an impairment or taking of water
rights, but rather be the technical limits of
water delivery and application under the
climatic, soil and othe agronomic con-
ditions of the area. These conditions for
water use would be tantamount to a
calculated "duty of water" ^ for the area in
light of the return flow quality problems.
Rationale — Under the water laws of each
western state, water is allocated under the
concept of beneficial use. This term is
generally not defined, it is normally nebu-
lous, but does in general meet the needs for
both allocating and distributing water
within the state. However, in certain areas
due to soil characteristics and water use
practices, irrigation return flow quality
problems do develop which are directly
related to the delivery and application of
water. Within designated areas, it is neces-
sary to develop specific standards and
3Duty of water means the quantity of water neces-
sary for effective use for the purpose to which it is
put under the particular circumstances of soil condi-
tions, method of conveyance, topography, climate
and crop grown (Water and Water Rights, Vol. 5,
Sec. 408.2, 1972).
criteria for beneficial use that still comply
with the concept of a water right under state
law.
Precedent — Under Nevada water law,
the State Engineer has discretion to deter-
mine the duty of water based upon the site
specific characteristics of the irrigated area,
the type of use to which the water will be
put, and the impact upon surrounding
water users. Utah is another state that has
applied the variable duty of water concept
in determining the appropriate amount of
water to allocate under a water right.
However, the state agency in Utah has had
to proceed through a judicial determination.
These two states are in contrast to the
standard concept of duty of water found in
many states — for example, Wyoming and
Nebraska — in which a fixed duty of water
has been adopted that is applicable state-
wide.
3. Introduce Incentives to Use Water More
Efficiently. Historically, the Agricultural Con-
servation Program administered under the
U.S. Department of Agriculture, with technical
assistance provided by the Soil Conservation
Service, has provided cost-sharing funds to
farmers and irrigation districts for irrigation
system improvements, most of which had water
quality benefits. This program has been
relatively inactive in recent years because of
lack of funds. However, this program should
play a very important role in the Influent
Control Approach, as a part of the federal-state-
local water users tripartite.
Action — Most western states have
revolving funds or low interest loan
programs for water resources planning and
development. Generally, these programs
require the applicant to be an irrigation
district or other corporate body. Where such
state programs exist, change in the legisla-
tion and/or regulations for participation
qualification should be changed to allow:
(1) individual irrigators to qualify;
(2) broaden the use of funds to include on-
farm improvement practices as well as im-
provement of delivery systems; and
(3) include in the objectives of the program
the improvement of water quality. When
states have no such programs, a low or no-
interest loan program containing the above
three components should be adoped in order
to cooperatively assist with the federal
429
-------
IMPLEMENTATION
government and local water users in achiev-
ing improved water management and agri-
culture practices.
In addition, dissemination of information
about other state and federal agency incen-
tive programs should be carried out by the
state water agencies, particularly to the
designated management areas and
cooperation extended to insure utilization of
such programs.
Other incentive programs, which may
require legislature enactments or agency
regulations, could include encouragement
of trading, leasing or selling of "saved"
water from more efficient practices as an
inducement to improve the delivery systems
and methods of application. State or local
water markets, under the direction of the
State Engineer (or equivalent state office),
could monitor or control the uses of these
waters.
To counter the traditional attack against
such an incentive program, it is highly
conceivable downstream juniors would be
the most likely to benefit, particularly if
they were given priority to pay for this
water.
Rationale — By providing incentives for
water users in designated areas, the
farmers will have an opportunity to volun-
tarily improve their water use practices,
which in turn will result in improved irriga-
tion return flow quality. This is consistent
with the philosophy of encouraging volun-
tary compliance versus forced or involun-
tary compliance. Further, if states are to
create standards and criteria for beneficial
use, it is the opinion that some mechanism
should be made available to the farmers
that will facilitate compliance with the new
criteria. Without it, irrigators are on solid
legal grounds to continue exercising this
water right as they have in the past. The
legal cost to the state and water users to
change this traditional practice may far
outweigh devising a process and procedure
by which water users can be encouraged to
improve their efficiency in water use for
both quantity and quality benefits, while
still protecting the downstream users.
Precedent — Funds have been made
available to irrigation districts and water
users through the federal Department of
Agriculture-Agricultural Conservation Pro-
gram, with the Soil Conservation Service
providing technical assistance. In addition,
several states have state-wide programs in
which low or no-interest monies are made
available to water users for improving their
delivery systems. In Wyoming, the funds
may be used for improving laterals and
application practices.
4. Include the Element of Water Quality in
Newor Transferred and Changed Water Rights.
Action— The water quality element
should be a general provision added to all
new water rights and requests for exten-
sions, changes in use and transfers, in order
to provide the necessary authority to state
water agencies for later setting and enforc-
ing of numerical standards (either with
respect to water application or return flows,
or both). Where water quality standards on
streams for beneficial use have been
realistically set, such standards can be
incorporated by reference to water rights
from that source of supply.
This action may require legislative en-
dorsement, but under the vast majority of
state law, it is conceivable that agency
regulations can initiate this component.
Rationale — The element of water quality
is only implicit in western state water laws,
with the exception of California which has
made it an explicit element in all water
rights since 1969. As a consequence, water
users must normally rely upon common law
doctrines and private litigation to protect
their water right where the quality has been
degraded to levels that hinder usage.
Because the quality element is not explicit,
as are the other elements of a water right —
quantity, source, point of diversion, type of
use, and place of use — the state agency
charged with adminstration of the water
laws and rights is not in a favorable posi-
tion to initiate action to prevent harm from
water quality deterioration, and thus the
management capability extends only to
quantity control.
Precedent — In California, there is a
general provision (which is added to all new
water rights, extensions on water rights,
and changes in ownership or type of use)
that the water will be used in such a manner
as not to unreasonably degrade the usage of
430
-------
INFLUENT CONTROL APPROACH
water for downstream users. In some in-
stances, the State Water Resources Control
Board has also applied numerical water
quality standards to particular water
rights.
5. Adopt and Enforce a Reporting and
Recording System for Water Rights.
Action —. Notice would be given to all
water users and water right claimants to
submit a report to the water right ad-
ministrative agency indicating their name,
address, basis for claiming right to use
water, use of water, source and beginning
data for water use. Water users may be
given notice by publication in local
newspapers. Many states have already in-
itiated a "tabulation of water rights"
program to acquire this data. It is
necessary, however, to also adopt a system
of annual reporting, indicating particularly
changes in ownership since other material
changes (e.g., transfer in type and place of
use) require state approval.
Rationale — Although this component is
more directly related to improving water
management within the states rather than
irrigation return flows only, it is related to
the ability of the state to manage water
quality because of the relationship between
the diversion and application of water and
the resultant return flow water quality.
Most western states have inadequate
knowledge of present ownership of water
rights, and thus: a) have procedural dif-
ficulties in notifying water users of matters
directly affecting their rights; b) are unable
to effectively remove "paper water rights"
from the records that are maintained and
thus making forfeiture provisions in the law
nearly useless; and c) would be hampered in
incorporating the element of water quality
to new, extended or changed water rights.
One of the major difficulties faced in
attempts to control irrigation return flow
quality is "contributor identification." A
data base of who the water right holders are
will greatly facilitate efforts to encourage
implementation of best management prac-
tices and best agricultural practices.
Precendent — Both Idaho and Oklahoma
have a system by which the current owners
of water rights are required to submit to the
state water right administration agency an
annual report (in the case of Oklahoma, this
is done on a computer card) which specifies
who the users are, where the water is used,
and approximately quantity. Failure to sub-
mit these annual reports serves as prima
facie evidence of non-use and could lead to
forfeiture of the water right.
6. Recognize Reasonable Degradation From
Agricultural Water Use.
Action — Legislative recognition of this
natural consequence of water use for irriga-
tion purposes is needed at the state and
federal levels.
