EPA810/1-85-999
Rural  Clean  Water  Program
    STATUS REPORT  ON THE CM&E PROJECTS
                      1985
        NATIONAL WATER QUALITY  EVALUATION PROJECT
        North Carolina Agricultural Extension Service
      Biological and Agricultural Engineering Department
             North Carolina State University
                Raleigh, North Carolina
                 In Cooperation With:

             U.S. DEPARTMENT OF AGRICULTURE
           U.S, ENVIRONMENTAL PROTECTION AGENCY

-------
       RURAL  CLEAN  WATER   PROGRAM


    STATUS  REPORT ON THE CH&E PROJECTS


                     1985


                      BY
        North Carolina State University

    National  Water Quality Evaluation Project

                   Personnel

  USOA  Cooperative Agreement - 12-05-300-472
  EPA   Interagency Agreement - AD-12-f-0-037-0

  Steven A. Dressing           Richard P. Maas
  Catherine A. Jamieson        Jean Spooner
       Michael D. Smolen - Principal Investigator
       Frank  J. Humenik  - Project Director

  Biological  & Agricultural Engineering Dept.
       North  Carolina State University
       Raleigh, North Carolina  27650


                      AND


       Economic Research Service (USDA)

                 Personnel

     C. Edwin Young - Project Leader
     Richard Magleby - Section Leader
  EPA PROJECT OFFICER      USDA PROJECT OFFICER
     James E. Meek            Fred N. Swader
 Implementation Branch       Extension Service
Water Planning Division      Natural Resources
    Washington, D.C.         Washington, D.C.

-------
                             PBBFACB
     This  document reports on the status of the five Rural Clean
Water Program (RCWP) projects which are conducting  comprehensive
monitoring and evaluation  (CN&E).  The emphasis is on describing
the  present  state of knowledge of the effects  of  agricultural
Best Management Practices (BMPs) and agricultural nonpoint source
(Ag NFS) control programs on water resources.   The status of the
CM&E  projects is evaluated in the context of how they contribute
to  answering the important questions relating to the control  of
Ag  NPS.   Using the preliminary results from the  CM&Es we  have
attempted   to synthesize a "bottom line" answer to twenty  ques-
tions  which  represent  the knowledge needed to control  Ag  NPS
through  a land treatment program.

     /he report consists of a cross-project summary section which
addresses major NPS control questions in the context of the  CM&E
projects,  economic analysis of the  CM&E projects,  and detailed
analyses  of  each   CM&E project.   We have used  both  our  own
analyses of the water quality data and those presented in project
reports.   We  did an extensive amount of additional analysis  on
the  Idaho project as well as moderate additional analysis on the
Illinois,  Pennsylvania,  and Vermont projects.   In all cases we
have attempted to distinguish data analysis done by  CM&E project
personnel  from  our own analyses.  Materials including the  1984
annual reports and personal communications with project personnel
were  used.   In  all  cases  we  have  discussed  the   analyses
extensively with  CM&E personnel.

     The  Economic  Research Service (ERS) has  made  significant
contributions  to this document in analysis of on-site  and  off-
site benefits/costs of the projects.   The ERS component provides
a  more  complete picture of the  CM&E program than would be  ob-
tained if water quality effects were considered alone.   We  have
integrated the BRS analysis of farm level costs into each project
chapter.   The  second chapter of this report was contributed  in
total  by ERS.   Other contributions for ERS are incorporated  in
the first chapter with explicit citation.

     The  status  and contributions of the CM&E projects will  be
reviewed again in 1987 and this report revised to reflect the new
information.   Thus,  the observations and conclusions  presented
here are interim.
                              m

-------
                     SIMMARY AND CONCLUSIONS

     The  experimental  Rural  Clean  Water  Program  (RCWP)   was
initiated in 1980 by P.L.   96-108.   Twenty-one RCWP projects  were
designated   nationwide,    with   five   projects,    located    in
Pennsylvania,  Idaho,  Illinois, South Dakota,  and Vermont,  desig-
nated  to conduct comprehensive monitoring and  evaluation  (CM&E)
of their water quality and economic impacts.

     The experimental  aspect of RCWP is intended to:

1.   determine  whether  the existing institutional structure  is
     adequate for obtaining the needed treatment of  agricultural
     nonpoint sources;

2.   determine  the. water  quality effects  of  Best  Management
     Practices (BMPs)  at the watershed and water resource level;  •

3.   determine  whether water resources impaired by  agricultural
     nonpoint  sources  can be improved by a concerted  voluntary
     land treatment program; and

4.   determine the economic impacts of treating nonpoint sources.

     Tentative  findings  from  four years  of  the  program   and
prominent  strengths  and  weaknesses of the   CM&E projects   are
summarized  below.   These  findings  and  observations  will  be
reviewed   and  revised  in  1987  as  more  information  becomes
available from the  CM&E studies.
Major Findings from  CM&E

1.   Field  studies  in the PA project indicate that  groundwater
     levels  respond  rapidly  in the permeable  soils  of  their
     critical area,  and therefore, rapid response of groundwater
     quality  to  manure and fertilizer  nutrient  management  is
     likely.

2.   Water  quality  monitoring in the ID project has shown  that
     significant   improvements  in  sediment  concentration   of
     irrigation   canals   have   been   achieved   through   BMP
     implementation.   Control of water quality in irrigated arid
     land appears to be much more rapid than in humid regions.

3.   The VT project has shown that by  eliminating  winter manure
     spreading  in northern climates,  significant reductions  of
     phosphorus and nitrogen loss to surface runoff are possible.
     However,  the experiment also showed that eliminating winter
     manure   spreading  increased  the  amount  of  runoff   and
     suspended solids from the field site.

4.   Analysis  of monitoring results from the ID project  suggest
     that  a 35-45 percent change in a water quality parameter is
     required for that change to be significant at the 95 percent
                               IV

-------
     confidence  level for an irrigated system.   The  amount  of
     change   needed  can  potentially  be  reduced  by  improved
     experimental  design.   The  amount of  change  required  in
     precipitation  driven systems will depend upon the extent to
     which meteorological variables can be accounted for.   NWQEP
     is  presently attempting to determine the amount  of  change
     needed  for statistical significance in precipitation-driven
     systems  with  and  without  correction  for  meteorological
     variability.

5.   Projected economic benefits compared with costs in three  of
     the five projects,  ID,  IL,   and PA,  would not justify the
     projects  except  as  experimental  efforts.    With   major
     redirection  in  BMP implementation,  the ID  project  could
     become economically justified.

6.   Fertilizer  management  (BMP 15) has been found  to  be  the
     least  cost  and  potentially  most  effective  approach  to
     address nutrient-related water resource impairments.

7.   In those projects where recommended BMPs are consistent with
     the farmers preference,  BMP adoption is high.  Examples are
     water  management (BMP 13) and sediment control (BMP 12)  in
     ID,  conservation  tillage (BMP 9) in SD,  and animal  waste
     management  (BMP  2)  in VT.   Those BMPs  not  preferred  by
     farmers  have relatively low rates of adoption.   An example
     is fertilizer management (BMP 15) in ID, SD, VT, and PA.

8.   NWQEP  has found that water quality  monitoring  information
     can  be  used  in  ID and II to refine  their  selection  of
     critical areas.       •
                 *                                            ,
9.   Economic  onsite  benefits projected over 50 years  for  the
     CM&E projects are as follows:

     ID:  Long term soil productivity enhancement ($814,000).
     IL:  Negligible  - Productivity  benefits  because  of  deep
          soils.
     PA:  Fertilizer   savings   and   productivity   enhancement
          (projections in process).
     SD:  Increase net farm income (possibly over $500,000)  from
          enhanced  yields  and  reduced  costs  of  conservation
          tillage.
     VT:  Fertilizer savings of up to $2 million.

10   Economic  offsite  benefits projected over 50 years for  the
     CM&E projects are as follows:

     ID:  Recreation  ($617,000  mostly fishing),  reduced  ditch
          cleaning ($185,000).
     IL:  Water   treatment  savings   ($225,000),   recreational
          fishing ($24,000).
     PA:  Negligible because of low farmer participation.
     SD:  Recreation   and  property  value  benefits  could   be
          relatively high (projections in process).

-------
     VT:   Recreation ($3.9 million),  property values ($1 Million).

Major Strengths of the RCWP-CM&E Program

1.   The    CM&Es   are  directed  specifically  to   track   BMP
     implementation  and  to  Monitor the  water  quality  impact
     before, during and after implementation.

2.   The  scope of the  CM&E projects ties together  field-scale,
     watershed-scale,   and  water  resource  BMP/water   quality
     information.

3.   New  information will be developed on the actual "in-stream"
     effectiveness  of agricultural Best Management Practices  at
     the watershed level.

4.   The   CM&E  Program will document the natural water  quality
     variability  from monitoring over a  10-year  period.   This
     result  will  be more precise than any previous  large-scale
     agricultural  water quality study.

5.   The  information on inherent water quality variability  from
     CM&E  projects can be used in the future to define how  much
     of a change in water quality is significant,  i.e.,  to show
     that  an  observed  change in water quality is  not  just  a
     random occurrence unrelated to the land treatment program.

6.   The  CM&E Program is providing  extensive information on the
     economics of NPS control at-the farm and public levels.
                .» ,
7.   Strong  provision was made from the very early stages of the
     program for independent cross-project analysis and synthesis
     of results.

8.   The two  CM&E groundwater projects,  SD and PA,  represent a
     major effort to tie land treatment to groundwater quality at
     the field level.

9.   The   CM&E projects are providing important new  information
     on appropriate criteria and procedures for selecting  criti-
     cal areas for control of water quality.

10.  Two   CM&E  projects  have  become  the  first  NPS  control
     projects  to  collect sufficient data to be able  to  select
     farm-level  critical areas on the basis of actual  in-stream
     water quality information.

11.  Economic  benefits will likely exceed costs for the  Vermont
     and South Dakota projects,  and, with major BMP redirection,
     could  possibly do so for the Idaho project.   However,  the
     five  projects were selected for their diverse problems  and
     to  determine  water quality impacts of  improved  practices
     rather than expectations of net economic benefits.

-------
12.  Although serious water quality problems existed in all  five
     projects,  potential offsite econonic benefits from improved
     water quality appear substantially higher in the Vermont and
     South  Dakota  projects  than  in the  other  three  because
     recreation   and  property  values  will  be  preserved   or
     enhanced,  and more people will be affected.   Consideration
     of potential economic benefits during project selection will
     contribute  to  maximizing benefits of  future  clean  water
     programs.

Major Weaknesses of the RCWP  CM&E Program
                                                i
1.   Only  two of the five  CM&E projects have yet  achieved  the
     level of farmer participation envisioned for the program.   A
     reasonable  level  of  participation  is  essential  to  the
     success of the program.

2.   The  water resource use impairments in several of the   CM&E
     projects  are  not  defined as precisely as in some  of  the
     other  RCWP  projects.  This limits the  potential  for  the
     projects  to  achieve "success" in terms  of  cost-effective
     treatment and reversal of their water quality impairments.

3.   Neither  of the two groundwater projects have  succeeded  in
     defining  and addressing the protection of their groundwater
     resource at the project-wide level.

4.   Only  one  of  the  projects  appears  to  be  doing  enough
     comprehensive  monitoring of land use activity to tie  water
     quality changes to specific land treatment activities at the
     watershed level.   Preliminary results from this one project
     show that this requires much more detailed tracking of  land
     use activities than anticipated.

5.   Not  all of the  CM&E projects have water resources that are
     amenable  to  reversal  of impairment by treatment  of  only
     agricultural NPS  within the project area.

6.   No  provision has yet been made to provide for inter-project
     analysis  of the  CM&E effort at the conclusion of the  data
     collection phase.   Most of the RCWP projects, including the
     CM&E  projects  are not likely to produce definitive results
     for  ten years or more,  and so it would be  appropriate  to
     plan for ovrall analysis of results after the water quality
     changes have had time to develop.
                               vn

-------
                        TABLE OF CONTENTS
                                                             Page

PREFACE	 iii

SUMMARY AND CONCLUSIONS	  iv

INTRODUCTION	   1

MAJOR QUESTIONS RELATED TO AGRICULTURAL NPS
CONTROL	   2
     Water Resource Treatment Feasibility	   2
     BMP Effectiveness and Cost	   8
     Critical Area Selection  and Implementation	  13
     Institutional/Organizational Considerations	  15
     Water Quality Monitoring	  16

RELATIVE CONTRIBUTION OF RCWP CM&Es TO AG NPS
KNOWLEDGE	  17

ECONOMIC IMPACTS OF THE CM&E PROJECTS	,	  25
     Total Project Costs.	  25
     Offsite Benefits	  27
     Water Quality Improvements	  27
     Benefit Estimates	  28
     Recreation Benefits	  28
     Water Storage Benefits	  30
     Property Value Benefits	;	  30
     Water Conveyance	  31
     Water Treatment	  31
   '  Onsite Benefits. .	  31
     Benefits versus Costs	  32
     Limitations	  34

                   ROCK CREEK, IDAHO  —  RCWP 3

INTRODUCTION	•	  35
     Background	  35
     Perspectives of the Project	  35
     Land Treatment Strategy	  37
     Water Quality Monitoring Strategy	  37

BMP IMPLEMENTATION ACHIEVEMENTS	  38

AN.-UYS IS OF FARM LEVEL COSTS	  38

WATStt QUALITY DATA ANALYSIS		,	 ,  .  42
     Summary of Project Results,,,	,.,,,	  42
     Further Analysis and Interpretations	  45

PROJECTIONS	  63
IMPLICATIONS	  63
                               VII 1

-------
                                                             Page
             HIGHLAND SILVER LAKE, ILLINOIS — RCWP 4

INTRODUCTION	  65
     Background	  65
     Perspectives of the Project	  65
     Land Treataent Strategy	  66
     Water Quality Monitoring Strategy	  66

BMP IMPLEMENTATION ACHIEVEMENTS	  69
     Infomation and Education	  69

ANALYSIS OF FARM LEVEL COSTS	  69

WATER QUALITY DATA ANALYSIS	  72
     Summary of Project Results	  72
     Further Analysis and Interpretations	  74

PROJECTIONS	  77

IMPLICATIONS	  78
               ST. ALBANS BAY, VERMONT — RCWP 12

INTRODUCTION	  79
     Background	*	  79
     Perspectives of the Project	  79
     land Treatment Strategy.	*	  80
     Water Quality Monitoring Strategy	  80

BMP IMPLEMENTATION ACHIEVEMENTS	  80

ANALYSIS OF FARM LEVEL COSTS	  81

WATER QUALITY DATA ANALYSIS	  84
     Summary of Project  Results	  84
     Further Analysis and Interpretations	  89

IMPLICATIONS	  94


          CONESTOGA HEADWATERS, PENNSYLVANIA ~ HCWP 19

INTRODUCTION	  96
     Background	  96
     Perspectives of the Projects	,	"	  96
     Land Treatment Strategy	,	  97
     Water Quality Monitoring Strategy	  97

BMP IMPLEMENTATION ACHIEVEMENTS	  98

ANALYSTS OF FARM LEVEL COSTS	 100
                               IX

-------
                                                              Page

WATER QUALITY DATA ANALYSIS	  104
     Sunmary of Project Results	  104

PROJECTIONS	  106

IMPLICATIONS	  108


        OAKWOOB LAKES - PQISSETT, SOUTH DAKOTA — RCWP 20

INTRODUCTION	  110
     Background	  110
     Perspectives of the Project	  110
     Land Treatment Strate^ 	  Ill
     Water Quality Monitoring Strategy	  112

BMP IMPLEMENTATION ACHIEVEMENTS	  113

ANALYSIS OF FARM LEVEL COSTS	  115

WATER QUALITY DATA ANALYSIS	  117
     Summary of Project Results	  117

PROJECTIONS	  118
     Land Treatment	  118
     Water Quality Effects	  119
  •
IMPLICATIONS	,	:	  120

REFERENCES	  122

-------
                       LIST OF TABLES

Table

  1.  Major questions addressed by RCWP CM&E	   3

  2.  Aggregate onsite impacts ($1,000) for five
     CM&E RCWP projects	  26

  3.  Offsite benefits for three RCWP projects	  29

  4.  Estimated benefits compared with costs for
     three RCWP projects with a 50 year planning
     horizon (PRELIMIRAHY)	  33

  5.  Impacts of RCWP on a typical farm in the Rock
     Creek, Idaho RCWP project area	  39

  6.  Summary of Idaho DOE Data Analysis for Rock
     Creek HCWP	  42

  7.  Calculated percent difference from 1981 to 1983 by
     two methods:  normalized mass export loading (MEL)
     and adjusted downstream concentrations (D*) for
     downstream subbasin stations.  (Compiled from pages
     40 and 48 of the 1984 Idaho Annual Report, DOE.)	  43

  8.  Calculated percent difference from 1980 to 1983
     by two methods:  normalized mass export loadings
     (MEL) and annual logarithmic mean concentration
     (ALMC) far the Rock Creek stations.  (Compiled
     from page 41 and 49 of the 1984 Idaho Annual
     Report, DOE . )	  45

  9.  Summary of further analyses of Rock Creek RCWP
     subbasin data,  1981 - 1984	  46

 10.  Analyses of variance of downstream logarithmic
     suspended sediment concentrations, multiple com-
     parisons among years 1981 to 1984 using concen-
     tration as a regression covariate. (Idaho RCWP)	  48

 11.  Changes in adjusted downstream concentrations
     over time (1981 - 1984), represented by scenarios
     dp<*cribed in Figure 11. (Idahn RCWP)	  58

 1 '•',  Ooocentrat ion differences between the mean
     upstream and the mean downstream values observed
     ia 1981 and 1984 and projected for 1990 for six
     water quality parameters if the 1981 to 1984 trend
     continues through 1990	  64

 13   Data collection schedule and parameter coverage	, ,,.  68
                               XI

-------
Table                                                        Page

 14.  Onsite iapacts of agricultural activities on a
     416-acre typical farm in the Highland Silver
     Lake watershed	  70

 15.  Regression analysis-to test for differences
     between water quality at gage site 2 and gage
     site3. (Illinois RCWP)	  75

 16.  The relationship of average annual gross erosion to
     observed total suspended solids yield.  (IL RCWP)	  76

 17.  Annual impacts of BMP adoption for two typical
     dairy farms in the Jewett Brook subwatershed of
     the St. Albans Bay RCWP project	.  82

 18.  Net returns, livestock numbers, alfalfa acreage,
     environmental losses for various storage/appli-
     cation systems, varying nitrogen loss constraints
     for a 60-acre farm in the Conestoga Headwaters
     RCWP project area	 102

 19.  The costs and effectiveness of conservation
     practices for continuous corn silage with daily
     spread on a 5 percent slope, 20 tons manure per
     acre for the Conestoga Headwaters RCWP project	 103

 20.  Goals and accomplishments	*	 114

 21.  Estimated impacts of RCWP on a 480-acre farm in
     the Oakwoods Lakes - Poinsett project area
     (Poinsett soil, 3 - 5* slope)	 116
                                xn

-------
                          LIST OF FIGfcftfcS
Figure                                                       Page

  1.  Relative information contributions concerning
     treatability of water resources impaired by NFS .........  20

  2.  Relative information contributions concerning
     BMP effectiveness and cost ..............................  21

  3.  Relative information contributions concerning
     NFS critical area identification ........................  22

  4.  Relative information contributions concerning
     effective organizational methods for Ag NFS
     control projects ................. . ......................  23

  5.  Relative information con* -ibutions concerning
     water quality monitoring of Ag NFS control pro-
     jects [[[  24

  6.  Map of the Rock Creek Rural Clean  Water Program
     study area,  Twin Falls County, Idaho ....................  36

  7.  Farmer benefits per acre for the Rock Creek
     RCWP project ............................................  41

  8.  Downstream vs.  upstream suspended  sediment concen-
     trations for subbasin pair 7-1,7-4 for each year of
     period 1981  - 1984.  A significant decrease in the
     midpoints* over years is seen.  (Idaho RCWP) ............ . .  49

  9.  (a) Upstream and downstream suspended solids
     concentrations and (b) differences between
     downstream and upstream sediment concentrations
     for subbasin 2, 1981 to 1984.  (Idaho RCWP) ..............  51

 10.  Annual mean  logarithmic sediment concentrations
     for upstream stations over time, 1981 - 1984.
     ( Idaho RCWP) ---- . ......................................  53

 11.  Diagrams representing the possible scenarios
     for the comparison of upstream and downstream
     linear slopes of concentrations vs.  time ................  55

 1?  Yearly geometric means of upstream and downstream
     suspended sediment, concentrations  by subbasin.
     Srror bars indicate two standard deviations
           the mean. (Idaho RCWP) ....,, ................... ...  57
 1,3,  Normalized sediment loading vs.  BMP implementation
     for Rock Creek RCWP.   (The data  were obtained from

-------
Figure                                                       Page

14.   The average percent decrease per year relative to
     the initial yearly geometric Bean downstream sedi-
     ment concentration required to detect a significant
     decrease over a 2-, 4-,  and 10-year monitoring scheme.
     The range over all subbasins is shown.   Twenty samples
     per year are assumed.  (Idaho RCWP)	  62

 15.  Monitoring sites of the  Highland Silver Lake
     Watershed (page 36 from  the Summary Report Fiscal
     Year 1984 . )	  67

 16.  Mean concentration of solids, phosphorus,  and
     nitrogen at the tributary and St. Albans Bay
     trend stations for two years.  (From Vermont 1984
     Summary Report, page V-2. )	  86

 17.  Estimated average annual phosphorus runoff
     reductions for Jewett Brook (Station 21) of the
     St. Albans Bay RCWP.  (From Vermont 1984 Summary
     Report, page 111-23. )	  88

 18.  Paired watershed treatment schedule. (From VT
     1984 Summary Report, page V-3.)	  89

 19.  Mean concentrations in runoff from the LaRose
     farm paired watersheds.   (From VT 1984 Summary
     Report V-5.)	  90
                                            «
 20.  Regression analysis of paired "observations of
     total suspended solids,  treatment vs. control
     watersheds, 1983 - 1984. (VT RCWP)	  91

 21.  Regression analysis of paired observations of
     total phosphorus, treatment vs. control watersheds,
     1983 - 1984. (VT RCWP)	  92

 22.  Regression analysis of paired observations of
     orthophosphate, treatment vs. control watersheds,
     1983 - 1984. (VT HCWP)	  93

 23.  Mass export of phosphorus and nitrogen from the
     LaRose paired watersheds.  (From VT 1984 Summary
     Report, page V-7.)	  95
                                 xiv

-------
                          INTRODUCTION

     The experimental Rural Clean Water Program (RCWP) was estab-
lished by P.L. 96-108 to provide longterm financial and technical
assistance  to  owners  and operators of  agricultural  lands  to
assist in control of nonpoint source pollution (NFS).   Of the 21
RCWP   projects,   five  projects  were  designated  to   include
comprehensive  USDA/EPA  joint project water quality  Monitoring:

     Idaho:        Rock Creek Project
     Illinois:     Highland Silver Lake Project
     Pennsylvania: Conestoga Headwater Project
     South Dakota: Oakwood Lakes - Poinsett Project
     Vermont:      St. Alban's Bay Project

The five projects,  which we refer to as  CN&E projects have  the
additional   mission   to   aonitor  "...basic   hydrologic   and
•eteorologic  factors, ...identify  and quantify changes it water
quality attributable to the installation of BNPs [best Management
practices],"   and  "wherever  possible,  identify  and  quantify
changes in land use, land use patterns and farming practices that
will affect the quantity,  quality,  or timing of nonpoint source
pollutants  reaching  an aquatic system."(Federal Register 7  CFR
Part 700.41).  The  CM&E projects are further expected to "detail
information  as to the number and location of  sampling  stations
and the frequency of sample collection [to document these aspects
successfully],"  and  to "identify the positive and negative  im-
pacts  on  landowners  in the  project  area,  and  estimate  the
community  or  off-site  benefits  expected  of  the  project  if
completed as planned."  The Idaho, Illinois, and Vermont project^
•were  initiated in late 1980;  the Pennsylvania and South  Dakota
projected were started one year later.

     The  ultimate  goal  of the RCWP  agricultural  NPS  c >ntrol
effort  is  to alleviate actual or potential water  resource  use
impairments  caused by agriculture.   The role of  CM&E,  in par-
ticular,  is  to provide the additional information to  establish
cause-effect  relationships.   This  report evaluates  the   CM&E
projects  and presents an analysis of how the  CM&E projects  and
the  CM&E Program have contributed to the state of current knowl-
edge in  the field of agricultural NPS control.

     The  NWQEP  1984 Annual Water Quality Report  proposed  nine
questions which would need to be answered in order to provide the
essential  information  to address the problem of Ag  NPS  affec-
tively    In this report we have broken these nine questions into
20 mure specific questions,   The questions are grouped into  the
following subject areas:

          - water resource  treatment feasibility
          - BMP effectiveness and cost
          - critical area  selection
          - institutional/organizational considerations
          - water quality monitoring.

-------
One  result  of this ^pfroach is an ov*r*iew of just bo** mtch  of
the total NFS control puzzle has been pieced together at present,
and what portion of this has cone from the  CM&E Program.

     Complete  summaries  and analyses of each  CM&E project  are
presented  as  separate  chapters of  the  present  report.   Our
emphasis  is  on  summarizing the findings of  the  projects  and
laying  out the analysis and interpretation of the water  quality
data.   A supplement to this report has been prepared  separately
to  present  the details of the analytic and statistical  methods
used for our analyses and synthesis.
       MAJOR QUESTIONS RELATED TO AGRICULTURAL NFS CONTROL

     Table 1. summarizes the present and projected  contributions
of  the   CM&E program to 20 major questions which  comprise  the
•ain  body  of knowledge needed to addres   national  NFS-related
water  quality  problems.   The questions are treated  in  detail
below.
      B§§P.urce Treatment Feasibility

1.   What  types  of  water resources  can  be  most  effectively
restored through land treatment efforts?

     Among.the impaired water resources being addressed in Ag NFS
control projects are streams, lakes, estuaries, groundwater aqui-
fers,  and 'irrigation canals.  All of these except esturaries are
present in the  CM&E projects.

     From the Idaho  CM&E project as well as several other irriga-
tion  projects,  it has become clear that irrigation canals  show
the  most  rapid water quality improvements in  response  to  BMP
implementation.   This  is  because of the  decreased  effect  of
meteorological  variables,   greater  control  options  over  the
management  of the water resource,  and the fact that agriculture
is  usually the only significant pollutant  source.  ' Because  of
their  short hydraulic residence time,  streams appear to be  the
next  most  treatable water resource,  although the  treatability
drops in direct proportion to the size of the watershed.  This is
attributable  to  an increasing variety of sources  and  the  in-
creasing time lag in flushing pollutants from the system.   Lakes
and  groundwater  are  proving  to be the  most  difficult  water
^sources  to restore;  lakes because of uutrient  recycling  and
because sediment filling can only be slowed,  not reversed.   The
difficulties  with  addressing groundwater impairments have  been
-•elated  either to the problem of defining the contributing  land
iurface area,  or to the potential conflicts between some surface
and  groundwater protection BMPs,   The r "suits of the  PA  field
studies,  however, indicate that their Intermediate depth ground-
water  (30-100  ft.) can be very responsive to surface  activity,
and thus is amenable to BMP treatment.

-------
Major  Questions
                                                    Table  1.
                                                      Addressed
                    by    RCWP   CM&E
Questions
Present by 1990
I I « 1
CoMents
Hater Resource  IreaUent Feasibility.
1.  What types  of water Resources can  be  lost
    effectively reversed through land treatment?

la. Wiat  types of Ag NFS use  iipairients   can
    be  most effectively restored through me-
    dial efforts?

ib. Mhat timeframe is required to observe utter
    quality results from land treatient pro-
    grais?

Ic. Do  groundwater resources  respond  rapidly     PA
    enough  to  reflect changes in land  manage-
    lent within a ten year timeframe?

Id. How iuch problei definition is needed to       *
    identify and develop a successful and cost-
    effective NPS control prograi?
    Are those Mater resources impaired by  irri-         IL
    gation  return flow tore readily  treatable         ID
    by BHPs than  those iipaired by stori  flow?         VT
BHP Effectiveness §nd Cost
2.  How  effective  are  the  various  BHPs   in     PA
    reducing pollutant inputs to water   resour-     VT
    ces?

2a. How do external (uncontrolled) factors such         VT
    as  leteorological  conditions  affect   the         PA
    ability  of BHPs to protect or restore   ii-         IL
    paired water resources?
?h. iihdt  JHPs  are aost cost-effective in
    .-Jrowni  qroundwater iipainents?
                 ad-
                  do gco'ifld*ater 8HPs
             *ith surface water SHPs?
PA
                         PA
                                    PA
                                    SO
                                    IL
                                    ID
                                    VT
                                             •Statistically  based answer  will  be
                                              derived using  all projects.

                                             ^Statistically  based answer  Hill  be
                                              derived using  all projects.
                                             *0nly projects  which achieve adequate
                                              implementation will contribute.
                    'Statistically based answer will be de-
                     rived.  See  ERS discussion of econo-
                     mic problem assessment.
                                               •
                     Statistical answer from all projects with
                     high implementation.
                                    PA  IL    ID project does not report infonation to
                                    VT        separate the effects of BHPs 12 and 13.
                                    VT        ID does not  address meteorologic variability.
                                    PA  IL
PA
SD

PA
SD
        A -•  Substantial contribution to  answer
        § :  Partial contribution to answer
      7ng of questions corresponds with  text.

-------
Table 1.  continued
                Questions

2d. What  degree of sediient reduction  can be
    achieved  by BHP itpleientation  at  a Hater-
    shed level?

2e. Mtat  degree of nutrient reduction  can  be
    achieved  by  IHPs  at a.  natershed level?
2f. Miat  degree of toct&riti r«fetftia« ea*  be
    achieved  through  various  lewis sf  K6f
    itpletentation.
3.
         A.rj § §§le.£jipn
    Where in a watershed should BHPs be  placed
    to   restore  or  protect  a  given   water
    resource?
3a. HOH  tuch  of a watershed aust  be  treated
    xith  IHPs  to restore or protect  a  given
    water resource?

3b. BhaV criteria are appropriate for selecting
    critical  areas for protection  of  ground-
    Hater frw DPS?
IC§Uiytionii/Qrganizitignal Considerations
4.  Hhat are the  lost cost-effective Mans
    obtaining farter participation?
                                             of
4a. Can  a  project effectively address ground-
    Mater and surface Hater iipainents  simul-
    taneously?
Hater gyality. Mgnitgring
Sa. Khat  are the groundHater levels of  pesti-
    cides  that  can be expected in areas  nith
    intensive agriculture?

