I www.epa.gov/ada
United States
Environmental Protection
Agency
Transport and Fate of Nutrients and
Indicator Microorganisms at a
Dairy Lagoon Water Application Site;
AN ASSESSMENT OF NUTRIENT
MANAGEMENT PLANS
Water status tools:
48 tcnsiometers (water head);
10 neutron access tubes (water content).
Soil solution tools:
48 solution samplers (Concentration):
8 tour-probe sensors (E('a).
Office of Research and Development
National Risk Management Research Laboratory, Ada, Oklahi
-------
-------
and of and
at a
AN ASSESSMENT OF NUTRIENT
MANAGEMENT PLANS
APS, US
Office of Research and Development
National Risk Management Research Laboratory, Ada, Oklahoma 74820
-------
Notice
This work was supported through Interagency Agreement DW-12-92189901-0 between
EPA's Ground Water & Ecosystems Restoration Division, National Risk Management
Research Laboratory (Stephen Hutchins, Project Officer) and USDA-ARS's U.S.
Salinity Laboratory (Scott Bradford, Principal Investigator). Although this work was
funded substantially by the U.S. Environmental Protection Agency, it has not been
subjected to Agency review and therefore does not necessarily reflect the views of
the Agency, and no official endorsement should be inferred.
Corresponding author. Scott Bradford, USDA, ARS, US Salinity Laboratory,
450 W. Big Springs Road, Riverside, CA92507; Tel. :951-369-4857; Fax: 951-342-4964;
E-mail: Scott.Bradford(@,ars.usda. gov
-------
Foreword
The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the Nation's land, air, and water
resources. Under a mandate of national environmental laws, the Agency strives to formulate and implement actions leading
to a compatible balance between human activities and the ability of natural systems to support and nurture life. To meet
this mandate, EPA's research program is providing data and technical support for solving environmental problems today
and building a science knowledge base necessary to manage our ecological resources wisely, understand how pollutants
affect our health, and prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory (NRMRL) is the Agency's center for investigation of technologi-
cal and management approaches for preventing and reducing risks from pollution that threatens human health and the
environment. The focus of the Laboratory's research program is on methods and their cost-effectiveness for prevention
and control of pollution to air, land, water, and subsurface resources; protection of water quality in public water systems;
remediation of contaminated sites, sediments and ground water; prevention and control of indoor air pollution; and resto-
ration of ecosystems. NRMRL collaborates with both public and private sector partners to foster technologies that reduce
the cost of compliance and to anticipate emerging problems. NRMRL's research provides solutions to environmental
problems by: developing and promoting technologies that protect and improve the environment; advancing scientific and
engineering information to support regulatory and policy decisions; and providing the technical support and information
transfer to ensure implementation of environmental regulations and strategies at the national, state, and community levels.
EPA currently requires that application of concentrated animal feeding operation wastes to agricultural fields follows
a Nutrient Management Plan (NMP). The tacit assumption is that a well-designed and executed NMP ensures that all
lagoon water contaminants (nutrients and pathogens) are retained or taken up in the root zone so that ground water is
inherently protected. This research was designed to test the assumption that appropriate NMPs are protective of ground
water and to address potential weaknesses in the land-application design and operation processes. A well-designed and
managed NMP was implemented on two 6-by-6-meter plots at a dairy farm in San Jacinto, California, for several years
using different forage and application patterns. The selected site was intensively characterized and instrumented for the
experimental studies. Lagoon water application rates were determined by following NMP guidelines from the National
Resource Conservation Service (NRCS).
Spatial and temporal variations in water, nutrient, and indicator microbe levels at the site were determined using a system
of nested tensiometers and soil solution samplers, neutron probe readings, weighing lysimeters, periodic soil coring, plant
tissue analysis, and measurements of apparent soil electrical conductivity. Along with the field experiments, laboratory
experiments were also conducted, using microbes, lagoon water, well water, and soils from the NMP field site. Microcosm
and batch studies were conducted to quantify microbial retention and survival. Transport experiments were conducted
to quantify the influence of water content, microorganism size, grain-size distribution, and lagoon water composition
on the movement and retention of microbes. Transport parameters were estimated by fitting numerical simulations to
experimental data.
This project is completed and has resulted in multiple journal articles that provide detailed information on the various
aspects of the project. This EPA report summarizes these results.
David G. Jewett, Acting Director
Ground Water and Ecosystems Restoration Division
National Risk Management Research Laboratory
-------
-------
Notice ii
Foreword iii
Contents v
Figures vii
Tables ix
Abbreviations x
Executive Summary xi
1.0 Literature Review 1
Background 1
Review 2
Environmental Contaminants 2
Nutrients and Organics 3
Salts 3
Pathogens 3
Land Application 5
Water Flow 7
Nutrient Transport 8
Pathogen Transport 10
Treatments 12
2.0 Site Characterization for Detailed NMP Studies 16
Background 16
Materials and Methods 16
Results and Discussion 18
Soft Data 18
Hard Data 19
Validation and Simulations 21
Bromide Travel Times 21
Averaging of Bromide Transport 21
3.0 NMP Results - Nutrients 25
Background 25
Materials and Methods 25
Field 25
Water and Nitrogen Mass Balances in the Root Zone 27
Results and Discussion 30
Management Considerations for Salinity 30
Management Considerations for Organic Nitrogen 33
Plant Available N, P, and K 35
Nitrogen Fixation - Alfalfa 2008 36
4.0 NMP Results - Indicator Microorganisms 40
Background 40
Materials and Methods 40
NMP Field Site 40
Analysis and Sampling for Indicator Microorganisms 40
-------
Laboratory Experiments 42
Results and Discussion 43
NMP Field Site 43
Laboratory Experiments 46
5.0 Laboratory Studies Investigating Mechanisms of Colloid Retention 52
Background 52
Materials and Methods 52
Colloids 52
Sand 53
Electrolyte Solution Chemistry 53
DLVO Calculations 53
Batch Experiments 54
Column Experiments 54
Colloid Transport Model for Saturated and Unsaturated Systems 54
Results and Discussion 55
6.0 Mathematical Models to Simulate Pathogen and Nutrient Fate 58
Background 58
Physical Chemical Nonequilibrium (PCNE) Model 59
Dual-Permeability Model 60
Stochastic Stream Tube Model 63
7.0 Summary and Conclusions 66
Quality Assurance 70
8.0 References 71
-------
Figure 1.1. A schematic of the processes and nitrogen species that are involved in the nitrogen cycle 9
Figure 1.2. A flowchart that illustrates the steps that may be used to assess and improve the
performance of lagoon water application sites 12
Figure 2.1. A schematic of the field site. Squares represent two 6x6 rn plots 17
Figure 2.2. A 4 ha map of the vertical component of the apparent soil electrical conductivity in the field
as measured using a mobile remote electromagnetic induction sensor system on a 5x5 m grid. 19
Figure 2.3. North-south and east-west transects of estimated values of Ks on plot 1 20
Figure 2.4. Field measurements of the relative bromide concentration over time for the bromide tracer
experiment 23
Figure 2.5. Field measurements (circle) and 2D-MIM simulation (lines and triangle) of relative bromide
concentration over time at -95, -126 and -155 cm 24
Figure 3.1. Schematic of the field site 26
Figure 3.2. The absolute value of the soil water pressure head (l/?l) in the soil profile as a function of day
after emergence (DAE) for the cyclic and blending water application strategies during the
sorghum 2007 growing season 31
Figure 3.3. Electrical conductivity of the soil solution (ECw) over depth and total dissolved solids (IDS)
load during 2007 for the blending and cyclic water application strategies 32
Figure 3.4. Nitrogen from ammonium (N-NH4) and nitrogen from combined nitrite and nitrate (N-
(NOp+NCL,)) concentration in the soil profile at the end (final) and beginning (initial) of two
consecutive growing seasons (triticale 2007/sorghum 2007 and sorghum 2007/alfalfa 2008).. . 33
Figure 3.5. Three conceptual approaches for nitrogen NMPs based on low (top figure-333 g of N-rrr2),
intermediate (middle figure-1666 g of N-rrr2), and high (lower figure-3333 g of N-rrr2) soil
organic N reservoirs 34
Figure 3.6. The composition, distribution, and amount of organic and inorganic N species (rng L1) in
dairy wastewater during the triticale 2007 growing season 35
Figure 3.7: Soil inorganic and organic N reservoirs ( N1^ and N°ojl) and cumulative values of N uptake
by plant ( N'olant ), exchange to/from organic and inorganic N forms (E0/), N loss to drainage
(N1, • ), supplied organic N ( N° .. ,. ). and supplied inorganic N minus loss to the
drainage " r-r o application ' r K D
atmosphere ( N1 ,. .. - N1. , ) in the root zone are presented over time 37
< application atmosphere '
Figure 3.8. Plant available phosphorus (P-POJ and potassium (K) in the soil profile at the beginning
(initial) of the triticale and end (final) of the sorghum growing seasons in 2007 for the
blending and cyclic water application strategies 38
Figure 3.9. Dry phytomass, cumulative water application and actual evapotranspiration throughout the
five growing cycles of alfalfa in 2008 38
Figure 4.1. Plots of the concentration of Enterococcus (Figure 4. la), fecal coliform (Figure 4.1b),
somatic coliphage (Figure 4.1c), and total E. coll (Figure 4.Id) in soil (S, N g4) as a function
of depth at several different days after emergence (DAE) during the winter triticale growing
season at the NMP field site 45
-------
Figure 4.2. Plots of the gravimetric water content (9D) (Figure 4.2a), and the normalized concentration
of Enterococcus (Figure 4.2b), fecal coliform (Figure 4.2c), somatic coliphage
(Figure 4.2d), and total E, coli (Figure 4.2e) in soil (S/CT| ) as a function of depth at
selected sampling times (initial, and times = 0, 14, 62, 134, and 206 h after wastewater
application) during the summer alfalfa growing season at the NMP field site 47
Figure 4.3. A semi-log plot of the relative total concentration (CT/CTi) as a function of time in batch
survival experiments under sterile (Figure 4.3a) and native (Figure 4.3b) conditions at 80%
water saturation for the various indicator microorganisms 48
Figure 4.4. Breakthrough curves (Figure 4.4a) and retention profiles (Figure 4.4b) for Enterococcus,
total E. coli, and §X174 in a column experiment packed with sterilized field soil 49
Figure 4.5. Plots of the normalized concentration of the indicator microorganisms in soil (S/CT/) as a
function of depth at a sampling time of 24 h after ponded infiltration of wastewater ceased
on the undisturbed soil core from the NMP site 50
Figure 4.6. A semi-log plot of the relative total concentration (CT/CT) of the various indicator
microorganisms as a function of time in the ponded infiltration undisturbed soil column
experiment 51
Figure 5.1 Observed and simulated breakthrough curves of colloids for various saturation levels in 360
urn sand at an ionic strength of 6 mM (Figure 5.la) and 30 mM (Figure 5.1b) 57
Figure 6.1. Plots of simulated breakthrough curves (Figure 6. la) and retention profiles (Figure 6.1b)
when qr was equal to 0.0918 crn min~! and the value of wq, was 0, 0.001, 0.002, 0.004,
and 0.008 cm min ! 63
Figure 6.2. Plots of the simulated breakthrough curves (Figure 6.2a) and retention profiles
(Figure 6.2b) when wq? was 0.001 cm min^and qt was equal to 0.024, 0.046, 0.091,
0.181, and 0.361 cm min l 64
Figure 6.3. (a) Plot of the relative flux concentration, < vC > /( < v > C), at a depth of 10 crn as a
function of time when i/and ka are both stochastic parameters and values of pv,d = -1,
-0.5, 0, 0.5, and 1. (b, c) Corresponding normalized solid phase colloid concentration,
< S > //V;c, and associated variance after 250 min with depth, respectively. Model
parameters that were employed in these simulations were D = 0.0313 cm2 rnin^1,
< v > = 0.313 crn min^1, < kr, > = 0.03 min^1, kr = 0.001 min^1, av, = 1, and CTC, = 1 65
-------
Table 1.1. A summary of measured total suspended solids (TSS), electrical conductivity (EC),
oxidation-reduction potentials (ORP), ammonium (NH4-N), nitrite and nitrate
(N02+N03-N), total Kjeldahl nitrogen (TKN), total phosphorus (TP), total organic carbon
(TOG), and potassium (K) concentrations from several swine, poultry, dairy, and beef
lagoon water samples 2
Table 1.2. Indicator microbial populations in whole lagoon samples from different CAFOs 4
Table 1.3. Nutrient uptake parameters for selected crops (Kellogg et al., 2000) 6
Table 1.4. Hypothetical blending ratio (well water to lagoon water) for various lagoon waters when
winter wheat and summer corn are grown using semi-arid and temperate climates 6
Table 1.5. The estimated environmental loading of total salts and indicator microorganisms when the
various lagoon waters were applied to a 1 rn2 area of agricultural field to meet the nitrogen
needs of corn during a 90 d summer growing season 7
Table 2.1. The hydraulic properties (8s is the saturated water content; 8r is the residual water content;
a is the reciprocal of the air entry pressure; and n is the pore size distribution parameter
of the van Genuchten model) of the three major layers in the upper soil profile of the field
plot. Soft data was estimated from particle size distribution and bulk density data using the
ROSETTA program 20
Table 2.2. Mean and standard deviation of bromide travel time (units in hours, h) at each sampling
depth and culvert pipe location, and associated average and variance values for each depth. . . 22
Table 3.1. Ammonia volatilization from the sprinkler irrigation system and soil surface 29
Table 3.2. Potential and actual evapotranspiration (ET), crop coefficient, rainfall and water application
during the growing season of triticale during winter 2007 (A) and sorghum during summer
2007 (B), DAE is day after emergence 31
Table 3.3. Salts and macro-nutrients of a raw and treated dairy wastewater (DWW). DWW was
sequentially treated by solid separator, sedimentation tank and sand filter 35
Table 4.1. Representative concentrations of indicator microorganisms in raw and treated dairy
wastewater, and the percent removal by treatment 43
Table 4.2. Actual evapotranspiration (ET), rainfall, well water application, wastewater application, and
drainage amounts during the winter (triticale) and summer (sorghum) growing seasons as
a function of day after emergence (DAE) 44
Table 4.3. Fitted parameters (A,O and a) from the time dependent decay model (Equation 4.2) for the
various indicator microorganisms under native conditions at 80% water saturation 48
Table 6.1. Expressions for the second- and third-order moments for breakthrough after a Dirac delta
input according to physical nonequilibrium (PNE), chemical nonequilibrium (CNE), and
PCNE models 61
-------
AWI air water interface
CAFO concentrated animal feeding operation
CML carboxyl modified latex
CNE chemical nonequilibrium
DAE day after emergence
DWW dairy wastewater
DLVO Derjaguin, Landau, Verwey and Overbeek
EC electrical conductivity
ET evapotranspiration
FC fecal coliform
IS ionic strength
MPN most probable number
NMP nutrient management plan
ORP oxidation-reduction potentials
PET potential evapotranspiration
PCNE physical and chemical nonequilibrium
PDF probability density function
PNE physical nonequilibrium
SWI solid water interface
TKN total Kjeldahl nitrogen
TOG total organic carbon
TP total phosphorus
TSS total suspended solids
-------
The Environmental Protection Agency (EPA) currently requires that application of Concentrated
Animal Feeding Operation (CAFO) wastes to agricultural fields follow an approved Nutrient
Management Plan (NMP). The tacit assumption is that a well designed and executed NMP
ensures that all lagoon water contaminants (nutrients and pathogens) are retained or taken up in
the root zone, so that groundwater is inherently protected. Recent research by the EPA, however,
has demonstrated that land application of CAFO lagoon water can cause nitrate contamination of
groundwater at significant depths in short time frames. The research outlined in this report was
designed to test the assumption that a well designed and executed NMP is protective of groundwa-
ter from nutrients, salts, and indicator microorganisms, and to address potential weaknesses in the
land application design and operation processes of NMPs.
The report is divided into seven chapters. Chapter 1 provides a detailed review of the literature
examining the reuse of CAFO wastewater on agricultural fields, with special emphasis on nutrients
and pathogens. A dairy farm in San Jacinto, California, was selected for our NMP research experi-
ments. A well designed and managed NMP was implemented on two 6x6 m plots at this site for
several years to study the fate of nutrients and indicator microorganisms. This study required
a high level of confidence in the water flow and solute transport behavior and therefore detailed
information on the associated soil properties was needed. Chapter 2 provides a description of
our field site and extensive experiments that were conducted to characterize the flow and trans-
port properties at this site. Chapter 3 provides a detailed description of the NMP implemented
at this site and results pertaining to the fate of nutrients and salts. Chapter 4 describes field and
laboratory research that was conducted to study the fate of indicator microorganisms under NMP
conditions. Many gaps still exist with regard to our understanding of microorganism transport and
retention in the field. Laboratory experiments and mathematical model development were there-
fore initiated to overcome some of these limitations. Chapter 5 outlines results from laboratory
studies that examined the coupled effects of pore structure, solution chemistry, and water veloc-
ity on colloid retention in saturated and unsaturated porous media, whereas Chapter 6 describes
the development and application of the mathematical models for colloid transport and retention.
Chapter 7 provides a summary of the research, and the main conclusions and recommendations.
Collectively, this research led to the development of recommendations to improve NMP per-
formance to protect groundwater under CAFO waste application sites from both nutrients and
indicator microorganisms. The main NMP lessons learned from the study include:
• Only minor differences in NMP performance were found between cyclic and blending application
strategies for dairy lagoon water.
« NMPs need to account for the soil organic N reservoir in the N mass balance. This may induce
difficulties in conservative NMP implementation that is protective of the environment due to: i)
difficulties in estimation of mineralization rates and their spatial variability; ii) delayed availability
of the organic N for plant uptake; and iii) continuous mineralization and potential nutrient leaching
during fallow periods.
« NMPs should be designed to meet plant N uptake and to minimize the migration of nutrients
below the root zone. This was achieved in this research by accurately applying inorganic N in dairy
wastewater (DWW) that was treated to remove most of the suspended solids, and minimizing the soil
-------
organic reservoir by applying only a fraction of the plant N uptake with DWW. Additional research is
needed to optimize the use of soil organic reservoirs under other NMPs,
« The use of leguminous crops, such as alfalfa, in NMPs is more complicated than cereal crops. Potential
advantages of leguminous crops are their deeper root system and greater plant N uptake. However,
atmospheric N fixation is a nutrient source for leguminous crops that is difficult to quantify, and this
creates potential challenges for implementing environmentally protective NMPs. Additional research is
therefore needed to optimize NMP performance with leguminous crops.
• DWW contain much higher quantities of salts than typical irrigation water. NMPs that precisely apply
water and DWW to meet evapotranspiration (ET) will therefore accumulate salts in the root zone that
may restrict plant growth, and water and nutrient uptake. If this reduction in ET is not considered at
NMP sites, additional leaching and contaminant migration will occur. This point is strongly dependent
on the salt tolerance of the crop, suggesting that NMPs should use only salt tolerant crops.
• The leaching timing of excess salts below the root zone is a crucial aspect in NMP design because of
continuous mineralization of organic N during fallow periods. In order to minimize N03~ leaching, pre-
irrigations should be scheduled at the end of the growing season, when the soil profile is depleted from
N03" by plant uptake.
« A comprehensive measurement of N mass balance in the root zone requires information on losses to
the atmosphere during irrigation. Atmospheric losses may be minimized by applying DWW during times
that are associated with low potential ET (i.e., early morning), or through drip systems that minimize the
exposure of DWW to the atmosphere.
« Differences in the concentration ratios of N, P, and K between DWW and plant uptake may lead to
accumulation of P and K in the root zone.
» Improve measurements of water and nutrient requirements by the crop to obtain accurate information
on the required timing and quantities for application and to minimize drainage (contaminant flux to
groundwater).
• Increase the water and N use efficiency by irrigating to meet plant uptake requirements using a
high uniformity application system. Minimize runoff and ponding conditions by matching the water
application rate to the soil infiltration rate.
« The transport potential of microorganisms can be significantly reduced by minimizing water leaching
below the root zone and surface water runoff. This can be achieved by: precise estimation of the ET
rate; uniform application of wastewater; and selecting water application timing and quantities based on
considerations of soil permeability and ET.
« Special caution is warranted for NMPs on coarse textured and structured soils, and during water flow
transients where enhanced microorganism transport potential has been reported in the literature.
• Timing of water application should allow for adequate die-off of microorganisms before leaching the root
zone by irrigation or natural precipitation.
» Develop a "hydrological sensitivity index" based on the soil and groundwater properties (depth, quality,
hydraulic properties, and mineralogy of the vadose zone and aquifer). This index should categorize high
and low potential zones of contamination from agricultural activity. Application of liquid and solid dairy
wastes in low sensitivity zones would be more flexible than in other zones.
-------
1.0
Literature Review
Background
New technological innovations and the
economic advantage of size have driven a struc-
tural shift from small to large concentrated
animal feeding operations (CAFOs) in the USA.
It has been estimated that confined livestock
and poultry animals in the USA generate about
453 million tonnes of manure annually (Kellogg
et al., 2000; USEPA, 2003). Water use at
CAFOs includes drinking water for animals and
water used in cooling facilities, sanitation and
wash down of facilities, and animal waste-dis-
posal systems. The United States Geological
Survey estimated that livestock water use
accounted for nearly 1% of the total freshwater
withdrawals (excluding thermoelectric power)
in the USA (Hutson et al., 2004). Daily use
of water for farm animals was reported to be
159 L d'1 for a milking cow, 18 L d4 for a hog,
and 18 L d 1 for 100 chickens. Large volumes
of animal manure-containing wastewater, wash
water, and storrn water runoff can be generated
at CAFOs (USEPA, 2001a). This manure-
contaminated water is typically collected and
stored in wastewater lagoons on farms (USEPA,
2001a).
Transportation, storage, and treatment of
manure and manure-contaminated water are
costly (Gleick, 2000). The large volume of
waste generated, and the lack of disposal area
at CAFOs, further limit the ability for effective
manure management. Manure and wastewa-
ter are, therefore, usually land applied within
about 16 km of CAFO facilities. When applied
to agricultural lands at agronomic rates,
manure and wastewater can be a valuable
fertilizer and soil amendment that can improve
the physical condition of the soil for plant
growth (Jokela, 1992; Kapkiyai etal., 1999),
reduce power required for tillage (Sommerfeldt
and Chang, 1985, 1987), and increase the
organic content of soil (Sommerfeldt et al.,
1988). The reuse of CAFO wastewater in
irrigated agriculture also provides a potential
means to reduce the demand for high qual-
ity water that is a scarce natural resource in
many arid and semi-arid regions (Pimentel
et al., 2004). Conversely, CAFO manure and
wastewater have also been reported to pose a
potential risk to environmental resources and
to human health (Thome, 2007; Burkholder et
al., 2007). Contaminants of potential concern
in animal wastes include excess amounts of
nutrients (Jongbloed and Lenis, 1998; Mallin,
2000), salts (Chang and Entz, 1996; Hao and
Chang, 2003), organics rich in biochemical
oxygen-demanding material (Webb and Archer,
1994), heavy metals (Barker and Zublena,
1996), microbial pathogens (Gerba and Smith,
2005; Schets et al., 2005), antibiotics (Shore
etal., 1988 and 1995; Nichols et al., 1997;
Peterson et al., 2000), and natural and syn-
thetic hormones (Hanselman et al., 2003;
Raman etal., 2004).
This review contains three sections, namely:
(i) Environmental Contaminants; (ii) Land
Application; and (iii) Treatments. The
Environmental Contaminants section reviews
the nutrients and pathogens that are found in
CAFO lagoon water and why they may pose a
risk to the environment and/or human health.
The Land Application section reviews the
current regulatory framework for land applica-
tion of CAFO wastewater that is based upon
Nutrient Management Plans (NMPs), the
implicit assumptions and possible weaknesses
in NMP design and application, illustrative
nutrient and pathogen loading rates at NMP
sites, and potential transport pathways for
nutrients and pathogens. Finally, the section
on Treatments discusses potential best man-
agement practices and lagoon water treatments
that may be needed before land application of
CAFO wastewater to minimize risks and dis-
semination of nutrients and pathogens in the
environment.
-------
Review
Environmental Contaminants
Three classes of potential lagoon water con-
taminants are considered in this section,
namely: (i) nutrients and organics, (ii) salts,
and (iii) pathogens. Tables 1.1 and 1.2 pro-
vide illustrative examples of concentrations
of the various nutrients, salts and indicator
microorganisms that were measured in CAFO
lagoon water from several farms. Considerable
variability is expected in the concentration of
these contaminants at different farms due to
differences in animal and waste management
practices. Hence, the concentration values
provided in these tables should be viewed as
only reflecting site specific conditions at the
indicated farm, and general trends should not
be ascribed to these data. Wastewater from
these lagoons, however, is actually used for
land application at these farms, and therefore
these data can be used to provide an estimate
of potential environmental concentrations for
the various contaminants.
Table 1.1 provides a summary of measured
total suspended solids (TSS), electrical con-
ductivity (EC), oxidation-reduction potentials
(ORP), ammonium (NH4-N), nitrite and nitrate
(N02+N03-N), total Kjeldahl nitrogen (TKN),
total phosphorus (TP), total organic carbon
(TOO, and potassium (K) concentrations
from several swine, poultry, dairy, and beef
lagoon water samples (Hutchins et al., 2007).
Concentrations of indicator microorganisms
in these lagoon water samples are provided
in Table 1.2. For these data, separate analy-
ses were conducted on the samples collected
as described previously (Hutchins et al.,
2007). Microbial indicators were analyzed
on whole lagoon samples using a commercial
most probable number (MPN) method (IDEXX
Laboratories, ME). More general information
on nutrient characteristics and other proper-
ties of various types of lagoon water samples is
also available in the literature (MWPS, 1993;
NCSU, 1994; USDA, 1996; ASAE, 1999;
USEPA, 2001a; Bradford et al., 2008).
Table 1.1. A summary of measured total suspended solids (TSS), electrical conductivity (EC), oxidation-re-
duction potentials (ORP), ammonium (NH4-N), nitrite and nitrate (N02+N03-N), total Kjeldahl nitrogen (TKN),
total phosphorus (TP), total organic carbon (TOO, and potassium (K) concentrations from several swine,
poultry, dairy, and beef lagoon water samples (mean with standard deviation, three locations for each lagoon).
CAFO
Type
Beef
Feedlot2
Dairy2
Poultry1
Poultry2
Poultry3
Swine Sow1
Swine
Finisher1
Swine
Nursery1
TSS
mg
L-1
212
±28
718
±150
847
±15
865
±79
253
±75
1230
±30
5310
±2430
4220
±1940
EC
dS
m-1
2.6
±0.0
3.1
±0.0
11.3
±0.2
8.0
±0.0
4.4
±0.0
12.6
±0.0
19.4
±0.0
21.5
±0.1
ORP
mV
73
±26
-277
±24
-331
±21
-337
±2
207
±28
-348
±16
-368
±2
-368
±5
NH4-N
mg
L-1
33
±3
84
±12
656
±34
289
±11
58
±5
944
±23
1630
±20
1370
±90
IM02+IM03-ISI
mg
L-1
0.2
±0.01
0.5
±0.1
1.2
±0.01
0.5
±0.01
10.4
±2.3
2.8
±0.04
8.0
±0.1
5.7
±0.8
TKN
mg
L-1
63
±1
185
±11
802
±7
407
±2
96
±0
1290
±15
2430
±40
2040
±60
TP
mg
L-1
14
±0
30
±1
50
±1
23
±0
30
±0
264
±5
324
±14
368
±35
TOC
mg
L-1
155
±6
576
±94
1050
±40
374
±15
114
±15
944
±304
4780
±280
1440
±130
K
mg
L-1
277
±2
178
±29
1430
±200
1490
±30
811
±14
137
±2
242
±3
4150
±210
• Primary lagoon;2 - Secondary lagoon;3 - Tertiary lagoon; CAFO - Concentrated Animal Feeding Operation
-------
Nutrients and Organics
Table 1.1 provides illustrative concentra-
tions of nutrients (NH4-N, TKN, TP, and K)
and organics (TOO that were measured in
several swine, poultry, dairy, and beef lagoon
water samples. Ayers and Westcot (1989)
reported severe restrictions on the use of
irrigation water that had total nitrogen values
greater than 30 mg L1 due to potential prob-
lems discussed below. These same authors
also reported that typical ranges for NH^-N,
P04-P, and K+ in irrigation water were 0-5,
0-2, and 0-2 mg L1, respectively. The EPA
has set the Maximum Contaminant Level
(MCL) in groundwater for nitrate (N03~) and
nitrite (NO-) at 10 mg L1 N03-N and^l mg L1
N02-N, respectively. The EPA-recommended
guideline for TOC in drinking water is
4 mg L4. Hence, Table 1.1 indicates that
nutrient and organic concentrations are very
high in these lagoon water samples, relative to
irrigation and drinking water standards.
Published literature also indicates that animal
wastes frequently contain high concentrations
of nutrients (Chang and Entz, 1996; Hao and
Chang, 2003) and organics rich in biochemi-
cal oxygen-demanding material (Webb and
Archer, 1994) that can adversely impact
soil and water quality (Jokela, 1992; Chang
and Entz, 1996; Craun and Calderon, 1996;
USEPA, 1997; USEPA, 2000). Potential
environmental problems from excess amounts
of nutrients and organics in water include
algal blooms, reduced biodiversity, objection-
able tastes and odors, and growth of toxic
organisms in surface waters that are used
for recreation and sources of drinking water
(Mallin, 2000; Burkholder et al., 2007).
These degraded conditions, especially the
associated hypoxia/anoxia and high ammonia,
have caused major kills of freshwater species
(Burkholder et al., 2007). High nitrate levels
in water have been associated with increased
risk of methemoglobinemia for infants (blue-
baby syndrome), as well as diarrhea and
respiratory disease (Ward et al., 2005).
Sate
Table 1.1 provides illustrative concentrations
of EC that were measured in several swine,
poultry, dairy, and beef lagoon water samples.
Ayers and Westcot (1989) reported severe
restrictions on the use of irrigation water that
had EC values greater than 3 dS nr1. The
salinity levels that are associated with these
lagoon waters are very high relative to the
irrigation quality guidelines.
Prolonged exposure of agricultural lands to
animal wastes that have high salinity levels
may alter many soil physical and chemi-
cal properties, and crop yields (Burns et al.,
1985; King et al., 1985; Chang et al., 1991).
For example, Chang et al. (1991) studied the
effects of 11 years of cattle manure addition
on soil chemical properties. Accumulation
of specific ions (soluble sodium, calcium,
magnesium, chloride, sulfate, bicarbonate,
and zinc), nutrients (total N, nitrate, total P,
available P), and organic matter increased
with increasing application rates of manure
and lagoon water. Near the soil surface the
EC and the sodium adsorption ratio (SAR)
increased and the pH decreased with increas-
ing addition rates of manure. Changes in these
soil chemical properties can in turn influence
soil hydraulic properties. For example, high
SAR values are frequently associated with
decreases in the saturated hydraulic conduc-
tivity (Ayers and Westcot, 1989) as a result
of dispersion of the clay (colloidal) fraction
of soils. High levels of EC as well as spe-
cific ions (sodium, chloride, boron, nitrate,
and bicarbonate) can also severely influence
crop yields (Maas and Hoffman, 1977; Maas,
1984) as a result of increases in osmotic
stress and specific ion toxicities.
Pathogens
Animal wastes frequently contain pathogenic
viruses, bacteria, and protozoa that pose a risk
to human and/or animal health (USDA, 1992;
USEPA, 1998; Gerba and Smith, 2005).
Although more than 130 microbial patho-
gens have been identified from all animal
-------
species that may be transmitted to humans
by various routes (USDA, 1992; USEPA,
1998), the most significant manure-borne
zoonotic pathogens are the protozoan parasites
Cryptosporidium pan/urn and Giardia duode-
nalis, and the bacterial pathogens Salmonella
spp., Catnpylobacter spp., Escherichia coli
0157: H 7, and Listen a monocytogenes.
Viruses of potential concern include: poliovi-
rus, coxsackie virus, echovirus, hepatitis A,
rotavirus, and Norwalk virus (Gerba and Smith,
2005),
The EPA has set drinking water goals for
pathogens to be no detection, because of
the low infectious dose for many patho-
gens (Loge et al., 2002) and variability of
individual responses to infection (Gerba,
1996). Regulations to protect public health
from pathogens, however, are largely based
on measured concentrations of indicator
microorganisms such as total or fecal (thermo-
tolerant) coliform (FC). For example, the World
Health Organization (WHO) recommended
standard for use of degraded water for irrigat-
ing crops eaten raw is <1000 FC per 100 ml
(WHO, 2006), whereas in the USA the stan-
dard for unrestricted urban use varies with
the state from non-detectable to 200 FC per
100 ml (USEPA, 2004). Table 1.2 provides
illustrative concentrations of various indica-
tor microorganisms of fecal contamination
(Total Coliforms, Fecal Coliforms, and Fecal
Enterococci) that were measured in several
swine, poultry, dairy, and beef lagoon water
samples. The concentration of indicator
microorganisms in the various lagoon water
samples is very high, and significantly exceeds
the recommended US standards for unre-
stricted irrigation (USEPA, 2004).
Surface and groundwater contamination
by pathogenic microorganisms is common
in many areas of the USA (USEPA, 1997).
Drinking water exposures may occur in vul-
nerable private wells, whereas recreation
exposures and illnesses can happen due to
accidental ingestion of contaminated water
and dermal contact during swimming. Surveys
of waterborne disease outbreaks frequently
demonstrate a farm animal source (Centers
for Disease Control and Prevention, 1998).
Pathogenic microorganisms in groundwa-
ter have been estimated to cause between
750,000 and 5 million illnesses per year in
the USA (Macler and Merkle, 2000). Greater
risks of serious illness occur for the very
young, elderly, pregnant women, the immu-
nocompromised, and those predisposed with
other illnesses (Gerba, 1996).
Table 1.2. Indicator microbial populations in whole lagoon samples from different CAFOs (mean with stan-
dard deviation, three locations for each lagoon).
