Advances in Agronomy, Volume 65

DVANCES IN gronomyy VOLUME 65 {y Advisory Board Martin Alexander Ronald Phillips Cornell University University ...

Author: Donald L. Sparks

31 downloads 1883 Views 18MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

DVANCES IN

gronomyy

VOLUME 65

{y

Advisory Board Martin Alexander

Ronald Phillips

Cornell University

University of Minnesota

Kenneth J. Frey

Larry P. Wilding

Iowa State University

Texas A&M University

Prepared in cooperation with the

American Society of Agronomy Monographs Committee Jon Bartels Jerry M. Bigham Jerry L. Hatfield N. B. Kirkham David M. Kral

William T. Frankenberger, Jr., Chairman Linda S. Lee Martin J. Shipitalo David Miller Diane E. Stott Kenneth J. Moore JeffreyJ. Volenec Donald C. Reicosky Dennis E. Rolston

DVANCES IN

gronomy Edited by

Donald L. Sparks Department of Plant and Soil Sciences University of Delaware Newark, Delaware

ACADEMIC PRESS San Diego London Boston New York Sydney Tokyo Toronto

This book is printed on acid-free paper.

@

Copyright 0 1999 by ACADEMIC PRESS All Rights Reserved. No part of this publication may be reproduced or transmitted 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 the Publisher. The appearance of the code at the bottom of the first page of a chapter in this book indicates the Publisher’s consent that copies of the chapter may be made for personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc. (222 Rosewood Drive, Danvers, Massachusetts 0 1923), for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Copy fees for pre-1999 chapters are as shown on the title pages. If no fee code appears on the title page, the copy fee is the same as for current chapters. 0065-21 13/99 $30.00

Academic Press a division of Harcourt Brace & Company

525 B Street, Suite 1900, San Diego, California 92101-4495, USA http://www.apnet.com

Academic Press 24-28 Oval Road, London NWl 7DX, UK http://www.hbuk.co.uk/ap/ International Standard Book Number: 0- 12-000765-7

PRINTED IN THE UNlTED STATES OF AMERICA 98 99 0 0 0 1 02 0 3 B B 9 8 7 6

5

4

3 2

1

Contents CONTRIBUTORS ........................................... PREFACE .................................................

ix

xi

ADVANCESIN FERTIGATION I. 11. EI.

IV V. VI. VII. VIII.

B. Bar-Yosef Introduction.. ............................................ Overview of Past Trends in Fertigation ........................ Principles of Fertigation .................................... Managing Crop Fertigation ................................. Modeling Fertigation. ...................................... Monitoring ............................................... Safety ................................................... Future Trends and Areas Needing More Research ............... References ...............................................

2 5 19 40 60 62 63 64 65

THEGENETICS, PATHOLOGY, AND MOLECULAR BIOLOGY OF T-CYTOPLASM MALESTERILITY IN MAIZE Roger P. Wise, Charlotte R. Bronson, Patrick S. Schnable, and Harry T. Horner I. Introduction .............................................. II. Cytoplasmic Male Sterility Systems ........................... 111. cms-T Causes Premature Degeneration of the Tapetum. . . . . . . . . . . w. Southern Corn Leaf Blight Epidemic of 1970 . . . . . . . . . . . . . . . . . . . V. Disease Susceptibility and Male Sterility ....................... VI. Nuclear-Cytoplasmic Interactions and Restoration of cms-T . . . . . . VII. Perspectives by cms-T Researchers ........................... VIII. Future Directions ......................................... References ...............................................

80 81 83 86 97 103 116 119 122

APPLICATION OF CAPILLARY ELECTROPHORESIS TO ANION SPECIATION IN SOILWATEREX~RACTS R. Naidu, S. Naidu, P. Jackson, R. G. McLaren, and M. E. Sumner I. Introduction .............................................. V

132

CONTENTS

vi

I1. General Principles ......................................... I11. Sample Introduction ....................................... Iv. Separations ............................................... v Detection ................................................ VI. Comparison with Other Analyucal Techniques . . . . . . . . . . . . . . . . . . VII . Implication for the Analysis of Soil Solutions.................... References ...............................................

134 136 137 143 143 146 146

ADVANCESINSOLUTIONCULTUREMETHODS FOR PLANTMINERAL NUTRITION RESEARCH David R . Parker and Wendell A. Norvell I . Introduction .............................................. I1. Soil Solutions and Nutrient Solutions ......................... I11. Advances in Solution Culture Methods for Controlling Nutrient Status ........................................... n! Summary and Future Outlook ............................... References ...............................................

151 154 160 201 203

RADIATIONUSEEFFICIENCY

Thomas R . Sinclair and Russell C. Muchow I . Introduction .............................................. II. Theoretical Analyses of RUE ................................ I11. Experimental Determination of RUE .......................... n! Experimental Measures of RUE .............................. v Sources of Variability in RUE ................................ VI. Conclusions .............................................. References ...............................................

215 220 226 233 248 258 259

THEEFFECTSOF CULTIVATIONON SOIL NITROGEN MINERALIZATION I. I1. I11.

n? v

Martyn Silgram and Mark A. Shepherd Introduction .............................................. Methods of Mineralization Measurement ...................... Cultivation Effects on Soil Physical Conditions . . . . . . . . . . . . . . . . . . Cultivation Effects on Nitrogen Mineralization . . . . . . . . . . . . . . . . . Conclusions ..............................................

267 270 275 280 297

CONTENTS

VI. Management Implications...................................

vii

References ...............................................

300 303

INDEX...................................................

313

This Page Intentionally Left Blank

Contributors Numbers in parentheses indicate the pages on which the authors’ contributions begin

B. BAR-YOSEF (1)Agi-icultural Research Organization,Institute of Soils and Water, Bet Dagan 50250, Israel CHARLOTTE R. BRONSON (79) Department of Plant Pathology, Iowa State University,Ames, Iowa SO01 1 HARRY T. HORNER (79) Department of Botany, Iowa State University,Ames, Iowa 5001 1 P. JACKSON (13 1) WatersAustralia Pg. Limited, Rydalmere, New South Wales 21 16, Australia’ R. G. MCLAREN (1 3 1) Department of Soil Science, Lincoln University, Canterbury, New Zealand RUSSELL C. MUCHOW (2 15) CSIRO, Tropical Ap’culture, Cunningham Laboratory, Brisbane, Queensland 4067, Australia R. NAIDU (13 1) CSIRO Land and Waterand CooperativeResearch Centrefor Soil and Land Management, Glen Omond, Adelaide, South Awtralia $064,Australia S. NAIDU (13 1) Panorama, Adelaide, South Australia 5064, Australia WENDELL A. NORVELL (15 1) USDA Plant, Soil, and Nutrition Laboratory, Ithaca, New York 14853 DAVID R. PARKER (15 1) Soil and Water Sciences Section, Department of Environmental Sciences, University of California, Riverside, California 92521 PATRICK S. SCHNABLE (79) Departments of Agronomy and Zoology 6 Genetics, Iowa State University,Ames, Iowa 5001l MARK A. SHEPHERD (267) ADAS, Gleadthorpe Research Centre, Mansjield, Nottinghamshire NG20 9Pe United Kingdom MARTYN SILGRAM (267) ADAS, “Woodthorne,” Wolverhampton W V 6 8TQ, United Kingdom THOMAS R. SINCLAIR (2 15) USDA-ARS Agronomy, Physiology 6 Genetics Laboratory, University of Florida, Gainesville, Florida 32611 M. E. SUMNER (13 1) Department of Crop and Plant Sciences, University of Georgia, Athens, Georgia 30602 ROGER P. WISE (79) Corn Insects and Crop Genetics Research Unit, USDAARS, Department of Plant Pathology, Iowa State University,Ames, Iowa 50011 ‘Present address: Dionex Corporation, Sunnyvale, California 94088.

ix

This Page Intentionally Left Blank

Preface Volume 65 includes six contemporary and comprehensive reviews of important agronomic topics. Chapter 1 provides a state-of-the-artreview of advances in fertigation. Topics that are covered include an overview of past trends in fertigation, principles of fertigation, managing crop fertigation, modeling fertigation, monitoring and safety, and future trends and areas needing more research. Chapter 2 is an authoritative review of the genetics, pathology, and molecular biology of T-cytoplasm male sterility in maize. The authors discuss the role that cytoplasmic male sterility systems play in facilitating the production of hybrid seeds, the effects of widespread planting of T-cytoplasm maize on the severe 1970 epidemic, the effect of a mitochondria1 gene on disease susceptibility and male sterility, the involvement of nuclear
xi

This Page Intentionally Left Blank

ADVANCES INFERTIGATION B.Bar-Yosef Agricultural Research Organization Bet Dagan, Israel 50250

I. Introduction A. Fertigation Expansion B. Potential Advantages of Microfertigation 11. Overview of Past Trends in Fertigation A. Crop Responses to Fertigation B. Water and Nutrient Distribution in Soils C. Soil Root Volume Effects D. Reducing Salinity Hazards E. Drip versus Microjet Fertigation E Subsurface Fertigation G. Avoiding Clogging 111. Principles of Fertigation A. Quantity Considerations B. Intensity Considerations C. Implication of the Relationshipsbetween Uptake F l u and Concentration D. Coupling Quantity and Intensity Factors E. Root Growth and Distribution in Soil F. Rhizospheric Processes G. Water Quality Considerations N. Managing Crop Fertigation A. N, P, and K Objective Consumption Functions B. Nutrient Concentrations in Irrigation and Soil Solutions C. Preplanting Broadcast Fertilization and Banding under Fertigation D. Choice of Fertilizers E. Temperature Effects F. Organic Matter G. Greenhouses V. Modeling Fertigation A. Models Simulating Transport and Uptake Processes B. Crop Models C. Auxiliary Models VI. Monitoring VII. Safety Vm. Future Trends and Areas Needing More Research References

1 Advances In Agronomy, Volume 61 Copyright 0 1999 by Academic Press. All riglia of reproduction in any form reserved. 0 0 6 m i 3 / 9 9 $moo

2

B. BAR-YOSEF

I. INTRODUCTION Irrigation and fertilization are the most important management factors through which farmers control plant development, fruit yield, and quality. The introduction of simultaneous microirrigation and fertilization (fertigation) (Goldberg er al., 1976; Dasberg and Bresler, 1985) opened up new possibilities for controlling water and nutrient supplies to crops and maintaining the desired concentration and distribution of ions and water in the soil. Even though combined irrigation and fertilization can be practices under flood, furrow, and sprinkler irrigation, the control under microirrigation is superior; therefore, microfertigation is discussed in more detail than other fertigation methods. Water transport in soil under various irrigation methods has been intensively studied and reviewed in the literature (e.g., for principles, Hanks and Ashcroft, 1980; Hillel, 1980: for drip irrigation, Bresler, 1977). Because of the above and other reviews, aspects related to irrigation, water flow in soil, and water uptake by plants are not discussed in the present review; attention is focused on fertigation principles and rhizospheric processes pertinent to fertigation, which have been but meagerly reported in the literature. The objectives of this work are: (i) To review past and current studies of cultural and theoretical aspects of fertigation that have contributed to the state of knowledge in this area; (ii) to discuss principles of fertigation with respect to crop production and environment sustainability; and (iii) to evaluate future trends in fertigation research and application.

A. FERTIGATION EXPANSION Agricultural areas under microirrigation are expanding almost linearily with time worldwide, including in the United States. In Mediterranean countries the expansion rate has decreased during the past decade, indicating that the use of microirrigation is approaching saturation. In the Far East (China, India, and Japan) and in Australia the expansion rate is still increasing (Fig. 1). In some countries, such as Israel, where water availability limits crop production, microirrigation supplies about 75% of the total irrigated area. Throughout the world, the percentage of microirrigation is considerably smaller (<1% in 1991), and in the United States it accounts for -3% of the total irrigated area (Bucks, 1995). In Israel, 75% of the microirrigated area was under fertigation in 1995 (Sneh, 1995) and this percentage probably applies to other advanced countries as well. In developing countries, fertigation involves a negligible fraction of the microirrigated area, and fertilizers are still being applied by broadcast dressing and banding (Bachchhav, 1995). In 1991 the distribution of microirrigation area by crop was orchards and vines,

3

ADVANCES IN FERTIGATION 2000 m

h

f

0

1500

v)

Q C

m

y

1000

0

s

Y

%

500

0 1970

1975

1985

1980

1990

1995

YEAR

(m),

Figure 1 Agricultural area under microimgation between 1973 and 1991 in the United States Australia Mediterranean countries (0).Far East (China, India, and Japan A),rest of the world (A),and total area worldwide (V).Data derived from Gustafson e? al. (1974) and Bucks (1995).

(v),

72%; vegetables, 16%; and field crops, 14% (Table I). The main reason for the relatively restricted use of microirrigation in crops such as corn and cotton is the high cost of microirrigation systems; the operating costs of trickle and furrow systems in cotton in the United States were within 5% of each other, but the annual fixed costs were $865 and $370 per hectare, respectively (Nakayama and Bucks, 1986). In developing countries, this margin is even wider (Saksema, 1995).The potential advantages of microfertigation over surface irrigation and broadcast fertilization (to be discussed later) are expected to be manifested in improved field crop yield and quality and to shift the benefit-to-cost ratio in favor of microfertigation. This trend was indeed confirmed in an economic evaluation of various irrigation methods in cotton in California (Boyle Engineering Corporation, 1994), and it is expected, therefore, that the area of microfertigated field crops will increase in the near future, with respect to the data in Table I.

B. POTENTIAL ADVANTAGES OF MICROFERTIGATION The main advantages of microfertigation over microirrigation combined with broadcast/banding fertilization can be summarized as follows. (i) Reduced time fluctuation in nutrient concentrations in soil in the course of the growing season, because of the flexibility in delivery of nutrients and water. According to the the-

4

B. BAR-YOSEF Table I MicroirrigationUse by Crops (1991)"

Tree Crops Citrus Deciduous Avocado + mango Other Vines Vegetables Field Greenhouse Field crops Cotton sugar Other Flowers Nursery Greenhouse

World (ha)

United States (ha)

230,000 295,550 51,275 171,000 234,660

95,000 131,000 8,000 99,000 94,000

157,000 65,000

58,500 8,500

36,630 45,100 42,330

15,000 42,000 20,000

9,275 17,685

6,500 4,500

"After Bucks (1995).

ory of Zaslavsky and Mokady (1 967), when curves of yield response to nutrients and water are convex and monotonic, time fluctuations of nutrient and water contents in soil cause reduction in yield (Bresler, 1977). The attenuated fluctuations under microfertigation therefore ensure higher and more consistent yields relative to broadcast fertilization. (ii) Easy adaptation of the amounts and concentrations of specific nutrients to crop requirements, according to the stage of development and climatic conditions. (iii) Convenient use of compound, ready-mix, and balanced liquid fertilizers with minute concentrations of minor elements that are otherwise very difficult to apply accurately to the field. (iv) The crop foliage remains dry, thus retarding development of plant pathogens (Yarwood, 1978) and avoiding leaf burn (Bernstein and Francois, 1975; Mass, 1986). (v) Precise application of nutrients according to crop demand, thus avoiding excess fertilizer concentrations in the soil and leaching out of the wetted soil root volume. (vi) Application of water and fertilizers to only a part of the soil volume; the addition of nutrients only to wetted area, where active roots are concentrated, enhances fertilizer use efficiency and reduces leaching of nutrients to deep underground water by seasonal rains. Drip fertigation has additional advantages. (vii) It is unaffected by wind and causes less runoff than overhead irrigation (see below). (viii) It reduces heavy tractor traffic in the field, associated with the broadcasting of fertilizers, and allows

ADVANCES IN FERTIGATION

5

easy application of nutrients via the water when top-dressing is expensive because of plant height. The above advantages of microfertigation are offset by expensive investment in fertilizer injectors, safety devices, and shipping and storage of large volumes of dilute liquid fertilizers. Steffen et al. (1995) evaluated the overall costs and benefits of trickle-fertigated, and sprinkler-irrigated, broadcast-fertilized, open-field tomatoes; the costs were $14,000 and $8500 per hectare, respectively. The difference in costs was not balanced by the higher gross income from the trickle-fertigated plots. However, this comparison was carried out under conditions of sporadic heavy rains, which adversely affected yields in the drip-fertigated treatment more than in the sprinkler plus broadcast fertilization treatment.

II. OVERVIEW OF PAST TRENDS IN FERTIGATION A. CROPRESPONSES TO FERTIGATION Numerous studies have been published during the past three decades on crop responses to fertigation and on comparisons between microfertigation, on the one hand, and microimgation coupled with broadcast and/or localized fertilization, on the other hand. A summary of a literature search on these subjects is presented in Table 11, which shows that the responses to fertigation of the majority of irrigated crops have been investigated in several locations. Orange and grapefruit trees in Israel and Florida were found to have higher fruit yield and N fertilizer use efficiency under microfertigation than under split N broadcast application (Dasberg et al., 1988; Boman, 1996). A h a and Mozaffari (1995) claimed that the main advantage of N microfertigation over broadcast-N fertilization in oranges was the reduced nitrate leaching below the soil root volume. Reduced NO, leaching under microfertigation, without a decline in fruit yield or quality has been also reported for annual crops (e.g., for tomato, Miller et al., 1976; for celery, Feigin et al., 1982). Several studies have shown that crop yield response to N was stronger under microfertigation than under microirrigation plus broadcast N application (e.g., in apples, Assaf et al., 1983; in tomatoes, Bar-Yosef, 1977; Clough et al., 1990; in lettuce, Bar-Yosef and Sagiv, 1982b). The data in Table I1 include much more information on specific crop responses to N fertigation, and on comparisons between different fertilization techniques; these specific subjects are not discussed further in the present review. Studies on P and K microfertigation (Table 11) showed remarkably improved crop responses to these elements, relative to broadcast fertilization. Rauschkolb et al. (1979) showed that P drip fertigation resulted in higher P content in tomato plants than P-banding with an identical P rate. Bar-Yosef et al. (1989) found that

6

B. BAR-YOSEF Table I1 Summary of Literature Search on Crop Response to Microfertigation

Crop

n e e crops Orange

Grapefruit

Apple

Peach Pecan Almond Prune Grapevine Grapevine Banana

Raspberry Papaya

System Studied

Location

Reference

N microfert.

Israel

Micro irr., fert. Drip fert. Drip vs microjet N fert. K, N microfert. N microfert. vs N broadc. Drip, flood irr., N fert. K, N microfert vs N, K broadc. N microfert. NO, drip fert. NO,,NH, drip fert. Drip fert. Drip vs microjet fert. Fert-soil acidif. Microfert. Microfert. N,P,K microfert. Microfert.

Spain Florida Israel

Dasberg et al. (1983, 1988) Dasberg (1987) Legaz et al. (1983) Koo and Smajstrla (1984) Dasberg ( 1995)

Israel Florida

Lavon et al. (1995) Alva and Mozaffari ( 1995)

Texas

Swietlik (1992)

Florida

Boman ( 1996)

Israel Israel Israel Holland France Canada New York Denmark Canada France

N drip fert. vs broadc. N N microjet fert. K drip fert. K drip fert. N,P,K drip fert. Drip fert. N,K drip fert., basin in: Drip irr.-fert. N drip fert.

Georgia

Assaf et al. (1983) Bar-Yosef et a!. (1988a) Klein and Spieler (1987) Delvert and Bolding (1988) Cassagnes (1988) Nielsen et al. (1994) Robinson and Stiles (1993) Dencker and Hansen (1994) Nielsen et al. ( 1995) Guennelon et al. (1981) Bussi et al. (1991. 1994) Worley et al. (1995)

California California California Israel Israel India

Kjelgren el al. (1985) Uriu et al. (1977, 1980) Christensen etal. (1991) Bravado and Hepner (1987) Lahav and Kalmar (1988) Hedge and Srinivas (199 1)

Denmark Hawaii

Callesen (1991) Awada et al. (1979)

Israel Israel

Rudich et al. (1982) Bar-Yosef (1977) Bar-Yosef et al. (1980b) Bar-Yosef and Sagiv (1982a) Kafkafi and Bar-Yosef (1980)

Field vegetables, flowers Tomato N drip fert. N drip fert., sand dunes N drip fert., calcareous soil

Israel

conrinues

ADVANCES INFERTIGATION

7

Table II-continued Crop

Potato

Sweet pepper

Lettuce

Celery Broccoli Squash Cauliflower Strawberry

Muskmelon Watermelon Chrysanthemum

System Studied

Location

N,K drip furrow fert N drip fert. P drip fert. vs banding N, P subsurf. drip fert. N,P drip fert., calcereous soil N,K fertil., irr. management N microfert. N,K drip fert. N drip fert. Fert. in drip, furrow irr. P drip fert. N drip fert. N fert. N drip fert. and fruit quality N drip fert. vs broad N,P,K drip fert. N,K drip fen. Drip fert. N,P,K microfert. N fert. vs broadc. N drip fert. N fert. N drip fert. Subsurf. N irr. interaction Irr. method, fert. Drip irr-K fertil. Drip fert., sched. N drip furrow fert. K dnp fert.

Florida

N drip fert. N irr. interaction in drip in: Microin.. fert.

Nebraska Virginia

Reference

Brazil California California

Canijoeral. (1983) Miller et al. (1976) Rauschkold et al. (1979)

California

Phene et al. (1982, 1986, 1990a)

California

Mikkelsen and Jarrel(l987)

South Carolina Karlen e f al. (1985)

Hawaii Florida Cyprus India Cyprus Puerto Rico Puerto Rico Puerto Rico

Coltman (1987) Clough et al. (1990) Papadopoulos ( 1988) Keshavaiah and Kumaraswamy (1993) Papadopoulos (1992) Crespo-Ruiz et al. (1988) Goyal et al. (1988) Goyal et al. (1989)

New Zealand

Haynes (1988)

Israel Georgia Florida Israel Holland Australia Holland California Arizona

Bar-Yosef (1 980c) Jaworski et al. (1978) Neary et al. (1995) Bar Yosef and Sagiv (1982b) Bakker et al. (1984) McPharlin et al. (1995) Slangen et aZ. (1988) Feigen et al. (1982) Doerge and Thompson (1995)

Florida Florida Florida Spain

Clough et al. (1992) Wall et al. (1989) Locasio et al. (1977) Pomares et al. (1992)

Albregts et al. (1996) Hochmuth er al. ( 1996) South Carolina Bhella and Wilcox (1985) Pier and Doerge (1995) Harbaugh and Wilfret (1980) Harbaugh et al. (1989) continues

B. BAR-YOSEF Table II-continued Crop

System Studied

Field crops Wheat Sugarcane Corn Sweet corn Peanut

Greenhouse Tomato

Lettuce Muskmelon Rose

Location

Reference

N fertigation N,K drip fert. Drip fert. Drip fert. P surf., subsurf. drip fert. N drip fert. Drip fert., saline water, sand dunes

Sweden Wisconsin Israel France

Flink et al. (1995) Ingram and Hilton (1986) Yanuka et al. (1982) Giradin et al. (1993)

Israel Israel

Bar-Yosef et al. (1989) Wallerstein et al. (1982)

Israel

Silberbush et al. (1985)

Drip irr-fert. relationship N drip fert. Drip irr.-fert. relationship N drip fert. P drip fert. N,P,K fert. P drip fert. K drip fert. Microfert.

Ohio

Bauerle (1975)

Cyprus Israel

Papadopoulos (1 987) Bar-Yosef (1988)

Australia Israel Netherlands Israel Israel France

Gastaldi and Sutton (1989) Bar-Yosef and h a s (1995) Sonneveld (1995) h a s and Bar-Yosef (1997) Bar-Yosef (1996b) Brun et al. (1993)

P drip-fertigated sweet corn gave a significantly higher yield than drip-irrigated sweet corn that also received preplant P fertilization. Enhanced response to microfertigated K has also been reported for citrus (Lavon et d., 1995), grapefruit (Boman, 1996), grapevine (Christensen et aZ.,1991), and sugarcane (Ingram and Hilton, 1986). In general, the response to P and K fertigation became more pronounced as N fertigation improved and became a nonlimiting growth factor.

B. WATERAND NUTRIENT DISTRIBUTION IN SOILS The governing equations that describe the flow of water and nonreacting solutes under point- and line-source irrigation were presented and discussed by Bresler (1977, 1978) and Dasberg and Bresler (1985). Bresler (1977) showed some applications of the analytical solution of the steady infiltration equation in planning emitter spacings and discharge rates as functions of soil hydraulic properties and desired matric water potential between emitters. Later studies (Warrick, 1986; Timlin et al., 1996) contributed additional tools for solving the two-dimensional convection-dispersion equation with emphasis on various agronomic applications.

ADVANCES IN FERTIGATION

9

Simulated and empirical results of water content distribution in soils under trickle irrigation emphasize two practical characteristics of microirrigation: (i) When a dry soil is irrigated, the localized wetted soil volume has a sharp wetting front that sharply separates the wet and dry soil domains. (ii) A major fraction of the wetted soil volume has a relatively uniform water content. These attributes (see for example Fig. 7 in Bresler, 1977)justify two important assumptions made in fertigation management: (i) The wetting front position defines the boundary of the plant’s soil root volume. (ii) The mean water content (6) and nutrient concentrations (C) in the soil root volume are reasonable approximations to the actual distributions of 9 and C in the root zone. If we accept these assumptions, the radius R (cm) of the wetting front can be evaluated from soil hydraulic properties, the emitter’s discharge rate, q (ml/h), and the duration of irrigation, t(h). A simple estimate under conditions of no evaporation and no water extraction is given by Eq. (1) (Ben Asher et al., 1986),

where soil properties are represented by 9 at field capacity (eft). Other semiempirical expressions for R(t) were reviewed by Zazueta et al. (1995). The governing equation describing two-dimensional (x, z ) transport of reactive ions in soil [Eq. ( 2 ) ]is presented in order to analyze physical and chemical factors that determine the movement of P, microelements, and exchangeable cations in the soil:

The equation accounts for convective and diffusive ion flows (q = flux of soil solution, Dh = sum of diffusion and mechanical dispersion coefficients) and the buffering capacity of the soil (b), which equals the slope of the ion adsorption isotherm at a given concentration in the soil solution (c).For axisymmetric cylindrical flow, which is used to describe transport from a single trickle emitter in the soil (Bresler, 1977), Eq. (2) is transformed into (3), r being the radial coordinate:

A solution of (3) for an inert salt ( b = 0) in loamy and silty soils is presented in Fig. 2. The irrigation water had a low solute concentration (C,) that miscibly dis-

10

B. BAR-YOSEF

Figure 2 Model computations [Eq.(3)] of two-dimensionalaxisymmetrical nonreactive salt concentration distribution during infiltration from a trickle source for two trickler discharges (Q,, Q,),and two soils. The cumulativeinfiltrationis 12 liters. The numbers labeling each curve indicate relative concentration expressed as (C - Co)/Cn.The numbers in parentheses are salt concentrations (c) in soil solution (mmol(+)/liter); Cois the inlet salt concentration, and cn is the initial c in the soil (mmol(+)/ liter). Heavy lines represent the wetting fronts. Reproduced from Bresler (1977) with permission from Academic Press.

placed a highly concentrated solution initially found in the soil (c,,). The emitter’s discharge rate was either 4 or 20 liters/h. In both soils, and for both trickle discharge rates, the soil solution concentration, c, of the saturated water entry zone was calculated to be identical to the concentration in the infiltrated water. The leached part of the soil in the vertical component of the wetted zone was deeper,

ADVANCES IN FERTIGATION

11

and in the radial component it was narrower, as the soil became coarser and the trickler discharge rate became smaller. As expected, the concentration increased toward the wetting front edges. The results presented can be used to evaluate the distribution in the investigated soils of nitrate added via the water at a concentration Co.In this case the zone (c-Co)/cn = 0 in Fig. 2 is the part of the wetted soil volume in which the concentration of the nitrate in the soil solution equals its concentration in the irrigation water. The distribution of C1- (b = 0) in a loamy sand soil under furrow irrigation [electrical conductivity (EC) = 3 dS/m] during infiltration and redistribution was simulated by Noborio et al. (1996) by solving Eq. (2). Characteristic profiles of minimum EC in the ridge and furrow slope regions, and the finding of an EC similar to the EC in the irrigation water at the furrow bottom, agreed quite well with experimental field data. Equation (2) predicts that strongly adsorbed nutrients (b)c)>> 1) have reduced mobility in soil, compared with unadsorbed ions. For a given c, b(c) of a clayey soil exceeds that of a sandy soil, therefore the mobility of adsorbed ions in finetextured soils is less than that in coarse-textured soils. For example, P concentration distributions in sandy and clayey soils determined experimentally after given amounts of P and water were applied via a point source are presented in Fig. 3 . Whereas NO, movement in both soils extended to a distance of 20 cm from the source (data not presented), P movement was restricted to distances of 12 and 7 cm from the emitter, in the sandy and clayey soils, respectively. Larger application rates of water and P increased the distance of the water from the emitter, but not that of P (data not presented). The values of b(c)at c = 2 mg P/liter were 35 and 1.O (mg P/kg soil) /(mg P/liter) in the clayey and sandy soils, respectively (BarYosef and Sheikholslami, 1976). Similar effects of soil type and P application rate on the P distribution in soil under drip fertigation can be found in studies by Rolston et al. (1979), Bacon and Davey (1982, 1989), and Keng et al. (1979). Rolston et al. (1979) showed that o-phosphate applied via trickle irrigation moved to a much greater distance into a clayey soil than had previously been observed for comparable application rates spread uniformly on the soil surface and irrigated by sprinkler irrigation. The reason for the difference is that in point source fertigation all P was applied over the small surface area of the solution entry zone, so that soil adsorption sites became saturated, b(c)decreased significantly, and the extent of P migration was greater than with broadcast P application,for which c was smaller and the high b(c)restricted P transport in soil. The difference in transport of nitrates between point source fertigation and broadcast application plus sprinkler irrigation is expected to be considerably smaller than that observed in P transport. The mechanism controlling K transport in soil is based on its rapid exchange with other cations in the soil (see Section IIIF). When the quantity of K in the soil is small relative to the soil cation-exchange capacity, adsorption is controlled mainly by variations in K concentration in the soil solution (c,). As c, increases

12

B. BAR-YOSEF

a 0 ,

-1 ..a J 0

'

'

'

2

4

6

'

'

'

'

'

'

8 I 0 12 14 16 18 20

DISTANCE (cm)

b

I:

,

,

2

4

,

,

6

8

,

,

,

,

-16 0

10 12 14 16 18

DISTANCE (CM) Figure 3 Concentration of NaHCO, extractable o-phosphate in soil as a function of depth and lateral distance from a 0.25 literlh emitter placed on top of the soil. (a) Sandy soil (field capacity 0.05, wlw), 21 h after terminating fertigation with 1 liter solution of 200 mg Plliter. (b) Clayey soil (field capacity 0.25, w / ~ )21 , h after terminating fertigation with 2.25 liters of solution a. Initial o-phosphate concentrations in soils a and b were 0 and 2 mg P/kg, respectively. Adapted from Bar-Yosef and Sheikholslami (1976).

around a point source, the K buffering capacity decreases and a deeper K movement is expected relative to sprinkler irrigation and broadcast K application (Uriu et al., 1977). Bar-Yosef and Sagiv (1985) showed that at the time of maximum K

ADVANCES IN FERTIGATION

13

uptake rate by crops with high demand for K, this element must be supplied through the water even if its concentration in the soil is sufficient, because the rate of sorbed K release to the soil solution becomes a rate-limiting step in K uptake. This constraint is particularly important under drip fertigation, where plant root volumes are restricted. The limited migration distance of strongly adsorbed ions in soil, with respect to the radius of the wetting front, implies that in many soils the distance between emitters strongly affects nutrient availability to plants. To reduce the impact of restricted mobility in soil, a combination of preplant broadcast fertilization and fertigation during the season must be practiced. The rate of preplant top-dressings should be based on routine soil test results multiplied by a factor (<1) to account for the extra supply via the irrigation water.

C. SOILROOTVOLUME EFFECTS Partial soil volume fertigation has been mentioned above as one of the advantages of microfertigation. For a given soil and emitter discharge rate, increasing the fertigation frequency reduces the time fluctuation in nutrient concentrations in soil solution, but decreases the soil root volume (Bar-Yosef et al., 1980~). The size of the soil root volume also depends on the initial soil water content. Phene et al. (199 1) reported that high-frequency-fertigated sweet corn had a root system that extended to a depth of 180 cm, whereas Martinez Hernandez et al. (1991) found under similar growth conditions that sweet corn roots were confined to the upper 50-cm soil layer. The main difference between the two studies was in the initial soil moisture: In the first study the 2-m soil profile was at field capacity, whereas in the second it was at air dryness. Plants with restricted root systems develop smaller canopies than plants with unrestricted root development (Bar-Yosef et al., 1988a; Plaut et al., 1988; Erez et al., 1992).Two mechanisms can explain this phenomenon: (i) reduced supply of root-synthesized plant growth regulators to the canopy (Itai and Birnbaum, 1996); (ii) decreased water and nutrient uptake rates limiting dry matter production by the plant. Smaller plants under root confinement conditions can be grown at a higher density in the field, thus maintaining the yield per unit land area (Plaut et al., 1988), and smaller plants, particularly trees, can be harvested more easily and sprayed more efficiently. Confined root systems also improve the carbon economy of plants, as the mass of thick roots leading to deep, active roots is reduced and more carbohydrates can be allocated for fruit production. Another feature of confined root systems is the quick and effective control of nutrient and salt contents in the root zone. For example, nitrate uptake by leafy crops and potatoes must be stopped 2-3 weeks before harvest to avoid the accumulation of toxic nitrate levels in edible plant organs. Leaching the nitrates from a small soil root volume is simple, and does not require much water, whereas in a

14

B. BAR-YOSEF

large soil root volume leaching is ineffective and the roots continue to absorb nitrate at a considerate rate, despite the reduced concentration in the soil. A disadvantage of confined soil root volumes is the limited ability to utilize mineralized soil organic N: The unexploited nitrates outside the root zone are prone to leaching by rains before the following crop can recover them from the soil.

D. REDUCING SALNTY HAZARDS Several studies (Bemstein and Francois, 1975; Singh et al., 1978; Meiri and Plaut, 1985; Mizrahi etal., 1988) indicate that irrigation water with total salt concentration of approximately 2 g/liter can be utilized in drip irrigation without significant yield loss relative to freshwater, whereas the use of such water for furrow and sprinkler irrigation caused decreases of 54 and 94%, respectively, in yield (Bemstein and Francois, 1975). The reduced salinity hazard under trickle irrigation can be related to the efficient displacementof salts to the periphery of the wetted soil volume (Fig. 2 ) and to the reduction in salt concentration because the higher irrigation frequency maintains a high soil water content. Salts accumulated in the margin of the top soil layer of the onion-shaped wetted volume (Fig. 2 ) may be leached into the main soil root volume by rains and cause a sudden osmotic shock to drip-irrigated plants. Irrigation during rainfall may avert such a stress. Another factor that may ameliorate salinity stress under trickle fertigation is the high and steady nutrient concentrations in the soil root volume. High nitrate concentration in solution may depress C1 influx by plant roots (Fried and Broeshart, 1967), although this mechanism was found insufficient to alleviate NaCl stress in tomato and melon plants (Feigin, 1990). Ravikovitch and Yoles (1971) and Chauhan et al. (1991) found a positive effect of P nutrition on the yield of clover and millet and of wheat, respectively, under saline conditions, but in other studies on the P-salinity relationship contradictory results were obtained (Feigin, 1990). High concentrations of Na+ inhibit K+ influx into roots, or increase K+ leakage from the cells. High Ca2+ concentrations reduce the influx of Na+ and thus maintain cell membrane integrity and alleviate salinity-induced K deficiency (BenHayyim et d., 1987). Alva and Syvertsen (1991) reported that citrus leaf K and Mg concentration were reduced by high salinity, but that an increase in leaf Ca concentration may minimize the effects of salt stress. Salinity reduces plant root length and surface area (see Section IIIE). Under such conditions high and steady nutrient concentrations in the fertigated soil volume may partially compensate for the expected decline in the rate of uptake by the plant (Kafkafi, 1984). Recently, a simple field technique was suggested for studying salinity-fertilization interactions under trickle irrigation (Demalach et al., 1996), by employing separate lines for saline water and nutrient solution. This technique is simple to operate and may help to elucidate salinity-nutrition relationships under field conditions.

ADVANCES IN FERTIGATION

15

The effective means of controlling solute concentrations in restricted soil root volumes allow growers to apply osmotic stress in microfertigated crops as a tool to improve fruit quality with minimal loss in yield. Fruit quality improvement under controlled NaCl stress was demonstrated in greenhouse tomatoes (Mizrahi et al., 1988; Plaut and Meiri, 1988). Data from Plaut and Meiri (1988) indicate that osmoticum alone cannot explain the fruit quality enhancement obtained, since NaCl was more effective than KCl in improving fruit taste and firmness.

E. DRIPVERSUS MICROJETFERTIGATION Direct comparisons between crop performance under trickle versus microjet fertigation have been meagerly reported in the literature. Differing crop responses are expected due to: (i) differing water and nutrient distribution in axisymmetrical and one-dimensional flow regimes; (ii) better uniformity of water and nutrient application in trickler than in microjet irrigation; and (iii) partial canopy wetting by microjets, resulting in damage to leaves. Bravdo et al. (1992) and Dasberg (1995) compared the responses of orange trees to trickle and microjet fertigation, for similar water and N application rates, fertigation frequency, and wetted soil volume. Neither yield nor fruit quality were significantly different in the two systems. Additional, yet inconclusive, information on fertigation by the two methods can be found in Cassagnes et al. (1984), Cassagnes (1988), and Orphanos and Eliades (1992).

F. SUBSURFACE FERTIGATION The high labor and energy requirements for spreading and collecting laterals every season, and the deterioration of drip lines exposed to solar radiation, animals, and heavy machinery jeopardizes further expansion of trickle imgation. Subsurface location of tricklers and laterals at the appropriate soil depth solves these problems. Moreover, recent advances in quality of fertilizer products, filtration devices, and quantification of chemical precipitation in fertigation systems have significantly reduced emitter clogging problems, which had been the primary reason for the slow expansion of subsurface drip fertigation in the past two decades. In addition to the aforementioned economic advantages, subsurface drip fertigation has some agronomic advantages over surface drip fertigation: i. Nutrients are supplied to the center of the root system, where the water content is relatively high and steady with time (Phene and Howell, 1984) and root activity is maximal. Nutrients introduced via subsurface tricklers can move in a spherical volume around the emitter, while transport in surface application is bound within a hemisphere below the point source (Phene et al., 1986). It is ex-

16

B. BAR-YOSEF Table HI Marketable Test Crop Yields in California and Israel in Response to Surface (SR) and Subsurface (SSR) Drip P Fertigation Israel

c,

mMP

SR

SSR

PO PI

0 0.16 0.40 0.64 0.77 1.29

22.1 22.4

22.9 24.3

23.1

24.9

26.3 23.5

28.9 25.2

P2 P3 Av

PO PI P2 P3 Av

PO P1 P2 P3 Av

0 0.43 0.65 0.77 1.30 2.00

0 0.40 0.64 0.80 1.30 1.90

California Av

SR

Sweet corn (tonha) 22.5 31.3 23.3 29.6 24.0 29.4 27.6 24.3** 30.1

***

SSR

Av

27.0

29.1

29.2

29.4

30.2

29.8

28.8

29.4 ns ns

159.3

Processing tomatoes (red fruit todha) 141.9 142.8 143.3 142.5 155.2 182.2 177.9 168.6

152.8 151.5 151.7

159.0 166.3 161.3

143.7

55.0

155.9 158.7 156.5*

**

Potatoes (tonha) 66.0 60.5

72.0

87.0

79.5

68.0 69.0 66.0

71.0 64.0 72.0

69.5 66.5 69.0**

**

149.2

162.3

Cotton lint (toniha) I .27 1.57 1.25 1.38

142.8 168.7

155.8*

*

1.42 1.31

1.22

I .39

1.30

1.25

1.45

1.35* ns

Note. Varieties: sweet corn, tomatoes, cotton, and potatoes were Jubilee, M82-1-8 in Israel and UC82B in California, GC-510, and Deseare. *, **, ***, ns: significant differences at P = 0.05, P = 0.01, and lack of significance, respectively. Source: Bar-Yosef et al. (1991).

pected that differences in uptake rates of mobile elements between surface and subsurface fertigation systems will be smaller than in those of immobile elements, as the former are distributed and exploited from a considerably larger soil volume than the latter. Data on the response of field crop yields to surface and subsurface P fertigation (Table 111) show that when N and K alone were added via the water and a sufficient concentration of P (20-25 mg NaHCO, extractable P/kg soil) was

ADVANCES IN FERTIGATION

17

found in the soil at seeding time, no differences in sweet corn and tomato yields between surface and subsurface fertigation were observed. In cotton and in potatoes the yields were higher under subsurface fertigation, even when P was not added via the water. When the P concentration in the water (C), varied between 0.2 and 1 mM, the fresh fruit yield was higher in all crops (excluding sweet corn in California) under subsurface than under surface P fertigation (Table 111). The Cp response curve was bell shaped, but the optimum Cp was higher under subsurface than under surface P fertigation (Table 111). The deviation of sweet corn in California from the general behavior stemmed from a high initial cp in the soil. Empirical results of P and N distribution in soil under surface and subsurface fertigation (Bar-Yosef et al., 1989; Martinez Hernandez et al., 1991) confirm the expected deeper P distribution under subsurface than under surface P fertigation and the smaller differences in N and water distribution between the two systems. ii. When emitters are placed below the soil surface in accordance with soil hydraulic properties, the top 3-5-cm soil layer can be kept dry during the whole season. This placement depth (d,)should be overridden in accordance with the plowing depth ($) if dp > d,. The dry top soil prevents weed germination (Bar-Yosef et al., 1989), thereby reducing herbicide use, and decreases water evaporation. Bar-Yosef et al. (1991) found delayed growth of tomato plants in a loamy soil when emitters were placed 40-50 cm below the soil surface as compared with 20-25 cm. The inhibited growth was attributed to the extra time the roots needed to pass through the relatively dry top soil layer and reach the center of the wetted soil volume near the emitter. Subsurface P fertigation may solve one of the major problems associated with no-tillage, namely, the frequent phosphorus deficiency, which stems from the limited downward movement of P applied as soil top-dressing and lack of soil mixing in the 0-40-cm upper soil layer, usually obtained by deep plowing. Unfortunately, this aspect of subsurface drip fertigation has not been investigated yet. iii. Roots grow deeper into the soil under subsurface than under surface drip fertigation (Martinez Hernandez et al., 1991). This stems from deeper water and nutrient distributions in the soil, and the response of root growth to variations in soil water content and nutrient concentrations (Bar-Yosef and Lambert, 1981; Hoogenboom and Huck, 1986; Timlin et al., 1996). Deeper root systems buffer roots against exposure to low and high soil temperatures. Low root temperatures at times of high radiation and high air temperature may cause plant collapse, as root water conductance is reduced (Dalton and Gardner, 1978) and water uptake cannot meet the prevailing potential transpiration demand. Kafkafi (1984) described such a collapse in surface drip-fertigated melons during mornings that followed cold nights; when the emitters were buried 40-50 cm below the soil surface, the melon plants survived. Adverse effects of high soil temperature on roots are discussed in Section IIIE. iv. Subsurface placement of tricklers may prevent soil crusting in sodic soils or

-

18

B. BAR-YOSEF

when using high sodium adsorption ratio irrigation water (see Section IIIG). This reduces surface runoff and improves water and nutrient distribution uniformity in the field. v. Secondary municipal effluents can be used to irrigate edible crops, provided that in no circumstances is a contact between aerial plant parts and irrigation water established. The only irrigation technique that meets this condition is subsurface trickle fertigation. The use of recycled solutions for fertigation is discussed in Section IIIG. From a wastewater management standpoint, buried emitters should be placed as close as possible to the soil surface to promote microbial decomposition of organic contaminants and pathogens and to minimize their leaching toward groundwater. There are some agrotechnical problems associated with subsurface fertigation: (i) Germination usually requires auxiliary overhead irrigation to maintain top soil moisture. (ii) Soils with low and spatially variable low hydraulic conductivity may reduce the effective discharge rate of subsurface emitters and cause nonuniform water distribution in the field. Although this problem has not yet been verified under field conditions, matching between the nominal emitter discharge rate and the effective soil hydraulic conductivity during irrigation is required. (iii) Tree main roots may compress subsurface laterals and reduce emitter discharge rates. The problem seems to be aggravated when laterals are placed in the soil shortly after tree planting, when main roots are growing.

G. AVOIDINGCLOGGING Plugging of emitters causes nonuniform distribution of water and nutrient in the soil. According to Bucks et al. (1982) the coefficients of variation (CVs) of eight different type of emitters increased after 4 years of operation with Colorado River water, from approximately 0.06 t 0.04 to approximately 0.50 ? 0.10, the main reasons being clogging and aging. When the tricklers were acid-treated the CV increased after 4 years to between 0 and 0.16, the larger value characterizing longpath, spiral-grooved tricklers. The fact that acid significantly reduced clogging indicates that the main cause of plugging is chemical precipitation, probably of Ca with HPO, (Imas et al., 1996) or CaCO,. The recommended acid treatment to dissolve Ca-P and carbonate precipitates comprises a 10 to 15-min flush with 0.6% HCl33%, followed by a 1-h wash with water. Other common causes of dripper clogging are suspended mineral and organic particles in fertigation solutions, biofilms formed by microflora inside tricklers and lateral tubes, and penetration of roots into drippers. Suspended particles are treated effectively by three types of filters: (i) layered gravel, which removes organic particles such as algae or solid residues; (ii) centrifugal filters, which remove mineral particles such as silt and sand exceeding a concentration of -3 mg/liter; and

ADVANCES IN FERTIGATION

19

(iii) screen filters, designed to remove particles found in commercial fertilizers and in corroded metal pipes. All filters cause significant head losses (2 to 10 m, depending on filter type and flow rate), which should be taken into account in planning fertigation systems. Automatic flushing devices maintain the head losses at relatively constant values. More details on dripper clogging can be found in Adin and Sacks (1991) and Ravina et al. (1992). Avoiding biofilm build-up inside tricklers and tubes requires chlorination every other week when colonial protozoa are the main reason for the clogging, and every 3 days when sulfur bacteria form the biofilm (Sagi et al., 1995). To minimize microorganism development in closed irrigation systems it is advisable to prevent fertigation solutions from remaining in tricklers and laterals between irrigations. This can be achieved by installation of end drains, running of water without fertilizers for the last 5-10 min of each fertigation, and routine flushing of mains and submains equipped with flushing valves. Clogging of tricklers by roots seems to vary with irrigation and cultivation techniques. Bar-Yosef et al. (1991) reported that after 4 years of sweet corn, tomato, and potato growth under surface and subsurface drip fertigation, the incidence of clogged drippers over an area of 1 ha was <2%. In lawns, golf courses, and field crops where heavy machinery traffic is general, clogging by roots may become a severe problem (personal observation). It is plausible that frequent traffic on dripirrigated fields causes a direct contact between emitters and the soil, which allows roots to grow into the orifices. In the absence of pressure on emitters, a cavity a few millimeters in diameter is formed around the orifice because of soil displacement by the outflowing water. When irrigation terminates, the water drains from the cavity and the hemispherical void around the orifice prevents root growth into the emitter. Root growth around the trickler orifice can be prevented by controlled herbicide release via the emitter. The most commonly used herbicide is trifluralin (e.g., Dornai et al., 1991), which has been incorporated into the plastic dripper casting by various manufacturers.

In. PRINCIPLES OF FERTIGATION Fertigation management is aimed at maximizing growers’ income and minimizing environmental pollution. Attaining these objectives depends on economic factors (input costs and produce value) and on crop growth, yield, and fruit quality. The incorporation of economic, environmental, and agronomic considerations into one management decision tool has not yet been attempted. The main problem in implementing such an integrated approach is the inability to predict plant response, in yield and fruit quality, to variations in daily input levels that stem from economic constraints (e.g., price of gas, fertilizers, or fruit). A few studies have

20

B. BAR-YOSEF

been initiated to link economic and agronomic considerations by optimizing CO, enrichment and air temperature in tomato greenhouses, on the basis of expected fruit price and input costs (Seginer, 1989; Aikman et al., 1996). Another attempt was reported by Wu (1993, who developed a simple model to optimize microirrigation on the basis of the price of water, the value of projected yield, and the cost of cleaning underground water, contaminated by deep leaching caused by inefficient water use. Doerge and Thompson (1995) used net value of yield and deep N leaching as criteria to determine recommended N fertigation levels in vegetable crops. The latter two studies allow ad hoc evaluation of field operations, but cannot yet be used as a real-time management decision tool. In the absence of an integrated agronomic-economic-environmental management approach, current fertigation recommendations are aimed at maximizing yield and fruit quality. The following sections discuss the question of how this objective can be met with respect to sustaining a clean environment.

A. QUANTITY CONSIDERATIONS It is assumed that yield and fruit quality are determined by two functions: (i) dry matter (DM) production and partitioning among plant organs as a function of time; and (ii) nutrient concentration (NC) in plant organs as a function of time. Unique DM and NC functions (both being independent of soil properties) determine unique crop yield and quality. The product DM(t) X NC(t) defines a cumulative consumption function Q(r).The DM(t) and Q(r)that result in maximum yield and fruit quality under given climatic conditions are termed the objective (target) DM and nutrient uptake curves. The practical importance of the objective Q(t)curves is that they indicate the minimal daily application rates of given nutrients necessary to maintain them at a steady-state concentration in the soil. The actual fertilization rate should account for the fertilizer use efficiency by the plant, EF, (EF < 1) and must, therefore, be Q/EF. Under good management EF exceeds 0.80 (Shevah and Waldman, 1989). An additional quantity factor in the soil-plant system is the water consumption during the course of the growing season that facilitates uninhibited plant development. The transpiration function depends on climatic conditions and on crop characteristics (Stanhill, 1985; Hatfield and Fuchs, 1990) and the actual irrigation rate should exceed transpiration as it must also compensate for evaporation from the soil surface and leaching of salts outside the root zone. The objective DM(r) and Q(t)curves are obtained empirically by growing crops under a certain range of fertigation regimes and climatic conditions, and determining the DM and NC in crop organs as a function of time, and the temporal crop yield and quality (e.g., Bar-Yosef, 1991). This is a very tedious procedure that ideally could have been replaced by crop-soil-atmosphere models and simulation of

ADVANCES IN FERTIGATION

21

growth, uptake, and yield under various fertigation regimes. Unfortunately currently available models (see Section V) do not predict fruit quality and, therefore, cannot replace empirical work.

B. INTENSITY CONSIDERATIONS There are two major intensity factors that must be known for proper fertigation management: (i) root concentration distribution in the soil; and (ii) nutrient concentration distribution in the soil solution. The two functions must be adjusted to enable plants to absorb nutrients according to the objective Q(t)function. Root and nutrient concentrations have complementary effects, since uptake rate is the integral of the flux times root surface area (or length) in a given soil subvolume, over the total number of subvolumes in the soil. The flux, F[mol (cm root)-' s-'1 is determined by the nutrient concentration in the soil solution at the root surface, Cr (mol liter- I ) , as shown by the Michaelis-Menten equation: F = Fmax Cr/(K,

+ CJ.

(4)

Here Fmax (maximum F') and K,,, (M> are plant coefficients determined in stirred nutrient solution experiments. Representative values of K, and Fmax for various crops are presented in Table IV. The integration of uptake based on Eq. (4) over the entire root zone is the basis for several models that simulate nutrient uptake in relation to plant growth (e.g., Jones and Kiniry, 1986; Marani et al., 1992; Fishman and Bar-Yosef, 1995; Abbas et al., 1996; Lafolie et al., 1997). When nutrients accumulate inside the root, the internal concentration CpImay affect F. In such cases (e.g., nitrate uptake by roses, Brun and Morisot, 1996) Eq. (4) may be replaced by F = KT (C, - C,,,),

(5)

where KT (cm s-') is the root ion transfer constant. An approach similar to (5) was used to simulate water influx into roots, with the concentration difference being replaced by root water head and water pressure head at the soil-root interface, and KT being replaced by a coefficient that accounts for the soil-root interface hydraulic conductance and geometry (Gardner, 1960). The mechanism of ion uptake based on root length and nutrient concentrations in the soil solution is challenged by another approach that maintains that as long as nutrient quantities in the root zone are sufficient, plants take up as much nutrients as their capacity for them demands, regardless of the nutrient concentrations in the soil solution and independently of plant root length (Steiner, 1996). The capacity depends on plant's physiological stage and is independent of the ionic composition of the nutrient solution (Andre et al., 1978). For the specific case of nitrate, for example, Drews et al. (1995) found that the nitrate uptake rate was

22

B. BAR-YOSEF Table IV and K,,,for NO,, P, and K of Several Plant Species Using Intact Plants Values of FmaX in Well-Stirred Solution Culture Fmax[(mol cm-'

Plant species Corn (Zea mays)

Soybean (Glycine man) Wheat (Tritium vulgare) Tomatod (Lycopersicon esculentum) Barley (Hordeum vulgare)

s-I)

x 1013]

NO,

P

K

1.16 -

0.50 6.1 4.0

5.02

-

K, [(mol liter-') x 10'7 NO,

P

10

3.0

-

1.o

40

70 10.3

1',6

0.09

-

2.0

-

0.18

0.88

-

6.0

6.6 -

-

600 258

436 -

-

34.9

-

-

-

16

5.5

0.10

5.1

Ref! I" 2 6

-

7.5

K

-

I

1

-

5

15

4

3

"Assuming root radius = 0.02 cm. "References: 1, Barber (1984); 2, Bar-Yosef (1971); 3, Ben Asher et al. (1982); 4,Glass and Perley (1980); 5 , h a s et al. (1997a) for P and h a s et al. (1977b) for NO,; 6, Williams and Yanai (1996). Tmln in Barber's data disregarded of NH, in tomato are 1.1 dData from ref. 3 were recalculated.According to Ref. 5 , K,,,and FmaX mM and 26 X 1013 mol cm-l s-I.

determined solely by the availability of soluble carbohydrates in the plant. The latter approach lacks a physicochemical and biological basis, in the present state of knowledge, and, therefore, is not further discussed in this review.

C. LMPLICATION OF THERELATIONSHIPSBETWEEN UPTAKE FLUXAND CONCENTRATION Even without management models, Eq. (4) can be used to define some threshold nutrient concentrations in irrigation water. When C, > Km, the increase in F due to an incrementalincrease in Crdiminishes rapidly, and it is inadvisable, therefore, to maintain at the root surface a concentration that sustains an F that exceeds 0.75Fm,, (i.e., C,= 3Km). Concentrations exceeding this threshold C,value may contributeto salinity and cause reduced influxes of other nutrients (Fried and Broe-

ADVANCES IN FERTIGATION

23

shart, 1967;Fishman and Bar-Yosef, 1995).Another application of Eq. (4) is to estimate, as a function of time, the minimal active root length (or weight) (R,) required to facilitate uptake rates according to Q(t):

R,

=

Q(NF,,,.

(6)

The concentration of a nonadsorbing nutrient in the irrigation water (Cw)is a first approximationto its concentration in the bulk soil solution (C,), but not to that at the root surface (C,). For adsorbing nutrients (e.g., P, K), Cwshould be corrected for adsorption. The difference between C, and Cb stems from the rapid depletion of nutrients by the root, and the slower transport of nutrients from the bulk soil to the root surface; this difference diminishes as the fertigation frequency increases. Assuming a constant rate of solution flow into the root, W[ml (cm root)-' s-l], and steady-state soil volumetric water content (0) and soil nutrient concentrations, the relationship among Cb,W, and C,., is given by Eq. (7), which is derived from Eqs. (14) and (4): C,/C,

= 1/A - B(l/A - l)/(K,

+ C,)].

(7)

b and a (cm) are the midway distances between roots and Here, A = (b/u)W'2TDp, root radius, respectively, B = F,,,/W, and Dp (cm2 s - l ) is the diffusion coefficient of the nutrient in the soil solution, defined as Dp = K Doexp(a 0) (Olsen and Kemper, 1968), in which Dois the diffusion coefficient in water and a(-10) and K(-0.001 to 0.005) are soil constants; K decreases as the soil surface area increases. More discussion of D p and its dependence on soil properties is given in Section IIIF. W is related to the solution velocity toward the root, v (cm s- I ) by W = v2Tae. Under high-transpiration conditions (e.g., 10 mm day-') and a crop with 9000 kg fresh roots/ha (characteristic of sweet corn, see Section IIIF) the average W is 0.014 ml (cm root)-' day-' (a = 0.02 cm), and v = 0.45 cm day-' (=5.2 E-6 cm s- I ) . At this velocity, the hydrodynamicdispersion coefficient is negligible and Dp is determined by the thermal diffusion coefficient (Olsen and Kemper, 1968). Assuming b = 2 cm, 0 = 0.25 cm3 cmP3, Dp = 6.0 E-8 cm2 s-I and Fmaxand K, of NO, for corn (Table IV), Eq. (7) yields a CJC, ratio of approximately 0.5, indicating nitrate accumulation (rather than depletion) at the root surface. Under lower 0 (=O. lo), transpiration is limited by soil hydraulic conductivity. Assuming that v = 2.3 E-9 cm s - I and that the other parameters are unchanged, Eq. (7) yields a Cb/Crratio of 1. Assuming a 10-fold greater Fmaxand 10-fold smaller K, (which represent the possible experimental variability in evaluating these parameters) gives Cb/Cr 10.A similar ratio is obtained for 8 = 0. l, the original values of Fmaxand K,, but smaller Dp (1.4 E- 10 compared with the reference value of 6.0 E-8 cm2 s-l). A CJC, value of -90 is obtained for 0 = 0.10, v = 2.3 E- 10 cm s- I , reference Fma and K,, and Dp = 1.4 E- 11. In several studies the parameters in Eq. (4) were determined in unstirred solu-

-

24

B. BAR-YOSEF Table V

MichaelisMenten Constants for N, P, and K for Pepper, Tomato, and Lettuce Plants Grown in Unstirred Solution Cultures and in Two Growth Substrates Fmsxr

Crop" Pepper Tomato

Lettuce

Nutrient Nb

P N P K P

[(mol cm-'

Kms

(phfl

System

Ref"

14.0 17.0 11.0 1.8

550 25 3000 320

1 1

3 .O

1000

2.3

136

Unstirred solution Unstirred solution Sand Rockwool Aerophonic Sand

SKI)X

loi3]

2 3 4

5

"Assumed root radius = 0.02 cm. hNH,-N:NO,-N in solution = 1:3 'References: 1, Bar-Tal et al. (1990); 2, Bar-Yosef (1991); 3, Bar-Yosef and [mas (1995); 4, BarTal er al. (1994); 5, [mas and Bar-Yosef (1997).

tions or in growth substrates. In such circumstances, K , is expressed as a concentration in the bulk soil solution (K,,,, Table V). The disadvantage of this approach is that Kms depends not only on the crop root properties but also on soil characteristics and fertigation regime. Comparison between Km,(N) of tomato plants grown in sand (Table V) and K , (Table IV) shows that Kmsis approximately 10-fold greater than K , and Fmax(tomato)in the soil is twice as great as Fmaxin stirred solution, which is a surprisingly good agreement. The K,,/Km ratio of 10 indicates that C, in sand is approximately one-tenth of Cb,or C,. This ratio is within the Cb/Crrange discussed above.

D. COUPLING QUANTITY AND INTENSITY FACTORS For best management results the quotient Q(t)/(daily irrigation rate) (=C) should equal a C, value that will sustain the F required by the roots. The daily irrigation rate (mm/day) is determined according to the estimated reference evapotranspiration multiplied by a time-dependent crop coefficient that accounts for the partial covering of the soil surface by the canopy (Hatfield and Fuchs, 1990). The irrigation scheduling is determined according to soil water status, monitored with tensiometers for matric potential (Richards, 1965), and neutron probe, time-domain reflectometry, or capacitance meter for volumetric water content (Gardner, 1986; Dalton et al., 1984; Bell et al., 1987). Reference daily evapotranspiration can be estimated from class A pan evaporation (Phene et al., 1990b) or calculated from meteorological data (Hatfield and Fuchs, 1990). When the irrigation rate is

ADVANCES IN FERTIGATION

25

high (e.g., summer time) and root systems are confined (e.g., under high-frequency drip irrigation), C may be too low to furnish the uptake rate required by the plants, as the integral of flux over the root length is too low. In such a case, the nutrient application rate should be raised above the target Q(t),which is, however, a wasteful and environmentally undesirable practice. To avoid such problems, plants can be grown with larger root systems, such that a lower Cr may be sufficient to maintain the target Q(t). Large soil root volumes have a high buffering capacity for water and nutrients, which reduces possible stresses arising from unexpectedly interrupted supplies. However, large root volumes cannot be rapidly enriched with or depleted of nutrients, so that the ability to control uptake according to timespecific plant needs is reduced.

E. ROOT GROWTHAND DISTRIBUTION IN SOIL Root growth depends on transport of carbohydrates and hormones from the canopy, and on the physical and chemical conditions that prevail in the soil. The mechanism controlling carbon supply partitioning between plant organs and roots is still unclear. Timlin et al. (1996), in their comprehensive crop-soil-atmosphere model, assumed that carbon allocation to roots is a function of leaf water potential (Acock and Trent, 1991). Fishman et al. (1984) suggested that carbon partitioning is determined by the relative sink power of plant organs, defined as the carbohydrate content of the organ divided by the total carbon content in plant. Hormone effects on root growth have not yet been treated quantitatively, and are not discussed in this review. Root growth models distinguish between potential and actual root growth rates. The potential root growth rate is described as a first-order reaction with respect to root length, L (cm), dLldt = PL, where P is a rate constant ( t - ' ) (Hoogenboom and Huck, 1986). The potential root growth rate for the entire plant is the integral of the individual rates in the soil subvolumes over the total number of subvolumes in the wetted soil domain. If the potential growth rate exceeds the actual rate of carbon supply by the canopy, allocation to individual soil subvolumes is proportional to their relative root growth rate. Soil factors affect root growth by influencing P (Bar-Yosef and Lambert, 1981) and by affecting root extension from a given soil subvolume into neighboring soil cells. The important soil factors are: (i) soil moisture content, which determines the impedance to root penetration into the soil at a given soil compaction (Bar-Yosef and Lambert, 1981); (ii) oxygen, P, and N concentrations in the soil (Bar-Yosef and Lambert, 1979); (iii) presence of elements that are toxic to roots (Marschner, 1995); and (iv) soil temperature. Factors i and ii explain how fertigation rate and frequency, which determine 8 and nutrient concentrations and distributions in the wetted soil volume, affect root growth and spatial and temporal distribution in the soil. Soil temperature may explain the

26

B. BAR-YOSEF

effects of soil depth on root growth and morphology; at low temperatures roots are smaller and less branched than at the optimal temperature for root growth (Top,). A list of Toptfor several field crops is given by McMichael and Burke (1996). The tap root growth rate of sunflower doubled as the root temperature increased from 10 to 20"C, increased by 20% when the temperature was further increased to 25"C, and then decreased by -35% as the temperature was raised to 35°C. In cotton, the tap root growth rate doubled as T was raised from 20 to 35°C and then fell by 35% as T was increased to 40°C. The root growth rate of maize increased by a factor of -3 as T was raised from 15 to 20"C, and by a factor of 2.3 as T was elevated to 30°C (McMichael and Burke, 1996). Ganmore-Neumann and Kafkafi (1980) found that tomato plants growing in a solution rich in NO, developed long, thin, branched roots whereas those grown in a solution rich in NH, developed short, thick, second- and third-order roots. The highest dry root weight was obtained when the solution temperature was between 16 and 24"C, for all studied ammoniurnhitrate ratios, and decreased when T was elevated to 34°C. The combination of 10 mM NH, without NO, at T > 24°C resulted in very strong decreases in dry root and shoot weights in comparison with 24"C, whereas at 16°C the root weight in the presence of NH, alone was only slightly lower than at NH,/NO, = 5 / 5 , which gave the maximum dry root weight. Similar adverse effects on root and shoot growth, of NH,/NO, ratios exceeding 7/3 at T > -25"C, were reported by Ganmore-Neumann and Kafkafi (1983) in strawberry. Due to the meager understanding of the mechanisms controlling root growth under varying soil conditions, this relationship has not yet been incorporated in soil-crop modeling. The modeling of three-dimensional root growth on the basis of root morphology and architecture (e.g., Clausnitzer and Hopmans, 1994) is too complex to be applied in fertigation problems and is outside the scope of this review. Sample root distributions of drip-fertigated, high-yield pepper, tomato, muskmelon, and sweet corn in soil are presented in Tables VI and VII. All root systems were restricted to a soil cylinder 40 cm in radius, which coincided with the horizontal water front position (data not shown). The depths of the tomato and pepper root systems were shallower than those of sweet corn and muskmelon, apparently because of inherent differences in root growth characteristics. The experimental root weight of tomato (Table VI) can be compared with the theoretical minimum root weight necessary to furnish the tomato plant with its maximum Q(N)[Eq. (6)]. Assuming a Q(N) of 2.5 kg N ha-' day-' (see below), FmaX(NO3) of (5E- 13 mol cm-' s-') (Table V), and 8% root dry matter content, a minimum dry root weight of 400 kg ha-' is obtained. The close agreement between the calculated and experimental tomato root weights indicates that at the time of peak N consumption rate, the roots of drip-fertigated tomato plants must absorb N at a flux that approaches its saturation value. To approach Fmax, the N concentration in the soil solution at the root surface must be >3Km.

ADVANCES IN FERTIGATION

27

Table VI Root Distribution in a Sandy Soil and Total Root Weight of Drip Fertigated Pepper and Greenhouse Tomato Plants That Gave Optimal Yields (Note the Different Units of Root Density, Presented in Their Original Forms)" Pepper

Tomato Lateral distance

Depth (cm)

0 10

11 20

21

31

0

30

40

10

5 6

320 45

87 12

4

45

2

28 22 1

34 45 60

11 20

21

31

30

40

41 50

% of total dry root wt

mg dry rootkg dry soil

in a sampled grid 0-10 11-20

21-30 3 1-40 41-50

15 6 5 2

12 11 8 1

-

51-60

-

-

I1

I 9 2

1 Total dry root weight (kgha)' 1900 400

39 63

43 60 15 1

9 36 85 15 28 1

103 103 62

63 22 1

"Sources: Pepper (cv. Maor), Bar-Yosef (1991); tomato (cv. F-144),Bar-Yosef et al. (1992). bFrom the stem toward the edge of the bed, perpendicular to the row. 'Plants stand: pepper 100,000, tomato 23,000 plantsha.

When calculating the minimal root weight for maximum N uptake rate by sweet corn (Q = 6 kg N ha-' day-') [Eq. (6)], if the available F,, of grain corn (Table IV) is used, a fresh root weight of 54,000 kg ha-' is obtained, which is appreciably higher than the sweet corn fresh root weight found under field conditions (9000 kg ha-'; Table VII). The discrepancy indicates that sweet corn, which has a much shorter growth season, might have a considerably higher F,, and lower K, than grain corn. A comparison between sweet corn root masses and distributions in soil under surface and subsurface drip fertigation, respectively, was presented by BarYosef et d.(1995a). In general, the results are similar to those in Table VII, with a distinct effect of deeper emitter placement (0, 30, and 45 cm) on root depth, which shifted downward by 15 and 30 cm, respectively, relative to surface placement. The presented root weight of pepper plants is appreciably greater than the dry tap root weight of chili pepper (600 kg ha-' at a comparable plant age) reported by Beese et al. (1982). However, the total dry matter production in the chili pepper was also considerably lower than in the current case. The root weights of tomato plants presented here are similar to previously published root weight data (BarYosef, 1991).

-

28

B. BAR-YOSEF Table VII Relative Root Density” Distribution in the Soil6 and Total Fresh Root Weighte of Drip Fertigated Sweet Corn and Muskmelon Plants That Gave Optimal Yields Sweet corn

Depth (cm)

0 10

11 20

0-10 11-20 2 1-30 31-40 41-50 5 1-70 7 1-90

48 100 90 67 30 11 10

94 84 78 42 50 24 5

Muskmelon

21 30

Lateral distance (cm)d 31 Depth 40 (cm)

Relative root density (W) 47 77 &I0 82 104 1 1-25 110 50 26-40 52 50 41-60 58 36 61-80 9 28 81-100 6 0 Total fresh root weight (kgha) 9000 I00

0 10

11 20

21 30

100 73 19 92 38 53

54 54 42 35 38 23

50 46 35 42 23 7

Source: Bar-Yosef and Sagiv (1985). “Relative root density = root density (mg dry rootkg soil) in a given soil cubehoot density in a reference soil cube. The reference root densities are 133 and 13 mg dry rootkg soil for sweet corn and muskmelon, respectively. The reference soil cube was 0-10 cm from the emitter. %weet corn in loess soil; muskmelon in sandy soil. ‘Total root weight was estimated by multiplying root density by soil weight represented by the soil cube and summing over the sampled soil volume. %om the stem toward the edge of the bed, perpendicular to the row.

F. RHIZOSPHERIC PROCESSES The part of the soil volume that is directly influenced by roots, called the rhizosphere, extends several millimeters from the root surface into the bulk soil (Mengel and Kirkby, 1987). The effect of the root is exerted by its uptake and release or organic and inorganic compounds, which influence biological, physical, and chemical processes in the soil. The released compounds include root debris, mucilage, and root exudates containing low-molecular-weight organic solutes (sugars, amino acids, organic acids), gases (CO, and ethylene), and protons (Marschner, 1995). Soil microorganisms utilize the released carbohydrates and amino acids, oxidize them by competing with plant roots for soil 0,, and release CO,, which reduces the soil pH. Certain rhizosphere microorganisms release phosphatases that mineralize organic P compounds (Jungk, 1966); others exude siderophores that chelate Fe3+ and enhance its transport in soil (Marschner, 1995). Protons, carboxylates, and phytosiderophores, released at differing intensities

ADVANCES IN FERTIGATION

29

by roots of various plants, play an important role in mobilizing P, Ca, and microelements in soil. Discussion on the role of root exudates in plant nutrition in relation to fertigation follows.

1. Carboxylate and Proton Release by Roots Excess of anion over cation uptake is accompanied by H+ influx into the root to preserve cell electroneutrality. In practice this occurs in the presence of nitrates in the root-surrounding solution, and results in a pH increase in the solution. Before being metabolized by plants, the absorbed nitrates are reduced in a reaction that takes place in both the shoots and the roots in a proportion that differs among species and depends on nitrate concentration in the rooting medium: NO,- + 8Hf + 8e- = NH, + OH- H,O (Marschner, 1995). The generated hydroxyls must be either excluded or neutralized, to prevent a pH increase in plant cells. Since OH- is immobile in the phloem, it has to be neutralized by carboxylation and formation of phloem-mobile anions of carboxylic acids (Marschner, 1995). Part of the carboxylates exude to the root-surrounding solution, thus replacing an equivalent influx of protons. When excess uptake of cations over anions takes place (e.g., under NH,+ or K + fertilization), electroneutrality is maintained by H+ efflux, which decreases the outer solution pH. It is apparent, therefore, that by varying the NH,/NO, ratio in fertigation solutions, a partial control over proton and carboxylate exudation by roots can be obtained, with direct effects on the external solution pH. The NH,/NO, effect is stronger under high than low fertigation frequency, due to the continuous supply of fresh NH, to the soil in the former case. The predominant carboxylates found in plant tissues are malate, citrate, oxalate, succinate, and malonate (Marschner, 1995). Their relative concentration varies among crops and within plant organs, and is affected by nutrient supplies. Carboxylates can complex divalent and trivalent cations at differing stability, and thus play a role in solubilizing Ca- and Al-P minerals. Carboxylates are also adsorbed by metal oxides and clay minerals which have pH-dependent charge, increase the negative surface charge, and desorb P and other oxyanions (Bar-Yosef, 1996b). Sorption-desorption reactions are discussed in more detail in Section IIIF2. The carboxylates effective in mobilizing Pin soils are, in descending order, citrate, oxalate, and malate (Bar-Yosef, 1996b). Rates of citrate exudation by tomato plants and variation in pH in nutrient solution in response to the NH,/(NO, + NH,) ratio (R) are presented in Figs. 4 and 5. Citrate exudation varied between 0 at R = 1 and 0.3 kmol/(pl 6 h) at R = 0. Assuming a root-affected soil solution volume equal to the fresh root weight of the plant, and a 6-day exudation period, the cumulative citrate concentration in this volume is 0.72 mM. The effect of this concentration on P mobilization depends on the adsorbing surface properties, total P concentration, and pH of the system. Under P deficiency conditions in most soils this citrate concentration might be very beneficial in providing P to plants (Bar-

+

30

B. BAR-YOSEF

0 .2

-

0.1

-

b=4.81 z1.78

*

u l-

a LI:

k 0

0.0 0.0

0.2

0.4

0.6

0.8

1.0

S 0 L UT 10N NHa/( NH4+N0,)

Figure 4 Citrate exudation rate (EX) by tomato roots as a function of NH,/(NH, + NO,) molar ratio (R)in nutrient solution. Plant age = 30 days. NH, + NO, = 7.5 mM. Root and shoot fresh weights were 19 5 2 and 30 IT 2 glpl, respectively. Reproduced from Imas et a!. (1997a). with kind permission from Kluwer Academic Publishers.

Yosef, 1996b).A similar relationshipbetween R and malate exudation has been reported for maize (Kraffczyk et al., 1984). The data in Fig. 5 show that the R value at which tomato plant did not change an initial solution pH of 5 was between 3 and 3.3, depending on plant age. When N was supplied solely as NO, or NH,, the solution pH increased to 8 or decreased to approximately 3.3, respectively. The maximal H+ efflux from the roots was 18 pmol H+/(pl 6 h) (Imas et al., 1997a). On the basis of the same assumptions as above, the accumulated H+ is shown to reduce the pH in the root-affected soil solution to -2.7. Similar variations in rhizosphere pH in response to R have been reported for other crops as well (Barber, 1984; Marschner, 1995).

2. Sorption-Desorption Sorption comprises three chemical reactions that coexist in soils: adsorption (and exchange), precipitation, and fixation. Together these reactions determine ion partitioning between the solid and liquid phases of soils. The three reactions differ in their retention mechanisms (see later) and kinetics. Sorption is characterized by an initial rapid reaction (adsorption) that takes minutes to hours to complete, followed by the slower precipitation and fixation reactions. Precipitation takes hours to several weeks, and fixation several weeks to several months to reach equilibrium (Sparks, 1986; Barrow, 1987).

ADVANCES IN FERTIGATION 9

31

DAP 30

-*-

8

I

-.-A-

7

37

44

I

P

6

5 4-

3 2 ”

0.0

0.2

0.4

0.6

0.8

1.0

S 0L UTI0 N NH4/(NH4+N 0,) Figure 5 Nutrient solution pH as a function of NH,/NH, + NO,) molar ratio ( R ) in the solution and tomato plant age (DAP). NH, + NO, = 7.5 mM. Root and shoot fresh weights were 19 ? 2 and 30 t 2 glpl, respectively. Reproduced from Imas et al. (1997a). with kind permission from Kluwer Academic Publishers.

Adsorption involves chemical and electrical forces induced by charged and chemically reactive surfaces. The chemical component of adsorption incorporates an inner sphere complexation between surface lattice atoms and surrounding free ions. The adsorbed ions are assumed to be organized in two or three discrete layers that form capacitors in parallel, the last one comprising a diffuse layer extending toward the solution phase. The sum of the capacitor charges balances the adsorbent surface charge, thus constituting the electrical component of adsorption. Such models are used to describe competitive adsorption of oxyanions and microelements by pH-dependent charged surfaces such as metal oxides and edge faces of clay mineral platelets in soil. Reviews on adsorption models and their application can be found in Sposito (1984) and Barrow (1987). Several studies have shown hysteresis in P adsorption-desorption by metal oxides and clay minerals in soil (Barrow, 1987). In all cases P release from those surfaces was slower than P retention, but the methods by which desorption was determined affected the ionic composition of the equilibrium solution (e.g., lattice dissolution due to large dilution), so that the actual hysteresis effect under field conditions is hard to evaluate. Barrow (1987) developed a pragmatic model that describes oxyanion desorption as a function of solution concentration and time, according to which desorption and adsorption do not match. The retention of macrocations in soil is governed by electrical forces exerted by the permanent surface charge of clay minerals. Simultaneous Ca, Mg, K, Na, and

32

B. BAR-YOSEF

NH, retention can be expressed by binary, ternary, and tertiary exchange isotherms, which take into account the soil cation-exchange capacity, specific exchange constants, and cation concentrations in the soil solution. These exchange models predict cation adsorption in soil suspensions quite well. More details about these models can be found in several reviews, for example, Sposito (1989). The aforementioned adsorption and exchange models are too complex for application in fertigation management or even in transport problems. The main limitations stem from the numerous parameters involved, the fact that several equations must be solved simultaneously, and the complex partial differentiation with respect to ion concentration and pH that defines the buffering capacity of the system. These limitations stress the need for simpler models that can be differentiated and that describe ion partitioning as a function of total ion concentration, concentration of competing ions, and pH. An example of such a model, which is suitable for P fertigation problems, is the modified Langmuir competitive adsorption model (Bar-Yosef et al., 1988b; Katou et al., 1996): n

A = T x Kj Mi/ [1 + j=l

x m

K jMi].

i=l

Here A is total P adsorption (mol/kg), K is the affinity of adsorbing species to the surface, M is activity in solution,j is the index of the n P species assumed to be adsorbed (H,PO,, HPO,, PO,, CaPO,), and i is the index of the n P species plus the (m-n) competing ions (e.g., OH, citrate). The parameter T is the maximum adsorption sites (mol/kg), and is defined as T = To exp[G (RPH/pH - l)], where To = T when the pH is equal to a reference pH (RPH), and G and RPH are soil-specific parameters. Under conditions of constant pH and absence of competing ions, Eq. (8) is the Langmuir adsorption isotherm ( A = K T M l [ 1 K W ) .For many soils the Langmuir model is sufficient to describe P partitioning between the solution and solid phases. Equation (8) was used to describe pH-dependent Zn adsorption by soils (Bar-Yosef, 1979) and was incorporated into a transport model describing Zn movement to roots (Bar-Yosef ef al., 1980b). To estimate K, NH,, and Na partitioning in fertigation problems, binary exchange of each of these cations against Ca should provide an acceptable solution, as long as Ca is the predominant cation in soil. A Gapon-type equation (9), with M denoting the monovalent ion, can be used for this purpose:

+

Here [ ] and ( ) stand for the concentrations of exchangeable cations in the solid (mmol +/lo0 g) and liquid (mM) phases, respectively, and KG ([mmol(+)/lOO g]

ADVANCES IN FERTIGATION

33

[ n M - $is the exchange constant. Calcium and Mg are summed due to their similar chemical characteristics. The value of KG for M = Na is 0.015 in a large number of soils in the western United States (Shainberg and Oster, 1978). KG data for K and NH, are not readily available, but a first approximation of 5-10 times greater than KG of Na might be useful (unpublished data). The time needed to attain equilibrium in adsorption and exchange reactions is usually much shorter than the residence time of ions moving in soil. This explains the common assumption made in transport models that ion partitioning in soil is instantaneous. Precipitation occurs when the solubility product of a given mineral (Ksp) is exceeded by the ionic activity product of the pertinent ions. Precipitation is affected by solution pH and ionic composition, through their influence on speciation and ion activity coefficients. Under alkaline conditions, particularly in calcareous soils, precipitation reactions may determine the P, Ca, Mg, Fe, and Zn ion concentrations in soil solution. In the case of P, the sequence of solid-phase transformations is monocalcium phosphate -+dicalcium phosphate-+octacalcium phosphate+hydroxyapatite. The critical pH for Ca-P precipitation is 7.2, which is the second pK of the o-phosphoric acid. Information about Ksp values of Ca-P minerals can be found in Lindsay (1979). In acid soils, the P minerals that may control P solution concentration are AIPO, and FePO, (Lindsay, 1979). The minerals Zn,(PO,), and Mn,(PO,), have very low K s p values and may compete with Zn(OH), and Mn(OH), as solid phases determining Zn and Mn ionic concentration in soil solutions. Precipitation poses severe problems in stock fertigation solutions and microfertigation emitters, particularly when the Ca concentration exceeds 0.1 mM and pH > 7. Under such conditions, CaHPO, may crystallize, followed by less-soluble Ca-P minerals according to the transformation sequence. CaCO, may precipitate at pHs approaching the second pK of H,CO, (= 10.3).Gypsum (CaSO, 2H,O log Ksp = -4.64) may precipitate in the presence of SO,*- (e.g., when ammonium sulfate is used in fertigation). Details on the solubility of some of the abovementioned salts are given in Section IVD. Precipitation kinetics depends on the rate of two processes occurring in series: crystal nucleation and crystal growth, both of which are prone to inhibition by certain components found in the plant rhizosphere (Inskeep and Bloom, 1986). In practice these processes cannot be separated, and the overall precipitation-dissolution kinetics is usually approximated by

dSldT

=

K (C, - C,).

(10)

Here S is the solid-phase concentration, and C, and C, are the solution-phase concentrations at time f and at equilibrium, respectively (Enfield et al., 1981). When C,> Ce precipitation takes place; when C, < C, dissolution occurs. Bar-Yosef et af. (1989) found that the rate of orthophosphate disappearance from solution, in a

34

B. BAR-YOSEF

+

loess soil suspension, obeyed the Elovich equation, C, = a b In t, where a and b are soil constants. The slowest reaction accounting for P and K disappearance from soil solution is occlusion (fixation). Barrow (1987) described P occlusion as diffusion-controlled P penetration into metal oxide and clay mineral lattices. Potassium is fixed in soils by penetrating into clay interlayer sites of expanded 2:1 clay minerals. Occlusion reactions are too slow to affect ion partitioning during fertigation and are not discussed in this review. The sequence of events following P fertigation can be summarized as follows: During the first few hours after termination of fertigation, P partitioning is determined by adsorption. If the resulting 0-P and Ca ion activity product (IAP) in solution exceeds the Ksp of any of the Ca-P minerals mentioned above, the 0-P concentration will decrease according to Eq. (10) until IAP < Ktp. Adsorption is rapidly adjusted to the new solution 0-P concentration. As the dissolution rate of P minerals is slow relative to P adsorption, the immediately available P is the P found in solution plus the adsorbed P. In acid soils the theoretical considerations are similar, only the predominant cation is A1 rather than Ca.

3. Transport of Ions in Soil toward Roots The transport of ions from the bulk soil to the root sink affects their concentration in the soil solution C at the root surface, and hence their flux of uptake by the root. The governing equation describing ion transport to a smooth, uniform cylindrical root under steady-state moisture is

[b(c) + el - =

- D

:r(

-

p:)

t(

+-

D p:)

+ -::r-(qoroC).

(11)

It is derived from Eq. (3) for the r coordinate only, with q,, r,, and Dp denoting solution velocity at the root surface, root radius, and diffusion coefficient in the soil solution, respectively. For q, values prevailing in soils, the mechanical dispersion coefficient [Eq. (2)]is negligible and Dhp= Dp(Olsen and Kemper, 1968). The boundary condition at the root surface that best complies with the fertigation principles discussed above is

The outer boundary condition,found either at the edge of the root-affected soil volume or midway between adjacent roots, whichever is smaller, (R) is aC,/ddr

= 0.

(13)

ADVANCES IN FERTIGATION

35

The steady-state solution of Eq. (1 1) (Olsen and Kemper, 1968,their Eq. [58]) subject to boundary conditions (12) and (13) is

Fm,, c, = W(k,/C, + 1)

(14)

where W = q, IT ro 8 and b = R. This equation was used to estimate the ion concentrations at the root surface from known bulk soil solution concentration [Eq. (7)]. Another application of Eq. (14) is to estimate nutrient uptake rate by plants, Q(kg m-’ h-I), in models that calculate temporal root weight (W) and nutrient concentration distributions in the soil. This is done by assuming at each time step that a quasi-steady-state nutrient concentration exists in the soil cylinder surrounding the root, which enables C, to be calculated according to (14) for given C , (= C,) and R. Once C, is known, F is calculated by inserting C, into Eq. (4) (e.g., Fishman and Bar-Yosef, 1995). The rate of uptake by the plant is Q = Z(WiFi),summed over the entire number of subvolumes in the soil domain. The use of Eq. (14) is more suitable than calculating fluxes by inserting the bulk soil solution concentration (C,) into the Michaelis-Menten equation [(4)], as the latter approach does not account for the rapid depletion of nutrient concentration at the root surface, caused by uptake. The diffusion coefficient of a given ion in soil solution, Dp (cm2 s - l ) , has been by three different emrelated to the diffusion coefficient of the ion in water (Do) pirical equations: Olsen and Kemper (1968) suggested Dp = Doaebewith b = 10 and a = 0.005 to 0.001 depending on soils; Nye and Tinker (1977) used Dp = D J 8 , wherefis a compound soil impedance factor (<1); and Mualem and Friedman (1991) found Dp = Do(OS Se2.5,where OS and Or are saturated and residual volumetric water content, respectively, and S, is the effective saturation degree [=(€I - 8,)/(8, - Or)]. The three equations stress the importance of 8 in determining ion transport in soil, and allow for specific soil characteristic effects on Dp.The soil impedance,f, decreases (higher soil tortuosity) as the clay content of the soil increases, and as 8 decreases. The dependence of D, on the soil bulk density (4 follows a bell-shaped curve: As d increases up to a certain value (e.g., by soil compaction) 8 increases and Dp increases too. After the optimum value of d is exceeded,fdecreases due to increased soil tortuosity while 8 does not change any more. More details on soil factors that affect D, can be found in a review by Jungk (1996). Representative Do(cm2 s- I ) values of some nutrients are: NO,- 1.9 E-5; H,PO,- 0.89 E-5; K+ 1.98 E-5; Ca2+ 0.78 E-5; Mg 0.70 E - 5 (Barber, 1984). Williams and Yanai (1996) compiled literature data on all the parameters that appear in Eq. (12) for various nutrients and crops. The data exhibit considerable variability, which stems from their derivation from a variety of experimental conditions and lack of standard experimental methods.

36

B. BAR-YOSEF

G. WATERQUALITY CONSIDERATIONS Irrigation water quality is determined by several factors: (i) Salinity, or total concentration of dissolved salts (TDS, mg/liter or EC, dS/m). The EC increases by approximately 2% per degree C increase in temperature. For mixtures of salts in the range of 1 to 10 dS/m, TDS = 640 EC; osmotic pressure (atm) = 0.36 EC; and EC = 0.10 C,, where C, = total cation or anion concentration in mmol(+)/liter. (ii) Sodicity, which is expressed as the sodium adsorption ratio [SAR = (Na)/(Ca + Mg)&],where ion concentration is given in mM, or modified SAR, which accounts for carbonate and bicarbonate ion pairs with Ca in solution (Bresler et al., 1982). (iii) Anionic composition of the water, particularly concentrationsof chloride, bicarbonate, carbonates, OH, and toxic anions such as B and F. When recycled wastewater is used for irrigation, additional quality criteria are applied: (iv) Biochemical oxygen demand (BOD g/liter), which is the quantity of oxygen required for microbial degradation of organic compounds in the water at 20°C. This value is lower than the chemical oxygen demand (COD g/liter), the difference depending on organic matter characteristics. (v) Total suspended solids in the water (TSS g/liter).

1. Use of Marginal Water in Fertigation a. Effect on Soil Physical Properties Increased SAR and decreased EC of soil solutions cause soil particle swelling and dispersion. Swelling reduces the pore size in soils, dispersion causes particle movement and pore blockage, and the overall effect is reduced soil hydraulic conductivity, Kh (cm/min) (Bresler et al., 1982). Recommended threshold values of SAR and EC in irrigation water, to prevent reduction of soil Kh are SAR < 5 mi%@ for any EC > 0.5 dS/m; SAR < 10 for EC > 1, and SAR = 20 for EC > 2 dS/m (Feigin et al., 1991). Note that these guidelines were determined under sprinkler and surface irrigation, not under drip fertigation. Soil crusting, another phenomenon that reduces the water infiltration rate, is affected by irrigation water SAR and EC, and by the mechanical impact of irrigation water and raindrops. Another reason for the sensitivity of seal formation to rain is that the EC is lower at the soil surface than deeper in the soil. Principles of soil surface seal formation, runoff problems associated with it, and means of alleviating soil crusting by chemical and physical treatments are discussed by Sumner and Stewart (1992). Soil crusting may possibly be avoided by using subsurface drip irrigation. b. Salinity Effects on Plant Growth Growth suppression due to salinity starts at some threshold value that varies with crop tolerance, environmental conditions, and the size of the root system and intensifies as salinity increases, until the plant dehydrates. The salt tolerance of

ADVANCES IN FERTIGATION

37

crops is usually expressed in terms of relative yield ( Y J , threshold salinity (a), and percentage decrement value per unit increase of soil saturated extract EC, (b, dS/m): Y,. = lOO-b(EC,-a).

(15)

Data on crop tolerance to salinity according to (15) under different growth conditions are presented by Mass (1986). These data were obtained under furrow and flood irrigation. Sprinkler-irrigatedcrops are potentially prone to extra damage due to foliar salt intake and leaf burn. Drip-irrigated crops exhibit enhanced tolerance to saline irrigation water (see Section IID). Variations in crop sensitivity to salinity at different growth stages are known (Mizrahi et al., 1988) and should be taken into account when allocating water sources varying in quality to several crops grown in different fields. c. Toxicity of Particular Element to Plants Boron becomes toxic to plants at a concentration of a few milligrams per liter in soil solution, depending on the crop. The toxicity is described in terms of a threshold value and yield decrement slope parameters as in Eq. (15) (Mass, 1986). Lemon, cauliflower, and celery have threshold concentrations in saturated soil extract of <0.5,4.0, and 10 mg B/liter, respectively, and represent crops that are sensitive, moderately sensitive, and tolerant to B. For most crops the threshold B concentration is 0.75- 1.O mg/liter. Chloride is particularly toxic to woody species and the toxicity is expressed as foliar bum. Tolerance levels to chloride of various crops are given by Mass (1986). The maximum permissible C1- concentrations in soil solution at field capacity without leaf injury in strawberry, avocado, grapefruit, sweet orange, and Thompson seedless grape are 12, 13, 50, 20, and 40 mM C1, respectively. The corresponding threshold concentrations in saturation extract should be half of these values. Sodium toxicity is attributed to its entry into the cytoplasm and interference with K-activated enzymes that control vital metabolic processes and growth in plants (Marschner, 1995). High Na concentrations also induce Ca replacement in the cell membrane, thus reducing membrane integrity and causing ion leakage (Ben Hayimm et al., 1987) and diminished uptake. Feigin et al. (1991) presented data classifying crops according to their tolerance to exchangeable sodium percentage (ESP) in soil. Whereas for woody crops the threshold ESP is 2-10, for most field crops it is >20.

2. Recycled Municipal Effluents The main problem associated with using recycled municipal effluents for microfertigation is the higher EC and B concentration relative to freshwater. The

38

B. BAR-YOSEF Table VIII

Chemical Compositionof Secondary Municipal Effluents and Freshwater Used for Irrigation Constituent

TSS BOD EC PH Total N NO,-N NH,-N Total P B HCO,CI Na Ca SAR

K Mg

Unit

Municipal effluents"

Freshwater

mgAiter mg/liter dS/m

10-100 (30) 10-80 (20) 1.4-2.2 7.3-8.4 10-85 0-10 2-68 6-17 0-1 6-13 2-14 2.5-16 0.5-3.5 3.0-8.0 0.2-1 .o 0.5-2.7

-

mgiliter mgAiter mgAiter mgAiter mgAiter mM mM mM mM mMl/2

mM mM

0.g1.6 7.6-8.3

0-4 0-10

0.1-0.2 2.5-5.3 3.0-10.0 1.0-10.0 0.7-3.8 1.5-5.5 0.1-0.2 1.0-2.3

Source: Feigin et al. (1991). "Common permissible levels are in parentheses.

main reason for the higher EC is a 50 to 200% increase in HC0,- and C1- concentrations, balanced mostly by Na+ (Table VIII). From a fertigation point of view, the higher bicarbonate concentration increases the pH-buffering capacity of the irrigation solution and more acid is required to decrease the pH below 7, which is necessary to avoid Ca-P precipitation. It also decreases the free Ca2+ activity due to formation of CaHC03+ (log Kstability= 3.1). Nutrients found in treated municipal effluents must be taken into account in fertigation recommendations. Of practical importance are NH4+-N and mineral P ( 5 0 4 0 % of total P) usually found in secondary effluents (Table VIII). Organic N (= total N - inorganic N) and organic P are also utilized by crops, subject to their mineralization by soil microflora. Boron might be the factor limiting effluent use in crops sensitive to B toxicity (threshold < 1 mg B/liter). In clayey soils that adsorb B the risk of B toxicity is smaller than in sandy soils. Another problem in secondary effluents, particularly when used in swelling soils, is the SAR, which is higher than that in freshwater (Table VIII). Recent studies have shown that organic matter in effluents might have adverse effects on underground water quality. Graber et al. (1995), for example, have shown that the transport of atrazine in soil was enhanced under irrigation with treated sewage effluent as compared with irrigation with freshwater. The suggest-

ADVANCES IN FERTIGATION

39

ed mechanism was that the adsorbed herbicide in soil was released into the aqueous phase, particularly into the dissolved organic C load of the effluent, and thus became more mobile. Humic substances cause dispersion of Na-montmorillonite,particularly around pH 8 (Tarchitzky et al., 1993). Because these substances are found in the effluent organic matter content, the use of effluents for irrigation may cause dispersion and reduction in soil hydraulic conductivity in montmorillonitic soils. The accumulation of organic matter in the plant rhizosphere stimulates oxygen consumption by soil microorganisms and decreases 0, availability to plants. The potential problems associated with the presence of organic matter in effluents call for caution in the use of effluents for irrigation, particularly in microirrigation. Problems in microfertigation are aggravated as the recycled water is added to a relatively small soil volume, thus creating higher local organic matter concentrations than occur under overhead irrigation. Several field studies on microfertigation with effluents, in which the aforementionedreactions and procezses must have played a role, were reported by Feigin et al. (1991). No immediate damage to cotton and wheat plants, to the yield, or to the soil occurred in these studies. Other studies showing good yield and no damage to cotton, apple, and grape plants irrigated with municipal effluents were reported by Oron et al. (1982) and Neilsen et al. (1989a,b).

3. Recycled Greenhouse Solutions Principles of fertigation should be similar in greenhouses and in open fields. Greenhouses however, pose a specific problem, stemming from the requirement not to dispose of leachates into the environment. To cope with this restriction, greenhouse fertigation solutions have to be recycled, which raises several management problems. (i) Salinity buildup in the circulated solution. No information is available in the literature to indicate whether the salinity threshold values under such conditions are similar to those obtained under fixed salinity levels (see Section IIIG). Several management strategies are available to maintain the EC within the permitted range and to postpone the replacement of solution with fresh nutrient solutions. These strategies include starting with solutions of minimum Na and C1 ion concentrations (ion-exchange column or reverse osmosis treatment of tap water) and dilution with salt-free rainwater collected from greenhouse roofs. (ii) Soil-borne root pathogens move with the recycled leachates and increase the incidence of root diseases. To minimize infection, solutions must be disinfected before being reintroduced into the greenhouse. Known treatments include heating to 70°C,slow-flow sand filtration,ultraviolet irradiation, and nonphytotoxic chemical treatments such as ozone, hydrogen peroxide, chlorine dioxide, and bromine hypobromous acid. These methods and their suitability for various crops and growth substrates were reviewed by Bliss (1996). (iii) Pesticides, root exudates, and growth substrate dissolution products accumulate in the recycled solution.

40

B. BAR-YOSEF

Most root exudates are expected to be consumed by nonpathogenic microflora, which keep the dissolved organic C at a stable level. Of concern are a number of phytotoxic root exudates like ferulic acid and other phenolic acids that cause toxic symptoms at minute concentrations, as low as 1 pA4 (Sundin et al., 1996). Exuded carboxylates, particularly oxalate and citrate, stimulate dissolution of growth substrates like pumice, tuff, and zeolites. At low pH, resulting from proton release by roots, A13+ activity in recycled solution may reach levels toxic to plants. (iv) Greenhouse recycled solutions must be adjusted frequently to maintain the required nutrient and oxygen concentrations. Current methods to control NO,, NH,, K, and Ca concentrations are based on on-line autoanalyzer and specific electrodes. The latter methodology was evaluated recently by Morard (1996). He concluded that calibration and fragility problems necessitate further development before electrodes can be commercially used in greenhouses. The autoanalyzer can be easily adapted for monitoring purposes (including P) but at a high price. Robust on-line sensors to measure oxygen concentration in solution (usually by polarography) are commercially available.

-

N.MANAGING CROP FERTIGATION A. N, P, AND K OBJECTIVE CONSUMPTION FUNCTIONS Daily nutrient uptake rates that result in optimum yield and product quality [objective uptake curves, Q(t)]are crop specific and depend on climatic conditions, but are independent of soil characteristics and irrigation technique. Objective functions of N, P, and K consumption rates versus time for several crops grown under specified conditions, are presented in Tables IX, X, and XI, respectively. Considerable differences in uptake rate and in the time at which maximum consumption rate occurs exist among crops and among varieties of the same species (e.g., processing, greenhouse, and open-field tomatoes). In some cases, the consumption function is not monotonic and exhibits sharp changes at critical physiological stages. Ignoring temporal variations in uptake rate may lead to overfertilization and, consequently, to salinity buildup, reduced intake of other nutrients (Fried and Broeshart, 1967), and contamination of the environment. Suboptimal supply may result in depletion of nutrients from the soil and inadequate uptake rates. Most of the data in Tables IX, X, and XI were okained in the Mediterranean area. Similar objective curves were published for east coast United States growth conditions by the North Carolina Cooperative Extension Service (1991). Extrapolation of the N, P, and K uptake data presented to environmental conditions much different from those specified (e.g., different temperatures or light conditions) should be done carefully and treated only as a first approximation.

ADVANCES IN FERTIGATION

41

B. NU-~IUENT CONCENTRATIONS IN IRRIGATION AND SOILSOLUTIONS The suitability of given nutrient concentrations in the irrigation water can be evaluated if the aforementioned root parameters are known. Let us consider as an example a fresh tomato (cv. 650) crop grown in a sandy soil, which has reached a growth stage of 100 days after planting, at the beginning of January. According to Table IXa, the target N consumption rate (Q,) of this cultivar at this time is 2.7 kg N ha- day- Suppose that the evaporation from a class A pan at that time is 3 mm day-' and the crop coefficient is 1 ( = 3 0 m3 ha-' day-'). Supplying the N through the water yields a concentration (=Cw) of 6.4 mM N. The question is whether this concentration is appropriate, that is, whether it allows the plants to absorb the amount of N that was added to the soil. To answer this question we need to estimate the N concentration at the root surface (C,) and to compare it with K , (eq. [4]). In sandy soils, a Cr/Cbratio of 0.1 has been shown to be a sound approximation, in which case, C, is expected to be -0.64 mM N, which is similar to K , of tomato (Table IV) (note that Cw = Cb) Recalling that C, should preferably be between K , and 3K,, it can be concluded that the concentration of 6.4 mM N in the irrigation water can be safely used. A more detailed analysis of the suitability of this concentration would involve the following steps: (i) Calculate F , [Eq. (4)] according to F,,, K (Table IV, Ref. 3 ) and the above C, ( F , = 4.4 X lo-' gN cm-' root day-'); ;i) evaluate tomato plant root weight from data in Table VI, an assumed root dry matter content (-5%), and an estimated root radius (0.02 cm) (root length = 6.4 lo9 cm root ha-'); (iii) multiply FN by root length to obtain the rate of N uptake (= 2.8 kg N ha-' day-'). The excellent agreement between the calculated uptake and the target consumption rate supports the previous conclusion regarding the suitability of a concentration of 6.4 mM N in the irrigation water. If the root system is smaller than in the example, and the calculated uptake deviates from the objective uptake rate by more than a prescribed value (e.g., 25%), the N concentration in the irrigation water should be elevated to increase F,, but it should not surpass 40K,. Immediate action should be undertaken to increase the plant root length, otherwise the excess of N application rate over target uptake rate (Q,) will cause environmental damage. If plant parameters are unavailable, one should take care that Q(t)/(daily irrigation rate) does not exceed the salinity threshold of the crop (see Section IIIG). An alternative approach to the direct evaluation of optimal Cwis to use empirical functions relating Cwto uptake rates of whole plants. These functions are specific to soil, crop, plant age, and irrigation regime. Two examples of such functions, for tomato and pepper, respectively, are presented in Figs. 6 and 7. According to Fig. 6, CJNO,) supplied to tomato plants grown in a sandy soil and having a root system bounded by a soil cylinder of 30 cm radius and 60 cm depth should not ex-

'

'.

Table Ma Daily Nitrogen Consumption Rate (kg N ha-' day-') by Various Field Crops Grown under Drip Fertigation as a Function of Time after Emergence or Planting Days after emergence or planting

&

1-10 11-20 21-30 31 4 0 4 1-50 5 1-60 61-70 71-80 8 1-90 91-100 101-110 111-120 121-1 30 131-150 151-180 181-220 Total (kg N ha-')

Bell pepper Processing tomatoes

Greenhouse tomatoes

Fresh tomatoes

a

b

Eggplant

Potatoes

0.10 0.50 1.00 2.80 4.50 6.50 7.50 3.50 5.00 8.00 -

1.00 1.00 1.00 2.00 2.50 2.50 2.50 2.50 1.50 1.50 1.00

0.05 0.10 0.20 0.25 3.20 2.90 0.25 0.25 0.25 0.25 0.25 1.20 2.40 2.60 2.30 1.90 290

0.25 0.35 0.40 2.10 2.00 2.10 2.90 2.20 1.40 1.50 0.80 1.00

393

0.10 0.60 2.30 4.00 4.50 5.50 6.00 2.00 1.OO 4.00 1 .00 7.00 380

0.10 0.50 1S O 1.60 1.70 1.60 1.70 2.60 2.80 2.50 2.50 1.50

1.50 1.50 4.00 2.00 450

0.30 0.30 0.30 0.40 0.40 0.45 0.50 1.70 2.80 1.30 2.70 4.60 3.90 2.70 250

1

.oo

-

205

170

Cotton" 0.25 0.25 0.25 0.25 0.36 1.05 2.65 3.55 5.70 5.80 1.82 0.13 0.13 0.13 224

Variety Seeding/ planting date Harvesting date Plants ha-' Soil Marketable yield (tons ha-') Reference

VF M82- 1-2

F- 144 Daniela

675

Maor

Mar 27b

Sept 25"

Sept 18=

Aug 26b

Jul 18 50,000 Clayey

Selective 23,000 Sandy

Selective 12,000 Sandy

Selective 90,000 Sandy

160 Dafne (1984)

195 Bar-Yosef etal. (1992)

127 Bar-Yosef eral. (1982)

65 Sagiv etal. (1977)

"Grown under sprinkler inigation and broadcast fertilization. bSeeding. 'Planting.

Black Oval

Desuea

Acala 4-42

Jul 14b

Sept 10"

Feb 19b

Apr 26

Selective 100,000 Sandy

Selective 12,500 Sandy

Jul 1

75 51 Bar-Yosef Bar-Yosef eral. (1980~) etal. (1981)

-

57 Feigin and Sagivetal. (1982)

Oct 5 60,000

4 Halevy (1976)

Table IXb Daily Nitrogen Consumption Rate (kg N ha-’ day-’) by Various Field Crops Often Grown under Drip Fertigationas a Function of Time after Emergence or Planting Days after emergence or planting

i2

1-10 11-20 21-30 3140 41-50 51-60 61-70 71-80 81-90 91-100 101-110 111-120 121-130 131-140 141-150 151-160 161-170 Total (kg N ha)

Variety Seeding/planting date Harvesting date Plants ha-’ Soil Marketable yield (tons ha-’) Reference

“Seeding. bPlanting.

Chinese cabbage

Lettuce

Celery

0.15 0.45 3.00 3.40 2.20 1.80 -

0.17 0.2 1 0.70 0.88 1.03 0.99 0.99 0.83 0.83 1.00 1.47 1.78 2.00 2.25 -

0.74 1.11 1.85 2.96 2.24 2.70 1.08 0.84 0.37

-

111 Kazomi Nov 4” Jan 19 80,000 hamy 82 Sagiv et al. (1992)

-

-

110 Iceberg Nov 5“ Jan 25 100,000 Sandy 45 Bar-Yosef and Sagiv (1982b)

150 Florida Oct 10” Feb 27 90,000 Loamy 65 Feigin et al. (1976)

-

0.30 0.07 -

Broccoli

Sweet corn

carrot

Muskmelon

0.02 0.07 1.08 1.22 1.75 1.04 3.02 3.41 2.79 2.09 0.93 0.20 0.18 0.15 0.06 -

0.50 1.00 1.50 3.50 4.50 6.00 4.00 3.00 -

0.45 0.87 0.54 0.56 0.93 0.71 1.19 1.09 1.20 1.18 1.54 2.03 2.23 2.34 3.83 3.80 3.47 279

0.15 0.20 0.35 0.90 1.30 2.50 4.30 2.40 1.20 1.00 0.50 0.30

-

202 Woltam Aug 306 Jan 17 33,000 Loamy 13 Feigin and Sagiv (1971)

-

-

240 Jubilee Apr 15b July 5 75,000 hamy 28 Sagiv et al. (1983)

Buror Oct 116 Apr 5 400,000 Loamy 85 Sagiv et al. ( 1995)

-

-

151 Galia Jan 14b Selective 25,000 Sandy 56 Sagiv et al. (1980)

Table Xa Daily Phosphorus Consumption Rate (kg P ha-' day-') by Various Field Crops Grown under Drip Fertigation as a Function of Time after Emergence or Planting Days after emergence or planting ~~

2

~

Bell pepper

Processing tomatoes

Greenhouse tomatoes

Fresh tomatoes

a

b

Eggplant

Cotton"

~

1-10 11-20 21-30 31 4 0 41-50 5 1-60 61-70 71-80 8 1-90 91-100 101-1 10 1 11-120

121-130 131-1 50 151-180 181-220 Total (kgPha-I)

0.02 0.05 0.16 0.19 0.75 0.80 1.80 0.50 0.50 0.89

0.10 0.10

0.10 0.20 0.40 0.60 0.30 0.30 0.30 0.10 0.10 0.10 0.20 0.35 0.50 0.30 65 ~

"Grown under sprinkler irrigation and broadcast fertilization.

0.01 0.02 0.03 0.03 0.03 0.04 0.04 0.18 0.22

0.01 0.10 0.25 0.35 0.40 0.20 1.oo 0.20 0.50

0.10 0.30 0.60 0.45 0.17

0.50

24

0.20 0.10 0.10 0.30

42

0.01 0.10

0.01 0.01

0.01 0.04

0.10 0.20 0.25 0.35 0.45 0.35 0.35 0.35 0.25 0.25 0.10

0.01 0.01

0.05 0. I 0.1 0.3 0.4 0.9 0.5 0.55 0.85 0.4 0.075

-

31

0.02 0.08 0.09 0.05 0.05 0.05

0.09 0.15 0.27 0.3 1 0.38 0.35 33

Table Xb Daily Phosphorus Consumption Rate (kg P ha-' day-l) by Various Field Crops Often Grown under Drip Fertigation as a Function of Time after Emergence or Planting Days after emergence or planting 1-10 11-20

21-30 3140 41-50 5 1-60 6 1-70 71-80 81-90 91-100 101-110 111-120 121-130 131-140 141-150 151-160 161-170 Total (kg P ha)

Lettuce

Celery

0.01 0.10 0.50 0.60 0.55 0.45

0.03 0.04 0.11 0.08 0.20 0.23 0.35 0.29 0.39 0.17 0.18 0.30 0.54 0.69

36

Chinese cabbage

0.10 0.16 0.31 0.5 1 0.87 0.81 0.45 0.28 0.28

29

Broccoli 0.00

0.01 0.12 0.13 0.20 0.13 0.36 0.46 0.38 0.32 0.18 0.09 0.09 0.04 0.01

26

Sweet corn

carrot

Muskmelon

0.10 0.15 0.20 0.55 0.85 1.15 0.80 0.20

0.06 0.16 0.12 0.12 0.19 0.20 0.29 0.27 0.27 0.24 0.30 0.59 0.58 0.91 1.32 0.88 0.81 73

0.03 0.03 0.07 0.18 0.25 0.25 0.35 0.45 0.43 0.27 0.13 0.07

-

40

25

Table XIa Daily Potassium Consumption Rate (kg K ha-' day-') by Various Field Crops Grown under Drip Fertigation as a Function of T i e after Emergence or Planting Days after emergence or planting 1-10 11-20 21-30 3140 41-50 5 1-60 61-70 71-80 81-90 91-100 101-110 111-120 121-130 131-150 151-180 181-220 Total (kg K ha-')

Bell pepper Processing tomatoes 0.10 0.30 2.00 2.30 8.00 8.50 9.00 4.50 9.20 9.00

Greenhouse tomatoes

2.00 4.00 3.50 3.50 5.50 5.50 6.00 4.00 6.00 0.10 0.10 1.oo

-

520

1.00 1.30 3.80 3.00 710

"Grown under sprinkler irrigation and broadcast fertilization.

