United States
Environmental Protection
Agency
Great Lakes
National Program Office
230 South Dearborn Street
Chicago, Illinois 60604
EPA-905/9-77-007-B
October, 1977
S-EPA
environmental impacf
of land use
on wafer quality
Final Report on the
Black Creek Project
-Technical Report
-------
The following P.L. 92-500, Section 108(a) reports are available through
the National Technical Information Service (NTIS) U.S. Department of
Commerce, Springfield, Virginia 22161. Prices listed for paper copy
and microfiche are prices given when placed on NTIS listing.
ENVIRONMENTAL IMPACT OF LAND USE ON WATER QUALITY
A Work Plan EPA-G005103
NTIS No. PB 227 112 Price: Paper $5.50, MF $2.25
ENVIRONMENTAL IMPACT OF LAND USE ON WATER QUALITY
Operations Manual EPA-905-74-002
NTIS No. PB 235 526 Price: Paper $9.25, MF $2.25
WATER QUALITY BASELINE ASSESSMENT FOR CLEVELAND AREA - LAKE ERIE
Volume I - Synthesis EPA-905-9-74-005
NTIS No. PB 238 353 Price: Paper $6.75, MF $2.25
WATER QUALITY BASELINE ASSESSMENT FOR CLEVELAND AREA - LAKE ERIE
Volume II - Fishes EPA-905/9-75-001
NTIS No. PB 242 747 Price: Paper $7.50, MF $2.25
ENVIRONMENTAL IMPACT OF LAND USE ON WATER QUALITY
Progress Report-1975 EPA-905/9-75-006
NTIS No. PB 248 104 Price: Paper $8.00, MF $2.25
NON-POINT SOURCE POLLUTION SEMINAR-NOVEMBER 1975
NTIS: PB 250 970 EPA-905/9-75-007
(266 pgs.) Price: PC 9.00/MF 2.25
IMPACT OF NON-POINT POLLUTION CONTROL ON WESTERN LAKE SUPERIOR
Red Clay Project-Work Plan EPA-905/9-76-002
NTIS No. PB 255 293 Price: PC 7.50/MF 3.00
CONFERENCE ON MUSKEGON COUNTY, MICHIGAN WASTEWATER TREATMENT
SYSTEM -9/17-18/75 EPA-905/9-76-006
NTIS No. PB 263 552/AS Price: PC A09/MF A01
WASHINGTON COUNTY PROJECT
Work Plan EPA 905/9-77-001 (Jan ?77)
NTIS No. PB 264 189 Price: PC A05/MF A01
BEST MANAGEMENT PRACTICES FOR NON-POINT SOURCES POLLUTION CONTROL SEMINAR
1976-Nov. 331 pgs. EPA-905/9-76-005
NTIS No. PB 265 731/owp Price: PC A15/MF A01
ENVIRONMENTAL IMPACT OF LAND USE ON WATER QUALITY
Progress Report-1976 EPA-905/9-76-004
NTIS No. PB Price: PC /MF
ENVIRONMENTAL IMPACT OF LAND USE ON WATER QUALITY-Summary Final Report
October 1977 95 pgs EPA 905/9-77-.007-A
NTIS No. PB 278 187/AS PC A06/ MF A01
-------
C.'l-
October, 1977 EPA-905/9-77-007-B
ENVIRONMENTAL IMPACT OF
LAND USE ON
WATER QUALITY
Final Report
on the
Black Creek Project
(Technical Report)
by
James Lake
Project Director
James Morrison
Project Editor
Prepared for
U.S. ENVIRONMENTAL
PROTECTION AGENCY
Great Lakes National Program Office
230 South Dearborn Street
Chicago, Illinois 60604
Ralph G. Christensen Carl D. Wilson
Section 108a Program Project Officer
UNDER U.S. EPA GRANT NO. G005103
to
ALLEN COUNTY SOIL & WATER
CONSERVATION DISTRICT
U.S. Department of Agriculture, SCS, ARS
Purdue University, University of Illinois
-------
DISCLAIMER
This project has been financed (in part) with Federal funds
from the Environmental Protection Agency under grant number
G-005103. The contents do not necessarily reflect the views
and policies of the Environmental Protection Agency, nor does
mention of trade names or commercial products constitute
endorsement or recommendation for use.
-------
TABLE OF CONTENTS
SECTION I
Introduction
1 . 1 HISTORY ................................................. 1
1 . 2 PROGRAM PARTICIPANTS .................................... 3
1.2.1 Environmental Protection Agency ......................... 3
1.2.2 Allen County SWCD ....................................... 3
1.2.3 Soil Conservation Service ............................... 3
1.2.4 Purdue University ....................................... 3
1.2.5 University of Illinois .................................. 4
1.2.6 Additional Assistance ................................... 4
1 . 3 CONCEPT OF THE PROJECT .................................. 4
SECTION II
Conclusions
2.1 FECAL POLLUTION OF BLACK CREEK .......................... 8
2 . 2 TOXIC SUBSTANCES ....................................... 9
2.3 FISH [[[ 9
2.4 CONSERVATION PRACTICE INSTALLATION AND WATER QUALITY ____ 10
2 . 5 NUTRIENT AND SEDIMENT TRANSPORT ............. . ........... 1Q
2 . 6 STREAM CHANNEL STABILITY ................................ 12
2 . 7 SEDIMENT POND ........................................... 12
2.8 IN CHANNEL DESILTING BASIN .............................. 13
2 . 9 TILLAGE ................................................. 13
2 . 10 TILE DRAINAGE , . ......................................... 14
2.11 WATER QUALITY BASED ON GRAB SAMPLE DATA ................. 15
2 . 12 ALGAL BIOASSAY STUDY .................................... 16
2.13 LABORATORY INCUBATION STUDIES ........................... 16
2 . 14 RAINULATOR .............................................. 17
2 . 15 WATERSHED MODEL ......................................... 18
2 . 16 REMOTE DATA ACQUISITION ................................. 18
2 . 17 SOCIO-ECONOMIC CONCLUSIONS .............................. 18
2.18 LAND MANAGEMENT ......................................... 19
2 . 19 ADMINISTRATION .......................................... 20
SECTION III
Technical Approach
3 . 1 SELECTION OF BLACK CREEK ................................ 21
3.1.1 Selection Criteria ...................................... 21
3.1.2 Comparison of Area, Maumee Basin ........................ 21
3.1.3 Land Capability Units ................................... 40
3 . 2 DESIGN OF LAND TREATMENT ................................ 47
3.2.1 Role of SCS ............................................. 47
-------
11
3 . 3 TECHNICAL APPROACH 51
3.3.1 Sampling 51
3.3.2 Laboratory Analysis 66
3.3.3 Data and Handling 76
SECTION IV
Study Results
4 .1 SIMULATED RAINFALL 87
4.1.1 Equipment 87
4.1.2 objectives !!!!!!!!!!!!!!!!!!!!!!!! 87
4.1.3 Summary of Results I!.!!!".!!.*!!!!!! 89
4 . 2 BANK STABILITY STUDIES .'..'.'..'..'.'...'..'. 116
4.3 SEDIMENT BASINS !!!!!!!!!!!!!!!!!! 118
4 . 4 MICROBIOLOGICAL STUDIES 120
4.4.1 Bacterial Counts at Low Flow '..'.... 12°
4.4.2 Contamination at Low Flow 121
4.4.3 Bacterial Counts, Maumee, Wann Ditch 125
4.4.4 Bacterial Counts in Tile Drainage Water 126
4.4.5 Fecal Coliform/Fecal Strep Ratios 127
4.4.6 Biochemical Oxygen Demand 128
4.4.7 Fish Kill Caused by Organic Pollution '..'.'....'.'.'.'.'.. 129
4.4.8 Pesticides and Heavy Metals 130
4 . 5 FISH STUDIES 132
4.5.1 Fish Sampling !!!!! 132
4.5.2 Habitat Structure .'....'.'...'......'......'... 132
4.5.3 Species Composition and Distributions 134
4.5.4 Seasonal Changes in Fish Communities 137
4.5.5 Habitat, Fish Community Diversity 138
4.5.6 Effects of Channel Modification '.[[ 138
4.5.7 Stream Disturbance and Fish Communities 140
4.6 TILE DRAINAGE .'..'..'.. 142
4.6.1 Initial Tile Sampling Program 142
4.6.2 Tile Monitoring Continued Program 144
4.6.3 Results of Automatic Tile Sampler 145
4.6.4 Pesticide and Herbicide Response 148
4.6.5 Stream Grab Sampling Program 148
4 . 7 TILLAGE STUDIES .'........ 152
4.7.1 Hoytville Silty Clay Loam * [ 155
4.7.2 Nappanee Silt Loam 156
4.7.3 Whitaker Loam 157
4.7.4 Morley Silt Loam !!!!!!!!!!!!!! I58
4.7.5 Haskins Loam 159
4.7.6 Chemicals Applied to Replicated Plots 159
4.7.7 Tillage Demonstrations 160
4 . 8 SOCIO-ECONOMIC STUDIES 161
4.8.1 Model: Sociological 162
4.8.2 Model: Economic 165
4.8.3 The Data Base !!!!!!!!! i66
4.8.4 Sociological Conclusion 173
4.8.5 Economic Conclusion 175
-------
Ill
4.8.6 Policy Implications 176
4 . 9 MODELING 177
4.9.1 ANSWERS 177
4.9.2 Tile Drainage Simulation Model 203
4 .10 STUDIES OF NUTRIENT AVAILABILITY 212
4.10.1 Effect of Incubation Temperature 212
4.10.2 Effect of Aeration Status and Shaking 213
4.10.3 Effect of Calcium Carbonate Addition 215
4.10.4 Addition of Soluble Inorganic Phosphorus 217
4.10.5 Addition of Soluble Inorganic Nitrogen 217
4.11 ALGAL STUDIES 221
4.11.1 Introduction 221
4.11.2 Materials and Methods 222
4.11.3 Results and Discussion 223
4 .11. 4 Conclusions 227
4 .12 STUDIES OF MAUMEE BASIN 229
4 .13 CONSTRUCTION, LAND USE DIFFERENCES 234
SECTION V
Discussion
5 .1 IMPACT OF LAND TREATMENT 237
5.1.1 Analysis of Costs 238
5.1.2 Arriving at BMP's 245
5.1.3 Engineering Observations 250
5.2 SOURCES OF SEDIMENT AND RELATED POLLUTANTS 252
5.2.1 Comparison of Subwatersheds 252
5 . 3 LOCAL VARIATIONS IN WATER QUALITY 272
5.3.1 Water Quality During a Drought 272
5.3.2 Effects of Channel Morphology 275
5.3.3 Variations in Water Quality 277
5.3.4 Effects of Construction Activities 279
5.3.5 Effects of Conservation Practices 279
-------
IV
LIST OF FIGURES
Figure 1 Maume Basin Map . . , , 27
Figure 2 Soil Associations - Maumee Basin . , , 28
Figure 3 Black Creek Watershed ,..,..,.., 34
Figure 4 Location of Study Area , 35
Figure 5 Soil Associations - Black Creek Watershed , 38
Figure 6 Land Capability Units - Black Creek Watershed . . , 42
Figure 7 Purdue Monitoring Sites ,.49
Figure 8 Illinois Monitoring Sites ..,,,,,..., , 58
Figure 9 Automatic Pumping Sampler , ... 65
Figure 10 Comparison Suspended Solids . ,, 73
Figure 11 Comparison of Ammonium N . . . , , 74
Figure 12 Comparison of Nitrate N 75
Figure 13 Configuration of ALERT Hardware ,,.,,., , ,.80
Figure 14 Data Processing Scheme ,,,,,,,,,,,. ,,.,......86
Figure 15 Flow Regimes in Watershed ,,,,.,,.,,.,,., 122
Figure 16 Frequencies of Levels of Contamination ,..124
Figure 17 Flow and Sediment Hydrograph for Automatic Tile
Sampler .,..,,.,,., ,,.,.., , ,..147
Figure 18 Nitrate Concentration vs Time for Weekly Grab
Samples ,..,,.,,.,,. , . . , 150
Figure 19 Suspended Solids vs. Stage , ..,,.., 151
Figure 20 Black Creek Sociological Model , , , . . , , 164
Figure 21 Surface and Subsurface Water Movement Relationships 182
Figure 22 Watershed Divided into Elements with Channel
Elements Shaded . . , ,..,,.. , 184
Figure 23 Partitionin of Overland Flow , 184
Figure 24 Transport Relationships Used in ANSWERS Model .,..,.191
Figure 25 Sediment Loss , ,,.,,....,, 197
Figure 26 Upper Black Creek Watershed. Local Net Sediment Lossl99
Figure 27 Effect of BMP
-------
LIST OF TABLES
Table 1 Land Capability Comparisons - Maumee Basin and Black
Creek Study Area 23
Table 2 Land Use Comparisons - Maumee Basin and Black Creek
Study Area 2J
Table 3 Soils Data By Land Capability - Black Creek Study
Area 46
Table 4 Land Treatment Goals and Estimated Installation Costs..4**
Table 5 Degree of Difficulty for Black Creek Conservation
Goals 50
Table 6 University of Illinois Analytic Methods 67
Table 7 Detection Limits (micrograms metal/ml water) 68
Table 8 Detection Limits in Micrograms Metal/Gram Dry Sediment.69
Table 9 Recovery of Compounds 71
Table 10 Erosion Losses 89
Table 11 Soil Loss As Aggregates 90
Table 12 Raindrop Energy vs. Surface Runoff 91
Table 13 Losses After Various Tillage Systems 91
Table 14 Losses After Various Tillage Systems 91
Table 15 Description of Experimental Conditions 93
Table 16 Nutrient Concentrations in Animal Wastes 94
Table 17 Amounts of Nutrients Added in Waste Applications
to Three Experimental Sites 94
Table 18 Losses of Sediment and Nutrients in Runoff From
Nappanee Soil 96
Table 19 Concentrations of Solids and Nutrients from Nappanee
Soil 97
Table 20 Proportions of Added Nutrients in Swine Waste Lost
in Surface Runoff 98
Table 21 Losses of Sediment and Nutrients From Morley Soil 99
Table 22 Concentrations of Solids and Nutrients From Morley
Soil 100
Table 23 Proportions of Nutrients Added Lost in Surface
Runoff From Morley Soil 101
Table 24 Losses of Sediment and Nutrient in Runoff From
Overgrazed Pasture 102
Table 25 Concentrations of Solids and Nutrients From an
Overgrazed Pasture 1°2
Table 26 Proportions of Added Nutrients Lost in Surface
Runoff From an Overgrazed Pasture 1°3
Table 27 Characteristics of Soils Used in the Investigation 104
Table 28 Losses of Soil Phosphorus Components in Surface
Runoff 1°5
Table 29 Losses of Nitrogen Components in Surface Runoff 1°6
Table 30 Percentage Distribution of Nitrogen and Phosphorus 1°8
Table 31 Percentage Nutrients 109
Table 32 Added Fertilizer Nitrogen and Phosphorus Lost in
Runoff HO
Table 33 Effect of Fertilization HI
Table 34 Analysis of Soil Size Fraction 114
-------
VI
Table 35 Percent Sand and Sizes in Desilting Basin Deposits
at Stations Listed 119
Table 36 Bacterial Counts From Surface Water Station 312,
May 31, 1977 120
Table 37 Stream Discharge Class By Date for Two Sampling Sites..121
Table 38 Total Coliform, Fecal Coliform and Streptococcus
Count - Low Discharge 121
Table 39 Surface Water Stations According to Stream Discharge
and Organic Pollution 123
Table 40 Coliform Counts at High Stream Discharge 125
Table 41 Coliform Counts at High Stream Discharge 125
Table 42 Coliform Counts for 6 Septically Polluted and 14
Non-septically Polluted Outlets I27
Table 43 The Number of FC/FS Observed and Expected FC/FS
Determinants 127
Table 44 BOD During Ascending Flow of Storm Event 128
Table 45 BOD and Ammonium N Concentrations During Fish Kill I30
Table 46 Pesticide and Heavy Metal Concentrations I30
Table 47 Summary of Pesticide and Heavy Metals Detected and
Not Detected 151
Table 48 Categories for Habitat Analysis i33
Table 49 Fish Species Collected in Black Creek i35
Table 50 Fish Community Density During April I38
Table 51 Percent Change - Number Species 4 Areas 141
Table 52 Summary of Initial Tile Sampling Program 142
Table 53 Concentration of Sediment and Nutrients in Tile
Effluent ^43
Table 54 Sediment and Nutrient Concentrations by Ditch - 1974 i43
Table 55 Values for Exponential Fit - Tile Flow and Sediment-
Nutrient Components 144
Table 56 Concentration of Sediment and Nutrients for 20 Tile
Sites 145
Table 57 Sediment and Nutrient Losses from Tile Effluent -
Automatic Sampling Site 145
Table 58 R-Squared Values for the Automatic Tile Sampling Site..146
Table 59 Concentration of Tile Effluent from Automatic
Sampling Site by Year 148
Table 60 Prediction of Loading From Complete Data Base 149
Table 61 Linear Correlation Results for Sites 2 and 6 154
Table 62 Exponential Correlation Results for Site 2 and 6 155
Table 63 Corn Response to Tillage System 155
Table 64 Soybean Response to Tillage System 156
Table 65 Corn Response to Tillage System 157
Table 66 Soybean Response to Tillage System 157
Table 67 Corn Response to Tillage System 158
Table 68 Soybean Response to Tillage System 158
Table 69 Corn and Soybean Response to Tillage System 158
Table 70 Corn Response to Tillage System 159
Table 71 Soybean Response to Tillage System 159
Table 72 Costs for Chemicals, Corn and Soybean 16°
Table 73 Corn Response to Tillage, Roger Ehle Farm 160
Table 74 Chemicals Applied at Planting 161
-------
Vll
Table 75 Sociological Model Variables
Table 76 Primary Tillage Practices
Table 77 Computer Coefficients - 370 Acre Farm
Table 78 Computer Coefficients - 580 Acre Farm
Table 79 Average Annual Soil Loss Coefficient
Table 80 Average Corn Yield and Moisture Coefficients
Table 81 Average Soybean Yield
Table 82 Production Costs
Table 83 Net Revenue, Returns, Slopes Greater Than Two Percent..
Table 84 Interception Constants Recommended by Horton
Table 85 Typical Surface Storage Coefficients
Table 86 Sediment Particle Characteristics
Table 87 Effect of Temperature on the Concentration of Soluble
N and P
Table 88 Effect of Aeration and Shaking on the Concentration
of Soluble N and P
Table 89 The Effect of Calcium Carbonate on the Level of
Soluble N and P
Table 90 Effect of SIP Addition on the Level of Soluble
Phosphorus
Table 91 Effect of Ammonium Nitrate Addition on the Concen-
tration of Soluble N
Table 92 Availability to Algae of Soluble Phosphorus in Stream
Water
Table 93 Variation of Available Soluble Phosphorus by
Adjoining Land Use Characteristics
Table 94 Proportion of Sediment Phosphorus Immobilized by Cells.
Table 95 Proportion of Sediment Inorganic Phosphorus Immobilized
Table 96 Proportion Immobilized by Cells Originating from
Certain Extractable Fractions
Table 97 Variation in Sediment Inorganic P Immobilized by
Source of P
Table 98 10-Year Annual Precipitation Sediment Yield and
Discharge
Table 99 Differing Parameters, Four Watersheds
Table 100 Sediment Concentration Indexes - Subwatersheds
Table 101 Cost by Practice
Table 102 Technical Assistance Cost SCS
Table 103 Cost of BMP Installation
Table 104 Estimated BMPs Needed at Beginning of Black Creek
Project
Table 105 Projected Costs of Ideal Black Creek Land Treatment....
Table 106 Estimated Costs - Maumee Basin
Table 107 Evaluation of Practices Applied in Black Creek Water-
shed
Table 108 Characteristics of the Areas Studied
Table 109 Rainfall and Runoff Amounts and Yields of Sediment
and Nutrients
Table 110 Nutrient Transport (kg/ha)
Table 111 Percent of Nitrogen Forms Transported
Table 112 Percent of Phosphorus Forms Transported..
Table 113 Partitioning of Runoff, Transported Sediment and
Nutrients - Smith-Fry Drain
165
165
167
168
170
171
172
173
175
186
187
192
213
214
216
218
219
224
225
226
226
227
228
229
234
235
237
238
242
243
243
244
247
255
257
258
259
259
260
-------
Vlll
Table 114 Partitioning of Runoff, Transported Sediment and
Nutrients - Dreisbach Drain 260
Table 115 Smith-Fry Drain - Sediment and Nutrient Concentrations
(1975) 261
Table 116 Smith-Fry Drain - Sediment and Nutrient Concentrations
(1976) 262
Table 117 Dreisbach Drain - Sediment and Nutrient Concentrations
(1975) 262
Table 118 Dreisbach Drain - Sediment and Nutrient Concentrations
(1976) 262
Table 119 Sources of Runoff, Transported Sediment and Nutrients
Smith-Fry Drain (1975) 265
Table 120 Sources of Runoff, Transported Sediment and Nutrients
Smith-Fry Drain (1976) 265
Table 121 Sources of Runoff, Transported Sediment and Nutrients
Dreisbach Drain (1975) 266
Table 122 Sources of Runoff, Transported Sediment and Nutrients
Dreisbach Drain (1976) 266
Table 123 Yearly Concentrations, Transported Sediment &
Nutrients - Smith-Fry Drain (75-76) 267
Table 124 Yearly Concentrations, Transported Sediment &
Nutrients - Dreisbach Drain (75-76) 267
Table 125 Yearly Concentrations of Total N and P Attached to
Sediment (1975-76) 268
Table 126 Estimate of Annual Applied Fixed Nitrogen 269
Table 127 Estimate of Percent Applied and Fixed Nitrogen Lost.... 269
Table 128 Characteristics, Groundwater, Channel Flows, Black
Creek 273
Table 129 Stations with Water - Dreisbach Drain 275
Table 130 Water Quality Characteristics - Manor Segments 278
-------
1.1 HISTORY
The problem of nonpoint source pollution of the nation's streams and
lakes has become a subject of increased attention because of the require-
ments of Section 208 of Public Law 92-500 for water quality management
planning.
Prior to the final passage of this act, the question of "agricultural
pollution" had been discussed in depth during a one-day conference on the
Maumee River held in Fort Wayne, Ind. in early January of 1972. This
conference, sponsored by Rep. J. Edward Roush (Ind.-4), provided one of the
first public discussions of the nonpoint source pollution problem in
northeastern Indiana and the Maumee River Basin.
As a result of the Maumee River Conference, local, state and federal
officials, representatives of agricultural and environmental protection
groups, and other interested citizens undertook a series of discussions of
the problems of erosion and sediment related pollution of the Maumee River
and ultimately Lake Erie. These discussions were arranged by the Allen
County Surveyor, William Sweet, and by then State Conservationist Thomas
Evans of the Soil Conservation Service, USDA.
As the result of these discussions, a proposal to study the problems
of agricultural pollution of the Maumee River was developed by the Allen
County Soil and Water Conservation District.
The proposal called for the cooperation of the District, Purdue
University and SCS in a demonstration project, supported by research. This
proposal, entitled "A Proposal for the Reduction of Sediment and Related
Pollutants in the Maumee River and Lake Erie," was submitted to Region Five
of the Environmental Protection Agency in June of 1972.
The proposal was assigned to the Office of Great Lakes Coordinator
where funding was available under Section 15 of the water quality amend-
ments of 1970. Section 15 became Section 108(a) with the adoption of the
1972 amendments. The Section 108 program, administered from the Region V
EPA office, is a special program created by Congress to demonstrate means
of improving water quality within the Great Lakes Basin. The act calls for
federal participation of 75 percent of the total cost for demonstration
projects. Research is funded under the program in support of the demons-
tration effort.
In October of 1972, almost simultaneously with the adoption of Public
Law 92-500, a six-month planning grant was awarded to the Allen County Soil
and Water Conservation District to design, in detail, the demonstration ef-
fort and related research.
This work was completed and is reported in Environmental Impact of
Lan<3 Use on Water Quality -- A Work Plan. Some of the descriptive material
about the area chosen for the study and its relation to the Maumee Basin
and to Lake Erie is reported in this volume.
This document constitutes the final technical report on the five years
of effort, beginning with the awarding of the planning grant and ending
with the close of the project, Oct. 16, 1977. It is one of four volumes
designed to provide a complete description of the work accomplished and the
INTRODUCTION
-------
results obtained during the course of the project. Other volumes include a
non-technical summary volume, a data volume in which information collected
during the project is summarized, and a volume made up of papers prepared
by individual researchers and administrators who worked on the project.
Section 1.2 lists a large number of individuals and organizations who
have been associated with the Black Creek project. Some observations about
the success with which this diverse group has operated and the ability of
this project to relate to individual landowners within the project area are
in order.
It is important to observe that the Black Creek project was developed
by interested persons in the local community and was not imposed from out-
side. The specific sequence of events was as follows:
(1) A problem was identified by speakers participating in an open
public meeting (the Maumee Conference).
(2) Interested governmental officials and private citizens discussed
the problem in a neutral forum. At this point, there was no implied
pressure to take any specific action, although a realization that a
problem as serious as this one had been described was likely to pro-
duce governmental action was present.
(3) Funding to attack the problem was made available by EPA.
(4) A program satisfying diverse interests was proposed.
(5) Funding was made available for a detailed plan for a program
which involved local control through the sponsorship of the Soil and
Water Conservation District.
(6) The program was funded and has been carried out.
Particular attention was paid to the need for keeping the general pub-
lic and affected landowners involved in the project. The Maumee Conference
was a public meeting. Although discussions at the Sweet-Evans meetings
were not extensively covered in the local press, they were not closed meet-
ings, and all interested citizens were allowed to attend.
The processes of proposal submission, revision, and final acceptance
was reported in the press. In addition, special meetings were held with
affected landowners, and contact was maintained by face-to-face communica-
tion and by letter.
As a result, rumors and hostilities that could have developed were
largely avoided. At the close of the project, both environmental and agri-
cultural interests remained supportive of the concept and the project.
Early in the conduct of the project, it was decided to hold monthly
"steering committee" meetings during which representatives of involved
agencies would be present. These meetings included both administrative and
research personnel. Minutes were prepared and distributed. Discussion was
open and covered all aspects of the project. As a result, communication
among researchers and communication between administrators and researchers
was excellent, and the project had a better focus than might otherwise have
been the case.
INTRODUCTION
-------
1.2 PROGRAM PARTICIPANTS
The success of the Black Creek project has depended on several key
elements. Most important has been the coordination of the efforts of
diverse administrative and technical interests in the project. Primarily,
agencies involved included the Environmental Protection Agency, which pro-
vided funding, the Allen County Soil and Water Conservation District, which
assumed overall responsibility for the project, the Soil Conservation Ser-
vice of USDA, which provided technical assistance, Purdue University and
the University of Illinois, which provided research support, and units of
local and state government, which provided needed assistance.
Key personnel from these agencies are listed in the following discus-
sion:
1.2.1 Environmental Protection Agency
Ralph Christensen, director Section 108 Programs, Great Lakes National Pro-
gram Office, Region V EPA.
Carl Wilson, project officer, nonpoint source coordinator, Planning Branch,
Region V EPA.
1.2.2 Allen County SWCD
Chairmen of the Board of Supervisors Ellis MacFadden, Roger Ehle.
Members of the Board of Supervisors Ellis MacFadden, Roger Ehle, Mic
Lomont, Eric Kuhne, John Hilger, Ray Arnold, Gilbert Whitsel.
Assistant Supervisor Don Rekeweg.
Employees of the Board of Supervisors:
James Lake project director
Dan Dudley aquatic biologist
Rex Journay tillage research
John Pidlisny technician
Alan Shupe technician
1.2.3 Soil Conservation Service
Dan McCain District Conservationist
John Dennison area technician
C. F. Polland area engineer
Planners and technicians Greg Woods, Gary Carlile, Doene Goettl, Bill
Howard, Stan Steury, Darrell Brown.
Area Conservationists Joe Branco, Ken Pyle
State Office Personnel Leon Kimberlin, Eugene Pope, Max Evans, Roy Ham-
ilton, Robert Bollman.
State Conservationists Thomas Evans, Cletus Gillman, Buell Ferguson.
1.2.4 Purdue University
Rolland Z. Wheaton, project coordinator, studies of channel stability and
the effectiveness of sediment basins
Richard E. Land, field coordinator
Jerry V. Mannering, simulated rainfall research
Don Griffith, tillage trials
Harry Galloway, tillage trials
Darrell Nelson, chemistry
INTRODUCTION
-------
Dr. Lee Sommer, chemistry
Edwin J. Monke, data handling and modeling
Larry F. Huggins, watershed simulation and data acquisition
Jack Burney, watershed modelling
Stephen Mahler, data acquisition
Ralph Brooks, socio-economic studies
William Miller, socio-economic studies
William McCafferty, aquatic biology
Jerry Hamelink, aquatic biology
James B. Morrison, project editor and technical writer
Dagmar Clever, editorial assistant
Graduate Instructors David Beasley, Adelbert Bottcher
Administrative Personnel Howard Diesslin, Director Indiana Cooperative
Extension Service; Bernard J. Liska, Director of the Indiana Experiment
Station; Ellsworth Christmas, Assistant Director of Cooperative Extension
Service; Gerry Isaacs, head, Department of Agricultural Engineering; Marvin
Phillips, head, Department of Agronomy; Paul Farris, head, Department of
Agricultural Economics.
1.2.5 University of Illinois
James Karr, aquatic biology and water quality
Illinois Natural History Survey Laboratory, sample analysis
1.2.6 Additional Assistance
William Sweet, Allen County Surveyor
William Jones, Allen County Highway Department
Elias Saamon, Northeastern Indiana Regional Coordinating Council
Ernest Lesiuk, Allen County Cooperative Extension Agent
Fort Wayne-Allen County Board of Health
Allen County Data Processing
Allen County Commissioners
Allen County Council
Agricultural Research Service, USDA, C.B. Johnson, technician
1.3 CONCEPT OF THE PROJECT
The Black Creek project has been misinterpreted by some observers as
an attempt to have a direct, measurable impact on water quality in the Mau-
mee River and Lake Erie. This goal, while a worthy one, has never been the
aim of the Black Creek project. In fact, it is doubtful that any measur-
able impact on water quality in the river and the lake could be obtained if
all of the runoff water from the 12,038-acre Black Creek area were diverted
from the system.
Instead, the basic concept of the Black Creek project has been as a
model. In fact, the Black Creek Watershed is similar to 200 to 300 agri-
cultural watersheds in the Maumee Basin. The question has been whether
techniques demonstrated in this watershed, if applied on a basin-wide
basis, would likely improve the quality of water in the Maumee River and
Lake Erie.
There have been two major thrusts simultaneously underway in the
watershed. The first has been the demonstration effort itself, a project
INTRODUCTION
-------
designed to demonstrate the effectiveness of traditional and innovative
soil and water conservation techniques for improving water quality. The
second has been the research effort, directed at understanding the mechan-
isms whereby the demonstrated techniques succeed or fail.
Initial project goals stated that it was desirable to know if a con-
centrated application of existing methods of land treatment in the Maumee
Basin could achieve a desired reduction in sediment. It was also desired
to estimate how much such a program would cost on a basin-wide basis.
Concurrently, the project was designed to discover what kind of pro-
gram might be carried out on a basin-wide basis which would convince indi-
vidual landowners to apply conservation practices for the improvement of
water quality. Specifically, it was asked whether this can be adequately
done on an incentive basis or whether some type of mandatory control of
pollution from nonpoint sources might be imposed with a reasonable chance
of success.
The purpose of the research conducted in conjunction with the demons-
tration project was to more fully understand the mechanisms whereby the
demonstration project could reduce the sediment load entering the Maumee
River and to utilize this understanding to project to the Maumee Basin an
accurate estimate of methods that can be employed to achieve a desired im-
provement in water quality and the costs of doing so.
Robert Schneider, EPA Great Lakes Coordinator summarized EPA's in-
terest in the Black Creek Project in light of additional developments over
the past five years, in a meeting at Fort Wayne, Ind. Oct. 26, 1977. His
remarks included the following observations:
"When the Environmental Protection Agency first authorized the project
back in May of 1973, the Agency was seeking guidance for the nonpoint
source pollution control program specifically as it relates to water quali-
ty. Under the provisions of the Federal Water Pollution Control Act Amend-
ments of 1972 (Public Law 92-500), the United States Environmental Protec-
tion Agency was given the primary responsibility for carrying out the
Federal role in water pollution control under section 208 of this Act.
Nonpoint source pollution control requires the development and implementa-
tion of area-wide waste treatment management plans that include procedures
to reduce pollution from nonpoint sources.
"We are acutely aware that this requirement introduces a new aspect to
our nation's pollution control efforts. Historically, pollution control
efforts have concentrated on collection and treatment of wastes that are
discharged from point sources. Nonpoint sources of pollution are not amen-
able to traditional methods of collection and treatment. Thus the approach
recommended is prevention of pollution before it occurs, EPA has gone on
record as advocating "Best Management Practices." The Black Creek Project
has the opportunity to answer this challenge.
"What was not anticipated, as the Black Creek Project progressed, was
a General Accounting Office report on Agricultural programs. I would like
to quote part of this report to you:
INTRODUCTION
-------
xlf the United States is to continue to meet its own food needs
and help alleviate world food shortages, it must maintain its top
soil. Estimates of soil losses for 283 farms GAO visited on a
random basis in the Great Plains Corn Belt, and Pacific
Northwest, indicate that topsoil losses are threatening continued
crop productivity.
"Soil scientists estimate that annual soil losses must be limited
to no more than 5 tons per acre in deep soils and one ton per
acre in shallow soils to maintain soil fertility and productivity
over time. According to Department of Agriculture technicians,
about 84 percent of the 283 farms were losing over five tons per
acre per year on cropland for which calculations were made. In
addition, soil erosion was creating water and air pollution and
highway maintenance problems.
"Agriculture's Soil Conservation Service and Agricultural Stabil-
ization and Conservation Service administers technical and finan-
cial assistance programs costing several hundred million dol-
lars annually designed to help farmers control erosion and
preserve topsoil.
NThese programs have not been as effective as they could be in
establishing enduring soil conservation practices and reducing
erosion to tolerable levels.'
"We in Environmental Protection Agency hope the Black Creek Project
results will help to resolve the issues in the General Accounting Office
report to Congress. To this end, let me state some of the major EPA expec-
tations from the Black Creek Project:
"1. That the study will develop NBest Management Practices.1
"2. That from a given set of practices, a specific level of wa-
ter quality can be expected.
"3. That a model institutional mechanism to implement a nonpoint
source program will be developed.
"4. That successful projects will be clearly identified and that
the ones that have failed will be similarly identified.
"5. That economic data will be developed that evaluates best
management practices in reference to water quality.
"6. That the social impacts from the study will be clearly iden-
tified.
"7. That best management practices will be developed so as to
have a beneficial effect on the biological system of a stream.
"8. That guidelines for cost sharing will be developed.
INTRODUCTION
-------
"These, then, are the highlights of EPA's expectations from this pro-
ject. I suggest that in the discussions over these two days, you give con-
siderations to these expectations, and any others you believe merit atten-
tion."
References
1. Environmental Impact of Land Use on Water Quality A Work Plan;
EPA Region V, 1973.
2. General Accounting Office, To Protect Tomorrow's Food Supply, Soil
Conservation Needs Priority Attention; 1977.
3. Schneider, Robert, "EPA Expectations of Black Creek" speech, Fort
Wayne, Indiana, 1977.
CONCLUSIONS
-------
II. CONCLUSIONS
The Black Creek demonstration and research project led to several gen-
eral conclusions which are reported in this section. Additionally, more
specific conclusions are included in the various sections of this document.
The conclusions reported in this section represent a consensus of the in-
vestigators and administrators in the Black Creek project. The order in
which they are presented is not intended to reflect the degree of impor-
tance with which the conclusion or subject is viewed.
2.1 FECAL POLLUTION OF BLACK CREEK
(See Section 4.4)
Fecal contamination and water quality criteria have significant impli-
cations for the swimmable fishable goal of public law 92-500. If that goal
is defined to mean water clean enough to allow whole body contact recrea-
tion, then fecal coliforms should not exceed 200 per 100 ml. This value
represents a threshold value above which the frequency of water borne human
pathogens greatly increases. Under current regulations for the Maumee
River and Black Creek, the standard is a weighted average of no more than
1,000 fecal coliforms per 100 ml, and a peak value of no more than 2,000
per 100 ml. These standards, in effect, regulate the water quality at a
level which insures only limited body contact recreation, such as fishing
and boating.
1. Fecal coliforms for the Black Creek in general exceed the water quality
standards for whole body contact (200 per 100 ml).
2. Fecal coliform concentrations were not demonstrated to surpass the ex-
isting Maumee River standard in the portions of Black Creek not influenced
by discharges (septic tank effluent) from the town of Harlan (pop. 600).
However, these standards were exceeded in portions of Black Creek influ-
enced by the town of Harlan.
3. Livestock sources of fecal pollution were detected during a runoff
event in March, 1977, in areas with unconfined livestock operations (pas-
ture and barnlot).
4. High flows carried higher concentrations of bacteria than low flows as
a result of two phenomena:
(1) contamination from human sources is flushed from fecally contam-
inated bottom sediment in channels and/or tile lines during high
flows, and
(2) contamination from livestock sources is carried by surface flow
to stream channels.
5. Biotic communities of stream reaches below septic tank outfalls and
below livestock waste discharges were affected in the following ways:
(1) Only the most pollution tolerant forms are present in areas of
long-term, gross pollution (at septic tank outfalls).
CONCLUSIONS
-------
(2) Modified invertebrate species composition and abundances are
recorded in areas of long-term, moderate to low levels of pollution
downstream of septic tank outfalls.
(3) The effect on aquatic life of short-term, moderate to high levels
of pollution during some runoff events has not been established.
(4) Short-term, gross pollution caused by a discharge of manure slur-
ry from an animal waste holding pit resulted in an extensive fish kill
(8 km of stream) and a complete disruption of the aquatic community.
6. In the vicinity of septic tank outfalls, aquatic communities are ad-
versely affected.
2.2 TOXIC SUBSTANCES
(See Section 4.4.8)
1. Spot checks of levels of contamination by a variety of toxic substances
(pesticides, PCB's, and heavy metals) in water, sediment, and fish tissue
show these to be at levels below those of concern for human health.
2.3 FISH
(See Section 4.5)
1. Thirty-five species of fish have been collected in the Black Creek
watershed. However, this number of species does not necessarily reflect
the decline in value of the fishery resource in Black Creek.
2. Many individuals and some species migrate into the Black Creek
watershed each year in search of spawning and nursery grounds.
3. Others reside permanently in the watershed, especially in more protect-
ed areas associated with forested sections of streams.
4. Fishes in unprotected areas tend to suffer high mortality and depend on
replacement from outside areas each year. Unprotected areas include re-
gions of straight stream channels and areas not shaded by overhanging vege-
tation. Typically they will also have uniform bottoms, often of sand or
silt.
5. Even relatively protected areas, such as wooded streams, will experi-
ence massive mortality in unusually dry or cold periods.
6. In Black Creek all areas are recolonized quickly (during the next
spring), even after mass mortality, unless channels are obstructed by con-
struction or other activities. Other activities include channel modifica-
tion which create unsuitable habitat high temperatures, low oxygen con-
tent, algal blooms, etc. However, the recolonization is often futile be-
cause the fishes will soon die in the poor habitat of unprotected areas.
7. The primary factor governing diversity of fish communities is complexi-
CONCLUSIONS
-------
10
ty of habitat. More diverse habitat structure will support fish communi-
ties of high diversity.
8. These results show that it will take more than improved water quality
to insure that waters are "fishable." Management for "fishable and swimm-
able" objective must also consider suitable habitat structure, trophic
structure of community, allocthonous energy inputs, and temperature re-
gimes.
2.4 CONSERVATION PRACTICE INSTALLATION AND WATER QUALITY
(See Section 4.15 and Section 5.3)
The project work plan did not call for detailed study of construction
impacts on water quality. It is generally agreed that erosion is greatly
accelerated by construction activities including the installation of many
permanent practices. (For example, stream bank protection, structures,
grassed waterways and PTO terraces.) The following fundamental points con-
cerning construction and water quality need to be stressed because they
were overlooked in the planning of Black Creek.
1. Construction activities associated with the installation of conserva-
tion measures can result in short term concentrated delivery of sediment to
stream channels and the alteration of biotic communities. Depending on the
magnitude of the construction, on-site and downstream water quality, stream
habitat, and biotic communities are altered for time periods of several
days up to several years.
2. Streams with aquatic resource potential need to be identified and pro-
tected from construction induced damages.
3. Programs that utilize soil conservation practices to improve water
quality demand careful planning and installation of the practices to reduce
sediment delivery to streams and other associated impacts on downstream and
on-site aquatic resources.
4. Planners and field technicians need to be trained in the ecological
principles that are the basis of understanding and recognizing sensitive
aquatic resources.
2.5 NUTRIENT AND SEDIMENT TRANSPORT
(See Section 4.6, Section 5.2, and Section 5.3)
1. Amounts of runoff, sediment, and nutrients discharged from a small
watershed are greatly affected by rainfall. Reductions in rainfall give a
greater percentage reduction in runoff which in turn gives a still greater
percentage reduction in sediment and nutrient transport.
2. During years with above average rainfall, land slope is clearly the
dominant factor affecting sediment yields. However, with below average
rainfall, the effect of land use on sediment yields becomes relatively more
important. This reflects the natural sequence of the rainfall-runoff event
CONCLUSIONS
-------
11
because rainfall must first meet the storage capabilities of the soil and
land surface before runoff begins.
3. With average or above average rainfall patterns, more than 90 percent
of the total phosphorus transported is sediment bound, while only about 50
percent of the total nitrogen transported is sediment bound. Models of P
transport can be based on the erosion sedimentation process. We recommend
that models of N transport not be based on the erosion-sedimentation pro-
cess only.
4. Transport of sediment and nutrients is strongly associated with large
storms which occur only a few times during a year.
5. In order to characterize nutrient and sediment loading from small
watersheds, runoff from large storm events must be well-monitored and, with
few exceptions, automated sampling is required.
6. During snow melt, the transport of soluble nutrients may be dispropor-
tionally high when compared with snow melt runoff.
7. Sediment and sediment-associated nutrient concentrations increase
markedly with large storm events.
8. Average concentrations of sediment and nutrients discharged from the
Black Creek Watershed are in line with measured concentrations in the Mau-
mee River.
9. Losses of soluble inorganic phosphorus, ammonium and nitrate-nitrogen
from the watershed are partially due to the input of these chemical forms
by precipitation.
10. Most of the sediment and sediment-bound nutrients originate in runoff
from the land surface. The discharge of ammonium is also largely associat-
ed with runoff from land surfaces. However, the discharge from septic tank
outlets contributes substantial amounts of soluble inorganic phosphorus
into the streams in the watershed.
11. Sediments in runoff are nutrient enriched. The least enrichment oc-
curs from subsurface drains and the most from septic tank outfalls. The
sediments in Black Creek have an enrichment factor of about three as com-
pared to the uneroded soils in the watershed.
12. Farming techniques common to the 19th century but still used by the
Amish did not seem to improve the water quality in streams from that area
of the Black Creek Watershed in which they reside.
13. The effects of agricultural nonpoint source pollution and point source
pollution on our water resources are sufficiently different that direct
comparisons between them cannot be made and separate objectives for their
evaluation and control are in order.
14. In the absence of surface runoff and tile flow, quality of ground water
flowing in channels downstream changes, demonstrating that stream channel
and riparian environments affect water quality.
15. During most flow conditions, large inputs of sediment and nutrients
CONCLUSIONS
-------
12
occur at BarIan.
16. Areas remote from Harlan generally have lower nutrient and sediment
concentrations than those at Harlan.
17. Water quality characteristic vary significantly over short stream dis-
tances. This emphasizes the importance of careful selection of monitoring
stations for monitoring water quality.
18. A meandering section of stream flowing through a woodlot acted like a
natural sediment trap.
19. Adjacent subwatersheds of similar size, topography and soil type pro-
duce runoff of varying quality due to differences in land use and applica-
tion of conservation practices. These results emphasize the need for de-
tailed experimental analysis of the determinants of water quality.
2.6 STREAM CHANNEL STABILITY
(See Section 4.2)
1. Total stream bank erosion is small although at the site of erosion it
may appear severe. There is no apparent correlation of bank erosion with
adjacent land use. Over 80 percent of the bank erosion was reported in the
areas containing Eel and Shoals type soils.
2. Channel bottom erosion produced unstable channel banks in several
areas. Channel stabilizing structures and bank stabilizing activities have
eliminated many severe problems. Rock drop structures were especially suc-
cessful for this purpose.
3. Seeding of the banks to establish vegetative cover is most successful
where slopes are 2:1 or flatter. The advantage of the flatter slopes jus-
tify the additional land area required to install them. Mulch is necessary
to establish a good seeding. Stone mulch materials are the most success-
ful. The stone mulch has the advantage of not being washed away during
periods of high flow.
2.7 SEDIMENT POND
(See Section 4.3)
1. The sediment pond had a measurable and beneficial impact on water qual-
ity. However, it is recognized that, in the long term, a pond which re-
moves a large quantity of sediment will require either a difficult cleanout
or will become ineffective.
2. In the first two years following construction of the sediment pond an
equivalent amount of sediment to 2.8 t/ha/yr (1.2 T/ac/yr) from the entire
watershed was collected in the pond. Most of this was due to one 50-year
recurrence interval storm which caused unusual erosion on the nearly level
watershed.
3. No measurable increase in sediment was observed from July 1976 through
CONCLUSIONS
-------
13
July 1977. During this period, runoff was small and very little erosion
was occurring on the watershed.
4. About 95 percent of the material reported above was particles less than
50 y m in diameter. This indicates the effectiveness of the pond in trap-
ping particles which control are difficult to control by land treatment
consistent with row crop production and suggests the value of sediment
ponds.
2.8 IN CHANNEL DESILTING BASIN
(See Section 4.3)
1. During the first two years following construction, the basin was ap-
proximately one-half filled due in part to channel construction activities
above the basin and to one greater than 50-year recurrence interval storm.
2. The largest amount of material was of the coarser particles.
3. Turbidity, total solids, and total phosphorus increase as the water
transits the basin during low flow conditions.
4. For continued operation, periodic clean out would be required.
5. Except in unusual circumstances, the use of an in-channel desilting
basin will have negative impacts on water quality and is not recommended.
2. 9 TILLAGE
(See Section 4.7)
1. On soils where conservation tillage for corn is adapted and where pests
are easily controlled, there should be little or no cost for the benefits
gained in erosion control and water quality.
2. Fall chiseling can replace moldboard plowing for corn after corn or
corn after beans, on most Black Creek soils without limiting production,
where weeds can be controlled.
3. Shallow tillage or no-till planting for corn after corn or corn after
beans should not limit production on well or moderately well drained soils
where perennial weeds are not a serious problem.
4. No-till planting for corn into a chemically killed sod should not limit
production, compared to moldboard plowing, where perennial weeds are not a
serious problem.
5. Perennial and herbicide resistant weeds, such as Canada thistle, field
bindweed and morning glory are more likely to be yield limiting factors for
soybeans than for corn with no-plow tillage.
6. Shallow tillage or no-till planting, is likely to lead to more serious
phytophthora root rot problems for soybeans on poorly drained soils, unless
CONCLUSIONS
-------
14
resistant varieties are available.
7. Farmers are not likely to adopt conservation tillage practices until
success can be proven and demonstrated in their area.
8. Some form or forms of conservation tillage that have proven to be
adapted to soils in an area should be a high priority BMP.
9. For corn, where perennial and resistant weeds are a problem, added her-
bicide costs and/or reduced yields with conservation tillage will reduce
profit. With herbicides currently available, the effect of problem weeds
on soybeans under conservation tillage is likely to be more severe than on
corn.
2.10 TILE DRAINAGE
(Based on Grab Samples)
(See Section 4.6)
1. Tile effluent yields small amounts of sediment and sediment-bound nu-
trients.
2. Tile effluent is relatively high in nitrate-N concentrations as com-
pared to these concentrations in the receiving streams.
3. Nitrate concentration in tile effluent increased with intensity of row
crop farming.
4. Septic contamination of tile outlets is common and greatly increases
the soluble inorganic phosphorus and ammonium yields from the Black Creek
Watershed .
5. Sediment and nutrient concentrations are not correlated with tile flow
for most of the tiles monitored.
6. Tile discharge must be well monitored during storm events to get mean-
ingful loading data. Data from grab samples are not sufficient to deter-
mine storm induced loadings.
7. The low degree of correlation emphasizes the inherent weakness with any
weekly grab sampling program especially for small watersheds.
(Based on Automatic Sampler)
(See Section 4.6)
1. BMP's which allow more water to flow through a tile drainage system
will significantly reduce total losses of runoff,sediment and nutrients.
Nitrate-N is the only nutrient which may increase in concentration, but the
CONCLUSIONS
-------
15
greater percentage reduction in runoff yields a lower total loss.
2. Rainfall during late winter and early spring produces a high percentage
of the annual water, sediment and nutrient loses from an agricultural tile
drainage system.
3. A positive correlation (R-squared = 0.1 to 0.86) is shown between sedi-
ment and nutrient concentrations and tile flow.
4. Initial flow following a storm yields higher sediment and nutrient con-
centration levels than equivalent flow levels when discharge is receding.
5. The monitored Hoytville soil is very responsive hydrologically to sub-
surface drainage and appears to have a higher hydraulic conductivity than
indicated in published soil survey reports. This may be due in part by
deep cracking of the soil.
6. Tile flow exhibits a diurnal variation during freeze-thaw periods of
the year.
7. Tile effluent yields an insignificant amount of phosphorus to streams
as compared to surface runoff.
8. Nitrate is the dominant nutrient in tile effluent. Losses of total ni-
trogen were on the order of 5 percent of nitrogen applied as fertilizer for
two rather dry years.
9. Both insecticide and herbicide were detected in the tile effluent
shortly after a surface application. Losses were of the order of .1 per-
cent of applied amounts. Deep cracking is believed to be the reason for
this pesticide movement.
2.11 WATER QUALITY BASED ON GRAB SAMPLE DATA
(See Section 4.6.5)
1. Annual loadings were not accurately predicted by use of weekly grab
samples alone (see Table 60) . Weekly grab sampling provides adequate
description only of base flow conditions .
2. For accurate loading information during storm events, grab sampling
must be accomplished at an interval significantly shorter than the time
response of a watershed.
3. An annual cycle exists for the concentration level of nitrate and to a
much less degree for ammonia. Suspended solid and phosphorus data did not
exhibit this behavior.
4. Total phosphorus and nitrogen concentrations have a positive correla-
tion (R-squared = 0.11 to 0.60) with suspended solids concentrations,
whereas soluble forms of phosphorus and nitrogen are not correlated with
suspended solids (see Table 61) .
5. Suspended solid and sediment bound nutrient concentrations have a posi-
tive correlation (R-squared = 0.092 to 0.41) with stage. Soluble com-
CONCLUSIONS
-------
16
ponents have no correlation with stage.
6. Water quality data for streams from small watersheds exhibits a high
degree of scatter, due in part to the wide variation in flow.
7. There is no simple function relationship between flow rate and suspended
solid or nutrient concentrations.
2.12 ALGAL BIOASSAY STUDY
(See Section 4.11)
1. Maximum algal growth occurred at soluble P levels of approximately 0.1
mg/1.
2. Not all soluble P in Black Creek drainage water was available to algae
in a four day incubation.
3. Phosphorus was the major limiting nutrient for algal growth in Black
Creek drainage water.
4. Addition of micronutrients increased algal growth and apparent P avai-
lability to algae suggesting that the micronutrient content of water limits
algal growth in sediment-free systems.
5. Approximately 20 percent of total sediment-bound P was available to al-
gae in a two week incubation. Samples taken in March and June yielded
similar results.
6. Approximately 30 percent of sediment-bound inorganic P was available to
algae in a two week incubation.
7. The ammonium flouride, sodium hydroxide, and HC1 extractable fractions
of sediment inorganic P all contribute to the pool of available P for al-
gae. The ammonium flouride and sodium hydroxide extractable sediment inor-
ganic P fractions were the most available to algae representing an average
of 43 and 38 percent of P taken up respectively.
2.13 LABORATORY INCUBATION STUDIES
(See Section 4.10)
1. The concentration of nitrate N in incubated Black Creek stream samples
increased with time as a result of mineralization and nitrification.
Treatments which increased microbial growth (high temperatures, aeration,
calcium carbonate amendment, and addition of inorganic N) increased the
rate of mineralization in samples.
2. The concentration of soluble inorganic P in incubated Black Creek
stream samples slowly increased with time. Increased temperatures and
aeration accelerated release of soluble inorganic P from sediment.
CONCLUSIONS
-------
17
2. 14 RAINULATOR
(See Section 4.1)
1. Soil losses from nearly level lake plain soils under fallow conditions
are much less than those of more sloping soils in the watershed. Even on
nearly level soils, significant differences in loss can occur. These are
related to soil structure.
2. Soil erosion is a highly selective process. Sediment shows distinct
clay enrichment and sometimes silt enrichment.
3. Effective measures for reducing erosion on nearly level soils should be
based on prevention of detachment of naturally occurring aggregates.
4. Protecting the soil surface from raindrop impact is one of the most ef-
fective means of minimizing sediment concentrations in runoff.
5. Soil losses are reduced by those tillage systems which leave appreci-
able crop residues on the surface.
6. Fall chiseling after corn, although not as effective as the no-til and
disc treatments, significantly reduces erosion compared to moldboard plow-
ing.
7. None of the conservation treatments are as effective following soybeans
as following corn.
8. Sediment loads can be reduced by sod buffer strips. The effectiveness
of this practice is dependent on composition of sediment; rate and depth of
flow; and vigor and height of sod.
9. Incorporation of applied annual waste will reduce nutrient losses in
runoff.
10. Waste application tends to reduce soil loss because of mulch effect.
11. Sediments eroded from animal waste treated areas is highly enriched
with nutrients because of manure particles in transported solids.
12. Waste application to untilled soil gave larger nutrient losses in
runoff than waste application on areas receiving some fall tillage.
13. Surface applications of commercial fertilizers to fallow soils in-
crease the concentration of soluble nitrate N, ammonium N, and soluble
inorganic P in runoff. However, the portion of added fertilizer lost with
intense rainstorms was low (less than 8% of added nutrients).
14. The majority of N and P in surface runoff from fertilizer cropland is
transported eroded sediment. The bulk of sediment-bound N and P is associ-
ated with the clay fraction.
15. The most successful approach for minimizing nutrient losses in surface
runoff from cropland is a combination of soil erosion control and incor-
poration of fertilizers after application.
CONCLUSIONS
-------
18
16. For the purpose of modeling agriculture systems, certain parameters of
surface runoff can be successfully estimated by analysis of the properties
of sediments or surface soils.
a. The concentration of soluble inorganic P in surface runoff can be
estimated by measurement of the equilibrium phosphorus concentration
or the extractable phosphorus content of either the eroded sediment or
original soil.
b. The concentration of soluble ammonium in runoff can be estimated
from the exchangeable ammonium content of the eroded soil and original
soil.
c. The concentration of nitrate in surface runoff was not related to
any runoff properties measured. Modelling of nitrate loss in runoff
will be difficult because nitrate losses are dependent upon a large
number of factors.
2.15 WATERSHED MODEL
(See Section 4.9.1)
1. A comprehensive, distributed parameter simulator, named ANSWERS, has
been developed.
2. ANSWERS offers unique and important capabilities for use during imple-
mentation phases of 208 planning.
3. The importance of applying non-point source control measures on a high-
ly site-specific basis has been clearly demonstrated by use of the ANSWERS
model on the Black Creek Watershed.
2J.6 REMOTE DATA ACQUISITION
(See Section 3.4.3.1)
1. The feasibility has been demonstrated of operating a real-time environ-
mental data acquisition system which supports remote field transducers and
computer activated sampling equipment under control of a general purpose
time-sharing system.
2. The concepts of the ALERT data acquisition system are applicable to
numerous non-point source monitoring requirements and result in several
operational benefits including: expanded data gathering, improved validity
of field data, more efficient utilization of field personnel and automated
early warning systems.
2.17 SOCIO-ECONOMIC CONCLUSIONS
(See Section 4.8)
1. Government agencies make an important contribution in encouraging the
adoption of conservation practices by providing information to farmers
about the practices.
2. The favorable attitude of farmers and the high level of participation
indicate adoption of practices can be achieved in most cases without coer-
CONCLUSIONS
-------
19
cive legislation.
3. A given level of water quality can be provided at the least cost to
participating farmers when they are provided as many alternatives as possi-
ble in selecting the management practices to achieve that required level of
quality.
4. Most economic costs are borne by those farmers who operate on the more
erosive land. Therefore, it may be appropriate to concentrate the cost
sharing assistance on these situations.
5. If the cost of a conservation practice is based only on yield reduc-
tions, the yearly dollar loss is dependent on crop pieces. When crop
prices are low, the cost of reducing soil loss is significantly lower than
occurs in periods of high farm product prices.
2 .18 LAND MANAGEMENT
(See Section 3.3 and Section 5.1)
1. The traditional approach to planning each acre is too methodical to
meet water quality objectives set out in 208 plans.
2. Selection of practices must emphasize the goal of improving water qual-
ity but minimize negative effects on production.
3. The land user must maintain practices. Field technicians must be able
to certify compliance on the basis of visual inspection (annual) during any
season of the year.
4. Without the use of a simplified format (BMP), conservation plans are
difficult for farmers to interpret and even more difficult for those trying
to certify compliance.
5. The most faithfully maintained conservation practices in Black Creek
were those that provided a "visual reminder" of the contractual committ-
ment. Most all permanent practices provide this visual mark (field bord-
ers, waterways, etc.). Conversely, most management practices (tillage, ro-
tations, etc.) do not leave an easily observed year round mark or reminder.
6. Planning to improve water quality alone will not leave a "showplace for
applied conservation." Concentrating on treating primarily the critical
areas on tillable land may be less apparent to the public.
7. It is not possible to conclude that BMP's are not necessary, when aver-
age gross erosion from a large area is demonstrated to be small. Neither
is it possible to conclude that because soil loss as measured by the USLE
is below the tolerable soil loss limits, water quality will be adequate.
Losses may be large enough from relatively small critical areas to have an
adverse effect on water quality. (See Section 4.9.1). This further em-
phasizes the need to concentrate treatment efforts on critical areas.
CONCLUSIONS
-------
20
2.19 ADMINISTRATION
(See Section 3.3 and Section 5.1)
1. The Allen County Soil and Water Conservation District demonstrated its
ability to deal with a complex program of water quality management. The
availability of district personnel, paid with local funds and responsible
to the board of supervisors, was important to this capability.
2. Public Information was vital to the success of the demonstration pro-
ject. Because all parts of the community were informed of plans and pro-
gress at all steps of the project, rumors and potential opposition to the
project did not develop.
3. The ability of the local administration to alter cost sharing rates was
important both in convincing landowners to apply certain practices and in
avoiding paying higher cost-share rates than necessary on practices that
proved to be popular.
4. Cost of treating a watershed such as the Black Creek can be categor-
ized. Necessary land treatment involves water quality improvement, soil
protection, increasing or maintaining agricultural productive capabilities,
and other conservation purposes. All of these goals should be considered
in a watershed program. However, attempts should be made to assign costs
to the appropriate category. All soil conservation costs for a particular
watershed cannot be considered costs of improving water quality.
5. Local leadership, such as a Soil and Water Conservation District, is in
the most favorable position to combine all state and local and federal pro-
grams into a total system which will balance the sometime conflicting water
quality improvement goals.
6. The most cost effective method of achieving improved water quality
through the best management practice approach is to concentrate remedial
efforts on those critical areas within watersheds where maximum benefit can
be obtained. It may not be necessary to treat every acre of every
watershed to achieve realistic water quality goals.
TECHNICAL APPROACH
-------
SECTION III
Technical Approach
Section Title
3.1 Selection of Black Creek
3.2 Design of Land Treatment
3.3 Technical Approach
Authors
J. Morrison
J. Lake
J. Lake
T. D. McCain
R. Land
Page
21
47
51
D. Nelson
E. Monke
A. Bottcher
L. Hugqins
S. Mahler
-------
-------
21
3.1 SELECTION OF BLACK CREEK
One of the most important tasks undertaken during the six-month plan-
ning phase of this program was the selection of a study area which would
accurately represent the Maumee Basin.
It was originally proposed that the area to be studied contain no more
than 20,000 acres. This represents less than one percent of the total land
area in the Maumee Basin. Because of the small size of the study area in
comparison to the basin, it was necessary to find a study area which was
similar to the basin in characteristics of soil type, land use, cultural
practices, and anticipated future land use.
It was also considered necessary to select a study area so that it
would be possible to both monitor gross results and to conduct small plot
experiments.
3.1.1 Selection Criteria
To facilitate the selection of the most representative study area, the
following general criteria were used:
1. The study area should include lake bed and upland soils which are rea-
sonably representative of much of the total basin.
2. Sufficient drainageways should be present so that monitoring stations
could be installed to evaluate erosion and sedimentation both from upland
areas as well as where the channel enters the Maumee River.
3. Existing land uses and cultural practices should be comparable to those
of the total Maumee Basin.
4. The anticipated future land uses should be typical of those expected
throughout the Maumee Basin.
5. The physiography of the study area should facilitate the separation of
runoff between agricultural areas and land under other uses.
6. It was considered desirable to have legal drains in the area with long
time records.
7. The study area should drain directly into the Maumee River.
8. The area should be up to 20,000 acres in size.
The area selected as most nearly satisfying these criteria was the
12,038 acres which drain into Black Creek in northeastern Allen County.
3.1.2 Comparison of Area, Maumee Basin
The Black Creek watershed contains both soils and land uses which are
representative of the Basin. Black Creek Study Area contains 36 percent
upland soils of the silty clay loam till of the Ft. Wayne moraine in the
TECHNICAL APPROACH
-------
22
Blount-Morley-Pewamo association. Soils are 39 percent Blount, 38 percent
Morley and 16 percent Pewamo with only 7 percent minor soils.
Below the upland, in a belt about 1-1/2 miles wide, on the lake plain,
is an apron of medium-textured sediments underlying the Rensselaer-
Whitaker-Oshtemo association comprising 25 percent of the watershed.
Poorly-drained Rensselaer and Whitaker make up 28 and 21 percent, respec-
tively, and excessively drained Oshtemo 6 percent. Soils such as well-
drained Martinsville and Belmore, comprise the remaining 5 percent.
Toward the outer edge of this apron is a small association, making up
5 percent of the watershed, where sandy loams overlie clays at less than 3
feet. This area, in the Haskins-Hoytville association, contains 34 percent
poorly-drained Haskins, 31 percent poorly-drained Nappanee, and 35 percent
minor soils.
On the main lake plain itself comprising 29 percent of the watershed
is the very level high clay (40-50 percent clay in subsoils) Hoytville-
Nappanee association. About 48 percent is dark poorly-drained Hoytville,
23 percent is light colored Nappanee and 29 percent is of minor soils.
Alluvial soils of overflow bottomlands comprise only 5 percent of the
watershed and occur mainly along the lower reaches of the Black Creek in
the four miles before it enters the Maumee River. Narrow bodies occur in
the upland as along Wertz Drain and the main stem of Black Creek southwest
of BarIan. In this Shoals-Eel association, Shoals soils comprise 44 per-
cent, Eel 20 percent and minor soils 27 percent.
These five soil associations comprise a range of soil conditions vary-
ing from those with 50 percent subsoil clay to those with less than 10 per-
cent. Surface soils range from silty clays to loamy fine sands.
Only the Paulding and Latty clay areas, having over 50 percent clay in
the subsoils and Ottokee and Granby, on the deep sand deposits of the north
part of the lake plain east of Archbold, are not represented in the Black
Creek Watershed. Data concerning these soil types is being collected in
another study being undertaken by Ohio State University.
Comparing percentages by land capability classes and subclasses for
the Maumee Basin with those for lands in the Black Creek Study Area reveals
how closely this watershed represents conditions in the Maumee Basin as a
whole. Table 1 illustrates this comparison.
The Maumee Basin is an area of intensive farming, producing corn, soy-
beans, wheat, sugar beets, speciality crops including tomatoes, and others
for canning. Amount of land in tillage-rotation varies from about 75 to 90
percent, being least in the more rolling counties and greatest in the coun-
ties which are mostly in the lake plain. Wooded land ranges from 5 to 19
percent among counties, being greatest in the sandiest ones, and permanent
pasture is generally low. The two most urbanized counties are Lucas
(Ohio), where 43 percent is occupied by Toledo and its environs, and Allen
(Indiana), where 12 percent is in Ft. Wayne and its surroundings.
More than 95 percent of the Black Creek Study Area is devoted to agri-
cultural uses. This includes nearly 81 percent in cropland, 4 percent in
pasture, 7 percent in woodland, 4 percent in other agricultural related
TECHNICAL APPROACH
-------
23
TABLE 1
Land Capability Comparisons - Mauraee Basin and Black Creek Study Area
Capability
Class
Subclass
I
He
Hie
IVe
IIw-IIIw
Hs-IIIs-IVs-VIe
Percent
Different
Maumee Basin
0.9
7.4
3.5
1.4
82.6
4.2
100.0
of Land Area In
Land Capabilities
Black Creek Area
2.4
12.6
3.0
1.3
79.6
1.1
100.0
uses and 4 percent in urban and built-up areas. This distribution of land
use compares favorably with the land use in the total Maumee River Basin as
shown in Table 2.
TABLE 2
Land Use Comparisons - Maumee Basin and Black Creek Study Area
Percent of Lands in Different Uses
Land Use
Cropland
Pasture
Woodland
Urban & Built-up
Other
Maumee Basin
73
4
8
9
6
100
Black Creek Study Area
80.7
4.3
7.1
3.6
4.3
100.0
As in the Maumee Basin, corn and soybeans are the major crops produced
with an estimated 7,000 acres devoted to these crops. Small grains and
meadow in rotation represent a correspondingly smaller amount of cropland
acreage.
The scattered woodlands and the relatively smaller acreages of pasture
and haylands in the Black Creek Study Area are typical of these land uses
in the Maumee Basin.
Urban and built-up acreages for the study area are less, on a percen-
tage basis, than for the total basin, since data for the basin includes the
large population centers of Toledo and Lima, Ohio, and Ft. Wayne, Indiana.
The Black Creek Study Area town of Harlan is fairly representative of the
small towns and villages found in the Maumee Basin.
The Maumee Basin was one of the last areas of the Lake Erie Basin to
be settled. Although Fort Wayne and Toledo were among the outposts esta-
blished around 1800, it was not until the Erie Canal opened an easy water
route to the region in 1825, that settlement of the Lake Erie region really
TECHNICAL APPROACH
-------
24
flourished. The "Great Black Swamp" was the last area to be settled. It
comprises the major portion of the Maumee Basin and represents the area of
the formerly glacial Lake Maumee.
It was primarily the German settlers, with their knowledge of farm
drainage, who brought the black soils of the former lake bed into produc-
tive use. By the middle of the nineteenth century, the dense forests of
this area had been cut, and the broad, flat lands now have one of the most
extensive farm drainage systems in the nation.
The Maumee Basin is today the largest and most productive agricultural
area within the entire Lake Erie region. Except for some suburban!zing in-
fluences in the Toledo, Lima, and Ft. Wayne areas, the Maumee Basin is al-
most entirely devoted to agricultural use.
The Maumee River Basin comprises 6,608 square miles, of which 1,283
are in northeastern Indiana, 4,862 in northwestern Ohio and 463 in southern
Michigan. Approximately 4,229,100 acres are involved in 26 counties: 17 in
Ohio, 6 in Indiana and 3 in Michigan. In Ohio, the Basin includes all of
Allen, Defiance, Henry, Paulding, Putnam, Van Wert, and Williams Counties;
substantial portions of Auglaize, Fulton, Hancock, Hardin, Lucas, Mercer,
and Wood Counties; and smaller areas of Seneca, Shelby, and Wyandot.
Within Indiana, the Basin includes substantial portions of Adams, Allen and
DeKalb Counties and smaller portions of Noble, Steuben and Wells Counties.
The Michigan portion includes portions of Hillsdale and Lenawee Counties
and a very small portion of Branch County.
The average annual rainfall for the Basin ranges between 28 and 36
inches. The mean annual temperature is about 50 degrees Fahrenheit, with
monthly means ranging between approximately 25-30 degrees in January and
February and 70-75 degrees in July and August. The mean length of the
freeze-free period ranges between 150 and 180 days for most of the Basin.
The basin is roughly circular in shape, measuring about 100 miles in
diameter. The Maumee River is formed at Ft. Wayne, Indiana by the conflu-
ence of the St. Joseph River and St. Mary's River. The St. Joseph River
rises in Hillsdale County, Michigan, and flows southwestward. The St.
Mary's River rises in Auglaize County, Ohio and flows in a northwestward
direction to Ft. Wayne, where it turns abruptly to a northeastward direc-
tion before joining with the St. Joseph River to form the Maumee River.
The Maumee River flows in a northeastward direction from Ft. Wayne, across
the Basin to Toledo and its entrance to the Maumee Bay of Lake Erie. Two
major tributaries, the Tiffin River and Auglaize River join the Maumee
River from the north and south respectively, at Defiance, Ohio.
Topography ranges from a nearly flat, featureless plain across much of
the center and eastern portion of the Basin to rolling hills around por-
tions of the Basin's periphery, especially in Michigan and Indiana. The
altitude ranges from nearly 1150 feet (mean sea level) in Hillsdale County,
Michigan to 570 feet at the mouth of the Maumee River. Local relief ranges
from a few tenths of a foot over much of the area to nearly 100 feet in the
rolling hills of Michigan and Indiana. The Maumee River flows in a tortu-
ous channel entrenched some 25 to 40 feet below the lacustrine plain. The
river is generally lacking any significant terrace or flood plain develop-
TECHNICAL APPROACH
-------
25
ment.
The erosion rates of the Maumee River Basin are often reported among
the highest in the Great Lakes Basin. The estimated annual gross erosion
exceeds 4-1/2 tons per acre. By contrast, the current estimated gross ero-
sion rate for the entire Great Lakes Basin is about 2 tons per acre. Sedi-
ment yields in the Basin are relatively large, as indicated by Waterville,
Ohio gage data. From 1951 to 1958, nearly 1-1/2 million tons of sediment
passed the Waterville gage annually. In addition, the sediment load in the
River fluctuates greatly. For example, during a 3-day period in February,
1959, nearly one-half million tons of sediment passed the Waterville gage.
Physiographically, the Maumee River Basin is essentially a nearly lev-
el plain that represents a portion of the abandoned floor of glacial Lake
Maumee which occupied the Lake Erie Basin in late Pleistocene time. Aban-
doned shoreline deposits diverge in a northeastward and southeastward
direction from Ft. Wayne. Dominant surficial deposits include lacustrine
clays and sands and reworked, wave-scoured lake-bottom till. Bedrock con-
sists predominantly of Silurian and Devonian limestones, dolomites and
shales. Depth to bedrock in the Indiana portion of the Basin ranges from
less than 50 feet to about 150 feet.
The Maumee River Basin is primarily agricultural, with more than 90
percent of the land in the Basin in agricultural use. Approximately 73
percent is in cropland, 4 percent in pasture, 8 percent in woodland, 6 per-
cent in other agricultural related uses and 9 percent in urban and built-up
areas. The principal crops grown are corn, soybeans, wheat, and oats, with
some sugar beets. There are also significant acreages of vegetable crops
and nursery stock produced within the Basin. Sales from livestock and
livestock products account for about one-fourth of the income from farm
sales.
Total population in the area is approximately 1,295,000 of which
50,000 reside in Michigan, 275,000 in Indiana, and the remaining 970,000 in
Ohio. Toledo, Ohio and Ft. Wayne, Indiana, are the major cities with Lima,
Findlay, and Defiance, Ohio, being the other major population centers. The
remainder of the Basin is primarily rural with a number of smaller
agriculturally-oriented communities.
The principal industries are machinery, electrical and transportation
equipment manufacture, metal fabrication, petroleum refining, and food pro-
cessing. Major industrial centers within the Basin are Toledo and Lima,
Ohio and Ft. Wayne, Indiana.
Toledo ranks as the nation's third largest railroad center, and the
city's port, which is the ninth largest in the United States, is the
world's largest shipper of soft coal. Major products passing through the
port include iron ore, farm products, machinery, and petroleum products.
Lima, Ohio is the center of an oil distribution system for the Great Lakes
and Eastern markets, while Toledo is the largest petroleum refining center
between Chicago and the Eastern Seaboard.
The drainage basins of the St. Joseph and St. Mary's River which join
at Ft. Wayne (where they reverse course and head toward Lake Erie), are
largely controlled by glacial features of the Lake Erie glacial lobe. This
lobe pushed across rocks mainly of limestone and shale and carried fine
TECHNICAL APPROACH
-------
26
till material into present day northwest Ohio, northeast and east central
Indiana and south central Michigan. During the last major stand of this
glacial lobe, in its retreat some 10,000 years ago, the Fort Wayne moraine
was deposited concentric to the front of the retreating lobe (see Figure
1), and this dammed up a great body of water between it and the eastward
retreating ice front of the lobe. This water body was named by geologists
"glacial lake Maumee," and the land area once covered by it is known today
as "the lake plain."