CAFO
Type
Beef Feedlot2
Dairy2
Poultry1
Poultry2
Poultry3
Swine Sow1
Swine Finisher1
Swine Nursery1
1 - Primary lagoon;
Total Coliforms
cfu per 100 mL
3.89E03±0.99E03
1.91E07±0.16E07
4.73E04±2.09E04
2.42E03 ± O.OOE03
5.83E04 ± 0.28E04
4.89E05±0.70E05
1.02E06±0.12E06
1.87E05±0.95E05
2 - Secondary lagoon; 3 - Tertiary
Fecal Coliforms
cfu per 100 mL
1.57E03±0.28E03
1.08E06±0.11E06
2.91E04±1.07E04
2.42E03 ± O.OOE03
2.03E01 ±0.06E01
3.34E05 ± 0.43E05
± 3.44E05
7.93E04 ± 8.30E04
lagoon; CAFO - Concentrated
Fecal Enterococci
cfu per 100 mL
2.94E03±1.07E03
1.53E05±0.23E05
1.41 EOS ±0.67E05
8.51E04±2.75E04
2.68E03±1.24E03
3.69E05±0.39E05
8.50E05±4.63E05
2.43E05±O.OOE05
Animal Feeding Operation
-------
Manure-contaminated water resources have
also been implicated in food-borne disease
outbreaks on a variety of fresh produce (Gerba
and Smith, 2005). The health impacts
of such outbreaks can be very significant.
For example, the 2006 outbreak of E.coli
0157:H7 on spinach from the Salinas Valley,
California, resulted in 205 illnesses and 3
deaths (USFDA, 2007). Furthermore, loss of
confidence in the safety of agricultural pro-
duce can have significant economic impacts.
For example, the 2005 spinach crop from this
same region was estimated to be worth around
million.
Land Application
The Environmental Protection Agency (EPA)
currently requires that CAFO waste application
to agricultural lands follow approved Nutrient
Management Plans (NMPs). The National
Resources Conservation Service (NRCS)
defines a NMP as, "Managing the amount,
source, placement, form and timing of the
application of nutrients and soil amendments"
(USDA, 2000), The purpose of a NMP is
to meet the nutrient needs of the crop to be
grown, while minimizing the loss of nutrients
to surface and groundwater (USEPA, 2003).
Nutrient management plans are developed to
adhere to state-specific NRCS guidelines. In
general, a NMP is developed by considering
all nutrient input sources (such as manure,
fertilizer, lagoon water, and well water), the
nutrient content at the soil surface, nutri-
ent volatilization losses to the atmosphere
(e.g., nitrogen losses as ammonia), nutrient
mineralization rates, and plant uptake rates
for nutrients. The nutrient concentrations
are typically measured directly for the input
sources, at the soil surface, and in plant
tissues, whereas nutrient volatilization and
mineralization rates are commonly estimated
from literature values. The amount of lagoon
water or other nutrient source that can be
used in a given irrigation cycle is determined
based on nutrient mass-balance consider-
ations for a limiting nutrient (e.g., nitrogen or
phosphorus). A review of several state NMP
guidelines shows that they are typically writ-
ten so as to be protective of surface water
resources, and only provide limited direction
on land application site characterization and
design to protect groundwater from contami-
nation (such as in regions with shallow water
tables).
A brief example of NMP implementation is
given below. In this example we assume
that nitrogen is the limiting nutrient for plant
growth, and use the CAFO lagoon water com-
position information for nutrients presented in
Table 1.1. Nitrate and nitrite only contributed
a small fraction to the total N in the lagoon
water samples (see Table 1.1) and were there-
fore neglected in this analysis. Table 1.3 lists
nutrient uptake parameters for selected crops
(Kellogg et al., 2000) that were used in this
example. For simplicity we also make the
following assumptions: N volatilization loss
of 20%, N mineralization (ammonification)
of 50% during the growing season, initially
zero N in the soil profile and in the well water,
and the only water source to the plants was
irrigation water. Table 1.4 presents the blend-
ing ratios that were calculated to meet the
nutrient (nitrogen) requirements for winter
wheat and summer corn for silage assum-
ing final yields of 27 and 56 tonne-hectare'1,
respectively, and 90 d of full cover crop. We
performed this calculation for two climatic
conditions, inducing different evapotranspira-
tion (ET) fluxes that correspond to semi-arid
and temperate areas. Table 1.4 indicates that
the higher ET flux produces a higher blending
ratio. Lagoon water with a low N concentra-
tion (Dairy and Beef) provides insufficient N
at low ET fluxes and therefore an additional N
source is required (fertilizer) in order to meet
crop needs.
-------
Table 1.3. Nutrient uptake parameters for select-
ed crops (Kellogg etal., 2000).
Crop
Corn for
silage
Sorghum
for silage
Barley
Wheat
Alfalfa
Grass
silage
Nitrogen
kg per ton of
product
3.6
7.4
18.6
20.6
25.2
6.8
Phosphorus
kg per ton of
product
0.5
1.2
3.8
3.6
2.4
0.8
Data in Tables 1.1-1.4 can be used to provide
an illustrative estimate of the environmental
loading for the various contaminants identified
in the lagoon water under NMP application
conditions. As an example, Table 1.5 pro-
vides the estimated environmental loading of
total salts (from EC values given in Table 1.1)
and indicator microorganisms (sum of Total
coliform and Enterococcus in a given row of
Table 1.2) when the various lagoon waters
(blended with well water) were applied to a
1 m2 area of agricultural field to meet the
nitrogen needs of corn during a 90 d summer
growing season with an evapotranspiration rate
of 10 mm d"1. In contrast to Tables 1.1-1.2,
Table 1.5 is dependent on both the contami-
nant concentration in the lagoon water and on
the blending ratio given in Table 1.4.
Table 1.4. Hypothetical blending ratio (well water to lagoon water) for various lagoon waters when winter
wheat and summer corn are grown using semi-arid and temperate climates.
CAFO
Type
Beef Feedlot2
Dairy2
Poultry1
Poultry2
Poultry3
Swine Sow1
Swine Finisher1
Swine Nursery1
Plant
available
N*
mg L'1
41
118
598
290
65
928
1704
1431
Winter wheat
blending ratio
Semi arid
climate
ET=5 mm d~1
insufficient N
0.05
3.88
1.37
insufficient N
6.58
12.93
10.70
Temperate
climate
ET=2.5 mm d~1
insufficient N
insufficient N
1.43
0.18
insufficient N
2.78
5.94
4.82
Summer
blending
Semi arid
climate
ET=10mmd-1
0.93
4.56
27.46
12.79
2.07
43.22
80.21
67.19
corn
ratio
Temperate
climate
ET=5 mm d~1
0.46
2.28
12.90
5.74
0.50
20.59
38.66
32.30
" Total N minus amount due to volatilization and ammonification; 1 - Primary lagoon;2 - Secondary lagoon;3 - Tertiary
lagoon; CAFO-Concentrated Animal Feeding Operation; ET - Evapotranspiration
-------
Table 1.5. The estimated environmental loading
of total salts and indicator microorganisms when
the various lagoon waters were applied to a 1 m2
area of agricultural field to meet the nitrogen needs
of corn during a 90 d summer growing season. An
evapotranspiration rate of 10 mm d^1 was assumed.
CAFO
Type
Total
Salts
gnr2
Total
Indicator
Microbes
number m-
Beef Feedlot2
Dairy2
Poultry1
Poultry2
Poultry3
Swine Sow1
Swine Finisher1
Swine Nursery1
1755.0
458.9
277.8
422.2
1434.8
196.8
163.3
216.0
6.1E+07
3.8E+10
6.2E+07
6.2E+07
2.7E+08
1.8E+08
2.1E+08
5.8E+07
- Primary lagoon;2 - Secondary lagoon;3 - Tertiary
lagoon; CAFO - Concentration Animal Feeding
Operation
Nutrient management plans assume that a
limiting nutrient is the primary environmen-
tal concern. In order to minimize surface
water runoff and leaching to groundwater,
NMPs also imply that water is applied to the
crop to meet the potential evapotranspira-
tion demands. The environmental loading
information presented in Table 1.5 indicates
that significant amounts of salts and indicator
microorganisms will be added to agricultural
lands where a NMP is being implemented.
An implicit assumption of a NMP is that all
CAFO contaminants will be taken up, inacti-
vated, retained, or degraded in the root zone,
so that surface and groundwater are inherently
protected. The validity of these assumptions
for all lagoon water contaminants has not yet
been thoroughly studied. Below we discuss
possible weaknesses in NMP design and appli-
cation, as well as potential transport pathways
and processes for nutrients and pathogens.
Water Flow
The basic premise of a NMP is that water
and nutrients are applied to the root zone
at agronomic rates so that both surface and
groundwater resources will be protected from
all contaminants in lagoon water. Potential
water inputs at a NMP application site include
irrigation water (well and/or CAFO lagoon
water), precipitation, surface water, snow
melt, soil water, and shallow groundwater
tables. The water application amount per
irrigation should be based on measurements
or estimates of potential evapotranspiration
(PET) and water-balance considerations at
the application site. If water input at a NMP
application site equals PET then surface and
groundwater contamination should theoreti-
cally be eliminated because all of the applied
water to the site is used by the crop.
When water in excess of PET is applied to a
NMP site then both surface and groundwater
contamination problems are possible. This
situation can arise from a number of differ-
ent scenarios that may be beyond human
control. Potential problems may arise due
to inaccuracies in the estimation of PET, or
from not accounting for all the sources of
plant available water (perched water tables,
precipitation, snow melt, and surface water
runoff). For example, significant amounts of
unexpected precipitation (a thunderstorm)
after application of nutrients may lead to
surface water runoff and/or leaching of con-
taminants. Other problems can occur due to
nonuniform water application practices during
flood or furrow irrigation or as a result of
spatial variability of soil hydraulic properties
(e.g., different amounts of water are applied to
specific locations in the field) and evapotrans-
piration. Changes in surface topography that
produce surface water runoff and lateral flow
to other locations in the field pose other NMP
difficulties.
Nutrient management plans implicitly assume
that water flow and lagoon water contaminant
transport are controlled by the soil matrix.
-------
Preferential flow is a potentially important
mechanism for water and contaminant trans-
port to bypass portions of the soil matrix and
root zone. Several recent reviews and other
papers provide detailed discussions of the
various processes and conditions leading to
preferential flow (Ritsema et a!., 1993; de
Rooij, 2000; Evans et al., 2001; National
Research Council, 2001; Bodvarsson etal,,
2003; Simunek et al., 2003; Wang et al,
2004), In summary, preferential water flow
can occur as a result of tunneling of water at
textural interfaces (capillary barriers), unstable
flow behavior that is induced by spatial and/or
temporal variations in capillary characteristics
(wettability or hysteresis), dynamic capillary
properties (nonequilibrium capillary pressure
- water content characteristics), macropores,
and fractured systems. All of these factors
can potentially accelerate the movement of
lagoon water contaminants through the root
zone to underlying groundwater. Some mecha-
nisms of preferential flow are also anticipated
to depend on the initial soil water content and
water application rate. For example, soil water
repellency is reported to occur below some
critical water content (Ritsema and Dekker,
2000), hysteresis is induced when infiltra-
tion stops and redistribution occurs (Wang et
al., 2003), and filling of macropores occurs
primarily near saturated conditions (Mohanty
etal., 1997). Improved NMP designs are
needed to provide guidelines to minimize
transport processes that bypass the root zone.
For example, it may be possible to minimize
preferential flow in macropores by applying
irrigation water at rates lower than the satu-
rated conductivity of the soil.
Nutrient Transport
It should be mentioned that NMPs are only as
good as the information on evapotranspiration
and nutrient mass balance. Errors in water
and nutrient application rates, or nonuniform
water application will likely lead to excesses
and/or shortages in lagoon water application
in certain locations. Such errors are poten-
tially responsible for some of the reported
environmental problems associated with
lagoon water application sites (Evans et al.,
1984; King et al., 1985; Sims et al., 1998;
Correll, 1998). Another potential problem
arises when the relative content of nitrogen
and phosphorus in lagoon water differs from
that in the crop. In this case, NMPs that are
designed to meet the nitrogen requirement
for crops may result in the over-application
of phosphorus. Conversely, NMPs based
on crop phosphorus needs may significantly
reduce the lagoon water application rates, and
thus require nitrogen fertilizers to meet plant
needs.
The biogeochemistry of nutrients at NMP
sites can be quite complex. For example,
Figure 1.1 shows a schematic of the processes
and nitrogen species that are involved in the
nitrogen cycle. The most important inorganic
forms of nitrogen are ammonium ions (NH4+),
ammonia gas (NH3), nitrite (N02~), nitrate
(N03"), nitrogen gas (N?), nitrous oxide (N?0),
and nitric oxide (NO). Organic nitrogen com-
pounds in lagoon water occur as urea, amino
acids, amines, purines, and pyrimidines, as
well as a fraction of the dry weight of plants,
microorganisms, detritus, and soils. Organic
and inorganic forms of nitrogen are continually
involved in biogeochemical transformations.
The major processes involved in the nitrogen
cycle are ammonification (mineralization of
organic nitrogen to ammonium), nitrification
(oxidation of ammonia to nitrate), denitrifica-
tion (reduction of nitrate to nitrogen gas),
nitrogen fixation (reduction of nitrogen gas to
ammonia), nitrogen assimilation (conversion of
inorganic nitrogen to organic forms), and nitro-
gen volatilization (conversion of ammonium
ion to ammonia gas).
The biogeochemistry of phosphorus at NMP
sites is also very complex. Most phosphorus
(63 to 92%) in manure and lagoon water
occurs in inorganic forms (Sharpley and
Moyer, 2000). Bohn et al. (1985) presents
a speciation diagram of phosphorus species
of dissolved orthophosphoric acid (H^PO,,)
-------
Inorganic N
Nitrogen
Assimilation
(Plants and
Microbes)
Runoff |
Figure 1.1. A schematic of the processes and nitrogen species that are involved in the nitrogen cycle.
Important inorganic forms of nitrogen are ammonium ions (NH4+), ammonia gas (NH3), nitrite (N02), nitrate
(N03), nitrogen gas (N2), nitrous oxide (N20), and nitric oxide (NO), The dashed lines denote secondary
transport processes.
as a function of pH. The three acid dis-
sociation constants for H3P04at 25°C are
7.5xlO-3 (pKal=2.1), 6.2xlO-8 (pKa2=7.2),
and 5.0xlO-13 (pKa3=12.3). The dominant
phosphate species in the pH range of 2.1-7.2
is H2P04~, whereas HP04~2 is the main spe-
cies in the pH range of 7.2-12.3. Both of
these species are available for plant uptake.
Phosphorus, however, forms complex minerals
with many elements (Arai and Sparks, 2007).
For example, phosphorus species frequently
precipitate out of solution with calcium, iron,
and aluminum ions (Yeoman et al., 1988;
Moore and Reddy, 1994). Furthermore,
phosphorus species can strongly adsorb to
positively charged surfaces in soils (Torkzaban
et al., 2006b; Arai and Sparks, 2007), such
as on metal oxide surfaces and clay edges.
The solubility of phosphorus in soil solution
is, therefore, a complex function of biogeo-
chemical variables and processes such as pH,
oxidation-reduction potential, the concentra-
tions of cations (such as Ca2+, Mg2+, AI3+, and
Fe3+) and phosphorus species, soil mineralogy
and organic matter content, and the mineral-
ization rate of organic (manure) compounds
(Moore and Reddy, 1994; Sharpley, 1995;
Arai and Sparks, 2007). Measurements
of total phosphorus in soils may, therefore,
not accurately assess the potential for plant
uptake of phosphorus and the potential for
subsurface transport losses (Brock et al.,
2007).
The subsurface fate of nutrient species at
NMP sites depends on a variety of transport
and mass transfer processes. For example,
dissolved inorganic nitrogen ions such as
ammonium, nitrite, and nitrate can be trans-
ported with flowing water (advection), as well
as by diffusive and dispersive solute fluxes.
Interface mass transfer processes may play
important roles between dissolved nitrogen
species and soil/sediment surfaces (e.g., sorp-
tion/desorption of ammonium) or the air phase
(e.g., volatilization of ammonia), and uptake
of nitrogen species by plants and microorgan-
isms is anticipated. The presence of colloidal
forms (manure suspension) containing organic
nitrogen and phosphorus that are mobile
in the water phase provide another nutrient
transport pathway.
Environmental losses of nutrient species
that are primarily associated with the solid
-------
phase (such as ammonium and phosphorus)
will occur mainly in conjunction with high
sediment loads during surface water runoff
(Sims et al., 1998; Correll, 1998), For a well
designed NMP that is conducted on most
mineral soils, little leaching and subsurface
transport of phosphorus is expected. Sims et
al. (1998), however, review the literature that
indicates that phosphorus species may also
be leached through soils under certain envi-
ronmental conditions (e.g., deep sandy soils,
high organic matter soils, or soils with high
soil phosphorus concentrations from long-
term over fertilization and/or excessive use of
organic wastes). Pautler and Sims (2000)
suggested that the degree of phosphorus satu-
ration, DPS (defined as the ratio of the amount
of phosphorus sorbed by the soil to the phos-
phorus sorption capacity), is a better indicator
for the potential losses to subsurface transport
than the total phosphorus. Increased addition
of manure has been reported to nonlinearly
increase the phosphorus sorption capacity of
the soil (Brock et al., 2007). When the DPS
was greater than around 30%, phosphorus
transport from soils increased dramatically
(Brock etal., 2007).
Pathogen Transport
The root and vadose zones at NMP applica-
tion sites play a critical role in protecting
water supplies from pathogen contamination.
Effective treatment relies on the retention and
inactivation of pathogens in unsaturated or
variably saturated porous media. Inactivation
of pathogenic microorganisms is commonly
assumed to occur within 60 d, although many
viruses and protozoa are known to be viable
for a longer duration (Schijven et al., 2006).
Considerable research has been devoted to
the fate and transport of microorganisms and
other colloids in porous media (reviews are
given by Herzigetal., 1970, McDowell-Boyer
et al., 1986; McCarthy and Zachara, 1989,
Ryan and Elimelech, 1996; Khilarand Fogler,
1998; Schijven and Hassanizadeh, 2000;
Harvey and Harms, 2002; Jin and Flury,
2002; Ginn et al., 2002; de Jonge et al.,
2004; DeNovio et al., 2004; Rockhold et al.,
2004; Sen and Khilar, 2006; Tufenkji et al.,
2006). Most of the published research per-
tains to saturated media, less is known about
microbe transport and retention in unsatu-
rated systems (Wan and Wilson, 1994b; Choi
and Corapcioglu, 1997; Wan and Tokunaga,
1997; Schafer et al., 1998ab; Saiers et al.,
2003; Saiers and Lenhart, 2003; de Jonge et
al., 2004; DeNovio et al., 2004). Below we
briefly review the literature about processes
and factors that will potentially influence the
transport and fate of pathogens and surrogates
(indicator microorganisms, latex rnicrospheres,
and other colloids) at NMP sites.
Microorganism retention mechanisms in the
vadose zone are more complicated than in
the saturated zone mainly due to the pres-
ence of air in the system. In unsaturated
porous media water flow is restricted by capil-
lary forces to the smaller regions of the pore
space and flow rates are relatively small.
Microorganism transport may be influenced
by increased attachment to the solid-water
interface (Chu et al., 2001; Lance and Gerba,
1984; Torkzaban et al. 2006a), attachment
to the air-water interface (Wan and Wilson,
1994ab; Schafer et al., 1998b; Torkzaban et
al., 2006b), straining of microorganisms near
multiple interfaces in the smallest regions of
the pore space (McDowell-Boyer et al., 1986;
Gushing and Lawler, 1998; Bradford etal.,
2006a), film straining in water films envel-
oping the solid phase (Wan and Tukonaga,
1997; Saiers and Lenhart, 2003), and reten-
tion at the solid-air-water triple point (Crist et
al., 2004 and 2005; Chen and Flury, 2005;
Zevi et al., 2005; Steenhuis et al., 2006).
Transients in water content and composition
during infiltration and drainage processes can
also significantly influence these unsaturated
retention mechanisms (Saiers et al., 2003;
Saiers and Lenhart, 2003; Torkzaban et al.,
2006W.
Size exclusion affects the mobility of micro-
organisms in the vadose zone by constraining
-------
them to more conductive flow domains and
larger pore networks that are physically acces-
sible (Ryan and Elimelech, 1996; Ginn,
2002). As a result, microbes can be trans-
ported faster than a conservative solute tracer
(Cumbie and McKay, 1999; and Harter et al.,
2000; Bradford et al., 2003). Differences
in the dispersive flux for microbes and a con-
servative solute tracer may also occur as a
result of size exclusion (Sinton et al., 2000;
Bradford etal., 2002).
Various combinations of well, surface, and
lagoon water, encompassing a range in solu-
tion chemistry, are used for irrigation at NMP
sites. Many chemical factors (i.e., pH, ionic
strength, surface charge, etc.) are known to
influence the transport behavior of patho-
gens. For example, using colloids as pathogen
models, colloids are stabilized when the elec-
trical double layers are expanded and when
the net particle charge does not equal zero
(Ouyang et al., 1996). Increasing electrolyte
concentration and ionic strength decreases
the double layer thickness and thereby pro-
motes aggregation and microorganism-porous
medium interactions. Aggregation will in
turn impact the colloid size distribution
and therefore straining. Aqueous phase pH
influences the net microorganism and solid
surface charge by changing pH-dependent
surface charge sites. McCarthy and Zachara
(1989) reported that colloids can be produced
through disaggregation when the ion balance
is shifted from one dominated by Ca2+ to one
dominated by Na+. These chemical factors
(pH, ionic strength, surface charge, and chem-
ical composition) are also known to influence
soil structure (disaggregation) and pore size
distribution (shrinking and swelling) when the
soil contains clays and other colloidal materi-
als (Ayers and Westcot, 1989).
Most transport experiments with pathogens
have been conducted in the absence of dis-
solved manure suspensions. Pathogens in
lagoon water, however, constitute only a
small portion of the colloidal suspension.
Large quantities of manure components
are also present. Several researchers have
examined the influence of organic matter or
manure suspensions on microbe transport.
Some researchers reported that dissolved
organic matter (DOM) or manure suspensions
enhanced microbe transport (Johnson and
Logan, 1996; Pieperetal., 1997; Powelson
and Mills, 2001; Guber et al., 2005ab;
Bradford et al., 2006cd). Blocking of favor-
able attachment sites by organic matter has
typically been used to explain this enhanced
transport (Johnson and Logan, 1996; Pieper
et al., 1997; Guber et al., 2005ab), although
filling of straining sites provides an alterna-
tive explanation (Bradford et al., 2006c).
Dissolved organic matter has also been
reported to sorb onto bacterial cell walls and
alter their electrophoretic mobility (Gerritson
and Bradley, 1987). Increasing the negative
charge of the bacterial surface diminishes
its attachment onto negatively charged
solid surfaces (Sharma et al., 1985). Other
researchers have reported that organic matter
inhibits microbe transport due to hydropho-
bic interactions between microbe and grain
surfaces that are coated with organic matter
(Bales et al., 1993). Adsorption of pathogens
onto mobile manure colloids could also facili-
tate their transport potential (Jin et al., 2000;
de Jongeetal., 2004).
Many microbiological factors will also play
important roles in pathogen migration at NMP
sites. Microbe growth, death, and inactiva-
tion are of special importance. The ability of
many pathogens to survive in manure (Wang
et al., 1996), lagoon water (Ibekwe et al.,
2003), and soil (Gagliardi and Karns, 2000
and 2002) has been well documented. The
survival and subsequent transport of viruses
have also been reported to be enhanced in the
presence of manure suspensions (Bradford
et al., 2006d), likely due to sorption of
organic components onto metal oxide surfaces
that would otherwise inactivate the viruses
(Sagripanti etal., 1993; Pieperetal., 1997;
Schijven et al., 1999; Chu et al., 2001; and
-------
Ryan et al., 2002). Some microbes can also
alter their surface characteristics and surround-
ing environment so as to promote attachment
or release (detachment) (Ginn et al., 2002).
Other mobile microbes possess the ability
to move in the direction of increasing con-
centration gradient of chemoattractants or a
decreasing concentration gradient of chemore-
pellents (Nelson and Ginn, 2001). Many of
these microbiological processes are expected
to be strongly coupled with temperature and
nutrient conditions at NMP sites.
Treatments
Figure 1.2 presents a flowchart that illustrates
the potential steps that may be used to assess
and improve the performance of lagoon water
application sites. To determine whether the
reuse of CAFO wastewater on agricultural lands
poses an acceptable environmental risk, it is
necessary to monitor the transport and fate
of relevant contaminants at these sites. At
present, land application of CAFO lagoon water
is based upon a nutrient balance. As men-
tioned earlier, other potential contaminants
of concern at NMP sites include salts, heavy
metals, pathogens, antibiotics, and hormones.
If the risk (measured concentration of a par-
ticular contaminant below or in the root zone)
is deemed to be unacceptable, then manage-
ment practices for animals will have to be
modified and/or wastewater treatment will have
to be implemented before this water can be
used in NMPs. Based on current technolo-
gies and associated costs, it is not practical for
this level of monitoring to be done for typical
farm operations. Additional field research is,
therefore, needed to determine appropriate
contaminant loadings for effective NMPs, and
to assess whether additional treatment steps
are needed. Below we discuss potential tools,
treatments, and management practices that
may be needed.
Animal and Waste Management
Lagoon Water
I
Identification of Key NMP
Contaminants and Processes
Lagoon Water Treatment
I
Design and Implementation of NMP
I
Monitor NMP Contaminants
I
Acceptable Risk
Figure 1.2. A flowchart that illustrates the steps that may be used to assess and improve the performance
of lagoon water application sites.
-------
Computer models are valuable tools that can
be used by researchers to study the transport
and fate of lagoon water contaminants under
various environmentally relevant scenarios
that are found at NMP sites. Conventional
models for contaminant transport are based
upon solution of mass conservation equa-
tions for flowing water (Richards equation)
and solute transport (advection-dispersion
equation), subject to a variety of equilibrium
or kinetic reactions (e.g., Simunek et al.,
2009). Improved NMPs could potentially be
designed with the aid of computer models to
maximize desirable effects such as nutrient
uptake by plants, soil retention of pathogens,
and to minimize the leaching and/or runoff of
contaminants.
To efficiently utilize CAFO lagoon water and
diminish the potential of manure contami-
nants to be transported to groundwater and
surface water bodies it is essential to plan and
implement proper management. Manipulation
of animal diet and veterinary practices are
likely to be some of the most simple and
cost effective ways to minimize the potential
release of some contaminants into the envi-
ronment. For example, it has been reported
that changing cattle diet from grain to forage
will reduce the numbers of pathogenic E. coll
0157:H7 in cattle manure (Russell et al.,
2000; Callaway et al., 2003). Selection of
CAFO locations away from vulnerable water
resources such as shallow water tables and
surface water, and on soils with relatively little
structure and moderate infiltration rates are
other common sense practices to minimize the
potential for water contamination. Sufficient
agricultural land should also be located adja-
cent to CAFOs so that the manure and lagoon
water that is generated may be applied to
these lands at agronomic rates. Actions to
further avoid water resource contamination
may include: proper timing of lagoon water
application (not during rainy seasons); leveling
the CAFO facilities to a small incline towards a
draining conduit and channeling runoff water
to a designated storage reservoir (lagoon);
and avoiding manure accumulation on CAFO
grounds, especially during rainy seasons.
Physical separation is commonly used as an
effective pretreatment method to reduce the
organic load to CAFO wastewater and therefore
decrease nutrient and pathogen concentra-
tions in lagoons. Separation methods to
remove organic solids from the wastewater
include using stationary inclined screens,
vibrating screens, rotary screens, a horizon-
tal decanter centrifuge, hydrocyclone, as
well as roller, belt, and screw presses (Zhang
and Westerman, 1997). The separated
manure sludge may in turn have to be treated
(composted) before it can be used as a soil
amendment and as a source for nutrients.
Lagoons for CAFO wastewater storage act in
a similar fashion as sedimentation basins in
municipal wastewater treatment. Given suf-
ficient retention time in lagoons, significant
reductions in total suspended solids are
expected (35-65%). Most lagoons operate
anaerobically. Aerated lagoons have received
less attention because of their higher costs
and their increased potential of transferring
reactive nitrogen to the atmosphere through
volatilization of ammonia. However, the
potential for decreased odor and increased
nitrogen reduction might increase their use
if economical treatment methods emerge
to mitigate ammonia loss. Typically, CAFO
wastewater storage in lagoons can reduce the
amount of nitrogen by 30 to 75% through
volatilization (depending on whether the
lagoon is anaerobic or aerobic) (Svoboda,
1995; USEPA, 2001a). Phosphorus tends
to be associated with the particulate fraction
of lagoon water, and nitrogen and potas-
sium are usually in dissolved form. Hence,
as particles settle out of solution the lagoon
sludge contains less nitrogen and potassium
but more phosphorus than the lagoon liquid.
Up to 80% of the phosphorus in lagoons can
accumulate in bottom sludge (MWPS, 1993;
Lander et al., 1998). The bottom of CAFO
lagoons should be properly sealed to prevent
-------
deep percolation of wastewater towards
groundwater resources. Most states have
regulations regarding permissible levels of
percolation rates from lagoon storage systems.
For example, regulations in Kansas stipulate
that percolation rates should be less than 0.3-
0.63 cm d-1.
Lagoon water from CAFOs may have to be
adjusted to become a stable, balanced, and
nutrient-rich product for reuse as irrigation
water and fertilizer during implementation of
NMPs. This may require active management
and treatments that are not currently prac-
ticed on farms. Many municipal wastewater
treatment techniques may be applied to CAFO
wastewater (Zhang and Lei, 1998; Chastain et
al., 2001; Vanotti et al., 2005ab; Loughrin et
al., 2006). For example, chemical treatments
may be used to promote flocculation of par-
ticles by neutralizing the electrostatic forces
that keep them apart, causing the particles to
form aggregate flakes (floes) that rapidly settle
out of solution. Commonly used flocculants
include various salts of multivalent cations
such as aluminum, iron, calcium or magne-
sium, and/or long-chain polymers (Dobias,
1993). Other factors such as pH, tempera-
ture, aeration, and salinity can be controlled
to induce flocculation (Lindeburg, 2001).
Given optimum amounts of chemical additives
and sufficient settling time, floes will form
and settle out of solution to clarify wastewa-
ter. Chemical flocculation treatments involve
equipment, chemicals, and maintenance that
are not typically found on CAFOs. Vanotti
and Hunt (1999) and Vanotti et al. (2002)
estimated it would cost an additional $1.27
to $2.79 per finished hog to remove 90-95%
of the suspended solids in swine lagoon
water using chemical treatments. Additional
research is needed to assess whether such
approaches will be economically viable to
implement.
Alternatively, it may also be possible to
adapt a variety of low cost soil treatments
for wastewater to remove excess amounts
of organics, nutrients, pathogens, and other
contaminants in CAFO lagoon water. Potential
soil treatments for wastewater include: infil-
tration ponds and galleries, sand filtration,
and subsurface flow wetlands (Schijven and
Hassanizadeh, 2000; Tufenkji et al., 2002;
Ray et al., 2002; Hunt et al., 2002; Weiss
et al., 2005). Treatment by passage through
sands can provide a physical treatment for
particles in wastewater due to size limita-
tions imposed by the pore sizes (Bradford
et al., 2006a), but also provides solid-water
and air-water interfaces where chemical and
microbiological reactions and transforma-
tions may occur (Tufenkji et al., 2002).
Furthermore, proper selection and design
of the solid surface chemistry of the porous
media can be used to target the retention and/
or degradation of specific contaminants of
interest. For example, metal oxide coatings on
solid surfaces can be created to adsorb and/
or inactivate a variety of microbial pathogens,
as well as inorganic and organic compounds
(Joshi and Chaudhuri, 1996; Ahammed and
Chaudhuri, 1999). Soil treatment forms the
theoretical basis for many permeable reactive
barrier technologies that have been developed
and applied to treat contaminated groundwater
plumes (Gu et al., 1999; Benner et al., 1999;
Scherer et al., 2000; Waybrant et al., 2002).
The collected CAFO wastewater effluent from
soil treatment could in turn be used in NMPs
to irrigate agricultural lands with mitigated
risk to the environment.
Maintaining soil and groundwater quality is
the key to the long-term successful use of
NMPs. If leaching is inadequate, then dis-
solved salts will concentrate and accumulate
in soil profiles when irrigation water that is
applied to crops subsequently returns to the
atmosphere via evapotranspiration. The total
dissolved solids of CAFO wastewater typi-
cally range between 1500 to 3500 rng L1,
equivalent to electrical conductivity levels of
2.4-5.5 dS m"1. Salt, sodium, and osmotic
effects of salinity can restrict crop estab-
lishment and growth due to changes in soil
-------
physical structure, tilth, infiltration and
permeability, sodium toxicity, and salt accu-
mulation (Ayers and Westcot, 1989; Chang et
al,, 1991). Selective use of gypsum, aggregate
stability imparted by crops, direct seeding of
crops, and careful irrigation timing of lagoon
water application under a site-specific man-
agement regime are potential methods that
may be used to ameliorate the negative effects
of CAFO lagoon waters on soil and groundwa-
ter quality (Ayers and Westcot, 1989). Salt
deposits can also be flushed away by adding
excess amounts of fresh water, but at a sig-
nificant cost (Bouwer, 2002). Furthermore,
desalination of wastewater and removing the
excess salt to the ocean can cost up to $2 per
1000 L. Selection of salt-tolerant crops and
timing of nutrient application are, therefore,
important considerations for NMPs. The salt
tolerance characteristics of various crops are
provided in Ayers and Westcot (1989).
-------
2.0
Site Characterization for
Detailed NMP Studies
Background
Field research was designed to study the
transport and fate of nutrients and indicator
microorganisms at a NMP site. To achieve a
high level of confidence in the water flow and
solute transport behavior at this site, detailed
information on the associated soil properties
was required and is discussed in this chapter.
The first phase of our approach was site selec-
tion. Since the transport and fate of nutrients
and microorganisms in the vadose zone are
very complex processes (Bradford et al., 2008;
Bradford and Torkzaban, 2008) measurements
of the apparent soil electrical conductivity
were used to select a highly uniform location
in a 4 ha field to reduce the level of com-
plexity and uncertainty. The second phase
of our approach involved an initial site char-
acterization to identify the soil stratigraphy,
approximate particle size distribution, and
bulk density information on soil cores and pits.
