Advances in agronomy volume 25

in subsurface and groundwater runoff will be influenced by the time of contact with any component in the soil profile capable of interacting with dissolved P in the percolating water and

AGRONOMY VOLUME 25

CONTRIBUTORS TO THIS VOLUME

AGRONOMY

Prepared under the Auspices of the

AMERICAN SOCIETY OF AGRONOMY

ACADEMIC PRESS New York San Francisco London

A Subsidiary of Harcourt Brace Jovanovich, Publishers

COPYRIGHT 0 1973, BY ACADEMIC PRESS, INC

ALL RIGHTS RESERVED

N O PART O F THIS PUBLICATION MAY B E REPRODUCED OR

T R A N S M I m E D IN ANY FORM OR BY ANY MEANS, ELECTRONIC

OR MECHANICAL, INCLUDING PHOTOCOPY, RECORDING, OR ANY INFORMATION STORAGE AND RETRIEVAL SYSTEM, WITHOUT

PERMISSION IN WRITING FROM T H E PUBLISHER

ACADEMIC PRESS, INC

111 Fifth Avenue, New York, New York 10003

United Kingdom Edition published by

ACADEMIC PRESS, INC (LONDON) LTD

24/28 Oval Road, London NWI

LIBRARY OF CONGRESS CATALOG CARD NUMBER: 50-5598

PRINTED I N T H E UNITED STATES OF AMERICA

CONTRIBUTORS TO VOLUME 25 , , ix PREFACE xi

PHOSPHORUS IN RUNOFF AND STREAMS

J C RYDEN, J K SYERS, A N D R F HARRIS

I Introduction 1

11 Terminology 2

111 Factors Affecting the Dynamics of Phosphorus in Runoff and Streams 4

IV Phosphorus Loads in Runoff and Streams 20

V Impact of Phosphorus Carried in Streams on Standing Waters 37

VI Present Status and Outlook , , , , , 38 References 41

Culture 57 Utilization 65

, 68 1 2 7 3

Conclusions

ZERO-TI LLAGE

K BAEUMER AND W A P BAKERMANS

1 Introduction: The Concept of Zero-Tillage 78

11 Comparison of Environmental Conditions in Tilled and Untilled Soils 80

111 Effects of Zero-Tillage on Plant Growth 95

IV Crop Husbandry , , , , , 103

V Evaluation of Zero-Tillage in Farming Systems 11 3

VI Conclusion 119 References 120

V

vi CONTENTS

J R QUINBY

I

11

111

IV

V

VI

VII

VIII

IX

X

XI

XII

The Floral Stimulus Implications to Plant Breeding References 160

ION ABSORPTION BY PLANT ROOTS T K HOLXES I Introduction 163

11 Overview of Nutrient Absorption by Roots 164

111 Energy-Dependent and Active Ion Transport 167

IV Kinetics and Selectivity of Ion Absorption 180

VII Summary 201

V Energetics of Ion Transport

VI Proposed Model for Ion Absorption by Roots 198

References 202

LODGING IN WHEAT, AND OATS: THE PHENOMENON, ITS CAUSES, AND PREVENTIVE MEASURES MOSHE J PINTHUS I 11 111 IV V VI VII VIII Introduction

Description and Causes

Effects of Lodging on Crop Development and Yield

Plant Characters Associated with Lodging

Environmental and Agronomic Factors Affecting Lodging

Prevention of Lodging

Breeding for Lodging Resistance

Increased Exploitation of Yield-Promoting Factors Due to the References Prevention of Lodging

210

21 1

217

223

23 1

23 8

246

254

256

I

11

111

IV

V

VI

VII

VIII

IX

I

11

111

IV

V

VI

VII

VIII

IX

X

XI

XII

XI11

C BLOOMFIELD AND J K COULTER

Introduction

The Formation of Sulfides

Oxidation of Sulfides

Mining and Corrosion Problems

Classification and Mapping

Conditions for Plant Growth

Management for Agriculture

Analysis of Pyritic Soils

Conclusions

References

MALTING BARLEY I N THE UNITED STATES G A PETERSON A N D A E FOSTER 266 267 278 290 292 296 3 08 3 15 318 319 Classification of Cultivated Barleys of the United States Quality Testing P cceptable Malting Barley Varieties Malting Barley Pr 364

References 375

AUTHOR INDEX 379

SUBJECT INDEX 398

This Page Intentionally Left Blank

Numbers in parentheses indicate the pages on which the authors' contributions begin

K BAEUMER (77), Faculty of Agriculture, University of Goettingen,

W A P BAKERMANS (77), Institute for Biological and Chemical Research

C BLOOMFIELD (265 ) , Rothamsted Experimental Station, Harpenden,

J K COULTER ( 2 6 5 ) , Rothamsted Experimental Station, Harpenden,

A E FOSTER (327), Department of Agronomy, North Dakota State

R F HARRIS ( 1 ) , Department of Soil Science, University of Wisconsin,

T K HODGES (163), Department of Botany and Plant Pathology, Purdue

E A HOLLOWELL (47), W.S Department of Agriculture, Beltsville,

W E KNIGHT (47), U.S Department of Agriculture, Mississippi State,

0 A PETERSON (327), Department of Agronomy, North Dakota State MOSHE J PINTHUS (209), The Hebrew University of Jerusalem, Faculty

J R QUINBY ( 125 ) , Pioneer Hi-Bred Company, Plainview, Texas

J C RYDEN (1 ), Department of Soil Science, University of Wisconsin,

J K SYERS ( I ) , Department of Soil Science, University of Wisconsin,

Goettingen, Federal Republic of Germany

of Field Crops and Herbage, Wageningen, The Netherlands

University, Fargo, North Dakota

of Agriculture, Rehovot, Israel

This Page Intentionally Left Blank

Dramatic reductions during the past two years in the world food supply have jolted a complacent world into the realization that the food-population race remains unquestionably the most critical problem facing mankind Population growth continues at alarming rates in those countries where food supplies are already inadequate Food shortages are plaguing not only the poor countries where hunger, malnutrition, and starvation are a way

of life, but have now reached the more affluent nations Even the United States which for a generation has sought through public programs to limit crop production is now concentrating on programs to increase food supply Once again tillers of the soil, and the crops and animals which supply our food have high national priorities In this time of international concern over food supply, reviews of scientific advancement such as those con- tained in this volume are most reassuring

Papers contained in this volume are concrete evidence of the contribution

of crop and soil scientists to mankind’s efforts to feed himself Four of the papers deal with crops One is concerned with research on crimson clover,

a legume grown in the southern part of the United States and a plant which is most important to a growing animal industry in this area Re- markable progress is reported on knowledge gained from the breeding of sorghum, a plant which is rapidly becoming a major crop in the semi-arid regions throughout the world Factors affecting the lodging of small grains

is the subject of one review Recent advances in research on malting barley, a crop of expanded acreage and of increasing quality expectations

is the subject of the fourth crops article

The reviews of advances in soil science are certainly not unrelated to crop production The mechanisms of ion absorption by plant roots are the subject of one review Plant root growth is one of the phenomena considered in the critical analysis of the practice of zero-tillage made by scientists who have devoted much of their research efforts to this cultural practice Phosphorus accumulation in streams and lakes fed by runoff from agricultural lands is the subject of another review The need to prevent environmental contamination from agricultural chemicals is considered The genesis and management of acid sulfate soils, which occupy millions

of acres of coastal areas in warm and hot humid climates are discussed These soils are important especially to the rice growing areas of the world The international focus of this journal is maintained not only by the nature of the subjects covered but by the selection of authors to write the reviews Food production is truly an international problem to which crop and soil scientists throughout the world are addressing their attention

