Photo by NRCS Phosphorus in Agricultural Watersheds A Literature Review by G. N. Zaimes, R. C. Schultz Department of Forestry, Iowa State University, Ames, Iowa January 2002
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Phosphorus in Agricultural Watersheds
A Literature Review by
G. N. Zaimes, R. C. Sch Department of Forestry, Iowa State
University, Ame
January 2002
Table of Contents
Executive Summary v
1. Context of Problem 1 1.1 Agriculture’s contribution of nonpoint
source phosphorus to surface waters 1 1.2 Phosphorus importance for
agriculture 3 1.3 Phosphorus negative impacts on surface waters 3
1.4 Total maximum daily loads program 4
2. Nonpoint Source Phosphorus Transport To Surface Waters 5
2.1 Terrestrial phosphorus processes 5 2.1.1 Phosphorus forms in
water 6
2.1.1.1 Importance 6 2.1.1.2 New systematic nomenclature 8
2.1.2 Transport Mechanisms 9 2.1.3 Pathways 10
2.1.3.1 Importance 10 2.1.3.2 Classification 10 2.1.3.3 Major
pathways 10
2.1.4 Hydrology 16 2.1.5 Terrestrial sources of phosphorus for
surface waters 17
2.1.5.1 Atmospheric Deposition 17 2.1.5.2 Soil sources 18 2.1.5.3
Agronomic sources 27
2.2 Aquatic Phosphorus Processes 30 2.2.1 Aquatic processes in
lotic systems 30 2.2.2 Flow regimes 32 2.2.3 Critical phosphorus
concentrations for surface waters 33
3. Land-use Practices and Best Management Practices 36
3.1 Land-use practices 36 3.2 Pastures 37 3.3 Cultivated Fields 41
3.4 Reducing phosphorus inputs to surface waters 45 3.5 Riparian
buffers 46
3.5.1 Filter Strips 49 3.5.2 Riparian forest buffer 49 3.5.3
Riparian management system 50
4. Other factors that influence phosphorus transport 53
4.1 Different Scale categories 53 4.2 Watershed scale studies
53
5. Phosphorus-Index System 55
6. Conclusion 60
7. References 62
APPENXIX I Case Studies: Phosphorus Losses Under Different Land-use
Practices 78
Appendix II -Table 1. Phosphorus losses under different land-use
management’s through different pathways and various P forms
79
Appendix I - References 98
APPENDIX II Original Phosphorus-Index 101
Appendix II -Table 1. The site characteristics, their weight
factors, and the five P loss levels of the
original P index system 102
Appendix II - Table 2. The site vulnerability chart for the
original P index system that indicates the potential of a site to
deliver P to surface waters 103
Appendix II - References 104
Acknowledgement This review was prepared with the support of the
Iowa DNR through a grant from USEPA under the Federal Nonpoint
Source Management Program (Section 319 of the Clean Water
Act).
i
List of figures
Figure 1. Phosphorus transport includes terrestrial and aquatic
processes. 5 Figure 2. Potentially mobile agricultural P inputs and
the hydrologic pathways that transport P to
reach surface waters. 6 Figure 3. A model describing how NPS P
reaches surface waters from terrestrial sources. Hydrology provides
the energy for P transport and the soil, atmospheric deposition,
and agronomic practices, are the sources. Phosphorus transport is
initiated by three mechanisms and P can reach the surface waters
through one or more of these pathways. 7 Figure 4. Hydrological
pathways for P transport in the (a) soil profile include matrix
and
preferential flow (b) slope/field scales include overland flow,
interflow and land drainage (e.g. tiles). 12
Figure 5. Mechanisms and forms of P transported in overland flow.
13 Figure 6. Phosphorus cycle in the soil with inputs, outputs, and
transformations that take place in
the soil. 20 Figure 7. Relationship between soil P test, crop
yield, and environmental problems due to
excessive soil P. 23 Figure 8. Percentage of soil samples testing
high P soil levels in 1989. 24 Figure 9. Phosphorus cycle in the
aquatic systems. In this cycle P moves from the water in the
bed and from the bed to the water while also transforming into
different forms. Terrestrial and streambank sources provide the
inputs. 31
Figure 10. The fourteen ecoregions in the United States. The
ecoregions were differentiated based on geology, land-use,
ecosystem type, and nutrient conditions. 35
Figure 11. Phosphorus concentrations (µg L-1), losses (g ha-1 yr-1)
and forms (TP (<0.45) and TP
(>0.45)) in surface runoff under different land-uses. 36 Figure
12. Nonpoint source P contributors in 1980. Cultivated, pasture and
range land are the
major contributors. 37 Figure 13. Phosphorus losses and its forms
RP (<0.45) RP (>0.45), and UP (unf.) in overland flow
as a function of different crops and different tillage practices on
several Southern Plain watersheds, averaged over five years.
43
Figure 14. Nontilled row-cropped fields have more overland flow and
less total
evapotranspiration in larger flow compared to riparian buffers that
reduce stormflow and increase baseflow due to higher infiltration
and evapotranspiration. Reducing overland flow and streambank
erosion will decrease P losses. 46
Figure 15. Phosphorus movement in riparian forest buffers. Sediment
and TP (>0.45) are filtered
from overland flow and TP (<0.45) can be taken up by biota of
the living filter. 47
ii
Figure 16. The riparian forest buffer. The first zone consists of
unmanaged forest, which protects the streambank and provides woody
debris. In the second zone, the forest is managed to maintain
nutrient uptake through vigorous plant growth. The third zone has
grasses with controlled grazing allowed under certain conditions.
50
Figure 17. The multi-species riparian buffer model. The first zone,
is located along consists of a managed tress and shrubs. They
provide bank stability, wildlife habitat, are a nutrient and
sediment sink, and modify the aquatic environment. The second zone,
consist of native grass and forbs that intercept overland flow,
increase infiltration and intercept NPS pollutants. 51
Figure 18. The Riparian Management System (RiMS) consists of five
practices: i) multi-species
riparian buffer with woody plants and warm-season grasses that
intercept NPS from adjacent land practices, ii) streambank
bioengineering that provide bank stability, iii)
constructed-restored wetlands that intercept and filter NPS from
subsurface tiles, iv) cool-season or warm-season grasses can
replace MRB that may be used for rotational grazing with stream
fenced out and v) instream structures like boulder weirs. 52
iii
List of tables Table 1. Percentage of NPS P inputs from agriculture
to surface waters for different countries. 2 Table 2. Limiting
nutrients for various surface water bodies. 4 Table 3.
Classification of P forms in water with the new suggested
classification and the
currently established terms. 9 Table 4. Definitions for hydrologic
pathways of P in the unsaturated zone. Terms are sorted
alphabetically. When there is more than one term describing a
pathway, the preferred term is indicated in italics. This
classification is based on spatial and temporal scales and planes
of water movement. Please note that this is a nominal
classification. 11
Table 5. Phosphorus concentrations in stream and drainage water
under various flow regimes.
Ranges are in the parenthesis when available. 17 Table 6. The
common chemical forms of P in soil and their characteristics, or
implication for
potentially mobile P. 19 Table 7. Soil P surpluses for different
developed countries under different agronomic land-use
practices. 22 Table 8. Percentage of TP (unf.) of A, B and C
horizons of several kinds of soil. Methods used
to measure TP (unf.) in the soil were not described. 24 Table 9.
Recommended agronomic and environmental soil P test threshold
values with the
appropriate P management recommendations. 26 Table 10. The amounts
of feces and phosphorus produce by different animals. 27 Table 11.
Phosphorus concentrations from headwaters to downstream reaches. 31
Table 12. Recommended critical or threshold P concentrations for
surface waters that may cause
eutrophication. 34 Table 13. Changes in physical and chemical
properties from the headwaters to the downstream
reaches and some biological implications. 34 Table 14. The EPA
recommended critical TP (unf.) concentrations for each some of the
nutrient
ecoregions for phosphorus. The ecoregions are shown in Figure 10.
34 Table 15. Phosphorus removal rates for different crops.
Phosphorus removal (kg ha-1) is
estimated by the mean P concentration and the mean yield of the
crop for the United States. All values are expressed on a fresh
weight basis. 42
Table 16. The effectiveness of different types of buffers with
different widths in removing NPS
P from overland flow. 48
iv
Table 17. The transport and source characteristics with their
respective weight factors, and the five P loss-rating levels of the
modified P index system. The transport include soil erosion, soil
runoff class and return period/distance. The source characteristic
include soil P test, P fertilizer rate, P fertilizer application
method, organic P application rate, and organic P application
method. 56
Table 18. The site vulnerability chart of the watershed-modified
P-index system that indicates the
potential of a site to deliver P to surface waters. 57 Table 19.
The site vulnerability rating (P hazard class) for the Iowa
P-index. 59
v
Executive Summary Phosphorus is a major nonpoint source pollutant
that causes eutrophication in surface waters. In the past interest
in phosphorus, as a nonpoint source pollutant was not as great as
for nitrogen because phosphorus is generally less mobile than
nitrogen in the agricultural landscape. Phosphorus is immobile
because it is easily adsorbed to soil particles. However, high soil
and streambank erosion can lead to increased amounts of phosphorus
in surface waters. Additionally, dissolved phosphorus contributions
are more significant than previously thought. These facts along
with the heightening concern of the impacts of poultry and
livestock manure on surface water quality has increased the
interest in phosphorus movement and management in the landscape. To
be able to assess the potential of nonpoint source phosphorus
pollution and to develop proper management strategies to reduce
phosphorus losses to surface waters it is essential to understand
the major processes involved in phosphorus transport. Both
terrestrial and aquatic processes are important. Terrestrial
processes are responsible for phosphorus inputs from upland and
riparian areas while aquatic processes include streambank inputs
and instream phosphorus inputs, outputs and transformations. A
model describing phosphorus transport from terrestrial sources
would include: the energy for transport, the specific sources, the
form and amount of phosphorus that is potentially mobile, the
mechanisms of transport and finally the specific pathways that
phosphorus will follow.
Precipitation provides the major source of energy for transport
through its effect on soil erosion and is dependent on watershed
morphology, hydrology and land cover and management. The major
sources of phosphorus include atmospheric deposition (wet and dry
fall), the soil (the phosphorus levels in the soil), and various
agronomic inputs (fertilizer, manure, animal litter, animal feces
and plant residue).