Rationale — It is commonly accepted that
any use of water for irrigated agriculture is
going to result in some degradation of the
quality of return flows. To pretend
otherwise is to either continue a process of
"playing the game" or will ultimately
remove irrigated agriculture with its ob-
vious adverse effects. Common knowledge
knows the latter will not occur, but a
tremendous and unnecessary cost to prove
it could be extended upon irrigators and the
public through the failure of legislatures to
recognize natural processes of water use.
Precedent — New Mexico has adoped a
specific provision in their statutes which
states that ". . . reasonable degradation of
water quality resulting from beneficial use
shall be allowed" (NW Rec. Stat.§75-39-ll).
Montana has arrived at the same conclu-
sion by defining "naturally occuring con-
ditions" in their water quality standards as
those "present from runoff or percolation
over which man has no control or developed
land where all reasonable land, soil and
water conservation practices have been
applied" (MAC 16-2.13(10)-S14480, Water
Quality Standards, § (3) Definitions).
Several state supreme court decisions also
recognize certain degradation from water
use, e.g., Ravndale v. North Fork Placers (91
p.2d 368 Idaho, 1939) where some con-
tamination from a mine will necessarily
occur to a stream.
7. Adopt an Agricultural Practices Act.
Action — Many of the 17 western states
have laws and programs requiring the
licensing of agricultural chemical dis-
tributors and applicators with the state
Department of Agriculture. The laws
431
-------
IMPLEMENTATION
and/or programs should be revised or new
legislation adopted to include the following:
a) Sediment and erosion control.
b) Licensing and control over application
of agricultural chemicals to include pest-
icides and artificial fertilizers.
c) Creation of an Agricultural Practices
Control Board consisting of representa-
tives from the agriculture, water quanti-
ty and quality, soil conservation (if
separate) and fish and wildlife agencies.
and appointed members of the public.
The board's functions would primarily
be establishing rules, regulations and
procedures for carrying out a) and b)
above and insuring functional im-
plementation through coordination and
designation of duties to appropriate
state agencies.
Rationale — Due to the impact upon
water quality resulting from the application
of herbicides, pesticides and fertilizers in
the agricultural sector, and the inability of
the state to control these practices, it is
highly recommended that an Agricultural
Practices Act be adopted which requires
licensing and monitoring of distributions
and applications of such chemicals and
control over harmful land management
practices contributing to erosion and subse-
quent sediment problems in receiving
waters. There may be many other
agricultural practices which could be in-
cluded under such an act.
It is essential to recognize the inter-
connection between these activities and
resulting water quality problems (which
may in turn contribute to downstream and
ground water supply problems) and the
usual division of jurisdiction and duties
between various state agencies. Short of
complete reorganization of state agencies to
insure interrelated activities all under one
agency (which may not only be impossible
but highly undesirable), an Agricultural
Practices Control Board (APCB) consisting
of action representatives from the various
involved state agencies and members of the
public could insure coordination and im-
plementation of their rules and regulations.
The current 208 planning bodies could be
designated to assume local implementa-
tion.
Precedent — Again, California has led
the way in licensing and monitoring of
commercial applicators of herbicides and
insecticides. Oregon has been considering
the appropriateness of such an act to
alleviate their major irrigation return flow
quality problems. However, a comprehen-
sive agricultural practices act has not been
prepared in any of the western states. Iowa
has adopted erosion control legislation that
even authorizes imposition of a fine on
those who fail to adopt approved practices.
8. Promote the Close Cooperation or
Integration of State Water Agencies.
Action — To facilitate the implementa-
tion of the Influent Control Approach to
irrigation return flow quality management,
it is important that close cooperation and
coordination exist between state water
agencies through operation of a liaison
board or committee, or integration of the
state water agencies under one department.
Rationale — It is difficult to provide the
necessary agency support to carry out any
new program, but even more difficult to
introduce a program of management and
control over an area of activity highly
sensitive to government intervention. In
addition to adding duties to agencies often
already burdened with heavy programs.
efforts to control irrigation return flow
quality meets with strong resistance from
an inflexible and institutionalized system
of property rights to the use of water in
which the state water quantity agency often
maintains a close relationship with the
water users. The result is potential polariza-
tion between the state agency carrying out
water quality control and the state water
quantity agency.
However, because of the interdependence
of water quantity and quality, particularly
as a natural process in water applied to
irrigation, it is inconsistent to promote the
goals of P.L.92-500 and not promote the
coordination or integration of agencies
charged with carrying out water quantity
and quality control.
Precedent — In 1969, California com-
bined the water quality and quantity agen-
cies under the Porter-Cologne Act in order to
specifically attempt to manage the
resources in a rational manner. In 1972,
432
-------
INFLUENT CONTROL APPROACH
Washington created the Department of
Ecology which encompasses all three of the
primary water functions; namely, water
adminstration, water quality control and
water resources development. Texas,
Oregon and Kansas are comtemplating an
integrated approach, while Oklahoma has
chosen to utilize an advisory coordinating
board to interface the various water ac-
tivities of numerous state agencies.
EPILOGUE
Based upon discussions with state water
quantity and quality personnel from all 17
western states, it is apparent that most states
feel a real credibility gap exists between the
Environmental Protection Agency and the
State agencies; that in those states where the
state agencies attempted to implement the
federal program, a credibility gap developed
between the water users and the state agency;
and, in several states the personnel expressed
the opinion that EPA let them down by backing
off after they attempted to carry out a control
approach they did not agree with in the first
place. In nearly every state, the water user
organizations and individuals have polarized to
combat the imposition of uncertain regulation
over their possible water use. A permit concept
is nothing new to them and they know that
eventual control can emanate from an initially
harmless permit. For this reason, the current
relationships between the three principals can
be graphically described as shown in Figure 3.
Missing are the prime ingredients of credibility
and understanding. It is considered necessary:
1. that public officials responsible for
carrying out laws appreciate the position of
the water users and the nature of irrigated
agriculture; and
2. that water users appreciate the water
quality problems caused and the often
awesome responsibilities and duties of the
federal and state water officials in carrying
out the legislative mandates.
If water users and state and federal agen-
cies will embark upon a cooperative under-
taking, a tripartite relationship can evolve that
will instill credibility and achieve improved
water quality. This would once again get those
directly involved with water use and with
administration of laws working together and
not through advocates. The tripartite
\
EPA )
/
/ STATE \
\ AGENCIES y'
\ /
s
I
WATER
\ USERS
Figure 3. Polarization between water users and
state and federal water quality agencies.
Figure 4. Tripartite relationship.
relationship (Figure 4) can develop if personal
objectives and discipline orientations are not
allowed to constrain a causal and flexible
approach. A cooperative approach will
facilitate the implementation of an action
program.
In Figure 5, the relationship between water
quality problems resulting from irrigation
return flows and the Influent Control Approach
is shown. Irrigation return flows can be in-
fluenced to successfully eliminate controllable
degraded discharges from irrigated agriculture
and improve the quality of out nation's waters
affected by such discharges. An Influent Con-
trol Approach is recommended based upon the
conclusion that BMP + BAP = IRFQC >
IWQ —> Cleaner Water.
433
-------
IMPLEMENTATION
THE PROBLEM
BtioB of ttoHf Quo%
Degraded D*cfcorgM
Irrigation
Retwrft
Ftovt
IRRIGATED AGRICULTURE
T
'CJaaMrWaMr
by
Elimination of
Controllable Degraded
Discharge*
This farm is symbolic
objectives of PL. 92
r» not to be interpreted
"Zero Dfecnarge"
of the
5OO a
Uee 0' Apr CHem.coH
r~^~~^^ \
C^',v£tiKCE
D..«r»,cr and
Dt ---*«fj Srttem
•^^^^ \CROPLAND
A0phco1ion.
tcnggt>:>" Methods
A Practices
Agronomic
Proct.CM
AppllCOtlOU .