N C BRADY

xi

This Page Intentionally Left Blank

J C Ryden,' J K Syers,' and R F Harris

Department of Soil Science, University of Wisconsin,

Madison, Wisconsin

I Introduction

11 Terminology

B Forms of

111 Factors Affecting in Runoff and Streams B Chemical Factors

Phosphorus Loads in Runoff and Streams A B Runoff from Forest Watersheds

C Runoff from Agricultural Watersheds D E Urban Runoff

Impact of Phosphorus Carried in Streams on Standing Waters

VI Prcsent Status and Outlook

References

IV Influence of Point Sources on Phosphorus in Streams

Runoff from Land Associated with Ani

V

1

2

2

4

4

4

7

20

21

22

25

32

33

37

38

41

I Introduction

Increasing evidence suggests that phosphorus ( P ) in surface waters is

a primary factor controlling the eutrophication of water supplies (Ohle, 1953; Mackenthun, 1965; Stewart and Rohlich, 1967; Vollenweider, 1968; Lee, 1970) Assessment of the relative contribution of the different

sources of P to surface waters (Fig 1 ) is of critical importance for imple-

mentation of control measures to prevent or reverse P-induced eutrophica- tion Although the importance of runoff and streams as major sources of

P to standing waters is well recognized, little attempt has been made to

differentiate between and quantify the P forms in runoff and streams which are of potential importance with respect to their impact on the biological productivity of standing waters Furthermore, little emphasis has been placed on the reactions that may occur between dissolved inorganic P and

Present address: Department of Soil Science, Massey University, Palmerston North, New Zealand

1

2 J C RYDEN, J K SYERS, AND R F HARRIS

the solid phases with which it is in contact in runoff and streams, as pointed out by Taylor ( 1967) and Biggar and Corey (1969)

Critical concentration limits have been suggested for P in surface waters which, if exceeded, will lead to excessive biological productivity (Sawyer,

1947; Mackenthun, 1968) In this review, however, rather than empha-

sizing critical concentrations, P in runoff and streams will be discussed mainly from the standpoint that any P load constitutes a potential increase

in the P fertility of surface waters

II Terminology

A HYDROLOGY AND PHOSPHORUS SOURCES

This review will use essentially the definitions proposed by Langbein and Iseri ( 1960)

Watershed (drainage basin; catchment area) A part of the surface of

the earth that is occupied by a drainage system, which consists of a surface stream, or a body of standing (impounded) surface water, together with all tributary surface streams and bodies of standing surface water

Stream A general term for a body of flowing water In hydrology the

term is usually applied to the water flowing in a natural channel

Stream flow The discharge (of water) that occurs in a natural channel Runoff That part of precipitation that falls on land and ultimately ap-

pears in surface streams and lakes Runoff may be classified fur- ther according to its source

Surface runoff (overland flow) That part of rainwater or snowmelt

which flows over the land surface to stream channels Surface runoff may also enter standing waters directly or be consolidated into artificial chan- nels, e.g., storm sewers in urban areas (urban runoff), before entering a stream or body of standing water

Subsurface runoff (storm seepage) That part of precipitation which in- filtrates the surface soil and moves toward streams as ephemeral, shallow, perched groundwater above the main groundwater level In many agricul- tural areas subsurface runoff may be intercepted by artificial drainage sys- tems, e.g., tile drains, accelerating its movement to streams

Groundwater run08 (base runoff) That part of precipitation that has passed into the ground, has become ground water, and is subsequently discharged into a stream channel or lake as spring or seepage water

In addition to runoff, the other potential contributors to streams and standing waters are precipitation incident on the water surface and indus-

trial and sewage effluents (Fig 1 )

McCarty (1967) and Vollenweider ( 1968) have made a useful division

of sources of P to surface waters based on the ease of quantification

Point sources enter at discrete and identifiable locations and are there-

fore amenable to direct quantification and measurement of their impact

on the receiving water Major point sources include effluents from indus-

FIG 1 Schematic representation of the relationships between phosphorus sources and runoff, streams, and standing waters

trial and sewage-treatment plants (Fig 1 ) Diffuse wurces may be defined

as those which at present can be only partially estimated on a quantitative basis and which are probably amenable only to attenuation rather than

to elimination Diffuse sources require the most investigative attention Vollenweider ( 1968) further divided diffuse sources into:

1 Natural sources such as eolian loading, and eroded material from virgin lands, mountains and forests

2 Artificial or semiartificial sources which are directly related to human activities, such as fertilizers, eroded soil materials from agricultural and urban areas, and wastes from intensive animal rearing operations The loads of P imparted to runoff and streams from natural diffuse sources provide a datum line against which the magnitude of P loads from artificial sources may be compared

4 J C RYDEN, J K SYERS, AND R F HARRIS

B FORMS OF PHOSPHORUS

In natural systems, P occurs as the orthophosphate anion (Pod3-) which may exist in a purely inorganic form (H2P0,- and HP0,2-) or be incor- porated into an organic species (organic P ) Under certain circumstances inorganic orthophosphate may exist as a poly- or condensed phosphate

A secondary distinction is made between particulate and dissolved forms

of P, the split conventionally being made at 0.45 pm

Other terminology used is as follows:

Total P All forms of P in a runoff or stream sample (dissolved and particulates in suspension) as measured by an acid-oxidation treatment (e.g., acid ammonium persulfate)

Dissolved inorganic P P in the filtrate after 0.45 pm separation deter- mined by an analytical procedure for inorganic orthophosphate

Organic P P that may be determined within the dissolved and particu- late fractions by the difference between total P and inorganic P

Ill Factors Affecting the Dynamics of Phosphorus in Runoff and Streams

Before evaluating the magnitude of various P sources in terms of the loads of P in runoff and streams, and the extent to which previous studies

of P loadings enable an adequate definition of P sources, it is important

to understand the physical and chemical factors affecting the dynamics of

P in runoff and streams These factors determine not only the movement

of P into runoff and streams, but also its distribution between the aqueous and particulate phases

in subsurface and groundwater runoff will be influenced by the time of contact with any component in the soil profile capable of interacting with dissolved P in the percolating water and by the concentration of dissolved

P that the soil components maintain in the soil solution Time of contact between the percolating solution and any soil component will in turn de- pend on the rates of infiltration and percolation into and through the soil

Some of the theories developed to describe water movcment in soils can

be applied to evaluate the potential loss of P from various soil types as

a result of subsurface runoff Gardner (1965) developed equations to de- scribe the movement of nitrate in the soil profile due to leaching The chemical interactions that occur between dissolved inorganic P and soil components (discussed later), when water percolates through the soil, must also be taken into consideration Inclusion of a term in the equations developed by Gardner (1965) to describe the relationship between P in particulate and aqueous phases is therefore necessary This could take the form of a linear adsorption isotherm relevant to the concentrations of dis- solved inorganic P maintained in the solution of a particular soil Biggar and Corey (1969) have also reviewed the literature on infiltration and percolation of water in agricultural soils as it pertains to nutrient movement

The movement of solid phase material in contact with natural waters

occurs during surface runoff and in streams The amounts of solid material capable of entering surface runoff will depend on the intensity of rainfall, physical and chemical attachment between various solid components, and the amounts and energy of runoff waters (Guy, 1970) It is the energy

of surface runoff or stream water, however, that governs the amounts of

a specific size fraction of particulate materials which will remain in suspen- sion during water flow

The primary source of particulate material to surface runoff and streams

is eroding soil (Guy and Ferguson, 1970), although in urban areas with little ongoing development, particulates may be dominated by specifically urban detrital material (e.g., street litter and dust) and organics derived from urban vegetation

