PBS6-173630
THE LUBBOCK LAND TREATMENT SYSTEM RESEARCH AND
DEMONSTRATION PROJECT.  VOLUME V:  EXECUTIVE SUMMARY
Lubbock Christian College
Lubbock, TX
Feb 86
                  U.S. DEPARTMENT OF COMMERCE
               National Technical Information Service

-------
&EPA
           United States
           Environmental Protection
           Agency
           • Robert S. Kerr Environmental
           Research Laboratory
           Ada OK 74820
EPA/600/2-86/027e
February 1986
           Research and Development
The Lubbock Land
Treatment System
Research and
Demonstration
Project:

Volume V.
Executive Summary

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing}
 1. REPORT NO.
   EPA/600/2-86/027e
                             2.
             3. RECIPIENT'S ACCESSIOf*NO.
                       173jb30/J.S
4. TITLE AND SUBTITLE
 THE LUBBOCK LAND TREATMENT SYSTEM RESEARCH
 AND DEMONSTRATION PROJECT
 VOLUME  V:   EXECUTIVE SUMMARY
             5. REPORT DATE
               February 1986
             6. PERFORMING ORGANIZATION CODE
            >B>  Georgej N-L> Altman, D.E.  Camann,. B.J.
 Claborn,  P.J.  Graham, M.N. Guntzel, H.J.  Harding, R.B.
 Harrist.  A.H.  Holauin. K.T. Kimball. N.A.  Klein, cont-lii
                                                           8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
 Lubbock Christian College Institute of
  Water Research
 Lubbock,  TX   79407
             10. PROGRAM ELEMENT NO.

             CAZB1B
             11. CONTRACT/GRANT NO.
             CS-806204 and
             CR-807501
12. SPONSORING AGENCY NAME AND ADDRESS
 U.S. Environmental  Protection Agency
 Robert S.  Kerr Environmental Research  Laboratory
 P.O. Box  1198
 Ada, OK   74820
             13. TYPE OF REPORT AND PERIOD COVERED
             Final  11/27/78 - 12/31/85
             14. SPONSORING AGENCY CODE
               EPA/600/15 .
15.SUPPLEMENTARY NOTES D.B. Leftwich, R.L. Mason,  B.E.  Moore, R.L.. Northrup,  C.  Becke^
 Popescu,  R.H. Ramsay, C.A. Sorber, R.M.  Sweazy
 Project Officers;   Lowell E. Leach and Walter  Jakubowski	
16. ABSTRACT
      The  Lubbock Land Treatment System  Research and Demonstration Project,  funded
 by Congress  in  1978 (H.R. 9375), was designed to address the various  issues
 concerning the  use of s-low rate land application of municipal wastewater.   The
 project involved the 1) physical expansion  of overloaded 40 year old  Lubbock
 slow rate land  treatment system; 2) characterization of the chemical,  biological
 and physical  conditions of the ground water,  soils and crops prior  to  and  during
 irrigation with secondary treated municipal wastewater; 3) evaluation  of the.
 health effects  associated with the slow rate  land application of secondary
 effluent  and  4) assessment of the effects of.  hydraulic, nutrient and  salt  mass
 loadings  on  crops, soil and percolate.
17.
                               KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS  C. COSATI Field/Group
13. DISTRIBUTION STATEMENT

 RELEASE TO  PUBLIC
19. SECURITY CLASS (This Report)

 UNCLASSIFIED	
21. NO. OF PAGES

  125-
                                              20. SECURITY CLASS (This page)

                                               UNCLASSIFIED
                                                                        22. PRICE
EPA Form 2220-1 (9-73)


-------
                                                        EPA/600/2-86/027e
                                                        February 1986
                      THE LUBBOCK LAND TREATMENT SYSTEM
                     RESEARCH AND DEMONSTRATION PROJECT

                                  VOLUME V

                              Executive Summary
                                     by

        D. B. George1, N.  L. Altman5, D.  E. Camann2, B.  J.  Claborn7,
               P. J. Graham5, M. N. Guntzel\  H.  0. Harding2,
        R. B. Harrist6. A. H. Holguin6, K.  T.  Kimball2,' N.  A. Klein1,
        D. B. Leftwichl, R. L. Mason2, B. E. Moore5, R.  L.  Northrup5,
       C. Becker Popescu5, R. H. Ramsey7, C. A. Sorbet5, R. M. Sweazy7


1Lubbock Christian College Institute of Water  Research,   Lubbock, TX 79407
            2Southwest Research Institute,  San Antonio,  TX 78284
            ^University of Illinois at Chicago, Chicago, IL 60680
         ^University of Texas at San Antonio,  San Antonio,  TX 79285
              ^University of Texas at Austin,  Austin, TX 78712
       6University of Texas School of Public Health, Houston, TX 77025
                  7Texas Tech University, Lubbock, TX 79409
                      EPA COOPERATIVE AGREEMENT CS806204
                             and CR807501


                              Project Officers

                               Lowell E. Leach
                              Jack L. Witherow
                              H. George Keeler
                              Curtis C. Harlin
                         Wastewater Management Branch
                R. S. Kerr Environmental Research Laboratory
                             Ada, Oklahoma 74820

                              Walter Jakubowski
                    Toxicology and Microbiology Division
                     Health Effects Research Laboratory
                            Cincinnati,Ohio 45268
              ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
                     OFFICE OF RESEARCH AND DEVELOPMENT
                    U.S. ENVIRONMENTAL PROTECTION AGENCY
                             ADA, OKLAHOMA 74820
                                     /•a-

-------
                       DISCLAIMER
     The  information in this document has  been funded in
part by  the  United States Environmental Protection  Agency
under assistance  agreement No. CS806204  to  the Lubbock
Christian  College Institute of Water Research and  by the
Health  Effects Research Laboratory, United States Environ-
mental Protection Agency  under  CR-807501 to Southwest
Research  Institute.  It has been subjected to the Agency's
peer and administrative review and has been  approved for
publication as an EPA document.  Mention of trade names or
commercial products  does not constitute  endorsement or
recommendation for use.
                          11

-------
                                 FOREWORD
     The U.S. Evironmental Protection Agency  was  established to coordinate
the administration of major  Federal  programs  designed to protect the qual-
ity of our environment.

     An important  part   of  the  Agency's  effort  involves the  search  for
information about. environmental  problems,  management" techniques,  and  new
technologies through  which   optimum  use  of  the  Nation's  land  and  water
resources can be  assured and the threat pollution poses  to  the welfare of
the American people can  be minimized.

     The U.S.  Environmental   Protection  Agency's  Office  of Research  and
Development conducts this search through  a nationwide network  of research
facilities.  As  one  of  these facilities,  the Robert  S.  Kerr Environmental
Research Laboratory is responsible for the management of programs including
the development and demonstration of soil  and  other natural systems for the
treatment and management of  municipal wastewaters.

     The slow rate land  treatment process  of municipal wastewaters uses the
unsaturated soil  profile and agricultural  crops  managed  as  the treatment
media.  The Lubbock  Land Treatment  System  Research  and  Demonstration Pro-
gram, funded by Congress in  1978  (H.R.  9375) was designed  to  address the
various issues  limiting  the   use  of  slow  rate  land application of municipal
wastewater.  The  project  involved  expansion of the  Lubbock  Land Treatment
System to 2,967  hectares; characterization  of the chemical,  biological and
physical condition of the ground water,  soils  and crops prior to and during
irrigation with  secondary treated  municipal wastewater;  and  evaluation of
the U.S.  Environmental  Protection  Agency's design  criteria  for  slow rate
land application.  Results  demonstrate  that,  where  such  systems  are cor-
rectly designed  and  operated,  they can  be  cost  effective alternatives for
municipal sewage treatment at sites  where  conditions  are favorable for low
hydraulic loading combined with cropping practices.

     This report  contributes  to  the knowledge which  is  essential  for the
U.S. Environmental Protection Agency to meet  requirements of environmental
laws and  enforce pollution   control  standards which  are  reasonable,  cost
effective and provide adequate protection  for  the American public.
                                    Clinton W. Hall, Director
                                    Robert S.  Kerr Environmental  Research
                                      Laboratory
                                    111

-------
                                ABSTRACT

     The Lubbock Land  Treatment System consists of two privately owned
farms.   The  Gray  farm comprises  1,489 ha and  has been reusing  treated
wastewater  for crop irrigation for  more  than 40  years.   In  1981  the land
application system was enlarged to  include the Hancock farm  which  had 1,153
ha under cultivation.  The primary  irrigation mode employed  by  both farms
was spray  irrigation using center pivot irrigation machines.   The  Lubbock
Land Treatment system Research and  Demonstration  Project involved  the 1)
physical expansion of the Lubbock Land Treatment System; 2) characteriza-
tion of the chemical, biological and physical  conditions  of  the  ground
water,  soils, -and crops  prior  to and during irrigation   with  secondary
treatment municipal wastewater; 3)  evaluation  of health effects of slow
rate land  application of  secondary effluent;  and 4)  assessment  of the
effects of  hydraulic, nutrient and  salt  mass loadings on crops,  soil  and
percolate.
     During the period when a portion of the treated wastewater  was  divert-
ed to the Hancock  farm, a decrease  in the ground-water level beneath  the
Gray farm  was  measured.   In  conjunction with the lowering of  the  ground-
water table was an increase in water  quality beneath most  of the farm (pri-
marily  the ground water underlying the  spray irrigated areas).   The  culti-
vation  of' alfalfa  in the spray irrigated areas  was probably the primary
factor  affecting the quantity and quality of percolate.
     Chemical and  nutrient constituents  in the treated wastewater applied
to the  Hancock  farm were removed  by the soil-crop matrix.   An  increase in
ground  water beneath the Hancock farm resulted  from deep  percolation of
surface  runoff  collected  in  moats  surrounding the reservoirs  and  excava-
tions constructed  to reduce flooding of crop land.  Deep  percolation of
surface  runoff  leached existing nitrate and salt deposits  within the soil
profile  to the  ground water;  thereby, causing increased ground-water
nitrate and total  dissolved solids  concentrations.
     The epidemiological study conducted on the populace  in  and  surrounding
the Hancock  farm  indicated  that  wastewater spray irrigation  produced no
                                 IV

-------
obvious disease during the  project period.   However, the rate  of viral
infections was  slightly higher among  participants who had a high degree of
aerosol exposure.  The polio  virus 1 infections during spring 1982  were
probably  related to this.exposure.
     Agricultural  studies showed that  cotton and grain sorghum  produced
greater yields with  increasing annual  hydraulic  loading rates  up to
3m.ha/ha.yr.   The highest alfalfa  yields  were obtained in test  plots irri-
gated with 365  and 434 cm.ha/ha.yr.   The  alfalfa test  plots appeared to
remove all  nutrients applied in the  wastewater stream.  Salts were leached
beyond 91  cm of soil in all plots receiving 60 cm.ha/ha.yr or greater.
     The Lubbock  Land Treatment System  Research and Demonstration Project
was conducted by Lubbock  Christian  College  Institute  of  Water  Research
(LCCIWR),  Southwest Research Institute (SwRI), University of Illinois  (UI)
University of Texas at San Antonio  (UTSA),  University of  Texas at Austin
(UT),  and Texas Tech University (TTU).   This report  was submitted in  ful-
fillment of CR807501 and CS806204 by   LCCIWR under primary  sponsorship of
the  U.S.  Environmental Protection  Agency.  This report covers a summary of
research  activities performed from  May 1,  1980 through December 31,  1983.
This work  was completed on June  30, 1985.

-------
                                 CONTENTS

Foreword	iii
Abstract	iv
Figures	viii
Tables	    ix
   1.  Introduction	1
          Background	  2
          Description of Land Application System, Expansion  	  2
          Effluent Quality	  6
          System Operation 	  13
   2.  Conclusions	19
   3.  Project Design	-.25
          Demonstration/Hydrogeologic Study  	  25
          Lubbock Infection Surveillance Study 	  29
          Percolate Investigation in the Root Zone	33
          Agricultural Research Studies  	  37
   4.  Summary of Findings
       Demonstration/Hydrogeologic Investigation . 	  41
       Lubbock Infection Surveillance Study  . 	  69
       Agricultural Research Studies 	 	  81
References	88
Appendices
   A.  Agricultural Cropping Patterns	90
   B.  Supplemental Figures and Tables for Section 3 	  94
   C.  Supplemental Figures and Tables for Section 4 	101
                                                   Preceding page blank

-------
                               LIST OF FIGURES
Number	Name	Page
1
2
3
4
5

6 '
7

8
9
10
11
Southeast Water Reclamation Plant Flow Diagram 	
Hancock Farm Hydraulic Distribution System 	
Precipitation During Project Period 	
Gray Farm Land Application Site 	
LISS Study Design: Time Frame of Monitoring in Relation to
Major Periods of Irrigation 	 '. .
Plan of Test Facility . 	
Nitrate Concentration (mg/1) in Well Water under Gray Farm,
Baseline Period, 1981-1982 	 ' 	
Illustration of Nitrite+Nitrate Lenses in Hancock Soil, 1981.
Inorganic Nitrogen in 183 cm Profile at the Hancock Farm. . .
Sampling Zones Comprising Study Area 	
Variation of Nitrate Concentration in Percolate and Accumu-
3
5
16
17

31
35

45
61
62
70

           lated Weight of Leached Nitrate in kg/ha with Time for Tube
           123 from July through December 1982'	   79
 12        Nitrogen Mass Balance for Trial 14000 Cotton Plots	   83
 13        Nitrogen Mass Balance for Trial 16000 Alfalfa Plots 	   85
 14        Soybean Seed Yield vs Hydraulic Loading - Trial 17000 ....   86
 15        Milo Whole Plant Yield vs Hydraulic Loading - Trial 17000 . .   86
                                    vm

-------
                               LIST OF TABLES
Number	Name	Page
  1        Characterization of Effluent Produced by Southeast Water
           Reclamation Plant in 1980 and 1981	   7
  2        Concentration of Trace Elements in Treated Wastewater ....   9
  3        Microorganism Concentrations in Wastewater Applied by
           Sprinkler Irrigation	  12
  4        Bacterial Screen—Hancock Reservoir	12
  5        Total Water Applied to Hancock Farm in 1983	15
  6        Gray Farm Hydraulic Loadings/Crop 	  18
  7        Type and Number of Underground Water Sampling Points by Site.  26
  8        Treatment Matrix Hydraulic Loading Rate for Trial 17000 ...  39
  9        Statistics of Depth to Water in Observation Wells at Gray
           Site During Project	42
 10        Percent of Gray Farm Well Water Samples Which Exceed or Equal
           Drinking Water Standards for the Following Parameters ....  44
 11        Trace Metals Mass Balance on Soils Collected from Flood
           Irrigated Area	51
 12        Statistics of Depth to Water in Observation Wells at Hancock
           Site During Project	54
 13        Percent of Hancock Farm Well Water Samples Which Exceed or
           Equal Drinking Water Standards for the Following Parameters .  55
 14        Metals Mass Balance for Hancock Farm	64
 15        Cotton Yields, Hancock Farm 	  67
 16        Elemental Shifts in Cotton Tissues Obtained from Hancock
           Farm 1981 vs 1983	  68
 17        Comparisons of Shift.in Farmers' Income Pre-effluent to
           Post-effluent 	  68

-------
                                SECTION 1

                               INTRODUCTION
     Agriculture is the  major  user of freshwater in the United States with
approximately 99 percent  of  the  agricultural water demand used for irriga-
tion  (Williams, 1982).   Increasing water demands by agriculture, industry
and municipalities  have created  severe water shortages  in various regions
of the United States and  the  world.  Application of municipal wastewater  to
agricultural lands  is a viable  alternative  to reduce  the  withdrawal  of
freshwater from surface  water  and ground-water sources.  In addition, land
application of wastewater is  a  cost-effective treatment alternative.   Slow
rate  wastewater application,  usually in the  form of spray irrigation,  is
the most widely used form of  land application.
     The Lubbock Land  Treatment System Research and Demonstration Program,
funded'by Congress  in 1978 (H.R. 9375), was designed to address the various
issues concerning the use of  slow rate land application of municipal  waste-
water.  The project involved  the 1) physical expansion of the Lubbock  Land
Treatment System; 2) characterization of the chemical, biological and phys-
ical conditions of the  ground water, soils and  crops prior  to and  during
irrigation with secondary   treated municipal wastewater; 3) evaluation  of
the health  effects  associated with the slow rate land application of secon-
dary  effluent; and 4)  assessment of the effects of hydraulic, nutrient and
salt mass loadings  on crops,  soil and percolate.  Results from the Lubbock
Land  Treatment  Research and Demonstration Project are published in four
volumes:
     1.    Volume I:  Demonstration/Hydrogeologic Study (George et al  1985);
     2.    Volume II: Percolate  Investigation  in the Root Zone  (Ramsey and
          Sweazy 1985);
     3.    Volume III: Agricultural Research Study (George et al 1985); and
     4.    Volume IV: Lubbock Infection Surveillance Study (LISS) (Camann
          et al 1985).

-------
BACKGROUND
     During the 1930s the  City  of  Lubbock entered into a contractual agree-
ment with Dr. Fred Standefer  to pump  all the sewage effluent to his  farm,
later  to be known  as the Gray farm.   As Lubbock grew, the Gray farm was
able to expand to encompass 1,489  ha.   Nonetheless, the Gray farm could not
adequately manage the hydraulic flow  pumped from the City of Lubbock.   Con-
sequently,  the farm was over-irrigated and ground water  accumulation oc-
curred beneath the farm with  associated water quality problems.
     In November 1980 construction commenced  to expand  the Lubbock  Land
Treatment System to include  the  Hancock farm located 25 km southeast of
Lubbock. and directly north of the  City  of Wilson, Texas.  The expansion was
designed to reduce  the hydraulic  and  nutrient overloaded condition of the
Gray farm. The combined area  of the Lubbock Land Treatment system was 2,967
ha (7,330 acres).

DESCRIPTION OF LAND  APPLICATION SYSTEM  EXPANSION
     Lubbock1s Southeast Water  Reclamation  Plant (SeWRP)  consists of two
trickling filter systems and  an  activated sludge system (Figure 1).  Un-
chlorinated effluent from  the two  trickling filter plants.was pumped to the
Gray and Hancock farms.
     A total  wastewater  discharge of  approximately 5.5 x 10\i-Vd (15 mgd)
was to be divided equally  between  the  Gray and Hancock  land application
sites.   Effluent from SeWRP  was conveyed to the Hancock land from a three-
pump, pumping station through 25 km of  0.69 m force main.
     The  diurnal flow variation within the wastewater treatment system due
to the management of water between the trickling filter  plants and the
activated sludge plant reduced  flow through the trickling filters from 2:00
a.m.  to 10:00 a.m.  each day  to 315 nvVhr (2.0 mgd).  The pump capacity and
sump were not designed to  absorb the  variations in flow from the trickling
filter  plant.  Consequently, the  dynamic nature of the effluent hydrograph
made it impossible to operate two  pumps for more than 16 hours each day.
     At the  northern boundary  of  the Hancock farm, the effluent was routed

-------
        Lime
Aeration
Treatment
                                             Supernatant
                             Primary
                            Clarifiers
                         Trickling   Secondary
                          Filters    Clarifiers
                                                               X	L_l
                            , , Anaerobic, Digesters
                                                       -*- Digested Sludge
                                             Trickling
                                             Filters
Hancock    Hancock
Lagoons    Farm
                                                Secondary
                                                Clarifiers
      Screens Grit
             Chamber
                         Return Sludge
                                            Plant Generator

                                         fteTurnSTudae'
                                                     Secondary
                                                    ' Clarifiers
                                                                              Chlorine
                                                                              Contact Chamber I
 I	t	SjJa^r^keILer_D^s^s^LO^tpige.sted Sludge	
                                                                                                            Power
                                                                                                            Utility
Figure 1.   Southeast Water Reclamation Plant  Flow  Diagram

-------
through  three 0.38 m  plastic irrigation pipelines to three  separate  reser-
voirs (Figure 2).  The reservoirs  were constructed on natural  playa  lakes.
The  reservoir capacity was adequate  to provide emergency   storage  during
rainfall  events,   and  to prevent the necessity of irrigating during periods
of cultivation,  seeding, and harvesting of crops.  Approximately  3.5  months
of storage were provided  by the three reservoirs.   Irrigation  pump sta-
tions  were provided at each reservoir.  Constant pressures  were  maintained
throughout the system  by a variable speed (lead) pump and a-constant  speed
(lag)  pump located on Reservoir 1.  Both pumps were controlled by  system
pressure  and discharge flow rate.
     The  hydraulic  distribution  system was designed to irrigate 1,153 ha
with 1,082 ha irrigated by electric drive center pivot irrigation machines.
Each center pivot was designed tg irrigate up to 15 cm in 20  days after
allowing  for 20 percent loss due to evaporation.   Without  the  use  of the
reservoirs, five to  six center pivots could be operated at the  same time,
utilizing the flow pumped  directly  from Lubbock1s wastewater  treatment
plant.  Each center pivot had a centrifugal booster pump.   The booster pumps
increased the line pressures to an operating level of 3.1 x  10°  pascals (45
psi).
     In-line  screens  were placed between the centrifuge booster pumps and
the center pivots  to reduce clogging of the spray nozzles.  On  each irriga-
tion machine Nelson® spray  nozzles were installed on drops located 3.2 m
apart.   The size of  the nozzles varied from  2.4 mm (3/32  in)  to  7.1 mm
(9/32  in).  Nozzles were positioned a distance of 1.2 m to  1.8 m above the
ground  which allowed easy maintenance of the nozzles. Each nozzle provided
a 360° umbrella  pattern with an effective wetted diameter  of  8.5 to 9.1 m
(28 to  30 ft)  to allow for the greatest application intensity.   The  energy
dissipating deflector incorporated into the nozzle assembly was a concave
plastic plate.  Water  discharged through the orifice was  deflected  upward
once it struck the deflector which enhanced the creation of  aerosols  during
the period of study  and increased drift and evaporation  of  water.  Convex
deflectors were installed on most nozzles after the LISS monitoring  period
ended (i.e., after October 1983) to direct the water downward.   This  change
reduced  aerosol  formation and drift.   In addition, end guns were provided

-------
                                                          Fore* Main
                                                          from S.WHP
                        Furrow Irrigation

                        Distribution Can
Figure  2.   Hancock Farm Hydraulic Distribution  System
                                   5

-------
on each pivot to  irrigate the corners.