So. Ho* tuch change in HPS pollution lust occur
    '0  be Detectable by a comprehensive  loni-
    $ on fig prograi?
                                                  Present
                                                       VT
                                                       ID
                                                       IL

                                                       VT
                                                       PA
                                                       ID
                                                       IL
VT
IL
ID

IB
                                                        PA
                                                    PA   SD
                                                        VT
                                                        IL
                                                        10
                                                        PA
       by 1990

        A   t

        VT
        ID
        IL

        ID  SD
        VT  IL
        PA
VT
IL
ID
                                                                   VT
                                                                   SD
                                                               ID  IL
                                                               VT
                                                               PA
        PA
        SD
        PA
        SD
        VT
        IL
        ID
                      Couents

           PA has too little land treatment to
           contribute to this question.  SD does
           not wnitor at watershed level.
                  Other RCUP projects are substantially
                  addressing this question.
Not addressed by PA or SD.
          Mot addressed by PA or SD.
            *    'Statistical  answer  Hill be derived  froi  all
                  projects.  See ESS  discussion of c/s rates.
          This  lay be the most definitive and  important
          result froi tfie  DttEs.
                                                                    PA
code:   A : Substantial contribution to answer
        B ; Partial contribution to answer
Kutbedng of questions corresponds with text.

-------
la.  What  types  of Ag NFS - caused use impairments can be  most
effectively restored through re»edial efforts?

     The most common impaired or potentially impaired uses  being
addressed  by  agricultural  NFS projects  are  domestic  supply,
reservoir storage, recreation, commercial fisheries and aesthetic
enjoyment.   The  primary impaired uses which exist in the   CM&E
projects are domestic supply (IL,  PA,  SD,), recreation (ID, IL,
SD,  VT), aesthetic enjoyment (ID, IL, .SD,  VT), fishery (ID), and
reservoir storage (IL).   The answer to what types of use impair-
ments  are most treatable will cone from a  statistical  analysis
using  all projects with identified uses impaired by agricultural
NFS.   At  the present time it appears that swimming  impairments
•ay  be the easiest to restore since often all that is needed  is
to  reduce  maximum  fecal  coliform  concentrations  below   200
mpc/lOOml.   In many systems this can be accomplished by treating
just  a  few of the most critical animal  production  operations.
Impairments of domestic water supply may be  -he next most readily
amenable  to restoration through BMPs.   This is at least  partly
because  the impairment is often caused by just a single physical
or chemical parameter,  and the extent of impairment can be quan-
titatively defined in terms of drinking water standards.   Hence,
once maximum nitrate levels, for instance,  fall below 10mg/l, the
use impairment has, by definition, been alleviated.  At the other
extreme  are fishery and aesthetic impairments which  are  highly
subjective in nature.  We are  finding that a claim of success in
treating  these  problems  can  lead to a  public  perception  of
improvement.  Thus,  to  some  extent  the  program  may  reverse
perceived  impairments,  without causing a  measurable  change  in
water quality.
Ib.  What time frame is required to observe water quality results
from land treatment programs?

     Almost  by  definition all RCWP projects,  which achieve  at
least  a moderate level of BMP implementation and  which  conduct
sufficient water quality monitoring, will contribute to answering
this question by the end of the program.   Of the  CM&E projects,
present  indications  are that only PA will lack  sufficient  BMP
implementation  to  contribute any  watershed  level  information
related to the question.

     Based  on  the  present status of the  CM&Es  the  following
statements related to the tine frame of wat^r quality results can


     <     Four  years  is sufficieat to begin observing  sediment
          reductions in irrigation cartaio when about half of  the
          land area is treated with BMPs,

     2.   Bringing  half of the manure under best management will
          not produce statistically significant stream phosphorus
          reductions in a three year  time frame.   (We  estimate
          that five years will probably be required.)

-------
     3.   Two and oft* half years is insufficient for documenting
          stream/lake  sediment or turbidity  changes due to BMP
          implementation on 20 percent of a  watershed  critical
          area.
Ic.   Do  groundwater resources respond rapidly enough to reflect
changes in land management within a ten year tine frame?

     The  PA field studies prove that Moderate depth (30-100  ft)
groundwater in the Conestoga Watershed is significantly recharged
by  aajor precipitation events and that almost complete  recharge
occurs  over a one-year period.   In this situation  we  conclude
that  the groundwater resource can respond quickly (probably  1-2
years) to changes in land Banageaent.   The SD project's plot and
field  studies wi^l provide substantial information  toward  this
question also,  buc that information is not presently available.
Id. How much problem definition is needed to identify and develop
a successful and cost-effective NFS control project?

     Our  answer  to this question is based primarily  on  deter-
mining the correlation between how thoroughly individual projects
have  conducted various aspects of problem definition  and  their
projected  degree  of success and efficiency in addressing  their
water  resource impairment.

     This  analysis (see 1985 RCWP  Cross-Project  Report,  NCSU)
indicated  that only three problem definition steps are necessary
to  maximize  the probability of a  successful  (improving  water
resource) and relatively cost-effective project:

     1.   Determine the extent to which a use or projected use of
          the water resource is impaired,  and make a rough esti-
          mate  of  the economic and aesthetic cost of  this  use
          impairment.

     2.   Estimate the'relative magnitude of sources that can  be
          treated  through the program and those that are  beyond
          the  program's jurisdiction,  such as point sources  or
          background.  (These estimates should be updated as more
          information is obtained.)

     3    Determine  what  pollutant(s) is actively  causing  the
          impairment,  and  estimate  how much reduction in  that
          pollutant(s)  is needed to achieve the desired  effect.
          (Subsequently  BMP  implementation goal?; should be  set
          to achieve this amount of reduction.)

     The  Economic  Research Service has provided  the  following
input  on the importance of assessing the economics of the  water
resource impairment as it relates to the  CM&E projects.

-------
     "Pre-project  assessment  of  economic  impairment  and
potential benefits would aid in project selection and  plan-
ning  since potential benefits vary considerably among areas
and can't be judged from simply looking at levels of  pollu-
tion.   For example,  the Idaho,  Illinois, and Vermont pro-
jects  all  had  highly polluted water,  but  ERS  (Economic
Research Service) estimated economic benefits to the  public
from  controlling  the pollution ranged from under  $250,000
for  the  Illinois project and $1.6 million  for  the  Idaho
project  up  to nearly $5 million for the  Vermont  project.
Considering  these benefits relative to project  costs,  the
Vermont  project  also has a much higher Benefit/Cost  ratio
than the other two. .  If RCWP funds had not been  sufficient
to fund all three projects, and if the goal of RCWP had been
to  maximize  public  benefits for  the  money  spent,  this
comparative  information  on economic impairment and  likely
benefits versus costs would have been invaluable to decision
makers.

     The  importance of tying agricultural  nonpoint  source
pollution  to water quality and determining an economic  im-
pairment can be illustrated in the case of the Illinois RCWP
project.   When the project was initiated, the loss of stor-
age  capacity  from deposition of sediment in  the  Highland
Silver  Lake  was  identified as the  principal  impairment.
Reductions in erosion in the watershed would reduce sediment
delivery  to  Highland Silver Lake,  the primary  source  of
drinking water for the City of Highland.   Substantial  off-
site  benefits  were envisioned through elimination  of  the
need  for dredging the lake or finding an alternative source
of water.  However, subsequent analysis of the Lake's* silta-
tion revealed that much of the sediment was not settling out
on  the lake  bottom but rather was remaining in  suspension
and passing through the lake.   Also the reservoir  capacity
was  large relative to future demand.   Thus,  there was  no
significant  problem in terms of lost water storage capacity
in  the  lake  and the primary benefit  identified  for  the
project had negligible economic value.

     A similar situation occurred in the Rock Creek  Project
in Idaho.   Reduced siltatioa of power-generation reservoirs
behind dams on the Snake River was identified as a potential
significant  benefit from the Rock Creek project.   However,
subsequent evaluation revealed that reductions in erosion in
the  Rock  Creek watershed were  unlikely  to  significantly
affect  the  water storage facilities 100 miles  downstream.
Although  measurable reductions in sediment delivery to  the
stream and subsequently to the Snake River occur,  the Snake
River itself will tend to pick up replacement sediment  from
streambanks and the river bottom.

     A  key  factor affecting the economic  impairment,  and
extent of potential benefits,  appears to be how many people
are  being affected,  particularly with regards  to  recrea-
tional  opportunities.    In  the Vermont project the  likely

-------
     recreatiui  b*?fi
-------
          manure $j}f^a4ing.  The differ^&ce in results appears to
          be  due to the fact that winter and early spring events
          produce the greatest pollutant yields in VT when Manure
          is spread during winter.

     3.   Terraces  -  Terraces  were installed at the  PA  field
          sites  during  the past year.    Thus,  some  definitive
          information  on the effects of terracing  on  pollutant
          surface  and subsurface transport is anticipated in the
          near  future.   On the basis of the  pre-implementation
          water  quality data,  the project monitoring  personnel
          have  made some projections on the effects of  terraces
          which  can  be viewed with a moderate level  of  confi-
          dence.   It  is projected that terraces  will  increase
          nitrate  concentrations  of water transported  to  both
          surface  and  groundwater because of increased  contact
          time  between manure and precipitation.   Increases  in
          surface runoff nitrogen loaud will be moderated by  the
          reduced  runoff volume from the terraced  field.   Sus-
          pended sediment and total phosphorus losses in  surface
          runoff should be significantly reduced.

     Our  analyses of the Idaho water quality data  show  conclu-
sively   that  a   reduction  of  drainage  canal  sediment  con-
centrations  has  occurred  as a result  of  BMP  implementation.
However,  both BMP-12 (sediment retention basins) and BMP-13 (ir-
rigation  management ) have been integrated in such a way that it
is  not  possible to separate the effects of the  two  practices.
Either practice may be affecting the majority of the improvement;
all  we can presently conclude is that the combined effect  is  a
significant reduction in sediment.

          The  Illinois and South Dakota projects have  not   yet
produced any new BMP-water quality ef ectiveness information.  It
is  anticipated that the IL' field studies will eventually provide
information-on the effectiveness of sediment and nutrient control
practices on natric (fine-textured) soils.   The SD project  will
eventually  produce strong recommendations concerning the  effec-
tiveness of conservation tillage, fertilizer management,  and pes-
ticide management.

     ERS  has provided the following economic perspective on  the
effectiveness  of  these BMPs as they are employed in  the   CM&E
projects:

          "In the case of the Idaho RCWP .project,  initial empha-
     sis was given to fairly costly structural BMPs which trapped
     •so U:men t  at  the end of the fie id or  improved  irrigation.
     This  emphasis resulted in very high costs for BMP imp!^men-
     tation and maintenance.   Estimated total cost of  implemen-
     ting  the  RCWP project as originally designed exceeded  $10
     million over a 50-year period, including both government and
     private costs.

          Alternative  BMPs were examined to determine  the  most

-------
     cost   effective  way of  reducing  sediment  delivery   to   Rack
     Creek.   One such BMP was conservation  tillage  (including no-
     till),   which  if  it could  be implemented  throughout   the
     watershed,   would  not only reduce  sediment  delivery  below
     that   projected for the  original  set of  BMPs,   but  it  would
     also   reduce costs.   However,  feasibility  of  conservation
     tillage  had   not  been  demonstrated   for  furrow   irrigated
     crops prior to 1980.  The preliminary budgets indicate  that
     farmers  using conservation tillage in place of conventional
     tillage  would receive a net  benefit through a reduction  in
     total operational costs.   In addition,  conservation tillage
     would  help retain  soil  in place  on the  field,   rather   than
     trapping  it at the bottom,   and  thus  produce  a larger   soil
     productivity benefit, also to the farmers' advantage.

          In the case of the  Conestoga Headwaters RCWP project in
     Pennsylvania,  the   critical   area  was redefined after   the
     project's  original  conception in  order  to  emphasize   the
     carbonate area as opposed to  the  triasic soils.   The deliv-
     ery  of pollutants  from  the carbonate  soils  was found to  be
     much  greater.   Also in  the Pennsylvania project, the mix of
     BMPs   is being reconsidered because of the unique high   con-
     centration  of animals per acre,  the  highest  in the United
     States.   In such a situation,  animal waste storage, gener-
     ally   a good BMP for reducing the discharge  of animal  waste
     constituents  to water bodies,  may be detrimental  to  water
     quality.   Animal  waste storage  conserves nutrients by  pre-
     venting volatilization.   Also,  it may  make it  easier for the
     farmer to increase  the number of  animals.  Thus, this BMP on
     some   farms  in  the project  area  could  result in larger
     amounts  of higher  nutrient manure  being applied at a  given
     time   than would otherwise occur,  and more  than can be  uti-
     lized by crops.  These  excess nutrients must  go somewhere.
     The  CREAMS  model  used  as part of  the  economic evaluation
     indicates  that  the excess nitrate-nitrogen tends   to   move
     downward  into  the  groundwater  and  subsequently to   the
     Conestoga River in  baseflow.
                                                    4
          A  preferable  BMP might  be one that involves short-term
     manure  storage  and an application   mode  which   increases
     nitrogen volatilization.  Other  alternatives  may be to   re-
     move   the  manure  from  the farm  to other  areas which   can
     utilize  the  nutrients,  or   to  reduce  the  incentives   for
     farmers in the project  area to have such high  concentrations
     of animals on their farm."
2a.   How do external (uncontrolled)   factors such as meteorologic
conditions  affect  the  ability of  BMPs to  protect  or  restore
         water resources?
     This  is a complex question which must consider whether  and
to  what extent large runoff events may overwhelm BMPs and  which
BMPs  are most effective over a wide range of meteorologic condi-
tions.   Clearly,  there will be a probabilistic coaponent to  the

-------
answer.    At  this  fdint the  CM&B provides only  very  Halted
information  relat*4 to this question.    Meteorologic  conditions
have only a Minor effect on the ID irrigation canals because flow
is  mechanically controlled and the project is some distance from
the source of water.   Therefore,   the effectiveness of the  BMPs
within the Rock Creek (ID) irrigation system has little relation-
ship  to meteorologic conditions except in the rare instance of a
major runoff event.   For projects emphasizing animal waste  man-
agement  (VT,  PA) the timing of major runoff events relative  to
spring  manure  spreading could greatly affect the  year to  year
water quality effectiveness of BMP 2.   The SD project will even-
tually provide information on the effectiveness of BMP-9 (consei—
vation tillage),  BMP-IS (fertilizer management) and BMP-16 (pes-
ticide  management ) mider a wide variety of aeteorologic  condi-
tions.
2b.  What BMPs are most cost-effective in addressing  groundwater
impairments?

     Substantial   information  relating  to  this  question  has
already been generated from the PA project.   These results  have
come  from monitoring of the field sites and from CREAMS modeling
done by ERS.

     Preliminary  results from the field site monitoring  clearly
indicate  that a nutrient management (both fertilizer and  animal
waste)  program,  which consists of soil testing  and  subsequent
matching of nutrient application to crop utilization  rates,  may
be the most effective BMP for reducing nitrogen inputs to ground-
water.   The  preliminary economic analysis suggests that,  where
export of manure is required to achieve this situation,   the cost
may be relatively high.   Exporting poultry manure is more  cost-
effective  than exporting cow manure.   The field site work  also
suggests  that  building manure storage structures which allow  a
more  timely  manure spreading will have little effect  on  total
groundwater nitrogen inputs.   This is because the increased crop
usage   efficiency,  achievable through manure storage  capabili-
ties,  is negated by the increased  amount of nitrogen  available
from stored versus daily spread manure.

     In  Pennsylvania,  the CREAMS  model predicted that  conser-
vation  tillage has no real effect but  terraces have a  negative
effect on nitrogen contamination in groundwater.

     The two PA field site experiments are set up to develop much
more  information  related to this question over the rjext two  to
three years.   The SD field sites will also provide a substantial
body of information related to the cost-effectiveness of  conser
vation  tillage systems for controlling groundwater nutrient  and
•pesticide inputs.
      To  what extent do groundwater BMPs ccnflict  with  surface
      BMPs?
                               11

-------
     Of  all  common agriculture related pollutants only  soluble
nitrogen   forms and soluble pesticides will generally present  a
potential conflict between reducing surface water and groundwater
inputs. In the case of soluble nitrogen forms the experience from
PA  and  from  other field studies  indicates  that  all  surface
runoff-reducing  practices  increase groundwater inputs  to  some
extent.  The  degree  of conflict depends on:   1)   the  relative
percentage of precipitation which becomes surface runoff, subsur-
face runoff, and groundwater recharge,  and 2) the degree to which
the  runoff-reducing  practice increases the time of contact  be-
tween water and the nitrogen source (i.e.  fertilizer or manure).
Practices  which  are  designed  to  match  nutrient  application
amounts  and timing to crop needs do not appear to  present  con-
flicts between ground and surface water objectives.

     It  is  anticipated that the SO field studies  will  provide
additional  information  related to thi:  question  provided  they
initiate surface runoff monitoring as proposed.
2d.   What  degree of sediment reduction can be achieved  by  BMP
implementation at the watershed level?

     The answer to this question will develop out of results from
all RCWP projects and other projects for various climates, topog-
raphy,  soil types and crops.  The greatest contribution from the
CM&Es  will come from the Illinois and Idaho projects.

     Idaho  has  shown  reductions in irrigation  canal  sediment
levels in the subbasins with high levels of sediment control BNPs
installed.   Our  analyses show that these reductions are in  the
range of approximately 40- 60  percent.     Additional land treat-
ment data are needed to tie,  unequivocplly,  the observed reduc-
tion to BMP application.

     The  overall level of BMP implementation is still too low in
the Illinois project to observe a change of in-stream or  in-lake
sediment  levels.   Projected levels of BMP  implementation along
with  the  extent  of sediment monitoring  indicate  that  useful
information related to this question will be produced.
2e.  What degree of nutrient reduction can be achieved by BMPs at
a watershed level?

     Considerable information on nutrient loading and  concentra-
tions  reduction from land treatment ha? been developed by the VT
•and PA projects.   The VT project has projected, based on the BMP
implementation level,  water quality data,  and modeling results,
that  total  P  loadings from its  ssost  extensively  implemented
subbasin  will  be  decreased  by 30  percent  over  the  project
fcimeframe.   A  57 percent decrease in dissolved P is  projected.
ft  appears  that these loading reductions would be even  greater
except  that  very high phosphorus levels have built  up  in  the


                               12

-------
soils  as  a result 
-------
  Recent  results ff&m the projects suggest a Beans of  improving
critical  area identification by analysis of the  monitoring  re-
sults.    Water  quality  data  presented by the ID  project  show
clearly  the location of at least one untreated  critical  source
and  one  subshed that will not respond with any substantial  im-
provement to further land treatment.   Similarly, data from the IL
project  suggest  that  one section of  the  project  contributes
disproportionately  to the total load of sediment to  the  reser-
voir.   These  examples show that critical area selection can  be
refined as water quality data beeCMC available.

     The IL,  ID, and VT  projects are expected to provide exten-
sive  information that can be used in developing  procedures  for
identifying critical areas and installing effective BMPs.  The PA
and  SD  projects are not likely to contribute  substantially  to
answering  this  question.  PA will contribute little because  it
does  not  follow its critical area guidelines due  to  low  farm
participation.   The  SD project also will contribute little  be-
cause it does not monitor the overall water quality effectiveness
of the project.
3a.   What  fraction of a watershed must be treated with BMPs  to
restore or protect a given water resource?

     The  monitoring  program associated with the ID project  has
shown a water quality improvement at the subwatershed level  that
appears  to  be  associated with treatment of 36 percent  of  its
critical  area.   Further documentation of the actual  extent  of
implementation,.  however,  is  needed  in order to  confirm  this
result.   The  IL project has treated 10 percent to 20 percent of
its  critical  area and has not yet observed  any  water  quality
response.  In this case treatment of the entire critical area may
not   have   sufficient   effect   to   improve   water   quality
significantly.

     Results  from ID and VT should provide a  substantial  basis
for answering this question by the completion of the project.
3b.  What  criteria are appropriate for selecting critical  areas
for protection of groundwater from nonpoint sources?

     This  question  has been considered extensively in  the  two
groun Iwater   CM&E projects,  PA and SD.   Both projects have de-
veloped guidelines for the selection procedure:  PA considers the
•exterl   of excessive nutrient application at the surface  (number
> r  animal units on the farm),  the soil  permeability,  and  the
location  of the impaired groundwater (whether carbonate or  non-
carbonate area);  SD considers proximity to regional groundwater,
thickness  of soil above aquifer,   and drainage  characteristics.
At  present PA has information from field monitoring which  shows
v.hat  67  percent of the groundwater wells in the carbonate  area
exceed 10 mg/1 nitrate-N,  while only 27 percent of the wells  in
h.b-"  ooncarbonate area exceed this standard.   These studies also
                               14

-------
show  that groundw8t«r nitrate levels respond to the  application
of  excessive amounts of Manure.    The PA results provide  direct
confirmation  that their critical areas were  selected  properly.
By  the  end  of the project,  data from both SD  and  PA  should
provide a substantial basis for answering this question.
                             Considerations
4.   What  are the »ost cost-effective Beans of obtaining  farmer
participation?

     The   CM&E projects will contribute some information  toward
answering  this question,  because a wide variety of  Information
and Education approaches including newsletters,  radio bulletins,
public meetings,  personal contacts between project personnel and
farmers, local newspaper articles, and on-farm demonstrations are
used.  In addition, the levels of incentives vary froa project to
project,  and in some cases,  such as SD, the incentive may be in
the  form of services such as scouting  programs.   Socioeconomic
analysis,  including  farmer workshops such as that conducted  in
the  SD project offer the potential for a great deal of  informa-
tion.  A complete analysis of the different institutional/organi-
zational  considerations in  CM&E and other RCWP projects  should
provide  worthwhile  lessons  that can be used  for  guidance  of
future implementation programs.

     ERS  has provided the following input .relating to how  cost-
sharing  resources could be used more efficiently in future  pro-
grams based on the experience of the  CM&Es.

          "An implication of the economic evaluation of the  CM&E
     projects  is that lower cost-share rates may be feasible  in
     some cases.   For example,   budgets for conservation tillage
     in the Idaho,  Illinois, and South Dakota projects all indi-
     cate  a  net cost savings on the average  over  conventional
     tillage.  This suggests that relatively more effort in these
     projects might go into information and education,  and  that
     financial  assistance levels may be reducible in the  second
     or third year after the farmer tries the improved practices.

          Another example is found in the Vermont project.    Here
     the  primary  BMP being implemented is animal waste  storage
     which conserves nutrients and reduces the need for purchased
     fertilizer.   The  economic  evaluation indicates  that  the
     =!fvings  in  fertilizer purchases over  a  20-year  planning
     hvrizon  allow  a faraer to recapture most of the  costs  of
     ixistalling  the animal waste storage  structure.    Thus,  if
     farmers understand this, they may be willing to adopt  manure
     storage  structures  at a lower government  coat-share  rate
     than  the  present 75 percent available in the  project.   A
     substantial  savings to the government may be feasible  with
     jainimal reduction in the overall level of implementation  of
     the animal waste storage BMP,
                               15

-------
          A  factor  which needs greater consideration in  estab-
     lishing cost-share rates for structural BMPs,  such as  ter-
     races  or animal waste storage structures,  is the potential
     income tax deduction for this type of investment.   In  some
     cases,  much  of the farmer's costs for structural BMPs  can
     be deducted from income taxes, which may make feasible lower
     cost share rates."
4a.   Can  a project effectively address groundwater and  surface
water impairments simultaneously?

     The  two groundwater  CM&E projects should  contribute  sub-
stantially to answering this question.   The experience in SD, to
date,  indicates  that  severe  problems can arise  in  nmltiple-
objective  projects  unless  cons-*erable effort  is  devoted  to
determining  how  to combine objectives properly.   The  lack  of
clarity  and  definition  of the surface  water  and  groundwater
objectives has slowed the development of the SD project.   Infor-
mation from the PA project shows that there are some  BMPs,  such
as  fertilizer  management,  that  can reduce  both  surface  and
groundwater impairments.  Heavy reliance on runoff-reducing prac-
tices can have negative effects on groundwater quality,  thereby,
solving surface water problems by impairing groundwater.
Water Quality Monitoring

5a.   What  are  the groundwater levels of pesticide that can  be
expected in areas with intensive agriculture?

     Both  the  SD project and tht, PA project  routinely  monitor
pesticide  concentrations in their  groundwater  samples.   South
Dakota  assays  for a wide range of herbicides  and  insecticides
that are used in their project area,  but too few data are avail-
able  at present to answer this question.   Herbicide  concentra-
tions  in  the  PA aonitoring wells  show  significant  increases
following field application periods in the spring.   The  concen-
trations observed in PA were consistently less than 1 ppb,  which
does not constitute a water use impairment.
5b,    How  much change in nonpoint source pollution must occur to
be detectable by a comprehensive monitoring program?

     i'hc monitoring data from the ID, PA, VT, and XL projects are
iur't ioiently detailed to develop good estimates of the background
variability  in  water quality in *»»eh of these  projects.   This
information  can  be used to assess future  projects  in  similar
areas  to  evaluate   a priori  whether a  monitoring  effort  is
feasible or not.  The data to estimate the natural variability of
water  quality,  and the minimum detectable water quality  change
may be two of the most valuable outcomes from the  CM&E program.
                               16

-------
    RELATIVE CONTRIBUTION OF RCWP  CM&Es TO AG NFS KNOWLEDGE

     The  preceding section provides an overview of how much   the
CM&E  projects  have and will contribute to the knowledge  needed
to effectively address Ag NFS pollution problems in coming years.
It  also summarizes present  CM&E knowledge contributions as  they
relate this "big picture".   It should be  noted,  however,   that
there  are  numerous  other  sources from  which  this  essential
information  is  being generated.   These include,  but  are   not
limited  to:   other HCWP projects,  on-going ACP  projects,   the
completed  bat  not  ful ly' analyzed MIP  projects,  the  EPA   108
program,  the PI 566 programs  recently initiated state-level  NFS
control  programs,  and  a large number of plot and  field  level
research experi. ^
     We  have identified over 70 watershed level Ag NFS  projects
nationwide  and have located published results of several hundred
plot and field studies related to Ag NFS.  A full analysis of the
contributions of these various efforts to answering the 20  major
Ag  NFS questions above is planned for future reports.   However,
at this point we can summarize the  CM&E contribution relative to
other Ag NFS information sources.

     Figures  1  through  5  illustrate  the  projected  relative
contribution  of the RCWP  CM&Es,  RCWP general  projects,  other
watershed  level projects and plbt/field studies  to  information
base  needed  to protect arid restore water resources impaired  by
agricultural  NFS  effectively.   We  have divided  this  Ag  NFS
knowledge into the broad categories of :

     1.  water resource treatment feasibility,
     2.  BMP effectiveness and cost,
     3.  critical area selection,
     4.  institutional/organization considerations, and
     5.  water quality monitoring.

The  percentage assigned to each information source is  based  on
our  knowledge  of how many on-going projects address each  ques-
tion.   The  percentage   for CM&E is based on whether  the  CM&E
projects  will contribute highly quality information or more  in-
formation relati/e to other projects on a given question.

     Figure  1 displays the relative contribution of  information
oij  wat^r resource treatment feasibility covered in Questions  1,
1A,  18, 1C,  ID, and IE.   As noted previously the answc-rs to somp
of thes- questions are developed from statistical analysis of all
projects which r jve a documented water quality problem caused  by
agricultural  NFS and which have a high level of BMP  implementa-
Hon.   On  this  basis we conclude that the  CM&E projects  will
 unti ;:.;*ite  to this subject area in proportion to the  number  of
projo-/>+s  which meet these criteria.   By 1990 wo project that an
adf--. i'> ^ '-'  knowledge  of  the  treatability  of  water   resources
afftc'.M  by  Ag  NFS  will have been  developed  with  the  RCWP
yrogr-v-i:* being the primary contributor.
                               17

-------
     Figure  2  shows the relative contributions to  the  overall
area of BMP effectiveness and cost as encompassed by Questions 2,
2A,  2B,  2C,   2D,   2E,   and 2F.    At present the Majority of our
knowledge on BMP effectiveness cones from plot and field studies.
However, as shown in Figure 2, we believe that the information to
be  gained  from  these studies is approaching its  upper  limit.
This  is  because  our major remaining need in this  area  is  to
determine  actual "in-stream" effects of BMPs at a  watershed  or
water  resource level.   This information can only come from  NPS
control projects with a high level of land treatment, and a water
quality  monitoring aystea capable of quantifying water  resource
response.

     A  largely uncharted area of NPS control is the  process  of
identifying  critical areas within watersheds for maximizing  the
water quality effects of BMPs.   As shown in Figure 3 our present
information  comes almost entirely from RCWP which represents the
first legitimate attempt direct BMP implementation on this basis.
Plot  and small field experiments provide very little insight  on
the subject.  We project that RCWP will have made substantial in-
roads  on  correct  critical area selection by  the  end  of  the
program.

     All  watershed  level  by  NPS  control  projects  face  the
problem  of obtaining farmer participation within the context  of
voluntary programs.   Although numerous approaches have been  and
are  being tried,  no vigorous studies of relative  effectiveness
have been conducted.   Thus,  as shown in Figure 4,  we have only
limited  knowledge  of  how  best and most  cost  effectively  to
achieve  BMP  adoption by landowners.   More  knowledge  will  be
gained over the next few years from practical experience and"from
analysis  of the social,  educational and economic factors  which
have produced success in RCWP and other programs.

     As  with  all public financial investments,  there exists  a
legitimate  concern  that  the returns be  measured  and  weighed
against  costs.    Thus,   the  feasibility  and  sensitivity  of
monitoring  programs to document water quality response  to  land
treatment has become an increasingly important issue in the field
of  NPS control.   Such documentation has proven to be  extremely
difficult  due   to  the natural variability of  aquatic  systems
affected by NPS and also because the nature of NPS water resource
impairments is often subtle,  chronic,  and transient.  As illus-
trated in Figure 5 determining which water resource responses can
'03 best detected, how much response is needed, and the time frame
required to document water quality benefits,  may be the greatest
achi.*<-eiaent  of  the RCWP  CM&B program.   Figure 5  shows  that,
while  other RCWP projects will also preside insight,  the   CM&E
ha^  and will continue to be our major source of  information  in
•• h: . important area.