Methods and
Practices
S«'ir,,ty
SOURCES OF RETURN FLOW
DEGRADING EFFEC
UPON WATER OUAL
krigotiaii Rete»n Flow!
An Influent Control Approach
ASSUMPTION:
Improved Agricultural Procticei + improved Wdtl
Invroved Water Quality
(IAP» IWM.IWQ}
CONCLUSION
B«st Uonegrnvnt Practices + Best AgrtcuNurol Pfaclioi •
Irrigation Return Flow Quality Control
(BMP* BAP-IRFOC— IWO)
PROGRAM:
I
INFLUENT CONTROL APPROACH 11CAI
t ACTION N.
BEST MANAGEMENT BEST AGRICULTURAL
PRACTICES1BMP] PRACTICES1BAPI
\ S Proper ApplicatiOH
Improved Local Water Propor Land of AgricvHural
Management (ILWMI UielPLUI Cntm.coli 1PAAC1
1
1
1
PROBLEM AREA 1
Designate area
and area entity.
D*. n6at4
and critena for
n wee
| COMPONENTS
| STATEWIDE !
3. Introduce incen-
tives tori LWM
to water rights
3. Add reporting and Sedi
1
STATEWIDE ]
7. Adopt Agricultural
Practices Act
1
|
nent Licensing and
recording for voter and control over
riahtt erosion application of
con
6. Recognize reason-
able degradation
from irrigation
return flOM.
rol agricultural
chemicals^
fertilizers •fttf
pesttcid*»
6. Promote doM cooporation of ttato vator
wd otter related fwcfioiu.
(Rodoeertct tt Stogerboe . I17T)
Figure 5. Relation between the Influent Control Approach and water quality degradation from irrigation
return flows.
434
-------
A Process for Identifying,
Evaluating and Implementing
Solutions for Irrigation Return
Flow Problems
EVAN VLACHOS, J. W. HUGH BARRETT, PAUL HUSZAR,
JAMES J. LAYTON, GEORGE E. RADOSEVICH, MELVIN SABEY,
GAYLORD V. SKOGERBOE, and WARREN L. TROCK
Colorado State University; Fort Collins, Colorado
ABSTRACT
The purpose of this study has been to
develop an effective process for implementing
technical and institutional solutions to the
problem of return flow pollution. The process
developed is to: (1) define the problem in terms of
its legal, physical, economic and social
parameters, (2) identify potential solutions in
relation to the parameters of the problem, (3)
test the implementability of these potential
solutions for diverse situations, and (4) specify
those solutions or groups of solutions which are
the most effective in reducing pollution and are
implementable. This process is initially concep-
tualized in the paper and then the general
results of its application to three study areas in
the western United States are summarized.
INTRODUCTION
Water quality control has been a broad
national objective since the enactment of P.L.
84-660, the Water Quality Act of 1956. However,
from 1956 until the late 1960's, the emphasis has
been almost entirely upon control of point
sources of discharge from municipalities and
industries. Obviously, these elements of pollu-
tion could be readily identified and various legal
and economic measures could be designed to
induce or compel elimination or reduction of
harmful discharges.
Contrasted to this concern, three different
conditions have produced a slow response by
state and local officials with regard to
agricultural pollution control programs. The
first condition is the relative invisibility of non-
point pollution. The second has to do with the
more or less localized nature of the adverse
effects from agricultural pollution and the dif-
ficulties of determing injuries in the absence of
obvious outfalls. And, finally, it is only recently
that a concerted effort by the federal govern-
ment has been undertaken in tackling in some
general way the problem of non-point pollution
and in interpreting the provisions of a very
complex law.
Thus, other than blatant violations (such as
direct discharge of animal waste and chemicals
into streams and rivers), control of pollution
from agricultural activities has been noticeably
lagging. In addition, agricultural uses of water
are controlled by state agencies which, at least
in most western states, are primarily concerned
with water allocation, distribution and ad-
ministration. In this respect, concern for
beneficial use of water, duty of water use and
wastages does not also include the degradation
of return flows from overapplication or misuse
of water.
Agricultural water quality control has
recently become a substantive part of dis-
cussions among the various states in the West.
Problems ranging from salinity and chemical
degradation, sedimentation and other problems
associated with suspended material have been
examined predominantly from a physical con-
trol perspective and technologies have been
developed which could alleviate or eliminate
such problems.
An examination of the total quantity and
diversity of nonpoint source pollutants, es-
435
-------
IMPLEMENTATION
pecially in the rural areas, indicates the com-
plexity of the problem that must be faced. It is
obvious that simply applying technological
solutions is not going to solve the nation's water
quality problems. Indeed, more and more it is
recognized that many of the gains made in the
point source area will not result in cleaner water
because of the failure to act in controlling the
highly significant non-point sources of pollu-
tion. (It is here that the NCVVQ has highlighted
the historical asynchrony that has developed
between implementation of sections of the Act
dealing with point source treatment alone —
Section 201 — and point and non-point source
management considered jointly — Section 208.)
In addition, however technically feasible they
may be. proposed solutions to problems of non-
point pollution have been met with substantial
resistance, despite economic and social analysis
demonstrating the long-range benefits — local,
regional and national — that would result from
the mitigation or elimination of water quality
degradation.
It is obvious that this problem is two-fold.
On the one side, there is disenchantment with
the physical difficulties in dealing with the
invisible irrigation return flows wherever they
exist. However, equally important is the lack of
a basic agreement as to resolving a problem of
immense complexity, namely the conflicting
and competing goals, objectives and priorities
of water resources management. At the same
time, there are also a number of factors that
influence the strategies for resolving this
problem such as the level of environmental
quality desired; the cost of achieving that quali-
ty; the equitable distribution of costs; the
presumed benefits to be derived from enhancing
environmental quality; and, finally, the means
for achieving that quality, including the host of
economic, legal, political, technical, as well as
social constraints.
The emphasis of the present approach is to
utilize existing technologies for improving
water management and alleviation of salinity
and sedimentation from irrigation return flows,
and introduce institutional alternatives which
will make possible implementation of a range of
solutions per specific cases.
In the discussion, "institutions" are defined
as those social mechanisms by which society
organizes, manages and directs its affairs.
"Institutional alternatives" are the whole range
of legal, economic, political, and cultural in-
stitutions (or crystallized ways of doing things)
which are used for meeting social needs. The
purpose of the analysis, therefore, is to view the
range of institutional alternatives that might be
incorporated into intervention schemes, which
will then be ultimately tested for feasibility and
acceptability in a number of irrigated areas
representing a spectrum of irrigation return
flow control problems. More than anything else,
the central concern has to do with the presenta-
tion of specific steps involved in the process of
building a basis for implementation and for the
adoption of desired changes.
The Research Approach
In general, the study undertaken by the
interdisciplinary team has attempted to further
the public and political decision-making process
involved in policies, law, standards, and
regulations for pollution control from
agricultural uses of water resources. In outlin-
ing the thrust of such an approach, the follow-
ing interrelated objectives have been identified
as necessary:
a. the identification of appropriate
technologies and the institutional alter-
natives that together may improve
irrigation return flow quality control;
b. the assessment of combinations of
technologies and institutions as to their
feasibility of implementation in selected
areas in the West, through field
responses and community feedback;
c. the analysis of the process of change and
decision-making as a basis for eventual
efforts of implementing return flow
quality control;
d. the consideration of trade-offs between
engineering, economic, legal, and
sociological alternatives within the con-
text of larger established policies and
goals.
In the proposed approach, there are five
interlocking steps, in a process of cumulatively
building experience with the problem and of
providing an analytical framework for
evaluating institutional alternatives.
1. The investigation of those current
technologies which could control quality
of irrigation return flow and the assess-
ment of their impact on the problem.
2. The generation of alternatives, or iden-
tification and analysis of various
436
-------
IDENTIFYING SOLUTIONS
technical and institutional solutions to
problems of quality control.
3. The assessment of those alternatives
and a critical analysis of total system
effects and of criteria for weighing alter-
natives.