EFFLUENT QUALITY
     During  1980 and 1981, Lubbock's SeWRP was producing an  effluent  from
the trickling filter  system which had a composition equivalent to  a  typical
medium  untreated domestic  wastewater  (Table 1).   The City of  Lubbock's
wastewater discharge  permit for  SeWRP required  the plant to produce an
effluent  with a  30-day-average 5-day biochemnical oxygen demand (8605) not
greater than 45 mg/1.  During the project monitoring period  the  effluent
BODj quality from SeWRP ranged from a monthly high of 260 mg/1 to  a  monthly
low of 27 mg/1:
                             Average Monthly Effluent 8005
                             _ Produced by Lubbock SeWRP
           Month
          January
          February
          March
          April
          May
          June
          July
          August
          September
          October
          November
          December

This poor quality effluent was mainly attributable to the malfunctioning of
the anaerobic digestion process. From June 1980 to February 1982,  the aver-
age effluent total  organic  carbon  (TOC)  produced from trickling filter
Plant  2 was 118.7 mg/1.  Total Kjeldahl Nitrogen (TKN) concentration aver-
aged 38.59 mg-N/1 of which 67 percent was ammonia-nitrogen (25.95  mg  N/l)
and 33 percent was organic nitrogen. Approximately 57 percent  of  the total
1982
mg/1
• 143
260
198
139
108
128
130
76
69
171
63
86
1983
mg/1
71
120
105.
65
30
39
49
27
43
31
63
49

-------
TABLE 1 .   CHARACTERIZATION OF EFFLUENT PRODUCED BY
SOUTHEAST WATER RECLAMATION PLANT IN 1980 AND 1981

Parameter
Alkalinity (mg CaCOyl)
Specific Conductance (pmhos/cm)
Total Dissolved Solids (mg/1)
pH
Chloride Ion (mg/1)
Sulfate Ion (mg/1)
Total Kjeldahl Nitrogen (mg N/l)
Nitrite plus Nitrate Nitrogen (mg N/l)
Ammonia .Nitrogen (mg N/l)
Total Phosphorus (mg P/l)
Orthophosphate Phosphorus (mg P/l)
Organic Phosphorus (mg P/l)
Chemical Oxygen Demand (mg/1)
Total .Organic Carbon (mg/1)
Concentration
Average
337
2216
1695
7.54
468
315
38.59
0.29
25.95
14.43
8.36
5.15
302
118

Standard
Deviation
34
290
537
0.21
55
43
.15.23
0.30
6.69
.4.27
2.03
4.20
136
45

-------
phosphorus  (14.43 mg/1) present in the effluent  from the trickling filter
plant  was orthophosphate  phosphorus  (PO^).   During the spring of 1982,
SeWRP placed on-line  additional anaerobic digesters  and  rehabilitated  the
primary  clarifiers and rotary distributors of  the trickling filter plants.
A much higher quality waste stream was pumped  to the Hancock and Gray farms
in 1982 through  1983. TOC levels at the terminus of  the forced main were 46
percent less than  the average concentrations measured in  trickling  filter
plant  effluent  samples obtained the previous  sampling periods.  No statis-
tically significant differences ( a = 0.05) were observed in   TKN  levels
measured in  the  waste streams from Plant #2 (38.59 mg N/l) and at the term-
inus of the  force  main (41.70 mg  N/l).   As SeWRP1s effluent reached  the
Hancock  farm, 62 percent  of the TKN was ammonia-nitrogen (25.80 mg N/l).
The data indicate no nitrogen transformations  through the force main.
Average  total phosphorus (TP) and organic phosphorus (Org P) levels (11.82
mg/1 and 1.6 mg/1) contained in water samples  obtained from the terminus of
the force main did decrease significantly from  baseline effluent concentra-
tions.  The  decrease  in TP appeared to be a result of a decrease in organic
phosphorus  mass loading from the trickling filter  plant.  As anticipated,
the bulk (71  percent) of the nitrogen contained in the water entering  the
Hancock  farm (41.77 mg-N/1) was lost within the reservoirs.  The reservoir
effluent average TKN concentration was 11.74 mg-N/1. The median nitrite
plus  nitrate (N02+N03 level in the reservoir  discharge  stream was 0.27 mg
N/l.                                                                .   .
     Approximately  85  percent  of  the total  phosphorus contained in the
effluent (11.82  mg/1) pumped to the Hancock farm was orthophosphate.  Aver-
age orthophosphate levels  decreased  from 8.43  to 4.85 mg/1  and as the
wastewater flowed  through the reservoirs total phosphorus concentrations
were  reduced by 47 percent  from 11.82 mg/1 to 6.31 mg/1.  Orthophosphate
removal in the reservoir accounted for about 65 percent of the decrease in
TP.
     The wastewater  effluent pumped to the  Gray  farm was significantly
affected by  the  activated sludge plant.  The nitrate-nitrogen concentration
averaged 3.45 mg/1.   Furthermore, the higher treatment efficiency of the
activated sludge plant decreased the TOC levels to an average of 52.6 mg/1.

                                   8

-------
Uptake of  phosphorus by suspended biomass, also, significantly reduced  the
average TP  concentration (9.18 mg/1)  in  the waste stream pumped to the Gray
farm.
     The sewage  treated by  SeWRP was primarily  derived.from domestic
sources with  less than  30  percent contributed from industrial sources.
Trace metal  levels contained  in SeWRP effluent reflected this low indus-
trial  wastewater  flow and presented no potential phytotoxicity problems.
Table 2 summarizes the concentration ranges of specific trace metals meas-
ured in treated wastewaters.   No significant differences  (  a= 0.05) In
trace metal  and  mineral  levels were determined  between any  irrigation
water  source  from February 1982 to October 1983.
     TABLE  2.   CONCENTRATION OF TRACE  ELEMENTS  IN TREATED WASTEWATER

Element
As
B
Cd
Cr
Cu
Hg
Mo
Ni
Pb
Se
Zn
Wastewater Ef
Range*
(mg/1)
<0. 005-0. 023
0.3-2.5
<0. 005-0. 22
<0. 001 -0.1
0.006-0.053
<0. 0002-0. 001
0.001-0.018
0.003-0.60
0.003-0.35
	
0.004-.35
fluent Median Concentration (mg/1)
Median*
(mg/1)
<0.005
0.7
<0.005
0.001
0.018
0.0002
0.007
0.004
0.008
	
0.04
SeWRP
<0.005
0.027
<0.0005
0.060
0.047
<0.0004
<0.003
0.065
0.032
<0.005
0.133
Hancock
Reservoir
<0.005
0.038
. <0.0005
0.006
0.033
<0.0004
<0.003
0.007
<0.005
<0.005
0.066 "
* Chang and  Page, Land Treatment of Wastewater, Vol. 1, pp.47

     The data indicate that minerals  may create salinity  and sodic prob-
lems within the upper soil  profile.   The effluent produced by SeWRP  was
slightly saline   (dissolved solids from 1,000   to 3,000 mg/1).  The low  hy-

-------
draulic loading to the Gray  and  Hancock, farm (20 to 60 cm) could contribute
to the accumulation of salts within  the upper soil profile.  Without proper
salt management,  salts could pose  future phytotoxicity problems to farmers.
The adjusted sodium adsorption ratio (SAR) of the effluent stream  from  the
trickling plant averaged  21.6.   Irrigation water with an adjusted SAR above
10 may create severe water penetration  problems and development of  alkali
soils  (Stromberg  and Tisdale  1979, EPA 1981, Loehr et al 1979).  Proper
management of salts contained in the irrigation water  was viewed  as  the
most  important task  which   would govern the long term success of 'the land
application system.
     Since  agriculture  is  the  major  industry in the Lubbock area, herbi-
cides (e.g., atrazine and propazine)  and by-products  produced  from  the
decomposition of  herbicides (e.g., 2,3-dichloroaniline and 3,4-dichloro-
aniline)  existed  in the SeWRP's  effluent.  Carbon tetrachloride, chloroben-
zene,  and diethylphthalate .levels  exceeded the respective organic concen-
tration range in  municipal wastewater treatment plants cited by Majeti  and
Clark (1981) and  Pettygrove  and  Asano (.1984).  A mean anthracene concentra-
tion of 6.1 ug/1,  4.0  yg/1 and 8,4 yg/1  was contained in the effluent  from
the trickling filter  plant;  wastewater  pumped to the Gray farm; and efflu-
ent at the terminus of the force main,  respectively.
     The  average  fecal coliform  concentration in the waste stream pumped to
the center pivot  irrigation  machine   exceeded EPA guidelines throughout  the
study period.  The guidelines state:
     "Biological  treatment by ponds  or  inplarit processes plus con-
     trol of fecal  coliform count to less than 1,000 MPN/100 ml -
     acceptable  for controlled  agricultural  irrgation except for
     human food crops  to  be  eaten raw."  (USEPA, 1981)
The actual  flow-weighted average fecal coliform concentrations  of  the
applied wastewater during the four major irrigation periods were:
                                  10

-------
                                      Fecal  coliform concentration
                                   colony forming units (cfu)/100 ml
              Spring 1982                        4,300,000
              Summer 1982                          840,000
              Spring 1983                          '  5,200
              Summer 1983                          120,000

     During the  first two years of the study (June 1980 to February 1982),
total enterovirus levels as measured on HeLa cell monolayers ranged  from
0.045 pfu/ml  to over 1.0  plaque forming  unit (pfu/ml) in the summer of
1980. The first  effluent sample from  the terminus of  the force main  was
obtained  at  the Hancock  farm in February  1982 and  represented a highly
atypical microbiological   sample.  Once a daily wastewater flow  to  the
Hancock  site was  established, the initial microbial and physical profile of
the wastewater delivered to the irrigation site  was not dissimilar from the
wastewater  previously characterized  at  the treatment plant.  In 24-hour
composite samples, maximal viral levels of  about 0.1 pfu/ml were detected
during  the  summer 1982 irrigation season. A similar pattern of enteric
viruses  enumerated on HeLa cell monolayers was observed during 1983.  While
viral levels in  effluent at the pipeline terminus did not  reach the highest
levels seen  in June 1982, the number of viruses recovered remained  rela-
tively constant  from late June through August 1983 at over 0.25 pfu/ml.
     A comparison of both indicator bacteria and virus  levels showed  that,
in general,  organism concentrations in reservoir water  were two to three
orders of magnitude lower than comparable wastewater  at pipeline terminus
(Table  3).   Of  the 19 samples of reservoir water concentrate which were
assayed  in two cell lines, enteroviruses were detected  in  only 12  samples
with  a  maximum  level of about 0.06 pfu/ml.   In most of the reservoir sam-
ples, viral  levels were at or below the detection sensitivity of the recov-
ery procedures employed.
     The most prevalent Enterobacteriaceae species encountered in waste-
water from  Lubbock included Citrobacter,  Enterobacter, Escherichia and
Klebsiella.  Aeromonas hydrophila was  the most abundant non-Enterobacteria-
ceae  member recovered, followed by Pseudomonas  species.   The effectiveness

                                 11

-------
TABLE 3.  MICROORGANISM CONCENTRATIONS IN WASTEWATER APPLIED BY SPRINKLER
                                 IRRIGATION
  Measurement by
irrigation period
                                                     Wastewater Source
 Pipeline
effluent
a
Reservoir
effluent0
FECAL COLIFORMS (colony forming units/ml)
  Feb-Apr 1982
  Jul-Sep 1982
  Feb-Apr 1983
  Jul-Sep 1983
ENTEROVIRUSES (plaque forming units/ml)
  43,000
  13,000
  20,000
  90,000
          130
           50
           30
. Feb-Apr 1982
Jul-Sep 1982
. Feb-Apr 1983
Jul-Sep 1983
0.04
0.05
0.07 .
0.17
0.003
<0.004
<0.004

Geometric mean
Geometric mean
of four to eight 24-hour composite samples.
of four to five grab samples.

             TABLE 4.  BACTERIAL SCREEN-- HANCOCK RESERVOIR
Organisms (10  cfu/ml)
          Sampling date
           Jul 26-27,
              1982
ENTEROBACTERIACEAE

Enterobacter cloacae
Klebsiella oxytoca
Klebsiella ozaenae

NON-ENTEROBACTERIACEA

Achromabacter xylosoxidans
Acinetobacter calcoaceticus var. Lwoffi
Aeromonas hydrophila
Alcaligenes sp.
CDC Group V E-2
Pseudomonas sp.
Psuedomonas cepacia
Pseudomonas maltophilia
              0.4
              0.1
              0.1
              0.9
              0.2
              4.3
              0.5
              0.1
              0.5
              0.1
              0.3
a.  Highest levels observed on either MacConkey agar of brilliant green agar
    and identified by API 20E bio-chemical tests.
                                  12

-------
of ponding  for the  reduction of microbial numbers was evident  both  by  the
lower levels and the  reduced diversity of organisms seen in  a single bac-
terial  screen completed on a sample from the Hancock reservoir (Table  4).
Since microorganism densities were much higher in the wastewater from  the
pipeline  than  from  the  reservoirs,  the exposure which most of the study
population received to  most microorganisms via the wastewater aerosol  was
greater in  1982 than in 1983.
     During  system operation, the fecal coliform concentration of the waste
stream from  SeWRP  and the discharge  from the  storage reservoirs greatly
exceeded  EPA guidelines, especially in 1982.  The effluent  8005 concentr-
ation produced by  SeWRP did not satisfy Texas permit, requirements until  May
1983.  The system, however, was operated below hydraulic design  capacity in
1982 and 1983.

SYSTEM OPERATION
Hancock Farm
     The Hancock slow rate system had the following alternative  operational
modes:
                                        ^
     1.   Direct irrigation with effluent from SeWRP;
     2.   Irrigation  with water only from reservoir; and
     3.   Combined direct irrigation with SeWRP effluent and  reservoir
          water.
     During  1982  the Hancock  farm was irrigated primarily  with secondary
effluent produced by  SeWRP.  Odorous compounds stripped  from the effluent
stream  as it was emitted  from the spray nozzles created public nuisance
conditions.   Consequently, the reservoirs were  used to oxidize the odor
compounds prior to  irrigation.  In 1983 practically all  of the water  ap-
plied to land was  pumped from the storage reservoirs.  Since  the same pipe-
line  distribution network  was used  to  provide water to  the center pivot
irrigation machines and transport Water to the reservoirs, main pipelines
had  to  be dedicated  to either irrigation from the reservoirs or transport-
ing water to the reservoirs.  Increased head losses resulting from closing
of valves to accomplish irrigation solely from the reservoirs,  in conjunc-
                                  13

-------
tion  with  the wastewater  management condition at  SeWRP, reduced the flow
pumped to the Hancock  farm. Consequently, the Hancock farm received only 28
percent (4,128,219  m3) of the total effluent produced from February through
December 1982.  In 1983,  19 percent (3,744,395 m3) of  the total effluent
was  pumped to the Hancock  farm from  January 1 through  October.   The
hydraulic  loading pumped  through each center pivot irrigation machine in
1982 and 1983 is  presented  in Table 5.  These hydraulic loadings  are very
low for a slow rate land application system.
     Due to the necessity to divert all SeWRP effluent to the reservoirs,
water management  at the farm was a problem.  A maximum of nine of 22 center
pivot machines were operated simultaneously.  Consequently, system hydraul-
ics was  the major factor  which governed irrigation  practices and not crop
requirements.   Cotton  was the primary crop grown at  the Hancock farm  prior
to 1982.   Rainfall and associated hail during the  months of May and June
1982 which  were the 24th and 25th  months of the  study monitoring  period
(Figure  3) destroyed  over  8.09 x 10^ ha (2 x 10^ acres)  of the cotton crop
in the South Plains of Texas.  Only 16.2 ha (40 ac)  of cotton remained  on
the Hancock farm.  The majority of the farmers planted grains to partially
recuperate   financial  losses.  Tenant farmers  at the Hancock farm  planted
approximately 552 ha (1365  ac) of grain sorghum,  162 ha (400 ac) of sun-
flowers,  and 257  ha (635 ac) of soybeans (Figure A.1).
     During the  summer  of 1983,  less than  2.5 cm (1  inch) of  rain was
recorded from the end  of 3une through mid-October.   Figure A.2 shows the
crops grown in 1983 at the  Hancock farm.
Gray Farm
     Secondary treated  effluent from SeWRP was delivered to the Gray farm
through three pipelines to  three storage reservoirs  (Figure 4).  The  esti-
mated hydraulic retention time of the ponds was 10 days.
     Prior  to 1982, with 75 to 80 percent of the  farm planted in cotton,
water  was  applied to the  cotton areas in early spring, February through
April (prewater); and  in the summer from June through August.  An estimated
70 cm  of water  was applied  to the  land designated for  cotton planting
(Table 6).   Any other  irrigation (the remaining six  months), with  no  stor-
                                 14

-------
TABLE 5.   TOTAL WATER APPLIED TO HANCOCK FARM IN 1983

Pivot No.
1
2
3
4
5
6
7
. 8
9
10
11
12 •
13
14
15
16
17
18
19
20
21
22
Total
(cm)
13.23
20.50
18.16
23.16
11.79
20.47
19.69
14.58
14.38
13.13
23.80
20.50
16.26
10.72
26.06
14.68
17.15
18.49
15.72
15.52
16.94
14.78
1982
(in)
5.21
8.07
7.15
9.12
4.64
8.06
7.75
5.74
5.66
5.17
9.37
8.07
6.40
4.22
10.26
5.78
6.75
7v28
6.19
6.11
6.67
5.82
Total
(cm)
20.0
34.0
. 27.0
48.0
40.9
50.0
31.8
26.9
38.4
29.0
29.2
31.0
33.3
22.9
43.9
35.3
29.2
29.0
30.2
17.5
20.6
27.4
1983
(in)
7.9
13.4
10.6
18.9
16.1
19.7
12.5
10.6
15.1
11.4
11.5
12.2
13.1
9.0
17.3
13.9
11.5
11.4
11.9
6.9
8.1
.10.8
                        15

-------
  o
  o
  C\J
                                                                     D Lubbock Airport
                                                                     O Normal Precipitation
                                                                     A Hancock Farm
                                                                    -j~ Gray Farm
   1.00
6.00
11.00      16.00      21.00     26.00
     MONTHS  (JUNE  1980-SEPT 1983)
31.00
36.00
m.oo
Figure 3.  Precipitation During Project Period

-------
                                                                         ••••-  Pipeline
                                                                             Water
                                                                         1 cm = 0.27 km
                                                                                                   SPRINGS
Figure  4.   Gray  Farm Land Application  Site

-------
age,  had  to be put  on winter  crop or grazing area (Figure A.3).  From two

to 4.5 m/yr was applied to  these  areas in  order to keep  the main economic

crop  (cotton) at  maximum production.  In the  spring of 1982 over 506 ha

(1,250 ac) of alfalfa (Figure  A.4) , 304  ha (750  ac)  of  wheat, and  121  ha

(300 ac) of soy beans were  planted on the  Gray farm.
               TABLE  6.   GRAY  FARM  HYDRAULIC LOADINGS/CROP*

Season

Prior to
1982
1982
1983
Flow
MGD

14
10
10
Cotton Milo Wheat
121 1

70 cm 65 cm — 465 cm
70 cm 230 cm
70 cm 207 cm
Soybeans Alfalfa
1 1 2

—
70 cm 70 cm 55 cm
70 cm 55 cm

* Approximate loadings applied  based   on maximum output  pivot and number
  of days of potential operation  during the growing season.

1 Row or flood irrigation  areas.    Row water crops  are based on one pre-
  water application and six  summer  time applications.
2 Center pivot irrigation  — Pivots are nozzled to deliver 15 cm/21 days,
  over six months application time  (7  on alfalfa).  This yields hydraulic
  loadings of 90 cm (105  cm on alfalfa).   Because only enough pivots are
  on hand to irrigate one-half  of the  acreage at any one time, these load-
  ings must be halfed to 45  cm/yr and  52 cm/yr, respectively.
                                  18

-------
                                SECTION 2
                         SUMMARY OF CONCLUSIONS

     The Lubbock Land  Treatment System consists  of two privately  owned
farms. -In  the past years,  the Gray farm suffered from  inadequate storage
and distribution piping network to  properly manage effluent produced by
Lubbock1s SeWRP.   Consequently, an increase  in ground-water elevation and
degradation of ground-water quality occurred beneath the  farm.   The  system
was expanded in 1981  to include the t,478 ha Hancock farm  which  is located
25 km southeast of Lubbock, Texas.   The expanded slow rate  land  application
system encompassed approximately 2,967 ha.  From June 1980  to October 1983,
both farms  were monitored to assess the impacts on  ground water,  soils  and
crops of:  1) reducing the hydraulic, chemical, and biological mass loading
for the Gray farm; and 2) spray irrigation of effluent to the Hancock farm
which was primarily a  dry land farm for  ten years  prior  to  1982.  Further-
more,  an epidemiologic study at the Hancock  farm was conducted  to  assess
the association  between human exposure to the wastewater  used  for irriga-
tion and the development of new infections.
     The findings of  the project  indicated  that the major  recharge of
ground water beneath the Gray farm was from  flood  irrigated wheat  areas.
Deep  percolation  of  irrigation water and precipitation continued in 1982
and 1983 in the  flood irrigated areas.   Physical limits of  irrigation
equipment, hydraulic  distribution  system, water storage, and crop cultiva-
tion eliminated the capabilities for proper  water  management.   With ade-
quate  winter storage  and the hydraulic capability  to distribute  more water
on the  alfalfa in  1982 and  1983, minimal deep  percolation  would have
occurred through the soil throughout the farm.   Comparison  of 1981 and 1983
ground-water elevation data indicated that the  ground-water  levels  beneath
the Gray farm decreased.
     During  the period from February 1982 to  October 1983, an  increase in
the ground-water quality also occurred beneath  most of the  Gray  farm.  Mass
balances conducted on  nutrient and  minerals indicated continued  leaching of
                                  19

-------
constituents through  a  soil depth of  183  cm beneath  the  flood irrigated
area; whereas,  most  of  the chemical constituents applied by  sprinkler irri-
gation  were retained  and/or removed through crop  uptake beneath the spray
irrigated areas.
     Statistically  significant  decrease in NQ.3-N levels were measured in
five of 27 monitoring wells from February 1982  to  October 1983.   In  gen-
eral, 17 of 27  wells experienced a decrease in ground-water  NOj-N levels. A
comparison of baseline  data (June 1980 to February  1982) and data collected
after  February 1982 indicated a decrease in the frequency of ground-water
NOj-N concentrations equaling or exceeding drinking water standards in nine
of 27 wells monitored.
     Wastewater  treated by  SeWRP was  primarily  derived  from domestic
sources with less  than  30 percent contributed by industrial  sources.   Con-
sequently, trace metals posed no potential  toxicity  problems to humans or
plants.
     Total  irrigation  at the  Hancock farm  varied from 16 cm to 20 cm in
1982 and 36 to  49  cm in 1983.  An overall increase  in  ground-water eleva-
tion occurred  beneath the  Hancock farm.  A maximum  rise of three to five
meters  was experienced  in ground-water wells  in close  proximity to  surface
runoff  collection  areas.  Increases in  ground-water  elevation beneath the
Hancock farm were  primarily due to percolation  of  surface runoff  through
coarse  material contained in moats surrounding the reservoirs and excava-
tions constructed  to reduce flooding of cropland and migration of percolate
through  material surrounding  poorly sealed well casings.  Increases in
ground-water elevation  commenced approximately two  months after heavy  pre-
cipitation events.
     Chemical constituents  contained in  the   treated  wastewater applied to
the Hancock farm were removed by the soil-crop matrix  from percolate water.
Increases in ground-water  chemical parameters appeared to be associated
with deep percolation of surface runoff contained in moats and excavation
pits constructed to contain surface runoff.  Existing salt and nitrate
deposits within the  soil profile were leached with  percolate to the ground
water;  thereby causing  increases in nitrate and total dissolved solids
(IDS) levels in several wells.
                                  20

-------
     In general,  no  significant changes in trace  metals or priority organic
pollutants occurred  in the ground water during  the  monitoring period. Based
on values cited  in  literature, trace elements  posed no public health prob-
lems.
     Salt accumulation occurred  in the upper 183 cm of the soil profile.
As expected,  salt accumulations were directly proportional to mass loadings
from  irrigation.   Insufficient water was applied  (less than 21 cm in 1982
and less than 50  cm  in 1983) to leach salts below the root zone.  Exchange-
able  sodium  percentage increased from two to  six  percent in the top 30 cm
of soil during the period from February 1982 to.October 1983.
     Cotton  and  grain sorghum  (milo) were the  primary crops grown on the
Hancock farm  in 1980, 1981 and 1983.  Due to severe weather in 1982,  sun-
flowers,  soybeans  and grain  sorghum were planted as alternative crops to
cotton.  While milo  yields were low due to late  planting  and trifluralin
damage,  sunflower and soybean yields were average  for the High Plains area
of Texas.  An improvement in cotton crop production occurred in 1983.  With
irrigation  with   effluent,  the cotton yields  for  the farm were 48 percent
greater than  the  Lubbock County average.  1983  cotton yields may have  been
limited  by  possible nutrient shortages,  boll  worm infestation, and. cool
weather during late  growing season.  Cotton production in 1983 ranged  from
353 to 740 kg/ha.
     Amortized system construction cost over  a  20 year period at ten per-
cent  annual  interest rate  would be.$167/1 ,000  m3 per year ($0.63/1,000
gal).  With  85 percent federal cost  sharing amortized  construction  cost
would have been reduced to $25/1,000 m3/yr ($0.10/1,000 gal).  Inclusion of
land cost would have increased annual capital  cost by  24  percent.   Total
operation and maintenance  (0 & M) costs associated with the Lubbock Land
Treatment System  Expansion were $156/1,000 m3 ($0.59/1,000 gal) in 1982 and
$139/1,000  m3 ($0.53/1,000  gal)  in  1983.  The City of  Lubbock  bore
$71/1,000 m3  of the  total 0 & M cost in 1982 and  $58/1,000 m3 in 1983.  The
farmer's portion of the 0 & M was $85/1,000  m3 ($0.32/1,000 gal)  and
$81/1,000 m3  ($0.31/1,000 gal)  in 1982  and 1983, respectively.  The economic
balance of cost expended and revenues received  showed  a  net negative  bal-
ance  each year  during the project period (1980  through 1983) ranging from