     The overall perspective that emerges from Figure 1-5 is that
s.S  iO  percent  of our current information on various aspects  of
agricultural NPS control is presently coming from RCWP.   By 1990


                               18
i

-------
RCWP  is  projected  to  contribute  for  40-70  percent  of  the
available information on these various aspects.
                               19

-------
                              YEflR-1985
                       CM&E
                         10%
               UNKNOWN
                    30%
OTHER  RCWP
30%
      PLOT I FIELD
      5%
                                         OTHER PROJECTS
                                         25%
                               YEflR-1990
                                OTHER RCWP
                                    45"
                                                 PLOT
                                                 5%
           & FIELD
                   UNKNOWN
                         5%
                                       OTHER PROJECTS
                                       30 %
figure 1.   Relative information contributions concerning
            treatability  of  water resources  impaired by  NFS.
                                 20

-------
                             YEflR-1985


                                    CM4E
               UNKNOWN
                    35
           OTHER PROJECTS
    OTHER RCWP
    15%
PLOT
30%
I FIELD
                              YEfiR-1990
                           CM&E
                            20%
              UNKNOWN
                   10%
       OTHER PROJECTS
                    15%
                                             OTHER RCWP
                                             20%
                                        PLOT
                                        35%
     FIELD
igure 2.   Relative information contributions  concerning BMP
           effectiveness  and cost,

                                21

-------
                              YEftR-1985
                                           OTHER
                                           30%
RCWP
                           UNKNOWN
                             60%
                              YERR-1990
                             OTHER RCWP
                                 50%
                    CM&E
                      20%
                                                OTHER PROJECTS
                                                10%
                                       UNKNOWN
                                       20%
Figure  3,   Relative information  contributions  concerning NFS
            critical area  identification.

                                 22

-------
                       YEfiR-lS85
                CM&E
                  10%
OTHER RCWP
30%
                                        OTHER PROJECTS
                                        10%
            UNKNOWN
                  50*
                       YEflR-1990
                         OTHER RCWP
                              45%
                                      OTHER PROJECTS
                                      20%
                  UNKNOWN
                    20%
4,  Relative  Information contributions concerning
    effective  organizational methods for Ag NPS  control
    projects.

                        23

-------
                       YEflR-1985
                             CM4E
                              20%
         UNKNOWN
               60*
                                        OTHER RCWP
                                        10%
 PLOT & FIELD
 5 *

OTHER PROJECTS
5%
                       YEflR-1990
            CM4E
               50%
                        UNKNOWN
                          20%
OTHER RCWP
15%
                                        PLOT & FIELD
                                        10%


                                     OTHER PROJECTS
                                     5%
5.  Relative  information  contributions concerning water
    quality monitoring of  Ag  NFS contrc ,  projects.

                        24

-------
             ECONOMIC IMPACTS OF THE  CMfcB PROJECTS

     The evaluation of the economic impacts of the  CM&E projects
is  being conducted by the Economic Research Service of the USDA.
Initial  analyses are summarized here and in subsequent  chapters
of this report.   The details of these analyses are presented  in
the projects' 1984 annual progress reports.   These analyses will
be expanded and refined in  1987.

      Project Costs
     The aggregate effects of RCWP on project areas are discussed
in  this  section.    Practices  implemented  or  planned  to  be
implemented  are ' evaluated in 1984 and in those instances  where
possible effects are projected to 1991 and 2031.

     The  cost  estimates  associated with  the  individual  RCWP
CM&E  project are presented in Table 2.   Costs are divided  into
direct  government  costs,  cost share costs and  private  costs.
Direct  government costs include information and education  costs
and  technical assistance costs.   The cost estimates  associated
with  private costs cannot be compared directly across  projects.
The  cost estimates include the changes in the operations of  the
farms  that are made in response to RCWP for the Highland  Silver
Lake,  Rock Creek and St. Albans Bay projects.  For the two other
projects   only  the  direct  costs  of  BMP  implementation  are
included, presumably farmers in these projects could make changes
in  their  operations  which  would  result  in  lower  costs  of
implementation.
                                    *                   *
     Nevertheless  there  is considerable variation in  the  cost
estimates  associated  with the five  projects  (Table  2).   The
estimated  impacts range from a high of $6.8 million for the Rock
Creek  Project  with  a 50-year planning horizon,   to  a  low  of
negative  $409,000  for the St.  Albans Bay project  when  it  is
projected to a 50-year planning horizon.   In the case of the St.
Albans  Bay  project adoption of manure storage structures  saves
nutrients  which  can be used to  replace  purchased  fertilizer.
This savings significantly reduces the total costs to the farmer.
In  the  long run (over 20 years) the cost savings  from  reduced
purchases  of  fertilizer may exceed the costs of installing  the
manure  storage structure.   Farmers in the Conestoga  Headwaters
project  area  may also benefit from reductions in  purchases  of
fertlli3er   associated  with  installation  of  manure   storage
st.r^f,- > 'sres .    The benefits to farmers in the Conestoga Headwaters
wo'Ud  be considerably lower than those for the  St.  Albans  Bay
project,  because  of excess amounts of nutrients available  from
       >aanures in the Conestoga Headwaters project area.
                               25

-------














*
n
o

o
&
E

*
v«
A
c
o


^H
*o
«
i-

i— t *^
m n
.P O
o u
0)
Jj Jj
28
«« O
£.
a.
.-i a> m
«£•*->

"o u o
H CV U

I


Q>

Sco
JJ"
u to

J %
jj a> n
Cj £ J-*
'« C W
U £- O
x G) O
0 >
o
o



•o
ai
fel
X O
c.
o.



















^>
u
CJ
«^
o
t.
O-









ro ro
\O CM
-"

O t~
CM CM



CTN ON
in sr
*- 0
»
^M.



t- t-

mvo




CM CM
JT CM










sr v-
rfN ^\
ON ON













if.
a.

A
n.
t.

re
3
•o
re
OJ
X

rt3
CO
o
^J
to
0)
c.
o
o

O ON
— O
CM ON


ro cd
in so
,— .sr
CM CO



in co

ON in
M
^B



vO ON
sr co
v£> O
^"



ON ST
sr CM
co in






•


ST r-
ff* ON
ON ON













^ J
>-H

*


•*^
CO

T3
C
ffl
^— (
JC
00
*»H
X
4-
O T-
m co
in JT
-"

CO
CO ^


4-
co *•
vO CO
sr sr
*
^M



CO ~-
~- CO
sr sr
CM* ^



in <

CM 2









CO ON
ON ON











a
,
I \
j %
07
M\
C

o
a.

i

a
0)

re


^^
o
o
3

(0
0

CM CO
sr \f\
sr t—
-•*>

o ^o
r"^\ oo
O m
• *
^- PO

CM CM
«- CM

«i
CO



CM ON
•- co

CM



•< CO
>» CO
SE ON









CO ro
ON O























a
^4

•
J^T
o
OJ

o

Ji^
o
o
cc

^** OO ^^
JT CO O
O e-a-
- ^

in co o
CM ON O
coco O
*+f* ^^ *
CM

CM »- -
t** CTf *
CO vO 0
» » »
^v ^_ «••



a- O O
t- O O
^- sr sr
•" •" *~



CO «- «-
ON CO CO
T- CM CM









CO ON CO
ON ON O
*- •— CM


















H


«
^
(0
CD

w
c
m

«— *
«^

•
^
V5





























'
•
c
o
,H
m

V
A)
W
O.

•^
o.
CD
t<
O

« .
« £
o ,2
u >
O -,
a; 2J
t. >

JM *J
fO
fl) ^
— « ^}
*J C

>, *

c ^
° -M


-------
     The estimated €4>3ts for the Rock Creek project are extremely
high.   The  BMP*  initially installed were primarily end of  the
field  practices.   While  treating the ends of  the  fields  and
trapping  sediment  once  it leaves the field  are  effective  in
limiting  sediment  delivery to the stream,  they are costly  op-
tions.  These types of BMPs take land out of production, reducing
income to the farmer.  The Rock Creek project is being redirected
to  emphasize  sediment control on the field through the  use  of
conservation tillage and no till.   The advantages to these  BMPs
may  include:   land is not taken out of production,  these prac-
tices  are less costly than conventional tillage,  and  the  soil
remains  on  the  field thus providing a  long  run  productivity
benefit.

        Benefits
     The purpose of the experimental Rural Clean Water Program is
to   determine  whether  or  not  agricultural  best   management
practices can be used to improve water quality for  society.   As
part of the economic evaluation of RCWP estimates of the value of
water  quality  improvement  were  developed  for  three  of  the
projects (Highland Silver lake, Rock Creek, and St. Albans Bay).

     When estimating offsite benefits for a program such as RCWP,
several  factors must be considered.   First,  a linkage  between
water  quality  and BMP implementation must be  established.   Is
water  quality likely to change and can the change be  attributed
to RCWP?   Second, what uses of water are being impacted by agri-
cultural  nonpoint source pollution and how will this use  change
with water quality improvement?   Finally, water quality benefits
do  not occur at the same time that BMPs are installed.   A  sig-
nificant  amount of time nay elapse between BMP installation  and
water  quality  improvement.   Once water  quality  improves  the
offsite benefits will c >ntinue over a long time period.

     The  initial  examinations of offsite benefits for the  RCWP
projects were not limited to examination  of measurable  economic
benefits,   rather  identification  of all potential benefits  was
addressed.    After  review  of potential benefits that  could  be
attributed  to  the  various projects,  an attempt  was  made  to
measure benefits in order to conduct benefit/cost evaluations for
each project.

              l5P.rovements
     jfl order to estiaate the offsiie benefits of RCWP  projects,
it is necessary to project the level of water quality improvement
J-.hnt  o*n be attributed to RCWP.    As a result of the Idaho  RCWP
?£•-"> j *s o t .   sediment  delivery  to  Rock  Creek  will  be  reduced.
Monitoring  data  indicated  some improvement in the  quality  of
wa!~«r in Rock Greek in i984.  The amount of reduction in sediment
rie'ivery will depend upon the type and number of BMPs  installed.
In Illinois the water quality impact will be a swal! reduction in
the  turbidity of Highland Silver Lake.   Monitoring and modeling
results indicate that deposition of sediment in the lake will  be
                               27

-------
reduced  slightly.   Since  the  natric soils tend to  remain  in
solution,  the  rate  of  sediment  deposition  in  the  lake  is
significantly lower than originally hypothesized.    Water quality
in St.  Albans Bay in Vermont is projected to improve to a  level
that   would   again   permit  water  contact   recreation   with
installation of the RCWP BMPs and the upgrading of the wastewater
treatment plant.  Water quality in the bay has improved, however,
the  improvement can not yet be tied to control  of  agricultural
nonpoint source pollution with available data.  The water quality
improvement  may be delayed for up to two years until  phosphorus
in the bay's sediments are used up.

!§2§£il Intimates

     Estimates  of  offsite benefits for three of the  CN&E  RCWP
projects  are presented in Table 3.   The estimates range from  a
low  of  $249,000 for the Highland Silver Lake  project  to  $4.9
million for tfee St. Albans Bay project.  The major differences in
the  estimates  relate  to two factors which are present  in  St.
Albans  Bay but are not present in the other two  projects.   The
first  and  most important factor is whether or not  RCWP  has  a
measurable  effect on water quality.   In the case of St.  Albans
Bay,  preliminary estimates indicate that water quality is likely
to  improve in the bay as a result of the  clean up effort.   The
second  major feature present in St.  Albans Bay that is vital to
generating estimates of offsite impacts is that the water quality
impairment has an impact on people.   If the quality of water  in
St.  Albans  Bay  is improved,  recreational activity in the  bay
could increase substantially.

     Offsite  benefits  were  not  estimated  for  the  Conestoga
Headwaters  and  Oakwood  Lakes - Poinsett projects except  in  a
qualitative sense.  Offsite benefits for the Conestoga Headwaters
project  were felt to be very small due to the low level  of  BMP
implementation in the project.   If,  however,  agricultural non-
point   source  discharges  were  controlled  in  the   Conestoga
Headwaters  and  surrounding areas,  substantial  benefits  would
result.   The  Conestoga  River  is part of  the  Chesapeake  Bay
drainage  basin.   The Conestoga Headwaters project area has been
identified  as  a primary contributor  of  agricultural  nonpoint
source pollutants to Chesapeake Bay.

     Potentially,  water  quality improvement in Oakwood lakes
Poinsett could result in substantial offsite benefits.   If  RCWP
..•kij.Tt^s  a  change  in  water  quality  in  the  lakes,   so  that
.-eci optional activity changes by 4 percent, recreational benefits

-------
                        Table 3.

          Offsite Benefits for Three HCWP Projects.
                        Rock Creek,       Highland  Silver     St. Albans
Benefit                   Idaho            Lake,Illinois        Bay,Vermont
                                            -$1,000-
Recreation                   $61?           $ 24               $3,886
Water Storage Facilities       $  0           $  0                N.A.
Property Values               N.A.           N.A.               $1,008
Water Conveyance
 Facilities                   $185            N.A.                N.A.
Water Treatment               N.A.          $225                N.A.
Other                        N.A.           N.A.               $   27
Total                        $596           $249               $4,921

N.A. = Not applicable or negligible


     Different   procedures were used to estimate  the recreational
benefits  for  the three RCWP projects.   A comprehensive study  of
fishing   in   the northwest was used as the basis  for the  benefit
estimates for the Rock Creek,  Idaho project.  Data  from the water
quality monitoring of the RCWP project were  used  to rate  fishing
success.   The   ratings were incorporated into  travel cost demand
functions  for   the  northwest to estimate an   average  value  of
$7.94 for a trip to Rock Creek.   With water quality  improvement
an  additional   7,075 user days per year are projected  for  Rock
Creek.    Recreational  benefits have not yet   been evaluated for
the Snake River.

     Fishermen   who currently use Highland Silver Lake were asked
how  much  they   would  be  willing  to  pay  for  water  quality
improvement.  These estimates of willingness to pay were expanded
to  represent the total population of fishermen  in the  Highland
Silver Lake area.   Total benefits of $24,000 were estimated.

     Survey data were also used to estimate  recreational benefits
for the St. Albans Bay project.  A questionnaire  was administered
to  a   elected   samplo of  recreationists   in  the  northeastern
;ortio,; of Lake  Champ lain.   Data were collected on  age,  household
i.ico*n^   occupafion,  family size, hoae address, local address (if
Hffer , t),   type  of recreation engaged in,  and views regarding
oecreating  at   St.   Albans Bay if water quality  were  improved.
Only individuals  familiar y»i tb St ,  Albans Bay were  interviewed.

     The  data were divided into present and former users of  St.
ulans Bay and separate travel cost models were  estimated.    The
!ata  were  segmented because there is no reason  to  assume  that
                                29

-------
those  who  continws  \f> use the bay and those who  have  stopped
using  it  will  receive the same benefits if  water  quality  is
improved.   Those  who have stopped using the bay have  indicated
that  they  have a different preference function than  those  who
continue using it.   For current users, benefits are estimated at
$106.52 per user,  while for former users benefits are $84.00 per
user.   Aggregate  benefits  for water  quality  improvement  are
estimated  to  be  $464,800  per  year.   Prevention  of  further
deterioration  of  water  quality in the bay is estimated  to  be
worth $118,300 per year.  Total recreation benefits are estimated
to  be  $3.9 Billion with a 50 year  planning  horizon,  a  7.875
percent  discount rate and water quality improvement occurring in
1989.

      Storage Benefits
     Sedimentation of water storage facilities was identified  as
the primary offsite impact for the Rock Creek and Hi^nland Silver
Lake  projects.   Sedimentation  of  storage  pools  for  several
hydroelectric generating plants on the Snake River was attributed
to  sediment losses from the Rock Creek project  area.   Highland
Silver  Lake  is  the  primary source of water for  the  City  of
Highland.    The storage capacity of the lake was estimated to  be
declining rapidly due to sediment delivered from the watershed.

     There  are no measurable benefits that can be attributed  to
either  of these projects for reductions in loss of water storage
capacity.    Reductions  in  sediment delivery to Rock  Creek  and
subsequently  to  the  Snake ^River were 'felt to  have  a  minimal
impact for two reasons.   1)* The first water storage facility of
any  size along the Snake River is approximately 120  miles  down
stream  from the confluence with Rock Creek.   2) Because of  the
hydrologic  features of the Snake River,  reductions in  sediment
delivered to  the river could be partially offset by stream  bank
and  channel erosion.   The potential benefits from a much larger
program along the Snake River were not evaluated.

     The situation in Highland Silver Lake is also  unique.   The
soil  characteristics  in the Highland Silver Lake watershed  are
such  that once soil particles enter solution,  they tend to  not
settle out.   Investigation of the rate of sediment deposition in
Highland Silver Lake revealed that sediment is being deposited in
the  bottom of the lake at a much lower rate than was  originally
hypothesized.
     Property value impacts were estimated for the St. Alb-'uject using a hedonic model.   With a hedonic model,  the value
of cv .loperty is assumed to be a function of the  characteristics
.-ol.ited  to the property,  one of which is access to high quality
•Mter for recreation.   Data on property characteristics and sale
-glues were collected from town records for the study.   Property
values> are estimated to increase by $4,300 per property if  wa^er
quality in St. Albans Bay improves.  Water quality i soro«eineu t in
                               30

-------
St.  Albans  Bay is ««p»cted to increase property values by  over
one-Million dollars.   The property value estimate for St. Albans
Bay  includes recreational benefits in addition to the  aesthetic
impacts  that can be attributed to the water quality improvement.
To avoid double counting of benefits owners of property  adjacent
to  the  bay  were  excluded from the  travel  cost  estimate  of
recreation benefits described previously.

      Conveyance
     The  effects  of  sediment  reduction  on  water  conveyance
facilities  were estimated for the Rock Creek Idaho and  Highland
Silver  Lake Illinois projects.   Irrigation return flow from the
Rock Creek project area is discharged into a system of drains and
lateral  canals.   The Twin Falls Canal Company  which  maintains
this   irrigation  network  incurs  costs  for  sediment  removal
throughout  its canal system.   Implementation of the Rock  Creek
. roject  as  originally proposed would result in a  reduction  in
sediment  deposition  in the irrigation canals  of  approximately
18,000  tons  a year for a cost savings  of  $185,000.   For  the
Highland Silver Lake Illinois project,  benefits are estimated to
be zero, since the sediment from the natric soils tends to remain
in solution and is not being deposited in the lake at a very high
rate.
     The  final  category  of offsite effects  evaluated  is  the
impact on water treatment costs.   Highland Silver Lake serves as
the  source for portable drinking water for the City of Highland.
A reduction in sediment delivered into Highland Silver Lake would
result  in  a  reduction of water treatment costs that  would  be
worth  $225,000,  which  could  be  attributed  to  the  project.
Although  the  magnitudes  of the benefits  were  not  estimated,
reductions in nitrate levels would generate water supply benefits
for the Conestoga Headwater and Oakwood Lakes-Poinsett projects.

       Benefits
     In two of the three projects, RCWP is generating some onsite
economic  benefits  from  preserving soil  productivity  or  from
reductions in farmers' costs of operations which more than offset
their installation costs of RCWP practices.

     In Idaho, planned implementation of conservation tillage and
>ther practices whic': help keep soil in place on the fields  will
reduce  long term soil productivity loss,  and generate  benefits
•-}' iaated at $814,000.  In this case, these productivity benefits
•*r>*  as great as the offsite benefits.

     In  the  Illinois  project,   conservation  tillage  is  the
principal BMP,  but because the soils are deep and fertile,  long
term productivity benefits are negligible.

     In the Vermont project,  the installation of improved animal


                               31

-------
waste  storage f»ciHti*» reduces manure ?»si«i«iiing and  fertilizer
costs  over  time  by  store than the farmers'  initial  share  of
putting in the system.  This negative cost of over $2 million can
be considered an onsite private benefit.   Note that it is  about
40 percent as large as the {public benefits.

                Costs      i
     Benefit/cost  ratios  are a traditional measure of  economic
efficiency for project evaluation.  A benefit/cost ratio compares
the present value of the stream of benefits to the present  value
of the stream of costs for a project.  Benefit/cost ratios can be
computed for three of the f|ive  CM&B RCWP projects.  The relative
magnitude of the benefit/cost ratio can also be discussed for the
two other projects.

     How   do  the  estimated  economic  benefits  in  the  three
RCWP/ CM&E  projects compare with the costs of  implementing  the
projects  to generate the benefits?   The answer to this question
depends on which benefits are compared with which costs.   First,
compare  total  economic  benefits,  including  both  public  and
private,  with  total costs,  including again both government  or
public  and  private.   The Vermont project with  a  benefit/cost
ratio of 1.8 to one,  is the only one of the three in which total
economic benefits exceed total costs (Table 4).  In the other two
projects,  total economic benefits are only one-fourth or less as
large as total costs.

     If  we are interested in just comparing public benefits with
public  costs,  and  include productivity benefits  as  a  public
benefit,  the  result changes slightly.    The Vermont project is
still  the only project with benefits exceeding  costs,  but  its
benefit/cost ratio drops to
1.3 while the ratio for the Idaho and
Illinois  projects improve slightly,  but still remain  low.   If
cost  share  rates  could  be reduced  without  affecting  farmer
participation  in  the Vermont project as the  economic  analysis
indicates, the benefit/cost; ratio including only government costs
would increase.            I
                4           I
     If  we  say that these projects were undertaken  to  improve
water quality and produce offsite benefits, and we are interested
in  how  much we are getting for the government  buck,  we  would
compare  offsite benefits against government costs.   When we  do
this,  the benefit/cost ratio for the Idaho project drops to  0.2
while  the others remain the same.   In ternss of strictly offsite
 -'.onoai.io  benefits,  the Vermont project remains by far 'th*;  aost
 •>nonomi OR! ly efficient of t tie f.hree.

     jfuo  benefit/cost ratios are considerably less than one  for
      «;k Creek and Highland
Silver Lake project,  it is
Silver Lake projects.   Because there
are   limited  potential  benefits associated  with  the  Highland
unlikely that this project could ever
have a benefit/cost ratio greater than one.   In the case of  the
Hock  Cre^k project,  however,  a benefit/cost ratio greater  than
one  could have been obtained with a redirection of the  project.


                               32

-------
if the project wer** f.-sd . »*-cted to emphasise conservation  tillage
and  erosion contrel,   ?he costs of the project would fall  while
benefits would increase.  The primary benefit from redirection  of
the  project  would be  onsite  productivity  enhancement.   Since
total sediment losses would be lower, recreational benefits would
also increase.

     Benefit/cost  ratios  were  not computed for  the  Conestoga
Headwaters and Oakwood  Lakes - Poinsett projects.   However,  the
relative  magnitudes  of  the ratios can be  projected  for   each
project.   For the Conestoga Headwaters project, the benefit/cost
ratio  would  be  very  small since there appears  to  be  limited
benefits associated with the project as it is being, implemented.


Table 4.  Estimated benefits compared with costs for three RCWP
       projects with a  50 ye-r planning horizon.  (PRELIMINARY)
ITEM
 IDAHO
PROJECT
Benefits
    Offsite               .8
    Onsite (Productivity) .8
      Subtotal Public

    Private Benefits

      Total Benefits
 1.6
 1.6
  ILLINOIS
   PROJECT

Million Dollars

     .2
VERMONT
PROJECT
     .2
     .2
  4.9


 "479"

  2.0
Costs
    Government Costsa/
    Private Costs

      Total Costs
    Total Benefits/
    Total Costs

    Public Benefits/
    "«•>••*.-rnment Costs

    •.) i !'site Benefits/
     ' •• "-jrnmers *• Cos ts
3.4
3.3
6.7

1.6
.3
1.9
— — 	 Pe>+ i n — — 	 — —
3 . 9b
— 	
3.9

   2


   5


   2
     .1


     .2


     .2
  1.8


  1.3


  1.3
1 •' I Deludes cost, share payments,  technical assistance, and infor-
   mation and education costs.
  [ncludes costs cf phosphorus wastewater treatment for the  City
   of St.  Albans,  VT
                               33

-------
The  benefit/cost  ratio is likely to exceed one for the  Oakwood
Lakes  - Poinsett  project.   As was previously  indicated,   a  4
percent  change in  recreational activity would be sufficient  to
obtain a benefit/cost ratio of one for this project.
     Several  limitations to the economic evaluations need to  be
pointed out.   These evaluations are preliminary and at best give
only  ball park numbers.    The estimates of load  reductions  are
based on modeling.   Corresponding Achievements in load reduction
have  not  yet  been  measured  in the  projects  to  verify  the
predictions .

     The  RCWP  projects  were  not  selected  on  the  basis  of
anticipated benefit/cost ratios,  but rather to experiment or try
ou* the program in different problem and geographical situations.
Al.aough the Idaho,  Illinois,  and some other RCWP projects  may
have  low benefit/cost ratios,  the information they provide will
be valuable for guiding future programs.

     In addition the RCWP projects are not representative statis-
tically of possible agricultural NPS projects in  general.   Thus
the  benefit/cost results should not be used to generalize  about
the economic efficiency of a future program.
                               34

-------
                       Rock Creek, Idaho

                            RCWP - 3

                          INTRODUCTION
     The  Rock Creek project, located in south central Idaho,  is
45,000 acres with 28,000 acres designated as critical.  There are
about 350 farm units in the area with emphasis on dry beans,  dry
peas,  sugar beets,  corn,  small grains, alfalfa, and livestock.
The reported pollutants are sediment,  phosphorus,  nitrogen, and
bacteria.   Recreation,  fishing,  and aesthetics are impaired in
Rock  Creek.   In addition Rock Creek empties a  disproportionate
load of sediment to the Snake River,

     Due to low annual  rainfall, irrigation is required for crop
production.   Water  is  supplied to crops  primarily  by  furrow
irrigation.   Irrigation  ditches,  which originate from main ca-
nals,  carry water to individual farms and eventually empty  into
Rock  Creek,  which  discharges into the Snake River (Figure  6).
Rock Creek has been reported to have poor water quality.   A 1975
report  by the Idaho Department of Health and Welfare  documented
the  water quality status of Rock Creek and recommended  clean-up
of  both point -and nonpoint sources.   Major sources of  nonpoint
pollution in the area are sediment and associated pollutants from
irrigation return flows.   Animal waste is another contributor to
the NFS problem.


?§£§E§ctives of the Project

     The  water  quality  goals of the Rock  Creek  project  were
redefined  in 1983 to improve the water quality of  the  effluent
from project subbasins rather than Rock Creek itself.   The water
quality  goals are to reduce:   (1) sediment by 70  percent,  (2)
phosphorus by 60 percent,  (3) nitrogen by 40 percent, (4) pesti-
cides by 65 percent,  and fecal coliform by 65 percent.  However,
there is not yet any indication whether these reductions would be
sufficient to reverse the impairment of Rock Creek.

     Several  questions  can  be addressed by  analysis  of  this
pro.jec'

     (!)  Have  any  significant water quality  changes  occurred
          over the past four years of the project?

     (2)  If  there  have been water quality  changes,  are  they
          attributable to changes in land management (i.e. BMPs)?
                               35

-------
        A STREAM  STATIONS
        • SUB8ASIN STATIONS
             scale
           1     ?     7    «
VS-WMM*.!.,',**-"*- *
 Fig..ti.; 6.  Map of the Rock Creek Rural Clean Water Program Study Area,
           Twin Falls County, Idaho,  (page 9 of the 1984 Idaho
           Annual Report, DOE)
                                       36

-------
     (3)  How  effective are sediment retention basins  (BMP  12)
          and  irrigation  management  systems (13)  at  reducing
          sediment    loading/concentrations    from    irrigated
          cropland?

The first question refers to changes in the identified pollutants
at either the subbasin level or in Rock Creek.  The project water
quality analyses as well as our own analyses will be discussed in
this report.  The last two questions,  which relate water quality
changes  to BMP implementation, are also monumental tasks of  the
project.   At  the present time,  the data are not sufficient  to
address this issue satisfactorily.


LiQd Tre§tment Strategy

     The project area has been divided into 10  subbasins,  which
have been prioritized by need of land treatment.    However,  most
irrigated  cropland is designated as critical and is not  priori-
tized  within  the subbasins.   Animal waste is  not    addressed
explicitly in this critical area scheme.

     The BMPs approved for land treatment include 1,  2,  9,  10,
11,  12,  13,  15,  and  16 (permanent vegetation,  animal  waste
management,  conservation tillage,  stream protection,  permanent
vegetation  on  critical areas,  sediment  retention,  irrigation
management,  and fertilizer and pesticide management, respective-
ly).   These  practices are appropriate for the problems  identi-
fied.

W§i§r. Quality Monitoring Strategy

     Monitoring  stations  have  been established on  Rock  Creek
since 1980 and in 6 of the 10 subbasins since 1981.  The subbasin
stations are located on irrigation ditches;  most  of these ditches
originate  from  the canals (Figure 6).   Some  of  the  subbasin
stations  have been positioned in pairs at the upstream and down-
stream  points  of the ditches within  the  subbasins,  with  the
downstream  stations  representing outlets from the subbasins  to
Rock Creek.

     Grab samples are taken biweekly during the irrigation season
at the Rock Creek stations;  the subbasins are sampled biweekly at
'ho beginning and end of the irrigation season and weekly  during
:"u  middle of the season (mid-May to early August ).  The samples
ifo  usually analyzed for total phosphorus,  dissolved ortho phos-
ph:,'o,   suspended solids,  fecal coliform, Kjeldahl nitrogen, and
< ii •_>. ,;inic  nitrogen.    Flow sneasurer^nt s are also taken  at  each
     e collection visit.
                               37

-------
                 BMP IMPLEMENTATION ACHIEVEMENTS

     Although  the subbasins have been prioritized,  the implemen-
tation   of  BMPs does not correspond well  with the priority  of
the subbasins, with the exception that subbasin 7 has the highest
reported  implementation and first priority.    Contracts for BMPs
have  been  established  for 62 percent  of  the  critical  area.
Approximately  36 percent of this area is receiving erosion  con-
trol benefits from BMPs, according to the 1984 Annual Report.

     The practices that have been installed as of September  1984
are  mostly  permanent  vegetation on  critical  areas,  sediment
retention,  and irrigation and water management (BMPs 11, 12, and
13).   There is little implementation of practices which are more
effective  at keeping soil and nutrients on the land.   Only  one
animal waste system (BMP-2) and 2! ' acres of conservation tillage
(BMP-9) have been implemented.  No implementation of BMPs 15  and
16,  fertilizer  and  pesticide management,  have been  reported.
Conservation tillage and animal waste management should have more
emphasis in order to address the water quality problems.