Results revealed that a uniform layer of sandy
loam was situated from the surface down to a
depth of about -70 cm, followed by a white
sand layer (-70 to -80 cm). The deeper pro-
file (less than -80 cm) was characterized by a
sandy loam layer, which incorporated various
sand and clay lenses at different depths. In
the third stage, pedotransfer functions and
interpolation algorithms were used to gener-
ate an initial conceptual model about the soil
profile, including initial estimates of hydraulic
properties (soft data). This conceptual model
guided the selection location for undisturbed
cores that capture the dominant water flow
behavior under study.
The hydraulic properties of and conservative
solute transport in undisturbed core samples
were subsequently measured in the labora-
tory (hard data) and used to further refine the
conceptual model of the site properties. The
next phase of site characterization was field
validation of the conceptual model of the soil
hydraulic and solute transport properties.
Water flow and solute transport experiments
were used for this purpose, and to ensure good
agreement between measured and simulated
data. A feedback between field experiments
and computer simulations was used to iden-
tify locations in the field where additional
measurements or computer simulations were
needed to improve the conceptual model and
level of understanding at the site.
and
A mobile remote electromagnetic induction
sensor system (Geonics EM-38DD) was used
to measure the apparent soil electrical con-
ductivity (ECa) across a 4 ha field (Corwin
and Lesch, 2005a). This system includes two
EM-38 units synchronized to operate simul-
taneously, enabling measurements of both
vertical and horizontal dipole conductivity.
Data was collected following the protocols
of Corwin and Lesch (2005b) to generate an
approximately 5x5 m grid ECa map of our field
site. The map coordinates were determined
using a global positioning system (GPS).
Selection of our experimental field plot for
nitrate and microorganism transport stud-
ies was based upon the generated ECa map.
The agricultural field in San Jacinto, CA
(33°50'22" North, 117°00'46" West) was
chosen for this purpose because it was asso-
ciated with minimal spatial variability in ECa
readings, which implies relatively low het-
erogeneity in soil texture. The experimental
plot is located next to a dry river bed and has
a shallow perched water table at a depth of
-220 cm.
-------
The experimental site consisted of two 6x6 m
plots (Figure 2.1). At the corners of each
plot a backhoe was used to expose the soil
profile and to install 120-cm diameter by
220-cm long culvert pipes vertically into the
soil (Figure 2.1 - circles marked with letters).
Soil profiles in each culvert pipe hole were
photographed and notes on soil stratification
were taken before installing each culvert pipe.
Each culvert pipe was instrumented with 6
tensiometers and 6 solution samplers to mea-
sure the soil water pressure and to extract soil
solution with depth. Tensiometers and solu-
tion samplers (90 cm in length) were installed
radially from the culvert pipe into the undis-
turbed soil profile (parallel to the soil surface).
A staggered configuration of the sensors in
the culvert pipe was selected to minimize the
potential for preferential flow and interference
from other sensors, and to maximize the area
of the profile that was sampled (Figure 2.1).
The arc in Figure 2.1 represents the area of
water potential and soil solution sampling.
Pressure transducers (MPX2100DP, Motorola
LTD., Denver, Colorado) and a data logger
(CR7, Campbell Scientific, Inc., Logan, Utah)
recorded the tensiometer readings every
15 min. Five neutron access tubes (220 cm
long) were installed vertically on each plot
(Figure 2.1 - circle marked with Roman
numerals). The water content with depth at
desired times was determined using a neu-
tron probe (503-DRHYDROPROBE®, CPN,
Martinez, CA) and an established calibration
curve in this soil profile.
To minimize the sampling time of the solu-
tion samplers, it was necessary to increase
the water flux through the soil solution sam-
pler. This was accomplished by replacing the
traditional ceramic cup with filter paper (MF-
0.45 urn, Millipore, Billerica, MA) that had a
high saturated conductivity (60 cm-lr1) and a
small thickness (180 urn).
The instantaneous profile method (Watson,
1966) was used to evaluate the hydraulic con-
ductivity of the upper 60 cm of the profile at
the field site. This method uses simultaneous
measurements of the volumetric water content
(9) and soil water pressure head (h) in a soil
profile during the course of drainage to deter-
mine the hydraulic conductivity.
6 m
(0,0)
Plotl
Plot 2
'b
B
'o b b
vo
15
Figure 2.1. A schematic of the field site. Squares represent two 6x6 m plots. Circles with letters are verti-
cal culvert pipes (120 cm diameter by 220 cm long) installed with six tensiometers. Circles with roman
numerals are vertical neutron probe access tubes. Arcs represent the area of water potential sampling. The
crossed circle in the center of Plot 2 denotes the origin of a radial axis (r).
-------
An additional set of 9 and h data was mea-
sured near saturation (h= 0 to -30 cm) during
redistribution of water immediately after
ponded infiltration ceased (Hillel, 1998). In
this case the value of 8 was acquired using a
time domain reflectometry (TDR) system (Trase
system, Soilmoisture Equip. Corp., Santa
Barbara, CA) as well as gravimetrically, and h
was monitored with a tensiometer (Tensimeter,
Soil Measurement Systems, Tucson, AZ).
A step pulse conservative tracer (Potassium
Bromide, KBr) experiment was conducted
under steady-state water flow on bare soil at
the plot, after plowing the upper 20 cm of the
soil. We employed a 0.2 x 0.2 m staggered
grid drip irrigation system (Typhoon 630,
Netafim, Fresno, CA) to uniformly apply small
water and solute fluxes to the soil surface that
induced unsaturated flow conditions in the
upper 120 cm of the soil profile. The surface
boundary condition was a steady infiltra-
tion rate of 0.275+0.025 cm-h4. The soil
surface was covered with a black nylon tarp
to avoid water evaporation during the tracer
experiment.
When a steady-state water flow condition was
established in the soil profile after 100 h,
the water application system was switched
between well water to a step pulse of KBr
solution for 47 h and then switched back to
well water for an additional 280 h. The KBr
solution had 2.4 g-L4 Br, and was mixed well
before and during application. Soil solution
samples were collected over an interval of
24 h and analyzed for Br concentration with
a Colorimetric system (O.I. Analytical, Flow
Solution IV, College Station, Texas). The Br
data from the field is presented at the median
time of the sampling interval.
Additional information was collected to evalu-
ate the transverse dispersivity in the root zone
(0 to -60 cm) and the effect of non-uniform
upper boundary condition on the soil water
and bromide application uniformity. These
experiments are described in detail in Segal et
al. (2009).
Undisturbed cores from the experimental site
were collected for measurements of bulk den-
sity (pb), hydraulic properties, and conservative
solute transport. Details on these experiments
and results are provided in Segal et al. (2008
and 2009).
The HYDRUS-2D code (Simunek et al., 1999)
was used to simulate the water flow in the
upper profile during steady-state infiltration
and drainage experiments. Other simulations
were conducted to determine the lateral water
flux that may occur directly above or through
the sand layer relative to the vertical surface
water flux. The two-dimensional water flow
and bromide transport at the field site during
the tracer experiment under a non-uniform top
boundary condition (drip) and when consider-
ing a heterogeneous distribution of hydraulic
properties in the lower profile (depths <-80
cm) was also simulated using this model.
Details on the simulation conditions and
parameters are in Segal et al. (2008, 2009).
and Discussion
Soft
In this section we discuss the use of soft data
to select a location for flow and transport stud-
ies and to develop a conceptual model of the
soil hydraulic properties at this field site.
Electromagnetic induction technology was
employed to quantify the field spatial vari-
ability of ECa. Figure 2.2 shows a map of the
vertical component of ECa. The experimental
study site was selected from this soft ECa data
to be located in a "relatively uniform" section
of the field with low ECa values (coarser tex-
tured material) as indicated in Figure 2.2.
Our initial conceptual model of the soil pro-
file stratigraphy at the experimental site was
developed from photographic information
(Segal et al., 2008). The particle size distri-
bution (PSD) and bulk density data (Segal et
al., 2008) were used in conjunction with the
ROSETTA program to estimate values of satu-
rated hydraulic conductivity (Ks) at sampled
-------
Vertical EC3:
Yellow < 0.49 dS m
Orange 0.49 - 0.64
Tan 0.64 - 0.80
Brown > 0.80 dS m'1
Riverbed
Figure 2.2. A 4 ha map of the vertical component of the apparent soil electrical conductivity in the field as
measured using a mobile remote electromagnetic induction sensor system on a 5x5 m grid.
locations. The nearest neighbor algorithm was
then used to interpolate values of Ks at other
locations that were not sampled. Figure 2.3
presents the estimated values of Ks on north-
south and east-west transects of the plot. The
dark uniform layer between the soil surface
and -70 cm was the cultivated layer. The light-
colored thin layer underneath the cultivated
layer was the sandy layer (-70 to -80 cm).
The lower soil profile (-80 cm to -210 cm)
was scattered with multiple lenses of different
texture. The data can be further analyzed in
terms of one-dimensional vertical conductivity
(Segal etal., 2008).
Hard Data
The soft data discussed above provided valu-
able information that was used to guide our
strategy for collecting additional hard data
in the field and the laboratory. This hard
data was used to further refine our concep-
tual model and to improve our quantification
of key soil hydraulic properties at the field
site. Furthermore, comparison of field and
laboratory data was used to identify potential
limitations of the various types of hard data,
and to assess the predicted hydraulic proper-
ties that were estimated from the soft data.
Undisturbed cores were collected for hydrau-
lic property analysis at several locations in
the root zone (0 to -70 cm), at the capillary
barrier (sand layer), and in the lower profile
(below -80 cm). These sampling locations
were chosen to improve our understanding
of water flow in areas of the site where dis-
tinct differences in estimated soil hydraulic
properties were identified (soft data shown
in Figure 2.3). The fitted hydraulic model
parameters to the measured soil core data are
provided in Table 2.1.
Field data collected during steady-state infil-
tration and redistribution were analyzed using
the instantaneous profile method to determine
hydraulic properties at depths of -30 and
-60 cm. Analysis of the entire profile with this
method was not possible due to the layering at
the site, which violated an assumption of this
approach (no lateral flow).
-------
East - West
North - South
100 200 300 400
100
200 300
400
Longitudinal distance (cm)
Figure 2.3. North-south and east-west transects of estimated values of Ks on plot 1, These transects were
generated from a three dimensional map of Ks that was estimated using particle size distribution and bulk
density data from locations I-V using ROSETTA, and the nearest neighbor interpolation algorithm. Circles
represent the undisturbed core sampling locations.
Fitted field-scale hydraulic properties param-
eters also are presented in Table 2.1. Field
and laboratory values of h(Se) matched rea-
sonably well. In contrast, field and laboratory
values of h showed considerable dissimilarity
and revealed limitations of using only lab
information to described field-scale hydraulic
properties. A potential explanation for this
difference could be entrapped air in the field.
Table 2.1. The hydraulic properties (6s is the saturated water content; 6r is the residual water content; a is
the reciprocal of the air entry pressure; and n is the pore size distribution parameter of the van Genuchten
model) of the three major layers in the upper soil profile of the field plot. Soft data was estimated from par-
ticle size distribution and bulk density data using the ROSETTA program. Hard data was measured on undis-
turbed core samples using the multistep outflow, pressure plate, and constant head permeameter techniques
or from TDR readings in the field site.
Hydraulic
property
Ks (cm.rr1)
6s
6r
a (crrr1)
n
Layer 1
Sandy loam
0 to -70cm
Soft data
1.95
0.37
0.033
0.018
1.45
Hard data
5
0.43
0.03
0.0085
1.6
Layer 2
Sand
-70 to -80cm
Soft data
9.25
0.35
0.04
0.042
2.41
Hard data
71.2
0.33
0.01
0.016
1.96
Layer 3
Silt loam
-80 to -90cm
Soft data
1.4
0.41
0.056
0.0047
1.70
Hard data
1.6
0.45
0.04
0.007
1.57
-------
Validation Simulations
In this section we compare soft and hard
estimates of hydraulic properties from several
locations in the field. Numerical simulations
were subsequently conducted to indepen-
dently "validate" our field scale hydraulic
properties where possible, and to further
improve our understanding of water flow
at specific locations in the field. Moreover,
the reliability of using soft and hard data
as input for simulations of water flow was
demonstrated.
Table 2.1 presents a comparison between the
hydraulic properties generated from soft and
hard data. This information was obtained for
three layers in the upper soil profile of the
field site (0 to -70 cm; -70 to -80 cm; and
-80 to -90 cm). In general, soft and hard
estimates of the hydraulic properties for the
silt loam layer were very consistent. In con-
trast, soft estimates of the hydraulic properties
were poorer for the coarser textured sandy
loam layer and especially the sand layer. In
general, the soft data provides a reasonable
prediction of soil textural characteristics, but
that the specific magnitudes of the hydraulic
properties are only approximately predicted.
Hence, soft data may likely be used to infer
trends in spatial variability in soil textural
properties, but not to deterministically simu-
late water flow at the site. Observed and
simulated steady-state water infiltration and
redistribution in the relatively uniform upper
soil profile (0 to -70 cm) support this state-
ment (Segal et al., 2008). In this case,
simulations which employed measured
hydraulic properties as model input provided
a better agreement of observed water flow and
drainage. Nevertheless, simulations which
employed the soft-data-derived hydraulic
properties did capture the general water flow
behavior and trends. Numerical modeling in
conjunction with measured hydraulic proper-
ties also proved to be an effective tool to gain
additional insight on field scale water flow.
For example, we used numerical simulations
to quantify the influence of water application
rate on the lateral water flux at a location of
pronounced soil textural discontinuity in the
field at a depth of -70 to -80 cm (Segal et al.,
2008).
Bromide Travel Times
The mean and standard deviation of the bro-
mide travel time for each BTC is presented in
Table 2.2, as well as average and variance of
these values at each depth. Minor changes
in the variance of the mean travel time were
measured in the upper profile (-32 and
-61 cm), due to high uniformity in soil texture
and water content (Segal et al. 2008). In
contrast, the variance significantly increased
immediately below the sand layer (-95 cm).
This high variability in travel time can be
attributed to three reasons, namely: (i) the
sand layer was associated with a lower water
content and consequently a higher pore water
velocity; (ii) low permeability lenses just below
the sand layer caused lateral flow of Br; and
(iii) variability in the soil permeability directly
below the sand layer generated variability in
water and bromide mass inputs.
The average standard deviation in travel time
(Table 2.2) tended to increase with depth as
expected due to dispersion. However, similar
values occurred at -95 and -126 cm, likely
due to greater variability in travel times at
-95 cm. The ratio between the variance and
the average values was much larger in the
mean travel time than in the standard devia-
tions of the travel times. This observation
suggests that local scale effects on the BTCs
were much less significant than plot scale
factors.
Averaging of Bromide Transport
The influence of soil heterogeneity, water
content, and location of the solution samplers
on the tracer data may be overcome by aver-
aging the 4 collected samples at the same
depth and time on a plot, and then analyzing
the depth averaged bromide transport behav-
ior. Bromide transport in the root zone and in
the lower profile will be separately discussed
in detail below due to the previously noted
-------
Table 2.2. Mean and standard deviation of bromide travel time (units in hours, h) at each sampling depth
and culvert pipe location, and associated average and variance values for each depth.
Depth
Location
A
B
C
D
Average mean travel time (h)
Variance of mean travel time (h2)
-32 cm
-61 cm
-95 cm
-1 26 cm
-155 cm
Estimated mean bromide travel time (h)
82
64
75
NA
73.6
82.3
120
110
130
110
117.5
91.6
280
160
200
145
196.2
3656.2
190
265
210
250
228.7
1206.2
250
250
300
270
267.5
558.3
Depth
Location
A
B
C
D
Average standard deviation of travel time (h)
Variance of standard deviation of travel time (h2)
-32 cm
-61 cm
-95cm
-126cm
-155 cm
standard deviation of
bromide travel time (h)
52
37
38
NA
42
70
50
46
56
46
50
22
69
61
66
62
65
13
60
62
68
66
64
13
80
63
79
76
75
62
differences in soil and water flow characteris-
tics in these regions.
Observed field measurements of the depth
averaged bromide transport data is presented
in Figure 2.4. The average relative bromide
concentration at a given time is denoted
using filled circles for the -32 cm depth and
with empty circles for the -61 cm depth.
Integration of the area under the breakthrough
curve provided very good mass balance at -32
and -61 cm (99.5-102%) which is presented
in the figure. The vertical bars represent
standard deviations and the horizontal bars
represent the sampling time range. Standard
deviations account for the inherent soil spa-
tial variability, and the variable location of
each solution sampler. The modeled break-
through curves, based on the mobile-immobile
(MIM) transport parameters evaluated in the
laboratory, are also depicted in Figure 2.4.
Additional information is provided in Segal
et al. (2009). The modeled curves provide a
reasonable prediction of the average measured
value of bromide at -32 and -61 cm and total
Br recovery. In general, the attempt to inde-
pendently predict the bromide transport in
the root zone by using only laboratory derived
model parameters was fairly successful at
this scale. In addition, both one- and two-
dimensional simulations (average value of a
cross section at a specific depth and time)
provided nearly an identical description of the
data, indicating that the water flow and solute
transport in the upper 60 crn of the soil profile
could be described as a uniform one-dimen-
sional process when using our drip irrigation
setup and measured values of dispersivity.
The water flow and solute transport pattern
in the lower profile (-70 to -200 cm) is much
more complex than in the root zone (0 to
-70 cm). In particular, the coarse sand layer
at about -70 cm generates high water veloci-
ties and a lateral component to water flow
(Segal et al. 2008). The presence of inter-
connected layers and fine textured lenses of
variable thickness below -80 cm produce a
-------
Br recovery
-32cm - 99.5%; -61cm -102%
q= 0.275 (cm.rf1)
1/2D MIM model -32cm
1/2DMIM model-61cm
Average -32cm
Average -61cm
100
200
Time (hours)
300
Figure 2.4. Field measurements of the relative bromide concentration over time for the bromide tracer
experiment. Data is presented at two depths in the field plot, -32 and -61 cm (circles). Each data point is
an average of the four locations at the same time and depth. The vertical bars are standard deviations and
the horizontal bars represent the sampling time range. The corresponding ID and 2D (depth averaged) MIM
simulation results are also presented in this figure based on transport parameters derived from the undis-
turbed core.
location specific pattern of solute transport
through the lower soil profile of the field plot.
Two-dimensional solute transport simulations
were conducted to study the relative impor-
tance of nonuniform boundary conditions and
soil heterogeneity on solute transport in the
lower profile. Results indicate that soil het-
erogeneity is the major contributor to solute
transport behavior in the lower soil profile.
Specifically, variations in bromide travel time
and total mass suggested that the sand layer
in conjunction with fine textured lenses slow
down the bromide propagation and generates
bypassing of water and solute through lower
resistance pathways. Yet, the lower water
velocities below this region (due to the raised
water table) decrease the variability of the
mean travel time (Table 2.2) and generate a
more uniform Br front. This finding coupled
with the fact that the exact soil type that is
associated with a particular soil solution sam-
pler is unknown, hampers the application of a
completely deterministic approach for describ-
ing solute transport in the lower profile.
Considering the above information, the
bromide transport data in the lower profile
is presented below in terms of the one-
dimensional average concentration (4 sample
locations) at a specific depth as a function
of time. Figure 2.5 presents measured and
simulated bromide breakthrough curves in the
lower profile at -95, -126, -155 cm and the Br
mass recovery under the BTCs. The horizontal
bars are standard deviations of the concentra-
tion measurements. It should be mentioned
that the one-dimensional modeling data that
are presented in this figure were generated
from a two-dimensional simulation that was
averaged over a specific depth at a given time.
A transect of spatially heterogeneous hydraulic
properties was generated from measurements
taken from the field plot that was used as
input for this simulation (Figure 2.3). Good
-------
agreement between measured and modeled
data was found (Correlation coefficient above
84%). In general, travel time and mass bal-
ance of the average bromide concentration
over depth is conserved (>88.4% bromide
mass recovery). However, the observed and
simulated standard deviations have the same
order of magnitude as the average value, sug-
gesting high variability of bromide at a specific
depth and time.
0.6
0.4
0.2
0.0
o
O
6
0.0
0.6
0)
;o
I 0.2
0.0
0.6
0.4 -
0.2
A / -95 cm
Br recovery- 89.6%
Correlation coefficient- 84%
Field data
2D model
B /-126cm
Br recovery- 88.4%
Correlation coefficient- 89.2%
C/-155 cm
Br recovery- 90.4%
correlation coefficient - 89.1%
100
200
Time (hour)
300
Figure 2.5. Field measurements (circle) and
2D-MIM simulation (lines and triangle) of rela-
tive bromide concentration over time at -95, -126
and -155 cm. Measured data is an average value
from four sampling locations at the same time and
depth. Simulated data is presented as an average
value at these same depths. The horizontal bars are
standard deviations.
-------
3.0
NMP Results - Nutrients
Background
NMPs are designed to meet the water and
nutrient needs of crops, while minimizing the
loss of nutrients to surface water and ground-
water (USEPA, 2003), However, researchers
with the USEPA have observed significant
migration of pollutants (e.g., nitrate) towards
surface water and groundwater bodies at NMP
sites (Hutchins and Bradford, 2005 - per-
sonal communication). These observations
suggest that implementing a NMP based on
current agronomic practices may not always
protect the environment. The literature review
(Chapter 1) indicates that potential problems
with NMP implementation include: (i) inac-
curate quantification of water and nutrient
mass balances due to inadequate information
on soil properties, climatic data, wastewa-
ter constituents or crop water and nutrient
uptake rates; (ii) inherent spatial and tempo-
ral variability in soil properties, manure and
nutrient distribution, irrigation uniformity,
and crop growth; and (iii) NMP management
constraints, such as quantifying the amount
of water and wastewater application and
managing the timing of the application. The
objective of this study was to measure the
fate of nitrogen, phosphorus, potassium and
salts from land application of dairy wastewater
under a well-designed and implemented NMP
in a semi-arid environment. We also pres-
ent key management practices that minimize
the potential leaching of nutrients and salts
toward groundwater.
and
Traditional NMP studies have been conducted
from an agronomic perspective to determine
the impact of specific NMP factors on a given
field. The statistical design for such experi-
ments include blocks and random repetitions
because the spatial variability of the field and
application systems is typically unknown. In
contrast, our experiment was a process-based
(mechanistic) study of flow and transport
processes under NMP conditions. Accurate
implementation of a process-based NMP
requires the determination of water, salt and
nutrient mass balances in the root zone, and
the ability to quantify flow and transport pro-
cesses. To collect this type of information
requires the use of measurement tools such as
weighing lysimeters, weather station, tensiom-
eters and solution samplers with depth, and
irrigation systems with a high level of unifor-
mity and precision. The traditional agronomic
design cannot achieve such high precision in
measuring and implementing a NMP on each
repetition due to economic constraints, and
this further increases the variability of tradi-
tional NMP studies. Below we highlight the
design of our process-based NMP that was
needed to overcome the limitations of the
traditional agronomic approach.
Field
Our field site was located in San Jacinto, CA
(33°50'22" North, 117°00'46" West) and was
chosen to be in a relatively homogenous part
of the field in order to minimize variability in
soil hydraulic properties. This objective was
achieved by acquiring preliminary informa-
tion on the field soil spatial variability using
a remote electromagnetic induction system
(Segal et al., 2008). This procedure elimi-
nated the need of repetitions scattered across
the field in traditional agronomic NMP stud-
ies, and allowed us to focus on within plot
variability. The experiment site was also
chosen to have no recent history of CAFO
wastewater application. The field was culti-
vated only during winter (triticale, TRICAL®
Resource Seeds, Inc.) and manure was
applied twice (the last manure application was
three years prior to this experiment) during the
last ten years prior to this experiment.
-------
The experimental site consists of two
6 m x 6 m plots (Figure 3.1, more detailed
description of the site is given in Segal et al.,
2008 and 2009). Within plot variability was
overcome by taking multiple measurements
over depth at several locations within the
plots. Briefly, one culvert pipe was installed at
each corner of both plots (8 pipes total). Each
pipe was instrumented with 6 tensiometers
and 6 soil solution samplers installed over
depth at approximately 30 cm intervals, which
were installed 90 cm horizontally from the cul-
vert pipe into the undisturbed soil profile. The
staggered configuration of the sensors (repre-
sented by the arc in Figure 3.1), was selected
to maximize the area of the profile that was
sampled. Each plot was also equipped with
five neutron probe (503-DRHYDROPROBE®,
CRN, Martinez, CA) access tubes for measur-
ing the water content with depth.
An intensive study on water flow and soil
hydraulic properties was conducted on the
two plots at our experimental site (Segal et
al., 2008) and no significant difference was
found between the plots. The soil texture of
the root zone (0-65 cm) was a sandy loam
(Grangeville fine sandy loam, a coarse-loamy,
mixed, superactive, thermic fluvaquentic
haploxeroll) with average contents of 55%
sand, 40% silt and 5% clay. The average bulk
density was 1.35 g-crrr3 and the saturated and
residual water contents were 0.43 and 0.03,
respectively. The average saturated hydraulic
conductivity was 2.35 cm-hr1, the longitudinal
dispersivity was 0.56 cm, the immobile water
content was 0.0978, and the mass transfer
coefficient between mobile and immobile
regions was 0.0035 tr1 (Segal et al., 2008
and 2009).
6 m
Flow Meter Mixing
Treated DWW jnM tank '
Well water
Blending plot
B
Pump Flow Meter
outlet
Plot
Weighing
Lysimeters
Weather station
Cyclic plot
H
18
Figure 3.1. Schematic of the field site. Squares represent two six-by-six meter plots. Circles with letters
(A-H) are 220 cm in length vertical culvert pipes installed with tensiometers and solution samplers. Arcs
represent the area of water potential sampling. The controlled mixing and application system is illustrated at
the top of the figure.
-------
Raw DWW was treated prior to land applica-
tion with a stationary inclined screen separator
(Zhang and Westerman, 1997), a sedimenta-
tion tank (Sukias et al,, 2001), and a sand
filter (Rodgers et al., 2005) packed with
crushed silica (0.45-0.5 mm in diameter,
AGF, Netafim, Fresno, CA). In contrast to tra-
ditional agronomic designs, our process-based
NMP minimized variability through uniform
application of water and nutrient with a high
level of precision. Specifically, well water and/
or DWW were uniformly applied to each plot
using a pump and nine emitters (R184CT,
Raindrip, Fresno, CA) in 3 m spacing. The
water application rates varied between 0.6 to
0.95 cm-h ] (25 to 40% of the soil saturated
hydraulic conductivity) with a Christiansen
uniformity coefficient of 94% under low wind
conditions. The water and wastewater appli-
cation system consisted of a mixing tank, a
controller, solenoid valves, pumps and two
electrical water meters with a resolution of
3.78 L + 1.5% (JSJ075, Carlon meter, Grand
Haven, Ml) to achieve desired blending ratios
for treated DWW and well water (Figure 3.1).
Dairy wastewater was applied using cyclic
and blending strategies. The blending strat-
egy employed a selected mixture of treated
DWW and well water to meet the needs of the
crop for N and water. Conversely, the cyclic
strategy applied separate irrigations of treated
DWW and well water to meet the crop needs
for N and water.
A NMP was implemented on winter and
summer crops during 2007, and a perennial
crop in 2008. Triticale (TRICAL® Resource
Seeds, Inc.) served as the 2007 winter crop
and was planted on February 10th, seedlings
emerged and established a full stand on
February 25th, and the crop was harvested
on May 8th. The 2007 summer crop was
NK 300 hybrid forage sorghum (Syngenta
Global) and was planted on May 30th, seed-
lings emerged and established a full stand
on June 5lh, and the crop was harvested on
August 13th. The 2008 perennial crop was
alfalfa (Grandslam cv. Western Farm Service,
CA). Alfalfa is an important crop for the dairy
industry due to its high yield, feeding value
and efficiency in N removal. Alfalfa has three
major differences relative to the crops used
in 2007, namely: i) it is a perennial crop with
multi-cuts; ii) it has a deeper root system;
and iii) it may assimilate N from the atmo-
sphere through nodules (symbiotic nitrogen
fixation). Alfalfa was planted on the blending
field plot at agronomical rate of 2.75 Kg-ha"1
on April 1st, emerged on April 10th, and
established full cover 30 days later. Five
consecutive growing cycles (five cuttings with
harvest interval between 37 to 45 days) were
achieved during 2008. The NMP was imple-
mented by dividing the growing season into
multi-cut segments that represent the growing
cycle. Each growing cycle was considered a
separate period with a new initial condition.
The first growing cycle started on May 10th
and lasted until June 17th. However, exten-
sive weed growth interfered with the normal
development of the alfalfa during this period.
Therefore, the plot was treated with herbicide
(Pursuit, BASF, NO at the rate of 150 ml-ha-1
on June 25lh. A fallow season (time between
harvest and planting that is associated with
minimal ET and nutrient uptake) occurred
between successive growing seasons for the
crops described above.
Blumenthal and Russelle (1996) reported that
atmospheric N fixation through nodules will be
less active during periods of high A/' .,. The
or e> soli
alfalfa NMP therefore attempted to maintain
high N concentrations in the root zone during
the growing season in order to minimize the
amount of N fixation and to maximize the
amount of DWW addition. No DWW was
applied between the last two alfalfa harvests
in order to deplete the soil profile of plant
available N.
Water Nitrogen in the Root
Zone
Plot scale water balance information in the
root zone over a given time interval was used
to determine the amount of applied irrigation
-------
water, / (ML3AT4) to meet crop ET (ML3A?1)
at the end of this interval as:
where D (ML^T1) is water loss due to drain-
age, Pw (ML"3AT"1) is the water input due to
precipitation, and AW (ML"3AT"1) is the change
in soil water storage (final - initial).
Water balance parameters in Equation [3.1]
were measured as described below. Potential
ET (£Tp), with a resolution of 0,1 mm, was
estimated using data from a weather station
(Penman, 1948) located between the plots.
Temperature, relative humidity, solar radia-
tion, wind speed and rain were recorded every
15 min. Actual ET (£Tactual) was estimated
from weighing lysimeters, 20 cm in diameter
and 100 cm length, installed at the perimeter
of each plot (Figure 3.1). The top of the lysim-
eter was at the soil surface connected to a
load cell that measured the total weight con-
tinuously (resolution of 50 g). A suction cup
(filter paper; MF-0.45 urn, Millipore, Billerica,
MA) was connected to the bottom of the lysim-
eter, where vacuum was applied to collect the
drainage. The crop coefficient (/(,) was cal-
culated as the ratio of £Tactual to ETp during a
given time period. The value of Pw (resolution
of 0.25 mm) was measured using a rain gauge
(CS700, Campbell Scientific, Logan, Utah).
The values of D and AM/were determined from
neutron probe and tensiometer readings in the
root zone and measured soil hydraulic prop-
erties. The value of / was verified from flow
meter readings.
In this study the following inorganic and
organic N mass balance equations for the root
zone were employed:
// i A // i A // i A A //
*"application + ^Ol =
plant '"drainage ~"~ '"atmosphere ~"~ ^'"soil
[3.2]
application
[3.3]
where A/',ro/_iopis the inorganic N applied to
the soil surface (ML3AT1), EOI (ML3AT1) is
the amount of N converted from/to organic
to/from inorganic forms, N'rlant (ML3AT"1) is
the inorganic N uptake by the plant, N1, .
° r J r i drainage
(ML3AT"1) is the inorganic N drained below
the root zone, /V^mosptere (ML-3AT1) is the
inorganic N lost to the atmosphere, A/V(.o/; is
the difference in inorganic N storage in the
root zone (final-initial), N%plicatjQn (ML3 A?1) is
the organic N applied to the soil surface and
A/V^., (ML"3AT1) is the difference in organic
N storage in the root zone (final-initial). The
total N mass balance is equal to the sum of
Equations [3.2] and [3.3]. Equation [3.3]
assumes that losses of organic N are only due
to mineralization (volatilization and drainage
of organic N are assumed to be negligible).
The N mass balance was calculated over the
upper 30 cm for the triticale, 60 cm for the
sorghum and 90 cm for the alfalfa, where
roots are most active in water and nutrient
uptake under irrigated conditions (Katterer
etal., 1993; Merrill and Rawlins, 1979;
Abdul-Jabbaretal., 1982). Nitrogen bal-
ance parameters were quantified as described
below. Total N and C in the solid phase of
the DWW, soil, and plant tissues were mea-
sured using the combustion method (Flash
EA 1112, Thermo-Finnigan, Waltham, MA).
Measurement of nitrogen from ammonium
(N-NH4) (Keeney and Nelson, 1982) and
nitrogen from combined nitrite and nitrate
(N-(N02+NO._!)) concentrations in soil solution
and DWW were performed using a colorimet-
ric system (O.I. Analytical, Flow Solution IV,
College Station, Texas) after filtering the
sample through a 0.22 urn filter. Values of
N1 , .. and N° , ,. were directly mea-
application application J
sured in the DWW before each application.
The value of AN1 ., was determined from
soil
sequential measurements of soil inorganic N
concentrations in the root zone before DWW
application events. The value of /V|_,,ra,n90e was
determined from measured inorganic N con-
centrations in soil solution below the root zone
and from information about the water fluxes.
N1. , accounts for volatilization of ammo-
aimosphere
nia (NHJ during application and from the
-------
soil surface. We assumed denitrification was
negligible under the low nitrate and unsatu-
rated conditions of the root zone (Luo et al.,
1999). The loss of NH3 during irrigation was
measured using the concentration ratio of
N-NH4 in the irrigation water at the emitter
outlet and at the soil surface. Volatilization
of NH3 from the soil surface was measured
following DWW wastewater application for a
period of one week during the 2007 winter
crop season using a standard chamber and
acid-trap technique to capture NH.^ emissions
(Black et al. 1985). The loss of N-NH4 during
irrigation and the ET are presented for two
application events in Table 3.1. The N-NH4
losses during irrigation varied between 9 and
32%, and higher rates were associated with
higher N-NH4 concentrations in the irrigation
water and higher ET. The atmospheric loss
of N-NH4 durin
irrigation was measured and
taken into account in the N balance before
each application. The loss of N-NH4 from
the soil after irrigation was measured to be
three orders of magnitude smaller than the
N-NH4 loss to the atmosphere during irriga-
tion (Table 3.1). These findings are consistent
with other data presented in the literature
(Cameron et al., 1995; Sharpe and Harper,
1997; Cameron et al., 2002; Hawke and
Summers, 2006).