                                  21

-------
$701,661.81  (1981) to $1,103,687.57 (1982).  Net costs were $267.35/1,000
m3 ($1.00/1,000 gal) in  1982  and $161.28/1,000 m3 ($0.61/1,000 gal)  in
1983.  Crop revenues offset costs by 18  and 47 percent  of total costs  in
1982 and 1983,  respectively.
     Spray irrigation of unchlorinated  wastewater piped from the treatment
plant was a more substantial source  of aerosolized microorganisms than
spray  irrigation of wastewater stored in reservoirs.  Enteroviruses were
regularly recovered in the aerosol  at 44 to 60 m downwind of irrigation
with piped treatment plant  wastewater.   The  geometric mean enterovirus
density in the  downwind air was 0.05 pfu/nr, although a much higher density
(17  pfu/m3)  was sampled  in  August 1982.   In addition, fecal streptococci
levels were detected at least 300 m  downwind,  and levels of fecal  coli-
forms,  mycobacteria and  coliphage were isolated at least 200 m downwind.
Organism levels downwind were also  significantly higher  than background
levels  in ambient air outside of participants'  homes:   fecal coliform
levels were higher beyond 400 m downwind, mycobacteria and coliphage levels
to at least 300 m and fecal streptococci levels to at least 200 m.
     The results indicate that a general association between exposure  to
irrigation wastewater and new infections existed, especially for 1982 when
there was exposure to higher levels of micro-organisms via wastewater
aerosol.   Poliovirus 1 seroconversions  were probably related to wastewater
aerosol exposure during the spring of 1982, even when the effects of  polio
immunizations  were controlled.   However, even during 1982, the strength  of
association remained weak and frequently was  not stable.   Wastewater  of
poor quality from the pipeline, comprised much of the irrigation water  in
1982. Of the  many infection episodes observed in the study population, few
appear  to have been associated with wastewater aerosol exposure, and none
resulted in serious illness.
     The lack  of a strong, stable association of clinical illness episodes
with the level  of exposure to irrigation•wastewater indicates that waste-
water spray irrigation produced  no obvious  disease during the study period.
However, when more sensitive indicators  of infection were used, the evi-
dence  indicates an association  existed,  especially for 1982.  A particular
concern from  a  public health standpoint  is  the evidence that the poliovirus

                                 22

-------
1 seroconversions were  related to wastewater  aerosol exposure during the
spring of 1982,   even when  the effects  of polio   immunizations  were con-
trolled.   Because of  the  low prevalence of poliovirus antibody observed
during the baseline period, the study population was  immunized, and  thus
was probably better protected against polio  than  other rural populations.
High concentrations of  both bacteria and enteric viruses  were observed  in
the 1982  poor  quality  wastewater applied as received  via pipeline directly
from the Lubbock  sewage treatment plant.  Exposure  would have been  reduced
by using wastewater from the reservoirs for irrigation rather than irrigat-
ing directly from the pipeline.
     Annual  hydraulic  loading  rates up  3 m.ha/ha.yr  did not  adversely
affect cotton,  grain sorghum, and alfalfa crop production.  Highest alfalfa
yields  were obtained in test plots irrigated  with  365 and 434 cm.ha/ha.yr.
Total dissolved solids  and associated sodium salts  were  leached beyond  91
cm soil  depth  within plots irrigated with 61  cm of treated sewage per year
or greater.   Bermuda yields were limited by transport of  macro and micro
nutrients past  the root zone.
     Soybeans  with  a  relatively  shallow  root system,  produced  highest
yields with more  frequent irrigation (i.e.,  one  irrigation per  week).
Soybeans  were  unable  to develop a deep root  system to utilize deeper soil
moisture during periods of water stress (one irrigation every four weeks  or
one irrigation  every eight weeks);  consequently, crop  yields were reduced.
     During  long  periods between irrigation events, the  deep root  system
developed by grain sorghum  enabled  the plant to utilize available soil
moisture and inorganic  nitrogen at greater depths.   Highest grain  sorghum
production  was achieved in plots irrigated 61 and  122 cm/yr at application
frequencies  of  once every four weeks and once  every eight weeks.
     Increasing  the  quantity of water applied  to  a crop transports sodium
salts deeper into the soil profile.  Soybean seed  and  stalk analysis  indi-
cated leaching  of sodium from the root zone commenced  almost immediately  at
the 122 cm/yr hydraulic  loading.   At the 61  cm/yr  loading, irrigation
events  must occur at  intervals of two weeks  or longer to promote leaching
of sodium.  Practically no leaching occurred even  at  the  one application
per eight weeks  frequency  at the effluent loading of 31 cm/yr.  With the

                                  23

-------
shorter  growing season  experienced in 1982, soybeans may have had a higher
water consumption rate than  the grain sorghum due to the  crop's maturity.
Higher water requirement of soybeans in conjunction with its shallow root
system may have  caused higher sodium accumulations in the upper 61 cm  than
observed  in grain sorghum test plots.
                                  24

-------
                                SECTION 3
                             PROJECT DESIGN

     The human  and  environmental monitoring portions  of the Lubbock Land
Treatment  Research and Demonstration Project  was divided  into two periods,
each monitoring period having a time span of approximately two years.  T-he
baseline monitoring period extended from June 1980  to  February 1982.  Dur-
ing this time  frame no effluent from SeWRP was pumped  to  the  Hancock farm.
In February 1982  treated  sewage from  the City of Lubbock1s wastewater
treatment facility  was pumped to the Hancock farm.  The second phase  of the
project monitoring  period encompassed the land application of wastewater at
the Hancock farm which began in February 1982  and continued through  Decem-
ber 1983.   The design of  each project  investigation is presented  in the
following text.

DEMONSTRATION/HYDROGEOLOGIC STUDY
     The objective of the environmental monitoring  program of the  Demon-
stration/Hydrogeologic Study was to establish a  data  base characterizing
conditions at  the  Gray and Hancock farms that would allow the detection of
any changes which might occur in the ground water, soils,  and crops  due to
reduction  of sewage effluent  loading  at the  Gray farm, and use of  sewage
effluent at the Hancock farm.
     Ground water  beneath the Gray and  Hancock farms  was  monitored each
year after  spring pre-irrigation (April), at the end of summer irrigation
(September) and in  winter (December).  Ground  water  samples were taken from
newly constructed monitoring wells, pre-existing irrigation  wells,   seeps
and springs at the Gray farms,  and drinking water wells  of residents on or
near the Hancock farm.  Table 7  gives the number and  types of ground-water
monitoring wells  at each  site.   Sampling locations were  selected using
hydrogeologic  data  in order to best  monitor  the  movement and quality of
water  on  the farms.  Figures B.1 and B.2 show the  ground-water monitoring
location for each site.

                                  25

-------
 TABLE 7.  TYPE AND NUMBER OF UNDERGROUND WATER SAMPLING  POINTS  BY SITE

Number of
Wells
1
5
3
15
23
1
5
11
10
1
11
Type of Well
HANCOCK FARM
Organic Non-contaminated well .30 cm (8 in)
Continuous water level recording wells 20 cm (12 in)
Non-continuous recording observation wells 10 cm (4
Pre-existing irrigation wells (currently in use)
Home drinking water wells
GRAY FARM
Organic non-contaminated well 20 cm (8 in)
Continuous water level recording wells 20 cm (8 in)
Non-continuous -recording observation wells 10 cm (4
Pre-existing irrigation wells (currently in use)
Multiple depth well [includes four, 12.5 cm (5 in)
Seeps, springs and retention pond overflows

in)
in)
wells]

     During each  sampling period depth to water measurements were made just
prior to taking water  samples.  Wells without  pumps  were sampled  using a
7.6  cm  (3  in) diameter, 122 cm (4 ft) long polyvinylchloride (PVC) bailer
with a neoprene check  valve connected to a 0.6 cm (1/4-in) diameter  cotton
rope.   The bailer was  cleaned between  use in each  well by immersion in
ethanol followed  by a  distilled water rinse.  The bailer removed approxi-
mately 4 1  (1  gal) of  water each time it was withdrawn  from the well.  Five
to 15 bails of water,  depending on depth of water in  the well's saturated
zone,  were wasted before samples were obtained.  The  waste volume .was ap-
proximately 60 to 80 percent of the water in the well.
     Twenty-four hour composite  water samples were  obtained from 1) the
Hancock and Gray  effluent pump stations at the SeWRP;  2) the flow  distri-
bution  can at the terminus  of the  force main prior  to the Hancock water
                                 26

-------
distribution system;  and 3) flow distribution cans which  divided  the  efflu-
ent from the reservoirs to various locations on the farm.
     Water  samples from the  Lubbock Land  Treatment System  Research  and
Demonstration Project  were analyzed for physical parameters,  priority
organic pollutants,  other organics, metals, other inorganics, and  indicator
bacteria. During the  project monitoring period, water samples were analyzed
for 104 parameters  listed in Table 8.1.
Soil Samples
     Soil sampling locations in the demonstration area  were determined by
first dividing  each  farm into  65 ha areas (one  quarter-section).  On the
Hancock  farm,   each pivot  encompassed  approximately 65 ha (one quarter-
section). Each  field  on the Gray farm was approximately 65 ha (one quarter-
section) .  Sampling  locations  were randomly selected within the 65 ha area.
Soil cores  were  pulled  from each location and  composited  at 0.3  m (1 ft)
increments  to make  one  sample  for each depth.  Soil samples were  cored just
after harvest for cotton (November)  and twice  a year for  double cropped
areas (April and November).
     During  the first  sampling period (March 1981) and final sampling per-
iod (November 1983),  1 .8 m cores were obtained  using a  10.2 cm  diameter,
1.2m  long  coring tube.  Cores were taken to only 0.91  m depth  during  the
intermediate sampling periods.
     Soil samples were  analyzed for the parameters shown  in Table  B.2. Soil
samples from the baseline years and final sampling period (winter  of 1983)
were  analyzed   for the complete list of parameters.  The 1982  soil samples
had only the top three,  30 cm sections analyzed for  pH, conductivity,
potassium,  total Kjeldahl  nitrogen, total phosphorus and priority  organ-
ics.
Crop Sampling
     The purpose of  crop sampling was to obtain plant samples  which  repre-
sented each farm, crop  and type of irrigation.  Sampling  locations for crop
samples  were determined in a  manner similar to that for  soil samples.   The
two farms were  divided  into approximately 65 ha (quarter-sections) sampling
                                  27

-------
areas, from which crop  samples were randomly collected.    Crop samples  were
collected at harvest  time  when the crops had developed maximum  maturity.
Normally, harvest occurred from mid-October through January.   Some  portions
of the farms had two  crops grown per year.  These  "double cropped" areas
were  harvested  and  sampled twice a year; mid-October  through January and
April through mid-May.
     Crop samples were  obtained for laboratory and  yield tests.  For labor-
atory analysis,  all the plants within a square meter area in  each sampling
location were removed and  composited into sterile, plastic  bags to obtain
one plant sample per  65 ha area (field or pivot).  Crop  samples were divid-
ed into specific plant  parts (i.e., seed, stem, leaves,  etc.)  at  the labor-
atory.
     The analyses performed on crop samples were determined  by the type of
plant and plant  part  (Table B.3).   Those parameters such  as metals  and
nutrients, which could be translocated from the  soil  to all parts of the
plant, were analyzed  on all plant  tissue samples.  Bacteria,  yeast  and
fungi,  which could contaminate exposed surfaces of crops, were analyzed on
all samples.
Irrigation Records
     Documentation of  the quantity of treated sewage applied  to  a  specific
area of the farm was  necessary for  the interpretation  of  soil and  crop
data.   Furthermore, the  records were  used in assessing the economics of
domestic wastewater reuse  through farming.  Finally, the irrigation records
were  used  in assessing the exposure of participants in the  Lubbock Infec-
tion Surveillance Study (LISS) to pathogens contained in aerosols generated
from  the spray  irrigation center pivot machines.  Each  farmer recorded the
daily amount of  effluent  pumped through a center pivot machine  and  the
starting and final position  of the machine in the field.
Economics
     The economics  of  land treatment were monitored through  operation and
maintenance cost records provided by the City  of Lubbock and  the  farmers
utilizing the land treatment system.  The city supplied  yearly summaries of
                                  28

-------
the cost  of treating  the water supplied to each farm and the  cost  of  oper-
ating  and  maintaining  the pump station delivering  water to  the  Hancock
farm.   Most of these  costs were divided into monthly subtotals.   Yearly,
the farmers turned  in  their  financial statements  including operation and
maintenance costs  of  the center- pivots, electrical costs of their  share of
the reservoir pumps, farming costs, and monetary returns.

THE LUBBOCK INFECTION  SURVEILLANCE STUDY (LISS)
     The LISS involved  a  four-year health watch  of people residing on or
near  the  Hancock  farm  and individuals farming the Hancock sites.   Further-
more,  organism levels  of the wastewater and  aerosols generated  through
spray irrigation were  monitored.
     The City of Wilson  was the nearest community to the Hancock farm.  It
was situated at the southern boundary of the farm.  The population of 576
(1980.census) occupied  181 residences ranging from small two bedroom stucco
or frame bungalows  to  large  all-brick homes.  Local commerce was based  pri-
marily on agriculture.
     The municipal water supply  for city residents was obtained  from six
wells which tapped  the  Ogallala aquifer.  A  water tower  and  underground
tank provided storage  facilities where the water was intermittently chlori-
nated manually prior to distribution.  Continuous chlorination of  the City
of Wilson water supply system commenced in March 1983.
     All but ten of the household within the city limits were  serviced  by  a
municipal  wastewater  collection and treatment system.  The treatment  plant
consisted of an Imhoff tank  preceded by a bar screen.  Plant  effluent was
allowed to evaporate  from  a series of lagoons while the settleable solids
were removed from  the  tank on a monthly basis  and placed in  an  adjacent
drying  bed.  Those households not  connected to the municipal system had
septic tanks.
     The rural portion  of the study area lay primarily in Lynn County  (1980
census population,  8,605), with a small portion above the northern  boundary
in Lubbock County. Approximately 130 households were located in this area
in 1980 with an estimated population of 450.
                                  29

-------
     Almost every rural  household obtained its drinking water from  a  nearby
private well which tapped the  Ogallala aquifer.   Treatment  of domestic
wastewater was  accomplished by septic  tank systems in half  of the rural
houses while the other half,  typically the older homes, utilized  cesspools.
     The Hancock site was" unique in that a typical rural community with  no
prior wastewater exposure  was challenged by the enteric agents  aetive in  a
much larger urban community (Lubbock).  Persons residing around the Hancock
site may have been exposed to infectious agents indigenous  in  the  Lubbock
population but  not circulating in the study area.   Thus, many  in the study
population may have been relatively  susceptible  to the  pathogens in the
wastewater.  A  health  watch of the rural community was maintained before,
during, and after periods  of  wastewater spray irrigation.  The  health watch
focused on infections detected serologically and through isolates recovered
from routine fecal specimens.   To-enhance the  likelihood of interpreting
observed  episodes  of infectionj the  likely routes of introduction and
transmission were monitored.
     Disease surveillance  was maintained to protect the population  from any
obvious untoward effects.  However, the study focused on infections and the
infecting  agents  rather  than illness in order to  obtain greater objectiv-
ity, sensitivity, specificity, and etiologic evidence.
     All participants  were  asked  to provide blood samples semiannually,
usually in June and December  (Figure 5).  Sera  were assayed for antibody
titers to .specific enteroviruses and other microorganisms known or  suspect-
ed to be present  in the sprayed wastewater.  A seroconversion, defined  as
the fourfold or greater  increase in agent-specific  antibody  titer in  simul-
taneously tested successive sera from one individual, was-considered  sero-
logic  evidence  that the individual had been infected by the  agent  during
the time interval between  the blood collections.  Since mycobacteria were
present in the wastewater, tuberculin skin tests were administered  annually
to give suggestive evidence of a non-tuberculosis mycobacterial infection.
     An adult  from each household  and any children under  13  years  of age
were designated as fecal donors.  Each donor, whether well or ill,  was ask-
ed to  submit routine stool  specimens  for microbiological testing  during
scheduled weeks which  spanned  each  major irrigation period  in 1982 and

                                 30

-------
         UJ
              I
         ui  —  »
8.0
6.O
4.0
2.0
         to
         I
              —   o.o
HEALTH WATCH
Blood Collections
Scheduled Fecal Specimens
Household Health Reports
and Illness Specimens
Skin Test
t
Activity Diary
ENVIRONMENTAL
MONITORING
Aerosol Sampling
Wastewater Sampling
Drinking Water Sampling
                         I  I  I  I  I   I  I
          1980
                       M|J|J|A|S|0|N|D
                                          I  I I   I  I
                                                         I   ii
1981
                      J|F|M|A|M|J|J|A|S|0|N|D

                                A
                              AA A
1982
               J|F|M|A|M|J|J|A|S|0|N|b
                                                    AA
                                                            AAA
                                      A
                                      A
                                                                    A A
1983
               J|F|M|A|M|J|J|A|S|0|N|D
                                                       AA
                                                      A A
                                                                LEGEND
                             Activity
                           , Irrigation from Pipeline
                        , Irrigation from Reservoir
    Figure 5.   LISS Study Design:  Time frame  of monitoring  in'relation to major periods of irrigation

-------
1983.   A  series of three  1-week fecal collection  sessions were scheduled
before, during, and near the end of  each irrigation period  to  detect
infection  events occurring in the interim.   Clinical bacteriological.analy-
ses were performed  to  isolate overt and opportunistic pathogens.   Clinical
virological analyses were performed to isolate enteric  viruses in the fecal
specimens  by tissue culture techniques.  Electron microscopic examination
was performed  on about one-fourth of the routine fecal  specimens to detect
a variety  of virus-like particles, many of which are  not recoverable by
tissue  culture techniques.   Detection of a specific  virus by laboratory
cultivation or  by electron microscopic examination was  considered  evidence
of a viral infection.  Each non-adenovirus viral infection was regarded to
be new, unless  the  same agent had been recovered  from the  individual in the
prior  six  weeks.
     Each  household  was contacted weekly  by  telephone  for a report of any
illnesses  during the prior week.  When a sufficiently recent respiratory or
gastrointestinal illness was reported,  the  ill participant was requested to
submit  a throat swab or stool specimen to  identify the causative  agent.
Weekly  self-reports  of  illness and  appropriate  illness specimens were ob-
tained  over the entire period of irrigation from  January 1982 until October
1983 and  over  baseline  periods corresponding to seasons of heavy irriga-
tion.
     The types  and  densities of potentially pathogenic bacteria and viruses
were monitored  in the  wastewater, wastewater aerosol,  and other environ-
mental  routes  of  introduction and transmission.   An effort was made to
determine  the fluctuations in levels  of every  measurable infectious  agent
utilized in the health watch.  However, the low densities of many agents in
environmental  samples necessitated  reliance on indicator organisms to
establish environmental patterns. -Wastewater samples of the effluent from
the pipeline and reservoirs to be utilized  for spray irrigation, and of the
Wilson  effluent,  were obtained and  analyzed for  indicator bacteria and
enteroviruses biweekly to span the major irrigation periods.  Identical
analyses  were  conducted on the Wilson sewage to ascertain if any signifi-
cant differences existed between Lubbock and Wilson  wastewaters.  Baseline
wastewater samples  had been obtained  with the  same frequency in 1981 and at

                                  32

-------
lesser  frequency in 1980  to  characterize  the  effluents.  Microbiological
screens of indigenous enteric bacteria were  conducted  on one sample  each
from the  pipeline and the reservoir per irrigation season.  The purpose of
the routine wastewater samples was to document  the presence, prevalence,
longitudinal  pattern, and passage through the study community of viral and
bacterial  pathogens  possibly introduced by the wastewater.  Extensive aero-
sol sampling was  conducted to characterize the aerosol density of indicator
microorganisms produced by the spray irrigation  of both pipeline and reser-
voir wastewater.  Virus runs were also conducted to measure the density and
diversity  of enteroviruses in aerosols emanating from  the sprinkler  rigs.
Drinking  water,  houseflies,  and dust storms also were  evaluated as other
means of introducing microorganisms into the study population.
     An aerosol exposure index (AEI) was devised to measure the degree of a
participant's  cumulative  exposure  to microorganisms  in the wastewater
aerosol, relative to all other study participants during  a given irrigation
period.  When  a number of similar  infection events were observed  either
serologically or microbiologically in the study population within a time
interval corresponding to an irrigation period,  this infection episode was
statistically analyzed for  association with  wastewater aerosol exposure
using AEI.  Infection incidence rates were  compared  among exposure  sub-
groups  and with baseline  rates to  determine  the relative risk of infec-
tion.

PERCOLATE  INVESTIGATION IN THE ROOT ZONE (PIRZ)
     The PIRZ  study investigated  the impacts  resulting from physical,
chemical,  and  biological activities in the soil  root zone on percolate flow
and quality.  To  monitor percolate flow and  possible quality changes  with
depth,  a  lysimeter system was installed at the Hancock and Gray farms so
that percolate could be obtained at three different levels within the  root
zone.   Measurements were  made of the precipitation and of the irrigation
waters applied to three cropping systems that were of  economic importance
to the  region and  which  exhibited a range of  water and nutrient require-
ments during growth. The crops grown on each test facility site were  cot-
                                  33

-------
ton, grain sorghum,  and  bermuda grass.
     Tray  and  tube  non-weighing lysimeters were employed in the percolate
collection system (Figure  6).  Vacuum devices were used to obtain percolate
from the  lysimeters.  The vacuum level maintained in the percolate collec-
tion system of  these units was adjusted to levels measured by a tensiometer
located in adjacent  undisturbed soil.
     Replicates  of  the two lysimeter technique for collecting soil perco-
late were  utilized on  each test plot.  Tubes containing  undisturbed soil
cores  and trays  containing disturbed soils placed in contact with undis-
turbed  soils were installed on each plot.  Three trays, located in cavities
108 degrees apart,  were placed at depths of 61, 122, and 183 cm.  Pairs  of
tube lysimeters were emplaced  with the upper surfaces 30 cm underground  so
that normal tillage  operations could be conducted.  Percolate was collected
from one pair of tube  lysimeters at a depth of  122 cm and  from the other
pair at 183 cm.   One  plot on  each farm area had an additional pair of tube
lysimeters which  collected percolate at the 244 cm depth.
     An underground chamber was installed at the center of each test plot.
These chambers  contained the necessary support equipment for the installed
vacuum  extractors.   The  lysimeters  were placed  in the soil  at varying
depths  radially around the chamber and at distances far enough from it  so
as not  to be influenced  by the chamber's interference with soil percolate
flow.
     A  traveling-gun irrigation system was employed to apply treated sewage
to approximately  2.5 ha  of land.  The amount of precipitation was monitored
with both  recording  and  non-recording rain gauges.
Hater Collection  and Analyses
     The quantity of treated wastewater applied during the irrigation event
was determined by the catch-can  procedure.   Lysimeters  were inspected
daily.  The percolate  volume collected in the 20 liter glass bottle which
functioned as the specific percolate storage unit for the  lysimeter was
drained and measured in  a graduated cylinder.   Water quality analysis was
limited by the  volume  of percolate collected.  Grab samples were collected
when a  lysimeter first began percolate production; started percolate col-
                                  34

-------
                                                         Extraction Tray
                                                         1.52m x .15m
                                                         (3 Trays will be emplaced at
                                                         each of the 3 specified depths)
Ul
         Vacuum Pump

             Manhole
     Sample Collection
          Units
        Tensiometers
                                                                         Tube Lysimeters
                                                                         2  - 1.07m  Length
                                                                         2  - 1.68m  Length
                                                                         Buried  0.3m below surface
         Figure 6.  Plan of Test Facility

-------
lection  again after  a period of  operation with  no percolate collection
activity; and, by protocol,  for a sample collection event.
     Weekly composite  samples  were  prepared by taking a portion of the dai-
ly collection volume from the  lysimeter and compositing  it  with the  water
composited from the  previous days.  The compositing of percolate samples
was accomplished by taking a constant volume sample from each daily perco-
late  collection event prior to April 1983.  After April, percolate samples
were composited on  a volume  weighted basis by taking 10 percent of the dai-
ly percolate volume for the  composite sample.
     The original water quality analysis is  presented in  Table B.4.   The
parameters that were measured  were  reduced to a priority listing based pri-
marily on the limited  sample volume collected.  In April,  1983 a sampling
schedule was developed to obtain more quality data from the collected per-
colate.  Based on the  amount of sample present,  the analyses to be  per-
formed  on each- type  of composite  sample were prioritized as follows.  For
the samples not acid-fixed,  the analyses were:  1) pH; 2)  COD; 3) TDS;  4)
50^;  and 5} alkalinity.  For  the acid-fixed samples they were: 1) N02+N03;
2) NH3;  3) TKN; 4) COD; 5) minerals (i.e.", Na, K, Ca, Mg); 6) Cl;  and 7)
TOC.
     Grab  samples  were taken  once  a week from each lysimeter that had col-
lected percolate over  the previous  24-hour  period.  The  sample analysis
priorities to be  followed  under the grab sample collection schedule  were:
1) conductivity;  2)  N02+N03; 3) orthophosphate; 4) bromide; 5) total pho-
sphorus;  and 6) organic phosphorus.  Also,  irrigation water applied  to the
test plots  were to be analyzed for 1) nutrients (TKN, N02/N03,  NH3,  Total
P, PO^, Organic P);.2)  minerals (i.e., Na, K, Ca, Mg); 3) COD; 4) TDS.
Soil Collection and Analysis
     Soil  cores extraction  and analysis followed the same protocol as pre-
viously described in the Demonstration/Hydrogeologic Study.   Three  1 .8  m
soil  cores were collected from each test plot.  Cores were taken only to a
depth of 90 to 107  cm  on the plots  at the Gray  site because of the  hard
caliche  layer below those depths.  A 7.6 cm x 120 cm flight auger was used
to obtain soil samples from  the bottom of the core hole to a 183 cm depth.
                                  36

-------
The cores were taken  within 6.0 m of the ends of the lysimeter trays.   Each
core was divided  into 30  cm sections.  Sections from the same depth  at  each
test plot were composited  to make a sample.  Soil samples obtained with the
auger were not subdivided.
Crop Collection and Analysis
     The procedure  used  to determine the yield and quality of crops grown
on each test plot was the  same as the protocol established  for crop moni-
toring  in  the Demonstration/Hydrogeologic Study.  Yield data on each  plot
were calculated from  crop  samples taken from three subplot  areas that  had
been laid  out parallel  with the path of the traveling-gun  irrigation  unit
and in  line with  the  lysimeter battery.