     It is difficult to determine the actual amount of  installed
practices for a given subbasin.  The 1984 Idaho DOE Annual report
included  the total acres benefited by all erosion control  prac-
tices in each subbasin; however, these numbers were determined by
adding  the benefits of all BMPs in the subbasin,  some of  which
have  overlapping benefits.   The Idaho Annual Report  (Executive
Summary) details the amount of land treated by components of BMPs
for each subbasin, but upon   close inspection the number of BMPs
implemented  was found to be inconsistent.  This problem has been
pointed  out  to the project and will be corrected  in  the  1985
report.

                 ANALYSIS OF FARM LEVEL COSTS

     Impacts  of RCWP on typical fields in the Rock Creek project
are  shown in Table 5.   The farm has 79 acres of cropland  in  7
fields  ranging in size from 6 to 37 acres and with slopes of 0.5
to  3.5 percent.   Five of the fields had dirt  ditch  irrigation
before  RCWP,  the  other two had concrete ditch or  gated  pipe.
None  had any sediment control practices.   Erosion rates  ranged
from 2 to 20 tons per acre per year.

     Umier RCWP,  improved irrigation was implemented on 6 of the
"  fields.  Dirt ditch irrigation systems were changed to concrete
or  gated pipe,  frequently in cor.iunction with irrigation  water
•aa««tnj8m«nt.  Sediment basins wore installed on 8 of the 7 fields.
fho  improved   irrigation practices reduced erosion rates  by  60
percent  on most fields.   Instal1-tion costs of the BMPs totaled
just uv-^r $13,000.

     \ssjuming the farmer maintains the improved practices over 50
years,  he  has a stream of increased costs partially  offset  by
cosh-share  payments  received and a stream of  improved  returns


                               38

-------
Jable 5.  Impacts of RCWP On a Typical Farm In the Rock Creek, Idaho RCWP
          Project Area.

Item
Acres
Slope (percent)
Irrigation practices;!/
Before
After
Sediment practices: 2/
Before
After
Erosion rate (tons/acre/
year)
Before
After
Installation costs ($000)
Irrigation practices
Sediment practices
5Q-year changes in costs
(discounted $000)
Irrigation system costs
Sediment practice costs
Less cost-share
payments
Total change in
costs
50-year changes in returns
(discounted $000)
Reduction in yield loss
50-year net change in re-
turns before income taxes
(di scour-.; -I $000)

1
37
3

3
106

0
0


20
12

5
0


9
0

^

6


3


(3
i... Cr' 'Cation system codes:
" « dirt ditch
> concrete ditch
'. ." * concrete ditch vith
£ » gated pipe

: 2 :
7
.5 3.5

3
6

0
6


20
18

.3 0
0


.5 0 '
1.0

-J- 0

.8 1.0


.8 0.2


.0) (O.S)



irrigation


3
6
1.5

3
104

0
6


7
4

1.3
0.4


2.3
0.9

-0.9

2.3


0.3


(2.0)



water

Field
: 4
6
0.5

3
3

0
6


2
2

0
0


0
0.9

0

0.9


0


(0.9)




: 5
8
1.0

3
104

0
6


2
1

5.0
0


8.9
1.2

-2.5

7.6


0.2


(7.4)




: 6
8
1.5

6
106

0
6


7
4

1.2
0

•
•2.2
1.2

-0.6

2.8


0.4


(2.4)



:
: 7 :
7
0.5

4
104

0
6


2
1

0
0


0
0

0

0


0.2


0.2



Total
farm
79












12. S
0.4


22.9
5.2

-6.7

21.4


5.1


(16.3)



manage me nt





106 •» gated pipe with irrigation water management
I. ^ef'tv^ne practice- codes:
0 - no practices
$ » sediment basins





















                                         39

-------
from  reductions in yield loss.   For all fields except one,  the
discounted  increased  costs  exceed  the  discounted   increased
returns,  resulting  in  a net loss before income taxes  of  some
$16,000  in  current dollars,  or just over $200 per acre on  the
average.

     Similar  projections were made for all 101  participants  in
the  RCWP  project  through  September  1983.    The  results  are
summarized  im  Figure 7.   Only about 40 percent of the  farmers
will gain economically or break even from their participation  in
RCWP.   The  other 60 percent will lose from a few dollars up  to
over  $200 an acre.   Why are they participating  then?   Several
reasons nay exist:

     1.   Tax   savings   may  make  participation   economically
          beneficial.

     2.   The budgets of costs and returns for the practices  may
          be  in  error or may inadequately consider the  reduced
          management   needed  under  the   improved   irrigation
          practices.  (Budgets will be further reviewed.)

     3.   The   farmer   places  a  high  noneconomic  value   on
          preserving his farm for future generations or on having
          an improved irrigation system.

     4.   The  farmer  does not realize the  economic  impact  of
          practices on his farm's future net returns.

     An  implication  for the project is that those  farmers'  not
gaining  or  breaking even from their participation will tend  to
drop the practices after the RCWP contract expires.

     Federal income tax regulations impact farmers' decisions  to
adopt  BMPs.   Cost-share  payments  are not  considered  taxable
income  and  thus  do not add to  a  participating  farmer's  tax
liability.   Total  costs  of annual practices can be counted  as
deductions.   Also 10 percent of the investment in BNPs involving
structures  with a life span of several years can be deducted  as
an  investment  tax credit from the amount of  taxes  owed.   The
balance of the capital investment can be depreciated,  permitting
an  additional  deduction  from  taxable  income.   In  the  best
circumstances,  particularly  for larger farms facing higher  tax
rates,  the  tax savings from BMP implementation may approach  or
iven  ixceed the noncost-shared part of BMP costs.   Thus,  it  is
conceivable  that  some  farms (in addition to those  where  cost
reduction and yield savings exceed the total of other costs)  may
actually make money from RCWP participation.  To be asore specific
than  this would require looking at individual tax circumstances,
which is beyond the scope of the analysis.
                               40

-------
                                             o
                                             T-l
                                                          u
                                                          01
                                                         *-»
                                                          o
                                                          M
                                                          a
                                                          o
                                                          o
                                                         en
                                                   fe   I
                                                          u
                                                          CO


                                                          Wi
                                                          V
                                                          o.
                                                          01
                                                          c
                                                          01
                                                          0)
                                                          u


                                                          60
                                               O
41

-------
                           QUALITY DATA  ANALYSIS
Summary  of Project  Results

      Included   in   the project 1984  Idaho DOE Annual Report   were
results   of  some   statistical  analyses  of  the  water   quality
monitoring data,  but data  for 1984  were not analyzed.    All  the
data  were transformed to  approach a  normal distribution by taking
their logarithms before other analytical techniques were  applied.
Table 6 summarizes the analyses for the Rock Creek and   subbasin
monitoring data from the  1984 Annual Report.

Table 6.   Summary of Idaho  DOE Data  Analysis for  Rock Creek  RCWP.
      Analysis
     Sampling
Stations Analyzed
       Explanation
Adjusted downstream
  concentration (D*)
     Subbasins
Normalized Mass
 export  loading
     Subbasins
    and Rock Creek
Upstream variability accounted
for  by  using  regression  of
upstream    vs.     downstream
concentration pairs to adjust
the downstream  values.   Per-
cent change from 1981-83 be-
tween the yearly averages re-
ported.
Normalized
    Load
MEL
      Or
                                                        Qi
                                         Where:
                                              MEL = annual Mass Export
                                                       Loading
                                              Qi = mean flow for year i
                                              QT= mean flow for all
                                                        years
                                         Reported change from  1981-83
                                         for the subbasins' downstream
                                         stations and from 1980-83 for
                                         Rock Cro.ek  stations.
        logarithmic
       concentration
      Rock Creek     Reported change  from  1980-
                    83, not adjusted for  upstream
                    or yearly variability.
                                  42

-------
      The  Idaho  DOE examined 6 pairs of upstream   and  downstream
subbasin  stations,  with the upstream stations representing  in-
coming  sources   to  the subbasins and  the  downstream  stations
representing   the outlets of the  subbasins.    Adjusted downstream
concentrations,    D*   ,    were  computed   for  each  subbasin  by
regression between paired upstream and downstream   concentrations
for    1981  to  1984   data.   This  procedure  adjusted  for  the
variability  of  incoming concentrations over  time.    Assumptions
associated  with  using  this procedure are (1)  the  samples  are
properly paired  with the up-  and downstream samples representing
the   same  parcel  of  water and (2) that  the  slope  relationship
between  up- and downstream concentrations is the  same  for   each
year.

   Results  from the Idaho DOE analysis of D* are  shown in   Table
7.   Significant  decreases of sediment D*  values from 1981 to 1983
were  reported for 5 out of 6 stations;   3 stations for  Kjeldahl
nitrogen;   2  stations  for total phosphorus,  and only 1 station
for   both dissolved orthophosphate and fecal  coliform  bacteria.
Inorganic  nitrogen  showed significant increases  in  both   down-
stream stations  of subbasin 4 and no significant decreases  in any
of the other subbasin  stations.
 Table 7.  Calculated percent difference froi 1981 to 1983  by two lethods:  norialized lass export loading (MEL)
        and adjusted dowistreai concentrations (D*) for dowistreai subbasin stations.  (Coipiled froi pages
        40 and 48 of the 1984 Idaho Annual Report, DOE.)
            »

                                                         Fecal        Inorganic
                                          _Qrthop.ho§fihate__  Qo.lifp.ri  IjeH 	N___
                                          8!  Norialized.HEL    B*      D*     D*
                                         —percent—
Subbasin  ..Suspended Sediient
Station    D*" No
                 . Tot§l Phosphorus.
                 IT"
 7-4

 5-2

 4-2

 4-3

 2-2
-51**2

-42**

-58**

-65**

-60

 -7
-73
-75
-46
-75
-59
+8
-29
-18
-15
-53**
-48**
+3
-28
-28
+14
-59
-38
+18
                                          +22

                                          +13

                                          +36**

                                          -42**

                                           +6

                                          -23
+62
-11
+38
-20
+22
-54
+12
+32
-15
-70**
-38
-20
-62**
-36
-21
-55**
-49**
-18
-45
+17
+292**
+131
+39
-19
 i ?*o statistical analyses of significance level was executed on norialized HEL
 2 Trsnds are indicated as folitms:

       +   - Increase
          = Decrease
       *   : Significant at p (0.05
       **  = Significant at p (0,01
                                 43

-------
     The  1984  Annual Report also presented flow  weighted  nor-
malized  loads  and a percent change in the  normalized  loadings
which  were  calculated  for both the Rock Creek  and  downstream
subbasin stations.   These results for the subbasins are shown in
Table 7.  The normalized load was obtained by dividing the annual
load (called MEL,  Mass Export Loading) by the mean flow of  that
year  and  multiplied  by the overall mean flow  for  all  years.
Loadings  were  calculated for suspended  sediment,  total  phos-
phorus,  and  dissolved  orthophosphate.    The percent change  in
normalized loadings representing the change over time compared to
the first year's load was computed as:

percent change in    Normalized Load* - Normalized  Loadissa *100
normalized load  =             Normalized Loadx

      Where:
             A = 1980 for Hock Creek Stations;
          or A = 1981 for subbasins stations.
The  changes in normalized loadings were not tested  for  statis-
tical significance.    On the subbasin level, the percent changes
of normalized loadings were similar to D*, with decreases for the
same 5 out of 6 stations for suspend sediment.   The results were
not as consistent,  however,   for total phosphorus and dissolved
orthophosphate.

     The   difference   between  the  annual   logarithmic   mean
concentrations (ALMC) between 1980 and 1983,  at stations on Rock
Creek, are  shown  in • Table  8.    Suspended   sediment,   total
phosphorus,  and  dissolved  orthophosphate  concentrations,  not
adjusted  for  variability  of upstream inputs nor  chanres  over
time,  showed  both  positive changes (representing increases  in
concentrations  over  time) and  negative  changes  (representing
decreases  in  concentrations  over  time).   ALMC  of  suspended
sediment,  total  phosphorus,  and dissolved orthophosphate  were
found to decrease significantly for station S-l, which is located
at   the  base of the project area  on  Rock  Creek.  Significant
changes  were not reported in ALMC over time for total phosphorus
and  dissolved orthophosphate for the other Rock Creek  stations.
Significant  ALMC decreases between the two years  for  suspended
sediment were also reported for stations S-3 and S-6.  S-6 is the
upstream station not affected by RCWP.   Therefore, the change in
S-l •;*<- oot be attributed definitively to RCWP.   The changes  in
 i.^-nnlieed loads for Rock Creek stations were not consistent with
 he  ALMC changes for any of the tested parameters.  In  summary,
th.~r--?  appears  to  have been a consistent decline  in  suspended
i«.j«u!  concentrations end a less consistent decline in sediment
loadings from Rock Creek and the subbasin stations.    1>3 analy-
sis did slot show consistently significant decre-ises in parameters
        sediment at other stations.
                               44

-------
 Table  8.  Calculated  percent  differences fro§ 1980 to 1983 by tito iethods:  norialized lass export loadings
         (HEL) and annual loganthiic lean concentration (ALHC) for the Rock Creek stations.  (Coipiled froi
         page 41 and 49 of the 1984 Idaho Annual Report, DOE.)


 Rock                                                          Fecal          Inorganic
 Creek    __StfSj>ended_Sedi»enJ_  	Tota]_Phosphorus    .Jrthoghosphate        Colifori   Kjel-N   _ N
 Station  Horialized.HEL  ALHC   HortaIized.HEL ALHC   Horialized~M ALHC     ALKG~~ ~   ALHC   ~ALHC"
        	 percent 	

 S-l          -502   -40**    *      +24  -10*         -26   -25**     +6     -48**    +10*

 S-3          -63    -67**          -23  -15          -10    -9       -63**    -53**    +18

 S-4          +55    -13          +143  +11          -17    +3       -72**    -56**    +41*

 S-6            0    -43**         +130  -12          +166   +20       -9     -77**    -82**
 1 Ho statistical analysis of significance level Has executed on normalized HEL
 2 Trends are indicated as follows:

       +   : Increase
           = Decrease
       *   : Significant at p (0.05
       **   = Significant at p (0.01
          Analyses and Interpretations
      We   performed   further   analyses on  project   data  for   the
subbasin   monitoring stations.    We included  the  1984   monitoring
data  which  were   omitted in previous   analyses.   We   also   used
different  analyses   to  determine  if trends could  be   established
from this expanded  database.     Table 9  describes  three groups of
analyses   that  we performed.    The first two  groups  used  paired
data   of  upstream   and  downstream  concentrations,    with   the
assumption  that the same parcel of water is  represented  in   the
ups*.r^-*ii   sample  as in  the  downstream sample of  the   pair.    The
thi- '   31 oup   of  tests   employed  unpaired  data   to   avoid   this
        * ^n.    Thorough details  of  thr-se  analyses  are given in the
        <* nt to  this  report,    "'J escribed below  ar?  the results   and
        > t- ions  ;>rom the additions!
                                     45

-------
Table 9.   Summary of  Further Analyses of Rock Creek  RCWP  Subbasin
             Data,  1981-84.
Analysis
       Analysis of Variance  (ANOVA)
     of downstream data over 4 years
     with upstream  sediaient concen-
     tration as a regres».i
-------
     The  subbasin suspended sediment data we're selected for  our
first  two  analyses since  water  quality changes would be  more
evident  on  the subbasin level than in Rock  Creek  and  because
sediment appears to be one of the more important pollutants being
addressed  by the RCWP project implementation.   The project area
has been divided into 10 subbasins;  however, these subbasins may
not be hydrologically unique, with water from one subbasin possi-
bly  passing into another from farm irrigation  systems  (Martin,
personal communication).  This complex hydrology of the area adds
to  the  difficulty of documenting water quality changes and  the
relationship between these changes and land management.   Due  to
the  uncertainty of separation between hydrologic units,  we  did
not  use loadings;  our analyses deal instead with the concentra-
tion data only,

     Seven  pairs  of stations located  upstream  and  downstream
within  subbasins  were chosen;  six of these are the same  pairs
described in the 1984 Idaho RCWP report along with an  additional
pair  (7-1  and  7-3) within subbasin  7 to avoid  the  input  of
another  ditch at site 7-6,   which enters below 7-3 but above 7-4
(Figure 6).

     These upstream and downstream suspended sediment  concentra-
tion  data  were tested for normality with the Shapiro-Wilk  test
(when n < 50) or the Kolomogorov-D statistic (when n > 50).  Most
of  the  data were found not to be normal.   The data  were  then
transformed  by adding 1 to each observation and taking the  log.
All of the transformed  data were feund to be normal,  except the
upstream  concentrations  and  positive  differences  (downstream
minus  upstream  concentrations) of subbasin 2 and  the  upstream
concentrations  of station 4-1.   For these- except ions,  however,
the  skewness and kurtosis were both greatly reduced by  the  log
transformations  (i.e.  kurtosis  reduced  from 52.9 to  4.8  for
station 2-1), and thus were much closer to a normal distribution.


An§!Y.§i§  of  Y§li§0ce  of  Downstream  Sediment   Concentrations
Between  Years  1981  to 1984 Using Upstream Concentration  as  a
H§^r§5§ioO  Qovariate^     Linear regression of sediment  concen-
trations  of upstream vs.   downstream stations were found  to  be
significant   (P<:.05) except  in subbasin 2 (Table 10.) This  rela-
tionship  suggests  that  the  downstream  concentration  can  be
adjusted for changes in upstream concentration.

     V set of analyses of variance (.ANOVA)  were performed on the
data  :~ o  test differences in downstream  concentrations  between
/ear;,  using  upstream  concentrations ^u a  covariate  for  each
j'J: '• --tri i n .   Two char act er i s i i es of these tests '7ere examined:   the
midpoints and slopes of the  annual regression lines.   The  rela-
tionship  between  the slopes  indicates  possible  changes  with
tiwa.   The   midpoint  represents the average of  the  downstream
•oncen i, rat i ons  for  a given ye'ir adjusted for  upstream  values.
This  analysis is similar  to ths parameter D* that was  presented
in <~he 1984  annual report,   Out  analysis-  however,   includes the

-------
Table 10. Analyses of Variance of downstream logarithmic suspended
          sediment  concentrations,   multiple  comparisons  among
          years  1981  to 1984 using upstream concentration  as  a
          regression covariate.  (Idaho RCWP)
Subbasin
1-1, 1-2
2-1,
4-1,
4-4,
5-1,
7-1,
7-1,
2-2
4-2
4-3
5-2
7-3
7-4
Significance of Among Years
Regression Be— — Test of Equal i
tween Downstream of
and Upstream " Midpoints1 S
-i- NS
NS *
+ *
+ *
+ *
+ *
+ *
4. ,T 	
ty -
lopes
NS
NS
*
*
*
NS
NS
1  = Test  for equality of midpoints for each year were
     performed using the appropriate model (i.e. common
     or unique slopes for each year).
NS = no significant differences                       •
*  = one or more significant differences exist
     between slopes or midpoints, with each regression
     line representing one irrigation season (P<.05).
+  = significant regression relationship between downstream
          and upstream concentrations exist (P<.05).
1984 data and the additional statistical tests for equality.  The
homogeneity   of  the slopes and midpoints for each of  the  four
years  in each subbasin were tested.   Results are shown in Table
10.  Subbasin  5  and both pairs of stations in subbasin  4  were
found to have significantly different slopes,  indicating that at
least  one of the four years had a slope that was different  from
M.i?  others.   The  other subbasins shewed no evidence  that  the
sK-.pes were significantly different between years.  Significantly
different  midpoints were also found for all  of the regressions,
except subbasin 1.  The difference of the midpoints signifies  at
leasf  one year had a different adjusted downstream Concentration
than  the  other  years for  that  particular  suhba.sin.    It  is
^ » >^jrest ing  to note that the downstream sediment  concentrations
if.  subbasin  7 did not show a significant decrease over time  if
chey   were  not  adjusted  for  upstream   concentration.    The
downstream vs.  the upstream sediment concentration  relationship
for  each year for subbasin pair (7-1,7-4) is sho^n in Figure  8.
                               48

-------
CO
oo
O)



D
D
D
0
• o
|

i- C

                                                                                                            c  s-
                                                                                                            o  u
                                                                                                           •i—  4)
                                                                                   -4->  C
                                                                                   c  re
                                                                                   QJ  U
                                                                                   U •!-
                                                                                   C <4-
                                                                                   o ••-
                                                                                   (_>  C
                                                                                   •a
                                                                                   ai  .  «

                                                                                      CO
                                                                                   •o cri
                                                                                   & "* x-.

                                                                                                            E -o  -S
                                                                                                            "5 O  .JL
                                                                                                            Oi ••-  "
                                                                                                            i- S.
                                                                                                            ^ C>   •
                                                                                                            i/> a. c
                                                                                                            a.    a)
                                                                                                            ID «4-  a>
                                                                                                               o  i/^

                                                                                                            i/i i_  on
                                                                                                            > (O  ••-

                                                                                                            E3U
                                                                                                            (O    LK
                                                                                                            QJ _C  t)
                                                                                                            t- O  CL)

                                                                                                            U1 QJ
                                                                                                            C    i-
                                                                                                            3 s_  a.'
                                                                                                            O O  >
                                                                                                            Q M-  O
    o
    o
    ••5
O
o
       CO

        OJ
                           (j/Buj) -QN03 1N2WG3S WV3yiSNMOQ
                                                         49

-------
The  supplement to thi» report has th« statistically  significant
relationships shown for all the subbasin pairs in similar fashion
to Figure 8.

     The conclusions from these tests are:

     1.  Downstream sediment  concentration is  significantly re-
         lated  to   upstream  concentration  for  the  subbasins
         tested,  except  that of subbasin 2.    (test of signifi-
         cance of regression)

     2.  Subbasin 1 has  the only pair of stations for which  the
         downstream sediment concentrations adjusted for upstream
         values were sot significantly different between  any  of
         the  years.   At least one year had a different adjusted
         downstream concentration for each of the other pairs  of
         stations,  (test of equality of midpoints)

     3.  Differences  among  slopes between years were found  for
         the regression of sediment concentrations  from subbasin
         5 and station pair (4-1,  4-2). (test of homogeneity  of
         slopes)
5§SE§§§i2S fif E§ir§§ B§t§i   P.ottR§tream Minus Upstream Concentra-
tions  vs...  Time.   In this analysis we  adjusted the  downstream
sediment  concentration by subtracting the upstream concentration
from the downstream concentration for each pair of samples within
each subbasin.  This   resulted  in  a  difference concentration,
which represents the portion contributed by the area between  the
two  sampling stations.  (Difference = Downstream Concentration -
Upstream  Concentration.)  The log of the difference   concentra-
tion was taken to reduce the skewness' and kurtosis. (It should be
noted  that this is not the same as using the difference  of  the
logs  of  down- and upstream concentrations which represents  the
ratio  of the two values).   Logarithms of the differences  could
not be taken when upstream concentration exceeded downstream con-
centration due to its negative value.

     The variations of the upstream,  downstream,  and difference
concentrations  over time for  subbasin 2 are shown in  Figure 9a
and b.   Similar  graphs for the other subbasins are presented in
the  supplement to this report.   The upstream suspended sediment
concentration appears to have an annual pattern  with  two  large
peaks:   one  at the beginning of the irrigation season  and  the
   influenced  by irrigation return flow from  other  drainages.
\ldo,  the  initial flush of the canals at the beginning  of  the
 jc'is-    may  contain sediment laden waters.   The peaks could  be
  .iK'od  to  land  use  practices,   because    many  fields  are
it-i '. ga^ed  just prior to planting to  enhance  germination.   The
                               50

-------
     10000 -
     iiroo -
     100  -
     10
                              SITES 2-1,2-2
                                                         (9a)
      1200 H
       800-
       100-
                                                        (9b)
      -400-
                                   WTE
Figure 9  (a)  Upstream  and  Downstream Suspended Solids Concentra-
              tions  and   (b)  Differeice   Between  Downstream   an!
              Upstream  Sediaent  Concentrations for  Subbasin  ?,
              1981 to 1984. (Idaho RCWP)
                                 51

-------
peak  in  August may coincide with the harvesting of many of  the
fields (Neubeizer,  personal communication).   Close inspection of
land  use activity during these peak concentration periods  could
reveal  useful  information  relative  to  Management  for  water
quality.

     There are periods when the upstream concentration is greater
than  the  downstream  concentration,  resulting  in  a  negative
difference  concentration.  The actual budget is  not  available,
because of unknown sources and losses of water.    Several factors
could  be  contributing to this phenomenon.   For  instance,  the
ditch ' could  be  at or near carrying capacity  at  the  upstream
station  so that a decrease in velocity would  cause  deposition.
Alternatively,  water  with less sediment could be  entering  the
system between the upstreaa and downstream stations.   The effect
of  BNPs would not be observed under such circumstances.   It  is
doubtful,   however,   that   this  phenomei.  n  is  due  to  BMPs
themselves  because  subbasin 1 has more occurances  of  upstream
concentration  exceeding  downstream concentration but has a  low
level of BMP implementation.   With this in mind, we examined the
case  where downstream concentrations were greater than  upstream
concentrations; this set of data we  called the "truncated"  data
because the negative differences were excluded.

     Regression analysis was used to test the relationship of the
difference concentrations against time in years.   Time in  years
was  used as a continuous variable for the four years rather than
a discrete variable for it is a stronger test statistically.

     The  truncated  difference concentrations were found  to  be
significantly correlated with time for each subbasin pair, except
subbasin  1.   The slopes of all of these significant regressions
were  negative,  indicating a decrease over v.ime.   None  of  the
data  tested  for subbasin 1 were significantly  correlated  with
time,  even  though  the annual means of the downstream data  de-
creased each year.  The upstream site in subbasin 1 has very high
sediment  concentrations, significantly higher than the other up-
stream sites (P <.001) as shown in Figure 10.  The source of high
incoming sediment concentration in subbasin 1 is presently  unex-
plained.  On-site investigation  would be useful to identify and,
if  possible,  treat  any  apparent  sources near  the  inlet  to
subbasin 1.

     The  mean of the upstream sediment concentrations were plot-
ted for each of the four years for stations  1-1,  2-1,  upstream
stations which feed from the low line canal (4-4,  4-1, and 5-1),
'n>1 '-'-I (Figure 10).   Station 11 had significantly higher means
than the other sites for all 4 years.   Stations 1-1 and 2-1  are
both   spring   fed,   but  have  extremely  different   sediment
con>,. ii trations.  Station 2-1 liad the "owest jpstream mean concen-
trations over the period  yf all upstream stations.  The upstream
stations  which  come from the low line canal  had  significantly
Mgher mean concentrations than station 7-1,  except in 1384 where
the means were not significantly different.
                               52

-------
1000-
 (_>
 o
o
LU
00
 UJ
O-
ZD
 10
  1 >
                                                                              •-O
                            _  _p._

      I — --— -—-  ' "
                           1982
                                        YEflR
*—*•  LOW LINE CflNflL
 >*-*-*  STATION 2-1
                              1983
                                                    <>--*. ,> STflTIQN  1-1
                                                    rs-o-a STP-TTOM  7-1
             10. Annual  Mean Logarithmic Seuiuient  Concentrations for
                 Upstrearr, Stations Over Time, 1981  to  1984.  (Idahr- PCWP)

-------
     In  summary,  there is evidence of a significant decrease in
suspended  sediment concentrations contributed by the  subbasins,
except for subbasin 1.  This test of differences vs. time is more
powerful  than  the previously discussed test  of  downstream  vs
upstream concentrations.  Nonetheless, evidence from  these tests
are  consistent  in  that:    (1) subbasin 1 does not  indicate  a
significant change in sediment concentration over the period, (2)
there appears to be a significant change for at least one year in
sediment  concentration for all other pairs of stations  and  (3)
significant  decreases  in sediment concentrations over time  for
subbasin 5 and the pair (4-1, 4-2) in subbasin 4.
                                                     i

B§gI§§§i2D of yBP.§ir.§
-------
Figure 11.  Diagrams representing the possible scenarios for the
           comparison of upstream and downstream linear slopes
           of concentrations vs. time.
                                                             downstream
                                                        	upstream

Scenario 1:    |
               u
               c
               o
               u
                     time
Scenario 2:
Scenario 3:
                                  55

-------
     The yearly a^a/i* **ith  2 standard deviation limits are shown
for  log  sediment concentrations by subbasin in Figure 12.   The
intervals  are  approximately equal to the 95 percent  confidence
intervals   on  the  geometric  Bean  values.    These   pictures
represent the application of the Figure 11 scenarios to the data.
Similar plots for the other parameters (Fecal coliform,  total-P,
TKN,  Inorganic-N and stream flow) are found in the supplement to
this report.