The total on-site losses of soil due to sheet and rill erosion are not neces- sarily delivered to streams The amount of sediment that travels from a

point of erosion to another point in the watershed is termed the sediment yield (Johnson and Moldenhauer, 1970) Consequently the Universal Soil Loss Equation used to predict field soil losses on an average annual basis (Wischmeier and Smith, 1965) must be corrected when used to predict sediment loads in streams because deposition of particulates may occur

on the land surface as a result of slope variations before surface runoff reaches a stream It is for this reason that estimates of soil loss in surface runoff from sites within a particular watershed cannot be translated into total P losses through a knowledge of the total P content of the soil, if the P loss is to be related to P enrichment of surface waters

An associated complication arises from the fact that soil P is primarily associated with the solid phase As soil erosion is a selective process with respect to particle size, selectivity has been observed for P loss in surface

6 J C RYDEN, J K SYERS, AND R F HARRIS

runoff The extent of the selectivity depends on the particle sizes with which most of the soil P is associated This observation has led to the concept of enrichment ratios (ER) , which for P are calculated as the ratio

of the concentration of P in the particulate phase of surface runoff to the concentration of P in the source of the particulate phase This effect was first considered by Rogers (1941), who observed ER values of 1.3 for total P and 3.3 for “0.002 N H,SO, extractable” P for a silt loam situated

on a 20-25% slope Other values range from 1.5 to 3.1 for total P

(Knoblauch et al., 1942; Neal, 1944; Stoltenberg and White, 1953), whereas Massey and Jackson (1952) observed values between 1.9 and 2.2 for “water-soluble plus pH 3 extractable” P for silt loams in Wisconsin The selective nature of surface runoff with respect to P is due to selective removal of fine soil particulates as a result of the energy limitations of

runoff and the fact that a large percentage of total soil P is frequently associated with clay-sized material (Scarseth and Chandler, 193 8; Williams and Saunders, 1956; Syers et al., 1969) Greater selectivity of fines and

consequently particulate P will occur as the energy of surface runoff de- creases Stoltenberg and White (1953) observed that as precipitation dis-

posed of through surface runoff decreased from 70 mm to 0.25 mm per hour, the clay content of eroded material from a soil with a clay content

of 16-18% increased from 25% to 60% This has obvious implications

in relation to the nature of the sediment load carried by a stream and the interactions of P between the solid and aqueous phases, particularly during periods of surface runoff It should be pointed out, however, that although the P content of the sediment load may increase as surface runoff dimin- ishes, as may be predicted from the work of Stoltenberg and White (1953), the total P load may not change, or may even decrease, owing

to lower sediment loads

The particulate material carried in streams may be divided into bed load and wash or suspended load The bed load, which may also have a con- tribution from existing stream sediment, is that which moves along or close

to the stream bed, whereas the wash load is maintained in the flow by turbulence (Johnson and Moldenhauer, 1970) By inference from the se- lectivity of surface runoff for fine soil particulates, the wash load will be high during surface runoff events Furthermore, Johnson and Moldenhauer (1970) suggested that the wash load travels at about the same velocity

as the water with which it is in contact Consequently, P associated with the clay- and silt-sized particulates constituting the wash load will move between any two points in the stream profile at the same speed as the ambient dissolved forms of P

Increased turbulence in streams during high flow, or arising from an increasing gradient, will tend to maintain in suspension particle sizes more

characteristic of the bed load, and may even resuspend existing stream bed sediment In a study of total P loads in the Pigeon River, North Caro- lina, Keup (1968) noted that an increase in gradient from 2.81 to 4.35 m/km, over which no tributaries entered the main stream, resulted in a

90.8 kg/day increase in the total P load carried

It appears that in the majority of cases a large proportion of particulate

P in streams arises from soil erosion Phosphorus may be stored in stream bed sediments, but unless the stream is actively aggrading, the amount of

P stored will be less than the inflow (Keup, 1968) That which is stored

is liable to resuspension and transport owing to turbulence during periods

a Inorganic P O n the basis of solubility product criteria, it has been

postulated that discrete phase crystalline Fe and A1 phosphates exist in

noncalcareous soils (Kittrick and Jackson, 1956; Hemwall, 1957; Chakra- vart and Talibudeen, 1962) The general occurrence of discrete Fe and A1 phosphates seems doubtful on the basis of the ion product data pre- sented by Bache (1964) and the experimental observations of Hsu (1964) It is now generally accepted that secondary inorganic P in many soils exists primarily in association with oxides and hydrous oxides of Fe and Al, as surface-bound forms or within the matrices of such components However, that discrete Fe and A1 phosphates are formed as temporary phases in the vicinity of phosphate fertilizer particles due to conditions

of localized high acidity and P concentration is well established (Lindsay and Stephenson, 1959; Huffman, 1969) Such compounds will not be stable as the dissolved inorganic P concentration in the soil solution or aqueous portion of other soil-water ecosystems decreases

The calcium phosphate mineral, apatite (Shipp and Matelski, 1960) and calcic fertilizer-soil reaction products (Huffman, 1969) have been identi- fied in soils The amounts of apatite are appreciable only in weakly

weathered soils (Williams et al., 1969), as predicted by the weathering

indices of Jackson ( 1969) Calcic fertilizer-sail reaction products may be present in neutral and calcareous surface soil horizons, and their impor- tance in maintaining high concentrations of dissolved inorganic P in soil-water ecosystems should not be overlooked

8 J C RYDEN, J K SYERS, AND R F HARRIS

Consequently three basic forms of inorganic P may exist in unfertilized

soils (Syers and Walker, 1969; Williams and Walker, 1969): apatite,

which is a discrete phase P compound; P sorbed on the surfaces of Fe,

Al, and C a soil components (nonoccluded); and P present within the matrices of Fe and A1 components (occluded) In fertilized soils, a variety

of P fertilizer-soil reaction products may exist as transient phases As the solubility product of pure apatite in water is low (0.03 pg per milliliter

at pH 7, Stumm, 1964) and the P held within the matrices of Fe and

A1 components is virtually chemically immobile, except under reducing conditions in the case of Fe, major emphasis should be directed toward the reactions involving P in solution and that sorbed on the surfaces of

Fe, Al, and Ca components as well as the release of P due to dissolution

of fertilizer-soil reaction products

b Organic P Elucidation of the composition of soil organic P is re- stricted by lack of extractants capable of removing organic P from soils

in a relatively unaltered form and by the inadequacy of current methods for mildly degrading extracted organic P-organic matter complexes Exist- ing data indicate that most of the organic P in soils is associated, in an ill-defined manner, with the humic and fulvic acid complex of soil organic

matter (Anderson, 1967) Of the specific forms of organic P that have

been identified in soils, inositol phosphates are present in largest relative

amounts, comprising up to 60% of the total organic P (Anderson, 1967; Cosgrove, 1967; McKercher, 1969) Other specific organic P compounds are present in soil in much lower quantities: nucleic acids account for 5-lo%, and other phosphate esters, such as phospholipids, sugar phos-

phates, and phosphoproteins, for less than 1-2% (McKercher, 1969)

2 Sorption of Dissolved P by Soils

Whenever water containing a particular concentration of dissolved P

comes into contact with soil material, there is a possibility for sorption, desorption, or dissolution reactions to take place The types of reactions are the same regardless of whether they occur under conditions existing

in the soil profile, surface runoff, or streams Although in some cases bio- logical assimilation may initially affect the distribution of P between dis- solved and particulate phases of soil-water systems, the distribution of P between these phases will be determined by the nature of the inorganic

particulates and the concentrations of dissolved P in solution (Keup, 1968;

McKee et al., 1970; Ryden et al., 1972b)

a Inorganic P It has been demonstrated that the uptake or sorption

of P from solution by soils is significantly related to the presence of short- range order (amorphous) oxides and hydrous oxides of Fe and A1 (Wil-

liams et al., 1958; Gorbunov et al., 1961; Bromfield, 1965; Hsu, 1964; Saunders, 1965; Syers et al., 1971) Furthermore, “pure” oxides and hy-

drous oxides of Fe and Al, and short-range order aluminosilicates have also been shown to be particularly effective in the sorption of inorganic

