United States
            Environmental Protection
            Agency
           Robert S. Kerr Environmental Research
           Laboratory
           Ada OK 74820
EPA-600/4-79-050
August 1979
            Research and Development
&EFK
Identifying
Sources of
Subsurface Nitrate
Pollution with
Stable Nitrogen
Isotopes

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology.  Elimination  of traditional grouping was  consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

      1.  Environmental  Health Effects Research
      2.  Environmental  Protection Technology
      3.  Ecological Research
      4.  Environmental  Monitoring
      5.  Socioeconomic Environmental Studies
      6.  Scientific and Technical Assessment Reports (STAR)
      7.  Interagency  Energy-Environment Research and Development
      8.  "Special" Reports
      9.  Miscellaneous Reports

This report has been assigned to the ENVIRONMENTAL MONITORING series.
This series describes research conducted to develop new or improved methods
and  instrumentation for the identification and  quantification of environmental
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia  22161.

-------
                                      EPA-600/4-79-050
                                      August 1979
        IDENTIFYING SOURCES OF SUBSURFACE
             NITRATE POLLUTION WITH
            STABLE NITROGEN ISOTOPES
                      by

    Timothy J.  Wolterink,  Hugh J. Williamson,
     David C. Jones, Thomas W. Grimshaw, and
                  W. F.  Holland
               Radian Corporation
              Austin, Texas  78758
           Contract No. 68-03-2450
                 Project Officer

                 Jack W. Keeley
          Ground Water Research Branch
Robert S. Kerr Environmental Research Laboratory
              Ada, Oklahoma  74820
ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
       OFFICE OF RESEARCH AND DEVELOPMENT
      U.S. ENVIRONMENTAL PROTECTION AGENCY
              ADA, OKLAHOMA  74820

-------
                          DISCLAIMER


     This report has been reviewed by the Robert S.  Kerr
Environmental Research Laboratory, U.S.  Environmental Protection
Agency and approved for publication.  Approval does  not signify
that the contents necessarily reflect the views and  policies of
the U.S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or
recommendation for use.
                               11

-------
                            FOREWORD


     The Environmental Protection Agency was established to
coordinate administration of the major Federal programs designed
to protect the quality of our environment.

     An important part of the Agency's effort involves the search
for information about environmental problems , management tech-
niques ,  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.

     EPA'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
to:  (a)  investigate the nature, transport, fate, and management
of pollutants in ground water; (b) develop and demonstrate methods
for treating wastewaters with soil and other natural systems ;
(c) develop and demonstrate pollution control technologies for
irrigation return flows; (d) develop and demonstrate pollution
control technologies for animal production wastes; (e) develop
and demonstrate technologies to prevent, control or abate pollu-
tion from the petroleum refining and petrochemical industries ;
and (f)  develop and demonstrate technologies to manage pollution
resulting from combinations of industrial wastewaters or indus-
trial/municipal wastewaters.

     This report contributes to that knowledge which is essential
in order for EPA to establish and enforce pollution control stand-
ards which are reasonable,  cost effective, and provide adequate
environmental protection for the American public.
                                   William C. Galegar
                                        Director
                               111

-------
                            ABSTRACT

     This report describes the methods,  results, conclusions,
and recommendations of an investigation of a technique to iden-
tify sources of nitrate ions in ground water.   A discussion of
the theoretical basis of the technique is also provided.   Over
300 soil and ground-water samples were collected for this study.
The samples are from numerous sites from within the United
States, representing a variety of environmental conditions.  The
nitrate ions in 66 of these samples were separated from other
nitrogen species,  converted to nitrogen (No) gas. purified, and
analyzed to determine the nitrogen-isotope ratio (15M/lttN).
These data were combined with the results of analyses performed
previously by Jones  (1) and Kreitler (2).   Standard statistical
techniques were used to analyze the observed variations in 615N
values, with respect to several nitrate-ion sources and various
environmental factors.  It was found that nitrate ions from feed-
lots, barnyards , and septic tanks can be distinguished from nat-
ural soil nitrate ions on the basis of the 615N values.  These
nitrate ions from various sources cannot,  however,  be distin-
guished  from each other.  Environmental factors contributed to
the observed variation in 615N values.

     This report was submitted in fulfillment of Contract No.
68-03-2450 by Radian Corporation under the sponsorship of the
U.S. Environmental Protection Agency.   This report covers the
period August 23,  1976 to March 31, 1978,  and work was completed
as of November 1,  1978.
                               IV

-------
                           CONTENTS

Foreword	  iii
Abstract.	  iv
Figures	  vii
Tables	  viii
Acknowledgements	  x

     1.   Introduction	  1
           Goals and Approach of This Study.	  1
           Significance of the Nitrate Problem	  2
     2 .   Conclusions	  4
     3 .   Recommendations	  6
     4.   Rationale for Tracing Nitrate Via Stable Isotope
         Ratios	  7
           Nitrogen Isotope Abundance	  7
           Isotope Fractionation Mechanisms	  9
           Nitrogen in the Environment	  13
           Nitrate Source Determination	  18
           Other Stable Isotopes As Tracers	  21
     5 .   Experimental Procedures	  26
           Selection of Sampling Sites	  26
           Sampling Procedures.	  28
           Sample Preparation and Analysis	  30
           Analysis of Standards	  37
     6.   Sampling Sites	  39
     7 .   Results and Discussion	  46
           Statistical Analyses of the Data.	  46
           Analytical Problems.	  68
           Summary and Conclusions.	  73
                               v

-------
                        CONTENTS (Continued)

References	 77
Appendices	 79
     A.  Reagents Used in Isotopic Analyses	 79
     B.  Description of Radian's Intensive Sampling Program,
         By Sites.	 81
     C.  Sample Numbering System	 122
     D.  Basic Data.	 125
     E.  Climate Classifications	 134
     F.  Soil Classifications	 136
     G.  Statistical Analyses of the Entire Data Base	 139
     H.  Separate Statistical Analyses for the Different
         Source Types	 144
                             VI

-------
                            FIGURES


Number                                                     Page

  1      Simplified Schematic of the Nitrogen Cycle	  16

  2      Mechanisms of Nitrate Formation	  19

  3      Distillation Apparatus Used for Removal of
         Interfering Nitrogen Species	  32

  4      Nitrogen Gas Generation and Purification
         Apparatus	  34

  5      Nitrogen Gas Generation Tube	  35

  6      Locations of Sampling Sites	  42

  7      Histograms of Del Values by Source Category	  50

  8      Illustration of the Meaning of Correlation
         Coefficient, r	  55

  9      Depth Profile of Nitrate Concentration and 615N
         for the Greeley, Colorado, Animal Feedlot Site,
         Hole H2	  71

 10      Nitrate and 615N Versus Depth Beneath a Barnyard
         With a Definite Animal Waste Contribution of
         Nitrogen, In Runnels County,  Texas	  72
                             VII

-------
                            TABLES


Number                                                     Page

  1      Nitrogen Isotope Ratios for Various Substances....  8

  2      Isotope Equilibrium Constants	  11

  3      Kinetic Fractionation Factors For Various Non-
         Equilibrium Reactions	  14

  4      Summary of Samples Collected In Radian's
         Intensive Sampling Program.	  41

  5      Key to Figure 6	  43

  6      Site Environmental Characteristics	  45

  7      Summary of Samples Included in the Final Data Base,
         By Source Category and Site.	  47

  8      Means and Standard Deviations of 615N For
         Different Nitrate Source Categories	  48

  9      Definitions of Variables Used In the Correlation
         Analyses	  52

 10      Value Required For A Correlation Coefficient (r)
         To Be Significant, As A Function of Significance
         Level and Sample Size	  57

 11      Variables Which Have Statistically Significant
         Correlations With DEL	'	  58

 12      Correlations Involving Variables Most Closely
         Related to DEL„	  58

 13      Variables Which Have Statistically Significant
         Correlations With PPM	  61

 14      Variables Which Have Statistically Significant
         Correlations With DEL:  Septic Tank Sources	  63

 15      Variables Which Have Statistically Significant
         Correlations With DEL:  Feedlot Sources	  64

                             viii

-------
                       TABLES (Continued)


Number                                                      Page

 16      Variables Which Have Statistically Significant
         Correlations with DEI,:   Irrigation Sources	 65

 17      Composite Relationships Between DEL and Predictor
         Variables	 67

 18      Variables Which Correlate Most Highly With PPM
         (At the  .01 Level of Significance) for the Various
         Source Categories	 68
                              IX

-------
                       ACKNOWLEDGEMENTS

      This  study was sponsored by the Environmental Protection
 Agency.  The Project Officer for EPA was Jack W. Keeley.  The
 investigation was carried out by Radian Corporation.  The
 Project  Director for Radian was David C. Jones.  Timothy J.
 Wolterink  directed the field investigations and sampling program,
 and oversaw the analytical activities.  Thomas W. Grimshaw as-
 sisted in  the sampling program.  Hugh J. Williamson conducted
 the statistical analyses and provided much valuable assistance
 in interpreting their results.  W.F. (Kirk) Holland provided
 many helpful suggestions and assisted in interpreting the results
 John T.  Robinson prepared the samples for isotopic analysis.

      Many people outside of Radian contributed much to the com-
 pletion  of  this study.   In particular,  Dr.  Charles W.  Kreitler,
 of  the Bureau of Economic Geology,  University of Texas at Austin,
 performed the isotopic analyses, provided isotope standards, and
 participated in many helpful discussions.

     A great many people provided invaluable assistance in ob-
 taining  samples or background information on various sites.
 Special  thanks go to Kirk W.  Brown (Department of Soil and Crop
 Sciences, Texas A&M University); Sam B.  Upchurch (Department of
 Geology, University of  South Florida at Tampa);  Karl Ford
 (Jefferson County,  Colorado,  Health Department);  Richard Bissell
 (Geological Survey Division,  Michigan Department of Natural
Resources); Thomas  B.  Bahr and Tom Burton (Institute of Water
Research, Michigan State University);  Dave Meier (Roscommon
County, Michigan,  Health Department);  Louise Sullenger (Houghton
Lake, Michigan);  and Bill Leseman (City of Tallahassee,  Florida).
                                x

-------
     Finally, thanks go to the personnel of the Oregon Depart-
ment of Environmental Quality; the West Virginia State Health
Department; the Minnesota Department of Health; and the Rolla,
Missouri, District Office of the U.S.  Geological Survey for
agreeing to provide samples via the kit sampling program.
                               XI

-------
                            SECTION 1
                          INTRODUCTION

GOALS AND APPROACH OF THIS STUDY

     This report describes the results of an investigation of a
technique to identify sources of nitrate (NOs) in ground water.
The investigation was funded by the U.S. Environmental Protection
Agency.

     The technique under investigation was developed previously
by Radian Corporation under EPA Grant 16060 HNI (1).  It is based
on the premise that nitrate in soils and ground water will have
a nitrogen-isotope ratio (15N/11+N) that is determined by (a) the
series of reactions which formed the nitrate, and (b) the nitrogen-
isotope composition of its precursors.  Previous work (1, 2) has
indicated that soil and ground-water nitrates that were derived
from animal wastes can be distinguished from other soil nitrates
on the basis of their nitrogen isotope ratios.  A detailed descrip-
tion of the technique is presented in Section 4 of this report.

     The overall objective of this study was to assess the general
feasibility of using stable nitrogen isotope ratios to identify
the source(s) of nitrate pollution in ground water.  Within this
context, two key questions were addressed:

     1.   For a given waste source, are the isotope
          ratios of nitrate-nitrogen statistically
          equivalent in different geographic and climatic

-------
         regions of the country? and

     2,   For different waste sources,  is the variance
         of the isotope ratio means adequate to dif-
         ferentiate between these sources?

Particular emphasis was placed on evaluating the technique with
respect to its ability to differentiate between domestic septic
tank, animal feedlot,  and municipal wastewater land-application
(irrigation) sources of nitrate.

     In pursuing this  study, soil and ground-water samples were
obtained from numerous sites around the United States.  Emphasis
was placed on acquiring sufficient numbers of samples to obtain
statistically significant results.  Samples were prepared and
analyzed using equipment and procedures previously developed by
Radian under EPA Grant 16060 NHI  (1).   These procedures are
described in Section 5 of this report.  Isotopic ratio determina-
tions were performed at the mass  spectrometer facility of the
Center for Research in Water Resources, University of Texas at
Austin.  Standard statistical procedures were then used to evaluate
the results of the analyses.

SIGNIFICANCE OF THE NITRATE PROBLEM

     The presence of excessive nitrate in drinking water sup-
plies is a well-recognized public health hazard.  The best known
health problem arising from nitrate consumption is methemoglo-
binemia,  a cyanosis which is brought about by the reduction of
nitrate (NOa) to nitrite (NOa) by bacteria in the digestive
tract, followed by absorption of the nitrite into the blood-
                                                        I I
stream.  There, the nitrite oxidizes the ferrous ion (Fe  ) in
                            I ||
hemoglobin to ferric ion (Fe   ),  thereby preventing the trans-
port of oxygen by the hemoglobin.   This results in a gradual
suffocation (cyanosis).  It is noteworthy that infants are most

-------
susceptible  since  the acidity of their stomach is considerably
less  than  that of  adults, resulting in a more favorable environ-
ment  for the nitrate-reducing bacteria.  The digestive system of
animals such as  cattle  (ruminants) also is conducive to this
bacterial  action.  Therefore, when a nitrate problem occurs, it
is first reflected in health problems of cattle or human infants

      It should also be  noted that the whole problem of methemo-
globinemia is poorly understood.  The nitrate or nitrite concen-
tration at which methemoglobinemia becomes a problem may vary
widely, and  is most likely influenced by some as yet unidenti-
fied  factor  or factors.  Because of the methemoglobinemia prob-
lem,  the U.S. Environmental Protection Agency has established
a maximum  contaminant level (MCL) for nitrate (as nitrogen)  of
10 milligrams per liter (mg/£).   This  MCL  was  established  for
drinking water supplies, under  the National Interim  Primary
Drinking Water Regulations  (3).

      Another potential  health effect of nitrate relates to can-
cer.  Nitrate and/or nitrite can react with secondary amines to
form  nitrosamines, many of which are carcinogenic.  Numerous
studies indicate a possible link between nitrate and cancer (4,
5,6).

      If ground waters become contaminated with nitrate as a
result of  man's  activities, a public health hazard could arise.
Moreover,  some ground waters naturally contain high levels of
nitrate.   The technique under investigation in this  study has
been  proposed as a means of identifying the source(s) of nitrate
in ground  water.   This  investigation was aimed toward identify-
ing any limitations and potential applications of the technique.

-------
                            SECTION 2
                           CONCLUSIONS

1.   Nitrates from septic tanks,  feedlots,  and barnyards can be
     distinguished from natural soil nitrates  on the basis of the
     isotopic compositions.

2.   Nitrates from septic tanks,  feedlots,  and barnyards cannot
     be distinguished from each other.

3.   The average nitrogen isotope ratio del value (615N) of
     nitrates from municipal wastewater irrigation sites falls
     between the average for feedlots  and septic tanks and the
     average for natural soil nitrates.

4.   The range of 615N values of nitrates from municipal waste-
     water irrigation sites overlaps the range for feedlots and
     septic tanks and the range for natural soil nitrates.

5.   Therefore,  nitrates from municipal wastewater irrigation
     sites cannot be  distinguished from nitrates derived from
     septic tanks, feedlots, barnyards, or from natural soil
     nitrates.

6.   The isotopic composition of nitrates from municipal waste-
     water irrigation sites suggests that these nitrates are a
     mixture of  nitrates derived from the waste and nitrates
     derived from natural soil processes.

7.   Environmental factors influence the 615N value of soil and
     ground-water nitrates.
                                4

-------
 8.   Only about 30 percent of the observed variation in 615N can
     be attributed to the environmental factors assessed in this
     study.

 9.   Identifying causal relationships between 6 * 5N values  and
     specific environmental factors will require  additional
     research.

10.   The ranges of 615N values for specific nitrate source
     categories overlap.

11.   Distinguishing between nitrates from animal  or human  wastes
     and natural soil nitrates will therefore require a compari-
     son of mean 6l5N values.

12.   Analysis of large numbers of samples will improve the abil-
     ity to distinguish between mean 615N values.

13.   The isotope technique for identifying sources of nitrate
     is not suitable for  use as a "routine" procedure,  because:

     a.  sophisticated analytical equipment,  which is not
         commonly available, is required to prepare and an-
         lyze samples;  and

     b.  a high level of  familiarity with the technique,
         the chemistry of nitrogen, and nitrogen  isotopic
         systematics is required on the part of the inves-
         tigator,  in order to  insure the proper analysis
         of samples and interpretation of results.

-------
                            SECTION 3
                         RECOMMENDATIONS

     Additional information is needed to make the nitrogen isotope
ratio technique a generally applicable tool for tracing nitrate
sources.  Studies should be performed to determine whether varia-
bles such as soil oxidation-reduction potential (Eh),  pH, and
bacteria exert a major influence on the del value.

     These studies also should include variables considered in
this program, e.g. precipitation, elevation, soils, etc.  Corre-
lation studies similar to those used in this study should be
performed on the larger data base to determine whether variations
in del value can be explained.

     If these expanded studies are successful in explaining the
del values, the results must be studied to develop a practical
scheme which would permit the sampling of a location to identify
the source of subsurface nitrate.  This might require on-site
measurements of parameters such as soil Eh and pH.

-------
                            SECTION 4
                  RATIONALE FOR TRACING NITRATE
                    VIA STABLE ISOTOPE RATIOS

NITROGEN ISOTOPE ABUNDANCE

     There are two naturally occurring, stable  (non-radioactive)
isotopes of nitrogen,  l "N and  15N.  The nucleus of a  * "*N atom
contains seven protons and seven neutrons, while the nucleus of
a  l 5N atom contains seven protons and eight neutrons.  The nitro-
gen gas in air is composed of 99 . 632 percent  1 4N and  0.368  percent
 15N.  Junk and Svec  analyzed air  samples  from many localities  ana
 found no  significant  variation  in  the  isotope composition  of  at-
mospheric nitrogen  (7).  In  other nitrogen- containing  compounds  the
 15N occurrence may be difiverent from tliat in  air, however.

     The  1 5N content  of any particular nitrogen species is ex-
pressed in terms of the nitrogen isotope  ratio  del value (6) ,
in parts per thousand (ppt) , defined as follows:

        i5  = 15N/^N  (sample) - l 5N/ 1 "N  (air)
                         1514
                          N/14N  (air)

     It can be seen that the nitrogen isotope- ratio in air is
used as a reference, and therefore has a 6 value of zero.  Sub-
stances with a higher  1 5N content than air will have a positive
6 value, while substances with a lower 1 5N content than air will
have a negative 6 value.  Del values of naturally occurring sub-
stances generally range from -20 to +20.   Selected nitrogen iso-
tope ratios determined for various substances are shown in Table  1

-------
    TABLE 1.  NITROGEN ISOTOPE RATIOS FOR VARIOUS  SUBSTANCES*
Sample (Source)
Inorganic Nitrogen
Chile nitratite (Chile)
Sal ammoniac (Mexico)
Rocks and Minerals
Pitchblende (Canada)
Granite (Maine)
Oil and Gas
Bitt No. 1 Oil (Oklahoma)
Natural Gas (Arkansas)
Plaisted No. 1 Oil (Oklahoma)
Peat and Coal
Peat (New York)
Bituminous coal (Pennsylvania)
Peat (Eire)
615N (ppt)

-2.6
+13.0

-2.3
-0.2

-11.5
-5.9
-3.5

-2.8
-0.9
+1.9
Animal  Protein
     Rat  liver  (Unknown)                              -3.3 to +9.3
     Lamb flesh  (Unknown)                                 +5.0
     Clam flesh  (Atlantic Ocean)                          +7.3
     Marine zooplankton  (Unknown)                        +10.0
     Marine fish flesh  (Unknown)                         +15.0

Plant Protein
     White Clover leaves  (Local)                          -6.5
     Dandelion leaves  (Local)                             -2.8
     Oats (Unknown)                                       +6.2
     Seaweed (Unknown)                                    +7.0

*From Jones (1)  and Kreitler  (2).
                                 8

-------
An isotope ratio mass spectrometer is used to measure 6 values.

ISOTOPE FRACTIONATION MECHANISMS

     Differences in the nitrogen isotope ratios of different
species occur because of the difference in mass between l5N and
14N atoms.  Since an atom of 15N contains an extra neutron, it
has a mass approximately 7  percent  greater  than a 14N atom.  This
 greater mass  causes 15N atoms  to behave slightly differently  from
 1^N atoms in  physical and chemical processes.   These behavioral
 differences  can produce variations in the  15N/1!4N ratios  of
 different nitrogen compounds (isotopic fractionation).

     The  isotopic ratio of a nitrogen  compound  is the result of:

     1.   Physical fractionation;
     2.   Chemical equilibrium  fractionation;
     3.   Chemical kinetic fractionation; and
     4.   The original isotopic composition of the material from
          which  the  compound was  formed.

     Physical fractionation, chemical  equilibrium fractionation,
and chemical kinetic fractionation are  all important  in the iso-
topic fractionation of nitrogen  in natural biogeochemical systems.
However,  their  relative importance is  only beginning  to be under-
stood.

     Physical fractionation can  occur  through diffusion,  evapora-
tion, and sublimation.  In diffusion the light  isotope  (ll*N) will
have a higher velocity.  In evaporation and sublimation the
lighter isotope will have a higher vapor pressure.

     Chemical equilibrium fractionation  (isotope exchange equilib-
rium) is  the tendency of an isotope to  concentrate in one species
of a chemical equilibrium reaction.  When two chemical  species
                                 9

-------
are in equilibrium, the isotopic distribution also reaches  an
equilibrium (i.e., the isotopic ratio in one species may not be
the same as the ratio in the the other species).  The  distribu-
tion of isotopes between the species is dependent on the differ-
ent energy levels of the molecules, and is described by the iso-
tope equilibrium constant  (K or a).  Isotope equilibrium con-
stants may be calculated or determined experimentally.

     The calculated isotope equilibrium constant is "K", where

                              [Products]
                        K =
                             [Reactants ]
     The experimentally determined isotope equilibrium  constant
 is  "a".  K and a are related as follows:
                               v-
                           a = K
where n is the number of equivalent exchangeable atoms  in  the
reaction .