AGRICULTURAL RESEARCH STUDIES
     Agricultural research activities conducted at the Hancock farm  focused
on crop management to minimize problems associated with high hydraulic,
nutrient, and salt loading rates.
     Experimental plots were established to evaluate the effect of hydraul-
ic loading rates  on various crops, soil, and percolate water.  Similarly,
research plots were  designed and  planted to determine the effect of hy-
draulic loading rates and  frequency of application on salt accumulation  in
soils and the ultimate impact on crops.
Hydraulic Loading Rate Studies
     Cotton was planted in 4.1 m x 13.7 m test plots.  Four  replicates  were
tested  for each treatment.  Treatments consisted of annu.al  treated  waste-
water  hydraulic  loading  rates ranging  from  zero  cm.ha/ha.yr  to  297
cm.ha/ha.yr.  A total of  13 hydraulic treatments were tested.  Similarly,
grain sorghum test plots  were irrigated with 0, 122, 183, 229 and 297 cm of
treated sewage/yr. Two replicate plots were  established for each  treat-
ment .
     The investigation not only considered the production of cotton and
grain sorghum but also certain high nitrogen and water consuming crops  such
as alfalfa and bermuda.   Certain alfalfa test plots were irrigated with 434
                                  37

-------
cm.ha/ha.yr  in 1983 while  specific common  bermuda test plots had 396 cm
ha/ha.yr applied  in  1983.  Three fresh  water  control  alfalfa plots  were
established  to differentiate the  affects  of  increasing hydraulic loadings
on crop yield  and  quality from the  growth response of crops and associated
quality resulting  from certain constituents  in the wastewater effluent.
Irrigation of  test plots was minimal until  a crop stand  was established.
Once  a  stand  was  established,  hydraulic  loadings  were increased to test
water tolerance of crops and nutrient utilization.   The alfalfa was  initi-
ally  watered  with  fresh water and sprinklers to get the best stand possi-
ble.  Irrigation of  seedlings with  slightly brackish  water (average  TDS
1,227 mg/1) may have presented germination  problems.
     Alfalfa was harvested every 28 to 30 days.  All test plots were  flood
irrigated.   Each month from 12.7  to 58.8  cm  of water had to be applied to
the corresponding  test plot per irrigation  period.

Hydraulic Application Frequency Study
     The average  total  dissolved  solids (TDS) level  in the wastewater
applied to the Hancock farm was approximately  1200 mg/1.  The major  cation
present in the waste stream was  sodium.  Sodium  adsorption ratio of the
effluent was about 10.  Accounting  for alkalinity (average  concentration
344 mg/1  as  CaC03), the adjusted  SAR was  approximately 22.   Management of
salt and water in  wastewater reuse  systems  can be greatly  affected  by  the
crop grown, hydraulic loading and the frequency  of wastewater applications.
Two operational parameters which potentially will aid in the management  of
salts are the  hydraulic loading rate and the frequency of application.
     A solid set sprinkler irrigation system was employed  which reflected
the method of irrigation  used at the  Hancock  farm and is the most widely
used throughout the  United  States.  Three  nozzles  spaced  6.1 m   (20  ft)
apart on  3.8  cm  (1.5 in)  PVC was moved  through  the field to attain the
necessary irrigation.  Crop samples, yield  and soil  samples  were obtained
from  within  the  circular  irrigated areas.   A  four frequency by three hy-
draulic loading rate matrix was established with two crops, soybeans  and
grain sorghum (Table 8).  Annual,  hydraulic loadings of 30,  61, and 122 cm
were scheduled with  applications made  at  one,  two, four  and eight  week

                                  38

-------
                  TABLE 8.   TREATMENT MATRIX  HYDRAULIC  LOADING RATE FOR TRIAL 17000

Application Treatment
Frequency Code
1

1

1

1

application 01
per week
application 02
per 2 weeks
application 03
per 4 weeks
application 04
per 8 weeks
30
(12
1
(0
2
(0
4
(1
8
. (3
cm/yr
")
.02
.4
.03
.8
.6
.6
.13
.2
Rate
cm
in)
cm
in)
cm
in)
cm
in)
Treatment 61 cm/yr
Code ' (24")
05 2
(0
06 4
(1
07 8
(3
08 16
(6
.03
.8
.06
.6
.13
.2
.26
.4
Rate
cm
in)
cm
in)
cm
in)
cm
in)
Treatment 122
Code (48
09 4.
(1.
10 8.
(3.
11 16.
(6
12 32.
(12
cm/yr
11 )
06
6
13
2
26
.4
51
.8
Rate
cm
in)
cm
in)
cm)
in)
cm
in)
VjJ

-------
intervals.   As the time  intervals between applications increased, the
amounts of water per application increased to maintain the scheduled yearly
loadings.
Sample Collection and  Analysis
     Irrigation water  was derived from three sources:  1)  -from  the distribu-
tion pipeline as the effluent was pumped to the Hancock farm  from  the  City
of Lubbock;  2)  from the reservoir wastewater discharge;  and 3)  from a well
used to provide ground water for the alfalfa fresh water  control plots.
     In 1982  approximately 95 percent of the effluent  applied  to test plots
was derived directly from the pipeline, prior  to  the reservoirs,  and 15
percent  of the effluent  came from the reservoirs.   The following year 80
percent of the effluent was obtained from the storage reservoirs' and the
remaining 20 percent  from  pipelines  prior to reservoir storage.  A well
located adjacent to research plots  was  used for  the  fresh water  source.
The  fresh water source was sampled in April,  August, and December of each
year.  The effluent water  from the  City of Lubbock  as  the  waste  stream
ar-rived  at the Hancock farm,  and from the reservoirs was sampled monthly
during the irrigation  seasons.  The exact position of  the effluent monitor-
ing  location was usually the effluent box at  the northern end  of the farm
where the effluent  force main from the City sewage treatment plant  entered
the  farm. Except  for trace organic compounds,  water  samples  were analyzed
for the same  constituents monitored in the Demonstration/Hydrologic  Study
(Table B.1).
     Soil and  crop samples were  collected from each trial.  The soil and
crop sampling, locations within replications (reps) of  a treatment were ran-
domly  selected.  Soil and  crop samples  were composited across reps to
obtain a representative sample of each treatment.
     Each year prior to pre-irrigation and after harvest, soil samples were
obtained representing  each treatment and crop within  a trial.  Depending
upon  the particular  number of treatments and  replicates per  trial, one to
three soil cores were  taken per plot, composited  within  a plot,  and  com-
posited  across reps.   Sample collection and constituent  analysis were sim-
ilar to procedures  stated previously. .

                                  40

-------
                                SECTION 4

                           SUMMARY OF FINDINGS

DEMONSTRATION-HYDROGEOLOGIC INVESTIGATION
Gray Farm
Ground-water Levels—
     An objective  of expanding  the Lubbock  land treatment  system was to
relieve the hydraulic  and nutrient mass overloading experienced  at  the Gray
land  application site.  Prior to diverting treated wastewater  to  the Han-
cock  farm,  the Gray  farm  received  up  to 57,000 m-Vday  of  SeWRP  effluent
which was used  to irrigate approximately 1,200 ha (Wells et  al 1979).  This
was an average  annual  hydraulic loading of 1.7 m.  Water percolation to the
ground water created  a ground-water mound beneath the Gray farm^   Depth to
water  ranged from 4.6  m  (15 ft) to 22.8 m (75 ft).  Ground-water quality
beneath the Gray farm  was degraded  due to water management  problems and
inappropriate  cropping patterns.  Figure C.1 presents ground-water  level
contours beneath the  Gray farm in December 1981.  Flow occurs  from  the Gray
site  toward the north, east, and south.  During 1983, the  water  levels in
most of the observation  wells dropped (Table 9).~. In 1982 only  25 percent
of  SeWRP's total effluent  (16,650,613  m3) to be•used for  irrigation was
transported to  the Hancock farm.  From January 1, 1983 to  October  31, 1983,
the Gray  farm  received 11,406,287  m3  which was  69 percent of  the total
effluent applied to land. During 1982 and 1983 ground water  withdrawal from
beneath  the Gray farm by the City of Lubbock to provide recreational water
for its citizens remained relatively constant with withdrawal  rates in 1980
and 1981.   Consequently, the decline in ground-water beneath  the  Gray farm
probably was only slightly affected by the transfer of water  to  the Hancock
farm.  The  change in  cropping pattern  to "alfalfa may have been  the major
contributing factor to the decrease in ground water level.
Ground-water Quality—
     Prior  to  1982  hydraulic and nutrient overloading applied  to  the farm

                                  41

-------
TABLE  9.   STATISTICS OF DEPTH TO WATER IN OBSERVATION WELLS
               AT GRAY SITE DURING PROJECT

Depth to Water m (ft)
Well No.
6880
6881
6882
6883.
6884
6885
6886
6887
6888
6889
6890
6891
6892
6893
6896
6895A
6895B
6895C
6895D
7000
Initial
19.9
21.5
22.9
18.9
8.5
8.2
15.2
10.7
19.8
13.7
4.6
5.1
8.5
12.1
12.6
13.8
7.9
7.7
7.7
6.9
(65.3)
(70.6)
(75.0)
(62.0)
(28.0)
(26.8)
(50.0)
(35.0)
(65.0)
(45.0)
(15.0)
(16.7)
(28.0)
(39.6)
(41.3)
(45.3)
(25.8)
(25.4)
(25.3)
(22.8)
Minimum
10.61
21.3
22.3
18.6
5.3
5.0
15.2
10.7
19.8
13.2
3.4
4.1
3.4
10.1
3.6
9.7
5.9
5.9
6.1
5.9
(34.9)
(34.9)
(73.0)
(61.1)
(17.3)
(16.4)
(50.0)
(35.0)
(65.0)
(43.3)
(11.3)
(13.4)
(11.0)
(33.0)
(11.9)
(31.7)
(19.4)"
(19.5)
(19.9)
(19.3)
Maximum
19.9
22.1
25.6
21.0
10.7
10.1
16.7
13.2
21.3
14.4
5.8
6.1
' 12.2
12.1
15.0
13.8
8.3
8.3
8.7
8.3
(65.3)
(72.5)
(84.1)
(68.8)
(35.1)
(33.0)
(54.7)
(43.2) .
(70.0)
(47.4)
(19.1)
(20.0)
(40.0)
(39.8)
(49.1
(45.3)
(27.2)
'(27.2)
(28.4)
(27.2)
Final
17.7
21.5
22.2
18.9
9.2
8.4
16.5
12.8
21.2
14.3
4.3
5.6
8.9
11.3
12.8
9.8
8.0
8.0
8.2
7.9
(58.1)
(70.5)
(73.0)
(62.0)
(30.2)
(27.6)
(54.1)
(42.0)
(69.5)
(46.9)
(14.0)
(18.5)
• (29.2)
(37.0)
(42.0)
(32.0)
(26.2)
(26.2)
(26.8)
(26.0)
                           42

-------
had increased the quantity of ground water at  the  expense of ground-water
quality.   Table 10  presents the percent of well  water samples collected
from the  Gray wells which  contained constitutents  equaling or exceeding
drinking  water standards.  Due to improper nitrogen  management on the farm,
sufficient nitrate-nitrogen was leached to the ground  water to create water
quality problems (>10 mg N/l) throughout the entire  aquifer beneath  the
farm.  Several water samples contain  Se concentrations greater than the
drinking  water maximum constituent level (MCI) (0.01 mg/1).   Iron and  man-
ganese levels consistently equaled or exceeded   recommended  secondary con-
stitutent  levels of 0.3 mg/1 and 0.05 mg/1, respectively.  During the base-
line period hydraulic overloading leached salt from and through the soil
profile thereby increasing IDS, SO^ and Cl levels in the ground water above
recommended secondary constituent  levels with  regard to drinking  water
sources.
     The range of  the average concentration of.specific water quality con-
stituents  contained in the ground water beneath  the Gray farm during  the
baseline  monitoring period (June 1980 to February 1982) and  the irrigation
monitoring period (February 1982 to October 1983) is presented in Appendix
C. (Table  C.1).
     Nitrogen—Prior  to pumping  water  to  the  Hancock farm, the average
N02+N03-N  levels in the ground water beneath the Gray  farm ranged from 5.05
mg/1 to 35.89 mg/1. Figure 7 illustrates that  the  higher ground-water
N02+N03 were experienced in areas which were row watered or flood irrigated
and wells  down gradient from these areas.  Once  SeWRP's effluent was pumped
to the Hancock farm, ground-water N02+N03-N concentrations beneath the Gray
farm ranged from 0.77 mg/1 to 33.43 mg/1. Comparison of the  average  base-
line and  irrigation monitoring period nitrate data showed a nitrate de-
crease in  17 of 27  monitoring wells  while five  wells (6852, 6855,  6885,
6864,  and 6883) had an increase in average nitrate levels.  Excessive pre-
cipitation (approximately 35 cm) in May and June 1982  caused  surface runoff
which inundated a few poorly sealed well casings producing a  pulse increase
in total  Kjeldahl nitrogen (TKN). During the baseline  period, TKN  concen-
trations  varied from 0.28 mg/1. to 6.97 mg/1.  Once  water was diverted from
                                   43

-------
TABLE 10.  PERCENT OF GRAY FARM WELL  WATER SAMPLES WHICH EXCEED OR EQUAL
         DRINKING WATER STANDARDS FOR THE FOLLOWING PARAMETERS

Total Number of Wells = 25
Maximum Constituent Level

Date
06/25/80
08/19/80
09/25/80
01/08/81
03/27/81
06/02/81
10/28/81
11/02/81
01/27/82
05/27/82
10/11/82
11/01/82
05/19/83
10/10/83
*Values
No. of
Wells
11
9
24
. 25 '
1
24
19
3
25
23
20
5
25
25

AS
0
0
0
0
0
0
. 0
0
0
0
0
0
4
0

BA
0
0
0 .
0
0
0
0
0
0
0
0
0
0
0

CD
9
0
0
0
0
4
0
- 0
0
0
0
0
0
0

CR
0
11
0
0
0
0
0
0
0
0
0
0 ..
0
0

PB HG
0 27
11 11
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0

N03*
73
100
79
84
100
83
74
33
84
74
55
60
60
52

SE
0
0
0
0
0
4
5
33
0
0
5
0
4
0

AG
0
0
0
0
0
0
0
0
0
0
0
0
0
0
reported as nitrogen
Recommended Secondary

Date
06/25/80
08/19/80
09/25/80
01/08/81
03/27/81
06/02/81
10/28/81
11/02/81
01/27/82
05/27/82
10/11/82
1 1 /01 /82
05/19/83
10/10/83
No. of
Wells
11-
9
24
25
1
24
19
3
25
23
20
5
25
25
Constituent

CL
55
89
63
84
100
83
68
33
80
78
65
80
76
76
Levels

CU
0
0
0
0
0
0
0
0
0
0
0
0
0
0


FE
18
0
0
12
100
13
26
33
12
43
35
100
4
72


MN
82
0
0
8
100
13
26
33
16
17
15
30
16
12


S04 TDS
64 100
7 100
4- 92
56 96
0 100
63 96
53 79
0 67
60 92
48 91
45 80
60 100
64 88
36 88


ZN
0
0
0
0
0
0
0
0
0
0
0
0
0
0


































                               44

-------
                                                                         Well
                                                                      £-:: Water
                                                                      1 cm = O.27 km
Figure 7.  Nitrate  Concentration (mg/1) in Weil Water  under  Gray Farm,  Baseline Period, 1981-1982

-------
the Gray farm to  the Hancock farm, average TKN  levels ranged from 0.20 mg/1
to 7.65 mg/1.
     Phosphor us--Aver age total phosphorus (TP)  levels  in the ground water
beneath the Gray farm  ranged from 0.10 to 3.49 mg/1  during the baseline
period (Table C.1).  During the irrigation period,  from February 1982  to
October 1983, ground wate'r TP levels generally  remained relatively stable
with average concentrations ranging  from <0.07 to  1.94 mg/1.  Eighteen
wells  contained water  which  dropped  in TP levels during the irrigation
period.
     Organic Carbon—Ground water COD measurements were  quite variable from
June 1980 through October 1982 and appeared to  stabilize during 1983.  Dur-
ing the baseline  and irrigation periods, average COD  values ranged  from
27.2 mg/1  to 125.4 mg/1 and 11.4 mg/1 to 100.3 mg/1, respectively.  Median
COD values were  from 12.1 mg/1 to  159.1 mg/1  from  June 1980 to February
1982 and 10.3 mg/1 to 46.9.mg/1 during the irrigation period.  Approximate-
ly 80 percent of  the Gray wells contained less than  average COD in  their
respective ground water during.the irrigation  period compared to the base-
line COD concentrations.  Similarly, ground water TOC  concentrations  de-
creased from baseline period through the irrigation  period.
     Minerals—The  salt management  procedure  employed at the Gray farm
prior  to transporting water to the Hancock farm  was  leaching.  The average
concentration of  total dissolved solids (TDS) in  the  ground water beneath
the Gray farm varied from 1,010 mg/1 to 2,271 mg/1.
     Calcium, magnesium, potassium, and sodium  salts  were the primary  con-
tributors  to the ground-water dissolved solids beneath  the Gray farm.  The
average ground-water calcium levels ranged from 47.7  mg/1 to 161.7  mg/1
during  the baseline period.   No significant (a = 0.05)  differences in
ground-water calcium concentrations were determined  when comparing the data
obtained prior to February 1982 and the data from February 1982 to October
1983.  A comparison of average ground-water calcium  levels computed for the
baseline and irrigation periods, however, does  indicate a slight increase
in calcium levels in 72 percent of the wells. Furthermore,  magnesium  con-
centrations increased slightly in 19 of 25 wells. Average magnesium concen-

                                  46

-------
trations  were  from 21.8 mg/1 to 137.6 mg/1 from  June  1980 to February 198.2
and from 38.5 mg/1 to 148.7 mg/1 during the irrigation period.
     Associated  with  the increase in Ca and Mg in the ground water was a
decrease in Na  levels in 15 of 25 wells. . The combined reduction  in  Na  and
increases in  Mg and Ca produced a decrease in SARacji in 17 wells.  During
the baseline  period the  average  SARacj-j of ground  water from  25 wells
exceeded  9 with well 6893  exhibiting  the highest  SARadj value of 23.9.
Severe sodic problems would develop in the soils  of  adjacent farms which
utilized  this  ground-water  source  for  irrigation.  As expected, ground
water  beneath the farm areas which were  historically  irrigated by flood or
ridge  and furrow  method (row water) contained the higher, sodium levels.  In
1982 the cropping pattern was changed and alfalfa became  the primary  crop
grown.   A more even and  reduced hydraulic distribution of water over the
entire farm associated with higher evapotranspiration  reduced leaching  of
sodium from the upper soil profile.
     During the  irrigation period, significant increases in ground-water
iron concentration were measured in several wells.  Saturated soil  condi-
tions  in May and  June 1982 caused reduction of iron to soluble ferrous iron
and subsequent  percolate water transported increasing  quantities of  ferrous
iron  to  the ground water.  Percolation resulting from heavy precipitation
events in May and August/September 1981  leached  sufficient quantities of
iron to the ground water to exceed or equal drinking water standards.
     Major anions associated with the salts were  chlorides and  sulfates.
Average chloride  concentrations ranged from 208 mg/1 to 535 mg/1 during the
baseline monitoring period and 154 mg/1  to 686 mg/1 during the irrigation
period.   Before effluent  was pumped to the Hancock  farm, average  ground-
water  sul fate concentrations varied from 149 mg/1 to  795 mg/1.   Once  the
hydraulic loading was reduced, 148  mg/1 to 399 mg/1   was the range of
average  ground-water sulfate (504) concentrations measured in the ground
water  beneath the Gray farm.  Use of the ground water  located beneath  the
flood  or  row water irrigated areas for sprinkler irrigation of alfalfa and
grain  sorghum may cause foliar injury.
     Trace Metals—The majority of trace metals analyzed in each water sam-
                                  47

-------
pie were  at low concentrations.  This was  anticipated since the irrigation
stream contained low levels of trace  metals and the  soil  matrix had  the
ability  to  remove most metals.  Nonetheless, certain metals did increase
significantly in the ground-water samples obtained  from certain wells and/
or equaled  or exceeded   drinking water  MCls.   The data show that trace
metals present in the ground water posed no potential  public health risk.
With  the  alkaline soils  which existed at  the  Gray farm,  anionic heavy
metals such  as As and Se were more soluble and consequently an apparent
association  was observed with rainfall events  and slight As and Se increase
in the ground water.  Colloidal particulates  entering  the well casing dur-
ing construction or after  heavy rainfall  events may have caused slight
pulse increases in Pb, Cr, Cd, and Co concentration in ground water   ob-
tained from  a few wells.
     Priority  Organic  Pollutants—The ground water beneath the Gray farm
contained very low levels  of specific priority  organic pollutants (POP)
assayed.   Contamination of ground water by  priority organic pollutants  was
not a  problem during the study period.  The  soil matrix was very efficient
in removing  and biologically degrading these  organics.
     Statistical  analysis of propazine'levels obtained from wells 6884  and
6892 showed  an increase in propazine  concentration during  the irrigation
period.   Propazine was a common herbicide  used to control weeds in grain
sorghum crop production.  Consequently,  the small increase in propazine
could  have  been attributable to leaching of the  herbicide during the  May
precipitation or transport of a small quantity of particulates containing
propazine directly into the well or  through  the gravel packing surrounding
the well  casing.
     Bacteriological  data—Bacterial indicator organisms, total coliform,
and fecal streptococci were assayed in each  water sample to determine  the
potential  presence of pathogenic organisms.   Salmonella  was isolated  in
four wells during the baseline period.  Three  of five Salmonella isolations
were measured during the fall 1981.  Contamination  of  these wells by Sal-
monella and the indicator  organisms was associated with heavy rainfall
events.  Similarly, heavy precipitation in  May and  June 1982 caused  flood-
                                  48

-------
ing of  well  6884 which  introduced Salmonella to water contained in the
poorly sealed well.   The  presence of Salmonella in  this water sample was
the only isolation measured from February 1982  to  October 1983.
Soils—
     The Gray  soils contained a higher  percentage  of coarse  material
throughout  the  upper  122  cm of  the profile than measured in the soil  pro-
file at ,the Hancock  farm.  The  predominate soil texture  within the upper 30
cm of the profile was sandy loam.  Sandy clay loams  existed  from a depth of
30 cm  to  91  cm.  Soil texture  at depths greater than 122 cm varied primar-
ily from clay to clay loams.  An  indurated caliche  soil was observed at
depths from 40  to 183 cm.
     Nitrogen—The  Gray farm  crops were predominantly cotton in 1981 and
alfalfa in 1982 and 1983.   Nitrate nitrogen concentrations were fairly
uniform throughout the entire 183 cm soil depth in 1981. Analysis of  soils
in 1983, however, indicated a decrease in nitrate  levels in  the upper  91 cm
which resulted  from  greater nitrogen uptake by  alfalfa.   Ammonia  nitrogen
existed primarily in  the  top 30 cm of soil.  No appreciable  change in  ammo-
nia concentrations was observed between soil samples  collected in  1981 and
1983.  The  bulk of the soil nitrogen was incorporated in organic matter.  A
reduction of organic  nitrogen from the 1981 levels was measured in 1983.  A
nitrogen  balance on  the  soil indicated that nitrogen uptake by crops  grown
beneath the center pivot  irrigation machine was the  major mechanism govern-
ing nitrogen losses.   Deep  percolation of inorganic nitrogen beneath the
center pivot machines was not a major mechanism of nitrogen  loss in 1982 or
1983.   The results  of  the  nitrogen mass balance  for  the  flood  irrigated
area showed deep percolation of nitrates to the  ground water remained a
significant process for the loss of nitrogen  from  the  soil profile in the
flood or row water areas. The  ground-water quality  data also substantiate
this conclusion.
     Phosphorus—Phosphate-calcite reactions were probably an  important
mechanism  for the  removal  of phosphorus from the  soil solution.  A  total
phosphorus mass balance  showed  annual phosphorus removal by  crops  in  1981,
                                 49