P from solution (Gastuche et al., 1963; Muljadi et al., 1966; Hingston

et al., 1969) The sorption of inorganic P by F e and A1 oxides and hydrous

oxides is known to be rapid, as is the sorption of P by soils Furthermore,

short-range order Fe and A1 oxides and hydrous oxides are ubiquitous in

soils (Hsu, 1964), their relative amounts depending on parent material, climatic and drainage conditions, and occur mainly as coatings on other soil components Shen and Rich (1962) and Jackson (1963) have noted

the occurrence of A1 hydroxypolymers and Dion (1944), and Roth et al

( 1969) have reported the presence of F e oxide and hydrous oxide coatings

on clay mineral surfaces Such coatings, in conjunction with the greater surface area of the clay fraction compared to that of the other particle-size fractions in a soil, explain the observation of Scarseth and Chandler (1938) that up to 50% of the total P in soils may be associated with the the clay fraction, as well as the enrichment ratio effect for P as a result

of soil erosion

Attempts have been made to correlate P sorption with the clay content

of soils (Williams et al., 1958) Correlations between P sorption and clay content after removal of Fe and A1 oxides and hydrous oxides often have been poor Better correlations may be expected if P sorption is related

to the content of water-dispersed clay The sorption of P by water-dis- persed clay and silt of soils has obvious implications to reactions occurring between dissolved and particulate P in surface runoff and streams Sorption of inorganic P by CaC03 has also been demonstrated (Cole

et al., 1953) The nature of the surfaces of calcite in calcareous soils may

be very different from those of pure calcite (Buehrer and Williams, 1936;

Lahav and Bolt, 1963; Syers et al., 1972)

The sorption of dissolved inorganic P by soils may be described by sorp- tion isotherms similar to that shown in Fig 2 Numerous workers have also shown that sorption may be described by some of the adsorption iso- therms developed to describe gas adsorption by solids (Russell and Pres- cott, 1916; Olsen and Watanabe, 1957; Rennie and McKercher, 1959;

Syers et al., 1973) Similar observations have been made for the sorption

of inorganic P by soil components such as kaolinite and short-range order

Fe and A1 oxides and hydrous oxides (Gastuche et al., 1963; Muljadi et

al., 1966; Kafkafi et al., 1967) Although these studies have been useful

in describing relationships between various soils and soil components with respect to their P sorption capacities, they have provided little information regarding P sorption behavior from solutions containing the low dissolved inorganic P concentrations characteristic of most soil-water ecosystems, largely because of the high levels of added P used (Ryden et al., 1972b) Furthermore, Syers et al (1973) obtained two linear Langmuir relation-

10 J C RYDEN, J K SYERS, AND R F HARRIS

sorbed ( f )

A P o n sol I

released (3

FIG 2 Typical isotherm for the sorption of added inorganic phosphorus by

a soil E = equilibrium P concentration (From White and Beckett, 1964.) ships which intersected at equilibrium P concentrations varying from 1.5

to 3.2 pg P/ml, for three contrasting soils-an observation that probably invalidates interpretations of P sorption made from many previous studies where high levels of added P were used

The study of White and Beckett (1964), conducted at initial dissolved inorganic P concentrations, comparable to those existing in soil-water eco- systems, provides a useful basis for understanding the interactions between

aqueous and particulate phases of P in runoff and streams Figure 2 illus-

trates the principle of the approach used White and Beckett (1964) de- fined the intersection of the P sorption isotherm and the abscissa, the

“equilibrium phosphate potential” ( 5 p C a + pH,PO,) , abbreviated to

“equilibrium P concentration” by Taylor and Kunishi ( 1971 ) The inter- section is equivalent to the inorganic P concentration in the ambient aque-

ous phase when there is no net sorption or release of P, i.e., AP = 0 This

is a point of reference which provides a predictive estimation of sorption

or release of P should the P concentration in solution change Furthermore, the average slope of the sorption curve over a given final P concentration range provides information on the ability of the soil to maintain the P concentration at the equilibrium P concentration The steeper the slope, the closer will the final P concentration be to the equilibrium P concentra- tion The slope of the curve, although not related to total P sorbed, is related to the extent to which that soil may sorb P over the concentration range considered The potential of this approach in predicting the chemical mobility of P in soil-water systems is clearly evident and has been used

with regard to streams by Taylor and Kunishi (1971) and Ryden et al

(1972a,b) for rural and urban soils, respectively

The desorption of sorbed P from soils is not as simple as may be in- ferred from the sorption-release relationships obtained by White and

Beckett (1964) In fact very few studies have been reported regarding

the desorption of sorbed P, and those reported by Syers et al (1970)

and Ryden et al (1972a), involved desorption following sorption of P

from solutions containing P concentrations in excess of those commonly found in soil-water ecosystems

In studies involving the sorption of P by kaolinite from solutions con- taining realistic inorganic P concentrations, Kafkafi et al (1967) observed

that initially all the sorbed P was isotopically exchangeable During a sub- sequent washing or desorption step, however, a portion of the sorbed P

became nonexchangeable, or “fixed,” this portion being dependent upon the amount of P sorbed, the number of washings, and the nature of the previous P sorption cycle Sorption of P was represented by either one- step sorption from a range of solutions of different initial P concentration

or by successive additions of small amounts of dissolved inorganic P Both these types of P sorption, as well as an effect analogous to washing, could occur in soil-water ecosystems

6 Organic P Although the mechanisms involved in the retention of organic P by soils have not been established fully, there is evidence that inositol hexaphosphate, and possibly other organic P compounds, are re- tained by a precipitation rather than a sorption reaction Nevertheless, re- moval of dissolved organic P from solution appears to be a rapid process

Pinck et al (1941 ) reported that many commonly occurring water-soluble organic phosphates, e.g., salts of glycerophosphate, hexose diphosphate, and nucleic acids, become nonextractable with water at almost the same rate and as completely as dissolved inorganic P

The retention of water-soluble organic P by sorption reactions may

occur by at least two basically different mechanisms (Sommers et al.,

1972) Goring and Bartholomew (1950) observed that removal of “free iron oxides” considerably reduced the amount of fructose 1,6-diphosphate sorbed by subsoil material, suggesting that the sorption of organic P may occur through orthophosphate groups by a similar mechanism to that for inorganic P It is also possible that organic P can be retained by interaction

of the organic moiety of the phosphate ester with inorganic soil compo- nents For example, nucleic acids and nucleotides are protonated at pH

5 (Jordan, 1955) and could consequently be retained on clay surfaces

by displacement of exchangeable cations Furthermore, physical adsorp- tion, also through the organic portion of the molecule, is possible, particu- larly if the molecular weight of the compound is high, as suggested by Greenland (1965) In such cases retention is weak and is accomplished

by van der Waals and ion-dipole forces Greaves and Wilson (1969) have implicated physical adsorption in the retention of nucleic acids by mont- morillonite It is also possible that retention occurs indirectly through other

12 J C RYDEN, J K SYERS, AND R F HARRIS

soil organic compounds such as fulvic and humic acids after interaction

of organic phosphates with these species (Martin, 1964)

The desorption of sorbed organic P has not been extensively studied The hypothesis that inorganic P added to soils displaces sorbed organic

P to solution (Latterell et al., 1971) was not supported by the data pre-

sented by Wier and Black (1968) Although organic P may be leached from soils, it appears that a large proportion of that removed may not

be in a dissolved form After incubating sucrose with ammonium nitrate

in the upper portion of a calcareous soil, Hanapel et al (1964) found that most of the organic P removed by leaching was present in a particulate rather than a dissolved form