Dissolved, particulate, organic and inorganic phosphorus are the
major terms used to describe various forms of phosphorus in the
environment. These terms describe the physical and chemical
conditions of phosphorus and are ambiguous because each term may
represent various different forms depending on the methods used to
extract and analyze them. As a result a new systematic nomenclature
has been suggested to describe phosphorus forms that incorporates
the methods used to differentiate them.
Dissolution, physical and incidental are mechanisms that are
responsible for the transport of phosphorus. Dissolution is the
transport of dissolved phosphorus as in soil leaching. Physical
mechanisms refer to the transport of particulate phosphorus as in
soil erosion, while incidental mechanisms are controlled by unique
conditions like the instantaneous, short-term transport of surface
applied phosphorus fertilizer after an intense rainfall.
The major pathways that moving phosphorus follows include overland
flow, matrix flow, preferential flow, interflow and land drainage
(ditches, tiles and moles). A model for aquatic processes includes
phosphorus inputs from streambank erosion, phosphorus outputs from
rivers and streams and phosphorus cycling within the aqueous
ecosystem. Phosphorus outputs from rivers and streams originate
from terrestrial inputs plus the addition of phosphorus from
streambank erosion and resuspension from the streambed. Instream
cycling includes phosphorus transformations between dissolved
phosphorus (more readily bioavailable) and particulate
phosphorus
vi
(less bioavailable) that regulate phosphorus bioavailability to
organism in the water column and the hyporheic zone. Additionally,
it is very important to recognize the impact of flow regimes on
aquatic processes and the critical phosphorus concentrations needed
for eutrophication. Land-use practices can have a major impact on
phosphorus losses to surface waters. The two major agricultural
land-use practices that contribute phosphorus to surface waters are
cultivation and grazing. Both increase phosphorus as they become
more intensive. Grazing increases phosphorus losses because of the
addition of phosphorus from animal feces, and trampling that can
increase overland flow, peak discharges and streambank erosion.
Cultivation in row-crop agriculture increases phosphorus losses
because much of the soil surface is bare and susceptible to direct
raindrop impact, is compacted from machinery and surface sealing
and may have added fertilizer and/or animal wastes. Best management
practices can reduce phosphorus losses, by reducing overland flow
and streambank erosion and should be concentrated in riparian areas
that may be the major contributors of phosphorus to surface waters
if they are cultivated or heavily grazed. Riparian areas should be
the focus for best management practices because the sediment that
is detached and moved in these areas has a greater probability of
reaching the stream than sediment that is detached in the uplands.
Best management practices in the riparian area include filter
strips, riparian forest buffers, and riparian management systems.
To further understand the processes and pathways of phosphorus
movement and the impacts of land-use on phosphorus losses, research
is needed at different scales and in different regions of the
country. More watershed scale studies should be conducted. However,
their costs are very high and results from more field and plot
scale studies should be used to extrapolate to the watershed scale
with the use of models.
The phosphorus index was developed by the USDA-NRCS to assess
potential losses at the filed scale. It has since been modified to
be used at the watershed scale. Additionally, most states,
including Iowa, have modified the original index for use at the
field scale in their respective states.
Nonpoint source phosphorus pollution is not an easy problem to
solve. It will take a combination of research, the modification of
existing and development of new best management practices and
rigorous application of those best management practices on the
landscape.
1
1. CONTEXT OF PROBLEM
1.1 Agriculture’s contribution of nonpoint source phosphorus to
surface waters
Nonpoint source (NPS) pollutants are the number one water quality
problem in the United States (USEPA, 1997a). The most common NPS
pollutants are sediment and nutrients (nitrogen (N) and phosphorus
(P)). Agriculture is the nation’s leading NPS contributor, and is
responsible for degrading 60% of the impaired river kilometers and
50% of the lake hectares in the country (USEPA, 1997a).
Cultivation, fertilizer and pesticide application, irrigation,
planting, harvesting, confined animal facilities, and grazing are
the major agricultural activities that cause NPS pollution (USEPA,
1997b). Agricultural activities in the U.S. annually contribute1.9
billion metric tons of sediment (USDA-NRCS, 1997a). In Iowa,
agriculture contributes 28.3 metric tons of sediment ha-
1 yr-1 (sheet and rill erosion and ephemeral gullies) 9.8 kg N ha-1
yr-1 and 0.3-2.0 kg P ha-1 yr-1 (USDA-NRCS, 1997b).
This review will concentrate on P as a NPS pollutant from
agricultural sources. Past perceptions were that NPS P movement to
surface waters was minimal, because P was primarily held by soil
particles. However, studies at both the plot and watershed scale
show evidence of significant P losses from agricultural fields
(Sharpley et al., 2000; Withers and Jarvis, 1998). These losses,
typically are below 1 kg total P ha-1 yr-1, and although negligible
from an agronomic point of view, can have a significant
environmental impact (Heckrath et al., 1995). In Illinois,
agricultural P contributions more than doubled in areas with
smaller urban population densities as P sewage effluent
contributions decreased, although riverine P loads were relatively
similar to the P loads in the higher urban population density areas
(David and Gentry, 2000).
Powlson (1998) mentions three factors that have led to a greater
realization of the importance of P movement from agricultural
fields. The first is that very low P concentration can cause
eutrophication. The second is that significant amounts of P losses
are through interflow and field drains that reach surface waters.
The third is a greater recognition of the importance of soil
erosion and overland flow to P transport. Even though P is strongly
held by soil, surface erosion mobilizes P when soil particles are
detached (Addiscott et al., 2000).
More evidence of the importance of agricultural P inputs to surface
waters is that the removal of point sources does not significantly
reduce P inputs to surface waters (Sharpley et al., 2000). In lakes
in Denmark and Ireland, reduction of point sources did lower P
concentrations but without any appreciable improvement in water
quality (Foy et al., 1995; Kronvang et al., 1993). This was
attributed to P inputs primarily from agricultural sources. The
easier identification and recent control of point sources has led
to the belief that NPS of P in agricultural runoff now contributes
a greater proportion to surface water inputs than point sources
(Sharpley et al., 1994) (Table 1).
The importance of agricultural P contributions to surface waters is
recognized in management plans for improving water quality. In
Illinois, regional rivers contributed a large fraction of the P
loads to the Mississippi River. To reduce these P loads, a
reduction of P from agricultural sources has to be addressed (David
and Gentry, 2000). To improve drinking water quality, New York City
decided to implement best
2
Table 1. Percentage of NPS P inputs from agriculture to surface
waters for different countries. Country Agriculture Share Reference
Year % Netherlands 24 1985 e Italy 33 1986 e Federal Republic of
Germany 38 1987 e North Sea catchment basin 25 1987 e Sweden 16
1987 e Denmark 70 1988 f Norway 27 1988 e United Kingdom 35 1995 d
U.S.A. 84 1998 c New Zealand 90 2000 b U.S.A., Illinois Illinois
River All other rivers
53 30 67
2000 a
a David and Gentry (2000), b Gilligham and Thorrold (2000), c
Carpenter et al. (1998), d Heckrath et al. (1995), e Isermann
(1990), f Chiaudani and Premazzi (1988). management practices
(BMP’s) to reduce NPS P from agricultural sources rather than build
a new water sewage treatment facility (Sharpley et al.,
2000).
Agricultural P inputs to surface waters have increased because of
intensive livestock grazing and because the combined fertilizer and
manure inputs in excess of crop requirements that have led to a
build up of soil P levels (Sharpley et al., 1994). These practices
have altered the soil and landscape hydrology. The final amount of
P entering a stream depends on agricultural management practices,
the soil type and its associated chemical, physical and biological
characteristics, the type of the runoff events and concentrations
of P in the runoff, and the nature of the receiving water (Abrams
and Jarrell, 1995). Because so many factors influence P inputs,
this literature review will concentrate on identifying agricultural
P sources, pathways and processes that control P transport, their
influence on P availability and their potential impact on surface
water quality (Edwards and Withers, 1998; Johnes and Hodgkinson,
1998).
Although research on NPS P has been conducted for the last 25
years, more studies are necessary to identify the environmental
impacts from agricultural practices (Johnson et al., 1997).
Sharpley and Tunney (2000) suggest four main areas of research: 1)
soil P testing for environmental risk assessment; 2) pathways of P
transport within a watershed: 3) BMP’s development and
implementation; and 4) strategic initiatives to manage P. They go
on to say that future research should be interdisciplinary and
involve soil scientists, hydrologists, agronomists, limnologists,
animal scientists, rural economists, and social scientists. In
addition, David and Gentry (2000) suggest more research is needed
to correlate P loads in streams to P loads from agricultural
production.
3
1.2 Importance of phosphorus in agriculture
Phosphorus is an essential macronutrient that is required to meet
global food requirements and make crop and livestock production
profitable (Hedley and Sharpley, 1998). The most important function
of P in plants is the storage and transfer of energy, and cell
division (Norfleet, 1998; Troeh and Thompson, 1993). Plant cells
need to have adequate P before they divide. Additionally, P
increases seed production, root growth, grain, fiber and forage
yield, enhances early plant maturity and stalk strength, and
promotes resistance to root rot disease and winter kill (Norfleet,
1998).
Phosphorus is the mineral with the most known biological functions
in animals (Beede and Davidson, 1999). Phosphorus is involved in
most energy transactions, in the acid-base buffer systems of the
blood and other fluids, and in cell differentiation of animals.
Bones and teeth contain almost 80% of the P found in the bodies of
animals. Every cell in the body also contains P as phospholipids,
phosphoproteins and nucleic acids. Growing cattle contain 6-8 mg
dl-1, while mature cattle contain 4-6 mg dl-1 in their blood
plasma. Whole blood contains 6-8 times more P than plasma.
