No title

Advisory Board Martin Alexander

Ronald Phillips

Cornell University

University of Minnesota

KennethJ. Frey

Larry P. Wilding

Iowa State University

Texas A&M University

Prepared in cooperation with the

American Society of Agronomy Monographs Committee William T. Frankenberger, Jr., Chairman P. S. Baenziger David H. Kral Dennis E. Rolston Jon Bartels Sarah E. Lingle Diane E. Storr Jerry M. Bigham Kenneth J. Moore Joseph W. Stucki M. B. Kirkham Gary A. Peterson

DVANCES IN

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 1998 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 01923). 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-1997 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. 0065-2 1 13/98 $25.00

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 Limited 24-28 Oval Road, London NW I 7DX. UK http://www.hbuk.co.uk/ap/ International Standard Book Number: 0- 12-000764-9 PRINTED IN THE UNl'IED STATES OF AMERICA 98 99 0 0 0 1 02 0 3 B B 9 8 7 6

5

4

3 2 I

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

ix xi

CYTOGENETICS AND GENETICS OF PFARL MILLET Prem P.Jauhar and Wayne W. Hanna I . Introduction .............................................. I1. Origin ................................................... I11. Taxonomic Treatment ......................................

rv. v.

VI. VII . VIII . Ix. x. XI . XI1. XIII .

m

Chromosomes. Karyotype. and Meiosis ........................ Genome Relationships...................................... Aneuploidy and Gene Mapping .............................. Molecular Markers and Gene Mapping ........................ Wide Hybridization with Pearl Millet ......................... Wide Hybridization and Genetic Enrichment for Fodder Traits .... Hybridization and Exploitation of Hybrid Vigor . . . . . . . . . . . . . . . . . Apomixis ................................................. Genetics of Qualitative Traits ................................ Genetics of Quantitative Traits ............................... Conclusion and Perspectives ................................. References ...............................................

2 3 4 6 8 10 10 11 11 13 16 18 19 19 21

ADVANCESINICP EMISSION AND ICP MASSSPECTROMETRY Parviz N . Soltanpour. Greg W.Johnson. Stephen M .Workman. J . Benton Jones. Jr., and Robert 0. Miller I . Introduction .............................................. 28 I1. ICP-AES and ICP-MS Instrumentation ....................... 31 I11. Spectrometers ............................................ 42 w. Analytical Capabilities ...................................... 44 v. ICP-AES Interferences ..................................... 78 VI. ICP-MS Interferences ...................................... 83

VII . Practical Applications ...................................... VZII . Quality Control Methods ...................................

V

91 99

vi

CONTENTS

IX. Summary ................................................ Appendix ................................................ References ...............................................

99 100 106

MANAGINGCOTTON NITROGEN SUPPLY

Thomas J . Gerik. Derrick M . Oosterhuis. and H . Allen Torbert I . Inuoduction .............................................. I1. Cotton Growth and Nitrogen Response ....................... I11. Soil Nitrogen Availability and Dynamics ....................... n! Foliar-Nitrogen Fertilization in Cotton ........................ V. Monitoring Cotton Nirrogen Status........................... VI. Managing Cotton Nitrogen Supply ........................... VII . Summary ................................................ References ...............................................

116 118 128 132 133 138 142 142

ARSENICINTHE Son. ENVIRONMENT: A REVIEW

E. Smith. R . Naidu. and A. M. Alston I . Introducdon .............................................. Position in the Periodic Table ................................ Background Sources ....................................... Anthropogenic Sources ..................................... AsToxicity ............................................... VI. Physiochemical Behavior of As in Soil ......................... VII . Soil As and Vegetation ...................................... VIII . Soil As and Microorganisms ................................. M . Conclusions .............................................. References ...............................................

I1. I11. n!. V.

150 1 SO 151 153 163 165 179 182 186 187

DRYLAND CROPPING INTENSIFICATION: A FUNDAMENTAL SOLUTION TO EFFICIENT USEOF PRECIPITATION H .J . Farahani. G. A . Peterson. D . G. Westfall I . Introduction .............................................. I1. Summer Fallow: A Second Look.............................. I11. Dryland Cropping Intensification............................. N. A Systems Approach to Intensification ......................... V. Conclusion ............................................... References ...............................................

197 201 203 213 221 222

CONTENTS

vii

How Do PLANT ROOTSACQUIREMINERAL NUTRIENTS? CHEMICAL PROCESSES INVOLVED INTHERHIZOSPHERE P. Hinsinger I . Introduction .............................................. I1. Definition of the Rhizosphere ................................ I11. Root-Induced Changes of Ionic Concentrations in the Rhizosphere ........................................... Iv. Root-Induced Changes of Rhizosphere p H ..................... V. Root-Induced Changes of Redox Conditions in the Rhizosphere . . . . VI . Root-Induced Complexation of Metals in the Rhizosphere . . . . . . . . . VII . Other Interactions Involving Root Exudates .................... VIII. Conclusion ............................................... References ...............................................

228 237 242 247 253 254 257

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

267

225 226

This Page Intentionally Left Blank

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

A. M. ALSTON (149), Department of Soil Science, University of Adelaide, Glen Omond, South Australia 5064, Australia H. J. FARAHANI (197), USDA-Agricultural Research Service, Great Plains Systems Research, Fort Collins, Colorado 80521 THOMAS J. GERlK (1 1S), Texas Agricultural Experiment Station, Blackland Research Center, Temple, Texas 76502 WAYNE W. HANNA (l), USDA-Ap‘cultural Research Service, Coastal Plain Experiment Station, Tifton, Georgia 31 793 P. HINSINGER ( 2 2 9 , Faculty of Agriculture, University of Western Australia, Nedlandr, Western Australia 6907, Australia PREM P. JAUHAR (I), USDA-Agricultural Research Service, Northern Crop Science Laboratoy, State University Station, Fargo, North Dakota 58105 GREG W.JOHNSON (2 7), Matheson Gas Products, Longmont, Colorado 80501 J. BENTON JONES, JR. (27), Macro-Micro Analytical Services, .fthens, Georgia 30607 ROBERT 0.MILLER (27), Department of Soiland Crop Sciences, Colorado State University, Fort Collins, Colorado 80523 R. NAIDU (149), CRCfor Soil and Land Management and CSIRO Division of Soils, Glen Omond, South Australia 5064, Australia DERRICK M. OOSTERHUIS (11S), Department of Agronomy, University of Arkansas, Fayetteville, Arkansas 72703 G. A. PETERSON (197), Department of Soil and Crop Sciences, Colorado State University, Fort Collins, Colorado 80523 E. SMITH (149), CRCfor Soil and Land Management and Department of Soil Science, University of Adelaide, Glen Omond, South Australia 5064, Australia PARVIZ N. SOLTANPOUR (27), Department of Soil and Crop Sciences, Colorado State University,Fort Collins, Colorado 80523 H. ALLEN TORBERT (1 1S), USDA-Ap‘cultural Research Service, Grassland Soil and Water Research Laboratoy, Temple, Texas 76502 D. G. WESTFALL (197), Department of Soil and Crop Sciences, Colorado State University, Fort Collins, Colorado 80523 STEPHEN M. WORKMAN (27), Analytical Technologies, Inc., Fort Collins, Colorado 80524

ix

This Page Intentionally Left Blank

Preface Volume 64 contains six contemporary and comprehensive reviews that will be of interest to plant and soil scientists, and to scientists in allied fields. Chapter 1 is concerned with the cytogenetics and genetics of pearl millet. The authors discuss taxonomy of pearl millet; chromosomes, karyotype, and meiosis; genome relationships; aneuploidy and gene mapping; molecular markers and gene mapping; wide hybridization and genetic enrichments for fodder traits; and exploitation of hybrid vigor, apomixis, and genetics of qualitative and quantitative traits. Chapter 2 is a comprehensive chapter on advances in ICP-emission and ICP-mass spectrometry. The review covers instrumentation spectrometers, analytical capabilities, ICP-AES and ICP-MS interferences, practical applications, and quality control methods. Extensive tabular data are included on prominent lines of the elements emitted by the ICP, isotope data for elements, detection limits, interelemental spectral interferences, and preparation of primary standard solutions. Chapter 3 discusses managing cotton nitrogen supply. Topics covered include cotton growth and nitrogen response, soil nitrogen availability and dynamics, foliarnitrogen fertilization in cotton, and monitoring cotton nitrogen supply. Chapter 4 is a timely and extensive review on arsenic (As) in the environment. The authors discuss sources of As, its toxicity and physicochemical behavior in soil, soil As and vegetation and plant uptake, and biotransformations of As. Chapter 5 deals with dryland cropping intensification and covers summer fallowing, dryland cropping intensification, and a systems approach to intensification. Chapter 6 is a thoughtful review on chemical processes involved in the rhizosphere. The author describes root-induced changes of ionic concentrations, pH, and redox conditions in the rhizosphere, and other interactions involving root exudates. I am most grateful to the authors for their first-rate reviews.

DONALD L. SPARKS

xi

This Page Intentionally Left Blank

CYTOGENETICS AND GENETICS OF PEARLLET Prem P. Jauhar' and Wayne W. Hanna* I USDA-Agricultural

Research Service Northern Crop Science Laboratory State University Station, Fargo, North Dakota 58105 WSDA-Agricultural Research Service Coastal Plain Experiment Station Tifton, Georgia 31793

I. Introduction

II. Origin

111. Taxonomic Treatment

IV

V. VI. VII. VIII. M. X.

XI.

A. Taxonomic Placement of Pearl Millet B. Wild Annual Relatives of Pearl Millet C. Perennial Relatives of Pearl Millet Chromosomes, Karyotype, and Meiosis A. Chromosomes as Multiples of 5 , 7 , 8 , and 9 and Size Differences B. Chromosomes of Pearl Millet and Other Penicillarias C. Evolution of the Chromosome Complement of Pearl Millet Genome Relationships Aneuploidy and Gene Mapping Molecular Markers and Gene Mapping Wide Hybridization with Pearl Millet Wide Hybridization and Genetic Enrichment for Fodder Traits A. Interspecific Hybrids B. Intergeneric Hybrids Hybridization and Exploitation of Hybrid Vigor A. Grain Hybrids B. Forage Hybrids C. Germplasm D. Types of Hybrids Apomixis A. Incidence of Apomixis in Pennisentm Species B. Genetics of Apomixis C. Harnessing Apomixis for Exploitation of Heterosis

Mention of a trademark or proprietary product does not constitute guarantee or warranty of the product by the USDA or imply approval to the exclusion of other products that also may be suitable. 1 Adumcx in Agmnmy, Volume 64

Copyright 8 1998 by Academic Press. All rights of repmducuon in any form reserved 0065-2113/98 $25.00

2

PREM P. JAUHAR AND WAYNE W. HANNA XI. Genetics of QualitativeTraits XIII. Genetics of Quantitative Traits xn! Conclusion and Perspectives References

I. INTRODUCTION Pennisetum is one of the most important genera of the family Poaceae. It includes such important species as pearl millet, Pennisetum glaucum (L.) R. Brown [ =Pennisetum typhoides (Bum.) Stapf et Hubb., Pennisetum americanum (L.) Schumann ex Leeke] (2n = 14),a valuable grain and forage crop; and its tetraploid relative Napier grass (I? purpureum Schum.) (2n = 4x = 28), prized for its fodder grown throughout the wet tropics of the world. Pearl millet is widely cultivated in different parts of the world. It is a multipurpose cereal grown for grain, stover, and green fodder on about 27 million hectares, primarily in Asia and Africa (ICRISAT, 1996). In terms of annual production, pearl millet is the sixth most important cereal crop in the world, following wheat, rice, maize, barley, and sorghum. Among the millets, it is second only to sorghum. Pearl millet is the only cereal that reliably provides both grain and fodder on poor, sandy soils under hot, dry conditions. It is remarkable that it produces nourishment from the poorest soils in the driest regions in the hottest climates. In the drier regions of Africa and Asia, the crop is a staple food grain. In more favored areas, however, pearl millet grain is fed to bullocks, milch animals, and poultry. In areas where other types of feed are not available, stover provides feed for cattle (ICRISAT, 1996). Pearl millet is also grown in several other countries. It was planted to almost 1 million hectares in Brazil in 1996. In the United States, it is grown as a forage crop on an estimated half a million hectares. It is also grown as a forage crop in tropical and warm-temperate regions of Australia and several other countries (Jauhar, 198la). Pearl millet is an ideal organism for cytogenetic and breeding research. Several favorable features of its chromosome complement--e.g., the small number and large size of chromosomes with distinctive nucleolar organizers-make pearl millet a highly suitable organism for cytogenetic studies. Because of its low chromosome number, pearl millet offers a particularly favorable material for aneuploid analysis and thereby elucidation of its cytogenetic architecture. Moreover, its protogynous flowers and outbreeding system make it ideal for interspecific hybridization and breeding work, particularly heterosis breeding. Pearl millet has also been found suitable for molecular studies. Although pearl millet has great agricultural importance and is a favorable organism for cytogenetic and molecular studies, it has not received the attention it deserves. Consequently, the information available on its genetics and cytogenetics is far less than that available for other agricultural crops. In a comprehensive

CYTOGENETICSAND GENETICS OF PEARL MILLET

3

review, Jauhar (1981a) compiled the available literature on the cytogenetics and breeding of pearl millet and related species. The purpose of this article is to summarize the information on cytogenetics and genetics of pearl millet mostly since the publication of Jauhar’s book (198 la).

II. ORIGIN Pearl millet originated in West Africa, where it grows in chronically droughtprone areas. Selection exercised by early cultivators within a variety of cultural contexts resulted in a multitude of morphologically diverse forms. The protogynous flowers of pearl millet facilitated the introgression of characters from related wild species to cultivated annual species. Although researchers generally agree that pearl millet is of African origin, pinpointing its specific region of origination has been controversial. Vavilov (1949-1950) placed pearl millet in the Ethiopian Center of Origin (particularly Abyssinia and Sudan), considering this the region of maximum diversity. However, the center of diversity is not always the center of origin (Harlan, 1971). In light of the great morphological diversity present in introductions from Central Africa, Burton and Powell (1968) inferred that pearl millet originated there. Another method used to pinpoint its center of origin is the occurrence of B chromosomes. Because B chromosomes frequently occur in primitive varieties but not in commercially bred cultivars, Muntzing ( 1958) suggested that their occurrence might indicate a crop’s center of origin. Therefore, based on the occurrence of B chromosomes in pearl millet collections, some researchers consider Sudan (Pantulu, 1960) and Nigeria (Powell and Burton, 1966; Burton and Powell, 1968) to be the crop’s centers of origin. However, drawing conclusions on the basis of occurrence of B chromosomes may not be scientifically sound (Jauhar, 1981 a), because several ecological and edaphic factors influence the occurrence of B chromosomes. In rye (Secafecereale), for example, the frequency of Bs is higher in rnaterial growing on acidic soils than on basic soils (Lee, 1966). Working on clonal plants of rye grown under different regimes of soil, temperature, and humidity, Kishikawa (1970) found that the frequency of Bs was lower in progeny derived from plants grown under high temperatures or dry soil conditions. Considering that the greatest morphological diversity of pearl millet occurs in West Africa, south of the Sahara Desert and north of the forest zone, and that the wild progenitor also occurs in the drier, northern portions of this zone, Harlan (197 1 ) suggested that the center of origin lies in a belt stretching from western Sudan to Senegal. Based on present-day distributions, the Sahel region of West Africa appears to be the original home of pearl millet (Brunken et al., 1977). The cultivated types show the highest level of morphological variability in this region (Clegg et al., 1984).

4

PREM P. JAUHAR AND WAYNE W. HANNA

Traditionally, characterization of genetic resources of crop plants has been accomplished through a combination of morphological and agronomic traits, e.g., growth habitat, earliness, and disease and pest resistance. Biochemical and molecular markers have also been used to obtain additional information on a crop plant’s center of domestication, the effect of domestication on genetic diversity, and potential gene flow between wild and cultivated types (Gepts and Clegg, 1989). However, using restriction fragment length polymorphisms (RFLPs) among chloroplast, nuclear ribosomal RNA, and alcohol dehydrogenase (ADH) sequences in a group of 25 wild and 54 cultivated accessions of pearl millet, Gepts and Clegg (1989) could not identify the precise pattern of its domestication. Brunken et al. (1977) hypothesized the existence of several independent domestications of pearl millet in the southern fringe of the Sahara. Based on polymorphisms in 12 genes coding for 8 enzymes in 74 cultivated samples and 8 wild samples from West Africa, the 82 samples were classified into three groups: (1) wild types, (2) early maturing cultivars, and (3)late cultivars (Tostain et al., 1987). The early maturing cultivars were found to have the highest enzyme diversity, whereas cultivars from Niger showed the most diversity. The high diversity of the early maturing group and its extensive divergence from West African wild millets further suggest multiple domestications.

III. TAXONOMIC TREATMENT A. TAXONOMIC PLACEMENT OF PEARL MILLET Pearl millet is the most important member of the genus Pennisetum in the tribe Paniceae. It has received a variety of taxonomic treatments, and its scientific binomials have been frequently shuffled by a variety of taxonomists. Consequently, it has had many Latin names, perhaps more than any other grass. In the post-Linnaean period from 1753 to 1809, pearl millet was treated as a member of at least six different genera, namely, Panicum, Holcus, Alopecuros, Cenchrus, Penicillaria, and Pennisetum (see Jauhar, 1981a,c). At the beginning of this century, pearl millet was commonly referred to as Pennisetum typhoideum, Penicillaria spicata, Panicum spicatum, and Pennisetum alopecuroides (Chase, 1921). By the mid-19th century, however, pearl millet was generally called Pennisetum typhoideum L. C. Rich, but this nomenclature was not widely accepted. The Latin name Pennisetum americanum given by K. Schumann (1895)-apparently based on the first name “Panicum americanum L.” used by Linnaeus (1753bwas accepted by Terrell (1976) and hence used by several American workers. However, this name is inappropriate and misleading because it inadvertently implies the American origin of pearl millet (Jauhar, 1981a,c).

CYTOGENETICSAND GENETICS OF PEARL MILLET

Stapf and Hubbard (1933, 1934) gave the name Pennisetum fyphoides (Bum.) Stapf et Hubb., which was accepted by several modem taxonomists, including Bor (1960), and used by most pearl millet workers outside the United States. In the 1960s, American workers joined the rest of the world in calling pearl millet Pennisetum ophoides (Burton and Powell, 1968).The name Pennisetum glaucum (L.) R. Br., based on Panicum glaucum (L.) R. Br., was adopted by Hitchcock and Chase (195 1) in Manual of the Grasses of the United States. Consequently, American scientists currently engaged in research on pearl millet use this name. All annual and perennial members of the section Penicillaria fall under the x = 7 group. They have typically penicillate anther tips. Whereas most penicillarias are diploid with 2n = 14 chromosomes, one, viz., Napier grass, is a perennial tetraploid.

B. WILDANNUAL RELATIVES OF PEARL MILLET Of the 32 species described by Stapf and Hubbard (1934) in the section Penicillaria of the genus Pennisetum, only two have been found outside Africa. There is considerable variation in seed and other characters both between and within different cultivars or races. Such variation could be attributed to independent domestications and migrational events resulting in geographical isolations. The protogynous nature of pearl millet and its intercrossabilitywith its wild relatives must have generated much of the existing genetic diversity. Meredith (1955) described four taxa, which he called “allied species,” closely related to pearl millet: Pennisetum americanum, I? nigritarum, I! echinurus, and I? albicauda. Since these are interfertile with pearl millet, they were merged into a single species with pearl millet (Brunken et al., 1977). However, for the sake of convenience, Brunken subdivided the morphologically heterogeneous pearl millet species he called “Pennisetum americanum” into three subdivisions: ( 1) ssp. americanum encompasses the wide array of cultivated pearl millets; (2) ssp. monodii includes all the wild and semiwild diploid races that are fully fertile with pearl millet and therefore form a single reproductive unit with it; and (3) ssp. stenostachyum is morphologically intermediate between the two preceding species. Amoukou and Marchais (1993) found some evidence of a partial reproductive bamer between wild and cultivated pearl millets. Crosses between 16 cultivated accessions (f? glaucum ssp. glaucum) (as female parents) and 11 wild accessions (f? glaucum ssp. monodii), from the whole range of diversity of the species, showed certain degrees of seed malformation and reduced 1000-grain-weightand germination ability. These are manifestations of a genetic imbalance between the cultivated and the wild groups, probably resulting from reproductive barriers that developed during the domestication process.

6

PREM P. J A W AND WAYNE W. H A N N A

C. PERENNIAL RELATIVESOF PEARL MILLET Elephant or Napier grass, Pennisetum purpureum (2n = 4x = 28), is a perennial relative of pearl millet (see Section V). It has typically penicillate anthers. Native to Africa, it is a robust perennial with creeping rhizomes. It was introduced into the United States in 1913. It is extensively grown in the humid tropics throughout the world.

N.CHROMOSOMES, KARYOTYPE, AND MEIOSIS A. CHROMOSOMES AS MULTIPLES OF 5,7,8, AND 9 AND SIZEDIFFERENCES The genus Pennisetum is a heterogeneous assemblage of species with chromosome numbers as multiples of 5 , 7, 8, and 9, for example, P. ramosum (2n = lo), P. ryphoides (2n = 14) and P. purpureum (2n = 28), P. massaicum (2n = 16,32), and P. orientale (2n = 18, 36, 54). The chromosome morphology is diverse and substantial size differences exist. A notable feature is that species with lower chromosome numbers have larger chromosomes. Thus, pearl millet (2n = 14) and P. ramosum (2n = 10) have relatively large chromosomes, larger than those of other members of the tribe Paniceae. In contrast, species with higher chromosome numbers, e.g., I? orientale (2n = 18), have strikingly smaller chromosomes than those of pearl millet (2n = 14) (Fig. 2C). A characteristicfeature of perennial species of Pennisetum is the occurrence of chromosomal races or cytotypes, e.g., P. orientale L. C . Rich. (2n = 18, 27, 36, 45, 54) and F! pedicellatum Tin. (2n = 36,45,54). However, no such cytotypes occur in the annual cultivated or wild pearl millets, all of which have 2n = 14 chromosomes.

B. CHROMOSOMES OF PEARL MILLET AND &HER

PENICILLARTAS

Rau (1929) was the first to determine the somatic chromosome number of pearl millet as 2n = 14, and he mentioned these chromosomes as being large. The chromosomes have median to submedian centromeres; the shortest chromosome pair is satellited, and during meiosis the shortest bivalent is associated with the nucleolus. The chromosomes of diploid taxa of the section Penicillaria are similar to those of pearl millet. Thus, I? ancylochaete, P. gambiense, I! maiwa, and I? nigritarum have 2n = 14 chromosomes, and their chromosome morphology is similar to one another and to chromosomes of pearl millet (Veyret, 1957). Not surpris-

CYTOGENETICS AND GENETICS OF PEARL MILLET

7

ingly, therefore, these taxa are interfertile with pearl millet, and there is no barrier to gene flow across these taxa. Pennisetum violaceum and R mollissimum, the two close wild relatives that form a primary gene pool with pearl millet, and I? schweinfurthii (a representative species of tertiary gene pool) were assessed for their genomic organization, using in situ hybridization with rDNA probes on somatic metaphase spreads and interphase nuclei (Martel et al., 1996). These studies showed chromosomal similarity of rDNA sequence locations in the three taxa in the primary gene pool. Pearl millet regularly forms seven bivalents at meiotic metaphase I. A characteristic feature is the rapid terminalization of chiasmata, such that at diakinesis mostly loose ring bivalents with two terminalized chiasmata each are observed. The annual, semiwild taxa also have regular meiosis with 7 11. They all have the genomic constitution AA. Recently, Reader et al. (1996) used fluorescence in situ hybridization (FISH) to characterize the somatic complement of pearl millet. A metaphase spread was hybridized with Fluorored-labeled rDNA (derived from plasmic clone pTa71; Gerlach and Bedbrook, 1979) and then stained with DAPI. In that double exposure. two large and two small NOR loci were observed. Napier grass is a perennial relative of pearl millet. Burton (1 942) determined its somatic chromosome number as 2n = 28 chromosomes. It is an allotetraploid (2n = 4x = 28) with diploidlike meiosis (see Jauhar, 1981a). It is genomically represented as AABB, the A genome being largely homologous to the A genome of pearl millet (see Section V).

C. EVOLUTION OF THECHROMOSOME COMPLEMENT OF PEARL MILLET Researchers generally believe that several crop species have evolved from species with lower basic chromosome numbers, with increase in chromosome number occurring by means other than straight polyploidy. Evidence supporting this view has been found by RFLP studies of maize (Helentjaris et al., 1986; Whitkus et al., 1992), brassicas (Slocum et al., 1990; Kianian and Quiros, 1992), and sorghum (Hulbert et al., 1990; Whitkus et al., 1992; Chittenden et ul., 1994). Based on cytogenetic evidence, Jauhar (1968, 1970a, 1981a) hypothesized that x = 5 may be the original basic number in Pennisetum and that pearl millet (2n = 14) may be a secondary balanced species as a result of ancestral duplication of chromosomes. If duplication of a part of the original genome occurred during the evolution of pearl millet, some duplicate loci should be observed in the present genome. Liu et al. (1 994) indeed detected several duplicate loci in their RFLP linkage map of the pearl millet genome. However, further studies are needed to fully characterize the duplicated regions of the genome.

8

PREM P. JAUHAR AND WAYNE W. HANNA

V. GENOME RELATIONSHIPS Knowledge of genome relationships between plant species is very useful in planning effective breeding strategies designed to transfer desirable genes or gene clusters from one species into another, thereby producing fruitful genomic reconstructions. Traditionally, the principal method of assessing the genomic affinities among species has been the study of chromosome pairing in their hybrids (Jauhar and Joppa, 1996). Genomic relationships are inferred from the degree of pairing between parental chromosomes.However, pairing in the hybrids may be due to allosyndesis (Le., pairing between chromosomes of the parental species) andor autosyndesis (i.e., pairing within a parental complement).Therefore, information on the nature of chromosome pairing is important for assessing the genomic relationships. The chromosomes of pearl millet are much larger than those of other species of Pennisetum (e.g., see Fig. 1). This size difference makes it possible to study intergenomic chromosome pairing relationships. A clearly distinguishable size difference between chromosomes of pearl millet (2n = 14 large chromosomes; AA genome) and those of Napier grass (2n = 28 relatively small chromosomes; AABB) makes it possible to study, in their hybrids (e.g., see Figs. 2A, 2B), the degree of allosyndetic and autosyndetic pairing (Jauhar, 1968). Based on pairing in triploid hybrids (2n = 3x = 21; AAB), it was inferred that the two species basically share a genome (A and A being very similar). However, the source of B genome remains unknown.

Figure 1 Somatic chromosomesof a hybrid between pearl millet and fountain grass, Penniseturn setaceurn (Forsk.) Chiov. Note the 7 large pearl millet chromosomes and 18 much smaller fountain grass chromosomes.

CYTOGENETICSAND GENETICS OF PEARL MILLET

L

;*

9

P

.;

'I)

* I .

A

C

D

Figure 2 Chromosome pairing in interspecific hybrids (2n = 3x = 21;AAB) between pearl millet (2n = 2x = 14;AA) and Napier grass (2n = 4x = 28;AAB). (A) Metaphase I showing 21 univalents-7 large ones from pearl millet (arrows) and 14 small ones from Napier grass. (B) Metaphase I with 7 11 (2 11 overlapping) + 7 I; the bivalents comprise 2 large, symmetrical bivalents within the A genome (hollow arrows), 1 heteromorphic intergenomic bivalent between chromosomes of A and A genomes (solidarrow),and 4 intragenomic bivalents within A and B genomes. Note 2 large univalents of the A genome. (C, D)Chromosome pairing in interspecific hybrids (2n = 16) between pearl millet (2n = 14) and P. orienrule (2n = 18). (C) Diakinesis with 16 univalents-7 large ones (arrows) from pearl millet and 9 small ones from orientale. Note the striking size differences among the parental chromosomes. (D) Metaphase I with 2 heteromorphic bivalents between pearl millet chromosomes and orientale chromosomes (solidarrows),and 1 autosyndetic bivalent within the orienrule complement (hollow arrow). (Reprinted from Jauhar, 1981a. by permission of the publisher.)

10

PREM P. JAUHAR AND WAYNE W. HANNA

Even more striking size differences exist between the chromosomes of pearl millet and those of oriental grass (Penniseturn orientale; 2n = 18) (Fig. 2C). The nature of chromosome pairing was analyzed in hybrids between these species (Patil and Singh, 1964;Jauhar, 1973,1981a,b). Association between chromosomes of the parental species resulted in the formation of conspicuously heteromorphic bivalents (Fig. 2D), suggesting an ancestral relationship between the two species. In addition to intergenomic pairing, intracomplement associations within the glaucum and the orientale complements were also observed.

VI. ANEUPLOIDY AND GENE MAPPING The establishment of a complete series of aneuploids is very useful in elucidating the cytogenetic architecture of a crop plant. Jauhar initiated work on the isolation of aneuploids of pearl millet. From the progeny of triploid X diploid crosses, he isolated two primary trisomics (2n + I = 15) (Jauhar, 1970b). Jauhar (198 la) summarized research on aneuploids in pearl millet. Over the years, there have been numerous reports on double trisomics, triple trisomics, double telotrisomics, ditertiary compensating trisomics, multiple interchange trisomics, and so on. Minocha et al. (1980a) described a set of primary trisomics and used them to assign genes to five of the seven chromosomes. Vari and Bhowal(1985) reported a set of primary trisomics distinguishable by morphological characteristics. Using trisomic analyses, Sidhu and Minocha (1984) located genes controlling peroxidase isozyme production on all seven chromosomes. Minocha et al. (1 982) described a translocation tester set of five translocation stocks, each of which involved two nonhomologous chromosomes. Rao et al. (1988) described various types of trisomics, some involving interchanges, and also reviewed some of the earlier work on aneuploids in pearl millet. However, it appears that little use has been made of these aneuploids and translocation stocks in genetic and breeding studies.

VII. MOLECULAR MARKERS AND GENE MAPPING An important aspect of genetic research is creating genetic maps that are useful to geneticists and plant breeders. DNA markers can be employed in the construction of genetic maps, which help determine the chromosomal location of genes affecting either simple or complex traits (Paterson et al., 1991). With these molecular methods, genetic maps of diploid plants can be developed more rapidly than those of polyploids. Pearl millet has a haploid (1C) DNA content of about 2.5 pg (Bennett, 1976).

CYTOGENETICS AND GENETICS OF PEARL MILLET

11

Using RFLP, Liu et al. (1994) constructed a linkage map of pearl millet. The RFLP map so generated is relatively dense, with a 2 cM distance between markers. However, specific chromosome regions with tightly linked markers are still evident. Using molecular markers, Jones ef al. (1995) assigned part of the genes controlling quantitatively inherited resistance to downy mildew to linkage group 1, 2,4, 6, and 7 of pearl millet. Busso et al. (1995) used RFLP markers to study the effect of sex on recombination in pearl millet. They found no differences in recombination distances at the whole-genome level; only a few individual linkage intervals differed, but all were in favor of increased recombination through the male. These results are contrary to those obtained with tomato. Using RFLP markers to compare male and female recombination in two backcross populations of tomato, De Vicente and Tanksley (199 1) reported a significantly higher recombination rate in female meiosis.

Vm. WIDE HYBRIDIZATION WITH PEARL MILLET In recent years, experimental hybridization has been effected between taxonomically distant taxa. Using pearl millet as a pollen parent in crosses with barley, Zenkteler and Nitzsche ( 1984) obtained globular embryos. In crosses between hexaploid spring wheat cv. Chinese Spring and the pearl millet genotype Tift 23 BE, Laurie (1989) observed fertilization in 28.6% of the 220 florets pollinated. Chromosome counts in zygotes confirmed the hybrid origin of the embryos; three embryos had the expected 21 wheat and 7 pearl millet chromosomes and a fourth had 21 wheat and 14 pearl millet chromosomes. However, the hybrid embryos were cytologically unstable and probably lost all of the pearl millet chromosomes in the first four cell division cycles. The elimination of pearl millet chromosomes at an early stage will limit the chances of gene transfer from pearl millet into wheat. In crosses between five cultivars of oat with pearl millet (as pollinator), Matzk (1996) obtained a hybrid frequency of 9.8%. However, the pearl millet chromosmes were lost during embryo or plant development. In one hybrid, 5 pearl millet chromosomes were retained with 21 of oat. Hybrids like this could offer an opportunity for transfer of pearl millet genes into oat or vice versa. Such hybrids could also help produce alien addition or substitution lines in the two crop plants.

M.WIDE HYBRIDIZATION AND GENETIC ENRICHMENT FOR FODDER TRAITS The potential for producing and using hybrids for forage production is greater in Pennisetum than in many other genera. A number of the species can be inter-

12

PREM P. JAUHAR AND WAYNE W. HANNA

crossed with various degrees of ease. Bridging species can be used to increase success of wide crosses. Pearl millet usually contributes vigor and high forage quality to wide hybrids, whereas the wild species contributes perennial growth habit and short-day sensitivity to extend the vegetative growing period. Successful propagation of hybrids will depend on commercial production of hybrid seed (usually in a frost-free or tropical area) (Osgood et al., 1997), vegetative propagation, andor apomictic seed production.

A. INTERSPECIFIC HYBRIDS The pearl millet (2n = 2.x = 14, AA genome) X Napier grass (2n = 4x = 28, AABB genomes) cross produces a vigorous, sterile triploid (PMN) hybrid (2n = 3x = 2 1, AAB). This hybrid can be produced by hand pollinations from which superior plants can be vegetatively propagated, or commercial hybrid seed can be produced on a cms (cytoplasmic-nuclear male sterile) pearl millet in a tropical area (Osgood er al., 1997). The interspecific hybrid needs to be produced in a frost-free area, because Napier grass is short-day sensitive and will not mature seed in the traditional pearl millet hybrid seed production areas. The PMN hybrids are perennial and extend the vegetative growing period into late fall. Muldoon and Pearson (1979) and Jauhar (1981a) published extensive reviews on most aspects of these hybrids. A 3-year study conducted by Hanna and Monson (1980) on 20 PMN hybrids showed that they can significantly out-yield the best pearl millet hybrids. Hanna and Monson also found that interspecific hybrids made with a tall cms pearl millet parent out-yielded those made with a dwarf parent. Napier grass genotypes varied in their combining ability with pearl millet to produce superior hybrids, and certain Napier grass pollinators produced varying amounts of seedling lethals in crosses with pearl millet. Schank and Hanna (1995) summarized reseal ch on the forage potential of derivatives of the PMN triploid hybrid. Doubling the chromosome number of the PMN triploid results in a seed fertile hexaploid (2n = 6x = 42, AAAABB) with excellent forage potential and which can be vegetatively or seed propagated. A vigorous leafy sterile tetraploid (2n = 4x = 28, AAAB) is produced when the fertile hexaploid is backcrossed to diploid pearl millet. This sterile tetraploid is perennial and can be vegetatively propagated. High-forage-yielding, leafy, perennial trispecific hybrids can be produced by pollinating the fertile hexaploid PMN hybrids with fertile apomictic hybrids from tetraploid pearl millet X apomictic I? squamulatum crosses (Hanna et al., 1989). Apomictic genotypes can be selected among the trispecific hybrids that combine germplasm from pearl millet, Napier grass, and P. squamulatum. Hussey er al. (1993) showed that 2n n hybrids from the P. jaccidum X P. mezianum cross have excellent forage potential. The pearl millet X P. squamularum hybrid has forage potential but does not appear to be as high-yielding as the preceding hybrids

+

CYTOGENETICS AND GENETICS OF PEARL MILLET

13

(Patil and Singh, 1980a; Hanna et al., 1989). Several other interspecific hybrids and derivatives (Patil and Singh, 1980; Jauhar, 1981a; Hanna et al., 1992) have been produced, but more research is needed to establish their forage potential. More potential exists for producing vigorous 2n n hybrids among the apomictic Pennisetum species.

+

B. INTERGENERIC HYBRIDS Jauhar (1981a) and Patil and Singh (1980) summarized studies on various intergeneric hybrids involving Pennisetum species. Most intergeneric hybrids are weak, and/or more information is needed to establish their usefulness. Hussey et al. (1993) reported on a 2n n Cenchrus ciliaris X t?orientale intergeneric hybrid that had excellent forage potential. It appears that more potential exists for producing vigorous hybrids between Cenchrus and Pennisetum species by taking advantage of the relatedness of these genera, apomixis, and the potential for 2n + n fertilization.

+

X. HYBRIDIZATION AND EXPLOITATION OF HYBRID VIGOR Heterosis is significant in pearl millet for both grain and forage production. Use of hybrids is increasing each year in all of the pearl millet growing areas except Africa and Pakistan. Most cultivars grown outside the major pearl millet growing areas of Africa, India, and Pakistan are hybrids and used for forage. However, there is an increased emphasis on production of grain hybrids in the United States. Researchers estimate that 40% of the cultivars in India are F, hybrids, but the areas planted to hybrids range from about 95% in Gujarat to about 10% in Rajasthan (Andrews, 1987; Dave, 1987). Reviews and summaries on the history and progress of inbred and hybrid development and breeding methods used to produce superior hybrids have been published by Andrews (1987), Andrews et al. (1989, Burton (1983), Jauhar (1981a), Anand Kumar and Andrews (1984), Rachie and Majmudar (1980), and Williams and Andrews (1983).

A. GRAINHYBRIDS Heterosis for grain yield in pearl millet was recognized in the mid-1940s. The first pearl millet hybrids released were X . 1 and X.2. These were single-cross grain hybrids produced by chance hybridization due to protogyny and yielded on average 45% more grain than the local types (Rao et al., 1951). It was recognized at

14

PREM l? JAUHAR AND WAYNE W. H A N N A

that time that a commercial method for producing 100% hybrid seed was needed, because the varying amounts of selfed seed produced by the chance method did not allow maximum expression of hybrid vigor at the low seeding rate for a commercially planted grain hybrid. However, Burton (1948, 1989) showed that up to 50% selfed plants in forage chance hybrids would not decrease forage yields at recommended seeding rates, which are higher than those for grain hybrids. Anand Kumar and Andrews (1984) found that research in the 1950s demonstrated the large yield increases possible with F, hybrids and that a crns system was needed to produce hybrids on a commercial scale. Tift 23A, a crns inbred, was made available to Indian pearl millet breeders in 1962 (Burton, 1965).Indian pearl millet breeders pollinated Tift 23A with Bil-3B, an Indian inbred, to produce HB 1, the first released pearl millet single-cross grain hybrid using the crns system. Hybrids using Tift 23A and Tift 18A as female parents and Indian inbreds as pollinators averaged 102% more grain production than the best available varietal checks in India from 1964 to 1967 (Rachie and Majmudar, 1980). Hybrids such as HB 1, using Tift 23A as the seed parent, eventually became susceptible to downy mildew (Sclerospora graminicola Sacc. Schroet.) and ergot (Clavicepsfusiformis Loveless). This initiated a concentrated effort to develop inbreds resistant to these diseases for production of resistant hybrids. The research is ongoing today. Scientists at ICRISAT (International Crops Research Institute for the Semi-And Tropics), India, have been exploring new sources of cytoplasmic male sterility for hybrid production (Sujata er al., 1994; Rai, 1995). The first release in India of a top-cross hybrid was announced in 1996 by government authorities in Madhya Pradesh. The hybrid named “Jawahar Bajra Hybrid 1 (JBHl)” has high grain-yield potential, medium height, nonbristled compact ears, and medium bold, globular grains. Both the hybrid and its top-cross pollinator are highly resistant to downy mildew. Similarly, Gujarat State Fertilizers Company Limited has developed a hybrid “Sardar Hybrid Bajra (SHB I),” which has about 20% more yield, has better quality grain, and matures earlier than the existing hybrids (SATNews, 19961997). Interest in producing pearl millet for grain in the United States and Australia has increased. HGM 100 was the first commercial grain hybrid released in the United States in the early 1990s (Hanna el al., 1993). The area planted to the crop was increasing in the southeastern United States until a new race of rust attacked the crop in late plantings. Pearl millet’s high-quality grain, drought resistance, and flexibility in rotation and multiple cropping systems have caused interest in it as a grain crop outside its traditional growing areas.

B. FORAGE HYBRIDS Gahi 1, the first commercial pearl millet forage hybrid-produced by harvesting all the seed from a field planted to a mixture of four inbreds that flowered at the same

CYTOGENETICS AND GENETICS OF PEARL MILLET

1s

time and gave high-yielding hybrids in all combinations-yielded 52% more than Common and 35% more than Stan: Gahi 3 replaced Gahi 1 and was the first singlecross forage hybrid produced using crns (Burton, 1983).Subsequentsingle-crosshybrids, such as Tifleaf 1, and Tifleaf 2, and a three-way hybrid, Tifleaf 3, have increased animal gains because of improved forage yields, leafiness, quality, andor disease resistance (Burton, 1983; Hanna et al., 1988; Hanna et al., 1997).

C. GERMPLASM Over 20,000 accessions of cultivated pearl millet and its wild relatives are stored in India and the United States. These accessions include landraces, improved populations and breeding lines, and wild relatives from the primary, secondary, and tertiary gene pools that are available to plant breeders. Most germplasm is in the primary gene pool. Objectives need to be clearly defined to effectively select and use the best germplasm. Principal component and cluster analyses can be used to help identify the genetic and phenotypic diversity needed in a breeding and improvement program (Wilson et al., 1991). Weedy relatives in the primary gene pool (Hanna et al., 1988; Hanna, 1989) and wild relatives in the secondary (Hanna, 1990) and tertiary gene pools (Hanna et al., 1993) are also potential sources of valuable genes (Hanna, 1987).

D. TYPES OF HYBRIDS Hybrids usually out-yield open-pollinated cultivars (Andrews, 1987; Burton, 1983). However, since all cross combinations may not always produce superior hybrids, inbreds with good general combining ability (GCA) and/or specific combining ability (SCA) need to be identified (Anand Kumar et al., 1992). Hybrids maximize yields and can be most easily made using crns in the seed parent (Anand Kumar and Andrews, 1984),especially if pollen-fertility restorer genes are present in the pollinator of hybrids grown for grain. Lack of complete male fertility restoration can result in poor grain yields and a higher incidence of smut and ergot diseases. Restorer genes are not needed (and probably undesirable) in pollinators of forage hybrids. Most pearl millet hybrids are single crosses. A single cross between two elite inbreds with high SCA is probably the best way to maximize yield. In addition to using crns in one inbred to produce single-cross F, hybrids, single-cross hybrids can also be made between two elite male fertile inbreds by taking advantage of naturally occumng protogyny in pearl millet. Protogyny can be used to make hybrids in at least two ways: (1) equal quantities of seed of two or more inbreds, equal in height and maturity, can be mixed, planted, and allowed to interpollinate; and (2) elite male fertile inbreds can be planted in adjacent rows and seed harvested from

16

PREM P. JAUHAR AND WAYNE W. HANNA

only one inbred.The inbred from which seed is harvested should flower 3 or 4 days earlier than the other inbred used to produce the hybrid. The use of protogyny to produce hybrids will result in some selfed and sibbed seed. The effects of selfed and sibbed seed can be overcome to some extent in the hybrid production field by increasing the seeding rate to crowd out the weaker plants. Seed from selfing and sibbing in grain hybrids may be more objectionable, especially when the hybrid grain is mechanically harvested. Seed yields can be increased in the hybrid seed production fields by producing three-way hybrids. Two inbreds are used to produce a cms F, hybrid, which is used as the seed parent and pollinated by a third inbred in hybrid production fields. The commercial forage hybrid Tifleaf 3 is produced by pollinating crns F, Tift 8593 (Hanna, 1997) with inbred Tift 383 (Hanna et al., 1997). Twice as much hybrid seed is produced on Tift 8593 as on crns inbred Tift 85D,A,, the seed parent of Tifleaf 2. Forage yields of Tifleaf 2 and Tifleaf 3 are similar. Inbred (crns or male fertile) X landrace hybrids may not maximize hybrid vigor but should increase yields and provide more genetic diversity in a hybrid population. These hybrids would maintain some of the agronomic characteristics of landraces preferred by farmers and provide more genetic diversity for diverse environmental growing conditions. Mean grain yields of crns inbred X open-pollinated variety crosses have been equal to or superior to the open-pollinated variety (Mahalskshmi et al., 1992). Landrace X landrace crosses seem to have the most potential for improving yield and reliability in harsh, variable climates. Ouendeba er al. (1993) showed that the better-parent heterosis for hybrids among five West African landraces ranged from 25 to 81% for grain yield.

XI. APOMMIS Apomixis is a reproductive mechanism that bypasses the sexual process and allows a plant to clone itself through seed. In Pennisetum, a chromosomally unreduced egg cell develops into an embryo in an embryo sac derived from a vegetative nucellar cell. This type of apomixis is called apospory. In addition to the egg cell developing into an embryo without fertilization by a sperm, pseudogamy or fertilization of the central cell is needed for endosperm and seed development. Apospory is the only type of apomixis confirmed in Pennisetum.

A. INCIDENCE OF h o r n s m Pennisetum SPECIES Apomixis is relatively common in the polyploid species of Pennisetum, especially those in the tertiary gene pool. Apomixis has been reported in polyploids

CYTOGENETICS AND GENETICS OF PEARL MILLET

17

(triploid and higher) of both the x = 8 and x = 9 chromosome groups. Only x = 7 chromosome species have been reported in the primary and secondary gene pools, and all are sexual. Likewise, tertiary gene pool species with the x = 5 and x = 7 chromosome groups and diploids with x = 8 or x = 9 have been reported to be sexual. Jauhar (1981a) listed at least nine species that have been reported to reproduce by apomixis. Additionally, F! squamulatum, F! polystachyon, and t! macrourum have been reported to be apomictic (Dujardin and Hanna, 1984). Apomixis may have played a role in building and maintaining new genome combinations in Pennisetum. Hanna and Dujardin (1991) summarized some of their research, which showed how apomixis was used in crosses among two sexual and three apomictic species in the x = 7 and x = 9 chromosome groups from the primary, secondary, and tertiary gene pools to develop and maintain more than 20 new chromosome and/or genome combinations. These were developed from sexual X apomictic crosses, parthenogenesis of a reduced gametophyte, and fertilization of an unreduced egg. Hussey er al. (1 993) and Bashaw ef al. (1 992) showed that facultative apomictic F! fiaccidum hybridized with Cenchrus setigerus, P. massaicum, F! mezianum, and P. orientale, as n + n and/or 2n + n hybridizations, produced new genome combinations.

B. GENETICSOF APOMIXIS The genetics of apomixis is difficult to study because sexual and apomictic counterparts are usually not available within the same species. Therefore, crosses need to be made between sexual and apomictic plants from different species. Genetic studies on apomixis are made more complex by facultative apomixis, lack of F, segregatingpopulations, and the limitation of having to use the apomictic plant as pollen parent in crosses. Asker and Jerling (1992) summarized the current status of the genetics of apomixis. Most researchers agree that it is probably under relatively simple genetic control. Both dominant and recessive gene actions have been reported. Crosses between sexual and apomictic Penniserum species indicate a major dominant gene and some modifiers (Hanna et al., 1993).

C. HARNESSING APOMIXIS FOR EXPLOITATION OF HETEROSIS Apomixis has tremendous potential for revolutionizingfood, feed, and fiber production around the world because it makes possible true-breeding hybrids through seeds. Apomixis not only would fix hybrid vigor but also could make possible commercial hybrids in seed-propagated crops lacking an effective male-sterility system for producing hybrids. The opportunities apomixis offers for developing

18

PREM P. JAUHAR AND WAYNE W. H A N N A

superior hybrids and simplifying hybrid production have been previously discussed (Hanna and Bashaw, 1987; Hanna, 1995). Probably more progress has been made in transferring the apomictic mechanism from wild I! squamulatum to cultivated pearl millet than in any other grain crop. The mechanism has been transferred to the BC, generation where high levels of apomixis have been maintained (Hanna et al., 1993; and unpublished data). However, a problem encountered has been the loss of 80-90% of the seed set postanthesis. Efforts are under way to transfer apomixis from Tripsacum dactyloides (L.) L. to maize (Savidan er al., 1993; Kindiger et al., 1996) and from Elymus rectisetus (Nees in Lehm.) to wheat (Carman and Wang, 1992). The greatest impact of apomixis may be realized by cloning and inserting the gene(s) controlling apomictic reproduction into various sexual species by molecular methods. To be useful, a transferred gene must express itself and be stable in an alien genome. The gene(s) controlling apomixis needs to be mapped before it can be cloned and used in other species. Molecular markers linked to apomixis are being developed in Pennisetum (Ozias-Akins et al., 1993; Lubbers et al., 1994).

XII. GENETICS OF QUALITATIVE TRAITS Numerous qualitative traits have been reported for pearl millet. Comprehensive reviews on the genetics of qualitative traits in pearl millet have listed at least 145 mutants (Koduru and Krishna Rao, 1983; Anand Kumar and Andrews, 1993). These consisted of chlorophyll deficiencies (26%), plant pigmentation (1 8%), earhead characters (14%), pubescence and plant form (each 7%), seed characters and reproductive behavior (each 6%), foliage striping and sterility (each 4%), leaf characters and disease resistance (each 3%), and earliness (1%) (Anand Kumar and Andrews, 1993). Other mutants have been described and not included in the preceding reviews. Some of these include a naked flower mutant (Desai, 1959) and a “spreading” mutant (Goyal, 1962). Most mutants are controlled by one or two loci and dominant or recessive gene action. Recently described qualitative characters include phylloid (Wilson, 1996), narrow leaf (Appa Rao et al., 1995), brown midrib (Gupta, 1995), and xantha terminalis (Appa Rao et al., 1992) mutants controlled by the phm phm, In In, bm, bm,, and xt xt genes, respectively. Hanna and Burton (1992) showed that two plant-color mutants, red (Rp,)and purple (Rp,), are allelic; and RpI is dominant over Rp2 and normal green, whereas Rp, is dominant over normal green. Uma Devi et al. ( 1996) observed linkage of semidwarf phenotype to interchange homozygosity. Most of the mutants have potential for mapping and various genetic and physiological studies. Some appear to have direct application in commercial cultivars. Dwarf genes, especially the d , locus, has been widely used to produce high qual-

CYTOGENETICSAND GENETICS OF PEARL MILLET

19

ity shorter forage hybrids and dwarf grain hybrids that can be mechanically harvested. The early genes have been effectively used to produce early grain hybrids. Forage quality could be rapidly increased with the brown midrib bm,gene, which can reduce lignin by 20% in the plant (Cherney et al., 1988). The trichomeless or tr locus could potentially have an effect on improving drought resistance, disease and insect resistance, and palatability. Loci controlling disease resistance are being used in both commercial grain and forage hybrids. Linkage relationships have been established for only a few of these mutants (Minocha et al., 1980b; Hanna and Burton, 1992, and summarized by Koduru et al., 1983; and Anand Kumar and Andrews, 1993). Minocha et al. (1980a) used trisomics to map genes to chromosomes 1,2,4,5,and 6 . Liu er al. (1994) placed 181 RFLP markers on a molecular map. The length of the linkage map for seven linkage groups was 303 cM, with an average map distance of 2 cM between loci.

Xm. GENETICS OF QUANTITATIVE TRAITS Burton (195 1, 1959) conducted some of the first quantitative genetic studies on various plant characters and yields of pearl millet. Virk (1988) published a comprehensive review on quantitative studies conducted on pearl millet. Both additive and nonadditive genetic variances are important in pearl millet. However, the nonadditive component tends to be more important, indicating the opportunity to successfully take advantage of hybrid vigor for both grain and forage production. This, in fact, has been the case in pearl millet (see Section X). Efforts have been made to identify qualitative characters linked to quantitative characters affecting forage yield. Burton et al. (1980) showed that three recessive mutants, T13 orange node, T18 early, and T23 stubby head, increased forage yields 34, 38, and 22%, respectively, when heterozygous in an F, hybrid. In another study involving crosses between nonlethal genetic markers and exotic pearl millet lines, the Rp, gene was associated with 1861% heterotic chromosome block heterosis (HCB), and the tr was associated with 1 7 4 % HCB heterosis (Burton and Werner, 1991). A similar approach used to identify HCBs in Burkina Faso landraces identified up to 5 1% HCB heterosis associated with the R p , locus in certain crosses (Burton and Wilson, 1995).