Fresh tomatoes

0.40 0.50 0.50 0.50 0.55 0.55 0.60 2.20 4.80 2.90 5.70 7.80 7.00 2.00

370

a

b

0.01 1.00 4.00 7.00 7.00 8.00 8.00 3.00 3.00 8.00 6.00 1.00 0.30 0.80

0.10 0.90 1.25 1.25 2.50 4.50 5.00 4.50 3.50 5.00 5.50 3.00

580

-

-

370

Eggplant

Cotton"

0.00 0.00 0.30 0.80 4.90 7.20 1.30 0.50 0.50 0.50 2.00 3.00 3.00 3.00 1.60 1.60 380

0.1 0.1 0.3 0.5 0.5 2.0 2.5 4.0 4.0 2.5 1.7 0 0 0

Table XIb Daily Pottasium ConsumptionRate (kg K ha-' day-') by Various Field Crops Often Grown under Drip Fertigation as a Function of Time after Emergence or Planting Days after emergence or planting 1-10 11-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90 91-100 101-110 111-120 121-130 131-140 141-150 151-160 161-170 Total (kg K ha)

Lettuce

Celery

0.20 0.50 5.10 7.80 8.20 3.20

0.21 0.24 1.33 1.52 2.56 2.78 4.11 4.05 5.56 4.04 5.00 8.60 8.50 10.35

-

-

1.70 2.80 4.50 7.20 5.25 5.52 1.37 0.01 -

-

-

-

-

-

-

-

250

-

Chinese cabbage

224

-

219

Broccoli 0.01 0.02 0.74 0.91 1.35 3.04 4.34 3.95 4.09 3.13 2.74 0.96 0.48

Sweet corn 1 .00 1.50 4.50 5.80 7.20 3.80 6.20 2.00 -

-

-

-

-

-

165

320

carrot 0.40 0.88 0.60

0.60 0.99 0.98 1.62 1.57 1.72 2.14 2.80 5.73 7.00 9.67 11.66 10.19 1.86 604

Muskmelon 0.10 0.25 0.60 1.45 3.00 6.00 7.00 8.00 7.50 3.50 1.00 0.05 -

385

49

ADVANCES IN FERTIGATION 140 DAYS AFTER SEEDING 0

8oo

r

70 DAYS AFTER SEEDING

A

N Concentration in soil solution (ppm) Figure 6 Mean daily nitrogen uptake determined from tomato plant analyses over the time intervals 60 to 73 and 138 to 165 days after seeding as a function of average NO,-N concentration in the solution of a soil cylinder bounded by a radius and depth of 30 and 60 cm, from the trickler, respectively, at specified times. The curves were hand fitted. Reproduced, with permission, from Bar-Yosef and Sagiv (1982a).

ceed 100 and 150 mg N liter-' at the ages of 70 and 140 days after seeding, respectively. These concentrations correspond to -2K,, and -3K,,,, (Table v),respectively. According to Fig. 7, the optimal Cw(N) of 76 to 96-day-old pepper plants grown in two sandy soils is -80 mg liter-'. The corresponding optimal Cw(K)of the same pepper is 100 mg K liter-'. The condition that C , X (daily irrigation rate) = Q(t)/EF must be fulfilled, also, when the empirical approach is used.

C. PREPLANTING BROADCASTFERTILIZATION AND BANDINGUNDER FERTIGATION For efficient preplanting fertilization one must know, for the particular soil and crop: (i) the relationship between nutrient application rate and nutrient availability level; and (ii) the response in terms of final yield to the nutrient availability lev-

50

B. BAR-YOSEF

-

-

-

1

1

1

1

1

1

A

1

1

1

I

-

I

76-96 days R;

0

c

r

0

0.80 104

0.75

A

1.02A

v

0.40

Horeva

c

n

=

Y

1

1

60

a-

-

K

41

1

I20

1

1

I80

I

I

I

,

j

240

z7*4-

40 80 120 160 200 Concentration in irrigation solution (mg /L

0

1

Figure 7 Rates of nitrogen and potassium uptake by pepper plants as a function of their concentration in the irrigation water 76 and 96 days after seeding. The results were obtained at two different sites, both with sandy soils. Ri is the ratio of seasonal overall irrigation to evaporation from a class A pan. Reproduced, with permission, from Bar-Yosef (1991).

el at seeding time. Discussion on preplanting fertilization as the only method of supplying nutrient to plants and a review of nutrient availability indexes can be found in Tucker and Hagin (1982). Under fertigation, preplanting fertilization is used to create the nutrient concentrations in the soil root volume that are required to allow growth and uptake during the initial growth stages, according to pertinent target uptake functions. To evaluate root depth during the first 2 weeks after emergence, a root elongation rate of 1.5 f 0.5 cm/day can be used (McMichael and Burke, 1996). For the small root volumes involved, banding seems to be more appropriate than broadcast fertilization for furnishing nutrients to young plants (BarYosef et al., 1995a). In the case of nitrogen, the preferred preplanting fertilizer is (NH,),SO,, which ensures minimal N leaching by technical emergence irrigations and rains. It also supplies sulfates to the soil that otherwise are seldom added via the water during the growing season. Under no-leaching conditions, urea (46% N) can be used as a base fertilizer. More details about the use of urea can be found in the next section.

ADVANCES IN FERTIGATION

51

Preplanting P fertilization has two aims: (i) to create the required initial P concentration in the soil solution within the seedling soil root volume (Gin); and (ii) to establish a sufficient overall P concentration in soil volumes where roots grow, but where P cannot be replenished via emitters. One example is P fertilizer incorporation in the 20 to 40-cm soil layer, which can be obtained by deep plowing following broadcast application. Another example is the replenishment of P to soil volumes between emitters that are outside the range of P transport from point sources. The application rate of either superphosphate (calcareous soils) or rock phosphate (acid soils) depends on the existing and target available P concentrations in the pertinent soil volumes, and the estimated P fixation rate in the soil. Objective i can be attained by banding soluble P (e.g., monopotassium phosphate, monoammonium phosphate, or ammonium polyphosphate) in the soil at seeding at a distance from the row and a depth of 10 cm. The combination of NH, and 0-P in preplanting fertilization is appropriate as NH, nitrification and uptake reduce the soil pH and delay P crystallization. Banding of either 0-P or polyphosphate in fertigated sweet corn resulted in similar yield and plant development (BarYosef et al., 1995a). To estimate the required quantity of P fertilizer to be banded, a dispersion radius of -10 cm around the band, and a target Cinof 1-2 mg P/liter can be assumed. From known soil P adsorption isotherm, the adsorbed P (A) in equilibrium with Cinis estimated (Section IIIF). Multiplying A by the soil weight in the 10cm-radius soil cylinder around the band gives the quantity of P to be banded (Q,). The initial quantity of adsorbed P should be subtracted from Qp. The short time between banding and plant emergence and the use of soluble P fertilizers justify the assumption that adsorption is the main mechanism determining Cinin this case.

-

D. CHOICE OF FERTILIZERS The data in Tables IX, X, and XI show the minimal application rates of N, P, and K that must be added to the soil via the water at any growth stage to satisfy plant demand and to maintain steady-state nutrient concentrations in the soil. The questions arise as to what is the recommended fertilizer to be used for this purpose and how various conditions in the system affect the decision regarding which fertilizer to choose. 1. Fertigation under Saline Conditions

According to the U.S. Salinity Laboratory (1954), irrigation water with EC exceeding 1.44 and 2.88 dS/m constitutes a moderate and a high salinization hazard, respectively. According to Tables IX,X, and XI, and assuming a daily irrigation of 5 mm, nitrogen and potassium concentrations in the irrigation water at the time

52

B. BAR-YOSEF

of maximum consumption rate may reach values of 15-20 mmol(+)/liter, which correspond to an EC of 1.5-2.0 dS/m. Under such conditions, and especially where the water EC > 1, which is common in arid zones, care should be taken to minimize the amount of accompanying ions added with the N or K. For example, KC1, which is a cheap source of K, should be replaced with KNO, and K,HPO,, while NH,NO, and urea should be preferred over (NH,),SO,. Chloride salinity is considered more toxic for the growth of most plants than isoosmotic concentrations of S0,2-(Marschner, 1995). Sodium-based fertilizers (e.g., NaNO, or NaH,PO,) are unacceptable sources because of the adverse effect of Na on soil hydraulic conductivity and plant functioning (see Section IIIG).

2. Fertigation Solution pH Different sources of N fertilizers have different effects on irrigation water and soil pH (Section IIIF). Alkaline pH in the irrigation water is undesirable, because Ca and Mg carbonate and 0-P may precipitate in the tubes and drippers. Also, high soil pH reduces Zn, Fe, and P availability to plants. Consequently, ammonia (fertilizer solution pH > 9) use in fertigation is not recommended, since it raises the pH when injected into irrigation water. Urea and ammonium nitrate liquid fertilizers have pH of 8.0 2 0.5 and 7 k 0.5, respectively. Urea increases soil pH upon hydrolysis; therefore, its application to soil together with superphosphate is undesirable. Compounds that may be used to reduce the irrigation solution pH are NHO,, H,PO,, H,SO,, and HCl. The last two acids are undesirable, because of their contribution to salinity. When the pH is depressed with acids, it should not be reduced below 5, at which soil CaCO, readily dissolves and Ca is leached outside the soil root volume. Higher acidity (pH < 4) is detrimental to root membranes and may increase the A1 and Mn concentrations in the soil solution to toxic levels by dissolution of clay minerals and metal oxides in the soil. The effect of the NH,/NO, ratio in irrigation water on soil pH, especially at the soil-root interface, was discussed above (Section IILF). For tomato and rose grown in tuff, a stable pH in growth substrate solution was maintained when the NH,/NO, molar ratio in the solution was between 1:4 and 1:3 (Feigin et al., 1979, 1986). Muskmelon grown in rockwool with NH, as the sole source of N decreased the leachate pH from -7 in the inflowing solution to -4 (Bar-Yosef et aE., 1995b). According to Ganmore-Neumann and Kafkafi (1980, 1983), NH,-N is an inappropriate source of nitrogen for tomato and strawberry plants at root zone temperatures >30°C, as it adversely affects carbon availability for root growth.

3. Nutrient Mobility in Soils Nitrogen spatial distribution in soil is strongly affected by the source of N added via the water. Ammonium is adsorbed by soil colloids and metal oxides and thus has a restricted mobility compared with the nonreactive NO,-. This means that

ADVANCES IN FERTIGATION

53

nitrification (see Section IVF) must be considered when evaluating N transport and distribution in the soil profile. To obtain rapid and uniform N distribution in the soil root volume, nitrate or urea should be used. Urea-the cheaper N source-is hydrolyzed within 2-3 days at 25°C (Black, 1968) and the ammonium obtained reduces the mobility until nitrification is completed. When choosing the P fertilizer for fertigation, care must be taken to avoid P-Ca and P-Mg precipitation in stock solutions, tubes, and emitters (Imas et al., 1996). From this standpoint, acidic P fertilizers (e.g., phosphoric acid and, to a lesser extent, monoammonium and monopotassium phosphate) are recommended. The use of polyphosphate in fertigation requires more knowledge than the aforementioned fertilizers. On the one hand, polyphosphate is capable of complexing Zn and Ca quite effectively but, on the other hand, it is susceptible to precipitation with Ca and it is adsorbed by soil similarly to 0-P, with considerable dependence on solution pH. In light of the variation in its properties with pH, the use of polyphosphate in fertigation systems is not recommended. More details on polyphosphate characteristics pertinent to fertigation can be found in Asher and Bar-Yosef (1982). The question of using controlled-release fertilizers in fertigation has not attracted much attention yet. At sufficiently high fertigation frequency, nutrient concentrations in the soil root volume can be maintained at the desired level with negligible time fluctuations. Under such conditions, the addition of controlled-release fertilizer to the soil would be superfluous. When the fertigation frequency is low (interval > 2 days), controlled-release fertilizer may attenuate time variations in concentration and reduce minor stresses stemming from transient plant starvation. The use of controlled-release fertilizers can be considered only if their rates of nutrient release under the prevailing environmental conditions are accurately known. The quantities released should be deduced from preplanned fertigation rates.

4. Ready-Mix Fertilizers The N, P, and K objective functions (Tables IX, X, and XI) may help to define the N:P:K weight ratio of a given compound fertilizer designed to supply those elements to a certain crop at a specific growth stage. The ingredients constituting a ready-mix fertilizer should be selected on the basis of the principles discussed in preceding sections and solubility characteristics. The supply tank from which the compound solution is injected into the mains should have a sufficiently large capacity to contain all the fertilizers needed for treating the entire service area without refilling. Mixing (mechanical or hydraulic) in the chemical supply tank is essential in order to avoid concentration gradients in the tank, which may cause variations in nutrient concentrations in the irrigation water over time. To choose a tank volume, VOL (m", one needs to know the fertilizer solubility in water, SOL (kg/m'), the fertilized service area, SA (ha), the amount of fertilizer to be supplied per application, Q, (kg ha- application-'), and the number of applications between successive refillings of the tank (n):

54

B. BAR-YOSEF

VOL = n Q, SA/SOL.

(16)

The solubility of common fertilizers in water at different temperatures is given in Table XII. Attention should be paid to the fact that solubility decreases considerably with temperature and it is, therefore, unsafe to leave concentrated fertilizer used in the summer for the winter period, since it may crystallize and block pipes connecting the tank and injection port. Incorporating microelements in stock fertilizer solutions and in fertigation water poses a problem, due to the low solubility of their hydroxides (e.g., Ksp of Fe(OH),, Zn(OH),, and Mn(OH), are 10-38.5,10-'5.5and 10- 12.8, respectively). To avoid precipitation at pH > 5 and to facilitate sufficient transport toward roots in soil, microelements are added in solution as chelates of organic ligands that have sufficient stability to avoid displacement by other cations and to prevent precipitation or adsorption by soils and growth substrates differing in chemical characteristics (Cadahia et al., 1988a). The main chelating agents used in fertigation systems are EDTA (ethylenediaminetetraacetic acid, C,,H,,08N2), DTPA (dieth-

Table Xn Solubility of Common Fertilizers in Water (kg fertilizer/m3) ~

~

~ _ _ _

Temperature range ("C) Fertlilizer Ammonium chloride Ammonium nitrate Monoammonium phosphate Diammonium phosphate Ammonium sulfate Potassium chloride Potassium nitrate Potassium sulfate Monopotassium phosphate Dipotassium phosphate Calcium nitrate Magnesium nitrate Monocalcium phosphate Phosphoric acid Urea

Formula

Cold 297 (0)" 1183 (0) 227 (0) 429 (0) 706 (0) 280 (0) 133 (0) 69 (0) -

Lukewarm

1950 (20) 282 (20) 575 (10) 760 (20) 347 (20) 316 (20) 110 (20) 330 (25)

Hot 758 (100) 3440 (50) 417 (50) 1060 (70) 850 (50) 430 (50) 860 (50) 170 (50) 835 (90)

1670 (20) 1020 (0)

-

780 (5)

Sources: Hodgman (1949) and Weast (1977). "Numbers in parentheses are solution temperatures, "C.

3410 (25) 423 (18) 18 (30) 5480 (25) 1193 (25)

3760 (99) 578 (90)

55

ADVANCES IN FERTIGATION Table XI11

Stability Constantsaof Several Common Macro- and Microelement Chelates and Complexes in Fertigation Systems and in Plant Rhizosphere Reaction

EDTA4-

DTPA4-

EDDHA4-

11.3 27.85 12.02 10.6 17.67 29.2 19.6 16.7 20.59

12.2 38.3 8.20 9.0 15.30 35.4 17.8

HCO;

Citrate3-

0xalate’-

log KO.,, H+L=HL 4H + L = H,L Ca2+ + L = CaL Mg2+ + L = MgL Fe2+ + L = FeL Fe3+ + L = FeL Zn2+ + L = ZnL Mn2+ + L = MnL A13+ + L = AIL

10.7 22.4 11.6 9.8 15.27 26.5 17.44 14.5

18.0

10.3 3.1

6.3 4.2

4.2

-

12.5 5.5 4.5 9.6

8.9 4.6 3.7 7.3

-

-

2.0

Sources: Lindsay (1979) and Norvel(l972). is defined as nM + L = M,L; K = (MnL)/[(M)” (L)].

ylenetriaminepentaaceticacid, C ,4H230 and EDDHA (ethylenediaminedio-hydroxyphenylaceticacid, C,,Hzo0,N2). Their stability constants with important cations are summarized in Table XIII. Also included in Table XI11 are stability constants of three anions normally exuded by roots: bicarbonate, oxalate, and citrate. In a well-aerated modified Hoagland solution (Ca = 1.5 mM, Mg = 0.8 mM, Fe-chelate = 0.1 mM, Zn = 1.5 cLM> FeEDTA becomes unstable above pH 6.5, and FeDTPA above pH 7.2, whereas FeEDDHA remains stable between pH 4 and 10 (Lindsay, 1979).This means that no cations in well-aerated hydroponic solutions are capable of displacing Fe3+ from FeEDDHA. The stability data show (Table XIII) that complexes of Fe with root-exuded carboxylic anions are considerably weaker than those with EDTA. Excessive chelate stability might be disadvantageous too, as the free cation activity in equilibrium with the chelate would be too low for effective uptake by plants. In such cases the plant must be able to reduce the Fe3+-chelate at the root surface and absorb the Fez+. Transferring the aforementioned Fe3+-chelate across the plasma membrane of root cells is very slow, even in crops having a specific transport system for phytosiderophores (Marschner, 1995). It is worthwhile to add to fertigation solutions a mixture of Fe”-DTPA (or EDTA) and Fe3+-EDDHA.When the latter chelate’s sites are saturated with Fe, Fe3+ activity in solution is controlled by DTPA, and Fe-EDDHA serves as a highly mobile Fe reserve that can be utilized in case the weaker chelate should lose its Fe3+ to stronger sinks in the soil.

56

B. BAR-YOSEF

The stability data for the Mn- and Zn-chelates (Table XIII) explain the fact that these elements cannot be effectively chelated in soils. Note that at pH < -3 the synthetic chelates (Table XIII) lose their capacity to retain microelements. This poses a problem in acid stock solutions, and it is advisable, therefore, to separate chelated microelements, keep them at pH > 5, and inject them into the water after the acid stock solution with the macroelements has been introduced.

E. TEMPERATURE EFFECTS Temperature affects all physical, chemical, and biological reactions in the soil-plant system. It is understandable, therefore, that the objective DM and uptake functions (Tables IX, X, and XI) are temperature specific. The effect of air temperature on evapotranspiration is taken into account in irrigation management considerations. A typical temperature profile of a wet soil at an air temperature of 25°C shows variations between 15 and -22°C at soil depths of 2 and 20 cm, respectively.After the same soil was dried for 25 days with heat lamps, the top soil temperature was -45"C, and at a depth of 20 cm it was 31°C. Below 50 cm the soil temperature approached 25°C in both the wet and dry soils (Hanks and Ashcroft, 1980). Such variations in soil temperature have a relatively small effect on chemical reactions. The diffusion coefficient in water (Do)is linearly related to absolute temperature T (Do = RTX,I[F,z,], where R is the gas constant, F is the Faraday, A is the equivalent conductivity, and z is the valence). The -log [ion activity coefficient],f, is positively related to T-3'2 (Debye-Huckel limiting law), so that changes in D, and f due to a 10°C increase in temperature (Q,,) are about 5%. Elevated temperatures increase ion adsorption by clay minerals and soils (Muljadi et al., 1966), with an estimated Q,, of 10-20%. The effect of T on fertilizer solubility was discussed in the preceding section. Meager information is available on the effect of root temperature on nutrient uptake rates. According to Marschner (1995) the Q,, of K + uptake by maize roots in the range 1O-2O0C, is about 300%, but further increase in T to 30°C resulted in only 10% increase in uptake rate. Between 30 and 35°C a decline in K' uptake rate was observed (Marschner, 1995). The Q,, for 0-P uptake rate by maize in the same temperature range was approximately one-half of the Q,, of K, and the decline in uptake rate occurred at 38°C. The reduced ion uptake at low temperature stems, according to Marschner (1993, from low membrane fluidity and a corresponding increase in membrane resistance to ion transfer. According to Hewitt (1966) the Q,, of nitrate absorption rate by maize seedlings in the temperature range 20-40°C is about 70%. The lack of data on temperature effects on Michaelis-Menten constants does not allow us to account quantitatively for the effect of variable soil temperature on nutrient uptake by fertigated crops.

-

-

ADVANCES IN FERTIGATION

57

F. O R G A N I C M A ~ R Mineralization of organic matter (OM) in soil must be taken into account in fertigation decision making as it contributes available nutrients to plants. The OM can be divided, for simplicity, into three pools: (i) OM added to soil (AOM); (ii) soil-indigenous OM (SOM); and (iii) OM found in the microbial biomass of the soil (BOM). The microbial biomass consumes or releases mineral N depending on whether the biomass is growing or decomposing, respectively. Biomass growth depends on the carbon availability and C/N ratio of the AOM. Net mineralization is equal to the overall mineralization minus the mineral N consumed (or released) by the microbial biomass (Jansson and Person, 1982). Comprehensive models of N transformations in soil (Molina et al., 1983) account for the three OM pools, but they also need to simulate carbon transformations and balance in the soil. The simulation of carbon transformations requires extra parameters that are hard to obtain, and it increases the complexity to the model. A simpler approach to estimating organic N mineralization, which seems to be appropriate for fertigation management, is presented in Fig. 8. This model is based on the assumption that N in the microbial biomass is steady with time, that is, that overall and net mineralizations are equal. Integration over time of the first-order mineralization rate equation (Fig. 8), and the condition -d(OM - N)/dt = d(minera1- N)/dt yield: [mineral - N],

=

OMol [ l

-

exp(-KoM,r)].

(17)

Here OMo, (=Ntotd,X foMI) is the initial SON or AON, Ntota,is total organic N concentration at t = 0 in the SON and AON pools,,foM, is the mineralizable N fraction, KoMl (=KOMoptgOMJis the rate constant, KOMoptis the rate constant under optimal soil 0 and T, and g is a correction function accounting for deviations of soil 8 and T from optimum values. The integrated nitrification equation has the same form, but OMo is replaced by the initial NH, concentration in the soil (fertilization) and the parameters are specific for this reaction. The correction term g (8, r ) is equal to g(0) X g(r). The g(T) function is an empirical expression that diminishes on both sides of the optimum T. The temperature at which mineralization and nitrification rate constants are maximal (g ( T ) = I ) is 30-35°C (Hadas et al., 1983; Tucker and Hagin, 1982). The correction term g(0) = [0 - 8thr)/(00ptOlhr)]m, where Othr and Oopt are 0 values for which g(8) = 0 and I , respectively, and m is a soil constant. Oop1 is 75-80% of the water-holding capacity of the soil (Legg and Meisinger, 1982). Another factor that strongly affects mineralization and nitrification kinetics is soil pH. The optimal range is 7.2-7.8; at pH 5 the activity of the oxidizing bacteria is significantly reduced. At the end of each year, the unmineralized AON is transferred into SON. Representative values of the abovementioned coefficients are given in Fig. 8. For demonstration, Eq. (17) was used to calculate the mineralization of SOM (400 mg organic N/kg soil), with dairy ma-

B. BAR-YOSEF

58

Mineralization of AON

~

+

MINERAL N POOL NH4

-

d.n/

NO3

Nitrification Mineralizatio of SON

1

Transition of

1

I

AON

Mineralization rate (i = AON, SON): d(ON,)/dt = -KOM,g,f, (ON,) Transition AON -SON: ON source

(AON,,,,

- AONm,&

. ”&. NO

SON Soil AON Dairy manure Poultry manure Dairy compost

f,

g/IOOg

l/d

Yo

0.02-0.05 1.5-2.5 3.0-4.0 2.5-3.0

0.006

15

0.2-0.3 30 1.5-2.5 55 0.0004 25

1 year time step

Reference Stanford&Smith 1972 Hadas et al. 1983 Hadas et al. 1983 Hadas&Portnoy 1994

d.n = denitrification Figure 8 Nitrogen transformations in soil that should be taken into consideration in fertigation management.The general first-orderequation is used for all transformations subject to pertinent parameter values (inset table). ON, = total organic N in pool i. Parameters are discussed in the text.

nure (AON) added to the soil at a rate of 8 tons/ha (2% N) in the top 0- to 20-cm soil layer (-2500 tons soil, yielding 64 mg N/kg) under conditions of optimum 0 and T. Thef,, and KOMof the SON and AON pools were assumed to be 15 and 30%, and 0.006 and 0.22 day-’, respectively. The cumulative mineralized N was compared with the objective QN ( t ) of processing tomatoes (Table IXa). During the first 20 days, the manure and indigenous organic N contributed 18 and 6 g mineral N m--2, respectively, while consumption by plants amounted to 2 g N m-2. At 65 days the corresponding cumulative mineralization and uptake figures were

ADVANCES IN FERTIGATION

59

20, 19, and 19 g N m-', and at 100 days 21,28, and 40 g N m-2. The excess of mineralization over consumption is detrimental to final yield and quality, and is prone to leaching outside the soil root volume. Another organic N pool-the plant residue incorporated into the soil-was disregarded in Fig. 8. Its mineralization rate is slower than that of AON, but since it is hard to evaluate amounts of plant residues added to soil this pool was disregarded in this review. In nitrification reactions f = 1 and Kopt = -0.1 day-' (Mengel and Kirkby, 1987). Denitrification may cause nitrate losses under fertigation, particularly after irrigation, when 8 > field capacity. However, as denitrification under irrigation has been estimated to be less than 5 - 10%of added N (Legg and Meisinger, 1982),this N loss is often disregarded in fertigation considerations.

G. GREENHOUSES The fertigation principles discussed so far are equally applicable to greenhouses and to open field. There are, however, some problems specific to greenhouses. (i) The partial control of temperature, light intensity, and CO, requires adaptation of the objective DM(t) and Q(t) functions to conditions prevailing in the greenhouse. (ii) Growth substrates in a greenhouse may be used at different volumes per plant. At small volume per plant, a certain fraction of the nutrients applied through the water to comply with the Q(t)function may be leached out of the substrate and become unavailable to plants. To control the leaching fraction under such conditions, fertigation scheduling, based on the water-retention characteristics of the substrate, becomes critical. The chemical reactions in small substrate volumes per plant are very intensive and may cause dissolution and release of toxic elements (e.g., A1 in tuff), and alter ion-retention characteristics and partitioning of nutrients between the solution and solid phases of the growth medium. Expected variations with time in substrate resistance to root growth may change the root distribution in the substrate and even the root morphology and uptake characteristics. (iii) In light of the difficulties in maintaining the appropriate supply rates of nutrients at the required concentrations in the water to small substrate volumes per plant, fertigation solutions in greenhouses are usually added in excess, resulting in large effluent volumes. To avoid underground pollution by effluent nitrates, and to save resources, greenhouse leachates should be recycled. Fertigation under recycling conditions, when salinity, pathogens, and substrate dissolution products accumulate in the solution, was discussed above (Section IIIG). Discussion of the physical and chemical characteristics of commercial growing media is outside the scope of this review. Data on the hydraulic properties, and ion adsorption and dissolution reactions of various substrates can be found, for example, in Adams et al. (1995).

60

B. BAR-YOSEF

V. MODELING FERTIGATION From a fertigation standpoint, modeling is important for the following reasons: (i) It may significantly improve management decision making, which is currently based on the simplifying assumption that water, nutrients, and roots are uniformly distributed in the wetted soil volume. Mathematical models that simulate distribution of water, ions, and roots in soil and calculate uptake from all soil subvolumes can improve water and nutrient uptake predictions and provide better estimates of solute leaching outside the soil root volume. (ii) Reliable crop-soil-atmosphere models can substitute for expensive field experiments in which site-specific, objective dry matter and nutrient consumption functions are determined. (iii) Models mentioned in ii, which stimulate crop growth, yield, and soil processes in response to given initial conditions, temporal water and nutrient supplies, and climatic conditions, may be transformed into management decision models. To serve as a management decision tool, the model must include a target yield versus time function. In case the target yield is not obtained under a projected fertigation regime and given climatic conditions, the simulation is reiterated for a modified regime, until the required yield function is obtained. This methodology does not conform with the economic-environmental-agronomic models discussed in Section 11, but it may significantly improve upon the simplified decision making process described above. Another group of models comprises auxiliary models that simulate specific processes, for example, optimizing the choice of fertilizers for ready-mix stock solutions, or scheduling irrigation according to real-time soil water and weather monitoring. A short review of available models in the above-mentioned categories follows.

A. MODELSSIMULATINGTRANSPORT AND UPTAKE PROCESSES Several mechanistic models that simulate transient two-dimensional nutrient and water transport in soil and uptake by plants have been published (Abbas et al., 1996;Timlin et al., 1996; Heinon et al., 1997; Lafolie et al., 1997). Of these, only the second simulates temporal root growth in relation to dynamic soil factors. None of the mentioned models treats N, P, and K concomitantly, but all of them include N transformations and concentration-dependent uptake by roots. None of the above models have been adjusted for drip fertigation, nor rigorously tested under diverse cultural conditions. The models do not account for root exudation and pH effects in soil (e.g., adsorption, mineralization), but have been reported to give a reasonable agreement between experimental and computed results, which encourages further evaluation of their performance.

ADVANCES IN FERTIGATION

61

Models focusing on subunits of the plant-soil-atmosphere system [for example, uptake by a single cylindrical root (Nye and Tinker, 1977; Bar-Yosef et al., 1980b); one-dimensional nutrient uptake under steady-state water content and predetermined root growth (Barber, 1984); and one-dimensional transient water and solute flow and nonmechanistic uptake (Wagenet and Hudson, 1987)l have been abundantly published during the past two decades, but are being used mainly in research.

B. CROPMODELS Few crop models have satisfactory routines that simulate processes occurring in soil and uptake by roots. One that may be applied to one-dimensional fertigation problems is CEREZ-Maize (Jones and Kiniry, 1986). In addition to plant processes and yield it simulates soil water and solute movements, N transformation in soil, and water and N uptake by the crop. Similar simulation capabilities are available in a potato model (Fishman and Bar-Yosef, 1995). Marani et al. (1992) adapted a cotton crop model (GOSSYM) to drip N fertigation including transport, subsurface emitter placement, and uptake by plants. The soybean simulator GLYCIN (Acock and Trent, 1991) has been coupled in a modular way with the two-dimensional soil simulator 2DSOIL (Timlin et al., 1996) to provide a potentially strong tool for fertigation simulation. It is hoped that more crop models (Rosenthal et al., 1989; van Keulen and Dayan, 1993; Sinclair and Muchow, 1995) will be coupled with this platform.

C. AUXILIARY MODELS Several simple models have been used for managing various aspects of fertigation. Models have been developed that determine when a critical soil moisture level has been reached, as a criterion for irrigation timing and rate (Fereres et al., 1981; Wu, 1995). A more comprehensive approach, which incorporates irrigation water price and fruit yield value, in addition to water and nutrient inputs and consumption balance, was suggested for fertigated orchards by Vera and de la Pena (1995). Despite the rough estimates of uptake and its relation to yield, this decision support system may be helpful in assessing fertigation needs under various growing conditions and economic environments. Recently, models that simulate crop growth and response to water have been used for irrigation scheduling (Hoogenboom et al., 1991). Breimer et al. (1988) presented a computerized fertigation program for greenhouse crops based on estimated evapotranspiration and nutrient consumption. An earlier program for greenhouse crops was presented by Oswiecimski (1984), with emphasis on the effects of growth media and container

62

B. BAR-YOSEF

volume on growth. A version of a balance-based fertigation model for surge irrigation was suggested by Ostermeier et al. (1992). Several computer programs that optimize the preparation of liquid compound fertilizers from available alternativeson the basis of price and plant preference are commercially available. Usually, such programs are available from big fertilizer distributorsand manufacturers. Technical considerations of how auxiliary models can be incorporated into automatic fertigation control systems are outside the scope of this review.

VI. MONITORING The principles discussed in the preceding sections allow one to fertigate during the growing season to sustain crop growth and uptake according to the objective functions, and thus to obtain an optimum yield. By monitoring plant organ dry weight and nutrient contents and comparing the results with the required, predetermined overall values (Tables IX, X, XI), one can determine whether the crop is developing and absorbing nutrients according to the objective functions. Any deviation between real and objective values exceeding a certain permitted error must elicit a correction measure, usually modification of nutrient and water application rates. The crop analysis should be done sufficiently early for the correction measure to be effective. Monthly crop monitoring seems to be adequate for characterizing crop development with respect to the reference target curves. The recommended entire plant analysis differs from the widely used leaf analysis. The latter is significantly cheaper, but it has the disadvantageof being an intensity factor (nutrient concentration in a diagnostic tissue) that cannot be translated directly into a correction measure. Another problem is that nutrient concentration in leaves are prone to daily fluctuations and reflect transient conditions in the plant environment. More practical experience is needed to compare the cost effectivenessof the entire plant analysis approach with that of the index leaf sampling procedure. The required nutrient concentrations in the soil solution that allow optimal uptake rates by plants were discussed in Section IVB. By means of soil tests, the deviation between prevailing and required concentrationsin the soil root volume can be determined, and measures to restore the needed concentrations can be undertaken. The soil water content must be maintained at a level that will not limit water and nutrient movement to the roots under any weather conditions and for any sink power of the crop. As stated above, discussion of the principles of irrigation management (rates and scheduling) and of optimization of soil 0 or water potential (4) is outside the scope of this review. For more information on microirrigation principles and control, readers are referred to reviews by Bucks et al. (1982) and Dasberg and Bresler (1985). To confirm that water status in soil is maintained

ADVANCES IN FERTIGATION

63

within the required range, soil 8 or $I must be monitored periodically (see Section IIID). Nutrient concentrations in the soil solution should also be periodically monitored to ensure that their deviations from the required optimum concentrations do not exceed a given permitted value. Soil monitoring can be done by two approaches: (i) soil sampling at one or more reference points in the soil root volume and extraction by standard solutions to determine soluble and sorbed nutrient concentrations in the soil samples; or (ii) direct sampling of the soil solution by means of vacuum cups inserted permanently in various locations in the soil, and chemical analysis of the solution samples for various nutrient concentrations. The vacuum extraction method is convenient, cheap, and only slightly alters the actual composition of nonadsorbing ions in the soil solution adjacent to the suction cup. Phosphorus and cations are adsorbed by ceramic cups and can be evaluated only by soil sampling. Commonly used vacuum cups and principles of their use are described in detail by Rhoades and Oster (1986). Soil extraction solutions are chosen in accordance with the ion and soil of interest: To obtain nutrient concentrations in the soil solution, saturated paste extract is recommended. For sorbed P in neutral and alkaline soils, the sodium bicarbonate extract is most suitable, while in acid soils the Bray extract is preferred. Exchangeable cations are determined by solution of a competing cation at high concentration, for example, 1 M ammonium acetate. Microelements in soil are determined by the chelating agent DTPA. Full accounts of soil extraction principles and methods are given by Page (1982). None of the above-mentioned extractions give the true nutrient concentration in the soil, but they determine a fraction of it that is closely related to its availability to plants. However, the extractable nutrient concentrationsare well correlated with the real concentrations in the soil and hence can show trends in nutrient status in sampled volume (depletion or accumulation) over time. Combined plant and soil monitoring is recommended, as the dual test can determine whether a certain deviation from the reference crop growth curve stemmed from under- or over supply of nutrients. When inhibited crop growth is not accompanied by suboptimal nutrient concentrations in the soil root volume, this indicates that other factors are limiting plant growth, for example, reduced light intensity or plant disease. In such cases, fertigation management should be modified to account for the reduced potential growth and development of the crop.

VII. SAFETY One problem with fertigation is potential contamination of the water source if proper antipollution devices are not in place. The most common possibilities for pollution are: (i) The injection system shuts off while imgation continues to oper-

64

B. BAR-YOSEF

ate. This could cause water to flow back through the chemical injection system and overflow the fertilizer supply tanks. (ii) The irrigation flow is shut off and the water-fertilizer mixture contained in the irrigation piping flows back into the irrigation water supply. (iii) As in ii, only the concentrated source fertigation solution flows into the irrigation water supply. Injection system safety devices and irrigation mainline backflow prevention equipment that are prerequisites for fertigation are described and discussed by Threadgill et al. (1990). According to this reference, regulations pertinent to chemigation (which includes fertigation) have been developed in several states in the United States in accordance with federal acts that apply to fertigation.

Vm. FUTURE TRENDS AND AREM NEEDING MORE RESEARCH Significant advances in microirrigation and fertilization equipment, automation, and understanding of basic processes have been made within the past three decades. Efficient utilization of available equipment is hampered by lack of data on optimum consumption rates of essential nutrients by important crops as functions of time. Additional data that are currently unavailable concern relationships between nutrient concentration and uptake flux and some basic soil parameters pertinent to ion transport in soil. The biological, chemical, and physical database presented in this review is still very limited, and simple extrapolation of the data to different climatic and soil conditions may lead to operational errors. The data presented should be regarded, however, as examples of the type of information needed to gain full benefit from advanced fertigation systems. Drip fertigation strongly affects plant root volumes. More research is needed to clarify soil physical and chemical effects on root growth, uptake, and excretion. An enhanced understanding of these phenomena will help us in using drip fertigation to produce desired root systems and thus to obtain plants that are more efficient in utilizing nutrients and water from the soil. It will also help us to design drip fertigation systems based on planning parameters that include root characteristics as well as soil hydraulic properties. Monitoring should be advanced on two fronts: (i) development of rapid and reliable methods to determine crop dry matter weight and nutrient contents, for comparison with the corresponding objective curves; and (ii) improving the methodology of determining nutrient concentrations in the soil solution, by allowing farmers to do it alone in the field, so that correction measures based on soil tests will be timely and effective. Available and specifically developed soil models should replace the currently crude calculations of nutrient uptake by roots, distribution in soil, and leaching

ADVANCES IN FERTIGATION

65

outside the soil root volume. A further research objective is to transform reliable soil-crop-atmosphere models into real-time fertigation management models, by introducing input optimization algorithms based on economic and environmental considerations. Achievement of this objective will depend on prior significant developments in fruit quality simulation in crop models, without which, crop yield value cannot be evaluated. Real-time management models should accept soil and crop monitoring data for continuous assessment of fertigation decisions. This poses another challenge for monitoring technology, as the sensing devices should be fully automated. More should be done to study the interrelationship between fertigation regimes and crop susceptibility to fungal and bacterial diseases (Jones et al., 1988).An understanding of the mechanisms involved in enhanced plant resistance to diseases as a function of nutrient status in the plant will add another dimension to fertigation optimization.

REFERENCES Abbas, C., Robert, J. L., and Parent, L. E. (1996). Mechanistic modeling of coupled ammonium and nitrate uptake by onions using the finite element method. Soil Sci. Soc. Am. J. 60, 1160-1 167. Acock, B., and Trent, A. (199 I). The soybean simulator, GLYCIN: Documentation for the modular version 91. Agric. Exp. Stn., University of Idaho, Moscow. Adams, P., Hidding, A. P., Kipp, J. A,, Sonneveld, C., and de Kreij, C. (Eds.) (1995). International symposium on growing media and plant nutrition in horticulture. Actu Hort. 401. Adin, A,, and Elimelech, M. (1989). Particle filtration for wastewater irrigation. J . Irrig. Drain. Div. ASCE 115,474-483. Adin, A., and Sacks, M. (199 I). Dripper clogging factors in wastewater irrigation. J. Irrig. Drainage Eng. 117,813-826. Aikman, D. P., Fenlon, J. S., and Cockshull, K. E. (1996). Anticipated cash value of photosynthates in the glasshouse tomato. Acta Hort. 417,47754. Albregts, E. E., Hochmuth, G. J., Chandler, C. K., Cornell, J., and Harrison, J. (1996).Potassium fertigation requirement of drip-irrigated strawberry. J. Am. Soc. Hort. Sci. 121, 164-168. Aha, A. K., and Mozaffari, M. (1995).Nitrate leaching in a deep sandy soil as influenced by dry broadcast or fertigation of nitrogen for citrus production. In “Dahlia Greidinger International Symposium on Fertigation,” pp. 67-77. Technion, Haifa, Israel. Aha, A. K., and Syvertsen, J. P. (1991). Irrigation water salinity affects soil nutrient distribution, root density, and leaf nutrient levels of citrus under drip fertigation. J. Plant Nu?,: 14,715-727. Andre, M., Massimino, D., and Daguenet, A. (1978). Daily pattern under the life cycle of a maize crop. 11. Mineral nutrition, root respiration and root excretion. Physiol. Plant. 44, 197-204. Asher, L. E., and Bar-Yosef, B. (1982). Effects of pyrophosphate, EDTA, and DTPA on zinc sorption by montmorillonite. Soil Sci. Soc. Am. J. 46,27 1-276. Assaf, R., Levin, I., and Bravdo, B. (1983).The response of apple trees to nitrogen fertilization regimes. Hassadeh 63,2586-2593. [in Hebrew] Awada, M., Wu, I. P., Suehisa, R. H., and Padgett, M. M. (1979).Effect of drip irrigation and nitrogen fertilization on vegetative growth, fruit yield, and mineral composition of the petioles and fruits of papaya, Technical Bulletin No. 103, p. 20. Hawaii Agric. Exp. Stn., University of Hawaii, Hilo.

66

B. BAR-YOSEF

Bachchhav, S. M. (1995). Fertigation in India-Astudy case. In “Dahlia Greidinger International Symposium on Fertigation,” pp. 11-24. Technion, Haifa, Israel. Bacon, P. E., and Davey, B. G. (1982). Nutrient availability under trickle imgation. I. Distribution of water and phosphorus in soil. Soil Sci. SOC.Am. J. 46,981-987. Bacon, P. E., and Davey, B. G. (1989). Nutrient availability under trickle irrigation-phosphate fertilization. Fertilizer Res. 19, 159-167. Bakker, M. J., Slangen, J. H. G., and Glas, W. (1984). Comparative investigation into the effect offertigation and of broadcast fertilization on the yield and nitrate content of lettuce (Lactuca sativa L.). Netherlands J. Agric. Sci. 32,330-333. Barber, S. A. (1984). “Soil Nutrient Bioavailability. A Mechanistic Approach.” Wiley, New York. Barrow, N. J. (1987). “Reactions with Variable-Charge Soils.’’ Nijhoff, Dordrecht. Bar-Tal, A., Bar-Yosef, B., and Kafkafi, U. (1990). Pepper transplant response to root volume and nutrition in the nursery. Agron. J. 82,989-995. Bar-Tal, A,, Pressman, E., and Rilska, E. (1994). Tomato response to K-Ca ratios and root restriction in solution culture. Annual Report, CALAR Tri-National (Egypt, United States, Israel) project. Bar-Yosef, B. (1971). Fluxes of P and Ca into intact corn roots and their dependence on solution concentration and root age. Plant Soil 35,589-600. Bar-Yosef, B. (1977). Trickle irrigation and fertilization of tomatoes in sand dunes: Water, N, and P distribution in the soil and uptake by plants. Agron. J. 69,486-491. Bar-Yosef, B. (1979). pH dependent Zn adsorption by soils. Soil Sci. SOC.Am. J. 43,1095-1099. Bar-Yosef, B. (1988). Control of tomato fruit yield and quality through fertigation. In “Optimal Yield Management” (D. Rimon, ed.), pp. 175-184. Avebury, Aldershot, UK. Bar-Yosef, B. (1991). Vegetable production and fruit yield and quality under combined trickle irrigation and fertilization. FAO/ECE Symposium on Methods and Concepts for the Use of Organic and Chemical Fertilizers, Geneva, R12, pp. 1-35. Bar-Yosef, B. (1996a). Greenhouse muskmelon response to K concentration in water and irrigation rate. Proc. ISOSC 9th Int. Congr. on Soilless Culture, pp. 35-50. Bar-Yosef, B. (1996b). Root excretions and their environmental effects-Influence on availability of phosphorus. In “Plant Roots-The Hidden Half (Y. Waisel, A. Eshel, and U. Katkafi, eds.), 2nd Ed., pp. 581-604. Dekker, New York. Bar-Yosef, B., and Imas, P. (1995). Tomato response to phosphorus nutrition via drip fertigation. Proc. 5th Int. Microirrigation Cong., Orlando, FL, pp. 200-208. Bar-Yosef, B., and Lambert, J. R. (1979). Corn and cotton root growth in response to osmotic potential and oxygen and nitrate concentration in nutrient solutions. In “The Soil-Root Interface” (J. L. Harley and R. S. Russell, eds.), pp. 287-299. Academic Press, New York. Bar-Yosef, B., and Lambert, J. R. (1981). Corn and cotton root growth in response to soil impedance and water potential. Soil Sci. SOC.Am. J. 45,930-935. Bar-Yosef, B., and Sagiv, B. (1982a). Response of tomato to N and water applied via a trickle irrigation system. I. Nitrogen. Agron. J. 74,633-637. Bar-Yosef, B., and Sagiv, B. (1982b). Trickle irrigation and fertilization of iceberg lettuce (Lacruca sativa), Proc. IXth Int. Plant Nutrition Colloq., Wanvick University, UK, pp. 33-38. Bar-Yosef, B., and Sagiv, B. (1985). Potassium supply to field crops under drip irrigation and fertilization. In “Proceedings of the K Symposium,” pp. 185-188. International Potash Institute, Pretoria. Bar-Yosef, B., and Sheikholslami, M. R. (1976). Distribution of water and ions in soils irrigated and fertilized from a trickle source. Soil Sci. SOC.Am. J. 40,575-582. Bar-Yosef, B., Stammers, C., and Sagiv, B. (1980a). Growth of trickle irrigated tomato as related to rooting volume and uptake of N and water. Agron. J. 72,815-822. Bar-Yosef, B., Fishman, S., and Talpaz, H. (1980b). Amodel of zinc movement to single roots in soils. Soil Sci. SOC.Am. J. 44,127 1- 1279.

ADVANCES IN FERTIGATION

67

Bar-Yosef, B., Kramer, S., and Ben Basat, S. (1980~).Trickle irrigation and fertilization in the Northern Arava Valley in Israel: Bell pepper (cv. Maor). Annual report. Agricultural Research Organization, Bet-Dagan, Israel. [In Hebrew] Bar-Yosef, B., Kramer, S., and Ben Basat, S. (1981). Trickle irrigation and fertilization in the Northern Arava Valley in Israel: Egg plant. Annual report. Agricultural Research Organization, Bet-Dagan, Israel. [In Hebrew] Bar-Yosef, B., Kramer, S., and Ben Basat, S. (1982). Trickle irrigation and fertilization in the northern Arava Valley in Israel: Fresh tomatoes (cv. 675). Annual report. Agricultural Research Organization, Bet-Dagan, Israel. [In Hebrew] Bar-Yosef, B., Schwartz, S., Markovita, T., Lucas, B., and Assaf, R. (1988a). Effect of root volume and nitrate solution concentration on growth, fruit yield and temporal N and water uptake rates by apple trees. Plant Soil 107,49-56. Bar-Yosef, B., Katkafi, U., Rosenberg, R., and Sposito, G. (1988b). Phosphorus adsorption by kaolinite and montmorillonite: I. Effect of time, ionic strength and pH. Soil Sci. SOC.Am. J. 52, 15801585. Bar-Yosef, B., Sagiv, B., and Markovitch, T. (1989). Sweet corn response to surface and subsurface trickle phosphorus fertigation. Agron. J. 81,443-447. Bar-Yosef, B., Phene, C. J., and Hutmacher, R. B. (1991). Plants response to subsurface trickle fertigation, BARD Project IS-I 116-86. Bet Dagan, Israel. Bar-Yosef, B., Matan, E., Levkovich, I., Assaf, A,, and Dayan, E. (1992). Response of greenhouse tomatoes (cv. F-144 and F-175) to irrigation and fertilization in the Besor area, Report on Project 30- 170-92. Agricultural Research Organization, Bet Dagan, Israel. [In Hebrew] Bar-Yosef, B., Sagiv, B., and Markovitch, T., and Levkovitch, I. (1995a). Phosphorus placement effects on sweet corn growth, uptake and yield. In “Dahlia Greidinger International Symposium on Fertigation,” pp. 14 I - 154. Technion, Haifa, Israel. Bar-Yosef, B., Keinan, M., Sternbaum, B., Levkovich, I., Markovich, T., Rosenberg, R., and Soriano, S. (1995b). Muskmelon (cv. Arava) response to N concentration in water, NH,:NO, ratio and greenhouse temperature, Report on Project 301-02 11-95.Agricultural Research Organization, Bet Dagan, Israel. [In Hebrew] Bauerle, W. L. (1975). The effect of fertilization and irrigation on yield and quality of greenhouse tomatoes. Research summary No. 82, pp. 13-15. Ohio Agricultural Research and Development Center, Wooster, Ohio. Beese, F., Horton, R., and Wierenga, P. J. (1982). Growth and yield response of chili pepper to trickle irrigation. Agron. J. 74,556-561. Bell, J. P., Dean, T. J., and Hodnett, M. G. (1987). Soil moisture measurement by an improved capacitance technique. Part 11. Field technique, evaluation and calibration. J. Hydrol. 93,79-90. Ben Asher, J., Gordon, J. M., Liner, A,, and Zarmi, Y. (1982). Nutrient uptake and supply to tomato plants in water culture system. Agron. J. 74,640-644. Ben Asher, J., Charach, A,, and Zemel, A. (1986). Infiltration and water extraction from trickle irrigation source: The effective hemisphere model. Soil Sci. SOC.Am. J. 50,882-887. Ben Hayyim, G., Kafkafi, U., and Ganmore-Neuman, R. (1987). The role of internal potassium in maintaining growth of cultured citrus cells on increasing NaCl and CaC1, concentration. Plunf Physi01. 85,434-439. Bernstein, L., and Francois, L. E. (1975). Effects of frequency of sprinkling with saline waters compared with daily drip irrigation. Agron. J. 67, 185-190. Bhella, H. S., and Wilcox, G. E. (1985). Nitrogen fertilization and muskmelon growth, yield and nutrition. Proc. 3rd Int. DriplTrickle Irrigation Cong., Fresno, CA, Vol. I, pp. 339-345. Bianchi, M. L., Burt, C. M., and Ruehr, T. A. (1985). Drip fertilization practices and soil permeability. Proc. 3rd Int. Drip/Trickle Irrigation Cong., Fresno, CA, Vol. I, pp. 357-364. Black, C. A. (1968). Soil-Plant Relationships. John Wiley & Sons, New York.

B. BAR-YOSEF Bliss, W. R. D. (1996).Control of water borne pathogens in hydrophonics using nylate. Proc. ISOSC 9th Int. Congr. Soilless Culture, pp. 505-513. Boman, B. J. (1 996).Fertigation versus conventional fertilization of flatwood grapefruit. Ferril. Res. 44,123-128. Boyle Engineering Corporation (1994).Demonstration of emerging irrigation technologies. Report submitted to the California Department of Water Conservation Office, Sacramento, CA, pp. 1-8. Bravdo, B. A., and Hepner, Y. (1987).Irrigation management and fertigation to optimize grape composition and vine performance. Acra Horr. 206,49-67. Bravdo, B., Salomon, E., Erner, Y., Shufman, E., and Oren, Y. (1992).Effect of drip and microsprinkler fertigation on citrus yield and quality. Proc. 7th Int. SOC.Citriculture, Acireale, Italy, Vol. 2, pp. 646-650. Breimer, T., Sonneveld, C., and Spaans, L. (1988).A computerized programme for fertigation of glasshouse crops. Acta Horr. 222,43-50. Bresler, E. (1977).Trickle-drip irrigation: Principles and application to soil-water management. Adv. Agron. 29,343-393. Bresler, E. (1978).Analysis of trickle irrigation with application to design problems. Irrig. Sci. 1,3-17. Bresler, E., McNeal, B. L., and Carter, D. L. (1982).“Saline and Sodic Soils.’’ Springer-Verlag, New York. Brun, R., and Morisot, A. (1996).Modeling the uptake of water and nitrate of rose plants in hydrophonic culture. Actu Horr. 417,55-63. Brun, R.,Paris, B., and Hammelin, I. (1993).Fertigation management of rose plants grown in greenhouse on rockwool. Adv. Horr. Sci. 7, 145-148. Bucks, D. A. (1995).Historical development in microirrigation. Proc. 5th Int. Microirrigation Cong., Orlando, FL, pp. 1-5. Bucks, D. A., Nakayama, F. S., and Warrick, A. W. (1982).Principles, practices and potentialities of trickle (drip) irrigation. Adv. Irrig. 1,220-298. Bussi, C., Huguet, J. G., and Defrance, H. (1991).Fertilization scheduling in peach orchards under trickle irrigation. J. Hort. Sci. 66,487-493. Bussi, C., Huguet, J., Besset, J., and Girard, T. (1994).Effect of nitrogen fertilization applied during trickle irrigation on the growth and fruit yield of peach. Eu,: J. Agron. 3,243-248. Cadahia, C., Garcia, P., Lucena, J. J., Sarro, M. I., and Garate, A. (1988a).Solubility of Fe, Mn, Cu and Zn fertilizers as a function of the substrate type in a drip irrigation system. Acra Horf. 222, 179- 186. Cadahia, C., Masaguer, A,, Garate, A,, and Sarro, M. J. (1988b).Nutrient solutions-substrates (rockwool and peat) interactions in drip-irrigation with highly saline waters. Acra Horr. 222, 173-178. Callesen, 0.(1991).Drip irrigation and fertigation of raspberry. Tidsskrifl Planteuvl95,411-415. Campbell, G. S . ( 1977).“An Introduction to Environmental Biophysics.” Springer-Verlag, New York. Carrijo, 0. A., Oliveira, C. A. S., Olitta, A. F. L., de Fontes, R. R., dos Reis, N. B. B., and Della Vecchia, P. T. (1983).Comparison between drip and furrow irrigation and NK fertilization in tomato (Lycopersicon esculentum Mill.). Hort. Brasileira 1,41-44. [In Portuguese] Cassagnes, P. (1988).Fertilization. Apple fertigation by micro-jet or trickle irrigation. Arboriculrure Fruitiere 406,39-46. Cassagnes, P., Bazaillas, M., and Fourcade, P. (1984).Leaf mineral composition of apple trees irrigated by microjets or by drip irrigation. Proc. VIth Int. Colloq. Optimization Plant Nutrition, Vol. 3, pp. 789-797. Chauhan, C. P. S., Sing, R. B., Minhas, P. S., Agnihotri, A. K., and Gupta, R. K. (1991).Response of wheat to irrigation with saline water varying in anionic constituents and phosphorus application. Agric. Water Manug. 20,223-23 1. Christensen, L.P., Peacock, W. L., and Bianchi, M. L. (1991).Potassium fertilization of Thompson Seedless grapevines using two fertilizer sources under drip irrigation. Am. J. Enol. Viric. 42, 227-232.

ADVANCES IN FERTIGATION

69

Clausnitzer, V., and Hopmans, J. W. (1994).Simultaneous modeling of transient three-dimensional root growth and soil water flow. Plant Soil 164,299-314. Clough, G. H., Locascio, S. J., and Olson, S. M. (1990).Yield of successively cropped polyethylenemulched vegetables as affected by irrigation method and fertilization management. J. Am. Soc. Hort. Sci. 115,884-887. Clough, G. H., Locascio, S. J., and Olson, S. M. (1992). Mineral concentration of yellow squash responds to irrigation method and fertilization management. J. Am. Soc. Hort. Sci. 117,725-729. Coltman, R. R. (1987). Yield and sap nitrate response of fresh market field tomatoes to simulated fertigation with nitrogen. J. Plant Nutz 10, 1699-1704. Crespo-Ruiz, M., Goyal, M. R., Baez, C. C., and Rivera, L. E. (1988). Nutrient uptake and growth characteristics of nitrogen fertigated sweet pepper under drip irrigation and plastic mulch. J. Agric. Univ. Puerto Rico 72,575-584. Dafne, 0. (1984). Nitrogen effect on dry matter production, mineral absorption, yield and quality of processing tomatoes. MSc thesis submitted to the Faculty of Agriculture, The Hebrew University of Jerusalem, Israel. [In Hebrew] Dalton, F. N., and Gardner, W. R. (1978). Temperature dependence of water uptake by plant roots. Agron. J. 70,404-406. Dalton, F. N., Herkelrath, W. N., Rawlins, D. S., and Rhoades, J. D. (1984). Time domain reflectometry: Simultaneous measurement of soil water content and electrical conductivity with a single probe. Science (Washington,DC) 224,989-990. Dasberg, S. (1987).Nitrogen fertilization in citrus orchards. Plunr Soil 100,l-9. Dasberg, S. (1995).Drip and spray irrigation of citrus oranges in Israel. Proc. 5th Int. Microirrigation Cong., Orlando, FL, pp. 28 1-287. Dasberg, S., and Bresler, E. (1985). Drip irrigation manual, Publ. No. 9. International Irrigation Information Center, Bet Dagan, Israel. Dasberg, S., Bilorai, H., and Erner, Y. (1983). Nitrogen fertigation of Shamouti oranges. Plant Soil 75, 41-51. Dasberg, S., Bar-Akiva, S., Spazisky, S., and Cohen, A. (1988).Fertigation versus broadcasting in an orange grove. Fertil. Res. 15, 147- 154. Delvert, P., and Bolding, P. J. (1988). Fertilization. Apple in the Netherlands. Fertigation and trickle irrigation of young orchards. Arboriculrure Fruitiere 406,65-73. [In French] Demalach, Y., Ben Asher, J., Sagiv, M., and Alert, A. ( I 996). Double emitter source (DES) for inigation experiments in salinity and fertilization. Agron. J. 88,987-990. Dencker, I., and Hansen, P. (1994). Flower initiation and node development of auxiliary buds, growth partitioning, and bleeding sap production of young apple trees as affected by fertigation. J. Horf. Sci. 69,869-876. Doerge, T. A,, and Thompson, T. L. (1995). Concurrent evaluation of agronomic, economic and environmental aspects of trickle irrigated vegetable production. In “Proceedings of the Dahlia Greidinger International Symposium on Fertigation,” pp. 263-274. Technion, Haifa, Israel. Dornai, D., Gerstl, Z., Chen, Y., and Mingelgrin, U. (199 I). Effect of trifluran on cotton growth in arid zone soils. Weed Sci. 31,375-384. Drews, M., Schondorf, I., and Krumbein,A. (1995).Nitrate, vitamineC,carotinandzuckergehaltvon kopfsalat im jahresverlauf beim anbau im gewchshaus (Lactuca sativa L.). Cartenbuuwiss 60,180-187. Enfield, C. G., Phan, T., Walters, D. M., and Ellis, R., Jr. (1981). Kinetic model for phosphate transport and transformation in calcareous soils: I. Kinetics of transformations. Soil Sci. Soc. Am. J. 45, 1059-1064. Erez, A,, Ran, Y., and Bar-Yosef, B. (1992).The effect of restricted root volume on the development, yield and dry matter partitioning of young fruiting peach trees. Acra Horf. 322, 199-214. Feigin, A. ( 1990) Fertilization of crops irrigated with saline water. In “Water, Soil and Crop Management Relating to the Use of Saline Water” (A. Kandiah, ed.), FA0 Publ. AGL/Misc/l6/90, pp. 137-151. FAO, Rome.

70

B. BAR-YOSEF

Feigin, A,, and Sagiv, B. (1971). Broccoli response to manure and N fertilization in a loessial soil in southern Israel, Report on Project 401- 10102. Agricultural Research Organization, Bet Dagan, Israel. [In Hebrew] Feigin. A,, and Sagiv, B. (1982). Dry matter and nitrogen accumulation in irrigated autumn and spring potatoes grown in arid zone. Proc. 9th Int. Plant Nutrition Colloq., Wanvick University, UK, Vol. 1, pp. 168-173. Feigin. A,, Sagiv, B., Aviram, H., and Zipilevich, Y. (1976). Celery response to manure and N fertilization in a loessial soil in southern Israel, Report on Project 301-046. Agricultural Research Organization, Bet Dagan, Israel. [In Hebrew] Feigin, A,, Zwibell, M., Rylski, I., Zamir, N., and Levav, N. (1979). The effect of ammonium/nitrate ratio in the nutrient solution on tomato yield and quality. Acra Hort. 98, 149-160. Feigin, A., Letey, J., and Jarrell, W. M. (1982). Celery response to type, amount and method of N-fertilizer application under drip irrigation. Agron. J. 74,971 -977. Feigin, A,, Ginzburg, C., Gilead, S., and Ackerman, A. (1986). Effect of NH,/NH, ratio in nutrient solution on growth and yield of greenhouse roses. Acra Horr. 189, 127-132. Feigin, A,, Shalhevet, Y., and Ravina, I. (1991). “Irrigation with Treated Sewage Effluent.” SpringerVerlag, New York. Fereres, E., Goldfine, R. E., Pruitt, W. 0..Henderson, D. W., and Hagan, R. M. (1981). The irrigation management program: A new approach to computer assisted irrigation scheduling. Proc. 4th Scheduling Conf. ASAE, Irrigation Scheduling for Water Conservation in the go’s, pp. 202-207. Fishman, S., and Bar-Yosef, B. (1995). Simulation of N uptake and partitioning in potato plants: Model description and sensitivity analysis. In “Potato Ecology and Modelling of Crops under Conditions Limiting Growth” (A. J. Haverkort and D. K. L. Mackerron, eds.), pp.147-166. Kluwer, London. Fishman, S., Talpaz, H., Dinar, M., Arazi, Y., Rozman, Y., and Varshavski, S. (1984). A phenomenological model of dry matter partitioning among plant organs for simulation of potato growth. Agric. Sysrems 14, 159-169. Flink, M., Pettersson, R., and Andren, 0. (1995). Growth dynamics of winter wheat in the field with daily fertilization and irrigation. J. Agron. Crop Sci. 174,239-252. Fried, M., and Broeshart, H. (1967). “The Soil-Plant System.” Academic Press, New York. Ganmore-Neumann, R., and Kafkafi, U.(1980). Root temperature and percentage NO,/NH, effect on tomato plant development. I. Morphology and growth. Agron. J. 72,758-761. Ganmore-Neumann, R., and Kafkafi, U. (1983). The effect of root temperature on NO,/NH, ratio on strawberry plants. I. Growth, flowering, and root development. Agron. J. 75,941-947. Gardner, W. H. (1986). Water content. In “Methods of Soil Analysis. Physical and Mineralogical Methods” (A. Klute, ed.), 2nd Ed., Vol. I, pp. 493-544. Am. SOC.Agron., Madison, WI. Gardner, W. R. (1960). Dynamic aspects of water availability to plants. Soil Sci. 89,63-73. Gastaldi, C. R., and Sutton, B. G. (1989). Optimizing nitrogen fertilization of vegetable crops by drip irrigation. Acra Hort. 247,217-221. Girardin, P. H., Trendel, R., Meyer, J. L., Birgaentzle, M., and Freyss, P. (1993). Effects of conventional and multiple N application by fertigation on maize grain yields and NO,-N residues. Optimization of plant nutrition. In “Proceedings of the 8th International Colloquium on Optimization Plant Nutrition, 1992, Lisbon,” pp. 411-415. Kluwer, Dordrecht. Glass, A. D. M., and Perky, J. E. (1980). Varietal differences in potassium uptake by barley. Potash Rev. 12, 1-9. Goldberg, D., Gornat, B., and Rimon, D. (1976). “Drip trrigation Principles, Design and Agricultural Practices.” Drip Irrigation Scientific, Kfar Shmariahu, Israel. Goyal, M. R., Crespo-Ruiz, M., and Rivera, L. E. (1988). Root distribution of nitrogen fertigated sweet pepper under drip irrigation. J. Agric. Univ. Puerfo Rico 72,5 1-55. Goyal, M. R., Guadalupa-Luna, R., de Hernandez, E. R., and Chao de Baez, C. (1989). Post harvest evaluation of nitrogen fertigated sweet pepper under drip. J. Agric. Univ. Puerfo Rico 73,109-1 14.

ADVANCES IN FERTIGATION

71

Graber, E. R., Gerstl, Z., Fischer, E., and Mingelgrin, U. (1995). Enhanced transport of atrazine under imgation with effluent. Soil Sci. SOC.Am. J. 59, 1513-1519. Guennelon, R., Habib, R., and de Cockborne, A. M. (1981). Aspects concerning the availability of N, Pand K following fertilization in drip irrigation of fruit trees. In “Seminars on Drip Irrigation and Similar Methods” (Commission des Communautes Europeennes, ed.). Univ. Bologna, Italy. [In French] Gustafson, C. D., Marsh, A. W., Branson, R. L., and Davis, S. (1974). Drip irrigation worldwide. Proc. 2nd Int. Drip Irrigation Cong., San Diego, CA, pp. 17-22. Hadas, A,, and Portnoy, R. (1994). Nitrogen and carbon mineralization rates of composted manures incubated in soil. J. Environ. Qual. 23, 1184-1189. Hadas, A., Bar-Yosef, B., Davidov, S., and Sofer, M. (1983). Effect of pelleting, temperature, and soil type on nitrogen release from poultry and dairy manures. Soil Sci. SOC.Am. J. 47, 1129- 1133. Halevy, J. (1976). Growth rate and nutrient uptake of two cotton cultivars under irrigation. Agron. J. 68,701-705. Hanks, R. J., and Ashcroft, G. L. (1980). “Applied Soil Physics.” Springer-Verlag, New York. Harbaugh, B. K., and Wilfret, G. J. (1980). Spray chrysanthemum production with controlled-release fertilizer and trickle irrigation. J. Am. Soc. Horf. Sci. 105,367-371. Harbaugh, B. K., Stanley, C. D., Price, J. F., and Jones, J. B. (1989). Irrigation and fertilization management of cut chrysanthemums. HortScience 24, 150. Hatfield, J. L., and Fuchs, M. (1990). Evapotranspiration models. In “Management of Farm Irrigation Systems’’ (G. J. Hoffman, T.A. Howell, and K. H. Solomon, eds.), ASAE Monograph, pp. 33-59. American Society of Agricultural Engineers, St. Joseph, MI. Haynes, R. J. (1988). Comparison of fertigation with broadcast application of urea-N on levels of available soil nutrients and on growth and yield of trickle-imgated peppers. Sci. Horf. 35, 189-198. Haynes, R. J., and Swift, R. S. (1987). Effect of trickle fertigation with three forms of nitrogen on soil pH, levels of extractable nutrients below the emitters and plant growth. PlanfSoil 102,211-221. Hedge, D. M., and Srinivas, K. (1991). Growth, yield, nutrient uptake and water use of banana crops under drip and basin irrigation with N and K fertilization. Tropical Agric. 68,331-334. Heinen, M. (1997). Dynamics of water and nutrients in closed, recirculating cropping systems in glasshouse horticulture. PhD thesis, Wageningen Agriculture University, Wageningen. Hewitt, E. J. (1966). “Sand and Water Culture Methods Used in the Study of Plant Nutrition,” 2nd Ed. Commonwealth Agriculture Bureaux, Kent, UK. Hillel, D. (1980). “Fundamentals of Soil Physics.” Academic Press, New York. Hochmuth, G . J., Albregts, E. E., Chandler, C. C., Cornell, J., and Harrison, J. (1996). Nitrogen fertigation requirement of drip-irrigated strawberries. J. Am. SOC. Horf. Sci. 121,660-665. Hodgman, C. (1949). “Handbook of Chemistry and Physics.” CRC Press, Cleveland, OH. Hoogenboom, G., and Huck, M. G. (1986). ROOTSIMU V4.0. A dynamic simulation of root growth, water uptake, and biomass partitioning in a soil-plant-atmosphere continuum, update and documentation, Agronomy and Soils Department Series No. 109. Auburn University, Auburn, AL. Hoogenboom, G., Hines, J. W., and Boote, K. J. (1991). A decision support system for prediction of crop yield, evapotranspiration, and irrigation management. Irrigation and drainage Proc. ASAE, pp. 198-204. h a s , P., and Bar-Yosef, B. (1998). Response of lettuce plants grown on different substrates to phosphorus fertigation. Acfa Horf. 458, 171-178. Imas, P., Bar-Yosef, B., Levkovich, I., and Keinan, M. (1996). Orthophosphate solubility in waters of different ionic composition. Ferfil. Res. 44,73-78. h a s , P., Bar-Yosef, B., Kakafi, U., and Ganmore-Neumann, R. (1997a). Release of carboxylic anions and protons by tomato roots in response to ammonium nitrate ratio and pH in nutrient solution. Planf Soil 191,27-34. h a s , P.,Bar-Yosef, B., Katkafi, U., and Ganmore-Neumann, P. (1997b). Phosphate induced carboxylate and proton release by tomato roots. Planr Soil 191, 35-39.

72

B. BAR-YOSEF

Ingram, K. T., and Hilton, H. W. (1986). Nitrogen-potassium fertilization and soil moisture effects on growth and development of drip-irrigated sugarcane. Crop Sci. 26, 1034- 1039. Inskeep, W. P., and Bloom, P. R. (1986). Kinetics of calcite precipitation in the presence of water-soluble organic ligands. Soil Sci. Soc. Am. J. 50, 1167-1 172. Itai, C., and Brinbaum, H. (1996). Synthesis of plant growth regulators by roots. In “Plant Roots-The Hidden Half” (Y. Waisel, A. Eshel, and U. Kafkafi, eds.), 2nd Ed., pp. 273-284. Dekker, New York. Jansson, S. L., and Persson, J. (1982). Mineralization and immobilization of soil nitrogen. Agronomy 22,229-252. Jaworski, C. A., Kays, S. J., and Smittle, D. A. (1978). Effects of nitrogen and potassium fertilization in trickle irrigation on yield of pepper and polebean. HortScience 13,477-478. Jones, C . A,, and Kiniry, J. R. (eds.) (1986). “CERES-Maize: A Simulation Model of Maize Growth and Development.” Texas A&M Univ. Press, College Station. Jones, J. B., Stanley, C. D., Csizinszky, A. A,, Kovach, S. P., and McGuire, R. G. (1988). K and N fertilization rates influence susceptibility of trickle irrigated tomato plants to bacterial spot. HorfScience23, 1013-1015. Jungk, A. 0. (1996). Dynamics of nutrient movement at the soil-root interface. In “Plant Roots-The Hidden Half” (Y. Waisel, A. Eshel, and U. Kafkafi, eds.), 2nd Ed., pp. 581-604. Dekker, New York. Kafkafi, U. (1984). Plant nutrition under saline conditions. In “Soil Salinity under Irrigation’’(I. Shainberg and J. Shalhevet, eds.), pp. 319-338. Springer-Verlag. Berlin. Kafkafi, U., and Bar-Yosef, B. (1980). Trickle irrigation and fertilization of tomatoes in highly calcareous soils. Agron. J. 72,893-897. Karlen, D. L., Camp, C. R., and Robbins, L. (1985). Fresh market tomato response to N and K fertilization and water management practices. Commun. Soil Sci. Plunr Null: 16,71-8 1. Katou, H., Clothier, B. E., and Green, S. R. (1996). Anion transport involving competitive adsorption during transient water flow in an Andisol. Soil Sci. SOC. Am. J. 60, 1368-1375. Keng, J. C. W., Scott, T. W., and Lugo-Lopez, M. A. (1979). Fertilizer management with drip irrigation in an oxisol. Agron. J . 71,971-980. Keshavaiah, K. V., and Kumaraswamy, A. S. (1993). Fertigation and water use in potato under furrow and drip irrigation. J. Ind. Potato Assoc. 20,240-244. Kjelgren, R., Goldhamer, D. A., Uriu, K., and Weinbaum, S. A. (1985). Almond tree response to variable nitrogen fertilization rates through low volume sprinklers. Proc. 3rd Int. DriplTrickle Imgation Cong., Fresno, CA, Vol. I, pp. 377-381. Klein, I., and Spieler, G. (1987). Fertigation of apples with nitrate and ammonium nitrogen under drip irrigation. I. Tree performance. Commun. Soil Sci. Plant Anal. 18,311-322. Koo, R. C. J., and Smajstrla. A. 0. (1984). Effect of trickle irrigation and fertigation on fruit production and juice quality of Valencia orange. Proc. Florida State Hort. SOC.,Vol. 97, pp. 8-10, Kraffczyk, I., Trolldenier, G., and Beringer, H. (1984). Soluble root exudates of maize: Influence of potassium supply and rhizosphere microorganisms. Soil Biol. Biochem. 16,3 15-322. Lafolie, F., Guennelon, R., and van Genuchten, M. Th. ( I 989). Analysis of water flow under trickle irrigation: 11. Experimental evaluation. Soil Sci. SOC.Am. J. 53, 1318-1323. Lafolie, F., Bruckler, L., de Cockbome, A. M., and Laboucarie, C. (1997). Modeling the water transport and nitrogen dynamics in irrigated salad crops. Irrig. Sci. 17,95-104. Lahav. E., and Kalmar, D. (1988). Response of banana to drip irrigation, water amounts and fertilization regimes. Commun. Soil Sci. Plant Anal. 19,25-46. Lavon, R., Emer, Y., Shapziski, S., and Mohel, E. (1995). The effect of K fertigation with different N forms on the yield and fruit size of “Shamouti” oranges. In “Proceedings of the Dahlia Greidinger International Symposium on Fertigation,” pp. 35-77. Technion, Haifa, Israel. Legaz, F., Ibanez, R.,Barreda, D. G. de, and Primo Millo, E. (1983). Influence of irrigation and fertil-

ADVANCES IN FERTIGATION

73

ization on productivity of the “Navelate” sweet orange. Proc. Int. SOC.Citriculture, 1981, Vol. 2, pp. 591-595. Legg, J. O., and Meisinger, J. J. (1982). Soil nitrogen budget. Agronomy 22,503-566. Lindsay, W. L. (1979). “Chemical Equilibria in Soils.” Wiley-Interscience, New York. Locasio, S. J . , Myers, J. M., and Martin, F. G. ( I 977). Frequency and rate of fertilization with trickle irrigation for strawberries. J. Am. Soc. Hort. Sci. 102,456-458. Marani. A., Cardon, G., and Phene, C. J. (1992). CALGOS, a version of GOSSYM adapted for imgated cotton. 1. Drip irrigation, soil water transport and root growth. Proc. Beltwide Cotton Conf., Nashville, TN, Vol. 3, pp. 1352-1357. Marschner, H. (1995). “Mineral Nutrition of Higher Plants,” 2nd Ed. Academic Press, New York. Martinez Hernandez, J. J., Bar-Yosef, B., and Kafkafi, U. (1991). Effect of surface and subsurface drip fertigation on sweet corn rooting, uptake, dry matter production and yield. Irrig. Sci. 12, 153-159. Mass, E. V. (1986). Salt tolerance of plants. Appl. Agric. Res. 1, 12-26. McMichael, B. L., and Burke, J. J. (1996). Temperature effects on root growth. In “Plant Roots-The Hidden Half” (Y.Waisel, A. Eshel, and U. Kafkafi, eds.), 2nd Ed., pp. 383-396. Dekker, New York. McPharlin, I. R., Aylmore, P. M., and Jeffery, R. C. (1995). Nitrogen requirements of lettuce under sprinkler irrigation and trickle fertigation on a Speanvood sand. J. Plant Nurr. 18,219-241. Meiri, A,, and Plaut, 2. (1985). Crop production and management under saline conditions. Plurzl Soil 89,253 -27 I . Mengel. K., and Kirkby, E. A. ( I 987). “Principles of Plant Nutrition,” 4th Ed. International Potash Institute, Bern, Switzerland. Mikkelsen, R. L., and Jarrel, W. M. (1987). Application of urea phosphate and urea sulfate to drip-irrigated tomatoes grown in calcareous soil. Soil Sci. Soc. Am. J . 51,464-468. Miller, R., Rolston, D. E., Rauschkolb, R. S., and Wolfe, D. W. (1976). Drip application of nitrogen is efficient. Cul(jbrnin Agric. 30, 16-18. Mizrahi, Y., Taleisnik, E., Kagan-Zur, V., Zohar, Y., Offenbach, R., Matan, E., and Golan, R. (1988). A saline irrigation regime for improving tomato fruit quality without reducing yield. J. Am. Suc. Hort. Sci. 113,202-205. Molina, J. A. E., Clapp, C. E., Shaffer, M. J., Chichester, F. W., and Larson, W. E. (1983). NCSOIL, a model of nitrogen and carbon transformations in soils: Description, calibration and behavior. Soil Sci. Soc. Am. J. 47,85-91. Morard, P. (1996). Possible use of selective electrodes for nutrient solutions in recirculated systems. Proc. ISOSC 9th Int. Congr. Soilless Culture, pp. 291 -298. Mualern, Y.,and Friedman, S. P. (1991). Theoretical prediction of electrical conductivity in saturated and unsaturated soil. Water Resour. Res. 27,277 1-2177. Muljadi, D., Posner, A. M., and Quirk, J. P. (1966). The mechanism of phosphate adsorption by kaolinite, gibbsite and pseudoboehmite. The effect of temperature on adsorption. J. Soil Sci. 17, 212-247. Nakayama, F. S., and Bucks, D. A. (eds.) (1986). “Trickle Irrigation for Crop Production,” p. 383. Elsevier, Amsterdam. Neary, P. E., Storlie, C. A,, and Paterson, J . W. (1995). Fertilization requirement for drip-irrigated bell peppers grown on loamy sand soils. Proc. 5th Int. Microirrigation Congr., Orlando, FL, pp. 187-193. Neilsen, G. H., Stevenson, D. S., Fitzpatrick, J. J., and Brownlee, C. H. (l989a). Nutrition and yield of young apple trees irrigated with municipal waste water. J. Am. Soc. Hort. Sci. 114,377-383. Neilsen, G. H., Stevenson, D. S., and Fitzpatrick, J . J. (1989b). The effect of municipal wastewater irrigation and rate of N fertilization on petiole composition, yield and quality of Okanagan Riesling grapes. Cun.J. Plant. Sci. 69, 1285- 1294. [In French]

74

B. BAR-YOSEF

Neilsen, G. H., Parchomchuk, P., Hogue, E. J., Wolk, W. D., and Lau, 0. L. (1994). Response of apple trees to fertigation-induced soil acidification. Can. J. Plant Sci. 74,347-351. Neilsen, G. H., Parchomchuk, P., and Berard, R. (1995). NP fertigation and irrigation affect potassium nutrition of newly planted apple trees. Acta Hort. 383,57-65. Nissen, P. (1996). Uptake mechanisms. In “Plant Roots-The Hidden Half” (Y. Waisel, A. Eshel, and U. Kafkafi, eds.), 2nd Ed., pp. 51 1-527. Dekker, New York. Noborio, K., McInnes, K. J., and Heilman, J. L. (1996). TWO-dimensional model for water, heat and solute transport in furrow-irrigated soil: II. Field evaluation. Soil Sci. SOC.A m J. 60,1010-1021. North Carolina Cooperative Extension Service (1991). Drip fertigation systems, Leaflet No. 33-D. Norvell, W. A. (1972). Equilibria of metal chelates in soil solution. In “Micronutrients in Agriculture” (J. J. Mortvedt, P. M. Giordano, and W. L. Lindsay, eds.), pp. 115-138. Soil Sci. SOC.Am., Madison, WI. Nye, €? H., andTinker, P. B. (1977). “Solute Movement in the Soil-Root System.”Blackwell Sci., Oxford. Olsen, S. R., and Kemper, W. D. (1968). Movement of nutrients to plant roots. Adv. Agron. 20,91-15 1. Oron, G., Ben-Asher, J., and DeMalach, J. (1982). Effluent in trickle irrigation of cotton in arid zones. J. Irrig. Drain. Div. Am. SOC.Civil Eng. 108, 115-126. Orphanos, P. I., and Eliades, G. (1992). Nitrogen fertigation of Valencia orange irrigated by drip or minisprinkler. Acta Hort. 365, 105-120. Ostermeier, K.A., Watts, D. G., Boldt, A. L., Eisenhauer, D. E., and Schepers, J. S. (1992). Verification of fertigation model for surge irrigation. Am. SOC.Ag. Eng. (No. 92-2519), 29. Oswiecimski, W. (1984). Effect of programmed fertilization on the growth of tomatoes cultivated in different substrates and size of containers. Acta Hort. 145,59-65. Page, A. L. (ed.) (1982). “Methods of Soil Analysis. Part 2. Chemical and Microbiological Properties,” 2nd Ed., Agronomy No. 9. Am. SOC.Agron., Madison, WI. Papadopoulos, I. (1987). Nitrogen fertigation of greenhouse-grown tomato. Commun. Soil Sci. Plant Anal. 18,897-907. Papadopoulos, I. (1988). Nitrogen fertigation of trickle-irrigated potato. Fertil. Res. 16, 157-167. Papadopoulos, I. (1992). Phosphorus fertigation of trickle-irrigated potato. Fertil. Res. 31,9-13. Phene, C. J., and Howell, T. A. (1984). Soil sensor control of high frequency irrigation. Trans. ASAE 27,392-396. Phene, C. J., Rose, J. L., Chavez, R. L., and Robb, D. J. (1982). Subsurface trickle irrigation and fertilization of processing tomatoes. In “DriplTrickle Irrigation. Winter 1982,” pp. 5-7. Agribusiness, Fresno, CA. Phene,C. J., Bar-Yosef, B., Hutmacher, R. B., Patton, S. H., Davis, K. R., andMcCormick, R. L. (1986). Fertilization of high yielding subsurface trickle irrigated tomatoes. Proc. 34th Annu. Calif. Fertilizer Conf., Fresno, CA, California Fertilizer Assoc., pp. 33-43. Phene, C. J., Hutmacher, R. B., Davis, K. R., and McCormick, R. L. (1990a). Water-fertilizer management of processing tomatoes. Act Hort. 277, 137-143. Phene, C. J., Reginato, R. J., Itier, B., and Tanner, B. R. (1990b). Sensing irrigation needs. In “Management of Farm Irrigation Systems (J. H. Hoffman, T. A. Howell, and K. H. Solomon, ed~.),pp. 207-261. Am. SOC.Agric. Eng., St. Joseph, MI. Phene, C. J., Davis, K. R., Hutmacher, R. B., Bar-Yosef, B., Meek, D. W., and Misaki, J. (1991). Effect of high frequency surface and subsurface drip irrigation on root distribution of sweet corn. Irrig. Sci. 12, 135-140. Pier, J. W., and Doerge, T.A. (1995). Nitrogen and water interactions in trickle-imgated watermelon. Soil Sci. SOC.Am. J. 59, 145-150. Plaut, Z., and Meiri, A. (1988). Effect of salinity in irrigation water on yield and fruit quality of greenhouse grown tomatoes during the winter. Israel Agric. Res. 2,79-97. [In Hebrew] Plaut, Z., Carmi, A., and Grava, A. (1988). Cotton growth and production under drip-irrigation restricted soil wetting. Irrig. Sci. 9, 143-156.

ADVANCES IN FERTIGATION

75

Pomares, F., Estela, M., and Tarazona, F. (1992). Nitrogen fertilization of strawberries in Valencia planted in the autumn with drip and furrow irrigation. Invest. Agraria Production Protection Vegetables 7 , 195-207. [In Spanish] Rauschkolb, R. S., Rolston, D. E., Miller, R. J., Carlton, A. B., and Burau, R. G. (1979). Phosphorus fertilization with drip irrigation. Soil Sci. SOC.Am. J. 40,68-72. Ravikovitch, S., and Yoles, D. (1971). The influence of phosphorus and nitrogen on millet and clover growing in soil affected by salinity. I. Plant development. Plant Soil 35,555-567. Ravina, I., Paz, E., Sofer, Z., Marcus, A., Schischa, A,, and Sagi, G. (1992). Control of emitter clogging in drip irrigation with reclaimed wastewater. Irrig. Sci. 13, 129-139. Rhoades, J. D., and Oster, J. D. (1986). Solute content. In “Methods of Soil Analysis. Part I” (A. Klute, ed.), 2nd Ed., Agronomy No. 9985-1006. Am. SOC.Agron., Madison, WI. Richards, S. J. (1965). Soil suction measurements with tensiometers. In “Methods of Soil Analysis. Part I.” (C. A. Black, ed.), Agronomy No. 9153-163. Am. SOC.Agron., Madison, WI. Robinson, T. L., and Stiles, W. C. (1993). Fertigation of young apple trees to maximize tree growth, fruit production and profitability. Compact Fruit Tree 26,61-65. Rolston, D. E., Rauschkolb, R. S., Phene, C. J., Miller, R. J., Uriu, K., Carlson, R. M., and Henderson, D. W. (1979). Applying nutrients and other chemicals to trickle irrigated crops, Bulletin No. 1893, University of California. Rolston, D. E., Miller, R. J., and Schulbach, H. (1986). Fertilization. In “Trickle Irrigation for Crop Production” (F. S. Nakayama and D. A. Bucks, eds.), pp. 317-344. Elsevier, Amsterdam. Rosenthal, W. D., Vanderlip, R. L., Jackson, B. S., Arkin, G. F. (1989). SORKAM: A grain sorghum growth model, TAES computer software documentation Ser. No. MP-1669. Texas Agric. Exp. Stn., College Station. Rudich, J., Sagiv, B., Geisenberg, H., Gera, G., Feigin, A,, Ohayon, M., Kalmar, D., and Albo-Yaron, A. (1982). Response of processing tomatoes to nitrogen fertilization applied through trickle irrigation. Hussadeh 63,52-56. [In Hebrew] Sagi, G., Paz, E., Ravina, I., Schicha, A., Marcus, A., and Yechieli, Z. (1995). Clogging of drip irrigation systems by colonial protozoa and sulfur bacteria. Proc. 5th Int. Microirrigation Congr., Orlando, FL, pp. 250-254. Sagiv, B., Bar-Yosef, B., and Kafkafi, U. (1977). “Drip Irrigation and Fertilization in Besor: Bell Pepper.” Agriculture Research Organization, Bet Dagan, Israel. [In Hebrew] Sagiv, B., Bar-Yosef, B., Kafkafi, U., and Mini, A. (1978). Fertilization and manuring on sprinkle-irrigated fields of pepper vs. fertilization via a trickle irrigation system, Report No. 763. Institute of Soils and Water, Agricultural Research Organization, Division of Scientific Publications, Bet Dagan, Israel. [In Hebrew] Sagiv, B., Bar-Yosef, B., and Eliah, E. (1980). Drip irrigation and fertilization of spring muskmelon at Besor, Pub. No. 17. Division of Scientific Publications, The Volcani Center, Bet Dagan, Israel. [In Hebrew] Sagiv, B., Bar-Yosef, B., Eliah, E., and Mini, A. (1983). Response of sweet corn to trickle irrigation and N fertilization under drip and sprinkle irrigation, Report No. 301/064. Agricultural Research Organization, Bet Dagan, Israel. Sagiv, B., Feigin, A,, Hadas, A,, Keinan, M., and Markovich, T. (1992). Response of drip imgated Chinese cabbage to manure and municipal refuse composts and inorganic N fertilization, Report No. 301- 154-90. Agricultural Research Organization, Bet Dagan, Israel. Sagiv, B., Bar-Yosef, B., Hadas, A., Soriano, S., and Levkovich, I. (1995). Response of drip imgated carrot (cv. Buror) to manure and municipal refuse composts and inorganic N fertilization, Report No. 301-504-90. Agricultural Research Organization, Bet Dagan, Israel. Saksema, R. S. (1995). Microirrigation in India-Achievement and prospective. Proc. 5th Int. Microirrigation Congr., Orlando, FL, pp. 353-358. Seginer, I. (1989). Optimal greenhouse production under economic constrains. Agric. Syst. 29, 67-80.

76

B. BAR-YOSEF

Shainberg, I. (1992). Chemical and mineralogical components of crusting. In ‘‘Soil Crusting. Chemical and Physical Processes” (M. E. Sumner and B. A. Stewart, eds.), pp. 33-54. Lewis, London. Shainberg, I., and Oster, J. D. (1978). “Quality of Irrigation Water.” International Irrigation Information Center, Bet Dagan, Israel. Shalhevet, J., Mantell, A,, Bielorai, H., and Shimshi, D. (1981). Irrigation of field and orchard crops under arid conditions, Publ. I . Int. Inig. Information Center (IIIC), Bet Dagan, Israel. Shevah, Y., and Waldman, M. (1989). Advances in management of fertilizer application and the resulting effect on the pollution load in soil. In “Inorganic Contaminants in the Vadose Zone” (B. Bar-Yosef, N. J. Barrow, and J. Goldshmid, eds.), pp. 179-189. Springer-Verlag. New York. Silberbush, M., Ben-Asher, J., Kaflcafi, U., and Lips, S. H. (1985). Fertigation of peanuts grown in sandy dunes and irrigated by trickling with saline water. In “Proceedings of the 3rd International DriplTrickle Irrigation Congress, Fresno, CA,” Vol. 2, pp. 652-658. Agribusiness, Fresno. Sinclair, T. R., and Muchow, R. C. ( I 99.5). Effect of nitrogen supply on maize yield: I. Modeling physiological responses. Agron. J . 87,632-641. Singh. S. D., Gupta, J. P., and Singh, P. (1978). Water economy and saline water use by drip irrigation. Agron. J. 70,948-95 1. Slangen, J. H. G., Titilaer, H. H. H., and Glas, W. (1988). The importance of fertigation for the improvement of N-fertilizer use efficiency in lettuce culture. Acra Horr. 222, 135-146. Sneh, M. (1995). The history of fertigation in Israel. In “Proceedings of the Dahlia Greidinger International Symposium on Fertigation,” pp. 1-10. Technion, Haifa, Israel. Sonneveld, C. ( 1995). Fertigation in the greenhouse industry. In “Proceedings of the Dahlia Greidinger International Symposium on Fertigation,” pp. 121- 140. Technion, Haifa, Israel. Sparks, D. L. (1986). Kinetics of reactions in pure and mixed systems. In “Soil Physical Chemistry” (D. L. Sparks, ed.). CRC Press, Boca Raton, FL. Sposito, G. (1984). “The Surface Chemistry of Soils.” Oxford Univ. Press, New York. Sposito, G. (1989). “The Chemistry of Soils.” Oxford Univ. Press, New York. Stanford, G., and Smith, S. J. (1972). Nitrogen mineralization potential of soils. Soil Sci. SOC.Am. Proc. 36,465-472. Stanhill, G. (1985). Water use efficiency. Adv. Agron. 39,53-85. Steffen, K. L., Dann, M. S., Harper, J. K., Fleischer, S. J., Mkhize, S. S., Grenoble, D. W., Macnab, A. A., Fager, K., and Russo, J. M. (1995). Evaluation of the initial season for implementation of four tomato production systems. J. Am. SOC.Hort. Sci. 120, 148- 156. Steiner, A. A. (1996). Principles of plant nutrition by a recirculated nutrient solution. Proc. ISOSC 9th Int. Congr. Soilless Culture, pp. 505-513. Sumner, M. E., and Stewart, B. A. (eds.) (1992). “Soil Crusting. Chemical and Physical Processes.” Lewis, London. Sundin, P., Waechter-Kristensen, B., Mari, S. Y., Gertsson, U. E., Hagton, U., Jensen, P., Lund, J., Knutsson, M., Jonsson, J., and Mathiasson, M. (1996). Phytotoxic organic compounds in the circulating nutrient solution of a closed, hydrophonic tomato culture. Proc. ISOSC 9th Int. Congr. Soilless Culture, pp. 523-533. Swietlik, D. (1992). Yield, growth and mineral nutrition of young “Ray Ruby” grapefruit trees under trickle or flood irrigation and various nitrogen rates. J. Am. Soc. Hort. Sci. 117,22-27. Tarchitzky, J., Chen, Y., and Banin, A. (1993). Humic substances and pH effects on sodium and calcium montmorillonite flocculation and dispersion. Soil Sci. Soc. Am. J. 57,367-372. Threadgill, E. D., Eisenhaur, D. E., Young, J. R., and Bar-Yosef, B. (1990). Chemigation. In “Management of Farm Irrigation Systems” ( J . H. Hoffman, T. A. Howell, and K. H. Solomon, eds.), pp. 749-780. Am. SOC.Agric. Eng., St. Joseph, MI. Timlin, D., Acock, B., and van Genuchten, R. (eds.) (1996). “2DSOIL-A Modular Simulation of Soil and Root Processes. Release 03.” Remote Sensing and Modeling Lab., USDA, Beltsville, MDlSoil Salinity Lab., USDA, Riverside, CA.

ADVANCES IN FERTIGATION

77

Tucker, B., and Hagin, J. (1982). “Fertilization ofDryland and Irrigated Soils.” Springer-Verlag, Berlin. Unu, K., Carlson, R. M., and Henderson, D. W. (1977). Application of potassium fertilizers to prunes through a drip irrigation system. Proc. 7th Int. Agric. Plastics Congr., San Diego, pp. 21 1-214. Uriu, K., Carlson, R. M., Henderson, D. W., Schulbach, H., and Aldrich, T. M. (1980). Potassium fertilization of prune trees under drip irrigation. Am. Soc. Hort. Sci. J. 105,508-510. U.S. Salinity Laboratory Staff. (1954). Diagnosis and improvement of saline and alkaline soils, USDA Agric. Handbook No. 60. USDA, Washington, DC. van Keulen, H., and Dayan, E. (eds.) (1993). TOMGRO-A greenhouse-tomato simulation model, CABO-DLO, Simulation Report No. 29. Center for Agrobiological Research, Wageningen Agricultural University, Wageningen. Vera, J., and de la Pena, J. M. (1995). FERTIGA-Acomputer program for fruit tree fertigation. Proc. 5th Int. Microirrigation Congr., Orlando, FL, pp. 194-199. Wagenet, R. J., and Hudson, J . L. (1987). “LEACHM: Leaching Estimation and Chemistry Model.” Center for Environmental Research, Comell University, Ithaca, NY. Wall, T. E., Hochmuth, G. J., and Hanlon, E. A. (1989). Calibration of Mehlich-I and -111 extractable potassium for polyethylene-mulched, drip-irrigated cauliflower. Pmc. Soil Crop Sci. Soc. Florida 48,46-49. Wallerstein, I. S., Bar-Yosef, B., Sagiv, B., Lobel, R., and Schiffmann, J . (1982). Effects of trickle irrigation rate and interval and of fertilization level on rhizobium-inoculated peanuts. Proc. Am. Peanut Res. Educ. SOC.,Vol. 14, p. 98. Warrick, A. W. (1986). Design principles. Soil water distribution. I n “Trickle Irrigation for Crop Production” (F. s.Nakayama and D. A. Bucks, eds.), pp. 93-1 16. Elsevier, Amsterdam. Weast, R.C. (1977). “CRC Handbook of Chemistry and Ph ’ CRC Press, Cleveland. OH. Williams, M., and Yanai, R. D. (1996). Multi-dimensional vity analysis of ecological implications of a nutrient uptake model. Planr Soil 180,31 1 -324. Worley, R. E., Daniel, J. W., Dutcher, J. D., Harrison, K., and Mullinix, B. G. ( I 995). A long term comparison of broadcast application versus drip fertigation of nitrogen for mature pecan trees. HorrTechnology 5,43-47. Wu, I. P. (1995). A simple optimal microirrigation scheduling. Proc. 5th Int. Microirrigation Congr., Orlando. FL, pp. 781 -786. Yanuka, M., Leshem, Y., and Dovrat. A. ( I 982). Forage corn response to several trickle irrigation and fertilization regimes. Agron. J. 74,736-740. Yarwood, C. E. (1978). Water and the infection process. In “Water Deficit and Plant Growth” (T. T. Kozlowski, ed.), Vol. 5 , pp. 141-165. Academic Press, New York. Zaslavsky, D., and Mokady, R. S . (1967). Nonuniforni distribution of phosphorus fertilizers: An analytical approach. Soil Sci. 104, 1-6. Zazueta, F. S., Clark, G. A., Smajstrla, A. G., and Carrillo, M. (1995). A simple equation to estimate soil-water movement from a drip irrigation source. Proc. 5th Int. Microirrigation Congr., Orlando, FL.. pp. 85 1-856.

This Page Intentionally Left Blank

THEGENETICS,PATHOLOGY, AND MOLECULAR BIOLOGY OF T-CYTOPLASM MALE STERILITYINMAIZE* Roger P. Wise, * Charlotte R. Bronson,2 Patrick S. S~hnable,~ and Harry T. Horner4 32

'Corn Insects and Crop Genetics Research Unit, USDA-ARS *Department of Plant Pathology 3Departments of Agronomy and Zoology & Genetics 4 D e p a r ~ e nof t Botany Iowa State University Ames, Iowa 50011-1020

I. Introduction 11. Cytoplasmic Male Sterility Systems 111. cms-T Causes Premature Degeneration of the Tapetum W. Southern Corn Leaf Blight Epidemic of 1970 A. The Rise and Fall of Race T B. The Exceptional Virulence of Race T C. Yellow Leaf Blight and the Insecticide Methomyl D. T-Toxin and PM-Toxin Disrupt Mitochondrial Function E. The Hazards of Genetic Homogeneity V. Disease Susceptibility and Male Sterility A. How Fungal Toxins Disrupt Mitochondrial Function B. How URFl3 Causes Sterility Remains a Mystery VI. Nuclear-Cytoplasmic Interactions and Restoration of cms-T A. Rfl, But Not Rj2, Alters the Expression of T-urf? B. Cloning of Nuclear Restorer Genes C. How Does the Rf2-Encoded ALDH Mediate Fertility Restoration? W. Perspectives by cms-T Researchers VIlI. Future Directions References

*Productnames are necessary to report factually on available data; however, the USDA neither guarantees nor warrants the standard of the product, and the use of the name by the USDA implies no approval of the product to the exclusion of others that may also be suitable. 79 Advnnres m Agronomy, Volume 65

Copyright 0 1999 by Academic Press. All rights of reproduction in any form reserved 0065-2 I 13/99 $30.00

80

ROGER P. WISE ET AL.

I. INTRODUCTION Texas cytoplasmic male sterility (cms-T or T cytoplasm)+ in maize has been studied intensely since its discovery in the 1940s. It is the most environmentally stable system for producing hybrid seed in maize and, as such, dominated the maize seed industry in the United States and elsewhere throughout the 1960s. In 1969, however, T-cytoplasm lines exhibited a high degree of susceptibility to two fungi, Mycosphaerella zeae-maydis and race T of Cochliobolus heterostrophus, which induce severe disease by the production of P-polyketol toxins. In 1970, an epidemic caused by C. heterostrophus effectively ended the use of cmsT for commercial maize production. The epidemic emphasized to breeders and others the real and potential dangers of genetic homogeneity in modem cropping systems. In contrast, the discovery that T-cytoplasm lines are acutely susceptible to the fungi and their toxins has proven a boon to research on mechanisms of male sterility in maize. Research stimulated by the epidemic has shown that male sterility is controlled by the same mitochondrial gene that is responsible for sensitivity to the fungal toxins. In fact, identification and cloning of the gene was facilitated by the use of toxin sensitivity, rather than male sterility, as an assay for the presence of a functional gene. This gene, T-urfl3, encodes an oligomeric protein, URF13, assembled in the inner mitochondrial membrane. Binding of thz fungal toxins to URFl3 causes membrane leakage and disruption of mitochondrial function; thus, the toxins were crucial in demonstrating that the URFl3 protein may have the capacity to form a pore spanning the inner mitochondrial membrane. The toxins may also provide insight into how T-urf1.3 causes male sterility. In cms-T plants, male sterility is associated with premature breakdown of the mitochondria-rich, tapetal cell layer of the anther; this layer is crucial to pollen production because it supplies nutrients to the developing microspores. The ability of the toxins to cause mitochondrial dysfunction suggests that male sterility may be due to an endogenous, toxin-like compound in maize that interacts with URF13 to cause mitochondrial dysfunction and subsequent death of the tapetal cells. Because of the wealth of molecular information available, cms-T now serves as a model for studies of male sterility and fertility restoration in plants. This review complements previous reviews on cms-T (Duvick. 1965; Edwardson, 1970; Ullstrup, 1972; Levings and Pring, 1979; Laughnan and Gabay-Laughnan, 1983; Kaul, 1988; Pring and Lonsdale, 1989; Levings, 1990, 1993; Levings +Nomenclature: According to the present maize nomenclature, loci and recessive alleles are designated by lowercase symbols: for example, the r f f allele of the rfl locus is a recessive mutant. Dominant alleles are designated by uppercase symbols; for example, the Rfl allele of the rfl locus is wild type. Lines that carry T cytoplasm (sterile or fertile) are referred to as T-cytoplasm lines. Male-sterile lines that carry T cytoplasm are designated cms-T. Restored T cytoplasm designates lines restored to fertility via the presence of nuclear restorer genes. N-cytoplasm lines are male fertile.

T-CYTOPLASM MALE STERILITY IN MAIZE

81

and Seidow, 1992; Williams and Levings, 1992; Dewey and Korth, 1994; Ward, 1995) by providing, in addition to the directly pertinent information regarding the mechanisms of toxin sensitivity and male sterility, information about the production of hybrids, the cytology of male sterility, the mechanisms of pathogenesis in C. heterostrophus and M. zeae-maydis, and recent breakthroughs in understanding fertility restoration in cms-T. This review should be useful, therefore, to students new to cms-T research, as well as to advanced investigators. At the end of the review are commentaries by prominent researchers that provide the reader with insights gained from years of work with cms-T.

11. CYTOPLASMIC MALE STERILITY SYSTEMS Heterosis plays a key role in applied plant breeding programs. F, progeny resulting from a cross between two (usually inbred) lines often exhibit enhanced expression of one or more positive characters relative to both parents. This improved condition is termed hybrid vigor or heterosis. Because selected maize hybrids display significant amounts of heterosis, they are widely used for agricultural production. To produce hybrids, pollen must be transferred from one parent (the male donor) and used to fertilize another (the female recipient). In addition, the female parent must not produce pollen that can compete with that provided by the male parent. Self-pollination of the female parent can be avoided by removing the male floral organs (emasculation), disrupting the development of these structures, or otherwise preventing the female parent from producing functional pollen. For example, the pollen-bearing tassel of maize can be removed from the female parent using machine or hand emasculation (Fig. 1, see color plate). However, these procedures can reduce the yield of F, seed from the female parent and they are expensive. Alternatively, the female parent can be provided with a genetic constitution that ensures that it will be male sterile. Male sterility results when plants do not form complete anthers; or when the internal male cells of the anther (microspore mother cells, microspores, or pollen) become nonfunctional during some stage of development; or when the anthers fail to release pollen. Mutations that cause these effects have been described in nuclear and mitochondria1 genes from many plant species (Laser and Lersten, 1972; Palmer el al., 19921, including a large number of crops, such as maize (Duvick, 1965). Although methods have been proposed to use nuclear male-sterile mutants to facilitate the production of hybrid seed (Jones et al., 1957), none have proved practical, at least in maize. However, it has been possible to develop efficient male sterility systems in maize by using nuclear genes (restorers) that complement cytoplasmic mutations inherited solely through the maternal line.

82

ROGER I! WlSE E l

HL.

Figure 2 Production of F, hybrid seed from two inbred lines using the cms-T maize system. An A line that is male sterile because it carries T cytoplasm, but not all of the nuclear restorers required for fertility restoration, is used as the female parent. An R line that is male fertile and carries the necessary restorers serves as the male parent. Because the A and R lines are planted in isolation from other maize, the only pollen available to fertilize the A line is derived from the R line. Hence, all seed set on the A line is hybrid. Although plants produced from this F, seed will carry the maternally transmitted T cytoplasm, they will be male fertile due to the paternal contribution of nuclear restorers (I?’).

Commercial maize hybrid seed production requires the use of three lines: an A line (the female parent of the F,) that is crns; a B line (the maintainer of the A line) that is isogenic relative to the A line, but that carries a normal (N) cytoplasm and is therefore male fertile; and an R line that carries the necessary restorer of fertility genes (designated Rf)and serves as the male parent of the F, (Fig. 2). The cmsA line is formed by backcrossing a male-fertile line to another line that carries a cms cytoplasm. One or both of these parents must lack a necessary nuclear Rf gene. This creates a new pair of A and B lines. An R line is formed by backcrossing Rf genes into the desired male parent of the F, hybrid. On a commercial scale, F, seed is produced by planting alternating rows (4-6) of the cms-T, male-sterile, femalefertile A line with one or two rows of the male-fertile R line. The F, hybrid seed is harvested from the A line. Three major groups of male-sterile cytoplasms have restorer capabilities and, hence, the potential to be used for the production of hybrid seed, S (USDA), C (Charrau), and T (Texas). These cytoplasms have different nuclear genes that suppress their associated male-sterile phenotype and restore normal pollen development (Duvick, 1965; Beckett, 1971; Gracen and Grogan, 1974; Laughnan and

T-CYTOPLASM MALE STERILITY IN MAIZE

83

Gabay-Laughnan, 1983). They can also be distinguished from each other by mitochondrial DNA (mtDNA) restriction endonuclease profiles (Levings and Pring, 1976; Pring and Levings, 1978; Borck and Walbot, 1982) and by characteristic polypeptide patterns resulting from [35S]methionineincorporation by isolated mitochondria (Forde et al., 1978; Forde and Leaver, 1980). The male-sterile C , S, and T cytoplasms produce fertile plants only in nuclear backgrounds carrying the appropriate restorer genes. These nuclear-encoded fertility-restoring genes compensate for cytoplasmic dysfunctions that are phenotypically expressed during microsporogenesis and/or microgametogenesis. Plants carrying S and C cytoplasms are restored to fertility by single dominant alleles of the $3 and $4 genes, respectively (Laughnan and Gabay, 1978; Kheyr-Pour et al., 1981; Laughnan and Gabay-Laughnan, 1983).In contrast, T cytoplasm is restored to fertility by the combined action of dominant alleles of the rfl and $2 genes (Duvick, 1956,1965; Laughnan and Gabay-Laughnan, 1983).Of the three types of cms found in maize, the cms-T type is the most stable under all environmental conditions and was the primary type used in maize hybrid seed production through 1970.

IJI. CMS-T CAUSES PREMATURE DEGENERATION OF THE TAPETUM In normal, male-fertile plants, pollen production is the result of a precisely timed sequence of events known as microsporogenesis and microgametogenesis (Mascarenhas, 1988, 1990; Koltunow et al., 1990; Goldberg et al., 1993). Disruption or incorrect timing of these events may lead to the failure of pollen development or the prevention of pollen release (Beals and Goldberg, 1997). It is now clear that male sterility in cms-T is caused by premature degeneration of the tapetum, a metabolically active cell layer of the anther. The processes of pollen formation in male-fertile maize and abortive development in cms-T maize are diagrammed in Fig. 3. Each anther consists of four microsporangia (sacs) held together by connective tissue (Palmer et al., 1992; Goldberg et al., 1995; Horner and Palmer, 1995). The earliest stage at which all of the anther tissues are distinguishable is the sporogenous mass stage. During this stage, the sporogenous cells undergo DNA synthesis, secrete callose walls around themselves, and become somewhat rounded. Meiosis occurs during the meiocyte, dyad, and tetrad stages. Following meiosis I at the dyad stage, a partitioning callose wall forms and separates the two resulting nuclei. Following meiosis I1 at the tetrad stage, two more callose walls form that separate the four resulting nuclei and their cytoplasms. Individual tetrads, each of which now consists of four microspores, are encased in callose. Athin microspore wall forms interior to the callose around each microspore, and within each thin wall a single pore is delimited.

S

Stage 1-7:

Microsporogenesis

stages 8-10 Microgametogenesis

Figure 3 Scheme showing the stages of nucrosporogenesis (1-7) and microgametogenesis (8-10) in fertile anthers of maize and stages that lead to tapetal degeneration and male-cell abortion in anthers of cms-T maize (5'-7') Each stage depicts only one of four locules in an anther and includes only affected cells, namely male cells and the surrounding tapetum cms-T tapetal cells show initial signs of degeneration at about the early mcrospore stage, and by the nud to late microspore t no pollen is produced by cms-T plants. stages the tapetum and male cells appear in vanous states of degeneration. In m o ~circumstances.

Figure 1 Photographs of emasculated maize fields and individual tassels of fertile and cms-T plants. (A) Machine-emasculated field. (B) A tractor with boom carries modified lawn mowers separated at distances to emasculate rows that serve as female parents. (C) Hand-emasculated field. (D) Hand removal of tassel (emasculation). (E) Tassel of male-fertile plant. (F) Tassel of cms-T plant (male sterile).

Figure 6 Aspects of southem corn leaf blight. (A) Lesions caused by race T of Cochliobolus hererostrophus on leaves of T-cytoplasm maize. Lesions, which may extend up to 27 mm long and 12 mm wide, are spindle-shaped due to their expansion beyond the leaf veins. In contrast, lesions of race T on N-cytoplasm maize, or race 0 on either N- or T-cytoplasm maize are small, generally less than 15 mm long and 1 to 3 mm wide; lesions are parallel sided due to their restriction by the leaf veins. (B) Conidia of C. heferosrrophus.Conidia, which measure 30 to 115 pm in length and 10 to 17 pm in diameter, are produced on the surface of infected maize leaves or infested debris under moist conditions. (Photograph by E. J. Braun, Iowa State University.) (C) Infection of maize by C. heferostrophus.In presence of free moisture, conidia germinate from both ends to form germ tubes which grow along the leaf surface until they reach depressions formed at the junctions between epidermal cells. At the ends of germ tubes, swellings, known as appressoria, form. Penetration of corn leaves occurs below appressoria. Conidia are shown on a plastic replica of a corn leaf. (Photograph by E. J . Braun, Iowa State University.)