3.1.2.1 GENERAL NATURE OF THE LAKE PLAIN
Glacial Lake Maumee did not remain long enough to influence all of the
lake plain uniformly. In the west end and along the south border it merely
reworked the glacial till beneath it, leveling the surface but leaving only
a thin deposit of fine lake-laid sediments. Similar areas occur in the
central part of tne bacin northeast and east of Defiance. There are a
number of areas where clays are overlaid by sandy or loamy sediments up to
3 feet thick.
In areas below the steep northeastern trending flank of the Ft. Wayne
moraine, deltas of loamy materials composed of eroded debris from the up-
lands were deposited in Lake Maumee. In this and similar border areas,
temporary lake stages were recorded as beach ridges. In these areas the
material deposited includes sandy and/or loamy beach ridges and deep loamy
sediments on level and depressed areas. Loamy sediments were deposited
only thinly over lake clays or till by action of water or wind.
Near the center of the glacial lake Maumee, fine sediments were depo-
sited most deeply as the retreating glacial lobe stood somewhat east of De-
fiance. Here in an east-facing crescent is an area known as the Paulding
Basin. These sediments in the Paulding Basin are higher in clay content
than any other part of the Maumee Basin. This area was the center of what
was once called the Maumee Swamp or Marsh. Beach ridges developed concen-
tric to the receding lake borders just as they did at the Fort Wayne end of
the glacial lake. Between Defiance and Toledo, clay loam till reworked by
waters of glacial lake Maumee lies east of the Paulding Basin. The north
flank of the lake plain is mantled with thick to thin sands. Sandiest
areas occur just west of Bowling Green, southwest of Toledo and in the
Wauseon vicinity. In these same areas, sandy loam and loam mantles only a
few feet thick over clayey till or in thin mantles of loamy sand over
clayey till or lake-laid clays. There is a high degree of local variation
in these areas in comparison with the more clayey parts of the lake plain.
3.1.2.2 GLACIAL MORAINES AND TILL PLAINS
Clay loam till, left by the receding glacial lobe of Lake Erie occurs
in parts of nine Michigan, Ohio and Indiana counties on the northwest flank
of the lake plain and twelve Indiana and Ohio counties on the southwest and
south. That part between St. Joseph River and the lake plain is perhaps
the most rolling with best expressed morainic features and is mostly part
of the Fort Wayne moraine. That southeast of Fort Wayne and east toward
Findlay lies lower and is less rolling, being mostly ground moraine.
Drainage of the southwest portion is through the St. Mary's River, which
parallels the south flank of the Fort Wayne moraine. The eastern portion
drains toward the Auglaize River and its tributaries, which flow north
through the lake plain. The northern part drains through the St. Joseph
TECHNICAL APPROACH
-------
27
M.WAYNE
\
\
RIVER BASIN BOUNDARY
Figure 1. Maumee Basin Map
River, which parallels the north flank of the Fort Wayne moraine, and
TECHNICAL APPROACH
-------
28
through Tiffin River, which flows south across the lake plain.
On the south flank, where the rise to the till plain is very gradual,
it is hard to determine the exact location of the lake plain boundary.
Since there was apparently less eroded debris from the uplands on the south
side, only a discontinuous apron of medium-textured deltaic deposits formed
on the southwestern flank. However, there are a number of local lake bed
deposits and muck areas in the till plain which occupy broader depressions.
Also, there are lake border ridges like that one followed by U.S. Highway
30 southeast of Fort Wayne, and broader deltaic strips fringe the lake
plain in the area north of Lima and east toward Findlay.
At the extreme north end of the St. Joseph River drainage in Michigan
the till is sandier and more elevated and more rolling. In this area there
are many valley train deposits along courses of glacial meltwater streams
which are often underlaid by sand and gravel.
3.1.2.3 SOILS ASSOCIATION MAUMEE BASIN
Each of the soil associations of the Maumee Basin is described below.
Associations 1 through 5 are soils dominantly formed in glacial till. As-
sociations 6 through 10 are soils predominantly formed in water-deposited
material, organic material, eolian material (see Figure 2).
3.1.2.3.1 Blount-Pewamo
Depressional to gently sloping, very poorly-drained to somewhat
poorly-drained soils that have clayey subsoils. The landscape in this as-
sociation consists of a glacial-ground moraine that is nearly level with
many narrow depressions. This soil formed in glacial till. This soil as-
sociation occupies about 26 percent of the watershed. Blount soils are
nearly level and gently sloping and are somewhat poorly-drained. They have
a surface layer of very dark grayish-brown and dark grayish-brown loam or
silt loam and a subsoil that is mostly dark-brown and dark grayish-brown,
mottled silty clay and clay. Pewamo soils are depressional, nearly level
and are very poorly-drained. They have a surface layer of a very dark gray
silty clay loam and a subsoil that is mostly dark gray or grayish-brown,
mottled silty clay or silty clay loam.
3.1.2.3.2 Morley-Blount-Pewamo
Depressional to moderately steep, very poorly-drained to moderately
well-drained soils that have clayey subsoils. The landscape in this asso-
ciation consists of a glacial moraine that is gently rolling with some
depressional areas near drainageways. The soils formed in glacial till.
This soil association occupies about 22 percent of the watershed. Morley
soils are gently sloping to moderately steep and are moderately well-
drained. They have a surface layer of very dark grayish-brown and grayish-
brown silt loam and a subsoil that is mostly dark yellowish-brown and brown
clay and is mottled in the lower part. Blount soils are nearly level and
gently sloping and are somewhat poorly-drained. They have a surface layer
of very dark grayish-brown and dark grayish-brown loam or silt loam and a
subsoil that is mostly dark brown and dark grayish-brown, mottled silty
clay and clay. Pewamo soils are depressional nearly level and are very
TECHNICAL APPROACH
-------
29
clay or silty clay loam.
3.1.2.3.3 Miami-Conover
Nearly level to moderately steep, well-drained and somewhat poorly-
drained soils that have loamy subsoils. The landscape in this soil associ-
ation consists of a glacial moraine that is gently rolling with some
depressional areas near drainageways. The soils formed in glacial till.
This soil association occupies about 2 percent of the watershed. Miami
soils are gently sloping to moderately steep and are well drained. They
have a surface layer of dark grayish-brown loam and a subsoil that is dark
brown clay loam. Conover soils are nearly level and are somewhat poorly-
drained. They have a surface layer of very dark grayish brown loam and a
subsoil that is mostly yellowish-brown and dark yellowish-brown, mottled
clay loam.
3.1.2.3.4 Hillsdale-Fox
Gently sloping to moderately steep, well-drained soils that have loamy
subsoils. The landscape in this soil association consists of glacial
moraines and valley trains that are rolling with nearly level areas at the
lower elevations. The soils formed in glacial till and outwash. This as-
sociation occupies about 1 percent of the watershed. Hillsdale soils are
gently sloping to moderately steep and are well-drained. They have a sur-
face layer of dark grayish-brown sandy loam and a subsoil that is dark
brown and dark yellowish-brown sandy loam and sandy clay loam. Fox soils
are gently sloping to moderately steep and are well drained. They have a
surface layer of dark grayish-brown loam and a subsoil that is dark brown
clay loam and gravelly loam.
3.1.2.3.5 Hoytville-Toledo-Nappanee
Depressional to gently sloping, very poorly-drained and somewhat
poorly-drained soils that have clayey subsoils. The landscape in this soil
association consists of glacial lake plain and glacial till plain that is
dominantly nearly level with occasional slight rises. The few sloping
areas in the landscape are near deeply dissected streams. Hoytville and
Nappanee soils formed in glacial till. Toledo soils formed in lacustrine
sediments. This soil association occupies about 17 percent of the
watershed. Hoytville soils are depressional and nearly level and are very
poorly-drained. They have a surface layer that is very dark gray silty
clay and a subsoil of dark grayish-brown, mottled silty clay. Toledo soils
are depressional to level and are very poorly drained. They have a surface
layer of very dark gray silty clay and a subsoil that is dark gray and
gray, mottled silty clay. Nappanee soils are nearly level to gently slop-
ing and are somewhat poorly-drained. They have a surface layer that is
dark gray and grayish brown silt loam or silty clay loam and a subsoil that
is mostly grayish brown, mottled clay.
TECHNICAL APPROACH
-------
30
10
LEGEND
SOILS DOMINATLY FORMED IN GLACIAL TILL
BLOUNT-PEWAMO ASSOCIATION: Depressional to gently sloping,
very poorly drained to somewhat poorly drained soils that have
clayey subsoils.
MORLEY-BLOUNT-PEWAMO ASSOCIATION: Depressional to
moderately steep, very poorly drained to moderately well-drained
soils that have clayey subsoils.
Ml A Ml-CO MOVER ASSOCIATION: Nearly level to moderately steep,
well-drained and somewhat poorly drained soils that have loamy sub-
soils.
HILLSDALE-FOX ASSOCIATION: Gently sloping to moderately steep,
well-drained soils that have loamy subsoils.
HOYTVILLE-TOLEDO-NAPPANEE ASSOCIATION: Depressional to
gently sloping, very poorly drained and somewhat poorly drained soils
that have clayey subsoils.
SOILS DOMINANTLY FORMED IN WATER-DEPOSITED
MATERIAL, ORGANIC MATERIAL, AND EOLIAN MATERIAL
CARLISLE-MONTGOMERY ASSOCIATION: Depressional and nearly
level, very poorly drained soils that have organic and clayey subsoils.
PAULDING-LATTY-ROSELMS ASSOCIATION: Depressional and
nearly level, very poorly drained and somewhat poorly drained soils
that have clayey subsoils.
HANEY-BELLMORE-MILLGROVE ASSOCIATION: Depressional to
strongly sloping, very poorly drained, moderately well-drained, and
well-drained soils that have loamy subsoils.
MERMILL-HASKINS-WAUSEON ASSOCIATION: Depressional and
nearly level, very poorly drained and somewhat poorly drained soils
that have loamy and clayey subsoils.
OTTOKEE-GRANBY ASSOCIATION: Depressional to sloping, very
poorly drained, poorly drained, moderately well-drained soils that
have sandy subsoils.
Legend Figure 2
TECHNICAL APPROACH
-------
31
Figure 2. Soil Associations - Maumee Basin
poorly-drained. They have a surface layer of very dark gray silty clay
loam and a subsoil that is mostly dark gray or grayish-brown, mottled silty
TECHNICAL APPROACH
-------
32
3.1.2.3.6 Carlisle-Montgomery
Depressional and nearly level, very poorly-drained soils that have or-
ganic and clayey subsoils. The landscape in this soil association consists
of a local lake plain that is flat and is surrounded by a glacial ground
moraine. Carlisle soils formed in organic materials. Montgomery soils
formed in lacustrine sediments. This soil association occupies about 1
percent of the watershed. Carlisle soils are depressional to nearly level
and are very poorly-drained. They have a surface layer of black muck and
underlying material that is black and dark-reddish brown muck. Montgomery
soils are depressional to nearly level and are very poorly-drained. They
have a surface layer of black silty clay loam and a subsoil that is dark
gray, grayish-brown, and gray silty clay loam and silty clay.
3.1.2.3.7 Paulding-Latty-Roselms
Depressional and nearly level, very poorly-drained and somewhat
poorly-drained soils that have clayey subsoils. The landscape in this soil
association consists of a glacial lake plain that is dominantly nearly lev-
el with occasional slight rises. A few sloping areas in the landscape are
near deeply dissected streams. The soils are formed in lacustrine materi-
al. This soil association occupies about 15 percent of the watershed.
Paulding soils are nearly level and are very poorly-drained. They have
surface layers that are dark gray clay and subsoil that is gray and olive
gray, mottled clay. Roselms soils are nearly level and are somewhat
poorly-drained. They have a surface layer of dark gray silty clay loam and
a subsoil that is light gray, brown , and grayish brown, mottled heavy
clay.
3.1.2.3.8 Haney-Belmore-Mi1Igrove
Depressional to strongly sloping, very poorly-drained, moderately
well-drained, and well-drained soils that have loamy subsoils. The
landscape in this soil association consists of long narrow sloping beach
ridges rising above the terrain and nearly level glacial deltas and lake
plain. The soils formed in glacial and beach ridge deltaic deposits and
lacustrine sediments. This soil association occupies about 10 percent of
the watershed. Any soil named in this association is more extensive than
the many soils of small extent not named. Although collectively, the Ha-
ney, Belmore, and Millgrove soils do not make up the majority of the asso-
ciation. Haney soils are gently sloping and sloping and are moderately
well-drained. They have a surface layer of dark grayish-brown loam and a
subsoil that is dark brown sandy clay loam and gravelly sandy clay loam.
Millgrove soils are depressional to nearly level and are very poorly-
drained. They have a surface layer of dark yellowish-brown loam and a sub-
soil that is dark brown sandy clay loam and gravelly sandy clay loam, very
dark-grayish-brown loam and a subsoil that is dark grayish brown and
grayish-brown, mottled sandy loam and sandy clay loam.
TECHNICAL APPROACH
-------
33
3.1.2.3.9 Mermi11-Haskins-Wauseon
Depressional and nearly level, very poorly-drained and somewhat
poorly-drained soils that have loamy and clayey subsoils. The landscape in
this soil association consists of a glacial lake plain and glacial ground
moraine that are nearly level with depressional areas and some gently undu-
lating rises. The soils formed in the outwash on glacial till or lacus-
trine sediments. This soil association occupies about 3 percent of the
watershed. Mermill soils are depressional and nearly level and are very
poorly-drained. They have a surface layer of very dark gray sandy loam.
There subsoil is mottled and is dark gray, gray, and grayish-brown. It is
a sandy clay loam in the upper part and a clay in the lower part. Haskins
soils are nearly level and are somewhat poorly-drained. They have a sur-
face layer of dark grayish-brown loam. The subsoil is mottled and is
yellowish-brown and light yellowish-brown. It is sandy clay loam, sandy
loam, and loam in the upper part and light clay in the lower part. Wauseon
soils are depressional and nearly level and are very poorly-drained. They
have a surface layer of very dark gray fine sandy loam. The subsoil is
mottled and is dark gray, grayish-brown, and gray. It is fine sandy loam
in the upper part and clay in the lower part.
3.1.2.3.10 Ottokee-Granby
Depressional to sloping, very poorly-drained, poorly-drained, and
moderately well-drained soils that have sandy subsoils. The landscape in
this soil association consists of beach ridges that are nearly level with
gently undulating rises. The soils formed in water-laid and eolian sedi-
ments. This association occupies about 3 percent of the watershed. Ot-
tokee soils are gently sloping and sloping and are moderately well drained.
They have a surface layer of very dark grayish-brown loamy fine sand and a
subsoil that is light yellowish-brown and yellowish brown, mottled loamy
fine sand. Granby soils are depressional and nearly level and are very
poorly-drained. They have a surface layer of black loamy sand and a sub-
soil that is dark gray and light brownish gray, mottled sand.
3.1.2.4 SOILS OF BLACK CREEK WATERSHED
The Black Creek study area comprises a drainage area of approximately
18.8 square miles (12,038 acres) in northeastern Allen County, Indiana.
The watershed is about 13 miles northeast of Ft. Wayne. Black Creek ori-
ginates about 2 miles north of the community of Harlan and flows in a
south-southeasterly direction for about 4 miles, where it turns to an eas-
terly direction for about 2 miles. Thence, after a number of abrupt
changes in direction the creek flows southward for about 1-1/2 miles to the
Maumee River. Black Creek is an entrenched stream throughout most of its
course particularly in the lowermost 2 miles when it flows about 25 to 30
feet below the general level of the lacustrine plain. Principal tribu-
taries are Smith-Fry Drain, Wertz Drain, Reichelderfer Drain and Upper
Gorrell Drain (see Figure 3).
The mean annual rainfall at Fort Wayne is 35.31 inches. The rainfall
is well distributed throughout the year, with the month of December having
the least (2.09 in.) and the month of June having the most (4.17 in.). The
mean annual temperature is 50.3 degrees Fahrenheit with a mean July tem-
TECHNICAL APPROACH
-------
34
LEGEND
Study Area Boundary
Highway
Road
Creek or Drain
Town
Figure 3. Black Creek Watershed
TECHNICAL APPROACH
-------
35
perature of 74.2 degrees and a mean January temperature of 27 degrees.
The altitude of the watershed ranges from about 710 to 850 feet above
mean sea level, a maximum relief on the order of 140 feet. Local relief
ranges from a fraction of a foot on portions of the lacustrine plain to as
much as 40 to 50 feet in the northernmost part of the watershed and in the
entrenched portion of Black Creek near the Maumee River.
The Black Creek study area is largely within the Maumee lacustrine
plain (see Figure 4). Surficial deposits consist largely of wave-scoured
lakebottom till. A narrow (about 1,000 foot) band of beach and shoreline
deposits parallels Indiana Route 37 through the watershed. These shoreline
deposits are bordered on the northwest by glacial till end-moraine deposits
and to the southeast by a rather narrow (approximately I mile wide) band of
lacustrine sands which grade into the wave-scoured lake-bottom tills.
Bedrock consists of Devonian limestone and dolomite generally less than 100
feet deep.
FT. WAYNE
\
\
Figure 4. Location of Study Area
The Black Creek Watershed area is entirely rural except for the small
unincorporated community of Harlan, which is located along Indiana Route 37
in the west central portion of the watershed. Land ownership is character-
TECHNICAL APPROACH
-------
36
ized by numerous small holdings. There are 176 individual ownership
tracts, of which 127 or 72% are less than 100 acres, 45 or 26% are from
100-249 acres, and only 4 (2%) are 250 acres or larger. The average value
of land and buildings is approximately $1,400 per acre.
The proximity of the watershed to Ft. Wayne provides excellent oppor-
tunities for employment in needy industry and results in high off-farm em-
ployment. It is estimated that nearly 2/3 of the farm operators work off
the farm. Of those operators who have off-farm employment, approximately
20% work less than 100 days off the farm, and 80% work more than 100 days
off the farm.
The average market value of agricultural products sold is approximate-
ly $27,000 per farm. This is about equally divided between the two
categories of cash crops and livestock, poultry and livestock and poultry
products.
Each of the soil associations are described below:
(See Figure 5)
3.1.2.4.1 Blount-Morley-Pewamo
Depressional to moderately steep, very poorly-to moderately well-
drained soils that have clayey subsoils; on uplands. The landscape in this
association consists of glacial-ground moraine and moraine that is nearly
level with many narrow depressions and is gently rolling with some depres-
sional areas near drainageways. The soils formed in glacial till. This
soil association occupies about 36 percent of the watershed. About 39 per-
cent is made up of Blount soils, 38 percent of Morley soils, 16 percent of
Pewamo soils, and 7 percent of minor soils. Blount soils are nearly level
and gently sloping and are somewhat poorly-drained. They have a surface
layer of very dark grayish-brown and dark grayish-brown loam or silt loam
and subsoil that is mostly dark brown and dark grayish-brown, mottled silty
clay and clay. Morley soils are gently sloping to moderately steep and are
moderately well-drained. They have a surface layer of very dark grayish-
brown and grayish brown silt loam and a subsoil that is mostly dark
yellowish-brown and brown clay and is mottled in the lower part. Pewamo
soils are depressional and nearly level and are very poorly-drained. They
have a surface layer of very dark gray silty clay loam and a subsoil that
is mostly dark gray or grayish-brown, mottled silty clay or silty clay
loam.
3.1.2.4.2 Shoals-Eel
Nearly level, somewhat poorly-and moderately well-drained soils that
have loamy subsoils; on bottom lands. The landscape in this association is
nearly level flood plains that are adjacent to streams. The soils formed
in alluvium. This soil association occupies about 5 percent of the
watershed. About 44 percent is made up of the Shoals soils, 29 percent of
Eel soils, and 27 percent of minor soils. Shoals soils are nearly level
and are somewhat poorly-drained. They have a surface layer of dark gray
and dark grayish-brown silty clay loam and a subsoil that is gray silty
clay loam. Eel soils are nearly level and are moderately well-drained.
They have a surface layer of dark grayish-brown and dark brown silt loam
and loam and a subsoil that is brown and dark yellowish-brown, mottled
light silty clay loam.
TECHNICAL APPROACH
-------
37
3.1.2.4.3 Hoytville-Nappanee
Depressional and nearly level, very poorly-and somewhat poorly-drained
soils that have clayey subsoils; on uplands. The landscape in this soil
association consists of glacial till plain that is dominantly nearly level
with occasional slight rises. The soils formed in glacial till. This soil
association occupies about 29 percent of the watershed. About 48 percent
is made up of Hoytville soils, 23 percent of Nappanee soils, and 21 percent
of minor soils. The Hoytville soils are depressional and nearly level and
are very poorly-drained. They have a surface layer that is very dark gray
silty clay and a subsoil of dark grayish-brown, mottled silty clay. Nap-
panee soils are nearly level and are somewhat poorly-drained. They have a
surface layer that is dark gray and grayish-brown silt loam or silty clay
loam and a subsoil that is mostly grayish-brown, mottled clay.
3.1.2.4.4 Rensselaer-Whitaker-Oshtemo
Nearly level to moderately sloping, very poorly, somewhat poorly, and
somewhat excessively-drained soils that have loamy subsoils; on uplands.
The landscape in this soil association consists of long narrow sloping
beach ridges above the terrain and nearly level glacial deltas and lake
plain. The soils formed in glacial deltaic and beach ridge deposits and
lacustrine sediments. This soil association occupies about 25 percent of
the watershed. About 28 percent is made up of Rensselaer soils, 21 percent
of Whitaker soils, 6 percent of Oshtemo soils, and 45 percent of minor
soils. Rensselaer soils are nearly level and are very poorly drained.
They have a surface layer of very dark brown loam, loam to silty clay loam
or mucky silty clay loam this is mottled in the lower part. The subsoil is
mostly gray or strong-brown, mottled sandy loam or sandy clay loam. Whi-
taker soils are nearly level and are somewhat poorly drained. They have a
surface layer of fine sandy loam, loam, or silt loam that is dark grayish-
brown in the upper part and pale brown in the lower part. The subsoil is
yellowish-brown and gray, mottled clay loam or silty clay loam. Oshtemo
soils are nearly level to moderately sloping and are somewhat excessively
drained. They have a surface layer that is dark-brown sandy loam or grav-
elly sandy loam.
3.1.2.4.5 Haskins, Hoytville
Depressional to gently sloping, somewhat poorly-and very poorly-
drained soils that have loamy and clayey subsoils; on uplands. The
landscape in this soil association consists of glacial-ground moraine that
is nearly level with depressional areas and some gently undulating rises.
Haskins soils formed in outwash on glacial till. Hoytville soils formed in
glacial till. This soil association occupies about 5 percent of this
watershed. About 34 percent is made up of Haskins soils, 31 percent of
Hoytville soils, and 35 percent of minor soils. Haskins soils are nearly
level or gently sloping and are somewhat poorly-drained. They have a sur-
face layer of dark grayish-brown loam. The subsoil is mottled and is
yellowish-brown and light yellowish brown. It is loam or sandy loam in the
upper part and light clay in the lower part. Hoytville soils are depres-
sional and nearly level and are very poorly-drained. They have a surface
layer that is very dark gray silty clay and a subsoil of dark grayish-
brown, mottled silty clay.
TECHNICAL APPROACH
-------
38
Figure 5. Soil Associations Black Creek Watershed
TECHNICAL APPROACH
-------
39
LEGEND
BLOUNT-MORLEY-PEWAMO ASSOCIATION: Depressional to moderately steep,
very poorly drained to moderately well-drained soils that have clayey subsoils;
on uplands.
SHOALS-EEL ASSOCIATION: Nearly level, somewhat poorly drained and moder-
ately well-drained soils that have loamy subsoils; on bottom lands.
HOYTVILLE-NAPPANEE ASSOCIATION: Depressional and nearly level, very
poorly drained and somewhat poorly drained soils that have clayey subsoils; on
uplands.
RENSSELAER-WHITAKER-OSHTEMO ASSOCIATION: Nearly level to moderately
sloping, very poorly drained, somewhat poorly drained, and somewhat excessively
drained soils that have loamy subsoils; on uplands.
HASKINS-HOYTVILLE ASSOCIATION: Depressional to gently sloping, somewhat
poorly drained and very poorly drained soils that have loamy and clayey subsoils;
on uplands.
Legend Figure 5
TECHNICAL APPROACH
-------
40
3.1.3 Land Capability Units
The land capability unit represents a grouping of soils which share
common limitations for agricultural uses and which respond to like treat-
ment under similar conditions of use. There are 58 different kinds of soil
in the Black Creek study area. These soils make up a total of 21 Land Ca-
pability Units which are used in determining land treatment needs (see Fig-
ure 6).
3.1.3.1 CAPABILITY UNIT I-I
This unit consists of deep, nearly level, well-drained, medium-textured
soils of the Martinsville and Rawson series. These soils have moderate in-
filtration and permeability and a high available moisture capacity. These
soils are productive and easy to manage and can be cropped intensively.
The proper use of crop residue maintains the content of organic matter and
helps to keep good tilth.
3.1.3.2 CAPABILITY UNIT L-2^
This unit consists of deep, nearly level, well-drained and moderately
well-drained, medium textured soils of the Eel and Genesee series. These
soils are flooded occasionally in the winter and spring. They have
moderate infiltration and permeability and high available moisture capaci-
ty.
3.1.3.3 CAPABILITY UNIT IIE-1
This unit consists of deep, gently sloping, well-drained, medium-textured
soils. These soils are of the Martinsville, Miami, and Rawson series.
They have moderate infiltration and permeability and high available mois-
ture capacity. Erosion control is the main management need. Contour farm-
ing, diversion terraces, sod waterways, and proper crop rotation and
minimum tillage are among the measures that can be used to control erosion.
3.1.3.4 CAPABILITY UNIT IIE-6
This unit consists of deep, gently sloping, moderately well-drained,
medium-textured soils of the Morley series. These soils have moderate in-
filtration, slow permeability, and high available moisture capacity. Their
natural fertility is moderate. Their content of organic matter is general-
ly moderate or low. Erosion is a hazard, particularly in intensively
cropped fields. Diversion ditches, contour tillage, stripcropping, and sod
waterways are among the measures needed for control of erosion. Crop resi-
due and intercrops help to maintain and increase the organic-matter con-
tent. Minimum tillage helps to maintain good tilth and control erosion.
Wet spots created by springs or by seepage can be drained with random tile
lines.
3.1.3.5 CAPABILITY UNIT IIE-9
This unit consists of gently sloping, well-drained soils of the Belmore
series. These soils are moderately deep and deep to gravel and sand. They
have moderately rapid infiltration, moderate permeability, and moderate
available moisture capacity. Erosion is a hazard. Contour farming and sod
waterways are among the measures needed for control of erosion. Proper
management of crop residue is important in maintaining the organic-matter
content.
TECHNICAL APPROACH
-------
41
3.1.3.6 CAPABILITY UNIT IIS-1
This unit consists of nearly level, well-drained, medium-textured soils of
the Belmore series. These soils are moderately deep to gravel and sand.
They have moderately rapid infiltration, moderate permeability, and
moderate available moisture capacity. Droughtiness is a major limitation
and crop residues should be left on the soil to maintain and increases the
content of organic matter.
3.1.3.7 CAPABILITY UNIT IIW-1
This unit consists of deep, level and depressional, very poorly-drained,
dark-colored, medium-textured to fine-textured soils. These soils are of
the Brookston, Hoytsville, Lenawee, Mermill, Pewamo, Rensselaer, Washtenaw,
and Westland series. They are waterlogged in periods of wet weather. They
have moderate infiltration, slow permeability, and high available water
capacity. Wetness is the main limitation. An adequate drainage system is
needed if the common crops are to be grown. Diversion terraces that inter-
cept runoff from adjacent uplands are beneficial. Sod outlets or structur-
al outlets for the diversion terraces are needed. Spring tillage should be
delayed until the plow layer is dry. Minimum tillage and crop residue
management help to maintain good tilth.
3.1.3.8 CAPABILITY UNIT IIW-2
This unit consists of deep, nearly level and gently sloping somewhat
poorly-drained, medium-textured or moderately coarse textured soils of the
Blount, Crosby, Del Rey, Raskins, and Whitaker series. These soils have
moderately slow or slow permeability and high available moisture capacity.
The gently sloping areas are erodible. Wetness is the main limitation. An
adequate drainage system is needed if the common crops are to be grown.
Diversion terraces that intercept runoff from higher areas are beneficial.
Grass waterways are needed. Other practices needed include minimum and
properly timed tillage and management of crop residues.
3.1.3.9 CAPABILITY UNIT IIW-7
This unit consists of nearly level, somewhat poorly drained and very
poorly-drained soils of the Shoals series. These soils are flooded occa-
sionally, and they have a fluctuating water table. They have moderate in-
filtration and permeability and high available moisture capacity. Wetness
is the main limitation. Adequate drainage is important. Other needed
practices include conservation cropping systems, crop residue management
and minimum tillage.
3.1.3.10 CAPABILITY UNIT IIIE-1
This unit consists of deep, moderately sloping, well-drained, medium-
textured soils of the Martinsville and Rawson series. These soils have
moderate infiltration, moderate permeability, and high available moisture
capacity. Erosion is the main hazard. Contouring is the erosion control
practice most applicable on the short slopes. On the few longer and more
uniform slopes, stripcropping can be used. Sod waterways are needed to con-
trol erosion in drainageways.
TECHNICAL APPROACH
-------
42
Figure 6. Land Capabilities Units Black Creek Watershed
TECHNICAL APPROACH
-------
43
LEGEND
CLASS I Land has few hazards that limit its use. It is nearly level, deep,
generally well-drained and suited for intensive cultivation with good management.
1 CLASS I Iw Land is very good cropland but has wetness hazards which may be
11 overcome with proper installation of drainage practices and use of conservation
' management systems.
^r-r^i CLASS He Land is very good cropland but has erosion hazards which may be
I IP £^^1 easi'y overcome with proper installation of practices and conservation manage-
LiiJ ment systems.
IPJ.PI CLASS Illw Land is good cropland but has severe wetness hazards that will
111 loo'j require careful soil management, and installation of more intensive drainage
IIIW \/fA systems to be good cropland. The hazards may limit the intensity and choice of
cultivated crops.
__ CLASS Hie Land is good cropland but severe wind and/or erosion hazards
11 I may 'im't the intensitY of use and ch°ice of crops. Carefully selected conserva-
11 I 6 Mmi tjon practices must be considered in use of this land for crops. The sandy soils
also have a droughty hazard.
.;.;..» CLASS Ills Land is considered good cropland but droughtiness is a severe
111 F&Wa hazard. Careful soil management including irrigation may be considered in the
* i use of this land for crops.
moat CLASS I Ve Land has very severe erosion hazard due to slope which limits
I\/P FWWI 'ts use ^or cu't'vatec' crops. It requires very careful management including
*""' special conservation practices when cultivated.
Legend Figure 6
TECHNICAL APPROACH
-------
44
3.1.3.11 CAPABILITY UNIT IIIE-6
This unit consists of deep, gently sloping and moderately sloping,
moderately well-drained, medium-textured soils of the Morley series. These
soils range from uneroded to severely eroded. They have moderate infiltra-
tion, slow permeability, and high available moisture capacity. Erosion is
a hazard, particularly in intensively cropped fields. Diversion ditches,
contour tillage, stripcropping, sod waterways, crop residue management and
minimum tillage are among the measures needed for control of erosion.
3.1.3.12 CAPABILITY UNIT IIIE-9
This unit consists of Fox loam, 6 to 12 percent slopes, moderately eroded,
a well-drained soil. This soil is moderately deep to sand and gravel. It
has moderate permeability and moderate available moisture capacity. This
soil occurs as small areas, many of which are managed along with less slop-
ing soils that can be used more intensively. As a result, considerable
erosion has taken place. Erosion is the main hazard. Contour tillage,
minimum tillage, mulch tillage, and a suitable cropping system help to con-
trol erosion.
3.1.3.13 CAPABILITY UNIT IIIE-11
This unit consists of deep, gently sloping, well-drained soils of the St.
Clair series. These soils range from uneroded to moderately eroded. They
have moderate infiltration, slow permeability, and high available moisture
capacity. Erosion is the main hazard. Maintaining good tilth and increas-
ing the content or organic matter are problems. Diversion terraces and
contour tillage help to control runoff and erosion. Permanent grassed wa-
terways are needed to prevent gullying of natural drainageways. Minimum
tillage, a suitable cropping system, and proper use of crop residue help to
improve tilth and to increase the content of organic matter.
3.1.3.14 CAPABILITY UNIT IIIE-13
This unit consists of deep, gently sloping and moderately sloping, well-
drained and somewhat excessively drained, moderately coarse textured soils
for the Belmore and Oshtemo series. These soils have moderately rapid in-
filtration, moderate and moderately rapid permeability, and low available
moisture capacity. Erosion is the main hazard, and droughtiness is a seri-
ous limitation. Contour tillage, crop residue management, and minimum til-
lage help to control erosion.
3.1.3.15 CAPABILITY UNIT IIIS-2
This unit consists of deep, nearly level, somewhat excessively-drained
moderately coarse textured soils of the Oshtemo series. These soils have
moderately rapid infiltration and permeability and low available moisture
capacity. Droughtiness is the main limitation to use.
3.1.3.16 CAPABILITY UNIT IIIW-2
This unit consists of deep, nearly level, very poorly-drained, dark-
colored, moderately fine textured or fine textured soils of the Montgomery
series. These soils become waterlogged in periods of wet weather and are
slow to dry out in spring. They have very slow infiltration and permeabil-
TECHNICAL APPROACH
-------
45
ity and high available water capacity. Wetness is the major limitation.
Maintaining good tilth is a problem. An adequate drainage system is need-
ed. It is necessary to keep tillage to a minimum. Crop residue management
in the cropping system is needed.
3.1.3.17 CAPABILITY UNIT IVE-6
This unit consists of deep, moderately sloping and strongly sloping,
moderately well-drained, medium-textured and moderately fine textured soils
of the Morley series. These soils have been eroded so severely that the
present surface layer consists almost entirely of material from the sub-
soil. They have slow to moderate infiltration, slow permeability, and high
available moisture capacity. Erosion is the main hazard. Contour cultiva-
tion, diversion terraces, and sod waterways help to control runoff and ero-
sion. Crop residue management and minimum tillage also improve tilth and
reduce runoff.
3.1.3.18 CAPABILITY UNIT IVE-11
This unit consists of St. Clair silty clay loam, 6 to 12 percent slopes,
moderately eroded, a deep, well-drained or moderately well-drained soil.
This soil has moderate infiltration, slow permeability, and high available
moisture capacity. Erosion is the main hazard. Permanent sod in natural
drainageways helps to control gully erosion. Contour cultivation, crop
residue management, and minimum tillage are effective in the control of
runoff and erosion.
3.1.3.19 CAPABILITY UNIT IVE-1
This unit consists of deep, nearly level and gently sloping well-drained,
coarse-textured soils of the Plainfield series. These soils have rapid
permeability and low available moisture-holding capacity. Droughtiness is
the main limitation. Crop residue management minimum tillage and cover
crops, help to control wind erosion.
3.1.3.20 CAPABILITY UNIT VIE-1
This unit consists of deep, strongly sloping, severely eroded, moderately
well-drained, medium-textured soils of the Morley series. These soils have
slow to moderate infiltration, slow permeability, and high available mois-
ture capacity. The soils are too steep and too erodible to be suitable for
cultivation, except what is necessary for the establishment of permanent
pasture. Erosion is the main hazard. A vegetative cover and protection
from overgrazing help to control erosion.
Major land uses in the Black Creek Study Area include cropland, 80.7
percent; grassland, 4.3 percent; woodland, 7.1 percent; wildlife and re-
creation, 2.7 percent; urban and built-up, 3.6 percent and farmstead, 1.6
percent. Lands categorized as urban and built-up include the acreages oc-
cupied by the town of Harlan as well as county roads, highways, schools,
and cemeteries. Table 3 shows present acreage for each capability unit in
the Black Creek Study Area.
The pattern of land use is expected to remain relatively stable over
the next five years. However, some minor changes can be anticipated as in-
dicated by recent trends and in response to planned land use. It is es-
TECHNICAL APPROACH
-------
46
TABLE 3
Soils Data By Land Capability - Black Creek Study Area
Land
Capability
Unit
1-1
1-2
IIe-1
IIe-6
IIe-9
Ils-l
IIw-1
IIw-2
IIw-7
IIIe-1
IIIe-6
IIIe-9
IIIe-11
IIIe-13
IIIs-2
IIIw-2
IIIw-6
IVe-6
IVe-11
IVs-1
VIe-1
TOTAL
Acres Major Soils Series
48 Martinsville, Rawson
239 Eel, Genesee
206 Martinsville, Miami, Rawson
1299 Morely
10 Belmore
3 Belmore
4435 Pewamo, Hoytville, Brookston
3698 Blount, Crosby, Haskins
384 Shoals
3 Martinsville, Rawson
127 Morley
5 Fox
50 St. Clair
171 Belmore, Oshtemo
102 Oshtemo
3 Montgomery
1074 Nappanee
137 Morley
24 St. Clair
5 Plainfield
15 Morley
12038
Major Hazard
None
Flooding
Erosion
Erosion
Erosion
Droughtiness
Wetness
Wetness
Wetness
Erosion
Erosion
Erosion
Erosion
Erosion, Droughtiness
Droughtiness
Wetness
Wetness
Erosion
Erosion
Droughtiness
Erosion
timated that urban and built-up acreage will increase by 118 acres as some
of the better-drained woodland and cropland along county roads and highways
are converted to residential use. A net decrease of 143 acres of cropland
and 70 acres of woodland is projected. The acreage used for wildlife and
recreation should increase by 118 acres.
TECHNICAL APPROACH
-------
47
3.2 DESIGN OF LAND TREATMENT
The specific goals of the Black Creek project for treatment of land in
the watershed were established during the development of the work plan in
late 1972 and early 1973.
The concept then was that no traditional practice of soil conservation
should be eliminated from the plan so that each would receive a fair
evaluation as to its potential impact on water quality. In all, 33 prac-
tices were specified for cost-sharing and potential application in the pro-
ject.
These practices and the goals established for them are reported in
Table 4, which is based on Table A-10 of the original work plan.
Assignment of quantities of practice within the general framework of
the project was based largely on the experience of members of the state and
local offices of the Soil Conservation Service. Personnel drew heavily on
experience in the design of small watershed projects and on experience in
previous programs of accelerated land treatment.
In fact, the practices described in Table 4, cover a variety of con-
servation purposes, a point which is developed more completely later in
this report. In general, practices can be classified as:
(a) those which primarily affect water quality,
(b) those which primarily result in protection of the soil resource,
(c) those which primarily lead to increased crop production, and
(d) those which primarily fulfill other conservation purposes.
A decision not to attempt to classify practices in this manner at the
beginning of the project, although most project personnel were aware that
practices could be so classified, represented both lack of experience in
developing plans primarily aimed at improved water quality and a desire to
give each practice a fair evaluation.
In fact, none of the four categories, listed above can fairly be la-
beled as undesirable categories and the relative importance of each must be
considered in planning a program for a specific watershed. In some cases,
practices are complementary. In other cases, there may be a conflict
between opposing uses, e:.cj., it may not be possible to achieve both maximum
water quality and maximum crop production simultaneously.
Similarly, while it may be desirable to consider all four categories
when designing land treatment for a specific watershed, policy decisions
may suggest that funds or technical assistance necessary to carry out cer-
tain practices be obtained from varying sources. This point is also furth-
er developed in Section 5,1.
3.2.1 Role of SCS
The Soil Conservation Service of the U.S. Department of Agriculture
has been responsible for technical assistance to landowners and to the Al-
len County SWCD during the Black Creek project. This assistance has taken
the form of both planning assistance the overall plan for watershed ap-
TECHNICAL APPROACH
-------
00
M
O
n
1
TABLE 4
Land Treatment Goals and Estimated Installation Costs
Land Trent -lent Goals an] r.st ir.ir te J Installation Hosts YEARLY c;3/\LS AND PROJfrT 1 NGTALLAT 1 ON COSTS (HOLLARS)
1 tc.-.i
Land Adequately Treated
Cropland to Grassland
Cropland to Woodlnn.1
Cropland to Wildlife "i Rcc.
Cropland to Other
District Cooperators
|j i s t r i c t Cooperators
Conservation Plans
Conservation Plans
Conservation Plans Revised
Lonservation Plans Revised
Conservation Cropping System
uon tou r Fa n i i ng
Critical 'irea Planting
Crop Residue f'anaf;erien t
Diversions
Farmstead '\ Feedlot Wi ndb reaks
Field Border
Field 1,'indb real-
Grade Stab i 1 i zat i on Structures
Grassed Waterway or Outlet
Molding Ponds u Tanks
Land SMOG t h i nt^
Livestock Exclusion
Livestock Watering Facility
ii i n i nium Till ajie
Pasture "> llayland Management
Pasture a Hayland Planting
Pond
I'rotection During Development
Recreation Area Improvement
Sediment Control Basin
Stream Channe 1 3 tab i 1 i zat i on ( 1 )
Stream Channel S tab i 1 i zat i on t 2 )
Streambank Protection
Str i pcroppi n^
Surface Drains
Terraces, Gradient
Terraces Parallel
Tile Urai ns (3 )
Tree Planting
Wildlife Habitat Management
Woodland .Improved Harvesting
Woodland Improvement
Wood! and Prun i n^
SUBTOTAL
TECHNICAL ASSISTANCE (SCS)
TOTAL INSTALLATION COSTS
I'n i t
Ac.
\c.
\r.
Ac.
Ac.
do .
Ac.
Mo.
Ac .
do .
Ac.
Ac.
Ac.
\c .
v~.
Ft.
Ac.
Ft.
Ft.
[Jo.
Ac.
do .
\c.
\c.
'] n ,
Ac.
Ac.
Ac.
'lo.
Ac .
ilo.
Ft.
Ft.
Ft.
Ac .
Ft.
Ft.
Ft.
Ft.
Ac.
Ac.
Ac.
Ac.
Ac.
Goal
1 ') , 5 7 3
li
11
nil
35
11(8
7,71* 7
170
10, CM
C
1 , U It J
7,I*1C
7b'J
13
7 , VJ 1
39,230
75
288,320
12,030
368
Go
11
330
215
28
7, CSC
1(02
501
39
118
J- ._
6
00,00"
6,000
122,000
300
"0, 500
11,000
11,000
200,300
10
222
200
610
50
Hn i t
Price <
1.50
2.0"
l)0o!oo
1.5"
0. 50
S3. 00
0,3n
".05
5nO. 0"
it 5 0 . 0 0
5, bOO.O"
75.0"
20.. 00
200.0"
G.5"
1 Z . 0 0
70.0"
2,500.00
100.00
200.00
5,000.00
0.50
C . 0 0
2.00
5.00
O.UO
0.25
0. 75
O.UO
80.00
70.00
15.00
20.00
30.00
Total
Tost
11,127
1,53T
i(,000
1 1 , 2 3 r
1 0 , f, " 0
r. , 0 " 0
f. r. , ii 9 G
roo
l£ii,0"0
30,G"0
G1,CO"
22,5""
i*,300
5 , G " 0
i! 0 , 7 G i»
7,23r
35,07"
97,5"0
11, ""0
2, 1*0"
3", 00"
1)5,000
3 G , 0 0 0
2'ii.,000
1,500
36,200
2,750
8,250
CO, 120
800
15,51(0
3,000
12,2"0
1,500
1,169,027
197, 3G1)
1 , 3 G 7 , 1 9 1
Oct. 72-Oct. 73
Goa 1
b/1
ItO
2, OJl*
lu
1,002
3
520
GOO
-
1
coo
2,uOO
-
25,000
-
30
j
1
-
20
2
300
20
50
2
3
-
C
3,000
-
10,000
-
3, 000
-
lu, Obo
-
-
20
20
10
Cost
900
-
'iOO
900
1,3"0
-
7,500
-
15,000
I 5 0
,, uOO
-
1)00
It 00
1,950
3GO
3,500
5,000
300
-
30,000
1,500
_
20,000
-
1,200
-
6,1,27
-
-
300
itOO
300
107,687
UG,500
15it,187
Oct. 73-Qct. 71;
(,oal
2,612
14
20
S
GJ
2 , 1 It 1
53
3,317
3
520
1,978
205
2
1 , 3 'J 7
10,000
21
5 1 , S C C
3,199
si;
18
9
80
1*7
5
2, OUl
£7
121*
8
28
3
23,591*
2,000
32,525
SO
21*, 127
2,933
2,933
50^360
3
59
1*0
11*3
10
(it)
Cost
2,9G7
1*10
800
2,995
5, 000
1,680
15,560
ICO
l)lt, 000
8,100
11,200
6,000
91*0
1,003
13, 2G6
1,566
8,G80
20,000
2,SOO
GOO
11,797
12,000
65,050
1*00
9,G51
733
2,200
20'!!*!*
21*0
l),130
GOO
2,S60
300
277,829
36,81)3
31.',, 672
Oct. 7.')-Oct . 75
Goal
3,533
10
29
11
29
1,538
57
3,589
2,637
273
i*
2,663
13,500
28
1 0 2 , 1) £ 8
1*,2G5
117
21*
1*
107
Gb
q
2,722
11*3
165
11*
1*2
i*
31,1)58
2,000
1*3,365
105
32,170
3,91"
3,911
68^360
1*
79
71)
217
13
Cost
3,955
5iiG
1,GQO
3,995
6,750
2, 2i|0
3 0 , 7 1* 7
213
58,500
10,800
22,1*00
8,025
1,320
1,800
17,691*
2,571)
11,550
35,000
1*,200
800
15,729
12,000
86,730
525
12,8GC
978
2,933
2 7 ', 3 1| 1*
320
5,530
1,110
l*,3!*0
390
395, 50G
55,291
1*50,797
Oct. 75-Oct.76
Goal
3,757
1(5
1C
19
977
'*!*
2,733
2,203
291
3
2,231
13,100
26
108, 9GG
2,536
133
17
1*
113
82
12
2,593
152
162
15
1*5
5
_
31,91,8 '
2,000
36,110
115
31,203
l*,157
1*, 1 56
G5'512
3
81)
66
230
17
Cost
3,305
582
1,200
3,31)6
R,550
2,080
32,6"9
227
G6,50"
7,650
22,1)00
8,1)75
1,61)0
2,1*00
16,851*
2,736
11,31*0
37,500
l),500
1,000
_
15,971)
12,000
72,220
575
12,l*,"l
1,039
3 117
26^205
21*0
5,880
990
1* , G 0 0
510
388,805
1*5,705
1)31*, 510
Oct.76-0ct.77
Goal
Cost
15,025
13,025
-------
49
plication, individual conservation plans, etc. and technical engineering
assistance, j^js., the design of practices to be applied.
Over the five years of the project, SCS estimates that 22,793 man-
hours were applied to the Black Creek project. Of these hours, 14,509 in-
volved planning or engineering assistance. This further can be broken down
into 6,099 for planning time, 4,400 hours in application, and 3,577 hours
of planning and application on projects which involved more than one lan-
downer. The balance represents administration and special sources.
Planners developed 133 individual conservation plans for landowners
during the project. This breaks down to about 46 hours per plan or a cost,
based on reimbursement to SCS, of nearly $340 per plan.
These time and cost figures are critical to understanding the approach
taken to planning for improved water quality in the Black Creek Watershed.
Individual conservation plans served as the basis for contracts with lan-
downers and resulted in the "total approach" of the Black Creek project as
opposed to a less coordinated "practice-at-a-time" approach which has been
criticized in other areas.
More detail on costs of technical assistance by SCS is included in the
section of report on costs.
3.2.2 Procedures of the SWCD Board
Soil and Water Conservation Districts have cooperated with county ASCS
Committees for many years in allocating funds for soil conservation pro-
grams under USDA projects. In the case of the Black Creek project, funds
were controlled directly by the Soil Conservation District. Cost sharing
rates were flexible, subject only to the requirement that the federal share
of the total project not exceed 75 percent of the total project cost.
As a result, an early responsibility of the Board of Supervisors was
to establish cost sharing rates for practices to be applied during the
Black Creek project. In order to do this, the Board ranked project goals
as easy, normal, moderately difficult, difficult or very difficult to
achieve. The initial rankings, as determined by the Board, are listed in
Table 5.
Cost-share rates were based on this evaluation. Higher rates were set
on practices which, it was assumed, would be difficult to convince farmers
to use. In general, the initial determination of the Board of Supervisors
proved to be accurate. There were exceptions, however. For example, field
borders, which the Board thought would be very difficult to apply, proved
to be popular with rural landowners. The initial cost-share rate, set for
field borders by the Board, was reduced before the end of the project.
In other cases, as much as 90 percent cost sharing was paid to achieve
cooperation. It is doubtful that the project would have been as successful
as it was if the Board had not retained this ability to make changes in
cost-share rates or other incentives.
To make sure that uniform procedures were followed in the conduct of
the project, an operations manual was prepared by the project staff and
adopted by the Board. The operations manual provided a guide for dealing
with anticipated problems that could arise during the program. It spelled
out procedures for actions to be taken in the event a contract was not fol-
TECHNICAL APPROACH
-------
50
TABLE 5
Degree of Difficulty for Black Creek Conservation Goals
District Cooperators
Conservation Plans
Land Adequately Treated
Conservation Cropping System
Critical Area Planting
Crop Residue Management
Diversions
Farmstead and Feedlot Windbreak
Field Border
Field Windbreak
Grade Stabilization Structure
Grassed Waterway or Outlet
Holding Ponds and Tanks
Livestock Exclusion
Livestock Watering Facility
Minimum Tillage
Pasture and Hayland Management
Pasture and Hayland Planting
Pond
Recreation Area Improvement
Sediment Control Basin
Stream Channel Stabilization
Streambank Protection
Stripcropping
Surface Drains
Terraces, Gradient
Terraces, Parallel
Tile Drains
Tree Planting
Wildlife Habitat Management
Woodland Improved Harvesting
Woodland Improvement
Easy
Normal
Normal
Difficult
Moderately Difficult
Difficult
Difficult
Normal
Very Difficult
Normal
Moderately Difficult
Difficult
Easy
Moderately Difficult
Moderately Difficult
Difficult
Difficult
Difficult
Easy
Normal
Difficult
Moderately to Very Difficult
Moderately Difficult
Very Difficult
Difficult
Very Difficult
Very Difficult
Difficult
Very Difficult
Moderately Difficult
Moderately Difficult
Difficult
lowed by participating landowners, what to do in the event of the death of
a cooperator, and what the responsibilities of all parties were in the
event that land ownership changed.
Adoption of the operations manual made it possible to avoid ad hoc de-
cisions. Even though all of the eventualities anticipated in the manual
did not, in fact, take place, the Board considers that the time spent on
its preparation was worthwhile.
Administrative control of the project was retained by the Board of Su-
pervisors, even though the day-to-day operation was delegated to the pro-
ject director. This control was maintained by requiring Board approval at
several critical stages in the process of working with cooperators. The
Board was required to approve each initial agreement in which a landowner
became a "cooperator" of the district. The Board further was required to
approve the conservation plan on which compliance and possible cost-sharing
TECHNICAL APPROACH
-------
51
was based. This approval came in the form of a contract with the landown-
er. Finally, the Board was required to approve payment of claims made
under the contract. In this way, the Board was not only regularly informed
of the progress being made under the project, but also had the opportunity
to make changes, if necessary, in procedures, payment rates, and certifica-
tion of compliance.
3.3 TECHNICAL APPROACH
Detailed monitoring of water quality and other parameters has been
carried out since the beginning of the Black Creek Project. This was
necessary to fulfill the purpose of understanding the mechanisms whereby
land use affects water quality. Monitoring was carried out by both Purdue
and University of Illinois personnel. Both carried out water quality moni-
toring, and each investigated other parameters. The monitoring scheme for
the Black Creek watershed is described in this section.
3.3.1 Sampling
3.3.1.1 MAJOR MONITORING SITES PURDUE
Purdue has collected regular grab samples at the sites listed here.
These locations are identified in Figure 7.
3.3.1.1.1 Water Quality Sampling Sites
1. Killian Drain at Notestine Road
2. Smith-Fry Drain at Notestine Road
3. Wertz Drain at Notestine Road
4. Gorrell Drain at Notestine Road
5. Richelderfer Drain at Notestine Road
6. Black Creek at Brush College Road
7. Lake Drain at Bull Rapids Road
8. Wertz Drain at St. Hwy. #37
9. Dreisbach at Trammel Road
10. Dreisbach at St. Hwy. #37
11. Fuelling Drain at Ward Road
12. Black Creek at Ward Road
13. Wann Drain at Killian Road
14. Maumee River at St. Hwy. #101
15. Fuelling Drain below sediment basin
16. Richelderfer Drain at Stopher Road
17. St. Mary's River at Ferguson Road
18. St. Joseph River at Van-Zile Road
19. Dreisbach at Cuba Road (end waterway)
Rainfall amount was measured regularly at the following seven loca-
tions. Periodically samples of rainfall were analyzed for water quality
parameters.
TECHNICAL APPROACH
-------
52
Figure 7. Purdue Monitoring Sites
3.3.1.1.2 Raingage Sites
TECHNICAL APPROACH
-------
53
1. Approximately 1/8 mile west of Cuba Road on Springfield Center
Road.
2. Southwest corner junction Springfield Center Road and Spencerville
Road.
3. Approximately 1/4 miles east of Spencerville Road on Grabill Road.
4. Northwest corner junction of St. Hwy. #37 and Ruppert Road.
8. Approximately 1/4 mile south of junction Notestine Road and Rup-
pert Road.
9. Fuelling Drain and Shaffer Road.
20. 1/4 mile east of St. Hwy. #101 and Gar Creek Road ('75).
Two sediment basins, the smaller being designated a desilting basin,
were constructed for study of the impact of this practice on water quality.
These were located the following locations.
TECHNICAL APPROACH
-------
54
3.3.1.1.3 Sediment Basin Study Sites
1. Fuelling Drain east of Shaffer Road
2. Desilting basin in Black approximately 3/4 mile east of Shaffer
Road and 1/4 mile south of St. Hwy. #101
Stage recorders and other methods of measuring stream flow were in-
stalled at nine locations in the watershed. Stream flow data is necessary
to convert concentrations measured of the various parameters into loadings.
These measurements were made at the following locations.
3.3.1.1.4 Stage Recorder, Flow Measuring Sites
1. Killian Drain at Notestine Road
2. Smith-Fry Drain at Notestine Road
3. Wertz Drain at Notestine Road
4. Gorrell Drain at Notestine Road
5. Richelderfer Drain at Notestine Road
6. Black Creek at Brush College Road
12. Black Creek at Ward Road
13. Wann Drain at Killian Road
15. Fuelling Drain at sediment basin outlet
Backwater due to flow restrictions in the lower reaches of Black Creek
has on occasion caused flow measurement problems at stations 2 and 12 and
to a lesser degree at other sites. Stream flow is determined by measuring
stage and using a flow rating curve; however, during periods of backwater
influence the stage vs. flow relationship can change unpredictably.
A deflection vane current meter was developed by the Department of
Agricultural Engineering to provide improved flow determination during
periods of backwater influence. The current meter measures water velocity
at a fixed depth near the center of the stream. The additional flow velo-
city information is combined with stage data to improve the accuracy of the
flow rate determination. The meter is also capable of measuring a limited
range of back-flow velocities should such conditions occur.
The current meter senses flow velocity by a spring-loaded vane which
is mounted perpendicular to the stream flow. A gearing mechanism provides
two digital lines which indicate each 3.6 degrees of vane deflection and
the direction of that movement. An analog signal proportional to vane de-
flection is also available. Of course, meter readings are not meaningful
until the deflection vane is fully inundated. A break-away mounting brack-
et is used to prevent damage in the event that large floating objects
strike the meter.
Current meters were installed at sampling sites 2, 6 and 12 during
January, 1978.
A question posed at the beginning of the study concerned the impact of
tile drainage on water quality. The following sites were utilized to col-
lect samples to measure concentrations and loadings from tile drains.
TECHNICAL APPROACH
-------
55
3.3.1.1.5 Tile Drain Outfall Studies (approximate outlet locations)
R29 Richelderfer Drain at Notestine Road
B59 Black Creek at junction Darling Road and Bull Rapids
L6 Lake Drain at Bull Rapids Road
Fll Fuelling Drain Along Ward Road 1/4 mile west junction Shaffer
Road
G13 200 feet east of Gorrell Drain and Notestine Road along Gorrell
Drain
Weo 600 feet east of Wertz Drain and Ruppert Road along Wertz Drain
SF2 Smith-Fry Drain and Notestine Road
K7 Killian Drain and St. Hwy. #101
SF26 Smith-Fry Drain and St. Hwy. #101
K65 Killian Drain and Antwerp Road
Wa2 300 feet west Warm Drain and Killian Road along Warm Drain
SF51 Smith-Fry Drain and Knouse Road
SF61 Smith-Fry Drain and Ruppert Road
SF79 Smith-Fry Drain and Springfield Center Road
We69 Wertz Drain and St. Road #37
G63 Gorrell Drain and Antwerp Road
R59 Richelderfer Drain and Hamm Road
D71 Dreisbach Drain and Antwerp Road
D122 Dreisbach Drain and Cuba Road at waterway outlet
D116 Dreisbach Drain at junction Grabill Road and Cuba Road
#20 1/4 mile east junction Gar Creek Road and St. Hwy. #101 along
Miller Ditch
The effect of bank slope and mulch on ditch bank erosion was investi-
gated systematically at two locations. These are identified and described
below.
3.3.1.1.6 Ditch Bank Mulch Study Sites
1. Joe Graber Farm
Slopes 2:1, 3:1, 4:1
Mulch Straw, stone, wood chips, no mulch, aquatain, saw dust
Seeding Mixture tall fescue and red top
2. Dick Yerks Farm
Slopes 2:1, 3:1, 4:1
Mulch Straw, stone, wood chips, no mulch
Seeding Mixture tall fescue and red top
The Agricultural Research Service rainfall simulator was used to con-
duct several tests of the relationship between parameters such as soil
type, slope, tillage, surface cover and erosion. Primary location of this
work are listed below.
3.3.1.1.7 Rainulator Study Sites
1. Virgil Hirsch Farm
Nappanee silt loam; 1% slope or less
2. Virgil Hirsch Farm
Hoytville silty clay loam; 1% slopes or less
TECHNICAL APPROACH
-------
56
3. Dick Yerks Farm
Haskins sandy loam; 1% slope or less
4. Dennis Bennett Farm
Morley silt loam; 5% slopes
5. Graber Farm
Morley silt loam; 5% slopes
Tillage research utilizing replicated plots and several combinations
of tillage was carried out at the following locations.
3.3.1.1.8 Tillage Study Sites
1. Max Woebbeking Farm Nappanee silty loam
2. Bill Shanebrook Farm Hoytville silty clay loam
3. Oliver Stieglitz Farm Whitaker loam
4. Roger Ehle Farm Rensellaer silty clay loam, Whitaker loam, Osh-
temo fine sandy loam
5. Dennis Bennett Farm Morley silt loam
6. Bill Schaefer Farm Haskins loam
Automatic (PS-69) pump samplers were utilized at four locations to ob-
tain data useful for describing variations in concentrations and loadings
during a storm event. These sites are listed here.
3.3.1.1.9 Automatic Water Sampling Stations
2. Smith-Fry Drain at Notestine Road
6. Black Creek at Brush College Road
12. Black Creek at Ward Road
20. 1/4 mile east junction Gar Creek road and State Hwy. 101 along
Miller Ditch (abandoned after six months due to vandalism)
An automated station capable of recording weather and stream flow data
for itself and of recording and reporting data from four remote substations
was installed. The locations for major elements of this system are listed
here.
3.3.1.1.10 Hydrological Remote Sensoring Site
1. Black Creek at Brush College Road
3.3.1.1.11 Substations
1. Smith-Fry Drain at Notestine Road
2. Wertz Drain at Notestine Road
3. Gorrell Drain at Notestine Road
4. Richelderfer Drain at Notestine Road
TECHNICAL APPROACH
-------
57
3.3.1.2 SAMPLE STATIONS U of !_
3.3.1.2.1 Fish Sample Stations
Twenty-nine sampling stations were established throughout the
watershed and in nearby streams for fish studies. All (except station 14)
included 100 m of stream channel. Some stations were dropped after initial
studies because of special stream conditions, intermittent nature of
stream, lack of fish or other reasons.
1. Killian Drain above Notestine Road
2. Smith-Fry Drain above Notestine Road
3. Wertz Drain above Notestine Road
4. Upper Gorrell Drain above Notestine Road
5. Richelderfer Drain above Notestine Road
6. Dreisbach Drain above Brush College Road
7. Lake Drain above Bull Rapids Road
8. Wertz Drain above State Highway #37
9. Dreisbach Drain above Trammel Road
10. Dreisbach Drain above State Highway #37
11. Fuelling Drain below Ward Road
12. Black Creek above Ward Road
13. Wann Drain above Killian Road
14. Maumee River at State Highway #101
15. Black Creek Downstream from entry of Smith-Fry Drain
16. Wertz Drain between Notestine Road and Black Creek
17. Black Creek immediately upstream from entry of Wertz Drain
18. Black Creek immediately below entry of Richelderfer Drain
19. Smith-Fry Drain between Indiana Highway #101 and Antwerp Road
(northeast corner of intersection)
20. Dreisbach Drain above Antwerp Road
21. Hamm Interceptor at Notestine Road
22. Black Creek below entry of Richelderfer Drain
23. Maumee River at entry of Black Creek
24. Black Creek above Maumee River
25. Black Creek below Ehle Road
26. Black Creek below Darling Road
28. Black Creek below Wertz Drain
29. Black Creek immediately downstream of Smith-Fry Drain
Stations I, 4, 5, 7, 8, 11, 14, 16, 21-25 were dropped after initial
studies.
3.3.1.2.2 Water Sample Stations
University of Illinois water quality monitoring stations were esta-
blished on the main stem of Black Creek, along Wertz drain, on Dreisbach
Drain, and at other locations in the watershed. These are listed below and
are located in Figure 8.
Dreisbach Drain
101 At Springfield Center Road, North side of road above drop structure.
Grass waterway.
TECHNICAL APPROACH
-------
58
/
/
/
/
s
Figure 8. Illinois Monitoring Sites
102 At Cuba Road, 50 feet above bridge east side of road; lower end of
grass waterway.
TECHNICAL APPROACH
-------
59
103 Ditch channel at Cuba Road east side of road.
104 Lower end of rip-rap at bridge west side of Cuba Road.
105 Along stream west of road at Cuba and Grabill Roads intersection; just
above several "rock ledges."
106 Stream channel at above intersection; several feet above entry of wa-
ter at drop structure from field to the west.
107 Water flowing over the drop structure at northwest corner of Cuba and
Grabill Roads.
108 Water flowing from drain tile just south of drop structure Tile D-116.
109 Stream above bridge at Grabill Road
110 At survey flag 100 feet below swimming pool and across stream from me-
tal tile.
Ill At fence post on edge of Joe Graber farm. Upstream about 40 feet from
marker flag 104+00.
Stations 110 and 111 on relatively straight channel with steep eroding bank
slope.
112 Just above bridge on Joe Graber farm and short distance above "tile"
outlet on west side. Downstream 30 feet from 107+00 marker flag. Channel
above this station has shallow slope bank.
113 About 10 feet above rock area above Trammel Road on Dreisbach Drain.
Channel above in grass with medium slope to bank.
114 40 feet upstream from bridge at Antwerp Road at middle to low end of
long slow pool.
115 At lower end of rip-rap on west side of stream.
116 40 feet below sewage treatment effluent tile from trailer park.
117 10 feet north of south end of "willow" thicket along stream.
118 Just above bend in stream 3/8 mile above Notestine Road.
119 At south end of fence row but slightly above outlets of three drain
tiles. Area with medium sloped bank.
120 40 feet above bridge. Bank with medium slope.
Stations 119 and 120 both on grassy channel with field border on both
sides.
121 Sample at middle of fish sample station number 6B; about 150 feet
above rip-rap.
122 At lower end of rip-rap.
TECHNICAL APPROACH
-------
60
123 Under bridge at Brush College Road.
Black Creek
124 Lower end of rip-rap below Brush College Road.
125 Fourth telephone pole downstream from bridge; end of straight channel
section with grass channel and field border.
126 Above rip-rap at bend to north; somewhat meandering channel; grass
channel with field border.
127 30 feet west of bridge at entry of Richelderfer Drain to Black Creek.
Meandering channel with grass banks.
128 On Richelderfer Drain 40 feet above bridge on Darling Road. Grass
channel with field border.
129 Downstream on Black Creek about 50 feet from entry of Richelderfer;
Station 18 of fish studies.
130 40 feet above rock (rip-rap) at bend.
131 40 feet above rock (rip-rap) area just before stream turns to east.
132 20 feet above entry of Wertz Drain to Black Creek.
133 At top of riffle area in Station 28 of fish studies.
134 At bottom of riffle area in Station 28 of fish studies.
135 Discharge from surface flow from field to north.
136 Discharge from tile outlet from terraces in field to west.
137 At survey flag #395 in area with forest along both banks.
138 At lower end of forest on both banks.
139 At lower end of field on one side of stream; forest on other bank.
140 Just above Schaeffer Road bridge; heavy sewage inflow at house above
Schaeffer»Road.
141 Black Creek about 15 feet above entry of Smith-Fry Drain.
142 At lower end of Station 15A; woodland on west side, field on east
side.
143 At lower end of rip-rap and steep erosion bank.
144 At lower end of wooded area both sides above sediment basin.
145 20 feet above bridge at sediment basin.
TECHNICAL APPROACH
-------
61
146 At lower end of sediment basin.
147 About 200 meters below the sediment basin.
148 About 180 meters below station 147, and about 100 feet above flag #470
which is where the stream turns to the west.
149 At flag #477.
Comment Large pool at flag #480.
150 Near south end of woods but above the small drainage channel entering
from the west.
Comment Flag #482 is about 300 to 400 feet south of edge of both sides
forest.
151 Just above rock (rip-rap) area in fields.
152 At upper end of Station 12 of fish studies.
153 At lower end of Station 12 just above Ward Road.
154 At north edge of thicket area.
155 At telephone pole in field above bridge at Ehle Road west side.
156 Downstream from bridge at Ehle Road about halfway to river
157 Maumee River upstream from Black Creek.
Wertz Drain
158 Wertz Drain just above Boger Road; at lower end above woodlot.
159 Above Springfield Center Road. Below area of grass waterway.
160 Above Spencerville Road.
161 Overflow from drop structure at lower end of grass waterway.
162 Flow from underground pipe of waterway; east pipe.
163 Station in forest at market flag #136; moved upstream of terrace #218
outlet.
164 At fence at flag #138; some locale below first rock drop.
165 Near Bull Rapids road; flag placed on east side of stream.
166 About 400 feet downstream from Bull Rapids Road; short distance
upstream from large shagbark hickory
167 Large pool at location of large tree on west bank.
TECHNICAL APPROACH
-------
62
168 At power line crossing.
170 Flag marks location of waterfall (at upper end); one station (169) 10
feet upstream from flag; other station (170) 30 feet downstream.
171 (Old Station 12) At upper end of pool about 100 meters downstream
from Knouse Road near metal fence post on east bank.
172 (Old Station 11) In pool where stream turns to east as you move
downstream. A pile of rocks and bricks are on the west bank at this point.
173 (Old Station 10) In pool about 200 meters upstream from bridge. No
real landmarks here so distance from forest should be measured.
174 (Old Station 9) In pool about 10-20 meters inside woods even
with double trunk tree laying in stream channel.
175 (Old Station 8) In riffle 10 meters below tree laying across
stream.
176 (Old Station 7) In small rocky pool 5 meters upstream of large
ditch on east bank even with triple trunk tree on west bank.
177 (Old Station 6) In first pool inside south end of forest; about 20
meters downstream of fallen tree across stream.
178 (Old Station 5) In pool in grove of trees 30-50 meters downstream
of woods at point below two large cotton woods on east side of bank. Also
fallen tree on field border on west side.
179 (Old Station 4) In pool at lower edge of grove of trees and even
with north edge of trees in field 100 meters east of stream.
180 (Old Station 3B) In pool at lower edge of grove of trees and even
with north edge of trees in field 100 meters east of stream.
181 (Old Station 3A) In pool about midway between large tile downstream
and grove of trees upstream; even with south edge of trees in field 100 me-
ters east of stream. Small stump on east bank.
182 (Old Station 2B) In pool about midway between large erosion
183 (Old Station 2A) In riffle area below large west bank erosion area
at point where trees are not present on east bank.
184 (Old Station 1C) In pool, even with small tree on west side of
stream below grove of trees.
185 (Old Station IB) In pool, even with small tree on west side of
stream below grove of trees.
186 (Old Station 1A) In a small pool adjacent to tree on east bank.
A more detailed map of water sample stations in Wertz Woods area is at-
tached .
TECHNICAL APPROACH
-------
63
187 100 feet downstream from fence.
188 Above bridge (some distance) above Ruppert Road.
189 Just above Notestine Road
190 Just above entry of Wertz Drain into Black Creek Smith-Fry Drain.
191 Just upstream from Highway 37 Bridge
192 At point where stream bends to east as one moves downstream.
193 Downstream about 20-30 meters from Ruppert Road.
194 Midway between Ruppert and Krouse Roads.
195 Just above Knouse Road.
196 Just above Dean Road
197 Just west of Highway 101, above Antwerp Road.
198 Just above Antwerp Road.
199 Just west of Highway 101 below Antwerp Road.
200 Just above fence at upper end of wooded area.
201 At upper end of old field thicket, pasture area, drain tile enters
between 200 and 201.
202 40 yards above farm bridge crossing and just above two tile inflows;
orange marker flag at site.
203 Just above eroding edge of field of east side of stream; 120 feet
above forest on east side; at lower end of pool.
204 In cleaned forest area about 20 meters upstream from the tile inflow.
205 Water from large tile inflow.
206 At lower end of cleared area on west side of stream; 10 to 15 meters
above fence; across from fallen tree.
207 In middle of forested, channeled section of stream, location to be
determined by Dan Dudley.
208 Lower end of woodland slightly upstream from large bank slippage area.
209 Lower end of eroded bank below woodland area.
210 At lower end of straight area 25 meters above upper end of station 2
fish sample.
211 Lower end of fish sample station.
TECHNICAL APPROACH
-------
64
212 Entry of Smith-Fry Drain into Black Creek.
Other Stations in Watershed
213 Wertz Ditch entry to Wertz Drain. From east side between Route 37 and
Knouse Road.
214 Smith-Fry Drain above bridge at Springfield-Center Road.
215 Warm Ditch above Knouse Road.
216 Small terrace just south of station 136 with drains about 20 acres.
217 Terrace area on Dreisbach Drainage located just north of Trammel Road
and east of stream channel.
218 Terrace on Wertz Drain just north of station 163.
219 Terrace on Dreisbach Drain north of station 102.
220 Black Creek at downstream end of channel modification below Ehle Road.
221 Warm Ditch located just upstream of first bend above station 215.
3.3.1.3 AUTOMATIC PUMPING SAMPLERS
Automatic pumping samplers (PS-69) were installed at stations 2, 6,
and 12 in the Black Creek Watershed and have been operating since February
22, 1975, March 17, 1975, and April 4, 1975, respectively. A sampler is
pictured in Figure 9. The stations are located in the lower section of the
watershed to allow sampler data to closely describe movement of water qual-
ity constituents from the watershed. The physical operation of the
samplers has been very satisfactory (one even continued to operate while
being inundated by a severe storm).
The samplers are energized only while the stage in their respective
stream is above the one foot level. While energized, the samplers take a
500 ml water sample every thirty minutes with a seventy-two sample capacity
(36 hours). The samples collected are analyzed for the same water quality
parameters as the "grab" samples, namely, ammonium, nitrate, total nitro-
gen, soluble nitrogen, inorganic phosphorus, total phosphorus, soluble
phosphorus, and suspended solids.
3.3.1.4 AUTOMATIC TILE SAMPLER
An automatic tile sampler has been operational since March, 1976. The
operation of the tile sampling station is unique in that discrete water
quality samples are collected proportionally to the tile outflow, which is
continuously monitored and recorded. The time in which a sample is col-
lected is also recorded on the flow hydrograph chart. The sampler has the
capability of collecting 72, 500 ml water samples. The sampling rate is
approximately 1 sample per 30 minutes at maximum flow.
Another feature of the tile sampling station is the prevention of the
tile outlet from becoming inundated. This is necessary to provide reliable
TECHNICAL APPROACH
-------
65
Figure 9. Automatic Pumping Sampler (PS-69)
data during storm events in which the ditch water level is above the tile
outlet. Two 200 GPM pumps used in conjunction with a sump maintain a free
TECHNICAL APPROACH
-------
66
water fall over the flow calibrated weir. The tile sampler is described in
below.
