This is to certify that the

thesis entitled

SPRINKLER IRRIGATION WITH
SIMULATED SECONDARY WASTEWATER
EFFLUENT OF A TILE DRAINED SOIL
CROPPED IN BROMEGRASS AND CORN

presented by

STEVEN ALEXANDER GRANT

has been accepted towards fulfillment
of the requirements for

M.S. degreein Crop and Soil Sciences

Major prof

Date November 14, 1980

0-7639

 

M-J

UnIVCl'uu'V

-,.

V...-

LIpra RY

‘- m
“‘1! ELQKV

,.__r- -~'

 

 

  

  

 

 

  

~ .nct

UVtRDUt rlflt):
25¢ per day per item

RETURNING LIBRARY MATERIALS:

Place in book return to remove
charge from circulation records

 

SPRINKER IRRIGATION WITH
SIMULATED SECONDARY WASTENATER
EFFLUENT OF A TILE DRAINED SOIL
CROPPED IN BROMEGRASS AND CORN

By
STEVEN ALEXANDER GRANT

A THESIS

Submitted To
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Crop and Soil Sciences

1980

ABSTRACT

SPRINKLER IRRIGATION WITH
SIMULATED SECONDARY NASTEWATER
EFFLUENT OF A TILE DRAINED SOIL
CROPPED IN BROMEGRASS AND CORN

By

Steven Alexander Grant

A tile drained soil cropped in bromegrass and corn was
irrigated with simulated secondary municipal effluent at 100
and 200 cm/year. The efficacy of additional N and K fertilization
was tested. The dry matter production, grain yield and nutrient
content of the crops were measured. The concentrations of
nutrients were determined in the soil and in water from drainage
tiles, wells and irrigation. The flow rates of water in the tile
drains were monitored continuously.

Corn took up more nutrients than bromegrass, mainly because
the dry matter yield of corn was greater. The tile drainage water
concentration of NOS-N and P, however, was typically less under
grass than that under corn. Crop yield and nutrient removal by
corn were increased with N fertilization. Fertilization with N

and K increased yield and nutrient removal by grass.

THIS THESIS IS DEDICATED

TO

ROGER NEWELL

ii

ACKNOWLEDGEMENTS

First, I would like to thank the Michigan State University
Agriculture Experiment Station for the funds to support this
research. Second I wish to record my appreciation of the generous
donation of time and expertise by the members of my advisory committee;
Drs. F. B. Dazzo, B. G. Ellis, B. D. Knezek and M. L. Vitosh. Special
appreciation and affection is expressed to the committee chairman
Dr. R. J. Kunze, whose patience and consideration was severely tested
by the author. Many other faculty have been greatly helpful, especially
Drs. J. M. Tiedje and F. H. Horne. All these men provided the
intellectual guidance but theory is worthless without experimental
evidence. Tremendous help in the laboratory was provided by Mrs.
E. Shields, Miss T. L. Hughes, Mr. J. C. Grove and Miss N. Leeman.
Expert and steady help was provided in the field by Mr. D. A. Hyde.
Finally, I wish to thank the cadre of graduate students with whom
I held innumerable, and some seeming to them interminable, conversa-

tions, especially Mr. T. E. Schumacher and Mr. G. Asrar.

TABLE OF CONTENTS

INTRODUCTION . .........................
Chapter
I. LITERATURE REVIEW .....................

Waste Disposal on Land ..................
Movement of Water and Ions Through the Soil ........
Flow of Water and Ions to Tile Drains ...........
Nitrogen Transformations in Soil .............
Management of Irrigated Soils to Maximize Nitrogen

Removal .........................

11 MATERIALS AND METHODS ...................

Introduction ........................
Description of the Site ..................
Size, Relief and Soil ..................
Plots and Subplots ....................
Tile Lines ........................
Sampling Wells .....................
Description of the Simulated Nastewater ..........
Preparation .......................
Irrigation ........................
Sampling. . ........................
Soil ...........................
Water in Tiles and Wells .................
Phosphate and Potassium vs. Pressure Experiment .....
Plants .........................
Laboratory Analysis ....................
Soil ...........................
Soil Nitrogen Determinations for the 1978 Growing
Season .................. . .....
Denitrification Incubation Experiment .........
Denitrification Assay by Acetylene Inhibition .....
Soil Organic Carbon ..................
Hater ..........................
Plants ..........................
Statistical Analysis ....................

III RESULTS AND DISCUSSION ...................
Concentration of Ions in Irrigation, Tile Drainage,
and Soil Water ......................

Concentration of Ions in Irrigation Water ........
Water and Ions in the Tile Drains ............

iv

56
56
59

Plant Response ...................... 77

Changes in the Soil .................... 87
The Depressive Effect of Grass Upon the NOS-N Concentration 90
IV SUMMARY AND CONCLUSIONS .................. 106

Table

10.

11.

12.

13.

14.

15.

LIST OF TABLES

Effluent Composition ............... . . . .

Mean concentration of ions in simulated wastewater and

in water draining from tiles in 1977 and 1978. .....
Annual water drainage from the tiles in 1977 and 1978. . .

Estimates of leaching fraction of the individual tiles . .

Linear regression and correlation coefficient for the

concentrations of PO4-P and K in tile drainage water . .

Calculated mean concentration of NOE-N in the solution
of soils sampled at indicated depths and water

from tile drains,_l978 .................

Influence of fertilization and irrigation rate on the

nutrient concentration of grass in two cuttings, l977. .

Influence of fertilization and irrigation rate on the

nutrient concentration of grass in two cuttings, l978. .

Influence of fertilization and irrigation rate on

nutrient concentration of corn silage in 1977 and 1978 .

Influence of fertilization and irrigation rate on dry

matter yield of bromegrass in 1977 and 1978 .......

Yield of corn, as silage and grain in T977 and 1978, as
affected by fertilization, effluent application rate

and corn hybrid. . . ..................

Effect of crop and corn variety on uptake of N, P and

K in 1977 and 1978 ...................

Effect of irrigation rate and additional fertilization
on uptake of N, P and K by corn and bromegrass in

1977 and 1978. . . . . . ................

Concentrations of extractable P and exchangeable K, Ca,

Mg and Na in soil samples at given depths (Fall, 1978) .

Extractable phosphorus concentrations in the soil profile

for Spring of 1973 and Spring and Fall of 1977 and 1978.

vi

Page
57

58
60
62

66

78

79

8O

82

83

85

16.

17.

18.

19.

20.
21.

Mean organic N concentration in soil sampled at three
depths and cumulatively to 60 cm ........... 95

Mean NH -N concentrations in soil sampled at indicated
depth? for three periods of the 1978 growing season. . 96

Estimated gains and losses of nitrogen by the soil in
1978 ......................... 100

Mean Organic carbon content of surface soil (0 - 15 cm)
in 1978 ........................ 103

Mean percent NOS-N lost after 7 days of incubation . . . 104

Comparison of denitrification rates as determined by
CZHZ inhibition .................... 105

vii

LIST OF FIGURES
Figure 1 Page

1 Theoretical distribution of streamlines and equi-
potential lines in a tile drained soil (After
Kirkham, 1956) .................... 23

2 Simulated secondary municipal effluent experimental area 42

3 Comparison of two equations to describe the relation
between height behind a weir and tile water flow rate 45

4 Plot of phosphate and potassium concentration in tile
water drainage for the 1977 growing season ...... 67

5 Changes in tile flow rate and concentrations of PO4-P

and K in the water draining from tile #11, for a

single flow event, 10-24-1978 ............. 68
6 Concentration of NO'-N in the tile drainage water under

two crops at two irrigation rates in l977 ....... 91

7 Concentration of NO'-N in the tile drainage water under
two crops at two rrigation rates in 1978 ....... 92

8 Mean NO'-N/C1 ratio in soil at indicated depths,
Sprin , 1978 . . ................... TOT

9 Mean NOS-N/Cl ratio in soil at indicated depths, Fall,
1978 . . . . . .................... 102

viii

INTRODUCTION

 

As every agronomist knows, a weed is a plant growing out
of place. Similarly, pollutants are chemicals which find their
way into environments where they do not belong. As with weeds,
man must determine the anomalousness of the chemicals which make
them pollutants. Unlike weeds, man determines, consciously or
not, the distribution of pollutants. With this role in deter-
mining the fate of pollutants goes both the responsibility and
the ability to separate pollutants from the environment.

One form of pollution is municipal wastewater. The general
problem of the disposal of municipal wastewaters is that when
released to rivers and lakes they stimulate growth of aquatic
flora. If the chemicals in wastewater could be used to fertilize
crops rather than aquatic algae, then effective land disposal
could have the additional benefit of increased crop production.
This is the strategy behind land application of municipal
wastes. Of course, systems in the real world do not always work
as they do in theory, but if the idea is sound, a place can be
found for the idea. This study is one small part in the process

of bringing this promising idea to fruition.

CHAPTER I
LITERATURE REVIEW

Waste Disposal on Land

Though the application of various kinds of sewage to land is
the object of current scientific investigation, the concept is hardly
new. The current interest in land application of sewage is reexamina-
tion of an older, though probably less well understood, practice.
Holman (1977) cited two earlier examples of land disposal: (l)
The use of night soil as fertilizer, an ancient practice that is still
observed widely in Asia; (2) in nineteenth century Europe many communi-
ties disposed of their sewage by irrigation of crop land. Virtually
none of these irrigation systems are in operation today. The encroach-
ment of the expanding cities upon the disposal sites was a factor in
their disappearance.

Though there has been some application of sewage waters to
crop land in the United States, the spread of the method was spurred
by Congress. In response to growing public concern about water pollu-
tion, the Congress enacted the 1972 Amendments to the Federal Water
Pollution Control Act which set a national goal to eliminate the
discharge of pollutants into navigable waters by 1985. One of the
several types of pollution the Amendments addressed was municipal
sewage, which is wastewater of residential and non-industrial commercial

origin (Metcalf and Eddy, Inc., 1972). The Amendments required

3

municipalities, if they were to receive federal funds for sewage
treatment, to consider land application as a possibility (Jacobs,
1977).

Many of the wastewaters put on croplands are treated prior to
application. There are two general objectives of wastewater treatment:
(1) primary treatment is the removal of particulate matter from the
wastewater; and (2) secondary treatment is the reduction of the bio-
logical oxygen demand (B.O.D.) of the organic compounds in the waste-
water by microbiological consumption (Jacobs, 1977).

Untreated wastewaters may be very different from city to city.
Once treated, however, these wastewaters are remarkably similar among
municipalities. For example, in a survey of 809 municipal wastewater
treatment plants, Gakstatter and Allum (1978) reported small standard
errors of the mean concentrations of phosphorus (P) and nitrogen (N)
in treated wastewaters. They identified two basic types of communities,
those which attempted to restrict P pollution and those which did not.
In communities which did not take measures to limit P pollution, the
mean phosphate P concentration was 4.5 ppm and the total P was 6.1 ppm.
In those that did, the mean concentration of phosphate P was 1.2 ppm
and the mean total P concentration was 1.8 ppm. In the same study,
the mean inorganic N content was 6.1 ppm, while the mean total N con—
tent was 15.0 ppm. Jacobs cited 25 ppm as a typical organic carbon
(C) content in treated municipal wastewater.

May and Feinmesser (1977) listed possible advantages of
irrigation with wastewater. Three advantages were presented: (1)

treated wastewater is an inexpensive source of irrigation water; (2) it

4

can also be a source of nutrients in addition to fertilizer; and
(3) wastewater irrigation is an inexpensive final treatment of
wastewater. They cite five disadvantages to this approach: (1)
whereas irrigation demand is seasonal, municipal wastewater supply
is continuous; (2) particulates in the wastewater may clog irriga-
tion equipment and soil pores; (3) some soluble constituents of
wastewater may be toxic to plants; (4) the range of crops which
receive sewage may be limited by health regulations; and (5) the
combination of the crop and soil may be unable to remove enough of

the nutrients to prevent contamination of the groundwater.

Movement of Water and Ions Through Soil

Almost all non-gaseous substances that move through soil do
so in solution. Hence, to understand the movement of various solutes,
including possible pollutants, the flow of water in soil must be
considered. In the following review, flow in water saturated soils
is emphasized since most solutions to tile drainage problems assume
water saturation of the soil. No doubt one motivation of this
assumption is to make the problem more tractable. Unsaturated flow
problems are more difficult to handle mathematically. It is doubtful
that the assumption of saturated flow is much in error, since drainage
tiles are inserted to reduce the height of the water table, below
which the soil is saturated. }

Water moves along gradients of potential energy. Taylor and
Ashcroft (1972) have summarized the basics of soil water energetics.
The total soil water potential energy (Wt), the relative ability of

the water to do work, is the sum of the water's chemical potential

5

(pw or WW) and the potential due to the water's height above a datum,
the gravitational potential (wt = WW + W2). The chemical potential,
in turn, is the sum of the matric, solute and pressure potentials

w m
of attraction between soil and water. Since these forces restrict

(v = W + vS-rwp). The metric potential (Wm) is due to the forces

the soil water, Wm is usually negative, and Wm becomes more negative
as the water content decreases. The solute potential (W5) is that
due to the presence of solutes. In general, the solute potential
decreases as the concentration of solutes increases. In the absence
of membranes, solutes have little effect on the total water potential
and water flow because their contribution, relative to the other
components of water potential, are usually small. In addition,

the ability of the soil to restrict the migration of solutes is limited
and the effect of solute concentration differences are lessened. The
pressure potential (WP) is due to the weight of the overlying water.
Unsaturated soils exert no pressure potential other than that due to
atmospheric pressure. Water in soils generally is thought to move
along gradients of hydraulic potential (Rh), the sum of the matric,

pressure and gravitational potentials (Rh = Wm + W + W2).

P
The cornerstone of the theory of water flow in soil is Darcy's

Law (Childs, 1957)

A‘l‘h
K3: ‘Kw'Zs— (1)

where: Q = quantity of water passed across A in time t
A = area
t = time
Kw = hydraulic conductivity
5 = distance

This may be rewritten as

dWh
V=-KWa'§—— (2)

where V is the darcian velocity (=Q/At) and the hydraulic gradient
is written in its derivative form. Equation (2) can be broken down
to its various vector components:

BWh 3Tb 3Tb
Vx--KW'a—x—,Vy-.-KW‘a‘y—gVz--KW8—z—. (3)

The equation of continuity can be written as

d_e=_EYE .15. .3_V_z_
t (3x By 32 ) (4)

where e is the soil moisture content. Substituting (3) in (4) assuming

that Kw is constant and that the soil is saturated (%%-= 0), then

azvh + azvh + azvh = 0 ' (5)

3X2 3y2 322

which is Laplace's equation (Kirkham and Powers, 1972). In deriving
Laplace's equation, three explicit assumptions have been made, the
violation of any brings to question the applicability of the equation
for flow through soils. It has been assumed that the soil in question
is saturated, that the hydraulic conductivity is uniform throughout
the soil and that Darcy's Law, with its associated assumptions, com-
pletely describes the flow of water in the soil. These assumptions
are true of few if any soils in either the laboratory or the field.

Since the equation is widely used to describe water flow through soil,

the benefits must outweigh the errors inherent in its use. The
advantages of using Laplace's equation are considerable. The equation
is used in many other disciplines, so that a problem with the same
boundary conditions as a soil problem may have already been solved.
Another benefit of the wide, interdisciplinary use of Laplace's
equation is that many different approaches are available to aid in

its solution to a specific problem. Some solutions of flow to drains
using Laplace's equation will be presented later.

In order to adequately describe the movement of solutes through
soils, a formula must contain an approximation of both the movement of
substances through soil without reaction and the interaction of the
ion with the soil. One way to do this is to add a term representing
this interaction to the differential equation describing flow through
porous media. One—dimensional flow of a non-reacting solute in porous

media may be represented by

— 21.9
JS - vC - D as

solute flux

darcian velocity
solute concentration
disperson coefficient
distance.

where:

MUO<L

Combining this equation with the equation for the conservation of mass

and adding a term to represent the interaction with soil yields

3c _ _3 ac __a-
5—- 35(D-5g) - aS(VC) + fn(cs S: taco-)3 (7)

where fn is some function which describes the release or removal of

the solute by the soil. Assuming that both dispersion and velocity

are uniform (§%-= O, %§-= 0) equation (7) reduces to the more familiar
equation:

§£=o——32C-v39 +f (8)

as as2 as n

Equation (8) can be solved numerically or analytically, depending
upon the boundary conditions and the nature of the term fn (Nielsen
et al., 1972).

The adequate description of the term fn is particularly elusive.
Certainly the variety of soils and the numerous minerals found in soils
make the job difficult. In addition, different solutes react by
different mechanisms with the same soil. Most soil-solute interactions
are between soil minerals and ions in solution. Two types of mechanisms
will be reviewed here,non-specific site adsorption andspecific site
adsorption. Non-specific site adsorption is the simplest soil-ion
interaction (Nye and Tinker, 1977). Three aspects of the ions and
the soil materials affect this interaction. (1) The sign and amount
of electrical charge carried by the soil mineral and the ion will
affect the behavior of the two together. All soils and most soil
minerals have a net negative charge and the retention by soil of
some anions, such as the chloride ion (Cl') and the nitrate ion (N03)
is almost nil. For this reason, in abiotic soils the two ions behave
almost identically. Because Cl' and N03 move similarly in soils, the
ratio of concentration of the two ions has been used to give an indication

of the movement and loss of NO3 in the soil profile. Soils vary

9

considerably in their amount of charge and the amount of cations
which will be adsorbed varies likewise. In general, divalent ions
are more greatly attracted to the soil than are monovalent ions.

(2) The chemical potential of both the ion and the mineral surface
affect adsorption. The chemical potential of either cannot be
measured directly or even calculated with a great degree of certainty.
Many authors have attempted to formulate a relation between ionic
chemical potential and adsorption similar to the thermodynamic

dissociation constant (Bolt, 1967). One such formula is

 

KN = “'22 “1"). <9)
N (Ao )
where: N+ = fractional amount of monovalent cations adsorbed
N++ = fractional amount of divalent cations adsorbed
Ao+ = activity of monovalent cations in solution
Ao++ = activity of divalent cations in solution ‘

KN = constant.
(3) To a large degree, the rate of adsorption is limited by the rate
of diffusion of the ion to the mineral surface and the geometry of the
soil matrix and the size of the hydrated ion will affect the diffusion
and adsorption of the ion. The ammonium ion (NHZ) has virtually the
same hydrated radius as the potassium ion (K+) and in sterile soil
the two ions behave almost identically (Bohn et al., 1979).

Unlike the interactions of soil cations and anions such as Cl'
and N03, the sorption of P by soil is site specific. Before reviewing
the details of the interactions between soil and P, I would like to

begin this section with a few general notes which point out the special

10

qualities of P both as a nutrient in the soil and a pollutant

in natural waters. Unlike N, no stable gaseous forms of P exist

in nature (Van Nazer, 1973). Thus water is the major carrier of P
from and within the soil. Unlike K, P makes Up only a fraction of
a percent of the earth's crust (Van Nazer, 1958). Further, in those
rocks which bear P, it is usually a minor constituent (McKelvey,
1973). Though P is capable of holding several valences, the quin-
quevalent is the only common form. The group of compounds known

as phosphates, which is any combination of molecules with the

P04.3 ion, make up nearly all natural P. Larsen (1967) in his review
of the soil P literature warned his readers that the words "phosphorus",
meaning the element, and "phosphate", meaning the group of compounds,
are often used interchangeably.