Figure 10 Isolated N-cytoplasm and T-cytoplasm maize leaf protoplasts, untreated or treated with the cationic potential-sensitive fluorescence dye 3,3’-dihexyloxacarbocyanineiodide to indicate metabolically active and inactive mitochondria when treated with T-toxin and respiratory inhibitors. (A) Untreated N-cytoplasm protoplast showing active mitochondria (bright dots). (B) Untreated T-cytoplasm protoplast showing active mitochondria (bright dots). (C) T-cytoplasm protoplast treated with carbonyl cyanide m-chlorophenylhydrazone, a respiratory inhibitor; no mitochondria fluoresce. (D) T-cytoplasm protoplast treated with T-toxin; no mitochondria fluoresce. Bars = 17 pm. N-cytoplasm protoplast mitochondria do not fluoresce when treated with the respiratory inhibitor, but they do when treated with T-toxin (not shown).

T-CYTOPLASM MALE STERILITY INMAIZE

85

Each pore is oriented toward the tapetum, the cell layer immediately surrounding the microspores. This layer, which consists of a mixture of uni- and bi-nucleate cells (Horner er al., 1993), functions in secretion, providing an array of substances needed by the developing male cells (Pacini et al., 1985; Pacini, 1997). The tapetum is thus considered critical for normal pollen development (Laser and Lersten, 1972). The very high metabolic activity of these cells is reflected in the number of mitochondria present. Tapetal cells have about twice the number of mitochondria as sporogenous cells and about 40 times the number of mitochondria as most maize somatic tissues (Lee and Warmke, 1979). The tapetal layers of N and cms-T lines are similar; however, cms-T lines have a consistently greater frequency of cells that are bi-nucleate versus uni-nucleate and, uni-nucleate tapetal nuclei of cms-T lines contain a higher level of DNA than their fertile counterparts (Figs. 4A and 4B; Horner er al., 1993). The significance of these differences is not clear. Following the tetrad stage, the callose around the tetrads is digested, releasing individual micropsores. All microspores remain pressed to the tapetum and, as the microspores round up and enlarge, the single nucleus moves to the microspore wall. Each microspore wall continues to thicken. These last stages are referred to as the early, mid, and late microspore stages. By the latter stage, the nucleus undergoes one mitotic division to form a pollen grain containing a generative cell and a tube cell. Interior to the microspore wall (exine), an inner wall (intine) is initiated, thus creating the pollen wall. At about this time, nutrient reserves form in each pollen grain as the tapetum degenerates. During the next two pollen stages, reserves continue to build up in each grain and the generative cell divides to form

Figure 4 Electron micrographs of flow cytometrically isolated fertile uni- and hi-nucleate tapetal protoplasts from developing maize anthers. The ratio of hi-nucleate to uni-nucleate tapetal cells is greater in cms-T anthers than in fertile anthers, from sporogenous mass stage through microspore stage. (A) Uni-nucleate tapetal protoplast. (B) Bi-nucleate tapetal protoplast. Bars = 5 pm. Published with permission from Springer-Verlag,Austria from Homer et ul. ( 1993).

86

ROGER P. WISE ETAL.

two sperm cells. Each pollen grain (three-celled, male gametophyte) is engorged with reserves at the time the pollen is shed for wind pollination (Fig. lE, see color plate). In cms-T anthers, degeneration of the tapetum occurs prematurely, beginning at the dyad (Yang, 1989) to the early microspore stages (Warmke and Lee, 1977). Mitochondria in the tapetal cells and middle wall layer show signs of structural disorganization, becoming enlarged with a less dense matrix and more diffuse cristae (Gengenbach et al., 1973; Watrud et al., 1975b; Gregory et al., 1977, 1978; Wannke and Lee, 1977; Yang, 1989). Dysfunction of the mitochondria precedes premature degeneration of the tapetum and abortion of the microspores (male cells) before they reach the pollen stage (Fig. 3). Thus, pollen does not form and the anthers do not exsert (Fig. lF, see color plate). The cause of premature tapetum degeneration in cms-T anthers is not known. However, given the very high metabolic activity of these cells, any mutation that disrupts mitochondrial function has the potential to result in premature cell death and male sterility (Wannke and Lee, 1978; Lee and Warmke, 1979; Levings, 1993). Mitochondria of cms-T tapetal cells differ from those of N-cytoplasm tapetal cells in the presence of the URFl3 protein in the inner mitochondrial membrane. URF13 thus has the potential to interfere with normal mitochondrial function. cms-T mitochondria are also sensitive to the fungal toxins that interact with URF13. Based on these observations, a model has been suggested for tapetum degeneration in which an unknown “Factor X,” acting similarly to the fungal toxins, disrupts mitochondrial function in the tapetal cells (Fig. 5; Flavell, 1974). Either Factor X is tapetal specific, or tapetal cells are especially sensitive to it, since the mitochondria of all cells in cms-T lines express URF13, but only the tapetum degenerates. Intriguing preliminary evidence for a factor in maize with these toxinlike properties is described by Levings in Section VIID.

Iv. SOUTHERN CORN LEAF BLIGHT EPIDEMIC OF 1970 T-cytoplasm maize is highly susceptible to two different fungi, Mycosphuerella zeae-maydis (asexual stage: Phyllosticta maydis), causal agent of yellow leaf blight, and race T of Cochliobolus heterostrophus (Drechs.) Drechs. [asexual stage: Bipolaris maydis (Nisikado & Miyake) Shoemaker = Helminthosporium maydis (Nisikado & Miyake)], causal agent of southern corn leaf blight. Both fungi produce toxins that induce susceptibility in male-sterile and fertility-restored Tcytoplasm lines by disrupting mitochondrial function. The toxin from M. zeae-maydis is known as PM-toxin; the toxin from C.heterostrophus is known as T-toxin.

T-CYTOPLASM MALE STERILITY INMAIZE

-

87

Stimulus

- Factor X Microspores

Tapetal cell

Figure 5 Model showing how cms-T tapetal cells might he triggered to break down prematurely, resulting in male-cell abortion and sterility. The model depicts just one of the two possible nuclei and just one of the many mitochondria in a tapetal cell. The URF13 protein is located in the inner mitochondrial membrane. An as yet unidentified, toxin-like “Factor X ’ interacts with URF13 to cause membrane leakage (Flavell, 1974).Factor X may be either imported from outside of the tapetal cell or produced within the tapetal cell as a result of gene expression following an unidentified stimulus from either within or outside of the tapetal cell.

Of the two fungi, the best known is C. heterostrophus because it caused a worldwide pandemic in 1970. The epidemic was most severe in the United States. The history of the epidemic has been reviewed previously (Tatum, 1971;Hooker, 1972; Ullstrup, 1972); however, it is worth repeating here because of its impact on both the utility of T cytoplasm as a source of male sterility and research into mechanisms of male sterility. Prior to 1970, C. heterostrophus was considered a moderate leaf-spotting pathogen on maize and teosinte in tropical and subtropical regions (Drechsler, 1925). The disease favors warm, wet environments and is generally of little concern to growers in temperate climates. In the United States, it is restricted primarily to the southeastern states (Leonard, 1987). Only a single race of C. heterostrophus, which we now call race 0, was known prior to the epidemic. Race 0 normally infects only leaf tissues, where it causes small, elongated, parallel-sided lesions up to 15 mm in length and 1-3 mm in width. The lesions are generally limited to a single interveinal region (Drechsler, 1925). Perhaps because the disease was not considered a threat to maize in temperate climates, the first published report that T-cytoplasm lines might be highly suscep-

88

ROGER P. WISE ETAL.

tible to C. heterostrophus went generally unnoticed. In 1961, Mercado and Lantican reported that T-cytoplasm lines, which had been introduced into the Philippines by 1957, were highly susceptible to C. heterostrophus (Mercado and Lantican, 1961). On T-cytoplasm lines, C. heterostrophus caused irregularly shaped lesions that coalesced, killing entire leaves and causing the plants to die prematurely. Efforts to backcross male sterility from these lines into resistant Philippine inbred lines resulted in the conversion of the Philippine lines to susceptibility and demonstrated that the high susceptibilitywas cytoplasmicallyinherited along with male sterility (Mercado and Lantican, 1961; Villareal and Lantican, 1965). The presence of restorer genes did not alter susceptibility (Villareal and Lantican, 1965). Despite these early indications that T-cytoplasm maize might be highly susceptible to C. heterostrophus, no signs of unusual susceptibility were noticed in other parts of the world and use of T-cytoplasm lines in temperate regions increased during the 1960s. By 1970, 85% of the hybrid maize in the United States was T cytoplasm (Ullstrup, 1972). In 1969, severe symptoms due to C. heterostrophus were reported in the United States in a number of midwestern and central states (Ullstrup, 1972; Scheifele et al., 1970; Leonard, 1987). The symptoms, which were attributed to a new race of the fungus called race T (Hooker et al., 1970b; Smith et al., 1970), included foliar lesions up to 27 mm in length and 12 mm in width. The lesions expanded beyond the leaf veins, giving them a spindle-shaped appearance (Fig. 6A, see color plate). The high, specific virulence of race T was attributed to its production of a toxin known as T-toxin, which specifically affects maize lines with T cytoplasm (Hooker et al., 1970a;Lim and Hooker, 1971, 1972). Warm, wet conditionscombined with inoculum of race T carried over from 1969 set the stage for a severe epidemic in the United States in 1970. The first reports of serious damage were in Florida in January 1970 (Moore, 1970). In addition to leaf infections race T attacked leaf sheaths, stalks, husks, shanks, ears, and cobs. Lodging due to secondary infections was also common (Hooker, 1972). The disease spread rapidly from the southern states into the midwestem states, northern states, and finally into southern Canada. Losses in the United States were estimated at 7 10 million bushels (Ullstrup, 1972). Losses were reported in other countries as well. Severe damage due to race T was noted in West Africa in 1970 (Craig, 1971). Fortunately less favorable weather conditions and reduced planting of Tcytoplasm lines resulted in only moderate damage in Brazil, thus reducing international losses (Ullstrup, 1972). In the United States, race T overwintered from 1970 to 1971 in corn debris on the soil surface and in refuse piles, resulting in localized outbreaks in 1971. However, drier, cooler weather combined with the reduced planting of T-cytoplasm lines in the spring of 1971 resulted in less severe disease and near record harvests. By the winter of 1971, sufficient seed of N-cytoplasm lines was available to meet demands for spring planting in 1972 (Ullstrup, 1972).

T-CYTOPLASM MALE STERILITY IN MAIZE

89

A. THERISEAND FALLOF RACE T The origin of race T is not known. It is clear that widespread planting of T-cytoplasm lines in the United States and elsewhere in the 1960s created a situation favorable for the rapid increase of race T. However, its source remains a mystery. Race T was not endemic, or at least not widespread, in most countries prior to 1970. The first known occurrence of race T in the United States was on seed stored in 1968 in Iowa (Foley and Knaphus, 1971), with subsequent occurrences in several midwestem and central states in 1969 (Leonard, 1987). In West Africa, race T was observed for the first time in 1970,even though T-cytoplasm lines had been grown there since 1963 (Craig, 1971). Race T was detected in Australia for the first time in 1972 (Alcorn, 1975). If race T had been endemic in these countries prior to 1970, it should have been detected earlier, shortly after T-cytoplasm lines were introduced. Epidemiological studies have provided further evidence that race T was not endemic in the United States. Iowa, the site of the first known occurrence of race T in the United States, is well outside the normal endemic range of C. heterostrophus. The subsequent occurrences in 1969 in several midwestern and central states were also outside of the normal range. If race T had been endemic in the United States, or had arisen by mutation from the race 0 population in the United States, its first appearance would have been more likely in the southeastern states because of their prevalent race 0 population and more favorable environmentalconditions. In addition, when race T first appeared in the United States, it was exclusively mating type I , whereas the endemic race 0 population had equal frequencies of both mating types (Leonard, 1971). If race T had been in the United States for more than a few years before 1968, it should have included both mating types due to intermating with race 0. Race T was also exclusively mating type 1 when it first appeared in Australia in 1972 (Alcorn, 1975). In contrast, race T appears to have been in the Philippines when T-cytoplasm maize was first introduced there in 1957 (Mercado and Lantican, 1961). The disease symptoms reported on T-cytoplasm lines were typical of those we now associate with race T, whereas N-cytoplasm lines showed normal symptoms, indicating that the high virulence was not simply a function of the favorable climatic conditions in the Philippines. In addition, the unusual susceptibility was inherited cytoplasmically.The fact that T-cytoplasm lines were not grown on a widespread basis before the susceptibility was observed suggests that race T may have been present in the Philippines, at least at low levels, at the time of the introduction of T cytoplasm. Given the available information, the simplest hypothesis is that race T was introduced into the United States from the Philippines or elsewhere in the mid-1960s and concurrentlyor subsequently spread to other countries. In fact, Leonard (1987) proposed that the epidemic in the United States arose from a single isolate of race T that became established in the Midwest during the mid- 1960s. This hypothesis

90

ROGER l? WISE ETAL.

is consistent with both the pattern of spread of race T and the distribution of mating types. Leonard further postulated that race T spread to Florida, where the epidemic began in January 1970, on contaminated seed from the Midwest planted in winter breeding nurseries (Leonard, 1987). The frequency of race T in the United States declined rapidly after the epidemic and the withdrawal of T-cytoplasm lines from widespread use (Leonard, 1974, 1977a, 1987). If race T is still present, it is at very low levels. For example, race T symptoms have not been observed in experimental fields planted each year in Iowa from 1988 to 1997, even though these fields have contained up to 100,000 T-cytoplasm plants each year (R. P. Wise and P. S. Schnable, unpublished results). The decline of race T has been attributed to its reduced virulence relative to race 0 on N-cytoplasm maize; race T produces lesions that are 10 to 38% shorter than those produced by race 0 on the same genotypes (Leonard, 1977b; Klittich and Bronson, 1986). Race T also grows more slowly as a saprophyte than does race 0;the reduction in growth measured as dry mass accumulation in 1 week is about 9% (Bronson, 1998). Both of these differences have been shown to be tightly linked to T-toxin production, suggesting that T-toxin production, or genes tightly linked to T-toxin production, reduce fitness when the fungus grows on substrates other than T-cytoplasm maize. A possible cause of the reduced fitness is the metabolic drain imposed by the synthesis of the toxin; T-toxin has been calculated to constitute 2% of the mycelial dry mass of race T (Tegtmeier et al., 1982). The low fitness of race T has raised questions about the timing of its evolution. If race T is as unfit everywhere as it is in the United States, it must have evolved shortly after T-cytoplasm lines became available as a host, otherwise, it would have never survived. Such a coincidental evolution at the time of widespread planting of T cytoplasm is possible, but unlikely. An alternative hypothesis is that race T can survive, at least at low levels, in some as yet undiscovered niche, for example, on a wild plant that is susceptible to P-polyketol toxins. Since race T was presumably present in the Philippines at the time of the introduction of T-cytoplasm lines, and since C. heterostrophus occurs predominately in the tropics and subtropics, regions such as the Philippines might be reasonable places to search for the origin of race T.

B. THEEXCEPTIONAL VIRULENCE OF RACE T The 1970 epidemic resulted in a variety of cytological, genetic, and molecular studies into how C. heterostrophus causes disease. This research has provided insight into the strategies used by races T and 0 to attack maize, and, more specifically, the mechanism by which race T causes severe disease on T-cytoplasm lines. As far as is known, races T and 0 behave identically prior to penetration into maize tissues. The primary infectious propagules of C. heterostrophus are asexu-

T-CYTOPLASM MALE STERILITY IN MAIZE

91

ally produced spores known as conidia (Fig. 6B, see color plate). The conidia, which are wind-dispersed, are released from lesions on infected tissue, usually leaves, and from colonies in debris on the soil surface. Conidia germinate within about 20 min after hydration by rain or dew. The first visible sign of germination is the production of a two-layered mucilaginous matrix on the spore tips (Evans et al., 1982). The inner layer is believed to be responsible for adhering the fungus to surfaces and thus preventing it from being washed off (Braun and Howard, 1994). The outer layer is thought to be necessary for lesion development once the fungus is inside the plant (Zhu et al., 1998). Germ tubes, surrounded by the two matrix layers, emerge a few minutes later from both ends of the conidia. About 2 h after hydration, swellings, known as appressoria, form at the ends of the germ tubes. Most of the appressoria form over the depressions on leaf surfaces at the junctions between epidermal cells. This preference seems to be a response to the physical shape of the surface, rather than to any difference in chemical composition; the fungus shows a similar preference for depressions whether the surface is a maize leaf or a plastic replica of a maize leaf (Fig. 6C, see color plate). Penetration generally occurs within 3-4 h by the production of a penetration peg below the appressorium. Penetration is believed to be enzymatic because of the relatively undifferentiated appressoria and the degraded appearance of cell walls at the sites of penetration. Hyphae are generally observed within mesophyll tissues 6 h after the hydration (E. J. Braun, personal communication). Little is known about the activities of C. heterostrophus after it enters maize tissues. Hyphae of both races T and 0 ramify through tissues, killing host cells and producing the characteristic necrotic lesions. The factor or factors responsible for host cell death or the ability of the fungus to resist host defense responses are unknown. The only gene currently proven to be needed by race 0 to produce lesions is CPSl, which encodes a cyclic peptide synthetase (Yoder, 1998).The substrate(s) of the peptide synthetase is not known. Race T presumably has all the genes needed for normal lesion development as does race 0. However, in addition, it has genes that confer an added level of virulence to maize lines with T cytoplasm. The high virulence of race T to T-cytoplasm maize has been shown to be due to its production of a host-selective toxin known as T-toxin (Yoder, 1976, 1980; Yoder et al., 1977). T-toxin is a family of long-chain, linear P-polyketols with a chain length of C,, to C,, (Kono and Daly, 1979; Kono et al., 1980,1981; Daly and Knoche, 1982; Suzuki et al., 1982, 1983, 1984) (Fig. 7). Because T-toxin specifically affects T-cytoplasm lines, whether male sterile or restored, it was of interest as a tool for plant breeding immediately after its discovery (Gracen et al., 1971;Comstock and Scheffer, 1972) and was used extensively between 1970-1971 to detectT cytoplasm in seed lots (Lim etal., 1971). Early attempts to determine the genetic control of T-toxin production indicated a single genetic locus difference between race T and race 0 (Lim and Hooker, 1971; Tegtmeier et al., 1982; Bronson, 1991); this locus was named Toxl. A few

92

ROGER P. WISE ETAL. OH OH 0

0 OH 0

OH 0

OH 0

O O H O

OH 0

OOHO

0

OH OH

PM-toxin O

Methomyl

CH3-C=N-0-CI S-C%

H

II 1

N-CH3

Figure 7 Structures of T-toxin, PM-toxin, and methomyl. Both T-toxin and PM-toxin are families of linear P-polyketols varying in chain length. Shown is one major component of each toxin: band 1 of T-toxin and band B of PM-toxin. Drawings of T-toxin and PM-toxin structures are adapted from Kono et al. ( 1985).

crosses produced segregation ratios suggestive of additional loci (Yoder and Gracen, 1975; Yoder, 1976); however, these ratios were later shown to be artifacts of a factor linked to Toxl that caused nonrandom abortion of progeny and the illusion of multiple-locus segregation (Bronson et al., 1990). It is now clear that two distinct loci, ToxlA and ToxlB, are required for production of T-toxin (Turgeon et al., 1995) and that these two loci absolutely cosegregate in crosses among naturally occurring strains due to their tight genetic linkage to the breakpoint of a reciprocal translocation (Bronson, 1988; Tzeng et al., 1992; Chang and Bronson, 1996). Race T and race 0 differ by a reciprocal translocation involving chromosomes 6 and 12; the breakpoint of the translocation is at or very near Toxl. Race T also contains, relative to race 0,one or more insertions totaling about 1.2 Mb; this extra DNA maps to the translocation breakpoint (Fig. 8; Chang and Bronson, 1996). Two genes involved in T-toxin synthesis have been cloned: ChPKSI, which maps to the ToxlA locus on chromosome 12;6, and DECI, which maps to the ToxlB locus on chromosome 6;12. ChPKSl was cloned by tagging using restriction enzyme-mediated integration (REMI) of nonhomologous DNA (Lu et al., 1994). ChPKSl contains a 7.6-kbopen reading frame encoding a polyketide synthase containing six catalytic domains (Table I; Fig. 9; Yang er al., 1996). The gene is flanked on both sides by A+T-rich, repeated, noncoding DNA. The other gene, DECI, encodes a decarboxylase required for the modification of T-toxin to its biologically active form (Rose et d., 1996; Yoder, 1998). Disruption of either ChPKSI or DECI results in the simultaneous loss of ability to make active T-toxin and loss of high virulence to T-cytoplasm lines. Sequences corresponding to ChPKSI and DECI are absent in race 0 and in other Cochliobolus species tested,

T-CYTOPLASM MALE STERILITY IN MAIZE

93

Figure 8 Model for the evolution of race T. Race T is hypothesized to have evolved from race 0 by the insertion of a I .2-Mb segment of alien DNA containing genes for T-toxin production. Concurrently or subsequently, a translocation event occurred that placed the genes on two different chromosomes. It is not known whether the segment initially inserted into chromosome 6 or chromosome 12. Chromosomes in this diagram are not drawn to scale. Adapted from an unpublished drawing by B. G. Turgeon and 0. C. Yoder, Comell University.

-

suggesting that these genes may have been acquired by horizontal transfer from another genus (Turgeon and Berbee, 1998; Yang et al., 1996; Yoder, 1998). The tight genetic linkage of ChPKSl and DECl to a chromosome rearrangement has led to a model for the evolution of race T, shown in Fig. 8 (Yang et al., 1996; Yoder, 1998). Given the nature of the chromosomal rearrangement, the simplest hypothesis is that race T evolved from race 0, and that it did so by an insertion into either chromosome 6 or 12 of a 1.2-Mb segment of alien DNA containing the genes for T-toxin production. Either simultaneously or subsequently, a translocation occurred between chromosomes 6 and 12 that placed ToxlA and ToxlB on two different chromosomes. The source of the alien DNA is not known. Evidence for additional genes involved in T-toxin production has been obtained recently by mutational analysis (Lu et al., 1995; Yoder, 1998) and by the analysis of quantitative genetic variation among naturally occurring strains (Bronson, 1998). These studies have identified genes unlinked to Toxl that control the amount of toxin produced by race T strains. These genes may be involved in the production of precursors, or the transport, regulation, or metabolism of T-toxin. As far as is known, T-toxin is the only host-specific toxin produced by C. heterostrophus. However, one impact of the 1970 epidemic was a lingering fear that C. heterostrophus might evolve the ability to produce additional toxins specific for other maize cytoplasms (Hooker, 1972). Species of the genus Cochliobolus produce at least three different host-specific toxins. In addition to T-toxin, there is HC-toxin, which allows C.carbonum to attack certain maize genotypes, and HV-

-

ROGER P. WISE ETAL.

94

Table I Genbank Accession Numhers for cms-T Related Gene Sequences Definition Genbank Acc. No. Reference Comment Definition Genbank Acc. No. Reference Definition Genbank Acc. No. Reference Comment Definition Genbank Acc. No. Reference Definition Genbank Acc. No. Reference Comment Definition Genbank Acc. No. Reference Comment Definition Genbank Acc. No. Reference Comment Definition Genbank Acc. No. Reference Comment Definition Genbank Acc. No. Reference Comment

Maize (cms-T) mitochondrial TURF 2H3 sequence containing 2 ORFs M12582 Dewey et al. (1986) Complete sequence of T-urfl3 and the co-transcribed orf221 Zea mays mitochondrial T mutant ORFl3 DNA M16366 Wise et al. (1987a) Zea mays mitochondrial mutant T-4 ORF8.3 DNA M 16268 Wise et al. (1987a) T-4 mutant has TCTCA tandem duplication in T-urfl3 reading frame CMS T maize mitochondrial DNA for T-URF25 N-terminus X0.5446 Rottmann er al.(1987) CMS T fertile revertant mit-DNA for T-URF25 N-terminus X05447 Rottmann et al.(1987) V3 mutant has lost T-urfZ3 via homologous recombination Maize B73 cms-T mitochondrion mRNA X60239 Ward and Levings (1991) orf221 but not T-urfl3 is subject to RNA editing Maize B73 X M017 mitochondrial mRNA X60238 Ward and Levings (1991) orf221 but not T-urfl3 is subject to RNA editing Zea mays T-cytoplasm male-sterility restorer factor 2 ( r f 2 ) mRNA U43082 Cui et al. (1996) Significant similarity to mitochondrial ALDHs Cochliobolus heterostrophus polyketide synthase (PKSI ) gene U68040 Yang et al. (1996) Currently designated ChPKSZ

toxin, which gives C. victoriae the ability to cause disease on certain genotypes of oats (Yoder et al., 1997). Thus, a healthy fear of this fungus is not unfounded. In fact, in 1988, researchers in China startled the corn-breeding community by announcing the discovery of a race of C. heterostrophus highly virulent to C cytoplasm (Wei et al., 1988). They reported that this new race, which they called “race C,” produced a toxin specific for C cytoplasm. Fortunately, these claims have not been confirmed by other laboratories. In fact, in experiments using the same fun-

T-CYTOPLASM MALE STERILITY INMAIZE

95

Figure 9 Polyketide synthase genes from C. heterostrophus and M. zeae-maydis. ChPKSI, a gene encoding a polyketide synthase involved in T-toxin production in C. hererostrophus race T. MzPKSI, a gene encoding a polyketide synthase involved in the production of PM-toxin in M. zeae-muydis. KS, f3-ketoacyl synthase; AT, acyltransferase; DH, dehydratase; ER, enoyl reductase; KR, P-ketoacyl reductase; ACP, acyl carrier protein. MzREDl and MzRED2 are reductases. Mztl through Mzr5 are transposons. Vertical arrows indicate the location of introns. Adapted from an unpublished drawing by B.G. Turgeon and 0. C. Yoder, Cornell University.

gal isolates as used by Wei et al., no differences in either virulence or toxin production were detected between “race C” and authentic race 0 isolates (0.C. Yoder, personal communication). Thus, to date, there is no evidence that C. heterostrophus has evolved the ability to attack other maize cytoplasms.

C. YELLOWLEAFBLIGHTAND

THEINSECTICIDE M~THOMYL

The popularity of T cytoplasm in the 1960s resulted in the selection of another fungus to which T-cytoplasm hybrids were susceptible. The fungus, Mycosphaerella zeae-maydis, was unknown as a pathogen on maize in the United States until 1968, when severe outbreaks of yellow leaf blight were reported in the northern maize-growing states and Ontario, Canada (Scheifele and Nelson, 1969; Scheifele et al., 1969; Ayers et al., 1970). Susceptibility was shown to be cytoplasmically inherited. Maize with N cytoplasm or forms of male sterility other than cms-T are resistant. The virulence of M . zeae-maydis is due to its production of a host-specific toxin, known as PM-toxin (after Phyllosticta maydis, the name applied to the asexual stage of the fungus) (Yoder, 1973). The fungus has no known races and all known naturally occurring isolates of M. zeae-muydis produce the toxin. Toxin-deficient mutants have been created by REMI mutagenesis and shown to be nonpathogenic on maize (Yun et al., 1998). PM-toxin is strikingly similar to T-toxin, in that it is a family of linear P-polyketols ranging in length from C,, to C,, (Fig. 7; Kono

96

ROGER P. WISE ETAL.

et al., 1983, 1985; Danko et al., 1984; Suzuki et al., 1988). Conformational studies suggest that PM-toxin assumes a helical or hairpin form in its native state (Suzuki et al., 1988). Three genes have been shown to be required for both the production of PM-toxin and virulence to T-cytoplasm maize (Yun et al., 1997; Yoder, 1998; 0. C. Yoder, personal communication).MzPKSl encodes a polyketide synthase and is similar to ChPKSl from C. heterostrophus in having six enzymatic domains and four introns (Fig. 9). However, MzPKSI lacks the A+T-rich flanking regions and has only 59% nucleotide identity to ChPKSI. In addition, immediately 5’ of MzPKSl areMzRED1 and MzRED2. MzREDl and MzRED2 have high similarity to the ketoreductases associated with plant and bacterial polyketide synthases. Disruption of MzPKSI, MzREDI, or MzRED2 results in loss of PM-toxin production. The differences in the organization and nucleotide sequences of genes involved in P-polyketol synthesis between M. zeae-maydis and C. heterostrophus suggest that the ability to make PMtoxin and T-toxin evolved separately. Thus, it was chance that two, unrelated fungi produced such similar compounds (Yun et al., 1997; Yoder, 1998). T-cytoplasm maize is also susceptible to damage from S-methyl-N-[(methylcarbamoyl)oxy]thioacetimidate (methomyl), the active ingredient in Lannate insecticide (Fig. 7). susceptibility was discovered accidentally in 1974 when a field containing T-cytoplasm lines was sprayed with Lannate. Methomyl caused necrotic bands on leaf tips and margins and, in severe cases, complete leaf necrosis. Because all other cytoplasms of maize are unaffected, methomyl treatment is a rapid and convenient method for identifying T-cytoplasm plants (Humaydan and Scott, 1977).

D. T-Tom AND P M - T o m DISRUPT MITOCHONDRIAL FUNCTION Consistent with the maternal inheritance of disease susceptibility in T-cytoplasm lines, mitochondria of T-cytoplasm lines are highly sensitive to T-toxin and PM-toxin. The toxins have a wide range of effects on tissues of T-cytoplasm plants, including promotion of ion leakage from cells, induction of stomata1closure, and inhibition of root growth, transpiration, photosynthesis, and dark CO, fixation (Arntzen et al., 1973a,b; Bhullar et al., 1975). However, the most direct effects are on mitochondrial function. Effects of the toxins on mitochondria include uncoupling of oxidative phosphorylation, inhibition of malate oxidation, stimulation of NADH and succinate oxidation, swelling and loss of matrix density (Miller and Koeppe, 1971;Gengenbach et al., 1973; Peterson et al., 1975; Aldrich et al., 1977; Hack et al., 1991). All of these effects are consequences of increases in mitochondrial membrane permeability, as evidenced by leakage of NADS., Ca2+, and H+ (Bervillt et al., 1984; Holden and Sze, 1984, 1987). The effect of the toxins on

T-CYTOPLASM MALE STERILITY IN MAIZE

97

membrane permeability can be detected visually with the florescent dye DiOC,, which is sensitive to mitochondrial cationic potential (Figs. 10A-D, see color plate; Yang, 1989). Methomyl, although structurally unrelated to T-toxin and PMtoxin (Fig. 7), induces the same physiological responses in T-cytoplasm mitochondria. The damage to mitochondrial function induced by the fungal toxins and methomyl is believed to be the cause of cell death. The presence of restorer genes in the nuclear genome modifies the response of maize, but not enough to eliminate disease susceptibility. Leaves, roots, and isolated mitochondria from male-sterile plants are highly sensitive to T-toxin; those from plants restored to fertility show an intermediate reaction; while those from N-cytoplasm plants are insensitive (Watrud et al., 1975a; Barratt and Flavell, 1975). The intermediate response of lines restored to fertility is clearly insufficient to confer resistance to C. heterostrophus and M . zeae-maydis, however, since the 1970 epidemic occurred on restored maize (Ullstrup, 1972).

E. THEHAZARDSOF GENETIC HOMOGENEITY Although it has been over 25 years since the southern corn leaf blight epidemic, its repercussions are still being felt. The susceptibility of T-cytoplasm lines to race T of C. heterostrophus has severely limited the use of cms-T as a source of male sterility for commercial maize production. The vast majority of hybrid seed is now produced at considerable expense by hand or machine emasculation. The most significant impact of the epidemic, however, has been the increased public, political, and scientific awareness of the potentially devastating consequences of genetic homogeneity. The southern corn leaf blight epidemic was highly publicized in both print and electronic media (Ullstrup, 1972); most well-informed individuals in the United States in 1970 would have been aware of its occurrence and the effects it was having, either directly or indirectly, on their lives. Fear of similar pandemics in the future motivated sober appraisals of the extent of genetic homogeneity in crop plants and a wide range of studies on ways to increase genetic diversity in modern agricultural systems (Day, 1977; Mundt and Browning, 1985; Wolfe, 1985).

V. DISEASE SUSCEPTIBILITY AND MALE STERILITY In many species, cms is associated with the expression of novel open reading frames in the mitochondrial genome (Schnable and Wise, 1998). Although each open reading frame is unique, the feature that these open reading frames have in common are one or more predicted hydrophobic domains, consistent with mem-

98

ROGER P. WISE ETAL.

Figure 11 Structure of T-urfl3 region in the mitochondria1genome of T-cytoplasm maize. Shown above the chimeric T-urfl3/0$221 transcriptional unit are the presumed progenitor sequences for these open reading frames. A 5-kb T-specific repeated region contains promoter sequences 5‘ to afp6. There are 15 nucleotides of unknown origin between the 5’ alp6 flanking region and the start of the 3’ rm26 flanking region. Sixty-nine nucleotides 3’ to the T-specific, 5-kb repeat junction is the T-urfl3 ATG start codon. The co-transcribed T-urfl3 and orJ221 reading frames most likely utilize a duplication of the arp6 promoter within the 5-kb repeat region.

brane-bound proteins. In cms-T, both male sterility and toxin sensitivity are attributed to T-urf13, a 345-bp reading frame that encodes a membrane-bound 13kDa protein found only in T-cytoplasm maize mitochondria (Fig. 11; Levings, 1993). One of the first indications that male sterility and toxin sensitivity are controlled by the same gene came from attempts to select for toxin-insensitive plants from tissue cultures of toxin-sensitive, cms-T maize (Gengenbach and Green, 1975; Gengenbach et al., 1977; Brettell et al., 1979). Cultures were formed from immature embryos of the T-cytoplasm line, A 188T. Tissue cultures from several cell lines were treated with a sublethal dose of T-toxin, and then passed through several more cycles with higher concentrations of the toxin. The survivors of these treatments were regenerated to yield T-toxin-resistant plants. Interestingly, all of the resulting toxin-resistant mutants were also male fertile and both of these traits were maternally inherited. Toxin-resistant mutants were also obtained from tissue cultures without T-toxin selection. All of these plants were also male fertile (Brettell et al., 1980; Umbeck and Gengenbach, 1983). Because all of the tissue-culture derived mutants resulted in the coordinate change in phenotype from toxin sensitivity to insensitivity and male sterility to male fertility, it was suggested that these two traits were controlled by the same or tightly linked mitochondrial genes (Gengenbachet al., 1981; Kemble et al., 1982; Umbeck and Gengenbach, 1983).

T-CYTOPLASM MALE STERILITY IN MAIZE

99

The gene controlling both traits, T-urjl3, was identified by two complementary methods. One approach took advantage of differential expression patterns of mtDNA regions between the N and T cytoplasms. A mtDNA region consisting of three contiguous Hind111 restriction fragments, designated TURF 2H3, revealed an abundant family of transcripts that was T specific. Sequence analysis of the 3547-bp TURF 2H3 revealed two open reading frames, T-urfl3 and the co-transcribed o f l 2 I (Table I; Dewey et al., 1986). An alteration in the accumulation of T-urfl3-specific transcripts was shown to occur in plants restored to fertility; subsequent genetic analysis identified the Rfl nuclear restorer gene as responsible for these alterations (Dewey et al., 1987; Kennell et al., 1987; Kennell and Pring, 1989; Wise et al., 1996). Concurrently, a map-based approach took advantage of the male-fertile, toxin-insensitive mutants described above. In 19 of 20 such mutants, a 6.7-kb XhoI mitochondrial DNA restriction fragment was altered due to a partial deletion (Gengenbach et al., 1981; Kemble et al., 1982; Umbeck and Gengenbach, 1983; Fauron et al., 1987). The remaining male-fertile mutant, T-4, retained the 6.7-kb XhoI fragment. Comparative restriction and sequence analysis of the 6.7-kb XhoI fragment from cms-T and T-4 revealed a tandem TCTCA repeat in the T-ulf13 coding region of the T-4 mutant. This tandem 5-bp repeat generates a frameshift, truncating the reading frame at nucleotide 222 (Wise et al., 1987a). In the deletion mutants examined, the T-urjl3 reading frame was excised via recombination through a 127-bp repeat (Rottmann et al., 1987; Fauron et al., 1990). The 127-bp repeat carries a 55-bp conserved core with 85% similar flanking regions and begins 6 bp 3' to the T-urfl3 TGA stop codon. Although the 127-bp repeat extends 56 bp into the co-transcribed orj921, this reading frame is unaltered in the deletion (Rottmann et al., 1987; Fauron et al., 1990) and T-4 frameshift mutants (Rocheford et d., 1992), providing rigorous genetic evidence that it is the T-ulfl3 gene that is responsible for both toxin sensitivity and male sterility. The physical organization of the T-urfl3/ofl21 complex appears to be the result of numerous recombination events among the 5' flanking region of atp6 (contained within the T-specific 5-kb repeat), the coding and 3' flanking region of rrn26, the cholorplast trnA gene, and several unidentified sequences (Fig. 11; Dewey et al., 1986; Wise et al., 1987a). Interestingly, the insertion of the tandem TCTCA repeat in the T-urfl3 coding region in the T-4 mutant creates 86 nucleotides of perfect identity with the 3' flanking region of the rrn26 sequence (Dale et al., 1984), consistent with the hypothesis that T-4 also arose by homologous recombination or gene conversion with similar sequences in the mitochondrial genome (Wise et al., 1987a; Pring et al., 1988). T-urfl3 encodes a 13-kDa mitochondrial protein (URF13), located in the mitochondrial membrane (Dewey et al., 1987). Initially identified by Forde et al. (1978), the identity of this protein was confirmed by immunoprecipitation and Western blot analysis with antiserum prepared against synthetic peptides derived from the predicted amino acid sequence of the T-urjl3 reading frame (Dewey et

100

ROGER P. WISE ETAL.

Figure 12 Irnmunocytochernical labeling with 20-nm gold particles (black dots) of URF13 protein associated with inner membrane of mitochondria in all cells of crns-T maize plants. (A) Portion of root-tip cell. (B) Portion of tapetal cell at early microspore stage. Bars = 0.6 pm (A) and 0.3 pm (B). There is no labeling of N-cytoplasm mitochondria (not shown).

al., 1987; Wise et al., 1987b). This protein is not synthesized by deletion mutants (Dixon et al., 1982), and is truncated to approximately 8 kDa in the T-4 frameshift mutant (Wise et al., 1987b). Immunogold localization further confirmed that U W l 3 is localized primarily in the mitochondrial inner membrane of mitochondria isolated from etiolated shoots, roots, and tapetal cells from the meiocyte to microspore stages (Figs. 12A and 12B; Yang, 1989; Hack et al., 1991).

A. How FUNGAL Toms DISRUPT MITOCHONDRIAL FUNCTION In addition to the genetic evidence described above, biochemical evidence that T-urfl3 is responsible for toxin and methomyl sensitivity comes from expression of the UW13 protein in heterologous systems, Escherichiu coli, yeast, and tobacco. When T-urfl3 is transformed into E. coli in an inducible expression vector, cells that express UW13 are sensitive to toxin and to methomyl. The E. coli cells not expressing URF13 or expressing a truncated version of UW13 are not sensitive to these compounds (Dewey et al., 1988). Expressed URF13 is associated with the E. coli plasma membrane, similar to its association with the mitochondrial membrane in T-cytoplasm maize. Likewise, the effects of toxin and methomyl treatment on T-urfl3-E. coli are similar to those previously observed from isolated T-cytoplasm mitochondria, including inhibition of whole cell respiration, in-

T-CYTOPLASM MALE STERILITY INMAIZE

101

duction of massive ion leakage, and swelling of spheroplasts (Miller and Koeppe, 1971; Barratt and Flavell, 1975; Matthews et al., 1979; Payne ef al., 1980; Holden and Sze, 1987; Dewey et al., 1988). The availability of E. coli expressing URF13 has resulted in the development of a rapid microbiological assay for Ttoxin and PM-toxin, thus facilitating research on toxin production by C. heterostrophus and M.zeae-maydis (Ciuffetti et al., 1992). Extensive site-directed mutagenesis experiments of the T-urj-13 sequence have provided insight into the structure-function relationships of URFl3 and toxin sensitivity (Braun er al., 1989). These studies demonstrated that approximately onequarter of the amino acids from the carboxyl end of the URF13 protein are not essential for toxin sensitivity. In maize, URF13 is truncated from 113 to 74 amino acids in the T-4 toxin-insensitive, frameshift mutant (Wise ef al., 1987b). Similarly, E. coli that has been transformed with T-urj-13mutations that result in truncated URF13 proteins of 73 or 82 amino acids in length are toxin insensitive, whereas E. coli with truncated URF13 proteins of 83 amino acids or longer are toxin sensitive. N,N’-Dicychlohexylcarbodiimide(DCCD) protection studies have revealed specific amino acid residues that are important in the URFl3/T-toxin interaction (Braun et al., 1989). DCCD binds covalently to acidic amino acids localized in hydrophobic environments. Hence, T-cytoplasm mitochondria or E. coli cells expressing URF13 preincubated with DCCD are protected against T-toxin. Use of site-directed substitutional mutations demonstrated that DCCD binds to the URF13 protein at two aspartate residues at positions 12 and 39. However, it is the aspartate at position 39 that is required for a functional URF13/T-toxin interaction because substitutional mutations at this position eliminate toxin sensitivity. [3H]PM-toxinbinds to wild-type URF13 in E. coli, but not to the 82-amino-acid truncated URF13 (Braun et al., 1990). This is consistent with the toxin sensitivity phenotypes in cms-T maize and insensitivity in the T-4 mutant. Additionally [3H]PM-toxin does not bind to the E. coli toxin-sensitive URF13 valine substitutional mutant at position 39. However, in the E. coli toxin-insensitive mutant missing amino acids 2 through 1I, the presence of significant toxin binding indicates that it is possible to be toxin insensitive even in the presence of toxin binding. Based on this finding, it was proposed that amino acids 2 through 11 are unnecessary for toxin binding, but essential for the toxin-URFl3 interaction that leads to membrane permeabilization. Hence, the binding to URF13 in E. coli cells is cooperative, suggesting that URF13 exists as oligomers in E. coli membranes (Braun et al., 1990). Based on the additional studies of Korth et al. (1991) and Rhoades et al. (1994), it has now been postulated that URF13 monomers containing three trans-membrane a-helices interact with T-toxin (or methomyl) and are assembled as tetrameric pore-forming structures spanning the inner mitochondria1 membrane. This pore-forming model is consistent with early observations on mitochondria] electrolyte leakage, uncoupling of oxidative phosphorylation, and chan-

102

ROGER P.WISE ETAL.

nel formation in the presence of T- and PM-toxins (Matthews et al., 1979; Bervil16 et al., 1984; Holden et al., 1985). Corroborating results were obtained when T-urfl3 was transformed into the yeast Saccharomyces cerevisiae and targeted to the mitochondrial membrane utilizing a Neurospora ATP synthase subunit 9 leader sequence (Huang et al., 1990; Glab et al., 1990, 1993). Yeast cells expressing T-urf13 were sensitive to both Ttoxin and PM-toxin, in addition to methomyl. These compounds inhibited growth of yeast cells and stimulated respiration by isolated mitochondria with NADH as a substrate, similar to the effects seen on isolated maize mitochondria. Yeast cells expressing URF13 without a targeting peptide do not exhibit sensitivity to the toxins or methomyl, demonstrating that localization of URF13 to mitochondria is essential for sensitivity in yeast. Similar effects are observed when the URF13 protein is expressed in tobacco. For example, it is essential that URF13 be targeted to mitochondria to cause toxin or methomyl sensitivity. Tobacco cells expressing URF13 without a targeting peptide do not exhibit sensitivity to the toxins or methomyl (von Allmen et al., 1991; Chaumont et al., 1995). However, none of the toxin-sensitive,transgenic tobacco plants produced to date has been male sterile. The failure to obtain malesterile plants suggests that correct tissue-specific expression and subcellular localization is required for URF13 to cause male sterility.

B. How URF13 CAUSES STERILITY R E ~A MYSTERY S Although the interaction of pathotoxins or methomyl with URF13 causes significant damage to URFl3-containing membranes, the mechanism by which URF13 causes tapetal degeneration and male sterility in the absence of pathotoxins and methomyl is still not clear. In particular, the selective nature of this degeneration is paradoxical, because URF13 is expressed in many, if not all, maize tissues (Hack et al., 1991). Two explanations for this paradox have been summarized by Levings (1993). One possibility (Flavell, 1974) is that there is a tapetumspecific, toxin-like compound (Factor X) that is a necessary prerequisite for URF13induced toxicity (Fig. 5). Consistent with this hypothesis, it is well established that mitochondrial proteins can be regulated in tissue-specific ways (Bedinger, 1992; Conley and Hanson, 1994). Alternatively,Wallace (1989) has proposed that tissuespecific degeneration could occur if tissues differ in their requirements for mitochondrial function. Microsporogenesis seems to be an energy-intensive process; the number of mitochondria undergoes a 40-fold increasein the tapetal cells (Lee and Warmke, 1979). In addition, antisense repression of the mitochondrial tricarboxylic acid (TCA) cycle enzyme, citrate synthase, results in the degenerationof ovary tissues of potato, apparently without affecting other organs (Landschiitze et al., 1995).Hence, reproductive tissues seem to be especially sensitive to perturbations in mitochondrial function. Consistent with the hypothesis that URF13 can cause

T-CYTOPLASM MALE STERILITY IN MAIZE

103

subtle perturbations in mitochondrial function even in the absence of pathotoxins and methomyl, there is some evidence that URF13 has a slight (but significant) deleterious effect on maize cells (Duvick, 1965; Pring et aL, 1988) (see also discussion by Duvick in Section VIIa). Hence, for the Wallace hypothesis to be correct, it would only be necessary that these perturbations in mitochondrial function be particularly deleterious to tapetal cells.

VI. NUCLEAR-CYTOPLASMIC INTERACTIONS AND RESTORATION OF CMS-T T-cytoplasm lines can be restored to male fertility by the action of dominant alleles of the rfl and rj2 nuclear genes. These genes are thought to suppress or compensate for cytoplasmic dysfunctions that are phenotypically expressed during pollen development.The mode of fertility restoration of cms-T is sporophytic; the genetic constitution of the diploid, sporophytic anther tissue, rather than the haploid, gametophytic pollen, determines pollen development. Therefore, a cms-T plant that is heterozygous for both restorer gene loci ( R f l h f l , Rj2/rj2), will produce all fertile pollen even though only one-fourth of the pollen grains carry both Rfl and Rf2 (Laughnan and Gabay-Laughnan, 1983). Rfs and Rf* are newly described restorers that can each (at least partially) substitute for Rfl (Dill et al., 1997). Such overlapping functions are either a consequence of duplications of gene function or an indication that multiple mechanisms can induce restoration. The allelic frequency of restorers in the maize gene pool can provide clues as to the origins and functions of these restorers. For example, the Rfl. Rf8, and Rf* alleles are all rare among maize lines (Table 11). This suggests that they are probably neomorphic alleles and do not have essential functions other than their fortuitous ability to influence restoration of cms-T. In contrast, the RF2 protein probably has a physiological function independent of its role as a nuclear restorer of cms-T (Schnable and Wise, 1994). The argument for this is that most maize lines have never been exposed to cms-T, but carry a functional Rj2 allele. This finding indicates that there has existed selective pressure for the RF2 protein that is independent of cms-T. Based on this reasoning it has been proposed that the RF2 protein was recruited in T-cytoplasm maize to ameliorate the mitochondrial lesion associated with T-urfl3 expression (Schnable and Wise, 1994).

A. &I,

BUT NOTRF2, ALTERSTHEEXPRESSION OF T-urn13

In many species, restoration of fertility is associated with the processing and possible post-transcriptionalediting of cms-associated mitochondrial transcripts. Typical examples of this are observed in oilseed rape, Petunia, common bean,

ROGER P. WISE ET AL.

104

Table I1 Restorer Genotypes in T-CytoplasmMaize Lines Line A188 A632 B37 B73 C103 33-16 38-1 1 8703” Ky2 1 Line C M017 Mo17 X B73 N6 R213 W23 W64A Wf9 Wf9-BG wxl-m8”

Restorer genotypes” 61

61 61 61 61 61 61 61 Rfl rfl 61 61 61 Rfl 61 61 61 Rfl 61

Rf2 NDc Rf2 Rf2 ND Rf2 R f2 Rf2 Rj2 Rj2 R.f2 Rf2 Rf2 62 Rt2 R.f2 62 Rf 2 Rf 2

68 $8 rf8 68 68 68

ND Rf 8” $8

68

rfs rf8 ND

6* rf*

6* rf*

6*

Rf * ND

6* 6* 6% rf * 6*

ND

68

6*

ND

ND

68 $8 $8 Rf8”

6* 6* rf*

rf*

OData derived from Dewey et a/. (1987). Kennell el al. (l987), Kennell and Pring ( 1 989). Rocheford et a!.(1 992), Gabay-Laughnan and Laughnan (1994). Wise er al. (1996). Dill et (I!. (1997). and P.S. Schnable and R. P. Wise (unpublished results). Unless otherwise indicated all lines are homozyous for the indicated allele. %ese stocks are not inbred lines, they segregate at many loci, including $8. “ND, no data.

rice, and sorghum, as well as cms-S and -T maize (reviewed by Schnable and Wise, 1998). In cms-T maize, seven major transcripts of T-urfl.3 and the co-transcribed orj221 range in size between 1.0 and 3.9 kb (Dewey et al., 1986; Kennell et al., 1987; Kennell and Pring, 1989; Wise et al., 1996; Dill et al., 1997). Many of these transcripts are presumably products of a series of processing events stemming from the 3.9-kb transcript. However, transcript capping experiments with guanylyl transferase identified the 1.85-kb transcript as a primary (initiated) transcript; hence, the 1.85-kb transcript is not the result of RNA processing (Kennell and Pring, 1989). Plants segregating for the Rfl restorer accumulate additional 1.6- and 0.6-kb T-urj13 transcripts as shown in Fig. 13 and Table I11 (Dewey et al., 1986, 1987; Kennell etal., 1987; Wise et al., 1996). Mitochondria1 RNAgel blot analyses have

rfl-m7212 I+ rf x Rfl R I RT rfI-B37 :I+ 7 rfl ,sy+

A

* rfl m3207

I 3+

P-

Rfl

-

La3

-

rflm3207 103+ rflm3207 lg3+

Figure 13 Effect of different restorer genotypes on the accumulation of T-urfl3 mitochondria1 transcripts. F, "F",and S denote male-fertile, partially fertile, and male-sterile plants, respectively. (A) Lanes 1-8 contain RNA from selfed progeny of plants carrying the mutant allele, rfl-m3207. The RNA in lanes 1-6 originated from plants that carried a dominant Rf/ allele, which is tightly linked to the morphological marker Lg3. Rfl mediates the processing of the larger 2.0-, 1.8-. and 1 .O-kb T-urj'I3 transcripts, resulting in the accumulation of 1.6- and 0.6-kb transcripts. As can be seen in lanes 7 and 8, RNA from plants homozygous for the rjl-m3207 mutant allele are deficient in these processing products. (B) RNA was isolated from progeny of a cross involving the rf/-m72/2 allele. Plants that carry the dominant Rf/ allele accumulate 1.6- and 0.6-kb transcripts, whereas those homozygous for the t f m7212 allele do not. However, all of the plants in this family carry the unlinked Rf* allele, which also mediates the processing of the larger 2.0-, 1.8-, and 1 .O-kb transcripts, but results in the accumulation of I .4- and 0.4-kb transcripts. Plants homozygous for rfl-m72/2 and that accumulate the 1.4-kb transcript are partially fertile or sterile (lanes 3-5) because of the weakly penetrant nature of the RJy restorer (Dill et al., 1997).Transcripts were detected with the T-urfl3 specific probe T-st308. Modified and published with permission from the Genetics Society of America, from Wise et a / . (1996).

106

ROGER P. WISE ETAL. Table III Effect of Nuclear Restorers on T-urfl3-Associated Mitochondria1 Transcripts

Restorer gene Rfl Rf 8 Rf *

Proposed mechanism

Additional accumulation of Rf-associated T-urfl3 transcripts

Concurrent effect on novel proteins

Transcript processing Transcript processing Transcript processing

+ 1.6, 0.6 + 1.42,0.42 + 1.4.0.4

Reduction of 1 3 - D a URF13 protein Reduction of 1 3 - D a URF13 protein Unknown

indicated that four independent mutant alleles of rfl isolated by Wise et al. (1996) condition reduced steady-state accumulations of the 1.6- and 0.6-kb T-urfl3 transcripts. This result demonstrates that these transcripts are Rfl dependent. Results from the guanylyl transferase capping experiments described above indicate that the Rfl-dependent T-urfl3 transcripts do not arise via novel initiation sites, but instead arise via processing from the transcripts synthesized and processed in all cms-T lines (Kennell and Pring, 1989). Even in the presence of Rfl, however, the steady-state accumulation of the larger T-urfl3 transcripts is not measurably reduced (Fig. 13). Similarly, novel 1.42- and 0.42-kb T-urfI3 transcripts accumulate in plants segregating for the partial fertility restorer, Rf8, and 1.4- and 0.4-kb Turf13 transcripts accumulate in plants segregating for Rf*. Extensive mapping of these transcripts via primer extension and Northern blot analyses indicates that the Rfl -dependent 1.6-kb, the Rf8-dependent 1.42-kb, and the Rf*-dependent 1.4-kb transcripts are all derivatives of the 1.8-,135, or 2.0-kb transcripts. Likewise, the smaller 0.6-, 0.42-, and 0.4-kb transcripts are likely derivatives of the 1.0-kb T~$13transcript (Fig. 14; Kennell er al., 1987;Kennell and Pring, 1989; Rocheford et al., 1992; Wise et al., 1996; Dill et al., 1997). Like the majority of the T-urfl3 and orj221 transcripts, atp6 transcripts are most likely initiated from identical promoter sequences within the T-specific 5-kb repeat (Fig. 11). However, differential processing of afp6-specific transcripts does not seem to occur in the presence of Rfl, Rf8, or Rf* (Kennell et al., 1989; Dill et al., 1997). As described in the next section, this may indicate sequence specificity in the processing events mediated by these restorer genes. Post-transcriptional RNA editing may also play a role in fertility restoration in some cms systems (Iwabuchi et al., 1993). Possibly, editing might shorten predicted cms-associated ORFs by creating UAA, UAG, and UGA stop codons, because the most prevalent example of editing in plant mitochondria1sequences is C-to-U (Hanson et al., 1996). Although the co-transcribed 0@21 is edited posttranscriptionally in cDNAs from cms-T mitochondria, T-urf13 transcripts are not. Additionally, no differences in editing are observed between the 013221 mRNA isolated from T-cytoplasm mitochondria and mRNA isolated from mitochondria

T-CYTOPLASM MALE STERILITY IN MAIZE

Mitochondrion

I

T-specific 5-kb repeat I X

H Hc

I I

107

H H

H

St

2.0,1.85,1.8 1.0 1.6 0.6 1.42 0.42 1.4 0.4

Figure 14 Effect of nuclear Rfgenes on T-urfl3 transcnpt processing. Rfl, Rfi, and Rf* each mediate processing of T-urfl3 transcripts at specific sites within T-urfl3 reading frame. Diagrammed (horizontal lines terminated by arrowheads or filled circles) are novel T-urfl3 transcripts that accumulate in the presence of the three nuclear restorers. Numbers in parentheses indicate 5’ termini (designated by the number of nucleotides 3’ of the start of the T-urfI3 reading frame) of each transcnpt as determined by primer extension experiments (Dill et ul., 1997). To the right of each group of transcripts are their respective molecular sizes as determined by Northern analyses. a, A M , H, HindIII; Hc, HincII; s, Suu3a; St, SstII; t, TayI; X, XhoI.

of plants restored by Rfl, thus, apparently ruling out the influence of RNA editing on restoration of cms-T maize (Table I; Ward and Levings, 1991).

1. Rf-Mediated T-ufl? Transcript Processing Sites Share a Conserved Motif Comparison of the sequences encompassing the 5’ termini of the Rfl-, Rf8-, and Rfx-dependent T-urfl3 transcripts and the Rf3-associated orf107 transcript from the A3 cytoplasm of crns sorghum has revealed a small conserved motif associated with fertility restoration and transcript processing (Fig. 15). The conserved motif 5’-CNACNNU-3’ overlaps a 5 ’-terminal U of each of these restorer-dependent transcripts. Notably, overlapping the 5’-terminal U of the Rfx-associated transcripts and the Rf3-associated orfl07 transcript from crns sorghum is the highly conserved sequence 5 ’-AC(U/C)ACAAUA-3‘, revealing striking similarities among restorer-associated processing sites in these two grasses (Tang et al., 1996; Dill et al., 1997). This mitochondria1 sequence may represent a recognition site for proteins regulated or expressed by the nuclear restorer genes. Thus, it appears that the Rfl, Rf8, and Rf* restorers may encode functionally similar gene products that mediate the specific modification of T-urf13 transcripts synthesized in all T-cytoplasm lines (Fig. 16). Motifs have been described in other crns systems, such as

108

ROGER P. WISE ET AL.

Figure 15 A conserved motif associated with fertility restoration and transcript processing. The sequence 5'-CNACNNU-3' overlaps the 5' termini of the 1.6- and 0.6-kb (Rfl-associated), 1.42- and 0.42-kb (Rj8-associated), I .4- and 0.4-kb (RF-associated), and the 380-nucleotide (Rjhssociated) transcripts from maize and sorghum. Letters highlighted by dark gray filled circles represent 5' termini of mtRNA transcripts associated with fertility restoration as determined by primer extension experiments (Tang et a/., 1996; Dill et a/., 1997). Conservation between Rj3- and Rf*-associated processing sites [AC(U/C)ACAAUA] is highlighted in the light-gray horizontal boxed area. The translation initiation codon for T-urfl3 is underlined. The line designated archaea 23s is a partial sequence from the pre-rRNA transcript of the 23s precursor processing stem from Suljolobus ucidoculdurius (Potter eta/., 1995); the encircled U and A have been shown to be in vivo processing sites. Modified and published with permission of the Genetics Society of America, from Dill et u/. (1997).

Figure 16 Molecular mechanisms by which restoration occurs are not well understood. Although the nuclear Rfl or Rj8 restorers are known to affect the accumulation of novel T-urfl3 transcripts in mitochondria, the mechanisms by which this occurs have not been established. The Rf2 restorer does not affect the accumulation ofT-urf13 transcripts. Although essential for restoration of cms-T, the molecular role of Rf2 in fertility restoration is not known. However, the recent discovery that this gene exhibits a high degree of sequence similarity to mammalian mitochondria1 aldehyde dehydrogenases has yielded a number of testable hypotheses. Adapted from an unpublished figure prepared by Xiangqin Cui, Department of Zoology & Genetics, Iowa State University.

T-CYTOPLASM MALE STERILITYIN MAIZE

109

the putative recognition sequence, 5’-UUGUUG-3’ within the orf224/atp6 transcriptional unit, downstream of the 5’-terminus of the Rfpl -polima-restorer-associated transcripts in Brassica napus (Singh et al., 1996). Yet, how or why rare nuclear (Rf) genes evolved to mediate recognition of distinct sites within specific transcripts is an unresolved question.

2. Rf-Mediated Processing of T-2463 Transcripts Is Accompanied by a Reduction in URF13 Rf-mediated processing of cms-associated mitochondria1 transcripts is often concurrent with a reduction in the accumulation of the cms-associated mitochondrial proteins. Although R J ~has no detectable effect on transcript processing or URF13 accumulation in maize (Dewey et al., 1987; R. P. Wise, C. L. Dill, and P. S. Schnable, unpublished results), the abundance of URF13 is reduced by approximately 80% in plants that possess the Rfl restorer (Forde and Leaver, 1980; Dewey et al., 1987). Areduction in URF13 can also be mediated by Rf8, however, the effect is not as pronounced as that mediated by Rfl, and may even be tissue dependent. For example, the difference in URF13 accumulation among plants segregating for Rf8 is much more evident in tassels as compared to ears. In contrast, plants that carry Rfl exhibit a marked decrease of URFl3 accumulation in both tassels and ears (Fig. 17; Dill et al., 1997). Rf* has not yet been tested for this capacity. The disparity that surrounds these observations is that there is a yet unknown mechanism for Rfmediated, URF13 reduction. As shown in Fig. 13, mature Turf13 transcripts still seem to be available for translation in the various Rf genotypes and an individual plant can accumulate an abundance of Rfl-, Rf8-, or Rf*associated transcripts, with no obvious decrease in the steady-state accumulation of the mature 2.0-, 1.8-, or 1 .O-kb transcripts (Wise et al., 1996; Dill el al., 1997). Therefore, the end result of Rfl and Rf8 function is likely post-transcriptional, that is, inhibition of translation or protein degradation, and not regulation of transcript accumulation. An excess of these processed transcripts might compete with the unprocessed transcript for translation, thereby reducing the accumulation of URFl3.

B. CLONING OF NUCLEAR RESTORERGENES As a step toward elucidating the molecular mechanisms by which restoration occurs, experiments were initiated to clone the rfl and rf2 genes via a transposon tagging’strategy (Figs. 18A and I8B). To assist in the generation of stocks necessary to transposon tag these restorer genes, the $2 locus on chromosome 3 (Duvick et al., 1961) and the rf2 locus on chromosome 9 (Snyder and Duvick, 1969) were positioned in reference to closely linked R E P and visible markers (Fig. 19; Wise and Schnable, 1994). Subsequently, four rfl-rn and seven $2-rn alleles were

110

ROGER P. WISE ETAL.

Rf8-8703 rf8-wx-rn8

x

rf8-W64A rf8- W64A

Figure 17 Effect of Rfl and Rfl on the accumulation of URF13 mitochondrial protein. Plants from families segregating for Rfl display a decrease in URF13 as compared to $1-containing plants, although variability has been observed among genotypes. Similarly, within families segregating for Rfl, three plants (lanes 3,6, and 7) that carried the Rfl allele accumulated less URF13 protein than two siblings that did not carry this restorer (lanes 4 and 5 ) (Dill et al., 1997).All plants, regardless of $3 genotype, accumulated less URF13 in tassels than in ears (data not shown). In addition, the difference in URF13 accumulation among plants segregating for Rfl was much more evident in tassels than in ears. In contrast, plants that carried Rfl exhibited a marked decrease of URF13 accumulation in both tassels and ears. Detection of URF13 was performed with an URF13 monoclonal antibody (a gift from C. S. Levings III, North Carolina State University) using goat anti-mouse IgG (H + L) alkaline phosphatase detection. Modified and published with permission from the Genetics Society of America, from Dill et al. (1997).

isolated from populations of 123,500 and 178,300 transposon-bearing plants, respectively (Schnable and Wise, 1994; Wise et al., 1996). To identify DNA fragments containing if1 and rf2 sequences, DNA gel-blot cosegregation analyses were performed. Mul-hybridizing EcoRI restriction fragments were identified that segregated absolutely with male sterility in large families segregating for the transposon-induced rfl -m3207 and $1 -m3310 alleles, suggesting that they contained Mu transposon insertions in the Rfl gene (Wise et al., 1996).These candidate $1 DNA fragments were cloned and corresponding cDNAs isolated (R. P. Wise, K. S. Gobelman-Werner, D. Pei, C . L. Dill, P. S. Schnable, unpublished results). However, further experimentation is necessary to establish whether these sequences are part of the $1 gene. Concurrently, a 3.4-kb Mu1 -hybridizing EcoRI-Hind111 restriction fragment

T-CYTOPLASM MALE STERILITY INMAIZE

111

-

Transposon tagging of Rf genes Functional gene = Fertile tassel

Mu'3

a Insertion mutant = Sterile tassel

Figure 18 Transposon tagging of Rf genes. (A) An insertional mutagenesis strategy employed to generate transposon-tagged mutants of the rfl and rfz genes. The following two crosses were used to generate large populations of plants that were screened for their inability to restore pollen fertility to cms-T maize (Schnable and Wise, 1994;Wise et ul., 1996).Cross 1: Screen for $1-m alleles (123,500 plants): T Rfl/Rfl, RfZ/RfZ (Murutor) X N rfl/rfl, Rfz/RfZ (inbred B37). Cross 2: Screen for rfz-m alleles (178,300plants): T R f l / R f l , Q/rj2 (inbred R213) X N rfl/rfl, RfZ/Rf2 (Mututor; Cy. Spm). These screenings resulted in the isolation of four heritable rfl-m alleles and seven heritable tjC2-m alleles. Subsequently, a cloned Mu1 transposon was used as a probe to recover the candidate rf-m DNA sequences. (B) A male-sterile rfl-m/rfl-B37plant identified by detasselling all male-fertile plants resulting from Cross 1 .

ROGER P. WISE ETAL.

112

Chromosome 3

--

Chromosome 9

WXl

umc92

bn15.10 umcl53

9.7

rf2 5.8

bn16.06, rgl 5.6

susl 1.3

--

umc95

bn15.37

Figure 19 Consensus genetic maps of the rfl and @ regions of chromosomes 3 and 9, respectively (Wise and Schnable, 1994; Wise et al., 1996). The rfl locus and/or linked visible markers were mapped in seven populations consisting of 729 individuals; the $2 locus was mapped in three populations consisting of 304 individuals. Data from individual populations were joined with the aid of JoinMap software. The resulting consensus maps place rfl between umc97 and umc92 on chromosome 3 and $2 between umc1.53 and susl on chromosome 9. Numbers to the left of the linkage lines indicate distances in cM. Modified and published with permission from the Genetics Society of America, from Schnable and Wise (1994), and Wise et al. (1996).

that cosegragated with male sterility in a family segregatingfor the $2-m8122 Mutam-induced allele was cloned and shown to contain a portion of the $2 gene via allelic cross-referencing experiments. This fragment was used to isolate several $2 cDNA clones. One of these clones seemed to be full length because its size corresponds well with that of the $2 transcript (ca. 2.2 kb), its translated sequence contains in-frame stop codons 5’ of two methionine residues, and it encodes a putative mitochondria1 targeting signal (Table I). Computer-based sequence similarity searches of various genome databases revealed that the predicted RF2 protein exhibits 60% identity and 75% similarity to Class I1 mammalian mitochondrial aldehyde dehydrogenases (ALDH) (Fig. 20; Cui et al., 1996).

T-CYTOPLASM MALE STERILITY IN MAIZE

Aldehyde

0

Acid

ALDH

ti

R-C-H

113

0

> R-C-OH

n

NAD+

NADH H+

Figure 20 The protein encoded by the Rj2 nuclear restorer exhibits a high degree of sequence similarity to mammalian mitochondria1 aldehyde dehydrogenases (ALDH, EC 1.2.1.3). These enzymes catalyze the oxidation of aldehydes to the corresponding acids; in so doing, they generate NADH.

The diverse array of potential substrates for an ALDH complicates the effort to establish the precise physiological role of the RF2 protein. ALDHs are capable of oxidizing a wide spectrum of aldehydes to their corresponding acids (Fig. 20; Lindahl, 1992). Aldehydes are highly reactive molecules due to the electrophilic nature of their carbonyl groups. Hence, one presumed physiological role for ALDHs in mammals is the detoxification of aldehydes produced as intermediates in various endogenous biochemical pathways. For example, aldehydes are produced during amino acid, carbohydrate, and lipid metabolism. They can also be produced as the result of membrane peroxidation such as may occur during oxidative stress when reactive oxygen species are produced. An additional complexity to assigning a physiological function for the RF2 protein in restoration relates to the fact that, as discussed above, it probably has a physiological function independent of its role as a nuclear restorer of cms-T. Hence, the physiological functions of the RF2 protein in normal maize and during fertility restoration of cms-T may differ. The difficulty in determining the physiological role of the RF2 protein in fertility restoration is exacerbated by the fact that only limited analyses have been conducted on plant ALDHs (Asker and Davies, 1985; op den Camp and Kuhlemeier, 1997).

C. HOW DOESTHE &’&ENCODED ALDH MEDIATE FER-MLITY RESTORATION? Two hypotheses have been articulated for the mechanism by which an ALDH could participate in fertility restoration (Cui et al., 1996).These hypotheses are referred to as the “metabolic hypothesis” and the “interaction hypothesis.” Central to both of these hypotheses, is the finding that URF13 seems to alter cell viabili-

114

ROGER P. WISE ETAL.

ty, even in the absence of toxin (Duvick, 1965; Pring et ul., 1988; Korth and Levings, 1993). According to the metabolic hypothesis, these perturbations in mitochondrial function create either a greater need for ALDH activity or the need for a novel ALDH activity. For example, if URF13 reduces the efficiency of mitochondrial function, the a-oxidation of lipids may become critical as a source of energy during microsporogenesis. In these circumstances, the role of the RF2 protein in fertility restoration would be clear, because a-oxidation requires ALDH activity. Alternatively, the RF2 protein might function to limit the concentration of acetaldehyde, an intermediate produced during ethanolic fermentation and ethanol metabolism (Fig. 21). Indeed, one of the major functions of Class I1 ALDHs in yeast, Drosophilu, and mammals, is the oxidation of acetaldehyde produced during ethanol metabolism.Alcohol dehydrogenase(ADH) and ALDH function in series to metabolize ethanol to acetate. ADH oxidizes ethanol to form acetaldehyde, which is then oxidized to acetate. Both of these oxidation reactions produce NADH. Indeed, it is this NADH-producing pathway that allows these organisms to survive (at least marginally) under aerobic conditions on diets in which ethanol comprises the sole carbon and energy source. Although plants are not normally faced with the need to metabolize exogenous ethanol, they do produce ethanol during anaerobic stress. The ATP molecules that play a central role in cellular metabolism are generated both directly by glycoly-

NAD+ Lactate

H+ NADH COA

NADH

co2

Pyruvate

Acetyl CoA

COP Acetaldehyde

TCA Cycle

"'YPDC 1

NADH Acetate

Ethanol

Figure 21 RF2 exhibits a high degree of sequence similarity to mammalian mitochondrial aldehyde dehydrogenases (ALDH, EC 1.2.1.3).According to the "metabolic hypothesis" T-urflhnediated perturbations in mitochondrial function create either a greater need for ALDH activity or a need for a novel ALDH activity. For example, if these perturbations affect partitioning of pyruvate between the pyruvate dehydrogenase complex (PDH, EC 1.1.1.27) and pyruvate decarboxylase (PDC, EC 4.1.1. I), the RF2 protein may function in the detoxification of acetaldehyde. ADH (EC 1.1,1.1),alcohol dehydrogenase.

T-CYTOPLASMMALE STERILITY IN MAIZE

11s

sis and by the oxidative phosphorylation of NADH that is produced by the TCA cycle. In the presence of oxygen, pyruvate generated via glycolysis is imported into the mitochondria by the action of pyruvate dehydrogenase (PDH), which oxidatively decarboxylates pyruvate to form acetyl-CoA (Fig. 21). Acetyl-CoA enters the TCA cycle, where additional NADH is produced during oxidation. The continued functioningof glycolysis and the TCA cycle requires that the NADH molecules generated by these pathways be reoxidized to NAD+.Under aerobic conditions, this occurs via oxidative phosphorylation, which generates a tremendous amount of ATP. However, under anaerobic conditions, oxidative phosphorylation is not possible. In this situation, pyruvate is shunted to lactate dehydrogenase (LDH) and pyruvate decarboxylase (PDC). LDH can regenerate a single NAD+ molecule for each molecule of pyruvate reduced to lactate. PDC generates acetaldehyde, which is subsequently reduced by ADH to ethanol, a process that also regenerates an NAD+ molecule. Hence, under anaerobic conditions, LDH and ADH function to regenerate NAD+ so the glycolysis can be sustained and thereby continue to produce some quantity of ATP, even in the absence of oxidative phosphorylation. Therefore, ethanolic fermentation allows plants to survive brief periods of anaerobic stress, such as might occur during transient flooding. However, ethanol is itself toxic. Hence, one possible role for ALDH in plants is in the second step of the oxidation of ethanol (through acetaldehyde) following the return to aerobic conditions after an anaerobic period. Of course, it is not likely that anthers experience anaerobic stress. However ethanolic fermentation can also occur under aerobic conditions. For example, it has been shown that PDC and ADH are expressed (and ethanol accumulates) during tobacco pollen development (Bucher et al., 1994, 1995; Tadege and Kuhlemeier, 1997). The activity of PDH determines the fate of pyruvate. Acetyl-CoA and NADH both inhibit PDH, while pyruvate and ADP activate it. Hence, although pyruvate preferentially enters the TCA cycle, when respiration is inhibited by some mechanism such as anoxia, or the presence of respiratory inhibitors (Bucher et al., 1994), this balance can be shifted in favor of ethanolic fermentation.It is therefore possible that URFl3-mediated perturbations in mitochondria1function could result in additional flux through the ethanolic fermentation pathway during microsporogenesis. Under these conditions, an absence of ALDH activity would result in the accumulation of acetaldehyde and/or ethanol. Because both of these compounds are toxic, such an accumulation could result in tapetal death and male sterility. Hence, one interpretation of the metabolic hypothesis would be that RF2 functions to correct an UFZF13-mediated metabolic disturbance by preventing the ethanol and/or acetaldehyde poisoning of the tapetum. This interpretation is currently being tested by the analysis of single and double mutants of genes involved in ethanolic fermentation. However, because, as indicated above, there are many physiological pathways that generate aldehydes, ethanol metabolism is only one of many pathways that may be involved in RFZmediated fertility restoration.

116

ROGER P. WISE ET AL.

Alternatively, it is also possible that the RF2 protein interacts indirectly with URF13 and thereby prevents the mitochondrial disturbance from occurring at all. This possibility is covered by the interaction hypothesis. Specifically, the RF2 protein could oxidize either an aldehyde component of the inner mitochondrial membrane or Factor X (assuming it is an aldehyde). These oxidation reactions could alter the binding of URFl3 to the inner mitochondria1membrane or inactivate Factor X, respectively. Either of these events could reduce the toxicity associated with the accumulation of URFl3. Hence, the cloning of the $2 gene has stimulated the formulation of several testable hypotheses regarding the specific molecular functions of the RF2 protein in fertility restoration.

VII. PERSPECTIVES BY CMS-T RESEARCHERS a. Donald N. Duvick, Department of Agronomy, Iowa State University For more than three decades I have wondered why maize hybrids in T cytoplasm are slightly less vigorous than their isogenic equivalents in normal cytoplasm. When isogenic hybrids in normal and in T cytoplasm were compared, plants in.T cytoplasm were 2 to 3% shorter during all growth stages, they had about 2% fewer leaves at all growth stages, and their grain yield was 2 to 3% lower (see Duvick, 1965, pages 20-26). These differences were statistically significant, and repeatable. They were independent of and different from the effect that pollen sterility per se can have on plant height and grain yield. Pollensterile plants are much shorter and yield more grain than pollen-fertile plants, particularly when plants are subjected to stress at the time of flowering, and particularly with genotypes that are highly susceptible to stress. These differences are only an indirect consequence of the changes induced by T cytoplasm. The yield gain probably is due directly to the energy-sparing effect of pollen sterility per se, and is expressed in any cytoplasmic male-sterile system (or even by detasselled plants in normal cytoplasm). I speculated that the size and yield reduction caused by T cytoplasm is due to a defective cytoplasmic gene or genes, independent of the genetic defect that causes cytoplasmic pollen sterility.Alternatively, fertility restoration might confer only partial correction of the basic defect in T cytoplasm, enough to restore pollen fertility to a nearly normal condition, but not enough to bring general background metabolism up to completely normal levels. But I was unable to test these hypotheses, 30 years ago. Perhaps today’s technology will enable design and execution of experiments to test these conjectures. The question needing an answer is: Why are pollen-fertile plants (plants with restorer genes) in T cytoplasm less vigorous than their nuclear isogenic equivalents in normal cytoplasm?

T-CYTOPLASM MALE STERILITY IN MAIZE

117

b. Olen C. Yoder and B. Gillian Turgeon, Department of Plant Pathology, Cornell University The uniqueness of the C. heterostrophus Tox genes to the genome of race T suggests that they were acquired horizontally rather than vertically. The source of these genes, however, remains a mystery and information gained from scrutiny of the gene structures is inconclusive. On one hand, these genes have introns, suggesting eukaryotic origin; on the other, they have highest sequence similarity to bacterial polyketide synthases, opening the possibility of a prokaryotic source. Of particular interest is the fact that M.zeae-maydis produces a polyketide toxin similar to T-toxin and has a functional homolog of the C. heterostrophus PKSl gene. In both fungi, multiple genes are required for toxin biosynthesis; in M. zeae-maydis these clearly constitute a gene cluster, whereas in C. heterostrophus it is likely that P K S l and DECl were initially linked but are now on two different chromosomes as the result of their linkage to the breakpoint of a reciprocal translocation. Gene clusters are commonly found to control secondary metabolite biosynthesis in fungi. An immediate goal in the continuing analysis of C. heterostrophus and M . zeae-maydis is to define the limits of the cluster in each fungus and to determine the function of each gene contained within the cluster. A long-term goal is to identify additional genes, not involved in polyketide production, whose products are essential for pathogenesis. Toward that end, we have generated several mutants of C. heterostrophus defective in pathogenesis, using the REMI procedure to induce tagged mutations. Analysis of one such mutation site has revealed a gene cluster containing an ORF encoding a cyclic peptide synthetase ( C P S I ) and several other ORFs whose putative products are known in other organisms to be involved in cyclic peptide biosynthesis. Disruption of C P S l results in loss of pathogenic capability. In contrast to the genes for T-toxin biosynthesis, which are unique to race T, CPSl is found in fungi generally and disruption of its homolog in C. victoriae, a pathogen of oats, leads to a drastic reduction in virulence. These results suggest the existence of a cyclic peptide required for general fungal virulence; host specificity of various pathogens is determined by unique factors such as T-toxin, PM-toxin, and in the case of C. victoriae, victorin. c. Daryl R. Pring, Crop Genetic & Environmental Research Unit, USDAARS and Plant Pathology Department, University of Florida The historic question concerning cms-T is the nature of the fertility-restoring genes and their recessive alleles. Since maize Rf2 may represent an aldehyde dehydrogenase, an assumed normal gene, determination of the nature of the $2 allele would be a major advance in our understanding of abnormal processes resulting in male sterility. What, then, might be the “normal” functions of R f l , Rf8, or Rf*? A key may be the site(s) of action of these genes. Each of the three 5 ’ -

118

ROGER P. WISE ETAL.

termini associated with RfZ, RfS, and Rf* occur within T-urfl3, in a region highly similar to the progenitor sequences located 3’ to rm26. Analyses of rm26 transcripts revealed three species (Maloney et al., 1989). The largest transcript represents the initiated species, and a processing event generates the 5’ mature species. The third transcript probably represents a precursor to 3’ processing, generating the mature rrn26. Sequences 3‘ to the mapped 3’-terminus of mature rrn26 (Kennell and Pring, 1989), which may include the terminus of the rm26 progenitor, are represented within the 317 bp shared by T-urfl3 and rm26 (Dewey et al., 1986). These T-urfZ3 sequences may then represent a template for processes associated with rrn26 maturation. The assignment of a 3’-terminus to the immature rm26 does not necessarily reflect transcript termination sequences, but rather, a probable intermediate, stable terminus. The complexity of 3’ transcript processing in chloroplasts, recently invoked for mitochondria, raises the distinct possibility that endo- and exonucleases, and regulatory factors, may be involved. Since such a scenario may invoke multiple mechanisms, Rfl and its analogs may represent components involved in rm26 maturation. Thus, it is conceivable that mechanisms operative in the putative 3’ transcription termination and processing of rm26 are also operative in cleavage within T-urfl3. The consensus sequence derived for 5’-termini associated with Rfl, RfS, and Rf* are retained in the progenitor sequence 3‘ to rrn26, albeit two sites exhibit 1-nt changes. The development of in vitro transcript processing assays, using extracts from lines differing in the candidate alleles, may allow the pursuit of these points. How the recessive alleles of the three Rf genes may fit a possible rm26/T-urfl3 processing model may emerge from such studies. Still, a more profound question remains: Why does the presence of truncated transcripts in Rfl, Rj8, or Rf* lines, which have abundant, whole-length T-urfl3 transcripts, apparently interfere with accumulation of UW13? The development of in vitro translation systems programmed with appropriate templates may allow the effect of the truncated transcripts on T-urf13 translation to be addressed. d. Charles S. Levings 111, Department of Genetics, North Carolina State University Despite substantial correlative evidence linking T-urfl3 and its polypeptide product U W l 3 with cytoplasmic male sterility (crns), the precise relationship between T-urfl3 and crns remains unresolved. Especially supportive of a causeand-effect relationship are the studies of crns-T revertants in which a mutation of deletion of T-urfl3 that knocks out URF13 synthesis results in the phenotypic reversion from male sterility and disease susceptibility to male fertility and disease resistance (Rottmann et al., 1987; Wise et al., 1987a; Fauron et al., 1990). Other data also suggest the importance of UW13 in causing crns, for example, the effect of RfZ on the transcriptional profile of T-urfl3 (Dewey et al., 1987; Kennell et al., 1987; Kennell and Pring, 1989).

T-CYTOPLASM MALE STERILITY INMAIZE

119

Firmly established is the cause-and-effect association between URFl3 and disease susceptibility (toxin sensitivity). In this instance, pathotoxins (called Ttoxins) produced by Bipolaris maydis race Tor Phyllosticta maydis interact with URFl3 to permeabilize the inner mitochondrial membrane, thereby dissipating the membrane potential, terminating oxidative phosphorylation, and causing the loss of mitochondrial function and cell death (Levings and Siedow, 1992). Expression of T-urfZ3 in Escherichia coli (Dewey et al., 1988; Braun et al., 1989), tobacco (von Allmen et al., 1991; Chaumont et al., 1995), yeast (Glab et al., 1990; Huang et al., 1990), and insects (Korth and Levings, 1993) also confers T-toxin and methomyl sensitivity, confirming the capacity of URFl3 to interact with toxin to permeabilize membranes in a variety of organisms. An attractive explanation for crns is that the mechanism for toxin sensitivity and crns are the same. In cms-T maize, URFl3 accumulates in the inner mitochondrial membrane. If an endogenous compound with toxin-like properties exists, it could interact with URF13 to permeabilize the inner mitochondrial membrane and cause cell death, When this compound occurs uniquely or predominately in the anther tissue, it could interact with URF13 to terminate pollen development and explain the crns phenomenon. Because URFl3 accumulates in all cms-T plant tissues, the endogenous compound must exhibit limited expression in non-anther tissues to avoid a lethal plant effect. In cms-T plants there is evidence that the T cytoplasm alters leaf number, internode length, and grain yield (Duvick, 1965). Limited expression of a toxin-like compound may account for these unique effects in plants carrying the T cytoplasm. Indeed, preliminary studies support the occurrence of an endogenous compound with toxin-like properties that interact with URF13 to cause crns (D. Cho and C. S . Levings 111, unpublished data). Treatment of cms-T mitochondria with an aqueous extract prepared from anther tissue results in the dissipation of membrane potential (AT)and the stimulation of respiration in the presence of NADH. Treatment of mitochondria from normal maize with the extract has no effect on membrane potential or respiration. These results, which are similar to those observed with cms-T and normal mitochondria treated with T-toxin or methomyl, suggest that a compound with toxin-like properties occurs in the anther tissue. Thus far, the specific identity of this compound remains unresolved. These initial studies support the view that the capacity of URF13 to interact with toxin or an endogenous compound accounts for disease susceptibility and crns, respectively.

VIII. FUTURE DIRECTIONS Substantial progress has been made in understanding the cms-T system of maize. Early studies revealed that both male sterility and disease susceptibility are maternally inherited, and that susceptibility is due to the sensitivity of T-cytoplasm

120

ROGER P. WISE ETAL.

mitochondria to P-polyketol fungal toxins. These studies provided a foundation for further research that resulted in the cloning of the T-urfZ3 and Rf2 genes from maize and the ChPKSl gene from C.heterostrophus, and the generation of models for the topology of URF13 in the inner mitochondrial membrane, Rfl-mediated processing of T-urfl3 transcripts, and the evolution of toxin biosynthesis in C. heterostrophus and M.zeue-muydis. Although these advances have provided significant insight into this interesting nuclear-mitochondria1 and host-pathogen interaction, much remains to be learned. For example: What is the molecular basis of male sterility? It is well established that fungal pathotoxins interact with URFl3 to cause mitochondrial disruption leading to cell death. However, it is still unclear how URF13 causes mitochondrial dysfunction leading to male sterility; this dysfunction occurs in the absence of the fungal toxins. Preliminary studies by D. Cho and C. S. Levings (Section VIId) support the existence of a tassel-specific, toxin-like compound (termed Factor X by Flavell, 1974) that could interact with URF13 to permeabilize the mitochondrial membrane and cause cell death. If confirmed, this finding would suggest that the mechanism of male sterility is similar to the mechanism of disease susceptibility. Although this hypothesis is intriguing, it is also possible that T-cytoplasm mitochondria are less efficient than N mitochondria and are therefore unable to meet the exceptional energy demands of microsporogenesis (Warmke and Lee, 1978; Lee and Warmke, 1979), thereby causing premature degeneration of the tapetum and hence male sterility. In support of this hypothesis is the observation by D. N. Duvick (Section VIIa; Duvick, 1965) that T-cytoplasm lines grow less vigorously than Ncytoplasm lines, even in the absence of disease. Of course, it is also possible that Factor X works in concert with the high metabolic rate of tapetal cells to cause sterility. How do Rfl and Rf 2 mediate fertility restoration? Although both Rfl and Rf2 are required to restore fertility to cms-T, they have very different functions. The molecular phenotypes mediated by Rfl and two similar restorers, Rfl and R F , indicate that they are involved in the differential processing of T-~$13 transcripts and the concurrent reduction of the URF13 protein. These processing events result in novel transcripts, but do not significantly decrease the levels of the original T-urfl3 transcripts. Hence, the mechanism by which T-urfl.3 transcript alteration influences the reduction of URF13 is not clear. The Rfl/Turf13 interaction has the potential to serve as a model for restorer-mediated processing of mitochondrial transcripts because this process occurs in many cms systems (e.g., sorghum, rice, petunia, and rapeseed). In addition, the analysis of this interaction may contribute to our understanding of fundamental processes of RNA processing in plants.The sequence of rf2 indicates that it encodes a mitochondrial ALDH. Identification of the physiologically significant substrate(s) of this enzyme may provide support for one of the several hy-

T-CYTOPLASMMALE STERILITY IN MAIZE

121

potheses that have been proposed to explain the role of an ALDH in fertility restoration. Why are restored T-cytoplasm lines acutely susceptible to C. heterostrophus and M. zeae-maydis? There is evidence that restoration can reduce the sensitivity of T-cytoplasm mitochondria to T-toxin (Watrud et al., 1975a; Barratt and Flavell, 1975). This is likely due to the ability of Rfl to reduce URF13 levels in the mitochondria and thus reduce the amount of substrate available for toxin binding (Braun et al., 1990; Hack et al., 1991). However, this reduction in mitochondrial sensitivity to the toxins does not provide significant resistance to the fungi (as evidenced by the 1970 epidemic on restored T-cytoplasm maize). Would further suppression of URF13 levels result in resistance? As the molecular mechanisms associated with Rfl- and Rf2-mediated restoration are elucidated, it may become possible to decrease (or even eliminate) disease susceptibility without the loss of the valuable characteristic of male sterility. Alternatively, if it is not possible to generate T-cytoplasm lines that are toxin resistant, are there ways to confer disease resistance to toxin-sensitive lines? Perhaps research on the mechanisms of pathogenesis in the fungi will identify ways to modify maize to better defend itself from fungal infection, such as by interfering with the function of fungal genes involved in disease induction, such as CPSI. Current efforts to identify and clone genes responsible for pathogenesis in these fungi may provide answers. Although it is clear that our understanding of cms-T and its associated diseases has advanced considerably, many significant questions remain to be answered. The challenge now is to build on the existing cytological, physiological, genetic, and molecular foundation to answer the remaining questions and assist in maize improvement.

ACKNOWLEDGMENTS The authors thank Drs. Don Duvick, Daryl Pring, Sam Levings, Olen Yoder, and Gillian Turgeon

for contributing their perspectives on cms-T; Xiangqin Cui and Drs. Ed Braun, Gillian Turgeon, and Olen Yoder for sharing unpublished figures and data; and Ms. Susan Aldworth for helping with the preparation of the photographs and drawings. Research on cms-T in the Wise and Schnable laboratories was supported in part by USDANRI/CGP Grant 9600804, Pioneer Hi-Bred International, and the Human Frontiers in Science Program. Research on Cochliobolus heterosrrophus in the Bronson laboratory was supported in part by USDA-NRI/CGP Grant 9401224. This chapter is a joint contribution of the Corn Insects & Crop Genetics Research Unit, USDAAgricultural Research Service and the Iowa Agriculture and Home Economics Experiment Station. Journal Paper No. J- 17752 of the Iowa Agriculture and Home Economics Experiment Station, Ames, Iowa, 50011, Project Nos. 3368. 3273, and 3390 and supported by Hatch Act and State of Iowa Funds.

122

ROGER P. WISE ETAL.

REFERENCES Alcorn, J. L. (1975). Race-mating type associations in Australian populations of Cochliobolus heterostrophus. Plant Dis. Rep. 59,708-71 1. Aldrich, H. C., Gracen, V. E., York, D., Earle, E. D., and Yoder, 0. C. (1977). Ultrastructural effects of Helminthosporium maydis race T toxin on mitochondria of corn roots and protoplasts. Tissue Cell 9,167-177. Arntzen, C. J., Haugh, M. F., and Bobick, S. (1973a). Induction of stomata1 closure by Helminthosporium maydis pathotoxin. Plant Physiol. 52,569-574. Arntzen, C. J., Koeppe, D. E., Miller, R. J., and Peverly, J. H. (1973b). The effect of pathotoxin from Helrninthosporium maydis (race T) on energy-linked processes of corn seedlings. Physiol. Plant Pathol. 3,79-89. Asker, H., and Davies, D. D. (1985). Mitochondria1 aldehyde dehydrogenase from higher plants. Phytochemistry 24,689-693. Ayers, J. E., Nelson, R. R., Koons, C., and Scheifele, G. L. (1970). Reactions of various maize inbreds and single crosses in normal and male-sterile cytoplasm to the yellow leaf blight organism (Phyllosticta sp.). Plant Dis. Rep. 54,277-280. Barratt, D. H. P., and Flavell, R. B. (1975). Alterations in mitochondria associated with cytoplasmic and nuclear genes concerned with male sterility in maize. Theor: Appl. Genet. 45,3155321, Beak, T. P., and Goldberg, R. B. (1997). A novel cell ablation strategy blocks tobacco anther dehiscence. Plant Cell 9, 1527-1545. Beckett, J. B. (1971). Classification of male-sterile cytoplasms in maize (Zea mays L.). Crop Sci. 11, 72 1-726. Bedinger, P. (1992). The remarkable biology of pollen. Plant Cell 4, 879-887. Bemill&, A., Ghazi, A,, Charbonnier, M., and Bonavent, J.-F. (1984). Effects of methomyl and Helminthosporium maydis toxin on matrix volume, proton motive force, and NAD accumulation in maize (Zea mays L.) mitochondria. Plant Physiol. 76,508-517. Bhullar, B. S., Daly, J. M., and Rehfeld, D. W. (1975). Inhibition of dark CO, fixation and photosynthesis in leaf discs of corn susceptible to the host-specific toxin produced by Helminthosporium maydis, race T. Plant Physiol. 56, 1-7. Borck, K. S., and Walbot, V. (1982). Comparison of the restriction endonuclease digestion patterns of mitochondria1 DNA from normal and male sterile cytoplasms of Zea mays L. Genetics 102, 109- 128. Braun, C. J., Siedow, J. N., Williams, M. E., and Levings, C. S., I11 (1989). Mutations in the maize mitochondrial T-urfl3 gene eliminate sensitivity to a fungal pathotoxin. Proc. Natl. Acad. Sci. USA 86,4435-4439. Braun, C. J., Siedow, J. N., and Levings, C. S., 111 (1990). Fungal toxins bind to the URF13 protein in maize mitochondria and Escherichia coli. Plant Cell 2, 153-161. Braun, E. J., and Howard, R. I. (1994). Adhesion of Cochliobolus heterostrophus conidia and germlings to leaves and artificial surfaces. Exp. Mycol. 18,211-220. Brettell, R. I. S., Goddard, B. V. D., and Ingram, D. S. (1979). Selection of Tcms-cytoplasm maize tissue cultures resistant to Drechslera maydis T-toxin. Maydica 24,203-213. Brettell, R. I. S., Thomas, E., and Ingram, D. S . (1980). Reversion of Texas male-sterile cytoplasm maize in culture to give fertile, T-toxin resistant plants. Theor: Appl. Genet. 58,55-58. Bronson, C. R. (1988).Ascospore abortion in crosses of Cochliobolus hererostrophus heterozygous for the virulence locus Toxl. Genome 30,12- 18. Bronson, C. R. (1991). The genetics of phytotoxin production by plant pathogenic fungi. Experientia 47,77 1-776. Bronson, C. R. (1998). The genetics of pathogenicity in Cochliobolus heterostrophus. In “Molecular

T-CYTOPLASM MALE STERILITY INMAIZE

123

Genetics of Host-Specific Toxins in Plant Disease” (K. Kohmoto and 0. C. Yoder, eds.). Kluwer, Dordrecht. Bronson, C. R., Taga, M., and Yoder, 0. C. (1990). Genetic control and distorted segregation of T-toxin production in field isolates of Cochliobolus heterostrophus. Phyroputhology 80,819-823. Bucher, M., Brkdle, R., and Kuhlemeier, C. (1994). Ethanolic fermentation in transgenic tobacco expressing Zymomonus mobilis pyruvate decarboxylase. EMBO ./. 15,2755-2763. Bucher, M., Brander, K. A., Sbicego, S., Mandel, T., and Kuhlemeier, C. (1995). Aerobic fermentation in tobacco pollen. Plunr Mol. B i d . 28,739-750. Chang, H.-R., and Bronson, C. R. (1996). A reciprocal translocation and possible insertion(s) tightly associated with host-specific virulence in Cochliobolus heterostrophus. Genome 39,549-557. Chang, T.-T. (1954). Pollen sterility in maize. M.S. thesis, Cornell University, Ithaca, N.Y. Chaumont, F., Bernier, B., Buxant, R., Williams, M. E., Levings, C. S., 111, and Boutry, M. (1995). Targeting the maize T-utfZ3 product into tobacco mitochondria confers methomyl sensitivity to mitochondrial respiration. Proc. Nutl. Acud. Sci. USA 92, 1167-1171. Ciuffetti, L. M., Yoder, 0.C., and Turgeon, B. G. (1992). Amicrobiological assay for host-specific fungal polyketide toxins. Fungal Genet. Newslett. 39, 18-19. Comstock, J. C., and Scheffer, R. P. (1972). Production and relative host-specificity of a toxin from Helminthosporium muydis race T. P lunt Dis. Rep. 56,247-25 I. Conley, C. A., and Hanson, M. R. (1994). Tissue-specific protein expression in plant mitochondria. Plant Cell 6,85-91. Craig, J. (1971). Occurrence of Helminthosporium muydis race T in West Africa. Plant Dis. Rep. 55, 672-673. Cui, X., Wise, R. P., and Schnable, P. S. (1996). The rj2 nuclear restorer gene of male-sterile T-cytoplasm maize. Science 272, 1334-1336. Dale, R. M. K., Mendu, N., Ginsburg, H., and Kridl, J. C. (1984). Sequence analysis of the maize mitochondrial26S rRNA gene and flanking regions. Plusmid ll, 141-150. Daly, J. M., and Knoche, H. W. (1982). The chemistry and biology of pathotoxins exhibiting hostselectivity. In “Advances in Plant Pathology” (D. s. Ingram and P. H. Williams, eds.), pp. 83-138. Academic Press, New York. Danko, S. J., Kono, Y.. Daly, J. M., Suzuki, Y., Takeuchi, S., and McCrery, D. A. (1984). Structure and biological activity of a host-specific toxin produced by the fungal corn pathogen Phyllosticru muydis. Biochemistry 23,759-766. Day, P. R. (ed.) (1977). The genetic basis of epidemics in agriculture. Ann. N. Y Acud. Sci. 287. Dewey, R. E., and Korth, K. L. (1994). Molecular aspects of cytoplasmic male sterility in maize. In “Biotechnology in Agriculture and Forestry: Maize” (Y. P. S. Bajaj, ed.), Vol. 25, pp. 402-416. Springer-Verlag. New York. Dewey, R. E., Levings, C. S., 111, and Timothy, D. H. (1986). Novel recombinations in the maize mitochondrial genome produce a unique transcriptional unit in the Texas male-sterile cytoplasm. Cell 44,439-449. Dewey, R. E., Timothy, D. H., and Levings, C. S., I11 (1987). A mitochondrial protein associated with cytoplasmic male sterility in the Tcytoplasm of maize. P roc. Nutl. Acud. Sci. USA 84,5374-5378. Dewey, R. E., Siedow, J. N., Timothy, D. H., and Levings, C. S., 111 (1988). A 13-kilodalton maize mitochondrial protein in E. coli confers sensitivity to Bipolaris muydis toxin. Science 239,293-295. Dill, C. L., Wise, R. P., and Schnable, P.S. (1997). Rj8 and Rf* mediate unique T-utfl3-transcript accumulation, revealing a conserved motif associated with RNAprocessing and restoration of pollen fertility in T-cytoplasm maize. Genetics 147, 1367-1 379. Dixon, L. K., Leaver, C. J., Brettell, R. I. S., and Gengenbach, B. G. (1982). Mitochondria1 sensitivity to Drechsleru maydis T-toxin and the synthesis of a variant mitochondria1polypeptide in plants derived from maize tissue cultures with Texas male-sterile cytoplasm. Theo,: Appl. Genet. 63, 75-80.

124

ROGER P. WISE ETAL.

Drechsler, C. (1925). Leafspot of maize caused by Ophiobolus heterostrophus, N. sp., the ascigerous stage of a Helminthosporiurn exhibiting bipolar germination J. Agric. Res. 31,701-726. Duvick, D. N. (1956). Allelism and comparative genetics of fertility restoration of cytoplasmically pollen sterile maize. Genetics 41,544-565. Duvick, D. N. (1965). Cytoplasmic pollen sterility in corn. In “Advances in Genetics” (E. W. Caspari and J. M. Thoday, eds.), Vol. 13, pp. 1-56. Academic Press, New York. Duvick, D. N., Snyder, R. J., and Anderson, E. G. (1961). The chromosomal location of Rfl, a restorer gene for cytoplasmic pollen sterile maize. Genelics 46, 1245-1252. Edwardson, J. R. (1970). Cytoplasmic male sterility. Bot. Rev. 36,341-420. Evans, R. C., Stempen, H., and Frasca, P. (1982).Evidence for a two-layered sheath on germ tubes of three species of Bipolaris. Phytoputhology 72,804-807. Fauron, C. M.-R., Abbott, A. G., Brettell, R. I. S., and Gesteland, R. F. (1987). Maize mitochondrial DNA rearrangements between the normal type, the Texas male sterile cytoplasm, and a fertile revertant cms-T regenerated plant Cur,: Genet. 11,339-346. Fauron, C. M.-R., Havlik, M., and Brettell, R. I. S. (1990).The mitochondrial genome organization of a maize fertile cmsT revertant line is generated through recombination between two sets of repeats. Genetics 124,423-428. Flavell, R. (1974).A model for the mechanism of cytoplasmic male sterility in plants, with special reference to maize. P l m t Sci. Lett. 3,259-263. Foley, D. C., and Knaphus, G. (1971). Helminthosporiurn muydis race T in Iowa in 1968. Plant Dis. Rep. 55,855-857. Forde, B. G., and Leaver, C. J. (1980). Nuclear and cytoplasmic genes controlling synthesis of variant mitochondrial polypeptides in male-sterile maize. Proc. Nutl. Acad. Sci. USA 77,418-422. Forde, B. G., Oliver, R. J. C., and Leaver, C. J. (1978).Variation in mitochondrial translation products associated with male-sterile cytoplasms in maize. Proc. Nut/. Acud. Sci. USA 75,3841-3845. Gabay-Laughnan, S., and Laughnan, J. R. (1994). Male sterility and restorer genes in maize. In “The Maize Handbook” (M. Freeling and V. Walbot, eds.), pp. 418-423. Springer-Verlag, New York. Gengenbach, B. G., and Green, C. E. (1975). Selection of T-cytoplasm maize callus cultures resistant to Helminthosporium muydis race T pathotoxin. Crop Sci. 15,645-649. Gengenbach, B. G., Miller, R. J., Koeppe, D. E., and Arntzen, C. J. (1973). The effect of toxin from Helminthosporiurn maydis (race T) on isolated corn mitochondria: Swelling. Can. J. Bor. 51, 2119-2125. Gengenbach, B. G., Green, C. E., and Donovan, C. M. (1977). Inheritance of selected pathotoxin resistance in maize plants regenerated from cell cultures. f roc. Nutl. Acad. Sci. USA 74,5 1 13-5 1 17. Gengenbach, B. G., Connelly, J. A,, Pring, D. R., and Conde, M. F. (1981). Mitochondrial DNA variation in maize plants regenerated during tissue culture selection. Theor Appl. Genet. 59,161- 167. Glab, N., Wise, R. P., Pring, D. R., Jacq, C., and Slonimski, P. (1990). Expression in Succhuromyces cerevisiue of a gene associated with cytoplasmic male sterility from maize: Respiratory dysfunction and uncoupling of yeast mitochondria. Mol. Gen. Genet. 223,24-32. Glab, N., Petit, P. X.,and Slonimski, P. P. (1993). Mitochondrial dysfunction in yeast expressing the cytoplasmic male sterility T-urfl3 gene from maize: Analysis at the population and individual cell level. Mol. Gen. Genet. 236,299-308. Goldberg, R. B., Beak, T. P., and Sanders, P. M. (1993). Anther development: Basic principles and practical applications. Plant Cell 5, 1217-1229. Goldberg, R. B., Sanders, P. M., and Beak, T. P. (1995). A novel cell-ablation strategy for studying plant development. Philos. Truns. R. SOC.London 350,5- 17. Gracen, V. E., and Grogan, C. 0. (1974). Diversity and suitability for hybrid production of different sources of cytoplasmic male sterility in maize. Agron. J. 65,417-421. Gracen, V. E., Forster, M. J., Sayre, K. D., and Grogan, C. 0. (1971).Rapid method for selecting resistant plants for control of southern corn leaf blight. Plant Dis. Rep. 55,469-470.

T-CYTOPLASM MALE STERILITYIN MAIZE

125

Gregory, P., Earle, E. D., and Gracen, V. E. (1977). Biochemical and ultrastructural aspects of southern corn leaf blight disease. In “Host Plant Resistance to Pests” (P. A. Hedin, ed.), ACS Symposium Series 62, pp. 90-1 14. American Chemical Society, Washington, D.C. Gregory, P., Matthews, D. E., York, D. W., Earle, E. D., and Gracen, V. E. (1978). Southern corn leaf blight disease: Studies on mitochondrial biochemistry and ultrastructure. Mycopathologia 66, 105-1 12. Hack, E., Lin, C., Yang, H., and Homer, H. T. (1991).T-URF13 protein from mitochondria of Texas male-sterile maize (Zeu mays L.). PIanr Physiol. 95, 861-870. Hanson, M. R., Sutton, C. A., and Lu, B. (1996).Plant organelle gene expression: altered by RNAediting. Trends Plant Sci. 1 , 5 7 4 4 . Holden, M. J., and Sze, H. (1984). Helminthosporium rnuydis T toxin increased membrane permeability to Ca’+ in susceptible corn mitochondria. Plant Physiol. 75,235-237. Holden. M. J., and Sze, H. (1987). Dissipation of the membrane potential in susceptible corn mitochondria by the toxin of Helminthosporium muydis, race T, and toxin analogs. Plant Physiol. 84, 670-676. Holden, M. J., Colombini, M., and Sze, H. (1985). Channel formation in phospholipid bilayer membranes by the toxin of Helminthosporium muydis, race T. J. Membrune Bid. 87, I5 1 - 157. Hooker, A. L. ( I 972). Southern leaf blight of corn-Present status and future prospects. J. Environ. Qual. 1,244-249. Hooker, A. L., Smith, D. R., Lim, S. M., and Beckett, J. B. (1970a).Reaction of corn seedlings with male-sterile cytoplasm to Helminthosporium maydis. Plant Dis. Rep. 54,708-7 12. Hooker, A. L., Smith, D. R., Lim, S. M., and Musson, M. D. (1970b). Physiological races of Helminthosporium maydis and disease resistance. Planr Dis. Rep. 54, 1109-1 110. Homer, H. T., and Palmer, R. G. (1995). Mechanisms of genic male sterility. Crop Sci. 35, 1527-1535. Homer, H. T., Hall, V. L., and Vargas-Olvera, M. A. (1993). Isolation, sorting, and characterization of uni- and binucleate tapetal protoplasts from anthers of normal and Texas cytoplasmic malesterile Zeu mays L. Protoplasma 173,48-57. Huang, J., Lee, S.-H., Lin, C., Medici, R., Hack, E., and Myers, A. M. (1990).Expression in yeast of the T-URF I3 protein from Texas male-sterile maize mitochondria confers sensitivity to methomyl and to Texas-cytoplasm-specific fungal toxins. EMBO J. 9,339-347. Humaydan, H. S., and Scott, E. W. (1977).Methomyl insecticide selective phytotoxicity on sweet corn hybrids and inbreds having the Texas male sterile cytoplasm. Horrscience 12,312-313. Iwabuchi, M., Kyozuka, J., and Shimamoto, K. (1993).Processing followed by complete editing of an altered mitochondria1 alp6 RNA restores fertility of cytoplasmic male sterile rice. EMBO J. 12, 1437-1446. Jones, D. F., Stinson, H. T., Jr., and Khoo, U. (1957). Pollen restoring genes. CTAgr: Exp. Stu. Bull. Immed. [email protected],2-43. Kaul, M. L. H. (1988). “Male Sterility in Higher Plants.” Springer-Verlag. Berlin. Kemble, R. J., Flavell, R. B., and Brettell, R. I. S. (1982). Mitochondria1 DNA analysis of fertile and sterile maize plants derived from tissue culture with the Texas male sterile cytoplasm. Them Appl. Gener. 62,213-217. Kennell, J. C., and Pring, D. R. (1989). Initiation and processing of utp6, T-uv13 and ORF22I transcripts from mitochondria of T cytoplasm maize. Mol. Gen. Genet. 216, 16-24. Kennell, J. C., Wise, R. P., and Pring, D. R. (1987).Influence of nuclear background on transcription of a maize mitochondrial region associated with Texas male sterile cytoplasm. Mol. Gen. Gener. 210,399-406. Kheyr-Pour, A., Gracen, V. E., and Everett, H. L. (198 I). Genetics of fertility restoration in the C-group of cytoplasmic male sterility in maize. Genetics 98,379-388. Klittich, C. J. R., and Bronson, C. R. (1986). Reduced fitness associated with TOXl of Cochliobolus heterostrophus. Phytopurhology 76, 1294- 1298.

126

ROGER P. WISE ETAL.