     The  magnitude  of  potential  improvements  in   downstream
sediment  concentration  can  be seen by examining the  plots  in
Figure 12.   Subbasin 1 has upstream concentrations almost as high
as  downstream concentrations,  which indicates that most of  the
sediment  is  not contributed by the  subbasin.   Improvement  of
downstream concentrations of subbasin 1 is not likely unless  the
incoming  source  is treated.    Downstream station 5-2 and   4-2
have    statistically   significant   decreases    in    sediment
concentration  with  time,  but  now appear  to  have  downstream
concentrations  approaching  the incoming source  concentrations.
Any  further  decrease  in these  2  subbasins'  contribution  of
suspended  sediment  will depend on improving the quality of  the
incoming water source.   However, subbasins which are  drained by
sites  2-2,   4-3,  7-3,  and  7-4  appear to  have  the  greatest
potential for improvement.  The upstream sediment  concentrations
for both upstream sites in subbasin 4 are approximately the same,
but outlet site 4-3 has higher sediment than the outlet site 4-2.
The  BMP implementation defined for the two drainages in subbasin
4  should be directed to emphasize  the sediment control  in  the
basin drained by site 4-3.
         *                                 *
                                f  ,
     Table  11  summarizes  the results of statistical  tests  to
place each of the six water quality parameters  examined in  each
of   the  seven  subbasin  station  p^irs  into  the  appropriate
scenario.   The test revealed that suspended sediment  concentra-
tions  decreased  significantly over time for all stations except
subbasin 1.    In fact, no significant changes over the period for
any  of the parameters were found for subbasin 1.   This  may  be
explained  by  the observation that the water quality  parameters
for subbasin 1 upstream are high.  Only pair (4-4, 4-3) was found
to  have a significant decrease in -fecal coliform  concentrations
over  time;   this corresponds with the reported removal  of  some
cows from the stream in 1983.  Fecal coliform concentrations were
much  higher  at  stations (2-2 and  4-2)  than  expected.   This
suggests  that major sources of fecal coliform msy exist in these
        -s,  and that these should be identified and treated.
     This  example shows clearly that water quality data fr.om the
   i tor ing  program  can provide  indi spensible  information  for
   MiH f ying critical areas and critical sources,
                               56

-------
CM
  I
IT)
t/i
CM
  I
CM
 I
CM
                                                                              i                  •
                                                                                       */*->»
                                                                                                                                                   I/)
                                                                                                                                                   tO
                                                                                                                                                  X5
                                                                                                                                                  X)
                                                                                                              O
                                                                                                             5
                                                                                                              ItJ
                                                                                                                                                   c
                                                                                                                                                   Ol
                                                                                                                                                   O
                                                                                                              c
                                                                                                              0)


                                                                                                             •o
                                                                                                              O)

                                                                                                                 Q.
                                                                                                             -O 3
                                                                                                              
                                                                                                             •a ex
                                                                                                              c
                                                                                                                                                      >c
                                                                                                                                                  3
                                                                                                                                                  »/»
                                                                                                                                                  a>   •
                                                                                                                                                  s-  c
                                                                                                                                                  C  E

                                                                                                                                                  o  01
                                                                                                                                                  (O  3
                                                                                                                                                      O
                                                                                                                                                  s-  i/>
                                                                                                                                                 •(->  C
                                                                                                                                                  wi  o
                                                                                                                                                  Q.-1-
«
ft)  E
i-  10

  i.
C 4->
3C  c^
o  a.
                                                                           CM
«a-


UJ
                                                                                                                                                  o  >
                                                                                                                                                      O)
                                                                                                                                                  (« -O

                                                                                                                                                  to -D
                                                                                                                                                  O)  i-
                                                                                                                                                  u  c
                                                                                                                                                 •r-  rtj
                                                                                                                                                  i- -M
                                                                                                                                                 4J  1/1
                                                                                                                                                  0)
                                                                                                                                                  e  o
                                                                                                                                                  o  2
                                                                                                                                                  a>  i->
                                                                                                                                                  >-, -w
                                                                                                                                                  r—  -  c
                                                                                                                                                      trt
                                                                                                                                                      5-
                                                                                                                                                  s-  s-
                                                                                                                                                  r  o
                                                                                                                                                  en s-
                                                                                                                                                  ••-  s_
                                                                                                                                                  U_ UJ
                                                                        57

-------
Table 11.  Changes in adjusted downstream concentrations over time
          (1981-1984),   represented  by  scenarios  described  in
          Figure 11.  (Idaho RCWP)
Subbasin
1-1,1-2
2-1,2-2
4-1.4-2
4-4,4-3
5-1.5-2
7-1,7-3
7-1,7-4
Suspended
Sediment
1
2
2
2
2
2
2
Fecal
Colif orm
1
1
1
2
1
1
1
Total-P
1
2
1
1
1
2
2
Ortho-P
1
1
3
1
3
1
1
TKN
1
1
1
1
1
1
1
Inorganic
N
1
1
1
1
1
1
1
   Scenarios:
       1= no change over tine
       2= adjusted downstream concentration decreased over time.
       3= adjusted downstream concentration increased over time.
     Significant  decreases  in total phosphorus for  the  period
were  identified  for  subbasin 2 and both pairs in  subbasin  7.
However,  significant increases in dissolved orthophosphate  con-
centration  .over time were found in subbasin 5 and the pair (4-1,
4-2).   Both of the station pairs where orthophosphate concentra-
tions  increased  had no significant change in  total  phosphorus
concentrations  over the same period.   The upstream stations  o-
riginating  from the low line canal (4-1,   4-4,  and 5-1) had the
same  general pattern of orthophosphate concentration  decreasing
              and 1983 and relatively no change between 1983  and
              downstream  orthophosphate  concentration  remained
               the same for stations 4-2 and 5-2,   but  decreased
              between 1982 and 1983 for station  4-3.   This  de-
               also be related to the removal of cows from  above
              however, this decrease was not enough to indicate a
between  1981
1984.    The
approximately
dramatically
crease  could
station 4-3;
significant
t ions
            change between the downstream and upstream concentra-
     No significant changes were noticed in either total Kjeldahl
or  inorganic  nitrogen.   There  was  much  less  nitrogen  data
reported than other parameters.   No data for TKN were  available
and /«ry few inorganic nitrogen data were reported for the second
year;   1982, and the variability of these data is extremely high,
masking any changes that could have occurred.

     In  summary  of this analysis,  (1) sediment  concentrations
were  found  to decrease significantly for all  subbasins  except
                               58

-------
subbasin  1,  (2} significant decreases in total  phosphate  were
identified in 3 of 7 pairs of stations, (3) only one station pair
(4-4,  4-3)  had a significant decrease in fecal coliform concen-
trations, and (4) two pairs (4-1,4-2) and  (5-1,5-2) were found to
have increases in orthophosphate concentrations over the period.
     j5fil§S§Si§ii2D If £ §£liY§D§§§ •  1° order to examine the rela-
tionship  of  water  quality to land treatment,  we  plotted  the
reported  percent  critical  area  with  implementation  of  BMPs
against  sediment  load  for each  downstream   subbasin  station
(Figure 13).   The  suspended sediment load at the  base  of  the
subbasins,  was normalized by adjusting for flow of that year and
for  the  average flow for the whole period  at  that  site.  The
relationships  between percent of critical area treated and water
quality were not the same from one subbasin to another,  ^>ut  for
sediment load vs.  BMP the trends were in' the expected directions
for all stations but 7-3.  The sediment load appeared to decrease
from  1981 to 1982 at site 7-3 and to increase from 1982 to 1983,
even though,  at the same time,  the area with implementation in-
creased.  Likewise, dramatic decreases in sediment loads were not
always  related  to  large increases in  implemented  area,  i.e.
stations 2-2 and 4-3.   Perhaps these areas received  their opti-
mal  effect  from BMPs at  low  implementation  levels,  although
imprecise  reporting of BMP implementation could also reduce  the
effectiveness of this analysis.
.Minimum Detectable Change in Water Quality. Parameters.- Variation
of the sediment concentration data were examined to estimate  the
magnitude  of changes in water quality needed to detect  signifi-
cant differences over time.  Variations in water quality measure-
ment are due to several factors including:

     (1) A  change  in  land  treatment  resulting  in  decreased
         loadings to receiving waters.

     (2) Sampling error

     (3) Analytical error

     (4) Changes in meteorologic and hydrologic conditions.

     (5") Changes in inputs to the system;  for example, changes in
         upstream  concentration can affect the downstream  water
         quality.

     Adjustment  should be ma-it- fo;  items 2-5 if  possible.  When
I...   -fitt'l just ed variability associated with these items is  large,
the  i'.-'juired change in water quality parameters for  statistical
significance is also large.  Several steps are involved in calcu-
lating  the amount of uncontrollable variability in a systnra  and
i.h..'>   the amount of changes in water quality necessary to  detect
significant trends that may be due to changes in land  treatment'.
                               59

-------
                                                        CO

                                                        a-



                                                        f
vi«»-
1
r r~
1 -)

1 •
1 -j






C-J
- o
• CD -*
OJ
0)
i.
: j^
o
i.
o
Oi
C£
|>«>
3
C
C


  • a>

=» ex

a> E

•r- u
•o >*-
re
o ~o
_J  ••-
C- rt3
(D 4->
           •a
           
                                                                          •r- 19

                                                                          •— X)
                                                                          (O
                                                               u o
                                                                           0)
                                                                           s-
                                                                           CD


                                                                          U_
(yv3A/SNOi) 9NIOV01 1N3WIQ3S
                                 60

-------
These steps are:

     (1)  Estimate  the    expected  variability  of  the sampling
          system for  the   unit  of  time that is to  be  compared,
          e.g.  the tine  period corresponding with the irrigation
          season.     This   should  include  a   year   to   year
          variability  component  which  addresses  changes    in
          meteorological   effects.     This  variability  can    be
          expressed as a  variance  which  corresponds to the mean
          square error (MSB)  from  the regression model:

            (a)  log downstream cone.  = A + A (year)
          or
            (b)  log downstream cone.  = A + A (year) + &(log upstream cone.)

          wher • fb  - intercept
                A  - regression coefficient for year
                &  - regression coefficient for concentration

                and (b) is  used if the downstream  concentrations
                are adjusted for upstream concentrations.

     (2)  For   comparison of one year  to  another,  a  multiple
          comparison  statistic  for  a minimum  mean  difference
          between  2 years can be  calculated.  (i.e.    a    least
          significance difference) .

     (3)  A more powerful test,   however,  is to perform a regres-
          sion  analysis.    For example,  the minimum  confidence
          limit  in water quality can be calculated   for  4- and
          10- year  experiments.    The  confidence limit in  this
          case  is  a function  of the standard deviation  of  the
                estimate.
     Figure 14 summarizes  the  average percent decrease  per  year
relative to the initial  yearly geometric mean downstream sediment
concentration  required  to detect  a significant difference over a
2-,  4 - and 10- year nonitoring scheme. (20 samples per year are
assumed.)  The means and ranges indicated in Figure 14 correspond
with  values from each subbasin.    These data indicate that 35 to
55  percent reduction over any given time period is  required  to
detect  a  real change.    The  project's stated  goal,  to  reduce
sediment  concentrations  by   70  percent,    should  therefore be
',;->' -yv5 able  if  accomplished.    Note thai the percent change  re-
quired to compare any 2  years  is very large,  but the change  per
/enr  required  over  a  4 to   10-year  period  using  regression
rtni lysis is much less.   Also,  the  change required per year over a
10  yoai  period  is one-third of  that required during a  4  year
pt :.-•<}    This indicates that  a longer time is required to detect
• \   w-:!.t;i«r  change.   Correction of downstream concen' -ation  with
upstream  concentrations improves  sensitivity as much as 10  per
•:ent   \ more detailed description of this  analysis can be  found
h: vhe supplement to this  report.
                                61

-------
      50 -
      40 -
 i- C
•M O
 Q) •»—
 0)
C 0)
T- O
   C
i- O
(C O
30-
S-  0)
o>  e
O.f-
  •o
01  0)
in  ui
1C
0)  E
1-  ~

20-
      10-
                                           No  adjustment for upstream
                                           sediment  concentration
                                           variation

                                           Adjustments  made for  .
                                           upstream  concentration
                                           variation
                                                                      10
                               Years of Monitoring
           14. The  average percent  decrease per year relative to the initial
               yearly geometric mean  downstream sediment concentration re-
               quired to detect a significant decrease over a 2 year, 4 year,
               and  10 year monitoring scheme.  The range over all subbasins
               is shown.  20  samples  per  year are  assumed.   (Idaho RCWP)

-------
     Similar  percent  reductions to those shown in Figure 14 for
fecal coliform are required to show a significant decrease.   The
reductions required for total-P,  ortho-P,  TKN,  and inorganic-N
are approximately 10 percent less due to  lower variability.
                          PROJECTIONS

     Selection  of  critical subbasins which could  benefit  from
BMPs can be partly performed by examining the water quality data.
Table 12  lists  the concentration differences between  the  mean
upstream and mean downstream values observed in 1981 and 1984 and
their  projections  to 1990 for the 6 water  quality  parameters.
The   mean  downstream concentration used for these 3  years  are
those  predicted by the linear regression of the  log  downstream
values  against the 4 discrete years of 1981 to 1984.   The  mean
upstream  concentrations  orrespond to the geometric mean for the
1981 to 1984 upstream concentrations.   The numbers correspond to
actual concentration differences so the reader can infer  whether
the  difference is practically important.   The 1990 values  were
obtained  by extending the linear trend existing for 1981 through
1984 under the assumption that the present trends will  continue.
In many cases,  this implies the downstream water quality will be
equal to the upstream water quality in 1990.   (the 0.0 values in
Table 12).  There  is  no  certainty that this  same  trend  will
continue,  however.   For each subbasin, Table 12 indicates which
parameters  show  a potential for water quality improvement  from
BMPs.  In general,  subbasins that are drained by sites 2-2, 4-3,
7-3,  and 7-4 deserve continued attention by increased BMP imple-
mentation that could decrease the sediment,  fecal coliform,  and
phosphorus  contribution  to the water.   The data  also  suggest
there is nothing to be gained by increased treatment on the areas
drained by stations 4-2 e -d 5-2, and there is little to be gained
in subbasin 1.
                         IMPLICATIONS

     The Rock Creek project has implemented BMPs approximately on
one-third (36*) of its critical area.   After four years of water
quality monitoring, significant sediment concentration reductions
have been found in 6 of the subbasins.   Additional documentation
of the relationship between land treatment and water quality will
be  helpful  to  establish  a cause-aad- effect of BMPs  and  water
     [r-igated  areas,  like  the Rock Creek project)  appear  to
         faster  to land treatment than do  other  non-irrigated,
•iu ,'i !  areas.    This is probably due to a relatively  low  varia-
bility  in the  nature of t\e irrigated system.  Further comparison
wif-h oMier projects will help to confirm this hypothesis.

     The percent decrease in water quality paraui^; ers required to
detect  .\ real  trend is about 30 to 50 percent. The projects water
quality reduction goals of 70 percent,  60 percent,  40  percent,
a iid 65  percent reduction for sediment,  phosphorus, nitrogi-n, and
*.-v-«1 coliforms, respectively,  would be detectable if obtained.

                                63

-------
BOO
0 OJ 0

•   0
0   44  .
0 T3 0 -
U B • O
44 0 0 O}
a
a>-4
  oo
B en
0-1
0
X B
  •H
0
J31J
44 0

B t.
0 0
0 0
fc.0

S°
A 0
   0
0 9
0 -i
O0
B >
0
t* •
0 0
*« 0
Vi U
•H 44
Q 0
   B
B X
O O
 0 B
 h 0
44 0
 B Z
 0
 O
 B 0

O +>
 0
i—i

 0
0 ~4
eu
  «

44 3
•H O
i-H U
0 JS
9 H
O

h 0
0 9
44 B
0 •*<

   B
(O O
  O

o -o
Vi B
   0
o u
05 H
O)

  00
U CO
O«-l
      0
     44 rl
      U 00
      0 at
 o
 i. 0
 a>j5
      B «N
      a -H

                    "*S

                    TIS!
                      :*,.

                                         g    *    «    -    -
                                         o    2    £    S    S
                                        CM


                                        S
          ^^   ^1   ^9   %W        ••<
           •    •    *    *         •


          0   S   8   £   5   R   S
                •    •••••



               *    *    ^   ^   ^   ^

          0   S   S   °   S   S   S
                *    •          t     •     *


          -   3   i   I   §   i   i
                       g      §    s
                         s   §   s
                       9s< &    ^    ^    -^    f>    M    «»O   9^
                       •Mtf I     C9    ^^    C3    Oi    O    *••   ^^
                     AcS! '     S
                                    i
                                    i    §
          *O
Sc5!f   2
                  K      %
*~*   +^   M»   «^
•-4   9>»   ^   ^
CO   «»   t«   «»



tO   »O   **   M»

R   R   CM   5?
                       —« I     MS    —<    f^.
                       004 I     M5    I-.    i»
                               e«    c^    CM    ft
                               JL,    CM    i    J.
                               —I    «M    «»

                                                                               |

                                                                               8
                                                                                    S
                                                                                    •s
                                                                                    trt
                                                                                    s
                                                                          -  u,-2   i
                                                                          -  g-   -S
                                                     £ £  S    «;
                                                        u —^     •
                                                     ~ g-s   -
                                                                             S S
                                                     - i  -    &
                                                      g-i  s   .s
                                                      o    ^
                                                        gc  **    i-
                                                        *•  "*   ,£

                                                        S.  "®   3
                                                      -*«   1

                                                      Ss?   I
                                                      «^  s
                                                      «?«
                                                        S  ai    ai
                                                      •> e  y>    a!
                                                     jc B  Z>   ^=
                                          64

-------
                 Highland Silver Lake, Illinois

                             RCWP 4

                          INTRODUCTION
     The  Highland Silver Lake project is located in southwestern
Illinois.  Recreational, water supply, and fish and wildlife uses
of  the  lake  are impaired by  high  turbidity  levels.   It  is
reported  that approximately 14 percent of annual water treatment
costs   are  due to the high turbidity of the  water.   With  the
present levels of phosphorus and nitrogen in the lake, these same
uses could he impaired by eutrophic  conditions if the  turbidity
problem were reduced,  allowing sore light to penetrate the waters
and promoting additional plant growth.

     The project area is 30,640 acres,  most of which (82*) is in
row  crops  (i.e.  corn,  soybeans,  and wheat).   Some  of  this
agricultural land is highly erosive,  composed of fine particles,
and  is  believed  to be an important  contributor  of  suspended
sediments  to  the  lake.   Streambank and  channel  erosion  are
nonagricultural NFS of sediment and are being studied to estimate
their  loadings  to the lake.    Livestock operations  within  the
watershed  are sources of nutrients which are also  addressed  by
the project.


Perspectives of the Project"

     The water quality goals of the Highland Silver Lake RCWP are
to:  (1) reduce turbidity and increase visibility to greater than
2  feet and  (2) reduce total  suspended solids concentration-s  to
an average that is less than 25 mg/1.  Goals for nutrient concen-
trations will be determined if a reduction in suspended sediments
yields a eutrophication problem.

     Broad questions that can  be addressed by this project are:

     1.   How much reduction in suspended sediment loading in the
          lake  can  be expected with treatment of 75 percent  of
          the critical area?

     2.   If suspended sediment loadings are reduced,  will there
          be a corresponding Deduction in turbidity of the lake?

     "•    Which  of  the  BMFs are  more  effective  at  reducing
          suspended sediment loadings;  (a)  from the fields,   (b)
          to the streams, and  (c)  tf1 the lake?

     \    If land treatment practices adequately reduce suspended
          sediment  loadings  and  turbidity  levels of the lake,
          will these same practices reduce nutrient loadings,  or
                               65

-------
          will  additional animal waste and fertilizer Management
          practices  be required to avoid eut rophication  of  the
          lake?

     5.    Have  any significant water quality changes occurred at
          the tributary and/or lake levels of the project?

The  answers to these questions and others can give  insights  to
enhance  the use of BMPs in order to obtain maximum benefits  for
given costs.

               Strategy.
     Land   treatment  practices  were  selected  to  reduce  the
detachment and transport of soil  particles, to maximize settling
of  suspended particles,  and to redrce nutrient  losses.   Thus,
BMPs  that  increase the ground cove* ,  decrease the velocity  of
surface  runoff,  and improved the management of livestock  waste
were   approved   for   the   project.    (This   includes   BMPs
1,2, 4, 5, 7, 8, 9, 10, 11, 12, 14, and 15.)

     Of the 30,640 acre watershed, approximately 6,525 acres have
been  designated  as critical.   Criteria for  the  selection  of
critical areas are:

     1.   crop  and  pasture lands composed of natric  soil  with
          slopes  >2% with fine particle size,  and  high  credi-
          bility, and

     2.   brop  and  pasture lands composed of  non-natric  soils
          with  slopes >5X with high erodibi-lity and proximity *o
          water course.

Feedlots  were identified as critical depending upon the distance
to  water  course and number of  animal  units.   These  criteria
appear to be appropriate to address selection needs.

£§£§£ Qualify. Monitoring Strategy

     Highland  Silver  Lake is an impoundment which  has  several
streams  contributing  to  it  and one spillway  as  a  point  of
discharge  (Figure 15).   The monitoring strategy  has  different
components to accommodate the hydrology of the watershed.  First,
the lake is sampled monthly at nine sites; 5 stations are located
in  ! 'he  main  lake  and 4 stations are  locaf^d  in  bay  areas.
^f >.'U1/   the  outflow  of the lake is sampled biweekly  at  the
srull'-uiy,  Three tributary sites witn continuous geging equipment
»i   sampled biweekly;  these sites are also sampled twice a year
for biological monitoring.   Tn addition,  8 field sites, ranging
from 29 to 332 acres in area,  are s-.mpled during runoff  events.
\  livestock  waste management practice is also being  monitored,
Tbo  water samples are analy^d for certain parameters as  speci-
'<«».; in Table 13.
                               66

-------
                                               GENERAL WATERSHED
                                                 MONITORING SITES

                                             Highland Silver Lake Watershed
F2 Field Site

S1 Stream Gage

R3 Raingage
        HIGHLAND
        / SILVER
         LAKE
                                           Stream Cross-section Number
                                  KILOMETERS
                                0       1
                                     MILES
Figure 15. Monitoring  Sites of the Highland Silver  Lake Watershed
           (p.36 from  the Summary Report Fiscal Year  1984).

                                     67

-------

r-
^^
UJ
UJ










>-
5c
^H.
^P
{"J
*





CNJ
0)
o> >-
re — i
0) UJ
> UJ
o 3
L.
ttl
+*
i
re
re
o.
•o
re
r2 U.
a 3
0
JC
u
CO
c:
O
^
 3
p- r-
Z Z

CO _J CO _J
>» ^» 5^» ^—
1— Q. H- 0.
CO Z 4/> Z
co ^*~ c/} NX







V)
P—
re

, i
o.
«t


_ICM
a. z
*• •
S CO
z z

* •
coz
1 1— p-
1
CO CX.
CO CO
p- Q


p-
z
CO
1 H-
CO*
CO
1—






V)
(U
^_
W» CO
a>

•P* &»
CO A
4J

'—• Ja
• o^ *^*
*P" ^m
U. «-






1 1



4J
•p»
C • •
••- -CM 3
P— CO CO p-
re co z z
p— p— . * V)
•*- < • _i cj o
«/> o •>•«-> a. o c
•^- t re f- re
re o. Q. > • • o»
*J -P- CO Z U
OJ CL •-»-»>• i^ O
E E •- o z P-
Sf 3
O. K- O "O » "O.
<: "^ o c co ros:
0 O 01 0 >• 3T 0
»- 5co vji— z —


w»
• CJ>
so. c
»— VO -P-
Z D -0
re
A * Q)
CO Z CM t-
> i^ CO
J- p- Z i-
JC
tfl _J CO O
co p- re • CO
co— a
2C
2Z * **
CO _l
-CO p-
CO h- Q.
u.
** *
• z o
CJ ^ O
U- p- 0










1
1











1






1







fc.
0>
4^
^ 2
u. •»-
CO
0)
> E
*r~ 0^
^ *J
« to
^J >n
ai co
o>
0)
>•




























.
>•>
i.
re
1
u.
c
re
>,
re
c
re
-3
O
«*-
•o
Cl
3
TO
a>
u
v>
4->
O
c

•£

en
c
•*•
?• -
a.
E
re
CO


•"•























>
oo
i-
fO
01
*~
*"•
u
en
T™
™~
t
s_
o
Q.
(U
a:
i.
g
CO
a*
re
s-
OJ
*
•f—
in
•a
c
10
j—
en
-7"
^^
QJ
-C
**
E
V.


C7t
(^O

O.

CVJ
68

-------
                 BMP IMPLEMENTATION ACHIEVEMENTS

     The rate of implements! i on is  lower than anticipated in the
original project plan.  As of September 1984, 2412 acres  (37%) of
the  critical  area  were under contract.   The  amount   o-f  BMPs
actually  implemented in the critical area was not clearly   indi-
cated  in the 1984 Annual Report,  but appears to be considerably
lower  than the implementation goals in the plan  of  work.   The
practices  that are emphasized are waterways,  stream protection,
terraces, diversions, conservation tillage, vegetative cover, and
waste management.    Six animal waste operations out of a  total of
14 are under contract, 3 of which have been installed.

     It  was reported that the BMPs applied thus far have  caused
an estimated sediment reduction of 12,073 tons/year.  There  is no
guarantee,   however,  that  a  reduction  in  suspended   sediment
loadings will yield a reduction in lake turbidity because:

     1.   the  finer size particles will still be transported  to
          the lake,

     2.   the  natric (sodium) nature of the soil keeps particles
          in suspension, and

     3.   turn-over   in  the  lake  may  cause  resuspension  of
          particles that have settled.

These  factors  need to be evaluated with  respect  to  hydraulic
retention  time,  settling rate and frequency of resuspension  to
determine  if  the present strategy of land treatment can  reduce
the  supply  of  fine  particles  sufficiently  to  reduce   lake
turbidity levels.

            §3!-} Education
     The  Illinois RCWP included several methods of I&E to inform
people  in the area and to encourage farmers to use  conservation
practices.   Among these methods were: (1) radio shows, (2) news-
paper articles to 13 regional papers,  (3) a semi-annual TV show,
(4) a bimonthly newsletter devoted to more specialized topics was
sent to farmers and agri-businesses, (5) slide presentations, and
(6)  no-till  demonstration.   The  cost-effectiveness  of  these
ift-.'"iHs was not reported.
                  ANALYSTS OF FARM LEVEL COSTS

     .,'ive  farms  were identified for analytical  purposes  using
 ',(,/, ''-near programming models.  These do not correspond to actual
ope-M*-ing  units although their boundaries do include farsn fields
in  production--some  of  which are  being  monitored  for  water
juaiity data.  Soils included on these farms are representative of
       crop production areas in the project watershed.  For  pur-
      .jf this  summary, only the  results  for one farm (Table 14)
                               69

-------
 o>
.*
 BJ
 V
to

•o

 10
t-t

 00
•H
X

 V
  I-.
  Q
  19
  U
 •H
  a.
  >«
 H

  01
  w
  U
 <
   I
 SO
       I-.
       01
      Vw
I      •
               en
              f4
               at
                       10
                       c •-*
                       
    «   U
    C   (3
   O  2
 «J-
 t-H

  U
 »-(
 A
  a
                                    u

                                     *

                                    10
                                   u-i   •
                                   •-I  O)
                                    (0  OO
                                   «*-  «3
                                   ^1 •-«
                                    «8 *-\
                                       •ft
                                    m  u
                                                                                                                            a
                                                                                                                         *  3
                                                                                                                        9 -O
                                                                                                                        o  at
                                                                                                                       •o  vi
                                                                                                                        a
                                                                                                                        01  n
                                                                                                                        E -H

                                                                                                                        s C
                                     O
                                     en    -
                                         a»
                                     m  eo
                                     ^i  «o
                                         f—i
                                     V)  i-t
                                                                                                                          e
                                                                                                                          w  »-«.
                                                                                                                          o  .*
                                                                                                                          U   M
                                                                                                                                00
                                                                                                                           u  c
                                                                                                                               iH
                                                                                                                            0)  «-
                                                                                                                            >  a.
                                                                                                                            O  ca
                                                  M
                                                  01
                                                                                                                             tn
                                                                                                                             erf
                                                                                                                             (U
                                                                                                                               ^i      *
                                       4)
                                      01
                                      00
                                    •t)
                                      e
                                      M

                                      e
                                      o
                                                   c
                                                   01
                                                   u
                                                   u
                                                   4)
                                                   a.
 o  c     <   e
 |   |   SJ *£   |
to  tn  u 3  X
3  3  I  w  5
in  in tn o  vi
o  u u u  u
                                                 a
                                                 o
                                                                           3
                                                                           C
                                                                                      4)
                                                                                      00
                                                                                      
 M   U
 tn   19
 o  -^
^H   tfl
     c
^   o
•^   M
 O
W!
                                                                                               0)
                                                                                               I-
                                                                                               U
                                                                                               <0
                                                                                                O
                                                                                                U
                                                                                           3
                                                                                           C
                                                                                                  0)
                                                                                       oi
                                                                                              «n
                                                                                              e
                                                                                               U
                                                                                               41
                                                                                               C
                                                                                               0)
                                                                                               CO  01
                                                                                               «  W
                                                                                               h  U
                                                                                               v  
.c  e
 u  o
    u
 u
 o  
-------
and general findings are discussed.

     Profit  maximizing  solutions without cost sharing   included
primarily  row  crops or permanent hayland  grown  under  no-till
resulting  in  estimated  erosion levels far below those  of  the
baseline  or preproject conditions (Table 14).  For most  typical
farms, these solutions did not exceed Illinois' allowable erosion
limits.  These profit maximizing activities involve a greater use
of  no-till and reduced-till systems with increased acreages  de-
voted  to  pasture and hayland.   According'to analyses   of  crop
budgets,  these activities were more profitable in both the short
and long run.

     Availability of ^ost share funds would have little impact on
land  use and treatment. '  Cost sharing would  induce  previously
profit  Bjaxisizing  operators to make only minor  adjustments  in
cropping activities during the ten years of project life.   These
adjustments  would aainly be shifts to no-till from  reduced-till
and  from permanent pasture to rotations including row crops  and
multiple  hay  crops.   These changes found in the first  10  year
period  would revert back in subsequent periods when  cost  share
payments were no longer available.

     The  University of Illinois SOILEC model  simulated  erosion
impacts of a 50 year planning horizon (Eleveld and Starr,  1983).
Average  erosion rates were virtually identical with and  without
cost sharing (a difference of .11 tons/acre).  The reason is that
less profitable activities, which are also slightly more  erosive,
become more profitable with the addition of cost share payments.

     The typical farm analytical models incorporated productivity
impacts  of soil erosion on yields.   The SOILEC model  estimated
crop  yields  by  soil  for crops and BMPs allowed  in  the  farm
analyses.  The  lower  sloping soils of primary concern   in  this
project  are  the most productive soils in  the  watershed.   The
erosion rates for these soils for a CSW/S rotation were estimated
to  be  very  low for  all   management  alternatives  considered.
Related   productivity  impacts  were  also  low  enough  to   be
negligible.

     In  several instances,  the availability of RCWP cost  share
paytneni-s created a different allocation of land use and treatment
>I;.T, ,, -) s&d to that created without such funds (Eleveld and  Starr).
This  .nJicates  a dilution of the impact of  transfer  payments.
That  i.s>  when  a lar.d use and/or SMF is selected as part of  an
"ff->-.ant resource allocation when cost sharing is available  but
ib  .i j t  selected when funds are not available,  the  farm  model
Selected  a  crop management alternative having a lower per  acre
Tot return without cost share payments.   Therefore,  some of the
^ r;>:••; fee payment is being used to compensate for a somewhai,  less
;>(ufi table  crop management system.   The iiepl i- ) ti on is  that de-
sicable  conservation impacts might be obtained with lower  total
tf•••'.lifer (cost share) payments.   This might be possible if opera-
tors could be convinced to  change to profit maximizing .aanagement
activities  suggested  by  analyses wh^n cost  sharing  wa?   not
                               71

-------
available .rather than those suggested when such funds were avail-
able.

     Cost share payments affect net income only during the  first
10 year period in the typical farm analyses.   The average annual
returns  for  the 416 acre farm illustrated in Table 14 in the 10
year  analysis suggests a sizeable impact of $13.28 per acre  per
year for production under cost sharing.