Phosphorus is essential to life; it is a fundamental constituent
of nucleic acids and ATP, which control, respectively, cellular replica-
tion and cellular energy transfer. Given P's scarcity and critical
position in biologic nutrition it is not surprising that small changes
in P concentration can have dramatic changes in biological productivity.
Algal growth in lakes is one example of this change in activity.
Schindler (1976) cited several workers who have shown a correlation
between P concentration in lakes and phytoplankton abundance, "
regardless of the latitude, size or trophic state of water bodies
studied or time period considered." The central role P can play in
proliferation of aquatic plants has made it a pollutant to natural
waters.

Since P loss by soil would damage the soil's ability to sustain

11

a crop and possibly contribute to the deterioration of natural waters,

how well the soil retains the nutrient is very important. Separate

studies have indicated that P is conserved, that is, less P leaves

than is added, whether the land is a farmland (Bargh, 1978), wood-

land (Bargh, 1977) or a sewage disposal site (Kardos and Hook, 1976).
Returning to the model of solute flow in soil (Equation 8)

fn’ the term describing the removal from or release to the solution,

of the solute, in this case P, may be affected by the other forms of

P in the soil matrix. There are three forms of P associated with

soil solids; mineral, organic and adsorbed.

The type of P minerals found in soil is affected by the pH

2 2 3

and the chemical activities in solution of Ca+ , Fe+ , Fe+3 and Al+ .
Most mineral P in unfertilized calcareous soils is apatite (Larsen,
1967). Fertilizer amendments may cause the precipitation of dical-

cium phosphate or octacalcium phosphate, but both of these are

transient and will degenerate into an apatite. The effect of sewage
effluent application upon soil P has hardly been investigated.

Lance (1977) found that the behavior of their soil receiving effluent
was consistent with an adsorption-precipitation reaction of calcium
phosphates, but he did not discover the P mineral precipitated. Another
group went to the trouble to discover the phosphate mineral precipi-
tated in a soil irrigated with sewage. Van Riemsdijk et al. (1978)
studied a soil receiving a sewage effluent for 50 years. Based on

the soil and sewage composition, they expected to find an increase

in soil calcium phosphate. They found, however, that an iron phosphate,

Sterrettite, had precipitated in the soil, probably from P in the
effluent.

12

Phosphate adsorbs to two different soil minerals. Experi-
mental evidence indicates that phosphate may adsorb to either the
surface sesquioxides (iron or aluminum oxides) of many soil minerals
or on surfaces of the calcite mineral. The adsorption of P to the
sesquioxides is more energetic than that to calcite (Holford and
Mattingley, 1975). Surface sesquioxides are found in many crystal-
line and amorphous soil minerals. In a ligand exchange reaction,
one of oxygens of the P04-3 ion displaces one of the hydroxyls
bonded to a surface sesquioxide (Parfitt, 1978). Once this bond
between the sesquioxide and the PO4'3 oxygen is established, a
second oxygen may bind to another sesquioxide, forming a bridging
complex between the P04”3 ion and two sesquioxide atoms on the
surface of the mineral (Parfitt et al., 1975). Apparently phosphate
can also bind by ligand exchange with sesquioxides chelated to the
soil organic matter (Parfitt, 1978).

Apparently the adsorption of phosphate on calcite surfaces
is followed quickly by the precipitation of calcium phosphate crystals.
Images from a scanning electron microscope showed the growth of
apatite crystals on the calcite exposed to phosphate solution (Stumm
and Leckie, 1970). Three steps are seen in the precipitation pro-
cess. (1) With the initial adsorption of phosphate, a monolayer
is formed on the calcite surface. Kuo and Lotse (1972) suggested
that this adsorption was caused by the displacement by the phosphate
ion of hydroxyl ions, water molecules and carbonates adosrbed on the
calcite surface. (2) In time, the adsorbed phosphate tends to

nucleate on the surface. (3) Following nucleation, crystallization

13

and the formation of a calcium phosphate mineral begins.

Since both the sesquioxide and calcite adsorption sites
will be found in soils, the relative abundance of each will deter-
mine the adsorption characteristics of the soil. In the development
of an equation to describe phosphate adsorption of soils, Holford
and Mattingly (1975) assumed the soil had two adsorption sites,
one "high energy" and the other "low energy." They felt the equa-
tion accurately predicted phosphate sorption capacity of soils.

The high energy site, they found, was closely correlated with
extractable Fe while the low energy site was best correlated with
soil calcite.

In most calcareous soils, the equilibrium concentration of
phosphate will be determined by solubility of calcium phosphate
minerals (Lindsay and Moreno, 1960). In most soils, if P has not
been added, the calcium phosphate mineral will be an apatite.

Since equilibrium is rarely achieved in soils, adsorbed P can have
a greater influence upon phosphate concentration in soil solutions.
Murrmann and Peech (1969) argued that the concentration of P in

soils shaken for 108 hr was dependent upon adsorbed, not mineral P.

It is impossible to accurately predict the effect soil
organic P has upon solution P. Surely some of the organic P is
mineralized during the course of a growing season, but it is diffi-
cult to say how much, by what organisms or at what rate (Cosgrove,
1977). A few things can be said with some certainty. Not surpris-
ingly, most of the soil organic P comes originally from plants and

animals. Contributions of organic P from these two sources are

14

rapidly assimilated by soil microorganisms, so that while virtually
all organic P comes from plants and animals the fraction of soil
organic P derived directly from them is quite small. Virtually
all organic P in soil comes after microbial metabolism. The minerali-
zation rate of these soil organic P compounds depends upon their water
solubility, which is usually very low (Cosgrove, 1977).

The chemistry of P in solution is no less complex than it is
in its mineral and organic forms. At neutral pH, solution phosphate
-2

-1 '
and H2P04 (Novozam-

sky and Beek, 1976). They each make up about half of the solution

is composed almost entirely of two ions, HPO4

phosphate. The range of the concentration of phosphate-P in soil
solution is typically 0.1 to 1.0 ppm (Larsen, 1967). Larsen, using
stability constants tabulated by Sillen and Martell (1964), suggested
that while experimental evidence was limited, some solution phosphate

3 +2

is present complexed with a metal ion, such as Fe+2, Fe+ , Ca or

Mg+2.
The soil solute interactions presented thus far have been
studied in soils not subject to water flow. Salt sieving, the
restriction of the passage of ions in solution flowing through soils,
has recently been the object of scientific interest. Salt sieving
is one example of a coupled phenomena. In order to understand salt
sieving, coupled phenomena should be briefly reviewed.
Energy may be dissipated by the movement of either energy or
matter. The flow of energy and matter is across potential gradients.

For example, these equations describe the flux of heat, a diffuser,

electricity and water (Katachalsky, 1963):

where:

J

Q
T
S
D
H
I
E
JV
p

Lnn

15

Q = L]](AT/AS) (10)
JD = L22(AH/AS) (11)
I = L33(AE/AS) (12)
Jv = L44(AP/As) (l3)
flow of heat
temperature
distance

movement of a diffuser

osmotic potential

flow of electrical current
electrostatic potential
flow of water

hydraulic pressure
constant

Equation(lcn is Fourier's Law, equation (11) is Fick's Law, equation

(12) is Ohm's Law and equation (13) is Darcy's Law.

For reasons of brevity that will be clear later, equations(10)

- (13) will

where Jn is

gradient.

be written:

L F

I
r
-n

a flux of either mass or energy and Fn is the appropriate

In 1808, Rous noted that an electrical potential applied to a

pair of porous electrodes strattling a water saturated clay would induce

the flow of water. His interest piqued, Rous applied a hydraulic

16

pressure gradient across the same system and the flow of electrical
current began. The combination of these two phenomena is known as
electrokinetic coupling. Darcy's Law and Ohm's Law do not completely
describe this sytem. The fluxes of electricity and water are coupled
and can be described, using the symbols from equations (10) - (13), by:
J

3 L33F3 I L34F4 (14)

J4 = L43F3 + L44F4 (15)

Similarly, the Seebeck and Peltier effects demonstrate, respectively,
the effects of temperature gradients on the flow of electrical current
and an electrical potential gradient upon heat flow. There are many
examples of coupled phenomena (Katalchatsky and Curran, 1965). The
two equations (14) and (15) which describe the electrokinetic effect,
can be written in matrix form:

J3 = L33L34 F3

J4 L43'44 F4 (16)
In 1931 Lars Onsager asserted that, for a set of coupled phenomena
for which the formulas are derived correctly, the matrix of L coefficients
is symmetric. For the electrokinetic case above, this means that
L43 2 L34°

The validity of this approach has been demonstrated repeatedly for a
variety of coupled phenomena (Miller, 1960).

As was stated earlier, salt sieving is a coupled phenomena
(Groenevelt and Bolt, 1969). In the terminology of coupled phenomena,
the important interactions in ion flow through soil can be represented

as:

17

Jv = LV Fv + LVD FD + LvE FE (17)
JD = LDv Fv + LD FD + LDE FE (18)
I = LEv Fv + LED FD + LE FE . (19)

where: = flux of water

J

JD = flux of salt relative to the flux of water
I = flux of electricity
F = -(AP/As)
F = -(AH/As)
F = (AE'/As)

AH = RTAC
= hydraulic pressure

= electrical potential
= distance

p

C = concentration

E

s
Groenevelt et al. (1978) calculated theoretical values for the L
coefficients to describe the flow of water, ions (Na+ and Cl') and
electrical current through a clay membrane. The flow of current was
restricted unless a wire connects one face of the membrane with the
other. With the wire absent, there was no current and I = 0. At
steady state flow, ionic balance was maintained and AE' = FE = O
(Nielsen et al., 1972). Using these two conditions, we can calculate
the reflection coefficient, 0, which is defined as the ratio of the
flux of ions relative to that of water. For the situation where

I = FE = O, the reflection coefficient is:

JD
0 = T: and (20)
V

 

ED
0 = LDv - l:Ev + LD (2])
L
ED
L - -——-+ L
v LEV vD

Using the values for the coefficients provided by Groenevelt et a1.

(1978) the value of the reflection coefficient may be determined:
3.70 8

 

 

o _ (6.20 x 10"?) - ‘710§'+ 2.39 x 10' _
' -10 3 70 4T0 ‘ ‘°°54 (22)
(1.09 x 10 ) - '105 + 6.20 x 10

This value for the reflection coefficient does not mean that the
concentration of salt on the effluent side of membrane is constant,
rather it means that the concentration drops as the rate of water
flow increases.

If the osmotic gradient is very small or the ions under
consideration are a small fraction of all ions in solution, we may
assume AH = FD = O. In this case, the reflection coefficient is:

-10

J .
ID = 5.69 (23)

o = _Q_= LDv = 6.20 x 10
Jv Lv 1.09 x 10'

 

 

which indicates that the ions under consideration will increase in
concentration as flow increases.

The experimental studies of salt sieving show that the effects
of hydraulic pressure vary depending on experimental conditions.
Kemper reviewed his own work in a collection of articles edited by
Nielsen et a1. (1972). Kemper forced .01 N NaCl solutions through clay
membrances with hydraulic pressure gradients of 5, 10 and 17 bars. He
found that the ionic concentration in the effluent decreased as the

pressure increased, which is predicted qualitatively by equation (22).

19

He noted that in investigations where higher pressure gradients were
used, the effluent ion concentrations increased rather than decreased
with increasing pressures.

Such a study is one performed by Kharaka and Smalley (1976)
who studied the effluent concentration in ion solutions of several
salts forced by pressures ranging from 7 to 180 bars. While the flow
of ions in all cases was restricted relative to the flow of water the
data showed that the effluent concentration of most types of ions
increased with increasing pressure. The authors argued that the
increased hydraulic drag on the hydrated ions at higher flow rates
explained the behavior. If this was the case, then LvD was dependent
upon Jv and, therefore, was dependent upon Lv, a violation of premises
of the general non-equilibrium model that the coefficients are inde-
pendent of the gradients (Groenevelt, 1971). Nonetheless, whether
the nonequilibrium model applies in all cases or not, the theory
and experimental evidence indicate that the concentration of ions
passing through soil is affected by the hydraulic pressure gradient.
Furthermore, whether the ionic concentration of the effluent increases
or decreases depends on the nature of the soil, hydraulic gradient,

ionic solution and water velocity.

Flow of Water and Ions to Tile Drains
What particular investigators attempt to understand about the
flow of water through soil to tile drains depends on the purpose of
the tile and the nature of the investigation. Soils are drained
for two reasons: (1) If a soil has poor natural drainage, a tile

system may be installed to drain the soil, lower the water table

20

and allow better root development by the crop; and (2) a sometimes
congruent purpose of tile drainage, especially in arid areas, is to
facilitate salt removal from the surface of the soil. Tile drains
are also studied for reasons other than those relating directly to
their agricultural use. One may want to study tile drainage because
the solution flowing in tiles drains may contribute to surface
water pollution. Phosphorous is a potential pollutant. The concen-
tration of P in soil solution decreases as its path length through
the soil increases (Reneau and Pettry, 1976). Ryden et a1. (1973)
suggested that tiles may increase the concentration in surface
waters because the path length through soil is shortened considerably.
In addition, many researchers have used the concentration of ions in
tile water as an estimate of the nature of the soil solution passing
through the root zone to the groundwater. The indications from the
literature on the efficacy of sampling tile water for this purpose
are contrary. Ivanov et al. (1977) wrote, "Drainage water is found
to reflect clearly the changes in the soil." Thomas and Barfield
(1974), however, contended that tile drainage water is an unreliable
indicator of N05 loss by leaching. Two articles (Sharpley et al.,
1977 and Karlen et al., 1976) have reported wide and apparently flow
dependent changes in the concentration of P in tile water. If only
because of convenience, the practice of using tile drainage water as
representative of the leaching water will probably continue.

Several groups of investigators have been forced to examine
limited aspects of tile flow. Since the answers for each of these

groups come only with great ingenuity and effort, it is natural that

21

knowledge about water flow to drain tiles is not comprehensive.
Many questions that would seem to be important are unanswered.
How much of the water which is applied finds it way to the tile?
What is the path of water to the drain? What is the nature of the
soil through which the water passes, saturated or unsaturated,
oxidized or reduced? How do macropores affect the behavior of
water flow? What are the dynamics of water flow in the tile drains
following irrigation or rain? On these questions, the literature
is either silent or equivocal. Most of what is known about water
in tile drained fields comes from three areas: (1) field observa-
tions, (2) mathematical analysis and (3) physical analogues.

Four things have been learned from field observations.
(1) In tiled soils the water table tends to be lowered. (2) The
water table in these soils is curved, with the maximum height at
the midpoint between two tiles and the lowest height directly above
the tiles (Engelund, 1957). (3) In fact, unless the tile is mal-
functioning, the water table above the tile is not higher than the
top of the tile (Luthin and Haig, 1972). (4) Recently, Meek et al.
(1978) analyzed the ionic and gaseous constitutents of tile water
entering from various points around the circumference of the tile.
They found, not surprisingly, that water entering the tile from above
reflected a more aerobic environment than the water entering the
tile from below.

In order to understand the possible chemical fluxes to the
tile water from the soil a dynamic picture of the contributions from

the various portions of the soil should be developed. Neither of

22

the two most widely used mathematical approaches gives a compre-
hensive picture. Since both approaches arrive at an analytic
solution which assumes an impermeable barrier below the tile,

theory requires all the water entering the surface must flow out

the tile, an assumption that is not always satisfied. The two
theories of water flow that have found wide use are potential theory,
based upon the applicability Laplace's equation, and Dupuit-
Forchheimer (D-F) theory, based upon Darcy's Law and the assumptions
that flow along all streamlines is horizontal and independent of
distance from the water table. While solutions of the D-F

theory of water flow to tiles have been capable of predicting
transient flow, the approach was incapable of describing realisti-
cally the path of water, the streamlines, to tile drains (Kirkham
and Powers, 1972). Solutions based upon potential theory suffer
from an opposite setback. While the streamlines are correct, the
approach has been incapable of analytically predicting transient
behavior.

Transient behavior in water flow to tiles has been approxi-
mated using potential theory by a series of steady states. This
approach has been found acceptable by Childs (1947), Kirkham (1964)
and Van Schilfgaarde (1974). Using a Fourier expansion of Laplace's
equation, Kirkham (1958) was able to solve analytically the stream-
line and hydraulic potential distribution for a tile drained soil.

A drawing of this distribution is presented in Figure 1. In a later
article, Kirkham (1964) assumed that the fall of the water table after

the cessation of water application was identical to a series of water

23

.Ammmp .Emsxewx Lmuw<v Fwom emcpmcc my?“ a cw mmcw— —mwpcmpoaw=cm
can mwchEmwcam mo cowusnwgumwu ~muwumcomsp .F weaned

one; Fmrpcmpoamzcm
weepsmmcpm

 

 

I _ u
I ’ a — m

a
/
I I \’\\4‘
1 I: z I

x -

~
Li \
I/x/
l/e
. mp

 

‘p-D‘

  
   

{. _ —

 
  

 

 

we
M ~23 emu—m3

 

“m. mummgsm pwom

24

table heights associated with ever decreasing water application rates.
The results of this method indicate that water flow with streamlines
originating furthest from the tile always make up the greatest propor-
tion of the tile flow. Because the height of the water table near the
tile is lower, the proportion of water flow from streamlines originating
nearest the tile, though starting high, quickly dissipate.

As with Kirkham, Childs, in an earlier article (1947), assumed
that a series of steady state solutions would simulate the falling water
table. As was discussed in the review of coupled phenomena,

Ohm's Law, which describes the flow of electricity, has exactly the

same mathematical form as Darcy's Law, which describes the flow of
fluids through porous media. Using this fact, Childs (1943) constructed
several electrical analogues of cross sections of tile drained soils.
This method gave the first precise description of the flow net of stream-
lines and equipotential lines about a tile drain. Because Childs

assumed a much higher water table initially, his results were contrary
to Kirkham's. Childs found that the first flow of water to the tile
after the cessation of water application was dominated by contributions
of streamlines originating close to the tile.

In general, laboratory models of tile drainage systems have not
been used to understand transient behavior of water flow to tile drains.
Rather, the method has been used to evaluate various materials and
methods which would be difficult to study in the field. A recent
article by Luthin and Haig (1972), for example, examined the effect of
tile diameter, the placement of holes in the tile and water table height

upon tile flow.

25

Three articles have found peculiarities in the concentrations
of ions in water flowing from tile drains. Thomas and Barfield (1974)
have suggested that the contribution of NOS-N from a tile drain to a
drainage ditch was many times greater than that from any other source
in a Kentucky field. They concluded that estimate of leachate loads
based upon the concentration of N03 in tile drains could be greatly
in error. Two articles have shown variations of the P concentration
in tile drainage water. Both Sharpley et al. (1977) and Karlen et a1.
(1976) observed that P concentration water from tile drains increased
with flow rate. Each presented a different explanation. Sharpley
et a1. stated that most of the P in the water was associated with
particulate matter and implied the increase in P concentration was due
to increased suspension of particulate matter in the tile drainage
water with increased flow. Karlen et al., citing great heterogeneity
in the soil at the experimental site, suggested that faster flow of
water through sand smears in the soil profile caused the peaks in the
P concentration.