1.3 Negative impacts of phosphorus on surface waters
Eutrophication has been recognized as the main cause of water
quality impairment (Environment Agency, 1998; European Environment
Agency, 1998; USEPA, 1996). Phosphorus has been identified as the
primary limiting nutrient causing eutrophication of many surface
waters (Table 2) (Daniel et al., 1998). Eutrophication increases
the growth of undesirable algae and aquatic plants that replace
benthic organisms and submerged macrophytes (Sharpley et al., 2000;
Carpenter et al., 1998; Norfleet, 1998; Kotak et al., 1994; Martin
and Cooke, 1994). The death and decomposition of algae cause oxygen
shortages that restrict water use for fisheries, recreation,
navigation, industry and drinking. Coarse, rapid-growing fish
replace high- quality edible fish. Undesirable odors and surface
scum are produced from the decaying algae, with mosquito and other
pest insect population increasing. Water transparency and the
aesthetic value of the surface water are decreased. Potentially
toxic dissolved compounds are produced that may harm livestock,
wildlife, and the cost of purification increases. Finally,
cyanabacteria blooms can lead to drinking water
unpalatability.
Land management practices can be used to effectively control the
movement of P into surface waters. Phosphorus transport is easier
to control than nitrogen and carbon (C), two other nutrients
involved in creating water quality problems (Sharpley et al.,
2000). It is very difficult to control the exchange of N and C
between the atmosphere and water, the fixation of N by some
blue-green algae, and the mobility of N in surface and subsurface
flow (Hession and Storm, 2000). Although P transport is easier to
control than N transport, very small P concentrations (as low as 10
µg L-1) cause eutrophic and hypereutrophic conditions (Sharpley et
al., 2000; Haygarth et al., 1998a; Powlson, 1998). Considering that
eutrophication exhibits high spatial and temporal variability and
that the natural background levels can be a significant contributor
of NPS P (28.4%) (Council of Environmental Control, 1989), solving
P pollution problems will not be simple (Edwards and Withers,
1998). Natural background levels of P refer to the P that would be
found in
4
Table 2. Limiting nutrients for various water bodies. a System N/P
ratio Limiting nutrient
Rivers and Streams Point source dominated Without P removal With P
removal
<<10 >>10
N P
Fresh water region Point source dominated Nonpoint source
dominated
>>10 <<10
P N
Brackish region ≈10 P or N Saline region <<10 N
Lakes Large Nonpoint source dominated
>>10
P
a Sharpley et al. (1994).
a water body in an undisturbed system (absence of human-induced
change) (Salloway, 2001; Taylor and Kilmer, 1980).
1.4 Total maximum daily loads program The United States
Environmental Protection Agency (USEPA) has developed the Total
Maximum Daily Loads (TMDL) program to regulate water pollution.
This program is based on section 303(d)(1) of the Clean Water Act
(CWA) (USEPA, 2000a). Under this program, States provide a list of
all their impaired waters to the EPA, and develop TMDL for sediment
and nutrients (N and P) for their impaired waters from point
sources and NPS. The USEPA (1999) defines TMDL as “ a calculation
of the maximum amount of a pollutant that a water body can receive
and still meet water quality standards, and an allocation of that
amount to the pollutant’s sources.” TMDL standards are based on
scientific criteria and the waterbody use (e.g. drinking, swimming,
fishing). The TMDL programs in the future will also require
implementation programs based on scientific data (USEPA, 2000b).
The TMDL program will have a significant impact on P management in
agricultural practices.
5
2. NONPOINT SOURCE PHOSPHORUS TRANSPORT TO SURFACE WATERS
To identify and quantify NPS P loads to water bodies it is
necessary to understand the
processes and pathways that control P transport (Johnes and
Hodgkinson, 1998). Phosphorus transport can be divided into
terrestrial and aquatic processes (Figure 1). Terrestrial processes
describe P inputs to surface waters from the upland and riparian
NPS sources, such as natural, agriculture and urban ecosystems, and
point sources (Pierzynski et al., 2000). Aquatic processes describe
P inputs from streambank erosion, P outputs from rivers, streams
and other water- courses, and the P cycling within the aqueous
ecosystem that regulates P bioavailability to aquatic organisms
(Pierzynski et al., 2000). This review will focus on how
agriculture P inputs reach and move through lotic systems. However,
it is very difficult to quantify, with any degree of certainty, the
terrestrial and aquatic processes (Edwards and Withers,
1998).
2.1 Terrestrial phosphorus processes
Terrestrial processes carry P above and below the soil surface
before it reaches surface waters (Figure 2). Potentially mobile P
describes P in the terrestrial landscape that is in forms that can
be transported by water and is a nonquantitative concept (Haygarth
and Jarvis, 1999).
Figure 1. Phosphorus transport includes terrestrial and aquatic
processes (Pierzynski et al., 2000).
6
The main sources of the potentially mobile P are fertilizers,
manure, including that in pastures and that associated with
confinement facilities, animal litter, plant residue, soil P and
atmospheric deposition (Figure 2). The potentially mobile P that is
actually transported by the various hydrologic pathways to surface
waters is called total transported P. Nonpoint source P transported
through these hydrologic pathways is accomplished by three
mechanisms (incidental, physical, dissolution) with hydrology the
main driving force (Figure 3). To better understand these
mechanisms the forms of P in water will be described first. 2.1.1
Phosphorus forms in water
2.1.1.1 Importance
The distinction between P forms is necessary because of their
differences in adsorption- desorption reactions, transport, and
potential bioavailability to aquatic organisms (Edwards and
Withers, 1998). For example, dissolved P primarily consists of
orthophosphate that is immediately available for algal uptake while
particulate P is a long-term source of P for aquatic biota
(Sharpley et al., 1994). A better understanding of P forms and the
mechanisms by which they are exported to surface waters is required
to minimize transport (Haygarth et al., 1998a).
Transport of P in the terrestrial environment can be via solution
or with sediment movement (Sharpley and Mendel, 1987). Dissolved
(soluble) P in runoff originates from the
Riparian zone
Agricultural crop
Mineralogy
Soil P
Plant residue
Matrix flow
Soil processes
Preferential flow
Hydrological pathways
Figure 2. Potentially mobile agricultural P inputs and the
hydrologic pathways that transport P to reach surface waters.
7
Hydrology
ENERGY
Potentially Mobile
MECHANISMS
P transported to surface waters
THE PROBLEM
Particulate P Dissolved P
Figure 3. A model describing how NPS P reaches surface waters from
terrestrial sources (modified from Haygarth and Sharpley, 2000).
Hydrology provides the energy for P transport and the soil,
atmospheric deposition, and agronomic practices, are the sources.
Phosphorus transport is initiated by three mechanisms and P can
reach the surface waters through one or more of these
pathways.
8
release of P from a thin zone of surface soil (1-2.5 cm) and/or
vegetative material that interacts with rainfall (Sharpley et al.,
1994; Sharpley et al., 1992; Sharpley, 1985). Dissolved P transport
depends on desorption-dissolution reactions (Sharpley et al., 1992;
Sharpley, 1985) and on the P content in surface soil (Sharpley at
al., 2000). Desorption, dissolution, and extraction of P from the
soil, crop residues, livestock manure or surface-applied fertilizer
lead to dissolved P movement in runoff. Particulate (suspended, or
sediment-bound) P is associated with soil and vegetative material
eroded during runoff. In most cases, particulate P is typically the
dominant form of P lost (David and Gentry, 2000; Vaithiyanathan and
Correll, 1992; Bottcher et al., 1981). In 116 agricultural
watersheds particulate P was on average 86% (ranged from 44-98%) of
the total P loads (kg yr-1) (Prairie and Kalff, 1986). However, to
get the full impact of P losses both dissolved P and particulate P
must be measured (Stevens et al., 1999).
Algal available P is dissolved P and the portion of particulate P
that is in equilibrium with dissolved P (Sharpley et al., 1996).
Sonzogni et al. (1993) defined algal available P as “the amount of
inorganic P a P-deficient algal population can utilize over a
period of 24 h or longer.” Algal available particulate P is
important because the excess of this form will cause
eutrophication. Algal available particulate P is a function of soil
loss, particle size enrichment, and chemical properties of the
eroded material that governs P adsorption and availability.
Reduction in total P loads may not lead to reduction of eutrophic
conditions in many cases, because it only reflects a decrease in
particulate P (Sharpley et al., 1992). Instead, the measurement of
algal available P is essential to accurately estimate the impact of
agricultural management practices on surface waters (Sharpley et
al., 1992).
2.1.1.2 New systematic nomenclature The characterization of the
various forms of P associated with water transport depends on
the filtration and chemical methods of analysis. Dissolved P and
particulate P is differentiated with filtration methods while the
inorganic P and organic P is differentiated with chemical methods.
There are many different methods that are used to differentiate P
fractions. Because the terms in the above section describe the
physical and chemical conditions of P and not the methods that are
used to differentiate the P fractions they are ambiguous. For
example to differentiate dissolved P and particulate P different
size filter membranes might be used. In this case the dissolved and
particulate fractions would differ depending on the filter membrane
size.
A new systematic nomenclature has been suggested based on the
operational definitions of the filtration and chemical methods
(Table 3). For P forms described by the filtration methods,
Haygarth and Sharpley (2000) suggest that samples be defined by the
filter size, with a suffix of the filter pore size or the term
(unf.) when they are not filtered (Table 3). Typically, a 0.45 µm
membrane filter is used. According to Haygarth and Sharpley (2000),
P that passes the 0.45 µm membrane filter is defined as dissolved
or soluble P (TP (<0.45)). The P that does not pass through the
membrane is the particulate, sediment-bound or suspended P (TP
(>0.45)).
From the chemical methods the Mo-blue reaction method (Murphy and
Riley, 1962), is most commonly used to estimate orthophosphate (an
inorganic P form). In reality the preparation of the sample causes
this method to overestimate orthophosphate. Comparing
orthophosphate estimated by the Mo method would be different from
estimates from other methods. The new terms for P, suggested by
Haygarth and Sharpley (2000), based on the Mo reaction method are
reactive (RP), unreactive (UP) and total P (TP) (Table 3).
Specifically, RP is reactive with Mo, while UP is unreactive with
Mo, and TP is the sum of RP and UP.