XIV. CONCLUSION AND PERSPECTIVES With world population currently growing at the alarming rate of more than 2% per year, meeting the ever-expanding need for food will be difficult in the near fu-

20

PREM P. JAUHAR AND WAYNE W. H A N N A

ture. The planet’s carrying capacity is not unlimited, and environmentalconstraints are ever increasing. Moreover, the balance of demographic power has shifted to the developing world, where about 78% of human beings live. Poverty is taking its toll, and more than 1 billion people today survive on less than a dollar a day. Immediate measures must be undertaken to provide quick and reasonable relief to this large segment of society. About one-sixth of the world’s population live in the semi-arid tropics encompassing parts of Asia, Africa, and Latin America-the regions typified by limited and erratic rainfall and poor soils. Pearl millet provides sustenance to a large proportion of poor people in these regions. It has the capacity to grow in some of the poorest soils in chronically drought-prone regions. The need for genetic improvement of pearl millet cannot, therefore, be overemphasized. Although its importance as a research tool in cytogenetics and breeding has been recognized, its potential as an economic crop has not been fully realized. This poses a challenge for cytogeneticists,breeders, agronomists, and biotechnologists. Pearl millet is endowed with an efficient C, photosynthetic pathway, and it responds well to fertilizers. Although it has a remarkable ability to grow on poor, depleted soils, nitrogen deficiency is a major factor limiting grain production. Therefore, genotypes with high-nitrogen-use efficiency should be produced. Fortunately, pearl millet responds extremely well to heterosis breeding. Utilization of hybrid vigor will, therefore, be the most efficient means of increasing both grain and forage production. If the vast pearl millet growing areas in Africa and Asia could be planted to improved hybrids, grain production would increase phenomenally. Apomixis provides a unique tool for reaping the fruits of heterosis over an extended period of time. If apomixis is transferred to hybrids with desired heterozygosity and superior gene combination, it can fix and help perpetuate heterosis, thereby obviating the need to produce hybrid seed year after year. Research in this area will be very rewarding. Developing a broad genetic base of hybrids is imperative to ensuring resistance to future diseases. With the availability of cytoplasmic-genic male-sterile lines in the mid- 1960s, several excellent hybrids were produced in India. Particularly promising among these was HB 3, which, because of its high yields, became widely accepted throughout India in the early 1970s. Soon afterward, however, the hybrid became vulnerable to downy mildew caused by the fungus Sclemsporu gruminicolu. The disease devastated the relatively genetically uniform hybrid crop. An effective solution to such an eventuality is to produce genetically broadbased male-sterile lines using disease-resistantgenetic resources. Recently, several male-sterile lines have been developed at ICRISAT, and thnx of these (ICMA 91113, ICMA 91114, and ICMA 91115) provide not only reasonable yields but also resistance to ergot, smut, and even downy mildew. Pearl millet is an important source of dietary protein for a sizable portion of those living in poverty in Africa and Asia. Therefore, the nutritional quality of the

CYTOGENETICS AND GENETICS OF PEARL MILLET

21

grain, particularly its protein content and amino acid balance, needs to be improved. With genetic enrichment of the quantity and quality of its proteins, pearl millet will be a more nutritional food source. Cytogenetic manipulations have no doubt been instrumental in producing superior cultivars of pearl millet. An exciting recent development is the availability of tools of modern biotechnology for crop improvement.The development and use of molecular markers-random amplified polymorphic DNA (RAPDs) and restriction fragment length polymorphism (RFLPs)-are beginning to revolutionize molecular mapping. For example, until recently, our knowledge of the inheritance of downy mildew resistance was limited. Resistance was generally believed to be monogenic dominant. However, molecular mapping has demonstrated that many genes contribute to downy mildew resistance and that these genes are scattered throughout the host genome. The use of DNA markers could help identify desired genotypes more precisely and hence assist in adopting appropriate breeding strategy for pearl millet. Pearl millet provides unlimited opportunities for both basic and applied research. With further cytogenetic manipulation and marker-assisted selection, combined with the exploitation of recent advances in biotechnological research, pearl millet may emerge as a leading economic crop that plays an ever-increasing role in the welfare of those living in poverty, particularly in the semi-arid tropics of the world.

REFERENCES Amoukou, A. I., and Marchais. L. (1993). Evidence of partial reproductive barrier between wild and cultivated pearl millets (Penniseturn gluucurn). Euphyrica 67, 19-26. Anand Kumar, K., and Andrews, D. J. (1984). Cytoplasmic male sterility in pearl millet [Penniserurn americunum (L.) Leekel-A review. Adv. Appl. Eiol. 10,113-143. Anand Kumar, K., and Andrews, D. J. (1993). Genetics of qualitative traits in pearl millet: A review. Crop Sci. 33, 1-20. Andrews, D. J. (1987). Breeding pearl millet grain hybrids. In “Hybrid Seed Production of Selected Cereal Oil and Vegetable Crops” (W. A. Feistzer and A. F. Kelly, eds.), pp. 83-109. FA0 Plant Production and Protection, Paper 82, Rome. Andrews, D. J., King, S. B., Witcomb, J. R., Singh, S. D., Rai, K. N., Thakur, R. P., Talukdar, B. S., Chavan, S. B., and Singh, P. (1985). Breeding for disease resistance and yield in pearl millet. Field Crops Res. 11,241-258. Appa Rao, S., Mengesha, M. H., and Rajagopal Reddy. C. (1992). Characteristics and inheritance of xantha terminalis in pearl millet. J. Hered. 83,6243. Appa Rao, S., Rai, K. N., Mengesha, M. H., and Rajagopal Reddy, C. (1995). Narrow leaf mutant: A new plant type in pearl millet. J . Hered. 86,299-301. Asker, S. E., and Jerling, L. (1992). “Apoxirnis in Plants.” CRC Press, Boca Raton, FL. Bashaw, E. C.. Hussey, M. A,, and Hignight, K. W. (1992). Hybridization (n + n and 2n + n) of facultative apomictic species in the Penniseturn agamic complex. fnf.J. Planr Sci. 15,466470. Bennett, M. D. (1976). DNA amount, latitude, and crop plant distribution. Environ. Exp. Bor. 16, 93- 108.

22

PREM P. JAUHAR AND WAYNE W. H A N N A

Bor, N. L. (1960). “Grasses of Burma, Ceylon, India, and Pakistan (excluding Bambuseae).” Pergamon Press, London. Brunken, J. N., de Wet, J. M. J.. and Harlan, J. R. (1977). The morphology and domestication of pearl millet. Econ. Bot. 31, 163-174. Burton, G. W. (1942). A cytological study of some species in tribe Paniceae. Am. J. Bot. 29,355-359. Burton, G. W. (1948). The performance of various mixtures of hybrid and parent inbred pearl millet, Pennisetum glaucum (L.) R. Br. J . Am. SOC.Agron. 40,908-915. Burton, G. W. (1951). Quantitative inheritance in pearl millet (Pennisetum glaucum) indicated by genetic variance component studies. Agron. J. 51,47948 l . Burton, G. W. (1959). Breeding methods for pearl millet (Pennisetum glaucum). Agron. J. 43,409417. Burton, G. W. (1965). Pearl millet Tift 23A released. Crops Soils 17, 19. Burton, G. W. (1983). Breeding pearl millet. Plant Breeding Reviews 1, 162-182. Burton, G. W. (1989). Composition and forage yield of hybrid-inbred mixtures of pearl millet. Crop Sci. 29,252-255. Burton, G. W., and Powell, J. B. (1968). Pearl millet breeding and cytogenetics.Adv. Agron. 20,49-89. Burton, G. W., and Werner, B. K. (1991). Genetic markers to locate and transfer heterotic chromosome blocks for increased pearl millet yields. Crop Sci. 31,576579. Burton, G. W., and Wilson, J. P. (1995).Identification and transfer of heterotic chromosome blocks for forage yield in short-day exotic pearl millet landraces. Crop Sci. 35, 1184-1 187. Burton, G. W., Hanna, W. W., and Powell, J. B. (1980). Hybrid vigor in forage yields of crosses between pearl millet inbreds and their mutants. Crop Sci. 20,744-747. Busso, C. S., Liu, C. S., Hash, C. T., Witcombe, J. R., Devos, K. M., deWet, J. M. J., and Gale, M. D. (1995). Analysis of recombination rate in female and male gametogenesis in pearl millet (Pennisetum glaucum) using RFLP markers. Theoc Appl. Genet. 90,242-246. Carman, J. G.. and Wang, R. R-C. (1992). Apomixis in Triticeae. In “Proc. of Apomixis Workshop,” pp. 26-29. National Technical Service, Springfield, VA. Chase, A. (1921).The Linnaean concept of pearl millet. Am. J . Bor. 8 , 4 1 4 9 . Cherney, J. H., Axtell, J. D., Hassen, M. M., and Anliker, K. S. (1988).Forage quality characterization of a chemically induced brown-midrib mutant in pearl millet. Crop Sci. 28,783-787. Chittenden, L. M., Shertz, K. F., Lin, Y-R., Wing, R. A., and Paterson, A. H. (1994). RFLP mapping of a cross between Sorghum bicolor and S.propinquum, suitable for high-density mapping, suggests ancestral duplication of Sorghum chromosomes. Theor:Appl. Genet. 87,925-933. Clegg, M. T., Rawson, J. R. Y., and Thomas, K. (1984). Chloroplast DNA variation in pearl millet and related species. Genetics 106,449461, Dave, H. R. (1987). Pearl millet hybrids. Proc. Intl. Pearl Millet Workshop, pp. 121-126. Desdi, M. C. (1959). A naked flower mutant in pearl millet. Sci. Culture 25,207-208. De Vincente, M. C., and Tanksley, S. D. (1991). Genome-wide reduction in recombination of backcross progeny derived from male versus female gametes in an interspecific cross of tomato. Theoc Appl. Genet, 83, 173-178. Dujardin, M., and Hanna, W. W. (1984).Microsporogenesis, reproductive behavior, and fertility in five Penniserum species. Theoc Appl. Genet. 67,197-201. Gepts, P., and Clegg, M. T. (1989). Genetic diversity in pearl millet (Pennisetumglaucum [L.] R. Br.) at the DNA sequence level. J. Hered. SO, 203-208. Gerlach, W. L., and Bedbrook, J. R. (1979).Cloning and characterization of ribosomal RNAgenes from wheat and barley. Nucleic Acids Res. 7, 1869-1885. Goyal, R. D. (1962).A “spreading” mutant in Bajra (Pennisetum typhoides S Kc H). Sci. Culture 28, 437438. Gupta, S. C. (1995). Inheritance and allelic study of brown midrib trait in pearl millet. J. Hered. 86, 301-303.

CYTOGENETICS AND GENETICS OF PEARL MILLET

23

Hanna, W. W. (1987). Utilization of wild relatives of pearl millet. Proc. Inrl. Pearl Millet Workshop, pp. 33-42. Hanna, W. W. (1989).Characteristics and stability of a new cytoplasmic nuclear male-sterile source in pearl millet. Crop Sci. 29, 1457-1459. Hanna, W. W. (1990). Transfer of germplasm from the secondary to the primary gene pool in Pennisetum. Theor:Appl. Genet. 80,20&204. Hanna, W. W. (1993).Registration of pearl millet parental lines Tift 8677 and A,/B I Tift 90D,El. Crop Sci. 33, 1119. Hanna, W. W. (1995).Use of apomixis in cultivar development. Adv. Agron. 54,333-350. Hanna, W. W. (1997). Registration of Tift 8593 pearl millet genetic stock. Crop Sci. 37, 1412. Hanna, W. W., and Bashaw, E. C. (1987).Apomixis: Its identification and use in plant breeding. Crop Sci. 27, 1136-1 139. Hanna, W. W., and Burton, G. W. (1992).Genetics of red and purple plant color in pearl millet. J. Hered. 83,386-388. Hanna, W. W., and Dujardin, M. (199 I). Role of apomixis in building and maintaining genome comKimber, ed.), pp. binations. In “Proc. 2nd Intl. symp. on Chromosome Engineering in Plants” (G. 112-1 17. University of Missouri, Columbia, MO. Hanna, W. W., and Monson, W. G. (1980).Yield, quality, and breeding behavior of pearl millet X Napier grass interspecific hybrids. Agron. J. 72,358-360. Hanna, W. W., Wells, H. D., Burton, G. W., Hill, G. M., and Monson, W. G. (1988). Registration of Tifleaf 2 pearl millet. Crop Sci. 28, 1023. Hanna, W. W., Dujardin, M.. and Monson, W. G. (1989).Using diverse species to improve quality and yield in the Pennisetum genus. Proc. Intl. Grassl. Congr: 16,403404. Hanna, W. W., Dujardin, M., Ozias-Akins, P., and Arthur, L. (1992). Transfer of apomixis in Pennisefum. In “Proc. of Apomixis Workshop,” pp. 30-33. National Technical Service, Springfield, VA. Hanna, W., Dujardin. M., Ozias-Akins, P., Lubbers, E., and Arthur, L. (1993).Reproduction, cytology, and fertility of pearl millet X Penniserurn squamularum BC, plants. J. Hered. 84,213-216. Hanna, W. W., Hill, G. M., Gates, R. N., Wilson, J. P., and Burton, G. W. (1997). Registration of ‘Tifleaf 3’ pearl millet. Crop Sci. 37, 1388. Harlan, J. R. (1971).Agricultural origins: Centers and noncenters. Science 174,468474. Helentjaris, T., Slocum, M., Wright, S. Shaefer, A,. and Neinhuis, J. (1986). Construction of genetic linkage maps in maize and tomato using restriction fragment length polymorphisms. Theor: Appl. Genet. 72,76 1-769. Hitchcock, A. S., and Chase, A. (1951). “Manual of the Grasses of the United States,” 2nd ed. U.S. Dept. Agric. Misc. Publ. 200, Washington, DC. Hulbert. S. H., Richter, T. E., Axtell, J. D., and Bennetzen, J. L. (1990).Genetic mapping and characterization of sorghum and related crops by means of maize DNA probes. Proc. Nut Acad. Sci. USA 87,425 14255. Hussey, M. H., Bashaw, E. C., Hignight, K. W., Wipff, J., and Hatch, S. L. (1993).Fertilization of unreduced female gametes: A technique for genetic enhancement within the Cenchrus-Pennisetum agamic complex. Proc. Intl. Grussl. Cong. 17,404-405. International Crops Research Institute for the Semi-Arid Tropics (ICRISAT). (1996). Improving the unimprovable: Succeeding with pearl millet. ICRISATReport, May. Jauhar, P. P. (1968). Inter- and intra-genomal chromosome pairing in an interspecific hybrid and its bearing on basic chromosome number in Pennisefum. Genefica39,360-370. Jauhar, P.P. (1970a). Haploid meiosis and its bearing on phylogeny of pearl millet, Pennisefum typhoides Stapf et Hubb. Geneticcr 41,532-540. Jauhar. P. P. (1970b).Chromosome behaviour and fertility of the raw and evolved synthetic tetraploids of pearl millet, Pennisetum typhoides Stapf et Hubb. Geneticu 41,407424.

24

PREM I? JAUHAR AND WAYNE W. H A N N A

Jauhar, P. P, (1973). Inter- and intra-genomal chromosome relationships in a Pennisetum hybrid. In “Proc. Intl. Cong. of Genetics.” Genetics 74, 126-127. Jauhar, P. P. (1981a). “Cytogenetics and Breeding of Pearl Millet and Related Species.” Alan R. Liss, New York. Jauhar, P. P. (1981b). Cytogenetics of pearl millet. Adv. Agmn. 34,407479. Jauhar, P. P. (1981~).The eternal controversy on the Latin name of pearl millet. Indian J. Bor. 4( I), 1-4. Jauhar, P. P..and L. R. Joppa. (1996). Chromosome pairing as a tool in genome analysis: Merits and limitations. In “Methods of Genome Analysis in Plants” (P. P. Jauhar, ed.), pp. 9-37. CRC Press, Boca Raton, FL. Jones, E. S., Lice, C. J., Gale, M. D., Hash, C. T., and Witcombe, J. K. (1995). Mapping quantitative trait loci for downy mildew resistance in pearl millet. Theor: Appl. Genet. 91,448-456. Kianian, S. F., and Quiros, C. F.(1992). Generation of Brassicu oleracea composite RFLP map: Linkage arrangementsamong various populations and evolutionary implications. Theor: Appl. Genet. 84,54&554. Kindiger. B., Sokolov, V., and Khatypova, I. V. (1996). Evaluation of apomictic reproduction in a set of 39 chromosome maize-Tripsacum backcross hybrids. Crop Sci. 36,1108-1 113, Kishikawa, H. (1970). Effects of temperature and soil moisture on frequency of accessory chromosomes in rye, Secale cereale. Jup. J. Breed. 20,269-214. Kodum, P. R. K., and Krishna Rao, M. (1983). Genetics of qualitative traits and linkage studies in pearl 90, 1-22. millet-A review. Z. P’nzenziichtz. Kumar, P., Kapoor, R. L., Dass, S., and Chandra, S. (1992). Genetics of days to heading and maturity in pearl millet. Huryana Agric. Univ. J. Res. 12,282-286. Laurie, D. A. (1989). The frequency of fertilization in wheat X pearl millet crosses. Genome 32, 1063-1 067. Lee, W.J. (1966). On accessory chromosomes in Secale cereale. III. Relationship between the frequency of accessory chromosomesin rye and soil properties. Korean. J. Bot. 9 ( 3 4 ) , 1-6. Linnaeus, C. (1753). “Species plantarum,” 1st ed. Stockholm. Liu, C. J., Witcombe, J. R., Pittaway, T.S., Nash, M., Hash, C. T., Busso, C. S., and Gale, M. D. (1994). An REP-based genetic map of pearl millet (Pennisetum glaucum). Theor: Appl. Genet. 89, 481487. and Ozias-Akins, P. (1994). Molecular markers shared by diLubbers, E. L., Arthur, L., Hanna, W. W., verse apomictic Pennisetum species. Theor: Appl. Genet. 89,636-642. Mahalakshmi, V., Bidinger, F.R., Rao, K. P., and Raju, D. S. (1992). Performance and stability of pearl millet topcross hybrids and their variety pollinators. Crop Sci. 32,928-932. Martel, E., Ricroch, A., and Sarr, A. (1996).Assessment of genome organization among diploid species (2n = 2x = 14) belonging to primary and tertiary gene pools of pearl millet using fluorescent in situ hybridization with rDNA probes. Genome 39,680687. Matzk, F.(1996). Hybrids of crosses between oat and Andropogoneae or Paniceae species. Crop Sci. 36, 17-2 1. Meredith, D. (ed.) (1955). “The Grasses and Pastures of South Africa.” Cape Times Ltd., Parrow CP, South Africa. Minocha, J. L., Gill, B. S., Sharma, H. L., and Sidhu, J. L. (1980a). Cytogenetic studies on primary trisomics in pearl millet. In “Trends in Genetical Research in Pennisetums.” (V. P. Gupta and I. L. Minocha, eds.), pp. 129-132. Punjab Agric. Univ., Ludhiana, India. Minocha. J. L.. Gill, B. S., and Sidhu, J. S. (1980b). Inheritance and linkage studies in pearl millet. In “Trends in Genetical Research in Pennisetums” (V. P. Gupta and J. L. Minocha, eds.), pp. 99-1 10. Punjab Agric. Univ.. Ludhiana, India. Minocha, J. L., Bra, D. S., Saini. P. S., Multani, D. S., and Sidhu. S. S. (1982). A translocation tester set in pearl millet. Theor: Appl. Genet. 62,3 1-33.

CYTOGENETICS AND GENETICS OF PEARL MILLET

25

Muldoon, D. K., and Pearson, C. J. (1979). The hybrids between Pennisetum americanum and Pennisetum purpureum. Herbage Absr. 49, 189-199. Miintzing, A. (1958).Accessory chromosomes. Trans. Bose Res. Insr. (Calcutta) 22, 1-15. Osgood, R. V., Hanna, W. W., and Tew, T. L. (1997).Hybrid seed production of pearl millet X Napier grass triploid hybrids. Crop Sci. 37,998-999. Ouendeba, B.. Ejeta, G., Nyquist. W. E., Hanna. W. W., and Kumar, A. (1993).Heterosis and combining ability among African pearl millet landraces. Crop Sci. 33,735-739. Ozias-Akins, P., Lubbers, E. L., Hanna, W. W., and McNay, J. W. (1993).Transmission of the apomictic mode of reproduction in Penniserum: Co-inheritance of the trait and molecular markers. Theo,: Appl. Genet. 85,632438. Paterson, A. H., Tanksley, S. D., and Sorrells. M. E. (199 I). DNA markers in plant improvement. Adv. Agron. 46,39-90. Patil, B. D., and Singh, A. (1964).An interspecific cross in the genus Pennisetum involving two basic numbers. Cur,: Sci. 33, 161-162. Patil, B. D., and Singh, A. (1980).Genetic and cytogenetic improvement of Penniserum and its allied species for grassland and pasture production. In “Trends in Genetical Research in Pennisetums” (V. P. Gupta and J. L. Minocha, eds.), pp. 112-122. Punjab Agric. Univ., Ludhiana, India. Pantulu, J. V. (1960).Accessory chromosomes in Penniserum ryphoides. Cur,: Sci. 29,28-29. Powell, J. B., and Burton, G. W. (1966). Nucleolus-organizing accessory chromosomes in pearl millet, Pennisetum ryphoides. Crop Sci. 6, 131-134. and Majmudar, J. V. (1980). “Pearl Millet.” Pennsylvania State Univ. Press, University Rachie, K. 0.. Park. Rai, K. N. (1995).A new cytoplasmic-nuclear male sterility system in pearl millet. Plunr Breed. 114,

445447. Rao, P. K., Nambiar, A. K., and Menon, P. M. (1951).Maximisation of production by cultivation of hybrid strains with special reference to cumbu (pearl millet). Madras Agric. J. 38,95-100. Rao, P. N., Nirmala, A,. and Ranganadham, P. (1988).Trisomic categories in pearl millet (Pennisetum umericunum (L.) Leeke). Theo,: Appl. Genet. 75,340-343. Rau, N. S. (1929). On the chromosome numbers of some cultivated plants of south India. J. Indian. Bot. Sci. 8, 1261 28. Reader, S . M., Miller, T. E., and Purdie, K. A. (1996).Cytological analysis of plant chromosomes using rapid in situ hybridization. Euphytica 89, 121-124. SAT (Semi-Arid Tropics) News. (19964997). ICRISAT. Patancheru, India, pp. 12-14. Savidan, Y.,LeBlanc, 0..and Berthaud, J. (1993).Progress in the transfer of apomixis in maize. Agron. Absr. 1993, 101. Schank, S., and Hanna, W. (1995).Usage of Pennisetum in Florida and the tropics. Proc. Intl. Conj: on Livestock in the Tropics. pp. 1-13. Schumann, K. (1 895). 11. Die Graser Ostafrikas und ihre Verwerthung. (The grasses of East Africa and their utilization.) In “Die Pflanzenwelt Ost-Afrikas und der Nachbargebiete,” R.B (A. Engler, ed.). pp. 51-59. Sidhu, S. S., and Minocha, S. L. (1984).Peroxidase isozyme patterns in primary trisomics of pearl millet. Theo,: Appl. Genet. 68, 179-182. Slocum, M. K., Figdore, S. S.. Kennard, W. C., Suzuki, J. Y., and Osborn, T. C. (1990).Linkage arrangement of restriction fragment length polymorphism loci in Brassica oleracea. Theo,: Appl. Genet. so,57-64. Stapf, O., and Hubbard, C. E. (1933).Notes on African grasses: XIII. Kew Bull. 1933,269-302. Stapf, 0.. and Hubbard, C. E. (1934).Pennisetum. In “Flora of Tropical Africa” (D. Prain, ed.), Vol. 9, Pt. 6, pp. 954-1070. Reeve & Co, Ltd. Ashford, Kent, England. Sujata, V., Sivaramakrishnan, S., Rai, K. N., and Seetha, K. (1994).Anew source of cytoplasmic male sterility in pearl millet: RFLP analysis of mitochondria1 DNA. Genome 37,482486.

26

PREM P. JAUHAR AND WAYNE W. H A N N A

Terrell, E. E. (1976). The correct names of pearl millet and yellow fox tail. Taxon. 25,297-304. Tostain, S., Riandey, M. F., and Marchais, L. (1987). Enzyme diversity in pearl millet (Penniserurn gluucurn). 1. West Africa. Theor Appl. Genet. 74, 188-193. Uma Devi, K., Narasinga Rao, P.S.R.L., and Krishna Rao, M. (1996). Linkage of semidwarf phenotype to interchange homozygosity in pearl millet. J. Hered. 87, 170-172. Vari, A. K., and Bhowal, J. G. (1986). Studies on the trisomics of Pennisetum urnericunum (L.) Leeke.-Morphological and cytological behaviour of primary trisomics. Cyrologiu 51,679492. Vavilov, N. I. (1949-1950). The origin, variation, immunity, and breeding of cultivated plants. Chronicu Borunicu 13, 1-366. Veyret, Y.(1957). Les chromosomes somatiques chez quelques esptces de Penniserum. (Somatic chromosomes in some species of Penniserurn.) Agron. Trop. 12,595-598. Virk, D. S. (1988). Biometrical analysis in pearl millet-A review. Crop Imprv. 15, 1-29. Whitkus, R., Doebley, J., and Lee, M. (1992). Comparative genome mapping of sorghum and maize. Genetics 132, 11 19-1 130. Williams, R. J., and Andrews, D. J. (1983). Breeding for disease and pest resistance in pearl millet. FA0 Plant Pro?. Bull. 31, 136-158. Wilson, J. P. (1996). A recessive homeotic mutant in pearl millet. J. Hered. 87,66-67. Wilson, J. P., Burton, G. W., Zongo, J. D., and Dicko, I. 0. (1991). Disease resistance and morphological traits of pearl millet landraces from south Burkina Faso. Crop Sci. 31,641-645. Zenkteler, M., and Nitzsche, W. (1984). Wide hybridization experiments in cereals. Theor:Appl. Gener. 68,3 11-3 15.

ADVANCESIN ICP EMISSION AND ICP M i s s SPECTROMETRY Parviz N. Soltanpour,' Greg W. Johnson,2 Stephen M. W ~ r k m a n , ~ J. Benton Jones, Jr.: and Robert 0. Miller' 'Department of Soil and Crop Sciences Colorado State University Fort Collins, Colorado 80523 2Matheson Gas Products Longmont, Colorado 80501 'Analytical Technologies, Inc. Fort Collins, Colorado 80524 '+Macro-MicroAnalytical Services Athens, Georgia 30607

I. Introduction 11. ICP-AES and ICP-MS Instrumentation A. ICP Generation B. Properties of ICP C. Sample Introduction Systems 111. Spectrometers A. ICP-Atomic Emission Spectrometry B. ICP-Mass Spectrometry n! Analytical Capabilities A. Selection of Wavelength B. Selection of Isotope C. ICP-AES Detection Limits D. ICP-MS Detection Limits v. ICP-AES Interferences A. Solute Vaporization B. Ionization C. Unwanted Radiation D. Correction for Interferences (ICP-ms) VI. ICP-MS Interferences A. Solids Deposition on Sampler and Skimmer Cones 27 Advrmrm in Agrorm~y,Volume 64 Copynght 0 1998 by Academic Press. All rights of reproduction in any form reserved. 0065-~11~/9n $ZS.OO

28

PAFWIZ N. SOLTANPOUR ETAL. B. NonspectroscopicInterferences C. Mass Discrimination D. Unwanted Ions E. Methods of Correction for Interferences(ICP-MS) VII. Practical Applications A. Grinding Soil Samples B. Obtaining Soil Extracts C. Digestion of Organic Matter and Dissolution of Silicates for Total Elemental Analysis D. Analysis of Soil Extracts and Digests E. Determination of Trace Levels of As, Se, and Hg Using the HydrideMercury Vapor Generator VIII. Quality Control Methods IX.Summary Appendix References

I. INTRODUCTION The application of inductively coupled plasma-atomic emission spectrometry (ICP-AES) to the analysis of soil was reviewed in 1982 and again in 1996 with inclusion of ICP-mass spectrometry (ICP-MS) (Soltanpour et al., 1982, 1996). In this review we treat ICP-MS more comprehensively and include a table for isotopes of elements (see Section 1V.B.) and an example for Ca, Fe, Ni, Zn, and Pb isotope selection for plant-tissue analysis (Appendix 1). New developments in ICP-AES include suspension-nebulization analysis of clays (Laird et al., 1991); interfacing ICP spectrometers with flow-injection analyzers for automatic dilution, calibration, separation, concentration, standard additions, and other operations (Greenfield, 1983; LaFerniere et al., 1985); interfacing ICP-AES with liquid chromatographs for concentration and speciation of elements (Roychowdhury and Koropchack, 1990); using high-salt nebulizers to prevent clogging of nebulizers (Legere and Burgener, 1985): successfully using concentration and reduction of spectral interference techniques such as chelation-solvent extraction (Huang and Wai, 1986; Bradford and Bakhtar, 1991); using computer programs such as orthogonal polynomials (Hassan and Loux, 1990), simplex optimization (Belchamber et al., 1986), and that recommended by Taylor and Schutyser (1986) to optimize spectrometer operating conditions and automatic correction for spectral interferences; and compiling ICP emission lines still in progress (McLaren and Berman, 1985; Boumans, 1984; Parsons et al., 1980). The ICP-MS method of analysis has been developed over the last 15 years. Houk et al. (198 1) showed suprathermal ionization in an ICP Ar plasma. Within the last 10 years the method has been applied to routine analytical concentration determinations. Several review articles document the ICP-MS developmental milestones (Beauchemin, 1989; Hieftje and Vickers, 1989; Douglas, 1989; Houk

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 29 and Thompson, 1988; Houk, 1986; Gray, 1985; Douglas and Houk, 1985). Between 1986 and 1988, ICP-MS enjoyed a surge of popularity. According to Cresser et al. (1988), the late A. R. Date attributed the success of ICP-MS to spectral simplicity, very high sensitivity, and isotope ratio capability; Date considered ICPMS “the greatest thing to happen to atomic spectroscopy since chopped light” (Date, 1986). Each year since 1986, papers published in the environmental area of atomic analysis, including ICP-AES and ICP-MS, have been reviewed by Malcolm S. Cresser and co-workers (Cresser et al., 1986; Ebdon et al., 1987; Cresser et al., 1988, 1989, 1990, 1991, 1992). Soil and biological material analysis is included in their scope. Another source of current literature on using ICP-MS to analyze geological and inorganic materials is the biennial review publication that appears in Analytical Chemistry (Jackson et al., 1989, 1991). Each January the ICP Information Newsletter publishes an annual bibliography of the ICP field (Barns, 1992) and, like the Cresser review, abstracts papers on ICP-MS presented at national and international conferences. A review concerned with inorganic mass spectrometry and X-ray fluorescence spectrometry with a section emphasizing developments in the ICP-MS field has been published yearly since 1988 (Ure er al., 1988; Bacon et al., 1989, 1990, 1991). Date and Gray (1989) edited a volume on applications of ICP-MS, and Holland and Eaton (1991) edited a volume containing 21 selected papers from the 2nd International Conference on Plasma Source Mass Spectrometry held at Durham University in September 1990. Isotopes of 71 naturally occurring elements can be monitored using conventional positive ion, solution nebulization ICP-MS. Accuracies of the concentrations estimated using these measurements at the Division of Agriculture and Natural Resources (DANR) Analytical Lab at the University of California, Davis, corrected for internal standard, are typically within 2.5% of the true concentrations in favorable cases. For about 70% of these elements, more than one stable isotope occurs in nature. Thus, they can be analyzed using isotope ratios and/or isotope dilution. Isotope ratios show precision of 0 . 1 4 3 % (Gregoire, 1989). Concentrations calculated using isotope dilution (Fassett and Paulsen, 1989) are generally within 1% of their true concentrations-an accuracy and precision rate higher than ICP-MS analyses done without the use of stable isotope addition (Viczian et al., 1990; Van Heuzen et al., 1989; Garbarino and Taylor, 1987; McLaren et al., 1987; Dolan et al., 1990). Concentrations for 13 other nuclei that are not naturally occurring can also be estimated using the ICP-MS, as indicated in P. G. Brown et al. (1988, tab. 2); Igarashi et al. (1990); and Kim et al. (1989a, 1989b, 1991). While elemental coverage and detection limits under relatively ideal conditions are excellent, there are some problem areas in ICP-MS that must be investigated before deciding whether or not the ICP-MS technique will work for you (Hiefje, 1992). Although most of the following problems have been overcome or circumvented to meet analytical needs in selected instances, the statements that follow are generally valid for a generic, normal resolution (i.e., peak widths between 0.5 and 1.O dalton), normal aqueous aerosol generation ICP-MS:

30

PARVIZ N. SOLTANPOUR ETAL.

1. Although ICP-MS has been found to be satisfactory for soil and biological tissue work, ICP-MS data are approximately three times less accurate and precise than ICP-MS data. However, for concentrations determined from isotope dilution-ratio measurements, precision and accuracy are somewhat better than concentrations determined by ICP-AES (Gregoire, 1989; Dolan et al., 1990). 2. Isobaric overlaps (spectral interferences) occur with some regularity for elements between approximately 28 and 80 daltons and do occur throughout the mass range. They are a result of ( 1 ) a common unit mass shared by more than one element; (2) doubly charged ions overlapping a singly charged isotope with half the unit mass of the doubly charged species (Vaughan and Horlick, 1986); (3) elemental oxide, elemental hydride, and/or elemental hydroxide ions overlapping isotopes of other elements (Vaughan and Horlick, 1986; Munro ef al., 1986; Date et al., 1987; Gray, 1986); and (4) background spectral problems (Vaughan and Horlick, 1986; Gray, 1986; Tan and Horlick, 1986). The isobaric interferences involving oxygen can be eliminated using techniques such as electrothermal vaporization (ETV), atomization, or laser-ablation sample aerosol production (Gregoire, 1989). 3. Ion response is significantly suppressed by concomitant concentrations. The threshold concomitant values are low compared to emission suppressions noted for ICP-AES. Nonspectroscopic interferences result from excessive dissolved solids in the test solutions. For a number of reasons, the analyte ion amval rate at the detector (i.e., analyte response) is suppressed under these circumstances (Beauchemin et al., 1987; Olivares and Houk, 1986; Douglas and Ken; 1988; Gregoire, 1987a,b; Hieftje, 1992). Although at the DANR Analytical Lab the onset of suppression is usually observed in the neighborhood of 100-500 mg liter-', Gregoire indicates somewhat higher levels using the same instrument model-manufacturer (Perkin-Elmer SCIEX 250, Gregoire, 1989). 4. The ICP, generated in Ar with normal aqueous solution nebulization, may be unable to produce measurable amounts of positive ions for some analytes that could be of interest, e.g., F, C1, and/or S . However, the halogens can be determined in the negative ion mode (Hieftje et al., 1988; Chisum, 1992), whereas sulfur can be detected if the water is removed from the sample prior to nebulization. Water vapor can be removed from the sample aerosol using a cooled spray chamber (Hutton and Eaton, 1987). Water can be completely separated from the sulfur using an electrothermal atomizer (Gregoire, 1989) or partially removed using nebulizationdesolvation equipment (Veillon and Margoshes, 1968). 5. The costs of instrumentation, operation, and maintenance for ICP-MS are generally higher than those for ICP-AES, leading to a higher cost-per-analyte-concentration determination. We calculate that the cost-per-analyte-concentration determination for an off-the-shelf ICP-MS is about 2.5 times that of a state-of-theart automated sequential scanning ICP-AES instrument using the same depreciation schedule for each instrument. Gregoire ( 1989) points out, however,

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 3 1 that the cost of analysis using ICP-MS is low relative to other methods capable of producing data on individual isotopes. Similarly, the sample throughput is greater by about a factor of five for ICP-MS than for other isotope methods. 6. Finally, while multielement capability exists for the ICP-MS, true simultaneous multielement analysis does not (Hieftje, 1992). For an ICP-AES simultaneous multielement system, adding more analytes does not require longer measurement times per sample to preserve detection limits. For the ICP-MS, however, adding additional analytical isotopes requires longer analysis time per sample to avoid detection limit and/or precision degradation.

JI. ICP-AES AND ICP-MS INSTRUMENTATION A. ICP GENERATION The ICP is produced by initially passing ionized Ar gas through a quartz torch located inside a Cu coil connected to a radio frequency (RF)generator. The RF generator provides up to 3 kW forward power (in most commercial units) at a frequency of 27.1 MHz. The high-frequency currents flowing in the Cu coil generate oscillating magnetic fields whose lines of force are axially oriented inside the quartz tube and follow elliptical closed paths outside the coil, as shown schematically in Fig. 1 (Fassel, 1977; Fassel and Kniseley, 1974). Electrons and ions passing through the oscillating electromagnetic field flow at high acceleration rates in closed annular paths inside the quartz tube space. The direction and strength of the induced magnetic fields vary with time, resulting in electron acceleration on each half cycle. Collisions between accelerated electrons and ions, and ensuing unionized Ar gas, cause further ionization. The collisions cause ohmic heating and, when measured spectroscopically, give thermal temperatures ranging from 6000 to 10,000 K (Fig. 2) (Fassel, 1977). However, with the advent of the ICP-MS, it is evident that the true thermal temperature of the plasma is much lower than this. For example, the Perkin Elmer SCIEX 500 that has been in the DANR Analytical Lab for over a year has run for hours with the “6000 K ’ region of the plasma shown in Fig. 2 striking the copper interface plate with no melting or etching of the copper metal surface. In addition, several ICP-MS laboratories use copper as the sampler cone metal (Hieftje and Vickers, 1989; Houk, 1986). Copper appears to give satisfactory results in this role unless sulfuric acid is present in the test solutions and the sampler cone aperture is relatively small (i.e., -0.4 mm); in which case, rapid erosion has been observed (Munro et al., 1986). Copper metal melts at 1356 K and boils at 2840 K (Weast and Astle, 1979). The quartz torch has three concentric channels. The outer channel conducts Ar

32

PARVIZ N. SOLTANPOUR ETAL.

H Figure 1 Magnetic fields (H) and eddy currents (shaded) generated by high-frequency currents (I) flowing through coil. (Adapted from Fassel and Kniseley, 1974.)

gas at about 15-17 liters min-’ to the plasma to sustain the plasma and to isolate the quartz tube from high temperatures. The innermost channel is for introducing sample into the plasma. The middle channel conducts the auxiliary Ar gas at about 1 liter min-l and is used in ICP-AES only when starting the plasma or for organic samples; it is routinely used for all types of samples for ICP-MS (Fig. 3). The ICP has an annular, or donut, shape when it is viewed from above. The hole has a lower temperature than the donut body and offers less resistance to the sample injection. The sample is injected into the plasma by using Ar carrier gas at a rate of about 1 liter min-’ for ICP-AES work. For ICP-MS work, the aerosol flow is approximately 1.5 liter min- l .

B. PROPERTIES OF ICP The ICP generater has unique physical properties that make it an excellent source for vaporization, atomization, ionization, and excitation of elements. For ICP-AES, the aerosol droplets containing the analyte are desolvated, the analyte salts-oxides are vaporized, and the analyte is atomized at the high temperature region of the plasma in the vicinity of the Cu coil (Fig. 2). An initial radiation zone (IRZ) has been defined by Koirtyohann et al. (1980) as the zone that begins in the sample aerosol channel inside the load coil for ICP-AES (Fig. 4). The IRZ extends upward to 1-2 mm above the load coil, taking on the appearance of an amber “bullet” during nebulization of many sample types related to agriculture. This is due

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 3 3

Temperature (K) (+lo%) 6000

v

.a

A f

25 20

6200

15

5

6500 6800 8000 (Estimate)

L?

0.OoO

Sample Aerosol I

I

Figure 2 Temperatures in the plasma as measured by the spectroscopic slope method. (Adapted from Fassel, 1977.)

to emission from CaO molecules on the surface of the “bullet,” with the color changing to a deep blue or purple further downstream as emission from calcium atoms and ions dominates. The blue-purple region is termed the normal analytical zone (NAZ) and is the region in which the analyte emission is observed by the spectrometer (Fig. 4). Color photographs illustrating the appearance of the IRZ and NAZ while nebulizing an elevated concentration of Y into an ICP have recently been published for ICP-AES (Winge ef al., 1988) and more clearly define these critical regions. The NAZ is 15-20 mm above the coil, or about 14-19 mm above the tip of the IRZ, in an environment relatively low in background emission. The background consists of Ar lines and some weak band emission from OH, NO, and CN molecules present in the plasma (Ward, 1978a). By the time the decomposition products of the sample reach the NAZ, they have had a residence time of about 2 msec at spectroscopically measured temperatures ranging from about 8000 to 5000 K (Fassel, 1977). The residence time and temperature experienced by samples introduced into the plasmas are about twice as large as those in the hottest flames, e.g., N,O-C,H,. The combination of high temperature and residence time, at the sample aerosol flow rates typically used in ICP-AES, leads to complete sample vaporization and atomization in contrast to flames that require releasing agents for refractory compounds (Larson et al., 1975). Once the free compounds, atoms, and ions are formed in ICP-AES, they are in a chemically inert environment in contrast to environments with highly reactive combustion flames. Ionization interferences are generally negligible in an ICP-AES experiment. Self-absorption (a phenomenon responsible for the flattening of the standard curve at high analyte

34

PARVIZ N , SOLTANPOUR ET AL.

Aerosol ' Carrier

argon flow

Figure 3 Typical quartz torch and inductively coupled plasma configuration. Flow A is auxiliary flow used for organic samples. (Adapted from Fassel and Kniseley, 1974.)

concentrations) is practically absent, which leads to a wide linear dynamic analytical range of 3-5 decades. No sampling or skimmer cones and lense stack or quadrupole rods are used in the ICP-AES, and, therefore, contamination from ablative processes off of them, e.g., secondary ion sputtering, is absent. For ICP-MS, the vaporization and atomization begin at approximately the same location relative to the load coil as do these processes in the ICP-AES, in a relatively hot region of the plasma in the vicinity of the Cu coil (Fig. 2). However, the flow rates of sample and/or auxiliary argon are increased for ICP-MS to obtain an analytically useful population of ions (Winge et al., 199 I), while keeping the sampling cone a safe distance from the load Cu coil to prevent arcing between the cone and the load Cu coil. The IRZ extends well beyond the downstream side of the load Cu coil (Fig. 4).The water droplets produced in a conventional concentric nebulizer, although apparently extremely few in number compared to the total number of aerosol droplets produced, can survive the rigorous desolvation-atomization conditions generated by the ICP (Winge et al., 1991).Although the downstream side of the load coil-to-IRZ tip distance varies from one lab to another, it is generally between 10 and 20 mm for ICP-MS. Unlike ICP-AES, with ICP-MS this leaves much of the analyte vaporization and atomization to be done in regions be-

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 3 5 A Ion-atom emission = NAZ Touch edge

Load coil Auxilliary tube Injector tube

B

Figure 4 Spatial nomenclature for ICP. In agronomic work, the initial radiation zone (IRZ) will likely appear red to pink due to emission of CaO and CaOH molecules. The location of the normal analytical zone (NAZ) depends on whether analytical measurements are being performed by (A) AES or (B) MS.Pressures: p, = 760 torr; pa = 1.2 torr; p3 = lo-’ to torr. Cones: S, = sampling cone; S, = skimmer cone. (Adapted from Koirtyohann et al.. 1980.)

yond the hottest parts of the ICP. The sampling cone orifice defines the NAZ in the ICP-MS and is another 2-10 mm downstream from the tip of the IRZ (Fig. 4). In the DANR Analytical Lab, the IRZ extends approximately 19 mm downstream from the spectrometer side of the load coil, and the sampler cone orifice is positioned another 3 mm downstream from the IRZ tip; which results in placement of the NAZ a total of 22 mm from the nearest surface of the load coil. Most of the particle beam is sucked through the sampling cone into the intermediate vacuum region of a differentially pumped aperture approximately 2-3 mm from the tip of the bullet. The tip of a second cone, called the skimmer, is immersed in what is termed a barrel shock (Gray, 1989) that results from supersonic expansion of the plasma gas as it passes from atmospheric pressure through the sampling cone orifice into a vacuum of about 1 torr. The kinetic temperature of the gaseous particles at the tip of the skimmer cone is 2200 K (Lim et al., 1989; Winge et al., 1991). Although the position of the sampler with respect to the extended IRZ of the ICP results in a maximum rate of ions per second at the detector, it is also sampling aerosol that has undergone solute vaporization and atomization reactions outside

36

PARVIZ N. SOLTANPOUR ETAL.

the hottest regions of the ICP. This is thought to contribute to the appearance of more molecular ions in the mass spectra and higher susceptibility to nonspectroscopic matrix effects than if the aerosol flow rate and/or auxiliary argon flow rate could be slowed down enough to put the IRZ back to within 1-2 mm of the downstream side of the load coil. However, this is not possible because of the arcing that occurs between the load coil and the metallic sampling cone in instances in which the cone is placed too close spatially to the load coil. We have unsuccessfully located descriptions of ICP-MS experiments designed to reduce molecular ion formation in the mass spectrum using a sampler constructed of a samplingcone that does not conduct electricity. Among the possibilities for nonconducting materials are high-tech ceramics that can withstand prolonged exposure to the highest temperature regions of the ICP. These include AIN, Sic, A1,0,, or zirconia ceramics.I The sampler could be placed so that the NAZ is in a region closer to local thermodynamic equilibrium (LTE) with respect to maximized ion populations while the analyte solute vaporization and atomization is allowed to proceed in the hottest parts of the plasma (Fig. 4). In general, the NAZ is much closer to the tip of the IRZ in ICP-MS (2-10 mm) than the NAZ is to the tip of the IRZ in ICP-AES (14-19 mm). The closer proximity used for the ICP-MS measurements increases the concentration of ions to an analytically useful level (Winge et al., 1991). Ideally, ions should be extracted from a region that approximates local LTE. Apparently, ion temperatures are sufficient to support high ion populations this close to the IRZ tip. Undoubtedly, the requirement for high ion density at a distance well downstream from maximum gas and excitation temperatures promotes formation of metal oxide ions and nonspectroscopic concomitant suppression effects that are observed in the ICP-MS. A number of modifications mentioned later, most involving the usual sample introduction techniques, have significantly reduced these problems.

C. SAMPLEINTRODUCTIONSYSTEMS 1. Nebulizers Nebulizers are devices used for the injection of the sample into the plasmas. There are three general types of nebulizers: pneumatic nebulizers, Babington style nebulizers, and ultrasonic nebulizers (USNs) (Thompson and Walsh, 1983).Pneumatic nebulizers use the Venturi effect to draw sample solutions into the spray chamber. Babington-style nebulizers require a pump to deliver the solution to a pinhole orifice from which argon gas is emerging at high velocity. The USN also 'Coors Ceramics, 600 Ninth St., Golden, CO 80401.

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 3 7 requires a pump to deliver the solution, this time to a vibrating plate. There are two common types of pneumatic nebulizers: cross-flow and concentric. For the cross-flow, as the solution emerges from the rigid capillary tube carrying the sample solution, another tube positioned at a right angle blasts argon past it to shear off fine aerosol particles. Cross-flow nebulizers are often made of highly corrosion-resistant capillary metal tubes, e.g., Pt-Ir alloy. One capillary carries Ar at approximately 1 liter min-l and the other carries the sample solution. The orientation of the tips is fixed by the manufacturer and may include a sapphire edge at the tip of the solution tube to produce a fine, uniform mist out of approximately 10% of the solution drawn in. The cross-flow systems in the authors' laboratories have held up to the most demanding applications for 2-3 years with no sign of degradation.The concentric-flow (Meinhard-type)glass nebulizers are routinely used at the DANR Analytical Lab for both ICP-AES and ICP-MS work. These are made entirely of glass in a T-type configuration. The main barrel of the nebulizer consists of a fine glass tube tapered to capillary size. The capillary portion carries the sample solution, is approximately 1 in. long, and is surrounded by a larger diameter tube carrying Ar. The Ar enters through a tube joined in a T shape to this barrel. The Ar pressure is 241.5-345 KPa (35-50 psi) and flowing at about 0.75-1.5 liters min-'. The open ends of the Ar tube and the capillary tube meet at a taper, and a fine mist is produced as the Ar flowing concentrically around the capillary shears off small fragments of water droplets at the capillary tip. These nebulizers are very steady and produce aerosol from about 10% of the solution going through the tip. The cross-flow and concentric nebulizers clog with high salt solutions. Soltanpour et al. (1979a) treated 1M NH,HCO,-O.OOSM DTPA (diethylenetriaminepentaacetic acid) soil extracts with 0.5 N HNO, to overcome clogging. However, the Colorado State University Soil Testing Laboratory (CSUSTL) currently uses a Legere2 Teflon nebulizer (Babington type) attached to a peristaltic pump that eliminates the need for acid pretreatment. Wolcott and Butler (1979) designed a pneumatic nebulizer that could aspirate solutions containing up to 36% suspended solids. To overcome differences in surface tension, density, and viscosity, the analyst can use a peristaltic pump to introduce sample solutions into the nebulizer (Beasecker and Williams, 1978). For concentric nebulizers, care must be taken to eliminate small insoluble particles from test solutions that would otherwise clog the capillary. If a particle becomes lodged in the capillary or between the capillary and the tapered tip, great care must be exercised while removing the blockage to avoid breaking the fragile glass tubing. One method is to carefully remove the nebulizer from the Ar and sample delivery tubes and squirt acetone from the nebuliz-

2Distributed by Burtec Instrument Corporation, P.O. Box 235, Delmar, NY 12054.

38

PARVIZ N. SOLTANPOUR ET AL.

er tip into the barrel while tapping with a finger, then force Ar through the tip backwards while continuing to tap.3 This has been used successfully by one of the authors to remove a microscopic piece of glass fiber that became lodged between the tip and the inner concentric capillary. In Babington nebulizers (Suddendorf and Boyer, I978), aerosol is produced when the solution is pumped into a V groove and is ruptured by gas coming from a small hole in the groove. Glass frit nebulizers and the like (e.g., the Hildebrand nebuli~er)~ use the same principle as in the Babington except with many orifices. These nebulizers can be used for high-salt solutions. Since no constricting orifices are needed to produce aerosol, they are relatively free of clogs. For pneumatic and Babington nebulizers, larger droplets settle out in the spray chamber and drain off, leaving the finer aerosol droplets suspended in the flow stream of Ar that is transported to the plasma. In USNs, transducers are used to produce the sample aerosol. Compared with pneumatic nebulizers, USNs improve the detection limit of ICP spectrometers by one to two orders of magnitude (Olson et al., 1977). A three to four order-of-magnitude improvement in ICP-MS detection limits has been noted using the USN with a high-resolution, double-focusing ICP-MS instrument (Tsumura and Yamasaki, 1991). The USNs are operated with a sample aerosol desolvation system that follows aerosol production by the transducer. The aerosol desolvation system is a heating assembly followed by a condenser column. Thus, factors involved in improved analytical performance of the ICP-MS with use of the USN observed in Tsumura and Yamasaki (1991) are ( 1 ) improved sample transport to the plasma, (2) reduced water vapor present in the aerosol introduced to the plasma (Hutton and Eaton, 1987), (3) reduced oxygen and hydroxide present as reactive species in the differentially pumped interface (Gregoire, 1989; Lim et al., 1989; Veillon and Marghoshes, 1968), and (4) reduced background as a result of reduced oxygen and hydroxide levels in the spectrum (Gregoire, 1989). Coupled with the high-resolution of the double-focusing mass spectrometer, detection limits achieved by Tsumura and Yamasaki (199 1) are in the low parts-per-quadrillion range. Ultrasonic, pneumatic, and Babington nebulizers can all be used with ICP-MS instrumentation. In fact, any nebulization system used for ICP-AES can be used for ICP-MS. Because of the severity of nonspectroscopic concomitant effects on analyte ion arrival rate at the detector-per-unit analyte concentration-i.e., analytical response (Houk and Thompson, 1988; Gregoire, 1987a,b; Beauchemin et al., 1987; Olivares and Houk, 1986; Douglas and Ken; 1988) encountered in routine aqueous nebulization ICP-MS-variations on the usual aqueous sample aerosol generation and introduction systems are more common in the ICP-MS area. Some 'Ken, Petrie. Precision Glassblowing of Colorado. 14775 East Hinsdale Ave., Englewood, CO 801 12. pers. comm. 4Leeman Labs Inc., Wentworth Dr., Hudson, NH 03051.