-------
1982  and  1983 was less than the mass applied through irrigation.  Only 33
percent of the applied  TP was consumed by crops.  The remainder  was  proba-
bly fixed  into the  soil matrix.
     The phosphorus  uptake by wheat (flood irrigated)  was  only  12 percent
of the estimated  1,072  kg/ha applied from 1981 through 1983.  Fixation of
phosphorus may have been the major mechanism which governed phosphorus re-
moval from the soil solution.
     Minerals—Total dissolved solids in the soil matrix beneath  the center
pivot  machines increased  gradually with depth in 1981;  whereas, frequent
leaching of salts in the wheat -areas produced a more uniform TDS  concentra-
tion  throughout  the entire  soil profile.  A TDS  mass  balance indicated
flood and  row irrigation and precipitation leached salts below  a depth of.
183 cm  during the  period from spring 1981 through 1983.  Chloride and sul-
fate mass  balances  further substantiated the deep percolation  of  anions
past  183  cm depth  in the flood irrigated areas.  The salts  applied to cot-'
ton and alfalfa areas,  however, were somewhat retained in  the soil profile.
     Increases in  TDS  levels measured in 1983 did not appear to  be associ-
ated with  increases in  sodium ion.  A slight increase in the average sodium
concentration was  measured at depths of 61  and 91 cm in 1981.   Sodium lev-
els were fairly uniform throughout the soil profile in 1983.   Based  on a
cation  exchange  capacity  (CEC) in the upper 30 cm of 22  meq/100g an ex-
changeable sodium percentage (ESP) of approximately seven  was computed for
sodium  in  the top  30 cm beneath the spray irrigated area.   Therefore, sod-
ium levels were maintained sufficiently low by  leaching to  prevent  sodic
conditions (ESP  <15)  in  the  soil.  Similarly, the ESP value for 1981 and
1983 in the upper 30 cm of the flood or row water areas  was  seven.
     Trace  Metals--Table 11 presents the  total average mass  of specific
trace metals applied by flood and row irrigation.  Over  a  three  year period
from 1981  through 1983, input of trace metals  through irrigation  to the
soil  was  low. Compared to  the concentration of metals  in the  soils, the
concentration of metals in the crop tissue was negligible.   Anionic  metals
such as arsenic and chromium (VI) probably were transported  by percolate to
depths greater than 91  cm within the alkaline soils.  In addition,  barium,

                                  50

-------
copper  and  nickel possibly leached beyond the  91  cm soil depth.  Possible
accumulation of  cadmium, cobalt and  lead occurred  in  the upper  91  cm of
soil in the  flood  irrigated areas.  Special variability had the most impact
on deficiencies  in trace metal levels measured in  1981  and 1983.
     Priority Drganics Pollutants—The majority of soil samples analyzed
for specific priority  organic compounds,  contained levels below their
respective detection limits.  In both the spray  and  flood  irrigated  areas,
solvents  such as  benzene and chloroform existed throughout the 183 cm soil
profile in 1981.
          TABLE  11.   TRACE METALS MASS BALANCE ON  SOILS  COLLECTED
                        FROM FLOOD IRRIGATED AREA

Trace
Metal
As
Ba
Cfl
Co
Cr
Cu
Pb
Ni
Mass Applied
(kg/ha)
1.887
26.96
0.276
0.358
4.820
5.167
2.962
4.235
Mass in Soil
(kg/ha)
1981
85.0
2351
2.56
36.8
593.7
202.2
34.4
175.0
Profile
1983
26.9 .
1267
4.40
90.1
189.4
83.3
101.4
118.6
Change in
Soil Profile
(kg/ha)
-58.1
-1084
+1.84
+53.3
-404.3
-118.9
+67.0
-56.4
Unaccounted
Mass (kg/ha)
-60.0
-11.11
+1.56
+52.9
-409.1
-124.1
+64.3
-60.6

Benzene  levels decreased  to levels barely exceeding  detection limits in
1983.  In 1983,  however, chloroform levels appeared to have  increased above
baseline levels.  Furthermore, chloroform concentration increased with
depth.
     Carbon tetrachloride and tetrachloroethylene were not measured at con-
centrations above  detection limits  in  1981, but  were  detected  at  levels
exceeding analytical  limits in practically all samples collected  from flood
and spray irrigated areas in 1983.
                                  51

-------
     Chlorinated  aniline, compounds such as  2,3-dichloroaniline and 3,4-
dichloroaniline  may  have  been derivatives of herbicides or fungicides.  In
general,  2,3-dichloroaniline was more prevalent  at  soil depths from 91  cm
to 183 cm beneath  the center pivot machines in 1983  than in 1981.  Further-
more,  less  amounts of chlorinatedanilines were detected in the flood irri-
gated soils.
     Acenaphthylene  and the various dichlorobenzene  forms in the soils may
have been  derived from application  of insecticides.  These  organic com-
pounds existed primarily in the upper 61  cm of the soil profile.
     The presence  of  trace organics in the soil profile  in 1981 and  1983
may  have  been  directly related to the mass loadings of specific organics
through flood irrigation  of wheat.   Decreased organic  mass loadings  by
spray irrigation,  however, reduced the potential of  residual levels of pri-
ority organic compounds in 1983 being derived solely  from irrigation.
    Bacteriological  Data—Total  coliform,  (-TC) fecal coliform (PC) and
fecal  streptococcus  (FS)  were primarily retained  in  the top 30 cm of soil.
Furthermore,  FS  was detected at greater concentrations  in  the spray  irri-
gated  areas  in  1983 than in 1981.  Certain  soil  samples obtained from the
flood irrigated areas  contained  concentrations of TC  and FS exceeding
detection limits at depths of 183 cm.
     Fungi and  actinomycetes concentrations within the soil profile were
relatively constant throughout the farm and the monitoring period.  Changes
in hydraulic  loading  or cropping patterns did not  appear to affect the num-
ber of actinomycetes  or fungi present  throughout the  entire 183 cm  soil
core.
Crops—
     Cotton yields  prior to 1982 equaled  or exceeded irrigated Lubbock
County yield  averages of 434 to 457 kg/ha in  1980  and 1981,  respectively.
Alfalfa  yields  in 1982 (crop  establishment year)  ranged from 1.8 to 2.7
metric tons/ha which  was relatively low compared to  a normal range  of 5.4
to 7.0  metric tons/ha.  The yield reduction  was partially due to 1) delays
in watering,  thereby  delaying regrowth; 2) poor stand establishment; and 3)
weed competition.   Weed  competition also affected  the quality of the hay

                                  52

-------
 harvested; thus lowering the crop's  marketability.  The 1983 alfalfa yields
 on  the Gray farm were again reduced  by  delays  in watering which reduced the
 number of cuttings.  Normally,  five  to seven cuttings of alfalfa are ex-
 pected in the Lubbock area (Texas  A  & M Extension Service).  The Gray  farm
 produced three  to four cuttings.   Alfalfa  yields were slightly improved
 over  1982 values.  Soybean yields  averaged  2,494 kg/ha  (37  bu/ac).  Wheat
 yields from the  Gray farm were difficult  to  determine since no grain har-
 vest  was planned and forage harvest  was accomplished by grazing.  Analysis
 of  crop tissue  indicated  macro  and  micro nutrients were  within normal
.ranges.  Potential toxic trace metals did not.  appear to  accumulate  in  the
 crop  tissue.
 Hancock Farm
 Ground-water Levels—
      The  saturated  thickness of the  aquifer beneath the Hancock farm was
 less  than'6.1 m (20 ft).  Depth to water varied from 24.4 to 36.6 m  (80 to
 120 ft).  In 1981 ground-water flow occurred  toward both the north and the
 south from a ridge through the center of the property (Figure C.2).
      Ground-water levels increased beneath  the Hancock site during the per-
 iod from January 1982 to October 1983 (Table  12).  Ground-water recharge
 from  surface runoff contained in  the moat  areas surrounding the reservoirs
 and through coarse material along  the fringes  of the playa lakes most like-
 ly  caused the water level rises beneath the farm.  Irrigation amounts dur-
 ing the study  period  were far less than  crop  evapotranspiration  rates;
 consequently  irrigation . does not  appear  to be a source of  ground-water
 recharge.
 Ground-water Quality—
      Prior  to land  application of  treated wastewater at the Hancock farm,
 the range of average concentration of constituents measured in the  ground
 water beneath  the  Hancock  farm is presented in Table C.2.  Several well
 water samples contained levels  of contaminants which exceeded or equaled
 drinking water maximum constituent levels (Table 13).
      Nitrogen--Associated with  the deep  percolation of rainfall surface
                                   53

-------
TABLE 12.  STATISTICS OF DEPTH TO WATER IN OBSERVATION WELLS AT
                  HANCOCK SITE DURING PROJECT

Depth to Water m (ft)
Well No.
10112
10211
10521
10721
10731
10821
10842
10931
10932
11032
20112
20243
20711
20721
20842
21141
21152
21234
21323
30312
40231
40331
Initial
34.0
37.0
24.2
25.9
24.7
21.9
26.9
21.5
20.6
28.4
35.7
41.5
31.4
27.4
30.5
27.4
26.3
26.3
29.6
32.0
30.2
29.6
(111.4)
(121.5)
( 79.3)
(- 85.0)
( 81.0)
( 72.0)
( 88.1)
( 70.4)
(67.7)
( 93.1)
(117.0)
(136.0)
(103.0)
( 90.0)
(100.1)
( 90.0)
( 86.2)
( 86.3)
( 97.1)
(105.0)
( 99.0)
( 97.0)
Minimum
27.9
36.5
22.9
19.1
19.0
18-. 9
20.6
17.4
13.4
27.2
33.7
41.5
31.1
24.8
25.3
25.0
22.0
22.7
29.0
31.4
29.6
28.2
- ( 91.6)
(119.9)
( 75.0)
( 62.6)
( 62.4')
( 67.7)
.( 62.0)
( 57.0)
( 43.9)
('89.3)
(110.7)
(136.0)
(102.0)
( 81.3)
( 83.0)
( 81.9)
( 72.1)
( 74.4)
( 95.0)
(103.1)
( 97.0)
( 92.7)
Maximum
34.4
40.3
25.8
28.1
25.2
27.3
27.3
21.5
20.7
28.7
36.1
44.6
32.4
29.5
30.7
27.7
26.3
26.6
30.2
32.4
30.8
29.6
(113.0)
(132.1)
( 84.8)
( 92.2)
( 82.8)
( 72.0)
( 89.7)
( 70.5)
( 67.9)
( 94.3)
(118.3)
(146.3)
(106.3)
( 96.7)
(100.8)
( 91 .0) "
( 86.4)
( 87.3)
( 99.1)
(106.3)
(101.0)
( 97.0)
I
32.6
36.6
22.9
19.1
19.0
18.9
24.9
18.3
15.5
27.4
34.2
41 .6
31.2
24.8
28.0
25.0
22.4
22.7
29.0
31.5
29.8
28.9
•inal
(106.8)
(120.0)
( 75.0)
( 62.6)
.( 62.4)
( 62.0)
( 81.8)
( 59.9)
.( 51.0)
( 90.0)
(112.2)
(136.4)
(102.2)
( 81.3)
( 92.0)
( 82.0)
( 73.5)
( 74.4)
( 95.1)
(103.4)
( 97.7)
( 94.8)
                         54

-------
TABLE 13.  PERCENT OF HANCOCK FARM WELL WATER SAMPLES WHICH EXCEED OR EQUAL
           DRINKING WATER STANDARDS FOR THE FOLLOWING PARAMETERS

Total Number of Wells =27
Maximum Constituent Level

Date
07/22/80
10/30/80
11/11/80
01/26/81
03/27/81
06/11/81
10/29/81
11/18/81
01/18/82
06/14/82
09/22/82
11/10/82
02/15/83
03/14/83
05/09/83
07/20/83
09/15/83
No. of
Wells
12
7
16
23
2
24
2
22
27
27
25
2
7
12
27
1
27

AS
0
0
0
0
0
0
0
5
0
0
0
0
0
0
0
0
0

BA
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

CD
17
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

CR
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

PB
8
0
0
0
0
0
0
0
0
4
0
0
0
0
0
0
0

HG
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

N03
8
0
25
0
0
13
0
27
4
11
12
0
57
8
15
0
7

SE
0
29
0
0
0
4
50
18
19
15
0
0
0
0
15
0
11

AG
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
                                                                              (continued)

-------
Table 13, continued
Recommended
Date
07/22/80
10/30/80
11/11/80
01/26/81
03/27/81
06/11/81
10/29/81
•* 11/18/81
01/18/82
06/14/82
09/22/82
11/01/82
02/15/83
03/14/83
05/09/83
07/20/83
09/15/83
Secondary Constituent Levels
No. of
Wells
12
7
16
23
2
24
2
22
27
27
25
2
7
12
27
1
27
CL
0
0
0
0
0
0
0
9
0
4
0
0
0
0
4
0
7
CU
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
FE
25
14
0
22
0
4
100
73
85
96
92
100
0
0
59
0
89
MN
67
71
25
30
100
42
50
45
52
52
44
100
0
0
30
0
48
so,
0
0
0
0
0
4
0
0
0
4
8
0
14
0
4
0
4
IDS
8
0
0
0
0
4
0
0
4
.7
0
0
29
0
7
0
11
ZN
0
0
0
0
0
4
0
0
0
0
0
0
0
0
0
0
n

-------
runoff captured in  playa  lakes or moat areas surrounding the  storage  reser-
voirs was the leaching  of nitrate nitrogen to  the  ground water.   Nitrate
levels  in water obtained  from seven wells, which  were adjacent to surface
runoff collection  areas,  increased after heavy precipitation  events.  The
rapid  increase in ground-water nitrate  concentrations may  have resulted
from migration  of  percolate around poorly sealed well  casings  to the
ground  water.   Most  of the ground-water samples collected during the  moni-
toring period contained less  than one ppm TKN.
     Phosphorus--During  the baseline monitoring period, average TP  levels
measured in the ground  water  ranged  from 0.19 mg/1  to 0.58 mg/1.  -The  aver-
age ground-water TP concentration from February 1982 through  October 1983
ranged  from 0.02  mg/1 to  0.41 mg/1.  The data indicate a decreasing  trend
in TP from baseline through irrigation.  This decrease was primarily  due to
a decrease in organic phosphorus in  the ground-water.
     There  appeared  to  be an increase in TP and PO^ in most  wells in June
1982.  During the  baseline  period ground-water P04  levels ranged from  <0.01
to 0.22 mg/1.  From February  1982 to October 1983,  the majority  of  water
samples analyzed contained  less than 0.10 mg/1 orthophosphate  phosphorus.
     Organic  Carbon—Median  COD  concentrations  during the baseline and
irrigation periods  ranged from 4.1  mg/1 to 94 mg/1  and 3.8 mg/1 to 88  mg/1,
respectively. High  COD  levels were primarily the result of soil  entering
the well during construction.  Ground-water organic carbon levels decreased
during the period  from  June 1980 to  October 1983.
     Minerals—Pripr to  transporting SeWRP's effluent to the  Hancock  farm,
the  average TDS in  the  ground water beneath the farm ranged  from 363 mg/1
to 989  mg/1.   As  was  observed with  increases in ground-water  nitrate
levels,  increases  in TDS in  water obtained from a  few wells  appeared  to be
associated with heavy precipitation  events.
     Comparison of baseline and irrigation period average sodium adsorption
ratio (SAP)  for the  water extracted  from each well indicate a  general
change in composition of  salts.  From June 1980 to  February 1982, the  aver-
age  adjusted SARs  (SARacy) of the ground water ranged from 3.0 to 8.4.  To
prevent the creation  of sodic soils, irrigation water should  have an  SARa(jj

                                  57

-------
below  six (Stromberg  and Tisdale  1971).  Eleven wells contain water -with
average SARacn less than six.   Increasing problems, however, may occur with
SARac|j  from six to nine.   Computed  SARacjj values for ground water obtained
from 24 wells were between six  and nine.   Therefore,  no  severe problems
with water penetration were indicated by the data.  During the irrigation
monitoring period, SARacjj  values  ranged from 2.1 to 11.0.  Approximately 60
percent of the wells demonstrated an increase in Ca and Mg, and Na  was  re-
duced  in  79 percent of the wells during the irrigation period.  Therefore,
in 18 wells the SARacjj  was lowered during the irrigation period.
     In general, no significant changes in ground-water  chloride ion  con-
centrations was  detected throughout the monitoring period.  A significant
increase in Cl ion levels  was observed in the monitoring wells adjacent  to
the moat  area surrounding reservoir 2.  During the baseline period, chlor-
ide levels ranged from  22  mg/1  (0.6  meq/1.) to 246 mg/1 (6.9 meq/1) and from
17 mg/1  (0.4 meq/1) to 345 mg/1 (9.7 meq/1) from February 1982 to October
1983.  Chlorides present in the ground water would not cause foliar  injury
to crops  grown on the  Hancock  farm  which were primarily cotton, grain sor-
ghum, and alfalfa.
     Trace Metals—As previously  stated, trace metals were not a major con-
cern due to the limited industrial waste water contribution to SeWRP's sew-
erage load and the "ability of the alkaline calcareous soil profile  to  ade-
quately  remove  and  render relatively insoluble most trace metals.  Trace
metals  posed no potential  agricultural or public health problems during the
project monitoring period.
     Priority Organic Pollutants—In general, slow rate land application of
organic  compounds contained in municipal waste water should pose no hazard
to ground water, soil microbial community and  vegetation  (Overcash  1983,
Davidson et al 1980).  The ground-water data confirms this statement. Major
priority  organic  compounds measured  in the  ground-water  samples  were
phthalates (i.e., dibutylphthalate,  diethylphthalate, diisooctylphthalate).
Phthalates are used as   plasticizers in polymers and migrate quite  readily
to the  surrounding environment.   Consequently, phthalate contamination may
have been an artifact of well construction, sample collection, and presence

                                  58

-------
of plastics within  the  laboratory.
     Ground water in  well  10521, 10931, 10721, 30312,  and 10721  experienced
a pulse increase in atrazine  in 1983.  Average atrazine concentrations in
wells 10521,. 10931, 10721, 30312, and 10731'were all less than  2.0  ppb dur-
ing the baseline period and 13.9 ppb, 2.5 ppb, 11.9 ppb,  10.8 ppb,  and 12.6
ppb, respectively,  from  February 1982 to October 1983.   Atrazine  was used
to kill weeds and grasses  in  borrow ditches, and around each  center pivot
irrigation machine.   In addition, patches of weeds surrounding  playa lakes
or extending into fields were treated with atrazine.
     Bacteriological data—During the baseline period,  indicator organisms
were isolated  in  water  from more  than  85 percent of the  wells.   A small
biochemical study identified  the fecal streptococci organisms to have been
possibly  S. faecalis  subspecies liquefaciens and not of human source.  A
potential  source  of Salmonella, fecal coliform and fecal streptococci was
most likely rodents which  burrowed beneath concrete pads  surrounding well
casings.   Detection of  Salmonella  in' water samples decreased during the
irrigation  period.  In general, very little difference  was observed  between
bacteriological  data  obtained during the baseline and  irrigation monitoring
periods.
Soils—
     At .the Hancock farm,  the soil texture within the  upper  30  cm (1  ft) of
the soil profile was  generally sandy clay loam.  Clay  to  clay  loams domi-
nated  the  soils from a depth of 30 cm to 122 cm (4 ft)  within  the  profile.
The majority of  soils from 122 cm to 183 cm (6 ft) were clays.   An  indurate
layer  of calcium  carbonate  (caliche) existed within  the soil  profile at a
depth of 61 cm to 183 cm throughout the farm.
     Soils  at the Hancock  farm were alkaline and calcareous  with pH values,
within  the  upper 183  cm, of seven  to eight.   Cation  exchange  capacities
(CEC)  were greater  than  20  meq/100 g (average 22.4 meq/100 g  + 3.7) which
were characteristic of  the clay/clay loam soils.
     Nitrogen—The bulk  of  the  nitrogen  in  the soil  profile was in the
organic form, which  appeared to decrease linearly through the  upper 152 cm
of the  profile.   Carbon to nitrogen (C/N) ratios of the  organic matter

                                  59

-------
ranged  from three to 47 with only three percent of 235  cores having a C/N
ratio greater than  20.   C/N ratios  less than 22  are associated  with  net
mineralization and ratios higher than 22 indicate net  immobilization.  The
average C/N ratios  of the effluent pumped to the farm and from  the  reser-
voirs  were 4.0 and 5.9,  respectively.   Therefore, net mineralization-of
organic nitrogen predominated within the soil profile.  Nitrate-nitrogen
was  the  major inorganic  nitrogen form  within the soil profile.  Nitrate
lenses were detected within the lower 91 cm of several  soil cores (Figure
8).   Low moisture conditions due to  the semiarid climate of  the South
Plains  may have inhibited decomposition of organic matter and denitrifica-
tion of nitrate-nitrogen.
     Nitrogen  mass balances  were  conducted on the soil-crop matrix.  The
spacial variability of the data in conjunction  with the  error  associated
with the assumptions imposed on the model produced highly variable results.
At the lowest annual hydraulic loading (42.2 cm) the processes included in
the nitrogen model  described most of the nitrogen transformation within the
soils (Figure 9). Increased nitrogen losses due  to denitrification,  vola-
tilization, or  possible  leaching were not accounted for in the mass balance
at the 52.2 cm/yr irrigation loading.
     The nitrogen  mass balance model when applied to  the area  of the  farm
irrigated with  an average hydraulic loading of 68.9 cm/yr, predicted  inor-
ganic  nitrogen mass in the  profile of  302 kg/ha  compared to a measured
average of 277  kg N/ha.  In general, the assumption that,  inorganic nitrogen
was  not  leached from the  upper 183 cm of the  profile appeared to be sub-
stantiated by the nitrogen mass balance.
     Phosphorus—Phosphate-calcite  reactions  were believed to be a major
factor  in  the  removal of phosphorus from the soil solution.  The soil pro-
files throughout the farm denote a general decrease in  TP from 1981  levels
to 1983 levels.
     A phosphorus mass balance on the soils at each hydraulic loading show-
ed that the crops utilized more phosphorus than  was applied in 1982.   In
1983,  the  mass of  phosphorus removed by crops was less than applied.  Fur-
thermore, the amount of  phosphorus removed by the cotton  was less  than  the
                                 60

-------
   o
   =»*.
   o
   o.