Thomas and Barfield argued that denitrification was less in
water flowing to tile drains because the soil in the region of the
tile was aerobic. At two times in the 1972 season, Thomas and Bar-
field measured the flow rate and N03 concentration in tile drains
and a ditch draining the experimental site. The authors observed only
two sources of water flowing to the ditch from the field, subsurface
drainage and tile drainage. The N03 concentration and volume of sub-
surface drainage water was calculated from the difference of both the

N03 and water loads from the ditch and the tile drains. Their results

26

seem to show that while the water from the tile was a minor con-
stituent of water flowing in the ditch, most of the N03 in the tile
drainage was many times greater than the N03 concentration measured

in the ditch and that calculated for the subsurface flow water. The
results suggested to the authors that water flowed to the tile drains
through portions of the soil that were mainly aerobic. The

N03 in this water was not lost by denitrification. Water flowing

to the drainage ditch by subsurface drainage entirely through the

soil passed through a more anaerobic environment conducive to denitrifi-
cation.

Thomas and Barfield attempted to show the uncertainty of relying
on tile concentrations of N03 for estimates of leachate concentration.
Results by other authors question whether changes in denitrification
rates were the cause of the differences in the N03 concentrations.

(1) Almost certainly much of the NOS flowing in the ditch was removed
from solution. A report on N03 pollution by the National Research
Council (1978) stated that unlike N03 ir1 groundwater, NO} in surface
waters is rapidly consumed by aquatic plants. (2) The field measure-
ments by Meek et a1. (1978) showed that much of water flowing into the
tile comes from reduced portions of the soil. (3) Denitrification

has been shown in aerobic soils in both the field and the laboratory
(Greenwood, 1961; Allison, 1966). (4) As noted by Thomas and Barfield,
anaerobiosis is only one of the environmental determinants of microbial
denitrification. The other main determinant is the supply of that
organic C which can be metabolized by the denitrifying bacteria. Most

of this C is located close to the soil surface. The concentration of

27

organic C at the depth of a typical tile drain is practically nil.

Thus, while Thomas and Barfield have suggested disquieting incon-

sistencies in the concentration of N03 in tile drainage water, they
were unable to adequately demonstrate the behavior or conclusively

explain its cause.

Two groups of investigators have found anomolies in the tile
drainage water P concentration. The soil at the experimental site
used by Karlen et a1. (1976) was very heterogenous. The concentration
of P in the tile drainage water varied greatly. A figure presented
by the authors showed a maximum of 0.75 ppm P and a minimum of 0.13
ppm during a 100 hour period. The maximum concentration was roughly
coincident with maximum flow following an irrigation. The authors
hypothesized the existence of sand smears in the soil profile. Some
of the irrigation water, they argued, flowed rapidly through the
smears. They assumed that little or none of the solution P was adsorbed
to the soil in the smear. The P peak was caused, they suggested, by
the rapid and unencumbered flow of P to the tile.

Sharpley et al. (1977) observed roughly the same behavior at
a New Zealand experimental site. The soil they used was much different
from the one used by Karlen and his coworkers. Because it often
indicates the relative abundance of sesquioxides and calcite, pH is
an important indicator of the amount and type of P adsorption in
soils. The soil at the site used by Karlen et al., a Conover 10am,
was derived from calcareous till and had a neutral to alkaline pH.

The pH of the soil used by Sharpley et al., a Tokomaru silt loam,

was low, normally between 5.2 and 5.7. It is likely that the two

28

soils had quite different P adsorption properties. Whereas, the
Conover soil used by Karlen et al. was very heterogenous. the Tokomaru
silt 10am, which was derived from a coarse loess was very homogenous
(During, 1972). In fact the group in New Zealand reported that the
entire experimental area, covering 112 ha, was all the same series.

Sharpley et al. examined the behavior of P in the tile in
more detail than did Karlen and his coworkers. The object of the
New Zealand study was to estimate how frequently an investigator
should sample the water in streams, subsurface drainage, tile drains
and surface runoff in order to get an accurate estimate of P loading.
Three types of P from all these sources were monitored: dissolved
inorganic P (DIP), total dissolved P (TDP) and total P (TP). DIP
and TOP were that P remaining after a sample was passed through a
0.45 pm filter. Particulate P was estimated by the difference between
TP and TOP.

The results presented in graphs in the article by Sharpley et
al. showed that TP varied with flow in streams, tile drains and sur-
face runoff. The difference between the maximum and the minimum con-
centration of TP was greatest in the tile drainage water. The
behavior of the TP concentration during a flow event was similar in
water sampled from tile drains and streams, where the highest concen-
tration of TP was at or near the maximum flow. Based on these results,
Sharpley et a1. argued that during maximum flow turbulence would
also be highest. Turbulent flow would bring particulate matter into
suspension. Much of the P and thus much of the increase in P concentra-

tion during rising flow rates, they argued, must be associated with newly

29

suspended particulate material. The behavior TP in streams and

the tile drains was not identical. Unlike the TP concentration in
water from the stream, the peak TP concentration and the peak water
flow in the tile drains did not coincide. The peak TP (and DIP)
concentration in the tile drainage water led the peak flow rate

consistently by about 1.5 hours.

Nitrogen Transformations in Soil

 

Five transformations in the N cycle may have an effect upon
N applied to croplands in sewage: (l) immobilization of inorganic

N, (2) mineralization of organic N, (3) volatilization of ammonia

+
4

N03. Volatilization of NH3 from water, and by implication from the

(NH3), (4) nitrification of NH to N03 and (5) denitrification of
soil solution, is controlled by pH. The ammonium (NHZ) stays in
solution. The loss of one proton from this ion forms NH3 which is
readily volatilized. The conversion of NH: to NH3 is fostered by
alkaline conditions. The rate and absolute amount of loss are
increased by increases in temperature. The collodial property of
soil limits volatilization of NH3. Losses by NH3 volatilization have
been reported from the water above rice paddy soils, from manures in
feed lots (Lauer at al., 1976), from sewage in sludge ammended soils
(King, 1973) and from plant surfaces (Denmead et al., 1974). Once
the NH3 source is incorporated into the soil, losses are lessened.
There is little possibility that NH3 volatilization losses are large
from sewage irrigated soils once the wastewater has entered the soil.
Mineralization is the conversion of organic N to inorganic

forms, usually NHZ. The organisms which mineralize organic N are

30

so numerous and diverse that the process probably occurs under
virtually all environmental conditions. By far the largest portion
of soil N is the organic fraction, indicating that in most soils the
process is slow. For reasons which are not clear, mineralization
of organic N is increased by alternate wetting and drying. The
addition of inorganic N also accelerates the breakdown of humus,
probably by stimulating microbial activity in general (Alexander,
1977). Apparently organic N added to soil as sewage is rapidly
mineralized. Both Lance and Whisler (1972) and Broadbent et al.
(1977) found that virtually all organic N in wastewater applied to
soil columns was mineralized.

In many secondary wastewaters,organic N is a minor constituent.
For example, 4% of the total N in wastewater used in a study by Lance
and Whisler (1972) was organic. Wastewater application may spur

15

mineralization of soil organic N. For example, using N Broadbent

et al. (1977) found that as much as half of the N03 leaving a soil
column continuously amended with simulated wastewater was derived from
soil organic N. Apparently this additional Nog-N came from organic

N which had been mineralized,then nitrified.

Immobilization of soil inorganic N is the opposite of minerali-
zation. Immobilization is the assimilation of inorganic N into
organic molecules as constituents of organisms. The balance in soil
between mineralization and immobilization is affected by the ratio
of organic C to organic N. Anything which increases the ratio, such

as the addition of organic C, tends to spur N immobilization while a

decrease in the ratio may bring about an increase in mineralization

31

(Alexander, 1977).
Nitrification, the conversion of NH: to N03, is also thought
to be complete in soils amended with secondary wastewaters (Broad-
bent et al., 1977). Nitrification is an aerobic process. Lance
and Whisler (1972) attributed the success of the process to the
ability of the soil to adsorb NH: during saturation. When a soil
was irrigated with wastewater and soil was saturated, the NH: was
retained by the soil exchange complex. When it drained, the soil
became aerobic again, and the sorbed NH: was nitrified.
Denitrification is the reduction of N03 to gaseous N. Bacteria
are the only organisms which have been shown to have the capacity to
denitrify. Fifteen genera of bacteria are capable of denitrification
(Payne, 1973). The bacteria are obligate aerobes which use N03 as
an oxidant when 02 is unavailable. Most of these bacteria are hetero-
trophs which use N03 as the terminal acceptor of electrons donated
by substrate organic C. Nitrate N is not assimilated in the denitri-
fication pathway. Denitrifying bacteria can be found in abundance
in most soils. The extent of denitrification in soils is to a large
degree dependent on the soil environment. In laboratory experiments
denitrification is favored by the presence of organic C, the absence
of 02, saturation by water, elevated temperatures and a neutral to
slightly alkaline pH (Nommik, 1956). Water saturation stimulates
denitrification because it severely restricts the diffusion of O2
in the soil. The 02 concentration in saturated soils is rapidly
depleted by biological respiration and oxidation of reductants (Howeler

and Bouldin, 1971).

32

Anoxia, the absence of 02, does not have to be complete in
soils for denitrification to occur. In a review of N balance studies
in soils typically below field capacity, Allison (1966) found un-
accounted losses which usually amounted to about 15% of applied N.

He attributed much of these losses to denitrification. The energetics
of electron transfer in the oxidation of organic C may help to explain
why denitrification can proceed in aerobic soils. The energy yields
from the transfer of one mole of electrons from organic C to the four
major inorganic oxidants are: to 02, 26.5 kcal; to N03, 18 kcal; to
504’2, 3.4 kcal; to 002, 2.4 kcal (Payne, 1973). By this approach
the energy yield of electron transfer to N03 is 68% as energetic as
that to 02. Another explanation for denitrification in aerobic soils
has been developed separately by Currie (1961) and Greenwood (1961).
They suggested that the restriction of O2 diffusion in soil aggre-
gates would create anoxic sites in the center of the aggregates where
denitrification could occur. The diffusion of 02 from the surface
through the soil is a factor in denitrification. Since 02 moves
primarily by diffusion in soil, the concentration is lower further
from the surface. In a simple experiment Jones (1974) measured the
N03 lost after 28 days at two depths in'a soil maintained at 90%
saturation. The soil section between the surface and 7.5 cm depth
lost no N03, while the section between 7.5 and 17.5 cm lost between
73% and 83% of the N03 present at the beginning of the experiment.

While anoxia may not restrict the ability of denitrifiers to
metabolize organictlsubstrate, anoxia can restrict the capacity of

microorganisms to transform complex organic compounds to a form which

33

denitrifers can readily use. Denitrifiers, like most heterotrophs,
rapidly metabolize simple carbohydrates. "Readily mineralizable"
organic C is that C quickly oxidized by aerobic heterotrophs. The
denitrification capacity of soils is strongly correlated with both
"glucose equivalent" organic C and "readily mineralizable" organic
C (Burford and Bremner, 1975). Some bonds in organic C can only be
broken under aerobic conditions (Focht and Verstraete, 1977). Perhaps
what limits denitrification is not organic C but forms of organic C
that denitrifiers can use under continuously anoxic conditions.
Denitrification is enhanced by the same things which increase
the amounts of simple carbohydrates in soil. A simple though inelegant
method to increase the carbohydrate concentration and denitrification is
to add glucose or methanol to the soil (Kohl et al., 1976; Lance et al.,
1976). Cycles of wetting and drying can also increase denitrification.
Reddy and Patrick (1975) found that the most frequent cycle of wetting
and drying removed the greatest amount of nitrogen by denitrification.
Apparently, this procedure allows enough aeration to break down organic
C to a more mineralizable form while preventing the 02 content from
rising high enough to inhibit denitrification. The rhizosphere, that
volume of soil immediately adjacent to the plant root, has higher
denitrification rates than the bulk soil. Woldendorp (1962), the
first author to investigate the phenomena in detail, concluded that
the stimulation of denitrification was due to the consumption of 02
around the root and the exudation of carbohydrates by the root. Studies
have shown considerable organic C exuded by roots. Barber and

Martin (1976) measured the C02 evolved from planted pots containing

34

sterile and non-sterile soil. The difference in C02 evolved was taken
to be due to the root exudation of organic C. The results indicated
that root exudation was equal to 18-25% of the total dry matter accu-
mulation by the plant. Using a different method, Johnen and Sauer-
beck (1977) estimated the annual production of carbonaceous root
exudate. They reported that in one year a hectare of wheat would
exudate some 650,000 kg of organic C. Palpably, the addition of
large quantities of organic C would stimulate heterotrophic micro-
bial activity, including denitrification if no other factor is limiting.
Many investigators have attempted to elicit the relationship
between plants and denitrification. The results have not been con-
sistent due, perhaps, to a lack of uniformity both in the plants
studied and in the methods used to assay denitrification. Garcia
(1975), measuring the activity of N20 reductase in pot soils, both
planted in rice and not planted, found greater denitrification in
the planted soils. He concluded the enhancement was due to the
development of an anaerobic zone about the roots, the presence of
root exudates and the presence of larger numbers of denitrifiers in
the rhizosphere. A study by Brar (1972) with excised roots found
that denitrification measured by N gas evolution was greater with
than without roots. He concluded that denitrification was enhanced
in the rhizosphere because the concentration of 02 was lowered near
the root. Volz et a1. (1976) studied two plots, one planted in

barley (Hordeum vulgare L.), the other fallow. The plots were

 

irrigated with water containing 100 ppm NOS-N and 46 ppm Cl'. At

various depths in the soil, they measured the concentration of glucose,

35

NOE-N and Cl' as well as the redox potential. They found that the
NOS-N loss was greater in the planted plot. They concluded that the
organic C exuded by the roots moved with the N03 in soil solution.
This organic C, they thought, stimulated denitrification by accelerating
microbial activity and reducing the redox potential.

Investigators measuring denitrification in soils containing
corn (Zea mays L.) roots have been less successful showing an enhance-
ment of denitrification. Bailey (1976) measuredlfllg loss and N gas
evolution in microcylinders containing soil with and without corn
seedling roots. His data showed no significant increase in denitrifi-
cation in the microcylinder with the corn seedling root. Patriguin
et al. (1978) measured denitrification by acetylene inhibition in
soils with and without corn roots. Denitrification was greatest before
the plant roots began to grow rapidly. After the plant root began to
grow denitrification dropped. They concluded that the roots maintained
the N03 concentration below that at which the bacteria could denitrify.

Management of Irrigated Soils
to Maximize N Removal

Nitrate pollution of the groundwater may have several undesirable
effects. (1) The presence of N03 in the groundwater will almost un-
doubtedly contribute to the N load of surface waters, with the atten-
dant problem of increased aquatic plant productivity. (2) Recently,
nitrosamines have been shown to be carcinogens. While the biochemical
connection between N03 and nitrosamines may be tenuous, until more is
known, it cannot be discounted (National Research Council, 1978).

(3) Nitrate has been shown to cause methemoglobinemia in human infants.

36

This sometimes fatal blood disorder has been caused by using well
water high in N03 to make infant formula (Lee, 1970).

Four factors have been shown to be important in decreasing N
content in water leached from cropland; the crop, the soil, irriga-
tion scheduling and, if the soil is sewage irrigated, chemical amend-
ments. Many authors have pointed out that some forage crops maintain
a low N03 concentration in the soil. Most researchers have attributed
the low concentration to either plant uptake or changes in the behavior
of the soil microorganisms. MacLean (1977) studied the uptake of

fertilizer“NO3-by.corn, timothy (Phleum pretense L.), bromegrass

 

(Bromus inermis L.) and orchardgrass (Dactylis glomerata L.). The above

 

ground portion of the corn crop contained less of the applied N than
any of the grasses. At the highest fertilization rate, 448 kg/ha, the
corn removed an average 19% of the applied N while bromegrass removed
43%. Warncke and Barber (1974) estimated the rate of N uptake by
roots of bromegrass, corn, sorghum (Sorghum bicolor L.) and soybean

(Glycine max L.). They also measured the minimum N03 concentration

 

maintained in solution about the roots of these plants. Bromegrass
roots maintained a lower N05 concentration than corn or any of the
other plants. Overman and Nguy (1975) studied the N uptake over a
season by corn, sorghum-sudan grass (Sorghum vulgare Pers. X Sorghum
sundanese Stapf) and kenaf (Hybiscus cannabinus L.) in a soil irrigated
at high rates with sewage effluent. While the N uptake of kenaf and
sorghum-sudan grass dropped rapidly after 50 days, the uptake of

corn remained consistent during the growing season.

Since early in this century, many authors have reported that

37

the concentration of N03 was less under grasses and forage crops
than that under row crops (Lyon et al., 1923; Kolenbrander, 1969;
Zwerman et al., 1972; Low, 1973). Lately, this fact has been used
to reduce the N03 leaching under irrigated soils. Avinimelech and
his co-workers (1978) found that No; pollution of the groundwater
was reduced under irrigated Rhodes grass (Chloris gayana Kunth)

and alfalfa (Medicago sativa L.). They speculated that the low
N03 concentration under these forage crops was due to denitrification
induced by low 02 tensions found under these crops. Levin and
Leshem (1978) concluded that the reduced N03 concentration under
forages was due to nitrification inhibition. Feigin et a1. (1978)
also found the N03 concentration lower under an irrigated Rhodes
grass sward. They thought that the grass had stimulated denitrifi-
cation.

The first explanation for the depression of N05 concentration
under grasses to receive much attention was nitrification inhibition.
Theron (1950) originally suggested that grass roots somehow acted
to inhibit the conversions of NH: to N03 in soils. The work of Munroe
(1976), Moore and Waid (1971) and Rice (1976) has shown that nitrifi-
cation was depressed in the presence of grass roots. This explanation
has been attacked, sometimes vituperously (Purchase, 1974), by
several authors including Theron himself. With ingeneousness rare
in scientific literature, Theron, in a later article (1963), repudia-
ted his own theory, and suggested the mechanism which depressed the
soil N03 concentration under grass was an inhibition of net soil

mineralization. This explanation has also been suggested by Huntjens

38

(1971a, 1971b) and by Kissel and Smith (1978).

Several aspects of the soil can affect the reduction of the
N03 concentration in irrigated lands. Avnimelech et a1. (1978)
found that soils high in clay reduced N03 leaching, presumably by
enhanced denitrification. Lance et a1. (1976) found that the
concentration of N03 in the soil leachate was inversely related to
infiltration rate of the soil. Lance and Whisler (1972) claimed
that NH: adsorption was important to facilitate nitrification,
though Broadbent et al. (1977) stated that nitrification in sewage
amended soils was almost always complete.