4. Evaluation through the help of affected
recipients and a juxtaposition of feasible
strategies and programs of quality con-
trol.
5. The building of the basis for implementa-
tion with special emphasis on the factors
that hinder or facilitate the adoption of
new measures.
By using this type of an approach and
through continuous interaction among
members of the research team, a consensus as to
critical findings has been established. The con-
cern throughout the conduct of the study has
been to provide concrete validation of the
theoretical processes described above, and, at
the same time, through interaction both within
the team as well as with water users in the
particular areas of concern, relate to actual
circumstances the critical findings concerning
solutions, constraints to implementation, and
the basis for developing strategies for con-
trolling irrigation return flow.
The problem does not reside exclusively on
the determination of an "appropriate solution,"
although the last has been a central point in
finding out what really can be done to com-
municate effectively the spirit of the law with
regard to problematic situations in a variety of
cases in the western United States. The concern
begins with the process of arriving at ap-
propriate solutions, in assessing in an inter-
disciplinary manner alternatives, and in outlin-
ing the steps for an eventual process of im-
plementation of whatever is the agreed-upon
"solution" or program.
However, it is important to underscore
again the centrality of the search for an "ap-
propriate" or "balanced" solution. A key ele-
ment and assumption of the study has been that
such a desired "appropriate" solution can be
reached by considering through an inter-
disciplinary analysis a variety of factors that
bring together what is technically sound,
economically viable, legally pertinent, socially
acceptable, and, finally, what is politically
feasible or implementable. This search for the
combination of a wide spectrum of conditions
leading to the "appropriate solution" is ar-
ticulated in the categories of Table 1.
In order to both further explicate this ap-
proach and, at the same time, summarize the
central argument of our study, the key dimen-
sions shown in Figure 1 may be used, where the
sequence of the approach involves:
a. setting the stage, bounding the problem
and considering potential solutions;
b. arriving at appropriate solutions and
determining alternative strategies; and
TABLE 1
Key Characteristics of an Appropriate Solution.
APPROPRIATE SOLUTION
Technically
sound
Economically
viable
Legally
pertinent
Socially
acceptable
Politically
feasible
Technologically practicable
Long-term benefits vs.
short-term gains
Ecologically non-damaging
Economically achievable
Providing efficient resource
allocation
Promoting equitable distribution
Protects vested rights
(due process)
Complies with legal criteria in
substantive law
Complies with administrative
procedures
Promotes beneficial use concept
Is characterized by flexibility
& predictability
Congruent with current practices
Consistent within cultural
context
Compatible with organizational
structure
Corresponding with desires
of people
Compatibility of interests
(local, regional, national)
Establishes priority of problems
Compares severity, intensity and
magnitude of potential effects
Aims at constituency
satisfaction
437
-------
IMPLEMENTATION
c. building the basis for implementation
and facilitating the acceptance of ap-
propriate solutions.
Figure 1 outlines also particular aspects or
dimensions in each of the above phases. Each of
these subdimensions has been intensively
analyzed as part of the desired interdisciplinary
synthesis aimed at building the basis for im-
plementing a given solution (or, for relating
appropriate to acceptable "solutions"). In addi-
tion, Figure 1 underscores the iterative steps
involved in such a process. Thus, in implement-
ing an appropriate solution (in making it accep-
table) monitoring and feedback may allow the
problems to be redefined (reexamine the stage,
SETTING Tn£ STAGE I
critical variables, law, or affected parties); the
appropriateness of the solution to be questioned
(especially with regard to trade-offs and local
sensitivity); and the acceptability of the propos-
ed solution to be reexamined in terms of the
degree of local involvement, availability of
implementation mechanisms and coordination
between all responsible agencies.
In summary, the process of implementation
brings together the objectives of the law and the
desires of the recipients in a compatible, com-
plimentary and negotiated scheme whose ul-
timate aim is the proper solution of problems of
pollution and irrigation return flow.
Determining affected parties
• responsible oroanizat
• affected individuals
• related aaencies
r,~
es ce:»-eer res Don s i D ie
Figure 1. A Sequential Paradigm for Building the Basis for Implementation
438
-------
IDENTIFYING SOLUTIONS
The essence of the approach developed is
that the problem requires consideration of a
number of alternatives leading to some solution.
The process of implementation brings together
problems and solutions, as well as an assess-
ment of the various alternative strategies. This
process is based on a juxtaposition of a set of
assumptions and of related programs as out-
lined in the following manner:
Assumptions
-Improved Agricul-
tural Practices
-Improved Water
Management
-Public Accept-
ability
Examples of
Intervention
Program
-Best Agricultural
•Incentives Practices
•Market
mechanisms +
•Legal
enforcement -Best Management
•Centralized Practices
demands
•Etc. +
-Public Mobilization
The key problem in this study was not so
much the repetition of the conditions in the
areas of concern that may hamper or facilitate
potential change and implementation of new
technologies (although this is a necessary part
of the problem); but, the focusing upon very
specific strategies and tactics required for a
dynamic process of effecting change. The im-
portant aspect is to develop a paradigm as to
how specific features of an implementation
process can be outlined and, at the same time,
formulate a particular plan for improving
irrigation return flow quality in areas of con-
cern.
Nature of the Problem
The following discussion deals with the
parameters of the irrigation return flow quality
problem. That is, it seeks to identify the par-
ticular causes of the problem in order to es-
tablish a basis for identifying potential solu-
tions.
Much of the focus in irrigated agriculture
has been to improve existing irrigation systems
by increasing the water supply, rather than
improving the use of existing water supplies to
more effectively produce crops and reduce the
quantity of return flows. Farmers generally
perceive the solutions to water problems as
revolving around more water supply; indeed,
many of the existing institutional mechanisms
for assisting irrigated areas facilitate this ap-
proach. As a consequence, many irrigated areas
are over-irrigated which results in large quan-
tities of irrigation return flows. In addition, in
many cases these return flows result in signifi-
cant water quality degradation. In such cases,
there is a direct relation between the inefficient
use of water and the resulting water pollution.
Therefore, alleviating water quality degrada-
tion from irrigated agriculture will, in most
cases, require increasing the efficiency of water
use, which involves improving water manage-
ment practices.
In improving water management prac-
tices, there exist a number of institutionalized
constraints which make the actual acceptance
of proposed practices difficult. Such practices
require that irrigation return flow quality con-
trol include both dimensions of the resource
problem — water quantity and water quality.
Separate categories of laws have evolved for
each dimension, each taking on characteristics
which contribute to the problem and compound
efforts to improve the quality of return flows.
Also, these practices must employ a program
that would incorporate cooperation between
organizational entities and the individual user,
something which is not now present. It is these
constraints that constitute a major part of the
problem at hand.
Problems of irrigation return flow quality
are compounded by the specific perceptions of
individual farmers regarding pollution and the
geographic significance of the problem.
Farmers know that using irrigation water will
cause some degradation, but the point at which
it becomes significantly detrimental and who is
responsible are a major source of contention.
Thus, many farmers either do not perceive the
consequences of their action or they believe that
with the existing means of irrigating (which are
the correct means), the level of pollution is
natural and, therefore, acceptable. In addition,
there is the lack of a broader perception in-
volving the regional nature of the problem,
since farmers are mainly concerned about their
own property. The critical point is that a water
user's perception of the farming situation and
the problem of water quality in particular
dictates how that person will accept any in-
novative technology to solve a given "problem."
The above discussion relates some of the
difficulties in gaining popular support for the
necessity to alleviate water quality degradation
from irrigated agriculture. However, the heart
439
-------
IMPLEMENTATION
of the matter and a major cause of the problem is
the use of too much water; thus, a central
constraint to improving water use efficiency in
the West is the present system of water law
administration. Water is allocated, distributed
and administered under a body of law which
grants to the user a water right synonymous to
the property right interest one can acquire to
land. The water right is not one of absolute
ownership, but rather one for the use of water
only and subject to specific conditions and
concepts which theoretically are prescribed to
protect the public and other users.