     Isotope equilibrium constants have been calculated from
partition functions and/or determined for several nitrogen equi-
librium reactions relevant to surface or near-surface  conditions
These are listed in Table 2.

     The meaning of the isotope equilibrium constant may be
illustrated using the second reaction in Table 2:

          1 'tMtr      i  1 5|arj     _  1 SVTTT       i ItMTj
                    +   NH3(g) -   NH4(aq) +   NH3(g)
This is the nitrogen isotope exchange equilibrium reaction when
ammonium is in chemical equilibrium with gaseous ammonia  (ammonia
volatilization) .   From the equation for the calculated equilibrium

                                10

-------
                              TABLE  2.   ISOTOPE EQUILIBRIUM CONSTANTS*
         Reaction
                                                  T°K
                   Isotope Equilibrium  Constant	
              Calculated (K)        Experimental (a)
         5N2) = !
                                                  298.1
                 1.023
         +15NH3   =15NH1|     +1J+NH3
                (g)       (aq)       (g)
                                                  298.1
                 1.035
1.034 ± 0.002
  NH1|+1 5NO=:
                                                  298.1
                 1.038
                                                    298.1
                                                                   1.012
1UNH3    +15NH3   =15NH3     +llfNH3
     (aq)       (g)       (aq)       (g)
                                                  298.1
                                          1.005
15
NH
         +14NH4    =11*NH3     +15NH!4
     (aq)       (aq)       (aq)      (aq)
298.1
    1.029
*From Kreitler  (2).

-------
constant,  the second reaction in the table is written:
    K - 1.035 =
                           (15NH3)      (15NH3/ll>NH3)
                       (aq)       (g)                (g)
 This  equation states that there should be a 35 parts per  thou-
 sand  (ppt) enrichment of 1 5N in aqueous ammonium compared to
 ammonia gas .

      In the  equations in Table 2 there is only one exchange  atom
 (i.e., n  = 1) , and the relation between K and a is therefore:

                             a = K.

 For the second reaction in Table 2,
                   a = 1.034 =
                               1 5M/ 1 if
N/14N (NH3)
 This  equation states that a 34 ppt enrichment of  l 5N  in ammonium
compared to ammonia,  was  determined.

      Chemical kinetic fractionation describes the  tendency  of
 the two isotopes to react at different rates when  undergoing
 the same reaction.  If all the reactants are not  converted  into
 products (because of side reactions or incomplete  reactions)
 the nitrogen isotope ratio will differ between  the reactants and
 products.  The particular isotope ratio for the products will
 depend on  (a) the initial isotope ratio of the  reactants, and
 (b) the particular reaction which occurred.
                               12

-------
     It should be noted that environmental parameters can
affect isotope fractionation.  For example, high temperatures
(hundreds of degrees) tend to lessen the tendency of the two iso
topes to react at different rates.  Thus, less isotope fraction-
ation would be expected at high temperatures „   This is because
the slight differences in energy levels  (due to the slightly
different masses of the nitrogen atoms) become less important
as the total energy of the system is increased by the added heat
(molecular motion) .   Bacterially-aided reactions are another
example of environmental influence on isotope fractionation.
Such reactions often exhibit significant fractionation.

     The kinetic fractionation factor B is used to describe
isotopic fractionations in nonequilibrium reactions.  Table 3
lists the values for B that have been determined for several
nitrogen reactions,  as summarized by Kreitler (2).  Since no
single convention for the usage of B has been established, the
equation defining B is given for each of the reactions listed in
Table 3.  For example, the equation for B for the first step of
the nitrification process (oxidation of ammonia to nitrite;
reaction #2 in Table 3) is written:

                     = ^N/^N (NH3) = ,
                            *N (N02)

The value given for B, 1.026, indicates that there is a 26 ppt
enrichment of l 5N in ammonia relative to nitrite.

NITROGEN IN THE ENVIRONMENT

The Nitrogen Cycle

     The earth nay be viewed as having a natural "metabolism",
in which materials are continually circulated through the atmo
sphere; to the oceans via streams; in and out of the biota via

                                13

-------
                 TABLE 3.  KINETIC FRACTIONATION FACTORS FOR
                           VARIOUS NONEQUILIBRIUM REACTIONS*
       Reaction
                                       Equation  for B
                                 Measured
                                Values of B
1.  Nonsymbiotic bacterial
   nitrogen fixation.
B =
                                     i S
                                     15
(atmospheric N2)
(fixed nitrogen)
                               1.0022-1.004
2.  Nitrification: two-step
   bacterial oxidation of
   ammonia to nitrate.
   Only the first step has
   been measured.
                                     1 5,
B =
    1 5 X
            (NH3)
            (N02)
                                  1.026
3.  Denitrification:  bacterial
   reduction of nitrate to
   nitrogen gas.
B =
            (NO 3)
            (N2)
                               1.0173-1.02
4.  Cation exchange of
   ammonium with exchange
   resins or kaolinitic
   clay.
B =
            (solution NHQ
    15N/1!tN (adsorbed
                             0.99926-0.99922
5.  Anion exchange of
   nitrate with exchange
   resin.
B =
    15N/1[tN (solution N03)

    15N/lltN (adsorbed N03)
                      1.0021
*Reported by Kreitler (2).

-------
photosynthesis, growth, and decay; and through the lithosphere
via sedimentation, weathering, and erosion.  Each element follows
a particular path or set of pathways through the natural system.
The paths are determined by the biogeochemical and physical prop-
erties of the elements.  Each has its own natural biogeochemical
cycle.

     The global nitrogen cycle is extremely complex, and many
of its aspects are poorly understood.  A simplified schematic
of the nitrogen cycle  is shown in Figure 1.  The figure shows
the various nitrogen reservoirs (e.g.,  the atmosphere, the bio-
sphere, sedimentary rocks, etc.) and the gross aspects of nitro-
gen transfer between these reservoirs.

     The major nitrogen reservoir is the atmosphere.  In this
respect, nitrogen is somewhat unique.  Other elements for which
global cycles have been investigated (e.g., carbon, sulfur,  and
phosphorus) have their major inventories in the sedimentary rock
reservoir (8).   Several nitrogen species are significant in the
atmosphere.  Of these, however, nitrogen gas is by far the most
abundant, accounting for more than 99.999 percent of all atmos-
pheric nitrogen.  Ammonia, ammonium, and oxides of nitrogen are
also present, but together they account for less than 0.001 per-
cent of the atmospheric nitrogen (8).

     The transfer of nitrogen from one reservoir to another is
indicated in Figure 1 by the various arrows.   The mechanisms by
which nitrogen is transferred are very complex, involving a
large number of pathways and many nitrogen species.  Many of
the mechanisms  are only beginning to be understood.  The situa-
tion is complicated by such factors as the recycling of nitrogen
in the biosphere,  the complex photochemical reactions involving
nitrogen oxides in the atmosphere, and the influence of man's
                                15

-------
Uplift,
Weathering
Deposition
Igneous
Rocks
Sedimentary
Rocks
    Figure 1.  Simplified Schematic of the Nitrogen Cycle,
                         16

-------
activities (e.g., fixation of atmospheric nitrogen gas for
fertilizer manufacture).

Nitrate In the Terrestrial Environment

     In the context of this study, the significant portions
of the nitrogen cycle are those that lead to the presence of
nitrate (N03) in soils and ground waters.

     The major input of nitrogen to the soil-water-biomass
subsystem of the nitrogen cycle is the fixation of atmospheric
N2.   This is accomplished primarily through natural bacterial
fixation.   Within the last 100 years, however,  the demand for
fixed nitrogen compounds as fertilizers has grown to the point
where the. amount of industrially fixed atmospheric N2 nearly
equals the amount fixed naturally  (8).

     Other somewhat less important inputs to the soil-water-
biomass system also exist.  Ammonium and nitrate are directly
deposited on the land surface by rainout of both natural and
pollutionally derived ammonium nitrate and ammonium sulfate.
Estimates of the rate of addition of nitrogen by this mechanism
are very uncertain, but some estimates suggest that it may be
significant.

     Nitrogen may also enter the soil-water-biomass system via
the weathering of nitrogen-bearing rocks.  This appears to be
a very minor source, however.  The rate of input, estimated on
the basis of global weathering rates and the average nitrogen
contents of rocks, is only about 2 percent of the rate of input
from atmospheric sources  (8) .

     Assuming the atmosphere to be the important source of
nitrogen,  several reaction paths for the production of nitrate
can be visualized.  The most reasonable paths are shown in Figure
                                17

-------
 2.  These include  (a) nitrate from fertilizer,  (b) nitrate due to
 fixation by free-living bacteria, (c) nitrate from decayed plants,
 (d) nitrate from legumes, (e) nitrate from animal wastes, and  (f)
 nitrate from human wastes.  Not all possible paths are shown,
 e.g., rainout of atmospheric ammonium nitrate.

     An examination of Figure 2 shows that the overall reaction
 schemes can be very complex due to possible recycling of the
 nitrate.  This is only true at depths within the root zone of
 plants, however.  Once the nitrate is flushed beyond the root
 zone it should be unavailable for further recycling.  It should
 be noted that many of the reactions shown are caused or aided
 by bacteria.

     It can be seen in Figure 2 that the ammonia and nitrate
 from human and animal wastes are not assumed to participate in
 recycling.  This is because these wastes are concentrated at
 particular sites (e.g., barnyards or septic tanks) where plants
 either do not grow or are not numerous enough to consume signif-
 icant amounts of the nitrate.

 NITRATE SOURCE DETERMINATION

     The technique under investigation in this study relies
 on the dependence of the nitrogen  isotope ratio of a particular
 compound (specifically, nitrate) on the type of reaction or series
 of reactions which formed that compound.  A complete knowledge
 of all possible nitrate formation pathways, and of the nitrogen
 isotopic fractionations that occur along those pathways, would
 enable one to predict the isotopic composition of any particular
nitrate sample.   Such detailed information on the nitrogen iso-
 tope behavior of nitrate and its precursors does not exist,
however.
                               18

-------
       Industrial fixation
    V
                                              Legumes plus
                                              rhizodium bacteria
 Bacteria:
 Azobacter,
Clostridium
    V
Commercial
Fertilizer
    v/
                                                      Animals
                                                                        Animal
                                                                        Wastes
                                                                                         V
                                                                              Human
                                                                              Wastes
                                                            Oxidation
                                                           Oxidation
                                                         \K
                                                        X*
                                                         NO 3
Oxidation
                                              C             D           E

                              Figure 2.   Mechanisms of Nitrate Formation.

-------
     The  technique  for  identifying sources of nitrate in ground
water  is  based  on more  of a "black box" approach.   It is sufficient
to know (a)  that isotopic fractionations exist  for  various reac-
tions  (such  as  those presented  in Tables 2 and  3),  and  (b) that
nitrate may  be  formed by different reaction pathways  (such as
those  shown  in  Figure 2) that depend, at least  in part, on the
source of the nitrogen.  As an  example, nitrate formed  during
the manufacture of  fertilizer should go through a different
series of reactions than nitrate formed from animal wastes, and
should therefore exhibit a different nitrogen isotope ratio.

     Jones (1)  and  Kreitler  (2) tested this hypothesis  in Runnels
County, Texas.  They found that nitrates derived from septic
tanks  and animal wastes had a different nitrogen isotope ratio
from nitrates derived from "natural"  processes   (i.e.,  bacterial
fixation and plant uptake,  growth,  and decay).  This  was  demonstrated
by obtaining samples of soil  nitrate  from sites  where  the source
of the nitrate  could be reasonably predicted.   Nitrate  derived
from the  decomposition  of animal wastes  (including  human wastes)
was found to have a characteristic isotopic composition (615N) of
from +10  ppt to +22 ppt, while  nitrate derived  from natural
soil-bacteria-vegetation systems had an isotopic composition of
from +2 ppt  to  +8 ppt.

     It is possible that these  results are applicable only to
the particular  systems  studied  in Runnels County, Texas.  Environ-
mental factors, such as climate and soil type,  could affect the
isotopic  composition of a soil  nitrate sample by affecting the
types  of  bacteria present and the reaction pathways leading  to
nitrate formation.

     One  purpose  of this  study  is to  determine  whether  environ-
mental factors  influence  the isotopic  composition of  nitrates
 derived from a  particular waste source  category (e.g.,  septic
 tanks).  This is  accomplished by  analyzing soil and ground-water
                                20

-------
samples obtained from various sites across the country, represent-
ing a range of environmental conditions.

     The other major goal of this study is to determine the ex-
tent to which nitrates from various sources can be differentiated
on the basis of their isotopic composition.  The work of Kreitler
and Jones indicated that for one geographic area nitrate from
animal and human wastes can be distinguished from nitrate derived
from "natural" soils or fertilizers.   This study addresses the
question of whether or not nitrate from various sources can be
distinguished on the basis of nitrogen isotope ratio for the
country as a whole.

OTHER STABLE ISOTOPES AS TRACERS

     It may be possible to identify sources of ground-water
contamination using stable isotopes of elements other than
nitrogen.  In the case of nitrate (N03),  there is only one other
element (oxygen) to consider.  Other contaminants (e.g., phos-
phates, sulfates and sulfides, detergents, and other organic
compounds) contain elements for which isotopic studies may
prove useful, however.  These elements include hydrogen, oxygen,
carbon, sulfur, chlorine, and some trace metals.

     The usefulness of these elements as waste-source indicators
is difficult to assess.  It should be remembered that for stable
isotope identification of waste sources to be successful, it is
necessary that relatively unique isotope fractionations exist
for given waste types or waste sources.  This has not been
established for the elements mentioned above.

     A complete theoretical evaluation of the potential for
using stable isotopes of elements other than nitrogen is beyond
the scope of this study.  Rather, brief discussions are provided

                               21

-------
for several elements that are considered to be "possible candidates"
for this use.

Hydrogen

     There are two stable isotopes of hydrogen,  JH and 2H (2H
is commonly called deuterium, and is designated by the symbol
"D").  The nucleus of a 1E atom contains a single proton, while
the nucleus of a D atom contains a proton and a neutron.

     The relative mass difference between :H and D is larger
than that of any other pair of stable isotopes of other elements:
the mass of the JH nucleus is essentially half that of the D
nucleus.  Because of this, hydrogen exhibits the largest fraction-
ations of any stable isotopes.  Fractionations of several hundred
parts per thousand have been observed for various earth materials
(9).

     The average natural abundances of the stable hydrogen iso-
topes are as follows:

                        1H:  99.9844%
                         D:   0.0156 %.

     Hydrogen is an extremely abundant element in natural bio-
geochemical systems.  It occurs in water, clay minerals, dis-
solved inorganic compounds and organics.  Appreciable fractiona-
tions have been observed in isotope exchange equilibrium reactions,
and significant kinetic fractionations seem possible in biochem-
ical processes (e.g., photosynthesis and bacterial production of
H2 and CHO(9) .
Carbon

     Carbon has two stable isotopes:  12C and 13C.  Carbon-12
accounts for 98.89 percent of natural carbon.  Natural frac-
                               22

-------
 tionations  of  greater  than  10  percent have been  observed  (9).
 Extensive fractionations  take  place  in  the biosphere,  during
 photosynthesis and metabolism.   Environmental  factors  (e.g.,
 temperature and pH)  have  also  been observed  to influence  carbon
 isotope  fractionations (9).

      Carbon occurs extensively in the environment  as organic
 compounds, carbon dioxide (C02)gas,  and bicarbonate (HCOa) and
carbonate (C03) ions.  Kinetic fractionations are significant in
the system:   atmospheric C02 - dissolved HCO~ ions -  Calcium car-
bonate (CaC03)   (9).

 Oxygen

      Oxygen has three  stable  isotopes.  Their natural  abundances
 are  as follows:

                         160:   99.7637.
                         170:    0.0375%
                         180:    0.1995%

 In oxygen stable  isotope  studies the  180/160 ratio  is  usually
 measured, due  to  the higher abundance and greater mass difference
 (9).

      Oxygen plays a  critical  role in natural biogeochemical pro-
 cesses.  It is the most abundant element on  earth  and  occurs in
 gaseous, liquid,  and solid  compounds.   Natural variations in the
 180/160  ratio  of  up  to about  10  percent have been measured.

      Isotope equilibrium  excnange fractionations are significant
 in inorganic geochemical  systems.  The  180/160 ratios  of  mineral
 pairs are temperature  sensitive, and oxygen  isotope studies have
 been  used extensively  to  estimate the "temperature  of  formation"
 of various  mineral samples.   Chemical kinetic  fractionations of
                               23

-------
oxygen  isotopes occur during biochemical processes  (e.g., photo-
synthesis  and respiration)(9).

Sulfur

     There are four  stable  isotopes of  sulfur:  32S,  33S,  3"S,
and  36S.   32S and  3"*S are most abundant:

                          32S = 95.02%
                          3*S =  4.21%

These are  the isotopes  that are generally  of  interest  in  sulfur
isotope studies.

     Sulfur is present  in nearly  all biogeochemical  environments,
in a number of forms including sulfides, organic  substances,
elemental  sulfur,  and sulfates.   Natural variations  in 31|S/32S
ratios  on  the order  of  15 percent have  been measured.   Kinetic
effects and isotope  exchange reactions  both produce  isotope
ratio variations.  Kinetic  effects are  particularly  noticeable
in biochemical reactions,  such as the bacterial reduction of  sul-
fate to hydrogen  sulfide (H2S).   Isotope exchange  between sulfate
and sulfide and between  different sulfide species  is significant
in geochemical systems  (9).

Chlorine

     The natural abrndances of the two  stable chlorine isotopes
are:

                          35C1:   75.53%
                          37C1:   24.47%

The  chloride ion is  a major anion in natural  waters, and  is
present in relatively high  concentrations  in  waste materials
                               24

-------
(see, for example, the analysis of septic tank effluent in Table
B-l, Appendix B).   Theoretically one might expect to see small
variations in the isotopic composition of chlorine.   No signifi-
cant fractionations have been observed, however (9).

Trace Elements

     Many of the more environmentally significant trace elements
have more than one stable isotope:  chromium (4 isotopes),  iron
(4 isotopes), nickel (5 isotopes), copper (2 isotopes), zinc
(5 isotopes), selenium (6 isotopes), silver (2 isotopes), cadmium
(7  isotopes), tin  (10 isotopes), mercury  (7 isotopes), and lead
C3 isotopes).  Some of these may be useful in tracing pollution
sources.  Little,  if any, previous work has been done in this
area with these elements, however.

Summary

     From a theoretical standpoint, stable nitrogen isotopes ap-
pear to be most useful for identifying pollutant sources that
contribute nitrate to subsurface water-soil systems.  The re-
stricted sources of nitrogen and the better defined pathways
reduce the uncertainty inherent in this approach.  Especially
helpful in this regard is the fact that most of the nitrogen
in natural soils is ultimately derived from a single source
(the atmosphere) which has a uniform nitrogen isotopic composi-
tion.
                               25

-------
                            SECTION 5
                     EXPERIMENTAL PROCEDURES

     The procedures followed in selecting sampling sites and in
collecting, preserving, transporting, storing, and analyzing
samples are discussed below.

SELECTION OF SAMPLING SITES

     Samples of soil and/or ground water were obtained from nu-
merous sites across the United States (see Section 6).  The pro-
cess of selecting these sites was based primarily on a review of
published literature relating to nitrate contamination of ground
waters.  Over 300 references were assembled in the course of this
review.  To the extent possible, these references were subdivided
as to suspected, presumed, or known sources of the nitrate and as
to their geographic distribution within the source categories.
The objective of the geographic classification was not to select
sampling sites in all areas of the country, but to identify sites
that were representative of a wide range of environmental con-
ditions (e.g.,,  climate, soil type, latitude, elevation).  This
classification enabled numerous candidate sites to be identified.

     Final site selection was based on an assessment of how well
the candidate sites met the set of selection criteria described
below.   In some cases,  inquiries were directed toward cognizant
local personnel to better define the suitability of a site for
sampling.
                                26

-------
     In order for a candidate site to be selected for sampling,
the site must first have been characterized as having a single,
identified source of environmental nitrate.  Three sources of
nitrate were of primary interest:  domestic septic tanks, animal
feedlots, and municipal wastewater land application operations.
Particular emphasis was placed on finding sites where domestic
septic tanks were considered to be the known or probable source
of any environmental nitrate.

     Additional criteria were that the final choice of sites
should provide a wide geographic representation and that the
geologic and soil conditions at the sites be well-characterized.
Finally, the sites had to be accessible for sampling.

     Fourteen sites were chosen for intensive sampling.  Samples
of soil and ground water were collected from these sites by
Radian personnel according to the procedures discussed below.  An
additional five sites were selected, to be sampled by local per-
sonnel of the United States Geological Survey, various state
agencies, or universities.  These sites were selected primarily
to provide additional geographic coverage, and were sampled less
intensively than the others.

     Data on five other sites were obtained from Kreitler (per-
sonal communication, 2).  These data provided additional geo-
graphic coverage and enlarged the final data base, particularly
for the animal waste source category.  They also provided cover-
age of "natural" sources, i.e., sites where the nitrates were
derived from the natural growth and decay of plants.

     The various sites and the samples obtained from them are
discussed in Section 6.  More detailed descriptions of the 14
intensively sampled sites are given in Appendix B.
                                27

-------
SAMPLING PROCEDURES

Ground-Water Samples

     Ground-water samples were collected from existing wells in
the vicinity of the sites whenever wells were available for sam-
pling.  In a few cases, water samples were also obtained from
augered soil sampling holes.  At two sites, well points were
driven to collect ground-water samples.

     The procedure for collecting samples from wells was straight-
forward.  Samples were collected from points as near to the
wellhead  as possible.  The presence of any treatment or storage
systems (e.g., water softeners or pressure tanks) between
the  wellhead and the sampling point was noted, along with any
other information about the well that was  available (such as
well type, use, date installed, diameter, and depth).  Prior to
collecting a sample, the well was pumped for a period of at least
five minutes to assure the collection of a fresh water sample.

     Information on well construction is needed because the man-
ner of well completion can affect the observed 615N value of
nitrate in the well water.  For example, a well completed in
two distinct aquifers may yield water containing nitrate from
two distinct sources.