Koltunow, A. M., Truettner, J., Cox, K. H., Wallroth, M., and Goldberg, R. B. (1990). Different temporal and spatial gene expression patterns occur during anther development. Plant Cell 2, 1201- 1224. Kono, Y.,and Daly, J. M. (1979). Characterization of the host-specific pathotoxin produced by Helminthosporium maydis, race T, affecting corn with Texas male sterile cytoplasm. Bioorg. Chem. 8,391-397. Kono, Y.,Takeuchi, S., Kawarada, A,. Daly, J. M., and Knoche, H. W. (1980). Structure of the hostspecific pathotoxins produced by Helminthosporium maydis race T. Tetrahedron Lett. 21, 1537-1540. Kono, Y., Takeuchi, S., Kawarada, A,, Daly, J. M., and Knoche, H. W. (1981). Studies on the hostspecific pathotoxins produced in minor amounts by Helminthosporium maydis race T. Bioorg. Chem. 10,206-218. Kono, Y., Danko, S. J., Suzuki, Y.,Takeuchi, S., and Daly, J. M. (1983). Structure of the host-specific pathotoxins produced by Phyllosticta maydis. Tetrahedron Lett. 24,3803-3806. Kono, Y., Suzuki, Y., Takeuchi, S., Knoche, H. W., and Daly, J. M. (1985). Studies on the host-specific pathotoxins produced by H. maydis race T and E maydis: Absolute configuration of PM-toxins and HMT-toxins. Agric. Biol. Chem. 49,559-562. Korth, K. L., and Levings, C. S., 111 (1993). Baculovirus expression of the maize mitochondrial protein URF13 confers insecticidal activity in cell cultures and larvae. Proc. Natl. Acad. Sci. USA 90, 3388-3392. Korth, K. L., Kaspi, C. I., Siedow, J. N., and Levings, C. S., I11 (1991). URF13, a maize mitochondria1 pore-forming protein, is oligomeric and has a mixed orientation in Escherichia coli plasma membranes. Proc. Narl. Acad. Sci. USA 88, 10865-10869. Landschiitze, V., Willmitzer, L., and Miiller-Rober, B. (1995). Inhibition of flower formation by antisense repression of mitochondrial citrate synthase in transgenic potato plants leads to a specific disintegrationof the ovary tissues of flowers. EMBO J. 14,660-665. Laser, K. D., and Lenten, N. R. (1972). Anatomy and cytology of microsporogenesis in cytoplasmic male sterile angiosperms. Bot. Rev. 38,425-454. Laughnan, J. R., and Gabay, S. J. (1978).Nuclear and cytoplasmic mutations to fertility in S male sterile maize. In “Maize Breeding and Genetics” (D. B. Walden, ed.), pp. 427-447. Wiley, New York. Laughnan, J. R., and Gabay-Laughnan, S. (1983). Cytoplasmic male sterility in maize. Annu. Rev. Genet. 17,27-48. Lee, S.-L. J., and Warmke, H. E. (1979). Organelle size and number in fertile and T-cytoplasmic malesterile corn. Am. J. Bot. 66, 141-148. Leonard K. J. (1971). Association of virulence and mating type among Helminthosporiurn maydis isolates collected in 1970. Plant Dis. Rep. 55,759-760. Leonard, K. J. (1974). Bipolaris maydis race and mating type frequencies in North Carolina. Plant Dis. Rep. 58,529-531. Leonard, K. J. (1977a). Races of Bipolaris maydis in the southeastern U.S. from 1974-1976. Plant Des. Rep. 61,914-915. Leonard, K. J. (1977b). Virulence temperature optima and competitive abilities of isolines of races T and 0 of Bipolaris maydis. Phytopathology 67, 1273-1279. Leonard, K. J. (1987). The host population as a selective factor. In “Populations of Plant Pathogens: Their Dynamics and Genetics” (M. S. Wolfe and C. E. Caten, eds.), pp. 163-179. Blackwell Sci., Oxford. Levings, C. S., 111 (1990). The Texas cytoplasm of maize: Cytoplasmic male sterility and disease susceptibility. Science 250,942-947. Levings, C . S., 111 (1993). Thoughts on cytoplasmic male sterility in cms-T maize. Planr Cell 5, 1285-1290. Levings, C. S., 111, and Pring, D. R. (1976). Restriction endonuclease analysis of mitochondrial DNA from normal and Texas cytoplasmic male-sterile maize. Science 193, 158-160.

T-CYTOPLASM MALE STERILITY IN MAIZE

127

Levings, C. S., 111, and Pring, D. R. (1979). Molecular basis of cytoplasmic male sterility in maize. In “Physiological Genetics” (J. G. Scandalios, ed.), pp. 171-193. Academic Press, New York. Levings, C. S., 111, and Siedow, J. N. (1992). Molecular basis of disease susceptibility in the Texas cytoplasm of maize. Plant Mol. Biol. 19, 135- 147. Lim, S. M., and Hooker, A. L. (1971). Southern corn leaf blight: Genetic control of pathogenicity and toxin production in race T and race 0 of Cochliobolus heterostrophus. Generics 69, 115- 117. Lim, S. M., and Hooker, A. L. (1972). Disease determinant of Helminrhosporium maydis race T. Phytopathology 62,968-971. Lim, S . M., Hooker, A. L., and Smith, D. R. (1971). Use of Helminthosporium maydis race T pathotoxin to determine disease reaction of germinating corn seed. Agron. J. 63,712-713. Lindahl, R. (1992).Aldehyde dehydrogenases and their role in carcinogenesis. Crit. Rev. Biochem. Mol. Biol. 27,283-335. Lu, S., Lyngholm, L., Yang, G., Bronson, C., Yoder, 0. C., and Turgeon, B. G. (1994). Tagged mutations at the Toxl locus of Cochliobolus heterostrophus by restriction enzyme-mediated integration. Proc. Natl. Acad. Sci. USA 91, 12649-12653. Lu, S. W., Turgeon, B. G., and Yoder, 0. C. (1995). New mutations involved in T-toxin production by Cochliobolus heterostrophus are unlinked to the Toxl locus. Fungal Genet. Newslett. 42A,43. Maloney, A. P., Traynor, P. L., Levings, C S., 111, and Walbot, V. (1989). Identification in maize mitochondrial26S rRNA of a short 5‘-end sequence possibly involved in transcription initiation and processing. Curr Genet. 15,207-212. Mascarenhas, J. P. (1988). Anther- and pollen-expressed genes. In “Temporal and Spatial Regulation of Plant Genes” (D. P. S. Verma and R. B. Goldberg, eds.), pp. 97-115. Springer-Verlag, New York. Mascarenhas, J. P. (1990). Gene activity during pollen development. Annu. Rev. Plant Physiol. 41, 317-338. Matthews, D. E., Gregory, P., and Gracen, V. E. (1979). Helminthosporium maydis race T toxin induces leakage of NAD+ from T cytoplasm corn mitochondria. Plant Physiol. 63, 1149-1153. Mercado, A. C., Jr., and Lantican, R. M. (1961). The susceptibility of cytoplasmic male sterile lines of corn to Helminthosporium maydis Nish & Miy. Philipp. Agric. 45,235-243. Miller, R. J., and Koeppe, D. E. (1971). Southern corn leaf blight: Susceptible and resistant mitochondria. Science 173,67-69. Moore, W. F. (1970). Origin and spread of southern corn leaf blight in 1970. Plant Dis. Rep. 54, 1104-1 108. Mundt, C. C., and Browning, J. A. (1985). Genetic diversity and cereal rust management. In “The Cereal Rusts” (A. P. Roelfs and W. R.Bushnell, eds.), pp. 527-560. Academic Press, New York. op den Camp, R. G. L., and Kuhlemeier, C. (1997). Aldehyde dehydrogenase in tobacco pollen. Plant Mol. Biol. 35,355-365. Pacini, E. (1997). Tapetum character states: Analytical keys for tapetum types and activities. Can. J. Bot. 75, 1448-1459. Pacini, E., Franchi, G. G., and Hesse, M.(1985). The tapetum: Its form, function and possible phylogeny in Embryophyta. Plant Sysr. Evol. 149, 155-185. Palmer, R. G., Albertsen, M. C., Homer, H. T., and Skorupska, H. (1992). Male sterility in soybean and maize: Developmental comparisons. Nucleus 35, 1- 18. Payne, G., Knoche, H. W., Kono, Y., and Daly, J. M. (1980). Biological activity of purified hostspecific pathotoxin produced by Bipolaris (Helminrhosporium) maydis, race T. Physiol. Plant Pathol. 16,227-239. Peterson, P. A,, Flavell, R. B., and Barratt, D. H. P. (1975). Altered mitochondria1membrane activities associated with cytoplasmically-inheriteddisease sensitivity in maize. Theor: Appl. Genet. 45, 309 -3 14. Potter, S., Durovic, P., Russell, A,, Wang, X., de Jong-Wong, D., and Dennis, P. P. (1995). Preribosoma1 RNA processing in Archaea: Characterization of the RNP endonuclease mediated processing

128

ROGER P. WISE ETAL.

of precursor 16s rRNA in the thermoacidophile Sulfolobus acidocaldarius. Biochem. Cell B i d . 73,813-823. Pring, D. R., and Levings, C. S., 111 (1978).Heterogeneity of maize cytopbsmic genomes among malesterile cytoplasms. Generics 89, 121- 136. Pring, D. R., and Lonsdale, D. M. (1989).Cytoplasmic male sterility and maternal inheritance of disease susceptibility in maize. Annu. Rev. Phyropnrhol. 27,483-502. Pring, D. R., Gengenbach, B. G., and Wise, R. P. (1988).Recombination is associated with polymorphism of the mitochondrial genomes of maize and sorghum. Philos. Trans. R. SOC.London B. 319, 187-198. Rhoads, D. M., Kaspi, C. I., Levings, C. S., 111, and Siedow, J. N. (1994).N,N’-dicyclohexylcarbodiimide cross-linking suggests a central core of helices I1 in oligomers of URF13, the pore-forming T-toxin receptor of cms-T maize mitochondria. Proc. Narl. Acud. Sci. USA 91,8253-8257. Rocheford, T. R., Kennell, J. C., and Pring, D. R. (1992).Genetic analysis of nuclear control of Turf13/orf221 transcription in T-cytoplasm maize. Theor: Appl. Genet. 84,891-898. Rose, M. S., Yoder, 0. C., and Turgeon, B. G. (1996).A decarboxylase required for polyketide toxin production and high virulence by Cochliobolus hererosrrophus. 8th Int. Symp. Mol. Plant-Microbe Int., Knoxville, TN, p. 1-49, Rottmann, W. H., Brears, T., Hodge, T. P., and Lonsdale, D. M. (1987).A mitochondrial gene is lost via homologous recombination during reversion of CMS T maize to fertility. EMBO J. 6, 1541- 1546. Scheifele, G . L., and Nelson, R. R. (1969).The occurrence of Phyibstictu leaf spot of corn in Pennsylvania. Plant Dis. Rep. 53, 186-1 89. Scheifele, G. L., Nelson, R. R., and Koons, C. (1969).Male sterility cytoplasm conditioning susceptibility of resistant inbred lines of maize to yellow leaf blight caused by Phyllosricra zeae. Plunt Dis. Rep. 53,656-659. Scheifele, G . L., Whitehead, W., and Rowe, C. (1970).Increased susceptibility to southern leaf spot (Helminrhosporium muydis) in inbred lines and hybrids of maize with Texas male-sterile cytoplasm. Plant Dis.Rep. 54,501-503. Schnable, P. S., and Wise, R. P. (1994).Recovery of heritable, transposon-induced, mutant alleles of the rf2 nuclear restorer of T-cytoplasm maize. Generics 136, 1 171- 1 185. Schnable, P. S.,and Wise. R. P. (1998).The molecular basis of cytoplasmic male sterility and fertility restoration. Trends Plant Sci. 3, 175-1 80. Singh, M., Hamel, N., Menassa, R., Li, X.-Q., Young, B., Jean, M., Landry, B. S., and Brown, G. G. (1996).Nuclear genes associated with a single Brnssica CMS restorer locus influence transcripts of three different mitochondrial gene regions. Generics 143,505-516. Smith, D. R., Hooker, A. L., and Lim, S.M. (1970).Physiologic races of Helminrhosporium maydis. Plant Dis.Rep. 54,819-822. Snyder, R. J., and Duvick, D. N. (1969).Chromosomal location of Rj2, a restorer gene for cytoplasmic pollen sterile maize. Crop Sci. 9, 156-157. Suzuki, Y.,Tegtmeier, K. J., Daly, J. M., and Knoche, H. W. (1982).Analogs of host-specifc phytotoxin produced by Helrninthosporium maydis, race T. Bioorg. Chem. 11,313-321. Suzuki, Y., Danko, S . J., Daly, J. M., Kono, Y., Knoche, H. W., and Takeuchi, S. (1983).Comparison of activities of the host-specific toxin of Helminthosporium maydis, race T, and a synthetic C,, analog. Plant Physiol. 73,440-444. Suzuki, Y., Danko, S. J., Kono, Y., Takeuchi, S., Daly, J. M., and Knoche, H. W., (1984).Synthesis of (+/-)-8,16,26,34-tetrahydroxy-6,I0,14, I8,24,28,32,36-octaoxohentetracontane as a potent synthetic analog of race T toxin produced by Helminrhosporium muydis. Agric. Biol. Chem. 48, 232 1-2329. Suzuki, Y., Danko, S . J., Kono, Y., Daly, J. M., Knoche, H. W., and Tdkeuchi, S.(1988).Studies on the conformations of PM-toxin, the host-specific corn pathotoxin produced by Phyllosricra maydis. Agric. Biol. Chem. 52, 15-24.

T-CYTOPLASM MALE STERILITY IN MAIZE

129

Tadege, M., and Kuhlemeier, C. ( 1997). Aerobic fermentation during tobacco pollen development. Plant Mol. Biol. 35,343-354. Tang, H. V., Pring, D. R., Shaw, L. C., Salazar, R. A., Muza, F. R., Yan, B., and Schertz, K. F. (1996). Transcript processing internal to a mitochondrial open reading frame is correlated with fertility restoration in male-sterile sorghum. Plant J. 10, 123- 133. Tatum, L. A. (1971). The southern corn leaf blight epidemic. Science 171, 1 113-1 116. Tegtmeier, K. J., Daly, J. M., and Yoder, 0. C. (1982). T-toxin production by near-isogenic isolates of Cochliobolus heterostrophus, races T and 0. Phytoputhology 72, 1492- 1495. Turgeon, B. G., and Berbee, M. L. (1998). Evolution of pathogenic and reproductive strategies in Cochliobolus and related genera. In “Molecular Genetics of Host-Specific Toxins in Plant Disease” (K. Kohmoto and 0. c. Yoder, eds.). Kluwer, Dordrecht. Turgeon, B. G., Kodama, M., Yang, G., Rose, M. S.,Lu, S. W., and Yoder, 0. C. ( I 995). Function and chromosomal location of the Cochliobolus heterostrophus TOXl locus. Can J . Bot. 73, S 1071-S 1076. Tzeng, T.-H., Lyngholm, L. K., Ford, C. F., and Bronson, C. R. (1992). A restriction fragment length polymorphism map and electrophoretic karyotype of the fungal maize pathogen Cochliobolusheteristrophus. Genetics 130,8 1-96, Ullstrup, A. J. (1972). The impacts of the southern corn leaf blight epidemics of 1970-1971. Annu. Rev. Phytopathol. 10,37-50. Umbeck, P. F., and Gengenbach, B. G. (1983). Reversion of male-sterile T-cytoplasm maize to male fertility in tissue culture. Crop Sci. 23, 584-588. Villareal, R. L., and Lantican, R. M. (1965). The cytoplasmic inheritance of susceptibility to Helminthosporium leaf spot in corn. Philipp. Agric. 49,294-300. von Allmen, J.-M., Rottmann, W. H., Gengenbach, B. G., Harvey, A. J., and Lonsdale, D. M. (1991). Transfer of methomyl and HmT-toxin sensitivity from T-cytoplasm maize to tobacco. Mol. Gen. Genet. 229,405-412. Wallace, D. C. (1989). Mitochondria1 DNA mutations and neuromuscular disease. Trends Genet. 5, 9-13. Ward, G . C. (1995). The Texas male-sterile cytoplasm of maize. In “The Molecular Biology of Plant Mitochondria” (C. S. Levings, 111, and I. K. Vasil. eds.). pp. 433-459. Kluwer, Dordrecht. Ward, G. C, and Levings, C. S., 111 (1991). The protein-encoding gene T-urfl3 is not edited in maize mitochondria. Plant Mol. Biol. 17, 1083- 1088. Warmke, H. E., and Lee, S.-L. J. (1977). Mitochondrial degeneration in Texas cytoplasmic male-sterile corn anthers. J. Hered. 68,213-222. Warmke, H. E., and Lee, S.-L. J. (1978). Pollen abortion in Tcytoplasmic male-sterile corn (Zea mays): A suggested mechanism. Science 200,561-563. Watrud, L. S., Hooker, A. L., and Koeppe, D. E. (1975a). The effects of nuclear restorer genes of Texas male-sterile cytoplasm on host response to Helminthusporium maydis race T. Phytopathology 65, 178-182. Watrud, L. S., Baldwin, J. K., Miller, R. J., and Koeppe, D. E. (197%). Induction of a sensitive response to Helminthosporium maydis race T toxin in resistant mitochondria of corn (Zeu mays L.) by removal of the outer mitochondria1 membrane. Plant Physiol. 56,216-221. Wei, J.-K., Liu, K.-M., Chen, J.-P., Luo, P.-C., and Lee-Stadelmann, 0. Y. (1988). Pathological and physiological identification of race C of Bipoluris maydis in China. Phytopathology 78,550-554. Williams, M. E., and Levings, C. S., 111 (1992). Molecular biology of cytoplasmic male sterility. In “Plant Breeding Reviews” (J. Janick, ed.), Vol. 10, pp. 23-5 I . Wiley, New York. Wise, R. P., and Schnable, P. S . (1994). Mapping complementary genes in maize: Positioning the if1 and rj2 nuclear restorers relative to visible and RFLP markers. Theor: Appl. Genet. 88,785-795. Wise, R. P., Pring, D. R., and Gengenbach, B. G. (1987a). Mutation to male fertility and toxin insensitivity in Texas (T)-cytoplasm maize is associated with a frameshift in a mitochondrial open reading frame. Proc. Natl. Acad. Sci. USA 84, 2858-2862.

130

ROGER P.WISE ETAL.

Wise, R. P., Fliss, A. E., Pring, D. R., and Gengenbach, B. G. (1987b). urjl3-Tof T cytoplasm maize mitochondria encodes a 13-kD polypeptide. Plant Mol. Biol. 9, 121- 126. Wise, R. P., Dill, C. L., and Schnable, P. S. (1996). Murator-induced mutations of the rjl nuclear fertility restorer of T-cytoplasm maize alter the accumulation of T-urfl3 mitochondria1 transcripts. Genetics 143, 1383-1394. Wolfe, M. S. (1985). The current status and prospects of multiline cultivars and variety mixtures for disease resistance. Annu. Rev. Phytopathol. 23,25 1-273. Yang, G., Rose, M. S., Turgeon, B. G., and Yoder, 0. C. (1996). A polyketide synthase is required for fungal virulence and production of the polyketide T-toxin. Plant Cell 8,2139-2150. Yang, H. (1989). Study of mitochondria of normal and Texas cytoplasmic male-sterile maize by conventional, immunocytochemical, and fluorescence microscopy. M.S. thesis, Iowa State University, Ames. Yoder, 0. C. (1973). A selective toxin produced by Phyllosricta maydis. Phytopathology 63, 1361-1 366. Yoder, 0. C. (1976). Evaluation of the role of Helminthosporium maydis race T toxin in southern corn leaf blight. In “Biochemistry and Cytology of Plant Parasite Interaction” (K. Tomiyama, J. M. Daly, I. Uritani, H. Oku, and S. Ouchi, eds.), pp. 16-24. Elsevier, Amsterdam. Yoder, 0. C. (1980). Toxins in pathogenesis. Annu. Rev. Phyroparhol. 18, 103-129. Yoder, 0. C. (1998). A mechanistic view of the fungal/plant interaction based on host-specific toxin studies. In “Molecular Genetics of Host-Specific Toxins in Plant Disease” (K. Kohmoto and 0.C. Yoder, eds.), Kluwer, Dordrecht. Yoder, 0.C., and Gracen, V. E. (1975). Segregation of pathogenicity types and host-specific toxin production in progenies of crosses between races T and 0 of Helminthosporium maydis (Cochliobo/us heterostrophus). Phytoparhology 65,273 -276. Yoder, 0. C., Macko, V., Wolpert, T., and Turgeon, B. G. (1997). Cochliobolus spp. and their hostspecific toxins. In “Plant Relationships” (G. C. Carroll and P. Tudzynski, eds.), Vol. 5, Part A of the “The Mycota” (K. Esser and P. A. Lemke, eds.), pp. 145-166. Springer-Verlag, Berlin. Yun, S.-H., Turgeon, B. G., and Yoder, 0. C. (1997). Comparative genetics of polyketide toxin production by Mycosphaerella zeae-maydis and Cochliobolus heterostrophus. Proceedings of the Fungal Genetics Conference, Asilomar, p. 39. Yun, S.-H., Turgeon, B. G., and Yoder, 0.C. (1998). REMI-induced mutants of Mycosphaerella zeaemaydis lacking the polyketide PM-toxins are deficient in pathogenesis to corn. Physiol. Mol. Plant Pathol. 52,53-66. Zhu, H., Braun, E. J., Perry, J. L., and Bronson, C. R. (1998). Identification, characterization, and mapping of Ecml, a locus affecting extracellular matrix production and lesion size in Cochliobolus heterostrophus. Genome 41, 11 1-1 19.

APPLICATIONOF CAPILLARY ELECTROPHORESIS TO ANION SPECIATION IN SOILWATEREXTRACTS R. Naidu,' S. Naidq2 P. J a ~ ks on,~ R. G. McLaren? and M. E. SumnerS 'CSIRO Land and Water and Cooperative Research Centre for Soil and Land Management Glen Osmond, South Australia 5064 2Panorama Adelaide, South Australia 5064 3Waters Australia Pty.Limited Rydalmere, New South Wales, Australia 4Department of Soil Science Lincoln University Canterbury, New Zealand SDepartmentof Crop and Plant Sciences University of Georgia Athens, Georgia 30602

I. Introduction

II. General Principles 111. Sample Introduction N Separations A. Electroosmotic Migration B. Longitudinal Diffusion C. Electrophoretic Migration D. Analyte Mobility E. Separation of Positively Charged Ionic Solutes F. Separation of Negatively Charged Ionic Solutes V. Detection VI. Comparison with Other Analytical Techniques VII. Implication for the Analysis of Soil Solutions References

131 Advancer in Agronomy, Volume 65 Copyright 0 1999 by Academic Press. All righbrs of reproduction in any form reserved 006S-Zl13/99 $30.00

132

R.NAIDU ETAL.

I. INTRODUCTION Accurate information on the ionic composition of soil solutions and water extracts is required for an understanding of ionic interactions in soils in relation to solid-phase chemistry, ionic mobility, and soil-plant transfer of nutrients. While cation analysis and speciation have been subjects of much research for more than 50 years, anion analysis and speciation gained prominence following the development of ion chromatography (IC) in the 1970s (Small et al., 1975; Gjerde and Fritz, 1979). Ion chromatography uses low-capacity anion-exchange columns for separation of ions and conductivity as the universal mode of detection. Ion chromatography was originally developed to analyze inorganic solutes such as F-, C1and SO,*-, but in recent years, its scope has been expanded to include a variety of other mono-, di-, and trivalent inorganic and organic anions and cations (Jones and Jandik, 1991). This has been achieved through major developments in the efficiency and selectivity of ion-exchange columns, in addition to the use of a wide variety of separation modes and detection methods. Consequently, IC is now widely applied to water, environmental, industrial, food, and clinical analyses. Capillary electrophoresis (CE) is a relatively new instrumental technique that was primarily developed for the analysis of proteins, peptides, and vitamins in biological and food samples. With the general exception of lipids, CE can now be used for analysis of most food components (Cancalon, 1995a; Lindeberg, 1995a,b; Morawski et al., 1993; Zeece, 1992). However, during the past 10 to 15 years, this technique has gained increasing application in the assay of inorganic anions and low-molecular-weight carboxylic acids in environmental samples such as wastewaters. More recently its application has been expanded to the analysis of drugs (Tennery and Wells, 1994) and soil water samples (Naidu et al., 1999). Unlike IC, which is based on ion-exchange separations, CE utilizes a narrow-diameter capillary (50- 100 km fused silica) and separates species on the basis of their ionic mobility under the influence of an applied potential. This leads to a different selectivity than IC. Additionally, CE results in extremely efficient separations, uses small sample volumes, and has very rapid analysis times. Because the instrumentation is relatively simple, the technique provides an attractive tool for the speciation of ions. Greater recognition that assessments of health hazards, toxicity, and bioavailability should be based on specific chemical forms rather than on total elemental levels has, in recent years, led to an increasing interest in chemical speciation procedures. Many different approaches have been adopted by scientists for the estimation of various species present in soil solution and water samples. For instance, anion speciation has been achieved using IC (Small et al., 1975; Gjerde and Fritz, 1979), gas chromatography (Daughtrey et al., 1975; Yu and Wai, 1991), supercritical fluid extractions (Laintz et al., 1992a,b), high-performance liquid chro-

CAPILLARY ELECTROPHORESIS AND ANION SPECIATION

13 3

matography (HPLC; Li and Li, 1993, inductively coupled plasma emission spectrometry after hydride generation (Cutter, 1985; Ericzon et al., 1989; Mentasti et al., 19891, and cathodic stripping voltametry (van den Berg and Khan, 1990). However, speciation using these methods is frought with difficulties and they are often subject to chemical interferences. Given these limitations, there has been increasing interest in the separation and determination of inorganic ions using CE. Capillary electrophoresis is now recognized as a potentially important analytical separation technique because it brings speed, quantitation, reproducibility, and automation to the inherently highly resolving technique of electrophoresis. Swaile and Sepaniak (1991) reported a novel method of separating and detecting metal cations by CE. For the cations Ca(II), Mg(II), and Zn(II), they reported limits of detection in the microgram per liter range and successfully applied the technique to the determination of Ca(I1) and Mg(I1) in blood serum using indirect fluorescence detection. Subsequently, numerous investigators published papers on cation speciation using CE under a variety of instrument and electrolyte conditions (Weston et al., 1992a,b; Koberda et al., 1992; Beck and Engelhardt, 1992; Baechmann et al., 1992; Morawski et al., 1993; Chen and Cassidy, 1993; Shi and Fritz, 1994). Anion speciation has also been successfully achieved by CE although the limits of detection for the majority of test ions have been relatively poor. Wildman et al. (1991) and Morin et al. (1992) separated arsenite and arsenate by CE with indirect ultraviolet (UV) detection, but only at relatively high concentrations. Poor detection in CE has been attributed to the ultra low sample volumes injected into the capillary. Li and Li (1995) suggest that by using stacking/pulse injections of samples, detection limits can be increased mainly by increasing the sample size. Based on this technique, sub microgram per liter level as could be detected in standard solutions. Capillary electrophoresis has been the subject of numerous reviews in the past few years (Monning and Kennedy, 1994). Its application to food analysis (Lindeberg, 1995a,b) and dairy research has also been reviewed (Olieman, 1993). Capillary electrophoresis offers several advantages over other analytical techniques, including simplicity, reduced matrix dependence, and enhanced separation efficiency, together with a separation selectivity different from that achievable with conventional ion-exchange separations (Jackson and Haddad, 1993). Valuable information on the nature of chemical species in soil extracts can be obtained through thoughtful use of CE. Because CE offers such a potential wealth of information on the chemical nature of ionic species in soil solutions, this chapter discusses its attributes and limitations to familiarize soil scientists with it. Thus the objective of this chapter is to review the principles, practices, and applications of CE as it relates to the determination of inorganic anions, cations, and low-molecular-weight organic solutes, with a view to using the technique for the analysis of soil and wastewater samples.

134

R. NAIDU ET AL.

II. GENERAL PRINCIPLES Capillary electrophoresis is a relatively simple analytical technique based on the separation of ions in charged, narrow-diameter capillaries and their subsequent detection, typically using either UV or fluorescence detectors. Capillary electrophoresis has several separation modes: capillary zone electrophoresis (CZE), isotachophoresis, micellar electrokinetic capillary chromatography, capillary gel electrophoresis, and isoelectric focusing. Of these techniques, CZE is by far the most applicable to the separation of inorganic ions; consequently only this approach is discussed in this review. Capillary zone electrophoresis has gained considerable attention over the past decade, particularly in the analysis of drugs, proteins, and more recently inorganic cations and anions. Figure 1 shows a schematic diagram of the basic CZE instrument, which consists of a high-voltage power supply, two buffer reservoirs, a narrow-diameter capillary, and a detector. The capillary containing the buffer solution is placed between two buffer reservoirs and a potential is applied across the capillary. Electroosmosis causes the flow of bulk solvent in the charged capillary when the potential is applied (Rice and Whitehead, 1965).The applied potential causes the ions to migrate at varying speeds in the homogeneous electrophoretic buffer, providing a discrete moving zone of the analyte. In a conventional CZE system, the electroosmotic flow is directed toward the cathode, where a detector is placed to quantify the solute. The separation mechanism is based on the differences in solute size and charge at a given pH, which

Figure 1 Schematic diagram of capillary zone electrophoresis.

CAPILLARY ELECTROPHORESIS AND ANION SPECIATION

135

cause different analyte mobilities. Because any charged species can move in an electric field, CZE has been used to separate widely different compounds ranging from small cations such as Li (Chen and Cassidy, 1993) to polymers with molecular weights of 2 million (Richmond and Yeung, 1993). Capillaries used for CZE are most commonly made of fused silica. In the presence of an appropriate buffer, the surface of the silica tube hydrolyzes, and at pH values above approximately 2.5, silica dissociates to form negatively charged silanol groups. This leads to the formation of a diffuse double layer consisting of the negatively charged surface, the immobile layer (Stem layer or Helmhotz layer), and the diffuse layer of cations adjacent to the surface of the silica (Fig. 2). The cationic counterions in the diffuse double layer extend into the mobile bulk electrophoretic buffer solution. Because of the polarity difference across the ends of the capillary, the cations in the diffuse layer adjacent to the surface of the silica migrate toward the cathode. Because these ions are solvated, they drag bulk electrolyte with them (Ewing et al., 1989). This flow of liquid through the capillary is called electroosmotic flow. The rate of flow is governed by the extent of the potential drop across the double layer. The electroosmotic flow is usually significantly greater than the electrophoretic mobility of individual ions in the injected samples, which enables the separation of both anions and cations in the same run. Cations are attracted toward the cathode and their speed is increased by the electroosmotic flow. Although anions are electrophoretically attracted toward the anode, they are swept toward the cathode along with the bulk flow of the electrophoretic medium. This results in separation of cations first, then unresolved neutral species, followed by anions. The basis of the separation is discussed in the following sections.

+ Figure 2 Schematic diagram of the double layer on the capillary surface.

136

R. NAIDU ETAL.

111. SAMPLE INTRODUCTION Injection modes in CE include hydrostatic (gravity) sample introduction, the application of either pressure or vacuum, and electromigration (Olechno er al., 1990). Hydrostatic injection involves immersion of the inlet end of the capillary into the sample solution and raising the height of the sample, causing it to enter the capillary by a siphoning effect. This mode of sample injection has been found to yield more reproducible results than either positive pressure or vacuum injection. In the latter modes, either a vacuum is applied to the outlet, or pressure is applied to the inlet, in order to introduce sample into the capillary. Sample introduction by the hydrostatic mode is the most commonly used injection mode for the CZE of ionic solutes, and peak area reproducibilities on the order of 1-3% RSD for ionic solutes at the sub milligram per liter level have been reported (Jones and Jandik, 1991;Grocott et al., 1992). Sample introduction by electromigration injection involves the application of voltage after the insertion of both the input end of the capillary and the electrode into the sample solution. Electromigration injection includes contributions from both electrophoretic migration of charged sample ions and electroosmotic flow of the sample solution. With this mode of injection, sample ions with a charge opposite to that of the detection electrode migrate into the capillary (Olechno et al., 1990) and the rate of migration is proportional to the total mobility of each ion (Huang er al., 1988). Contrary to hydrostatic injection, where reproducibility is high, electromigration injection is subject to variability, largely due to the bias associated with the differences in the mobilities associated with the analyte ions. This stems from the increased injection of the more mobile relative to the less mobile species into the capillary when the voltage is applied. Additionally, differences in the conductivity between the sample solution and the operating buffer can also cause bias (Li, 1992). In the electromigration mode of injection, another phenomenon known as sample stacking can result, especially when the conductivity of the injected sample is lower than that of the surrounding buffer, which is typically the case. This results in the focusing of the analyte ions into zones within the capillary (Vinther and Soeberg, 1991a; Burgi and Chien, 1991; Jones and Jandik, 1991). This compression of the analyte zone has been attributed to the inverse dependence of the electric field on specific conductivity, because lower ionic strength increases the resistivity of the sample leading to increased electric field strength. Such an effect leads to on-column concentration of the analyte at the sample-buffer interface (Burgi and Chien, 1991; Jackson and Haddad, 1993). This effect has been described as sample stacking and has been successfully used to analyze anions and organic acids at low microgram per liter levels in ultrapure water (Jones and Jandik, 1991) and environmental samples (Jackson and Haddad, 1992; Li and Li, 1995). Sample stacking has been reviewed by Li (1 992).

CAPILLARY ELECTROPHORESIS AND ANION SPECIATION

13 7

Iv. SEPARATIONS There are considerable differences in the modes of separation between CZE and ion chromatography. Whereas separations by the latter technique are typically based on ion-exchange interactions, the selectivity of CZE is based on the ionic mobilities of the solutes. Consequently, the elution sequences for CZE are often significantly different from those attained for the same solutes using IC. The separation in CZE also differs from IC in that net elution time is the result of two mobility vectors, that of the bulk electroosmotic flow and the electrophoretic mobility of individual solutes.

A. ELECTROOSMOTIC MIGRATION Electroosmotic mobility causes bulk fluid flow (Olechno et al., 1990), which arises from the zeta (5) potential at the interface between the surface of the silica capillary and the electrolyte where, at most pH values, an excess of anionic species exists in the static double layer (Fig. 2). The zeta potential across this layer, which is often referred to as either the Stern or Helmhotz layer, is given by the Helmhotz equation,

5

= (47FT)Feok

(1)

where 7 is the viscosity, E is the dielectric constant of the electrolyte solution, and pe0 is the coefficient for electroosmotic flow (Hjerten et al., 1965).The presence of an anionic charge on the silica capillary and the static diffuse double layer produces an excess of cations in the bulk solution that migrate toward the cathode on the application of a potential. Because these ions are solvated, they drag the bulk solvent/electrolyte with them, producing a flat velocity distribution across the capillary diameter (Kuhr, 1990),as shown in Fig. 3. The velocity of the electroosmotic flow (u)is given by u = ~E5/4nq,

A Flat Profile

(2)

B Parabolic Profile

.q .q Figure 3 Comparison of flow profiles in (A) capillary electrophoresis and (B) high-performance liquid chromatography.

138

R.NAIDU ETAL.

where E is the applied potential field strength. Since electroosmoticflow originates at the inner wall of the capillary, the appearance of the flow profile changes from parabolic to flat as the capillary radius becomes greater than seven times the double layer thickness, d (Stevens and Cortes, 1983). The thickness of the double layM (Ewer ranges from 3 to 300 nm for electrolyte concentrationsof lo-* to ing et al., 1989). This flat flow profile results in a high separation efficiency compared to HPLC, which is based on pressure-driven pumping and has a parabolic flow profile (Li, 1992).

B. LONGITUDINAL DIFFUSION Because of the narrow diameter of the capillary and the absence of a pressuredriven, parabolic flow, band broadening in CZE capillaries is minimized. Moreover, electroosmoticcharge separation within the narrow-diametercapillaries (see below) results in very thin double layer thickness, leading to a flat electroosmotic flow profile (Fig. 3). According to Stevens and Cortes (1983), such profiles are expected when the capillary radius is greater than seven times the double layer thickness. However, Jorgenson and Lukacs (1981a,b) report that even with flat electroosmotic flow, longitudinal diffusion could still lead to some zone broadening. Li (1992) reported that if the only contribution to band broadening is longitudinal diffusion, then the variance (m2) of the migrating zone width is given by

u2 = 2Dt

(3)

= 2DL2/ke,V,

(4)

where D is the diffusion coefficient of the solute, V is the applied voltage, L is the length of the capillary, and t is time. The diffusion coefficient of the solute is inversely related to the number of theoretical plates (N), which is given by

N = pe,V/2D = L2/u2. From Eq. ( 5 ) , it can be seen that the number of theoretical plates is directly proportional to the applied voltage. Thus high voltages can result in a very large number of theoretical plates. Normal operating conditions routinely used in CZE produce N values on the order of 50,000 to 250,000 plates. Therefore, large applied voltages favor efficient separations, but in practice, there is a limitation due to the Joule heating that occurs when an electric current passes through the electrolyte. However, capillary tubes with small inner diameters have a high surface to volume ratio that facilitates heat dissipation. Because the number of theoretical plates is extremely high for CZE, its resolving power and peak capacity are better than those of other analytical separation techniques, such as HPLC.

CAPILLARY ELECTROPHORESIS AND ANION SPECIATION

139

C. ELECTROPHORETIC MIGRATION Electrophoretic separations resolve ions on the basis of charge; anions move toward the anode and cations toward the cathode. Mobility is based on particle charge and Stokes radius. In the absence of electroosmosis, the electrophoretic velocity and time taken for migration are given by u = p ePE = p,,VlL t = L2/pe,V,

where pepis the electrophoretic mobility.

D. ANALXTEMOBILITY Because the applied voltage and the capillary size are the same for all analytes, the factor that determines the separation is the mobility of a particular analyte, which is the net result of the vectors described by both electrophoretic and electroosmotic mobilities. Thus total analyte velocity is described by = (Fee

+ II.,,)V/L

(8)

and t = L2/peo + pep)V,

(9)

where p is analyte mobility in cm2V-' s-I and peois electroosmotic mobility. Analyte mobility is the most important parameter determining separation efficiency and selectivity. Factors such as pH, ionic strength, potential drop across the capillary, and temperature can be manipulated to improve separations through changes in analyte mobility. Jorgenson and Lukacs (1981~)studied the effect of pH and current density on electroosmotic flow (EOF). They found that the rate of electroosmotic flow was highest under conditions that increased the zeta potential or double layer thickness or decreased the solution viscosity. The EOF was also increased in the direction of the cathode with increasing pH, the result of greater surface charge density due to increasing dissociation of surface silanols at higher pH values. Because electroosmosis is influenced by the status of the capillary inner surface (Lambert and Middleton, 1990), it is important not only to keep the buffer characteristics under control but also to ensure good maintenance of the capillary. Electrolyte pH can be controlled by selecting a buffer with adequate buffering capacity at the appropriate pH and by replenishing or frequently changing the inlet and the outlet electrolyte reservoirs (McLaughlin et al., 1992). Our experience shows that rinsing of the capillary between runs using sodium hydroxide, wa-

140

R.NAIDU ETAL.

ter, buffer, detergents, etc., helps maintain the inner silica surface in a standardized state. The use of internal standards in CZE have also been advocated for the calculation of corrected migration time. Altria et al. (1994) report that the precision that can be achieved is dependent on how well matched the mobilities of the analyte and the internal standard are. Analyte mobility can also be modified through changes in electroosmotic flow by the introduction of additives to the buffer solution. Such additives act by alterating the zeta potential developed across the capillary-solution interface. Some examples include the addition of cationic surfactants such as cetyltrimethyl ammonium bromide (CTAB), cation complexing agents (8-hydroxyquinoline-5-sulfonic acid), along with Triton X- 100 and putrescine. However, analyte velocity can also be influenced by a number of other experimental parameters. For instance, Ewing et al. (1989) reported that in normal or large-scale electrophoresis, heating of the solution due to the current carried between the electrodes can impair separation. Such heating can result in density and temperature gradients that can increase both zone broadening and evaporation of solvent. This, in turn, affects analyte velocity, and hence migration time. This can be minimized considerably by the use of a narrow-bore capillary, which enhances heat dissipation through the capillary wall. The high inner surface area to volume ratios in narrow-bore capillaries provide more efficient heat dissipation relative to traditional electrophoresis systems. Such capillaries allow the use of a high potential field strength (100-900 Vcm- '), which provides faster migration and EOF rates leading to rapid, highly efficient separations.

E. SEPARATION OF POSITIVELY CHARGED IONICSOLUTES While cations, neutrals, and anions can often be separated in one run in CZE, the very high (opposite) mobilities of small inorganic cations and anions typically dictate that different approaches are required for each solute type. Cationic solutes are typically separated by allowing them to migrate in the same direction as the electroosmotic flow, which often results in highly efficient separations. The electroosmotic flow is toward the cathode at most pH values and detection is carried out at this end of the capillary. Unfortunately, the differences in ionic mobilities between metal cations of similar charge are not generally sufficient to permit their separation (Jones and Jandik, 1991; Jackson and Haddad, 1993). Additional separation selectivity in CZE can be achieved through the addition of a complexing agent to the electrolyte (Jackson and Haddad, 1993). As shown in Eq. (lo), complexing agents can be used to control the net (effective) charge of metal ions depending on the metal-ligand conditional formation constants [Eq. (1 l)], while the observed electro-phoretic mobility of the free metal ion and various complexes is described by Eq. ( I 2):

CAPILLARY ELECTROPHORESIS AND ANION SPECIATION Mm+

PML w e , Obs

=

+ L l - eML(t2-I)

[ML'"- " ] / [ M " + ] [ L ' - ]

= 'mPM

+ 'MLIJ'ML

+ xMLZkML27

141

(10) (11) (12)

where wM, p,ML, and kML2are the mobilities of free metal ion, 1:1, and 1:2 M-L complex and X M , XML, and XML2are the mole fractions of each species in the capillary. These mole fractions are a function of electrophoretic buffer parameters that influence the metal-ligand equilibrium described in Eq. (10). Organic acids such as lactic, citric, oxalic, and hydroxyisobutyric acids have been shown to be suitable complexing agents (Foret et al., 1990; Weston et al., 1992b; Chen and Cassidy, 1992; and Simunicovi et. al., 1994). A key factor in the use of a complexing agent is the presence of a chromophore (typically an aromatic base) in the electrolyte solution as a detection probe. As most metal cations have no intrinsic absorbance, indirect UV absorbance is the most commonly used mode of detection (Jackson and Haddad, 1993). Using tropolone as a selectivity modifier, Weston et al. (1992b) obtained separation of K, Ba, Sr, Ca, Na, Mg, and Li in the milligram per liter range (see Fig. 5). A much better sensitivity was obtained by Swaile and Sepaniak (1991) who used 8-hydroxyquinoline-5-sulfonic acid in the mobile phase as a complexing agent to enhance both the selectivity and sensitivity of certain divalent metal cations when using a laser-based fluorescence detector. They reported that by controlling mobile-phase parameters that affect the complexation reaction, the observed electrophoretic mobility of the metal can be manipulated. Using this technique, limits of detection in the microgram per liter range were achieved for Ca2' and Mg2+. One of the major limitations of the CZE technique is the low detection limits of cations, particularly for trace metals of environmental significance. For example, trace metals such as Cd, Cu, Zn, Pb, which often exist in concentrations that span the milligram to microgram per liter range, are below the detection limits of most of the detectors used by CZE. However, the addition of complexing agents as reported by numerous investigators (Swaile and Sepaniak, 1991; Timerbaev et al., 1994; Lin et al., 1993) has considerable merits for trace metal ion determination. Details on the effects of various electrolyte parameters, including the use of various types of capillaries, have been reported by numerous investigators, and readers are directed to the work of Weston et al. (1992a,b) and Chen and Cassidy (1992).

F. SEPARATION OF NEGATIVELY CHARGED IONICSOLUTES Anion speciation and analyses have gained greater popularity than cation analyses with CZE, as is also the case with IC. As with cations, anion separation is based on mobility differences between analyte species in the capillary following the ap-

R. NAIDU ET AL.

142

A ElectrophoreticMobility, pm

+

------++

-+

-4-

I

B Electroosrnotic Mobility, 10 -

+

-

-

I

-

-

-

+++++++++ +++++++++

-

+

-

I

C Direction of Analyte Migration

Krn and PO.

+

----

-+

++ I

Figure 4 The direction of analyte migration is the result of electrophoretic and electroosmotic mobility.

plication of potential. In the presence of an applied potential, anionic solutes migrate toward the anode, that is, in the opposite direction to the electroosmoticflow (Fig. 4) and away from the detection end of the capillary. Such a migration fits in with the separation of large molecules such as proteins and peptides, where the bulk electrolyte flow toward the cathode pushes the analyte through to the detector. Thus the net analyte velocity, which is the vector sum of the electroosmotic flow and electrophoretic mobility [Eq. (8)], is in the direction of electroosmotic flow. However, in the case of highly mobile anions such as C1- and NO,-, the electrophoretic mobility can exceed the electroosmotic flow, and therefore such analytes do not reach the detector. As discussed above, analyte mobility can be controlled by inclusionof additives.Jackson andHaddad (1993)reported that, for highly mobile anionic solutes, addition to the electrolyte buffer solution of a cationic surfactant, such as CTAB, has the effect of reversing the direction of electroosmotic flow so that it flows from the negative to the positive electrode. Changing the polarity of the capillary by the addition of cationic surfactant and reversing the applied potential polarity, such that the cathode is at the inlet and the anode at the outlet end, ensures that the anionic solutes move in the same direction as the electroosmotic flow. Such a setup allows highly mobile anionic solutes to reach the detector (Huang et aE., 1993; Altria and Simpson, 1987), leading to rapid, highefficiency separationsof inorganic anions and organic acids (Romano et al., 1991).

CAPILLARY ELECTROPHORESIS AND ANION SPECIATION

143

This setup has been successfully used for anion analyses in pure and environmental samples, such as wastewaters and C1- in urine (Romano et al., 1991; Romano, 1993) and C1- and lactic acid in wine (Cotter et al., 1995).

V. DETECTION As is the case with HPLC, the most widely used detection approach in CZE is UV absorbance. However, fluorescence, electrochemical, radiometric, and even mass spectrometric measurements are also used. Jackson and Haddad (1993) report that in the case of ionic solutes, conductivity, indirect fluorescence, and indirect UV absorbance have been employed, although detection is normally carried out using indirect or “vacancy” spectroscopic methods. In CE, UV absorption detection is typically performed on-line with the capillary itself (after removal of the polyimide coating) acting as the detector cell. Although this is a very straightforward approach, its sensitivity is limited in that the capillary diameter (generally 50- 100 Fm) defines the path length. One of the constrains of using CZE in samples with ultra low concentrations is the limitation of the existing detection techniques, which make it relatively insensitive compared to techniques such as ICP-MS and HPLC-MS. On-column combination of CZE with either of these analytical techniques offers the potential to make it one of the most powerful techniques in contaminant analysis.

VI. COMPARISON WITH OTHER ANALYTICAL TECHNIQUES The technique of CZE has often been compared with IC for inorganic solutes and HPLC for organic solute analyses. There are distinct differences in the modes of separation of analyte species among these three techniques. For instance in CZE, if the polarity favors the separation of inorganic anions, cations do not participate during separation, since they travel in the direction opposite to the anions and appear at the cathode. As the common inorganic anions have greater conductances than weak acid anions, they migrate more rapidly and are therefore well resolved (Jones and Jandik, 1990). Neutral solutes are carried along by the electroosmotic flow and have appreciably longer migration times than anionic species. Separations in IC are determined by the type of ion-exchange used; these are generally specific for either anionic or cationic species. Thus selectivities for IC are predominantly based on the composition of the column’s substrate and the ionexchange components either covalently bound or dynamically coated to the sur-

144

R. NAIDU E T A .

face, Additional selectivity refinements are also provided by the eluent (Haddad and Jackson, 1990; Jones and Jandik, 1992). This mode of separation can be a problem, especially with ions of similar exchange capacity and also for weakly retained ions. Such separations can be subject to interference from other weakly retained solutes, such as carboxylic acids (Hannah, 1986). One example of such an interference is demonstrated by the analysis of FF,NO,-, and oxalate using ionexclusion chromatography. These three anions are weakly retained and elute near the void peak where they are subject to interferences from other organic acids, thus making detection of the ions difficult. The simultaneous analysis of inorganic anions and carboxylic acids is clearly one area where CZE is superior to IC and HPLC. Separations in CZE can be manipulated by controlling the buffer electrolyte characteristics. The optimum resolution and detection sensitivity occur when the peak shapes are symmetrical, which is achieved only when the electrolyte ion mobility is similar to analyte ion mobilities (Mikkers et d.,1979). Solute anions with a higher mobility than the electrolyte anion exhibit “fronting” while those with a lower mobility exhibit “tailing.” Figure 5 shows an electropherogram illustrating the differences in peak shapes that can occur in CE. Peak 1 has greater mobility than the electrolyte cation; hence it exhibits fronting. Peak 7 has lower mobility than the electrolyte cation; hence it exhibits tailing. The mobility of Peak 5 is similar to that of the electrolyte cation; hence it is symmetrical. Electrolyte buffer solutions used for optimum separation of analyte species are discussed by Lindeberg (1995a,b). CZE is potentially more attractive than HPLC because of its ability to achieve very high separation efficiencies. The major reason for this is the characteristical-

5

mg 7

3.3

Potassium Barium Strontium Calcium Sodium Magnesium 7 Lithium 1 2 3 4 5 6

L-1

1.0 2.0 1.5 0.7

0.6 0.4 0.2

5.7 Minutes

Figure 5 Electropherogram illustrating peak shape variations in capillary electrophoresis. Reprinted from Weston et al., “Optimization of detection sensitivity in the analysis of inorganic cations by capillary ion electrophoresis using indirect photometric detection,” 1992a. pp. 395-402, with permission from Elsevier Science.

CAPILLARY ELECTROPHORESIS AND ANION SPECIATION

145

ly flat flow profile attained in CZE (Fig. 3). Because HPLC involves pumping fluids under pressure, a parabolic flow profile is maintained under normal operating conditions. This difference in the flow profile shape is the fundamental reason for narrow peaks and potentially better resolution in CZE than HPLC. According to Koh et al. (1993), the ease with which closely related compounds are separated by CZE often facilitates inclusion of internal standards, thus diminishing the need for extensive sample cleanup. However, CZE still lags behind IC and HPLC as an analytical tool because of the overall precision, ruggedness, and large number of accepted (and regulated) methods developed for these techniques. One of the major limitations of CZE is the poor detection limit. Since the capillary itself is used as the detection cell the light path is very short. Until recently, the detection systems were based on conventional UV, visible absorbance, and fluorescence methods. The development of indirect UV systems has increased the range of analyte species capable of determination by CZE (Beck and Engelhardt, 1992); however, sensitivity still remains the major problem, especially with the analysis of trace metals such as Cd, Cu, Zn, and Pb. More recent research has led to the development of optical systems with greater path lengths (Albin et al., 1993), high-intensity UV laser- and fluorescence-based detectors, electrochemical detectors, and on-line mass spectrometry. These new developments make CZE considerably more attractive than the traditional IC and HPLC techniques, as illustrated by the recent investigations of Jones et al. (1995). These researchers showed that simultaneous direct conductivity and UV detection enables quantification of molecules as diverse as bromide, sulfate, acetate, and benzoate in a single analysis. The development of suppressed conductivity detection for CZE offers the ability to achieve sub microgram per liter detection limits for inorganic anions, although this technique has yet to become commercially available (Harrold et al., 1995). One of the greatest advantages of CZE is its tolerance to variations in the nature of the sample matrix, making it useful for the determination of ionic solutes in sample types that have traditionally been difficult to analyze by IC or HPLC, such as petroleum refinery extracts (Romano et al., 1991) and highly alkaline samples (Salomon and Romano, 1992). Compared to the IC and HPLC techniques, CZE does not require any extensive sample preparation, such as derivitization, pH adjustments, or complex mobileimmobile phase preparations. Direct injections of ultra low volumes (nl) make it even more attractive, especially for analysis of those environmentally important samples for which sample size is small. One major limitation of CZE is the loss of resolution and peak distortion that occurs when one component is present in large concentrations compared to others. For example, the presence of large concentrations of C1 in soil water samples has been shown to produce peak distortions (R. Naidu, unpublished results), while total phosphate concentrations in excess of 2 p.g literp2 cause peak displacement. Cancalon (1995a,b) also reported similar constraints during the analysis of citric acid and other organic components in a sample of orange juice. However, such peak displacements and interferences are eas-

146

R.NAIDU ETAL.

ily overcome following dilution (R. Naidu, unpublished results) provided that analyte concentrations do not fall below the detection limit of the instrument.

VII. IMPLICATION FOR THE ANmYSIS OF SOIL SOLUTIONS It is now generally accepted that speciation, in addition to total concentrations, of elements in soil solution is required to give a complete understanding of elemental bioavailability, toxicity, or mobility in soils. It would also be fair to say that our current understanding of such processes is limited by a relative lack of data on speciation of both nutrient and pollutant elements, such as heavy metals, in natural soil solutions. There are probably two main reasons for this lack of data. The first is the difficulty of isolating volumes of soil solution (as distinct from water extracts) in sufficient volume for the complete analysis and speciation of ions. The second is that for many elements suitable analytical techniques capable of directly determining individual ionic species are lacking. Soil scientists interested in speciation are often faced with using speciation schemes that do little more than distinguish broad groups of ionic species, for example, using techniques based on the reactivity of species with exchange resins (Tills and Alloway, 1983; Holm et al., 1995).An alternative approach has been the use of computer speciation models such as GEOCHEM or MINTEQ. Compared to existing techniques and procedures, CZE appears potentially to have distinct advantages in that it enables the rapid direct determination of ionic species using extremely small volumes of solution (<1 pl). Thus, the use of this technique could have considerable implications for improving our understanding of soil solution/solid-phase equilibria, bioavailability and/or toxicity of ionic species, and ion mobility in soils. The availability of accurate speciation data could well provide a significant boost to the various nutrient cycling and solute transport models being developed at present. In some cases, the ability to check the predictions of computer speciation models could also be of considerable benefit. However, a note of caution is perhaps warranted. Analysis by CZE is in its infancy as far as its application to soil solution is concerned. A considerable amount of development work is required to establish the range of species, their limits of detection and possible interferences, for which the technique can be used. Nevertheless, we consider that there is great potential for CZE in soil solution studies; it remains for soil scientists to exploit that potential and develop reliable protocols for its routine use.

REFERENCES Albin, M., Grossman, P. D., and Moring, S. E. (1993). Sensitivity enhancement for capillary electrophoresis. Anal. Chem. 65,489A-497A.

CAPILLARY ELECTROPHORESIS AND ANION SPECIATION

147

Alhia, K. D., and Simpson, C. F. (1987). High voltage capillary zone electrophoresis:Operating parameters effects on electroendosmoticflows and electrophoreticmobilities. Chromatographia 24,527-532. Altria, K.D., Clayton, N. G., Hart, M., Harden, R. C., Hevizi, J., Makwana, J. V., and Portsmouth, M. J. (1994). An inter-company cross validation exercise on capillary electrophoresis testing of dose uniformity of paracetamol content in formulations. Chromatographia 39, 180- 184. Baechmann, K., Boden, J., and Haumann, I. (1992). Indirect fluorimetric detection of alkali and alkaline earth metal ions in capillary zone electrophoresis with cerium (111) as carrier electrolyte. J. Chromatog,: 626,259-265. Beck, W., and Engelhardt, H. (1992). Capillary electrophoresis of organic and inorganic cations with indirect UV detection. Chromatographia 33,313-316. Burgi, D. S., and Chien, R. L. (1991). Optimization in sample stacking for high-performance capillary electrophoresis. Anal. Chem. 63,2042-2047. Cancalon, P. F. (1995a). A new tool in food analysis: Capillary electrophoresis. J. Assoc. OfJ Anal. Chem. 78, 12-15. Cancalon, P. F. ( I995b). Capillary electrophoresis: A useful technique for food analysis. Food Technol. 49,52-58. Chen, M., and Cassidy, R. M. (1993). Separation of metal ions by capillary electrophoresis. J. Chromatogr: 640,425-43 I . Chien, R. L., and Burgi, D. S. (1992a). Sample stacking on extremely large injection volume in highperformance capillary electrophoresis. Anal. Chem. 64, 1046. Chien, R. L., and Burgi, D. S. (1992b). On-column sample concentration using field amplification in CZE. Anal. Chem. 64,489A-496A. Cotter, R. L., Benvenuti, M., and Krol, J. (1995). The use of capillary ion analysis in the analysis of organic acids in potable liquids. Pittsburgh Con$ Abst. 1365. Cutter, G. A. (1985). Determination of selenium speciation in biogenic particles and sediments. Anal. Chem. 57,295 1-2955. Daughtrey, E. H., Fitchett, A. W., and Mushak, P. (1975). Quantitative measurements of inorganic and methyl arsenicals by gas-liquid chromatography. Anal. Chim. Acta 79, 199-206. Ericzon, C., Petterson, J., Andersson, M., and O h , A. (1989). Determination and speciation of selenium in end products from a garbage incinerator. Environ. Sci. Technol. 23, 1524-1528. Ewing, A. G., Wallingford, A. R., and Olefirowicz, T. M. (1989). Capillary electrophoresis.Ana1. G e m . 61,292-303A. Gjerde, D. T., and Fritz, J. S. (1979). Effect of capacity on the behaviour of anion-exchange resins. J . Chromatog,: 176,199-206. Grocott, S. C., Jefferies, L. P., Bowser, T.,Camevale, J., and Jackson, P. E. (1992). Effect of electrolyte composition on the separation of inorganic metal cations by capillary ion electrophoresis.J. Chromatogr: 602,249-256. Haddad, P. R., and Jackson, P. E. (1990). Ion chromatography: Principles and applications. J . Chromatog,: Libr: 46,22-25. Hannah, R. E. (1986). A HPLC anion exclusion method for fluoride determinations in complex effluents. J. Chromatog,: Sci. 24,336-339. Harrold, M., Stillian, I., Bao, L., Rocklin, R., and Avdalovic, N. (1995). Capillary electrophoresis of inorganic anions and organic acids using suppressed conductivity detection: Strategies for selectivity control. J. Chromatogr 717,371-383. Hjerten, S., Jerstedt, S., and Tiselius, A. (1965). Electroplactic particle sieving in polyacrylamide gels as applied to ribosomes. Anal. Chem. 11,211-218. Holm, P. E., Christensen, T. H., Tjell, J. C., and McGrath, S. P. (1 995). Speciation of cadmium and zinc with application to soil solutions. J. Environ. Qual. 24, 183-190. Huang, X., Gordon, M. J., and Zare, R. N. (1988). Current-monitoring method for measuring the electroosmotic flow rate in capillary zone electrophoresis. Anal. Chem. 60, 1837-1840.

R. NAIDU ETAL. Jackson, P. E., and Haddad, P. R. (1992). Optimization of injection technique in capillary ion electrophoresis for determination of trace-level anions in environmental samples. J. Chromatog,: 640, 48 1-487. Jackson, P. E., and Haddad, P. R. (1993). Capillary electrophoresis of inorganic ions and low-molecular-mass ionic solutes. Trends Anal. Chem. 12,23 1-238. Jones, W. R., and Jandik, P. (1990). New methods for chromatographic separations of anions. Am. Lab. 22,s 1-55. Jones, W. R., and Jandik, P.(1991). Controlled changes of selectivity in the separation of ions by capillary electrophoresis. J. Chromatog,: 546,445-458. Jones, W. R., and Jandik, P. (1992). Various approaches to analysis of difficult sample matrices of anions using capillary ion electrophoresis. J. Chromatog,: 608,385-393. Jones, W. R., Soglia, J., and McGlynn, M. (1995). Method development approaches for capillary ion electrophoresis using conductivity and uv detection. Pittsburgh Con$ Abst. 706. Jorgenson, J. W., and Lukacs, K. D. (1981a). Zone electrophoresis in open-tubular glass capillaries. Anal. Chem. 53,1298-1502. Jorgenson, J. W., and Lukacs, K. D. (1981b). High-resolution separtins based on electrophoresis and electroosmosis. J. Chromatog,:218,209-216. Jorgenson, J. W., and Lukacs, K. D. (1981~).High resolution chromatography. Chromatogr.Comrnun. 4,230-23 1. Koberda, M., Konkowski, M., Yangberg, P., Jones, W. R., and Weston, A. (1992). Capillary electrophoretic determination of alkali and alkaline-earth cations in various multiple electrolyte solutions for parenteral use. J. Chromatog,: 602,235-240. Koh, E. V., Bissell, M. G., and Ito, R. K. (1993). Measurement of vitamin C by capillary electrophoresis in biological fluids and fruit beverages using a steroisomer as an internal standard. J. Chromafog,: 633,245-2.50. Kuhr, W. G. (1990). Capillary electrophoresis. Anal. Chem. 62,403R-414R. Laintz, K. E., Yu, J. J., and Wai, C. M. (1992a). Separation of metal ions with sodium bis(trifluoroethy1) dithiocarbamate chelation and supercritical fluid chromatography. Anal. Chem. 64,311-3 15. Laintz, K. E., Sjieh, J. M., and Wai, C. M. (l992b). Simultaneous determination of arsenic and antimony species in environmental samples using bis(trifluoroethy1) dithiocarbamate chelation and supercritical fluid chromatography. J. Chromafog,:Sci. 30, 120- 123. Lambert, W. J., and Middleton, D. L. (1990). pH hysteresis effect with silica capillaries in capillary zone electrophoresis. Anal. Chem. 62, 1585- 1587. Li, K., and Li, S. F. Y. (1995). Speciation of selenium and arsenic compounds in natural waters by capillary zone electrophoresis after on-column preconcentration with field-amplified injection. Analyst 120,361-366. Li, S. E Y. (1992). “Capillary Electrophoresis: Principles, Practice and Applications.” Elsevier, New York. Lin. T. I., Lee, Y. H., and Chen, Y. C. (1993). Capillary electrophoretic analysis of inorganic cations: Role of complexing agent and buffer pH. J. Chromatog,: 654, 167-176. Lindeberg, J. (1995a). Capillary electrophoresis in food analysis. Food Chem. 55,73-94. Lindeberg, J. (1995b). Addendum to “Capillary electrophoresis in food analysis.” Food Chem. 55, 95-101. McLaughlin, G. M., Nolan, J. A., Lindahl, J. L., Palmieri, R. H., Anderson, K. W., Morris, S. C., Morrison, J. A., and Bronzert, T. J. (1992). Pharmaceutical drug separations by HPCE: Practical guidelines. J. Liquid Chrornatogr: 15,961- 1021, Mentasti, E., Nicolotti, A., Porta, V., and Sarzanini, C. (1989). Comparison of different preconcentration methods for the determination of trace levels of arsenic, cadmium, copper, mercury, lead and selenium. Anulyst 114, 11 13-11 17. Mikkers, E E. P., Everaerts, F. M., and Verheggen, Th. P. E. M. (1979). High performance zone electrophoresis. J. Chromatog,: 169, 11-20.

CAPILLARY ELECTROPHORESIS AND ANION SPECIATION

149

Monning, C. A., and Kennedy, R. T. (1994). Capillary electrophoresis. Anal. Chem. 66,280R-314R. Morawski, J., Alden, P., and Sims, A. (1993). Analysis of cationic nutrients from foods by ion chromatography. J. Chrornarogr:640,359-364. Morin, P., Amran, M. B., Favier, S., Heimburger, R., and Leroy, M. (1992). Separation of arsenic anions by capillary zone electrophoresis with UV detection. Fresenius J. Anal. Chem. 342, 357-362. Naidu, R., Smith, J., McLaren, R. G., Stevens, D. P., Sumner, M. E., and Jackson, P. E. (1999). Capillary zone electrophoresis. Speciation of arsenic in natural soil water samples. Accepted for publication. Olechno, J. D., Tso, J. M. Y.,Thayer, J., and Wainright, A. (1990). Capillary electrophoresis: A multifaceted technique for analytical density 2. Separation and injection. Am. Lab. 22,30-37. Olieman, C. (1 993). Capillaire elektroforese bij zuivelonderzoek. Voedingsmiddelentech.26, 10- 1 1 . Rice, C. L., and Whitehead, R. (1965). Electrokinetic flow in a narrow cylindrical capillary. J. Phys. Chem. 69,4017-4024. Richmond, M. D., and Yeung, E. S. (1993). Development of laser-excited indirect fluorescence detection for high molecular weight polysaccharide in capillary electrophoresis. Anal. Biochem. 210, 245-248. Romano, J. P. (1993). Capillary ion analysis: A method for determining ions in water and solid waste leachates. Millipore Waters Chromatography No. TI 6 I , American Laboratory. Romano, J., Jandik, P., Jones, W. R., and Jackson, P. E. (1991). Optimization of inorganic capillary electrophoresis for the analysis of anionic solutes in real samples. J. Chromatogr: 546,411-421. Salomon, D. R., and Romano, J. (1992). Applications for capillary ion electrophoresis in the pulp and paper industry. J. Chromatogr: 602,219-225. Shi, Y. C., and Fritz, J. S. (1994).New electrolyte systems for the determination of metals by capillary zone electrophoresis. J. Chromatogr:671,429-435. Simunicov&,E., Kaniansky, D.. and Loksikova, K. (1994). Separation of alkali and alkaline earth metal and ammonium cations by capillary zone electrophoresis with indirect UV absorbance detection. J. Chrornatogr. 665,203-209. Small, H., Stevens, T., and Bauman, W. (1975). Novel ion exchange chromatographic method using conductimetric detection. Anal. Chem. 47, 1801-1809. Stevens, T. S., and Cortes, H. J. (1983). Electroosmotic propulsion of eluent through silica based chromatographic method. Anal. Chem. 55, 1365-1 370. Swaile, D. F., and Sepaniak, M. J. (1991).Determination of metal ions by capillary zone electrophoresis with on-column chelation using 8-hydroxyquinoline-5-sulfonicacid. Anal. Chem. 63, 179- 184. Tennery, V. C., and Wells, R. J. (1994). The analysis of illicit heroine seizures by capillary zone electrophoresis. J. Chromatogr: Sci. 32, 1-6. Tills, A. R., and Alloway, B. J. (1983).The use of liquid chromatography in the study of cadmium speciation in soil solutions from polluted soils. J. Soil Sci. 34,769-781. Timerbaev, A. R., Semenova, 0. P., Jandik, P., and Bonn, G. K. (1994). Metal ion capillary electrophoresis with direct UV detection: Effect of a charged surfactant on the migration behavior of metal chelates. J. Chrornatogr: 671,419-427. van den Berg, C. M. G . , and Khan, S. H. (1990). Determination of selenium in seawater by adsorptive cathodic stripping voltammetry. Anal. Chim. Acta 231,221-229. Vinther, A,, and Soeberg, H. ( 199la). Mathematical model describing dispersion in free solution capillary electrophoresis under stacking conditions. J. Chrornatogc 559,3 -26. Vinther, A., and Soeberg, H. (199 I b). Temperature elevations of the sample zone in free solution capillary electrophoresis under stacking conditions. J. Chromarogr:559, 27-42. Weston, A,, Brown, P. R., Jandik, P., Heckenberg, A. L., and Jones, W. R. (1992a).Optimization of detection sensitivity in the analysis of inorganic cations by capillary ion electrophoresis using indirect photometric detection. J. Chrornatogr: 608,395 -402.

150

R.NAIDU ETAL.

Weston, A,, Brown, P. R., Jandik, P., Jones, W. R., and Heckenberg, A. L. (1992b).Factors affecting the separation of inorganic metal cations by capillary electrophoresis. J. Chromatogs 593,289-29s. Wildman, W. 3.. Jackson, P. E., Jones, W. R., and Alden, P. G. (1991).Analysis of anion constituents of urine by inorganic capillary electrophoresis. J. Chromatogs 546,459-466. Yu, J. J., and Wai, C. M. (1991).Chromatographic separation of arsenic species with sodium bis(trifluoroethy1)dithiocarbamatechelation. Anal. Chem. 63,842- 847. Zeece, M. (1992).Capillary electrophoresis: A new analytical tool for food science. Trends Food Sci. Technol. 3,6-10.

ADVANCESIN SOLUTIONCULTURE METHODSFOR PLANT~ R A L NUTRITION RESEARCH David R. Parker’ and Wendell A. Norvel12 ‘Soil and Water Sciences Section Department of Environmental Sciences University of California Riverside, California 9252 1 2USDA Plant, Soil and Nutrition Laboratory Ithaca, New York 14853

I. Introduction 11. Soil Solutions and Nutrient Solutions IU. Advances in Solution Culture Methods for Controlling Nutrient Status A. Periodic Replacement or Addition of Nutrients B. Stability Derived from Large Solution Volumes and/or Flowing Culture C. Buffering p H and Metal Ion Activities with Soluble Ligands D. Buffering with Ion-Exchange or Chelating Resins E. Buffering with Inorganic Solid Phases IV Summary and Future Outlook References

I. INTRODUCTION Since the middle of the 19th century, the experimental culture of plants in soilless media has become a widespread approach to the study of a host of physiological phenomena, especially those centered around root activity and function (Asher and Edwards, 1983). The methods can be collectively termed “solution culture” or “water culture” because mineral nutrients are, in the main, provided as dissolved constituents of the aqueous phase. In addition to the culture of plants with roots completely immersed in solution, other applications include the culture of roots in inert sand or gravel through which the nutrient solutions are perfused, as well as “mist” or “aeroponic” culture, in which the roots are suspended in moist air and are periodically or continuously sprayed with fine droplets of nutrient solution. IS1 Vdwm 6 5 Copyright 0 I009 by .4carlemic Press. All rights of reproduction in any form reserved o n h w I I 3/99 $Jn.oo Adumuu

11) Aronomy,

152

DAVID R. PARKER AND WENDELL A. NORVELL

Solution culture methods are widely used for studying root physiology, and are especially critical to the study of plant mineral nutrition. They provide the obvious advantage of a well-defined, homogeneous, and fully controllable medium in terms of its elemental composition. Moreover, solution culture methods allow ready examination and recovery of the roots themselves and minimize populations of pathogens and other microbes. Solution culture methods are also widely and successfully used in studies of other edaphic factors such as plant-water relations, root anoxia/hypoxia, and elemental phytotoxicities (Asher and Edwards, 1983). As with most experimental techniques, solution culture has drawbacks as well as advantages. Notable among the former is the fact that solution culture must necessarily be an imperfect simulation of plant growth in soil. Among the obvious differences between soil and solution environments is the fundamental difference in the mechanical support of the plant. In the absence of soil (or at least sand or gravel), the “propping” effect of the roots and crown must be accomplished in other ways (Asher and Edwards, 1983). Further, in most solution culture systems, the roots are fully submerged and the solutions are well mixed and aerated, which further alters both the chemical and physical environment in the rhizosphere. In such systems, the diffusive fluxes of 0,, water, and nutrients that occur in soils are not easily simulated. Considering these differences, it is not surprising that root morphology often differs between soil- and solution-grown plants. Some physiological characteristics of roots differ as well. There is convincing evidence that root exudation of organic solutes is higher in the presence of root mechanical impedance, so that study of strictly solution-cultured plants can lead to underestimates of the quantities of exudates (Marschner, 1995). Microbial populations are almost certainly smaller in “clean” solution culture, further accentuating differences in the rhizosphere environment relative to field conditions. However, the “failure” of solution culture to simulate the soil environment does not negate its advantages. Instead, the differences should serve to instill caution in interpreting results from solution culture experiments and to remind researchers that such methods may not be suitable for the study of all edaphic factors (Asher and Edwards, 1983). It is worth noting that not only are solution-grown roots morphologically different than soil-grown roots, but also there are marked differences in root morphology among plants grown under otherwise similar conditions in solution culture, sand culture, or mist culture. The lack of literature on this subject is surprising. Our purpose in this chapter is to review some of the advances in solution culture methodology that have occurred in the 15 years since Asher and Edward’s (1983) comprehensive review of the topic. We focus primarily on methodological advancements for the specific study of mineral nutrition, but recognize that the methods may be useful for other types of physiological research as well. We have chosen not to discuss the commercial production of horticultural crops using water culture methods, often termed “hydroponics”; reviews of this topic can be found in Resh (1989), FA0 (1990), and elsewhere.

ADVANCES IN SOLUTION CULTURE METHODS

153

More specifically, we devote much of this review to describing methods for overcoming one of the most fundamental limitations of traditional solution culture, the absence of chemical “buffering” of nutrient concentrations. In soils, the soil solution is depleted of nutrients by plant uptake, but continuously replenished via chemical and biological reactions of the solid phase (Asher and Edwards, 1983). Simulation of these phenomena in solution culture has been a challenge to plant nutritionists for many years. Since Asher and Edward’s (1983) review, there have been a number of attempts to buffer solution composition against plant depletion using synthetic resins, specific mineral phases, and soluble ligands, and we review these efforts in some detail. In addition, we revisit and update the uses of flowing solution cultures, discussed in detail by Wild et al. (1987), and we provide an overview of methods wherein nutrients are continuouslyor intermittently added in proportion to plant growth (Section IIIB). For convenience, abbreviations and chemical names for the chelating agents and pH buffers discussed throughout this paper are summarized in Table I.

Table I Abbreviations and Common Chemical Names for the Synthetic Chelators and Biological pH Buffers Discussed in this Chapter

Chemical name

Abbreviation(s)

Chelators BPDS CDTA (DCTA) DTPA EDDA EDDHA (EHPG,APCA) EDTA EGTA FZ = Ferrozine HBED HEDTA (HEEDTA) HEIDA (HIDA) IDA NTA

Bathophenanthrolinedisulfonate frans-1,2-Cyclohexylenedinitrilotetraacetate Diethylenetrinitrilopentaacetate Ethylenediaminediacetate Ethylenediiminobis[(2-hydroxyphenyl)acetate] Eth ylenedinitrilotetraacetate Ethylenehis(oxyethlenetrinitri1o)tetraacetate 3-(2-Pyridyl)-5,6-diphenyl-I ,2,4-triazine-p,p’-disulfonate

N,N’-Bis(2-hydroxybenzyl)ethylenedinitrilodiacetate N-(2-Hydroxyethyl)ethylenedini trilotriacetate N-( 2-Hydroxyethy1)iminodiacetate lminodiacetate

Nitrilotriacetate

Buffers

ACES ADA BES HEPES MES MOPS TAPS TES

N-( 2-Acetamido)-2-aminoethanesulfonate

N-( 2-Acetamido)-2-iminodiacetate N,N-Bis(2-hydroxyethyl)-2-aminoethanesulfonate N-(2-Hydroxyethyl)piperazine-N’-2-ethanesulfonate 2-(N-Morpholino)ethanesulfonate 3-(N-Morpholino)propanesulfonate N-Tris(hydroxymethyl)methyl-3-aminopropanesulfonate N-Tris(hydroxymethyl)methyl-2-aminoethanesulfonate

154

DAVID R. PARKER AND WENDELL A. NORVELL

11. SOIL SOLUTIONS AND NUTRIENT SOLUTIONS The chemical composition of a nutrient solution should, ideally, mimic the composition of the soil solutions that are relevant to the problem under study. Moreover, it is often desirable to impose low but stable nutrient concentrations that simulate growth-limiting values likely to occur in soil solution. For these purposes, it is instructive to compare reported concentrations of nutrients in soil solutions with those traditionally used in nutrient solution culture. In doing so, however, we recognize that the range of nutrient concentrations reported in soil solutions is large, depends on a host of soil characteristics and management practices, and may vary substantially among solutions obtained and stored using different methods (Ross and Bartlett, 1990; Walworth, 1992; Jones and Edwards, 1993; Lawrence and David, 1996). With these caveats, we now consider some published surveys of soil solution composition. Reisenauer (1966) provided a survey of many previously published papers, and we summarize his results in Figs. 1 and 2. The median value for NO,-N in soil solution was 1.5 mM, with about 90% of the values falling between 0.4 and 5 mM (Fig. 1). Both nitrate and ammonium concentrations are expected to be highly variable, even within a single soil, due to the dynamic nature of the N cycle and the numerous possible biological reactions (mineralization, im-

-

0.1

1.o

10.0

Concentration, mM(NO,-N and K)or m(P)

Figure 1 Distribution of soil solution concentrations of NO,-N, P, and K reported in a survey by Reisenauer (1966). Dashed lines indicate estimated median value for each nutrient.

ADVANCES IN SOLUTION CULTURE METHODS

155

Figure 2 Distribution of soil solution concentrations of S, Ca, and Mg reported in a survey by Reisenauer (1966).Dashed lines indicate estimated median value for each nutrient.

mobilization, nitrification, denitrification). For phosphorus, Reisenauer’s (1966) survey provides a median soil solution concentration of -0.75 pV2 (Fig. l), but with a very wide range, which is not surprising because P solubility is so markedly affected by soil pH and fertilizer practices. In other surveys, Barber et al. (1962, cited in Barber, 1995) reported a modal value of 1.6 pM ( n = 134), while Kovar and Barber (1988) gave a median value of -3.5 for 33 diverse U.S. soils. Thus, the vast majority of soil solution P concentrations are <5 pM, and a majority are probably < 1 pM. For potassium, the Reisenauer (1966) survey provided a median concentration of 1.1 mM (Fig. I), with about 90% of the values falling between 0.3 and 4 mM. The median value of 0.35 mM (n = 33) reported by Kovar and Barber (1990) was lower, as was the modal value of 0. I5 mM reported by Barber et al. (1962, cited in Barber, 1995). For Ca, Reisenauer’s (1966) survey found -20% of the values below 1 mM, with most of the remainder between 1 and 10 mM, and a median of value -1.7 mM (Fig. 2). Barber et al. (1962, cited in Barber, 1995) reported a model Ca concentration of about 0.9 mM. Similarly, Reisenauer’s (1966) survey showed that -90% of soil solution Mg concentrations fell between 1 and 8 mM with a median concentration of about 3.0 mM (Fig. 2), while the Barber et al. study provided a modal value of about I .3 mM (Barber, 1995). Data for soil solution S are seemingly more scarce, but the Reisenauer (1966) survey provides a median value of -0.45 mM(Fig. 2). In the main, the soils in the above-cited surveys were probably dominated by fertile agricultural soils. Native soils that are nutrient poor, especially highly weathered and acidic soils, may have substantially lower soil solution concentra-

-

156

DAVID R. PARKER AND WENDELL A. NORVELL

tions of essential nutrients (Chapin, 1988). These differences are especially obvious with the cations K, Ca, and Mg (Gillman and Bell, 1978; Bruce et al., 1989; David and Lawrence, 1996). Data for the micronutrient metals (Cu, Fe, Mn, Ni, Zn) in soil solution are scarce, in part because the concentrations are so low that they are often difficult to measure in uncontaminated soils. Data summarized by Lindsay (1991) and Welch (1995) suggest that, in aerated and uncontaminated soils, the total concentrations of Cu, Fe, Zn, and Ni rarely exceed 1 p.M, while soluble Mn usually does not exceed 100 p M . Concentrations associated with deficiencies of these metals are orders of magnitude lower. Largely based on experiments using chelator-buffered nutrient solutions (Section IIIC2), the critical free metal ion activities for normal growth have been estimated to be as low as about M for Cu2+, 1O-I’ to M for Fe3+, M for Mn2+, M for Ni2+, and M for Zn2+ (Welch, 1995; Parker et al., 1995a). Over the years, literally hundreds of “recipes” for nutrient solution composition have been employed by various researchers, often with very satisfactory results. In the United States, two of the more popular formulations have been the so-called “Hoagland” and “Johnson” solutions (Hoagland and Amon, 1950; Johnson, 1957). Both usually contain nitrate-N, but can be formulated to include ammonium. These very concentrated solutions are most often used in a more dilute form, that is, at one-quarter or one-half strength as shown in Table 11. In the United Kingdom, the similarly “rich’ Long Ashton solution (Hewitt, 1966) has enjoyed widespread use. Other, modern versions of these “conventional” macronutrient solutions are usually not very different, as exemplified by the “Hohenheim” solution that is included in Table 11. The most notable exceptions to these formulations are the very dilute solutions usually employed in flowing solution culture, a method that has been used to simulate very infertile soils and to identify critical, external concentrations of nutrients (Section IIIB). Two example compositions are provided in Table I1 to illustrate that most nutrient concentrations are at least one order of magnitude lower than those used in conventional Hoagland- or Johnson-type solutions. If we compare the conventional nutrient solutions in Table I1 with the “typical” soil solutions described previously, it is obvious that the nutrient solution concentrations tend to be higher. However, this discrepancy would seldom exceed about one order of magnitude for the major nutrients N, K, Ca, and Mg, and with S it would often be lower. Phosphorus, on the other hand, is several orders of magnitude higher in conventional nutrient solution than in soil solution (Fig. 1, Table 11; also, Asher and Edwards, 1983; Barber, 1995). The other notable discrepancies in chemical composition are the rather high micronutrient concentrations used in many conventional nutrient solutions. We note also that Ca/Mg ratios are typically 2 to 4 in most nutrient solutions, values that are consistent with some soil solution data (Wright et al., 1987; Qian

157

ADVANCES IN SOLUTION CULTURE METHODS Table I1 Composition of Some Commonly Used Nutrient Solutions

Element NO,-N NH,-N P K Ca

Mg S CI

Half-strength Hoagland"

7000 500 500 3000 2000 1000 1000

cu Fe Mn Zn B

Mo

9 0.15 12.5 4.5 0.4 23 0.05

Half-strength Johnson"

7000 1000 1000

3000 2000 500 500 25 0.25 2 2.5 1

12.5 0.05

Hohenheim No. 1'

FSC No. I d

4000 I00 1600 2000 500 1200 I00 0.2 4- I00 0.5 0.5 1-10 0.01

750 100 0.04-25 250 250 I00 I00 100 0.1 2 1 0.5 3 0.02

FSC No. 2' 700

50 1-33 42W70 100 100 125 0.02 5 0.5 0.05 2.5 0.005

Note. The first three are for conventional solution culture with small pots. The latter two are for flowing solution culture (FSC) with very large reservoirs (see Section IIIB). All concentrations are in

w.

Hoagland and Arnon's (1950) No. 2 solution with all nutrients at one-half strength. Johnson er ul.'s (1957) solution with all nutrients at one-half strength. Treeby et u1. (1989); Walter et ul. (1994); von Wirt5n et al. (1996). Asher and Loneragan ( 1967). "Wild et ul. (1974). "

<'

and Wolt, 1990; Walworth, 1992; Curtin and Smilie, 1995). But the large survey data sets summarized above suggest that Ca/Mg is often <1 in soil solution. In general, the reported range in Ca/Mg is at least 0.1 to 10 (Menzies and Bell, 1988; Jones and Edwards, 1993), which suggests that a reasonably wide range in culture solution is also quite acceptable. Very low Ca/Mg values are associated with induced Ca deficiency and/or Mg toxicity (Proctor, 1970). Interestingly, it is difficult to find any explicit rationale for the composition of most traditional nutrient solutions. It is apparent, however, that the composition of nutrient solutions is often more reflective of the chemical composition of plant shoots than of soil solutions. The fact that traditional nutrient solutions are generally more concentrated than soil solutions arises in part because more dilute nutrient solutions are subject to rapid depletion, especially if they are not frequently replaced (Epstein, 1972; Asher and Edwards, 1983). Soil solutions are better buffered by ion-exchange, adsorption-desorption, and dissolution-precipitation reactions, as well as by nutrient cycling and mineralization of organic matter (Ep-

158

DAVID R. PARKER AND WENDELL A. NORVELL

stein, 1972; Lindsay, 1979; Asher and Edwards, 1983). The absence of comparable buffering capacity in traditional nutrient solutions contributes to the use of artificially high concentrations to avoid unforeseen and unwanted nutrient deficiencies in experimental studies of mineral nutrition. Finally, many early formulations specifically attempted to use just three or four salts as macronutrient sources (Hewitt, 1966; Epstein, 1972), and this too restricts the range of feasible nutrient concentration ratios. Because plants regulate their uptake of nutrients effectively, the traditional, high-concentration nutrient solutions have probably persisted because they perform adequately, at least most of the time. That said, it is important to note that scattered evidence has accumulated over the years showing that elemental toxicities can occur in conventional solution culture. Certain species of genotypes can be especially sensitive to such toxicities; some representative examples are summarized in Table 111. Particularly striking is the apparent toxicity of P at very low concentrations relative to those added in conventional Hoagland-type solutions. The presence of relatively high concentrations of micronutrient metals (Cu, Mn, and Zn) can also cause problems, although some “protection” is afforded, albeit inadvertently, by the common use of chelated Fe(III), for example, FeEDTA. At pH values near 6 or above, it is relatively easy for other trace metals to displace Fe(II1) from most chelating ligands because of the low solubility of solid-phase Fe(OH),. The “liberated” chelating agent effectively reduces the bioavailability of other trace metals (Chaney, 1988a; Norvell, 1991; Parker et al., 1995a).This is especially true for Cu (see Section IIIC2), which, in short-term studies using simple growth solutions containing only CaCl, as a major salt, is acutely toxic to root growth at concentrationsbelow 1 p M (D. R. Parker, 1998,unpublished data). Comparable and even higher concentrations are widely used in solutions containing Fechelates, usually without problems. It is also possible that the relatively high concentrationsof Ca and Mg in most nutrient solutions offer some additional protection against toxicities of Cu, Mn, or Zn, perhaps through competitive inhibition of uptake (Robson and Loneragan, 1970; Wainwright and Woolhouse, 1977). Careful and meaningful nutritional physiology research requires stable solution concentrations, often at levels much lower than those in Table 11. In the past two decades, research has been conducted to improve upon the classical solution culture methods developed in the 1940s and 1950s. This research has largely been driven by the following objectives: 1. To realistically simulate the composition of soil solutions to produce nutrient stress similar to that occurring in the field. Benefits include more stable but reduced growth rates, with improved “physiological constancy” in nutrient deficiencies. 2. To avoid elemental toxicities and nutrient imbalances, many of which do not occur in nature. Also to avoid unwanted shifts in nutrient status (e.g., from exces-

S Examples of Nutrient Solution Concentrationsof Essential Elements That Have Been Observed to Cause Toxicity in Sensitive Plant Species or Cultivars

Element P P

Toxic concentration 1600 25

(m

Species Glycine m a (L.) Merr. Lupinus digitatus Forsk., Trifolium subterraneum L., Vulpza myuoros (L.) Gmel. Trifolium subterraneum L.

B cu

2.5 46 2.5

cu

0.2

Hordeum vulgare L. Zea mays L. Triticum aesrivum L.

Mn Mn

9 0.9

Hordeum vulgare L. Hordeum vulgare L.

Mn

0.3

Medicago trunculata, M. tomata

Zn Zn

0.5

Arachis hypogaea L. Trifolium subterraneum L.

P

1

Comments Large genotypic variation Some evidence of incipient toxicity at 5 p M P Correctable by lowering to 1 pM Standard Hoaglands concentration Based on both short- and long-term root growth in absence of Fe-chelate Standard Hoagland’s concentration Toxic threshold depends on environmental conditions Toxic only at neutral pH and low Ca

0.3 pM was marginally toxic

Reference Howell and Bernard (1961) Asher and Loneragan (1967) Asher and Loneragan (1967) Asher and Loneragan (1967) Kim et al. (1985) Williams and Vlamis (1957a) Dragun et al. (1976) D. R. Parker (1988, unpublished results) Williams and Vlamis (1957a) Williams and Vlamis (1957b) Robson and Loneragan ( 1970) Asher (1991) Carroll and Loneragan (1968)

160

DAVID R. PARKER AND WENDELL A. NORVELL

sive to adequate or deficient) due to poor buffering capacity of many nutrient solutions. 3. To successfully produce deficiencies of nutrients for which plant requirements are low (e.g., Cu or Ni), or which are ubiquitous contaminants in the laboratory (e.g., Zn). 4. To develop culture solutions that allow the study of more complex systems under controlled, but realistic, conditions (e.g., mycorrhizal associations or N, fixation by rhizobia). The following section is organized according to type of method, but our emphasis throughout is on advances in controlling and buffering (i) solution pH; (ii) major element concentrations, for which P is an especially problematic case; and (iii) microelements, most notably Cu, Fe, Mn, and Zn. We also consider bicarbonate-CO, chemistry, which is important for Fe nutritional studies. Finally, we discuss some examples of improved solution culture methods for environmental research.

III. ADVANCES IN SOLUTION CULTURE METHODS FOR CONTROLLING NUTRIENT STATUS Culture solutions in which the nutrients are not periodically renewed are of limited utility in mineral nutrition research, because nutrients are depleted and concentrations are unstable (Asher and Edwards, 1983). Highly concentrated solutions resist depletion, at least for a time, but cause other problems, including (i) insolubility of some of the major ions (e.g., Ca-PO, salts); and (ii) concentrations that are unrealistic, unnecessary, or phytotoxic. The use of nonrenewed solutions of limited volume should probably be restricted to undemanding tasks such as for the qualitative demonstration of nutrient essentiality and deficiency symptoms. Our general experience has been, however, that nonrenewed culture is not always satisfactory even for these tasks.

A. PERIODIC REPLACEMENT OR ADDITION OF NUTRIENTS The simplest way to overcome limitations of nonrenewed solutions is through “intermittent renewal” (Asher and Edwards, 1983). The most common approach to such renewal is to periodically transfer the plants to a second set of pots containing fresh solution and to discard the old solutions. The key variables for this approach are (i) the growth rate of the plants, (ii) the solution volume per plant, (iii) the frequency of solution replacement, and (iv) the concentration of each nutrient.

ADVANCES IN SOLUTION CULTURE METHODS

161

Asher and Edwards ( 1 983) presented a simple equation for calculating the frequency with which solutions must be replaced to limit depletion of initial concentrations to some set value (e.g., 5 or 10%). Their calculation can be represented as

where F = required frequency of replacement (in hours); D = the tolerable degree of nutrient depletion (%); V = the solution volume (in liters) per pot (or per plant); W , = the fresh weight of the roots per pot (or per plant); Ci = the initial nutrient concentration (in pM); and U = the expected nutrient uptake rate in pmol g Fwh-’. Because nutrient demand increases over time along with plant size (often in an exponential fashion), the usefulness of results is dependent on the accuracy of the estimates of U and W,. Final plant size may be used to estimate these parameters, but then the necessary frequency of solution change is greatly overestimated early in the experiment. Asher and Edwards (1983) presented example calculations for P and Zn, two elements that are difficult to maintain at stable levels in conventional nutrient solutions. Their calculations yielded F values ranging from 72 s to 117 h, with most of the values falling well below 24 h (i.e., more than one replacement per day). Moreover, if root size becomes large, it is clear that either the frequency of replenishment must increase or much larger pots must be used to lower V /WR. Their calculations represent something of a worst case scenario for the intermittent renewal approach. Our experience has been that the prospects for intermittent nutrient renewal are not as dire as the example above suggests. Major elements other than P can usually be satisfactorily provided at modest concentrations such as those in a “standard” one-half- or one-quarter-strength Hoagland’s or Johnson’s solutions (Section 11). Using solution volumes ranging from 0.5 to 7 liters, we have conducted many growth chamber studies where N, K, Ca, Mg, and S levels were reasonably stable in solution and neither excessive nor deficient in plants. Our objective in many of these studies was to maintain reasonable levels of these macronutrients while micronutrient availability was varied separately. The key to success with intermittent renewal is to conduct experiments of short duration (a few weeks) to limit plant size and to replace the nutrient solutions every few days. The primary challenge for this approach is to maintain P concentrations that are both stable and realistic, as we discuss later in this chapter. We have used chelating agents to regulate the availability of micronutrient metals (Section IIC2), and have stabilized pH using the organic buffer MES (Section IIIC1). Boron and Mo are often relatively “forgiving” in that concentrations that are adequate, nontoxic, and not subject to excessive depletion are also realistic for many soil solutions.

’

162

DAVID R. PARKER AND WENDELL A. NORVELL

As an example of intermittent renewal protocols, we consider experiments in which wheat (Triticum aestivum L.), maize (&a mays L.), tall wheatgrass (Elytrigia pontica [Podp.] Holub), and other species were grown in growth chambers in the senior author's laboratory (Bell et al., 1992; Parker et al., 1997). Solution volumes were typically 3 liters pot-', and the number of plants per pot was varied from 2 to 10 depending on expected growth rate. The target for final dry weights was generally just 2 to 3 g DW pot-', which is adequate to provide high-quality growth data and plant material for elemental analysis. The frequency of solution replacement was varied to reflect the exponential nature of the cumulative growth curve. In a typical experiment lasting 20 days (after seedling germination and transfer to the test solutions), we changed solutions on Days 6, 11, 15, and 18 (i.e., replacement intervals of 6,5,4, and 3 days, followed by harvest 2 days later). The desired P concentration was typically 20 kA4, which, while still high relative to most soil solution values, is much more reasonable than the values of 500 to 2000 FM often employed. Exactly 24 h after each solution replacement, a small sample was taken from each pot and rapidly analyzed for P. Based on comparison with the initial value of 20 p N , daily demand (in Fmol pot-') was calculated and the appropriate amount of P promptly added to each pot. At subsequent 24-h intervals, the addition was repeated until the next solution change, after which a new 24-h analysis was conducted. The addition rate was thus redetermined after each solution change, and it was observed to increase significantly in response to the increase in demand as the experiment progressed. In this example, only four P analyses were required, on Days 7, 12, 16, and 19. Although P concentrations are not highly stable using this approach, deficiency can be avoided and the possibility of P toxicity can be limited to more sensitive species. If P supply per se is the variable of interest, more rigorous approaches are required (see Sections IIIB and IIIE). Studies of Zn deficiency are a special case because of evidence that Zn-deficient plants can hyperaccumulate P to toxic levels from nutrient solutions with P concentrations as low as 10 FM (Parker, 1997). The other macronutrients can often be quite stable using the intermittent renewal approach as outlined here, and the number of required solution changes can be kept to a minimum. Rapid, multielement analyses (i.e., by inductively coupled plasma emission spectroscopy) can be used to verify stability in concentrations of many nutrients. We have found also that electronic spreadsheets are quite useful for calculating hypothetical depletion as a function of estimated plant weight, along with nutrient concentrations in both solution and plant tissue. These simulations allow for the rapid testing of various scenarios that allow us to optimize experimental designs, and forewarn us of likely depletion problems that should be monitored. A simple example for use of large solution volumes is given in Section IIIB. Two closely related techniques, programmed nutrient additions and relative addition rates, involve frequent additions of small quantities of nutrients to each pot of nutrient solution (Asher and Blarney, 1987; Ingestad and Lund, 1986). The ad-

ADVANCES IN SOLUTION CULTURE METHODS

163

ditions may be daily (or every other day), or can be made more-or-less continuously via very slow metering of a stock solution into each vessel. A key feature of both methods is that quantities of nutrients added at each interval increase progressively, usually in an exponential fashion, to match the increase in daily demand caused by exponential plant growth (Fig. 3). To obtain “optimal” growth, one must therefore have considerable foreknowledge of the growth rates under the conditions of the experiment. The programmed nutrient addition (PNA) technique, developed at the University of Queensland, was introduced by Asher and Cowie (1970) and has been subsequently reviewed by Asher and Edwards (1983) and by Asher and Blarney (1987). The method is designed to manipulate plant nutritional status (deficient, adequate, or toxic) by varying the daily additions of one or more nutrients to each pot based on anticipated growth rates in conjunction with “normal” tissue nutrient concentrations (Fig. 3). The nutrient concentration in solution, however, are neither well known nor tightly controlled (Asher and Blarney, 1987).With “one-time’’ additions made every 1 or 2 days, the solution concentration fluctuate widely when plants become large relative to the solution volume. The magnitude of these fluctuations also depends on how accurately the daily additions match the daily demand. With this approach, it is hoped that the solution concentrations remain above the physiological minimum required by the species under study, and this generally seems to be the case (Asher and Blarney, 1987).

5

10

15

Time, days

20

25

Figure 3 Cumulative amounts of an added nutrient according to the programmed nutrient addition or relative addition rate technique. In this example, there are three relative addition rates of 0.10, 0.015, or 0.20 per day, and the additions to the pots are made once per day. The nutrient content of the seedlings at time zero was assumed to be 4 mg per pot, and this value determines the magnitude of all subsequent additions.

164

DAVID R. PARKER AND WENDELL A. NORVELL

The PNA approach avoids the initially high (and potentially toxic) solution concentrations of nonrenewed solutions and similarly avoids changes in nutrient status (i.e., from excessive to adequate to deficient) that can occur even with intermittently renewed solutions. When successful, the nutrient status of the plant can be held more-or-less constant over prolonged periods, or can be manipulated at a specific growth stage if desired (Malik et al., 1978; Asher and Blarney, 1987). Applications of the PNA method have included the production of nutrient-deficient or intoxicated plants, including relatively mature plants, for the purpose of describing and photographing the associated foliar symptoms (Asher and Blarney, 1987). The method is suitable also for establishing the critical plant tissue concentrations associated with deficiency and toxicity (Forno et af., 1979; Howeler et al., 1982). And it has been advocated as a method for supplying P at very low solution concentrations in studies of A1 toxicity, thus minimizing opportunities for Al-phosphate precipitation (Asher and Blarney, 1987). Simplified versions of the PNA approach are widely used by researchers who supplement their culture solutions with one or two nutrients at rates intended to meet plant demands between complete changes of the nutrient solution. The closely related relative addition rate (RAR) technique was developed and evaluated by Ingestad and co-workers in Sweden (Ingestad and Lund, 1979; Ingestad, 1981; Ericsson and Ingestad, 1988), and has been reviewed and touted for its superiority on numerous occasions (Ingestad, 1982; Ingestad and Lund, 1986; Ingestad and Agren, 1988, 1992). In this method, the rate of addition of nutrient is proportional to the amount already accumulated by the plant. This is most frequently done during exponential growth, under conditions where most of the added nutrients are being actively absorbed and utilized by the plant. Instead of attempting to simply match the rate of nutrient uptake, the RAR approach is to consciously set the addition rate at one or more levels as imposed treatments, which in turn control relative growth rates. Almost all of the published studies have been done using N as the growth-limiting nutrient, and many of the early ones used seedlings of deciduous tree species as the test plants. For example, in the work of Ingestad and Lund (1979), a stock solution containing N was automatically metered from a burette into a mist-culture (aeroponic) chamber every 2 h. Every 24 h any small corrections in the total addition rate for that day were made manually, and the addition rate for the subsequent 24-h period was adjusted (by changing the duration of each addition, the stock concentration, or both). In this fashion, the amount of N added each day increased exponentially such that the relative addition rate is constant on a daily basis (Fig. 3). Nutrients other than N are added via (i) automated metering in of mixed stock solutions in response to changes in solution conductivity; or (ii) addition along with, and in fixed proportion to, the added N (Ingestad and Lund, 1979; Ingestad, 1981; Mattson et al., 1991). Implementations of the method in other laboratories have entailed addition methods that are either less (Stadt et al., 1992; Oscarson and Larsson, 1986) or more (Macduff et af., 1993) automated.

ADVANCES IN SOLUTION CULTURE METHODS

165

As advocated by Ingestad, the principal advantage of the RAR method is that it deemphasizes external concentration as a determinant of plant nutritional status, and emphasizes a progressive and exponential increase in the size of the “available pool” of nutrients (Ingestad and Lund, 1979, 1986; Ingestad, 1982). The latter is accomplished via the increasing size of the daily additions and is purported to simulate the exponential increase in soil volume that is explored as a plant root system expands in three dimensions, a time-related phenomenon not otherwise possible in a well-stirred liquid medium. This argument seems to ignore, however, the fact the exponential increase of a root system in a well-buffered nutrient solution would also lead to an exponential increase in root surface area and thus in capacity to absorb nutrients. Other key features of the RAR method include the observation that, after an initial acclimation period of about 1 to 2 weeks, during which the plant adjusts to the newly imposed nutrient status, plants exhibit a relative growth rate (RGR) that is numerically equal to the RAR (Ingestad and Lund, 1979, 1986; Ingestad, 1981, 1982; Oscarson and Larsson, 1986). Classic deficiency symptoms tend to appear during the acclimation phase, but usually disappear once the new, stable growth rate has been established. With N, these “steady-state” symptoms include a slightly paler color in all foliage and longer primary roots with less branching, but not the “classic” chlorosis and senescence of the older leaves (Ingestad and Lund, 1979; Ingestad, 1981; Oscarson and Larsson, 1986; Oscarson et al., 1989). Similarly, the tissue N concentration tends to be constant for a given value of RAR (and RGR), further supporting the notion that the intrinsic nutrient stress level is constant once the acclimation period is complete. In one study, similar results have been obtained with P as the limiting nutrient in birch (Betula pendula Roth.) seedlings (Ericsson and Ingestad, 1988). To our knowledge, however, other nutrients have not been specifically studied using the RAR approach. Ingestad (1982) and Ingestad and Lund (1986) have argued that the acclimation period, adjustment in growth rate, and subsequent constancy in nutrient status and growth rate represent the “norm” under field conditions. This may be the case for many “wild” plants growing in natural ecosystems, although Chapin (1988) has described a somewhat different paradigm for slow-growing perennials wherein the relative growth rate is low and inflexible, and higher nutrient supply leads primarily to luxury consumption. Regardless, Ingestad’s model is not universally descriptive of the responses of crop plants. In many agricultural settings, initial growth (often sustained by seed reserves) is vigorous but declines as plants become larger and the soil-root system cannot sustain adequate rates of nutrient uptake for normal growth. Deficiency symptoms do indeed develop and, with many nutrients, remobilization and translocation occur in the phloem to meet the demand associated with new growth (Marschner, 1995). Attempts to utilize the RAR method in some other laboratories have also resulted in equivocal results. Mattson et al. (1991) reported that barley (Hordeurn vulgare L.) supplied with nitrate using the RAR approach exhibited only “incom-

166

DAVID R. PARKER AND WENDELL A. NORVELL

plete” steady-state growth and fluctuating tissue N concentrations. Results with pea (Oscarson et al., 1989) suggested that the period of correspondence between RAR and RGR is rather short in fast-growing annual plants. Stadt et al. (1992) adapted the method to more traditional (and manually controlled) solution culture for the rearing of wheat, but did not obtain the expected equality in RAR and RGR. This discrepancy likely reflected inaccuracies in the estimated initial N content of the seedlings, upon which all subsequent additions are based (Stadt et al., 1992), and serves to emphasize the importance of investigator experience using this technique (Hellgren and Ingestad, 1996). Macduff et al. (1993) compared the RAR approach to constant, low external concentration (maintained by flowing solution culture, Section IIIB) of NO,- and their effects on unnodulated pea plants. These authors concluded that most of the physiological parameters describing the plant response to N status were comparable in the two systems, and were thus method independent. Hellgren and Ingestad (1996) strongly criticized this work on the grounds that the steady-state, internal concentration was not assiduously maintained in the RAR plants and that insufficient data were collected after the acclimation period for both sets of plants. Further studies are needed to resolve this dispute. The results of Macduff et al. (1993) do suggest, however, that at least a part of Ingestad’s original objection to “conventional” solution culture methods is really the same objection made by others, that is, that both solution concentration and plant nutrient status are subject to excessive variation over the course of an experiment. Thus, at least partial relief is offered by methods in which concentrations are well buffered at low, realistic levels. On balance, the relative merits of the RAR approach may be subjective and may also reflect differing philosophies concerning nutrient limitations to plant growth in soil. Soil chemists tend to view nutritional problems as involving limited solubility, especially for nutrients that are abundant but very insoluble in soil (i.e., P, Fe); they thus tend to gravitate toward methods that emphasize chemical control of nutrient availability (i.e., solubility). When more mechanistic views of nutrient supply and uptake are involved (Barber, 1995), other, more dynamic factors such as ion diffusion rate also become important. Even more holistic views consider soil nutrient pools that are cyclically depleted and replenished, with consequent fluctuations in solution concentration, but with plant “dampening” of these fluctuations by subtle shifts in root/shoot partitioning and in root morphology (Stadt et al., 1992). The RAR approach represents one means of simulating these factors along with the exponential increase in root system size. Because other nutrients are often added only in proportion to the limiting nutrient (usually N), the RAR approach often fails to simulate a very common soil solution condition in which some nutrient concentrations are at limiting values and readily depleted, while others are supraoptimal. The nutrient interactions, including competitive inhibition of ion uptake, that are likely in such circumstances may not be accurately simulated by the RAR method. The presence or absence of these

ADVANCES IN SOLUTION CULTURE METHODS

167

interactions confounded some of the early experiments using the RAR method (Ingestad and Lund, 1979; Ingestad, 1981, 1982).