Until the past few decades, water used for
irrigation was not considered a type of use that
required strict application and enforcement of
the law to achieve water quality goals. In fact,
many of the concepts and conditions provided
guidelines for allocation and distribution, with
implementation carried out when the water
right was granted and thereafter only when
severe abuse occurred or another user com-
plained.
The primary elements of the water quantity
law which contribute to both the problem of
degraded return flows and efforts to improve the
quality are:
• Failure to enforce legal conditions for
water use, namely beneficial use and non-
waste.
* Constraints in the law which prevent the
transfer of excess and saved,water to
other lands or users where it could more
effectively be used.
• Lack of adequate recognition of the legal
duty to include water quality control as an
attribute of the water right to be enforced
particularly by irrigation districts.
• Restrictions or deficiency in the law on
the use of low-cost funding from
state/federal programs for water quality
control.
These four factors provide the explanation
for water user conduct as well as constraints to
adoption of more efficient physical and
technical solutions that may not only improve
the quality of return flows, but also increase
crop production.
The doctrine of prior appropriation has led
water users into a continual diversion of their
full "water right" for fear of loss of this right if
the full amount were not used. Therefore, users
have been unwilling to sell, rent or lease any
portion of their water right, which could have
led to economic benefits to both parties, as well
as more efficient use of the resource. Thus, there
has been no market for reallocating irrigation
water. In addition, farmers have been able to
pass on to downstream water users part of the
costs of production in the form of pollution.
The present institutional arrangement
allocates water on the basis of a priority of
rights rather than on the value of use. The price
of water is generally the cost of its conveyance
to the farm and does not represent the value of
opportunities foregone. The result is that the use
of water is not competitive; it is not allocated to
its highest valued use; and its relatively low
price causes it to be excessively applied.
With the exception of irrigation water, farm
inputs are allocated through markets. Labor
and capital, for example, are allocated and
priced through markets according to the value
of their use. Consequently, water tends to be
relatively cheap, so that profit-maximizing
farmers rationally substitute water for capital
and labor (i.e., water management) in the
production process. The result is an over-
application of water with associated return flow
pollution.
Irrigation return flow pollution also results
from the avoidance by farmers of some costs of
production. The profit-maximizing farmer
attempts to minimize production costs. In so
doing, however, he may select production
methods and techniques which are low cost to
him, but polluting to downstream water users.
Alternative production methods and techniques
may be less polluting but higher cost to the
farmer. By selecting the lowest cost methods
and techniques, the farmer passes on part of the
costs of production to downstream water users
in the form of water pollution.
Irrigated agriculture is a collective enter-
prise involving all of the users and improving
existing water management practices, whether
to alleviate water quality degradation or more
effectively utilize existing water supplies to
increase crop production, certainly requires
collective action. There exist a number of
organizational entities that administer irriga-
tion, but, generally, there is a lack of explicit
rules established for the management of this
resource with regard to quality. There also
exists a lack of communication and coordina-
tion between agencies and districts, and the
440
-------
IDENTIFYING SOLUTIONS
farmers with regard to how the water should be
managed.
As a consequence of the lack of an explicit
institutional framework surrounding this
problem and certain individual perceptions that
do not enhance a specific water quality manage-
ment ethos, implementing a program of irriga-
tion return flow quality management can be
expected to be a very difficult task, further
complicated by the economic and legal con-
ditions outlined above.
Identification of Potential Solutions
The range of possible solutions to irrigation
return flow pollution is, of course, a function of
the parameters of the problem identified in the
previous section. Thus, potential solutions are
discussed in this section in terms of the causes of
the problem.
There are a number of potential solutions
for controlling the quantity and quality of
irrigation return flow. The irrigation system
may be subdivided into the water delivery
subsystem, the farm, and the water removal
subsystem. The use of efficient practices in the
delivery canals and pipelines, as well as im-
proving on-farm water management, will
minimize the problems in the water removal
system. In most cases, the key to minimizing
irrigation return flow quality problems is to
improve water management practices on the
croplands.
The water delivery system can be improved
by lining canals and laterals, using closed
conduits for water transportation, providing
adequate control structures, and installing flow
measuring devices.
Improved practices that can be used on the
farm include judicious use and application, or
placement, of fertilizers, use of slow-release
fertilizers, controlling water deliveries across
the farm, use of improved irrigation application
methods (e.g., subsurface application, sprinkler
irrigation, or trickle irrigation), control of soil
evaporation, use of a pumpback system to allow
recycling of surface return flows, erosion control
practices (e.g., contour farming), and irrigation
scheduling to insure that the proper amounts of
water are applied at the times required by the
plants.
In the water removal subsystem, open
drains and tile drainage can be used to collect
return flows, which can then be subjected to
treatment on a large area or basin-wide basis, if
necessary.
Identifying appropriate technological
solutions must be related to the nature of the
problem, i.e., water quality degradation as a
result of surface return flows or subsurface
return flows, or both. Knowing the sources of
pollution, then potential solutions can be iden-
tified. The appropriateness of such solutions
will be related to other "site specific" physical
parameters, as well as historical irrigation
methods and practices in the area, and the
perception of the users regarding the necessity
for change. In addition to informing the water
users of the existing irrigation return flow
problems, it becomes necessary to demonstrate
appropriate technologies in order to gain farmer
acceptance. This phase, as well as area-wide
implementation, could easily be hampered by
the lack of sufficient technological assistance
and by the legal constraints on the use of low-
cost government funding to achieve water quali-
ty improvements at the farm level.
Improved irrigation water management
practices will almost invariably result in re-
duced demand for water diversions. The real
difficulty in gaining water user acceptance lies
in solving the problem of who benefits from the
saved water.
At the present time, the irrigator cannot
benefit from the water saved by improved
irrigation water management practices. Conse-
quently, little progress in water quality control
of irrigation return flows can be expected until
the water right issue is addressed.
One of the viable alternatives for producing
a positive incentive for water users to benefit by
improving their irrigation system is to establish
a market for irrigation water. In order to
minimize the disruption of the present in-
stitutional arrangement, the market form iden-
tified as having the most potential is a water
rental market. The demand for rental water
would represent its addition to the total value of
output per additional unit of water. The market
supply schedule would represent the water right
holder's increasing opportunity cost of using
the water himself rather than renting it. The
market equilibrium price would be greater than
the current costs of conveyance. Those demand-
ing and supplying would have an economic
incentive to use water more efficiently. That is, a
rental market would increase the price of water
to its marginal value in production and would,
441
-------
IMPLEMENTATION
thus, encourage more use of labor and capital
(i.e., water management) in combination with
the water, thus reducing return flow pollution.
Such an arrangement would take as given
the present structure of water rights and
allotments. Those with water rights or
allotments, however, would be allowed to rent
surplus water to other users without jeopar-
dizing their rights or allotments in the future. In
most states, such a market could be created by
removing the legal and physical uncertainties
associated with such transfers under the pres-
ent system. Transfers within irrigation dis-
tricts ofexcess or saved waters requires cnanges
in both federal reclamation and some state
laws.
A market allocation, however, might not be
sufficient to correct all return flow pollution.
Farmers would still have a profit motive for
externalizing all possible production costs, in-
cluding the costs of controlling pollution.
Waters users (or irrigation districts and com-
panies) should be required to internalize costs
imposed on other water users, public or private,
through adoption of standards and criteria for
beneficial use and creation of programs (volun-
tary incentives) and penalties (compulsory com-
pliance) that could be employed by water ad-
ministrators to protect adversely affected par-
ties. In economic terms, this means the imposi-
tion of taxes and subsidies. The particular form
and application of such taxes and subsidies can
only be specified for particular cases. In general,
however, taxes can be utilized to adjust the price
of water to approximate a market price, thus
inducing farmers to be more efficient in its use.