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 39 of the alternate methods of sample aerosol production and/or sample injection are as follows: 1. Hydride generators (Workman and Soltanpour, 1980; Thompson er al., 1978a,b; Ek et al., 1991). 2. Laser ablation (Denoyer, 1991; Hager, 1989; Abell, 1991; Denoyer et al., 1991b; Pearce et al., 1992). 3 . High-performance liquid chromatography (HPLC, including ion chromatography) (Braverman, 1992). 4. Liquid-liquid solvent extraction (Plantz, 1989; Serfass et al., 1986). 5. Flow-injection (FI) analysis (Thompson and Houk, 1986; Dean et al., 1988, Denoyer et al., 1991a; Denoyer and Stroh, 1992). 6. ETV (Gregoire, 1989). 7. Aerosol-desolvation apparatus (Veillon and Margoshes, 1968). 8. Direct-injection nebulizers (DIN).5 9. Direct-insertion devices (Gervais and Salin, 1991). 10. USN systems (Olson et al., 1977). Hydride generators (Workman and Soltanpour, I980), laser-ablation systems, DIN, and flow-injection principles are discussed in the following sections and the other systems have been described in the above references.

2. Hydride-Mercury Vapor Generator Certain elements, when reduced by NaBH,, form gases that can be directly introduced into the plasma. Arsenic (As), Sb, Bi, Se, and Te are thus reduced to form hydrides, and Hg is reduced to Hg vapor. This method of sample introduction (Fig. 5) greatly improves the detection limits of these elements compared with pneumatic nebulization because of an improvement in sample delivery and a decrease in matrix effect. Thompson et al. (1978a,b) simultaneously determined As, Sb, Bi, Se, and Te by use of ICP-AES and a hydride generator. Studies at CSUSTL indicate that by reducing As and Se to their hydrides and Hg to its vapor form and introducing these gases into the ICP, they can be quantitatively detected at 1.O, 0.5, and 0.5 pg liter-' of these elements (Workman and Soltanpour, 1980). Recently, Ek et al. (199 1 ) have used an analogous system with ICP-MS instrumentation to improve Se detection limits to 0.05 pg liter- I .

3. Laser Sampling of Solids Many solid samples are difficult or time-consuming to put into solution, e.g., soils and ceramics. Sometimes the elemental composition of grain features and 'Transgenomic/CETAC Technologies, Inc., 5600 South 42nd St., Omaha, NE 68107.

40

PARVIZ N. SOLTANPOUR ETAL.

small inclusions in the solid are of greater interest than the overall composition, e.g., minerals. To save time in sample pretreatment and to permit feature analysis, surface-sampling methods using a laser have been developed (Denoyer, 1991; Hager, 1989;Abell, 1991; Denoyer et al., 1991b; Pearce ef al., 1992). Laser ablation can be used in conjunction with ICP-AES, but most frequently the ablated aerosol is injected into an ICP, and the ions produced are subsequently detected using a mass spectrometer. Two of the manufacturers of ICP-MS instrumentation market a laser-ablation accessory.6 The accessory is equipped with an NdYAlG (neodymium-yttrium-aluminum-garnet) laser and an ablation stand. The ablation stand has an X-Y-Z translational specimen stage that is moved under computer control. Vendor software supports time-resolved data acquisition and semiquantitative analytical reports. Laser repetition rates are adjustable from a single shot to hundreds of bursts per second. Beams can be used in a defocused mode to cover approximately 1 mm of surface area or sharply focused to less than 0.02 mm (Pearce et al., 1992). The time durations and number of repeating shots are operator selectable. The amount of energy per pulse is variable. A threshold energy is required to fire the laser. The upper limit on repetition rate and energy per pulse is set either by the limitations of the laser output or by the window-material-degradation threshold.A typical pulse can be as short as a few nanoseconds (Q-switched) and delivers approximately 0.1 J of energy. In operation, the sample Ar flow to the ICP is momentarily interrupted while the ablation stage cover is removed, the sample specimen placed on the Teflon ablation stage, and the ablation cover replaced. The sample Ar flow is then resumed, and the portion of the sample to be ablated is located within the ocular of a light microscope. The specimen is focused using the X-Y-Z movement of the sample stage. The computer is notified of impending analysis, and the laser is fired. Preablation times, Laser repetition rates, and laser-power per pulse are important variables. The ablation stage is disk-shaped, with the circular top surface used to support the sample. A metal tube protrudes through the disk and serves to supply an Ar flow into the sample area. A groove in the side of the disk is used to seat an 0 ring. The ablation cover makes a gas-tight seal with the 0 ring. The cover, resembling an upside-down glass beaker, is approximately 5 cm in diameter and height. A glass sample aerosol exit tube protrudes from the side, toward the top, of the sample cover. The cylindrical side of the sample cover and sample aerosol exit tube are constructed of heavy gauge glass, and the top surface of the sample cover is made of a glasslike material that is transparent to laser light. A relatively low-power light microscope is used for viewing the specimens, requiring a high-intensity lamp inside the ablation stand next to the sample cover to illuminate the specimen. %G Elemental Inc., 27 Forge Parkway, Franklin, MA02038; Perkin-Elmer Sciex, 761 Main Ave., Norwalk, CT 06859-0012.

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 41 A video camera is sometimes used to project the ablation process onto a television-type screen. The laser, light microscope, and video camera can all be made to focus on the same point in space. In many applications, however, the light microscope and video camera focuses are set to coincide with some point on the sample surface, and the laser focus is set 1-5 mm deeper. The detection limits of metals in the solid are usually less than 1 pg g-' with the laser setup, and the elemental coverage is superior. The dry-sample aerosol produced by the laser bursts is free of many of the recombination polyatomic ions that would ordinarily accompany a major element (M) in a nebulized sample (e.g., MO+, MH+, MOH+; see Date et al., 1987). Argide polyatomic ion species, e.g., MAr+, may persist, however. The sample analysis rate can be rapid, but it depends on the analytical objectives and the variability between samples. The accuracy of the analyses is highly dependent on the availability of certified materials of composition similar to the sample. At ultra low concentrations, memory effects must be considered. For example, assume that a gold nugget is to be ablated to determine approximate elemental composition. On the next sample an elemental assay for gold content is requested on a metallic inclusion in a piece of quartz. To reduce the gold background between the two samples, the entire ICP-MS system should be shut down to permit thorough cleaning of the sample and skimmer cones, the ICP torch, the aerosol carrier line from the Laser stand, and the interior of the glass sample stage cover. Cleaning the glass sample stage cover is probably the most critical process, because the interior of the laser-ablation window becomes coated with a metallic film of elemental composition generally representative of the ablated sample, and re-ablation of the film can occur during ablation on subsequent samples. Thus, for the analysis problem at hand, the total analysis time can be a few minutes or a few hours, depending on whether the quartz piece can be run ahead of the gold nugget and, more generally, on the detection limit and accuracy requirements.

4. Direct Injection Nebulizers Direct injection nebulizers (DIN) provide for the direct injection of microvolume aqueous liquid samples into the base of torch plasma using fused silica capillary tube and a high-pressure HPLC type pump. A DIN can be used on either ICPAES or ICP-MS instrumentation. It is uniquely suited for determining nebulizer memory-prone mass isotopes of B, Hg, I, C, S, and Br or where sample volume is limited. Smith et al. (1991) have shown it capable of detecting as little as 1 ng/g of B in biological materials using ICP-MS. Powell and Boomer (1995) have shown the technique under optimal conditions accurately capable of detecting Cr at the 30 ngfliter concentration range for multiple Cr"' and Cr"' species. In addition, the technique provides for fast sample washout and high sample throughput.

42

PARVIZ N. SOLTANPOUR ETAL.

In. SPECTROMETERS A. ICP-ATOMIC EMISSION SPECTROMETRY Atoms of elements in a sample when excited emit light of characteristic wavelengths with an intensity directly proportional to the element concentration. The light is focused on the entrance slit of the spectrometer to illuminate the diffraction grating. The diffraction grating separates light into its component wavelengths of lines (spectrum). The spectral line of an analyte passes through the aperture of an exit slit and strikes a photomultiplier tube. Photomultiplier tubes produce signals directly proportional to the intensity of the spectral line. The signal is fed to the readout system, which displays intensities, concentrations, or both. Readout systems are computer controlled. The computer stores the intensities of standards and uses these data to calculate the concentrations of unknowns. Systems are available that check calibration-curve accuracy periodically, so that if the quality control (QC) limits are exceeded, the system automatically updates the calibration.' If the system is equipped with tandem nebulizers,8 one nebulizer could be shut down by the computer if clogging or another irrecoverable error has occurred, leaving the second nebulization system to finish running the samples. If the sensitivity is degraded beyond prescribed limits, or if the run is finished, there are commercially available systems that automatically shut down the ICP generator, Ar flow, and other system functions. Two types of spectrometers are commonly used (Slavin, 1971): (1) direct-reading polychromators (direct readers) and (2) scanning monochromators. Some systems are equipped with both spectrometers. Direct readers are designed to reduce the possibility of unwanted light reaching the photomultiplier tubes. The refractor plates used for fine alignment of the spectral lines are also filters that exclude stray light. The exit-slit assemblies of the photomultiplier are protected by a light shield, and the internal surfaces of the spectrometer are blackened to reduce reflections. Scanning monochromators use a variety of techniques to make a wide range of useful analytical wavelengths accessible. Fixed or movable gratings, single or multiple detectors, and movable entrance and exit slits are a few of the options available from a variety of manufacturers. The scanning is computer-controlled, fast, and accurate. In a recent demonstration for the DANR Analytical Lab, one manufacturer was able to produce 150 elemental concentrations per hour on a set 7Thom Zalinski, Thermo Jarrell Ash/Baird Corporation, 27 Forge Parkway, Franklin, MA 02038, pers. comm. 8Jim O'Dell, Leeman Labs, Inc., 6 Wentworth Dr., Hudson, NH 03051, pers. comm.

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 43 of samples that the lab analyzes for three elements per solution. Over the 2-hour run, 300 analytical concentration determinations were made, and all concentration measurements on several control checks were within +5% of the true values. Direct readers have the advantage of being faster if concentrations are being determined on more than a few elements per sample, and a smaller sample volume is required in these circumstances compared with scanning monochromators. The disadvantage of direct readers is their fixed wavelengths. Scanning monochromators, in contrast, allow the analyst to scan the entire spectrum and choose the most useful line. For laboratories engaged in both routine and research activities, a spectrometer with both a scanning monochromator and a polychromator is the best system. The manufacturers of spectrometers usually provide the required software (computer programs) for the operation of the spectrometer. This software enables the computer to do many tasks automatically. Through software commands, modem spectrometers are able to perform standardization, normalization of standard solution readings, correction of the interelemental spectral interferences, printing data out, etc. When a spectrometer is purchased, such factors as computer size, available software, printer speed, automatic interelemental spectral interference corrections, and other computer-related factors should be considered in addition to optical system factors.

B. ICP-MASS SPECTROMETRY There are at least three manufacturers of ICP-MS instruments. All produce quadrupole spectrometers, and one manufactures a high-resolution double-focusing mass spectrometer as well. The normal peak widths on quadrupoles are typically 0.5-1 dalton across the mass range; 6-250 daltons. This is sufficient to separate the baseline isotopes differing in one atomic mass unit that may be up to a factor of 2 X lo7 different in concentration in the sample. The high-resolution double-focusing electrostatic-magnetic sector mass spectrometers are capable of achieving a resolution of 50,000. Note that the resolution, R, is defined as being equal to M, the mass of interest, divided by AM, the peak width at 5% of the peak height, or R = M/AM. The resolution of the double-focusing system can be used to avoid isobaric overlap in many instances and costs about three times as much as the quadrupole-equippped unit. All manufacturers make extensive use of computers for instrument control and data processing. Among the common features of commercially available systems are the following: 1. An ICP is used as the ionization device. 2. The ions are sampled at atmospheric pressure and detected at high vacuum,

44

PARVIZ N. SOLTANPOUR ETAL.

requiring a differentially pumped interface at an intermediate vacuum-typically 1 torr (1/760 atm). 3. The pressure is very low inside the spectrometer that produces the mass-tocharge separation-typically 10-4-10-7 torr. 4. The systems are highly automated, with computers being used for instrument control and data processing. 5. All systems have rapid sequential multi-isotope capability and are able to quantitatively analyze isotopes for more than 70 elements. 6. Measurements are sequential in nature. Spectrometers do not as yet exhibit true simultaneous multielement capability. 7. Detection limits are in the low part-per-trillion range (ng liter-l; see Table 11) for generic ICP-MS units and low part-per-quadrillion range (pg liter-l; e.g., see Tsumura and Yamasaki, 1991) for many elements using high-resolution and ultrasonic nebulization but degrade as a result of several factors, including the number of elements in the analytical suite, the complexity of the sample composition, and the amount of dissolved solids in the analytical test solutions. Several add-on accessories are also available for ICP-MS and ICP-AES, i.e., USN, DIN, HPLC systems, FI accessory, hydride-generation equipment, ETV accessory, and laser-ablation solid-sampling equipment.

W. ANALYTICAL CAPABILITIES A. SELECTION OF WAVELENGTH The number of wavelengths of spectral lines generated after atom excitation occurs will vary depending on the number of electrons in the atom of an element and the number of energy steps in electron shell movement. Elements such as Fe and Co generate many spectral lines, whereas an element such as B generates very few. The theory and explanation of wavelength concepts make for exciting reading but are beyond the scope of this discussion. Those wishing to explore spectral theory more thoroughly can read Boumans’s book (1966) on the subject. For the analyst using the spectrometric technique, line selection involves finding the most useful line; i.e., a line sufficiently intense to be easily detected with a minimum of spectral interference from other spectral lines and background. Line selection can be a difficult process requiring careful examination of the spectrum. In some instances, the most useful lines may lie outside the spectral range of the spectrometer or fall in areas of high background. For some elements, only one or two useful lines are available, whereas other elements offer several useful lines.

ADVANCES IN ICP EMISSION AND ICP M A S S SPECTROMETRY 45 Winge et al. (1979) determined the relative intensities of atomic and ionic lines of elements excited in ICP. This information is partially reproduced in Table I.

B. SELECTION OF ISOTOPE Ideally, the most abundant isotope is selected for analytical work. It will produce the highest gain involving analytical measurements for the element. Thus, it is likely to offer the lowest level of detectable concentration, the best probability for analytical accuracy, and the best sensitivity among the isotopes available for the element. However, the analytical isotope selection process can be complicated both by the presence of isobaric interference from ions in the background spectra that are characteristic of the plasma, solvent, and reagents used to prepare the test solutions and by the presence of nonanalyte sample constituents in elevated concentrations relative to the analyte in test solutions (Vaughan and Horlick, 1986; Munro et al., 1986; Date et al., 1987; Gray, 1986; Tan and Horlick, 1986). The following considerations must be taken into account in the isotope selection process: analytical isotope abundance, background isobaric species, and isobaric species resulting from the sample and dependent on test-solution composition. Relative abundances of the isotopes are given in Table I1 for all naturally occurring elements (Date and Gray, 1989). For elements that do not occur naturally but are present as a result of human activity, the isotope with the longest half-life has been tabulated. Individual isotope masses are also given (Holden and Walker, 1972). These are useful when combined with the relative abundances for accurately calculating atomic masses of the elements and by themselves for those considering high-resolution, double-focusing ICP-mass spectrometerexperiments. Table I1 also has ionization energy data and some of the more common isobaric interferences that are possible at normal resolution for individual isotopes. Detection limits listed for ICP-MS were determined using the Semiconductor Equipment and Materials International (SEMI) ClO-94 Protocol’ (see Section 1V.D).

C. ICP-AES DETECTION LIMITS Detection limit is defined as the analyte concentration equivalent to two times the standard deviation of the background beneath the analyte line. However, concentrations five times the detection limits are generally required for quantitative measurements. Hence, the latter is referred to as the “quantitative detection limit” (Skogerboe and Grant, 1970). 9Serniconductor Equipment and Materials International, 805 East Middlefield Rd., Mountain View, CA 94043.

PARVIZ N. SOLTANPOUR E T A .

46

Table I Prominent Lines of Elements Emitted by ICPa.h

Ionization state'

Wave length (nm)

Ag

I I I1 I1 I1 I1 I1 I1 I1

A1

I I I I 1 I I I I I I I I

Element

As

B

I"I1,"

(CLdml)

Estimated detection limitf (pg/ml)

328.068 338.289 243.779 224.641 241.3 18 21 1.383 232.505 224.874 233.137

38.0 23.0 2.5 2.3 1.5 0.9 0.7 0.6 0.5

10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0

0.007 0.0 I 0.1 0.1 0.2 0.3 0.4 0.5 0.6

309.271 309.284 396.152 231.335 237.312 226.992 226.9 10 308.215 394.40 1 236.705 226.346 221.006 257.510

13.0 13.0 10.5 10.0 10.0 9.0 9.0 6.6 6.3 5.8 5.0 4.8 4.0

10.0

0.02 0.02 0.03 0.03 0.03 0.03 0.03

193.696

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

0.05 0.08 0.08 0.1 0.1

10.0 10.0 10.0 10.0

0.005 0.006 0.0 I 0.01

Concentration'

-

I I I I I I I I I I

228.8 12 200.334 189.042 234.984 198.970 200.919 278.022 199.048

56.0 39.0 36.0 25.0 22.0 21.0 16.0 6. I 5.7 5.5

I I I I

249.773 249.678 208.959 208.893

63.0 53.0 30.0 25.0

197.197

10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0

0.04 0.05 0.05 0.06 0.06 0.08

Comments8

OH band, NRh OH band, NR

NR

NR NR NR OH band

0. I

0.2 0.5 0.5 0.5

continues

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 47 Table I-continued

Element

Ba

Ionization statec

II

10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0

0.00 1 0.002 0.004 0.004 0.03 0.04 0.06 0.08 0.2 0.2

313.042 234.861 3 I 3. I 07 249.473 265.045 2 17.510 2 17.499 332.134 205.590 205.601

I 10.0 96.0 41.0 8.0 6.4 2.5 2.5 I .4 0.7 0.7

1.o

I .o I .o I .o 1.o I .o 1.o I .o 1.o

0.0003 0.0003 0.0007 0.004 0.005 0.01 0.01 0.02 0.04 0.04

I I I1 I I I

223.061 306.772 222.825 206. I70 195.389 227.658 190.241 213.363 289.798 2 I 1.026

87.0 40.0 36.0 35.0 14.0 12.0 10.0 10.0 9.0 7.8

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

I1 I1 I1 I

393.366 396.847 317.933 422.673

89.0 30.0 1.5 1.5

0.5 0.5 0.5 0.5

0.0002 0.0005 0.01 0.01

I1 I I1 I I

2 14.438 228.802 226.502 361.051 326.106

120.0 110.0 89.0 I .3 0.9

10.0 10.0 10.0 10.0 10.0

0.003 0.003 0.003 0.2 0.3

I1 I I1 I I I 1 1

I I I I 1

Ca

Cd

Estimated detection limitf(pg/ml)

230.0 130.0 75.0 73.0 9.1 7.8 5.2 3.7 2.0 1.9

I

Bi

Concentration' (kg/mI)

455.403 493.408 233.527 230.424 413.066 234.758 389.178 489.997 225.473 452.493

I1 I1 I1 I1 I1 I1 I1 I1 I1 Be

Wave length (nm)

1.o

0.03 0.08 0.09 0.09 0.2 0.3 0.3 0.3 0.3 0.4

Comments8

H 388.905

OH band OH band Group NR Group NR NR NR Group NR' NR NR

OH band

H 397.007 OH band

continues

48

PARVIZ N. SOLTANPOUR ETAL.

Table I-continued Ionization statec

Wave length (nm)

Cd

I 1 I

346.620 231.284 479.992

co

II

Element

Concentration' (pg/ml)

Estimated detection Iimitf(pg/rnI)

0.7 0.5 0.5

10.0 10.0 10.0

0.4 0.6 0.6

238.892 228.616 237.862 230.786 236.379 231.160 238.346 231.405 235.342 238.636 234.426 231.498 234.739

50.0 43.0 31.0 31.0 27.0 23.0 21.0 18.0 17.0 14.0 14.0 13.0 13.0

10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0

0.006

49.0 42.0 42.0 42.0 35.0 31.0 22.0 21.0 16.0 15.0 15.0 14.0 14.0 13.0 13.0

10.0 10.0 10.0 10.0 10.0

I1 I1 I1 I1 I1 I1

205.552 206.149 267.716 283.563 284.325 206.542 276.654 284.984 285.568 276.259 286.257 266.602 286.511 286.674 357.869

I I1 I I I1 I I I1 I1

324.754 224.700 219.958 327.396 213.598 223.008 222.778 221.810 219.226

56.0 39.0 31.0 31.0 25.0 23.0 19.0 17.0 17.0

10.0 10.0 10.0 10.0 10.0 10.0 10.0

n

I1 11 I1 I1 I1 I1 I1 I1 I1 I1 I1

Cr

I1 I1 I1 I1 I1 I1 I1 I1

n

cu

IJI/

10.0 10.0

10.0 10.0 10.0 10.0

10.0 10.0

10.0 10.0

10.0 10.0

10.0 10.0

Comments8

0.007 0.01 0.01 0.01 0.01 0.01 0.02 0.02 0.02 0.02 0.02 0.02 0.006 0.007 0.007 0.007 0.009 0.01 0.01 0.0 I 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.005 0.008 0.01 0.01 0.0 I 0.0 1 0.02 0.02 0.02

OH band

continues

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 49 Table I-continued Ionization state"

Wave length (nm)

cu

I I

Fe

I1 I1

Element

In/I,d

Concentratione (pg/d)

Estimated detection limitf(pg/mt)

217.894 221.458

17.0 13.0

10.0 10.0

0.02 0.02

65.0 59.0 48.0 29.0 27.0 24.0 24.0 23.0 23.0 20.0 20.0 19.0 19.0 19.0 16.0 15.0 15.0

10.0 10.0 10.0 10.0 10.0 10.0 10.0

0.005 0.005 0.006 0.01 0.01

10.0 10.0 10.0 10.0 10.0 10.0

n

238.204 239.562 259.940 234.349 240.488 259.837 261.187 234.8 I0 234.830 258.588 238.863 263.105 263.132 274.932 275.574 233.280 273.955

0.01 0.01 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02

I I I I I I1 I I I I I

294.364 417.206 287.424 403.298 250.017 209.134 245.007 294.418 27 1.965 233.828 265.987

64.0 45.0 38.0 27.0 16.0 11.0 10.0 9.4 5.7 3.9 3.6

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

I1 I

120.0 49.0

100.0 100.0

I I 1 I

194.227 253.652 296.728 435.835 265.204 302.150 365.483

0.7 0.6 0.3

100.0 100.0 100.0 100.0

I1 I

230.606 325.609

47.0 25.0

100.0 100.0

I1 I1 I1 I1 I1

n U I1

n

I1 I1 I1 I1 I1

Ga

Hg

1

In

1.7 1.1

10.0 10.0 10.0 10.0

100.0

100.0

100.0

Comments8

0.01 0.01

NR NR

NR NR

0.05 0.07 0.08 0.01 0.2 0.3 0.3 0.3 0.5 0.8 0.8 0.03 0.06 1.8 2.7 4.3 5.O 10.0

0.06 0.1 continues

50

PARWZ N. SOLTANPOUR ETAL.

Table I-continued

Element

K

Ionization state"

Wave length (nm)

Concentration' (Fglml)

Estimated detection limit.f(kg/rnl)

I I I I I I I I I I

303.936 451.131 410.176 271.026 325.856 207.926 256.015 293.263 197.745 175.388

20.0 16.0 6.4 5.4 5.0 4.2 4.2 2.0 1.7 1.6

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

0.2 0.2 0.5 0.6 0.6 0.7 0.7 1.5 1.8 I .9

I I

404.721 404.414

0.7 NM

I 000.0 1000.0

42.9 NM

460.286 323.263 274.118 497.170 256.231 413.262 413.256

3.5 2.8 1.9 1.4 0.7 0.4 0.4

100.0 100.0 100.0 100.0 100.0 100.0 100.0

0.9 1.1 1.6 2. I 4.3 1.5 1.5

I1 I1 I I1 I I1 I I I I1

279.553 280.270 285.213 279.806 202.582 279.079 383.826 383.231 277.983 293.654

195.0 100.0 19.0 2.0 1.3 1.0 0.9 0.7 0.6 0.5

1.o 1.o 1.o 1 1 .o

I1 I1 I1 I1

257.610 259.373 260.569 294.920 293.930

220.0 190.0 145.0 39.0 29.0

10.0

279.482 293.306 279.827 280.106

24.0 22.0 18.0 14.0

10.0 10.0

Li

Mg

Mn

11

Mn

I 11

I I

.o

I .o I .o 1.o 1 .o 1.o

10.0 10.0 10.0 10.0

10.0

10.0

Comments#

H 410.174

Ar 404.442

OH band

NR NR

0.0002 0.0003 0.002 0.02 0.02 0.03 0.03 0.04 0.05 0.06 0.001 0.002 0.002 0.008 0.0 I

0.01 0.0 I 0.02 0.02 continuer

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 5 1 Table I-continued

Element

Ionization stateC I

Wave length (nm)

Concentration' In/I/ (pg/ml)

Estimated detection hnit'(pghn1)

I I1

403.076 344.199 403.307 I9 I .510

6.8 6.6 6.3 5.8

10.0 10.0 10.0 10.0

0.04 0.05 0.05 0.05

Mo

I1 I1 I1 I1 I1 I1 I1 I1 I1 I1 11

202.030 203.844 204.598 28 1.615 201.551 284.823 277.540 287.15 1 268.4 14 263.876 292.339

38.0 24.0 24.0 21.0 16.0 15.0 12.0 11.0 10.0 8.0 8.0

10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0

0.008 0.01 0.01 0.01 0.02 0.02 0.03 0.03 0.03 0.04 0.04

Na

I I I I I I I1

588.995 589.592 330.237 330.298 285.301 285.28 1 288.114

101.0 43.0 I .6 0.7 1.1 1.1 0.6

100.0 100.0 100.0 100.0 1Ooo.o I000.0 1Ooo.o

0.03 0.07 I .9 4.3 27.3 27.3 50.0

I1 I1 I1 I1 I1 I1 I1 I1 I1 I1 I1 I1

83.0 75.0 60.0 43.0 42.0 41.0 40.0 40.0 34.0 31.0 31.0 30.0 28.0

100.0 IOO.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

I1

309.418 3 16.340 313.079 269.706 322.548 3 19.498 295.088 292.781 27 1.662 288.3 I8 2 10.942 272.198 287.539

100.0

0.04 0.04 0.05 0.07 0.07 0.07 0.08 0.08 0.09 0.1 0.1 0.1 0. I

I1 I I1 I1

22 1.647 232.003 23 1.604 216.556

29.0 20.0 19.0 17.0

10.0 10.0 10.0 10.0

0.01 0.02 0.02 0.02

11

Nb

Ni

Commentss

Ar 588.859

NR

OH band OH band OH band OH band OH band

continues

52

PARVIZ N. SOLTANPOUR ETAL.

Table I-continued

Element

Ionization state'

Wave length (nm)

(IJ.im)

Estimated detection Iirnitf(kg/mi)

Concentratione

I1 I I1 I I

217.467 230.300 227.021 225.386 234.554 239.452 352.454 341.476

13.0 13.0 12.0 12.0 9.5 7.8 6.6 6.2

10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0

0.02 0.02 0.03 0.03 0.03 0.04 0.05 0.05

P

I I I I I I I I I I

213.618 214.914 253.565 213.547 203.349 215.408 255.328 202.347 215.294 253.401

39.0 39.0 11.0 8.5 7.4 7.2 5.2 3.8 3.4 3.0

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

0.08 0.08 0.3 0.4 0.4 0.4 0.6 0.8 0.9 1.o

Pb

n I I I I I I 1 I I I I

220.353 216.999 261.418 283.306 280.199 405.783 224.688 368.348 266.316 239.379 363.958 247.638

70.0 33.0 23.0 21.0 19.0 11.0 9.0 8.6 7.7 6.3 5.2 5.1

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

0.04 0.09 0.1 0.1 0.2 0.3 0.3 0.3 0.4 0.5 0.6 0.6

S

I I

180.669 181.979

30.0 30.0

100.0 100.0

0.1 0.1

Sb

I I I I I I

206.833 217.581 231.147 252.852 259.805 259.809

91.0 68.0 49.0 28.0 28.0 28.0

100.0 100.0 100.0 100.0 100.0 100.0

0.03 0.04 0.06 0.1 0.1 0. I

I1 I1

II

Comments 8

Vac linej Vac line

NR

NR continues

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 53 Table I-conrinued Ionization stater

Wave length (nm)

Sb

I I I I I I I I I

217.919 195.039 213.969 204.957 214.486 209.841 203.977 220.845 287.792

Se

I I I I I

Si

Sn

Element

Concentration' (CLg/ml)

Estimated detection 1imitf(@n1)

19.0 18.0 16.0 15.0 12.0 8.7 6.6 6.5 4.7

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

0.2 0.2 0.2 0.2 0.3 0.3 0.5 0.5 0.6

196.026 203.985 206.279 207.479 199.511

40.0 26.0 10.0 1.9 0.6

100.0 100.0 100.0 100.0 100.0

0.08 0. I 0.3 1.6 5.0

I I I I I I I I I I I I I I I

25 1.611 212.412 288.158 250.690 252.851 251.432 252.411 221.667 251.920 198.899 221.089 243.515 190.134 220.798 205.813

250.0 180.0 110.0 100.0 95.0 19.0 75.0 72.0 61.0 50.0 47.0 36.0 23.0 23.0 23.0

100.0 100.0 100.0 100.0 100.0 IOO.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

0.01 0.02 0.03 0.03 0.03 0.04 0.04 0.04 0.05 0.06 0.06 0.08 0.1 0. I 0.1

I1 I I I

189.989 235.484 242.949 283.999 226.891 224.605 242.170 270.651 220.965 286.333 317.505

120.0 31.0 31.0 27.0 25.0 25.0 19.0 18.0 16.0 14.0 14.0

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

0.03 0.1 0.1 0.1 0.1 0. I 0.2 0.2 0.2 0.2 0.2

1

I I I I 1

I

Commentsg

OH band continues

PARVIZ N. SOLTANPOUR ETAL.

54 Table I-continued

Wave length (nm)

In/I/ 72.0 39.0 36.0 29.0 13.0 8.8 4.8 4.4 2.9 2.4

10.0 10.0 10.0 10.0 10.0 10.0

I1

407.771 421.552 216.596 215.284 346.446 338.071 430.545 460.733 232.235 416.180

Te

I I I I I I I I I I

214.281 225.902 238.578 214.725 200.202 238.326 208.116 199.418 225.548 226.555

73.0 17.0 17.0 14.0 12.0 11.0 11.0 6.3 2.7 2.6

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Ti

I1 I1 I1 I1 I1 I1 I1 I1 I1 I1 11

334.941 336.121 323.452 337.480 334.904 308.802 307.864 338.376 323.657 323.904 368.520

79.0 57.0 56.0 45.0 40.0 39.0 37.0 37.0 30.0 29.0 26.0

10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0

I1 I I I I I I I I

190.864 276.787 351.924 377.572 237.969 291.832 223.785 352.943 258.014

74.0 25.0 15.0 13.0 7.0 2.9 2.2 1.7 1.7

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Element Sr

Ionization state" I1 I1 I1 11

I1 I1 I1

I

I1

TI

Concentration' (pglml) 1.o 1 .o

10.0

10.0

Estimated detection limitf(pgh1)

Comments8

0.0004 0.0008 0.008 0.01 0.02 0.03 0.06 0.07 0.1 0.1 0.04 0.2 0.2 0.2 0.3 0.3 0.3 0.5 1.1 1.2 0.0004 0.0005 0.005 0.007 0.008 0.008 0.008 0.008 0.01 0.01 0.01

OH band

OH band OH band

OH band OH band

0.04 0.1 0.2 0.2 0.4 1.o

I .4 1.8 1.8 continues

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 5 5 Table I-conrinued

Element V

tJI/

ConcentrationP (pg/ml)

Estimated detection limitf(pg/ml)

309.311 310.230 292.402 290.882 311.071 289.332 268.796 311.838 214.009 312.528 327.612 292.464 270.094

60.0 47.0 40.0 34.0 30.0 29.0 29.0 25.0 20.0 20.0 19.0 18.0 17.0

10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0

0.005 0.006 0.008 0.009 0.01 0.01 0.01 0.01 0.02 0.02 0.02 0.02 0.02

I1 I1 I I I I I I I I

213.856 202.548 206.200 334.502 330.259 481.053 472.216 328.233 334.557 280.106 280.087

170.0 75.0 51.0 2.2 1.3 1.3 0.7 0.6 0.4 0.4 0.4

10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0

0.002 0.004 0.006 0.1 0.2 0.4 0.4 0.5 0.8 0.8 0.8

I1 I1 I1 I1 I1 I1 I1 I1 I1 I1 I1 I1 I1 11 I1 I1 I1

343.823 339.198 257.139 349.621 357.247 327.305 256.887 327.926 267.863 272.261 273.486 274.256 270.013 350.567 355.660 348.115 256.764

42.0 39.0 31.0 30.0 30.0 25.0 22.0 21.0 20.0 16.0 14.0 14.0 12.0 12.0 12.0 12.0 11.0

10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0

0.007 0.008 0.01 0.01 0.01 0.01 0.01 0.02 0.02 0.02 0.02 0.02 0.03 0.03 0.03 0.03 0.03

Ionization state"

I1 I1 I1 I1

II I1 I1 I1 I1

LI I1 I1 I1 Zn

Zr

1

Wave length (nm)

Comrnentsg OH band OH band

OH band

OH band OH band

NR NR

continues

56

PARVIZ N. SOLTANPOUR ETAL.

Table I-continued

Element Zr

Ionization stateC I1 I1 I1 11 I1

I1 Il I1

Wave length (nm)

272.649 330.628 316.597 318.286 328.471 274.586 275.221 357.685

I,,/Ibd 11.0 11.0 11.0

11.0 10.0 10.0 10.0 10.0

Concentratione (p,g/ml)

Estimated detection Iimitf(pg/mI)

10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0

0.03 0.03 0.03 0.03 0.03 0.03 0.03 0.03

Comments#

OH band OH band

"Adapted from Winge et al., 1979. bElements are arranged alphabetically; lines of each element are in order of decreasing I,/&, ratio. =Romannumerals I and II indicate that the spectral lines originate from the neutral atom or the singly ionized state. "Ratio of net analyte intensity to background intensity. 'Concentration of the single-element analyte solution used for the wavelength scans from which the prominent lines were determined. 'Detection limits estimated from the IJI, ratios using the formula DL = 0.03C/(In/lb),where C is the concentration of analyte. #Includes interference information when a component of the background spectrum overlaps an analyte line ( e g , the Ba 389.178 nm line is located on the H 388.905 nm line) or when an analyte is located in a complex molecular band system ( e g , the OH 306.3 nm system) where band components may cause spectral interferences. The notation of molecular bands does not preclude the use of analyte wavelengths within the band region. "Not resolved, indicating components of an unresolved pair of lines. 'Components of an unresolved group (three or more lines). Only the wavelength of the strongest line is listed. Vacuum lines for S require either a vacuum or an Ar-purged spectrometer; see Section 1V.A. J

Detection limits are a function of excitation source and sample matrix as they affect the intensity of background radiation. The detection limits are also affected by sample-delivery efficiency. Detection limits are further affected by operational parameters of plasma spectrometers such as power, height of observation, and flow rate of the sample carrier gas. The optimum operational parameters often differ for different groups of elements, and therefore, one should select compromise operational parameters for simultaneous multielement analysis. Winge et af. (1979) calculated the ICP detection limits of different elements (Table I), assuming that the standard deviation of the background is approximately 1% of the background signal level and that the detection limit is three times the standard deviation of the background. The formula used for these calculations is given in a footnote to Table I. These values can be used as first approximation in the absence of other measured values for detection limits.

Table Il Isotope Data for the Elements”

Atomic no.

Isotope mass

Relative Atomic abundance (’36) mass

Ionization energy

(ev)

first

second

Detection limit p,g liter-’

Element

Symbol

Mass no.

Hydrogen” Deuterium

H D

1 2

1.007825 2.014 102

99.9855 0.0145

1.008

13.60

-

Helium

He

3 4

3.016030 4.002603

0.00014 99.99986

4.003

24.59

54.41

Lithiumb

Li

6 7

6.015 12 7.01601

7.50 92.50

6.939

5.39

75.64

0.19

Beryllium

Be

9

9.01218

100.00

9.012

9.32

18.21

0.15

Boron”

B

10

10.01294 11.00931

19.78 80.22

10.811

8.30

25.15

0.37

I1

Carbon”

C

12 13

12.00000 13.00336

98.888 1.112

12.011

11.26

24.38

Nitrogenb

N

14 15

14.00307 15.00011

99.633 0.367

14.007

14.55

29.61

Oxygen”

0

16 17 18

15.99492 16.99913 17.99916

99.759 0.037 0.204

15.999

13.62

35.15

-

Comments

H,O decomposition

Response hysteresis

-

C, N,and 0 are entrained from the atmosphere surrounding the ICP and present as contaminants in the Ar.

-

H,O decomposition

continues

Table II-continued

Atomic no.

9

Element Fluorine

Symbol

Mass no.

Isotope mass

F

19

18.99841

Relative Atomic abundance(%) mass 100.00

18.998

Ionization energy

(eV)

first

second

Detection limit k g liter-'

17.42

34.99

4 w ?d

10

Neon

Ne

20 21 22

19.99244 20.99385 21.99139

11

Sodium

Na

23

22.98977

12

Magnesiumb

Mg

24 25 26

23.98504 24.98584 25.98259

13

Aluminum

Al

27

14

Silicon

Si

15

Phosphorous

16

Sulfurb

Comments 19( IH180)+, 19(

-

20.183

21.56

41.08

22.990

5.14

47.30

0.46e

&Ca+

78.70 10.13 11.17

24.312

7.65

15.03

0.35

48ca++ 24 I 2

26.98 154

100.00

26.982

5.99

18.83

0.21

28 29 30

27.97693 28.97650 29.97377

92.21 4.70 3.09

28.086

8.15

16.34

13f

P

31

30.97376

100.00

30.974

10.48

19.72

78f

S

32 33 34 36

3 1.97207 32.97146 33.96787 35.96708

32.064

10.36

23.4

90.92 0.257 8.82 100

lHlHIHl60)+

2.5 ppb background

+

m

m

95.018 0.760 4.215 0.0 14

5w

, {

c 12CI'

17

18

Chlorineh

Argonh

CI

Ar

35 31

34.96885 36.96590

75.53 24.47

35.453

13.02

23.80

36 38

0.337 0.063 99.600

39.948

15.76

27.63

40

35.96755 37.96273 39.96238

3' 0.08d

19

Potassium"

K

39 40 41

38.96371 39.974 40.96183

93.10 0.0118 6.88

39.102

4.34

3 1.82

0.42'

20

Calcium

Ca

40 42

39.96259 4 1.95863

96.97 0.64

40.08

6.11

11.87

0.28f

43 44 46 48

42.95878 43.95549 45.95369 47.95253

0.145 2.06 0.003 0.185 100.00

44.956

6.56

12.80

0.15

47.90

6.84

13.64

21

Scandium

sc

45

44.956

22

Titanium

Ti

46 41 48

45.95263 46.95177 47.94795

7.93 7.28 73.94

49 50

48.94787 49.94478

5.5 1 5.34

50 51

49.94716 50.9440

0.24 99.76

23

Vanadium

V

in units of pg/secd

0.49

50.942

6.74

14.65 0.25

continues

Table II-continued

Atomic no. 24

Element Chromium

Symbol

Mass no.

Cr

50

Isotope mass

Relative Atomic abundance (%) mass

52

49.9460 5 1.9405

4.3 I 83.76

53

52.9407

9.55

54

53.9389

2.38

51.996

Ionization energy

(ev)

first

second

6.77

16.49

Detection limit p,g liter- I

Comments

0.21

25

Mangansese

Mn

55

54.9381

100.00

54.938

7.43

15.64

26

Iron

Fe

54

53.9396

5.82

55.847

7.90

16.18

56 57 58 59

55.9349 56.9354 57.9333 58.9332

91.66 2.19 0.33 100.00

58 61 62 64

57.9353 59.9308 60.93 11 61.9283 63.9280

63 65

62.9296 64.9278

0.26

m 0

27

Cobalt

co

28

Nickel

Ni

60

29

Copper

cu

0.489

58.933

7.87

17.05

0.23

67.88 26.23 1.19 3.66 1.08

58.71

1.63

18.15

0.49

69.09 30.9 1

63.54

1.73

20.29

0.34

63 { 2 3 ~ ~ 4 0+h 63 } [ 3 I PI65 [

2SMg40h) + , I 3 q a +

160)+

30

Zinc

Zn

64

63.9291

48.89

65.37

9.39

17.96

0.78

.

6 4 ~ i +64 24 { Mg40Ar)+,

64{48ca'60)+,64(31PlH160160)+,

64(32~160160)

66

65.9260

+,

64(32~32~]+,

64(63C~1H)+ [ 26Mg40Ar] ,66 { I P3'C1 ] +, w S 1 6 0 1 6 0+, ) 66{65C~'H]+.

27.81

+

l32~~++

67

66.9271

67 ( 3Sc~I60160) +,

4.11

67{IH34sI600'60)+, 1%.13SBa++

68

67.9248

68( 3 2 ~ 3 6 h+, ) 681 3lp37c1)+,

18.57

68{ 3sC11H'60160]+, 135.136,137Ba++ 136ce+ +

70

69.9253

Tie+,I%++,

0.62

I4oce++

31

Gallium

Ga

69

68.9256

60.4

71

70.9247

39.6

70

69.9243

20.52

72

71.9221

27.43

73

72.9235

7.76

74

73.9212

36.54

76

75.9214

7.76

"'1 37CI'H'60'60)+, 141p,.++

69.72

2 32

Germanium

Ge

72.59

continues

Table II-continued

AtomiC no.

33

34

Element Arsenic

Seleniumb

Symbol

Mass no.

As

75

Se

74

Isotope mass

74.9216

73.9225

Relative Atomic abundance(%) mass

100.0

0.87

74.922

78.96

Ionization energy

(eV)

first

second

9.82

18.63

9.75

Detection limit pg liter-'

0.55

21.5

Comments 75( 35C140&)

+, ISWd+ +,

149. I soSm++ 1s 1Euf

7%+,

+

l48Nd++

147.148, 149sm++

76

75.9192

9.02

77

76.9199

7.58

78

77.9173

76Ge+,76(36APAr)+,

2.1

lS2sm++ I S l . l 5 3 b + + , I%?Gd++ 77[ 37C140kj +, 154Srn++ , 153Eu++ 154,155Gd++

23.52

7 8lSS.IS6,1S7Gd++, ~ r +7, 9 ~ ~ A P O 1 A, ~++ 156D 80Kr+, 80{4oAFOAr}+, 80(40Ca40Ar)+,ImGd++, ImJ6IDY 82Kr+, 82(42Ca40Ar)+,

80

79.9165

49.82

82

81.9167

9.19

+

Y

++

163,IMDy++

35

Bromineb

Br

79

78.9183

50.537

79.909

11.85

21.6

11

79 ( I H38M

IS%++

81

80.9163

49.463

sl(

165Ho++

O k j + , 157,158Gd++

. lS8D Yf +

lH4OflAr}+, 81 ( 3 2 ~ l ~ l 6 0 1 6 0 1 6 0 ) + ,

161.162,163~++, 162&++, Y Response hysteresis

36

Krypton

Kr

78

77.9204

0.35

83.80

14.00

24.57

8 0 s e + , R O ( 4 0 ~ O h J + , 160Gd++

80

79.9164

2.27

82

81.9135

11.56

83

82.9141

11.55

84

83.9115

56.90

%r+, ~ ( 4 4 C a 4 0 h ) + ,

86

85.9106

17.37

86Sr+, 86(46Ca40h)+, 171.172.1 73yb+ +

85 87

84.9118 86.9161

72.15 27.85

85.47

84

83.9134

0.56

87.62

160,161DY++ 82Se+,82( 42Ca40Ar]+, 163,164Sm++ 165H0++ 164f$++ 83 ( 4 3 c a 4 0 h ) +, 165~,,++ 166.167Er++

167.168~++I T m + +

37

38

Rubidium

Strontiumb

Rb

Sr

4.18

27.5

0.13

17qr++

l68yb++

I 6 q m + + 170.17lyb+f

87sr+ I73,174yb++

5.69

11.03

84=+

I67,168&.++ I6Tm++

I68yb++

86 87 88

85.9093 86.9089 87.9056

39

Yttrium

Y

89

88.9059

40

Zirconium

Zr

90 91 92 94 96

89.9047 90.9056 91.9050 93.9063 95.9083

41

Niobium

Nb

93

92.9064

86=+

9.86 7.02 82.56 100.0 5 1.46 11.23 17.11 17.40 2.80 100.0

17 1.172,173yb++

87Rb+ 173.174yb++

0.20 88.905

6.53

12.23

0.15

91.22

6.95

13.13

0.26

92.906

6.88

14.0

0.14

I76ybt+

continues

Table II-conrinued

no. 42

2

Element

Symbol

Molybdenum

Mo

Mass no.

Isotope mass

92 94 95 96 97 98 100

9 1.9068 93.9051 94.9058 95.9047 96.9060 97.9054 99.9075 96.9 98.9

Relative Atomic abundance(%) mass 15.84 9.04 15.72 16.53 9.46 23.78 9.63

95.94

Ionization energy

(W

first

second

7.10

16.15

97 99

44

Ruthenium

Ru

96 98 99 100 101 102 104

95.9076 97.9053 98.9059 99.9042 100.9056 101.9043 103.9054

5.51 1.87 12.72 12.62 17.07 31.63 18.58

101.07

102.905

7.45

18.07

106.4

8.33

19.42

103

102.9055

100.00

46

Palladium

Pd

102

101.9056 103.9040 104.9051 105.9035

0.96 10.97 22.23 27.33

104

105 106

+

0.27

Tc

Rh

92a+

%a+, %Ru+

Technetium

Rhodium

Comments

94zr

43

45

Detection limit pg liter-'

7.28

15.26

7.36

16.76

t l R = 2.6E6 years"' t1,2 = 213,ooO years'

%a+, %Mot y8Mo+

0.20

-J

0.49

47

48

Cadmium

108

107.9039

26.71

110

109.9052

11.81

Ag

107 109

106.9051 108.9048

51.817 48.183

107.87

7.58

Cd

106

105.9065

1.22

112.40

8.99

108

107.9042

0.88

110

109.9030

12.39

111 112

110.9042 111.9028

12.75 24.07

113 114 116

112.9044 113.9034 115.9048

12.26 28.86 7.58

21.48

0.37

I07 { 9 Izr160 ) + 1 0 9 ( 9 3 ~ b l 6 0 )+

losPd+,

16.90

los{%160)+, 1061 8 9 y l ~ I 6 0 ) +

lOSPd+

108 92 1

{

I6

l08( 92M0160 ) +

OI+.

IlOpd+, I l 0 ( 9 ~ 1 6 0 ~ + , I LO( 94M0160 ] +

0.39

1 I I (95M0160) +

I I 2sn+, 1 121% a 1 6 0 ) + ,

I I2( 96M0160 ) + I I 3 h + , I 1 3 ( 97Mo160) + I 14Sn+ I l4{ 98Mo160) + ll.5sn+, 116( l ~ o l Q ] + , 2 3 ~ + +

49

Indium

In

113 115

112.9041 114.9039

4.28 95.72

114.82

5.79

50

Tin

Sn

112

111.9048

0.96

118.69

7.34

114 115 116 117 118

113.9028 114.9034 115.9017 116.9030 117.9016

0.66 0.35 14.30 7.61 24.03

18.86

I I3cd+, I 1 3 ( 97M0160) i

0.10

llsSn+

14.63

continues

Table 11-continued

Atomic no.

51

52

Element

Antimony

Tellurium

Symbol

Sb

Te

OI

Mass no.

Isotope mass

119 120 122 124

118.9033 119.9022 121.9035 123.9053

8.58 32.85 4.72 5.94

121 123

120.904 122.9042

57.25 42.75

121.75

120 122 123 124 125 126

0.089 2.46 0.87 4.61 6.99 18.71 3 1.79 34.48

127.60

130

119.9040 121.9031 122.9043 123.9028 124.9044 125.9033 127.9045 129.9062

128

53

Iodine

I

127

126.9045

54

Xenon

Xe

124 126 128 129 130 131 132

123.9061 125.9043 127.9035 128.9048 129.9035 130.0951 131.9042

Relative Atomic abundance (%) mass

100.0

0.096 0.090 1.919 26.44 4.08 21.18 26.89

Ionization energy

(eV)

Detection

first

second

7.85

16.5

limit pg liter-'

Comments

0.20 1 2 3 ~ ~ i

9.01

IZOS,+

18.6

12zsn+

123Sb+ 124Sn+

0.33

124xe+

+

1 26xe

128xe+,128( 1 1 2 ~ , , 1 6 0 ) + l 3 0 ~ ~1+ 30xe+l 3 O (

126.904

10.46

19.09

131.30

12.13

21.21

5.1

l14Sn160)+

Response hysteresis 124Sn+ 124Te+ 12

6+ ~

~

1 2 8 ~ ~12x1 + I 12Snl6 0 )+ 129{*9y@Ar] + 130Ba+ 1 3 q e + ,

130(114~n16~]+

I31 { I I5snI60] + 1 3 2 ~ ~132( + 116~n160)+

s

m

m

m

m

m a

67

2

-0

a 0

'pable II-continued

Atomic no.

Mass no.

Isotope mass

Relative Atomic abundance(%) mass

Ionization energy

(eV)

first

second

Element

Symbol

61

Promethium

Pm

145 147

144.9 146.9

62

Samariumb

Sm

144 147 148 149 150 152 154

143.9121 146.9149 147.9149 148.9172 149.9173 151.9198 153.9222

3.09 14.97 11.24 13.83 7.44 26.72 22.71

150.35

5.63k

11.07

63

Europium

Eu

151 153

150.9199 152.9213

47.82 52.18

151.96

5.64

11.25

64

Gadolinium

Gd

152 154 155 156 157 158

151.9198 153.9209 154.9226 155.9221 156.9240 157.9241

0.20 2.15 14.73 20.47 15.68 24.87

157.25

6.16

12.1

160

159.9271

21.90

159

158.9254

100.00

158.925

5.98

11.52

%

65

Terbium

Tb

Detection limit pg liter-'

Comments tllZ = 18 years,' 14sNd+ t,,, = 2.623 years,d 147Sm+ IwNd+

0.26

159( 143NdI60)+

66

Dysprosium

DY

156 158

155.9243 157.9244

0.052 0.090

160

159.9252

2.29

161 162

160.9270 161.9268

18.88 25.53

163 164

162.9288 163.9292

24.97 28.18

100.00

67

Holmium

Ho

165

164.9304

68

Erbium

Er

162

161.9283

0.136

164

163.9292

1.56

166 167 168

165.9303 166.9321 167.9324

33.41 22.94 27.07

170

169.9355

14.88

100.00

.c m

69

Thulium

Tm

169

168.9343

70

Ytterbium

Yb

168

167.9339

0.135

170

169.9348

3.03

162.50

5.93

11.67

164.930

6.02

11.80

167.26

6.10

11.93

168.934

6.18

12.05

173.04

6.22

12.17

0.27

l69( 153Eu160)+

continues

Table 11-continued

Mass

Atomic

no.