B. STABILITYDERIVED FROM LARGE SOLUTIONVOLUMES AND/OR FLOWING CULTURE Depletion of nutrients in solution can, to some degree, be avoided simply by employing a very large volume of nutrient solution relative to the expected plant size at harvest (see Asher et al., 1965, for an early review). Some obvious limitations to this approach exist, however. For example, conventional growth chambers usually preclude the use of a pot size much in excess of 10 liters. In the greenhouse, space limitations are less of a problem and individual tanks in the range of 100 or more liters are usable. At the senior author’s institution, there is a dedicated greenhouse section containing 96 plastic tanks, each with a volume of 120liters. Originally used as reservoirs for a recirculating sand culture system, these have been modified for use with solution culture, and each tank supports three plastic inserts so that three different species can be conveniently reared together (Parker, 1997). All components are made of inert plastics, and a centralized compressedair manifold allows convenient aeration and mixing of the nutrient solutions. The system has been used, along with periodic Panalyses, to successfully maintain stable P concentrations of 10 p M (Parker, 1997). More recent experiments have examined the inhibitory effect of P at 2,20, or 200 p,M on selenite uptake by plants (Hopper, 1997). Significant depletion problems were experienced only late in the experiment, and only at the lowest P concentration. The likelihood of significant depletion using these large tanks can be readily ascertained by using an electronic spreadsheet such as that depicted in Fig. 4. With the exception of P, major nutrients in this example are present in millimolar concentrations. Although these concentrations are not representative of very nutrient-poor soils, they are reasonable for more fertile ones and are unlikely to cause major complications in the study of other growth-limiting factors such as trace element deficiency or toxicity. As indicated in Fig. 4, dry matter yields of up to 20 g per tank can be achieved without significant (15%) depletion of any macronutrient other than P. Somewhat lower macronutrient concentrations could be readily maintained (P excepted) via periodic (i.e., weekly) analyses of the solutions and adjustment with concentrated stocks. With chelator buffering of micronutrient metals (Section IIIC2), the total quantities of Cu, Fe, Mn, Ni, and Zn present in each tank we usually large relative to the expected plant contents, and depletion of these metals is insignificant. A small concentration of the pH buffer MES (0.5 or 1 mM; see Section IIICI) coupled with weekly measurement and adjustment of pH is usually adequate to regulate solution pH in these larger vessels. In many nutritional investigations it is necessary to maintain very stable con-

168

Nutrient

N P K Ca Mg S

DAVID R. PARKER AND WENDELL A. NORVELL

[C] initial mM

5 0.02 1 2 0.5 0.5

[C] plant Atomic Wt. mglg glmol 15 2 10 5 2 1

14 31 39 40 24 32

1

Assumed dry matter yield (glpot) 2 5 10 % depletion

--------0.18 2.7 0.21 0.05 0.14 0.05

-

0.36 5.4 0.4 0.10 0.28 0.10

0.9 13.4 1.1 0.26 0.7 0.3

-

1.8 27 2.1 0.5 1.4 0.5

I

20 _

-

~

3.6 54 4 1.o 2.0 1.o

Figure 4 Example of simple electronic spreadsheet used to calculate depletions of major nutrients in 120-liter vessels for various assumed plant dry matter yields. Shoot tissue concentrations are typical values for monocotyledonous plants (Epstein, 1972).

centrations of major nutrients at concentrations much lower than those depicted in Fig. 4. Examples include many types of studies pertinent to low-fertility soil solutions in which all nutrients are often present at < l mM, as well as efforts to identify the minimum external concentrations required for normal plant growth. In these cases, the continuously flowing solution culture (FSC) method is without peer for maintaining the most stable concentrations at the lowest values (Asher and Edwards, 1978, 1983; Wild et al., 1987). Many FSC systems combine the advantages of very large solution volumes (sometimes > 1000 liters) with frequent (sometimes even continuous) renewal of plant nutrients. In most FSC systems, large volumes of very dilute nutrient solution are automatically pumped or otherwise delivered to individual pots of modest volume (1 to 4 liters). Flow rates vary with design, but are typically 1 liter per pot per minute such that the solution bathing the roots is usually completely replaced every 1 to 5 min (Asher and Edwards, 1978, 1983; Temple-Smith and Menary, 1977; Bhat, 1980; Wild et al., 1987). Several distinct categories of FSC have been delineated in the literature. For example, Edwards and Asher (1974) and Asher and Edwards (1978) described two types. In the first, the concentration of each nutrient in the pot is held constant by continuous adjustment of the inflow rate of a concentrated stock solution such that the combined losses due to uptake and pot overflow are exactly matched. Relatively low flow rates and small pots can be used in these systems, but they are difficult to engineer and not amenable to studies needing more than a few pots. In Asher and Edwards’ second category of FSC systems, the nutrient concentrations entering the pot are simply present at the “target” level desired. Very high flow rates are required to minimize depletion, but these flow rates need not be accurately known or controlled (Edwards and Asher, 1974). This approach has been more widely used, and is more suitable for use with many pots, as long as the pumping requirements can be met. In their review of FSC techniques, Wild et al. (1987) categorized systems as either nonrecirculating, wherein the nutrient solution is pumped through pots and directly to waste, or recirculating, where a large reservoir is continuously readjust-

-

-

ADVANCES IN SOLUTION CULTURE METHODS

169

ed and recirculated. Both systems often utilize clusters of pots (six or more) that receive the same treatment solution. Most systems are for use in a greenhouse, but modifications have been described that allow FSC methods to be used in growth cabinets (Bhat, 1980). With nonrecirculating systems, nutrients in concentrated stock solutions may be simply injected by pumping into the main influent line carrying deionized water. The relation between the two flow rates determines the final concentrations of the nutrients. The principal advantage of these systems is their simplicity, especially the lack of need for chemical monitoring of solution composition (Gutschick and Kay, 1991). The primary drawback is the wasteful consumption of water and nutrients. Recirculating FSC systems of varying degrees of sophistication have been described in the literature. Early versions often entailed manual sampling (once or twice daily) and chemical analysis using whatever laboratory techniques were required for the elements and concentrations of interest (Asher and Edwards, 1978). More elaborate versions have since appeared that include automated analysis of certain nutrients in each solution reservoir and computerized additions of nutrient stock solutions. The system at Hurley in the United Kingdom permits automated monitoring and adjustment of NH4+ and/or NO,-, K, Na, P, and pH (Wild et al., 1987). Nutrients that are not monitored and automatically adjusted are either (i) slowly metered into the reservoirs based on periodic analyses (e.g., Clement et al., 1978); or (ii) added along with, and in fixed proportion to, a nutrient ion that is monitored continuously and adjusted (Woodhouse et al., 1978). To accommodate studies at very low nutrient concentrations, the reservoirs for recirculating FSC systems are often quite large, ranging from a few hundred to >2000 liters each (Asher and Edwards, 1978, 1983; Wild et al., 1987). With both recirculating and nonrecirculating systems, nutrient uptake can be calculated based either on nutrient depletion in solution or on serial plant harvests that reveal temporal changes in plant nutrient content (Wild et al., 1987). The principal advantage of FSC systems is their ability to expose roots to concentrations of nutrients that are both very low and very stable. This capability is especially useful for long-term studies (i.e., more than a few weeks), in which plant biomass becomes large with consequent increases in nutrient demand (Asher and Edwards, 1983). In comparison to nonrenewed culture techniques, flowing culture methods provide a better simulation of dilute soil solutions, especially those of highly weathered, infertile soils (Asher and Edwards, 1978). One of the more common usages of FSC methods has been to define “physiological minima” for various nutrients, that is, the lowest external concentration that will support maximum growth or yield, or the lowest concentration leading to an elemental toxicity. Many of these values are surprisingly low, and some published results for NO,--N, P, K, and Ca are summarized in Table IV. Many of these “critical concentrations” are much lower than those identified using nonrenewed or intermittently renewed solutions. Presumably, the high flow rates used in FSC cause

170

DAVID R. PARKER AND WENDELL A. NORVELL

turbulent flow, which minimizes the unstirred layers around roots where ion uptake is limited by depletion and diffusion. But, as noted by Clement et al. (1974), FSC methods cannot eliminate this effect entirely, and the minimum concentrations observed will always show at least some dependence on flow rate. We also note that the minimum required P levels in Table IV seem to exceed many reported soil solution concentrations (Fig. 1). This discrepancy may reflect the importance of mycorrhizae in many field soils and/or differences in root morphology between soil- and solution-cultured plants. In an earlier review, Asher and Edwards (1983) identified several other potential uses for FSC beyond merely identifying minimum external concentrations required by various plant species and calculating the associated uptake rates. Included was the interaction of temperature and concentration on nutrient uptake and growth, and some progress has been made in this area with N (Bloom and Chapin, 1981; Macduff et al., 1987, 1994). Flowing culture also offers excellent prospects for studying nodulated legumes under controlled chemical conditions (Asher and Edwards, 1983), and recent studies have used the technique to examine how low levels of NO,- and/or NH4+ inhibit N, fixation rates (Macduff etal., 1996; Svenning et al., 1996), as well as the interaction of this inhibition with temperature (Macduff et al., 1989; Macduff and Dhanoa, 1990; Svenning and Macduff, 1996). Curiously, the potential of FSC for studying mycorrhizal effects at suitably low P concentrations (Asher and Edwards, 1983) seems to have been little realized. Howeler et al. (1981) obtained results with casava that seemed to be in excellent agreement with expectations based on soil-grown plants but, to our knowledge, no other instances of this usage of FSC have been published. The method has also enjoyed only limited use for screening different genotypes (or “ecotypes”) within a species (Hommels et al., 1990; Svenning et al., 1996). One of the more significant uses of FSC techniques has been physiological studies of nutrient uptake, especially N, P, and K, and how uptake is coupled to other physiological processes such as leaf CO, flux (Wild et al., 1987). The method allows resolution of short-term (i.e., diurnal) fluctuations in uptake, as well as much longer-term, ontogenetic changes in the complex interdependence between external concentration, relative growth rate, and uptake rate (Wild et al., 1987). Flowing culture has also provided insights into the seemingly fine regulation of nutrient uptake. Uptake rates often change within hours to days in response to new external concentrations so that plant tissue concentrations are rather insensitive to wide variations in solution concentration (Wild et al., 1987). Ion transport rates and related physiological parameters can also be studied using FSC (Macduff and Jackson, 1992), and can often be studied at more-or-less constant physiological “stress levels” (Macduff et al., 1993).Wild et al. (1987) identified other topics such as trace metal partitioning between roots and shoots, and root effluxhnflux of Hf using pH-stat techniques as amenable to FSC methods. Despite the power of FSC techniques, several drawbacks seem worthy of dis-

Table IV Minimum, S a c i e n t Concentrationsof Macronutrientsas Determined Using Flowing Solution Culture Techniques Element NO;-N NO;-N NO;-N

P

P

Concentration (pM) 1400 10 10 25 5 5 1

3.8 0.5

P K

K

K

K Ca

0.4-1.6 24 94 .lo00 6 2 32 1 3 (1.3 <1.3-1.9 2.5 10 100 1000

Ca

50 100 2500 12 or 2500

Species

Lolium perenne Pisum sativum Trifolium repens Hypochoeris glabra; Medicago tribuloides Cryptostemma calendula, Bromus rigidus Erodium botrys; Lupinus digitatus Trifolium subterraneum; Vulpia myuros Latuca sativa Brassica olearea var. capitata Lolium perenne Lolium perenne; Pisum arvense Trifolium subterraneum; Trifolium hirtim Hordeum vulgare Manihot esculenta Crantz Helianthus annus Zea mays L. Anthoxanthum odoratum Dactylis glomerata; Trifoliumpratense Lolium perenne Hordeum vulgare; Raphanus sativus Lolium perenne.; L. albus Triyolium subterraneum; Secale cereale Lycopersicon esculentum; Triticum aestivum Medicago trunculata; Bromus rigidus Arachis hypogaea Glycine mar Cyamopsis tetragonoloba; Cajanus cajan (L.) vigna unguiculata

Comments 14 pM gave 90% of control

Reference Clement er al. (1978) Macduff et al. (1993)

100, 1000 only other concentrations Macduff et al. (1996) Asher and Loneragan (1967) Asher and Loneragan (1967) Asher and Loneragan (1967) 25 p M P was toxic Asher and Loneragan (1967)

representative species; 9 others in study

All levels adequate; suboptimal only at <2 22 other species; these are representative

Cultivar dependent

Temple-Smith and Menary (1977) Temple-Smith and Menary (1977) Breeze et al. (1984) Asher and Ozanne (1967) Asher and Ozanne (1 967) Asher and Ozanne (1967) Spear et al.. (1978) Spear et al. (1978) Spears et al. (1978) Wild et al. (1974) Wild et al. (1974) Woodhouse et al. (1978) Woodhouse et al. (1978) Loneragan ef al. (1968) Loneragan et al. (1968) Loneragan et al. (1968) Loneragan et al. (1968) Bell et al. (1989) Bell et al. (1989) Bell et al. (1989) Bell et al. (1989)

172

DAVID R. PARKER AND WENDELL A. NORVELL

cussion. First, the more sophisticated recirculating systems, such as that at Hurley, represent expensive and specialized capital investments that may be beyond the reach of most investigators. Only a handful of such systems are operating worldwide, and more are unlikely to be built given the costs relative to present priorities in scientific funding. Second, the number of distinct experimental “units” (i.e., a solution reservoir and its associated control system) is generally small (i.e., 59), thus limiting the number of treatments that can be imposed. Often no true replication is possible, but multiple pots, adjacent in space and fed by a single unit, are used instead as pseudoreplicates (Hurlbert, 1984). This limitation can make it difficult or impossible to make meaningful treatment comparisons and thus to obtain results that are statistically rigorous. Other problems arise when a concentration of interest is very low and difficult to analyze. Pretreatments such as solvent extraction or other concentration steps may be required, increasing costs and slowing turnaround time so that it is difficult to maintain nutrients in a timely manner (Asher and Edwards, 1978, 1983). These problems are exacerbated with the trace elements, and other techniques such as chelator buffering (Section IIIC2) are often more convenient. Automated analysis and adjustment add to capital costs and engineering complexity, and only a few nutrients have been successfully handled this way. The prospects for better ion-selective electrodes and improved control systems envisioned by earlier authors (Clement et al., 1974; Asher and Edwards, 1983) seem not to have been fully realized at this time. Some authors recommend drainage and complete replacement of each solution reservoir weekly (Asher and Edwards, 1978), or even as frequently as every 2 days (Bhat, 1980). Such labor needs greatly diminish the value of recirculating FSC methods relative to nonrecirculating FSC or even to the simple large-volume systems described earlier in this section. One putative advantage of FSC systems is that continuous monitoring of solution composition permits a similarly continuous view of nutrient uptake rates (Wild et al., 1987). Although ostensibly true, there is an obvious and inherent disadvantage in relying on nutrient drawdown to calculate uptake rates, because the approach runs contrary to the overall objective of maintaining stable experimental conditions. Obviously, the most accurate estimates of uptake will be obtained when concentration differences between successive samples are large. A final and more philosophical criticism of FSC methods was originally made by Carroll and Loneragan (1968). As usually practiced, FSC techniques represent one extreme case in nutrient buffering: Solution concentrations may be both very low and stable. At the other extreme, nonrenewed solutions usually start with concentrations that are relatively high, but then are depleted because they are completely unbuffered. Neither extreme accurately mimics how nutrient deficiencies develop in the soil-plant system where, in fact, both mechanisms are likely to be involved. For example, during germination and emergence, soil solution concentrations may be (marginally) adequate for normal growth, but become depleted to

ADVANCES IN SOLUTION CULTURE METHODS

173

suboptimal levels when plants enter exponential growth (Carroll and Loneragan, 1968). Thus, soil solution concentrations are buffered, but inadequately to maintain them at values required for optimal growth. Theoretically, progressively lower soil solution concentrations might be established as different solid-phase pools of varying lability are exhausted (Asher and Edwards, 1978). Relative growth rates might thus decline, and symptomology worsen, but never with the rapidity that occurs in unbuffered, nonrenewed solution cultures.

C. BUFFERING PHAND METALIONACTIVITIES WITH SOLUBLE LIGANDS 1. pH Buffering

Soils are comparatively well buffered with respect to pH, while nutrient solutions traditionally are not. Control of nutrient solution pH is a universal concern because pH may influence ion uptake directly, that is, physiologically, through participation of protons in ion co-transport, or indirectly in a variety of ways (Parker et al., 1995a). Examples of indirect effects include changes in the solubility or aqueous speciation of cations and anions. Increases in solution pH can readily lead to precipitation of Fe(III), even when it has been added to solution in a chelated form such as FeEDTA (Chaney, I988a). The precipitated Fe-hydroxides may drop out of solution and/or may coat root surfaces, altering the availability of Fe and perhaps other trace metals (Bell et al., 1991a). Moreover, the precipitation of Fe effectively “liberates” some of the added chelator (such as EDTA) so that it is free to complex Cu, Zn, and Mn, thus lowering the plant availability of these metals (Chaney, 1988a; Norvell, 1991; see Section IIIC2). Other effects of pH on speciation and availability include pH-dependent changes in complexation by ligands such as OH- (e.g., the hydrolysis of Al) or carbonate species (e.g., complexation of Cu at neutral or alkaline pH). Changes in solution pH are caused primarily by imbalances in the uptake of cationic and anionic nutrients. When cation uptake exceeds anion uptake on an equivalent charge basis, the solution is acidified, while alkalinization occurs when the anion uptake is greater (Asher and Edwards, 1983; Marschner, 1995). The source of N is particularly influential, because N is usually the mineral nutrient taken up in the greatest molar quantity. Ammonium-based nutrient solutions usually lead to pronounced acidification, while those where N is supplied as NO,tend to become alkaline (Asher and Edwards, 1983). Exceptions occur, however, especially with certain species that tend to acidify regardless of solution composition (Marschner, 1995). Efforts have been made to maintain pH by providing NH4+ and NO,- in prescribed ratios (e.g., I :8; Clark, 1982), or by making small daily additions of an NH4+ salt to a NO,--based solution (Bell et al., 1991b).

174

DAVID R. PARKER AND WENDELL A. NORVELL

However, such methods are not very precise, require foreknowledge of plant responses, and are unlikely to be successful during longer-term experiments because of ontogenetic changes in cation-anion balance (Table 15.4 of Marschner, 1995). Simple measurement and manual adjustment of pH, perhaps daily or twice daily, can be used to moderate changes in pH, but this approach is labor intensive and may be inadequate to maintain pH within a narrow range. Difficulties in maintaining pH are particularly likely late in experiments when plants become large relative to nutrient solution volumes. Automated measurement and control is possible (Webb, 1993), but the cost and technical requirements may place this approach beyond the reach of many investigators. The high concentrations of phosphate used in many conventional nutrient solutions have, perhaps inadvertently, provided some degree of pH buffering in countless nutritional studies. But the use of P at the millimolar concentrations required to effectively buffer pH is undesirable from a number of standpoints: P is readily taken up by the plant leading to possible reductions in buffering capacity; P may be phytotoxic at these concentrations (Howell and Bernard, 1961; Ismande and Ralston, 1981); high P levels may alter the uptake of other nutrients (Brown, 1972); root hair formation may be inhibited (Foehse and Jungk, 1983); and P may precipitate with Ca or Mg, particularly at pH values 27.0 (Parker et al., 1992). Moreover, only the second dissociation constant of phosphoric acid (the three pKa values are 2.2,7.2, and 12.4) is in a range useful for plant culture. If a pH value of 5.0 or 5.5 is desired, phosphate provides rather weak pH buffering (Bugbee and Salisbury, 1985). Finally, the use of P to buffer pH requires concentrations of soluble P that are orders of magnitude larger than those that occur in soil solution.

a. Good’s Buffers The problems discussed above led researchers to seek out alternative buffers to help stabilize solution pH. In 1966, Good et al. introduced a set of “biologically inert” buffers that have become widely used for stabilizing solution pH for in vitro biochemical experiments. The suitability of these buffers is enhanced by (i) high solubility in water and low solubility in organic solvents (to minimize diffusion across membranes), (ii) low reactivity with cationic metals, and (iii) resistance to enzymatic degradation (Good et al., 1966; Good and Izawa, 1972). These characteristics coincide well with the desirable features of a soluble buffer for nutrient solution work. The buffering capacities of the Good’s reagents are due to the protonation behavior of a secondary or tertiary amine group that dissociates more readily than primary amines (such as glycine) that have pK, values >9.0 (Good et al., 1966). Many also contain a strong-acid sulfonate group that is fully dissociated at all relevant pH values. Thus, whenever the solution pH is significantly less than the pKa of the amine, most Good’s buffers are zwitterions containing both -NHxf and -SO,- functional groups. At higher pH values, the amine group dissociates such that an anionic species predominates. The range in available pK, val-

ADVANCES IN SOLUTION CULTURE METHODS A+ 0 N-CH,CHzSO3W H MES (pK, = 6.1)

0

N-CH&HZCH,SO, W H MOPS (pK, = 7.2)

+A

NCH2CH,S03H w HEPES (pK, = 7.5)

-OOCCH,\+

+

HOCH,CH,/

N-CH,CH,SO; H

BES (pK, = 7.1) HOCH,

n+

HOCCH,CH,N

HOCH,CH,\

175

f NCHZ-C-NH,

-OOCCH,’ ADA (pK, = 6.6)

\ + HOCb- C-N-CH,CH,SOi HOCH, / H, TES (pK, = 7.4) HOCH,

\ + HOCH,-C-N-CH,CH,CH,SO~ HOCH, / H, TAPS (pK, = 8.4) 0 II

+

H,N-C-CH,-N-CH,CH,SO3H, ACES (pK, = 6.8)

Figure 5 Eight Good’s biological buffers that have been evaluated for use in nutrient solution culture of plants. The upper six buffers have been shown to be nontoxic at concentrations up to ca. 2-5 mM. ADA and ACES are acutely phytotoxic and are not recommended.

ues is currently about 6 to 10, and the eight Good’s buffers that have been evaluated for use in nutrient solutions are depicted in Fig. 5, along with their respective pKa values. Ismande and Ralston (198 1) were the first to report on the use of a Good’s buffer (MES) for controlling nutrient solution pH. Either 1 or 2 mM provided reasonably stable pH values of 6.2 with soybeans (Glycine max [L.] Merr.) that were supplied with nitrate or were nodulated and fixing atmospheric N,. The stability in pH was somewhat lessened by the use of NH,NO, as an N source. No toxicity or other adverse effects of MES were observed. Subsequent reports have indicated that MES can be safely used with a variety of crop species with concentrations ranging from 1 to 5 mM (Bugbee and Salisbury, 1985; Rys and Phung, 1985; Miyasaka et al., 1988). Concentrations of 10 mM have been reported as both harmful (Rys and Phung, 1985) and innocuous (Bugbee and Salisbury, 1985), but are usually not necessary to provide good buffering in the pH 6.0 range. Concentrations of MES from 1 to 5 mM should be adequate depending on plant species and size, nutrient solution volume, N source, and other factors. In addition to the specific evaluations of its suitability for nutrient solution buffering, MES has recently enjoyed some widespread use in mineral nutrition

176

DAVID R. PARKER AND WENDELL A. NORVELL

studies (Bell et al., 1991a,b, 1992; Norvell and Welch, 1993; Gries et al., 1995; Parker, 1997).With a pKa of 6.1, MES provides excellent buffering in the pH range 5.5-6.5. To date, the other seven buffers depicted in Fig. 5 have been less studied and little utilized. Working with Trifolium repens (both N, fixing and supplied with NH,), Rys and Phung (1985) reported that, in addition to MES, BES and MOPS were nontoxic at concentrations up to at least 2 and 12 mM, respectively. The buffers ADA and ACES, however, were acutely toxic at 2 mM, and are not recommended for use as nutrient solution buffers. Wehr et al. (1986) reported no toxicity problems with HEPES up to 2.5 mM for solution culture of algae. Preliminary data (D. R. Parker, 1998, unpublished data) suggest that TES and TAPS at 5 5 mM can be used to buffer at pH 7.0 and 8.0, respectively, and cause no toxicity problems with alfalfa (Medicago sativa L.). In general, inclusion of MES or other buffers does not seem to significantly alter uptake of mineral nutrients (Bugbee and Salisbury, 1985). In a few cases, however, high concentrations (i.e., 5-10 mM) of MES have resulted in somewhat altered shoot levels of some nutrients, but without any effect on dry matter yields (Miyasaka et al., 1988; Stahl et al., 1996). Note that leaf S concentrations are usually elevated due, apparently, to uptake and translocation of the intact MES molecule (Stahl et d., 1996). All of the Good’s buffers depicted in Fig. 5 exhibit minimal complexation with Ca and Mg, and with transition metals such as Cu and Zn (Good et al., 1966; Good and Izawa, 1972; NIST, 1993). Both BES and TES form weak complexes with Cu(I1) (KML= lo4;Good et al., 1966; NIST, 1993), so that MOPS may be the preferred buffer when pH values near 7.0 are desired. Angle et al. (1992) reported reductions in Cd toxicity toward Rhizobia in liquid culture when concentrations of MES or HEPES reached or exceeded about 5 mM. These authors assumed that the reduced Cd toxicity was due to complexation by the buffer, but did not confirm this complexation with independent chemical measurents. Except when axenic culture techniques are used, nutrient solution cultures will always contain at least modest populations of heterotrophic bacteria that may potentially degrade any organic buffer in solution. Based on the unchanged titration behavior of a nutrient solution that had been in use for 21 days, Bugbee and Salisbury (1985) concluded that MES was largely resistant to microbial degradation, as originally asserted by Good et al. (1966). Solutions that had been used continuously for 69 days showed significant degradation; however, such long growth periods without solution replacement are rarely, if ever, suitable for experimental plant culture. Other evidence for resistance to degradation is provided by Wehr et al. (1986), who attempted to grow four aerobic heterotrophs using MES or HEPES as the sole carbon source. No growth occurred on either buffer during the 7-day experiment, further demonstrating the recalcitrant nature of the Good’s buffers. In summary, several of the Good’s buffers, and MES in particular, seem to be suitable for buffering nutrient solution pH when used at concentrations 5 5 mM,

ADVANCES IN SOLUTION CULTURE METHODS

I77

with 1-2 mM often being adequate. The upper six buffers depicted in Fig. 5 are all nontoxic, and are all zwitterions containing one amine and one sulfonate group; metal-binding tendencies and biodegradability are both low. It is not known why ACES and ADA are acutely phytotoxic, but both contain carbamyl groups (Fig. 5). At this time, there is a regrettable lack of a proven buffer for use at pH values of -5.0 or below. Such pH values are of interest to numerous researchers studying acid-soil infertility problems, including the phytotoxicities of A1 and Mn. Identification and evaluation of one or more buffers suitable for studies in the pH range 4.0 to 5.5 is an area worthy of future research. b. Bicarbonate Bicarbonate is present naturally in all solutions exposed to atmospheres containing CO,. This highly soluble anion is an important pH buffer in natural waters, including the soil solutions of near-neutral and alkaline soils (Bradfield, 1942; Stumm and Morgan, 1996). Although bicarbonate may be considered the most “natural” pH buffer for solution culture, it has been used most often simply to exacerbate Fe chlorosis in susceptible crops. Soluble bicarbonates have been recognized for many years as a contributor to “lime-induced chlorosis,” a common problem for some crops on alkaline soils (Harley and Lindner, 1945; Wadleigh and Brown, 1952). Concentrations of NaHCO, ranging up to 40 mM or more have been added to culture solutions to induce symptoms of chlorosis (Coulombe et al., 1984a; Viti and Cinelli, 1993; Shi et al., 1993). Much less common has been the purposeful use of bicarbonate in combination with controlled partial pressures of CO, to buffer pH at specific levels. Bicarbonate may serve as an effective pH buffer for nutrient solutions in the slightly acid to alkaline pH range, providing that the interactions among pH, bicarbonate, and CO, are understood and utilized properly. These three components are linked through the well-known equilibria shown below, where H2C03*represents the sum of “true” carbonic acid plus dissolved CO,(aq), and CO,(gas) is expressed in atmospheres: CO,(gas) H,C03*

+ H,O P H,CO,* P H+ + HC0,-

log K = -1.46 log K = -6.36

At pH values near the pK of 6.36, or above, bicarbonate effectively resists acidification by reacting with protons to form H2C03*.Any excess H2C03* will dissociate into CO, and water, and CO, will be lost to the atmosphere until equilibrium is reestablished between CO, in solution and CO, in the gas phase. Conversely, increases in pH are resisted by the dissolution of CO, from the atmosphere to release protons in a reversal of these same reactions, which causes HC0,- to accumulate in solution. The direct involvement of a gas adds some complexity to the use of bicarbonate as a pH buffer because gas phase composition

178

DAVID R. PARKER AND WENDELL A. NORVELL

plays an important role. A short discussion of the interactions among CO,, soluble carbonate species, and pH is given by Lindsay (1979), and a more extensive analysis is presented by Stumm and Morgan (1996). Note that although CO,,participates in these reactions, its contribution to buffering is generally not important in a pH range suitable for plant growth. Figure 6 shows the relations between pH and the partial pressure of CO, for several concentrations of bicarbonate at a temperature of 25°C. Conditions are restricted here to a general range suitable for plant culture in solution. For any specific level of bicarbonate, pC0, and pH are uniquely linked at equilibrium along an iso-concentration line. For example, if the bicarbonate concentration is 1 mM and the solution is in equilibrium with atmospheric CO, of 0.0003 atm (or 0.03% in air), then the pH will be relatively high at about 8.3. However, if the CO, partial pressure is higher at 0.01 atm (1% in air), then the equilibrium pH will be about 6.8. This latter solution is actually well buffered because both HC0,- and H,CO,* are present in significant amounts and H2C03* is easily maintained by aeration. Higher levels of CO, and bicarbonate can be combined to provide even stronger pH buffers in a pH range still suitable for plant growth. One of the earliest efforts to use C0,-HCO, equilibria to regulate pH in solu-

h \

\30

Carbonate Alkalinity, mM (Ionic Strength 0.05)

E

c

6 B

0.0001-I 6

I

I

6.5

7

,

I

7.5 8 8.5 PH Figure 6 Relationships between the partial pressure of CO, and solution pH for five concentrations of carbonate alkalinity (mmol,-, liter-'. Relationships are shown as dashed lines above pH values at which precipitation of CaCO, would be expected to alter concentrations of Ca and alkalinity in a one-quarter-strength Hoagland's solution with Ca, = 1 mM. For simplicity, a constant ionic strength of 0.05 was assumed.

ADVANCES IN SOLUTION CULTURE METHODS

179

tion culture was that of Porter and Thorne (1955) in a study of the causes of chlorosis in bean (Phaseolus vulgarus L.) and tomato (Lycopersicon esculentum L.). The pH was controlled at different levels between 7.7 and 8.3 by aerating solutions containing 0.3 to 50 mM NaHCO, with air enriched with CO, up to a level of 5%. They concluded that chlorosis increased with bicarbonate concentration even when pH was held constant. Brown et al. (1959), using a similar approach, studied the role of bicarbonate in chlorosis of soybean. They too recognized the interrelationship of CO, and HC0,- and chose to control pH at 7.8 2 0.2 by aerating solutions containing 1.5 to 20 mMNaHCO, with air containing CO, at 0.15 to 2%. Despite the apparent success of these early studies, the use of C0,-pH-bicarbonate equilibria to regulate pH in nutrient solutions received little additional attention until the approach was revisited by Chaney, Coulombe, and co-workers (Coulombe et al., 1984b; Chaney er al., 1992b; Dragonuk et al., 1989b). Various concentrations of NaHCO, were added to solutions, in combination with aeration by 2.5 or 3% CO,, to screen soybeans and several other crops for resistance to Fe chlorosis. Chaney er al. (1992b) provided detailed instructions for setting up and operating a C0,-HCO, buffered system. In addition to using 3% CO, and 10 mM HCO, to buffer pH at about pH 7.3, they raised the concentrations of Ca and Mg to 5 mM in their nutrient solutions to better reflect the composition of the soil solution in many calcareous soils (Inskeep and Bloom, 1984). Recently, Norvell et al. (1995) modified the method of Chaney et al. (1992b) by substituting Mg(HCO,), for NaHCO, as the source of alkalinity. This was done to allow the effects of bicarbonate to be studied separately from the effects of high levels of Na+. This change also permits NO,- concentrations to be reduced from the higher levels used by Chaney et al. (1992b). Both changes make the final solution even closer in composition to the soil solutions of calcareous soils. This modified method was tested against the original and was found to provide the same separation of soybean genotypes with respect to their chlorosis sensitivity. Bond (1998) adopted the approach for studies of the chlorosis sensitivity of seedlings of oak (Quercus spp.), because of the reported susceptibility of some oaks to injury by high soluble Na. In this latter study, bicarbonate at 1 or 5 mM was combined with aeration by air containing approximately 3% CO, to buffer pH at about 6.8 or 7.4, respectively.

2. Chelator or Ligand Buffering of Microelement Metals Jacobson (195 1) first advocated the use of a chelating agent such as EDTA as a means of maintaining Fe in a soluble form in nutrient solutions, and this has become an almost universal practice in most laboratories (Parker et al., 1995a). This and other early studies (DeKock and Mitchell, 1957; Brown et al., 1961; Wallace, 1962) provided tantalizing clues about the potential utility of chelators for manipulating the plant availability of the micronutrient metals, Cu, Fe, Mn, Ni, and Zn.

180

DAVID R. PARKER AND WENDELL A. NORVELL

For example, Brown et al. (196 1) observed that increasing the molar ratio of DTPA to Fe(II1) from 0.16 to 18 (at constant total Fe) resulted in marked declines in shoot concentrations of Fe, Cu, and Mn, and they attributed the results to increased “competition” between the chelator and plant roots. An astute insight at the time, the same result would now often be couched in terms of the relative unavailability of chelated/complexed metals for root absorption, as compared to the free, aquated metal ion (Parker et al., 1995a and below). Jacobson (1951) and Brown et al. (1961) showed that the ability to acquire adequate Fe in the presence of a molar excess of EDTA or DTPA was very species dependent, a result that is fully understandable only in light of recent advances in our understanding of Fe acquisition by dicots versus graminaceous monocots (Marschner and Romheld, 1994; Parker et al., 1995a;Welch, 1995). In parallel with the above-cited nutritional studies, soil chemists were adapting chemical equilibrium models for describing the aqueous speciation of soil solutions (Lindsay et al., 1966; Adams, 1971) and nutrient solutions (Halvorson and Lindsay, 1972). The latter study was seminal as it opened the door to a number of subsequent studies of metal-chelate equilibria in nutrient solution (Halvorson and Lindsay, 1977; Lindsay, 1979; Cline et al., 1982; Chaney, 1988a; Schwab and Lindsay, 1989a,b; Norvell, 1991). These reports contributed to a much more sophisticated view of nutrient solutions, including a strong emphasis on the chemical forms of the trace metals rather than on just their total concentrations. Prior to these advances, most researchers were generally unaware of the importance of chemical speciation in nutrient culture solutions. Shifts in the chemical form of ions were an uncontrolled source of variation in trace metal availability (Chaney, 1988a; Norvell, 1991). We illustrate the importance of chemical speciation with calculations, using GEOCHEM-PC (Parker et al., 1995b),to evaluate two conventional nutrient solutions, one utilizing FeHEDTA and the other FeCDTA (Table V). Free ion activities of Cu, Mn, and Zn would be expected to be relatively high ( 10-5.9-10-6.5 M ) in the absence of any added Fe-chelate. In the presence of the Fe-chelates,even without losses of Fe by precipitationof Fe(OH),, the free Cu2+ activity drops dramatically with added HEDTA, but much less so with CDTA. Activities of Mn2+ and Zn2+ are only minimally affected (Table V). This result reflects the ease with which Cu(I1) can often displace Fe(II1) from commonly used chelating agents. If even small amounts of Fe(OH), (s) precipitate, however, activities of the other three metals are markedly decreased (Table V). These calculations illustrate the concurrent strengths and weaknesses of conventional nutrient solutions.On the one hand, precipitation of Fe(OH), liberates chelator, which reacts with Cu, Mn, and Zn reducing the likelihood that they might pose toxicity problems (Parker et al., 1995a). On the other hand, the activities of all the micronutrient metal ions decline unpredictably. In the late 1980s, Chaney and others popularized the term “chelator-buffered nutrient solutions” (CBNS) to describe nutrient solutions to which an excess of

Table V Calculated Trace Metal Speciation in a Conventional Nutrient Solution”at pH 6.0 as a Function of Fe-Chelate Added and of the Assumed Log K for the Reaction Fe3+ + 3H,O P Fe(OH),(s) + 3H+

No Fe-chelate 10 pA4 FeHEDTA ‘‘-m,3

-3.5 -2.5 -1.5 ! FeCDTA 10 p4 “-m”

-3.5 -2.5 -1.5

6.2

6.5 8.9 10.8 11.8 12.6

100 100 100 100

6.3 7.7 8.7 9.5

6.6 7.1 7.9 8.9

18 72 96

6.2 6.2 6.2 6.4

100

5.9 26 97 100 100 0.1 1 7 42

5.9 5.9 6.0 6.3

0 2 17 58

12.5 14.5 15.5 16.5

5.9 5.9 5.9 5.9

0

13.4 14.5 15.5 16.5

0 0.1 1

15 21 45

4 6 9

a 2 mM Ca(NO,),, 1 mM KNO,, 0.5 mM MgSO,, 100 pkf KCl, 20 pA4 NaH,PO,, 10 p V H,BO,, 0.1 pA4 Na,MoO,, 0.5 pA4 CuCl,, 2 pbf MnCI,, and 1 pkf ZnCI,. Percentage of the total metal in solution complexed by the chelating agent. Percentage of Fe predicted to precipitate as the hydroxide solid phase.

182

DAVID R. PARKER AND WENDELL A. NORVELL

chelator is intentionally added (Chaney et al., 1989; Bell et al., 1991a,b).This approach has been greatly facilitated by easy-to-use computerized chemical equilibrium programs, such as GEOCHEM-PC, mentioned above (Parker et al., 1995a,b). The basic principles and advantages of this approach have been thoroughly reviewed elsewhere (Chaney ef al., 1989; Parker ef al., 1995a), but are summarized here: 1. A molar excess of EDTA, HEDTA, or a similar ligand is included. This excess is usually 25 to 50 p M (occasionally 100 pA4) greater than the sum of the Cu, Fe, Mn, Ni, and Zn concentrations. Free metal ion activities are generally about lop8 for Mn2+,and substantially lower for the other metals, and are thought to be more representative of those occurring in solutions of circumneutral uncontaminated soils (Parker et al., 1995a). 2. Root concentrations of the trace metals are much more consonant with shoot concentrations and physiological “norms” (Bell et aE., 1991a). 3. Precipitation of Fe(II1) as colloidal hydroxides that coat pot walls and root surfaces is either eliminated or greatly reduced. This leads to better constancy in metal speciation and plant availability. 4. Contaminating levels of Cu, Zn, and, to a lesser degree, Mn have traditionally made it difficult to achieve deficiencies of these metals, especially in young seedlings. Longer experiments, often without solution replacement or renewal, have sometimes been required to achieve substantial growth reductions due to deficiencies (Parker et al., 1995a). The CBNS approach overcomes this problem and is usually unimpaired by modest levels of trace-metal contamination. In light of earlier reviews (Chaney et al., 1989; Norvell, 1991; Parker et al., 1995a), our coverage of the CBNS method is confined to a few key points and some newer utilizations of the method. First, we note that plant responses to buffered trace-metal activities do not seem to conform exclusively to a free ion model of metal availability (Campbell, 1995; Parker et al., 1995a). Such a model predicts that growth responses and tissue concentrations should only depend on the free ion activity, not on the total metal concentrations or the specific chelator used to buffer the activity. Comparatively small deviations from this model have been noted for Cu, Mn, and Zn (Bell et al., 1991a; Parker etal., 1995a).Much larger deviations occur with Fe, and these relate both to Fe acquisition mechanisms by plants and to the proper selection of a chelating agent for CBNS experiments. Dicots and nongraminaceous monocots, the so-called “Strategy I” plants, take up Fe exclusively as Fe(I1) (Welch, 1995). Membrane-bound, extracellular reductases reduce Fe(II1) to Fe(I1) prior to transport into the cytoplasm, and the activity of these reductases increases, often dramatically, when plants are Fe-stressed (Welch, 1995). This reductase activity, in concert with acidification of the apoplastic space, helps to destabilize chelated Fe(II1). Thus, with an “Fe-efficient” species such as tomato, chelators with extremely high affinities for Fe(II1) such as HBED

ADVANCES TN SOLUTION CULTURE METHODS

183

can still be used effectively as Fe sources (Chaney, 1988b). On the other hand, a useful chelator should have a relatively low affinity for Fe(I1) so that, once Fe(II1) is reduced, the Fe(I1) can be transported into the plant rather than bound to the ligand (Parker et al., 1995a). With less efficient plants, the list of candidate ligands becomes more restrictive, but reasonably strong Fe(II1) chelators such as DTPA and CDTA may still be appropriate (Parker et al., 1992, 1995a; Chaney et al., 1992a,b). Because of the external reduction step and the “competition” between the root and the chelator for Fe(II), plant availability of Fe to dicots and non-grass monocots in CBNS systems is not necessarily related to the calculated Fe3+ activity. In the graminaceous monocots (Poaceae), the so-called “Strategy 11” plants, reduction and uptake of Fe(I1) is also thought by some authors to occur at low rates (Welch, 1995). However, when Fe-stressed (a situation we view as “normal” in many solution cultures), the grasses excrete nonproteinaceous amino acids called phytosiderophores with high affinities for Fe(II1) (Marschner and Romheld, 1994; Welch, 1995).These compounds scavenge and chelate Fe(III), which is then transported, without reduction, into the root cells where it is metabolized and translocated. In the CBNS system, the molar excess of ligand greatly limits precipitation of inorganic Fe-hydroxides on the root surface and in the apoplastic space, a putatively important source of Fe to grasses growing in more conventional nutrient solutions (Welch, 1995). Consequently, the exuded phytosiderophores may have to compete directly with the exogenous synthetic chelating agent for Fe, and this greatly limits the selection of chelators suitable for CBNS work with grasses (Parker et al., 1995a). Any ligand significantly stronger than EDTA is unlikely to succeed, but the somewhat weaker Fe chelates of HEDTA and perhaps NTA are likely to be suitable. Chelators weaker than NTA are usually unsuitable because they cannot prevent precipitation of Fe-phosphates and -hydroxides (Parker et al., 1995a). In addition to ligand strength, the availability of Fe to the grasses shows a marked dependence on total Fe in solution (Bell et al., 1991a) and is probably only secondarily related to the calculated free Fe3+ activity. Moreover, the total Fe level required is often higher than with dicots, and Fe concentrations of 50 to 100 pA4 are not unusual (Rengel and Graham, 1995; Gries etal., 1995; Parker, 1997).These considerations are even more crucial with grasses that are Fe-inefficient, such as sorghum (Sorghumhicolor L.) and maize, than they are with more efficient species such as wheat and barley (Marschner and Romheld, 1994). Across both groups of plants (Strategy I and 11), the chelators that have been most used to date include DTPA (Parker et al., 1992; Swietlik and Zhang, 1994), HEDTA (Norvell and Welch, 1993), EDTA (Bell et al., 1991a, 1992), and to a lesser degree CDTA and NTA (Bell et al., 1991a). Other candidate ligands exist (see Table I; Parker et al., 1995a) but generally do not have the most desirable metalbinding properties. Because of its moderate affinity for Fe(III), HEDTA can main-

184

DAVID R. PARKER AND WENDELL A. NORVELL

tain Fe in solution while also supplying adequate Fe to Poaceae. Moreover, relative to its affinity for Fe, HEDTA is quite a strong chelator of Cu and Zn and seems suitable for imposing deficiencies of either metal. Consequently, HEDTA has become something of the "chelator of choice," especially in studies of Zn deficiency (Norvell and Welch, 1993; Welch and Norvell, 1993; Parker, 1993, 1997; Rengel and Graham, 1995, 1996). It has also been used successfully to impose Mn deficiencies (Webb et al., 1993; Huang et al., 1993). Metal concentrations and activities in an HEDTA-buffered solution typical of those used in our laboratories are summarized in Table VI. It utilizes a chelator excess of 25 IJ.M (Norvell and Welch, 1992; Gries et al., 1995), but excesses of 50 pM have also been used effectively (Parker, 1993, 1997). Note that, except for Fe, which is necessarily high, the total metal concentrations are modest ( < l o I.LM), thus avoiding the artificially high shoot-tissue metal levels that have sometimes arisen with stronger ligands such as DTPA (Parker et al., 1992; Parker, 1997). On the other hand, the total concentrations of metals are generally high with respect to anticipated depletion, and their activities thus are well buffered in these solutions (Parker et al., 1995a). If the solution described in Table VI was used with a pot volume of 3 liters, and the plant tissues contained the typical concentrations of Cu, Fe, and Zn shown, then total dry matter yields of 1 g pot- would result in depletions of 1 3 % , assuming no solution replacement. With intermittent replacement of solutions, depletion would of course be even less. Manganese is more problematic because MnHEDTA is such a weak chelator that only very low total concentrations can be added to buffer (Mn2+) at the appropriate level (Webb et al., 1993). For the HEDTA-buffered solution described above, approximately six solution changes would be required to maintain depletions of total Mn at <-5% (Table VI). Thus, HEDTA generally exhibits excellent buffering of metals with reTable VI Total Concentrations and Free Ion Activites of MicronutrientMetals in a Chelator-Buffered Nutrient Solution

wid

MT Metal

(M

cu2Fe3+ Mn2+ Zn2+

2 24 0.6 8

log(M"+)

- 13.2 - 17.0 -1.1 -9.8

Plant tissue concentration (Pi2 g-I) 10

80 30 30

Depletion in & % (g DM)-' 2.6 2.0 30.3 1.8

Nore. For the typical plant tissue concentrations given, percentage depletion in total metal concentration per gram of dry matter produced was calculated. Metal-ion buffering was provided by a 25 pkf excess of HEDTA; assumed pot volume was 3 liters. Solution composition is from Gries et al. (1995) (2 mMCa(NO,),, 1 mM KNO,, 0.5 mM MgSO,, 80 pI4 NaH,PO,, 10 WH,BO,, 0.1 pkfNa,MoO,).

ADVANCES IN SOLUTION CULTURE METHODS

185

spect to depletion by plant uptake, except for Mn, for which considerably more care is needed to maintain constant Mn2+ activities (Webb et al., 1993). A second issue of solution stability or buffering has to do with how the CBNS system responds to changes in solution pH. All chelator-bufferedsolutions exhibit better stability in speciation than do conventional solutions at high pH (>6); in the latter, Fe precipitation problems are even more profound than those depicted in Table V. We previously presented computations showing that calculated activities of Cu, Mn, Ni, and Zn were very stable in EDTA-buffered solutions, but showed more variability as a function of pH when DTPA was used (Parker et al., 1995a). Similar calculations for the HEDTA solution presented in Table VI have been made and are presented in Fig. 7. The “break” in each solid line at pH -7.3 corresponds to the onset of Fe(OH), precipitation. The dashed lines depict the hypothetical speciation in the absence of Fe precipitation and are included for comparison. Stabilities in Cu, Mn, and Zn activities are intermediate between those in EDTA- and DTPA-buffered solutions (Parker et al., 1995a). The continuous decline in Fe3+ with increasing pH (Fig. 7) reflects the ease with which Fe(II1) hydrolyzes to form soluble hydroxo complexes and is common to all CBNS. Gries et al. (1995) used HEDTA-buffered solutions to impose deficiencies of Fe, as well as of Cu, Mn, and Zn, in young barley seedlings. Deficiencies of the latter three metals were obtained simply by omitting each from the solution such that only background concentrations were present (i.e., S O . 1 I.M).Thus, free met-

-22

4.0

f

5.0

I

I

6.0

,

I

7.0

Solution pH

I

I

8.0

9.0

Figure 7 Changes in computed free metal ion activities as a function of pH for a nutrient solution buffered with a 25 pM excess of HEDTA. The solution is the same as that depicted in Table VI.

186

DAVID R. PARKER AND WENDELL A. NORVELL

Figure 8 Changes in relative yield (whole plant) (a) and root weight ratio (b) over time for barley (Hordeurnvulgare L. cv CM-72) seedlings grown in chelator-buffered solutions using a 25 f l e x cess of HEDTA. The control solution was that depicted in Table VI except that total Fe was 75 pM [log (Fe3’) = - 16.51. For the -Cu, -Mn,and -Zn treatments the respective metal was simply omitted (see text). Reproduced from Gries et al. (199% with permission from Kluwer Academic Publishers.

a1 activities were extremely low, although not precisely known, and severely deficient plants were produced (Fig. 8a). Relative yields after 18 days in the treatment solutions were 34, 61, and 49% in the absence of Cu, Mn, and Zn, respectively (Fig. 8a). This result with Cu is in marked contrast with those reported by Bell et al. (1991b), who suggested that Cu deficiency in maize (and presumably other Poaceae) could not be achieved in chelator-buffered solutions due to (i) contaminating Cu levels, and (ii) the requirement for a weak chelator to provide adequate Fe to Strategy I1 plants. Instead, they suggested a more cumbersome “doublebuffered” solution where the Cu-specific ligand BPDS was used in conjunction with HEDTA to supply other nutrients. Gries et al. (1995) solved this problem, at least in barley, by chemically “scrubbing” the macronutrient stock solutions for contaminating levels of Cu and by working in clean environments (i.e., filtered growth cabinets to eliminate dust problems). A further advantage of chelator-buffered solutions is that they seem capable of providing relatively constant levels of physiological stress (Chaney et al., 1992a;

ADVANCES IN SOLUTION CULTURE METHODS

187

Parker et al., 1995a).When a nutrient such as Fe is simply withheld from a nutrient solution, plants become increasingly chlorotic and eventually necrotic. Tissue Fe concentrations decline as well, and this is not a realistic simulation of Fe deficiency as occurs under field conditions, where the degree of chlorosis and reduced relative growth rate can be sustained for some time (Chaney et al., 1992a).Recent experience with barley growing in HEDTA-buffered solutions suggests that intermediate levels of deficiency stress can be adequately simulated in growth chamber or greenhouse. Figure 9a shows that, after about a 10-day adjustment period, relative growth rates tend to stabilize at values related to the imposed Fe levels. (These data are somewhat “noisy” because separate subgroups of plants had to be destructively harvested at each sampling date.) Even more striking is the constancy in shoot Fe concentrations from Day 5 onward (Fig. 9b). These results contradict the assertions of Ingestad (1982) and others (Stadt et al., 1992) that the RAR technique is the only means to achieve constant physiological stress levels, a method that would be quite difficult to apply to the micronutrient metals. In our view, the CBNS approach, where free metal activities are well buffered at suitably

7J

sE

r

3 POl

log (Fe”) -16.5 -17.0 -17.5 V -18.0 -19.0

‘

I

I

I

,

I I

I

I

0.3 0 A

0.2

a,

2 0.1

-m

d

I

a

0.0

960

1

-

€----*/--r \*--* = = : : * ; I - --+

c

a,

40

-

20

-

a, 0

,

-+ ---*------* - ---

0

“

a 0

0

0

b I

I

I

0 I

Figure 9 Changes in relative growth rates (a) and shoot Fe concentrations (b) of barley (cv Morex) seedlings grown at five Fe stress levels in solutions buffered with HEDTA. Solution composition was identical to that given in Table VI, but with total Fe concentrations of 0.24, 2.4, 7.5,24, or 75 pM to provide the computed free Fe3+ activities shown. Adapted from Wu (1996).

188

DAVID R. PARKER AND WENDELL A. NORVELL

low levels, offers greater promise and is deserving of further investigation with respect to growth rates and patterns. Because of this apparent constancy in deficiency symptoms, leaf Fe concentrations, and growth rates, Chaney and co-workers have advocated CBNS methods for screening of germplasm for resistance to Fe stress (Chaney et al., 1989).This method can provide a practical tool for plant breeders working with crops prone to Fe chlorosis, such as soybean and chickpea (Cicer arietinum L.) (Dragonuk et al., 1989a). The Fe(I1)-specificchelator ferrozine has been proposed as a means of buffering Fez+ activities in nutrient solutions (Parker et al., 1995a). Such a method would be useful for simulation of flooded soils, as well as for physiological investigations of Fe(I1) transport into root cells. Continued experience with this ligand has convinced us, however, that its utility is somewhat limited. Despite its affinity for Fe(II), ferrozine cannot prevent oxidation to Fe(II1) in a well-aerated solution, and this oxidation is greatly enhanced if a second ligand such as EDTA is included to buffer other metals (W. A. Norvell and D. R. Parker, 1998, unpublished results). Ferrozine has, however, proved useful for studying the short-term (560 min) unidirectional influx of Fe(I1) into pea roots (Fox et al., 1996).Among the significant findings of this work is that both Fe(I1) influx and translocation rates were elevated in Fe-deficient plants, suggesting that it is not only the activity of the extracellular Fe(II1)-chelatereductase that is stimulated by Fe stress. Chelator buffering has been recommended for environmental research with trace elements such as Cd (Chaney et al., 1989). The chelator EGTA is comparatively Cd specific and may be useful for buffering Cd2+ activities at low and environmentally relevant levels (Parker et al., 1995a). In their study of how complexation with C1 affects Cd uptake by Swiss chard (Beta vulgaris L.), Smolders and McLaughlin (1996b) used NTA or EGTA to buffer Cd2+ at about lop9 or lo-" M,respectively. McKenna et al. (1993) studied Zn-Cd interactions in lettuce (Latuca sativa var. longifalia) using solutions containing an excess of EDDHA. Although Cd2+ and Zn2+ activities were undoubtedly buffered at low values in these solutions, calculated activities were not reported by the authors. This is appropriate because commercially available EDDHA is actually a combination of the racemic mixture and the meso isomer (Bannochie and Martell, 1989). The rac and meso forms have significantly different stability constants for the trace metals, making accurate calculation of the metal speciation difficult. As a consequence, we would not recommend the use of EDDHA for most chelator-buffered applications. To our knowledge, the only other published reports using CBNS to study Cd uptake by higher plants have been abstracts (Chaney and Green, 1994; Norvell et al., 1995), although chelator-buffered media have been used to screen strains of Rhizobium spp. for tolerance to Cd, Cu, Ni, and Zn toxicities (Angle and Chaney, 1989; Chaudri et al., 1993).

ADVANCES IN SOLUTION CULTURE METHODS

189

One issue that has not been heretofore discussed in detail is the possible biodegradation of chelators, particularly if used in long experiments with infrequent or no solution replacement. In small-pot studies, frequent solution replacement is usually necessary and greatly minimizes the possibility of significant chelator degradation. This may not be the case, however, in large-tank systems such as the one described in Section IIIB. Degradation can be quite rapid in soil (Norvell, 1991), but has not been studied in nutrient solutions. One barrier to such studies is the lack of simple and convenient analytical methods for unambiguously quantifying the chelators in solution. Progress in this area would greatly facilitate studies of chelator degradation in nutrient solutions.

3. Buffering of EH @E) Although not a nutritional factor per se, oxidation-reduction potential (expressed as E, or p e ) is a parameter of interest for those studying plant response to poorly drained or wetland soils. Recently, some progress has been made in chemical buffering of E, at low levels in nutrient solutions by including a Ti(II1)-citrate buffer (DeLaune et al., 1990). The more common practice of bubbling an inert gas such as N, instead of air through the solution only excludes oxygen from the system. The redox potential can still be quite high (EH 350 mV) in the absence of 0,, in contrast to E, values as low as -250 mV in anaerobic wetland soils (DeLaune etal., 1990). Originally developed for the culture of anaerobic bacteria, the Ti(1II)-citrate buffer in conjunction with an 0,-free bubbling solution can buffer nutrient solutions at E, values of -200 to -300 mV. The gradual oxidation of Ti(II1) to Ti(1V) scavenges any trace 0, and provides a sink for oxygen that may be transported from shoot to roots via aerenchyma (DeLaune et al., 1990; Brix and Sorrell, 1996). This reaction was used by Sorrell and Armstrong (1994) to measure actual 0, release rates from roots of wetland species in response to the highly anaerobic nutrient solution. Brix and Sorrell (1996) showed that two wetland plant species were unaffected by growth in nutrient solutions that were merely deoxygenated. Only when the solutions were made truly “reducing” with Ti(II1)-citrate did these species exhibit hypoxia with consequent reductions in growth rate, net photosynthesis, and root concentrations of adenine nucleotides. Inclusion of Ti(1V)-citrate in aerated control solutions has allowed investigators to discount Ti toxicity as a contributor to reduced growth in the anaerobic, Ti(II1)-buffered solutions (Brix and Sorrell, 1996). An additional convenience of this method is the intense blue color of the Ti(II1)-citrate complex, which disappears upon oxidation to Ti(1V)-citrate (DeLaune et al., 1990).Thus, one could easily monitor and compensate for the gradual loss of the E, buffer by making daily absorbance readings using a simple spectrophotometer or colorirneter.

-

190

DAVID R. PARKER AND WENDELL A. NORVELL

D. BUFFERINGWITH ION-EXCHANGE OR CHELATING RESINS Hewitt (1966) reviewed the early work on using synthetic resins, zeolites, and clays to supply nutrients and to moderate pH changes in solution and sand culture. Much of the initial interest in these materials was related to the need in sand culture to add substantial quantities of nutrients, while avoiding osmotic stress and minimizing losses from precipitation or leaching. Not surprisingly, the earliest work focused on providing the nutrients needed in the greatest quantities, that is, nitrate from anion exchangers and macroelement cations from cation exchangers (Schlenker, 1940; Converse et al., 1943; Jenny, 1946). The many uses and effects of exchange materials in nutrient cultures are not easily classified, because the inclusion of solid-phase exchangers increases the number of interactions that may occur among solution components. However, work with the synthetic ion-exchange resins can be grouped into three general areas: buffering of pH, supplying exchangeable macronutrients, and regulating micronutrients or toxic metal ions.

1. pH Buffering Efforts to buffer pH changes in nutrient culture solutions have generally used synthetic ion-exchange resins with weakly acidic carboxyl groups. Mixtures of Ca-form and H-form resin are often used, but the manner in which solution and resin are brought into contact varies. Hageman et al. (1961) were among the first to use Amberlite IRCJO (Rohm and Haas, Philadelphia, PA) to control pH in solution culture. Mixtures of Ca-form and H-form resin were simply added to solutions to maintain desired pH values during growth of maize. They found that different combinations of these resins were successful in holding pH in the ranges 4-4.5, 5.5-6.0, or 6.2-7.4. Similarly, Bugbee and Salisbury (1985) used direct addition of resin in their study of pH buffering by IRC-50 and MES. When enough resin was added to buffer pH effectively, however, significant losses of Mg and Mn from solution occurred. Checkai et al. (1987a,b) also added mixtures of Caform and H-form resins to nutrient solutions, but they placed the resin within membrane-filter pouches to separate the resin from roots. The pHs of their solutions were controlled within a few tenths of a pH unit by the weak-acid resin Bio-rex 70 added at a rate of about 12 g per liter of culture solution. Harper and Nicholas (1976) attempted to improve equilibrium between nutrient solution and weak-acid resin by recirculating the solution through a column of resin contained within the pot. The pressure differential created by the aeration stream was used to pull the solution through the resin and expel it from the column. This recirculator improved the control of pH by IRC-50 in their studies of soybean nodulation, nitrogen fixation, and nitrogen uptake. Further studies of nitrogen nutrition were carried out by Polisetty and Hageman (1985) using the same

ADVANCES IN SOLUTION CULTURE METHODS

191

system to control pH. Nicholas and Harper (1993) later used this technique to compare buffering by weak-acid resin with the soluble pH buffer, MES. Miyasaka et al. (1988) devised a simpler and much smaller aeration-driven circulator for use with another weak-acid resin (Bio-rex 70), but flow rates were not high enough for effective pH control, and the rates declined further during operation because of plugging by microbial debris. Checkai et al. (1987~)and Checkai and Norvell (1992) overcame resistance to flow by pumping the nutrient solution through the weak-acid resin. In their studies with tomato, the mixture of Ca-form and H-form resin was placed in an external column, and solution from the pot was recirculated at a rate of about six pot volumes per hour. The control of pH was excellent, with pH rising from about 6.1 to only 6.4 over a period of 6 weeks. Despite this success with pH control, the circulating system required considerable attention because of mechanical problems with pumps, heat generated by the pumps, and leaky tubing connections. Recently, Righetti e l al. (1991) reported that a new type of amphoteric resin buffer with acrylamido functional groups may have promise for pH control in culture solutions. The proportions of acidic and basic groups on this resin may be adjusted during synthesis to vary the isoelectric point and thus to establish a number of effective buffering ranges. A resin with an isoelectric point of pH 5.6 provided excellent control of pH during growth of endive (Cichorurn endivia) in solutions containing either NH,-N or NO,-N. Although plants grew well in the presence of this resin, no information is available concerning possible effects on solution characteristics other than pH.

2. Major and Minor Elements Amon and Grossenbacher (1947) appear to have been the first to use synthetic exchange resins to supply all of the macronutrient elements (as well as several of the micronutrients). Mixtures of cation- and anion-exchangeresins were added to sand to provide the nutrients required to grow tomato for several months. Tomato was grown successfully in these resin systems, but the growth was not as good as in conventional solution culture. These authors observed and discussed several important characteristics of ion exchange, including the need for counterions in solution to permit effective release of adsorbed ions and the decrease in availability of an ion as its relative abundance on the exchanger decreases. With few exceptions (Welch etal., 1954; Skogley and Dawson, 1963),there was little effort after the 1940s to regulate multiple elements in nutrient culture with synthetic exchange resins until the work of Checkai and co-workers (Checkai et al., 1987a,b,c).Checkai etal. (1987a,b) used a mixture of soluble components and several types of resins to supply and regulate concentrations of all macronutrient cations, all micronutrient cations, and P. The macronutrient cations Ca, Mg, and K were loaded individually onto aliquots of strongly acidic cation-exchange resin

192

DAVID R. PARKER AND WENDELL A. NORVELL

(Dowex 50W-X4). An additional aliquot of the strong acid resin was saturated with Al, partially neutralized, and treated with P to provide a large source of sorbed P for release to nutrient solutions. As noted in Section IIID1, a pH buffer was prepared by loading additional Ca (along with H) onto a weakly acidic cation-exchange resin by equilibration with Ca acetate at pH 6. A chelating resin (Chelex 100) was selected to regulate the micronutrient cations Cu, Zn, Mn, Ni, and the heavy metal cation Cd. The ratios of these metals and Ca were adjusted to provide metal activities suitable for plant growth and for study of Cd uptake. Ferrous iron also was added to the chelating resin (where, at the solution pH of 6, it presumably was oxidized and sorbed as amorphous ferric hydroxides). This Fe provided a reserve to maintain concentrations of FeEDDHA in solution. Checkai etal. (1987a,b) placed portions of each prepared resin in tubular pouches made of plastic membrane filters, and then placed the filled pouches in the pot to equilibrate with the nutrient solution. Tomato seedlings were grown at four different treatment levels of Cd for 25 days using this culture system. The chelator EDTA was added to some pots to increase the total concentration of soluble micronutrient metals. Electrolyte levels in solution were maintained by addition of dilute nutrient solution to replace water lost by evapotranspiration, and one supplement of KNO, was added at 20 days. The pH and concentrations of macronutrient and micronutrient cations were well maintained in this culture system. Ionic activities of the micronutrient metals were held at levels similar to those in soil solutions and much lower than typically found in traditional nutrient solutions. Even P was well maintained in the concentration range of 4 to 10 pM, which also is more typical of soil solutions. Root hair development was prolific, probably in response to the more realistic levels of soluble P. Checkai et al. (1987b) used this mixed resin system to compare the uptake of chelated and ionic trace metals and concluded that uptake was regulated primarily by ionic activities rather than by total concentrations in solution. Mullins and Sommers (1986), Checkai et al. (1987c), and Checkai and Norvell (1992) modified the system described above by using pumps to circulate solution from pots through columns containing the resins. This was done in part to overcome the limitations imposed by slow diffusion out of the membrane pouches used by Checkai et al. (1987a). Mullins and Sommers (1986) combined all resins required for a particular treatment in a single small column and studied the concentration-dependent kinetics of Cd uptake by maize. Checkai et al. (1987~)and Checkai and Norvell (1992) modified the method further by placing the four major types of resins in separate columns to permit the resins to be recharged with nutrients. The four columns were connected in series (Fig. 10): (1) strong-acid resin for regulation of ratios of major cations, (2) weak-acid resin for pH control, (3) Al-treated resin with adsorbed P to supply P, and (4) chelating resin for regulating ratios of micronutrient metals. In addition, a guard column was included to extend system life by filtering out debris and microorganisms. Magnetically dri-

ADVANCES IN SOLUTION CULTURE METHODS

193

Figure 10 Components of a recirculating resin-buffered system: (1) magnetically coupled pump, (2) guard column, (3) strong-acid resin column for exchanging major cations, (4)weakacid column for pH buffering, ( 5 ) Al-treated resin column for releasing sorbed P,(6) chelating resin column for exchanging micronutrient metals, (7) plant culture container with tomato plants in polyethylene cups, and (8) root-zone temperature control cabinet for regulating solution temperatures. Reprinted from Checkai and Norvell (1992). A recirculating resin buffered hydroponic system for controlling nutrient ion activities. J. Plant nut^ 15,871-892, p. 879 by courtesy of Marcel Dekker, Inc.

ven pumps with ceramic pumping heads were used to recirculate solutions. All columns, tubing, and fittings were made of polyethylene or polypropylene to prevent contamination of solutions with metals, and all parts were opaque or shielded to prevent exposure of solutions to light. The resin preparation methods of Checkai et al. (1987~)and Checkai and Norvell ( 1 992) were generally similar to those of Checkai et al. (1987a) except that commercial-grade resins were chosen to lower costs (but then required extensive cleaning before use). Using relatively large resin columns to permit longer periods of operation, Checkai and Norvell(l992) grew tomato plants for 6 weeks in treatments with and without Ni in an effort to induce and characterize Ni deficiency. Water or dilute solutions of nitrate salts of K or Ca plus Mg were added occasionally to compensate for evapotranspiration. Throughout the growth period, no nutrient solution had to be changed. Whereas the resin-buffered system held nutrient concentrations in ranges generally similar to soil solutions, the control of trace-metal concentrations was not as effective as desired. Several opportunities

194

DAVID R. PARKER AND WENDELL A. NORVELL

to improve resin preparation and composition were discussed by Checkai and Norvell(l992). Finally, it is worth noting that while resin buffering of multiple nutrients is conceptually straightforward, the mechanical systems involved can be troublesome. For example, Checkai and Norvell(l992) experienced occasional difficulties with power interruptions, leaking tubing connections, pump failures, and heat generated by pumps. Indeed, the generated heat necessitated placing the pots in root-zonetemperature tanks to cool the solutions and then locating the entire experimental system inside a large environmentally control chamber. Unfortunately, mechanical failure of the cooling system caused premature termination of the entire experiment because of serious overheating of the chamber during an overnight breakdown. Several other investigations of resins as controls for nutrients or metals in plant culture solutions have occurred recently. Without plants, Kerven et ul. (1993) measured the chemical desorption of major nutrient ions from five ion-exchange resins (two cation and three anion exchange) that might be used to regulate nutrients in culture solutions of low ionic strength. A nitrate-selective resin, Wofaiit SN 36-L, showed potential to regulate NO, with SO, as a counterion, but none of the anionexchange resins were particularly effective in buffering SO,. They, like earlier workers such as Arnon and Grossenbacher (1947), found that cation-exchange resins were ineffective in maintaining soluble Ca and Mg via exchange with relatively low concentrations of K. Smolders and McLaughlin (1996a) circulated a complete nutrient solution through columns of chelating resin loaded with Cu, Zn, Mn, and '09Cd to provide buffering of ionic Cd for studies of Cd complexation by C1. The highly Fe-specific chelating agent HBED was used to provide soluble Fe, and pH was controlled with MES. They observed that elevated concentrations of C1 increased the uptake of Cd by Swiss chard, even when the activity of ionic Cd was little changed. Although the mechanism of this increase is not certain, they concluded that the most likely cause was some direct uptake of Cd in complexes with C1. This experiment is another example of successful use of chelating resins to provide control of the ionic activities of micronutrient and trace-metal levels. A boron-specific resin, Amberlite IRA-743, was used by Asad et al. (1997a) to supply buffered concentrations of B to canola (Brussica napus L.). The concentration of B was well maintained at a low level of 4 IJ.M, or at a moderate level approximately 25 times higher. Separate elution studies showed that the resin equilibrated rapidly with solution and maintained B concentrations in solution quite well, despite decreases in the saturation of the resin. Asad et al. (1997b) used this resin to investigate the external solution requirement of canola for soluble B. In summary, the most successful uses of resins in plant nutrient solutions have been the use of weak-acid resins to buffer solution pH, and the use of chelating resins to maintain low but well-buffered activities of micronutrient or trace-metal ions. The weak-acid resins provide an alternative to soluble pH buffers such as

ADVANCES IN SOLUTION CULTURE METHODS

195

MES or HCO,, and the chelating resins may provide an alternative to chelator buffering as a means of regulating metal activities. The recent use of a B-specific resin in culture solutions is encouraging and should permit continuing studies of B nutrition over a wide range of better-stabilized concentrations. In all cases where exchange resins are being considered for regulation of nutrient solution composition, the comment of Converse et al.( 1943) remains pertinent: “The time required to set up the original cultures is somewhat greater than for solution cultures, because it is necessary to prepare different ion-exchange materials in rather large quantities in advance.” To which we add that the simpler hydraulic systems for equilibrating solution with resin have generally been much more successful than the complex.

E. BUFFERINGWITH INORGANIC SOLIDPHASES 1. pH Buffering Calcium carbonate has been added to nutrient solutions for at least a hundred years to moderate pH changes, supply Ca, reduce ammonia toxicity, or exacerbate micronutrient stress (Hewitt, 1966). The amounts added are commonly in the range 0.1 to 1.O g CaCO, liter-’. The pH values of culture solutions containing CaCO, are not precisely controlled, but they are generally in the range 6.5 to 8.0 when the solutions are well mixed and aerated with ambient air. The presence of solid CaCO, in a culture solution provides substantial resistance to pH decreases. Dissolution of CaCO, releases alkalinity in the form of C0,’- into solution, which consumes protons and increases pH and the concentration of bicarbonate. Bicarbonate itself provides a fast-reacting buffer in solution, and the solid-phase CaCO, serves as a large reservoir of additional alkalinity to resist acidification. At least in principle, a solution saturated with CaCO, can also provide resistance to increases in pH. The precipitation of CaCO, from solution removes alkalinity, decreases bicarbonate concentrations, and releases protons to resist pH increases. However, significant precipitation of CaCO, is unlikely unless large amounts of alkalinity are introduced and the pH is much higher than is normally tolerated in culture solutions. Moreover, supersaturation of solutions with respect to CaCO, is common (Suarez et al., 1992), and supersaturation would certainly be expected in the complex ionic environment of plant nutrient solutions. Thus, precipitation of CaCO, per se is unlikely to be an effective means of pH control in plant culture solutions. There is, of course, resistance to pH increases in any alkaline solution that is in equilibrium with CO, in the gas phase, because protons are released by dissolution and hydrolysis of CO, (see Section IIIC1). However, this pH-buffering reaction is not a specific benefit of CaCO, additions.

196

DAVID R. PARKER AND WENDELL A. NORVELL

Additions of CaCO, have successfully maintained near-neutral or alkaline pH values in nutrient solutions in a wide variety of studies. For example, Wilson and Reisenauer (1963) and Johnson and Youngblood (1991) used CaCO, at 1 g liter-’ to maintain pH between about 6.5 and 7.5 in studies of nitrogen uptake or fixation. Similarly, CaCO, has been added commonly to culture solutions used to screen plants for iron efficiency or for resistance to chlorosis (Byron and Lambert, 1983; Hershey, 1991; McKenzie et al., 1984; Stevens et al., 1993; Wallace et al., 1978), or to help induce other micronutrient stresses (Halvorson and Lindsay, 1977; Swietlik and Zhang, 1994). Although it is clear that CaCO, can provide a measure of pH control in nutrient solution, the presence of this solid phase also creates some difficulties in regard to control of other constituents. Several nutrients, including phosphate and many micronutrient metals, may be lost from solution as a result of precipitation, adsorption, or oxidation on carbonate surfaces. Thus, the benefit of modest stability in pH is offset by unpredictable instability in concentrations of other constituents. Sometimes, NaHCO, or KHCO, is added to solutions in addition to CaCO, to increase pH further and improve pH stability (Coulombe et al., 1984a; Alcantara et al., 1988; Tang et al., 1996). In a few studies, additions of CaCO, and soluble bicarbonate have been combined with elevated pressures of CO, to regulate pH at lower values in the presence of high bicarbonate (Coulombe et al., 198413; Jalil and Carlson, 1993). Although these combinations improve the control of pH, they retain the disadvantages of having an unnecessary and reactive solid phase present in the culture solution.

2. Nutrient Buffering A seemingly obvious solution to the absence of buffering capacity for nutrients in small-volume solution culture is to include inorganic solid phases similar to those that regulate nutrients in soils. Such buffering might be achieved by using discrete, sparingly soluble mineral phases, or by using nutrients sorbed to an appropriately reactive colloidal surface, so that rapid dissolution or desorption would maintain nutrient concentrations as they are depleted by plant uptake. As we shall see, both approaches have been tried, but the distinction between them is somewhat blurred and perhaps semantic. To our knowledge, all efforts to date using this approach have focused on the buffering of P at concentrations approximating those found in soil solution. Cassman et al. (1981) utilized P sorbed into goethite to culture Rhizobiu in the absence of any host plants. The goethite was confined in pouches made of dialysis tubing such that bacterial cells were not in direct contact with the goethite. Consequently, P had to desorb from the goethite surface and diffuse across the dialysis membrane barrier prior to uptake. This system provided adequate P for growth of the bacteria at aqueous P concentrations as low as 0.03 FM, but only at cell den-

ADVANCES IN SOLUTION CULTURE METHODS

197

sities Ilo6ml- ‘.At higher densities the rate of Pdemand during log-phase growth outstripped the ability of the goethite-dialysis membrane system to replenish solution P (Cassman et ul., 198l). Further evaluation of this system has not been conducted. The most-studied system for buffering solution Pat low levels involves the sorption of orthophosphate on “alumina” that has been coated onto washed quartz sand. First introduced by Coltman et al. (1982), the method usually entails sand culture of the plants, but a few studies have utilized solution culture of the plants with the nutrient solution being recirculated through a column of sand-alumina (Elliott et al., 1983). The “alumina” used in all of these studies has been a single commercial product (Alcoa F-1; 300-600 pm dia.). Gourley et a1.(1993) used X-ray diffraction and electron microscopy to show that this product is a mixture of crystalline boehmite (yAIOOH) and its slightly more hydrated analog, pseudoboehmite (Hsu, 1989). The sand-alumina method has successfully imposed different phosphorus regimes on a variety of plant species (Coltman et al., 1982; Elliott et al., 1983; Pereira and Bliss, 1987; Buso and Bliss, 1988; Gourley et al., 1993). Coltman et al. (1985) used the method to separate seven tomato genotypes according to their “P efficiency.” In a subsequent study, these authors were able to select different P regimes that resulted in physiologically “equivalent stress” levels in six genotypes, and showed that genotypic characteristics associated with efficiency were still evident despite the similarity in growth reductions (Coltman et al., 1986). Pereira and Bliss (1987) employed the method to study the P responsiveness of three bean genotypes, including effects on nodule size and N, fixation. Using a similar approach, Lynch et ul. (1990) were able to grow beans to maturity and to examine the responsiveness of this species to mycorrhizal infection when grown at low P status. The lowest solution P concentrations that have been established by this method have generally been between 1 and 5 p M , with intermediate and higher values generally ranging from 10 to 50 p M (see Table 2 in Gourley et al., 1993). These values are high in comparison to typical soil solution concentrations (Section 11), and this fact is further evidenced by reported plant responses. For example, field studies reported by Nishimoto et al.( 1977) showed that 75% of maximum yield was reached at soil solution P concentrations ranging from 0.3 to 6 pA4, depending on crop species and soil type; for one-half of the 10 soil-crop combinations less than <2 p M was needed. In contrast, relative yields of 75% are reached at about 8 to 15 pLM P in sand-alumina culture (Coltman et al., 1982, 1986; Gourley et al., 1993). Similarly, Lynch et al. (1990) obtained bean yields at 1 pM P in sand-alumina culture that were 25% of those obtained at 27 pM P. With mycorrhizal inoculation, relative yields were doubled at 1 p M , suggesting that the absence of mycorrhia in most sand-alumina cultures may help account for the seemingly high P concentrations required (as compared to soil-grown plants). Alternatively, root

198

DAVID R. PARKER AND WENDELL A. NORVELL

morphology in sand-cultured plants may be sufficiently different than in soil to alter P requirements (Barber, 1995). Our experience with sand culture has been that it produces stunted, thickened roots without extensive lateral formation, a morphology that tends to minimize surface areas available for P uptake. One obvious drawback to the sand-alumina technique is its largely empirical nature. In contrast with the previously discussed chelator-buffered solutions for manipulating trace-metal availability, solution P concentrations cannot be predicted but must be arrived at through trial and error. Prior experience helps, but problems with reproducibility have been noted even when identical protocols were followed (Coltman et al., 1982; Gourley et al., 1993). Another disadvantage of sand-alumina culture is that the prepared pots require extensive leaching for 2 to 4 weeks to obtain stable solution P concentrations, and the required leaching period seems to vary from batch to batch (Gourley et al., 1993). Much of this inherent variability is likely due to the nature of the P sorption chemistry. Gourley et al. (1993) found that when treated with P, the pseudoboehmite phase of the alumina is lost. In its place, an amorphous A1-K-P mineral is formed, probably a precursor of potassium tarakanite [H,K,Al,(PO,),~ 18H,O]. This poorly ordered precipitate seems to coexist with amorphous Al(OH), in a solid solution, the relative amounts depending primarily, but not exclusively, on the P concentration of the “loading solution” (Gourley et al., 1993). This ill-defined solid-solution chemistry probably accounts for the undesirable variability encountered across independent experiments (Gourley et al., 1993), but it is responsible for the desirable ability to vary “equilibrium” P concentrations among the imposed treatments. It likely also accounts for the significant reduction in P availability caused by aging of P-loaded alumina (Coltman et al., 1982). Other common difficulties with the sand-alumina method may include rapid and wide fluctuations in solution pH within experiments (Coltman et al., 1982; Pereira and Bliss, 1987). Such fluctuations are particularly undesirable given the effects of solution pH itself on P uptake rates by roots (Marschner, 1995). Also, at lower pH values (i.e., 6 . 0 ) there is the possibility of induced Al toxicity due to enhanced solubility of the alumina substrate (Lynch et al., 1990), although such an effect has not been documented. Other limitations to the sand-alumina method are more subjective and require careful consideration of experimental goals. As evidenced by declines in solution P concentrations, buffering is often inadequate at the lower P levels, particularly in long-term experiments when plants get large (Elliott et al., 1983; Gourley et al., 1993). Lynch et al. (1990) partially overcame this obstacle by including P at its three different treatment concentrations in the nutrient solutions used to flush the pots twice each day. But, such an approach begins to blur the distinction between the sand-alumina method and FSC, with a loss in the convenience of the former method. Other data indicate that, at a constant solution P concentration, plant P status shows a marked dependence on the total P loading on a per pot basis (i.e., the

ADVANCES IN SOLUTION CULTURE METHODS

199

grams of P-loaded alumina per kilogram of sand) (Coltman et al., 1982). Thus, the method does not show the precise control of nutrient availability afforded by methods such as FSC and chelator buffering, where the (available) nutrient concentration or activity is more invariant and stable. On the other hand, these features of the sand-alumina system are more reminiscent of the behavior and availability of Pin soil (Elliott et al., 1983). Specifically, the severity of P deficiency seems to depend on both quantity and intensity factors, as predicted by mechanistic models of nutrient uptake from soils (Barber, 1995). Accordingly, Coltman et al. (1982) and Gerloff (1987) have referred to the “diffusion limited” uptake of P in the sand-alumina systems. To date, the demonstrated utility of the sand-alumina system has been for imposing different P regimes upon one or more plant species and for investigating the nature of differential plant tolerance of low-P soils. The method has not enjoyed any use as a means of imposing “realistic” P levels while concurrently studying other nutritional problems. An obvious candidate would be Zn deficiency because of the extensively studied, albeit oft-confusing, interactions between P and Zn in plants (Parker et al., 1992; Webb and Loneragan, 1988). The highly reactive, P-rich surfaces of the alumina would, however, make it quite difficult to control the chemistry and bioavailability of the trace metals. Zinc could sorb to the alumina surface, thus reducing its concentration (Coltman et al., 1982), or, more likely, might be present as a trace contaminant in the alumina that would preclude attainment of Zn deficiency (Parker et al., 1995a). The sand-alumina method is most probably incompatible with chelator buffering because A13+ would compete for the chelating ligand, releasing other metals into solution in unpredictable amounts. Control of the activities of Cu, Mn, and Zn would thus be unachievable. This shortcoming has led to some investigations of the potential of hydroxyapatite [HAP; Ca,(PO,),OH] for maintaining P availability in studies of Zn deficiency. Specifically, attempts have been made to use HAP to provide adequate P for normal growth, while simultaneously preventing hyperaccumulation of P to toxic concentrations in Zn-deficient plants (Parker, 1993; Gries et al., 1995). Such accumulation rarely occurs in field-grown plants, presumably because the solubility of P is low and the rate of P uptake is diffusion limited in soils. Thus, P toxicity is almost exclusively an artifact of solution culture experiments (Parker, 1993, 1997). The least soluble of the common Ca-phosphates, HAP is conveniently available as a slurry of crystalline particles of uniform size. Parker (1993) used chelatorbuffered solutions with commercial HAP enclosed in a pouch constructed of dialysis tubing. Dissolution and diffusion out of the pouch occurred as plants depleted solution P. Maize was adequately supplied with P using this approach, although the control plants had shoot P concentrations that may have been only marginally adequate or slightly deficient (<2 mg g-I). Older leaves of acutely Zn-deficient maize contained normal P levels (ca. 3 mg g-I), in contrast to values of >40 mg

2 00

DAVID R. PARKER AND WENDELL A. NORVELL Table VII

Responses of Wheat (cv Yecora Rojo) Seedlings to 14 Days of Growth in Chelator-Buffered Nutrient Solutions Where P Was Supplied as a Suspensionof Hydroxyapatite Contained within a Pouch Made of Dialysis Tubing (12,000-14,000 MWCO) Relative yield

Shoot Zn concentration

Shoot P concentration

log (Zn2+)

(%)

(w g - 9

(wg-‘)

6

-11.7 -11.0 - 10.5 - 10.2

44.8 2 1.6 72.6 C 4.9 103.1 ? 7.0 102.1 C 2.5

10

- 10.0

100.0 ? 1.3

14.6 19.4 30.6 45.2 48.8

9.7 4.7 3.2 3.0 3. I

Total Zn

(W) 0.2 1

3

Nore. The methods were reported by Parker ( I 993) but the data are previously unpublished

’

g- observed with a “conventional” P supply at just 10 pM P (Parker, 1993,1997). With HAP as the P source, shoots of Zn-deficient wheat exhibited slightly elevated P concentrations of about 10 mg g- but Zn-sufficient plants contained normal levels of 3 mg g-’ (Table VII). Again, these shoot P levels are far less than those observed in wheat grown in a 10 p M P solution (Parker, 1997). Using barley, Gries et al. (1995) observed results very similar to those depicted in Table VII, but Parker (1993) reported that two dicots, tomato and snapbean, were too P-inefficient to grow normally with HAP as the sole P source. Thus, the HAP-dialysis pouch system has enjoyed only limited success to date. It seems to provide adequate P to cereals, but care must be used to avoid deficiencies in large plants during rapid growth (Parker, 1993). The system greatly limits, but does not completely eliminate, hyperaccumulation of P in Zn-deficient cereals, but has thus far not been successfully utilized with dicots. As with the sandalumina method, a priori predictions of P availability from the HAP pouches cannot be readily made. Factors such as the molecular weight cutoff of the dialysis tubing, the size and number of pouches per pot, and the amount of HAP per pouch all must be arrived at via trial and error (Parker, 1993). This greatly increases the amount of preliminary experimentation that must be done before nutritional studies can actually be conducted. These drawbacks partially negate the obvious advantage of HAP relative to the sand-alumina systems: It is a discrete, crystalline solid phase of definable solubility. Unfortunately, HAP placed “loose” in the nutrient solution tends to oversupply P (D. R. Parker, 1998, unpublished data), while containment with dialysis tubing creates kinetic limitations due to dissolution and/or diffusion that are not readily predicted. Further research is needed to surmount these difficulties in using HAP to supply P to different species under varying growth conditions.

’,

ADVANCES IN SOLUTION CULTURE METHODS

201

W. SUMMARY AND FUTURE OUTLOOK Since Asher and Edwards (1983) last reviewed the topic, there have been significant advances in the sophistication of solution culture methods employed for mineral nutrition research. Many of these developments have been aimed at overcoming the poor buffering of traditional nutrient solutions with respect to both nutrient concentrations and pH. Many nutrients, especially P and the micronutrient metals, need to be maintained at extremely low concentrations to accurately simulate soil solutions and to avoid unwanted elemental toxicities in sensitive species. With the trace metals (notably Cu and Zn), attainment of deficiencies has often been hampered by contamination derived from macronutrient salts and/or laboratory apparatus. Even small amounts of contamination can meet a large portion of the plant growth requirements. Both excessively high initial nutrient concentrations and severe depletion of those concentrations can often be avoided by frequent and complete replacement of the nutrient solutions. Simple calculations, facilitated by electronic spreadsheets, can be used to estimate the needed frequency of replacement, as long as reasonable estimates of growth rate and plant nutrient levels are available. Related, but more complex, approaches include the programmed nutrient addition and relative addition rate methods wherein exponentially increasing quantities of nutrients are frequently (i.e., daily) added to solution. This approach has been most extensively used in studies of N nutrition wherein maintenance of constant, but reduced, relative growth rate has been considered desirable; future applications to other nutritional problems may reveal additional advantages. The low buffering capacity of nutrient solutions can be partially overcome simply by using, where practical, very large solution volumes relative to the total plant biomass produced. At low nutrient concentrations, superior control of solution composition is provided by continuously flowing solution culture, especially if combined with large reservoir volumes to further enhance stability. Drawbacks of FSC include the large capital costs to construct even partially automated systems, the very limited number of independent treatments and/or replicates that can be imposed, and the need for some rather demanding and labor-intensive chemical analysis to monitor and maintain low nutrient concentrations. Major advances have been made in the use of soluble ligands to buffer selected nutrient solution parameters, including pH. Several of the common Good’s biochemical buffers, most notably MES, have proven quite suitable for buffering pH at ca. 5.5 or above without plant toxicity or other adverse consequences. At pH values of ca. 6.5 to 8.5, bicarbonate may suitably buffer pH as long as the relationships among pH, bicarbonate concentration, and atmospheric CO, are properly understood and manipulated. The latter approach is particularly useful for simulating calcareous soil conditions. There is still a need for a convenient buffer that

202

DAVID R. PARKER AND WENDELL A. NORVELL

would provide adequate pH stability in the range 4.0 to 5.5, so that acid-soil infertility problems could be more conveniently studied in solution culture. Soluble chelating agents such as EDTA and HEDTA have been successfully employed to buffer free ion activities of Cu, Fe, Mn, and Zn at levels low enough to attain severe deficiencies of these metals. These chelator-buffered solutions offer vastly superior control of trace-metal nutritional status, in large measure because background contamination is usually rendered inconsequential. Chelator buffering offers the additional convenience of predictability: Treatment levels can be accurately selected based on some preliminary data and simple speciation calculations such as those afforded by the GEOCHEM-PC program. Environmentally oriented research may be facilitated by this method, although such use has so far been confined to a few studies of Cd. Additional applications might include physiological studies of trace element ion transport processes, a field of study that has historically been hampered by the use of unrealistically high metal concentrations. Studies of the physiological roles of Ni, seemingly required at extremely low levels, might also benefit from the chelator-buffering approach. Ion-exchange resins have been successfully used to buffer pH and concentrations of both plant-essential elements and environmentally important trace elements such as Cd. Desirable features of these resins are their chemical purity, their ability to provide a large elemental “reservoir,” and their seemingly rapid reaction kinetics. Obstacles to their convenient use include the need for dependable mechanical systems for reliable recirculation of solution through resin columns and the need for more accurate predictions of equilibrium concentrations so that tedious empirical determinations of resin loadings can be minimized. Well-defined mineral solid phases offer some potential for the buffering of sparingly soluble nutrients, most notably P. The sand-alumina system, containing P loaded onto commercially available boehmite, has been extensively studied for its ability to impose varying degrees of P deficiency. Although generally successful, the method requires significant empirical testing to arrive at the proper treatment levels and seems to suffer from poor reproducibility. The P does not remain as a surface-adsorbed species on the boehmite, but instead forms a poorly ordered Ktarakanite that probably contributes to the undesirable variation encountered using this method. Crystalline hydroxyapatite, a rather insoluble Ca-phosphate, offers an alternative means of buffering P at low levels. Additional study is needed to develop a convenient and predictable means of establishing treatment P levels. Continued technological improvements in solution culture methodology are likely, especially for the trace elements and for P. Increasing knowledge of the underlying chemistry of culture solutions has facilitated the development of controls of nutrient availability, as exemplified by the chelator-buffering approach to manipulating trace-metal availability. Similar advances in our ability to predict the equilibrium and/or kinetic relationships should facilitate the use of “model” ion-

ADVANCES IN SOLUTION CULTURE METHODS

203

exchange resins and solid-phase minerals to more effectively buffer nutrient solution composition at realistically low concentrations.

REFERENCES Adams, F. (I97 I). Ionic concentrations and activities in soil solutions. Soil Sci. SOC.Am. Proc. 35, 420-426. Alcantara, E., Romera, F. J., and de la Guardia, M. D. (1988). Genotypic differences in bicarbonateinduced iron chlorosis in sunflower. J. Plant Nuts 11,65-75. Angle, J. S., and Chaney, R. L. (1989). Cadmium resistance screening in nitrilotriacetate-buffered minimal media. Appl. Environ. Microsc. 55,2101 -2104. Angle, J. S., McGrath, S. P., and Chaudri, A. M. (1992). Effects of media components on toxicity of Cd to Rhizobia. Water Air Soil Pollut. 64,627-633. Arnon, D. I., and Grossenbacher, K. A. (1947). Nutrient culture of crops with the use of synthetic ionexchange materials. Soil Sci. 63, 159-192. Asad, A., Bell, R. W., Dell, B., and Huang, L. (1997a). Development of a boron buffered solution culture system for controlled studies of plant boron nutrition. Plant Soil 188,21-32. Asad, A., Bell, R. W., Dell, B., and Huang, L. (1997b). External boron requirements for canola (Brassica napus L.) in boron buffered solution culture. Ann. Bor. 80,65-73. Asher, C. J. (1991). Beneficial elements, functional nutrients, and possible new essential elements. In “Micronutrients in Agriculture” (J. J. Mortvedt, F. R. Cox, L. M. Shuman, and R. M. Welch, eds.), 2nd Ed., Soil Sci. SOC.Am. Book Ser. No. 4, pp. 703-723. Soil Sci. SOC.Am., Madison, WI. Asher, C. J., and Blamey, F. P. (1987). Experimental control of plant nutrition status using programmed nutrient addition. J. Plant Nurr 10, 1371-1380. Asher, C. J., and Cowie, A. M. (1970). Programmed nutrient addition: A simple method for controlling the nutrient status of plants. Proc. Aust. Plant Nuts Con$ Mt. Gambier l(b), 28-32. Asher, C. J., and Edwards, D. G. (1978). Relevance of dilute nutrient solution culture studies to problems of low fertility tropical soils. In “Mineral Nutrition of Legumes in Tropical and Subtropical Soils” (C. S. Andrews and E. J. Kamprath, eds.). CSIRO, Melbourne. Asher, C. J., and Edwards, D. G. (1983). Modem solution culture techniques. Encycl. Plant Physiol. New Ser 15A, 94-1 19. Asher, C. J., and Loneragan, J. F. (1967). Response of plants to phosphate concentration in solution culture: I. Growth and phosphorus content. Soil Sci. 103,225-233. Asher, C. I., and Ozanne, P. G. (1967). Growth and potassium content of plants in solution cultures maintained at constant potassium concentrations. Soil Sci. 103, 155-161. Asher, C. J., Ozanne, P. G., and Loneragan, J. F. (1965). A method for controlling the ionic environment of plant roots. Soil Sci. 103, 155-161. Bannochie, C. J., and Martell, A. E. (1989). Affinities of racemic and meso forms of N,N‘-ethylenebis[2-(o-hydroxyphenyl)glycine] for divalent and trivalent metals. J. Am. Chem. SOC. 111, 4735-4742. Barber, S. A. (1995). ‘‘Soil Nutrient Bioavailability,” 2nd Ed. Wiley, New York. Bell, P. F., Chaney, R. L., and Angle, J. S. (1991a). Free metal activity and total metal concentrations as indices of micronutrient availability to barley [Hordeurn vulgare (L.) “Klages”]. Plant Soil 130, 51-62. Bell, P. E, Chaney, R. L., and Angle, J. S. (1991b). Determination of the coppes’ activity required by maize using chelator-buffered nutrient solutions. Soil Sci. SOC.Am. J. 55, 1366-1374. Bell, P. F., Chen, Y., Potts, W. E., Chaney, R. L., and Angle, J. S. (1991~).A reevaluation of the Fe(III),

2 04

DAVID R. PARKER AND WENDELL A. NORVELL

Ca(II), Zn(II), Cu(I1) and proton formation constants for 4,7,-diphenyl-l ,lo-phenanthrolinedisulfonate. B i d . Trace Elem. Res. 30, 125-144. Bell, P. F., Parker, D. R., and Page, A. L. (1992). Contrasting selenate-sulfate interactions in selenium-accumulating and nonaccumulating plant species. Soil Sci. Soc. Am. J. 56, 1818-1824. Bell, R. W., Edwards, D. G., and Asher, C. J. (1989). External calcium requirements for growth and nodulation of six tropical food legumes grown in flowing solution culture. Aust. J. Agric. Res. 40,85-96. Bhat, K. K. S. (1980). A low-cost, easy-to-install flow culture system suitable for use in a constant environment cabinet. J. Exp. Bot. 31, 1435-1440. Bloom, A. J., and Chapin, F. S. (1981). Differences in steady-state net ammonium and nitrate influx by cold- and warm-adapted barley varieties. Plant Physiol. 68, 1064- 1067. Bond, G. A. (1998). Screening oak species for iron efficiency. M. S. thesis, Cornell University, Ithaca, Ny. Bradfield, R. (1942). Calcium in the soil: I. Physico-chemical relations. Soil Sci. Soc. Proc. 6, 8-15. Breeze, V. G., Wild, A., Hopper, M. J., and Jones, L. H. P. (1984). The uptake of phosphate by plants from flowing nutrient solution. 11. Growth of Loliurn perenne L. at constant phosphate concentrations. J. Exp. Bof. 35, 1210-1221. Brix, H., and Sorrell, B. K. (1996). Oxygen stress in wetland plant species: Comparison of deoxygenated and reducing root environments. Funct. Ecol. 10,521-526. Brown, J. C. (1972). Competition between phosphate and the plant for Fe from Fez+ ferrozine. Agron. J. 64,240-243. Brown, J. C., Lunt, 0. R., Holmes, R. S., and Tiffin, L. 0. (1959). The bicarbonate ion as an indirect cause of chlorosis. Soil Sci. 88,260-266. Brown, J. C., Tiffin, L. O., Specht, A. W., and Resnicky, J. W. (1961). Iron absorption by roots as affected by plant species and concentration of chelating agent. Agron. J. 5 3 , 8 1 4 5 . Bruce, R. C., Warrell, L. A,, Bell, L. C., and Edwards, D. G. (1989). Chemical attributes of some Queensland acid soils. I. Solid and solution phase composition. A m . J. Soil. Res. 27, 333-351. Bugbee, B. G., and Salisbury, F. B. (1985). An evaluation of MES (2(N-rnorpholino)ethanesulfonic acid) and Amberlite IRC-50 as pH buffers for nutrient solution studies. J. Plant N u a 8,567-583. Buso, G. S. C., and Bliss, F. A. (1988). Variability among lettuce cultivars grown at two levels of available P. Plant Soil 111,67-73. Byron, D. F., and Lambert, J. W. (1983). Screening soybeans for iron efficiency in the growth chamber. Crop Sci. 23,885-888. Campbell, P. G. C. (1995). Interactions between trace metals and aquatic organisms: A critique of the free-ion activity madel. In “Metal Speciation and Bioavailability in Aquatic Systems’’ (A. Tessier and D. R. Turner, Eds.), pp. 45-102. John Wiley & Sons, New York. Carroll, M. D., and Loneragan, J. F. (1968). Response of plant species to concentrations of zinc in solution: l. Growth and zinc content of plants. Aust. J. Agric. Res. 19,859-868. Cassman, K. G., Munns, D. N., and Beck, D. P. (1981). Growth of Rhizobium strains at low concentrations of phosphate. Soil Sci. Soc. Am. J. 45,520-523. Chaney, R. L. (1988a). Metal speciation and interaction among elements affect trace element transfer in agricultural and environmental food-chains. In “Metal Speciation Theory, Analysis, and Application” (J. R. Kramer and H. E. Allen, eds.), pp. 219-260. Lewis, Chelsa, MI. Chaney. R. L. (1988b). Plants can utilize iron from Fe-N,N‘-di-(2-hydroxy-benzoyl)-ethylenediamineN,N‘-diacetic acid, a ferric chelate with lo6 greater formation constant than Fe-EDDHA. J. Planr NutL 11, 1033-1050. Chaney, R. L., and Bell, P. F. (1987). Complexity of iron nutrition: Lessons for plant-soil interaction research. J. Plant Nutc 10,963-994. Chaney, R. L., and Green, C. E. (1994). Effect of Zn-excess and Zn-deficiency on Cd uptake and translocation by wheat sources. Agron. Absrc, 324. Chaney, R. L., Coulombe, B. A., and Lloyd, C. A. (1984). Bicarbonate induced Fe-chlorosis with normal plant phosphorus levels. Agron. Absts, 201.

ADVANCES IN SOLUTION CULTURE METHODS

205

Chaney, R. L., Bell, P. F., and Coulombe, B. A. ( 1 989). Screening strategies for improved nutrient uptake and utilization by plants. HorrScience 24,265 -272. Chaney, R. L., Chen, Y., Green, C. E., Holden, M. J., Bell, P. F., Luster, D. G., and Angle, J. S. (1992a). Root hairs on chlorotic tomatoes are an effect of chlorosis rather than part of the adaptive Fe-stress response. J. Plunz Nutr: 15, 1857-1875. Chaney, R.L., Coulombe, B. A., Bell, P. F., and Angle, J. S. ( 1992b). Detailed method to screen dicot cultivars for resistance to Fe-chlorosis using FeDTPA and bicarbonate in nutrient solutions. J. Plant Nurr 15,2063-2083. Chapin, F. S., I11 (1988). Ecological aspects of plant mineral nutrition. In “Advances in Plant Nutrition, Volume 3” (B. Tinker and A. Lauchli, eds.), pp. 161-191. Praeger, New York. Chaudri, A. M., McGrath, S. P., Giller, K. E., Angle, J . S., and Chaney, R. L. (1993). Screening of isolates and strains of Rhizobium leguminosarum biovar trifolii for heavy metal resistance using buffered media. Environ. Toxicol. Chern. 12, 1643- 1651. Checkai, R. T., and Norvell, W. A. (1992). A recirculating resin-buffered hydroponic system for controlling nutrient ion activities. J. Planr Nutr: 15,871-892. Checkai, R. T., Hendrickson, L. L., Corey, R. B., and Helmke, P. A. (1987a). A method for controlling the activities of free metal, hydrogen, and phosphate ions in hydroponic solutions using exchange and chelating resins. Plant Soil 99, 321-334. Checkai, R. T., Corey, R. B., and Helmke, P. A. (1987b). Effects of ionic and complexed metal concentrations on plant uptake of cadmium and micronutrient metals from solution. PIanz Soil 99, 335-345. Checkai, R. T., Norvell, W. A., Welch, R. M., and Brown, P. H. (1987~).Using exchange resins to impose nutrient treatments on hydroponically grown seedlings from germination to transplanting. J. Plant Nutr: 10, 1447-1455. Clark, R. B. (1982). Nutrient solution growth of sorghum and corn in mineral nutrition studies. J. Plant Nurr: 5, 1039-1057. Clement, C. R., Hopper, M. J., Canaway, R. J., and Jones, L. H. P. (1974). A system for measuring the uptake of ions by plants from flowing solutions of controlled composition. J. Exp. Eot. 25,81-99, Clement, C. R., Hopper, M. J., and Jones, L. H. P. (1978). The uptake of nitrate by Loliumperenne from flowing nutrient solution. I. Effect on No,- concentration. J. Exp. Eoz. 29,453-464. Cline, G. R., Powell, P. E., Szaniszlo, P. J., and Reid, C. P. P. (1982). Comparison of the abilities of hydroxamic, synthetic, and other natural organic acids to chelate iron and other ions in nutrient solution. Soil Sci. SOC.Am. J. 46, 1158-1 164. Coltman, R. R., Gerloff, G. C., and Gableman, W. H. (1982). A sand culture system for simulating plant responses to phosphorus in soil. J. Am. Soc. Hort. Sci. 107,938-942. Coltman, R. R.,Gerloff, G. C., and Gableman, W. H. (1985). Differential tolerance of tomato strains to maintained and deficient levels of phosphorus. J. Am. Soc. Hort. Sci. 110, 140-145. Coltman, R. R., Gerloff, G. C., and Gableman, W. H. (1986). Equivalent stress comparisons among tomato strains differentially tolerant to phosphorus deficiency. J. Am. Soc. Hort. Sci. 111, 422-426. Converse, C. D., Gammon, N., and Sayre, J . D. (1943). The use of ion exchange materials in studies on corn nutrition. Plant Physiol. 18, 114-121. Coulombe, B. A., Chaney, R. L., and Wiebold, W. J. (1984a). Bicarbonate directly induces iron chlorosis in susceptible soybean cultivars. Soil Sci. Soc. Am. J. 48, 1297- 1301. Coulombe, B. A., Chaney, R. L., and Wiebold, W. J. (1984b). Use of bicarbonate in screening soybeans for resistance to iron chlorosis. J. Plant Nutr: 7,4 I 1-425. Curtin, D., and Smilie, G. W. (1995). Effects of incubation and pH on soil solution and exchangeable cation ratios. Soil Sci. SOC.Am. J. 59, 1006-101 I. David, M. B., and Lawrence, G. B. (1996). Soil and soil solution chemistry under red spruce stands across the northeastern United States. Soil Sci. 161,314-328.

206

DAVID R. PARKER AND WENDELL A. NORVELL

DeKock, P. C., and Mitchell, R. L. (1957). Uptake of chelated metals by plants. Soil Sci. 84,55-62. DeLaune, R. D., Pezeshki, S. R., and Pardue, J. H. (1990). An oxidation-reduction buffer for evaluating the physiological response of plants to root oxygen stress. Environ. Exp. Bot. 30, 243-247. Doney, R. C., Smith, R. L., and Weibe, H. H. (1960). Effects of various levels of bicarbonate, phosphorus, and pH on the translocation of foliar-applied iron in plants. Soil Sci. 80,269-275. Dragonuk, M. B., Fehr, W. R., and Jessen, H. J. (1989a). Effectiveness of nutrient-solution evaluation for recurrent selection for Fe efficiency of soybean. Crop Sci. 29,952-955. Dragonuk, M . B., Fehr, W. R., and Jessen, H. 1. (1989b). Nutrient-solution techniques for evaluation of iron efficiency of soybean. J. Plant Nutr 12,871-880. Dragun, J., Baker, D. E., and Risius, M. L. (1976). Growth and element accumulation by two singlecross corn hybrids as affected by copper in solution. Agron. J. 68,466-470. Edwards, D. G., and Asher, C. J. (1974). The significance of solution flow rate in flowing culture experiments. Plant Soil 41,161-175. Elliott, G. C. (1989). Evaluation of sand-alumina-P media for studies of P nutrition. J. Plant Nurr: 12, 265 -278. Elliott, G. C., Carlson, R. M., Lauchli, A,, and Rosen, C. J. (1983). A solid-phase buffer technique to maintain concentrations of phosphate in nutrient solutions. J. Plant Nutr 6,1043-1058. Epstein, E. (1972). “Mineral Nutrition of Plants: Principles and Perspectives.” Wiley, New York. Ericsson, T., and Ingestad, T.(1988). Nutrition and growth of birch seedlings at varied relative phosphorus addition rates. Physiol. Plant. 72,222-235. Eskew, D. L., Welch, R. M., and Norvell, W. A. (1984). Nickel in higher plants: Further evidence for an essential role. Plant Physiol. 76,691-693. Foehse, D., and Jungk, A. (1983). Influence of phosphate and nitrate supply on root hair formation of rape, spinach and tomato plants. Plant Soil 74,359-368. Food and Agriculture Organization of the United Nations (FAO) (1990). “Soilless Culture for Horticultural Crop Production,” FA0 Plant Production and Protection Paper No. 101. FAO, Rome. Fomo, D. A., Asher, C. J., and Edwards, D. G. (1979). Boron nutrition of cassava, and the boron X temperature interaction. Field Crops Res. 2,265-279. Fox, T. C., Shaff, J. E., Grusack, M. A., Norvell, W. A., Chen, Y., Chaney, R. L., and Kochian, L. V. (1996). Direct measurement of 59Fe-labeled Fe2+ influx in roots of pea using a chelator buffer system to control free Fez+ in solution. Plant Physiol. 111,93-100. Gerloff, G. C. (1987). Intact-plant screening for tolerance of nutrient-deficiency stress. Plant Soil 99, 3-16. Gillman, G. P., and Bell, L. C. (1978). Soil solution studies on weathered soils from tropical North Queensland. Aust. J. Soil Res. 16,67-77. Good, N. E., and Izawa, S.(1972). Hydrogen ion buffers. Methods Enzymol. 24(Pt. B), 53-68. Good, N. E., Winget, G. D., Winter, W., Connoly, T. N., Izawa, S., and Singh, R. M. M. (1966). Hydrogen ion buffers for biological research. Biochemistry 5,467-477. Gourley, C. J. P., Allan, D. L., Russelle, M. P., and Bloom, P. R. (1993). Evaluation and improvements of a sand-alumina culture technique to screen plants for low phosphorus tolerance. Soil Sci. SOC. Am. J. 57,103-110. Graham, R. D., Welch, R. M., Grunes, D. L., Carey, E. E., and Norvell, W. A. (1987). Effect of zinc deficiency on the accumulation of boron and other mineral nutrients in barley. Soil Sci. SOC.Am. J. 51,652-657. Grattan, S . R., and Maas, E. V. (1984). Interactive effect of salinity and substrate phosphate on soybean. Agron. J. 76,668-676. Gries, D., Brunn, S., Crowley, D. E., and Parker, D. R. (1995). Phytosiderophore production in relation to micronutrient metal deficiencies in barley. Plant Soil 172,299-308.

ADVANCES IN SOLUTION CULTURE METHODS

2 07

Gutschick, V. P., and Kay, L. E. (1991). Nonrecirculating hydroponic system suitable for uptake studies at very low nutrient concentrations. Plant Physiol. 95,1125-1 130. Hageman, R. H., Flesher, D., Wabol, J. J., and Storck, D. H. (1961). An improved nutrient culture technique for growing corn under greenhouse conditions. Agron. J. 53, 175-180. Halvorson,A. D., and Lindsay, W. L. (1972). Equilibrium relationships of metal chelates in hydroponic solutions. Soil Sci. SOC.Am. Proc. 36,755-761. Halvorson, A. D., and Lindsay, W. L. (1977). The critical Zn2+ concentration for corn and the nonabsorption of chelated zinc. Soil Sci. SOC.Am. J. 41,531-534. Harley, C. P., and Lindner, R. C. (1945). Observed responses of apple and pear trees to some irrigation waters of North Central Washington. Proc. Am. SOC. Hort. Sci. 46,35-44. Harper, J. E., and Nicholas, J. C. (1976). Control of nutrient solution pH with an ion exchange system: Effect on soybean nodulation. Physiol. Plant. 38, 24-28. Hellgren, 0..and Ingestad, T. (1996). A comparison between methods used to control nutrient supply. J. Exp. Bof.47, 117-122. Hendrickson, L. L., Turner, M. A., and Corey, R. B. (1982). Use of Chelex-100 to maintain constant metal activity and its application to characterization of metal complexation. Anal. Chem. 54, 1633- 1637. Hershey, D. R. (1991). Iron deficiency stress response of Tolmiea menziesii. J. Plant Nutr. 14, lI45-1l50. Hewitt, E. J. (1966). “Sand and Water Culture Methods Used in the Study of Plant Nutrition,” Rev. 2nd Ed., Tech. Commun. No. 22. Commonwealth Bureau of Horticulture and Plantation Crops, East Malling. Hoaglund, D. R., and Amon, D. I. (1950). The water-culture method for growing plants without soil, California Agric. Expt. Stn. Circ. No. 347. Hommels, C. H., Kuiper, P. I. C., and Telkamp, G. P. (1990). Study on nutrient stress tolerance of nine Taraxmum microspecies with contrasting mineral ecology by cultivation in a low nutrient flowing solution. Physiol. Plan?. 79,389-399. Hopper, J. L. (1997). Factors affecting selenium uptake to toxic concentrations in forages in the central Coast Range of California. M.S. thesis, University of California, Riverside. Howeler, R. H., Edwards, D. G., and Asher, C. J. (1981). Application of the flowing solution culture techniques to studies involving mycorrhizas. Plant Soil 59, 179-1 83. Howeler, R. H., Edwards, D. G., and Asher, C. J. (1982). Micronutrient deficiencies and toxicities of cassava plants grown in nutrient solutions. I. Critical tissue concentrations. J. Plant Nutr. 5, 1059-1076. Howell, R. W., and Bernard, R. L. (1961). Phosphorus response of soybean varieties. Crop Sci. 1, 31 1-313. Hsu, P. H. (1989). Aluminum oxides and hydroxides. In “Minerals in Soil Environments, 2nd Ed.” (J. B. Dixon and S. B. Weed, eds.), Soil Sci. SOC.Am. Book Ser. No. 1, pp. 331-378. Soil Sci. SOC.Am., Madison, W1. Huang, C., Webb, M. J., and Graham, R. D. (1993). Effect of pH on Mn absorption by barley genotypes in a chelate-buffered nutrient solution. P lanr Soil 155/156,437-440. Hurlbert, S. H. ( 1984). Pseudoreplication and the design of ecological field experiments. Ecol. Monogr. 54,187-211. Ingestad, T. (1981). Nutrition and growth of birch and gray alder seedlings in low conductivity solutions and at varied rates of nutrient addition. Physiol. Plunt. 52,454-466. Ingestad, T. (1982). Relative addition rate and external concentration: Driving variables used in plant nutrition research. Plant Cell Environ. 5,443-453. Ingestad, T., and Agren, G . I. (1988). Nutrient uptake and allocation at steady state nutrition. Physiol. Plant. 72,450-459.

208

DAVID R. PARKER AND WENDELL A. NORVELL

Ingestad, T., and Agren. G. I. (1992). Theories and methods on plant nutrition and growth. Physiol. Plant. 84, 177-184. Ingestad, T., and Lund, A.-B. (1979). Nitrogen stress in birch seedlings. I. Growth technique and growth. Physiol. Plant. 45, 137-148. Ingestad, T., and Lund, A.-B. (1986). Theory and techniques for steady state mineral nutrition and growth of plants. Scand. J. Forest Res. 1,439-453. Inskeep, W. P., and Bloom, P. R. (1984). Acomparative study of soil solution chemistry associated with chlorotic and nonchlorotic soybeans in Western Minnesota. J. Plan? Nur,: 7,513-53 1. Ismande, J., and Ralston, E. J. (1981). Hydroponic growth and the nondestructive assay for dinitrogen fixation. Plant Physiol. 68, 1380-1384. Jacobson, L. (1951). Maintaining iron supply in nutrient solutions by a single addition of femc potassium ethylenediamine tetra-acetate. P Ian? P hysiol. 26,4 11-4 13. Jalil, A,, and Carlson, R. M. (1993). Potassium uptake by Marianna plum under limited oxygen and carbon dioxide levels in the root atmosphere. J. Plant Nurr 16,723-737. Jenny, H. (1946). Adsorbed nitrate ions in relation to plant growth. J. Colloid Sci. 1,33-47. Johnson, C. M., Stout, P. R., Broyer, T. C., and Carlton, A. B. (1957). Comparative chlorine requirements of different plant species. Plant Soil 8,337-353. Johnson, G . V., and Youngblood, G. J. (1991). Responses of nitrogen-fixing and nitrate-supplied alfalfa (Medicago sativa L.) to iron chelates in an alkaline hydroponic medium. In “Iron Nutrition and Interactions in Plants” (Y. Chen and Y. Hadar, eds.), pp. 259-264. Kluwer, Dordrecht. Jones, D. L., and Edwards, A. C. (1993). Effect of moisture content and preparation technique on the composition of soil solution obtained by centrifugation. Commun. Soil Sci. Plan? Anal. 24, 171-186. Kerven, G. L., Asher, C . J., and Dethlefs, L. (1993). Release characteristics of ion-exchange resins as buffers in low ionic strength nutrient solutions. In “Plant Nutrition-From Genetic Engineering to Field Practice.” pp. 265-268. Kluwer, Dordrecht. Kim, M.-K., Edwards, D. G., and Asher, C. J. (1985). Tolerance of Trifolium subterraneum cultivars to low pH. Aust. J. Agric. Res. 36,569-578. Kovar, J. L., and Barbar, S. A. (1988). Phosphorus supply characteristics of 33 soils as influenced by seven rates of phosphorus addition. Soil Sci. Sue. Am. J. 52, 160-165. Kovar, J. L., and Barber, S. A. (1990). Potassium supply characteristics of thirty-three soils as influenced by seven rates of phosphorus. Soil Sci. SOC.Am. J. 54,1356-1361. Lawrence, G. B., and David, M. B. (1996). Chemical evaluation of soil-solution in acid forest soils. Soil Sci. 161,298-3 13. Lindsay, W. L. (1979). “Chemical Equilibria in Soils.” Wiley, New York. Lindsay, W. L. (1991). Inorganic equilibria affecting micronutrients in soils. In “Micronutrients in Agriculture” ( J . J. Mortvedt, F. R. Cox, L. M. Shuman, and R. M. Welch, eds.), Soil Sci. SOC.Am. Book Ser. No. 4,2nd Ed., pp. 89-112. Soil Sci. SOC.Am., Madison, WI. Lindsay, W. L., Hodgson, J. F., and Norvell, W. A. (1966). The physico-chemical equilibrium of metal chelates in soils and their influence on the availability of micronutrient cations. In?. Soc. Soil Sci. Trans. 214,315-316. Loneragan, J. F.,and Webb, M. J. (1993). Interactions between zinc and other nutrients affecting the growth of plants. In “Zinc in Soils and Plants” (A. D. Robson, ed.), pp. 119-134. Kluwer, Dordrecht. Loneragan, J. F., Snowball, K., and Simmons, W. J. (1968). Response of plants to calcium concentration in solution culture. Aust. J. Agric. Res. 19,845-857. Loneragan, J. F., Grove, T. S., Robson, A. D., and Snowball, K. (1979). Phosphorus toxicity as a factor in zinc-phosphorus interaction in plants. Soil Sci. Sue. Am. J. 43,966-972. Loneragan, J. F.,Grunes, D. L., Welch, R. M., Aduayi, E. A,, Tengh, A,, Lazar, V. A,, and Cary, E. E.