3.3.2 Laboratory Analysis
3.3.2.1 PURDUE
Analysis of samples at Purdue University has been carried out accord-
ing to the following procedures.
1. Suspended solids: solids in an aliquot of the water sample (250 ml)
are collected on a tared Nucleopore membrane (47 mm diameter, 0.4 14m mean
pore size), dried at 105 degrees C, and filter is weighed. Reported as mg
suspended solids per liter of water.
2. Total nitrogen: The method of Nelson and Sommers (1975) is used to
determine total N in filtered and unfiltered aliquots of water samples.
3. Ammonium -N: The method of Bremner and Keeney (1965) was used to
determine ammonium -N in filtered water samples.
4. Nitrate plus nitrite -N: The method of Bremner and Keeney (1965) was
used to determine (nitrite + nitrate)-N.
5. Soluble inorganic phosphorus: Reactive inorganic phosphorus was deter-
mined in filtered water samples by the method of Murphy and Riley (1962).
6. Total phosphorus: Total P in filtered and unfiltered water samples was
determined by the perchloric acid digestion method of Sommers and Nelson
(1972).
7. Total organic carbon: Total organic C in filtered and unfiltered water
samples were determined with a Dohrman Envirotech DC-50 Carbon Analyzer.
8. Selected alkali, alkaline earth, and transition metals: Ca, Mg, Na, K,
Cu, Ni, Zn, Cd, and Pb were determined in filtered water samples by atomic
absorption spectrophotometry.
9. Other water parameters: temperature, dissolved oxygen, alkalinity, pH,
and conductivity were determined in situ or immediately after sampling by
methods described by the American Public Health Association (1971) .
Calculations used in estimating constituents in water samples:
Soluble organic N - Total N(filtered) - (nitrate + nitrite + ammonium)-N
Sediment N = Total N(unfiltered) - Total N(filtered)
Soluble organic P = Total P(filtered - soluble inorganic P
Sediment P = Total P(unfiltered) - Total P(filtered)
Sediment organic C = Organic C(unfiltered - Organic C(filtered)
References
American Public Health Association, 1971. Standard Methods for the Exami-
nation of Water and Wastewater. 13th ed. American Public Health Associa-
TECHNICAL APPROACH
-------
67
tion, Washington D.C.
Bremner, J.M. and D.R. Keeney, 1965. Stream Distillation Methods for
Determination of Ammonium, Nitrate, and Nitrite. Anal. Chem. Acta.
32:485-495.
Murphy, J. and J.P. Riley, 1962. A modified single solution method for the
determination of phosphate in natural water. Anal. Chem. Acta. 27:254-
267.
Nelson, D.W. and L.E. Sommers, 1975. Determination of total nitrogen in
natural waters. J. Environ. Qual. 4:465-468.
Sommers, L.E. and D.W. Nelson, 1972. Determination of total phosphorus in
soils: A rapid perchloric acid digestion procedure. Soil Sci Soc. Amer.
Proc. 36:902-904.
3.3.2.2 UNIVERSITY OF ILLINOIS
Analysis at the University of Illinois was carried out at the Illinois
Natural Survey according to the procedure listed in Table 6.
TABLE 6
University of Illinois Analytic Methods
"ParametersMethods
Total Alkalinity (as CaC03) Metrohm Autotitrator to pH 4.6*
Total Dissolved lonizable Solids By Calculation from Specific Conductance
(as NaCl) Table
EDTA Hardness (as CaC03) EDTA Colorimetric Method (Autoanalyzer)
Turbidity (JTU) Monitek Model 150 Turbidimeter
Total Phosphorus (as P) Stannous Chloride Method*
Soluble Orthophosphate (as P) Ascorbic Acid Method (Autoanalyzer)*
Nitrate (as N) Cadmium Reduction Method (Autoanalyzer)*
Nitrite (as N) Diazotization Method (Autoanalyzer)
Ammonia (as N) Berthelot Reaction Method (Autoanalyzer)
Organic Nitrogen (as N) Modified Berthelot Reaction Method
(Autoanalyzer)
Sulfate (as S) Turbidimetric Method*
Residue, Total Constant Weight Upon Drying @ 180 C,
Unfiltered*
* Standard Methods
American Public Health Association, American Water Works Association,
and Water Pollution Control Federation.
1975.
Standard methods for the examination of water and wastewater,
14th ed. Washington, D. C. 1193 pp.
MERCURY - WATER
Procedure: A 50-ml aliquot of unfiltered water is treated with 10-ml of 5%
TECHNICAL APPROACH
-------
68
potassium sulfate and digested for 3 hours at room temperature. A 10-ml
aliquot of the digested sample is transferred to the reduction vessel of
the Mercury Analyzer. One ml of 5% tin chloride is added to the vessel and
a 2 minute reaction time is allowed before the Hg vapor is swept through
the absorption cell of the analyzer. Quantitative results are obtained by
comparison with standard Hg solutions of 1, 10, and 50 ppb Hg treated in a
like manner to the samples.
Equipment: Fisher Mercury Analyzer, Varian Model 485 Digital Integrator
Detection Limit: 0.2 ppm (yg Hg/ml water)
Reference: A. A. El-Awady, R. B. Miller, M. J. Carter, Analytical Chemis-
try 48 (1), 110-117 (1976).
WATER CATIONS - (Ag, Al, As, B, Ba, Be, Cd, Co, Cr, Cu, Fe, Hg, Mg, Mn, Mo,
Ni, Pb, Sb, Se, Si, Sn, V, Zn.)
Procedure: The sample was analyzed by emission spectrometry using a
radio-frequency inductively-coupled argon plasma as the source of radia-
tion. Standard solutions of the elements desired are used in calibrating
the instrument which is controlled by a mini-computer.
Equipment: Sartorius Membrane Filter Holders, Jarrell-Ash Model 975 Plasma
AtomComp
Detection limits are presented in Table 7.
TABLE 7
Detection Limits (micrograms metal/ml water)
Ag
Al
As
B
Ba
Be
.007 ppm
.051
.045
.004
.001
.001
Cd
Co
Cr
Cu
Fe
Fe
.004 ppm
.006
.007
.004
.020
.020
Mg
Mn
Mo
Ni
Pb
Pb
.001 ppm
.002
.007
.011
.033
.033
Se
Si
Sn
V
Zn
Zn
.039 ppm
.030
.028
.007
.009
.009
Reference: U. S. Environmental Protection Agency, Methods for the Analysis
of Water and Wastewater, 1974.
TOTAL Hg - SEDIMENT
Procedure: A sample of sediment is centrifuged at 3,000-rpn for ten
minutes and the interstitial water decanted. A 2-g sub-sample is then
placed in a 250-ml Erlenmeyer flask to which 25-ml of aqua regia is added.
The flask and sample are heated to boiling and boiled for two minutes.
After it has cooled to room temperature, the flask has 25-ml of 5% potassi-
um promangenate and 2-ml of 5% potassium sulfate added to it. The sample
is then digested for 30 minutes in a 95 degree C water bath. The digested
solution is treated with 10% ammonium hydrochloride to reduce the excess
TECHNICAL APPROACH
-------
69
potassium promangenate. The liquid is transferred to a 100-ml volumetric
flask, the residual sediment washed several times with deionized water, the
washing added to the volumetric flask, and the solution diluted to volume
with deionized water.
A 10-ml aliquot of the digested samples is transferred to the reduc-
tion vessel of the Mercury Analyzer. One ml of 5% tin chloride is added to
the vessel, and a two minute reaction time is allowed before the Hg vapor
is swept through the absorption cell of the analyzer. Quantitative results
are obtained by comparison of the values obtained with the results of stan-
dard Hg solutions of 1, 10, and 50 ppb Hg treated in a like manner to the
samples.
Equipment: precision 67390 Centrifuge, Blue M Model 1130A Water Bath, Fish
Mercury Analyzer, Varian Model 485 Digital Integrator
Detection Limits: 2 ppb (ygHg/g dry sediment)
Reference: I. K. Iskandar, D. R. Keeney, Environmental Science and Tech-
nology, 8 (2), 165-170 (1974).
L. W. Jacobs, D. R. Keeney, ibid, 8 (3), 267-268 (1974).
ACiD EXTRACTABLE CATION - SEDIMENTS AND SOILS (Ag, Al, As, B, Ba, Be, Cd,
Co, Cr, Cu, Fe, Hg, Mg, Mn, Mo, Ni, Pb, Se, Si, Sn, V, Zn.)
Procedure: A 5-g representative aliquot of the sample is weighed into a
50-ml plastic centrifuge tube, and 20-ml of 0.05 N acid mixture (0.01 N
sulfuric acid + 0.04 N HC1) is added. The tube is shaken mechanically for
one hour and then centrifuged at 20,000-rpm for ten minutes. The superna-
tant is carefully decanted and analyzed by emission spectrometry using a
radio-frequency inductively-coupled argon plasma as the source of radia-
tion. Standard solutions of the elements desired are used in calibrating
the instrument which is controlled by a mini-computer.
Equipment: Reciprocal shaker, Beckman Model J-2IB Centrifuge, Jarrell-Ash
Model 975 Plasma AtomComp
Detection Limits are presented in Table 8.
TABLE 8
Detection Limits in Micrograms Metal/Gram Dry Sediment
Ag
Al
As
B
Ba
Be
.03 ppm
.21
.18
.02
.004
.004
Cd
Co
Cr
Cu
Fe
Hg
.16 ppm
.02
.03
.02
.08
2.0
Mg
Mn
Mo
Ni
Pb
Sp
.004 ppm
.01
.03
.04
.13
.19
Se
Si
Sn
V
Zn
.16
.12
.11
.03
.60
ppm
Reference: Perkin-Elmer Analytical Methods for Atomic Absorption Spectro-
photometry, 1976.
TECHNICAL APPROACH
-------
70
MERCURY - SOFT TISSUES
Procedure: The tissues are freeze-dried and pulverized so that a relative-
ly homogeneous 0.5-g sample can be removed for analysis. The sample is
weighed into a 125-ml Erlenmeyer flask, which is placed in an ice bath, and
to which is added, in order, 20-ml of concentrated sulfuric acid, 10-ml of
concentrated nitric acid, and 20-ml of 5% potassium promangenate. The sam-
ples are then allowed to stand 15 minutes at room temperature after which
10-ml of 5% potassium sulfate is added. The samples are then digested two
hours in a 95 degree C water bath. Small amounts of solid potassium
promangenate are added as needed to maintain an oxidizing environment.
After digestion and cooling, the sample is diluted to volume in a 100-ml
volumetric flask with deionized water.
A 10-ml aliquot of the digested sample is transferred to the reduction
vessel of the Mercury Analyzer. One ml of 5% tin chloride is added to the
vessel and a two minute reaction time is allowed before the Hg vapor is
swept through the absorption cell of the analyzer. Quantitative results
are obtained by comparison of the values obtained with the results of stan-
dard Hg solutions of 1, 10, and 50 ppb Hg treated in a like manner to the
samples.
Equipment: Blue M Model 1130A Water Bath, Fisher Mercury Analyzer, Varian
Model 485 Digital Integrator
Detection Limits: 4 ppb (yg Hg/g wet tissue)
Reference: W.L. Anderson, K. E. Smith, Environmental Science and Technolo-
gy, 11 (1), 75-80 (1977).
Method of Analysis of Pesticides and PCB's in Water, Sediments, and Fish
Water: The volume was measured and placed in a separatory funnel. For
PCB's, 100 ml of 10 percent ether in hexane was added with shaking,
separated, dried over anhydrous sodium sulfate and reduced with a 3 ball
Snyder column and the volume adjusted for injection into GLC. For atrazine
and 245-T, the extraction was effected with three portions of 100 ml of
methylene chloride which was combined, dried over sodium sulfate, reduced
with a 3 ball Snyder column and exchanged with ethyl acetate for injection
by flame thermionic ionization.
Sediment: Fifty grams of mixed sample were extracted by the addition of
100 ml of ethyl acetate, stirring for an hour with magnetic stirs. The
ethyl acetate was decanted with rinsing, reduced in volume and dried with
the addition of hexane, over anhydrous sodium sulfate. Final results were
reported on a dry weight basis.
Fish: The whole fish was extracted with 100 ml acetonitrile in a blender.
The acetonitrile was swirled with 10 ml hexane to remove fish oil, blown
dry with nitrogen and ethyl acetate added for GLC analysis. After initial
injection, they were saponified and re-extracted with hexane for confirma-
tion of PCB's.
Samples suspected of containing 245-T were acidified by addition of a
drop of hydrochloric acid before extraction and ester ified before GLC in-
TECHNICAL APPROACH
-------
71
jection.
Detection Systems
Lasso, Furadan, simazine, atrazine, and malathion were analyzed by
flame thermionic ionization on a Varian 2100 gas chromatograph using rubi-
dium sulfate for detection of nitrogen and phosphorus containing compounds.
Column temperature 210 degrees, Port temperature 235 degrees, Nitrogen flow
15 ml per minute through a 6 foot glass column containing 3 percent OV-17
on 100-120 mesh Gas Chrom Q.
PCB's were measured under the same conditions using 3 percent OV-210
and 1.5 percent OV-17 on 100-120 mesh Gas Chrom Q in a glass column and a
Ni 63 detector.
TABLE 9
Recovery of Compounds
Water and sediment: Nitrogen containing compounds 95%+
Phosphorus containing compounds 97.5%
PCB's 98%
Fish: Nitrogen containing compounds 90%
Phosphorus containing compounds 92%
PCB's 98%
Lower Limits of Detection
Flame thermionic ionization: Nitrogen containing compounds 10 nanograms
Phosphorus containing compounds 50 picograms
Electron Capture: PCB's 100 picograms
TRACE METAL ANALYSIS
Scope: Total Cations - Cd, Fe, Zn.
Sample: Soft tissues
Procedure; A 0.5 - 1 - g sample of freeze dried tissue is weighed into a
50-ml Vycor crucible, covered, and ashed at 450 degrees C for 16-hours.
After cooling, the ash is wetted with a few drops of water, and 5-ml cone.
nitric acid is added. The crucible is slowly heated to dryness on a hot-
plate and then returned to the muffle furnace for 0.5 hours. When cool,
5-ml cone, hydrochloric acid (nitric acid if Ag is to be determined) is ad-
ded and the sample heated on a hot plate to reflux under the inverted cru-
cible cover until dissolved. The sample, crucible and cover are rinsed
into a 50-ml volumetric flask and diluted to volume with deionized water.
The solution is analyzed by emission spectrometry using a radio-frequency
inductively-coupled argon plasma as the source of radiation. Standard
solutions of the elements desired are used in calibrating the instrument
which is controlled by a mini-computer.
Equipment: Muffle furnace, Hotplate, Jarrell-Ash Model 975 Plasma AtomComp
TECHNICAL APPROACH
-------
72
Detection Limits: ppm (ug metal/9 wet tissue), Cd 0.09 ppm, Fe 1.0 ppm,
Zn 0.5 ppm
Reference: K. E. Smith, Illinois Natural History Survey, in-house.
3.3.2.3 OTHER PARAMETERS
BOD analysis performed by Pollution Control Systems, Inc. (Laotto,
Ind.) following procedures of Standard Methods (APHA 1971).
COLIFORMS Samples were collected in sterilized 100 ml glass bottles
by immersing them at the water's surface. Testing procedures were initiat-
ed five hours or less after sample collection. The Fort Wayne Allen County
Board of Health Laboratory conducted the testing. The membrane filter
technique was used following Standard Methods (American Public Health Asso-
ciation, 1971) and the recommendations of the Millipore Company. After
serial dilution of the samples, plates were inoculated. Endo Millipore,
fecal MF-C Millipore and bacto KF dehydrated streptococcus broth (Difco La-
boratories) were used for total coliform, fecal coliform and fecal strepto-
coccus tests, respectively. The incubation and counting procedures out-
lined in Standard Methods were followed. Throughout this report the bac-
terial concentrations are expressed on the basis of counts per 100 ml of
water.
3.3.2.4 COMPARISON OF ANALYSES
Comparisons of water quality analysis results as determined by the
Purdue and Illinois Natural History Survey laboratories are shown in Fig-
ures 10 through 12 for suspended solids, ammonium N and nitrate N. Water
samples were collected at 30-minute intervals during a rainstorm at three
locations in the watershed. After collection, the water samples were
frozen and then alternate samples were transported to the laboratories,
where they were analyzed according to standard procedures used by the
respective laboratory.
The concentration values for suspended solids, ammonium N, and nitrate
N suggest that methods used by the two laboratories give comparable
results. There was very close agreement between laboratories for nitrate N
concentrations. With ammonium N, close agreement was obtained for one
site, but for the other.two sites, the concentrations obtained by the Pur-
due laboratory were higher than the concentrations obtained by the Illinois
Natural History Survey Laboratory at one of these locations and lower at
the other. Good agreement was obtained for suspended solid concentrations
when the level was below 1400 mg/1. However, the values obtained by the
Purdue laboratory were as much as 25 percent lower than those obtained by
the Illinois Natural History Survey Laboratory at high suspended solid con-
centrations.
3.3.3 Data and Handling
The data collection procedures and equipment utilized in the Black
Creek Project can best be considered in three separate categories: 1) the
collection of socio-economic data, 2) data collected primarily for fishery
and biological studies and 3) monitoring the effects of agricultural opera-
tions and loadings of various non-point source pollutants in the Black
Creek and its tributaries. The first two categories are discussed in Sec-
TECHNICAL APPROACH
-------
200
100
(-3
M
O
ffi
n
Station 2
Sample Number
2000
1800
1000
800
400
200
21 22 23 24 25 26 27 28 29 30 31 32 33 34
Station 6
Sample Number
1700
1500
1300
900
700
14 16 18 20 22 24
Station 12
Sample Number
9 10 12 14 16 18
Purdue
0
Illinois > 0 < 0
O
PS
Figure 10. Comparison of Suspended Solids
U)
-------
74
ja»n jad swejBi||!l/\|
CM
o
'
CO
E
3
o
o.
in
M
in
a
J81H Jad siuejBi||||/\|
c
o
CO
£
3
Z
_w
a.
I
-------
W
O
ffi
2!
H
o
15
14
13
12
11
Station 2
Sample Number
22 24
26
15
14
13
12
11
10
28 29 30 31
Station 6
Sample Number
20
18
16
14
12
10
10
14 16 18 20 22 23
Station 12
Sample Number
10 12 14 16
18
Purdue
Illinois
o
Figure 12. Comparison of Nitrate N
Ol
-------
76
The concentration values for suspended solids, ammonium N, and nitrate
N suggest that methods used by the two laboratories give comparable
results. There was very close agreement between laboratories for nitrate N
concentrations. With ammonium N, close agreement was obtained for one
site, but for the other two sites, the concentrations obtained by the Pur-
due laboratory were higher than the concentrations obtained by the Illinois
Natural History Survey Laboratory at one of these locations and lower at
the other. Good agreement was obtained for suspended solid concentrations
when the level was below 1400 mg/1. However, the values obtained by the
Purdue laboratory were as much as 25 percent lower than those obtained by
the Illinois Natural History Survey Laboratory at high suspended solid con-
centrations.
3.3.3 Data and Handling
The data collection procedures and equipment utilized in the Black
Creek Project can best be considered in three separate categories: 1) the
collection of socio-economic data, 2) data collected primarily for fishery
and biological studies and 3) monitoring the effects of agricultural opera-
tions and loadings of various non-point source pollutants in the Black
Creek and its tributaries. The first two categories are discussed in Sec-
tion 0.0 and Section 0.0. The third category is further broken into two
major subcategories. The network of instrumentation installed at the be-
ginning of the project was based upon a combination of grab sampling pro-
cedures and independent strip-chart recorders for measuring rainfall and
stage. Three stage-initiated, clock driven pumping samplers were installed
to provide a more comprehensive picture of water quality conditions in the
streams during runoff events. The specific equipment, its location and
operation are discussed elsewhere in this report. At about the mid-point
of the project an effort was initiated to develop and install an automated
data acquisition system to supplement this other data collection network.
A comprehensive discussion of the philosophy and nature of the resulting
system is the subject of the following material.
3.3.3.1 THE ALERT SYSTEM
One useful method of classifying data collection systems is according
to the controller to which the basic sensors are attached. This approach
allows automated systems to be divided into two broad classes: data loggers
and real-time computer controlled. The primary emphasis of the following
discussion is on the latter.
Traditional methods for studying natural phenomenon involve data-
logging. A data-logging system is one in which the primary function is to
record the data supplied from multiple sensors to a recording media. The
storage is a printed copy of values, a strip chart or, in some newer units,
computer readable punch tape or magnetic tape. These methods imply (1) a
substantial delay between data recording and its availability to be
analyzed, or (2) that media other than strip charts and printed copy be
utilized to (a) reduce labor and errors associated with transcribing to
computer readable media, (b) eliminate time registration errors, and (c)
allow more channels of data to be recorded.
When newer technology is utilized many benefits occur. A real-time
computer is one which not only collects data from an array of transducers,
TECHNICAL APPROACH
-------
77
but has the ability to analyze that information in a sufficiently short
time to effect a controlled response, dependent on the results of analyzing
current conditions. Thus, the real-time computer must be able to communi-
cate with all transducers at all times (this is called "on-line") through a
communications system.
Real-time data acquisition inherently provides all the capabilities of
data logging systems plus those resulting from the expanded analytical
capabilities of an associated on-line computer. While real-time systems
are generally more expensive, the additional costs are normally a small
percentage of the total outlay for an environmental data acquisition net-
work. In view of the substantially increased benefits, they can often be
easily justified.
The most important factor which determines the success of a data col-
lection system is the fidelity of the data collected. The fidelity of a
data base refers to its ability to accurately portray the complete behavior
of the system it purports to characterize. Data base fidelity is dependent
upon: 1) proper positioning of adequate numbers of sensors to permit con-
tinuous inference of the complete state of the processes under study and 2)
the operational reliability of all components of the data acquisition sys-
tem. Selection of sensor location is highly process dependent and outside
the scope of this discussion. However, the fact that operational reliabil-
ity is strongly influenced by system organization, as well as transducer
hardware selection, is inadequately appreciated.
Assembling a system to collect data from a dispersed network of unat-
tended instruments which must operate over wide environmental extremes is
not easy. It requires both careful selection of reliable instruments as
well as proper system configuration. This involves a configuration which
provides redundancy and cross-checking measurements. An on-line computer
contributes to both of these areas.
First, an on-line computer affords a direct means of providing a
redundant data recording system. No real-time data acquisition which is
intended to maintain a continuous historical data file should solely rely
on the on-line computer to record incoming data. A backup data logging ca-
pability, preferably battery powered, which requires neither the computer
nor the communication link between the computer and the field instruments
should be a part of the backup system. Secondly, the analytical capacity
of an on-line computer makes it feasible to institute sophisticated trans-
ducer error detection schemes. In addition to the simple alarm limit ap-
proach, one can incorporate tests for rates of change on single and corre-
lated variables. Cross checking can also be program controlled. For exam-
ple, air temperature readings can be combined with net solar radiation data
to yield independent estimates of soil temperatures adequate to detect a
questionable operational status for a soil temperature transducer.
Since the communication link to an on-line computer is a two-way path,
it is a comparatively minor task to implement operational control of field
equipment which is dependent upon environmental conditions. Consider the
data acquisition needs associated with monitoring nonpoint source pollution
in a stream. Many of these pollutant problems are storm related and in-
volve rapidly changing concentrations. A real-time computer, monitoring
hydrometeorological conditions in a watershed, can control the activation
of remote pumping water samplers to collect frequent samples during rapidly
TECHNICAL APPROACH
-------
78
changing conditions, but infrequent samples during slowly changing condi-
tions. Such an approach simultaneously improves the fidelity of the data
base and reduces the total number of water samples which must be collected
and analyzed.
An integral part of any data collection program is the "permanent"
storage of data in a format suitable for subsequent intended uses. This
function is virtually unchanged between real-time and off-line systems.
However, the virtually instantaneous availability of current as well as
historical data files with on-line systems has many ramifications for util-
izing this information.
The real payoffs for real-time systems almost all are direct conse-
quences of immediately applying the phenomenal analytical capability of a
general-purpose computer to data arriving from a remote sensor network. It
is the elimination of the time lag between the acquisition of data and its
analysis/interpretation which makes real-time systems attractive. Of
course, any benefits to be realized are totally dependent upon the ingenui-
ty of the persons responsible for developing the computer programs which
must analyze all incoming data.
3.3.3.1.1 Configuration
In July of 1976, a major amendment to the original Black Creek Project
was approved. The additional activity funded under this amendment was the
development of an automated data acquisition system and an associated dis-
tributed parameter hydrologic model. The subsequent development of a sys-
tem designed to accomplish the Acquisition of Local Environmentally Related
Trends, ALERT, was intended to accomplish two primary objectives: (1) to
automate the process of collecting hydrometeorological data from the Black
Creek catchment in order to reduce data transcribing delay and labor while
expanding the scope of data collected and (2) to demonstrate a real-time
acquisition system capable of providing a data base to permit hydrologic
simulation of watershed responses concurrently with naturally occurring
storm events. It should be noted that the ALERT system is a combination of
both hardware (transducers, etc) and computer software.
An integral part of the second objective for ALERT involved using the
computer to generate operational commands to control pumping samplers which
collect periodic water quality samples during a runoff event. The availa-
bility of simultaneous data from a network of sensors dispersed over the
watershed together with the predictive capabilities of the on-line computer
were intended to improve the fidelity of these water samples. This im-
proved fidelity is a result of using short sampling intervals when pollu-
tion concentrations are likely to be changing rapidly and much slower sam-
pling rates when conditions are stable. The rapid sampling rates make pos-
sible an accurate evaluation of total quantity of pollutant in the runoff
while slow rates during stable conditions reduce the number of samples col-
lected and the cost of subsequent laboratory analyses. While the ALERT
system was designed to satisfy the objectives of a specific project, the
requirements were of such a nature that the resulting system is directly
applicable to a large percentage of environmental data acquisition applica-
tions.
The ALERT hardware was constrained by three major requirements: bat-
TECHNICAL APPROACH
-------
79
tery powered operation, low cost, and field maintainability.
The system hardware must function in areas where 110 VAC line power is
not available. This requirement resulted in the use of solar chargers on
NICAD* batteries. This type of energy source resulted in limited power
availability which restricted the amount of electronics possible at each
site. Second, wherever possible CMOS# logic elements were used. CMOS log-
ic has the attribute of using very low power when it is not active. Due to
the relatively low rates of change involved in environmental monitoring a
great deal of power is conserved. * NICAD is an acronym for Nlckel-
CADmum. NICAD batteries exhibit long shelf-life, high current capability,
small size, and are rechargeable. # CMOS is an acronym for Complementary
Symmetric Metal Oxide Silicon.
One objective of the project was to develop a data acquisition system
which would have general applicability to other environmental monitoring
applications. Therefore, it was especially desirable to develop a system
of low unit cost. A remote site with a standard set of transducers (for
our applications) cost on the order of $1500.00. This requirement did
somewhat restrict the choice of power supplies described above and preclud-
ed reliance on commercially available complete acquisition systems. In ad-
dition, certain desirable features had to be held to a minimum. A rather
large reduction in data transfer redundancy was made to cut power consump-
tion and unit cost. Naturally, this had some adverse influence on data in-
tegrity. Since design of the remote stations, a CMOS microprocessor has
become available. The designers believe that another project in this area
should carefully check this possibility of having at low power consumption
a very high level of intelligence (as needed to implement redundance in
data transfer).
The distance between the designers and the watershed dictated that the
system be constructed of modular pieces. All circuit modules plug into a
mating connector. This allows field personnel with limited electronics
background to change systems modules and maintain an operational status.
The configuration of the ALERT hardware as implemented is shown in
Figure 13.
As in all data acquisitions systems the information begins at the (1)
transducers. The ALERT hardware is based on an Incremental Increment
Recording system. This system was developed by Goodspeed and Savage
(1966), evaluated in detail by Langham (1971) and compared to the more
widely known Time Increment Recording system by Wong, et al (1976). IIR
transducers indicate whenever a fixed change occurs in current conditions,
e_.g_., a drop in temperature of .5 degrees. This technique allows the
conversion from analog to digital data at the earliest possible moment
which increases data integrity, ease of transmission and results in an ex-
tremely efficient data storage format.
Each transducer reports any incremental change of conditions to its
(2) remote station. A transducer may be up to 200 meters away from its re-
mote station. A remote station then waits for permission from the central
site to transmit its data. Permission to transmit occurs about once every
2 seconds. After a remote station sends the transducer data to the central
site, the remote site signals the transducer that data has been sent and it
TECHNICAL APPROACH
-------
Figure 13. Configuration of ALERT Hardware
fd
O
i- 1
n
o
K
FT. WAYNE
TERMINAL IN
FIELD OFFICE
PURDUE
DEDICATED
LONG-LINE
~250 km
REAL-TIME
VIINI-COMPUTER
.m
DIAL-UP
MODEM
TERMINAL FOR
PROGRAMMING
AND DISPLAY
TERMINAL
oo
o
-------
8JL
is ready for further information at any time.
A remote site receives its permission to transmit data (polling) and
sends data to the central site on the (3) party line. The line consists of
2 separate circuits. (3a) One circuit takes the polling information and
distributes it to the remotes. (3b) The other circuit accepts data from
the remotes and routes it to the central site. All remotes are connected
in series with one another. This implies if one remote fails the communi-
cation link fails, but it saves an enormous amount of power the remotes
could not supply due to power restrictions. The party line was a special
telephone circuit that by-passed the dialing equipment in the watershed
area.
Both lines terminate at the (4) central site, which is physically lo-
cated at watershed Site 6. The central site performs several functions in-
cluding: (4a) sending continuous polling information to all remotes which
prevents two remotes from simultaneously transmitting data; (4b) monitoring
the long lines (see below) and interspersing control commands with polling
commands; (4c) receiving data from the remote sites and (4c,a) sending the
data to a modem* which impresses the data as serial bits on the long lines
and (4c,b) to a paper tape punch as a backup in case of long line failure.
*MODEM is an acronym for modulator-demodulator. Operating in pairs, these
devices provide a standard means of converting the voltage or currents gen-
erated and required by computers into tones suitable for transmission over
telephone lines.
The (5) long lines are dedicated telephone circuits electrically
separate from the party line. The long lines stretch 250 km from a modem
on Black Creek Watershed to a modem in the Agricultural Engineering Depart-
ment at Purdue University. The telephone circuits link the modems 24 hours
a day and by-pass normal dialing equipment.
The modem at the Purdue campus connects to a (6) mini-computer with
real-time software. The data from the watershed appears to come from one of
up to 20 users on the computer. As data are received various software
packages (inter-related programs) are executed to perform the desired func-
tions of the system as described above. The computer also has the ability
to establish communications with other systems on campus (see below).
Also, the machine can answer its own telephone and connect that telephone
line to a modem (see below). The real-time mini-computer has a disk
storage system for storing data and programs in addition to various termi-
nals for data display and programming.
The Black Creek installation is seen by the computer as simply one of
several simultaneous users active on the system. The operating program
which controls communication with the Black Creek station has four primary
responsibilities: (1) assembling the incoming data into suitable files and
permanent storage of these files on magnetic disk and/or tape, (2) mainte-
nance of a dynamic file of the instantaneous level of all variables being
monitored in the watershed and the operational status of all transducers,
(3) providing a preliminary analysis of water stage data in order to issue
feedback control commands to operate the water sampling equipment, and (4)
detection of storm conditions in the watershed that indicate the need to
activate a complete real-time simulation of the hydrologic behavior of the
catchment. During a runoff producing storm, a simulation model can be ac-
tivated which combines historical data files describing physical charac-
teristics of the catchment with real-time, dynamically changing data con-
cerning rainfall intensity distribution and stream stage to estimate height
TECHNICAL APPROACH
-------
82
and times of peak flows at all points in the drainage network. If the geo-
graphical location warranted such action, the computer could be programmed,
based on predicted conditions, to automatically ALERT responsible authori-
ties in the event of impending dangerous flood levels.
The ability of the computer to answer a telephone allows a "reverse
information circuit." The field office in Ft. Wayne is approximately 25 km
from the watershed. Using a computer terminal and a modem, the staff can
phone the real-time mini-computer at W. Lafayette. At that point the Ft.
Wayne office has the ability to perform any function available on the sys-
tem. This includes the display of current conditions or of archive files.
This capability greatly enhances the efficiency of field personnel while it
increases the overall integrity of the data base by very quickly identify-
ing malfunctions while simultaneously eliminating unnecessary field inspec-
tions of correctly working sensors. The current status report allows quick
response by field personnel to thundershower events which may occur in the
catchment, but not in the vicinity of their office.
Another feature available through the software allows the real-time
mini-computer to link to other computers. A linkage is available to the
system which logs the most recent messages from the NOAA weather service.
This service provides the most recent Ft. Wayne area forecast and Maumee
River predictions, as supplied via NOAA, as a supplement to the tools
available to the Ft. Wayne staff.
Operational experience with the system has been only partially suc-
cessful. The primary meteorological station at Site 6, the battery operat-
ed paper tape punch, the long-lines to W. Lafayette and the on-line mini-
computer have operated together very satisfactorily for over two years.
The area that has involved the greatest difficulties is a low power method
to couple the remote stations to the "party line". Devices that exhibit
high reliability in transfer of data have also shown a low resistance to
outside destruction. The destruction usually comes in the form of a light-
ing strike in the watershed area, although it is possible for routine tele-
phone repair procedures in the vicinity to damage the interface boards.
Interface boards, which had protection circuits for the above problems, ex-
hibited a markedly lower reliability in transmitting the data. Techniques
which utilize magnetic coupling to provide isolation between the telephone
lines and the remote station hardware are currently being designed.
Further projects in this area should have the following additional
goals. The first goal should be to improve on the reliability of the party
line interfaces. A new product on the market has been earmarked for inter-
face testing by the designers. The new device will hopefully raise the
data transfer and destruction protection reliability far above those re-
quired. The use of HI-VHF transceivers should be incorporated to support
data transmission from sites where phone service is not available or where
the service is unreliable or too costly. The most extensive goal would be
to increase the real-time warning/simulation phase of ALERT and increase
the ability of the software to function with less operator intervention.
The hardware is of no value without a software package to support it.
The ALERT software presently consists of no fewer than 18 FORTRAN programs.
Commands to the mini-computer are fairly simple (although restart pro-
cedures are relatively lengthy). For example, to activate the software
display of current conditions the user (at Purdue or the field office)
TECHNICAL APPROACH
-------
83
types dec (Display Current Conditions). The terminal then responds with a
display for each transducer that informs the user of: (1) the device name;
(2) current value or rate; (3) trend indicator (increasing, steady, de-
creasing); (4) today's high and low value or rate; (5) the time of today's
high and low; and (6) a series of cross checking counters. There is an op-
tion on the dec program to also display element (4) and (5) for yesterday,
the day before, and monthly totals back 2 months on those transducers that
can utilize it. Other programs allow retrieval of data, testing of phone
circuits, scanning of data files for illegal data, running special statis-
tics, and editing of data files.
The data storage format is very space conserving considering the re-
quirements put on the data base. The data storage technique allows a ra-
port generated at any time interval (resolution of 1 second). Overall the
raw data is stored as a site number, a channel number, and the time since
the last data arrived. The values are stored "unformatted" (non-human
readable format). Programs read the raw data and reconstruct current con-
ditions by starting at a known point in time with known conditions (a
benchmark), and stepping forward through the data. A benchmark of the con-
ditions at the most recent midnight are kept in an easily accessible form
to reduce the overhead in computing current conditions which, in the case
of intelligent decision making (for control), occurs many times per day.
3.3.3.1.2 Conclusions
The collection of comprehensive environmental data is an essential re-
quirement for rational planning of nonpoint pollution control measures and
for subsequent enforcement and post-planning evaluation activities.
Several examples of such activities underway in the Black Creek Study area
have been described. Tne dramatic impact of utilizing real-time computers
to collect environmental data has been outlined.
Proper transducer selection and data network configuration allow ex-
isting time-sharing computer systems to serve as real-time systems for most
environmental data requirements with no additional hardware or system level
software changes. This approach provides the benefits of an on-line com-
puter with no capital outlay beyond those associated with a data logging
system of greatly reduced capability. While operating costs will be
slightly higher for the real-time system, these extra costs are primarily
proportional to the degree of utilization of the on-line features of the
system and are, therefore, subject to cost/benefit considerations and ad-
ministrative control. Furthermore, many of these associated benefits are
sufficient to significantly influence the economic justification of the
network of field transducers required for any degree of automation of data
collection procedures.
References
Goodspeed, M.J. and J.V. Savage. 1969. A multi-channel digital event-
recorder for field applications. J. Sci. Instr. v2 pp 178-182.
Langham, E.J. 1971. New approach to hydrologic data acquisition. Proc.
Am. Soc. Civil Eng. J. Hydro. Div. HY12:1965-78.
TECHNICAL APPROACH
-------
84
Wong, G.A., S.J. Mahler, J.R. Barrett, Jr. and L.F. Huggins. 1976. A sys-
tematic approach to data reduction using GASP IV. Proc. Winter Simulation
Conf. PP. 403-410.
3.3.3.2 PURDUE DATA HANDLING
In the Black Creek Watershed, rainfall data was collected from up to
seven recording rain gauges and water stage data was collected from up to
nine pressure-actuated stage records. Water quality samples were collected
both manually and mechanically.
Three pumping samplers, each capable of collecting 72 consecutive sam-
ples, are located at junctions of two primary drains into Black Creek and
on Black Creek approximately one and a half miles from its confluence with
the Maumee River. The pumping samplers were storm-actuated. Grab samples
were taken at all stage recorder sites, or strategic locations upstream
from the stage recorder sites, and at selected tile outfalls. Grab samples
were collected weekly and during storm events.
Rainfall or water stage data and water quality samples have been col-
lected since early 1973. An enormous amount of information was made avail-
able for various kind of analysis. In order to put the data into useful
form for future analysis, a procedure as illustrated by Figure 14 was ini-
tiated. Raw data as represented by rainfall charts, water stage charts,
grab samples, and automated pump samples were processed largely by comput-
ers and then stored to be used by researchers connected with the project
and researchers outside of the project who are interested in the regional
aspects of the data.
Figure 14 is a schematic diagram of data processing for the Black
Creek Watershed study. Steps in this process as indicated on the figure
are as follows:
Step 1: Water stage and raingage charts are read on a chart reader
and the data punched on paper tape.
Step 2: Data on paper tape are read into the digital computer file.
Step 3: Rainfall data, which are accumulated inches of rainfall, are
transferred into rates in on/hr.
Step 4: Areal rainfall is calculated by taking the weighted average
of the rainfall data on an area basis between adjacent sites.
Step 5: Rainfall data as well as water stage and water quality data
are stored by year and by the site number.
Step 6: Water stage data are edited for comma and characters and then
stored by year and by the site number as in Step 5.
Step 7: Grab sample data are verified for punching errors and correc-
tions made.
Step 8: Grab sample data are then sorted out by time, date, and site
TECHNICAL APPROACH
-------
sr
Figure 14. Data Processing Scheme
CHART
READER
LAB
ANAL-
YSIS
I--3
t?J
O
hi
H
O
GRAB
SAMPLES
CHART
READER
PUMP
SAMPLES
(BY
STORM)
ffi
LAB
ANAL-
YSIS
DIGITAL
COMPUTER
FILE
2
DIGITAL
COMPUTER
FILE
2
KEY-
PUNCH
2 DECKS
STAGE
DATA
ADDED
12
VERIFY
FOR
PUNCH
ERROR 7
CHANGE FROM
CUMULATIVE
DATA TO RATE
DATA 3
SORTED
BY
TIME &
SITE
KEY-
PUNCH
2 DECKS
VERIFY
FOR
PUNCH
ERROR7
AVERAGE TWO
RAINFALL
SITES ON
AREA BASIS
REFORMAT
FILE
ERROR
CHECK
9
SORTED
BY
TIME &
SITE 8
6
BEST
ESTI-
MATE
10
ERROR
CHECK
9
STOR
YEAR
E DATA
3Y
&SITE
5
STORE DATA
BY
YEAR & SITE
5
STAGE
CONVER
SIGN
11
STORE
DATA BY
YEAR &
SITE
BEST
ESTI-
MATE
10
STORE
DATA BY
YEAR&
SITE
COMBINE
PREPROCESSED
DATA FILES
INTO SINGLE
FILE BY YEAR
AND SITE
13
COMPRE-
HENSIVE
DATA
FILES
oo
Ui
-------
86
Step 9: Grab sample data and also pump sample data are checked for
errors and omissions such as poor response from a site, unrealistic dates
and times, unreadable characters, abnormally high values, and bad values of
N or P constituents.
Step 10: Best estimates are made for missing data or for water quality
parameters, which are flagged for possible error in analysis or for wrong
entries in the data log. If errors are due to faulty analysis, rules for
obtaining the best estimate are:
Let soluble N = Nitrate + Ammonium, if Nitrate + Ammonium > soluble N
Let total N=soluble N, if soluble N > total N
Let soluble P=inorganic P, if inorganic P > soluble P
Let total P=soluble P, if soluble P > total P
Step 11: The distance from a benchmark to the water level is converted
to depth of water for the stage record water the grab samples. The grab
sample data are now stored as in Step 5.
Step 12: As in Step 11 for the grab samples, stage data are added to
the pump sample file. Stage data are necessary to calculate for loadings.
The pumping sample data then go through the same steps as for grab sample
data and are also stored as in Step 5.
Step 13: The data files are now combined and sorted according to time
and location and then placed on disk into a comprehensive data base.
STUDY RESULTS
-------
87
4.1 SIMULATED RAINFALL
4.1.1 Equipment
A portable rainfall simulator, developed by the Agricultural Research
Service of USDA at Purdue University, produces artificial storms of approx-
imately the kinetic energy of high intensity natural rainfall. The simula-
tor was used to compare the soil loss, water loss, and infiltration rates
of treatments on standard-size rectangular runoff plots.
Simulated storms may be applied to treatments being studied under any
condition at any time and as often as desired throughout the year, except
on tall crops and during freezing weather.
Factors well suited to tests with the simulator include: soil erodi-
bility, length of slope, percent of slope, past erosion, crop cover at
various stages of growth, rotations, fertility level, tillage practices,
and residue management.
Soil and water loss evaluations by natural rainfall usually take 10 to
25 years yet it is often vital to have such information much sooner.
The rainfall simulator frequently makes more rapid evaluations possible.
4.1.2 Objectives
Environmental Imgact of Land Use on Water Quality -- A Work Plan, a
plan to utilize the rainfall simulator in Black Creek watershed, wal~pub-
lished in April of 1973. Objectives were as follows:
(a) To determine the base values for the sediment contributions of
the major soil capability units in the study area.
(b) To determine runoff and sediment composition (physical and chemi-
cal) from the major soil capability units.
(c) To determine the relative importance of raindrop impact and sur-
face runoff in detaching soil material from nearly level lake plain
soil.
(d) To compare the runoff and soil erosion effects of presently used
cultural practices to those conservation cultural practices recommend-
ed by the Soil Conservation Service. (Several forms of Conservation
tillage compared to fall plowing, effects of crop rotations, effects
of various methods of residue management, effects of winter cover, ef-
fects of over-grazing, effects of fertilizer and manure applications
on cropland and pastures).
Work was carried out in all of these areas during the 1973-1976 pro-
ject period. In each case, the following test storms were used:
The simulated rainfall program was started in the summer of 1973, and
approximately six weeks of field testing was committed to this study each
year for four years. The individual studies are outlined below.
STUDY RESULTS
-------
88
4.1.2.1 BASE EROSION LOSSES
Values for the major soils in the watershed were obtained during the
summer of 1973. Thirteen cm (5 in) of simulated rainfall were applied to
fall plots under uniform test conditions on four different soils. Runoff,
infiltration, sediment concentration, and total soil loss were obtained in
each study.
4.1.2.2 PARTICLE SIZE IN SEDIMENT
Sediments in runoff from all four soils in the 1973 tests were
analyzed for particle size distribution (five sand fractions, silt, total
clay, colloidal clay, and organic matter content). These values have been
compared to the values that occur in the soil in place.
4.1.2.3 SOIL LOSS AS AGGREGATES
Sediment occurring in runoff from four soils (each soil with fall
plow, fall chisel, fall disk, and no tillage treatments) were analyzed for
soil loss in aggregated form as constructed to that occurring as primary
particles. Field and laboratory work was completed during the 1975-1976
project year.
4.1.2.4 FERTILIZER LOSS IN RUNOFF
The effects of surface applied nitrogen and phosphorus fertilizer on
nutrient content of runoff were obtained under fallow plot conditions in
1973 and under four tillage systems (fall plow, fall chisel, fall disk, no
tillage) in 1974 and 1975. In some instances, the tests were conducted on
soybean land, and in other instances, the tests were conducted on corn
land. In all instances, runoff from fertilized plots was compared to
runoff from not-fertilized plots.
4.1.2.5 RAINDROP ENERGY VS. SURFACE RUNOFF
The relative importance of raindrop energy and runoff in the soil ero-
sion process on both nearly-level and sloping soils was measured in 1973.
The tests were conducted on fallow plots on four major soils in the
watershed. The results were reported in the first annual report.
4.1.2.6 TILLAGE AND CROP RESIDUE
Soil Erosion was determined from four basic fall land treatments (fall
plow, fall chisel, fall disk, no tillage) following both corn and soybeans
during 1974, 1975, and 1976. Runoff, infiltration, and sediment concentra-
tion of the runoff were also obtained. Percent surface covered by crop
residues were determined for all treatments. A portion of the data was
analyzed and reported in the 1975 and 1976 progress reports.
4.1.2.7 APPLICATION OF ANIMAL WASTE
The effects of animal waste application to land, both on runoff and
soil loss, as well as on water quality, were tested during the spring of
1976. Individual tests were:
(a) Spring application of liquid and solid swine waste (surface ap-
STUDY RESULTS
-------
89
plied and incorporated) on corn stalk land.
(b) Spring application of solid swine waste on corn stalk land that
had four different fall treatments (plow, chisel, disk, no tillage).
(c) Spring application of solid cattle waste to closely grazed pas-
tures .
4.1.2.8 SOD BUFFER STRIPS, WATER QUALITY
The effects of sod buffer strips in reducing the sediment load of
runoff water was a preliminary investigation and results obtained are at
best an indication of the efficiency of the system.
4.1.3 Summary of Results
4.1.3.1 BASE (BARE, FALLOW CONDITION) EROSION
Runoff and soil loss values for four of the major soils in the
watershed are presented in Table 10. Thirteen (13) cm of simulated rain-
fall were applied to fallow plots under uniform test conditions.
TABLE 10
Erosion Losses
Soil Type %
Haskins sandy loam
Nappanee clay loam
Hoytville silty clay
Morley clay loam
Slope
0.2
0.7
0.5
5.1
Runoff (cm)
8.6
6.6
7.7
9.2
Soil Loss (t/ha)
10.1
4.6
6.4
34.5
Soil erosion losses from nearly level lake plain soils under fallow
conditions are low when compared to the more sloping soils in the
watershed. For example, a comparison of soil losses from nearly level Nap-
panee and Hoytville soils versus sloping Morley soils shows the latter to
be 6 to 7 times as erosive. Therefore, it can be concluded that the major
soil erosion problem does not occur on the nearly level lake plain soils,
but rather on the sloping, glacial till soils.
However, even on nearly level soils, significant soil loss differences
can occur. Note that even though the Haskins soil had the least slope,
soil loss values were approximately twice those on the Hoytville and Nap-
panee soils. Laboratory tests show both Hoytville and Nappanee soils are
better structured than the Haskins, having major shrinkage cracks which in-
crease infiltration. The Haskins soil was prone to surface sealing, caus-
ing increased runoff, as evidenced by the runoff data in Table 10.
Therefore, it is concluded that soil structure has a pronounced influ-
ence on soil loss the better the structure, the lower the soil loss.
STUDY RESULTS
-------
90
4.1.3.2 PARTICLE SIZE DISTRIBUTION
Results show the erosion process to be highly selective with the sedi-
ment showing distinct clay enrichment and, in some instances silt fraction
enrichment as well. Total clay content was often 30 percent or higher in
sediment than in the original soil. Sand fraction content was appreciably
lower in the sediment. These relationships occurred on both nearly level,
lake plain soils and sloping glacial till soils. On nearly level soils,
the clay enrichment of sediment remained somewhat constant from initial
runoff until the end of the storm. On sloping soils, clay enrichment was
greatest in the initial runoff, but decreased as the test storm continued.
Selective erosion, particularly clay enrichment of sediment, is occur-
ring from cropland fields in the Black Creek Watershed. This has definite
implications regarding sediment and related nutrient transport.
4.1.3.3 SOIL LOSS AS AGGREGATES
Sediment in runoff was analyzed to determine the percentage of soil
loss occurring as aggregates greater than 210 microns in diameter. The
results are shown in Table 11. Values presented are averages obtained dur-
ing a 6.4 cm simulated rainstorm applied to fall-turn-plowed land.
TABLE 11
Soil Loss As Aggregates
Soil Type % Aggregates in Sediment>210 micron
Haskins sandy loam 17
Nappanee clay loam 22
Hoytville silty clay 17
Morley clay loam 15
These results show that less than 20 percent of the soil transported
in runoff occurs as aggregates greater than 210 microns in diameter on
plowed soils. Therefore, on these soils, effective measures for reducing
erosion, at least on nearly level soils, should be based on prevention of
detachment and dispersion of naturally occurring aggregates by raindrops
since low-velocity runoff is not capable of transporting much of the soil
as large aggregates.
4.1.3.4 RAINDROP VS. SURFACE RUNOFF ENERGY
The relative importance of raindrop energy and runoff energy in the
soil erosion process is shown in Table 12. These tests were conducted on
fallow soils, and sediment concentrations are reported for raindrop-induced
runoff and erosion vs. inflow-induced runoff and erosion.
On all four soils tested, rainfall-induced runoff contained approxi-
mately 7-12 times the sediment concentration as that found in the inflow
produced runoff. These results provide further evidence that protecting
the soil surface from raindrop impact is one of the most effective means of
minimizing sediment concentrations in runoff.
STUDY RESULTS
-------
91
TABLE 12
Raindrop Energy vs Surface Runoff
Soil Type
Haskins sandy loam
Nappanee clay loam
Hoytville silty clay
Morley clay loam
Slope
\
0
0
0
5
I
;>
2
.7
.5
.1
Sediment
with
1.
.
1.
3.
Concentration
of
rain with
18
98
37
81
0.
.
.
*
Runoff
inflow
18
08
14
42
4.1.3.5 TILLAGE AND CROP RESIDUE
Soil erosion losses determined from four basic fall land treatments
(fall plow, fall chisel, fall disk and no tillage) following both corn and
soybeans are reported in Tables 13 and 14 12.7 cm of simulated rainfall
were applied to the tillage treatments prior to seedbed preparation in the
spring. The study was performed on the four major soils listed earlier,
but results from only the Hoytville and Morley soils are reported below.
TABLE 13
Losses After Various Tillage Systems
Tillage System
No till (check)
Disk
Chisel
Plow
(Hoytville Silty Clay
Surface Cover (%) Soil
After After After
Corn Soybeans Corn
78 24 1.1
77 12 .9
57 9 1.7
4 1 4.3
0.8% Slope)
Loss ( t/ha)
After
Soybeans
7.8
6.9
9.3
5.3
From these results several significant conclusions can be drawn.
a) Soil losses are greatly reduced by those tillage systems that leave
appreciable crop residues on the surface. Generally, there is an inverse
relationship between surface residue cover and erosion. Particularly ef-
fective are no till (checks) and disk treatments following corn since large
amounts of surface residue remain through the winter and into spring.
b) Fall chiseling following corn, although not as effective as the no
till and disk, significantly reduces erosion compared to plowing. The de-
gree of erosion control is dependent upon the amount of surface cover
remaining as well as the roughness of the surface. Fall chiseling of corn
land would be expected to be much more effective in reducing erosion if
performed across slope rather than up and down slope (this study evaluated
STUDY RESULTS
-------
92
TABLE 14
Losses After Various Tillage Systems
(Morley Clay Loam 4.0% Slope)
Tillage System
No till (check)
Disk
Chisel
Plow
Surface
After
Corn
69
70
25
7
Cover (%)
After
Soybeans
26
17
12
1
Soil
After
Corn
2.4
2.5
15.0
21.8
Loss (t/ha)
After
Soybeans
13.4
12.4
30.1
40.9
up and down slope tillage).
c) None of the conservation tillage treatments are as effective fol-
lowing soybeans as following corn primarily because of reduced surface cov-
er after soybeans. Chiseled soybean land can be particularly erosive when
chisel marks run up and down slope.
d) The major soil losses occurred on the sloping land. Therefore, it
is much more important to apply conservation tillage on the more sloping
portions than the nearly level portions of the watershed to achieve signi-
ficant soil erosion reductions and resultant sediment concentration in
drainageways.
e) Although these tests were conducted over a relatively brief period
of the erosion year, they illustrate the major influence various crop
species can have on soil erosion.
f) In the only direct comparison of spring moldboard plowing and fall
moldboard plowing, soil losses from 12.7 cm of simulated rain were 6.3 t/ha
and 17.6 t/ha for spring and fall, respectively. It should be noted that
the tests were made on a Raskins loam soil shortly after completing spring
plowing.
4.1.3.6 SOD BUFFER STRIPS AND WATER QUALITY
In a preliminary investigation the influence of 50 feet of bluegrass
sod in reducing sediment load of runoff water was investigated. Sediment
concentration of runoff decreased from 1% to 0.46% (54% reduction) when
passed over the sod. Although this study demonstrated that sediment load
can be effectively reduced by sod strips, it is recognized that the effec-
tiveness of the practice is dependent upon many factors. These would in-
clude original composition of sediment, rate and depth of plow, vigor and
height of sod, and many others.
4.1.3.7 APPLICATION OF ANIMAL WASTE
Three experimental sites were used in the conduct of the study (Table
15). Experiment 1 was conducted to evaluate the effects of two forms of
swine waste (liquid and solid) and two rates of solid swine waste addition
STUDY RESULTS
-------
93
(90 and 238 t/ha) on losses of nutrients in surface runoff from corn resi-
due covered surface. Experiment 2 was conducted to evaluate the interac-
tion of solid swine waste application with tillage methods on losses of nu-
trients in surface runoff from a rolling Morley soil which was initially
nearly saturated. Experiment 3 was conducted to determine the effects of
solid cattle waste application rate (0, 90, and 180 t/ha) on losses of nu-
trients in surface runoff from an overgrazed pasture. The conditions under
which the experiments were conducted are summarized in Table 15.
TABLE 15
Description of Experimental Conditions
Experiment Soil Tillage
No. Type Used
1 Nappanee None (residues)
None
None
Disk (after
waste added)
2 Morley No-til
No-til
Chisel
Chisel
Disk
Disk
Fall Plow
Fall Plow
3 Morley None (pasture)
None
None
Animal
Waste Used
None
Swine
Swine
Swine
None
Swine
None
Swine
None
Swine
None
Swine
None
Cattle
Cattle
(liquid)
(solid)
(solid)
(solid)
(solid)
(solid)
(solid)
(solid)
(solid)
Rate of
Waste
t/ha
0
95.6
90
238
0
90
0
90
0
90
0
90
0
90
180
In most cases, tillage was performed on the plot area, animal wastes
were applied at rates indicated in Table 15, and three standard simulated
rainstorms were applied (see Section 4.1 for details of rainstorm se-
quence, intensity, and duration). In Experiment 1, plots treated with the
high rate of solid swine waste were disked prior to application of rain-
storms, and in Experiment 3, only two rainstorms were applied (total of 90
minutes of rain). The concentrations of solids and nutrients in animal
wastes applied in each experiment are given in Table 16. The total amounts
of nutrients added with each waste-treatment are given in Table 17.
No waste-treatment added significant amounts of nitrate N because of the
low nitrate content of the wastes.
Samples of surface runoff were collected, frozen, and stored at -10
degrees C until analyzed. All chemical analyses were performed by methods
detailed in Section 3.3.2 Flow data from runoff plots was collected and
calculated as described in Section 3,3.3 Computer techniques were
used to integrate flow and concentration data, to computeloadings. Aver-
age flow-weighted mean concentrations of nutrients were determined for each
STUDY RESULTS
-------
94
TABLE 16
Nutrient Concentrations in Animal Wastes
Type Waste
Liquid swine
Solid swine
Solid cattle
Solids
%
8.1
39.0
23.4
Amm.-N
3165
5750
1288
Nit.-N
T-vm /
Ppfu (we
27
82
38
Org. N
i i __ ' __ \
T- DaSlSj
2580
10190
3535
Total P
1370
7000
690
TABLE 17
Amounts of Nutrients Added in Waste Applications to Three Experimental Sites
Experiment Type Waste
No.
1
2
3
Applied
Swine (liquid)
Swine (solid)
Swine (solid)
Swine (solid)
Cattle (solid)
Cattle (solid)
Rate Waste
Appl ied
t/ha
95.6
90.0
238
90.0
90.0
180.0
Nutrients Applied
Amm.-N
303
524
1350
524
116
332
Nit.-N
\f-t
Kg/
3
8
18
8
3
6
Org. N Total P
At *i
246 131
970 730
2244 1396
970 730
318 62
636 124
rainstorm and for all three rainstorms applied to each plot. All values
are averages derived from data obtained from duplicate plots. Values for
amounts of waste-derived nutrients lost in runoff were calculated as the
difference in nutrient loss between waste-treated and control plots. Per-
centages of added nutrients lost in runoff were calculated by dividing
amounts of waste-derived nutrients lost by amounts of nutrients added. All
values for amounts of animal waste added are on a net basis. Values for
concentrations of nutrients (N & P) in eroded sediment are on a moisture-
free basis.
4.1.3.7.1 Experiment ^
Table 18 presents data on losses of sediment and nutrients in runoff
from untilled Nappanee soil as affected by type and rate of swine waste ap-
plication. On the average, waste application had little effect upon the
amounts of water running off the soil or on the loss of nitrate in runoff.
The loss of sediment was reduced, whereas the amounts of ammonium N soluble
organic N (SON), sediment N, soluble inorganic phosphorus (SIP), soluble
organic phosphorus (SOP), and sediment P were markedly increased by swine
waste addition. Significantly higher losses of ammonium-N, SON, SIP, SOP,
and sediment P were obtained with solid swine waste application (90 t/ha)
as compared to liquid swine waste application (95.6 t/ha). This was ex-
pected due to higher amounts of nutrients added in the solid swine waste
application (Table 17). Losses of soluble nutrients from plots receiving
high amounts of solid swine waste (238 t/ha) disked-in were lower than
plots receiving the low rate (90 t/ha) of waste. Incorporation of applied
STUDY RESULTS
-------
95
waste is apparently very significant in reducing losses of soluble nu-
trients. Losses of sediment-bound nutrients were similar at both applica-
tion rates of solid swine waste.
The concentrations in runoff of all soluble nutrients except nitrate
were greatly increased by swine waste application (Table 19). The concen-
trations of ammonium-N and SIP in runoff were high enough to constitute
significant environmental problems. Disking-in solid swine waste lowered
the concentrations of SON, SIP, and SOP in runoff as compared to runoff
from undisked solid swine waste-treated plots. The concentrations of total
N and P in sediment running off the plots were markedly increased by waste
addition (four to ten times for N and three to sixteen times for P). The
high total N and P contents of eroded sediment suggests mass movement of
some waste particles as well as enrichment of eroded soil particles with
waste-derived N and P.
Calculations indicate that substantial proportions of nutrients added
in animal waste are lost in runoff if wastes are not incorporated (Table
20). From 34 to 44 percent of added ammonium-N, 20 to 25 percent of added
organic N, and 7 to 16 percent of added total P was lost from plots receiv-
ing 95.6 and 90 t/ha of liquid and solid swine waste, respectively. Lower
proportions of added nutrients were lost from plots having a high rate of
solid swine waste (238 t/ha) incorporated by disking.
STUDY RESULTS
-------
en
i-3
a
O
Kj
en
G
cn
TABLE 18
Losses of Sediments and Nutrients in Runoff from Nappanee Soil
Till age
ilone
None
lione
Ui sk
Waste
App 1 i cat i on
.lone
Liquid was te
(95.6 t/ha)
Sol i d \/as te
(90 t/ha)
Sol i d was te
(237.6 t/ha)
R a i n s t o r n
1
1
3
Total
1
O
3
Total
1
O
3
Total
1
2
3
Total
!hO
Runoff
en
4.30
2.17
2.44
8.91
4.41
2.14
2.19
8. 73
3.83
2.31
2.88
a.03
2.98
2.12
2.39
7.49
t\.
0.28
n.in
0 . 08
0.4G
87.22
10.88
4.88
102.98
184.82
24.86
21.95
231.64
138.37
21.95
1C. 77
177.09
no^-r.
3.43
1.60
i.7n
G.73
3.19
1.41
1 . On
5. GO
2.12
0.43
0.68
3.23
1.59
0.42
0.71
2.72
GOf!
O.G7
0.25
0.07
n.99
22.65
2.09
1.62
31.96
125.28
17.51
14.05
156.84
37.33
11. 7C
9.09
58.18
Se ' 1 M
5.19
3.89
3.46
12.54
26.11
F.7?
9.37
42. 20
21.49
] 4 . 9 3
14.09
50.51
28. 2C
13.03
13.35
54.58
SI P
0.024
O.U06
0.005
0.035
11.111
1.246
0.971
13.328
20.306
4.291
4.837
29.434
10.547
2.430
2.847
15.824
SOP
0.024
O.OOG
0.009
0.039
2.711
9.486
0. 219
3.416
4.165
1.21E
1.492
6.873
1.943
1.130
0.557
3.630
Sed P
1.470
0.5G6
0.444
2.480
5.398
O.G85
0.855
6.938
7.646
3.199
3.121
13.9G6
5.984
2.898
2.433
11.315
Sedi ment
1859
563
589
3011
1811
457
307
2575
5GO
268
223
1051
849
412
250
1511
-------
TABLE 19
Concentrations of Solids and Nutrients from Nappanee Soil
Till age
iJonc
.June
None
Disk
Waste
Appl i cat i on
None
Liquid svvi no
Sol i d swi ne
( 1 ow )
Sol i d swi ne
(high)
Ra i nstorn
1
2
3
All
1
2
3
All
1
o
L.
3
All
1
0
3
All
Concentration in Runoff
Sol i ds ,'IH + it-N
1+323 0.7
2591* 0.5
2U11* 0.3
3379 0.5
1*116 193.2
2136 50. C
11*02 22.3
2950 118.0
1UG2 1*82.6
11GO 107.6
77k 7C.2
lift* 256.5
2 C k 9 i* C i* . 3
1943 103.5
10'iG 70.2
2017 230. '*
MO-3-.'l
IDS/ 1
8.0
7.1*
7.0
7.G
7.3
P. 6
U.G
G.l*
5.5
1.9
2.1*
3.G
5.3
2.0
3.0
3.G
so:1
l.G
1.2
0.3
1.1
51.5
9.8
7.1*
3C.G
327.1
75.8
1*0.8
173.7
125.3
55.5
38.0
77.7
SI P
0.06
".03
0.02
0.0'*
25.25
5.82
!;. 1*3
15.27
53.02
1C.2C
16.80
32.60
35.39
ll.i* 6
11.91
21.13
son
0.06
0.03
".01*
O.Oi*
6.16
2.27
1.00
3.91
10.87
5.2G
5.18
7.G1
6.52
5.33
2.33
l*.85
Concentration in Sediment
P
2790
6910
5870
U1G5
11*1*20
11*81*0
30520
1G1*10
38375
55709
G318U
1*8059
33216
31G26
531*00
36122
N
fl.fr
' '^ti
790
1005
75'i
821*
2980
1500
2785
2691*
13650
11937
13996
13288
701*8
7031*
9732
71*88
en
H
a
a
en
G
to
-------
98
TABLE 20
Proportions of Added Nutrients in Swine Waste Lost in Surface Runoff
Waste
Tillage
None
None
Disk
Application
Liquid
Solid
Solid
(low)
(high)
Amm.
33.
44.
13.
Nutrients
-N
Nit.
in
-N
% of added
8
1
1
0.
0.
0.
0
0
0
Swine Waste
Org
lost in
24
20
4
N
rune
.7
.0
.4
Lost
Total
^ffic
16
6
2
.1
.5
.0
P
* Calculated as:
(Nutrient from treated plot - nutrient from control plot) x 100
" "-» _ i ___________ . 1 .._ _ ._ _ | T _
Nutrient added in waste
4.1.3.7.2 Experiment 2
The conditions used in Experiment 2 (rolling Morley soil nearly sa-
turated prior to waste application) were designed to test the effects of
fall tillage practices and waste application on nutrient loss during winter
and early spring situations. The data (Table 21) indicates that swine
waste application reduced the amounts of runoff water and amounts of eroded
sediment for all tillage treatments. However, waste application greatly
increased the losses of ammonium-N, SON, sediment N, SIP, SOP, and sediment
P. The amounts of sediment lost from waste-treated plots decreased in the
following order of tillage method: no-till > fall plow > chisel > disk;
whereas from control plots the order was: fall plow > chisel > no-till >
disk. The amounts of soluble N compounds lost in runoff from waste treated
plots decreased in the following order of tillage method: no-till > chisel
> disk > fall plow; whereas for soluble P components the order was: chisel
> fall plow > disk not equal no-till. Sediment N and P losses from waste-
treated plots were highest for no-till and least for fall plow (N) and
chisel (P) tillage treatments.
The solids content of surface runoff from all tillage methods was re-
duced by waste addition. The greatest reduction was observed with fall
plow tillage (Table 22). Waste addition increased the ammonium-N, SON,
SIP, and SON concentrations in runoff water. On the average, similar con-
centrations of ammonium-N, SON, and SOP were observed in runoff from
waste-treated plots for all tillage treatments. Higher SIP concentrations
in runoff from waste-treated plots were found with chisel and fall plow
tillage as compared to disk or no tillage. The N and P concentrations in
eroded sediment were markedly increased by waste addition for each tillage
treatment. The highest enrichment of N and P in eroded sediment occurred
with the disked plots where the N and P contents increased 13 and 29 times,
respectively, as a result of waste addition. The high N and P concentra-
tions in eroded solids from waste-treated plots suggest that mass movement
of waste particles occurred during runoff. High BOD loadings on recurring
streams are also possible under conditions of mass transport of waste
solids.
The proportion of added ammonium-N lost in runoff from waste-treated
plots varied from 19 percent (fall plow tillage) to 32 percent (no til-
lage) , whereas none of the added nitrate-N was lost from any tillage treat-
ment (Table 23). From 4.5 percent to 15.6 percent of the waste-derived or-
STUDY RESULTS
-------
TABLE 21
Losses of Sediment and Nutrients From Morley Soil
Till age
Chi sel
Disk
Fal 1
Plow
lio-ti 1
Waste
Rate
4- / 1
t/lia
90
0
90
0
90
0
90
0
Ra i n storm
1
O
£_
3
Total
1
2
3
Total
1
o
3
Total
1
o
I.
3
Total
1
2
3
Total
1
2
3
Total
1
2
3
Total
1
2
3
Total
H20
Runoff
cm
2. 86
1.49
2. CO
6.95
3.81
2.30
2.58
8.69
2.25
1.25
2.34
5.815
3.31
2.36
2. GO
8.27
2.11
l.On
2.01
5.12
5.40
2.81
2.97
11. 1C
3.70
1.25
2.14
7.09
4.18
2.58
3.18
9.94
Sed i merit
726
224
330
1280
4980
2706
2873
10559
347
°,8
157
592
1092
1270
123C
3598
909
207
301
1417
11079
5358
5475
21912
1324
141
139
1604
780
6232
2092
4104
NHj-N
109.47
24.032
23.15
156. G5
0.27
0.09
O.OG
0.42
80.07
16.44
20.94
117.45
0.05
0.07
0.07
0.19
67.07
13.59
17.42
98.03
0.60
0.14
0.11
0.85
127.62
19.73
20.24
167.59
0.12
0.09
0.06
0.27
N07 -M
-* \,
1.43
0.09
0.17
1.69
1.07
0.53
0.44
2.04
0.50
0.08
0.13
0.71
0.34
0.46
0.34
1.14
0.78
0.09
0.18
1.05
1.17
0.64
0.34
2.15
0.80
0.11
0.10
1.01
0.53
0.51
0.48
1.52
SON
«. / !i -,
g/na
64.99
6.57
9.11
CO. 67
n.3l
0.13
0.22
0.66 .
48.12
11.41
11.27
70. SO
0.11
0.04
O.OG
0.21
48.57
6.31
8.73
63.61
0.42
0.37
0.16
0.95
82.85
13.40
11.42
107.67
0.22
0.18
0.04
0.44
Sed \\
23.97
12.37
7.19
43.53
12.22
7.75
8.67
28.64
20.68
6.59
8.31
35.58
3.34
4.16
li. 4 7
16.97
IS. 73
7.48
8.84
35.10
27.52
10.58
15.59
53.69
46.66
F.96
5.23
58.85
6.60
3.31
5.15
15.06
SI P
16.449
5.232
5.752
27.433
0.009
0.006
0.006
0.021
3.121
3.296
2.907
9.324
0.023
0.025
0.030
0.078
10.438
2.614
4.387
17.439
.009
.003
.001
0.013
1.888
2.203
4.354
8.445
0.007
0.013
0.014
0.034
SOP
7.587
0.300
0.281
8.168
O.U14
0.007
0.007
0.029
6.397
0.027
1.467
7.891
a. 014
0.011
0.010
0.035
3.453
U.41S
0.641
4.512
0.020
0.012
0.010
0.042
b.989
1.597
0.511
9.097
0.013
0.007
0.010
0.030
Sed P
4.520
1.930
2.555
9. 005
3.242
1.808
1.440
6.490
11.489
2.036
3.640
17.165
1.685
0. 702
1.1R5
3.572
8. 212
1.69k
1.824
11.730
5.789
2. 716
2.547
11.052
26.105
1.758
2.131
29.994
1.795
0. 609
0.490
2.894
01
K
C
i
-------
en
o
o
TABLE 22
Concentrations of Solids and Nutrients From Morley Soil
c
§
Till age
Chi sel
Disk
Fal 1
P 1 ow
iJo-t i 1
Waste
Rate
t/lia
90
0
90
0
90
0
90
Ra i n storm
1
2
3
All
1
2
3
Al 1
1
2
3
Al 1
1
2
3
All
1
2
3
Al 1
1
2
3
Al 1
1
o
3
Al 1
1
2
3
Al 1
Concentration in Runoff
Sol ids
2538
1503
1269
1841
13071
117C5
11136
12151
1542
704
C71
101C
3299
53C1
4754
1*351
4308
2070
1498
2768
20517
190CS
18434
19599
3578
1128
577
22C2
1866
';775
6579
4129
NH + -i)
3S3.76
161.29
89.08
225.40
0.71
0.39
0.23
0.48
355.87
131.52
8 9 . 4 9
201.98
0.15
0 . 30
0.27
0.23
317.87
135.90
86.67
191. 5C
1.11
0.50
0.37
0.76
344.92
157.84
83.98
236.33
0.29
0.35
0.19
0.27
ilO;-Fi | COM
. ___mrr / 1
1 1 1 (_ j /
5.00
0.60
0.65
2.43
2.81
2.30
1.71
2.35
2.22
0.64
0.56
1.22
1.03
1.95
1.31
1.38
3.70
0.90
0.90
2.05
2.17
2.23
1.15
1.92
2.16
0.88
0.42
1.43
1.27
1.98
1.51
1.53
i
227.24
44.09
35.04
IIP. 07
0.81
0.57
0.85
0.7G
213.87
n i . 2 "
4s!lC
121.75
0.33
0.17
o.23
0.25
230.19
63.10
43.43
124.24
0.78
1.32
0.54
0.85
223.92
197.20
47.39
151.86
0.53
n. 70
0.13
0.44
51 P
57.71
35.11
22.12
39.47
0.02
0.03
0.02
0.02
13.87
26.37
12.42
16.03
0.07
0.11
0.12
0.09
49.47
26.14
21.83
3 4 . 0 fi
".02
0.01
n.003
0.01
5.10
17.62
18.07
11.91
0.02
o.05
0.04
0.03
SOP
26.53
2.01
1.08
11.75
0.04
0.03
0.03
0.03
28.43
0.22
6.27
13.57
0.04
0.05
n.04
0.04
16.37
4.18
3.19
8.81
0.04
0.04
0.03
0.04
18.89
12.78
2.12
12.83
0.03
0.03
0.03
0.03
Concentration in sediment
N 1 P
_r-,^r/'-
rng/
33020
55220
21790
34010
2450
2860
3018
2712
59600
74890
52930
60100
7637
3276
3617
4717
20660
36135
29370
24770
2484
1975
2847
2450
35240
49360
37630
36690
8460
2690
2462
3670
>-&
6230
8620
7740
7035
651
668
501
615
33110
23140
23185
28995
1543
553
959
993
9034
8184
6060
8280
523
507
465
504
19720
12470
15330
18700
2300
494
234
705
-------
101
ganic N added was lost in soluble form or as components of eroded sediment.