Results on the effect of irrigation scheduling upon soil
N05 are contradictory. Reddy and Patrick (1975) found that frequent
flooding (every 2 days) maximized N loss by denitrification. Lance

and Whisler (1972) found that while all of the NH:4

and organic N

was transformed to N03 during a similar schedule, the maximum N

loss was found in soils flooded for longer periods, up to 23 days.
Apparently the amount and quality of organic C found in

secondary sewage effluent is insufficient to denitrify much of the

N applied to the soil. Lance and Whisler (1976) found that the

addition of glucose or methanol to the effluent before irrigation

would increase N loss. Lance et a1. (1976) found that mixing sewage

effluent with high N03 soil filtrate stimulated denitrification of
the filtrate in the soil.

CHAPTER II

MATERIALS AND METHODS

Introduction

 

This report details an investigation made as part of an
ongoing study of the irrigation of cropland with stimulated secondary
wastewater effluent. This investigation, begun in 1973 and ended in
1978, was supported by the Michigan State University Agriculture
Experiment Station with special funding from the Michigan Legis-
lature. Descriptions of the experimental site for the study and
the initial results may be found in Karlen (1975) and Karlen et a1.
(1976). Here I report the data collected during the 1977 and 1978
growing seasons. While some experiments spurred by field observa-
tions were conducted in the laboratory virtually all the information
came from samples or data collected at the experimental site. Since
the project spanned several years, the experiment changed as time
progressed. As more was learned about the response of the soil and
plants to irrigation, the crops, the irrigation rate and the chemical
application rates were changed appropriately. In addition, from
year to year different types of data were collected in the field
as questions were resolved and new lines of inquiry were developed.
Because the experiment changed with time, the description of the
materials and methods may be confusing. To avoid this possibility,
it will be noted in the text when a procedure was done and why a new

one was initiated. A brief explanation will be given at that time

39

40

for these transitions.

Description of the Site

Size, Relief and Soil

 

The same experimental site was used in all five years of the
wastewater study. The site was 0.68 ha block, 185.3 m long by 36.6
m wide. The longer dimension was oriented directly east-west. In
the morning, when sprinkler irrigation commenced, the winds were
generally from the southwest. If the wind velocity exceeded 24 kil-
ometers per hour, the drifting of water became excessive and the
irrigation was stopped. The site was slightly rolling. The highest
point, running most of the width of the plot, was located in the
center of the plot. The maximum change in relief was less than 3
meters over the whole site. Even though considerable amounts of
water were applied to the site runoff and erosion were observed
infrequently. The soil was heterogeneous. Nominally, the soil
at the experimental site was of the Conover series, a fine loamy
mixed mesic Udollic ochraqualf. This soil was formed from calcareous
till. The Ca content in the soil at the site was very high
throughout the profile. Soils of the Conover series are somewhat
poorly drained. In 1970, the experimental area was tile drained
at a depth of 1.05 m. The tiles ran along the width of the plot

and were spaced every 15 m.

Plots and Subplots
Throughout the simulated wastewater experiment, four treatments

were always in effect. In 1973 and 1974, the entire area was planted

 

41

in corn which was irrigated at four rates; 25 cm, 50 cm, 100 cm
and 200 cm annually. Little or no potential contamination of the
groundwater was found at the 25 and 50 cm rates. Beginning with
the 1975 season and continuing to the end of the experiment, two
crops, corn and bromegrass, were irrigated at two rates, 100 and 200
cm/year. There were 12 plots located at the experimental site (see
Figure 2). Each treatment was conducted on three adjacent plots.
Prior to 1973, a bromegrass sod was established at the experi-
mental site. In the spring of that year, a preplant mixture of
paraquat and atrazine was applied to the experimental area and the
entire site was planted in corn. In 1974, the site was again planted
with corn. For these two seasons, plots 1, 2 and 3 were irrigated
at 25 cm/year, plots 4, 5 and 6 received 200 cm/year, plots 7, 8 and
9 received 100 cm/year while plots 10, 11 and 12 were irrigated at
50 cm/year. In 1975, the plots 1, 2, 3, 10, 11 and 12 were planted
with bromegrass. Plots 1, 2, 3 received 100 cm/year. Plots 4, 5
and 6 continued in corn and continued to receive 200 cm/year in
simulated secondary effluent. Plots 7, 8 and 9, in corn, again
received 100 cm/year. Plots 10, 11 and 12, while in bromegrass,
received 200 cm/year. Each year corn, but not bromegrass, received a
starter fertilization consisting of 37 kg/ha N, 18 kg/ha P and 22 kg/
ha K. In addition, all twelve plots were divided into four subplots
to test the efficacy of additional K and N fertilization on the yield
of the two crops. In each plot, two subplots received 112 kg/ha of
N fertilizer in addition to any starter fertilization; the other

two subplots received no additional N. Similarly, two subplots in

 

42

 

E oém

 

.wmcm qucmemcmaxm ucmapemm FQQwowcze acmvcoumm umpmpzewm .N mcamwd

 

NF

: o— m w m o m e m N —
a a a a a n. n. a n. n. a me
. . _ . . . . . . _ .
. . . . . _ . . . . . .
u o _ o u .1 o ” AH m A. o _ o q . q _ o .
_ . . . . . _ . _ . u
_ _ u . _ _ . . . _ .
. _ . . . . _ _ . . . .
. _ - _ . . _ . — - —
. _ . . . _ . _ . . . .
_ _ — _ - . . _ . - - —
_ Eu. . _ . . . . . . . .
. com " . Ea so. u . so com . _ as co. "
. o o o o o . o o a
. mm<¢e . " zeou " " zeOu " . mm<¢u "
. _ . .
. . . u . . u . u . _ .
. . . _ . . . . _ . .
. . . _ . . n . . _ .
. _ _ . . . . . . . .
. . _ . _ . _ _ _ . .
- - . . . - . _ — . . —
. _ . . . _ . . . . . _
_ . _ L . F _ . L b . L

 

 

 

 

 

 

 

game.
ee_uaom meeeaupea: nu

e_aao a__e .1..-
__a3 me_PaEam a

 

43

each plot, one which received additional N, one which had not,
received an additional fertilization of 93 kg/ha K. The other

two subplots received no additional K.

Tile Lines

 

As was alluded to earlier, the tiles, in addition to draining
the site, were also used to estimate leaching losses of ions and
water. Tile lines were located below the center of each plot at a
depth of 1.05 m. Before being intercepted by a major tile which
carried the water off the site each tile line passed through a
subsurface monitoring station. In the station, the tile line was
exposed to reveal a special plexiglass tile section with two features.
(1) A stoppered port in the plexiglass tile section allowed samples
to be taken of the tile water. These samples were removed to the
laboratory for chemical analysis. (2) The tile section contained
a 600 weir, which allowed the water flow rate to be estimated
by measuring the height of water behind the weir. The height of
the water in the tile behind the weir was measured continuously
by a Stevens type F, Model 61 water stage recorder. Each weir

was calibrated individually by regression to one of two formulas:

in 1977: F = ahb (24)
in 1978: F = (c + dh)5/2 (25)
where: F = water flow rate (liters/min)

h

height of water behind the weir
a,b,c,d = constants.

Equation 25 described the increase in flow with tile water

 

44

height more accurately than did Equation 24. The correlation co-
efficient between flow rate and the height of water behind the weir

2

using Equation 25 was r = 0.998 or better. Qualitatively, the fit

(Figure 3) was excellent.

SamplingWells

 

Since the concentration of N03 in the tile drainage water
may have been a poor indicator of leaching losses, 26 sampling wells
were established on the perimeter and at various locations within the
plot areas in 1975. An estimate of the water table height was formed
by measuring the height of water standing in the wells. A sampling
well consisted of a 5-cm diameter non-perforated PVC pipe inserted
in an augur hole dug to a depth of 2 t0 3 m. Figure 2 shows the

location of the wells.

Description of the Simulated Waste Water

 

Preparation

 

Secondary treated municipal wastewater was not available at
the experimental site. A facsimile of East Lansing's secondary
effluent was made at the site. Fresh water, containing about 25 ppm
Cl, was piped to the site from an irrigation well. A solution con-
taining a combination of NaCl and 12-6-6 liquid fertilizer was injected
into the pipe supplying water to the sprinklers on the site. Ammonium
and urea made up most of the N in the liquid fertilizer. It was noted
that the injection solution produced a white precipitate when mixed
with the well water. While we did not find out what it was, it was

probably a phosphate salt. The concentration of a phosphate P in

 

45

“save; :mmzumn cowumch one onwcommc op mcomumscm oz“ co commcmaeou .m mczmwm

Asov;

 

1m

N\m

Ee+£uo

 

.mums zopm Lopez up?“ new Lem: a ucwgmn
.P
. . a q _ i. . .J.

.o_
48

Acws\msmumpv

i seem
..om
1 ca

 

46

the simulated wastewater was supposed to be 2.2 ppm. The typical
concentration of phosphate P in the simulated wastewater in 1977
and 1978 was 1.3 ppm. While not conclusive, this drop in concen-
tration suggests that some of the phosphate added to the irrigation

water did not remain soluble.

Irrigation

 

Irrigation water was brought to each plot by a single pipe
set down in the middle of the plot. Elevated sprinklers were
fastened to each pipe at 9.1 m intervals. Two types of sprinklers
were used; three full-circle Rainbird 9/64“ nozzles in the interior
of the plot and two Rainbird 3/32" semi-circle nozzles at either end.
The irrigation water was applied at a rate of 0.78 cm/hr for 3 hours.
Thus, the plots received 2.5 cm of stimulated effluent per irrigation.
In practice, the plots receiving 200 cm/year were irrigated 6 times

weekly, plots receiving 100 cm/year, 3 times weekly.

Sampling

m
0

do
u—J

 

The soil from the experimental site was sampled to a depth of
274 cm twice annually from 1973 through 1978. Samples were taken at
the following depths: 0-15 cm, 16-30 cm, 31-60 cm, 61-90 cm, 91-122 cm,
123-152 cm, 153-213 cm, 214-274 cm. The soils were taken from the
field, air dried and stored until analysis.

In 1977 and 1978, an investigation on the effects of supple-
mental fertilizer was initiated. After 1977 surface soil samples were

taken in each subplot. To maintain consistency, after the initiation

47

of the fertilizer experiment in 1977, the deeper semi-annual samples
from each plot were taken from the check subplot, to which no
additional fertilizer was added. In the fall of both years, soil
was taken from the top 15 cm of all 48 subplots, air dried and stored
until analysis.

In 1978, 3 forms of N were determined in soil samples taken

to a depth of 274 cm. The procedure is presented in a later section.

Water in Tiles and Wells

 

Irrigation was sometimes halted to harvest crops or because
of breakdowns in the supply of well water to the site. If irrigation
was not interrupted, tile water samples were collected weekly. Tile
water was withdrawn by suction through sampling ports in the plexi-
glass tiles at the monitoring stations which also housed the water
stage recorders.

Water in sampling wells was collected as time permitted in 1975,
1976 and 1977, but not in 1978. Well water was drawn by applying a
suction to a collection flask attached to a 3.5 m length of tygon tubing.
The height of water in the well was measured during this operation.
All water samples were stored in 250 ml polyethylene bottles. Within
two hours after sampling, the water samples were brought to the lab
and stored at 4°C until they were analyzed.
Phosphate and Potassium vs.
Pressure Experiment

In an effort to discover a relationship between effluent con-

centrations of phosphate and K with hydraulic head, a soil column

48

experiment was set up. Soils from the experimental site were collected
from 2 depths, 0-O.5 m and 0.5-1 m. These soils were air dried, passed
through a 10 mesh screen and packed in the order as taken from the
field to a depth of l m in four 10 cm 1.0. PVC pipes. Twenty liters

of solution with phosphate and K concentrations similar to the simu-
lated effluent were passed through the soil in each pipe. The hole

at the bottom of each pipe was stoppered and according to a randomized
scheme heads of water were maintained above each soil at 5, 10, 20 and
50 cm. After 8 hours, the stoppers were removed and water samples

were taken from each tube every five minutes for 50 minutes. The
samples were stored at 4°C until the end of the experiment and then

analyzed for phosphate and K.

Blew.

A 15 m section of one row of each hybrid in each corn sub-
plot was cut in the fall. The corn plants were weighed, a subsample
for chemical analysis was oven dried then ground to pass through a
40 mesh sieve.

The grass in each subplot was out twice yearly. The size of
the area cut varied, but in general, an area 13.5 m2 was cut. These
samples were also weighed, subsampled, dried and ground to pass through

a 40 mesh screen.

Laboratory Analysis
Soil

A11 soil samples, with the exception of those analyzed for N

in 1978, were analyzed at Michigan State University Soil Testing Service

49

Laboratory. Samples were air-dried, ground to pass through a 10
mesh screen and analyzed for extractable P, exchangeable Na+, K+,

+
a+2 and Mg 2

C and soluble Cl'. Phosphorus was extracted from the
soil with 0.03 N NH4F and 0.025N HCl at a 1:8 Soil-solution ratio,
shaken for 5 minutes. Color development was accomplished by

using ammonium molybdate and 1, 2, 4-aminonapthosulfonic acid
(Jackson, 1958). Cations were extracted with lN NH40Ac solution
(pH 7). Potassium and Na+ concentrations were determined using a
Coleman flame photometer. A Perkin-Elmer 303 atomic absorption
spectrophotometer was used to measure the concentration of Ca+2 and

Mg+2.

Samples to be measured for soluble Cl' were shaken with
saturated CuSO4 for 30 min. After filtration, the concentration
of Cl' was measured using an Orion specific ion electrode.
Soil Nitrogen Determinations of
the 1978 Growing Season

Field soil samples were collected three times for N analysis.
The soil was sampled to a depth of 274 cm or as far as possible.
These depths were as follows: 0-15 cm, 31-60 cm, 61-90 cm, 91-
122 cm, 123-152 cm, 153-213 cm and 214-274 cm. 0-15 and 15-30
samples were collected using a soil probe with which perhaps 10 sub-
samples in a plot were mixed to make up one sample. The samples taken
from 30 cm and below were obtained from the same hole with an auger.
Often, the lower samples could not be brought up because the sampling
depth was below the water table and the soil was not retained by
the auger.

All samples were placed in plastic bags and taken to the

50

laboratory. Here three subsamples were weighed and analyzed. One
subsample, weighing about 20 g, was used for NH: determinations.
These samples were shaken with 20 ml of 2N KCl, centrifuged and the
supernatant analyzed for NH: by micro-Kjeldahl distillation. A
second subsample was weighed, shaken with 20 ml of saturated CaSO4
solution and filtered. The filtrate was analyzed for NOE-N using
a Technicon Auto-Analyzer. The third subsample was weighed, placed
in an oven and reweighed to determine the water fraction in the soil.
Denitrification Incubation
Experiment

Air dried soils were sampled from each block at 3 depths; 0-15,
15—30, 30-60 cm were used in the experiment. Two subsamples were
taken from each sample; one subsample was wetted and dried before
analysis and the other subsample was used directly. Other than
wetting and drying, all subsamples were treated identically. Fifteen
g of soil were placed in a 125 ml plastic bottle to which 75 ml of
200 ppm N as KNO3 were added. The bottle was then capped and shaken.
The bottles were then kept in the dark at 25°C for 7 days. Following
the 7 days of incubation, the bottles were shaken and the contents
filtered. The filtrate was analyzed for N03 using an Orion specific
ion electrode.
Denitrification Assay by
Acetylene Inhibition

The measurement of denitrification by acetylene (CZHZ) inhibition

of nitrous oxide reduction was conducted on several soil samples in

51

September 1980. Since N20 is readily measured with a gas chromatograph,
the method has been shown to be an effective measure of denitrifica-
tion (Smith et al., 1978). The rate of N20 evolution from soil in
an anoxic environment in the presence of CZH2 was measured for 2 hours
following the addition of N03 solution. Denitrification during
this period has been designated "Phase I" by Smith and Tiedje (1979),
representing the denitrification rate before the derepression of
nitrate reductase synthesis following extensive periods of anoxia.
Soils were sampled two years after the cessation of the simula-
ted secondary sewage effluent project. The experimental site had
laid bare for the time after the end of the project. Soils at five
different locations in and around the experimental site were sampled
to a depth of 15 cm. One sample was taken in a plot which was planted
in corn, another sample was taken in a plot which had been cropped
in bromegrass. Both plots had been irrigated at 100 cm/year. These
soil samples were designated "Corn 100" and "Brome 100" respectively.
A third sample was taken from a bromegrass sod 20 m from the experimental
site. Two samples were taken in an irrigated corn plot 30 m from
the original site. One sample, called "Corn A" was taken right next
to the corn stalk. Another sample, called "Corn 8" was collected at
mid-row, 37.5 cm from the stalk. All soils were the same type,
Conover loam. The soil was near field capacity with a moisture con-
tent (by weight) of 22% to 28%. The samples were passed through a
4 mm sieve and stored without drying for two weeks in sealed plastic
bags at 4°C.

Three replicates of each soil sample were assayed for

52

denitrification. Each replicate (11.76 g oven dry weight) was put
into a 70 m1 serum bottle. The bottle was stoppered and 1 ml
chloramphenicol solution (365 mg/l) was injected into the bottle.
The chloramphenicol was added to inhibit protein synthesis. The
bottle was evacuated and flushed with argon (Ar) gas five times.
Nitrate solution (4.2 ml) of varying concentrations was injected
into the bottle to bring the total aquous concentration to 22.8
ppm NOS-N. Immediately, 5 m1 CZH2 was injected into the bottle
and the bottle was placed on a rotary shaker at 250 rpm. Every 24
minutes for two hours a 0.5 ml sample was taken of the serum bottle
head space with a Pressure-lock syringe (Precision Sampling Corp.,
Baton Rouge, La.).

The N20 concentration in the head space sample was measured
using a Perkin-Elmer 910 gas chromatograph. This machine was
equipped with a 2 m long, 0.32 cm diameter column filled with 8100
mesh Porapack Q. The gas chromatograph was equipped with a 63Ni
electron capture detector. The carrier gas was 5% methane, 95%

Ar at a flow rate of 15 ml/minute.

The total N20 content was calculated using the Bunsen
absorption coefficient (Wilhelm et al., 1977), an index of the
water solubility of the gas. The linear regression of the plot of

N20 content vs. time was used to calculate the rate of N20 evolution.

Soil Organic Carbon
To get an indication of the relative potential for denitrifica-

tion of the soils cropped in corn and bromegrass, the organic C content

53

of the surface soil was measured. Carbon was determined in soil
by ignition of 1 g oven-dried samples. The CaCO3 content was estima-
ted from the previously determined Ca content of the soil. Organic

C was calculated by the difference.

Water
Field samples of water from drainage tiles, wells and irriga-

a+2’ Mg+2

tion lines were analyzed for Na+, K+, C , Cl', NOS-N and phos-
phate. Na+ and K+ were both determined with a Coleman Flame Photo-
meter. These samples were diluted 1:10 to reduce interference by the

2 and Mg+2,

high concentration of salts in solution. Samples for Ca+
which were analyzed by using a Perkin-Elmer 303 atomic adsorption
spectrophotometer, were also diluted. The dilution ratio was 1:50 in
a solution to which La203 was added.

The Technicon Auto-Analyzer, an automated colorimeter, was
used to estimate the solution concentration of NOS-N and phosphate P.
Karlen (1975) found only nominal amounts of NOé-N in solution, so
NOé-N was measured colorimetrically after reduction to NOE-N. The
reduction is achieved by passing the sample solution through a cadmium
copper column. Ammonium molybdate and ascorbic acid were used for

color development to measure phosphate, also with the Technicon

Auto-Analyzer.