Taxes can also be used to penalize farmers lor
return flow pollution, but monitoring costs are
typically prohibitive. Generally, subsidies in
the form of direct payments or technical
assistance and capital improvements appear
most applicable for improving on-farm manage-
ment of water. Finally, taxes and subsidies may
be jointly applied as, for example, with a tax on
water to approximate its market value and the
revenues from this tax being used to subsidize
farmers to adopt less polluting methods and
techniques of water use.
The argument made above concerning legal
considerations involved in problems of water
quality has set the stage for potential solutions
improving the law's sensitivity and ability to
address such problems. Essentially, identifiable
solutions include: adoption and enforcement of
criteria for beneficial use, waste and water duty;
removal of constraints concerning transfer of
excess or saved waters within irrigation dis-
tricts; promotion of low-cost funding; inter-
nalization of costs through adoption of stan-
dards for water use, creation of programs, or
compulsory compliance, all of which could be
employed by water administrators to protect
adversely affected parties.
Finally, the various alternatives must take
into consideration both individual attitudes and
the organizational structure that provides the
rules and mechanisms which influence in-
dividual behavior. For the individual, potential
solutions must involve an awareness of the
critical character of irrigation return flow. On
the other hand, for the organizational context,
solutions revolve around the creation of
mechanisms and practices which can facilitate
the adoption of means for acceptable water
quality standards.
Awareness by the farmer entails two con-
ditions. First, the awareness should be towards
specific on-farm management procedures which
enhance water quality. In addition, a holistic
approach as to why improved water quality will
enhance not only his neighbor's operation, but
also his own, must be explicated. Clean water
should be seen as beneficial to the farmer and
this benefit must be viewed as one that can be
attained only through an area-wide involve-
ment of water users.
Instilling this individual behavior as a
public good can only be accomplished when the
organizational structure supporting the farmer
embraces broader and firm commitment to
cleaner water. This can be done by using the
existing structure; by changing or restructuring
present arrangements; or by adding to the
existing framework. The use of existing
mechanisms such as the extension service, SCS,
the local mass media, Co-ops, etc., can provide
communication and information networks from
which the individual farmer can become aware
of the problem and the solution. Yet, some
valleys do not have a well-organized system of
communication by which water quality con-
ditions are adequately processed, investigated,
and disseminated. Such an organization can be
established either by modifying the existing
structure or creating a new one altogether. This
organization should work with farmers, agen-
cies and the public in such a manner that it
serves as a focal point for water quality informa-
442
-------
IDENTIFYING SOLUTIONS
tion; and as a nodal point through which flow
and exchange of information from above and
from below can be transferred into coherent,
collectively arrived at policies.
It should be noted that in the juxtaposition
of individual and organizational approaches,
the assumption was made that acceptance of
new management procedures will follow dis-
semination of knowledge and awareness.
However, this is not always the case. Since there
is not a necessary or sufficient causal
relationship between "appropriate" and "accep-
table" solutions, systems of rewards and
penalties must also be established in order to
provide support, to reinforce desired behavior,
and, generally, in order to make sure that
proposed solutions will not die because of
neglect or lack of sustained implementation by
the users.
Yet, as with all other solutions, these
proposed monitoring and enforcement
mechanisms must have an effective say or part
in any decision-making process regarding
irrigation return flow quality control. For, at the
end, the pursuing of an effective water policy is
part of a larger commitment and of an ethos
that combines individual motivation and
economic opportunities within an organiza-
tional context that makes possible collective
social action and timely technical interven-
tions.
Generally, the alternative solutions pro-
posed for evaluation ranged from those which
were wholly technical (e.g., rehabilitation of
distributive systems) to those which were insti-
tutional (e.g., creation of water markets). Some
were combinations of technical and institution-
al measures which would cause improvements
in quality of return flows (e.g., cost sharing
arrangements for improved irrigation
facilities). They can be generally classed as: 1)
those directed to sources of water, generally,
those which would increase supply; 2) those
concerned with the influent, i.e., the water
diverted to agriculture; 3) those associated with
the management of land and water on farms;
and 4) those which were concerned with the
effluent, i.e., the return flow.
Testing the Implementability of Potential
Solutions
Following the identification of potential
solutions for return flow quality problems, the
team directed its efforts to evaluation. It was
understood that alternative solutions would be
more or less acceptable (and thus implemen-
table) depending on their impacts on the
affected parties. Testing procedures were de-
vised to determine technical, economic, politi-
cal, and social acceptability of alternative solu-
tions. They involved: 1) the project team; 2)
state and federal agency personnel; 3) irriga-
tion water managers; and 4) water users.
The first evaluation was done by the project
team. Composed as it was of engineers,
economists, sociologists, and an attorney, the
team was able to judge alternative solutions in
terms of technical, economic, legal, and social
feasibility (per criteria outlined in Table 1).
Obviously, inappropriate and ill-advised
solutions were weeded out, though the number
was not great. Alternatives with potential for
significant impacts on the quality problem and
those without prohibitive costs were retained for
evaluation by others. The team wished to pre-
sent the widest possible range of alternatives to
succeeding evaluators.
A second evaluation was accomplished by
federal and state agency personnel, chiefly
those presently or prospectively involved in
administration of quality improvement
programs. The alternative solutions were thus
screened by those with technical and legal
expertise, a group with a special concern for
administration of laws and programs. This
group tended to sort out those solutions which
did not fit within the framework of existing
laws, rules and regulations and which would
therefore be difficult to implement. The list of
alternatives was reduced, but not so as to
exclude some solutions which would be possible
with changes in laws, rules and regulations.
The third evaluation was completed by
managers of water supply agencies (e.g., irriga-
tion companies and districts) and their boards
of directors. These were individuals having
responsibility for distribution of water among
farms of members and patrons and for
maintenance of system facilities. Because they
are potentially responsible for administration
of revised rules governing diversions and use of
water, they tended to resist measures of control.
But they were aware of water quality problems;
they were generally convinced of possibilities
for improved use of water; and they tended to
favor quality control measures located and
administered at their level rather than at higher
or lower levels.
443
-------
IMPLEMENTATION
Finally, the fourth evaluation was done by
the farmers who use water in irrigation of crops.
They were interviewed separately; there was
discussion of the return flow quality problem:
and potentially useful solutions were outlined
and discussed. These individuals, though
alarmed by present efforts to control their use of
water, showed both ability and willingness to
comprehend problems of water quality and deal
with them. They were very practical in their
judgments of implementability of the various
alternative solutions, and they tended to favor
those measures aimed at improved use of water
in agriculture. It was these measures over which
they had some control.
The overall response to all such solutions
depended somewhat on who was doing the
evaluating. Administrators were more inclined
to favor the technical solutions which were most
familiar to the agency personnel. They were
inclined to prefer measures that they could
control and administer, since their experience
was largely with water development and water
treatment. Users tended to prefer those
solutions which emphasized management of
water in agriculture. They were aware of some
inefficiencies in water use. some non-
conservative uses of water and land, and they
knew of possibilities for improved management.
Managers of distribution systems wrere aware of
inadequacies in their systems and liked
proposals for improvement. They tended to
favor the influent control measures, i.e..
solutions affecting diversion and allocation of
water among users. Farmers understood these
solutions, too, but were understandably con-
cerned about possible reductions in their annual
allotments.
Probably the greatest support was found for
those solutions that dealt with improved
management of water in agriculture. There was
appreciation in most of the project areas for the
efficacy of those measures that affected on-farm
use. But there was also appreciation for
solutions proposing new controls on diversions
and use, for in two of the project areas water
allocations are usually large, i.e., there is an
abundant supply. The managers of distribution
systems and farmer-users of water know that
greater efficiencies in water use can be a-
chieved. Their concern is for loss of rights which
have been long held and carefully guarded.
There was some interest in water markets, as
means for reallocating supplies, but un-
familiarity with such a measure in some areas
prevented enthusiastic support.