A/;, , was determined from measurements of
plant
dry phytomass and its N content. Before each
water application, aim long row of plants of
the triticale and sorghum (0.2 to 0.4 m2) or
0.3 by 0.3 m of the alfalfa was collected for
N analysis from the middle of the plot, where
minimal effects from the measuring devices
are expected. Since the root system was not
removed during the harvesting, it was not con-
sidered as an N sink.
For triticale and sorghum during 2007 the
value of fn, accounts for the net exchange
(_//
due to mineralization. In contrast, for alfalfa
(a legume) during 2008 the value of EOI
also accounts for N fixation from the atmo-
sphere. The value of EOI was determined from
Equation [3.2], while all other parameters
were measured, and this information was used
in conjunction with Equation [3.3] to deter-
mine the changes in N° ., using measured
& so/I &
average values of the initial N° , at the field
o soil
site. The exchange rate was subsequently
calculated as EOI divided by the initial /V°i/y for
a given time period. Final total N (dominated
by AP,,,) and its spatial variability in the plots
were measured on March, 2008. Ten soil
cores, 30 cm long by 1.25 cm in diameter,
were sampled from each plot in random loca-
tions. Three samples were taken from each
Table 3.1. Ammonia volatilization from the sprinkler irrigation system and soil surface.
N-NH3 volatilization from sprinkler irrigation
Late
winfpr
DAE
35
35
57
57
Late
wintpr
DAE
29-36
Application
Strategy
Blending
Cyclic
Blending
Cyclic
Application
Strategy
Cyclic
N-NH4 in Irrigation water
mg-L1
11.1
39.45
92.57
157
N-NH3 volatilization from
N-NH4 in Irrigation water
mg-nr2
2407.82
Volatilization Loss
%
15
9
32
22
soil surface
N-NH3 volatilization
mg-rrr2
1.614
ETP
mm-h"1
0.446
0.142
0.68
0.24
Volatilization Loss
%
6.7-1 0-6
-------
core at depths of 5, 15 and 25 crn (total of
30 samples, 10 at each of these depths). Due
to the small soil volume used in the combus-
tion method, each sample was divided into
three subsamples (total of 90 subsamples)
that were analyzed for their total N and C
content.
In practice, N1 , .. was calculated from
r ' application
Equation [3.2] to meet the projected A/'/anf
and fo;during the subsequent time inter-
val. The projected plant uptake for each
time interval was determined from potential
N uptake curves for crops under optimum
growth conditions (Belanger and Richards,
2000;Gibson et al., 2007; Rahman et al.,
2001), and the projected mineralization rate
was estimated from literature values (Stenger
et al., 2001) or from the previous time step.
The blending ratio before each application
was determined by matching simultaneously
/ ,. , and N1 , ,
application application
where / ,. t. =/„.,...+
application DWW
le//and NlapPl,cst,on= N(,el, '.ell + ^/^ and
the subscripts DWW and well denote the water
source.
Suspended sediments concentration (SSC)
of the DWW were measured by centrifug-
ing a known volume at 2040 times gravity
for 20 min, decanting the liquid phase and
measuring the remaining solid after drying at
60°C for 48 h (ASTM D 3977-97 - Method A).
The TDS was assumed to be correlated to the
EC (1 dS-m ' = 10 meq-L1) and was mea-
sured with an EC meter (M33.1, Agricultural
Electronics, Montclair, CA). The concentra-
tions of plant available K (Olsen bicarbonate
method) and P (ammonium acetate method)
in the soil profile were determined at the
beginning and the end of the 2007 growing
season, whereas salt concentrations (EC and
TDS) in the root zone were measured before
each irrigation event.
The T-Statistic (T-Test) was used to evaluate
significant differences (P<0.05) between
cyclic and blending application strategies
during 2007 (Sigmaplot 11, Systat Software
Inc., CA).
and Discussion
Results from triticale (winter 2007) and sor-
ghum (summer 2007) data are discussed
below in sections entitled Management
Considerations for Salinity,
Considerations for Organic Nitrogen, and Plant
Available N, P and K. These two cereal crops
do not form nodules to fix atmospheric N, and
E0| therefore reflects exchange due to miner-
alization. In contrast, alfalfa (summer 2008)
is a legume that forms nodules and may fix
atmospheric N. The additional N source from
fixation poses additional challenges for effi-
cient NMP implementation that is discussed
below in a separate section entitled Nitrogen
Fixation - Alfalfa 2008.
Management Considerations for Salinity
Water balance information for 2007 winter (A)
and summer (B) growing seasons are pre-
sented in Table 3.2. Rainfall occurred during
winter (total of 28.75 mm) but was absent
during summer. The value of ETp was lower
during winter than during summer. The final
water application amounts were adjusted to
include a leaching factor of 0.2 for the first
30 days and 0.1 for later times in order to
leach excess salts from the root zone and
to minimize downward migration of N03\ A
system malfunction, however, delivered an
extra 55.4 mm of well water to the cyclic plot
on day 58 of the summer growing season. The
total drainage flux (average of the blending
and cyclic strategies) below the root zone was
23.9 mm and 56.4 mm throughout the winter
and summer growing seasons, respectively.
Figure 3.2 presents the absolute value of soil
water pressure head (\h\) in the soil profile
as a function of day after emergence (DAE)
for the blending and cyclic strategies during
the 2007 summer growing season. Changes
in \h\ were restricted only to the upper 60 cm
of the soil profile due to the accurate water
mass balance on both plots. In general, \h
followed the water application events: decreas-
ing after irrigation (soil becomes wetter) and
increasing with time between irrigations (soil
-------
Table 3.2. Potential and actual evapotranspiration (ET), crop coefficient, rainfall and water application
during the growing season of triticale during winter 2007 (A) and sorghum during summer 2007 (B), DAE is
day after emergence.
A
DAE
15-29
30-36
37-50
51-58
59-65
66-72
B
DAE
5-28
29-35
36-44
45-49
50-58
59-62
63-70
potential ET
mm
100.2
30.1
84.4
36.7
39.8
40.1
potential ET
mm
258.6
86.7
94.0
47.1
86.0
33.5
59.8
Crop coefficient and
leaching factort
0.5
0.5
0.9
1.1
1.2
1.2
Crop coefficient and
leaching factort
0.5
0.75
0.9
1.0
1.0
1.05
1.1
Actual ET
mm
50.1
15.05
75.96
40.37
47.76
48.12
Actual ET
mm
129.3
65.02
84.6
47.1
86.0
35.17
65.78
Rainfall
mm
7.25
0.00
4.25
4.00
13.25
0.00
Rainfall
mm
0
0
0
0
0
0
0
Water application
mm
42.9
15.0
71.7
36.3
34.6
48.0
Water application
mm
135.1
62.4
90.5
48.5
86.6
35.3
66.6
800
600 •
_400
E
SMM
•d
ro
4, 0-1
9
in
•
£ 500 -
Q.
•g 40°
l300^
200 -
100 -
o -
Blending 2007
^—^ 30cm
60cm
.__. .90cm
-120cm
-150cm
— •-• -180cm
Cyclic 2007
fCrop coefficient was measured based on water mass balance in the root zone by using weighing lysimeters. The
leaching factor was 0.2 during the first 30 days and 0.1 through the rest of the growing season.
becomes drier). Figure 3.2 shows that aerobic
conditions were maintained during the major-
ity of the season. The soil water pressure head
below the root zone (i.e. <-90 cm) was gener-
ally steady throughout the growing season.
The system malfunction on the cyclic plot
at day 58, however, caused a decrease in \h\
below the root zone (<-60 cm).
The increase in the EC of soil solution (ECw)
of the root zone (-30 and -60 cm) over time
due to the use of irrigation water with high
TDS, a low leaching factor, and concentra-
tion of salts by ET is presented in Figure 3.3.
The ECw of the root zone increased from
1 to 3 dS-rrr1 for the blending strategy and
from 1 to 2.5 dS-rrr1 for the cyclic strategy (no
significant difference in the level of P<0.05
were found between the final values of EC of
w
each treatment). The extra water application
on the cyclic plot produced greater leaching
and hence a lower final value of ECw. Only
minor changes in the electrical conductivity of
the soil solution (ECw) were detected below the
root zone during the growing seasons (data is
not presented), due to the low leaching factor
20
40
60
DAE
Figure 3.2. The absolute value of the soil water
pressure head (\h\) in the soil profile as a function
of day after emergence (DAE) for the cyclic and
blending water application strategies during the
sorghum 2007 growing season.
-------
C
O
O 3
s 8
I 's
^
LLJ
3 -
2 -
1 -
• -30cm
O -60cm
V -90cm
A -120cm
Blending 2007
4
— Water application
8
1
[
»
5
i
i
f
! 1 i5
1
t
4
's
5
_«
i
^
i
/
»
)
X
).
Cyclic 2007
^H Extra water application
3
I
T * "
fir*
9 1 lc
nir
I
5*
)
III 1
Feb-07
Apr-07
Jun-07
Aug-07 Feb-07
Time
Apr-07
Jun-07
Aug-07
250
200 ~ C
O •
3
"O —
150 8? o
>. VI
ii
100 =5 o
5 *"
50
§
Figure 3.3. Electrical conductivity of the soil solution (ECw) over depth and total dissolved solids (TDS)
load during 2007 for the blending and cyclic water application strategies.
that was implemented at these sites. The
measured values represent the salt load under
a conservative NMP approach that applied
only a fraction of the total N that was required
by the plant with DWW. If 100% of the plant
N had been applied by DWW, then the accom-
panying salts would increase the ECw in the
root zone to higher levels. High salt levels
in the root zone may restrict plant growth,
and accordingly water and nutrient uptake.
If this reduction in ET is not considered at a
NMP site, additional leaching and contami-
nant migration will occur. An optimum point
likely exists between the benefits of nutrient
application and the detrimental effects of
salt accumulation on crop yield. This point
is strongly dependent on the salt tolerance of
the crop, suggesting that NMP should use only
salt tolerant crops.
Minimizing the potential adverse effects of
salts on plant growth is commonly achieved by
leaching excess salts below the root zone. The
timing of salt leaching may be a crucial man-
agement decision in NMPs because organic
soil N continues to be converted to inorganic
N forms (NH4+, N02-and N03-) during the
fallow season. A pre-irrigation at the beginning
of a new growing season, or seasonal rains
during the fallow season may result in migra-
tion of inorganic N, especially NO -, below the
root zone towards groundwater (Feng et al.,
2005; Woodard etal., 2002).
Figure 3.4 demonstrates this scenario by
presenting the concentrations of N-(N02+N03)
and N-NH4 in the soil profile at the end (final-
after harvesting) and beginning (initial- after
seasonal rains and pre-irrigation) of consecu-
tive growing seasons. The graphs of triticale
2007/sorghum 2007 and the sorghum 20077
alfalfa 2008 fallow seasons show that the soil
profile at both strategies was depleted from
inorganic N at the beginning of the fallow
seasons. Conversely, high concentrations of
N-(N02+N03) were found along the soil pro-
files at the end of the fallow seasons. No
significant difference (P<0.05) between strat-
egies was found in the initial and final values
of N-NH4 and N-(N02+N03) in the soil profile.
A mass balance of the inorganic N in the pro-
file revealed that 9.67 and 17.95 g of N-rrr2
was mineralized for the blending strategy and
3.40 and 19.5 g of N-rrr2 for the cyclic strat-
egy during the triticale 2007/sorghum 2007
and the sorghum 2007/alfalfa 2008 fallow
periods, respectively. These values are equiva-
lent to mineralization rate of 2.37E-04 day"1
and 4.25E-04 day-1 for the blending strategy
and 2.93E-04 day-1 and 4.72E-04 day-1 for
the cyclic strategy. Leaching excess salts is
therefore recommended right after harvesting,
-------
Blending- Triticale 2007 / Sorghum 2007
-100
-150
U
0.
-------
figures. Cumulative amounts of potential
N1
plant
EOI, and the sum of
En. and N1 .. ,
01 application
as a function of day of the year (DOY) are
presented for each N°sgjl level. The year is
divided into two growing seasons (winter
and summer) and two fallow periods (fall
and spring). For winter and summer crops
the cumulative amounts of potential N1 lgnt
were assumed to be 16 and 22 g of N-rrr2
and mineralization rates were assumed to be
2E-04 and 3E-04 day-1, respectively. The
error bars reflect the assumed variance (40%)
in the mineralization rate. Differences in the
predicted cumulative amounts of EOI are due
to differences in the initial organic reservoir.
^'application 's determined from the difference
between potential N'plant and EOI, while consid-
ering the uncertainty.
For the case of a low N° ,
so/I
-01
N1 .. and N1 .. ,
son application
will under apply N1
reservoir, match-
ing between potential N'plant and N'application is
practical with low deviations due to the minor
amounts of N°soil and N°application. This well
controlled NMP condition is similar to fertiga-
tion. The second scenario, intermediate N°ojl
reservoir, is representative of our field study
during 2007 and requires the consideration of
This second scenario
,plant when uncertainty in
the mineralization rate is considered and the
N°ojl reservoir will deplete over time and shift
the system to the well controlled NMP condi-
tion shown in the upper graph. In the third
scenario, the entire N1 lgnt is dependent on
the N°sgjl pool and mineralization rates. When
considering the high uncertainty in EOI, sig-
nificant under-application of N is likely with
corresponding yield reductions. Consequently,
the N°sojl reservoir will be depleted over time.
If uncertainty in the mineralization rate is not
considered with scenarios 2 and 3, then yield
reduction can be minimized but the risk for N
migration below the root zone toward ground-
water increases.
Figure 3.5 indicates that the potential
migration of N below the root zone can be
minimized when N°sojl is low. The majority of
the SSC from the DWW was therefore removed
as part of the NMP implementation with an
inclined screen separator, a sedimentation
tank and a sand filter. In addition to this
NMP consideration, lower SSC in the DWW
40
30
40
O
D)
O
1
3
E
3
O
30
20
10
Dominant by Inorganic N
Combined Inorganic - Organic N
^ Potential plant uptake
Predicted mineralized N
—• Predicted mineralized N &„
Inorganic N application
Dominant by Organic N
40
30
20
10
0 50 100 150 I 200 250 300 350 I
DOY
Winter Fallow Summer Fallow
crop season crop season
Figure 3.5. Three conceptual approaches for
nitrogen NMPs based on low (top figure-333 g
of N-m2), intermediate (middle figure-1666 g of
N-m2), and high (lower figure-3333 g of N-m2)
soil organic N reservoirs. Cumulative amounts of
potential plant N uptake (16 and 22 g of N-m2 for
winter and summer crops), predicted mineralized
N, and the sum of mineralized N and applied inor-
ganic N are presented as a function of the day of
the year (DOY). Mineralization rates were 2E-04
and 3E-04 day1 for winter and summer seasons.
The error bars reflect the assumed variance (40%)
in the mineralization rate.
-------
allowed us to use a water application system
with a high uniformity (micro-sprinkler). The
inorganic and organic N species and their
distribution in applied DWW during the 2007
winter growing season (triticale) are presented
in Figure 3.6. The organic N was relatively
constant throughout the growing season (6.12-
7.78 mg L4), and the inorganic N fraction
increased from 90% to 97% of the total N.
Figure 3.6 also indicates that N-NH4 was the
dominant N form in the DWW (Campbell-
Mathews et al., 2001; Cameron et al., 2002;
Wang et al., 2004). Temporal variability
in the fraction and amount of N-NH4 and
N-(N02+N03) in the DWW is due to NH3
volatilization and nitrification during storage
(Bussinkand Oenema, 1998). The DWW
treatment was found to also have an effect
on many chemical properties (Table 3.3). A
small reduction in EC was measured and was
attributed to adsorption (Rodgers et al., 2005)
and removal of suspended solids.
200
_ 150
100
Organic N
N-(NO2+NOJ
N-NH,,
30
72
Figure 3.6. The composition, distribution, and
amount of organic and inorganic N species (mg L1)
in dairy wastewater during the triticale 2007 grow-
ing season,
Plant Available N, P, and K
NMPs need to be designed to maintain soil
fertility and to minimize the risk of N migra-
tion below the root zone. A conservative NMP
approach was implemented at our field site
for triticale and sorghum during 2007 in an
attempt to transition the soil organic reser-
voir to an optimum value of /V°o/;. During
Table 3.3. Salts and macro-nutrients of a raw and
treated dairy wastewater (DWW). DWW was sequential-
ly treated by solid separator, sedimentation tank and
sand filter.
Category
General
Salts
(mg-L-1)
Macro-
Nutrients
(mg-L1)
Component
EC (dS-nr1)
SSC (mg-L1)
PH
Na
Ca
Mg
Cl
S-S04
HC03
N-(NH4+N02+N03)
Organic N
K
Total P
Raw
DWW
3.7
1611.4
7.57
182.5
378.3
243.1
174.2
109.1
2163.1
157.4
55.5
404.9
39.0
Treated
DWW
3.2
199.1
8.16
149.9
299.3
231.4
175.8
70.4
1829.0
145.2
6.9
380.5
27.9
this transition phase, the N° , reservoir was
r ' so/I
depleted over time in order to minimize the
migration of N03 below the root zone. Once
an optimal value of N°ojl is achieved then it
can be maintained by matching rates of domi-
nant N sources (DWW addition) and sinks
(plant uptake).
NMP results for triticale and sorghum during
2007 are summarized in Figure 3.7. Here
N'soii ' N°son and cumulative values of N'plant ,
EN1 N1 - N1 in the
01' drainage ' application atmosphere
root zone are presented over time (no signifi-
cant difference at P<0.05 level were found
between cyclic and blending treatments).
Initial low N1 , and minor changes in N' .
soil ° drainage
over time were measured for both cyclic
and blending strategies and crops. Average
N1 , , - N1. , for the triticale and sor-
application atmosphere
ghum crops were 48% and 28% of the total
N'plant , hence the dominant N source was EOI
(mineralization). An additional factor influ-
encing N'a lication was the low ET values during
winter. Implementation of the cyclic strategy
during these low ET conditions could not add
sufficient N to match plant requirements for
-------
several weeks, without adding DWW in excess
of ET,
Decreases in N°oj, over time are shown in
Figure 3.7 for both triticale and sorghum
plots as a result of mineralization. Average
mineralization rates were 1.23E-03 day1
during the triticale growing season and
2.54E-03 day^1 during the sorghum growing
season. The elevated rates during summer
were associated with higher soil temperatures
(the average soil temperatures for the winter
and summer growing seasons at -15 cm were
15.6°C and 24.5°C, respectively) that acceler-
ated the mineralization process (Watts et al.,
2007). These mineralization rates are lower
than earlier studies using DWW (Feng et al.,
2005; Stenger et al., 2001). This is likely
due to the low organic content of the applied
DWW that was treated, and the fact that most
of the organic N in the soil was plant residuals
(low C to N ratio). Mineralization continues
during fallow periods (Figure 3.4) and this
lead to differences in N°sojj between growing
seasons, especially for the long fallow period
(241 days) between the harvest of sorghum
and emergence of alfalfa. A comparison
between measured and calculated (N mass
balance) values of N°soj! at the beginning of
winter 2008 revealed no significant differ-
ence. The final calculated and measured N°soil
reservoir at the beginning of winter 2008 was
found to be 267.9±45.1 and 220.2±44.6 g
for the blending strategy and 276.6±39.8 and
236.3±38.9 g of organic N-nr2 for the cyclic
strategy.
The typical concentration ratios of N, P and
K in the applied DWW were 10:1.93:26.2
(Table 3.3). Direct measurements of P and
K contents in plant tissues were not made
during the course of the growing season.
However, plant uptake rates of N, P, and K by
wheat and sorghum have been reported to be
10:1.9:11.5 and 10:1.4:8.75, respectively
(Bar-Tal et al., 2004; Vanderlip and Reeves,
1972). If these nutrient uptake ratios are
assumed, then P and K will be applied in
excess and accumulated in the root zone
when a NMP based on N is implemented to
meet crop demand. Excess application of
P becomes an environmental concern when
surface water runoff and shallow water tables
can mobilize the P into surface water bodies
(Wang et al., 2004). Yet, arid and semi-arid
environments are mostly associated with
deep water tables, a high capacity of mineral
soils for P adsorption and limited runoff to
nearby surface water due to efficient water
application.
Figure 3.8 presents concentrations of plant
available K and P (phosphorus from phos-
phate, P-P04) in the soil profile at the
beginning of the 2007 winter growing season
(initial) and the end of the 2007 summer
crop season (final) for both blended and cyclic
dairy wastewater application strategies. The
distributions of K and P exhibited changes
primarily in the upper 50 cm of the soil pro-
file, where roots were most active in water and
nutrient uptake. The actual deficit N1 , ,
apfjUCaiiOn
(Figure 3.7) corresponded to 5.05 g of P-rrv2,
relative to an estimated 8.65 g of P-rrr2 taken
up by the crop. Therefore, the final soil P con-
tent with both strategies is less than the initial
content. However, if N'app!jration is matched to
N1 lant as the main N sink, P will be applied in
excess and accumulate in the soil. Similarly,
excess amounts of K in the DWW and the large
pool of K in the soil profile causes accumula-
tion of K in the root zone for both cyclic and
blending strategies.
Nitrogen Fixation - Alfalfa
Since no significant differences in cyclic and
blended treatments were observed during
2007, only the blended treatment was imple-
mented on alfalfa during 2008. In contrast to
triticale (winter 2007) and sorghum (summer
2007) crops discussed above, alfalfa (summer
2008) is a legume and may also obtain N
through fixation. The implications of N fixa-
tion on efficient NMP implementation will be
discussed below.
-------
o
O)
_>
'•?
m
3
|
O
50
40
30
20
10
£ 30
20
10 -
Triticale Blending
ff-i
N'plant
soil
~^ ™ ^ supplied'^ atmosphen
—A.. \|O
1N supplied
""* ~ N drainage
-D- EOI
-*- N°SOil
fftfi
Sorghum Blending
O <
400
300
1
V H
>
> [
1V
f [
-4
40
60
100
75
50
25
Alfalfa Blending
HI> EOT and Symbiotic fixation
$ Measured N s
100
150
200
" - 400
- 300
- 200
- 100
DAE
Figure 3.7: So/7 inorganic and organic N reservoirs ( N1^ and N°sojl) and cumulative values of N
uptake by plant ( N1 lant ), exchange to/from organic and inorganic N forms (EOI), N loss to drainage
( N!drainage ), supplied organic N (N°applicatjon), and supplied inorganic N minus loss to the atmosphere
( N application ~ ^atmosphere ^//? ^e root zone are Presented over time. Data is presented as g of N-rrr2 for triticale
(30 cm root zone), sorghum (60 cm root zone) and the alfalfa (90 cm root zone). Error bars represent
measured or calculated standard deviations. The value of EOI was obtained by closure of the mass balance
(Eqs. 3.2 and 3.3), and reflects changes due to mineralization for triticale and sorghum, and mineralization
and N fixation for alfalfa.
-------
o
P-PO4(ppm) K (ppm)
0 10 20 30 40 50 60 70 0 50 100 150 200 250 300 350
-25 -
-50
Q. -75 H
0)
Q
-100 -
-125 -
Blending initial
-o Blending final
-T- Cyclic initial
•••&- Cyclic final
Potassium
Figure 3.8. Plant available phosphorus (P-P04) and potassium (K) in the soil profile at the beginning (ini-
tial) of the triticale and end (final) of the sorghum growing seasons in 2007 for the blending and cyclic water
application strategies.
Similar to triticale and sorghum data shown
in Table 3.2 and Figure 3.2, an efficient
water mass balance was implemented for
alfalfa during 2008. Dry phytomass, cumula-
tive water application and cumulative ETactual
throughout the five growing cycles of alfalfa
in 2008 are presented in Figure 3.9. The
error bars represent an estimation of the
measuring error associated with the sampling
procedure of alfalfa (10%). Dry phytomass
ranged between 325 g-rrr2 at the last grow-
ing cycle, to 750 g-rrr2 at the second growing
cycle (ending July 24th). The relatively slow
growth rate during the last cycle was due to
seasonal variation in climate; i.e., the aver-
age maximum and minimum air temperatures
and solar radiation during this period were
relatively low with values of 23.1°C, 3.07°C
and 265.3 W-rrr2, respectively. In contrast,
the low dry phytomass of the first cycle was
related to the extensive weed growth that
interfered with the normal development of the
alfalfa during this period. The ratio between
the total water applications (1626.7mm) to
the total ET t , (1455.5 mm) corresponded to
3CTU3!
a leaching fraction of 11.7%. The calculated
crop coefficients, Kc, were consistent with
published data (Allen et al., 1998) and varied
between 0.4 after harvesting to 1.2 when full
foliage cover was reached. Similar to 2007
data shown in Figure 3.2, changes in \h\ were
restricted only to the upper 60 cm of the soil
profile due to the accurate water balance.
_ 1000
^
E
g, 600
Q.
| ™
I
• Actual evapotranspiration
D Water application
-•- Phytom
^ / / .
_
II
Sep Nov
Time (month)
Figure 3.9. Dry phytomass, cumulative water ap-
plication and actual evapotranspiration throughout
the five growing cycles of alfalfa in 2008.
The alfalfa 2008 growing season started after
a long fallow period (Nov. 2007-April 2008).
A total of 25.7 cm of rainfall occurred during
the fallow period that leached salts further
into the soil profile. Similar to the 2007 data
presented in Figure 3.3, the TDS increased
over the growing season. The TDS in the
soil profile (0 to -170 cm) at the beginning
-------
of the growing season (May 2008) was
937.7±338 g-rrr2 and at the end (Dec. 2008)
was 1391.7± 265 g-nr2. This increase was
pronounced at all depths and especially the
upper 30 cm (98.8 g-rrr2 versus 254.3 g-nr2).
Figure 3.7 shows the various components of
the N balance that were measured for alfalfa
during 2008. Since alfalfa can fix atmo-
spheric N through nodules, the value of EOI
now accounts for the net exchange between
organic and inorganic N forms due to both
mineralization and atmospheric N fixation.
Due to this additional complication, only
initial and final values of measured N° , are
so//
shown in Figure 3.7 for alfalfa. Recall that a
very conservative NMP approach was imple-
mented for triticale and sorghum crops in
2007 that depleted the soil organic N. In
contrast, a less conservative NMP strategy was
implemented for alfalfa during 2008 due to a
greater value of N!, t , a deeper root system,
° plant ! r J J
and the desire to suppress atmospheric N
fixation. Specifically, the cumulative value of
A/', , was 104 g of N-m2 for alfalfa in com-
plant °
parison to 60 g of N-nr2 for both triticale and
sorghum. The depth of the root zone, where
roots are most active in water and nutrient
uptake under irrigated conditions, for alfalfa
was approximately 90 cm in comparison to 30
and 60 cm for triticale and sorghum, respec-
tively. A larger root zone allows nutrients to
be extracted over a larger area and to imple-
ment a more flexible schedule for lagoon water
application. In addition, by maintaining a
high N concentration in the root zone, we can
minimize the need of the plant to seek for
alternative N sources and theoretically apply
more DWW. The value of N1 , .. - N' .
application atmosphere
for the alfalfa was selected to be 76% of the
total N'lant . This implies that the remaining
24% of AM|ant comes from mineralization or
atmospheric N fixation.
Several implications for the less conserva-
tive NMP strategy that was implemented on
alfalfa during 2008 are discussed below.
First, an increase in measured A/0., is shown
" C/l//
in Figure 3.7 for alfalfa. This observation
can be attributed to three factors, namely:
i) intensive root growth of the alfalfa that
increased the total A/'^,; ii) the less conser-
vative NMP approach yielded a higher ratio
of A/; , .. - N1, . to the total A/;. •
application atmosphere plant J
and iii) the mechanism of N fixation through
nodules reduced the need for mineralized
N°saj, and induced immobilization of inor-
ganic N. Amounts of N in excess of plant
uptake requirements were apparently gener-
ated due to these factors. Consequently, 15%
of the supplied inorganic N was measured
to be steadily drained below the root zone
throughout the growing season under the
applied leaching factor of 11.7%. Additional
research is therefore needed to optimize NMP
performance with leguminous crops.
-------
4.0
NMP Results-
Indicator Microorganisms
Background
NMPs implicitly assume that pathogenic
microorganisms in CAFO lagoon water will
be retained and inactivated/degraded in the
root zone, so that food and water supplies are
protected. This assumption has not yet been
thoroughly tested. The objective of this work
was to test the hypothesis that fecal indicator
microorganisms (Enterococcus, total E, coli,
fecal coliforms, and somatic coliphage) will
be retained and die-off in the root zone of a
NMP application site. To this end, the fate
of indicator microorganisms at a field NMP
site was monitored during winter and summer
growing seasons. Additional information
was obtained in well controlled laboratory
experiments in order to: (i) identify mecha-
nisms that controlled the fate of the indicator
microorganisms at the NMP site, and (ii) to
develop recommendations for improved NMP
performance. Specifically, batch experiments
were conducted to quantify the survival char-
acteristics of the indicator microorganisms.
Small-scale packed column experiments were
conducted to study the influence of the soil
matrix on transport and retention behavior.
Finally, a large undisturbed column experi-
ment was conducted to assess the influence
of soil structure on indicator microorganism
transport and survival under a worst case sce-
nario of ponded infiltration and redistribution.
and
Site
The experimental field site is located
in San Jacinto, CA (33°50'22" N,
117°00'46" W), next to a dry river bed and
has a shallow perched water table at a depth
of -220 cm. The experiments were conducted
on a 6x6 rn plot that was described in detail
in Chapters 2 and 3. Soil cores were periodi-
cally collected from the NMP plot boundaries
for microbial analysis as described below.
Before initiating the NMP, extensive field and
laboratory studies were conducted to charac-
terize the soil hydraulic properties, the water
flow behavior, and the transport of a conser-
vative solute tracer on these plots. Detailed
information is provided in Segal et al. (2008
and 2009) and in Chapter 2.
For accurate NMP implementation it was
necessary to conduct a treatment on the
dairy wastewater using an inclined screen, a
sedimentation tank, and sand filtration. This
treatment significantly decreased the organic
load of the wastewater and thus increased
the proportion of plant available inorganic N
relative to organic N, which was necessary for
efficient management of the NMP (Segal et al.
2010). The indicator microbial composition
of the dairy wastewater before and after this
treatment was measured as part of this work.
Detailed information on the NMP design and
implementation is provided in Segal et al.
(2010) and in Chapters.
Analysis and Sampling for Indicator
Microorganisms
Representative viral (somatic coliphage) and
bacterial (total E. coli, fecal coliforms, and
Enterococcus) indicator microorganisms were
monitored at the NMP application site and
in laboratory experiments discussed below.
These microbes are commonly associated
with fecal contamination and are typically
found in high concentrations in animal wastes
(Havelaar, 1986; Bradford eta!., 2008).
Somatic coliphage, such as $X174, are very
persistent bacterial viruses that resemble
pathogenic viruses in their fate and behavior
-------
(Schijven and Hassanizadeh, 2000), but
are harmless to humans and can be pre-
pared and quantified easily. Total E. coll and
Enterococcus are frequently used as indicators
of bacterial pathogens that are found in the
intestinal tract of humans and warm-blooded
animals.
The concentration of indigenous somatic
coliphage in aqueous samples was deter-
mined using the double agar overlay Method
1601 (USEPA, 2001W with bacterial host
Escherichia coliCN-13 (ATCC 700609). In
brief, 1 ml of log phase host culture and
1 ml sample was added to borosilicate test
tubes (Thermo Fisher Scientific, Waltham,
MA) containing 4 ml of trypticase soy agar
(TSA) (BD Diagnostic Systems, Sparks, MD)
supplemented with 1% nalidixic acid (NA)
(Sigma-Aldrich, St. Louis, MO). The mixture
was poured onto sterilized 100 x 15 mm
plastic petri plates (Thermo Fisher Scientific,
Waltham, MA) and allowed to solidify for
15 min, after which, they were inverted and
incubated for 16 hat 36°C±1.0°C. The
number of plaque forming units (PFU) in
the plates was determined by counting the
plaque density under a darkfield colony
counter (Leica, Buffalo, NY). All coliphage
assays were run in duplicate and diluted as
necessary.
Aqueous concentrations of total E. coli, fecal
coliforms, and Enterococcus were determined
using the conventional spread plating method
(Clesceri et al., 1989). A 100 ul sample
was plated on Chromagar ECC (CHROMagar
Microbiology, Paris, France) plates for total
E.Coli, on mFC agar (BD Diagnostic Systems,
Sparks, MD) plates for fecal coliforms, and on
KF agar (EM Science, Gibbstown, NJ) plates
for Enterococcus. The plates were inverted
and incubated at 37°C for total E. Coli (24 h)
and Enterococcus (24-48 h), and at 44.5°C
for fecal coliform (24 h). The bacterial colony
forming units (CFU) were then counted. All
bacterial assays were run in duplicate and
diluted as necessary.
Dairy wastewater samples were analyzed
directly according to the procedures outlined
above. Soil samples were collected and
weighed under field and oven dried condi-
tions. For microbial analysis, a 10 g sample
of the field soil from each depth increment
was placed in a 50 ml sterile polypropylene
centrifuge tube (Thermo Fisher Scientific,
Waltham, NJ), 20 ml of phosphate buffered
solution or de-ionized water was added, and
the solution was gently mixed on a Eberbach
shaker (Eberbach corporation, Ann Arbor,
Michigan) for 30 min, and then the soil solu-
tion was allowed to settle for 5 min. This
solution was subsequently analyzed for
microbial concentrations using the outlined
procedures, and the concentrations were cor-
rected for the amount of soil and solution in
each depth increment. Results from column
transport and batch survival experiments
discussed below demonstrate that reason-
able mass balance and microbe recovery was
achieved when using this protocol with soil
from the field plot and the considered indica-
tor microorganisms.
For laboratory transport experiments that
utilized spiked concentrations of indica-
tor microorganisms the following protocols
were followed. Stock bacteria, E. coli (ATCC
11775); Enterococcus faecalis (ATCC 19433),
and somatic coliphage, fyX174 (ATCC 13706-
Bl) were grown 24 h prior to the laboratory
experiments. E. co//was propagated in trypti-
case soy broth (TSB) (BD Diagnostic Systems,
Sparks, MD) and Enterococcus faecalis was
grown in LB broth (FisherBiotech, Fair Lawn,
NJ). Both bacterial strains were incubated
overnight at 37°C ± 1.0°C. The somatic
coliphage host, E. co//(ATCC 13706) was
incubated in TSB for 18 h at 37°C + 1.0°C.