The highest percentage loss of added organic N was obtained with plots hav-
ing no tillage performed. Limited percentages of waste-derived total P (3
to 6 percent) were lost in surface runoff from waste-treated plots. The
highest percentage of added P lost was obtained with no-till plots and the
lowest percentage was observed with fall plow tillage plots.
TABLE 23
Proportions of Nutrients Added Los in Surface Runoff From Morley Soil
Nutrients in Waste Lost
Tillage
No-til
Chisel
Disk
Fall Plow
Amm.-N
0
31.9
29.8
22.4
18.6
Nit.-N Org N
Total P
^ff* -
0.0 15.6 6.1
0.0 9.8
0.0 9.2
0.0 4.5
5.2
4.2
3.1
* Calculated as:
(Nutrient from treated peat - nutrient from control plot) x 100
_____ __,__ _ ____ _ _,_-.____.____ ______________
Nutrient added in water
4.1.3.7.3 Experiment 3_
Experiment 3 was conducted to evaluate the nutrient losses in surface
runoff from a heavily manured, overgrazed pasture. The results (Table 24)
from two rainstorms indicate that sediment losses were very low for all
treatments. However, waste-treated plots had higher soil losses than con-
trol plots. This finding is likely the result of mass movement of manure
particles during intense rainstorms. As compared to the previous results
from the other waste experiments, the losses of all nutrient forms were
low. However, amounts of ammonium-N, nitrate-N, SON, sediment N, SIP, SOP,
and sediment P increased with increasing rate of waste application. These
results likely are due to high infiltration of applied rainwater and the
resultant low transport capacity of water running from the plots. It is
particularly noteworthy that ammonium-N and SIP losses were low for all
waste-treated plots.
Table 25 provides data on the average concentrations of solids and nu-
trients in runoff and nutrients in eroded sediment from the overgrazed pas-
ture site. Solids and nutrient concentrations in runoff were low as com-
pared to previous studies with waste-treated tilled soils. The concentra-
tions of nitrogenous components were particularly low. The concentrations
of nutrients in eroded sediment from control plots were higher than those
from control plots of tilled soil because of extreme enrichment during
selective erosion of pasture land. The N and P concentrations in eroded
sediment were markedly increased by waste application (three-fold for N and
two- to three-fold for P). Selective erosion of fine clay and mass move-
ment of manure solids probably explain the high N and P content of eroded
sediment.
Low percentages of added ammonium-N (1 to 29 percent), nitrate-N (2 to
4 percent), organic N (3 percent), and total P (4 percent) were lost in
surface runoff from waste-treated plots (Table 26). The proportion of ad-
ded nutrients lost in runoff decreased with increasing waste application
STUDY RESULTS
-------
102
TABLE 24
Losses of Sediment and Nutrient in Runoff From Overgrazed Pasture
Cattle
Waste Rain-
Applied storm
t/ha
0
90
180
1
2
Total
1
2
Total
1
2
Total
H20
Runoff
cm
2.07
2.29
4.36
1.61
1.99
3.60
0.93
2.98
3.91
Sed-
iment
24
21
45
63
47
110
61
103
164
Amm.-N
0.
0.
0.
1.
1.
2.
1.
2.
3.
04
10
14
34
05
39
23
54
77
Nit.-N
0.03
0.04
0.07
0.07
0.11
0.18
0.07
0.11
0.18
SON
-kg/ha
0.26
0.39
0.65
2.17
1.48
3.65
2.31
4.32
6.63
Sed N
0.44
0.66
1.10
3.96
4.51
8.47
2.48
11.50
13.98
SIP
0.073
0.010
0.083
1.122
0.793
1.915
0.920
2.232
3.152
SOP
0.013
0.007
0.020
0.061
0.047
0.108
0.132
0.108
0.340
Sed P
0.073
0.080
0.153
0.232
0.488
0.720
0.176
1.327
1.503
TABLE 25
Concentrations of Solids and Nutrients From an Overgrazed Pasture
Cattle
Waste Rain-
Applied storm
0
90
180
1
2
Ave.
1
2
Ave.
1
2
Ave.
Concentrat
Solids Amm.-N Nit
116
92
103
391
236
306
656
346
419
0.
0.
0.
8.
5.
6.
13.
8.
9.
19
44
32
32
28
64
22
52
64
0.
0.
0.
0.
0.
0.
0.
0.
0.
ion
.-N
mg
15
18
16
44
55
50
75
37
46
in Runoff
SON SIP
1.26
1.70
1.49
13.48
7.44
10.14
24.84
14.50
16.96
0.353
0.044
0.190
6.969
3.985
5.319
9.892
7.490
8.061
SOP
0.063
0.031
0.046
0.379
0.236
0.300
1.419
0.362
0.870
Concent
in Sec
N
mg/
18330
31430
24440
62860
95960
77000
40660
111650
85240
:ration
J intent
P
'kg
3040
3810
3400
3680
10380
6545
2885
12880
9165
rate. These findings suggest that runoff from heavily grazed or heavily
manured pasture land does not represent a large threat to water quality if
soil infiltration rates are reasonably high. The limited runoff coupled
with trapping of waste particles by grass plants apparently resulted in low
nutrient and solids loss in surface runoff.
STUDY RESULTS
-------
103
TABLE 26
Proportions of Added Nutrients Lost in Surface Runoff From an Overgrazed Pasture
Cattle Waste
Added
t/ha
90
180
Nutrients in Cattle Waste Lost
Amm.-N
c
1.9
1.1
Nit.-N
1 of added lost
3.7
1.8
Org N Total
CC +
in runorr
3.3 4.0
3.0 3.8
P
* Calculated as:
(Nutrient from treated peat : nutrient from control plot) x 100
Nutrient added in waste
4.1.3.7.4 Conclusions
1. Surface runoff losses of animal waste-derived nutrients (N and P) will
be a problem if intense rainstorms occur soon (two to three days) after
waste application.
2. Incorporation of applied annual waste will reduce nutrient losses in
runoff.
3. Waste application tends to reduce soil loss because of mulch effect.
4. Sediment eroded from animal waste-treated areas is highly enriched with
nutrients because of manure particles in transported solids.
5. Concentrations of soluble N and P compounds in runoff from waste treat-
ed areas are high enough to create water quality problems.
6. Waste application to untilled soil gave larger nutrient losses in
runoff than waste application on areas receiving some fall tillage.
7. Large nutrient losses in runoff were observed where soils were nearly
saturated before waste application due to low infiltration and high runoff.
8. Sediment and nutrient losses from waste-treated and untreated over-
grazed pasture were low because of the rapid infiltration of applied rain-
water into the pasture soil.
4.1.3.8 EVALUATION OF FERTILIZER LOSS
The rainfall simulator was used to evaluate the loss of fertilizer and
native soil nitrogen and phosphorus from four soil types found in the Black
Creek watershed. The characteristics of the four soils used are shown in
Table 27. All plots of each soil type were fertilized with 50 Ibs.
phosphorus/ac (54 kg/ha) of triple superphosphate and 150 Ibs. nitrogen/ac
(168 kg/ha) as ammonium nitrate. Rainstorms were applied at an intensity
of 2.50 in/hr (6.35cm/hr). Two rainstorms were applied to the larger
plots, one of 60 minute duration and one of 30 minute duration. The small
plots had one storm applied of 45 minute duration.
STUDY RESULTS
-------
104
TABLE 27
Characteristics of Soils Used in the Investigation
Characteristic*
Slope, %
Clay, %
Silt, %
Sand, %
Total N, ppm
Total P, ppm
Ext. P, ppm
EPC, ppb
Raskins
Loam
0.1-0.3%
12.5
44.5
43.0
1021
363
46
50
Soil
Nappanee
0.7-0.8%
29.5
41.5
28.9
1557
706
44
45
Type
Morley
Clay
4.7-5.2%
33.0
43.5
23.5
1240
366
12
21
Hoytville
Silty Clay
0.3-0.7%
43.8
42.0
14.2
2969
1364
116
115
* Slope is average slope of experimental area; ext. P is amount of
dilute acid soluble P (Bray PI) in soil.
The treatment conditions in this study are very severe and should re-
veal the losses of nutrients under the worst possible conditions for nu-
trient loss. High rates of nitrogen and phosphorus fertilizer were applied
to the surface of the soil just prior to the initiation of a 2.50 in/hr
(6.35 on/hr) rainstorm. This type of situation presents the greatest po-
tential for nutrient loss in runoff water.
The runoff losses of phosphorus and solids are shown in Table 28, and
the losses of nitrogen are shown in Table 29. The losses of soil and nu-
trients were low from these gently sloping soils as compared with losses
reported for other rainulator studies. It has been shown previously that
soil losses of 9.4 and 11 tons/ac (21.38 and 24.74 tons/ha) resulted from
two successive storms applied to a conventionally tilled bedford silt loam
on a 8.3 percent and 12.4 percent slope. Soil losses from rainstorms ap-
plied to the soils in this study ranged from .07 (.15) for the nearly level
Nappanee clay loam to .97 tons (2.18t) for the Morley clay loam soil having
5 percent slope. Soil losses were probably low because of the higher clay
content and reduced slope of these soils as compared to the Bedford soil.
The sediment nitrogen and phosphorus losses were lower than or equal to
those found on the Bedford silt loam. Sediment phosphorus losses on the
Bedford soil were 6.7 to 10 Ibs/ac (7.52 to 11.25 kg/ha) as compared to .39
to 5.3 Ibs/ac (.38 to 5.90 kg/ha) for soils in this study. Sediment nitro-
gen losses were also less for the soils representative of the Maumee River
basin 1.4 to 22 Ibs/ac (1.52 to 24.22 kg/ha) as compared to 19 to 26 Ibs/ac
(21.84 to 28.59 kg/ha) for the Bedford silt loam. Since the sediment
losses from the bedford soil were high, it is logical that sediment nu-
trient losses should also be greater. Plowing down fertilizers on Bedford
silt loam reduced the losses of soluble nitrogen and phosphorus to levels
equal or lower than the four soils used in this study. The concentration
of soluble inorganic phosphorus was reduced the most by incorporation; how-
ever, the concentrations of nitrate and ammonium were also reduced markedly
by plowing. Even though surface application of fertilizer without incor-
poration increases the loss of soluble nitrogen and phosphorus, nutrient
STUDY RESULTS
-------
105
losses, when expressed as a percentage of fertilizer applied, remain quite
low.
H-l
o
Q)
8
IW
Vj
3
cn
|
I
00 O
CN a
|8
w
D
U
O
t
o
fi
i-H
H
cS
w
0)
CO
CO
in
n- E
O »-
O
E u-
oo CL
Q_
ft\
\Lf
z. ro
« r-
r ^.
O tC M
O
Q.
r .
t;
O 4-
/ r\ ^
- J vj
c
tO <4-
-O 14-
O
C
O 3
CO G:
Q
r *4 _T~;
ra 4- ^
4-1 O :/>
O c c
H- 3 0
c: *->
4-1
C
o
[ 1
re
CJ
1-
c
z o ^
V- .
O 4-> c:
4-1 ra
co i- E
a
5 ^
^ 1 Q^
to _c
c E
ro :r
Or
to r-~ oo
-3" rA 1C
j- rA rA
CM rA rH
000
I 1 rH J-
^t CD rA
I-- LA i-H
r-! CM rH
O rH CD
O 0 0
cn CM r*--
i-H O rH
j rA o
F o o o
re
C
i
in
c
CJ O O
^ CO LA LA
I/) CM O CM
rs CM CM rH
LA co r^«
O O "si
rA rA CM
i 1 1 *
_ J Ui_ )
C3 O :A
LO IO J"
rH i 1
CD rA rA
CD CD CM
CO CD rA
CM rH rH
000
cn LO o
J- O CM
O CM CO
O i-H O
CD O O
000
O t-l LA
O CM rH
J- 0 rH
000
O C O
o -3- r-»
O CM r-l
rH rH rH
LA CD -3"
-=t rA LO
CM CM CM
1 1 i 1 1
I-L, __J LX.
LA O O
j- hA rA
^
E
(3
O
_J
ra
^
C-3
C
a,
c
fC
c.
a.
-5
^
o co LA
CO CM _3-
^ r^ r-»
O 0 0
CD CM OO
CO C CM
rA o i-^
f~> ^-^ f~*j
rA m CM
cr rH cr
CD CD O
O 0 0
rA o LA
rH J- rH
O rH O
O O CD
O L-i O
C-J rH rA
rA -it r^
CO CD CM
LO rH f-^
rH CM rH
1 Lt ~~i
' -*- J
O CD LA
CD CO -3"
i-H rH
CJ CM
I 1 CD
r-< j-
rH O
CO CD
LA r~-
j- rA
_ i £^
J- r-t
0 rH
0 O
0 O
0 C J
if, r-:
C-J O
0 C
CD O
rH ^1
rA CM
i-H LA
IA cn
rH rH
t 1 " ^
U_ _y
-A 0
j- rA
CM
LA
cn
CD
0
rH
rA
CD
^^
rH
rA
CO
0
rH
LA
0
O
o
r
r^
i-H
LA
CJ
1 1
1_L.
o
rA
to o
0 CM
01 00
O rH
CM -3-
CM cn
00 LA
O C
rA CM
o o
O 0
5
o
-^ <.O
->LA O
4-1 0 CM
CO
(/)
a;
r
1
._
> o o
4-1 rH CO
>^ j* r^.
_O
^
TA rA
LA rA
CM CJ
_ J LL>
0 0
CD CC
rH
rH
to
-*
rH
CD
O
&
0
0
o
o
CM
LA
O
0
O
oo
CM
L3
CD
CM
LA
J-
rH
r^ cxs cn
cn rA oo
LA J- O
rH rH CM
tO CO rH
cn r-- cn
rA rA cn
O rH O
TA rH r-l
0 0 O
O 0 0
1A O CO
to ^- co
rH 0 0
000
000
O OO CO
r~* '^o oo
CM tC. rA
LO LA CO
rH CJ CM
LA O O
-^ rA rA
CJ
CD CD f^
CM J- CM
r^ o cn
CM to rH
r~\ LA r^
rH O rH
r^ cn cn
CNJ L/i r"H
CM O U3
rH CJ 0
O 0 O
000
E -i- LA -if
TC 0 rH 0
O 0 rH 0
0 0 O
>,
rc
r_
(_j
>,
C 0 0 0
i j- co cn
\- cn rA cn
0 rA-* rH
LA CD -^-
O 03 tA
rA rA CM
CD CD LA
tO LO J"
rH rH
O lii -*
rA rH cn
j- CM rA
CM CM rA
LA OO 1^
LA O rA
CM CM rA
C^l £xj |^*
0 LA rH
O 0 O
O O O
O O O
LA rA tc
r-~ c LA
rH O O
O 0 O
O CD O
cvi cn o
^? rH r>.
rH rA rA
N~> CD CD
O CO CO
CJ CM CN
LA 0 0
J- rA rA
CJ
STUDY RESULTS
-------
106
TABLE 29
Losses of Nitrogen Components in Surface Runoff
Rainstorm Storm
Number Duration Treatment Amm.-N
Nit.-N
Haskins Loam
1 60
60
1 45
45
2 30
30
1 60
60
1 45
45
2 30
30
1 60
60
1 45
45
2 30
30
U
F
U
F
U
F
U
F
U
F
U
F
U
F
U
F
U
F
0.067
0.510
0.051
0.750
0.091
0.547
Nappanee
0.030
0.863
0.880
0.607
0.023
0.810
Hoytville
0.054
0.863
0.051
0.607
0.023
0.810
0.706
0.547
0.714
0.750
0.091
0.258
Clay Loam
0.198
0.295
0.298
0.864
0.074
0.732
Sol.
Org.N
kg/ha
0.088
0.270
0.000
0.000
0.248
0.193
0.092
0.353
0.000
0.000
0.131
0.000
Sed.N
5.270
5.420
3.650
2.690
3.026
3.200
1.700
4.130
1.520
3.100
1.310
3.670
Sum of
N forms
6.131
6.747
4.415
4.441
3.539
4.206
2.020
5.640
1.906
4.571
1.538
5.212
Silty Clay
1.708
3.027
0.350
4.930
0.340
0.546
0.000
0.048
0.086
0.000
0.033
0.000
3.850
3.780
2.980
3.920
3.950
5.430
5.612
7.718
3.467
3.467
4.346
7.596
Morley Clay Loam
1 60
60
1 45
45
2 30
30
U
F
U
F
U
F
0.107
1.629
0.029
0.628
0.066
1.407
0.367
0.777
0.159
0.463
0.348
0.725
0.024
0.000
0.169
0.015
0.000
0.266
20.090
24.220
10.620
7.910
11.370
13.320
20.588
26.626
20.588
9.016
11.784
15.718
It is evident from looking at the results that fertilizer application
does lead to increased nutrient loss from soils. The increases are the
greatest in the soluble nitrate, ammonium and inorganic phosphorus frac-
tions. Sediment phosphorus also appears to increase with fertilizer addi-
tion, probably as a result of sorption of added inorganic phosphorus by the
clay fraction in soil. The loss of sediment nitrogen does not seem to be
markedly affected by fertilization.
The percent of the various types of nitrogen and phosphorus found in
the runoff as compared to the total amounts of nitrogen and phosphorus in
runoff is listed in Table 30. In unfertilized plots, the large majority of
the nitrogen found in runoff is in the sediment. On the contrary, with
fertilized plots the proportion of sediment nitrogen in runoff decreased as
the ammonium-nitrogen and nitrate-nitrogen originating from fertilizer in-
creased. However, the fraction of sediment nitrogen in fertilized plot
runoff was at least 41 percent and, in most cases, greater then 50 percent
of the total nitrogen. In unfertilized plots, almost all the phosphorus in
runoff is sediment phosphorus. When plots are fertilized the percentage of
total phosphorus in runoff present as sediment phosphorus decreases but
STUDY RESULTS
-------
107
stays at relatively high levels. Data from the fertilized plots reveals
that at least 85 percent of the total phosphorus in runoff was in the sedi-
ment phase.
From the above data, it can be concluded that the most effective way
to control loss of phosphorus and, to a lesser extent nitrogen, is to con-
trol soil erosion. Substantial decreases in the total nutrient load would
have occurred if soil erosion were decreased. Most likely the amounts of
soluble nitrogen and phosphorus in runoff would decrease if the fertilizer
were incorporated in a way that would also minimize erosion.
The majority of soluble inorganic nitrogen and inorganic phosphorus in
runoff is derived from fertilizer (Table 31). In most cases the majority
of sediment nitrogen is derived from the soil. The Nappanee clay loam was
an exception since the sediment nitrogen derived from fertilizer sources
was quite high. This finding may be due to the high clay content of the
soil which would increase the probability of large amounts of added ammoni-
um present on the exchange sites of the eroded soil. Substantial propor-
tion of the sediment phosphorus appears to be derived from the added fer-
tilizer. This finding is further substantiated by the increased concentra-
tion of phosphorus in the sediment from fertilized plots as compared to
sediment from unfertilized plots. The increases in sediment phosphorus and
nitrogen in runoff resulting from fertilizer are apparently due to the at-
tachment of ammonium to cation exchange sites and sorption of phosphate to
the clay mineral surfaces shortly after fertilization. These nutrients are
then carried from the plots as components of the sediment during erosion.
Total inorganic nitrogen removed in runoff from the two storms varied
from 0.4 percent to 2.2 percent of that applied (Table 32). losses of ap-
plied inorganic nitrogen from Haskins soils were substantially lower than
the other three soils. This finding may have been due to the high infil-
tration rate in Haskins soil because of higher amounts of sand and lower
clay content than other solids, thus permitting ammonium and nitrate to be
moved deeper into the soil making it less susceptible to runoff. The
amount of fertilizer nitrogen lost in all forms varied from 0.8 percent to
5.9 percent of that added (Table 32). The losses of added nitrogen from
Haskins soil again were much lower than other soils which seems to indicate
clay content of the soil may be the difference since the fertilizer nitro-
gen lost in the sediment phase was probably largely in the form of ammonium
on the cation exchange sites.
The percentages of soluble inorganic phosphorus removal by runoff wa-
ter varied from 0.3 percent to 0.7 percent of that added, and the amount of
added phosphorus lost in all forms in runoff varied from 2.0 percent to 8.0
percent (Table 32). The losses of fertilizer phosphorus were larger than
nitrogen losses as compared to the amounts of fertilizer nitrogen and phos-
phorus applied. The greater losses of applied phosphorus are apparently
from the phosphate sorbed on soil surfaces since the largest amount of
phosphorus in runoff is carried by the sediment. The quantities of fertil-
izer nutrients lost in runoff from the four soils do not represent signifi-
cant monetary losses to the farmer. The nutrient losses are low consider-
ing the severity of the experimental conditions. The incorporation of the
fertilizer would have likely substantially reduced losses.
The average concentrations of sediment phosphorus and extractable sed-
iment phosphorus in runoff increased as a result of fertilization (Table
STUDY RESULTS
-------
o
00
TABLE 30
Percentage distribution of Nitrogen and Phosphorus
Ra i n storm
Number
1
1
2
1
1
2
1
1
2
1
1
2
Storm
Durat i on
(min. )
GO
GO
45
45
30
3U
GO
GO
45
45
30
30
60
60
45
45
30
30
GO
60
45
45
30
30
Treatment
U
F
U
F
U
F
U
F
U
F
U
F
U
F
U
F
U
F
U
F
U
F
U
F
% of Total M in Runoff as:
MH+4-M
NO-3-N
Raskins Loam
1.1
7.6
1.1
16.9
2.6
13.0
11.5
8.1
16.2
22.5
4.9
6.1
Nappanee Clay Loam
1.5
15.3
4.6
13.2
1.5
15.5
9.8
5.2
15.6
18.9
4.8
14.0
Hoytvi lie S 1 ty Clay
1.0
11.2
1.5
6.4
0.5
10.7
30.4
39.2
10.1
52.1
7.8
7.2
Morley Clay Loam
0.5
6.1
0.3
7.0
0.6
9.0
1.8
2.9
1.4
5.1
3.0
4.6
Sol .
Org.ri
1.4
4.0
0.0
0.0
7.0
4.6
4.5
6.3
0.0
0.0
8.5
0.0
0.0
0.6
2.5
0.0
0.8
0.0
0.1
0.0
1.5
0.2
0.0
1.7
Sed.M
86.0
80.3
82.7
CO. 6
85.5
76.1
84.2
73.2
79.7
67.8
94.9
70.4
68.6
49.0
86.0
41.5
90.9
71.5
97.6
91.0
96.7
87.7
96.5
84.7
% of Total P in Runoff as:
Soil
Inorg.P
0.8
9.1
1.3
13.8
1.9
8.7
3.2
13.5
2.0
14.6
3.0
5.1
6.0
11.3
3.6
10.3
3.6
4.2
0.1
1.9
0.2
7.2
0.1
1.6
Sol .
Org.P
0.7
3.8
0.8
0.0
1.1
0.6
0.8
1.4
0.3
0.2
2.7
1.3
3.3
1.1
0.0
2.2
0.8
0.5
0.4
0.3
0.3
0.0
0.2
0.0
Sed.P
98.5
87.2
98.0
86.2
97.2
90.7
96.0
85.1
97.7
85.2
94.3
93.6
90.7
87.6
96.4
87.4
95.8
95.3
99.4
97.8
99.5
92.8
99.6
98.3
-------
TABLE 31
Percentage Nutrients
Soi 1
Type
Haskins Loam
iJappanee Clay Loam
Hoytvi lie Si 1 ty Loam
Morley Clay Loam
Rn i nstorm
Number
1
2
1
2
1
0
L.
1
2
Form of \\ in Runoff
Sol .
1 norg . fj
% of \\ Der
26.9
67.1
80.3
93.7
5k. 7
73.2
80.3
80.7
Sed.N
ived fn
2.8
5.1*
58.8
P4.3
n.o
27.3
17.1
Ik. 7
Sum of Al 1
N Forms
am Fertilizer
9.1
15.8
6k. 2
70.5
27.3
42.8
22.7
25.0
Form of P in Runoff
Sol .
Inorg. P
% of P Per
93.7
81.7
90.7
76.5
73.8
44.3
9U.5
9U.6
Sed.P
ived frc
17.2
11.7
75.0
59.3
48.5
30.8
54.1
33.8
Sum of All
P Forms
>m Ferti 1 izer
26.7
17.4
61.9
59.6
50.2
31.2
54.8
34.7
* Calculated by subtracting the nutrient loss from the un-
treated plot from that of the fertilized plot, dividing the
difference by the nutrient loss from the fertilized plot,
and multiplying the resultant value by 100.
en
I
o
vo
-------
11Q
TABLE 32
Added Fertilizer Nitrogen and Phosphorus Lost in Runoff
Added N lost in
Soil
Type
Rainstorm
Number
runoff
Sol.
Inorg.N
as:
All N
Forms
% of added N
Haskins Loam
Nappanee Clay Loam
Hoytville Silty
Clay
Morley Clay Loam
1
2
Total
1
2
Total
1
2
Total
1
2
Total
0.2
0.2
0.4
0.6
0.9
1.5
1.3
0.6
1.9
1.2
1.0
2.2
0.4
0.4
0.8
2.1
2.1
4.2
1.3
1.9
3.2
3.6
2.3
5.9
Added P lost in
runoff as:
Sol. All P
Inorg . P Forms
% of added P
0.5 1.6
0.2 0.4
0.7 2.0
0.2 1.1
0.1 1.1
0.3 2.2
0.3 1.6
0.1 1.2
0.4 2.8
0.2 5.9
0.1 2.1
0.3 8.0
* Percent of added nutrients lost in runoff was calculated by subtracting
the nutrient loss from untreated plots from that of the fertilized plot,
dividing the difference by the amount of nutrient added, and multiplying
the resultant value by 100.
33). Addition of superphosphate decreased the proportion of total phos-
phorus in runoff present as sediment phosphorus from 96.6 percent to 91.8
percent. On the average, the total phosphorus content of the sediment in-
creased 269 ppm, and the extractable phosphorus content of the sediment in-
creased 97 ppm as a result of fertilization. Superphosphate addition in-
creased the proportion of sediment phosphorus which was extractable with
the Bray PI solution from 20.6 percent to 25.2 percent suggesting that a
higher percentage of added phosphorus associated with the sediment was ex-
tractable than native phosphorus associated with the sediment. The finding
that in excess of 90 percent of the total phosphorus in runoff is sediment
phosphorus agrees with previous work and suggests that control of soil ero-
sion can greatly reduce the levels of total phosphorus in surface runoff.
The concentration of sediment nitrogen in runoff was not markedly af-
fected by fertilization although the amount of sediment exchangeable am-
monium increased as a result of ammonium nitrate addition. The proportion
of total nitrogen in runoff present as sediment nitrogen decreased from 86
percent to 71 percent as a result of fertilization. The fact that sediment
nitrogen makes up the bulk of the nitrogen in runoff suggests that control
of soil erosion can greatly reduce the total amounts of nitrogen in runoff
although the more available soluble nitrogen components would still be
present. Fertilization increased the average Kjeldahl nitrogen content of
the sediment about 280 ppm and the exchangeable ammonium content about 81
ppm. The proportion of Kjeldahl nitrogen in the sediment present as ex-
changeable ammonium increased from 1.2 percent to 5 percent with fertiliza-
tion.
The soluble organic carbon content of runoff averaged 19.6 mg/1 for
unfertilized plots and 28.3 mg/1 for fertilized plots. The values are
somewhat higher than the soluble organic carbon contents of streams and
STUDY RESULTS
-------
Ill
TABLE 33
Effect of Fertilization
Effect of Fertilizati<
Composition of Sedimen
to
Soil Treat- Rain Sol.
Type ment* storm P
Haskins
Nappanee
Morley
Hoytville
Average
U
F
U
F
U
F
U
F
U
F
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
530
454
520
643
685
661
267
422
258
547
707
828
553
513
477
665
697
823
1230
1355
1202
1375
1466
1414
607
876
Dn on
t in
Simi
c
ppn
76
65
94
213
150
178
113
153
127
199
167
155
25
31
34
196
146
161
231
248
217
368
320
331
118
215
the Phosphorus
Runoff from Fo
ilated Rainstor
Ext. Sed.
5ed.P EPC
i ppb P-
(14.7)
(14.3)
(18.1)
(33.1)
(21.9)
(26.9)
(42.3)
(36.0)
(49.2)
(36.4)
(23.6)
(18.7)
(4.5)
(6.0)
(7.1)
(29.5)
(20.9)
(19.6)
(18.8)
(18.3)
(13.1)
(26.8)
(21.8)
(23.4)
(20.6)
(25.2)
28
35
30
455
180
175
74
44
44
300
136
136
4
10
9
104
54
48
15
86
94
284
138
138
39
179
3, Nitrogen and Carbon
ur Soils and Subjected
ms
Sed . Sed . Sed .
Total N exc.NH4-N Org.C
VWMt O.
1330
1710
1590
1750
1610
1620
790
2570
1040
1840
1830
1590
1860
1650
1590
2070
1890
2030
2350
3190
3160
2870
3570
3490
1903
2180
ft*"
46
47
60
105
116
73
0
0
0
125
89
104
26
28
20
95
100
80
0
0
0
118
114
103
19
102
(3.5)
(2.7)
(3.8)
(6.0)
(7.2)
(4.5)
(0.0)
(0.0)
(0.0)
(6.8)
(4.9)
(6.5)
(1.4)
(1.7)
(1.3)
(4.6)
(4.8)
(3.9)
(0.0)
(0.0)
(0.0)
(4.1)
(3.2)
(2.9)
(1.2)
(5.0)
t
0.
1.
1.
1.
1.
2.
3.
3.
3.
2.
5.
2.
1.
1.
1.
1.
1.
2.
2.
5.
3.
5.
3.
2.
2.
2.
53
13
00
18
00
64
48
89
76
50
11
64
45
18
04
68
38
35
13
11
78
81
81
90
37
62
* U-unfertilized; F-fertilized
Values in parenthesis are extractable P as a % of sediment total P and
changeable ammonium as a % of sediment total N.
ex-
rivers, which are usually about 10 mg/1. The average sediment organic car-
bon concentration in runoff was about 200 mg/1, thus the organic carbon to
Kjeldahl nitrogen ratio (C/N) for surface runoff was approximately 11.
Soils normally have a C/N ratio of 9 to 12, so it can be seen that sediment
carbon and sediment nitrogen are being eroded in roughly the same propor-
STUDY RESULTS
-------
112
tion that they occur in the soil.
There was a strong relationship between the solids content of runoff
and the sediment phosphorus concentration in the runoff. A similar rela-
tionship was observed between solids and the sediment nitrogen content of
runoff. These findings suggest that the solids content of surface runoff
provides a good indication of the relative amounts of sediment nitrogen and
phosphorus present in runoff. It may be possible to obtain a semi-
quantitative estimate of the sediment nitrogen and phosphorus content of
runoff based upon solids content in a given watershed if most soils in the
watershed are similar.
The solids content was also significantly correlated with the soluble
organic phosphorus concentration, the soluble organic carbon content, and
the exchangeable ammonium concentration in the sediment. Thus, it appears
that the solids content of surface runoff from these soils is a key factor
in determining the concentrations of certain forms of nitrogen, phosphorus,
and carbon in runoff. However, solids content may not always be related to
sediment nitrogen and phosphorus content since soil type, slope of the
soil, natural fertility and other soil associated factors can affect the
composition of the sediments inducing variability. Therefore, the nitrogen
and phosphorus concentration in sediment may change as related to the ori-
ginal soil because of selective enrichment by smaller size fractions, espe-
cially clay. When prediction of nutrient content of sediments is made,
these factors need to be considered.
Clay content of the runoff was related to solids, soluble organic
phosphorus, sediment total phosphorus and organic carbon in the sediment.
It appears that the clay fraction carried a large portion of these nu-
trients contained in the runoff. Clay particles especially the related
amorphous material, have a high affinity for phosphate ions and organic
molecules.
A very high correlation coefficient (r=0.91) was obtained for the re-
lationship between the soluble inorganic phosphorus concentration in runoff
and the equilibrium phosphorus concentration (EPC) of runoff sediment. It
appears that the concentration of soluble inorganic phosphorus in runoff
may be accurately estimated by determining the EPC of the sediment. It is
interesting to note that the relationship between soluble inorganic phos-
phorus and sediment EPC was observed even when results from several soils
were combined and when the solids content of runoff varied from 0.17 per-
cent 2.74. These findings suggest that the relationship between the sedi-
ment EPC and the soluble inorganic phosphorus concentration in water is
similar for different soil types.
The original soil EPC, when compared to soluble inorganic phosphorus
concentration of the runoff from unfertilized plots, resulted in a correla-
tion coefficient of 0.81, which indicates under certain conditions the ori-
ginal soil EPC may be useful in predicting the soluble inorganic phosphorus
in runoff. The nature of the equilibrium between the soil and solution
makes it possible for the soil to buffer solution phosphorus. Soil, when
associated with solutions of low phosphorus status, may desorb phosphorus
whereas under conditions of high concentrations of solution phosphorus the
soil may sorb phosphorus. If the composition of the runoff is known, pred-
ictions can be made to assess the contribution of sediment phosphorus in
STUDY RESULTS
-------
113
runoff to the phosphorus status of lakes and streams.
Sediment phosphorus in runoff was related to the concentration of
soluble phosphorus in runoff and to the concentration of soluble organic
carbon and sediment organic carbon in runoff. As expected the concentra-
tion of extractable phosphorus in the sediment was directly related to the
concentration of total phosphorus in the sediment. The concentration of
extractable phosphorus in sediment was also related to sediment EPC. This
finding suggests that the EPC technique and the Bray P-l extraction pro-
cedure measure a similar fraction of soil phosphorus which determines the
concentration of orthophosphate iji equilibrium solutions. It has been ob-
served that EPC measures the buffering capacity of a soil for phosphate,
whereas Bray P-l extraction measures the phosphate potential of the soil.
The soluble ammonium nitrogen concentration in runoff was related to
the exchangeable ammonium nitrogen concentration in runoff and to the ex-
changeable ammonium nitrogen content of the sediment. This finding sug-
gests that an equilibrium exists between the soluble ammonium in runoff and
the exchangeable ammonium present on solids in runoff. Thus, it may be
possible to predict one of these forms of ammonium in runoff if the other
form is known. Soluble organic nitrogen in runoff is related to the con-
centration of sediment nitrogen in runoff and to the concentration of ex-
changeable ammonium in the sediment. It appears likely that a given pro-
portion of organic nitrogen in these soils is solubilized during runoff
events and therefore soils containing high organic nitrogen contents have
runoff higher in soluble organic nitrogen than soils containing low concen-
trations of organic matter.
The four soils used in the study were separated into their sand, silt,
and clay fractions for chemical analysis. The fractionation procedure gave
fractions which represented 94.3 percent of the weight of the original sam-
ple. A higher recovery could have been achieved if the large volumes of
suspended clay could have been air-dried in a short period of time. The
freezing process made recovery much faster since a major portion of the su-
pernatant could be decanted.
The results of the chemical analysis of the soil constituents and the
original soils are presented in Table 34. The analyses of samples for to-
tal phosphorus, total nitrogen, and extractable phosphorus all indicate
that the highest concentrations of nitrogen and phosphorus in soil are as-
sociated with the clay fraction.
Phosphorus adsorption on clays has been related to the presence of
amorphous oxides and hydrous oxides or iron and aluminum. The presence of
iron oxide and hydrous oxide coatings on clay mineral surfaces has been re-
ported. Such coatings, along with greater surface area, may explain the
fact that a large proportion of the total phosphorus in the soil is associ-
ated with the clay fractions. Since silt and clay are, in many cases, pre-
ferentially eroded, higher losses of sediment nitrogen and phosphorus would
be expected than if the soil were eroded in mass. This observation has led
to the use of enrichment ratios where the concentration of sand, silt, and
clay in the particulate phase of surface runoff is compared to concentra-
tions in the soil. Previous researchers have observed increases in clay
content of eroded materials. For a soil containing 16 percent to 18 per-
cent clay, clay percentage in the eroded material increased from 25 percent
to 60 percent as runoff diminished from 2.7 inches to .01 inches (70 mm to
STUDY RESULTS
-------
114
TABLE 34
Analysis of Soil Size Fraction
Soil Type
Fraction
Percent
of Soil
Total
N
P
u
Haskins
Loam
Morley
Clay Loam
Hoytville
Silty Clay
Nappanee
Clay Loam
whole soil
sand
silt
clay
whole soil
sand
silt
clay
whole soil
sand
silt
clay
whole soil
sand
silt
clay
100.0
43.0
44.5
12.4
100.0
23.5
43.4
33.0
100.0
14.2
42.1
43.7
100.0
28.9
41.6
29.5
1021
166
710
4406
1240
225
835
2165
2969
424
1794
4466
1557
182
972
3231
364
168
240
1135
366
90
127
739
1241
704
756
1364
706
399
335
1109
Ex tractable
P*
/g
46
29
36
155
12.4
10.5
10.5
16.1
117
49
102
166
44
21
34
75
EPC**
ng/ml
5
3
28
0
0
0
0
7
44
1
2
9
* Bray P, (Jackson, 1970).
** EPC - equilibrium phosphorus concentration (Taylor and Kunishi, 1971).
0.25 mm) of runoff per hour. Runoff from storms of lesser intensity may
carry as much sediment phosphorus as intense storms due to greater enrich-
ment during low runoff.
If selective erosion occurs, the potential for enriching water with
nutrients is increased because the concentration of nutrients in the clay
fraction is much higher than the whole soil. It is apparent that selective
erosion occurred in these soils because the nitrogen and phosphorus concen-
tration of the sediment from unfertilized plots are higher than those of
the whole soil and in some cases higher than the concentration in the clay
fraction of the soil. This finding suggests that the fine clay particles
are being eroded preferentially and could be a potential source of nitrogen
and phosphorus to the solution due to the higher nutrient concentration of
the fine clay.
Phosphorus adsorption isotherm relations were determined for the three
factions of each soil. The isotherms indicate clay has the greatest abili-
ty to buffer the soluble phosphorus concentration. The clays, when sub-
jected to concentrations of soluble orthophosphate greater than the EPC,
will sorb phosphorus from solution. Clay in stream and lake systems can
serve as a phosphate sink during periods of high phosphorus concentration.
Even though clay may holding the largest fraction of phosphorus in streams
and lakes it also has the ability to maintain the concentration of phos-
phorus in solution at levels lower than in runoff which enters the water.
The concentration maintained is most likely dependent upon the EPC of the
STUDY RESULTS
-------
115
eroded material and the EPC of the material present on the stream bank and
lake bottoms.
4.1.3.9 CONCLUSIONS
Fertilizers increase the soluble nitrate-nitrogen, ammonium nitrogen
and orthophosphate-phosphorus content of runoff. These losses are small
(none greater than 8 percent) as compared to the amount of fertilizer nu-
trient added to the soil. The soluble inorganic nitrogen and phosphorus in
runoff are readily available for algal growth; and therefore, reduction of
the concentration of nutrients in runoff is desirable.
Sediments contain the majority of nitrogen and phosphorus in runoff.
Sediment nitrogen and phosphorus are not immediately available to aquatic
plants but these forms may serve as potential nutrient sources. The
greatest portion of nitrogen and phosphorus associated with sediment ap-
pears to be in the clay fraction. Soils susceptible to selective erosion
may yield sediment which contain more nitrogen and phosphorus than the ori-
ginal soil since clay and silt percentages of sediment generally increase
relative to the soil during erosion. The best approach for reducing nu-
trient loss from cropland appears to be erosion and runoff control.
Practices such as fertilizer incorporation are important as indicated
by this study and previous investigations. Conservation tillage methods
need to be evaluated with respect to control of erosion and runoff and sui-
tability to present day farming practices.
The findings of this study allow one to make certain inferences about
the composition of runoff from soils of the Upper Maumee River watershed.
It appears that the solids and clay content of the runoff are the most im-
portant parameters in controlling the concentrations of several forms of
nitrogen and phosphorus in runoff. The data suggests that control of soil
erosion and proper incorporation of fertilizers (to prevent mass movement
of fertilizer in runoff water) can greatly reduce the concentrations of nu-
trients in runoff from the soils studied.
The concentration of soluble inorganic phosphorus in runoff (possibly
the most important single parameter in eutrophication) can be estimated by
measurement of the EPC or the extractable phosphorus content of eroded ma-
terial. The EPC value of unfertilized soil or soils in which phosphorus
fertilizers have been incorporated can be used to estimate the soluble
inorganic phosphorus concentration in runoff. These relationships should
be very useful in development of models for prediction of nutrient loss in
the Maumee River watershed.
The concentration of soluble ammonium in runoff can be estimated from
the exchangeable ammonium content of the runoff sediment. It also appears
likely that the exchangeable ammonium content of the soil is related to the
soluble ammonium concentration in runoff. The concentration of nitrate in
the runoff was not related to any runoff properties which were measured.
This is to be expected in that nitrate is water soluble and is not associ-
ated with the solid phase of the soil. This finding suggests that develop-
ment o a model for loss of nitrate in runoff water will be very difficult
because nitrate losses are dependent upon a large number of hydrologic and
soil factors.
STUDY RESULTS
-------
116
References
First Annual Report, Black Creek Project, Allen County Soil and Water Con-
servation District.
Schroeder,Steve, Soil Aggregates Transported in Runoff From Cropland and
Their Relationship to Total Soil Loss, Purdue University, May 1976.
Kaminsky, Jr., D. B., Nitrogen and Phosphorus in Surface Runoff from Agri-
cultural Soils, Purdue University, May 1975.
4.2 BANK STABILITY STUDIES
Two study areas were selected to determine the effect of slope and
mulching materials on the revegetation of the stream banks and of the ef-
fects of the mulching materials in controlling erosion during the revegeta-
tion. Site 1 on the upper end of the Dreisbach Drain (on the Joe Graber
farm) was installed in September-October, 1973. Site 2 on the Wertz Drain
between Notestine Road and Black Creek (on the Dick Yerks farm) was in-
stalled in April, 1974. Three slopes, 2:1, 3:1, and 4:1 were used at each
site. Mulch materials of stone, straw, and wood chips along with a check
or no-mulch were used on both locations. In addition, aquatain and sawdust
were used on the Joe Graber farm.
The final evaluation on May 2, 1975, showed that all mulches were ef-
fective in controlling erosion and in helping to establish cover. There
was no consistent difference in the mulch material effectiveness with the
exception of the stone mulch. Stone appeared to be slightly superior in
controlling erosion resulting from high water, and resulted in as good or
better grass cover than other mulch materials. In May of 1974 both the
wood chip and straw mulch materials were washed away during high water flow
in the Wertz Drain. While there is not a totally consistent advantage of
one mulch material over the other, all mulches improve the establishment of
the grass and help control erosion during the establishment period.
The 3:1 slopes appeared to be slightly better than either the 2:1 or
the 4:1 slopes. This can vary with local conditions. For example, on the
Graber farm the 3:1 slopes are far superior to the others but this is com-
plicated by the fact that there was more good soil on the 3:1 slopes than
on any of the other two slopes. In fact, a year and a half after the es-
tablishment and planting of grass, the 2:1 and 4:1 slopes on the Graber
farm still showed evidences of low fertility. Grass cover helped control
erosion in all cases.
Bank stability studies were a part of original plan of work. To
determine if a correlation might exist between bank cover, particularly
trees vs. grass, and bank stability. The reported data of the SCS study in
Black Creek has been reviewed. While this data shows a strong correlation
between soil type and bank erosion it is not possible to relate erosion and
cover in the published data.
Three sites were initially selected for studies to determine if the
channels were agrading or degrading. Two of the sites have since been
reconstructed and revegetated. Observation of the third site (Wertz Drain
south of the woods and north of Antwerp Road) indicates that the channel is
STUDY RESULTS
-------
117
relatively stable with little or no accumulation or loss.
High water flow during May of 1974 deposited one to two inches of silt
and sand on the mulch study plots on the Wertz Drain. Other evidences of
such deposition during high flows has been observed throughout the
watershed. There is also evidence of continued cutting of some of the
channels where velocities of flow are high. This seems to be especially
true where velocities are high during normal to low flow such as occurs on
the mulch study area on the Dreisbach Drain. Observations of channel scour
are in agreement with stability studies which showed that many of the lower
bank and channel bottoms were unstable for existing flow and slope condi-
tions throughout many areas of the Maumee Basin.
Soil mechanics studies have identified several locations where channel
bottoms are potentially unstable. The most likely reason for instability
is excess channel slope and a less resistant soil material in the profile
near the channel bottom. It is evident that if a channel bottom degrades
even stable banks eventually must become unstable.
Four sites were selected for study of channel bottom stability. One,
on the Joe Graber farm, was known to have lowered 30 to 60 cm (1 to 2 feet)
following revegetation of the banks. In this area, small rock drop struc-
tures were installed in 1975.
The Black Creek channel at Notestine Road was surveyed for a distance
of 30 m (150 ft.) upstream and 65 m (200 ft.) downstream from the bridge.
This is an area where rock was used for channel training. It is also an
area where soil mechanics studies indicated a potentially unstable channel
at flood flow. This channel was shown to be unstable because of the soil
material in the channel bottom and also the slope, 0.25 percent. This 115
m (350 ft.) degraded approximately 42 cm (1.4 ft.) between May of 1974 and
August of 1976. The channel has considerable grass and other water type
vegetation in the bottom.
Another site on the Gorrell drain along Notestine Road for a distance
of 165 m (500 ft.) downstream from the monitoring site shows the ditch bot-
tom to be almost identical with the original. This section has an average
slope of 0.2 m per 100 m (0.20 percent). This is the smallest slope of any
of the four sites studied.
Wertz drain between Notestine Road and the main channel of the Black
Creek, a distance of approximately 305 m (1000 ft.), was a site of the bank
slope-mulch studies. This channel reach has an average slope of 0.4 per-
cent (0.4 meters per 100). Earlier observations had indicated that the
channel bottom was eroding in several sites. The survey conducted in Au-
gust of 1976 showed that with the exception of a section about 200 meters
below the Notestine Road all of the channel had eroded. For the first 160
m (500 ft.) an average lowering of approximately 30 cm (1 ft.) occurred
between March of 1974 and August of 1976. The last 70 meters (200 ft.)
above the main Black Creek channel eroded approximately 45 cm (1.5 ft.).
There are several areas in this section where erosion of the channel bottom
has caused the top of the banks to slip into the channel.
These survey results plus other observations indicate that there are a
number of sections throughout the Black Creek watershed where channel bot-
tom erosion is producing unstable bank conditions.
STUDY RESULTS
-------
118
Survey of Streambank Erosion
As a part of the International Joint Commission PLUARG studies,
streambank erosion is being surveyed under the direction of a Mr. W. F.
Mildner of the Soil Conservation Service. These studies are generally made
on a statistical sampling basis, but because of the intensity of the Black
Creek Study and the size of the basin, a complete survey was made in this
particular watershed.
The study was conducted in the summer of 1975 and the data has recent-
ly been made available to the author by Mr. Mildner. The survey covered
29.3 miles on stream for a total bank miles of 58.6. The following meas-
urements are listed in bank miles. The drainage density was determined to
be 1.56 miles of channel per square mile. It was determined that there are
7.2 eroding bank miles producing approximately 400 tons of sediment per
year. It was estimated that only 6.3 miles of simple treatment and 0.9
miles of armoring would solve this problem. At the present time there are
18.4 miles of simple treatment and 1.7 miles of armoring in the watershed.
There does not appear to be any correlation of the bank erosion with
use of the adjacent land or with whether or not the banks are fenced. It
is interesting to note that over 80 percent of the total tons of streambank
erosion are produced by two soil types, Eel 59.4 percent and Shoals 25.1
percent. Yet the same soils account for only 18.7 and 7.3 percent respec-
tively of the total miles of streambanks.
While this survey shows a relatively small amount of sediment produced
by streambank erosion, nevertheless at the site of occurrence the erosion
may be quite severe. Often the eroding sections can be controlled with a
comparatively small amount of simple treatment. The need to control bank
erosion emphasizes a need for a good maintenance program in any plan to
reduce sedimentation.
4.3 SEDIMENT BASINS
The sediment pond, a small sediment basin, was constructed on the Vir-
gil Hirsch farm in the fall of 1973. It filled to overflowing in November
of that year. It serves a drainage area of 185 ha (460 acres). The soil
types are Hoytville'and Nappanee. The land slopes are generally less than
one-half per cent. Sediment depositions were determined by fathometer and
by probing. Sediment was examined for determination of particle size.
Sediment samples for laboratory analysis were collected.
Sediment deposits were found to be uniform is depth throughout the
pond area with an average accumulation of 6 cm (0.2 ft.). Particle sizes
were uniform, being primarily in the clay and silt fractions, with a small
amount of fine sand.
Laboratory analysis of the samples confirmed that the sediment is a
silty clay texture. The range of the sample analysis were as follows:
Silt 52.1 63.9%
Clay 31.9 42.0%
Sand 4.2 5.9%
The sediment pond accumulated 1880 cu m (2400 cu yd.) of sediment.
STUDY RESULTS
-------
119
Assuming a dry weight of 857 kg per cu meter (55 pounds per cubic foot)
this amounts to an average of 2.8 ton per ha per year (1.2 tons of sediment
per acre per year). However, this figure cannot be projected as a long term
average because of two factors:
1) The area immediately to the north of the pond site was in a tran-
sition stage and was very subject to erosion until the conservation
practices on it were completed in 1975. Thus, this area may have con-
tributed above a normal amount of sediment in this three year period.
There has also been some construction activity on the west end of the
pond site.
2) In May of 1975, a nearly 100-year frequency storm was received.
This storm produced the highest runoff volume and sediment concentra-
tions yet measured at many of the stations. It produced between 1/3
and 1/2 of the 1975 annual sediment transport at some of the measuring
stations.
The Desilting Basin on the main stream of the Black Creek was con-
structed in September of 1974 and was first surveyed on July 30, 1975. A
second survey was conducted July 7, 1976. Sediment samples were collected
from this basin for particle size determination.
The first survey covering a period of approximately nine months showed
an accumulation of 770 cubic meters (980 cubic yards) of material. The
second survey showed an additional accumulation of 416 cu meters (530 cubic
yards) in approximately a one year additional time. Sediment sample
analysis is shown in Table 35. This table shows only the per cent sand by
size fraction and does not include the finer silt and clay fractions. It
is only at stations 460 and 461 where less than one-half of the sediment
accumulated was in the sand size fraction.
TABLE 35
Percent Sand and Sizes in Desilting Basin Deposits at Stations Listed
Size Range in mm
Station
457+000
458+000
459+000
460+000
461+000
>2
26.58
7.94
.85
.38
0
2tol
11.68
5.15
1.09
.46
0
lto.5
17.36
11.52
2.41
.19
.13
.5to.25
27.46
46.07
14.88
1.28
.88
.25to.l
6.67
16.43
33.01
17.02
13.59
.lto.05
1.48
2.44
13.93
21.24
21.08
Total
%*
Sand
91.23
89.55
66.17
40.57
35.68
Percent sand is based on total dry oven weight of sample.
This would indicate that much of the material being trapped by this
desilting basin is bed-load. To date, no evidence has been seen of addi-
tional scour of the channel immediately below the desilting basin. The
first 50 meters (150 ft.) of this basin is nearly full. If it continues to
trap material at the present rate, it will have to be cleaned out to remain
effective.
STUDY RESULTS
-------
120
4.4 MICROBIOLOGICAL STUDIES
Microbiological studies were conducted on grab samples collected at
stations throughout the watershed.
To discover the variability associated with grab sampling and analysis
of the water samples, triplicate samples were collected from station 312 on
April 31, 1977. Agreement between the samples was good (see Table 36).
However, the estimated standard deviations are, in many cases, as large as
the differences between stations (except where septic tank effluent greatly
elevated the counts). Due to this level of variability and an incomplete
data record at many stations, it was decided not to examine the data from
individual stations. Instead, the general patterns of bacterial contamina-
tion in the Black Creek watershed were defined by grouping similar stations
together. The results, during conditions of low stream discharge, are
presented in Section 4.4.1.
TABLE 36
Bacterial Counts From Surface Water Station 312, May 31, 1977
Total col i form
Mean
Standard
Deviation
2,700
1,600
1,700
2,000
608
Fecal col i form
400
700
100
400
300
Fecal streptococcus
100
100
200
133
64
4.4.1 Bacterial Counts at Low Flow
On five of the six sampling dates, the stream discharge in Black Creek
was considered to be low (see Table 37). This distinction was not made on
the basis of an arbitrary numerical discharge rate but rather on the pres-
ence or absence of surface runoff from agricultural land. At low stream
discharge there was no surface runoff, while at high stream discharge there
was substantial surface runoff.
A total of 65 observations from 20 surface water stations are avail-
able to estimate the bacterial counts expected in the Black Creek watershed
given normal conditions of low stream discharge. Table 38 summarizes this
information. The mean values are unrealistically high because of unusual
environmental conditions, such as nearby septic tank outfalls, that created
a few extremely high values. The median values of 3,500 total coliforms,
1,000 fecal coliforms and 200 fecal streptococcus per 100 ml of water serve
as our best estimates of bacterial contamination in the Black Creek
watershed as a whole during low stream discharge. However, these figures
do not accurately describe the fecal pollution that existed at many locales
in the Black Creek watershed as some stations were far more polluted and
others had coliform concentrations which approached the most stringent
clean water guidelines.
STUDY RESULTS
-------
121
TABLE 37
Stream Discharge Class By Date for Two Sampling Sites
Date
3-19-76
6- 7-76
8-23-76
11-29-76
3-28-77
5-31-77
Average Daily Discharge m"3/sec
Site 2
.048
.008
.004
.000
1.007
.001
Site 6
.025
.004
.000
.002
.775
.001
Discharge
class
low
low
low
low
high
low
TABLE 38
Total Coliform, Fecal Coliform and Streptococcus Count - Low Discharge
Total coliform Fecal coliform Fecal streptococcus
Count/100 ml wate-i?
Range 100-2,600,000 0-2,600,000 0-890,000
Median 3,500 1,000 200
Mean 164,668 109,114 19,114
4.4.2 Contamination at Low Flow
Because reliable estimates of bacterial contamination at specific sam-
pling stations could not be made, the stations were grouped into six
categories based on levels of discharge and organic pollution. When not
carrying storm runoff, some streams in the Black Creek watershed maintained
a base discharge from groundwater. Others ceased to have any discharge and
in the summer months became a series of isolated pools. A few sections of
stream had an intermittent flow arising from domestic waste effluent. Fig-
ure 15 outlines the location of these three flow regimes in the Black Creek
watershed. The major source of sewage was the town of Harlan, but indivi-
dual septic tank outfalls occurred throughout the watershed.
The grouping of stations was done by assessing the locality in rela-
tion to both the proximity of sewage outfalls and the stream discharge re-
gime. Table 39 shows these groupings along with mean bacterial counts
plus/minus the standard deviation for each group. The large standard devi-
ations reflect the highly variable nature of the data caused in part by the
wide range of environmental conditions encountered during low stream
discharge. There were no significant differences between the group means
due to the high variances. However, the mean bacterial counts did show
trends consistent with a subjective assessment of organic pollution in each
of the groups.
In Figure 16, the data for each group is presented to show the fre-
quency of six levels of bacterial contamination. Groups B and F, stations
with high levels of organic pollution, had broad distributions of total
STUDY RESULTS
-------
122
Base Flow
"Harlen" Flow
"No Flow"
Figure 15. Flow Regimes in Watershed
STUDY RESULTS
-------
TABLE 39
Surface Water Stations According to Stream Discharge and Organic Pollution
123
Group
Number
of
stations Discharge Organic Mean +- stand deviation
in group Regime Pollution (Count per 100 ml water)
Total Fecal Fecal
coliform coliform
A
B
C
D
E
F
2
2
4
4
6
2
Isolated
pool
Intermittent
flow
Intermittent
flow
Base
flow
Isolated
pool
Isolated
pool
Moderate
High
Moderate
Low
Low
High
5
+-2
554
+-473
16
+-5
7
+-14
2
+-2
537
+-9 30
,333
,754
,900
,800
,010
,903
,243
,398
,316
,640
,600
,260
3
+-4
671
+-954
3
+-7
4
+-13
+-1
29
+-47
,666
,038
,450
,069
,850
,424
,079
,508
895
,030
,656
,500
streptococcus
200
+-346
116,190
+-278,800
480
+-539
300
+-568
216
+-432
7,422
+-19,000
coliform counts and median values of 435,000 and 59,000 respectively. The
moderately polluted groups, A and C, also had wide distributions but median
total coliform counts were 4,000 and 3,450 respectively. The remaining two
groups, D and E, had low levels of pollution, a narrower distribution of
total coliform counts and median values of 2,850 and 1,750 respectively.
The distribution of fecal coliform counts was very similar to that just
described for total coliforms. The median values were as follow: highly
polluted groups, 165,000 (3) and 7,200 (F); moderately polluted groups,
1,800 (A) and 950 (C); slightly polluted groups, 500 (D) and 400 (E). Fe-
cal streptococcus contamination was slight except in the highly polluted
areas where median counts were 5050 (Group B) and 1,600 (Group F). The
remaining groups of stations had median counts less than 200 fecal strepto-
cocci per 100 ml water.
It is evident that, at low stream discharge, the degree of fecal con-
tamination was dependent upon the proximity to sewage outfalls and the type
of flow in the stream. The fecal coliform counts indicate that sections of
the Black Creek drainage receiving septic waste have fecal contamination
far in excess of any public health standards. Other areas of the drainage
removed from nearby septic tank pollution were found to occasionally meet
federal public health standards for fecal coliform contamination but gen-
erally these areas had twice the allowable limit of 200 fecal coliforms per
100 ml of water.
On 28 March, 1977, there was considerable surface runoff from agricul-
tural land, and stream discharge in Black Creek was high (.775 cubic
meters/sec). Table 40 summarizes the data collected during this runoff
event. The mean and median values were not greatly different because sep-
tic tank effluent in the badly polluted stream sections was being diluted.
The median values for total coliform, fecal coliform and fecal strepto-
coccus counts were 5, 3, and 17 times greater, respectively, than the medi-
an values during low stream discharge (see Table 38).
The increase in total and fecal coliform counts was caused by higher
counts at stations on Black Creek downstream of Harlan, The tributary
drains not influenced by Harlan showed only a slight increase in coliform
STUDY RESULTS
-------
124
5
5
5
1O
5
1O
5
5
l
l
11
P
i
l
l
l
i
n
0
1
i
m
,
n
i
i
i
i
i
n i
i
L
A
n \MI sn
| | |
c
1 HI n
D
^ M 1 1 ^
I
n^ra n
E
ii n^ m
lifll In
1 00 1 00 - 1 ,000 1 ,000 - 1 0,000 1 0,000 - 50,000 50,000 - 500,000 500,000 - greater
,., RflPtprifll rnnnt
Cl Total Colifor
Fecal Col iform
Fecal Streptococcus
Figure 16. Frequencies of Levels of Contamination
STUDY RESULTS
-------
TABLE 40
Coliform Counts at High Stream Discharge
125
Total Fecal Fecal
coliform Coliform Streptococcus
count/100 ml water
Range 600-92,000 0-36,000 0-10,000
Median 18,000 3,350 3,500
Mean 25,960 4,865 4,220
contamination over low stream discharge levels (see Table 41).
TABLE 41
Coliform Counts at High Stream Discharge
Stations down-
stream of Harlan
Stations not
influenced by
Harlan
Total
low
discharge
2,850
1,800
coliform
count per ]
high
discharge
32,000
2,500
Fecal
On ml Tiia-HoK-
low
discharge
600
500
coliform
high
discharge
5,300
700
These observations led us to two conclusions:
1) Storm runoff from agricultural land in the Black Creek watershed
was not substantially different from low stream discharge runoff in the
levels of total and fecal coliform contamination, and,
2) Highly organic, contaminated material deposited at septic tank out-
falls was scoured and flushed downstream during the storm event.
The increase in fecal streptococcus contamination during high flow was
observed at all stations on the Dreisbach Drain. A grassed waterway and an
open ditch in this section of the watershed carried between 5 and 10
thousand fecal streptococci per 100 ml of water. A slightly lower level of
contamination (3,000-9,000) was maintained along the lower Dreisbach drain,
the Richelderfer Drain and Black Creek. The remaining tributary drains had
from 0 to 3,500 fecal streptococci per 100 ml of water. Thus the Upper
Dreisbach Drain agricultural area and the town of Harlan were the major
identifiable areas of fecal streptococcus contamination at high stream
discharge. The dominant type of livestock handling in the Dreisbach
subwatershed is open grazing in pastures and confinement in small barnlots.
In these areas the contamination was over an order of magnitude greater
than the levels of pollution observed at low stream discharge. - The other
monitoring stations in the Black Creek watershed showed the storm runoff to
be only slightly more contaminated than low stream discharge runoff. It is
concluded that some livestock operations were responsible for substantial
fecal pollution of storm runoff in the Black Creek watershed.
STUDY RESULTS
-------
L26
4.4.3 Bacterial Counts,Maumee,Warm Ditch
Surface water samples were collected from the Maumee River and the
Warm Ditch to facilitate a comparison of the coliform contamination in
Black Creek with the coliform contamination in its receiving body of water
(Maumee River) and a similar creek adjacent to the Black Creek watershed.
The Warm Ditch lacks any concentrated urban area and is comparable to the
tributary drains of Black Creek not influenced by the town of Harlan (sam-
pling Group E). As expected, the counts were similar to those reported
from stations in Group E. The trend for increasing fecal streptococcus
counts at high stream discharge was also observed on the Warm Ditch.
Counts observed in the Maumee River were quite wide-ranging, probably
being a function of discharge, as the highest counts were recorded in the
spring. Fecal coliforms were generally fairly low (200-500 counts per 100
ml), but at high discharge, the counts increased an order of magnitude
(400-15,000). Fecal streptococci were detected only once at the Maumee
River station during a period of high stream discharge (March 28,1977).
Compared to the waters of the Black Creek drainage, the Maumee River had
approximately the same concentration of total coliforms but lower concen-
trations of fecal coliforms and fecal streptococcus (excluding periods of
high stream flow).
4.4.4 Bacterial Counts in Tile Drainage Water
Twenty tile drainage systems were chosen for sampling on the basis of
the soil types being drained. The data record is too incomplete to deter-
mine if soil type was a factor affecting the level of fecal pollution from
tile systems. Of 120 attempted sample collections, 55 could not be made
because the tile systems were not flowing. In addition, six of the tile
lines were found to be connected to septic tank systems. As a result, con-
clusions about fecal pollution from tile systems are rather limited and
have been based on the following observations.
The tile drainage stations are broken down into the septically pollut-
ed tiles and the non-septically polluted tiles in Table 42. A further
division is made between the low flow and high flow sampling dates.
Predictably, the septically-polluted tiles had very high average values
during low flow when there was very little dilution of septic waste by na-
tural drainage water. At high flows these same tiles carried far lower
concentrations of all three bacteria because of the dilution by sub-surface
drainage water. The non-septically polluted tile systems had fairly low
levels of coliform contamination that remained unaffected by discharge
rates. The total coliform counts were very similar at high and low
discharge and a small increase in fecal coliform concentrations at high
flow resulted from two tile stations with surface water inlets having
higher values.
The increase in fecal streptococcus contamination of non-septically
polluted tile systems at high flow was a general occurrence recorded for
the majority of the tiles sampled. At low flow, only 1 of 14 stations had
a fecal streptococcus count above 100, while at high flow 12 of 14 stations
had counts over 100. However, the magnitude of increase was not generally
as great as the average figures in Table 42 indicate. At high flow, 10 of
the 14 stations had fecal streptococcus counts between 100 and 500, while
two tiles with surface water inlets had counts an order of magnitude
STUDY RESULTS
-------
127
TABLE 42
Coliform Counts for 6 Septically Polluted and 14 Non-septically Polluted Outlets
Low
Flow
mean
Septically Polluted
Non-septically Polluted
TC
FC
FS
TC
FC
FS
329,
60,
29,
2,
500
683
662
327
60
12
High
n
21
21
21
24
25
25
TC
FC
FS
TC
FC
FS
9
2
5
2
2
mean
,740
,340
,880
,286
430
,150
Flow
n
5
5
5
14
14
14
greater. Non-septically polluted tile drainage systems were an identifi-
able source of fecal pollution at high stream discharge, although the de-
gree of contamination was not great.
4.4.5 Fecal Coliform/Fecal Strep Ratios
An indication of the source of fecal pollution is given by the fecal
coliform/fecal streptococcus ratio (FC/FS). A ratio greater than 4 indi-
cates human sources of pollution, a ratio less than 1 indicates an animal
source and a ratio between 1 and 4 indicates combined human and animal fe-
cal pollution. In the Black Creek watershed, at low flow, the dominant
source of fecal contamination was the human septic waste effluent, but
there was a significant (p<.01) shift from human to mixed human and animal
sources of pollution when discharge increased (see Table 43). This obser-
vation further supports the conclusion that livestock operations had a sub-
stantial impact on the fecal contamination of storm water runoff. Stations
in the upper Dreisbach agricultural area exhibited the greatest degree of
fecal contamination from livestock, while stations in other areas showed
either livestock sources or mixed human and animal sources.
TABLE 43
The Number of FC/FS Observed and Expected FC/FS Determinants
Human sources
Mixed sources
Livestock sources
FC/FS > 4
FC/FS 1-4
FC/FS < 1
x2=24.44,
Low flow
45(36.7)
14(15.3)
6(13)
p<.01
High flow
3(11.3)
6(4.7)
11(4)
4.4.6 Biochemical Oxygen Demand
STUDY RESULTS
-------
128
4.4.6.1 BOD
Samples were collected for biochemical oxygen demand (BOD) analyses
during a major storm event on June 30, 1977 (7.1 cm rainfall). Grab sam-
ples were collected in a 2 liter polyethylene containers and refrigerated
until laboratory set-up was initiated the next day. The grab sample loca-
tions are noted in Table 44. In addition, composite samples were made from
the water collected by automated pump samplers at sites 2 and 6. At stream
stages above one foot, the automated pump samplers collected a water sample
every 30 minutes during the course of the storm event. Composite samples
for BOD analysis were made at the end of the event by combining 50 ml of
water from each sample taken by the pump sampler. Laboratory analysis for
BOD was done by Pollution Control System, Inc. (Laotto, Indiana) following
the procedures of Standard Methods (American Public Health Association,
1971).
TABLE 44
BOD During Ascending Flow of Storm Event
Station
Grab samples
296
297
127
128
165
310
Composite samples
2
6
BOD (mg/1)
12.0
6.6
14.0
16.0
16.0
480.0
6.3
9.3
Predominant Feature of Watershed
Small watershed, Amish farming
Small watershed- conventional farming
Dreisbach Drain, urban buildup
Richelderfer Drain, urban buildup
Wertz Drain, Amish farming
Tile Drain, confined feeding operation
Smith-Fry Drain, conventional farming
no urban buildup
Dreisbach Drain, Amish farming, urban buildup
4.4.6.2 RESULTS
All grab samples were taken during the ascending climb of the storm
hydrograph. Results are listed in Table 44. Two samples were collected
from adjacent small watersheds (60 acres) in the rolling uplands. The
watershed with Amish farming practices had twice the BOD concentration in
runoff water as the area under a conventional farming operation. It ap-
pears the unconfined livestock in the Amish area substantially increased
the amount of organic matter in surface runoff.
Grab sample 310 was taken from a large tile outlet known to have sur-
face inlets in the vicinity of a large confined livestock feeding opera-
tion. Although this operation is equipped with properly designed animal
waste holding facilities, surface runoff from the barnyard area does reach
the tile drain and eventually Black Creek. The BOD at station 310 was
exceedingly high, 480 mg/1, equivalent in strength to raw sewage (Hynes,
1960). The tile outlet was sampled during the initial phase of the storm
and surface runoff was just beginning so the volume of discharge was small.
It is doubtful that BOD concentrations remained this high during peak storm
STUDY RESULTS
-------
129
runoff because dilution would be greater. In terms of total BOD loading,
such a confined feeding operation cannot be considered a major source
within the watershed, but it is the type of source that raises BOD concen-
trations in stream above the concentrations found in runoff from cropland.
Also, highly concentrated organic matter delivered to stream in this manner
could be damaging to the localized section of stream at the tile outlet if
rainfall and runoff are not sufficient to dilute and flush away the organic
matter.
The composite samples collected from the Smith-Fry drain (site 2) and
the Dreisbach drain (site 6) also illustrate the effect of urban build-up
and Amish farming practices on BOD concentrations. The Smith-Fry watershed
lacks any urban influence and is farmed predominantly by conventional
methods while the Dreisbach watershed contains the town of Harlan and a
large number of Amish farmers. Composite BOD concentrations for site 2 and
6 were 6.3 and 9.3 mg/1 respectively. Precise BOD loading data has not
been calculated yet, but to arrive at a first approximation, composite con-
centrations were multiplied by the average daily discharge computed for
June 30, 1977. Estimated loadings of BOD for the storm event were 320 kg
and 220 kg at sites 2 and 6 respectively. More rainfall and runoff in the
Smith-Fry watershed accounted for the greater loadings at site 2. Averag-
ing data for both watersheds the rate of BOD export from the Black Creek
watershed for this storm was approximately one-third kilogram per hectare.
Grab samples were taken on three major drains of Black Creek prior to
peak flows on June 30. BOD concentrations were similar in all three sam-
ples ranging from 14 to 16 mg/1. Two of the samples were taken downstream
from the town of Harlan and the other from a predominantly Amish farming
area. The urban area with its septic effluent and the Amish area with a
large number of unconfined livestock appear to be the factors that created
high BOD concentrations in stream flows (14-16 mg/1) compared to runoff
from conventional cropland (6.6 mg/1).
4.4.7 Fish Kill Caused by Organic Pollution
On 28 September 1977 several thousand gallons of manure slurry were
accidentally discharged into Black Creek when an animal waste holding
lagoon was emptied directly onto adjoining cropland. The slurry entered a
subsurface tile network through broken tile lines and/or surface inlets and
was delivered to the stream with very little dilution (Table 45). The im-
pact at the outfall was devastating and low stream flows were inadequate to
dilute the pollutant to non-toxic levels. The material moved downstream as
a slug which could be visually detected. Three downstream samples had very
high BOD (130-300 mg/1) even prior to the arrival of the main slug of pol-
lutant. Ammonia N concentrations were also greatly elevated (Table 45).
Fish mortality was severe in the entire 9 kilometers of stream below
the spill. Mortality probably resulted from low oxygen levels and/or an
ammonia toxicity. Accidental or intentional discharge of organic pollu-
tants from animal waste holding facilities can create gross organic pollu-
tion in streams when discharge is low and potential damage is the greatest.
Wide spread organic pollution from other sources (septic tanks and uncon-
fined livestock) is greatest at high stream discharge when dilution reduces
the impact on aquatic life.
STUDY RESULTS
-------
130
TABLE 45
BOD and Ammonium N Concentrations During Fish Kill
Sample
Location
Upstream
Source (tile line)
Downstream 100 m
Downstream 580 m
Downstream 1720 m
Downstream 2440 m
BOD
mg/1
2.1
28,000
7,200
130
220
300
Ammonia N
mg/1
1.2
2,400
600
11
18
20
4.4.8 Pesticides and Heavy Metals
Several spot checks were made for the presence of pesticide residue
and for heavy metals in sediment, in fish and in water samples according to
the procedures outlined in Section 0.0.2. Results of this analysis are
presented in Table 46 and are summarized in Table 47. Sampling was done
under low-flow conditions and may not reflect the situation during storm
events. Overall, a low level of contamination was found.
TABLE 46
Pesticide and Heavy Metal Concentrations
Location Dieldrin
Sediment Samples
401 ND
402 ND
403 ND
404 ND
405 ND
406 ND
407 ND
408 ND
409 ND
410 ND
411 ND
412 ND
413 ND
414 ND
DDE
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
PCB's
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
2-4-5-T Cd
0.77
0.66
0.64
0.94
1.08
ND 1.51
2.18
ND
.34
.34
.37
ND 1.15
1.97
ND 2.65
Hg
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
.003
ND
.005
Zn
2.69
2.24
2.28
1.85
1.55
1.66
2.90
2.46
ND
1.55
1.40
1.56
4.46
4.47
Fish Samples
415
416
417
418
419
420
421
422
0.012
0.013
0.027
0.031
0.031
0.047
0.022
0.013
0.012
0.017
0.021
0.027
0.031
0.021
0.014
0.007
0.117
0.070
0.265
0.140
0.182
0.142
ND
ND
ND
ND
ND
0.18
0.16
ND ND
ND
ND
0.092
0.042
0.008
0.009
0.036
0.021
0.051
0.077
14.4
20.2
21.8
19.1
19.2
12.2
16.6
18.2
STUDY RESULTS
-------
131
Table 46 (continued)
423
424
425
426
427
428
429
430
431
432
Water
433
434
435
436
437
438
439
440
0
0
0
0
.086
.010
.011
.004
0.009
0
0
0
0
0
.014
.021
.015
.017
.009
0
0
0
0
0
0
0
0
0
0
.014
.015
.010
.009
.023
.008
.012
.012
.018
.010
0.