9
Table 3. Classification of P forms in water with the new suggested
classification and the currently established terms. a New
classification Established terms b
TP (unf.) Total P (TP) -from a unfiltered sample - TP (<0.45)
Total dissolved P (TDP), soluble P TP (>0.45) Particulate P
(PP), sediment-bound P,
suspended P RP (unf.) Total reactive P (TRP), UP (unf.) Total
organic P RP (<0.45) Molybdate-reactive P (MRP),
dissolved-
reactive P (DRP), soluble reactive P (SRP), dissolved
molybdate-reactive P, orthophosphate, inorganic P, phosphate
UP (<0.45) Dissolved organic P (DOP), soluble organic P (SOP),
dissolved nonreactive P (DNRP)
RP (>0.45) Molybdate-reactive particulate P (MRPP), particulate
reactive P
UP (>0.45) Particulate organic P a Modified from Haygarth and
Sharpley (2000). b May not necessarily be inclusive.
Finally, Haygarth and Sharpley (2000) recommend that the terms
bioavailable P and available P should not be used because of their
inherent value judgment. Instead, P should be described by the
organism to which it is bioavailable. For example, for P in surface
waters that can potentially cause eutrophication the proper term is
algal-available P (AAP). 2.1.2 Transport Mechanisms The three main
P transport mechanisms are dissolution, physical, and incidental
(Haygarth and Jarvis, 1999). Dissolution describes the transport of
TP (<0.45) from the soil particle or adsorption site to the soil
solution. Dissolution is a micro-scale soil profile mechanism that
is determined by chemistry. Examples of dissolution are
mineralization, enzyme hydrolysis, adsorption-desorption,
solubilization of P from saturated soils, and leaching (Haygarth
and Sharpley, 2000; Haygarth and Jarvis, 1999). Leaching is the
elluviation of solutes through soil. Although it is a process term,
it commonly is used inappropriately to denote a pathway. Adsorption
and desorption continue to occur once the TP (<0.45) is in
solution (Sharpley et al., 1994). The magnitude and direction of P
transformations depend on TP (<0.45), TP (>0.45) and sediment
concentration in the runoff. Total P (>0.45) is primarily moved
by physical mechanisms. As the intensity of the physical mechanism
increases, TP (>0.45) concentration in runoff increases
(Sharpley et al., 1992). In contrast to dissolution, this mechanism
is a macro-scale process (Haygarth and Jarvis,
10
1999). Examples are soil erosion, and displacement and entrainment
of colloids and submicron- sized material (Haygarth and Sharpley,
2000).
Fertilizer and manure P inputs to the soil are retained by smaller
particles, so the added P is not redistributed uniformly through
the whole soil profile (House et al., 1998). Soil P levels are
higher in the top 5 cm of the surface soil. Soil detachment and
transport in surface runoff preferentially erode finer particles.
This results in eroded material with higher TP (>0.45) content
in the runoff compared to the soil in the source area (Sharpley et
al., 2000). This phenomenon is referred to as enrichment. The
smaller, lighter particles are transported greater distances and
are more likely to enter surface waters (House et al., 1998).
Because of the selective erosion of finer material from the surface
few centimeters, TP (>0.45) losses are much larger than that
predicted from whole soil profile analysis (House et al.,
1998).
The incidental mechanism is conceptually different than the
dissolution and physical mechanisms, because it is controlled by
instantaneous unique conditions (Haygarth and Jarvis, 1999).
Examples are short-term transport of farm amendments like P
fertilizer, manure or animal feces. This short-term transport
occurs after effective rainfall removes the amendment shortly after
it has been applied. Land-use management factors and hydrology have
major impacts on all three P transport mechanisms (Sharpley et al.,
2000). 2.1.3 Pathways
2.1.3.1 Importance To effectively mitigate P transport to surface
waters the pathways of P movement must
be identified (Sharpley et al., 2000). The pathways that P follows
depend on the form and quantity of P that is transported to the
surface waters (Sharpley and Syers, 1979a; Foy and Withers, 1995).
For example, RP (unf.) dominates in surface pathways because it
reacts with the upper surface horizons (Haygarth et al., 1998a). In
contrast, more than 50% of the P in subsurface (tile) drained flow
is UP (unf.) because it moves more freely in the soil
profile.
2.1.3.2 Classification Classification of the hydrologic transport
pathways is difficult because of the large
variability that exists in the spatial and temporal scale of water
flow. The latest classification of pathways, by Haygarth and
Sharpley (2000), is organized around spatial and temporal
characteristics (Table 4). The spatial characteristics considered
are the plane and scale of water movement. The plane refers to the
vertical and lateral direction of water flow. The spatial scale is
divided into: i) the soil profile (commonly in the vertical plane),
ii) the slope/field (commonly in the lateral plane), and iii) the
watershed/catchment (the largest scale) (Figure 4). The temporal
scale can be minutes, hours, or days depending on local
conditions.
2.1.3.3 Major pathways The most important pathways for P transport
from agricultural landscapes are (Figure 4):
i) matrix flow, ii) preferential flow, iii) overland flow iv)
interflow, and v) land drainage. These P pathways are interrelated
during an event. For example, the pathway down a slope can be
11
Table 4. Definitions for hydrologic pathways of P in the
unsaturated zone. Terms are sorted alphabetically. When there is
more than one term describing a pathway, the preferred term is
indicated in italics. This classification is based on spatial and
temporal scales and planes of water movement. Please note that this
is a nominal classification. a Term Scale Time Plane Definition
By-pass flow Soil Min/h Vertical See preferential flow. Interflow
Slope/field Min/h Lateral Lateral flows below the soil surface.
Land drainage Subcatchment
& Slope/field Min/h Lateral Water and solute (+entrained
solids) export to
catchment resulting from land drainage practices:
anthropogenic.
Leakage Slope/field Not applicable Both General non specific term
for describing water and chemical movement.
Macropore flow Soil Min/h Vertical See preferential flow; must be
tightly defined when used, otherwise best avoided.
Saturated (soil) flow
Soil Days Lateral See matrix flow, but the plane is lateral not
vertical.
Seepage Slope/field Not applicable Lateral General nonspecific term
describing water sampled from the soil environment by whatever
means-not a pathway.
Soil solution Soil Not applicable Both Nonspecific term describing
water movement, implies emergence at or near the ground
surface.
Subsurface flow Slope/field Min/h Lateral General lateral flow
below the soil surface. Surface runoff Slope/field Min/h Lateral
See overland flow. Throughflow Soil and
slope/field Not applicable Both See percolating water.
Unsaturated flow
Slope/field Min/h Lateral See preferential flow, but occurring
laterally over capped, compacted, or slowly permeable
horizons.
Matrix flow Soil Days Vertical Implies a type of soil water
movement-in this case uniform vertical movement downward, common in
very porous media, such as sandy textured soils.
Overland flow Slope/field Min/h Lateral Movement of water
exclusively over the soil surface during heavy rain.
Percolating water
Soil Not applicable Both General nonspecific term describing water
movement.
Pipe flow Slope/field Min/h Lateral Lateral subsurface preferential
flow. Piston flow Soil Not applicable Vertical See for matrix flow.
Preferential flow Soil Min/h Vertical Implies a type of soil water
movement-in the
case of vertical movement along larger subsoil pathways, e.g.,
wormholes and fissures, often occurring in unsaturated
conditions.
Return flow Slope/field Min/h Lateral Where a subsurface flow
pathway emerges at the soil surface.
Runoff Slope/field Min/h Lateral General hydrological term
describing the lateral movement of water off land above- and below-
ground, causing a short-term increase in flow at the catchment
outlet. Can refer to qualified pathway (e.g., surface runoff), but
also has been used to describe process and water media.
Vertical saturated flow
Vertical unsaturated flow
a Haygarth and Sharpley (2000).
12
Groundwater flow
Water table
Figure 4. Hydrologic pathways for P transport in the: (a) soil
profile including matrix and preferential flow (Haygarth and
Jarvis, 1999) and (b) slope/field scales including overland flow,
interflow and land drainage (e.g. tiles). either overland flow, or
deep and shallow interflow, with potential interchanges between
them as the degree of soil saturation, slope and infiltration
capacity changes (Johnes and Hodgkinson, 1998). Runoff, a very
common term used as a hydrologic pathway should not be used to
describe a specific pathway (Haygarth and Sharpley, 2000). If used,
it should be used in a very general sense to describe all the
pathways at the slope/field scale. The five pathways mentioned
occur in the unsaturated zone. Groundwater flow (saturated
conditions) can also be a major pathway by which water reaches
streams. Not many studies have looked at the P contributions
13
of groundwater flow. The first reason is that groundwater P
contributions are small compared to P transported in the
unsaturated zone (Peterjohn and Correll, 1984). However, Phillips
et al. (1982) found that RP (<0.45) concentrations in
groundwater did not change with depth and over a number of years
but were 10-20 µg L-1, concentrations that are high enough to cause
eutrophication. The second reason is that groundwater nutrients are
difficult to measure. The nutrients are typically transported in
anoxic and hypoxic environments and when the groundwater reaches
the surface waters it is oxygenated and the nutrients precipitates
in the surface sediments (Kalff, 2001).
i) Matrix flow: Matrix flow is the uniform downward movement of
water through the macro- and micro-pores of the soil. The scale is
the soil profile, the plane is vertical, and time is measured in
days. Phosphorus is immobile and easily adsorbed in the soil. This
is especially true for the subsoil horizons that are P deficient
(Sharpley et al., 2000). Typically, there is little movement of P
in the matrix flow pathway because the surface soil accumulates
most of it (Mozaffari and Sims, 1994; Oldham, 1998). Sandy, acid,
organic soil with low P fixation and holding capacities can have
increased P in matrix flow (Sharpley et al., 2000).
ii) Preferential flow: Preferential flow is also the downward
movement of water but in larger subsoil pathways, such as fissures
and cracks, burrows, and wormholes. The scale is the soil profile,
the plane is vertical, and time is measured in minutes or hours.
Water moves faster with less chance of adsorption and therefore
greater transport of TP (<0.45) and TP (>0.45) (Addiscott et
al., 2000). Preferential flow can occur in heavy, clay-rich soils
because of the cracks that develop in response to drying and
wetting, and where there is a heavy population of earthworms and
other burrowing organisms.
iii) Overland flow: Overland flow is the downslope movement of
water over the soil surface during heavy rainfall events (Figure
5). The scale is the slope or field, the plane is lateral, and time
is measured in minutes or hours. It has been traditionally
considered as the major transport pathway for P in agricultural
landscapes (Sharpley et al., 1993; Oldham, 1998).