     Terracing and contouring although included in the farm model
alternatives were not selected as part of any solution.  This was
in part due to the expense of these practices and in part due  to
the  relatively  low  levels of erosion inherent in both  of  the
alternative future conditions under the profit maximizing assump-
tion.

     None of the results of these analyses suggests  differential
economic impacts due to farm size.  Identifiable differences seem
more  related  to  the quality of the resource endowment  of  the
typical farms analyzed.


                   WATER QUALITY DATA ANALYSIS

Summary of Project Results

          Extensive analyses of the Illinois RCWP data  performed
by project staff have been reported in the 1984 annual report and
previous  reports.    The  project  should be commended  for  its
efforts  in data analysis and also for its clarity  in  reporting
the results of these analyses.

     For most analyses performed by the project,  the  data  were
stratified by two different factors,  time and flow.   The yearly
data were divided into 3 periods based on agricultural activities
and condition of the land surface:

     period 1 (Pi)  April to June:  fertilizer, seed bed prepara-
                    tion, and  crop establishment

     period 2 (P2)  July to November: reproduction and maturation
                    of crop

     period 3 (P3)  December to May:  residue.

Clio  flow  was classified by bassflow and event  flow   for  data
in ""-.is purposes.

     The  data  analysis in the 1984 annual report  Included  2.25
/V-JA'-S  of small watershed  study data and 3 years of   lake  water
jualily data.   The standard deviations of the concentrations for
stream  baseflow  and  event  flow  are   quite  high,  sometimes
exceeding  the means.   This high variation makes it difficult to
document significant water quality changes over a short period of
t. I..UIP    However,  a few significant changes were reported.  Total
                               72

-------
suspended solids (TSS) and total phosphorus concentration at  the
spillway  were  found to be significantly less than that at  gage
site 1,  located immediately above the lake.   This suggests that
the lake acts as a trap for TSS and phosphorus.   TSS  concentra-
tions  during  events in periods P3 and Pi were generally  higher
than those during P2 ,  this may be related to precipitation  and
land use activities.  No significant differences were found among
stream   stations for concentrations of TSS or turbidity.   Event
concentrations  were found  to be significantly higher than base-
flow concentrations for both these parameters.

     Several  observations  from  the field site  data  indicated
that,  TSS yields from 7 of the 8 fields were composed mostly  of
inorganic  soil  particles,  and phosphorus  concentrations  were
highly  correlated  with  TSS  concentrations  during  PI,   less
correlated during P2,  and least correlated during P3.   Nutrient
concentrations  from-  feedlot  runoff were also  reported  to  be
significantly higher than from croplands.

     The  monthly  loads  measured  at the  stream  and  spillway
stations also had a high amount of variation.  Loadings during P3
were  generally  highest;   probably due to  snowmelt  and  spring
runoff.    Unit area loadings of TSS from stream gage site 3, the
smaller watershed,  were about twice as high as at sites 2, except
during P3-1981.  This disproportionately high unit area load from
stream  site  3  suggests  that  this  subbasin  may  contain  an
important untreated source.

     Two levels of- modeling are planned for this project:   field
scale  and watershed. • -Different land use practices were modeled
on the field scale using CREAMS.   To summarize the results,  (1)
the most effective practice was no-till with 80 percent  residue,
(2)  spring  moldboard  and fall chisel plowing  were  found   to
produce approximately the same change in water quality, (3) grass
waterways  reduced  concentrated flow erosion,  and (4)  sediment
loadings were reduced by dry dams.    CREAMS is also being used to
model two field sites which are also monitored.   The results  of
comparisons  between simulated and observed data for the 2  sites
is  planned for the 1985 Annual Report.   Results from  watershed
modeling,  performed  with  the  AGNPS model,  has not  yet  been
reported.

     Several  regression analyses were performed and reported  in
attempts   to   establish  the  relationship  between   different
var i -I'nl.is .    Land  use (i.e.   type of crop) was not found  to  be
aigni t'iuantly  related  to sedioienl and nutrient yields   on  the
watershed-scale.   However, this comparison  lid not consider other
'^oportant  characteristics  such as  dinnagement  practices,  soil
',ype, slope of field, farm location, etc,
             on  of  nutrient and jodiment  yields  vs,  rainfall
energy  and rainfall/runoff ratios indicated that rainfall energy
ind  the amount of runoff were related to nutrient yields at  the
throe stream sites,  but rainfall energy and runoff were  related
'••o suspended solids at only site 3.
                               73

-------
     To  document  I ; h* relationship bet**^**ii *»ater quality of  the
stream  (site  1) «o4 J«ke,  regression analysis  was  used.   No
significant correlation was found.  However,  only baseflow  data
that  were not stratified into periods (due to small sample size)
were examined.   Perhaps after additional data are  collected,  a
relationship  between stream and lake water quality will be  more
apparent .

     Duncan's  multiple range test was used to compare the  means
of  the lake water quality data.   The variation in the lake data
was  high,  but  not  as high as  the  stream  data.   Generally,
however, Secchi transparency- improved in the summer (P2) probably
due to stratification of the lake.   Approximately 70 percent  of
TSS  is  non-volatile.   TSS is correlated to turbidity,  with  a
better  correlation  found  in  main lake  samples  than  in  bay
samples,   however,   neither  the  significance  level  nor  the
regression  coefficients were reported.   Nutrient levels in  the
lake  were high,  especially during PI.  Chlorophyll a concentra-
tions were low, probably due to limited light penetration.

     The following spatial trends in the lake data were observed,
although they were not significant at 95 percent confidence.   In
a  direction  going  toward  the  dam,  Secchi  transparency  and
dissolved  oxygen  increased and  total  suspended  solids,  non-
volatile solids, and alkalinity decreased.

     Simple  regression of lake water quality parameters  against
time  indicated  significant  increases  in  inorganic  nitrogen,
dissolved  phosphorus,  and  Secchi over the three  ye'ar  period.
However,  upon brief inspection of the reported Secchi  data,  it
appe'ars  that  Secchi transparencies in the. lake were  decreasing
over time.   Further analysis of these data,  with stratification
of data by time period, may clarify this apparent discrepancy.
        Analysis §od isi§r.p.r.etations

     Additional data were presented after the 1984 Annual  Report
in a report from the Illinois State Water Survey (Report No. 357,
February  1985).   We  examined  these data to determine  if  the
differences  in water quality at gage sites 2 and 3  were  indeed
significant.  Although the report covers the period from February
1982  to April 1985,  it included only 8 storms where loads  were
estimated  at  both gage sites 2 and 3.   Total storm runoff  and
runoff  ratios were estimated for 32  events.   Additional  storm
•-infrfl orobnbly exist but were not reported.

     #e examined the data by regression analysis.  Storm loads of
't-Si  ;>nd  event  mean concentrations of TSS  and  turbidity  were
t ra.\ s f vrmed  to their logarithms.   Paired observations from  the
two  subbasins were then regressed by linear least  squares,  and
the  simultaneous hypotheses that the intercept is equal to  zero
*nd slope is equal to one were tested.

     Results  from the regiession analysis are sho*n in Table 15.
.1   .lear indication of differences in water quality  between  the
                               74

-------
two subbasins was observed.   At 90 percent confidence,  we found
that  TSS  concentration  and TSS unit area  loading  rates  were
higher  in  subbasin 3,  the smaller basin,  than in subbasin  2.
Turbidity  was  not  significantly different at  the  90  percent
confidence level.

     Because  both  loading  and concentration were found  to  be
significantly  different,  we  considered  whether  storm  runoff
volume  and  rainfall/runoff  ratio  might  also  be   different.
Results  are shown in Table 15.  A higher level of confidence was
possible  in  this  analysis because there  were  more  stormflow
observations  that could be paired between gages than there  were
water  quality  observations that could be  paired.   The  runoff
ratio  was not transformed to its log.   The  analysis  indicated
that  runoff  volume  froa the two sites were  not  significantly
different,  but  the  rainfall/runoff  ratio  in  subbasin 3  was
significantly higher than that of subbasin 2 (p=.95).

     These  results  suggest  that the runoff  response  and  the
concentration of sediment in stormflow from subbasin 3 is greater
than  that of the larger subbasin.   Subbasin 3 is generally  not
included in the designated critical area of the  project.   These
observations suggest that  the  water  quality  data  should   be
considered  in a reevaluation of critical  areas.   Specifically,
subbasin  3  should be examined for unique  characteristics  that
make it contribute more than is expected.
Table 15.  Regression analysis to test for  differences  between
           water  quality  at  Gage  Site 2  and  Gage  Site  3*
           (Illinois RCWP)
             Parameter
N
                                      Conclusion  Confidence
TSS EMC**
TSS Load
Turbidity
Runoff
Runoff/Rainfall
8
8
8
32
32
S3
S3
S3
S3
S3
> S2
> S2
= S2
= S2
> S2
90*
90*


95*
  Data from Sta'e Water Survey Report 357.
 *SM!' is Event faean Concentrations.
     Other  analyses  froai the available data may  be  useful  in
.i-lading  the  project  and  producing early  projections  of  the
results.    Sepa at.ing  the bay water quality data from  the  mean
lakr-  water  quality data would allow analyses of each data  set,
The  Says may respond to land treatment at a different rate  than
the  main  lake.    Since  all  of the Innd  around  the  bays  is
designated as critical,   data from the bay stations may  indicate
other   sources  of  loading,    Mso,   a
                                                    the
                                                    may
                                            relationship
                              between
                               75

-------
concentration  in  •torn  samples  and  discharge  rate  can   be
developed  and  applied  to  storms with  incomplete  records  to
utilize more of the information in the monitoring data.

     Many  of the water quality projections and analyses  of  the
project are based on a delivery ratio developed from analysis  of
soils,  slopes, and.distance to stream network (Davenport, 1984).
Delivery ratios from this analysis varied from 22 percent  to 100
percent for one field site.   In general, they concluded that the
site specific delivery ratio was even higher than 47 percent, the
previously estimated overall delivery ratio.

     To  consider the validity of a delivery ratio of 47  percent
or more,  we compared an estimate of annual TSS load from  gaging
sites 1, 2, and 3 with an estimate of gross erosion.  Results are
shown in Table 16,   This is a rough estimate, because the actual
gross  erosion for the period of study *ay be different from  the
average annual gross erosion eatianated by the USLE.  The observed
sediment  delivery  ratio was about one-half of that used by  the
project.   It  seems  likely that the delivery ratio  for  larger
particles  may  be low,  while the ratio for fine  particles  may
approach  100 percent.   An accurate delivery ratio is  extremely
important  in getting a correct projection of water  quality  re-
sults from this project.
Table 16.  The Relationship  of Average Annual Gross  Erosion  to
           Observed Total Suspended Solids Yield. (Illinois RCWP)


      Land Use/Cover      Percent of Area     Average Erosion
                                              Rate (t/acre/yr)
      Cropland                 82.3*              3.8**
      Pasture/Hayland           5.4               0.9
      Woodland    ,              4.1               0.2
      Urban                     0.7               0.7
      Feedlots                  0.3              17.4
      Other                     7.2               1.6
      Overall                 100.0               3.30
                                          ***
        Site    Months*   Average TSS load   Deliv. Ratio
                            t/acre/yr         percent

                                                 23
                                                 14
                                                 21


* data
** data
***only
GS 1
GS 2
OS 3
from
from
mon t


22
20
19



0.
0.
0.
779
452
707
Davenport, 1984
1982 Annual Report
hs with complete records

used
                               76

-------
                           PROJECTIONS

     The  absence of significant water quality improvement  could
be due to any single factor or combination of factors such as the
following:

     1.   The period of record is still too short.

     2.   There  is  not  enough BMP implementation  to  cause  a
          detectable change in water quality.

     3.   BMPs  are  not adequate to reduce these  water  quality
          problems.

     4.   Critical areas may not be correctly defined.

     5.   Other phenomena are masking the ef %«5cts of BMPs.

     The  water  quality data collected thus far have  relatively
high  variation and represent a relatively short period of  time.
Thus far, only 37 percent of the critical area is under contract,
and roughly 10 to 20 percent of the critical area has implementa-
tion  of BMPs.   A change in water quality might be detected when
there  are  higher  levels  of  BMP  implementation.    The  fine
particles that are responsible for the high turbidity in the lake
and •  the  natric  (sodium)  nature  of  the  soil  may  not   be
appropriately addressed by the BMPs that have been  used.   These
BMPs  may  have the potential of reducing erosion,  but  may  not
substantially  reduce the fine particles concentration in  runoff
water.    «In  the unique case of natrijc soils,  practices designed
specifically to reduce the volume of runoff may be more- effective
than those which have primary effectiveness for erosion control.

     Other  phenomena,   such as lake turnover,  may also mask the
effects of the BMPs.   Lake turnover causes resuspension of  fine
particles from the bottom sediments,  but this process also helps
purge  the  lake  of sediment build up.   At  this  time,  it  is
difficult  to  determine exactly which factor or  combination  of
factors is predominant,  but it is clear that considering the low
level   of  BMP-implementation and the short period of  monitoring
record,  we would not expect to observe a water quality effect in
the lake.

     The  1984 Annual Report projected a 67 ->ercent reduction  of
;*> upended  solids  load  at  gage site  I   inder  the  following
."xiremely  optimistic conditions:    (1) all cropland  is  managed
•rii'-h  conservation tillage,  (2) pastures ard woodlands are  well
managed,  (3) waterways  and structures are implemented where needs
indicate,  (4) rainfall runoff is reduced by 50 percent,  and (5)
rainfall  energy  is  reduced by 45 percent,   No  projections  of
changes in lake water quality w?re attempted.   The complexity of
th*' ^y«?t. em makes all of these projections rather tenuous.
                               77

-------
                          IMPLICATIONS

     The Highland Silver Lake project had much advance  planning.
Critical  areas were defined and a sound monitoring approach  was
developed.   Yet there is no certainty at this time whether or not
the  water quality impairment can be reversed.  The main question
remaining  to be  answered is:  "Can BMPs effectively  reduce  the
erosion  of  fine particles of sediment from natric soils  suffi-
ciently to reverse the impairment of Highland Silver  Lake?"

     The  field study data should contribute substantially to the
answer to this question.  These data should be analyzed to deter-
mine  which BMPs are best for treating natric soils.   With  this
information,  the more effective BMPs could be promoted for   im-
plementation.    The   water  quality  data  should .be  used  to
reevaluate  where critical areas and important sources of  pollu-
tants  are  located.   Th-  relationship between stream  and  lake
water quality needs to be well established.   Characteristics  of
the lake (i.e. hydraulic retention time and the rates of sediment
settling  and resuspension) would help to estimate the  potential
effects  that  the  recommended  land  treatment  would  have  on
Highland Silver Lake.
                               78

-------
                     St. Albans Bay, Vermont

                             RCWP 12

                          INTRODUCTION
    The St. Albans Bay Project is located in northwest Vermont on
Lake Champlain.  Agricultural activity in the 33,344 acre project
area  consists  primarily of dairy operations with an average  of
330 acres and about 100 dairy cows.

     The water resources use impairments are related to eutrophic
conditions  in  the  Bay.   Excessive  macrophytic  plant  growth
.npairs boating;  algal conditions impair swimming, fishing, aes-
thetic  enjoyment,  and shore-line property  values.  Phosphorus,
contributed  from the sewage treatment plant (76%) and from  non-
point  sources (24%),  is believed to be the  limiting  nutrient.
Bacteria  from  these same two sources impair use of  waters  for
swimming.

E§r§B§£:tiyes of the Project

     The  following questions related to agricultural  management
and water quality are relevant to the project situation.

1.   What  degree of phosphorus,   nitrogen and  sediment   loading
     reductions can be accomplished through treatment of  80  per-
     cent  of   the  critical areas and sources in  an  intensive
     dairy farming area?

2.   What  are the most effective manure management practices for
     reducing phosphorus loading in northern U.S. climates?

3.   What  is the relationship  between total phosphorus   loading
     reductions   and  orthophosphate  reductions  from  improved
     animal waste management?

4.   To  what  extent and how quickly will a  lentic  water  body
     respond  to  significant pollution loading  reductions  from
     both point and non-point sources?

o.   What will be the effect on stream nutrient loading of elimi-
     nating  the  application  of  manure  to  frozen  ground  in
     northern U.S. cLiraates?

6,   How  well  does  the CREAMS  model predict  actual pollutant
     losses  from  various  agricultural  management  systems  in
     northern U.S. climates?

7     What   is  the  minimum  water  quality  change  which   can
     be  detected by a  well-designed,  comprehensive  monitoring
     program?
                               79

-------
     What  role do?»3 *«diment-bound phosphorus play in  improving
     the  trophic  status of a  phosphorus-limited  lentic  water
     body?

               Strategy.
     All  RCWP  BMPs except BMP 13 and BMP 16 have been  approved
for  the  project.   The emphasis has been  on  BMP 1  (permanent
vegetative cover),  BMP 2 (animal waste management), BMP 8 (crop-
land  protective  system),   and BMP 15  (fertilizer  management).
The greatest need has been for BMP 2 which has been part of every
RCWP water quality plan written in the project.

     The  project has developed a relatively rigorous method  for
selecting  critical areas.    The criteria include present  manure
management  practices,   water  resource  accessibility,  on-site
evaluation,  and  water  quality monitoring needs.   A  total  of
15,257  acres or 45.8 percent of the total project area has  been
designated as critical.

     Information/Education    efforts   in   the   project   have
been extensive.    Information about the project has reached every
resident  within the watershed,  and the effectiveness  of  these
activities  is evidenced by the fact that nearly every  landowner
in the critical  area has signed up for the program.


Water "Quality. Monitoring Strategy

     A  thorough and complex water quality monitoring (WQM)  pro-
gram  is  an  integral part of the project.   The design  can  be
briefly summarized as follows:

     1.   monitoring of St. Albans Bay.  (This  includes  special
          studies  such as  bay  circulation patterns  and  phos-
          phorus remobilizat ion from bay sediments.)

     2.   automated  sampling of streams to  determine  pollutant
          loading trends.

     3.   randoa? grab sampling of streams to determine concentra-
          tion trends.

     •1 .   monitoring  at the edges of  t'.co smaH paired watershed
          sites.

     5.   monitoring of the St. Albans Bay sewage  treatment plant
          effluent.
                 BMP IMPLEMENTATION ACHIEVEMENTS

     The  project has been quite successful in attaining its
implementat ion goals.  Credit for this achievement belongs to
                               80
BMP
all

-------
of the agricultural Bgc/*vy personnel involved who have created  a
high  level  of  farmer awareness of both the water  quality  and
economic benefits of participation in the project.

     BMP  implementation  is occurring under both  RCWP  and  ACP
contracts.  Of  85 dairies and 15,257 acres identified as needing
treatment,   69  dairies (56 - RCWP,  13 - ACP) and 12,762  acres
(10,330 - RCWP,  2392 - ACP) were under contract as of  September
1984.   The  project  estimates that 13,442 acres and 74  dairies
will eventually receive treatment.  This represents 87 percent of
the critical dairies and 88 percent of the critical acreage.

     Project  personnel believe that the awareness brought  about
by  the  RCWP project has contributed an impetus to  other  dairy
operations for initiating barnyard improvements under ACP and for
achieving  a  general  reduction in fertilizer usage  as  farmers
understand better the nutrient values of manure.  Of the 56 far.;s
under RCWP contract as of September 1984, treatment had been com-
pleted on 29.   Almost exactly half of the animal manure produced
in the project was under best management as of this  date.   This
percentage  is  much  higher if only the critical  area  is  con-
sidered.    Nearly  all  cost-share  funds  (96*)  have  gone  for
managing animal waste.   This includes cost-sharing directly  for
BMP 2 (86%) and also for BMP 12 (10*) which has actually involved
barnyard runoff improvements.

     Overall  the  project personnel estimate that 74 percent  of
all BMPs contracted have been completed.   This indicates clearly
the farmer enthusiasm for the project.
                                                  •
     The  project requested and received'supplemental funding for
FY 1985 to cost-share additional requests and to modify  existing
contracts.   In addition to these pending applications,   howeve ,
there  remain several dairy operations in the critical area which
have not applied for RCWP.    It is somewhat unclear whether these
farms  could  be persuaded to participate  if  cost-sharing  were
available.
                  ANALYSIS OF FARM LEVEL COSTS

     Impacts  at  the individual farm level  are  illustrated  in
Table 17,  which shows the relevant financial and physical impacts
associated with BMP adoption.  Two representative farms, a medium
ind l-^rge  size,  are depicted in the table.  In each instance the
daily  spreading' situation is presented which reflects  the  pre
projet. l  condition.   The  project effects can be  determined  cy
•roaipar I ?ig   the various manure storage alternatives with the daily
3prea
-------
Table 17.  Annual  Impacts of BMP Adoption for Two Typical  Dairy Farms in the Jewett
          Brook Subwatershed of the St. Albans Bay RCWP Project.
•
*
Impacts "
FinnncJnl F.ffects
Cross Revenue * i
Vre-tax Met, Income-
Cost Share (Covrt)
Cost Share (Farmer)
Environmental Effect
Cross Erosion (Tons)
Delivered Sediment (Tons)
K Loss
Adsorbed (Ibs)
Dissolved (Ibs)
Total
P Loss
Adsorbed (Ibs)
Dissolved (Ibs)
Bionvailable
«

Farm Operations
Cows Milked '
Replacement: Bought
Raised
Crops
Corn (Acres
Alfalfa (Acres)
Manure Spread (Tons)
Spring
Summer
Fall
Vintcr
frV-rrUiz-i- Purchased (Tons
«-abor Hired ^Hrs)
Spring
Sunncr
Fall
Win?. .;;'
Total

w/o RCVP
Daily
Spread
109,470
27,427
0
0

102.6
40.0

554
171
725

30.8
141.0
171.8


58
32
9

48.0
60.4

450
450
450
450
20.5

290

276
214
780
Medium Fam
\ W/
^Semi-Solid
* A-Frame
•
•
107,619
29,785
27,862
8,966

68.0
26.7

407
93
500

22.6
23.0
45.6 .


58
20
21

53.1
60.4

915

'915

9.04

228

194
138
560
n
P.C'-f?
Liquid
' (Earthen
! Pit)
•
108,948
28,752
11,312
17,783

68.2
26.7

407
126
633

22.6
72.0
94.6
* „

58
29
12

53.1
60.4
-
904

904

9.7

224

190
129
543
•
Large
7 w/o RCWP :
; Daily
• •
• *
• •
191,500
42,675
0
0

171.0
66.6

9-41
302
1243

51,6
239! 0
. 290,6


100
70
0

87.4
114.0

761
761
761
761
39.8

414
108
383
303
1208
Farn ^P
w/ RCVP
Liquid
(Earthen
Pit)
191,500 ,
45,800
16,005
22,633

114.3
44.8

699-
226
925

38,0
123.0
161%0 ^
m
V
100
70
0

87.4
114.0

1522

1522

22.0

304

240
167
TiT
          net  income  to  land, management, and owner  labor.   Value,  are adjusted aftc_,
     years when  the facility is  totally amortized  and  repair and replacement 'costs arc
     	j  to be  50 percent of original costs.

-------
     Pre-tax  net  fara  income is used «s a measure  of  project
efficiency  since taxes tend to mask what is occurring  with  re-
spect to the allocation of resources.   For example, rapid depre-
ciation schedules do not provide a realistic measure of equipment
life.   Also, consider two pre-tax situations that are identical,
but  because of different exemptions the net incomes can be mark-
edly different.   Consequently,  pre-tax net farm income is  used
when examining economic efficiency.

     Examination  of Table 17 reveals that in each case  the  in-
stallation  of  a nanure storage facility with a 75 percent  RCWP
cost sharing of eligible costs results in an increase in  pre-tax
net  farm  income when compared with the pre-BMP state  of  daily
spreading.   In the case of the medium size farm,  the semi-solid
and  liquid  storage facilities provide increases in pre-tax  net
farm income of $2,358 and $1,325,  respectively.  The; adoption of
a  liquid  manure  pit yields an increase in  pre-tax  income  of
$3,125  in the large farm setting.   After 20 years these  values
increase  somewhat  as  the structures are assumed  to  be  fully
amortized  at that time.   However,  it is further  assumed  that
repair  and replacement will be 50 percent of the original  cost.
Thus,  the  annual pre-tax net farm income figures for the medium
size farm after 20 years will be $2,887 and $2,355  respectively,
and $4,435 for the large farm liquid pit.   Although there exists
an  increase  in  current income it must be recognized  that  the
simulation performed by the IP model accounted only for the farms
financial  commitment,  i.e.,  25 percent of eligible RCWP  costs
plus some noneligible costs, not the government's cost share.

     After tax income was not used to evaluate project efficiency
because it does not properly account for resource use,  but  also
because  the  number  of possible tax scenarios was too  many  to
incorporate into the analysis.  However, the reality of every day
decision  making dictates their consideration.   Although  it  is
difficult to make a specific statement about the tax implications
it can be generalized that the Federal and Vermont state tax laws
are  such that they will reduce the negative impacts of BMP adop-
tion.  The farmer has several tax incentives applicable to the 25
percent  fans cost share and other RCWP associated  expenditures.
An  investment  tax credit can be claimed on the  manure  storage
facility if it is more than an earthen structure.  This of course
would apply to additional spreaders,  tractors,  and pump invest-
ments.    Further,  manure  pits that do not qualify as  soil  and
wa^er  conservation  expense and can therefore be  deducted  from
>• -».xah 1 e income ,

     Vermont's income, property,  and sales taxes also have provi-
       that  can affect conservation and environmental  programs.
      t' s  income tax schedule in based on a straight  percentage
,2P percent)  of the federal tax.   As such, all tax considerations
jpp] xr-^ble  to BMPs for federal tax purposes also apply  to  Ver-
mont    In addition,   manure storage facilities are also exempted
from property taxes.
                               83

-------
                   WAfiSt DUALITY DATA ANALYSIS

Summary of Project Results

§1.1 Albans Bay  The Bay monitoring program consists of:

     1.   periodic (approx.  20/year) grab sampling at four sites.

     2.   some biological monitoring at these four sites.

     3.   sediment phosphorus release studies.

     4.   a  study of bay circulation which"involves using  wind,
          water,  current and concentration data to determine the
          effect of bay circulation on water  quality.

     The  bay concentration data show no obvious trends over  the
1982-1984 period.  Mean concentrations of total suspended solids,
total P,  and  TKN  were slightly higher in *83-'84 than '82-'83.
This  may  be  associated with higher than  normal  precipitation
during '83-'84,  although it must be remembered that most of  the
nutrients in the bay itself come from the sewage treatment plant.
Ortho-P concentrations were about the same both years.

     No  biological monitoring results from the bay were included
in the 1984 report materials.   However,  extensive biomonitoring
information from the five tributary sites was reported.

     The sediment phosphorus release studies  appear to have  been
completed.  The following results are evident from this study:

     1.   Phosphorus  is  released from bay sediments under  both
          aerobic and anaerobic conditions;  however,  release is
          .more rapid under anaerobic conditions.

     2.   Phosphorus release rates increase with both temperature
          and flow rate at the sediment-water interface.

     3.   Over  extended  periods of time phosphorus can  be  re-
          leased  frost  deeper  sediments as  well  as  the  upper
          several centimeters.

     4.   The  sediment column contains sufficient phosphorus  to
          be a long-term source for continued supply.

     These results have 'several possible implications relating to
t->otenH"l water quality improvement in the bay.  Most importantly
they j-'j'gest that improvement of the bay's trophic status may lag
behind  anticipated  reductions in phosphorus input.   It is  not
• •1..MC  iYom information contained in project Annual Reports,  just
hov? large a sediment phosphorus reservoir exists.

     Another implication is that,  as volatile suspended  solids,
fluttients,  and oxygen-demanding dissolved organic loads are sub-
§t-antin]ly  reduced,  the  incidence  of bay sediment  conditions

-------
becoming anoxic may tv rr^luced.   In this case flushing of  sedi-
ment phosphorus may t*k« longer, but the bay may exhibit improved
trophic conditions in the meantime.

S§Y.  Circulation  Studies   No  results of  the  bay  circulation
studies  were  included  in the  1984  annual  progress  reports.
Earlier  reports  indicated that detailed computer models of  bay
circulation patterns have been developed and are ready for use.

Tributaries  The tributary (level 2) studies include:

     1.   automated  stream  sampling to  determine  loading  and
          concentration trends in the four main tributaries.

     2.   biological monitoring of fish species,  benthos,  other
          invertebrates,  and periphyton.

     3.   meteorological  monitoring  to relate  .tributar •  water
          quality changes to meteorological variables.

     4.   detailed  land  use monitoring (manure spreading  logs,
          dates  of field operations,  etc.) to relate  tributary
          water quality observations to land use activity.

     Only  two full years of tributary monitoring had  been  com-
pleted,  analyzed  and  reported as of September 1984.   The  WON
personnel  .are,  thus,  reluctant  to project any  water  quality
trends based on this short period of record.  Also, BMP implemen-
tation was in progress throughout this period,  although completed
impleme.ntation was significantly higher during .the second year.
      ,                      *                                  •
                                                  f ,
     The  analysis  conducted by the project shows several  'note-
worthy spatial .trends.   The subwatershed (Jewett Brook) with the
greatest  intensity  of agricultural activity shows  the  highest
concentrations and loads of agricultural NFS  pollutants.   Phos-
phorus loads are about 20 times the average for U.S. agricultural
watersheds.   It is this subwatershed which receives the greatest
BMP   coverage,  and thus,   has the most potential to show  water
quality improvement during the RCWP timeframe.