3 Chemical Aspects of P in Subsurface and Groundwater Runoig

Losses of P in subsurface and groundwater runoff have been considered minimal in the past, but, as will be discussed later, such losses can amount

to a significant proportion of losses from agricultural land, and possibly

a major proportion from forest lands The supposition that P losses in sub- surface and groundwater runoff are low probably stems from the concept

of P immobility based on the P sorption properties of soils using added inorganic P concentrations far in excess of those normally present in the soil solution

It is of interest to note that many of the reported mean concentrations

of dissolved inorganic P in subsurface runoff are within the range of values expected to be maintained in the soil solution Pierre and Parker (1927) reported values ranging from 0.020 to 0.350 pg P/ml, with an average

of 0.090 pg/ml, for several surface soils from the southern and midwestern states of the United States These workers also noted that dissolved inor- ganic P concentrations could be maintained at a fairly constant level Bar-

ber et al (1963) reported similar values for the upper 1 5 cm of 87 soils

from the midwestern United States, with an average of 0.180 pg of P per milliliter; the frequency distribution of the values obtained, however, sug- gested a mode of between 0.040 and 0.060 pg of P per milliliter

As water percolates through the soil profile, there tends to be a “chemi- cal sieving” of dissolved inorganic P (Black, 1970) This arises as a result

of the sorption of inorganic P by soil components The low concentrations

of P found in groundwater runoff, which has experienced the maximum effects of deep percolation with concomitant increase of contact with P-deficient particulates of the subsoil, are undoubtedly a direct result of the chemical sieving effect The principle of this effect is illustrated by

other data presented by Barber et al (1963) For the same 87 soils men-

tioned previously, the average dissolved inorganic P concentration at a depth of 46-61 cm was 0.089 pg/ml, less than half that for the upper

0-15 cm Another illustration is observed in results presented by Ryden

et al (1972a) for the P sorption properties of successive soil horizons

of a Miami silt loam profile The concentrations of dissolved inorganic

P maintained in solution after shaking with solutions of different initial added inorganic P concentrations at a solution: soil ratio of 40: 1 are given

in Table I

Dissolved Inorganic Phosphorus (P) Concentrations Maintained

by Soil IIorizoiis of Miami Silt Loam after Equilibration

with Solutions of Different Initial Added Inorganic

P Concentrationsn Initial P conc Final P coiic

B1 15-38 0 4 7 1 0 0 3 0

3C1 56-66 0 0 3 0 0 0 0 7

~

Data extrapolated from Ryden et (11 (197‘2a)

The concentration of dissolved inorganic P in subsurface and ground- water runoff will depend on the nature and amounts of P-retaining com- ponents in the profile, the surface area exposed to percolating waters, and the ease of percolation which affects the contact time of dissolved inorganic

P with the retaining components In studies of P leaching through columns

of organic soils in the laboratory, Larsen et al (1958) observed that P

retention, measured by srP autoradiographs, was closely correlated with the total hydrous Fe and A1 oxide (“sesquioxide”) content Similarly, losses of P due to leaching through a deep siliceous sandy soil were demon-

strated in W Australia by Ozanne (1963) When 225 kg/ha of 32P-labeled

superphosphate was broadcast during winter on a fallow sandy soil, over 50% of the P had penetrated to more than 1 m below the surface within

38 days, during which 230 mm rain had fallen Ozanne (1963) also dem-

onstrated that the potentially large losses of P to subsurface and ground- water runoff from sandy soils compared to that from loamy soils were due to quantitative rather than qua1itativ.e differences in P-retaining components

Although major emphasis has been placed on P losses in surface runoff,

it appears that losses of P to subsurface and groundwater runoff, although

of little significance from an agricultural standpoint, may under certain conditions constitute a significant loss of P from agricultural watersheds

in terms of the P enrichment of surface waters, as will be discussed

14 J C RYDEN, J K SYERS, AND R F HARRIS

later Losses of P to subsurface and groundwater runoff are even more difficult to evaluate than those in surface runoff and demand further inves- tigative attention

4 Chemical Aspects of P in Streams

As discussed previously, surface runoff from agricultural land constitutes

a heterogeneous and relatively short-lived system Any attempt to consider the distribution and chemical mobility of P between solid and aqueous phases before entry into the receiving stream would be pointless as a new and more homogeneous system is rapidly established Surface runoff in urban areas is somewhat different because in most cases it is channelized shortly after origin by alteration of surface drainage patterns; under such circumstances it is analogous to a stream in an artificial channel Conse- quently, the chemical mobility of P will be discussed from the standpoint

of the stream environment

The potential of suspended particulates derived from eroding soil to modify the dissolved inorganic P concentration of streams has been sug- gested by Taylor ( 1967) and Biggar and Corey (1969) Wang and Brabec (1969) also implied that inorganic P was sorbed by suspended particulate material from observations of dissolved inorganic P concentrations in the Illinois River at Peoria Lake

An evaluation of the possible effects of eroded soil materials on the dis- solved inorganic P concentrations of streams may be obtained from P

sorption studies (Taylor and Kunishi, 1971; Ryden et al., 1972a,b) It

is essential, however, that conditions realistic of those existing in streams

are used if meaningful results are to be obtained (Ryden et al., 1972a)

Widely differing interpretations can be made as solution: soil ratios and initial dissolved inorganic P concentrations are changed from those conven- tionally used in P sorption studies to those realistic in terms of the stream environment (Fig 3 a - c ) The data in Fig 3a suggest that inorganic P released from the A1 horizon, which contained a P fertilizer-soil reaction product, would be largely resorbed by the noncalcareous B1 horizon and

to some extent by the calcareous 3C1 horizon, should the horizons erode together Sorption studies employing low initial added inorganic P concen- trations and a wide (400: 1 ) so1ution:soil ratio (Fig 3c) indicate that

the B1 horizon has a much lower ability to remove dissolved inorganic

P from solution than expected, this being equal to or only slightly greater than that of the 3C1 horizon In fact for mixtures of varying ratios of

A1 and B l , and A1 and 3C1 horizons, it was found (Ryden et al., 1972b)

that the latter mixtures were able to maintain lower dissolved inorganic

P concentrations than the former The conditions used by Ryden et al

(1972a,b) to predict the potential of eroding soils to modify the dissolved

Final dissolved inorganic P Concentration Wgll)

FIG 3 Sorption of added inorganic phosphorus by horizons of a Miami silt

loam profile from solutions of varying initial dissolved inorganic P concentrations

and at varying so1ution:soil ratios ( a ) High added P (0-6 pg/ml) and narrow so1ution:soil ratio (50: 1 ) ( b ) Low added P (0-0.2 pg/ml) and narrow solution:soil

ratio ( 4 0 : l ) (c) Low added P (0-0.2 pg/ml) and wide so1ution:soil ratio ( 4 0 0 : l )

[From Ryden et al (1972a), reproduced with permission of the American Society

of Agronomy.]

inorganic P concentrations of streams, gave results comparable to those obtained in simulated stream systems using a solution: soil ratio of 1000: 1 This is equivalent to a sediment concentration of 1000 mg/liter, which lies well within the range of values cited by Guy and Ferguson (1970) and Johnson and Moldenhauer ( 1970)