0.
0.
0.
0.
0.
0.
086
108
111
ND
ND
ND
173
211
121
113
_
-
-
ND
-
-
-
-
-
ND
ND
ND
ND
ND
0.15
ND
ND
ND
ND
ND
0
0
0
0
0
0
0
0
0
ND
.023
.041
.050
.047
.030
.076
.102
.174
.078
13.3
24.2
17.5
23.4 i
23.9
16.4
15.7
22.6 :
17.3
22.3
Samples
ND
ND
ND
-
ND
ND
ND
ND
ND
ND
ND
-
ND
ND
ND
ND
0.
0.
0004
ND
0002
-
ND
ND
ND
ND
__._
-
-
0.003
-
ND
-
-
0.001
0.113
0.099
0.052
-
0.037
0.029
0.026
0.038
0
0
0
0
.0002
.0002
.0002
-
.0002
ND
ND
ND
0.029
0.022
0.021
-
0.020
0.017
0.010 1
0.014
j
Note: Concentrations expressed in ppm
fish /g wet weight
water /g ml
ND = not detected
- = not determined
TABLE 47
Summary of Pesticide and Heavy Metals Detected and Not Detected
Substances Not Substances
Medium
Water
Sedi-
ment
Fish
Detected
Atrazine
Alachlor
Carbofuran
Malathion
Dieldrin
DDE
Atrazine
Alachlor
Carbofuran
Malathion
Dieldrin
DDE
PCB's
2, 4, 5,-T
Atrazine
Alachlor
Carbofuran
Malathion
2, 4, 5,^T
Detected
2, 4, 5,-T
PCB's
Cadmium
Mercury
Zinc
Cadmium
Mercury
Zinc
Dieldrin
DDE
PCB's
Cadmium
Mercury
Zinc
Number
Samples
2
7
7
7.
-»
/
14
14
14
18
18
18
18
18
18
Number Max.
Min.
Mean
Occur ences Concentrations
2
2
7
4
7
12
2
13
18
18
13
3
17
18
0.003
0.0004
0.113
0.0002
0.029
2.65
0.005
4.47
0.086 0
0.031 0
0.265 0
0.18 0
0.174 0
24.2 12
0.001
0.0002
0.026
0.0002
0.010
0.34
0.003
1.55
.004
.007
.070
.15
.008
.2
0.002
0.0003
0.056
0.0002
0.019
1.19
0.004
2.39
0.022
0.016
0.140
0.16
0.053
18.8
STUDY RESULTS
-------
132
Trace levels of 2, 4, 5,-T in water was found to persist after ditch
bank maintenance with herbicide sprays. Higher concentrations may exist
immediately after the maintenance procedure.
Low levels of organochlorines were detected in fish, but not in sedi-
ment. This suggests that there is currently no significant loading of
these insecticides. Fish maintain a small amount of pesticide residue, but
this is well below acceptable levels for human consumption as determined by
the Food and Drug Administration. Effects on the fish life cycle are con-
sidered minimal. The level of PCB's detected in fish tissue were low, and
are not an immediate cause for concern to humans or to fish life. However,
PCB's were also found in water which presents evidence of continued loading
to the aquatic system.
Heavy metals were found in low concentrations and are considered to be
not much different than background levels. The concentrations are not a
concern to human health.
4.5 FISH STUDIES
4.5.1 Fish Sampling
Sample stations (see Section 3.3,1 were chosen to include rif-
fles and pools and were paced off to 100 m lengths. In the Wertz Woods a
set of three pools with combined length of approximately 30 m was sampled.
Minnow seines (1.2 x 4.5 m) with 6 mm mesh were used in most sampling. A
seine was placed at the lower end of a sample area and at least two seine
sweeps were made downstream through the area. For more complex areas such
as Wertz Woods stations were broken into smaller segments and seined until
capture rates declined to near zero. Typically this required three to six
seine hauls in the most complex streams sampled.
Fish samples were preserved in 20% formalin solution until a synoptic
collection was developed. Thereafter, most were released immediately after
field tabulation. Identification of fishes followed classification of
Trautman (1957) and Nelson and Gerking (1968).
4.5.2 Habitat Structure
Stream habitat structure was measured in June and September 1975, and
March 1976 for three dimensions: depth, bottom type, and current. Four
depths, nine bottom substrates, and five current categories were recognized
(Table 48). Stream depth categories were chosen as representative of habi-
tats found in the small streams sampled: 0-5 cm corresponded to shallow
edges and riffles, 5-20 cm to riffles and shallow pools, 20-50 cm to pools,
and greater than 50 cm to deep pools. Bottom types were categorized into
physical and biotic structures. Among the physical forms, categories 1-5
corresponded to alluvial material of increasing size from silt to rocks
with category 6 as clay parent material (clay pan). Biotic categories (7
and 8) included vegetation (aquatic plants and filamentous algae) and
litter (leaves, twigs and branches). A miscellaneous category, 9, was
reserved for unusual items such as bedrock slabs or large tree trunks.
Currents were correlated with specific water velocities and were gauged by
observing the movement of water about a measuring pole.
Point samples were taken in a regular fashingin each study area. Be-
ginning 10-20 cm from the left bank, points were taken at 1 m intervals
across the stream. For very narrow stream (maximum 1 m wide) such as Limbo
Creek and Wertz Woods at low flow, 0.33 m intervals were used. Repeated
STUDY RESULTS
-------
TABLE 48
Categories for Habitat Analysis
D imens ion
Depth
Range (cm)
Descr i pt i on
1
0-5
very
Category "lumber and Description
2
5-20
shal low
3
20-50
node rate
it 5 6789
>50
deep
s h a 1 1 ow
Current
Bottoms
Flow Velocity
(m/sec )
Descr i pt ion
Diameter (mm)
Descr i pt ion
<.05
very s
<.05
silt
.05-. 2
low slow
.05-2
sand
.2-.lt
moderate
2-10
gravel
.it-1.0 >1.0
fast torrent
pebble rock clay-pan vegetation litter misc.
-------
134
sets of points were taken across the stream at 5 m intervals moving
upstream. Sample data showed that a minimum of 90 to 160 points were re-
quired at structurally complex sites (Wertz Woods and Indian Creek) and 60
to 80 points for simple areas (Black Creek) for adequate measurement of ha-
bitat diversity. Habitat sampling was generally conducted on the same day
as fish sampling with the exception of September 1975 when Black Creek ha-
bitat samples were taken one week after fish sampling. No changes in
weather conditions occurred during this interval.
Habitat diversity was calculated using the Shannon-Weiner equation.
Diversities were calculated for each habitat dimension alone and then in
combinations of depths and bottoms (36 categories), currents and bottoms
(40 categories), and finally depths, bottoms, and currents (180
categories). Only certain combinations of these dimensions existed in
these streams. For instance, we found no rock-bottom torrents greater than
50 cm deep, or silt-bottom riffles.
Combinations of habitat dimensions were tried to determine which di-
mension or combinations were better predictors of fish species diversity.
The effect of combining dimensions increased the number of categories
geometrically and thereby increased the habitat diversity index additively
since this index is an exponent.
Mean habitat indexes were calculated for each dimension by averaging
the number of values for each category. This allowed each stream site to
be characterized as shallow, sandy, slow, etc. In the bottom-type dimen-
sion only the alluvial categories (1-5) represent continuous variation, so
calculation of mean habitat indexes was restricted to those categories.
4.5.3 Species Composition and Distributions
A total of 35 species of fish representing 24 genera and 11 families
have been collected in the Black Creek watershed basin (Table 49). This
represents about 21 percent of the species and 41 percent of the families
known from the State of Indiana (Nelson and Gerking 1968). A recent IJC
report (1975) on the fishes of the Maumee River lists 91 species from the
Maumee Basin in a survey by the Ohio Department of Conservation and Traut-
man (1957). Therefore about 35 percent of the fishes known from the basin
have been collected in Black Creek. It is significant that most species
found in Black Creek show either stable population trends (32 species) or
are decreasing in the Maumee Basin (3 species). The three declining
species are Northern Pike, Redear Sunfish, and Creek Chubsucker. The sunf-
ish in Black Creek escaped from a farm pond and the Pike were captured only
in the spring of 1974. The Chubsucker was common in Wertz Woods but it has
disappeared following siltation after bank stabilization in upstream areas.
About a dozen species collected in Black Creek seem to be declining in
abundance although they are not experiencing declining populations in the
Maumee Basin according to the IJC report. Other species known from Black
Creek are the most tolerant forms to be expected in first, second, and
third order streams in an agricultural watershed. The minnow family
(Cyprinidae) is represented by the largest number of species (12) with the
sunfish family (Centrarchidae) being represented by the second largest
number of species (8). The minnows make up the largest number of individu-
als in the basin but in terms of biomass the sucker family (Catastomidae)
is the dominant group, especially during the late spring and early summer
migration period.
STUDY RESULTS
-------
135
TABLE 49
Fish Species Collected in Black Creek
Scientific (Common) Name Status in the Black Creek Basin
Distribution
20 18 20 Wertz Wd
Aplodinotus grunniens
(Freshwater Drum)
Umbra limi
(Central Mudminnow
Semotilus atromaculatus
(Creek Chub)
Pimephales promelas
(Fathead Minnow)
P. notatus
(Bluntnose Minnow)
Ericymba buccata
(Silverjaw Minnow)
Campostoma anomalum
(Stoneroller)
Notropis cornutus
(Common Shiner)
N. stramineus
(Sand Shiner)
N. spilopterus
(Spotfin Shiner)
N. umbratilis
(Redfin Shiner)
Phenacobius mirabilis
(Suckermouth Minnow)
Notemigonus chrysoleucos
(Golden Shiner)
Cyprinus carpio
(Carp)
Very rare fall migrant
(One specimen)
One specimen
Distributed throughout the area; C V V V
some movements, especially influx
in spring.
Most abundant in upstream areas. V V V V
Common throughout the basin but C V V V
most abundant in small streams
below P. promelas areas.
Common throughout the basin V V V V
especially in silty to sandy area.
Uncommon but distributed U U C
throughout.
Abundant to common throughout; R U C V
some migration in spring.
Generally restricted to areas in C V
the main channel of Black Creek;
move upstream to some extent in
late spring and early summer;
often found in same areas as
E. buccata.
Common below station 12; uncommon C C V R
in rest of basin except in fall
when large numbers invade.
Rare throughout the basin, except C U C R
in early summer.
Rare in Black Creek below station R R
15.
Uncommon below station 12; resident R C
population in Wertz Woods.
Common and often large in main R U
Black Creek channel-
STUDY RESULTS
-------
136
Erimyzon oblongus
(Creek Chubsucker)
Catastomus commersoni
(White Sucker)
Carpiodes cyprinus
(Quillback Carpsucker)
Moxostoma sp.
(Redhorse)
Fundulus notatus
(Black-striped Topminnow)
Percina maculata
(Black-sided Darter)
Etheostoma nigrum
(Johnny Darter)
E. spectabile
(Orangethroat Darter)
Lepomis cyannellus
(Bluegill)
L. macrochirus
(Green Sunfish)
L. microlophus
(Redear Sunfish)
L. humilis
(Orange-spotted Sunfish)
L. gibbosus
(Pumpkinseed)
Pomoxis nigromaculatus
(Black Crappie
P. annularis
(White Crappie)
Micropterus salmoides
(Largemouth Bass)
Labidesthes sicculus
(Brook Silverside)
Ictalurus natalis
(Yellow Bullhead)
Uncommon in basin except common in C
Wertz Woods.
Abundant in spring migration (esp. R R U V
large individuals); permanent
residents throughout Wertz Woods.
Common in Black Creek near Maumee U
River; sporadic in rest of area,
most abundant in late summer and
fall.
One specimen
Common to abundant, esp. in areas C C R
with dense growth of aquatic plants
and in late summer and early fall.
One specimen. R
Common, esp. in rocky areas. R U
Uncommon u R
Common throughout R R C U
Common throughout V U V V
Collected occasionally C U C R
Rare
Rare R
Rare
Rare summer iramatures R
Young sporadic in several areas R R R R
Also widely distributed in 1977.
One individual caught in Black
Creek near Maumee River.
Small individuals sporadic U U R
throughout basin; some large
residents in Wertz Woods
STUDY RESULTS
-------
I. melas and below station 12. R
(Black Bullhead) 137
Esox lucius Several large individuals seen or
(Northern Pike) captured in spring 1974.
Dorosoma cepidianum Common near Maumee River; many R U
migrate upstream in late summer
and fall.
Distribution at first (Station 20), second (18) , and third (12) order
stations of the watershed plus Wertz Woods using the following abundance
classes:
V Very Common > 85% of samples
C Common 50 to 80% of samples
U Uncommon 20 to 50% of sampels
R Rare < 20% of samples
Predictably, the similar physiognomies and geographic proximity of
Black Creek and Wann Creek resulted in nearly identical fish faunas. How-
ever, the fauna at the sample station on Wann Creek was exceptionally rich;
17 species were common (caught more than 50 percent of the samples) while
at Black Creek's station 18 (of similar size and stream order), only 7
species were common. Overall, one station in the Wann Creek drainage had
more common species than any area on Black Creek and more species were cap-
tured per sampling period. The first order stream in the Wertz Woods was
also surprisingly rich in comparison the rest of Black Creek; 11 species
were common in Wertz Woods while at station 20 in Black Creek only 8
species were common. Only station 12 had more common species (12).
Overall, the number of species captured per sample in Wertz Woods was simi-
lar to.station 12 but more consistent. The Wertz Woods yielded the highest
fish species diversity for the Black Creek drainage. Typically each sta-
tion on Black Creek was dominated by a few species, resulting in low fish
species diversity, while Wertz Woods and Wann Creek had consistently more
species and more equitable abundance distributions even though Wertz Woods
was first order and Station 12 was a third-order stream.
4.5.4 Seasonal Changes in Fish Communities
The average number of fishes per station peaks in the Black Creek
watershed in early spring but the magnitude of peaks varies from year to
year depending on the activities of man and rainfall amounts (Karr and Gor-
man 1975). Spring peaks (March-April) are associated with spawning migra-
tions of fishes from downstream areas, including the Maumee River. Summer
declines in Black Creek are associated with disappearance of larger fishes
due to low water levels. An increased capture rate in late summer and fall
results from increased catchability of young of the year and some fall mi-
grations. Late winter sampling show that densities decline as a result of
some downstream migration and natural mortality during the winter. Larger
fishes were relatively uncommon in upstream areas except during the spring
migration period of March and April. Communities in headwaters were dom-
inated by small minnow species such as Pimephales spp., Ericymba buccata,
and juveniles of Semotilus atromaculatus. Fishes at downstream stations
tended to be larger and less numerous. Overall, upstream areas had higher
densities of smaller fishes such that headwater and downstream areas sup-
port similar biomasses per meter of stream.
STUDY RESULTS
-------
138
4.5.5 Habitat,Fish Community Diversity
There is a significant relationship between habitat diversity and fish
species diversity when Black Creek data are combined with data from two
other stream systems (Gorman and Karr 1978). Significantly the relative
importance of the three dimensions bottom, current, and depth varies
among stream systems. Neither bottom nor depth diversities are good pred-
ictors of fish species diversity of Black Creek because of the uniform na-
ture of the habitat components due to ditching activities. Algal blooms
characteristic of dry periods with low flow conditions are also correlated
with exceptional low fish diversities. These blooms are associated with
stream reaches lacking shading vegetation. Further, water temperatures
tend to be very high in such areas: 28 degrees C as compared to 19 degrees
C in the same stream where it is shaded in a woodlot (Karr and Gorman
1975).
4.5.6 Effects of Channel Modification
The effects of channel and bank modifications can most easily be
demonstrated by examining changes in the fish communities at station 15 on
Black Creek.
The first 50-m sample from station 15 was made on 12 April 1974 when
210 fish were captured. Fifty-nine of these fish had average total lengths
of over 150 mm indicating the high biomass in that section of stream.
Larger individuals were creek chubs, white suckers, common shiner, and
green sunfish. Mean weight per individual was 25.85 gms and biomass densi-
ty was 108.6 g per meter of stream. After bottom dipping and bank modifi-
cations in late spring 1974 fish densities declined and have remained low.
Table 50 summarizes the fish community data for April.
TABLE 50
Fish Community Density During April
April 1974
April 1975
April 1976
Number of
fish per meter
4.2
.34
.05
Mean Weight (g)
per individual
25.85
2.10
<0.5
Biomass (g)
per meter
108.6
1.0
Fish densities in 1977 were also quite low at station 15 although they
were high in nearby upstream areas with little or no channel modification.
At the time of the April 1974 samples the stream showed the effects of
earlier channel modification but erosion and deposition areas with some
segregation of particle sizes was evident. The mean depth of the stream
was near 50 on while following channel modification in 1974 water depth
averaged only about 10 on. The decline in fish populations at this station
seems primarily attributable to channel modifications reducing depth and
bottom diversity. Similar patterns of change in fish densities have oc-
STUDY RESULTS
-------
139
curred at other Black Creek sample areas.
These changes in streams modified by man's activities are in sharp
contrast to the more damped oscillations of fish abundances in Wann Creek
and in several pools associated with the Wertz Woods along the Wertz Drain.
In Wertz Woods as a number of species characteristic of downstream areas
persisted throughout the year. Fish marked with a cold-branding technique
were commonly recaptured in Wertz Woods while fish marked in other areas in
the Black Creek watershed were rarely recaptured. It is not clear from our
data whether fish outside Wertz Woods disappeared due to high mortality or
leaving the watershed; both factors are probably very important.
Even relatively protected sections of stream like that in Wertz Wood
may be subjected to high mortality. The very dry period in late 1976
resulted in almost total mortality throughout upstream areas in Black
Creek. Because of deep pools fish in Wertz Woods persisted later than fish
in other areas. However, most were killed before the onset of winter and
the few remaining individuals were killed by the severe cold weather of
early 1977.
Upstream migrations in 1977 were similar to those in earlier years and
resulted in recolonization of the whole watershed. However, fish have not
been able to reestablish resident populations in Wertz Woods. As demon-
strated in 1975 (Karr and Gorman 1975) the meandering pool and riffle to-
pography of Wertz Drain in Wertz Wood resulted in deposition of about 28
percent of sediment carried by Wertz Drain. Although we do not have direct
evidence, indirect evidence suggests that declining sediment loads are due
to physical processes (Unit Stream Power of hydrology) rather than action
of biotic systems (Karr and Schlosser 1977).
Construction on the Wertz Branch upstream from Wertz Woods was instru-
mental in halting the recolonization of Wertz Woods in 1977. The construc-
tion activity itself and subsequent erosion of unstable bank and channel in
Wertz Branch yielded large amounts of sediment. Grab sampling of several
runoff events revealed that sediment concentrations from the Wertz Branch
were 4-8 times greater than from undisturbed channels. As this sediment
moved downstream it was deposited in downstream areas (especially Wertz
Woods). This deposition has reduced the diversity of depth, bottom and
current characteristics in Wertz Woods and prevented reestablishment of
resident fish populations. Monitoring is continuing to determine how long
this problem will persist.
It should be clear from these results and the more detailed presenta-
tions of Karr and Schlosser (1977) and Gorman and Karr (1978) that attempts
to improve water resources must involve broad based, multi-purpose program.
Management plans which produce clean water in the absence of suitable habi-
tat diversity will not improve the biota of waterways. Low quality stream
biotas may result from low water quality, poor habitat diversity, seasonal
low flows and other factors. All of these problems must be addressed be-
fore incremental improvement can be expected in a wide range of water
resource characteristics.
STUDY RESULTS
-------
140
4.5.7 Stream Disturbance and Fish Communities.
Stream environments are naturally unstable; i_.e., due to seasonal
variation in rainfall, flow volumes do not remain constant in most streams.
However, flow volumes are relatively predictable in most natural streams.
Environmental extremes, such as high water and droughts, occur predictably
from season to season and year to year. Fishes have evolved physiologies
and behaviors to minimize mortality during these extremes, ( e_.g_. migra-
tion to more suitable habitat, moving into pools during dry periods, etc.).
Structurally diverse natural streams typically have a great deal of
buffering capacity: meanders tend to moderate the effects of floods, pools
offer excellent refuges for fishes during dry spells, and tree shade de-
creases heat loads and minimizes the oxygen-robbing effects of decomposing
and extensive algal blooms.
The reliability of stream environments is reduced by man's modifica-
tions of these systems to suit his needs. Ditching increases stream gra-
dients by meander removal and channel shortening. Also, bottoms are
dredged to create a uniform, pool-less, unstable substrate. These attempts
to increase drainage efficiency result in little buffering from floods and
droughts and increases their severity.
Other stream modification procedures which enhances instability is the
removal of shade-producing vegetation, and the sloping of banks. This max-
imizes solar heating of the water and increases problems from algal blooms.
Also, massive deforestation of drainage basins enhances the impact of
floods and causes silt pollution problems via increased soil erosion (see
review by Karr and Schlosser 1977). The discharge of sewage effluents may
also be detrimental to fish, but in areas below the septic zone, the more
constant flow of water may help to stabilize the downstream fish popula-
tions during drought periods (Karr and Dudley 1976).
Some indication of the effect of these factor can be obtained by com-
paring studies in Black Creek with work from other stream areas. Most
stream reaches in Black Creek have been modified significantly in the past
few years. The only significant exception is Wertz Woods. A sample sta-
tion on Wann Creek was modified about a decade ago but has developed some
pool and riffle characteristics along with a diverse vascular plant commun-
ity in and near the stream channel. Finally, Indian Creek in Tippecanoe
County west of Lafayette, Ind. has also been studied. Indian Creek is an
agricultural watershed but most of the stream has not been ditched. Thus,
natural stream topography has been maintained and most of the channel is
shaded by trees. The degree of buffering present among our study streams
was evident in the structure and stability of the resident fish communities
(Table 51). Community structure in Black Creek was simple in each area and
unstable. Wann Creek usually had a more diverse community but seasonal
stability was low. Indian Creek and Wertz Woods, however, had both diverse
and relatively stable communities.
Even though Wann Creek was subject to the same unstabilizing influ-
ences as Black Creek it showed remarkable recovery. The stream's habitat
structure was more diverse, with greater pool formation and dense bulrush
thickets which stabilized the stream channel and provided partial shading.
This stream structure supported a more diverse community but not a more
stable one. The unstabilizing influences of massive stream and watershed
STUDY RESULTS
-------
141
TABLE 51
Percent Change - Number Species 4 Areas
Stream
Black Creek
Wann Creek
Wertz Woods
Indian Creek
Number of
Fish Species
5(1)*
14(2)
9(1)
10(1)
Number of
Stations
5(1,2,3)*
1(2)
KD
2(1,2)
Mean Percent
June-Sept.
68
70
40
33
Change
1975
* Numbers in parenthesis indicate stream order(s).
modification tend to strip'the stream ecosystem of some equilibrium capa-
bility (homeostatic mechanism). Thus, Black Creek and Wann Creek biotic
communities and habitats changed dramatically seasonally.
In contrast, Indian Creek and Wertz Woods had remarkable complexity
and stabilized habitat structure allowed the stream some degree of homeos-
tatic equilibrium and was reflected in the stability of the fish community.
From this study of an assortment of streams it is evident that natural
processes and structures enhance the reliability of a basically unstable
system. Habitat complexity increases community diversity and environmental
stability appears to. control community stability. In the succession of a
ditched stream after disturbance, the habitat and community diversity re-
cover first while stability requires a longer period or may never be
achieved as long as overlying unstabilizing influences persist on the
watershed.
Since large-scale watershed modifications in the U.S. have continued
for perhaps two centuries, it is now difficult to evaluate the extent to
which stream communities have been altered. Larimore and Smith (1963)
found considerable changes in the fish fauna in Champaign, County, Illinois
in the past 60 years. The changes they found were associated with large-
scale modification of watersheds. Recent deterioration of the Great Lakes
fisheries has been tied to over-exploitation of fish populations and to
massive deforestation and modification of the watersheds over the last cen-
tury (Smith 1972).
References
Gorman, O.T. and J.R. Karr. 1978. Habitat Structure and Stream Fish Com-
munities. Ecology. In press.
Karr, J.R. and O.T. Gorman. 1975. Effects of Land Treatment on the Aquat-
ic Environment. "Non-Point Source Pollution Seminar" Section 108a Demons-
tration Projects. Technical Report EPA-906/9-75-007.
Dudley, D.R. 1976. Determinants of water quality in the Black Creek
watershed. "Best Management Practices for the Control of Non-Point
Sources." Section 108a Demonstration Projects. Technical Report EPA. 14p.
STUDY RESULTS
-------
142
Schlosser, I.J. 1977. Impact of nearstream vegetation and stream morphol-
ogy on water quality and stream biota. Ecological Research Series U.S. En-
vironmental Protection Agency, Athens, GA. 90pp. (EPA-600/3-77-097).
Larimore, R.W. and P.W. Smith. 1963. The Fishes of Champaign County, Il-
linois, as Affected by 60 Years of Stream Changes. Bull. 111. Natur. Hist.
Surv. 28: 299-382.
Nelson, J.S. and S.D. Gerking. 1968. Annotated Key to the Fishes of Indi-
ana. Indiana University, Bloomington, Indiana.
Smith, S.H. 1972. Factors of ecologic succession in oligotrophic fish
communities of the Laurentian Great Lakes. J. Fish. Res. Bd. Canada 29;
717-730.
Trautman, M.B. 1957. The Fishes of Ohio. Ohio State University Press,
Columbus, Ohio.
4.6 TILE DRAINAGE
4.6.1 Initial Tile Sampling Program
Approximately fifty percent of the Black Creek Watershed has subsur-
face drainage. This was determined by a tile monitoring program in 1974
which entailed each tile outlet in the watershed being located and sampled
if flowing. Table 52 is a summary of this sampling program. Using an es-
timate of drained area per tile (based on its size) an estimate of 6000 Ac
of drained land was obtained.
TABLE 52
Summary of Initial Tile Sampling Program
Ditch
Black Creek
Dreisbach
Richeldfer
Gorrel
Wertz
Smith-Fry
Killian
Lake
Fuelling
Warm*
Total
Black Creek
No.# of
Tile Outlets
Located
88
126
139
95
85
81
69
21
43
102
only 747
No.# with Known
Surface Drain
Inlets
11
16
8
9
2
46
Tile Outlets
Sampled
29
36
29
29
49
12
10
7
0
65
201
*Not within Black Creek Watershed (adjacent)
The above tile samples were analyzed for sediment and nutrient concen-
trations. Table 53 provides the concentration statistics of all the col-
STUDY RESULTS
-------
143
lected samples. The water quality of the tiles were not significantly dif-
ferent between ditches except for the soluble nutrients (see Table 54).
The Dreisbach and Richelderfer drains were very high in ammonia and soluble
phosphorus but relatively low in nitrate. Wertz and Smith-Fry showed oppo-
site characteristics. The high ammonia and soluble phosphorus levels are
directly attributed to the high septic input (from Harlan) to the tiles in
the Dreisbach and Richelderfer drains. The high nitrate levels in the
Wertz and Smith-Fry drains result from the more intense row crop agricul-
ture in these watersheds as compared to the other drains.
TABLE 53
Concentration of Sediment and Nutrients in Tile Effluent
Component
Sediment
Ammonia
Nitrate
Sol. Org. N
Sed. N
Sol Inorg. P
Sol Org. P
Sed. P
Range G
634
73
43
58
53
7.6
1.45
13.5
eo. Mean
- Ttig/
79
.63
5.9
3.17
6.26
.02
.01
.10
Mean
^
88
2.33
8.95
6.21
9.35
.31
.05
.56
S.D.
52.5
7.9
7.3
10
7.8
1.04
.14
1.75
TABLE 54
Sediment and Nutrient Concentrations by Ditch - 1974
Mean Concentration (mg/1)
Ditch
Black Creek
Dreisbach
Gorrel
Killian
Wertz
Smith-Fry
Richelderfer
Lake
Wann*
SS
101
101
84
91
74
98
88
110
86
Amm
2.7
9.8
.56
.55
.51
1.0
2.0
5.8
.50
Nit
9.8
7.5
4.6
7.0
10
9.6
10
6.0
10
SON
5.9
4.2
4.1
3.9
3.2
4.7
6.1
26
7.2
SED.N IP
6.3
8.3
7.8
5.2
12
5.0
11
7.4
11
.36
1.3
.05
.06
.02
.15
.19
1.0
.12
SOD
.06
.14
.01
.05
.02
.09
.05
.01
.02
SED.P
.45
.80
.08
.05
.04
.16
.22
2.7
.87
*Not in Black Creek Watershed (adjacent)
No significant difference in sediment or nutrient concentration was
observed between tiles of different flow. This is expected because the
higher flows are achieved when larger areas are drained and not by a la-
teral system design or moisture difference. It should be noted that the
sampling program took several weeks to complete, but the general area
remained dry for the entire period.
STUDY RESULTS
-------
144
4.6.2 Tile Monitoring Continued Program
After the analysis of the 1974 sampling program it was decided that
tile effluent in Black Creek needed further study. The initial study did
not provide for any time dependence within the data. Therefore, twenty
representive tile outlets were selected for a continued monitoring program.
Septic influence, flat and rolling land, intensity of farming, artesian wa-
ter, size of drainage area, etc. were all considered.
This additional tile data served to verify the conclusions of the ini-
tial sampling program and also indicated some significant time dependent
effects. The concentration of all sediment and nutrient components varied
dramatically with time. A trend could not be determined for the concentra-
tions of any component response to tile flow. Table 55 shows the R-squared
values of the correlation study. It is obvious that the flow does influ-
ence concentration but not always in the same way. Table 56 shows the
overall statistical results for two years of data for the twenty sites com-
bined. The high range in the data is evidence of the flow effects. The
high range in the 1974 data (initial study) is due to the variety of tiles,
not flow.
TABLE 55
Values for Exponential Fit - Tile Flow and Sediment-Nutrient Components
Tile R-Squared-Value
Site SS Amm Nit TN SN IP TP SP
1 -.076 -.086 .029 -.005 -.010 -.049 -.044 -.046
2 .127 .004 .000 .026 -.020 .116 .187 .081
3 .004 .003 .000 .003 .000 .009 .056 .020
4 .083 .064 .050 .070 .056 .208 .195 .185
5 .030 .239 .131 .167 .166 .151 .310 .335
6 .090 .066 .014 .026 .019 .229 .225 .088
7 .118 .051 .013 .058 .020 .406 .253 .279
8 .014 .220 -.050 -.017 -.026 .385 .449 .553
9 -.025 .000 .106 .093 .121 .046 .456 .034
10 .009 .020 -.013 .000 .000 .010 .219 .013
11 .323 .053 .083 .015 .027 -.248 -.269 -.195
12 .004 .029 .018 .038 .019 .138 .156 .108
13 .033 .030 -.026 -.001 -.005 .545 .449 .522
14 -.001 -.022 .057 .002 .000 .077 .068 .093
15 -.017 -.009 .003 -.056 -.058 .000 .000 .000
16 .240 -.077 -.014 -.003 -.174 -.036 -.001 -.036
17 -.178 -.021 -.046 -.039 -.036 .037 .154 .031
18 .054 .158 -.001 .010 .000 .544 .502 .506
19 -.012 .065 .081 .067 .081 .201 .264 .089
20 .273 .013 -.106 .000 -.109 .409 .394 .112
Number of samples = 34/site
STUDY RESULTS
-------
145
TABLE 56
Concentration of Sediment and Nutrients for 20 Tile Sites
Component
Sediment
Ammonia
Nitrate
Sol. Org. N
Sed. N
Sol Inorg. P
Sol Org. P
Sed. P
Range
12,900
41
49
151
47
52
9.
39
Geo. Mean
mg/1-
76
.35
6.16
.07
.31
.03
3 .03
.07
Mean
Ill
1.70
8.72
.96
.93
.34
.06
.37
S.D.
496
4.85
6.80
6.83
2.74
2.28
.37
1.90
4.6.3 Results of Automatic Tile Sampler
An automatic tile sampler was developed to help determine the time
dependent relationships in tile flow. The data will also be helpful in the
calibration of a tile flow model which is being developed (see Section
0.0.2).
The automatic tile sampler has been operational since March, 1976 on a
43 acre Hoytville soil. The tile drainage system is uniform with no sur-
face inlets. The two years of record are for a below average tile flow
period, because of the relatively low rainfall during late winter and early
spring which is normally the high flow period for tiles. 1976 was the dri-
est of the two years. Table 57 shows the results of the sediment and nu-
trient losses from the tile site. It is estimated that a more normal out-
flow figure would be between 5-10 cm/year. Note that in both years the ni-
trate losses greatly exceeded all other losses except for sediments.
TABLE 57
Sediment and Nutrient Losses from Tile Effluent - Automatic Sampling Site
Component
Sediment
Ammonia
Nitrate
Sol. Org. N
Sed. N
Sol Inorg. P
Sol. Org. P
Sed. P
Outflow
Rainfall
1976
\rn
Kg,
20.5
.011
.678
.053
.112
.002
.005
.019
1.22 cm
65.70 cm
1977*
53.9
.27
10.6
3.14
.65
.051
.019
.141
6.91 cm
45.50 cm
*Through 7/6/77
STUDY RESULTS
-------
146
The concentration data did vary with flow but not as dramatically as
some other tiles in the Black Creek Watershed. Sediment and nutrient com-
ponents all tended to increase with increased flow. This response is not
linear and tends to be more dramatic during the leading edge of a hydro-
graph then during the trailing edge (see Figure 17). An initial flashing
action seems to be present. Flow vs. concentration correlation results are
given in Table 58. Table 59 shows the average concentration of the tile
effluent for the two years.
TABLE 58
R-Squared Values for the Automatic Tile Sampling Site
Variables
R-Squared
Independent Dependent Linear Fit
Stage SS
Aram
Nit
TN
SN
IP
TP
SP
Flow SS
Amm
Nit
TN
SN
IP
TP
SP
Sus. Solids Amm
Nit
TN
SN
IP
TP
SP
No.# of Samples
1976
.676
.200
.809
.801
.798
.611
.860
.538
.699
.179
.802
.795
.784
.632
.866
.524
.104
.611
.606
.592
.620
.850
.537
47
1977
.152
.336
.040
.318
.332
.408
.143
.391
.182
.371
.036
.359
.372
.454
.168
.438
.197
-.004
.203
.184
.328
.353
.330
180
Exponential Fit
1976
.715
.006
.866
.830
.841
.538
.886
.544
.733
-.007
.840
.806
.809
.545
.867
.522
-.009
.649
.565
.626
.446
.787
.567
47
1977
.121
.304
.045
.253
.261
.516
.268
.437
.145
.325
.044
.281
.286
.525
.292
.465
.148
.002
.141
.113
.169
.366
.203
180
The R-squared values may be misleading because of times
series dependence which exists in the data.
It would be more realistic to assume the number of degrees of freedom
to be less than fifteen which would yield a 5% significant
level of R-squared = 0.3.
The field monitored is nearly flat and has raised field borders. This
prevents any surface runoff and could be considered an ideal BMP for ero-
sion control and soil conservation. Because of this unique situation, we
were able to measure total field losses by simply monitoring the tile ef-
fluent. As a result of forcing all the water through the soil profile,
STUDY RESULTS
-------
suinsan AOOLS
SUSPENDED SOLIDS (100 X MG/LITER)
f
ro
1
CO -j
1
P* en
L. .,_ I
FLOW (LITERS/SEC)
O
ro
CO m
o
3=
-<
CO
O en o
RAIN (MM/HR)
-------
348 TABLE 59
Concentration of Tile Effluent from Automatic Sampling Site by Year
Component
Sediment
Ammonia
Nitrate
Sol. Org. N
Sed. N
Sol. Inorg. P
Sol. Org. P
Sed. P
1976
mg/i
170
.09
5.6
.44
.92
.02
.04
.22
1977*
78
.38
15.3
4.5
.93
.07
.03
.20
*Through 7/6/77
sediment and nutrient losses were greatly reduced. This is due to both the
reduced water loss and the reduced concentration of the nutrient and sedi-
ment, except for nitrate-N whose concentration would not necessarily de-
crease. However, the reduced water loss was much more significant than the
change in nitrate-N concentration. See Table 57 (Also Table 0 in Section
0.0.0) for a comparison of results.
The amount of fertilizer nutrients lost was not large. The field mon-
itored received amounts of fertilizer (180 kg/ha -N and 70 kg/ha -P) in ex-
cess of the Black Creek average. Even at this higher fertilizer rate, the
annual losses of N and P as compared to the amount applied were less then 6
and .2 percent respectively for the wetter of the two years monitored.
The hydrologic response of the tile system on the Hoytville soil is
ver^ ^jd. Peak flows occurred within 5 to 7 hours of all rain storms and
generally returned to zero flow within a week. Deep cracking may L'i a con-
tributing factor to this rapid hydrological response of the tile system.
4.6.4 Pesticide and Herbicide Response
Both insecticide and herbicide contamination was detected in the tile
effluent shortly after a surface application of these chemicals. The in-
secticide used was Furadan and the herbicide used was Lasso II. The accu-
mulative losses of Furadan and Lasso II per hectare were .014 and .003 ki-
lograms respectively with the peak concentration in ppm reaching .213 and
.042 respectively. Herbicide and pesticide losses were of the order of .1
percent of applied. The movement of these pesticides is due mostly to the
deep cracking, which gives water an open channel to flow to the tiles.
4.6.5 Stream Grab Sampling Program
A stream grab sampling program was started in the Black Creek
Watershed in March, 1973. Nineteen sites were selected for monitoring.
(See Section 0.0.1.1.) The Maumee, St. Mary and St. Joseph rivers were
among these selected sites. Samples are collected at least once a week at
each site. Grab sites 2, 6 and 12 were monitored more intensively with the
installation of an automatic pumping sampler.
The additional data obtained at sites 2 and 6 showed that the grab
samples alone could not adequately describe the loadings of sediment and
nutrients from the site's associated watershed. Grab samples alone greatly
under predicted the annual loading for all sediment associated components.
Soluble components were better described by the grab samples, but their as-
STUDY RESULTS
-------
149
sociated error would still be high. Table 60 gives a quantitative view of
the inherent weakness of a grab sampling at an interval greater than the
time response of the watershed. The hydrological time response of the
Black Creek Watershed is of the order of hours compared to a weekly sam-
pling rate.
TABLE 60
Prediction of Loading From Complete Data Base
Loading Prediction (kg/Ha)
Weekly Grab Only*
Component
Sus. Solids
Ammonia
Nitrate
Sol. Org. N
Sed. N
Sol. Inorg. P
Sol. Org. P
Sed. P
Site
1975
572
1.21
14
1.4
3.2
.09
.09
.61
2
1976
178
.59
2.95
.63
.91
.04
.04
.20
Site
1975
869
2.4
9.4
2.0
6.7
.26
.10
1.4
6
1976
142
1.3
1.3
.88
.75
.21
.04
.27
Complete
Site
1975
2130
1.5
19.0
1.7
31
.14
.097
5.2
2
1976
640
.60
5.5
.31
3.9
.064
.034
.98
Data File
Site
1975
3740
1.8
12.0
2.3
28.0
.34
.12
4.5
6
1976
380
.85
2.4
.53
2.8
.18
.040
.73
Flow data for computation does use the complete
flow record, not just that associated with the given grab
sample.
Using the qrab flow record only would cause even larger errors
A strong annual cycle existed for the concentration of nitrate in
streams as seen in Figure 18. To a much lesser degree, the cycle also ex-
ists for ammonia. The nitrate and ammonia concentrations peaked during
late winter or spring. The high percent of tile flow and fertilizer appli-
cation during this period could account for some of the increase in nitrate
levels. Also the biomass uptake of nitrogen would tend to lower the con-
centrations of both nitrate and ammonia in the summer months. Phosphorus
and suspended solids showed no annual variation.
Correlation analysis of the grab and automatic pumping sampler data
combined showed a positive correlation between flow and sediment associated
components and also between suspended solids and sediment associated nu-
trients. The soluble nutrients are generally not correlated with either
flow, stage or sediment. Table 61 gives a summary of the R-squared values
from sites 2 and 6 for the years 1975-76. The low R-square values are in-
dicative of the high degree of scatter observed in the data. Figure 19
shows a plot of stage vs. suspended solids. Similar scatter was observed
for the nutrients. An exponential curve fit was also tried but did not sig-
nificantly improve the R-squared values as see in Table 62. The low R-
squared values also show that functional expressions would not well
represent the data.
The correlation of the sediment and sediment-bound nutrients may be a
little misleading because of strong time series dependence in the samples.
STUDY RESULTS
-------
m
Figure 18. Nitrate Concentration vs. Time for Weekly Grab Samples
25.00
20.00-
15.00-
CE
10.00-
5.00
.00-
-6.00
N03-N
SITE 2
1973 - 1977
156. 312. 468.
624. 780.
TIME- DflYS
936. 1092. 1248. 1404. 1560.
-------
4.000
3.625-
3.250-
0_
Q_
CO
a
2.875 -
2.500-
Q
UJ
O
~Z.
UJ
Q_
CO
ID
CO
CD
IED
2.125-
1.750 -
1.375 -
1.000
SITE 2. 1975
Figure 19. Suspended Solids vs Stage
.000 1.000 2.000 3.000
STflGE (FTJ
4.000
151
5.000
A more realistic number of degrees of freedom would be 20 or less. If the
different hydrological conditions of the watershed between the ascending
and receding sides of the hydrograph were considered separately, an im-
provement is expected in the correlations. This is indicated by the two
distinct scatter patterns in Figure 19.
STUDY RESULTS
-------
152
4.7 TILLAGE STUDIES
It was predicted, prior to the undertaking of the Black Creek project,
that alteration of tillage systems could have a significant impact on water
quality in Black Creek and by extension in the entire Maumee Basin. These
initial expectations were confirmed by simulated rainfall investigations
during the project (see Section
The fact that water quality can be improved by tillage systems which
increase surface cover and residue and which minimize moldboard plowing,
does not guarantee that such systems will be economically feasible or that,
if feasible, they will be voluntarily adopted by landowners in the Maumee
Basin. The tillage studies undertaken in the Black Creek project had the
dual purpose of investigating potential costs, yields, and profits of con-
servation tillage systems on the soils of the study area and of providing
demonstrations of these tillage systems for area landowners.
Tillage studies were begun in 1974 on individual farms with the work
conducted by the landowners. This procedure was not entirely satisfactory,
as the experimental work often conflicted with regular farm operations on
which the landowners were dependent for their livelihood. As a result, ex-
perimental work suffered. Beginning with the 1976 crop year, more tradi-
tional tillage investigations utilizing replicated plots farmed under
closely controlled conditions were begun.
One of the original farmer-tillage demonstrations was carried on for
three years, 1974-76. Yields were compared on three different soil types
in this field. Average yields for the three years across all soils showed
both chisel and disk tillage systems to be very competitive with either
fall or spring moldboard plowing.
Five typical soil types were selected on which to perform replicated
studies. Other areas were used as needed to obtain additional information.
All study areas were fertilized and limed according to Purdue soil testing
laboratory recommendations. Nitrogen was applied to the corn ground at the
rate of 170 Ibs actual N per acre, using 28 percent liquid nitrogen. Phos-
phorus and potassium were broadcast on all plots in the fall ahead of fall
tillage every other year. Furadan was band applied for insect control on
all corn plots. Herbicides were applied pre-emergence at planting. All
chemicals, costs, and rates are reported later.
All 1976 tillage work was done in the spring due to late acquisition
of the land and equipment. Primary tillage for the 1977 season was com-
pleted in the fall.
The corn hybrid used was Pioneer 3780, and the bean variety was Amsoy
71 in 1976 and Shawnee in 1977. Corn seeding rate was 24,000 per acre for
all plots. Soybean seeding rate was 48 Ibs/Ac for all plots. Corn yields
were corrected to 15 1/2 percent moisture, and bean yields to 13 1/2 per-
cent moisture.
Four tillage systems were investigated:
1. NO-TIL the only pre-plant activity was chopping the stalks.
2. CHISEL the soil was chisel plowed 8" to 10" deep and disked
STUDY RESULTS
-------
153
once or twice prior to planting, as needed for a satisfactory seed-bed.
3. DISK two diskings to a 4" depth prior to planting were used
with no primary tillage.
4. CONVENTIONS! the soil was moldboard plowed and disked twice
before planting.
In 1976 corn was not cultivated, but all beans were cultivated, except
the no-til plots. In 1977 all plots except no-til were cultivated. All
tillage systems were repeated for corn after corn, corn after beans, and
beans after corn. For the first year, 1976, the previous crop was the same
for all systems. It was necessary to change row direction on two of the
experiments. On these, early spring disking was necessary for the no-til
plots in 1976.
Although replicated tillage comparisons were begun in 1976, the effect
of previous crops on tillage success could not be determined until 1977 and
is not included in this report. For corn, the three conservation tillage
systems yielded as well as or better than conventional tillage on 4 of the
5 soils in 1976. Yields were not checked on Nappanee silt loam in 1976 due
to extreme variation in the experimental area.
1977 corn harvest was not yet complete as this report was prepared.
Yield data gathered so far shows all conservation tillage systems below
conventional on poorly drained Hoytville and Nappanee soils. Corn yield
with chisel tillage equalled conventional corn on better drained Whitaker
and Haskins soils. (Data from Morley soils not yet available.)
Soybean yields were not significantly different among tillage systems
in 1976 on upland Morley and Haskins soils. Disk and no-till soybean
yields were slightly reduced on Whitaker loam and severely reduced on Nap-
panee silt loam. Increased phytophthora root rot with disk and no-till
systems appeared to be the major cause of reduced yields.
Canada thistle, morning glory, and field bindweed were a problem in
spots in the experimental areas but had little influence on yields.
In 1977, dry weather after planting caused reduced and delayed germi-
nation leading to more variation than usual in plant stands among treat-
ments. Phytophthora root rot was much less severe than in 1976, but was
again more prevalent in disk and no-till soybeans. Canada thistle, morning
glory, and field bindweed were greater problems in 1977, were more diffi-
cult to control in soybeans than in corn, and were usually more severe with
no-plow tillage systems.
In 1977 bean yields with all conservation tillage systems were greatly
reduced on Hoytville and Nappanee soils, slightly reduced on Whitaker, and
equal to or greater than conventional yields on Morley and Haskins soils.
Previous research in Indiana has identified the major factors which
influence corn response to tillage.
These are:
a. Soil drainage the better drained the soil, the less tillage
STUDY RESULTS
-------
154
TABLE 61
Linear Correlation Results for Sites 2 and 6
R-Squared Value
Variables
Independent
Stage
Dependent
SS
Amm
Nit
TN
SN
IP
TP
SP
Suspended Sol. Aram
Flow
Nit
TN
SN
IP
TP
SP
SS
Amm
Nit
TN
SN
IP
SP
No.# of Samples
Site
1975
.246
.000
-.030
.271
-.028
.013
.367
.010
-.001 -
-.106
.333
-.103 -
.033 -
.519
.022 -
.212
-.001 -
-.075
.222
-.072
.001 -
.001 -
646
2
1976
.175
.000
.078
.264
.057
.002
.272
.004
.064
.000
.170
.006
.063
.276
.080
.164
.005
.012
.127
.005
.001
.001
396
Site
1975
.413
-.030 -
-.010
.232
-.014 -
.023 -
.092
.026 -
-.007 -
-.078
.349
-.068 -
-.018 -
.114
-.010 -
.393
-.007 -
-.028
.037
-.025 -
-.006 -
-.005 -
461
6
1976
.191
.096
.050
.187
.006
.028
.173
.030
.023
.000
.600
.030
.064
.570
.065
.247
.068
.010
.173
.023
.047
.050
409
needed.
b. Previous crop shallow or no-tillage planting is more likely to
be successful for corn in rotation.
c. Latitude shallow or no-tillage planting is more likely to be
successful in areas with a longer growing season.
d. Pest control where weeds, insects and diseases cannot be ade-
quately controlled with chemicals, no-plow tillage is not likely to be suc-
cessful .
Research so far in the Black Creek study tends to support these
points.
1976 and 1977 yield results are presented in the following tables. It
should be noted that tillage systems must be evaluated over several years
to accurately determine their success on different soil types.
STUDY RESULTS
-------
155
4.7.1 Hoytville Silty Clay Loam
TABLE 62
Exponential Correlation Results for Site 2 and 6
R-Squared Value
Variables
Independent Dependent
Stage SS
Anrn
Nit
TN
SN
IP
TP
SP
Suspended Sol. Amm
No.
Nit
TN
SN
IP
TP
SP
Flow SS
Amm
Nit
TN
SN
IP
TP
SP
# of Samples
Site 2
1975
.439
.001
-.003
.328
-.007
.097
.466
.041
.000
-.048
.384
-.055
.154
.761
.157
.295
-.004
-.035
.189
-.049
.016
.307
.009
646
1976
.308
.002
.188
.349
.141
.042
.307
.038
-.046
.072
.279
.024
-.010
.338
-.033
.238
-.003
.045
.137
.025
.013
.190
.009
396
Site 6
1975
.617
-.012
.000
.301
-.005
.051
.429
.046
-.006
-.030
.430
-.041
.035
.626
.045
.218
-.007
-.018
.035
-.019
-.004
.058
-.006
461
1976
.242
-.073
.097
.241
.000
-.023
.186
-.022
-.025
.031
.484
-.012
-.062
.398
-.073
.239
-.081
.025
.173
-.013
-.040
.149
-.043
409
TABLE 63
Corn Response to Tillage System
Tillage
No-til (b)
Double disk
Chisel
Plow
(a) Any yields
Plants/Ac
1976 1977
21,500
22,015
18,937
19,640
followed by the
8
8
11
17
,500
,600
,200
,200
Bu/Ac
1976 (a) 1977
134.9 b
139. 2ab
144. la
124.8 c
same letter are not
different at the 10% level.
(b) These plots
were disked once
in 1976
to change
92.0
90.8
105.0
120.4
significantly
row direction.
STUDY RESULTS
-------
156
TABLE 64
Soybean Response to Tillage System
Bu/Ac
Tillage 1976(a) 1977
No-til(b) 31.8a 35.7
Double disk 32.4a 35.2
Chisel 31.Sab 33.4
Plow 28.3 b 40.4
(a) Any yields followed by the same letter are not significantly
different at the 10% level.
(b) These plots were disked once in 1976 to change row direction.
Planting dates: CornMay 5, 1976; SoybeansMay 25, 1976
May 11, 1977 May 25, 1977
This experimental area was in soybeans in 1975. In 1976, chisel pro-
vided highest corn yield and conventional the lowest, with shallow tillage
intermediate. Conventional bean yields were also lowest in 1976. The con-
ventional system was at a disadvantage on this soil because of the poor
seedbed and delayed emergence associated with spring plowing.
In 1977 the situation was essentially reversed, with all no-plow sys-
tems having greatly reduced germination in the early season drought. Con-
ventional tillage gave best stands and yields for both corn and soybeans.
4.7.2 Nappanee Silt Loam
TABLE 65
Corn Response to Tillage System
Plants/Ac
Tillage
No- til
Double disk
Chisel
Plow
1976
19,000
19,000
20,625
20,000
1977
16,000
18,700
19,500
21,800
Bu/Ac
1976
No Data
No Data
No Data
No Data
1977
76.6
105.7
111.2
126.6
The experimental area was in corn in 1975. Due to variation in
drainage and nitrogen deficiency, no corn yields were taken in 1976. In
1977, stands with the conservation tillage systems were reduced slightly,
for corn, but yield reductions were greater than would be indicated by
stand losses on this very poorly drained soil.
Soybean yields with no-til and disk tillage were severely reduced in
STUDY RESULTS
-------
157
TABLE 66
Soybean Response to Tillage System
Tillage
No- til
Double disk
Chisel
Plow
(a) Yields followed
different at the
Bu/Ac
1976 (a)
7.8 c
7.9 c
17.2 b
29. 8a
by the same
90% level.
1977
28.5
32.3
36.7
48.7
letter
are not significantly
j Planting dates: CornMay 11, 1976; SoybeansMay 25, 1976
i May 12, 1977 May 20, 1977
1976. Yield with chiseling was intermediate between no-til and convention-
al. The same yield pattern occurred in 1977, but reductions were much
less. Much of the yield loss appeared to be due to phytophthora root rot.
This disease is known to be more damaging in poorly aerated soils, but the
great difference in infection related to tillage depth was surprising.
4.7.3 Whitaker Loan
TABLE 67
Corn Response to Tillage System
Tillage
No-til(b)
Double disk
Chisel
Plow
Plants/Ac
1976 1977
29,843
20,953
20,859
14,937
19
19
22
23
,100
,800
,300
,200
1976
142
145
150
142
.9
.7
.3
.6
Bu/Ac
(a) 1977
131.
145.
153.
155.
2
7
7
6
(a) Yields were not significantly different at the 90% level.
(b) Plots were disked once in 1976 to accomodate change in row
direction.
i
This experimental area was in soybeans in 1975. There was no signifi-
cant difference in corn yield among tillage treatments in 1976, but no-til
yields were reduced in 1977. Shallow seed placement in 1977 no-til plots
led to delayed germination, the most likely cause of yield loss.
Deep tillage, plowing and chiseling apparently improved soybean yields
in 1976 and no-til yields were also reduced in 1977. Again, phytophthora
STUDY RESULTS
-------
158
TABLE 68
Soybean Response to Tillage System
Tillage
No-til (b)
Double disk
Chisel
Plow
Bu/Ac
1976 (a) 1977
34.2 c 42.0
39.8 b 45.6
41. lab 44.3
44. 8a 48.8
(a) Yields followed by the same letter are not significantly
different at the 90% level.
(b) Plots were disked once in 1976 to accomodate change in row
direction.
Planting dates: CornApril 23, 1976; SoybeansMay 21, 1976
May 10, 1977 May 26, 1977
root rot appeared to contribute to these yield losses.
This is one of the better drained soils in the watershed and is a soil
where we would expect disk and no-til yields to be competitive with deep
tillage for corn. There is insufficient experience to make similar predic-
tions for soybeans.
4.7.4 Morley Silt Loam
TABLE 69
Corn and Soybean Response to Tillage System
Tillage (b)
No- til
Chisel
Plow
Corn,
Plant/Ac
20,281
18,812
17,609
1976 (a)
Bu/Ac (c)
91.1
89.5
88.2
Soybeans ,
1976 (c)
23.2
21.7
24.3
Bu/Ac
1977
35.1
30.1
29.1
(a) 1977 corn plots were not harvested when this report was prepared.
(b) Due to a space limitation the disk treatment was omitted at this
site.
(c) Yields were not significantly different at the 90% level.
Planting dates: CornMay 13, 1976; SoybeansMay 27, 1976
May 17, 1977 May 23, 1977
The 1975 crop at this site was corn. Yield differences for both corn
and soybeans were not significant in 1976. No-til bean yields were highest
in 1977. It is important to note that conservation tillage systems were
STUDY RESULTS
-------
159
equal to plowing on this erosive soil.
4.7.5 Haskins Loam
TABLE 70
Corn Response to Tillage System
Plants/Ac
Tillage
No-til
Double disk
Chisel
Plows
1976
21,156
20,453
21,000
20,500
1977
16,500
20,400
21,200
22,600
Bu/Ac
1976 (a)
121. 9a
105.8 b
111.7 b
111.5 b
1977
110.8
118.7
126.7
129.0
(a) Yields followed by the same letter are not significantly
different at the 90% level.
TABLE 71
Soybean Response to Tillage System
Bu/Ac
Tillage
No-til
Double disk
Chisel
Plow
1976 (a)
24.0
24.4
20.8
25.7
1977
43.3
43.5
46.5
44.5
(a) Yields not significantly different at the 90% level.
Planting dates: CornMay 12, 1976; SoybeansMay 27, 1976
May 16, 1977 May 23, 1977
Corn was grown in this experimental area in 1975. No-til corn yield
was better than other systems in 1976, showing less drought stress in mid-
summer. Moisture conserved by the surface residues was important during
this period. In 1977, no-til stand and yield were both lower than for oth-
er systems for corn. There was little difference among tillage systems in
soybean yield in either year.
4.7.6 Chemicals Applied to Replicated Plots
The following chemicals and rates were used both years.
were typical retail prices in the area.
Prices used
STUDY RESULTS
-------
160
TABLE 72
Costs for Chemicals, Corn and Soybean
For corn
Aatrex 4L, 1 qt/Ac @ 4.08/qt
Bladex 4L, 1 qt/Ac @ 4.00/qt
Lasso, 2 qts/Ac @ 3.87/qt
Furadan, 10#/Ac @ .76/lb
Paraquat, 1 pt/Ac @ 4.99/pt
Total
For soybeans
Lorox, 2#/Ac @ 3.70/lb
Lasso, 2 qt/Ac @ 3.87/qt
Paraquat, 1 pt/Ac @ 4.99/pt
Total
Cost for plows,
chisel, disk
$4.08
4.00
7.74
6.70
$22.52
7.40
7.74
$15.14
Cost for
no-til
$4.08
4.00
7.74
6.70
4.99
$27.51
7.40
7.74
4.99
$20.13
4.7.7 Tillage Demonstrations
This continuous corn tillage trial on the Roger Ehle farm was one of
the original farmer-cooperator demonstrations started in 1974. Although
not replicated, the trial is interesting because the tillage plots cross
three soil types in the same field, and because the years ranged from very
dry (1974), to very wet (1975), with 1976 being closer to "average."
Yields from the soil types reflect the differences in seasons.
Response to tillage is not so clear-cut. Two things stand out in the data.
First, fall chisel yields were as good as, or better than, other systems
across all conditions. And, second, no-til yields tended to be less under
all conditions. We cannot fully explain the no-til response. Surface
residue systems are usually more successful on droughty soils such as Osh-
temo.
TABLE 73
Corn Response to Tillage, Roger Ehle Farm, Bu/Ac
Rensselaer CL 1.
74 75 76
No-Til 128 115 125
Disk, Spring & Fall 187 138 164
Fall Chisel 184 143 162
Fall Plow 118 137
Spring Plow 168 139 161
Whitaker 1.
74 75 76
81 129 101
94 152 116
111 164 122
80 143 -
72 151 134
3 Year
Average
Oshtemo FSL All Soil
74 75 76 Types
27 137 93
30 159 96
54 166 102
64 167 -
21 166 90
104
126
134
122
STUDY RESULTS
-------
161
In 1976, corn was planted into standing red clover 8"-10" in height,
an intercrop from the previous year's wheat. Soil types were Haskins and
Morley silt loam. Planting date was May 20 and final population was
19,500/Ac. Fertilizer applied was the same as for replicated plots. The
following chemicals were applied at planting:
TABLE 74
Chemicals Applied at Planting
Aatrex 4L, 1.2 qt/Ac @ 4.08/qt = $4.90
Bladex 4L, 1.2 qt/Ac @ 4.00/qt = 4.80
Lasso, 2.4 qt/Ac @ 3.87/qt = 9.29
Paraquat, 2 pts/Ac @ 4.99/pt = 9.98
Furadan, 10#/Ac @ 0.67/lb = 6.70
Stand was reduced in several places in the demonstration due to rodent
damage. The legume growth was greater than is recommended for no-til sod
planting. Either taking an early hay crop or chemically killing the sod
when about four inches in height would have allowed easier coulter penetra-
tion and provided less cover for rodents. Machine harvest of this four-
acre demonstration yielded 98.2 bu/Ac, about equal to conventionally tilled
corn in the area. Stress symptoms were much less severe in the sod-planted
corn during summer drought.
Corn was no-til planted in soybean stubble in 1976. This was a seven
acre demonstration near one of the replicated tillage experiments, with
chemicals and fertilizer applied as in the replicated study. Corn was
planted on April 22 on erosive Morley and Haskins silt loam soils. The
machine harvest yield was 102.9 bu/Ac. The demonstration was quite suc-
cessful, with an excellent stand and good weed control.
4.8 SOCIO-ECONOMIC STUDIES
Socio-economic studies in the Black Creek Watershed involved the col-
lection of primary data from farmers in the watershed in order to determine
changing attitudes of farmers toward soil conservation practices and to
determine the cost of alternative soil conservation practices. The diver-
sity of soil types, farm sizes, cultural situations and management prac-
tices in the watershed provided information on a variety of farming situa-
tions. The following material summarizes the approach of the study, the
survey instruments used for data collection, the model design and results.
The approach to the sociological portion of the study involves
analysis of those factors which influence the decision of the farmer to
adopt conservation practices. These factors are related to those aspects
of the Black Creek Project which are policy instruments which can be al-
tered to influence the rate and extent of adoption of conservation prac-
tices .
The approach to the economic portion of the study includes analysis of
the economic gain or loss to individual farmers that results from utilizing
STUDY RESULTS
-------
162
alternative conservation practices rather than currently prevalent methods
of crop production. The results for the individual farmer are aggregated
to determine the relative cost of instituting alternative conservation
practices in the watershed.
Since data were needed from the farmers in the watershed, a survey in-
strument was prepared to obtain these data. The instrument was employed at
two different time periods two years apart. This was necessary to obtain
estimates of the changes in attitudes of farmers toward conservation during
the operation of the project. The data were obtained for the survey in-
strument by personal interview of farmers in the watershed.
The model design of the study is divided L, j two sections. One sec-
tion describes the model used in sociological research while the second
section describes the model used in the economic research.
4.8.1 Model: Sociological
The questionnaires which were administered (in 1974 and 1976) were
designed to measure characteristics of the farming operation, as well as
various social, psychological, and economic characteristics of the farmer,
including his attitudes toward pollution, government and toward the project
itself. The specific variables used in this study are described below:
XI Education (EDUCA) is the level of education completed by the farmer
in actual years.
X2 Socio-economic status (SES) is the gross annual income of the farm-
er, in dollars, plus the total number of acres in his farm.
X3 Perceived need for innovation (NEED) is an index of 8 questions
aimed at assessing the farmer's perception of the adequacy and effec-
tiveness of pollution control efforts prior to the project. The lowest
value possible was 8 and the highest possible was 32.
X4 Off-farm employment (OFFFARM) is measured by the percent of the
farmer's 1973 family income which came from his off-farm employment.
X5 Leadership score (LEADER) is the actual number of sociometric
choices the farmer received when respondents were asked who was a
well-respected farmer in the area.
X6 Ethnic group (ETHGRP) is a dichotomous variable which indicates
whether the farmer is Amish or Non-Amish. It is coded (1) for Amish
and (2) for Non-Amish.
X7 Advice from leader (ADVICE) indicates whether the farmer has sought
advice about farming practices from the person he selected as a well-
respected farmer referred to here as "LEADERS", the variable
representing community leadership. It is coded (0) for "no" and (1)
for "yes".
X8 Agency Contact (AGENCYCT) is an index of the amount of contact the
farmer has had with each of the agencies involved in the project during
the past year. For each agency the actual number of contracts was cod-
STUDY RESULTS
-------
163
ed unless the number exceeded four. The scores for each agency are
then summed to yield a total score for each farmer.
X9 Knowledge of the project (KNOW) is an index of two questions indi-
cating whether the farmer had knowledge of the project and was familiar
with its intentions as of January, 1974. It is scored from (0) for no
familiarity to (4) for much familiarity.
Xll Participation in the project (TAKEPART) indicates whether and to
what extent the farmer has participated in the project as of March
1976. It is measured by an index of 2 questions designed to measure
the farmer's perception of how much he had participated in the project
since its inception.
These variables, along with their means and standard deviations, are
presented in Table 75. They are presented as they fit the stages of the
Black Creek Model in Figure 20.
STUDY RESULTS
-------
RECEIVER VARIABLES
1. PERSONAL
EDUCATION
PERCEIVED NEED
2. SOCIAL
SES
OFF-FARM EMPLOYMENT
LEADERSHIP SCORE
SOCIAL SYSTEM VARIABLES
AMISH/NON-AMISH
KNOWLEDGE
OFBCP
PERSUASION
TOWARD BCP
Figure 20. Black Creek Scciological Model
PARTICIPATION
IN BCP
CO
EH
i-q
D
W
W
CO
-------
165
TABLE 75
Sociological Model Variables
Variable
XI Education of the farmer
X2 Socio-economic status
X3 Perceived need for innovation
X4 Off-farm employment
X5 Leadership score
X6 Amish or Non-Amish
X7 Advice from leader
X8 Agency contact
X9 Knowledge of the project
X10 Persuation toward the project
Xll Participation in the project
Mean
9.573
20275.281
27.225
77.494
2.528
1.640
1.337
6.006
2.989
14.584
8.325
Standard
Deviation
3.285
12411.972
5.011
104.067
6.455
.483
.475
7.969
.805
6.210
3.254
4.8.2 Model: Economic
A linear programming model was formulated to create a static equili-
brium cropping plan. The model maximizes profit by selecting the most pro-
fitable combinations of crops and tillage practices while complying with
resource constraints. Proper timing of operations and resource availabili-
ty was specified in the land preparation, planting, and harvesting phases
of production for each tillage practice. The four primary tillage prac-
tices were moldboard plow, chisel plow, double disking, and no-tillage
preparation (see Table 76). The importance of timeliness in modern crop
production is built into the model by adjusting yields to correspond to
planting and harvest dates. In addition, cropping operations are specified
to insure proper sequence.
TABLE 76
Primary Tillage Practices
Operation
Moldboard plow
Chisel
Disk or Field Cultivate
Disk
Plant
No-till Plant
Tillage Tillage Tillage Tillage
Practice Practice Practice Practice
1234
X
X
X
X
X
X
X
X
X
The analysis was conducted on four sets of data. These data sets
correspond with: (1) a 580 acre farm with less than an average two percent
land slope, (2) a 370 acre farm with less than average two percent land
slope, (3) a 580 acre farm with greater than an average two percent land
STUDY RESULTS
-------
166
slope, and (4) a 370 acre farm with greater than an average two percent
land slope.
The model solutions for each data set were examined at several levels
of soil loss which constrained the productive activities below levels per-
mitted in the optimal unconstrained solution. The results of the modeling
process were evaluated at selected prices for the crop output. For exam-
ple, the model solutions were compared for wheat prices at both $3.20 per
bushel and $2.46 per bushel. The results were further evaluated for policy
alternatives which either eliminated fall tillage or prohibited the use of
the moldboard plow.
The analysis was performed first for the individual farm data sets to
examine the change in the crop management practices and subsequent changes
in net returns to labor and management. From these results, costs of each
policy alternative were aggregated to represent modern crop management in
the Black Creek Watershed.
The land, labor, and machinery resources of the two hypothetical farms
used in this study were based on representative findings obtained in a sur-
vey of modern operations in the Black Creek Watershed. Model coefficients
pertaining to labor, machinery, and time resources used in crop production
were calculated according to procedures outlined in publications pertaining
to crop budgeting (Brink et al. 1976, Doster et al. 1976). These calcu-
lations used labor and machinery information taken from a survey of the
watershed and secondary data pertaining to rates of land, labor, and
machinery use (Doster e_t al. 1976). Product prices and production cost
estimates also came from secondary sources (Agricultural Prices, 1974).
Estimates of soil losses from agricultural cropland in the Black Creek
Watershed were derived from the Universal Soil Loss Equation (Ohio CES,
1975). Crop management and yield information for these calculations was
obtained from both agricultural data collected in the survey of the
watershed and the secondary information concerning soil characteristics and
crop production. Yield differences associated with different tillage prac-
tices were based on previous studies (Mannering et al. 1976, Griffith et
al. 1973). Similarly, secondary information showed that different tillage
practices cause different amounts of soil loss (Mannering et al. 1976,
Griffith et al. 1973).
4.8.3 The Data Base
Computer coefficients which specify the resource requirements per acre
are given in Table 77 and Table 78. A descriptive sample of the calcula-
tions used to derive these coefficients and a reprint of the field and la-
bor time requirements chart from the Purdue Crop Budget Model B-93 (Doster
et al. 1976) are given in Appendix B. Field rates from this chart were
used~to insure consistency among field and labor time requirements.
Coefficients that specify wage rates are based on figures used in the
1976 Purdue Crop Budget Model B-93 and were $4.00 an hour for labor hired
in and $2.00 an hour for labor hired out of the operator's own crop activi-
ties. Part-time labor is assumed to be only 80 percent as efficient as the
STUDY RESULTS
-------
TABLE 77
Computer Coefficients - 370 Acre Farm
^ahor Tractor Harvest Field
Requirements Field Prep- Field Prep- Field Plant- Tield Hours
(Man per field aration Hours aration Hours ing Hours Hours Available
hours/acre/ per Acre- per Acre- per per ' per
!i£iLL> fell Spring Acre Ac,e Arwi
Corn Preparation
Till a
Till 92
Till ,"3
Com Production
Till PI
Till 22
Till j?3
Soybean Preparation
Till SS\
Till 12
Till S3
Till 34
Soybean Production
Till f'l
Till 02
Till C3
Till (74
Wheat Production
Till (71
Till i'2
Till ?3
Till 54
.6486 .3093
.5097 .2062
.4516
1.1312
1.1812
1.1C12
-.4507 .3093
.3477 .2062
.2856
.9494
.9494
.9494
.9023
1.2579 .3093
1.1549 .2063
1.0963 .1667
.7912 .1586
.1833 .6521
.1833 5221
.1833 .4673
Plant Cult.
.1993 .1993 .1325
.1993 .1993 .1325
.1993 .1993 .1325
Harvest - .3333
1333 .uis
1233 .3303
.1833 .2840
PJnnt Cult.
.1993 .1993 .1325
.1993 .1993 .1325
.1993 .1993 .1325
.1586 .1586 .1325
Harvest - .3333
.6180
.5150
.4507
.5536
.3545
.3545
.3545
.2146
.2146
.2146
.2146
.4000
.4000
.4000
.4000
D
K
en
G
cn
-------
00
TABLE 78
Computer Coefficients - 580 Acre Farm
Labor
Requirements Field Prep-
(Ilan per field aratior. Hours
hour/acres/ per Acre-
hour) Fall
Corn Preparation
Till Si
Till 1?2
Till 63
Corn Production
Till Si
Till 12
Till 03
Soybean Preparation
Till £1
Till ?2
Till 03
Till #4
Soybean Production
Till i'l
Till £2
Till #3
Till 04
Wheac Production
Till 9i
Till i?2
Till £3
Till l?4
.5452 .1415
.4638 .1031
.3307
1.1312
1.1SI2
1.1812
13332 .1415
.3037 .1031
.2037
.9494
.9494
.9494
.9023
1.2216 .1415
1.1140 .1031
1.0259 -1072
7^12 .1536
Field Prep- Field Plant- Field
aration Hours ing Hours Hours
per Acre- per per
Srrin" Acre Acre
.1500
.1500
.1500
Plant
.1057 .1057
.1057 .1057
.1057 .1057
Harvest
.1500
.1500
.1500
Plnnt
.1057 .1057
.1057 .1057
.1057 .1057
.1536 .1556
Harvest
.5035
.4487
.1057
Cult.
.1325
.1325
.1325
- .3333
.3535
.3388
.1997
Cult.
.1325
.1325
.1325
.1325
- .3333
.4722
.4653
.3653
.5586
Hours
»v»i1»>h:e
par
Acre
.3545
.3545
.3545
.2146
.2146
.2146
.2146
.4000
.4000
' A^r*
*+ v ^J v
.4000
a
tr1
i-3
en
-------
operator's or full time labor.
169
Based on Information contained in Farm Planning and Financial Manage-
ment (USD&, 1975), machine operation direct cost per acre is assumed to be
$18.00 for corn, $8.75 for soybeans, and $6.25 for wheat. Indirect
machinery and equipment costs per acre are assumed to be $24 for corn, $20
for soybeans, and $10 for wheat. Indirect machinery costs are deducted
from the objective function value only after the basis has been determined
and are used in calculating returns to labor and management.
Based on a land capability analysis of the Black Creek Study Area as
given in Table A-3 of the Environmental Impact of Land Use on Water Quality
(A Work Plan), two slope categories for the watershed can be developed.
The first category includes all the land capability units with land slopes
less than two percent. Of the 12,038 acres in the watershed, 6,293 defin-
itely fit this class. The other more erosive category, land capability un-
its with slopes greater than two percent, includes 2,047 acres. The II W-2
land capability unit is characterized by land with slopes from zero to two
percent as well as land with slopes from two to six percent. The types of
soils in the watershed characterized by these two ranges in slope are
equally divided. Thus, the 3,698 acres in the II W-2 capability group were
assumed to be equally divided between the two slope categories. This
resulted in 8,142 total acres, or 68 percent of the watershed acres, being
categorized as less than a two percent slope. The remaining 3,896 acres,
or 32 percent of the watershed, are assumed to have slopes greater than two
percent.
For the purposes of this study, an estimated 7,500 acres of cropland
are characterized by modern management practices. The estimated 5,100
acres with land slopes less than two percent are assumed to have a univer-
sal soil loss equation length slope (SL) factor of .128. The 2,400 acres
with greater than two percent slopes are assigned an SL factor of .6. The
importance of land slope to this study becomes apparent from the following
discussion of the universal soil loss equation.
Soil loss coefficients are derived from calculations based on
Wischmeier and Smith's Universal Soil Loss Equation (Wischmeier et al.