Plants
Both corn and bromegrass samples were sent to the Ohio State
University Plant Analysis Laboratory where they were analyzed

spectrographically for P, K, Ca, Mg, Na, Cu, Fe, Zn, 8, Mn, Al and

54

Ba. At Michigan State University the samples were analyzed for N

by the semimicro-Kjeldahl method.

Statistical Analyses

 

Most statistical analyses presented here were performed on
Michigan State University's Control Data Corporation 6500 computer.
Most of the statistics were calculated using the Statistical
Package for the Social Sciences (Nie et al., 1975).

Five specific subprogram routines were used. These were as
follows:

1. BREAKDOWN calculated the sums, means, standard deviations
and variances of a dependent variable.

2. PEARSON CORR computed Pearson product-moment correlations
for pairs of variables.

3. SCATTERGRAM presented a two-dimensional plot of data
points coinciding with the values of two variables being considered.
Slopes and intercepts along with their significance were also computed.

4. ANOVA calculated an analysis of variance for factorial
designs.

5. ONEWAY was used to calculate the Least Significant
Difference (LSD) of a number of dependent variables due to one
independent variable.

Those analyses not conducted with SPSS were calculated using
M.S.U.'s STAT data package.

The calculation of the correlation coefficients between water

height and tile flow was done using a Hewlett-Packard 25 programmable

55

calculator. These data were plotted with the aid of a Hewlett-

Packard 9845 desktop computer with hardcopy capacity.

CHAPTER III
RESULTS AND DISCUSSION

Concentration of Ions in Irrigation,
Tile Drainage, and Soil Water

 

 

Concentration of Ions in
the Irrigation Water

 

We attempted to approximate the concentration of the major
inorganic ions found in East Lansing's secondary treated effluent
by injecting a salt-fertilizer mixture into the irrigation water
at the site. Table 1 presents an analysis of East Lansing's
secondary treated sewage along with the projected concentration
in the simulated effluent. Comparison with the data in Table 2
shows how closely the concentrations of ions in the irrigation
water approached the concentration of ions in the real wastewater.

The concentrations of all ions in the stimulated wastewater were
reasonably close to what was intended to be manufactured at the
site.

Since Cl' will be given special attention in the body of this
report, the nature of this anion will be investigated more extensively.
The concentration of irrigation water C1' in 1977 must be taken with
caution. Difficulties with the injection system caused huge variations
in the Cl' concentration. In some samples of irrigation water Cl'
concentrations were in excess of 1000 ppm. The concentrations of

C1' in the irrigation water were much more variable in 1977 than 1978.

56

57

Table l. Effluent composition. (From Karlen, 1975)

 

 

 

 

Constituent East Lansing's Secondary+ Simulated
apmt—

Total N 15.2 15.0
Nitrate N 3.1 3.5
Total P 4.9 3.3
Soluble P 1.1 2.2
K 8.6 6.2
Na 130 1301
Ca 100 200
Mg 25 62
01 260 2001

 

+In 1973 the simulated effluent contained 210 ppm Na and 324 ppm 01.

This was the amount of Cl added. The irrigation water had an additional
25 ppm C1 from natural sources.

58

cmumzmpmmz poc .Lmumz mmmcwmcc up?“

Lee ae_a> omsfi

.Eaa mp mm:

cmpmzmummz Low 2 Peace +

 

 

 

 

 

 

 

Ammv _NN New oem mom ANN _u
A__v mmp mpp omp mpp amp az
A~.mv m.em “.mm otmm m.ee ~.Ne a:
Amy _m_ Nap amp Ne, map ac
Aa.ov N.e m.m ~.N o.m e.a x
Aao.ov Ne.o N~.o ep.o ep.o am.o a-eoa
Amm.Fv mm.w mm.a Ne.m mN.m mp.m z-moz
.MMmH
Aemv ~_N mom _e~ New mpm _u
Ao_v PF. ao_ mm, N_P wep az
Am.mv m.mm F._e e.~e 3.5m e.,e a:
Amzv mom New m_N N_N mom a8
AN._V ~.m a.m m.m a.m ~.m x
Am_.ov ee.o me.o om.o mm.o Ne.P a-eoa
AN_._V ea.m Ne.e e~.m oo.~ +me.~ z-moz
“mmwfl
AEQQV
+Amo.oc ems Ea oom-eaau so oo_-eeou Ea oo~-mmaau so oop-mmaee
4, cope: ummcwmco whee cmumzmgmmzi
.wea_ eea Reap

cw mm—wp Eocw mcwcwecu Lopez cw use copezoummz umme:E_m cw meow mo mcowumcucmucoo com: .N opnmh

59

Since Cl' was used as a tracer to estimate the leaching fraction
leaving the root zone of the soil, this variability was unwelcome.
Hence, the leaching fraction estimates in 1978 must be considered
more valid than those in 1977.

As was noted earlier, virtually all N passing through the
drainage tiles was in the form of N03. The majority of the N in
the irrigation water was non-NOS-N, primarily urea and MHz. The
makeup and amount of the non-NOE-N in the irrigation water was not
verified and this is a major failing in any interferences or con-
clusions made about the behavior of this nutrient once applied to
the soil. The indications in the literature are that the urea
in soil was rapidly hydrolized to NH: (Tabatabai, 1972) and that
NH: was completely and quickly nitrified to NO} in sewage amended
soils (Broadbent et al., 1977). In this thesis, it is assumed
that the concentration of total N applied was 15 ppm and that the
N present as urea and NH: was rapidly converted to N03.

The concentration of phosphate P in the simulated secondary
wastewater was below the projected soluble P concentration of 2.2 ppm.
This discrepancy may have been due to precipitation of P when the
salt, fertilizer and water were mixed. How much of the P which was
applied to the soil either as a phosphate precipitate, as non-phosphate

precipitate or as non-phosphate soluble P is conjecture.

Water and Ions in the Tile Drains
The calculated volume of water draining through the tiles in

1977 was less than 1978 (Table 3). The difference was substantial and

60

Table 3. Annual water drainage from the tiles in 1977 and 1978.

 

 

Plot or Block 1977 1978
liters
1 729,173 1,102,998
2 198,909 156,300
3 586,360 937,632
Grass - 100 cm 1,514,442 2,196,930
4 509,099 1,062,535
5 40,977 48,956
6 100,334 110,382
Corn - 200 cm 650,410 1,221,873
7 64,640 77,561
8 407,151 405,758
9 771,707 1,175,193
Corn - 100 cm 1,243,498 1,658,512
10 565,652 968,220
11 535,171 477,438
12 448,023 791,992
Grass - 200 cm 1,548,846 2,237,650
Total 4,957,196 7,314,965

 

 

 

 

61

cannot be explained by the use of different calibration formulas

for the two years, because over the typical range of flow the two
formulas are very similar. Another possibility is that the positioning
of the floats used to measure tile water height was altered. This
explanation is less probable because the increases in 1978 were con-
sistent in most tiles with 1977 flow, which suggests an arithmetic
variation during data manipulation rather than a physical variation

in the field. The results indicate that between 50% to 75% of water
applied by irrigation left via the tiles.

The theoretical development of tile flow does not allow lateral
movement of subsurface water from the area about one tile to another
tile or flow beyond the domain of the tiles altogether. Not much
lateral movement out of the experimental area was indicated by the
data. The high volumes of flow in tiles #1 and #12 did not indicate
great amounts of water loss by flow east or west. That the flow
east and west was probably small suggests that water movement north
and south was also limited. There were wells placed north of the
boundary of the experimental area. The water level in these
wells was considerably less than that in the wells in the experimental
area. Apparently, the loss of irrigation water by lateral movement
out of the experimental site was limited.

The leaching fraction calculated using the ratio of Cl' con-
centration is presented in Table 4. The leaching fraction, the volume
of water leached below the tile drains during the portion of the
growing season when water was sampled in the tile drains, was calculated

by the following formula:

Table 4.

62

Estimate of the leaching fraction of the individual tiles

(liters/year)

 

Number

1977

1978

 

0'1wa

oxooowcn

11
12

Total

 

704,347 (123)+

739,130 (129)
733,096 (128)
1,490,527 (130)
1,952,266 (170)
1,718,741 (150)
925,817 (161)
863,502 (150)
831,521 (145)
1,589,456 (138)
1,454,320 (127)
1,436,868 (125)

14,439,591 (140)

 

liters

669,953 (117)
634,522 (111)
616,732 (107)
1,113,761 ( 97)
1,293,928 (113)
1,189,015 (104)
695,456 ( 83)
592,425 ( 97)
. 573,829 (100)
1,061,941 ( 93)
1,079,864 ( 94)
1,057,553 ( 92)

10,587,979 (103)

 

+Number in parenthesis is leaching fraction expressed as a
percentage of the amount of irrigation water applied over the year.

63

C1
v=1v
L CIL I
where: VL leaching fraction (liters)

<
N

I total irrigation volume for the Season (liters)

n
.....1
II

I mean concentration of C1' in the irrigation water (ppm)

0
._a
I

L - mean concentration of C1. in the tile drainage water (ppm)
The estimates of the leaching fraction during 1977 should be viewed
with caution. If we accept the estimates of leaching in 1978, then
the amount leached was approximately equal the volume applied. If
this is true, then the amount transpired from a plot was about equal
to rainfall, which makes sense intuitively. The estimates of
leaching volumes were used to calculate nutrient loss by leaching.

2 2, Na+ and Cl',

Seven ions: phosphate, N03, K+, Ca+ , Mg+
were monitored weekly in the tile water. Ammonium was evaluated
intermittently. The typical NHZ-N content was well below 1 ppm

and apparently not above the error inherent in the analytical method.
The concentrations of the other seven ions in the tile water were
analyzed statistically by irrigation rate and crop. This analysis

is presented in Table 2.

The differences in the concentration of C1' in the tile water
were not consistent from year to year. Given the fact that Cl' is
poorly retained by the soil and hardly taken up by crops, attempts
to draw conclusions from this data is bound to be spurious. On the
other hand, the tendency of Na+ in tile water is clear. The concen-

tration of Na+ under plots irrigated at 200 cm per year was higher

than those irrigated at 100 cm. This behavior is probably due to

64

the higher amount of the ion which passed through the exchange

complex of the soil.

2 2

The concentration of Mg+ , Ca+ , and K+ was significantly
different in plots 1, 2 and 3, which were cropped in bromegrass

and received 100 cm/year, than in the rest of the experimental area.
There is no similar behavior of the other plots irrigated at 100 cm
or the other grass plots. Under these three plots, Mg+2 and Ca+2
tended to be lower, whereas the concentration of K+ was substantially
elevated.

The concentration of NOS-N was less in tile water under
grass and also less at the lower irrigation rate. The depression
of NOS-N concentrations under grass will be examined in more detail
later in this thesis.

The concentration of phosphate in tile water was less under
grass than under corn. As will be shown later, the corn crop took
upmore P than grass and one can only speculate on the mechanisms
responsible. Lance (1977), in a study of the removal of sewage water
P from soil columns, suggested that plant roots act somehow to keep
P compounds in solution. Lance did not demonstrate or suggest how
the plant roots were capable of doing this.

We have observed, as Karlen (1975) first demonstrated, that
the concentration of phosphate in the tile drainage water rose as
the flow rate of water in the tile increased. This behavior is con-
trary to any model in which sorption-desorption reactions of P in

the soil are time dependent. This behavior was so peculiar that I

investigated it further. Here, I will present the facts about the

65

[dynamic behavior of phosphate—P in the tile water. Following that, I
will evaluate various hypotheses to explain this egregious phenomenon.

1. The phenomenon was the increase and decrease of phosphate
P concentration during the tile flow subsequent to an irrigation.

2. The rate of water flow in the tile was highest about 30
minutes after the cessation of irrigation.

3. In 9 of the 12 plots the concentration of phosphate P in
the tile drainage water was linearly correlated with K+. (see
Table 5 and Figure 4)

4. The concentration of phosphate P in the irrigation water
was not similarly related with K+.

5. The concentration of phosphate P and K+ rose together
as the flow event progressed. The maximum concentration of both ions
fell after the peak. The maximum flow rate was not coincident
with the phosphate P and K+ concentration maxima (Figure 5).

‘ 6. This behavior was observed in tiles under both bromegrass

and corn.

7. Both the linear correlation between phosphate P and K+
and the flow related vagaries of the two ions, though marked, were
less noticable early in the season and became progressively more
distinct as the year progressed.

8. Returning to Table 6, the correlation was not significant
between the concentration in tile water under plots 1, 2, and 3.
The slopes of the regression lines relating the concentration of
phosphate P with that of K for these tiles are much less than the

other tiles. Unlike most other tile lines in the experiment, we

Tabl

66

e 5. Linear regression and correlation coefficient for the
concentrations of PO4-P and K in tile drainage water
(umoles liter")

 

 

Tile No. of Samples A+ B r2
1977

l 12 -TTTO4 .007 .011

2 13 - 5.32 .074 .304*
3 13 11.11 -.005 .003

4 l4 - 27.44 .373 .820**
5 13 - 5.10 .242 .644

6 ll - 7.10 .220 .723**
7 12 9.00 .085 .380*
8 13 4.75 .113 .435**
9 10 - 0.90 .157 .871**
10 ll - 4.97 .188 .614**
11 ll - 5.03 .258 .659**
12 12 — 3.54 .234 .700**

1_Z§_

1 9 .10 .027 .543*
2 9 - .08 .021' .234

3 9 .50 .016 .244

4 9 3.29 .023 .006

5 8 - 3.18 .182 .441*
6 8 - 29.80 .361 .721**
7 7 - 17.74 .240 .688*
8 9 - 11.03 .202 .895**
9 9 - 4.73 .129 .862**
10 8 - 6.71 .145 .940**
11 8 - 2.17 .118 .571
12 8 - 5.65 .169 .653

 

+(P04) = A + B (K)

67

.commmm mchocm Rump may

coo emacwmcu mPTP cw cowpmcucmucoo sawmmmuoq ucm mumcamosa we pope .v mczmwd

 

“Enavx
o.o_ o.m_ o.m I o.¢
_ . d o
co 00 %@m m m
GOOD 0
. . a. .
am 6 an N o Ens
D DD 0
o o a given.
0
o
o L
DU 0 .Voo
$628 : .62 2.; - a
AER: m .02 3:. i o
O L

 

m.o

' 50
4O
TILE FLOW
30
(Liters/min)
20
10
0
1.0
TILE
P04-P
(ppm) 0

Figure 5.

 

 

68

Changes in tile flow

 

NOD-$50103

I TILE

(ppm)

rate and concentrations of

Fog-P and K in the water draining from Tile #11, for a single flow
ev

nt, 10-24-1978.

69

cannot say that there is a linear relationship between phosphate
P and K+ in tile drains l, 2 and 3. Perhaps part of the lack of
correspondence between the two ions in these tile drains is due
to the higher concentration of K+ in the tile drainage water.

I was unable to find out what caused this behavior of phos-
phate P in the tile drainage water. Four hypotheses exist to describe
this phenomenon and I will critically examine them below. Whatever
mechanism (or mechanisms) is at work clearly involves the soil.

The correlation between the concentrations of phosphate and K+ were
closer coming out of the soil in the tile drainage water than it

was in the irrigation water. In addition, the behavior became more
distinct as the year progressed, and the cumulative load of the

ions increased. Thus, I discount hypotheses which were based on

the thought that water moved through portions of the soil profile
with little or no reaction. While the mechanism need not have been
instantaneous, it must have accounted for the coincidence of the
cessation of irrigation and the peak ion concentration. Adsorption-
desorption reactions in soil can be fast, but as far as I know,

there is no specific relation between the sorption of P and K. While
there are phosphate minerals which contain K, reactions of these
minerals tend to be very slow. Since the precision of the behavior
increased over the year, I am not inclined to discount the involvement
of some sort of adsorption mechanism.

Several lines of inquiry can be dismissed quickly. Whatever
caused the phenomenon, it was not due to changes in the electrical

charge balance of the soil solution. Phosphate and K+ were minor

70

constituents of the soil solution and had opposite charge. Any

change in the electrochemical balance that would cause one to increase
would tend to decrease the other. The close behavior of the two ions
suggests some sort of associated complex with K and phosphate.

If an associated complex was the cause, the regression equations
suggests that virtually all the P was associated with K and that

the molar ratio of phosphate P to K was between 1:6 to 1:5. The

2

association constants for HPO4' and H2P0; with K+ presented by

Larsen (1967) suggest only a small fraction of phosphate P would
be associated with K. The valences of the three ions, K+, HP04'2,
and H2P0£ are not consistent with a ratio of P:K higher than 1:2.
Several other suggestions cannot be dismissed so readily.
The relative hydrated radius of the two ions would be expected to
stay fairly constant. Thus if the two were moving in response
to a nonelectrical gradient, one would expect constancy in the ratio
of the two ion concentrations. The two ions may be complexed by
a larger molecule, possibly organic, which dragged the two ions through
the soil solution. Because the organic molecule was larger than most
other ions in solution, its flow from soil to tile could be velocity
dependent.
The first hypothesis to explain the behavior of phosphate P
in tile drainage water was developed by Karlen (1975). He argued
that the rapid flow of water through sand smears to the drains
explained the peak behavior of phosphate in tile drains. In the

soil, the argument went, there were sections with a lower ionic

adsorption capacity and higher hydraulic conductivity than the rest of

71

the soil. In these smears, the concentration of phosphate and

K+ in the soil solution was higher than in the rest of the soil.
Since the conductivity was higher, as the irrigation proceeded the
contribution of the solution from the smear to the tile drainage
water increased. Thus, the concentration of the two ions in the

tile water increased during an irrigation. The sand smear hypothesis
is consistent with the flow of water through soils which are composed
of two commingled portions, each with different hydraulic conduc-
tivities. The flow to the tiles, where the maximum flow was found
after the cessation of water application is virtually identical

to that of flow through macropores in soil (Thomas and Phillips,
1979). In theory, flow through macropores and through sand smears

is similar in that the dynamics of flow is governed by two sectors
of the soil, each having greatly different conductivities.

The hypothesis also poses many considerable failings. The
sand smear hypothesis depends upon a convenient geometry in the
subsoil. While only two tile drains were studied for the dynamic
behavior of phosphate in the water 9 of the 12 tile drains showed
significant correlations between phosphate and K indicating that,
if the two behaviors were linked, the behavior was common in the
field. Therefore, in a very heterogenous soil, if they were the
cause of the behavior, sand smears must make a sizable portion of
the flow to most of the tiles. Further, the soil used by Sharpley
et al. (1977) in which the same behavior is observed was homogeneous
probably without smears, sandy or otherwise. The model provides

nothing to explain the correspondence of the concentrations of

72

phosphate and K in the tile. Lastly, the model does not explain
the repetitiveness of the behavior. To show why the model does
not, the behavior of hypothetical model will be examined. Begin
with two sectors of the soil, A and B. Both are connected to

the tile drain and water flows through each to the tile drain.
Assume that the conductivity and ion concentration is greater in A.
Also assume that the volume of B is substantially greater than that
of A. After one irrigation, the behavior would be that observed

in the field, increasing flow and ion concentration in the tile.
After irrigation stops, though, we can expect much of the solution
in B to replace that in A. Thus, with the second irrigation, no
appreciable change in the ion concentration would be observed. There-
fore, the explanation is neither internally consistent nor does it
adequately explain much of the behavior which has been observed.