Bacteriophage §X174 was then added into
the broth culture of host bacteria and allowed
to propagate overnight. The host, E, coli,
was removed from the suspension, contain-
ing bacteriophage §X174, by centrifugation
at 17329 x g for 10 min. The other bac-
teria were harvested from the growth broth,
-------
and washed twice in PBS and centrifuged at
17329 x g before using them in the transport
experiments,
Laboratory Experiments
Batch - Batch experiments were conducted
to assess the survival characteristics of the
indicator microorganisms. Experiments were
conducted by placing 7,5 g of field soil into
a 20 ml glass scintillation vial. Two condi-
tions were studied to determine the influence
of biotic factors on the survival of the indica-
tor microorganisms. One set of experiments
employed native soil and wastewater, and the
other used sterilized (autoclaved) soil and
wastewater spiked with indicator microorgan-
isms. Lagoon water (0.43 ml) and deionized
water were added to the vials to achieve a
water saturation of 80%. The porosity of the
soil in the vials was estimated from the soil
volume and mass to be 0.43. The deionized
water and wastewater were then thoroughly
mixed with the soil using a metal spatula, and
the vial was capped (a pin hole was used for
aeration) and wrapped in aluminum foil. The
vials were subsequently stored at 25°C and
analyzed for indicator microorganism concen-
trations at desired times. Separate vials were
used for each time (0, 6, 10, 24, 48, 80,
144, and 168 h). Hence, a total of 8 vials
were independently analyzed for native and
sterilized batch experiments.
Microorganism death/inactivation is typi-
cally assumed to follow a simple first-order
decay model (Yates et al., 1987; Schijven
and Hassanizadeh, 2000). However, virus
inactivation and bacterial growth in soils have
been observed to exhibit more complex sur-
vival behavior than the simple first-order decay
model. The following model was proposed by
Sim and Chrysikopoulos (1996) to account for
observed time dependent inactivation:
dCT
dt
= -X0 exp(-ocOCr
[4.1]
length], is the total microorganism concen-
tration, t [T, denotes units of time] is time,
\t [T1] is the initial rate of die-off/inactiva-
tion, and a [T1] is the resistivity coefficient to
die-off/inactivation. The solution to Equation
[4.1] is given as (Sim and Chrysikopoulos,
1996):
where CT [N L~3; where N is the number
of microorganisms, and L denotes units of
Ti
where CTi [N L3] is the initial total microorgan-
ism concentration. The parameters A,o and a
will be fitted herein to measured batch survival
data using a nonlinear least squares optimiza-
tion routine.
Packed Column - Packed soil column experi-
ments were conducted using top soil from the
field site to assess the transport potential of
Enterococcus, somatic coliphage, and E, coli
under well controlled conditions. Sterilized
field soil and wastewater that was spiked with
indicator microorganisms were used for this
purpose,
Kontes Chromaflex chromatography columns
(Kimble / Kontes, Vineland, NJ) made of
borosilicate glass (15 cm long and 4.8 cm
inside diameter, equipped with an adjustable
adapter at the top) were used in the transport
studies. The columns were packed dry with
oven-dried soil from the surface of the NMP
site. Well water from the field was slowly
pumped upward through the vertically oriented
columns at a steady flow rate for several pore
volumes (PV) to saturate the column. A peri-
staltic pump (Cole-Parmer, Vernon Hills, ID
was used for this purpose. Microorganism
transport experiments were subsequently
conducted by pumping tracer solution at a
steady rate through the column for several
PV, after which a three-way valve was used to
switch to the well water for several additional
PV. Effluent samples were continuously col-
lected in 16 x 150 mm borosilicate test tubes
(Thermo Fisher Scientific, Waltharn, NJ) using
an auto sampler (Isco, Inc., Lincoln, NE).
-------
Following completion of the transport
experiments, the concentration of indicator
microorganisms in the effluent and retained in
the soil was determined using the previously
outlined protocols,
Undisturbed Column - To study the influence
of soil structure on the transport and survival
of indigenous indicator microorganisms in
wastewater a larger undisturbed soil column
(length was 65 cm and the internal diameter
was 24 cm) was taken from the field site just
outside of the plot. This column was also used
in water flow and bromide transport experi-
ments, and details on the collection, handling,
instrumentation, hydraulic properties, and
solute transport characteristics of this column
are provided in the literature (Segal et al.
2009). In brief, the soil core was encased
in an acrylic cylinder that was pushed into
the soil with a hydraulic piston and then dug
out. The bottom of the column was placed
on a ceramic plate with a bubbling pressure
of 100 kPa and a hanging water column was
used to control the water pressure and to
collect effluent samples. The soil surface
boundary condition was selected to mimic a
worst case transport scenario of ponded infil-
tration in which saturated conditions enable
water flow and pathogen transport through
macropores and soil structure. In this case,
wastewater was instantaneously added to the
soil surface to a depth of 7 cm and allowed to
infiltrate into the profile. Following applica-
tion and infiltration of the wastewater, 2 cm
diameter and 65 cm long soil cores were
collected vertically from the column surface
at selected times and analyzed for concen-
trations of indicator microbes. The sample
locations were subsequently plugged with a
similar sized PVC tube that was capped. This
experiment was conducted at 25°C in a tem-
perature controlled room.
Results and Discussion
Site
The sequential treatment (inclined screen
separator, sedimentation tank and sand filter)
of the dairy wastewater at the field site had
a large effect on many microbial properties.
Table 4.1 provides representative information
on the concentration of indicator microorgan-
isms before and after this treatment. The
treatment produced almost a 2 log decrease
in the concentration of the various indicator
microorganisms. Concentrations of indica-
tor microorganisms in the treated wastewater,
however, still significantly exceeded the rec-
ommended U.S. standards for unrestricted
irrigation (USEPA, 2004). If pathogenic
microorganisms behave similarly to indica-
tors, then care must be taken in using dairy
wastewater as a source for irrigation water. To
protect surface and groundwater supplies it is
therefore essential that pathogenic microor-
ganisms in the dairy wastewater be removed
by treatment, or be retained and inactivated/
degraded in the root zone of NMP sites.
Table 4.1. Representative concentrations of
indicator microorganisms in raw and treated dairy
wastewater, and the percent removal by treatment.
Treatment included solid separator, sedimentation
tank and sand filter.
Indicator
Entemcoccus
(cfu ml-1)
fecal
coliform
(cfu ml-1)
somatic
coliphage
(pfu ml-1)
total £. co//
(cfu ml-1)
Raw
1.56E+06
2.47E+05
7.50E+03
9.30E+04
Treated
6.10E+04
6.00E+03
5.00+02
3.00E+03
%
Removal
96.1
97.6
93.3
96.8
Water mass balance information for the winter
triticale and the summer sorghum grow-
ing seasons are presented in Table 4.2 as a
function of day after emergence (DAE). The
accurate water application quantities (well
water + lagoon water) relative to the actual
-------
plant evapotranspiration (ET) restricted
changes in water pressure to the upper 60 cm
of the soil profile (Segal et al,, 2010). Hence,
the amounts of drainage below the root zone
(Table 4.2) were very small throughout the
winter and summer growing seasons and
most of this drainage occurred with well
water application (57.6 and 86.9% during
winter and summer, respectively). This find-
ing indicates that there exists little potential
for advective transport of the indicator micro-
organisms below the root zone under these
NMP management conditions. If the indica-
tor microorganisms survive in the root zone,
however, they may still be transported below
the root zone when water in excess of ET is
applied to this NMP site such as during the
fallow period with high precipitation events or
during pre-irrigations to leach salts. Hence,
pathogen survival in the root zone may pose a
risk to food and water resources at NMP sites.
Figure 4.1a-d present plots of the concentra-
tion of Enterococcus, fecal coliform, somatic
coliphage, and total E. coll in soil (S, N g"1)
as a function of depth at several differ-
ent DAE during the winter triticale growing
season, respectively. Table 4.2 also includes
the amount of well water and dairy wastewa-
ter that was applied to the NMP study site.
Irrigation water was applied at the end of the
intervals provided in Table 4.2, except for
the first interval in which the irrigation water
was applied periodically throughout the inter-
val. Low concentrations of total E. coll, fecal
coliform, and somatic coliphage were initially
present in the surface soil of the NMP site.
The Enterococcus concentrations, however,
were initially more abundant. After wastewa-
ter addition on DAE 36, the concentrations of
the indicator microorganisms in the soil core
increased dramatically. The highest concen-
trations occurred near the soil surface (top
Table 4.2. Actual evapotranspiration (ET), rainfall, well water application, wastewater application, and
drainage amounts during the winter (triticale) and summer (sorghum) growing seasons as a function of day
after emergence (DAE).
Winter
DAE
15-29
30-36
37-50
51-58
59-65
66-72
Actual ET
mm
40.1
13.5
68.4
36.5
43.0
43.2
Rainfall
mm
7.2
0.0
4.2
4.1
13.2
0.0
We!! Water
mm
42.9
0.0
0.0
36.3
0.0
48.0
Waste Water
mm
0.0
15.0
71.7
0.0
34.5
0.0
Drainage1
mm
10.0
1.5
7.6
4.1
4.8
4.8
Summer
DAE
5-28
29-35
36-44
45-49
50-58
59-62
63-70
Actual ET
mm
104.0
58.5
76.2
42.4
77.4
31.7
59.1
Rainfall
mm
0.0
0.0
0.0
0.0
0.0
0.0
0.0
We!! Water
mm
135.0
62.0
0.0
48.6
86.6
35.3
66.7
Waste Water
mm
0.0
0.0
91.51
0.0
0.0
0.0
0.0
Drainage1
mm
26.0
6.5
8.5
4.7
8.6
3.5
6.6
1- Value estimated from water balance
-------
(a)
0.0
3.0
f 6'°
Q.
Q 9.0
Enterococcus
15.0
a
d>
Q
(b)
0.0
3.0
9.0'
Fecal Coliform
• DAE = 36
— » — DAE = 60
•-••¥-- DAE = 65
....+ ... DAE = 71
4.0E+004
S (N g-1
15.0
r' )
I ,,'
'
• DAE = 36
— »— DAE = 60
,_,.*_.. DAE = 65
....*... DAE = 71
S(Ng-i)
(C)
0.0
Somatic Coliphage
—9 DAE = 36
— •• DAE = 60
...*_,. DAE = 65
.,,>.., DAE = 71
E
5
(d)
0.0
6.0
8.0E+002 1.2E+003
S (N g-1)
9.0
12.0
Total E. coli
» 7
f«-
/
• DAE = 36
— »— DAE = 60
•-••¥-•- DAE = 65
<••'»"• DAE = 71
O.OE+OOO 3.0E+003 6.0E+003 9.0E+003 1.2E+004 1.5E+004
S (N g-1)
Figure 4.1. Plots of the concentration of Enterococcus (Figure 4. la), fecal coliform (Figure 4.1b), somatic
coliphage (Figure 4.1c), and total E. coli (Figure 4.Id) in soil (S, N g1) as a function of depth at several dif-
ferent days after emergence (DAE) during the winter triticale growing season at the NMP field site.
4-6 cm) and then rapidly decreased with soil
depth. Over time concentrations of the indica-
tor microorganisms at the soil surface tended
to decrease, and by DAE 71 approached the
low levels that were found before wastewater
application (DAE 36).
Similar NMP experiments were conducted
during the summer on sorghum. Soil cores
were periodically collected from the field
site during the summer season and analyzed
for indicator microorganism concentrations.
Similar to the winter crop, only low levels of
Enterococcus were found in the surface soil at
the beginning and end of the summer season.
A more detailed NMP experiment was initi-
ated the following summer growing season
(alfalfa, Medicago sativa, was the crop) to
better understand the transport and fate of the
indicator microorganisms. In this case, the
sampling frequency (initial, and times=0, 14,
62, 134, and 206 h after wastewater appli-
cation) and number of soil cores (composite
of 3 cores) were much higher after a single
wastewater application event. In addition,
the gravimetric water content with depth was
measured at selected times after wastewater
application. Initial concentrations in soil were
low for the somatic coliphage and total E.
-------
co//(Crwas 0 CPU cm"3). Conversely, initial
concentrations for Enterococcus (Crwas 5E4
CPU cm"3) and fecal coliform (Crwas 1.3E4
CPU crrr3) in soil were higher. Approximately
10.5 cm of wastewater was applied to the
field site at an average rate of 0.95 cm h ].
Changes in the water content occurred over
depth and time (Figure 4.2a) as a result of
infiltration, redistribution, and then evapo-
transpiration in the root zone (top 60 cm of
soil). Figure 4.2b-e present plots of the nor-
malized concentration of Enterococcus, fecal
coliform, somatic coliphage, and total E. coli
in soil (S/CTi; cm3 g"1) as a function of depth
at selected sampling times after wastewater
application. Here the concentrations of a
given microorganism in soil were normalized
by Cj-for the entire soil core taken at time=0
h, i.e., denoted as C/;. The concentration
of indicator microorganisms in the soil was
high at time = 0 h, and reflected initial condi-
tions, transport during wastewater infiltration,
and retention of the indicator microorgan-
isms in the soil. At later sampling times
(greater than or equal to 14 h) infiltration has
ceased; redistribution and evapotranspiration
of water depletes water from the root zone
(Figure 4.2a). In general, the concentration
of the indicator microorganisms in the soil
cores is controlled at earlier sampling times
(t=14 and 62 h) by initial conditions and
microbe retention, and at later times (t=134
and 206) by microbe survival and/or inactiva-
tion. Microbe retention tends to be highest
at the soil surface and to rapidly decrease
with soil depth. Survival of fecal coliforms,
somatic coliphage, and total E. coli was appar-
ently short lived under these NMP conditions,
with concentrations in soil approaching low
background levels within 134 h after waste-
water application. This result is at least partly
due to the high summer soil temperature
which averaged 22.6°C at a depth of 7 cm
during this experiment. Cooler temperatures
that were measured during the winter grow-
ing season (20.3°C at a depth of 7 cm) are
expected to enhance microorganism survival
(Reddy et al., 1981; Yates et al., 1987) shown
in Figure 4.1. Conversely, final concentrations
of Enterococcus in soil were more pronounced
than the other indicator microbes. This may
occur as a result of differences in survival or
due to differences in the initial concentra-
tion in these microbes in the soil (C7 was 5E4
and 1.3E4 CPU crrr3 for Enterococcus and
fecal coliform, respectively) and wastewater
(C. in the wastewater was 8.9E04 and 1.7E3
CPU ml"1 for Enterococcus and fecal coliform,
respectively).
A more quantitative determination of the
transport and survival of the indicator micro-
organisms at the NMP site was not attempted
due to the relatively low initial concentration
levels of these microbes in the irrigation water
and our analytical detection limits, potential
spatial variability in microbe concentrations in
the field, incomplete control of the lower water
flow boundary conditions, and transients in
water content that were induced by infiltration,
redistribution, and ET. Laboratory experiments
that are described below were initiated to over-
come many of these limitations.
Laboratory
Figure 4.3 presents results from the batch
survival experiments conducted under sterile
(Figure 4.3a) and native (Figure 4.3b) condi-
tions at 80% water saturation for the various
indicator microorganisms. A semi-log plot
of CT/CT! as a function of time is shown in
this figure. The considered bacterial indica-
tor microorganisms had much greater die-off
under native than sterile conditions. This
indicates that biotic factors such as predation
and competition in native soil and wastewater
had a significant influence on the survival of
the considered bacterial indicator microor-
ganisms. Other researchers have reported
similar observations and this literature has
been reviewed by Van Veen et al. (1997).
Conversely, the somatic coliphage exhibited
similar survival/inactivation under both sterile
and native conditions, suggesting that abiotic
factors controlled die-off/inactivation for this
coliphage. The native condition is expected
-------
Fecal Coliform
60.0
30.0
Enterococcus
Somatic Coliphage
0.0
5.0
£ 15.0
Q.
O
a
20.0
25.0
5.0
10.0
15.0
20.0
25.0
(b)
;r
I
r>
30.0
O.OE'
30.0'
O.OEtOOO
1.0E-002
S CT.-1 (cm" g-1)
(d)
5.0E-003 1.0E-002
SC -1(
Total E. coli
30.0
-•— Time = 0 h
»— Time = 14 h
^-- Time = 62 h
•»... Time = 134 h
^ Time = 206 h
1.0E-002 1.5E-002
S CT.-1 (cm3 g-1)
Figure 4.2. Plots of the gravimetric water content (Qg) (Figure 4.2a), and the normalized concentration of
Enterococcus (Figure 4.2b), fecal coli form (Figure 4.2c), somatic coliphage (Figure 4.2d), and total E. coli
(Figure 4.2e) in soil (S/CT.) as a function of depth at selected sampling times (initial, and times=0, 14, 62,
134, and 206 h after wastewater application) during the summer alfalfa growing season at the NMP field
site. Microbial concentrations were determined from a composite of 3 soil cores at the indicated depth and
time, and normalized by the total concentration of a given microorganism measured in the entire core at
time=0 (C).
-------
0.010
0.001
Enterococcus
Fecal Conform
Somatic Coliphage
Total E. coli
0 25 50
100
rime (h)
1.000
0.100
0.010
0.001
(b)
Enteroi
— Fecal Coliform
-,. Somatic Coliphage
... Total E. coli
75 100
Tiine(h)
Figure 4.3. A semi-log plot of the relative total concentration (CT/CTi) as a function of time in batch survival
experiments under sterile (Figure 4,3a) and native (Figure 4,3b) conditions at 80% water saturation for the
various indicator microorganisms.
to be most representative of field conditions
at the NMP site. Table 4.3 therefore provides
a summary of the fitted die-off/inactivation
rates (Equation 4.2) and the standard error
coefficient on these parameters for the vari-
ous indicator microorganisms under native
conditions at 80% water saturation. The P
values for statistical significance of the cor-
relation coefficients were less than 0.00002.
It is interesting to note that fecal coliforms
had a lower die-off rate than Enterococcus
under native conditions (Figure 4.3b and
Table 4.3). This finding implies that the
observed differences in final concentrations
of these microbes in soil shown in Figure 4.2
was likely due to differences in the initial
concentrations of these organisms in soil and
wastewater. Packed soil column experiments
were conducted using sterilized soil (sandy
loam) from the root zone of the NMP field
site to better understand the transport of the
indicator microorganisms. The tracer solution
consisted of sterilized (autoclaved) wastewater
spiked with a known concentration of specific
microorganisms, and sterilized well water was
employed as the resident and eluant solutions.
Figure 4.4a presents breakthrough curves for
Enterococcus, total E. coli, and $X174 in the
packed soil column. Relevant experimental
Table 4.3. Fitted parameters (ko and a) from the
time dependent decay model (Equation 4.2) for
the various indicator microorganisms under native
conditions at 80% water saturation. The goodness
of model fit is quantified by the coefficient of linear
regression (R2) on natural log transformed data, and
standard error coefficient of the fitted model param-
eters is provided in brackets.
Indicator
Enterococcus
fecal coliform
somatic
coliphage
total E. coli
\
(h-1)
3.52E-02
(1 .31 E-03)
1.90E-01
(4.03E-02)
1.61E-01
(2.05E-02)
7.29E-02
(1 .22E-02)
a
(h-1)
3.54E-11
(4.32E-04)
7.81 E-02
(1 .86E-02)
2.95E-02
(4.65E-03)
1 .25E-02
(3.55E-03)
R2
0.992
0.933
0.973
0.963
conditions are provided in the figure caption.
The coliphage §X174 exhibited much greater
transport potential (87.1%) than the bacte-
ria Enterococcus (0.6%) or E. coli (0.3%).
This is likely due to differences in the size
and surface chemistry of these microorgan-
isms. In particular, §X174 is much smaller
than Enterococcus or E. coli (0.5-2 microns).
Straining is therefore expected to play a more
dominant role in retention of the bacterial
-------
1E-001
(a)
0.0
—•— Enterococcus
— »— Somatic Coliphage
•-•¥-- Total E.coli
1.5 2.0 2.5
Pore Volume
Enterococcus
Somatic Coliphage
Total E. col
(b)
1.00
S C,-< (cms g-<)
Figure 4.4. Breakthrough curves (Figure 4.4a) and retention profiles (Figure 4.4b) for Enterococcus,
total E. co/;, and §X174 in a column experiment packed with sterilized field soil. In Figure 4.4a log relative
concentrations (C/C.; where C is the aqueous concentration and C. is the initial concentration in the influent
suspension) are plotted as a function of pore volumes. In Figure 4.4b log relative concentrations in soil,
(S/C), is plotted as a function of distance away from the inlet. Relevant experimental conditions include
the following parameters: porosity=0.50, column length=14.4 cm, tracer pulse duration=112.5 min, and the
Darcy water velocity=3.36 cm h~1.
cells than the coliphage based on size con-
siderations (Bradford et al., 2003). It is
interesting to note that the breakthrough curve
for both Enterococcus and E. coll exhibited a
similar shape and magnitude in the relative
concentrations, with initially low effluent con-
centrations that slowly increased with time.
This observation suggests the potential for
low but persistent amounts of bacteria trans-
port that were likely due to slow release of
cells from the solid phase, i.e., detachment,
hydrodynamic shearing, or diffusion from low
velocity regions (Bradford and Torkzaban,
2008).
After recovery of the breakthrough curve,
the soil in the column was excavated to
recover the remaining cells in the column.
Figure 4.4b presents the microorganism
retention profiles for Enterococcus, total
E. coll, and $X174 in the packed soil column.
Consistent with the breakthrough curve
information the total amounts of microbes
recovered in the soil were 7.8, 83, and 109%
for (|)XI 74, E. coll, and Enterococcus, respec-
tively. Reasonable mass balance was achieved
in this packed column experiment (83 to
109%). Batch results shown in Figure 4.3a
indicate that little death or growth of the
indicator microorganisms was likely to occur
during the column experiments (<6.25 h).
Hence, the mass balance information supports
our methodology for determining microorgan-
ism concentrations in the field. It should also
be mentioned that the collected retention
profiles for the microorganisms were not log
- linear with depth; i.e., the rate of retention
was not first-order. The observed nonmono-
tonic shape of the retention profiles for the
bacteria cells likely reflected the release and
continued slow migration of cells through
the soil. Other researchers have reported
nonmonotonic profiles for bacteria in porous
media and this shape has been reported to
be sensitive to the soil grain size, the system
hydrodynamics, and the sampling time (Tong
et al., 2005; Bradford et al., 2006b). In
contrast, the retention profile for $X174 shows
enhanced retention near the column inlet that
decreased with distance. In the literature this
type of depth dependent retention has been
attributed to straining (Bradford et al., 2006a)
as well as chemical heterogeneity of the
microorganisms (Li et al., 2004; Tufenkji and
Elimelech, 2005).
Collectively, Figures 4.4a and 4.4b dem-
onstrate the complexity of microorganism
-------
transport and retention behavior in soil from
the NMP site. In this work we do not attempt
to quantify and simulate the exact retention
mechanisms, as this would require additional
experiments and model development that
are beyond the scope of this manuscript. In
addition, application of laboratory reten-
tion information collected under saturated
conditions to predict transport of the microor-
ganisms at the NMP site is complicated by the
sensitivity of microorganism retention to water
content, transients, and fluid velocity. The
interested reader is referred to the following
literature that discusses current challenges in
modeling microorganism retention in porous
media (Bradford and Toride, 2007; Bradford
and Torkzaban, 2008).
Wastewater application under ponded infil-
tration conditions represents a worst case
scenario for microorganism transport,
because more conductive pore spaces and
soil structure are water filled under satu-
rated conditions. An additional transport
experiment was therefore conducted on an
undisturbed column from the NMP site (core
length was 65 cm and the internal diameter
was 24 cm), with wastewater instantaneously
added to the soil surface to a depth of 7 cm.
Figure 4.5 presents plots of the normalized
concentration of the indicator microorganisms
in soil as a function of depth at a sampling
time of 24 h. In this case, the infiltrate
front has already passed through the column
and water is slowing redistributing (data not
shown). High concentrations of the indicator
microorganisms occur near the soil surface
and then tend to rapidly decrease with depth.
This mainly reflects microbe retention behav-
ior. Figure 4.5 also indicates the presence
of isolated low concentrations of bacterial
indicators at several deeper depths (27.94
and 53.34 cm). The concentrations at these
locations likely reflect the influence of soil
structure or low levels of bacteria release
and migration as previously identified in
Figure 4.4.
E
.<£.
I
• Enterococcus
— *— Fecal Conform
-•*-- Somatic Coliphage
...... Total E.coli
5.0E-003 1.0E-002 1.5E-002 2.0E-002
SCT.-'(cm'g-1)
Figure 4.5. Plots of the normalized concentra-
tion of the indicator microorganisms in soil (S/CTi)
as a function of depth at a sampling time of 24 h
after ponded infiltration of wastewater ceased on
the undisturbed soil core from the NMP site. Here
the concentration in soil is normalized by the total
concentration of a given microorganism measured
in the entire core at time = 0 (CTi),
It should be mentioned that analysis of the
column effluent and soil at the column outlet
from the undisturbed core did not reveal any
indicator microorganisms for all core sam-
pling times (24, 48, 96, and 176 h). Soil
cores taken at later sampling times primarily
reflected the influence of indicator micro-
organism survival and may be analyzed in a
similar fashion to the batch survival experi-
ments by plotting the log of C/C^ for the
entire core as a function of time in Figure 4.6.
Survival information obtained from batch
experiments under native conditions and 80%
water saturation were generally consistent with
results shown in Figure 4.6. After 176 h the
batch and undisturbed column values of C^/CTi
were 0.003 and 0.006 for Enterococcus,
and 0.003 and 0.005 for somatic coliphage,
respectively. The total concentrations of fecal
coliform and total E. coll in the undisturbed
column could only be detected up to 96 and
48 h, respectively, but were similar in mag-
nitude to native batch experiments at these
times (compare with Figure 4.3b).
-------
0.001 L
0
—•— Enterococcus
-«•— Fecal Conform
-<*-<• Somatic Coliphage
:>... Total E. coli
75 100
Tims (h)
Figure 4.6. A semi-log plot of the relative total
concentration (CT/CTi) of the various indicator mi-
croorganisms as a function of time in the ponded
infiltration undisturbed soil column experiment.
Here the total concentration in soil is normalized
by the total concentration of a given microorganism
measured in the entire core at time = 0 (CTi),
-------
5.0
of
Background
Recent experimental and theoretical research
has indicated that the pore structure and
the solid-water-air triple point can play an
important role in colloid and microorganism
retention under unfavorable attachment condi-
tions (Gushing and Lawler, 1998; Bradford et
al., 2002, 2003, 2004, 2005, 2006abcd; Li
et al., 2004, 2006ab; Tufenkji et al., 2004;
Bradford and Bettahar, 2005 and 2006;
Foppen et al., 2005). Pore spaces occurring
at grain contacts and triple points provide
optimum locations for colloids that are weakly
associated with the solid water interface (SWI)
or the air-water interface (AWI) to be retained
because of reduced hydrodynamic forces, size
limitations, and adhesive interactions from
multiple interfaces (Hoek and Agarwal, 2006).
Enhanced retention of colloids in the smallest
regions of the soil pore space formed adjacent
to grain-grain contacts and the triple point
has been referred to as straining (Gushing
and Lawler, 1998; Bradford and Torkzaban,
2008).
Most published research on straining has
focused on the role of physical factors such as
the relative size of the colloid to the median
grain diameter, and little attention has been
given to the potential interrelated influence
of solution chemistry and hydrodynamics on
straining. Attachment under unfavorable
conditions is known to be highly dependent
on solution chemistry (e.g., Li et al., 2004;
Tufenkji and Elimelech, 2004 and 2005a),
and system hydrodynamics (Wang et al.,
1981; Tan et al., 1994; Kretzschmar et al.,
1997; Compere et al,, 2001; Li et al., 2005).
Under unfavorable attachment conditions,
hydrodynamic forces may be sufficient to over-
come weak adhesive interactions (Bergandahl
and Grasso, 1999; Torkzaban et al., 2007).
In this case, attached colloids may be lifted
from the solid surface and detach, or they
may roll, skip or slide down gradient on the
SWI and/or AWI to locations where the hydro-
dynamic shear is less significant. It is logical
to anticipate that some of these mobilized
colloids will be retained in small pore spaces
formed at grain-grain contacts and the triple
point. One can therefore expect that solution
chemistry and hydrodynamic forces will play
an important interrelated role in straining.
The objective of this research is to inves-
tigate the role of solution chemistry and
system hydrodynamics on colloid transport
and straining. Negatively charged latex
microspheres and quartz sands were used in
batch experiments and packed column stud-
ies that encompassed a range of solution
ionic strength, Darcy water velocity, grain
sizes, and water saturation. All experiments
were conducted using electrolyte solution
buffered to a pH of 10 to ensure highly unfa-
vorable attachment conditions. Data analysis
and interpretation was aided through inter-
action energy calculations, mass balance
computation, mathematical modeling, and
experimental determination of breakthrough
curves and deposition profiles. Details on the
saturated colloid transport experiments and
results are given in Bradford et al. (2007),
whereas Torkzaban et al. (2008a) discusses
findings from the unsaturated colloid transport
experiments. Only an abbreviated discussion
is provided below.
and
Colloids
Yellow-green fluorescent latex microspheres
(Molecular Probes, Eugene, OR) were used as
model colloid particles in the experimental
studies (excitation at 505 nm, and emis-
sion at 515 nm). Two sizes of microspheres
were used in the transport experiments,
-------
1.1 and 3.0 urn. The uniformity of the col-
loid size was verified using a Horiba LA 930
(Horiba Instruments Inc., Irvine, CA 92614)
laser scattering particle size and distribution
analyzer and by inspections of suspensions
under an epi-fluorescent microscope. The
microspheres had carboxyl surface functional
groups, a density of 1.055 g crrr3, and are
reported as hydrophilic by the manufacturer.
The initial influent concentration (C.) for the
1.1 and 3.0 urn colloids for the experiments
was 2.7xl010 and l.SxlO9 Nc L1 (where Nc
denotes number of colloids), respectively.
Several experiments with the 1.1 urn colloids
were also conducted at a lower initial concen-
tration of C.=6.8xl08 NCU.
Aquifer material used for the column experi-
ments consisted of various sieve sizes of
Ottawa sand (U.S. Silica, Ottawa, IL). The
porous media were selected to encompass a
range in grain sizes, and are designated by
their median grain size (dv/) as: 360, 240,
and 150 urn. Specific properties of the 360,
240, and 150 urn sands include: the coef-
ficient of uniformity (d60/dw; here 10 and
60% of the sand mass is finer than dlfj and
d60, respectively) of 1.88, 3.06, and 2.25;
and intrinsic permeabilities of 6.37xlO~n,
1.12x10-", and 4.68xlQ-12 m2, respectively.
Ottawa sands typically consist of 99.8% Si02
(quartz) and trace amounts of metal oxides,
have spheroidal shapes, and contain relatively
rough surfaces. An estimate of the pore-size
distribution for these sands can be obtained
by using Laplace's equation of capillarity
and measured capillary pressure - saturation
curves presented by Bradford and Abriola
(2001).
Electrolyte Solution Chemistry
The background electrolyte solutions uti-
lized in the column studies consisted of
deionized water with a buffered pH of 10
achieved with 1.7 mM NaHC03 and 1.7 mM
Na2C03 (Cherrey et al., 2003). This solution
chemistry was chosen to create a stabilized
mono-dispersed suspension with the selected
colloids. In particular, at a pH of 10 quartz
and iron oxides possess a net negative charge
(Tipping, 1981; Redman etal., 2004),
and any attractive electrostatic interactions
between the colloids and porous media is
expected to be minimized at this pH. The
ionic strength (IS) of the eluant, resident, and
tracer solutions in the transport experiments
was varied between 6-106 mM using this
same pH 10 solution by changing the amounts
of added NaCI, KCI, or NaBr. Unless specifi-
cally indicated, the concentrations of resident,
tracer, and eluant solutions were maintained
over the course of an experiment.
DLVO Calculations
Derjaguin, Landau, Verwey and Overbeek
(DLVO) theory (Derjaguin and Landau, 1941;
Verwey and Overbeek, 1948) was used to
calculate the total interaction energy (sum
of London-van der Waals attraction and elec-
trostatic double-layer repulsion) for our 1.1
and 3 urn colloids upon close approach to
quartz surfaces (assuming sphere-plate inter-
actions) for the various solution chemistries
(pH=10, and IS=6-106 mM). In these cal-
culations, constant-potential electrostatic
double layer interactions were quantified using
the expression of Hogg et al. (1966) and
zeta potentials in place of surface potentials.
Retarded London-van der Waals attractive
interaction force was determined from the
expression of Gregory (1981) utilizing a value
of 4,Q4xlQ~21 J for the Hamaker constant
(Bergendahl and Grasso, 1999) to represent
our polystyrene latex-water-quartz system.
The zeta potential of the quartz at pH 10 and
for IS ranging from 6 to 106 mM was esti-
mated using results presented in Elimelech et
al. (2000) and Redman et al. (2004), Zeta
potentials for our 1.1 and 3 urn colloids in
the various solution chemistries that were
used in the DLVO calculations were calculated
from experimentally measured electropho-
retic mobilities using a ZetaPals instrument
(Brookhaven Instruments Corporation,
Holtsville, NY).
-------
Batch Experiments
Batch experiments were conducted by plac-
ing 10 g of sand and 10 ml of a known initial
concentration of colloid suspension into a
polypropylene centrifuge tubes with the tem-
perature kept at approximately 20° C. Three
different ionic strength solutions (6, 30,
60 mM) buffered at pH 10 were used for
making the colloid suspension. The suspen-
sion and sand were allowed to equilibrate for
2 h by gently rotating the tubes end over end
(15 rpm) on a tube rotator (Fisher Scientific,
San Diego, CA), The 2-h equilibration time
was chosen to mimic the duration of the
column experiments. A control experiment
without colloids was also run for measur-
ing the background concentration of colloids
introduced from the sand. The initial and final
concentrations of colloids in the suspension
were determining using a spectrophotometer
(Perkin Elmer LC95 UV/VIS spectrometer,
Irvine, CA) after setting the tube to rest for a
few minutes. All experiments were performed
in duplicate.
Column
Procedures and protocols for the saturated
packed column glass (15 cm long and 4.8 cm
inside diameter, equipped with an adjustable
adapter at the top) experiments are reported
in detail by Bradford et al. (2002 and 2007),
whereas details on the unsaturated column
(10 cm long and 5 cm inside diameter)
experiments are provided in Torkzaban et
al. (2008a). The columns were wet packed
(water level kept above the sand surface) with
the various sands. For saturated systems,
the colloid suspension was pumped upward
through the vertically oriented columns at a
steady flow rate, after which a three-way valve
was used to switch to the background solution.