CONCLUSIONS
The control of water quality raises two basic
issues (which are also present in any water
resources management scheme), namely, incen-
tives and enforcement. Questions here include:
What organizational structure is going to make
the rules and regulations and also enforce
them? How is the present problem of insuf-
ficient control going to be alleviated? How can
the marginal value of excess water be
operationalized to a water market? What is the
situation with regard to intra-system, inter-
state and inter-basin transfers of water? Broad
as these questions may be, they are also part of
the general considerations necessary for the
eventual control of irrigation return flow.
As repeatedly stated, three basic dimen-
sions of the irrigation system are central in all
strategies of return flow control: water delivery,
the user and removal efforts. With regard to
delivery, the critical point is that of control at
the inlet. It has been early agreed that the thrust
of the various alternatives should focus on the
user. Improvements in delivery systems, use of
better technology, improvements in removal,
etc.. should all be built around the user. The
critical point of this dimension is the manner of
applying the water on the land, i.e., on-farm
water management. Constraints to better on-
farm management include such factors as lack
of information technical assistance, lack of
control over the water, existing water rights,
lack of physical facilities for use of water, and
the lack of institutional facilitators (tradition,
value system, education, etc.). Simply, the
problem is one of motivating users to internalize
better management techniques.
Recognizing the thrust of the present study
in outlining the process for identifying,
evaluating and building the basis for implemen-
tation, the following conclusions summarize the
central findings:
1. The roots of the problem are the in-
stitutional arrangement for allocating
water, i.e., legal rather than market
adjustment.
2. The most appropriate solutions deal
with the diversions and use of water
rather than treatment of irrigation
return flows, i.e., the cause rather than
the symptom of the problem.
444
-------
IDENTIFYING SOLUTIONS
3. Solutions mutually beneficial to both
water quality and the farmer are most
implementable, such as:
a. subsidies for on-farm physical im-
provements;
b. providing technical assistance;
c. water rental markets.
4. Various means of improving on-farm
practices remain a useful approach for
controlling irrigation return flow.
5. Irrigation districts play a major role as
part of existing organizations in im-
plementing solutions.
6. Informational and educational
programs to assist individual farm
operators must be instituted early; be
imaginatively conceived; and be con-
tinuously monitored, modified and up-
graded if motivation for change is to be
encouraged.
7. There must be a clear definition as to
who has authority, control and respon-
sibility for specific tasks associated
with return flow control.
8. Major technological breakthroughs
should not be relied upon for providing
return flow control; instead, emphasis
should be on a combination of current
technologies and of institutional
measures.
9. State-wide and regional advisory com-
mittees should be part of a continuous
effort for cooperation, coordination and
combination of efforts and resources.
10. Structural opportunities for im-
provements should be utilized with sen-
sitivity to local conditions and through
local mobilization as part of a slow,
iterative and long-range process of im-
plementation.
11. Credibility and trustworthiness of
federal and state agencies in the eyes of
water users provide the important final
ingredient in understanding the need
for change; in motivating individuals
for accepting appropriate solutions;
and, in creating a climate of coopera-
tion and credence as to the need and
ultimate usefulness of a larger social
policy concerning "cleaner water."
In summary, the final approach to irriga-
tion return flow quality management requires
an imaginative combination of physical
methods, implementation measures and in-
stitutional arrangements. The success of such a
synthesis will be ultimately based on a gradual,
if not hierarchical, testing of alternative solu-
tion packages, on sensitivity to local conditions,
and on a committed, open process of com-
munication linking appropriate authorities
with individual users.
445
-------
Appendix
-------
Published Reports
from EPA's Research Program
on Irrigation Return
Flow Quality
Report No.
EPA-600/2
76-219
EPA-R2-
73-23.r>
EP A-1 3030-
05/69
EPA-13030
ELY 12/69
EPA-600/2-
76-2H7
EPA-13030
KLY04/71-OH
Assessment of Irrigation Return Flow Models
BY: W. R. Walker, Colorado State University,
Fort Collins, CO
ORDER FROM: GPO
NTIS-PB263897/AS
(83 pages)
Cation Transport in Soils and Factors Affect-
ing Soil Carbonate Solubility
BY: J. J. Jurinak, Sung-Ho Lai, and J. J.
Hassett, Utah State University, Logan,
UT
ORDER FROM: GPO-EP1.23/2:R2-73-235
NTIS-PB222006
(87 pages)
Characteristics and Pollution Problems of
Irrigation Return Flow
BY: Utah State University Foundation,
Logan, UT
ORDER FROM: GPO
NTIS-PB204817
(237 pages)
Collected Papers Regarding Nitrates in
Agricultural Waste Waters
BY: Federal Water Quality Administration,
CA; Reclamation, U.S. Bureau of Sacra-
mento, CA; California Department
of Water Resources, San Joaquin Valley,
CA
ORDER FROM: GPO
NTIS-PB197595
(186 pages)
Control of Sediments, Nutrients, and Ad-
sorbed Biocides in Surface Irrigation Re-
turn Flows
BY: D. L. Carter and J. A. Bondurant, U.S.
Department of Agriculture, Snake River
Conservation Research Center, Kimberly,
Idaho
ORDER FROM: GPO
NTIS-PB263610/AS
(53 pages)
Denitrification by Anerobic Filters and Ponds
BY: Roberts. Kerr Water Research Center,
USEPA, Ada, OK
ORDER FROM: GPO
NTIS-PB213719
(67 pages)
EPA 13030 Denitrification by Anaerobic Filters and
ELY06 71-14 Ponds — Phase II
BY: Robert S. Kerr Water Research Center,
USEPA, Ada, OK
ORDER FROM: GPO
NTIS-PB 218 413/3
(34 pages)
EPA 13030 Desalination of Agricultural Tile Drainage
ELY05/71 -12 BY: Robert S. Kerr Water Research Center,
USEPA, Ada, OK
ORDER FROM: GPO
NTIS-PB 213 890
(32 pages)
EPA 600 2 Effects of Irrigation Methods on Groundwater
7H-291 Pollution by Nitrates and Other Solutes
BY: C. W. Wendt, A. B. Onken, and O. C.
Wilke, Texas Agricultural Experiment
Station, Lubbock; R. D. Lacewell, Texas
A & M University, College Station, TX
ORDER FROM: GPO
NTIS-PB
( pages)
EPA -R2- Evaluation of Canal Lining for Salinity Con-
72-047 trot in Grand Valley
BY: G. V. Skogerboe and W. R. Walker,
Colorado State University, Fort Collins,
CO
ORDER FROM: GPO-EPl.23/2:RS-72-047
NTIS-PB 214 113
(199 pages)
EPA-660 2 Evaluation of Drainage for Salinity Control in
74-084 Grand Valley
BY: G. V. Skogerboe, W. R. Walker, R. S.
Bennett, J. Ayars, and J. Taylor, Colora-
do State University, Fort Collins, CO
ORDER FROM: GPO-EP1.23/2:660/2-74-084
NTIS-PB 240213/AS
(100 pages)
EPA-660/ 2- Evaluation of Irrigation Scheduling for
"4-052 Salinity Control in Grand Valley
BY: G.V. Skogerboe, W.R. Walker, J. H.