   OJ
   o
   tO
J—o
Q_ c\J
o
  o
  oo
  o
  =f_
  o
  o
                                             Hancock Farm
                                                                             D Soil Core 02003


                                                                             O Soil Core 05071


                                                                             A Soil Core 06043
    0.00
Figure 8.
    0.03
0.07
0.10
o.m
0.17
  I

0.21

  -i
                        NITRITE + NITRATE (MG N/G)*1O

Illustration of Nitrite+Nitrate  Lenses in Hancock Soil, 1981
0.2U
0.28

-------
                                    N  in Root Zone  in 19U1
                                    N  from Organic  N in Root Zone
                                    N  Applied in Effluent
                                    Removed by Crop
                                    Denitrification
                                    Measured level  in Profile 1983
                                    Difference between Measured  and
                                    Predicted - Potential Leaching
400

300

(0
^200

c
0
O)
o 100
4-1
z
'c +


O
£

100

200
300
400
, 42.2cm ^52. 2 cm , 68.9cm



9
\
I !
! **
i
i
•
h
1
!Si

1
1
1


'



S
•I
| S \

. _
IS
• S



i! 1




NITROGEN
' SOURCES







1
1
1
I

NIT
LO
1

1












ROGEN
SSES


Figure 9.   Inorganic  Nitrogen in 183 cm  Profile  at the Hancock Farm
                                     62

-------
normal crop requirement of the 15 kg/ha.yr to 34 kg/ha.yr.
     Minerals—In general,  salts accumulated in the  upper  122 -cm of the
profile.   In the double cropped areas (68.9 cm hydraulic  loading)  IDS lev-
els increased at  the  152 cm and 183 cm depths, where no increase  at  these
depths  was detected  in soils irrigated with less amounts  of water.  Assum-
ing negligible crop uptake of salts, a mass balance of the IDS in  the  soil
profile,  indicated  the majority of applied salts were  retained within the
183 cm soil zone.
     Sodium salts composed most of the salt load to the  Hancock farm.  A Na
mass balance indicated that sodium was retained in the soil profile.   The
exchangeable sodium  percentage (ESP) was approximately  two in the  upper 30
cm in 1981  and was increased to  a maximum of  six in the  double  cropped
areas.   Future use of SeWRP's effluent without proper management of sodium
in the soil profile may produce sodic soil (ESP > 15) in the upper  30 cm.
     Both  Cl  and SO^ ions accumulated within  the upper 122 cm of  the soil
profile.  Average Cl  levels ranged from 12 to 70 mg/1 in soil  samples  col-
lected in 1981 and 38 to 162 mg/1 in soils obtained in the fall of  1983 and
winter of 1984.  The  Cl concentration increased throughout  the entire 183
cm soil profile  beneath  areas where  the highest  Cl mass loading was ap-
plied.  The majority  of soils analyzed contained Cl levels at  the lower end
of the normal range of 50 to 500 mg/1.
     Trace Metals—Due to the low levels of metals in the  wastewater, trace
metals  were not  considered to pose a problem to crops or  public health.   A
mass balance computed on the trace metals (Table 14) indicated the change
in metals levels  from 1981 to 1983 was not attributable to irrigation.
     Priority Organics--The majority of samples analyzed for priority or-
ganics  in  1981 and 1983 contained organic compounds at  levels below detec-
tion levels of the analytical procedure.  Atrazine, common in herbicides
used  on the  farms,  was measured in some soil samples in  1981  but  was less
than detection levels in 1983.  Both 2,3-dichloroaniline and 3,4-dichloro-
aniline were detected in a few samples in the upper 30 cm of soil.  These
organic compounds were probably degradation  products of  the trifluralin
herbicide  which  was commonly used  on the farm.   Benzene and chloroform

                                  63

-------
                        TABLE 14.   METALS MASS BALANCE FOR  HANCOCK  FARM



42.2 cm
As
Ba
Cd
Co
Cr
Cu
Tl
Pb '
Ni
Se
Zn
52.2 cm
As
Ba
Cd
Co
Cr
Cu
Ti
Pb
Nl
Se
Zn
68.9
As
Ba
Cd
Co
Cr
Cu
Tl
Pb
Ni
Se
Zn
Total Mass
Applied
(kg/ ha)
a
0.026
0.165
0.009
0.138
0.032
0.055
0.021
. 0.091
0.06
0.021
0.474
0.32
0.203
0.012
0.026
0.045
0.068
0.026
0.111
• 0.073
0.026
0.585
0.041
0.257
0.012
0.034
0.054
0.087
0.034
0.117
0.073
0.026
0.585
Soil Profile Mass
(kg/ha)
1981
b
108.0

1.79
56.32
317.9
147.1
24.7
64.1
288.89
4.694

152.55
6815 '
1.750
62.26
377.65
140.95
22.62
71.77
184.8

1181.0
122.30
9532.9
1.37
56.37
218.35
143.08

48.09
184.8

1181.0
1983
c
170.4

2.13
91.87
206.3
93.9
22.6
81.3
172.82
6.4

124.30
1284
2.688
93.84
158.14
75.53
12.80
72.41
137.8

957.9
337.83
1566.1
2.65
118.37
200.00
81.29

124.43
137.8

957.9
4 in Profile
(kg/ha)
d = c - b
+62.4

+0.34
+35.55
-111.6
-53.2
-2.1
+ 17.2
-110.07
+I.706

-28.25
-5531
• +0.938
+31.58
-219.51
-65.42
-9.82
+0.64
-47.0

-223.1
+215.53
-7966.8
+1.28
+62
-18.35
-61.79

+76.34
-47.0

-223.1
Unaccounted
Mass
( kg/ha)
e = d - a
+58.8

+0.331
+35.41
-111.6
-53.3
-2.1
+17.1
-110.13
+1.685

-28.57
-5531
+0.926
31.55
-219.55
-65.49
-9.84
+0.53
-47.

-224
+215.49
-7967.1
+ 1,27
+62
-18.40
-61.88

+76.22
-47.

-224
Percent Error
e
(b + a)
54

18
63
35
36
8
27
38
36

19
81
53
51
58
86
43
0.7
' 25

19
176
84
92
110
8
43

158
25

19

Mass Balance Computed  on  91  cm  of Soil
                                                64

-------
within  the profile  in 1981 and 1983 was most likely used as  a solvent  for
herbicides sprayed  on the land.  Another solvent, tetrachloroethylene  and
carbon tetrachloride,  was measured above detection limits in 1983.   Organic
compounds used  as insecticides (i.e., acenaphthylene, m-dichlorobenzene,
p-dichlorobenzene,  and p-dichlorobenzene) were also isolated in the  upper
30 cm of soil in 1981.  The mass of each organic in the  irrigation stream
contributed very  little to  the  mass  detected  in  the soil  profile.

     Bacteriological  Data—Irrigation with effluent apparently did  increase
the concentration of  coliform bacteria in the -upper 30 to 61 cm of the soil
profile.  Similarly,  fecal streptococcus was detected in the upper  91  cm at
levels greater  than analytical limits in 1983 in soils collected from  areas
receiving 52.2  cm and 68.9 cm of treated sewage per year.  Due to the fre-
quency  of isolation  of  fecal coliform above detection limits in samples
from the irrigation wastewater and soil, fecal coliform  appears to have  a
higher die-off  rate than fecal streptococcus.
     Average actinomycetes levels within the soil profile ranged  from  10^
to 10^2 counts  per  gram of soil.  During the irrigation period, actinomyce-
tes within the  upper  91 cm experienced a one-to-two log increase in concen-
tration.   Since an  increase in actinomycetes normally follows increased
bacterial and mold  growth, the rise in actinomycetes indicated a general
increase in biological activity in the upper 91 cm.
Crops—
     In 1982 the Hancock farm was planted in three crops:  sunflowers, soy-
beans, and grain sorghum.  For late planted crops, sunflower yields were in
the  normal range recorded for the High Plains, 1,124 to 2,247 kg/ha (1,000
to 2,000 Ibs/ac).   Soybean production ranged from 2,036 kg/ha (30.2 bu/ac)
under Pivot 19  up to  2,697 kg/ha (40 bu/ac) on Pivot 2 which was within  the
High Plains soybean range of 1,685 to 2,697 kg/ha (25 to 40 bu/ac)  cited by
Texas  A & M Extension Service.  Grain sorghum yields were less than  ex-
pected and ranged from 3,307 kg/ha (2,943 Ib/ac) beneath  Pivot 9 to  6,691
kg/ha (5,944 Ib/ac) under Pivot 19.
     The  1980,  1981  and 1983 cotton yields obtained from the Hancock farm
                                  65

-------
are  presented in  Table  15.  A definite improvement in crop production  was
experienced in .1983.   Regardless of irrigation, an increase  in production
was  anticipated since 1'983 was the only year during the study when  a nat-
ural disaster did not  affect crop planting and establishment.  In general,
cotton  production in 1983 exceeded average yields obtained from irrigated
land in Lubbock  County.
     The overall decrease of nitrogen  and phosphorus in tissues from 1981
to 1983 was partially  due to the failure of the irrigation  stream to meet
crop nutrient requirements.  Originally, irrigation was expected to  provide
sufficient  nitrogen and phosphorus to satisfy crop needs; however,  due  to
odor problems the effluent  was transported through the reservoirs  before
application to the  soil.  Nitrogen concentrations in the  irrigation  water
were reduced from  42  mg/1 to approximately 12 mg/1 by passing the effluent
through the reservoirs.   The nitrogen reduction, in conjunction with a  50
percent  reduction in the  total hydraulic loading resulted in a nitrogen
deficiency  in many  of  the fields.  In addition,  no accumulation of  trace
metals  appeared  to have  resulted from land application of the City  of Lub-
bock's  wastewater (Table  16).
Economics—
     In the Lubbock land  treatment situation, the Hancock farm was private-
ly owned utilizing  tenant farmers.  The farm owner could not make land pay-
ment by  $41,265 in 1980, $16,959 in 19"81, and $45,063 in 1982; but  gained
$84,171  in  1983.  Therefore, the landowner has experienced a  four year net
loss of $19,116.   In addition, the tenant farmers as a group netted  $30,584
in 1980,  $65,564  in 1981, and $202,364 in 1983; but lost  $15,584 in  1982.
The farmer's operational  cost did not include living expenses for the farm-
ers.  The  average farmers  net  income  during  1980  through  1983 was
$70,715/yr.  If each of   nine tenants received an equal share of the  profit
then each tenant  would have received an average net income of $7,857/yr  for
their families to live on.
     A  comparison  of the shift  in  tenant farmers income between pre-  and
post- irrigation  periods  is given in Table 17.  The table shows in general
that the years  the farmers had irrigation water (1982 and  1983) they made

                                 66

-------
                                 TABLE 15.  COTTON YIELDS, HANCOCK FARM
ON
-vj

Tenant Ha.
Farmer A 56.7
Farmer B 87.8
Farmer C 120.2
Farmer E 171.2
Farmer F 100.8
Farmer G 82.2
Farmer H 116.1
Farmer I 87.4
Lubbock County***
Lubbock County***
1980*
Ac. kg/ha
140
217
297
423
249
203
287
216


70
137
234
100
85
417
393
109
156*

Ib/ac
62
122
208
89
76
371
350
97


Ha.
56.7
143.3
172.0
213.7
100.8
.82.2
116.1
87.4


1981*
Ac . kg/ha
140
354
425
528
249
203
287
216


222
118
178
131
274
316
196
246
373*

Ib/ac
198
105
158
117
244
281
174
219


Ha.
56.7
142.5
65.6
207.2
50.6
82.2
116.1
29.9


1983*
Ac . kg/ha .
140
352
162
512
125
203
287
74


615
579
676
560
740 .
572
353
i
597
(312)*
395**
Ib/ac
547
515
602
498
659
509
314
531



      * Dryland
     ** Irrigated Land
    *** Texas A & M Extension Service, Lubbock, Texas

-------
 TABLE 16.  ELEMENTAL SHIFTS IN COTTON TISSUES OBTAINED FROM HANCOCK FARM
                               1981 vs. 1983
         Concentration
          1981 > 1982
Concentration
 1981   1983
Concentrations
 1981 < 1983
TKN Stalk
TP Stalk
Ca Stalk
Ca Seed
K Stalk
Na Stalk
Na Seed
Cd Stalk
Cd Seed
Cu Stalk
Cu Stalk
Pb Stalk
TKN Seed
TP Seed
K Seed
Fe Stalk
Fe Seed
Ba Seed
Cr Stalk
Cu Seed*
Cu Seed**



            Pb Seed
 * Overall the same but one or two sharp rises or drops show inconsist-
   ency in trend
** Pivots 3 and 11 show substantial drops but overall trend is same or a
   little increase
            TABLE 17.  COMPARISONS OF SHIFT IN FARMERS'
                       PRE-EFFLUENT TO POST-EFFLUENT
                       INCOME
                   1980 + 1981
                   Net Average
                  $Income/acre
*$/acre x 2.17 = $/hectare
       1982 + 1983
       Net Average
      $Income/acre*
    Pre to Post
    Irrigation
   $Net acre/yr*
Farmer A
Farmer B
Farmer C
Farmer D
Farmer E
Farmer F
Farmer G
Farmer H
Farmer I
-5.22
+2.82
-0.59
-9.83
-0.96
-9.01
+15.72
+42.30
+25.58
+66.58
+56.57
+ 0.32
	
+21.40
+105.20
+15.34
-44.79
+89.05
+71.37
+59.39
+ 1.22
___
+22.36
+114.21
+ 9.63
+ 2.49
+63.48

                                   68

-------
more net  income per hectare than the baseline period (1980 and 1981)  when
there was  no  effluent water applied.   All  of the farmers  had a positive
four year  average net income primarily due  to the income received in 1983.
     The City of Lubbock paid for all costs associated with the treatment
of the  municipal  wastewater  and transporting  the treated sewage to the
Hancock farm. Assuming Lubbock paid 15 percent of the construction costs,
the  City's   cost  per 1000 m3 would have been $98 ($0.37/1000 gal)  and $83
($0.31/1000 gal) for 1982 and 1983, respectively.

LUBBOCK INFECTION SURVEILLANCE STUDY (LISS)
     The rectangular area within 4.8 kilometers (3 miles) to the north, 4.0
km (2.5 .mi)  to the south,  and 3.2 km (2  mi) to the east and west of the
perimeter  of  the spray irrigation rigs on the Hancock farm  was designated
as the  study area. This area,  which includes the small city of Wilson,
Texas and  the rural areas north, northwest,  and  northeast  of Wilson,  was
divided into six  sampling  zones (Figure  10).  The rectangular Zone 1 in-
cluded all rural households located  on  the Hancock farm  and within 0.5
miles  of  its perimeter. Zone 2 contained the households located within 0.5
miles of the  Hancock site boundary within Wilson.  Included in Zone 3  were
all  rural residences located  from 0.5 to  1.0  (E  and W) or 1.5 (N or S)
miles from the Hancock farm.  Zone  4  consisted  of the  Wilson households
which  were located 0.5 to  1.0 miles from the site.  Zone 5 contained the
rural households which were located from  1.0 or 1.5 to 2  miles (E and W) ,
2.5 miles (S) and 3 miles  (N) of the Hancock farm boundary.  Zone 5 was
extended to approximately 3 miles .north of  the farm due  to  the prevailing
southerly winds.   The households of the  small number of Hancock farm work-
ers who resided outside the study area were placed in Zone 6.
     The LISS  monitored four major periods of wastewater irrigation at the
Hancock farm.  These periods were  termed spring 1982 (February 16-April
30,1982), summer  1982 (July 21-September  17, 1982), spring 1983 (February
15-April 30,  1983)  and summer 1983 (June  29-September 20, 1983).  The qual-
ity of  the wastewater used for irrigation varied substantially by irriga-
tion period.   All of the irrigation wastewater  was obtained via pipeline
                                 69

-------
KEY
    Rural household participating
    during irrigation period(s).

    Figure 10.  Sampling Zones  Comprising the  Study Area

                                         70

-------
directly from the  Lubbock SeWRP in the spring 1982 irrigation  period, since
operation of the  reservoirs had not been approved at  that  time.   The qual-
ity of  this pipeline effluent was similar to that of a  low  quality,  primary
effluent.   Pipeline wastewater  comprised 64%, 0?o and 1%,  respectively, of
the total  applied by  spray  irrigation  in the  summer of  1982, spring of
1983,  and summer  of 1983 irrigation periods. There was some  improvement in
the pipeline wastewater quality during summer 1982 and  spring 1983, but it
did not reach the  quality expected of secondary effluent until summer 1982.
Reservoir wastewater was more consistently of secondary  effluent  quality in
all three of these periods.   This  observation  is  important,  since the
majority of irrigation  wastewater  used during 1982  came via pipeline
directly from the  SeWRP, while essentially all the wastewater  applied  dur-
ing 1983 was from  the irrigation reservoirs.
     The wastewater utilized at the Hancock farm contained  a broad spectrum
of enteric  bacteria and viruses.  Spray irrigation  of  wastewater received
via pipeline directly from the Lubbock SeWRP was found to  be a substantial
aerosol   source of each  group of microorganisms monitored in the  aerosol
sampling (i.e., fecal coliforms, fecal streptococci,  mycob'acteria, Clos-
tridium  perfringens,  coliphage, and enteroviruses).  Microorganism levels
in air  downwind of spray  rigs using pipeline wastewater  were found to be
significantly higher than upwind levels:  fecal streptococci  levels to at
least  300m  downwind, and levels of fecal coliforms,  mycobacteria and coli-
phage  levels to at least 200 m downwind.  The  downwind levels  were  also
significantly  higher than the background levels in ambient  air outside the
home of participants.   Fecal coliform levels  were higher  than background
levels  to  beyond 400  m downwind, mycobacteria and  coliphage levels to at
least  300 m  downwind, and fecal streptococci levels to at  least 200 m down-
wind.   Operation at night  and at high  wind speeds appeared  to  elevate
microorganism levels to greater downwind distances.  Enteroviruses  were
recovered in the  aerosol at 44 to 60 m downwind of irrigation  with pipeline
wastewater on each of four virus runs. The geometric  mean  enterovirus  den-
sity in air  was 0.05 pfu/m-*, although a much higher density  (17 pfu/m-*) was
sampled on one  run in August 1982. Spray irrigation of reservoir  wastewater
was also found to be a source of aerosolized fecal coliforms, fecal strep-

                                  71

-------
tococci and coliphage, sometimes to downwind distances of at least 125 in.
     Since microorganisms densities were much higher  in  the wastewater from
the pipeline than  from the reservoirs, the exposure which most of the study
population received  to most microorganisms via the wastewater aerosol  was
greater  in 1982 than in 1983.  The irrigation period in which aerosol ex-
posure at a given  distance downwind was estimated  to  be  highest was:   sum-
mer 1982 for enterov iruses,  summer  1982  for fecal coliforms, and spring
1982 for fecal  streptococci.  For  each of the microorganism groups  with
adequate aerosol  and wastewater monitoring  data,  summer 1982 was the irri-
gation  period when  most of the  more highly exposed  study population
received either their largest or their second largest cumulative dose from
the wastewater  aerosol.
Findings from Self-reported Illness Data
     Disease surveillance did not disclose any obvious connection between
illness and degree of wastewater exposure.    Self-reports of illness  are
always  subject to respondent bias.  Nevertheless, the  participants in the
high exposure level  (AEI>5)  reported the highest  rate  of illness  shortly
after  the onset of wastewater irrigation,  both in  spring 1982 and in summer
1982.   The e*xcess  total acute illness among  high exposure level partici-
pants  over the spring 1982  irrigation  period occurred primarily during
February 14-27, 1982, in  the initial two weeks of wastewater irrigation.
The high exposure level  participants also reported a significant excess of
total  acute illness  in August 1982, primarily during August  15-28  (after
more than three weeks of  wastewater  irrigation had  elapsed).  The high
exposure level  participants did not  report  a comparable excess of  acute
illnesses during either  irrigation period in 1983.   This pattern of excess
illness during both  irrigation periods in  1982 is consistent with  the
hypothesis of an  association of illness with exposure  to wastewater irri-
gation:  the pattern appeared both upon initial wastewater exposure  and  in
the summer 1982 irrigation period which produced highest exposure to micro-
organisms in the wastewater aerosol.  However, the patterns did not persist
throughout either irrigation period in 1982.  In addition, the effects of
known  risk factors such as age and  social economic  status have not  been
                                  72

-------
taken into  account.  A small  excess rate  of  illnesses might  have  been
associated with the initial and heaviest  periods of microorganism  emission
from wastewater  irrigation.   Since the agents  which the LISS  monitored
clinically and serologically show a very  high  proportion  of asymptomatic
infection,  it  is difficult  to correlate the  findings for self-reported
illness  with  those  for the  clinically and  serologically  detected  infec-
tions.
Findings from Seroconversion Incidence Densities
     An  overview of  the association of serologically detected infections
with exposure to wastewater aerosols was  obtained by  comparison of the
seroconversion incidence densities for serum donors in the three levels (or
two groups)  of aerosol exposure, for both the entire baseline (June 1980-
Oanuary  1982) and the entire irrigation (January 1982-October 1983)  periods
of observations.  The high exposure level participants had  a higher inci-
dence density  of coxsackievirus B4 infections versus intermediate level
participants  during the entire irrigation period.   In contrast,  partici-
pants in the high exposure level had no  elevated infection incidence dens-
ity to specific agents in the baseline period.   Based  on  test-based 95%
confidence  intervals for  the  crude incidence  density ratios,  the  high
exposure group (AEIX5) had a significantly greater incidence of infections
to coxsackievirus 82 and  echovirus 11  over the irrigation period, but  a
significantly greater infection incidence only to one agent, echovirus .9,
during the baseline period.
     During  the baseline period, individuals in the high exposure  level had
the lowest infection incidence densities  of  the  three exposure levels to
all of  the  adenoviruses tested, to all  coxsackie B viruses tested,  and to
all echoviruses tested.  In the irrigation period, individuals in  the  high
exposure level  had the highest incidence densities of infection  by all
coxsackie B  viruses tested and by all echoviruses tested.  Moreover,  in the
irrigation  period the high  exposure level'also had the highest  incidence
density  of infections to all of the tested viruses which had been  recovered
from the irrigation wastewater.  These crude incidence densities  suggest  a
probable association between  seroconversions  (especially  to viruses
                                  73

-------
recovered  from the  wastewater) and wastewater  aerosol exposure.  The crude
incidence density  ratios of the high exposure level  to.the  intermediate and
low exposure levels during the irrigation period  were 1.8  and 1.5, respec-
tively, for the viruses recovered  from the wastewater, indicating  some
excess risk of viral infection from wastewater  aerosol exposure.
Evidence  of Association  of  Specific  Infection Episodes with Wastewater
Aerosol Exposure
     Specific  infection episodes which displayed  good or marginal evidence
of association with  wastewater aerosol exposure were identified by compar-
ison  of  results from  four methods of  investigation;  i.e., confirmatory
statistical analysis, exploratory logistic regression (ELR) analysis,  con-
fidence intervals  of incidence density ratios,  and risk ratio scoring.
     Some  excess risk of viral infection was  associated with wastewater
aerosol exposure,  based  on  comparison  of'crude seroconversion incidence
densities  by aerosol exposure level and by irrigation vs. baseline period.
A symmetric risk ratio score approach  provided evidence of a stable and
dose-related association between infection events -and  wastewater aerosol
exposure  in the infection episodes observed by  the LISS.  Furthermore, some
infection  episodes appear to have been re-lated  to wastewater aerosol
exposure,  because more  statistically significant associations than expected
were found in the confirmatory analysis of independent  infection episodes
using  a one-sided Fisher's exact test.  An exploratory logistic regression
analysis  found significant (p<0.05) associations between  presence of infec-
tion  and  degree of aerosol  exposure while controlling  for the effects of
extraneous variables in four infection episodes.
     Additional evidence was considered regarding recovery of the infectious
agent from the irrigation wastewater, seasonal  correspondence of the infec-
tion  response to aerosol  dose, association  with  contaminated drinking
water, alternative risk factors identified by  ELR,  and  within-household
transmission of infections.  The eight, infection episodes with good or mar-
ginal evidence of wastewater aerosol exposure  association were placed in
three  categories based on the likelihood of causal association of the
infection  events with wastewater aerosol exposure:
                                 74

-------
     1)    More plausible  alternative explanation identified:
          o    Episode  of Klebsiella infections in summer 1983
               —alternative:  eating food prepared at local restaurant A
          o    Spurious control  episode of  echovirus 9  seroconversions
               in the baseline period
               —alternative:  within-household spread
     2)    Both aerosol exposure and identified alternative explanation(s)
          are plausible risk  factors (evidence inconclusive):
          o    Episode of clinical viral  isolates excluding adenoviruses
               and immunization-associated polioviruses in summer 1982
               —alternative:  eating food prepared at local restaurant A
          o    Episode  of echovirus 11 seroconversions in 1982
               —alternatives:  o  contaminated drinking water
                               o  Caucasian, large household
          o    Episode of seroconversions to viruses isolated from waste-
               water  in summer 1982
               —alternatives:  o  contaminated drinking water
                               o  low income, Caucasian
          o    Episode of seroconversions to viruses isolated from waste-
               water  in 1982
               —alternative:  farmer, history of pneumonia
          o    Episode of seroconversions in  summer 1982  to all  serum
               neutralization-tested viruses
               —alternative:  contaminated drinking water
     3)    Strong  evidence of aerosol exposure association and no alterna-
          tive explanation identified:
      o     Episode of poliovirus 1 seroconversions in spring 1982
      All five of the infection episodes  in Category 2 relate primarily to
echo  or  coxsackie B viral infections observed in summer 1982 and to agents
recovered from the wastewater at  that time.   Hence, it  is reasonable to
consider these  to be  five manifestations of a single nonpolio enterovirus
episode  centered  on the summer 1982 irrigation season. With the heavy rain-
fall,  widespread drinking water  contamination  and other  unusual circum-
                                  75