     In many cases,  however, details of well construction were
lacking.   All wells sampled were assumed to tap a single aquifer,
based on observations of depth, well diameter, use, and pump
size.

     Water samples from augered soil sampling holes were obtained
with a small peristaltic pump or by bailing with a small bucket.
The holes  were bailed out several times before a sample was col-
lected.   Water samples obtained with well points were collected
                              28

-------
with a  small peristaltic pump.  The pump was run  for several
minutes before a  sample was collected to flush the pump and hoses,
All water  samples were collected  into one-liter polyethylene
bottles.

Soil Samples

     Most  soil samples were collected using a 4-inch diameter
soil auger.  Some near-surface  samples were obtained using a
shovel.  The depth  intervals  from which the samples were col-
lected  were recorded.  Soil samples were placed in plastic bags
for shipping and  storage.

     At one site, a  set of 1.25-inch diameter cores was obtained.
These samples were  obtained by  driving a core-barrel fitted with
a  tapered  cutting bit.  The core  barrel was lined with 1.25-inch
thin-wall  PVC pipe.  When the maximum depth had been reached,
the PVC pipe, with  the core-sample inside, was extracted.  The
core was cut into convenient  lengths, and  the ends of the PVC
tube were  sealed.

Sampling Kits

     Sampling kits were sent  to local personnel at the five sites
not visited by Radian personnel (see Section 6).  These kits were
designed to supply  all of the materials required  for collecting
a  sample.  Each kit  consisted of  a styrofoam-lined shipping con-
tainer, water sample bottles, labels, data sheets, vials of sam-
ple preservative, plastic soil  sample bags, and an instruction
sheet.

Sample  Preservation, Transport, and Storage

     Water  samples were  preserved  by  adding 10 milliliters  (ml)  of
1  normal (N)  sulfuric acid  (t^SO^) to  each  liter  of  sample.  The
                              29

-------
samples were refrigerated,  frozen,  or placed on ice as soon after
collection as possible.   Deviations from this procedure are noted
in the descriptions of the sampling sites given in Appendix B.

     No preserving agents were applied  to  the  soil  samples.   These
 samples were  refrigerated and/or frozen as  soon after collection
 as possible.  Usually this was within an hour, and  in no  case
 were samples  left unrefrigerated for more  than eight  hours.
 Coolers and dry ice were taken to  the sites when  cold storage
 facilities  were unavailable.

     The  samples were shipped ,  by  either air or surface carrier,
packed in dry ice.    Upon receipt  in Austin some  frozen samples
 were found  to have thawed in transit, but no samples  had  reached
 ambient temperature.

     All samples were kept frozen after  they reached Austin.  Samples
were not thawed until immediately prior  to  analysis.  Frozen samples
were stored for up  to nine months before being  analyzed.

 SAMPLE PREPARATION AND ANALYSIS

     Isotope  analysis is most conveniently performed  on nitrogen
 gas (N2).   This means that all nitrate  samples must be reduced
 to nitrogen gas, and this conversion must be performed without
 introducing any isotope fractionation during the  sample prepara-
 tion.   Once the nitrate is converted to nitrogen  gas,  the  gas
must be purified to remove species such as oxygen which might
 damage the  filaments in the mass spectrometer.  The techniques
which were  developed for sample preparation were  as follows.

Soil Sample Leaching

     Prior  to analysis,  the soluble nitrogen species were  ex-
tracted from  the soil samples by leaching.   A one-kilogram
                               30

-------
aliquot of  soil was placed  in  a 4.2  liter polyethylene wide
mouth bottle.  One  liter  of deionized water was added and the
bottle was  placed on  an orbital shaker  table  for  four hours.

     After  four hours  of  shaking,  the sample  was  removed and
allowed to  settle for  15  to 20 minutes.  A 50 ml  aliquot of
the  supernatant liquid was analyzed for nitrate with a specific-
ion electrode.  Approximately 65 milligrams (mg) of silver sulfate
were added to minimize chloride interference during the  nitrate
concentration determinations.

     Those  samples  that yielded a  leachate nitrate  concentration
of less than  5 mg/£ were  not analyzed further.  A concentration
of at least 5 mg/Jl was found to be necessary  in order to obtain
a sufficiently large nitrogen  gas  sample for  isotopic analysis.

     Those  soil samples yielding a leachate nitrate  concentration
greater than  5 mg/£ were  filtered  using an 11-centimeter Buchner
funnel and a 1,000-ml vacuum flask attached to an  aspirator.  Di-
atomaceous earth was added as a filter aid.   The subsequent  sample
preparation procedures were identical for  soil sample leachates
and ground-water samples.

Sample Preparation

     The  procedures followed in preparing samples  for nitrogen
isotopic  analysis were essentially the  same as  those described
by Krietler  (2).  Since only the isotopic ratio of  the nitrate
was  of interest, other species of  nitrogen had  to be removed.
These included ammonia and  organic nitrogen.  Removal of these
species and extraction of nitrate  was accomplished  using a two-
stage distillation process.  The apparatus used for  this process
is illustrated in Figure  3.
                              31

-------
      DISTILLATION
            FLASK
        HOT PLATE
         SUPPORT
            PLATE
MAGNETIC
 STIRRER
                                      BLOCK
                                     SUPPORT
                                                DISTILLATION
                                                COLUMN
      Figure 3.  Distillation Apparatus Used for Removal
                 of Interfering Nitrogen Species.
                             32

-------
     The pH of the sample (soil leachate or ground water) was
first raised to between 9.5 and 10.0 by addition of powdered
magnesium oxide (MgO).   The solution was agitated to suspend
the MgO (see Appendix A), transferred to a two-port round bottom
distillation flask, and distilled until 50 ml of distillate
were collected.  This caused the conversion of NH4 to NH3 and
the breakdown of any labile organic compounds to ammonia.  At
this stage, interfering forms of nitrogen were removed from
the sample, having been collected in the distillate.

     The hot plate was then replaced with the magnetic stirrer,
and the exit port of the water-jacketed distillation column was
immersed in a beaker containing 50 ml of 0.1 normal hydrochloric
acid (HC1) (see Appendix A).  Five grams of finely ground De-
varda's Alloy  (see Appendix A) were added through the side port
of the flask.  The flask was immediately restoppered and the
solution magnetically stirred for 10 minutes without heating.
This step reduced the nitrate in the sample to ammonia.  Finally,
the solution was redistilled until 75 ml of distillate were col-
lected in the 0.1 HC1 solution.  Relatively concentrated HC1
was used to insure that all ammonia was absorbed, producing an
acidic ammonium chloride solution.  The volume of this solution
was reduced to 5 to 10 ml by heating on a hot plate at 70°C to
80°C.

Nitrogen Gas Generation

     A special nitrogen gas generation and purification apparatus
was required to further prepare a sample for isotopic analysis.
A schematic of the apparatus is shown in Figure 4.  The appara-
tus consists of the following:  a nitrogen gas generation tube,
a freeze-thaw manifold, two furnaces (containing copper and
copper oxide, respectively), a Toeppler pump, a gas collection
flask, two mechanical vacuum pumps, and a mercury diffusion pump.
                                33

-------
             VACUUM
          THERMOCOUPLE
                                        FREEZE-THAW MANIFOLD
                                                           GAS
                                                        COLLECTION
                                                           FLASK
 MERCURY
DIFFUSION
  PUMP
                                                                                     NITROGEN
                                                                                        GAS
                                                                                     GENERATION
                                                                                       TUBE
                                                        Figure  4.   Nitrogen Gas  Generation
                                                                    and  Purification Apparatus
                  MECHANICAL VACUUM PUMPS

-------
     After the ammonium chloride solution was concentrated to
an appropriate volume  (5 to 10 ml), as described above, it was
transferred to the large volume side of a nitrogen gas generating
tube (see Figure 5).  Approximately 5 ml of sodium hypobromite
(NaOBr)  solution (see Appendix A)  were added to the small  side-
arm.   This assembly was attached to the freeze-thaw manifold of
the vacuum preparation apparatus.
                    NH/.CL
NAOBR
           Figure 5.   Nitrogen Gas Generation Tube
     The  two  separate solutions were frozen by immersing the
generating  tube  in an equilibrated  dry  ice-acetone mixture.
Once frozen,  the stopcock on  the reaction tube was opened to
the vacuum  system allowing the atmosphere above the frozen
solutions to  be  withdrawn.  The stopcock was reclosed, and the
                              35

-------
solutions were allowed to thaw.  This sequence was carried out
three times to completely degas the solutions and thereby elim-
inate atmospheric nitrogen contamination.

     The nitrogen gas generation tube was transferred to the
sample injection port below stopcock D (see Figure 4).  With the
gas generation tube isolated from the rest of the system (i.e.,
stopcock D closed) , the side-arm was rotated 180° to drain the
sodium hypobromite solution into the ammonium chloride solution.
The resulting reaction liberates nitrogen gas:

                 2NH3 + 3NaOBr - SNaBr + 3H20 + N2

N2 Gas Purification

     When the N2 gas generation reaction was complete, stopcocks
D and F were opened and the Toeppler pump was activated (stop-
cocks E, G, and J closed).  The nitrogen gas was pumped from
the generation tube, through the furnaces, and into the tubing
segment bounded by stopcocks G, I, and J.  When a pressure of
3 to 5 mm Hg was accumulated (as indicated by vacuum thermocouple
#2) , stopcock F was closed and stopcock G was opened.  A closed
loop was thus created through the furnaces and the Toeppler pump.

     The nitrogen gas was circulated through the hot copper
furnace (T = 400°C), the hot copper oxide furnace (T = 800°C),
and a liquid nitrogen cold trap for one hour.  This eliminated
any oxygen, carbon monoxide, and gaseous oxides of nitrogen that
may have been present.

     To collect the purified nitrogen sample the three-way stop-
cock I was turned, which redirected the purified sample into the
gas collection flask.  The stopcock on the collection flask as
well as the stopcock on the collection port (0) were then closed
and the gas collection flask removed.
                              36

-------
Isotopic Analysis
     Purified nitrogen gas samples were analyzed for their nitro-
gen isotope compositions by personnel of the Center for Research
in Water Resources of the University of Texas at Austin.  The
isotope ratio was measured by comparing mass 28 (14N14N) to mass
29 (15N14N) in a double-collector, gas-source, mass spectrometer,
using atmospheric nitrogen as a standard.

ANALYSIS OF STANDARDS

     The analytical technique used in this study was perfected
by Kreitler (2),  who found that the experimental error in the
technique was about ±1 ppt.  Careful adherence to the details
of the procedure is necessary to achieve this level of precision,
however.  Deviations from the procedure can introduce isotopic
fractionations in the sample.

     An additional check on the reliability of the analytical
procedures was made in this study by running standards of known
isotopic composition.  A set of potassium nitrate (KN03) standards
was supplied by Kreitler.  Four of these were run through a dis-
tillation, preparation, and purification system.  The resulting
nitrogen gas samples were returned to Kreitler for isotopic anal-
ysis.  Kreitler compared the 615N values of the investigator's
standards with the results of a set of analyses that were per-
formed on the same material in Kreitler's laboratory.  Kreitler
concluded that the 615N values were in agreement, within the
limits of the mass spectrometer, and that the investigator's
preparation techniques were acceptable.

     Additional standards were periodically run during the course
of the program.  These standards were prepared from a stock solu-
tion of KNOs provided by Kreitler.  A sample of atmospheric
nitrogen was also run.
                               37

-------
     All of these standards, except one, yielded 615N values
within ±0.5 ppt of the known 615N values.   The one exception was
a KN03 standard (615N = -0.7) which yielded an apparent 615N value
of -10.7 ppt.   Calculations involving the mass of nitrogen in the
standard, the mass of Devarda's  Alloy in the nitrate-to-ammonium
reaction, and the mass of NaOBr  in the ammonium-to-nitrogen gas
reaction indicated that incomplete conversion reactions were not
the cause of this discrepancy.   This data point is viewed as an
analytical "outlier," since none of the known opportunities for
isotopic fractionation in the analytical procedures could be
blamed for the discrepancy.

     As an additional check on the procedure, the measured
<51 5N values of the samples prepared by Radian were examined as
a function of the time during the program in which the samples
were analyzed.  Most of the analytical difficulties experienced
(see Section 7) occurred in the  early stages of the program.
The examination of 615N versus time was to determine whether
there was any consistent and/or  progressive error in the proce-
dures as the program progressed.   No relationship was found
between 6l5N and the time a sample was analyzed.
                               38

-------
                           SECTION 6
                        SAMPLING SITES

     The intensive field sampling program carried out by Radian
personnel resulted in the acquisition of 324 discrete soil and
water samples, plus a total of about 45 feet of 1%-inch core.
Two hundred forty-three of the discrete samples were soils;
eighty-one were water samples.

     Fourteen sites were visited in the intensive program:

     !•  Texas A&M University  (TAMU) at College Station, Texas;

     2-  Rockdale , Texas ;

     3*  Tampa, Florida  (2 sites);

     4-  Indian Hills, Colorado;

     5-  Pueblo, Colorado;

     6-  Houghton Lake, Michigan  (2 sites);

     7*  Tallahassee, Florida;

     8-  Abilene, Texas;

     9-  Michigan State University (MSU) at East Lansing, Michigan;

    10-  San Angelo, Texas;

                               39

-------
    !!•  Greeley,  Colorado;  and

    12*  Dimmitt,  Texas

These sites are associated with three nitrate source categories:
domestic septic tank systems (7 sites),  municipal wastewater
land application systems (5 sites) ,  and cattle feedlots (2 sites)
A breakdown of the samples  collected from each of these sites
is given in Table 4.

     Sampling kits were sent to an additional 5 sites, all of
which are associated with septic tank nitrate sources.  Provi-
sion was made in each kit for the collection of  3  samples.

     In addition to these 19 sites,  data were obtained from Dr.
C.W. Kreitler, of the Texas Bureau of Economic Geology, for
samples that he obtained from 5 other sites.   The Queens County,
New York, and Grand Cayman Island sites are believed to represent
septic tank sources of nitrate.  The Macon County,  Missouri, and
Lockhart-Taylor,  Texas,sites are believed to represent animal-
waste nitrate.  In Runnels  County, Texas, data for septic tank,
animal waste  (barnyards), and natural nitrate sources were ob-
tained.   The geographic locations of the 24 sites are shown in
Figure 6.  A key identifying each site shown in the figure is
given in Table 5.

     Environmental factors  were assessed for each of the 24 sites
The environmental variables considered were related to climate
and soil type:

        mean annual precipitation,

        mean annual air temperature,
                              40

-------
                TABLE 4.   SUMMARY OF SAMPLES COLLECTED IN
                          RADIAN'S INTENSIVE SAMPLING PROGRAM.
Location
TAMU
Rockdale
Tampa
Indian Hills
Houghton Lake
(A)
Houghton Lake
(B)
Pueblo
Tampa
Tallahassee
Abilene
San Angelo
MSU
Greeley
Dimmitt

N03 Source
Septic Tank
Septic Tank
Septic Tank
Septic Tank
Septic Tank
Septic Tank
Septic Tank
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Feedlot
Feedlot

No. of Soil
Samples
8
23
64
5
8
11
10
21*
19
16
15
16
13
14_
243
Total Combined
Depth
Interval
2.0-4.5 ft.
1.0-5.0 ft.
0.0-11.5 ft.
0.5-5.1 ft.
0.0-4.9 ft.
0.0-5.7 ft.
0.1-10.9 ft
0.0-20.0 ft.
0.0-9.5 ft.
0.0-7.2 ft.
0.5-6.4 ft.
0.0-6.2 ft.
0.0-6.0 ft.
0.0-10.3 ft.

No. of Water
Samples
12
0
8
6
0
0
1
29
6
6
4
7
1
1
81
*Plus four 1 1/4-inch cores from the surface to a maximum depth of 20 feet.
                                    41

-------
Figure 6.   Locations of sampling
           sites (see key in Table 5)

-------
                         TABLE 5.   KEY TO FIGURE 6.
Map
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Location
Texas A&M University
Rockdale, Texas
Tampa, Florida
Indian Hills, Colorado
Houghton Lake, Michigan, Site A
Houghton Lake, Michigan, Site B
Pueblo, Colorado
Tampa, Florida
Tallahassee, Florida
Abilene, Texas
San Angelo, Texas
Michigan State University
Greeley, Colorado
Dimmitt, Texas
Wahkon, Minnesota
College Station, Texas
Jefferson County, West Virginia
Joplin Area, Missouri
Portland Area, Oregon
Lockhart-Taylor Area, Texas
Grand Cayman Island
Macon County, Missouri
Queens County, New York
Runnels County, Texas
N03 Source
Septic Tank
Septic Tank
Septic Tank
Septic Tank
Septic Tank
Septic Tank
Septic Tank
Wastewater
Wastewater
Wastewater
Wastewater
Wastewater
Feedlot
Feedlot
Septic Tank
Septic Tank
Septic Tank
Septic Tank
Septic Tank
Barnyard
Septic Tank
Barnyard
Septic Tank
Septic Tank,
Barnyard , and
"natural"
Sampled
By*
A
A
A
A
A
A
A
A
A
A
A
A
A
A
B
B
B
B
B
C
C
D
D
D
^Samples by:   A = Radian Intensive Sampling Program;  B =  Radian  Kit  Sampling
              Program;  C = Kreitler (unpublished data); D =  Kreitler (2).
                                     43

-------
        Koppen Climate Classification,

        regional soil type,

        elevation,  and

        latitude.

The values of these parameters  are listed in Table 6 for each
of the sites.  The  Koppen Climate Classification system and Soil
Classification system are described in  Appendices E and F,
respectively.
                               44

-------
                                  TABLE  6.    SITE ENVIRONMENTAL  CHARACTERISTICS
Site Location
Texas A&M University
Rockdale, Texas
Tampa, Florida (2 sites)
Indian Hills, Colorado
Houghton Lake, Michigan, Site A
Houghton Lake, Michigan, Site B
Pueblo, Colorado
Tallahassee, Florida
Abilene, Texas
San Angelo, Texas
Michigan State University
Greeley, Colorado
Dimnitt, Texas
Wahkon, Minnesota
College Station, Texas
Jefferson Co. , West Virginia
Joplin Area, Missouri
Portland Area, Oregon
Lockhart-Taylor Area, Texas
Grand Cayman Island
Macon Co., Missouri
Queens Co. , New York
Runnels Co., Texas
Sampled
ByL
A
A
A
A
A
A
A
A
A
A
A
A
A
B
B
B
B
B
C
C
D
D
D
Average
Annual
Temperature
68
67
72
42
43
43
51
58
64
65
47
46
56
39
68
52
57
54
68
82 1!
52
50
66
Average
Annual Koppen
Precipitation Climate
(Inches) Classification*
36
34
52
24
30
30
12
56
24
20
31
12
18
28
36
40
40
48
32
80//
36
44
24
Caf
Caf
Caf
H
Dbf
Dbf
BSk
Caf
BSk
BSh
Daf
BSk
BSk
Dbf
Caf
Daf
Caf
Cb
Caf
Au
Daf
Daf
BSh
Elevation
(Feet)
200
490
50
7,200
1,150
1,270
3.500//
250)?
1,700
1,800
870
4,630
3,900
1,500*
200
500*
1,500*
250*
500
50*
750l?
1000
1,600
Latitude
(Degrees)
30.50
30.75
28.00
39.75
44.25
44.25
38.25
30.25
32.50
31.50
42.75
40.50
34.50
46.00
30.50
39.50
37.00
45.50
30.25
19.00
39.75
40.75
31.75
Soil
Type4
5
5
3
]0
1
1
6
3
6
6
2
6
6
1
5
2
3
10
5
4
5
2
6
 Sampled by:  A = Radian Intensive Sampling Program; B = Radian Kit Sampling Program;
            C = Kreltler  (unpublished data); D = Kreitler  (2).
*See Appendix E for definitions of Climate Classes.
+See Appendix F for definitions of Soil Classes.
^Estimated value.

-------
                           SECTION 7
                     RESULTS AND DISCUSSION

STATISTICAL ANALYSES OF THE DATA

     The analytical data used in this study are tabulated in
Appendix D.  This data base includes the results of the isotopic
analyses performed under this study, and also includes both pub-
lished  (2) and unpublished data provided by Dr. C.W.  Kreitler
of the Texas Bureau of Economic Geology.  The distribution of
the data by nitrate source category and site is presented in
Table 7,  It should be noted that not all of the samples collected
in Radian's sampling program were analyzed, and that some sites
are not represented in the data base.  The reasons for this are
discussed in the following subsection on Analytical Problems.

     This data base was analyzed using several statistical
techniques.  These included an analysis of means and variances
of 615N for the various waste source categories, preparation of
histograms, and correlation and regression analysis.

     Correlation, multiple correlation, and partial correlation
analyses were performed for the entire data base and for each
separate nitrate source category.  The analysis of the entire
data base examines the "behavior" of 
-------
                 TABLE 7.   SUMMARY OF  SAMPLES  INCLUDED IN THE  FINAL
                             DATA  BASE,  BY SOURCE CATEGORY  AND SITE.
NOa Source Category
Septic Tanks











Animal Wastes (Feedlots
and Barnyards)




Municipal Wastewater
Irrigation


Natural Soil-Plant Systems

Site
Houghton Lake, Michigan, Site A
Houghton Lake, Michigan, Site B
Puehlo, Colorado
Tampa, Florida
Indian Hills, Colorado
Rockdale, Texas
Texas A&M University
Grand Cayman Island
Macon County, Missouri'
Queens County, New York
Runnels County, Texas
Subtotal
DJmmltt, Texas
Greeley, Colorado
Lockhart-Taylor Area, Texas
Macon County, Missouri
Runnels County, Texas
Subtotal
Abilene, Texas
San Angelo, Texas
Michigan State University
Subtotal
Runnels County, Texas
TOTAL, ALL SOURCE CATEGORIES AND SITES
Number of
Analyses
6
4
6
13
5
4
3
4
1
2
4
52
8
7
11
10
16
52
2
5
_3
10
J*
178
Analyzed
By*
A
A
A
A
A
A
A
B
C
C
C

A
A
B
C
C

A
A
A

C

*Analyzed by: A = Radian; B = Kreitler, unpublished data; C = Kreitler (2).