2

71

Element

Lutetium

Symbol

Lu

no.

Isotope mass

Relative Atomic abundance (%) mass

171 172 173 174

170.9364 171.9364 172.9382 173.9389

14.31 21.82 16.13 31.84

176

175.9426

12.73

175 176

174.9408 175.9427

97.41 2.59

174.97

Ionization energy

(eV)

first

second

6.15

13.9

Detection limit p g liter-'

0.21

Comments

175{ 159TbI60]+ 176yb+

I 76 {

72

Hafnium

Hf

174

173.9401

0.18

176

175.9414

5.20

177 178 179 180

176.9433 177.9437 178.9485 179.9466

18.50 27.14 13.75 35.24

180

179.9476

0.012

181

180.9480

99.988

178.49

6.65'

176m+

176(

160GdlhO}+,

'WYI6OI +

1 7 4 y b + , 174( lS8Gdl60]f,

14.92'

1741 l.5XDy160 } + I 7 6 y b C , 176 { 160G d 1 6 0 ] + ,

1761 I6ODYI6O) +

73

Tantalum

Ta

177 { I6 I Dy I 6 0 ] + l78{ 1 6 2 D y 1 6 0 ] + 179{

0.50

, 178 { 162F P O ] +

163Dy160)+

l S C ~ a + , Isow+, 180 { 1 6 4D Y ' ~ O ) + ,

I SO{ I WEr160} +

180.948

7.88

I80Hf+, 18OW+,

16.2

180 { 164Dy1601+,

1 x 0 { 1 6 4Er'"OJ+

0.25

181{ 1 6 S H o 1 6 0 ) +

74

Tungsten

W

180

76

Rhenium

Osmiumb

186

185.9544

185 187

184.9530 186.9558

37.07 62.93

186.2

184

183.9526

0.02

190.2

185.9539

1.59

187 188 189 190

186.9558 187.9559 188.9582 189.9585

1.64 13.3 16.1 26.4

192

191.9615

41.0

Ir

191 193

190.9606 192.9626

37.3 62.7

Pt

190

189.9600

Re

0s

78

Platinum

192

191.9611

7.98

1 8 o ~ f f 18tva+, IXO( 16-IDy160)+,

17.7

180{ I 64Erl 6 0 ) t

182 { 1 6 6 ~ ~ 1 6+0 )

28.41

-4 e

Iridium

183.85

26.41 14.40 30.64

186

77

0.14

181.9483 182.9503 183.9510

182 183 184

75

179.9467

1x3 { 167Er160)+

IX40st, 184{168ErlhO]+ 1 8 4 { I6Xybl60

0.013

)+

1860s+, IH6{ 170Er160]+, 1 8 6 { I70yb 160 ) +

7.87

I85 { I 6 q m 1 6 0 ) +

16.6 0.15

8.73

lX70s+

, 187 ( 171Y b ' 6 0 ) +

184w+, I S { 1 6 8 ~ ~ 1 6 0 ) + ,

17.0

184 { 168Y b ' 6 0 )

+

186w+, 186{170Er160)+, 186{ 170y b l 6 0 ) + 187Re+, l87{ 171yb160)+ 188{ 172ybI6o)C

189 { 173ybl60) 1-t

,

+

190 174

{

yb'601+,

1W{1 7 4 ~ f I 6 0 ] + 1 9 2 ~ + ,IY2(176ybl60)+,

19211 7 6 ~ f 1 6 0 ) +192{ , 176~,16o)+

192.2

191{ 175~~l60)+

9.12' 0.75

195.09

8.96

18.56

I 9 3 [ 177~fI60)+

I 900s+ 190( I74yb160) + 1 9 0 ( 174HfI60)+

0.78

1Y20s+,

192( 176ybI60)+,

I%?[ 176Hf'Cf))+,I%?( 176blCf))+

continues

'Igble II-continued

Atanic no.

Element

Symbol

Mass no.

Isotope mass

Relative Atomic abundance(%) mass

194 195 196

193.9627 194.9648 195.9650

198

197.9679

7.21

Ionization energy

(ev)

first

second

9.23

20.5

limit pg liter-'

Comments

32.9 33.8 25.3

79

Gold

Au

197

196.9666

100.00

80

Mercuryb

Hg

196

195.9658

0.146

198 199 200 201 202 204

197.9668 198.9683 199.9683 200.9703 201.9706 203.9735

10.02 16.84 23.13 13.22 29.80 6.85

203 205

202.9724 204.9744

204 206 207 208

203.9731 205.9745 206.9759 207.9767

209

208.9804

196.967 200.59

10.44

18.75

29.5 70.5

204.37

6.11

20.42

1.48 23.6 22.6 52.3

207.9

N -4

81

82

83

Thallium

Leadb

Bismuth

TI Pb

Bi

100.0

203( 1 8 7 ~ ~ l 6 0i )

0.11 7.42

15.03

2wHgi

0.33 208.98

7.29

16.68

0.15

-4

w

84

Polonium

Po

209

208.9824

8.43

-m

t,, = 102 yearsh

85

Astatine

At

210

209.9870

9.5

-"

t,,, = 8.1 hours"

86

Radon

Rn

222

222.0176

-m

tlCl = 3.824 days"

87

Francium

Fr

223

223.0198

4

-m

t I n = 22 minutes"

88

Radium

Ra

226

226.0254

5.28

10.144

-rn

t,,, = 1600 yearsh'

89

Actinium

Ac

227

227.0278

6.9

12.1

-

t,,

= 21.77 years"

90

Thorium

Th

232

232.0381

6.08'

11.5

-m

t,,

= 1.4E10 yearsh

91

Protactinium

Pa

231

231.0359

-m

t,, = 32,500 years"

92

Uranium

U

234 235 238

234.0410 235.0439 238.0508

0.0057 0.72 99.27 99

93

Neptunium

NP

237

237.0482

94

Plutonium

Pu

238 239

238.0496 239.0522

240 244

240.0538 244.0642

10.746

100.00

232.04

238.03

6.05k

m

14.72* 0.03 1

-m

+

5. I

t,,, = 247,000 years t,, = 7.1E8 years tll, = 4.51E9 years t,/, = 2.14E6 yearsh.'

t,, = 87.8 years, 238U+ t,,, = 2.439E4 years, 239 238

I 1 UHJ+

t,,

=

6540 years +

zM(238~160)

'tV2

=

8.3E7 yearsh 95

Americium

Am

241

241.0568

t,12 = 433 years continues

Table II-continued

Atomic no.

96

Element

Curium

Symbol

Cm

Ionization energy

(ev)

first

second

Detection limit pg liter-

Mass no.

Isotope mass

243

243.0614

-rn

tl,* = 7370 yearsh'

244 247

244.0628 247.0704

-m

t,,

Relative Atomic abundance (a) mass

'

Comments

tin = =

17.9 years 3.54E7 years"'

"The authors gratefully acknowledge permission to reproduce much of the isotope data from Dr. A. L. Gray, the surviving editor, and Blackie and Sons, the publisher (Date and Gray, 1989). bElements for which isotope variations in nature are known or suspected (Gregoire, 1989). 'Isotope(s) monitored in negative ion mode (Heiftje ernl., 1988; Chisum, 1992). '%otope(s) monitored using positive ion helium microwave induced plasma as a detector for a gas chromatographic set up (Brown et al., 1988). 'Flame atomic emission. QCP atomic emission with ultrasonic nebulization. ZGraphite furnace atomic absorption. 'Isotope of element having longest half-life (Weast and Astle, 1979; Holden and Walker, 1972). 'Most commonly available long-lived isotope of element (Weast and Astel, 1979). hotope used as internal standard in authors' lab; thus, the detection limit has not yet been determined. 'A. L. Gray, pers. comm. regarding revised ionization potentials. 'Tungsten is routinely present in our test solutions at 100 pg/mI; thus, there is a relatively high residual background that inordinately inflates the method detection limit (MDL) determination. "'Isotopes of element not available for instrument calibration of MDL determination. "MDL not yet evaluated.

ADVANCES rN ICP EMISSION AND ICP MASS SPECTROMETRY 75 Table I11 gives the measured ICP detection limits of some elements in pure water, in 10% HCl, and in a solution containing major elements of a typical arid-region soil digest (assuming a dilution factor of 50). The pure-water values were obtained from literature (Robin, 1979). Other values were determined at CSUSTL. These blank or zero concentration solutions were analyzed 10 times to determine the standard deviation of the background signal at the wavelength of the elements of interest. These standard deviations were multiplied by 2 and changed to their apparent concentration equivalents using appropriate calibration curves. These values are called detection limits by definition. The detection limits in 10%HCl solution are given for those elements for which the CSUSTL ICP spectrometer has channels. The simulated soil digest was prepared in a 10% HCI solution containing Al, Fe, K, Na, Cu, Mg, Ti, Mn, and P; therefore, the detection limits for these elements are not given in the third column of Table 111. It should be emphasized that the last two columns of Table I11 represent detection limits that are more realistic and could be obtained under routine conditions. The detection limits close to those in 10% HCI are probably obtainable for soilwater extracts, NH,HCO,-DTPA extracts, and other extracts of low background. However, for total soil digests, the last column of Table I11 is more realistic. It seems that detection limits decrease by one order of magnitude when going from soil extracts to total soil digests. Table 111 reveals the deterioration of the detection limits due to interelement spectral interferences.For example, in the case of As, a detection limit of 0.7 ppm is shown. This detection limit will preclude determination of As in total soil digests using direct nebulization. In this case, As is separated from major soil constituents by the hydride-generation technique to be discussed later (see Section VILE). When interelemental interference is not as severe as in As, other correction techniques, to be discussed later, can be used.

D. ICP-MS DETECTION LIMITS Detection limits are one to three orders of magnitude lower by ICP-MS than by ICP-AES for most elements measurable by both techniques. However, a few key analytes of agricultural-agronomical interest exhibit better detection limits by ICP-AES than by ICP-MS, including sulfur and calcium. Alist of conservative detection limits appropriate under relatively ideal conditions are given in Table 11. The detection limits in Table I1 were determined using SEMI C10-94 protocol. For these measurements, solutions containing analyte concentrations differing by a factor of 10 were prepared. The lower concentration produced a measurable response above the zero-concentrationblank. The 6-242 dalton mass range was selected for the scans. Eight integration intervals per unit mass were taken, with response effectively integrated for a total of 0.27 sec per unit mass. One scan

76

PARVIZ N. SOLTANPOUR ETAL. lsble III Detection Limits in an Ideal Solution (PureWater), a 10%HCI Solution, and a SimulatedArid-Soil Digest ICP Ideal solution"

Element

(dm0

Ag Al AS Au B Ba Be Bi Ca Cd

0.004

Ce

co Cr

cu Fe Ga Hf Hg In

K La Li Mg Mn Mo Na

Nb Ni P Pb

Pt

Sb Se Si Sn Sr Ta Ti

U

0.0002 0.04 0.04 0.0007 o.ooo02

O.OOO4 0.05 0.m2 0.002 0.0007 0.003 0.0003 0.0001 0.0003 0.0006 0.01 0.001 0.03 O.ooOo5 0.0003 0.m1 0.00006 0.0002 0.0002 O.ooo07 O.OOO4 0.02 0.002 0.08 0.2 0.01 0.03 0.m2 0.002 0.0007 0.03

10%HClb (dml)

Simulated soil digest'

-

-

0.03 0.02

(~dml)

0.7

-

-

0.002 0.001

0.03 0.001

-

-

0.002 0.001

-

0.01 0.003 0.003 0.003

0.006

0.02

0.01 0.01

0.001

0.2 -

0.004

0.02

0.02

0.1 O.OOO4

0.01

0.005 0.03 0.03

0.02

-

-

0.009

0.07 -

0.2

-

-

0.0006

0.007

-

-

0.002

-

-

continues

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 77 Table 111-continued ICP

Element

V

w Zn Zr

Ideal solution" (PLglml)

10% HCl"

Simulated soil digest"

(Pdml)

(Pdd)

0.0002 0.0007 0.002 O.OOO4

"Data from Robin, 1979. "Data from Soltanpour et al., 1982. 'Data from Soltanpour er al., 1982. Simulated soil digest contained 10% HCI and the following elements at the given concentrations in micrograms per milliliter: A1 = 1500, Fe = 500, K = 400, Ca = 200, Na = 200, Mg = 100, Ti = 60, Mn = 20, P = 15.

required 105 sec, and an average of two successive scans were used for one response measurement. Ten duplicates were done while nebulizing the lower concentration solution and ten while using the higher concentration. Standard deviations at both concentration levels were calculated, pooled, and converted to concentration using the slope of the response versus concentration curve calculated from the mean response at both levels and the known concentration difference. Results were multiplied by a scale factor of 3.7, resulting in a method-detection limit (MDL) at the 99.87% level of confidence. Measurements could be improved by a factor of ten by narrowing the mass scan range to 7-10 analytical isotopes in the same solution by ICP-MS over a 1 min 45 sec period. For fewer isotopes per solution run for the same length of time, the detection limits would improve because of the increased duty cycle on the analytical mass. Some of the detection limits in Table I1 can be improved if necessary by adopting strict cleanliness procedures to reduce analyte contamination in reagents and glassware. Detection limits can also be improved by analytical isotope observation using mass spectrometers capable of resolution equal to or better than 3500 (Tsumura and Yamasaki, 1991; Bradshaw et al., 1989; Appendix). In addition, detection limits could be reduced by increasing the duty cycle on the analytical mass. This could be done by using time-of-flight mass spectrometers instead of quadrupole mass spectrometers (Hieftje, 1992). Analytical gains for generic ICP-MS, in units of analyte-response-per-unit-analyte concentration, decrease more rapidly with increasing concomitant concentration and begin degrading at lower concomitant concentrations in ICP-MS than in ICP-AES (Beauchemin, 1989; Houk and Thompson, 1988; Houk, 1986; Gregoire, 1989; Beauchemin et al., 1987; Douglas and Kerr, 1988; Gregoire, 1987a,b).Thus, comparison of detection limits for the two methods is accurate for describing analysis of test solutions with total dissolved solids up to approximately 100-500

78

PARVIZ N. SOLTANPOURETAL.

mg liter-', with the range depending on several factors, including the mass(es) and ionization potential(s) of the concomitant(s) (Gregoire, 1989).This fact, coupled with the outstanding detection limits exhibited by the ICP-MS, makes it a more natural choice for a chromatographicdetector than the ICP-AES. With continuing interest in chemical speciation, much literature has appeared in the area of ICP-MS involving ion exchange, HPLC, andor liquid-liquid solvent extraction prior to detection. Comparing detection limits of ICP-MS and ICP-AES must be done on an analysis-by-analysisbasis. If digestion of solid materials is involved, the detection limits between ICP-MS and ICP-AES could be about the same because the ICP-AES can tolerate 10-100 times more dissolved solids than the ICP-MS before the analytical sensitivity becomes adversely affected. Many of the apparatus mentioned later are designed to allow the concentration of concomitant in the sample solutions to be increased while maintaining the analytical response.

V. ICP-AES INTERFERENCES A. SOLUTEVAPORIZATION In emission spectrometry, refractory compounds such as calcium phosphates or calcium aluminatesare vaporized in the excitation sources. These compounds may not dissociate in some emission sources and hence interfere with analysis. For example, Johnson (1979) showed that A1 suppresses the Ca signal in direct-current plasma (DCP). The solute vaporization interference is negligible in ICP (Larson et al., 1975).

B. IONIZATION When atomic or ionic species of an element in a plasma emit their characteristic line radiation, any shift in the ratio between these two species causes a shift in the intensity of the atomic and ionic lines. Johnson et al. (1979b, 1980) reported the enhancement effect of Cs and Li on K, Na, Ba, Al, Cr, etc. in DCP. This enhancement effect is negligible in ICP when using recommended parameters for power input, observation height, and carrier-gas-flow rate (Larson et al., 1975).

c. UNWANTED RADIATION Unwanted radiation refers to radiation other than the analyte radiation reaching the analyte detector. In any emission system, the analyte signal consists of the

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 79

wanted analyte radiation and the unwanted radiation. The latter may be divided into the following categories (Ward and Myers, 1979): 1. Source background. 2. Extracting solution background. 3. Stray radiation in the spectrometer, 4. Spectral line or band interference.

Source background refers to the radiation originating, for example, from Ar. This background radiation is very stable in Ar plasmas. Extracting solution background refers to the continuum originating from the extracting solution. Stray-light radiation in some direct readers creates the most serious error in determining trace metals (USDCNTIS, 1977).The three main sources of stray light are (1) grating scatter; (2) reflections and scatter in the secondary optics, i.e., the region between the exit slits and photomultipliers; and (3) general scatter from reflections by internal surfaces of the direct readers. Grating scatter is due to grating imperfections and has been discussed by Larson el d.(1976). The degree of stray-light interference of the latter two types depends on the engineering design of the direct readers and could be reduced significantly by using nonreflective coatings, light traps, baffles, and general optical design (USDCNTIS, 1977). The spectral line or band interferences arise when there are spectral overlaps between the analyte and concomitant species. In some instances, concomitant species may elevate the intensity of the background continuum. The method used to correct for spectral overlaps, background elevations, and stray-light interferences follows.

D. CORRECTION FOR INTERFERENCES (ICP-AES) To correct for solute-vaporization effect the analyst should add a releasing agent to both sample and standards. For example, to reduce the effect of A1 on Ca, one may add Sr to both the sample and standard solutions. The Sr will combine with A1 and reduce its effect on Ca. To suppress ionization interference, an easily ionizable element such as Cs or Li is usually added to standard and sample solutions. These sample pretreatments are not necessary for ICP-AES. To correct for source and extracting solution background, the analyst will zero the spectrometerwith the blank solution made up of the extracting solution (blank correction). The interferences due to stray light, background elevation, and spectral overlaps could be corrected for if the blank and the sample solutions were identical in composition except for the analyte. This ideal solution is beyond the practical realm, especially when a multielement analysis is desired. However, if the samples are rather uniform in major interfering species, these could be added to the blank and to the standards to compensatefor their interference. But addition of ma-

80

PARVIZ N. SOLTANPOUR ETAL.

jor interfering species to the sample solutions prevents analysts from simultaneously determining these elements with other elements. This dilemma is resolved by using a scheme known as interelemental spectral interference correction. Interelemental interference is observed when the analyte detector (channel) receives signals from the interfering elements. When the soil-water extracts, NH,HCO,-DTPA extracts, dilute-acid extracts, and other extracts with low concentrationsof interfering elements are analyzed by ICP, the degree of interelemental interference is usually small. However, a soil analysis for total elemental content results in high concentrationsof interfering elements and correspondingly large interelemental interferences. In the latter case, one should determine significant interelemental interferences in sample solutions and correct for them. To determine the degree of interelemental interference, the spectrometershould be standardized, a pure solution of the interfering (affecting) elements aspirated, and the apparent concentration of the affected elements determined. When the sample is analyzed, the concentrationof the interfering elements is determined and the necessary corrections made on the apparent concentration of the affected elements. Computer programs are available for automatically correcting the interelemental interference (Dahlquist and Knoll, 1978). The following example is given to show the use of the interelemental interference correction method. A synthetic solution containing 1.O ppm of Pb read 3.66 ppm of Pb when Al, Fe, K, Ca, Na, Mg, Ti, Mn, and P were added to the Pb solution at 1500,500,400,200,200, 100,60,20,and 15 ppm concentrations, respectively. When pure solutions of these elements at the same concentrations were aspirated into the plasma, the apparent concentrations of Pb were 2.62, 0.164, and 0.038 for Al, Fe, and Ti solutions, respectively. Other elements did not produce any noticeable unwanted radiation at the Pb wavelength (220.3 nm). Subtracting the preceding interferencesfrom 3.66 gave a Pb value of 0.84 ppm, which is much closer to the true value of 1.O ppm than the uncorrected value of 3.66. In this case, the spectrometer was not restandardized before the high-background Pb solution was read. This may explain the reason for obtaining the reading 0.84 ppm instead of 1 .O ppm for Pb. However, when a solution containing all the previously mentioned element concentrationsexcept Pb was aspirated, it gave an apparent Pb concentration of 2.80 ppm, which is almost identical to the sum of the apparent Pb concentrations in Al, Fe, and Ti solutions. This example shows the validity of the interelemental interference correction method. Some precautions to be observed in interelemental spectral interference correction are discussed in the remainder of this section. Care must be taken to ensure that an adequate rinse is performed between the introduction of each interferant solution. Pure chemicals such as SpecPure reagents should be used to determine interlemental interferences. If the chemicals are not pure, an impurity of analyte in these chemicals will create rather large errors in the results. For soil HF-HCIO,

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 8 1 digests, the interference of major soil constituents on other elements should be determined. Soil extracts should also be examined for possible interelemental interferences. These interferencesare specific for a given instrument, depending on the wavelength used for each element and effective use of baffles and black interiors to reduce stray-light interferences. In the instrument (Jarrell-Ash Model 975 ICPAtomComp) used at CSUSTL, it was found that the following elements significantly interfere with some trace elements in HF-HClO, soil digests: Al, Fe, Mg, Mn, and Ti. However, one should be on guard against interferencefrom other elements that may be found in large quantities in contaminated soils. The interelemental spectral interferences found in the ICP system are shown in Table IV. In the instrument used at CSUSTL, Ca interference with other elements is very low as compared with published reports on another instrument (USDCNTIS, 1977). The order of interelement correction is important. An example given by Marciello and Ward (1978) is cited here. In the Jarrell-Ash 975 ICP AtomComp, the spectral bandpass of an exit slit is typically 0.03 nm. This means that the detector views a wavelength region of approximately 0.015 nm on each side of the analytical wavelength. Therefore, any elemental line that falls within this region increases the analyte signal. For example, Co emission at 238.892 nm is being monitored in a solution containing 100 ppm of Fe, 10 ppm of Co, and 1 ppm of element X. The Fe emits a line at 238.863 nm, which is practically within the bandpass of the Co exit slit. When a pure Fe solution containing 100 ppm of Fe was monitored, it produced a signal at the Co detector equivalent to 5.5 ppm of Co. Suppose that in the preceding example, in which 1 ppm of Fe affects Co by a factor of 0.055, Co also affects element X by a factor of 0.1. If the Co interference on X is corrected before the effect of Fe on Co is corrected, the X value will be 0.45 instead of 1 ppm. But, if the effect of Fe on Co is corrected first, then a value of 10 ppm of Co is used for correcting its effect on X,and a value of 1 ppm will be obtained for X. As a general rule, interelement correction should be programmed into the computer in the following order: 1. Major matrix components as interfering elements. 2. Order of magnitude of interference effect for minor matrix components.

When correcting for interelemental interference,remember that interference per unit concentration of the interfering element may not be linear. This has been demonstrated in the particular instrument used for the Ca interference on As, Se, Pb, and Sn (USDCNTIS, 1977). In this event, curves should be plotted showing the apparent concentration of the affected element as a function of the concentration of the affecting elements. These curves should then be used to correct for interelemental interferences.Computer programs and computers interfaced with the direct readers capable of performing these tasks will make the interelemental interference corrections easier and much faster.

Table IV

Examples of Some Interelemental Spectral Intereferences Observed in an ICP Spectrometer at CSUSTL" Wavelength (nm) Affecting element

Concentration (pglml)

324.7 Cu

206.2X2 213.6X2 Ni Zn

202.0 Mo

214.4 Cd

228.6 Co

267.7 Cr

407.7

Sr

249.1 B

455.4 Ba

220.3

Pb

253.6 Hg

193.6 As

196.0

-

Se

Apparent concentration of affected elements (pglml)

K

4006

-

Ca

50 2006 500 50 1006 200 2006 156

-

-

606

0.026

2ob 100

-

Mg N m

Na P Ti Mn Fe

0.007 0.012 0.024

0.009 0.012 0.038

-

-

-

0.006

,ooo

0.005 0.014 0.025

500 1.ooo 1,5006

0.002 0.002

0.088 0.176 0.261

0.046

0.295

0.080

5006

High background solution

-

0.002 0.01 1 0.048 0.093 0.002 0.013 0.016

1

A1

-

-

0.021

0.040

0.002 0.004 0.004 0.008 0.013

-

-

-

0.003 0.003 0.0 10

-

0.111

-

-

0.082 0.253 0.5 17 0.760

0.008 0.046 0.088 0.080 0.164 0.242

0.0 16 0.072 0.140

0.807

0.288

0.044

-

0.003 0.015 0.039

-

-

-

-

0.024 0.079 0.024 0.039 0.066

-

0.004

-

-

-

-

-

0.030 0.028

-

-

-

0.038

0.197

0.006 0.010

0.008 0.01 8

0.008 0.010 0.004 0.024 0.046 0.008 0.0 16 0.024

0.053 0.106 0.156

0.285 1.32 2.51 0.006 0.020 0.020

0.200

0.074

0.172

1.26

-

"Data from Soltanpour er al., 1982. hHigh background solution was made from the elements and concentrations marked by

h. All

-

0.018

-

-

-

0.064

-

0.012 0.018

0.164 0.302 0.866 1.76 2.62

0.023

2.80

0.004

0.006

solutions were 10% in HCI.

-

0.043 0.083

-

1.36 0.053 6.96 0.250 13.1 0.494 0.019 5.96 0.030 12.2 0.054 18.0 6.84

18.2

-

0.054 0.087 0.168

-

0.083 0.564 1.07 1.46 2.90 4.28 4.98

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 83 Another scheme for interference correction, called background correction by some, is to use a spectrum shifter. The spectrum shifter measures the background radiation close to the analyte exit slit. This radiation is assumed to originate from nonanalyte sources and to increase the radiation at the analyte wavelength. To correct for the interference, the average background radiation signal is subtracted from the signal observed at the analyte wavelength. This correction method may be used when the analyst is not aware of all the constituents of the sample or is using samples of varying matrices. This correction method is obviously not valid when the background radiation received at the analyte wavelength is grossly different from the one measured by the spectrum shifter. Hydride generators are used to concentrate the hydride-forming analytes and to eliminate the matrix effect (see Section VI1.E).

VI. ICP-MS INTERFERENCES In a particularly lucid explanation of interference effects in ICP-MS, Gregoire (1 989) broadly divided the subject of interference effects in the ICP-MS into three areas: isobaric interference, nonspectroscopic interference, and mass discrimination. The isobaric problems were further subdivided into two categories: molecular ion interferences and spectral interferences due to other elements and oxide species. The nonspectroscopic suppression effects were discussed in the context of space charge and ionization suppression effects. Mass discrimination effects were viewed in two categories: instrumental mass discrimination and matrix-dependent mass discrimination. To the discussion of nonspectroscopic interference will be added suppressions due to solids deposition on the skimmer cones, to solute vaporization, and to collisional-dependent de-excitation.

A. SOLIDSDEPOSITION ON SAMPLER AND SKIMMER CONES Deposition of solids on the skimmer and sampler can cause unwanted changes in the analytical response, i.e., reductions in the quantity (ion-arrival rate at the detector per unit analyte concentration in the test solutions) (Douglas and Kerr, 1988).At the DANR Analytical Lab, long runs involving analytical measurements on plant digests prepared using a microwave bomb technique (Sah and Miller, 1992) have been made using the ICP-MS instrumentation. During a run, coatings of calcium sulfate and oxide on the sampler and skimmer cones occur. Unless the dilution factors and sample nebulization times of the test solutions are carefully controlled, the solid depositions on the sampler and skimmer cones can result in serious suppression of analytical response. The coating of the skimmer cone near

84

PARVIZ N. SOLTANPOURETAL.

the tip, to the point of obscuring the orifice, causes the most serious decrease in analytical response; this effect persists independent of concomitant concentration in the test solution for the duration of the analytical run. The suppression can be at once eliminated by cleaning the skimmer cone.

B. NONSPECTROSCOPIC INTERFERENCES Nonspectroscopic interference is the general term adopted in ICP-MS for describing reduction in analytical response with increasing concomitant in the test solution. Nonspectroscopic interference is a complex issue, with several factors contributing to the suppression(s) observed. For application of ICP-MS to soils, the major element content of the solutions will vary widely and must be anticipated. Among the factors discussed below are solute vaporization, ionization suppression, space-charge, and collision-dependentde-excitation. Solute vaporization interference occurs in analytical atomic excitation-ionization sources in instances in which the solute does not have sufficient time and/or the source does not have sufficient energy to dissociate the solute before the analytical species moves into and through the region of observation. Typical manifestations of the interference are suppression and/or increased variability of the analyte signal as a function of increased concomitant concentration. For example, Johnson et al. (1979a) showed that A1 suppresses the Ca signal in a direct-current plasma. Winge et al. (1991) published high-speed photographs indicating that species generally associated with low spectroscopic temperatures can persist through the central channel of the ICP to enter the sampling cone of the ICP-MS and they discussed similar reports of unvaporized solvent doing the same. Ionization interferences can be noted for ICP-MS. Partially ionized elements in an ICP, e.g., Au and B, are susceptible to ion suppression from fully ionized interferants (greatereffective interferant ion-analyte ion molar ratio) and less effective in causing ion suppression for fully ionized analytes (smaller interferant ion-analyte ion molar ratio) (Gregoire, 1989). In ICP-MS measurements, ionization interferences cause suppression of analytical response (Houk and Thompson, 1988; Houk et al., 1981; Tan and Horlick, 1986). A prominent feature of easily ionized element concomitant interference is the trends that have been recognized in the ICP-MS data. For a given concomitant, the analytical response is suppressed more for the lighter atomic mass isotopes than for the heavier ones. For a given analyte, lighter atomic mass concomitants suppress the response less than do heavier ones. These trends are consistent with what would be expected from a space charge effect (Gregoire, 1989; Hieftje, 1992). Tracing the course of the ions from the point of ion production in the plasma, the ions that move through the sampler and skimmer, through the ion optics, and through the quadrupole are reflected away from the deflector and accelerated onto the detector. As the particle beam exits the plasma, it becomes increasingly more

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 85 positive as the electrons diffuse out of the beam more quickly than the positive ions; this process is called ambipolar diffusion (Ahearn, 1972). As a consequence of this process, columbic repulsion spreads the ion beam. The larger ions stay on course better than the lighter ones. Equivalently, the trajectories of the lighter ions are more affected by the ions of the heavier mass isotopes than the trajectories of the heavier ions are affected by the lighter mass isotopes. In addition to these factors, a mechanism is needed to account for observations made during analytical runs at the DANR Analytical Lab on the Perkin-Elmer SCIEX 500 hardware 5000 software ICP-MS instrument. For most ICP-MS work done at DANR, Be, In, and Bi are added to the test solutions immediately prior to nebulization as internal standard elements. On many occasions the analytical response on Be is suppressed more than on In, which in turn is suppressed more than on Bi for measurements made on sample solutions and those calibration standards to which Y was added versus calibration standards prepared without Y. This is consistent with the space-chargeinterference mechanism; the interfering elements are the Ca and Mg in the test solutions and the yttrium added to half of the analytical response standards and two of the three 0 ppb calibration solutions. However, approximately 40% of the time, the reverse is noted; i.e., Be is suppressed less than In, which is either suppressed about the same or less than Bi. Clearly, an additional mechanism is required to reconcile these observations with theory. Currently under consideration is a collisionally induced de-excitation mechanism. In a recent run, test-solution concomitants typical of plant extracts (Ca, Mg, K, S, and P) in a nutrient solution with sucrose and sequestrene (sodium femc ethylenediaminediL-hydroxyphenyl acetate) caused a two-fold reduction in the Be analytical response, a five-fold reduction in the In analytical response, and a six-fold reduction in the Bi analytical response. Suppression of production of analyte ions normally occurring as a result of Penning ionization is indicated here. The Penning ionization process is described as a neutral atom collision with metastable Ar to produce an ion and an electron from the atom and a ground state Ar atom from the metastable Ar. The de-excitation cross sections of metastable Ar, and equivalently the ionization cross sections of the collision partner, have been shown to be proportional to the polarizability of the collision partner (Bourene and Le CalvC, 1973). Thus, the trend in the suppression can be explained by de-excitation of the Ar metastable population by interaction with the concomitant in the sample (sucrose, sequestrene, Ca, Mg, K, S, and/or P) or Y in the spiked calibration standards, resulting in fewer ions of the larger, more polarizable analyte and/or internal standard atoms (e.g., In and Bi). Penning ionization collisional processes with Ar metastable species are responsible. For ICP-MS the IRZ is the position in which the sampler cone orifice is placed in the plasma. The NAZ must be closer to the tip of the IRZ for ICP-MS than for ICP-AES to permit an analytically useful population of ions to be observed (Winge et al., 1991). However, to avoid arcing between the load coil and the sampling cone, higher flow rates of the Ar stream cawing the sample aerosol and the aux-

86

PARVIZ N. SOLTANPOUR ETAL.

iliary Ar are used to push the IRZ tip away from the load coil (Fig. 4). This pushes the IRZ tip to a point away from that which would be required if the ions were taken from a region of approximate LTE between ionization, excitation, and gas energy. Thus, measurements using the ICP-MS are subject to a higher degree of collision-dependent de-excitation of metastable-state Ar (Bourene and Le Calve, 1973) as well as collision-dependent de-ionization between ions and electrons, where the electrons are provided by concomitant atom ionization (Beauchemin, 1989; Houk and Thompson, 1988; Houk, 1986; Beauchemin et al., 1987; Douglas and Kerr, 1988; Gregoire, 1987a). In practice, separating the nonspectroscopic interferences is difficult. These interferences can be compensated for by matrix matching, but this seriously limits the range of concomitant-level variability between test solutions within the run. Another approach used successfully by many is the internal standard calibration method, discussed later.

C. MASSDISCRIMINATION Gregoire (1989) defines mass discrimination as bias in ion transmission to the detector the magnitude of which is dependent on the mass of the analytical isotope. Furthermore, the effects can be divided into two categories, depending on the origin of the mass bias. The first category is called the instrumental mass discrimination effects, which are interference effects caused by mass discrimination occurring at the interface (sample and skimmer cones), ion lenses, quadrupole mass filter, and detector. Correction factors for instrumental mass discrimination are normally found by comparing measured isotope ratios to the known isotope ratio for a substance of known or certified isotopic composition and applying the correction to the samples run during the same time. Instrumental mass discrimination can range from 50% per dalton for light elements to 2%per dalton for heavy elements. The other type of mass discrimination results from the presence of concomitant elements, has been reported only twice, and effects only Li and B. Briefly, the magnitude of the effect is dependent on five factors: Absolute mass of the analyte. Degree of ionization of the analyte. Difference in mass between the two isotopes. Mass of the concomitant. 5 . Degree of ionization of the concomitant.

1. 2. 3. 4.

These factors are very similar to the space-charge interpretation of analyte-response suppression, so for all practical purposes matrix-dependent mass discrimination can be considered a special case of the more general space-charge effect; for more details, see Gregoire (1989).

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 87

D. UNWANTEDIONS Unwanted ions refer to that part of the ion beam other than the analyte ion part reaching the detector. In a mass spectrometer system, particularly one operated with peak widths close to 1 dalton, the analyte signal may be accompanied by unwanted ions (Table 11). These unwanted ions occur as a consequence of several factors (Gray, 1985; Date and Gray, 1989; Gregoire, 1989; Vaughan and Horlick, 1986; Munro et al., 1986; Date et al., 1987; Gray, 1989; Tan and Horlick, 1986), including the following: 1. Elemental ions (NM+)of the same unit mass and charge as the analyte isotope (NA+). 2. Element hydride (N{N-'MIH}+)molecular ions of the same unit mass and charge as the analyte isotope (NA+). 3. Element oxide (N{N-'6M'60)+)molecular ions of the same unit mass and charge as the analyte isotope (NA+). 4. Element hydroxide (N{ N-17M'60'H}+)molecular ions of the same unit mass and charge as the analyte isotope (NA+). 5. Elemental (2NM++)ions that are doubly charged and twice the unit mass of the analyte isotope (NA+). 6. Elemental argide (N{ N-36M36Ar]+, N ( N-38M38Ar}+,N{ N40M40Ar}+)molecular ions of the same unit mass and charge as the analyte isotope (NA+). 7. Elemental hydrogen argide (N{N-37M'H36Ar}+ , { -39M1H38Ar}+, ( N41M'H40Ar}+) ions of the same unit mass and charge as the analyte isotope (NA').

Many examples of these generalized isobaric problems are cited in the "Comments" column of Table 11.

E. METHODS OF CORRECTION FOR INTERFERENCES (ICP-MS) The two predominant types of interference in ICP-MS work, spectral and nonspectroscopic, can generally be corrected to yield measurements that are within 10% of the true concentration under most conditions. Several assumptions accompany this statement: ( 1) the analyte is present at a concentration level 10 times higher than the detection limit, (2) the spectroscopic interference is no more than half the gross response at the mass-charge ( d e ) ratio of the analyte, (3) nonspectroscopic interferences suppress and/or enhance the signal by no more than about a factor of two, (4) an internal standard can be found with ionization characteristics similar to those of the analyte, and/or (5) the method of internal standardization in use accurately accounts for changes in analyte gain with changes in concomitant level.

PARVIZ N. SOLTANPOUR ETAL. Isobaric (spectral) interferencecorrections to the data may be required for both quantitativeanalysis and isotope-ratiomeasurements. These are usually performed before the corrections for nonspectroscopic interferences. Isotope-ratio measurements usually do not require correction for nonspectroscopic interference other than a multiplicative constant to correct for detector-response changes as a function of isotope mass and/or matrix-dependent mass discrimination effects. Correction of quantitative concentration determination data for nonspectroscopic interferences is almost always required. Application of spectroscopic interference correction is done as a last resort to ICP-MS work. Using an analytical isotope that is free of spectral interference is always desirable; and if high resolution is available, identifying an isotope free of spectral overlap becomes much more probable (see Appendix) than if a quadrupole spectrometer with unit resolution is used. If an analytical isotope that is free of spectral interference cannot be located, then calibration-subtraction type corrections can be applied to the measurements, an extraction step can be used to separate the analytical elements from the sample concomitant element(s), or another, more suitable analytical method can be used. Calibration-subtraction procedures similar to those discussed for ICP-AES can be used for isobaric-interferencecorrection if the concentration of the interfering species can be determined using measurements on an alternate mass. However, if the interfering species concentration cannot be determined, a valid correction may still be possible if the interfering species can be monitored at more than one (massto-charge) spectral position. For example, using a unit-resolution spectrometer, corrections due to the presence of an unspecified concentration of CI in test solutions affecting the As concentrations may be necessary. The presence of sufficiently elevated C1 concentrations in solutions can produce 75{35C140Ar)+ions in the spectrum; presenting an isobaric interference on As concentrationsdetermined using the only naturally occurring isotope of As at 75 daltons. To correct for the effect, the apparent Se concentration is measured using both the 82Se+isotope and the 77Se+isotope. The Se concentration measurement at 82 daltons is subtracted from the apparent Se concentration at 77 daltons to determine the apparent concentration of Se due to 77{ 37C140Ar)+ions. As a first approximation, the difference in the apparent Se concentrations can be multiplied by 7.58 (the naturally occurring abundance of 77Se)and by 75.53 (the naturally occumng abundance of 35Cl)and divided by 100 (the naturally occurring abundance of 75As)and by 24.47 (the naturally occurring abundance of 37Cl)to arrive at a term that is subtracted from the apparent As concentrations. More exact correction procedures include spiking a test solution with a small volume of perchloric acid in the absence of both Se and As and experimentally determining the multiplicative correction factor used to multiply the net apparent Se concentration to arrive at the apparent As concentration that is subse-

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 89 quently subtracted from the gross As concentration measurement. The CIAP correction factor is determined by subtracting the apparent concentration of Se at 82 daltons made while nebulizing the synthetic test solution containing perchloric acid from the corresponding apparent concentration at 77 daltons and dividing the difference into the apparent concentration measured for As. Then, for the samples, the Se concentrations measured using the 82 dalton mass is subtracted from the apparent Se concentration measured using the 77 dalton mass; this difference is multiplied by the CIAr+correction factor. The product is then subtracted from the apparent As concentration measurement. Corrections for nonspectroscopic interferencesare done using the internal standardization method (Thompson and Houk, 1987). The basis for the method is formed by adding a constant concentration of a nonanalyte isotope to all the test solutions, calibration standards, analytical blanks, and sample solutions. In addition to the requirement that the internal standard isotope not be an analyte, other desirable characteristicsof the internal standard include a negligible concentration of the internal standard isotope in the samples other than what is added, ionization behavior similar to the analyte, and absence of significant isobaric interference problems on the mass of the internal standard isotope. The internal standardization procedure in its simplest form corrects analytical concentration measurements by the multiplicative factor (the internal standard response measured during nebulization of the calibration standard divided by the internal standard response measured on the current sample). If more than one calibration standard is run, then the analyte calibration curve is constructed using a plot of the ratio of the analyte response divided by the internal standard reading on the ordinate (y-axis) versus the known analyte concentration on the abscissa (xaxis). The analyte concentration in the test solutions is then found by applying the ratio (the analyte response divided by the internal standard response) to the calibration curve. Generally, the accuracy of the analytical results is improved by subtracting the capability of a blank response from the gross analytical response before taking the ratio and by the capability to select one of three or more possible internal standards isotopes, all of which must be added to the test solutions either before the analytical run or at least prior to the point when the test solution is introduced into the nebulizer. Variations on the internal standard method include simultaneously applying measurements performed on two internal standards to analyte response measurements that are weighted according to atomic mass (Doherty, 1989) and using linear-regression statistical models to predict individual analyte-gain factors as a function of one-three response signals from internal standards (Johnson el al., 1992a,b).In the latter two studies, Y was added to half the standards and two of three solutions that were used to determine the 0 ppb level. The concentration of Y added to the calibration standard solutions was in the

90

PARVIZ N. SOLTANPOUR ETAL.

100-500 pg ml- range and was added to induce a suppression of not only the internal standard response but also the analyte response per unit concentration (i.e., analytical gain). After fitting the 0 ppb concentration responses (dependent variables) to the corresponding internal standard measurements (independent variables) made on the 0 ppb solutions using linear-regression statistics, estimates of the appropriate response to subtract from the individual analyte measurements in the sample solutions were determined from the internal standard responses on each sample solution. These were subtracted from the gross analyte responses measured on the sample solutions, and the net response was used to determine the analyte concentration. To do this, a regression model was developed using net analyte response per unit analyte concentration as the dependent variable (analyte gain, plotted on the y-axis) and the response of the internal standard(s) as the independent variable(s) (plotted on the x-axis in the case of one internal standard). Then, plugging the response of the internal standard@)measured on the sample solution into the regression equation, an analytical gain was determined. The gain was divided into the net analyte response to determine the analyte concentration. From one to three internal standards could be used; if more than one are used, then the corrections for internal standard response could be applied sequentially or simultaneously. If they are applied simultaneously, then three internal standard measurements with interaction between the three internal standard responses in a second-order linear-regression format and an intercept term required for inversion of a 10 X 10 matrix could be used to determine the regression coefficients appearing in the calibration equation. It was found that the accuracy of the measured versus the true concentrations of the calibration standards improved as the complexity of the regression models increased; in addition, some analyte concentrations were more highly correlated with one or two of the internal standards than were the other(s), but not all analytes were correlated to the same extent with the same internal standard(s). In another experiment, one set of calibration standardization solutions was prepared with Ca, additions, and another set was prepared with Y addition. The concomitant was spiked into half the calibration standards and two-thirds of the 0 ppb concentration solutions. Also, spike-no spike sample-solution pairs were used to calibrate the analytical gains as a function of the response of the internal standard(s). The samples were naturally high in Ca concentration because they were plant-material digests prepared using a microwave digestion method (Sahand Miller, 1992). It was found that analytical accuracy as determined by spike recovery was improved if the analytical gains were predicted using the regression models found with spike-no spike sample-solution pairs and the Ca-no Ca containing calibration standards rather than the regression models determined using the spike-no spike sample-solution pairs and the Y-no Y containing calibration standards. The method of correction selected for internal standardization depends on the analytical objectives of the study. Generally, 2 10% is sufficiently accurate; in which case, the software supplied by the instrument manufacturer is adequate.

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 9 1

VII. PRACTICAL APPLICATIONS A. GRINDINGSOILSAMPLES Soil samples should be air-dried as soon as possible after sampling (drying and grinding of soil samples is not recommended for Mn and Cr). Soil samples may contain clods or large aggregates, which must be ground or crushed to reduce subsampling error. Many laboratories use automatic grinders to crush soil. Studies have shown that the amount of extractable micronutrients from soils is affected largely by the degree of grinding (Soltanpour ef al., 1976,1979a); therefore, grinding variables such as force and time should be standardized. When soils are ground for extractable microelements, care must be taken to avoid excessive grinding. It is important to use grinders that do not contaminate the soil. At CSUSTL, a highdensity aluminum oxide auger made by the Coors Porcelain Co., Golden, Colorado, is attached to a Nasco-Asplin automatic grinder. The grinder is equipped with a 2 mm stainless-steel sieve. This grinder minimizes the degree of soil contamination with trace elements. In soil analysis, passing the soil through a 2 mm sieve after mild grinding is a standard procedure; then all analytical results are based on a 2 mm soil. For total elemental analysis, the 2 mm soil may be further ground so that all of it passes through a 100-mesh PVC sieve.

B. OBTAINING Son. EXTRACTS For simultaneous multielement determination, single-element extraction solutions are not useful. Therefore, Soltanpour and Schwab (1977) developed a 1M NH,HCO,-O.OOSM DTPA (AB-DTPA) solution for simultaneous extraction of P, K, Zn, Fe, Cu, Mn, and nitrates from soils. This test was modified by Soltanpour and Workman (1979) to omit carbon black, which sometimes contaminated the sample and adsorbed metal chelates. The above test is routinely used by CSUSTL to assess soil fertility of Colorado farms. After extraction, ICP-AES is used to simultaneously analyze these extracts for P, K, Zn, Fe, Cu, and Mn. Experience has shown that AB-DTPA solution should be acidified to eliminate the carbonate-bicarbonate matrix in order to prevent clogging of the capillary tip (Soltanpour ef al., 1979b). However, using high-salt nebulizers (Legere and Burgener, 1985) has obviated the use of acid pretreatment (Soltanpour, 1991 ). Soil-water extracts and DTPA extracts (Lindsay and Norvell, 1978) can be analyzed by ICP-AES. We are analyzing the soil saturation extracts simultaneously for Ca, Mg, Na, and K and then calculating the Na absorption ratio. Plant digests are also analyzed by ICP-AES. When analyzing mine overburden and mine spoil materials to determine their potential toxicity to plants and consumers, CSUSTL and other environmental labs

92

PARVIZ N. SOLTANPOUR ETAL.

in the western United States use AB-DTPA and ICP-AES to screen for P, Zn, Cu, Mn, B, Cd, As, Se, Mo, Pb, Ni, and other elements (Soltanpour, 1991). The ABDTPA extract is low in Ca, Mg, Al, Fe, and Mn, which cause interelemental interference, and, therefore, it is well adapted to ICP-AES analysis. Obviously, water extracts are ideal for ICP-AES determinations, but the concentrations of some elements in water extracts are below the ICP-AES detection limits. Jones (1977) found Ca, K, Mg, and P simultaneously in the double-acid extracts of Georgia soils. The type of vessel and shaker and the speed of the shaker may affect the amount of some extractable elements, but their effect is small compared with the grinding variables (Soltanpour et al., 1976). To make a 1M NH4HCO,-O.OO5M DTPA solution, add 1.87 g of DTPA to 800 ml of distilled-deionized water (DDW). Add approximately 2 ml of 1:1 NH,OH to facilitate dissolution and to prevent effervescence. Shake until most of the DTPA is dissolved. Then add 79.06 g of NH,HCO,, and stir gently until dissolved. Adjust the pH to 7.6 with NH,OH. Dilute the solution to I .O liter with DDW. The pH of the solution is unstable; if the solution is stored under about 3 cm of mineral oil, the pH remains stable. However, using a fresh solution is preferable. Put 10 g of a 2-mm soil into a 125-ml conical flask. Add 20 ml of AB-DTPA solution. Shake on the reciprocal shaker for 15 min at I80 cycles/min with flasks kept open. Filter the extracts through Whatman no. 42 filter paper or its equivalent. Take a 2-ml aliquot of the extract, and add 0.25 ml of concentrated HNO,. Shake for 10 min to eliminate the carbonate-bicarbonate matrix and to prevent clogging of the capillary tip in cross-flow or concentric nebulizers. This solution is now ready for simultaneous multielement determination. With high-salt nebulizers (Babington type) the acid pretreatment is not necessary.

C. DIGESTION OF ORGANIC MATTERAND DISSOLUTION OF SILICATES FOR TOTAL ELEMENTAL ANALYSIS Digestion (oxidation) of organic matter and dissolution of silicates are necessary steps for bringing all elements into solution. In this article we refer to these processes as digestion. Certainly, it is advantageous to use methods that yield themselves to multisample rather than single-sample digestion. Another important consideration in choosing a method is preserving the easily volatilized elements in the sample. Fusion and other high-temperature methods of digestion result in losing volatile substances such as As, Se, Sb, and Hg. The following method avoids the loss of As, Se, and Sb in the presence of silicates (Bajo, 1978); therefore, we recommend this method. Other methods, such as HF digestion of siliceous material in capped polyethylene bottles, may be used (Odegard, 1979; Langmyhr and Paus, 1968).An all-Teflon bomb (Bernas, 1968; Lechler and Leininger, 1979) has been used for analysis of siliceous material.

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 93

Put 1 .O g of the 100-mesh (0.15-mm) soil into a 100-ml Teflon beaker. Add 10 ml of HNO, and 10 ml of HCIO,. Cover with a Teflon watch cover, and heat at 200°C for 1 hour under a perchloric hood. Remove the cover and continue heating until the volume is 2-3 ml. Cool the sample, add 5 ml of HClO, and 10 ml of HF, cover with a Teflon watch cover, and heat overnight at 200°C. Overnight digestion is for convenience,but digestion may be terminated as soon as all siliceous material has been dissolved. Remove the cover, and continue heating until the volume is 2-3 ml. Cool the digest, add 10 ml of 50% HCI, cover, and heat at 100°C for 30 min. Remove the digest from the hot plate, and allow it to cool. Transfer the solution quantitatively into a 50-ml volumetric flask, and bring to volume. The solution is then ready for ICP determinations. The reason a 100-mesh (0.15 mm) soil sample, instead of a 2 mm sample, is used in digestion is to speed up the breakdown of silicates.At CSUSTL, 2 mm soil samples have been digested with no difficulty when soils were digested overnight in the presence of HF-HCIO,. The perchloric hood should be washed periodically to remove perchlorates and to avoid the danger of explosion. Do not let HClO, solution get dry; anhydrous HClO, is explosive.

D. ANALYSISOF SOILEXTRACTS AND DIGESTS Several different after-manufacture add-ons are available for sample aerosol production andor introduction for ICP-AES and ICP-MS systems. These include USN, DIN, HPLC systems (Braverman, 1992), flow-injection (FI) accessory (Thompsonand Houk, 1986;Dean et al., 1988;Denoyer and Stroh, 1992;Denoyer et al., 1991a), hydride-generation equipment (Workman and Soltanpour, 1980), ETV accessory (Gregoire, 1989), and laser-ablation solid-sampling equipment (Denoyer, 1991; Hager, 1989; Abell, 1991; Denoyer et al., 1991b; Pearce et al., 1992). One of the most versatile add-ons is the FI accessory. There are many possible physical and chemical procedures that can be combined with the ICP-MS and ICPAES through the FI accessory, including online dilution, isotope dilution, standard additions, hydride generation, cation exchange, anion exchange, and electrothermal vaporization. For direct analysis of solutions, the FI involves introducing a discrete sample aliquot into a flowing carrier stream. Sample volumes for a typi, to 1-3 cal ICP-MS-FI analytical measurement range from 50 to 500 ~ 1 compared ml for continuous solution aspiration. Full mass scans can be performed on a quadrupole in about 100 ms, or 10 scans per second. The high scanning speed of the quadrupole allows transient signals generated by flow injection to be captured and measured. Relative standard deviations are 2-5% on replicate injections. Rinse-out times are considerably shorter for FI-equipped systems than for inter-

94

PARVIZ N. SOLTANPOUR ET AL.

mittent nebulization. An impressive advantage of FI is that it allows analysis of solutions containing approximately 50 times more concomitant than conventional nebulization equipment for comparable reductions in analytical response due to nonspectroscopic matrix-dependent interferences, putting the acceptable level of dissolved solids in the test solutions in the 0.5 g per 100 ml to 2.5 g per 100 ml range. Using the FI accessory also reportedly prevents clogging of sampler and skimmer cones while reducing the amount of solids deposited on them per sample solution. Four-fold increases in sample throughput per unit time have been reported using FI (Dean et al., 1988).