Taylor, and R. S. Bennett, Colorado State
University, Fort Collins, CO
ORDER FROM: GPO-EP1.23/2:660/2-74-052
NTIS-PB 235 633/AS
(86 pages)
449
-------
EPA-R2- Herbicide Contamination of Surface Runoff
73-266 Waters
BY: J. O. Evans and D. R. Duseja, Utah State
University, Logan , UT
ORDER FROM: GPO-EP1.23 2:R2-73-266
NTIS-PB 222 283
(99 pages)
Influence of Trickle and Surface Irrigation on
Return Flow Quality
BY: P. J. Wierenga. New Mexico State Uni-
versity. Las Cruces. NM
ORDER FROM: GPO
NTIS-FB
(175 pages I
EPA-600 2- Irrigation Management Affecting Quality
:t> s* and Quantity of Return Flow
BY: L. S. Willardson and R. J. Hanks. Utah
State University. Logan. UT
ORDER FROM: GPO
NTIS-PB
(206 pages)
EPA R.' Irrigation Management for Control of Quality
73-air. of Irrigation Return Flow
BY: L. G. King and R. J. Hanks, Utah State
University. Logan. UT
ORDER FROM: GPO-EP1.23 2:R2-73-265
NTIS-PB 222 773
(307 pages)
Managing Irrigated Agriculture to Improve
Water Quality — Proceedings of Nation-
al Conference on Managing Irrigated
Agriculture to Improve Water Quality. May
16-18. 1972
BY: USEPA. Robert S. Kerr Water Research
Center. Ada. OK; Colorado State Univer-
sity. Fort Collins, CO
ORDER FROM: GPO (Not Available)
NTIS-PB 220149 9
Graphics Management
Corporation
1101 16th Street, N.W..
Washington, D.C.
(306 pages)
EPA^6t> -j- Management Practices Affecting Quality and
7MX6 Quantity of Irrigation Return Flow
BY: L. G. King and R. J. Hanks. Utah State
University, Logan, UT
ORDER FROM: GPO
NTIS-PB 242 827 AS
(155 pages)
EPA-i3o*' National Irrigation Return Flow Research and
GJfri- '', Development Program
BY:J. P. Law, Jr., Robert S. Kerr Water
Research Center, USEPA. Ada, OK
ORDER FROM: GPO
NTIS-PB 209 857
(23 pages)
EPA600. 2-
76-158
EP.VIW.SU
EIAUfi 71 -m
EPA \MM
ELY05 72-11
EPA-R2-
73-168
EPA-13030
ELY04 71-07
EPA-13030
ELY06 71-13
EPA-13030
ELYOo 71-06
EPA-13030
- 11 71
Nitrogen and Irrigation Management to
Reduce Return-Flow Pollution in the Colum-
bia Basin
BY: B. L. McNeal and B. L. Carlile,
Washington State University, Pullman,
WA
ORDER FROM: GPO
NTIS-PB 259 328 AS
(141 pages)
Nutrients from Tile Drainage Systems
BY: California Department of Water Re-
sources. San Joaquin Valley. CA
ORDER FROM: GPO
NTIS-PB 216 552
(90 pages)
Possibility of Reducing Nitrogen in Drainage
Water by On Farm Practices
BY: Reclamation, U.S. Bureau of Scaramento,
CA
ORDER FROM: GPO
NTIS-PB 221 482
(83 pages)
Prediction Modeling for Salinity Control in
Irrigation Return Flows
BY: A. G. Hornsby, Robert S. Kerr Envi-
ronmental Research Laboratory. USEPA.
Ada, OK
ORDER FROM: GPO-EP1.23/2:R2-73-168
NTIS-PB 221 647
(55) pages
Removal of Nitrate by an Algal System
BY: California Department of Water Re-
sources, San Joaquin Valley. CA
ORDER FROM: GPO
NTIS-PB 205 425
(132 pages)
Removal of Nitrate by an Algal System —
Phase II
BY: California Department of Water Re-
sources. San Joaquin Valley, CA
ORDER FROM: GPO
NTIS-PB 216 552 0
(132 pages)
Removal of Nitrogen from Tile Drainage — A
Summary Report
BY: California Department of Water Re-
sources, San Joaquin Valley, CA; USEPA,
Robert S. Kerr Water Research Center, Ada,
OK
ORDER FROM: GPO
NTIS-PB 215 417
(30 pages)
Research Needs for Irrigation Return Flow
Quality Control
BY: G. V. Skogerboe and J. P. Law, Jr..
Colorado State University, Fort Collins,
CO; USEPA, Robert S. Kerr Water Research
Center, Ada, OK
ORDER FROM: GPO
NTIS-PB 219 979
(90 pages)
450
-------
Salinity Control Technology — A Report to the
Steering Committee of the Salinity Control
BY: Office of Research and Development,
USEPA, October 1972
ORDER FROM: (Not Available)
EPA-600/2- Scientific Irrigation Scheduling for Salinity
75-064 Control of Irrigation Return Flows
BY: M. E. Jensen, U.S. Department of Agri-
culture, Kimberly, ID
ORDER FROM: GPO
NTIS-PB 249 114/AS
(91 pages)
EPA-R2- Selected Irrigation Return Flow Quality
72-094 Abstracts 1968-1969
BY: G. V. Skogerboe, W. R. Walker, and V. T.
Sahni, Colorado State University, Fort Col-
lins, CO
ORDER FROM: GPO-EP1.23/2:R2-72-094
NTIS-PB 214 105
(221 pages)
EPA-R2- Selected Irrigation Return Flow Quality
73-271 Abstracts 1970-1971
BY: G. V. Skogerboe, Colorado State Univer-
sity, Fort Collins, CO
ORDER FROM: GPO-EP1.23/2:R2-73-271
NTIS-PB 222 796
(285 pages)
EPA-660/2- Selected Irrigation Return Flow Quality
744)49 Abstracts 1972-1973
BY: G. V. Skogerboe, W. R. Walker, R. S.
Bennett, and B. J. Zakely, Colorado
State University, Fort Collins, CO
ORDER FROM: GPO-EP1.23/2:660/2-74-049
NTIS-PB 235 385
(409 pages)
Selected Irrigation Return Flow Quality
Abstracts 1974
BY: G. V. Skogerboe, W. R. Walker, and
S. W. Smith, Colorado State University, Fort
Collins, CO
ORDER FROM: GPO
NTIS-PB 253 664/AS
(220 pages)
Selected Irrigation Return Flow Quality
Abstracts 1975
BY: G. V. Skogerboe, S. W. Smith, and W. R.
Walker, Colorado State University, Fort
Collins, CO
ORDER FROM: GPO
NTIS-PB
(254 pages)
EPA-13030 Techniques to Reduce Nitrogen in Drainage
ELY06/71-10 Effluent During Transport
BY: Reclamation, U.S. Bureau, Region II,
Sacramento, CA
ORDER FROM: GPO
NTIS-PB
(49 pages)
EPA-600/2-
76-019
EPA-13030
ELY08/71-09
EPA-13030
DYY06/69
The Effects of Agricultural Waste Water Treat-
ment on Algal Bioassay Response
BY: USEPA, San Francisco, CA
ORDER FROM: GPO
NTIS-PB 213 891
(57 pages)
Water Quality Management Problems in Arid
Regions
BY: J. P. Law, Jr., and J. L. Witherow, Federal
Water Quality Administration, Robert S.
Kerr Water Research Center, Ada, OK
ORDER FROM: GPO
NTIS-PB 198 125
(105 pages)
Managing Saline Water for Irrigation —
Proceedings of the International Conference
on Managing Saline Water for Irrigation:
Planning for the Future, August 16-20,
1976
BY: Texas Tech University, Lubbock, TX
ORDER FROM: GPO (Not Available)
NTIS-PB
Dr. Harold Drenge, Director
International Center for Arid
and Semi-Arid Land Studies
P.O. Box 4620
Texas Tech University
Lubbock, Texas 79409
(618 pages)
GPO Address:
Superintendent of Documents
Government Printing Office
Washington, D.C. 20402
NTIS Address:
National Technical Information Service
U.S. Department of Commerce
5285 Port Royal Road
Springfield, Virginia 22151
RSKERL Address:
Robert S. Kerr Environmental
Research Laboratory
U.S. Environmental Protection Agency
P.O. Box 1198
Ada, Oklahoma 74820
If no GPO or NTIS numbers are shown for report, the report
may be ordered by EPA number and title from the ap-
propriate source. An asterisk by report indicates report has
been forwarded for printing.
451
------- |