The P sorption studies reported by Taylor and Kunishi (197 1) and

Ryden et al (1972a,b) involved closed systems, i.e., soil in contact with

the same aqueous phase This may be justified on the grounds that the wash load of a stream travels at the same velocity as the water in which

it is suspended (Johnson and Moldenhauer, 1970), as discussed previously

1 6 J C RYDEN, J K SYERS, AND R F HARRIS

Sorption studies may be used to provide reasonable estimates of dis- solved inorganic P concentrations in streams, under various flow condi- tions, draining rural watersheds Taylor and Kunishi (1971 ) observed that dissolved inorganic P concentrations during base flow of a stream draining

a small agricultural watershed in Pennsylvania, were in the range of 0.040

to 0.060 pg P/ml, values which were close to those predicted from P sorp- tion studies using stream bank sediment and subsoil material During pe- riods of surface runoff, predicted dissolved inorganic P concentrations would be in excess of 0.200 pg of P per milliliter for the surface soil used

by Taylor and Kunishi (1971) and 0.100 pg of P per milliliter for that used by Ryden et al (1972a) due to release of P from eroded surface soil; however, predictions from the work of Taylor and Kunishi (1971) are based on the use of a narrow (10: 1 ) so1ution:soil ratio The ability

of eroding stream bank material or resuspended stream bed sediment to resorb inorganic P released to solution should not be ignored (Taylor and Kunishi, 197 1 )

In a more recent study of the same watershed in Pennsylvania, Kunishi

et al (1972) observed that during a heavy summer rainstorm only 31%

of the total “available” P (total dissolved plus resin-extractable P on the suspended sediment) was in the resin-extractable form in a stream draining

an agricultural subwatershed At the outflow of the main watershed, how- ever, over 50% of the total “available” P was in the resin-extractable form Kunishi et al (1972) suggested that for this watershed, as suspended ma- terial moves downstream and mixes with material from other parts of the watershed as well as that eroded from the stream banks, dissolved P is actively sorbed During a second less intense storm, however, when stream bank erosion was less severe, the proportion of total “available” P asso- ciated with the sediment was virtually the same at both monitoring stations

A similar hypothesis might also explain the observation of White ( 1 972)

at Taita, New Zealand, that the concentration of dissolved inorganic P at the outflow of small watershcds during base flow was lower than that re- corded for groundwater seepage giving rise to the stream flow

It is important to distinguish between the quantities of various types

of soil materials expected to enter streams in urban as opposed to agricul- tural surface runoff In agricultural areas, surface runoff will carry pri- marily surface soil material to receiving streams Surface soils may contain

P fertilizer-soil reaction products capable of producing significant increases

in dissolved inorganic P concentrations, due to their dissolution (Ryden

et al., 1972a) In urban areas, however, land under development, which

is prone to severe erosion, is frequently graded, exposing some or all hori- zons of the area profile to potential erosion Dissolved inorganic P concen- trations of receiving streams in urban areas may be sufficiently high that

the addition of eroded soil material may cause a reduction in the dissolved inorganic P concentration

An approach similar to that used by Taylor and Kunishi (1971) and

Ryden et al (1972a,b) could be used to identify other diffuse sources

of potential P enrichment within a watershed The approach would be par- ticularly useful for estimating the potential of various forms of urban detrital material to influence the dissolved inorganic P concentrations of surf ace runoff

One diffuse source of considerable importance is the leachate from leaves, particularly during the autumn An appreciable percentage of the total P in leaf tissue may be in a water-soluble form Ash leaves may con- tain 62% of total P as water-soluble inorganic P (Nykvist, 1959) Cowen

and Lee (1972) observed that 44 and 120 pg soluble inorganic P per gram

air-dry weight of fallen oak and poplar leaves, respectively, could be leached by 1 liter of distilled water percolating at a rate of 8.4 ml per minute Greater amounts of P were released from oak leaves during con- secutive leaching cycles and after fragmentation of whole leaves Similar

experiments were conducted by Timmons et al (1970) using agricultural

crop residues These were leached in a fresh condition and after drying, and freezing and thawing cycles The data suggest that the leaching of crop residues is most likely to contribute to the dissolved inorganic P concentra- tion of streams during spring thaw in certain areas when, after numerous freezing and thawing cycles, the residues will be carried over frozen ground

in surface runoff When greater infiltration can occur, a portion of the leached P may be retained in the soil due to sorption

5 P Chemistry of Stream-Bed Sediment

Little is known of the chemistry of stream-bed sediment although it is conceivable that it is similar to that of the subsoil of the surrounding area (Taylor and Kunishi, 1971 ) Consequently, P sorption studies using sub- soil material may provide some information on the role of stream-bed sedi- ment in regulating the dissolved inorganic P concentration due to its sus-

pension during turbulence This would be particularly true in watersheds with little contribution to stream-bed sediment as a result of surface runoff

In watersheds where surface runoff is a regular occurrence, however, stream-bed sediment is expected to have a significant contribution from surface horizon soil material, and the latter could contribute to base flow concentrations of inorganic P

Care should be taken, however, in the extension of the P sorption prop-

erties of field soils to stream-bed sediment Hsu (1964) observed that the

amount of inorganic P sorbed by soil after storage for 1 year in a con-

tinuously wet condition, increased from 69 to 99 pg of P per gram of soil

18 J C RYDEN, J K SYERS, AND R F HARRIS

The increased sorption was attributed to release of Fe to solution from crystalline phases due to the development of localized reducing conditions during storage, and reprecipitation of “ferric hydroxide” on contact with more aerobic conditions The redox status of stream-bed sediments does not appear to have been studied, but it is reasonable to suggest that reduc- tion occurs at depth in the sediment with the possibility of crystalline ferric components being transformed to short-range order ferrous forms The im- portance of short-range order oxides and hydrous oxides of Fe in the sorp-

tion of inorganic P has already been discussed The possible transformation

of Fe from crystalline to short-range order forms represents the first stage

of the more aggressive transformations which occur in lake sediments

under anaerobic conditions (Shukla et al., 1971)

The observation of Kafkafi et al (1967) that the washing of kaolinite,

on which P had been sorbed, produces a “pool” of nonexchangeable P

is also of direct relevance to the P chemistry of stream-bed sediments, as- suming a similar effect occurs Stream-bed sediment with associated sorbed

P could undergo a series of steps equivalent to sequential washing due

to resuspension and settling as a result of minor turbulence The observa-

tions of Kafkafi et al (1967) suggest that sorbed P could become progres- sively less exchangeable and may constitute an essentially permanent re- moval of dissolved inorganic P from streams

When stream-bed sediment contains eroded fertilized soil materials,

however, a different situation may prevail Ryden et al (1972a) showed

that release of P from a surface soil horizon by repeated washing with

P-free 0.1 M NaCl initially followed first-order kinetics, suggesting that

release was due to the dissolution of solid phase P, probably a fertilizer-soil reaction product

6 Forms of P in Runoff and Streams

In many studies concerned with various aspects of P in runoff and streams there has been a tendency to measure total P The measurement

of total P discharged by streams does not provide any indication of the amounts of P available for aquatic plant growth Consequently, the forms

of P measured in streams that enter a lake or reservoir are of direct impor- tance in assessing the impact of runoff- and stream-derived P on a body

of standing water Dissolved inorganic P is one of the obvious choices be- cause this form of P is directly available for biological utilization Objec- tions to the measurement of dissolved inorganic P, as it is conventionally determined, have been raised by Frink (1971 ) on the basis that distinction between dissolved and particulate forms is based on filtration through a

0.45 pm filter Although it is possible that filtration does not strictly differentiate between dissolved and particulate P, it provides a more

realistic measure in terms of the effects of runoff- and stream-derived P

on the biological productivity of standing waters than the measurement

of total P

Vollenweider ( 1968) has also pointed out the necessity to distinguish

between total P and dissolved forms of P because it is possible that P ex- ports from some watersheds occur mainly in biologically unavailable forms, such as apatite This work showed that P exports from the Alpine portion

of the Rhine Basin amount to 1.45 kg/ha per year As this is mainly in

the form of apatite, however, the contribution of biologically available P

to Lake Constance is relatively small In other regions it appears that a high proportion of particulate inorganic P in streams draining urban and rural watersheds may in fact be apatite Eroding urban soils in the Lake Mendota watershed, Wisconsin, contain between 6 and 80% of the total

inorganic P as apatite, with amounts exceeding 50% in the lower B and C horizons (J K Syers, J C Ryden, and J G Thresher, unpublished data) For the same soil materials, Sagher and Harris (1972) observed that algal

cultures suffered P starvation when the sole P source in the growth medium was C horizon material, indicating the very low biological availability of the P present in apatite