1972).
Table 79 shows the average annual soil loss coefficients that
correspond to the various tillage practices, crop rotations, crop manage-
ment factors, and two length-slope (LS) factors that are associated with
the given soil type. Soil losses were specified to correspond with common
three year combinations of crop rotations and tillage practices.
Corn yields are assumed to average 118 bushels, soybeans average 42
bushels and wheat averages 52 bushels. Variations in yields and moisture
correspond to different planting and harvesting dates. Table 80 presents
these variations.
Corn yields are reduced for delays in planting at 1 bushel/acre/day
from May 10 to May 23, and 2 bushels/acre/day from May 24 to June 6 (Doster
et al. 1972). A two percent harvest field loss is assumed between the
first and second harvest periods and a three to six percent loss between
the last two periods. Similar yield penalties exist for soybeans. In ad-
dition to planting and harvesting dates, corn and soybean yields are af-
fected by tillage practices.
Corn yield variations for soil types similar to those located in Black
STUDY RESULTS
-------
170
TABLE 79
Average Annual Soil Loss Coefficient
Three-Year ,
f\ r\
3asc Combination '
CTP1-CTP1-SBTP1
CTP2-CTP2-SBTP2
CTP1-SBTP1-SBTP1
CTP2-CTP2-SBTP2
CTP1-CTP1-SBTP3
CTP1-SBTP1-WTP1
CTP2-SBTP2-WTP2
CTP1-SBTP3-WTP3
CTP1-CTP1-SBTPI
CTP2-CTP2-SBTP2
CTP1-CTP1-SBTP3
CTP1-SBTP1-SBTP1
CTP2-SBTP2-SBTP2
CTP1-SBTP3-SBTP1
CTP1-SBTP4-S3TP1
CTP1-WTP1-SBTP1
CTP2-WTP2-SBTP2
CTP1-WTP3-SBTP3
SBTP1-SBTP1-WTP4
CTP1-CTP1-WTP1
CTP2-CTP2-WTP2
CTP1-CTP1-WTP3
CTP1-SBTP1-WTP1
CTP2-SBTP2-WTP2
SBTP1-SBTP3-WTP1
SBTP1-SBTP4-WTP1
SBTP1-WTP1-VTPI
SBTP2-WTP2-WTP2
SBTP3-WTP3-WTP1
SBTP3-WTP4-WTP1
Crop
Management
Factor (C)
.3519
.1720
.1362
.1720
.2368
.3000
.1320
.1501
.3519
.1720
.2368
.4362
.1694
.3204
.3204
.3000
.1320
.1501
.1549
.2227
.0952
.2328
.3000
.1320
.1549
.1549
.2234
.0847
.0847
.0847
Three Year
Average Annual
Soil Loss (Tons/Acre)
LS = .128
2.365
1.155
2.931
1.155
1.591
2.017
.887
1.008
2.365
1.155
1.591
2.931
1.125
2.153
2.153
2.016
.887
1.008
1.041
1.497
.640
1.564
2.017
.887
1,041
1.041
1.515
.555
.555
.555
LS .6
11.086
5.414
13.739
5.414
7.459
9.450
4.159
4.725
11.086
5.414
7.458
13.739
5.273
10.092
10.092
9.450
A. 159
4.725
4.880
7.017
3.000
7.331
9.450
4.159
4.880
4.880
7.102
2.602
2.602
2.602
C » Corn
SB = Soybeans
W «= Wheat
TP1 " Tillage Practice 1
TP2 = Tillage Practice 2
TP3 = Tillage Practice 3
TP4 = Tillage Practice 4
Refer to Table 3.1 for explanation.
STUDY RESULTS
-------
TABLE 80
Average Corn Yield and Moisture Coefficients
Harvest
Period
Sept 27
to
Oct 17
Oct. 18
to
Nov. 7
Nov. 8
to
Nov. 28
Tillage-
Planting
Treatment
TP1
TP2
TP3
TP1
TP2
TP3
TP1
TP2
TP3
Planting Period
Production item
Corn Yield
Moisture Points9
Corn Yield
Moisture Points3
Corn Yield
Moisture Points3
Corn Yield
Moisture Points3
Corn Yield
Moisture Points
Corn Yield
Moisture Points
Corn Yield
Moisture Points3
Corn Yield
Moisture Points
Corn Yield
Moisture Points
April 26
to
May 2
321
1210
109
1090
108
1080
119
595
107
535
106
530
117
468
105
420
104
416
May 3
to
May 9
121
1452
109
1308
108
1298
119
838
107
749
106
742
117
585
105
525
104
520
May 10
to
May 16
114
1596
102
1428
101
1414
112
1008
100
900
99
891
110
668
98
588
97
582
May 17
to
Mav 23
105
1260
94
1128
94
1128
103
824
92
736
92
736
May 24
to
Mav 30
98
1568
88
1A08
90
1440
86
1056
56
946
80
968
tn
C
tr1
1-3
Moisture points above 14 point base = moisture points above base x bushels/acre.
-------
Creek were estimated based on conversations with Mannering and Griffith.
Similarly, reductions in soybean yields are assumed to occur when
moldboard preparation is replaced. Adoption of chisel or double disc
preparation is assumed to cause a five percent reduction in yield and no-
tillage planting is assumed to cause a 10 percent reduction. Soybean pro-
duction activities and corresponding yields are also presented in Table 81.
TABLE 81
Average Soybean Yield
Harvest
Period
Sept. 13
to
i
Sept. 26
Tillage
Planting
Treatment
TP1
TP2
TP3
Item
Soybean
Soybean
Soybean
Yield
Yield
Yield
May 3
to
May 9
43
41
41
May 10 May 17 May 24 May 31
to to to to
May 16 May 23 May 30 June 6
43
41
41
TP4 Soybean Yield 39 39
Sept. 27
to
Oct. 17
TP1
TP2
TP3
TP4
Soybean
Soybean
Soybean
Soybean
Yield
Yield
Yield
Yield
41
39
39
37
41
39
39
37
40
38
38
36
38
36
36
34
34
32
32
30
Elevator drying costs for corn are taken from the Purdue Crop Budget
B-93 and are assumed to be $.16 per every ten moisture points dried down.
Fertilizer, insecticide, herbicide, and credit and miscellaneous costs are
presented in Table 82.
Original product prices are assumed to be $2.30 per bushel of corn,
$5.50 per bushel of soybeans, and $3.20 per bushel of wheat. Based on the
survey findings, land market value was assumed to be $1,500 per acre. In-
terest in land at eight percent per acre and an $8.00 change per acre for
taxes and land maintenance were taken from Farm Planning and Financial
Management (USDA, 1975). Thus, $128.00 per acre is deducted from the ob-
jective function value when returns to labor and management are derived.
The right-hand-side values of the linear program specify the resource
limit. the number of productive acres for any one crop was never allowed
to exceed 2/3 of the total available acreage. This permitted the use of a
three-year crop rotation. The right-hand-side value for the soil loss con-
straint equalled the sum of the allowable soil loss per acre multiplied by
the number of acres.
STUDY RESULTS
-------
173
TABLE 82
Production Costs
o
Costs per Acre of Production
Item
Fertilizer
Insecticide
Herbicide
Seed Cost
Credit & Miscellaneous
Corn
$44.00
7.00
11.00
11.00
8.00
Soybean
$12.00
0
13.00
8.00
6.00
Wheat
$34.00
0
0
12.00
5.00
Based on cost figures contained in Purdue Crop Budget Model B-93, p.
9-12.
4.8.4 Sociological Conclusions
The sociological model has several implications for government pro-
jects which attempt to introduce innovations, in terms of understanding the
process of developing participation and adoption among farmers, and the
role that the project agencies and informal social relationships play in
this process.
To a large extent, the model indicates that agencies played their
principal role in simply informing farmers about the project. Fanners who
knew about the project tended to develop a favorable attitude toward it.
This is important, because it means that it is not necessary to coerce
farmers to participate in projects of this nature.
Another important implication of the model is indicated by the size-
able affect of ETHGRP at all three stages of the process. The model also
indicates that the Amish are more likely to have a higher level of
knowledge, greater favorability, and more participation, than the non-
Amish. It is recognized that the project and its subsidies provide special
circumstances for the adoption of environmental practices, and there are
several specific considerations worth noting.
(1) The Amish have a very integrated, effective communication system
within their community, so that if contacts are made with only a few Amish
STUDY RESULTS
-------
174
fanners, many of them will know about it in a very short time. This is
very different from the situation among the non-Amish, where farmers tend
to know fewer of their neighbors, have less frequent contact with them, and
would be less likely to inform others.
(2) The Amish might also have been more likely to have received
knowledge of the project early (in 1974) because their land was generally
poorer and had more erosion problems. Thus, they would have been more
likely to have been contacted by project personnel than non-Amish, who are
situated on better land, have larger farms and higher SES.
Farmland and its surrounding roads constitute over 590 million acres
of land in the United States, of which about 387 million acres are cropped
(Seneca et al., 1974). This farmland borders on virtually all major water
sources. Given this unmense amount of land, it is readily apparent that
agricultural pollution could affect all major waterways in this country.
The type of effort represented by the Black Creek Project provides a
way for area farmers to become involved in decision making. It gives them
an opportunity to discuss agricultural options open to them with their
neighbors. Farmers who share their opinions on certain issues can work to-
gether to help speed solutions to their problems.
The project also provides farmers with an opportunity to work together
to help solve a problem of national significance. By participating in this
project they can make an effective contribution to the reduction of pollu-
tion, while only slightly reducing their profits from farming. This is
particularly true in the present case due to the existence of the subsidies
for those adopting the innovations.
Because of this, the outlook for continuation of innovations by area
farmers after withdrawal of cost-sharing is good. Furthermore, the gen-
erally favorable reaction reported by the farmers to the project should
serve as a positive indication that many farmers are quite willing to ac-
cept and use technological innovations if they are made available and their
benefits are carefully explained. The high level of participation in the
project among the Amish is a particularly good sign in this respect.
Although the long term effects of the Black Creek Project on the
farmer's attitudes and behavior will not be known for several years, the
conclusion that the existence of the project led to a voluntary alteration
of behavior and attitudes by farmers in the area is inescapable. This is
consistent with much of the research using behavior modification, which has
found that one way to modify behavior is to offer explicit rewards in early
stages, but that in many cases after a sufficient period of time the re-
wards for the modified behavior become implicit, so that the behavior will
continue even after the explicit rewards are no longer provided.
It should also be noted that this research can be used to illustrate
the major role which technological innovations can play in generating so-
cial development in rural areas. By initiating community meetings and en-
couraging participation in decision-making the existence of a project such
as the Black Creek Project is also likely to lead to an increase an indivi-
duals' involvement in the affairs of the local community.
STUDY RESULTS
-------
175
4.8.5 Economic Conclusions
Analysis of the results show that far less than proportional amounts
of sediment come from the relatively flat, less erosive, crop land. Based
on the unconstrained optimal solutions to the hypothetical farms, only 31
percent of the soil loss from modern crop production came from 68 percent
of the total acres. Therefore, primary emphasis of controlling soil loss
should be placed on those acres with steeper slopes where, according to the
analysis, 69 percent of the watershed's soil loss from modern crop produc-
tion came from only 32 percent of the acres.
At the unconstrained optimal solutions, farmers prepared corn and soy-
bean land in the fall with the moldboard plow. One-third of the acreage
was put into corn production while two-thirds went into soybeans. As the
soil loss constraints on the steeper land slopes became more restrictive,
tillage practices tended to shift from moldboard preparation to increasing-
ly greater amounts of chisel preparation (see Table 83).
TABLE 83
Net Revenue, Returns; Slopes Greater Than Two Percent
Farm
Size
(Acres)
580
580
580
580
580
370
370
370
370
370
Soil
Loss
Tons/Acre
12. 8a
10
8
6
4
12. 8a
10
8
6
4
Net
Revenue
Per Acre
$177.62
175.55
172.90
170.25
166.34
179.89
177.77
175.12
172.47
166.34
Acres of
Corn
Preparation
Moldboard Chisel
197
0
0
0
123
0
0
0
197
197
197
342
123
123
123
230
Acres of Soybean
Preparation
Moldboard Chisel
383
325 58
185 198
45 338
238
247
208 39
119 128
30 217
0 140
a Soil loss is unconstrained.
Results of restricting the use of moldboard plow preparation indicate
the same crop production patterns as in the unconstrained optimal solu-
tions. However, corn and soybean preparation was performed with a chisel
plow on all of the 370 acres and most of the 580 acres. Double disking was
used for some soybean preparation on the 580 acre farm.
Overall the larger farm seemed to show greater flexibility in comply-
ing with the policy constraints on fall preparation and the use of mold-
board preparation than the smaller farm. At least the economic impacts, in
terms of costs per acre when fall preparation was constrained and cost per
ton of reduced soil lost when moldboard tillage was constrained, were less
severe for the larger farm. A comparison of a soil loss constraint policy
with a crop management constraint policy revealed that the same amount of
soil loss could be achieved at a lower economic cost when soil loss con-
STUDY RESULTS
-------
176
straints are implemented. The flexibility to select the most appropriate
crop management combination for a particular farm and sediment problem al-
lows more timely operations and higher revenues. For example, achieving a
specified level of soil loss by restricting the use of a moldboard plow has
a higher cost per acre than achieving that level of soil loss without res-
tricting the use of the moldboard plow.
4.8.6 Policy Implications
The control of nonpoint agricultural water pollution in general, and
soil losses in particular, will have a significant impact on crop manage-
ment. According to this study, sediment reduction can more efficiently be
achieved by policies that constrain soil loss than by policies that_con-
strain crop management practices. However, the application of a uniform
soil loss constraint policy would more severely affect operators on more
erosive land. Since these soils are generally less productive anyway, the
disparity between these and better, less erosive land would only be in-
creased. Consequently, operators on the more erosive, steeper sloping land
would likely suffer greater economic impacts from (1) changes in crop
management that cause reduced yields or less profitable crop rotations, and
(2) a further decline in relative land values.
The exclusive reliance on crop management practices to meet the more
restrictive soil loss constraints would have negative economic conse-
quences. Other methods of control, such as structural remedies, jl.e. ter-
races, holding ponds, and field borders, should be assessed to supplement
changes in crop management. However, even with alternative control
methods, soil loss reduction on more erosive land would have to be subsi-
dized for these operations to remain at a competitive profit level with
larger farms on flat land.
References
1. Agricultural Prices, November 30, 1976, p.4.
2. Allen County Soil and Water Conservation District, Environmental Impact
of Land Use on Water Quality (A Work Plan) Black Creek Study Area,
ApriT7~1973.
3. Brink, Lars, Bruce McCarl and D. Howard Doster,, Methods and Procedures
in the Purdue Crop Budget (Model B-9): An Administrator's Guide, Station
Bulletin No. 121, Department of Agricultural Economics, Agricultural Exper-
iment Station, Purdue University, West Lafayette, Indiana, March, 1976.
4. Doster, Howard and Bruce McCarl, Purdue Crop Budget Model B-93_ Depart-
ment of Agricultural Economics, Purdue University, West Lafayette, Indiana,
1976.
5. Griffith, D.R., J.V. Mannering, and Richey, "Energy Requirements and
Areas of Adaptation in Eight Tillage-Planting Systems for Corn," paper
presented to the Engineering and Agriculture Conference on Energy, St.
Louis, Missouri, 1976.
STUDY RESULTS
-------
177
6. Mannering, J.V., C.B. Johnson, and R.Z. Wheaton, "Conservation Tillage
Effects on Crop Production and Sediment Yield," Paper presented to the
American Society of Agricultural Engineers, Chicago, Illinois, December
14-17, 1976.
7. Ohio Cooperative Extension Service, Ohio Erosion Control and Sediment
Pollution Abatement Guide Bulletin 594, Columbus, Ohio, 1975.
8. Seneca, Joseph J. and Michael K. Tanssig, Environmental Economics,
Prentice-Hall, Inc., Englewood Cliffs, New Jersey, 1974.
9. U.S. Department of Agriculture, Farm Planning and Financial Management,
Cooperative Extension Service, Purdue University, West Lafayette, Indiana,
Revised 1975.
10. Wischmeier, Walter H. and Dwight D. Smith, Predicting Rainfall Erosion
Losses from Cropland East of the Rocky Mountains, U.S. Department of Agri-
culture, Agriculture Handbook No. 282, Washington, D.C., 1972 reprint.
4.9 MODELING
4.9.1 ANSWERS
Planning is a process of evaluating the relative effectiveness and
costs of alternative courses of action designed to attain stated goals.
One of the most effective methodologies appropriate for planning non-point
source pollution control programs is known as simulation. This approach
utilizes a mathematical model of the system under consideration to evaluate
the consequences of alternative control strategies.
The validity of pollution control programs developed using the
simulation approach is entirely dependent upon the accuracy with which the
model can predict the effectiveness of the control strategies being
considered. Therefore, a model capable of accurately simulating a wide
range of alternatives is essential if this approach to planning is to be
successful. One major effort of the Black Creek Project was the
development of a comprehensive model designed to characterize the behavior
of natural watersheds during storm events and to evaluate the impact of
alternative land use practices on water quality.
In selecting a structure for an agricultural non-point source
pollution model, the pollution control planning process was characterized
as a two-stage effort. The first step, as evidenced by the current efforts
in 208 planning agencies, was considered to be an assessment phase. The
budgetary and time constraints and the magnitude of the task precluded any
other approach. This assessment phase is a process of determining the
relative magnitude of the pollution problems and identifying the primary
geographic regions requiring urgent attention, i.e. one of setting
priorities.
The second stage of non-point source planning can be defined as an
implementation phase. During this phase, planning decisions will be
required concerning expenditures of funds to achieve effective treatments.
This involves the application of specific treatments to individual, small
land units. In order to be economically viable and politically acceptable,
these planning decisions must be accurate and applied with great
STUDY RESULTS
-------
178
selectivity to only those areas which will yield significant improvements
in water quality.
With these considerations in mind, project personnel decided to
develop a model designed to be of general usefulness to the second phase of
208 planning. The result was a distributed parameter model called ANSWERS,
Areal Non-point Source Watershed Environment Response Simulation.
A distributed parameter watershed model is one which incorporates the
influences of the spatial variation of controlling parameters, e^.g_.
topography, soil types, vegetation, etc., in a manner internal to its
computational algorithms. In contrast, the more commonly employed lumped
model approach is one which incorporates, to whatever degree they are not
ignored, these effects by an a priori analysis on a case by case basis. In
other words, the lumped approach uses some type of averaging technique to
generate an "effective" coefficient(s) for characterizing the influence of
specific non-uniform distributions of each parameter. The influence of
this distribution is then represented by the "lumped" coefficients, and the
resulting model is treated as a mathematical transformation of input into
output, ^i.e., a "black box," for the subsequent simulation.
A primary advantage of a distributed parameter analysis is its
potential for providing a more accurate simulation of natural catchment
behavior. The term potential is used because increased accuracy is by no
means a direct consequence of using a distributed analysis; rather, it is
realized only if the model is designed to take advantage of removing the
constraints imposed by lumped parameters.
Lumped models almost invariably employ some weighting function to
account for the spatial variability of watershed parameters, such as soil
type, cover and slope steepness. Such weighting functions, regardless of
how elaborate, are applied to the catchment prior to modelling runoff. Tnis
constrains the parameter values to be independent of the magnitude and
temporal distribution of the storm event. Such a constraint is valid only
for linear systems. Thus, the assumptions and limitations of behaving, at
least to some degree, as a linear system are subtly imposed on lumped
models.
Another linear system assumption implicit to almost all weighting
functions used with lumped models results from ignoring the influence of
geographic placement of spatially varying factors within the catchment
boundaries. The magnitude of error associated with such approximations has
been demonstrated by Huggins, et al. (1973).
A second major advantage of a distributed model is its inherent
ability to simultaneously model conditions at all points within the
watershed. Tnis readily permits the simulation of processes that change
both spatially and temporally throughout the catchment. The accuracy with
which interacting processes can be modelled is thereby increased. In
addition, a great deal more information about the simulated process is
available to planners.
Finally, distributed models greatly facilitate incorporation of
relationships developed from small scale "plot-size" studies to yield
predictions on a watershed scale. It is much easier to formulate the
individual processes being modelled as independent equations applicable at
STUDY RESULTS
-------
179
a point, letting the subsequent model integration process incorporate
effects of spatial and temporal variability, than to develop an elaborate
weighting function for each process. This approach also directly accounts
for process interactions that would otherwise be ignored or require complex
modifications of weighting functions.
The development of a comprehensive watershed model such as ANSWERS can
be subdivided into a number of steps or phases. A useful subdivision
consists of three stages to model building: conceptualization, definition
of quantitative (mathematical) component relationships, and verification.
Model concepts determine what is required in the way of component
relationships and define the ultimate accuracy limits of the model. The
process of developing component relationships for the various parameters
included in a model is generally an iterative one. In other words, the
verification process is used to modify and refine the component
relationships of the model. Ultimately, the accuracy and generality of the
resulting model are limited by the adequacy of model concepts and the
availability of field data to validate various component relationships
incorporated into the model.
Initial concepts of a distributed parameter watershed model were
reported by Huggins and Monke (1966) and by Huggins, et al. (1973). That
work was restricted to predicting runoff hydrographs, although the unique
applicability of the approach as a framework for non-point source pollution
models was noted, Huggins, Podmore and Hood (1976). Beasley (1977) greatly
expanded this basic hydrologic model to incorporate tile inflow, channel
flow routing and the most important aspect of non-point source pollution
from cropland, soil erosion. Henceforth it became known as the ANSWERS
model. Beasley utilized the GASP simulation language, with its inherent
implicit solution algorithms, to code the original version of the ANSWERS
model. Dr. J. R. Burney developed a much improved method of incorporating
channel flow effects, included the ability to study spatially and/or
temporally variable rainfall and refined the infiltration and soil erosion
relationships used in the model. He also recoded the model using FORTRAN
in order to obtain a major reduction in the amount of computer memory
required and some improvement in execution speed. This version of the
model employs an explicit integration algorithm. The detailed structure of
the ANSWERS model, developed under the Black Creek Project, is presented
below.
4.9.1.1 MODEL CONCEPTS
ANSWERS is a deterministic model based upon the fundamental hypotheses
that:
"At every point within a catchment a functional relationship exists
between the rate of surface runoff and those hydrologic parameters
which influence runoff, e.g., rainfall intensity, infiltration,
topography, soil type, etc. Furthermore, these surface runoff rates
can be utilized in conjunction with appropriate component
relationships as the basis for modelling other transport-related
phenomenon such as soil erosion and chemical movement within that
watershed."
An important feature of the above hypothesis is its applicability on a
"point" basis. In order to apply this approach on a practical scale, the
point concept is relaxed to refer instead to a watershed "element". An
element is defined to be an area within which all hydrologically
STUDY RESULTS
-------
180
significant parameters are uniform. Of course, this process of going from
a point to an elemental area could be extended indefinitely until one
assumed the entire watershed was composed of a single element with
"averaged" parameter values, i.e., a lumped model. The actual geometric
size of an element is not critical because there is no finite-sized area
within which some degree of variation in one or more parameters does not
exist. The crucial concept is that an element must be sufficiently small
that arbitrary changes of parameter values for a single element have a
negligible influence upon the response of the entire catchment.
A watershed to be modelled is assumed to be composed of elements,
square in shape for computational convenience, with all hydrologic
parameters being uniform within each element. Parameter values are allowed
to vary in an unrestricted manner between elements; thus, any degree of
spatial variability within the watershed is easily represented. Individual
elements collectively act as a composite system because of supplied
topographic data for each element delineating flow directions in a manner
consistent with the topography of the watershed being modelled. Element
interaction occurs because surface flow (overland and channel), flow in
tile lines and groundwater flow from each element becomes inflow to its
adjacent elements. In all other respects, the elements are hydrologically
independent.
Mathematically, individual elemental responses are combined into a
watershed system response by integration of the continuity equation:
1 - ° af <"
where:
I = inflow rate to an element from rainfall and adjacent
elements,
Q = outflow rate,
S = volume of water stored in an element,
t = time.
This equation may be solved when it is combined with a stage-discharge
relationship, e_.g_. Manning's equation.
For the watershed model to be complete, it is necessary to incorporate
component relationships for the remaining hydrologically significant
processes which influence runoff and any related pollutants. Specific
forms of relationships included have no effect whatsoever on the
integration algorithm or distributed model concepts. However,
relationships chosen have a marked impact upon the accuracy with which the
model can characterize real watershed behavior. The significant
consequence is that to substitute one component relationship for another
when subsequent research on component processes develops improved
relationships is a relatively trivial task.
Hydrologic processes for which quantifying component relationships
must be developed are shown qualitatively in Figure 21. After rainfall
begins, some is intercepted by the vegetal cover until such time as the
interception storage potential is met. When the rainfall rate exceeds the
interception rate, infiltration into the soil begins. Since the
STUDY RESULTS
-------
181
infiltration rate decreases in an exponential manner as the soil water
storage increases, a point may be reached when the rainfall rate exceeds
the combined infiltration and interception rates. When this occurs, water
begins to stand on the surface in micro-depressions.
Once surface retention exceeds the capacity of the micro-depressions,
runoff begins. Water accumulated in order to produce flow across the
surface is surface detention. Subsurface drainage begins when the pressure
potential of the groundwater surrounding a tile drain exceeds atmospheric
potential. A steady-state infiltration rate may be reached if the duration
and intensity of the rainfall event are sufficiently large.
When rainfall ceases, the surface detention storage begins to
dissipate until surface runoff ceases altogether. However, infiltration
continues until depressional water is no longer available. Subsurface
drainage continues as long as there is excess soil water surrounding the
drains. The long recession curve on the outflow hydrograph, typical of
tile drained areas, is then produced.
Soil detachment, transport, and deposition are very closely related to
concurrent hydrologic processes in a watershed. Detachment and transport
can both be accomplished by either raindrop impact or overland flow.
Detachment by rainfall occurs throughout a storm even though overland flow
may not occur. Thus, most of the soil particles detached prior to flow
initiation are deposited and to some extent, reattached. Detachment of
soil particles by overland flow occurs when the shear stress at the surface
is sufficient to overcome the gravitational and cohesive forces of the
particles. Whether or not a detached soil particle moves, however, depends
upon the sediment load in the flow and its capacity for sediment transport.
The transport of chemical pollutants irom a land area is also highly
related to the hydrologic behavior of a catchment and, for certain
chemicals such as phosphorus and cadmium, the soil erosion that occurs.
These processes can readily be incorporated into a distributed model by
developing component relationships for individual elements.
Natural rainfall events do not exhibit the steady appearance shown in
Figure 21. Furthermore, uniformity of coverage over a watershed will
usually vary during an event. In addition, hydrologic responses of various
areas within a watershed may vary greatly. Hence, the resultant hydrograph
for the entire watershed will contain at least some of the effects of all
of these highly complex, unsteady, non-uniform interactions. For these
reasons, a distributed model must be designed and utilized as a means of
describing and quantifying these processes.
Every hydrologic or erosion component of the ANSWERS model is
expressed as a rate. Thus, infiltration and interception rates can both be
subtracted from the rainfall rate to provide the excess rainfall rate used
in satisfying surface retention and detention. The difference between the
inflow and outflow rates is integrated to provide a volume. When divided
by the elemental area, this yields the average depth of water over an
element. Tne depth, in turn, is used to determine an outflow rate by
applying a runoff function that accounts for both runoff and detention.
The thrust of the project's modelling effort was the development of a
new approach that could be used as a planning tool. Father than devote
STUDY RESULTS
-------
182
UJ
o:
ct
UJ
Q_
LU
h-
<
o:
_ ' Figure 21. Surface and Subsurface Water Movement Relationships
large amounts of time and money toward basic research on hydrologic
component relationships, the focus was on integrating existing, accepted
STUDY RESULTS
-------
183
relationships into a simulation 'structure that had a high degree of
versatility and reliability.
The specific component relationships selected for the current version
of the ANSWERS model are outlined below. All were selected so that
modularity was preserved. Modification or replacement of component
relationships such as infiltration or erosion does not affect the
computational algorithms of other components. In other words, the
component relationships are sufficiently independent from each other that
user-supplied subroutines may easily be substituted for those originally
used in the ANSWERS model. This framework also permits users to append
additional component relationships to simulate other processes important to
specific applications.
4.9.1.2 COMPONENT RELATIONSHIPS
4.9.1.2.1 Flow Characterization
Within its topographic boundary, a catchment is divided into an
irregular matrix of square elements, as shown in Figure 22. Each element
acts as an overland flow plane having a fixed slope and direction of
steepest descent. Channel flow is analyzed by a separate pattern of
channel elements (referred to hereafter as channel segments) , which
underlie the grid of overland flow elements. Elements designated to have
channel flow may, therefore, be viewed as dual elements. These elements act
as ordinary overland flow elements, with the exception that all overland
flow out of that element goes into its "shadow" channel segment. Flow out
of a channel segment goes into the next downslope channel segment. This
downslope channel segment will also receive flow from any other channel
segments which flow into it and from its own overland flow element.
Overland and tile outflow from an element is assumed to be
proportioned as separate surface and tile line inflow into adjacent row and
column elements according to the direction of the slope of the element.
The slope direction is designated on input as the angular degrees counter-
clockwise from the positive horizontal (row) axis. For the example shown
in Figure 23 below, slope direction is in the fourth quadrant. The slope
direction angle equals 270 degrees plus angle "a".
The fraction of outflow going into the adjacent row element, RFL, is:
|a)_ if. a ± 450f and
RFL = tana)_
RPT - i tan(90-a) . o o
Ri-L - 1 - - - - !f: 45^ < a < go"
with the remaining outflow going into the adjacent column element. Since
everything, including surface slope, within an element is assumed to be
constant, this method of partitioning overland flow seems intuitively
obvious. Such is not the case for tile flow. In general, records are
seldom available which delineate the layout of tile systems. However,
limitations on feasible installation depths mean that tile slopes must
with only temporary deviations, follow the general topography. Therefore
the use of an element's slope properties would seem to be a close
STUDY RESULTS
-------
184
Figure 22. Watershed Divided into Element with Channel Elements Shaded
A2 1
Al + A2
Q
Direction of
steepest slope
Figure 23. Partitioning of overland flow.
STUDY RESULTS
-------
185
approximation which eliminates the need for additional input information.
The flow relationship utilized in conjunction with the continuity
equation to perform the overland flow routing is Manning's equation. The
hydraulic radius is assumed equal to the average detention depth in an
element. Detention depth is calculated as the total volume"of surface
water in an element, minus the retention volume, divided by the area of the
element. This implies that the entire specified retention volume of an
element be filled before any water becomes available for surface runoff.
Although the channel system is unrestricted in direction and
branching, it is necessary that it be continuous and that each element
contain only one channel segment. To achieve greater definition it is
necessary to assume a smaller element size for all elements, with a
consequent increase in core storage and execution time.
As all flow leaving the overland section of a dual element is
constrained to enter the channel segment, the slope direction of a dual
element is irrelevant in terms of division of outflow. However, the slope
direction is of vital importance in establishing channel continuity. All
outflow from a channel segment enters an adjacent channel segment located
in one of the eight directions of the cardinal axes or the diagonals. The
specific direction is determined by the slope direction assigned to the
dual element. This slope direction must be within 22 degrees of the
direction of the intended receiving channel segment, and in view of the
above it is recommended that only 45 degree increments in slope direction
be specified for dual elements.
It is permissible to specify the slope, width and Manning's roughness
coefficient for each channel segment independent of the corresponding
values for its overland flow element. Typically, rather than having a
unique set of values for each channel segment, they are grouped into
reaches with similar coefficients. Manning's equation is again used as the
flow relationship required, in conjunction with the continuity equation, to
perform the routing calculations.
4.9.1.2.2 Interception
Interception encompasses the total volume of water removed from the
incoming rainfall by raindrop contact with and retention by the vegetal
canopy, and by evaporation during the storm event. The water retained by
the vegetation, i.e., interception storage, is held primarily by surface
tension forces. This portion of the total interception is quickly
satisfied, particularly in more intense storms. Since a dense vegetal
cover can expose an immense surface area to rainfall, the amount of
moisture evaporating from this exposed area can be considerable, even
during high humidity conditions of a storm. However, for storm intensities
of primary interest from the standpoint of non-point pollution from
cropland, interception is a relatively minor hydrologic component. In
order to reduce simulation costs, interception was assumed to be uniform in
rate and total volume over each type of vegetation.
Horton (1919) did a great deal of work in the area of estimating the
amount of and mechanisms controlling interception. He studied the water
intercepted by several species of trees as well as some economically
important crops. Values of from 0.5 millimeter to 1.8 millimeters of
interception storage volume were found to exist for trees and nearly as
STUDY RESULTS
-------
186
much for well developed crops. Horton recommended the following
relationship as a means of describing the interception for a particular
event:
V = A(B + CP)h
where:
V = interception volume per unit area mm,
P = precipitation mm,
h = height of crop m,
A, B and C are constants depending on the type of vegetal
cover.
The constants referred to above are listed for several crops in Table 84.
The B coefficient is a measure of the interception potential of the crop.
The C coefficient attempts to describe evaporation potential.
TABLE 84
Interception Constants Recommended by Horton
Crop
Oats
Corn
Grass
Pasture and meadow
Wheat, rye, and barley
Beans, potatoes, and cabbage
A
3.3
h/3
7
3.3
3.3
.8h
B
.18
.13
.13
.13
.13
.5
C
.07
.005
.08
.08
.05
.15
4.9.1.2.3 Retention
Surface retention is a component that can have a pronounced effect on
surface runoff and drainage characteristics of a watershed. Rough ground
can store large amounts of water. Huggins and Monke (1966), using several
field surfaces, developed a dimensionally homogeneous relationship
describing the surface retention storage potential of a surface as a
function of the water depth in the zone of micro-relief. The resulting
relationship is of the form:
s_ = a/*L.r (5)
su
where:
s = volume of stored water,
su = hu times area of element,
h = height above datum,
hu = height of maximum micro-relief,
A and B are characteristic parameters.
STUDY RESULTS
-------
187
In order for this equation form to yield a correct surface area as h
approaches hu it is necessary that the product of A and B equal 1. Values
of hu, A, and B for several field surface conditions, appear in Table 85.
TABLE 85
Typical Surface Storage Coefficients
Condition
Plowed Ground
Spring smooth
Spring normal
Spring rough
Fall smooth
Fall normal
Fall rough
Disked and Harrowed
Very smooth
Rather rough
Corn stubble
hu mm
100
130
130
60
70
130
30
60
110
A
.53
.48
.59
.37
.33
.45
.42
.43
.59
B
1.9
2.1
1.7
2.7
3.0
2.2
2.4
2.3
1.7
4.9.1.2.4 Subsurface Waters
Currently, three subcomponents of subsurface water movement are
incorporated into the ANSWERS model: infiltration, tile drainage and base
flow. The effort devoted to modelling each roughly corresponds to their
relative importance in simulating non-point pollution from cropland. Thus,
infiltration processes are simulated in greater detail than tile flow while
base flow is modelled only crudely.
Infiltration can greatly affect the hydrologic response of a
watershed. Although many years of research have been conducted on
infiltration phenomena, there is still no universally accepted method for
describing infiltration on a watershed scale. The method chosen for the
ANSWERS model was developed by Hoitan (1961) and Overton (1964). In a
dimensionally homogeneous form it can be expressed as:
/s - P\P
f fc + A(Sr) (6)
where:
f = infiltration rate at a particular time,
fc = final or steady state infiltration rate,
A = maximum possible infiltration rate,
S = storage potential of soil within the control zone
(total porosity minus antecedent soil moisture),
F = total volume of water infiltrated,
Tp = total porosity within the "control depth",
P = diraensionless coefficient relating the rate of
decrease in infiltration rate with increasing soil moisture
content.
STUDY RESULTS
-------
188
This form uses the soil water content, rather than time, as the independent
variable. It is computationally advantageous when modelling with rainfall
intensities less than the infiltration capacity and for simulating recovery
due to drainage during brief periods without rainfall.
During periods of zero rainfall rate any infiltration which occurs
must be supplied by the volume of water stored as either retention or
detention. Since the surface of an element is seldom entirely inundated,
the computed infiltration capacity is reduced in direct proportion to the
percent of the soil surface not submerged. Tne area submerged for any
given volume of water in an element can be computed by differentiating the
relationship developed above to characterize surface retention.
Hoitan's equation requires six infiltration parameters to be specified
for a given soil type: total porosity, field capacity, depth of the control
zone, steady-state infiltration rate (fc), and the two unsteady-state
coefficients (A and P). Data from both large plot-sized simulated rainfall
tests, Skaggs, et al. (1969) and field rainulator tests conducted in
cooperation with USDA-ARS as a part of the Black Creek Project have been
used to estimate parameter values. All of these tests indicated that
surface crusting conditions have a major impact on the observed
infiltration relationship. Experience with Holtan's equation has indicated
the influence of crusting can be modelled by adjustment of the specified
depth of the control zone. Crusting requires the use of a shallower
control zone.
According to Holtan's conceptualization of the infiltration process, a
"control zone" depth of soil determines the infiltration rate at the
surface. He defined the depth of this control zone as the shallower of the
depth to an impeding soil layer or that required for the hydraulic gradient
to"reach unity. Extending this same concept somewhat, the ANSWERS model
maintains an accounting of water that leaves this control zone.
The rate of water movement from the control zone is a function of the
moisture content of that zone. The two conditions which can exist are
handled according the following rules:
(1) when the moisture content of the control zone is less than field
capacity, no water moves from this zone,
(2) when the control zone moisture exceeds field capacity, the water
moves-from this zone according to the equation:
(7)
where:
Dr = drainage rate of water from the control zone,
Piv = volume of water that could still potentially
infiltrate into the control zone,
Gwc = gravitational water capacity of the control zone,
i^.e. total porosity minus field capacity,
N ~~ = parameter controlling rate of movement, usually
assumed equal to 3.
This relationship satisfies the continuity requirement that at saturation,
STUDY RESULTS
-------
189
when Piv = 0, the drainage rate from the control zone equals the steady
state infiltration rate, fc. In addition, it exhibits intuitively
desirable properties of decreasing moisture movement rapidly as the soil
dries from saturation and of asymptotically approaching a zero drainage
rate.
Water leaving the control zone contributes to tile drainage, if the
element is tiled, or to base flow. In both cases the water is assumed to
re-emerge into the channel segments. Water moves from the infiltration
control zone into the "pools" available for tile and/or base flow at a rate
equal to Dr.
Individual elements may selectively be designated as being tile
drained. In addition to water coming from the control zone, tile inflow
may be occurring from adjacent tiled elements. The sum of these two rates
constitutes the rate of subsurface inflow into an element. Subsurface
water moves out the element's tile at this inflow rate up to a maximum
outflow rate equal to the tile drainage coefficient. Whenever the rate of
subsurface inflow to an element exceeds its drainage coefficient, that
excess water is diverted to baseflow storage. Elements which are not tiled
have a drainage coefficient of zero.
Subsurface water entering an element at rates in excess of its
drainage coefficient is stored in a single "pool". To simulate base flow
it is released directly into channel segments at a rate proportional to the
volume of water in storage. The proportionality constant, referred to as
the base flow release fraction, is specified in the input data file. Water
is released at an equal rate to each channel segment. For small catchments
having no defined channels, only overland flow will appear at the outlet.
Base flow will be non-zero only for watersheds with channel segments.
4.9.1.2.5 Sediment Detachment and Movement
Soil erosion as it relates to non-point source pollution can be viewed
as two separate processes, detachment of particles from the soil mass and
transport of these particles into the streams and lakes. Detachment of
either primary soil particles or aggregates can result from either rainfall
or flowing water. These same factors can cause detached particles to be
transported to the water supply network. Thus, there are four processes
for which quantifying relationships must be developed.
The detachment of soil particles by water is accomplished by two
STUDY RESULTS
-------
190
processes. The first involves dislodging soil as a result of the kinetic
energy of rainfall. Rainfall is the major detachment process on relatively
flat watersheds. The second process involves the separation of particles
from the soil mass by shear and lift forces generated by overland flow.
Detachment of soil particles by raindrop impact is calculated using
the relationship described by Meyer and Wischmeier (1969):
DR = .0270K-Ai-I2 (8)
where:
DR = rainfall impact detachment rate, kg/min,
C = cropping and management factor (from Universal Soil Loss
Equation), Wischmeier and Smith (1965),
K = soil erosivity factor (from Universal Soil Loss
Equation), T/A/EIunit,
Ai = area increment, sq m,
I = rainfall intensity, mm/min.
The detachment of soil particles by overland flow was described by
Meyer and Wischmeier (1969) and modified by Foster (1976) as follows:
DF » ,018-C-K-Ai-S-Q (9)
where:
DF = overland flow detachment rate, kg/min,
S = slope steepness,
Q = flow rate per unit width, sq m/min.
Once a soil particle has been detached, sufficient energy must be
available to transport it or the particle will be deposited. The transport
of sediment by overland flow is self-regulating, soil particle detachment
by_ overland flow does not occur unless there is excess energy available in
addition to the amount required to transport suspended sediments. However,
detachment by rainfall impact often occurs when there is little or no flow
available for transport.
After a literature study which included Yalin (1963), Meyer and
Wischmeier (1969), Foster and Meyer (1972), and Curtis (1976), as well as
an inspection of soils data, a relationship for particle transport in
overland flow was chosen as shown in Figure 24.
The two portions of the curve generally represent the laminar and turbulent
flow regions. Equations and their region of application are:
STUDY RESULTS
-------
191
K31-
LAMINAR FLOW
T=I46 S Q-5
TURBULENT FLOW
T= 14600 SO2
IO
q = rnVmin
Figure 24. Transport Relationship Used in ANSWERS Model
T = 146 (S/Q)
T = 14600(SQ)
if Q < .74 sq m, and
if Q > .74
(11)
where: T = potential transport rate of sediment, kg/min/m.
Table 86 lists the four particle sizes that were considered as
representative in making the transport calculations.
The erosion portion of the ANSWERS model was simplified further by the
following assumptions:
1. Subsurface or tile drainage produces no sediment. (Data indicate
around two percent of the average annual loading originated from
subsurface systems on the Black Creek Watershed).
STUDY RESULTS
-------
192
TABLE 86
Sediment Particle Characteristics
Particle
Group
I
II
III
IV
Mean
Diameter-mm
.01
.05
.15
.30
Specific
Gravity
2.65
2.65
2.00
2.00
2. Sediment detached at one point and deposited at another is
reattached to the soil surface.
3. Re-detachment of sediment requires the same amount of energy as
required for the original detachment.
4. For channel segments rainfall detachment is assumed to be zero and
only deposited sediment is made available for flow detachment, ^.e.
original channel linings are not erodible.
Although these assumptions were made primarily to reduce the computational
cost of using the model, some were also required because little or no data
were available in the literature to quantify the particular process.
After consideration of the relative magnitude of the four
detachment/transport processes the transport of soil particles by rainfall
was assumed negligible.
Combining the above equations and assumptions gives a composite soil
movement model wherein soil particles are dislodged from the soil mass by
both rainfall and flowing water. Detached solids then become available for
transport by overland flow. Within an element the material available for
transport is the combination of that detached within the element and that
which enters with inflow from adjacent elements.
Once the available detached sediment within an element is known, the
transport capacity is computed. If it is insufficient to carry the
available material, the excess is deposited in the element. The overall
accounting relationship for this process is the differential form of the
continuity equation as applied above to water flow. Sediment carried out
of an element is apportioned between adjacent elements in direct relation
to overland flow.
4.9.1.3 USER CONSIDERATIONS
The selection of a specific model for use as a planning tool is a
difficult process complicated by the large number of different models
developed in recent years. The most appropriate model to use will depend
most of all on the intended application, the type of input data available
and the suitability of the output information generated. The accuracy of a
model's simulation should also be an important consideration, but
unfortunately this is very difficult to judge. Primarily this must be done
intuitively by a through study of the relationships incorporated into a
STUDY RESULTS
-------
193
model.
ANSWERS represents an attempt to develop a comprehensive model
intended to be of use in quantitatively evaluating the importance of non-
point source pollution in an ungaged catchment and in determining the
relative effectiveness of alternative corrective measures. One of its
primary strengths arises from the use of a distributed parameter type of
analysis which inherently accounts for the importance of the areal
distribution of the many relevant factors. The distributed analysis
provides a very complete characterization of hydrologic response and
erosion/deposition occurring at all points in the watershed throughout a
storm event.
In its present form sediment is the only pollutant for which ANSWERS
gives a direct numerical estimate. This is the result of time limitations
and the importance of sediment to agricultural non-point pollution. Of
course any pollutant which is closely correlated with sediment loss is also
easily predicted once erosion losses are known. The modular structure of
the model together with the advantages of a distributed modelling concept
make the addition of any other pollutants for which component relationships
can be developed rather easy.
It was anticipated that the primary use of ANSWERS would be to
simulate isolated storm events. This decision was based on the general
consensus of erosion research, confirmed by data from Black Creek, that the
majority of annual sediment losses result for only a few of the larger
storms which occur in a year. In order to use the model for continuous
simulation it would be necessary to add component relationships for
evapotranspiration and a different integration algorithm for use during
periods with no surface runoff.
Selection of an appropriate element size to use with the ANSWERS model
is a significant decision from the standpoint of the cost of using the
model. Obviously, the larger the element size the lower the cost of input
data file preparation (for an area which has not previously been modelled)
and the computer costs. Therefore, it is desirable to chose a size which
is as large a possible without serious degradation of the accuracy of the
subsequent simulation. The most suitable size will depend upon the
accuracy requirements of a specific application and on the degree of non-
uniformity of topography and soils in the watershed. For the Black Creek
Project an element size smaller than considered optimum for ordinary use
was deliberately chosen in order to evaluate the influence of element size
as a parameter. From that experience it is recommended that an element
size in the range of 2-5 ha will be satisfactory for most applications in
areas with a variability comparable to Black Creek.
The primary effort required in preparing a data base to use the
ANSWERS model on a particular watershed concerns characterizing the
topography and soil type of each element. Where computer compatible data
files with such information are not available, U.S. Geological Survey
Topographic Maps and County Soil Survey Maps must be used. While these
sources of information are quite adequate, the effort required to digitize
the information is not trivial.
A frequent complaint voiced by potential users of comprehensive
watershed models is the large volume of data required concerning watershed
STUDY RESULTS
-------
194
characteristics. This comment is often directed at distributed parameter
models because a large data base is usually required. However, one of the
fundamental strengths of comprehensive models is their potential ability to
characterize the many processes for which input coefficients must be
specified. When data are not available to quantify some coefficients,
assumed values can be supplied. While one is never comfortable in such a
situation, it is easy to use a comprehensive model to evaluate the
sensitivity of the prediction to changes in assumed values.
In contrast, "simpler" models that require less input have
incorporated at creation time implicit assumptions concerning all variables
for which explicit numerical values are not demanded from the user.
Because these assumptions are implicit the user has neither control over
them nor any means of evaluating their importance. Thus, while it is
desirable to have available hard data to quantify all parameters of a
comprehensive model, it is better to assume values for a model than to
submit to the rigidity of implicit assumptions inherent with simpler
models.
It is anticipated that future versions of ANSWERS will contain default
values for all parameters so that completely inexperienced personnel would
not be required to make assumptions about initial parameter values in the
absence of hard data.
4.9.1.4 RESULTS
Verification of the accuracy with which a model such as ANSWERS can
simulate the behavior of natural watersheds is, in a strict sense,
impossible. This situation results from the number of degrees of freedom,
i.e. coefficient values, of the model. It is true that optimization
techniques could, given enough computer time, yield a set of coefficients
which would give an "accurate" simulation of almost any gaged event(s).
However, since it is not feasible to obtain field data to authenticate the
correctness of each of these optimized coefficients, such an effort would
not prove the model is accurate.
Fortunately, from the standpoint of application of a model for
planning analysis or design, one needs to simulate hypothetical future
events rather than accurately reproduce historical records. For such uses
massive amounts of data concerning antecedent conditions (with their
associated coefficient values) are irrelevant. Coefficient values for such
hypothetical situations must be determined from probabilistic relationships
rather than measured conditions. Therefore, massive data collection
efforts are not required for operational use of the model.
Despite the impossibility of absolute model verification, a real need
exists to provide some measure of the accuracy of a model's simulation
during its developmental period. Ultimately this comes down to comparing
its output with gaged data from specific events on natural watersheds.
Several complex storm events which occurred during 1975 and 1976 were
simulated for two gaged subcatchments of the Black Creek. These
subcatchments were 714 and 942 ha (1765 and 2328 A) in size. Beasley
(1977) gives a detailed discussion of these results. While they varied
somewhat from one storm to another and on the basis of evaluation criteria,
the results were generally within 30 percent of gaged values for all
STUDY RESULTS
-------
195
criteria.
From a pollution control planning standpoint the most relevant
question is "What needs to be done to achieve stated water quality
standards for a particular area?" The first step necessary to rationally
respond to such a question is a quantitative assessment of current
conditions. The strength of using a distributed model for this purpose is
best illustrated with an example.
Figures 25 and 26 illustrate some of the more applicable outputs
available from an ANSWERS simulation. They were produced by simulating the
behavior of a 714 ha subcatchment of the Black Creek Watershed using 1 ha
elements together with cropping, management and rainfall data for the
specified date.
Figure 25 gives output typical of a lumped model, runoff and sediment
concentration hydrographs at the watershed outlet. The simulated volume
was within 9 percent of the gaged amount (19 mm) and the total sediment
yield within 13 percent of the observed amount (325000 kg). These
quantities were produced from a storm with 64 mm of rainfall.
Any benefits of using a distributed parameter model instead of a
lumped one are not obvious from Figure 25. While it was claimed above that
the distributed approach makes possible a more accurate simulation, a
single example of close agreement between gaged and simulated results is
totally inadequate to judge to validity of such a claim. Furthermore, even
an extensive set of comparative simulations using ANSWERS and the best of
available lumped models could establish only the relative merit between the
two specific models. This would not offer conclusive evidence concerning
which of the two fundamental philosophies was superior.
Figure 26 clearly illustrates one major advantage of a distributed
model, more comprehensive output information. The "contour" lines on the
map result from connecting points within the watershed which experienced
equal soil detachment during the storm. Thus areas with closely spaced
lines correspond to regions of intense erosion. Such maps readily identify
those regions where control measures should first be considered.
Figure 26 indicates the bulk of the erosion occurred in the upland
portion of the watershed (the most steeply sloping region) with two small
areas experiencing a loss of more than 14000 kg/ha. The general location
of severe erosion could certainly have been predicted by any person
familiar with the area and reasonably knowledgeable of erosion processes.
The reason for modelling the area's behavior is not to identify the
location of problem areas, but to obtain a quantitative estimate of both
the amount of soil eroded and of its impact on water quality.
Maps similar to Figure 26 can be produced upon request for other
factors in addition to eroded soil. For example, a deposition map or
concentration hydrographs at any point within the watershed are also
available to a planner or designer.
It is worth emphasizing that ANSWERS' prediction of both field erosion
and sediment concentrations in flowing water is based upon
detachment/transport relationships used in conjunction with topographic
data about the watershed. That approach eliminates the need to use a
STUDY RESULTS
-------
196
.00-,
20.00-
UJ
I
cr
40.00-
CE
o;
60.00-1
60.00
50 .00 -
40.00 -
30.00 -
cc
o;
20 .00 -
10.00-
SED.
RUN.
540. 1080. 1620.
2160. 2700.
TIME - MINUTES
12.00
- 10.00
- 8.00
6.00
CD
CJ
- 4.00
- 2.00
.00
324Q. 378Q. 4320.
Figure 25. Runoff Rate vs Time in Minutes
STUDY RESULTS
-------
197
Contours indicate kg/ha.
Management Practice = 2
Figure 26. Sediment Loss
STUDY RESULTS
-------
198
"delivery ratio" concept which is difficult (some would say impossible) to
accurately quantify for the infinite variety of situations which occur in
natural catchments.
The second step in developing a program to meet water quality
standards for an area involves formulating and evaluating the effects of
alternative strategies. The detailed output of a distributed model is
ideally suited to this iterative step of planning.
Figure 27 shows what ANSWERS predicts would be the effect of a
hypothetical change in tillage practice for the entire catchment. The
actual tillage practice used on almost all grain fields in the watershed is
fall moldboard plowing. Figure 26 was generated with that tillage practice
specified. Figure 27 was generated under the assumption that moldboard
plowing would be replaced by fall chisel plowing for all cropland in the
catchment.
In contrast to the moldboard plow, the chisel plow leaves more crop
residue on the surface and a rougher micro-relief which tends to enhance
infiltration. Comparison of the two simulation results shows the impact of
such a management change on the resulting erosion pattern. Integration of
the sediment concentration hydrograph at the watershed's outlet indicates
only 1/3 of the sediment yield simulated for current management practices.
The purpose of this example is not to praise the benefits of a
specific tillage practice. In fact, the cost effectiveness, political
acceptability and long-term consequences (such as unforseen weed or pest
problems) of making significant management changes on a widespread scale
are often questionable. Such a course of action for this catchment would
never be recommended if the information from the simulation shown in Figure
26 is fully utilized. That figure clearly shows the major erosion occurs
from few a localized regions. It is on these specific areas that control
measures should first be evaluated.
Figure 28 shows an ANSWERS simulation of the effect of changing to
chisel plowing in only two of the highest erosion regions, those enclosed
by broken lines.
The total area of these two regions is only 32 ha of the watershed area of
714 ha. Integration of the outflow hydrographs indicates that changing
tillage on only these two small areas would achieve 40 percent of the
sediment yield reduction that could be achieved by changing the management
of the entire watershed.
It is the ability to be very site-specific concerning implementation
plans and to quantitatively demonstrate the effects of hypothetical control
measures on both upland regions and water quality conditions throughout the
watershed that makes a distributed parameter model such an effective
planning tool.
4.9.1.5 CONCLUSIONS
A comprehensive, non-point source watershed simulator, named ANSWERS,
has been developed. It was designed around a distributed parameter concept
with the intention of giving an accurate, comprehensive description of a
watershed's behavior during and immediately following storm events. The
purpose of this effort was to develop a model for use during the
STUDY RESULTS
-------
199
Contours indicate kg/ha.
Management Practice = .3
Figure 27. Upper Black Creek Watershed, Local Net Sediment boss
STUDY RESULTS
-------
200
The primary strengths of the distributed parameter approach are its
inherent accuracy, especially for ungaged situations such as evaluating the
influence of hypothetical changes to a watershed, and its detailed
description of the behavior of all interior points within a catchment1s
boundaries. The primary disadvantage of the approach is that the cost, for
both data preparation and computer time, to use it increases somewhat
proportionally to the area being simulated.
While it is not possible to give rigid limits as to the maximum area
that can be simulated without resorting to subdivision into smaller units,
current computer technology would suggest an upper limit in the range of 50
to 200 sq km. This is not considered to be a serious handicap to the
utility of the ANSWERS model to 208 implementation planning because its
intended use is on relatively small areas where detailed information is
needed. Other coarser procedures can be used on a "first pass" basis to
identify subregions of a large basin which have significant non-point
pollution problems and therefore warrant the use of a distributed model
prior to spending public funds to alleviate the problem. Solomon and Gupta
(1977), using a model structure identical to that employed in the ANSWERS
model, but with entirely different component relationships have developed a
distributed model intended for use on a river basin scale.
Extrapolation of unit cost data for various non-point pollution
control measures and the watershed simulation example discussed above lead
to the same conclusion. In order to be feasible, any non-point source
program must be highly site specific. Attempts to treat large areas with a
uniform set of practices or regulations will so dilute available funds that
the program will have little chance of being effective or publicly
accepted. ANSWERS offers a unique planning tool to help formulate site
specific non-point source programs for agricultural areas.
Annual loading data presented in earlier Black Creek reports and
elsewhere in this volume have shown values which, when presented on a per
hectare basis, are relatively low (on the order of 1000 kg/ha). This value
has been, by some individuals, contrasted with values for "tolerable soil
loss" used in conjunction with the Universal Soil Loss Equation to draw
incorrect implications concerning the potential effectiveness of BMP's to
improve water quality. Such comparisons are incorrect for several reasons.
First, it is almost always meaningless to use values (either for soil
losses or unit costs) based on averages per nectar over an entire watershed
to make extrapolations to another watershed, especially of a much larger
size. Secondly, the USLE tolerable soil loss values are for gross soil
erosion on upland areas, not net transport at the outlet of a watershed.
Because of this there is no reason to believe that those loss values,
developed with the sole criteria of preserving long-term productivity, are
directly applicable from the standpoint of water quality. Finally, the
ANSWERS example presented above clearly demonstrates the falsity of drawing
a conclusion to the effect that "because loadings (on a per hectare basis
over the watershed) are significantly below some xacceptable level1 it is
either unnecessary or maybe futile to consider a BMP." The model indicated
that for a specific storm with an average sediment yield of 460 kg/ha more
than 25 percent of the yield could have been prevented by a different
tillage practice on only 32 ha of the 714 ha catchment.
The ANSWERS model is operational and available to anyone interested in
using it. Because of the limited number of geographic regions on which it
STUDY RESULTS
-------
201
Figure 28. Effect of BMP's on Only Critical Areas
has been verified it must still be considered to be in a stage of
development rather than a fully operation model ready for widespread use.
STUDY RESULTS
-------
2Q2
The primary reason for this status is the lack of parameter values tested
for numerous regions. However, it should be emphasized that parameter
values required by the model have been designed to be ones with a known
physical analogy rather that "black box" coefficients requiring extensive
calibration in order to establish values. This means it is generally
possible to obtain parameter values which are regionally dependent from
available publications, e.g. Soil Survey Reports and publications
containing USLE soil erodibility and cropping factor coefficients. Because
of the availability of such data and because of the great flexibility of
the model, it is considered applicable to a broad range of conditions and
areas. Finally, if hard data are not available for a few parameters it is
possible to estimate their values relative to published values for similar
conditions and then use the model to evaluate the sensitivity of the
watershed to a plausible range for those values.
Two kinds of materials are available to assist users of the ANSWERS
model: a user's manual for persons interested it the direct utilization of
the current version of the model in a planning/design application 'and a
manual describing concepts and the inner structure of the model for persons
interested in changing component relations used or in adding relationships
to model additional processes. These materials are available from the
Department of Agricultural Engineering, Purdue University, W. Lafayette, IN
47907.
References
1. Beasley, D.B. 1977. ANSWERS: a mathematical model for simulating the
effects of land use and management on water quality. Ph. D. Thesis.
Purdue Univ.
2. Curtis, D.C. 1976. A deterministic urban storm water and sediment
discharge model. Proceedings of the National Symposium on Urban Hydrology,
Hydraulics, and Sediment Control. Univ. of Kentucky. Lexington, KY. July
26-29.
3. Foster, G.R. and L.D. Meyer. 1972. Transport of soil particles by
shallow flow. Trans. ASAE. 15(1):99-102.
4. Foster, G.R. 1976. Sedimentation, general. Proceedings of the
National Symposium on Urban Hydrology, Hydraulics, and Sediment Control.
Univ. of Kentucky. Lexington, Ky. July 26-29.
5. Holtan, H.N. 1961. A concept for infiltration estimates in watershed
engineering. Agricultural Research Service, U.S. Department of
Agriculture. ARS-41-51. 25 p.
6. Horton, R.E. 1919. Rainfall interception. Monthly Weather Review.
47:603-623.
7. Huggins, L.F. and E.J. Monke. 1966. The mathematical simulation of
small watersheds. Water Resources Research Center, Purdue Univ. Tech.
Rept. 1. 130 p.
STUDY RESULTS
-------
203
8. Huggins, L.F., J.R. Burney, P.S. Kundu and E.J. Monke. 1973.
Simulation of the hydrology of ungaged watersheds. Water Resources
Research Center, Purdue Univ. Tech. Rept. 38. 70 p.
9. Huggins, L.F., T.H. Podmore, and C.F. Hood. 1976. Hydrologic
simulation using distributed parameters. Water Resources Research Center,
Purdue Univ. Tech. Rept. 82. 59 p.
10. Meyer, L.D. and W.H. Wischmeier. 1969. Mathematical simulation of
the processes of soil erosion by water. Trans. ASAE. 12(6):754-758.
11. Overton, D.E. 1965. Mathematical refinement of an infiltration
equation for watershed engineering. Agricultural Research Service, U.S.
Department of Agriculture. ARS-41-99. 11 p.
12. Skaggs, R.W., L.F. Huggins, E.J. Monke and G.R. Foster. 1969.
Experimental evaluation of infiltration equations. Trans. ASAE. 12(6):
822-828.
13. Solomon, S.I. and S.K. Gupta. 1977. Distributed numerical model for
estimating runoff and sediment discharge of ungaged rivers. II. Model
development. Water Resources Research. v!3, n3. pp. 619-629.
14. Wischmeier, W.H. and D.D.Smith. 1965. Predicting rainfall-erosion
losses from cropland east of the Rocky Mountains. Agric. Handbook' 282.
Agric. Res. Serv., USDA. 47 p.
15. Yalin, Y.S. 1963. An expression for bed-load transportation.
Hydraulics Division, ASCE. 89(HY3):221-250.
4.9.2 Tile Drainage Simulation Model
A computerized simulation model has been developed to provide a
predictive tool for determination of sediment losses from tile effluent.
The model provides a flow hydrograph with associated sediment loading as a
function of the input variables (rainfall and initial soil moisture
content). The model will have the capability of being modified to
represent different tile system designs and soil types.
The need for concern of tile drainage influence on water quality is
shown by the significant contribution it has to stream flow. Approximately
50 percent of the Black Creek Watershed is drained by subsurface tile
systems. A tile system can contribute anywhere from 10 to 100 percent
(typically 30 percent) of the total runoff of a drained area. This
indicates that approximately 15 percent of the runoff per year from Black
Creek is tile effluent. During non-storm periods tile effluent is the
major source for stream flow in agricultural areas. The influence of tile
flow on stream flow may vary greatly depending on the annual rainfall
distribution.
An estimate of the sediment, phosphorus and nitrogen going into the
Maumee River from Black Creek tile effluent is approximately 19, .036 and
2.7 kilograms per hectare per year, respectively. This is based on the
previous flow assumption and tile effluent data collected by grab sampling
on 266 tile outlets in the Black Creek Watershed. The loading rates of
localized areas can be much larger as shown by G. 0. Schwab (1973). He
STUDY RESULTS
-------
204
measured annual sediment losses from tiles as high as 5400 kg/hect/year.
His results indicate that in some critical areas the tile effluent may be
the dominant effect on stream water quality.
4.9.2.1 BACKGROUND
Glacial tilled soils of the Midwest seem to be susceptible to erosion
losses through tiles. The soils drained by tiles generally have high silt
and clay contents. These fine particles are able to be detached and
transported within the soil profile by forces exerted on them by flowing
water. The actual detachment and transport mechanisms within a soil
profile are not well understood, but many studies have been done in closely
related areas such as piping effects and force balance relationships within
soils.
A model by D. Zaslavsky (1965) describes the force balance of
particles in cohesive soils. This model also shows the implied inter-
relationship of flow gradients to fine particle movement. Particularly it
indicates that for a given particle size a threshold flow level must be
reached before particle movatient will occur. The effect of the flow
channel size on the threshold flow is also provided. Zaslavsky's model
uses these relationships to obtain an expression which relates the critical
(threshold) flow for particle movement to a given particle size assuming a
mean pore channel size. To extend the use of Zaslavsky's model for an
erosion yield model it becomes necessary to attach a probabilistic
detachment model to the basic force balance relationships. Thus a
probability is associated with the critical flow and has a functional
relationship to flow above the critical flow. So for a given particle size
and flow rate it is possible to show a distribution of detachment potential
for a particle size distribution. The probabilistic approach used by H. A.
Einstein (1950) provided excellent results for particle detachment and
transport in open channels.
The particle detachment model described above requires knowledge of
the water flux distribution within the soil profile. Several tile flow
models (4) are available, but none are uniquely suited for a tile erosion
model. Therefore a two-dimensional porous medium flow model has been
developed to provide the necessary water flux distribution within the soil
profile. The flow in the unsaturated profile region is determined by
Darcy's Law which is the tension-conductivity method. The flow at the tile
will be tentatively determined by Toksoz and Kirkham's formula (1961) using
the watertable height above the tile. This tile flow formula was developed
for a constant infiltration rate passing through an unsaturated layer into
a saturated layer. This indicates continuity at the watertable which is
required to effectively model across this transition layer. Continuity is
expressed as:
Change in water storage = Inflow - Outflow
Using the assumption that a known geometric flow pattern exists near the
tile, the magnitude of water movement near the tile can be generated as a
function of R and 0 (radian distance and angular direction from tile,
respectively). Flow nets are available for several different soil profiles
above tiles (1957). The water flux is then used to determine the relative
volume of soil which is experiencing a certain erosion potential. These
volumes are then summed for all erosion potentials for given particle sizes
STUDY RESULTS
-------
205
to determine total erosion potential. As indicated, the sediment loss is
determined as a distribution shape and therefore absolute magnitudes are
not directly provided by this approach. Field data is needed to quantify
the sediment loss distribution.
4.9.2.2 MODEL FOR PARTICLE DETACHMENT
The forces acting on a soil particle can be summarized as:
Fg gravitation force
Fc cohesive force (attraction between particles)
Fh hydraulic forces (caused by water movement)
Fp point forces (caused by physical contact with other
particles)
To get particle movement, the hydraulic forces Fh must exceed the sum of
all the other forces, that is,
F > F + F + F
h g P
4.9.2.2.1 Gravitation Force
The effective gravitation force is the submerged weight of a particle.
Fg = V
-------
206
therefore,
+ AD)
f(D)dD (16)
(D. - AD)
The variables in the above expressions are described as follows: A is area
of test sample, gt is failure stress of test sample,ai is a geometric factor
for the i-th particles shape, A^ is the i~th particle area which is
influenced by the shear-stress, a^is the stress of the i-th particles, n is
the number of particles in the test layer, D is a particles size, "^is the
fraction of particles in the i-th particle interval which has a 2£Dwidth.
The density function f(D) should be proportional to the square inverse of
the particle size and directly proportional to the particle distribution.
4.9.2.2.3 Hydraulic Forces
A particle experiences drag and lift forces when flowing water passes
over it. The sum of these forces is equal to the overall hydraulic force,
that is,
F = F + F
h lift drag (17)
Drag forces are given by Stokes Law:
2
(18)
Parameters are described as follows: Ap is the effective area factor
(accounts for particle size exposed to the stream flow y fR.eis Reynold's
Number, and p is the density of water. ^
The lift force on a particle is developed when the water flows faster
over one side of the particle than the other. The lift force for an
attached particle can be expressed as a function of its exposed surface and
the velocity of water across its surface.
FL = CLV2
Parameters not previously described are as follows: Cj, lift coefficient
and F-^is exposed surface factor.
4.9.2.2.4 Ratio of Non-Point Forces
For particle detachment the hydraulic force must exceed the sum of the
cohesive and gravitational forces or expressed as a ratio.
F + F
IS S. = R < 1 (20)
Fh
STUDY RESULTS
-------
207
If the ratio R is less than unity then the potential for erosion exists.
However, the above expression assumes that all particles of the same size
experience the same cohesive force. This was not the case, so a normal
distribution was assumed such that a probability exists for particle
detachment for R's slightly greater than one. This serves to smooth out
the computer summation of the erosion potentials.
4.9.2.2.5 Point Forces
The point forces are impossible to describe for any one particle. The
magnitude of these forces can be very large, but intuitively one can reason
that only the particles near a free surface will have any possibility of
small or zero point forces. Within a soil profile only the smaller
particles (clay and silt) will see free surfaces in the pore cavities and
channels. A probability p for detachment must be associated with each
particle size D to account for the point forces. Now the probability of
detachment is the probability that the point force F.is less than all the
other forces, that is,
Prob [Fp < Fh(l - R)l - P ^
This functional relationship of p will vary by soil type and depth in
profile.
4.9.2.3 DISCUSSION
Several of the relationships presented are not easily determined
experimentally. Therefore, some assumed relationships will be used
initially until reliable laboratory data is available. The accuracy of the
model to predict actual erosion losses will serve to validate or disprove
assumed relationships. It should be noted that all forces are vector
quantities and therefore, the numeric analysis of the erosion model will be
more complex than indicated above.
The time dependence of the detachment mechanism is represented by the
probabilistic relationships of the point hydraulic forces. In order to
change the erosion potential of a particle either the flow must change or
another particle near it must move. So over time the loss of surface
particles increases the probability of detachment of particles beneath the
surface layer. Armoring will occur over time, but for this model it is
assumed that natural soil weathering will periodically recharge the smaller
particles in the pore cavities and channels.
To use the particle detachment model, it is convenient to combine the
probabilistic and force balance relationships into one erosion potential
expression for a tile system. . The detachment model provides a fraction
of particles detached per unit area per unit time for the particle
distribution interval D±AD at a given water flux. Define erosion
potential as the ratio or erosion rate at tile flow TQ to the erosion rate
at maximum tile flowf It is now possible to compute the erosion
potential at a TQ by making the assumptions of geometric flow net and the
STUDY RESULTS
-------
208
same soil texture throughout the profile. The computation requires an
integration over the entire soil profile for each particle size interval
used and then the addition of the results. For Zaslavsky's relationship
and a radial flow assumotion the erosion potential is given as:
EP =
(22)
The parameter QC;JJ,S the critical flow for a-'given particle size .
4.9.2.4 COMPUTER MODEL
The tile model is programmed in the GASP IV Simulation Language. GASP
IV was selected because of its advanced time stepping and differential
equation solving techniques. The computer model breaks the soil profile
above the tile into N layers (see Figure 29).
Soil Surface
Tile Depth
Figure 29. Tile system Layout
Hydraulic conductivity for each layer can be provided separately. This
gives tremendous latitude in the types of soil profiles which can be
analyzed. Flow between each layer is determined for each time step by use
of Darcy's Law.
q = k
9(T + Z)
3Z
At
(23)
Continuity at the watertable is determined by comparison of the flow into
STUDY RESULTS
-------
209
the layer in which the watertable is located and the flow out the tile as
determined by Toksoz and Kirkham's method.
TQ (tile flow) = M_ (24)
Parameters described above are as follows: q is vertical water flux, K is
hydraulic conductivity, T is tension head, Z is elevation head, t is time,
H is height of watertable above tile center, S is tile spacing, and F is
geometry coefficient for the tile system layout. Initial attempts will use
Zaslavsky's piping relation, and associated erosion potential. The
probabilistic model will be added when fully developed.
The computer model solves the above relationships for any rainfall
distribution provided. The output of the model is a plot and table of tile
outflow and sediment loading rate as a function of time. Also, at any time
during the simulation a moisture plot can be obtained for the soil profile
above tile.
4.9.2.4.1 Discussion
A similar problem exists for the hydraulic model as in the detachment
model, that is, some of the parameters are not readily available and when
available are in a graphic form. Graphs are empirically hard to represent
in a computer program if they do not have a functional relationship
(equation) which will represent them. This is the problem with many of the
required parameters for the tile erosion model. The relationship of
tension and hydraulic conductivity exhibit hysteresis which further
complicates exact determination. Because of these determination problems
linear assumptions are made for some of the relationships. It will be done
so that the parameter's representation will be within obtainable
experimental error. Linearization also greatly increases the efficiency of
the computer program.
4.9.2.5 CALIBRATION USING TILE SAMPLER DATA
Field data is essential to calibrate and verify the computer model.
As indicated the sediment loss potential as determined by the model does
not provide absolute magnitudes of the sediment loss directly. To
calibrate this potential distribution at least one water quality sample is
needed during a significant flow period. This in itself does not assure
that the computed shape of the potential distribution is correct.
Therefore, it is necessary to have water quality data for as many flow
conditions as possible in order to compare the distribution shapes of both
the actual and simulated sediment loss curves. To obtain this data base an
automatic pumping tile sampler was installed on a tile draining forty- three
acres of a typical soil type (Hoytville) of the Black Creek Watershed.
4.9.2.6 AUTOMATIC TILE SAMPLER
An automatic tile sampler has been operational since March, 1976. The
peak flow periods for tiles are during the winter and spring months.
Therefore, the equivalent of two years of data for calibration and
verification of the tile model should be available by late spring of 1977.
This pump sampler data has also been analyzed to provide loading rates
STUDY RESULTS
-------
210
directly for' the determination the tile effluents effect on water quality.
The fertilizer nutrients have also been looked at closely to define their
loss rate through the tile system.
The operation of the tile sampling station is unique in that discrete
water quality samples are collected proportionally to the tile outflow,
which is continuously monitored and recorded. The time at which a sample
is collected is also recorded on the flow hydrograph chart. The sampler
has the capability of collecting 72-500 ml water samples. The sampler
rates are approximately 1 sample per 30 minutes at maximum flow decreasing
to 1 sample per 12 hours at low flow. Another feature of the tile sampling
station is prevention of the tile outlet from becoming inundated. This is
necessary to provide reliable data during storm events in which the ditch
water level is above the tile outlet. Two 200 GPM pumps used in
conjunction with a sump maintain a free water fall over the flow calibrated
weir. Pump "on" times are also recorded to provide a check for the water
volume passing through the station. See Figure 30 for more detail of the
tile monitoring station.