The unsaturated column experiment was
designed to accurately establish steady state
flow and uniform saturation conditions. In
this case, the sand in the column was drained
to the desired water saturation level by reduc-
ing the inflow water rate to the hydraulic
conductivity corresponding to that saturation.
Simultaneously, the pressure head at the
bottom of the column was gradually reduced
until unit hydraulic gradient conditions were
achieved (the only driving force for water flow
was gravity), which implies a constant capil-
lary pressure and saturation along the column.
After establishing a specified saturation and
water chemistry, colloid transport experi-
ments were initiated by introducing a colloid
suspension and then rinsing with background
electrolyte.
Effluent samples from the column experi-
ments were collected and analyzed for colloid
concentration using a Turner Quantech
Fluorometer (Barnstead/Thermolyne, Dubuque,
IA) or a spectrometer (Perkin Elmer LC95
UV/VIS spectrometer, Irvine, CA). The average
of three measurements were used to determine
each colloid concentration (reproducibility was
typically within 1% of C.).
Following completion of the colloid trans-
port experiments, the spatial distribution of
retained colloids in each packed column was
determined. The sand was carefully excavated
into tubes containing excess eluant solution
of the same IS and pH that was used in the
transport experiment. The tubes were slowly
shaken to liberate reversibly retained colloids.
The concentration of the colloids in the excess
aqueous solution was then measured with the
fluorometer. The volume of solution and mass
of sand in each tube was determined from
mass balance.
A colloid mass balance was conducted at the
end of each column experiment using effluent
concentration data and the final spatial distri-
bution of retained colloids in the sand. The
calculated number of colloids in the effluent
and retained in the sand was normalized by
the total number of injected particles into a
column.
Colloid Transport Model for Safe/rated and
Unsaturated Systems
The HYDRUS-1D code (Simunek et al., 2009)
was used to simulate colloid transport and
retention in the column experiments. Details
-------
are given by Torkzaban et al. (2008a). In
brief, the model includes provisions for trans-
port by advection and dispersion, and two-site
kinetic colloid retention/release with various
time and depth dependent blocking functions.
Colloid retention and release parameters for
given conditions (e.g., ionic strength, water
velocity and saturation, porous medium size,
and colloid size) were obtained by nonlinear
least squares optimization to the experimental
data.
Results and Discussion
Details on the experimental results are
presented in Bradford et al. (2007) and
Torkzaban et al. (2008a). In this section we
focus our discussion on the interpretation and
implications of these findings.
Experimental evidence in this study demon-
strated that attachment to the SWI was not
the dominant mechanism responsible for
colloid retention. The first piece of indirect
evidence was from the batch experiments
in which no significant colloid attachment
to the sand was observed. Additionally, the
experimental protocol for determining the
deposition profiles was based upon the rapid
release of colloids into solutions, and high
recoveries were not consistent with the low
detachment rates that were observed in the
column experiments. If the colloids were
irreversibly attached to the SWI, they should
have remained on the grain surface even after
suspending the sand in the same solution that
was used for the transport experiments. The
next piece of evidence is the fact that col-
loid retention occurred in the column, even
under conditions when DLVO calculations
indicate the presence of a substantial energy
barrier to colloid - SWI interaction. Finally,
the colloid deposition profiles resulting from
the column experiments were not consistent
with first-order attachment-detachment model
predictions (e.g., non-exponential).
Non-exponential deposition profiles in
saturated conditions have been attributed
to charge variability of the porous media
(Johnson and Elimelech, 1995), heterogene-
ity in colloid surface charge characteristics
(Bolster et al., 1999; Li et al., 2004), and
deposition of colloids in a secondary energy
minimum (Redman et al., 2004; Tufenkji and
Elimelech, 2005b). These factors may have
been involved; however, these mechanisms
cannot fully explain the experimental data.
The experiments were conducted in a solution
at a pH 10, which should minimize any effect
of charge heterogeneity on the sand surface
because the isoelectric points of most metal
oxides fall below this pH (Elimelech et al.,
2000). The results of our batch experiments
and DLVO calculations do not support the
existence of any significant heterogeneity on
the grain or colloid surfaces. Furthermore, the
colloids may be held in the secondary minima;
however, this cannot explain the observed
depth and sand size dependent deposition.
Therefore, another retention mechanism must
also be involved.
Retention of colloids under unsaturated condi-
tions has also been attributed to partitioning
at the AWI and/or film straining (e.g. Wan
and Wilson 1994a; Wan and Tokunaga 1997;
Saiers and Lenhart, 2003a; Lenhart and
Saiers, 2002). However, attachment to the
AWI is not believed to play a significant role in
the reported unsaturated experiments, as both
the AWI and colloids were negatively charged
even in the highest ionic strength solution. It
has been well documented that hydrophilic-
negatively charged colloids are unlikely to
attach to the negatively charged AWI (e.g. Wan
and Tokunaga 2002; Crist et al. 2004, 2005;
Torkzaban et al. 2006a, 2006W.
It is also unlikely that film straining is playing
a major role in the reported colloid reten-
tion as the 1.1 urn latex colloids are too
large to be bound within a thin-film. These
thin water films have calculated thicknesses
around 50 nm during steady-state water flow
- over an order of magnitude thinner than the
diameter of the colloids used in this current
study. Moreover, film straining is theoreti-
cally independent of solution ionic strength
-------
and flow velocity (Wan and Tokunaga, 1997),
In contrast, notably more colloid retention is
observed in higher ionic strength solutions and
for lower water velocities.
Straining of colloids provides a plausible
explanation and mechanism for the observed
colloid deposition behavior. Colloids can
become trapped in small pores where the
flow velocity is reduced. These nearly immo-
bile regions include the small pores formed
by narrow wedges that are defined by grain-
to-grain contacts, dead-end pores, and the
three-phase contact line of the solid-water-air
interface (in case of unsaturated conditions).
It is proposed that colloids are weakly associ-
ated with the SWI interface via the secondary
energy minimum; hence, they are subject
to fluid drag forces that can translate and/or
funnel the colloids to the low velocity regions
of the pore structure or "straining sites".
Earlier studies have observed similar trends by
fluorescent microscopy and x-ray microtomog-
raphy that colloids accumulate in the narrow
region of the pore spaces near the contacts of
irregularly shaped sand grains under unfavor-
able attachment conditions (Bradford et al.,
2005, Li et al., 2006a, and Bradford et al.,
2006bJ. In these studies, pore-space con-
strictions apparently served as locations for
colloid retention by straining, whereas few
colloids appeared to be immobilized far from
the grain-to-grain contacts. Colloid retention
in the pore network is further supported by
Hoek and Agarwal (2006) who reported that
DLVO forces act on colloids up to 5 times
more in small pores than with a single flat sur-
face. Straining, as the mechanism of colloid
retention, is also supported by mass balance
calculations and the fact that once the sand
is resuspended in the solution and the pore
structure eliminated, the colloids return to
solution.
As noted previously, colloids that are weakly
associated with the SWI via the secondary
energy minima experience significant hydro-
dynamic forces (e.g., fluid drag, lift forces)
due to fluid flow. The shear force acting
on the colloid surface held in the second-
ary energy minimum is different than that
acting on the colloid surface in the bulk fluid.
Consequently, at close range to the SWI the
colloids experience a combination of forces
(hydrodynamic forces, electrical double-
layer repulsion, and London-van der Waals)
which creates a torque and hence rotation on
the surface (Bergendahl and Grasso 2000;
Gushing and Lawler 1998). It is therefore
hypothesized that colloids associated with
secondary energy minima can be translated
to regions of low velocity where there are no
significant hydrodynamic forces present and
they can remain attached even by a weak
attraction force. Based on this argument,
more colloids will be retained in a secondary
energy minimum as ionic strength is increased
and these colloids will subsequently be tun-
neled to the straining sites. It is believed that
colloids retained in straining sites can aggre-
gate together depending upon the solution
chemical conditions as it has been observed
by Bradford et al. (2005), Li et al. (2006),
and Bradford et al. (2006b). This hypothesis
can also explain the observed decrease in col-
loid retention with increasing fluid velocity. In
this case, enhanced hydrodynamic forces may
overcome the electrostatic and van der Waals
interactions and detach colloids from the grain
surface. It is also possible that the extent
of stagnation regions decreases as the water
velocity increases. This is consistent with our
observations and modeling results presented
by Gushing and Lawler (1998),
It was anticipated that the magnitude of
straining would be enhanced with decreas-
ing sand size and water saturation level due
to a greater amount of low velocity regions.
Both of these expected trends were verified
by our experimental observations. The extent
of retention of colloids increased with ionic
strength and soil grain size for all of the vari-
ous saturation levels tested (Figure 5.1). With
decreasing soil grain size, the peak effluent
concentration was lower, indicating a greater
-------
removal of colloids. This is especially true
for the higher ionic strength conditions when
very little breakthrough of colloids occurred.
Comparison of the breakthrough curves for
all of the ionic strength conditions reveals
a decrease in peak effluent concentrations
with water content. All these results provide
substantial direct and indirect evidence that
straining is an important removal mechanism
for the CML colloids in the Ottawa sands
investigated.
(a)
Saturation (%)
D 100
O 70
O 50
A 40
o
SJ
U
(b)
Saturation (%)
D 100
O 70
O 50
A 40
4 8
Pore Volume
Figure 5.1. Observed and simulated breakthrough
curves of colloids for various saturation levels in
360 urn sand at an ionic strength of 6 mM (Fig-
ure 5. la) and 30 mM (Figure 5.1b)
The comparison of experimental and mod-
eled results also suggests that straining is the
underlying mechanism in colloid deposition in
both the saturated and unsaturated systems
tested. The concentration of retained colloids
enumerated after the column experiments
was found to exhibit hyper-exponential spatial
distribution through the length of the column.
This indicates the deposition coefficient was
depth-dependent, a trend that has previously
been associated with straining (e.g., Bradford
etal., 2002, 2003, and 2006aJ.
All the observations discussed above provide
convincing evidence that straining was the
primary mechanism of colloid retention in
our column experiments. The straining rate
in a given porous medium is apparently a
complicated mechanism, coupled with such
parameters as pore size distribution, hydrody-
namics, solution chemistry and water content.
Specific findings are highlighted below.
• Straining increases in magnitude with
increasing ionic strength due to an increased
force (secondary energy minimum) and
number of colloids that are tunneled to and
retained in small pores formed adjacent to
grain-grain junctions.
• Straining increases in magnitude with water
content due to an increase in the extent
of stagnant regions of flow within the pore
structure in lower water content.
• Increasing the flow rate of a system tends to
decrease the amount of straining as a result
of the increased fluid drag force that acts on
weakly-attached colloids on the SWI and also
decreased flow stagnation regions.
• The shape of the colloid deposition profile is
highly sensitive to the physical (grain size,
water content, and flow rate) and chemical
(solution IS and pH) properties of a system
due to the interrelation of these parameters
on colloid straining.
Additional research is required to better
understand and quantify the coupling of phys-
ical and chemical processes that influence
colloid straining in saturated and unsaturated
porous media. This information is believed to
be essential for predicting colloid transport
and fate in many natural environments.
-------
6.0
and
Background
Most models for microorganism retention in
porous media have been based on filtration
theory (Yao et al., 1971, Rajagopalan and
Tien, 1976), These models have typically only
considered the average pore-water velocity in
the porous medium and a single first-order
attachment rate coefficient which predicts that
the profile of retained colloids in porous media
will decrease exponentially with distance.
Under unfavorable attachment conditions,
these models have commonly been found to
be inadequate to predict the amount of reten-
tion, the shape of the deposition profile, and
the dependence of colloid retention on veloc-
ity, solution chemistry, and grain and colloid
size (Bradford et al., 2003, 2006a, 2007;
Tufenkji and Elimelech, 2005ab; Tong et al.,
2005; Li and Johnson, 2005; Li et al., 2005;
Johnson et al., 2007; Torkzaban et al., 2007
and 2008a).
Recent research has demonstrated that colloid
and microorganism retention in porous media
does not solely depend on the strength of the
adhesive interaction (Bradford et al., 2007
and 2009; Bradford and Torkzaban, 2008b;
Johnson et al., 2007; Torkzaban et al., 2007
and 2008; Duffadar and Davis, 2008; Shen et
al., 2008; Kim et al., 2009ab). Findings sug-
gest that colloid retention in porous media is a
strongly coupled process that depends on the
chemistry of the aqueous and solid phases,
as well as the pore structure and surface
roughness, the colloid size and concentra-
tion, and water velocity. In particular, it has
been reported that weakly associated colloids
with the solid-water interface via the second-
ary minimum or nanoscale heterogeneity may
experience significant hydrodynamic forces
due to fluid flow that may result in rolling,
sliding, skipping, or detachment of colloids
on/from the collector surface (Bradford and
Torkzaban, 2008b; Torkzaban et al., 2007 and
2008b; Duffadar and Davis, 2008). Some
of these weakly associated colloids can be
translated and/or tunneled by fluid drag force
to low velocity regions in small pore spaces
and "eddy zones" which occur near some
grain-grain contacts and surface roughness
locations where they can be retained (Bradford
and Torkzaban, 2008b; Torkzaban et al., 2007
and 2008b). Indeed, recent experimental
evidence by Kuznar and Elimelech (2007)
demonstrates that weakly interacting colloids
can be translated along the collector surface
via hydrodynamic forces and be retained in
regions near the rear stagnation point.
The above literature indicates that, under
unfavorable attachment conditions, colloid
retention will be enhanced in the low velocity
regions of the porous medium. It is therefore
not surprising that models that consider only
the average pore water velocity and a single
attachment coefficient have been found to be
inadequate to predict colloid retention behav-
ior in many instances. Below, we discuss
alternative model formulations that can be
used to account for different colloid retention
mechanisms in the various regions of the pore
space. Specific model formations that will be
discussed include: a combined physical-chem-
ical nonequilibriurn model (Leij and Bradford,
2009), the dual permeability model (Bradford
et al., 2009), and the stochastic stream tube
model (Bradford and Toride, 2007). A brief
description of the applications, implications,
and limitations of these models to characterize
colloid/microorganism transport and retention
will be discussed. Although specific examples
are provided below for colloid transport in
homogeneous porous media, these same
-------
models are well designed to simulate colloid
transport in the field,
Physical Chemical Nonequilibrium (PCNE)
Model
The two governing equations for transport of
a colloid in a porous medium with mobile and
immobile regions of water are defined as:
dt
dt
-=a(Cm-C,J
[6.2]
where 8 is the volumetric water content, the
subscripts m and im refer to the mobile and
immobile region, C is colloid concentration in
aqueous phase [N/L3], S is solid phase con-
centration of colloid from either the mobile or
immobile region per mass of dry soil [N/M],
pb is dry soil bulk density [M/L3], D is disper-
sion coefficient [L2/T], v is pore-water velocity
[L/T], a is a physical nonequilibrium (PNE)
coefficient for mass transfer between mobile
and immobile region [1/T], z is depth [L], and
t is time [T],
The aqueous phase consists of mobile and
immobile regions. The solid phase of the soil
is either in contact with the mobile or immo-
bile aqueous region, colloid may be retained
or released from the solid phase. Without
considering the actual mechanism for colloid
retention, the following equalities hold:
[6.3a,b,c,d]
emDm=6D
Sm+S,m=S
The chemical nonequilibrium (CNE) compo-
nent is introduced by further distinguishing
between equilibrium (type 1) and kinetic
(type 2) retention:
+ S
m>2
S,m=S,m>1
+S,m>2[6.4a,b]
Colloid retention on type-1 sites is an equilib-
rium process described by:
, S!ml- K!mlC!m
[6.5a,b]
with K as a "distribution" constant expressed
as volume of aqueous phase (either the
"mobile" or "immobile" region) per mass of
dry soil [L3/M], Retention on type-2 sites is
governed by a first-order rate equation:
dt
[6.6a,b]
The retention rate is proportional to the dif-
ference in (eventual) equilibrium and actual
solid phase concentration with p as the pro-
portionality constant [1/T], These coefficients
are likely to be different for retention from the
mobile and immobile regions. For both regions
a ratio of "equilibrium" to "total" sites may be
defined. At equilibrium these follow from the
solid phase concentrations or the distribution
coefficients according to:
/- _
Jm ~
f _
Jim ~
[6.7a,b]
The total concentration, which is given as
mass per volume of bulk soil, is defined as:
+ S
[6.8]
At equilibrium, the total concentration is given
by:
[6.9]
with R = l+pt>K/Q
and K = Km + Kim
where R is a retardation factor.
The system of mathematical equations that
needs to be solved is as follows:
0 'm ,m /Q _
[6. Ida]
[6.10W
-------
~\ o
~ fi
im ~ S; 2]
[6.10c,d]
The problem will be rewritten with only Cm as
dependent variable. The selected mathemati-
cal conditions involve a zero initial condition
and a time-dependent input C(t):
cm(z,o) =
[6.11]
[6.12a,b]
The advection-dispersion equation was
adapted to separately account for physical and
chemical nonequilibrium during transport of
solutes and colloids in porous media. In the
resulting physical-chemical nonequilibrium
(PCNE) model the aqueous phase is parti-
tioned into an immobile and a mobile aqueous
region. Based on equilibrium or nonequilib-
rium interaction of colloids with the solid
phase, four types of solid domains may be
distinguished.
An analytic solution for the PCNE model
described above was obtained by Leij and
Bradford (2009). Expressions for the first
three time moments of the solutions are pre-
sented in Table 6.1. These may be used to
elucidate the impact of transport parameters
on the mean, variance, and skewness of break-
through curves. There are several reasons
to have analytical tools available to quantify
PNCE transport. Analytical methods are
useful to verify numerical methods, elucidate
the role of different model parameters, and to
approximately quantify transport such as for
longer time or spatial scales. The derived ana-
lytical tools for the PCNE model also offer the
flexibility to independently quantify the impact
of a "chemical" and "physical" nonequilib-
rium process on colloid transport.
The sensitivity of the breakthrough curves to
model parameters was illustrated by Leij and
Bradford (2009) for different types of non-
equilibrium using the analytical solution for
the PCNE model. The simplest cases involve
the dependency of the curve on the PNE
parameter 9m/9 and the CNE parameters fm
and f.m in the presence of physical nonequi-
librium. The curves exhibit the characteristic
features of earlier breakthrough and more tail-
ing with increased nonequilibrium. However,
the shape of the curves is not very sensitive
to the parameter values. On the other hand,
the shape of the curve will change with differ-
ent combinations of the PCNE parameters a
and p. The additional parameters in the PCNE
allow greater flexibility to generate different
types of breakthrough curves. The moment
results of Table 6.1 were used to predict
contours of mean, variance, and skewness
of the colloid breakthrough curve as a func-
tion of either log a and log /cma or 9im and log
kma. This information illustrates the utility of
having a model with independent physical and
chemical nonequilibrium terms.
Colloid transport will be affected by physical
nonequilibrium because pores are not (readily)
accessible and by chemical nonequilibrium
due to (different) attachment and detachment
rates. These nonequilibrium phenomena are
intertwined because attachment/detachment
rates depend on the flow regime. The PCNE
model, with its ability to independently model
physical and chemical nonequilibrium, was
therefore applied to four colloid breakthrough
experiments by Bradford et al. (2002). It
should be noted that independently quantify-
ing physical and chemical nonequilibrium
processes will become more useful for
transport in natural porous media where non-
equilibrium phenomena may no longer be
ignored.
Dual-Permeability Model
Dual-permeability models have commonly
been used to study preferential and non-
equilibrium flow and solute transport in
-------
Table 6.1. Expressions for the second- and third-order moments for breakthrough after a Dirac delta input
according to physical nonequilibrium (PNE), chemical nonequilibrium (CNE), and PCNE models. The first-
order moment is equal to Rz/vtor all the models.
Model
Moment
Expression
PNE
9fl2
aOv
CNE
M,
\ r\ , _ r- . 2rv1
z — 2D + vz ) + —±
I v (3v
PCNE
^2 A D A. O
\ • U • i\ • -i U i\ -i
x/m_ /m im,2 . m m,2
a p,f
PNE
f.C& p2 / p
•6Dvz + v2z2)+ ;m ;m —
a6v I v
Q/m^/m
a
CNE
^(12D2 + 6Dvz + v2z2) + -^ 4(2D + vz) + |
PCNE
v
6v a I v
emRm,TR
P™ «
*""*" ' + ^^^1^(20 + vz) + —
Pm lv2
structured soils and fractured rocks (Simunek
et al., 2003; Gerke, 2006; Simunek and
van Genuchten, 2008). In this case, the
dual-permeability model partitions the pore
space into two regions that have fast (frac-
ture) and slow (matrix) rates of advective and
dispersive transport of solutes. In contrast to
previous work, Bradford et al. (2009) used the
dual-perrneability model described herein to
simulate the different colloid retention mecha-
nisms that occur in fast (larger pore spaces
- region 1) and slow (small pore spaces, dead
end pores, and grain-grain contact points-
region 2) velocity regions of homogeneous
porous media. The dual-permeability model
has not been previously used to gain insight
into enhanced colloid retention processes in
low velocity regions of homogeneous porous
media. The approach is somewhat analogous
to multiphase flow and transport models that
partition the pore space to regions accessible
for the wetting (small pore spaces) and non-
wetting (large pore spaces) phases. The small
pore spaces in the dual-permeability model
are assumed to maintain continuity by slow
flow adjacent to the solid phase, in crevice
sites near grain-grain contacts, and in small
pores in the same way as the wetting phase in
multiphase systems.
-------
As indicated in the Background section, col-
loids that collide with solid surfaces in fast
regions of the pore space experience differ-
ent hydrodynamic forces than colloids in slow
regions. The higher hydrodynamic forces in
region 1 act to remove colloids from the solid
surface, thus causing region 1 to be associ-
ated with lower rates of colloid retention.
Colloid exchange in the aqueous phase may
occur to and from "slow" water (region 2). In
addition, in this work we also consider the
potential for colloid exchange on the solid
phase from fast to slow regions due to either
rolling or sliding of colloids on the solid sur-
face. The governing equations for water flow
in the dual-permeability model are well known
and are available in the literature (Gerke and
van Genuchten, 1993a,b; Simunek and van
Genuchten, 2008). The corresponding one-
dimensional dual-permeability equations for
local scale colloid transport and retention are
as follows:
a(ewlc,)
r.
l-w
a,
^l-p/ArtlS,-
dt
[6.14]
[6.15]
[6.16]
w
Ls = W^-W)Vm^2-^) [6J7]
where z [L] is distance, t [T] is time, C, and
C2 [Nc L3; Nc denotes the number of colloids]
are the liquid phase concentrations of col-
loids in regions 1 and 2, sl and s2 [Nc M -1] are
the solid phase concentrations of colloids in
regions 1 and 2, 6w] and 8w2 are the volumetric
water contents in regions 1 and 2 [-], pbl and
pb2 are the bulk densities in regions 1 and 2
[M L3], kj and k2 [T1] are the first order colloid
retention rate coefficients in regions 1 and 2,
kdetl and kdet2 [T1] are the first order detach-
ment coefficients in regions 1 and 2, Jj and J2
[Nr L'2 T1] are the total solute fluxes (sum of
the advective and dispersive flux) for colloids
in regions 1 and 2, to [T1] is a coefficient for
colloid exchange between liquids in regions 1
to 2, k12 [T1] is a coefficient for transfer of col-
loids from solid phase region 1 to 2, and w is
the ratio of the volume of region 2 to the total
volume (volume of region 1 + volume of region
2). The term Fs [N L3 T1] accounts for aque-
ous phase mass exchange of colloids between
regions 1 and 2.
Equations [6.13H6.17] were written in terms
of local scale mass balances of regions 1 and
2. To formulate the equations in terms of the
total pore space, the mass balance equations
for regions 1 and 2 need to be multiplied by
(l-w) and w, respectively. The relationship
between several variables at the local scale
and total pore space is provided below. Under
steady-state flow conditions, the total water
flux (qt; LT1) is defined as:
[6.18]
where ql (LT1) and q2 (LT1) are the local
scale water fluxes in regions 1 and 2, respec-
tively. Expressions for the total water content
and bulk density are written in an analogous
manner as Equation [6.18], The total flux
concentration of colloids (Ct, Nc L3) is given as
(Simunek and van Genuchten, 2008):
wq2C2+(l-w)qlCl
wq2+(l-w)q1
[6.19]
and the total solid phase colloid concentration
(st, Nc M-1) as:
Wpb2S2+(l-W)pb,Sl
wpb2+(l-w)pbl
[6.20]
The dual-permeability model outlined above
has been implemented into the HYDRUS-1D
computer model (Simunek et al., 2007; and
Simunek and van Genuchten, 2008). The
-------
code employs the Galerkin-type linear finite
element method for spatial discretization of
the governing differential equations, and a
finite difference method to approximate tem-
poral derivatives. A Crank-Nicholson finite
difference scheme was used to solve the
outlined transport equations sequentially (with
equations for region 1 solved first). Complete
details about the numerical techniques are
provided in the HYDRUS-1D technical manual
(Simunek et al., 2007). For the simulations
discussed below, a third-type boundary condi-
tion was used at the inlet, and a concentration
gradient of zero was fixed at z equal to the
outlet depth. The initial concentration of the
simulation domain was zero.
Bradford et al. (2009) performed a sensitiv-
ity analysis with the dual-permeability model
parameters. Simulation results indicated
that low amounts of advective transport to
low velocity regions of the pore space had a
dramatic effect on the shape of the retention
profile, especially near the inlet boundary
(Figure 6.1). The total water flow rate was
also shown to have a significant influence on
the shape of the colloid retention profile near
the inlet boundary (Figure 6.2). In this case,
higher water velocities were found to produce
less colloid retention near the inlet boundary
because greater amounts of colloids bypassed
the low velocity regions. Both of these pre-
dictions are consistent with experimental
observations that have been reported in the lit-
erature. Away from the inlet boundary, colloid
retention was controlled by the deposition rate
in the higher velocity region, and the aque-
ous and solid phase exchange rates. Example
figures are shown below.
Published colloid transport and retention
data that were obtained for unfavorable
attachment conditions for a range of colloid
and sand sizes, water velocities, and solu-
tion chemistries were described using the
dual-permeability model. Fitted model param-
eters exhibited systematic trends with grain
size, velocity, and solution chemistry. The
dual-permeability model provided a plausible
interpretation for the experimental observa-
tions, and a reasonable approximation of the
pore-scale physics controlling colloid retention
under unfavorable attachment conditions.
Stochastic Stream Tube Model
The CXTFIT program (Toride et al., 1995)
served as the foundation for our stochastic
modeling effort. This code includes the analyt-
ical solution for the one-dimensional advective
dispersion equation with one-site kinetic
chemical nonequilibrium deposition subject to
various initial and boundary conditions. This
model formulation is equivalent to the well-
known first-order attachment and detachment
model that is commonly employed to describe
colloid transport and deposition (e.g., Harvey
and Garabedian, 1991; Corapcioglu and Choi,
wq2=0
wq2=0.001
wq2=0.002
wq2=0.004
wq =0.008
(a)
2 4
Pore Volume
wq2=0.0
wq =0.001
wq =0.002
wq2=0.004
. wq =0.008
(b)
Figure 6.1. Plots of simulated breakthrough
curves (Figure 6. la) and retention profiles (Figure
6.1b) when qt was equal to 0.0918 cm mm1 and
the value of wq2 was 0, 0.001, 0.002, 0.004, and
0.008 cm min1.
-------
//
I/
r
i/
q=0.024
q=0.046
q -0 091
\ qj=0.181
qt=0.361
(a)
\
1
^
234
Pore Volume
q =0.024
qt=0.046
q ,=0.091
qt=0.181
q =0.361
Figure 6.2. Plots of the simulated breakthrough
curves (Figure 6.2a) and retention profiles (Fig-
ure 6.2b) when wq2 was 0.001 cm min^and qt was
equal to 0.024, 0.046, 0.091, 0.181, and 0.361
cm mm1. Other model parameters are provided in
the manuscript text.
1996; Bolster et al., 1999; Schijven and
Hassanizadeh, 2000). This analytical solution
was used in conjunction with the stochastic
stream tube model in CXTFIT to explore col-
loid transport and deposition. Details are
provided in Bradford and Toride (2007).
The value of first-order deposition coefficient
(kd) is typically assumed to be constant in
colloid transport models. In the stochas-
tic modeling approach, parameters may be
defined by probability density functions. If kd
is considered to be stochastic, we assume a
log-normal probability density function (PDF)
that is defined as:
where ad is the standard deviation of the
log-normal probability density function, and
Yd is the normalized log-transformed variable
defined as:
[6.22]
Here ud is the mean value of the log-normal
probability density function defined as
ud = ln( < kd> )-0.5ad2, where < kd > is the
ensemble average of kd. The subscript d is
used on ad, Yd, and ud to identify parameters
associated with the deposition coefficient.
Subscripts rand i/are used in a similar fash-
ion to identify parameters associated with the
release coefficient and the pore water velocity,
respectively.
If both kd and kr are assumed to be log-normal
stochastic parameters that are correlated, then
a joint probability density function is defined
as:
1 f Yr2-2pdrYrYd+Yd2
= exp
2(1-
[6.23]
The parameter pdr is the correlation coefficient
between Yd and Yr and is defined as:
P* = (Y
-------
concentrations determined from the analytical
solution of the advective dispersion equation
with one-site kinetic chemical nonequilibrium.
The variance in solid phase colloid concen-
trations is given as < S(z,t)S(z,t) >-< S(z,t)
> 2 when using the two parameter stochastic
model. Alternatively, if kd and i/are stochastic
and kr is constant, Equation [6.23-6.26] can
be rewritten by replacing kr and v.
Simulation results presented in Bradford
and Toride (2007) indicate that variations
in the deposition coefficient and the average
pore water velocity can both produce hyper-
exponential deposition profiles. Bimodal
formulations for the PDF were also able to
produce hyper-exponential profiles, but with
much lower variances in the deposition coef-
ficient. The shape of the deposition profile was
found to be very sensitive to the correlation
of deposition and release coefficients, and
to the correlation of pore water velocity and
deposition coefficient (example simulations
are shown below in Figure 6.3). Application of
the developed stochastic model to a particular
set of colloid transport and deposition data
indicated that chemical heterogeneity of the
colloid population could not fully explain the
observed behavior. Alternative interpretations
were therefore proposed based on variabil-
ity of the pore size and the water velocity
distributions.
0.8
f" 0.6
_V_
7r
f 0.4
0.2
i
.» 6.0
i
» 6.0
50 100 150
Time (min)
0.1 0.2
Pva—O.S
p.,-0.0
pvd-0.3
0.05 0.1 0.15
Variance S/N. (g-*|
Figure 6.3. (a) Plot of the relative flux concentra-
tion, /(Cj), at a depth of 10 cm as
a function of time when v and kd are both sto-
chastic parameters and values of pvd = -1, -0,5,
0, 0,5, and 1, (b, c) Corresponding normalized
solid phase colloid concentration, < S > /Nic, and
associated variance after 250 min with depth,
respectively. Model parameters that were employed
in these simulations were D = 0,0313 cm2 min'1,
< v > = 0,313 cm min'1, < kd> = 0,03 min'1,
kr = 0.001 min1, vv= 1, and vd= 1.
-------
7.0
Summary and Conclusions
Concentrated animal feeding operations gen-
erate large volumes of manure-contaminated
water that are typically stored in lagoons
before land application. In Chapter 1 we
review the current level of understanding on
the environmental impact and sustainability of
concentrated animal feeding operation (CAFO)
wastewater reuse in agriculture. Specifically,
we address the source, composition, applica-
tion practices, environmental issues, transport
pathways, and potential treatments that are
associated with nutrients and pathogens in
CAFO wastewater.
When applied to land at agronomic rates,
CAFO lagoon water has the potential to be a
valuable fertilizer and soil amendment that
can improve the physical condition of the
soil for plant growth and reduce the demand
for high quality water resources. However,
excess amounts of nutrients, organics with
high biochemical oxygen demand, salts, heavy
metals, pathogenic microorganisms, and phar-
maceutically active compounds (antibiotics
and hormones) in CAFO lagoon water can all
adversely impact soil and water quality, as has
been well documented in the literature. The
current regulatory framework is based on nutri-
ent management plans (NMPs) to ensure that
lagoon water is applied to agricultural lands at
agronomic rates to meet the nutrient demands
of crops. Other lagoon water contaminants are
implicitly assumed to be retained, inactivated,
or degraded in the root zone. Additional
studies are needed to thoroughly test these
hypotheses.
Potential problems and weaknesses associated
with the implementation of NMPs were iden-
tified in Chapter 1. One problem for proper
NMP implementation is due to differences
in the nutrient composition of lagoon water
and the nutrient uptake rates by plants. A
second more critical issue concerns the accu-
rate estimation and delivery of water to meet
plant water demands. Potential errors in water
balance may occur as a result of inaccuracies
in estimates of potential evapotranspiration
(PET), due to not accounting for all sources
of plant available water, spatial variability
of soil hydraulic properties and evapotrans-
piration (ET), nonuniform irrigation water
application, and preferential water flow that
bypasses the root zone. If water in excess of
PET is applied (due to irrigation inefficiency or
unforeseen natural causes) to or preferential
water flow occurs at specific locations on a
NMP site, lagoon water contaminants may also
be transported below the root zone towards
groundwater resources. Of special concern are
contaminants that have limited ability to sorb
to solid surfaces such as nitrate, or contami-
nants that can be associated with the mobile
colloidal fraction of soil solution. Water in
excess of PET will also lead to potential sur-
face water contamination issues, especially for
contaminants that are associated with sedi-
ments in runoff water such as phosphorus. If
water application (from all water sources) is
less than or equal to PET and preferential flow
does not occur, then all lagoon water con-
taminants should remain in the root zone. In
this case, potential environmental problems
may occur due to accumulation of salts and/or
survival of pathogens.
Mitigation of the environmental risks associ-
ated with the reuse of CAFO lagoon water
on agricultural lands may require planning
to prevent contamination, as well as various
lagoon water treatments before use in NMPs.