-------
stances  which  occurred during this summer,  it  is not surprising that frag-
mentary evidence  of various alternative explanations surfaced for this non-
polio enterovirus episode.
     All the evidence,  however,  supports  the finding that the episode of
poliovirus  1  seroconversions in spring 1982  was associated with wastewater
aerosol  exposure.  The results indicate that a general association between
exposure to  irrigation wastewater  and new   infections  existed  for 1982.
However, even  during 1982,  the strength of association remained weak and
frequently  was  not stable.  Wastewater, directly from  the  pipeline,  com-
prised  much  of the  irrigation  water  in 1982.  The isolation  of entero-
viruses from  pipeline wastewater was greater than that  observed with  the
wastewater that  had been  retained in  reservoirs.  The methods employed
resulted in the observation of a large number of infection  episodes,  none
of which resulted in serious illness.  The voluntary nature of participa-
tion and the  unrepresentative circumstances  of  the study area make general-
ization  of the results  unwise. A larger sample size with greater compar-
ability of  the  exposure groups on the basis  of  drinking  water sources  and
frequency  of visiting the  same eating establishments would have reduced
their confounding effects.
     From  the  public health standpoint, the lack of a strong, stable asso-
ciation of  clinical illness episodes with the level of exposure to  irriga-
tion  wastewater  indicates  that wastewater spray irrigation produced no
obvious disease during the  study  period.   However, when  more sensitive
indicators of  infection were used,  a  general association  was found to
exist, especially for 1982.  A particular concern is  that statistical
interpretation of the data indicates that the  poliovirus 1  seroconversions
were probably related to wastewater aerosol  exposure during the spring of
1982,  even when  the effects of polio immunization were controlled. Because
of the low  prevalence of poliovirus antibody observed during the baseline
period,  the  study population was immunized,  and thus was. probably better
protected against polio than other rural populations.
                                  76

-------
PERCOLATE INVESTIGATION  IN THE ROOT ZONE (PIRZ)
Percolate Quantity
     During the  PIRZ study  several problems  such as  flooding  of service
manholes, improper  vacuum system performance, inflexibility in  irrigation
schedule,  etc.,  limited the amount of percolate quantity  and quality data
obtained from  the root zone investigation.  Therefore the objectives of the
PIRZ  study were not achieved.  The  percentages of the design  irrigation
application rates that were actually applied in 1982 to  bermuda,  cotton and
grain  sorghum were 27 percent (99 cm), 114.6 percent (57 cm) and 98.2 per-
cent (99 cm),  respectively, at the Hancock site; and 26.1 percent  (97 cm),
83.5  percent  (38 cm) and 60.7 percent (48 cm) at the Gray  site.  Only 25 of
the 41 lysimeters on the Gray site contributed  percolate during  the five
months of the  1982  growing season.  At the Hancock site, 26  of the 46 units
had contributed percolate by  the end of September.
     With  increased amounts  of irrigation in August and September and with
decreased evapotranspiration  requirements due to crop maturity  during  the
latter part of the  growth period, percolate flow increased  in lysimeters on
the crop plots.  The increase of moisture in  the  soil profile  during  the
fall of 1982 led  to decreased air leakage in the vacuum  systems  at all man-
holes and reduced daily  vacuum pump operations.   In late November  and dur-
ing December, the  number of  lysimeters contributing percolate began to in-
crease at both sites.  Fall irrigation, fall precipitation, and  decreased
evapotranspiration  improved soil moisture conditions in  the  profile and
subsequently led  to percolate generation.   During the 1983  growing season,
the percentages of  the design application rates actually applied to the
plots for the  bermuda, cotton and grain sorghum at the Hancock  site were
44.8  (164  cm), 17.5 (9  cm) and 18.4 (19 cm) percent, respectively.  At the
Gray site,  the respective rates were 15.7 (55 cm), 46.6 (21 cm)  and 24.4
(19 cm)  percent of  design loadings for bermuda, cotton and grain sorghum.
Failure to  apply  the design hydraulic loading had a major impact  on  perco-
late  collection amounts during the 1983 season.  The amounts of percolate
captured by the lysimeters at the Hancock and Gray sites during  the  study
period  are presented in Tables C.3 and C.4.  In September  1983  the bermuda
                                  77

-------
and cotton  test plots were  flood irrigated  with approximately  40 cm of
water. The tray  lysimeters on the bermuda  plot at the Hancock  site were
kept  operative  until November 16,  1983.   Percolate was collected  from the
three trays  at the 61 cm  depth and two trays  at the 183  cm  depth.  The
depths of percolate collected by these  trays in' November were:
                Tray                         Depth in mm
                108                                <1
                102                                 4.6
                103                            .     5.3
                108                                 4.2
                109                                19.4
Tray 109 had  contributed  percolate in only  one previous  event over the
project  period.  If the water had  been in the profile, there  was  a  possi-
bility  that  more lysimeters would have  contributed during  the  study
period.
Percolate Quality
     Many of the water quality samples were of such small volume that only
a few parameters could be measured.   Comparison of the geometric means of
parameter concentrations generally revealed a decrease in nutrient  levels
in the percolate samples.   Levels of total  Kjeldahl nitrogen,  ammonia
nitrogen, total phosphorus,  orthophosphate phosphorus,  and organic phos-
phorus decreased by a factor of 10 or greater.  The levels  of nitrite plus
nitrate-nitrogen (N02+N03)   increased in the  percolate.   Most of this
nitrogen was  assumed to be in the form  of nitrates.  The  increase  in con-
centration  of  nitrate resulted from  the oxidation of the other  nitrogen
compounds present in the applied wastewater as  well as from  the  mobiliza-
tion by the  percolate of nitrates stored in the profile.
     Examination of sample data sequentially taken over the project period
showed  a general decrease in the concentrations of the chemical constitu-
ents recorded as  the total volume of percolate collected  by the unit in-
creased. This was easily noticeable  in  sample values obtained over  time for
TDS, N03 (Figure 11), NH3, TP, and PO^.  Decreases in conductivity, Cl, and

                                 78

-------
I
340


320



300


280



260


240
                      160
                       120


                       100
                                                                                             300

                                                                                             —o
                                                                                             200 .
                                                                                              100 •
      _JU
    JL
AUG
1ST
.SEPTEMBER
OCTOBER
NOVEMBER
DECEMBER
   Figure 11.   Variation of Nitrate Concentration in Percolate and Accumulated Weight of
                 Leached Nitrate  in kg/ha with Time for  Tube 123 from July through December
                 1982

-------
504 also occurred  during the monitoring period.
Soils
Nitrogen and Phosphorus—
     As observed in  the Demonstration/Hydrogeologic Study,  organic  nitrogen
was the primary nitrogen form in the soil profile.  The increase  in soil
moisture  at the Hancock test site resulting from irrigation  created  favor-
able conditions for  the mineralization of organic nitrogen  and  consequently
a decrease  in organic nitrogen.  At the Gray site, the average  TKN  of  the
three plots at the  90 cm  depth  for  each sampling period  showed little
change.
     Nitrate-nitrogen was the major inorganic nitrogen form  present  in  the
soils at both  test sites.  The nitrate content in the soil  profiles at  the
Hancock site were  generally higher at both the start and end  of the project
than the levels measured at the Gray test site.  The higher initial nitrate
levels  at the Hancock test site probably resulted from the drier soil con-
ditions which  historically have been experienced at this site.   The  drier
soils  will  retain  water-soluble nitrogen forms until excess soil  water is
available to leach them from the profile.  The average profile  nitrate con-
tent  for  the  three sampling periods  in the top 90 cm showed a general
decrease over  time.   High  concentrations of  nitrates were found in  the
initial percolate  volumes intercepted by many lysimeter units.
     Average orthophosphate phosphorus values in the profile  decreased over
the project  period in  the upper 90 cm depth at both sites.   The addition of
this nutrient  in the irrigation water would have offset some  of the losses
caused  by vegetative growth.   The total phosphorus consumed  by the crops
was less than  the  amount  applied  in  the irrigation water; therefore,  a
portion of the orthophosphate was transformed to more insoluble compounds.
Priority Organics  Pollutants—
     The  composition of the project soils at the beginning  and  end  of  the
project period showed  reductions in most of the priority organics measured
at both sites.   Increases  had occurred in the concentration  of nine com-
pounds:  carbon tetrachloride, dibutylphthalate, hexadecane, methylhepta-
decanoate,  methylhexadeconate, octadecane, phenol, propazine,  and  tetra-
                                  80

-------
chlorethylene.   The greatest increase  in  the soil profile occurred in  the
levels of carbon  tetrachloride, hexadecane,  and dibutylphthalate.    The  two
former  compounds are solvents.  The wastewaters pumped to the Hancock site
during the project period contained an average  concentration of 4.7 pg/1
for carbon tetrachloride  (0.145 kg/ha);  <2.0  yg/1 for hexadecane (0.05
kg/ha);  and 104 ug/1 for dibutylphthalate  (2.6 kg/ha).   In  water  going to
the Gray  site the values  were  3.2 ug/1  for carbon tetrachloride (0.066
kg/ha) and 140 pg/l for dibutylphthalate (1.98 kg/ha).
Bromide—
     Sodium bromide was applied to the surface of the sub-plots annually in
order  to trace the movement of percolate  through the  soil  profile.   In
1982, the bromide moved about  1.5 cm down through the profile for each
centimeter of  water applied at this site.   Variability in movement was much
greater" at the  Gray site.  The bermuda  sub-plot,  with a  much  larger
hydraulic loading, showed a bromide ion  accumulation in  the  60 and 90 cm
level which corresponded to the  depth of the indurated caliche layer.   The
movement of bromide on the other  two plots  was  approximately 1 .95  cm  per
centimeter of  applied water.
     In the fall of 1983, the cotton and grain sorghum sub-plots  still  had
bromide  present in the surface layer.  Some  translocations were evident in
the bermuda plots because of the heavier hydraulic loading on these plots.
At the Gray test  site, the effect of the caliche layer on  the bromide  was
evidenced by  build-up of  bromide in the  45 to 60 cm layer on the bermuda
plots.

AGRICULTURAL RESEARCH STUDIES
Hydraulic Loading Study
Crops—
     Crop  yield  data  indicated  grain  sorghum yields increased as treated
wastewater hydraulic loading increased  to approximately 3 m.ha/ha.yr.
Average  cotton  lint yields  obtained  from test  plots irrigated with  122
cm.ha/ha.yr to 297 cm.ha/ha.yr ranged from  1,300 to 1,538  kg/ha.   Cotton
test  plots having less than 122  cm.ha/ha.yr of treated sewage applied pro-

                                81

-------
duced an average lint yield  of  100 to 925 kg/ha.  In general alfalfa irri-
gated with 365 and 434 cm.ha/ha.yr of treated sewage generated the highest
yields  during each  cropping period.  Furthermore, the alfalfa production
was greatest during the month of June.  Common bermuda grass production  in
test  plots receiving  treated  sewage was greater than dryland plots.   During
June  and September  1983, the  highest bermuda yield (9368. -+2327.  kg/ha)
was harvested from the lowest annual effluent hydraulic loading of 152 cm.
      Chemical analysis of the bermuda plant tissue indicated that the  crop
had  several macro and micro nutrient deficiencies.  With the shallow root
system of bermuda increased irrigation may have leached nitrogen past the
root  zone; thereby,  limiting  nitrogen availability to the crop.   Other
nutrients such  as phosphorus,  zinc,  potassium and iron  appeared  to  be
deficient in the bermuda tissue.
   •   Grain  sorghum and  cotton  tissue experienced less nitrogen content  in
1983  compared to 1982.  Protein  content in alfalfa ranged from 24  to  28
percent.  Alfalfa irrigated  with 137 cm.ha/ha.yr or greater contained more
than  26 percent protein.  Phosphorus and potassium, however, may have been
deficient in the alfalfa tissue.
Soils —
      Soil  texture within  the upper 30 cm (1 ft) of the soil profile were
generally sandy clay  loam.  Clay to clay loams dominated  the soils  from  a
depth of 30 cm to 122 cm (4  ft)  within the profile.  The majority of soils
from  122 cm to 183 cm (6 ft)  were clays.  Cation exchange capacities  (CEC)
within the test plots were  greater than 20 meq/100 g (average 23.6 meq/100 g
+3.3)  which were characteristic of the clay/clay loam soils.
      Nitrogen--As observed  throughout the Hancock farm, organic nitrogen
was the primary nitrogen from within the soil profile.  In general,  nitrate
nitrogen was the major form of inorganic nitrogen.  Leaching of inorganic
nitrogen  past  a  soil depth of 91 cm appeared to have occurred as annual
hydraulic loadings of 61  cm or greater within the cotton test plots (Figure
12).   Nitrogen  mass balances  indicated that nitrates were leached beyond
183 cm of soil in cotton, grain  sorghum and bermuda test plots irrigated
with  122 cm.ha/ha.yr  or greater.

                                  82

-------
en




*<0
^c
5
o
V.
*»
IS
eO
'c
2k
2*
o
c

300



2OO ~
100
—
100


200
0 cm 20 cm 41 cm 51 cm 61 cm

i i i I i I || i




»
IB :



1
IS, -
* IB! ^
*


1
ISl1
( i^i
. •' •• ••


i.

s *
Si ;
is! 1
•• i1
i" •• • i- =i
i §l
! i
•
••
••••• N Root Zone Pre- irrigation 1983
^KfSt N From Organic N in Root Zone
— — — N Applied in Effluent
69 cm

I I




#
a *
I
i
iB: 1
if
-i
ii
i



86 cm 102 cm

( I I I


*
* *
s- * «' *
I 1
I i !
n n
i1. !i
"l 1
1 1



122

I



i
i
1
ill






cm

1
*
*.
*
*
*
*
*
J
•r
•1
1
1
1
1


Hydraulic
Loading











         300
                        **
ii N  Removed by Crop
•i N  Removed by Denitrification
• N  Measured in Profile Post-irrigation  1983
# N  Difference between Measured and Predicted
       Figure  12.   Nitrogen Mass Balance for Trial  14000 Cotton Plots

-------
     All nitrogen mass  applied to the alfalfa plots, however,  was  consumed
and nitrogen fixation  was  a  source of inorganic nitrogen for the  crop  (Fig-
ure 13).   Nitrite plus  nitrate-nitrogen  (N02+NQ-3)   lenses were  measured
beneath  the entire research area.  Nn2+NQ-3 levels decreased with  depth as
hydraulic loadings increased.   Inorganic nitrogen apparently was  not leach-
ed beyond depth of 183 cm.
     Minerals--Accumulations of total dissolved solids (IDS)  were limited
through  leaching  of  IDS  through the soil profile in test plots irrigated
with 122 cm.ha/ha.yr or  more.   In the grain sorghum plots the exchangeable
sodium  percentage (ESP)  was less than seven in the top 30 cm of soil.  As
the annual  hydraulic and salt mass loading increased in the cotton plots
(122  cm.ha/ha.yr to  297  cm.ha/ha.yr)  and  alfalfa test  plots  (137
cm.ha/ha.yr to 434 cm.ha/ha.yr) the ESP values within the top 30  cm of soil
increased.  In the  fall of  1983, the ESP values within the top 30  cm of soil
collected from cotton  test plots irrigated with 183 cm and  297 cm/yr  were
9.2 and  8.1, respectively.  ESP levels were less than 10  within the  upper
61 cm of soil. Sodic conditions may have existed within the upper 30 cm of
the plots  irrigated  with 434 cm/yr.   High water  consumption  of alfalfa
caused accumulations of  sodium within  1.8 m of the soil profile  at sched-
uled hydraulic loadings  of 137, .198, and 259 cm/yr.
Hydraulic Application  Frequency Study
Crops—
     Yield  data (Figure 14) indicated soybean production  was  highest with
more frequent wastewater application  (i.e., one  irrigation/week and one
irrigation/2 weeks); whereas, grain sorghum yields were significantly  high-
er with longer periods between irrigation (1  irrigation/4 weeks and 1  irri-
gation/8 weeks)(Figure 15).   Soil moisture within the profile corresponded
to the type of root system.  Less soil moisture  was measured  beneath the
grain  sorghum at  soil depths of 122, 152, and 183 cm.  These soil  moisture
differences possibly were due to:   (1)  the more  extensive  grain sorghum
root  system and  (2)  the  fact that growing season for soybeans ends with a
complete shutdown  of the plant while the milo  plants stay  green and  con-
tinue to extract moisture  from the soil until frost kills the crop.

                                  84

-------
CD



800





600

I 400

o>
^ 200
O> -|-
V,
1 20°
| 400
600


800


Effluent Water
I
137 198 259 3O5 365 43'
1
t
t
t
t
1
1 1
1 1
I 1
1 1
1 1
! i i
i !
• '•
i ' •
iiii!
: : : ! !
*• s : i i Si
Si Si i V 5 8
IS1 B ' ^ ' fi ' 5' 8 '
iSi . l^i. . iS: • !0: I iM . .Si
i! ii! iii :* JU
•* ii i ! !«
i • i 1 1

: i





Well Water
| 1 1
4 0 365 3O5 259 cm/











• i5 • .B! • i5l • .Si 1
"•"!?"" |l • | w I T — —••*•- • i If
i i i i I i
i t jt •*
m I
i ;
i
*•••• N Root ^one Pre- irr iijyt inn 19BJ
^fftfQI N From Organic N id Knot Zone
•••————— N ADD 1 i eel in Ef t'l ijtnt.
^ • •• • M Removed by Crop
• •• •• N Keinuveci hy Dertitr i f'icat ion
IHH^ N t-Vj;isure<1 in Prof i l« Post- i rr itjat ion 1985
* # '* N Dtflurenct: between Mcauured and Predicted
           Figure 13.  Nitrogen Mass Balance for Trial 16000 Alfalfa plots

-------
      a
      UJ<
       'S
      "
      O
                                                                   Application  Frequency
                                                                    D1  Application/wk
                                                                    O1  Application/2 wks
                                                                    A 1  Application/4 wks
                                                                    -f- 1  Application/8 wks
         0.00
                  20.00
                           UO.OO
                                     60.00     80.00     100.00
                                   HTDRflULJC LORDING -(CM)
                                                                 120.00-
                                      1110.00
                                                160.00
 Figure  14.   Soybean Seed  Yield  vs  Hydraulic  Loading  - Trial  170GO
     "og
       "8-I
      — o
      a -
                                      Application Frequency
                                       D - 1 Application/wk
                                       O - 1 Application/2 wks
                                       A - 1 Application/4 wks
                                       + - T Application/B wks
         0.00
                  20.00
                            UO.OO
   I         1         I
 60.00      60.00     100.00
HYORflULJC LOflDING (CM)
                                                                 120.00
                                                                          mo. oo    160.00
Figure  15.   Milo  Whole Plant  Yield vs Hydraulic  Loading -  Trial 17000
                                               86

-------
     At an  annual hydraulic  loading of 30 cm/yr, symbiotic nitrogen fix-
ation apparently provided  inorganic nitrogen  to the soybeans.  Once  the
hydraulic loading  was increased to 61 cm/yr, sufficient nitrogen mass was
applied to soybeans to inhibit  nitrogen fixation.  In  the soybean  test
plots  inorganic  nitrogen leached  through the 183 cm soil profile when 122
cm of effluent/yr was applied  at irrigation intervals greater than once per
week.   Similarly,  nitrogen was leached from the 91 cm profile within the
grain sorghum test plot irrigated with 122 cm/yr.
     Sodium increased in the grain  sorghum at low irrigation rates.  Leach-
ing of Na within the plots resulted when greater quantities of water  were
applied per irrigation period.   Consequently, less Na was available for the
crop.  Grain sorghum, with a deeper more extensive root  system, was  not
affected  by minor changes in  loadings or frequencies of application,  while
a small shift of either of these factors on the  soybean  crop changed  the
availability of sodium.
Soil —
     The ability of the crop to  adapt to water stress conditions was a fac-
tor which influenced Na accumulation within the soil profile.  Due to  the
shorter  growing  season in 1982,  soybeans with a more shallow root system
and possibly greater water requirements than sorghum, utilized water within
the upper 61 cm of soil.   Consequently, the greatest Na levels as a percent
of base saturation was observed  in  the upper 61 cm of the  soil.  In  addi-
tion, higher water application frequency appeared to increase ESP values in
the upper soil profile.
     Water utilized by grain sorghum caused an accumulation of Na at great-
er soil depths than soybeans.  Upward migration of water  due to capillary
action  during water stress periods (increased time intervals between irri-
gation) may have caused an increase in Na in the upper profile.
                                  87

-------
                               REFERENCES

 1.  Camann,  D.  et al.  Health Effects Study for the  Lubbock Land Treat-
     ment:  Lubbock  Infection Surveillance Study.  Vol. IV of the Lubbock
     Land Treatment System Research and Demonstration  Project. EPA.   U.S.
     EPA Health  Effects Research Laboratory, Cincinnati, Ohio, 1985.

 2.  Chang, A. C. and A. L. Page.'  Toxic Chemicals Associated with  Land
     Treatment  of Wastewater.  International Symposium, State of Knowledge
     in Cold  Regions  Research and  Engineering Laboratory,  Hanover,  New
     Hampshire,  1978.  pp. 47-57.

 3.  EPA.  Process Design Manual  for Land Treatment of Municipal Waste-
     water.   EPA 625/1-81-013, U.S. EPA Center  for Environmental Research
     Information, Cincinnati, Ohio, 1981.

 4.  George,  D.  B., N. A. Klein, and D. B. Leftwich.   Agricultural Research
     Studies:   Vol.  Ill of the Lubbock Land Treatment System Research and
     Demonstration Project.  EPA.   U.S. EPA.  Robert S. Kerr Environmental
     Research Laboratory.  Office of Research  and Development, Ada,  Okla-
     homa,  1985.

 5.  George,  D. B. et  al.  Demonstration/Hydrogeologic Study:  Vol.  I of
     the Lubbock Land Treatment System Research  and Demonstration Project.
     EPA.  U.S. EPA,  Robert S.  Kerr Environmental  Research Laboratory,
     Office of Research and Development, Ada, Oklahoma, 1985.

 6.  Loehr, R.  C., W. 0. Jewell,  G. D. Novak, W. W. Clarkson, G. S. Fried-
     man.  Land  Application of Wastes.  Vols. 1  and 2. ' Van Nostrand Rein-
     hold,  New York, 1979.

 7.  Majeti,  V. A. and  C. S.  Clark.  Health  Risks  of Organics  in  Land
     Application.  In:  Journal of the Environmental Engineering Division,
     Proceedings of the American Society of Civil Engineer, ASCE, Vol.  107,
     No. EE2, 1981.

 8.  Overcash,  M. R.  Land Treatment of Municipal Effluent and  Sludge:
     Specific Organic Compounds.  In:  Utilization of  Municipal Wastewater
     and Sludge on Land.  University of California, Riverside, California,
     1983.  pp.  199-231.

 9.  Pettygrove, G. S. and T." Asano (ed).  Irrigation  with Reclaimed Munic-
     ipal Wastewater - A Guidance  Manual.  Report No.  84-1  wr, California
     State  Water Resources Control Board, Sacramento,  California, 1984.

10.  Ramsey,  R.  H. and R. M. Sweazy.  Percolate  Investigation in the  Root
     Zone:  Vol. II of the Lubbock Land Treatment System Research and  Dem-
     onstration  Project.   EPA.  U.S.  EPA, Robert S. Kerr Environmental
     Research Laboratory, Ada, Oklahoma, 1985.

-------
11.  Stromberg,  L.  K. and  S.  L.  Tisdale.  "Treating Irrigated Arid- Land
     Soils with Acid-Forming  Sulphur Compounds".  Tech.  Report No.  24,  The
     Sulphur Institute, Washington, D.C., 1979.

12.  Wells, D.  M,  R.  M. Sweazy,  and G. A. Whetsone.  Long Term Experiences
     with  Effluent  Reuse.   Journal of the Water Pollution Control  Federa-
     tion, Vol. 51,  No. 51,  1979.  pp. 2641-2648.