ADVANCES IN SOLUTION CULTURE METHODS

2 09

(1982). Phosphorus accumulation and toxicity in leaves in relation to zinc supply. Soil Sci. Soc. Am. J. 46,345-352. Lynch, J., Epstein, E., Lauchli, A., and Weigt, G. I. (1990). An automated greenhouse sand culture system for studies of P nutrition. Planr Cell Environ. 13,547-554. Macduff, J. H., and Dhanoa, M. S. (1990). N, fixation and nitrate uptake by white clover swards in response to root temperature in flowing solution culture. Ann. Bot. 65,325-335. Macduff, J. H., and Jackson, S. B. (1992). Influx and efflux of nitrate and ammonium in Italian ryegrass and white clover roots: Comparisons between effects of darkness and defoliation. J. Exp. Bot. 43,525-535. Macduff, J . H., Hopper, M. J., and Wild, A. (1987). The effect of root temperature on growth and uptake of ammonium and nitrate by Brassica napus L. in flowing solution culture. J. Exp. Bor. 38, 42-52. Macduff, J. H., Gordon, A. J.. Ryle, G. J. A., and Powell, C. E. (1989). White clover N,-fixation in response to root temperature and nitrate. J. Exp. Bor. 40,517-526. Macduff, J. H., Jarvis, S. C., Larsson, C.-M., and Oscarson, P. (1993). Plant growth in relation to the supply of NO,-: A comparison between relative addition rate and external concentration as driving variables. J. Exp. Bor. 44, 1475-1484. Macduff, J. H., Jarvis, S. C., and Cockburn, J. E. ( I 994). Acclimation of NO,- fluxes to low root temperature by Brassica napus in relation to NO,- supply. J. Exp. Bot. 45, 1045-1056. Macduff, J. H., Jarvis, S. C., and Davidson, I. A. (1996). Inhibition of N, fixation by white clover (Trifolium repens L.) at low concentrations of NO,- in flowing solution culture. Plant Soil 80, 287-295. Malik, M. N. A., Evenson, J . P.,and Edwards, D. G. (1978). The effect of level of nitrogen nutrition on earliness in upland cotton (Gossypium hirsutum L.). Aust. J. Agric. Res. 29, 1213-1221. Marschner, H. (1995). “Mineral Nutrition of Higher Plants,” 2nd ed. Academic Press, San Diego. Marschner, H., and Romheld, V. (1994). Strategies of pli)nts for acqu on of iron. Plant Soil 165, 261-274. Martell, A. E., and Smith, R. M. (1976-1989). “Critical Stability Constants,” 6 vols. Plenum, New York. Mattsson, M., Johansson, E., Lundborg, T., Larsson, M., and Larsson, C.-M. (1991). Nitrogen utilization in N-limited barley during vegetative and generative growth. I. Growth and nitrate uptake kinetics in vegetative cultures grown at different relative addition rates of nitrate-N. J. Exp. Bot. 42, 197-205. McKenna, I. M., Chaney, R. L., and Williams, F. M. (1993). The effects of cadmium and zinc interactions on the accumulation and tissue distribution of zinc and cadmium in lettuce spinach. Environ. Pollur. 79, 113 - 120. McKenzie, D. B., Hossner, L. R., and Newton, R. J. (1984). Sorghum cultivar evaluation for iron chlorosis resistance by visual scores. J. Plant Nutr: 7,677-685. Menzies, N. W., and Bell, L. C. (1988). Evaluation of the influence of sample preparation and extraction on soil solution composition. Ausr. J. Soil Res. 26,45 1-464. Miyasaka, S. C., Checkai, R. T., Grunes, D. L., and Norvell, W. A. (1988). Methods for controlling pH in hydroponic culture of winter wheat forage. Agron. J. 80,213-220. Mullins, G. L., and Sommers, L. E. (1986). Cadmium and zinc influx characteristics by intact corn (&a mays L.) seedlings. Plant Soil96, 153-164. National Institute of Standards and Technology (NIST) (1993). “NIST Critical Stability Constants of Metal Complexes, Ver. 1 .O,”NIST Standard References Database 46. NIST, Gaithersburg, MD. Nicholas, J. C., and Harper, J. E. (1993). Effect of MES [2(N-morpholino)ethanesulfonic acid] and Amberlite IRC-50 resin on nutrient pH control and soybean growth. J. Plant Nutr: 16,895-909. Nishimoto, R. K., Fox, R. L., and Parvin, P. E. (1977). Response of vegetable crops to phosphorus concentrations in soil solution. J. Am. SOC.Hort. Sci. 102,705-709.

210

DAVID R. PARKER AND WENDELL A. NORVELL

Norvell, W. A. (1991). Reactions of metal chelates in soils and nutrient solutions. In “Micronutrients in Agriculture” (J. J. Mortvedt, E R. Cox, L. M. Shuman, and R. M. Welch, eds.), Soil Sci. SOC. Am. Book Ser. No. 4,2nd Ed., pp. 187-227. Soil Sci. SOC.Am., Madison, WI. Norvell, W. A., and Welch, R. M. (1993). Growth and nutrient uptake by barley (Hordeum vulgare L. cv Herta): Studies using an N-(2-hydroxyethyl)ethylene-dinitrilotriaceticacid-buffered nutrient solution technique. I. Zinc ion requirements. Plant Physiol. 101,619-625. Norvell, W. A., Adams, M. L., Dufour, A. P., and Bassuk, N. L. (1994). Screening soybean and oak seedlings for resistance to iron chlorosis in hydroponic solutions containing magnesium or sodium bicarbonate. Agron. Abstx. 324. Norvell, W. A., Hart, W. J., Welch, R. M., and Kochian, L. V. (1995). Effect of zinc deficiency on cadmium accumulation by young wheat seedlings. Agron. Abstx, 263. Oscarson, P., and Larsson, C.-M. (1986). Relations between uptake and utilization of NO,- in Pisum growing exponentially under nitrogen limitation. Physiol. Plant. 67, 109-1 17. Oscarson, P., Ingemarsson, B., and Larsson, C.-M. (1989). Growth and nitrate uptake properties of plants grown at different relative rates of nitrogen supply. I. Growth of Pisum and Lemna in relation to nitrogen. Plant Cell Environ. 12,779-785. Parker, D. R. (1993). Novel nutrient solutions for zinc nutrition research: Buffering free zinc2+ with synthetic chelators and P with hydroxyapatite. Plant Soil 155/156,461-464. Parker, D. R. (1997). Responses of six crop species to solution zinc2+ activities buffered with HEDTA. Soil Sci. SOC.Am. J. 61, 167-176. Parker, D. R., Zelazny, L. W.. and Kinraide, T. B. (1987). Improvements to the program GEOCHEM. Soil Sci. Soc. Am. J. 51,488-491. Parker, D. R., Aguilera, J. J., and Thomason, D. N. (1992). Zinc-phosphorus interactions in two cultivars of tomato (Lycopersicum esculentum L.) grown in chelator-buffered nutrient solutions. Plant Soil 143, 163-177. Parker, D. R., Chaney, R. L., and Norvell, W. A. (1995a). Chemical equilibrium models: Applications to plant nutrition research. In “Chemical Equilibrium and Reaction Models” (R. H. Loeppert et al., ed.), SSSASpec. Publ. No. 42, pp. 163-200. ASAISSSA, Madison, WI. Parker, D. R.. Norvell, W. A,, and Chaney, R. L. (1995b). GEOCHEM-PC: A chemical speciation program for IBM and compatible personal computers. In “Chemical Equilibrium and Reaction Models” (R. H. Loeppert et al., ed.),SSSA Spec. Publ. No. 42, pp. 253-269. ASAISSSA, Madison, WI. Parker, D. R., Tice, K. R., and Thomason, D. N. (1997). Effects of ion pairing with calcium and magnesium on selenate uptake by plants. Environ. Toxicol. Chem., 16,565-571. Pearson, J. N., and Rengel, Z. (1995). Uptake and distribution of 65Zn and 54Mnin wheat grown at sufficient and deficient levels of Zn and Mn. I. During vegetative growth. J. Exp. Bot. 46,833-839. Pereira, P. A. A., and Bliss, F. A. (1987). Nitrogen fixation and plant growth of common bean (Phaseolus vulgaris L.) at different levels of phosphorus availability. Plant Soil 104,79-84. Polisetty, R., and Hageman, R. H. (1985). Effect of media pH on nitrate uptake, dry matter production and nitrogen accumulation by corn (&a mays L.) seedlings grown in solution culture. Biol. Plantarum 27,451-457. Porter, L. K., and Thorne, D. W. (1955). Interrelation of carbon dioxide and bicarbonate ions in causing plant chlorosis. Soil Sci. 79,373-382. Proctor, J. (1970). Magnesium as a toxic element. Narure (London) 227,742-743. Qian, P., and Wok, J. D. (1990). Effects of drying and time of incubation on the composition of displaced soil solution. Soil Sci. 149,367-374. Reisenauer, H. M. (1966). Mineral nutrients in soil solution. In “Environmental Biology” (P. L. Altman and D. S. Dittmer, eds.), pp. 507-508. Fed. Am. SOC.Exp. Biol., Bethesda, MD. Rengel, Z., and Graham, R. D. (1995). Wheat genotypes differ in Zn efficiency when grown in chelatebuffered nutrient solution. I. Growth. Plant Soil 176,307-3 16.

ADVANCES IN SOLUTION CULTURE METHODS

211

Rengel, Z., and Graham, R. D. (1996). Uptake of zinc from chelate-buffered nutrient solutions by wheat genotypes differing in Zn efficiency. J. Exp. Bor. 47,2 17-226. Resh, H. M. (1989). “Hydroponic Food Production: A Definitive Guidebook of Soilless Food Growing Methods,” 4th ed. Woodbridge Press, Santa Barbara, CA. Righetti, P. G., Chiari, M., and Crippa, L. (1991). Macroreticulate buffers: A novel approach to pH control in living systems. J. Biotechnol. 17, 169-176. Robson, A. D., and Loneragan, J. F. (1970). Sensitivity of annual Medicago species to manganese toxicity as affected by calcium and pH. Aust. J. Agric. Res. 21, 223-232. Ross, D. S., and Bartlett, R. J. (1990). Effects of extraction methods and sample storage on properties of solutions obtained from forested spodosols. J. Environ. Qual. 19, 108-1 13. Rys, G. J., and Phung, T. (1985). Nutrient solution pH control using dipolar buffers in studies of Trifolium repens L. nitrogen nutrition. J. Exp. Bot. 36,426-432. Schlenker, F. S. (1940). Plant growth in culture solution and availability of ions adsorbed in permutit and aniline black. Am. J. Bot. 27,525-529. Schwab, A. P., and Lindsay, W. L. (1989a). A computer simulation of Fe(II1) and Fe(I1) complexation in limited nutrient solution: I. Program development and testing. Soil Sci. Soc. Am. J. 53,29-34. Schwab, A. P., and Lindsay, W. L. (1989b). Acomputer simulation of Fe(II1) and Fe(I1) complexation in nutrient solutions: 11. Applications to growing plants. Soil Sci. Soc. Am. J. 53,34-38. Shi, Y.,Byrne, D. H., Reed, D. W., and Loeppert, R. H.(1993). Iron chlorosis development and growth response of peach rootstocks to bicarbonate. J. Plant Nutr: 16, 1039- 1046. Skogley, E. O., and Dawson, J. E. (1963). Synthetic ion-exchange resins as a medium for plant growth. Nature 198, 1328-1329. Smolders, E., and McLaughlin, M. J. (1996a). Chloride increases cadmium uptake in Swiss chard in a resin-buffered nutrient solution. Soil Sci. Soc. Am. J. 60, 143-1447. Smolders, E., and McLaughlin, M. J. (1996b). Effect of CI on Cd uptake by Swiss chard in nutrient solutions. Plant Soil 1 7 9 , 5 7 4 4 . Sorrell, B. K., and Armstrong, W. (1994). On the difficulties of measuring oxygen release by root systems of wetland plants. J. Ecol. 82, 177-183. Spear, S. N., Asher, C. J., and Edwards, D. G. (1978). Response of cassava, sunflower,and maize to potassium concentration in solution. I. Growth and potassium concentration. Field Crops Res. 1,347-361. Stadt, K. J., Taylor, G. J., and Dale, M. R. T.(1992). Control of relative growth rate by application of the relative addition rate technique to a traditional solution culture system. Plunr Soil 142,113-1 33. Stahl, R. S., Greenhalgh, M. R., Grossi, P. R., and Bugbee, B. G. (1996). Is the buffer 2-[N-morpholino]-ethanesulfonic acid (MES) phytotoxic? Agron. Abstr:, 245. Stevens, W. B., Jolley, V. D., and Hansen, N. C. (1993). Modified procedures for commercial adaptation of root iron-reducing capacity as a screening technique. J. Plant Nutr: 16,2507-2519. Stumm, W., and Morgan, J. J. (1996). “Aquatic Chemistry,’’ 3rd ed. Wiley, New York. Suarez, D. L., Wood, 3. D., and Ibrahim, I. (1992). Reevaluation of calcite supersaturation in soils. Soil Sci. Soc. Am. J. 56, 1776-1784. Svenning, M. M., and Macduff, J. H. (1996). Low root temperature retardation of the mineral nitrogen induced decline in N, fixation by a northern ecotype of white clover. Ann. Bot. 77,615-621, Svenning, M. M., Junttila, O., and Macduff, J. H. (1996). Differential rates of inhibition of N, fixation by sustained low concentrations of NH4+ and NO,- in northern ecotypes of white clover (Trifolium repens L.), J. Exp. Bot. 47,729-738. Swietlik, D., and Zhang, L. (1994). Critical zincfz activities for sour orange determined with chelator-buffered nutrient solutions. J. Am. Soc. Hour. Sci. 119,693-701. Tang, C., Adams, H., Longnecker, N. E., and Robson, A. D. (1996). A method to identify lupin species tolerant of alkaline soils. Aust. J. Exp. Agric. 36, 595-601. Temple-Smith, M. G., and Menary, R. C. (1977). Growth and phosphate absorption in lettuce and cabbage plants in dilute nutrient solution culture. Aust. J. Plant Physiol. 4,505-513.

2 I2

DAVID R. PARKER AND WENDELL A. NORVELL

Treeby, M., Marschner, H., and Romheld, V. (1989).Mobilization of iron and other micronutrient cations from a calcareous soil by plant-borne, microbial, and synthetic metal chelators. Plant Soil 114,217-226. Viti, R., and Cinelli, F. (1993).Line-induced chlorosis in quince rootstocks: Methodological and physiological aspects. J. Plant Nut[ 16,631-64 1. von Wirbn, N., Marschner, H., and Romheld, V. (1996).Roots of iron-efficient maize also absorb phytosiderophore-chelated zinc. Plant Physiol. 111, 1119-1 125. Wadleigh, C. H., and Brown, J. W. (1952).The chemical status of bean plants afflicted with bicarbonate-induced chlorosis. Boran. Gazette 113,373-392. Wainwright, S. J., and Woolhouse, H. W. (1977).Some physiological aspects of copper and zinc tolerance in Agmstis tenuis Sibth: Cell elongation and membrane damage. J. Exp. Bot. 28,1029-1036. Wallace, A. (1962).“A Decade of Synthetic Chelating Agents in Inorganic Plant Nutrition.” Los Angeles. Wallace, A,, Mueller, R. T., and Alexander, G. V. (1978).Influence of phosphorus on zinc, iron, manganese, and copper uptake by plants. Soil Sci. 126,336-341. Wallace, A,, Wallace, G. A,, and Alexander, G. V. (1983).Effect of excess chelating agent in nutrient solution at low levels of iron, zinc, copper, and manganese. J. Plant Nurr: 6,507-5 11. Walter, A., Romheld, V., Marschner, H., and Mori, S . (1994).Is the response of phytosiderophores in zinc-deficient wheat plants a response to impaired iron utilization? Physiol. Plant. 92,493-500. Walworth, J. L. (1992).Soil drying and rewetting, or freezing and thawing, affects soil solution composition. Soil Sci. SOC.Am. J. 56,433-437. Webb, M.J. (1993).A multichannel pH controller for solution culture systems. Plant Soil 155/156, 50 1-504. Webb, M. J., and Loneragan, J. F. (1988).Effect of zinc deficiency on growth, phosphorus concentration, and phosphorus toxicity of wheat plants. Soil Sci. SOC.Am. J. 52, 1676-1680. Webb, M. J., Norvell, W. A., Welch, R. M., and Graham,R. D. (1993).Using a chelate-buffered nutrient solution to establish the critical solution activity on Mn2+ required by barley (Hordeum vulgare L.). Plant Soil 153,195-205. Wehr. J. D., Brown, L. M., and Vanderelst, I. E. (1986).Hydrogen ion buffering of culture media for algae from moderately acidic, oligotrophic waters. J. Phycol. 22,88-94. Welch, H.V., Jr., Wallace, A,, and Mueller, R. T. (1954).Influence of factorially combined levels of cations and nitrate ions adsorbed on ion-exchange resins on the nutrient absorption by plants. Soil Sci. SOC.Am. Pmc. 18,137-140. Welch, R. M. (1995).Micronutrient nutrition of plants. CRC Crit. Rev. Plant Sci. 14,49-82. Welch, R. M., and Norvell, W.A. (1993).Growth and nutrient uptake by barley (Hordeum vulgare L. cv Herta): Studies using an N-(2-hydroxyethyl)ethylene-dinitrilotriaceticacid-buffered nutrient solution technique. I. Role of zinc in the uptake and root leakage of mineral nutrients. Plant Physiol. 101,627-631. Wild, A,, Skarlou, V., Clement, C. R., and Snaydon, R. W. (1974).Comparison of potassium uptake by four plant species grown in sand and in flowing solution culture. J. Appl. Ecol. 11,801-812. Wild, A., Jones, L. H. P., and McDuff, J. H. (1987).Uptake of mineral nutrients and crop growth: The use of flowing nutrient solutions. Adv. Agron. 41, 171-219. Williams, D. E., and Vlamis, J. (1957a).Manganese and boron toxicities in standard culture solutions. Soil Sci. SOC.Am. P roc. 21,205-209. Williams, D. E., and Vlamis, J. (1957b).Manganese toxicity in standard culture solutions. Plant Soil 21,205-209. Williams, R. F. (1946).The physiology of plant growth with special reference to the concept of net assimilation rate. Ann. Bot. 10,41-72. Wilson, D. 0.. and Reisenauer, H. M. (1963). Cobalt requirements of symbiotically grown alfalfa. Plant Soil 19,364-373.

ADVANCES IN SOLUTION CULTURE METHODS

213

Wolt, J., and Graveel, J . G. (1986). A rapid routine method for obtaining soil solution using vacuum displacement. Soil Sci. SOC.Am. J. 50,602-605. Woodhouse, P. J., Wild, A., and Clement, C. R. (1978). Rate of uptake of potassium by three crop species in relation to growth. J. Exp. Bor. 29,885-894. Wright, R. J., Baligar, V. C., and Wright, S. F. (1987). Estimation of phytotoxic aluminum in soil solution using three spectrophotometric methods. Soil Sci. 144,224-232. Wu, C.-L. (1996). Adaptations of barley to iron deficiency. M. S. thesis, University of California, Riverside. Yang, X., Romheld, V., Marschner, H., and Chaney, R. L. (1994). Application of chelator-buffered nutrient solution technique in studies on zinc nutrition in rice plant (Oryzn sntiva L.). Plant Soil 163, 85 -94.

This Page Intentionally Left Blank

RADMTIONUSEEFFICIENCY Thomas R. Sinclair' and Russell C. Muchow2 'USDA-ARS Agronomy Physiology & Genetics Laboratory University of Florida Gainesville, Florida 3261 1-0965 *CSIRO, Tropical Agriculture Cunningham Laboratory Brisbane, Qld. 4067, Australia

I. Introduction A. Time-Based Growth Analysis B. Light-Based Growth Analysis C. Terminology 11. Theoretical Analyses of RUE A. Initial Analysis of Crop Productivity B. Leaf Photosynthetic Rates and RUE C. Radiation Environment and RUE D. Conclusions from Theoretical Analyses 111. Experimental Determination of RUE A. Determination of Crop Mass B. Determination of Solar Energy C. Calculation of RUE IV Experimental Measures of RUE A. C+Species B. C , Species V. Sources of Variability in RUE A. Species B. CO, Assimilation Rate C. Seasonal Variation D. Radiation Environment VI. Conclusions References

I. INTRODUCTION Light levels have a profound influence on plant growth, as readily observed in plants in the shady areas of a garden compared with plants in areas of full sunlight. Somewhat surprisingly then, only in recent times has considerable attention been 215 Adwmmr in Armnomy, Vuhme 65

Copyright 0 IW9 by Academic Press. tUI rights of reproduction in any form reserved 0065-2113/90 $30.00

2 16

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

given to investigating the quantitative relationship between crop growth and light levels. The influence of light levels on crop photosynthesis and mass accumulation was not considered explicitly until the late 1950s (DeWit, 1959). Prior to this time, the focus on classical crop growth analysis seems to have inhibited a more mechanistic appraisal of crop mass accumulation.

A. TIME-BASEDGROWTH ANALYSIS Early investigators sought to analyze crop growth as a function of time (Blackman, 1919), although it was clear that growth is not dependent on time at all. Consequently, the accepted and popular approach of classical crop growth analysis was to describe growth by the change in crop mass (m) as a function of time (f). Crop growth rate (dm/df) could be readily estimated from successive harvests through the growing season. Crop growth rates were calculated and compared for various crops and locations in spite of the fact that these results were confounded by varying amounts of light intercepted by the growing crop as a result of differences in incident light or changing crop leaf area. The misleading approach of expressing crop growth as a function of time was muddled further in crop growth analysis by calculating the derived variables of relative growth rate and net assimilation rate. Relative growth rate (RGR), as defined by Watson (1952) in the following equation, was a more generalized expression of the efficiency index proposed by Blackman (1919): RGR = l/m dm/dt. The RGR equation required the additional assumption that crop mass increased as an inverse function of current crop mass. This type of expression is used to describe, for example, the growth of microbes in an environment of unlimited resources. Such an assumption in the analysis of crop growth requires, of course, that there is a constant relationship between plant mass and leaf area. Additionally, it is implicit that there is no self-shading among leaves. Clearly, the RGR assumption can only be approximated during the early stages of plant development. Blackman (1919) recognized for individual plants that RGR declines as plant mass increases. According to Watson (1952), in 1917 Gregory, recognizing light interception as critical in plant growth, developed net assimilation rate (NAR) as an approach to incorporate leaf area (L) explicitly in crop growth analysis: NAR

=

l/L dm/dt.

In addition to the problem of assuming that growth is a function of time, this equation failed to consider the uneven pattern of light interception by leaves in a crop canopy. The NAR calculation assumed a uniform light distribution over leaves.

RADHTION USE EFFICIENCY

2 17

This assumption grows worse as shading in the leaf canopy increases. In spite of the potential for instability resulting from the assumption of uniform light distribution, fairly stable values of NAR were obtained in much of the early research using this approach (Watson, 1952). The basis for this stability is likely traced to the fact that much of the data was from situations where leaf area index was low for much of the season and the amount of shading was not great. In modern crops with high plant densities and rapid development of leaf area, there is a high degree of leaf shading. Calculations of NAR do not offer much mechanistic insight about the development of crop mass and yield.

B. LIGHT-BASEDGROWTH ANALYSIS Some of the first, cogent analyses of increases in crop mass in response to the amount of light available to a crop were done by DeWit (1959) and by Loomis and Williams (1963). DeWit (1959) assumed a constant efficiency of light use at the leaf level up to a saturating light value and examined the consequences of light distribution in a leaf canopy. Canopy photosynthetic capability was calculated to be proportional to the amount of solar radiation incident to the canopy. Loomis and Williams (1963) advanced this mechanistic perspective by considering the quantum nature of light and by attempting to express the efficiency in the use of light in terms of accumulated plant mass. Assuming a maximum value for leaf quantum efficiency of 10 quanta per CO, fixed, they calculated that the limit for crop efficiency was 3.34 g CH,O MJ-' solar radiation. Early experimental data also confirmed a close linkage between the amount of light intercepted by a crop and its growth. Shibles and Weber (1965) found a linear relationship between dry matter increase and the fraction of radiation intercepted through the entire growing season for soybean. Williams et al. (1965) found for maize that 1.71 g of plant mass was produced for each MJ of intercepted solar radiation. During the 1970s, however, several studies analyzed efficiency on the basis of incident radiation rather than intercepted radiation. These estimates were used by ecologists and agronomists to compare plant productivity under different systems of land use and management and in different climates. For example, Cooper (1970) estimated an efficiency in the use of incident radiation by assuming that 45% of the total incoming radiation is in the visible spectrum and available for photosynthesis, and that the production of 1 g of dry matter corresponds to the fixation of 4250 calories of chemical energy. He calculated that the average efficiency in the use of incident radiation for pasture production of up to 3% was possible over the whole year with up to 5% during the most productive part of the growing season. Cooper also estimated annual dry matter production potential in contrasting climatic regions based on differences in annual energy input. Analysis based on in-

218

THOMAS R. SINCLAIR AND RUSSELL C . MUCHOW

cident radiation was also done by Austin et al. (1978), who found that sugarbeet was less efficient than sngarcane. Jong et al. (1982) showed that grain yield of maize in Hawaii was linearly related to incident radiation and this accounted for 78.5% of the yield variation. The limitation in relating productivity to incoming solar radiation is that only a proportion of the incident radiation is intercepted by plants during their life cycle and available for photosynthesis (Squire, 1990). Differences in leaf canopy development among crops and climates can confound comparison of efficiency based only on incident radiation. In fact, Monteith (1972) commented that classifying ecosystems on the basis of efficiency in relation to incident radiation is becoming a popular form of taxonomy but it contributes little to our understanding of how plants respond to their environment. His analysis went a stage further, relating efficiencies of dry matter production to the physical and biological factors that determine growth rates such as the fraction of radiation intercepted by a leaf canopy, the irradiance of individual leaves, the diffusion resistance of stomata, and the behavior of the photochemical system. Monteith (1972) concluded that the interception efficiency, defined as the ratio of the actual rate of gross photosynthesis to the maximum rate estimated for a stand of identical plants with enough leaves to intercept all the incident light, emerges as a major discriminate of dry matter production. This efficiency on the one hand accounts for differences in productivity under different conditions of climate and management and, on the other, for differences between the mean and maximum rates of production within a particular stand. In 1977 John Monteith published a paper that fully established both experimental and theoretical grounds for the relationship between accumulated crop dry matter and intercepted solar radiation. He concluded from a composite of experimental results obtained under good growth conditions that for most crops approximately 1.4 g of crop mass was accumulated per MJ of intercepted solar radiation. Monteith also presented a theoretical curve for the response of radiation use efficiency (RUE) to changes in maximum rate of leaf photosynthesis. While RUE was highly sensitive to maximum leaf photosynthesis rates at low rates, the sensitivity of RUE to photosynthetic rates was much less in the usual range of leaf photosynthetic rates for nonstressed crops. Monteith’s 1977 paper was particularly important in pointing to RUE as a robust and theoretically appropriate approach for describing crop growth. Subsequent to Monteith’s paper, a number of studies incorporated estimates of radiation interception so that RUE could be calculated. Several papers summarized RUE estimates obtained under a range of conditions (Gallagher and Biscoe, 1978; Gosse et al., 1986; Kiniry et al., 1989). Gallagher and Biscoe (1978) compared the growth of cereals under a range of conditions at two locations in the United Kingdom. Unfortunately, the amount of intercepted radiation was not measured but estimated based on crop leaf area. Nevertheless, RUE was essentially stable and

219

RADIATION USE EFFICIENCY

equivalent for wheat and barley when grown under good conditions. They reported that the RUE value was about 3 g MJ-’ of photosynthetically active radiation (PAR; or equivalently, 1.35 g MJ- intercepted solar radiation as discussed in Section 111). Unfertilized and drought-stressed conditions resulted in a decrease in RUE. In a comparison among various crop species, Gosse et al. (1986) concluded that, in fact, important differences in RUE existed among species. They found that C, species had the highest RUE, followed by nonleguminous species, and leguminous species had the lowest. Variation among species was also reported by Kiniry et al. (1989) in their comparison of five crop species. In their study also, RUE was generally calculated from estimates of light interception based on leaf area. Maize had the highest RUE at 3.5 g MJ-’ of intercepted PAR (or 1.75 g MJ-’ intercepted solar radiation), and sorghum had only 2.8 g MJ- (or 1.4 g MJ- solar radiation) intercepted PAR. The estimates of RUE for sunflower, rice, and wheat were 2.2, 2.2, and 2.8 g MJ- intercepted PAR (or 1.1, 1.1, and 1.4 g MJ- solar radiation), respectively.

’

’

C. TERMINOLOGY “Conversion efficiency” has sometimes been used as the preferred terminology instead of radiation use efficiency (Horie et al., 1997). The concept of solar energy conversion arose from thermodynamic considerations where ecosystems are likened to machines supplied with energy from an external source, that is, solar energy. The available energy input in any environment is determined by the seasonal distribution of solar radiation, and provided that water and nutrients are not limiting, this sets the ultimate limit to productivity (Cooper, 1970; Loomis et al., 1971; Monteith, 1972).Dividing the useful energy of a thermodynamic process by the total energy involved gives a figure for the efficiency of the process and this procedure has been used to analyze the flow of energy in ecosystems, particularly for pastures that are grown primarily as a source of digestible energy and other nutrients for the ruminant. This approach was used with grain crops in evaluating performance in terms of intercepted radiation (Allen and Scott, 1980; Muchow and Coates, 1986; Marshall and Willey, 1983). The terminology of “radiation conversion efficiency,” however, is inappropriate when measuring plant mass accumulated per unit of radiation because it implies that there is a process of direct conversion of radiant energy to mass. This is, of course, an erroneous view because photosynthesis and plant mass accumulation involve no direct conversion of energy to plant mass. At the energy levels of solar radiation, it is not even possible to directly condense radiant energy to atoms such as carbon. The actual photosyntheticprocess involves the absorption of radiant energy by pigments resulting in the establishment of new energy levels in these mol-

220

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

ecules. The higher energy states of these pigments are then used to assimilate carbon dioxide and synthesize plant constituents. Hence, crop mass accumulation relative to light levels is appropriately referred to as radiation use efficiency.

II. THEORETICAL ANWYSES OF RUE Before examining in detail the various experimental measures of RUE, those studies that examined theoretically the nature of RUE will be reviewed. These theoretical studies help to give a background and framework for evaluating potential sources of variation in experimental measures of RUE. Specifically, theoretical studies help to identify variations in the environment and crops that might influence RUE.

A. INITIALANALYSISOF CROPPRODUCTMTY As discussed previously, one of the first attempts to estimate theoretically crop productivity based on solar radiation incident to a crop was published by Loomis and Williams (1963). The basis of their calculation was an assumed quantum efficiency for individual leaves of 10 mol per mole of assimilated CO,. Loomis and Williams calculated for the crop that 14 pg CH,O could be assimilated per calorie of incident solar radiation. Assuming a 0.7 conversion factor from CH,O to plant mass, then the equivalent RUE on a total solar radiation basis was 2.34 g MJ-’. Because of the essentially optimum value assumed for the quantum efficiency, this theoretical estimate offered a maximum limit to RUE. DeWit (1965) expanded substantially on the calculation of Loomis and Williams by considering the geometry of light interception by a leaf canopy. The model of DeWit accounted for variations in solar elevation, leaf angles, and fraction of diffuse radiation. Using a Michaelis-Menten equation to express the response of individual leaves to light level, an approximately linear increase in canopy assimilation was calculated in response to absorbed light by the leaf canopy. In addition, DeWit’s analysis indicated that an increasing diffuse component in the incident light resulted in increased canopy productivity. Calculations of RUE from the results presented by DeWit (DeWit’s Tables 6 and 7) for daily canopy CO, assimilation gives RUE values that are essentially stable through the growing season (i.e., time of year) and over latitude. Increasing maximum leaf photosynthetic rate resulted in increasing values in calculated RUE. Goudriaan (1982) subsequently expanded DeWit’s analysis on the latitude response and showed little variation in RUE over a wide range of latitudes. Duncan et al. (1967) also undertook an analysis of canopy photosynthesis based

RADIATION USE EFFICIENCY

22 1

on the geometry of light interception. Unfortunately no data were presented that allow direct calculations of RUE. They did, however, analyze the importance of leaf angle on daily net photosynthesis. They compared canopies with extreme, uniform leaf angles of 0 and 80" as compared to the more usual angle of 40". Lower leaf angles were advantageous below a canopy LAI of 3 to 4, and more erect leaf angles were advantageous above this LAI. Overall, there was little difference between canopies of differing leaf angles when LA1 was less than 6. Duncan (197 1) extended this analysis by considering canopies with leaf layers that were either completely horizontal or vertical. An advantage in daily photosynthesis resulting from a hypothetical mixture of leaves with vertical and horizontal leaves occurred only when LAI was greater than 4. For most crop species, these analyses indicated that RUE would not be sensitive to leaf angle even with extreme leaf angles. Overall, these early analyses gave important suggestions about the relative importance of individual variables on RUE. A number of variables were indicated to have only a small influence on RUE, including solar altitude, latitude, time of year, and leaf angle. Quantum efficiency and maximum leaf photosynthetic rate appeared to have the potential to alter RUE values substantially.

B. LEAFPHOTOSYNTHETIC RATESAND RUE As mentioned in the Introduction, Monteith (1 977) presented the first theoretical analysis leading explicitly to predictions of RUE. Although the details of his model were not presented, the input conditions included a leaf quantum efficiency of 10 mol PAR per mole CO, fixed, a photorespiration rate equivalent to 0.3 of photosynthesis rate, and a dark respiration rate equal to 0.4 of photosynthesis. Essentially a linear relationship was calculated between dry matter accumulation and intercepted solar radiation (Monteith's Fig. 3) for any given maximum leaf photosynthesis rate. A curvilinear response of RUE to maximum leaf photosynthesis rate, as anticipated in the analysis of DeWit (1965), was predicted (Monteith's Fig. 4). RUE increased from 0 at 0 g (CH,O) m-* h- leaf photosynthetic rate to about 1.6 g MJ-' at 5 (CH,O) mp2 h p ' maximum leaf photosynthetic rate. At a leaf photosynthetic rate of 3 g (CH,O) mP2h- I , RUE was calculated to be 1.4 g MJ-', which was the value Monteith (1977) concluded was generally representative of C, crops. Murata (1981) calculated RUE using a single equation for calculating daily gross canopy CO, assimilation. RUE was calculated based on differing leaf photosynthetic rates of various crop species and varying levels of solar radiation. In a comparison of species, higher rates of leaf photosynthesis for C, species resulted in higher calculated values of RUE (average of 2.48 g mass MJ- solar radiation) than for C, species (average 1.82 g MJ-I). Murata's estimates of RUE, however, seem somewhat inflated because he assumed a very high conversion coefficient

222

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

from CO, to plant dry matter (0.61). A more realistic conversion decreases the RUE values to about three-fourths of his original estimates. Considering the conversion of CO, to CH,O is 0.68 (30/44) and that the conversion of CH,O to plant material high in carbohydrate is roughly 0.7 (Sinclair and DeWit, 1975), a more realistic coefficient is approximately 0.48 (= 0.68 X 0.7). Consequently, more realistic estimates of RUE using Murata's results are seemingly 1.95 g MJ- for C, species and 1.43 g MJ-' for C, species. Sinclair and Horie (1989) used a model of canopy photosynthesis that calculated separately the CO, assimilation of leaves exposed to direct beam radiation and those in the shade in the canopy. Leaf photosynthetic rates were calculated using an asymptotic exponential equation that was defined by a quantum efficiency of 5 g CO, MJ- solar radiation, and by the light-saturated CO, assimilation rate. The main intent of this model was to examine the influence of light-saturated photosynthetic rates and leaf nitrogen contents on RUE. Results of calculations from this model were similar to previous analyses in that RUE was closely linked to leaf carbon exchange rate (Fig. 1). In addition to variations among species as a result of differences in leaf photosynthetic rates, variations in the energy content of the plant products also were calculated to alter RUE. For example, soybean with plant products high in proteins and lipids was calculated to have a lower RUE at equivalent leaf photosynthetic rates than cereal crops (Fig. 1). The saturating nature of the RUE response to leaf photosynthetic rate at high leaf photosynthetic rates was concluded to be important in explaining potential sta-

'

'

CER (mg C02 m-2d) Figure 1 Calculated RUE as a function of light-saturated leaf photosynthetic rate (Sinclair and Hone, 1989). The cereals, maize and rice, were plotted separately from soybean to account for differences in the energy content of the plant mass.

223

RADIATION USE EFFICIENCY

bility in RUE. Large variations in leaf photosynthetic rate at the high values were required to result in substantial variations in RUE (Fig. 1). Any factor, however, that decreased leaf photosynthesis to the range of low photosynthetic rates had a direct consequence in lowering RUE. Therefore, low photosynthetic rates resulting from various stresses were predicted to result in decreased RUE. For example, Sands (1996) calculated that temperature might influence RUE depending on leaf photosynthetic response to temperature. The sensitivity of leaf photosynthetic rates to changes in leaf nitrogen was explored by Sinclair and Hone (1989) as a potentially important source of variation in RUE. In their model, maximum leaf CO, assimilation rate was argued to be a direct function of leaf nitrogen content per unit leaf area. They assumed that the leaf nitrogen content was uniform throughout the leaf canopy. From this model, Sinclair and Horie (1989) calculated substantial differences among crop species in the relationship between RUE and leaf nitrogen content (Fig. 2). At high leaf nitrogen contents, the calculated RUE response curves were saturated and there were only small changes in RUE with changes in leaf nitrogen. The maximum values of RUE from this analysis for each species were about 1.8 g MJ-' for maize, 1.5 MJ-' for rice, and 1.3 g MJ-' for soybean. At more realistic levels of leaf nitrogen, the RUE estimates were slightly less than these maximal values. For situations where leaf nitrogen content decreased to lower levels, the estimates of RUE became very sensitive to these changes in nitrogen content (Fig. 2). Certainly, under conditions where leaf nitrogen content is fairly low, changes in

"

00.2

0.6

1.0

1.4 1.8 2.2

2.6

3.0

Leaf N (g m-') Figure 2 Calculated RUE as a function of the mean leaf N per unit area for a crop canopy (Sinclair and Horie, 1989).The differences in crop species result from differences in the energy content of the plant mass and the relationshipbetween leaf photosynthetic rate and leaf N content.

224

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

these nitrogen contents were predicted to have the possibility of resulting in large changes in RUE. Both Hammer and Wright (1 994) and Sands (1996) calculated a similar sensitivity of RUE to leaf nitrogen content. Sinclair and Shiraiwa (1993) expanded the investigation of the sensitivity of RUE to leaf nitrogen content by investigating the sensitivity of RUE to a nonuniform distribution of nitrogen in the leaf canopy. Commonly, the leaves at the top of the canopy, which are exposed to higher light levels, tend to have the highest leaf nitrogen contents. The calculations of Sinclair and Shiraiwa (1993) showed important increases in RUE as a result of the nonuniform distribution of nitrogen, particularly at low mean leaf nitrogen levels. The analysis of Hammer and Wright (1994) showed a similar improvement in RUE from a nonuniform leaf nitrogen content at low average leaf nitrogen levels. In summary, theoretical analyses have consistently indicated a dependence of RUE on leaf photosynthetic activity. The example presented in Fig. 1 is illustrative of most of the calculated relationships between RUE and maximum leaf photosynthesis rates. Increasing leaf photosynthesis at high rates results in only small increases in RUE. For C, species, maximum RUE was generally calculated to be in the range 1.4 to 1.5 g MJ- intercepted solar radiation. Depressed leaf photosynthesis rates, however, were calculated to result in large decreases in RUE.

'

C. RADIATIONENVIRONMENT AND RUE The original calculations of DeWit (1965) indicated that RUE might be sensitive to the fraction of the diffuse component in the incident radiation but stable under differing latitudes. Murata (1981) included as a variable in his calculations of RUE the levels of total daily radiation. Decreasing solar radiation resulted in increased RUE, particularly at daily radiation levels that are less than half of that on a bright day. Hammer and Wright (1994) provided a detailed analysis of the importance of the radiation environment on RUE. In their model they simulated changes in radiation level by altering the atmospheric transmission ratio. They concluded that the atmospheric transmission ratio had the greatest effect on RUE of any abiotic variable they examined. RUE increased by about 0.4 g MJ-' when going from a clear day to a very cloudy day. In a less sophisticated analysis, Sinclair et al. (1992) had previously calculated about the same increase in RUE when the amount of diffuse light was held constant and the total amount of incident light was increased by increasing only the direct component. In segregating the influence of the change in proportion of diffuse radiation, Hammer and Wright (1994) concluded that the increasing fraction of diffuse component on the cloudy day accounted for a RUE increase of 0.15 g MJ-I.

RADIATION USE EFFICIENCY

225

Norman and Arkebauer (199 1) calculated the response of RUE on an hourly basis in response to the changing radiation conditions. When they plotted the calculated RUE values against the diffuse:direct radiation ratio of the incident radiation, they obtained nearly a linear increase in RUE as the diffuse:direct ratio increased. These results were confounded, however, by decreases in total solar radiation associated with an increasing diffuse component. Analyses of the influence of solar elevation, and therefore latitude, on RUE have indicated no substantial influence, consistent with the original analysis of DeWit (1965). The lack of sensitivity of RUE to latitude was noted in the RUE calculations of Sinclair and Horie (1989) and Hammer and Wright (1994). Finally, the sensitivity of RUE to varying LA1 has been examined theoretically. Horie and Sakuratani (1985) showed little response in calculated RUE for rice as a result of varying LAI. Similarly, Sinclair and Horie (1989) found that RUE was stable over a range of LAI, except at LA1 less than 1.O. At the low LAI, RUE estimates were calculated to decrease. Sands (1996) also reported a lack of sensitivity in RUE to varying LAI. Overall, the radiation environment has been calculated to be important in determining RUE but the effects are fairly small. Higher RUE values were estimated for low-radiation, high-diffuse component conditions than for high-radiation, low-diffuse component conditions. These differences in RUE are evident during the diurnal cycle and might be of some significance when comparing RUE across environments.

D. CONCLUSIONS FROM THEORETICAL ANALYSES Overall, the various studies of RUE have been consistent in the analysis of RUE sensitivity to various variables. Clearly, maximum leaf photosynthetic capability is the main theoretical variable that influences RUE. Important differences among crop species and locations may well result from differences in leaf photosynthetic activity. Theoretical analyses of RUE also indicated that stress resulting in decreased leaf photosynthetic rate would result in major decreases in RUE. Consequently, in reviewing and comparing experimental results, it is likely that the more meaningful comparisons among experiments will be those where the crops were not subjected to stresses. Among the remaining factors that potentially influence RUE, the level of radiation was calculated to have an important, albeit a fairly modest influence on RUE. Potential sources of variation in RUE among locations indicated by the theoretical analyses are the level of solar radiation and the fraction of diffuse radiation. Other factors such as latitude and LA1 appear to have only a very minor role in explaining variation in RUE.

226

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

III. EXPERIMENTAL DETERMINATION OF RUE RUE is a derived variable based on measurements of the ratio of accumulated crop mass and intercepted radiation. Whereas this definition of RUE seems rather straightforward,there has been a wide range of experimental approaches used to estimate RUE. Similarly, a wide range of units have been used to express RUE. For example, RUE estimates differ because i. some are expressed on a PAR basis while others are expressed on a shortwave radiation basis and the transmission of PAR through canopies differs from that of short-wave radiation (Marshall and Willey, 1983); ii. some are expressed on an intercepted basis while others are on an absorbed basis; iii. some are based on net aboveground dry matter production with variable leaf losses while others are expressed on a total dry matter production basis including roots; iv. some are based on differences between two discrete samplings at different stages of crop growth and are subject to large sampling errors in contrast to those based on the fitted slope from many crop growth samplings. To further complicate measurements of RUE, the various experimental approaches used in measuring accumulated crop mass and radiation are subject to a number of errors and bias. Therefore, in comparing RUE among experiments, the reliability in the RUE estimates can vary to a large extent. In the following sections some of the experimental issues associated with measuring the components of RUE are discussed.

A. DETERMINATION OF CROP MASS The crop mass used in the calculation of RUE is usually based on net aboveground biomass production. Inclusion of roots will result in higher RUE but it is difficult to estimate root biomass. In cereals, root mass at anthesis is commonly 10 to 20% of the total crop mass (Gregory, 1994). A recent study on sugarcane in Hawaii showed that belowground mass decreased from 17% of total crop mass at 6 months to 11% from 12 to 24 months (Evensen et al., 1997). Variable losses of mass due to leaf senescence and even shoot death under some stress situations can lead to an underestimate of RUE. In sugarcane, for example loss of trash material can lead to underestimates of up to 15% in RUE (Muchow et al., 1997), and stalk death associated with lodging further reduces RUE (Muchow et al., 1994). Of more importance in the precision of biomass estimates, however, is field and sampling variability. Given that RUE estimates are based on consecutive field

RADIATION USE EFFICIENCY

227

samplings, RUE estimates can be highly variable in nonuniform stands especially when based on the difference between two discrete samplings. For well-managed crop stands, crop mass can usually be estimated to an accuracy of 5 to 15% (Gallagher and Biscoe, 1978). However, use of small plots and small sampling areas can lead to edge effects and biased estimates. For example, Monteith (1978) found that several reports of high maximum growth rates were associated with significant amounts of lateral radiation interception in small plots. Laboratory standards such as calibration of balances and care in partitioning biomass samples contribute to the rigor in estimates of crop biomass and hence RUE. When comparing different crop species, it is also necessary to consider the energy content of the plant mass. For example, Muchow et al. (1993) adjusted soybean biomass upward to take account of the higher energy content of the grain with the energy content of the grain taken as 1.3 times that of the vegetative material (Sinclair and DeWit, 1975; Penning de Vries et al., 1983). In the Muchow et al. (1993) study, energy contents of the vegetative material and of the grain in mung bean and cowpea were assumed the same. Other studies have corrected for the presence of energy-rich substances that accumulate during reproductive growth in several species including peanut (Bell et al., 1992; Wright et al., 1993) and sunflower (Hall et al., 1995; FlCnet and Kiniry, 1995). The variation in nutrients in plant tissue contributes little to estimates of RUE (Squire, 1990). In summary, the use of crop biomass as the basis of RUE calculation may need special attention in dealing with a productive oil crop.

B. DETERMINATION OF SOLARENERGY The amount of radiation intercepted as used in the calculation of RUE has two components (i) the input of solar radiation and (ii) the fraction that is intercepted by the leaf canopy. Incoming radiation can be expressed as either total solar (0.4 to 3 pm) or that in the wavebands of PAR (0.4 to 0.7 pm). Radiation absorbed by green foliage is sometimes substitutedfor interceptedradiation. An important consideration is how incident and intercepted radiation are measured.

1. Incident Radiation Outside the earth’s atmosphere, a surface kept at right angle to the sun’s rays receives energy at a mean rate of 1.36 kJ m-* s-’, a figure known as the solar constant (Monteith, 1972).The proportion of this radiant energy received at the earth’s surface is determined by the geometry of the earth’s surface with respect to the sun and depends on latitude and season. The annual average value of this geometrical factor decreases from about 0.3 in the tropics to 0.2 in temperate latitudes (Mon-

228

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

teith, 1972). The greatest annual input to solar radiation occurs in subtropical regions in latitudes 20 to 30°,in climates with little cloud cover and correspondingly low rainfall (Cooper, 1970). Humid tropical regions show somewhat lower values while temperate Oceanic regions, which currently have the most intensive grassland production, are distinguished by remarkably low radiation input. Solar radiation penetrating the earth's atmosphere is absorbed and scattered by gases, by clouds, and by aerosols in the form of soil and salt particles, smoke, insects, and spores. Accordingly, the incident radiation available for crop growth varies from day to day and data are usually obtained from direct measurements using solarimeters or from indirect measurements based on sunshine hours. For a regularly calibrated and well-maintained weather station, the uncertainty in incident radiation will be about 22%. At sites where it is necessary to estimate incident radiation from records of cloudiness or sunshine hours, the uncertainty in monthly averages of incident radiation will be on the order of ? 10% (Monteith, 1972). However, given the short crop cycles of many annual crops and the fact that as leaf area develops the proportion of incident radiation that is intercepted increases nonlinearly, aberrations in RUE estimates can occur unless reliable daily estimates of incident radiation are obtained from correctly sited and calibrated sensors. Attention to incident radiation data is frequently overlooked in many experimental measures of RUE. The ratio of PAR to total radiation in the direct solar beam is between 0.44 to 0.45 when the sun is more than 30" above the horizon (Moon, 1940) and a figure of 0.45 has often been used by biologists to calculate the receipt of PAR from the flux of total radiation recorded with a solarimeter (Monteith, 1965; Meek et al., 1984).This estimate ignores a contribution of diffuse radiation, which is scattered by gas molecules in the atmosphere and which contains a much higher proportion of PAR than the direct beam. When the solar elevation exceeds 40°, the estimated ratio of PAR to total radiation and the diffuse component is about 0.60. Combining the direct and diffuse components in appropriate proportions, the ratio of PAR to total solar radiation is close to 0.5. Monteith (1972) suggests an average of 0.5 is probably appropriate in the tropics as well in the temperate latitudes. Chlorophyll absorbs very strongly in the blue and red regions of the spectrum so that the light reflected and transmitted by leaves is predominantly green. Integrating over the whole spectrum from 0.4 to 0.7 nm, the fraction of PAR absorbed by leaves is usually between 0.80 to 0.90, the precise figure depending on factors such as the amount of chlorophyll per unit area of lamina. Gallagher and Biscoe (1978) used a figure of 0.90. Using an average figure of 0.85, the fraction of total solar radiation absorbed by green leaves is therefore about 0.5 X 0.85 = 0.425. Accordingly, to convert RUE values to a total solar radiation basis, it is appropriate to multiply those estimates based on intercepted PAR by 0.5, and those based on absorbed PAR by 0.425.

229

RADIATION USE EFFICIENCY

2. Intercepted Radiation The amount of radiation intercepted by the leaf canopy can be determined from the radiation that it receives and transmits. This can be measured using tube solarimeters beneath the crop canopy compared to a measurement of the incident radiation above the canopy (Szeicz et al., 1964; Monteith et al., 1981). The tube solarimeter is a device for measuring irradiance received by a thermopile that provides an electrical output proportional to the difference between the temperature of black and white segments of its surface. This electrical output can be logged to provide continuous diurnal measurement of the fraction of incoming radiation that is intercepted (4) (Muchow and Davis, 1988). Care must be taken in solarimeter placement to ensure that the portions of the canopy “seen” by a set of replicated instruments are not anomalous. When calculating RUE, it is important that the interception is that from the green leaf canopy and not overestimated due to interception by dead leaves. A good practice is to remove dead leaves at weekly intervals from areas where tube solarimeters are placed (Muchow and Davis, 1988; Muchow et al., 1993, 1994). Uniform stands are essential to get reliable cost-effective measures of radiation interception using tube solarimeters, but even in a well-managed nonuniform crop such as sugarcane, reliable estimates of intercepted radiation required four solarimeters per plot (Muchow et al., 1994). The need for replication, the cost and fragility of tube solarimeters, and the need for continuous data-logging have resulted in many surrogate measurements for the fraction of incident radiation intercepted by the crop canopy Commonly, spot measurements are taken below the canopy using a line sensor around solar noon and compared to incident radiation above the canopy (Gallo and Daughtry, 1986). Spot measurements are usually confined to sunny days to avoid measurement difficulties associated with transient clouds. Unfortunately, restricting measurements to sunny days with high fractions of direct radiation results in biased4 estimates that are low. Interpolation between spot measurements also contributes to error inversely proportional to the frequency of measurement. Monteith (1994) also highlights the temporal error associated with spot measurements when done at one time of the day, usually midday, because4 is a function of time of day. Charles-Edwards and Lawn (1 984) have reported that& is underestimated by up to 10%when measurements are taken in the middle of the day instead of being integrated during the day, and FlCnet et al. (1996) found that east-west row orientation minimized time-of-day effects. Muchow (1985) measured diurnal variation inf; in a range of grain legumes and showed that the lower the value off, at solar noon, the greater its diurnal variation. The underestimate in& based on noon measurement decreased from 40 to 3% when& at solar noon increased from 0.2 to 0.9. RUE estimates based on spot measurements must be examined cautiously.

q).

230

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

A second method used for estimating& is based on measuring the reflected radiation from the leaf canopy in two spectral bands. These techniques are based on a comparison of the reflected radiation in a red wavelength band (630 to 690 nm) and a near infrared wavelength band (760 to 900 nm). Generally, a normalized difference vegetation index (NDVI) is calculated as the difference in reflectance between these two wavelength bands (near infrared minus red) divided by the sum of the reflectance in these two wavelength bands. Empirical relations between & and NDVI have been obtained for wheat (Garcia et al., 1988) and for maize and soybean (Daughtry et al., 1992). Major et a/. (1991) calculated& for maize from the reflectance data in these two wavelength bands by using a model that considered canopy radiative properties in detail. Unfortunately, Major er al. (1991) offered no direct comparison betweenfi calculated using the reflectance data in the model and& obtained directly from radiation interception results. All experiments using spectral reflectance data to estimate fi have focused on essentially cloud-free days near solar noon. Therefore, the problem of one-time, spot measurements also arises with attempts to estimate interception from the spectral composition of reflected radiation (Garcia et al., 1988; Rudorff et al., 1996; Major et al., 1992). Extensions of this technique to estimate& with NDVI estimated from satellites are similarly difficult because they are spot measures that are taken at various angles relative to the crop surface (Hanan et al., 1995). A third method of estimatingfi is to estimate the foliage cover by photographing the canopy either from above or looking upward from the soil surface. This is done by fitting a standard camera with a fisheye lens (Anderson, 1971). In addition to being less expensive than tube solarimeters, the technique has the advantage that photographs sample a relatively large area, may be taken rapidly in the field, are easily reproducible, and provide a permanent record of the state of the crop. Gregory and Marshall (1980), as cited by Monteith et al. (1981), compared estimates of& using a fisheye lens to the value calculated from tube solarimeters placed above and below canopies of millet and peanut. The hemispherical photographs were taken from the soil surface and the estimates agreed well both in the early and late stages of development. However, there was a discrepancy of about 20%during the period when the canopies were developing rapidly, probably as a consequence of the unavoidable positioning of the camera between the rows. Steven et al. (1986) compared foliage cover estimates of& with& estimates from solarimeters in sugarbeet, field beans, and barley with variable results, which were particularly poor for barley. Haverkort et al. (1991) compared several indirect methods for determining&, and observed that ground cover was associated with fewer errors than methods based on leaf area index and extinction coefficient, and on infrared reflectance. These surrogate measures, while useful in some circumstances, must be treated with caution in calculating RUE given the spatial and temporal errors discussed above.

RADIATION USE EFFICIENCY

23 1

C. CALCULATION OF RUE There are many methods used for estimating the components of RUE with many associated errors. Regularly calibrated and well-sited sensors are required to measure daily incident radiation. Fractional interception is best measured continuously using tube solarimeters in uniform crop stands to avoid temporal errors. Dead leaves should be removed periodically so that interception by green leaves is measured. Similarly biomass needs to be well defined (i.e., shoot plus or minus roots with or without senesced material) and taken from representative areas without edge effects. There are also important issues in how the component data are used to calculate RUE. RUE can be calculated from the difference in biomass between two consecutive harvests divided by the corresponding amount of radiation intercepted. This method suffers from large errors associated with calculated differences. A more appropriate measure is to fit a linear relationship between cumulative biomass accumulation and cumulative radiation interception, with RUE calculated as the slope of the linear relationship. This method has been widely used. This cumulative approach has been criticized by Demetriades-Shah et d. (1992) but the criticisms were appropriately refuted by Arkebauer et al. (1994), Kiniry (1994a), and Monteith (1994). Care needs to be taken in obtaining RUE estimates by regression, as in some species RUE declines during reproductive growth, even under high input conditions (Muchow and Davis, 1988), associated with N mobilization to the grain and subsequent reduction in leaf photosynthetic capacity (Muchow and Sinclair, 1994). Perhaps most important is the need for the establishment of good crop stands and good husbandry with well-recorded inputs, as a prerequisite for obtaining reliable estimates of RUE. Given the wide variation in experimental methods used to measure RUE, we now present a case study on calculating RUE over a range of environmental conditions. Muchow (1994) reports a study in which maize was grown in both a tropical and a subtropical environment at different sowing dates under different N fertilizer rates. The same hybrid (Dekalb XL82) was sown with the same row spacing, and identical management inputs with the same sprinkler irrigation regime in all crops. Crops were thinned to the same density and biomass accumulation was determined by sampling 2 m2 every 7 to 10 days from thinning to maturity. Radiation interception was measured by placing a tube solarimeter diagonally across the two inner rows of each replicate plot at ground level. These solarimeters were used to record at 2-min intervals the radiation (0.35 to 2.4 km) transmitted through the crop canopy, using the data collection system described by Muchow and Davis (1988). Dead leaves (>50% blade area senesced) were removed at about weekly intervals from plants around the tube solarimeters, so that radiation transmission through green leaf area only was recorded. Another tube solarimeter was placed above the crop and the incident radiation was recorded. Dai-

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

232

ly totals and individual tube calibration factors were used to calculate the fraction of incident radiation intercepted (&) in each plot. Since the readings from individual solarimeters were found to vary by up to 20%from the nominal calibration, the absolute incident radiation (S) was recorded with a regularly calibrated pyranometer. The amount of radiation intercepted (Si) was calculated as the cumulative product of the daily& and S. Radiation use efficiency was calculated both as the average value from sowing to maturity and as the maximum value before photosynthetic capacity declined during grain-filling. The average RUE was calculated as the ratio of net aboveground biomass at maturity to cumulative Si from sowing to maturity. The maximum RUE was derived as the fitted slope of the linear relationship between net aboveground biomass and Si (Fig. 3) using a stepwise regression procedure (Muchow and Sinclair, 1994). Starting at crop maturity, data points were progressively removed from the fit until no further improvement was gained in the proportion of variance accounted for by the regression. Under high N supply, maximum RUE was relatively stable across sowings and environments, indicating no response to temperature, absolute incident radiation level, or water vapor saturation deficit for the range of conditions experienced (Fig. 3). The difference in biomass production could be largely explained by differences in Si. Where N supply limited yield, the decrease in biomass production was associated with a much larger decrease in RUE than in Si. Under both low and high N sup-

“0

A

0

600 1200 1800 Intercepted radiation (MJ m-’)

600 1200 1800 Intercepted radiation (MJ m-’)

B

Figure 3 Relationship between net aboveground biomass accumulation and radiation interception for maize where 0 and 24 g N m-2 were applied for (A) 29 January 1986 sowing at Katherine and (B) 28 August 1990 sowing at Lawes (Reprinted from Field Crop Res., Vol. 38, R. C. Muchow, “Effect of nitrogen on yield determination in irrigated maize in tropical and subtropical environments,” pp. 1-13, with permission from Elsevier Science).The slope of the fitted linear relationship is the maximum radiation use efficiency and values are (A) 0.87 g MJ-’ for 0 g N m-2 and 1.65 g MJ- I for 24 g N m-2 and (B) 0.52 g MJ-l for 0 g N m-* and 1.64 g MJ-’ for 24 g N m-*.

RADIATION USE EFFICIENCY

233

ply, the grain demand for N could not be met solely by soil N uptake during grain filling, and there was significant mobilization of vegetative N to grain N. Consequently leaf N and RUE declined during grain filling in all situations, highlighting the difference between maximum and seasonal average RUE.

W. EXPERIMENTAL MEASURES OF RUE The objective of this section is to compare RUE reported for various species and experiments. The focus is particularly on data collected under optimum, usually control conditions, in order to compare observations on potential RUE from each study. The collection of a baseline of potential RUE data could serve as a useful reference in developing a realistic perspective on the upper limits of RUE that might be reasonably expected for individual crop species. Fortunately, in the past 10 years there has been a large expansion in the number of studies that included estimations of RUE, so there is a large data resource on which to make these comparisons. This large data resource also allows some selectivity in the data to be included in the comparison among species. As indicated in the previous section, there are a number of possible sources of error or bias as a result of various methodologies used in estimating RUE. We have established two criteria in selecting data sources to be included in this comparison. Our narrowing of the data sources to be included in this comparison does not mean, however, that the excluded data are necessarily in error or biased. We have introduced this selectivity among the published data sources in order to focus on those reports that used more exhaustive methodology in developing estimates of RUE. The two criteria used to select data for this comparison of RUE are based on experimental issues discussed previously. One criterion was that we included only estimates of RUE based on direct measures of canopy radiation interception during the growing season. Several studies calculated radiation interception based on measures of leaf area index and an assumed radiation extinction coefficient. Errors in measuring leaf area index and uncertainty of transferring an extinction coefficient to a new situation can introduce substantial uncertainty in RUE estimates. The second criterion was that RUE must have been obtained from several periodic observations through the growing season. RUE values that were calculated based only on the difference between two observations were not included because of the greater level of uncertainty in the RUE estimates when only two measures are made during the growing season. After selecting those data sources that met the two above criteria, there was also a problem in making comparisons among various experiments because of differences in how RUE was expressed. To facilitate comparison of RUE among varous sources, an attempt was made to convert estimates of RUE to a common unit.

234

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

The common unit selected for comparing RUE was mass (usually shoot only) per unit of intercepted solar radiation. Indeed, most experiments were directly dependent on measurements obtained in these units. Deviations in the expression of RUE in many cases resulted from decisions by investigators to convert the measurement units into alternate units. Fortunately, many papers reported the methods used to make these conversions so that it was straightforward to recalculate the units as shoot mass per unit of intercepted solar radiation. In those cases where the conversion technique is not explicitly available in a paper, there were two assumptions that may have been invoked. The first assumption was used to convert those reports of RUE based on intercepted PAR into units of total solar radiation. As discussed in the previous section, it was assumed that the level of PAR was 0.5 of total solar radiation. The second assumption was used to convert from absorbed PAR radiation to intercepted PAR radiation. When a conversion was unavailable in the original data source, it was assumed that 0.85 of intercepted PAR radiation was absorbed, as discussed previously. The estimates of RUE were collated by crop species (Table I) because of the important differences that have previously been suggested among species (Gosse et al., 1986; Kiniry et al., 1989).An attempt was made to include some of the key information about each experiment in Table I, and this table is the basis for making comparisons of potential RUE among species.

1. Maize

There have been extensive reports on RUE in maize. An important consideration arising from these data is that maximum RUE occurs during vegetative growth and there is a tendency for RUE to decrease during grain filling associated with mobilization of leaf nitrogen to the grain and consequent reduction in RUE (Muchow and Davis, 1988). The maximum value reported was 1.86 g MJ-' by Otegui et al. (1995). However, radiation interception was measured by spot readings around solar noon at 14-day intervals and since the fraction of radiation intercepted varied from 0.59 to 0.79 over the 2 years of the study, this may result in an underestimate of radiation interception and an overestimate of RUE. A number of studies have shown maximum RUE in the range 1.6 to 1.7 g MJ-' (Table I). In fact, there is a great deal of consistency around these values for a large number of studies. Interestingly, the study of Andrade et al. (1992, 1993) showed lower maximum RUE associated with the lower temperature environment. Also, Tollenaar and Aguilera (1992) observed higher RUE in new compared with old hybrids. The main outlier in Table I for maize is Bolanos and Edmeades (1993) with a maximum RUE of 1 g MJ-'. The authors noted that while the low RUE values were

Table I Summary of Maximum RUE Reported for Various Crop Species, Including an Estimate of These RUE in the Common Units of Plant Mass per Unit of Intercepted Solar Radiation Mas,

Source

Lacation

Expenmental variables

Samplearea

Stage

hdiauon Shoot\

Spot

Intercepted

or

01

01

total

continuous

RUE

absorbed

PAR or solar

Maximum reported value

Intercepted

Solar

1.6Og MJ-' I..U)gMJ-'

Estimated ddjusment (gMJ;;)

Comments

Maim Katherine, Australia

N. species

Muchow and Davis (1988) Tollenaar and Bruulsema (1988)

Elora, Canada

Hybrids. density

2.0.2.3, and and 2.1 mz

Muchow (1989a)

Katherine, Ausualia

Sowing date, species

Andrade er al. (1992)

Balcarce, Argentina

Year,CUltiVar

Daughvy er 01 ( 1992) Tollenaar and Aguilera (1992) Andrade er 01. (1993)

West Lafayette, IN,USA Elora. Canada

Balcarce, Argentina

Bolanos and Edmeades (1993)

Tlaltiapan. Mexico

2 m-?

Vegetative. season Vegetative

Shoots Shoots

Continuous

Absorbed

PAR

3.46gMJ-'

1.40 1.47

2 m2

Vegetative. season

Shoots

Continuous

Intercepted

Solar

1.59 g MJ-' 1.27gMJ-'

1.59 1.27

10 plants (6.1

Vegetative. season

Shoots

Spot (?)

Intercepted

PAR

3.03 g MI2.96gMJ-'

0.54111~

Season

Shoots

Absorbed

PAR

4.26gM.-'

1.81

Hybrids. density

3.4 to 3.6 m2

4-6 weeks postsilking

Shwts

Spot (reflected) Continuous

Absorbed

PAR

3.78 g MI-'

1.61

Year.cultivar.

10 plants

Vegetative

Shoots

Spot (-15-day intervals)

Intercepted

PAR

3.17 g MJ-'

1.52

2.25 m2

Anthesis, maturity

Shoots

Spot(10- to 12day intervals)

Intercepted

PAR

1.99gMJ-' 1.48gMJ-'

1.00 0.74

to 9.1 plants

Continuous

1.60

'

1.45 1.42

m-Z)

sowing date

Selection cycles, water regime

Reduction in RUE under N stress. Growth rate 30.5 g m-2 day-'. PAR absorbed 4.07 mol photon W Z day-', RUE = 3.46 PAR absorbcd 0.425 = 1.47. RUE during vegetative growth fitted by regression. Decrease In RUE after silkiing;RUE maize higher than for sorghum. Low-temperature study with mean 15 to 18°C during vegetative gmwth.

Silking to 6 weeks postsilking; higher DM in new cf old hybrid atmbuted to higher RUE. Max RUE ranged from 2.27 to 3.17gM.-'PARover5years dependent on temperature; RUE varied with sowing date (temperature). Noted low RUE under well-watered conditionchigh VPD. Water applied only every 10 days.

continues

Table I-confinued Mass

Source Kiniry (19946)

Muchow (1994) Otegui el a/.

Lacation Temple, TX, USA

w N

m

Coates (1986) Muchow and Davis (1988) Hammer and Vanderlip (1989) Muchow (1989a)

Year, species, competition

Sample area -0.3 mz

Katherine & Lawes. Australia R o p . Argentina

Location. year. N

2 m-2

Hybrids, sowing date

1 m2

ICRISAT. India

Year, soil type. hybrids Sowing date. cultivars. row spacing, density N. species

Periodic. 3 mL Penodic,

(1995) Sorghum Sivakumar and Huda (1985) Muchow and

Experimental variables

Kununurra, Ausvalia Katherine. Australia Manhattan, KS. USA

Katherine, Austrdlia

Greenhouse. genotype X temperature Sowing date, species Location. year. N

Stage Vegetative

Radiation Shoots

Spot

01

Or

total

continuous

Shoots

Spot ( I - to 23day intervals)

Periodic. 2

m2

PAR or solar

Maximum reported value

Estimated adjustment

Intercepted

PAR

3.42 g MJ-'

1.54

Comments

(g

(1991)

3.75gM1-' (1992) 1.68 g MI-' 1.67 g MI-' 4.14 g MJ-' 3.39 g h%-'

1.69

Shoots

Continuous

Intercepted

Solar

Shoots

Spot (14-day intervals)

Intercepted

PAR

Season

Shoots

Intercepted

PAR

2.74 g MI-'

1.37

Season

Shoots

Spot (7- to 10day intervals) Spot (7-day intervals)

Intercepted

PAR

2.40 g MJ-'

1.20

Vegetative. season Vegetative

Shoots

Continuous

Intercepted

Solar

Shoots

Continuous

Intercepted

PAR

I .25 1.12 2.16

Shoots

Continuous

Intercepted

Solar

Vegetative

Shoots

Continuous

Intercepted

Solar

1.25 g MI-' I.I2gMJ-' 4.31 g MI-' at 25°C 2.99 g MI-' at 17°C 1.29 g id-' I.IOgW-' 1.26 g MJ-'

Vegetative, season

1.68 1.67 1.86 1.53

1.29 1.10 1.26

Katherine & Lawes. Australia

Westgate PI a/. (1997)

Moms, MN, USA

Hybrids, row spacing

Periodic, 1 m2

Vegetative

Shoots

Spot (7- to 10 day intervals)

Intercepted

PAR

3.02 g MJ-'

1.51

Ayr. Australia

Growth analysis, plant crop

Periodic.

Season

Shoots

Continuous

Intercepted

Solar

1.75 g MJ-'

1.75

2 m2

15 mz

Reduction in RUE under N stress. No IoCation effect on max RUE. Fraction PAR intercepted varied from 0.59 to 0.79 over 2 years, 4 sowing dates. and 4 hybrids.

Decrease in RUE after anthesis. Cultivar difference.

IS O

Muchow and Sinclair (1994)

Sugarcane Muchow er al. ( 1994)

Periodic.

Intercepted or absorbed

Vegetative, season Vegetative. season

I m' Periodic, 2 m' 15 potc

RUE

Decrease in RUE after anthesis. RUE independent of temperature, solar radiation. and water vapor saturation deficit.

If assume 15%underestimate of biomass due to nonrecovery of all trash (Evensen er a/., 1997) then RUE = 2.0 g W - ' .

Robenson cr a/.

Munchow el al. (1997)

Potato Allen and Scott (1980) Burstall and Hams ( 1986) Jeffenes and Mackerron ( 1989) Kenaf Muchow (1992)

RUE lower during late growth due to winter temperature and biomass loss due to stalk death. a h .

1.87

Includes estimates of 15% wash loss in biomass estimates for Ausualia: Hawaii measured biomass estimates for 1st-year growth.

Variety, plant, ratoon crop

Periodic. I5 m2

Season

Shoots

Continuous

Intercepted

Solar

Kunia. Hawaii, USA; Ingham and Ayr. Australia

Variety

Periodic. 18.9 m2 and 15 m2

Season

Shoots

Contiuous (and Intercepted calculated using k = 0.4)

Solar

Sutton Bonnington. UK Sonning-on-Thames, England Scotland

NA

Periodic, NA Penodic. 1.4rn' Periodic, 1.08 m2

Season

Total

Continuous

Intercepted

Solar

I .6 g MI-'

1.6

Season

Total

Continuous

Intercepted

Solar

1.76 g M.-'

1.76

Varietal differences in RUE.

Total

Continuous

Intercepted

Solar

1.75 g MI-'

I75

Includes tubers

Katherine. Australia

Water. N

2 m2

Season

Shoots

Continuous

Intercepted

Solar

1.20 g h W '

1.20

RUE decreased more than RI under water and N stress.

Lincoln. New Zealand

Season, cultivar

Season

Shoots

'?

Intercepted

PAR

2.38 g MI-'

1.19

All results gave common RUE.

Sutton Bonnington.

N. year

-0.2

Vegetative

Total

Continuous

Absorbed

PAR

1.28

N. stage location

12 plants

Vegetative and reproductive

Shoots

Spot (reflected)

Absorbed

PAR

4.07 mmol hex. mo1-I 3.82 g MI-'

1.62

RUE decreaed with decreased N application. No location difference. Decrease = 1 low N. Decreaye in reproductive stage.

0.525 m2

Vegetative

Shoots

Intercepted

PAR

1.46g MJ-'

0.73

Decrease during reproductive p w h .

N

0.3 m2

Vegetative

Shoots

Spot(6 occasions) Spot (9- to 20day intervals)

Absorbed

PAR

3.36gM.-'

1.51

Penh. Ausrralia

Cultivar

0.32 m2

Vegetative

Shoots

Continuous

Intercepted

PAR

2.93 g MI-'

I .46

No difference in N treatment identified. RUE increased up to 10 days after anthesis. Cultivar and season difference.

East Beverley. Australia

Season. sowing dare

1.068 m2

Season

Shoots

Contmuour

Intercepted

PAR

I .68 g MI-'

0.84

Variety Water supply

1.72 g MI-' plant 1.59gM.-' ratoon 1.87 g MI-' Hawaii 1.96 g MI Australia

I .72 1.59

Ingham. Australia

(19%)

~

'

I .96

w N

Wheat Wilson and lamteson (1985) Green ( 1987) Garcia el a/. (1988)

Gregoory er a/

(1992) Fischer (1993)

Yunusa er a / (1993) Gregoory and Eastham (1996)

UK Mandan. ND. USA; Manhattan, KS, USA; Lubbock, TX.USA East Beverley. Austmlia Griffith. Australia

m2

per ueatmenl

~~

~

continues

Table I-continued Mass

Source

Calderini era/. (1997) Barley Gregory el al. (1992) Goyne n al (1993) Jamieson el a/. (1995)

w N Q)

Rice Hone and Sakuratani (1985) Inhapan and Fukai (1988) Sunflower Trapani cr o/. (1992)

Experimental variables

Lncation

Buenos Aires. Argentina

Cultivar

East Beverley. Australia Warwick, Australia

CUltiVZS

Lincoln, New Zealand

Cultivars, irrigation Irrigation

Tsukub4 Japan

Sowing date, shading cultivar

Redland Bay, Australia

CUltiVar.

Buenos Aires, Argentina

Cultivar

Samplearea

0.075 m2

0.525 mz

Radiation Shoots or total

Vegetative and reproductive

Shoots

Spot (3- 104-

Vegetative

Shoots

Stage

RUE

Spot

Intercepted

01

Or

continuous

absorbed

PAR or solar

Intercepted

Solar

day intervals)

Spot (6 wca-

Maximum reported value

1.25 g MI-' I .02 g MI-

'

Estimated adjustment (g MJLd,)

I .02

Cultivar variation. RUE decreased post-anthesis.

1.25

Intercepted

PAR

1.79 g MJ-'

0.90

Decrease during reproductivegrowth.

Absorbed

PAR

2.9OgMJ-'

1.30

Some cultivar differences

Intercepted

PAR

2.33 g MJ-'

1.16

Some irrigation differences

0.36 m2

Season

Shoots

0.1 m2

Season

Shoots

sions) Spot(-7-day intervals) Spot (once)

Season

Shoots

Continuous

Absorbed

PAR

3.28gMJ-'

1.39

Shading increased RUE

0.15 m2

Entire season

Shoots

Spot (-74.9 intervals)

Intercepted

Solar

0.93 g MI-'

0.93

No cultivar difference. RUE decreased for dry.

2-3 plants

Vegetative

Shoots and total

Spot(4- to7day intervals)

Intercepted

PAR

3.13g,,, MI-'

1.56 I .63

Decreased at emergence and photo synthesis. No cultivar difference.

1.14

Decreased during early stages. Derreased with low N. No effect of density. Decreased with less N. Decreased late in season. No N difference.

irrigation

Gimenez el 01. (1994)

Cordoba, Spain

N. density

4 4 plants

Vegetative

Shoots

Spot (3 occasions)

Intercepted

PAR

3.26 g_, MI-' 2.29 g MI-'

Halletal. (1995)

Buenos Aires. Argentinia Temple, TX.USA

N. density

3-4 plants

Season

Total

Intercepted

Solar

1.24gMJ-'

1.24

N

1.4 m2

Season

Shoots and total

Spot(l0a'casions) Spot (9 a'casions)

Intercepted

PAR

1.77 g5,, MI-'

0.88 0.97

1.33

Henet and Kiniry (1995)

Bange et al. (1997a) Bange et al.

Ganon College, Australia Ganon College.

Comments

N

7 plants

Season

Shoots

Continuous

Intercepted

Solar

I .94 &md MJ-' 1.47 g MJ-'

Shading

7 plants

Season

Shoots

Continuous

Intercepted

Solar

1.33 g MJ-'

I .47

Decreased with decreased N treatmentto 1.25gMJ-'. Higher RUE with shading.

soybean Nakaseko and Gwth (1983) Leadley ef ol. (1990) Daughuy ef a/.

Sapporo. Japan Raleigh. NC. USA Beltaville. MD

(1992)

Muchow et al. (1993) Sinclair and Shiraiwa (1993) Rochette pf a/. (1995)

Peanut Bell erol. (1987)

Katherine, Australia Gainesvtlle, FL, USA, Shiga, Japan Ottawa, ON, Canada

Kununurra, Australia

Species

Season

Shoots

Spot (3 occa-

Intercepted

PAR

sions) Continuous

Solar

I I .8 mg kcal-' 0.86 g MJ-I

Intercepted

0.86

spot (reflected) Continuous

Absorbed

PAR

2.34gMJ-'

0.99

Intercepted

Solar

0.86 g MJ-'

0.86

spot (7-daY intervals)

Intercepted

Solar

0.66 g M.-' 1.15gMI-l

0.66

Shoots and total

Continuous

Intercepted

2.04 g
1.02

2.09 g,, MJ-'

I .I4

Total

Spot (7day intervals)

1.37

0.98

Increased greatly from shading.

0,. inside open-

4 plants

Season

Shoots

top chambem Row spacing and direction Season. location

0.45 or 1.9 mL

Season

Shoots

1.70r2rn2

Seaon

Shoots

CUlt!"ar. location

0.56 or I rnz

Comparison of biomass harvest to co2flux density

0.48 m2

Density

6 plants

Shoots

Season

PAR

Kingaroy and Bundaberg, Australia Gainesville. FL. Bennett er n/ (1993) USA Kingaroy and Wright er ol. Bundaberg, (1993) Australia Bell el 01. (1994~) Delhi, Canada Faba bean Fasheun and Dennett (1982) Slim and Saxena (1992) Madeira er ul. (1994)

Differences among cultivm.

1.15

Only small vanation in RUE when w > 2

Shading

10 plants

Season

Total

Spot (10 occa. sions)

Intercepted

Solar

Location, cultivar

1.2or1.8rn'

SCaSO"

Total

spot ( 7 h Y

Intercepted

Solar

1.12 g MJ-'

1.12

No cultivar difference. Location difference.

Intercepted

Solar

1.01 g MJ-'

1.01

No cultivar difference.

Intercepted

Solar

1.07 g MI-'

I .07

RUE sensitive to leaf N content.

Intercepted

PAR

a Hyderabad, India

Decreases with increasing 0,.

3.04 g M1- I (veg.1 2.27 g M.-' (sea.) 0.98 g M.-'

Vegetative and season

w N

Stirling er ol. (1992) Bell erol. (1992)

1.26

No density difference.

1.02

intervals) Cultivar

1.22 m2

Season

Total

N

1.2 or I .8 rn2

Season

Total

Cultivar, season

0.6 m2

Season

Total

spot (Fday intervals)

Intercepted

Solar

2.24 g MJ-'

1.12

Small cultivar differences.

Reading, UK

Faba bean

3 plants

Season

Shoots

Continuous

Absorbed

PAR

4.8gMJ-1

2.04

Difference with sowing date.

ICARDA. Syria

Faba bean. cultivar, season, density Faba bean disease

0.225 m2

Season

Shoots

Spot (7-day

Intercepted

PAR

2.06 g MJ-'

1.03

Shoots

intervals) Continuous

Intercepted

Solar

1.45 g M1-'

I .45

Cultivar and sowing density variation. Disease decreased RUE.

Sutton, Bonnington, UK

10 plants

Season

spot (7-day intervals) Spot (7-day intervals)

continues

Table I-continued Mass

Source Other legumes Hughes er nl. (1981) Muchow and CharlesEdwards (1982) Nakaseko and Gotoh (1983) Heath and Hebblethwate (1985) Leach and Beech (1988) Singh and Sri Rama (1989) McKenzie and Hill (1991)

Location

Experimental variables

Trinidad

Pigeon pea

ffimbedey, Australia

Mung bean

Sapporo, Japan

Phaseolus. Azuki

Samplearea

Stage

Radiation Shoots or total

Season

Shoots

I m'

Vegetative

Shoots

?

Season

Shoots

bean

Spot

RUE Maximum repofled value

Estimated adjusment (gMJ;I)

Intercepted or absorbed

PAR or solar

Intercepted

Solar

1.91%

1.09

Variation due to watering.

Absorbed

PAR

2.17 g MJ-'

0.92

RUE less during @-filling.

Spot (3 occasions) Spot (2- to 3day intervals)

Intercepted

PAR

0.98

Species differences.

Intercepted

PAR

9.1 mg kcd1.46 g MJ-'

1.46

RUE decreased by water shortage at one location.

Spot (14- to 28day intervals) Spot (3- to 4day intervals) Spot(14day intervals)

Intercepted

PAR

1.53 g h W '

0.76

Intercepted

Solar

0.67 g MJ-'

0.67

Cultivar variation lower for wide row spacing. Water deficit decreased RUE.

Absorbed

PAR

2.14gMJ-'

0.96

Sowing date difference

Some differences between species.

Or

continuous

Spot(peri0dically) Spol(7-day intervals)

'

Comments

P a

0.25 m2

Season

Shoots

Chickpea cultivar. spacing Chickpea

7

Season

Shoots

0.3 m2

Season

Shoots

Lentil. sowing date. cultivar. irrigation Cowpea. mung bean Lupin, season

0.1 m2

Season

Shoots

1.7 or 2 m2

Season

Total

Continuous

Intercepted

Solar

1.09 g MJ-'

1.09

1.068 m2

Season

Shoots

Continuous

Intercepted

PAR

1.16 g MJ-'

0.58

Madrid, Spain

Pea

0.3 mz

Vegetative

Shoots

Intercepted

PAR

1.43 g MJ-'

0.72

No difference between two cultivars.

Merredii, Ausualia

Various swcies

0.5 m2

Vegetative

Shoots

Spot (7day intervals) Spot(lCday intervals)

Intercepted

PAR

1.09 g MJ-'

0.54

Some species differences.

Suuon Bonnmgton,

UK Dolby, Australia Hyderabad. India Cunterbury. New Zealand

Muchow et a/.

Katherine, Ausualia

(1993) Gregoty and Eastham

East Beverley,

Austda

(1996)

Manin er nl. (1%)

Thomson and Siddique (1997)

Nore. The only experimental reports included were those that (1) measured intercepted radiation and (2) measured plant mass accumulation at several times during the growing season.

RADIATION USE EFFICIENCY

241

obtained under apparently well-watered conditions, a high saturation deficit may have contributed to water stress where water was applied only every 10 days. There have been a number of published reports with maximum RUE in maize much higher than those shown in Table 1. Jones (198 1) cited by Kiniry et al. (1989) reported 4.5 g MJ-' PAR (or 2.25 g MJ- solar) but radiation interception was estimated from an extinction coefficient and not measured directly. Also, this value was obtained at a density of 6 plants m-*, whereas at 4 plants m--2 RUE was 3.6 g MJ-' PAR (or 1.8 g MJ-' solar), which is close to the maximum values reported in Table I. Indeed, it is difficult to explain such a marked response with a relatively small change in density, especially since the greater effect is more likely on the amount of intercepted radiation. Similarly, Cabelguenne (1987), as cited by Kiniry et al. (1989), and Kiniry (1987), as cited by Kiniry et al. (1989), obtained high RUE values of 2.05 and 1.95 g MJ- I , respectively. Here, either the extinction coefficient was estimated or there was high variability in RUE across plant densities where radiation interception was measured with a line sensor only in the middle of the day. Sivakumar and Virmani (1984) measured maximum RUE of 0.82 g per mole and Kiniry et al. (1989) converted this to 3.8 g MJ-' PAR. Here PAR interception was measured only at midday on noncloudy days and the conversion from PAR to solar was not given. If a conversion of 0.5 is used, this translates into a RUE of 1.9 g MJ-I solar. If a conversion of 0.45 is used, then RUE is 1.7 g MJ-', which is in line with the maximum reported values in Table I. In summary, the maximum RUE for maize is consistent in the range 1.6 to 1.7 g MJ-I during vegetative growth and the RUE during reproductive growth decreases such that the seasonal RUE ranges from 1.3 to 1.7 g MJ-' (Table I).

'

2. Sorghum Fewer studies were available where maximum RUE was measured in grain sorghum. Values seem to be consistently in the range 1.2 to 1.4 g MJ-' during vegetative growth with seasonal values being slightly lower (Table I). Sivakumar and Huda (1985) reported a maximum value for RUE of 1.37 g MJ-', which may be slightly overestimated because PAR interception was only measured at midday on noncloudy days. An important outlier for sorghum is the study by Hammer and Vanderlip (1989) where a maximum of 2.16 g MJ-' was recorded. However, this study was conducted in a glasshouse and radiation interception was measured using a line sensor at solar noon. Additional reflected radiation that was not measured and an increased proportion of diffuse radiation may have contributed to the high RUE in this study. Several higher maximum RUE values have been reported for sorghum by Kiniry et al. (1989). Again, they cite Sivakumar and Virmani (1984) at 2.9 g MJ-I PAR. At a conversion of PAR to solar of 0.5, this gives a RUE of 1.45 g MJ-'. Other

2 42

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

higher values cited by Kiniry et al. (1989) are from studies where radiation interception was estimated based on an extinction coefficient. The exception is Steiner (1986) where Kiniry et al. (1989) quoted a value of 3.0 g MJ-I but no values are given in the actual paper by Steiner (1986). Rosenthal et al. (1993) also recorded a maximum RUE of 3.46 g MJ- PAR absorbed, or as converted 1.47 g MJ-'. However, here absorbed PAR was estimated from leaf area and there was a highly scattered relationship between biomass and absorbed PAR (? = 0.68). In summary, it appears that the maximum RUE for sorghum is in the range 1.2 to 1.4 g MJ-', with RUE from sowing to maturity being less. This range of potential RUE for sorghum is less than that of maize. In direct comparisons of RUE in sorghum and maize, sorghum had the lower RUE (Muchow and Davis, 1988; Muchow, 1989a; Muchow and Sinclair, 1994). Since sorghum and maize both have the C , photosynthetic pathway and similar potential leaf photosynthetic rates (Muchow and Sinclair, 1994), it is surprising that sorghum has not been commonly observed to have RUE values equivalent to maize.

'

3. Sugarcane Very few studies have measured RUE in sugarcane. Recent work by Muchow et al. (1994) and Robertson et al. (1996) have shown maximum values in the range 1.7 g MJ- based on net aboveground biomass. Sugarcane has a very long growth period (12 to 36 months) compared to most crops, and loss of mass in the senesced leaves (trash) is a major concern in the determination of RUE. Where all trash has been recovered, the value is closer to 2 g MJ-' (Muchow et al., 1997). Ratoon crops have a lower maximum RUE compared to plant crops (Robertson et al., 1996). These studies for sugarcane are particularly interesting given the long growth duration and large standing mass of sugarcane. In the study of Muchow et al. (1994), a maximum RUE of 1.75 g MJ-' was obtained where biomass was linearly related to intercepted radiation up to a biomass of 72 tons h-l. Not all the trash was recovered in this study and if it is assumed that the trash accounts for 15% of the biomass (Evensen et al., 1997), then the RUE is close to 2.0 g MJ-I as estimated for Hawaiian crops by Muchow et al. (1997). The reason that sugarcane has a higher RUE value than does maize has not been fully explored. One possibility, however, is that the ultimate product in sugarcane, sucrose, has a lower energy content than the seeds of maize. Maize seeds contain protein and lipids such that 0.7 1 g seed is produced per gram of photosynthate (Sinclair and DeWit, 1975). In sugarcane, the storage of sucrose is likely to result in even greater than 0.83 g carbohydrate produced per gram of photosynthate (estimated by Penning De Vries, et al., 1974). Therefore, much of the advantage in RUE of sugarcane over maize may simply be a result of the difference in energy content of the plant product of these two species. In all three of the sugarcane studies in Table I, RUE declined during late growth

RADIATION USE EFFICIENCY

243

due both to lower temperature during winter and to biomass loss associated with stalk death and lodging. Also, for Hawaiian crops, the apparent RUE during the second 24 months of growth was lower and this is likely to be associated with lodging and stalk death. However, it is extremely difficult to accurately measure radiation interception in lodged crops, and hence estimates of RUE in large lodged crops must be viewed with caution. In summary, the maximum RUE for sugarcane appears higher than those for maize or grain sorghum with values approaching 2 g MJ-' when the majority of dead leaf is recovered. Consequently, the RUE results from the three C, species indicate a large range among these species in the expression of maximum RUE.

1. Potato

The several studies that examined RUE in potato have consistently obtained maximum values in the range 1.6 to 1.75 g MJ-' (Table I). These values seem particularly high for a C, species. Similar to comparison between sugarcane and maize, the reason for the high RUE values in potato relative to other C, species may be attributed partly to the biochemical composition of the plant product in potato. The tuber is up to 80% of the total plant weight (Burstall and Harris, 1986; Jefferies and Mackerron, 1989) and the tuber is extremely high in starch content. The conversion of photosynthate to carbohydrate in potato may be higher than 0.83 g carbohydrate per gram photosynthate because starch is the main carbohydrate product rather than cellulose (Penning De Vries, 1975). In contrast, seeds of wheat contain approximately 14% protein and have a conversion from photosynthate to seed mass of 0.7 1 (Sinclair and DeWit, 1975). This difference in the photosynthate costs for production of plant mass (17% advantage to potato) is consistent with potato having a high RUE. In contrast to variations in RUE through the season observed in other species, the three reported studies with potato all indicate an essentially constant RUE through the growing season. This may result from the fact that the growth of the tubers begins at an early stage of crop development and continues at essentially a constant fraction of total growth (Allen and Scott, 1980; Burstall and Harris, 1986; Jefferies and Mackerron, 1989). Consequently, there is no strong differentiation between vegetative and reproductive development stages in potato that might impose an important influence on the use of assimilate and nutrients within the plant. One study on potato in Hawaii (Manrique et al., 1991) obtained RUE values of only 1.17 g MJ-'. Manrique et al. (1991) suggested that the low RUE was associated with high vapor pressure deficit and high absolute solar radiation in the Hawaiian environment. However, radiation interception was estimated based on a

244

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

fixed extinction coefficient of 0.57 and hence it is difficult to be certain of the RUE values and the associated reasons for the apparent lower RUE. Overall, observed RUE values in potato are very high, with a likely explanation based on the high fraction of starch in the plant mass. As a basis for comparing to grain species, potential RUE for potato appears to be in the range of about 1.45 to 1.7 g MJ-'.

2. W h e a t There are more experimental reports for RUE of wheat than for any other species. Therefore, the range of locations and experimental conditions probably offers the greatest variation in growth conditions of any species. Not surprisingly, there is a wide range in the estimates of RUE (Table I). Two studies in Australia (Gregory et al., 1992; Gregory and Eastham, 1996) reported RUE values that were substantially lower than others, and may not reflect the potential RUE for wheat. Among the six other studies, the mean RUE was 1.38 g MJ-'. The two highest values of RUE were 1.51 and 1.62 g MJ- but these were based on spot measures of radiation and are likely to be overestimates. The values of 1.46 g MJ-I (Yunusa et al., 1993) based on continuous measures of radiation may more accurately reflect potential RUE for wheat. A number of factors were found to result in variation in potential RUE. One of the more important variations was the stage of crop development.As discussed later, during early vegetative growth and during reproductive growth, RUE was lower than during the middle and late stages of vegetative growth (Green, 1987; Fischer, 1993; Calderini et al., 1997). Cultivar differences in RUE have also been reported. Yunusa et al. (1993) compared three cultivars with a wide range of year of release and found that the most recently released cultivar had the greatest RUE. On the other hand, Calderini et al. (1997) compared seven cultivars released over the period from 1920 to 1990 and found no trend in a change in RUE with year of release.

',

3. Barley The three RUE values for barley (Table I) tend to be somewhat lower than those reported for wheat. However, the two reports with the lowest values of RUE for barley (Gregory et al., 1992; Jamieson et al., 1995) both included a comparison with wheat, and the RUE of barley was equivalent to or greater than that of wheat. Therefore, it seems unlikely that the RUE of barley is inherently inferior to that of wheat. The study of Goyne et al. (1993) reported the highest value of RUE for one cultivar of barley of 1.30 g MJ-', which is comparable to several studies with wheat. RUE of barley appeared to be reasonably stable throughout the growing season. No variation in RUE through the growing season was indicated in the study of

RADIATION USE EFFICIENCY

245

Goyne et af. (1993). On the other hand, Gregory et al. (1992) showed a decrease in RUE after flag leaf emergence as compared to before flag leaf emergence. Little variation in RUE among barley cultivars has been identified. RUE was equivalent among cultivars in the study of Gregory et al. (1992) and in the study of Goyne et af. (1993), except for the superiority of the cultivar Gilbert.

4. Rice Only two references could be identified for rice that included both measurements of radiation interception and biomass accumulation (Table I). A partial explanation for such a limited number of investigations may be the difficulty of measuring intercepted radiation in a paddy field. That is, measurement of radiation transmittance under a crop canopy at the water surface in a paddy field is more difficult than positioning an instrument on a soil surface. In any event, the maximum RUE value of 1.39 g MJ- reported by Horie and Sakuratani (1985) compares favorably with the RUE values obtained for other cereal crops. There was substantial consistency in RUE among cropping seasons in the study of Horie and Sakuratani (1985), although there was an indication of differences in RUE between genotypes. Kiniry et al. (1989) estimated RUE values for rice from four studies. Excluding the largest value, which was roughly twice the other three, the mean value was 1.1 g MJ-I.This estimate is consistent with the two direct measures of RUE. Nevertheless, there remains a need to document RUE fully in rice. Despite the difficulties of measuring intercepted radiation under paddy conditions, measures of RUE under a range of conditions need to be obtained as have been done in other species.

'

5. Sunflower

The mean value of RUE from the six experimental reports for sunflower (Table I) was 1.27 g MJ-I,the highest value being 1.56 g MJ-'measured in Argentina (Trapani et al., 1992) during the period of rapid growth before anthesis. No difference was found among the tested cultivars. Also, Bange et al. (1997a) observed RUE values in sunflower of 1.6 g MJ- or greater in individual replicates. There is no obvious explanation for these high values of RUE obtained during the period of rapid vegetative growth. Considerable variation in RUE for sunflower during the growing season was found in several studies. Trapani et al. ( I 992) found-thatRUE from emergence to a leaf area index of 1.7 was 45% of that from leaf area index of 1.7 to anthesis. Also, they reported that RUE during the postanthesis phase was 43% of the preanthesis stage. Similarly, Gimenez et af. (1994) and Bange et af. (1997a) obtained low RUE during the early periods of sunflower development. The most detailed description of the changes in RUE through the growing season was presented by

246

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

Hall et al. (1995). Their data clearly indicated a decreased RUE during crop establishment and during the later stages of seed fill. The high seasonal variation in RUE in sunflower might, consequently, explain some of the variation in RUE reported among the various investigations. Nitrogen fertilization of sunflower has also been considered for its influence on RUE. The sensitivity to the application of nitrogen has been variable and may reflect the original nitrogen status of the soil. Hall et al. (1995) and FlCnet and Kiniry (1995) found little influence of nitrogen application on RUE. On the other hand, Gimenez et al. (1994) and Bange et al. (1997a) found substantially decreased RUE in treatments of no nitrogen fertilization. Nitrogen availability also appears to be an important source of RUE variation among experiments.

6. Soybean RUE for soybean (Table I) tends to be lower than that reported for species that have already been reviewed. The mean maximum RUE value in the six studies is 1.02 g MJ-'. Curiously, the two highest RUE values (1.26 and 1.15 g MJ- I ) were both reported from Japan. The basis for the overall lower RUE values reported for soybean as compared to other C, species seems to result from two factors. First, the energy content of the constituents of the soybean plant, particularly in the seed, necessarily results in a decreased RUE (Fig. 1). The photosynthate costs for nitrogen accumulation by either symbiotic nitrogen fixation or nitrate reduction are high. Second, the maintenance of a high leaf photosynthetic rate in soybean to sustain a high RUE may be difficult because of the high nitrogen requirements in the leaves (Sinclair and Horie, 1989). Within individual studies a fairly high level of stability in RUE throughout the growing season has, however, been found. Plotting cumulative plant mass and radiation interception through the season has not indicated large changes in RUE. Only the detailed study of Rochette et al. (1995) indicated that RUE was low early in the season (LA1 < 2), and then remained fairly stable after that. Interestingly, including roots in the estimate of RUE, as in the study by Rochette et al. (1993, increased the seasonal estimate by only 8%.

7. Peanut Agreement in seasonal RUE among the six experimental reports for peanut is quite high (Table I). The range of RUE from these experiments is only from 0.98 to 1.12 g MJ-' with a mean value of 1.05 g MJ-'. These RUE values are more similar to those reported for soybean than for the other C, species. An especially important factor in evaluating the RUE of peanut is correction of the mass during seed fill for the energy content of the seed. Bell et al. (1987) and

RADIATION USE EFFICIENCY

247

Bennett et al. (1993) showed the sharp decrease in RUE during seed fill if the plant mass during this period is not corrected for seed energy content. When a correction is done for energy content, then RUE remains constant through much of the season. Energy content of the plant is also likely to be important in explaining the relatively low RUE of peanut. The vegetative component of the peanut plant is high in protein content, which results in a high requirement for assimilate and nitrogen input. The importance of nitrogen accumulation in determining peanut RUE was shown directly in the results of Wright et al. (1993) where RUE in a warm climate was closely linked to the leaf nitrogen content. A great deal of stability in RUE has also been found among peanut cultivars. No differences among cultivars were detected in comparisons of two cultivars (Bell et al., 1992), of three cultivars (Wright et al., 1993), and of four cultivars (Bennett et al., 1993). Bell et al. (1994c), however, did find that the cultivar Chico had a significantly lower RUE than the five other tested cultivars. Plant density also did not result in variation in RUE (Bell et al., 1987).

8. FabaBean The three reports of RUE in faba bean give highly divergent results (Table I). The values of 1.03 g MJ-' (Silim and Saxena, 1992) and of 1.45 g MJ- (Madeira et al., 1994) are in the range of RUE values reported for other C, species. In contrast, the RUE estimate presented by Fasheun and Dennett (1982) converts to 2.04 g MJ-'. This extremely high value for RUE was based on plant samples of only three plants, which may have resulted in an overestimation of plant mass per unit area. The calculated crop growth rate of 30 g m-* day-' obtained in this study also seems to be a very high estimate. Clearly, additional data are needed to resolve the potential RUE in faba bean.

9. Other Grain Legumes Overall, the values of RUE for a number of grain legumes (Table I) are consistent with the range of values obtained in soybean and peanut. The maximum RUE obtained in these additional studies was 1.46 g MJ- reported for pea (Heath and Hebblethwaite, 1985), followed by 1.09 g MJ-' reported for pigeon pea (Hughes et al., 1981) and for cowpea (Muchow et al., 1993). In a number of cases, RUE values were reported to be less than 0.8 g MJ-I. Such low estimates of RUE in many of these grain legumes seems inconsistent with the low energy content (i.e., high carbohydrate levels) of their grain (Sinclair and DeWit, 1975). Based on the energy content, it seems many grain legumes should have RUE values that are substantially greater than those of soybean and peanut. The experimental results to date indicate a need for intense studies of RUE in a

'

248

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

number of these grain legumes. Important questions exist concerning whether RUE is inherently low within some of these grain legume species, or whether the cultivars that have been tested or the environment of the experiments limited the expression of a higher RUE. Compared to soybean and peanut, it appears that there is an opportunity to increase the RUE of some of these other grain legume species.

V. SOURCES OF VARIABILITY IN RUE Both the theoretical analyses and the experimental results clearly demonstrate that there is not a constant universal RUE value. The theoretical analyses indicate several factors that could be important in causing variation in the RUE that is achieved. In contrast to the conclusion of Demetriades-Shah et al. (1992), the variation in RUE has been found to be quantitatively linked to plant traits and environmental conditions. The intent of this section is to examine the important factors associated with variability in RUE.

A. SPECIES Table I and the discussion in the previous section indicate important variations among crop species. While variations in measured RUE within a species are noted, to a great extent there is reasonable consistency in the estimates of RUE within a species. C, species tend to have the higher RUE values, with sugarcane being the highest of all species with RUE values approaching 2.0 g MJ-' (Table I). Sugarcane is a good candidate for maximum RUE values because of its use of the C, photosynthetic pathway and its production of sucrose as the final plant product. Maize has RUE values somewhat lower than those of sugarcane, in the range 1.5to 1.7 g MJ- (Table I). Surprisingly, sorghum has fairly modest RUE values. The reason for the lower values of sorghum within the C, species is unknown, but the theoretical studies indicate the importance of the expression of leaf photosynthetic rate on RUE. Muchow and Sinclair (1994) found that the lower value of RUE in sorghum compared to maize was associated with a lower leaf nitrogen content. An important challenge is to understand the low values of RUE obtained for sorghum relative to other C, species. Important differences in experimental measures of RUE also exist among the C, species. Clearly, potato has RUE values substantially greater than any other C, species, in the range 1.6 to 1.75 MJ-' (Table I). As discussed previously, the fact that much of the plant product in potato is starch allows a higher accumulation of plant mass per unit of photosynthate than in most other species. The cereals (wheat, barley, and rice) seem to have similar RUE values in the

2 49

RADIATION USE EFFICIENCY

range of about 1.3 to 1.5 g MJ-I, as originally proposed by Monteith (1977). One important outcome of this review, however, is the fact that very few experiments with measures of intercepted radiation have reported RUE for rice. Considering its importance, it seems that there is a need for more extensive measurements of RUE in rice. Among those species that produce energy-rich seed components, sunflower has been found to have high RUE values, although the highest reported values for RUE in sunflower are restricted to the vegetative stage of plant development. Trapani et al. (1992) found potential RUE during vegetative growth to be 1.56 g MJ- which is equal to or greater than those of the C , cereal crops. Also, Bange et al. (1997a) presented data from individual replicates in which RUE was 1.6 g MJ-' or greater. These high RUE values during vegetative growth for a C , species are particularly intriguing. There appears to be no ready explanation for these high RUE values, which indicates a need for more intensive investigation. Over the whole season, however, sunflower is reported to have RUE in the range 1.25 to 1.35 g MJ(Table I). Soybean and peanut have RUE values lower than that of sunflower, approximately 1.05 g MJ-' 10% (Table I). It is likely that the lower RUE value for soybean and peanut is due to the high energy content of the constituents of both vegetative tissues and seeds. In particular, both of these species are composed of tissues with high protein contents. Grain legumes other than soybean and peanut could be expected to have higher RUE values than these two species because other grain legumes tend to produce plant material of lower energy content. Surprisingly, this is not the case, with most measures of RUE in these other grain legumes being less than 1.O g MJ- * (Table I). Only the experiment of Heath and Hebblethwaite (1985) with pea resulted in a large estimate of RUE (1.46 g MJ-') that is comparable to the C , cereals. Overall, these results indicate a potential for increasing RUE in many of the grain legumes. At the least, additional research would be helpful to document and to develop a better understanding of their apparently low RUE values. Only a few experiments have directly compared differences in RUE between C, and C , species. Of course, an inherent difficulty in such species comparisons is the possibility that all species in the comparison may not be fully adapted to the environmental conditions of the test. As a consequence, a particular species may not be able to fully express its potential RUE. Nevertheless, such comparisons generally have produced results that are consistent with comparisons across experiments. Inthapan and Fukai (1988) compared the RUE of rice against that of maize and sorghum. RUE values of maize and sorghum were equivalent, but the RUE value for these two C, species was 42% greater than that of rice. In a comparison of maize and soybean, Daughtry et al. (1992) found that RUE was 1.81 and 0.99 g MJ-I, respectively. While this RUE for maize is a bit higher than that i.n many

',

'

*

250

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

studies, this difference between species seems to be fully consistent with other data. Rudorff et al. (1996) compared wheat and maize, but unfortunately the study was done in open-top chambers. Particularly for wheat, the potential for altered radiation environment because of the plastic enclosure could have resulted in the measurement of an abnormally high RUE of 1.71 g MJ-' for wheat. Consequently, Rudorff et al. (1996) detected no difference in the RUE between wheat and maize. Only two direct comparisons have been made between species that have plant tissue differing in biochemical compositions.Muchow et al. (1993) compared the RUE of soybean, mung bean, and cowpea and demonstrated the importance of the difference in the energy content of soybean relative to the other two species when comparing the measured RUE values. A comparison of wheat and lupin was done by Gregory and Eastham (1996). While the RUE of both species tended to be low in this particular study, wheat had a maximum RUE that was 45% greater than that of lupin. This difference in RUE between species, however, is larger than expected based on biochemical composition alone. In summary, important differences in RUE have been confirmed among crop species. There is a tendency for potential RUE in C, species to be greater than those in C , species as indicated in the theoretical analyses. There are, however, overlaps between C, and C, species so that the clear distinction as identified by Gosse et al. (1986) has important exceptions.An important contributor to the variation in RUE species is differences in the energy content of the biochemical constituents of the plant products. For example, sugarcane and potato have low energy products (sucrose and starch, respectively) that appear to result in exceptionally high RUE compared to other species with the same photosynthetic capabilities. Therefore, it is necessary to define separately a potential RUE for each crop species.

B. CO, A~SIMILATION RATE Theoretical analyses have all shown a clear dependence of RUE on leaf CO, assimilation rate (e.g., Fig. 1). Increasing leaf photosynthesis rate is directly linked to increasing RUE, although the response is curvilinear, approaching a maximum RUE at high photosynthesis rates. Due to the difficulty in experimentally relating CO, assimilation to RUE, only a few studies have generated data to examine this relationship. Bennett et al. (1993) measured the leaf CO, exchange rate through the growing season for peanut in conjunction with RUE measures. Both CO, exchange rate and RUE were found to be essentially constant through the bulk of the growing season. Rochette et al. (1995) did a careful examinationof the link between canopy CO, assimilation rate and RUE for a soybean crop. Canopy CO, flux density was mea-

RADIATION USE EFFICIENCY

251

sured hourly by the eddy correlation technique over 60 days during the growing season. Daily values of net CO, exchange were linear for daily intercepted PAR radiation between 4 and 11 MJ mP2 day-'. Use of the linear relationship over the season introduced only a 4% error compared to use of a nonlinear relationship. Hence, the basic premise of RUE that a linear relationship exists between CO, assimilation and light interception was confirmed. Further, Rochette et al. (1995) calculated RUE for intervals through the growing season by summing the daily measures of CO, assimilation and light interception, and converting to an estimate of dry weight accumulation. They found that the data obtained by direct plant harvest agreed to within about 10% of the CO, assimilation estimate of RUE over the periods in the season when the LA1 was greater than 2.