Erosion of TP (>0.45)
Overland Flow TP (<0.45) and TP (>0.45)
Figure 5. Mechanisms and forms of P transported in overland flow
(Daniel et al., 1994).
14
Overland flow can carry P in solution (TP (<0.45)) and the P in
the eroded soil (TP (>0.45) (Sharpley et al., 2000), and is
generated by intense rainfall where infiltration capacities cannot
keep up with rainfall intensity (Johnes and Hodgkinson, 1998).
Although it contributes only periodically to stream flow, the
quantities of P losses are considerable (Sharpley et al., 1976).
Overland flow is efficient, first, because the largest
concentration of P is in the surface layers of the soil, and
second, because the greatest concentrated hydrologic energy is on
the soil surface (Haygarth et al., 1998a). Soil physical
properties, vegetation cover and slope steepness determine the
efficiency of this pathway (Sharpley and Syers, 1976a). Phosphorus
losses from overland flow can vary threefold as site hydrology,
relative drainage volumes, and soil P release characteristics
change (Sharpley and Tunney, 2000).
iv) Interflow: Interflow is the lateral movement of water in the
soil. The scale is the slope or the field, the plane is lateral,
and time is measured in minutes or hours. Typically, interflow P
losses are much lower than overland flow. For three corn-cropped
watersheds, the average RP (<0.45) losses from overland flow
were 0.105, 0.250 and 0.381 kg ha-1, respectively, while for
interflow the P loses for the same watersheds were 0.028, 0.030 and
0.031 kg ha-1 over a ten year period (Alberts and Spomer,
1985).
Stevens et al. (1999) found that interflow could not be ignored in
many cases. In their research, overland flow was the major pathway
for a soil with a shallow silty clay loam A horizon and a
medium-heavy clay loam B horizon. But for a soil with a sandy A
horizon and a heavy clay B horizon with low hydraulic conductivity,
interflow was the major pathway.
Interflow provides the major proportion of flow in many streams and
can be the major contributor of P, although it typically has low P
concentrations (Sharpley et al., 1976). Preferential flow and P
forms less susceptible to adsorption also may enhance P losses in
interflow (Heckrath et al., 1995). High soil P levels can also
increase P loss from interflow (Sharpley and Tunney, 2000).
Typically, sandy Spodosols (Sharpley et al., 1994), Histosols
(Izuno et al., 1991; Duxbury and Peverly, 1978) and poorly drained
soils in arable regions (Johnes and Hodgkinson, 1998; Deal et al.,
1986) can have significant P losses in interflow.
v) Land drainage: Land drainage includes a number of management
practices that result in increased water and solute export from
watersheds. The scale is the subcatchment, the plane is lateral,
and time is measured in minutes or hours. These practices can be
divided in: open channels, tiles, and moles. Their purpose is to
remove water from the upper part of the soil improving
trafficability and decreasing waterlogging that damages the crops.
Artificial drainage reduces the overland flow volume and the
amounts of P forms and sediment overland flow transports (Sharpley
et al., 1976; Turtola and Paajanen, 1995; Nelson et al., 1996).
Note that wetland drainage should be avoided because wetlands are
nutrient sinks (Nelson et al., 1996). Recent research indicates
that land drainage can contribute significantly to and cause
eutrophication, although its contributions are less than from
overland flow (Sims et al., 1998).
Open channels: Open channels include open ditches and canals
(Schwab et al., 1993). In this review, only ditches will be
discussed because most of them drain to surface waters. In
contrast, canals carry irrigation water. Open ditches are primarily
used in flat land with impermeable subsoil and shallow topsoil
because subsurface drainage is not economical and practical. There
are three types of open ditches (Schwab et al., 1993): i)
bedding-shallow depressions that are plow deadfurrows, ii) field
ditches-drains that are more widely spaced than deadfurrows, and
iii) parallel open ditch-drains that machinery cannot cross.
Campbell et al. (1985) found that furrows and overland flow had
similar P concentrations (0.30 and 0.29 mg/L, respectively).
However, Izuno et al. (1991) found that P concentration
15
decreased as the water moved to larger ditches due to dilution,
assimilation, and adsorption. Less contact with the surface soil,
compared to overland flow, also suggests that water in ditches
could decrease NPS P contributions. Although P concentration might
decrease, TP (unf.) loads may still be high enough to cause
eutrophication. Sallade and Sims (1997a and b) found that in
ditches in the Delaware Inland Bay watershed, 42 % of the winter
and 53 % of the spring samples had TP (<0.45) concentrations
that could cause eutrophication.
The quality of drainage waters in ditches is also influenced by the
crops that are present, management practices that are used and
field conditions (Izuno et al., 1991). Crops with high P uptake in
soils and high P mineralization can have very low P losses. In
contrast, Coale et al. (1994a) found TP (<0.45) and TP (unf.)
losses between sugarcane and fallow plots from drainage events were
not different. Faster field drainage rates lead to smaller P losses
(Coale et al., 1994b). Under slower drainage rates the soil becomes
saturated faster and for a longer period that leads to higher TP
(<0.45) losses although TP (>0.45) losses do not decrease
compared to fast field drainage rates.
Tiles: Tiles can be separated according to their inlets to: i)
surface inlets and ii) blind inlet (French drain) (Schwab et al.,
1993). The surface inlet tiles have a surface intake structure that
removes surface water from potholes, road ditches and other
depressions. In contrast, the blind inlet tiles have no surface
intake and are constructed by backfilling the trench. Blind inlet
tiles should be more successful in reducing P concentrations
because the water that reaches these tiles travel through the soil
and P can be adsorbed. For surface inlet tiles, P losses could be
higher, since the water originates as overland flow. Surface inlet
tiles could be successful in reducing P losses if the water spends
less time on the surface and causes less surface erosion, thereby
minimizing TP (>0.45) losses.
Many studies do not identify the type of tiles under investigation,
but the ones that do have focused primarily on blind inlet tiles.
In this review, both types of tiles will be treated similarly.
Sharpley and Syers (1979b) found that tiles contributed less P to
streams than overland flow, but more than interflow. Haygarth et
al. (1998a) found that TP (unf.) losses were reduced by 30 % in
tile drained lands compared to overland flow, while Schwab et al.
(1980) found a reduction of 45 %, because only a small portion of P
originated from the surface when soils were drained. Tiles were
also much more successful in reducing P compared to surface
drainage ditches (Campbell et al., 1985, Sims et al., 1998). Tiles
primarily reduced TP (>0.45) (Bottcher et al., 1981).
Other studies have found large amounts of sediment in tile
effluent, as much as 2,800 kg ha-1 (Logan and Schwab, 1976).
Particulate matter carrying P can be transported from the topsoil
to the tiles within hours (Laubel et al., 1999). As the water flows
vertically through the soil (matrix flow) to the drains adsorption
decreases P concentration (Haygarth et al., 1998b). However, when
preferential flow occurs, adsorption is more limited because of the
reduced contact time between percolating waters and the soil, and
sediments will be more easily transported (Sharpley and Syers,
1979a).
Powlson (1998) believes that more P moves by preferential flow than
previously thought and this is why tile waters can promote
eutrophication. Sims et al. (1988) report that ridge tillage had
higher P losses in tiles than conventional tillage because of the
enhanced formation of macropores that increased preferential flow.
Similarly, fertilized forages had greater TP (<0.45) losses than
continuous corn because of the better established macropore
network.
Even if tile water has low P concentrations, it can be a major
pathway, because of the large areas that have been tiled, that
contribute significant amounts of water to stream flow. Tile
16
drain effluent can also contain greater P concentrations during
flooding events that create anaerobic or reducing conditions that
lead to increased P mobility (Sawhney, 1978).
Phosphorus losses are significantly influenced by agricultural
management practices (Sims et al., 1998), crop cover (Sawhney,
1978; Phillips et al., 1982) and surface soil P. Greater soil cover
over tile lines decreases P concentrations (Bottcher et al., 1982).
Phosphorus concentrations in tiles increased substantially after
the soil reached 57 mg kg-1 Olsen P even though the subsoils were
P-deficient (Heckrath et al., 1995; Hesketh and Brookes, 2000).
This soil P value is called the change-point. In soils with P
values above the change-point, tile effluent has elevated P losses
because of P transport in preferential flow. Additionally when the
soil P is above the change-point it is well above what is needed
for plants optimum yield. Moles: Moles are cylindrical artificial
channels in the subsoil similar to tile but not lined and not as
deep (Schwab et al., 1993). They have been successfully used in
England and New Zealand, but are not commonly used in the United
States. They are a temporary method of drainage and deteriorate
within the first few years. Sharpley and Syers (1976a) found that
moles decreased overland flow and P transport. Doubling the mole
spacing increased P losses (Addiscott et al., 2000). As the spacing
and the travel distance of the water increased, water reaching the
moles had greater amounts of P-enriched sediment. 2.1.4
Hydrology
Hydrology is the main driving force and carrier for P transport
(Sharpley and Tunney,
2000; Haygarth and Jarvis, 1999; Edwards and Withers, 1998). Gburek
et al., (2000) mention that to assess water quality problems the
hydrologic pattern (wet vs. dry periods) should be considered.
Phosphorus concentrations in surface waters do not follow a
consistent seasonal pattern (Owens et al., 1989). Instead annual
and seasonal P concentrations are higher with higher precipitation
(Correll et al. 1999; Shirmohammadi et al., 1997) or peak
discharges (Kolpin et al. 2000). Major rainfall events can account
for up to 90% of annual P losses (Sharpley et al., 2000; Edwards
and Withers, 1998; Schuman et al. 1973) with the lowest P
concentration in surface water during baseflow (House et al., 1998)
(Table 5). Rainfall intensity, duration, and timing as well as
snowmelt events have major influences on P transport and loss
(Truman et al., 1993; Culley and Bolton, 1983) (Table 5). High
antecedent soil water content also increases P losses because of
higher runoff volumes (Edward and Withers, 1998; Coale et al.,
1994a).