1. Preparation of Stock Standard Solutions Any emission-spectrometric method compares the emission signals from the sample solutions with those of the standard solutions to determine the composition of the sample solution. Therefore, extreme care must be taken when preparing standard solutions-it is recommended that Specpure reagents be used. All acids used for dissolution should be of high purity, such as Hi-Pure or Ultrex grade. Water used for dilution should be distilled-deionized. Table V, adapted from Ward 1978b, can be used to prepare solutions containing 1000 ppm of an element.

2. StandardizationProcedures In calibrating the ICP spectrometer, one should consider the concentration range to be used, the interelement interference correction, and the stability of the standards. Avoid mixtures of chemicals that cause precipitation. McQuaker et al. (1979) devised a calibration scheme for 30 elements that satisfies the needs of researchers in soil, water, tissue, and particulate-matter analysis. In soil analysis, one set of secondary standards is required for each extracting solution and one for the total soil digest. When preparing a multielement standard solution, avoid preparing those containing high concentrations of affecting and low concentrations of affected elements (see Table IV). Interelemental effects are measured by using single-element solutions prepared with Specpure chemicals. Appropriate computer software is used to correct for interelemental effects. In case of nonlinear interference, computer software should be able to store correction curves for interelemental corrections. Secondary standards should be made in such a way that standard solutions match the sample solutions in concentration of acids. For AB-DTPA extracts, standards should be made in 1M NH4HCO,-O.OO5M DTPA solution that has been neutralized with concentrated HNO, (Soltanpour and Workman, 1981). The HNO, neutralization is not necessary if a high-salt nebulizer is used. For HN0,-HC10,HF digests, standard solution should be made in solutions that contain 5 % (vol/vol) HClO, and 10% (vol/vol) HCl.

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 95 Table V Primary Standard Solution Preparation"," Element

Compound

Weight (8)

Al

Al AICI,*6H,O Sb SbCI, As As,O, BaCI,' BaCO,' BaNO, Be Be0 Bi

1.m 8.948 1 1.m 1A736 1.m 1.3203 1.1516 1.4369 1.9029 1.m 2.7753 1.OoOo 1.1149 2.32 1 I 1.OoOo 5.7195 1.oooo I . I423 2.4972 5.8920

Sb As Ba

Be Bi

B Cd Ca Cr co

cu In Fe Pb

Li Mg Mn Hg Mo Ni

Bi20, Bi(NO,),.SH,O B H,BO, Cd CdO CaCO, Ca(N0,);4H,0r Cr CrC1,(6H,O) co CoCI;6H20

4.0373

cu

1.m

CUO In Fe %O, Pb PbO Pb(NO,), Li,CO, LiCl MgO MgC1,~6H,O' Mn MnO, HgCI,

1.2518 1.0000 I .m 1.4297 1.0772 2.6758 5.8241 6.1092 1.6581 8.3625 I .m 1.5825 1.3535

Mo MOO, Ni

1.5003 1.oooo

1.oooo

5.1244 1

.ooo

I .m

1

.oooo

Solvent 6M HCI IM HCI Aqua regia IM HCI 4M HNO, 4M HCI Water 0.05M HNO, Water 0.5M HCI 0.5M HC1 4M HNO, 4M HNO, 1M HNO, 4M HNO, Water 4M HNO, 4M HNO, 0.5M HNO, Water 4M HCI Water 4M HCI Water 4M HNO, 4M HNO, Aqua regia 4M HCI 4M HCI 4M HNO, 4M HNO, Water IM HCI Water 0.5M HCI Water 4M HNO, 4M HNO, Water + Ig (~,),S,OL7 Aqua regia Aqua regia 4M HCI continues

PARVIZ N. SOLTANPOUR ETAL.

96 Table V-continued

Element

Compound

Weight (g)

NiO NiCI2.6H,O

1.2725 4.0489 1.4305

Nb

Nb205

K

NaH,PO, NH,H,PO, KCI K2C03

Se Si Ag Na Sr

Te TI Sn Ti

V Zn

SeO, Na,Si0,.9H20r Ag Ag,O NaCl Na,CO,

srco, Sr(NO,), TeO,

TlCl Sn SnC1;2H20 Ti V Zn ZnO Zn(N0,),-6H20

3.8735 3.7137 1.9067 1.7673 1.4053 10.1190 1.m 1.0742 2.5421 2.3051 I .6849 2.4152 1.2508 1.1174 1.1735 1.m 1.9010 1 .m 1.m

1.m

I .2@8 4.5506

Solvent 4M HCI Water Minimum quantity of HF, add IM HCI Water Water Water IMHCI Water Water 4M HNO, 4M HNO, Water 1M HCl I M HNO, Water 4M HCI 4M HCI Water 4M HCI 4M HCI 4M HCI 4M HNO, 4M HNO, 4M HNO, Water

“Adapted from Ward, 1978b. bUse 100-150 ml of solvent to dissolve and bring to a liter volume to give a concentration of loo0 ppm of element. ‘Not Specpure materials.

3. Comments The following operational parameters have been used with the ICP spectrometer (Jarrell-Ash 975 AtomComp) at the CSUSTL: sample flow rate, 0.5 ml/min; Ar pressure, 690 kPa (100 lb/inch2);aerosol carrier Ar flow rate, 1 litedmin; Ar plasma support flow rate, 19 literjmin when creating the plasma; nebulizer types, Legere high-solid nebulizer (Babington type); height of observation above coil, 15 mm;incident power, 1.25 k W , reflected power, <10 W. The operational procedure of different ICP spectrometers is beyond the scope of this article. When a particular ICP spectrometeris purchased, usually the manu-

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 97

facturers will train the analysts in operational procedures and in solving any operational problems that may arise.

E. DETERMINATION OF TRACE LEVELS OF As, Se, AND Hg USING THE HYDRIDE-MERCURY VAPORGENERATOR The hydride-mercury vapor generator (Fig. 5) is used to determine trace levels of As, Se, and Hg in soil extracts and total soil digests (Workman and Soltanpour, 1980).Pneumatic nebulization is not useful when levels of these elements are lower than 100 pghter. In contrast, the hydride-mercury vapor generator will enable the analyst to quantitatively measure levels as low as 1.O pghiter. Soil extracts must be pretreated before introduction into the hydride generator. The oxidation state of the analyte is critical. Selenium'" is readily reduced to the hydride, but SeV1is not. Arsenic"' is more readily reduced than As". Also, organic constituents in the extracts interfere with hydride generation and should be destroyed. At present, no pretreatment procedure has been found to simultaneously pretreat extracts for Se and As analysis. The Se pretreatment is not effective in reducing As" to As"', and the As pretreatment reduces Se to the metal form, which will not form the hydride. The total digest solutions resulting from the procedure given in Section VI1.C should contain only Se'" so that no pretreatment for Se analysis is needed. However, for As determinations the digest solution must be pretreated. Mercury may be determined in solutions pretreated for Se or As analysis. It is

NaBH, solution t

t 2-m

I+ Sample or 0.5N HCI

To

plasma

t

t

60 ml coarsefritted disc

Buchner funnel

II

Drain

Pumps

t

P&on 1llmin

Figure 5 The hydride-mercury vapor generator.

98

PARVIZ N. SOLTANPOUR ETAL.

recommended that spike recovery studies be performed on the particular type of solution to be analyzed to verify the accuracy of the technique.

1. Description of Generator The system (Fig. 5) consists of three peristaltic pumps and a gas-liquid separator. PumpA(head no. 7014,5.3 ml/min flow rate) (Curtin Matheson Scientific, Inc., Denver, Colorado) pumps the NaBH, (in 1 liter of DDW, first dissolve 0.5 g of NaOH, then 2.0 g of NaBH,) solution continuously. Pump B (head no. 7016, 20 ml/min flow rate) pumps either 30 sec ( 10 ml) aliquots of pretreated sample or 30 sec of 0.5 N HCl rinse between samples. Pump C (head no. 7015,42 ml/min flow rate) drains the gas-liquid separator. Argon is forced through the base of a 60 ml coarse-fritted disc Biichner funnel to strip the gases from the liquid and to carry them into the plasma. A signal is produced whenever a detectable amount of analyte is introduced into the plasma. Ten-second integrations of the maximum signal are used to calculate concentrations. To begin the integration at the proper time, a chart recorder is useful initially to observe the signal timing. One sample per minute is usually analyzed unless a very high concentration is encountered and a long rinse period becomes necessary. Total Se determination in highly contaminated soils may need special attention. Gary Banuelos'O noticed copperlike precipitates, after which Se determinations were not reproducible. The NaBH, might have been used up by elements other than Se. Banuelos recommended sample dilution.

2. Pretreatment of Soil Extracts to Be Analyzed for Se The following procedure is used to reduce SeV' to SeIVand to destroy organic constituents in aqueous solutions before generating hydrides. Place 15 ml of extracting solution into a 50 ml digestion tube. Add 1 ml of fresh 30% H,O,, and heat for 20 min in a boiling-water bath. Add 10 ml of concentrated HCl, and heat again for 20 min. Cool the solution and bring it to a final volume of 25 ml with DDW. The solution is now ready to be analyzed for Se using the hydride-mercury vapor generator. The solution is approximately 4.8 N HCI. Standards should contain the same acid concentrations.

3. Pretreatment of Soil Extracts or Total Soil Digests to Be Analyzed for As The following procedure is used to reduce AsV to As"' before generating hydrides. "Gary Banuelos, U.S. Dept. ofAgric., Water Management Research Lab., 2021 South Peach Ave., Fresno, CA 93727-595 I , pers. comm.

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 99 Add 3 ml of potassium iodide-ascorbic acid solution (in 100 ml of DDW, dissolve 10 g of KI and 1 g of ascorbic acid) to 10 ml of either soil extracts previously treated using the procedure given in Section V1I.B. or total soil digests from the procedure given in Section V1I.C. Wait at least 10 min. The solution is now ready for analysis using the hydride-mercury vapor generator. Standards should contain the same acid concentration and be treated in the same way as the samples.

VIII. QUALITY CONTROL METHODS In ICP-AES and ICP-MS, as in other methods of analysis, quality control should be an integral part of the procedures used for analysis. The quality of the analytical work may be checked in several different ways. Use of blind duplicates, check sample, standard reference materials, recovery of added elements, and interlaboratory comparisons are recommended. Blind duplicates are samples that are introduced into the laboratory by the laboratory supervisor. The analysts analyze these samples without any knowledge of the previous analytical results. A check sample is a sample that has been analyzed many times by different laboratories and is available in large quantities. This sample is analyzed with every batch of samples to see if gross contamination or error has occurred. Standard reference materials are materials that have certified analytical values, such as National Institute of Standards and Technology samples. Canadian Land Resource Institute, Agriculture Canada, and the Canada Center for Mineral and Energy Technology have prepared a few reference soil samples with recommended values for several elements. One of the best ways to evaluate analytical results is to have them analyzed by several reputable laboratories and then to compare the values. Another method of quality control is to add known amounts of elements to soil digests or extracts and to determine the percentage recovery of the added element. For a discussion of quality control methods, see Skogerboe and Koirtyohann (1976). In the case of trace metal analysis, the best guard against poor results is maintaining a high standard of cleanliness throughout the laboratory.

Ix.SUMMARY In this article we have discussed ICP-AES and ICP-MS. Our experience and that of others using ICP spectrometry for routine analysis show that ICP is a generally superior method because of its accuracy, precision, detection limit, freedom from

100

PARVIZ N. SOLTANPOUR ETAL.

interferences, and dynamic range. Analysis is facilitated by automatic samplers, large computers, and the necessary software. The ICP-AES can analyze a solution for many elements in 1 min; thus, large amounts of data can be generated quickly. Isaac and Johnson (1983)found that by using ICP-AES one technician is able to do the work that previously required four. Data handling and processing should be carefully considered. Interfacing the instrument with more powerful computers is a must if large volumes of data are anticipated. It is recommended that users of ICP-AES and ICP-MS subscribe to the ZCP-Znfomation Newsletter (Department of Chemistry, University of Massachusetts, Amherst) and other newsletters available from manufacturers to keep abreast of new developments in ICP-AES and ICP-MS. Journals such as Applied Spectroscopy, Analytical Chemistry, Analytical Chimica Acta, and others referred to in this article are also good sources of information. Additional sources of information have been described in the text.

APPENDIX: EXAMPLE OF ISOTOPE SELECTION IN PLANT TISSUE In the following example we outline the analytical isotope selection process used to determine total sample concentrations for Ca, Fe, Ni, Zn, and Pb in plant tissue digests. Assume that the plant digests were prepared using nitric acid and hydrogen peroxide to digest the plant material in closed Teflon digestion bombs as described by Sah and Miller (1992). The isotope selection process is illustrated for two ICP mass spectrometers-one with normal mass resolution and the other with a high resolution (HR) instrument, with resolution of 50,000(Tsumura and Yamasaki, 1991; Bradshaw et al., 1989).

CALCIUMISOTOPE SELECTION Case 1: Normal Resolution

For Ca, a glance at Table I1 reveals several anticipated difficulties in identifying a Ca isotope that will be analytically useful. The most abundant isotope, 40Ca+, cannot be used because of an overwhelming abundance of 40Ar+in the ion source. Degradation and recombination products of the common atmospheric gases N,, 0,, and CO, reduce the analytical utility of 42Ca+,43Ca+,44Ca+,and 46Ca+.Although the abundance of the 48Caisotope is only 0.18596, it is selected as the analytical isotope for Ca concentration estimates. The 48Ca isotope is the same unit

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 101 mass as the 4x[31P'H160}+ molecule (Table V). Thus, a systematic study of the phosphorous content range expected for the sample solutions and the maximum response due to this concentration at 48 daltons are necessary. The maximum response must include a predigestion spike of the expected phosphorous in the typical samples, because the oxide levels are thought to change from one sample to another (Gregoire, 1989; Vaughan and Horlick, 1986). If the contribution by 48[ 31P1H160)+ is found to be significant with respect to the gross response at 48 daltons, then a decision based on the outcome of the predigest spike study will determine if adding the 31P+ion to the list of selected isotopes and performing the appropriate calibration-subtraction corrections will or will not work. If not, measures must be taken to reduce the amount of 48[ 31P1H160}+ in the mass spectrum, including sample aerosol processing (Munro et al., 1986; Lam and McLaren, 1990; Evans and Ebdon, 1989) or aerosol desolvation based on evaporation-condensation systems (Veillon and Marghoshes, 1968).In addition, the most abundant isotope of Ti overlaps the 48 Ca+ at unit resolution. Although the Ti concentration in plants is typically thousands of times less than the Ca concentration, the effect of the Ti ion must be taken into account to accurately determine the Ca concentration in these samples. In fact, for every unit of Ti concentration in the sample extracts, approximately 400 units of apparent Ca concentration is subtracted from the gross Ca concentrations.To do this, we select the 49Ti isotope to determine the appropriate quantity due to 48Tito subtract from the 48Cameasurements. For plant digests, the spectral interferencedue to Ti has been amounting to about 10%of the measured Ca concentrations. For commercial ICP-MS systems, the manufacturer's software will do these corrections automatically. However, the analyst must use good judgment when specifying which isotope of the element causing the isobaric interference should be monitored for purposes of making the correction. In the case of foliar digests, nitric acid is used to dissolve the ashed residue. In this example, the 48Ti isotope is not a good choice to determine the correction on 48Ca due to 48Ti,because 47 { 1H16016014N}+, a degradation-recombination product of HNO, produced in respectable quantities by the plasma, interferes with the 47Ti measurement (see Table 11). In addition, the 47{ 31P160)+could cause significant problems with efforts to assess the Ti concentration using 47Ti measurements. Thus, selection of the 48Ca as the analytical isotope for Ca using a spectrometer capable of normal resolution requires selection of the 49Ti isotope, and possibly measurement of the 31P+isotope, for purposes of isobaric correction. Case 2: High Resolution

Note from Table VI that resolving the 48Caions at mass 47.95253 daltons from the 48{ 31P1H160}+ molecule at mass 47.976500 daltons and the 48Ti+ at mass 47.947949 daltons, differences of +0.023970 and -0.004581 dalton, should be no problem using a mass spectrometercapable of 50,000 resolution.

102

PAFWIZ N. SOLTANPOUR ETAL. Table VI Resolution Requirementsfor Analysis of Ca, Fe, Ni, and Zn in Plant Tissue

Analyte isotopes 4xCa

Separation from analyte to interfering species (daltons)

Interfering species 4 X ( 3 I PI H 1 6 0 ) +

4xTi+ s6Fe

+ +

S 6 [ 1 6 0 4 0 )~ ~

56( 4 O ~ ~ l h 0 )

SRN~

"Zn

sxFe sx142Ca'60)+ 581 1R040Ar)+ 581 40CalXOJ+ "Ni +"4XCaW]+ 64 ( 24Mg40Ar) + ffl(3lplHl60l60) + M(32s I 6 0 1 6 0 ) + 641 32S32SJ +

"( 6 3 C ~ ' H )

+

+0.02397 -0.004581 +0.022365 +0.022573 -0.002061 +0.018207 +0.026208 +0.026416 -0.001 184 +O.O 18305 +0.018288 +OM2275 +0.032763 +0.015006 +0.008275

Status"

R 0 R R 0

R R R 0 R R R X X X

('Disposition of the potential isobaric interference, assuming that a mass spectrometer resolution of 3500 or greater is available. 0 = correct using interelement calibration-subtraction technique. X = not probable in this application; if it turns out to be important, consider including an ashing step in the sample extract preparation protocol. R = set resolution to remove (potential) spectral problem.

IRONISOTOPESELECTION Case 1: Normal Resolution

Aglance at Table I1 indicates that there will be difficulties in choosing an Fe isotope for this analysis. The most abundant isotope is 56Fe.The background equivalent concentration of iron at the 56Fe isotope due to isobaric overlap by s6{40Ar'60)+ is approximately 200 pg liter-' using standard aqueous aerosol introduction (Lichte et al., 1987). There is also an isobaric overlap on 56Fe due to the presence of Ca in the test solutions in the form of the 56(40Ca'60)+molecule. The amount of 56(40Ca'60}+produced in the plasma depends on the concentration of Ca in the test solutions and is usually in the neighborhood of 1 4 % of the total Ca in the solution (Lichte et al., 1987).This is a very significant amount relative to the 56Fe+produced for the plant sample solutions. The next most abundant Fe isotope is s4Fe. If aerosol processing using N, addition to the Ar stream is being used to suppress molecular ion formation (Munro et al., 1986; Lam and McLaren, 1990; Evans and Ebdon, 1989), then the s4{ 14N40Ar)+will overwhelm

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 103

the background and another isotope should be considered for Fe concentration determinations. However, assuming N, is not being added, the isobaric interference by s4Cr+ must be considered. Unless the analyst is willing to assume that the Cr concentrations in the samples are negligible with respect to the iron concentration, an estimate of the Cr content of the samples relative to the Fe concentration must be made. To obtain this estimate, measurements for 50Cr+,52Cr+,andor 53Cr+ions on the sample solutions are generally made. Among these, the s3Cr+ is the most reliable because concentration estimates performed using measurements on it are more free of isobaric interference due to test-solution composition than the other two isotopes (see Table 11).

Case 2: High Resolution Using data from Table 11, the Fe isotope with the greatest naturally occurring abundance can be separated from both isobaric molecular species using a highresolution mass spectrometer. The atomic mass of 56Fe is 0.022365 daltons less than the molecular mass of 56(40Ar'60}+.It is 0.022573 daltons less than the molecular mass of s6(40Ca'60}+.

NICKEL ISOTOPESELECTION Case 1: Normal Resolution The naturally occurring relative abundance of the s8Ni isotope is 67.88%. It is subject to isobaric overlap by 58Fe+and 58(42Ca'60}+. The relative abundance of 60Ni, the next most abundant Ni isotope, is 26.23%. While free of isobaric problems due to Fe, it too is subject to interference by an oxide of Ca, 60(44Ca160}+. The three remaining Ni isotopes constitute about 6% of the total Ni in a naturally occurring sample and are subject to potentially serious isobaric interferences, as noted in Table 11. Thus, measurements using both 58Ni+and 60Ni+are made and corrections applied as discussed in Section V1.D. Using this correction method requires selecting an Fe isotope (discussed earlier).

Case 2: High Resolution The atomic mass of 58Ni is 57.935336 daltons. The atomic mass of s8Fe is 57.933275 daltons, or 0.002061 dalton smaller than the 58Ni atomic mass. The molecular mass of s 8 ( 42Ca160]+is 0.018207 dalton greater than 58Ni. Similarly, the molecular masses of 5 8 ( 18040Ar)+and 5 8 ( 40Ca'80}+are 0.026208 dalton and 0.026416 dalton greater than the atomic mass of 58Ni. Thus, resolving these species using a high-resolution mass spectrometer should be no problem.

104

PARVIZ N. SOLTANPOUR ETAL.

ZINCISOTOPESELECTION Case 1: Normal Resolution

The isotope selection process for determining the Zn concentration in plant tissue requires, like that for Ca, Fe, and Ni, more than simply choosing the most abundant isotope. According to Table 11, the two most abundant isotopes of Zn are subject to isobaric problems associated with the Mg, Ca, P, s, and/or CI content of the test solutions. The @Zn isotope is the most abundant naturally occumng isotope of Zn. It is subject to interferences by 64Ni+, @(24Mg40Ar}+,@(48Ca160}+, 61( 31P1H160160}+, @(32S160160)+, 32S32S)+,and @{63C~1H}+ (see Table VI). For analysis of plant tissue using the 64Zn isotope, aqueous nebulization, and normal mass resolution, the apparent Zn concentration due to @(24Mg40Ar}+ is usually the highest. The 66Zn isotope is the next most abundant naturally occurring isotope of Zn. It is subject to possible isobaric interferences by 66( 26Mg40Ar}+, 66( 31P35C1]+,132Ba++, 66( 34S160160}+,and 66{65C~1H}+. For analysis of plant tissue using the 66Zn isotope, aqueous nebulization, and normal mass resolution, the apparent Zn concentration due to 66( 26Mg40Ar}+is usually the highest. The Zn content of plant material is sufficiently high to allow use of either the 68Zn or 67Zn isotopes. Note that Table I1 indicates that to avoid isobaric interference by 68( 31P37C1}+,68( 35C11H160160}+,and 67( 35C1160160}+, the digestions should probably be done so that little or no perchloric acid and/or salts of perchlorate remain in the test solution. In addition, if appearance of sulfurcontaining molecules is a problem on the spectrometer, then the 68Zn+should be used even though it may be subject to spectral overlap by 68{ 32S36Ar}+.Calibration-subtraction methods may be needed to correct for this effect if it appears to be a problem. Finally, either the concentration of Ba in the test solutions must be known to be low, or the concentration of Ba must be measured and the appropriate correction applied. Doubly charged Ba ions are less than 2% of the total Ba in the plasma, so the 137Ba+is satisfactory for estimating corrections on measurements made using 68Zn or 67Zn isotopes. A sample pretreatment that includes ashing at 450-500°C can effectively remove much of the S from the plant sample. Thus, any possibility of an isobaric interference from 68{32S36Ar}+ on 68Zn+could be removed in this way. Case 2:

High Resolution

Resolving @Znfrom the most likely interfering species listed in Table VI should not be a problem using a high-resolution mass spectrometer. The atomic mass of @Zn is 63.929140 daltons, or 0.001184 dalton greater than @Ni, requiring a resolution better than what is commercially available. However, the Ni concentration in plant tissue is generally small compared to the Zn content; so this spectral in-

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 105 terference can, in most cases, be ignored. The molecular mass of 64{48Ca160}+ is 0.018305 dalton greater than the atomic mass of "Zn; the molecular mass of 64[ 24Mg40Ar}+ is 0.018288 dalton greater than the atomic mass of 64Zn;and the 64{32S160160}+, a{32S32S}+, and molecular masses of @ { 31P1H160160}+, 6 3 C ~ ' H } are + 0.042275, 0.032763, 0.015006, and 0.008275 daltons greater than the atomic mass of "Zn.

LEADISOTOPESELECTION Case 1: Normal Resolution There are four naturally occurring Pb isotopes. The three most abundant are not subject to isobaric interference from any species that would be present in plant tissue extracts. The most abundant Pb isotope is selected because the concentration levels of Pb in plant tissue are usually low, below 10-20 kg g-l. As discussed in Section VI.E, isotope dilution can be conveniently used for Pb concentration determinations, in which case measurements using 207Pband/or *06Pb isotopes must also be made.

Case 2: High Resolution There are no isobaric problems associated with measuring the three most abundant Pb isotopes: 208Pb,207Pb,and 206Pb.The 204Pbisotope is separated from the *04Hg isotope by 0.00045 dalton. A resolution of over 450,000 is necessary to resolve the 204Hgand 204Pbunder these conditions.

APPENDIX SUMMARY

In summary, the isotope selection process for normal-resolution quadrupole mass spectrometers generally requires more than choosing the most abundant naturally occurring isotope for analytical measurement. A knowledge of the major elemental components in the sample is necessary in most situations. Reagents that are to be used in the sample pretreatment must also be taken into account. In some instances, it is not possible to select an isotope completely free of isobaric interference that will permit analytical detection and/or accuracy requirements to be achieved. In these cases, the simplicity of the mass spectrum generated by the ICP-MS instrumentation is a limitation of the technique and, unless a modification of the sample introduction system is devised, e.g., adding a desolvation apparatus after the nebulizer to reduce oxide and/or sulfide populations to tolerable levels (Veillon and Marghoshes, 1968), an alternate method of measurement will be necessary for the affected analyte(s).

106

PARVIZ N. SOLTANPOUR ETAL.

The isotope selection process is greatly simplified for high-resolution, doublefocusing mass spectrometers with magnetic and electrostatic sectors. These spectrometers are capable of 50,000 resolution. However, a resolution setting of 400, selected to give peaks 0.5 dalton wide at 200 daltons, is used for maximum analyte gain, i.e., ion response per unit concentration. The response is inversely proportional to the resolution setting. For example, taking the signal at resolution 400 as 100% for a given isotope, a resolution of 2500 would only result in 10%of this signal (Bradshaw et al., 1989). Each isobaric interference would be evaluated in terms of the severity of the interference in units of apparent concentration of analyte species resulting for the highest (expected) concentration of the species causing the problems. The lowest resolution, while still allowing reduction of the most serious isobaric interferences to acceptable levels, should be used. In this way, the resulting analyte gain will be the one most likely to result in useable analyte signal. In the example presented here, a resolution of 3500 was selected (see Table VI). This resulted in analytical peak heights in these high-resolution spectral segments that are approximately 7.1% of the peak heights observed using 400 resolution for the same solution. Also, a word of caution to the reader: Kinetic-energy spreading of the ions can cause smearing of the isotope masses under some conditions (Ahearn, 1972). We do not assume responsibility for manufacturers’ resolution claims. We suggest an assessment of the resolution, under anticipated standard-operating conditions, with the manufacturer(s) prior to purchase.

REFERENCES Abell, I. D. (1991). Performance benefits of optimization of laser ablation sampling for ICP-MS. In “Applications of Plasma Source Mass Spectrometry,” (G. Holland and Andrew N. Eaton, eds.), pp. 209-217. The Royal Society of Chemistry, London. Ahearn, A. J. (Ed.) (1972). “Trace Analysis by Spark Source Mass Spectrometry.’’Academic Press, New York. Bacon, J. R.,Ellis, A. T., and Williams, J. G. (1989). Atomic spectrometry update-Inorganic mass spectrometry and x-ray fluorescence spectrometry. J. Anal. Ar. Specrrom. 4, 199R. Bacon, J. R., Ellis, A. T.,and Williams, J. G. (1990). Atomic spectrometry update-Inorganic mass spectrometry and x-ray fluorescence spectrometry. J. Anal, At. Spectrom. 5,243R. Bacon, J. R.,Ellis, A. T.,and Williams, J. G. (1991). Atomic spectrometry update-Inorganic mass spectrometry and x-ray fluorescence spectrometry. J. Anal. A?. Spectrom. 6,2298. Bajo, S. (1978). Volatilization of arsenic (HI, V), antimony (III, V), and selenium (IV, VI) from mixtures of hydrogen fluoride and perchloric acid solution: Application to silicate analysis. Anal. Chem. 50,649-65 I . Barnes, R. M. (ed.) (1992). “ICP Information Newsletter.” Univ. of Massachusetts, Amherst, MA. Beasecker, D. R., and Williams, L. L. (1978). An improved sample delivery system for ICAP analysis. Jarrell-Ash Plasma Newsler. 1(3), 5-9. Beauchemin, D. (1989). Early experiences with inductively coupled plasma mass spectrometry. J. Anal. At. Specirom. 4,553. Beauchemin, D., McLaren, J. W., and Berman, S. S. (1987). Study of the effects of concomitant elements in inductively coupled plasma mass spectrometry. Spectrochim. Acra 42B,467.

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 107 Belchamber, R. M., Betteridge, D., Wade, A. P., Cruickshank, A. J., and Davison, P. (1986). Removal of a matrix effect in ICP-AES multi-element analysis by simplex optimization. Spectruchim.Acta. 41B, 503-505. Bernas, B. (1968). A new method for decomposition and comprehensive analysis of silicates by atomic absorption spectrometry. Anal. Chem. 40,1682-1686. Boumans, P.W.J.M. (1966). “Theory of Spectrochemical Excitation.” Plenum Press. New York. Boumans, P.W.J.M. ( 1984). “Line Coincidence Tables for Inductively Coupled Plasma Atomic Emission Spectrometry,” 2nd ed. Pergamon, New York. Bourene, M., and Le CalvC, J. (1973). De-excitation cross sections of metastable argon by various atoms and molecules. J. Chem. Physics 58, 1452. Bradford, G. R., and Bakhtar, D. (1991). Determination of trace metals in saline irrigation drainage waters with ICP-OES after pre-concentration by chelation/solvent extraction. Envi,: Sci. and Tech. 25, 1704-1708. Bradshaw, N., Hall, E. F. H., and Sanderson, N. E. (1989). Inductively coupled plasma as an ion source for high-resolution mass spectrometry. J. Anal. At. Spect. 4,801. Braverman. D. S. (1992). Determination of rare earth elements by liquid chromatography separation using inductively coupled plasma mass spectrometric detection. J. Anal. At. Spect. 7,43. Brown, P. G., Davidson, T. M., and Caruso, J. A. (1988). Application of He microwave induced plasma mass spectrometry to the detection of high ionization potential gas phase species. J. Anal. At. Spect. 3,763. Brown, R. M., Long, S. E., and Pickford, C. J. (1988). The measurement of long-lived radionuclides by non-radiometric methods. Sci. Total Envis 70,265. Chisum, M. E. (1992).Applications of negative ion analyses on the ELAN 250 ICP/MS. At. Spect. 12, 155. Cresser, M. S., Ebdon, L. C., McLeod, C. W.. and Burridge, J. C. (1986). Atomic spectrometry update-Environmental analysis. J. Anal. At. Spect. 1, 1R. Cresser, M. S., Ebdon, L. C., and Dean, J. R. (1988). Atomic spectrometry update-Environmental analysis. J. Anal. At. Spect. 3, 1R. Cresser, M. S., Ebdon, L. C., and Dean, J. R. (1989). Atomic spectrometry update-Environmental analysis. J. Anal. At. Spet,: 4, IR. Cresser, M. S., Ebdon, L. C., Armstrong, J., Dean, J. R., Ramsey, M. H., and Cave, M. (1990). Atomic spectrometry update-Environmental analysis. J. Anal. At. Spect. 5, IR. Cresser, M. S., Armstrong, J., Dean, J. R., Ramsey, M. H., and Cave, M. (1991). Atomic spectrometry update-Environmental analysis. J. Anal. At. Spect. 6, IR. Cresser, M. S., Armstrong, J., Dean, J. R., Watkins, P., and Cave, M. (1992). Atomic spectrometry update-Environmental analysis. J . Anal. At. Specr. 7, IR. Dahlquist, R. L., and Knoll, J. W. (1978). Inductively coupled plasma-atomic emission spectrometry. Analysis of biological materials and soils for major, trace, and ultra-trace elements. Appl. Spect. 32, 1-29. Date, A. R. (1986). ICP-MS: The best thing in analytical chemistry since chopped light? In “SAC 86/3rd Biennial National Atomic Spectroscopy Symposium,” pp. 2G26. University of Bristol, UK. Date, A. R., and Gray, A. L. (Eds.) (1989). “Applications of Inductively Coupled Plasma Mass Spectrometry.” Blackie and Sons, London. Date, A. R., Cheung, Y. Y..and Stuart. M. E. (1987). The influence of polyatomic ion interferences in analysis by inductively coupled plasma source mass spectrometry (ICP-MS). Spectruchim. Actu 42B, 3. Dean, J. R., Ebdon, L., Crews, H. M., and Massey, R. C. (1988). Characteristics of flow injection inductively coupled plasma mass spectrometry for trace metal analysis. J. Anal. At. Speci. 3, 349. Denoyer, E. R. (1991). Analysis of powdered samples by laser sampling ICP-MS. In “Applications of Plasma Source Mass Spectrometry” (G. Holland and A. N. Eaton, eds.), pp. 199-208. The Royal Society of Chemistry. London.

108

PARVIZ N. SOLTANPOUR ETAL.

Denoyer, E. R., and Stroh, A. (1992). Expanding ICP-MS capabilities using Row injection. Am. Lab. (Feb.), 74. Denoyer, E. R., Qinghong, L., Stroh, A,, Vollkopf, U.,and Thomas, R. (199la). “Extending ICP-MS Capabilities with Flow Injection Analysis.” Paper presented at the Elemental Mass Spectrometry Workshop. Am. SOC.for Mass Spect. San Francisco, CA. Denoyer, E. R., Fredeen, K. J., and Hager, J. W. (1991b). Laser solid sampling for inductively coupled plasma mass spectrometry. Anal. Chem. 63,445A. Doherty. W. (1989). An internal standardization procedure for the determination of yttrium and the rare earth elements in geological materials by inductively coupled plasma-mass spectrometry. Spectrochim. Acta 44B, 263. Dolan, R., Van Loon, J., Templeton, D.. and Paudyn, A. (1990). Assessment of ICP-MS for routine multielement analysis of soil samples in environmental trace element studies. Fresen. J. Anal. Chem. 336,99. Douglas. D. J. (1989). Some current perspectives on ICP-MS. Can. J. Spect. 34,38. Douglas, D. J., and Houk, R. S. (1985). Inductively-coupled plasma mass spectrometry (ICP-MS). Prog. Anal. At. Spect. 8, 1. Douglas, D. J., and Ken; L. A. (1988). Study of solids deposition on inductively coupled plasma mass spectrometry samplers and skimmers. J. Anal. At. Spect. 3,749. Ebdon, L. C., Cresser, M. S., and McLeod, C. W. (1987). Atomic spectrometry update-Environmental analysis. J. Anal. At. Spect. 2, IR. Ek, P. G., Huldkn, S. G., Johansson, E., and Liljefors, T. (1991). A continuous hydride generation system for ICP-MS using separately nebulized internal standard solutions. In “Applications of Plasma Source Mass Spectrometry” (G. Holland and A. N. Eaton, eds.), pp. 178-198. The Royal Society of Chemistry. Evans, E. H.,and Ebdon, L. C. (1989). Simple approach to reducing polyatomic ion interferences on arsenic and selenium in inductively coupled plasma mass spectrometry. J. Anal. Ar. Spect. 4,299. Fassel, V. A. (1977). Current and potential applications of inductively coupled plasma (ICPtatomic emission spectroscopy (AES) in the exploration, mining, and processing of materials. Pure Appl. Chem. 49,1533-1545. Fassel, V. A., and Kniseley, R. N. (1974). Inductively coupled plasmas. Anal. Chem. 46,1155A-I 164A. Fassett, J. D., and Paulsen, P. J. (1989). Isotope dilution mass spectrometry for accurate elemental analysis. Anal. Chem. 61,643A. Garbarino, J. R., and Taylor, H. E. (1987). Stable isotope dilution analysis of hydrologic samples by inductively coupled plasma mass spectrometry. Anal. Chem. 59, 1568. Gervais, L. S., and Salin, E. D. (1991). Inserted injector tubes for inductively coupled plasma spectrometry. J. Anal. Ar. Spect. 6,493. Gray, A. L. (1985). The ICPas an ion source-origins, achievements, and prospects. Spectrochim. Acta 40B, 1525. Gray, A. L. (1986). Mass spectrometry with an inductively coupled plasma as an ion source: The influence on ultratrace analysis of background and matrix response. Spectrochim. Acta 41B, 15 1. Gray. A. L. (1989). Visual observation of shock waves in an inductively coupled plasma mass spectrometry expansion stage. J. Anal. At. Spect. 4,371. Greenfield, S. (1983). Inductively coupled plasma-Atomic emission spectroscopy (ICP-AES) with Row injection analysis (FIA). Spectrochim. Acta. 38B, 93-105. Gregoire, D. C. (l987a). The effect of easily ionizable concomitant elements on non-spectroscopic interferences in inductively coupled plasma-mass spectrometry. Spectrochim. Acta 42B, 895. Gregoire, D. C. (l987b). Influence on instrument parameters on nonspectroscopic interferences in inductively coupled plasma-mass spectrometry. Appl. Specr. 41,897. Gregoire, D. C. (1989). Application of isotope ratios determined by ICP-MS to earth science studies. Prog. Anal. Spect. 12,433. Hager, J. W. (1989). Relative elemental responses for laser ablation-ICP-MS. Anal. Chem. 61, 1243.

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 109 Hassan. S. M., and Loux, N. T. (1990). Elimination of spectral interferences in inductively coupled plasma-atomic emission spectroscopy using orthogonal polynomials. Specfrochim. Acra 45B, 7 19-729. Hieftje, G. M. (1992). Toward the next generation of plasma-source mass spectrometers. In “Winter Conf. on Plasma Spectrochemistry,” p. PL6. San Diego, CA. Hieftje, G. M., and Vickers, G. H. (1989). Developments in plasma sourcelmass spectrometry. Anal. Chim. Acra 216, I . Hieftje, G. M., Vickers, G. H., and Wilson, D. A. (1988). Detection of negative ions by ICP-MS. Anal. Chem. 60(17), 1808-1812. Holden, N. E., and Walker, F. W. (1972). “Chart of the Nuclides,” 1Ith ed. Knolls Atomic Power Laboratory, operated by General Electric under direction of Naval Reactors, U S . AEC. Educational Relations, General Electric, Schenectady, NY. Holland, G., and Eaton, A. N. (1991). “Applications of Plasma Source Mass Spectrometry.”The Royal Society of Chemistry, London. Houk, R. S. (1986). Mass Spectrometry of Inductively Coupled Plasmas. Anal. Chem. 58,97A. Houk, R. S., and Thompson. J. J. (1988). Inductively Coupled Plasma Mass Spectrometry. Mass Specr. Rev. 7,425. Houk, R. S . , Svec, H. J., and Fassel, V. H. (1981). Mass spectrometric evidence for suprathermal ionization in an inductively coupled argon plasma. Appl. Specr. 35,380. Huang, Y. Q.. and Wai, C. M. (1986). Extraction of arsenic from soil digests with dithiocarbamates for ICP-AES analysis. Commun. Soil Sci. Plant Anal. 17, 125-133. Hutton. R. C., and Eaton, A. N. (1987). Role of aerosol water vapour loading in inductively coupled plasma mass spectrometry. J. Anal. At. Spect. 2,595. Igarashi, Y., Kim, C. K., Takaku, Y., Shiraishi, K., Yamamoto, M., and Ikeda, N. (1990). Application of inductively coupled plasma mass spectrometry to the measurement of long-lived radionuclides in environmental samples. A review. Anal. Sci. 6, 157. Isaac, R. A., and Johnson, W. C. (1983). High-speed analysis of agricultural samples using inductively coupled plasma-atomic emission spectroscopy. Spectrochim. Acra 38B, 277-282. Jackson, L. L., McKown, D. M., Taggart, J. E., Jr., Lamothe, P. J., and Lichte. F. E. (1989). Geological and Inorganic Materials. Anal. Chem. 61, 109R. Jackson, L. L., Fries,T. L., Grossman, J. N., King, B. S. W., andLamothe, P.J. (1991). Geological and inorganic materials. Anal. Chem. 63 (biennial Applications Review issue): 33R-48R. Johnson, G. W. (1979). Trace analysis with a direct-current plasma Echelle spectrometer system. Ph.D. diss. Colorado State Univ., Fort Collins, CO. Johnson, G. W., Taylor, H. E., and Skogerboe, R. K. (l979a). Evaluation of solute vaporization interference effects in a direct current plasma. Anal. Chem. 51,2403. Johnson. G. W., Taylor, H. E., and Skogerboe, R. K. (1979b). Evaluation of spectral interferences associated with a direct current plasma-multielement atomic emission spectrometer (DCP-MAES) system. Appl. Specr. 33,45 I. Johnson, G. W.. Taylor, H. E.,and Skogerboe, R. K. (1980). Characterization of an interelement enhancement effect in a dc plasma atomic emission spectrometry system. Appl. Spect. 34, 19. Johnson, G. W., Littlefield, E. S., and Miller, R. 0. (1992a). Botanical tissue analysis using the USN ICP-MS technique. In “Pacific Conf. on Chem. and Spect.,” paper no. 64,p. 49. Foster City, CA. and Brown, P. (1992b). ICP-MS for plant analysis. In “Winter Conf. on Johnson, G . W., Miller, R. 0.. Plasma Spect.,” paper no. F9,p. 2 12. San Diego. CA. Jones, J. B., Jr. (1977). Elemental analysis of soil extracts and plant tissue ash by plasma emission spectroscopy. Commun. Soil Sci. Plant Anal. 8,349-365. Kim, C. K., Oura. Y., Takaku, Y., Nitta, H., Igarashi, Y., and Ikeda, N.(1989a). Measurements of 24”Pu: 239Puratio by fission track method and inductively coupled plasma mass spectrometry. J. Radioanal. Nucl. Chrm. 136(5). 353. Kim, C. K.. Takaku, Y., Yamamoto, M., Kawamura, H., Siraishi, K., Igarashi, Y., Igarashi, I., Takaya-

110

PARVIZ N. SOLTANPOURETAL.

ma, H., and Ikeda. N. (1989b). Determination of neptunium-237 in soil samples using inductively coupled plasma mass spectrometry. Radioisotopes 38, 153. Kim, C. K., Seki, R., Morita, S., Yamasaki, S. I., Tsumura, A.. Takaku, Y.,Igarashi, Y., and Yamamoto, M. (1991). Application of a high-resolution inductively coupled plasma mass spectrometer to measurement of long-lived radionuclides. J. Anal. At. Spect. 6,205. Koirtyohann, S. R., Jones, J. S., and Yates, D. A. (1980). Nomenclature system for low-power inductively coupled plasma. Anal. Chem. 52, 1965. LaFerniere, K. E., Rice, G. W., and Fassel, V. A. (1985). Flow injection analysis with inductively coupled plasma-atomic emission spectroscopy: Critical comparison of conventional pneumatic, ultrasonic, and direct injection nebulization. Spectrochim. Acta 40B, 1495-1504. Laird, D. A., Dowdy, R. H., and Munter, R. C. (1991). Suspension nebulization analysis of clays by inductively coupled plasma-atomic emission spectroscopy. Soil Sci. Soc. Am. J. 55, 274-278. Lam, J. W., and McLaren, J. W. (1990). Use of aerosol processing and nitrogen-argon plasmas for reduction of oxide interference in inductively coupled plasma mass spectrometry. J. Anal. At. spect. 5,419. Langmyhr, F. J., and Paus, P. E. (1968). The analysis of inorganic siliceous materials by atomic absorption spectrophotometry and the hydrofluoric acid decomposition technique. Anal. Chem.Acta 43,50&507. Larson, G. F., Fassel, W. V., Scott, R. H., and Kniseley, R. N. (1975). Inductively coupled plasma4ptical emission analytical spectrometry. A study of some interelement effects. Anal. Chem. 47,238. Larson, G. F., Fassel, V. A., Winge, F. K., and Kniseley, R. N. (1976). Ultratrace analyses by optical emission spectroscopy: The stray light problem. Appl. Specr. 30,384-391. Lechler, P. J., and Leininger, R. K. (1979). Analysis of black shale by ICAP spectroscopy. Jarrell-Ash Plasma Newsletter 2( I), 8-10, Legere, G., and Burgener, P. (1985). Elimination of high-salt interference effects caused by lithium metaborate using a new Teflon high-salt nebulizer. ICP Information Newsletter 11,447-456. Lichte, F. E., Meier, A. L., and Crock, J. G.(1987). Determination of rare earth elements in geological materials by inductively coupled plasma mass spectrometry. Anal. Chem. 59, 1150. Lim, H. B., Houk, R. S., Edelson, M. C., and Carney, K. P. (1989). Some fundamental characteristics of a reduced-pressure plasma extracted from an inductively coupled plasma. J. Anal. At. Sprct. 4, 365. Lindsay, W. L., and Norvell, W. A. (1978). Development of a DTPA soil test for zinc, iron, manganese, and copper. Soil Sci. SOC.Am. J. 42,421-428. Marciello, I., and Ward, A. F.(1978). Interelement corrections for spectral line interferences. JarrellAsh Plasma Newsletter 1(2), 12-13. McLaren, J. W., and Berman, S. S. (1985). Wavelength selection for trace analysis by ICP-AES. Spectrochim. Acta 40B, 21 7-225. McLaren. J. W., Beauchemin, D., and Berman, S. S. (1987).Application of isotope dilution inductively coupled plasma mass spectrometry to the analysis of marine sediments. Anal. Chem. 59,610. McQuaker, N. R., Kluckner, P. D., and Chang, G. N. (1979). Calibration of an inductively coupled plasma-atomic emission spectrometer for the analysis of environmental materials. Anal. Chem. 51, 888-895. Munro, S., Ebdon, L. C., and McWeeny, D. J. (1986). Application of inductively coupled plasma mass spectrometry (ICP-MS) for trace metal determination in foods. J. Anal. At. Specr. 1,211. Odegard. M. (1979). Determination of major elements in geological materials by ICAP spectroscopy. Jarrell-Ash Plasma Newsletter 2( I), 4-7. Olivares, J. A.. and Houk, R. S. (1986). Suppression of analyte signal by various concomitant salts in inductively coupled plasma mass spectrometry. Anal. Chem. 58,20-25. Olson, K. W., Haas, W. J., Jr., and Fassel, V. A. (1977). Multielement detection limits and sample nebulization efficiencies of an improved ultrasonic nebulizer and a conventional pneumatic nebulizer in inductively coupled plasma-atomic emission spectrometry. Anal. Chem. 49,632-637.

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 11I Parsons, M. L., Forster, A,. and Anderson, D. (1980). “An Atlas of Spectral Interferences in ICP Spectroscopy.” Plenum, New York. Pearce, N. J. G., Perkins, W. T., Abell, I., Duller, G. A. T., and Fuge, R. (1992). Mineral microanalysis by LASER ablation inductively coupled plasma mass spectrometry. J. Anal. At. Spect. 7,53. Plantz, M. R., Fritz, J. S., Smith, F. G., and Houk, R. S. (1989). Separation of trace metal complexes for analysis of samples of high-salt content by inductively coupled plasma mass spectrometry. Anal. Chem. 61, 149. Powell, M. J., and Boomer, D. W. (1995). Determination of chromium species in environmental samples using high-pressure liquid chromatography direct injection nebulization and inductively coupled plasma mass spectrometry. Anal. Chem. 67( 14), 2474-2478. Robin, J. (1979). Emission spectrometry with the aid of an inductive plasma generator. I C P Information Newsletter 4( l l ), 495-509. Roychowdhury, S. B., and Koropchak, J. A. (1990). Thermospray enhanced inductively coupled plasma atomic emission spectroscopy detection for liquid chromatography. Anal. Chem. 62,484-489. Sah, R. N., and Miller, R. 0. (1992). Spontaneous reaction for acid dissolution of biological tissues in closed vessels. Anal. Chem. 64,230. Serfas, R. E., Thompson, J. J., and Houk, R. S. (1986). Isotope ratio determinations by inductively coupled plasmdmass spectrometry for zinc bioavailability studies. Anal. Chim. Acta 188,73. Skogerboe, R. K., and Grant, C. L. (1970). Comments on the Definitions of the Terms Sensitivity and Detection Limit. Spect. Lett. 3,25. Skogerboe, R. K., and Koirtyohann, S. R. (1976). Accuracy assurance in the analysis of environmental samples. Nat. Bureau of Standards Publ. 422. Washington, D.C. Slavin, M. (I97 1 ). Emission spectrochemical analysis, pp. 53-80. Wiley-Interscience, New York. Smith, F. G., Wiedrin, D. R.,and Houk, R. S. (1991). Measurement of boron concentrations and isotope ratios in biological samples by inductively coupled plasma mass spectrometry with direct injection nebulization. Anal. Chim. Acta 248,229-234. Soltanpour, P. N. (I991 ). Determination of nutrient availability and elemental toxicity by AB-DTPA soil test and ICPS. Adv. Soil Sci. 16, 165-190. Soltanpour, P. N., and Schwab, A. P. (1977). A new soil test for simultaneous extraction of macro- and micro-nutrients in alkaline soils. Commun.Soil Sci. Planr Anal. 8, 195-207. Soltanpour. P. N., and Workman, S. (1979).Modification of the NH,HCO,-DTPA soil test to omit carbon black. Commun.Soil Sci. Plant Anal. 10, 141 1-1420. Soltanpour, P. N.. and Workman, S. (198 I). Soil testing methods used at Colorado State University Soil Testing Laboratory for the evaluation of fertility, salinity, sodicity, and trace element toxicity. Colorado State Univ. Exp. Stn., Technical Bull. 142. Fort Collins, CO. Soltanpour. P. N., Khan, A,, and Lindsay, W. L. (1976). Factors affecting DTPA-extractable Zn, Fe, Mn, and Cu from the soils. Commun.Soil Sci. Plant Anal. 7,797-821. Soltanpour. P. N., Khan, A,. and Schwab. A. P. (1979a). Effect of grinding variables on the NH,HCO,DTPA soil test values for Fe, Zn, Mn, Cu, P, and K. Commun.Soil Sci. Plant Anal. 10,903-909. Soltanpour, P. N., Workman, S. M., and Schwab, A. P. (1979b). Use of inductively coupled plasma spectrometry for the simultaneous determination of macro- and micronutrients in NH,HCO,DTPA extracts of soils. Soil Sci. Soc. Am. J. 43,75-78. Soltanpour, P. N., Jones, J. B., Jr., and Workman, S. M. (1982). Optical emission spectrometry. In “Methods of Soil Analysis,”2nd ed. (A. L. Page, ed.). Part 2, pp. 29-65. Agron. Monogr: 9. SSSA, Madison, WI. Soltanpour. P. N., Johnson, G. W., Workman, S. M., Jones, J. B., Jr., and Miller, R. 0. (1996). Inductively coupled plasma emission spectrometry and inductively coupled plasma mass spectrometry. In “Methods of Soil Analysis” (D. L. Sparkes, A. L. Page, P. A. Helmke et al., eds.), Part 3, Chemical methods, pp. 9 I- 139. Agron. Monogr. 9. SSSA, Madison, WI. Suddendorf, R. F., and Boyer. K. W. (1978). Nebulizer for analysis of high-salt content samples with inductively coupled plasma emission spectrometry. And. Chem. 50, 1769-177 1.