1. Nitrogen Response Interestingly, the more exhaustive examination of the influence of CO, assimilation influence on RUE has resulted from the linkage between leaf CO, assimilation rate and leaf nitrogen content. In their review, Sinclair and Horie (1989) concluded that variation in leaf CO, was commonly associated directly with changes in leaf nitrogen content per unit leaf area. The theoretical study of Hammer and Wright (1994) provided a detailed analysis of the linkage between RUE and leaf nitrogen content. A number of experimental studies have examined the response of RUE to leaf nitrogen content. In soybean, Sinclair and Shiraiwa (1993) found a positive, curvilinear relationship between RUE and specific leaf nitrogen within each of their two experimental locations. The mean canopy leaf nitrogen content in each case was fairly low (less than 1.6 g mP2). In contrast, in peanut Wright et al. (1993) found little variation in RUE with specific leaf nitrogen except for a nonnodulating line grown in a warm environment. The specific leaf nitrogen content in their study was, however, generally high (greater than about 1.5 g mP2) in those situations where RUE was essentially constant. Therefore, these two contrasting studies are consistent with the theoretical conclusion that there is a saturating response in RUE to increasing leaf nitrogen content (Fig. 2). In sunflower, Gimenez et al. (1994) found that increased soil nitrogen fertility resulted in both an increased specific leaf nitrogen content and increased RUE. In a more detailed study with sunflower, Hall et al. (1995) measured average canopy leaf nitrogen content and RUE for seven periods during the growing season for two levels of soil nitrogen fertility. A curvilinear relationship between RUE and leaf nitrogen content was obtained when using all the data collected through the season, except for the first harvest of the season when RUE values were low. For much of the season RUE was essentially constant, but the last two harvests during seed-fill had both decreased RUE and decreased leaf nitrogen content. Bange et

252

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

al. (1997a) undertook a similar study in sunflower with essentially the same results. In the period from bud visible to anthesis, leaf nitrogen content was high (greater than 1.5 g mP2) and RUE was fairly stable. Following anthesis when leaf nitrogen content decreased, RUE was also observed to decrease. The combined data of Bange et al. (1997a) revealed a curvilinear increase in RUE with leaf nitrogen content. In contrast to the above studies, in sunflower FlCnet and Kiniry (1995) found no changes in RUE among three levels of soil nitrogen applications during most of the growing season. Only during seed growth, when there was no further accumulation of plant mass in the 0 N application treatment, was there a difference in RUE among the treatments. Muchow and Davis (1988) subjected maize and sorghum to a range of nitrogen fertility treatments. During the later stages of vegetative growth, they found a close correlation between RUE and average canopy leaf nitrogen content per unit leaf area. A common linear relationship fitted both species over the range of leaf nitrogen values observed. Muchow and Sinclair (1994) examined the relationship between RUE and leaf nitrogen content in maize and sorghum over a wider range of leaf nitrogen content per unit leaf area. Their results showed a nonlinear, saturating response of RUE to increasing leaf nitrogen. Consistent with these results, Fischer (1993) was able to develop relationships for wheat between RUE and the percent N concentration in the tops of the plants for individual growth periods. Nearly all results, therefore, indicated that RUE achieves a saturated value at high leaf nitrogen contents and decreases curvilinearly with decreasing leaf nitrogen content below the saturating leaf nitrogen content. It is not surprising then that investigations of RUE under treatments of differing soil nitrogen fertility have shown important variations in RUE (Fig. 3). These observations have been done in a number of species, including wheat (Green, 1987; Garcia et al., 1988; Fischer, 1993), sunflower (Gimenez et al., 1994; Hall et al., 1995; Bange et al., 1997a), and maize and sorghum (Muchow and Davis, 1988; Muchow, 1994).

2. Drought Response Soil water deficits can have a major influence on leaf photosynthesis, and, consequently, it is hypothesized that RUE is also decreased under drought conditions. It seems likely that some of the studies that have reported decreased RUE may have been a result of periods of soil water deficit. A few studies that made direct comparisons of RUE under well-watered and drought-stressed conditions have been reported. Muchow (1985) compared the influence of soil water deficits on RUE measured in several grain legumes. Drought stress in all cases resulted in a decreased RUE relative to the well-watered treatments. The effect of the drought stress was particularly marked in a treatment when the crops were established with irrigation for the first 6 weeks and then received no further irrigation. In all comparison the

RADIATION USE EFFICIENCY

253

drought-stressed treatments had RUE values that were only one-third or less of the controls. Inthapan and Fukai (1988) measured the RUE of rice, maize, and sorghum subjected to irrigated and water-deficit treatments. All three crops had a sharp decrease in RUE in the water-deficit treatment, with rice having the greatest reduction. Muchow (1989b) compared the RUE of maize, sorghum, and pearl millet in a drought study involving three experiments. Only maize consistently had RUE decreased in the water-deficit treatment of each experiment. In two of the experiments, RUE of sorghum and pearl millet was found not to be statistically decreased by the water-deficit treatments. These contrasting responses in RUE to drought stress were interpreted by Muchow (1989b) to indicate differences in the rate of water use from the soil and the severity of drought that was ultimately imposed on each species. The fact that RUE did not decrease in an experiment even though irrigation was withheld was concluded to indicate that a significant soil water deficit had actually not developed. Jamieson et al. (1995) also found that the timing and duration of a drought treatment were important in interpreting RUE. They compared the RUE of barley subjected to 12 irrigation treatments. RUE was decreased in those treatments with early drought such that there was a linear decline in RUE with a water deficit calculated from potential evapotranspiration. Those treatments designed to impose drought stress in middle or late periods during the growing season did not exhibit a decrease in RUE. Unfortunately, soil water content was not measured directly in this experiment so variation in the severity of the soil water deficit could not be compared among the treatments. Singh and Sri Rama (1989), however, did an especially useful RUE experiment in which they measured soil water content of irrigated and nonirrigated plots of chickpea. They generated a relationship between RUE and fraction of extractable soil water whereby RUE was independent of fraction of extractable soil water when extractable soil water was greater than about 30%. At fraction of extractable soil water of less than 30%, however, there was a marked decline in RUE that was represented by an exponential equation. These results, therefore, demonstrated that the response of RUE to water deficits was quantitatively dependent on the severity of the soil water deficit.

3. Other Environmental Factors RUE seems to be fairly stable across environments under optimal growing conditions. For example, Muchow et al. (1993) concluded that the RUE values measured for three grain legumes in two locations were similar in spite of the fact that the environments varied in terms of temperature, solar irradiance, and vapor pressure. Nevertheless, there are circumstances where direct influences of the environment on leaf CO, assimilation rate have been linked to changes in RUE.

2 54

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

The effect of temperature on the photosynthetic activity of leaves has the potential to result in altered RUE in some species. Andrade et al. (1992) concluded that low RUE values for maize grown at Balcarce, Argentina, were a result of low temperature. In additional experiments with maize using varying sowing dates, Andrade et al. (1993) found a linear decrease in RUE associated with a decrease in mean temperature from 21 to 16°C. They found variation in RUE during vegetative growth of 1.05 to 1.52 g MJ-' over five years that was associated with yearly differences in temperature. The influence of temperature on RUE is especially well documented for peanut. The leaf photosynthetic rate of peanut decreases in an approximately linear relationship with night temperature lower than 16°C (Sinclair et al., 1994; Bell et al., 1994a,b). This response is consistent with the decrease in RUE associated with minimum daily temperature measured at a number of locations, in that there was a linear relationship between the decrease in RUE and night temperature of less than about 20°C (Bell ef al., 1992).In addition, differences in peanut RUE between two locations in Australia (Wright et al., 1993) and between two years in Canada (Bell et al., 1994c) have been attributed to differences in the night temperature environment. Vapor pressure deficit has been indicated as having the potential for influencing RUE (Stockle and Kiniry, 1990). To the extent that a large vapor pressure deficit may result in decreased photosynthetic rates, then a decrease in RUE would be expected. The influence of vapor pressure deficit on leaf photosynthetic rate tends, however, to develop at fairly high vapor pressure deficits (greater than 2 Wa), but even then the decreases in photosynthetic rate are fairly modest (Fig. l b of Stockle and Kiniry, 1990). Nevertheless, Stockle and Kiniry (1990) presented regressions for sorghum and maize using a number of data sources for RUE showing a decrease in RUE with increasing vapor deficit. Their analysis indicated a large effect that was greater than predicted by leaf behavior. Since decreasing vapor pressure deficit is likely to be associated with a number of other environmental variables (e.g., level of solar radiation and fraction of diffuse radiation), a simple regression of RUE against vapor pressure deficit is confounded. In comparing RUE of barley between two seasons, Goyne et al. (1993) reached a conclusion concerning the influence of vapor pressure deficit on RUE similar to Stockle and Kiniry (1990). In one season of 1.06 kPa vapor pressure deficit a RUE of 0.67 g MJ-' was observed, while in another season of 0.68 kPa the RUE was 1.30 g MJ-I. Again, the large difference in RUE between seasons was much greater than anticipated response in leaf photosynthesis at these low values of vapor pressure deficit. Other factors seem likely to have contributed to the large differences in RUE observed between seasons. In contrast to the above, Muchow and Sinclair (1994) found no variation in RUE with vapor pressure deficit for maize and sorghum across six environments of differing vapor pressure deficit. Consideringthe reasonable stability observed in most

RADIATION USE EFFICIENCY

255

crops across environments, it seems likely that vapor pressure deficit has only a small influence on RUE in most cropping situations. Other factors that inflict damage on leaves have also been shown to decrease RUE. Infection of faba bean plants with Ascochyta fabae resulted in decreased RUE (Madeira et al., 1994). Exposure of leaves to ozone has been found to be associated with decreases in RUE of soybean (Leadley et al., 1990) and of wheat and maize (Rudorff et al., 1996).

C. SEASONAL VARIATION Considering the importance of leaf photosynthetic rates on RUE, variation in photosynthetic capacity through the growing season would be anticipated to have an important influence on variation in RUE. In a number of cases, maximum RUE values are reported for the vegetative period when it is anticipated that the production of new leaves results in much of the radiation intercepted by young leaves with relatively high photosyntheticcapacity. There have been, in fact, several studies that have examined specifically the question of variation in RUE through the growing season. Two periods that potentially may have relatively lower RUE because of diminished leaf photosynthetic capacity are during crop establishment and during seed growth.

1. Crop Establishment Early in crop establishment, when the first leaves may have decreased photosynthetic capacity relative to later leaves, there exists the possibility that RUE may be decreased. This hypothesis is, however, difficult to study experimentally. Substantial errors are possible in measuring radiation interception in crop canopies composed of small plants. The plants can be widely dispersed, and it is difficult to position sensors under the leaf canopy when the leaves are developing near the soil surface. Nevertheless, several studies offer observations on RUE in the early stages of crop development. In a study of wheat RUE, Garcia et al. (1988) found that RUE was lower for the period from double ridge to terminal spikelet than for periods later in the growing season. Similarly, data presented by Fischer (1993) for the wheat cultivars with the highest nitrogen content revealed that growth intervals early in crop development had RUE lower than that achieved later in the season. In sunflower, Trapani et al. (1992) found a lower RUE during the emergence phase than during the period following this period. The point where there was an increase in RUE occurred when LA1 reached about 1.7. They attributed the difference in RUE between the two development periods to differences in leaf photosynthesis capability.Gimenez et al. (1994) also found a lower RUE in sunflower

256

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

during the period from 9 to 42 days after sowing compared to that during the periods including 43 to 7 l days after sowing. These results were further confirmed by Hall et al. (1995), who found that the first harvest period for sunflower yielded a RUE lower than that measured in subsequent harvest periods prior to seed growth. Similar to other species, Rochette et al. (1995) found for soybean that during the early stages of crop development when LA1 was less than 2, RUE was lower than that achieved later in growth. They attributed this lowered RUE to a lowered leaf photosynthetic capacity.

2. SeedGrowth The dynamic interaction between nitrogen storage in leaves and the development of seeds commonly leads to a decrease in leaf nitrogen through the seed growth period. The close linkage between leaf nitrogen and photosynthetic capacity, consequently, raises the important possibility that RUE will decrease during the seed growth period. Fortunately, there have been a number of studies that have segregated RUE data for the vegetative and reproductive growth periods so that the effect of this hypothesis can be examined. Muchow (1985) found that RUE from emergence to 41 days after sowing in six grain legumes was greater than that from 42 days after sowing to maturity. The decrease in RUE was attributed to a loss in leaf photosynthetic activity during grain fill for each of these crops. On the other hand, the results of Leadley et al. (1990) revealed that RUE of soybean grown in open-top chambers was greater during reproductive development than during vegetative development. Lower RUE during the postanthesis period relative to the rapid growth phase prior to anthesis was observed in sunflower by Trapani et al. (1992). Subsequently, Hall et al. (1 995) also reported for sunflower a dramatically decreased RUE at the end of the seed growth period. Bange et al. (1997a) confirmed these previous observations and related the decrease in RUE during seed growth to a decrease in leaf nitrogen content and decreased leaf photosynthetic activity. In maize and sorghum, there are several studies that clearly demonstrated a decline in RUE following anthesis (Muchow and Davis, 1988; Muchow 1989a, 1994). This is illustrated in Table I where much higher RUE values are reported for vegetative growth than for seasonal growth. Calderini et al. (1997) found that RUE for the postanthesis period in wheat was substantially lower than RUE measured for the preanthesis period.

D. RADIATIONENVIRONMENT Theoretical analyses indicate that differences in RUE may result from differences in the radiation environment. Hammer and Wright (1994) concluded from

257

RADNTION USE EFFICIENCY

such a theoretical analysis that the radiation level on a cloudy as compared to a clear day could result in a 0.4 g MJ-' increase in RUE. An increasing fraction of diffuse radiation on a cloudy day contributed a 0.15 g MJ- I increase in RUE. There are several experimental studies that offer direct evidence concerning the influence of the radiation environment on RUE. Horie and Sakuratani (1985) imposed shading treatments on rice that resulted in shading of 46 and 72%. The resulting RUE values were 1.26 g MJ-' for the unshaded treatment, 1.66 g MJ-' for the 46% shade, and 1.98 g MJ-' for the 72% shade. In an experiment with peanut, Stirling et al. (1990) positioned bamboo weave over plots to give about a 75% shade throughout reproductive development. As a result of this severe shading, the RUE increased from 0.98 g MJ-' in the unshaded treatment to 2.36 g MJ-' for the shaded treatment. Consequently, in experiments with both rice and peanut, severe shading resulted in dramatic increases in RUE. Bange et al. (1997b) placed plastic films over sunflower giving differences in the radiation environment that more realistically matched natural variations in the radiation environment. One of the plastic films resulted in a radiation level that was 86% of incident radiation and an increase in the proportion of diffuse radiation relative to unshaded conditions of 14% on clear days. The second plastic resulted in a radiation level of 80% of incident radiation and a 13% increase in diffuse radiation. In these treatments of fairly modest adjustments in the radiation environment, RUE was increased only by 0.15 and 0.19 g MJ- (1 1 and 14%) in the two treatments, respectively. Plants grown in controlled environments have also resulted in unusually high estimates of RUE. Hammer and Vanderlip (1989) measured RUE of sorghum grown in a greenhouse at 2.16 g MJ-I, which substantially exceeded all other sorghum measures (Table I). Similarly, Rudorff et al. (1996) reported a RUE of 1.71 g MJ- for wheat grown in open-top chambers, which is much greater than other measures of RUE for wheat (Table I). The decrease in total radiation and increase in the proportion of diffuse radiation resulting from the surrounding greenhouse or chamber structure are likely to have been important factors in contributing to elevated RUE in these studies. Overall, the theoretical and experimental results indicate that variations in the natural radiation environment can have an important influence on RUE. In particular, substantial decreases in the total radiation level can result in increased RUE values. Since decreased radiation is almost always associated with an increased proportion of diffuse radiation, the increase in RUE is further enhanced. In the natural environment, however, daily variations in irradiance are averaged by the plant over long time periods so that the small daily variations induced in RUE are not apparent. Only when comparing large, sustained differences in irradiance level are there likely to be differences in RUE. This possibility was examined by Sinclair and Shiraiwa (1993) in a comparison of RUE for soybean measured in Japan and in Florida. The lower radiation environment in Japan, and apparently greater dif-

'

'

258

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

fuse component,compared to Florida was offered as the explanationfor the greater RUE measured for soybean in Japan.

VI. CONCLUSIONS In little more than 20 years, radiation use efficiency has become an important, if not essential, approach for understanding crop growth and yield. Because crop yield is linked directly to the capacity of the plants to use the intercepted solar energy in the accumulation of crop mass, RUE provides the measure that directly reflects the efficiency in the use of radiant energy. Considerable confusion has resulted about the RUE concept simply because a number of experimental approaches and units of expression have been presented. As with all experimental results, considerable care is needed to ensure the collection of data with a minimum of error and bias. Among some of the important considerationsin various experimental approachesare direct measures of radiation interception and adequate sampling. considerable caution is needed in considering RUE estimates that are not obtained from primary data. Periodic sampling through the growing season is important so that data are available to do regression analyses and that changes in RUE through the growing season can be detected. The results of both theoretical analyses and experiments are consistent about the major factors that influence RUE. Certainly, RUE varies among species because of differences in the biochemical components of the plant products and in photosynthetic capacity. Among species with similar photosynthetic rates, those plants that produce energy-rich plant products have lower RUE because of the limitation on the production of plant mass. Those species with high leaf photosynthetic activity, especially species with the C, photosynthetic pathway, tend to have higher RUE. There are, however, important exceptions to a clear distinction in RUE between C, and C, species. Sorghum is commonly found to have RUE that is substantially lower than other C, species, with RUE values in the same range as those for C , cereals. In addition, potato and sunflower during vegetative development have been found to have high RUE that are equivalent to maize. These exceptions to a clear distinction in RUE between C, and C, species may be especially important in gaining additional insights about factors influencing the expression of RUE. An important use of the RUE concept is to understand those circumstances in which RUE may not match the expected potential RUE value. Certainly, theoretical analyses have highlighted the importance of decreased photosynthetic activity resulting in decreased RUE. A number of factors, therefore, result in decreased RUE, including water deficits, lowered leaf nitrogen content, and temperature. Changes in leaf photosynthetic capacity through the life cycle of the crop can have a direct influence on changes in RUE. In addition, the environment can have im-

RADIATION USE EFFICIENCY

2 59

portant consequences on RUE other than the direct influence on photosynthesis. The radiation environment in particular has been shown to influence RUE. Although not usually a major factor, some of the variation in RUE among various locations may be a consequence of differences in the radiation level or the amount of diffuse radiation. In general, RUE is an independent measure that can be used to benchmark crop performance and highlight yield limitations. RUE is a valuable approach for interpreting large variations in crop yield from season to season and across locations resulting from climatic variations, particularly cumulative solar radiation incident to a crop. Estimates of RUE can also identify crop management and husbandry limitations and can be used to assess the scope for yield improvement in different cropping systems. An important context in which to assess environmental factors on crop growth and RUE is crop models that are based on RUE. The use of RUE in models offers a relatively simple, mechanistically based description of the key factors influencing the accumulation of crop mass. For example, Sinclair (1986) presented one of the first crop models that was fully based on RUE. The model was designed to simulate soybean development and growth by assuming a constant potential RUE of 1.2 g MJ-'. Importantly, the actual RUE in the model was computed daily from potential RUE multiplied by functions that decreased RUE based on simulated leaf N per unit leaf area and soil water content. Hence, the influences of inadequate N or water-deficit stress on RUE were simulated in the model and offered a basis for assessing experimental results (Muchow and Sinclair, 1986). We anticipate that RUE will continue to have increasing importance in the assessment of crop performance and will lead to enhanced efficiency in the conduct of field experimentation. Certainly, measurements of RUE can help to resolve whether intercepted radiation is being used by the crop at its potential efficiency, or whether other factors are limiting crop growth. Careful data collection for evaluations of RUE offers a powerful tool in further understanding of crop growth and yield.

ACKNOWLEDGMENTS The assistance of Heidi Vogelsang (CSIRO, Brisbane,Australia) and Annette Prasse (ARS-USDA, Gainesville, FL, USA) in searching the literature and in organizing and preparing this paper are gratefully acknowledged.

REFERENCES Allen, E. J., and Scott, R. K. (1980).An analysis of growth of the potato crop. J. Agric. Sci. Cambridge 94.583-606.

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW Anderson, M. C. (1971). Indirect estimates of radiation and stand structure. In “Plant Photosynthetic Production, Manual of Methods’’ (Z. Sestak, J. Catsky, and P. G. Jarvis, eds.), pp. 447-45 1. Junk, The Hague. Andrade, F. H., Uhart, S. A,, Arguissain, G. G., and Ruiz, R. A. (1992). Radiation use efficiency of maize grown in a cool area. Field Crops Res. 28,345-354. Andrade, F. H., Uhart, S. A., and Cirilo, A. (1993). Temperature affects radiation use efficiency in maize. Field Crops Res. 32, 17-25. Arkebauer, T. J., Weiss, A,, Sinclair, T. R., and Blum, A. (1994). Discussion: In defense of radiation use efficiency: A response to Demetriades-Shah et al. (1992). Agric. Foresr Mereorol. 68,221-227. Austin, R. B., Kingston, G., Longden, P. C., and Donovan, P. A. (1978). Gross energy yields and the support energy requirements for the production of sugar from beet and cane: A study of four production areas. J. Agric. Sci. 91,667-675. Bange, M. P., Hammer, G. L., and Rickert, K. G. (1997a). Effect of specific leaf nitrogen on radiation use efficiency and growth of sunflower. Crop Sci. 37, I20 1- 1207. Bange, M. P., Hammer, G. L., and Rickert, K. G. (1997b). Effect of radiation environment on radiation use efficiency and growth of sunflower. Crop Sci. 37, 1208-1214. Bell, M. J., Muchow, R. C., and Wilson, G. L. (1987).The effect of plant population on peanuts (Arachis hypogaea) in a monosoonal tropical environment. Field Crops Res. 17,91- 107. Bell, M. J., Wright, G. C., and Hammer, G. L. (1992). Night temperature affects radiation-use efficiency in peanut. Crop Sci. 32, 1329-1335. Bell. M. J., Michaels, T. E., McCullough, D. E.,and Tollenaar, M. (1994a). Photosynthetic response to chilling in peanut. Crop Sci. 34, 1014-1023. Bell, M. J., Gillespie, T. J., Roy, R. C., Michaels, T. E., and Tollenaar, M. (1994b). Peanut leaf photosynthetic activity in cool field environments. Crop Sci. 34, 1023-1029. Bell, M. J., Roy, R. C., Tollenaar, M., and Michaels, T. E. (1994~).Importance of variation in chilling tolerance for peanut genotypic adaptation to cool, short-season environments. Crop Sci. 34, 1030-1039. Bennett, J. M., Sinclair, T. R., Ma, L., and Boote, K. J. (1993). Single leaf carbon exchange and canopy radiation use efficiency of four peanut cultivars. Peanut Sci. 20, 1-5. Blackman, V. H. (1919). The compound interest law and plant growth. Ann. Bot. 33,353-360. Bolanos, J., and Edmeades, G. 0. (1993). Eight cycles of selection for drought tolerance in lowland tropical maize. I. Responses in grain yield, biomass and radiation utilisation. Field Crops Res. 31, 233-252. Burstall, L., and Harris, P. M. (1986). The physiological basis for mixing varieties and seed ‘‘ages’’ in potato crops. J. Agric. Sci. Cambridge 106,411-418. Calderini, D. F., Dreccer, M. F., and Slafer, G. A. (1997). Consequences of breeding on biomass, radiation interception and radiation-use efficiency in wheat. Field Crops Res. 52,27 1-281. Charles-Edwards, D. A., and Lawn, R. J. (1 984). Light interception by grain legume crops. Plant Cell Environ. 7,247-25 1. Cooper, J. P. (1970). Potential production and energy conversion in temperate and tropical grasses. Herbage Abst,: 40, 1-13. Daughtry, C. S. T., Gallo, K. P., Goward, S. N., Prince, S. D., and Kustas, W. P. (1992). Spectral estimates of absorbed radiation and phytomass production in corn and soybean canopies. Remote Sens. Environ. 39, 14 1- 152. Demetriades-Shah, T. H., Fuchs, M., Kanemasu, E. T., and Flitcroft, I. (1992). A note of caution concerning the relationship between cumulated intercepted solar radiation and crop growth. Agric. Forest Mereorol. 58, 193-207. DeWit, C. T. (1959). Potential photosynthesis of crop surfaces. Nerherlands J. Agric. Sci. 7 , 141-149. DeWit, C. T. (1965). Photosynthesis of leaf canopies. Agricultural Research Rep. No. 663. Institute for Biological and Chemical Research on Field Crops and Herbs, Wageningen, The Netherlands.

RADIATION USE EFFICIENCY

261

Duncan, W. G. (1971). Leaf angles, leaf area, and canopy photosynthesis. Crop Sci. 11,482-485. Duncan, W. G., Loomis, R. S . , Williams, W. A., and Hanau, R. (1967).A model for simulating photosynthesis in plant communities. Hilgardia 38, 181-205. Evensen, C. I., Muchow, R. C., El-Swaify, S. A., and Osgood, R. V. (1997). Yield accumulation in imgated sugarcane: I. Effect of crop page and cultivar. Agron. J. 89,638-646. Fasheun, A., and Dennett, M. D. (1982). Interception of radiation and growth efficiency in field beans (Viciajaba L.). Agric. Meteorol. 26,221-229. Fischer, R. A. (1993).Imgated spring wheat and timing and amount of nitrogen fertilizer. 11. Physiology of grain yield response. Field Crops Res. 33,57-80. FICnet, F., and Kiniry, J. R. (1995). Efficiency of biomass accumulation by sunflower as affected by glucose requirement of biosynthesis and leaf nitrogen content. Field Crops Res. 44, 119-127. FICnet, F., Kiniry, J. R., Board, J. E., Westgate, M. E., and Reicosky, C. (1996). Row spacing effects on light extinction coefficients of com, sorghum, soybean, and sunflower. Agron. J. 88,185- 190. Gallagher, J. N., and Biscoe, P. V. (1978).Radiation absorption, growth and yield of cereals. J. Agric. Sci. Cambridge 91,47-60. Gallo, K. P., and Daughtry, C. S. T. (1986).Techniques for measuring intercepted and absorbed photosynthetically active radiation in corn canopies. Agron. J. 78,752-756. Garcia, R., Kanemasu, E. T., Blad, B. L., Bauer, A,, Hatfield, J. L., Major, D. J., Reginato, R. J., and Hubbard, K. G. (1988). Interception and use efficiency of light in winter wheat under different nitrogen regimes. Agric. Forest Meteorol. 44, 175- 186. Gimenez, C., Connor, D. J., and Rueda, E (1994). Canopy development, photosynthesis and radiationuse efficiency in sunflower in response to nitrogen. Field Crops Res. 38, 15-27. Gosse, G., Varlet-Grancher, C., Bonhomme, R., Chartier, M., Allirand, J.-M., and Lemaire, G. (1986). Maximum dry matter production and solar radiation intercepted by a canopy. Agronomie 6,47-56. Goudriaan, J. (1982). Potential production processes. In “Simulation of Plant Growth and Crop Production” (F. w. T.Penning de Vries and H. H. van Lax, eds.), pp. 98-109. Pudoc, Wageningen, The Netherlands. Goyne, P. J., Milroy, S. P., Lilley, J. M., and Hare, J. M. (1993). Radiation interception, radiation use efficiency and growth of barley cultivars. Aust. J. Agric. Rex 44, 1351-1366. Green, C. F. (1987).Nitrogen nutrition and wheat growth in relation to absorbed solar radiation. Agric. Forest Meteorol. 41,207-248. Gregory, P. J. (1994). Root growth and activity. In “Physiology and Determination of Crop Yield” (K. J. Boote et al., eds.), pp. 65-93. Am. SOC.Agron., Madison, WI. Gregory, P. J., and Eastham, J. (1996). Growth of shoots and roots, and interception of radiation by wheat and lupin crops on a shallow, duplex soil in response to time of sowing. Ausr. J. Agric. Res. 47,427-447. Gregory, P. J., and Marshall, B. (1980). A study of an intercrop of pearl millet and groundnut with special reference to the interception of radiation and the growth of root systems, Report No. 3, pp. 11-28. ODA Microclimatology Unit, University of Nottingham School of Agriculture, Nottingham, UK. Gregory, P. J., Tennant, D., and Belford, R. K. (1992). Root and shoot growth, and water and light use efficiency of barley and wheat crops grown on a shallow duplex soil in a Mediterranean-type environment. Ausr. J. Agric. Res. 43,555-573. Hall, A. J., Connor, D. I., and Sadras, V. 0. (1995). Radiation-use efficiency of sunflower crops: Effects of specific leaf nitrogen and ontogeny. Field Crops Res. 41,65-77. Hammer, G. L., and Vanderlip, R. L. (1989). Genotype-by-environment interaction in grain sorghum. I. Effects of temperature on radiation use efficiency. Crop Sci. 29,370-376. Hammer, G. L., and Wright, G. C. (1994). A theoretical analysis of nitrogen and radiation effects on radiation use efficiency in peanut. Aust. J. Agric. Res. 45,575-589. Hanan, N. P., Prince, S. D., and BCguC, A. (1995). Estimation of absorbed photosynthetically active ra-

2 62

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

diation and vegetation net production efficiency using satellite data. Agric. Forest Mereorol. 76, 259-276. Haverkort, A. J., Uenk, D., Veroude, H., and Van De Waart, M. (1991). Relationships between ground cover, intercepted solar radiation, leaf area index and infrared reflectance of potato crops. Potato Res. 34,113-121. Heath, M. C., and Hebblethwaite, P. D. (1985). Solar radiation interception by leafless, semi-leafless and leafed peas (Pisum sarivum) under contrasting field conditions. J. Appl. Biol. 107,309-318. Hone, T., and Sakuratani, T. (1985). Studies on crop-weather relationship model in rice. ( I ) Relation between absorbed solar radiation by the crop and the dry matter production. Jpn. Agric. Meteo1~1.40~331-342. Horie, T., Ohnishi, M., Angus, 1. F., Lwein, L. G., Tsukaguchi, T., and Matano, T. (1997). Physiological characteristics of high-yielding rice inferred from cross-location experiments. Field Crops Res. 52,55-61. Hughes, G., Keatinge, J. D. H., and Scott, S. P. (1981). Pigeon pea as a dry season crop in Trinidad, West Indies. 11. Interception and utilization of solar radiation. Trop. Agric. Trinidad 58, 191-199. Inthapan, P., and Fukai, S. (1988). Growth and yield of rice cultivars under sprinkler irrigation in southeastern Queensland. 2. Comparison with maize and grain sorghum under wet and dry conditions. Aust. J. Exp. Agric. 28,243-248. Jamieson, P. D., Martin, R. J., Francis, G. S., and Wilson, D. R. (1995). Drought effects on biomass production and radiation-use efficiency in barley. Field Crops Res. 43,77-86. Jefferies, R. A., and Mackerron, D. K. L. (1989). Radiation interception and growth of irrigated and droughted potato (Solanum tubersum). Field Crops Res. 22, 101-1 12. Jong, S. K., Brewbaker, J. L., and Lee, C. H. (1982). Effects of solar radiation on the performance of maize in 41 successive monthly plantings in Hawaii. Crop Sci. 22, 13-18. Kiniry, J. R. (1994a). Discussion: A note of caution concerning the paper by Demetriades-Shah et al. (1992).Agric. Forest Meteoroi. 68,229-230. Kiniry, J. R. (l994b). Radiation-use efficiency and grain yield of maize competing with johnsongrass. Agron. J. 86,554-557. Kiniry, J. R., Jones. C. A,, O’Toole, J. C., Blanchet, R., Cabelguenne, M., and Spanel, D. A. (1989). Radiation-use efficiency in biomass accumulation prior to grain-filling for five grain-crop species. Field Crops Res. 20,5 1-64. Leach, G. J., and Beech, D. F. (1988). Response of chickpea accessions to row spacing and plant density on a vertisol on the Darling Downs, south-eastern Queensland. 2. Radiation interception and water use. Aust. J. Exp. Agric. 28,377-383. Leadley, P. W., Reynolds, J. F., Flagler, R., and Heagle, A. S. (1990). Radiation utilization efficiency and the growth of soybeans exposed to ozone: A comparative analysis. Agric. Forest Meteorol. 51,293-308. Loomis, R. S . , and Williams, W. A. (1963). Maximum crop productivity: An estimate. Crop Sci. 3, 67-12. Loomis, R. S., Williams, W. A,, and Hall, A. E. (1971). Agricultural productivity. In “Annual Review of Plant Physiology” (L. Machlis, W. R. Briggs, and R. B. Park, eds.), pp. 431-441. Annual Reviews, Palo Alto, CA. Madeira, A. C., Clark, J. A., and Rossal, S. (1994).Growth and light interception in field bean (Vicia faba) infected by Ascochyta fabae. J. Agric. Sci. Cambridge 123,225-232. Major, D. J., Beasley, B. W., and Hamilton, R. I. (1991). Effect on maize maturity on radiation-use efficiency. Agron. J. 83, 895-903. Major, D. J., Janzen, H. H., Sadasivaiah, R. S., and Carefoot, J. M. (1992). Morphological characteristics of wheat associated with high productivity. Can J. Plant Sci. 72,689-698. Manrique, L. A., Kiniry, 1. R., Hodges, T., and Axness, D. S. (1991). Dry matter production and radiation interception of potato. Crop. Sci. 31, 1044-1049.

RADIATION USE EFFICIENCY

263

Marshall, B., and Willey, R. W. (1983). Radiation interception and growth in an intercrop of pearl milletlgroundnut. Field Crops Res. 7, 141-160. Martin, I., Tenorio, J. L., and Ayerbe, L. (1994). Yield, growth, and water use of conventional and semileafless peas in semiarid environments. Crop Sci. 34, 1576-1583. McKenzie, B. A., and Hill, G. D. (1991). Intercepted radiation and yield of lentils (Lens culinaris) in Canterbury, New Zealand. J. Agric. Sci. Cambridge 117,339-346. Meek, D. W., Hatfield, J. L., Howell, T. A,, Idso, S. B., and Reginato, R. J. (1984). A generalized relationship between photosynthetically active radiation and solar radiation. Agron. J. 76,939-945. Monteith, J. L. (1965). Light and crop production. Field Crop Abstr 18,213-219. Monteith, J. L. (1972). Solar radiation and productivity in tropical ecosystems. J. Appl. Ecol. 9, 747-766. Monteith, J. L. (1977). Climate and the efficiency of crop production in Britain. Philos. Trans. R. Soc. London B 281,277-294. Monteith, J. L. ( I 978). Reassessment of maximum growth rates for C, and C, crops. Expl. Agric. 14,l-5. Monteith, J. L. (1994). Discussion: Validity of the correlation between intercepted radiation and biomass. Agric. Forest Meteorol. 68,213-220. Monteith, J. L., Gregory, P. J., Marshall, B., Ong, C. K., Saffell, R. A,, and Squire, G. R. (1981). Physical measurements in crop physiology. I. Growth and gas exchange. Expl. Agric. 17,113-126. Moon, P. (1940). Proposed standard radiation curves. J. Franklin Inst. 230,583-617. Muchow, R. C. (1985). An analysis of the effects of water-deficits on grain legumes grown in a semiarid tropical environment in terms of radiation interception and its efficiency of use. Field Crops Res. 11,309-323. Muchow, R. C. (1989a). Comparative productivity of maize, sorghum and pearl millet in a semiarid tropical environment. I. Yield potential. Field Crops Res. 20, 191-205. Muchow, R. C. (1989b). Comparative productivity of maize, sorghum and pearl millet in a semiarid tropical environment. 11. Effect of water-deficits. Field Crops Res. 20,207-219. Muchow, R. C. (1992). Effect of water and nitrogen supply on radiation interception and biomass accumulation of kenaf (Hibiscus cannabinus) in a semi-arid tropical environment. Field Crops Res. 28,28 1-293. Muchow, R. C. (1994). Effect of nitrogen on yield determination in irrigated maize in tropical and subtropical environments. Field Crops Res. 38, 1-13. Muchow, R. C., and Charles-Edwards, D. A. (1982). An analysis of the growth of mung beans at a range of plant densities in tropical Australia. I. Dry matter production. Ausr. J. Agric. Res. 33,41-51. Muchow, R. C., and Coates, D. B. (1986). An analysis of the environmental limitation to yield of irrigated grain sorghum during the dry season in tropical Australia using a radiation interception model. Aust. J. Agric. Res. 37, 135-148. Muchow, R. C., and Davis, R. (1988). Effect of nitrogen supply on the comparative productivity of maize and sorghum in a semi-arid tropical environment. 11. Radiation interception and biomass accumulation. Field Crops Res. 18, 17-30. Muchow, R. C., and Sinclair, T. R. (1986). Water and nitrogen limitations in soybean grain production. 11. Field and model analyses. Field Crops Res. 15, 143-156. Muchow, R. C., and Sinclair, T. R. (1994). Nitrogen response of leaf photosynthesis and canopy radiation use efficiency in field-grown maize and sorghum. Crop Sci. 34,721-727. Muchow, R. C., Robertson, M. J., and Pengelly, B. C. (1993). Radiation-use efficiency of soybean, mung bean and cowpea under different environmental conditions. Field Crops Res. 32, 1-16. Muchow, R. C., Spillman, M. F.,Wood, A. W., and Thomas, M. R. (1994). Radiation interception and biomass accumulation in a sugarcane crop grown under irrigated tropical conditions. Aust. J. Agric. Res. 45,37-49. Muchow, R. C., Evensen, C. 1.. Osgood, R. V., and Robertson, M. J. (1997). Yield accumulation in irrigated sugarcane: 11. Utilization of intercepted radiation. Agron. J. 89,646-652.

2 64

THOMAS R. SINCLAIR AND RUSSELL C. MUCHOW

Murata, Y. (1981). Dependence of potential productivity and efficiency for solar energy utilization on leaf photosynthetic capacity in crop species. Jpn. J. Crop Sci. 50,223-232. Nakaseko, K., and Gotoh, K. (1983). Comparative studies on dry matter production, plant type and productivity in soybean, azuki bean and kidney bean. VII. An analysis of the productivity among the three crops on the basis of radiation absorption and its efficiency for dry matter accumulation. Jpn. J. Crop Sci. 52,49-58. Norman, J. M., and Arkebauer, T. J. (1991). Predicting canopy photosynthesis and light-use efficiency from leaf characteristics. In “Modeling Crop Photosynthesis-From Biochemistry to Canopy,” CSSA Special Publ. No. 19, pp. 75-94. Am. SOC.Agron. Crop Sci. SOC.Am., Madison, WI. Otegui, M. E., Nicolini, M. G., Ruiz, R. A., and Dodds, P. A. (1995). Sowing date effects on grain yield components for different maize genotypes. Agron. J. 87,29-33. Penning De Vries, F. W. T. (1975). Use of assimilates in higher plants. In “Photosynthesis and Productivity in Different Environments” (J. P. Cooper, ed.), pp. 459-477. Cambridge Univ. Press, London. Penning De Vries, F. W. T., Brunsting, H. M., and van Laar, H. H. (1974). Products, requirements and efficiency of biosynthesis: A quantitative approach. J. Theor Biol. 45, 339-377. Penning De Vries, F. W. T., van Laar, H. H., and Chardon, M. C. M. (1983). Bioenergetics of growth of seeds, fruits and storage organs. In “Proceedings of the Symposium of Potential Productivity of Field Crops under Different Environments,” pp. 37-59. IRRI, Los Banos, The Philippines. Robertson, M. J., Wood, A. W., and Muchow, R. C. (1996). Growth of sugarcane under high input conditions in tropical Australia. I. Radiation use, biomass accumulation and partitioning. Field Crops Res. 48, 11-25. Rochette, P., Desjardins, R. L., Pattey, E., and Lessard, R. (1995). Crop net carbon dioxide exchange rate and radiation use efficiency in soybean. Agron. J. 87,22-28. Rosenthal, W. D., Gerik, T. J., and Wade, L. J. (1993). Radiation-use efficiency among grain sorghum cultivars and plant densities. Agron. J. 85,703-705. Rudorff, B. F. T., Mulchi, C. L., Daughtry, C. S. T., and Lee, E. H. (1996). Growth, radiation use efficiency, and canopy reflectance of wheat and corn grown under elevated ozone and carbon dioxide atmospheres. Remote Sensing Environ. 55, 163-173. Sands, P. J. (1996). Modelling canopy production. 111. Canopy light-utilisation efficiency and its sensitivity to physiological and environmental variables. Aust. J. Plant Physiol. 2 3 103-1 14. Shibles, R. M., and Weber, C. R. (1965). Leaf area, solar radiation interception and dry matter production by soybeans. Crop Sci. 5,575-577. Silim, S. N., and Saxena, M. C. (1992). Comparative performance of some faba bean (viciafaba)cultivars of contrasting plant types. 2. Growth and development in relation to yield. J. Agric. Sci. Cambridge 118,333-342. Sinclair, T. R. (1986). Water and nitrogen limitations in soybean grain production. I. Model development. Field Crops Res. 15, 125-141. Sinclair, T. R., and DeWit, C. T. (1975). Photosynthate and nitrogen requirements for seed production by various crops. Science 189,565-567. Sinclair, T. R., and Horie, T. (1989). Leaf nitrogen, photosynthesis, and crop radiation use efficiency: A review. Crop Sci. 29,90-98. Sinclair, T. R., and Shiraiwa, T. (1993). Soybean radiation-use efficiency as influenced by nonuniform specific leaf nitrogen distribution and diffuse radiation. Crop Sci. 33, 808-812. Sinclair, T. R., Shiraiwa, T., and Hammer, G. L. (1992). Variation in crop radiation-use efficiency with increased diffuse radiation. Crop Sci. 32, 1281-1284. Sinclair, T. R., Bennett, J. M., and Drake, G. M. (1994). Cool night temperature and peanut leaf photosynthetic activity. Proc. Soil Crop Sci. SOC.Florida 53,74-76. Singh, P., and Sri Rama, Y. V. (1989). Influence of water-deficit on transpiration and radiation use efficiency of chickpea (Cicer arietinum L.). Agric. Forest Meteorol. 48,317-330.

RADIATION USE EFFICIENCY

265

Sivakumar, M. V. K., and Huda, A. K. S. (1985). Solar energy utilisation by tropical sorghums. Agric. Forest Meteorol. 3 5 , 4 7 4 7 . Sivakumar, M. V. K., and Virmani, S. M. (1984). Crop productivity in relation to interception of photosynthetically active radiation. Agric. Forest Meteorol. 31, 13I - I4I . Squire, G. R. (1990). Dry matter production by interception and conversion of solar radiation. In “The Physiology of Tropical Crop Production,” pp. 71 -102. CAB International, Wallingford, UK. Steiner, J. L. (1986). Dryland grain sorghum water use, light interception, and growth responses to planting geometry. Agron. J. 78,720-726. Steven, M. D.. Biscoe, P. V., Jaggard, K. W., and Paruntu, J. (1986). Foliage cover and radiation interception. Field Crops Res. 13,75-87. Stirling, C. M., Williams, J. H., Black, C. R., and Ong, C. K. (1990). The effect of timing of shade on development, dry matter production and light-use efficiency in groundnut (Aruchis hypogaea L.) under field conditions. Aust. J. Agric. Res. 41,633-644. Stockle, C. 0..and Kiniry, J. R. (1990). Variability in crop radiation-use efficiency associated with vapor-pressure deficit. Field Crops Res. 25, 17I - 18I . Szeicz, G., Monteith, J. L., and dos Santos, J. M. (1964). A tube solarimeter to measure radiation among plants. J. Appl. Ecol. 1, 169-174. Thomson, B. D., and Siddique, K. H. M.(1997). Grain legume species in low rainfall Mediterraneantype environments. 11. Canopy development, radiation interception, and dry-matter production. Field Crops Res. 54, 189-199. Tollenaar, M., and Aguilera, A. (1992). Radiation use efficiency of an old and a new maize hybrid. Agron. J. 84,536-541. Tollenaar, M.. and Bruulsema, T. W. (1988). Efficiency of maize dry matter production during periods of complete leaf area expansion. Agron. J. 80,580-585. Trapani, N., Hall, A. J., Sadras, V. O., and Vilella, F. (1992). Ontogenetic changes in radiation use efficiency of sunflower (Heliunrhus unnuus L.) crops. Field Crops Res. 29,301-316. Watson, D. J. ( I 952). The physiological basis of variation in yield. In “Advances in Agronomy” (A. G. Norman, ed.), pp. 101-145. Academic Press, New York. Westgate, M. E., Forcella, F., Reicosky, D. C., and Somsen, J. (1997). Rapid canopy closure for maize production in the northern US corn belt: Radiation-use efficiency and grain yield. Field Crops Res. 47,249-258. Williams, W. A., Loomis, R. S., and Lepley, C. R. (1 965). Vegetative growth of corn as affected by population density. I. Productivity in relation to interception of solar radiation. Crop Sci. 5,211-219. Wilson, D. R., and Jamieson, P. D. (1985). Models of growth and water use of wheat in New Zealand. In “Wheat Growth and Modelling” (W. Day and R. K. Atkin, eds.), pp. 211-216. Plenum, New

York. Wright, G. C., Bell, M. J., and Hammer, G. L. (1993). Leaf nitrogen content and minimum temperature interactions affect radiation-use efficiency in peanut. Crop Sci. 33,476-481. Yunusa, I. A. M., Siddique, K.H. M., Belford, R. K., and Karimi, M. M. (1993). Effect of canopy structure on efficiency of radiation interception and use in spring wheat cultivars during the preanthesis period in a Mediterranean-type environment. Field Crops Res. 35, 113-122.

This Page Intentionally Left Blank

THEEFFECTSOF CULTIVATION ON SOILNITROGEN MINERALIZATION Martyn Silgram’ and Mark A. Shepherd2 ‘ADAS Wolverhampton “Woodthorne” Wolverhampton WV6 8TQ, United Kingdom *ADAS Gleadthorpe Meden Vale, Mansfield, Nottingham NG2O 9PF, United Kingdom

I. Introduction 11. Methods of Mineralization Measurement A. Soil Microbial Biomass B. Microbial Respiration C. Isotopic Labeling D. Temporal Changes in Soil Mineral Nitrogen E. Control Plots F. Which Method? 111. Cultivation Effects on Soil Physical Conditions A. Cultivation Techniques B. Cultivation Effects on Soil Physical Properties n! Cultivation Effects on Nitrogen Mineralization A. Interactions with Soil Texture B. Effects on Soil Fauna C. Effects on Yield D. Previous Crop and Residue Management E. Timing and Frequency of Cultivation F. Soil Mineral Nitrogen and Nitrate Leaching V. Conclusions VI. Management Implications References

I. INTRODUCTION Nitrogen (N) is one of the most important plant nutrients in arable agriculture. The nitrogen cycle in soils is largely microbially mediated, and a major compo267 Advances in Agronomy, Volume 65 Copyright 0 1999 by Academic Press. All rlghrs of reproduction in any form reserved. 0065-2113/99 $30.00

MARTYN SILGRAM AND MARK A. SHEPHERD

nent involves the transformation of organic N into plant-available mineral forms, primarily nitrate (NO,-N) and ammonium (NH,-N). The transformation from organic N into ammonium N is termed mineralization, while the oxidation of ammonium into nitrate is termed nitriJication. Ammonium levels are usually relatively low in arable soils as it is mineralization that is typically the limiting step (Wild, 1988) and hence forms the focus of this review. The opposing process of immobilization essentially involves the conversion of mineral N into organic N by microorganisms, with the balance between mineralization and immobilization processes (net mineralization) determining the effect on the magnitude of the soil mineral nitrogen (SMN) pool. Inputs of organic N to the soil N system may derive from manures, nitrogen fixation, and crop residues while mineralization-immobilization processes also involve native soil organic matter. The mineralization-immobilization balance is of pivotal importance as it controls the supply to and magnitude of the plant-available mineral N pool. Fertilizer N additions are used to supplement this plant-available N in the soil. It is recognized by many that a major limitation to improving fertilizer recommendations is our ability to determine the magnitude and dynamics of soil N supply reliably (Sylvester-Bradley er al., 1987; Jarvis et al., 1996).Thus an enhanced understanding of the factors regulating the mineralization process will enable improvements to be made to the current imprecise nature of fertilizer recommendations. In addition to such agronomic benefits, environmental concerns have arisen in recent years regarding the susceptibility of soils to nitrate leaching. Loss of nitrate from soil systems via leaching depletes soil N reserves over the long term and can therefore carry an agronomic cost as a result of the decline in underlying soil fertility. Leached nitrate also plays a contributory role in the eutrophication of freshwater bodies, while high levels of nitrates have been linked to human health issues (McLaren and Cameron, 1990). In response to concerns over nitrate concentrations in drinking water, there is now worldwide political and legislative pressure to decrease nitrate lost from the soil via leaching. To maintain such levels of control on drinking water quality, strategies are being developed aimed either at preventing nitrate inputs to water courses through improved land management and limitations on atmospheric pollutant emissions or at treating nitrate inputs once they have reached water courses (Burt et al., 1993). The UK Department of the Environment (DOE)concluded that effective land-use management and pollution abatement measures would prove to be the more economical of these two options (DOE, 1988). The impact of cultivation practices on soil physical properties (e.g., structure, water retention, aggregate stability, aeration, bulk density) and soil biochemical processes (e.g., mineralization, nitrification) is of considerable importance in the effective sustainable management of soils as an international resource. Cultivation plays an important role in influencing the physical condition of the soil: It may be

CULTIWATION EFFECTS ON SOIL N MINERALIZATION

269

used to obtain a good seedbed, to kill weeds, to undo the damage caused by previous traffic over the land, or to increase the permeability of the surface or subsoil layers, which will allow better aeration and drainage in the soil, improving root penetration and influencing water retention properties (Wild, 1988; McLaren and Cameron, 1990). La1 (1991) identified the judicious use of tillage as a powerful tool to overcome problems associated with low infiltration, surface crusting, poor drainage, soil compaction, burial of weeds and surface debris, and pest management. Physical processes that disrupt the soil structure, such as cultivation, may influence soil N mineralization and nitrification due to their effects on soil porosity, aeration, and hydraulic conductivity, and because the physical disruption may bring microbial populations into contact with fresh, previously unavailable substrate. The effects of cultivation on soil per se and its effect on the mineralization of crop residues are inexorably linked, and it is therefore difficult to discriminate between effects on the decomposition of indigenous N and crop residue N. The effects of particular cultivation techniques may differ according to soil type, previous crop, and cultivation method and timing, and literature reports have been far from consistent regarding the presence, magnitude, and persistence of any detectable effect of different cultivation techniques on N supply to succeeding crops (Balesdent et al., 1990; Radford et al., 1992; Ekeberg and Riley, 1996; Kapusta et al., 1996). Tillage may also increase the flux of CO, from soils through enhanced biological oxidation of soil carbon by increasing microbial activity (e.g., due to residue incorporation), and this has implications for fluxes of this greenhouse gas and the role of soil as a sink in the global C balance. Of particular importance to agriculture and fertilizer recommendations is the question of whether the disturbance caused by cultivation results in a change in the absolute magnitude of net N mineralization, or whether it simply modifies the temporal dynamics of the release of N into plant-available forms. For recommendations, there is also an inevitable trade-off between what is ideal in principle and what is agriculturally practical, for example, in terms of cultivation timing. Jarvis et al. (1996) highlighted the importance of time and type of cultivation on the release of nitrate from soils. They stated that there was much qualitative but less quantitative information on specific cultivation effects. However, the scope of their review was necessarily wide, so that a more detailed review of cultivation research results is justified. Our aims are to summarize available methods of measuring mineralization, assess the effects of cultivation on soil physical properties (which will influence biochemical processes), and review evidence for the effects of cultivation on N mineralization. Within the review, we aim to quantify the size (or reported ranges) of cultivation effects on the mineral N pool, including the effects on the overall magnitude and temporal dynamics of N release; to examine the reasons for cultivation apparently increasing mineralization; and to provide information on areas where further research is required.

2 70

MARTYN SILGRAM AND MARK A. SHEPHERD

11. METHODS OF MINERALIZATION MEASUREMENT A range of direct and indirect methods have been used to monitor decomposition and the release of organically bound N in soil systems.

A. SOILMICROBIAL BIOMASS The soil microbial biomass is a key feature mediating nutrient cycling in soil systems and represents an important reservoir for potentially available plant nutrients: Jenkinson (1990) considered it as the eye of the needle through which virtually all nitrogen must pass. Saffigna et al. (1989) emphasized the value of measuring C in the soil microbial biomass, defined as that living part of the soil organic matter excluding plant roots and fauna larger than amoeba i.e., >5000 pm3 (Jenkinson and Ladd, 1981), as a sensitive indicator of changes in soil organic matter following contrasting cultivation practices. However, larger organisms such as earthworms and beetles also play a vital role in nutrient cycling and can also serve as indicators of the degree of soil disturbance. Direct measurement of soil microbial biomass populations involves counting numbers and sizes of organisms and is exceptionally tedious and open to some contention as it requires assumed chemical composition and density values (Jenkinson et al., 1976). Indirect methods are more popular, one of the most frequently used being the chloroform fumigation-extraction technique (Brookes et al., 1985). This involves fumigating a soil sample with CHC 1 and comparing the N mineralized in the fumigated soil with the N mineralized in an unfumigated control. The flush in mineralization typically observed following fumigation is due to the recolonizing microbial population decomposing the cells killed by the fumigant. Assuming that 68% of the N in the original microbial biomass is mineralized (Shen et al., 1984), then the difference in N mineralized between fumigated and unfumigated soils provides a measure of soil microbial biomass N, as under most conditions the decomposability of other soil organic matter fractions is little, if at all, affected by the CHCl, fumigation (Jenkinson and Powlson, 1976). A similar method can be used as a measure of microbial biomass C from CO, release following fumigation and inoculation (Jenkinson and Powlson, 1976). However, these methods have sometimes failed to identify changes in microbial biomass C or N concentrations in spite of contrasting management regimes (Ritz and Robinson, 1988) and the techniques at best provide only a crude assessment of biomass C and N, and hence some qualitative assessment of mineralization processes. Researchers are becoming increasingly interested in biomass community structures, and the tools for studying the effects of perturbations on such structures are

CULTIVATION EFFECTS ON SOIL N MINERALIZATION

27 1

now available. For example, biomarkers such as sterols can be used to monitor fungal biomass and lipid phosphorus can monitor bacterial biomass (O’Donnell, 1997). Such novel techniques can help to increase our understanding of the soil biomass and how factors such as cultivation can affect it.

B. MICROBIAL RESPIRATION Soil respiration is the sum of all respiratory activity within the biologically active soil layers, with the primary sources of CO, evolution being microbial and root respiration. As the mineralization of organic materials is a microbially mediated process, measuring CO, evolution (and neglecting any contribution from root respiration) can serve as an indirect measure of microbial activity in response to the disturbance caused by cultivation practices. Measurements of CO, efflux from soil have traditionally been made using alkali (e.g., NaOH, KOH) traps to quantify the cumulative gas respired in a closed chamber and hence infer the size and activity of the microbial biomass. The CO, absorbed is then determined by titrating the resulting solution against a dilute acid, usually HC1. However, such chemical absorption techniques can underestimate the gas efflux and are only capable of providing a single integrated measurement. Any laboratory incubations have the advantage of allowing the researcher greater control over abiotic conditions (moisture, temperature, redox) than is possible in the field, but depend on creating an artificial environment that may mean that results bear little relationship to processes occurring under undisturbed field conditions. A novel technique, substrate induced respiration (SIR), uses patterns of utilization of contrasting C substrates to assess the functional biodiversity and activity of soil organisms (Garland and Mills, 1991; Garland, 1996). Recent research has found that differences in SIR responses between substrates gradually decline with increasing soil disturbance from pasture through ley to arable soils (Degens and Harris, 1997), with higher topsoil SIR rates (and greater microbial biomass) under minimum tillage compared with conventionally tilled soils (Kandeler and Bohm, 1996). Results suggest that differences in SIR between management regimes reflect the smaller microbial biomass in arable compared with grassland soils and arise from differences in the composition of mineralizable soil organic matter (Degens and Harris, 1997). However, the range of microorganisms cultured in this technique can be much smaller than the whole soil microbial community and therefore may not provide an accurate indicator of changes in the activity and diversity of the greater microbial community under field conditions. A variety of closed or open chamber methods are available for use in the field (King, 1997), with the most widely used method of measuring CO, concentrations being infrared gas analysis (King and Harrison, 1995). However, although CO, efflux provides a means of characterizing microbial activity in the soil that may be

272

MARTYN SILGRAM AND MARK A. SHEPHERD

influenced by cultivation practices, measurements of CO, evolution may be somewhat confounded by the release of CO, from the roots of test crops growing in the same sampled soil volume, and such determinations do not actually quantify net N mineralization of the incorporated residue material, which must be inferred from the change in the temporal and spatial dynamics of microbial activity. Furthermore, there is evidence that short-term CO, flux from tilled soils is influenced more by mass flow processes related to a tillage-induced change in porosity than to changes in current soil microbial activity (Reicosky et al., 1997).

C. ISOTOPIC LABELING Isotopic labeling of 15N has proved a useful, if relatively expensive, technique with which to monitor the mineralization of organic N. One approach, the “isotope dilution” method (Barraclough and Pun, 1995), involves quantifying the dilution of a labeled ammonium solution injected into the soil as the proportion of labeled N present in the soil mineral nitrogen pool decreases over time due to the mineralization of unlabeled organic matter, including residue material. 15N/14N isotope ratios are then typically determined by mass spectrometry. However, this method assumes that the basal N mineralization is the same in the presence or absence of residue material: If, following residue incorporation, part of the soil microbial biomass switches from decomposing indigenous soil organic matter to decomposing the fresh residue, then the N mineralization resulting from residue decomposition will be underestimated as the basal mineralization rate will have dropped (Watkins and Barraclough, 1996). This technique allows field or laboratory measurement of gross rather than net mineralization and presents the opportunity to study gross N mineralization dynamics unconfounded by the processes such as nitrification and plant uptake, which can consume NH,. However, this means that immobilization of N, and hence net N mineralization, is not determined, and yet it is this net result that will ultimately determine the soil nitrogen supply to any succeeding crop. An alternative approach is to label (enrich) either a crop residue or fertilizer with 15N and monitor its movement through the soil-plant system. This can prove particularly useful as part of an N budget approach where major losses are quantified in addition to changes within the soil N pool. This can be achieved by measuring test crop recovery of labeled N, together with using either lysimeters or porous pots to quantify solute fluxes. However, gaseous losses via denitrification and volatilization may account for a significant component of the labeled N applied (e.g., 10-20%; Dowdell and Webster, 1984), and when these fluxes are not measured this results in incomplete recovery of 15Nin measured soil and plant components.

CULTWATION EFFECTS ON SOIL N MINERALIZATION

2 73

D. TEMPORAL CHANGESIN Son. MINERAL NITROGEN Another method involves monitoring the change in SMN (the total of NH,-N and NO,-N) and the change in plant N uptake of a test crop over time following contrasting cultivation techniques and compared to a no-till or direct drilled control. This method is very widely used and assumes that other pathways for N loss, such as denitrification and volatilization, are negligible during the monitored period, so that all the N released by mineralization will be reflected in the change in soil nitrogen supply (SNS = SMN plant N). Under rain-fed conditions, this method may require the estimation of nitrate leaching losses over winter. A recent development has been the use of soil cores incubated in situ under field conditions to determine net N mineralization. This technique was originally developed to measure net N mineralization of native soil organic matter (SOM) in grasslands (Hatch et al., 1990, 1991) but has recently been adapted for arable soils (Bhogal and Shepherd, 1996). Six duplicate pairs of soil cores are taken, with one from each pair being bulked and extracted immediately while the others are incubated for a week in the field in sealed Kilner jars injected with 2% acetylene to inhibit nitrification and denitrification. Jars are incubated in a covered trench to ensure that their temperature approximates that in the undisturbed soil. The difference in NH,+ concentration between pre- and postincubation samples yields a measure of net N mineralization. As the method measures absolute changes over time, it is less prone to the limitations associated with the more usual “snapshot” approach to measuring SMN, and has the advantage of representing actual temperature conditions in the field. However, soil cores are contained within airtight jars and thus are not subject to precipitation inputs or evaporative losses, while the disturbance caused by sampling can lead to overestimates of net mineralization compared to an N balance approach (Bhogal et al., in press). The time over which net mineralization may be measured is also limited to around 7 days: Longer periods may yield spurious results as microbial activity becomes progressively restricted by oxygen depletion in the sealed jars.

+

E. CONTROLPLOTS Field-based methods of indirect determination of N mineralization invariably include a (usually zero-tilled or direct drilled) control treatment that may be bare fallow or have the same test crop as the cultivated plots. The effect of cultivation on N mineralization is often determined from comparisons of SNS, yield, grain %N, and N offtake at test crop harvest from cultivation and control treatment plots: Determination of SNS is required rather than SMN alone to make allowance for crop N uptake. If a bare fallow control is used, then there are potential difficulties

2 74

MAEUYN SILGRAM AND MARK A. SHEPHERD

associated with contrasting temperature and/or soil moisture conditions under the bare fallow control plots, rendering them different from treatment plots, while the lack of a test crop may modify nitrate leaching losses and influence mineralization rates as there will be no crop N uptake “sink’ for SMN and no rhizosphere interactions associated with the release of root exudates. Thus results of research comparing cultivated treatment plots with bare fallow plots should be interpreted with care, as differences in SMN will be due to the combined effect of differences in nitrate leaching and any additional N mineralized following the disturbance caused by cultivation. From a scientific standpoint, a further problem is that it is difficult, if not impossible, to discriminate between physical and biological effects, that is, between (i) the effect of cultivation on physical structure and water retention properties that may influence crop growth and rooting patterns (e.g., increased N uptake in a ploughed vs a direct drilled control due to a more permeable soil structure in the former assisting root penetration and hence nutrient availability), and (ii) the disturbance caused by the cultivation process stimulating additional N release via net mineralization.

F. WHICHMETHOD? It is clear that no one method is capable of providing all of the information typically required in field studies, with each method having associated limitations. This is undoubtedly why relatively little progress has been made in understanding and quantifying soil N mineralization in comparison with the large amount of time and effort dedicated to the task in recent years. Techniques for studying soil biomass are available but are generally crude, difficult to interpret, and have limited applicability in cultivation studies, although newer methods may reveal more about population dynamics in detailed process studies. The measurement of CO, efflux can prove a useful guide for characterizing the effects of cultivation on soil microbial activity, particularly if field-based and automated, but root respiration can confound the interpretation of such results. Substrate induced respiration can be a valuable laboratory tool for determining population functionality but cultures will represent only a small component of the soil microbial community, and therefore results may not provide an accurate indictor of changes in the activity and diversity of the greater microbial community under field conditions. I5N techniques are expensive and require careful interpretation, but remain a useful tool in detailed process studies. In contract, the N balance approach is a more straightforward method giving snapshots of soil N status, though spatial variability may necessitate considerable replication and there may be a need for other input/output fluxes to be quantified. Recently developed incubation methods are more labor-intensive, but can provide much greater informa-

CULTIVATION EFFECTS ON SOIL N MINERALIZATION

275

tion on the temporal dynamics of N mineralization in the field. However, care is required as the technique ultimately chosen should not itself alter the rate or amount of mineralization. This use of control plots is necessary in all experiments, but more than one type of control (e.g., barehndisturbed and bare/cultivated plots) may be required to factor out the effects of tillage per se from other treatment effects. In short, it is necessary to choose the method that best suits the objectives of the experiment: Literature reports have included all the above methods in experiments designed to characterize and quantify cultivation effects on N mineralization with its associated environmental and agronomic implications for land management and the maintenance of long-term soil fertility.

111. CULTIVATION EFFECTS ON SOIL PHYSICAL CONDITIONS A. CULTIVATION TECHNIQUES Soil cultivations are used to control weeds, destroy and bury residues, level the soil surface, remedy compaction in subsoil layers, and create a seedbed with a suitable tilth for the next crop (Chamen and Parkin, 1995). It is important to understand fully the cultivation methods used in soil tillage since these ultimately affect the environment for biological activity and hence N mineralization. Primary cultivation techniques either completely invert the soil (mouldboard ploughs) or mix the soil down to a working depth (rotary cultivators and disks). The traditional method of obtaining a good seedbed is to plough the land with a mouldboard plough, which turns a furrow slice, then to work this furrow slice down into a suitable tilth for the seedbed using secondary tillage implements such as cultivators, harrows, and rolls (Wild, 1988). In contrast, disk ploughs have large typically vertically mounted concave disks instead of shares and mouldboards and, unlike mouldboards, they do not completely invert the soil. However, the mouldboard plough and disks do have limitations, especially on finer-textured soils: They may compress soil in the furrow slice if the soil is wet, leaving it in larger clods requiring further cultivations to break them down into a suitable seedbed, and they can create a compacted plough pan beneath the cultivated layer, which may subsequently restrict drainage, aeration, and root penetration. More recent innovations include the use of rigid or spring tine and chisel cultivators, which do not invert the soil but generally cause less subsoil compaction than conventional mouldboard ploughs and disks: These have proved most useful on heavier soils, reducing the cost and time required to produce a good seedbed for the next crop (Wild, 1988). Rotary cultivators have a series of blades rotated by a shaft set orthogonal to the direction of travel, with the degree of pulveriza-

276

m T Y N SILGRAM AND MARK A. SHEPHERD

tion controlled by the speed of rotation of the blades, the speed of the tractor, and the position of the shield that recirculates clods back through the rotor. Such implements have a great advantage of being capable of producing a good tilth in only one operation, although this can lead to greater damage to the soil through misuse in wet conditions. Light rollers may also be used to crush soil clods, break surface caps, and consolidate the uppermost few centimeters of the soil to keep soil in contact with seeds and plant roots and help conserve moisture (McLaren and Cameron, 1990). Zero, reduced, or minimum tillage may be used as an alternative to conventional ploughing, and typically leads to soil conditions that differ markedly from those under more conventional arable systems: Compared with conventional cultivations, effects of reduced tillage typically include greater bulk density, root penetration resistance, structural stability, and pore connectivity, but lower porosity and soil nitrogen availability, with often little effect on overall crop yield (McLaren and Cameron, 1990; Wild, 1988). In minimum tillage, the soil is lightly worked with a cultivator or harrow before drilling, whereas in direct drilling the seed is drilled straight into the undisturbed soil: Such simplified cultivation techniques are being adopted by cereal growers in many countries (Cannell, 1985). Direct drilling requires much smaller energy inputs to plant new crops (35-80 MJ ha-') compared with conventional ploughing techniques (200-360 MJ ha-'): It also has the advantages of allowing greater flexibility due to the reduced workload and improving soil and water conservation (Douglas et al., 1986; Wild, 1988). However, minimal cultivation techniques may not be feasible on lighter textured soils because of their tendency to slump. Furthermore, any economy in time and fuel cost is offset by the necessity to use herbicides or extra machinery and labor in addition to a small amount of supplementary N, which may be required to maintain crop yields. The accumulation of plant residues near the soil surface, which would otherwise be buried in conventional ploughing operations, can also interfere with drilling, germination, and seedling growth such as through the anaerobic fermentation of straw releasing substances toxic to seedlings (Harper and Lynch, 1981). Without conventional cultivations, organic matter and nutrients such as N tend to accumulate at or near the soil surface, and this may restrict mineralization rates in the soil beneath (Wild, 1988; Chamen and Parkin, 1995). In conclusion, the choice of cultivation is strongly influenced by soil type. Greater flexibility is possible on heavier soils, which may be managed under reduced tillage systems with rotational ploughing.

B. CULTIVATION EFFECTSON SOILPHYSICAL PROPERTIES Ploughing and cultivation increase soil aeration (Granli and B@ckman,1994), and the physical disruption caused by intensive cultivation can result in excessive

CULTIVATION EFFECTS ON SOIL N MINERALIZATION

2 77

breakdown of soil aggregates and produce a tilth that is very fine and loose (Scullion et al., 1991). The reduced aggregate stability often identified in ploughed soil is frequently associated with increased porosity and aeration and a decrease in soil bulk density within the plough depth (Ekeberg, 1992; Stokes et al., 1992), and such changes in bulk density have been found to be inversely related to rates of N mineralization (Kemper, et al., 1971). In contrast to the effects of ploughing, minimum cultivation can lead to a loss of soil pores, and this will reduce the rate at which water enters (infiltration) and drains through (hydraulic conductivity) a soil. A key effect of cultivation practices is to modify a soil's drainage characteristics by disrupting the connectivity and continuity of soil pores, especially the larger, intrapedal pores that may connect the near-surface region to deeper layers in the profile, in addition to the possible development of a smeared plough pan restricting rooting and drainage below the cultivation depth (Chamen and Parkin, 1995). Any such modification to soil moisture status can influence redox conditions and hence influence rates of soil microbial activity, including mineralization. The amount of organic matter in soil is critical for maintaining the stability of soil aggregates, and this is also influenced by cultivation techniques. Intensive cultivation increases organic matter decomposition and can thus lead to a decrease in a soil's organic matter content: Thus in addition to minimum- or zero-tillage systems generally possessing higher bulk densities and more water-stable aggregates near the soil surface, such cultivation systems also tend to have greater organic matter contents (Hill, 1990: Kladivko et al., 1986). Changing from a conventional to a zero-tillage system can improve soil structure as organic matter content and soil organic C content increase, both of which appear strongly related to soil aggregate stability (Kladivko et al., 1986; Havlin et al., 1990; Carter, 1992), and this increased structural stability can greatly reduce soil erosion as well as having agronomic benefits such as decreasing labor and machinery costs (Featherstone et al., 1991). This is supported by Rasmussen and Collins (1991), who reported from 10 different studies comparing the effects of noninversion compared with conventional tillage over periods ranging from 5 to 44 years in duration, and found that topsoil C and N contents increased by an average of 1-2% in noninversion tillage systems compared with conventional cultivation. In a long-term study into tillage effects on soil properties in Ohio, USA, Mahboubi et al. (1993), growing maize (Zea mays L.), sampled a silt loam under continuous mouldboard ploughing, chisel ploughing, or zero tillage for the 28-year period beginning in 1962. Results indicated that increasingly intensive cultivation resulted in less organic C, higher porosity, few water-stable aggregates, a smaller mean weight diameter' of water-stable aggregates, and lower saturated hydraulic conductivity. Such results concur with those of Arvidsson and HAkansson (1996), 'The sum of the mass fraction of soil remaining on a sieve after sieving multiplied by the mean aperture of the adjacent meshes (Besnard e t a / . , 1996).

278

MARTYN SILGRAM AND MARK A. SHEPHERD

who concluded that ploughing largely restores the macroporosity of the of the soil within the plough layer, with the main effect of no tillage and soil compaction being on the interaggregate bonding. Evidence suggests that in addition to changing the aggregate size and stability, tillage changes a soil’s total porosity, pore size distribution, and the continuity and connectivity of soil pores (Addiscott and Dexter, 1994). For example, Dowdell et al., (1979) found 3% (v/v) greater oxygen concentrations at 15 cm depth in a direct drilled compared with a ploughed clay soil, which was attributed to the development of a system of continuous large pores and channels that would otherwise have been disrupted by ploughing. As the size of soil aggregates increases, this also tends to increase the range of pore sizes, and thus tillage operations that result in a finer tilth will typically reduce the pore size in a given soil. In a long-term experiment comparing mouldboard ploughing and direct drilling for 22 years, Ball et al. (1996) found the surface of a Gleysol and a Cambisol to be more stable, less compactable, and have greater plasticity limits under direct drilling compared with conventional ploughing, with these differences correlated to total carbon and carbohydrate concentrations with depth and tillage treatment. Timing of cultivations with respect to soil moisture conditions can also have a significant influence on soil physical conditions and hence on microbial processes. Cultivations such as ploughing can result in serious soil compaction if conducted when the soil is too wet, and this can result in short-term effects relating to the bulk density of the plough layer, structural effects that persist after ploughing, and subsoil compaction including plough pan formation. In a review of 21 longterm field experiments in Sweden, Arvidsson and HAkansson (1996) compared standard (control) seedbed preparation (mouldboard ploughing in autumn) with three harrowings in spring and compacted (extra traffic in autumn) soils over 259 site-years. The extra traffic significantly ( P < 0.05) decreased the porosity and the proportion of large pores, increased the tensile strength of dry aggregates, caused a mean yield loss of 1 1.4%, and on clay and loam soils also decreased the proportion of fine aggregates and the gravimetric water content in the seedbed. Plant N uptake was lower ( P < 0.05) in the compacted treatments, and yield loss was mainly influenced by soil type, being 20% on clay soils. Such evidence indicates that any beneficial effect of cultivation in creating a good seedbed, burying trash, and stimulating microbial activity to release mineral N must be counterbalanced by the risk that excess traffic and overcultivation, especially on heavier soils, not only will damage soil structure, but also can cause serious subsurface compaction and ultimate yield reduction. Given the overwhelming evidence of physical changes in soil properties due to contrasting tillage practices, one important issue is the persistence of such effects. The duration of cultivation effects was studied by Arvidsson and HAkansson (1996), who reported that within 4-5 years after the termination of their traffic treatments, the yield loss had disappeared and yields had returned to the control

CULTIVATION EFFECTS ON SOIL N MINERALIZATION

279

level. This suggests that cultivation effects on soil structure, and possibly mineralization, may be relatively short-lived, and these issues are discussed further in Section IV. The method of cultivation can have an important impact on the magnitude and pattern of water use through its effects on the rooting patterns and water distribution within the soil profile. Early work found that approximately 10% more water was stored in arable soils that had been direct drilled compared to those that had been conventionally ploughed, which enabled a winter wheat crop to extract up to 22 mm more water from the direct drilled soil (Goss et al., 1978). This is consistent with more recent research by Shepherd and May (1992), who reported that a loamy sand soil dried out significantly more after ploughing than after direct drilling. For vegetable crops, deeper cultivations have been shown to be beneficial because they allow roots to penetrate deeper into the subsoil and extract more water from the topsoil region (Wild, 1988), resulting in greater utilization of water held at greater depths, which can feed through to give higher yields (Rowse and Stone, 1980). Direct drilled soil typically has a higher surface reflectance coefficient (albedo) and higher thermal diffusivity than ploughed soil (Hay et al., 1978) and this is one reason unploughed soil is often cooler than ploughed soil (Ekeberg, 1992; Fortin et al., 1996; Ekeberg and Riley, 1996). This can result in a delay in planting and maturation, which could be either an advantage or a disadvantage: Areas with a long growing season could have greater yields at reduced cost, while in other areas this could result in the need to harvest under unsatisfactory (wet and cold) conditions or before the crop has matured sufficiently (Ekeberg and Riley, 1996). However, in some circumstances a large accumulation of residues at the surface of reduced tillage soils can have an insulating effect, leading to slightly (e.g., 1°C) higher temperatures at sunrise compared with ploughed soils (Franzluebbers et al., 1995). Any difference in temperature between cultivated and uncultivated soils will modify rates of microbial processes such as mineralization, which researchers generally consider to follow a Q , , response pattern with microbial activity typically doubling for a 10°C change in soil temperature (Quemada and Cabrera, 1995, 1996). Furthermore, the different temperature conditions typically reported for reduced tillage compared with conventional cultivations can influence crop root growth, which research has found to be 2.6 to 5.1 times greater at a soil temperature of 25°C than at 18°C (Mackay and Barber, 1984). Cultivation therefore has major effects on soil structure and physical characteristics. Tillage typically increases porosity and aeration, but with associated decreases in the mean diameter and structural stability of soil aggregates. It reduces soil bulk density and pore connectivity and continuity, which can lead to a smaller water holding capacity and lower saturated hydraulic conductivity. Ploughed soil may also be slightly warmer than unploughed soil, with resulting feedback ef-

280

MARTyNSILGRAM AND MARK A.

SHEPHERD

fects on soil evaporation, microbial activity, and crop growth (see below). Increasingly intensive cultivation practices tend to decrease soil organic matter status (soil organic C and N), with implications for long-term soil fertility/sustainability and erosion risk. By modifying the physical environment and abiotic conditions (water content and temperature), cultivation practices have direct effects on soil microbial activity, which governs nitrogen cycling processes, including mineralization, crop growth, and N uptake.

IV. CULTIVATION EFFECTS ON NITROGEN MINERALIZATION A. INTERACTIONS WITH SOILTEXTURE In his classic work, Hans Jenny (1941) noted that, in general, soil organic matter levels tended to increase with increasing clay content of soils, and thus for a given climate, topography, and vegetation, fine-textured soils generally had more organic matter and, therefore, total (predominantly organically bound) N compared to their coarser-textured counterparts. This organic matter accumulation is thought to be the combined result of the effect of clay in stimulating microbial growth and activity (Bondietti et al., 1971; Martin et al., 1976) and the development of organoclay complexes that may have a reduced susceptibility to biodegradation (Stevenson, 1982). Cultivation generally leads to a temporary increase in soil mineral nitrogen, most probably because the soil disturbance thus caused leads to a larger pool of carbon substrates being made available to support greater microbial activity (Wild, 1988).This is thought to occur as the physical disruption of soil aggregates caused by tillage practices results in the exposure of microsites where organic matter was previously physically protected from microorganisms or their enzymes (Adu and Oades, 1978). Physical protection of organic matter in soils is thought to be associated with encrustation by clay particles (Tisdall and Oades, 1982) and/or entrapment in small pores within soil aggregates that may be inaccessible to microbes (Elliot and Coleman, 1988). Organic matter bound to the <2-pm clay particle fraction has been shown to concentrate microbial biomass and its metabolites (Amato and Ladd, 1980) and may contribute to a temporary or transient pool of organic material (Tisdall and Oades, 1982), although this will be dependent on its relationship to larger aggregates constituting soil structure, with laboratory incubations revealing that short-term C and N mineralization is faster from macroaggregates than from microaggregates (Gupta and Gennida, 1988; Gregorich et al., 1989; Cambardella and Elliot, 1994).

CULTWATION EFFECTS ON SOIL N MINERALIZATION

2 81

There are thought to be two main features that account for differences in soil organic N mineralization in different aggregate size classes: (i) differences in pore sizes, which means that organic material may be protected from degradation within soil aggregates (Jocteur-Monrozier et d., 1991; Ladd et al., 1992); (ii) differences in the “quality” (e.g., relative decomposability, as measured e.g., by C/N ratio) of particulate organic matter in different aggregate size classes, which means that N turnover in macroaggregates may be faster because of the presence of large, fresh plant debris promoting more rapid microbial activity (Balabane, 1996). It is these larger macroaggregates that may be most vulnerable to physical disruption during conventional cultivation. Indirect evidence for the physical protection of organic matter in soil includes (a) drying of soil samples and disruption of soil aggregates before incubation can increase organic C and N mineralization during the first few weeks after the start of the incubation (Cabrera and Kissel, 1988; Gregorich et al., 1989); (b) net mineralization of soil organic matter and the decomposition of added plant material appears more rapid in sandy compared with clay soils (Ladd et al., 1990; Hassink et al., 1990; Verberne et al., 1990). The lower net N mineralization in the heavier clay soils is thought to be caused by this greater physical protection of soil organic matter and microbial biomass (Verberne et al., 1990), with the most recent fractionation studies of fresh and indigenous soil organic matter suggesting that mineralization appears to be more strongly influenced by this physical position and protection rather than by the substrate’s chemical composition (Balesdent, 1996). As a major influence on soil structure, evidence suggests that clay content is particularly important in influencing N mineralization as organic N is often intimately associated with the clay fraction in soils, with more clay-rich soils often possessing higher total %N concentrations. This is confirmed by Franzluebbers er al., 1996), who found the amount of mineralizable C and N per unit soil microbial biomass C decreased with increasing clay content, indicating that the soil microbial biomass was more active in undisturbed coarse-textured soil than in undisturbed fine-textured soils. However, once soils are disturbed the opposite is true, as grinding soil samples of differing textures results in much greater mineralization in clayrich soils due to the release of readily mineralizable organic N previously inaccessible to microbial degradation (Haynes, 1986;Hassink, 1992).A major hypothesis to account for the retention of such organic matter is that the soil mineral matrix protects soil organic matter against faunal predation (biodegradation), either through adsorption of substrates to mineral surfaces or by sequestration in aggregates at sites inaccessible to microbes (Van Veen and Kuikman, 1990). Physical protection is likely to be less in cultivated than in uncultivated soils because tillage periodically breaks up soil aggregates and exposes previously protected soil organic matter (Balesdent et al., 1990). This physical protection may explain why

2 82

MARTYN SILGRAM AND MARK A. SHEPHERD

fine-textured (>40% clay) soils have been found to contain 1.2 to 1.5 times more soil organic C (SOC) and 2.5 to 3.5 times more soil microbial biomass C compared with coarse-textured (<15% clay) soils (Van Veen et al., 1985; Van Gestel et al., 1991), although the higher SOC content in finer-textured soils could also be due to differences in C input, rather than long-term decomposition dynamics, since they tend to be more fertile than their coarser-textured counterparts (Franzluebbers et al., 1996). In field soils, Adu and Oades (1978) concluded that as much as 90% of the organic matter may be inaccessible to the soil microflora and extracellular enzymes, with physical disruption caused by mechanical cultivation resulting in a flush of microbial respiratory activity measured as 14C0, evolution as microorganisms are brought into contact with fresh, previously unavailable substrate. Contemporary research by Besnard et al., (1996) studied the decomposition of a labile pool of particulate organic matter (POM), including plant residues, in a loamy forest soil introduced into maize cultivation, and found that cultivation decreased the mean weight diameter of stable aggregates from 2.55 to 2.04 and 1.23 mm after 7 and 35 years of continuous maize production, respectively, with the proportion of stable aggregates decreasing from 78 to 47% of the soil mass after cultivation, primarily due to a loss in macroaggregates >200 pm. Cultivation reduced the C/N ratio in these macroaggregates from 30 in the forest soil to 24 and 21 after 7 and 35 years of continuous maize production, respectively. Results suggested that cultivation resulted in the loss of C from outside the aggregates, with the POM fraction occluded within microaggregates (50-200 pm) found to turnover more slowly. Other research has shown that young POM from maize residues can act to stabilize soil aggregates in cultivated silty soils (Puget et al., 1993, with cultivation resulting in a decrease in the relative proportion of carbohydrates in SOM and an increase in carboxyl C, phenolic C, and aromaticity of the SOM (Lessa et al., 1996). Research by Hassink (1992) investigated the hypothesis that disrupting soil structure increases mineralization rates in loams and clays more than in sandy soils, and that this increase can be used to estimate the fraction of physically protected organic matter that might be made available following cultivation. C and N mineralization was measured in undisturbed and in finely and coarsely sieved moist or driedlremoistened soil. N mineralization rates were significantly ( P < 0.05) lower in the undisturbed samples, with a regression analysis of the data from 1991 and 1992 revealing that variations in the proportion of the clay+silt component (i.e., <50 pm) between different soils explained 71% of the observed variation in N mineralization rates over 84-day monitoring periods. This is consistent with I5N research by Balabane (1996), who found that recently immobilized N associated with the clay fraction was rapidly sequestered in microaggregates < 100 pm. Skjemstad et al. (1993) also found that a considerable proportion of soil organic matter was physically protected within clay- and silt-sized aggregates, with

CULTIVATION EFFECTS ON SOIL N MINERALIZATION

283

material external to clay- and silt-sized aggregates being largely proteinaceous in contrast to internal material, which more closely resembled humic acids. From their research, Hassink (1992) found that fine sieving caused a temporary increase in mineralization that was much larger for N than for C and for loams and clays compared with sandy soils, with relative increases in N mineralization averaging 150% for loams and clays but only 5% for sands in the 2 weeks after fine sieving relative to coarsely sieved control soils: These findings are consistent with other research by Cabrera and Kissel (1998) and Schoder et al. (1989). In loams and clays, small pores constitute a higher percentage of the total pore space compared with sandy soils, and Hassink (1992) found that (a) the fraction of pores C1.2 pm and (b) the clay content both significantly affected N mineralization rates. The first finding (a) was thought to be because most small voids in soil are filled with carbohydrates, many of which are attached to clay particles (Foster, 1986), but this material cannot be reached by microorganisms and was therefore physically protected against decomposition, with the physical disruption caused by sieving exposing part of this fraction to rapid microbial degradation. As the mineralization flush following disturbance was significantly greater for N than for C, this suggests that any physically protected organic matter had a lower C/N ratio than the rest of the soil organic matter, which is consistent with this being more readily mineralizable organic matter susceptible to rapid mineralization following physical disruption of the soil structure. The sandy soils studied by Hassink (1992) had generally higher C/N ratios than the loams and clay soils, possibly, as Chichester ( 1 969) and Cameron and Possner (1979) reported, because C/N ratios decrease with decreasing particle size because organic material coated with clay particles has a better physical protection than organic material around sand particles. Such results are in agreement with earlier research by Tisdall and Oades (1982), who found that aggregates between 2 and 20 pm in diameter contain most of the micropores and soil C and N: Such aggregates contain bacteria surrounded by clay particles, cell-wall remnants, and other microbial decay products. If substrates made available by physical disruption are of microbial origin, they will have a narrow C/N ratio, which may also explain why the relative increase Hassink (1992) detected in N mineralization due to sieving was much larger than the relative increase in C mineralization. N mineralization was 0.04-0.06%/day for sieved but 0-0.02%/day for undisturbed sand, and 0.02-0.06%/day for sieved but 0-0.02%/day for undisturbed clay and loam soils over the first 2 weeks following disturbance; C mineralization was between 2.4 and 6.3 times greater than corresponding N mineralization rates. However, this flush of N mineralization was relatively short-lived, with Hassink (1992) reporting that values were significantly (P< 0.01) higher than undisturbed control soils only during the first 5 days (loams and clays) or the first 2 weeks (sands) following physical disturbance. Other researchers have reported equally ephemeral effects of structural disturbance, such as cultivation, on C and N

2 84

MARTYN SILGRAM AND MARK A. SHEPHERD

mineralization rates (Richter et al., 1982; Dowdell et al., 1983; Nordmeyer and Richter, 1985), again suggesting that the compounds that are made accessible are very easily decomposable. However, although there does appear to be evidence suggesting a significant effect of soil type on the effects of cultivation on N mineralization rates, no clear relationship has ever been described between pore-size distribution and the physical protection of organic matter from microbial decomposition, although researchers such as Hassink (1992) suggest that it is the small pores that contain physically protected organic matter with a low C/N ratio that can be mineralized very quickly following physical disruption such as cultivation. Therefore, two explanations may account for the observed effects of cultivation inducing a “flush” of microbial activity resulting in increased soil nitrogen mineralization: (i) The physical protection of soil organic matter, as the physical disturbance caused by cultivation fractures soil peds and brings microorganisms and soil fauna into contact with fresh, previously unavailable (physically protected) substrate; (ii) the modified soil environment conditions (aeration, water content, temperature) induced by cultivation, which will directly influence growth and activity of the soil biomass. Cultivation disrupts soil structure, decreasing aggregate size and the C/N ratio in macroaggregates. The soil mineral matrix protects soil organic matter against biodegradation either through adsorption of substrates to mineral surfaces or by sequestration in aggregates at sites inaccessible to microbes. There is an interaction between mineralization potential and soil type: Clay-rich soils possess larger amounts of physically protected soil organic matter within structural aggregates and are characterized by greater mineralization potentials. This accounts for net mineralization of soil organic matter and the decomposition of added plant material being more rapid in sandy than in more clay-rich soils under undisturbed conditions, whereas the reverse is true after conventional cultivations due to the release of readily mineralizable organic N previously inaccessible to microbial degradation. Effects of cultivation on mineralization are typically relatively short-lived, with differences from uncultivated controls detectable only for several weeks following tillage.

B. EFFECTSON SOILFAUNA In a study comparing 13 years of conventional mouldboard (CT) ploughing to 15 cm depth versus no tillage (NT) on a clay loam in the southeastern United States, Beare et al. (1994) found an 18% increase in organic C in the plough lay-

er of the NT relative to the CT treatment, which they attributed to differences in the assimilation and decomposition of SOM under the different tillage regimes.These authors reported far more variable temperature and moisture conditions in the surface soils of CT treatments that would influence microbially mediated processes, while the greater biological activity near the soil surface of NT

CULTIVATION EFFECTS ON SOIL N MINERALIZATION

2 85

treatment soils helped to incorporate particulate organic matter within macroaggregates and hence increase their structural stability. These results concur with the findings of Ball et al. (1996), who identified increased organic C and carbohydrate concentrations near the surface of direct drilled plots compared to the more uniform distribution in ploughed plots of a Cambisol and a Gleysol under winter barley in southeastern Scotland, and this accumulation of organic matter near the surface of no-till soils has been reported by other researchers (Blevins et al., 1983; Douglas and Goss, 1982). The accumulation of organic material at the soil surface in direct drilled plots may lead to greater biological activity compared with ploughed soil (Douglas, 1977; Hoffman et al., 1996a,b) and this is reflected in the greater microbial biomass measured in direct drilled compared to ploughed soils (Lynch and Panting, 1980, 1982). In particular, direct drilled soils have been found to possess significantly more fungal propagules, but not bacteria, in the 0- to 5-cm layer compared to unploughed controls (Barber and Standell, 1977): Fungi are frequently the largest component of the soil biomass (Anderson and Domsh, 1975). This is consistent with other research by Lee and Pankhurst (1992), who concluded that the initial breakdown of plant tissues was predominantly mediated by bacteria in ploughed systems, whereas fungi dominated under direct drilled conditions: This may be encouraged by the development of acidic surface layers of decomposing crop residues in direct drilled systems. However, in contrast, Campbell et al. (1989) actually found a narrowing of the biomass C/N ration from 9 to 5 following 6 years of zero tillage, which was attributed to a shift from a microbial population dominated by fungi to one with a greater preponderance of bacteria and actinomycetes. Thus, tillage practices can also influence decomposition processes controlling N cycling and mineralization by modifying the soil faunal population. The agronomic significance of soil fauna depends on the intensity of cropping systems, since grazing, tillage, fertilization, and pesticides generally tend to reduce the species complement and their population densities (Anderson, 1988). Other research has found that noninversion tillage systems can increase populations of gamasid mites, earthworms, and Collembola in experiments conducted in Germany and on a range of soils from sandy loams to clays in the United Kingdom (Edwards and Lofty, 1982; El Titi and Landes, 1990). This is consistent with other research by Rovira et al. (1987) and experimental work on a silt loam at Rothamsted, UK, which revealed more earthworms and soil arthropods such as surface predatory beetles, springtails, and insects, but fewer mites and slugs after 6 years of direct drilling compared with conventional ploughing in fields sown to winter wheat (Patterson et al., 1980). Similarly, Carter (1991b) reported an increase of 140-160% in earthworm biomass in rotary harrowed or direct drilled plots compared to conventional ploughing of a sandy loam soil. Earthworms can represent the primary agent for incorporating residues in untilled soils, and their burrows may enhance deeper storage of soil water (Mackay

2 86

MARTYN SILGRAM AND MARK A. SHEPHERD

and Kladivko, 1985), while earthworm casts are active microsites for denitrification and N,O production (Elliot et al., 1990; Knight et al., 1992).Nagel etal. (1993) found that CO, production and nitrate release were enhanced by mesofauna, but lowered by earthworms due to a transient immobilization of nutrients in microorganisms that increase dramatically during the passage through the earthworm gut (Edwards and Fletcher, 1988). Thus a decrease in earthworm biomass has been associated with an increase in soil inorganic N (Lee and Pankhurst, 1992), which is consistent with reports of greater soil nitrate concentrations and nitrate leaching losses in soils under conventional cultivation compared with those under reduced tillage systems. Despite this greater faunal activity in direct drilled soils, there is evidence suggesting that such minimum tillage regimes may restrict nitrate availability as the presence of a mulch of dead vegetation, often with a wide C/N ratio, tends to stimulate immobilization near the surface of direct drilled or untilled soils (Kitur et al., 1984). This would explain why untilled soils can contain significantly less nitrate-N than cultivated ones, particularly during autumn (Dowdell et al., 1983), although other causes of lower levels of nitrate-N in untilled soils may include increased leaching due to improved pore connectivity (Goss et al., 1978) and greater denitrification losses of N,O and N, associated with higher soil moisture contents and an additional source of readily available C in the soil (Aulakh et al., 1984). Mineralization should therefore be considered as the net effect of all soil biological processes, including the contribution from macrofauna such as earthworms and soil arthropods, rather than purely a function of the population dynamics of the soil microbial biomass alone. The physical disruption caused by cultivation has a profound effect on soil faunal populations, activity, and location in the soil profile. Earthworm and beetle populations decrease when a direct drilled soil is ploughed, with ploughing leading to a shift in microbial populations from one dominated by fungi to one with a greater preponderance of bacteria and actinomycetes.

C. EFFECTSON YIELD Given the typically smaller soil nitrate levels under minimum tillage systems, it is not surprising that long-term cropping experiments on corn, wheat, and barley have shown that without added N, or with N at low rates, crop N uptake and yields are often lower on untilled compared to tilled soils, although with adequate fertilizer N supply similar or greater yields can be achieved without tillage (Kitur et al., 1984; Smith and Howard, 1980; Ellis etal., 1982; Haynes, 1986; Dick etal., 1992). However, such conclusions have not been universally observed (Ekeberg and Riley, 1996; Kapusta et al., 1996) with crop yields sometimes showing no differ-

CULTIVATION EFFECTS ON SOIL N MINERALIZATION

2 87

ence following establishment by direct drilling or conventional ploughing (Webb et al., 1991). This variable response may be a result of prevailing weather conditions during research studies, as other research by Carter (1991a) found that shallow tillage and direct drilling may produce similar grain yields as mouldboard ploughing only when environmental conditions were optimal, with wet or dry seasons favoring ploughing and direct drilling, respectively, due to their effects on soil moisture status. In a separate study, Ekeberg and Riley (1996) compared the effects of tillage systems on the yield and nutrient uptake of potato (Solanurn tuberosum L.) in Norway from 1987 to 1993. In their research, the effects on potato yield of a conventional, labor-intensive treatment using autumn mouldboard ploughing and two passes with a spring-time harrow in spring were contrasted with planting directly into untilled barley stubble with straw removed. Results on a morainic, stony loam soil indicated a pattern of distinct yield curves for the different cultivation treatments expressed as functions of harvest date: The yield curve for direct planting was steeper, crossing that of conventional tillage on 10 September and thus predicting higher tuber yield for direct planting when harvesting after this date, but lower tuber yields compared with conventional tillage in the case of early harvesting. This difference was largely attributed to cooler soil and delayed growth and hence maturation in the case of direct planting. Nutrient uptake of the plants was consistently greater (57, 11, and 41 kg N/ha) with direct planting, despite the same fertilizer being applied to both treatments, and was attributed to the direct planted crops making better use of the applied fertilizer or the increase in topsoil organic matter, and hence nitrogen, found after a number of years without ploughing (Ekeberg, 1992). In a novel approach making use of the natural difference in I3C content between C, and C, (maize) plants, Balesdent et al. (1990) measured C and 13C/'*C ratios and found that mineralization of the indigenous soil organic C was approximately double on conventionally ploughed compared with zero-tilled experimental plots. However, less organic carbon accumulated in the zero-tilled plots after harvest due to lower yields of maize in these plots. Saffigna et al. (1989) also used measurements of organic C in a study of the effects of cultivation practices on soil organic matter dynamics. Following 6 years of growing sorghum on an Australian Vertisol in central Queensland, these authors detected an increase in biomass C of 14-21% as a result of zero compared with minimum tillage (composed of 10 cm depth diskkine cultivation). In a detailed study of the long-term effects of contrasting cultivation systems on the yield of corn over 20 years on a silt loam in Illinois, USA, Kapusta et al. (1996) found that yields were equal in conventional till, reduced-tillage (chisel cultivator), alternate-tillage (cycle of NT 2 years, CT 1 year), and zero-tillage systems all receiving broadcast NPK fertilizer. These results are consistent with earlier work by Griffith et al. (1988), although Kapusta et al. (1996) also found that yields were 15-18% lower in NT treatments compared to the other tillage systems if no

288

MARTYN SILGRAM AND MARK A. SHEPHERD

fertilizer or N only was applied in the spring. Dick et al. (1992) studied the effects of cultivating fields under no tillage for the previous 5 years on maize yield and N uptake using six levels of fertilization on a silty loam soil. Grain N concentrations were consistently lower for maize grown under NT compared to CT, yet grain yields were generally higher, especially at high fertilizer N rates. This increased efficiency of N for grain production was attributed to the greater amount of water measured in the soil under NT management in all 5 years of the study. Spring ploughing also affected N availability to maize, although this was more pronounced in the second crop rather than the first crop after ploughing. Overall, grain yields and N uptake resulted in low fertilization rates proving more effective for the CT treatment, whereas high fertilization rates increased grain yields and N uptake for NT so that it equalled or exceeded those from CT treatments: This is a pattern reported by a number of other researchers (Meisinger et al., 1985; Rice et al., 1986; Thiagalingam et al., 1991). Tillage effects are also profoundly affected by prevailing weather conditions. Radford et al., (1992) reported that zero tillage with stubble retention stored the most water during the fallow periods, but the least soil mineral nitrogen. As a result, this treatment outyielded all other treatments during dry years but produced one of the lowest yields during the wettest year. There is also evidence that the accumulation of nutrients in the uppermost 5 cm of NT soils can result in taller plants compared to other tillage systems, possibly due to the enhanced water and nutrient availability (Griffith et al., 1988; Dick et al., 1991). A number of other researchers have found that minimum-tillage systems promote soil water retention and can reduce soil temperature (Bennett et al., 1973; Johnson and Lowery, 1985; Wilhelm et al., 1986), which can mean that crop emergence and early season growth may be delayed (Imholte and Carter, 1987) even though root extraction of water and nutrients may be greater with zero tillage than under conventional ploughing (Hargrove, 1985). However, such effects mean it can be difficult to identify the precise reasons for observed differences in N mineralization, SMN, and crop yield and N uptake due to contrasting cultivation practices: Differences could be (i) due to physical effects on soil structure and hydraulic properties due to contrasting soil water and temperature status influencing crop growth and development and acting as abiotic controls on microbial activity; (ii) due to the accumulation of C and N in residues near the surface of direct drilled plots modifying soil N cycling processes (in contrast to the more uniform residue distribution in ploughed soil); or (iii) due to the physical disturbance caused by cultivation making previously physically protected substrate available for microbial decomposition. Observed tillage effects are most probably the net result of all of these different influences. Due to their effects on soil water and temperature conditions, minimum- or zerotillage cultivations are usually most successful on well-drained soil, rather than on poor or imperfectly drained soil, especially under wet soil conditions (Griffith et

CULTIVATION EFFECTS ON SOIL N MINERALIZATION

2 89

al., 1988; Herbek et al., 1986). Poorly drained soils with relatively high organic matter contents cropped to corn have been found to yield generally less with NT compared with CT because of lower soil temperatures and excess moisture retention in the spring (Kapusta er al., 1966). On well-drained soils, crop yields are generally similar under direct drilled and conventional ploughing systems under optimal fertilization regimes and weather conditions. However, direct drilled crops typically require a small amount (ca. 20 kg/ha) of additional fertilizer N to achieve the same yields as crops grown following conventional ploughing, largely as a result of the greater immobilization of mineral N near the surface of no-till soils due to the accumulation of carbonaceous crop residues.

D. PREVIOUS CROPAND RESIDUEMANAGEMENT The mineralization dynamics of N from the preceding crop’s harvest residues will largely depend on the nature of the organic N added, C/N ratio and N content of the residue, residue placement, the degree of contact with the soil matrix, tillage and cropping practices, as well as soil temperature, moisture, and aeration (Iritani and Arnold, 1960; Frankenberger and Abdelmagid,l985; Smith et al., 1987; Breland, 1994; Kuo et al., 1996). Cereal straw residues may be baled and removed, incorporated by cultivation, or left on the soil surface. Legislation has now outlawed burning as a method of residue disposal in the United Kingdom. Such residues’ relatively high C/N ratio (e.g., 70-100 for cereal straw) may promote rapid immobilization of soil mineral N as microbial populations are unable to satisfy their N demand from such carbonaceous substrates. Leaving residues on the soil surface can reduce SMN supply to a succeeding crop as the limited N supply can severely limit decomposition, while research suggests that the alternative of incorporating residues by cultivating can greatly increase their decomposition rate (Brown and Dickey, 1970). It is difficult to assess the true significance of the management of previous crop residues such as stubble on crop growth, as it may not be possible to discriminate between any phytotoxic effects of stubble in reduced cultivation systems from other adverse effects on plant growth such as the immobilization of N caused by stubble incorporation and the often lower levels of nitrate in the soil profile under reduced-tillage systems (Thompson, 1992). Stubble retention modifies soil moisture (and hence redox) conditions by storing incident precipitation and mitigating soil evaporative loss, which accounted for a difference of 5-10 mm after successive rainfall events in the weighing lysimeter study of Freebairn et al. (1987). The increased risk of localized anaerobicity may be exacerbated by direct drilling into zero-till plots where planting equipment does not fully close the drill slot after seeding (Scott Russell et al., 1975), and this in turn may result in problems in seedling emergence due to fungal or bacterial colonization generating phytotox-

2 90

MARTYN SILGRAM AND MARK A. SHEPHERD

ins such as the antibiotic patulin or acetic acid (Thompson, 1992).For example, in a long-term experiment on an Australian Vertisol, Thompson (1992) reported that stubble retention caused far greater depression of vegetative growth of barley (Hordeum vulgare L.) in zero tillage compared with mechanical tillage of the fallow. Swift et al. (1979) noted that the surface area and volume of detritus particles will significantly influence their susceptibility to enzymes and to ingestion by soil animals, and hence to decomposition. In general, more finely divided, macerated, or ground plant material decomposes more quickly than coarse material as it exposes a greater surface area for microbial colonization and enzymatic activity (Moore, 1974; Haynes, 1986). Thus crop residue management such as chopping and incorporation can have a significant influence on N mineralization and soil mineral nitrogen supply in the postharvest period, with chopping materials with a wide C/N ratio promoting the rapid immobilization of inorganic mineral N (Smith and Sharpley, 1990). This is supported by research on maize stalk pith that found that finely chopped particles immobilized six times as much inorganic N in the first month after incorporation compared with more coarsely chopped particles (Sims and Frederick, 1970). In such situations, much of the residue N is retained by incorporation into microbial cells with some of this later converted into recalcitrant humic substances while, in contrast, the C present is progressively reduced via CO, evolution so that the C/N ratio of the residue narrows as decomposition proceeds, eventually resulting in a net release of N via mineralization. The critical N content above which net mineralization will occur is generally considered to be in the range 1.4-1.8%N or at a C/N ratio of <25-30 (Haynes, 1986). The magnitude and temporal duration of such microbial immobilization is of considerable agronomic importance, as by temporarily reducing soil mineral nitrogen supply the incorporation of such cereal residues may reduce a soil’s susceptibility to overwinter nitrate leaching by up to 25 kg N/ha (Nicholson et al., 1997) and/or restrict N supply to a succeeding crop (Thompson, 1992). Research by Smith and Sharpley (1990) detected a depression in net mineralization in the first 14 days after incorporating residues of alfalfa, corn, oat, peanut, sorghum, soybean, and wheat with C/N ratios of 16, 64,40,27, 36, 54, and 58, respectively, into eight contrasting soil types. This depression in net mineralization was enhanced when residues were mechanically incorporated into the soil rather than being left on the soil surface, and was greater for the residue N compared with the older, indigenous soil N. Other research by Goss et al., (1993) found that, although residues from oats produced an amount of mineralized nitrogen similar to wheat, oilseed rape residues released N equivalent to that from cereal residues to which 26 kg N/ha had been added. These authors concluded that the nature of the previous crop was particularly important in determining the nitrate leaching loss over the winter period, with most of the enhanced nitrate leaching losses observed under conven-

CULTIVATIONEFFECTS ON SOIL N MINERALIZATION

291

tional ploughing compared to direct drilling being derived from the mineralization of residues from the previous crop, that is, from the more labile organic N fraction. Other studies have investigated the effects of cultivation on different crop genera with characteristically lower C/N ratio residues.For example, in a study examining the effect of antecedent legume crops and tillage on the distribution and dynamics of N in a sandy loam soil in Canada, Soon and Clayton (1996) used zero tillage and conventional tillage in a randomized block design with three crop sequences: pea-wheat, red clover green manure-wheat, and wheat-wheat. No significant differences were detected in plant N uptake or yield due to tillage method or previous crop, and there was no significant difference in the amount and distribution of soil mineral nitrogen due to tillage method. Francis et al. (1992) investigated the effects of different tillage practices on changes in SMN after cultivating a temporary leguminous pasture, and similarly found no difference between the effects of mouldboard ploughing and chisel ploughing on the accumulation of mineral N or the amount of nitrate leached over winter. However, timing of cultivation proved important, with overwinter nitrate leaching losses totalling 78, 40, and 5 kg N/ha for cultivations carried out the previous March, May, and July, respectively: The experimental site was in New Zealand and therefore these months represented early autumn, late autumn, and winter, respectively. Such data concur with current UK recommendations delaying cultivations until as late as practicable to reduce nitrate leaching risk (MAFF, 1991). Hoffman et al. (1996a,b) recently quantified the effect of soil tillage on net N mineralization under sugar beet and found that although the cumulative amount of N mineralized over the growing season and the period of highest N mineralization did not differ between conventional and reduced cultivation, tillage method did affect the mineralization rate in the uppermost 30 cm of a sandy loam soil profile. The N mineralization rate was higher at 0- 10 cm in the reduced tillage soil, whereas for the conventionally tilled treatment it was consistently greater at 10-20 cm depth. Research has also compared the effect of minimal tillage of a soil previously under rotational fallow: Campbell et al. (1989) found cumulative net N mineralization over 16 weeks to be 40 kg g-I soil more under direct drilling compared with the same Chernozem soil retained under fallow. Overall, attempts to quantify the magnitude of any mineralization flush associated with cultivation exposing readily mineralizable organic material, which was previously physically protected, may be somewhat confounded by the effect of the chemical composition and management of the previous harvest’s residues, which may restrict nitrate availability, and any such effect will be more pronounced after incorporating residues with wider C/N ratios. Therefore, the chemical composition and relative abundance of N in crop residues, which may be crudely expressed using C/N ratios, play critical roles in

2 92

MARTYN SILGRAM AND MARK A. SHEPHERD

governing the residues’ relative decomposability and the consequences for soil mineral nitrogen supply. Long-term retention of stubble in no-till systems can increase the size and turnover of the microbial biomass, the respiration of organic C, and the gross mineralization of N compared with conventional cultivations, due to the accumulation of organic material at (or close to) the soil surface and modifications to prevailing soil moisture conditions. However, this may translate into less net N mineralization under zero-till conditions, especially in cereal cropping systems characterized by carbonaceous residues. More finely divided or macerated residues will decompose more rapidly as a result of the greater surface area exposed for microbial colonization and decomposition. Similarly, leaving residues on the soil surface tends to reduce their rate of decomposition compared with residue incorporation.

E. TIMING AND FREQUENCY OF CULTIVATION Soils under no-tillage management may sometimes be ploughed for crop rotation purposes or to correct a pest or soil management problem. Pierce et al. (1994) investigated whether soil properties created by long-term no-till management were retained after a single ploughing and return to no tillage for a loam soil in the United States. Compared with long-term NT, both conventional tillage CT and a single ploughing of no-tillage areas decreased bulk density by 0.17 to 0.28 Mg m-3, increased total porosity from 0.03 to 0.10 m3 m-3,increased macroporosity by 0.05 to 0.13 m3 m-3, and decreased microporosity by 0.03 to 0.05 m3 m-3. The single ploughing of no-till areas enhanced N mineralization over both CT and NT by 9.8 to 18.4 g m-3 in the surface 5 cm, and the residual effects of this single cultivation were still evident 1 year after ploughing. However, after 4 or 5 years after the single cultivation, most soil properties had returned to levels similar to those prior to disturbance, although C and N concentrations in the uppermost 5 cm of the soil still remained lower than those that had accumulated under long-term NT. Other research has indicated that the effect of cultivation on mineralization is strongly dependent on the number of years under the tillage method considered (Carter and Rennie, 1982; Staley et al., 1988), with smaller proportions of physically protected soil organic matter in soil that was ploughed annually compared with the same soil that had remained untilled for many years (Balsedent et al., 1988, 1990) because soil under long-term ploughing will have reached a stable equilibrium typically characterized by lower total C and N contents. The corollary to this is that soil recently brought into reduced tillage will typically have higher fertilizer N demand during the first few years due to initial immobilization processes (Baeumer and Kopke, 1989), with Manzke et al. (1992) reporting that 15 years were required before a new steady state between immobilization and mineralization was obtained after reverting to a zero-tillage system.

CULTNATION EFFECTS ON SOIL N MINERALIZATION

293

There are profound differences in the effect of cultivation with depth, however, with Doran (1 987) reporting that microbial biomass and potentially mineralizable nitrogen in the surface layer (0-7.5 cm) of no-till soils were 34% higher than those of ploughed soils, although the opposite was true at 7.5-15 cm depth. A similar pattern was reported by Carter and Rennie (1982), who also identified clear differences in mineralization with depth under contrasting cultivation regimes. These authors found that for 2-week incubations of four different soils (two clay loams, one loam, and one silt loam) under optimal water and temperature conditions, zero tillage resulted in cumulative C mineralization averaging +43% at 0-5 cm depth but -30% at 5-10 cm depth while cumulative N mineralization averaged +53% at 0-5 cm depth but - 18% at 5-10 cm depth, with all changes relative to measurements under conventional cultivation. This provides strong evidence that, compared to conventional cultivation, reduced tillage results in a substantial increase in mineralization potential within the near-surface zone, although this may be partly compensated for by a concomitant decline in mineralization potential in deeper subsurface layers. The timing of cultivations may also be an important consideration in governing mineralization and N supply to a succeeding crop. Cultivation when the soil is warm and moist typically leads to the greatest amount of N mineralization, with mouldboard and chisel ploughing operations resulting in reduced nitrate leaching if they were delayed until late winter (Watson et al., 1989; Vinten et al., 1994). However, this is not always a practical option, especially on heavier soils that may be waterlogged this late in the year: On such soils, delayed cultivations would risk later spring sowings and soil damage such as compaction and plough pan development due to the wet conditions (Chamen and Parkin, 1995). In a study of the effect of type and timing of cultivations on N mineralization in a shallow calcareous loam overlying chalk in Lincolnshire, UK, Stokes et al. (1992) found that a 7-week delay in soil disturbance after harvest of the preceding vining pea crop reduced nitrate concentration at 0-30 cm depth from 88 to 55 kg N/ha in mid September, with a similar experiment showing that a 3-week delay in cultivation after oilseed rape reduced nitrate levels from 262 to 150 kg N/ha at 0-30 cm depth in late October. Thus the timing of cultivation may be crucial in governing the magnitude of N release and hence its availability to subsequent crops and susceptibility to nitrate leaching. However, there is a trade-off, because if delaying cultivations leads to late establishment of the next winter sown crop, then yields may be compromised and the later crop establishment may lead to increased nitrate available for overwintering leaching. Physical protection of soil organic matter-which is thought to be an important factor in controlling mineralization potential-is increased under no-till regimes (Section IV). This will increase the size of the pool of labile organic N and C, which is physically protected but may subsequently be made available following rotational ploughing or periodic cultivation. No-till systems have significantly greater

2 94

M A R T " SILGRAM AND MARK A. SHEPHERD

potentially mineralizable N and microbial biomass in the near-surface zone, but less in subsurface layers, compared to conventionally ploughed systems. Delaying cultivations can be a successful means of reducing soil nitrate concentrations and hence limiting overwinter nitrate leaching losses, but such environmentally motivated advice must be balanced with the practical agronomic requirement to avoid adversely restricting crop establishment and final yield. Thus the timing of cultivations and their relationship to crop establishment and N demand are critical factors governing the fate of N from any mineralization flush resulting from the tillage process.