Hydrologic effects can be further separated into temporal and
spatial effects. The temporal effects have two major levels of P
transport. Level 1 is associated with baseflow, and level 2 is
associated with stormflow (Haygarth and Jarvis, 1999). Typically,
in the temperate region 7.5 mm d-1 of rain may cause erosion
(Evans, 1978), while 5 mm h-1 of rain or more is considered high
intensity that can cause severe erosion (Boardman and Robinson,
1985). Spatial hydrologic effects can be divided into soil profile,
slope/field and catchment effects. Soil profile describes the
vertical water transport pathways (matrix and preferential flow)
over centimeters to meters of depth. Slope/field describes lateral
water pathways (overland flow, interflow, and land drainage) over
hectares to square kilometers. Finally, catchments (watersheds)
include land-use variables from square kilometers and upward.
17
Table 5. Phosphorus concentrations in stream and drainage water
under various flow regimes. Ranges are in the parenthesis when
available. Flow regimes TP (unf.) TP (< 0.45) Land-use mg L-1
Stormflow 1.5 a
0.49 (0.14-2.37) c 1.20 (0.16-4.30) c 1.43 (0.50-5.21) c
0.46 (0.07-3.30) d
0.13 (<0.01-0.36) d
Pasture with dairy operations, stream water. Grassland, tile
drains. Pasture, stream water. Riparian pine afforested 1, stream
water. Riparian pine afforested 2’ stream water. Cropland (91%),
stream water.
Snowmelt 0.27 (0.09-0.90) d
Baseflow 0.8 a
0.071 (0.02-0.61) d
0.031 a (<0.01-0.23) d
Pasture with dairy operations, stream water. Grassland, tile
drains. Pasture, stream water. Riparian pine afforested 1, stream
water. Riparian pine afforested 2’ stream water. Cropland (91%),
stream water.
a Shirmohammadi et al. (1997); b Stamm et al. (1997); c Smith
(1992); d Culley and Bolton (1983). 2.1.5 Terrestrial sources of
phosphorus for surface waters
Major sources of P that get into surface water in agricultural
watersheds are atmospheric deposition, the soil, and agronomic
inputs. Phosphorus in the soils originates from weathering of
parent materials, as well as from external inputs of fertilizers,
manure, animal feces and animal litter. While the agronomic sources
of P provide inputs to the soil they may also serve as direct
sources to surface water in surface runoff events.
2.1.5.1 Atmospheric deposition The importance of atmospheric
deposition depends on the ration of the drainage area of
the watershed to that of the actual surface water. In most lotic
systems the drainage area is much greater than the water surface
area. As a result, the low concentrations of P in atmospheric
deposition are not a very important source compared to agronomic
inputs in lotic surface waters (Johnes and Hodgkinson, 1998; Ryding
et al., 1990).
Atmospheric deposition (wet and dry) is 0.22 kg TP (unf.) ha-1yr-1
in England, 0.11 kg TP (unf.) ha-1yr-1 in Scotland (Haygarth et
al., 1998b), while in Illinois wet atmospheric deposition is much
lower, 0.02 kg TP (unf.) ha-1yr-1 (David and Gentry, 2000). Jordan
et al. (1995) found in the Rhode watershed in Maryland even smaller
amounts of atmospheric deposition (wet and dry) (0.0002 kg TP
(unf.) ha-1yr-1). For the whole Rhode watershed, the atmospheric
deposition (wet and dry) amounts were 0.94 kg TP (unf.) yr-1 while
the rest of the load from the watershed was 1749 kg TP (unf.) yr-1
(Correll et al., 1992).
Gibson et al. (1995) summarized a number of studies and found that
wet atmospheric inputs typically range from 0.05-0.40 kg TP (unf.)
ha-1yr-1 and are not a significant source. However, others have
found that rainfall can contribute up to 25 % of annual P inputs
(Johnes et al., 1996) to freshwater and be a significant source of
P eutrophication (Schindler, 1977; Lee,
18
1973). Schwab et al. (1980) found in a crop field in Ohio that rain
water averaged 4.6 kg TP (unf.) ha-1yr-1 while surface drains and
deep pipe drains averaged only 1.9 and 1.6 kg TP (unf.) ha-1yr-1,
respectively.
2.1.5.2 Soil Sources Soil P sources have a major influence on
dissolution and physical modes of transport.
They control the release of P from the solid phase to the solution
(dissolution) phase, and the amount of P in the particulates that
are transported by water or wind (physical) (Haygarth and Jarvis,
1999). Although total soil P concentrations in the soil, range from
0-0.4 % with an average of 0.05 % (kg/ha in the furrow slice)
(Ryan, 1983), soil may be the major source of P to surface waters
(House et al., 1998). To understand the reasons for this seeming
discrepancy it is necessary to describe the different forms of P in
the soil, the P cycle in soil, current soil P levels, the
influences of different soil textures, soil microorganisms and the
current soil P test methods that are used to identify the P in the
soil.
Soil P forms: Phosphorus in the soil can either be in soil solution
or in the soil matrix.
Typically, soil P is described in terms of the following
relationship (Larsen, 1967):
Soil solution P ↔ labile soil P ↔ nonlabile soil P In general, out
of 1000 kg ha-1 of P in the soil only 1 kg ha-1 is in solution
(Troeh and
Thompson, 1993). Soil typically contains 100-3000 mg orthophosphate
kg-1 (Frossard et al., 2000) the form that plants primarily take
up. The dominant orthophosphate form changes with the pH
(Pierzynski et al., 2000):
pH 2.2 7.2 12.3 orthophosphate form H3PO4 ↔ H2PO4
- ↔ HPO4 -2 ↔ PO4
-3
The labile and nonlabile soil P describes the condition of P in the
soil matrix (Troeh and Thompson, 1993). Labile soil P refers to the
P in the soil that readily replaces P lost from the soil solution
(e.g. plant uptake) within a period of several days or a few weeks
to maintain equilibrium. In contrast, nonlabile is not readily
available.
The main forms of P in the soil matrix are inorganic P and organic
P (Table 6) (Bowman, 1989; Tate and Newman, 1982; Syers, 1974) and
they can be both labile and nonlabile. Intensively managed
agricultural soil with long-term fertilizer and manure inputs has
higher proportions of inorganic P (Hawkins and Scholefield, 1984)
ranging from 8 to 72% of the total P (Haygarth and Jarvis, 1999).
Inorganic P is easily released in solution and is primarily
considered labile. Major inputs of inorganic P in soils are
weathering (mineral apatite Ca5(PO4)3F) and fertilizers. Little is
known about the role of organic P in P transport although its
importance is increasingly acknowledged (Haygarth et al., 1998a;
Hannapel et al., 1963a). Organic P is derived from plant residues
and excreta from above- and below-ground organisms (Haygarth and
Jarvis, 1999). The proportion of P in organic forms can range from
30 to 65 % of the total soil P (Frossard et al., 2000). The most
common organic compounds are phytic acid, nucleic acids and
phospholipids that range from 35, 2 and 1% (Anderson, 1980) to 50,
2 and 5% (Haygarth and Jarvis, 1999), respectively of the total
organic P in the soil matrix.
19
Table 6. The common chemical forms of P in soil and their
characteristics and potential mobility. a Form Characteristics or
Implication INORGANIC in solution Orthophosphate (H3PO4) Readily
mobile but easily
adsorbed/immobilized INORGANIC in the soil matrix Apatites
Hydroxyapatite (Ca10(PO4)6OH2) Fluorapatite (Ca10(PO4)6F2) Sodium
phosphates Pyrophosphate (Na4P2O7
.H2O)
Very low solubility. Tend to be present more in nonacid soils than
in acid. Soluble. No known information on the implications
Polyphosphate (Na3PO3)n Other calcium phosphates Monocalcium
phosphate (Ca(H2PO4)2
.H2O) Dicalcium phosphate (Ca(HPO4)
.5H2O) Aluminum phosphates Variscite (AlPO4
.2H2O) Taranakite (H6K3Al5(PO4)8
.8H2O) Strengite (FePO4
.H2O) Surface-adsorbed P
Tend to form when fertilizers are added to nonacid soils. Tend to
form when fertilizers are added to acid soils that have aluminum.
Tend to form when fertilizers are added to acid soils that have
iron. Adsorbed on calcium, iron and aluminum compounds.
ORGANIC in soil matrix Phytic acid or Inositol hexaphosphates
(IHP:(C6H6O6)(PO3)6)
Not generally thought to be easily mobile except in sandy soils:
maybe highly sorbed but have been noted in lake sediments
Phosphate diesters (nucleic acids, DNA, RNA, phospholipids)
Probably mobile but confirmation is required
Glucose P (D-Glucose-6-phosphate; D-Glucose-1- phosphate)
Leaches through sandy soils
Phosphonates (R-PO3) No known information on the implications
Polyphosphanates (ATP,AMP) No known information on the
implications
a Haygarth and Jarvis (1999); Troeh and Thompson (1993); Anderson
(1980); McClellan and Gremillion (1980); Sample et al.
(1980).
20
The established terms for P forms in the soil, organic, inorganic,
labile and nonlabile, are ambiguous because they are physical and
chemical definitions (Haygarth, 2001; Sharpley, 2001). As with the
P forms in water, Haygarth and Sharpley (2000) suggest that an
operational definition should be incorporated with each term
describing soil P. The classification of P in solution is similar
to the classification of P in water (Table 3). For the
classification of P forms in the soil matrix, the operational
definitions of the procedure used to measure P is incorporated
(Bray-P, Mehlich-P, Olsen-P). More work is needed for the
classification of P in the soil.
Soil P-Cycle: The P cycle in the soil includes inputs and outputs
of P, as well as internal cycling of P (Figure 6). A point to note
about the soil P-cycle is that there is no process to remove P to
the atmosphere like in the N-cycle (denitrification) (Cooper and
Gilliam, 1987). Removal of P can occur by erosion of enriched
sediment, desorption by moving water (runoff, leaching, tiles), or
crop and animal removal. The major internal P inputs are weathering
and the major external P inputs to soil include fertilizers,
agricultural wastes, plant residues, atmospheric deposition, and
municipal/industrial by products. This review deals with NPS
pollutants and will not discuss municipal/industrial by-product
inputs. Finally, the internal cycling processes include
immobilization-mineralization, adsorption-desorption, and
precipitation-dissolution.