112

PARVXZ N. SOLTANPOUR ETAL.

Tan, S. H., and Horlick, G. (1986). Background spectral features in ICP-MS. Appl. Spec?.40,445. Taylor, P., and Schutyser, P. (1986). Description of a computer program for quantitative spectral analysis of ICP-AES spectra generated with a high-resolution computer-controlled monochromator. Spectrochim.Acta. 41B, 81-103. Thompson, J. J.. and Houk, R. S. (1986). Inductively coupled plasma mass spectrometric detection for multielement flow injection analysis and elemental speciation by reversed-phase liquid chromatography. Anal. Chem. 58,2541. Thompson, J. J.. and Houk, R. S. (1987). Study of internal standardization in inductively coupled plasma-mass spectrometry. Appl. Spec?.41,801. Thompson, M., and Walsh, J. N. (1983). “A Handbook of Inductively Coupled Plasma Spectrometry.” Blackie and Sons, London. Thompson, M., Pahlavanpour, B., and Walton, S. J. (1978a). Simultaneous determination of trace concentrations of arsenic, antimony, bismuth, selenium, and tellerium in aqueous solution by introduction of the gaseous hydrides into an inductively coupled plasma source for emission spectrometry. Part I. Preliminary studies. Analyst (London) 103,568-579. Thompson, M., Pahlavanpour, B., and Walton, S. J. (1978b). Simultaneous determination of trace concentrations of arsenic, antimony, bismuth, selenium, and tellurium in aqueous solution by introduction of the gaseous hydrides into an inductively coupled plasma source for emission spectrometry. Part 11. Interference studies. Analyst (London) 103,705-71 3. Tsumura, A,, and Yamasaki. S. (1991). Determination of ultra-trace levels of rare earth elements in terrestrial water by high resolution ICP-MS with an ultrasonic nebulizer. In “Applications of Plasma Source Mass Spectrometry” (G. Holland and A. N. Eaton, eds.), pp. 119-129. The Royal Society of Chemistry, London. U.S. Department of Commerce National Technical Information Service (USDCNTIS). (1977). “Evaluation of an Inductively Coupled Plasma, Multichannel Spectrometric Analysis System.’’ Anal. Chem. Branch PB-270-918. Ames Lab., Ames, IA; Envir. Res. Lab., Athens, GA. Ure, A. M., Ellis, A. T., and Williams, J. G. (1988). Atomic spectrometry update-Inorganic mass spectrometry and x-ray fluorescence spectrometry. J. Anal. At. Spect. 3, 174R. van Heuzen, A. A., Hoekstra, T., and van Wingerden, B. (1989). Precision and accuracy attainable with isotope dilution analysis applied to inductively coupled plasma mass spectrometry: Theory and experiments. J. Anal. A?. Spect. 4,483. Vaughan, M. A., and Horlick, G. (1986). Oxides, hydroxide, and doubly-charged analyte species in inductively coupled plasmahass spectrometry. Appl. spect. 40,434. Veillon, C., and Margoshes, M. (1968). A pneumatic solution nebulization system producing dry aerosol for spectroscopy. Spectrochim.Acta 23B, 553. Viczian, M., Lasztity, A., Wang, X., and Barnes, R. M. (1990). On-line isotope dilution and sample dilution by flow injection and inductively coupled plasma mass spectrometry. J. Anal. At. Spect. 5, 125. Ward, A. F. (1978a). Inductively coupled argon plasma spectroscopy. Development, technique, and applications. Am. Lab. 10,79-87. Ward, A. F. (1978h). Stock standard preparation. Jarrell-Ash Plasma Newsletter 1(2), 1 4 1 5 . Ward, A. F., and Myers, R. B. (1979). The effect of concomitant species upon the analyte signal observed from a modem ICAP direct reading spectrometer (Abstract). I C P Information Newsletrer 4(9), 403. Weast, R. C., and Astle, M. J. (Eds.) (1979). “CRC Handbook of Chemistry and Physics,” 59th ed. CRC Press, Boca Raton, FL. Winge, R. K.,Peterson, V. J., and Fassel, V. A. (1979). Inductively coupled plasma-atomic emission spectroscopy. Prominent lines. Appl. Spect. 33,206-21 9. Winge, R. K., Eckels, D. E., DeKalb, E. L.. and Fassel, V. A. (1988). Spatiotemporal characteristics of the inductively coupled plasma. J. Anal. At. Specr. 3,849.

ADVANCES IN ICP EMISSION AND ICP MASS SPECTROMETRY 1 13 Winge, R.K., Crain, J. S., and Houk. R.S. (1991). High-speed photographic study of plasma Ructuations and intact aerosol particles or droplets in inductively coupled plasma mass spectrometry. J. Anal. At. Spect. 6,601.

Wolcott, J. F., and Butler, C. S.(1979). A simple nebulizer for the inductively coupled plasma source (Abstract). ICP Information Newsleffer 4( lo), 461. Workman, S. M., and Soltanpour, P. N.(1980). Importance of pre-reducing selenium (VI) to selenium (IV) and decomposing inorganic matter in soil extracts prior to determination of selenium using hydride generation. Soil Sci. Soc. Am. J. 44,1331-1333.

This Page Intentionally Left Blank

MANAGING COTTONNITROGEN SUPPLY Thomas J. Gerik,' Derrick M. Oosterhuis,2and H. Allen Torbert3 'Texas Agricultural Experiment Station Blackland Research Center Temple, Texas 76502 *Department of Agronomy University of Arkansas Fayetteville, Arkansas 72703 USDA-Agricultural Research Service Grassland Soil and Water Research Laboratory Temple, Texas 76502

I. Introduction 11. Cotton Growth and Nitrogen Response A. Plant Growth Habit

B. Plant Response to Nitrogen Deficiency Interactions between Growth Habit and Nitrogen D. Plant Nitrogen Requirements 111. Soil Nitrogen Availahility and Dynamics A. Inorganic Soil Nitrogen B. Soil Organic Nitrogen C. Nitrogen Additions to the Soil Tv. Foliar-Nitrogen Fertilization in Cotton V. Monitoring Cotton Nitrogen Status A. Fertilizer Tests B. Soil Testing C. Plant-Tissue Analyses D. Petiole Nitrate Analysis E. Total Leaf Nitrogen Assay F. Nitrate Reductase Activity G . Chlorophyll Determination VI. Managing Cotton Nitrogen Supply A. Computer Models B. Managing Nitrogen Supply with Crop-Water Use VII. Summary References

<:.

115 A d i ~ n t t r ain ,-lpmnt~y,Volrintr 64

Copyright 0 I9 Y X hy Academic Press. NI right$of reproducoon in any fonn reserved 0065-21 13/98 $25.00

116

THOMAS J. GERIK ETAL.

I. INTRODUCTION Maintaining soil fertility is important in sustaining cotton (Gossypium hirsutum L.) productivity and profitability. Of the three macronutrients, nitrogen (N), phosphorous (P), and potassium (K), nitrogen is applied to cotton in the greatest quantity (Table I). Yet the complexity of N cycling in the soil and the indeterminate growth habit of cotton complicate our ability to estimate fertility requirements. In the U.S. Cotton Belt the timing and method of N fertilization differs greatly among regions (Table 11). Nitrogen is applied as preplanting (before planting) and postplanting (after planting) applications in most states, but less than 10%of U.S. cotton acreage receives N at planting. Most N is uniformly applied over the field as preplanting and postplanting applications that are broadcast or injected directly into the soil, but combining N fertilizer with irrigation (i.e., chemigation) is popular in the Arizona and California deserts. Less than 5% of cotton acreage received N as a foliar treatment. Although in 1994 most cotton growers typically used multiple N applications (Table I) to reduce losses associated with leaching, denitrification, or immobilization and to minimize risk of salinity injury to seedlings, growers in most states did not use soil or plant-tissue analysis for crop-fertilization decisions (Table 111). Only 19-53% of the cotton acreage was tested in 1994 for soil N, and 1-33% of the cotton acreage was tested with tissue-analysis procedures in representative U.S. cotton-growing states (Taylor, 1995). Yet growers that used soil andor tissue testing valued the information, since they overwhelmingly followed the resulting recommendations. The N requirement and utilization for cotton is more complex than for other major field crops. The question is, Why do many cotton growers in the United States Table I Fertilizer Use and Planted Cotton Acreage in Different Regions of the U.S. Cotton Belt in 1994‘

Arizona F e ~ u S e d ( t 0 n s XIOOO) Nitrogen 90 Phosphorous 30 Potassium I Nitrogen-use change 0.23 from 1993 (%) Planted cotton acres 313 (X Nitrogen application 220 Annual rate (Ib/acre) Averagetreatmentsperane 2.8 “Data from Taylor, 1995.

Adansas

California

Louisiana

Mississippi

Texas

303 79 I20 0.13

549 179 162 0.02

188 50 78 0.17

172 64 104 -0.08

1040

980

1100

900

110

2.3

188 1.9

157 2.3

1280

122 2.0

273 159 0.11 5450

71 1.4

117

MANAGING C O m O N NITROGEN SUPPLY Table I1 Timing and Method of Application to Cotton Acreage in Different Regions of the U.S. Cotton Belt in 1994a Treated acres (%)’

Nitrogen timing Fall, before planting Spring, before planting Spring, at planting Spring, after planting Fertilizer application method Broadcast (ground) Broadcast (air) Chemigation Banded Foliar Injected (with knife)

Arizona

Arkansas

California

Louisiana

Mississippi

Texas

15

23 52 9 60

44 21 13 86

10 45 10

63

9 54 8 12

41 47 5 32

90

32 5 32 25 4 64

46 18 NR‘ 29 NR 56

64 12 NR 19 2 12

64 I 5 NR 35

22 15 95 17 8 43 22 2 62

10 1 10

I 29

19

uData from Taylor, 1995. ’Percentages may exceed 100, because an acre may be treated more than once. “NR, not reported.

Table 111 Timing and Method of N Application to Cotton Acreage in Different Regions of the U.S. Cotton Belt in 1994u Planted acres (%) Nitrogen testing Soil Acreage tested Recommendation applied Greater than recommendation applied Less than recommendation applied Tissue Acreage tested Recommendation applied Greater than recommendation applied Less than recommendation applied “Data from Taylor, 1995. ”NR, not reported.

Arizona

Arkansas

California

Louisiana

Mississippi

Texas

27 78

36 85

40 92

53 96

38 82

19 68

17

15

4

4

18

6

5

NRb

4

NR

NR

26

23

15 100

33 95

22

100

100

20 97

98

NR

NR

NR

NR

NR

NR

NR

NR

5

NR

3

2

1

118

THOMAS J. GERIK ETAL.

use multiple N applications but remain reluctant to evaluate soil and plant-N status in determining the fertility needs of the crop? Our objective is to review cotton-N response and requirements, soil-N cycling, and soil- and plant-testing procedures.

II. COTTON GROWTH AND NITROGEN RESPONSE A. PLANTGROWTHHABIT The growth habit of a plant defines the timing of phenological events and the duration of important growth stages. The perennial growth habit and indeterminate nature of cotton is characterized by five growth stages that are interdependent and overlap (Table IV) (Mauney, 1986; Oosterhuis, 1990). These phenological growth stages are emergence, first square (floral bud), first flower, first open boll, and harvest. The timing and duration between each stage is closely associated with temperature. The growth habit of cotton is often described in terms of growingdegree-days or thermal units (Mauney, 1986). Leaf and fruit appearance follow a predictable pattern in the early stages of development (Mauney, 1986). Unless nutrient, water, or biotic stresses interfere, the plant grows unimpeded by producing a series of reproductive branches (also called sympodial branches) beginning at the sixth or seventh main-stem node. A mainstem leaf subtends each sympodial branch, and a leaf (called a sympodial leaf) subtends each fruit formed on successive nodes. New main-stem nodes and sympo-

Table IV Range of Published Growing Degree Days for Morphological Periods and Growth-StageEvents of Cotton Using a Base Temperatureof 15.3 “C” Phenological events and morphological periods Emergence Nonreproductive period First square Square period First flower Peak bloom period Boll period First open boll Harvest “Data from Mauney, 1986.

Duration of period (days)

Seasonal sum to phenological events (days)

45- I 30

45-130

350-450

480-530 740- 1 150 -

250-500 200-800 910-950 -

-

1690-2050 2550-4600

MANAGING COTTON NITROGEN SUPPLY

119

dial branches form approximately every 40 thermal units, and fruit appears on reproductive branches every 60-80 thermal units, depending on the cultivar (Hesketh et af.,1972; Jackson er af.,1988). Under ideal growing conditions (e.g., average air temperature of 30°C), successive main-stem nodes with sympodial branches usually appear every 3 days, and successive fruit on each sympodial branch appears every 6 days (McNamara er al., 1940; Kerby and Buxton, 1978). Thus, the growth habit results in a four-dimensional growth pattern in time and space (Mauney, 1986). Although the growth habit of cotton is indeterminate, fruit formation does not continue indefinitely-even in the absence of water, nutrient, and biotic stresses. Cessation of fruiting, commonly called cutout, typically occurs about 90 days after planting and is usually associated with the appearance of flowers in the upper canopy. Bourland ef al. (1992) found that white flower appearance on the fifth main-stem node from the apex of normal fruiting cotton plants signals the development of the last harvested boll of acceptable size and quality. Thus, five nodes above white flower ( 5 NAWF) may be definitive criteria for identifying cutout in cotton.

B. PLANTRESPONSE TO NITROGEN DEFICIENCY Plant response to N deficiency usually begins with limitations in uptake. Cotton only uses inorganic forms of N, either as nitrate (NO;) or ammonium (NH;). Nitrate is the principal source of N, since ammonium is quickly transformed in the soil solution to nitrate through nitrification when typical weather conditions for cotton prevail. Like most higher plants, cotton absorbs nitrate through the roots and transports it directly to the leaves in the transpiration stream. Once in the leaf, nitrate is reduced to ammonium and combined with organic acids to form amino acids and proteins. These processes require considerable energy in the form of reductants, like NADH, and a ready supply of organic acids from carbon assimilation. Up to 55% of the net carbon assimilated in some tissues is committed to N metabolism (Huppe and Turpin, 1994). Most attention has focused on the relationship between photosynthetic rate and leaf N (Fig. 1 ) (Natr, 1975; Radin and Ackerson, 1981; Radin and Mauney, 1986; Wullschleger and Oosterhuis, 1990). This probably arises from the most obvious visual symptom of N deficiencies-chlorosis, which increases with increasing N deficiency. Yet no direct evidence supports the hypotheses that lower chlorophyll content limits normal photosynthesis (Benedict et al., 1972). Nevertheless, N reduction and carbon assimilation processes are so interdependent that Huppe and Turpin (1994) concluded that neither could operate to the detriment of the other. For example, when N deficiency occurs, photosynthetic efficiency declines and assimilated carbon accumulates in the plant as starch and oth-

120

THOMAS J. GERIK E T A . h

8

'-

4.0

N -

% 3.5 F

3.0

d 2.5

?!t

Q)

F

2

::

0"

2.0 1.5

1.0

0.5 0

10

20

30

40

50

Leaf N content tma Nlka) Figure 1 The relationship between leaf-N content and the carbon exchange rate of greenhousegrown cotton (T.J. Gerik. unpub. data).

er carbohydrates (Rufty et al., 1988; Foyer et al., 1994). Carbohydrate accumulation in N-deficient plants is often greater in the roots than in other parts of the plant. When leaf-N supplies are replenished, there is a rapid increase in respiratory activity and starch degradation through the oxidative pentose phosphate pathway and tricarboxylic acid cycle (Smirnof and Stewart, 1985). Both pathways supply reductant and ketoacids (Ireland, 1990) for nitrate reduction and amino acid synthesis. Once available starch reserves are depleted, photosynthetic activity increases ( E M and Turpin, 1986; Syrett, 1981). Thus, leaf carbohydrate and N appear to signal the priority of metabolic substrates used in the different pathways controlling carbon and N assimilation. Thus, the suppression in photosynthetic activity by N deficiency appears to be transient. Equally important to declining photosynthesis are reductions in leaf expansion and leaf area and increased sensitivity to water stress when N deficiency occurs. Physiological responses of N-stressed cotton are similar to those encountered with water stress (Radin and Mauney, 1986). Similar to water stress, N stress decreases stomata1 and mesophyll conductance of CO, (Radin and Ackerson, 1981), decreases hydraulic conductivity, e.g., water uptake and transport in the plant (Radin and Parker, 1979; Radin and Boyer, 1982), reduces leaf expansion and leaf area (Radin and Matthews, 1989), increases starch and soluble carbohydrates in roots (Radin ef al., 1978), and decreases leaf osmotic and turgor potential (Radin and Parker, 1979). Given these similarities, scientists argue that the behavior of Nstressed plants prolongs plant survival after the onset of drought in three ways: (1)

MANAGING COTTON NITROGEN SUPPLY

121

by conserving water; (2) by redirecting assimilates from leaf to roots, thereby enhancing root growth at the expense of leaf growth; and (c) by redirecting assimilates into the formation of metabolites for osmotic adjustment (Radin et al., 1978; Radin and Parker, 1979). Yet these gains in water conservation or changes in assimilate allocation have limited value in sustaining economic yield, since early stomata1 closure reduces the plant’s photosynthetic capacity and ability to accumulate dry matter. Like most other higher plants, cotton absorbs more nitrate than is required to satisfy its metabolic requirements. It stores additional N as nitrate in leaf vacuoles (Smirnof and Stewart, 1985) and as additional leaf protein, such as ribulose 1,5biphosphate carboxylase, which is often found in prodigious quantities in chloroplasts. Because N is mobile, the change in plant-N status as soil-N supplies become limited is not abrupt, but is gradual as the plant uses its reserve N to satisfy the plant’s requirements when soil-N supplies are limited. Cotton’s response to N deficiency is well understood and falls into three areas: ( I ) altered photosynthetic rate, (2) altered leaf expansion, and (3) altered responses to water stress. Although altered photosynthesis has received the most attention, all three responses contribute to alterations in plant growth and yield. However, the mechanisms underlying reductions in carbon assimilation (photosynthesis) and growth of N-deficient cotton remain unclear. These mechanisms and yield can be greatly affected by different environmental conditions that occur each year (McConnell er al., 1993).

C. INTERACTIONS BETWEEN GROWTH HABITAND NITROGEN Although N deficiency does not usually influence the timing of phenological events, it has a negative effect on the number of leaves and fruit formed, thereby reducing the duration of the squaring and flowering period (Radin and Mauney, 1986). In this way, the plant-N status has a big impact on the accumulation and partitioning of dry matter during each morphological period. For example, Jackson and Gerik (1990) found that N fertility was highly correlated with leaf area and boll number but not with boll weight or the number of main-stem nodes (Fig. 2 ) . Vegetative growth, as evidenced by increases in the length and cross-sectional area of main-stem internodes, increases with applied N, resulting in greater plant height and weight without increases in boll number or yield (Wadleigh, 1944; Tucker and Tucker, 1968). Thus, large applications of N combined with excessive moisture early in the season can overstimulate vegetative growth, causing problems with mechanical harvest, increased shielding of floral buds, (squares) and small bolls (Walter et al., 1980) and contributing to delayed maturity and rot of lower bolls. But growth regulators, such as mepiquat chloride, retard internode

122

THOMAS J. GERIK ETAL.

.-E X

2

60

'c

0

CI

n

+Boll weight 20

+Main-stem nodes 6 Boll number

v Leaf area

elongation (Walter et al., 1980; Reddy et al., 1992), thereby counteracting some of the adverse effects of high N and excessive soil moisture on plant height and internode length. Boll number is the most important factor correlated with yield (Morrow and Krieg, 1990), with the number of bolls per plant and the number of plants per square meter contributing equally. Leaf development (e.g., vegetative growth) and reproductive developments (e.g., square production) occur simultaneously (Fig. 3). Jackson and Gerik (1990) found that the leaf area and boll canying capacity were linearly related, with approximately 0.1 m2 of leaf area required to maintain a boll in greenhouse-grown plants. Yet the leaf area required for each boll may be lower in the field. More recent studies suggest that only 0.02 m2 of leaf area per boll may be required under optimum field-growing conditions (Bondada et al., 1996). Vegetative growth in terms of leaf number and leaf area and boll number are tightly coupled-in the formation of fruiting sites and by providing N and carbon assimilates to support growing bolls. Nitrogen deficiency dui ing the critical fruiting period from first square (i.e., >2 mm) to peak flowering (e.g., typically 40-85 days after planting; see Fig. 3) has a large adverse affect on cotton yield. Furthermore, excessive N contributes to excessive leafiness, which may adversely partition carbon assimilate away from bolls within the canopy-specially in cloudy weather midseason. Cotton bolls have high N requirements. The seed, which accounts for about half of the total dry weight of the boll, contains almost twice the N concentration as corn-3.3% N in cotton compared with 1.75% in corn. However, bolls have low

‘:I 1 MANAGING C O n O N NITROGEN SUPPLY

123

A

xi’

+Total canopy

30

8 X

Sympodial leaves Main-stem leaves

-2+3-

I

I

I

I

I

I

I

1

0

20

40

60

80

100

120

140

160

0

20

40

60

80

100

120

140

160

180

1000 800

v

600 0

ti

400

200

s

o

5 180

Days after planting Figure 3 Typical leaf growth (A) and fruit development (B) of cotton grown in the United States (Oosterhuis, 1990).

nitrate reductase activity for reducing NO,-N to ammonium for protein synthesis (Radin and Sell, 1975). The sympodial leaves, which subtend bolls, and the vegetative leaves, attached to the main-stem, are the primary sources of assimilated N for growing bolls (Oosterhuis et al., 1989). Even when N supplies are sufficient in relation to the developing boll load, cessation of vegetative growth and fruit formation ( e g , cutout) occurs, provided the plant is well “fruited” with bolls. In the mid-South region of the United States, the cessation in growth occurs 90-100 days after planting or when the first flower appears on the fifth main-stem nodes from the apex (Oosterhuis, 1990).This hiatus in growth is thought to be associated with the decline in leaf N and carbon assimilation capacity due to leaf age (Table V, Zhu and Oosterhuis, 1992; Wullschleger and Oosterhuis, 1992) and the increasing assimilate requirements of developing bolls for assimilated carbon and N. The declining N and carbon assimilation capacity and increasing boll assimilate requirement temporarily halt vegetative growth and fruit formation. The plant’s inability to fully satisfy the assimilate requirements of growing bolls suggests that weaker sinks, like squares and small bolls (e.g., bolls < 10 days in

124

THOMAS J. GERIK ETAL. Table V Percentage of Leaves within a Cotton Canopy by Age and Level of PhysiologicalActivity at Three Growth Stages" Days after planting 60

Leaf age (days after unfolding)

0-14 15-28 + 29

( 1st square)

Physiological activity Sink Strong source Declining source

90 (anthesis)

120 (boll fill)

Leaves (%) 36 38 26

II 21 68

3 II 87

OData from Oosterhuis, 1990.

age), might be shed, thereby reducing boll number and cotton yield. Plant maps of Wadleigh ( 1944) support this deduction, whereby N deficiency slightly increased the percentage of fruit shed on the first sympodial nodes and drastically increased fruit shed on the remaining nodes; whereas sufficient N increased the number of harvested bolls at these nodes. However, Jackson and Gerik (1990) found that shedding of squares and young bolls was proportional to the ratio of the number of actively growing bolls to the plant's boll carrying capacity (Fig. 4). The data

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1

Growing bolls / Carrying capacity Figure 4 The relationship between boll shedding and the ratio of actively growing bolls to the plant carrying capacity (e.g., maximum boll number) (Jackson and Gerik, 1990).

MANAGING COTTON NITROGEN SUPPLY

125

2000 ,

-

r 0

u)

.-

18001700 -

4-

C

3 1600-

c

m

1500

0

h

f

4600

50

100

150

200

250

-

MOO-

B

2 4200.

6 4000iE 3800+

~

-rr8 Q)

360034003200 30002800 0

Applied nitrogen (kg/ha) Figure 5 The effects of applied N on the time to harvest and cotton yield (reprinted from McConnell eta!., 1993, by permission of the publisher).

suggest that N deficiency does not increase fruit shedding beyond the levels already imposed through reductions in leaf area and the number of fruiting sites. Nitrogen can have a big impact on crop maturity by prolonging vegetative growth and fruit formation (Jackson and Gerik, 1990; Gerik et al., 1994) and by delaying harvest. Yet delay in maturity due to excessive N does not always result in higher yield (Fig. 5 ) . Data from McConnell et al. (1993) illustrate that N rates beyond the N optima (e.g., 112 kg N per hectare in this case) delayed harvest without an increase in yield. It is important to note that the maturation period of individual bolls (e.g., the time from anthesis to open boll) does not appear to be altered by N deficiency (Table VI). Therefore, delays in harvest beyond the point representing the yield maximum and N optima are a function of continued vegetative growth-fruit formation beyond the limitations in growing season for maturation of young fruit. Thus the length of the growing season, plant density, water supply, and the cultivar’s yield potential must be in balance with the N supply to maximize yield (Maples and Frizzell, 1985; Morrow and Krieg, 1990).

126

THOMAS J. GERIK ETAL. Table VI

Average Boll Periods for Bolls That Flowered at 10-Day Intervals for the Cotton Cultivar Stoneville 213 Grown at Five N Fertilization Levels in a Greenhouse" Days from flower appearance Applied nitrogen (mM/week)

58

0 18 36 72 144 Pooled error

43.8 44. I 45.3 44.5 45.3 1.8

68

78

88

Mean

-

44.1 43.6 44.8 43.3 43.5 2.4

Boll period (days)

44.3 43.6 44.8 43.6 42.7 2.4

44.0 43.5 43.6 42.7 42.4 2.4

38.3 40.6 1.5

"Data from T. J. Gerik. unpub.

D. PLANTNITROGEN REQUIREMENTS It is evident from the preceding discussion that a balanced interdependence exists between cotton growth and N uptake. Nitrogen uptake is proportional to the plant's photosynthetic capacity and dry-matter accumulation. Nitrogen requirements and distribution within the plant have been studied (Wadleigh, 1944; Bassett er al., 1970; Halevy, 1976; Oosterhuis er al., 1983; Mullins and Burmester, 1990). Most recent reports indicate that cotton requires 16-20 kg of N per 100 kg of lint (Table VII). This translates into N use efficiencies of 5.0-6.6 kg of lint per kilogram of N. However, large discrepancies exist between recently published reports (Table VII). This suggests that further research is needed to establish the plant-N requirement and factors responsible for this variation. Dry-matter accumulation and yield of cotton are highly correlated to seasonal evapotranspiration (Orgaz er al., 1992).Thus, water supply is critical in maintaining the crop's photosynthetic capacity and growth. Morrow and Krieg ( 1990)studied the interaction of water and N supply on cotton yield in a short-season environment. Although they reported that increasing N supply from 0 to 100 kg N/ha at peak flowering increased yield regardless of irrigation intensity, sufficient water supply during the fruiting period was the most important factor leading to increases in boll number and yield. Furthermore, they found that maximum cotton yields were obtained when N and water were applied in a ratio of 0.25 kg N ha-' mm-I H,O. However, the ratio of N to water received should not change unless cultural practices or climate substantially alter evaporative water loss or unless losses in soil N occur as a result of denitrification or leaching. Similar studies in long-season environments (>150 days) have not been reported.

MANAGING COTTON NITROGEN SUPPLY

127

Table VII Reported N Requirements for Cotton Lint Production ~

Location

Lint yield requirement (kg N/100 kg lint)"

Source

Texas Mississippi Georgia California Israel Arkansas Unknown Zimbabwe Texas Alabama