-------
separate correlation analyses examine the extent to which this
observed variation in 615N for specific nitrate sources can be
explained in terms of the various environmental factors.
Descriptive Statistics for j>
15N
     The means and standard deviations of the isotope ratios
of the samples were computed for each of the nitrate source
categories.  These statistics are presented in Table 8.

     The means and variances of 615N for the different source
groups were compared using a one-way analysis of variance  (F-
test)  and  other  statistical tests  (t-tests) .  This analysis in-
dicated that the mean  615N values  for the septic tank  (+10.9),
animal waste (+12.4),  and municipal wastewater irrigation  (+9.5)
source categories do not differ significantly (i.e., these sources
of nitrate cannot be distinguished on the basis of 615N values).
Moreover,  the mean 615N values for the irrigation source  (+9.5)
and  "natural" nitrate  sources  (+7.3) do not differ significantly.

       TABLE 8.  MEANS AND STANDARD DEVIATIONS OF 615N
                 FOR DIFFERENT NITRATE SOURCE CATEGORIES
                            Mean     Standard      Number of
       N03 Source           (PPt)    Deviation    Data Points
Septic Tanks                +10.9       9.8           52
Animal Wastes               +12.4       9.4           52
(Feedlots and Barnyards)
Municipal Waste              +9.5       7.2            10
Water Irrigation
Other*                       +7.3       5.4           64

^"Natural" soil systems:  primarily unfertilized cultivated
 fields.
                                48

-------
     The analysis indicated, however, that the mean 6l 5N values
of both the septic tank (+10.9) and the animal waste (+12.4)
source categories differ from the mean 615N value for "natural"
soil nitrates (+7.3) at the 0.05 level of significance.   (The
"level of significance" is discussed below in the section on
correlation analysis.)  This supports Kreitler's conclusion (2)
that nitrates derived from animal wastes (including human wastes)
can be distinguished from natural soil nitrates on the basis of
their nitrogen-isotope ratios.

     The distribution of 6l 5N values for the various N0$ source
categories are plotted as histograms in Figure 7.  These histo-
grams (and the standard deviations given in Table 8) show that
the isotopic composition for each source category varies over a
wide range of values.  These ranges overlap considerably (in
contrast to Kreitler's (2) findings).

     The variation in del within the source categories was ex-
amined using correlation analysis.  This analysis was carried
out to determine whether environmental factors associated with
the sampling locations were the cause of the variations observed.

Correlation Analysis

Definition of Variables--
     The environmental variables associated with the sampling
sites were described in Section 6 (see Table 6).   They cover
such environmental factors as latitude, climate class, soil
type, and elevation.  Some of the variables are "continuous."
For example, the mean annual precipitation at a site may logi-
cally assume any real number value, even though the actual
values used were integers ranging from 12 to 80.   Other variables
are "binary."  These variables are assigned a value of either
zero or one.  A value of "1" means that a sample is character-
ized by that variable.  For example, climate zone Caf (see
                               49

-------
  0.3- .
  0.2- •
 u. 0. 1 • •
         SEPTIC TANK
  0.0-
          nn
    -20      -10
                            10      20       30      40
                      Del value (ppt)
                                                                  0.3- -
                                                                  0. 1- •
                                                                 0.0.
                                                                         IRRIGATION
                                                                   -20      -10
                                                                                     Del value (ppt)
                                                                                           10       20       30      40
  0.3- •
  0.2- '
i o.i- •
  0.0
                                                                  0.3--
                                                                  0.2- -
                                                                i o.i--
                                                                  o.o-
                                                                         OTHER
                                                                    -20      -10
i   n   p
                                                                                            10       20       30      40
                      Del value (ppt)
                                                                                      Del value  (ppt)
               Figure  7.   Histograms  of Del  Values  by  Source  Category.

-------
Appendix E) is designated by the variable name "C CAF."  Thus, a
sample acquired from a site located in climate zone Caf will have
a value of C CAF = 1, while a sample acquired from a site located
in climate zone Aw (see Appendix E) will have a value of C CAF = 0
The definitions of the variables used in the correlation analyses
are given in Table 9.  These definitions are provided because
coded variable names (e.g., C CAF) are used in the following
discussion of the results of the correlation analyses.

Meaning of the Correlation Coefficient--
     The correlation coefficient is a very useful measure of the
extent to which two variables are related.  The correlation
coefficient r is defined as follows:
                 N
            r =
z
i = 1
(X± - X)
(Y. - Y)

N
r
i = 1
av\ 2
— A )
1
N
Z (Y. - Y)2
j = 1
        where
             N =  number  of data points,
             X =  average of the X values,  and
             Y =  average of the Y values.
                                51

-------
      TABLE 9.   DEFINITIONS OF  VARIABLES  USED  IN THE
	CORRELATION ANALYSES
A.  Variables with Continuous  Values
                         1 5
        DEL             6 N value,  in parts per thousand.
        PPM             N03  concentration, in parts per million (mg/&).
        D*SOIL           Depth  from which a particular  soil sample was ob-
                          tained, in meters.
        TEMP            Mean annual  air temperature at a site, in de-
                          grees F.
        PRECIP           Mean annual  precipitation at a site, in inches.
        ELEV            Elevation of a site, in feet above mean sea
                          level.
        LAX             Latitude of  a site, in decimal degrees.
        SEQNCE           Sequence number of samples analyzed in this
                          study.
B.  Variables with Zero-One Values'5'
    1.  Miscellaneous  Variables
        RADSAM          The sample was prepared by Radian under this study.
        SOIL            The sample was a soil sample.
    2.  Nitrate Source Variablest
        SEPTIC          The sample was from a septic tank site.
        FEEDL           The sample was from an animal waste (feedlot or
                         barnyard) site.
        IRRIG           The sample was from a municipal wastewater irri-
                         gation site.
    3.  Climate VariablesT
        C AW            The sample was from a site in climate zone Aw.
        C BSK           The sample was from a site in climate zone BSk.
        C CAF           The sample was from a site in climate zone Caf.
        C DAF           The sample was from a site in climate zone Daf.
        C DBF           The sample was from a site in climate zone Dbf.
        C H             The sample was from a site in climate zone H.
                                                       (Continued)

                                   52

-------
           TABLE 9.   DEFINITIONS OF VARIABLES USED IN THE
                       CORRELATION ANALYSES  (Continued)
    4.   Soil Type Variables§
        SOIL 1          The sample was  from a  site with  type  1 soils.
        SOIL 2          The sample was  from a  site with  type  2 soils.
        SOIL 3          The sample was  from a  site with  type  3 soils.
        SOIL 4          The sample was  from a  site with  type  4 soils.
        SOIL 5          The sample was  from a  site with  type  5 soils.
        SOIL 6          The sample was  from a  site with  type  6 soils.
        SOIL 10         The sample was  from a  site with  type  10 soils.
    5.   Geographic Variables
        HLMA            The sample was  from Houghton Lake, Michigan, site A.
        HLMB            The sample was  from Houghton Lake, Michigan, site B.
        PU              The sample was  from Pueblo, Colorado.
        TAM             The sample was  from Tampa, Florida.
        IH              The sample was  from Indian Hills, Colorado.
        RT              The sample was  from Rockdale, Texas.
        TAMU            The sample was  from Texas A&M University.
        ABL             The sample was  from Abilene, Texas.
        SNGLO           The sample was  from San  Angelo,  Texas.
        MSU             The sample was  from Michigan State University.
        DI              The sample was  from Dimmitt, Texas.
        GR              The sample was  from Greeley, Colorado.
        LOCTAY          The sample was  from the  Lockhart-Taylor area, Texas.
        GNDCMN          The sample was  from Grand Cayman Island in the
                          Caribbean.
        MACON           The sample was  from Macon County, Missouri.
        QUEENS          The sample was  from Queens County, New York.
        RUNLS           The sample was  from Runnels County, Texas.

*A value of "1" indicates that sample is characterized by a particular
 variable; a value of "0" indicates that it is not.
tSamples of "natural" soil nitrate were indicated by a value of "0" for each
 of the three nitrate source  variables.
tClimate classifications are  defined  in Appendix E.
§Soil classifications are defined  in  Appendix F.
                                     53

-------
     The correlation coefficient is a number between -1 and +1,
and r2 can be interpreted as the proportion of the variation in
Y that can be "explained" or "predicted" in terms of X.  For ex-
ample, the correlation coefficient between DEL (Y) and ELEV (X)
is (from Table 11)

     r = .31.

Since

     r2 = (-.31)2 = .096,

it follows that 9.6 percent of the total variation in DEL can
be. explained in terms of ELEV.

     Figure 8 illustrates the meaning of a non-negative corre-
lation coefficient.  A negative correlation between two variables
occurs if one increases as the other decreases.

     The multiple correlation coefficient is a measure of the
strength of the composite relationship between a single variable
Y (e.g., DEL) and a set of predictor variables.   The square of
the multiple correlation coefficient is the proportion of the
total variation in Y that can be explained in terms of the set
of predictor variables.   For example, consider the multiple
correlation analysis of the data for the septic tank source
category.   The correlation coefficient between DEL and the set
of predictor variables (D*SOIL, ELEV, and SOIL)  is (from Table 17)

     r = .56.

Since

     r2 =  (.56)2 = .314,

                                54

-------
                      y-


                      6-


                      4-


                      2-
   x and y are
perfectly related
                    -2-
                    • 4-
                    -6-
                    -8-1
r = 0
                        -2
                             -1
No relationship exists
  between x and y
        y ~i
        6-
             =  7
       -6-
                x and y are
            imperfectly related
       -8-1  •
 Figure 8.   Illustration of the
              Meaning of  Correlation
              Coefficient,  r.
                    55

-------
it follows that 3.14 percent of the variation in DEL can be
explained in terms of the composite influence of D*SOIL, ELEV,
and SOIL.

     The partial correlation coefficient is a measure of the
strength of the relationship between two variables given that
one or more other variables are fixed.

Criterion for Choosing Variables Which Are Correlated--
     If two variables had no true relationship,  their correla-
tion computed from a finite data sample would still almost
certainly be non-zero.  The question,  then, is how large should
the correlation coefficient be to indicate a true or repeatable
relationship between the two variables?

     There is a small chance (one in ten) that an r value as
large as  .123 in magnitude would be obtained if the two variables
being considered were unrelated.  (The threshold correlation
.123 was computed as a function of the sample size, 178).  Thus,
if two variables have a correlation coefficient larger than
.123 or less than -.123,  there is a high probability that a
true relationship exists between them.  In this case, the cor-
relation  is said to be "statistically significant."

     The probability  (.10 in the example above) of erroneously
concluding that a true correlation exists is called the level
of significance, or "confidence" level.  Values of the sample
correlation coefficient (r) required to achieve statistical
significance for different levels of significance and sample
sizes used in this study are given in Table 10.
                               56

-------
     Correlation, multiple correlation, and partial correla-
tion analyses were performed for the entire data base and for
each separate nitrate source category.   The analysis of the
entire data base examines the "behavior" of DEL (and PPM) as a
function of all other independent variables, including SEPTIC,
FEEDL, and IRRIG (see Table 9).
          TABLE 10.   VALUE REQUIRED FOR A CORRELATION
                     COEFFICIENT (r) TO BE SIGNIFICANT,
                     AS A FUNCTION OF SIGNIFICANCE LEVEL
                     AND SAMPLE SIZE.
Sample Size
(N)
10
52
64
178
Value of r at
.10
.539
.231
.207
.123
Significance
.05
.619
.273
.246
.146
Level
.01
.746
.354
.320
.191
     Separate correlation analyses were performed within each
of the four nitrate source categories.  The values of 615N (DEL)
were shown above to vary over fairly wide ranges (see Table 8
and the histograms in Figure 7).   The separate analyses examine
the extent to which this observed variation in DEL for specific
nitrate sources can be explained in terms of the various environ-
mental factors.

Correlations in the Entire Data Base

     The most significant results of the correlation analysis
of the entire data base are presented in Tables 11 and 12.  Ad-
ditional details of the statistical analyses are presented in
Appendix G.
                               57

-------
TABLE 11.  VARIABLES WHICH HAVE STATISTICALLY
           SIGNIFICANT CORRELATIONS WITH DEL

TEMP
PRECIP
ELEV
FEEDL
C AW
C BSK
C DAF
C H
SOIL 4
SOIL 5
SOIL 6
SOIL 10
HLMB
IH
DI
LOCTAY
GNDCMN
MACON
RUNLS
Correlation
with DEL
.13
.29
-.31
.19
.15
-.13
.18
-.18
.15
.27
-.26
-.18
-.13
-.18
-.19
.22
.15
.15
-.15
Level of
Significance
.10
.01
.01
.05
.05
.10
.05
.05
.05
.01
.01
.05
.10
.05
.05
.01
.05
.05
.05
TABLE 12.   CORRELATIONS INVOLVING VARIABLES
           MOST CLOSELY RELATED TO DEL

DEL
PRECIP
ELEV
SOIL 5
SOIL 6
LOCTAY
DEL
1.
.29
-.39
.27
-.26
.22
PRECIP
.29
1.
-.61
.21
-.69
.08
ELEV
-.31
-.61
1.
-.35
.35
-.21
SOIL 5
.27
.21
-.35
1.
-.58
.58
SOIL 6
-.26
-.69
.35
-.58
1.
-.33
LOCTAY
.22
.08
-.21
.58
-.33
1.
                    58

-------
     Table 11 lists the significant correlations with DEL and
the levels of significance of those correlations.  None of the
other variables listed in Table 9 are significantly correlated
with DEL in the entire data base.  Table 12 presents the complete
correlation matrix for DEL and the five variables which corre-
late most highly (at the .01 level) with DEL.  Table 12 indicates
the extent to which these independent variables are correlated
with each other.  Note especially the relatively high correla-
tion between PRECIP and ELEV (-.61); these are the two variables
with the highest correlations with DEL.  These correlations
suggest that, for the sites and samples considered in this study,
615N values tend to be higher in areas of high precipitation
and/or low elevation.

     It is impossible to determine, however, whether a true
relationship exists between DEL and PRECIP, between DEL and
ELEV, or between DEL and the combination (PRECIP + ELEV),  solely
on the basis of the correlations presented in Table 12.  The
fact that PRECIP and ELEV are correlated with each other may
be causing the observed correlation between DEL and PRECIP or
between DEL and ELEV.

     For example, it is possible that a true relationship exists
between DEL and PRECIP (i.e., that the mean annual precipitation
influences the biogeochemical environment of soil/ground-water
systems, and thereby influences the nitrogen isotope ratio).
This would show up as a significant correlation.  The relation-
ship between DEL and ELEV,  however, may be caused by the cor-
relation of PRECIP and ELEV.  This correlation may be real,
or may be the result of a sampling bias (e.g.,  no samples were
obtained from low elevation deserts).  In either case,  ELEV
could be linked to DEL through PRECIP.  The fact that sites
with high precipitation are at relatively low elevations (in
this study) would yield a correlation between DEL and ELEV,

                               59

-------
even if the elevation, in and of itself, has no influence on
615N.  Additional research will be required to clarify the
influence of these environmental variables on 615N in nitrates.

     The most significant correlation in Table 11 is the posi-
tive (although small) correlation between FEEDL and DEL.  The
DEL value is higher in the feedlot (and barnyard) data than
in the data for septic tank, irrigation, and natural ("other")
nitrate sources as a whole.  This difference is primarily due
to the difference between feedlot DEL values and natural ("other")
DEL values (see the earlier discussion of mean 615N values for
the various source categories).

     The correlation between FEEDL and DEL cannot be "explained"
by other environmental factors.   An analysis of partial correla-
tions  (see Appendix G) indicates that the correlation between
DEL and FEEDL changes very little as the influence of other
variables (including ELEV) is removed.  Thus,  a true relation
between animal waste nitrate sources (feedlots and barnyards)
and higher DEL values appears to exist.

     The correlations between nitrate concentration (PPM) and
various other variables were also examined.   The variables
which have significant correlations with PPM are listed in
Table 13.  Six variables are correlated with PPM at the .01
level:   RUNLS (.36), RADSAM (-.36), FEEDL (.29), C CAF (.25),
SEPTIC (-.25), and SOIL 6 (.23).

     Thus,  for the data examined in this study, higher nitrate
concentrations were found in Runnels County, Texas; in feedlots
and barnyards; in climate zone CAF; and in areas with type 6
soils (Runnels County soils are classified as type 6).

     Samples collected for this study (RADSAM = 1) generally
had lower nitrate concentrations than samples analyzed previously

                               60

-------
TABLE 13.  VARIABLES WHICH HAVE STATISTICALLY
           SIGNIFICANT CORRELATIONS WITH PPM
Variable
TEMP
PRECIP
LAT
RADSAM
SOIL
SEPTIC
FEEDL
IRRIG
C BSK
C CAF
SOIL 3
SOIL 6
TAM
RUNLS
Correlation
with PPM
.17
-.14
-.13
-.36
-.16
-.25
.29
-.13
-.15
.25
-.14
.23
.14
.36
Level of
Significance
.05
.10
.10
.01
.05
.01
.01
.10
.05
.01
.10
.01
.10
.01

-------
by Kreitler (2) .   This situation had been observed prior to the
statistical analyses:  few sites had samples with nitrate con-
centrations as high as those found in Runnels County, Texas,
by Jones (1) and Kreitler (2).

     Finally,  samples obtained from around septic tank laterals
tended to have relatively low nitrate concentrations (compared
primarily to samples from feedlots and barnyards).

Correlations Within Source Categories

     Correlation analyses were also performed to determine
whether any of the observed variation in DEL within a particular
source category (see histograms in Figure 9) can be "explained"
by environmental factors.  Only the more significant results of
these analyses are presented here.  Additional details of the
source-by-source analyses are presented in Appendix H.

Correlations With Del:  Septic Tank Sources--

     The significant correlations with DEL in the septic tank
source category are listed in Table 14.   The two variables that
correlate most highly (at the .01 level) with DEL are TEMP (.35)
and ELEV (-.39).   These variables are also correlated with each
other, and with some of the other variables that are less highly
correlated with DEL (PRECIP and LAT).   Thus, it appears that the
615N values of nitrates from septic tanks are higher in warm,
moist, low elevation and low latitude areas than in cool, dry,
high elevation, and/or more northerly areas.  The specific
causal relationship between DEL and these four factors cannot
be identified.  A true relationship may exist between DEL and
any one of these variables,  between DEL and a combination of
these variables,  or between DEL and a fifth unidentified variable
which is also correlated with TEMP, PRECIP, ELEV, or LAT.
                              62

-------
            TABLE 14.  VARIABLES WHICH HAVE STATISTICALLY
                      SIGNIFICANT CORRELATIONS WITH DEL:
                      SEPTIC TANK SOURCES
Variable
TEMP
PRECIP
ELEV
LAT
RADSAM
C H
SOIL 10
HLMB
IH
Correlation
with DEL
.35
.33
-.39
-.31
-.27
-.32
-.32
-.25
-.32
Level of
Significance
.01
.05
.01
.05
.10
.05
.05
.10
.05
NOTE:  C H,  SOIL 10, and IH occur simultaneously in this data subset
      and,  therefore, are indistinguishable for the purposes of
      statistics.
     It should  be  noted that the variables  C H,  SOIL 10,  and
IH occur simultaneously in the septic tank  source  category data.
That is, the  correlation between each of these  three variables
is r = 1.0.   They  are therefore indistinguishable  in this analy-
sis .

     The variable  RADSAM is negatively correlated  with DEL
(-.27) in this  data subset.  This means that the <515N values
of samples collected by Radian in this study were  generally
lower than the  515N values for septic tank  samples provided by
Kreitler (2).

Correlation With DEL:  Feedlot Sources--
     The significant correlations with DEL  in the  feedlot (and
barnyard) source category are listed in Table 15.   The variables
which correlate most highly with DEL are DI, RADSAM,  C BSK,  and
ELEV.  All of these correlations are negative (i.e.,  these
variables are inversely related to DEL).  These  correlations
are strongly  influenced by the fact that six of  the eleven
                               63

-------
negative 615N values that were  determined in this study are in-
cluded in the feedlot data  subset.   Five  of these six are as-
sociated with the DI (Dimmitt,  Texas)  site variable.   It is
interesting to note that nitrates  from feedlots had the highest
average 615N values of any  source  category (see the discussions
of means and correlations in  the entire data base above) in
spite of this relatively high prevalence  of negative  <515N values
             TABLE 15.  VARIABLES WHICH HAVE STATISTICALLY
                      SIGNIFICANT CORRELATIONS WITH DEL:
                      FEEDLOT SOURCES
Variable
PRECIP
ELEV
SEQNCE
RADSAM
C BSK
C CAF
SOIL 5
SOIL 6
DI
LOCTAY
Correlation
with DEL
.32
-.37
-.30
-.38
-.38
.25
.32
-.32
-.44
.27
Level of
Significance
.05
.01
.05
.01
.01
.10
.05
.05
.01
.10
     The negative correlation of climate  class  BSk with 615N
can largely be attributed to the DI  samples.  The Dimmitt site
is located in climate zone BSk.

     The negative correlation with RADSAM indicates that the
del values of samples collected and  analyzed  for this study are
lower than the del values reported by Kreitler  for barnyards in
Runnels County, Texas, the Lockhart-Taylor Area,  Texas, and
Macon County, Missouri.

Correlations With DEL:  Irrigation Sources--
     Only one variable (SOIL) was correlated  with DEL in the
waste water irrigation data subset (see Table 16).   This is a
                               64

-------
fairly "weak" negative  correlation (i.e., the level of  signifi-
cance is  .10).   It  indicates the two water samples had  higher
615N values than the  eight soil samples that were analyzed.

Correlations With DEL:   Other Sources--
     There were  no  meaningful correlations of variables with
DEL in the "other"  nitrate source category.  It is important
to note that all of the points in this data subset are  from
Runnels County,  Texas.   Thus, the only meaningful variables
that could be correlated with DEL are PPM, D*SOIL, and  SOIL.
The fact  that no significant correlations were found indicates
simply that the  influence of environmental factors (e.g.,  PRECIP,
TEMP, etc.) was  not assessed for this source category in  this
study
            TABLE 16.  VARIABLES WHICH HAVE STATISTICALLY
                      SIGNIFICANT CORRELATIONS WITH DEL:
                      IRRIGATION SOURCES
                              Correlation         Level of
      Variable                 with DEL          Significance
      Soil                      -.55               .10

NOTE:  This data set included only 10 points, so a correlation of at
      least .54 was required for significance at the .10 level.