Internal Soil P Inputs Weathering: Phosphorus in the soil
originates primarily from weathering of minerals and
other more stable geologic materials (Pierzynski et al., 2000) and
typically is a minor input for soil in agricultural landscapes
(Johnes and Hodginkson, 1998). However, Abrams and Jarrell (1995)
suggest that native soil P is a potential source for NPS P
pollution. In other studies,
Atmospheric deposition
Plant residues
Figure 6. Phosphorus cycle in the soil with inputs, outputs and
transformations that take place in the soil (modified from
Pierzynski et al., 2000).
21
geology is the most important factor correlating P concentrations
in the stream water compared to land-use practices (Thomas et al.,
1992; Thomas and Crutchfield, 1974).
External P inputs into the soil Atmospheric deposition: Wet
atmospheric inputs typically range from 0.05-0.40 kg TP
ha-1yr-1 (Gibson et al., 1995) and are a minor input for soil
(Haygarth et al., 1998b). Agricultural wastes: Animal manure,
litter from animal confinements, and feces are the
major agricultural wastes. Animal manure is animal waste from
confinement facilities that are applied to cropfields. Cultivated
soils that have repeated manure applications have consistently
higher soil P (Mehlich I) than the field borders that do not
receive manure (Mozaffari and Sims, 1994). Animal litter includes
bedding mixed with animal manure. Soils with broiler litter applied
for twenty years had 86 times more Mehlich-III P in the plow layer
than unamended soil (Oldham, 1998). Typically, manure or animal
litter applications are designed to meet crop N levels. Low N:P
ratios in manure and animal litter lead to over application of P
compared to the needs of crops (Sharpley et al., 1996). Animal
feces is animal waste that is deposited by animals in pastures and
rangelands, and can be a significant input on pastures, especially
if they are overgrazed. Haygarth et al., (1998b) found feed
supplements were a large source of high P levels in dairy manure
with 70% of P intake excreted in animal feces (Tamminga,
1992).
Fertilizers: For decades P fertilization rates have exceeded the
amount of P removed by crops and that has resulted in increased
soil P (Sharpley et al., 1994).
Plant residue: Plant material that is not harvested and is left in
the fields can increase P levels, especially if P fertilization is
not incorporated in the soil profile (Sharpley et al., 1994). On
no-till fields, that had surface applied fertilizer P the soil
total P levels increased by six times within a couple years
(Griffith et al., 1977).
Internal Cycling Adsorption-desorption, and
precipitation-dissolution are the major soil inorganic P
transformations (Pierzynski et al., 2000) (Figure 6). Adsorption
refers to the chemical bonds that form between orthophosphate
anions (H2PO4
-, HPO4 -2) and soil colloids. Clays, oxides and
hydroxides of Fe, cations of Al and calcium carbonate, and organic
matter are the main reactive soil phases. Desorption, is the
release of P from the solid phase into the soil solution. Plant
uptake, runoff, and leaching cause desorption. Precipitation is the
formation of insoluble compounds in soils and dissolution is the
reverse process. Orthophosphate anions reacting with Ca, Al, and Fe
are the most common precipitated P products.
Mineralization-immobilization is the major soil organic-inorganic P
transformation (Pierzynski et al., 2000) (Figure 6). Mineralization
is the decomposition of organic matter by microbes that results in
the release of inorganic P and is highly dependent on soil
conditions. Immobilization, in contrast, is the conversion of
mineral (inorganic) P to biochemical compounds (organic P) by soil
microorganisms and is part of the active soil biomass fraction of
the soil. Internal cycling is the major replacement mechanism of P
in the soil solution with labile soil P. The replacement can either
occur with desorption of adsorbed P ions on the soil or dissolution
of P compounds.
Soil P levels: Similar soils may have different P losses to surface
waters because of different soil P levels (Pote et al., 1999). Soil
can be a source of P for surface waters when P levels are high
(Abrams and Jarrell, 1995). Phosphorus can build up in soils
because it is fixed in soils and not easily leached (Haygarth et
al., 1998a). As soil P levels get higher in the surface
22
soil, the risk of P moving to the surface water by soil erosion and
overland flow becomes greater (Mitchell, 1998). Important
relationships between P losses to surface waters and soil P levels
have been observed in grasslands (Smith et al., 1995), arable lands
(Heckrath et al., 1995), forest, and cropped watersheds (Sharpley
et al., 1994). Vaithiyanathan and Correll (1992) found that soil P
content and soil loss in runoff were highly correlated (r2 = 0.96)
for both forested and cropped watersheds in the Atlantic Coastal
Plains.
Intensive agriculture has maintained a P surplus in the soil (Table
7) from P fertilizers and animal waste inputs that exceed P outputs
(Sharpley et al., 1994). Producers have applied P at rates
exceeding crop uptake (Pote et al., 1996). The high P levels in the
top 5-10 cm with the rapid decrease of P below this depth,
indicates that P has accumulated in the soil from external sources
(Haygarth et al., 1998a; Cooper and Gilliam, 1987). Globally, from
1950 to 1995 ~600 x 106 Mg of fertilizer P were applied on the
earth’s surface (primarily in croplands) while only ~250 x 106 Mg
were removed from croplands through harvest (Carpenter et al.,
1998). Additionally, animal manure has added 50 x 106 Mg that leads
to a net addition of P during this period of 400 x 106 Mg
(Carpenter et al., 1998). Once soil P levels exceed crop P
requirements, the potential of P losses with runoff and erosion is
greater than any agronomic benefit (Figure 7). In Ireland, the
average soil P test has increased 10-fold in the last 45 years
(Tunney et al., 1997). Table 7. Soil P surpluses for different
developed countries under different agronomic land- use practices.
Country Management Surplus kg P ha-1 yr-1 Denmark b cropfields 49
Federal Republic of Germany c cropfields 55 Germany Democratic
Republic c cropfields 71 Europe b cropfields 17 Italy b cropfields
11 Lombardia, Italy b cropfields 50 Netherlands (two references) b
cropfields 94 United Kingdom b cropfields 3 England U.K. b
cropfields 2 Belgium b grassland, dairy farm 38 Denmark b
grassland, dairy farm 11 Europe b grassland, dairy farm 24
Brittany, France b grassland, dairy farm 92 Italy b grassland,
dairy farm 21 Lombardia, Italy b grassland, dairy farm 32
Netherlands b grassland, dairy farm 49 United Kingdom b grassland,
dairy farm 14 England U.K. b grassland, dairy farm 14 Devon, U.K. b
grassland, dairy farm 27 Northern Ireland, U.K. b grassland, dairy
farm 24 Pennsylvania, U.S.A. b grassland, dairy farm 11 Scotland,
U.K. b grassland, extensive sheep 0.24 Delaware, U.S.A. a
poultry-grain farm 34 Inland Bay watershed, Delaware, U.S.A. a
various 52 Sussex County, Delaware, U.S.A. a various 70 Delaware,
U.S.A. a various 30 a Pierzynski et al. (2000); b Haygarth and
Jarvis (1999); c Isermann (1990).
23
Soil P Test (mg kg-1) Figure 7. Relationship between soil P test,
crop yield, and environmental problems due to excessive soil P
(Pierzynski et al., 2000).
In Illinois, large inputs of fertilizer and manure occurred from
1965 to 1990 with a net input of 230 kg P ha-1 yr-1 (David and
Gentry, 2000). Since 1990, the P cycle has become balanced because
P fertilization decreased to the point where net P export matches
inputs. But most of the P surpluses from the last 25 years are
still in Illinois soil because once high levels of soil P have been
attained, considerable time is required for significant depletion
(Sharpley et al., 1994).
Phosphorus accumulations (Figure 8) in the United States have
reached a point were soils are considered ‘over-fertilized’ because
their P levels exceed crop needs (Sharpley and Smith, 1989a).
Thomas and Crutchfield (1974) found no direct correlation between
stream water P concentrations and concentrations of fertilizers
applications. Halving P fertilizer applications did not
significantly lessen P losses in mole drain flow (Addiscott et al.,
2000). In both cases, the fertilizers had no effect because the
potentially mobile P levels in the soil were very high and soils
are the primary contributor to P losses in runoff (House et al.,
1998; Pote et al., 1996).
Soil Type: Soil textural differences may account for differences in
TP (unf.) in the soil (Table 8) but also in different P losses to
surface waters (Stevens et al., 1999; Heckrath et al., 1995).
Organic soils have greater P losses than mineral soils (Duxbury and
Peverly, 1978). Sandy A-horizon soils and clay-cultivated layers
can have very high P losses (Catt et al., 1998). These differences
are due to different soil dispersion and soil chemical
transformations. In many cases soil with higher soil P levels are
less threatening to surface water because of low dispersivity and
dissolution, and high adsorption. Low dispersivity and dissolution,
and high adsorption decrease the chances of soil P transport.
Soil dispersivity is promoted with high pH, high percentage of
larger particles, low salt concentrations and high ratio of
monovalent to divalent and trivalent cations. The major affect of
high soil dispersivity is that it increases soil erosion and TP
(>0.45). However, soil dispersion
24
Figure 8. Percentage of soil samples with high soil P levels in
1989 (Sharpley et al., 1996). Table 8. Percentage of TP (unf.) in
A, B and C horizons of several kinds of soil. Methods used to
measure TP (unf.) in the soil were not described. a
Percentage P (%) (kg ha-1 in furrow slice)
Soil Type Soil Order A horizon B horizon C horizon Miami Silt loam
Alfisol 0.035 0.031 0.035 Barnes silt loam Mollisol 0.100 0.065
0.065 Holdrege silty clay loam Mollisol 0.087 0.092 0.096 Mackburg
silty clay loam Mollisol 0.061 0.072 0.086 Marshall silt loam
Mollisol 0.052 0.044 0.070 Otley silty clay loam Mollisol 0.055
0.055 0.070 Richfield clay loam Mollisol 0.044 0.017 0.039
Sharpsburg silty clay loam Mollisol 0.059 0.073 0.081 Nipe clay
Oxisol 0.250 0.140 0.140 Becket fine sandy loam Spodosol 0.057
0.035 0.031 Davidson clay loam Ultisol 0.044 0.087 0.105 Maury silt
loam Ultisol 0.136 0.158 0.240 Sassafras sandy loam Ultisol 0.048
0.035 0.031 Houston black clay Vertisol 0.065 0.074 0.039 a Troeh
and Thompson (1993). also influences losses from fertilizer and
manure. Phosphorus losses from fertilizers are greater in more
dispersible soils (Addiscott et al., 2000; Catt et al., 1998). In
dispersible soils, it is recommended that the fertilizer or manure
be applied when soil moisture content is below field capacity (Catt
et al., 1998).