Fraps, 1919 McHargue, 1926 Olson and Bledsoe. 1942 Bassett ef al., 1970 Halevy, 1976 Maples et al., 1977 Olson and Kurtz, 1982 Oosterhuis ef al., 1983 Morrow and Krieg, 1990 Mullins and Burmester, 1990

~~~~~

Lint yield efficiency (kg lint/kg N)"

25.0 16.0 29.0 10.0 13.2

4.00 6.25 3.45 10.00 7.50 10.00 6.25 4.9 I 6.60 5.00

10.0 16.0 20.3 16.6 19.9

OData from MacKenzie et al., 1963. "Data from Gardner and Tucker, 1967.

The N requirement of the cotton plant varies with the growth rate and growth stage (Fig. 6 ) . Before flowering, cotton leaves contain 60-85% of the total N, but the N content declines after flowering as it translocates from leaves to developing bolls. At maturity, the fiber and seed removed in harvest contain almost half of the total N accumulated in the shoot during the growing season (i.e., about 42% of the total above-ground N) (Oosterhuis et al., 1983).Thus, cotton N requirements 160

1

140 -

. g r"

h

120 100 -

a

5 +

80 -

3

60

a

ctures

-

C

.-P z

40-

c

20-

Reproductive structures

0-

I

I

I

I

1

I

I

I

I

I

0

20

40

60

80

100

120

140

160

180

Days after planting Figure 6 Cumulative N uptake for the total plant and the vegetative and reproductive structures during the growing season (Oosterhuis ef al.. 1983).

THOMASJ. GERIK ET AL.

128

are highest during the latter growth stages, when N supplies typically diminish and root activity is less.

JII. SOIL NITROGEN AVAILABILITY AND DYNAMICS Most N for cotton growth is supplied from soil nitrate, although the plant uses ammonium when available. Less than 4% of the cotton acreage in the United States receive foliar N application (Table 11). Therefore, the interacting chemical, physical, and biological functions in the soil are extremely important in determining N supply to the cotton plant. Taken as a whole, the interacting biological processes in the soil are termed the N cycle. A simplified version of the N cycle is depicted in Fig. 7; a more detailed description of N processes in agricultural soils can be found in Stevenson (1982a,b), Important functions of the soil-N cycle are the N-transformation process, such as (1) NH; adsorption, (2) nitrification, (3) immobilization,and (4) mineralization; the N-input processes, such as (1) N fertilization and (2) biological N, fixation; and the N-outjlow processes, such as (1) denitrification,(2) leaching, and (3) plant

I

Fertilizer Application

Ads

Soil Organic Matter Including Microorganisms and Crop Residues Figure 7 The fate and utilization of applied-N fertilizer within the plant-soil continuum

I

MANAGING COTTON NITROGEN SUPPLY

129

uptake. These processes are interdependent and influenced by soil type and environmental conditions such as soil temperature and moisture content.

A. INORGANIC Son, NITROGEN Nitrogen in the NH; form is present either in soil solution or adsorbed to soil clay particles (Nommik and Vahtras, 1982). The adsorption of NH; is a chemical process that depends on soil cation exchange capacity (CEC). The amount of NH; in soil solution depends on the interaction between the CEC and the total cation concentration adsorbed and in soil solution and the NH; concentration adsorbed and in soil solution. In general, as the quantity of NH; in soil solution declines, the more NH; will be released into the soil solution from clay particles. While NH; is adsorbed, it is not susceptible to the soil N transformations, including plant N uptake. In some soils, NH; can be bound so tightly to clay particles that it cannot be readily released into the soil solution for plant uptake. When this occurs, the NH;t is considered to be fixed or nonexchangeable. Ammonium in solution is subject to immobilization or nitrification as well as plant uptake. Immobilization is a microbial process of converting N from the inorganic form into organic forms of N (Jansson and Persson, 1982). This occurs by the uptake of NH; by microorganisms to be used in their growth process. Microorganisms, like any living organism, use N in the synthesis of DNA, proteins, and other organic constituents. The source of N for microorganisms (similar to plants) is from inorganic N (NO; and NH;) in the soil solution. Nitrification is the microbial process of converting NH; into NO; with a release of energy to the microorganisms (Schmidt, 1982). The process is carried out primarily by Nitrosomonas bacteria, which convert NH: to NO,, and by Nitrobacter bacteria, which convert NO; to NO;. Since these two bacteria depend on these processes for energy, the conversion of NH; to NO; proceeds very rapidly as long as warm soil temperature and moisture are adequate. Nitrate in soil solution is the most prevalent form of N associated with plant N uptake, immobilization, leaching, and denitrification within cotton production systems. Immobilization of NO;, as with NH;, is the utilization of soil nitrate N by the soil microorganisms for their normal growth processes (Jansson and Persson, 1982). Most plant N uptake is in the form of NO;, because warm and moist growing conditions favor rapid conversion of NH; to NO;. Because of this, many plants have adapted to NO;. Leaching is the physical movement of NO; through the soil (Stevenson, 1982a). As water moves through the soil, so does the NO; in solution. In soils where water moves rapidly through the soil profile and where water (either from rainfall or irrigation) exceeds evapotranspiration, nitrate is commonly found below the rooting zone and is no longer available for plant uptake.

130

THOMASJ. GERIK ETAL.

Denitrification is a microbial process of converting NO; into gaseous N,O or N, (Firestone, 1982). This process occurs when soil is water saturated and microorganisms no longer have ready access to 0,. Most microorganisms depend on 0, for energy conversion by utilizing 0, as the last electron acceptor, thereby converting 0, to CO,. However, certain microorganisms can also utilize NO; in the same way as 0, in anaerobic conditions (due to water saturation), thereby utilizing NO; as the last electron acceptor when converting NO; to N,O or N,. Large amounts of NO; can be lost from the soil to the atmosphere through this process. Losses from irrigated soils in California ranged from 95 to 233 kg N ha-' year-' (Ryden and Lund, 1980).

B. SOILORGANIC NITROGEN While cotton plants primarily use inorganic N, over 90% of soil N is usually held in the organic form (Stevenson, 1982b).This is part of the natural soil organic matter, which includes a mixture of plant residue (such as leaves and roots) in various stages of decomposition and microorganisms (both living and dead). As the organic matter decomposes, NH: is released into the soil solution through mineralization. In certain years, organic N is a key source. As organic matter progressively decomposes, it supports fewer microorganisms, because the energy content of the remaining organic materials declines and becomes more difficult to decompose. The result is greatly reduced microbial activity and highly decomposed organic materials, called humus. However, N-rich and biologically active phases of soil organic matter are continually renewed through the addition of plant roots and other crop residues. The result is a continuous process called mineralization immobilization turnover (MIT) (Jansson and Persson, 1982), whereby plant material and dead microorganisms are being decomposed, resulting in mineralization, and new NH; and NO; are being assimilated into microorganisms, resulting in immobilization. The MIT plays a pivotal role in the N nutrition of the cotton plant. It is estimated that only 10-15% of the total soil organic N is subject to the biologically active MIT processes. Since only a small portion of the total N required by the plant is available in the soil solution at any time during the life cycle, the mineralization process must continually replenish the soil solution with NH; to meet the plant demand. In heavily fertilized cotton fields in the humid southern United States, the fertilizer N supplied only half of the N found in the plant, with the other half coming from N already present in the soil from organic and previous fertilizer additions (Torbert and Reeves, 1994). For a short-season cotton grown in a semi-arid environment, Morrow and Krieg (1990) found that the residual soil N accounted for 35% of the final yield.

MANAGING COTTON NITROGEN SUPPLY

131

The fertilizer requirement of cotton is dependent on weather and differs between soil type and year, in large part because of variation in MIT and other components of the N cycle. The complexity, resulting from the interaction of N-cycle components and soil type and climate, makes the prediction of N fertilization very difficult and necessitates long-term N-rate and crop-response experiments to optimize N application with crop yield. Most commercial fertilizers are either nitrate or ammonium in some form or mixture. For example, application ammonia nitrate (NH,NO,) will dissociate into the NO; and NH1; ions in soil solution. Once applied, fertilizer N is rapidly incorporated into the N cycle, where it functions under the same constraints as the N already present.

c. NITROGEN ADDITIONS T O THE SOIL Commercial fertilizer N consists of many different forms of inorganic N, but most is applied as one or in combination with the following forms: anhydrous ammonia, ammonium sulfate, ammonium phosphate, urea, ammonium nitrate, and potassium nitrate (Jones, 1982).Anhydrous ammonia, ammonium sulfate, and ammonium phosphate rapidly dissociate in soil solution and enter the soil-N cycle as ammonium. The enzyme urease dissociates urea and forms two gaseous molecules of ammonia. If this conversion occurs in soil solution, the NH, quickly forms NH1; and enters the soil-N cycle; however, if urea is applied to the soil surface or to plant residue, substantial gaseous losses of the NH, can occur (Jones, 1982).Fertilizer applications of ammonium nitrate will dissociate in soil solution and enter the soil N cycle in both the NH1; and the NO; form, while potassium nitrate will dissociate into the NO7 form. Nitrogen applications are often also made in organic N forms, including foodprocessing waste, municipal waste, and animal manures (Jones, 1982). While these materials contain some portion of NO; and NH:, a substantial portion of the N will be in the organic N form and will enter the N cycle in the organic pool (active MIT pool) and have to be converted through mineralization into NH1; before plant uptake. The makeup and content of various N forms with animal manure and municipal waste depend on both the source and the handling characteristics of the material before application. Another method of adding N to soil for cotton production is using crop rotation with legumes (Reeves, 1994). Legume species are capable of fixing atmospheric N, by means of a symbiotic relationship with soil microorganisms. Therefore, legumes grown as a cover crop or in rotation with cotton provide an additional source of N. The N attributed to soil from the legume is estimated as N-fertilizer equivalents-i.e., the amount of fertilizer N likely to be replaced by legume

132

THOMASJ. GERIK ETAL.

N, fixation. The N-fertilizer equivalents typically range between 60 and 100 kg ha-' but depend on the legume species and the climate conditions during the legume and cotton growing season (Reeves, 1994).

W. FOLIAR-NITROGEN FERTILIZATION IN COTTON Foliar application of nitrogen to cotton (Gossypium hirsutum L.) has frequently been used mid-to-late season across the U.S. Cotton Belt to supplement plant N requirements. It has been suggested that foliar-applied N may serve as an N supplement to alleviate N deficiency caused by low soil-N availability, to provide cotton plants with the N required by the rapidly developing bolls, and to avoid possible hazards of excessive vegetative growth resulting from excessive soil N (Hake and Kerby, 1988; Miley, 1988). However, reports on the effects of foliar-applied urea on cotton yields have been inconsistent (Anderson and Walmsley, 1984; Smith et al., 1987), and information on the absorption and translocation of foliarapplied N in cotton is limited. Urea is the most popular form of N used for foliar fertilization in cotton production. Yamada (1962) reported that the greater effectiveness of urea when applied to foliage resided in its nonpolar organic properties. Urea, containing 15Nlabel, has been employed to measure rates of absorption and translocation of foliar-applied N, because it permits direct determination of the uptake and translocation of foliar-applied N. Oosterhuis et al. (1989) and Baolong (1989) reported that the sympodial leaf rapidly took up foliar-applied N. They found that 30 and 47% of applied N was recovered within 1 hour and 24 hours after application, respectively. Approximately 70% of the foliar-applied urea N was absorbed by 8 days after application. There have been similar reports for olive (Olea europaea L.) (Klein and Weinbaum, 1984), coffee (CofSeaarabicu), cacao (Theobroma cacao), and banana (Musa acuminafa) (Cain, 1956), greenhouse-grown tobacco (Nicotiana tabacum L.) (Volk and McAulliffe, 1954), and soybean (Glycine max L.) (Vasilas etal., 1980). Foliar-applied I5N to cotton was rapidly translocated from the closest treated leaf to the bolls and was first detected 6 hours after application (Baolong, 1989). The increase in I5N in cotton bolls coincided with a progressive decline in percentage of 15N recovery in the treated leaves. Rapidly developing fruits were the major sinks of foliar-applied N with bolls closer to the site of application (firstfruiting position on the sympodial branch) being a much stronger sink than the next closest boll (second position) along the branch. Baolong (1989) found about 70% of the total foliar-applied 15Nurea was found in the cotton bolls, with less than 5% remaining in the leaves, petioles, bracts, and branches.

MANAGING COTTON NITROGEN SUPPLY

133

Foliar-applied urea solution will reach many different cotton organs of varying ages when applications are made to the canopy. Baolong (1989) also showed that main-stem leaves, sympodial leaves, and bolls were all capable of absorbing foliar-applied urea, regardless of physiological age. Bondada et al. (1997) demonstrated correlation between increasing leaf cuticle thickness as the leaf aged and decreased absorption of foliar-applied "N. Absorption was reported to be more rapid in young leaves than in old leaves for coffee, cacao, and banana (Cain, 1956) and apple (Maluspumila) (Miller, 1982), although Boynton et al. (1953) found no significant difference in the absorption of foliar-applied urea applied to apple leaves of different age. Many factors can affect the uptake of foliar-applied urea, including the condition of the leaf and the prevailing environment. It has been shown that the leaf water status affects the physical structure of the cotton leaf cuticle (Oosterhuis et ul., 1991) and consequently affects the absorption of the foliar-applied nutrients (Wittwer et al., 1963; Boynton, 1954; Kannan, 1986). Baolong (1989) showed that water deficit stress impeded the absorption of foliar-applied urea N by sympodial leaves, as well as the subsequent translocation within the branch. Furthermore, applications made either late afternoon or early morning was more effectively absorbed than those made at midday, and this was more pronounced for waterstressed plants. This was associated with crystallization of the urea on the leaf surface and also with changes in the cuticle caused by water stress. Bondada et al. (1994) demonstrated the importance of the size of the developing boll load in determining plant response to foliar-N fertilization. The location of the foliar-N spray within the canopy affected uptake and lint yield in cotton (Oosterhuis et al., 1989). There was a significant increase in yield when 15N-urea was applied to the top of the canopy compared to the lower canopy. This was probably due to the larger N requirement of developing bolls in the upper canopy late in the season. These results show that N-deficient cotton can benefit from foliarapplied N. However, indiscriminate application of N without due consideration of soil N availability, plant-N status, and environmental conditions can be wasteful.

V. MONITORING COTTON NITROGEN STATUS The uncertainty in soil-N availability generated by the N cycle and seasonal changes in plant-N utilization makes it difficult to predict the crop's N requirements. Methods of monitoring soil and plant-N status have been developed to alleviate these difficulties. These methods include both direct and indirect measurements of soil and plant mineral N and plant response to fertilizer. For soils, direct methods include fertilizer tests and measurement of soil nitrate, ammonium, or-

134

THOMASJ. GEIUK ETAL.

ganic matter, and carbon mineralization. For plants, direct methods include measurements of petiole nitrate concentration, total leaf N, and nitrate reductase activity. Indirect methods include measurements of leaf chlorophyll content and use of physiologically based crop models.

A. FERTILIZER TESTS Soil fertilizer tests are the oldest and most widely used method of determining fertilizer-N requirements and developing recommendations for cotton. These tests indirectly account for the mineralization potential of soil, N leaching, denitrification, N immobilization, availability of fertilizer N, and climatic variability on N uptake and yield. They involve empirical measurement of yield response to increasing levels of applied N fertilizer from experiments conducted at specific locations over several years. A fertilizer-response equation (typically curvilinear) developed with data from these experiments enables the user to estimate the amount of fertilizer required to attain an anticipated or projected yield (Fig. 5B). When tests are conducted in combination with crop rotation, manure application, or legume rotation, N credits are estimated and used to adjust predicted N requirements for crop history or organic fertilizer. Yet changes in weather (i.e., precipitation and temperature) and cultural practices (i.e., tillage, variety, fertilizer formulation, plant density, row spacing, crop rotation, and pest management) from those experienced during the fertilizer test limit the accuracy with which the equation can predict future fertilizer requirements. Ideally, an independent evaluation of the crop yield and fertilizer-N response should be conducted whenever changes in edaphic and crop-management factors occur. At best, fertilizer tests are retrospective estimates of the crop’s N requirement and should be viewed as a calibrated response of crop yield to the applied N fertilizer for the soil type under the prevailing growing conditions. Although the analyses are often generally applied to estimate N fertilizer needs of cotton and other crops, their application should be confined to the location and soil type where testing was performed. They should be conservatively used to estimate the N requirements of cotton.

B. SOILTESTWG Soil analyses are direct measurements of the soil-N status. Several approaches have been adopted to directly assess soil N (Stanford, 1982). In western states, where arid conditions prevail, soil NO; analyses have been successfully used to determine existing N levels and to adjust N-application rates. Other locations have

MANAGING COTTON NITROGEN SUPPLY

135

adopted a preseason soil NO; test for adjusting fertilizer-N rates based on existing NO; levels. As with fertilizer tests, soil analyses should be confined to the general location and soil type where testing was performed. Methods to assess the organic-N mineralization potential of soil have been developed (Nadelhoffer, 1990; Torbert and Wood, 1992).These procedures typically require soil incubation to assess microbial biomass or microbial activity by measuring the CO, evolution from the soil. Yet adoption of these methods by soil-testing laboratories has been limited by inefficiency and high costs resulting from cumbersome soil-handling procedures, sample turnaround time, and procedure inaccuracy. Recent improvements in methods of assessing the N-mineralization potential of soil samples may eliminate some of the impediments (Franzluebbers et al., 1996).

C. PLANT-TISSUE ANAL,YSES Plant-tissue analyses were developed to overcome variation inherent in fertilizer tests and soil analyses. For cotton, tissue analyses supplement information from soil analyses and from soil mineral and organic N analyses, enabling growers to better manage the crop-N content after flowering. From a practical perspective, the procedure must be economical and simple, and the results must be quickly available. Establishing critical reference points is the first step in diagnosing the N deficiency using tissue analyses. However, identifying the critical value that imparts N-deficiency response is difficult. Because plant-N levels are dynamic-they change over the growing season; differ between years; differ among organs, plant age, and growth stage; and differ between genotypes-the critical point cannot be considered a single value but must be interpreted as a range of values within which to work. In the following sections, we discuss tissue-analysis procedures that have been used to monitor the N status of cotton.

D. PETIOLE NITRATE ANALYSIS Petiole nitrate analysis is the most popular plant-tissue assay to ascertain the N status of cotton (Tucker, 1965; Gardner and Tucker, 1967; Miley and Maples, 1988). Its popularity arises from the speed and simplicity of analysis. Because cotton absorbs more nitrate than any other source of N, the petiole nitrate test measures the nitrate levels in xylem vessels in the petiole, estimates the flow of N from the root to the leaf, and indirectly estimates the nitrate levels in the soil solution. Tissue samples for petiole nitrate analysis usually comprise 20-30 petioles from

136

THOMAS J. GERIK ETAL. Table VIII Qpical Cotton Petiole Nitrate Concentrations Reported in the U.S. Cotton Belt

Arizona' Growth stage First square First Rower Fust large boll,midflower Fust open boll,late Rower

California Acala"

Texasb

Acala

DPL 16

Arkansasd

16,000 8000 2000

16,000 8000 2000

15,000 12,000 6000-8000 4000

18,000 14,000 8000-10,000 4000

12,000-28,M)O 8000-24,OOO 5000-15,OOO 200(M000

"Data from MacKenzie et al., 1963. bData from Longenecker et ab, 1964. 'Data from Gardner and Tucker, 1967. "Data from Sabbe and Zelinski, 1990.

young, fully expanded main-stem leaves collected from the third or fourth mainstem node from the apex. Nitrogen-deficiency symptoms do not usually appear, nor will growth decline until petiole nitrate levels fall below 2,000 Fg/g (Hearn, 1986). Petiole nitrate analyses cannot determine the total amount of N used by the plant prior to sampling but reflect the amount of nitrate N taken up by the plant from the soil solution. Petiole nitrate levels must be used and interpreted with care, because they vary with cultivar, growth stage, soil type, weather, and insect damage (Table VIII; MacKenzie etal., 1963;Longeneckeret al., 1964; Gardner andTucker, 1967; Baker et al., 1972; Oosterhuis and Morris, 1979). Because water and nitrate uptake occur simultaneously, petiole nitrate samples should be collected when soil moisture or sunlight does not limit leaf gas exchange and transpiration. Zhao (1997) showed that petiole nitrate N in cotton increased by 50%after 1 day of simulated overcast weather (i .e., 60% reduction in incident radiation). Petiole nitrate levels decrease during the growing season, typically decreasing from about 18,000 to 1000 Fg/g from early square to maturity (Fig. 8 and Table VIII). These ontogenic changes are associated with declines in root-uptake activity, increased N demand of growing bolls, and lower soil nitrate levels. In western Texas, Sunderman er al. (1979) found that petiole nitrate variation was lowest and yields were best correlated when plants were sampled at flowering. Weekly measurements have been recommended during the important growth stages to reduce the variability associated with petiole nitrate analyses (Maples et al., 1990). Care should be taken to assess and report the crop-water status, growth stage, plant-yield status (i.e., boll load), and efficiency of insect control at the time of sampling (Maples et al., 1990).

MANAGING COTTON NITROGEN SUPPLY

137

.-0..125 kg N/ha

-2

-1

0

1

2

3

4

5

6

7

8

Weeks after first flower Figure 8 Comparison of four applied-N rates with petiole nitrate (NO,) concentration at sequential time over the flowering and boll-maturation period of cotton (William Baker, Univ. of Arkansas,

pers. comm.).

E. TOTAL LEAFNITROGEN ASSAY Determining total N content of the most recent fully expanded cotton leaves in the upper canopy is probably one of the most reliable methods to ascertain the plant’s N status. It is a direct measure of leaf-N status and provides an estimate of N accumulated prior to sampling, given the mobility of N within the plant. As with the petiole test, total leaf N vanes with cultivar, growth stage, soil type, and weather and can be influenced by insects (if boll damage is severe). On a dry-weight basis, leaves are usually considered deficient in N if they contain less than 2.5% N, low in N if they contain 2.5-3.0% N, sufficient in N if they contain 3.0-4.5% N, and very high or excessive in N if the N content exceeds 4.5% (Sabbe et al., 1972; and Sabbe and MacKenzie, 1973). However, total leaf-N assays do not have the ease of sampling and handling that petiole sampling have, and the increased cost and time required has discouraged its use and limited its acceptance as a tool in monitoring commercial cotton fields.

F. NITRATE REDUCTMEACTIVITY Nitrate reductase is the enzyme that catalyzes the first step in reduction of nitrate N to organic forms within the plant, and it is thought to reflect the level of N activity in leaves (Beevers and Hageman, 1969; Lane et al., 1975). In comparing leaf nitrate reductase activity with petiole nitrate concentration, Oosterhuis and

138

THOMAS J. GERIK ETAL.

Bate (1983) found that the nitrate reductase assay was a more sensitive and reliable indicator of plant-N status. However, the nitrate reductase assay is too expensive and time consuming to be used routinely for assessing N levels of commercially grown cotton.

G. CHLOROPHYLL DETERMINATION Chlorophyll, an N-rich pigment molecule in leaves, converts light into the chemical energy needed to drive photosynthesis. Scientists have long known that leaf chlorophyll and N content were correlated. However, chlorophyll determination has not been considered practical for commercial plant-N analyses because it requires timely extraction of fresh leaf tissue with volatile organic solvents. The development of the SPAD-502 chlorophyll meter by Minolta Camera Co., Ltd., Japan, has renewed interest in the use of chlorophyll content as an indicator of plant-N status. This hand-held device nondestructively estimates the chlorophyll content of leaves by measuring the difference in light attenuation at 430 and 750 nm. The 430 nm wavelength is the spectral transmittance peak for both chlorophyll a and b, whereas the 750 nm wavelength is in the near-infrared spectral region where no transmittance occurs. The chlorophyll meter provides the means to indirectly determine plant-N status without destructive sampling and laboratory analysis. Recent reports by Tracy et al. (1 992) and Wood et al. (1 992) were encouraging and confirmed that leaf chlorophyll contents measured with the SPAD-502 and leaf-N contents were correlated for field-grown cotton. However, further research is needed to determine the strengths and limitations of this new technique.

VI. MANAGING COTTON NITROGEN SUPPLY A. COMPUTER MODELS Soil-N analyses and fertilizer tests provide retrospective assessments of the soil and plant-N status, and tissue analyses are instantaneous “snapshots” of the plantN status. Crop-simulation models are the only tools that simultaneously integrate the interacting soil, plant, and weather factors important in determining soil-N availability and crop demand for estimating current and future N needs. Several crop-simulation models have been developed and documented to assist in N management of cotton. These models include GOSSYM (Baker et al., 1983), OZCOT (A. B. Hearn, pers. comm., CSIRO, Narrabri, NSW, Australia), EPIC (Williams et al., 1989) and ALMANAC (Kiniry etal., 1992). GOSSYM is the most widely used and accepted cotton model (Albers, 1990). Albers (1990) conducted a survey of

139

MANAGING COTTON NITROGEN SUPPLY 1000

1

I

m

h

f 800 0, 5

5.-r.

600 -

c.

.-c

400 -

0

"5

200 -

0 0

Observed Simulated Observed: Y= 700 + z.zx - o.oo8x2; P= 0.39

-

GOSSYM users and found that 76% of farmers who used the model changed their N-management practices. The estimates of the crop-N utilization, yield, and soil-N availability have been tested with independent field measurements for GOSSYM but not for the other models. Stevens et al. (1 996) reported that GOSSYM overestimated soil-N availability by 10-30 kg N ha-', overestimated fertilizer N recovery, and underestimated cotton yield (Fig. 9). However, GOSSYM does not currently simulate MIT processes or ammonia-volatilization losses from soil or plants (Boone et al., 1993, which could explain the overprediction of fertilizer-N recover. EPIC and ALMANAC have the ability to simulate the N MIT processes, leaching, and volatilization from the soil (Williams et al., 1989; Kiniry et al., 1992), but N uptake or the response of cotton yield to N fertilizer has not been validated. Although crop-simulation models have potential to assist in making fertilizerN decisions, most have not been validated to determine their accuracy and precision in estimating plant uptake and soil-N availability. Validation studies must be conducted to ensure confidence in the accuracy of the simulated estimates under varied environmental conditions and to identify areas needing improvement.

B. MANAGINGNITROGEN SUPPLYWITH CROP-WATER USE Basing N fertilization on crop-water use may be another means of balancing the N demand of the crop with supply. It is well established that seasonal evapotranspiration is highly correlated with dry-matter accumulation and yield of cotton

140

THOMAS J. GERIK ETAL.

6

I

I

I

I

I

I

I

I

I

0

20

40

60

80

100

120

140

160

180

120,

halvest

B

I 0

20

40

60

80

100

120

140

160

180

Days after planting Figure 10 Comparison of cumulative water use (A) (Grimes and El-Zik, 1982) with plant-N uptake (B) (Olson and Bledsoe, 1942) during the growing season.

(Orgaz et af., 1992) and most other field crops. Furthermore, the cumulative cropwater use and N uptake of cotton follows a similar pattern (Fig. 10). The findings of Morrow and Krieg (1990) from the Texas high plains support this concept. Their data illustrate the curvilinear response of cotton lint yield to water supply and N, but they found a linear decline in the water-use efficiency of lint production with water supply during the fruiting period (Fig. 11). They reported that lint production increased 0.016 kg lint mm-' H,O for each additional kilogram of N per hectare applied during the fruiting period. Although, their growing season is shorter than most other U.S. cotton-growing regions, sufficient thermal time is available (e.g., 1250 thermal units with a threshold of 15°C) to achieve potential yields of 1000 kg lint ha-'. Because Morrow and Krieg (1990) obtained maximum yield by applying 400 mm water and 100 kg N ha-' during the fruiting

MANAGING COTTON NITROGEN SUPPLY

141

10 0

8

ON I

POSTN=O POSTN=BO

6

c

'E €

4

c

.C

9

2 0

-2

I

I

I

I

I

0

100

200

300

400

500

Water supply 61-120 DAP (mm) Figure 11 The effect of water supply and N on the water use efficiency of cotton lint production during the critical fruiting period, 61 to 120 days after planting (DAP) (reprinted from Morrow and Krieg, 1990, by permission of the publisher).

period, they concluded a ratio of 0.25 kg N ha-' mm-' H,O during the fruiting period was necessary to obtain maximum cotton production in their environment. Earlier, Grimes ef al. (1969) and Grimes and El-Zik (1982) reported a curvilinear response of cotton lint yield to irrigation and N and found that the water-use efficiency of lint production improved, in some cases, with applied N. Yet, Grimes et al. ( 1969) did not account for the total N supplied to the cop (e.g., the residual soil N supplied to the crop), nor did they consider that cotton growth stages might influence the water-N response as did Morrow and Krieg (1990). Morrow and Krieg's interpretation has merit because it parallels our fundamental understanding of the interaction of water and N on physiological and morphological processes associated with cotton yield. A relationship between N-fertilizer rate and irrigation was also demonstrated by McConnell et al. (1989), although this was also related to irrigation method. Applying N in irrigation water is often the most convenient and cheapest method of fertilization, and technology is rapidly improving for measuring crop-water use and in applying fertilizers through irrigation systems. Perhaps the time has come to more closely examine the concept of managing crop-N supply on the basis of crop-water supply.

142

THOMAS J. GERIK ETAL.

vn. SUMMARY Cotton growth is sensitive to N supply. Physiologically, N uptake and carbon assimilation are so interdependent that neither can operate without detriment to the other. This interdependence transcends the obvious impact on photosynthesis and alters other physiological and morphological processes, including water uptake, leaf expansion, assimilate partitioning, and the duration of morphological periods associated with fruit formation by changing the time to harvest. Optimizing N supply during the fruiting period is critical for promoting vegetative growth (e.g., leaf development), maintaining photosynthetic activity, and maximizing the plant’s boll carrying capacity and lint yield. The mobility and dynamic nature of N in the plant-soil continuum complicate its availability to the crop. Plant uptake must be balanced with soil N and water supply. Most analytical methods for measuring soil- or plant-N status provide antecedent estimates of N uptake or availability, and empirically derived fertilizer tests rely on previous experience to estimate the fertility needs of the crop. Basing N fertilization on crop-water use has potential in imgated production systems. Crop-simulation models hold considerable promise for estimating crop-N consumption and future needs. Several models have been developed to simulate N uptake of cotton and to predict future growth and final yield. The accuracy of these models relies, in part, on our knowledge of plant-N requirements. Most models have not been validated for N uptake or must be improved to accurately simulate N recovery and yield. Both analytical and empirical methods provide valuable information in determining the crop’s fertilizer needs, but knowledge of the plant-N requirement and future growth are needed to estimate the fertilizer required for the remainder of the growing season. Bondada et al. (1994) showed that boll load had a major influence on plant-N requirements and response to foliar N. Substantial discrepancies exist in published estimates of cotton’s N requirement for lint production. Are these discrepancies due to variation in water supply, or are they due to variation in soil N and the mineralization-immobilization turnover; or do they reflect differences in cultivar due to differences in source sink relations (i.e., boll load and leaf area), soil type, or climate? Research is needed to rectify these discrepancies-to accurately determine the plant-N requirement for cotton and to identify the factors responsible for this variation.

REFERENCES Albers, D. W. ( 1990). GOSYM-COMAX N recommendations: Strengths and weaknesses. In “Nitrogen Nutrition of Cotton: Practical Issues” (W. N. Miley and D. M. Oosterhuis, eds.), pp. 53-58. Proc. I st Ann. Workshop for Practicing Agronomists. ASA, Madison, WI.

MANAGING COTTON NITROGEN SUPPLY

143

Anderson, D. J., and Walmsley, M. R. (1984). Effects of eight different foliar treatments on yield and quality of an unfertilized short season cotton variety in the Texas Coastal Bend. In “Proc. Beltwide Cotton Production Research Conferences” (J. M. Brown, ed.), pp. 128-130. National Cotton Council, Memphis, TN. Baker, D. N., Lambert, J. R., and McKinion, J. M. (1983). “GOSSYM: A Simulation of Cotton Crop Growth and Yield.” South Carolina Agric. Exp. Sta., Tech. Bull. 1089. Clemson, SC. Baker, J. M., Reed, R. M.. and Tucker, B. B. (1972). The relationship between applied nitrogen and the concentration of nitrate nitrogen in cotton petioles. Comm. Soil Sci. and Plant Anal. 3,345. Baolong, Z . (1989). The absorption and translocation of foliar-applied nitrogen in cotton. M.S. thesis, Univ. of Arkansas, Fayetteville. Bassett. D. M., Anderson, W. D., and Werkohoven, C. H. E. (1970). Dry matter production and nutrient uptake in imgated cotton (Gossypium hir.suturn L.). Agron. J. 62,299-303. Beevers, L., and Hageman, R. H. (1969). Nitrate reduction in higher plants. An. Rev. Plant Physiol. 20, 495-522. Benedict, C. R.. McCree, K. J., and Kohel, R. J. (1972). High photosynthetic rate of a chlorophyll mutant cotton. Plant Physiol. 49,968-971. Bondada. B. R., Oosterhuis, D. M., and Baker, W. H. (1994). “Cotton Yield Response to Foliar N Fertilization in Relation to Boll Load. Petiole, and Fertilizer Nitrogen.” Ark. Agric. Exp. Sta. Research Series 436 (W. E. Sabbe, ed.), pp. 99-91. Fayetteville, AR. Bondada, B. R., Oosterhuis, D. M., Norman, R. J., and Baker, W. H. (1996). Canopy photosynthesis, growth, yield, and boll IsN accumulation under nitrogen stress in cotton. Crop Sci. 36, 127-133. Bondada, B. R., Oosterhuis, D. M., and Norman, R. J. (1997). Cotton leaf age, epicuticular wax, and nitrogen-15 absorption. Crop Sci. 37,807-81 1. Boone, K., Porter, D., and McKinion, J. (1995). Rhizos-1991: A simulator of row crop rhizosphere. USDA-ARS Crop Simulation Res. Unit, USDA ARS- I 1 3. Mississippi State Univ. Bourland, F. M., Oosterhuis, D. M., and Tugwell. N. P. (1992). Concept for monitoring the growth and development of cotton plants using main-stem node counts. J. Prod. Agric. 5, 532-538. Boynton, D. (1954). Nutrition by foliar application. Ann. Rev. Plant Physiol. 5,31-54. Boynton, D., Margolis, D., and Gross, C. R. (1953). Exploratory studies of nitrogen metabolism by McIntosh apple leaves sprayed with urea. Proc. Arne,: Sue. Horr. Sci. 62, 135-146. Cain, R. ( 1956).Absorption and metabolism of urea by leaves of coffee, cacao, and banana. Proc. Arne,: Soc. Hort. Sci. 67,279-286. Cappy, J. J. (1979). The rooting pattern of soybean and cotton throughout the growing season. Ph.D. diss. Univ. of Arkansas, Fayetteville. Elrifi, I. R.. and Turpin, D. H. (1986). Transient photosynthetic responses of nitrogen addition. Ma,: Ecol. frog. Ser: 20,253-258. Firestone, M. K. ( 1982). Biological denitrification. In “Nitrogen in Agricultural Soils” (F. J. Stevenson e t a / . . eds.), pp. 289-326. Agron. Monogr. 22. ASA and SSSA, Madison, WI. Foyer, C. H., Lescure, J. C., Lefebvre, C., Morot-Gaudry, F. F., Vincentz, M., and Vaucheret, H. ( 1994). Adaptations of photosynthetic electron transport, carbon assimilation, and carbon partitioning in transgenic Nicotiuna plumhagir$olia plants to changes in nitrate reductase activity. Plant Physi01. 104, 171-178. FranAuebbers, A. J., Hancy, R. L., Hons, E M., and Zuberer, D. A. (1996). Determination of microbial biomass and nitrogen mineralization following rewetting of dried soil. Soil Sci. Soc. Am. J. 60, 1133-1 139. Fraps. G. S. (1919). “The Chemical Composition of the Cotton Plant.’’ Texas Agric. Exp. Sta., Bull. 247. Temple, TX. Gardner, B. R., and Tucker, T. C. (1967). Nitrogen effects on cotton. 11. Soil and petiole analyses. Proc. Soil Sci. Soc. A m 31, 785-79 I . Gerik, T. J.. Jackson. B. S..Stockle, C. 0..and Rosenthal. W. D. (1994). Plant nitrogen status and boll load of cotton. Agron. J. 86,5 14-5 18.

144

THOMASJ. GERIK ETAL.

Grimes, D. W., and El-Zik, K. M. (1982). “Water Management for Cotton.” Univ. of California Coop. Exten. Bull. 1904. Grimes, D. W., Yamada, H., and Dickens, W. L. (1969). Functions for cotton (Gossypiurn hirmrum L.) production from irrigation andnitrogen fertilization variables. I. Yield and evapotranspiration. Agron. J. 61,769-733. Hake, K., and Kerby, T. (1988). Nitrogen fertilization. In “Cotton Fertilization Guide,” pp. 1-20. Univ. of California, Bakersfield, CA. Halevy, J. (1976). Growth rate and nutrient uptake of two cotton cultivars grown under irrigation. Agron. J. 68,701-705. Hauck, R. D. (1982). Nitrogen: Isotope-ratio analysis. In “Methods of Soil Analysis” (A. L. Pace, P.M. Miller, and D. R. Keeney, eds.), 2nd ed., Amer. SOC.Agron. Madison, WI. Hearn. A. B. (1986). Effect of preceding crop on the nitrogen requirements of irrigated cotton (Gossypium hirsuturn L.) on a vertisol. Field Crops Res. 13, 159-175. Hesketh, J. D., Baker, D. N.. and Duncan, W. G. (1972). Simulation of growth and yield in cotton: 11. Environmental control of morphogenisis. Crop Sci. 12,436-439. Huppe, H. C., and Turpin, D. H. (1994). Integration of carbon and nitrogen metabolism in plant and algal cells. Annu. Rev. Plant Physiol. Plant ,4401. Biol. 45,577-607. Ireland, R. (1990). Amino acid and ureide biosynthesis. In “Plant Biochemistry, Physiology, and Molecular Biology” (D. T. Dennis and D. H. Turpin. eds.), pp. 407-421. Longman Sci. Tech., Essex. Jackson, B. S.. and Gerik. T. J. (1990). Boll shedding and boll load in nitrogen-stressed cotton. Agron. J. 82,483-488. Jackson, B. S., Arkin, G. F., and Hearn, A. B. (1988). The cotton simulation model “COTTAM”: Fruiting model calibration and testing. Trans. ASAE. 31,846854. Jansson, S. L., and Persson, J. (1982). Mineralization and immobilization of soil nitrogen. In “Nitrogen in Agricultural Soils,” pp. 229-252. Agron. Monogr. 22. ASA and SSSA, Madison, WI. Jones, U. S. (1982). “Fertilizers and Soil Fertility.’’ Reston Publishing Co., Reston, VA. Kannan, S. (1986). Foliar absorption and translocation of inorganic nutrients. CRC Criricul Re~j.in Plan? Sci. 4(4), 341-375. Kerby, T. A., and Buxton, D. R. (1978). Effect of leaf shape and plant population on rate of fruiting position appearance in cotton. Agron. J. 70,535-538. Kiniry. J. R., Williams, J. R., Gassman, P. W., and Debaede. P. (1992). A general, process oriented model for two competing plant species. Trans. ASAE. 35, 801-8 10. Klein, I., and Weinbaum, S. A. (1984). Foliar application of urea to olive: Translocation of urea nitrogen as influenced by sink demand and nitrogen deficiency. J. Amer: Soc. Hort. Sci. 109(3). 358-360. Lane, H. C.. Thompson, A. C., Hesketh, J. D., and Sloane, C. (1975). Some observations on nitrate reduction in cotton. In “Proc. Beltwide Cotton Production Res. Conf.” (J. Brown, ed.), p. SO. National Cotton Council, Memphis, TN. Longenecker, D. E.. Thaxton, E. L., and Lyenly, P.(1964). “Nutrient Content and Nutrient Ratios as Affected by Irrigation Frequency, Water Quality, and Other Factors.” Texas Agric. Exp. Sta., MP728. p. I I . Temple, TX. MacKenzie, A. J., Spencer, W. F., Stockingen. K. R.. and Krantz, B. A. (1963). Seasonal nitrate-nitrogen content of cotlon petioles as affected by nitrogen application and its relationship to yield. Agron. J. 55,55-59. Maples, R., and Frizzell, M. (1985). “Effects of Varying Rates of N on Three Cotton Cultivars.” Arkansas Agric. Exp. Sta.. Bull. 882. Fayetteville, AR. Maples, R. L., Keogh, J. G.. and Sabbe. W. E. ( 1977). “Nitrate Monitoring for Cotton Production in Loring-Calloway Silt Loam.” Arkansas Agric. Exp. Sta., Bull. 825. Fayetteville, AR. Maples. R. L.. Miley, W. N., and Keisling, T. C. (1990). Nitrogen recommendations for cotton based Strengths and limitations. In “Nitrogen Nutrition of Cotton: Practical Issues” (W. N. Miley and D. M. Oosterhuis, eds.), pp. 59-64. American SOC.of Agron., Madison. WI.

MANAGING COTTON NITROGEN SUPPLY

145

Mauney, J. R. (1986). Vegetative growth and development of fruiting sites. In “Cotton Physiology” (J. R. Mauney and J. McD. Stewart, eds.), pp. 11-28. The Cotton Foundation, Memphis, TN. McConnell, J. S., Frizzell, 8 . S., Maples, R. L., Wilderson, M. H., and Mitchell, G. A. (1989). “Relationships of Irrigation Methods and Nitrogen Fertilization Rates in Cotton Production.” Arkansas Agric. Exp. Sta., Report Series 310. Fayetteville, AR. McConnell, S., Baker, W. H., Miller, D. M., Frizzell, B. S., and Varvil, J. J. (1993). Nitrogen fertilization of cotton cultivars of differing maturity. Agron. J . 85, 1151-1 156. McHargue, J. S. (1926). Mineral Constituents of the Cotton Plant. J. Am. SOC.Agron. 18, 1076-1083. McNamara, H. C., Hooten, D. R., and Porter, D. D. (1940). “Differential Growth Rates in Cotton Varieties and Their Response to Seasonal Condition at Greenville, TX.” USDA, Cir. 401. Miley, W. N. (1988). Foliar nitrogen can be good supplement. Delra Farm Press. 45(27), 7. Miley, W. N., and Maples, R. L. (1988). “Cotton Nitrate Monitoring in Arkansas.” Univ. of Arkansas Coop. Exten. Ser., Cotton Comments 2-88. Fayetteville, AR. Miller, R. H. (1982). Apple fruit cuticle and the Occurrence of pores and transcuticular channels. An. Boi. 50,355-360. Morrow, M. R., and Krieg, D. R. (1990). Cotton management strategies for a short growing season environment: Water-nitrogen considerations. Agron. J. 82,52-56. Mullins, G. I., and Burmester, C. H. (1990). Dry matter, nitrogen, phosphorous, and potassium accumulation by four cotton varieties. Agron. J. 82,729-736. Nadelhoffer, K. J. ( 1990). Microlysimeter for measuring nitrogen mineralization and microbial respiration in aerobic soil incubations. SoilSci. SOC.Am. J. 54,411-415. Natr, L. (1975). Influence of mineral nutrition on photosynthesis and the use of assimilates. In “Photosynthesis and Productivity in Different Environments” (J. P. Cooper, ed.), pp. 537-555. Cambridge University Press, Cambridge. Nommik, H., and Vahtras, K. (1982).Retention and fixation of ammonium and ammonia in soils. In “Nitrogen in Agricultural Soils” (F. J. Stevenson ef d,, eds.), pp. 123-171. Agron. Monogr. 22. ASA and SSSA, Madison, WI. Olson, L. C., and Bledsoe, R. P. (1942). The chemical composition of the cotton plant and the uptake of nutrients at different stages of growth. Georgia Agric. Exp. Sta., Bull. 222. Olson, R. A,, and Kurtz, L. T. (1982). Crop nitrogen requirements, utilization, and fertilization. In “Nitrogen in Agricultural Soils” (F. J. Stevenson, ed.), pp. 567-604. Agron. Monogr. 22. ASA, CSSA, and SSSA, Madison, WI. Oosterhuis, D. M. (1990).Growth and development of a cotton plant. In “Nitrogen Nutrition of Cotton: Practical Issues” (W. N. Miley and D. M. Oosterhuis, eds.), pp. 1-24. Am. Soc. of Agron., Madison, WI. Oosterhuis, D. M., and Bate. G. C. (1983).Nitrogen uptake of field grown cotton. 11. Nitrate reductase activity and petiole nitrate concentration as indicators of plant nitrogen status. Expl. Agric. 19, 103- 109. Oosterhuis. D. M., and Morris, W. J. (1979).Cotton petiole analysis as an indicator of plant nitrogen status in Rhodesia. Rhod. Agric. J. 76, 37-42. Oosterhuis, D. M.. Chipamaunga, J., and Bate, G. C. (1983). Nitrogen uptake of field grown cotton. 1. Distribution in plant components in relation lo fertilization and yield. Expl. Agric. 19,91-102. Oosterhuis, D.M., Zhu. B.. and Wullschleger, S. D. (1989).The uptake of foliar-applied nitrogen in cotton. In “Proc. Arkansas Cotton Res. Meeting and Summaries of Cotton Res. in Progress” (0.M. Oosterhuis, ed.), pp. 23-25. Arkansas Agric. Exp. Sta., Special report 138. Fayetteville, AR. Oosterhuis, D. M., Hampton, R. E., and Wullschleger, S. D. (I991 ). Water deficit effects on cotton leaf cuticle and the efficiency of defoliants. J. Prod. Agric 4,260-265. Orgaz, F.. Mateow, L.. and Fereres, E. ( 1992). Season length and cultivar determine optimum evapotranspiration deticit in cotton. Agron. J. 84,70&706. Radin, J. W., and Ackerson. R. C. (1981).Water relations of cotton plants under nitrogen deficiency.

146

THOMAS J. GERIK ETAL.

111. Stomata1 conductance, photosynthesis, and abscisic acid accumulation during drought. Plant Phv.TiOl. 67, 115-1 19. Radin. J. W., and Boyer, J. S . (1982). Control of leaf expansion by nitrogen nutrition in sunflower

plants: Role of hyduaulic conductivity and tugor. Plant Physiol. 69,771-775. Radin, J. W., and Matthews, M. A. (1989). Water transport properties of cortical cells in roots of nitrogen- and phosphorous-deficient cotton seedlings. Planr Physiol. 89,264-268. Radin, J. W., and Mauney, J. R. (1986). The nitrogen stress syndrome. In “Cotton Physiology” (J. R. Mauney and J. McD. Stewart, eds.), pp. 91-105. The Cotton Foundation, Memphis, TN. Radin, J. W., and Parker, L. L. (1979). Water relations of cotton plants under nitrogen deficiency. I. Dependence upon leaf structure. Plant Physiol. 64,495498. Radin, J. W., and Sell, C. R. (1975). Some factors limiting nitrate reduction in developing ovules of cotton. CropSci. 15,713-715. Radin, J. W., Parker, L. L., and Sell, C. R. (1978). Partitioning of sugar between growth and nitrate reduction in cotton roots. P h i Physiol. 62,550-553. Reddy, V. R., Trent, A,, and Acock, B. (1992). Mepiquat chloride and irrigation versus cotton growth and development. Agron J. 84,930-933. Reeves, D. W. (1994). Cover crops and rotations. In “Advances in Soil Science: Crops Residue Management” (J. L. Hatfield and B. A. Stewart, eds.), pp. 125-172. Lewis Publishers, CRC Press, Boca Raton, FL. Rufty, T. W., Huber, S. C., and Volk, R. J. (1988). Alterations in leaf carbohydrate metabolism in response to nitrogen stress. Plant Physiol. 89,457463. Ryden, J. C., and Lund, L. J. (1980). Nature and extent of directly measured denitrification losses from some irrigated vegetable crop production units. Soil Sci. Soc. Am. J . 44,505-5 1 1. Sabbe, W. E., and MacKenzie. A. J. (1973). Plant analysis as an aid tocotton fertilization. In “Soil Testing and Plant Analysis” (L. M. Walsh and J. D. Beaton, eds.), rev. ed., pp. 299-3 13. Soil Sci. SOC. Am., Madison, WI. Sabbe, W. E., and Zelinski, L. J. (1990). Plant analysis as an aid in fertilizing cotton. In “Soil Testing and Plant Analysis” (R. L. Westerman, ed.), 3rd ed., pp. 469490. Soil Sci. Soc. Am. Madison, WI. Sabbe, W. E.,Keogh, J. L., Maples, R., and Hileman, L. H. (1972). Nutrient analysis of Arkansas cotton and soybean leaf tissue. Arkansas Farm Res. 21( l), 2. Schmidt, E. L. (1982). Nitrification in soil. In “Nitrogen in Agricultural Soils’’ (F. J. Stevenson ei al., eds.), pp. 258-288. Agron. Monogr. 22. ASA and SSSA, Madison, WI. Smirnof, N.. and Stewart, G.R. (1985). Nitrate assimilation and translocation by higher plants: Comparative physiology and ecological consequences. Physiol. Plani. 64, 133-140. Smith, F., Malm, N., and Roberts, C. (1987). Timing and rates for foliar nitrogen application of cotton. In “Proc. Beltwide Cotton Production Res. Conf.” (J. M. Brown, ed.), pp. 61-64. The National Cotton Council, Memphis, TN. Stanford, 0.(1982). Assessment of soil nitrogen availability. In “Nitrogen in Agricultural Soils” (F. J. Stevenson et al., eds.), pp. 651-688. Agron. Monogr. 22. ASA and SSSA, Madison, WI. Stevens, W. E., Varco, J. J., and Johnson, J. R. (1996). Evaluating cotton nitrogen dynamics in the GOSSYM simulation model. Agron. J. 88, 127-132. Stevenson, F. J. (1982a). Origin and distribution of nitrogen in soil. In “Nitrogen in Agricultural Soils” (E J. Stevenson et al., eds.), pp. 1-42, Agron. Monogr. 22. ASA and SSSA, Madison, WI. Stevenson, F. J. (1982b). Organic forms of soil nitrogen. In “Nitrogen in Agricultural Soils” (F. J. Stevenson et a/.. eds.), pp. 67-122. Agron. Monogr. 22. ASA and SSSA. Madison, WI. Sunderman, H. K.. Onken, A. B.. and Hossner, L. R. (1979). Nitrate concentration of cotton petioles as influenced by cultivar, row spacing, and N application rate. Agron. J . 71,73 1-737. Syrett, P. J. (1981). Nitrogen metabolism of microalgae. Can. Bull. Fish. Aquat. Sci. 210, 182-210. Taylor, H. ( 1995). “1994 Nutrient Use and Practices on Major Field Crops.” ARE1 Updates: Nutrient

MANAGING COTTON NITROGEN SUPPLY

147

Use and Management, 2. Natural Resources and Environmental Division, Econ. Res. Ser., USDA, Washington, D.C. Torbert, H. A., and Reeves. D. W. (1994). Fertilizer nitrogen requirements for cotton production as affected by tillage and traffic. Soil Sci. Soc. Am. J. 58, 14161423. Torbert. H. A., and Wood, C. W. (1992). Effects of soil compaction and water tilled pore space on soil microbial activity and N losses. Comm. Soil Sci. Plant Anal. 23, 1321-1331. Tracy, P. W.. Hefner, S. G., Wood, C. W., and Edmisten, K. L. (1992). Theory behind the use of instantaneous leaf chlorophyll measurement for determining mid-season cotton nitrogen recommendation. In “Proc. Beltwide Cotton Conf.” (D. J. Herber and D. A. Richter, eds.), pp. 1099-1 100. The National Cotton Council, Memphis, TN. Tucker, T. C. (1965). “The Cotton Petiole, Guide to Better Fertilization.” Plant Food Rev., pp. 9-1 1. Univ. of Arizona, Dept. Agric. Chem. Tucson, AZ. Tucker, T. C.. and Tucker, B. B. (1968). Nitrogen nutrition. In “Advances in Production and Utilization of Quality Cotton: Principles and Practices” (F. C. Elliot, M. Hoover, and W. K. Porter. Jr., eds.), pp. 185-208. Iowa State University Press, Ames, IA. Vasilas, B. L., Legg, J. 0..and Wolf, D. C. (1980). Foliar fertilization of soybeans: Absorption and translocation of iSN-labeledurea. Agron. J. 72,27 1-275. Volk, R. J., and McAulliffe, C. (1954). Factors affecting the foliar absorption of ”N-labeled urea by tobacco. Soil Sci. Soc. Arne,: Proc. 18,308-3 12. Wadleigh, C. W. (1944). “Growth Status of the Cotton Plant as Influenced by the Supply of Nitrogen.” Arkansas Agric. Exp. Sta., Bull. 446. Fayetteville, AR. Walter, H., Gausman, H. W., Rittig, F, R., Namkin, L. M., Escobar, D. E., and Rodriguez, R. R. (1980). Effects of mepiquat chloride on cotton plant leaf and canopy structure and dry weights of its components. In “Proc. Beltwide Cotton Prod. Res. Conf.” (J. M. Brown, ed.), pp. 32-35. The National Cotton Council. Memphis, TN. Williams, J. R., Jones, C. A,, Kiniry, J. R., and Spanel, D. A. (1989). The EPIC crop growth model. Trans. ASAE. 27(1). 129-144. Wittwer, S. H., Bukovac, M. J., andTukey, H. B. (1963). Advances in foliar feeding of plant nutrients. In “Fertilizer Technology and Usage” (M. H. McVickar, G. L. Bridger, and L. B. Nelson, eds.), pp. 429-455. Amer. SOC.Agron., Madison, WI. Wood, C. W., Tracy, P. W., Reeves, D. W.,and Edmisten, K. L. (1992). Determination of cotton nitrogen status with a hand-held chlorophyll meter. J. Plant Nut,: 15, 143-1448, Wullschleger, S. D., and Oosterhuis, D. M. (1990). Canopy development and photosynthesis of cotton as influenced by nitrogen fertilization. J . PIunr Nuts 13, 1141-1 154. Wullschleger, S. D., and Oosterhuis, D. M. (1992). Canopy leaf area development and age-class dynamics in cotton. Crup Sci. 32,45 1 4 5 6 . Yamada, Y.(1962). Studies on foliar absorption of nutrients using radioisotopes. Ph.D. diss. Kyoto Univ.. Kyoto, Japan. Zhao, D. (1997). Floral bud development of cotton (Gossjpium hirsutum L.) and responses to shade and PGR-IV application. Ph.D. diss., Univ. of Arkansas, Fayetteville. Zhu, B., and Oosterhuis. D. M. (1992). Nitrogen distribution within a sympodial branch of cotton. J. PImt Nut,: 15. 1-14.

This Page Intentionally Left Blank

ARSENICINTHE SOILENVIRONMENT: A REVIEW E. Smith,'?' R. Naidu,'y3* and A. M. Alston2 'CRC for Soil and Land Management Glen Osmond, South Australia 5064 Australia *Department of Soil Science University of Adelaide Glen Osmond, South Australia 5064 Australia 'CSIRO Division of Soils Glen Osmond, South Australia 5064 Australia

I. Introduction 11. Position in the Periodic Table 111. Background Sources A. Background Concentrations of As in Soils n! Anthropogenic Sources A. Industry B. Mining C. Other Sources D. Agriculture V. AsToxicity A. Accumulation in Biota B. Human Exposure to As VI. Physiochemical Behavior of As in Soil A. Inorganic As Compounds B. Organic As Compounds C. The Soil Solution D. Adsorption-Desorption Processes E. Kinetics of As Adsorption-Desorption VII. Soil As and Vegetation A. Soil As and Plant Uptake VIII. Soil As and Microorganisms A. Biotransforination of As in Soils IX.Conclusions References *Corresponding author. 149 .4hmrc.r in /lgronotnq, Vaiumr 64 Copyright 0 1098 by Academic Press. All rights of rrproducnon in m y form reserved.

1 1 0 6 s -ziim

moo

150

E. SMITH ET AL.

I. INTRODUCTION Indiscriminate use of arsenical pesticides during the early to mid- 1900s has led to extensive contamination of soils worldwide. Contamination in excess of 1000 mg As kg-' has been recorded at many sites throughout Australia. Similar contaminated sites also exist in the United States, Africa, and other parts of the world. The persistence of As residues in soil and toxicity to both plants and animals is of concern. Consequently, many investigators have studied the chemistry of As at contaminated sites. Arsenic shows no indication of being an essential element in biological processes, although some organic As compounds have been used in low concentrations as food additives to poultry and swine supplements (Leonard, 1991). The toxicity of As compounds depends on a number of factors, including the chemical form present: inorganic As forms are more toxic than organic, and arsenite (As"') is more toxic than arsenate (As") (World Health Organization, 1981). In humans, the LD,, has been estimated from the incidents of poisoning to range from 1 to 5 mg As kg-' (Fowle 111 et al., 1991). Evidence of As poisoning of humans has been reported in India (Das et al., 1996). Thousands of humans in Bengal have developed symptoms of As toxicity either through consumption of Ascontaminated groundwater or through ingestion of As-containing food crops. The latter pathway seems unlikely given the low soil-plant transfer of As. Much research needs to be done on the dynamics of As in the soil-plant system under field conditions that assess plant uptake of As. The presence of As in the environment may be due both to background and to anthropogenic sources. The soil environment is an important sink for As compounds. Arsenic deposited in the soil may accumulate rapidly since it is only slowly depleted through plant uptake, leaching, methylation, or erosion. Because of the known toxicity of As to human and animal systems and the presence of contaminated sites throughout the world, there has been renewed interest in studying the dynamics of As with a view to developing management strategies. For this reason, numerous reviews have been published in recent years describing the behavior of As in the soil environment (Nriagu, 1994). Although these reviews consider As in the soil system, they lack details on the major sources of As, the rates of inputs, and the chemistry of As in the soil environment. This review addresses some of these issues.

11. POSITION IN THE PERIODIC TABLE Arsenic (atomic number 33; atomic mass 74.9216) has an outer electronic configuration of 4s2 4p3 and belongs to subgroup V of the periodic table. The decrease

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

1.51

in electronegativity that is found on descending this group is not sufficient to give As a metallic character, and it is often described as a metalloid. In soils, the chemical behavior of As is in many ways similar to that of phosphorus (P), especially in aerated systems, where the As" ion generally resembles orthophosphate ion closely (Walsh et al., 1977). However, under conditions normally encountered in soils, As is more mobile than P and unlike P can undergo changes in its valence state.

III. BACKGROUND SOURCES The main source of As in soils is the parent materials from which the soil is derived (Yan-Chu, 1994). The native As content may vary considerably within an area and is often determined by the geological history of the region (Wild, 1993). Arsenate and As"' are the dominant As species in soils (Deuel and Swoboda, 1972; Walsh and Keeney, 1975), and anthropogenic sources of As pollution have enhanced the background concentrations of these species. The As content of rocks depends on the rock type, with the sedimentary rocks containing much higher concentrations than the igneous rocks (Bhumbla and Keefer, 1994). Although discernible differences exist between rock groups, the range of As concentrations within a rock type may vary considerably. Generally, the mean As concentrations in igneous rocks range from 1.5 to 3.0 mg As kg-I, whereas the mean As concentrations in sedimentary rocks range from 1.7 to 400 mg As kg-I. Atmospheric deposition contributes significantly to the geochemical cycle of As (O'Neill, 1990). Chilvers and Peterson (1987) estimated a global atmospheric flux value of 73,540 t year-', with a 6 M O split between natural and anthropogenic sources. This compares with Nriagu and Pacyna (1988), who estimated a ratio of 70:30, with an anthropogenic As input of 18,800 t year-'. About 60% of the atmospheric As flux has been estimated to be due to low-temperature volatilization, with volcanic activity the next most important natural source (Chilvers and Peterson, 1987). However, on a localized scale, volcanic activity may be the dominant source of atmospheric deposition (O'Neill, 1990).

A. BACKGROUND CONCENTRATIONS OF As IN SOILS The distribution of As in soils may vary with soil type, depending on the nature of the parent material. Background concentrations do not generally exceed 15 mg As kg-l (National Research Council of Canada [NRCC], 1978), although concentrations ranging from 0.2 to 40 mg As kg-l soil have been reported (Walsh et

152

E. SMITH ETAL. Table I

Arsenic Concentrationsfrom Noncontaminatedand ContaminatedSoils in North Americaa ~~~~

~~

Total As content

Sampling site Colorado Florida Idaho Indiana Maine Maryland New Jersey New York North Carolina Nova Scotia Ontario Oregon Washington

Wisconsin

Noncontaminated soil (mg kg-')

Contaminated soil (mg kg-')

Crop

1.3-2.3 8

13-69 18-28 138-204 56-250 10-40 21-238 92-270 9M25 1-5 10-124 10-121 17-439 4-103 106-830 106-2553 48 6-26

orchard potato orchard orchard bluebeny orchard orchard orchard tobacco orchard orchard orchard orchard orchard orchard orchard potato

0-10

2-4 9 19-41 10 3-12 4 0-7.9 1.1-8.6 2.5-14 3-32 613 8-80 4-13 2.2

"Reprinted with permission from Walsh and Keeney, 1975, 0 1975 American Chemical Society.

al., 1977). Dudas and Pawluk (1980) reported background As concentrations that averaged 5 mg As kg-' in 78 chernonzemic and luvisolic soil samples in Alberta. Much higher As concentrations have been reported in acid sulphate soils developed on pyritic parent material. For instance, Dudas (1987) attributed elevated As concentrations that ranged from 8 to 40 mg As kg-I in Canadian acid sulphate soils to the weathering of pyrites in the parent material. Other studies have reported a similar variability among soils from various regions. Reviewing literature on As concentrations in nonpolluted and polluted soils, Walsh and Keeney (1975) concluded that nonpolluted soils in North America (Table I) rarely contain more than 10 mg As kg-' soil. Similarly, the NRCC (1978) report on the effects of As in the Canadian environment concluded that background As concentrations in soils rarely exceed 15 mg As kg-' soil. A limited number of similar studies have been reported in Australia. Merry et al. (1983) studied 15 surface (0-150 mrn) soils from South Australia and 6 from Tasmania that were considered unlikely to have received anthropogenic sources of As. The median As concentrations in South Australian and Tasmanian soils were 3.9 mg As kg-' (+2.0) and 0.6 mg As kg-' (+0.55), respectively. Tiller (1992)

ARSENIC IN THE SOIL ENVIRONMENT:A REVIEW

153

~~

Total As (rng kg-I) Melbourne Hobart Sydney Adelaide

< 0.2-8.1 2-45 0.6-1 1

0.2-16

“‘Iiller, 1m;reprintedby permission of CSIRO Aushlia

has also compared background As concentrations from several studies of urban soils (Table 11) demonstrating a wide range in soil-As concentrations. In contrast to studies by Merry e?al. (1983) and Tiller ( 1992), Fergus ( I 955) reported elevated As concentrations in soils derived from weathered quartzite (70-100 mg As kg-’ at 0-75 mm) that resulted in restricted growth and toxicity symptoms on the leaves of banana palms in Queensland. These regional variations in As concentrations in soils highlight the wide variability in soil As.

IV. ANTHROPOGEMC SOURCES Anthropogenic activities that contribute As to the soil environment originate from primary and secondary industries. These varying sources add As that differs widely in nature and composition. Such variations in the composition and nature of As have implications for biological availability as well as the mobility of As in soils. The following sections briefly consider the sources and forms of As entering the soil environment through anthropogenic activities.

A. INDUSTRY Arsenic trioxide (As,O,) is the major form of As that is produced for industry. Industrial uses include the manufacture of ceramics and glass, electronics, pigments and antifouling agents, cosmetics, and fireworks (Leonard, 1991). Arsenic is also added as a minor constituent to Cu and Cu-based alloys to raise the corrosion resistance of the metal(s) (Nriagu, 1994). Arsenic trioxide is recovered from the smelting or roasting of nonferrous metal ores or concentrates (Loebenstein, 1993). From the limited data available (Fig. 1 ), the world production of As,O, appears to have remained relatively constant

154

E. SMITH E T A . n v)

n v)

g 60000 -

L: -

-4 CI

c.

mg + 50000

:\

l! 40000 -s30000 3 20000 -

v

C

1800 g 1600 L 1400 5 E 1200 1000 - 800 - 600 - 400 = a 2 - 200 'j

7

0

3

Q

8

+World

10000 -

s o

'c

production

4-Australian imports

I

I

I

I

I

I

1

I

0

2

Year Figure 1 World production of As,O, and Australian imports per year (after Loebenstein, 1993, and Australian Bureau of Statistics, 1995).

from 1985 to 1990 at approximately 50,000 t year-' (Loebenstein, 1993). This contrasts with the declining use of As compounds in agriculture, which in the late 1970s to the early 1980s made up approximately 70% of the world As consumption (Hillier, 1980). This decline in As use in the agricultural sector has probably been offset by the increasing use of As in the timber treatment industry. Arsenic has excellent wood-preserving properties and is used in the timber industry in conjunction with Cu and Cr. World usage of As as a wood preservative is increasing at approximately I-2% a year (Loebenstein, 1993). Tentative estimates of the anthropogenic As fluxes between land, oceans, sediments, and the atmosphere have been calculated (Nriagu and Pacyna, 1988; Chilvers and Peterson, 1987). Nriagu and Pacyna (1988) estimate that the total worldwide anthropogenic As discharge onto land was 64,000-132,000t year-' (Fig. 2). They estimated the major sources of As discharged onto land originated from commercial wastes (about 40%), coal ash (about 22%), and atmospheric fallout from the production of steel (about 13%). Other anthropogenic sources of pollution associated with the mining industry (about 16%)also greatly contribute to As emissions onto land.

B. MINING Arsenic is a natural component of Pb, Zn, Cu, and Au ores. Consequently, contamination of the atmosphere, soils, sediments, streams, and groundwaters is possible during mining and/or smelting processes. Although As-contaminated agri-

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

15s

Figure 2 Worldwide As discharges onto soils (after Nriagu and Pacyna, 1988).