The second explanation can be called the particulate hypothe-
ses. According to this hypothesis, the behavior of P in tile lines
was based upon the nature of flow in the tile drain. Phosphorus and
K were adsorbed by minerals in the sediment of the tiles. As the
flow rate in the tiles increased, so did turbulence, and sediment was
suspended in the tile drainage water. Phosphorus and K previously
adsorbed upon mineral surfaces were released into solution and the
concentration of the two ions increased. This hypothesis was implied
by Sharpley et a1. (1977) to explain the increase in P concentration
with the increase in tile drainage water flow. The hypothesis was
presented obliquely. They noted that several investigators had

observed a correlation between flow rate in rivers and runoff and

73

the concentration of ions, including P. This was believed to be due
to an increase in turbulence concomitant with the increase in flow.
Turbulence caused the sediment to be brought into suspension. Their
presentation may be faulted on two counts. (1) One of the articles
they cite about the correlation of flow and ion concentration (the
only one I was able to locate) showed a decrease in ion concentration
in a river with an increase in flow. (2) While they present the
general observation that "...an increase in suspensionate was obser-
ved in the tile," they failed to demonstrate that P associated with
particulate matter was responsible for the increase in P concentra-
tion in tile water. This criticism is due largely to the fact that
the Sharpley article is the only one on this aspect of tile flow to
appear, and I have scrutinized it on an aspect fairly far removed from
the main theme of the paper. The hypothesis was presented in passing
and probably was not considered in great detail by the authors. Aside
from the objections of the hypothesis inherent in the presentation of
the article, there are two other criticisms. (l) The maximum tile
flow rate, and I assume the maximum turbulence, and the peak concentra-
tions of P and K do not coincide. (2) Since the chemistry of the ad-
sorption or precipitation of P and K in sediments has not been shown
to be related, the correspondence in the tiles draining the Conover
study soil is unexplained.

A third hypothesis of the ion behavior in the tiles is based
upon different path lengths of the soil water must have traveled to
the tiles drains. Water and phosphate flowing from portions of the

soil near the tile moved quickly into the tile. The concentration

74

of phosphate in the soil solution flowing from portions near the

tile drain was higher than in solutions flowing from portions further
from the tile drain. This quickly flowing, high concentration solu-
tion caused a spike in the tile water concentration of phosphate during
an irrigation.

This theory, like the sand smear theory, depends upon con-
tributions of water from different sectors of the soil with different
P and K concentrations. Two facts weigh in its favor. (l) P concen-
tration in soil solution decreases with the increase in distance
traveled and (2) according to models presented by Childs (1947) and
Kirkham (1964) most of the water which is delivered to the tile along
the shortest path lengths enters the tile soon after the end of water
application. There is nothing to explain such persistent correlation
between the phosphate and K+ in the tile drainage water.

The last hypothesis to explain the peak beahvior of phosophate
assumes that the soil about the tile drain acted as a semipermeable
membrane allowing water to pass more freely than ions in solution.

In the soil near the outside of the tile, the water pressure dropped
dramatically from that due to the hydrostatic head to atmospheric
pressure. This difference in water pressure increased as the irri-
gation proceeded and the water table rose. Increases in the pressure
difference decreased the ability of the soil to act as a membrane,

to restrict the passage of ions relative to water. More ions passed
through the portion of soil close to the tile relative to the flux
of water and the concentration of the ions in the tile drainage

water increased. All ions behaved this way, only the behavior of

75

phosphate and K were discernable because their concentrations were
so much less than the rest.

Because little is known about the subject, it is difficult
to evaluate the possibility that the soil about the tile acted as
a semipermeable membrane. But, if one discounts macro-physical
explanations such as the sand smear hypothesis and is unable to
find satisfactory explanations in the sorption behavior of phosphate,
there are few avenues left to follow. Studies in soils and clays
have shown changes in the relative flux of ions to the flux of
water due to changes in hydraulic head. To test this possibility,
I measured the concentration of phosphate and K in leachate from
soils maintained at four different hydraulic heads. No tendency
was observed. The soil had been screened, the phosphate adsorption
capacity had increased and the phosphate concentration in the effluent
was very low. The field data has shown that the behavior was more
pronounced the longer the soil has been irrigated with phosphate
solution. Perhaps not enough P had passed through the soil in the
laboratory to really test if hydraulic head has an effect upon ion
concentration in the effluent water. I suspect that the variable
head experiment did not adequately test the effect of pressure on
phosphate and K+ concentration in the effluent. A second fact
supporting membrane hypothesis is that the concentrations of phos-
phate and K+ were strongly and consistently related. No other expla-
nation presented here, save the existence of a complexing molecule,
can explain this fact. We can expect that the hydrated radii of

2

HP04' , H2P04 and K+ remained consistent in the soil solution and if

76

they are moved by non-electrical gradients, the ratio between them
.in solution would be consistent. If the cessation of irrigation
affected the concentration of ions in the tile, and all information
I have gathered supports this argument, then whatever mechanism

was responsible, it must explain the speed and consistency of the
soil system response. Looking only at the chemistry of P in soil
we see that mineral dissolution and chemical desorption are

too slow, as is chemical adsorption. Only physical adsorption or
possibly membrane behavior of the soil react in times as quickly

as we observed in the field.

Several formidable arguments go against this explanation.

It has not been shown that the membrane behavior of soils is affected
by pressure gradients as small as would be found at the tile drain.

It has not been shown that soils act as a membrane with mixed
electrolyte solutions. Indeed, it has not been shown that soils

act as membranes at all, only pure clays have been used in experiments
thus far. It has not been shown that the movement of phosphate is
affected by this aspect of soil behavior. In fact, few anions have
been investigated. The membrane hypothesis and the other three,

are unable to adequately explain the peak behavior of phosphate and

K+ in the tile drainage water.

As noted above, the mean concentrations of N05 in tile drainage
water under grass was lower than in those under corn. Thomas and
Barfield (1974) have contended that, because flow to tile lines is
through aerobic soil, denitrification is reduced and the concentration

of N03 is greater in tile drainage water than in other subsurface water.

77

Using the measured soil N03 concentration and the moisture content
of soils collected in 1978, I calculated the N03 concentration in
the soil solution. The soil solution N03 content calculated by
this method should be more accurate than with most other ions

in solution because N03 is weakly attracted by most soils. The
mean concentration of N03 in soil solution in 1978 are presented
in Table 6, as well as the mean drainage water concentration for
the ion that year. The tile drainage water concentration were
consistently less than those calculated for the soil solution.
While these data do not establish the drainage water as a reliable
estimate of leachate N03 concentration, they do run counter to the
argument by Thomas and Barfield that the N03 concentration is
typically much higher in the tile drainage water than in leachate

from the soil.

Plant Response

 

The application of sewage to cropland has two aims; to remove
potential pollutants and to produce a usable crop. The plant response
in our experiment must be judged by similar criteria. Here three
categories of crop response to irrigation rate and additional fertil-
izer will be presented; nutrient quality of the crop, crop production
and annual nutrient removal by the crop.

The nutrient concentration of the bromegrass was consistently
affected only by the irrigation rate (Tables 7 and 8). The calcium
concentration of the bromegrass was higher in the plots irrigated

at 200 cm/year.

 

78

Table 6. Calculated mean concentration of NOR-N in the solution of

soils sampled at indicated depths a
drains, 1978.

d in water from tile

 

NOS-N Concentration

 

 

 

 

Depth Grass-100 cm Grass-200 cm Corn-100 cm Corn-200 CEL_
—( cm)-— (ppm)

0- 15 14. 25.8 27.9 27.7
15- 30 16. 19.0 26.1 37.5
30- 60 12. 12.8 18.4 29.0
60- 90 10. 11.4 15.0 19.6
90-120 13. 9.8 12.9 18.2

120-150 11. 8.4 11.5 18.5
150-210 10. 7.8 11.4 16.6
210-270 11. 9.7 12.3 17.7
Tile drains

(120 cm) 3. 5.4 6.6 8.5

 

79

Table 7. Influence of fertilization and irrigation rate on the
nutrient concentration of grass in two cuttings, 1977.

 

Nutrients

 

Treatment N P K . Ca Tfig

 

 

 

I - __
First cutting (percent of dry weight)

Check 1.81 .49 1.77 .64 .24
N 2.28 .55 2.29 .72 .34
K 1.73 .43 1.70 .62 .23
NK 2.43 .52 2.16 .69 .31
LSD (0.05) (.47) (.07) (.31) (NS) (NS)
100 cm 2.02 .47 1.65 .58 .20
200 cm 2.11 .53 2.70 .76 .36
LSD (0.05) (NS) (NS) (NS) (NS) (NS)

Second cutting

Check 1.97 .67 1.54 .72 .16
N 2.06 .64 1.52 .70 .16
K 1.95 .65 1.64 .76 .18
NK 2.10 .64 1.56 .70 .16
LSD (0.05) (NS) (NS) (NS) (NS) (NS)
100 cm 1.86 .60 1.48 .64 .16
200 cm 2.17 .70 1.05 .80 .17

LSD (0.05) (.19) (NS) (NS) (.05) (NS)

 

 

80 '

Table 8. Influence of fertilization and irrigation rate on the
nutrient concentration of grass in two cuttings, 1978

 

Nutrients

 

Treatment N P K Ca ’Mg

 

 

 

(percent of dry weight)
First cutting

Check 1.26 .27 2.16 .30 .12
N 1.25 .27 2.10 .29 .10
K 1.23 .26 2.20 .29 .11
NK 1.33 .26 2.13 .28 .10
LSD (0.05) (NS) (NS) (NS) (NS) (NS)
100 cm 1.26 .27 2.15 .28 .10
200 cm 1.28 .26 2.15 .30 .11
LSD (0.05) (NS) (NS) (NS) (.02) (NS)

Second cutting

*Check 1.26 .42 2.38 .54 .19
N 1.25 .40 2.44 .67 .26
K 1.23 .43 2.13 .56 .20
NK 1.33 .39 2.32 .61 .22
LSD (0.05) (0.07) (.02) (0.28) (0.07) (.07)
100 cm 1.26 .41 2.45 .57 .23
200 cm 1.28 .41 2.18 .62 .21

LSD (0.05) (NS) (NS) (0.20) (0.05) (NS)

 

 

81

As shown in Table 9, additional fertilizer had only one
effect upon nutrient concentration in corn. Nitrogen fertilization
increased N concentration in the crop. Phosphorus concentration
was higher in corn grown in plots receiving the higher irrigation
rate.

The dry matter yield of bromegrass was increased by
additional irrigation and fertilizations. Table 10 shows that only
the combination of N and K fertilization consistently increased
dry matter production.

Irrigation rate did not have a significant effect on dry
matter production or grain yield of corn, as shown in Table 11.
Of the two corn varieties grown in the experiment, Pioneer 3780
yielded more silage and grain than did Funks 4444. Nitrogen fertili-
zation, but not additional K, increased grain and silage production.

The major determinant of nutrient removal was dry matter yield
of the crop. Corn took up more N, P and K than did bromegrass
(Table 12). Increased irrigation rate stimulated removal of N, P
and K by bromegrass and increased the removal of P by corn (Table
13). Phosphorus removal was unaffected by additional fertilization.
Of the two corn varieties, Pioneer 3780 removed more of the nutrients
than did Funks 4444. More N was added to the crops as fertilizer
than was removed by the induced increase in production. The increase
in N uptake by corn in subplots receiving additional N fertilization
was a 40.9 kg/ha, the increase in subplots receiving both N and K
was 52.9 kg/ha. In subplots cropped in bromegrass the increases

in N removal for these treatments were 30.5 and 65.2 kg/ha

 

 

 

82

Table 9 . Influence of fertilization and irrigation rate on nutrient
concentration of corn silage in 1977 and 1978.

 

 

 

 

 

Nutrients
Treatment N P K . Ca Mg
1977 -percent of dry weight
Check .96 . .48 1.15 .31 .20
N 1.08 .43 1.14 .29 .19
K .94 .43 1.11 i .28 .20
NK 1.08 .39 1.13 .27 .18
L50 (0.05) (.05) (.05) (NS) (NS) (NS)
100 cm 1.00 .40 1.14 .27 .20
200 cm 1.03 .47 1.12 .30 .19
LSD (0.05) (NS) (.07) (NS) (NS) (NS)
1978
(Check 1.08 .33 1.22 .34 .24
N 1.09 .32 1.11 .32 .25
K 1.03 .33 1.21 .33 .24
NK 1.12 .34 1.23 .36 .25
LSD (0.05) (0.07) (NS) (NS) (NS) (NS)
100 cm 1.02 .27 1.11 .28 .21
200 cm 1.14 .38 1.27 .39 .27

LSD (0.05) (0.05) (.05) (.14) (.05) (.03)

 

 

83'

 

 

 

 

 

Table'H). Influence of fertilization and irrigation rate on dry
matter yield of bromegrass in 1977 and 1978.
1977 1978
First Second First Second
Treatment cutting cutting, Total cutting cutting Total
(metric tons/ha)

100 cm
Check 2.69 1.40 4.09 4.95 2.32 7.27
N 4.52 1.77 6.29 5.31 3.66 8.98
K 2.63 1.22 3.85 4.64 2.69 7.33
NK 7 4.68 1.59 6.17 5.68 4.15 9.83
LSD (0.05) (1.59) (NS) (NS) (NS)

200 cm
Check 5.50 1.59 7.08 5.80 4.34 10.14
N 5.68 1.34 7.02 5.37 4.40 9.77
K 4.34 1.65 5.98 5.86 4.82 10.69
NK 8.06 1.53 9.59 5.74 5.19 10.93
LSD (0.05) (1.59) (NS) (NS) (NS)

Fert. Treat. Av.
Check 4.09 1.47 5.56 3.60 5.07 8.67
N 5.07 1.53 6.59 4.52 4.88 9.40
K 3.48 1.22 4.70 3.66 5.31 8.97
NK ; 6.29 1.59 7.88 4.88 5.50 10.38
LSD (0.05) (1.10) (NS) (0.55) (0.55)

Water Treat. Av.
100 cm 3.60 1.53 5.13 5.13 3.18 8.30
200 cm 5.86 1.40 7.27 5.68 4.70 10.38
LSD (0.05) (1.59) (NS) (0.55) (0.55)

 

 

Table 11. Yield of corn, as silage and grain in 1977 and 1978, as
affected by fertilization, effluent application rate and

corn hybrid.

 

Silage Yield+

 

Grain Yield

 

 

Treatment 1977 1978 1977 1978
—(metric ton/ha)- ——-(quintal/ha)——-—
Fertilization
Check 14.9 13.4 85 58
N 17.3 16.7 107 97
K 15.7 13.8 85 59
LSD (0.05) (1.3) (1.4) ( 8) (4)
Water rate
100 cm 17.0 15.3 99 75
200 cm 16.3 15.4 95 8O
LSD (0.05) (NS) (NS) (NS) (NS)
Corn variety
Pioneer 3780 17.6 16.4 107 86
Funks 4444 15.7 14.2 86 7O
LSD (0.05) (1.3) (1.0) (3) (5)

 

+Dry weight.

 

85

Table 12. Effect of crop and corn variety on uptake of N, P
and K in 1977 and 1978.

 

Nutrient Uptake

 

 

 

 

Treatment N P - K
(kg/ha)
Crop (1977)
Bromegrass 121.0 31.6 113.2
Corn 170.8 71.2 185.9
LSD (0.05) (27.1) (5.8) (18.3)

Variety (1977)

Pioneer 3780 178.8 68.4 172.6
Funks 4444 162.8 74.1 199.1
LSD (0.05) (17.4) (6.7) (15.4)

Variety (1978)
Pioneer 3780 179.0 54.7 195.7
Funks 4444 154.6 46.3 169.9
LSD (0.05) (20.5) (7.4) (23.9)

 

 

86

Table 13. Effect of irrigation rate and additional fertilization on
uptake of N, P and K by corn and bromegrass in 1977 and 1978.

 

Crop Uptake

 

 

 

 

 

Corn ‘ Brome rass
Treatment N P K N E K
1977 (kg/ha)
Irr. Rate
100 cm 169.3 67.1 178.0 93.4 24.3 76.5
200 cm 172.2 75.4 193.7 144.0 37.8 143.7
LSD (0.05) (NS) (6.4) (NS) (38.2) (7.3) (29.2)
Fertilization
Check 144.8 70.5 169.2 102.2 31.1 99.7
N 188.4 74.4 196.5 132.7 34.9 129.8
K 149.5 67.3 169.2 80.5 22.8 79.3
NK 200.6 72.8 208.6 167.4 38.3 144.3

LSD (0.05) (15.1) (NS) (17.6) (47.8) (5.3) (52.7)

1978

Irr. Rate
100 cm 157.9 41.6 169.3
200 cm 175.7 59.4 196.3
LSD (0.05) (NS) (5.7) (23.8)

Fertilization
Check 145.3 44.4 165.2
N 183.4 52.8 183.2
K 143.3 45.6 168.0
NK 195.3 59.2 214.8

LSD (0.05) (24.5) (10.0) (32.1)

 

 

 

87

respectively. Since both crops received 112.1 kg/ha N as additional
fertilizer 50% or less of the additional N was not taken up by the
crops. The efficacy of additional N fertilization must be judged
by two factors: (1) the load of fertilizer N finding its way to

the groundwater and (2) the economic benefit from the increased
crop production. As will be shown in the next section, much of the
added N not removed by the crop did not find its way to the ground-
water. This disparity was most pronounced under grass. The danger
of N pollution may be less than indicated by incomplete N removal

by the crop.

Changes in the Soil

 

During the 1977 and 1978 seasons, the chemical components measured

*2 and Ca+2 and soluble 01’)

(extractable P, exchangeable Na+, K+, Mg
changed little. Table 14 presents the concentration of these ions
in Fall 1978. The distribution of extractable P in the soil profile,
shown in Table 15, varied greatly from time to time but no real trend
was discernible in 1977 and 1978. Comparing the concentrations of
extractable P in 1977 and 1978 with those in spring 1973, it is clear
that considerable amounts of P had accumulated in the soil profile
since the beginning of the experiment. The only obvious change in
the profile distribution between 1977 and 1978 was a small increase
in the P concentration in the lower depths.

Calcium and Mg+2 concentrations varied greatly from sample to

sample but showed no tendency for accumulation or loss. Potassium,

on the other hand, had a higher concentration in the soil profile

 

88

Tablelll. Concentrations of extractable P and exchangeable K, Ca,
Mg and Na in soil samples at given depths (Fall, 1978).

 

 

 

 

Depth (cm) P K Ca Mg Na
{09/9 soil)

0 - 15 126 72 1655 242 140
15 - 30 36 69 1351 234 135
30 - 60 14 71 1596 291 147
60 - 9o 7 62 2410 314 175
90 - 120 3 49 3092 279 129

120 - 150 2 31 3072 201 96
150 - 210 1 30 3164 194 95
210 - 270 1 35 4200 193 89

 

 

89

Table 15 , Extractable phosphorus concentrations in the soil (profile '
for Spring of 1973 and Spring and Fall of 1977 and 1978.