Prevention measures may include the wise
choice of locations for CAFOs, proper manure
management and modification of animal
diets to minimize specific contaminants (e.g.,
pathogens). Potential treatments to CAFO
-------
lagoon water include: use of solid-liquid sepa-
rators; optimization of the design of lagoon
water storage basins to remove sediment loads
and nutrients (aerobic compared to anaero-
bic); and the adaptation of various municipal
wastewater treatments such as chemical floc-
culants and/or soil treatments.
A nutrient management plan (NMP) was
implemented in a semi-arid environment
on a triticale and sorghum rotation during
2007 and on alfalfa during 2008, where
nitrogen (N) was assumed to be the primary
environmental concern. Chapter 2 discusses
extensive research that was conducted to
characterize flow and transport properties at
this site. Chapters 3 and 4 discuss results
pertaining to nutrients and indicator micro-
organisms, respectively. The NMP field site
was designed to accurately measure the water
and N mass balances in the root zone and
to determine the fate of N, phosphorus (P),
potassium (K) and other salts. Cyclic and
blending dairy wastewater (DWW) applica-
tion strategies, varying in nutrient application
timing, were investigated during 2007. Only
minor differences were found between these
two strategies, therefore the key findings dis-
cussed below were valid for both conditions.
This NMP study was initiated under the condi-
tions of high levels of soil organic N, which
may be representative of historic manure and
DWW application at many sites. Such condi-
tions require that the soil organic N reservoir
be taken into account in the N mass balance
and this may induce difficulties in conserva-
tive NMP implementation that is protective
of the environment due to: i) difficulties in
estimation of mineralization rates and its
spatial variability; ii) delayed availability of
the organic N for plant uptake; iii) DWW can
only be applied to meet a fraction of plant
N uptake; and iv) continuous mineralization
and potential nutrient leaching during fallow
periods.
Based on these insights we implemented a
conservative NMP in 2007 that depleted the
soil organic reservoir. This was achieved by
applying only a fraction of the plant N uptake
with DWW and using DWW that was treated
to remove most of the suspended solids con-
tent (SSC). One advantage of lowering the
SSC of DWW is the ability to irrigate using
water application systems with high water use
efficiency (sprinkler or drip). The outcome
of this transition was an accurate application
of plant available inorganic N that minimized
the migration of nutrients below the root zone.
After the soil organic reservoir is depleted, the
total amount of DWW that can be applied per
crop is expected to increase.
In contrast, a more aggressive NMP (higher
N application rates) implemented in 2008
demonstrated potential problems associated
with the use of treated DWW to grow alfalfa.
Specifically, the soil organic N reservoir
increased and leaching of nitrate below the
root zone (15%) increased during the multi-
cut alfalfa growing season, in spite of the
much larger plant uptake of N by alfalfa than
the cereal rotation in 2007. These observa-
tions were attributed to fixation of atmospheric
N, increased root density, and applying a
higher fraction of plant N uptake with DWW.
The precise water application in conjunction
with the high content of salts in the DWW
caused salt accumulation in the root zone over
time. The measured values represent the salt
load under a conservative NMP approach that
applied only a fraction of the total N that was
required by the plant with DWW, If 100% of
the plant N had been applied by DWW, then
the accompanied salts would increase the
ECw in the root zone to higher levels. High
salt levels in the root zone may restrict plant
growth, and accordingly water and nutrient
uptake. If this reduction is not considered at
NMP sites, additional leaching and contami-
nant migration may occur. An optimum point
likely exists between the benefits of nutrient
application and the detrimental effects of
salt accumulation on crop yield. This point is
strongly dependent on the salt tolerance of the
-------
crop, suggesting that NMPs should use only
salt tolerant crops.
The leaching timing of excess of salts below
the root zone is a crucial aspect in NMP
design. In this study, the soil profile was found
to be depleted from inorganic N at the end of
the 2007 growing seasons, yet mineralization
of organic N was observed to continue during
fallow periods, and high levels of N-(N02+N03)
(nitrogen from the sum of nitrite and nitrate)
were found in the soil profile at the begin-
ning of the sorghum 2007 and alfalfa 2008
seasons. This fallow period poses a greater
contamination risk to ground water due to
planned (pre-irrigation) or natural (precipita-
tion) water application in excess of ET. In
order to minimize nutrient leaching, pre-irri-
gations should be scheduled at the end of the
growing season.
A comprehensive measurement of N mass
balance in the root zone requires information
on both organic N (soil reservoir) and inor-
ganic N (DWW) sources. During the growing
seasons, the major N sink was plant uptake
and losses to the atmosphere during irriga-
tion. Atmospheric losses may be minimized
by applying DWW during times that are associ-
ated with low potential ET (i.e. early morning),
or through drip systems that minimize the
exposure of DWW to the atmosphere.
Differences in the concentration ratios of N,
P, and K between DWW and plant uptake may
lead to accumulation of P and K in the root
zone. However, the conservative NMP imple-
mented in this study applied only a fraction
of the plant N uptake with DWW because it
accounted for the organic N reservoir in the
soil. Similarly, in 2007 the P in DWW was
applied in deficit to plant uptake, and this
resulted in a depletion of soil P. In contrast,
K was applied in excess and needed to be
leached with the salts.
Additional longer term research is needed
on both cereal and legume crops to optimize
NMPs, the soil organic reservoir, and irriga-
tion methods to simultaneously: (i) Prevent
excessive nitrate leaching even after multiple
years of implementation and through crop
rotations; (ii) Maximize the plant uptake of N
in DWW per unit land; (iii) Provide economic
returns on the forage harvest; and (iv) Prevent
excessive salt build up.
NMPs implicitly assume that pathogenic
microorganisms in the lagoon water will be
retained and die-off in the root zone. To test
this assumption with regard to groundwater
protection, the transport and fate of indicator
microorganisms were monitored during winter
and summer growing seasons at the NMP field
site. These results are discussed in detail in
Chapter 4.
In order to efficiently implement the NMP the
dairy wastewater was treated (screen, sedi-
mentation tank, and sand filter) to minimize
the organic load in the wastewater. This treat-
ment also produced close to a 2-log reduction
in concentrations for Enterococcus, fecal
coliform, total E. coli, and somatic coliphage.
When well-water and treated wastewater were
applied to the field site to meet ET and plant
nutrient requirements, little advective trans-
port of the indicator microorganisms occurred
below the root zone (60 cm). The remaining
concentrations of these indicator microor-
ganisms in the root zone died-off during the
winter and summer growing seasons. These
observations support the hypothesis that a
well-designed and implemented NMP at this
site will protect groundwater supplies from
microorganism contamination.
Additional experiments were conducted in the
laboratory to better quantify microorganism
transport and survival in the field soil. Batch
survival experiments revealed much more
rapid die-off rates for the bacterial indicator
microorganisms in native than in sterilized
soil, suggesting that the biotic factors played a
dominant role in survival behavior. Saturated
column experiments with packed field soil,
demonstrated much greater transport potential
for somatic coliphage than bacterial indicators
(Enterococcus and total E. coli}. Retention
-------
rates for the indicator microorganisms were
not log-linear with depth (not first-order with
respect to concentration), demonstrating the
complexity of the microorganism transport
behavior, A worst case transport scenario
of ponded infiltration on a large undistrib-
uted soil column from the field was also
investigated. Concentrations of the indica-
tor microorganisms were not detected in the
column outflow and in the soil at a depth
of 65 cm. The remaining concentrations of
the indicator microorganisms in the column
rapidly decreased toward low concentrations
within 1 week at a temperature of 25°C. Both
of these observations were consistent with
field NMP results and conclusions.
Results from field and laboratory experiments
demonstrate that the fate of microorganisms
at NMP sites will depend on both trans-
port and survival characteristics. Although
transport and survival of microorganisms in
NMP soils were shown to be complicated and
are likely to be site specific, a few recom-
mendations for NMP implementation can
be developed. The transport potential of
microorganisms can be significantly reduced
by minimizing water leaching below the root
zone and surface water runoff. This can be
achieved by: (i) precise estimation of ET
rate; (ii) uniform application of wastewater;
and (iii) selecting water application timing
and quantities based on considerations of
soil permeability and ET. Special caution is
warranted in coarse textured and structured
soils and during water flow transients where
enhanced microorganism transport potential
has been reported in the literature (Natsch et
al., 1996; Bradford et a!., 2003; and Saiers
et al., 2003). Survival characteristics of
microorganisms are also likely to be site spe-
cific, but may be quantified through relatively
simple batch type experiments that mimic
natural conditions. Timing of water appli-
cation should allow for adequate die-off of
microorganisms before leaching the root zone
by irrigation or natural precipitation. Finally,
the potential for groundwater contamination
will increase with shorter travel times and
distances. The water table depth is therefore
another important consideration for environ-
mentally protective NMPs. Implementation
of these NMP recommendations will undoubt-
edly be technically challenging, and may not
be economically feasible in some instances.
Nevertheless, these recommendations provide
guidance to minimize the potential risks of
pathogen contamination of water resources at
NMP sites.
In Chapter 5 we discuss experimental and
theoretical studies that were undertaken to
explore the coupling of physical and chemi-
cal mechanisms of colloid retention under
unfavorable attachment conditions (pH=10).
Negatively charged latex microspheres (1.1
and 3 urn) and quartz sands (360, 240,
and 150 urn) were used in packed column
studies that encompassed a range in sus-
pension ionic strengths (6-106 mM), Darcy
water velocities (0.1-0.45 cm min'1), and
water saturations (40-100%). Derjaguin-
Landau-Verwey-Overbeek (DLVO) calculations,
moment analysis, and batch results suggest
that attachment of colloids to the solid-water
and air-water interfaces was not a significant
mechanism of deposition for the selected
experimental conditions. Breakthrough curves
and hyperexponential deposition profiles were
strongly dependent on the solution chemistry,
the system hydrodynamics, and the colloid
and collector grain size. A mathematical
model, accounting for time- and depth-
dependent straining, produced a reasonably
good fit for both the breakthrough curves and
final deposition profiles. Greater deposition
occurred for increasing ionic strength, lower
flow rates and water saturations, and larger
ratios of the colloid to the median grain diam-
eter. Increasing the solution ionic strength is
believed to increase the adhesive force and
number of colloids in the secondary minimum
of the DLVO interaction energy profile. These
weakly associated colloids can be funneled to
small pore spaces formed adjacent to grain-
grain junctions and the air-water-solid triple
-------
point. For select systems, the ionic strength
of the eluant solution was decreased to 6 mM
following the recovery of the breakthrough
curve. In this case, only a small portion of
the deposited colloids was recovered in the
effluent and the majority was still retained in
the sand. These observations suggest that
the extent of colloid removal by straining is
strongly coupled to solution chemistry.
Experimental research has demonstrated
that much greater microorganism retention is
possible in the smallest regions of the pore
space that are associated with lower flow
rates. We have therefore developed models
that are based upon physical and chemical
nonequilibrium (PCNE), dual permeability,
and stochastic stream tube formulations. This
research is discussed in Chapter 6. Although
specific examples were provided for colloid
transport in homogeneous porous media, these
same models are well designed to simulate
microorganism transport and retention in the
field at NMP sites.
The PCNE model partitions the pore space
into "mobile" and "immobile" flow regions
with first-order mass transfer between these
two regions (i.e, "physical" nonequilibrium or
PNE). Partitioning between the aqueous and
solid phases can either proceed as an equilib-
rium or a first-order process (i.e, "chemical"
nonequilibrium or CNE) for both the mobile
and immobile regions. An analytical solution
for the PCNE model was obtained using iter-
ated Laplace transforms. The PCNE model
allows greater flexibility to describe experimen-
tal breakthrough curves and retention profiles
than the traditional CNE model. Expressions
for moments and transfer functions were
developed to facilitate the analytical use of the
PCNE model.
The dual-permeability model accounts for
different rates of advective and disper-
sive transport, as well as first-order colloid
retention and release in fast and slow veloc-
ity regions of the pore space. The model
also includes provisions for the exchange
of colloids from fast to slow regions in the
aqueous phase and/or on the solid phase. A
sensitivity analysis performed with the dual-
permeability model parameters indicated that
low rates of advective transport to low-velocity
regions had a pronounced influence on colloid
retention profiles, especially near the inlet.
The stochastic model for colloid transport
and retention is based on the conventional
advective dispersion equation that accounts
for first-order kinetic deposition and release
of colloids. One or two stochastic parameters
can be considered in this model, including
the deposition coefficient, the release coef-
ficient, and the average pore water velocity.
In the case of one stochastic parameter, the
probability density function (PDF) is charac-
terized using lognormal, bimodal log-normal,
or a simple two species/region formulation.
When two stochastic parameters are consid-
ered, then a joint log-normal PDF is employed.
Simulation results indicated that variations in
the deposition coefficient and the average pore
water velocity can both produce hyperexponen-
tial deposition profiles. Bimodal formulations
for the PDF were also able to produce hyper-
exponential profiles, but with much lower
variances in the deposition coefficient.
Quality Assurance
This project was designed to evaluate and
select basic options, and/or to perform pre-
liminary assessment of unexplored areas with
regard to NMPs. A quality assurance plan was
followed in all experimental phases of this
study. The data was obtained using standard
physical, chemical, and/or microbiological
laboratory measurements, analytical proce-
dures, bench studies, and field evaluations
outlined in this report. The data presented
in this report met the project's data quality
requirements. Conclusions and recommen-
dations made in this report are supported by
this data, and are scientifically, but not legally
defensible. Caution is warranted when extrap-
olating this information and findings to other
conditions.
-------
8.0
References
Abdul-Jabbar, A.S., Sammis, T.W., and Lugg,
D.G. 1982. Effect of moisture level on
the root pattern of alfalfa. Irrig. Sci,
3:197-207.
Ahammed, M.M. and Chaudhuri, M. 1999. A
low-cost home water filter. J. Water, SRT-
Aqua. 48:263-267.
Allen, R.G., Pereira, L.S., Raes, D., and
Smith, M. 1998. Crop evapotranspiration-
guidelines for computing crop water
requirements. FAO - Food and Agriculture
Organization of the United Nations,
irrigation and drainage paper 56.
Arai, Y. and Sparks, D.L. 2007. Phosphate
reaction dynamics in soils and soil compo-
nents: a multiscale approach. Adv. Agron.
94:135-179.
ASAE. 1999. Manure production and charac-
teristics. AS Data: AS D384.1. St. Joseph,
Ml. American Society of Agricultural
Engineers,
Ayers, R.S. and Westcot, D.W. 1989. Water
quality for agriculture, FAO Irrigation and
Drainage Paper 29, Rev. 1, FAO, Rome,
Italy,
Bales, R.C., Li, S., Maguire, K.M., Yahya,
M.T., and Gerba, C.P. 1993. MS-2 and
poliovirus transport in porous media:
hydrophobic effects and chemical pertur-
bations. Water Resour. Res. 29:957-963.
Barker, J.C. and Zublena, J.P. 1996. Livestock
manure nutrient assessment in North
Carolina, in: Proc. 7th International
Symposium on Agriculture and Food
Processing Wastes, ASAE. pp. 98-106.
Bar-Tal, A., Yermiyahu, U., Beraud, J., Keinan,
M., Rosenberg, R., Zohar, D., Rosen, V.,
and Fine, P. 2004. Nitrogen, phosphorus,
and potassium uptake by wheat and their
distribution in soil following successive,
annual compost applications. J. Environ,
Qua/, 33:1855-1865.
Belanger, G., and Richard, J.E. 2000.
Dynamics of biomass and N accumulation
of alfalfa under three N fertilization rates.
Plant and Soil. 219:177-185
Benner, S.G., Blowes, D.W., Gould, W.D.,
Herbert Jr., R.B., and Ptacek, C.J. 1999.
Geochemistry of a permeable reactive
barrier for rnetals and acid mine drainage.
Environ. Sci. Technol. 33:2793-2799.
Bergendahl, J., and Grasso, D. 1999.
Prediction of colloid detachment in a
model porous media: thermodynamics.
AiChEJ. 45:475-484.
Bergendahl, J., and Grasso, D. 2000.
Prediction of colloid detachment in a
model porous media: hydrodynamics.
Chem. Eng, Sci. 3:1523-1532.
Black, A.S., Sherlock, R.R., Smith, N.P.,
Cameron, K.C., and Goh, K.M. 1985.
Effect of form of nitrogen, season and urea
application rate on ammonia volatilization
from pastures. New Zealand J, Agric. Res,
28:469-474.
Blumenthal, J.M., and Russelle, M.P. 1996.
Subsoil nitrate uptake and symbiotic
dinitrogen fixation by alfalfa. Agron. J,
88:909-915.
Bodvarsson, G.S., Ho, C.K., and Robinson,
B.A. 2003. Yucca Mountain Project,
special issue. J. Contam. Hydrol.
62-63:1-750.
Bohn, H., McNeal, B., and O'Connor, G.
1985. Soil chemistry. 2nd Ed. Wiley, New
York, NY.
Bolster, C.H., Mills, A.L., Hornberger, G.M.,
and Herman, J.S. 1999. Spatial distri-
bution of deposited bacteria following
miscible displacement experiments
in intact cores. Water Resour. Res,
35:1797-1807.
Bouwer, H. 2002. Integrated water man-
agement for the 21st century: problems
-------
and solutions, J. Irrig, Drain, Engrg,
128:193-202.
Bradford, S.A., and Abriola, L.M. 2001.
Dissolution of residual tetrachloroethylene
in fractional wettability porous media:
incorporation of interfacial area estimates.
Water Resotir. Res. 37:1183-1195.
Bradford, S.A. and Schijven, J. 2002.
Release of Cryptosporidium and Giardia
from dairy calf manure: impact of solu-
tion salinity. Environ. Sci. Technoi.
36:3916-3923.
Bradfordj S.A., Yates, S.R., Bettahar, M.,
and Simunek, J. 2002. Physical fac-
tors affecting the transport and fate
of colloids in saturated porous media.
Water Resour. Res., 38(12), 1327,
doi:10.1029/2002WR001340.
Bradford, S.A., Bettahar, M., Simunek, J.,
van Genuchten, M.T. 2004. Straining and
attachment of colloids in physically het-
erogeneous porous media. VadoseZoneJ,
3:384-394.
Bradford, S.A., Simunek, J., Bettahar,
M., Tadassa, Y.F., van Genuchten,
M.T., and Yates, S.R. 2005. Straining
of colloids at textural interfaces.
Water Resour. Res. 4LW10404,
doi:10.1029/2004WR003675.
Bradford, S.A., Simunek, J., Bettahar, M., van
Genuchten, M.T., and Yates, S.R. 2003.
Modeling colloid attachment, straining,
and exclusion in saturated porous media.
Environ. Sci. Technoi. 37:2242-2250.
Bradford, S.A., and Bettahar, M. 2005.
Straining, attachment, and detach-
ment, of Cryptosporidium oocysts in
saturated porous media. J. Environ. Qual.
34:469478.
Bradford, S.A., and Bettahar, M. 2006.
Concentration dependent colloid transport
in saturated porous media. J. Contam,
Hydro I. 82:99-117.
Bradford, S.A., Simunek, J., Bettahar, M., van
Genuchten, M.T., and Yates, S.R. 2006a.
Significance of straining in colloid deposi-
tion: evidence and implications. Water
Resour. Res. 42:W12S15.
Bradford, S.A., Simunek, J., and Walker, S.L.
2006b. Transport and straining of E. coli
0157:H7 in saturated porous media.
Water Resour. Res., 42.W12S12.
Bradford, S.A., Tadassa, Y.F., and Pachepsky.
Y. 2006c. Transport of Giardia and manure
suspensions in saturated porous media. J.
Environ. Qual., 35:749-757.
Bradford, S.A., Tadassa, Y.F., and Jin, Y,
2006d. Transport of coliphage in the pres-
ence and absence of manure suspension.
J. Environ. Qual. 35:1692-1701.
Bradford, S.A., and Toride, N. 2007. A
stochastic model for colloid trans-
port and deposition. J. Environ. Qual.
36:1346-1356.
Bradford, S.A., Torkzaban, S., and Walker,
S.L. 2007. Coupling of physical and
chemical mechanisms of colloid straining
in saturated porous media. Water Res.,
41:3012-3024.
Bradford, S.A., Segal, E., Zheng, W., Wang,
Q., and Hutchins, S.R. 2008. Reuse
of concentrated animal feeding opera-
tion wastewater on agricultural lands. J.
Environ. Qual. 37: S97-S115.
Bradford, S.A. and Torkzaban, S. 2008.
Colloid transport and retention in unsatu-
rated porous media: a review of interface-,
collector-, and pore-scale processes and
models. Vadose Zone J. 7-.667-681.
Bradford, S.A., Torkzaban, S., Leij, F.J.,
Simunek, J., and van Genuchten,
M.T. 2009. Modeling the coupled
effects of pore space geometry and
velocity on colloid transport and reten-
tion. Water Resour. Res. 45:W02414.
doi:10.1029/2008WR007096.
Brock, E.H., Ketterings, Q.M., and Kleinman,
P.J.A. 2007. Measuring and predict-
ing the phosphorus sorption capacity
of manure-amended soils. Soil Sci.
172(4):266-278.
Burkholder, J., Libra, B., Weyer, P., Heathcote,
S., Kolpin, D., Thome, P.S., and Wichman,
M. 2007. Impacts of waste from con-
centrated animal feeding operations on
water quality. Environ. Health Perspect.
115:308-312.
-------
Burns, J.C., Westerrnan, P.W., King, L.D.,
Cummings, G.A., Overcash, M.R., and
Goode, L. 1985. Swine lagoon effluent
applied to 'Coastal' berrnudagrass: I.
Forage yield, quality, and element removal.
J. Environ. Qua/. 14:9-14.
Bussink, D.W., and Oenema, 0. 1998.
Ammonia volatilization from dairy farm-
ing systems in temperate areas: a review.
Nutrient Cycling Agroecosyst. 51:19-33.
Campbell-Mathews, M., Frate, C. Harter,
T., and Sather, S. 2001. Lagoon water
composition, sampling, and field analy-
sis. Proceedings of the 2001 California
Plant and Soil Conference, Fresno, CA. pp
43-51B.
Cameron, K.C., Rate, A.W., Carey, P.L., and
Smith, N.P. 1995. Fate of nitrogen in pig
effluent applied to a shallow stony pasture
soil. NZJ. Agric. Res. 38:533-542.
Cameron, K.C., Di, H.J., Reijnen, B.P.A., and
Li, Z. 2002. Fate of nitrogen in dairy fac-
tory effluent irrigated onto land. NZ J.of
Agric. Res. 45:207-216.
Callaway, T.R., Elder, R.O., Keen, J.E.,
Anderson, R.C., and Nisbet, D.J. 2003.
Forage feeding to reduce preharvest
Escherichia coli populations in cattle, a
review. J. Dairy Sci. 86:852-860.
Centers for Disease Control and Prevention.
1998. Surveys of waterborne disease
outbreaks compiled by the Center for
Disease Control and Prevention from 1986
to 1998. Morbidity and Mortality Weekly
Reports. Atlanta, GA, CDC.
Chang, C. and Entz, T. 1996. Nitrate leach-
ing losses under repeated cattle feedlot
manure applications in southern Alberta.
J. Environ. Quai. 25:145-153.
Chang, C., Sommerfeldt, T.G., and Entz, T.
1991. Soil chemistry after eleven annual
applications of cattle feedlot manure. J.
Environ. Qua!. 20:475-480,
Chastain, J.P., Vanotti, M.B., and Wingfield,
M.M. 2001. Effectiveness of liquid solid
separation for treatment of flushed dairy
manure: a case study. Appl. Eng. Agric.
17:343-354.
Chen, G. and Flury, M. 2005. Retention of
mineral colloids in unsaturated porous
media as related to their surface proper-
ties. Colloids Surf. A: Physicochem. Eng.
Aspects. 256:207-216.
Cherrey, K.D., Flury, M., and Harsh, J.B.
2003. Nitrate and colloid transport
through coarse Hanford sediments
under steady state, variably saturated
flow. Water Resour. Res. 39:1165.
doi:10.1029/2002WR001944.
Choi, H. and Corapcioglu, M.Y. 1997.
Transport of a non-volatile contaminant
in unsaturated porous media in the pres-
ence of colloids. J. Contam. Hydro!.
25:299-324.
Chu, Y., Jin, Y., Flury, M. and Yates, M.V.
2001. Mechanisms of virus removal during
transport in unsaturated porous media.
Water Resour. Res. 37:253-263.
Clesceri, L.S., Greenberg, A.E., and Trussell,
R.R, 1989, Standard methods for the
examination of water and wastewater.
American Public Health Association,
Washington, D.C.
Compere, F., Porel, G., and Delay, F. 2001.
Transport and retention of clay particles in
saturated porous media: influence of ionic
strength and pore velocity. J. Contam.
Hydro!. 49:1-21.
Corapcioglu, M.Y, and Choi, H. 1996.
Modeling colloid transport in unsaturated
porous media and validation with labo-
ratory column data. Water Resour. Res.
32:3437-3449.
Correll, D.L. 1998. The role of phosphorus in
the eutrophication of receiving waters: a
review. J. Environ. Qual. 27:261-266.
Corwin, D.L., and Lesch. S.M. 2005a.
Apparent soil electrical conductivity mea-
surements in agriculture. Computers and
Electronics in Agriculture. 46:11-43.
Corwin, D.L. and Lesch, S.M. 2005b.
Characterizing soil spatial variability
with apparent soil electrical conductiv-
ity. I. Survey protocols. Computers and
Electronics in Agriculture. 46:103-133.
-------
Craun, G.F., and Calderon, R, 1996,
Microbial risks in groundwater systems:
Epidemiology of waterborne outbreaks.
in: Under the microscope: Examining
microbes in groundwater. Denver, CO, Am.
Water Works Assoc. Res, Foundation.
Crist, J.T., McCarthy, J.F., Zevi, Y., Baveye, P.,
Throop, J.A., and Steenhuis, T.S. 2004.
Pore-scale visualization of colloid transport
and retention in partly saturated porous
media. Vadose Zone J. 3:444-450.
Crist, J.T., Zevi, Y., McCarthy, J.F., Throop,
J.A., and Steenhuis, T.S. 2005. Transport
and retention mechanisms of colloids in
partially saturated porous media. Vadose
Zone J. 4:184-195.
Cumbie, D.H. and McKay, L.D. 1999.
Influence of diameter on particle transport
in a fractured shale saprolite. J. Contam.
Hydro/, 37(1-2):139-157.
Gushing, R.S. and Lawler, D.F. 1998.
Depth filtration: fundamental inves-
tigation through three-dimensional
trajectory analysis. Environ. Sci. Technol.
32:3793-3801.
de Jonge, L.W., Kjaergaard, C., and Moldrup,
P. 2004. Colloids and colloid facilitated
transport of contaminants in soils: an
introduction. Vadose Zone J. 3.-321-325.
de Rooij, G.H. 2000. Modeling fingered
flow of water in soils owing to wetting
front instability: a review. J. Hydro/,
231-232:277-294.
DeNovio, N.M., Saiers, J.E., and Ryan, J.N.
2004. Colloid movement in unsaturated
porous media: recent advances and future
directions. Vadose Zone J. 3:338-351.
Derjaguin, B.V., and Landau, L.D. 1941.
Theory of the stability of strongly charged
lyophobic sols and of the adhesion of
strongly charged particles in solutions of
electrolytes. Acta Physicochim, USSR
14:733-762.
Dobias, B. Coagulation and flocculation.
Mercel Dekker, Inc. [v. 47]. 1993. New
York, NY. Surfactant Science Series,
Duffadar, R.D., and Davis, J.M. 2008.
Dynamic adhesion behavior of micrometer
scale particles flowing over patchy surfaces
with nanoscale electrostatic heterogeneity.
J. Coll. Interface Sci. 326:18-27.
Elimelech, M., Nagai, M., Ko, C.-H., and
Ryan, J.N. 2000. Relative insignificance
of mineral grain zeta potential to colloid
transport in geochemically heterogeneous
porous media. Environ. Sci. Technol.
34:2143-2148.
Evans, D.D., Nicholson, T.J., and Rasmusson,
T.C. 2001. Flow and transport through
unsaturated fractured rock. 2nd ed. AGU,
Washington, DC, Geophysical Monograph
42.
Evans, R.O., Westerman, P.W., and Overcash,
M.R. 1984. Subsurface drainage
water quality from land application of
swine lagoon effluent. Trans. ASABE
27:473-480.
Feng, G.L., Letey, J., Chang, A.C., and
Campbell-Mathews, M. 2005. Simulating
dairy liquid waste management options as
a nitrogen source for crops. Agric. Ecosyst.
Environ. 110:219-229.
Foppen, J.W.A., Mporokoso, A., and Schijven,
J.F. 2005. Determining straining of
Escherichia coli from breakthrough curves.
J. Contam. Hydro/. 76:191-210.
Gagliardi, J.V. and Karns J.S. 2000. Leaching
of Escherichia coli 0157:H7 in diverse
soils under various agricultural manage-
ment practices. Appl. Environ. Microbiol.
66:877-883.
Gagliardi, J.V. and Karns, J.S. 2002.
Persistence of Escherichia coli 0157:H7
in soil and on plant roots. Environ.
Microbiol. 4(2):89-96.
Gerba, C.P. 1996. What are the current micro-
biological and public health issues in
drinking water? in: Under the microscope:
Examining microbes in groundwater.
Denver, CO. Am, Water Works Assoc. Res.
Foundation.
Gerba, C.P., and Smith Jr., J.E. 2005. Sources
of pathogenic microorganisms and their
fate during land application of wastes. J.
Environ. Qual. 34:42-48.
-------
Gerke, H.H., and van Genuchten, M. T.
1993a. A dual-porosity model for simu-
lating the preferential movement of
water and solutes in structured porous
media. Water Resour, Res. 29:305-319,
doi:10.1029/92WR02339.
Gerke, H.H., and van Genuchten, M.T. 1993b.
Evaluation of a first-order water transfer
term for variably saturated dual-porosity
flow models. Water Resour. Res. 29:1225-
1238. doi:10.1029/92WR02467.
Gerke, H.H. 2006. Preferential flow descrip-
tions for structured soils. J. Plant Nutrition
SoilSci. 169:382-400. doi: 10.10027
jpln. 200521955.
Gerritson, J. and Bradley, S.W. 1987.
Electrophoretic mobility of natu-
ral particles and cultured organisms
in fresh waters. Limnol. Oceanogr.
32(5):1049-1058.
Gibson, L.R., Nance, C.D., and Karlen, D.L.
2007. Winter triticale response to nitro-
gen fertilization when grown after corn or
soybean. Agron. J. 99:49-58.
Ginn, T.R. 2002. A travel time approach to
exclusion on transport in porous media.
Water Resour. Res. 38(4):1041.
Ginn, T.R., Wood, B.D., Nelson, K.E.,
Schiebe, T.D., Murphy, E.M., and Clement,
T.P. 2002. Processes in microbial trans-
port in the natural subsurface. Adv. Water
Res. 25:1017-1042.
Gleick, P.H. The world's water. 2000.
Washington, DC. Island Press.
Gregory, J. 1981. Approximate expressions
for retarded van der Waals interaction. J.
Colloid Interface Sci. 83:138-145.
Gu, B., Phelps, T.J., Liang, L, Dickey, M.J.,
Roh, Y., Kinsall, B.L., Palumbo, A.V.,
and Jacobs, G.K. 1999. Biogeochemical
dynamics in zero-valent iron columns:
implication for permeable reactive barri-
ers. Environ.Sci.Techno!. 33:2170-2177.
Guber, A.K., Shelton, D.R., and Pachepsky,
Y.A. 2005a. Effect of manure on
Escherichia coli attachment to soil. J.
Environ. Quai. 34:2086-2090.
Guber, A.K., Shelton, D.R., and Pachepsky,
Y.A. 2005b. Transport and retention of
manure-borne coliforms in soil. Vadose
Zone J. 4:828-837.
Hanselrnan, T.A., Graetz, D.A., and Wilkie,
A.C. 2003. Manure-borne estrogens
as potential environmental contami-
nants: a review. Environ. Sci. Techno!.
37:5471-5478.
Hao, X. and Chang, C. 2003. Does long-term
heavy cattle manure application increase
salinity of a clay loam soil in semi-arid
southern Alberta? Agric. Ecosyst. Environ.
94:89-103.
Harter, T., Wagner, S., and Atwill, E.R.
2000. Colloid transport and filtration of
Cryptosporidium parvum in sandy soils and
aquifer sediments. Environ. Sci. Technol.
34:62-70.
Harvey, R.W., and Garabedian, S.P. 1991.
Use of colloid filtration theory in modeling
movement of bacteria through a contami-
nated sandy aquifer. Environ. Sci. Technol.
25:178-185.
Harvey, R.W. and Harms. H. 2001. Transport
of microorganisms in the terrestrial subsur-
face: In situ and laboratory methods, in:
Manual of Environmental Microbiology, 2nd
Edition: ASM Press. Washington, DC. pp.
753-776.
Havelaar, A.H. 1986. F-specific bactriophages
as model viruses in water treatment pro-
cesses. Univ. of Ultrecht, Ultrech, The
Netherlands.
Hawke, R.M., and Summers, S.A. 2006.
Effects of land application of farm dairy
effluent on soil properties: a literature
review. NZ J of Agric. Res. 49:307-320.
Herzig, J.P., Leclerc, D.M., and LeGoff, P.
1970. Flow of suspension through porous
media: application to deep filtration. Ind.
Eng. Chem. 62(5): 129-157.
Hillel, D. 1998. Environmental soil physics.
Academic Press.
Hoek, E.M.V., and Agarwal, G.K., 2006.
Extended DLVO interactions between
spherical particles and rough surfaces. J.
Colloid Interface Sci. 298:50-58.
-------
Hogg, R., Healy, T.W., and Fuerstenau,
D.W. 1966. Mutual coagulation of col-
loidal dispersions. Trans, Faraday Soc.
62:1638-1651.
Hunt, P.G., Szogi, A.A., Humenik, F.J., Rice,
J.M., Matheny, T.A., and Stone, K.C.
2002. Constructed wetlands for treatment
of swine wastewater from an anaerobic
lagoon. Trans. ASAE. 45:639-647.
Hutchins, S.R., White, M.V., Hudson, P.M.,
and Fine, D.D. 2007. Analysis of lagoon
samples from different concentrated
animal feeding operations for estrogens
and estrogen conjugates. Environ. Sci.
Techno!. 41:738-744.
Hutson, S.S., Barber, N.L., Kenny, J.F.,
Linsey, K.S., Lumia, D.S., and Maupin,
M.A. 2004. Estimated use of water in the
United States in 2000. USGS Circ. 1268.