The soil chemical transformations that are most affected by soil
type are adsorption- desorption. These transformations control P
transport between the solid phase and soil solution and
subsequently the vulnerability of soil P to losses in water as TP
(<0.45) (Frossard et al., 2000). Soils with the highest P
adsorption are usually, acidic, and high in clay or Fe/Al oxides,
particularly amorphous oxides (Pierzynski et al., 2000). Large
areas with high clay content are
25
important P sinks (Cooper and Gilliam, 1987). High levels of
reactive Fe and Al in the soil can also reduce P transport (Abrams
and Jarrell, 1995). In general, weathered soils that have more clay
and Al/Fe oxides adsorb greater amounts of P (Pierzynski et al.,
2000). In soils with high pH (calcareous) the existence of CaCo3
and Fe oxides increase P adsorption (Pierzynski et al., 2000).
Adsorption maxima give the long-term capacity of soil to retain P
(Sharpley et al., 1994).
Sandy soils are more susceptible to P leaching when fertilizer is
added because of low P adsorption (Pierzynski et al., 2000). In
soils with high organic matter (peats and heavily manured soils), P
mobility is enhanced because the colloidal surfaces responsible for
P adsorption are coated by soluble organic matter, and because
organic P complexes leach faster and to greater depths. Soils with
high equilibrium P concentration at zero adsorption have a greater
tendency to desorb TP (<0.45) in runoff (Wolf et al.,
1985).
Soil organisms: Soil organisms can influence P transport because
they play a major role in P immobilization-mineralization processes
that control the transformations of P between inorganic and organic
forms (Frossard et al., 2000). Organic P forms are mineralized
through ingestion/excretion by soil organisms while inorganic forms
are immobilized by retention within their biomass (Patra et al.,
1990; Brookes et al., 1982). Plant roots and soil microbes can
cause hydrolysis of organic P. Hannapel et al. (1963b) found that
the addition of a microbial energy source increased mobility of P
by 38 times. Treatments with formaldehyde suppressed microbe
activity and reduced P mobility. When inorganic P was added in the
residue, P transport was not significantly increased. This
indicated that the mobilization of P by the microbial population
was the most important factor in P transport (Hannapel et al.,
1963a).
Earthworms increase soil P levels because their casts contain finer
soil particles that can release 4 times more inorganic P into
solution than surface soil (Sharpley and Syers, 1976b). However,
absence of earthworms would reduce infiltration and the litter
incorporation into the soil thereby increasing overland flow and P
losses (Radke and Berry, 1993; Sharpley et al., 1979).
Some microorganisms can solubilize certain P forms and cause
leaching (Illmer et al., 1995), while others can excrete P
(Haygarth and Jarvis, 1999). Passively, soil organisms can also
transport P as it is attached or contained within them (Haygarth
and Jarvis, 1999). Finally, when these organisms die,
wetting/drying and/or freezing/thawing cause cell lysis of the dead
organisms that releases P that is available to the soil solution
(Haygarth and Jarvis, 1999).
Soil P test: It is very difficult to rectify P losses in surface
waters when soil P is high.
This is because reduction of surplus P in soil needs long-term
planning (Sharpley et al., 2000). Before meaningful planning can be
conducted soil P analysis must be conducted. Most of the P
accumulation in the soil is in the upper few centimeters that are
the most critical for P runoff to surface waters (Ahuja,
1986).
The soil P tests (Bray-I, Olsen, and Mehlich-I and -III) are simple
and inexpensive tools that can assure optimum crop production and
optimum nutrient availability for crops (Maguire et al., 1998).
From an environmental prospective, however, the agronomic soil P
tests are not very useful. These tests are based on crop nutrient
requirements not on runoff water quality. In many cases, they might
underestimate potentially mobile P that could be transferred to
surface waters (Haygarth and Jarvis, 1999). For example, Olsen P
represents only 1-5% of the total soil P. However, the remaining
95-99% of the soil P might also be potentially mobile. The
sampling, analytical, interpretive, and educational role of soil P
testing should be re-evaluated to meet
26
environmental goals (Sims, 1993). Alternative soil testing
approaches should be considered for soil P release to runoff. These
approaches include the estimation of algal available P (AAP), or
water extractable P and the iron-oxide impregnated paper method
(Pote et al., 1996; Sharpley, 1993). Another approach gaining
acceptance is the P adsorption saturation of the top 5 cm of soil
(more effective than top 15 cm) (Sharpley et al., 1996).
Threshold soil P test values have been recommended (Table 9).
However, different soils can have different susceptibilities to P
losses irrespective of soil test P (Sharpley et al., 2000).
Phosphorus management recommendations for soils with similar P, but
contrasting land-use management and topography, could lead to
extreme differences in P losses. Pote et al. (1996)
Table 9. Recommended agronomic and environmental soil P test
threshold values with the appropriate P management recommendations.
a
State Agronomic b Environmental Soil P Test Management
recommendations for water
Threshold values method quality protection mg kg-1 Arkansas 50 150
Mehlich-III At or above 150 mg kg-1 soil P:
Apply no more P, provide buffers next to streams, overseed pastures
with legumes to aid P removal, and provide constant soil cover to
minimize erosion.
Delaware 25 50 Mehlich-I Above 50 mg kg-1: Apply no more P unless
soil P is significantly reduced.
Idaho 12 50 and 100 Olsen Sandy soil-above 50 mg kg-1: Silt loam
soils-above 100 mg kg-1: Apply no more P unless soil P is
significantly reduced.
Ohio 40 150 Bray-I Above 159 kg mg-1: Reduce erosion and reduce or
eliminate P additions
Oklahoma 30 130 Mehlich-III 30 to 130 mg kg-1 soil P: Half P rate
on slopes > 8% 130-200 mg kg-1 soil P: Half P application rates
and reduce surface runoff and erosion Above 200 mg kg-1 soil P: P
rate not to exceed crop removal
Michigan 40 75 Bray-I Below 75 mg kg-1 soil P: P application not to
exceed crop removal Above 75 mg kg-1 soil P: Apply no P from any
source
Texas 44 200 Texas A&M
Above 200 mg kg-1 soil P: P addition not to exceed crop
removal
Wisconsin 20 75 Bray-I Below 75 mg kg-1 soil P: Rotate to P
demanding crops and reduce P additions Above 75 mg kg-1 soil P:
Discontinue P applications
a Sharpley and Tunney (2000). b Agronomic threshold concentrations
are average values for nonvegetable crops; actual values vary with
soil and crop type. In addition, vegetables have higher agronomic P
requirements.
27
found that although soil P test values for three sites ranged only
from 285-295 mg kg-1 (using Mehlich-III), P losses in overland flow
were 0.05, 0.16, 0.35 kg/ha during a 30-min rainfall simulation due
to runoff volume variations (Sharpley and Tunney, 2000). Pionke et
al. (1997) reported that most annual P export comes from a small
area of the landscape. Therefore, threshold values have little
meaning if only the soil P test is used without defining the site’s
potential for overland flow and erosion. The P-index (discussed in
more detail in a following section) provides a more reliable tool
to estimate the risk of P losses than soil P tests.
2.1.5.3 Agronomic Sources Under the right conditions, agronomic P
sources of animal feces, fertilizers, manure,
animal litter, and plant residue can make their way to surface
waters very quickly making them the most influential for the
incidental transport mechanism.
Animal feces: Phosphorus from animals is predominantly excreted in
feces (Betteridge
et al., 1986). Grazing animals can directly deposit feces in the
surface waters (Shirmohammadi et al., 1997). Feces can also move to
surface waters by runoff. The best ways to reduce these sources is
to fence or restrict animal entrance into the stream, and decrease
grazing densities in riparian areas (Sharpley et al., 1994).
Finally, different animals produce different amounts of feces and
have different amounts of P in their feces (Table 10). Table 10.
The amounts of feces and phosphorus produce by different animals.
a
Animal Size Total Manure Production Phosphorus oxide (P2O5)
Phosphorus oxide (P2O5)
kg kg d-1 kg d-1 kg yr-1 Dairy Cow 68 5.4 0.0104 4128 113 9.1
0.0204 6804 227 18.6 0.0372 13608 454 37.2 0.0753 27670 635 52.2
0.1502 52618 Beef Cattle 227 13.6 0.0576 20412 340 20.4 0.0866
30845 454 27.2 0.1134 41278 567 34.0 0.1442 51710 Swine Nursery Pig
16 1.0 0.0054 1950 Growing Pig 29 1.9 0.0223 3720 Finishing Pig 68
4.4 0.0500 8618 Finishing Pig 91 5.9 0.0680 11340 Boar 159 5.0
0.2676 9979 Sheep 45 1.8 0.0068 2495 Poultry-Layers 2 0.10 0.00011
422 Poultry-Broilers 1 0.06 0.00006 195 Horse 454 20.4 0.0476 17690
a Midwest Plan Service (1985).
28
Fertilizers-Manure-Animal Litter: Fertilizers, manure, and animal
litter applied to
cropfields can be a significant source of P for surface waters.
They can be directly deposited in stream water with improper
application, but usually end up in surface water bodies with
runoff. In the United States 4100 million metric tons of phosphate
fertilizer and 1.2 million metric tons of organic phosphorus were
applied in 1992 (USDA-NRCS, 1997c).
When fertilizers are applied to cropfields stream P levels are
often elevated (Oldham, 1998; Sharpley and Rekolainen, 1997).
Single additions of P fertilizers increase concentrations and
amounts of P in overland flow (Sharpley and Syers