Regression Models for DEL--
     Multiple  regression models for DEL were constructed for
each of the nitrate  source categories.  The regression  models
were developed by using stepwise regression analysis.   This
method adds variables to and deletes variables from a model one
at a time  until no  further significant improvement in the  pre-
dictive accuracy can be made.

     The primary purpose of this analysis is not to develop equa-
tions to be used to  estimate DEL, but to investigate  the strength
                               65

-------
of the relationship between DEL and the prediction variables.
The set of candidate predictor variables included all except
the site variables (HLMA, HLMB,  etc.), RADSAM, and SEQNCE.

     The variables that were included in the final regression
model equations are not necessarily the ones that are most
highly correlated with DEL in Tables 14, 15, and 16.  If two
predictor variables are highly correlated with each other, it
is unlikely that both variables  will be selected for inclusion
in the equation.  This is due to the fact that if one of the two
variables is already in the equation, the other variable will
make very little additional contribution to the predictive
accuracy.

     When there are many candidate predictor variables which
are correlated with each other (as is the case in this analysis)
the set of variables which is included in the equation is a
complicated function of the nature of the correlations.   The
variables included in an equation for DEL and the list of vari-
ables which have the highest correlations with DEL may therefore
be very different.

     The results of the multiple regression analyses are sum-
marized in Table 17.   Additional details of the analyses, in-
cluding the forms of all four equations for DEL and the analysis
of variance tables, are given in Appendix H.

     It is seen from Table 17 that the multiple correlations
for septic,  feedlot,  and irrigation sources are all between .54
and .56;  thus,  only about

                         C.55)2  = .30

or 30 percent of the total variation in DEL can be explained in
terms of the predictor variables used in this study.  In the
                               66

-------
                 TABLE 17.  COMPOSITE RELATIONSHIPS BETWEEN
                            DEL AND PREDICTOR VARIABLES
Source
Septic

Feedlot



Irrigation
Other

Terms
in Multiple
Modelt Correlation
D*SOIL . 56
ELEV
SOIL
ELEV .54
SOIL
C BSK
SOIL 5
SOIL .55
PPM .23
SOIL
Significant Standard
at Level Error
.01 8.3

.01 8.3



.10* 6.4
.25 5.4

t
 These variables were selected for inclusion in the equations by using
 the stepwise regression procedure.

-'Borderline.  The small sample size (10) accounts for the lower level
 of significance in this case.

-------
case of other  sources,  only about 5 percent  can be explained.
It should be remembered that the only meaningful predictor
variables in the  "other" data subset are PPM,  D*SOIL, and SOIL.

Correlations With PPM By Source--
     The correlations between nitrate concentration (PPM) and
the various environmental factors were also  examined.  Complete
tables of the  significant correlations with  PPM are included in
Appendix H.  A summary of the variables which  correlate most
highly with PPM is presented in Table 18.

          TABLE 18.  VARIABLES WHICH CORRELATE MOST HIGHLY WITH
                   PPM (AT THE .01 LEVEL OF SIGNIFICANCE) FOR
                   THE VARIOUS SOURCE CATEGORIES
Source
Septic Tank


Feedlot






Irrigation
Other

Variable
RADSAM
SOIL 6
RUNLS
TEMP
LAT
SEQNCE
RADSAM
C BSK
C CAP
RUNLS
SNGLO
D*SOIL
SOIL
Correlation
with PPM
-.41
.39
.64
.46
-.40
-.38
-.40
-.40
-.52
.63
.77t
-.34
-.33
 This data set included only 10 points, so a correlation of at least
 .75 was required for significnace  at the .01 level  (see Table 10),

ANALYTICAL PROBLEMS

     When sample  analysis  activities began,  several problems
occurred with  the preparation and purification apparatus  (the

                                 68

-------
"prep-line" is described in Section  5.  The problems, which  were
related to equipment failure,  included a  ruptured  Toeppler pump,
a cracked mercury diffusion pump,  and numerous  leaks.   Leaking
S tnnrnr'ks TA7PTP> rort^a t-^rll \T  t--rnnl-> 1 o a/-vmo
a. \_o_ a.v_jxcu uicj-L;U..L y U.J.JL±. us JLUII puny-i, ana
stopcocks were repeatedly troublesome.
     These various problems  led,  in  some  cases,  to  the  loss  of
samples.  Cumulatively,  these problems  resulted  in  considerable
lost time.  Because of this, although over  320 discrete soil and
ground-water samples were obtained in our intensive sampling pro-
gram, only about 140 samples were leached and had their nitrate
concentrations determined.   Only  87  of  these  (62 percent)  contained
sufficient nitrate to warrant further processing and isotopic
analysis  (see below).  Of these 87,  only  66 were successfully
analyzed  for their isotopic  compositions.   The remainder were
either  contaminated with interferring species such  as carbon
monoxide  (CO) or atmospheric nitrogen  (N2)  gas,  or  produced  too
little  N2 gas for a successful isotopic analysis.

     Another problem was that none of the samples obtained via
sampling  kits were received  in time  to  be analyzed  in this study.
The exclusion of these samples from  the final data  base reduced
the geographic coverage  and  the size of the "septic tank" source
data subset.

     Another problem was related  to  the nitrate  concentration
of the  samples.  The configuration of the preparation apparatus,
and the size of the samples, were such  that a nitrate concentra-
tion of at least 5 mg/£  was  required in the soil leachates and
ground-water samples in  order to  generate sufficient N2 gas  for
an isotopic analysis.  Because of this  constraint,  none of the
samples obtained from the Tallahassee,  Florida,  or  Tampa,  Florida
spray irrigation sites were  analyzed.   This limited both the
size of the data base and the geographic  coverage for the
"Irrigation" waste source category.
                                69

-------
     Table 7 presented a summary of the numbers of samples suc-
cessfully analyzed for each site, according to nitrate source
category.  Also included in the table are the numbers of samples
for which data was provided by Kreitler.

     A final problem is a limitation of the method of study
rather than an analytical problem.  It is possible that the total
history  of a sample is unknown.  Thus, the nitrate in a sample
could conceivably have been derived from a variety of sources.
During the sampling, every effort was made to minimize this
possibility by sampling as close to the waste source as possible
(e.g., within an active pen of a cattle feedlot).   The results
of the nitrate concentration and 6l 5N analyses are tabulated in
Appendix D.

     Profiles of nitrate concentration and 615N versus  depth were
examined for the various sampling holes.   An  example  of such a
profile is  presented in Figure  9.   As  can be  seen  in  the  figure,
both nitrate concentration and del value vary over a fairly wide
range with depth.  This degree of variation was fairly typical
for the  profiles developed in this study.  In the previous work
by Kreitler  (2), however, the profiles of del value were much
more constant with depth.  A representative profile prepared by
Kreitler is reproduced in Figure 10 for comparison.

     This difference in the behavior of <515N between different
sites with similar waste sources can be accounted for by two
possible explanations.  The first explanation*is that some un-
identified deviation from the standard analytical procedure
produced a variation in 615N values.  Due to the time losses
discussed above, duplicate analyses of samples were not per-
formed.  Standards of known isotopic composition were run,
however.  The results of these analyses did not indicate any
systematic analytical errors (see Section 5).
                                70

-------
4-1
0)
0)
MH
CX
0)
P
                                                                    20       25  615N (PPt)

                                                                   105      120 ppm NOj
                                                                      __.-.--I
Nitrate Concentration
                                                                '    615N
                                               Figure 9.  Depth Profile of Nitrate Concentration
                                                         and 615N for the Greeley,  Colorado,
                                                         Animal Feedlot Site,  Hole H2.

-------
                2-
                3-
              )M	
                4-
                    200  400
                   NO (mq/kq)
         -5  0   5   10
             6N'5(%0)
                                              20
         Figure 10.
Nitrate and 615N Versus Depth
Beneath a Barnyard With a Definite
Animal Waste Contribution of
Nitrogen, In Runnels County, Texas
[From Kreitler (2)].
     The second explanation is that the differences are real.
Differences in the biogeochemical environments of the sites
could be causing the observed variations in isotopic behavior.

     Several depth-related factors could be involved.  In par-
ticular, variations in the Eh-pH conditions of the soil column
could have a strong influence on the depth distribution of
both nitrate concentration and 615N value.  The variations
could also be related to changes or differences in soil bac-
teria populations, clay content or mineralogy, chemical composi
tion  or soil texture.
                               72

-------
     It is impossible to explain the differences in isotopic be-
havior illustrated in Figures 9 and 10 with the data generated
in this study.  Detailed, site-specific studies, conducted under
a variety of biogeochemical conditions, will be required to fully
explain the N-isotope fractionations which occur in soil columns.

SUMMARY AND CONCLUSIONS

     The objectives of this study were to:

     1.  determine whether 615N values of nitrates
         derived from a given waste source are con-
         sistent in different geographic  regions  of
         the  country and to

     2.  determine whether 615N values of nitrates
         from different waste sources are sufficiently
         different to discriminate between the sources.

     In order to meet these objectives, samples of soil- and
ground water-nitrate were obtained from different regions of
the country, from sites where the source of the nitrate could
be reasonably predicted.  Four nitrate source categories were
examined:

     1.  nitrates from domestic septic tanks;

     2.  nitrates from f eedlots and barnyards ;

     3,  nitrates from municipal wastewater irrigation
         sites; and

     4   "natural" nitrates from unfertilized soil-
         bacteria-vegetation systems  (primarily cul-
         tivated fields).

                                73

-------
     The nitrate contents and the N-isotope compositions of
these samples were determined.   The results were examined using
standard statistical procedures.

     The means and standard deviations of the 615N values were
computed for each of the nitrate source categories.   These statis-
tics were compared using a one-way analysis of variance (F-test)
and other tests.  This analysis indicated that nitrates from sep-
tic tanks,  feedlots, and barnyards can be differentiated from
natural soil nitrates on the basis of their isotopic compositions.
The mean 615N values for septic tank nitrates and for feedlot/
barnyard nitrates were +10.9 and +12.4,  respectively,  while the
mean 6l5N value for natural soil nitrates was +7.3.

     The analysis of variance of the means also indicated that
nitrates from septic tanks and from feedlots and barnyards can-
not be differentiated from each other.   Finally,  nitrates from
municipal wastewater irrigation sites cannot be distinguished
from nitrates from any of the other sources.   The mean 615N
value for irrigation nitrates was +9.5.

     It should be remembered when considering these  results that
the data for natural soil nitrates are all from Runnels County,
Texas, samples.  Additional research is required to  verify
that the mean 6:5N value for natural soil nitrates (+7.3)  is
representative for other regions of the country.

     The finding that nitrates  from wastewater irrigation sites
cannot be differentiated from nitrates from natural  soils, septic
tanks, or feedlots and barnyards may be attributable to the
particular set of conditions that occur at wastewater irrigation
sites.  Wastewater irrigation operations are oriented toward
removal of nutrients from the wastewater, via uptake by plants.
The water generally receives some treatment before irrigation,
and is then sprayed or flooded over the land surface.   Two of
                               74

-------
the three sites included in the irrigation data base were actively
cultivated.  The one site which was not had the highest 615N
values.  These factors tend to suggest that the nitrates at
wastewater irrigation sites represent a mixture of nitrates
derived from the waste and nitrates derived from natural soil-
plant interactions.

     Correlation analyses of the entire data base were carried
out.  The results of these analyses confirmed the ability to
distinguish nitrates from feedlots and barnyards from natural
soil nitrates (the variable FEEDL was positively correlated with
DEL).   This analysis also indicated that some of the variation
in 615N values can result from environmental influences.  The
6l5N value was correlated with mean annual precipitation (high
515N values occurred in areas of high precipitation) and negatively
correlated with elevation (high 615N values occurred at low eleva-
tions).  Identifying specific causal relationships between 615N
and environmental variables will require additional study.

     Correlation analyses were also carried out within each
nitrate source category.  These analyses explored the variation
in 615N values determined for each nitrate source.  The source-
by-source analysis also included multiple regression analyses.
These analyses indicated that about 30 percent of the variation
in 615N for septic tank, feedlot and barnyard, and irrigation
sources can be "explained" in terms of the environmental factors
considered in this study.

     As regards natural soil nitrates, only about 5 percent of
the variation in 615N can be explained.  The only meaningful
variables in this source category are nitrate concentration (PPM),
the depth from which soil samples were obtained (D*SOIL),  and
whether the sample was a soil sample or a ground-water sample
(SOIL).
                                75

-------
     It is fairly clear that environmental factors influence
615N values.   Additional research will be required to precisely
identify the nature of this influence.   It is also apparent,
however, that this influence is not so great as to completely
prevent the identification of nitrate sources by nitrogen iso-
tope studies.

     It is important to note that the mean 615N values of the
nitrate from septic tanks and from feedlots and barnyards were
statistically different from the mean value for natural soil
nitrates.   The ranges of S15N values overlap considerably.   Thus,
it will be impossible to identify the source of the nitrate in
any single sample unless the 615N value is greater than about
+24 ppt.  This is the mean for natural soil nitrate,  plus three
standard deviations:  7.3 + 3(5.4) = 23.5.  The more samples
that are analyzed, the more confident one would be of whether
or not the samples represent contamination by a particular
waste source or mixture of sources.
                               76

-------
                          REFERENCES

1.    Jones,  B.C.   An Investigation of the Nitrate Problem in
     Runnels County, Texas.   EPA-R2-73-267,  U.S.  Environmental
     Protection Agency,  Washington, D.C., 1973.   214 pp.

2.    Kreitler,  C.W.   Determining the Source  of Nitrate in
     Ground Water by Nitrogen Isotope Studies.  Report of
     Investigations  No.  83,  Bureau of Economic Geology,  Austin,
     Texas,  1975.  57 pp.

3.    U.S.  Environmental Protection Agency.  National Interim
     Primary Drinking Water Regulations.   Title 40,  Code of
     Federal Regulations,  Part 141.  Promulgated December 24,
     1975.

4.    Gelperin,  A.  The Health Effects of Nitrate in  Water.
     Proc.  12th Sanitary Engineering Conference.   University
     of Illinois, 1970.   pp.  51-52.

5.    Asahina,  S., et al.  Acute Synergistic  Toxicity and Hepatic
     Neurosis  Following Oral Administration  of Sodium Nitrite
     and Secondary Amines  to Mice.  Cancer Research. 31,  p.
     1201,  1971.

6.    Alam,  B.,  et al.  Synthesis of Nitrosopiperidine From
     Nitrate and Piperidine in the Gastro-Intestinal Tract of
     the Rat.   Nature, 232,  p. 199, 1971.
                                77

-------
 7.   Junk,  H.  and G.  Svec.   The Absolute Abundance of the
      Nitrogen Isotopes in the Atmosphere and Compressed Gas
      From Various Sources.   Geochimica and Cosmochimica Acta,
      14.  p.  234,  1958.

 8.   Garrels,  R.M.,  F.T.  Mackenzie,  and C.H.  Hunt.  Chemical
      Cycles  and the Global  Environment:  Assessing Human
      Influences.   William Kaufman,  Inc.,  Los  Altos,  California,
      1975.   206 pp.

 9.   Hoefs,  J.   Stable Isotope Geochemistry.   Springer-Verlag,
      New York,  1973.   140 pp.

10.   Brown,  K.W.,  J.  Frank  Slowey,  and H.W. Wolf.   Accumulation
      and Passage of Pollutants In Domestic Septic  Tank Disposal
      Fields,  2nd Annual Report.   College  Station,  Texas,  Texas A&M
      University,  Civil Engineering Department and  Soil and Crop
      Sciences  Division,  1976.

11.   Trewartha,  G.T.,  A.H.  Robinson,  and  E.H.  Hammond.   Funda-
      mentals  of Physical  Geography.   McGraw-Hill,  New York,  1961.
      409  pp.
                                78

-------
                          APPENDIX A
              REAGENTS USED IN ISOTOPIC ANALYSES

1.  Magnesium Oxide - Finely ground "Baker Analyzed" reagent
    grade magnesium oxide was used directly from the bottle.

2.  Devarda's Alloy - Reagent grade alloy from Baker composed
    of 50 percent copper, 44.5 percent aluminum, and 5.5 percent
    zinc was ground in a rotary ball mill using carborundum
    balls until it would pass through a 100-mesh screen and
    75 percent would pass through a 200-mesh screen.  In Radian's
    Norton rotary ball mill, two pounds could be processed to this
    specification after 48 hours of operation.   It is absolutely
    necessary that the Devarda's Alloy be ground to a fine pow-
    der.  If it is not, reduction of N03 may be incomplete and
    isotopic fractionation may occur.

3.  0.1 Normal Hydrochloric Acid - This solution was prepared
    from a 0.1 normal hydrochloric acid Baker "Dilut-It" con-
    centrate.

4.  Silver Sulfate - Approximately 65 milligrams was added to
    minimize chloride interference during nitrate (NO^)  analyses
    with the specific ion electrode.

5.  Sodium Hypobromite-Iodide Solution - 200 grams of sodium
    hydroxide were dissolved in 300 milliliters of deionized
    water and cooled to 0°C in ice-sodium chloride bath.  Half
    of the cooled solution was transferred to a 500 milliliter
    Erlenmeyer flask immersed in crushed ice.   Sixty milliliters

                              79

-------
of bromine (Br2) were added over a period of 30 minutes.  The
solution was stirred vigorously during the addition, and the
rate of addition regulated so that the temperature of the
solution did not exceed 5°C.   When the addition of bromine
was complete, the remainder of the sodium hydroxide solution
was added and stirred.  The solution was then stored in a
•refrigerator for six days.  During the period of cold storage
a precipitate of sodium bromide formed.   The solution was
then filtered through a glass-fibre filter and an equal
volume of potassium iodide solution (2 grams KI in one liter
of water) was added.  The sodium hypo-bromite-iodide solution
was stored in a tightly stoppered amber bottle at 4°C.   One
milliliter of this solution will oxidize five to six milli-
grams of NHij to N2 .

-------
                          APPENDIX B
                 DESCRIPTION OF RADIAN'S INTENSIVE
                    SAMPLING PROGRAM,  BY SITES

TEXAS A&M UNIVERSITY EXPERIMENTAL SEPTIC TANK

     The experimental septic tank system at Texas A&M was lo-
cated on the TAMU Research Farm, about 12 miles west of the
university campus.  A plan of the site is shown in Figure B-l.
Operation of the site ceased in 1977.

     The site consisted of three tile  fields and a lysimeter
complex.  These four basic elements were tied into a central
effluent distribution system.   The effluent was derived from a
septic tank serving several houses located about a mile from
the site.  It was delivered to the site via pipeline.   The
average composition of the effluent is presented in Table B-l.

     The lysimeter complex consisted of three lysimeters, each
measuring about 7 feet long, 6 feet across, and 6 feet deep.
Each lysimeter contained a different soil type:  Lakeland
sand; Norwood sandy loam; and Miller clay.  Effluent was ap-
plied at a constant rate by peristaltic pumps.

     There were three tile fields at the site:  the dump field;
the Norwood field; and the Miller field.  The dump field con-
sisted of four parallel laterals, each of which was 50 feet
long.  They were spaced about 10 feet apart.  The dump field
was in Norwood sandy loam soil and was used to dispose of ex-
cess effluent.  No record was kept of the effluent application
rate to this field.
                              81

-------
    TO SEPTIC TANK*
                     5000'
                               DUMP FIELD
                                           DESIGN
                                           RATE
                                           NORWOOD
                                       J  FIELD PLOT
       LYSIMETERS UNDER
       RAIN-SHELTERS
         DEMAND.
         RATE
                                                     100
            -DESIGN
            RATE
Figure  B-l.
Plan of TAMU Experimental
Septic  Tank  System
[From Brown  (10) ]
                           82

-------
   TABLE B-l.   TAMU EXPERIMENTAL  SEPTIC TANK EFFLUENT QUALITY
               FOR 14 MONTH PERIOD

pH
T alkalinity
Conductivity
Hardness
COD
T-PO^-P
0-PO^-P
Organic N
NH3-N
N03-N
N02-N
Cl
SOtt
TOC
IOC
Fecal Coliform
Total Coliform
TC
TN
Mean
7.5
952
2,346
42
292
8.3
6.6
9.2
23.4
1.55
<0.55
166
19
128
125
1.8 x 106
1.8 x 106
220
30.4
Standard
Deviation
± 0.3
± 246 mg/1
± 438 mhos/cm
± 30 mg/1
± 212 mg/1
± 4.5 mg/1
± 3.0 mg/1
± 9.5 mg/1
± 9.4 mg/1
+ 0.44 mg/1

± 64 mg/1
± 12 mg/1
± 81 mg/1
± 39 mg/1
± 1.03 x 106 /100 ml
± 1.03 x 106 /100 ml
± 9.6 mg/1
± 13.6 mg/1
Source:  Brown,  et al.  (10).

-------
     The Norwood field consisted of two parallel, 50-foot long
laterals in Norwood sandy loam.   One lateral was a design rate
line which received effluent at  a standard 40 gallons per day.
The other was a demand rate line.  They were located about 50
feet apart at a depth of 45 cm.   A stand pipe was connected
to the demand field via a 2 cm diameter tube which was posi-
tioned below the level of the line so that the system remained
sealed from the air.   A water stage recorder was positioned on
top of the stand pipe to record  the changes in water level within
the field.  Three stainless steel rods were used in the pipe as
electrodes.   These activated the pump when the water level
dropped to 5 cm from the bottom  of the field and turned it off
when the level rose to within 5  cm of the top of the field.
This demand field began operation near the end of April,  1976.

     The Miller field was set up the same as the Norwood field.
It was located in Miller clay.  The Miller clay is less cap-
able of accepting effluent than  is Norwood sandy loam.  The
Miller design field was therefore set at an application rate
of 20 gpd.