4.9.2.7 SUMMARY
The hydrologic model is working for the assumptions previously
mentioned and for a modified non-linear moisture-tension relationship which
was determined by laboratory experiments. The model's output is in close
agreement with field data. The hydraulic conductivity-moisture
relationship is not yet fully developed, therefore further refinement of
the model is expected. The sediment detachment model will be attached to
the hydrologic model when the hydrologic model is fully calibrated.
The sampler station has functioned well for the past two years. The
collected samples have been analyzed for sediment and nutrient
concentrations. Results of this analysis are given in Section 0.0.
References
1. Schwab, G.O. and E.O. McLean. 1973. Chemical and Sediment
Movement from Agricultural Land into Lake Erie. Ohio Water Resources
Center. Ohio State University. Report No. 390X.
2. Zaslavsky, D. and G. Kassiff. 1965. Theoretical Formulation of
Piping Mechanism in Cohesive Soils. Geotechnique Vol. 15. No. 3.
Institution of Civil Engineers. Haifa, Israel.
3. Einstein, H.A. 1950. The bed-load Function for sediment
transportation in Open Channel Flow. U.S. Dept. Agric., Tech. Bull.
No.1026.
4. Luthin, J.N. 1966. Drainage Engineering. John Wiley and Sons,
Inc.
5. Toksoz, S. and D. Kirkham. 1961. Graphical Solution and
Interpretation of a New Drain-spacing Formula. J_. Geophys. Res. 66:509-
516.
6. Luthin, J.N. 1957. Drainage of AGricultural Lands. Published by
Am. Soc. Agro. Madison, Wise.
STUDY RESULTS
-------
211
Flow Baffles
Pump Float Switches
(2) 200 GPM Pumps
Figure 30. Automatic Tile Sampling Station
STUDY RESULTS.
-------
212
4.10 STUDIES OF NUTRIENT AVAILABILITY
Laboratory incubation experiments were conducted to evaluate the
transformations of nitrogen and phosphorus in water systems. These
transformations are important in determining the fate of N and P entering
water resources from agricultural land.
4.10.1 Effect of Incubation Temperature
The temperate zone of the U.S. is subject to wide variations in air
temperature throughout the year and thus, the temperature regime in surface
waters will vary diurnally and seasonally. The effects of incubation
temperature on the soluble phosphorus and inorganic nitrogen composition of
creek water are given in Table 87. The concentration of soluble inorganic
phosphorus increased significantly with time at all incubation
temperatures; however, the increases were greater at higher temperatures.
The increase in inorganic phosphorus was approximately one-third greater
for samples incubated at 33 degrees C as compared to those incubated at 5
degrees C. The soluble organic phosphorus concentration in samples
decreased with time, but the rate of decrease was slower at the lowest
incubation temperature.
The concentration of total soluble phosphorus generally increased with
time of incubation. This finding suggests that the increase in soluble
inorganic phosphorus in samples is not entirely the result of
mineralization of soluble organic phosphorus, but may in part result from
desorption of inorganic phosphorus from the sediment or mineralization of
organic phosphorus in the sediment. During the 12 weeks of incubation 1.8,
2.1, 3.75, and 8.25 yg of phosphorus per sample were released from the
sediment to the solution phase of samples incubated at 5, 15, 23, and 33
degrees C respectively.
In all samples except those incubated at 33 degrees C, the ammonium-
nitrogen content increased during the first week, probably as_a result of
mineralization of organic nitrogen. Nitrification was rapid in samples
incubated at all temperatures as evidenced by the increase in nitrate
content with time. The nitrate concentration in incubated samples
increased steadily with time and reached approximately the same level in
all treatments after 12 weeks of incubation. After 12 weeks of incubation
188, 224, 218, 231 yg of nitrogen per sample were released from the
sediment to the solution phase in samples incubated at 5, 13,23, and 33
degrees C, respectively. The increase in soluble inorganic nitrogen is
likely the result of mineralization of organic nitrogen present in the
sediment phase of creek water.
As stated previously, temperature has a definite effect on the rate
with which insoluble nutrients become available and on the amount of
nutrients that are converted to water soluble forms. The increasing
amounts of nitrogen and phosphorus converted to water soluble forms were
directly proportional to the incubation temperature. This is a significant
finding since aquatic plants exhibit accelerated growth rates during
periods of elevated water temperature, provided the temperature is within
the optimum range for growth. If water temperatures increase in a lake or
pond in which algal growth is a problem, the data obtained from incubation
experiments indicate that increased amounts of nitrogen and phosphorus will
become available to aquatic plants. In lakes where the temperature does
STUDY RESULTS
-------
213
TABLE 87
Effect of Temperature on the Concentration of Solution N and P
Treatment
Degrees C
5
13
23
33
Change in
Incubation concentration
time Concentration of soluble during incubation
(weeks) Inorg-P Org. P Amm.-N Nit.-N Sol-P Sol.inorg-N
**<^. />
0
1
3
6
12
0
1
3
6
12
0
1
3
6
12
0
1
3
6
12
.189
.161
.182
.280
.286
.189
.210
.261
.288
.295
.189
.220
.258
.309
.314
.189
.279
.280
.315
.340
.104
.132
.123
.093
.019
.104
.076
.017
.043
.012
.104
.090
.024
.018
.004
.104
.080
.018
.024
.008
0.54
0.97
1.20
0.76
0.10
0.54
1.16
0.24
0.10
0.07
.054
1.15
0.21
0.22
0.10
0.54
0.19
0.16
0.22
0.07
i»y/A
2.88
2.92
3.19
3.86
4.57
2.88
2.92
4.28
4.59
4.84
2.88
2.97
4.36
4.84
4.77
2.88
3.82
4.10
4.93
4.89
__
.000
.012
.080
.012
_
-.007
-.015
.038
.014
__
.017
-.003
.034
.125
_
-.066
.005
.046
.055
_
.47
.97
1.20
1.25
_
.66
1.10
1.27
1.49
_
0.70
1.15
1.64
1.45
__
0.59
0.84
1.73
1.54
not exceed 5 degrees C, the nutrient status would be much lower and thus,
reduce the potential for the growth of aquatic plants.
4.10.2 Effect of Aeration Status and Shaking
Sediments are subjected to various levels of aeration during
transportation and deposition in natural waters. Aeration is known to have
a great effect upon the soluble nitrogen and phosphorus in water-sediment
systems. It is essential to determine the role of sediments in supplying
soluble nutrients in natural water systems.
The effects of aeration status and shaking on the soluble nitrogen and
phosphorus components of creek water incubated for 12 weeks is given in
Table 88.
It was anticipated that shaking would increase the degree of aeration in
samples; however, there was little difference in dissolved oxygen content
of static or shaken samples. The dissolved oxygen measurements taken after
incubation of helium purged samples show that that anaerobic conditions
were not maintained throughout the incubation period; however, the
dissolved oxygen content was much lower than in those samples which were
STUDY RESULTS
-------
214
a.
4-1 O
C ft C
U 3
C O i
O C O >s.
o --loo w>
c w
c
.- o.
D i_ i
cfl 3
C T3 O
(0 00
JZ
r:
1
CJ 1 rA
r- O
.0 2
,_
O
IT) "^"
c rrf
0 c.
4-1 .
fO to
1 1 f~.
*-' O ~-^
OJ p-
u e-
t bu
O i_
«-> 0
c
73
CJ
> 4-1
. c
o c<
LO CM 4-1
to o c
o
C2 O
c
O '-
to
4-J C -^
tD E (U
_Q ._ QJ
o ^
c
c
(U
4J
fO
(U
V.
O LA J- LA
r~. rH to -a-
1
O rH rH rH
r-^ fA J- LA
rH O |vA CM
1 O O O CD
O O CD CD
co r-~ tD J" r~-
oo en IA co r-
CM CM J" J" -=T
J- LA rH CM e
LA rH CM CM rH
C rH CD O CD
-3- O -rj- CO -3"
CD CTJ CM i 1 CD
rH O 0 O O
O O O O CD
Cl O 00 CD J"
CO CM LA O rH
rH CM CM hA hA
O O O O O
CD LA O LA
1 tO CO J" LA
.
CO CO CD CD
O rH tA CD CM
rH
O O
.Q +->
o ro
L J j*
0) LO
r^ to o
r~> LA CM
i . . i
O rH CM
CM CD CO
cn j- to
1 O O 1 O
o o o
CO CM CD CO
CO rH 1 CM
CM j- -a- LA
j- r^* cn -3*
LA O rH tA
CD 0 O O
J- O O O
O rH O CD
rH rH O 1 O
O O O O
CD LA CM rH
03 r~- -J- to
rH CM ^A 1 rA
0 C 0 0
0 LA 0
i r>> r-^ i IA
CO CO CD
CD rH hA tO OJ
-u._|
o
._ o
.Q _i^
o ra
v. ^z
0) to
r^> cn cn to
r^- CM j- to
i
O rH rH rH
CO l-~ tO 1
rA 0 J- LA
1 O O O O
0 O 0 0
co cn CM to rH
co j- to r>- CD
CM I-A -3- J- -=T
j- o CD LA r^.
LA r«- O rH rH
00000
J- hA O CD O
O CD CD CD O
rH rH O 0 0
CD CD O O O
CD OO CD 1 tO
CO CM CD -3" rA
rH CM tA CM CM
O C3 O CD CD
CD CD O CD
1 rA CD tO LA
CM CM CM CD
*
O rH I>A tO CM
~0
E CJ 0)
3 bO -^
. i_ ro
3 -C
01 Q. to
STUDY RESULTS
-------
215
not helium purged. The soluble inorganic phosphorus concentration increased
with time in aerobic samples; however, in helium purged samples the soluble
inorganic phosphorus increased for the first three weeks and then decreased
during the last nine weeks of incubation. The levels of soluble inorganic
phosphorus were significantly higher in aerobic samples which were shaken
as compared to those which were static. The soluble organic phosphorus
content of all samples tended to decrease with time; however, the rate of
decrease of soluble organic phosphorus content during the 11 weeks of
incubation was greatest in samples which were shaken. The total soluble
phosphorus content of all samples tended to reach a peak value after one
week of incubation and then remain relatively constant or decline with time
thereafter.
In aerobic samples, the decline in total soluble phosphorus content
resulted from more rapid decreases in soluble organic phosphorus than
increases in soluble inorganic phosphorus, whereas in helium purged samples
soluble organic phosphorus rapidly decreased and soluble inorganic
phosphorus increased for three weeks and then decreased giving a net
decrease in total soluble phosphorus. During 12 weeks of incubation 3.75 yg
and 10.2 g of phosphorus were released to the solution phase from sediment
in aerobic samples which were static and shaken, respectively. In helium
purged samples 8.55 yg phosphorus per sample was removed from the solution
phase during 12 weeks of incubation.
The ammonium content of aerobic static samples and helium purged
samples increased during the first week of incubation, decreased to a low
level during the next two weeks, and then remained low during the remainder
of the incubation period. The ammonium concentration in aerobic shaken
samples decreased during the first week of incubation and then slowly
increased during the remaining 11 weeks of incubation. The nitrate content
of all samples increased rapidly during the initial one to three weeks of
incubation and then increased slowly during the remainder of the incubation
period. The finding that nitrification was occurring in helium purged
samples is good evidence that the samples were not anaerobic and that
nitrification is not limited by relatively low dissolved oxygen contents in
the water.
The soluble inorganic nitrogen content of samples tended to increase
with time throughout the incubation. During 12 weeks of incubation 218,
330, and 249 yg of inorganic phosphorus per sample were released from
sediment to the solution phase in aerobic static, aerobic shaken, and
helium-purged shaken sample, respectively. This finding suggests that
shaking may increase the mineralization of organic nitrogen and that a
reduction of dissolved oxygen in water may decrease the mineralization of
organic nitrogen.
4.10.3 Effect of Calcium Carbonate Addition
Sediments may contain various concentrations of calcium carbonate and
it has been reported that calcium carbonate has the ability to sorb
phosphate. However, it has been found that the phosphorus sorption
capacity of lake sediments tended to be inversely related to calcium
carbonate content. Since eroded soil materials may contain significant
amounts of calcium carbonate, it is desirable to study the effects of
calcium carbonate on solubility in sediment-water systems.
The effects of adding calcium carbonate on the soluble nutrient levels
in creek water samples incubated for 12 weeks is given in Table 89.
STUDY RESULTS
-------
216
TABLE 89
The Effect of Calcium Carbonate on the Level of Soluble N and P
Incubation
Treatment
None
CaC03*
time
(weeks)
0
1
3
6
12
0
1
3
6
12
Concentration of soluble
Inorg-P
0.189
0.220
0.258
0.309
0.314
0.185
0.203
0.260
0.270
0.289
Org.P Amm.-N
nn/1
mg/i
0.104 0.54
0.090 1.15
0.024 0.21
0.018 0.21
0.004 0.10
0.117 0.57
0.050 0.89
0.013 0.14
0.015 0.29
0.000 0.36
Nit.-N
2.88
2.97
4.36
4.84
4.77
2.97
3.11
4.81
4.64
5.20
Change in
concentration
during incubation
Sol-P Sol
_
0.017
0.003
0.034
0.025
_
0.049
0.029
0.017
0.013
inorg-N
_
0.69
1.14
1.63
1.44
_
0.46
1.41
1.39
2.02
2 grains of CaC03 added per sample.
The finding that addition of large amounts of calcium carbonate did not
significantly decrease the initial levels of soluble inorganic phosphorus
or soluble organic phosphorus in water samples suggests that calcium
carbonate does not sorb soluble phosphorus in samples of Black Creek water.
The levels of soluble inorganic phosphorus increased with time in both
calcium carbonate amended and unamended samples. The soluble organic
phosphorus concentration decreased with time in both amended and unamended
samples, but the rate of decrease was more rapid in calcium carbonate
amended samples. The total soluble phosphorus content of calcium carbonate
amended samples decreased during the first week of incubation and then
increased slowly during the remainder of the incubation period. The total
soluble phosphorus content of unamended samples increased for the first
week of incubation, decreased during the next two weeks, and then increased
to a near constant level for the remainder of the incubation period.
During the 12 week incubation period, 3.75 yg of phosphorus per sample were
released to the solution phase of unamended sample, whereas in calcium
carbonate unamended samples, 1.95 yg of phosphorus were removed from
solution. The effect of calcium carbonate on the total soluble phosphorus
content of water samples was likely due to decreased desorption of
inorganic phosphorus from sediment or to decreased mineralization of
organic phosphorus by the sediment.
The soluble ammonium content of calcium carbonate amended amples
increased during the first week of incubation and then declined to a low
level for the remainder of the incubation period. The nitrate
concentration increased with time in both calcium carbonate amended and
unamended samples, amended and unamended samples. There was little effect
of calcium carbonate addition on the apparent nitrification rate in
samples. The total soluble inorganic nitrogen content of unamended and
STUDY RESULTS
-------
217
calcium carbonate samples increased at about the same rate during the
first six weeks of incubation. However after twelve v/eeks of incubation
216 yg nitrogen were released to the solution in unamended samples whereas
303 y'g nitrogen were released in calcium carbonate amended samples.
4.10.4 Addition of Soluble Inorganic Phosphorus
Table 90 gives data on the effects of the initial inorganic phosphorus
concentration on the level of soluble phosphorus in creek water subjected
to laboratory incubation. Replacement of the liquid phase of the creek
water samples with distilled water markedly lowered the initial soluble
inorganic phosphorus content and increased the soluble organic phosphorus
content of the samples relative to the untreated samples. Addition of
organic phosphorus (34 or 84 yg per sample) markedly increased the soluble
inorganic phosphorus content of water samples (91 percent and 63 percent of
the added inorganic phosphorus remained in solution at the low and high
addition rates, respectively). Addition of 34 yg of inorganic phosphorus
per sample decreased the level of soluble organic phosphorus, whereas
addition of 84 y g phosphorus per sample increased the soluble organic
phosphorus concentration. The soluble inorganic phosphorus content of
samples whose liquid phase was replaced by water increased with time up to
three weeks of incubation and then remained relatively constant for the
remainder of the incubation period. The soluble inorganic phosphorus
content of samples amended with inorganic phosphorus tended to remain
relatively constant with time during incubation. The soluble organic
phosphorus content of most samples decreased rapidly during the first three
weeks of incubation and then o.cureased slowly thereafter. In samples
amended with 34 i-g of inorganic phosphorus the soluble organic phosphorus
content increased rapidly during the first week of incubation and then
decreased rapidly thereafter. The total soluble phosphorus content of
samples whose liquid phase was replaced by distilled water decreased
throughout the period of incubation. The total soluble phosphorus content
of samples amended with 34 yg of inorganic phosphorus remained relatively
constant throughout the incubation period, however, in samples amended with
84 yg of inorganic phosphorus the total soluble phosphorus content
decreased markedly for the first three weeks and remained constant.
The fact that the soluble organic phosphorus content of amended
samples was decreasing faster than the soluble inorganic P content was
increasing suggests that at least a portion of the decrease in organic
phosphorus from solution is the result of sorption of organic phosphorus
compounds by sediments. During 12 weeks of incubation, the amount of
phosphorus lost from solution (sorbed by sediments or immobilized in
microbial cells) was 11.4, 2.25, and 22.95 yg of phosphorus per sample for
samples with liquid phase replacement, amendment with 34 ug of phosphorus,
and amendment with 84 ug phosphorus respectively.
4.10.5 Addition of Soluble Inorganic Nitrogen
The effects of the initial levels of soluble inorganic nitrogen on the
concentrations of ammonium and nitrate in creek water samples incubated for
12 weeks is given in Table 91. Replacement of the liquid phase of samples
with distilled water markedly reduced the initial concentrations of
ammonium and nitrate in incubated samples. Addition of ammonium nitrate
(1500 or 6000 g nitrogen per sample) increased the initial concentration
of both forms of inorganic nitrogen in solution, although from 12 to 24
STUDY RESULTS
-------
218
TABLE 90
Effect of SIP Addition on the Level of soluble Phosphorus
Incubation
Treatment
Deionized
Water*
None
34 g P
added
84 g P
added
time
(weeks)
0
1
3
6
12
0
1
3
6
12
0
1
3
6
12
0
1
3
6
12
Concentration Change in
concentration
of soluble of total soluble P
Inorg.P
0.052
0.125
0.174
0.171
0.157
0.189
0.220
0.258
0.309
0.314
0.396
0.404
0.412
0.401
0.433
0.545
0.466
0.513
0.551
0.544
Org.P during
^ j /i
mg/i
0.181
0.083
0.009
0.005
0.000
0.104
0.090
0.024
0.018
0.004
0.055
0.040
0.025
0.008
0.002
0.159
0.174
0.029
0.014
0.007
incubation
_
.026
.059
.076
.076
_
.017
.013
.035
.025
_
.050
.013
.041
.015
.064
.162
.139
.153
Entire liquid phase of sample replaced with deionized
water.
percent of the added ammonium was apparently removed from solution by
cation exchange reactions with sediment. The ammonium concentration in.
samples having their liquid phase replaced with distilled water remained
low throughout 12 weeks of incubation. The ammonium concentration in
samples treated with ammonium nitrate remained constant for one week and
then decreased to very low values during the next two weeks of incubation.
The nitrate content of samples having their liquid phase replaced with
distilled water tended to increase slowly during the incubation period.
The nitrate content of samples amended with ammonium nitrate increased,
slowing during the first week of incubation, then increasing rapidly during
the next two weeks, and then remained relatively constant throughout the
remainder of the incubation period.
The total soluble nitrogen content of samples whose liquid phase was
replaced with distilled water tended to increase slightly as a result of
incubation indicating that mineralization was slow in that system. The
total soluble nitrogen content of samples amended with ammonium nitrate
increased significantly during the first three weeks of incubation and then
STUDY RESULTS
-------
219
TABLE 91
Effect of Ammonium Nitrate Addition on the Concentration of Soluble N
Incubation
Treatment
De ionized
Water*
None
1500 g N
added
6000 g N
added
time
(weeks)
0
1
3
6
12
0
1
3
6
12
0
1
3
6
12
0
1
3
6
12
Concentration
of soluble
Amm.-N
0.26
0.21
0.14
0.20
0.17
0.54
1.15
0.21
0.22
0.10
4.36
4.34
0.19
0.22
0.32
18.12
18.92
0.22
0.24
0.22
Nit.-N
0.28
0.11
0.80
0.93
0.80
2.88
2.97
4.36
4.84
4.77
7.14
7.88
13.33
13.34
13.22
21.64
22.89
43.06
43.20
42.95
Change in concentration
of soluble inorganic N
during incubation
_ _TYlf"T ~\
mg/i -
_
-0.21
0.40
0.58
0.44
_
0.72
1.17
1.66
1.64
0.71
2.00
2.05
2.03
_
2.14
3.60
3.86
3.50
* Entire liquid phase of sample replaced with deionized water,
remained relatively constant for the remainder of the 12 week period. The
net amounts of inorganic nitrogen formed during 12 weeks of incubation were
66, 246, 305, and 525 yg per sample for samples with liquid phase
replacement, no treatment, addition of 1500 g nitrogen, and addition of
6000 yg with liquid phase replacement as compared to untreated samples
suggests that a portion of the nitrogen mineralized is soluble organic
nitrogen which was lost when the liquid phase was removed. The finding
that more nitrogen is mineralized in samples treated with ammonium nitrate
than in untreated samples suggests that a "priming effect" of inorganic
nitrogen on mineralization may be significant in aquatic systems.
From the data collected during the incubation of the stream samples
under various environmental conditions, it can be concluded that there are
both long-term and short-term transformations occurring which influence the
amounts and forms of nitrogen and phosphorus found in the solution phase.
Short-term transformations appear to be sorption or desorption of both
organic and inorganic phosphorus and possibly the release of sorbed
phosphorus from soil minerals and dissolution of phosphorus occluded in
RESULTS
-------
220
iron and aluminum. The short-term transformations influencing the nitrogen
and organic phosphorus contents of water are much more dependent upon the
factors affecting microbial growth since transformations are largely
carried out by microorganisms. Microbiological transformations may occur
in very short periods of time if conditions are favorable for organisms or
long periods when conditions are unfavorable. The short-term processes
involving inorganic phosphorus transformations are more physiochemical in
nature and therefore not so dependent on environmental factors.
The long-term processes involved in nitrogen and phosphorus
transformations in water systems are likely to be the mineralization of
organic nitrogen and phosphorus components in solution and in the sediment.
Based on data obtained in this study, mineralization processes appear to be
major long-term processes leading to higher soluble nitrogen and phosphorus
concentrations in incubated samples.
The soluble inorganic phosphorus concentration in most samples
increased slowly with time during the incubation period. It appears this
increase in soluble inorganic phosphorus was due mainly to the
mineralization of soluble and sediment organic phosphorus. During the
mineralization process the sorption-desorption process controls the
equilibrium obtained between the sediment and solution.
In aquatic systems amended with calcium carbonate the concentration of
total soluble phosphorus remained relatively constant indicating that
calcium carbonate may decrease mineralization of sediment organic
phosphorus. Increased content of dissolved oxygen and shaking of samples
increased the concentration of soluble inorganic phosphorus and decreased
the concentration of organic phosphorus. The overall effect was an
increase in total soluble phosphorus during the incubation.
Addition of inorganic phosphorus to samples released organic
phosphorus into solution most likely by replacement of organic phosphorus
compounds sorbed on soil colloid surfaces by added inorganic phosphorus.
The total soluble phosphorus content of samples treated with inorganic
phosphorus decreased during the 12 week period indicating that soluble
organic phosphorus was sorbed by the sediments. The samples in which the
solution was replaced by deionized water were unable to attain the original
concentration of inorganic phosphorus by desorption processes and part of
the soluble organic phosphorus present in the liquid phase was sorbed by
the sediment.
A 33 degree C incubation temperature increased the rate of
mineralization of phosphorus and increased the final total soluble
phosphorus concentration by a significant amount. A 5 degree C incubation
temperature decreased the rate of mineralization of organic phosphorus.
The treatments applied which would increase microbial growth also
enhanced the mineralization of nitrogen and as a final result the
concentration of nitrate-nitrogen in solution increased. Aeration, calcium
carbonate amendment, increased temperature and the addition of ammonium
nitrate all had positive effects on the mineralization of nitrogen and the
final total soluble inorganic nitrogen concentration. Low temperature
seemed to have the greatest negative effect on the concentration of soluble
inorganic nitrogen, but there were still increases in the solution
concentration over the 12 week incubation period.
STUDY RESULTS
-------
221
4.11 ALGAL STUDIES
4.11.1 Introduction
The importance of phosphorus contamination of natural waters has
received an enormous amount of attention in past years. One reason for
this amount of investigation has been the "algal bloom" effect of
increasing phosphorus additions to water which normally did not contain
concentrations of phosphorus capable of supporting excessive photosynthetic
biomass. The short-term effects of an algal bloom range from a
degeneration in the quality of drinking water by algal exudates and
physical fouling of water treatment facilities, to the impairment of
enjoyment of recreational waters by odorous decomposition of senesced
cells, direct inducement of fish kills, and depletion of aquatic aesthetic
value. The long-term effect of an algal bloom is the acceleration of
eutrophication.
The dissolved orthophosphate ion is the most available form of P to
algae (Vollenweider, 1970). However, in the past, the comparison of
chemical analysis of orthophosphate to algal availability of the ion has
given variable results. Therefore, the determination of algal available
soluble P is the only dependable method of ensuring knowledge of a water
sample's potential for supporting biomass. The most frequently used method
for accurately determining available soluble P is the algal bioassay method
developed by the U. S. Environmental Protection Agency called the
Provisional Algal Assay Procedure Bottle Test (PAAP). (1971).
In lake systems the amount of soluble orthophosphate in the water
phase is a function of the equilibrium between the inorganic P bound on the
sediment and the interstitial dissolved inorganic P (Syers et al, 1973).
Dissolved inorganic P can, therefore, be released to the overlying water
column when the concentration of interstitial dissolved inorganic P exceeds
that in the water column. Because of the dynamic equilibrium existing
between sediment and solution, the sediment of a lake system is ultimately
a P reservoir capable of replenishing inorganic P taken up by aquatic
organisms.
Sediments normally contain large quantities of amorphous, hydrated
iron oxides capable of sorbing P. Sorbed P in sediments is capable of
mobility within the system and rapidly sorbing or desorbing in response to
small changes in P concentration. Therefore, the potential for P release
in response to uptake within the system is of importance and in essence
will dictate the quantity of supportable biomass.
Determination of the fraction of sediment P available to algae is very
difficult. Although dialysis systems have been used in the past with
limited success, a method developed by Sagher and Harris (1975) incubates
algal cells intimately with sediment and enables the determination sediment
P fraction capable of supplying P for algal uptake. The Sagher-Harris
method was selected for use in this study.
The objectives of this study were:
1. TO determine the availability to algae of soluble and sediment-
STUDY RESULTS
-------
222
bound P in water samples collected from the Black Creek watershed, and
2. To determine the sediment P fraction (ammonium flouride, NaOH, and
HC1 extractable) of sediment-bound P which is most available to algae.
4.11.2 Materials and Methods
4.11.2.1 PAAP BOTTLE TEST
An algal bioassay was conducted to determine the available portion of
soluble P in water samples obtained from seven sites within the Black Creek
watershed on March 28 and June 30, 1977. The bioassay method used was the
PAAP Bottle Test (USEPA, 1971). The method finds its roots in Liebig's Law
of the Minimum which states that "growth is limited by substance that is in
minimal quantities in respect to the needs of the organism." That is, if
one nutrient is lacking, the organism's limit for growth (in this case,
reproduction) is determined by parameters set by the amount of limiting
nutrient. Therefore, the PAAP method used created a reference curve which
related the number of cells (or growth rate) of a test alga to the
concentration of the nutrient being tested, phosphorus, in an inoculated.
growth medium in which no other factor is limiting. The cell numbers of
the inoculated water sample were determined and the numbers compared to
those found in the reference medium to ascertain the corresponding
concentration of available P. In addition to inoculating the unamended
water samples, three additional treatments were used to attempt a more
complete evaluation of limiting nutrients in Black Creek drainage water.
The treatments used were (1) a P spike of 0.1 mg/1 to ensure that P would
not be limiting maximal cell reproduction, (2) a micronutrient spike equal
to that found in the PAAP reference medium to measure the response of the
test organism to added micronutrients, and (3) a combination spike of 0.1
mg P/l plus micronutrients equal to that found in the reference medium in
order to create a non-limiting growth medium.
The bioassay was conducted in a 250 ml Erlenmeyer flask containing 60
ml of the PAAP medium or the water sample. The standard PAAP reference
medium contained between 0.0 and 0.2 mg P/l (0.000, 0.005, 0.015, 0.050,
0.075, 0.100, 0.200 mg P/l). Cell counts were made initially and at 24
hour intervals until a constant, positive relationship existed between cell
numbers and P concentration (after three days of incubation for all studies
conducted) at which time the cell densities of inoculated water samples
were counted and the available P concentration determined by reference to
the standard growth curve.
4.11.2.2 AVAILABILITY OF SEDIMENT-BOUND P
In determining the biologically available sediment-bound P, studies
were conducted on the amounts of sediment-bound P sequentially extractable
with ammonium flouride, NaOH, and HC1 before and after incubating with
algal cells for two weeks under optimum growth conditions. Decreases in
extractable P in sediments resulting from incubation were assumed to be
that P immobilized into cells during algal growth. All sediment P values
were corrected for the amounts of algal P in the system which were
extracted during the sequential extraction procedure.
All sediment bioassay studies were carried out in 250 ml Erlenmeyer
flasks using 60 ml of standard medium containing sufficient sterile
STUDY RESULTS
-------
223
sediment to provide 0.62 mg P/l. Flasks were inoculated with Selenastrum
capricornutum (10,000/ml), incubated at 24 degrees C for 2 and 4 weeks,
cell densities determined, and the entire contents of duplicate flasks were
subjected to sequential extraction and P determination.
4.11.3 Results and Discussion.
4.11.3.1 AVAILABILITY OF SOLUBLE P
A study was conducted to determine the availability of soluble
phosphorus to algae in water samples from seven sites within the Black
Creek watershed. The method employed was a modification of the Provisional
Algal Assay Procedure Bottle Test (USEPA, 1971) to provide a rapid assay of
available soluble P.
S. capricornutum exhibited a typical sigmoid growth rate in the
reference medium at P concentrations of 0.05 mg/1 or greater. The
stationary phase of growth began after 96 hours of incubation for all
treatments, but at lower cell densities for each decrease in P
concentration.
The reference curves produced by plotting cell numbers after three or
four days versus initial P concentration gives a leveling off of cell
numbers at P concentrations greater than 0.1 mg/1. The leveling off above
0.1 mg P/l may be looked upon within the experimental system as the
critical level of P or that concentration at which nearly maximum cell
production takes place.
Table 92 presents concentrations of soluble inorganic P (SIP) and
total soluble P (TSP) analytically determined in the March and June
samples, as well as the amount of available P detected by the bioassay
using cell counts as the determinative criteria in amended and unamended
samples. In all but one sample, the quantity of soluble orthophosphate
determined by chemical analysis was greater in March as compared to June
samples.
The same trends were observed for the quantity of TSP in water
samples.
Also indicated is the fact that in no case did the quantity of P
determined as available by the bioassay procedure exceed that detected
chemically as SIP. Tne chemical analysis of SIP, therefore, detected P
components which were unavailable to S_. capricornutum in unamended water
samples during three to four days of incubation. These components may
soluble polyphosphates or hydrolyzable organic P esters.
The available P detected in the unamended samples averaged 0.096 mg/1
(nearly maximal biomass production) in March samples and 0.033 mg/1 in June
samples (indicating cell production was limited by P). Tne range in
available P was 0.076 to 0.128 mg P/l and 0.012 to 0.049 mg P/l for March
and June samples, respectively.
Results of March samples receiving the P spike seem to bear out the
fact that P was not limiting cell production. However, data from the June
samples indicate a response to P addition, but in most cases cell densities
did not reach maximal values observed in reference curves. The finding
STUDY RESULTS
-------
224
TABLE 92
Availability to Algae of Soluble Phosphorus in Stream Water
P Concentration Available P in Water Samples as
Site #
March:
2
3
4
5
6
12
14
Ave
June:
2
3
4
5
6
12
14
Ave
in Water
SIP
.106
.121
.121
.171
.259
.135
.131
.149
.069
.038
.045
.053
.072
.047
.161
.069
TSP
.123
.150
.139
.173
.280
.153
.148
.166
.100
.063
.075
.072
.091
.062
.190
.093
Determined from Cell Count Bioassy of:*
U
.080
.076
.116
.128
.102
.086
.083
.096
.035
.016
.033
.040
.045
.012
.049
.033
P
^g F/TQl -
.084
.044
.098
.106
.068
.068
.094
.080
.018
.036
.042
.080
.099
.081
.045
.048
MN
094
106
116
099
096
116
132
108
041
019
043
046
044
020
257
067
PMN
.090
_
.081
_
_
.085
.061
.088
_
.085
.032
.066
U, unamended water sample; P, water sample spiked
with 0.1 mg P/l; MN, water spiked with micronutrients;
PMN, water samples spiked with phosphorus (0.1 mg P/l)
and micronutrients.
could be an indication that another nutrient was limiting algal growth.
The averages of the March and June samples spiked with micronutrients
gave growth responses over both the unamended and P spiked samples. That
is, the organisms incubated in the water sample were able to utilize more P
as a result of being supplied with additional micronutrients. This growth
stimulation from micronutrient addition occurred in four of seven March
samples and two of the seven June samples.
The P plus micronutrient spike gave positive growth responses in five
of six samples treated, but in no case was the available P detected by
bioassay equal to the 0.1 mg P/l added. This is an indication of a
limiting concentration of micronutrients or P, or both, which upon adding P
and micronutrients disappeared until growth was limited by another
nutrient.
A comparison was made (Table 93) between amounts of available soluble
P detected by bioassay in water samples from groups of sites which differ
in their land use characteristics. In the March sample, the soluble P
determined by bioassay to be available was greatest in the rural-urban
STUDY RESULTS
-------
225
portion of the watershed. The June samples showed the greatest available P
concentration in samples collected from the rural-urban portion of the
watershed and the Maumee River.
TABLE 93
Variation of Available Soluble Phosphorus by Adjoining Land Use Characteristics
Site #
2
3
4
Ave
5
6
Ave
12
14
*Ave
Sampling Time
March June
Available Soluble P(yg/l)
Rural Portion
.080 .035
.076 .016
.116 .033
.091 .028
Rural-Urban Portion
.128 .040
.102 .045
.115 .042
Entire Watershed
.086 .012
Maumee River
.083 .049
.096
.033
Includes available soluble P from all sites sampled.
4.11.3.2 AVAILABILITY OF SEDIMENT-BOUND P
A study was conducted to evaluate the availability of sediment-bound
phosphorus to algae in suspended sediments collected from seven sites
within the Black Creek watershed. A modified version of the method
developed by Sagher and Harris (1975) was employed. This method is a
combination chemical and biological assay which fractionates the forms and
quantities of available P present. Only data from the two week incubation
time is reported.
The proportion of total sediment P immobilized by algae from each
sample during a two week incubation and the final density in each sample
are shown in Table 94. In March samples an average of 18.1% of total
sediment P was available to algae and the range in availability was from
10.5 to 27.7%. The available P in the June samples averaged 21.1% of the
total sediment P with a range from 10.5 to 30.0%. These results tend to be
higher than some other previously measured values (Wildung and Schmidt,
1973) and lower than others (Sagher and Harris, 1975).
Table 95 presents the final cell densities and the available P as a
percent of the sediment inorganic P (Pi) after incubation for two weeks.
STUDY RESULTS
-------
226
TABLE 94
Proportion of Sediment Phosphorus Immobilized by Cells
Site #
2
3
4
5
6
12
14
Ave
Cell
Density
xl 0-6 A»l
8.529
9.599
4.242
5.225
6.500
-
-
6.819
Sampling
March
Available P
as % of Total
14.7
19.4
10.5
27.7
20.6
-
-
18.6
Time
Cell
Density
xlO-6A»l
5.175
8.551
5.954
5.000
6.591
5.900
8.408
6.511
June
Available P
as % of Total
10.5
20.2
16.8
16.6
29.0
24.6
30.0
21.1
The averages of the March versus June samples show an increase (26.9 vs.
33.1%, respectively) in the percent of sediment Pi of which become
available after two weeks of incubation. This suggests that a higher
percentage of sediment Pi is available to algae during storm events with
low sediment transport (June sample) as compared to more intense storms
with more soil erosion.
TABLE 95
Proportion of Sediment Inorganic Phosphorus Immobilized
Sampling Time
March
Site #
2
3
4
5
6
12
14
Ave
Cell
Density
xlO-6/ml
8.529
9.599
4.242
5.225
6.500
-
6.819
Available P
as % of Pi
26.6
27.9
15.0
34.7
30.7
-
-
26.9
June
Cell
Density
x!0-6A»l
5.175
8.551
5.954
5.000
6.591
5.900
8.408
6.511
Available
as % of
26.7
29.0
34.1
31.1
37.7
32.7
40.9
33.1
P
Pi
The average proportions of Pi to become available are again slightly
higher than Wildung and Schmidt (1973) reported in lake sediments. The
percentage of the total available P derived from each inorganic Pi fraction
is presented in Table 96.
STUDY RESULTS
-------
227
TABLE 96
Proportion Immobilized by Cells Originating from Certain Extractable Fractions
Sampling Times
March
Inorganic P Fraction
Extracted By
Site #
2
3
4
5
6
12
14
Ave
Available
P
pg P/flask
3.45
4.70
2.62
6.84
4.84
-
-
4.49
Amm.F
% of
33.1
43.7
22.9
36.6
52.0
-
-
37.7
NaOH
Available
39.2
47.0
0
38.8
40.7
-
-
33.1
HC1
Pi
27.7
9.3
77.0
24.5
7.2
-
-
29.1
Available
P
y g P/flask
4.03
4.54
4.63
3.73
8.21
6.51
6.40
5.44
June
Inorganic P Fraction
Extracted By
Amm.F
% of
73.3
28.2
32.5
51.6
43.6
32.4
61.8
46.2
NaOH
Available
7.0
36.6
67.5
48.2
39.0
49.4
28.0
40.8
HC1
Pi
20.6
35.1
0
0
13.6
18.2
10.2
14.0
On an average, in both March and June samples, the highest proportion
(37.7 and 46.2%, respectively) of the available Pi originated in the
ammonium flouride extractable fraction (range was 22.9 to 57.0%, and 28.2
to 73.3% respectively). Ranking second in availability is NaOH extractable
fraction which contributed an average of 33.1% and 40.8% of the available
Pi in March and June samples, respectively. The HC1 extractable fraction
also contributed to available P in both the March and June samples (29.1
and 14.0% respectively). This finding suggests that all Pi fractions in
sediment are at least partially available to algae, however, the aluminum
and iron bound P fractions (ammonium flouride and NaOH extractable) seem to
be the most important.
, Table 97 compares the proportion of Pi which becomes available after
two weeks in sediments from sites with differing land use characteristics.
There was a higher proportion of Pi which was available in the samples from
the rural-urban portion of the watershed as compared to those from the
rural portion (over 9% greater in March and 4% in June samples). In
addition, the highest proportion of available Pi was observed in the June
samples taken from the Maumee River site. Most of the available sediment
Pi originated in the ammonium flouride extractable fraction (Table 95).
4.11.4 Conclusions
4.11.4.1 AVAILABILITY OF SOLUBLE PHOSPHORUS
To study the availability of soluble P to algae a modification of the
PAAP Bottle Test was used and provided a consistent, positive relationship
between cell numbers and P concentration.
The soluble P determined as available by bioassay for all sites
sampled and both sampling times was always less than that measured
chemically as SIP and TSP. Algal growth in June samples was limited by low
STUDY RESULTS
-------
228
TABLE 97
Variation in Sediment Inorganic P Immobilized by Source of P
Site #
2
3
4
Ave
5
6
Ave
12
14
*Ave
Sampl ing
March
Available P as
Rural Portion
26.6
27.9
15.0
23.2
Rural-Urban Portion
34.7
30.7
32.7
Entire Watershed
-
Maumee River
-
26.9
Time
June
% of inorg.P
26.7
29.0
34.1
29.9
31.1
37.7
34.4
32.7
40.9
33.1
* Includes values available from all sites.
concentrations of available P, but in most cases some nutrient or factor
other than P or micronutrients also limited growth. In addition, the
rural-urban portion of the watershed contained higher levels of available P
than did the rural portion in both March and June sampling periods.
4.11.4.2 AVAILABILITY TO ALGAE
Algal growth was shown to be limited by P in treatments in which the
sediment was the sole source of P. However, the sediments did supply in
most cases sufficient P to support substantial growth of S. capricornutum.
The proportion of total sediment P which became available in the March and
June samples were 18.6 and 21.1%, respectively, and the proportion of Pi
which became available was 26.9 and 33.1%, respectively. The largest
portion of the available P originated in the ammonium flouride extractable
fraction, while both the NaOH and HC1 extractable fraction contributed to
the total available P pool. Sediments from the rural-urban portion of the
watershed had higher proportions of the Pi available for algae as compared
to agricultural portions. Maumee River sediment from the June sampling
displayed the highest (40.9%) proportion of Pi as available to algae.
References
United States Environmental Protection Agency. 1971. Provisional Algal
STUDY RESULTS
-------
229
Assay Procedure: Bottle Test. Washington, D.C. Eutrophication Research
Program.
Volleneweider, R.A. 1970. Scientific Fundamentals of the Eutrophication
of Lakes and Flowing Waters, With Particular Reference to Nitrogen and
Phosphorus as Factors in Eutrophication. Organizations for Economic
Cooperation and Development, Paris.
Syer, J.K., R.E. Harris, and D. E. Armstrong. 1973. Phosphate Chemistry
in Lake Sediments. Journal of Environmental Quality. 2* 1-13.
Sagher, A. and R. Harris. 1975. Availability of Sediment Phosphorus to
Microorganisms. Water Resources Center (Also Technical Report WIS WRC 75-
01), Madison, Wisconsin.
Wildung, R.E. and R.L. Schmidt. 1973. Phosphorus Release from Lake
Sediments. Office of Research and Monitoring. United States Environmental
Protection Agency (EPA-660/3-73-015).
4.12 STUDIES OF MAUMEE BASIN
The annual precipitation for the Maumee River Basin and the annual
sediment yield and discharge from the Maumee River into Lake Erie for a
ten-year period, October 1961 to October 1972, are presented in Table 98.
TABLE 98
10-Year Annual Precipitation Sediment Yield and Discharge
Precipitation* Discharge
Year
1970-71
1969-70
1968-69
1967-68
1966-67
1965-66
1964-65
1963-64
1962-63
1961-62
Average
(mm)
739
898
938
980
835
807
826
716
656
683
808
(mm over basin)
198
267
320
345
348
181
196
150
109
186
230
Ratio of Discharge
to Precipitation (%)
27
30
34
35
42
22
24
21
17
27
28
Sediment
Yield (kg/ha)
304
609
662
989
763
202
516
427
159
323
495
*Average of three locations: Fort Wayne, Defiance and Toledo.
This table is based on data provided by the U.S.
Environmental Data Service and the U.S. Geological
Survey.
The annual precipitation for the 10-year period seems to be part of
some cyclical pattern. Less than normal precipitation occurred from 1961
to 1966 and above normal precipitation from 1967 to 1971. The years with
the highest precipitation had the highest runoff and also the highest
STUDY RESULTS
-------
230
percentage of precipitation occurring as runoff. As a consequence, the
average annual sediment yield for the last five years was 665 kg/ha, more
than twice the 325 kg/ha yield for the first five years.
Monthly precipitation patterns at Fort Wayne in the western part of
the Maumee River Basin, at Defiance in central part, and at Toledo in the
eastern part near the mouth of the river are shown in Figure 31 for the
ten-year period. Low precipitation months are October, February and for
the growing season, August. High precipitation months are usually April,
May, June and July.
The average monthly discharge of the Maumee River given as a depth
over the basin and the sediment yield on an area basis are shown in Figure
32.
Comparison of the monthly patterns of discharge and sediment yield
shows a disproportionally high sediment yield occurring in December.
Usually the first winter storm of any magnitude carries a high
concentration of sediment. The winter storms are generally frontal storms
which may be basin-wide. Much of the sediment is already deposited in the
channels or lower portions of slopes from localized, convective-type storms
which occurred in the spring and early summer months. While the intensity
of the localized storms may be high, they often do not have an areal
coverage which is great enough to cause significant flow in the major
streams of the basin. During the six-month period, June through November,
only 16 percent of the total flow and 8 percent of the sediment yield
occurred for the 10-year period. On the other hand, the six-month period,
December through May, had 84 percent and 92 percent of the total flow and
sediment yield, respectively.
In general, those months with the highest discharge or runoff are also
the months with the highest sediment yield. For the years with relatively
high precipitation, the highest sediment yield usually occurs in December
or early January or later in February and early March; for the years with
low precipitation, the highest sediment yields are likely to occur in March
or April.
The average annual sediment yield of nearly 500 kg/ha from the Maumee
River into Lake Erie is not large when compared to many river basins.
However, considering the special nature of the receiving body of water
yields of this magnitude may still have a detrimental impact. Although
this yield is caused by much larger field erosion rates, these rates may
yet be within the natural restoration limits of most of the basin soils.
The sediment load also consists mostly of suspended particles with a high
colloidal fraction. Streams in the basin tend to remain murky even during
low flow. And sediment plumes often extend 60 km or more into Lake Erie.
4.12.1 Sources of Sediment
The major sources of sediment from agricultural watersheds in the
Maumee Basin are cultivated fields which are subject to rainfall impact and
runoff turbulence. Ditch bank sloughing and channel scouring probably
account for less than five percent of the total sediment load. Tile drains
in some sections of the Maumee Basin have been reported to discharge
substantial amounts of sediments into ditches or streams. However, in the
Black Creek Watershed, only about one to two percent of the total sediment
STUDY RESULTS
-------
231
120
100
| 80
Z
560
£
5 40
LU
tr
Q_
20
0
FT. WAYNE
120
100
| 80
Z
o
P 60
2
0 tO
LU
or
a.
20
0
DEFIANCE
>!
MM
ONDJFMAMJJAS
ONDJFMAMJJAS
i<:u
100
£ 80
I"
540
LU
o:
CL
20
o
TOLEDO
_-
_
PRECIPITATION (mm)
ONDJFMAMJJAS
IOJ
100
80
60
40
20
0
AVERAGED FT. WAYNE,
DEFIANCE $ TOLEDO
._.
__
:
-
_..
ONDJFMAMJJAS
Figure 31. Monthly Precipitation Pattern Fort Wayne, Defiance, Toledo
STUDY RESULTS
-------
232
60
50
to
o
UJ
O
or
x
o
c/5
Q
tr
<
or
UJ
30
10
o
150
-125
o>
50
UJ
Q 25
UJ
ONDJFMAMJJAS
ONDJFMAMJJAS
(a) Average monthly discharge (left) and sediment yield (right)
71
70
69
68
67
66
65
64
63
62
2
2
1
1
-
6
-
-
1
2
4
12
3
3
7
3
1
1
1
5
9
4
9
25
32
15
1
-
1
2
3
3
27
7
4
13
7
2
1
18
30
14
15
17
II
8
16
1
2
14
21
16
5
9
21
12
31
28
58
44
4
24
15
8
II
8
30
49
17
6
15
15
II
13
10
22
8
9
4
4
8
3
6
9
2
3
2
7
8
2
2
4
6
4
1
7
1
1
4
1
1
2
1
3
1
2
1
1
2
1
1
1
1
1
-
1
2
-
1
1
71
70
or 69
UJ 68
>-67
-------
233
yield originates from tile drainage systems. Even though yields from tile
drains are relatively small, it should be pointed out that the discharged
sediments are mostly colloidal and are delivered directly to ditches or
streams for ready transport.
The potential for rainfall and field runoff to cause erosion
correlates closely with a statistic consisting of the product value of two
rainstorm characteristics, namely, the total kinetic energy of a storm and
its maximum 30-minute intensity. This statistic, called the erosion index,
allows for comparisons of the annual erosion potential of rainfall and
field runoff between different regions and for the determination of the
distribution of the annual erosion potential within a region. For
instance, the annual erosion potential for the Maumee River Basin is only
one-fourth of the erosion potential for lands in the United States near the
Gulf of Mexico. Of greater interest, however, is the distribution of the
annual erosion potential for the Maumee River Basin. The distribution, in
general, follows an S-shaped curve in which 80 per cent of the annual
erosion potential occurs in the five-month period, May through September.
May and June are particularly critical months as far as soil erosion
is concerned. This is the time when seed beds are prepared and new crops
established. In the Black Creek Watershed, over 70 percent of the annual
sediment load into the Maumee River occurred during these months in 1975.
For the year, the total sediment yield from the watershed was about 2400
kg/ha. However, 1976 was an abnormal year with very little runoff during
May and June. Only about 2 percent of the annual sediment yield occurred
during these months and quite predictably the sediment yield for 1976 was
low averaging about 500 kg/ha.
The erosion-sedimentation process in a river basin can, in general, be
divided into upland and in-channel phases. The upland phase is closely
related to rainfall events and the nature of the receiving land area.
Major variables are the degree of slope and slope length; surface cover;
soil type, tilth, and water content; and the rainfall pattern, intensity,
duration, and amount. Temperature may also be an important variable in
areas where a significant number of freeze-thaw cycles occur. Erosion is a
selective process. Runoff from eroded land will usually contain a
progressively higher percentage of smaller particles than the original soil
mass. The in-channel phase is little influenced by rainfall and is
dependent mostly on the nature of the bed material and the transport
capacity of the flow. In this phase, the channel flow can usually
transport all the fine material supplied.
The Black Creek Watershed is an upland area. Sediment yield has been
monitored and some portions of the watershed have been modeled. However,
neither the actual or modeling results can be extended to the Maumee Basin
as a whole. On a long term basis, we might reasonably assume the major
streams in the basin to be in dynamic equilibrium so that the annual
sediment discharged into Lake Erie, except for man-made traps, equals the
average annual sediment yield from the upland areas into the main channels.
The average sediment yield into the main channels of the Maumee Basin might
be predicted if more watersheds similar to the Black Creek Watershed are
selected so that the soil, topographic, and land use conditions of the
basin are adequately represented and then a random generation of storms
representative of an average year is applied to the Basin.
STUDY RESULTS
-------
234
4.13 CONSTRUCTION, LAND USE DIFFERENCES
Two sets of data are included in this section. The first illustrates
water quality differences arising because of small scale differences in
land use. The second shows the impact of construction activities on water
quality.
On 28 March 1977 four discrete watershed areas (3 at 100 acres, 1 at
500 acres) of the Dreisbach Drain were sampled following a 1.5 inch
rainfall. At the outlet of each watershed, four grab samples were
collected at one-half hour intervals. No estimates of discharge were made,
and only concentration data are reported here. Samples were analyzed by
the Illinois Natural History Survey Laboratory. General land use, soils
and slopes were as nearly identical as could be expected under field
conditions. However, there were differences in conservation practices
applied to the four areas: watersheds A and B had no conservation
practices applied, watershed C was treated with an extensive terrace system
and the largest watershed, D, contained a grass waterway and a variety of
other practices, including terraces.
Water quality parameters that differed significantly among the four
watersheds are shown in Table 99. Dissolved solids and nitrate were lowest
from watershed A, moderate from B and D, and highest from C. The tile
outlet of the terrace system was enriched with dissolved ions and soluble
nitrogen. Factors responsible for this observation may include the greater
leaching of soluble ions due to the subsurface tile drainage acting alone
or in combination with heavier fertilizer application within this
watershed.
TABLE 99
Differing Parameters, Four Watersheds
Watershed
A
B
C
D
Dissolved
Solids
126 +- 4
185 +- 10
250 +- 7
170 +- 10
Nitrate
mg/1
2.2 4- 2.0
6.4 4- 1.1
8.8 +- .2
3.8 +- .5
Sediment
(Total Residue)
300 +- 65
1700 +- 458
700 4- 87
500 +- 71
Sediment (total residue) concentrations also were lowest from
watershed A, moderate from C and D, and highest from watershed B. Again
the factors responsible for these differences are complex and cannot be
precisely identified. In this situation other land use factors besides
conservation practices seemed to exert the greatest influence on water
quality. Watershed A had a high percentage of cropland with vegetative
cover, including the field immediately upstream from the sampling point,
and had the lowest sediment concentration. Conversely, watershed B had a
lower percentage of cropland with vegetative cover, including a fall plowed
field immediately upstream from the sampling point, and had the highest
sediment concentrations. The two watersheds with conservation practices
installed had sediment concentrations between these two extremes. To
STUDY RESULTS
-------
235
conclude, this small data set illustrates differences in the concentrations
of some water quality parameters resulting from small differences in land
use. Such data cannot establish cause-effect relationship between land use
and water quality, but are useful tests of the principles established
through more controlled experimentation.
Ten discrete watershed areas (350-550 acres) were sampled on 3 March,
4 March, 28 March and 23 April, 1977. At the outlet of each watershed two
samples were collected at two hour intervals on each of the sampling dates.
All sampling was done during moderate runoff events caused by one to two
inches of rainfall. Again only concentration data are reported and
analysis was conducted by the Illinois Natural History Survey Lab. Soils
and slopes differed considerably among the ten watersheds while general
land use and application of conservation practices differed to a lesser
degree. All these watershed characteristics certainly influence surface
runoff water quality, so an evaluation of the interactions of such a
complex set of variables with a small data set is difficult. However, the
data reveals a striking parallel between high sediment concentrations and
recent construction activities (jobs completed in September 1976).
To facilitate an easier pair-wise comparison of sediment
concentrations among the watersheds a sediment index was formed. Watershed
1 was consistently assigned an index value of 1.00. Index values for other
watersheds (2-13) were formed by dividing the sediment concentrations of
watersheds 2-13 by the sediment concentration of watershed 1. This was
done separately for all eight sampling intervals. The average sediment
indexes and standard deviations of the ten watersheds are shown in Table
100. Well established grass waterways tended to have the lowest sediment
indexes (1.0-1.6) and therefore had relatively low sediment concentrations.
Stable open ditches with good vegetative cover had higher index values
(1.6-3.0).
TABLE 100
Sediment Concentration Indexes - Subwatersheds
Watershed
Name
Dreisbach
Wertz
Richelderfer
West Gorrell
East Gorrell
Smith-Fry
Lake
Killian
Wertz Branch
Fuelling
No.
1
5
2
3
4
7
11
13
6
12
Channel
Classification
Grassed waterway
(upland)
Stable open ditch
(upland)
Stable open ditch
(lake bed)
Reconstructed open
ditch
Average
Sediment
Index
1.00
1.62
1.59
2.01
3.04
1.91
2.09
2.32
7.85
3.71
S. D.
0.00
0.40
0.30
0.30
1.19
0.65
0.53
1.36
3.33
1.71
STUDY RESULTS
-------
236
The East Gorrell watershed had the highest sediment index among the
stable open ditches. It is significant to not that a major in-channel
erosion control structure (drop structure) washed out on this ditch in the
spring of 1977. Failure of this structure resulted in rapid erosion of the
upstream channel as the excessive grade attempted to re-stabilize. The
data indicate that this process raised the sediment concentration of the
stream by 50% compared to the adjacent West Gorrell drainage. Recently
reconstructed open ditches with poor vegetative cover had the highest index
values (3.7, 7.9). The ditch banks were an obvious source of sediment and
the sampling reveals that in-stream sediment concentrations were elevated
4-8 times above the levels in surrounding streams. The large sediment
export from the Wertz Branch Ditch altered fish populations in the Wertz
Woods.
It is generally agreed that construction activities greatly
accelerates soil erosion. Evaluating the degradation of water quality by
construction site runoff was not specifically addressed in the Black Creek
project. However, in a grab sampling program of ten watersheds,
construction associated disturbances were an identifiable influence on
water quality (sediment concentrations). These findings raise concerns
about temporary negative effects on water quality created by the
installation of some soil conservation measures.
In the situations described above, the post-construction precautions
against erosion (seeding, mulching, etc.) were certified to meet SCS
specifications. However, an extremely dry fall resulted in very poor
vegetative cover so that erosion was greater than normal. Regardless, the
concerns about water quality degradation remain justified. Both near-site
and downstream changes in the stream ecology could damage aquatic
resources.
To lessen the environmental impact of conservation practice
installation, planners and field technicians need to be aware of stream
reaches with aquatic resource potential. In cases where aquatic resources
are judged worth protecting, more stringent precautions should be taken to
minimize construction damages. Personnel training programs stressing basic
ecological principles may be the shortest route to achieving maximum water
quality benefits from soil conservation practices.
DISCUSSION
-------
237
5.1 IMPACT OF LAND TREATMENT
The cost of land treatment in the Black Creek Watershed was not trivi-
al.
Costs by practice are included in Table 101. Total costs listed in
the table include contributions of individual landowners. These totaled
nearly $150,000 for the project.
Also included must be the cost of technical assistance furnished under
the grant by the Soil Conservation Service. Soil Conservation Service
costs are listed in Table 102.
TABLE 101
Cost by Practice
PRACTICE
District Cooperators
Conservation Plans
Landowner-District Contracts
uroup Contracts
Land Adequately Treated
Land Adequately Protected
Conservation Cropping System
Contour Farming
Critical Area Planting
Crop Residue Management
u i ver s i ons
Farmstead Windbreaks
Field Border
Field Wi ndbreak
Grade Stabilization Structure
(Including tile outlet CMP)
Grassed Waterway
Holding Ponds it Tanks
Land Smoothing
Livestock Exclusion
Livestock Watering Facility
Minimum Till age
Pasture Management
Pasture Planting
Pon.l
Protection Burins Developnent
Recreation Area Improvement
Sediment Control Basins
Stream Channel Stabilization
Streahbank Protection
St r i pcropp i ng
Surface D ra i ns
Terraces
Tile Drains
Tree PI an t i n^
Wildlife Habitat Management
Woodland Improved Harvesting
Woodland Improvement
Woodland Pruning
TOTAL (WATERSHED)
UMIT
No
Ho
No
Ho
Ac
Ac
Ac
Ac
Ac
Ac
Ft
Ac
Ft
Ft
!Jo
Ac
ilO
Ac
Rd
;io
Ac
Ac
Ac
.'In
Ac
Ac
Mo
Ft
Ft
Ac
Ft
Ft
Ft
Ac
Ac
Ac
Ac
Ac
Ac
GOAL
148
170
148
10,573
7, 'tis
7G9
10
7,491
39,200
75
288,370
12,003
3G8
C8
11
300
2,050
28
7 , D 5 G
1(02
501
39
118
12
0
6,000
122,000
300
90,500
22,000
200,300
10
222
200
BIO
50
12,038
ACCOMP-
Ll SMMEHT
I'll
133
11D
19
7,975
10,025
G,5iiS
10
15
2,952
1,860
I*
132, CCS
0
51C
6U
10
0
15,8GQ
7
P8P
17
n:
10
li
10
J
IB, 093
D9,30lj
0
t
22
20
3
?3
50
268
31
0
10
23 U
r-7
0
G7
9
0
0
TOTAL COST
( 1 nc 1 udes
Landowner Costs)
55,91(8
20
3,41(0
3,857
1,906
386
38,071
0
105,527
1(8,332
1(5,120
n
12,795
1,715
2,670
1 , !( 1 G
10,695
20,71(0
313
1,187
5,931
127,660
7S,CS2
0
3,71(1
39,031
152, 08G
0
1,952
0
0
0
$709.791
% OF
COST
SHARE
70
80
80
65
70
75
70
70
75
80
50
70
70
70
80
65
70
60
65
50
75'
80
70
SO
65
90
70
70
GO
65
70
70
HI STRICT
PAYMENT
518,876.59
513,876.59
i(,16it.OO
16.00
2,752.57
2,507.15
1,331(.81
289.79
26,G50.3C
0.00
79,11(5.91
38,666.21
22,560.09
0. 00
8,956.88
1,200.50
2,126.65
920.40
7,486.72
12,444.21
203.62
593.81
4,448.90
102,128.51
55,077.itO
0. 00
2,431.99
35,127.95
106,460.79
0.00
1,171.37
0.00
0.00
0.00
Av.$73 518,876.59
TOTAL
UMIT
COST
0.90
2.00
229.38
1.30
1.02
96.59
0.30
0. 00
294.51
755.19
4,512.91
0.00
0. 80
245.00
3.38
14.59
95.49
2,074.03
78.31
118.76
1,977.29
7.93
0.79
0.00
n.39
0. 75
1.13
0. 00
13.19
0.00
0. 00
0.00
DISTRICT
UNIT
COST
65.05
51.75
0.63
1.60
183.50
0.84
0.71
72.44
0.20
0.00
153.38
604.15
2,256.00
0. 00
0. 56
17l'.50
3.10
9.48
66.84
1,244.42
50.90
59.38
1,482.96
6.34
0.55
0. 00
0. 25
0.68
0. 79
0. 00
7. 91
0. 00
0. 00
0.00
DISCUSSION
-------
238
TABLE 102
Technical Assistance Cost SCS
Assistance
Professional
Subprofessional
Total
Planning time
Application time
Group application time
Cartographic
Soils Mechanics Lab
Travel and support
GRAND TOTAL
Hours
14,509
8,284
22,793
6,099
4,400
3,557
Dollars
$159,345.47
70,746.90
230,092.37*
44,736.44
33,677.28
29,639.92
1,223.53
1,761.75
2,147.35
235,225.00
*Includes employee benefits and overhead.
Costs for land treatment thus total nearly $910,000 for the 12,000
acre Black Creek watershed. The analysis of this cost and its relationship
to the Black Creek watershed, the Maumee Basin, and other agricultural
watersheds is included in the next section.
5.1.1 Analysis of Costs
As previously stated, a total conservation approach to any watershed
must involve more than a consideration of water quality. The initial plan-
ning for the Black Creek project listed 33 categories in which the costs
could be incurred for specific conservation practices. These were outlined
in Table A-10 of the initial work plan according to estimated costs per
year. Table A-10 is reproduced here as Table 101.
The practices listed in Table 101 were described in detail in the work
plan as follows:
5.1.1.1 CONSERVATION CROPPING SYSTEMS
Growing crops in combination with needed cultural and management meas-
ures. Cropping systems include rotations that contain grasses and legumes
as well as rotations in which the desired benefits are achieved without the
use of such crops.
5.1.1.2 CONTOUR FARMING
Farming sloping cultivated land in such a way that plowing preparing
and planting, and cultivating are done on the contour. (This includes fol-
lowing established grades of terraces, diversions, or contour strips.)
DISCUSSION
-------
239
5.1.1.3 CRITICAL AREA PLANTING
Stabilizing silt-producing and severely eroded areas by establishing
vegetative cover. This includes woody plants, such as trees, shrubs or
vines, and adapted grasses or legumes established by seeding or sodding to
provide long-term ground cover. (Does not include tree planting mainly for
the production of wood products.)
5.1.1.4 CROP RESIDUE MANAGEMENT
Using plant residues to protect cultivated fields during critical ero-
sion periods.
5.1.1.5 DIVERSIONS
A channel with a supporting ridge on the lower side constructed across
the slope.
5.1.1.6 FARMSTEAD AND FEEDLOT WINDBREAKS
A belt of trees or shrubs established next to a farmstead or feedlot.
5.1.1.7 FIELD BORDER PLANTING
A border or strip of perennial vegetation established at the edge of a
field by planting or by converting from trees or crop production to herba-
ceous vegetation or shrubs.
5.1.1.8 FIELD WINDBREAKS
A strip or belt of trees or shrubs established to reduce wind erosion.
5.1.1.9 GRADE STABILIZATION STRUCTURE
A structure to stabilize the grade or to control head cutting in na-
tural or artificial channels. (Does not include stream channel improve-
ment, streambank protection, diversion, or structure for water control.)
5.1.1.10 GRASSED WATERWAYS
A natural or constructed waterway or outlet shaped or graded and esta-
blished in vegetation suitable to safely dispose of runoff from a field,
diversion, terrace or other structure.
5.1.1.11 HOLDING PONDS AND TANKS
A fabricated structure or one made by constructing a pit dam or em-
bankment for temporary storage of animal or agricultural wastes, associated
runoff and waste water.
5.1.1.12 LAND SMOOTHING
Removing irregularities on the land surface by use of special equip-
ment.
DISCUSSION
-------
240
5.1.1.13 LIVESTOCK EXCLUSION
Excluding livestock from an area where grazing is not wanted.
5.1.1.14 LIVESTOCK WATERING FACILITY
A trough or tank with needed devices for water control to provide
drinking water for livestock.
5.1.1.15 MINIMUM TILLAGE
Limiting the number of cultural operations to only those that are
properly timed and essential to produce a crop and prevent soil damage.
5.1.1.16 PASTURE AND HAYLAND MANAGEMENT
Proper treatment and use of pastureland or hayland.
5.1.1.17 PASTURE AND HAYLAND PLANTING
Establishing and re-establishing long-term stands of adapted species
of perennial, biennial, or reseeding forage plants. (Includes pasture and
hayland renovation; does not include grassed waterway or outlet on crop-
land.)
5.1.1.18 PONDS
A water impoundment made by constructing a dam across a water-course
or a natural basin, or by excavating a pit or "dugout". (Such ponds do not
include spring development or irrigation reservoirs.)
5.1.1.19 PROTECTION DURING DEVELOPMENT
Treatment based on a plan to control erosion and sediment during
development for residential, commercial-industrial, community services,
transportation routes or utility uses.
5.1.1.20 RECREATION AREA IMPROVEMENT
Establishing grasses, legumes, vines, shrubs, trees, or other plants
or managing woody plants to improve an area for recreation.
5.1.1.21 SEDIMENT CONTROL BASINS
A barrier or dam constructed across a waterway or at other suitable
locations to form a silt or sediment basin.
5.1.1.22 STREAM CHANNEL STABILIZATION
Stabilizing the channel of a stream with suitable structures.
5.1.1.23 5TREAMBANK PROTECTION
Stabilizing and protecting banks of streams or excavated channels
against scour and erosion by the use of vegetative or structural means.
DISCUSSION
-------
241
5.1.1.24 STRIPCROPPING
Growing crops in a systematic arrangement of strips or bands on the
contour to reduce erosion.
5.1.1.25 SURFACE DRAINS
A graded channel for collecting excess water within a field. This
does not include grassed waterway or outlet.
5.1.1.26 TERRACE, GRADIENT
An earth embankment or a ridge and channel constructed across the
slope at a suitable opening and an acceptable grade to reduce erosion dam-
age and pollution by intercepting surface runoff and conducting it to a
stable outlet.
5.1.1.27 TERRACE, PARAT.T.FL
An earth embankment or a ridge channel constructed in parallel across
the slope at a suitable spacing and acceptable grade to reduce erosion and
pollution and provide a more farmable terrace system.
5.1.1.28 TILE DRAINS
A conduit, such as tile, pipe or tubing, installed beneath ground sur-
face and which collects and/or conveys drainage water. The project goal
was approximately 200,300 lineal feet needed only for erosion and sediment
control of surface drains and grassed waterways.
5.1.1.29 TREE PLANTING
Planting tree seedlings or cuttings.
5.1.1.30 WILDLIFE HABITAT MANAGEMENT
Retaining, creating, or managing wildlife habitat for both upland and
wetland.
5.1.1.31 WOODLAND IMPROVED HARVESTING
Systematically removing some of the merchantable trees from an imma-
ture stand to improve the conditions for forest growth.
5.1.1.32 WOODLAND IMPROVEMENT
Improving woodland by removing unmerchantable or unwanted trees,
shrubs or vines.
5.1.1.33 WOODLAND PRUNING
Removing all or parts of selected branches from trees.
The Black Creek project was an experimental project. It was begun as
an attempt to evaluate several practices. Of an initial list of 32 prac-
DISCUSSION
-------
242
tices, previously described, 12 have been selected as best management prac-
tices. These are listed in Table 103, along with total unit cost of in-
stallation. These costs are based on the actual contracts awarded in Black
Creek.
TABLE 103
Cost of BMP Installation
Practice
Field Border
Holding Tanks
Sediment Basins
Contour Farming
Critical Area Planting
Crop Residue Management
Grassed Waterways
Livestock Exclusion
(fencing)
Reduced Tillage
Pasture Renovation
and Planting
Terraces
Unit
Mile
Each
Each
(1)
Acre
(2)
Acre
Foot
(3)
Acre
Foot of terrace
(with tile)
Cost
1,584
5,600
5,000
(1)
400
(2)
1,200
0.50
(3)
100
1.75
(1) very little application in Black Creek Area or
Maumee Basin.
(2) can be applied by management techniques without
additional cost.
(3) considered only on soils where reduced tillage should not
result in significant yield penalties.
At the completion of the project new estimates were made of the
amounts of each of these practices which should have been installed to meet
water quality goals.
It should be emphasized that these amounts do not represent totals ac-
tually installed, but rather the amounts that should have been installed if
the designers of the project could have known the results from the Black
Creek project when it began. In some cases the amount is greater than the
amount actually applied. These amounts as estimated for Black Creek
watershed are set out in Table 104.
These amounts can be multiplied by the cost of land treatment practices to
derive estimated costs for the project if all recommended practices had
been applied and if no practice were applied that were not needed. These
results are presented in Table 105.
Amounts of tnese practices can be projected to the total basin on to
the basis of the similarities between the Black Creek Watershed and the
basin. This projection is made in the following discussion, not as a plan
for the Maumee, but as an example of how Black Creek cost figures can be
DISCUSSION
-------
243
TABLE 104
Estimated BMPs Needed at Beginning of Black Creek Project
Practice
Field Borders
Holding Tanks
Sediment Basins
Critical Area Planting
Grassed Waterways
Livestock Exclusion
(fencing)
Pasture Renovation
and Planting
Terraces
Amount Needed
40 miles
10
6
10 acres
68 acres
15,000 feet
400 acres
44,000 feet
TABLE 105
Projected Costs of Ideal Black Creek Land Treatment
Practice Cost
Field Border 63,360
Holding Tanks 56,000
Sediment Basins 30,000
Critical Area Planting 4,000
Grassed Waterways 81,600
Livestock Exclusion
(fencing) 7,500
Pasture Renovation 4,000
Terraces 77,000
Total 323,460
applied. The following rationale was used.
Field Borders: In Black Creek, a field border is defined as a 16-foot
strip of vegetation (usually sod) placed along a water course in cropland.
It is estimated, based on figures supplied by the Soil Conservation Service
in Ohio, that 13,800 miles of such stream bank exists in the basin. The
total amount of field border needed would thus be 13,800 miles times two or
27,000 miles. This amount multiplied by a unit cost of 1584 yields an es-
timated basin cost of $42.7 million.
Holding Tanks: In Black Creek, about 1/5 to 1/4 of the livestock
operations needed holding tanks to improve water quality. If the same ra-
tio holds in the Maumee Basin, the amount needed would be 800 to 1000.
Sediment Basin: There were six sites in Black Creek for sediment
basins. Multiplication this by the ratio of Black Creek area to basin area
would result in 2100 sites. Since the erosive area of the watershed is
DISCUSSION
-------
244
only 2/3 that of the basin, the number of sites could be as low as 1400.
Critical Area Planting: It was estimated that 10 acres of critical
area planting were needed in the Black Creek area. Projecting this amount
to the total basin by the previously described method results in an esti-
mate of 2500 to 3500 acres for the basin.
Grassed Waters: Applying the same types of estimates to grassed wa-
terways results in the following estimates: 16,000 - 24,000.
Livestock Exclusion: Applying these factors plus a correction based
on the fact that livestock is more common in Black Creek than in the Basin
by about a 3/2 ratio, results in an estimate of 500 miles of fencing need-
ed. This is higher by a factor of 10 than the amount estimated in other
studies, which may indicate that extensive fencing has already been done.
Pasture Planting: In Black Creek, it was determined that about 4/5 of
the existing pasture needed replacement or renovation. Applying this to
the 125,000 acres of pasture results in an estimated 100,000 acres.
Terraces: Applying the standard corrections to the 44,000 feet of
terrace needed in Black Creek results in 12,000,000 to 15,000,000 feet
needed in the Basin.
These estimates, multiplied by the Black Creek costs give a cost esti-
mate for the basin. This is shown in Table 106.
TABLE 106
Estimated Costs - Maumee Basin
Practice Cost
Field Borders 43,700,000
Holding Tanks 4,480,000-5,600,000
Sediment Basins 7,000,000-10,500,000
Critical Area Planting 1,000,000-1,400,000
Grassed Waterway 19,200,000-28,000,000
Livestock Exclusion 1,300,000
Pasture Planting 10,000,000
Terraces 21,000,000-26,250,000
Total 107,480,000-126,250,000
In the preceding analysis, no costs are assigned for systems related
to conservation tillage, although crop residue management and conservation
tillage are both considered to be important water quality management prac-
tices in the Black Creek Watershed and the Maumee Basin. In fact, it is
estimated that most of the approximately 4,600 acres of land in the Black
Creek Watershed which has an erosion problem and which is utilized for row
crop production and most of the 1.4 million acres in the Maumee Basin which
is in crop production and which is subject to erosion, would benefit from a
water quality standpoint from conservation tillage practices. However, it
is further believed that careful selection of the areas where this is un-
DISCUSSION
-------
245
dertaken and a willingness to phase-in new tillage systems can result in
achieving this practice at virtually no cost.