     Although  the  WQM personnel are hesitant to claim  a  water
quality  improvement in Jewett Brook,  we believe that there  are
already  some  good  indications of a  significant  reduction  in
nutrient concentrations.    Referring to Figure 16, large (approx.
50%) reductions in mean orthophosphate and total Kjeldahl  nitro-
gen  concentrations  are  evident between Year 1 and  Yeai   2  of
lonitoring.  As noted above,  there were substantially more manure
        nt  systems completed in Year1 2.   The observed reduction
    orthophosphate concentration occurred even though  precipita-
      was  30 percent greater in Yeai 2 which would tend  to  in-
    se NPS nutrient concentrations.    This gives further  support
    the  suggestion that  the manure management BMPs are having  a
positive effect.   It is anticipated that these trends will become
         given a longer  monitoring  period and the high level of
                              85
o

-------
    75T
              TRIBUTARY STATION
          21      22      23     24
       JEWETT STEVENS  RUGG   MILL
    1.0-r
         75
        S,P
       6.24.8
                                      4.44.8
                     BAY STATION
                        12
             OUTER  INNER
                                                       11
                                                 0.1-r
                                12
                                24
         25
                11
           12
,-T  10-T
                                     18.317.1
          21      22      23
      JEWETT STEVENS PUGG
  24
MILL
 "5
STP
   II
OUTER
  12
INNER
Figure  16.  Mean Concentrations  of  Solids, Phosphorus and Nitrogen at
             the Tributary and  St. Albans Bay Trend Stations  for Two
             Years.  (From VT  1984 Summary Report, page V-2.)
                              86

-------
BMP  implementation \%Q -30*) projected.   Based on the percentage
of  manure  brought  uttder  best  management,   the  project  has
developed  models which estimate that total P runoff losses  have
been  reduced about 21 percent and dissolved P runoff  losses  by
about 57 percent in the Jewett Brook drainage (Figure 17). Reduc-
tions  by  1990  are projected to be 30 percent  and  86  percent
respectively. The 1984 report suggests that the large majority of
P  inputs  after bringing manure under BMP is from soil  erosion.
It would appear,  therefore,   that a large excess of P has built
up in some project soils.  These results emphasize again the need
for good fertilizer nutrient management to achieve maximum reduc-
tion in P loadings.   The project has noted that further  refine-
ments  of  these  models  are underway  and  should  improve  the
accuracy of the projections.

     Figure 16 also provides a graphic portrayal of the magnitude
of  the  nutr ent inputs from the  sewage  treatment  plant. (STP)
Plants constructed or up-graded in the past ten years are expect-
ed to produce final effluent with total P concentrations of  less
than  1 mg/1.   This illustrates once again the importance of the
STP upgrading to obtaining improvement in the Bay itself.

     The  biological monitoring information presented in the 1984
Report reflects essentially pre-BMP or baseline conditions.    The
tributaries  were found to have basically the fauna  expected  of
small  Vermont  warm-water  streams moderately impacted  by  NPS.
Jewett  Brook  exhibited a unique fish  species  composition  and
distribution  which  may  change significantly as NPS  loads  are
reduced.
         Watershed"   Field  Sites  The objective  of  a  "paired
watershed"  experiment was to show the effects of land management
on  water quality by effectively controlling  for  meteorological
and other tim,. - related variation.   The paired watershed design
conducted by the project is shown in Figure 18.

     Essentially  the  design involved doing the opposite of  the
RCWP  BMP  implementation:   (1) two corn fields  initially  were
cropped under best manure spreading practices during the calibra-
tion  period;   (2) management on one field reverted to the  pre-
RCWP practice of field spreading manure through  the winter. Thus,
the  expected  result from comparison between fields was  an  in-
crease  in  pollutant losses from the winter  spread  field  that
should  be approximately equal to  the reductions accomplished by
installing animal waste BMPs on poorly managed fields.
                               87

-------
L
                                        JO

-------
               Paired Watershed Treatment Schedule

                                    PHASE:

                         Calibration     Treatment
Watershed:
     Control
     Treatment
          Best
         Manure
         Management
         Best
         Manure
         Management
Best
Manure
Management
Winter
Spreading
Figure 18.
Paired  watershed treatment schedule
Summary Report V-3, 1984)
        (from  Vermont
        Analysis and Interpretations

     The  project monitoring personnel have  conducted  extensive
analysis   of  tfee  paired  watershed  experiment.    A  somewhat
surprising  result  was  that wint.er  manure  spreading  actually
decreased   runoff  concentrations  and  mass  export  of   total
suspended solids (TSS) by 68 percent and 50 percent, respectively
(Figures 19 and 20).  The  reduction in TSS was attributed  to  a
"mulching   effect"  of  the  winter-applied  manure.    Volatile
suspended solids (VSS) concentrations were al'so reduced by winter
spreading.   Our analysis of the TSS and VSS data confirmed these
results.   The  divergence of the regression lines in  Figure  20
suggests  that the mulching effect is most pronounced at high TSS
concentrations.

     In contrast to TSS and VSS, concentration increases of total
phosphorus, ortho-phosphate, total Kjeldahl N and  ammonia N have
been  observed  as a result of winter spreading (See Figures  19,
21,  and 22).   In general,  the elimination of winter  spreading
reduced  concentrations more than mass export because the  winter
manure  application reduced runoff volume from the treatment site
by 78 percent.   This is presumably due to improved  infiltration
conditions.

     The  project, found that total mass export increased from the
winter ?pread field of  ortho-P,  1500 percent,  total P, 11 per-
ceril,  TKN,  148 percent, and annnonia N, 618 percent, even though
th  -us,*,jit of runoff decreased (Figure 23),   This gives a direct
inul cation of the probable effectiveness of eliminating the prac-
»-ice v f winter spreading.   It should be note--!, that BMP 2 in this
project also reduced manure-derived stream inputs by  eliminating
losses   from  stacked  manure  ami  by  improving  barnyard  and
•ailkhouse  manure conditions.   Thus,  the expected  improvements
from treatment of a farm with BMP 2 might be different from  what
.<>as indicated by the paired watershed experiment
                               89

-------
    g  40-r   CALIBRATION
    O

    HP
    Si 20
    ww
    <2
    P-J
    bo
    HW «
                          TREATMENT

                          -68%
                      '"
ED
          LOWER UPPER    LOWER  UPPER
ATILE SUSPEND
IDSIMG/LI

M *.
SO

O
    §
    I -2t
  <3>^
  18.
•
I


"^
-24
I
%
r



 ~ ^CTJ
 ££1,
    IU

  ol_-'
  X W-J
  = §5
  Oa*o
    iu
  i8
  !^ff 2.

  2!=
  < -J^-
  gxo
  ess0.

                    * * *
JH
                         •••1480%


                         +114%
                    **
                                              UPPER
                                              CONTROL
                                              LOWER
                                              TREATMENT
                                              OBSERVED-
                                              PREDICTED
                                              TREATMENT
                                           **  ps 0.01

                                           * * * PS 0.001
2uJ
Z'S-.
003
s*^
*H-O
«C »»«r
t £«=
O
                         4-576%

          LOWER  UPPER

                        LOWER  UPPER
?igure 19,
           Mean  concentrations  in  runoff  froa  the LaRose  farm
           paired watersheds.  (Fron VT  1984 Summary Report, pg,
                             90

-------
     1000  -
"a

as
10
0)
~   100   -
oo

o
o
z
o
o
o
oo

£   10
                                                 D /
                                              *  /
                                            * /
                                               D
                            /*O
,'  *


 *
                                             *--*-^.  1983  (Calibration)



                                             a  a o  198«  (Post-Treatment)
            I                 10                100              1000

                  TOTAL  SUSPENDED SOLIDS CONCENTRATIONS (Control Watershed)
                                                           10000
    rc  20.  Regression  analysis  of paired observations of total suspended solids
            concentrations,  treatment vs. control watersheds, 1983-1984. (VT RCWP)


                                       91

-------
    10  -I
-o
OJ
JZ
I/I
s_
0)
-M
ITJ
 c
 O)
 O)
 i-
     i  -i
ex
t—

UJ
t_)

o
cs:
o
     01 -J
                           a
                            a
                                                 ^n
                                                        a  /
                                                       * X
                                                      /  a
                                                   /*
                                                  /*
                                             *   /*

                                              ./**
                                   U°°o   if
                                  ffr,  ° *St   *
                                      •A
                                  **?
                                                          -+  1983 (Calibration)
                                                      ODD
                                                                  (Post-Treatment)
        001
                          .01                1                 1


                  TOTAL  PHOSPHORUS rONCEf'TKAflOKS  (Control Watershed)
10
                Regression analysis jf paired observations of total phosphorus

                •-oncentrations,  treatment vs. control  wa'^rshejs, '5-CJ-198'1-fvr RCWP)

-------
    10
•o
a;
a>
-*->
c
a>

     Oj   -
z
O
LU
LU


g


Q.


O


n.
     001  -
          0001
                                                 * /
                                                  f
                                                  f
                                                 s

                                                 ' *

                                     *
                       / *
                                                      *-^-* 1983 (Calibration)



                                                      e-B-s 199^ (Post-Treatment)
                                                                                  10
               .001             01             1              1


         ORTHO HHOSPH/JL CONCENTRATIONS (Control Watershed)


Figure 22.  Regression analysis of paired observations of ortho phosphate

            concentrations, treatment vs, control watershed, 19G3~1984.(V7  RCWP)


                                93

-------
                          IMPLICATIONS

     This  project  is making several important contributions  to
our  knowledge of agricultural NFS control.    These  are  briefly
summarized below:

     1.    A  properly  conducted  paired  watershed  experimental
          design  can document water quality effects of  specific
          BMPs within a two year timeframe.

     2.    A  very detailed land use information base is necessary
          to  attribute  water quality changes to  specific  land
          management activities.

     3.    Eliminating the practice of winter manure spreading  in
          northern  U.S.    climates  would  have  the  effect  of
          ^creasing  suspended sediment losses but would  reduce
          surface  losses of total  phosphorus,  orthophosphorus,
          and total nitrogen.

     4.    The  major  portion of surface pollutant  transport  in
          northern  dairy areas is associated with  winter  thaw,
          spring snow-melt or spring precipitation events.

     5.    Some  modification  of the CREAMS  model  is  needed  to
          describe  field   losses of agricultural pollutants  in
          northern U.S climates.
                               94

-------
          CALIBRATION
 TREATMENT
iS 0.54-
  ;x
  °-  0
         LOWER UPPER
    0.1 -p

  ^0.05


II
£z
Oo.  o •
3j

2--
LOWER  UPPER



•M500%
                           +148%
                      UPPER
                      CONTROL
                  I
                      LOWER
                      TREATMENT
OBSERVED-
PREDICTED
TREATMENT
                                            **  p = 0.01
     •3T

     .2--
         LOWER  UPPER
                           +618%
LOWPR  UPPER
Figure 23. Mass  export of phosphorus  and nitrogen  from  the  LaRose
           Paired watersheds.   (Frow  \T 1£84 Summary  Report,  pg. V-7)
                                95

-------
              Conesioga Headwaters,  Pennsylvania

                             RCWP 19


                          INTRODUCTION

Background

     The  Conestoga  Headwaters project is located  in  Lancaster
County,   which  is  acknowledged  to be the most  intensive  non-
irrigated  agricultural  county in the  U.S.    The  intensity  of
agricultural  production in the project area is even greater than
for the  county as a whole.   This production takes the form of row
cropping  as  well as intensive animal  production  (approx.  two
animal  units (a.u.) per acre).   As a result severe  groundwate.
(bacteria,  nitrates)  and surface water  impairments  (sediment,
phosphorus, nitrogen, bacteria) have been documented.  From among
the  multitude of agriculture-related impairments(fishery,  water
supply,  contact recreation, aesthetics and downstream eutrophica-
tion)   the  project  has chosen to  focus  increasingly  on  the
drinking   water  impairment  caused  by  excessive  nitrates  in
groundwater.   In the opinion of NWQEP,  this is the most serious
water  resource impairment because it impairs the drinking  water
supply  for  175,000 people both within and outside  the  project
area.   Several  cases of methemoglobinemia have been reported by
the project for infants in the project area.   There has also been
increasing concern about surface transport of nitrogen forms  and
herbicides in the context of the Chesapeake Bay studies.


Perspectives of the Project

     The  following questions related to agricultural  management
and water quality are relevant to the project situation.

1.   Can  animal manure be managed sufficiently to protect  water
     quality in an area where the nutrient content of this manure
     exceeds crop needs?

2.   Is  there an inherent trade-off between practices designed to
     reduce  surface transport of nitrogen and those designed  to
     minimize nitrogen transport to groundwater?

.:;     5-;   groundwater  resources  in karst areas  respond  rapidly
      ;'-,ough  to reflect changes in land management within  a  ten
     year tiraeframe?

'\    Can groundwater nitrate levels be reduced below lOag/1 in an
     area  with  an animal density of 2 a.u./acre with  a  public
     investment of approximately $60/acre of agricultural land?
     How  will groundwater nitrate levels change- in  response  to
     the  construction of manure storage facilities which  permit
                               96

-------
     better timing of nanure application?

6.   Do  herbicides  impair  groundwater in  a  karst  area  with
     intensive row cropping?

7.   What  BMPs are the most cost-effective for reducing  ground-
     water impairments caused by agricultural activity?

L§Qd Treatment Strategy

     All  HCWP  BMPs  except BMP 13 have been  approved  for  the
project.   The  emphasis has been on BMP 2 (animal waste  manage-
ment),   BMP 4 (terraces),  BMP 7 (grassed waterways), BMP 9 (con-
servation tillage) and BMP 15 (fertilizer management).

     Targeting  of cost-sharing to critical arpis has been under-
mined greatly by lack of farmer participation.   The project  has
established  three  priority areas based on water  quality  moni-
toring needs and groundwater nitrate levels.

     Priority 1 - The Little Conestoga watershed (3700 acres)

     Priority 2 - Other lands within the carbonate area  (areas
                  with highly permeable soils).

     Priority 3 - Non-carbonate area.

     .Information/Education  efforts during FY84 'included a  news-
letter sent to 25 percent of farms, visits to 12 percent o'f farms
by SCS and ASCS, availability of two no-till  corn planters, fifty
RCWP posters,  1500 pamphlets, public meetings, and meetings with
Amish church leaders.

W§t§r Quality. Monitoring IWQM^ Strategy.

     There  are three levels of WQM being performed.   The  first
level  is  a regional network which monitors  ground  and  surface
waters in the entire project area.  The network includes 2 stream
gauges  which monitor major storms,  four baseflow sites  sampled
monthly, and 43 groundwater sites sampled quarterly.

     The  second level involves more detailed monitoring  of  the
Little Conestoga watershed.   (2 stream gauges - major storms;   7
base f*iow  sites  - 17  times/year;   5-10  grorndwater  sites
quarterly).   The  watershed has two paired watersheds within it,
one  designated for a high level of BMP implenjentat ion  (nutrient
          . ) and the other to serve as a control.
     The  third level is the two 25 acre field sites  which  have
-.nterisive   monitoring  of  surface  and  grounawater   pollutant
transport:.   These sites are scheduled to have nutrient management
and/or erosion control BMPs installed after a suitable background
•nonitoring   period.     One   site  is  presently  in   the   BMP
implementation  phase  while  the other is still in  the  pre-BMP
<. • h f\ s e .

-------
                 BMP IMPLEMENTATION ACHIEVEMENTS

     The  actual implementation of BMPs through the RCWP  program
has been far below anticipated levels.   Out of 1250 farms in the
project  area  only 96 have filed an RCWP-1  application  through
9/30/84.   During this same period 51 contracts have been  signed
which  cover less than 4 percent of the project area.   Only  ten
contracts were signed dmring FY84. * The original project goal was
to obtain contracts on 300 farms;  however, this has been revised
to 90 farms in the face of low farmer participation.

     The  reasons for the lack of participation are numerous  and
complex.  The Economic Research Service (ERS) reported a 40 page,
indepth  analysis of the situation in the 1984  Progress  Report.
In the opinion of NWQEP, the  Tsic problems are that:

1.   Animal  production  is  so intensive that  manure  nutrients
     exceed  crop utilization potential,  and thus  the  economic
     incentive to properly manage animal waste is lost.

2.   The BMPs which potentially benefit the farmers by protecting
     their  soil  resource base are not the same BMPs  which  are
     needed to address the groundwater nitrate problem.

3.   The  cost-share rates for several key BMPs have been set too
     low to expect significant participation.

4.   Farmers have had inadequate access to soil nitrogen* cont&nt
     information  to make cost-efficient judgments on  fertilizer
     usage.

5.   Over  50  percent  of tht farmers are Atnish  with  a  strong
Over  50  percent  of tnt farmers are
cultural history of self-sufficiency.
     The  underlying  assumption throughout the RCWP program  has
been  that  farmers  will  have  inherent  economic  interest  in
utilizing  animal waste to its fullest potential in meeting  crop
nutrient requirements.   Hence cost-sharing BMP 2 was intended to
assist  with  the  relatively large initial  capital  outlay  for
constructing  a storage and application  system.   However,  when
manure  nutrients  exceed  crop needs,  there  is  no  longer  an
economic  incentive to efficiently store and apply manure.   From
this p'.-rspect ive it is not surprising that there has been  little
fg-roer  interest  in constructing manure management systems at  a
30  percent  cost-share rale.   Even in RCWP projecfs where  crop
 lull-lent requirements exceed manure nutrients, and manure storage
has  greater value,  a 75 percent cost-share rate has  often  not
insured the desired level of Dinner participation.

     For  farmers  with  an excess  of  manure  nitrogen,  manure
 .fcorage  systems  should  be designed to  allow  maximum  ammonia
 volatilization and denitrification.   The ERS report states  that
 in uncovered, 6-month storage systeru is optimal for this purpose.
                               98

-------
These systems are now ter'isg emphasized in the project.  It should
be  noted that maximua {nitrogen volatilization takes place  under
conditions of daily spreading.   Manure storage systems partially
compensate  for  this  disadvantage by enabling  the   farmer  to
refrain   from   spreading  on  frozen  ground  and   immediately
preceding  storm conditions  which increase nitrogen transport to
ground or surface water resources.   The net result is that some-
what  more  nitrogen  is  percolated  to  groundwater  under  the
storage-application system than under the daily spreading  system
when  manure  is  applied at the nutrient-excessive  rate  of  40
tons/acre.   When  manure  is applied at 20  tons/acre  (i.e.  an
amount  better Batched to crop needs) storage systems are predic-
ted  to  reduce  losses to groundwater because  of  the  improved
application timing options.

     The   previous  discussion  eepbasizes  the  importance   of
nutrient  management  to address the groundwater nitrate  impair-
ments  in  the  project  area.   The  typical  situation  at  the
beginning  of the project was that excess manure  nutrients  were
being  applied to cropland and then chemical fertilizer  was used
in addition.   The economic analysis of BMPs performed by the BRS
shows  clearly  that BMP 15 (fertilizer management) is  the  most
cost-effective BMP for reducing losses of nitrate to groundwater.
Considerable  progress has been made toward reducing  unnecessary
chemical fertilizer usage primarily through: educational efforts,
obtaining a 50 percent cost-share for BMP 15, requiring BMP 15 in
all contracts with an animal density of 1.5 animal units per acre
or greater, and increasing the accessibility to nitrogen soil and
manure tests.

     Even  with these efforts,  however,  only $16,185 (7.5*)  of
RCWP  cost-share  funds  have been contracted for  BMP  15.   The
proj ct personnel estimate that BMP 15 has reduced edge of  field
nitrogen loss by 38,880 pounds which is 56 percent of the total N
reduction.   At  the other extreme 49- percent of cost-share funds
have  gone  for terraces which have been responsible for  only  6
percent of nitrogen loss reductions.   If the primary goal of the
project  is to cost-effectively address the  groundwater  nitrate
impairment,  it  is clear that an increasing emphasis on nutrient
management and a decreasing emphasis on expensive erosion control
practices is needed.

     It should be noted that the project estimates that more BMPs
have  he^n  implemented exclusive of the RCWP program than  under
?OWP >  v.ract since the project began.   This phenomenon has been
j i. tr lo-• ted  to  tho restrictions built into  the  RCWP  contracts
vn '.  ••  ?-j.sentially require that any land under contract must hnve
,'•  -• '   "as  which reduce erosion down to "T".   The  ERS  analysis
5h-;.»   clearly  that   the most cost-efficient  surface  transport
.T:duo i, 1 ons of nitrogen,   phosphorus and sediment are accomplished
by  practices which are not sufficient to reduce erosion to  "T",
but   '-^her reduce erosion by 30-60 percent at very low  per-acre
'^osfc.    The adoption  of these practices, including contour strip--
cropping,  cover crops,   reduced tillage, diversions,  and grassed
            through  ACP and without cost-sharing  indicates  the
                               99

-------
increased acceptability of these practices to farmers relative to
the RCWP contracts being developed.

     The total nutrient losses to project surface and groundwater
resources  can  only  be  reduced to a  somewhat  limited  extent
(approx.  30-35*) through optimal management of nutrients in  the
watershed  and traditional soil and water conservation practices.
It is recognized that further reductions can be achieved only  by
either:

     1.   reducing the cumber of livestock

     2.   exporting nutrients from the project area.

The  project is continuing to explore possibilities  for exporting
nutrients.   These include hauling manure to be used on  cropland
outside the project area and drying/comporting and bagging manure
for retail sale.  The 1984 Annual Report stated that at least one
such facility has begun operation in the project area.

     Another aspect of the manure management situation which  has
been overlooked involves poultry manure.   While poultry accounts
for only about 13 percent of total manure production its nitrogen
content  is such that it accounts for approximately 40 percent of
manurial  nitrogen  in  the project.   Thus  40  percent  of  the
manurial  nitrogen  could be removed by exporting poultry  manure
with minimal transport costs,  compared to costs of hog or cattle
manure  export.   The poultry manure also has the highest  market
value as fertilizer.  Thus, although the ERS analysis showed that
cow manure hauling would be very expensive per pound of  nitrogen
removed,  the analysis should be about three times more favorable
for poultry manure.


                  ANALYSIS OF FARM LEVEL COSTS

     Two  approaches were used to model the farm level impacts of
participation  in  the  Conestoga  Headwaters  RCWP  project.   A
representative  farm linear programming model was used to  assess
the  impacts of alternative manure storage and handling  systems.
The  impacts  of field level BMPs were evaluated  using  budgets.
Environmental effects were modeled using the CREAMS model in both
approaches.

      l:j
-------
storage were the ao%t n^-.i. -effective sys.?e«.j» for preventing field
losses  of  nitrogen,  lj.o*h because of lo**»»r capital outlays  for
these  two  systems - and the lesser  amount  of  plant  nutrients
available compared to other types of storage, particularly the 6-
month uncovered solid storage (Table 18).   A "typical" Lancaster
County  farmer  can reduce nitrogen losses by 10 percent with  no
loss in income by applying manure evenly on all farm fields in an
environmentally sound manner for a given crop and by reducing the
rotation  intensity  of  some land—a  clear  indication  of  the
utility of the fertilizer management BMP.

     Cost  information is combined with field losses of soil  and
plant  nutrients in Table 19 to illustrate the cost effectiveness
of field BMPs for reducing losses.   Costs shown are total costs,
with  the life span of nonstructural practices being 5 years  and
structural  practices requiring maintenance for 10 years as  part
of  RCWP contract requirements by farmers.   The costs shown  are
estimated  average  costs  and  do not include  cost  sharing  to
farmers or other considerations (such as tax incentives).

     Among  the nonstructural practices,  conservation  tillage—
reduced  tillage and no-till—are effective in reducing pollutant
losses  at  a cost assumed to be zero for purposes of  RCWP  cost
sharing.   Other  benefits  such as reduced soil  compaction  and
increased  moisture-holding  capacity make this  BMP  a  critical
element  in  any economically sound nutrient and soil  management
program.   The  deep,  well-drained soils typical in the  project
area should provide equal crop yields compared with  conventional
tillage in normal years (Bepper,  et al., 1981).  Crop yields may
actually,  be  better  for conservation-tilled  land  relative  to
conventionally  tilled  land  in  those years  when  rainfall  is
significantly below average.   The only clear disadvantages  are:
1) the capital outlays required to convert tillage equipment;   2)
the  possibility of increased reliance on herbicides  to  control
weed  problems normally controlled by deep tillage;   3) conserva-
tion  tillage,  especially  no-till,  has been found  to  require
greater  fertilization to attain equal yields (although this  may
not  be a major concern for many farmers who have  considered  or
are  using conservation tillage.   The use of herbicides that are
incorporated into the soil during tillage or  planting,  combined
with  the  reduced runoff associated with  conservation  tillage.
The  use of herbicides that are incorporated into the soil during
tillage or planting,   combined with the reduced runoff associated
with  conservation  tillage,  may help to  alleviate  the  latter
o r • •!< 1 em .

     Permanent  vegetative cover will obviously control a  number
of pollutant problems,  but should only be considered in critical
areas  due  to its high cost to both farmers and  government  (50
percent cost sharing is provided).  Other nonstructural practices
include residue management(  stripcropping,   and contouring.    It
has been assumed in the RCWP project area that farmers' costs are
Iho  same  with  contouring and  stripcropping  (largely  due  to
protection   of  agricultural  productivity)  as  they  are   for
conventional  practices  once  the strips and contours have  been
                               101

-------
   v^
    y
 n  ra
 o  i
iw  O
    NO
 Crt
 0)  (0
 in
 ui  w
 o  o
i-i  m

4J  C
 C  -H
 V  <3
 E  i-
 C  "
 O  W
 u  c
•H  O
 >  o

u  tn

  •  o
 v  -j

 t!3  C  tfl
 0)  0)  4)

 00*5
<  l-
 (QUO
it,  c  U
^  -^ CM
•<  >>
    U Ou
  - nj 3
 tn  > u
 h*     a:
 a)   -
.o  tn  tn
 E  E  u
 3  V  V
2  u  JJ
    0)  «
Ji  >,  3
 O  WJ T3
 O      CJ
 •U  C  V
 m  o S
    to  60
    U  O
    O  01
    o.  c:
    rj  O
   -v., CJ
    0)
    N)  11
    rj £
 -  O
 a  u  n
 35 VJ  -H
CO
j—<

 i)
.—4

 re
                    w    *•*.
                    s     v
                    i*     w    in CM  ON ••*  %^ *if  \o co   in  o in  ON o*  in
                    O w  u    •ff^cocMooeo-^ovcMc-iOOOON-HincM—»O«-<

                    CL O "*••    f** *J  ON in  NO ^  CM %o  r^ O  ON NO  c^ O  f""* rx  N£> NC  «J ON
                    !«_!««•
                    C    J3
                   j:    •-<
                   a.    ^^




                       K  V
                       W  W
                       O  u    mcM«»r»-«r!«n^.fi^«naOf>»eoininO'iinincM
                      ^|   «C    »-lr»NONC^-<—l-^««-Hi-lOCM-->-ICOCM~(>-IONO

                           W    p*NO«tfCMr>.r».NO«ffCxNOincor«.rx,»o«ffr».rxNONO

                       O  C
                       W "



                                    (S
                       a /-v        eo  co            co H  19



                                            eo         PNI •»         ON O         en CM

                          S     mineocMinmcMninmineoinincooNinincMCM
                          O     «a>^cnn«jr«S'»»cn«»cM^^(-icM«»«»sj(n
                          U
                   111
                   3  L. ~2
                   -o «~i  a
                   5      °
    u>
    e
u  w
U  3
2  +>
                           C
                       W  O
                       c -*
                       0) *J

                       U  3
                       4) **3
                       O.  V
                           W
                           S

                           n

                           \n
            I  in  CM —•
            i  r«»  o in
            I  CM  O CM
O m eo
C CM ON
NO m m
                     en >*   i  in  ON NO  i
                      - . r».   |  CM
                      m O   I  r>.

                      ~t r*.
                                       ON ON  I  CM  oo O
                                       —I NO  i  CM  —4 ON
                                                                     •n
                                                                                I  en ON

                                                                                   r>
                                                                                                 O m
                                                   CM fx> r«*  *r\ CM —H o c** c^ in O  oo vo  r** ON
                                                   c^ en in.  ^\ ^o ^** »j ^^ CD ^n ON  ^j f**  ^y co
                                 ec  co sr
                                                      \o ~*
                                                                 co *^ i*x r^ rx.
                                                                                      ON
                                                                                                 NO O
           oooooooocooooooooooo
              ~+  n m
                            «—inm     «-tcnm    »Hom
                                                                       «-i r>  m
              "3 "3
               n  o
               4)  U
               U  W
             .  c.  a
               n  m

             .  >s
              r-t <
                                                        jar,  xi J3

               •O-O'O -13               •O'O'O'a'C'O'TStJ
               «H"i~'i-('^<*O"O'X3'O"H<»^'^^<'«-<  H-J-'-H
               ^^^^-••«^-l-^3333333  3
                          033330-o-Cr^-O-C-crcr
                          w  o*tr~cr--<-H-^<-^'-<'-"'^*^
            n>
           a
                                      a Q
^s  o
 n  to
 o
 I*  "O
 o. v
 m  u
    ai

-H  O
•r*  y

O  S3
 O  O
 n  m

•O  t5
 il  4)
 U  \-
 4J  ftA

 O  O
 u  u
 c  c
                                £ JC £  *-  •-• ^
                                                                                   J= X
                                                                         weccecccc
                              ccccoooooc
                                     o  o  o
                                     o  E  -
                                     C  I
                                     O  O  £
                                        e
                                        i
                                    \0
eest
T  T  i  CM
e  e   E  s  e
 i   i   i   i   i
CM CM  CM CM CM
                                O  O
                                e  s
                                i   i
                                                                                                             (3
                                                                                                             U
                                                                                                             o
                                                                                                              0)
                                                                                                              tA
                                                                                                             ^3
                                                                                                              C
                                                                                                              .o
                                                                                                              •/I
                                                                                                              V
                                                                                                                    •1)
                                                                                                                    u
                                                                                                                    U
                                                                                                                    01

-------
c
0
•o
re
41
1-
o.
to
^.
^\
^H
**4
CQ
O
•G •
1 1 »l
4J *J
•H U
» 41
•»*"l
01 0
60 fci
re cu
^H
•H CU
to 3
U
c ce
M
O CO
U li
4)
CO *J
3 CJ
0 S
3 -a
c re
•H ft)
u 5S '
C
O C
u to
o
U iJ
O M
•M V
C
CO O
V U
u
•H 41
*-" .c
U u
n
u u
cu o
lj .
*w
c
o cj
^H iU
4J o
«8 <
>
ki u
ft) O
U) CU
O •
a n
jj o
o
H V
*J U
C 0
4> £
o a
t_ n
V 0
a. c
6.

TJ
rl 4)
re >
*j re
0 (O
c
*> a>
f~ f«O
S 0
0 °
t- J->
£?5





•o
c a-
«4 ^>
^^ ^
re
c to
4>
60 C.
O «-
L. O
*> C
>d

*> c
c c
Scs
<-
t. t-
41 3
a. to



ft.
-H (.
«•" O
0 <
n
t_
jJ 4*
C Cw
4)
o -r
(. 4>
4) >
e. re
11




T>
ft) JJ
JJ 4J —
%$£
— o
*J !_
n c
W a.





o
— 4)
i> O
re —
• > *»
^ t. 0
on:


l -/
,tS
A3
H
n t.
c cu
o
o.