Chemical fractionation schemes have been used to determine the forms

of inorganic P in soils These schemes evolved from the observations of Chang and Jackson (1957) that certain chemical reagents were able to

solubilize inorganic P contained in various synthetic phosphates and phos- phate minerals Recent workers (Bromfield, 1967; Williams et al., 1967,

1971a,b; Syers et al., 1972) have questioned the validity of the separation

of inorganic P into Al-, Fe-, and Ca-bound forms, as proposed by Chang and Jackson ( 1957) Providing that the problems inherent in inorganic

P fractionation schemes are recognized, useful interpretations may be drawn from the data obtained The form of particulate inorganic P which

is expected to have the greatest potential impact on the biological produc- tivity of standing waters is that which is nonoccluded Part of the nonoc- cluded and even some of the occluded inorganic P associated with ferric components is released into solution when anaerobic conditions develop subsequent to sedimentation Appropriate inorganic P fractionation schemes applied to suspended stream sediments may provide a more mean- ingful measure of the forms and amounts of particulate inorganic P carried

in streams As pointed out by Taylor et al (1971), suspended sediment

concentrations are frequently not high enough to provide adequate amounts of xaterial in a manageable volume of water Evaluation of the forms of P in soil materials which are known to be transported to streams

in surface runoff may overcome this problem to some extent In the case

of eroding soil materials, the inorganic P fractionation schemes should not

20 J C RYDEN, J K SYERS, A N D R F HARRIS

be applied to the whole soil, due to the ER effect resulting from erosion Water-dispersed particle-size separates should be used

In spite of the possible errors involved in a dissolved-particulate P split based on filter pore size, it seems that in the majority of cases the most meaningful and useful measurements of P in runoff are dissolved forms, particularly dissolved inorganic P Frequently dissolved forms of P account for a major percentage of total P (Sylvester, 1961; Sullivan and Hullinger,

1969), whereas dissolved inorganic P in many cases constitutes a major proportion of the total dissolved P It should be noted that dramatic changes can occur in the concentration of dissolved inorganic P and other

P fractions after sample collection, even after only a short period of time

(Ryden et al., 1 9 7 2 ~ ) In some cases when samples are not analyzed im- mediately after collection, the only valid P parameter that can be measured

is total P

IV Phosphorus loads in Runoff a n d Streams

The P content of precipitation reflects the amount of P subject to wash- out from the atmosphere at the time of the precipitation event The

amounts of P carried in precipitation rarely exceed 1 kg/ha per year as

total P (Miller, 1961; Weibel et al., 1966; Allen et al., 1968; Gore, 1968)

Weibel et al (1966) reported that the average concentration of total acid-

hydrolyzable P in precipitation falling on Cincinnati, Ohio, was 0.080

pg/ml, whereas Taylor et al ( 197 1 ) reported an average concentration

of 0.020 pg/ml for total dissolved P in precipitation collected at rural Coshocton, Ohio

Data for the P content of precipitation should be viewed with some skepticism unless adequate precautions have been taken to guard against

contamination of the collection vessel (Gore, 1968; White, 1972) White

(1972) found that although rainwater collected over an extended period

indicated a mean dissolved inorganic P concentration of 0.020 pg/ml, a mean concentration of 0.003 pg/ml, based on specific showers, might be

a more accurate estimate

It is difficult to evaluate the effect of P carried in precipitation on P

loads in runoff and streams Phosphorus contained in precipitation which becomes a part of any soil-water ecosystem may undergo considerable change in form, depending primarily on the chemical factors discussed pre- viously, and will become an integral part of the P forms in runoff and streams

Surface runoff water is the carrier of not only the P initially present

in precipitation but also any P which enters surface runoff water because

of chemical interactions or the energy of the water itself Several factors affect the amount and energy of surface runoff water at any particular loca- tion and, therefore, the amount of additional P entering and carried by

it These include nature of land use, extent of vegetative cover, slope, in- tensity of rainfall, and permeability of the land surface

The quantity of precipitation entering subsurface and groundwater runoff is inversely related to that disposed of in surface runoff and evapo- transpiration It is consequently affected by the factors listed above for surface runoff The major portion of P in subsurface and groundwater run-

off is expected to be in dissolved forms If subsurface runoff is accelerated

by artificial drainage systems, however, soil colloids, with associated P, may appear in the water as it enters streams

The P load carried by a stream under given flow conditions will repre- sent the relative contribution of P loads in each of the runoff components,

as well as the influence of any point source of P

A INFLUENCE O F POINT SOURCES ON PHOSPHORUS IN STREAMS

Estimates of the contribution of P to surface waters from domestic wastes in the United States range from 91 x loo to 227 x l o G kg

per year with total P concentrations ranging from 3.5 to 9.0 pg/ml

(McCarty, 1967; Ferguson, 1968) Weibel et al ( 1964) estimated that

P discharged as raw sewage from combined storm sewers in Cincinnati, Ohio, amounted to 3.4 kg/ha per year as total dissolved P In the area

of Madison, Wisconsin, the per capita contribution of P to surface waters

from treated domestic waste was estimated to be 0.544 kg/capita per year

(Sawyer, 1947), whereas an estimate of 1 kg/capita per year was given

by Metzler et al (1958) for Chanute, Kansas The difference between

the estimates of Sawyer (1947) and Metzler et al (1958) may reflect

the increased use of P in domestic products, particularly detergents

The impact of sewage outfall on the dissolved inorganic P concentration

of streams and rivers was studied by Brink and Gustafsson (1970) Their

results are summarized in Table 11 Obviously the impact of the outfall

is dependent on factors which include flow rate of the receiving stream and the P content of the effluent

Under certain agricultural management conditions animal excrement may constitute a point source of P to streams Excrement may enter

streams during surface runoff from feedlot operations or by the cleaning

of milking sheds into open drains The magnitude of these sources of P

will be discussed later

McCarty (1967) was unable to estimate the magnitude of contributions

of P made to streams from industrial wastes The amounts of P discharged

22 J C RYDEN, J K SYERS, AND R F HARRIS

Effect of Sewage Outfall on tllc Dissolved Inorganic Phosphorus

Concentration of the Receiving Water"

IXssolved inorganic P concentration ( p g P/ml)

Data from Brink and Gustafsson (1970)

to streams will depend on the industrial process concerned and local legis-

lation covering the discharge of industrial effluent Mackenthun et al

(1968), for example, estimated that a potato canning factory and a woollen mill contributed 3446 and 835 kg of P per year, respectively, to

the East Branch of the Sebasticook River, Maine

Domestic and many industrial wastes not only supply large amounts of total P to streams but also have a pronounced effect on the concentrations

of dissolved forms of P in the receiving stream Because domestic and in- dustrial wastes are point sources, they are easily recognized within a water- shed and are amenable to direct manipulation

B RUNOFF FROM FOREST WATERSHEDS

A compilation of data from several studies of the quantities of P lost

in streanis from stable forest and woodland watersheds is presented in

Table 111 Exports of P in streams from long-established and stable forest watersheds provide a useful datum line against which losses of P from

other land-use areas may be compared The data in Table I11 show a con-

siderable degree of uniformity Total P losses range from 0.68 to 0.02 kg/ha per year with three out of the four values being less than or equal

to 0.1 kg/ha per year

Only a few measurements have been made of the dissolved inorganic

P concentration of stream water in forested watersheds The values re-

ported by Brink and Gustafson (1970) in Sweden show a mean of 0.015

pg/ml, with this fraction amounting to 33% of the total annual loss of

P The data suggest that total P and dissolved inorganic P concentrations

rarely exceed 0.1 15 and 0.025 pg/ml, respectively Two interesting points arise from the data in Table 111 From the study of a stream draining a