 

Depth (cm) 4-73 4-18-77 11-11-77 4-18-78 11-22-78

 

 

 

(Hg P/g soil)

0 - 15 9 107 98 109 126
15 - 30 8 77 44 65 36
30 - 60 5 7 10 13 14
60 - 90 3 4 4 6 7
9O - 120 3 2 3 5 ' 3

120 - 150 2 1 3 3 2
150 - 210 2 1 2 2 1
210- 270 2 1 2 2 1

 

 

90

each spring than the subsequent fall. The concentration of K+ in
the soil in spring 1978 were greater than those in fall 1977.

Sodium behaved the opposite of K+. The concentration was higher
in fall than spring. This is consistent with a build-up of Na+
during the irrigation season fellowed by leaching during the winter
and spring snow melt. Surface samples taken in each subplot showed
little real variation. Each year, the subplots received the same
fertilizer treatment so any effects which were additive would have
been able to accumulate. The irrigation rate affected K+ accumulation
in samples taken in 1978. The K+ concentration was less in the plots

a+2 and

irrigated at 200 cm/year. In 1978, the concentration of K+, C
Mg+2 were lower in the corn plots. Consistently, the K+ concentration

was greater in soil from subplots receiving additional K fertilizer.

The Depressive Effect of Grasses Upon
the Concentration of N05

 

As the data presented in Table 2 show, the concentration of NO}
in tiles under grass was less than that under corn. These mean values,
while statistically significant, do not tell the whole story. As can
be seen from Figures 6 and 7, the concentration of N03 in tiles under
corn was much more variable than that under grass. The concentration
of N03 under grass was fairly constant throughout the growing season,
whereas the concentration of the ion under corn was typically quite
high in the spring and fall and only approached the concentration
found in the tiles under grass during the middle of the season, when
the growth of corn was greatest. The mean concentrations of N03

in wells under grass and corn point out the disparity between the

 

 

 

.Kemp ea mapaa cave
leave»? 03» pm maoeu oz» cone: Lopez mmmcwmcu mp?» mg» cm Zimoz mo cowpmcpcmucou .o acumen

 

    
   

 

 

 

 

Dam gem omm omm DEWITY _oap cup om, om_
d u d - - - a -
\‘IIIIIIIII'II'I’ l-IIIIII.II.III-Ill\“'llll.lullllll|l
\\ (III-II
1
llllmm<mw
mp0: Eu OON . zmoo
mu
0mm CNN 0mm 0mm o_.N 02. ct. cm—
1 u d u q a u a
ll." I'll-5",, \I’ IIII|.-I.I."’
I'lllu'lul‘“ 1‘ I lull-II“ [I'llllll'llll
(II-III
L

 

111.1 mmgw
muoE Eu oo_.

 

zmou

 

o
m
o_ m”
N
_
3
m
ON
0
1m,
p
.oF _mw
N
.
3
m

om

 

92

 

.mmmp cw move; :0?“

 

 

 

 

 

 

-mmwccw 03» pm maocu 03¢ cone: Laue: mmmcmmcu mp?“ on» cw zimoz mo cowumcpcwucoo .n mesmwm
mama cmwpaa
omm cum omm omm oFN cop oNF omp omp
_ - _ fl - q _ -
still
4
mpoFa Eu com
alllwmgu
zmoo g
omm cum omm omm opm omp cup omp om_
- . 141 q q . . .
l'Illll'll'IIIIlllllllll" \r’lllu‘Il-II-
ll, \
IIIIIII.\
4
1111 mm<¢¢
mpopa Eu coy
zmou 1

 

op

ow

op

om

N03” (P131111

"03'" (Ppm)

93

two concentrations even more dramatically. In 1977, the mean con-
centration of NOS-N in water taken from wells located in grass
plots was 1.04 ppm whereas in samples taken from wells in the corn
plots the mean concentration was 11.15 ppm. By implication the
soil water leached to the groundwater from soil cropped in corn
contained substantially more N05 than that from soil under grass.

There are four possible explanations for this disparity in the
solution concentration of N03. (1) The corn may have simply taken
up less of the nutrient than grass. (2) As mentioned in the litera-
ture review, many authors have suggested that nitrification inhibition
acts to reduce the N03 content below grasses. Perhaps this was the
cause. (3) Theron (1965) has suggested that plants, specifically
grasses, are capable of reducing the net mineralization of soil
organic N. If mineralized organic N was a contributor to the soil
solution N then such an inhibition mechanism by grasses could reduce
NOS-N loss. (4) Soils cropped with grass have been shown to have a
greater denitrification potential than soils under other crops.
Perhaps enhanced denitrification under grass was the cause of the
relatively lower N03 concentration under grass.

As was pointed out earlier, corn took up more N than bromegrass.
This is opposite of what one would expect if crop uptake was the
cause of the disparity. In fact, the differences in mean N03 tile
water concentrations were more pronounced in 1977 than in 1978. And
while corn took up more N than grass in both years, the difference

was greater in l977. Cr0p uptake is a highly unlikely explanation

 

for the differences in tile water N03 concentration.

94

I was unable to adequately evaluate the possibility that increased
net mineralization of N was responsible for this disparity. There was
substantially more organic N in the soil than either of the major

15N

inorganic N forms, N03 and NH2. Broadbent et a1. (1977) using
in a study of nitrification-denitrification relations in wastewater
amended soils, has shown that mineralized organic N is leached from
irrigated soil columns as NOS-N. Small changes in the net minerali-
zation rate could easily surpass any affects by the transformations

of NH: and N03. One method to evaluate net mineralization is

to monitor the changes in organic N over the course of the season.

This was done during the 1978 season. The changes in N03 concentra—
tion which were to be explained were small, hence high sensitivity in
discerning the concentration of organic N was required. The variability
of organic N unfortunately is very large. For example, in order to
detect a real difference in the concentration of the mean organic
nitrogen of 25 ug N/g soil, one would have to evaluate 125 samples

of soil for each treatment. Thus, while the tendency of the organic

N from the surface horizon seems to indicate that the loss of organic

N was greatest in the corn soil, the results for the entire profile

are not statistically significant (Table 16). A circuituous indica-
tion of the relative mineralization rate could be inferred from

the invariability and paucity of NH: in the soil profile (Table 17).

NH: is the major product of mineralization. Small changes will tend

to indicate small rates of mineralization. Generally, the distribution

of NH: in the soil profile was homogeneous.

 

 

95

Table 16. Mean organic N concentration in soil sampled at three
depths and cumulatively to 60 cm+.

 

Mean Organic N Concentration

 

 

 

 

Depth Grass - 100 Grass - 200 Corn - 100 Corn - 200 LSD(0.05)
(cm) (kg/ha)
0 - 15 2433 2754 1952 2079 299
15 - 30 907 1420 1160 1142 NS
30 - 60 725 614 1002 1184 NS
4114 _4405

Z 4065 4788

 

+$611 was sampled 10-77, 4-78, 9-78.

. nh'n.
la
_.... _—_...._.....---

 

 

 

Amzv 0.. o.~ cum- opm

 

 

 

 

 

 

m.o m._ ~.N m.o m._ ~.m Amzv m.o m.o OFN- om_
Amzv m.F ~._ Amzv m._ m.~ Amzv _._ m.F amp- om,
% as E 3 , $5 5 3 35 3. 9o 87 8
Amzv N._ N.N Amzv o.F ~.~ Amzv m.o c.F om - om
Amzv N.N ¢.N Amzv _.N m.m Amzv _.F m.F co - om
Amzv N.m m.~ Amzv o.N m.m Amzv ¢.P m.o om - m_
Amzv N.P a.e Amzv ~.~ _.m Amzv N.F m._ m_ - o
APPOm m\z-wxz an“ IAEUYI
Amo.ovom4 ccou mmacu Amo.ovom4 ecou macaw Amo.ovom4 ccou mango guamo
Fpmu Lossam mcwggw

 

.commmm mcwzosm
mnmp any we mnomgma mecca cow mcuamu vmumuwucw um cmpasmm Fwom cw mcowpmgpcwucou znezz com: .mp mFth

97

Though the subject is controversial in the literature, I think
the evidence holds that some plants have a depressive effect upon
nitrification, but that this caused the lower N03 concentration here
is doubtful. First, Broadbent et a1. (1977) has stated that under
sewage irrigation nitrification of NH: was essentially complete.
Second, the changes in NH: concentration in the soil above the tile
are not statistically significant. Furthermore, the trends of the
average concentrations of NH: indicate enhanced nitrification under
corn, not bromegrass. Lastly, the changes, in whatever direction,
are too small to account for the measured differences in N03 concen-
tration.

Denitrification in field soils may be evaluated in many ways.
Five of these approaches were used in the simulated wastewater study
experimental site. Two of the estimates of N03 loss used the
concentration of C1' in samples in the irrigation water, tile
drainage water and soil. Because N03 and Cl' have the same charge
and hydrated radii, the physical behavior of the two ions in soils
is virtually identical. Changes in the concentration of each ion
in the soil due to physical causes, especially dilution and con-
centration due to wetting and drying, will be the same. Chloride
was used to estimate the volume of N03 leached from the site in
calculating a N balance for the site. The method used was similar
to that used by Pratt et a1. (1978). Their method was based on
the assumption that Cl' is a conservative ion in the soil water
of irrigated fields. They also assumed that Cl' flowed concommitant

with water through the soil profile. If these assumptions are true,

 

98

then

VIC]I = VLClL

where: VI volume of irrigation water

('3
_.I
II

I concentration of C1' in the irrigation water

<
ll

L volume of leachate water

0
_a
ll

L concentration of C1' in leachate

The amount of N03 lost in the leachate was calculated by multiplying
the NOS-N concentration in the tile drainage with the leachate volume,
VL‘ There are several possible problems with this approach. (1)
Natural sources of 01' would increase the estimate of leachate

volume. Since the mean 01' concentration in the irrigation water

was quite high (270 ppm) this was probably not a problem. (2) Anion
exclusion would make the Cl' move more rapidly in the soil than the
carrier water. Since N03 and Cl' move identically, while the estimate
of VL may be in error, the amount of N03 calculated by this method
should be correct. (3) If a steady state in the flow of the ions

in the leachate water has not been achieved, erratic or misleading
results would be calculated. Obviously, a steady state is rarely
found in the ion flux through field soils. Unfortunately, there

are no criteria for what variation in the N03 and Cl' concentrations
may be allowed for the steady state assumption to be invoked.

In their study, Pratt et a1. (1978) stressed the importance of
determining the variability of inorganic N in soil samples before

a N study is established, but gave no comprehensive guidelines.

 

99

A soil N balance for each crop at each irrigation rate was
calculated for the 1978 growing season. This balance is presented
in Table 18. Soil accumulation (or loss) or N was calculated using
the NOE-N and NHZ-N concentrations in the top 120 cm of the soil
in spring and fall of 1978. "Leaching" losses were calculated
using the method of Pratt et a1. (1978). The calculated accumula-
tion of inorganic N during the 1978 growing season was greater
in the soil cropped in bromegrass than that in corn. The unaccounted
losses of N were greater in grass soils than the corn soils.

The second use of C1" in field studies of N03 movement is to
calculate the ratio of the concentration of the two ions in soil
samples collected at various depths in the soil profile. Focht (1978)
in his review of the various methods to assay denitrification argued
that this method was more qualitative than quantitative. He thought
that the method could be used to indicate denitrification in soils
where N balance studies found unaccounted N losses. Plots of
Nog-N/Cl' ratios in soils sampled at various depths in 1978 are
presented in Figures 8 and 9. In both the corn and bromegrass
plots, the concentration of N03 dropped relative to the concentration
of Cl'. This loss substantiates the N loss in the soil profile indica-
ted by the N balance study, possibly due to denitrification. The
relative constancy of the ratios lower in the profile, especially
later in the year, indicates that the steady state assumption, used
in the calculation of the N balance, may have been valid.

He can expect that the soil at the experimental site was saturated

much of the time. If this is true, then the readily available organic

 

 

100

Table 18. Estimated gains and losses of nitrogen by the soil in 1978.

 

 

 

 

~Grass Corn
100 cm 200 cm 100 cm 200 cm

Irrigation 152 305 152 305
Fertilization 56 56 93 93

Total Application+208 361 245 398
Plant Uptake 105 (50)1 133 (37) 153 (62) 159 (46)
Soil Accumula-

tion (Loss) 23 (11) 4o ( 6) (l6) (-7) 5 ( 1)
Tile Loss 41 (19) 76 (21) 76 (31) 65 (16)
"Leaching" 37 (18) 102 (28) 7o (29) 180 (45)
Total Loss§ 165 (79) 275 (76) 207 (84) 344 (86)
Unaccounted Loss++ 43 (21) 86 (24) 38 (16) 54 (14)

 

+Sum of irrigation and fertilization gains.

INumbers in parenthesis represent N loss expressed as a percentage of
total application.

§Sum of Plant Uptake, Soil Accumulation and Tile losses.

++Total Application less Total loss.

 

101

NO3-N/Cl Ratio

 

50 L

 

D - Under grass
A - Under corn

Depth 100 _
(cm)

 

 

150 _.

 

 

618

 

200 -

Figure 8. Mean N03-N/Cl Ratio in soil at indicated
depths, Spring-1978.

102

N03-N/Cl Ratio

 

  
 

0.10 0.20 0.30
0 l I 1
50 _
DEPTH ..
(cm)

0 - Under grass

100 A - Under corn
r
150 _
200 -

 

Figure 9. Mean N03-N/Cl ratio in soil at indicated
depths, Fall-1978.

103

C would be a main determinant of denitrification rate. Hence,

by a third method, several authors have shown a correlation between
soil organic C and potential denitrification. With this in mind,
the organic C content of the surface soils was determined. The
results are presented in Table 19. The organic C content of

the grass soil was greater than the corn soil.

Table 19. Mean organic carbon content of surface soil (0 - 15 cm)
in 1978.

 

Organic Carbon Content

 

Grass Corn LSD (0.05)
100 cm 200 cm 100 cm 200 cm

 

 

1.76 1.87 1.28 1.47 0.40

 

The correlation between denitrification and organic C is generally
weak. A more direct (fourth) method is to incubate a soil saturated
with a N03 solution. Since assimilation of NOS-N by microorganisms
in saturated soils is generally small, the disappearance of Nog-N
is a measure of potential denitrification of the soil (Focht, 1978).
The potential denitrification in the top 60 cm of the soil collected
in the spring and fall of 1978 was determined. ‘This data is presented
in Table 20. The potential denitrification measured by this method
was greatest in the grass soils. The amount of denitrification
represented by the difference of the denitrification potential of
the two soils was sufficient to explain the differences in N unaccounted

for in the calculation of the N balance.

104

Table 20. Mean percent NOE-N lost after 7 days of incubation.

 

 

 

 

 

 

NOS-N Loss
*Grass Corn
Depth Spring Fall Spring, Fall LSD (0.05)
%
0 - 15 54 23 36 11 8
15 - 30 9 4 7 3 2
30 - 60 2 5 2 1 2

 

The measurement of N20 production from soil following CZHZ
injection is a fifth means of monitoring denitrification. In the
summer of 1980 Phase I denitrification assays were conducted on five
soil samples. Two samples were collected from the abandoned simu-
lated secondary wastewater experimental site. One sample was
collected in what had been a corn plot, another from what had been
a bromegrass plot. As can be seen from Table 21 there was no
significant difference in the denitrification rate in these two
soils presented as "Corn 100" and "Brome 100" respectively. The
denitrification measured in soil collected from an established
bromegrass sod was larger than that in any other soil assayed.

Two soil samples were taken from an irrigated corn plot showed
different denitrification rates. The soil collected at mid-row
("Corn 8") had a significantly higher denitrification than the
soil sampled next to the corn stalk ("Corn A"). The implication,

again, is that denitrification in_situ was greater in soils cropped

105
in bromegrass than soils cropped in corn.

Table 21. Comparison of denitrification rates as determined by
C2H2 inhibition.

 

 

Soil Denitrification Rate
ugN.g soil-1. hr-1
Bromegrass 11.74 a
Corn A 1.08 b
Corn 8 2.51 c
Brome 100 4.35 c
Corn 100 1.82 bc

 

+Rates followed by same letter are not statistically
different at the 5% level of probability as determined by the
Mann-Whitney test (Snedecor and Cochran, 1967).

Thus, I discount nitrification inhibition and crop uptake as
the major cause of the difference in N03 concentration in drainage
water. Although indications are that net mineralization of organic
N was not responsible fbr the differences in N03 concentration, the
possibility cannot be discounted. Evidence from several lines
of inquiry suggests that enhanced denitrification under grass may

have been responsible for most of the differences in N03 concentra-

tion under the two crops.

CHAPTER IV

SUMMARY AND CONCLUSIONS

Simulated secondary municipal effluent was applied by sprinkler
irrigation to a tile drained soil cropped in bromegrass and corn.
The effluent was applied at two rates, 100 and 200 cm/season. The
efficacy of additional N and K fertilization was also tested.

The yield and nutrient concentration of the crops were measured.
The nutrient concentration was determined in the soil, tile drainage
water, well water and irrigation water. The flow rate of water in
the tiles was monitored continuously.

l. The concentration of both phosphate and N03 in water
draining from tiles under grass was less than that under corn.

2. Between 50 and 75% of the irrigation water was lost through
the tile drains.

3. There was a strong linear relation between the concentra-
tions of K and phosphate P in the tile drainage water. The concentra—
tion of these ions in the tile water varied during an irrigation.
This peak was roughly 1/2 hour before peak flow in the tile.

4. The concentration of N03 in the tile water was found to
be less than that calculated for the soil solution.

5. Mainly because its dry matter production was greater, the
nutrient removal by corn was greater than that by bromegrass.

6. Additional N, but not K, fertilization increased yield and

nutrient removal by corn.

106

107

7. The combination of additional N and K fertilization increased
yield and nutrient removal by bromegrass.
8. The yield and nutrient removal of bromegrass, but not corn,

was increased at the greater irrigation rate.

BIBLIOGRAPHY

Alexander, M. 1977. Introduction to soil microbiology. 2nd ed.
John Wiley and Sons, New York.

Allison, F. E. 1966. The fate of nitrogen applied to soils.
Advances in Agronomy 18:219-258.

Avnimelech, Y., S. Dasberg, A. Harpaz and I. Levin. 1978.
Prevention of nitrate leachate from the Hula basin Israel:
A case study in watershed management. Soil Science 125(4):
233-239.

Bailey, L. D. 1976. Effects of temperature and root of denitifi-
cation in a soil. Canadian Journal of Soil Science 56(2):
79-87.

Barber, 0. A. and J. K. Martin. 1976. The release of organic
substrates by cereal roots into soil. New Phytologist 76(1):
69-80.

Bargh, B. J. 1977. Output of water, suspended sediment and
phosphorous and nitrogen forms from a small forested catch-
ment. New Zealand Journal of Forestry Science 7(2):l62-l7l.

Bargh, B. J. 1978. Output of water, suspended sediment and
phosphorous and nitrogen fOrms from a small agricultural
catchment. New Zealand Journal of Agricultural Research
21 1 :29-385.