13.  Williams,  R. B.  1982.   Wastewater Reuse - An Assessment of the Poten-
     tial and Technology.  Water Reuse.  Edited by E. 3. Middlebrooks.  Ann
     Arbor Science Publ.,  Inc., Ann Arbor, Michigan,  pp. 87-136.
                                   89

-------
           APPENDIX A

Demonstration/Hydrogeologic Study
 Agricultural Cropping Patterns
                 i.
-------
                        Reservoir
                    -1-2-1 Playa
                        Soybeans
                    + + + Sun Flowers
Figure A.1.  Summer 1982 cropping pattern for Hancock Farm
                               90

-------
                I   I COTTON




                i&SS MIIO




                    WHEAT » MILO
                f f -r



                •fffff. WHEAT « BEANS




                I-"-""-' WHEAT » ALFALFA
                             ROW IRRIGATION
                             DISTRIBUTION CAN




                             DISTRIBUTION LINE
             HANCOCK  LAND


             DISPOSAL  SITE
Figure A.2.  Summer  1983 cropping pattern, Hancock  Farm



                                    91

-------
vo
N>
                                                                                          Grazing
                                                                                      ::-:>-:: Water

                                                                                      1 cm = O.27 km
        Figure  A.3.  Winter  1982 crop  and grazing areas at Gray Farm

-------
                                           WINTER CROP
                                           1982-1983
                                          ALFALFA
                                         1982-1983
Figure  A.4.  Cropping patterns  on Gray Farm  during 1982  and 1983
                                    93

-------
          APPENDIX B

Supplemental Figures and Tables
for Section 3 (Project Design)

-------
                        TABLE B.1.  WATER ANALYSIS
Alkalinity, mg/1  CaOIj
Total Organic Carbon (TOC), mg/1
Specific Conductance, umhos/cm
Total Dissolved Solids (TDS), mg/1
pH
Chloride (Cl), mg/1
Total Kjeldahl Nitrogen (TKN), mg N/l
Nitrite plus Nitrate (N0.2+N03), mg N/l
Ammonia (Nh^), mg N/l
Total Phosphorus (TP), mg P/l
Orthophosphate Phosphorus (P04), mg P/l
Organic Phosphorus (Org P), mg P/l
Chemical Oxygen Demand (COD), mg/1
Sulphate (504), mg S04/1
Total Coliform (TO/100 ml
Fecal Coliform (FC)/100 ml
Fecal Streptococci (FS)/100 ml
Salmonella/300 ml
Aluminum (Al), mg/1*
Arsenic (As), mg/1*
Barium (Ba), mg/1*
Boron (B),  mg/1*
Calcium (Ca), mg/1*
Cadmium (Cd), mg/1*
Cobalt (Co), mg/1*
Chromium (Cr), mg/1*
Copper (Cu), mg/1*
Iron (Fe),  mg/1*
Lead, (Pb), mg/1*
Magnesium (Mg), mg/1*
Nickel (Ni) , mg/1*
Potassium (K), mg/1*
Selenium (Se), mg/1*
Silver (Ag), mg/1*
Sodium (Ag), mg/1*
Thallium (Tl), mg/1*
Zinc (Zn), mg/1*
Anthracene/phenathrene (yg/1)
Atrazine (yg/1)
Benzene (yg/1)
Benzeneacetic acid, (yg/1)
4-t-butylphenol, (yg/1)
Carbontetrachloride, (yg/1)
4-chloroaniline, (yg/1)
Chlorobenzene, (yg/1)
Chloroform, (yg/1)
2-chlorophenol, (yg/1)
1-chlorotetradecane, (yg/1)
Dibutylphthalate, (yg/1)
2,3-dichloroaniline, (yg/1)
3,4-dichloroaniline, (yg/1)
Dichlorobenzene M,P,0, (yg/1)
Dichloromethane, (yg/1)
2,4-dichlorophenol, (yg/1)
Diethylphthalate, (yg/1)
Diisooctylphthalate, (yg/1)
Dioctylphthalate, (yg/1)
Dodecanoic acid, (yg/1)
Ethyl Benzene, (yg/1)
Heptadecane, (yg/1)
                                                               (continued)

-------
TABLE B.1, continued
Manganese (Mn), mg/1*
Mercury (Hg), mg/1*
Molybdenum (Mo), mg/1*
Methylhexadecanoate, (yg/1)
1-methylnaphthalene, (yg/1)
2-methylphenol, (y'g/1)
4-methylnaphthalene, (yg/1)
Naphthalene, (yg/1)
Octadecane, (yg/1)
    Hexadecane, (yg/1)
    Hexadecanoic acid, (yg/1)
    Methyheptadecanoate, (yg/1)
    Phenol, (yg/1)
Propazine, (yg/1)
ct-terpineol, (yg/1)
Tetrachloroethylene, (yg/1)
Toluene, (yg/1)
Trichloroethylene, (yg/l)
*Total and Dissolved
                                  95

-------
                                    TABLE B.2.  SOIL ANALYSIS
Alk, mg/g as
TOC, mg/g
Specific Conductance, pmhos/cm
IDS, mg/g
pH
Cl, mg Cl/g Total
TKN, mg N/g Total
N02+N03, mg N/g
NH3, mg N/g
P, mg P/g
PO^, mg P/g
50^, mg S/g
CaC03, mg/g as CaC03
Cationic Exchange
Anionic Exchange
Organic Matter
Buffer Capacity
Solution Cations, mg/g
Sulfur, mg/g
Specific Gravity
Texture
Bulk Density
Consistency
Color
Humus, mg/g
Total Coliform/g
Fecal Coliform/g
Fecal Strep/g
Actinomycetes/g
Fungi/g
Al, mg/g*
As, mg/g*
Ba, mg/g*
B, mg/g*
Ca, mg/g*
Cd, mg/g*
Co, mg/g*
Cr, mg/g*
Cu, mg/g*
Fe, mg/g*
Pb, mg/g*
Mg, mg/g*
Mn, mg/g*
Hg, mg/g*
Mo, mg/g*
Ni, mg/g*
K, mg/g*
Se, mg/g*
Ag, me/g*
Na, mg/g*
Tl, mg/g*
Zn, mg/g*
Acenaphthylene, (pg/kg)
2-chlorophenol, (pg/kg)
Atrazine, (pg/kg)
Benzene, (pg/kg)
Benzeneatic acid,  (pg/kg)
4-t-butylphenol, (pg/kg)
Carbontetrachloride, (pg/kg)
4-chloroaniline,  (yg/kg)
Chlorobenzene
Chloroform, (pg/kg)
Anthracene/phenanthrene, (pg/kg)
1-chlorotetradecane, (pg/kg)
Dibutylphthalate, (pg/kg)
2,3-dichlorotetradecane, (pg/kg)
3,4-dichlorotetradecane, (pg/kg)
Dichlorobenzene, ( g/kg) M,P,0 (3)
Dichloromethane, (pg/kg)
2,4-dichlorophenol, (pg/kg)
Diethylphthalate, (pg/kg)
Diisooctylphthalate, (pg/kg)
Dioctylphthalate, (pg/kg)
Oodecanoic acid, (pg/kg)
Ethylbenzene, (pg/kg)
Heptadecane
Hexadecane, (pg/kg)
Hexadecanoic acid, (pg/kg)
Methylheptadecanoate, (pg/kg)
Methylhexadecanoate, (pg/kg)
1-methylnaphthalene (pg/kg)
2-methylphenol, (pg/kg)
4-methylphenol, (pg/kg)
Napthalene, (pg/kg)
4-nonylphenol, (pg/kg)
Octadecane, (pg/kg)
Phenol, (pg/kg)
Propazine, (pg/kg)
a-terpineol, (pg/kg)
Tetrachloroethylene, (pg/kg)
Toluene, (pg/kg)
Trichloroethane, (pg/kg)
Trichloroethylene (pg/kg)
*Total and Available
                                             96

-------
          TABLE B.3.  CROP ANALYSIS PROTOCOL
                        COTTON

Lint, Seed, Burs, Stems:
      TC, PC, FS
      TKN, TP
      K, Ca, Mg, Na, Zn, Mn, Fe, B, Al, Cd, As

Seed:
      Protein, Cl, Oil

                 GRAIN SORGHUM (MILD)

Grain, Stalk, Leaf:
      TC, FC, FS
      TKN, TP, Cl
      K, Ca, Mg, Na, Zn, Mn, Fe, B, Al, Cd, As

Stalks, leaf:
      HCN, Fiber

Grain:
      Protein, Starch, Oil

                   ALFALFA, BERMUDA

Whole Plant:
      TC, FC, FS
      TKN, TP, Protein, Cl
      K, Ca, Mg, Na, Zn, Mn, Fe, B, Al, Cd,
      Fiber

                 SOYBEANS, SUNFLOWERS

Leaf, Stem, Seed:
      TC, FC, FS
      TKN, TP, Cl
      K, Ca, Mg, Na, Zn, Mn, Fe, B, Al, Cd, As

Seed:
      Protein, Oil

                      WHEAT, DATS

Leaf, Stem, Seed:
      TC, FC, FS
      TKN, TPS, Protein, Cl
      K, Ca, Mg, Na, Zn, Mn, Fe, B, Al, Cd, As

Seed:
      Starch
As
                           97

-------
    TABLE. B.4.  WATER qUALITY ANALYSIS SCHEDULE FOR PERCOLATE SAMPLES
                                                       Group
               Parameter                           Classification
               Alkalinity                                A
               COD                                       A
               IDS                                       A
               Specific Conductance                      A
               pH                                        A
               Total Kjeldahl Nitrogen                   A
               NH3                 •                      A
               N02 + N03                                 A
               Total Phosphorus                          A
               Organic Phosphorus                        A
               Orthophosphate Phosphorus                 A
               TOC                                       A
               Ca                                        B
               Cl                                        A
               K                                         A
               Mg                                        B
               Na                                        A
                                                     .    B
               Heavy Metals                              C
                 Ag, As, Ba, Cd, Cr, Cu, Fe, Hg, Ni
                 Pb, Se, Zn, Co, Al, Mn, Tl, Mo, B
               Trace Organics                            D
               Fecal Coliform                            C
               Viruses                                   C
Key:  A - weekly;  B - Monthly;  C - Quarterly;  D - Yearly
                                  98

-------
MD
                                                                                    •—•  Boundary
                                                                                        Road
                                                                                    °   Well
                                                                                    :    Water
          Figure  B.1.   Gray Farm ground water monitoring locations

-------
                                                                                  DRINKING WATER
O
o
                     GROUND WATER
                                            10212
                                                                      A             / Road
                                                                         Location     '
       Figure B.2.  Hancock  Farm Ground Water  and Drinking Water Monitoring Locations

-------
               APPENDIX C

        Supplemental Figures and
Tables for Section 4 (Summary of Findings)

-------
TABLE C.1.   GROUND WATER QUALITY  -  GRAY  WELLS

Parameter
Alkalinity (mg CaC03/l
Conductivity (umhos/cm)
TDS (mg/1)
pH
Cl (mg/1)
S04 (mg/1)
TKN (mg N/l)
N02 + N03 (mg N/l)
NH3 (mg/1)
TP (mg P/l)
P04 (mg P/l)
Org P (mg P/l)
COD (mg/1)
TOC (mg/1)
TC (Counts/100 ml)
FC (Counts/100 ml)
FS (Counts/100 ml)
Metals, Dissolved (mg/1)
Al
As
Ba
B
Ca
Cd
Co
Cr
Cu
Fe
Pb
Monitoring
June 1980-Jan. 1982
224-402
1 ,244-2,882
1,010-2,271
7.10-7.72
208-680
149-795
0.28-6.97
5.05-35.89
0.02-2.05
0.10-3.49
<0. 01-0. 84 . .
0.08-2.31
27.2-125.4
12.2-38.6
0-300,000,000
0->3603
0-4,501

0.080-3.530
<0. 058-0. 243
0.058-0.243
0.521-3.671
47.7-161.7
<0. 001 -0.004
<0. 005-0. 012
<0. 005-0. 025
<0. 005-0. 106
0.021-0.556
<0. 002-0. 060
Period
Feb. 1982-Dec.1983
140-439
1,098-3,846
723-2,812
7.28-8.13 •
154-686
148-1,067
0.20-7.65
0.77-33.43
<0.01-6.96
<0. 01-1 .94
<0. 01-1. 68
<0. 01-0. 23
11.4-100.3
1.1-9.9
0->31,700
0-30,000
0-26,660

0.153-1.477
<0. 005-0. 038
0.019-0.250
<0. 100-1. 174
44.1-175.1
<0. 001 -0.004
<0. 005-0. 030
<0. 005-0. 023
<0. 005-0. 020
0.021-1.105
0.003-0.041
(continued)
                  101

-------
Table C.1,  continued
Parameter
'Mg
Mn
Hg
Mo
Ni
K
Se
Ag
Na
Tl
Zn
Metals, Total (mg/1)
Al
As .
Ba
B
Ca
Cd
Co
Cr
Cu
Fe
Pb
Mg
Mn
Hg
Mo
Ni
K
Monitoring
June 1980- Jan. 1982
22-138
0.002-0.260
<0. 001 -0.003
<0. 003-0. 055
<1. 005-0. 077
8-40
<0. 005-0. 009
<0.005
73-476
<0. 005-0. 007
0.027-0.591

0.146-36.340
<0. 005-0. 018
0.028-4.750
<0. 100-5. 10
30-394
<0. 001-0. 011
<0. 005-0. 096
0.003-0.096

-------
Table C.1,  continued
Parameter
Se
Ag
Na
Tl
Zn
Organics (ppb)
Acenaphthylene
Anthracene
Atrazine
Benzene
4-t-butylphenol
Carbon tetrachloride
4-chloroaniline
Chlorobenzene
Chloroform
2-chlorophenol
1 -chlorotetradecane
Dibutylphathalate
2 ,3-Dichloroaniline
3 ,4-Dichloroaniline
Dichlorobenzene meta
Dichlorobenzene para
Dichlorobenzene ortho
Dichloromethane
2,4-Dichlorophenol
Diethylphthalate
Diisooctylphthalate
Dioctylphthalate
Monitoring
June 1980- Jan. 1982
<0. 005-0. 017
<0. 005-0. 031
93-434
<0. 005-0. 012
0.061-0.923

<5.0
<2.0-2.9
<2.0-5.7
<1.0-11.9
<2.0-3.4
<5.0
<10.0
<1.0-1 .2
<1. 0-8.4
<2.0-3.3
<2.0-5.7
<2. 0-14.1
<5. 0-8.1
<2.0-2.6
<2.0-3.2
<2. 0-2.1
<2.0-4.2
NR
<2.0-4.6
<2.0-11.7
<2. 0-757. 7
<2.0-34.4
Period
Feb. 1982-Dec.1983
<0. 002-0. 013
<0.001
118-573
<0.005
<0. 020-1 .568

<2.0-4.3
<2.0-5.2
4.7-10.0
<1.0-6.2
<1. 0-3.0
2.4-6.5
<10.0
<1.0
<1.0-2.5
<1. 0-3.0
<2.0-4.9
<1. 0-8.8
<2.0-8.9
<2.0-2.6
O.O-1.7
<1.0-1.7
0.0-1.9
NR
<2. 0-3.0
<2 .0-862.1.
<2. 0-475.0
<2.0
                                                               (continued)
                                 103

-------
Table C.1,  continued
Parameter
Ethyl Benzene
Heptadecane
Hexadecanoic. Acid
Methylheptadecanoate
Methylhexadecanoate
1 -Methylnapthalene
2-Methylphenol
4-Methylphenol
Naphthalene
4-nonylphenol
Octadecane
Phenol
Propazine
o-terpineol
Tetrachloroethylene
Toluene
Trichloroethane
Trichloroethylene
Monitoring Period
June 1980-Jan. 1982 Feb.
<1 .0-2.1
<2. 0-3.0
9.0-111.0
<2. 0-14.1
<2.0-6.9
<2.0-2.9
<2.0-2.7
<5.0-5.2
<2.0-7.5
2.09
(1 data point) (1
<2.0-4.4
<10.0
<2.0-2.3
<2.0-2.3
<1.0-5.4
<1 .0-4.6
<5.0
1.4-4.6

1982-Dec.1983
<1. 0-2.1
<2.0
NR
<2.0-2.3
<2.0
<1.0-1.9
<1.0-1 .9
<2.0=-31.0
<1 .0-4.0
<2.0
data point)
<2. 0-4.0
<1 .0-10.5
<1. 0-2.1
<1. 0-2.1
<1.0-5.3
<1.0-2.2
<5. 0-10.4
0.0-7.2
                                   104

-------
TABLE C.2.  RANGE OF AVERAGES WATER QUALITY CONSTITUENTS -
                       HANCOCK FARM

Parameter
Alkalinity (mg CaC03)
Specific Conductivity (umhos/cm)
TDS (mg/1)
pH
Cl (mg/1)
S04 (mg/l>
Total N (mg N/l)
N02 + N03 (mg N/l)
NH3 (mg N/l)
Total. P .(mg P/l)
Ortho P (mg P/l)
Org P (mg P/l)
COD (mg/1)
TOC (mg/1)
Total Coliform (counts/100 ml)
Fecal Coliform (counts/100 ml)
Fecal Strep (counts/100 ml)
Metals-Dissolved (mg/1)
Al
As
Ba
B
Ca
Cd
Co
Cr
Cu

Monitoring
6/80-1/82
230-349
504-1287
363-989
7.32-7.2
22-246
32-243
0.10-4.21
0.79-10.90
0.03-1.36
0.08-0.58
<1. 01 -0.29
<0. 01-5. 50
6.6-116.7
7.1-53.3
0-200,004,032
'0-150,002,752
0->4,604

0.124-9.180
<0. 005-0. 018
0.042-0.369
0.188-1.186
24-150
<0. 001 -0.004
<0. 005-0. 016
<0. 005-0. 031
<0. 005-0. 090

Period
2/82-12/83
238-382
694-1933
462-1250
7.33-8.14
17-345
44-271
0.11-24.89
0.10-14.99
0.03^15.15
0.02-1.02
<0. 01-0. 85
<0. 01 -0.08
5.8-84.3
0.9-15.4
65-43,275
0-45,005
8-19,200

0.173-4.656
<0. 010-0. 010
0.022-0.236
0.510-2.175
29-118
<0. 001-0. 003
<0. 005-0. 008
<0. 005-0. 008
<0. 005-0. 126
(continued)
                          105

-------
Table C.2.,  continued
Parameter
Fe
Mg'
Mn
Mo
Ni
K
Se
Ag
Na
Tl
Metals, Total (mg/1)
Al
As
Ba
B
Ca
Cd
Co
Cr
Cu
Fe
Pb
Mg
Mn
Mo
Ni
K
Se
Ag
Na
Monitoring
6/80-1/82
0.170-4.688
21-70
0.006-0.650
0.004-0.058
<0. 005-0. 162
7-13
<0. 005-0. 011
<0.005
30-148
0.028-0.247

0.114-6.650
<0. 003-0. 020
0.059-0.482
<0. 100-0. 792
29-103
<0. 001-0. 008
. 0.003-0.005
0.002-0.047
0.004-0.112
0.113-3.710
<0. 002-0. 022
37-85
0.003-1.230
0.010-0.085
<0. 005-0. 133
8-15
0.001-0.009
<0.005
78-686
Period
2/82-12/83
0.337-8.885
28-69
0.009-1.642
NR
<0. 005-0. 190
6-17
<0. 005-0. 134
<0. 001-0. 005
34-181
<0. 005-0. 007

0.379-2.850
0.002-0.011
0.059-0.150
NR
20-137
NR
<0.005-
<0.005-
<0. 005-0. 006
0.580-6.410
<0. 002-0. 006
31-55
0.020-0.132
NR
<0.005
11-18
<0. 005-0. 009
<0.001
78-159
                                                             (continued)
                                  106

-------
Table C.2.,  continued
Parameter
Tl
Zn
Qrganics (ppb)
Acenapththylene
Anthracene
Atrazine
Benzene
4-t-Butylphenol
Carbon Tetrachloride
4-Chloroaniline
Chlorobenzene
Chloroform
2-Chlorophenol
1 -Chlorotetradecane
Dibutylphathalate
2 ,3-Dichloroaniline
3 ,4-Dichloroaniline
Dichlorobenzene meta
Dichlorobenzene para
Dichlorobenzene ortho
2,4-Dichlorophenol
Diethylphthalate
Diisooctylphthalate
Dioctylphthalate
Ethylbenzene
Heptadecane
Hexadecane
Hexadecanoic Acid
Methylheptadecanoate
Monitoring
6/80-1/82
0.001-0.009
0.072-0.0436
-
<5.0-5.9
<2.0-4.4
<2.0-38.9
<1.0-2.3
<2.0-75.9
<5. 0-193.0
<10. 0-11.1
<1.0
<1.0-7.4 •
<2.0
<2.0-52.2
<2. 0-111 .0
<5.0-9.2
2.0-7.1
<2. 0-3.9
<2.0-9.6
<2.0-3.6
<2.0-5.8
<2.0-16.7
9.9-223.8
<2.0-62.2
<1.0-2.6
<2.0-3.7
<2.0-2.5
17.6-36.0
<2. 0-36.0
Period
2/82-12/83
<0.005
<0. 020-0. 096

<3.0-3.9
<2.0-7.2
4.7-13.9
<1.0-3.7
1.3-2.5
2.8-5.4
<10.0
<1.0-1.6
<1. 0-12.2
<2.0
<2.0-8.9
2.6-16.4
2.8-8.2
<2.0-8.3
<2.0
<2.0
<2.0
<2.0-5.8
3.0-62.3
<2. 0-378.0
<2.0-2.9
<1.0-1 .8
<2. 0-3.1
<2.0
<1.0
<2.0-3.4
                                                               (continued)
                                 107

-------
Table C.2.,  continued
     Parameter
                                                  Monitoring Period
 6/80-1/82
  2/82-12/83
     Methylhexadecanoate
     1-Methylnaphthalene
     2-Methylphenol
     4-Methylphenol
     Naphthalene
     Octadecane
     Phenol
     Propazine
     a-terpineol
     Tetrachloroethylene
     Toluene
     Trichloroethane
     Trichloroethylene
 <2.0-560.0
 <2.0-2.1
 <2.6-2.8
 <5.0-6.8
 <2.0-7.0
 <2.0-6.7
<10.0-24.9
 <2.0-4.5
 <2.0-4.4
 O.O-
 0.0-12.3
 <5.0-11.5
 <1.0-8.9
 <2.0-4.3
 2.0-3.2
 <2.0
 <5.0
 <2.0-3.4
 <2.0-6.1
<10.0
  5.6-9.8
  2.0
 <1.0-4.9
 <1 .0
 <5.0
 0.0-5.8
                                      108

-------
   TABLE C.3.  DEPTH OF PERCOLATE  INTERCEPTED  BY LYSIMETER  UNITS  OVER  STUDY  PERIOD  AT  THE  HANCOCK  FARM

Location
Tray
0.61 m

1.22 m


1.83 m

Tube
1.22 m


1.83 m

2.44 m

Controls
1.22 m

1.83 m


Bermuda
Units
101a
102
103
Avg.b
104
105
106
Avg.
107
108
109
Avg.
111
112
Avg.
113
114
Avg.


121
122
Avg.
123
124
Avg.
Depth
(cm)
5.1
12.2
4.4
7.2
0.9
0.9
Neg.
1 .1
0.3
4.1
2.2

0.1
0.1
108.1
43.1
75.6


171.3
155.3
163.3
113.3
114.9
114.1
Grain Sorghum
Units
201
202
203
Avg.
204
204
206
Avg.
207
208
209
Avg.
211 .
212
Avg.
; 213
214
Avg.
215
216
Avg.




Depth
(cm)
0.8
1.7
1.25
0.8
0.8
0.8
0.3
Neg
Neg
0.3

0.5
0.5
Neg
1.4
1.4
	
	





Cotton
Units
301
302
303
Avg.
304
304
306
Avg.
307
308
309
Avg.
311
312
Avg.
313
314
Avg.







Depth
(cm)
Neg
3.9
3.9
3.5
3.5
3.2
0.9
1 .6
0.7
2.9
0.8
1.5
10.8
1.2
6.0
2.0
2.0
1.2








f?Unit code — the first digit identifies the plot  and  the  next  two  identify  lysirneter  type  and  depth.
DAverage of producing units.

-------
    TABLE C.4.  DEPTH OF PERCOLATE INTERCEPTED BY LYSIMETER UNITS OVER STUDY PERIOD AT THE GRAY SITE

Bermuda
Location Units
Tray
0.6 m 101a
102
103
Avg.b
1.22 m 104
105
106
Avg.
1.83 m 107
108
109
Avg.
_» Tube
0 1.22 m 111
112
Avg.
1.83 m 113
114
Avg.
2.44 m 115
116
Avg.
Depth
(cm)

	
Neg
6.1
6..1
0.3
1.8
0.6
0.9
3.3
4.1
1.5
3.0
31.1
10.8
21.0
10.1
14.1
12.1
28.9
25.1
27.0
Cotton
Units

201
202
203
Avg.
204
205
206
Avg.
207
208
209
Avg.
211
212
Avg.
213
214
Avg.



Depth
(cm)

12.3
3.6
6.6
• 7.5
5.2
0.6
4.3
3.4
2.9
0.2
2.0
1.7
5.8
6.7
6.3
0.4
0.4
0.4



Grain Sorghum
Units

301
' 302
303
Avg.
304
305
306
Avg.
307
308
309
Avg.
311
312
Avg.
313
314
Avg.



Depth
(cm)

17.2
33.2
15.5
22.0
1.7
7.2
	
4.5
1.3
1.0
4.3
2.2
5.0
18.8
11.9
17.0
33.1
25.1




aUnit code — the first digit identifies the plot and the next two identify lysimetr type and depth.
DAverage of producing units.

-------
                       CONTOUR ELEVATIONS IN FEET (.3048 M/ft)




Figure C.1.  Water Level Contours,  December  1981,  Gray Farm

-------
     CONTOUR ELEVATIONS IN FEET
        (.3048 METERS/FT)
Figure C.2.  Water Level Contours, December 1981,  Hancock Site
                             112

-------