     Sampling of the various fields was accomplished in a vari-
ety of ways.   The lysimeters contained tubes terminating in
porous ceramic suction cups through which soil water samples
were obtained.  The cups were located in an array at different
depths and distances from the line of inflow.  Effluent was
applied through a horizontal perforated pipe enclosed in a
gravel pack.   The pipe was located at a depth of 45 cm along
one wall of the lysimeter.  Soil water samples were obtained
by applying a vacuum to the suction cup tubes.

     Twelve water samples were obtained from lysimeters at TAMU.
Three samples were obtained from a lysimeter containing Lakeland
                             84

-------
sand, 4 were obtained from a lysimeter containing Norwood sandy
loam, and 5 samples were obtained from a lysimeter containing
Miller clay.

     Sampling along a dump field lateral was achieved through
hand augering.  Eight soil samples were obtained from the dump
field.  The positions of the sampling holes are shown in Figure
B-2.  The depth intervals sampled from each hole are shown in
Figure B.-3.  No samples were obtained from the Norwood field or
the Miller field, because sampling these fields would have inter-
fered with the TAMU experimental program.

ROCKDALE, TEXAS, SPETIC TANK SITE

     A septic tank system located approximately 5 miles east
of Rockdale, Texas, was sampled.  The system has an overall
length of 100 feet.  The. current effective length of the lateral
is only about 30 feet, however.

     The system was installed in April, 1966,  and has been in
continuous operation since then.  The lateral extends to the
southeast from the south side of a single residence.   It is
buried at a depth of about 18 inches and is surrounded by a
gravel pack one foot thick.

     At a distance of about 30 feet from the tank, there is
a 3 foot break in the lateral.  The tiles were removed from
this section of the lateral in October of 1976, when they be-
came clogged with roots.  No effluent passes this break at
present.

     Twenty-three soil samples were collected from around the
30 feet  section of lateral still in operation.   They were col-
lected using a 4-inch, hand operated soil auger.  They were placed
in plastic bags and frozen upon return to Austin (within 8 hours).

-------
                               INFLOW
                                    2.0'
                                   /
                                    1.5'
                                   X
                                    2.0'
                                    2.0'
                                                 "°
                                                -10
                                                -20
                                                     DISTANCE
                                                       FROM
                                                      INFLOW
                                                      (FEET)
                                                —30
                                                  40
      THESE  THREE  LINES
         NOT IN  USE
Figure B-2.   Locations of Soil  Sampling  Holes  H1-H4,
             TAMU Experimental  Septic  Tank Dump Field.
                              86

-------
                GROUND
        GRAVEL
       ENVELOPE
        SEPTIC
        LINE
                                              GROUND
                                                         H2
      SURFACE
 GRAVEL
ENVELOPE    [„	Ii5,

 SEPTIC
 LINE
                                                SAMPLED ,
                                                INTERVALS^
                                                         ^
                                                                   -T 0
                                                                    - I
                                                                    - 2
                             - 3
                                                                    -  4
                                                                   -1- 5
               GROUND
                            H4
     GROUND
                 H3

GRAVEL
ENVELOPE

SEPTIC
LINE





SURFACE

^-
\
_r\
\J




- 2.0' 	





/
/
SAMPLED /
INTERVALS\




•1



•TT1
^


1

GRAVEL
ENVELOPE

SEPTIC
LINE





SURFACE

H —
\
r^
\J



-2.0' 	




SAMPLED .•
INTERVALS\


\


H



I


I

                                                                   -r 0
                                                                    - I
                                                                    - Z
                                                                    - 3
                                                                    - 4
                                                                   -1- 5
Figure B-3.   Depth Intervals Sampled in  Holes  H1-H4,
                TAMU Experimental  Septic Tank Dump Field
                                    87

-------
     The positions of the sampling holes and cross sections
showing the depth intervals sampled are shown in Figure B-4
and B-5, respectively.

TAMPA, FLORIDA, SEPTIC TANK SITE

     A plan of the Tampa septic tank site showing the locations
of the sampling holes is presented in Figure B-6.  The site is
owned by a private homeowner.   The septic tank handles the
wastewater load of about seven people.   Disposal of the tank
effluent is to the. ground via a recharge pit.   This occupies
the area marked "Input" on Figure B-6.

     Soil samples were collected at discrete depth intervals
from auger holes at the six locations indicated in Figure B-6.
The maximum depth sampled was  11.5 feet.   A total of 64 soil
samples were acquired.

     Water samples were collected from auger holes H3, H4, H5,
and H6.   When these holes encountered the water table, the sam-
ples were collected using a small peristaltic pump.  In addition,
samples of lake water and water from a well located about 275
feet up-gradient (southeast) from the. recharge pit were collected
The ground-water gradient is to the northwest, toward the lake.

INDIAN HILLS, COLORADO, SEPTIC TANK SITE

     The Indian Hills site was visited twice,  once for water
samples and a second time for a suite of soil samples.  Indian
Hills is an old mountain community (development) about 20 miles
southwest of Denver that has been occupied since early in the
century.  Although a public water supply system has been in-
stituted in recent years, most residences are on individual
water wells and septic tank systems.   The aquifer supplying the
                               88

-------
       5H
   0)
   01
      10 H
   IP  20-
   OO
   t-H
   Q
      15 H
      25 H
      3(H
                           HI
              H2
             rf
                           H3
              H4
             rf-
                           H5 H6 H7
                  +
                            H8
Figure B-4.
Locations of Soil Sampling Holes,
Rockdale, Texas, Septic Tank  Site,
                        89

-------
             HI
            H2
            H3
 GROUND SURFACE
GROUND SURFACE
GROUND SURFACE
                                                                  GROUND SURFACE.
GRAVEL
ENVELOPE


SEPTIC-""
LINE





K-l


r^\
^






o1-










--)


\v

\\^



\\v

GRAVEL
ENVELOPE 	 1

A
SEPTIC--"
B LINE



C

K-i-


r^
^
-------
LAKE
                   H6,
                         H5'
                                                  H3
   0
 _SCALE

10    20
30 FEET
                                                H2
                                                    HI
         Figure  B-6.
             Plan of the Tampa, Florida,
             Septic Tank Site Showing
             Locations of Sampling Holes
                            91

-------
domestic water supplies is a typical aquifer in crystalline
rock terrain.  The porosity is fracture porosity, and the water
is at shallow depths.  The community is now apparently recycling
part of the water inadvertently through the septic tanks and
wells, with resulting high levels of nitrate.  The Jefferson
County Health Department is aware of the situation and is moni-
toring the water quality in wells and regulating septic tank
design.  An intensive study has been conducted in cooperation
with the U.S.G.S.

     During the first visit,  a total of six water samples were
collected from water wells having the highest nitrate concentra-
tions.  Most of these were collected from outside lawn spigots
after letting them run for 5 to 15 minutes.

     During the second trip to Indian Hills,  a likely candidate
septic field was selected for soil sampling.   A plan of the site
is shown in Figure B-7.   The log of the hole showing the sampled
intervals is shown in Figure B-8.   No previous work has been
done specifically at this site.  A more ideal sample point would
have been within the septic field, but the Jefferson County
Health Department advised against this, due to the fact that the
investigation might "put someone in violation of county regula-
tions" by penetrating an engineered septic field structure.  Five
soil samples were collected.

PUEBLO, COLORADO, SEPTIC TANK

     This site is located in a rural area about 10 miles north of
Pueblo, Colorado.  No known previous studies have been made at
the site, which is located on the alluvial bottomlands of
Fountain Creek.
                               92

-------
GENERAL DIRECTION
   OF SLOPE
Figure B-7.   Plan of  the  Indian Hills, Colorado,
              Septic Tank  Site (not to scale).
                        93

-------
GROUND
                                      SURFACE
        LU
        LU
        UL
                   LlJ
                   Q

                               SAMPLED
                               INTERVAL
Figure B-8.
  Depth Intervals Sampled at the Indian Hills,
  Colorado,  Septic Tank Site.
                 94

-------
     A plan of the sample site is shown in Figure B-9.  The
sewage disposal system at the site consists of three parts:
a "grease trap", a septic tank, and a cesspool with laterals.
The samples were collected near one of the laterals leading
away from the cesspool.  A simplified sample log showing the
depths at which samples were collected is shown in Figure B-10.
The soil was moist and appeared to have been exposed to effluent,
although no odor was detected.

     A total of 10 soil samples were collected from the sample
hole.  In addition, a water sample was collected from the water
well, which is located about 50 feet from the septic tank.   This
well is a shallow well (about 30 feet deep) in the Fountain
Creek alluvium.

HOUGHTON LAKE, MICHIGAN..  SEPTIC TANK SITES
     Houghton Lake is located in the north-central part of
Michigan's Lower Penninsula.   The area surrounding the lake is
a popular resort community.   Nearly all of the 27 mile shore-
line is developed.   Permanent residences are interspersed with
summer homes  and cottages.  Until recently, all of the resi-
dences were serviced by individual water wells and septic
tank systems.

     In 1971,  Michigan State University and the State of
Michigan Department of Natural Resources began an intensive
study of the deteriorating water quality in Houghton Lake.   This
study identified septic tank disposal systems as a primary source
of excess nutrient loading in the lake.   Construction of a cen-
tralized sewage collection and treatment facility has been under-
way for several years.   At present about 75 percent of the im-
mediate shoreline community is connected to this system.
                               95

-------
  N
  GARAGE
\
              WATER
               WELL
                                  CESSPOOL
                                     SAMPLE
                                     POINT
Figure B-9.   Plan of the Pueblo, Colorado,
             Septic Tank Site (not to scale)
                    96

-------
        GROUND

             LU
             LU
            X
            I—
            Q_
                6j


                7


                8


                9-1
               10 H



               SURFACE
                                SAMPLED
                                INTERVAL
Figure B-10.
Depth Intervals Sampled at the Pueblo,
Colorado,  Septic Tank Site.
                         97

-------
     Soil samples were collected adjacent to two active septic
tank/tile field systems in the Houghton Lake area.   A combined
total of 19 soil samples were collected from these two sites.

     Site A is located about one-fourth mile from the lake in
the community of Houghton Lake Heights.  A plan of the site is
shown in Figure B-ll.   The system serves a single family resi-
dence and has been in operation for about six years.   Soil sam-
ples were collected at several depth intervals from a hole at
the low end of the drain field.   The depth intervals sampled
are indicated in Figure B-12.  In addition,  a single near-surface
sample was obtained right at the edge of the drain field (see
Figure B-ll).

     Site B is located about one-half mile from the lake on a
glacial moraine.   The site is about 135 feet above  the level
of the lake.   The water table is fairly deep (>20')  and the.
soils are composed of sand and gravel.   The  septic  tank system
at Site B serves  a nursing home  with about 25 to 30 occupants.
The system consists of three tanks  in series.   The  tank effluent
is discharged to  two drain fields.   The system is overloaded
and effluent was  observed seeping from the ground at  one of
the tile fields.   A plan of the  site is shown in Figure B-13.

     One of the tile fields at this site is  functioning pro-
perly.   A suite of soil samples  was obtained from a point
immediately downslope from this  field.   This sample point is
designated B-l in Figure B-13.  The depth intervals sampled
at this point are shown in Figure B-12.  Sampling was only car-
ried out to a depth of 3.2 feet because a dense gravel layer
was encountered.
                              98

-------
ROAD






1
f




1
I














EXTENT OF
GRAVEL FIELD








LATERALS 	 «=















SEPTIC
TANK











/
/




1
1







/







J
;

*
~^.
\ «







HOUSE






SURFACE
SAMPLE


) ©
V
SAMPLE
HOLE











o
a




Figure B-ll.  Plan of Houghton Lake, Michigan,
              Septic Tank Site A (not to scale)
                      99

-------
O
O
                SITE A

        GROUND          SURFACE
               1 -
               4-
      SITE B-l

GROUND          SURFACE
                                               1-
                                             O)
                                             Ol
                                              2-
                                            a.
                                            LU
                                            Q
                                -  SAMPLED INTERVAL
       SITE B-2

GROUND             SURFACE
                                      1-
                                   OJ
                                   O)
                                   Q.
                                   LU
                                   O
                                                                             3-
                                                                             4 -
                                                                             5-
                 Figure B-12.   Depth Intervals Sampled  at the Houghton Lake,
                                 Michigan Septic Tank Sites.

-------
                                                   SAMPLE
                                                   POINT B-2
LU
s
o
x
CD
CO
en.
                                                    SURFACE
                                                     SAMPLE
                                 GENERAL DIRECTION
                                      OF SLOPE
          N
                      TILE
                     FIELD
                  SAMPLE POINT
                      B- 1
         Figure B-13.
Plan of Houghton Lake, Michigan,
Septic Tank Site B  (not to  scale)
                            101

-------
     The other tile field at the nursing home site is over-
loaded.   The exact location of the tile field was not delineated
because of the heavily overgrown,  spongy nature of the soil.
Effluent was observed seeping from the soil surface and flowing
for a distance of about 10 feet down the side of the moraine.
Samples were acquired from a point immediately downs lope from
the point where the effluent reentered the soil.   This sample
point is designated B-2 in Figure B-13.   The sampled intervals
for Point B-2 are also indicated in Figure B-12.

     It was obvious that during periods of heavy use and/or
rainfall, the area of seepage to the surface is much greater
than that observed on the sampling data.   A near-surface (0.2-
0.6 feet) sample was,  therefore, taken at a point about five
feet downslope from Point B-2.   This sample point is also in-
dicated in Figure B-13.

CITY OF TAMPA, FLORIDA,  SPRAY IRRIGATION SITE
     A plan of the Tampa Fowler Avenue Spray Irrigation Site,
showing the locations of the monitoring wells that were sampled,
is presented in Figure B-14.   Duplicate samples were obtained
from each of the 11 wells.   The wells range from 38 to 103 feet
in depth.  The wells were pumped for about five minutes prior
to collecting the water samples.  Ten ml of IN HaSOit were added
to each sample as a preservative.   The samples were stored at
4°C prior to shipment (14 days).  They were frozen upon receipt
in Austin.

     Soil samples were obtained at three locations in Spray Lot
7, adjacent to wells IB and 1C (see Figure B-14).  Spray Lots
6 and 7 have been in operation since September of 1973.
                              102

-------
LT
UNIVERSITY OF SOUTH  FLORIDA
      WELL 6C •
       EAST FOWLER AVE.
  WELL  10

     LOT 3

RAILROAD R/H
            0.1 MILE
                             LOT 4
                WELL 15

                  LOT 5
                                             "RAILROAD T7₯
                                              LOT 11

                                       • WELL 13
                  LOT 12

                WELL 14
LOT 6
                                                                   WELL  1B,,.WELL1C
                                                                                      LOT 7
                                          T7ULROAD  R/U
                                                            UTILITY EASEMENT
                                  FOWLfUjAVENU
                                    WASTEWATER
                                 TREATMENT PLANT
                                LOT  19
                                                            WELL 3B
                                                            LOT 20
                  Figure  B-14.
        Plan of Tampa,  Florida,  Wastewater
        Spray  Irrigation  Site.

-------
     The sample collected at Location A consists of a single  1.25
inch diameter core.  The core sample extends from the surface
to a depth of 10   feet.  Those collected at Location B consist
of both core and augered samples.  The core extends to a maximum
depth of 20 feet.  Augered samples were collected from discrete
intervals to a depth of about   7   feet.

     Samples collected at Location C consist of two cores and a
suite of auger samples.  One core extends to a depth of 20
feet, and the other to a depth of 1.9 feet.  The augered samples
were collected from discrete intervals to a depth of  8    feet.
A total of 21 discrete augered samples were collected from the
three locations.

TALLAHASSEE, FLORIDA, SPRAY IRRIGATION SITE
     The Tallahassee spray irrigation site is located adjacent
to the Thomas P.  Smith Wastewater Renovation Plant.   Figure B-15
is a plan of the plant and spray fields, showing the locations
of the three soil sample holes and four of the six wells that
were sampled.  These four wells are maintained by the City of
Tallahassee.  The average depth is about 40 feet.  The two other
wells that were sampled are maintained by the U.S. Geological
Survey.  They are located about 0.1 and 0.5 miles south  (down-
gradient) of the irrigation fields.  Well No. 17 is 152 feet
deep and Well No. 21 is 247 feet deep.  Water levels vary over
the region and during the year from about 15 feet to 35 feet
below land surface.

     The four city wells  are  completed in sand overlying the St.
Marks Limestone (Floridan Aquifer).   The two  U.S.G.S.  wells are
cased through the surficial  sand deposits and are open to the
limestone of the  Floridan Aquifer.   The sand  and the underlying
Floridan Aquifer  are hydraulically connected.
                              104

-------
o
Ol
                   SINGLE-GUN

                   AREA

                                     TALLAHASSEE CITY LIMIT
                                                   \  \
                                                    \  \
                                                     \  \
                                             CAPITAL  CIRCLE  SR 261
                                                                                   \
O  O  O  On
                                                  I
                                             PONDS
IRRIG
FIE
112 •
HI
• CB8D
•e-3 •
AT I ON
LDS


           • WELL

           m AUGER HOLE

           A SPRAY NOZZLE


               Figure B-15.
                                                    500 FEET
               0
150  METERS
   Plan of Tallahassee, Florida,  Spray Irrigation Site

   Showing Locations of Sampled Wells and  Auger Holes.

-------
     Samples from the six wells were obtained by pumping the
wells for about five minutes,  or until the water level dropped
below the pump.  Two one-quart samples were then collected.  Ten
ml of IN H2SOIt  were added to the samples collected from Wells
4-2, 6-3, and 17.   No preservative was added to the samples col-
lected from Wells CB8D,  4-1, and 21.  These samples were placed
in a freezer within one  hour of the time they were collected,
and had frozen within five hours.

     The soil samples were collected using a   4   inch diameter
soil auger.  The samples were taken from measured depth incre-
ments and were sealed in plastic bags.  They were put on dry
ice within three hours of the time they were collected.  The
locations of the three soil sampling holes are shown in Figure
B-15.  The depth intervals sampled in each hole are shown in
Figure B-16.  A total of 19 soil samples were collected.

ABILENE, TEXAS, LAND APPLICATION SITE
     The Abilene wastewater irrigation farm is located in Jones
and Shackleford Counties, Texas, about eight miles northeast of
the city off State Route 351.   A map of the site is presented
in Figure B-17.  The treatment plant, immediately adjacent to
the farm, handles an average wastewater flow of slightly more
than ten million gallons per day.   After receiving secondary
treatment,  the water is  stored in several large holding ponds.
The water is delivered to the fields via ditches.   Irrigation
is by flooding (overland flow and/or ridge and furrow).  When
no irrigation is required, the water is discharged from the
ponds to the Deadman Creek drainage system, a part of the Brazos
River Basin.  The irrigation schedule is determined by the farm-
ers who lease the land from the city.

     A total of 16 soil  samples were acquired from two locations
at the farm.  The samples are from specific depth intervals from
the surface to a maximum depth of about seven feet.  The sampling
                               106

-------
  TAL - HI
             TAL - H3
                                                              TAL - H2
GROUND SURFACE
      S
                  GROUND  SURFACE
                               GROUND SURFACE
                                                                           D.
                                                                           U-l
Figure E-16.
Depth Intervals Sampled at  the  Tallahassee, Florida,
Wastewater  Spray Irrigation Site.

-------
      EXPLANATION

           HOLDING PONDS

           TREATMENT PLANT

       	 FARM BOUNDARY
Figure  B-17.  Plan  of the Abilene Municipal Wastewater
               Land  Application Site,  Showing Sampling
               Locations.
                             108

-------
locations, designated Hi and H2,  are indicated in Figure B-17.
The depth intervals sampled from each site are shown in
Figure B-18.

     Six water samples were obtained at the Abilene site.  Two
of the samples are from wells located within the irrigated area.
The wells are shallow (about 25 and 50 feet).  One is not in
use.  The other is used for watering a lawn and small vegetable
garden.  Both are completed in a sand and gravel bed.

     Two of the water samples are of the irrigation water
(treated effluent).  These samples were taken from ditches
delivering water to the fields.  The last two water samples
were obtained from soil sampling hole HI.  Water was encountered
in HI at a depth of 5.2 feet.  The hole was extended to 6.3 feet
and a sample of water was collected.  The hole was then bailed
out four times, over a period of about 30 minutes, and another
sample was acquired.  The locations from which the water samples
were obtained are shown in Figure B-17.

SAN ANGELO, TEXAS, LAND APPLICATION SITE
     The San Angelo wastewater irrigation farm is located about
five miles east of the city off Texas Farm/Market Road 380.  A
map of the site is presented in Figure B-19.  At present, waste-
water receives primary treatment only.  The plant is currently
being expanded to provide secondary treatment,  -pj^g method of
irrigation is flooding.

     Soil samples were collected at four locations on the farm.
These are indicated in Figure B-19.  At two of the sampling
locations (HI and H3) surface samples were collected.  Samples
were collected from discrete depth intervals at the other two
locations (H2 and H4) to a maximum depth of 6.4 feet.  The depth
intervals sampled at these two sites are shown in Figure B-20.
The total number of samples collected was 15.

                             109

-------
            ABL-H1
                    ABL-H2
GROUND SURFACE
            1 -
            2 -
            4 -
            6-
                   M.T.
  SAMPLED
  INTERVAL
                 Figure  B-18.
        GROUND SURFACE
                                                    3 -
                                                 O-
                                                 LoJ
                                                    7-
                                                       \\
Depth  Intervals Sampled At the
Abilene  Land Application Site.

-------
EXPLANATION
      HOLDING PONDS

    — FARM  BOUNDARY
                         0-"> EFF 1
               FARM
              BUILDINGS
                        SECONDARY TREATMENT PLANT
                        (UNDER CONSTRUCTION)
  Figure B-19.
Plan  of the San Angelo, Texas,  Municipal
Wastewater Land Application Site,  Showing
Sampling Locations.

             Ill

-------
             SNGLO-H2
                          SNGLO-H4
GROUND SURFACE
              1 -
              2 -
          a.
          LU
          a
             4 -
             6 -
 SAMPLE
 INTERVAL
              GROUND SURFACE
                           1-
                       a.
                       LU
                       Q
                                                      3-
                           4 -
                           6-
Figure B-20.
Depth Intervals Sampled  at the
San Angelo,  Texas, Land  Applica-
tion Site.

-------
     Four water  samples were collected in San Angelo.  Three
are well-water samples from wells in the irrigated areas.   The
fourth is a sample of the irrigation water.