Beek, J., F. A. M. deHaan and H. H. Van Riemsdijk. l977. Phosphates
in soils treated with sewage water: I. General information
on sewage farm, soil, and treatment results. Journal of
Environmental Quality 6(1):4-7.

Bohn, H. L., B. L. McNeal and G. A. O'Conner. 1979. Soil chemistry.
Hiley-Interscience, New York.

Bolt, G. H. 1967. Cation-exchange equations used in soil science -
A review. Netherlands Journal of Agricultural Science 15(2):
81-103.

Bolt, G. H., M. G. M. Beuggenwert and A. Kamphorst. 1976. Adsorption
of cations by soil. p. 54-90. In_G. H. Bolt and M. G. M.
Bruggenwert (eds.). Soil Chemistry. Elsevier Scientific,

New York.

Brar, S. S. 1972. Influence of roots on denitrification. Plant
and Soil 36(3):713-715.

108

109

Broadbent, F. E., D. Pal and K. Aref. Nitrification and
denitrification in soils receiving municipal wastewater.
p. 321-348. In F. M. D'ltri (ed.). Wastewater renovation
and reuse. MSFcel Dekker, New York.

Burford, J. R. and J. M. Bremner. 1975. .Relationships between the
denitrification capacities of soil and total, water-soluble
and readily decomposable soil organic matter. Soil Biology
and Biochemistry 7(6):389-394.

Cameron, 0. R., C. G. Kowalenko and K. C. Ivarson. 1978. Nitrogen
and chloride leaching in a sandy field plot. Soil Science
126(3):l74-l79.

Childs, E. C. 1943. The water table, equipotentials and streamlines
in drained land. Soil Science 56:317-330.

Childs, E. C. 1947. The water table, equipotentials and streamlines
in drained land: The moving water table. Soil Science 63:361-376.

Childs, E. C. 1957. The physics of land drainage. p. 1-78. In
J. N. Luthin (ed.). Drainage of agricultural lands. AmeriEan
Society of Agronomy, Madison, WI.

Cosgrove, D. J. 1977. Microbial transformations in the phosphorous
cycle. Advances in Microbial Ecology 1:95-134.

Currie, J. A. 1961. Gaseous diffusion in the aeration of aggregated
soils. Soil Science 92(1):40-45.

Denmead, O. T., J. R. Simpson and J. R. Freney. 1974. Ammonia flux
into the atmosphere from a grazed pasture. Science 185:609.

During, C. 1972. Fertilizers and soils in New Zealand farming. A. R.
Shearer, wellington, New Zealand.

Engelund, F. 1957. The water table in equilibrium with rainfall or
irrigation water. p. 113-138. In_J. N. Luthin (ed.). Drainage
of agricultural lands. American Society of Agronomy,

Madison, WI.

Feigin, A., H. Bielorai, Y. Dag, T. Kipnis and M. Giskin. 1978. The
nitrogen factor in the management of effluent-irrigated soils.
Soil Science 125(413248-254-

Focht, D. D. 1978. Methods for analysis of denitrification in soils.
p. 433-490. IQ_D. R. Nielsen and J. G. MacDonald (eds.). Nitrogen
in the environment. Vol. 2. Academic Press, New York.

Focht, D. D. and N. Verstraete. 1977. Biochemical ecology of
nitrification and denitrification. Advances in Microbial
Ecology 1:135-214.

110

Gakstatter, J. H. and M. O. Allum. 1978. A survey of phosphorous
and nitrogen levels in treated municipal wastewater. Journal
of the Water Pollution Control Federation 42:718-722..

Garcia, J. L. 1975. Effect rhizosphere du riz sur la denitrification.
Soil Biology and Biochemistry 7(2):l39-14l.

Gast, R. G., N. N. Nelson and J. M. MacGregor. 1974. Nitrate and
chloride accumulation and distribution in fertilized tile-
drained soils. Journal of Environmental Quality 3(3):209-213.

Greenwood, D. J. 1961. The effect of oxygen concentration on the
decomposition of organic materials in soil. Plant and Soil
14(4): 360- 376.

Groenevelt, P. H. 1971. Onsager's reciprocal relations. Search
2(8):264-267.

Groenevelt, P. H. and G. H. Bolt. 1969. Non-equilibrium thermo-
dynamics of the soil-water system. Journal of Hydrology 7:358-
388.

Groenevelt, P. H., O. E. Elrick and T. J. M. Blom. 1978. Coupling
phenomena in saturated homo-ionic montmorillonite: III.
Analysis. Soil Science Society of America Journal 42(5):67l-
674.

Holford, I. C. R. and G. E. G. Mattingly. 1975. The high- and
low-energy phosphate adsorbing surfaces in calcareous soils.
Journal of Soil Science 26: 407- 417.

Howeler, R. H. and Bouldin, D. R. 1971. The diffusion and consumption
of oxygen in submerged soil. Soil Science Society of America
Proceedings 35(2):202-208.

Huntjens, J. L. M. 1971a. Influence of living plants on immobiliza-
tion of nitrogen in permanent pastures. Plant and Soil 34:393-404.

Huntjens, J. L. M. 1971b. The influence of living plants on
mineralization and immobilization of nitrogen. Plant and
Soil 35:77-94.

Ivanov, Yu. 1., Ye. M. Karlov and T. Kasatkina. l977. Drainage
water as an indicator of changes in drained soils. (In Russian)
Pochvovedeniye 5:101-104.

Jackson, M. L. 1958. Soil chemical analysis. Prentice-Hall, Engle-
wood Cliffs, N.J.

111

Jacobs, L. H. 1977. Sewage wastewaters - characteristics and
management. IQ_L. H. Jacobs (ed.). Utilizing municipal
sewage wastewaters and sludges on land for agricultural
production. U.S.D.A. North Central Regional Extension
Publication No. 52.

Johnen, B. G. and 0. R. Sauerbeck. 1977. A tracer technique for
measuring growth, mass and microbial breakdown of plant roots
during vegetation. In Soil organisms as components of eco-
systems. Ecology Bulletin (Stockholm) 25:366-373.

Jones, L. H. 1974. Effect of soil depth on denitrification. Soil
Science 118(4):280-281.

Kardos, L. T. and J. E. Hook. 1976. Phosphorous balance in sewage
effluence treated soils. Journal of Environmental Quality
5 1 :87-90.

Karlen, D. L. 1975. Renovation of simulate municipal wastewater
through intensive irrigation of corn grown on a tile drained
Conover loam. M. S. Thesis, Michigan State University.

Karlen, D. L., M. L. Vitosh and R. J. Kunze. 1976. Irrigation
of corn with simulated municipal sewage effluent. Journal of
Environmental Quality 5(3):269-273.

Katchalsky, A. 1963. Nonequilibrium thermodynamics. International
Science and Technology. p. 43-49.

Katchalsky, A. and P. F. Curran. 1965. Nonequilibrium thermodynamics
in biophysics. Harvard University Press, Cambridge, MA.

Kharaka, Y. K. and H. C. Smalley. 1976. Flow of water and solutes
through compacted clays. American Association of Petroleum
Geologists Bulletin 60:973-980.

King, L. D. 1973. Mineralization and gaseous loss of nitrogen in
soil-applied liquid sewage sludge. Journal of Environmental
Quality 2(3):356-358.

Kirkham, D. 1958. Seepage of steady rainfall through soils into
drains. Transactions of the American Geophysical Union 39:829-
908.

Kirkham, D. 1964. Physical artifices and formulas for approximating
water table fall in tile-drained land. Soil Science Society
of America Proceedings 28(5):585-590.

Kirkham, D. and N. L. Powers. 1972. Advanced soil physics. Wiley-
Interscience, New York.

112

Kissel, D. E. and S. J. Smith. 1978. Fate of fertilizer nitrate
applied to coastal bermudagrass on a swelling clay soil.
Soil Science Society of America Journal 42(1):77-80.

Kohl, D. H., F. Vithayathil, P. Hhitlow, G. Shearer and S. H. Chien.
1976. Denitrification kinetics in soil systems: Significance
of good fits of data to mathematical forms. Soil Science
Society of America Proceedings 40(2):249-253.

Kolenbrander, G. J. 1969. Nitrate content and nitrogen loss in
drainwater. Netherlands Journal of Agricultural Science
17:246-255.

Kuo, S. and E. G. Lotse. 1972. Kinetics of phosphate adsorption by
calcium carbonate and Ca-kaolinite. Soil Science Society of
America Proceedings 36(5):725-729.

Lance, J. C. 1977. Phosphate removal from sewage water by soil
columns. Journal of Environmental Quality 6(3):279-284.

Lance, J. C. and F. D. Whisler. 1972. Nitrogen balance in soil
columns intermittenly flooded with secondary sewage effluent.
Journal of Environmental Quality 1(2):181-186.

Lance, J. C. and F. D. Whisler. 1976. Stimulation of denitrification
in soil columns by adding organic carbon to sewage waters.
Journal of the Hater Pollution Control Federation 48:346-356.

Lance, J. C., F. D. Whisler and R. C. Rice. 1976. Maximizing
denitrification during soil filtration of sewage water. Journal
of Environmental Quality 5(1):102-107.

Larsen, S. 1967. Soil phosphorous. Advances in Agronomy 19:151-210.

Lauer, D. A., D. R. Bouldin and S. D. Klausner. 1976. Ammonia
volatilization from dairy manure spread on the soil surface.
Journal of Environmental Quality 5(2):l34-14l.

Lee, D. H. K. 1970. Nitrates, nitrites,and methmoglobinemia.
Environmental Research 3(5-6):484-511.

Levin, I. and Y. Leshem. 1978. Using forage crops to reduce nitrate
accumulation in Hula peat soils. Plant and Soil 50(2):419-426.

Lindsay, H. L. and E. C. Moreno. 1960. Phosphate phase equibria in
soils. Soil Science Society of America Proceedings. 24(3):
177-182.

113

Low, A. J. 1973. Nitrate and ammonium nitrogen concentration in
water-draining through soil monoliths in lysimeters cropped
with grass or clover or uncropped. Journal of the Science
of Food and Agriculture 24(12):1489-1495.

Luthin, J. N. and A. Haig. 1972. Some factors affecting flow
into drainpipes. Hilgardia 41(10):235-245.

Lyon, T. L., J. A. Bizzell and P. D. Wilson. 1923. Depressive
influence of certain higher plants on the accumulation of
nitrates in soil. Journal of the American Society of
Agronomy 15(11):457-467.

MacLean, A. J. 1977. Movement of nitrate nitrogen with different
cropping systems in two soils. Canadian Journal of Soil
Science 57(1):27-33.

McKelvey, V. E. 1973. Abundance and distribution of phosphorous
in the lithosphere. pp. 13-31. In E. S. Griffith, A.
Beeton, J. M. Spencer and D. T. MTEchell (eds.). Environ-
mental phosphorous handbook. Wiley-Interscience, New York.

Meek, B. 0., L. 0. Grass and A. J. MacKenzie. 1978. Chemical
characterization of the gaseous and liquid environments of
subsurface drain systems. Soil Science Society of America
Journal 42(5):693-698.

Metcalf and Eddy, Inc. 1972. Wastewater engineering. McGraw-Hill,
New York.

Miller, D. G. 1960. Thermodynamics of irreversable processes.
Chemical Review 60:15-37.

Moore, 0. R. E. and J. S. Waid. 1971. The influence of the washings
of living roots on nitrification. Soil Biology and Biochemistry
3 1 :69-83.

Munroe, P. 1976. Inhibition of nitrifiers by grass root extracts.
Journal of Applied Ecology 3:231-238.

Murrman, R. P. and M. Peech. 1969. Relative significance of labile
and crystalline phosphates in soil. Soil Science 107(4):249-255.

National Research Council. 1978. Nitrates: An environmental
assessment. National Academy of Sciences, Washington, D.C.

Nie, N. H., C. H. Hull, J. G. Jenkins, K. Steinbrenner and D. H.
Bent. 1975. Statistical package for the social sciences.
2nd ed. McGraw-Hill, New York.

114

Nielsen, D. R., R. D. Jackson, J. W. Cary and D. D. Evans (eds.).
1972. Soil water. American Society of Agronomy, Madison, WI.

Nommik, H. 1956. Investigations on denitrification in soil. Acta
Agriculturae Scandinavica 6:195-227.

Novozamsky, I. and J. Beek. 1976. Common solubility equilibria in
soils. p. 96-125. In_G. H. Bolt and M. G. M. Bruggenwert
(eds.). Soil Chemistry. Elsevier Scientific, New York.

Noy, J. and A. Feinmesser. 1977. The use of wastewater for agricul-
tural irrigation. p. 73-92. In_H. I. Shuval (ed.). Water
renovation and reuse. Academic Press, New York.

Nye, P. H. and P. B. Tinker. 1977. Solute movement in the soil-
root system. Blackwell Scientific Publications, Oxford.

Overman, A. R. and A. Nguy. 1975. Growth response and nutrient
uptake by forage crops under effluent irrigation. Communications
in Soil Science and Plant Analysis 6(1):81-93.

Parfitt, R. L. 1978. Anion adsorption by soils and soil materials.
Advances in Agronomy 30:1-49.

Parfitt, R. L., R. J. Atkinson and R. S. C. Smart. 1975. The
mechanism of phosphate fixation by iron oxides. Soil Science
Society of America Proceedings 39(6):837-84l.

Patriguin, D. G. J., C. MacKinnon and K. I. Wilkie. 1978. Seasonal
patterns of denitrification and leaf nitrate reductase activity
in a corn field. Canadian Journal of Soil Science 58(2):283-285.

Payne, W. J. 1973. Reduction of nitrogenous oxides by microorganisms.
Bacteriology Review 37(4):409-452.

Pratt, P. F., L. S. Lund and J. M. Rible. 1978. An approach to
measuring leaching of nitrate from freely drained irrigated
fields. p. 223-256. In D. R. Nielsen and J. G. MacDonald
(eds.). Nitrogen in thé'environment. Vol. 1. Academic Press,
New York.

Purchase, 8. S. 1974. Evaluation of the claim that grass root
exudates inhibit nitrification. Plant and Soil 41(3):527-
539.

Reddy, K. R. and W. H. Patrick, Jr. 1975. Effect of alternate
aerobic and anaerobic conditions on redox potential, organic
matter decomposition and nitrogen loss in a flooded soil.
Soil Biology Biochemistry 7(2):87-94.

115

Reneau, Jr., R. B. and D. E. Pettry. 1976. Phosphorous distribution
from septic tank effluent in coastal plain soils. Journal of
Environmental Quality 5(1):34-39.

Rice, E. L. 1976. Allelopathy and grassland improvement. pp. 90-
111. IQ_J. R. Estest and R. J. Tyrl (eds.). The grasses and
grasslands of Oklahoma. Ann. Okla. Acad. Sci. No. 6,

Noble Foundation, Ardmore, OK.

Ryden, J. C., J. K. Syers and R. F. Harris. 1973. Phosphorous in
runoff and streams. Advances in Agronomy 25:1-45.

Schindler, D. W. 1975. Biogeochemical evolution of phosphorous
limitation in nutrient enriched lakes of the Precambrian
Shield. p. 647-664. In J. O. Nriagu (ed.). Environmental
biogeochemistry. Vol. 27' Ann Arbor Science, Ann Arbor, MI.

Sharpley, A. N., J. K. Syers and P. W. O'Conner. 1977. Phosphorous
inputs into a stream draining an agricultural watershed I:
Sampling. Water, Air and Soil Pollution 6(1):39-52.

Sillen, L. G. and A. E. Martell. 1964. Stability constants of
metal-ion complexes. Chemical Society, London.

Smith, M. S., M. K. Firestone and J. M. Tiedje. 1978. The acetylene
inhibition method for short-term measurement of soil denitrifica-
tion and its evaluation using 13N. Soil Science Society of
America Journal 42(4):6ll-615.

Smith, M. S. and J. M. Tiedje. 1979. Phases of denitrification
following oxygen depletion in soil. Soil Biology and Bio-
chemistry 11:261-267.

Snedecor, G. W. and W. G. Cochran. 1967. Statistical methods.
p. 128-131. Iowa State University Press, Ames.

Stumm, W. and J. O. Leckie. 1970. Phosphate exchange with sediments:
Its role in the productivity of surface waters. Advances in
Water Pollution Research. Vol. 2. III-26/l-l6.

Tabatabai, M. A. 1972. Michaelis constants of urease in soils and
soil fractions. Soil Science Society of America Proceedings
37(5):707-710.

Taylor, S. A. and G. L. Ashcroft. 1972. Physical Edaphology. W.
H. Freeman and Co., San Francisco.

Theron, J. J. 1950. Agric. Res. Inst. Ser., Dept. Agric. Union of
S. Africa. No. 16 Science Bulletin No. 288.

116

Theron, J. J. 1963. The mineralization of nitrogen in soils
under grass. South African Journal of Agricultural Science
6:155-164.

Thomas, G. W. and B. J. Barfield. 1974. The unreliability of tile
effluent for monitoring subsurface nitrate-nitrogen losses
from soils. Journal of Environmental Quality 3(2):183-185.

Thomas G. W. and R. E. Phillips. 1979. Consequences of water
movement in macropores. Journal of Environmental Quality
8(2):l49-152.

Van Riemdijk, W. H., F. A. Weststrate and J. Beek. 1976. Phosphates
in soils treated with sewage water: III. Kinetic studies
on the reaction of phosphate with aluminum compounds.
Journal of Environmental Quality 6(1):26-29.

Van Schilfgaarde, J. 1974. Nonsteady flow in drains. In J. Van
Schilfgaarde (ed.). Drainage for agriculture. AmeFTcan
Society of Agronomy, Madison, WI.

Van Wazer, J. R. 1958. Phosphorous and its compounds. Vol. 1.
Wiley-Interscience, New York.

Van Wazer, J. R. 1973. The compounds of phosphorous. pp. 169-177.
In E. G. Griffith, A. Beeton, J. M. Spencer and D. T. Mitchell
(Eds.). Environmental phosphorous handbook. Wiley-Interscience,
New York.

Volz, M. G., M. S. Ardakani, M. S. Schult, R. K. Stolzy and A. D.
McLaren. 1976. Soil nitrate loss during irrigation, enhance-
ment by plant roots. Agronomy Journal 68(4):621-627.

Warncke, D. D. and S. A. Barber. 1974. Nitrate uptake effectiveness
of four plant species. Journal of Environmental Quality 3(1):
28-30.

Wilhelm, E., R. Battino and R. J. Wilcock. 1977. Low-pressure
solubility of gases in liquid water. Chemical Reviews 77:219-
262.

Woldendorp, J. W. 1961. The quantative influence of the rhizosphere
on denitrification. Plant and Soil l7(2):267-270.

Wolman, A. 1977. Public health aspects of land utilization of
wastewater effluents and sludges. Journal of the Water
Pollution Control Federation 49(11):2211-2218.

Zwerman, P. J., T. Greweling, S. D. Klausner and D. J. Lathwell.
1972. Nitrogen and phosphorus content of water from tile
drains at two levels of management and fertilization. Soil
Science Society of America Proceedings 36(1):134-137.

 

 

     

   

mCHmAN
[111111
3

   

   

STATE UNIVE

11 1 Ullllllillllll“

1293 030