U.S. Geological Survey, Reston, VA.
Ibekwe, A.M., Grieve, C.M., and Lyon, S.R.
2003. Characterization of microbial com-
munities and composition in constructed
dairy wetland wastewater effluent. Appl.
Environ. Microbiol. 69(9): 5060-5069.
Jin, Y. and Flury, M. 2002. Fate and transport
of viruses in porous media. Adv. Agron.
35:39-102.
Jin, Y., Pratt, E., and Yates, M.V. 2000.
Effect of mineral colloids on virus trans-
port through saturated sand columns. J.
Environ. Qual. 29:532-539.
Johnson, W.P. and Logan, B.E. 1996.
Enhanced transport of bacteria in porous
media by sediment-phase and aqueous-
phase natural organic matter. Water Res.
30(4):923-931.
Johnson, P.R., and Elirnelech, M. 1995.
Dynamics of colloid deposition in porous
media: blocking based on random sequen-
tial adsorption. Langmuir 11:801-812.
Johnson, W.R, long, M., and Li, X. 2007. On
colloid retention in saturated porous media
in the presence of energy barriers: The
failure of a, and opportunities to predict r).
Water Resour. Res. 43:W12S13.
Jokela, W.E. 1992. Nitrogen fertilizer and
dairy manure effects on corn yield and soil
nitrate. Soil Sci. Soc. Am. J. 56:148-154.
Jongbloed, A.W. and Lenis, N.P. 1998.
Environmental concerns about animal
manure. J. Anim.Sci. 76:2641-2648.
Joshi, A. and Chaudhuri, M. 1996. Removal
of arsenic from groundwater by iron oxide-
coated sand. J. Environ. Eng. (ASCE).
122(8):769-771.
Katterer, T., Hansson, A-C., and Andren, 0.
1993. Wheat root biomass and nitrogen
dynamics-effects of daily irrigation and
fertilization. Plant and Soil 151: 21-30.
Kapkiyai, J.J., Karanja, N.K., Qureshi, J.N.,
Smithson, P.C. and Woomer, P.L. 1999.
Soil organic matter and nutrient dynamics
in a Kenyan nitisol under long-term fertil-
izer and organic input management. So/7
Biol. Biochem. 31:1773-1782.
Keeney, D.R., and Nelson, D.W. 1982.
Nitrogen-inorganic forms, p.643-698.
A.L. Page (Eds.) Methods of soil analysis,
part 2. Agron. Monogr. 9. ASA and SSSA,
Madison, Wl.
Kellogg, R.L., Lander, C.H., Moffitt, D.C. and
Gollehon, N. 2000. Manure nutrients
relative to the capacity of cropland and
pastureland to assimilate nutrients spatial
and temporal trends for the United States.
USDA National Resources Conservation
Service and Economic Research Service.
Khilar, K.C. and Fogler, H.S. 1998. Migration
of fines in porous media. Dordrecht, The
Netherlands. Kluwer Academic Publishers.
Kim, H.N., Bradford, S.A. and Walker, S.L.
2009a. Escherichia co//0157:H7 trans-
port in saturated porous media: role of
solution chemistry and surface mac-
romolecules. Environ. Sci. Techno!.
43(12):4340-4347.
Kim, H.N., Hong, Y., Lee, I., Bradford, S.A.
and Walker, S.L. 2009b. Surface char-
acteristics and adhesion behavior of
Escherichia coll 0157:H7: role of extracel-
lular macromolecules. Biomacromolecules
10: 2556-2564.
-------
King, L.D., Westerman, P.W., Cummings, G.A.,
Overcash, M.R. and Burns, J.C. 1985.
Swine lagoon effluent applied to 'coastal'
bermudagrass: II. Effects on soil. J.
Environ. Qual. 14:14-21.
Kretzschmar, R., Barmettler, K., Grolimund,
D., Yan, Y.D., Borkovec, M., and Sticher,
H. 1997. Experimental determination
of colloid deposition rates and collision
efficiencies in natural porous media. Water
Resour. Res. 33(5): 1129-1137.
Kuznar, Z.A. and Elimelech, M. 2007. Direct
microscopic observation of particle deposi-
tion in porous media: role of the secondary
energy minimum. Colloids Surf. A:
Physicochem. Eng. Aspects 94:156-162
Lance, J.C. and Gerba, C.P. 1984. Virus
movement in soil during saturated
and unsaturated flow. Appl. Environ.
Microbiol. 47:335-341.
Lander, C.H., Moffitt, D. and Alt, K. 1998.
Nutrients available from livestock manure
relative to crop growth requirements.
Resource Assessment and Strategic
Planning Working Paper 98-1. U.S.
Department of Agriculture, Natural
Resources Conservation Service.
Leij, F.J. and Bradford, S.A. 2009. Combined
physical and chemical nonequilibrium
transport model: Analytical solution,
moments, and application to colloids. J.
Cont. Hyd. 110:87-99.
Lenhart, J.J., and Saiers, J.E. 2002. Transport
of silica colloids through unsaturated
porous media: experimental results and
model comparisons. Environ. Sci. Technol.
36:769-777.
Li, X., Scheibe, T.D. and Johnson, W.P. 2004.
Apparent decreases in colloid deposition
rate coefficient with distance of transport
under unfavorable deposition conditions:
a general phenomenon. Environ. Sci.
Technol. 38:5616-5625.
Li, X., and Johnson, W.P. 2005. Non-
monotonic variations in removal rate
coefficients of microspheres in porous
media under unfavorable deposi-
tion conditions. Environ. Sci. Technol.
39:1658-1665.
Li, X., Lin, C.L., Miller, J.D., and Johnson,
W.P. 2006a. Pore-scale observation
of microsphere deposition at grain-to-
grain contacts over assemblage-scale
porous media domains using X-ray
microtomography. Environ. Sci. Technol.
40:3762-3768.
Li, X., Zhang, P., Lin, C.L., and Johnson,
W.P. 2005. Role of hydrodynamic drag
on microsphere deposition and re-
entrainment in porous media under
unfavorable conditions. Environ. Sci.
Technol. 39:4012-4020.
Li, X., Lin, C.L., Miller, J.D., and Johnson,
W.P. 2006b. Role of grain-to grain con-
tacts on profiles of retained colloids
in porous media in the presence of an
energy barrier to deposition. Environ. Sci.
Technol. 40:3769-3774.
Lindeburg, M.R. 2001. Civil engineering refer-
ence manual for the PE exam. Professional
Publications, Inc. Belmont, CA.
Loge, F.J., Thompson, D.E. and Call, D.R.
2002. PCR detection of specific patho-
gens in water: A risk-based analysis.
Environ. Sci. Technol. 36:2754-2759.
Loughrin, J .H., Szogi, A.A. and Vanotti, M.B.
2006. Reduction of malodorous com-
pounds from liquid swine manure by a
multi-stage treatment system. Appl. Eng.
Agric. 22:867-873.
Luo, J., Tillman, R.W., and Ball, P.R. 1999.
Factors regulationg denitrification in a soil
under pasture. So/7 Biology and Biochem.
31:913-927.
Maas, E.V. and Hoffman, G.J. 1977. Crop salt
tolerance - current assessment. J. Irrig.
Drainage Div. ASCE 103:115-134.
Maas, E.V. 1984. The handbook of plant
science in agriculture. CRC Press. Boca
Raton, FL.
Macler, B.A. and Merkle, J.C. 2000. Current
knowledge on groundwater microbial
pathogens and their control. Hydrogeol. J.
8:29-40.
Mallin, M.A. 2000. Impacts of industrial-scale
swine and poultry production on rivers and
estuaries. American Scientist 88:26-37.
-------
McCarthy, J.F. and Zachara, J.M. 1989.
Subsurface transport of contaminants,
Environ. Sci. Techno!. 23:496-502.
McDowell-Boyer, L.M., Hunt, J.R. and Sitar,
N. 1986. Particle transport through
porous media. Water Resour. Res.
22:1901-1921.
Merrill, S.D. and Rawlins, S.L. 1979.
Distribution and growth of sorghum roots
in response to irrigation frequency. Agron.
J. 71:738-745
Mohanty, B.P., Bowman, R.S., Hendrickx,
J.M.H., and van Genuchten, M.T. 1997.
New piecewise-continuous hydraulic func-
tions for modeling preferential flow in an
intermittent-flood-irrigated field. Water
Resour. Res. 33:2049-2063.
Moore Jr., P.A., and Reddy, K.R. 1994. Role
of Eh and pH on phosphorus geochemistry
in sediments of Lake Okeechobee, Florida.
J. Environ. Qual. 23:955-964.
MWPS. 1993. MWPS-18: Livestock waste
facilities handbook. Midwest Plan Service,
Iowa State University. Ames, IA.
NCSU. 1994. Livestock manure production
and characterization in North Carolina.
North Carolina Coop. Ext. Service, Raleigh,
NC.
National Research Council. 2001. Conceptual
models of flow and transport in the frac-
tured vadose zone. National Academy
Press. Washington, DC.
Natsch, A., Keel, C., Troxler, J., Zala, M.,
Von Albertini, N. and Defago, G. 1996.
Importance of preferential flow and soil
management in vertical transport of a bio-
control strain of Pseudomonas fluorescens
in structured field soil. Appl. Environ,
Micmbiol. 62:33-40.
Nelson, K.E. and Ginn, T.R. 2001. Theoretical
investigation of bacterial chemotaxis in
porous media. Langmuir. 17:5636-5645.
Nichols, D.J., Daniel Jr., C.P., Moore, P.A.,
Edwards, D.R. and Pote, P.M. 1997.
Runoff of estrogen hormone 17 p-Estradiol
from poultry litter applied to pasture. J.
Environ. Qual. 26:1002-1006.
Ouyang, Y., Shinde, D., Mansell, R.S. and
Harris, W. 1996. Colloid-enhanced
transport of chemicals in subsurface envi-
ronments: a review. Crit. Rev, Environ, Sci,
Technol, 26:189-204.
Pautler, M.C. and Sims, J.T. 2000.
Relationships between soil test phospho-
rus, soluble phosphorus, and phosphorus
saturation in Delaware soils. Soil Sci. Soc.
Am. J. 64:765-773.
Penman, H.L. 1948. Proceedings of the Royal
Society of London.
Peterson, E.W., Davis, R.K. and Orndorff,
H.A. 2000. 17p-estradiol as an indica-
tor of animal waste contamination in
mantled karst aquifers. J. Environ. Qual.
29:826-834.
Pieper, A.P., Ryan, J.N., Harvey, R.W., Arny,
G.L., Illangasekare, T.H. and Metge, D.W.
1997. Transport and recovery of bacterio-
phage PRD1 in a sand and gravel aquifer:
Effect of sewage-derived organic matter.
Environ. Sci, Technol. 31:1163-1170.
Pimentel, D., Berger, B., Filiberto, D., Newton,
M., Wolfe, B., Karabinakis, E., Clark, S.,
Poon, E., Abbett, E. and Nandagopal, S.
2004. Water resources: agricultural and
environmental issues. Biosci. 54:909-918.
Powelson, O.K. and Mills, A.L. 2001.
Transport of Escherichia coli in sand col-
umns with constant and changing water
contents. J. Environ. Qual. 30:238-245.
Rahman, M., Fukai, S, Blarney, F.P.C. 2001.
Forage production and nitrogen uptake
of forage sorghum and maize as affected
by cutting under different nitrogen
levels. Proceedings of the 10th Australian
Agronomy Conference, Hobart, Australia.
Rajagopalan, R., and Tien, C. 1976. Trajectory
analysis of deep-bed filtration with the
sphere-in-cell porous-media model.
AlChEJ., 22:523-533. doi:10.10027
aic.690220316,
Raman, D.R., Williams, E.L., Layton, A.C.,
Burns, R.T., Easter, J.P., Daugherty,
A.S., Mullen, M.D. and Sayler, G.S.
2004. Estrogen content of dairy and
swine wastes. Environ. Sci. Technol,
38:3567-3573.
-------
Ray, C,, Soong, T.W., Lian, Y.Q. and Roadcap,
G.S, 2002, Effect of flood induced chemi-
cal load on filtrate quality at bank filtration
sites. J. Hydro!. 266:235-258.
Reddy, K.R., Khaleel, R., and Overcash,
M.R.1981. Behavior and transport of
microbial pathogens and indicator organ-
isms in soils treated with organic wastes.
J. Environ. Qua!. 10:255-266.
Redman, J.A., Walker, S.L. and Elimelech, M.
2004. Bacterial adhesion and transport
in porous media: role of the secondary
energy minimum. Environ. Sci. Technol.
38:1777-1785.
Ritsema, C.J. and Dekker, L.W. 2000.
Preferential flow in water repellent sandy
soils: principles and modeling implica-
tions. J. Hydrol. 231/232:308-319.
Ritsema, C.J., Dekker, V., Hendrickx, J.M.H.,
and Hamminga, W. 1993. Preferential flow
mechanism in a water repellent soil. Water
Resour. Res. 29:2183-2193.
Rockhold, M.L., Yarwood, R.R. and Selker,
J.S. 2004. Coupled microbial and trans-
port processes in soils. Vadose Zone J.
3:368-383.
Rodgers, M., Healy, M.G., and Mulqueen, J.
2005. Organic carbon removal and nitrifi-
cation of high strength wastewaters using
stratified sand filters. Water Research
39:3279-3286.
Russell, J.B., Diez-Gonzalez, F. and Jarvis,
G.N. 2000. Effects of diet shifts on
Escherichia coll in cattle. J. Dairy Sci.
83:863-873.
Ryan, J.N. and Elimelech, M. 1996. Colloid
mobilization and transport in groundwater.
Colloids and Surfaces A: Physicochem.
and Eng. Aspects. 107:1-56.
Ryan, J.N., Harvey, R.W., Metge, D.,
Elimelech, M., Navigato, T. and Pieper,
A.P. 2002. Field and laboratory investiga-
tions of inactivation of viruses (PRDland
MS2) attached to iron oxide-coated quartz
sand. Env. Sci. Techno!. 36.-2403-2413
Sagripanti, J.L., Routson, L.B. and Lytle, C.D.
1993. Virus inactivation by copper or iron
ions alone and in the presence of peroxide.
Appl. Environ. Microbiol. 59:4374-4376.
Saiers, J.E., Hornberger, G.M., Gower, D.B.
and Herman, J.S. 2003. The role of
moving air-water interfaces in colloid
mobilization within the vadose zone.
Geophy. Res. Lett. 30:2083.
Saiers, J.E. and Lenhart, J.J. 2003. Colloid
mobilization and transport within
unsaturated porous media under tran-
sient-flow conditions. Water Resour. Res.
39(1):1019.
Schafer, A., Harms.H., and Zehnder, A.J.B.
1998. Bacterial accumulation at the air-
water interface. Environ. Sci. Technol.
32:3704-3712.
Schafer, A., Ustohal, P., Harms, H., Stauffer,
F., Dracos, T., and Zehnder, A.J.B. 1998.
Transport of bacteria in unsaturated porous
media. J. Contain. Hydrol. 33:149-169.
Shen, C., Huang, Y., Li, 6., and Jin, Y.
2008. Effects of solution chemis-
try on straining of colloids in porous
media under unfavorable conditions.
Water Resour. Res. 44:W05419, DOI:
10.1029/2007WR006580.
Scherer, M.M., Richter, S., Valentine, R.L.
and Alvarez, P.J. 2000. Chemistry and
microbiology of permeable reactive barriers
for in situ ground water cleanup. Crit. Rev.
Microbiol. 26:221-264.
Schets, F.M., During, M., Italiaander, R.,
Heijnen, L, Rutjes, S.A., van der Zwaluw,
W.K. and de Roda Husman, A.M. 2005.
Escherichia col'10157:H7 in drinking
water from private water supplies in the
Netherlands. Water Res. 39:4485-4493.
Schijven, J.F., Hoogenboezem, W.,
Hassanizadeh, S.M. and Peters, J.H.
1999. Modeling removal of bacterio-
phage MS2 and PRD1 by dune recharge
at Castricum, Netherlands. Water Resour.
Res. 35:1101-1111,
Schijven, J.E, Mulschlegel, J.H.,
Hassanizadeh, S.M., Teunis, P.F. and de
Roda Husman, A.M. 2006. Determination
of protection zones for Dutch groundwater
wells against virus contamination - uncer-
-------
tainty and sensitivity analysis, J. Water.
Health. 4:297-312.
waste and the land-water interface, Lewis
Publ. Boca Raton, FL. pp. 155-162.
Schijven, J.K, and Hassanizadeh, S.M.
2000. Removal of viruses by soil pas-
sage: overview of modeling, processes,
and parameters. Crit. Rev, Environ, Sci,
Technol. 30:49-127.
Segal, E., Shouse, P., Poss, J. A., Crohn,
D. M. and Bradford, S. A. 2010.
Recommendations for nutrient manage-
ment plans in a semi-arid environment.
Agric. Ecosyst. Environ., 137:317-328.
Segal, E., Bradford, S.A., Shouse, P.,
Lazarovitch, N. and Corwin, D.L. 2008.
Integration on of hard and soft data to
characterize field-scale hydraulic proper-
ties for flow and transport studies. Vadose
Zone J. 7-.878-889.
Segal, E., Shouse, P. and Bradford, S.A.
2009. Deterministic analysis and
upscaling of bromide transport in a het-
erogeneous vadose zone. Vadose Zone J,
8:601-610.
Sen, T.K. and Khilar, K.C. 2006. Review on
subsurface colloids and colloid associated
contaminant transport in saturated porous
media. Adv. Coll. Interface Sci.l 19:71-96.
Sharma, M.M., Chang, Y.I. and Yen, T.F.
1985. Reversible and irreversible surface
charge modification of bacteria for facili-
tating transport through porous media.
Co//, Surf. 16:193-206.
Sharpe, R.R., and Harper, L.A. 1997.
Ammonia and nitrous oxide emissions from
sprinkler irrigation applications of swine
effluent. J. Environ. Qua!. 26:1703-1706.
Sharpley, A.N. 1995. Soil phosphorus dynam-
ics: agronomic and environmental impacts.
EcoI. Engr, 5:261-279.
Sharpley, A.N. and Moyer, B. 2000.
Phosphorus forms in manure and compost
and their release during simulated rainfall.
J. Environ. Qual. 29:1462-1469.
Shore, L.S., Correll, D.L. and Chakraborty,
P.K. 1995. Relationship of fertilization
with chicken manure and concentrations
of estrogens in small streams. In: Animal
Shore, L.S., Shemesh, M. and Cohen,
R. 1988. The role of o-Estradiol and
o- Estrone in chicken manure silage
in hyperoestrogenism in cattle. Aust.
Veterinary J. 65:68.
Sim, Y., and Chrysikopoulos, C.V. 1996. One-
dimensional virus transport in porous
media with time-dependent inactiva-
tion rate coefficients. Water Resour. Res.
32:2607-2611.
Sims, J.T., Simard, R.R., and Joern, B.C.
1998. Phosphorus loss in agricul-
tural drainage: historical perspective
and current research. J. Environ. Qual.
27:277-293.
Sirnunek, J., Jarvis, N.J., van Genuchten,
M.Th., and Cardenas, A. 2003.
Nonequilibrium and preferential flow and
transport in the vadose zone: review and
case study. J. Hydro/. 272:14-35.
Sirnunek, J., van Genuchten, M.Th., and
Sejna, M. 2007. The HYDRUS-1D
software package for simulating the one-
dimensional movement of water, heat,
and multiple solutes in variably saturated
media. HYDRUS Software Ser. 3, 315
pp., Dept. of Environ, Sci,, Univ.of Calif.,
Riverside.
Sirnunek, J., van Genuchten, M. T. and Sejna,
M. The HYDRUS-1D software pack-
age for simulating the one-dimensional
movement of water, heat, and multiple
solutes in variably saturated media, ver-
sion 4.08, HYDRUS Software Ser. 3, 332
pp., Dept of Environ. Sci., Univ. of Calif.,
Riverside.
Sirnunek, J. and van Genuchten, M. T. 2008.
Modeling nonequilibrium flow and trans-
port using HYDRUS. Vadose Zone J.,
7:782-797, doi:10.2136/vzj2007.0074.
Sinton, L.W., Noonan, M.J., Finlay, R.K.,
Pang, L. and Close, M.E. 2000. Transport
and attenuation of bacteria and bacte-
riophages in an alluvial gravel aquifer.
New Zealand J, Mar. Freshwater Res.
34:175-186.
-------
Sommerfeldt, T.G., Chang, C, and Entz, T.
1988. Long-term annual manure appli-
cations increase soil organic matter and
nitrogen, and decrease carbon to nitrogen
ratio. Soil Sci. Soc. Am. J. 52:1667-1672.
Sommerfeldt, T.G. and Chang, C. 1985.
Changes in soil properties under annual
applications of feedlot manure and differ-
ent tillage practices. So/7 Sci. Soc, Am. J.
49:983-987.
Sommerfeldt, T.G. and Chang, C. 1987. Soil-
water properties as affected by twelve
annual applications of cattle feedlot
manure. So/7 Sci. Soc. Am. J. 51:7-9.
Steenhuis, T.S., Dathe, A., Zevi, Y., Smith,
J.L., Gao, B., Shaw, S.B., DeAlwis, D.,
Amaro-Garcia, S., Fehrman, R., Ekrem
Cakmak, E., Toevs, I.C., Liu, B.M., Beyer,
S.M., Crist, J.T., Hay, A.G., Richards,
B.K., DiCarlo, D. and McCarthy, J.F. 2006.
Biocolloid retention in partially saturated
soils. Biologia61 (Suppi. 19) S229-S233.
Stenger, R., Barkle, G.F., and Burgess, C.P.
2001. Mineralization and immobilization of
C and N from dairy farm effluent (DFE) and
glucose plus ammonium chloride solution
in three grassland topsoils. So/7 Biology
and Biochem. 33:1037-1048
Strong, D.T., Sale, P.W.G., Helyar, K.R. 1999.
The influence of the soil matrix on nitro-
gen mineralisation and nitrification. IV.
Texture. Australian Journal of Soil Research
37:329-350.
Svoboda, I.F. 1995. Nitrogen removal from
pig slurry by nitrification and denitrifica-
tion. ASAE. Proc. 7th Int. Symp. Agric. and
Food Proc. Wastes (ISAFPW95) St. Joseph,
Ml. pp. 24-33.
Sukias, J.P.S., Tanner, C.C., Davies-Colley,
R.J., Nagels, J.W., and Wolters, R. 2001.
Algal abundance, organic matter, and
physico-chemical characteristics of
dairy farm facultative ponds: implica-
tions for treatment performance. New
Zealand Journal of Agricultural Research
44:279-296.
Tan, Y., Cannon, J.T., Baveye, P., and
Alexander, M. 1994. Transport of bacte-
ria in an aquifer sand—experiments and
model simulations. Water Resour. Res,
30:3243-3252.
Thorne, P.S. 2007. Environmental health
impacts of concentrated animal feeding
operations: anticipating hazards - search-
ing for solutions. Environ, Health Perspect,
115:296-297.
Tipping, E. 1981. The adsorption of aquatic
humic substances by iron oxides.
Geochimica et Cosmochimica Acta
45:191-199.
Tong, M., Li, X., Brow, C.N. and Johnson,
W.P. 2005. Detachment-influenced trans-
port of an adhension-deficient bacterial
strain within water-reactive porous media.
Environ. Sci. Technol. 39:2500-2508.
Toride, N., Leij, F.J. and van Genuchten, M.Th.
1995, The CXTFIT code for estimating
transport parameters from laboratory and
field tracer experiments. Research Rep.
No. 137, U.S. Salinity Lab., ARS, USDA,
Riverside, CA.
Torkzaban, S., Hassanizadeh, S.M., Schijven,
J.F. and de Roda Husman, A.M. 2006a.
Virus transport in saturated and unsatu-
rated sand columns. Vadose Zone J.
5:877-885.
Torkzaban, S., Hassanizadeh, S.M., Schijven,
J.F. and van den Berg, H.H.J.L 2006b.
Role of air-water Interfaces on retention
of viruses under unsaturated conditions.
Water Resour. Res. 42:W12S14.
Torkzaban, S., Bradford, S.A. and Walker, S.L.
2007. Resolving the coupled effects of
hydrodynamics and DLVO forces on colloid
attachment to porous media. Langmuir.
23:9652-9660, doi:10.1021/la700995e.
Torkzaban, S., Bradford, S.A. and Walker, S.L.
2008a. Colloid transport in unsaturated
porous media: the role of water content
and ionic strength on particle strain-
ing. J. Contam. Hydro!. 96:113-127.
doi:10.1016/j.jconhyd.2007.10.006.
Torkzaban, S., Tazehkand, S.S., Walker, S.L.,
and Bradford, S.A. 2008b. Transport and
fate of bacteria in porous media: Coupled
effects of chemical conditions and pore
space geometry. Water Resour. Res.,
44:W04403.
-------
Tufenkji, N., Ryan, J.N. and Elirnelech, M.
2002, The promise of bank filtration.
Environ. Sci. Techno!. 36:A422-A428.
Tufenkji, N., Elimelech, M. 2004. Deviation
from the classical colloid filtration theory
in the presence of repulsive DLVO interac-
tions. Langmuir. 20:10818-10828.
Tufenkji, N., Miller, G., Ryan, J., Harvey,
R.W., and Elimelech, M. 2004. Transport
of Cryptosporidium oocysts in porous
media: Role of straining and physico-
chemical filtration. Environ. Sci. Technol.
38:5932-5938.
Tufenkji, N., and Elimelech, M. 2005a.
Breakdown of colloid filtration theory: role
of the secondary energy minimum and
surface charge heterogeneities, Langmuir
21:841-852.
Tufenkji, N., and Elimelech, M. 2005b.
Spatial distributions of Cryptosporidium
oocysts in porous media: evidence for dual
mode deposition. Environ. Sci. Technol.
39:3620-3629.
Tufenkji, N., Dixon, D.R., Considine, R. and
Drummond, CJ. 2006. Multi-scale
Cryptosporidium/sand interactions in water
treatment. Water Res. 40:3315- 3331.
United States Department of Agriculture.
1992. Part 651 (210-AWMFH, 4/92).
National engineering handbook:
Agricultural waste management field hand-
book. USDA. Washington, DC.
United States Department of Agriculture.
1996. NRCS. National Engineering
Handbook: Agricultural Waste Management
Field Handbook, U.S. Dept. Commerce,
/V7/S, Springfield, VA.
United States Environmental Protection
Agency. 1997. National Water Quality
Inventory, Report to Congress, EPA841-R-
97-008. Washington, DC, USEPA.
United States Environmental Protection
Agency. 1998. Environmental impacts
of animal feeding operations. Office of
Water, Standards and Applied Sci. Div.
Washington, DC, USEPA.
United States Environmental Protection
Agency. 2000. National Primary Drinking
Water Regulations: Ground Water Rule,
Proposed Rule.
United States Environmental Protection
Agency. 2001a. Development document
for the proposed revisions to the national
pollutant discharge elimination system
permit regulation and effluent limitation
guidelines and standards for concentrated
animal feeding operations (CAFOs). EPA-
821-R-01-003. 2001, Washington, DC,
USEPA,
United States Environmental Protection
Agency. 2001b. Method 1601: Male
specific (F+) and somatic coliphage in
water by two-step enrichment procedure.
EPA-821-R-01-030. Office of Water,
Engineering and Analysis Div., Washington,
DC, USEPA.
United States Environmental Protection
Agency. 2003. National pollutant
discharge elimination system permit regu-
lation and effluent limitation guidelines
and standards for concentrated animal
feeding operations (CAFOs). Federal
Register 69(29)-.7176-7274.
United States Environmental Protection
Agency. 2004. Guidelines for water reuse.
EPA 625/R-04/108. Cincinnati, OH,
USEPA.
United States Food and Drug Administration.
2007, FDA finalizes report on 2006 spin-
ach outbreak, FDA News.
United States Department of Agriculture.
2000. National Nutrient Management
(Code 590), Natural Resource
Conservation Service, Conservation
Practice Standard.
Valenzuela-Solano, C. and Crohn, D.M. 2006.
Are decomposition and N release from
organic mulches determined mainly by
their chemical composition? So/7 Biology &
Biochemistry 38:377-384.
Vanderlip, R.L., and Reeves, H.E. 1972.
Growth stages of sorghum [Sorghum
bicolor, (L.) moench]. Agron. J. 64:13-16.
Van Veen, J.A., van Overbeek, L.S. and van
Elsas, J.D. 1997. Fate and activity of
microorganisms introduced into soil.
Microbiol. Mol. Bio I. Rev. 61:121-135.
-------
Vanotti, M.B. and Hunt, P.G. 1999. Solids
and nutrient removal from flushed swine
manure using polyacrylamides. Trans.
ASAE. 42:1833-1840.
Vanotti, M.B., Millner, P.O., Hunt, P.G. and
Ellison, A.Q. 2005a. Removal of pathogen
and indicator microorganisms from liquid
swine manure in multi-step biological and
chemical treatment. Bioresour, Techno!,
96:209-214.
Vanotti, M.B., Rashash, D.M.C. and Hunt,
P.G. 2002. Solid-liquid separation of
flushed swine manure with PAM: Effect
of wastewater strength. Trans. ASAE.
45:1959-1969.
Vanotti, M.B., Rice, J.M., Ellison, A.Q., Hunt,
P.G., Humenik, F.J. and Baird,C.L. 2005b.
Solid-liquid separation of swine manure
with polymer treatment and sand filtration.
Trans. ASAE. 48:1567-1574.
Verwey, E.J.W., and Overbeek, J.Th.G. 1948.
Theory of the stability of lyophobic col-
loids. Elsevier, Amsterdam.
Wan, J. and Tokunaga, T.K. 1997. Film strain-
ing of colloids in unsaturated porous
media: Conceptual model and experi-
mental testing. Environ. Sci, Techno!,
31:2413-2420.
Wan, J. and Wilson, J.L. 1994a. Colloid trans-
port in unsaturated porous media. Water
Resour. Res. 30:857-864.
Wan, J. and Wilson, J.L. 1994b. Visualization
of the role of the gas-water interface on
the fate and transport of colloids in porous
media. Water Resour. Res. 30:11-23.
Wan, J.M., and Tokunaga, T.K. 2002.
Partitioning of clay colloids at air-
water interfaces. J. Coll, Interf. Sci.,
247:54-61.
Wang, D.S., Gerba, C.P., and Lance, J.C.
1981. Effect of soil permeability on virus
removal through soil. Appl. Environ.
Microbiol, 42:83-88.
Wang, G., Zhao, T. and Doyle, M.P. 1996.
Fate of enterohemorrhagic Escherichia coll
0157-H7 in bovine feces. Appl. Environ.
Microbiol. 62:2567-2570.
Wang, I., Jury, W.A., Tuli, A. and Kirn, DJ.
2004. Unstable flow during redistribution:
controlling factors and practical implica-
tions. Vadose Zone J, 3:549-559.
Wang, Z., Wu, L., Harter, T., Lumia, D.S. and
Jury, W.A. 2003. A field study of unstable
preferential flow during soil water redis-
tribution. Water Resour. Res. 39: 1075,
doi:10.1029/2001WR000903
Ward, M.H., deKok, T.M., Levallois, P.,
Brender, J., Gukus, G., Nolan, B.T. and
VanDerslice, J. 2005. Workgroup report:
drinking-water nitrate and health—recent
findings and research needs. Environ.
Health Perspect. 113:1607-1614.
Watson, K.K. 1966. An instantaneous profile
method for determining the hydraulic con-
ductivity of unsaturated porous materials.
Water Resour. Res. 2:709-715.
Watts, D.B., Torbert, H.A., and Prior, S.A.
2007. Mineralization of nitrogen in soils
amended with dairy manure as affected
by wetting/drying cycles. Communications
in Soil Science and Plant Analysis
38:2103-2116.
Waybrant, K.R., Ptacek, C.J. and Blowes,
C.W. 2002. Treatment of mine drainage
using permeable reactive barriers: column
experiments. Environ. Sci. Technol.
36:1349-1356.
Webb, J. and Archer, J.R. 1994. Pollution
of soils and watercourses by wastes
from livestock production systems, hi:
Pollution in livestock production systems.
CABI Publishing. Oxfordshire, UK. pp.
189-204.
Weiss, W.J., Bouwer, E.J., Aboytes, R.,
LeChevallier, M.W., O'Melia, C.R., Le,
B.T. and Schwab, K.J. 2005. Riverbank
filtration for control of microorganisms
results from field monitoring. Water Res.
39:1990-2001.
WHO, 2006, WHO guidelines for the safe
use of wastewater, excreta and greywa-
ter. Geneva, Switzerland, World Health
Organization.
Woodard, K.R., French, E.G., Sweat, L.A.,
Graetz, D.A., Sollenberger, L.E., Macoon,
B., Portier, K.M., Wade, B.L., Rymph, S.J.,
-------
Prine, G.M., and Van Horn, H.H. 2002.
Nitrogen removal and nitrate leaching for
forage systems receiving dairy effluent. J.
Environ. Qual. 31:1980-1992.
Yao, K.M., Habibian, M.T., and O'Melia, C.R.
1971. Water and waste water filtration—
concepts and applications. Environ. Sci.
Techno!. 5:1105-1112.
Yates, M.V., Yates, S.R., Wagner, J. and Gerba,
C.P. 1987. Modeling virus survival and
transport in the subsurface. J. Contam.
Hydrol. 1:329-345.
Yeoman, S., Stephenson, T., Lester, J.N. and
Perry, R. 1988. The removal of phospho-
rus during wastewater treatment: a review.
Environ. Pollut. 49:183- 233.
Zevi, Y., Dathe, A., McCarthy, J.F., Richards,
B.K. and Steenhuis, T.S. 2005.
Distribution of colloid particles onto inter-
faces in partially saturated sand. Environ.
Sci. Techno!. 39:7055-7064.
Zhang, R.H. and Lei, F. 1998. Chemical
treatment of animal manure for liquid
separation. Trans. ASAE. 41:1103-1108.
Zhang, R.H. and Westerman, P.W. 1997. Solid-
liquid separation of animal manure for odor
control and nutrient management. Appl.
Eng. Agric. 13:657-664.
-------
-------
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGES FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
Recycled/Recyclable
Printed with vegetable-based ink on
paper that contains a minimum of
50% post-consumer fiber content
processed chlorine free
------- |