     Well No.  1  serves the  farm  building complex.  The well is
 78  feet  deep.  The water  level at the time of sampling was
 about  35 feet  below  land  surface.   It is operated with an elec-
 tric submersible pump.  The sample  was collected at  the well
 head after pumping for a  period  of  20 minutes.  This well may
 penetrate more than  one water-bearing zone.

     Well No.  2  is located  adjacent  to a small corral and is
 used to  supply a watering trough.   It is 48 feet deep.  The
 water  level was  at 22 feet.  The sample was collected in the
 same manner as at Well No.  1.

     Well No.  3  is a  shallow well (exact depth unknown) used to
 maintain a stock watering pond.  It  is operated by a windmill
 and pumps nearly continuously at about 4 gpm.  The sample was
 collected from the pipe where the well discharges to the pond.

     The sample  of irrigation water  was collected as the water
 was being applied to  a field.  The  delivery system is made up
 of  buried 12-inch pipe along the boundaries of the fields.
 Water  is applied directly from the holding ponds.   The water
 sampling locations are indicated in  Figure B-19.

 MICHIGAN STATE UNIVERSITY WASTEWATER LAND APPLICATION SITE
     Michigan  State University began its program to investigate
 the effects of land application  of municipal wastewater in 1973.
 A plan of the  irrigation site is shown in Figure B-21.   The site
was once farmed but has been lying idle for about 15 years.   The
 vegetation cover is mainly old-field vegetation.   About one-
 fifth of the site is  comprised of a  Beech-Maple woodlot.
                               113

-------
                                         •FELTON  DRAIN
                                                                 Typical arrangement
                                                                 of  surface aluminum
                                                                 irrigation laterals
  UNDERGROUND
  PIPELINE
t
  REMOTELY
  CONTROLLED
  VALVES
IRRIGATION
BOUNDARY
                                              BUFFER ZONE
                                                                 100m
     Figure B-21.   Plan  of the MSU Experimental Wastewater  Application
                    Site,  Showing Sampling Locations  (A,B,C).

-------
     Samples were collected from three locations at the MSU
spray site  (see Figure B-21).  Sample  Point A is located in an
old-field area receiving 4 inches of effluent per week.  Sample
Point B is  located in a control area which receives no effluent.
The third soil sampling, Point C, is located in the Beech-Maple
woodlot.  This area receives 4 inches of effluent per week.  The
depth intervals sampled at each point are indicated in Figure
B-22.  Sixteen soil samples were collected in all.

     In addition to these soil samples, seven water samples were
collected.  Five of these are of soil water obtained from lysi-
meters in the old-field and woodlot areas.  The sixth is from
a pond located in the woodlot, and the seventh is a sample of
ground water.  This was obtained from the top of the water table
(depth = 5  feet) at sample Point C.

GREELEY, COLORADO, CATTLE FEEDLOT SITE
     Soil samples were collected from a feedlot near Greeley,
Colorado.  No known previous work has been done for this site,
with the exception of soil mapping by the Soil Conservation
Service.  The feedlot is an operation consisting of about 40,000
cattle.   Part of the feedlot is on the bottomland of the Cache
la Poudre River (a tributary of the South Platte),  and part is
on an adjoining terrace.  Runoff from the feedlot is apparently
impounded to prevent it from entering the Cache la  Poudre River.

     Two sites were selected for sampling at this feedlot.   Both
are in cattle pens.   The first site (A) was chosen  in the older
part of the feedlot,  which has been operating since the 1930's.
This site is on a gravel terrace.   The log for the  sample hole
is shown in Figure B-23.  A total of 5 soil samples were re-
tained for transport back to Austin.   Gravel that could not be
penetrated with the soil auger was encountered at a depth of
                              115

-------
POINT A
POINT B
POINT C
3ROUND
SURFACE
1 -
2-
i—
UJ
UJ
— 3-
in
(—
Q_
UJ
Q
4-
5 -
6-
i

^^

is
\\x

b
Vv\

S^ SAMPLED
\
-------
only two feet, so the sample point was considered only a partial
success.  A second sample site (B) was therefore chosen on the
bottomlands, as a backup.  This site has been used as a feedlot
for about 12 years.   The log of this sample hole is also given
in Figure B-23.  Seven samples were collected from this hole to
a depth of 6 feet.

     In addition to these soils samples, one water sample was
collected from a shallow well near the feedlot.   No data are
available for this well, but it is apparently a shallow well
in the Cache la Poudre River alluvium.  The water has a foul
odor and appears to be contaminated, but part of the contamina-
tion may be from surface runoff.

DIMMITT, TEXAS, CATTLE FEEDLOT SITE
     The Dimmitt Feedlot was selected as the representative
feedlot for the North Texas High Plains.  Little or no previous
work has been done at the site except for soil mapping by the
Soil Conservation Service.   The site is located in the Ogallalla
Formation.  The feedlot is built concentrically around one of
the "playas" that are typical of the Llano Estacado.   Drainage
from the feedlot is inward into the central playa lake.  The
feedlot has been in operation about eleven years.   A plan of
the site is shown in Figure B-24.

     Two sites were selected for soil sampling at the Dimmitt
Feed Yard.  The first site (A) was selected in a cattle pen in
the oldest part of the feedlot.  A log of the sample hole is
shown in Figure B-25.  The hole was augered to a depth of only
48 inches because a hard, impenetrable caliche layer was en-
countered at that depth.
                              117

-------
                            POINT A
                POINT  B
CO
            GROUND SURFACE
            SAMPLED
            INTERVAL
                         LU  -,
                         UJ  i
                        a_
                        LU
                        Q
                            2 H
GROUND SURFACE
            Figure B-23.   Depth Intervals Sampled at
                            the Greeley,  Colorado Feedlot
                            Site.
               1 -
                                                                         o 	
            i—
            LU
            UJ
            Q-
            UJ
            Q
                5 ~'

-------
   N
                                         TEXAS
    SAMPLE
    POINT A
WATER WELL
SAMPLE POINT
                                     FEEDLOT
                                     BOUNDARY
    Figure  B-24.
Plan of  the Dimmitt, Texas,
Feedlot  Site,  Showing  Sampling
Locations.
                         119

-------
                POINT A
GROUND SURFACE
   SAMPLED
   INTERVAL
                1-
             Q.
             UJ
             Q
                3-
          Figure B-25
                                 POINT  B
                 GROUND SURFACE
Depth Intervals Sampled
at the  Dimmitt, Texas
Feedlot Site
                                 1-
                                                          2-
                                 3-
                                                          4-
                                                          5-
                                                       o_
                                                       UJ
                                 6-
                                                          7-
                                                          8-
                                                          9-
                                                         10
                                                             >0\\

-------
     Experience at the first sample hole and at the site at
Greeley indicated that little precipitation probably infiltrates
within the cattle pens.  The tightly packed manure mat in the
pens appears to act as a virtual pavement, and most precipita-
tion probably goes to surface runoff.  Consequently, it was
decided to collect a suite of samples at the edge of the playa
lake at the Dimmitt site.  The log of this sample hole (B) is
also shown in Figure B-25.  The upper 8 inches at this site is
black, spongy, odoriferous washed-in manure.  Fourteen samples
were collected from these two holes.

     A single water well sample was collected from an irrigation
well near the site (Figure B-24).   No data were obtained for
this well, but the pump size indicated the well is several hun-
dred feet deep.
                               121

-------
                           APPENDIX C
                     SAMPLE NUMBERING  SYSTEM

      All samples  used in  the  statistical  analysis were  assigned
 a sample code number for  identification.   The  sample numbers
 consist of several parts  separated by hyphens.

      The first part of the sample  number  identifies the location
 from which the sample was obtained.   Location  codes are defined
 in Table C-l.

      The second part of the sample number identifies the sampl-
 ing hole, in the case of  soil samples,  or the  well number,  in
 the case of ground-water  samples.   For samples collected by
 Radian in the intensive sampling program,  these are related to
 the holes identified in the various sections of Appendix B.

      Soil samples have a  third part in their sample identifica-
 tion numbers.   This part  is a number  indicating the level within
 a particular sampling hole from which the sample was obtained.
 This number is roughly related to  depth:   deeper samples have
 higher numbers.   An example of the numbering system follows:

                      a sample is numbered:
                            HLMA-H2-3

 HLMA indicates that the sample is  from Houghton Lake, Michigan,
site A. H2 indicates that  the  sample is from the second  sampling
                               122

-------
     TABLE C-l,
LOCATION CODES USED IN SAMPLE
IDENTIFICATION NUMBERS
ID Code
          Sampling Location
TAMU


RT

TAM

IH

PU

HLMA

HLMB

TAL

ABL

SNGLO

MSU


GR

DI

LT

GC

MCM

QNY

RU
 Texas A&M University; College Station,
   Texas

 Rockdale, Texas

 Tampa, Florida

 Indian Hills, Colorado

 Pueblo, Colorado

 Houghton Lake, Michigan, Site A

 Houghton Lake, Michigan, Site B

 Tallahassee, Florida

 Abilene, Texas

 San Angelo,  Texas

 Michigan State University;  East
   Lansing,  Michigan

 Greeley, Colorado

 Dimmitt, Texas

 Lockhart-Taylor Area, Texas

 Grand Caymen Island  (Caribbean)

 Macon County, Missouri

 Queens County, New York

 Runnels County, Texas
                          123

-------
hole completed at  that  site,  and  the  "3"  indicates  that the
sample is the third sample collected from the sampling hole
H2.

     These sample code numbers may be used to relate the basic
data given in Appendix D to the sites identified in Section 6
and Appendix B.
                              124

-------
                          APPENDIX D
                          BASIC DATA

     The analytical data upon which this report is based are
presented in Tables D-l and D-2.   Table D-l contains soil
sample data, and Table D-2 contains ground-water sample data
                              125

-------
TABLE D-l.   SOIL SAMPLE DATA
Sample^
Number
HLMA-H1-1
HLMA-H2-2
HLMA-H2-3
HLMA-H2-5
HLMA-H2-6
HLMA-H2-7
HLMB-H1-1
HLMB-H1-3
HLMB-H1-4
HLMB-H2-1
PU-H1-1
PU-H1-4
PU-H1-7
PU-H1-8
PU-H1-9
PU-H1-10
IH-Hl-4
RT-H5-1
RT-H5-2
RT-H8-2
RT-H8-3
TAMU-H1-2
TAMU-H2-1
TAMU-H4-1
TAMU-H2-3
Source*
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
Depth
(meters)
0.2
0.4
0.7
1.2
1.4
1.5
0.1
0.7
0.9
0.1
0.1
1.0
2.0
2.4
2.8
3.3
1.2
0.5
0.8
0.7
1.0
1.1
0.7
0.8
0.6
NO 3
Concentration 615N ,
(ppm) (ppt) Reference
22.1
26.1
46.8
30.8
52.3
25.8
71.5
18.1
9.4
8.1
106.4
34.9
30.0
21.0
48.2
18.9
8.3
5.0
1.8
20.1
8.7
10.9
10.3
5.7
15.0
+10.0
+20.5
+12.3
+ 7.4
+13.9
+10.5
-12.8
+ 7.6
+ 7.8
+ 8.0
-12.7
+12.5
+ 7.3
+ 6.8
+ 6.7
+23.8
-20.0
+17.0
+19.0
+ 7.3
+10.0
+13.0
+ 7.2
+11.0
+ 8.0
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
                               (continued)
              126

-------
TABLE D-l (continued)
Sample^
Number
TAM-H4-1
TAM-H4-3
TAM-H4-4
TAM-H4-5
TAM-H4-6
TAM-H4-7
TAM-H4-8
TAM-H4-9
TAM-H5-2
TAM-H5-4
TAM-H5-11
TAM-H4A- 1
DI-H1-1
DI-H1-3
DI-H2-4
DI-H2-5
DI-H2-6
DI-H2-7
DI-H2-8
DI-H2-9
GR-H2-2
GR-H2-4
GR-H2-6
GR-H2-8
GR-H2-10
GR-Hl-4
GR-H1-7
Source*
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
ST
AW
AW
AW
AW
AW
AW
AW
AW
AW
AW
AW
AW
AW
AW
AW
Depth
(meters)
0.1
0.4
0.8
0.9
1.1
1.2
1.5
1.9
0.4
0.7
3.0
3.0
0.1
0.9
0.7
1.1
1.5
1.9
2.2
2.6
0.2
0.4
0.5
0.8
1.1
0.3
0.6
NO;
Concentration 615N
(ppm) (ppt) Reference
18.0
56.0
23.2
14.2
28.8
10.4
19.3
31.7
34.2
23.8
8.0
102.6
36.2
16.8
40.4
46.8
142.9
48.8
91.9
100.7
30.1
44.0
31.5
47.0
82.0
43.8
30.5
+ 3.9
- 6.3
+ 5.8
+18.8
- 0.4
+21.0
+26.7
+ 3.8
+24.6
+25.9
+30.3
+ 8.6
+ 8.7
- 0.2
- 5.0
+10.5
- 3.9
- 5.7
+23.4
- 5.2
+27.7
+26.4
+13.4
+18.6
+14.2
+ 3.5
-22.0
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
                             (continued)
          127

-------
TABLE D-l (continued)
Sample@
Number
MSU-H1-1
ABL-H1-4
ABL-H1-7
SNGLO-H2-1
SNGLO-H2-2
SNGLO-H2-3
SNGLO-H2-4
SNGLO-H2-6
LT-H1-1
LT-H2-1
LT-H3-1
RU-HA-1
RU-HA-2
RU-HA-3
RU-HA-4
RU-HA-5
RU-HA-6
RU-HA- 7
RU-HB-1
RU-HB-2
RU-HB-3
RU-HB-4
RU-HB-5
RU-HC-1
RU-HC-2
RU-HC-3
RU-HC-4
Source*
MWW
MWW
MWW
MWW
MWW
MWW
MWW
MWW
AW
AW
AW
0
0
0
0
0
0
0
AW
AW
AW
AW
AW
0
0
0
0
Depth
(meters)
0.1
0.5
1.2
0.2
0.5
0.9
1.2
1.9
0.3
0.3
0.3
0.6
0.9
1.2
1.5
2.1
2.4
2.7
0.3
0.9
1.2
1.8
2.3
0.3
0.9
1.5
2.1
NO;
Concentration <515N ,
(ppm) (PPt) Reference
10.4
12.6
6.3
17.1
16.5
22.8
30.8
17.0
1116.0
37.0
25.0
348.0
458.0
381.0
237.0
105.0
104.0
44.0
42.5
25.0
19.0
43.0
32.0
43.0
65.0
75.0
104.0
+ 6.9
+16.7
+ 3.5
+ 6.4
+ 8.5
+ 0.2
+10.4
+ 8.3
+17.3
+37.6
+17.6
+ 4.4
+ 3.8
+ 4.5
+ 3.6
+ 2.7
+ 2.0
+ 2.5
+15.3
+17.9
+11.8
+11.6
+ 7.0
+ 3.3
+ 3.9
+ 5.0
+ 3.8
a
a
a
a
a
a
a
a
b
b
b
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
                          (continued)

-------
TABLE D-l (continued)
S amp 1 e(§
Number
RU-HD-1
RU-HD-2
RU-HD-3
RU-HD-4
RU-HD-5
RU-HD-6
RU-HE-1
RU-HE-2
RU-HE-3
RU-HG-1
RU-HH-1
RU-HH-2
RU-HJ-2
RU-HK-1
RU-HK-2
RU-HK-3
RU-HK-4
RU-HK-5
RU-HL-1
RU-HL-2
RU-HL-3
RU-HM-1
RU-HN- 1
RU-HN-2
RU-HN- 3
RU-HO-1
RU-HO-2
Source*
AW
AW
AW
AW
AW
AW
0
0
0
AW
0
0
AW
0
0
0
0
0
0
0
0
AW
0
0
0
0
0
Depth
Cmeters)
0.3
0.9
1.2
1.5
1.8
3.0
0.6
1.2
2.1
0.6
0.8
1.4
0.9
0.5
1.1
1.8
2.1
2.4
0.3
0.9
1.5
0.9
0.3
0.9
1.5
0.6
0.9
NO 3
Concentration 6 * 5N ,
(ppm) (ppt) Reference
180.0
1682.0
1300.0
429.0
706.0
700.0
30.0
126.0
36.0
135.0
10.0
25.0
805.0
13.0
36.0
58.0
49.0
45.0
78.0
36.0
32.0
1032.0
37.0
32.0
19.0
225.0
450.0
+13.6
+16.5
+15.2
+13.7
+14.1
+14.0
+24.6
+31.0
+19.2
+21.9
+12.4
+ 7.8
+ 9.7
+ 5.8
+ 6.5
+ 5.6
+ 4.6
+ 4.9
+ 4.4
+ 5.7
+ 6.2
+14.0
+ 3.3
+ 8.6
+ 6.2
+ 6.7
+ 5.7
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
                           (continued)
         129

-------
                     CABLE D-l  (continued)
Sample(§
Number
RU-HO-3
RU-HO-4
RU-HP-1
RU-HP-2
RU-HQ-1
RU-HR-1
RU-HS-1
RU-HT-1
RU-HU- 1
RU-HU-2
Source*
0
0
ST
ST
AW
0
0
AW
ST
ST
Depth
(meters)
2.1
2.7
0.6
0.9
1.5
2.0
0.0
0.9
0.6
0.9
NO 3
Concentration 615N
(ppm) (ppt) Reference
65.0
70.0
33.0
33.0
17.0
10.0
631.0
306.0
711.0
523.0
+ 6.4
+ 6.2
+17.8
+13.0
+15.9
+10.4
+ 9.8
+14.6
+10.3
+12.4
c
c
c
c
c
c
c
c
c
c
-Sample numbers are explained in Appendix C.

k
 Sources:  ST = Septic Tank; AW = Animal Waste; WWW = Municipal
 Wastewater;  0 = Other.
t
 References:   a = This work; b = Kreitler, unpublished data-
 c = Kreitler (2) .
                              130

-------
TABLE D-2.  GROUND-WATER SAMPLE DATA
Sample^
Number
IH-W1
IH-W2
IH-W4
IF-W5
MSU-W4
MSU-W5
LT-W1
LT-W2
LT-W3
LT-W4
LT-W5
LT-W6
LT-W7
LT-W8
GC-W1
GC-W2
GC-W3
GC-W4
MCM-W1
MCM-W2
MCM-W3
MCM-W4
MCM-W5
MCM-W6
MCM-W7
MCM-W8
MCM-W9
Source*
ST
ST
ST
ST
MWW
MWW
AW
AW
AW
AW
AW
AW
AW
AW
ST
ST
ST
ST
AW
Aw
AW
ST
AW
AW
AW
AW
AW
NO;
Concentration <515N ,
(ppm) (ppt) Reference
108.0
66.0
90.0
92.0
14.0
8.0
16.0
341.0
313.0
200.0
147.0
226.0
271.0
12.0
27.3
29.0
12.3
11.9
75.0
62.0
105.0
260.0
330.0
60.0
95.0
145.0
215.0
+ 6.0
+ 7.5
+10.0
+ 3.3
+ 8.0
+26.0
+18.2
+17.2
+17.8
+ 7.3
+18.4
+11.8
+ 9.8
+16.6
+21. 7
+14.2
+19.0
+18.0
+15.9
+10.8
+18.8
+14.5
+13.7
+16.6
+13.5
+16.2
+14.8
a
a
a
a
a
a
b
b
b
b
b
b
b
b
b
b
b
b
c
c
c
c
c
c
c
c
c
                                 (continued)




                  131

-------
TABLE D-2 (continued)
Sample
Number
MCM-W10
MCM-W11
QNY-W1
QNY-W2
RU-W15
RU-W16
RU-W18
RU-W67
RU-W105
RU-W211
RU-W165
RU-W-201
RU-W233
RU-W234
RU-W309
RU-W366
RU-W369A
RU-W386
RU-W388
RU-W419
RU-W421
RU-W506
RU-W551
RU-W551A
RU-W552
RU-W728
RU-W865A
RU-W865B
Source*
AW
AW
ST
ST
AW
0
AW
0
0
0
0
0
0
0
0
0
0
0
AW
0
0
0
0
0
0
0
0
0
NO 3
Concentration S 1 5N ,
(ppm) (ppt) Reference
180.0
145.0
20.0
30.0
200.0
243.0
288.0
154.0
21.0
211.0
78.0
96.0
416.0
290.0
390.0
32.0
257.0
366.0
978.0
186.0
105.0
174.0
184.0
470.0
94.0
61.0
1360.0
1260.0
+12.2
+15.2
+21.3
+12.1
+ 3.3
+ 2.0
+ 6.4
+ 7.5
+ 7.4
+ 7.5
- 1.1
+ 2.2
+ 7.0
+ 3.3
+ 4.5
+ 2.5
+ 7.0
+ 5.8
+10.3
+ 5.3
+ 8.6
+ 5.2
+ 6.6
+ 8.3
+ 1.2
+ 2.7
+14.1
+13.0
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
                         (continued)
         132

-------
                     TABLE D-2  (continued)
Sample^-
Number
RU-W867A
RU-W868
RU-W1002
RU-W1003
RU-W1034
RU-W1004
RU-W1005
Source*
AW
0
0
0
AW
0
0
NO;
Concentration & l 5N
(ppm) (ppt) Reference
579.0
21.0
223.0
225.0
1238.0
250.0
75.0
+10.4
+13.1
+ 5.0
+ 6.3
+12.0
+10.0
+ 6.8
c
c
c
c
c
c
c
^-Sample numbers are explained in Appendix C.

*
 Sources:   ST = Septic Tank; AW= Aniipal Waste; MWW = Municipal
 Wastewater; 0 = Other.


 References:  a = This work: b = Kreitler, unpublished data;
 c - Kreitler (2).
                               133

-------
                          APPENDIX E
                    CLIMATE CLASSIFICATIONS

     The sites in this study were classified with respect to
climate using the Koppen system,  as described by Trewartha and
others (11) .   Five great climatic groups are defined under this
system as follows:

     A -- Tropical Humid Climates:   temperature of coolest
          month over 64.4° (18°C) .

     B -- Dry Climates:  potential evaporation exceeds
          precipitation.