Jaworslii and Hetting (1970) Potomac River Basin Total P 0 1

24 J C RYDEN, J K SYERS, A N D R F HARRIS

woodland area at Coshocton, Ohio, to which no fertilizer P had been ap-

plied for over 30 years (Taylor et al., 1971), it would appear that the woodland is conservative of P The average total soluble P content of rain- fall was 0.020 pg/ml, whereas that in the stream draining the watershed was 0.015 pg/ml The extent of addition of total dissolved P to the wood-

land can be calculated from precipitation data given by Taylor et al (1971 ) ; a value of 0.17 kg/ha per year is obtained This value is more than three times greater than the annual P loss in the stream The con- servative nature of forests for P is further borne out by the fact that the annual contributions of P to the land surface in precipitation, quoted previ- ously, are in most cases considerably greater than annual exports of P in streams from forest watersheds In many cases there is an order of magni- tude difference This hypothesis assumes that data covering the P content

of precipitation are correct

The second point of interest relates to the “background” P concentration

in forest streams The data suggest only minor seasonal fluctuations in P

concentrations, particularly that of dissolved inorganic P As a major por- tion of streamflow is considered to have a groundwater origin (Biggar and Corey, 1969; Johnson and Moldenhauer, 1970), it is conceivable that the dissolved inorganic P load in streams of forested areas is primarily due

to that in groundwater runoff If the reported mean P concentrations of forest streams are compared to those for groundwaters, a marked similarity

is observed Juday and Birge ( 193 1 ) found that the total dissolved P con- centrations of 19 wells in northern Wisconsin, an extensively forested area, ranged from 0.002 to 0.197 pg/ml, with an average of 0.018 pg/ml when the highest value is omitted This mean value is, if anything, slightly higher than the mean concentrations for dissolved fractions of P reported in Table

111 The higher mean concentrations of total P probably arise from sus-

pended inorganic and organic solids that enter streamflow due to turbu- lence, especially during high flow

The minor fluctuations in P concentrations reported for forest streams suggest that P export is minimally affected by P input from surface runoff Amounts of surface runoff in forest watersheds will be low owing to the protection afforded by canopy cover and/or forest floor vegetation The

“background” P export in forest streams is a direct reflection of the chemi- cal and physical factors that affect P concentrations in groundwater and subsurface runoff Because larger amounts of stream flow from forest watersheds will arise from groundwater and subsurface runoff, the “chemi- cal sieving” action of the soil plays a major role in maintaining the con- sistently low dissolved inorganic P concentrations in forest streams and may also account in part for the apparent conservative nature of forest watersheds for P

C RUNOFF FROM AGRICULTURAL WATERSHEDS

The loss of P in streams draining agricultural (in most cases arable) watersheds is far less well defined than that for forest streams This is prob- ably due to the fact that in studies designed to estimate this loss, little differentiation has been made with respect to the forms of runoff Conse- quently, there are major problems in estimating P loss from agricultural watersheds using many of the data presented in the literature Losses of

P from agricultural land have not only been based on analyses of streams

draining a specific watershed (Campbell and Webber, 1969; Taylor et al.,

1971 ), but have also been estimated from data obtained in surface runoff

studies (Timmons et al., 1968) Many previous reviews of this subject

have relied on such data (Taylor, 1967)

Losses of P in streams draining various agricultural watersheds are sum- marized in Table IV The lowest loss of total P is from rangeland in On- tario, Canada (Campbell and Webber, 1969) which had received no P fertilizer in living memory This loss is very similar to losses of total P

from forest watersheds, suggesting a minimal contribution if P from sur- face runoff Similarly, the total P carried in the base flow, primarily at-

tributable to groundwater runoff, of several streams draining arable agricul- tural watersheds in S.W Wisconsin (Minshall et al., 1969) is also little different from total P loads in streams draining forest watersheds Minshall

et al (1969) reported the total P loss in base flow to be less than 0.12

kg/ha per year If stream flow during periods of surface and subsurface runoff is included, however, the estimated annual loss of total P increases

by one order of magnitude, as indicated by the data of Witzel et al ( 1 969)

for the same area of S.W Wisconsin (Table I V )

These studies suggest that the groundwater runoff or base-flow compo- nent of streams draining agricultural watersheds is little different from the total P load of forest streams It is therefore necessary to estimate the ex- tent to which the P load of streams draining agricultural watersheds may

be augmented by P loads of surface and subsurface runoff

The major factors affecting the loads of P in surface runoff from agricul- tural land include time, amount, and intensity of rainfall, rates of infiltra- tion and percolation, slope, soil texture, nature and distribution of native soil P, P fertilization history, cropping practice, crop type, and crop cover density

A selection of reported losses of P in surface runoff from arable land

of various slopes and cropping practices is summarized in Table V Losses range from the extremely high values of 67 kg/ha per year to almost zero Losses of P in all studies listed in Table V have been based on the collec- tion of surface runoff (water and particulates) from small experimental

TABLE I V Losses of P in Streams Draining Agricultural Watersheds

S.W Wisconsin Silt Total P

plots frequently no larger than 30 x 6 m, with subsequent analysis for one

or more forms of P Although this approach was originally developed to investigate soil fertility losses due to soil erosion, it is still used to estimate

P loads in surface runoff as it relates to the fertility of surface waters (Tim-

mons et al., 1968; Nelson and Romkens, 1969)

It is difficult to make any generalizations regarding the P loads carried

in surface runoff or to draw conclusions from them in terms of how agricul- tural practices and natural variables affect P loads in streams draining agri- cultural watersheds This is due to the differences in forms of P measured and the lack of comparative studies with respect to slope, soil texture, cropping, and climatic variables

One of the few studies from which meaningful interpretations of P loss

in surface runoff can be made in relation to degree of slope and cropping

practice is that by Massey et al (1953) in Wisconsin (Table V ) As ex-

pected, greater “available” (water-soluble plus pH 3 extractable) P losses

to surface runoff occurred on the steeper slopes when cropping practice was kept constant The introduction of two years hay into the rotation reduced the P loss by a factor of approximately four The value of “im- proved” or “conservative” agricultural practices in reducing the magnitude

of P losses is illustrated in the studies at Coshocton, Ohio (Weidner et

al., 1969) and at Lafayette, Indiana (Stoltenberg and White, 1953) It

should be noted, however, that although the “improved” practice reduced the total amounts of acid-hydrolyzable P lost in surface runoff at Coshoc- ton, the concentration of this fraction during surface runoff increased from 0.43 to 0.59 pg/ml

Attempts have been made to measure the relative contributions of the

aqueous and particulate fractions of surface runoff to the total loss of a

measured form of P In a plot study using simulated rainfall, Nelson and Romkens ( 1969) obtained dissolved inorganic P concentrations of 0.05,

0.30, and 0.50 pg of P per milliliter in the aqueous phase of surface runoff

from fallow plots 12 days after 0, 56, and 1 1 2 kg of P per hectare, respec-

tively, had been disked into the soil, with only slight decreases in concen-

trations up to 75 days after fertilizer application Although very high arti- ficial rainfall rates were employed (up to 73.5 mm/hr), indications are

that high concentrations of dissolved inorganic P may be maintained in

surface runoff water Timmons et al (1968) determined the distribution

of total P loss in surface runoff between the aqueous and particulate phases from plots under natural precipitation Although these workers did not report P concentrations, losses of total P in the aqueous phase of surface runoff arising from snowmelt far o.utweighed those in the particulate phase

In contrast, total P loss in the aqueous phase varied in most cases between

5 and 40% of the loss in particulates in surface runoff arising from rainfall