Capital costs associated with changes of tillage such as an esti-
mate of the cost of chisel plows or no-till planters that would be needed
have been included in some estimates of cost of conservation application
in the Maumee Basin. However, if the purchase of this equipment is allowed
to be phased-in so that it can be obtained as a part of normal equipment
replacement system of individual farms, the costs are greatly reduced.
Landowners regularly replace equipment to maintain complex modern farms.
The cost of electing to purchase a new chisel plow rather than a new mold-
board plow is not a significant one.
Other costs associated with changes in tillage relate to normal
operating costs and any yield reduction that might be associated with the
new tillage.
In the Maumee Basin and Black Creek it can be expected that certain
operational costs those associated with depreciation of equipment, fuel
for tillage operations and labor in tillage operations will be reduced by
reducing the amount of tillage done. On the other hand, certain costs,
largely those associated with chemical weed or pest control, will be higher
in reduced tillage systems. As a rule of thumb, for both corn and soybe-
ans, it can be assumed that the savings on one hand and the increased chem-
ical costs on the other will likely balance, leaving the impact on yield as
the cost most associated with conservation tillage.
In the total basin, a relatively small amount of the land, about
170,000 acres, can be considered suitable for greatly reduced tillage or
no-till systems. This represents cropland that can be expected to have an
erosion problem, but which does not have a wetness hazard. About 1.2 mil-
lion acres of the basin has both an erosion hazard and a wetness hazard.
Tillage systems which leave large quantities of surface residue will reduce
drying of these soils in the spring and will slow warming of the soil.
During years with wet and cool springs, significant yield reductions can be
anticipated on this type of land.
In the Black Creek project, conservation tillage was defined as a til-
lage system which included one primary tillage operation (spring or fall),
one secondary tillage operation (spring, no more than one field cultiva-
tion, and no moldboard plowing. This system reduced erosion and was, in
some cases, superior to conventional tillage.
It is believed that this type of tillage system could be extended into
most of the erosive areas of the Maumee Basin without danger of significant
yield reduction, particularly in corn and soybean rotations. In any long-
term attempt to introduce conservation tillage, however, the option to
periodically moldboard plow to aid in weed or disease control should be
given landowners.
5.1.2 Arriving at BMP's
The best management practices recommended for water quality improve-
ment in the preceding discussion were selected in relationship to water
quality. Water quality is not, however, the only goal that needs to be
considered in the planning of the best land use within an agricultural
DISCUSSION
-------
246
watershed.
Conservation practices in the Black Creek project have been arbitrari-
ly categorized in four separate classification:
1. Those which benefit water quality
2. Those which protect the soil resource
3. Those which enhance production capability
4. Those which accomplish other conservation purposes
The wise use of the land resource requires that all of these objec-
tives be considered when planning is undertaken. In fact, a watershed
which has been properly managed will have elements of each of these
categories and an ideally managed watershed can be considered one in which
all of the four classifications is maximized. Unfortunately, in the real
watershed which must be considered, these classifications can be conflict-
ing. It may not, for example, be possible to have maximum crop production
and maximum water quality at the same time. It may also not be possible to
have maximum water quality, as defined by concentrations of nutrients, and
maximum wildlife production in the same watershed at the same time.
For example, as has been previously discussed, sediment may be reduced
by placing large quantities of stone on eroding channel banks and degrading
channel bottoms. This, on the other hand, may interfere with fish communi-
ties through the reduction of habitat diversity.
Conversely, some practices may serve more than one of the four clas-
sifications, h practice may, for example, serve to both preserve the soil
resource and improve water quality. A practice such as converting erosive
crop land into a wildlife habitat might serve the purpose of soil protec-
tion, improvement of water quality, and the additional conservation purpose
of wildlife protection.
While this is very useful from an environmental viewpoint, it is much
less simple from an administrative viewpoint. Precisely, if a practice can
be said to have both a water quality benefit and a soil protection benefit,
and some agency is funding water quality improvement programs only, does
that practice meet the requirement of water quality improvement. Similar-
ly, if a project improves water quality, but has an adverse effect on
wildlife, and these two aspects are being regulated by different agencies,
how is a conflict to be resolved? In order to obtain some insight into
these kinds of administrative questions, all of the practices in the Black
Creek project were evaluated on the basis of each of the four classifica-
tions. These results are included in Table 107.
This analysis was then utilized to assign Black Creek actual land
treatment costs to each of the four categories. The results are included
in Figure 33.
DISCUSSION
-------
247
As can be seen from Table 107 and Figure 33, a large share of funds
spent in the Black Creek demonstration project went for practices that were
not directly related to water quality. On the other hand, according to the
TABLE 107
Evaluation of Practices Applied in Black Creek Watershed
Practi ce
Conservation Cropping Cystcn
Contour Farming
Critical Area Planting
Crop Residue Management
Di vers i ons
Farmstead Windbreaks
Field Border
Field Windbreaks
Grade Stabilization Structures (1)
(a) aluminum toewall overfall str.
(b) surface water inlet pipes
(c) tile outlet protection pipes
Grassed Waterway
Holding Pond and Tanks (Animal '/aste)
Land Smoothing
Livestock Exclusion
Livestock Watering Facility
Mi n imum Till age
Pasture Management
Pasture Planting
Pond
Protection During Development
Recreation Area Improvement
Sediment Control Has ins
Stream Channel Stabilization (2)
Strear.ibank Protection (3)
Str i pcroppi r\r.
Surface Drains
PTO Terraces
Ti le Drains ( k )
(a) it inch
(b) b inch
(c) G inch
(d) 8 inch
(e) 10 inch
(f) 12 inch
(g) Ik inch
(h) 15 inch
Tree Planting
Wildlife Habitat Management
Woodland Improved Harvesting
Woodland Improvement
Woodland Pruning
Water
Qual i ty
2
1
1
1
2
1>
1
2
o
L.
1
1
k
1
2
1
1
1
2
2
2
1
2
3
2
3
1
3
2
3
3
3
3
So] 1
Protect ion
1
1
1
1
1
i
)
1
1
1
3
3
1
2
1
1
1
2
1
2
3
1
2
1
3
1
2
2
3
3
2
2
Crop
Product i on
j
3
3
2
3
i*
it
3
3
3
2
1
1*
It
3
2
2
it
It
it
it
3
1
2
1
2
1
k
k
2
2
2
Other
Conservation
Uses
2
3
2
2
1*
1
o
2
it
3
it
it
2
3
2
3
2
1
2
1
2
3
2
2
it
2
3
1
1
1
1
1
(footnotes Next Pagel
DISCUSSION
-------
248
Pi rect Benefi t
1 -- Primary Benefit 2 -- Secondary Benefit 3 -- Very little, if
any Benefit k -- Ho Benefit
(1) Includes both over-fall structures and surface inlet pipes.
If we only considered over-fall structures we might justify
primary benefit (1) to water quality, however, surface pipes
are definately not primary for water quality.
(2) Stream channel stabilization includes all use of rip-rap
stone for stabilizing channel banks and bottoms. Rock drop
structures are included here. The practice should be limit-
ed to critical eroding areas only.
(3) Streambank protection involved using mechanical hoes to pull
channel banks back to stable slope and vegetating with grass
mixtures. In some case material v/as removed from the chan-
nel bottom to improve drainage capacity of the channel.
(U) Tile drains are primarily for removing subsurface water to
improve crop production by reducing the wetness problems
however, tile drains are a necessary part of some water
quality BMP's such as grass waterways and terraces.
* Total cost for W/Q BMP's Applied = 211,383.49.
** Total District Payment for water quality BMP applied =
152,229.88.
Total BMP cost per acre in watershed (12,000 Ac) over five years
= 17.bl or 3.52/yr.
District Payment for BMP per acre in watershed over 5 years =
12.68
-------
1. Field Border
CO
WATER
QUALITY
ONLY
WATER
QUALITY
AND
SOIL
RESOURCE
PROTECTION
SOIL
PROTECTION
ONLY
PRODUCTION
CROP
OTHER
CONSERVATION
USES
Black CreekCost Shai
(Total per Practice)
26,650.36
4,448.90
38,666.21
70,020.21 (with
71,200.45 (with
1,334.81
78.822.00
55,077.40
District Cost for
$53.659.35
Water Quality Only
District Cost for
Practices
with Primary
$66,003.53
Benefit to Both
Water Quality
and Soil Protection
District Cost for
$186,652.94
Soil Protection Only
District Cost for
Crop Production
$96,537.18
District Cost for
$13.327.81
Other
Conservation Uses
Cost
by Catagory
Funding Sources
EPA 208 Funds
Culver Amendment
Clean Water Grants
EPA 208 Funds
Culver Amendment
Clean Water Grants
ACP Funds
Great Plains Funds
566 Watershed Funds
State Cost-Share
Fund for Erosion
Control Programs
ACP Funds
Great Plains Funds
566 Watershed l-unds
State Funds
USDA Loans
Production Credit
Loans
Small Business Loans
Department of
Interior Funds
Natural Resource
Funding
Fishery Incentives
Wildlife Incentive
Programs
Technical and
SWCD's SCS
Extension, EPA
Water Quality Agencies
EPA
SWCD's
SCS
Extension
ASCS
State Conservation
Agencies
Water Quality
Agencies Others
SWCD's
SCS ASCS
Extension State
Conservation
Agencies
SWCD-SCS
Extension. ASCS
Private Lending Firms
Department of Interior
State Natural
Resources
SWCD's SCS
Extension, Others
mplrance Mechanism
Cost-Sharing Incentives
Education. Tech. Ass.
208 Regulations
Cost-Sharing Incentives
Education Technical
Assistance
State and County
Erosion Control Laws
Cost-Sharing Incentives
Education
Technical Assistance
Erosion Control
Regulation
Production Stimulus
Education
Financing Profit
Cost-Sharing Incentive:
Education
Technical Assistance,
Ordinances
Watershed
System
Adequately
Treated
Figure 33. Assignment of Costs to Categories
NJ
-------
250
previous analysis, only slightly more than $320,000 is now thought to have
been necessary to have achieved water quality objectives through land
treatment. The primary reason for the substantial differences between
these two cost figures is the research and demonstration nature of the pro-
ject.
One weakness of all of the above analyses from the standpoint of water
quality is the inability at this time to precisely relate costs of land
treatment to levels of water quality. It is not possible, for example, to
state that the $212,000 spent on water quality practices in the Black Creek
Watershed improved the quality of water in Black Creek by some percentage.
It is also not possible to project what spending $323,000 on water quality
practices, would have accomplished; and it is not possible to make an esti-
mate of how much the Maumee River or Lake Erie would benefit by carrying
out all of the recommended practices.
The ANSWERS model, which is discussed in detail elsewhere in this re-
port, provides a mechanism by which these benefit and costs can be predict-
ed for small watersheds, and through the judicious selection of sample
watersheds (as the Black Creek area was selected to represent the Maumee
Basin) to much larger areas. The ANSWERS model will, in its final form, be
capable of predicting improvements in water quality accomplished by specif-
ic land treatment practices. By a technique similar to that with which
Black Creek costs were projected to the entire basin, water quality im-
provements can be projected from carefully modelled sample watersheds.
5.1.3 Engineering Observations
Three engineering practices were investigated in the Black Creek
Watershed which were previously not common practices within the Soil Con-
servation Service in Indiana. Two of these practices involved the use of
stone for channel protection. The third involved the application of ter-
races in areas where terraces had not been normally be applied in the past.
Terraces applied in the Black Creek project were of the parallel type
with tile outlets, PTO. A typical arrangement is shown in Figure 34.
These terraces, with tile outlets, differ from graded terraces in that
they are laid out parallel to each other across the water course graded
slope. They use tile to carry away the water. They serve the dual purpose
of reducing slope length (and thereby erosion) and temporary retention of
surface water which in turn allows some sediment and attached pollutants to
settle out of the water before it enters the subsurface drain.
PTO terraces as used in Black Creek are, in general, shorter than con-
ventional PTO terraces. They are not necessarily constructed perpendicular
to the slope but are perpendicular to the minor water courses. A series of
terraces are installed, designed according to SCS standards so as to
prevent overtopping during a 10 year frequency storm event.
The number and size of terraces needed is dependent on the size of the
drainage area being treated. Outflow from the terrace is accomplished by a
riser pipe, upstream from the terrace. An orifice plate determines the
rate at which water will be discharged from the ponding behind the terrace
into the subsurface outlet. The size of the orifice determines maximum
flow. Outlet tile need be designed only large enough to accomodate the
DISCUSSION
-------
251
Figure 34. Typical PTO Terrace Arrangement
flow through this metered orifice. Orifice size and tile size are deter-
mined on the basis of that necessary to allow discharge of all of the pond-
ed runoff from the design storm in a 24 hour period.
Longer versions of the PTO terrace serve the same purpose as do con-
ventional grass-back or broad-based parallel terraces,perpendicular to the
field slope. In some cases, terraces were installed in the Black Creek
area specifically for their sediment control functions. These terraces
generally involve smaller areas of steeper side slopes and function pri-
marily as temporary sediment basins.
Rock was used in the Black Creek project, both as an attempt to remove
velocity from areas of rapid fall in stream channels and in unstable soil,
DISCUSSION
-------
252
to prevent undermining of the channel at the seepage line.
Details of one of these structures, intended for protection of un-
stable channels are included in Figure 35. In Black Creek, stone for pro-
tection was established to elevations above the seepage line of one, two
and three feet for experimental purposes. Cost of this installation was
roughly $3 per lineal foot for the one foot series, $6 for the two foot
series, and $9 for the three foot series. Observations have indicated that
all three types of installations are equally satisfactory in reducing chan-
nel instability. It is therefore as successful to use the one foot eleva-
tion as it is the three foot elevation resulting in a potential practical
saving of $6 per linear foot in areas suited to the practice application.
Structures called rock-drop structures were included in the project at
areas where rapid change in channel elevation produced erosion or bank sta-
bility problems through the high velocity of the water.
5.2 SOURCES OF SEDIMENT AND RELATED POLLUTANTS
5.2.1 Comparison of Subwatersheds
5.2.1.1 INTRODUCTION
The Maumee River is usually a gently flowing river and even in flood
periods the velocities are relatively low because the gradient is not
steep. However, the waters in the Maumee River never appear clear because
of its suspended sediment load. For a 10-year period of record, 1961-71,
the sediment rate from the Maumee River into Lake Erie averaged about 500
kg/ha (a kilogram per hectare is roughly equivalent to a pound per acre)
annually (1,2). However, sediment yields even of this rather low order of
magnitude may be important in lowering the water quality in Lake Erie be-
cause they are composed mostly of colloidal-sized particles. For a small
volume, the sediments can carry a high nutrient load, particularly phos-
phorus. Unknown at this time, however, is the contribution which phos-
phorus attached to sediments actually makes to the eutrophication rate of
Lake Erie. It is implicitly assumed in this study that some of the at-
tached phosphorus becomes available to the eutrophication process occurring
in Lake Erie.
This section reports on the sediment and nutrient yields from the
Black Creek Watershed into the Maumee River. In particular, it compares
the data collected from two major drainage areas within the watershed.
However, the reported results are based only on a two-year period of
record, 1975 and 1976, which is a rather short hydrological period on which
to make definite conclusions. As a consequence, the conclusions reached
are subject to further modification with succeeding years of record.
5.2.1.2 STUDY AREA AND METHODS
The Black Creek Watershed was chosen as being fairly representative of
the soils and agricultural practices of the Maumee Basin although it is
only 4950 ha in size compared to 1,711,500 ha for the Maumee Basin. An
outline map of the experimental watershed together with instrumentation and
sampling sites is shown in Figure 36.
DISCUSSION
-------
253
Channel Bottom
«* * #1 Crushed Stone
« * -«
o° «' o
.. .%.«*.*.
#1 Crushed Stone
#1 Crushed Stone
Figure 35. Details of Channel Protecting Structure
DISCUSSION
-------
254
Kaingage sites
Stage recorder sites
A Sampling sites
^ FUELLING DRAIN
MAUMEE
RIVER
Figure 36. Instrumentation Sites
The soils in the Black Creek Watershed can be roughly divided into two
categories those soils which were formed entirely in glacial till and
DISCUSSION
-------
255
those soils which were also influenced at various stages in their formation
by shallow water cover or by wave action. These soil categories are subse-
quently referred to as glacial till soils and lake plain and beach ridge
soils, respectively. Indiana Highway 37 as shown in Figure 36 divides
these soils fairly well with the glacial till soils to the north and the
lake plain and beach ridge soils to the south. The glacial till soils are
gently rolling while the lake plain soils are nearly level.
Two of the major drainage areas which empty into Black Creek, that of
the Dreisbach Drain and Smith-Fry Drain, were studied intensively. The
Dreisbach Drain is located along the western boundary of the watershed and
the Smith-Fry Drain is located along the eastern boundary. Their drainage
areas, of comparable size, represent the greatest contrast in soils and
land use within the watershed. The drainage area of the Dreisbach Drain
contains 74 percent rolling and 26 percent nearly level topography while
that of the Smith-Fry Drain contains only 29 percent rolling and 71 percent
nearly level topography. The land use is also quite different with 35 per-
cent of the drainage area of the Dreisbach Drain in row crops as compared
to 63 percent for the drainage area of the Smith-Fry Drain. The drainage
area of the Dreisbach Drain also contains the town of Harlan which has an
effect on water quality in that stream. Characteristics of these two
drainage areas as well as those for the Black Creek Watershed are given in
Table 108. Note that characteristics of the total watershed are very simi-
lar to those for the drainage area for the Smith-Fry Drain.
TABLE 108
Characteristics of the Areas Studied
Characteristics
Drainage area:
Soil groups:
Lake plain & beach ridge
Glacial till
Land use:
Row crops
Small grain & pasture
Woods
Urban, roads, etc.
Homes :
Black Creek
Watershed
4950 ha
64%
36%
58%
31%
6%
5%
Smith-Fry
Drain
942 ha
71%
29%
63%
26%
8%
3%
28
Dreisbach
Drain
714 ha
26%
74%
35%
48%
5%
12%
143
Sediment and nutrient yields from the Black Creek Watershed and the
drainage areas for the Dreisbach Drain and Smith-Fry Drain were determined
by integrating sediment and nutrient concentrations with flow rates.
Stage-discharge relationships were developed for the outlets of these study
areas to give flow rates. Water stages were recorded continuously at these
locations with a pressure-type stage recorder (Model 12 Flow Recorder, Fox-
boro). (Product descriptions and manufacturers are given for reader infor-
mation and should not be construed as endorsements.)
Water samples for determining the concentrations of sediment and nu-
DISCUSSION
-------
256
trients were collected either manually or with automated samplers. Grab
samples were collected each week and also during storm events. The au-
tomated samplers were triggered at a set minimum stage and then continued
to operate automatically until the sample storage was exhausted or the
stage fell below the set minimum stage. The water samples were normally
collected and the automated sampler reset before the sample storage was ex-
ceeded.
Three automated pumping samplers (PS-69, U.S. Interagency Sedimenta-
tion Project) were installed at the junctions of the Dreisbach and Smith-
Fry Drains with Black Creek and on the main stem of Black Creek near its
entrance into the Maumee River. Each sampler was capable of automatically
collecting 72 samples of 500 ml each at a chosen time interval.
After the samples were collected, they were frozen within 24 hours.
Before analysis, the samples were thawed and one-half of the sample fil-
tered. Suspended sediment was determined by passing 200 ml of runoff
through a tared membrane filter (0.40 pore diameter, Nucleopore) and then
weighing the collected solids after oven drying at 105 degrees C for 24
hours.
The nutrients analyzed were nitrogen and phosphorus and their consti-
tuent forms. Ammonium and nitrate in the filtrate were determined by the
method of Bremner and Kenney (3). Total nitrogen in the filtered and un-
filtered samples were determined by the method of Nelson and Sommers (4).
Soluble inorganic phosphorus in the filtrate was analyzed by procedures
outlined by Murphy and Riley (5). And total phosphorus was determined by
the method described by Sommers and Nelson (6). A detailed description of
the analyses procedures is given in the Section 0.0.2.1 of this volume.
5.2.1.3 RESULTS AND CONCLUSIONS
The results and conclusions associated with sediment and nutrient
yields are based on two years of record 1975 and 1976. Precipitation
for 1975 was about 20 percent over normal, but for 1976 it was about 20
percent less than normal. Fortunately these two years represent as wide a
variation in precipitation amounts and patterns as will likely occur over a
lengthy period of record.
For better comprehension of the data, the reader should remember that
1975 was a wet year and 1976 was a relatively dry year. Also the reader
should remember that the drainage area for the Dreisbach Drain was more
rolling and less intensively farmed than the drainage area for the Smith-
Fry Drain. Also as an aid to the reader, conclusions are reported under
each subheading while the information upon which they are based is still
apparent to the reader.
5.2.1.3.1 Values of Important Parameters
The rainfall, runoff, sediment yields and total nitrogen and phos-
phorus yields for the Black Creek Watershed and the drainage areas of the
Dreisbach Drain and Smith-Fry Drain are given in Table 109.
Note under the column for Black Creek Watershed that a 40 percent
reduction in rainfall from 1975 to 1976 resulted in 60 percent reduction in
runoff, a greater than fourfold reduction in sediment yield, a greater than
DISCUSSION
-------
257
TABLE 109
Rainfall and Runoff Amounts and Yields of Sediment and Nutrients
Parameter
Rainfall
Runoff
Sediment
Total N
Total P
Year
1975
1976
1975
1976
1975
1976
1975
1976
1975
1976
Black
Creek
Watershed
112
70
27.5
11.2
2370
530
48.7
8.6
5.2
1.1
cm
cm
cm
cm
kg/ha
kg/ha
kg/ha
kg/ha
kg/ha
kg/ha
Smith-Fry
Drain
112
70
29.1
12.4
2130
640
53.2
10.3
5.4
1.1
cm
cm
cm
cm
kg/ha
kg/ha
kg/ha
kg/ha
kg/ha
kg/ha
Dreisbach
Drain
112
70
26.0
10.1
3740
380
44.1
6.6
5.0
1.0
cm
cm
cm
cm
kg/ha
kg/ha
kg/ha
kg/ha
kg/ha
kg/ha
fivefold reduction in total nitrogen discharged, and about a fivefold
reduction in total phosphorus discharged from the watershed. These data
clearly demonstrate the adverse effects of excess rainfall.
Conclusion: Amounts of runoff, sediment, and nutrients discharged from
a small watershed are greatly affected by rainfall. Reductions in rainfall
give a greater percentage reduction in runoff which in turn gives a still
greater percentage reduction in sediment and nutrient transport.
The runoff and sediment yield data for the drainage areas of the
Smith-Fry Drain and the Dreisbach Drain present an interesting comparison.
The runoff from the drainage area for the Smith-Fry Drain was greater
than that for the Dreisbach Drain in both 1975 and 1976. Although the
drainage area for the Smith-Fry Drain was more level than the drainage area
for the Dreisbach Drain, it also had better subsurface drainage and base
flow also seemed to be sustained for longer periods of times from interflow
through the ditch banks.
In 1975, the year with above normal rainfall, the sediment yield at
the outlet of the Dreisbach Drain was about twice that at the outlet of the
Smith-Fry Drain. However, in 1976, the reverse was true. Although all
values were much lower, the sediment yield at the outlet of the Smith-Fry
Drain was about twice that at the outlet of the Dreisbach Drain. The
better land use in the drainage area of the Dreisbach Drain was apparently
sufficient to retard runoff and subsequent erosion more than in the
drainage area of the Smith-Fry Drain with its more intensive land use.
However, when rainfall was excessive, as in 1975, greater runoff and subse-
quent erosion occurred on the drainage area of the Dreisbach Drain than on
the drainage area of the Smith-Fry Drain because of steeper slopes on the
drainage area of the Dreisbach Drain.
Conclusion: During years with above average rainfall, land slope is
clearly the dominant factor affecting sediment yields. However, with below
average rainfall, "the effect of land use on sediment yields becomes rela-
tively more important. This reflects the natural sequence of rainfall-
DISCUSSION
-------
258
runoff events because rainfall must first meet the storage capabilities of
the soil and land surface before runoff begins.
5.2.1.3.2 Nutrient Transport
The total nutrient yields for 1975 and 1976 from the Black Creek
Watershed and the drainage areas of the Smith-Fry and Dreisbach Drains were
given in Table 109. Their constitutive forms in kg/ha are shown in Table
110.
TABLE 110
Nutrient Transport (kg/ha)
Smith-Fry
Component
Soluble inorganic P
Soluble organic P
Sediment P
Ammonium N
Nitrate N
Soluble organic N
Sediment N
1975
0.14
0.10
5.2
1.5
19
1.7
31
1976
0.06
0.03
0.98
0.60
5.5
0.31
3.9
Dreisbach
1975
0.34
0.12
4.5
1.8
12
2.3
28
1976
0.18
0.04
0.73
0.85
2.4
0.53
2.8
All of the forms of nitrogen and phosphorus transported greatly de-
creased from 1975 to 1976 due primarily to decreased runoff. The amounts
transported from one drainage area are roughly similar to the other except
for soluble inorganic phosphorus and perhaps nitrate nitrogen.
The larger amounts of soluble inorganic phosphorus from the drainage
area of the Dreisbach Drain were caused by the larger number of houses
discharging domestic sewage into the Dreisbach Drain as compared to the
Smith-Fry Drain. On the other hand, larger amounts of nitrate-N on a area
basis were being discharged from the Smith-Fry Drain in comparison to the
Dreisbach Drain very likely because of more extensive subsurface drainage
and row crop farming in the drainage area of the Smith-Fry Drain.
Percentages of nitrogen and phosphorus forms which were transported
past the outlets of the two drains are given in Tables 111 and 112.
Of interest in Table 111 is the shifting which takes place between the
various nitrogen forms from 1975 to 1976. Sediment nitrogen was high in
1975 because excess rainfall caused more runoff during that year. In 1976,
the relative dry year, ammonium and nitrate-N constituted a greater percen-
tage of the total nitrogen transported than in 1975.
With regard to phosphorus transport, most of that which was transport-
ed was sediment-bound, as shown in Table 112.
Conclusion: With average or above average rainfall amounts, more than
90 percent of the total phosphorus transported is sediment-bound. However,
only about 50 percent of the total nitrogen is sediment-bound. Models of
DISCUSSION
-------
259
TABLE 111
Percent of Nitrogen Forms Transported
Form of N transported
Drain
Smith-Fry
Dreisbach
Year
1975
1976
1975
1976
Amm.-N
2.8
5.8
4.1
12.9
Nit.-N
35.7
53.4
27.2
36.5
Sol Org N
3.2
3.0
5.2
8.1
Sed N
58.3
37.8
63.5
42.5
TABLE 112
Percent of Phosphorus Forms Transported
Form of P transported
Drain
Smith-Fry
Dreisbach
Year
1975
1976
1975
1976
Sol Inorg P
2.6
5.9
6.9
19.0
Sol Org P
1.8
3.2
2.4
4.2
Sed P
95.6
90.9
90.7
76.8
P-transport can be based on models of the erosion-sedimentation process.
We recommend that models of N-transport should not be based on models of
the erosion-sedimentation process only.
5.2.1.3.3 Partitioning of Certain Parameters
The runoff and transported sediment and nutrients were separated into
different categories dependent essentially on the size of the runoff event.
These categories were: base flow, small events, large events, and snow
melt.
Snow melt was a lumped flow event of base flow and small events which
occurred during thaw periods. In many instances, the outlet drains were
then blocked by ice and snow and the recorded water stages were high.
Under these conditions, the snow melt flow reported is our best estimate
based on the depth of snow over the drainage area and the amount of rain-
fall which may have occurred simultaneously with the thaw. Unfortunately,
the actual water stage and its uncorrected runoff were reported in some
earlier project reports. The data below supersedes that reported earlier.
Base flow can easily be identified from the runoff hydrograph because
it is a longtime, steady (depth does not vary with time) flow event. A
large event was arbitrarily established as an event which produced more
than 2.5 cm (1 inch) of runoff from an entire drainage area. Small events
were those flow events which occurred between base flow and large flow
DISCUSSION
-------
260
events.
The results of partitioning the runoff and transported sediments and
nutrients for the Smith-Fry and Dreisbach Drains are given in Tables 113
and 114.
TABLE 113
Partitioning of Runoff, Transported Sediment and Nutrients - Smith-Fry Drain
Component
Runoff
Sediment
Sol Inorg P
Sol Org P
Sed P
Amm.-N
Nit.-N
Sol Org N
Sed N
Base
1975
17
3
7
14
1
15
13
18
1
Flow
1976
23
5
13
12
5
20
13
18
8
Small
1975
35
19
43
38
26
34
40
18
18
Event
1976
14
6
15
14
6
29
21
20
9
s Large E
1975
ii_ j
L.L cuiSpOL uGQ
34
73
37
34
69
40
27
37
77
vents
1976
54
86
41
62
80
36
60
41
80
Snow
1975
14
5
13
14
4
11
20
27
4
Melt
1976
9
3
31
12
9
15
6
21
3
TABLE 114
Partitioning of Runoff, Transported Sediment and Nutrients - Dreisbach Drain
Component
Runoff
Sediment
Sol Inorg P
Sol Org P
Sed P
Amm.-N
Nit.-N
Sol Org N
Sed N
Base
1975
10
1
o
8
2
13
8
9
1
Flow
1976
22
6
22
23
7
20
10
31
4
Small
1975
37
14
49
42
32
35
37
27
23
Event
1976
11
4
5
12
4
29
7
18
3
s Large E
1975
. , «,». I- J
ur cinspOL ueo
38
78
26
33
58
28
34
48
68
vents
1976
56
84
41
52
78
31
75
36
83
Snow
1975
15
7
19
17
8
24
21
16
8
Melt
1976
11
6
32
13
11
20
8
15
10
The large flow events only occurred a few times during either 1975 or
1976. In 1975 only three storms produced over 2.5 cm of runoff from an en-
tire drainage area and in 1976 only two such storms occurred. Yet these
storms accounted for the major sediment and sediment-bound nutrients tran-
sported from the two drainages areas. Less than 6 percent of the sediment
was transported by base flow and over 70 percent by large flow events.
Conclusion: Transport of sediments and sediment-bound nutrients is
strongly associated with large storms which occur only a few times during a
DISCUSSION
-------
261
year.
With small watersheds, the runoff period is relatively short. The
time from the beginning of a storm to the time peak runoff occurs at the
outlets of the two drains is between two and three hours. With only a grab
sample program it would be very likely that all or parts of the major storm
events would have been missed. As a consequence, the sediment and nutrient
yields from a small watershed may be grossly underestimated if based on
grab sample data. In our case if the large events had been completely
missed by a sampling program only about one-third of the actual sediment
yield would have been reported. High concentrations in grab samples, if
sustained, will still indicate a serious pollution problem. However, an
analysis such as we are presenting here would be impossible because of data
gaps.
Conclusion: In order to characterize nutrient and sediment loadings
from small watersheds, runoff from large storm events must be well-
monitored and, with few exceptions, automated sampling is required.
During certain snow melt periods, it was noticed that high loadings of
soluble inorganic phosphorus frequently occurred. This may have been
caused by a release from decayed vegetative matter at this time. Whatever
the exact causes, however, the levels of soluble inorganic phosphorus as
well as the other soluble forms of phosphorus and nitrogen were higher dur-
ing snow melt events than the percentage of runoff during these events
would indicate.
Conclusion: During snow melt, the transport of soluble nutrients may
be disproportionally high when compared with snow melt runoff.
5.2.1.3.4 Sediment and Nutrients in Runoff
The average concentrations of sediment and nutrients given as a total
and also partitioned according to runoff events are given in Tables 115
through 118. Tables 115 and 116 are for the Smith-Fry Drain for 1975 and
1976, respectively, Table 117 and 118 are for the Dreisbach Drain also for
these two years.
TABLE 115
Smith-Fry Drain - Sediment and Nutrient Concentrations (1975)
Component
Sediment
Sol Inorg P
Sol Org P
Sed P
Amm.-N
Nit.-N
Sol Org N
Sed N
Total
730
0.05
0.03
1.8
0.52
b.5
0.58
11
Base
Flow
130
0.02
0.03
0.11
0.46
2.1
0.61
0.40
Small
Events
««_ /I
mg/i
400
0.06
0.04
1.3
0.51
7.4
0.39
5.6
Large
Events
1570
0.05
0.03
3.7
0.61
5.2
0.63
25
Snow
Melt
260
0.04
0.03
0.51
0.41
9.3
1.1
3.1
DISCUSSION
-------
262
TABLE 116
Smith-Fry Drain - Sediment and Nutrient Concentrations (1976)
Component
Sediment
Sol Inorg P
Sol Org P
Sed P
Amm.-N
Nit.-N
Sol Org N
Sed N
Total
520
0.05
0.03
0.79
0.48
4.5
0.25
3.2
Base
Flow
110
0.03
0.01
0.17
0.42
2.5
0.20
1.1
Small
Events
-mg/j.
230
0.06
0.03
0.35
1.0
6.8
0.36
2.1
Large
Events
830
0.04
0.03
1.2
0.32
5.0
0.19
4.7
Snow
Melt
170
0.19
0.04
0.79
0.80
3.0
0.58
1.1
TABLE 117
Dreisbach Drain - Sediment and Nutrient Concentrations (1975)
Component
Sediment
Sol Inorg P
Sol Org P
Sed P
Amm.-N
Nit.-N
Sol Org N
Sed N
Total
1430
0.13
0.05
1.9
0.70
4.5
0.88
11
Base
Flow
140
0.08
0.04
0.38
0.91
3.6
0.79
1.1
Small
Events
rnn/l
mg/j.
540
0.17
0.05
1.6
0.74
4.5
0.64
6.8
Large
Events
2900
0.09
0.04
2.9
0.52
4.0
1.1
20
Snow
Melt
670
0.16
0.05
1.0
1.1
6.3
0.94
5.9
TABLE 118
Dreisbach Drain - Sediment and Nutrient Concentrations (1976)
Component
Sediment
Sol Inorg P
Sol Org P
Sed P
Amm.-N
Nit.-N
Sol Org N
Sed N
Total
380
0.18
0.04
0.72
0.84
2.4
0.52
2.8
Base
Flow
120
0.18
0.04
0.23
0.75
1.1
0.73
0.51
Small
Events
/-i
mg/i
140
0.08
0.04
0.26
2.2
1.5
0.85
0.76
Large
Events
580
0.13
0.04
1.0
0.47
3.2
0.33
4.2
Snow
Melt
210
0.52
0.05
0.72
1.5
1.8
0.71
2.6
DISCUSSION
-------
263
These tables reinforce the previous conclusions with regard to the ef-
fect of large storm events on sediment and nutrient transport. In addition
to amounts, concentrations also increase for large events.
Conclusion: Sediment and sediment-associated nutrient concentrations
increase markedly with large storm events.
In 1975, these concentrations (sediment, sed P, sed N) were approxi-
mately double that of the average yearly concentrations in the total
discharge. Even in 1976, which was a relatively dry year, these concentra-
tions were approximately 50 percent higher than the average yearly concen-
trations. However, the concentrations of the soluble forms of nitrogen and
phosphorus for the large events were in general about equal to the average
yearly concentrations in the total discharge. This was probably due to di-
lution by large runoff volumes. On the other hand, the concentrations of
soluble forms in snow melt were in general higher than the average yearly
concentrations in the total discharge.
It would be difficult just by looking at these average concentrations
to judge whether the streams in the Black Creek Watershed are meeting the
"fishable and swimmable" criteria set forth in PL 92-500, the clean water
bill. During critical low flow periods which normally occur in late summer
and fall, the flow remaining in the streams often collects into shallow,
stagnating pools. If sunlight is available, considerable blue-green algal
growth may occur. What through-flow occurs is sometimes augmented by sep-
tic tank discharge which should actually be classified and treated as point
source pollution. Septic tank discharge may keep the streams in the
watershed from drying up completely, but on the other hand, it provides the
nitrogen and phosphorus necessary for algal growth.
It is difficult to compare our records with those for the Maumee
River. In 1976, the sediment loss from the Black Creek Watershed was about
the same as the longterm average from the Maumee River as measured at Wa-
terville, Ohio. In 1975, however, the sediment loss from the Black Creek
Watershed was over four times the longterm average recorded at the station
in Waterville. The total phosphorus concentration in the Maumee River at
New Haven which is just above the entrance of Black Creek into the Maumee
River as measured by the Water Pollution Control Plant of Fort Wayne aver-
aged about 0.45 mg/1 for 1975-76. Their average measured sediment 'concen-
tration for these two years was around 80 mg/1. Our measurement of total
phosphorus concentrations at the State Route 101 bridge across the Maumee
River and just below the entrance of Black Creek into the river were 0.48
mg/1 and 0.43 mg/1 for 1975 and 1976, respectively. Our average measured
sediment concentrations were 240 mg/1 and 140 mg/1 for these two years.
All of the water samples, those collected by the Water Pollution Control
Plant of Ft. Wayne and by us, were grab samples and the concentrations were
not flow weighted. We collected more samples particularly during storm
events. This could account for much of the difference between the values
for sediment concentrations.
The total phosphorus concentrations agree closely. Both sets of data
also indicate that total phosphorus concentrations in the Maumee River are
not correlated closely to sediment concentrations. Our data for the Maumee
River show that 21 percent of the total phosphorus concentration in 1975
and 42 percent of the total phosphorus concentration in 1976 was soluble
DISCUSSION
-------
264
phosphorus. The total phosphorus concentrations in the Maumee River seem
to be greatly influenced by soluble phosphorus; on the other hand, the to-
tal phosphorus concentrations in the Black Creek Watershed are greatly in-
fluenced by sediment-bound phosphorus.
The average suspended sediment concentration at Waterville, Ohio, near
the entrance of the Maumee River into Lake Erie is approximately 200 mg/1.
This corresponds to the sediment concentrations reported above and indi-
cates that if comparisons with our data can be made they should be with our
base flow and some part of our small flow events and certainly not with ei-
ther our total flow or large flow event categories. If we did this, then
the concentrations of total phosphorus and suspended sediments seem to be
in rough agreement.
Conclusion: Average concentrations of sediments and nutrients
discharged from the Maumee River are in line with measurable concentrations
in the Maumee River.
This conclusion seems to skirt the issue of the source of the nu-
trients in the Maumee River. However, there is a scarcity of data, espe-
cially of data which were gathered and analyzed the same way, for valid
comparisons to be made. Certainly, we do not wish to give the impression
that reductions of nutrient loadings from agricultural watersheds into the
Maumee River are not desirable. On the other hand, the relatively large
concentrations of soluble phosphorus in the Maumee River would indicate
that industries and municipalities still have a job to do also.
5.2.1.3.5 Runoff,Sediment,Nutrient Sources
The probable sources of runoff and transported sediment and nutrients
in the Smith-Fry and Dreisbach Drain for 1975 and 1976 are given in Tables
119 through 122. The waste treatment facility for almost all non-Amish
homes in the watershed is a septic tank which is discharged directly with a
tile drain or a stream. The sewage column reflects the contribution of
outfalls for septic tanks. Calculations were made by assuming the
discharge from each home to be 100 gallons per day. The column for rain
shows those chemical components which were measured in rainfall as a per-
centage of the total amount which was discharged for that particular year.
It does not represent the proportion of the soluble forms which was contri-
buted by rainfall, but it merely indicates that rainfall is also a poten-
tial source for these phosphorus or nitrogen forms. The contribution of
soluble inorganic phosphorus by rainfall would be included in the column
for land, and the contribution of ammonium or nitrate-nitrogen by rainfall
might be included under the tiles, interflow or land columns.
Conclusion: Losses of soluble inorganic phosphorus, ammonium and
nitrate-nitrogen from the watershed are partially due to the input of these
chemical forms by precipitations.
The land surface was the major source of the sediment and nutrients
which were transported in the Smith-Fry and Dreisbach Drains during 1975.
Over 90 percent of the sediment and sediment-bound nutrients originated
from the land surface during that year. This also held true for the
Smith-Fry Drain in 1976. However, in the Dreisbach Drain during 1976,
sewage outflow which constituted only 3 percent of the runoff contained
substantial percentages of the soluble nutrients as well as some sediment-
DISCUSSION
-------
265
TABLE 119
Sources of Runoff, Transported Sediment and Nutrients - Smith-Fry Drain (1975)
Component
Runoff
Sediment
Sol Inorg P
Sol Org P
Sed P
Amm.-N
Nit.-N
Sol Org N
Sed N
Tiles
15
1.3
4
8
0.6
9
25
9
0.2
Interflow
17
5
9
.
10
28
10
Sewage
1 «»*k^V -1- J-l
tr anspor tea
0.3
0.2
27
5
6
5
1.1
0.4
Land
64
97
64
78
93
76
46
81
99
Rain*
(115)
(340)
(35)
Percentage of soluble nutrient as measured in the storm which could
have been supplied by rainfall.
TABLE 120
Sources of Runoff, Transported Sediment and Nutrients - Smith-Fry Drain (1976)
Component
Runoff
Sediment
Sol Inorg P
Sol Org P
Sed P
Amm.-N
Nit.-N
Sol Org N
Sed N
Tiles
8
1.4
2
15
0.4
5
18
11
0.3
Interflow
18
5
16
12
43
13
Sewage
4- if it v^nnxs.
-------
266
from the Black Creek Watershed.
Conclusion: Most of the sediment and sediment-bound nutrients ori-
ginate in runoff from the land surface. The discharge of ammonium is also
largely associated with runoff from the land surface. However, the
discharge from septic tank outlets contributes substantial amounts of solu-
ble inorganic phosphorus into the streams in the watershed.
5.2.1.3.6 Average Yearly Concentrations
TABLE 121
Sources of Runoff, Transported Sediment and Nutrients - Dreisbach Drain (1975)
Component
Runoff
Sediment
Sol Inorg P
Sol Org P
Sed P
Amm.-N
Nit.-N
Sol Org N
Sed N
Tiles
23
1.1
7
15
0.9
7
25
13
0.8
Interflow
8
2.4
5
2.4
9
5
Sewage
J_ r\f^ *- A- r^s3
CL cinspo L ueu
1.2
0.4
37
14
5
13
6
1.1
Land
68
98
54
66
94
78
60
82
98
Rain*
(46)
(280)
(56)
* Percentage of soluble nutrient as measured in the stream which
could have been supplied by rainfall.
TABLE 122
Sources of Runoff, Transported Sediment and Nutrients - Dreisbach Drain (1976)
Component
Runoff
Sediment
Sol Inorg P
Sol Org P
Sed P
Amm.-N
Nit.-N
Sol Org N
Sed N
Tiles
12
2
3
10
1
3
24
11
2
Interflow
g.
t>
13
3
11
4
26
12
Sewage
4- *- ai"^G7"V\ Y" ^fis^
UL cinbpoi. neu
3
7
71
44
35
28
30
16
Land
72
91
23
35
64
65
20
77
82
Rain*
(54)
(380)
* Percentage of soluble nutrient as measured in the stream which
could have been supplied by rainfall.
at
The average yearly sediment and nutrient concentrations in the runoff
the outlets of the Smith-Fry and Dreisbach Drains are given in Tables
DISCUSSION
-------
267
123 and 124, respectively.
TABLE 123
Yearly Concentrations, Transported Sediment & Nutrients-Smith-Fry Drain (75-76)
Component
Sediment
Sol Inorg P
Sol Org P
Sed P
Amm.-N
Nit.-N
Sol Org N
Sed N
Stream
630
0.05
0.03
1.3
0.50
5.5
0.42
7.1
Tiles
60
0.01
0.02
0.04
0.31
11
0.35
0.12
Interflow
«%~*. /T
-nig/i
0.01
0.02
0.31
11
0.35
Sewage
470
4.2
0.57
8.4
8.0
24
15
Land
1060
0.05
0.03
2.1
0.59
5.9
1.0
13
Rain
0.014
0.46
.59
TABLE 124
Yearly Concentrations, Transported Sediment & Nutrients-Dreisbach Drain (75-76)
Component
Sediment
Sol Inorg P
Sol Org P
Sed P
Amm.-N
Nit.-N
Sol Org N
Sed N
Stream
910
0.16
0.05
1.3
0.77
3.5
0.70
6.9
Tiles
70
0.04
0.03
0.07
0.22
4.9
0.48
0.36
Interflow
__-nrt /I
"KJ/-L
0.01
0.02
0.22
4.9
0.48
Sewage
470
4.2
0.57
8.4
8.0
24
15
Land
1570
0.14
0.04
1.7
0.97
3.6
0.75
10
Rain
0.014
0.46
0.59
The concentrations in the column for stream are the average flow-
weighted concentrations from the sources: tiles, interflow, sewerage, and
runoff. The concentrations of soluble nutrient forms in the column -for
rain are given as reference values only.
The ye rly concentrations are averages of the concentrations for a re-
latively wet and a relatively dry year and so might well be good estimates
of the longtime yearly concentrations in the Smith-Fry and Dreisbach
Drains. As such they tend to reinforce the conclusion that the average
concentrations seem to be in line with the measurement of similar concen-
trations in the Maumee River.
5.2.1.3.7 Average Nutrients on Sediments
The average yearly concentrations of total nitrogen and phosphorus in
suspended sediments from the Black Creek Watershed are shown in Table 125.
Also shown are these concentrations with respect to relatively undisturbed
soils in the watershed.
DISCUSSION
-------
268
TABLE 125
Yearly Concentrations of Total N and P Attached to Sediment (1975-76)
Soil and Sediment Total N Total P
g/1
Watershed Soils 1760 680
Sediment:
Stream 8900 1800
Tile drains 3600 950
Septic tanks 24000 19000
Surface runoff 8800 1600
The average yearly concentrations of the nutrients attached to sedi-
ments in Black Creek, in the outflow from tile drainage underlying portions
of the watershed, in the effluent from septic tanks into the watershed
streams, and from surface runoff can be compared to these concentrations
which are attached to relatively undisturbed soils in the watershed.
Conclusion: Sediments in runoff are nutrient enriched. The least en-
richment occurs from subsurface drainage systems and the most from septic
tank outfalls. The sediments in Black Creek have an enrichment factor of
about three as compared to the uneroded soils in the watershed.
5.2.1.3.8 Estimate of Added Nutrients Lost
The estimation of nutrients which were added as commercial fertilizers
or manures lost from the Black Creek Watershed is a tenuous undertaking but
certainly one which should be addressed. Estimates of the application
rates for fertilizers and manures were based on a questionnaire (8).
A nitrogen balance for the entire watershed subdivided between Amish
and non-Amish farms is shown in Tables 126 and 127. The amount of nitrogen
applied as commercial fertilizer and manure and fixed by legumes was re-
garded as the same for both 1975 and 1976. We assumed that mineralization
of nitrogen was balanced by that which was fixed in the soil mass. The in-
put of nitrogen caused by that portion of the precipitation which infil-
trated into the soil was added to the subtotal for both years. The total
applied or fixed nitrogen was then estimated as 329,500 kg for 1975 and
316,500 in 1976.
The applied or fixed nitrogen lost from the watershed was the measured
amount of soluble inorganic nitrogen discharged from the watershed minus
the contribution from septic tank outfalls and that portion of the percipi-
tation which occurred as runoff. The loss of applied or fixed nitrogen
from the Black Creek Watershed was then estimated as 66,000 kg for 1975 and
10,700 kg for 1976.
If we assume that the same proportion of the applied or fixed nitrogen
was lost, then 20 percent of the nitrogen applied as commercial fertilizer
and manure was lost in 1975, a very wet year, and 3 percent was lost in
1976, a relatively dry year. The loss of soluble inorganic phosphorus ori-
DISCUSSION
-------
269
TABLE 126
Estimate of Annual Applied Fixed Nitrogen
Nitrogen Source
Commercial fertilizer or manure:
Corn
Soybeans
Small grain
Hay and pasture
Fixation:
Soybeans
Forage legumes
Total
Amish
12,200
-
5,400
13,800
-
11,000
42,400
Non-Ami sh
IT,.,
Kg
153,200
5,700
23,400
3,000
57,900
1,200
244,400
Total
165,400
5,700
28,800
16,800
57,900
12,200
286,800
TABLE 127
Estimate of Percent Applied and Fixed Nitrogen Lost (1975-76)
Nitrogen Input or Loss
Applied and fixed nitrogen inputs
Input due to infiltrated precipitation
Total nitrogen input
Total nitrogen loss (measured)
Loss due to precipitation runoff
Loss due to septic tank discharge
Applied and fixed nitrogen loss
Percent applied and fixed nitrogen lost
1975 1976
kg
286,800
42,700
329,500
87,400
14,100
7,300
66,000
20%
286,800
29,700
316,500
22,700
5,800
6,200
10,700
3%
ginating as commercial fertilizer or manure was estimated to average about
0.3 percent per year for 1975 and 1976.
The estimated loss of soluble inorganic nitrogen from commercial fer-
tilizers or manures is in line with estimated losses from other areas with
similar application rates. The loss of added phosphorus in a soluble form
is relatively low and suggests that most of it was immediately tied up in
the soil complex. However, more phosphorus could be expected to be lost
through the erosion process. We did not attempt to evaluate this because
of our imprecise knowledge of where erosion was occurring in relation to
the areas which received phosphorus fertilization.
The total nitrogen loss per hectare from the Smith-Fry Drain was only
about 20 percent higher than from the Dreisbach Drain (see Table 109).
However, the percentage of the drainage area of the Smith-Fry Drain in row
crops was about double that for the Dreisbach Drain. It is probably rea-
sonable to expect that most of this difference was due to the increased
usage of commercial fertilizers on the lands contributing runoff to the
DISCUSSION
-------
270
Smith-Fry Drain. Improved fertilizer management techniques may still de-
crease the amounts of soluble inorganic nitrogen reaching the streams in
the Black Creek Watershed, but the level to which it can be reduced is cer-
tainly bounded and the effect of improved techniques may not be all that
noticeable in loadings from the watershed into the Maumee River.
5.2.1.3.9 Another Factor The Farming Community
The Black Creek Watershed is somewhat atypical of similar sized areas
in the Maumee Basin in relation to its comparatively larger settlement of
Amish farmers. Most of them have farms on the glacial till soils which oc-
cur in the northern part of the watershed. The land farmed by the Amish in
the drainage areas of the Smith-Fry and Dreisbach Drains is roughly in pro-
portion to the percentages of glacial till soils in these two areas (see
Table 108). Some aspects of their farming operations which would have a
beneficial effect on the level of non-point source pollution from agricul-
tural lands were (1) a rotational type agriculture which included pasture
and small grains and (2) limited usage of commercial fertilizers. Some
negative factors were (1) fall plowing of erosive soils, (2) overgrazing of
pastures, (3) low crop yields which would increase the erosion potential
from their farms, and (4) use of streams for watering livestock.
The lands which are farmed by the Amish are in general mostly rolling.
Practices for reducing nonpoint source pollution are particularly effective
on these lands. As reported elsewhere in this volume, good cooperation was
generated with the Amish community and many beneficial practices including
parallel tile outlet terraces have been installed on their farms.
During the period of record reported on here, however, and in particu-
lar for 1975, very few of these practices had been installed. Based on our
records and also on observation of lands on which severe erosion was occur-
ring, we have concluded that the 19th century farming techniques as used by
the Amish did not seem to improve the water quality in streams from that
area of the Black Creek watershed in which they reside. On the other hand,
we expected the greatest reduction of nonpoint source pollution from that
area in the future. We have in the Black Creek Watershed a unique oppor-
tunity to study "old and modern" agriculture and their relative effects on
the quality of our water resources. Further analyses will be made based on
our water sampling program and also on our modeling effort as reported in
Section 0.0.
5.2.1.3.10 Effect of Agricultural NFS Pollution
The actual impact of sediment, phosphorus and nitrogen discharged from
the Black Creek Watershed and other agricultural watersheds in the Maumee
Basin on the eutrophication process in the Maumee River and Lake Erie is
speculative and needs further study. Phosphorus is largely sediment-bound
and would have to be released in order to enter into the nutrient cycle of
the algal biomass. In a study reported elsewhere in this report only about
15 percent of the phosphorus which was bound to sediments could become
available for algae growth. This was for a laboratory situation where
light was not limiting as it would in fact be in a lake environment. Re-
garding nitrogen, a large percentage of the total nitrogen discharged oc-
curs in the late winter and spring months during high flows. Just how much
would remain and be available later in the year for usage by the algal
DISCUSSION
-------
271
biomass is also largely unknown at this time.
Nonpoint source pollution from the Black Creek Watershed occurs pri-
marily from large storm events. These large storm events usually occur
during the spring and early summer months. Also the chances are that the
Maumee River is at a moderately high stage during the period when the large
storm event occurs. On the other hand, the input of point source pollution
is most critical in the Maumee River during its low flow periods. While
there may be some residual chemicals remaining from high flow periods, the
effect of agricultural nonpoint source pollution by-and-large should not be
evaluated using low flow criteria as is characteristic for point source
pollution.
Conclusion: The effects of agricultural nonpoint source pollution and
point source pollution on our water resources are sufficiently different
that direct comparisons between them can not be made and separate objec-
tives for their evaluation and control are in order.
5.2.1.4 SUMMARY
An analysis of rainfall, runoff, and transported nutrients from the
Black Creek Watershed for 1975 and 1976 was made. Rainfall was about 20
percent over normal in 1975 and, to the other extreme, about 20 percent
less than normal in 1976. We were fortunate from a data collection stand-
point to have experienced these extreme rainfall conditions since there is
no guarantee that a much longer period of record would have produced these
rather extreme weather conditions. Future years of record will be used to
verify our conclusions.
References
1. Geological Survey. 1971. Water Resources Data for Ohio. Part 1.
Surface Water Records. U.S. Department of the Interior, Washington, D.C.
223p.
2. Monke, E.J., D.B. Beasley, and A.B. Bottcher. 1975. Sediment Contri-
butions to the Maumee River. EPA-905/9-75-007, Proc. Non-Point Source Pol-
lution Seminar, November 20, 1975 in Chicago, IL. pp.71-85.
3. Bremner, J.M. and D.R. Keeney. 1965. Steam distillation methods for
determination of ammonium, nitrate, and nitrite. Anal. Chem. Acta 32:485-
495.
4. Nelson, D.W. and L.E. Sommers. 1975. Determination of total nitrogen
in natural waters. ,J. Environ. Quality 4:465-468.
5. Murphy, J. and J.P. Riley. 1962. A modified single solution method
for determination of phosphate in natural waters. Anal. Chem. Acta
27:254-267.
6. Sommers, L.E. and D.W. Nelson. 1972. Determination of total phos-
phorus in soils: A rapid perchloric acid digestion procedure. Soil Sci.
Soc. Amer. Proc. 36:902-904.
DISCUSSION
-------
272
7. Schwab, G.O., E.G. McLean, A.C. Waldron, R.K. White, and S.W. Michner.
1973. Quality of drainage water from a heavy textured soil. Trans. ASAE
16(6): 1104-1107.
8. Brooks, R.M. and D.L. Taylor. 1975. Leadership important to accep-
tance of conservation. EPA-905/9-76-006, Environmental Impact of Land Use
on Water Quality, pp. 155-196.
5.3 LOCAL VARIATIONS IN WATER QUALITY
Water quality in the Black Creek watershed varies strikingly over
space and time. Spatial variation results from differing typographies,
soil types, land use, and channel characteristics. Temporal variation in-
cludes changes due to the seasons with added variation due to changing flow
regimes. The following discussion illustrates several patterns of varia-
tion.
5.3.1 Water Quality During a Drought
The Black Creek watershed experienced a prolonged dry period in summer
1976 and, thus, provided an excellent opportunity to determine water quali-
ty characteristics at base flow. During this period rainfall events were
generally less than one inch and were quickly absorbed by dry soil result-
ing in very little or no surface runoff. Most suspended solids and nu-
trients carried in Black Creek were from within streams, drain tiles, and
domestic sewage sources. Furthermore, by late summer most of the streams
in the watershed were reduced to isolated pools with no surface flow.
Because of the low flow conditions on the Black Creek watershed during
July, August, and September 1976 it was possible to divide the watershed
into several distinct units (Figure 37). The lower reaches of Black Creek
had a base flow maintained by groundwater all summer. Estimated flow rates
during these base flow conditions ranged from 0.0005 to 0.0040 cubic
meters/sec. Most of the water for this flow originates at two distinct lo-
cations; Gorrell Drain at Notestine Road and about 150 m downstream from
station 131 (east of Bull Rapids road on Black Creek). Water samples of
this outflow reflect lower phosphorus, ammonia, and turbidity levels than
found at downstream stations (Table 128). Data for Table 128 were collect-
ed on August 24 and 25 after a period of at least one week without flows
from any tributary drain or from the upper reaches of Black Creek. Since
there was no surface runoff or upstream channel flow, the changes in water
chemistry indicated by Table 128 result from the accumulation of materials
from the stream channel and riparian environment of Black Creek. A hog wa-
tering facility a short distance above station 132 may have been responsi-
ble for some of the changes in concentrations of some nutrients. Unfor-
tunately, we did not have any sample stations between the groundwater
source and the hog lot.
Two sources of pollutants in the lower section of Black Creek were
tile lines near stations 140 and 154 which carried domestic sewage ef-
fluent. No significant shifts in water chemistry could be detected at
these locations, suggesting that these sources were of little consequence
in affecting water chemistry at base flow conditions.
Immediately following the few rainfall events of summer 1976 Dreisbach
and Richelderfer Drains experienced increased runoff rates while other tri-
DISCUSSION
-------
273
TABLE 128
Characteristics, Groundwater, Channel Flows, Black Creek
Parameter (c)
Alkalinity
Conductivity
Dissolved Solids
Hardness
Turbidity
Total Phosphorus
Soluble Orthophosphate
Nitrate
Nitrite
Ammonia
Organic Nitrogen
Total Residue
(Suspended Solids)
Sulfate
Groundwater
267
702
512
357
27
0.047
0.003
0.01
0.01
0.01
0.29
532
88.1
Stagnant Stream Water (a)
268.7 +- 6.35
783.7 +- 35.01
571.3 +- 25.00
358.3 +- 3.51
58.0 +- 4.00(b)
0.64 +- 0.02(b)
0.04 +- 0.008(b)
0.01
0.01
0.087 +- 0.005(b)
0.313 +- 0.095
578.7 22.7
88.8 1.15
(a) Stations 132-134, Mean +- standard deviation
(b) Groundwater and channel flow concentrations significantly
different at p<0.05
(c) mg/1 except turbidity in Jackson Turbidity Units
butaries were little affected by the rains. These intermittent flows iden-
tify a second major area of the watershed: areas with very low flow or in-
termittent flow through most of the summer but with an occasional spate
originating in Harlan (Figure 37). We do not have any regular sampling
stations on Richelderfer Drain near Harlan but a series of stations on
Dreisbach Drain yields interesting results. Virtually all water chemistry
parameters increase sharply as the stream passes Harlan (Stations 115 and
116). As the stream continues south beyond Harlan all parameters decline
DISCUSSION
-------
274
Base Flow
"Harlen" Flow
"No Flow"
Summer, 1976
Stagnant Pools
Organic Pollution
Clean Water
Figure 37 Flow Regimes in Watershed
DISCUSSION
-------
275
and reach levels similar to those above Harlan. This generally happens at
or slightly below station 118 about 1 km south of Harlan. This seems to be
a result of settling out of organic matter between stations 115 and 118.
Apparently, this particulate matter cannot be metabolized by the stream
ecosystem during low flow periods. It accumulates in anaerobic sediments
and is later washed downstream during rainfall events (see discussion of
microbiological sampling). A population of fish is able to persist at sta-
tion 118 but very few persist above that sample station.
The dynamics of nutrients in this section of stream are less clear.
Perhaps the decline in nutrients is due to incorporation into algal
biomass, which settles to the bottom. It is then flushed out with other
organic matter during runoff events.
Observations of Richelderfer Drain in late September and October indi-
cate that channel flow from Harlan is considerably more turbid than that in
the main Black Creek channel.
The third major situation in the watershed existed upstream of Harlan
on the Dreisbach and throughout Wertz and Smith-Fry Drains. Generally,
flow conditions in these areas were very low or stagnant. Upper Dreisbach
(above station 113) remained dry through much of the summer (Table 129).
Some flow occurred in Smith-Fry Drain in late June and July and flow rates
in Wertz Drain were lower and more intermittent. Most areas along the tri-
butaries were reduced to standing pools many of which dried up completely
by the end of the summer. This resulted in progressive concentration of
fishes and ultimately death for many as habitat deterioration made them
especially susceptible to raccoon and other predators. The most severe
condition occurred in early September when the isolated pools could be
classed into two groups. Many pools were maintained by septic tile flows
which commonly contained significant amounts of organic effluents (Figure
37). Pools in areas not receiving such effluent maintained relatively
better water quality.
TABLE 129
Stations with Water - Dreisbach Drain
Stream Segment
Upper Segment
(Springfield
Center to
Antwerp Roads
Lower Segment
Total Number
of Stations
11
9
No
June
10
2
0
. of Stations Without Water
June July July Aug
29 13 26 9
4899
0010
On
Aug
24
10
4
5.3.2 Effects of Channel Morphology
The general program of water sampling in the Black Creek Study in-
volves routine and often automated sampling of water at a number of loca-
tions around the watershed. This sampling protocol has been supplemented
DISCUSSION
-------
276
with detailed surveys of water quality throughout the watershed. Initial-
ly, we sampled the segment of Wertz Drain between Knouse and Antwerp Roads.
Between those two roads there is about 1800 m of stream channel, of which
about 550 m meanders through a small patch of mixed forest. Tree dominants
in the forest include oaks (Quercus), hickories (Carya), maples (Acer), and
beech (Fagus). Within the 1800 m section of stream, 12 sample stations
(later expanded to 16) were established. Three upstream stations (#10-12)
are in a region of grass waterway bordered by agricultural land. Stations
6-9 are within the main woodlot area and 4-5 are in a small extension of
forest bordering the stream. The remaining stations (1-3) are in a region
of grass waterway bordered by agricultural land. Station 2 is on a bend in
the stream in an area with several large cottonwoods (Populus). A steep
badly eroded bank is located just above station 2. Inside the forest no
grass stabilizes the bank, and the stream forms a complex of pools and rif-
fles meandering widely through the forest. Just below station 7 a badly
eroded tile drain enters the main channel.
Since February 1975 we have collected water samples regularly from the
Wertz Drain. Briefly our results demonstrate that sediment loads decline
as the water flows through the area. Furthermore, very soon after the
stream leaves the forest, sediment load increases and stabilizes near a
level characteristic of the Wertz Drain above the forest.
Variations within the agricultural and forested section of the stream
are also of interest. For example, a typical increase at station 5 is ap-
parently due to the nearby entry of a large drainage tile from adjacent
fields. Considerable erosion is evident at that point. Note that even the
presence of a small finger of forest extending south from station 5 results
in a decline of sediment load to station 3. As the stream continues south
to station 2 sediment loads increase significantly until they approach lev-
els above those in the forest.
A series of t-tests to determine the significance of variation in sed-
iment loads shows that stations above the forest are not significantly dif-
ferent in suspended solid content from the lower two stations. Stations
10-12, and 1-2 are significantly higher than stations at the lower end of
the forest (stations 4 and 6) indicating that the forest acts as an effec-
tive sediment trap in reducing suspended solids by about 28%.
At very high stream flows, the forest apparently has little impact on
reducing sediment loads. We obtained water samples shortly after a very
heavy rain (about 10 cm in 2 hours) in 20 May 1975. From station 12 to 1
there was a gradual increase in suspended solids through the Wertz Drain
Study Area. Furthermore, the increased load of sediment seems to be a gen-
eral phenomenon as the load increased from the headwaters of the streams to
near the junction of Black Creek and the Maumee River. These results sug-
gest that the forest acts as a very efficient trap for removing suspended
. solids during most flow rates (>95%, Beasley, pers. comm.) but at very high
flow rates the forest has no value as a sediment trap.
Initially, we were not able to determine the factors responsible for
declining sediment loads. However, we are now convinced that sediment
reductions are due to the physical processes obtaining in the forested sec-
tion of the watershed. From unit stream power theory sediment reductions
on the order of 25% are expected in a meandering pool-and-riffle channel
like that in Wertz Woods and observed sediment declines were 28%. The
DISCUSSION
-------
277
roughness coefficient (n) is the likely factor responsible for the de-
creased sediment loads since the slopes (S) are lower (.25 ) above and
below the woodlot than in the woodlot (.40).
5.3.3 Variations in Water Quality
From an initial sampling base of 12 stations on a small section of the
Wertz Drain we have expanded to an intensive series of 120 sample stations
throughout the Black Creek watershed (Figure 0). These stations were
selected to allow intensive sampling on four major areas of the watershed.
1. Dreisbach Drain. Since the initiation of the Black Creek Sediment
Control Project, Dreisbach Drain has been the subject of intense efforts to
improve agricultural and conservation practices. A series of 20 channel
stations between Springfield Center and Brush College Roads has been esta-
blished to monitor the impact of changes in agricultural and conservation
practices on water quality.
2. Wertz Drain. Our original sampling included 12 stations between
Knouse and Antwerp Roads. The number of stations in that reach has been
expanded to 16 and an additional 18 stations have been located in the
remainder of the Wertz Drain between Boger Road and Black Creek. This ex-
pansion is designed to clarify the changes in stream quality resulting from
several planned conservation activities initiated in the summer of 1976.
3. Smith-Fry Drain. Little or no conservation planning work had been
initiated on the Smith-Fry Drain by summer 1976. As a result, water quali-
ty in this channel may reflect that of other Black Creek tributaries before
initiation of the Black Creek Sediment Control Project. A total of 23
channel stations is located along the Smith-Fry Drain.
4. Black Creek. Thirty-two stations have been located along the
Black Creek channel between Brush College Road and the Maumee River. The
deeper waters and increasing flow volumes of the Black Creek channel will
help to clarify the dynamics of sediment and nutrient movements in larger
streams.
In addition to these stations a number of other sites have been
selected to measure such areas as tile outflows from fields with and
without parallel tile outlet terrace systems. The sample sites for this
expanded effort have been selected to sample areas of different stream mor-
phology, land use, vegetation cover, and other factors.
As demonstrated by earlier studies of the Wertz Woods area, local
variation in streamside vegetation, channel morphology, and land use affect
water quality characteristics. Water quality varies strikingly within and
between waterways within the Black Creek watershed (Table 130). At present
all of the cause and effect relationships which account for this variation
are not known.
However, some small scale variation can be accounted for. At station
189 at the lower end of Wertz Drain suspended solids loads increased by
about 200/mg/l as a result of domestic ducks feeding in the channel at that
location.
In the case of phosphorus spikes are especially high at Harlan where
DISCUSSION
-------
278
TABLE 130
Water Quality Characteristics - Major Segments
Parameter
Phosphorus
Total Phosphorus
Soluble
Or thophosphate
Nitrogen
Nitrate
Nitrite
Ammonia
Organic Nitrogen
Other Parameters
Turbidity
Total Residue
(Susp. Solids)
Alkalinity
Hardness
Conductivity
Dissolved Solids
Sulfate
Dreisbach
High
High
Very High
Very High
Very High
Very High
I
High
High
High
High
High
High
Stream Segment Index
Wertz Smith-Fry Black Creek Value*
Low
Very Low
Very High
I
Low
I
Low
Low
High
I
Low
Low
Low
Very Low
Very Low
High
High
Low
Low
Low
Low
I
High
Low
Low
Low
I
I
I
I
I
I
I
I
I
I
I
I
I
(D)**
(D)
(D)
(D)
(D)
(D)
(D)
0.70
0.15
0.01
0.03
0.40
0.50
50
600
140
220
700
500
30
** I=Index Value
(D) = Decreasing trend through this segment of the watershed.
* Concentration in water in mg/1 except for turbidity which is measured
in Jackson Turbidity Units.
concentrations are up to 10 and 100 times higher than the index value for
total phosphorus and soluble orthophosphate, respectively.
We are conducting analysis and continuing monitoring efforts to demon-
strate how specific land use and channel characteristics affect water qual-
ity parameters. However, one point is clear. Variation in sediment and
nutrient loads over short stream distances may be striking for a large com-
plex of reasons. Clearly, when sediment and nutrient loads may vary by
orders of magnitude over relatively short distances, it is important to use
caution in the selection of sampling localities for monitoring of water
quality. This is obviously true where point-source inputs are present.
However, evidence that channel structure and other factors have effects on
transport and deposition suggests that care should be taken in selection of
monitoring sites even where non-point sources predominate. Ideally, a set
of recommendations should be developed to aid researchers in selection of
monitoring sites. At the present time it is only possible to suggest that
caution be exercised. Concentration and loading estimates will be affected
DISCUSSION
-------
279
by carelessness in site selection.
5.3.4 Effects of Construction Activities
Short term effects of construction may cause significant changes in
water quality. Activities of heavy equipment in pulling channel banks and
bottom-dipping commonly created visible changes in water quality. This is
especially true of suspended solids which increase markedly downstream from
construction areas.
5.3.5 Effects of Conservation Practices
As discussed in Section 0.0.6, channel modification activities on the
Wertz Bank upstream from Wertz Woods on Wertz Drain has had a significant
affect on downstream channel characteristics. Erosion of unstable bank and
channel bottom after construction in Wertz Branch yielded large amounts of
sediment. As this sediment moved downstream it was deposited in downstream
areas (especially Wertz Woods). This deposition has all but obliterated
many of the pool and riffle areas. Implementation of this conservation
practice has over the short term increased sediment loads and affected down
stream biotic communities by destroying several habitat types (pools and
riffles) and by creating very unstable substrates (shifting sand and silt
bottoms) in a number of downstream reaches.
Another example of a conservation practice affecting water quality
comes from the desilting basin constructed on the main Black Creek Channel
below Station 15. While large amounts of sand have been deposited in this
basin since its construction our data show that during low flows total
residue, total phosphorus and turbidity are significantly higher at the
outlet than at the entry of the basin (Table 131). Total residue is in-
creased by 8% while total phosphorus and turbidity increase by 58% and 41%
respectively. The ratio of total residue to turbidity declines from sta-
tion 145 to 146. This ratio can be viewed as coefficient of fineness of
suspended materials where a decreasing value indicates increased proportion
of fine particles in suspension. The reason for this cannot be definitely
established but may be due to such factors as increased turbidity due to
wind fetch or feeding activities of fish although feeding activities of
fish seems unlikely as ammonia levels do not increase in the desilting
basin. Another possibility is that algal populations have increased.
Without knowledge of carbon fractions, it is difficult to distinguish that
possibility from the more likely effect of wind fetch. The desilting basin
seems to be functioning as a turbid pond.
The accumulation of sand size particles in the desilting basin suggest
that bed load is reduced by deposition in the basin but evidence described
above shows that fine particle material increases. The rapid rate of fil-
ling the basin with sands shows that regular maintenance activities will be
required to maintain an efficient operation. Such costly, repetitive
maintenance combined with the high sediment movement associated with con-
struction and increased fine particle concentrations associated with the
basin suggest that this may not be a sound practice for improvements in wa-
ter quality.
DISCUSSION
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-905/9-77-007-B
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Environmental Impact of Land Use on Water Quality-
Final Report on the Black Creek Project (Volume 2-
Tehnical Report)
5. REPORT DATE
1 Q77
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
James B. Morrison- Technical Writer and Editor
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Allen County Soil and Water Conservation District
Executive Park, Suite 103
2010 Inwood Drive
Fort Wayne, Indiana 46805
10. PROGRAM ELEMENT NO.
2RA645
11. CONTRACT/GRANT NO.
EPA Grant G005103
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agecy
Great Lakes National Program Office
230 South Dearborn Street
Chicago. Illinois 60604
13. TYPE OF REPORT AND PERIOD COVERED
Final Report-1972-1977
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
Carl D. Wilson- EPA Project Officer
Ralph G. Christensen- Section 108 (a) Program Coordinator
16. ABSTRACT
This is the Final Technical Report of the Black Creek sediment control project. This
project is to determine the environmental impact of land use on water quality and has
completed its four and one half years of watershed activity. The project, which is
directed, by the Mien County Soil and Water Conservation District, is an attempt
to determine the role that agricultural pollutants play in the degradation of water
quality in the Maumee River Basin and ultimately in Lake Erie.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Sediment
Erosion
Land Use
Water Quality
Nutrients
Socio-Economic
Land Treatment
18. DISTRIBUTION STATEMENT
19. SECURITY CLASS (ThisReport)
21. NO. OF PAGES
Document is available to the public
through the National Technical Infor-
mation Service, Springfield, VA 22161
20. SECURITY CLASS (Thispage)
22. PRICE
EPA Form 2220-1 (Rev. 4-77)
PREVIOUS EDITION IS OBSOLETE
* U.S. GOVERNMENT PRINTING OFFICE: 1978753-066
------- |