     C -- Humid Mesothermal Climates:   temperature of
          coldest month between 64.4°  (18°C) and 26.6°C
     D -- Humid Microthermal Climates:   temperature of
          coldest month under 26.6° (-3°C) ;  warmest month
          over 50° (10°C) .

     E -- Polar Climates:   temperature of warmest month
          under 50° (10°C) .

     In addition, a sixth climate group (H)  is used to refer to
undifferentiated highland climates.  The great complexity of
mountain climates depends  on contrasts in altitude and in ex-
posure to sun and winds.   Highland areas are therefore grouped
together under a single designation.
                              134

-------
     The five main climate groups are subdivided into a number
of finer classifications by adding a one or two letter qualify-
ing code to the letter designation of a climate group (e.g.,
Caf).   The definitions of the various qualifying codes are given
in Table E-l.
          TABLE E-l.   QUALIFYING CODES OF THE KOPPEN
                      CLIMATE CLASSIFICATION SYSTEM*

     Code                      Definition

      S              Steppe

      a              Temperature of the warmest month over
                     71.6° (22°C).

      b              Temperature of the warmest month below
                     71.6° (22°C).

      h              Hot:   average annual temperature over
                     64.4° (18°C).

      k              Cold:  average annual temperature below
                     64.4° (18°C).

      f              Moist throughout the year:  no month
                     with less than 2.4 inches of rain.

      w              Dry season in winter, or low sun period:
                     at least one month with less than 2.4
                     inches of rain.
*0nly those codes used in this study are given.
                              135

-------
                         APPENDIX F
                     SOIL CLASSIFICATIONS
     The soils at the various sites were classified according
to the regional scheme presented by Trewartha and others (11).
Ten basic soil categories are used in this scheme, as shown in
Table F-l.

     Podzol and tundra soils (Classes 1 and 9) are the typical
zonal soils of humid, high-latitude regions.   Podzols occur in
forested areas, while tundra soils lie in the treeless areas
closer to the pole.   These soil classes are largely confined
to the northern hemisphere.   The surface layer of a podzol soil
typically is rich in partially decomposed organic material.  Be-
low this layer, the A horizon is strongly leached.  Material
leached from the A horizon accumulates in the B horizon.  The
C horizon is often composed of sandy glacial drift.

     Podzolic soils  (Class 2) tend to be similar to the true
podzols, but their A horizons are not so heavily leached.  Podzolic
soils are generally more fertile than the true podzols.

     Podzolic-latoso.lic soils (Class 3) occur in the southeastern
United States.  The upper horizons are weathered and clay rich.
They contain a leached zone as a result of podzolic processes.

     Latosolic soils (Class 4) occur in areas of high temperature
and precipitation.  They are commonly very deeply weathered, and
are rich in clay and oxides of iron and aluminum.  These com-
monly occur in a layer of hardpan (laterite).
                              136

-------
            TABLE F-l.   MAJOR CLASSES OF SOILS
              Soil  Class            Identifying Number
     Podzol  Soils                            1
0) en

^|   Podzolic  Soils                          2
>W h4

°T)   Podzolic-latosolic Soils               3
to -H

3I   Latosolic Soils                         4
O X

     Tundra  Soils                            9
  w
 II T)
£. C

u,3    Chernozemic  Soils                       5
M-l

 °!2    Chernozemic-desertic Soils             6
 to S

^j:    Desertic  Soils                          7
 O ,n
w 3
  CO
     Alluvial  Soils                          8

     Complex Soils  of Areas of

       High Local Relief                   10
                             137

-------
     Chernozemic soils (Class 5) develop in dryer areas  than
those above.  They typically contain a deep surface layer which
is rich in organic  materials  and nutrients.   They are not
severely leached.

     Chernozemic-desertic soils  (Class 6) are transitional be-
tween the chernozemic soils and the true desertic soils.  They
may develop a hardpan, commonly of lime (caliche) due to the
low precipitation and high evaporation.

     The true Desertic soils (Class 7) are low in organic con-
tent and typically contain decomposing rock fragments.  Due to
the very low rainfall, these soils are frequently immature and
horizon differentiation is poor.

     Two basically non-zonal categories of soil are the alluvial
soils (Class 8) and the soils of areas of high local relief
(Class 10).

     The characteristics  of alluvial soils are highly dependent
on the characteristics of the parent alluvium.  Little charac-
terization of soils in the areas of high local relief can be
made because of the complex influences of slope and local cli-
mate.
                             138

-------
                           APPENDIX G
          STATISTICAL ANALYSES OF THE ENTIRE DATA BASE

     In this and the following appendix, statistical tables which
supplement those given in Section 7 are presented.  These two
appendices are presented for completeness of documentation and
for the benefit of those who are interested in further details
of the statistical analyses.  The discussions in these two appen-
dices are brief, however, since the results are discussed in Sec-
tion 7 .

     In this appendix,  statistical analyses of the entire data
base, including septic tank, feedlot,  irrigation, and other
sources,  are presented.   Separate analyses of the four source
classes are discussed in the following appendix.

     The  primary objective is to analyze the relationship between
the dependent variable DEL and the various independent variables.
Various approaches to multivariate problems of this type exist,
such as the following:   (1) correlation, partial correlation,
and regression, (2) automatic interaction detection, (3) factor
analysis, and  (4)  path analysis.

     While all of these approaches have merits, the first was
selected for several reasons.  This approach requires relatively
few a priori assumptions and allows one to explore the data
freely.   Correlations and regression,  moreover, allow the analy-
sis of continuous variables as continuous variables, whereas
automatic interaction detection involves discrete subdivisions
of a data set.  Also, the output from the selected method is

                                139

-------
relatively easy to interpret, especially compared to results from
a factor analysis.

     Various data plots, especially plots of regression residuals,
were produced to determine whether a nonlinear form of any of the
variables should be analyzed.  No evidence of nonlinear trends
was seen, however.

     Some of the tables presented in this and in the following
appendix are also presented and discussed in the text.  The com-
plete set of statistical correlation tables is included here for
the convenience of the reader.

     Table G-l indicates the variables which correlate signifi-
cantly with DEL, and Table G-2  presents the complete correlation
matrix for DEL and the variables which correlate most highly
with DEL.  Table G-2 indicates  the extent to which these inde-
pendent variables are correlated with each other.  Note especial-
ly the relatively high correlation between PRECIP and ELEV; these
are the two variables with the  highest correlations with DEL.

     Table G-3 presents the significant correlations of PPM with
other variables.

     Tables G-4 and G-5 present the correlations and partial
correlations of the variables SEPTIC, FEEDL, and IRRIG with DEL.
The partial correlation is the  correlation of one variable with
another given that  one or more  other variables are fixed.   The
most important conclusions from these tables are as follows:

     (1)  The DEL value is higher in the feedlot data than in
         data from septic tank,  irrigation, and other sources
         as a whole.   This is indicated by the positive (al-
         though small)  correlation between FEEDL and DEL.
                               140

-------
     (2) The correlation between  FEEDL and DEL cannot be
         "explained" by the  influence  of other factors;
         as other variables  are fixed,  the correlation
         between FEEDL and DEL changes very little.

     (3) The correlations for the septic  and irrigation
         variables also change by a  fairly small amount
         as other variables  are fixed.

     It should be recalled that the  mean  values for  DEL  for the
four sources are analyzed in Section 7.   The difference  between
the DEL value for feedlots and for the  three other sources as
a whole is  due primarily to  the difference between the DEL values
for feedlots as opposed to those in  the fourth,  or "other" cate-
gory.

             TABLE G-l.  VARIABLES WHICH  HAVE STATISTICALLY
                       SIGNIFICANT CORRELATIONS WITH DEL

TEMP
PRECIP
ELEV
FEEDL
C AW
C BSK
C DAF
C H
SOIL 4
SOIL 5
SOIL 6
SOIL 10
HLM B
IH
DI
LOCTAY
GNDCMN
MACON
RUNLS
Correlation
with DEL
.13
.29
-.31
.19
.15
-.13
.18
-.18
.15
.27
-.26
-.18
-.13
-.18
-.19
.22
.15
.15
-.15
Level of
Significance
.10
.01
.01
.05
.05
.10
.05
.05
.05
.01
.01
.05
.10
.05
.05
.01
.05
.05
.05
                                141

-------
 TABLE G-2.  CORRELATIONS INVOLVING VARIABLES
             MOST CLOSELY RELATED TO DEL

DEL
PRECIP
ELEV
SOIL 5
SOIL 6
LOCTAY
DEL
1.
.29
-.31
.27
-.26
.22
PRECIP
.29
1.
-.61
.21
-.69
.08
ELEV
-.31
-.61
1.
-.35
.35
-.21
SOIL 5
.27
.21
-.35
1.
-.58
.58
SOIL 6
-.26
-.69
.35
-.58
1.
-.33
LOCTAY
.22
.08
-.21
.58
-.33
1.
TABLE G-3.   VARIABLES WHICH HAVE STATISTICALLY
            SIGNIFICANT CORRELATIONS WITH PPM
Variable
TEMP
PRECIP
LAT
RAD SAM
SOIL
SEPTIC
FEEDL
IRRIG
C BSK
C CAP
SOIL 3
SOIL 6
TAM
RUNLS
Correlation
with PPM
.17
-.14
-.13
-.36
-.16
-.25
.29
-.13
-.15
.25
-.14
.23
.14
.36
Level of
Significance
.05
.10
.10
.01
.05
.01
.01
.10
.05
.01
.10
.01
.10
.01
                      142

-------
            TABLE  G-4.   CORRELATION OF  SEPTIC, FEEDL, IRRIG WITH DEL

SEPTIC
FEEDL
IRRIG
Correlation
.07
-.01
Partial
Correlation*
-.03
.15**
-.02
  *The partial correlations were computed with the following variables fixed:
   PPM, ELEV, SOIL, C AW, SOIL 2, SOIL 3, SOIL 5.  These were selected by
   stepwise regression analysis.

 **Statistically significant at the .05 level.
           TABLE G-5.  CORRELATION OF SEPTIC, FEEDL, IRRIG WITH DEL
                       WITH VARIOUS INDEPENDENT VARIABLES FIXED

                              Variables Fixed*






SEPTIC
FEEDL
IRRG





NONE
.07
.19
-.01





ELEV
.05
.24
-.02




SOIL 5
ELEV
.06
.18
-.00 ••



C AW
SOIL 5
ELEV
.03
.18
.01


SOIL 2
C AW
SOIL 5
ELEV
.03
.19
-.03

SOIL 3
SOIL 2
C AW
SOIL 5
ELEV
-.02
.19
-.03
PPM
SOIL 3
SOIL 2
C AW
SOIL 5
ELEV
-.01
.16
-.01
SOIL
PPM
SOIL
SOIL
C AW
SOIL
ELEV
-.03
.15
-.02

3
2

5




*The variables ELEV, SOIL 5, C AW, SOIL 2,  SOIL 3,  PPM,  and SOIL were selected
in that order by stepwise regression analysis for inclusion in a model to
predict DEL.
                                     143

-------
                           APPENDIX H
                 SEPARATE STATISTICAL ANALYSES
                 FOR THE DIFFERENT SOURCE TYPES

     In this and the preceding appendix,  statistical tables which
supplement those given in Section 7 are presented.   In this ap-
pendix, the results done separately for the four different source
types are given.  Some of the correlation tables presented here
are also discussed in the text.   The complete set is given here,
however, for the convenience of the reader.

     Tables H-l through H-4 present the significant correlations
of the other variables with DEL.   Tables  H-5 through H-8 present
similar results for the variable PPM.

     Tables H-9 through H-13 present the  results of multiple re-
gression analyses to develop models to predict DEL.  Table H-9
summarizes the results, and Tables H-10 through H-13 present fur-
ther details, including the forms of all  four equations and the
analysis of variance tables, not presented in the text.

     The primary purpose of this analysis is not to develop equa-
tions to be used to estimate DEL, but to  investigate the strength
of the relationship between DEL and the prediction variables.
This point is discussed further below.

     The regression models were developed by using stepwise re-
gression analysis.   This method adds variables to and deletes
variables from a model one at a time until no further significant
improvement in the predictive accuracy can be made.  The set of

                                144

-------
candidate predictor variables included  all except the site var-
iables  (HLM A,  HLM B,  etc.), RADSAM,  and SEQNCE.   Thus, the mul-
tiple correlations presented can be  thought of as indicating  the
strength of the relationship between DEL and the set of candidate
physical predictor variables.

     It is  seen (from Table H-9) that the multiple correlations
for septic,  feedlot, and irrigation  sources are all between  .54
and .56; thus,  only about
                           (.55)2 =  .30

or 30% of  the  total variation in DEL  can be explained in terms
of the predictor variables used in  this  study.   In the case of
other sources,  only about 5% can be explained.
             TABLE H-l.  VARIABLES WHICH HAVE STATISTICALLY
                        SIGNIFICANT CORRELATIONS WITH DEL:
                        SEPTIC TANK SOURCES
Variable
TEMP
PRECIP
ELEV
LAT
RADSAM
C H
SOIL 10
HLM B
IH
Correlation
with DEL
.35
.33
-.39
-.31
-.27
-.32
-.32
-.25
-.32
Level of
Significance
.01
.05
.01
.05
.10
.05
.05
.10
.05
  NOTE:   C H, SOIL 10, AND IH occur simultaneously in  this data subset and,
         therefore, are indistinguishable for the purposes of statistics.
                                 145

-------
                 TABLE H-2.  VARIABLES WHICH HAVE STATISTICALLY
                             SIGNIFICANT CORRELATIONS WITH DEL:
                             FEEDLOT SOURCES
Variable
PRECIP
ELEV
SEQNCE
RADSAM
C BSK
C CAF
SOIL 5
SOIL 6
DI
LOCTAY
Correlation
with DEL
.32
-.37
-.30
-.38
-.38
.25
.32
-.32
-.44
.27
	 •-••••
Level of
Significance
.05
.01
.05
.01
.01
.10
.05
.05
.01
.10
                 TABLE H-3.  VARIABLES WHICH HAVE STATISTICALLY
                             SIGNIFICANT CORRELATIONS WITH DEL:
                             IRRIGATION SOURCES
                                      Correlation            Level of
     Variable                          with DEL            Significance

   SOIL                                 -.55                   .10
NOTE:  This data set included only 10 points, so a correlation of at least
       .54 was required for significance at the .10 level.
                 TABLE H-4.   VARIABLES WHICH HAVE STATISTICALLY
                             SIGNIFICANT CORRELATIONS WITH DEL:
                             OTHER SOURCES
                      None at .10 level of significance
NOTE:  TEMP, PRECIP, ELEV,  LAT are all constants within this data set.  Only
       one climate classification (C CAF),  one soil type (SOIL 6), and one
       geographical location (RUNLS) were represented.
                                     146

-------
             TABLE H-5.  VARIABLES WHICH HAVE STATISTICALLY
                         SIGNIFICANT CORRELATIONS WITH PPM:
                         SEPTIC TANK SOURCES

Variable
SEQNCE
RADSAM
SOIL 6
MACON
RUNLS
Correlation
with PPM
-.26
-.41
.39
.23
.64
Level of
Significance
.10
.01
.01
.10
.01
             TABLE H-6.  VARIABLES WHICH HAVE STATISTICALLY
                         SIGNIFICANT CORRELATIONS WITH PPM:
                         FEEDLOT SOURCES
Variable
TEMP
ELEV
LAT
SEQNCE
RADSAM
C BSK
C CAF
DI
GR
RUNLS
Correlation
with PPM
.46
-.25
-.40
-.38
-.40
-.40
-.52
-.26
-.26
.62
Level of
Significance
.01
.10
.01
.01
.01
.01
.01
.10
.10
.01
             TABLE H-7.  VARIABLES WHICH HAVE STATISTICALLY
                         SIGNIFICANT CORRELATIONS WITH PPM:
                         IRRIGATION SOURCES
                                   Correlation            Level of
  Variable                          with PPM            Significance

PRECIP                               --63                   .05
SNGLO                                 -11                   .01
                                  147

-------
              TABLE H-8.   VARIABLES WHICH HAVE STATISTICALLY
                          SIGNIFICANT CORRELATIONS WITH PPM:
                          OTHER SOURCES
                                    Correlation           Level of
   Variable                          with PPM           Significance

    D*SOIL                            -.34                  .01
    SOIL                              -.33                  .01
                TABLE H-9.   COMPOSITE RELATIONSHIPS BETWEEN
                            DEL AND PREDICTOR VARIABLES
Source
Septic


Feedlot



Irrigation
Other
Terms
in Multiple
Modelt Correlation
D*SOIL .56
ELEV
SOIL
ELEV .54
SOIL
C BSK
SOIL 5
SOIL .55
PPM .23
SOIL
Significant Standard
at Level Error
.01 8.3


.01 8.3



.10* 6.4
.25 5.4
 These variables were selected for  inclusion in the equations by using
 the stepwise  regression procedure.
'^Borderline.   The small sample size  (10)  accounts  for  the lower level of
 significance  in this case.

                                   148

-------
         TABLE H-10.  REGRESSION MODEL FOR DEL:  SEPTIC TANK SOURCES
   Multiple Correlation:  .56          Standard Error:  8.3
   Regression Model:  DEL = 19.8 + 4.2 (D*SOIL) - .0024 (ELEV) - 11.3 (SOIL)
    Source
   Regression
   Residual
ANALYSIS OF VARIANCE
        Degrees of    Sum of
         Freedom      Squares
             3         1542
            48         3308
   Mean
  Square    F
    514   7.46*
     69
^Easily significant at the .01 level.
          TABLE H-ll.  REGRESSION MODEL FOR DEL:  FEEDLOT SOURCES
    Multiple Correlation:  .55          Standard Error:  6.4
    Regression Model:  DEL = 17.0 - 9.4 (SOIL)
     Source
    Regression
    Residual
ANALYSIS OF VARIANCE
        Degrees of     Sum of
         Freedom       Squares
             4          1327
            47          3152
 Mean
Square    F
  332   4.95*
   67
^Significant at the .01 level.
                                    149

-------
         TABLE H-12.   REGRESSION MODEL FOR DEL:  IRRIGATION SOURCES

Multiple Correlation:  .55           Standard Error:  6.4
Regression Model:  DEL = 17.0 - 9.4 (SOIL)
                           ANALYSIS OF_ VARIANCE
                                   Degress of    Sum of     Mean
   Source                           Fredom       Squares   Square     F _
  Regression                           1           141       141    3.45*
  Residual                             8           327        41

*Significant at  .10 level (borderline).
           TABLE H-13.  REGRESSION MODEL FOR DEL:  OTHER SOURCES

 Multiple Correlation:  .23            Standard Error:  5.4
 Regression Model:  DEL = 5.2 + 0.0036 (PPM) +2.4 (SOIL)
                           ANALYSIS OF VARIANCE

Source
Regression
Residual
Degrees of
Freedom
2
61
Sum of
Squares
95
1750
Mean
Square
47
29

F
1.65*

*Significant at .25 level.
                                    150

-------
                              TECHNICAL REPORT DATA
                        (Please read Instructions on the reverse before completing}
  REPORT NO.
   EPA-600/4-79-050
             3. RECIPIENT'S ACCESSION-NO.
  flTLE AND SUBTITLE
   IDENTIFYING  SOURCES OF SUBSURFACE  NITRATE
   POLLUTION WITH STABLE NITROGEN  ISOTOPES
             5. REPORT DATE

              August  1979
 issuing date
             6. PERFORMING ORGANIZATION CODE
 7. AUTHOR(S)
   T. J. Wolterink,  H.  J. Williamson,  D.  C.
   Jones, T. W.  Grimshaw, and W. F. Holland
             8. PERFORMING ORGANIZATION REPORT NO
               DCN 78-200-155-15
9. PERFORMING ORGANIZATION NAME AND ADDRESS

   Radian Corporation
   8500 Shoal  Creek Boulevard
   Austin, Texas  78758
                                                  10. PROGRAM ELEMENT NO.
             11. CONTRACT/GRANT NO.

                 68-03-2450
 12. SPONSORING AGENCY NAME AND ADDRESS
 Robert S. Kerr Environmental Research Laboratoi
 Office of Research & Development
 U.S. Environmental Protection Agency
 Ada, Oklahoma  74820
             13. TYPE OF REPORT AND PERIOD COVERED
             y  Final  (8/76 -  3/78)
             14. SPONSORING AGENCY CODE

                 EPA/600/15
 15. SUPPLEMENTARY NOTES
 16. ABSTRACT
       This report  describes the methods, results,  conclusions,  and
  recommendations of  an  investigation of a  technique to identify
  sources of nitrate  in  ground water.  A discussion of the theoretical
  basis of the technique is also provided.   Over 300 soil and ground
  water samples were  collected for this study.   The samples are  from
  numerous sites around  the United States,  representing a variety  of
  environmental conditions.  The nitrate in 66  of these samples  was
  separated from other nitrogen species  converted  to N2 gas, purified,
  and analyzed to determine the ratio   N/1I+N.   These data were  com-
  bined with the results of analyses performed  previously by Jones  (1)
  and Kreitler (2).   Standard statistical techniques were used to
  analyze the observed variations in 61 5N values, with respect to  sev-
  eral nitrate sources and various environmental factors.  It was  found
  that nitrates from  feedlots,  barnyards and septic tanks can be dis-
  tinguished from natural soil nitrates on  the  basis of their 6l 5N
  values.  They cannot,  however, be distinguished from each other.
  Environmental factors  contributed to the  observed variation in 6 * 5N
  values.
                           KEY WORDS AND DOCUMENT ANALYSIS
               DESCRIPTORS
                                       b. IDENTIFIERS/OPEN ENDED TERMS
                        c. COSAT1 Field/Group
  Ground Water
  Waste Disposal
  Nitrates
    Septic  Tanks
    Feedlots
    Land Application
    48 F
 3. DISTRIBUTION STATEMENT

   RELEASE  TO PUBLIC
  19. SECURITY CLASS (This Report)
     Unclassified.
21. NO. OF PAGES
   163
                                       20. SECURITY CLASS (This page)
                                          Unclassified.
                                                             22. PRICE
EPA Form 2220-1 (9-73)
151
                                                        V US GOVERNMENT PRINTING OFFICE 1979 -657-060/54^0

-------