1 & O
| _ (V, pr-)
1 O m »-







1 A J3
1 in CO CO
1 =T •-

















1 £ jo
1 (^ 0 —
1 to tr\ —
















1
1 in ON a
i ON m —








o *x
* •-!
^ 0 —
o < re
t) CO U *s. 3
•^ — v. o c
z v> z r.' c
••» re



1 i
n TJ 4J
*» C t£
4) i re re
eo ex c
•-< o> t. -< t. re
re « > o t_ 4- jr
c 4) > *> eo >
OOjJOCOCOO)
—•—•CO — < O 3
*-» JJ 4) j_ CX •C
COc4.'3Q.t.-<
4>rere>oO4>n
>t_E_j>tJ->4)
CCX(_4JCOCK
O 4i O — «
« a. u 3s



•C £
f- a- e\j o «s CM M
PI m vo a z m co







J3 J3
O CO CJ a «c ON m
t\J m CO (\j Z f u a

















A JO
a o v£> (M «c ox CO
m in in a- z a t-


•













re
a- co M fo < a- ON
a \o i*— a1 Z vo co







• *™^ • • •
*> O *> 4-> JJ
h. < (JU U. U. 0
•N. x v. »^ v. «e
S«n o in m >•
t- o «\» ?- m
O U . fO . . . tsi
"•«. v. ^ v» _ _ —
Z 2 •*» -_x •* *» •**«»•


E
4> 4)
4) X3 C *J
CO e ••* O 85
v^re o>3: « xo>
eoi-i e -»-> c. *j v) ta
C^H 41 to«ft> o re
•* ~* *> X*->*> 4) X^
JH co M^.^( jj fl^.
O x l ~-« O 1 — •
<->T3 CO C O U. U f(-
CL. 4) o (M O. E 4>
O~*0> — < COQ. 4>«JT3
JJr-i3^O «JCcO-«B.«->r34>
C4>t3-ire i_^-»ret.re(r)io
4JC04)*>t. 0 -4 t. A> 4) x 3
e — ^it- >3Otoutox>-o
£ O 4> — < JJ o ft)
O z f- o to to ee










































.


























l
. D
n to eo -o
Sc re o
« -* c •»
re eo a re
v. c a. E -o
t. — O ex
4> L. t. 4) re «
fc> 3 CJ 3 J
O Q. T3 • t.
£ ij ^ _ o 4>
** n t. to c *i
— « o *-> 4> a.1 m
> O « J- Si 5























O
•i
fa
j
CO
f
w
0)
jC
^}
Q

JJ
4J
5

^
o;
jj
1C

o
0
M
t **
n
c
R
u
xj
n
E
«M
jj
n
4)

• to
^ £
n <
4) LJ
o e
-» u
t.
ex ^
c
o re
co
o> to
C 8
~H
r»*i
eo — '
o> —
— c
ff>
to" c
0 0
to
«< T?
V f
<: et
c c
to j=
T3
4>
• • jJ
4) C
U E
l_ -•
3 JJ
o n
to us
re jo
103

-------
 established.   Inasmuch  as  farmers,  who  have  signed  contracts with
 no  cost  sharing,   have  adopted  these  two  practices,  they are both
 cost   effective   since  thftir  assumed  annual  cost   is  zero;   i.e.
 benefits  to  farmers  are perceived  as  matching the costs.

      Structural   practices  are  not  applied on a  "per acre"   basis
 and are  slightly  more difficult to  evaluate.   However,   they are
 more   expensive than nonstructural  practices,  with the  exception
 of  permanent  vegetative cover.   The  stream  protection system,  or
 livestock  fence,   is designed to keep livestock  out of the  stream
 and thus  from adding nutrients  directly to waterway.   Savings  of
 nitrogen,  phosphorus,   and  fecal  coliform  bacteria from  fences
 would depend  on the  number  of animals and could  be  substantial.

      Other  practices   include   terraces,    diversions,   and  sod
 waterways.    Examining   the  costs  of the three   systems,  it  is
 evident   ,at  terraces are  much  more expensive.    Estimated   costs
 per  acre  for terraces are based  on  examination  of contracts  in
 the  project  .    However,   terraces do  retain nutrients  and  soil
 more  uniformly throughout  the field,   and are effective  in   those
 cases where   a   great  deal of  control  is needed  in  addition  to
 those obtained through  nonstructural  practices.

      Practices which   reduce runoff  on the  field  also   conserve
 valuable  nutrients for  crop production.   For example, looking  at
 Table 19,  the full set  of  BMPs  modeled  for continuous corn  silage
 can be installed  at  a one  time  cost of  $425  per  acre (which has  a.
 life  of  10 years  with proper  maintenance).    At  20  tons  of  manure
 per acre  per  year,   36  pounds of nitrogen per acre  could be saved
"each  year  at  a value of $10.00  ($0.28/lb. • N),  plus  the savings  of
 phosphorus,   potash,  and   organic  matter.    Without the cost  of
 expensive  structural practices, particularly terraces,  the bene-
 fits   of   nutrient and  organic  matter retention  on  the field  may
 outweigh   the  costs of  the  practices.   Reduced   tillage,  for
 example,   saves 23 pounds  total nitrogen  and 11  pounds phosphorus
 per acre  per  year when  manure is applied  at  30 tons per  acre.
                    WATER QUALITY DATA ANALYSIS
,i..\
      The  water quality results  from the project to date have come
     \ r 'i y  from the Zimmerman farm,  22 acre experimental site with
       additional   results  from the little  Conestoga  3700  acre
  r- l .-»r-s ' i1.   All of the water quality results should be considered
  »   '.:  :^MP implementation,   However, the documentation of spatial
  i'i  leiaporal  trends as woll as  thr-  coincidence of land activities
  i t '-   ob' erved water quality suspenses lend considerable  insight
  ot<-  • h?  potential effects  of BMPs  in the watershed.

      C;.sted  below  are some of the major observations  that  the
  t •. ,}pf;!.  monitoring personnel have reported.     These observations
   O   ^he'i  discussed in more detail  in relation to  interpretation
                                104

-------
and implications for R-.'^j-' program goals.  &** also present som« of
our.own analysis and i«t?rpretation of the water quality data.

1.   Of  the 43 wells in the regional network 67 percent  in  the
     carbonate areas exceeded 10mg/l.  nitrate,  while 27 percent
     in  non-carbonate  area exceeded the standard.   This  trend
     persisted  throughout  the year with  somewhat  higher  mean
     concentrations during the summer.

2.   Most of the nitrogen transported in surface waters is in the
     nitrate  form  during baseflow periods but organic  nitrogen
     predominates during storm flow.

3.   The   water   levels   in  field   site   monitoring   wells
     (carbonate  area)  respond  quickly  (i.e.  days)  to  major
     precipitation events.

4.   Suspended sediment runoff concentrations were lowest  during
     frozen ground and maximum crop cover conditions.

5.   The  highest  surface  runoff nitrogen  concentrations  were
     observed during snowmelt where the water-manure contact time
     was longer.

6.   Groundwater nitrate levels at the field site respond rapidly
     to the combination of manure application and precipitation.

     Our  own  analyses of  the project's  raw  data  corroborate
these findings.   As will be discussed subsequently each of these
results  has  important implications relating to the  effects  of
BMPs  and  the  potential for the project to  address  its  water
resource  impairments.   Our  analysis of the water quality  data
indicates  some  qualification  of  results  #3,  #5 and #6.   In
regard  to # 3 it should be noted that the field site  wells  not
only  respond quickly to precipitation events but also show large
rises in water table levels relative to the amount of  precipita-
tion.  For instance, a one-inch infiltration appears to raise the
water  table by about a foot.   This indicates a porosity of  the
groundwater  aquifer  of less than 0.1.   This would be a  common
value in many areas but is significantly lower than we would have
expected   for  this  karst  topography  which  clearly   has   a
preponderance of direct infiltration routes.

     The projects' analysis and interpretation of results  #5 and
:*C,   while  correctly  indentifying  trends  and  causal  factors
ro'.;l:?d  to  groundwater  nitrate  levels,    give  the   somewhat
misleading  impression  that  groundwater nitrate levels  show  a
i»r£e response to precipitation, snow-melt and manure application
p.-^ terns.    While  the trends summarized in # 5 and # 6 are  evi-
dent,  these  results should be placed into the context that  the
groundwater nitrate concentrations at the field site are actu.illy
fairly  constant,   and the responses to precipitation and  manure
spreading  are  relatively smnll.    For the  site  analyzed  most
thoroughly  the total range of nitrate concentrations was only 5-
1R  mg/1 and 80 percent of the observations fell between 10.0 and
                               105

-------
16.0  mg/1.    It  would h>sve been easy for the  project  to  have
dismissed this data »« showing no trends or causal relationships,-
so  it is to their creUit that they performed the level of analy-
sis needed to show that concentrations increased following manure
spreading combined with precipitation events.   However,   the im-
plication  of our analysis is that a substantial period  with  no
additional  manure  application  would probably be  required  for
nitrate levels to flush to a significantly lower level.

     The  project.  has  also attempted to  develop  a  regression
equation  which relates nitrate concentration to manure  applica-
tion and precipitation.  The purpose of the equation is to enable
prediction  of  future (post-BMP)  nitrate  concentrations.   The
regression  equation developed:  N03 = 0.035 Manure Load  - 0.073
water  level rise + 9.995,  while generating results of the  same
general  trend  as  observed results,  is probably not  the  best
regression equation to explain the effects of these two variables
on  nitrate levels.   First,  the statistical equation  indicates
that  water level change has a negligible  effect.   Second,  the
effect  of  recent manure applications appears to  be  underesti-
mated.   One  reason for this may be the use of the previous  120
day period for calculating manure application.   A shorter period
would  probably provide a better predictive  equation.   Also,   a
temporal staggering of water level and nitrate level would appear
to explain better the relationship of these two variables.
                           PROJECTIONS
                                                        *
     A  number  of interpretations of the present  water ' quality
data  and projections to the future effects of BMPs can be  made.
Some  of these have already been postulated by the project  water
quality monitoring personnel.  The following includes some speci-
fic  BMP projections based on the observed water qualit.,  results
to  date.   For  the most part these are fairly  similar  to  the
projections developed by ERS through the use of the CREAMS model.

1)   Terraces

     Terracing will greatly increase the time  of contact between
     applied manure and precipitation which infiltrates the soil.
     This  will increase the nitrate concentrations of  infiltra-
     ting  water.    In addition a slightly  larger percentage  of
     precipitation will reach groundwater. The combined effect is
     an  increase  on groundwater  nitrate  concentratiins,    The
     terraces  would be expected to increase  dissolved  nutrient
     concentrations in surface runoff but somewhat decrease total
     surface  nutrient losses.   Sedimenf loads would be  reduced
     -significantly (70-80%).

2)   Animal Waste Management

     The   pre-BMP  management  strategy  is  to   spread  manure
     daily   when   field  conditions   permit.   Generally  this
     means  that  no  spreading is done  during  June,  July  and
                               106

-------
      August   when  e*»rfl is  in  the field.    The animal  waste  BMP
      involves 6 month  open  storage so that  manure is applied when
      crops   can make  maximum use of the nutrient (i.e.   May  for
      field  crops and September for cover crops).

      The  projected effect  is that  ground  and  surface  water
      nitrate   concentrations would increase over present  values
      for  the periods following application  and would be lower the
      rest of the year.   Hence,  it is probably the range of values
      which  will be most  affected.  The loss of nutrients to water
      sources  would be reduced by more timely plant  uptake,   but
      this would be counteracted by the increased  nutrient content
      of stored versus  daily-spread manure.

 3)    Nutrient Management

      We   see  nutrient manag. "ent in this project as having  four
      basic  components:

      a)   Soil  and manure  testing to insure that no  unnecessary
          chemical fertilizer is used.

      b)   Growing  crops  which use rather than fix nitrogen  (i.e.
          corn) .

      c)   Applying  manure  nutrients  when  they can best be  used
          by the crop.

      d)   Exjporti-ng  manure  where nutrient supply  greatly  exceeds
» .        crop needs.

      The  nutrient management BMP is projected to greatly  reduce
 both  surface and groundwater  nutrient concentrations.     Fluctua-
 tions wo.uld still be  observed as a function of application dates
 and precipitation  events.   However,   the mean concentration   and
 the   frequency  of  nitrate values greater  than 10mg/l   would  be
 reduced.   We  estimate  that  the reduction  would  be greater  than
 the percentage reduction in nutrients applied because  of a closer
 match of  application and plant  usage rates.
             Considerable  monitoring  for  herbicides  in  surface  and
 groundwater  is  being  conducted  in  the  project.   Some analysis  and
 interpretation   of  this   data   has  been   done   by  the   project
 including  tabular  and  graph4 ?  presentations  of  the concentration
 la»:a.

     Our  inspections and analysis of  the  data  reveal the  following
 i nf ornnn t i on   relevant  to  pesticide  surface   and   groundwater
 tin "sport  in the project  area.

 1.   Herbicide   concentrations  in  groundwater at  the field  sites
     rise    to    barely   above   detectable    levels    following
     appl icat ion .

 2.   Herbicide  concentrations  in groundwater  throughout  the   pro-


                                107

-------
     ject  area are surprisingly low (generally less than 1 ug/1)
     given the amount of usage and the karst topography.

3.    Herbicide  concentrations  in stream baseflow are also  very
     low (less than 1 ug/1)  but show small though clear increases
     during the summer months.

4.    Moderate concentrations of alachlor (2-5 ug/1),  atrazine(7-
     17  ug/1),  cyanazine (7-16ug/l) and metolachlor  (l-18ug/l)
     were  observed during a storm event in the Little  Conestoga
     basin following spring herbicide application.

5.    High concentrations (greater than 50 ug/1) were observed  in
     surface runoff from the field site during first runoff event
     following  application.   Elevated concentrations appear  to
     persist for about 2-3 months following application.

6.    The  surface water pesticide data from the Little  Conestoga
     Basin  strongly  suggest  that there is  another  source  of
     herbicide  input beside field application (elevated  concen-
     trations  yeai—round and consistent observation of  elevated
     levels  of  herbicides no longer in  common  use).   Further
     investigation revealed the presence of a pesticide  disposal
     area in the upper part of the watershed.  Further investiga-
     tion of the impacts of this source are underway.

                          IMPLICATIONS

     It is possible,  although unlikely, that nutrient management
and- other  BMPs will be implemented to an extent  sufficient  to
significantly reduce area-wide surface and groundwater  pollutant
loads within the project timeframe.   However,  since it has been
shown  that  shallow  groundwater is recharged by the  land  area
immediately  (with a few hundred feet) up-gradient,  BMP  ground-
water  relationships  for  the project can  be  established  with
relative accuracy from the field sites.

     Thus,  at  this point it appears that the projects'   primary
contributions  to  overall BMP water quality  understanding  will
come  from the field sites and possibly from the 3700 acre  small
watershed  site.   Relevant information from the overall  project
will  probably relate primarily to economic,  social and institu-
tional factors which affect the success of the voluntary  project
Approach.  These contributions are summarized below:

1.    To   gain  farmer  participation and to  select  appropriate
     BMPs  and critical areas a project needs to decide early  on
     which water resource impairment is to be given top priority.
     This is particularly true when both groundwater and  surface
     water impairments exist.   All subsequent project activities
     need to be consistent with this decision.

.2    A  50  percent cost-share rate for animal  waste  management
     structures will probably be inadequate  to gain farmer parti-
     cipation in areas with excess manuria? nutrients.
                               108

-------
3.   Land conservation contracts which are developed to meet  'T'
     may  not  ke a cost-efficient method of  meeting surface  or
     groundwater quality goals.

4.   There  may  be inherent trade-offs between BMPs designed  to
     improve surface water and groundwater quality.

5.   In  an  area   where subsurface flow  re-emerges  as  stream
     baseflow,  BMPs which reduce surface runoff losses of nitro-
    .gen  will generally increase stream baseflow nitrogen  loads
     proportionately.

6.   The project illustrates what may be a very common  situation
     which is that pesticide disposal practices and sites consti-
     tute  at  least  as great a source of  pesticides  as  field
     applications.

7.   The application timing advantages provided by manure storage
     versus daily spreading are partially or completely nullified
     by increased manurial nutrient availability.

8.   Groundwater  nitrate  concentrations  from  the  field  site
     increased  significantly  in  response  to  periodic  manure
     applications.

9.   Precipitation  infiltration,  which raises the water  table,
     can  have  either  an  increasing or  decreasing  effect  on
     groundwater  nitrate  concentrations depending on  the  time
     interval  since  manure application and on the  quantity  of
     manure  applied.    A common situation observed at the  field
     site  is that precipitation first has a diluting  effect  on
     nitrates.   However,  as slower percolating water,  which has
     had   longer soil nutrient contact time,   reaches the  water
     table, nitrate concentrations increase.

10.   Used properly as  a BMP, nutrient management reduces nitrogen
     inputs to both surface and groundwater.
                               .109

-------
             Oakwood Lakes - Poinsett,  South Dakota

                             RCWP 20


                          INTRODUCTION
     The Oakwood Lakes - Poinsett RCWP is a 106,000-acre  project
located in the glacial lakes region of east .central South Dakota.
The  project  area includes about 80 percent crop and grass  land
with the balance state/federal lands or lakes.   Area soils are a
mix  of relatively impermeable till and  less  interspersed  with
highly  permeable areas of sand and grav 1 outwash.   Use impair-
ments have been documented for groundwater affected by  excessive
nitrate  levels and for eutrophic recreational lakes that receive
excessive  plant nutrients and sediment.   Both impairments  have
been  attributed  to loss of soil and fertilizer  nutrients,  from
cropland and grasslands.

       The  impaired surface waters of the project include  three
large lakes and a number of smaller lakes.   The lakes are gener-
ally shallow, average depth from 4 to 10 feet.  The three largest
lakes  are Lake Poinsett,  7,868 acres draining 32,452 acres  (83
percent  cropland),  Oakwood Lake,  2,184 acres  draining  52,856
acres  (5Q% cropland),  and Lake Albert,  2400 acres,  within the
drainage of Lake Poinsett. '  Average depth of Lake Albert is only
4 feet.   These lakes have very high recreational value which  is
impaired primarily by eutrophication.

Perspectives of the Project

     Because  this  project has both documented  groundwater  and
surface water impairments, there are four general questions to be
answered analytically:

1.   What  BMPs are most cost-effective for reducing the  impair-
     ment of    groundwater by agricultural activity?

2.   What  BMPs are most cost-effective for reducing the  impair-
     ment of surface water  by agricultu %al activity?
     To  what  extent do BMPs designed to protect
     exacerbate groundwater impairments0
4-
What  will  be  the
quality  impairment
efficiently?
economic benefits  from
and  how can  *;hese  be
                                              surface  water
reduced  water-
achieved  most
     fc  date,  the project has developed a monitoring program  to
investigate the impact of agricultural management on  groumhv >' er
but" -as not monitored surface water    Therefore,  there has 'oecn
                               110

-------
no contribution from tJ-;i.^ &>o ject toward the answers to questions
2  or 3.   A monitoring effort to evaluate the overall effective-
ness  of  the  RCWP  in reducing  the  impairment  of  the  water
resources  has not been attempted because there are no continuous
streams in the project area, and the subsurface hydrology is very
complex.  Specifically, the  interconnections  of  the  surficial
sand  and  gravel aquifers with the  major  groundwater  resource
of  the  area,  the  Big Soux aquifer,  or with  the  lakes,  are
generally unknown.

     Project  monitoring efforts have concentrated on  evaluating
the  relative  effectiveness of management practices in  reducing
nitrate and pesticide leaching.

     Specific questions addressed by monitoring include:

1.   Do conservation t;llage  (no till or chisel plowing)  reduce
     the downward flux of nitrate-N?

2.   Does  residue management affect the nitrogen content of  the
     soil  profile through  denitrification,   leaching,   or  crop
     uptake?

3.   Does  the quality of groundwater beneath agricultural  lands
     with   conservation  tillage  or  fertilizer  and  pesticide
     management differ from that of areas with conventional farm-
     ing practices?

4-   What is the relationship between soil characteristics  (gla-
     cial  till  or'  sand  and gravel  'outwash) .  and  groundwater
     quality under alternative agricultural management practices?

5.   Is  the  quality of groundwater  beneath  agricultural  land
     different  from t..iat of a control area where there has never
     been any agricultural activity?

     Operationally,  the  monitoring component of the project  is
investigating  whether or not agricultural BMPs can  prevent  ni-
trate  and  pesticide contamination of groundwater.   Plans  have
been  developed to instal.1 runoff gaging and sampling devices  at
each  groundwater monitoring site to look  at trade-offs  between
protection of surface and subsurface waters,  but these plans have
not  been implemented.    Although the land treatment component of
f:h.^  project  is concerned with reducing  sediment  and  nutrient
-:. •  inGport to the lakes,  no clear experiment has been developed to
     > I-rat e this type of effect.

     Treatment Strategy

     before the project was initiated, 52 percent of the crop and
     lind was considered adequately treated for erosion.    There-
fore,   the  focus  of the RCWP land treatment effort has been  on
reducing the contamination of ^roundwater by fertilizer  nitrogen
and  preventing  their contamination by pesticides.  Because  the
'»ii<~facial aquifer occurs irregularly throughout the project area,
                               111

-------
no contribution fro* ti>i<  project toward the answers to questions
2  or 3 .    A monitoring effort to evaluate the overall effective-
ness  of  the  RCWP  in reducing  the  impairment  of  the  water
resources  has not been attempted because there are no continuous
streams in the project area,  and the subsurface hydrology is very
complex.   Specifically, the  interconnections  of  the  surficial
sand  and  gravel aquifers with the  major  groundwater  resource
of  the  area,  the  Big Soux aquifer,   or with  the  lakes,  are
generally unknown.

     Project  monitoring efforts have concentrated on  evaluating
the  relative  effectiveness  of management practices in  reducing
nitrate and pesticide leaching.

     Specific questions addressed by monitoring include:

1.   Do conservation t'llage   (no till  or chisel plowing)  reduce
     the downward flux of nitrate-N?

2.   Does  residue management affect the nitrogen content of  the
     soil  profile through  denitrif ication ,  leaching,  or  crop
     uptake?

3.   Does  the quality of groundwater beneath agricultural  lands
     with   conservation  tillage  or  fertilizer  and  pesticide
     management differ from that of areas with conventional farm-
     ing practices?

4.   What is the relationship between soil characteristics  (gla-
     cial  till  or  sand  and gravel  outwash)  and  groundwater
    "quality under alternative agricultural management practices?
          *                   •
5.   Is  the  quality of groundwater  beneath  agricultural  land
     different  from t-iat of  a control  area where there has never
     been any agricultural activity?

     Operationally,  the  monitoring component of the project  is
investigating  whether or not agricultural BMPs can  prevent  ni-
trate  and  pesticide contamination of groundwater.   Plans  have
been  developed to install runoff gaging and sampling devices  at
each  groundwater monitoring  site to look  at trade-offs  between
protection of surface and subsurface waters, but these plans have
not  been implemented.   Although the land treatment component of
bh^  project  is concerned with reducing  sediment  and  nutrient
I; Disport to the lakes, no clear experiment has been developed to
      t- rat e this type of effect.

               Strategy
     before the project was initiated, 52 percent of the crop and
     land was considered adequately treated for erosion.   There-
fore,  the  focus  of the RCWP land treatment effort has been  on
     ing the contamination of groundwater by fertilizer  nitrogen
     preventing  their contamination by pesticides.  Because  the
.•snrficial aquifer occurs irregularly throughout: the project area,
                               111

-------
no contribution from thi* {••• oject toward ihe answers to questions
2  or 3.   A monitoring effort to evaluate the overall effective-
ness  of  the  RCWP  in reducing  the  impairment  of  the  water
resources  has not been attempted because there are no continuous
streams in the project area, and the subsurface hydrology is very
complex.  Specifically, the  interconnections  of  the  surficial
sand  and  gravel aquifers with the  major  groundwater  resource
of  the  area,  the  Big Soux aquifer,  or with  the  lakes,  are
generally unknown.

     Project  monitoring efforts have concentrated on  evaluating
the  relative  effectiveness of management practices in  reducing
nitrate and pesticide leaching.

     Specific questions addressed by monitoring include:

1.   Do conservation tillage  (no till or chisel plowing)  reduce
     the downward flux of nitrate-N?

2.   Does  residue management affect the nitrogen content of  the
     soil  profile through  denitrif ication ,   leaching,   or  crop
     uptake?

3.   Does  the quality of groundwater beneath agricultural  lands
     with   conservation  tillage  or  fertilizer  and  pesticide
     management differ from that of areas with conventional farm-
     ing practices?

4.'   What is the relationship' between soil characteristics  (gla-
     cial  till  or  sand  and 'gravel  outwash)  and  groundwater
     quality under alternative agricultural . management practices?

5.   Is  the  quality of groundwater  beneath  agricultural  land
     different  from t.iat of a control area where there has never
     been any agricultural activity?

     Operationally,  the  monitoring component of the project  is
investigating  whether or not agricultural BMPs can  prevent  ni-
trate  and  pesticide contamination of groundwater.   Plans  have
been  developed to install runoff gaging and sampling devices  at
each  groundwater monitoring site to look  at trade-offs   between
protection of surface and subsurface waters,  but these plans have
not  been implemented.    Although the land treatment component of
thf  project  is concerned with reducing  sediment  and  nutrient
'•, •. t.jsport to the lakes   no clear experiment has been developed to
            this type of effect.
     Treatment Strategy

     •Jefore the project was initiated, 52 percent of the crop and
     land was considered adequately treated for erosion.   There-
fore,   the  focus  of the RCWP land treatment effort has been  on
rednr.ug the contamination of groundwater by fertilizer  nitrogen
and  preventing  their contamination by pesticides.   Because  the
          aquifer occurs irregularly throughout the project area,
                               111

-------
no contribution from thi* f-foject toward the answers to questions
2  or 3.   A monitoring effort to evaluate the overall effective-
ness  of  the  RCWP  in reducing  the  impairment  of  the  water
resources  has not been attempted because there are no continuous
streams in the project area, and the subsurface hydrology is very
complex.  Specifically, the  interconnections  of  the  surficial
sand  and  gravel aquifers with the  major  groundwater  resource
of  the  area,  the  Big Soux aquifer,  or with  the  lakes,  are
generally unknown.

     Project  monitoring efforts have concentrated on  evaluating
the  relative  effectiveness of management practices in  reducing
nitrate and pesticide leaching.

     Specific questions addressed by monitoring include:

1.   Do conservation tillage  (no till or chisel plowing)  reduce
     the downward flux of nitrate-N?

2.   Does  residue management affect the nitrogen content of  the
     soil  profile through  denitrification,  leaching,  or  crop
     uptake?

3.   Does  the quality of groundwater beneath agricultural  lands
     with   conservation  tillage  or  fertilizer  and  pesticide
     management differ from that of areas with conventional farm-
     ing practices?

4.   What is the relationship between soil characteristics  (gla-
     cial  till  or  sand  and gravel  outwash)  and  groundwater
     quality und*er alternative agricultural management practices?

5.   Is  the  quality of groundwater  beneath  agricultural  land
     different  from t.iat of a control area where there has never
     been any agricultural activity?

     Operationally,  the  monitoring component of the project  is
investigating  whether or not agricultural BMPs can  prevent  ni-
trate  and  pesticide contamination of groundwater.   Plans  have
been  developed to install runoff gaging and sampling devices  at
each  groundwater monitoring site to look  at trade-offs  between
protection of surface and subsurface waters, but these plans have
not  been implemented.   Although the land treatment component of
thr>  project  is concerned with reducing  sediment  and  nutrient
•!. •: lusport to the lakes  no clear experiment has been developed to
 lomo-strate this type of effect.

Oanci Treatment Strategy

     Before the project was initiated, 52 percent of the crop and
a'ra»s"iind was considered adequately treated for erosion.   There-
fore,  the  focus  of the RCWP land treatment effort has been  on
reducing the contamination of groundwater by fertilizer  nitrogen
*nd  preventing  their contamination by pesticides.  Because  the
-."rficial aquifer occurs irregularly throughout the project area,
                               111

-------
 no  contribution  from  t Js i ?•-  /-s eject  toward the answers  to questions
 2   or  3.    A  monitoring  effort  to  evaluate  the overall  effective-
 ness   of   the RCWP   i t»  reducing  the  impairment   of  the  water
 resources   has not been  attempted  because  there are no  continuous
 streams in  the project area,  and the  subsurface hydrology is very
 complex.   Specifically,  the   interconnections   of   the   surficial
 sand   and   gravel aquifers with  the   major   groundwater  resource
 of  the  area,   the   Big Soux aquifer,   or  with  the   lakes,   are
 generally  unknown.

     Project  monitoring efforts have concentrated  on  evaluating
 the  relative effectiveness  of  management  practices  in  reducing
 nitrate and pesticide  leaching.

     Specific questions addressed  by  monitoring include:

 1.   Do conservation  t;llage  (no  till  or chisel plowing)   reduce
     the downward flux of nitrate-N?

 2.   Does  residue management affect  the nitrogen content  of   the
     soil  profile through  denitrification,   leaching,   or   crop
     uptake?

 3.   Does  the quality of groundwater beneath  agricultural   lands
     with   conservation  tillage  or   fertilizer   and   pesticide
     management differ from that of areas with  conventional  farm-
     ing practices?

 4-   What is  the relation-ship between soil characteristics   (gla-
     cial  till  or  sand  and gravel   outwash)  and  groundwater
     quality  under alternative agricultural management  practices?

 5.   Is  the  quality of groundwater  beneath   agricultural   land
     different  from t-iat of  a control  area where there  has never
     been any agricultural activity?

     Operationally,   the  monitoring  component  of the project  is
 investigating  whether or not agricultural BMPs can   prevent   ni-
 trate  and  pesticide contamination of  groundwater.   Plans   have
been  developed to install runoff  gaging and sampling devices  at
each  groundwater monitoring site  to  look  at trade-offs  between
protection of surface and subsurface waters, but these  plans  have
not  been implemented.   Although  the land treatment  component of
 thp  project  is  concerned with  reducing  sediment  and  nutrient
 ;. i lusport  to  the  lakes, no clear experiment has been  developed to
     •> t rate this  type of effect.

     Treatment Strategy

     Before the project was initiated,  52 percent of  the crop  and
     land  was considered adequately • trea ted for erosion.   There-
fore,   the  focus  of the RCWP land treatment effort  has been  on
reducing the contamination of groundwater by fertilizer  nitrogen
ami  preventing  their contamination by pesticides.    Becausfe   the
 -,"rficial  aquifer occurs  irregularly throughout the project area,
                               111

-------