E SOILMINERAL NITROGEN AND NITRATE LEACHING Cultivation is an oxidative process since it typically promotes good aeration (provided the soil is not excessively wet) and the rapid decomposition of soil organic matter, and consequently promotes increased mineralization of organically bound nitrogen (Campbell, 1978; Haynes, 1986). Decomposition of soil organic matter is generally most rapid in the first 25 to 50 years of cultivation and reaches steady-state conditions within 50 to 100 years after conversion to arable cropping (Allison, 1973). Thus, artifact effects from previous site management such as the ploughing up of grassland may account for the higher mean SMN concentrations found in some shallow soils overlying chalk, which otherwise would be expected to have generally lower total topsoil N contents and hence N availability (Williams et al., 1996). The effect of cultivation in promoting a mineralization flush conveys considerable argonomic and environmental importance, as research on clay soils indicates that crops grown after conventional mouldboard ploughing have between two and four times greater soil solution nitrate concentrations at 30 and 60 cm depths compared to direct drilled crops, and this difference persists throughout the winter months (Dowdell and Cannell, 1975). The potential for overwinter nitrate leaching is therefore greater in ploughed soils, while direct drilled crops frequently require greater fertilizer N inputs compared to crops grown following ploughing. This is supported by Colbourn (1989, and in more recent research by Catt et al. (1992), who found that tillage increased overall nitrate leaching losses by 24% compared to direct drilling in hydrologically isolated plots in the Brimstone Farm experiment on a clay soil in Oxfordshire, UK. At the same site, Goss et al. (1993) found that conventional ploughing (20 cm depth) for autumn-sown cereal crops increased overwinter nitrate leaching losses by 21% relative to direct drilling, mainly as a result of the enhanced mineralization of soil organic matter. However, in the spring, direct drilling actually increased nitrate leaching losses following fertilizer applications (Goss et al., 1988), probably as a result of greater bypass (macropore) flow in direct drilled plots com-

CULTIVATION EFFECTS ON SOIL N MINERALIZATION

295

pared to ploughed ones, because ploughing would have disrupted the continuity and connectivity of macropore channels involved in solute transport.This slightly greater loss of spring-applied N has been proposed as one reason why larger applications of spring fertilizer N are often required in direct drilled systems to achieve yields comparable to similar fields under conventional cultivation (Cannell, 1985). In contrast, research has found that soil-derived N from the mineralization of organic N is often distributed in fine pores within the soil matrix and so is less readily displaced by water flowing in macropores (Youngs and LeedsHarrison, 1990; Goss et al., 1988, 1993). In the study reported by Goss et al. (1993), mean nitrate leaching losses for 1981-1988 (excluding 1984) under wheat were 30 kg N/ha in ploughed soil but only 23 kg N/ha in direct drilled plots, with corresponding values under oats of 43 and 30 kg N/ha, and under oilseed rape of 41 and 27 kg N/ha, respectively. Over the entire period 1981-1988, winter leaching losses of NO,-N were consistently greater by 1-2 1 kg N/ha from ploughed plots compared to those that had been direct drilled (Fig. l), although a portion of this N could have been residual fertilizer applied the previous spring rather than the mineralization of indigenous organic N resulting from the cultivation process per se. To discriminate cultivation effects with greater confidence, Goss et al. (1993) calculated apparent net mineralization using an N budget approach taking account of changes in plant N and SMN over time plus any loss of N in monitored mole drains. Apparent net mineralization was variable, but consistently less under direct drilling. For plots under winter wheat after an oat harvest, cultivation had little effect in autumn and winter 1988 when fluxes were 26 and 3 1 kg N/ha for direct drilled and ploughed plots, respectively, although a larger effect was evident when fluxes were summed over

Figure 1 Effect of cultivation on overwinter nitrate leaching. Adapted from Goss et al. (1993)

2 96

MARTYN SILGRAM AND MARK A. SHEPHERD

the whole year with only 67 kg N/ha mineralized in direct drilled plots but 83 kg N/ha in plots under conventional cultivation. Radford et al. (1992) also found that zero tillage and stubble retention reduced water use efficiency compared with conventional cultivations, with results indicating higher soil water status but lower soil nitrate concentrations under zerotillage systems. Ammonium concentrations were always negligible, but levels of NO,-N showed a reasonably consistent pattern of tillage effects over the 3-year study on a Typic Natrustalf sown to wheat, with conventional tillage with disks responsible for 14-38 kg/ha more NO,-N at sowing and 20-49 kg N/ha more NO,-N at harvest compared with zero tillage (Table I): This additional nitrate left at harvest would be vulnerable to leaching the following winter. Cameira et al. (1996) also reported lower soil nitrate concentrations under minimum tillage compared to conventional Cultivation of irrigated maize on a Fluvisol in Portugal, and attributed this to lower nitrification and mineralization in the minimum-tillage system as well as a larger loss of nitrate by denitrification due to the less aerobic soil conditions (Doran, 1980). This is plausible, as denitrification losses are often greater from undisturbed compared with ploughed soil (Linn and Doran, 1984; Staley et al., 1990), with Burford et al. (1981) reporting fluxes of 5.4-8.6 kg N,O-N ha-’ a-I from no-till plots compared with only 0.9-5.6 kg N,O-N ha-’ a-I from ploughed plots. The results reported by Goss et al. (1993) and Radford et al. (1992) are consistent with the findings of monthly SMN measurements after cultivation of five adjacent fields in an replicated trial on a clay soil in Cambridgeshire, which also sugTable I Effect of Tillage Practices on Nitrate-Nitrogen at 0 4 0 cm Depth at Sowing and Harvest(’ NO,-N, 0-60 cm depth at sowing (kg Nha) Tillage treatment

Z RB RD CB CD LSD. P = 0.05

NO,-N, 0-60 cm depth at harvest (kg Nha)

1985

I986

1987

I986

1987

37 49 69 58 71 21

44 49 60 69 82 20

16 58 90 75 90 ns

51 69 88 68

22 30 54 29 42 20

100 ns

Nore. Z, zero tillage; R, reduced tillage; C, conventional tillage; B, blade plough; D, disk plough; LSD, least-significant difference; ns, not significant. Reprinted from Soil & Tillage Research, 22, Radford et al., “Fallowing practices, soil water storage, plant-available soil nitrogen accumulation and wheat performance in South West Queensland,” p. 86, 1992, with kind permission of Elsevier Science-NL, Sara Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands.

CULTIVATION EFFECTS ON SOIL N MINERALIZATION

297

gested that cultivation may typically release up to 20 kg N/ha in the 2 months following ploughing or deep tine cultivation (ADAS, unpublished data). On a similar soil type, Dowdell et al. (1983) also reported that SMN levels in January were 12-65 kg N/ha greater at three sites subject to autumn ploughing compared with direct drilling. Despite this evidence, the disturbance caused by cultivation has not always been found to increase SMN status (Soon and Clayton, 1996), and the ultimate fate of any additional soil mineral N will be strongly dependent on soil type, which will control its susceptibility to nitrate leaching. If crops are to make use of any flush of additional mineral N made available as a result of the cultivation process, they must be planted as early as possible to enable them to use this SMN before it is leached out of the soil. Measurements by Dowdell and Cannell(l975) revealed no differences in denitrification or leaching losses of N from soils with conventional or direct drilled crops; crop N offtake explained only a third of the recorded differences in soil mineral N concentrations, with the authors concluding that the remaining differences were due to less mineralization of organic nitrogen in the direct drilled plots. Any effect of reduced tillage in conserving N may, however, only be transient. Catt et al. (1992) reported results from comparisons of direct drilled, shallow tine, and deep plough cultivations on a heavy clay soil in Oxfordshire, UK. Shallow tine cultivations decreased nitrate leaching losses and associated drainflow concentrations, but only for the first year, while once plots that had been direct drilled for 8 consecutive years were disturbed and deep cultivated they lost 23-43% (13-8.3 kg N/ha) more nitrate overwinter than plots conventionally ploughed throughout the same 8-year period. Such effects only lasted for 1 year, but influenced crop yield and N offtake for 2 years, with results indicating that although direct drilling and shallow tine cultivation can carry short-term (1-2 years) benefits, they only succeed in slowing mineralization, reducing nitrate leaching, and storing organic N temporarily with subsequent disturbance hastening mineralization rates once again.

V CONCLUSIONS Soil cultivation may influence the magnitude and temporal and spatial dynamics of mineralization by altering (i) soil temperature, (ii) soil structure, aeration, and hydraulic properties, (iii) the amount and distribution of organic residues with depth, and (iv) the degree of physical protection preventing a proportion of soil organic matter from being microbially degraded. All four of these factors are important influences on the magnitude and activity of the soil biomass. The bulk of evidence indicates that increasingly intense cultivation practices

298

MARTYN SILGRAM AND MARK A. SHEPHERD

tend to reduce the stability of soil structure. Any such effects are exacerbated by high soil water status and hence are potentially worse in finer-textured soils. Research indicates that cultivation techniques can significantly modify soil physical properties, with more intensive cultivation leading to less organic C, higher porosity, and fewer water-stable aggregates. In contrast, the accumulation of residues at or near to the soil surface means that soils under reduced- or zero-tillage systems are often cooler, wetter, and more compact than those under conventional tillage such as mouldboard ploughing, and this increased soil water retention and lower temperature can mean that crop emergence and early season growth may be delayed. Ploughing and cultivation tend to decrease aggregate size and introduce a more oxidized environment, which will accelerate organic matter mineralization. Research suggests that this is linked to a more easily mineralizable fraction of particulate organic matter associated with macroaggregates that are fractured by the physical disruption caused by tillage, allowing microorganisms to gain access to fresh, previously unavailable (physically protected) substrate. In addition to minimum- or zero-tillage systems generally possessing higher bulk densities and more water-stable aggregates near the soil surface, such cultivation systems also tend to have greater organic matter contents. There is a significant interaction with soil type: Loams and clay-rich soils with larger pools of organic C and N are characterized by much greater amounts of physically protected (and readily mineralizable) soil organic matter, some of which will be made available as a result of cultivation. Most importantly, contrasting tillage practices result in fundamentally different depth distributions of organic residues in soils, and this is largely responsible for minimum-tillage systems typically leading to an accumulation of organic C and N near the soil surface. The increase in total organic C detected after 3-10 years of NT management has been attributed to immobilization of fertilizer and crop residue N. Rates of CO, evolution appear greater following mouldboard ploughing compared to disking, chisel ploughing, or zero tillage, with small but consistent differences in efflux between treatments still evident 19 days after initial tillage. No-till plots have been found to possess 7 1- 132% greater mineralization potentials in the 0- to 5-cm surface zone compared to conventionally ploughed soils, but the opposite was the case in subsoil layers due to the incorporation of crop residues through ploughing. This increased pool of labile N under reduced tillage has implications for the rotational ploughing of such soils. However, minimum tillage and stubble retention typically reduce the levels of soil NO,-N primarily due to increased nitrogen immobilization, especially near the soil surface. The greater continuity and connectivity of vertically oriented macropores has been proposed as an explanation for the slightly greater leaching of spring fertilizer N sometimes reported from direct drilled plots. However, in general, differences in the pattern of water movement under no-till or conventional ploughing regimes

299

CULTIVATION EFFECTS ON SOIL N MINERALIZATION

typically lead to less cumulative nitrate leaching under reduced cultivation, as research suggests that the nitrate derived from mineralization of organic nitrogen is distributed in fine pores within the soil matrix and so is less readily displaced by water flowing in macropores. Conversely, tillage experiments often reveal that the more frequent and intensive the cultivation technique, the greater the level of NO,-N during the subsequent months. By reducing the mean aggregate diameter, cultivation leaves a greater surface area exposed for microbial colonization and enzymatic activity (Haynes, 1986), and the resulting enhanced mineralization can lead to greater nitrate leaching losses (if autumn ploughed) and a concomitant decline in soil organic matter content. Table I1 presents a summary of a number of reported changes in net mineralization, SMN, and nitrate leaching losses under conventional compared with reduced-tillage systems. The cultivation of arable soils induces a mineralization flush typically responsible for between two- and four-fold increases in soil solution nitrate concentrations. Reported research consistently indicates that this release of labile organic N is responsible for 5-65 kg N/ha more SMN following conventional ploughing compared with minimum cultivation, with the effect often only detectable during the first 1-2 years following cultivation. The fate of this additional pool of mineral N is strongly site-dependent (soil/crop type, weather), and may lead to an increase of 20-50% (up to 25 kg/ha/a) in nitrate leaching losses following ploughing compared with the same soil left uncultivated or direct drilled (Table 11). In contrast, net N mineralization under minimum-tillage management may be 5-25 Table I1 Overview of Reported Changes in Net N Mineralization, SMN,and Overwinter Nitrate Leaching Losses in ConventionallyPloughed Compared with Direct Drilled Soils Increase in net N mineralization (kg N/ha/a)

Increase in soil mineral N concentration

Catt er al. ( 1992)

-

-

Dowdell and Cannell (1975) Dowdell ef al. (1983) Radford et af. (1 992)

-

200-400% NO,-N

-

12-65 kgha (January) 14-38 k g h a (sowing): 20-49 kg/ha (harvest)

Powlson (1980) Goss ef al. (1993)

6 9

-

16

-

1-21 kg N/ha (21%)

ADAS (unpublished data)

20

-

-

Reference

Increase in nitrate leaching losses 2-8 kg Niha (2348%)

3 00

MARTYN SILGRAM AND MARK A. SHEPHERD

kg N/ha/a less than that found in conventionally ploughed soils, partly because the accumulation of carbonaceous crop residues at or near the surface of direct drilled soils promotes greater immobilization of available N leading to lower soil solution nitrate concentrations. Although conventional cultivation typically results in greater SMN concentrations than direct drilling, any increased SMN availability due to more mineralization under conventional ploughing is not necessarily translated into detectable differences in crop yield. Furthermore, N uptake by crops under different tillage systems do not appear to differ substantially, possibly due to the typically larger levels of SMN in the conventionally ploughed soils making them vulnerable to greater nitrate leaching losses during the winter months. However, when soil moisture is not limiting, cereal yields are generally greater under conventional rather than under a zero- or minimum-tillage system when zero or low levels of N fertilizer are applied to well-drained soils, although when higher rates of N are applied grain yields are often similar. Thus no-till soils tend to require slightly (ca. 20 kg N/ha) increased fertilizer N additions for crop yields to match those following conventional cultivations. The greater treatment effects under limited fertilization regimes are probably indicative of the greater soil mineral nitrogen supply following cultivation due to increased rates of N mineralization, with I5N research revealing the tendency for enhanced N immobilization in reduced tillage fields due to the accumulation of relatively carbonaceous residues near the soil surface. Research suggests that the mineralization flush caused by intensive cultivation can persist for several years following reversion to reduced or zero-till systems, enhancing SMN levels, increasing nitrate leaching risk, and potentially contributing additional N to subsequent crops. However, the release of N into mineral forms following the spring cultivation of fields previously under reduced-tillage regimes may not be so great as to permit reductions in fertilizer N inputs to spring-sown crops. There is limited evidence suggesting that by enhancing N mineralization in the short term, repeated conventional cultivations could lead to net N mineralization being reduced over the longer time scale (compared with zero-tillage systems) as soil organic N reserves may become progressively depleted.

VI. MANAGEMENT IMPLICATIONS In terms of environmental policy, there are indications that more intensive cultivations, such as ploughing, can increase N mineralization and soil mineral nitrogen levels by up to 65 kg N/ha in the months following tillage, compared with zero- or reduced-tillage systems. This additional pool of soil mineral nitrogen will be primarily in nitrate form and thus is potentially vulnerable to leaching, especially during the winter months and on the lighter soil types, with research suggesting increased losses of up to 25 kg N/ha/a when a soil under reduced cultiva-

CULTIVATION EFFECTS ON SOIL N MINERALIZATION

3 01

tion is conventionallyploughed. The widespread adoption of direct drilling or other alternative reduced cultivation practices would therefore result in a reduction in nitrate leaching losses in the short term, and this would lead to smaller loadings of nitrate being leached into ground and surface water bodies. Research suggests that such reduced cultivation practices would not, in general, result in significant losses in crop yield if minimal tillage is restricted to appropriate soil types, with the majority of evidence suggesting that the effects of such contrasting cultivation practices on yield are generally negligible. Research indicates that the mineralization flush resulting from the cultivation process is influenced by soil type and by the timing and frequency of cultivation practices. Recent evidence indicates that any increase in N mineralization following cultivation may be relatively short-lived, and this suggests that sowing should be undertaken as soon as possible after deep cultivations to maximize the utilization of additional mineral N by the developing crop. Further work is required to characterizethe timing of N release following different cultivation methods so that the dynamics of the flush of additional SMN can be synchronized as far as possible with crop N demands: Efficient utilization of the additional SMN released by cultivation has agronomic importance through the improved precision associated with the use of fertilizer N, and environmental significance by restricting the potential for nitrate leaching. This review supports current UK advice for decreasing nitrate leaching by delaying cultivations in the autumn until as late as practicable (MAW, 1991). Although minimum cultivation methods will decrease nitrate leaching losses in the autumn and winter months, this technique may not be appropriate on lighter-textured soils with relatively high proportions of silt and fine sand particles due to their propensity for structural instability and compaction. However, innovative techniques have been developed for establishing sugar beet on light soils by direct drilling (for erosion control) and such approaches should be explored for autumn crops. Cereals following the harvest of potatoes provide another opportunity for a rapid establishment of the next crop to decrease nitrate loss. This review has shown that there can be large differences in soil N supply following the two extremes of cultivation (no till and mouldboard ploughing) of as much as 65 kg N/ha: This is sufficiently large to be agronomically significant, and allowance should therefore be made for this additional pool of plant-available N when planning future fertilizer recommendations.However, there is relatively little published information on the effect of the rotational ploughing of no-till fields (which may be undertaken periodically to resolve a pest or soil management problem) or the effect of the reversion of no-tillage sites back into conventional cultivation on temporal patterns of N mineralization and the consequent implications for soil and water quality and soil fertility. The influence of cultivation date after set-aside* also warrants further investigation. *Land temporarily taken out of crop production under the EU Common Agricultural Policy.

3 02

MARTYN SILGRAM AND MARK A. SHEPHERD

There is considerable evidence to suggest that due to typically greater N immobilization and slightly increased denitrification losses, reduced-tillage systems may require a small amount (up to 25 kg/ha) of additional fertilizer N in order to equal the yields under conventional cultivations: This need for supplementary seedbed N was acknowledged in earlier versions of UK fertilizer recommendations, but was not included in the most recent update (MAW, 1995). Additional practical research is required to improve the mineralization component of fertilizer recommendations for these two extremes of cultivation (no till vs ploughing); The need for such research has been recently acknowledged (Shepherd et al., 1996) and would have agronomic benefits in terms of more efficient use of fertilizer N and associated environmental benefits by reducing nitrate leaching risk. However,refinements for cultivation systems within these two extremes would not be feasible given the current state of knowledge on the temporal dynamics of soil mineral nitrogen supply. An overview of the key findings of this review are summarized in Table 111. Reduced cultivation systems can significantly reduce nitrate leaching losses and may also contribute to reduced erosion risk. However, one of the most pertinent questions is the medium- to long-term sustainability of such minimum-tillage systems and the ultimate fate of the observed increases in soil organic C and N. Is this Table III Summary of the Typical Effects of Conventional and Reduced Cultivations Conventional tillage

Reduced tillage

Reduced aggregate stability Greater aggregate stability Reduced mean aggregate diameter Increased mean aggregate diameter Higher porosity Lower porosity Lower bulk density Higher bulk density Reduced soil pore continuitykonnectivity Greater soil pore continuitykonnectivity Reduced physical protection of SOM Greater physical protection of SOM Increased erosion risk Reduced erosion risk Warmer, drier soil Cooler, wetter soil Biomass dominated by bacteria Biomass dominated by fungi Fewer earthworms and soil arthropods More earthworms and soil arthropods Lower organic matter content Higher organic matter content Less total organic C and N More total organic C and N Residue C and N incorporated to depth Residue C and N accumulated at surface Lower mineralization potential Greater mineralization potential Increased nitrogen mineralization Increased nitrogen immobilization Increased SMN concentrations Reduced SMN concentrations Increased nitrate leaching risk Reduced nitrate leaching risk Follow standard ferilization guidelines Small additional fertilizer N required Similar yields achievable under optimal moisture and fertilization regimes

CULTIKATIONEFFECTS ON SOIL N MINERALIZATION

303

additional N incorporated into more stabilized forms with a beneficial effect on soil structure and aggregate stability, or does it remain more labile and simply serve as a temporary store for a pool of readily mineralizable N that may pose future environmental problems when it is remineralized once the land is ploughed again? This sustainability issue is an important one that needs to be addressed before greater confidence can be placed in management solutions, such as reducedtillage systems, aimed at reducing nitrate leaching risk while maintaining underlying soil fertility and yield potential.

ACKNOWLEDGMENT We are grateful to the Ministry of Agriculture, Fisheries and Food, London, for providing funding for the preparation of this review.

REFERENCES Addiscott, T. M., and Dexter, A. R. (1994). Tillage and crop residue management effects on losses of chemicals from soils. Soil Tillage Res. 30, 125-168. Adu, J. K.,and Oades, J. M. (1978). Utilisation of organic materials in soil aggregates by bacteria and fungi. Soil Biol. Biochem. 10, 117-122. Allison, F. E. (1973). “Soil Organic Matter and Its Role in Crop Production.” Elsevier, Amsterdam. Amato, M., and Ladd, J. N. (1980). Studies of nitrogen immobilization and mineralization in calcareous soils. V. Formation and distribution of isotope-labelled biomass during decomposition of I4C and I5N labelled plant material. Soil Biol. Biochem. 12,405-41 1. Anderson, J. M. (1 988). The role of soil fauna in agricultural systems. In “Advances in Nitrogen Cycling in Agricultural Ecosystems” (J. R. Wilson, ed.), Proceedings of the Symposium on Advances in Nitrogen Cycling in Agricultural Ecosystems, Brisbane. CAB International, Wallingford, UK. Anderson, J. P. E., and Domsh, K. H. (1975). Measurement of bacterial and fungal contribution to respiration of selected agricultural and forest soils. Can.J. Microbiol. 21, 314-322. Arvidsson, J., and Hikansson, I. (1996). Do effects of soil compaction persist after ploughing? Results from 21 long-term field experiments in Sweden. Soil Tillage Res. 39,175-197. Aulakh, M. S., Rennie, D. A,, and Paul, E. A. (1984). Gaseous nitrogen losses from soils under zerotill as compared with conventional till management systems. J. Environ. Qual. 13, 130-136. Baeumer, K., and Kopke, U. (1989). Effects of nitrogen fertilization. In “Energy Saving by Reduced Tillage” (K. Baeumer and W. Ehlers, eds.). Brussels. Balabane, M. (1 996). Turnover of clay-associated organic nitrogen in the different aggregate-size classes of a cultivated silty loam. Eur: J. Soil Sci. 47,285-291. Balesdent, J. (1996). The significance of organic separates to carbon dynamics and its modelling in some cultivated soils. Eur: J. Soil Sci. 47,485-493. Balesdent, J., Wagner, G. H., and Mariotti, A. (1988). Soil organic matter turnover in long-term field experiments as revealed by the carbon- 13 natural abundance. Soil Sci. Soc. Am. J. 52, 1 18- 124. Balesdent, J., Mariotti, A,, and Boisgontier, D. (1990). Effect of tillage on soil organic carbon mineralization estimated from I3C abundance in maize fields. J. Soil Sci. 41,587-596. Ball, B. C., Cheshire, M. V., Robertson, E. A. G., and Hunter, E. A. (1996). Carbohydrate composition in relation to structural stability, compactability and plasticity of two soil in a long-term experi‘ . ment. Soil Tillage Res. 39, 143-160.

3 04

MARTYN SILGRAM AND MARK A. SHEPHERD

Barber, D. A., and Standell, C. J. (1977). Preliminary observations on the effect of direct drilling on the microbial activity of soil. Agricultural Research Council Letcombe Laboratory Annual Report for 1976, pp. 58-60. Barraclough, D., and Pun, G.(1995). The use of I5N pool dilution and enrichment to separate the heterotrophic and autotrophic pathways of nitrification. Soil Biol. Biochem. 27, 17-22. Beare, M. H., Hendrix, P. F., and Coleman, D. C. (1994). Water-stable aggregates and organic matter fractions in conventional and no-tillage soils. Soil Sci. SOC.Am. J. 58,777-786. Bennett, 0.L., Mathias, E. L., and Lundberg, P. E. (1973). Crop responses to no-till management practices on hilly terrain. Agron. J. 65,488-49 1. Besnard, E., Chenu, C., Balesdent, J., Puget, P., and Arrouays, D. (1996). Fate of particulate organic matter in soil aggregates during cultivation. Eur: J. Soil Sci. 47,495-503. Bhogal, A., and Shepherd, M. (1 996). Effect of residue incorporation and nitrogen fertilizer on the measurement of nitrogen mineralization in field soils. Transacrions of the 9th Nitrogen Workshop, Braunsweig, Germany, September, pp. 185- 187. Bhogal, A., Hatch, D. J., Shepherd, M. A., and Jarvis, S. C. (1998). Comparison of methodologies for field measurement of net N mineralization in arable soils, method development. Plant and Soil (in press). Blevins, R. L., Thomas, G.W.. Smith, M. S., Frye, W. W., and Cornelius, P. L. (1983). Changes in soil properties after LO years non-tilled and conventionally tilled corn. Soil Tillage Res. 3, 135-146. Bondietti, E., Martin, J. P., and Haider, K. (1971). Influence of nitrogen source and clay on growth and phenolic polymer production by Stachybotrys spp., Hendersonula toruloidea and Aspergillus sydowi. Soil Sci. SOC.Am. Proc. 35,917-922. Breland, T. A. (1994). Enhanced mineralization and denitrification as a result of heterogeneous distribution of clover residues in soil. Plant Soil 166,l-12. Brookes, P. C., Landman, A., Pruden, G.,and Jenkinson, D. S. (1985). Cholorform fumigation and the release of soil nitrogen, a rapid direct extraction method to measure microbial biomass nitrogen in soil. Soil Biol. Biochem. 17,837-842. Brown, P., and Dickey, D. D. (1970).Losses of wheat straw residue under simulated field conditions. Soil Sci. SOC.Am. P roc. 34, 1 1 8- 121. Burford, J. R., Dowdell, R. J., and Crees, R. (I98 I). Emission of nitrous oxide to the atmosphere from direct drilled and ploughed clay soils. J. Sci. FoodAgric. 32,219-223. Burt, T. P., Heathwaite, A. L., and Trudgill, S. T. (1993). “Nitrate Processes Patterns & Management.” Wiley, Chichester, UK. Cabrera, M. L., and Kissel, D. E. (1988). Potentially mineralizable N in disturbed and undisturbed soil samples. Soil Sci. SOC.Am. J. 52, 1010-1015. Cambardella, C. A., and Elliot, E. T. (1994). Carbon and nitrogen dynamics of soil organic matter fractions from cultivated grassland soils. Soil Sci. SOC.Am. J. 58, 123-130. Cameira, M. C., Magalhles, M. C., and Pato, R. L. (1996). Nitrogen dynamics on a soil with different tillage systems. In “Progress in Nitrogen Cycling Studies” (0.Van Cleemput et al., eds.), pp. 31-35. Kluwer, London. Cameron, R. S., and Possner, A. M. (1979). Mineralizable organic nitrogen in soil fractionated according to particle size. J. Soil Sci. 30,565-577. Campbell, C. A. (1978). Soil organic carbon, nitrogen and fertility. In ‘‘Soil Organic Matter” (M. Schnitzer and S. U.Khan, eds.), pp. 174-271. Elsevier, New York. Schnitzer, M., Selles, F., and Zentner, R. P. (1989). Effect of six Campbell, C. A., Biederbeck, V. 0.. years of zero tillage and N fertilizer management on changes in soil quality of an Orthic Brown Chemozem in Southwestern Saskatchewan. Soil Tillage Res. 14,39-52. Cannell, R. Q. (1985). Reduced tillage in north-west Europe-Areview. Soil Tillage Res. 5,129-177. Carter, M. R. (1987). Physical properties of some Prince Edward Island soils in relation to their tillage requirement and suitability for direct drilling. Can. J. Soil Sci. 67,473-487.

CULTIVATIONEFFECTS ON SOIL N MINERALIZATION

305

Carter, M. R. (1991a). Evaluation of shallow tillage for spring cereals on a fine sandy loam. I. Growth and yield components, N accumulation and tillage economics. Soil Tillage Res. 21,23-35. Carter, M. R. (1991b). Evaluation of shallow tillage for spring cereals on a fine sandy loam. 11. Soil physical, chemical and biological properties. Soil Tillage Res. 21,37-52. Carter, M. R. (1992). Influenceof reduced tillage systems on organic matter, microbial biomass, macroaggregate distribution and structural stability of the surface soil in a humid climate. Soil Tillage Res. 23,36 1-372. Carter, M. R., and Rennie, D. A. (1982). Changes in soil quality under zero tillage farming systems, distribution of microbial biomass and mineralizable C and N potentials. Can. J. Soil Sci. 62,587-597. Carter, M. R., and Rennie, D. A. (1985). Spring wheat growth and 15N studies under zero and shallow tillage on the Canadian prairie. Soil Tillage Res. 5,273-288. Catt, J. A., Christian, D. G., Goss, M. J., Harris, G. L., and Howse, K. R. (1992). Strategies to reduce nitrate leaching by crop rotation, minimal cultivation and straw incorporation in the Brimstone Farm Experiment, Oxfordshire. Aspects Appl. Biol. 30,255-262. Chamen, W. C. T., and Parkin, A. B. (1995). The impact of tillage practices on the soil and its linked environment, with particular reference to arable cropping in the United Kingdom, ADAS-SRI Joint Contract Report No. CR/678/95/0209. Chichester, F. W. (1969). Nitrogen in soil organo-mineral sedimentation fractions. Soil Sci. 107, 356-363. Colbourn, P. (1985). Nitrogen losses from the field, denitrification and leaching in intensive winter cereal production in relation to tillage method on a clay soil, 1984-5, Divisional Note DN1318, Silsoe Research Institute, Silsoe, Bedfordshire, UK. Degens, B. P., and Harris, J. A. (1997). Development of a physiological approach to measuring the catabolic diversity of soil microbial communities. Soil Biol. Biochem. 29, 1309- 1320. Department of the Environment (DOE) (1988). “The Nitrate Issue.” HMSO, London. Dick, W. A,, McCoy, E. L., Edwards, W. M., and Lal, R. (1991). Continuous application of no-tillage to Ohio soils. Agron. J. 83,65-73. Dick, W. A., Edwards, W. M., Stehouwer, R. C., and Eckert, D. J. (1992). Maize yield and nitrogen uptake after established no-tillage fields are ploughed. Soil Tillage Res. 24 (I), 1-10. Doran, J. W. (1980). Soil microbial and biochemical changes associated with reduced tillage. Soil Sci. Soc. Am. J. 44,765-77 1. Doran, J. W. (1987). Microbial biomass and mineralizahle nitrogen distributions in no-tillage and ploughed soils. Biol. Ferril. Soils 5, 68-75. Douglas, J. T. (1977). The effect of cultivation on the stability of aggregates from the soil surface. Agricultural Research Council Letcombe Laboratory Annual Report for 1976, pp. 46-48. Douglas, J. T., and Goss, M. J. (1982). Stability and organic matter content of surface soil aggregates under different methods of cultivation and in grassland. Soil Tillage Res. 2, 155-175. Douglas, J. T., Jarvis, M. G., Howse, K. R., and Goss, M. J. (1986). Structure of a silty soil in relation to management. J. Soil Sci. 37, 137- I5 1. Dowdell, R. J., and Cannell, R. Q. (1975). Effect of ploughing and direct drilling on soil nitrate content. J. Soil Sci 26,53-61. Dowdell, R. J., and Webster, C. P. (1984). A lysimeter study of the fate of fertilizer nitrogen in spring barley crops grown on a shallow soil overlying Chalk: Denitrification losses and the nitrogen balance. J. Soil Sci. 35, 183-190. Dowdell, R. J., Crees, R., Burford, J. R., and Cannell, R. Q. (1979). Oxygen concentration in a clay soil after ploughing or direct drilling. J. Soil Sci. 30,239-245. Dowdell, R. J., Crees, R., and Cannell, R. Q. (1983). A field study of effects of contrasting methods of cultivation on soil nitrate content during autumn, winter and spring. J. Soil Sci. 34,367-379. Edwards, C. A,, and Fletcher, K. E. (1988). Interactions between earthworms and microorganisms in organic-matter breakdown. Agn‘c. Ecosystems Environ. 24,235-247.

3 06

MARTYN SILGRAM AND MARK A. SHEPHERD

Edwards, C. A., and Lofty, J. R. (1982). Effects of direct drilling and minimal cultivation on earthworm populations in agricultural soils. J. Appl. Ecol. 19,723-734. Ekeberg, E. (1992). Reduced tillage on loam soil. Soil investigation. Norwegian Agric. Res. 6,223-244. Ekeberg, E., and Riley, H. C. F. (1996). Effects of mouldboard ploughing and direct planting on yield and nutrient uptake of potatoes in Norway. Soil Tillage Res. 39, 131-142. Elliot, E. T., and Coleman, D. C. (1988). Let the soil work for us. Ecol. Bull. (Copenhagen) 39,23-32. Elliot, P., Knight, D., and Anderson, J. M. (1990). Denitrification in earthworm casts and soil from pastures under different fertilizer and drainage regimes. Soil Biol. Biochem. 22,601-605. Ellis, F. B., Graham, J. P., and Christian, D. G. (1982). Interacting effects of tillage method, nitrogen fertilizerand secondary drainageon winter wheat production on a calcareousclay soil. J. Sci. Food Agric. 34,1068-1076. El Titi, A,, and Landes, H. (1990). Integrated farming system of Lautenbach, a practical contribution toward sustainable agriculture in Europe. In (C. A. Edwards, R. Lal, P. Madden, R. H. Miller, and G. House, eds.) “Sustainable Agricultural Systems”, pp. 265-286. Soil & Water Conservation Society, Iowa. Featherstone, A. M., Fletcher, J. J., Dale, R. F,, and Sinclair, H. R. (1991). Comparisons of net returns under alternative tillage systems considering spatial weather variability. J. Prod. Agric. 4, 166- 173. Fortin, M. C., Rochette, P., and Pattey, E. (1996). Soil carbon dioxide fluxes from conventional and no.tillage small-grain cropping systems. Soil Sci. SOC.Am. J. 60, 1541-1547. Foster, R. C. (1986). In situ identificationof organic materials in soils. Quesriones Entomol. 21,609-633. Francis, G. S., Haynes, R. J., Sparling, G. P., Ross, D. J., and Williams, P. H. (1992). Nitrogen mineralization, nitrate leaching and crop growth following cultivation of a temporary leguminous pasture in autumn and winter. Fertil. Res. 33,59-70. Frankenberger, W. T., Jr., and Abdelmagid, H. M. (1985). Kinetic parameters of nitrogen mineralization rates of leguminous crops incorporated into soil. Planr Soil 87,257-27 1. Franzluebbers, A. J., Hons, F. M., and Zuberer, D. A. (1995). Tillage-inducedseasonal changes in soil physical properties affecting soil CO, evolution under intensive cropping. Soil Tillage Res. 34, 41-60. Franzluebbers, A. J., Haney, R. L., Hons, F. M., and Zuberer, D. A. (1996). Active fractions of organic matter in soils with different textures. Soil Eiol. Biochem. 28, 1367-1372. Freebaim,D. M., Hancock, N. H., and Lott, S. C. (1987). Soil evaporation studiesusing shallow weighing lysimeters, techniques and preliminary results. Trans. Mech. Eng. Inst. Aust. 12,67-72. Garland, J. L. (1996). Patterns of potential C source utilization by rhizosphere communities.Soil Biol. Biochern. 28,55-63. Garland, J. L., and Mills, A. L. (1991). Classification and characterization of heterotrophic microbial communitieson the basis of patterns of community-levelsole-carbon-source utilization.Appl. Environ. Microbiol. 57,2351-2359. Goss, M. J., Howse, K. R., and Harris, W. (1978). Effects of cultivation on soil water retention and water use by cereals in clay soils. J. Soil Sci. 29,475-488. Goss, M. J., Colbourn,P.,Harris, G. L., and Howse, K. R. (1988). Leaching of nitrogen under autumnsown crops and the effects of tillage. In “Nitrogen Efficiency in Agricultural Soils (D. s.Jenkinson and K. A. Smith, eds. ) pp. 269-282. Elsevier, London. Goss, M. J., Howse, K. R., Lane, P. W., Christian, D. G., and Hams, G. L. (1993). Losses of nitrate-nitrogen in water draining from under autumn-sown crops established by direct drilling or mouldboard ploughing. J. Soil Sci. 44,35-48. , C. (1994). Nitrous oxide from agriculture.Norwegian J. Agric. Sci. (SupGranli, T., and B ~ k m a n0. PI. 12), 56-57. Gregorich,E. G., Kachanoski, R. G., andvoroney, R. P.(1989). Carbon mineralizationin soil size fractions after various amounts of aggregate disruption. J. Soil. Sci. 40,649-659. Griffith, D. R., Kladivko, E. J., Mannering, J. V., West, T. D., and Parsons, S. D. (1988). Long term

CULTIVATION EFFECTS ON SOIL N MINERALIZATION

307

tillage and rotation effects on corn growth and yield on high and low organic matter, poorly drained soils. Agron. J. 80,599-605. Gupta, V. V. S. R., and Germida, J. J. (1988). Distribution of microbial biomass and its activity in different soil aggregate size classes as affected by cultivation. Soil BioLBiochem. 50,777-786. Hargrove, W. L. (1985). Influence of tillage on nutrient uptake and yield of corn.Agmn. J. 77,763-768. Harper, S. H. T., and Lynch, J. M. (198 1). J. Soil Sci. 32,627. Hassink, J. (1992). Effects of soil texture and structure on carbon and nitrogen mineralizationin grassland soils. Biol. Fertil. Soils 14, 126-134. Hassink, J., Scholefield, D., and Blantern, P. (1990). Nitrogen mineralization in grassland soils. In “Proceedings of the 13th General Meeting of the European Grassland Federation” (N. Gaborcik, V. Krajcovic, and M. Zimkova, eds.), Vol. 11, Banscka Bystrica, Tsjechoslovakia.pp. 25-32. Hatch, D. J., Jarvis, S. C., and Philipps, L. (1990). Field measurement of nitrogen mineralization using soil core incubation and acetylene inhibition of nitrification. Plant Soil 124,97-107. Hatch, D. J., Jarvis, S. C., and Reynolds, S. E. (1991). An assessment of the contribution of net mineralization to N cycling in grass swards using a field incubation method. Plant Soil 138,23-32. Havlin, J. L., Kissel, D. E., Maddux, L. D., Claasen, M. M., and Long, J. H. (1990). Crop rotation and tillage effects on soil organic carbon and nitrogen. Soil Sci. SOC.Am. J. 54,448-452. Hay, R. K. M., Holmes, J. C., and Hunter, E. A. (1978). The effects of tillage, direct drilling and nitrogen fertilizer on soil temperature under a barley crop. J. Soil, Sci. 29, 174-183. Haynes, R. J. (1986). “Mineral Nitrogen in the Soil-Plant System,” PhysiologicalEcology-A series of monographs, texts, and treatises. Academic Press, London. Herbek, J. H., Mudock, L. W., and Blevins, R. L. (1986). Tillage systems and date of planting effect on yield of corn on soils with restricted drainage. Agron. J. 78,824-826. Hill, R. L. (1990). Long-term conventional and no-tillage effects on selected soil physical properties. Soil Sci. SOC.Am. J. 54, 161-166. Hoffman, C., Lindtn, S., and Koch, H.-J. (1996a). Influence of soil tillage on net N mineralization under sugar beet. Z. Pjunzenernahr Bodenk. 159,79-85. Hoffman, C., Platte, H., Lickfett, T., and Koch, H.-J. (1996b). The role of microbial biomass for the N mineralization in a field incubation experiment with reduced tillage. Transactions of the 9th Nitrogen Workshop, Braunsweig, Germany, September, pp. 243-246. Imholte,A. A., and Carter, P. R. (1987). Planting date and tillage effects on corn following corn. Agron. J. 79,746-75 1. Iritani, W. M., and Arnold, C. Y. (1960). Nitrogen release of vegetable crop residues during incubation as related to their chemical composition. Soil Sci. 89,74-82. Jarvis, S. C., Stockdale, E. A,, Shepherd, M. A,, and Powlson, D. S. (1996). Nitrogen mineralization in temperature agricultural soils: Processes and measurement. Adv. Agron. 57, 187-237. Jenkinson, D. S. (1990). The turnover of organic carbon and nitrogen in soil. Philos. Trans. R. SOC. London B 329,361-368. Jenkinson, D. S., and Ladd, J. N. (1981). In “Soil Biochemistry” (A. E. Paul and J. N. Ladd, eds.), Vol. 5 , p. 415. Dekker, New York. Jenkinson, D. S., and Powlson, D. S. (1976). The effects of biocidal treatments on metabolism in soil. V. A. method for measuring soil biomass. Soil Biol. Biochem. 8,209-321. Jenkinson, D. S., Powlson, D. S., and Wedderburn, R. W. M. (1976). The effects of biocidal treatments on metabolism in soil. 111. The relationship between soil biovolume, measured by optical microscopy, and the flush of decomposition caused by fumigation. Soil Biol. Biochem. 8,189. Jenny, H. (1941). “Factors of Soil Formation.” McGraw-Hill, New York. Jocteur-Monrozier. L., Ladd, J. N., Fitzpatrick, R. W., Foster, R.C., and Raupach, M. (1991). Components and microbial biomass content of size fractions in soils of contrasting aggregation. Geodenna 49,37-62. Johnson, M. D., and Lowery, B. (1985). Effect of three conservation tillage practices on soil temperature and thermal properties. Soil Sci. SOC. Am. J. 49, 154771552,

3 08

MAEtTYN SILGRAM AND MARK A. SHEPHERD

Kandeler, E., and Bghm, K. E. (1996). Temporal dynamics of microbial biomass, xylanase activity, N mineralization and potential nitrification in different tillage systems. Appl. Soil. Ecoi. 4, 18 1 - I9 1. Kapusta, G., Krausz, R. F., and Matthews, J. L. (1996). Corn yield is equal in conventional, reduced and no tillage after 20 years. Agron. J. 88,812-817. Kemper, W. D., Stewart, B. A,, and Porter, L. K. (1971). Effects of compaction on soil nutrient status. In “Compaction ofAgricultural Soils’’ (K. K.Barnes, W. M. Carleton, H. M. Taylor, R. I. Throckmorton, and G. E. van den Berg, eds.), pp. 178-1 89. American Society ofAgricultura1 Engineers, St. Joseph, MO. King, J. A. (1997). The development of methods for the measurement of carbon fluxes from soils and vegetation, final report to MAFF on seedcorn project No. VLAFDIAI2. ADAS Boxworth, Boxworth, Cambridge, UK. King, J. A., and Harrison, R. (1995). An automated chamber system to measure soil respiration. Aspects Appl. Biol. 43,87-94. Kitur, B. K., Smith, M. S.,Blevins, R. L., and Frye, W. W. (1984). Fate of I5N depleted ammonium nitrate applied to no-tillage and conventional tillage corn. Agron. J. 76,240-242. Kladivko, E. J., Griffith, D. R., and Mannering, J. V. (1986). Conservation tillage effects on soil properties and yield of corn and soybeans in Indiana. Soil Tillage Res. 8,277-287. Knight, D., Elliot, P. W., Anderson, J. M., and Scholefield, D. (1992).The role of earthworms in managed permanent pastures in Devon, England. Soil B i d . Biochem. 24, 1511-1517. Kuo, S., Sainju, U. M., and Jellum, E. (1996). Winter cover crop influence on nitrogen mineralization, presidedress soil nitrate test, and corn yields. Biol. Fertil. Soils 22,3 10-3 17. Ladd, J. N., Amato, M., Jocteur-Monrozier. L., and Van Gestel, M. (1990). Soil microhabitats and carbon and nitrogen metabolism. In “Proceedings of the 14th International Congress of Soil Science, August 12-18 1990, Kyoto, Japan,” Vol. III, pp. 82-87. Ladd, J. N., Jocteur-Monrozier, L., and Amato, M. (1992). Carbon turnover and nitrogen transformations in an Alfisol and Vertisol amended with U-I4C glucose and I5N ammonium sulphate. Soil Biol. Biochem. 24,359-371. Lal, R. (1991). Tillage and agricultural sustainability. Soil Tillage Res. 20, 133-146. Lee, K. E. and Pankhurst, C. E. (1992). Soil organisms and sustainable productivity. Ausrr: J. Soil Res. 30,855-892. Lessa, A. S. N., Anderson, D. W., and Chatson, B. (1996). Cultivation effects on the nature of organic matter in soils and water extracts using CPIMAS 13CNMR spectroscopy. Plant Soil 184,207-21 7. Linn, D. M., and Doran, J. W. (1984). Effect of water-filled pore space on carbon dioxide and nitrous oxide production in tilled and nontilled soils. Soil Sci. Soc. Am. J. 48, 1267-1272. Lynch, I. M., and Panting, L. M. (1980).Cultivation and the soil biomass. Soil Biol. Biochem. 12,29-33. Lynch, J. M., and Panting, L. M. (1982). Effects of season, cultivation and nitrogen fertilizer on the size of the soil microbial biomass. J. Sci. FoodAgric. 33,249-252. Mackay, A. D., and Barber, S. A. (1984).Soil temperature effects on root growth and phosphorus uptake by corn. Soil Sci. SOC.Am. J. 48,818-823. Mackay, A.D., and Kladivko, E. J. (1985). Earthworms and rate of breakdown of soybean and maize residues in soil. Soil Biol. Biochem. 17,851-857. Ministry of Agriculture Fisheries and Food (MAFF) (1991). “Code of Good Agricultural Practice for the Protection of Water.” HMSO, London. Ministry of Agriculture Fisheries and Food (MAFF) (1995). “Fertilizer Recommendations for Agricultural and Horticultural Crops,” Reference Book No. 209. HMSO, London. Mahboubi, A. A., Lal, R., and Faussey, N. R. (1993). Twenty-eight years of tillage effects on two soils in Ohio. Soil Sci. SOC.Am. J. 57,506-512. Manzke, F., Claupein, W., and Baeumer, K. (1992). Ermittlung der Stickstoff-Nettomineralization im Feld in unterschiedlichen Bodennutzungs-systemen. Mitf. Ges. Pjanzenbauwiss. 5,221 -224. Martin, J. P., Filip, Z., and Haider, K. (1976). Effect of montmorillonite and humate on growth and metabolic activity of some actinomycetes. Soil Biol. Biochem. 8,409-413.

CULTIVATION EFFECTS ON SOIL N MINERALIZATION

309

McLaren, R. G., and Cameron, K. C. (1990). “Soil Science.” Oxford University Press, Oxford. Meisinger, J. J., Bandel, V. A., Stanford, G., and Legg, J.O. (1985). Nitrogen utilisation of corn under minimum tillage and mouldboard plough tillage. I. Four year results using labelled N fertilizer on an Atlantic coastal plain soil. Agron. J. 77,602-61 1. Moore, A. W. (1974). Availability of rhodegrass (Choris gayana) nitrogen in tops and roots added to soil. Soil Biol. Biochem. 6,249-255. Nagel, R. F., F r o m , H., and Beese, F. (1993). The influence of earthworms and soil mesofauna on the C and N mineralization in agricultural soils-A microcosm study. Acta Zool. Fenn. Nicholson, F. A., Chambers, B. J., Mills, A. R., and Strachan, P. J. (1997). Effects of repeated straw incorporation on crop fertilizer nitrogen requirements, soil mineral nitrogen supply and nitrate leaching losses. Soil Use Management 13, 136-142. Nordmeyer, H., and Richter, J. (1995). Incubation experiments on nitrogen mineralization in loess and sandy soils. Plant Soil 83,433-445. O’Donnell, A. G. (1997). Microbial diversity and the maintenance of ecosystem services: Towards an understanding of community structure:function relationships in soils. In “Proceedings of British Soil Science Society Golden Jubilee Meeting (R. Shiel, ed.) pp. 41-42. University of Newcastle upon Tyne, UK. Patterson, D. E., Chamen, W. C. T., and Richardson, C. D. (1980). Long term experiments with tillage systems to improve the economy of cultivation for cereals. J. Agric. Eng. Res. 25, 1-35. Pierce, F. J., Fortin, M. C., and Staton, M. J. (1994). Periodic ploughing effects on soil properties in a no-till farming system. Soil Sci. Soc. Am. J. 58, 1782-1787. Powlson, D. S. (1980). Effect of cultivation on the mineralization of nitrogen in soil. Plant Soil 57, 151-153. Puget, P., Chenu, C., and Balesdent, J. (1995). Total and young organic carbon distributions in aggregates of silty cultivated soils. Eur J. Soil Sci. 46,449-459. Quemada, M., and Cabrera, M. L. (1995). CERES-N model predictions of nitrogen mineralized from cover crop residues. Soil Sci. Am. J. 59, 1059-1065. Quemada, M., and Cabrera, M. L. (1996). Temperature and moisture effects on C and N mineralization from clover residues. Transactions of the 9th Nitrogen Workshop, Braunsweig, Germany, September, p. 303. Radford, B. J., Gibson, G.. Nielsen, R. G. H., Butler, D. G., Smith, G. D., and Orange, D. N. (1992). Fallowing practices, soil water storage, plant-available soil nitrogen accumulation and wheat performance in South West Queensland. Soil Tillage Res. 22,73-93. Rasmussen, P. E., and Collins, H. P. (1991). Long-term impacts of tillage, fertilizer and crop residue on soil organic matter in temperate semiarid regions. Adv. Agron. 45,93-134. Reicosky, D. C., Dugas, W. A., and Torbert, H. A. ( I 997). Tillage-induced soil carbon dioxide loss from different cropping systems. Soil 7i:llage Res. 41, 105-1 18. Rice, C. W., Smith, M. S., and Blevins, R. L. (1986). Soil nitrogen availability after long-term continuous no-tillage and conventional tillage corn production. Soil Sc. SOC.A n . J. 50,1206-1210. Richter, J., Nuske, A,, Habenicht, W., and Bauer, J. (1982). Optimized N mineralization parameters of loess soils from incubation experiments. Plant Soil 68,379-388. Ritz, K., and Robinson, D. (1988). Temporal variations in soil microbial biomass C and N under a spring barley crop. Soil Biol. Biochem. 20,625-630. Rovira, A. D., Smettem, K. R. J., and Lee, K. E. ( I 987). Effect of rotation and conservation tillage on earthworms in a red-brown earth under wheat. Austr J. Agric. Res. 38,829-834. Rowse, H. R., and Stone, D. A. (1980). Deep cultivation of a sandy clay loam, I. Effects on growth, yield and nutrient content of potatoes, broad beans, summer cabbage and red beet in 1977. Soil Tillage Rex 1,57-68. Saffigna, P. G., Powlson, D. S., Brookes, P. C., and Thomas, G. A. (1989). Influence of sorghum residues and tillage on soil organic matter and soil microbial biomass in an Australian Vertisol. Soil Biol. Biochem. 21.759-165.

3 10

MARTYN SILGRAM AND MARK A. SHEPHERD

Schoder, D., Gehlen, P., and Peters, H. (1989). Bodenmikrobielle Aktivitat in gesiebten, aggregierten und ungestorten Proben, Kongrepband 101, VDLUFA-Kongrep Bayreuth, VDLUFA-Schriftreihe 3011990.461-466. Scott Russell, R., Cannell, R. Q., and Goss, M. J. (1975). Effects of direct dnlling on soil conditions and root growth. Ourlook Agric. 8,227-232. Scullion, J., Mohammed, A. R. A., and Ramshaw, G. A. (1991). The effect of cultivation on structure in soils affected by opencast mining for coal. J. Sci. Food Agric. 55,327-339. Shen, S. M., Pruden, G., and Jenkinson, D. S. (1984). Mineralization and immobilizationof nitrogen in fumigated soil and the measurement of microbial biomass nitrogen. Soil Biol. Biochem. 16, 431-444. Shepherd, M. A., and May, M. J. (1992). Practical aspects of cover crop establishment,choice of species and cultivations. Aspects Appl. Biol. 30,331-334. Shepherd, M. A., Stockdale, E. A,, Powlson, D. S., and Jarvis, S. C. (1996). The influence of organic nitrogen mineralization on the management of agricultural systems in the UK. Soil Use Management 12,76-85. Sims, J. L., and Frederick, L. R. (1970). Nitrogen immobilization and decomposition of corn residue in soil and sand as affected by residue particle size. Soil Sci. 109,355-361. Janik, L. J., Head, M. J., and McClure, S. G. (1993). High energy ultraviolet photoSkjemstad, J. 0.. oxidation, A novel technique for studying physically protected organic matter in clay- and siltsized aggregates. J. Soil Sci. 44,485-499. Smith, K. A., and Howard, R. S. (1980). Field studies of nitrogen uptake using 15Ntracer methods. J. Sci. FoodAgric. 31,839-840. Smith, M. S., Frye, W. W., and Varco, J. J. (1987). Legume winter cover crops. Adv. Soil Sci. 7,95-139. Smith, S. J., and Sharpley, A. N. (1990). Soil nitrogen mineralizationin the presence of surface and incorporated crop residues. Agron. J. 82, 112- 116. Soon, Y. K., and Clayton, G. (1996). Influence of antecedent legume crops and tillage on the distribution and dynamics of nitrogen in a sandy loam soil. Transactions of the 9th Nitrogen Workshop, Braunsweig, Germany, September, pp. 235-238. Staley, T. E., Edwards, W. M., Scott, C. L., and Owens, L. B. (1988). Soil microbial biomass and organic component in a no-tillage chronosequence.Soil Sci. SOC.Am. J. 52,998- 1005. Staley, T. E., Caskey, W. H., and Boyer, D. G. (1990). Soil denitrification and nitrification potentials during the growing season relative to tillage. Soil Sci. SOC.Am. J. 54, 1602-1608. Stevenson (1 982). “Humus Chemistry: Genesis, Composition,Reactions.” Wiley, New York. Stokes, D. T., Scott, R. K., Tilston, C. H., Cowie, G., and Sylvester-Bradley, R. (1992). The effect of soil disturbance on nitrate mineralization.Aspects Appl. Biol. 30,279-282. Swift, M. J., Heal, 0. W., and Anderson, J. M. (1979). “Decomposition in Terrestrial Ecosystems.” Blackwell, Oxford. Sylvester-Bradley, R., Addiscott, T. M., Vaidyanathan, L. V., Murray, A., and Whitmore, A. (1987). Nitrogen advice for cereals: Present realities and future possibilities. Proceedings 263 of the Fertiliser Society of London, December 10, 1987, pp. 4-31. Thiagalingam,K., Gould, M., and Watson, P. (1991). Effect of tillage on rainfed maize and soybean yield and the nitrogen fertilizer requirements for maize. Soil Tillage Res. 19,47-54. Thompson, J. P. (1992). Soil biotic and biochemical factors in a long-term tillage and stubble management experimenton a Vertisol. I. Seedling inhibition by stubble. Soil Tillage Res. 22,323-337. Tisdall, J. M., and Oades, J. M. (1982). Organic matter and water-stable aggregates in soils. J. Soil Sci. 33,141-163. Van Gestel, M., Ladd, J. N., and Amato, M. (1991). Carbon and nitrogen mineralizationfrom two soils of contrasting texture and microaggregatestability,influence of sequential fumigation,drying and storage. Soil Biol. Biochem. 23,313-322. Van Veen, J. A., and Kuikman, P. J. (1990). Soil structural aspects of decompositionof organic matter by microorganisms. Biogeochernistry 11,213-223.

CULTIVATION EFFECTS ON SOIL N MINERALIZATION

3 11

Van Veen, J. A,, Ladd, J.N., and Amato, M. (1985). Turnover of carbon and nitrogen through the microbial biomass in a sandy loam and a clay soil incubated with I4C(U) glucose and 15(NH,),S0, under different moisture regimes. Soil Biol. Biochern. 17,747-756. Verberne, E. L. J., Hassink, J., de Willigen, P., Groot, J. J. R., and Van Veen, J. A. (1990). Modelling organic matter dynamics in different soils. Netherlands J. Soil Sci. 38,221-238. Vinten, A. J. A,, Vivian, B. J., Wright, F., and Howard R. S. (1994). A comparative study of nitrate leaching from soils of differing textures under similar climatic and cropping conditions. J. Hydrol. 159, 197-213. Watkins, N., and Barraclough, D. (1996). Gross rates of N mineralization associated with the decomposition of plant residues. Soil Biol. Biochem. 28, 169-176. Watson, C. A,, Fowler, S. M., and Wilman, D. (1989). Nitrate leaching following the ploughing of grasslclover leys in organic farming systems. First Research Meeting at the Welsh Agricultural College, September 13-15, 1988, Session IV,Paper No. 3. British Grassland Society, Hurley, UK. Webb, J., Froment, M., and Sylvester-Bradley, R. (1991). Effects of cultivation and nitrification inhibitor on soil nitrogen availability after a grass ley and on the response of the following cereal crop to fertilizer nitrogen. J. Agric. Sci. (Cambridge) 117,9-21. Wild, A. (ed.) (1988). “Russell’s Soil Conditions & Plant Growth,” 1 lth Ed. Longman ScientificIJohn Wiley, Harlow, UK. Wilhelm, W. W., Doran, J. W., and Power, J. F. (1986). Corn and soybean yield response to crop residue management under no-tillage production systems. Agron. J. 78, 184- 189. Williams, J., Harrison, R., Silgram, M., and Chambers, B. J. (1996). Soil mineral nitrogen supply in arable cropping rotations. Aspecrs Appl. Biol. 47,59-66. Youngs, E. G., and Leeds-Harrison, P. B. (1990). Aspects of transport processes in aggregated soils. J. Soil Sci. 41.665-675.

This Page Intentionally Left Blank

Index A ACES buffer chemical formula and pKa value, 175(table) chemical name, 153(table) Acetaldehyde, RF2-mediated fertility restoration and, 114, 115 ADA buffer chemical formula and pKu value, 175(table) chemical name, I53(table) in solution culture, 176, 177 Adsorption, in rhizosphere, process of, 30,3 1-33 Aeroponic culture, 151, see also Solution culture Albedo, of direct drilled soils, 279 Aldehyde dehydrogenases, mitochondria1 fertility restoration and, 113-116 RF2 protein and, 1 12- 113 Alumina, 197, see also Sand-alumina culture Aluminum toxicity, in sand-alumina culture, 198 Amberlite ion-exchange resins, 190, 194 Ammonium concentration in nutrient solutions, 157(table) concentration in soil solutions, 154-1 55 in effluent-derived irrigation water, 38 mobility in fertigated soils, 52 in nitrogen cycle, 268 partitioning in soil, Gapon equation for, 32-33 in solution culture, effects on solution pH, 173- 174 Ammonium nitrate fertilizers, pH considerations and, 52 Ammonium sulfate fertilizers, preplanting application, used with fertigation, 50 Ampboteric resin, in pH buffering, I9 1 Anaerobic stress, in pollen development, cytoplasmic male sterility and, 114-1 15 Anion-exchange resin buffers, 191- 195 Anion speciation in capillary electrophoresis, 133, 135, 14I - 142 electrophoretic migration and, 139

ion chromatography and, 132 methods in, 132-133 Anthers, male sterility and, 81, 83-86 Appressoria, of southern corn leaf blight, 91 Atmospheric transmission ratio, radiation use efficiency and, 224 Atrazine, in effluent-derived irrigation water, 38-39

B Banding fertilization, preplanting, used with fertigation, 50.51 Barley, radiation use efficiency in, 238(table), 244-245,253,254 Bell peppers, see Peppers BES buffer chemical formula and pKu value, 175(table) chemical name, 153(table) in solution culture, 176 Bicarbonate, in pH buffering, 177-179, 195, 20 I Bio-rex 70 resin buffer, 190 Bipolaris rnaydis, 86, see also Southern corn leaf blight Boehmite, 197,202 Boron in solution culture ion-exchange resins and, 194, 195 nutrient solution concentrations, I57(table) periodic replacement approach, 16 1 toxic levels, 159(table) toxicity, irrigation water quality and, 36, 37, 38 Boundary conditions, in rhizosphere ion transport, 34-35 BPDS chemical name, 153(table) copper deficiency studies, 186 Broadcast fertilization compared to microfertigation, 3-5 preplanting, used with fertigation, 13.49-51 Broccoli, nutrient consumption rates, 44(tabIe), 46(table), 48(table)

313

3 14

INDEX

Buffering, see also Chelator-buffered nutrient solutions; Nutrient buffering; pH buffers nutrient solutions and, 157 of oxidation-reduction potential, 189 in soil solutions, 157 Buffers, see also Good’s buffers; Ion-exchange resin buffers; pH buffers in capillary electrophoresis, effects on ion separation, 144 Bulk flow, in capillary electrophoresis, 137

C C, crops, see also individual crops radiation use efficiency in, 237-240(table), 243-250 C, crops, see also individual crops radiation use efficiency in, 234, 235-237(table), 241 -243,248-250 Cadmium, chelator buffering, 188 Cadmium toxicity, use of Good’s buffers in, 176 Calcium complexation with Good’s buffers, 176 concentration in nutrient solutions, 157(table) concentration in soil solutions, 155 effects of sodium toxicity on, 37 ion-exchange resins and, 194 physiological minima defined for, 171(table) in salt stress dynamics, 14 Calcium : magnesium ratio, in nutrient and soil solutions, 156-157 Calcium carbonate, in pH buffering, 195-196 Calcium-phosphate minerals, see Calcium precipitates Calcium precipitates emitter clogging and, 18 in soil, 33 Canopy, see Leaf canopy Canopy shading, see Shading Capillary electrophoresis advantages of, 133 analyte mobility, 139- 140 anion speciation, 133, 141-142 applied to soil solutions, 146 capillary layering in, 135 cation speciation, 140-141 compared with high pressure liquid chromatography, 144-145 compared with ion chromatography,

143-144, 145 detection in, 143, 145 electroosmotic migration, 137- 138 electrophoretic migration, 139 general principles of, 134-135 limitations of, 145 longitudinal diffusion, 138 overview of, 132 sample introduction methods, 136 Capillary zone electrophoresis, 134, see also Capillary electrophoresis Carbon : nitrogen ratio, in crop residues, nitrogen mineralization and, 290,291-292 Carbon, organic, in no tillage systems, 298 Carbonate precipitates, microfertigation emitter clogging and, 18 Carbon dioxide in bicarbonate pH buffering, 177-179, 195 in soil effects of cultivation on, 269 measuring microbial respiration by, 271-272 Carbon dioxide assimilation rate, see also Leaf photosynthesis radiation use efficiency and, 221-222, 250-255 Carbon mineralization effects of cultivation on, 287 effects of soil disturbance on, 283 Carbon partitioning, in root growth, 25 Carboxylates, from exudates in recycled greenhouse solutions, 40 rhizosphere pH and, 29-30 Carrots, nutrient consumption rates, 44(table), 46(table), 48(table) Cation-exchange resin buffers, 191-195 Cation speciation in capillary electrophoresis, 135, 140-141 electrophoretic migration and, I39 C cytoplasm, 82,83 CDTA in chelator-buffered nutrient solutions, 183 chemical name, 153(table) iron buffering, 180,183 Celery, nutrient consumption rates, 44(table), 46(table), 48(table) Cereals, see also Barley; Rice; Wheat radiation use efficiency in, 248-249 Cereal straw residues, nitrogen mineralization and, 289

INDEX CEREZ-Maize model, fertigation management and, 62 Cetyltrimethyl ammonium bromide (CTAB), 140, 142 Charrau cytoplasm, 82, 83 Chelating resin buffers, 191-195 Chelator-bufferednutrient solutions, 202 cadmium studies, 188 chelator biodegradation and, 189 chelators commonly used in, 183-184 deficiency and stress studies, 185- 188 ferrozine in, 188 germplasm screening and, 188 HEDTA solutions and, 184-185, 187 iron buffering and, 182- 183, 188 overview of, 180, 182 solution pH and, 185 Chelators, see also Chelator-buffered nutrient solutions chelating resin buffers, 191-195 chemical names for, 153(table) in fertigation systems, 54-56 in protection from micronutrient toxicity, 158 in solution culture, 179-189,202 Chickpea, radiation use efficiency, 253 Chili peppers, root characteristicsunder microfertigation, 27 Chinese cabbage, nutrient consumption rates, 44(table), 46(table), 48(table) Chisel cultivators, effects on soil, 275 Chloride distribution in soils under microfertigation, 11 identifying with capillary electrophoresis, 142 Chloride toxicity fertigation solutions and, 5 1 from irrigation water, 37 Chlorine, concentration in nutrient solutions, 157(table) Chloroform fumigation-extractiontechnique, 270 Chlorophyll, absorbance of, 228 Chlorosis, caused by bicarbonate buffering, 177, 179 ChPKSl gene, 92-93.96 Citrate in recycled greenhouse solutions, 40 in root exudation, 29 Clay soils nitrogen mineralization in, 281,283,284,298

315

protection of soil organic matter by, 280, 282-283 cms-T, see Maize, T-cytoplasm lines Cochliobolus spp, host-specific toxins of, 93-94, 117 Cochliobolus carbonum, 93 Cochliobolus heierosirophus, 86 1970 epidemic, 87-90.97 race 0 CPSl gene, 91 virulence of, 87,90-91 race T 1970 epidemic, 88-90 current levels of, 90 evolution in, 93 low fitness of, 90 origins of, 89-90 T-cytoplasm maize and, 80 T-toxin genes and, 117 T-toxin production in, 91-93 virulence of, 90-95 Cochliobolus victoriae, 94 Colonial protozoa, microfertigationemitter clogging and, 19 Conidia, of southern corn leaf blight, 91 Controlled released fertilizers, in fertigation, 53 Control plots, in methods of measuring nitrogen mineralization, 273-274,275 Conversion efficiency, 219, see also Radiation use efficiency Copper complexation with Good’s buffers, 176 concentration in nutrient solutions, 156, 157(table) concentration in soil solutions, 156 HEDTA buffering, 184 iron chelators and, 180, 18l(table) toxic levels in nutrient solutions, 158, 159(table) Copper deficiency, chelator-bufferingmethods and, 185-186 Corn, see Maize; Sweet corn Cotton, under fertigation nutrient consumption rates, 42(table), 45(table), 47(table) responses to, 16(table), 17 Cowpea, radiation use efficiency in, 240(table), 247 CPSl gene, 91,117

3 16

INDEX

Crop(s) nutrient consumption rates, 40,42-48(tables) quality, factors affecting, 20-21 radiation use efficiency in C, crops, 237-240(table), 243-250 C, crops, 234,235-237(table), 241-243, 248-250 variability of, 248-258 responses to microfertigation, 8(table) Crop canopy, see Leaf canopy Crop growth, see ulso Plant growth allocation of nitrogen during, 232-233 establishment phase, radiation use efficiency and, 255-256 light-based analyses, 217-219 light levels and, 215-216 seed growth phase, radiation use efficiency and, 256 time-based analyses, 216-217 Crop management cultivation and nitrogen mineralization in, 300-303 use of osmotic stress in, 15 Crop mass estimations of, 226-227.231-232 in time-based growth analyses, 216 Crop models based on radiation use efficiency, 259 fertigation management and, 62 Crop residue management, nitrogen mineralization and, 289-292 Crop yield effects of cultivation on, 286-289 factors affecting, 20-21 CTAB, see Cetyltrimethyl ammonium bromide Cultivation, see ulso No tillage; Reduced tillage effects on nitrogen mineralization crop management and, 300-303 crop residues and, 289-292 cultivation depth and, 293 mineralization flush, 280,283,294, 299-301 nitrate leaching and, 294-297, 301 overview of, 297-300 soil fauna and, 284-286 soil mineral nitrogen and, 294-297, 300, 30 1 soil organic matter decomposition and, 294 soil texture and, 280-284

timing and frequency of cultivation, 292-294 yield and, 286-289 effects on soil physical properties, 268-269, 275-280 techniques in, 275-276 Cumulative consumption function, defined, 20-21 Cyclic peptide synthetase, fungal virulence and, 117 Cytoplasmic male sterility, see ulso Maize, T-cytoplasm lines characteristics of, 81 fertility restoration and, 103 in maize, 81 mitochondria1 open reading frames and, 97-98

D DCCD, see N,N'-Dicychlohexylcarbodiimide DCTA, chemical name, 153(table) DECl gene, 92-93, 117 Denitrification effects on fertigation, 59 in no tillage systems, 296 Developing countries, use of fertigation in, 2 N,N'-Dicychlohexylcarbodiimide,101 Diffuse radiation photosynthetically active, 228 radiation use efficiency and, 224-225 Diffusion, longitudinal, in capillary electrophoresis, 138 Diffusion coefficients, rhizosphere ion transport and, 34,35 Direct drilling, see also No tillage crop yield and, 287,289 effects of cultivation on, 297 effects on soil fauna, 285, 286 effects on soil physical properties, 276, 279 effects on soil water, 279 nitrate leaching and, 294-295, 301 nitrogen immobilization and, 293 nitrogen mineralization and, 29 I , 292 Disk plows, effects on soil, 275 Drinking water, nitrate leaching and, 268 Drip fertigation, see ulso Microfertigation advantages of, 4-5 compared to microjet fertigation, 15 research areas in, 64

3 17

INDEX Drip irrigation, see also Microirrigation saline irrigation water and, 37 Drought, radiation use efficiency and, 252-253 Dry matter production effects on crop quality and yield, 20-21 light-based analyses of, 217-218 DTPA in chelator-buffered nutrient solutions, 183 in fertigation solutions, 54-55 iron buffering, 183 in solution culture, 180

E Earthworms effects of cultivation on, 285,286 effects on nitrogen mineralization, 285 -286 EDDA, chemical name, 153(table) EDDHA cadmium buffering, 188 chemical name, 153(table) in fertigation solutions, 55 EDTA in chelator-buffered nutrient solutions, 182, 183,202 chemical name, 153(table) in fertigation solutions, 54, 55 iron buffering, 183 in protection from micronutrient toxicity, 158 in solution culture, 179 Effluents, municipal, in fertigation, 18, 37-39 Eggplant, nutrient consumption rates, 42(table), 45(table), 47(table) EGTA, cadmium buffering, 188 Electrolyte buffers, in capillary electrophoresis, effects on ion separation, 144 Electromagnetic injection, in capillary electrophoresis, 136 Electroosmosis in capillary electrophoresis, 134, 135 factors influencing, 139- 140 Emitters, in microfertigation, clogging of, 18-19 Escherichiu coli, T-urf-13 expression, 100- 101 Ethanol metabolism, RF2-mediated fertility restoration and, 114-1 15 Evapotranspiration, see ulso Transpiration effects on daily fertigation rates, 24-25 Exine, 85

F Faba bean, radiation use efficiency in, 239(table), 247 Ferrozine in chelator-buffered nutrient solutions, 188 chemical name, 153(table) Fertigation, see also Microfertigation choice of fertilizers for, 5 1-56 crop monitoring, 62,64 crop responses to, 5-8 cumulative consumption function, 20-2 I in developing countries, 2 future trends, 64-65 goals of, 19-20 in greenhouses, 59 irrigation rates, defining, 24-25 modeling of, 60-62,64-65 nutrient concentrations in, 22-24,41,49 nutrient consumption rates for crops, 40, 42-48( tables) nutrient uptake dynamics, 2 1-25 plant disease and, 65 preplanting fertilization and, 13,49-5 I rhizospheric processes and, 28-35 root growth and distribution, 25-28 safety issues, 63-64 soil distribution of water and nutrients, 8-13 soil monitoring, 62-63.64 soil organic matter and, 57-59 soil root volume effect.., 13-14 subsurface, advantages of, 15- 18 temperature effects in, 56 water quality considerations, 36-40 world-wide use of, 2-3 Fertigation fertilizers for saline conditions, 5 1-52 soil nutrient mobility and, 52-53 solubility of, 54(table) solution pH and, 52 using ready-mixes for, 53-56 Fertility restoration, see also Nuclear restorer genes sporophytic, 103 in T-cytoplasm maize, 103-116 Fertilization, see ulso Fertigation compared to microfertigation, 3-5 coupled with irrigation, 2 preplanting, in fertigation, 13.49-5 1

318

INDEX

Fertilizers, see also Nitrogen fertilizers in fertigation, 51-56 in reduced tillage systems, 302 solubility of, 54(table) Ferulic acid, in recycled greenhouse solutions, 40 Field crops, see Crop(s) Field variability, crop mass estimations and, 226-227 Filter systems, in microfertigation, 18- 19 Fixation, in rhizosphere, process of, 30,34 Flowers, field grown, responses to microfertigation, 7(table) Flowing solution culture applications of, 169-170 limitations of, 170, 172,201 sand-alumina culture and, 198-199 soil solution dynamics and, 172-173 techniques in, 168-169 Flux, see Nutrient uptake Forest soils, effects of cultivation on, 282 Fruit quality, osmotic stress and, 15 Fungal toxins, see also PM-toxin; T-toxin susceptibility of T-cytoplasm maize to, 80, 86 Fungi no tillage systems and, 285 soil biomass, monitoring, 271 T-cytoplasm maize and, 80, see also Cochliobolus heterostrophus;Mycosphaerella zeae-maydis virulence of, cyclic peptide synthetase and, 117

G Gapon equation, for estimating cation partitioning in soil, 32-33 Generative cell, 85 GEOCHEM-PC, 180, 182,202 Germplasm screening, for iron stress, 188 Germ tubes, of southern corn leaf blight, 91 GLYCIN model, fertigation management and, 62 Goethite, in phosphorus buffering, 196-197 Good’s buffers, see also individual buffers; pH buffers chemical formulas and pKa values, 175(table) microbial degradation and, 176 in solution culture, 174-177

GOSSYM model, fertigation management and, 62 Grain legumes, see Legumes Grapefruit, responses to microfertigation, 5, 6(table) Greenhouse crops, responses to microfertigation, 8(table) Greenhouses, principles of fertigation and, 59 Greenhouse solutions, recycled, in fertigation, 39-40,59 Growth, see also Crop growth; Plant growth; Relative growth rate; Root growth; Seed growth light-based analyses, 2 17-2 19 time-based analyses, 216-217 Gypsum, precipitation in soil, 33

H Hawaiian crops, see also Sugarcane radiation use efficiency in, 243 HBED chemical name, 153(table) iron buffering, 182-183 HC-toxin, 93 HEDTA in chelator-buffered nutrient solutions, 182, 184-185.202 chemical name, 153(table) in deficiency stress studies, 187, 188 iron buffering, 180, 183 HEIDA, chemical name, 153(table) Helmhotz layer, in capillary electrophoresis, 135, 137 Helminthosporiummaydis, 86, see also Southern corn leaf blight HEPES buffer chemical formula and pKa value, 175(table) chemical name, 153(table) Herbicides in effluent-derived irrigation water, 38-39 to remedy microfertigation emitter clogging, 19 Heterosis, male sterility systems and, 8 1 High pressure liquid chromatography (HPLC), compared with capillary electrophoresis, 144-145 Hoagland solution, composition of, 156, 157(table)

INDEX Hohenheim solution, composition of, 156, 157(table) HPLC, see High pressure liquid chromatograPhY Humic substances, in effluent-derived irrigation water, 39 HV-toxin, 93-94 Hybrid vigor, male sterility systems and, 81 Hydrostatic injection, in capillary electrophoresis, 136 Hydroxyapatite, 202 in phosphorus buffering, 199-200 8-Hydroxyquinoline-5-sulfonicacid, in capillary electrophoresis, 141 Hyphae, of southern corn leaf blight, 91 Hysteresis, in phosphorus rhizosphere sorptiondesorption, 3 1

I IDA, chemical name, 153(table) Incident radiation in analyses of crop growth, 217-218 estimations of, 227-228 Insecticides, T-cytoplasm maize and, 96 Intercepted radiation in calculations of radiation use efficiency, 231-232 efficiency, defined, 218 estimations of, 229-230 Intine, 85 Ion chromatography compared with capillary electrophoresis, 143-144.145 overview of, 132 Ion-exchange resin buffers, 202 for nutrients, 191-195 for pH, 190-191. 194-195 Ion transport, see also Nutrients; Nutrient uptake in rhizosphere, description of, 34-35 Iron in chelator-buffered nutrient solutions, 182-183.188 concentration in nutrient solutions, 156, 157(table) concentration in soil solutions, 156 ferrozine and, 188 precipitation in solution culture, 173, 180, 18l(table), 182, 185

3 19

Iron chelators EDTA, 179 ferrozine and, 188 in fertigation solutions, 55 other micronutrients, 180 in solution culture, 182-183, 188 Iron chlorosis caused by bicarbonate buffering, 177, 179 chelator-buffering methods and, 187, 188 Iron hydroxides, see Iron, precipitation Irrigation, see also Fertigation; Microfertigation; Microirrigation in crop management, 2 salinity and, 14 water quality considerations, 36-40 Isotope dilution technique, for measuring soil nitrogen mineralization, 272,274

J Johnson solution, composition of, 156, 157(table)

K Kenaf, radiation use efficiency in, 237(table)

L Langmuir competitive adsorption model, 32 Lannate, T-cytoplasm maize and, 96 Latitude effect on incident radiation, 227 radiation use efficiency and, 225 Leaching, see Nitrate leaching; Soil leaching Leaf angle, radiation use efficiency and, 221 Leaf area, in time-based growth analyses, 216-217 Leaf area index, radiation use efficiency and, 221,225,233 Leaf bum, irrigation water quality and, 37 Leaf canopy, see also Shading estimation of intercepted radiation and, 229-230 in light-based growth analyses, 217 nitrogen distribution in, radiation use efficiency and, 224 Leaf nitrogen carbon dioxide assimilation rate and. 25 1

320

INDEX

Leaf nitrogen (continued) radiation use efficiency and, 223-224, 25 1-252 during seed growth phase, 256 Leaf pathogens, radiation use efficiency and, 255 Leaf photosynthesis, see also Carbon dioxide assimilation rate radiation use efficiency and, 21 8,220, 22 1-224,225,250 seasonal variation, 255-256 temperature and, 254 vapor pressure deficit and, 254 Leaf quantum efficiency, radiation use efficiency and, 2 17,220.22 I Legumes, radiation use efficiency in, 239-240(table), 246-248 compared to other crops, 249 during seed growth phase, 256 soil water deficits and, 252-253 Lettuce, nutrient consumption rates, 44(table), 46(table), 48(tahle) Ligand buffering, in solution culture, 179- I89 Light, see also Radiation crop growth and, 215-216 Lipid phosphorus, in monitoring soil microbial biomass, 27 1 Loams, nitrogen mineralization in, 283, 298 Lodged crops, see also Sugarcane radiation use efficiency in, 243 Long Ashton solution, 156 Longitudinal diffusion, in capillary electrophoresis, 138 Lupin, radiation use efficiency in, 250

M Macronutrients, ion-exchange resin buffers and, 191-195.202 Magnesium complexation with Good’s buffers, 176 concentration in nutrient solutions, 157(table) ion-exchange resins and, 194 pH resin buffers and, 190 Magnesium bicarbonate, in pH buffering, I79 Maize, see also Sweet corn commercial seed production, 82.97 male sterility systems, 82-83 photosynthate conversion in, 242 radiation use efficiency in, 234, 235-236(table), 241

calculations of, 231 -233 compared to other crops, 248,249-250 leaf nitrogen and, 252 during seed growth phase, 256 soil water deficits and, 253 root uptake, effects of temperature on, 56 yield, effects of cultivation on, 287-288 Maize, T-cytoplasm lines abortive pollen development in, 83-86 fungal susceptibility, 80,86 linked with male sterility, 80,97-100 mitochondrial dysfunction and, 96-97, 100-102 restorer genes and, 97 southern corn leaf blight epidemic, 87-90, 97 yellow leaf blight, 95-96 future studies in, 119-121 methomyl sensitivity, 96,97, 100. 102 mitochondrial genes and, 80 nuclear restorer genes, 82.83, 103-1 16 reduced vigor of, 116 researchers on, 116-1 19 URF13 protein and male sterility, 102-103 Malate, in root exudation, 29, 30 Male sterility, see also Maize, T-cytoplasm lines cytoplasmic, overview of, 81 maize systems, 82-83 URF13 protein and, 102-103 Malonate, in root exudation, 29 Manganese chelators, 56 concentration in soil solutions, 156 precipitation in soil, 33 in solution culture deficiency studies, chelator-buffering methods and, 185-186 HEDTA buffering, 184-185 iron chelators and, 18l(table) nutrient solution concentrations, 156, I57(table) pH resin buffers and, 190 toxic levels, 158, 159(table) Marginal water, see also Wastewater in fertigation, quality issues, 36-37 MES buffer chemical formula and pKa value, 175(table) chemical name, 153(table) in solution culture, 167, 175-176,201 Methomyl, T-cytoplasm maize sensitivity, 96, 97,100.102

INDEX Michaelis-Menten constants, in fertigation nutrient uptake dynamics, 21.24 Microelements, see Micronutrient metals Microfertigation, see also Fertigation advantages of, 3-5 clogged emitters, avoiding, 18-19 cost-benefit analysis of, 5 crop responses to, 5-8 drip and microjet techniques compared, 15 in greenhouses, 59 nutrient consumption rates for crops, 40, 42-48(tables) salinity and, 14-15 soil distribution of water and nutrients in, 8-13 soil root volume effects, 13-14 subsurface, advantages of, 15-18 temperature effects in, 56 use of municipal effluents in, 37-39 water quality considerations, 36-40 world-wide use of, 2-3 Microflora, microfertigation emitter clogging and, 18, 19 Microgametogenesis, abortive, in T-cytoplasm maize, 83-86 Microirrigation. see also Drip irrigation; Fertigation; Microfertigation agricultural areas under, 2 compared to microfertigation, 3-5 crops used on, 2-3 soil water distribution and, 9 Microjet fertigation, see also Microfertigation compared to drip fertigation, 15 Micronutrient deficiencies, chelator-buffering methods and, 185-186 Micronutrient metals concentration in soil solutions, 156 in fertigation solutions, 54-56 in solution culture calcium carbonate pH buffering and, 196 chelator buffering, 179-189 ion-exchange resin buffers and, 191-195, 202 iron chelators and, 180 large solution volumes and, 167 nutrient solution concentrations, 156, 157(table) toxicity levels, 158, 159(table) Microorganisms, see also Soil microorganisms microfertigation emitter clogging and, 18, 19

32 1

Microsporogenesis, abortive, in T-cytoplasm maize, 83-86 Mineralization, see Carbon mineralization; Nitrogen mineralization Mist culture, 151, see also Solution culture Mitochondria in normal tapetal development, 85 in T-cytoplasm maize linkage of fungal susceptibility with male sterility, 80, 97-100 susceptibility to fungal toxins, 80, 86, 96-97, 100-102 tapetal degeneration and, 86 URF13 protein and male sterility, 102-103 Mitochondria1 genes in male sterility systems, 8 1 in T-cytoplasm maize, 80, see also T-urfl3 gene Moldboard plows, effects on soil, 275 Molybdenum concentration in nutrient solutions, 157(table) in solution culture, periodic replacement approach, 161 Montmorillonitic soils, effects of municipal effluents on, 39 MOPS buffer chemical formula and pKa value, 175(table) chemical name, 153(table) in solution culture, 176 Municipal effluents, in fertigation, 18, 37-39 Muskmelons nutrient consumption rates, 44(table), 46(table), 48(table) root characteristics under microfertigation, 26, 28(tabIe) Mycosphaerella zeae-maydis,see also Yellow leaf blight PM-toxin, 86.95 -96 f3-polyketol toxins, 80 T-cytoplasm maize and, 80, 86, 95-96 toxin genes, 117 MzPKSI gene, 96 MzREDI gene, 96 M7RED2 gene, 96

N N cytoplasm fungal toxins and, 97 methomyl and, 97 yellow leaf blight and, 95

322

INDEX

Net assimilation rate, limitations of, 216-217 Nickel, concentration in soil solutions, 156 Nitrate concentration in soil solutions, 154-155 identifying with capillary electrophoresis, 142 in nitrogen cycle, 268 in no tillage soils, 286 physiological minima defined for, 171(table) rhizosphere pH and, 29 in solution culture effects on solution pH, 173-174 ion-exchange resins and, 194 nutrient solution concentrations, 157(table) relative addition rate techniques and, 165-166 uptake dynamics, 21-22 Nitrate leaching crop residues and, 290-291 effects of cultivation on, 293,294-297.301 under microfertigation, 13-14 problems of, 268 in reduced tillage, 299 Nitrification fertigation and, 53,57-59 in nitrogen cycle, 268 Nitrogen, see also Leaf nitrogen allocation during crop growth, 232-233 in effluent-derivedirrigation water, 38 in fertigation, 16, 17 crop consumption rates, 40,42-44(tables) crop responses to, studies on, 5, 6-8(table) isotopic labeling of, 272, 274 organic, in nitrogen cycle, 268 physiological minima defined for, 17I(tab1e) radiation use efficiency and, 232-233 soil levels, root growth and, 25 in solution culture effects on solution pH, 173-174 relative addition rate techniques and, 164, 165- 166 Nitrogen cycle, overview of, 267-268 Nitrogen deficiency, symptoms, created in solution culture, 165 Nitrogen fertilizers crop yield and, effects of cultivation on, 286-289 mineralization dynamics and, 268 mobility in fertigated soils, 52-53 preplanting application, used with fertigation, 50-5 1

Nitrogen fixation, flowing solution culture studies and, 171 Nitrogen immobilization direct drilling and, 293 in nitrogen cycle, 268 reduced tillage and, 298 soil microorganisms and, 290 Nitrogen mineralization effects of cultivation on, 269 crop management and, 300-303 crop residues and, 289-292 cultivation depth and, 293 cultivation timing and frequency, 292-294 mineralization flush, 280,283,294, 299-301 nitrate leaching and, 294-297 overview of, 297-300 soil fauna and, 284-286 soil mineral nitrogen and, 294-297 soil texture and, 280-284 yield and, 286-289 effects on fertigation, 57-59 methods of measuring, 270-275 in nitrogen cycle, 268 in no tillage systems, 276, 289,291, 292, 293-294,298 Normalized difference vegetation index, 230 No tillage, see also Direct drilling; Reduced tillage crop yield and, 286-289 cultivation of no till soils, 292, 297 fertilizer needs in, 300 initial nitrogen needs of, 292 nitrate leaching and, 296 nitrogen mineralization and, 276,289,29 1, 292,293-294,298 soil characteristics,296,298 soil fauna and, 284-285,286 soil moisture and, 288, 289 soil organic matter and, 293 soil physical properties and, 276,277 soil temperature and, 288 soil types suitable for, 288 subsurface fertigation and, 17 surface zone, 289,293,294,298 MA cadmium buffering, 188 in chelator-buffered nutrient solutions, 183 chemical name, 153(table) iron buffering, 183

INDEX Nuclear restorer genes in male sterility systems, 81 in T-cytoplasm maize, 82-83 cloning of, 109-113 conserved motif associated with, 107- 109 determining “normal” function of, 117118 fertility restoration and, 103- 109 fungal susceptibility and, 97 methomyl susceptibility and, 97 Nutrient buffering, see also Solution culture with chelators, 179-189, 202 cadmium studies, 188 chelator biodegradation and, 189 chelators commonly used in, 183-1 84 deficiency and stress studies, 185- 188 ferrozine in, 188 germplasm screening and, 188 HEDTA solutions and, 184-185, 187, 202 iron buffering and, 182- 183, 188 overview of, 180, 182 solution pH and, 185 with inorganic solid phases, 196-200 with ion-exchange resins, 191-195,202 Nutrient consumption rates, for crops under fertigation, 40,42-48(tables) Nutrients, see also Soil nutrients in fertigation water defining threshold values for, 22-24 determining concentrations for, 41,49 wastewater-derived, problems of, 38 in plants, effects on crop quality and yield, 20-21 Nutrient solutions, see also Solution culture chelator buffering, 179-189,202 composition of, 156-157 elemental toxicity and, 158 ion-exchange resin buffering, 191-195.202 physiological minima defined for, 169-170, 171(table) Nutrient uptake, see also Ion transport dynamics of, fertigation management and, 21 -25 effects of temperature on, 56 flowing solution culture studies and, 171 modeling of, fertigation management and, 60-62 Nutritional stress studies, chelator-buffering methods in, 186-188

323 0

Occlusion, see Fixation Open reading frames, mitochondrial, in cytoplasmic male sterility, 97-98 Oranges, responses to microfertigation, 5, 6(table) orf221,99, 104, 106 Organic matter, in effluent-derived irrigation water, 38-39 Orthophosphate, 197, see also Sand-alumina culture Osmotic stress, in crop management, 15 Oxalate in recycled greenhouse solutions, 40 in root exudation, 29 Oxidation-reduction potential buffering, in solution culture, 189 Oxygen effluent-derived irrigation water and, 36,39 in oxidation-reduction potential buffering, 189 soil levels, root growth and, 25 Ozone, radiation use efficiency and, 255

P PAR, see Photosynthetically active radiation Particulate organic matter, see also Soil organic matter effects of cultivation on, 282 Pathogens leaf, radiation use efficiency and, 255 in recycled greenhouse solutions, 39 Pea, radiation use efficiency in, 240(table), 247 Peanut, radiation use efficiency in, 239(table), 246-247,249,25 1 Pearl millet, radiation use efficiency in, 253 Penetration pegs, of southern corn leaf blight, 91 Peppers, under fertigation nutrient consumption rates, 42(table), 45(table), 47(table) root characteristics, 26, 27 pH, see also Soil pH effects on chelator-buffered nutrient solutions, 185 effects on electroosmotic flow, 139 in preparation of fertigation solutions, 52

324

INDEX

pH (continued) of rhizosphere, root exudates and, 29-30 in sand-alumina culture, 198 pH buffers, see also individual buffer.s chemical names for, 153(table) in solution culture, 173-179,201-202 calcium carbonate, 195- I96 ion-exchange resins, 190-191, 194-195, 202 large solution volumes and, 167 Phenolic acids, in recycled greenhouse solutions, 40 Philippines, Cochliobolus heterosrrophus race T, 88.89 Phosphate precipitates, microfertigation emitter clogging and, 18 in solution culture, pH buffering and, 174, I96 Phosphorus in fertigation, 16- I7 crop consumption rates, 40.45 -46(tables) crop responses to, studies on, 5-8 in effluent-derived water, 38 uptake dynamics, 22(table) physiological minima defined for, 170, 17l(table) in soil concentrations, 155, 156 distribution, 11 fixation, 34 precipitation, 33 root exudates and, 29-30 root growth and, 25 salt stress and, 14 in solution culture inorganic solid phase buffering, 196-200 ion-exchange resin buffers, 192 large solution volumes and, 167 nutrient solution levels, 156, 157(table) periodic replacement approach, 161, 162 pH buffering and, 174 physiological minima defined for, 170, 17l(table) toxic levels, 159(table) Phosphorus fertilizers, in fertigation preplanting application, 5 1 soil mobility of, 53 Photosynthate conversion in maize, 242

in potato, 243 in sugarcane, 242 in wheat, 243 Photosynthesis, see also Leaf photosynthesis radiation use efficiency and, 2 18 Photosynthetically active radiation, conversions for, 228, 234 Phyllosticta maydis, 86, see also Yellow leaf blight Physiological stress studies, nutritional, chelator-buffering methods in, 186- 188 Phytosiderophores, 55, 183 Pigeon pea, radiation use efficiency in, 240(table), 247 PKSl gene, 117 Plant growth, see also Crop growth; Relative growth rate; Root growth; Seed growth effects of salinity on, 36-37 light-based analyses, 217-219 soil solution dynamics and, 172-173 time-based analyses, 216-217 Plant mass energy content of, crop mass estimations and, 227 in time-based growth analyses, 216 Plant residues, no tillage and, 276 Plow pan, see also Soil compaction cultivation techniques and, 275 PM-toxin, 86 effects on T-cytoplasm maize mitochondria, 96-97. 101-102 genetic control of, 96 T-cytoplasm maize and, 95 Pollen development, cytoplasmic male sterility and, 83-86, 114-115 P-Polyketol toxins, see also PM-toxin; T-toxin T-cytoplasm maize and, 80 Polyphosphate, mobility in fertigated soils, 53 Potassium concentration in nutrient solutions, 157(table) in fertigation crop consumption rates, 40,47-48(tables) crop responses to, studies on, 5-8 soil transport, 11-13 uptake dynamics, 22(table) physiological minima defined for, 17 1(table) in salt stress dynamics, 14 in soil fixation, 34 partitioning, Gapon equation for, 32-33

INDEX Potassium bicarbonate, in pH buffering, 196 Potassium tarakanite, 198, 202 Potato nutrient consumption rates, 42(table) radiation use efficiency in, 237(table), 243-244,248 responses to subsurface fertigation, 16(table), 17 yield, effects of cultivation on, 287 Precipitation, in rhizosphere, process of, 30, 33-34 Pressure injection, in capillary electrophoresis, 136 Programmed nutrient addition technique, 162-164 Pseudoboehmite, 197, 198 Putrescine, in capillary electrophoresis, 140 Pyruvate dehydrogenase, I15

R Radiation, see ulso Diffuse radiation; Incident radiation; Intercepted radiation: Photosynthetically active radiation radiation use efficiency and factors influencing, 224-225,256-258 measurements of, 227-230 reflected, used in estimating intercepted radia tion, 230 Radiation conversion efficiency, 2 19-220 Radiation use efficiency among species, variation in, 219,248-250 calculation of, 23 1-233 carbon dioxide assimilation rate and, 250-255 during crop establishment, 255-256 crop models based on, 259 defined, 226 drought and, 252-253 estimations of crop mass and, 226-227 estimations of solar energy and, 227-230 experimental measures, 233-248 for C, crops, 237-240(table), 243-248 for C, crops, 234,235-237(table), 24 1-243 introduction of, 218-219 leaf nitrogen and, 25 1-252 leaf photosynthesis and, 218,220,221-224, 225 overview of, 258-259

325

radiation variables and, 224-225.256-258 seasonal variation and, 255 -256 during seed growth phase, 256 temperature and, 253,254 terminology for, 219-220 theoretical analyses of, 220-225 units for, 226, 233-234 vapor pressure deficit and, 253,254-255 variability in, sources of, 219,248-258 Reciprocal translocations, Cochliobolus heterosrrophus toxin genes and, 92 Reduced tillage, see also No tillage fertilizer needs in, 302 initial nitrogen needs of,292 nitrate leaching and, 297, 299 nitrogen immobilization and, 298 soil characteristics under, 296, 298 soil nitrate levels and, 296 soil physical properties and, 276,277 soil temperature and, 288 soil types suitable for, 288 surface zone nitrogen mineralization, 293 sustainability of, 302-303 Relative addition rate technique, 162, 164-167, 187 Relative growth rate general equation for, 2 16 solution culture renewal techniques and, 165, I 66 Resin buffers, ion-exchange, see Ion-exchange resin buffers Restorer genes, see Nuclear restorer genes Rfl gene, 99 conserved motif associated with, 107-109 determining “normal” function of, 117- 118 fertility restoration in T-cytoplasm maize, 103- I09 $1 gene, 83, 109-1 10 Rf2 gene, 103, 109 $2 gene, 83, 109-110, 112 RF2 protein fertility restoration and, 103, 113-1 16 mammalian mitochondria1 aldehyde dehydrogenases and, 112-113 r j 3 gene, 83 rj4 gene. 83 Rf8 gene, 103, 106, 107, 109 determining “normal” function of, 117-1 18 Rf* gene, 103, 106, 107, 109 determining ‘‘normal’’ function of, 11 7-1 18

326

INDEX

Rhizosphere effects on fertigation, 28-35 pH in, carboxylate release and, 29-30 sorption-desorptionin, 30-34 transport of ions in, 34-35 Rice, radiation use efficiency in, 238(table), 245,249-2.50.253.257 Rock phosphate, preplanting application, used with fertigation, 51 Rollers, in cultivation, 276 Root(s), see also Rhizosphere effects of salt stress on, 14 effects of solution culture on, 152 under fertigation, 17 characteristics of, 13-14,25-28 emitter clogging and, 18, 19 physical characteristics, nutrient uptake dynamics and, 21 Root exudation of carboxylates, 29-30 effects of solution culture on, 152 of protons, 29 in recycled greenhouse solutions, 39-40 soil microorganisms and, 28 Root growth under microfertigation, 26-28 soil factors and, 25-26 Root ion transfer constant, 21 Root mass crop mass estimations and, 226 nutrient uptake dynamics and, 21 Root morphology effects of solution culture on, 152 in sand culture, 198 Root pathogens, in recycled greenhouse solutions, 39 Root temperature, effects on nutrient uptake, 56 Rotary cultivators, effects on soil, 275-276 rrn26. 118

S Succharomyces cerevisiae, T-urf-I3 expression, 102 Salinity fertigation solutions for, 5 1-52 microfertigation and, 14- 15 of waste-derived irrigation water, 36-38,39 Salt stress, microfertigation and, 14-15

Sample stacking, in capillary electrophoresis, 136 Sampling variability, crop mass estimations and, 226-227 Sand-alumina culture, 197-199.202 S cytoplasm, 82, 83 Season, effect on incident radiation, 227 Seed growth, radiation use efficiency during, 256 Seeds, energy content, maize, 242 Shading net assimilation rate and, 217 radiation use efficiency and, 257 Sodicity, irrigation water quality and, 36, 38 Sodic soils, subsurface fertigation in, 17-18 Sodium, partitioning in soil, Gapon equation for, 32-33 Sodium adsorption ratio, of waste-derived irrigation water, 36, 38 Sodium bicarbonate, in pH buffering, 177-179, 196 Sodium toxicity in bicarbonate buffering, 179 from irrigation water, 37 Soil(s), see also Rhizosphere factors affecting root growth, 25-26 under fertigation distribution of water and nutrients in, 8-13 monitoring, 62-63 nitrogen supply, monitoring of, 273 type, cultivation methods and, 276 Soil aggregates carbon : nitrogen ratio in, 283 effects of cultivation on, 277,278, 279,280, 284,298 effects of wastewater on, 36 soil organic matter and, 281,282-283 Soil arthropods, effects of cultivation on, 285, 286 Soil bulk density, effects of cultivation on, 277, 279 Soil carbon, effects of cultivation on, 269 Soil compaction, cultivation and, 275,278 Soil crusting effects of wastewater on, 36 subsurface fertigation and, 17-18 Soil erosion, effects of cultivation on, 276, 280 Soil extractions, in monitoring soils under fertigation, 63 Soil fauna, effects of cultivation on, 284-286

INDEX Soil impedance, ion transport and, 35 Soil leaching, under microinigation, 9- 10 Soil microorganisms biomass, measuring of, 270-271,274 effects of cultivation on, 279,280, 284-285, 286 nitrogen immobilization and, 290 respiration, measuring, 27 1-272, 274 root exudates and, 28 solution culture and, 152 temperature and, 279 Soil mineral nitrogen crop residues and, 289, 290-291 effects of cultivation on, 294-297, 300 mineralization-immobilization balance and, 268 monitoring, 273,274-275 Soil moisture, see also Soil solution(s) crop residue management and, 289 distribution, under fertigation, 8- 13 effects of cultivation on, 279, 288, 289, 298 effects on root growth, 25 effects on salt stress, 14 under fertigation, monitoring of, 62-63 radiation use efficiency and, 252-253 soil root volume and, 13 timing of cultivation and, 275,278 Soil nutrients distribution of, 8-13, 15-17 mobility of, 52-53 modeling of, fertigation management and, 60-62 monitoring of, 63 salt stress and, 14 uptake dynamics, 21 Soil organic matter carbon : nitrogen ratio in, effects on, 283 decomposition, cultivation and, 294 effects of cultivation on, 277,280, 284 effects on fertigation, 57-59 in no tillage, protection of, 293 physical protection of, nitrogen mineralization and, 280-284 soil porosity and, 284 Soil pH, effects of fertigation solutions on, 52 Soil physical properties effects of cultivation on, 268-269.275-280, 298 effects of wastewater on, 36

327

Soil porosity, see also Soil tortuosity effects of cultivation on, 277,278,279 effects of wastewater on, 36 nitrogen mineralization and, 28 I , 283 soil organic matter and, 284 Soil root volume effects on microfertigation, 13-14,25 salt stress and, 14 Soil solution(s), see also Soil moisture buffering processes in, 157 capillary electrophoresis and, 146 composition of, 154-156 plant growth and, 172-173 Soil temperature effects of cultivation on, 279,288, 298 effects on plants under subsurface fertigation, 17 root growth and, 25 Soil texture, cultivation and, effects on nitrogen mineralization, 280-284 Soil tortuosity, see also Soil porosity ion transport and, 35 Soil vacuum cups, 63 Soil water, see Soil moisture; Soil solution(s) Solar constant, 227 Solar energy, see also Radiation measurements of, 227-230 Solarimeters, measurement of incident radiation, 229,23 1-232 Solution culture, see also Nutrient solutions applications of, 151-152 buffering, 153 with bicarbonate, 177-179, 195,201 with chelators, 179-189, 202 with inorganic solid phases, 195200 with ion-exchange resins, 190-195, 202 of oxidation-reduction potential, 189 of pH, 173-179, 190-191,195-196, 20 1-202 effects on root characteristics, 152 elemental toxicity and, 158 flowing solution method, 168-173,201 limitations of, 152 nutrient solution composition, issues in, 154- I60 overview of, 201-203 periodic replacement methods in, 160-167, 20 1

328

INDEX

Solution culture (continued) programmed nutrient addition technique, 162-164 relative addition rate technique, 162, 164-167 research objectives in, 158, 160 simulation of soil nutrition, perspectives on, I66 use of large solution volumes in, 167-168, 20 1 Sorghum, radiation use efficiency in, 236(table), 24 1-242 compared to other crops, 248,249-250 leaf nitrogen and, 252 radiation environment and, 257 during seed growth phase, 256 soil water deficits and, 253 Sorption, in rhizosphere, process of, 30-34 Southern corn leaf blight, see also Cochliobolus heterostrophus epidemic of 1970,87-90.97 Soybean, radiation use efficiency in, 239(table), 246 compared to other crops, 249 leaf nitrogen and, 25 1 radiation environment and, 257 during seed growth phase, 256 Sperm cells, 86 Stem layer, in capillary electrophoresis, 135, 137 Sterols, in monitoring soil fungal biomass, 27 1 Stokes radius, 139 Stress studies, nutritional, chelator-buffering methods in, 186-188 Stubble retention, see Crop residue management Substrate induced respiration, 271 Subsurface fertigation advantages of, 15-1 8 problems with, 18 Succinate, in root exudation, 29 Sugarcane, radiation use efficiency in, 236-237(table), 242-243,248 Sulfate, ion-exchange resins and, 194 Sulfur concentration in nutrient solutions, 157(table) leaf levels, use of MES buffer and, 176 Sulfur bacteria, microfertigation emitter clogging and, 19 Sunflower, radiation use efficiency in, 238(table), 245 -246

compared to other crops, 249 during crop establishment, 255-256 leaf nitrogen and, 25 1-252 during seed growth phase, 256 shading and, 257 Superphosphate, preplanting application, used with fertigation, 5 1 Suppressed conductivity detection, used with capillary electrophoresis, 145 Surface reflectance coefficient, of direct drilled soils, 279 Surfactants, in capillary electrophoresis, 140 Sweet corn, under fertigation nutrient consumption rates, 44(table), 46(table), 48(table) responses to, 16(table), 17 root characteristics, 26, 27, 28(table) studies on, 8(table)

T Tapetum, in T-cytoplasm maize, 80,83-86, 102- 103 TAPS buffer chemical formula and pKu value, 175(table) chemical name, 153(table) in solution culture, 176 T cytoplasm, see Maize, T-cytoplasm lines Temperature, see also Soil temperature effects on fertigation, 56 radiation use efficiency and, 223,253,254 TES buffer chemical formula and pKa value, 175(table) chemical name, 153(table) in solution culture, 176 Texas cytoplasmic male sterility, see Maize, T-cytoplasm lines Theoretical plates, in capillary electrophoresis, 138 Tillage, see Cultivation Time-based growth analyses, 216-217 Tine cultivators, effects soil, 275 Titanium-citrate buffer, 189 Tobacco, T-urj-Z3 expressing, 102 Tomatoes, under fertigation nutrient consumption rates, 41,42(table), 45(table), 47(table), 49 responses to, 5, 6(table), 16(table), 17 rhizosphere pH and, 30 root characteristics, 26,27(table)

329

INDEX ToxIA locus, 92.93 ToxlB locus, 92.93 Toxicity, from irrigation water, elements in, 37 Transpiration, see also Evapotranspiration effects on crop quality and yield, 20 effects on fertigation nutrient uptake dynamics, 23 Tree crops, responses to microfertigation, 6(table) Trickle irrigation, see Microirrigation Trifluralin, microfertigation emitter clogging and, 19 Triton-X 100, in capillary electrophoresis, 140 Tropolone, 141 T-toxin, 86, 88 effects on T-cytoplasm maize mitochondria, 96-97, 100-102 production, genetic control of, 91 -93 reduced fitness of Cochliobolus heterostrophus and, 90 Tube cell, 85 Tube solarimeters, measurement of incident radiation, 229, 23 1-232 T-urfl3 gene control of fungal susceptibility and male sterility, 98 cytoplasmic male sterility and, 1I8 encoded protein (URF13), 99-100 in fungal toxin sensitivity, 100-102 identification of, 99 nuclear restorer genes and, 103-109 organization of, 99 relationship to nuclear restorer genes, 118 in T-cytoplasm maize, 80

U Ultraviolet detection systems, used with capillary electrophoresis, 141, 143, 145 United States, southern corn leaf blight epidemic, 87-90.97 Urea mobility in fertigated soils, 53 pH considerations and, 52 preplanting application, used with fertigation, 50 URF13 protein, 80.86 cytoplasmic male sterility and, 102- 103, 1 18 effects on cell viability, 113-1 14 fungal susceptibility and, 100-102, 119

identification and localization of, 99- 100 nuclear restorer genes and, 109 possible interaction with RF2 protein, 116 USDA cytoplasm, 82.83 UV detection systems, see Ultraviolet detection systems

V Vacuum injection. in capillary electrophoresis, I36 Vapor pressure deficit, radiation use efficiency and, 253,254-255 Vegetables, field, responses to microfertigation, 6-7(table)

W Wastewater in irrigation, quality issues, 36-40 in subsurface fertigation, 18 Water, see Soil moisture; Wastewater Water culture, see Solution culture Water quality effects on irrigation and fertigation, 36-40 nitrate leaching and, 268 Water transport, modeling of, fertigation management and, 60-62 Weak-acid resin buffers, 190- 19 I Weed control, in subsurface fertigation, 17 Wetting front, under microinigation, 9 Wheat photosynthate conversion in, 243 radiation use efficiency in, 237-238(table), 244 compared to other crops, 250 during crop establishment, 255 radiation environment and, 257 Wofaiit SN 36-L ion-exchange resin buffer, 192

Y Yellow leaf blight, 86, see also Mycosphaerella zeae-maydis T-cytoplasm maize and, 95-96 Yield effects of cultivation on, 286-289 factors affecting, 20-2 1

330

INDEX Z

Zero tillage, see No tillage Zinc chelators, 56, 184 concentration in soil solutions, 156 precipitation in soil, 33 in solution culture iron chelators and, lll(tab1e)

nutrient solution levels, 156, 157(table) periodic replacement approach, 161 toxic levels, 158, 159(table) Zinc deficiency studies chelator-buffering methods and, 185-186 hydroxyapatite and, 199-220 sand-alumina methods and, 199 solution culture methods and, 162 Zwitterions, Good’s buffers and, 174