cultural soils have been the subject of numerous detailed investigations, limited studies have been carried out on the detailed nature and dynamics of As in mine spoils. Peterson et al. (1979) investigated the total As concentrations and nature of As in grossly contaminated mine spoils of southwest England. Total As concentrations in the spoils were greater than 20,000 mg As kg- with the maximum concentration at a depth of 20-40 mm, and very high As concentrations being detected to the lowest depth sampled (33,750 mg As kg-' at 250-300 mm). These investigators also assessed bioavailability of As with deionized water. They found that water-soluble As concentrations (0.3 1 mg As kg- at 0-20 mm) were generally less than 1% of total As. Arsenate and As"' were the major As forms present in the water-soluble soil extracts, although dimethylarsine was also detected in surface samples. Although As" was the predominant form throughout the soil profile, similar concentrations of As"' were reported in the surface samples. Even though these soils had been relatively undisturbed for 70 years, the contaminated sites were largely barren and supported only a limited number of plant species that generally covered less than 1% of the contaminated area. Grasses were the predominant species present (Agrostis stolonifera and A. tenuis),with As concentrations in leaves greater than I000 mg As kg- (dry weight). Wild ( 1 973-74) reported that 72 plant species were found on 15 Rhodesian arsenical mine dumps with total As concentrations ranging from 200 to 30,000 mg As kg-'. As may be expected, the number of plant species and their density was found to increase as As content decreased. Of the mine dumps surveyed, the Banshee mine dump (30,000 mg As

',

'

'

156

E. SMITH ETAL.

kg-l) was affected the worst and was incapable of supporting vegetation. Generally, weed species were found to be the most important species in terms of plant numbers on the mine dumps. Ffaveriatrinervia (Gaika Weed) was often the dominant or most important weed species present on the mine dumps. Of the grass species, Cynodon dactylon was the most important species, being present in soils with As concentrations ranging from 200 to 30,000 mg As kg- *.While As and other heavy metal concentrations at some mine dumps may inhibit stabilizing soil vegetation from establishing, some plants, such as Cydon dactylon, can tolerate high As concentrations and may be useful for stabilizing mine dump soils. They may also offer a long-term, low-cost solution for remediating mine dumps when other remediation techniques are impractical. There have been a number of reported incidences of atmospheric As release during the smelting of Pb, Zn, Au, and Cu ores (Crecelius et al., 1974; Ragaini et al., 1977; Li andThornton, 1993). Crecelius et aE. (1974) reported that a large Cu mine near Tacoma, Washington, released approximately 300 t of particulate matter into the atmosphere per year. Dust containing approximately 20-30% As contaminated the soil (0-30 mm) within a 5-km radius of the smelter, with up to 380 mg As kg-' occurring at some of the sites sampled. Li and Thornton (1 993) studied the As contamination of soil from three ore smelting areas in England-Derbyshire, Cornwall, and Somerset. They reported that As concentrations in the topsoil (0-1 50 mm) were elevated above background measurements (7.69-8.97 mg As kg-l) and ranged between 16 and 925 mg As kg-', depending on the sampling area. Although most mining and smelting in these regions ceased at the end of the 19th century, As contamination in some areas is still particularly high. This emphasizes the general long-term problem posed by the recalcitrant nature of compounds associated with soil contamination from industrial sources. Unlike many heavy metals such as Cr, Cd, and Hg, As has been detected in groundwaters especially at sites contaminated by mill tailings. Bernard (1 983) investigated the contamination of groundwater and the subsequent contamination of Lake Moira, Canada, and found that haphazard disposal of mill tailings and other slag wastes resulted in considerable leaching of As from these sites. Water samples collected from around the tailings and As storage areas in a hydrological investigation of groundwater had As concentrations ranging from 600 to 2200 mg As liter-'. Extensive mitigation methods have been required to alleviate the high As concentrations in the lake. Similarly, Leblanc et a f .(1996) reported that the dissolved As content of an acidic stream (pH 2.2-4) originating from a waste mine dump of the Camoulbs Pb-(Zn) mine in Gard, France, was extremely high (average 250 mg As liter-'). Leblanc et al. (1996) also observed that As was precipitated and concentrated in Fe-As bacterial stromatolites. It was proposed that the accumulation of As was through direct or induced microbial action. Rittle el al. (1995) also reported that the immobilization of As"' into a Fe-As-S solid phase was also linked to microbial activity.

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

157

The mitigation of polluted As sites is an important aspect of the remediation process and is a research area that is receiving considerable attention. However, discussion of these remediation processes is beyond the scope of this review.

c. OTHERSOURCES 1. coal Fly ash presents an increasing waste problem worldwide because of the continuing demand for coal-fired power stations. In addition to directly releasing As into the atmosphere, coal combustion produces fly and bottom ash containing As. The physical and chemical properties of fly ash restrict the general utilization of fly ash, resulting in a large proportion of the ash being used as land fill (Beretka and Nelson, 1994). Generally, As concentrations in coal vary from 2 to 82 mg As kg-I, depending on geological origin (Adriano et al., 1980). However, very high concentrations of As (1500 mg As kg- I ) have been recorded in brown coal from the former Czechoslovakia (Bencko and Symon, 1977). An important feature of fly ash is the variation of elemental concentration with particle size. Natush er al. (1975) have observed that the concentrations of As and other metals in fly ash tend to increase as particle size decreases. Smaller particles of fly ash may escape emission-control devices and therefore may have greater impacts on biological systems in the vicinity of the emission source. The oxidation state of As in coal ash and its leachate are of concern, since As"' is considerably more toxic than As". However, there is a lack of information on the dominant redox state of As present in fly ash. Turner (1981) reported that pond effluent from 12 ash disposal systems contained quantities of total dissolved As of less than 0.5-1 50 pg As literp1. Arsenate was the dominant As species present, but the more mobile and toxic As"' accounted for between 3 and 40% of the dissolved As. This highlights the fact that, although As" may be the dominant As species present, conditions may be favorable for the presence of As"' as well. In contrast to the relatively low As"' concentrations in the ash porewater, interstitial As"' concentrations collected from two wells ranged from 1.2 to 550 pg As liter-' at4-6 m and43-1480 pg As liter-' at 10-12 mdepth, respectively (Turner, 1981). These relatively high concentrations of As"' may be of environmental and human and animal health concerns if contamination of surface andor groundwater were to occur. The disposal of fly ash in Australia is typical of that found in many Western countries. Australia currently produces approximately 7.7 X lo6t (1992) of fly ash annually from major power stations (Beretka and Nelson, 1994). Of this, approximately 7.7 X lo5 t were used in the production of cement, sand replacement in

158

E. SMITH ETAL.

concrete, and the manufacture of blended Portland fly-ash cement. Small quantities are also used as a mineral filler in asphalt, and the remaining fly-ash (about 90%) is disposed of either as mine fill or as a codisposal material in waste dumps (Beretka and Nelson, 1994). This cheap and therefore very attractive form of disposal may be of environmental concern given the relatively high concentrations of potentially toxic and mobile As"' in interstitial porewaters that have been reported.

2. Tannery Wastes Arsenic was historically used as a pesticide in the treatment of animal hides (Sadler et al., 1994). Sadler et al. (1994) investigated the As status of surface and subsurface soil contamination by long-term disposal of tannery wastes in Queensland, Australia. For over 81 years, liquid wastes from a tannery were pumped or transported by tanker to a site in Brisbane. The liquid waste was disposed of either by burial or spray irrigation methods. Surface-soil contamination (0-125 mm) displayed considerable variation, ranging from less than 1 to 435 mg As kg- soil across the site. Similarly, subsurface soil (250-375 mm) contamination also varied considerably, ranging from less than 1 to 1010 mg As kg-' soil. Although the amount of As contamination of the soil is not as high compared with mining and other reported sites, contamination at this site extended to considerable depth (600-725 mm) in the soil. This may be due both to the nature of As in the wastes and to the soil characteristics. Sodium arsenite was the active ingredient in the pesticide formulation used extensively to treat animal hides (Sadler ef al., 1994). Presence of As at considerable depth in sandy soils may be attributed to the high mobility of As"' in such soils ( T a m e s and de Lint, 1969).

3. Forestry The wood preservative industry is the major market for As in the United States (Loebenstein, 1993), and in 1990 this industry accounted for approximately 70% of the domestic As demand in the United States (Loebenstein, 1993).Although the wood preservative industry is a major end user of imported As,O,, there are few reported incidences of contamination. Nevertheless, Lund and Fobian ( 1991 ) reported elevated As concentrations in two soil types (typic haplorthod and typic hapludalf) due to spillage of chemicals used in impregnating wood. Arsenic concentrations in the haplorthod were highest in the surface soil (3290 mg As kg-' soil) and showed a general decline with increasing soil depth. Similar trends were evident for the hapludalf (surface sample approximately 380 mg As kg- soil), but there were large variations in the profile that could not be explained by the composition of the soil horizons (Lund and Fobian, 1991). Generally, As was retained in the A and B horizons of both profiles. In the A horizon, the retention of As was

'

ARSEMC IN THE SOIL ENVIRONMENT: A REVIEW

159

associated with high organic-matter content, whereas retention in the B horizon may be associated with adsorption by Mn, Fe, and A1 oxides (Lund and Fobian, 1991). The mechanism of As adsorption and the role of oxidic materials is further discussed in the section on adsorption mechanisms. McLaren et al. ( 1 994) investigated the leaching of Cu, Cr, and As (CCA) solution through free-draining, coarse-textured surface and subsurface soils (typic ustipsamment and udic ustochrept) using undisturbed soil lysimeters. Cumulative amounts of As leached through the lysimeters ranged from 4 to 30% of the total CCA solution applied (90 mg Cu, 157 mg Cr, and 130 mg As). The large amount of As leached is probably due to As being present as a simple salt (H,AsO,) in the CCA solution and therefore represents an increased leaching potential in comparison with metals in sewage sludge, which are in relatively immobile forms (McLaren et af.,1994).

D. AGRICULTURE Agricultural inputs such as pesticides, desiccants, and fertilizers are the major sources of As in soils (Jiang and Singh, 1994). Numerous cases of As contamination of agricultural soils have been recorded (Bishop and Chisholm, 1961; Woolson et al., 1971a; Hess and Blanchar, 1976; Merry et al., 1983).

I. Pesticides From the late 1800s and until the introduction of dichlorodiphenyltrichloroethane (DDT), Pb arsenate (PbAsO,), calcium arsenate (CaAsO,), magnesium arsenate (MgAsO,), zinc arsenate (ZnAsO,), zinc arsenite (Zn(AsO,),), and Paris green [C~(CH,COO);~CU(A~O~)~] were used extensively as pesticides in orchards (Anastasia and Kender, 1973; Merry et al., 1983). The resultant pollution of orchard soils by inorganic Pb and As pesticides has been extensively reported in the literature(BishopandChisholm,1961;Franketal., 1976;Merryetal., 1983; Peryea and Creger, 1994). Bishop and Chisholm (1 96 1) investigated As soil pollution on 25 Annapolis Valley orchards (mostly sandy loams; pH 6.2-6.7). It was found that the use of of arsenical pesticides had resulted in the accumulation of 9.8-124 mg As kg-' in the topsoil (0-1 SO mm) (Table 111). The considerable variations in As concentrations were attributed to different spraying practices at each orchard. Frank et al. ( 1976) reported similar findings in apple orchards to which PbAsO, sprays were applied for periods ranging from S to 70 years. The mean As concentrations were 54.2 2 25.8 mg As kg-' and 20.9 ? 13.6 mg As kg-*, respectively, in the 0-150-mm and 150-300-mm layers of soil. A comparison of the age of orchards versus As concentration in the surface soil showed an increase of 7-121 mg As kg-' after 70

E. SMITH ETAL.

160

Table 111 Arsenic Contaminationof Orchard Sites Total soil As (&I50 mm)

Apple orchards Torbrook Kentville Woodville Morristown Cornwallis Mean of 3 1 orchards Mean of 98 orchardsb ~

Background concentration (mg kg-')

Contaminated concentration (mg kg-')

4.2

124.4 15.0

4.5 -0

1.0 trace 6.21 3.9'; 0.6d

53.8 30.4

9.8 54.2 29'

Source Bishop and Chisholm, 1961 Ibid. bid. Ibid. bid. Frank et al., 1976 Merry et al., 1983

~ ~ _ _ _ _ _ _ _~ _ ~ ~ _ _ _ _ _ _ _ _ ~

"No data available. bTen apple and pear orchards from South Australia and 60 from Tasmania. CMeanof 15 soils from South Australia. %lean of 6 soils from Tasmania. 'Samples from depth of 0-100 mm.

years of pesticide applications. Increases in the As concentration were also evident in the 150-300-mm layer, although the concentrations were much lower. However, comparison of the Pb-As ratio between the untreated and treated topsoils indicates that there may have been considerable loss of As from the surface soil (Frank et af.,1976). This loss is reflected to some extent in the accumulation of As in the 150-300-mm horizon. Merry et af. (1983) reported that although there was considerable accumulation of As in 98 surface soils of apple and pear orchards in South Australia and Tasmania (Table III), there was evidence of loss of As from the surface soil at some sites. Translocation by leaching in soil solution or colloids in suspension were suggested as possible mechanisms for losses (Merry et al., 1983), although losses ofAs compounds through volatilization may also be an important but difficult pathway to quantify. Barrow (1974) and Davenport and Peryea (1 99 1) have reported that P amendments to soil may contribute to the displacement of As in soils, and leaching may be accentuated in sandy soils (Tammes and de Lint, 1969). Peryea and Creger (1994) found that the movement of Pb and As was greater in soils with low clay and organic contents, high irrigation rates, and high application rates of Pb arsenate pesticide. This study highlights that As mobility is a result of complex interactions between soil and solution factors that influence the leaching of As from the surface soil. Arsenical pesticides were also widely used in livestock dips to control ticks, fleas, and lice (Vaughan, 1993). In Australia, As-based pesticide solutions were widely used in Queensland and northern New South Wales from the early 1900s

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

161

Table N Average As Residues in Aggregated Soil in and around Cattle-Dip Sites" As (mg kg-' soil) Location within dip site Adjacent to dip bath Draining pen Scooping mound Disposal pit

Mean

Range

290 436 720 467

0-1636 2-870 15-3000 (L2600

"After DIPMAC Report, 1992.

to 1955 (DIPMAC Report, 1992) to control ticks in cattle. Investigations of possible polluted sites have identified 1607 known cattle dip sites, of which 1041 sites were still in operation in 1990 (DIPMAC Report, 1992). High concentrations of As residues have been identified at some dip sites (Table IV), especially in the immediate area around the dip bath and draining pen (DIPMAC Report, 1992; Barzi et al., 1996). Arsenic residues (<50 to >3000 mg As kg- l ) have also been identified in subsurface layers at depths exceeding 500 mm (Naidu et al., 1995). Some of the As concentrations present in the contaminated soil around livestock dips are comparable to As concentrations present in mine spoils but are perhaps more toxic because of the soluble nature of the As compounds present. Residential development of the contaminated sites may pose a considerable risk to human health. Similar problems also exist at former sheep dip sites due to the use of As-based pesticides by pastoralists in Australian Capital Temtory, Australia. Similar contaminated sites, resulting from animal dips, exist in Africa and many parts of the United States.

2. Herbicides Since the late 19th century, inorganic arsenical compounds were used as nonselective soil sterilants and weed killers (Vaughan, 1993). Because of their persistence in the soil and toxicity to humans and stock they were superseded by organoarsenical herbicides (McMillan, 1988). Monosodium methanearsonate (MSMA) and disodium methanearsonate (DSMA) have been used extensively as preemergence and postemergence herbicides in cotton and turf grasses (Sachs and Michael, 1971). Although both MSMA and DSMA are effective as selective grass suppressors, MSMA has been used almost exclusively in Australia (McMillan, 1988).The use of MSMA and DSMA in

162

E. SMITH ETAL.

the cotton industry has declined since the development of more effective contact herbicides. Despite the low mammalian toxicity of the methanearsonates, there is some concern with the ultimate fate of these As compounds, because they may persist in the soil environment andor be phyto-available (Hiltbold et al., 1974). Hiltbold et al. ( 1974) investigated the distribution of MSMA after repeated application to three soil types. After 6 years of applying MSMAat various rates (0-40 kg ha- year- I ) , the cumulative As concentration in the soils (0-900 mm) ranged from 10 to 3 1 mg As kg-' in a Decatur silt loam (rhodic paleudult), 2.4 to 23.6 mg As kg-' in a Hartsells fine sandy loam (typic hapludult), and 5.2 to 2 1.2mg As kg- in a Dothan loamy sand (plinthic paleudult). There was no apparent decline in the cotton yield over the 6-year period, and only low concentration of As were detected in the cotton seed (<0.2 mg As kg-'). Gilmor and Wells (l980), in contrast, reported that the residual effects of MSMA increased the sterility of rice cultivars grown in a Crowley silt loam soil (typic albaqualf). It was observed that the occurrence of straighthead in the rice cultivars increased with increasing application rates of MSMA (1.1-11.2 kg As ha-'). No sterility or yield decrease was found with either midseason draining and drying treatment, which is a common management practice to prevent straighthead, or at low rates of As application to the rice cultivars (Gilmor and Wells, 1980). Although many soils have repeatedly received applications of As-based herbicides, the highest reported concentrations of As residues have been observed in soils that have been treated with arsenical-based pesticides. This is mainly due to the differences between herbicide and pesticide application rates, with herbicides being applied at significantly lower rates (Vaughan, 1993).

'

3. Fertilizers Information on the effect of P fertilizers on the As content of soils is limited. Goodroad and Caldwell(l979) reported that there was no increase in As concentrations in a Nicollet clay loam soil (aquic hapludoll) and Port Byron silt loam soil (typic halpudoll) after receiving various P-fertilizer treatments. Phosphate fertilizer (concentrated superphosphate; about 20% P) was applied in two parts at various rates with a cumulative total of 0-8,888 kg ha-' of superphosphate applied to the Nicollet soil, while the Port Byron soil received five annual treatments of no P, 99 kg ha-' of concentrated superphosphate, 73 kg ha-' of calcium metaphosphate, 82 kg ha- * of phosphoric acid, and 352 kg ha-' of southern rock phosphate. Soil samples were collected from the Ap horizon (0-250 mm, Nicollet clay loam; and 0-225 mm, Port Byron silt loam). Charter et al. (1995) analyzed commercial phosphorus fertilizers marketed in Iowa and phosphate rocks used in the production of P fertilizers worldwide for

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

163

trace metal contamination. The concentrations of As and Mo were greater and more variable than other trace metals analyzed for in concentrated superphosphate, monoammonium phosphate (MAP), diammonium phosphate (DAP) and phosphate rocks (PRs). Arsenic concentrations in the samples were in the range 2.4-18.5 mg As kg-' (TSP), 8.1-17.8 mg As kg-I (MAP), 6.8-12.4 mg As kg-I (DAP), and 3.2-32.1 mg As kg-I (PRs).

V. ASTOXICITY For many soil pollutants, regulatory criteria for remediation or health are usually based on the effects on human health as an endpoint, but certain pollutants are toxic to plants and soil biota at concentrations that do not affect animal or human health. Arsenic is a good example of one such pollutant. The speciation of As in the environment is of critical importance because organic and inorganic compounds differ largely in their toxicity (Leonard, 1991). The toxicity of As is also related to the rate that it is metabolized from the body and the degree to which it accumulates in the tissues. The general pattern of toxicity is ASH, > As"' > As" > RAs-X (Fowler, 1977). Therefore, as a rule, inorganic arsenicals are more toxic than organic arsenicals, and the trivalent oxidation state is more toxic than the pentavalent oxidation state (NRCC, 1978). Although the pharmacokinetic aspects of As toxicology are important, they lie largely outside the scope of this review, and no attempt has been made to assess the vast amount of literature on this topic.

A. ACCUMULATION IN BIOTA Arsenic is present in many plants and plant products (Fowler, 1977) but typically does not exceed 1 mg As kg-' (Kiss et al., 1992). MacLean and Langille (1981) studied the uptake of As in apples grown on orchard sites and found that the As concentration in the peel and pulp of the fruit did not exceed 0.36 and 0.30 mg As kg-I, respectively. The uptake of As by radishes and silverbeet has been studied by Merry et al. (1986) in a glasshouse experiment with eight soils (3 X typic rhodoxeralf, 2 X ultic palexeralf, typic pelloxeralf, dystic xeropsamment, lithic xeropsamment). They found that with soil-As concentrations ranging between 26 and 260 mg As kg- none of the plants grown in these experiments contained As that exceeded currently accepted health limits for human consumption of 1 .O mg As kg-' (dry weight) (National Food Authority, 1993). The general assumption, therefore, is that in many situations the soil-plant transfer of As is low.

',

164

E. SMITH E T A .

Although the biochemical role of As in animals has been studied extensively (WHO, 1981; Petito and Beck, 1990), little is known about the biochemical role of As in plants (Kabata-Pendias and Pendias, 1992). Arsenic induces phytotoxicity that effectively protects humans from As poisoning. Phytotoxicity also results in restricted plant growth, which is undesirable (Sheppard, 1992). The effects of phytotoxicity have been reported to vary with soil type, with As being more toxic in sandy soils than in clay soils (Sheppard, 1992), which may be attributable to the greater As bioavailability in sandy soils. However, Jacobs et al. (1970) reported that As may stimulate plant yields at low concentrations in the soil. Similarly, microorganisms have been shown to display a range of sensitivities to As compounds, with the responses being dependent on soil type, the nature of the As species, and the concentration of As in the soil (Maliszewska et al., 1985).

B. HUMANEXPOSURE TO As Humans may be exposed to As from a variety of environmental sources, but food constitutes the largest source of As intake, with smaller contributions from air and drinking water (Chen and Lin, 1994). “Normal” As concentrations in human whole blood and urine have been reported to be about 100 and 15 pg As liter-’ (Fowler, 1977), respectively, but this may vary widely depending on environmental exposure. Approximately 5-1 5% of As ingested by humans is absorbed (NRCC, 1978), and As compounds are distributed in the liver, kidney, lungs, spleen, and the wall of the gastrointestinal tract within 24 hours of absorption. Some As may also be deposited in the bones, hair, nails, and skin (Leonard, 1991). Children may be exposed to higher amounts of As through the direct ingestion of soil. Effects of acute and chronic As poisoning in humans vary depending on the sex, age, dose, and duration of exposure (Fowler, 1977) and the chemical form and oxidation state of the As compound (NRCC, 1978). The acute effects caused by the ingestion of inorganic As compounds, mainly As,O,, are well documented in the literature. The fatal human dose for ingested As,O, ranges between 70 and 180 mg (WHO, 1981). Induction of cancer appears to be the most common long-term effect of chronic exposure to inorganic As. However, most animal experiments have not been able to demonstrate a direct relationship between As and carcinogenicity (Leonard, 199l), although epidemiological studies have demonstrated a causal relationship between environmental, occupational, and medicinal exposure of humans to inorganic As and cancer of the skin and lungs (National Academy of Sciences, 1977). Organoarsenic compounds that accumulate in marine seafood appear to pose little health risk to animals and humans, because the As compounds ingested are rapidly excreted in unchanged forms (Tamaki and Frankenberger, 1992). There are many clinical manifestations of As poisoning,

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

165

but the most commonly observed symptoms are conjunctivitis, melanosis, and hyperkeratosis (Das et al., 1996). Chronic As poisoning has been intermittently reported in the literature and is most commonly associated with As groundwater contamination (Lu, 1990). The most recent reported incident of chronic As poisoning induced by groundwater was in six districts of West Bengal, India. More than 800,000 people in this region are drinking As-contaminated water (total As range: 0.05-3.7 k g As liter-l), and at least 175,000 people show skin lesions caused by As poisoning (Das et al., 1996). The high As concentration in groundwater is geological in origin, and the water demands in the region are met mainly by groundwater resources. The main concern with contaminated soils is that the presence of As may pose an immediate or, more likely, a potential long-term hazard to the health of plants and animals, including humans. Various guidelines have been developed to provide a framework for the prevention, assessment, cleanup, and management of existing and future contaminated sites. For example, the Australian and New Zealand Environment and Conservation Council (ANZECC) have recommended maximum As concentrations in soils for the assessment and management of contaminated sites. These guidelines are generally followed by most Australian states. 0ther countries have developed their own soil-contamination criteria; however, many countries have not, using accepted soil contamination criteria from other countries (Tiller, 1992). The soil-contamination criteria of the U.S. Environmental Protection Agency, Britain, or the Netherlands are often quoted. All regulatory measures are based on total As concentration regardless of the well-accepted fact that the most toxic fraction is that which is bioavailable. Indeed, the term "bioavailability" is abused and shows lack of clarity by consultants, regulatory bodies, and researchers. Toxic substances such as As need special attention to delineate between total and bioavailable fractions.

VI. PWSIOCHEMICAL BEHAVIOR OF As IN SOIL Arsenic forms a variety of inorganic and organic compounds in soils (Vaughan, 1993 ) and is present mainly as inorganic species, either AsV or As"' (Masscheleyn et al., 1991). Under oxic soil conditions (Eh > 200 mV; pH 5-8), As is commonly present in the +5 oxidation state. However, As"' is the predominant form under reducing conditions (Masscheleyn et al., 1991 ;Marin el al., 1993). Both AsV and As"' species have been reported to be subject to chemical and/or microbial oxidation-reduction and methylation reactions in soils and sediments (Braman and Foreback, 1973; Brannon and Patrick, 1987). Many different As compounds have been identified in the soil environment (Table V), and they may be classified into two major groups: inorganic As compounds and organic As compounds.

166

E. SMITH ETAL. Table V Arsenic Compounds of Environmental Importance Name

Synonym

Inorganic As, trivalent As trioxide Arsenenous acid Arsenite As chloride As sulfide Arsine Inorganic As, pentavalent As oxide As acid Arsenenic acid Arsenate Organic As Methylarsonic acid Dimethylarsinic acid Trimethylarsine oxide Methylarsine Dimethylarsine Trimethylarsine Arsenobetaine Arsenocholine Arsanilic acid Cu acetoarsenite

Formula

As trioxide, arsenous oxide, white oxide Arsenious acid Salts of arsenous acid As trichloride As trisulfide, orpiment -

HAsO, H,AsO,-. HASO,,-, or As0,'AsCI,

As pentoxide Orthoarsenic acid Metaarsenic acid Salts of arsenic acid

As,O, H,AsO, HAsO, H,AsO,-,

Methanearsonic acid, or monomethylarsonic acid Cacodylic acid

CH,AsO(OH),

-

4-aminophenylarsonic acid Paris green

As203

As2S3 ASH,

or HAsOd2-, AsOd3-

(CH,),AsO(OH) (CH,),AsO CH,AsH, (CH,),AsH (CH,),As (CH,),As+CH,COOH (CH,),As CH,CH,OH H,NC,H,AsO(OH), Cu(CH,CO0),~3Cu(AsO2), +

"After Vaughan, 1993.

A. INORGANIC As COMPOUNDS Among the As species found in the soil environment, compounds of AsV and As"' are the most important inorganic As species in the soil, because their compounds are highly soluble in water (Vaughan, 1993) and may change valency states depending on the pH (Masscheleyn et al., 1991) and redox conditions (Marin et al., 1993). The equilibria for arsenic acid (As") and arsenous acid (As"') in aqueous solutions are given in Eqs. 1-6 (O'Neill, 1990). Arsenic acid

H,AsO,

+ H,O

H,AsO,

+ H,O+

pKa 2.20

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

167

+ H,O - w HASO:- + H,O+ + H,O - w AsO2- + H,O+

pKa 6.97

(2)

pKa 11.53

(3)

+ H,O H H,AsO; + H,O+ H,AsO; + H,O w As0;- + H,O+ HAsOS- + H,O w As0:- + H,O+

pKa 9.22

(4)

pKa 12.13

(5 )

pKa 13.4

(6)

H,AsO; HASO:Arsenous acid

H,AsO,

Geochemical systems are commonly interpreted in terms of their response to redox potential (Eh) and pH. The most thermodynamically stable species over the normal soil pH range 4-8 are H,AsO, (As"'), H,AsO;, and HASO:- (As").

B. ORGANIC As COMPOUNDS Organic As compounds (Table V) exist in both the trivalent and pentavalent states in soils (Vaughan, 1993). Microbial methylation of the As oxyanions may occur, forming methylarsenic compounds such as monomethylarsonics and di- and trimethylarsines (O'Neill, 1990; Vaughan, 1993), and ultimately may lead to the formation of arsine gas (NRCC, 1978). Different microorganisms vary in their ability to methylate inorganic As compounds present in the soil (NRCC, 1978). The methylation pathway for bacteria and fungi differ, with biomethylation of As by bacteria proceeding only to dimethylarsine, which is stable in the absence of oxygen. In comparison, fungi are able to transform inorganic and organic As compounds into volatile methylarsines (Cullen and Reimer, 1989; Tamaki and Frankenberger, 1992). Some microorganisms can methylate As compounds over a wide range of soil conditions, whereas other microorganisms are limited by the As substrates they can methylate and the degree of methylation of those substrates (NRCC, 1978). The equilibria for methylarsonic acid and dimethylarsinic acid in aqueous solution are given in Eqs. 7-9 (O'Neill, 1990). Few studies have reported the presence of organoarsenical compounds in soil systems, but this is probably due to the analytical difficulties of determining trace levels of organoarsenic species. Monnrneth.vlar.ronic acid

+ H,O w (CH,)AsO,(OH)- + H,O+ (CH,)AsO,(OH)- + H,O H (CH,)AsO:- + H,O+

(CH,)AsO(OH),-

pKa 4.19

(7)

pKa, 8.77

(8)

pKa 6.27

(9)

Dimethylarsinic acid

(CH,),AsO(OH)

+ H,O

w (CH,),AsO;

+ H,O'

E. SMITH E T A .

168

C. THESOILSOLUTION Limited information is available on the concentration and nature of As in soil solutions under field conditions. However, considerable information is available on the solubility and nature of As species in soils under known reducing conditions simulated in the laboratory (Deuel and Swoboda, 1972; Masscheleyn et al., 1991; Marin et al., 1993; Onken and Hossner, 1995).These studies reveal that under moderately reducing conditions As"' is the predominant species in the soil solution. Deuel and Swoboda (1972) found that there was an increase of As"' in soil solution over time under flooded soil conditions, which they attributed to the release of As during dissolution of iron oxyhydroxide minerals that have a strong affinityfor AsVunder aerobic conditions. Thus, minerals such as FeAsO, and other forms of Fe"' are reduced to the soluble Fe", and sorbed As" is released into solution (Takamatsu et al., 1982).These reactions are in general agreement with processes that have been observed in groundwaters. Intermittent incidents of As contamination of groundwater and the consequential As poisoning of people have been reported (Lu, 1990; Das et al., 1996). In a recently reported case of As poisoning in six districts encompassing an area of 34,000 km2 in West Bengal (Das et al., 1996),As concentrations were found to be above the maximum permissible limit established by the World Health Organization of 0.05 mg As liter-'. It has been proposed that As enters the groundwater through changes in the geochemical environment produced by the high withdrawal rate of groundwater. It is likely that the major mechanism of As release is through the decomposition of arsenopyrite according to Eq. 10 (Rimstidt et al., 1994).

+

FeAsS + 13Fe3+ 8H20

14Fe2++ SO$-

+ 13H+ + H,AsO,(aq)

(10)

Onken and Hossner ( 1995)identified the concentration and nature of As species present in the solution from two flooded soils (entic pelludert and typic ochraqualf) treated with sodium arsenate or sodium arsenite (rates of 0 4 5 mg As kg-I) in a glasshouse study. In soils treated with sodium arsenite (25 mg As kg- I), As"' was the major As species present in aqueous solution at day 0, but by day 10 conversion to As" had occurred (Midland silt loam, about 50%of total As present as AsV; Beaumont clay, about 20% of total As present as As") due to the relatively high redox potential of the soils. Similarly, a Midland silt loam treated with sodium arsenate (25 mg As kg-') contained no As"' at day 0, but by day 10 about 80% of total As was present as As"' in solution. The increase in As"' in aqueous solution resulted from the conversion of As" to As"' as the redox potential in the flooded soils declined. However, complete reduction of As" to As"' was not observed. Masscheleyn et al. (199 I ) studied the influence of redox potential and pH on As speciation and solubility in a contaminated soil (aeric ochraqualf). Changes in the redox potential and pH greatly affected the As species present in the soil solution. At higher soil redox potential (500-200 mV), As solubility was low, and AsV was

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

169

the predominant species in solution. Increasing pH or reducing AsV to As"' increased the concentration of As species in the solution. Based on these studies they concluded that the solubility of As under moderately reduced conditions was controlled by the dissolution of Fe hydroxides, which is consistent with the observations of Deuel and Swoboda (1972). More recent studies by Marin et al. (1993) support the earlier findings. The transition of As" to As"' is not surprising given that As"' is thermodynamically more stable than AsV under reducing soil conditions; i.e., where free electron activity (pE) + pH < 8, and pH < 6 (Sadiq et al., 1983). Despite the thermodynamic stability of As"' relative to As", numerous investigators (Masscheleyn et al., 1991; Marin et al., 1993) have reported the presence of As" in aqueous solution under reducing conditions. Masscheleyn et al. (199 l ) suggested two possible reasons for this: (1) competition of Fe"' as a terminal electron acceptor in microbial respiration (Eq. l l), or (2) presence of manganeseIv oxides (Eq. I2), which have been shown to be effective oxidants of As"' (Oscarson et al., 1981). Fe20, MnO,

+ 4H+ + AsOi+ 2H+ + AsOi-

+ + 2H,O e Mn2+ + As0;- + H,O 2Fe2+ As0;-

E" = 0.21 V

(1 1)

E" = 0.67 V

(12)

The redox potential of soil depends on the half-cell potentials (E") of all the reducing and oxidizing systems in the soil, and due to the heterogenetic nature of soils, these relationships are very complex.

D. ADSORPTION-DESORPTION PROCESSES As with many other contaminants, the concentration of As in the soil solution concentration is controlled both by soil physical and soil chemical properties that influence adsorption-desorption processes. Compared to the large volume of literature on metal adsorption by pure silicate and oxidic mineral systems, little information is available on As adsorption and transport in soils. Studies on pure systems suggest that As has a high affinity for oxidic surfaces, although reactivity of oxides may vary considerably, depending on pH, charge density, and soil solution composition. Soil texture (Wauchope, 1975; Frost and Griffin, 1977), nature of constituent minerals (Walsh er al., 1977; Pierce and Moore, 1980), pH, and the nature of competing ions have all been shown to influence adsorption processes. Few researchers have investigated the mechanisms involved in As sorption. The studies that have been conducted have generated considerableevidence for the formation of inner sphere complexes (specific adsorption) with soil components (Hingston et al., 1971 ;Anderson and Malotky, 1979). Direct evidence for the formation of AsV inner sphere complexes have been obtained using extended X-ray absorption fine structure (EXAFS) spectroscopy (Waychunas et al., 1993) and wide-angle X-ray scattering (Waychunas et al., 1996) on the ferrihydrite, infrared

170

E, SMITH ETAL.

spectroscopy by Lumsdon et al. (1984) on goethite, and on other hydrous Fe oxides by Harrison and Berkheiser (1982). Waychunas and co-workers (1993,1996) have postulated that AsV adsorbs onto ferrihydrite by forming binuclear, inner sphere complexes. However, monodentate complexes were also observed and accounted for approximately 30% of all As-Fe correlations (Waychunas et al., 1993), and monodentate AsV-Fecomplexes were comparable to the number of bidentate AsV-Fecomplexes at low total As concentrations. Similarly, Fendorf et al. (1997) investigated the surface structure of AsV and chromate sorption on goethitc. They concluded, from EXAFS spectroscopy examination of the surface, that AsV formed three different complexes on goethite. A monodentate complex was favored at low surface coverage, whereas the bidentate complexes were favored at higher surface coverages. Indirect methods have also been used to study sorption mechanisms. Specific anion adsorption produces a shift in the zero-point charge (pH ) of the adsorbent. Pierce and Moore (1980) investigated the sorption of AsIfiZnto amorphous Fe hydroxide and observed that the pHzF decreased with increasing addition of As"'. This was assumed to be indicative of As"' being specifically sorbed to the hydrous Fe hydroxide surface. Theoretically, all the adsorbed metal may be desorbed from the soil constituents. However, investigations to date report that substantial proportions of trace metals sorbed by soil constituents are not readily released into the soil solution. Few studies to date have investigated desorption of As from soil constituents. Phosphate has been reported to displace adsorbed As from soils (Woolson et al., 1973; Peryea, 1991). Heavy additions of P to As-polluted soils have been reported to displace approximately 77% of the total As in the soil, with the water-soluble As fraction being redistributed to lower depths in the soil profile (Woolson et al., 1973). Peryea (1991) observed that although P increased As solubility, desorption of As was dependent on the soil type, with the As concentration in soil solution from a volcanic soil not altering after the addition of P. These volcanic soils have high anion-fixing and pH-buffering capacities due to the presence of allophanic minerals; this implies that only large additions of P to high anion-fixing soils may affect As solubility. In leaching experiments with columns of repacked soils collected from PbAs0,-contaminated apple orchards, the addition of P in the form of MAP or monocalcium P (MCP) significantly increased the amount of As leached from the columns (Davenport and Peryea, 1991). Although As"' has been recognized as being more mobile and toxic than As", there have been few reported desorption studies of As"' from the soil. Tammes and de Lint (1969) found that even after extremely high concentrations of As"' applications, symptoms of phytotoxicity in potatoes gradually disappeared over time, which probably indicates the leaching of As"' from the soil root zone. Elkhatib et al. (1984), in contrast, reported that As"' sorption was not reversible, since only a small amount of the sorbed As"' was released after five desorption steps. Considering the importance of desorption processes in controlling As concen-

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

171

trations in soil solution, further studies in this area are urgently needed for a more complete understanding of desorption processes.

I . Soil Properties Among the factors that influence sorption of As, soil properties have been most extensively studied. These investigations show that both the amount of clay and the nature of constituent clay minerals control As adsorption in soils. Johnson and Hiltbold ( 1969) reported that approximately 90% of As present in the soil was associated with the clay fraction after 4 years of repeated applications of MSMA, monoammonium methanearsonate (MAMA), and DSMA to turf. Livesey and Huang (198 1 ) studied retention of As" in four Saskatchewan surface soils (orthic dark gray Carrot River, orthic black Melfort, orthic black Oxbow, low-humic eluviated gleysol Oxbow) and reported that at dilute As concentrations adsorption-desorption processes controlled retention of As. They found that sorption was linearly related to ammonium oxalate-extractable Al and, to a lesser extent, to clay and ammonium oxalate-extractable Fe. Wauchope ( 1975) investigated the adsorption of As", P, DSMA, and the sodium salt of dimethylarsinic acid by 16 Mississippi River alluvial flood-plain soils. Sorption of the two organoarsenical herbicides was strongly correlated with As" and P sorption, and all four As species were found to be correlated @ < 0.01) with the clay and Fe oxide contents of the soils. This suggests that Fe and Fe coatings on clay surfaces may be important in controlling As adsorption-desorption processes in soils. Gustafsson and Jacks (1995) examined As solid-phase relationships in forest soil profiles. They reported that As" was the dominant As species present in the soil (entic haplocryod, typic halpocryod, typic cryorthent). On addition of As" to the soil, it was found that adsorption of As by imogolite-type materials and fenihydnte were the key properties that determined As" concentration in solution. Clays may often be coated with Fe and Al oxides (Shuman, 1976; Schutless and Huang, 1990; Naidu et al., 1990; Naidu etaf., 1994), and this may modify As-clay interactions. Fordham and Nomsh (1979) reported that clay minerals were relatively unimportant in comparison with Fe oxides and, to a lesser extent, titanium oxides in the adsorption of As" in several acidic soils. Fordham and Norrish ( 1983), in a later study, reported that the adsorption of As" by a lateritic podzol surface soil (palexeralf) was controlled mainly by Fe oxides of approximately 50 nm diameter. Iron oxides were associated with other soil components, forming surface deposits on larger kaolin flakes or microaggregates with smaller flakes. Titanium oxides competed with Fe oxides for As" and were able to dominate As" adsorption when Fe oxides were chemically removed. Similarly, Elkhatib et al. (1984) examined As"' sorption in the A and B horizons of five major West Virginian soils (typic hapludults, typic haplualf, and fluventic dystrocrept). Iron OXide and pH were the soil properties most related to As"' sorption.

172

I -Mn

E. SMITH ETAL.

I

I

-0 -Mn -OH

-Mn-

I I -MnI 0 I -MnI

I I -Mn-0 -Mn-OH I I 0 0 I I -Mn-0 -Mn-OH I I I I

0

0

0

surfaces

I I 0 I

E

0 -Mn

0-

I Mn -OH I

Release of Mnll

I I -Mn -0-Mn-OH I I OH 0 I Mn2+ -Mn-OH I OH 0 I I -Mn -0 -Mn-OH I I

,OH

-As

0 I

I I 0 I -MnI 0 I

I I 0 OH I I 0 -Mn -As= I I 0 OH I -Mn-O- Mn-OH I I

-Mn-

0- Mn -OH

'OH

0 -Mn -OH

0

D

I I -Mn0- Mn -OH I I OH 0 I I -Mn-0-Mn-OH I I OH 0 I I -Mn0- Mn -OH I I

OH

I I

0-Mn=O OH

Figure 3 Proposed schematic representation of the cross-section of the surface layer of a Mn'" oxide (birnessite) and the proposed As"' adsorption and subsequentAs" oxidation and release (reprinted with permission from Scott and Morgan, 1995.0 1995 American Chemical Society).

Oscarson et al. ( I 983a) reported that Mn oxides may play an important role in the adsorption of As"' and As" from soil solution as well as the oxidation of the more toxic and mobile As"' to As" (Oscarson et al., 1981). Sorption of As by Mn oxides after the addition of As"' to soil solution (pH 7) was reported to be in the order: cryptomelane (a-MnO,) > birnessite (6-Mn0,) > pyrolusite (p-MnO,) (Oscarson et al., 1983a). The amount of As adsorbed by Mn oxides appears to be related to the pHZ and the surface area of the oxides, as well as to the oxidation of As"' to As". TLs implies that in some environments that have been contaminated with As"', the presence of Mn oxides such as cryptomelane or birnessite in the system may decrease the potential toxicity of As"' by converting As"' to the less toxic As" and the subsequent adsorption of this species. Scott and Morgan (1995) proposed that the surface redox reactions between As"' and a Mn'" oxide (6-Mn0,) occurs through a multiprocess mechanism (Fig. 3 ) . Oscarson et al. (1983b), in a later study, reported that Fe and Al oxides and CaCO, coatings deposited on Mn oxides affected the adsorption of As from solution. The coatings

ARSENIC IN THE SOIL ENVIRONMENT:A REVIEW

173

evidently masked the electron-accepting sites on Mn dioxides for converting As"' to AsV (Oscarson et al., 1983b).

2. Effect of pH The effect of pH on As sorption has been studied widely using both pure mineral systems and soils (Frost and Griffin, 1977; Pierce and Moore, 1980; Xu er d., 1988, 1991). These investigations showed that the pH of the soil solution has a large influence on adsorption of As. Generally, the effect of pH on sorption varies with the As species. Frost and Griffin ( 1977)reported that As" sorption by the layer silicate minerals kaolinite and montmorillonite exhibited a maximum pH of 4-6. Arsenite, in contrast, was adsorbed steadily from pH 4 to 9 on kaolinite and peaked at pH 7 on montmorillonite. Goldberg and Glaubig (1988) also investigated sorption of AsV on montmorillonite, kaolinite, and calcite. The shape of the sorption curves closely agreed with those found by Frost and Griffin (1977). However, in contrast to Frost and Griffin (1977), similar amounts of As" were sorbed onto both kaolinite and montmorillonite. This may be attributable to the similar surface areas of the kaolinite and montmorillonite clays (Table VI) used by Goldberg and Glaubig (1 988), compared with those used by Frost and Griffin (1 977). Perhaps importantly, Goldberg and Glaubig (1988) found that carbonates play an important role in As" sorption in the pH 9-12 range. Xu and his co-workers (Xu et al., 1988, 1991) studied the adsorption of AsV, monomethylarsonic acid (MMAA), dimethylarsinic acid (DMAA), and As"' on alumina, hematite, and quartz. The adsorption of all four forms of As was strongly influenced by pH (Fig. 4), and this was attributed to the pH-dependent charge and the distribution of As species in soil solution. Based on stability constants, H,AsO; and HASO:- are the main AsV species, H,AsO, the main As"' species, CH,AsO,OH- the main MMAA species, and either (CH,),AsO(OH) or (CH,),AsO; the main DMAA (pKa 6.2) species present in the pH range (pH 4-9). The pHzpcfor alumina and haematite is approximately pH 6.5-7, and the solid surfaces are therefore negatively charged at a pH above this, which may explain why the adsorption of AsV (in deprotonated form) rapidly decreases above pH 7.

Table VI Surface Area of Kaolinite and Montmorillonite Clays Surface area (rn2g-1) Kaolinite 34.2 20.5

Montrnorillonite

Source

86.0 18.6

Frost and Griffin, 1977 Goldberg and Glaubig, 1988

174

E. SMITH ETAL. -0-As' -0- As"' -M- MMAA

0 DMAA

PH Figure 4 Adsorption of As". As"'. MMAA, and DMAA on alumina as a function of pH (As", As"') = 10-hM: MMAA, DMAA = 10-8M; adsorbent-solution = 25 g liter-') (Xu et al., 1991; reprinted with kind permission from Kluwer Academic Publishers).

The P H : ~of quartz is approximately 2, and surfaces are negatively charged, thus depressing the adsorption of As". Similar reasoning may be used to explain the adsorption of As"' and organoarsenic species to the solids. However, Xu er al. (1991) noted adsorption discrepancies in the sorption maxima of DMAA and MMAA for both compounds and suggested that other factors may also affect the sorption of these ions.

3. Effect of Competing Ions Appreciable quantities of both inorganic and organic ligand ions are present in many soils and aquatic systems. This is especially true for Australian soils where over 30% of the soils are affected by salt (Naidu et al., 1993) and the ligand ions generally include C1-, SO:-, PO:- ions (Naidu and Rengasamy, 1993). In addition, soils contain organic ligands arising from both plant root exudates and decomposing plant residues (Harter and Naidu, 1995). Competition for adsorption sites between some of these ligand ions and As can appreciably affect the amounts of As sorbed. Phosphate is known to displace sorbed As from soils (Woolson et al., 1973).Applications of relatively high rates of P fertilizers (about 8-12 mmol P kg-' soil [Peryea, 19911 and 0-48.6 mmol P kg-' soil [Melamed el al., 19951) have been shown to enhance As mobility in laboratory columns (Melamed er al., 1995) and As solution concentration in laboratory batch studies (Peryea, 1991). The presence of P in the equilibrating solution has been reported to suppress the adsorption of As, whereas the addition of CI-, NO;, and SO:- to the equilibrating solution had little significant effect on As adsorption (Livesey and Huang,

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

175

1981). Roy et ul. ( I 986) reported P and molybdate (Mo) suppressed the adsorption of As on a Cecil clay (typic hapludult), with P being more effective than Mo at suppressing AsV adsorption. Competitive adsorption equations of the Freundlich type developed by DiGiano et al. (1978) and Shiendorf et al. (198 1) for the adsorption of dilute organic compounds by activated carbon appear to be potentially useful for describing competitive interactions of AsV with other ions on clay (Roy et al., 1986). Both equations appeared to apply only in situations where the equilibrium concentration of the competing P or Mo anion was much less than that of AsV (Roy et al., 1986), which is not the case in many soil solutions. Recent studies of competitive adsorption interactions of anions on pure mineral systems (Xu et al., 1988) suggests that at pH < 7, the SO:- (20 mg liter-') anion decreased the adsorption of AsV on alumina (Fig. 5). Increasing the SO:- concentration (40 mg liter-') has little effect on AsV adsorption, indicating the sorption mechanisms of AsV and SO:- are not identical (Xu et al., 1988). The presence of fulvic acid greatly affected the adsorption of AsV on alumina at a pH between 3 and 7.5 (Xu et al., 1988). Fulvic acid may be adsorbed on alumina by coulombic attraction (Xu et al., 1988), or fulvic acid may react directly with As (Thanabalasingam and Pickering, 1986), which tends to decrease the adsorption of the corresponding As complex (Xu et al., 1988). Few studies have investigated the adsorption of As by organic matter. Thanabalasingam and Pickering (1986) reported that adsorption of AsV and As"' by humic acid was pH-dependent. This trend was more apparent when a high-ash-containing humic acid was used. The maximum adsorption of As" occurred at approximately pH 5.5, whereas As"' maxima occurred at a much higher pH of 8.5. Adsorption of As"' was less than AsV, which is a trend that has been noted by other authors (Frost and -0 so:-0- SO:-D SO:-

= 0.0 mg

4-

b 3 -

-H- SO:-

=

v

0

5

6 2 Y

0

P i -

= =

liter-' 20.0 mg liter-' 40.0 mg liter-' 80.0 mg liter-'

176

E. SMITH ETAL.

Griffin, 1977).The general behavior of the As species was largely attributed to humic acids becoming more soluble as pH increased (more alkaline), which decreased their ability to remove As from solution. Alternatively, the observed pH effect could reflect the changes in the protonation of both the adsorbent and absorbate (Thanabalasingam and Picketing, 1986).

4. Other Effects on As Adsorption Although numerous studies have investigated the effects of pH and ion competition on adsorption behavior of As in soil systems, few studies have considered other factors, such as ionic strength and index cations (e.g., Na and Ca). Many of these other factors, however, have been studied with other anions such as P and S . There is considerable data showing that when P is adsorbed by soil, or a soil constituent, adsorption varies with the concentration and nature of the background solution, although the underlying mechanisms are open to debate. The effects of different cations have been attributed to a number of mechanisms, including the formation of surface P complexes with divalent cations (Heylar et af., 1976), the formation of insoluble Ca-P compounds (Freeman and Rowell, 198l), and differences in surface electrostatic potential (Barrow, 1983; Curtin et al., 1992). Similarly, differences in ionic strength have been shown to affect P (Barrow, 1984; Bolan et af., 1986), S (Bolan et al., 1986; Ajwa and Tabatabai, 1995), and other anions in soil solution. In the case of P, increasing ionic strength has been shown to decrease the adsorption of P below the zpc and to promote the adsorption of P above the zpc on variable-charged surfaces (Barrow et al., 1980; Bolan et al., 1986). Researchers suggest that the effects of ionic strength operate through its effect on the electrostatic potential in the plane of adsorption (Bolan et al., 1986). Therefore, at a pH above the zpc of a variable-charge surface, increasing ionic strength decreases the negative potential in the plane of adsorption, whereas at pH less than zpc it decreases the positive potential in the plane of adsorption (Barrow et al., 1980).Although this type of adsorption behavior for P is displayed over a wide range of ionic strengths, other anions, such as S , behave differently on variable-charge surfaces. Although little information is available about these affects of soil solution composition on the adsorption of As, studies (Woolson et al., 1973; Barrow, 1974; Peryea, 1991) have shown that P and AsV behave very similarly in soils. This suggests that the affects of ionic strength and different index cations on the behavior of As" adsorption are similar to the adsorption behavior of P. Current unpublished data in our laboratory confirm this conclusion.

E. KINETICSOF As ADSORPTION-DESORPTION Adsorption and desorption processes are the principal factors affecting the transport, degradation, and biological availability of compounds in soils. Numer-

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

177

ous studies have accumulated a large amount of data from anion adsorption and interpreted these results with adsorption models. Application of these models to soil systems is not simple, because a number of soil processes complicate the measurement and interpretation of the results. However, it is generally agreed that in most cases anionic adsorption is bimodal, since it occurs in fast and slow stages. The adsorption of both AsVand As'" on pure minerals and soils has been well studied (refer to earlier sections), and researchers have found that the initial adsorption of both As species is rapid (Anderson ef al., 1976; Pierce and Moore, 1980; Elkhatib et al., 1984; Scott and Morgan, 1995). Anderson et al. (1976) reported that the rate of AsV adsorption on aluminium hydroxide was initially rapid, with over 90% of the adsorption reaction (75 mg As liter-' added) taking place before the sample could be collected. After 1 hour the rate of AsV adsorption slowed considerably, but As" adsorption continued at a slow rate for the duration of the observational period (70 hours). Similarly, Elkhatib et al. (1984) reported that the initial reaction of As"' with five surface (A) and subsurface (B) soils having a range of chemical and physical properties (coarse-loamy, mixed, mesic typic hapludults; fine-loamy, mixed, mesic typic hapludults; coarse-loamy, mixed, mesic fluventic dystrochrepts; fine, mixed, mesic typic hapludalfs) was rapid, with more than 50% (5-500 mg As"' liter-') of the original As"' present being adsorbed in the first 30 min. Conventional methods (batch and flow methods) are too slow to observe the kinetics of most surface chemical reactions. Pressure-jump @-jump) relaxation technique allows the determination of extremely rapid surface chemical reactions. Grossl ef al. (1997) investigated the rapid adsorption-desorption of AsV and chromate on goethite using the p-jump technique. From information elucidated using this technique, Grossl et al. (1997) proposed that the rapid adsorption of AsV on goethite was a two-step process that resulted in the formation of an inner sphere bidentate complex. This was in general agreement with EXAFS spectroscopy data obtained by Fendorf et al. (1997). For many of these batch studies, the apparent equilibrium between the solution and the solid was assumed to be reached in a few days. However, few studies have investigated the long-term adsorption behavior of As in soils, and the magnitude of this slow fraction is unknown. Examples from long-term studies with other elements indicate that the apparent adsorption distribution coefficient (Kd)can increase as much as 10-fold between short contact times (1-3 days) and long times (Pignatello and Xing, 1996). In contrast to adsorption studies, little information is available on the desorption of As or other elements from soils. Elkhatib et al. (1984) reported that as As"' desorption was quite hysteric and only slowly desorbed from five soils where As"' had been in contact with the soils for 24 hours. Carbonell Barrachina et al. (1996) have also studied the desorption behaviour of As"' and found that As"' sorption was a reversible process from three freshly saturated soils (aridisol gypsiorthid torriothent, inceptisol haplumbret dystochrepts, and entisol torriorthent haplargid calciorthid). Carbonell Barrachina et al. (1996) have suggested that the

178

E. SMITH ETAL.

difference between their results and Elkhatib et al. (1984) may be explained through the different sorption capacities of the soils, since the soils used by Elkhatib et al. (1984) had a greater sorption capacity than those used by Carbonell Barrachina et d.(1996). Other ions, such as P, exhibit desorption behavior similar to that ofAs. Many desorption studies often reveal a fast desorbable fraction followed by a slow fraction (Garcia-Rodeja and Gil-Sortes, 1995; Lookman et al., 1995). Garcia-Rodeja and Gil-Sortes (1995) studied the desorption of P from 3 1 surfacesoil samples collected from northwest Spain that had been spiked with varying amounts of P (200-2000 mg P kg-') and maintained at 75% field capacity for 1 year. They reported that Pdesorption was initially very rapid but over time became progressively slower. Lookman et al. (1995) also studied the desorption of P from 44 soil samples in long-term spiking trials. They reported that the desorption of P could be described by considering that P occupies two discrete pools-in one pool P was readily available, and in the other pool P was strongly fixed and desorbed slowly. The desorption of P from the fast and slow pools could be described by a two-component first-order model (Eq. 13), where Ql,oand Q2.0are the amounts of P initially present in the labile pool, k , and k, are the rate of desorption from each descrete pool, and t is the time. Q,,,(t) = Q,,o(l - e P k l . 7 + QJl

- e-k2.t)

(13)

However, equilibrium and thermodynamic considerations make it difficult to visualize the presence of such discrete pools of P. Mathematical equations of the nature described here assist us in explaining the trends in desorption, but they often fall short of the mechanism of interactions. Kinetic studies are becoming increasingly important in clarifying adsorptiondesorption processes (Pignatello and Xing, 1996). Adsorption data may be fitted to any number of equations, ranging from zero- and second-order equations to the parabolic diffusion law, the Elovich equation, and the modified Freundlich equation (Table VII). The particular equation used to describe adsorption-desorption rates is the one that best fits the data. Few researchers have investigated the adsorptiondesorption kinetics of As, but Elkhatib et al. ( I 984) found that the Elovich and modified Freundlich equations described the sorption kinetics of As"' by 10 surface and subsurface soils, whereas the modified Freundlich equation described the desorption kinetics of As"'. The fact that the Elovich equation did not describe the desorption rate of As"' may be indicative of the premises on which the equation is based and may limit the application of the equation across a broad spectrum of soils. A number of studies have investigated the adsorption of AsV and As"' by pure minerals and soils. Many of these studies employed the Freundlich or Langmuir equations, but these equations do not adequately describe the adsorption of As to surfaces. Surface-complexation models are chemical models that give a general molecular description of adsorption using an equilibrium approach (Goldberg,

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

179

Table VII Summary of Equations Used to Describe Various Kinetic Models" Equation Zero order First order Second order Parabolic diffusion Two-constant rate Elovich-type equation Modified Freundlich

Formulah C, = C, + Kr InC, = InC,, + Kt VC,= K,, - Kt C, = C, + Kdt InC, = InC,, + K In t C, = C, K In t InC, = InC, + p, In t

+

+ p2C+ &T

"Garcia-Rodeja and Gil-Sortes, 1995: reprinted by permission of the publisher. "C, = final concentration of adsorbate as a function of contact time; C, = initial concentration of adsorbate; K = rate constant; t = time; p and Tare independent variables.

1992). Only a few studies (Goldberg, 1986; Goldberg and Glaubig, 1988; Belzile and Tessier, 1990; Manning and Goldberg, 1996) have investigated surface complexation modeling as a means of quantifying the adsorption of AsV by different surfaces. The constant-capacitance model (CCM) has been used to describe AsV adsorption on pure systems (Goldberg, 1986; Manning and Goldberg, 1996) and soils (Goldberg and Glaubig, 1988).Goldberg and Glaubig (1988) showed that the CCM adequately described AsV adsorption on an Imperial soil series (fine, montmorillonitic, hyperthermic vertic torrifluvent) up to pH 9 but was unable to describe As" adsorption in the pH 9-12 range. Similarly, Manning and Goldberg ( 1 996) reported that the CCM adequately described adsorption envelopes of AsV, P, and Mo by goethite and gibbsite. However, the authors applied the CCM model using both the one-site (monodentate) and two-site (monodentate bidentate) conceptualization of the oxide surface. The CCM, using both these approaches, gave similar descriptions of the experimental data, indicating that the present understanding of anion adsorption on mineral surfaces is not complete (Manning and Goldberg, 1996).

+

VII. SOIL As AND VEGETATION Arsenical compounds have been widely used as pesticides and herbicides in agriculture. Attention has therefore focused on the accumulation of As in agricultural soils and the possible toxic effects on plant production. Extensive research on the effects of As on plant production are well documented (Jacobs ef al., 1970;

180

E. SMITH ETAL.

Steevens et al., 1972; Anastasia and Kender, 1973; MacLean and Leangille, 1981; Jiang and Singh, 1994).

A. SOILAs AND PLANTUPTAKE The accumulation of As in the edible parts of most plants is generally low (Vaughan, 1993; O'Neill, 1995). Plants seldom accumulate As at concentrations hazardous to human and animal health because phytotoxicity usually occurs before such concentrations are reached (Walsh and Keeney, 1975). Thus, the major hazard for animal and human systems is ingesting As-contaminated soils or consuming contaminated water. Uptake of As by plants occurs primarily through the root system, and the highest As concentrations are reported in plant roots and tubers (Anastasia and Kender, 1973; Marin et al., 1993). Therefore, tuber crops (e.g., potatoes) could be expected to have higher As concentrations than other crop types when grown in polluted soils. This appears not to be the case, since potatoes grown in a sandy soil that received As additions ranging from 45 to 720 kg As ha-' accumulated only 0.5 mg As kg-' in the potato tuber (Jacobs et al., 1970). In contrast, the external potato peels had As concentrations of up to 84 mg As kg- I . This was attributed to contamination from soil adhering to the surface peels. 1. Crop Tolerance to As A considerable variation in plant sensitivity to As exists among plant species (Jacobs et al., 1970; Jiang and Singh, 1994). Vegetable crops grown in three soils (Lakeland loamy sand, Hagerstown clay loam, and Christiana clay loam) in a greenhouse trial exhibited a range of sensitivities to sodium arsenate (0-500 mg As kg- I ) . Plant sensitivity followed the order green beans > lima beans = spinach > radish > tomato > cabbage (Woolson, 1973). The symptoms of phytotoxicity may vary between species. Tomato plants grown in soils with high As background concentrations (100-130 mg As kg- I ) showed leaf dieback from the tip and poor-quality fruit set (Fergus, 1955). Fruit trees grown on replanted orchard sites commonly exhibit retarded early growth, to which As toxicity may contribute (Davenport and Peryea, 1991). Similarly, rice grown on former cotton-producing soils that had a history of repeated MSMA applications showed indications of susceptibility to straighthead disease (abnormally developed or sterile flowers resulting in low grain yields) under flooded soils conditions (Wells and Gilmor, 1977). The range of soil-As concentrations that may be phytotoxic is summarized in Fig. 6. Although the data are not extensive, they highlight both the broad range of concentrations of soil As over which toxicity symptoms may occur and the narrow margin that exists between background con-

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

181

A

IE I

I

,TF I&

Figure 6 Range of As concentrations in soils at which crops may exhibit phytotoxic symptoms. (A) Woolson, 1973; ( B ) Jacobs ef a/., 1970; (C) Woolson ef a/., 1973; (D) Woolson et al., 1971b; (E) Steevens ef al., 1972; (F)Wells and Gilmor, 1977.

centrations of As (<40 mg As kg- in most soils) and phytotoxic concentrations. As discussed earlier, the total concentration of As in soil is a poor indicator of its bioavailability. The concept of bioavailability generally refers to some fraction of the total amount of As present in the soil that best correlates with plant availability (Pierynski et al., 1994). Woolson et al. (1971a) related plant-available As to the total soil As and found that the amount of water-soluble As was better correlated with plant growth than with total soil As. Plants absorb nutrients and metals from the soil solution, and the “bioavailable” As soil pool may be a better indicator than total As concentration of plant phytotoxicity in the soil (Woolson et al., 1971; O’Neill, 1995). Sadiq (1986) reported that As concentration in maize was correlated with the water-extractable As but not with the total As concentrations in calcareous soils. Jiang and Singh (1994) found a similar relationship between the As concentrations of barley and ryegrass plants grown in greenhouse experiments and soils (typic udipsamment and humaquept) fortified with As. Reviewing the literature on As, Sheppard (1992) concluded that soil type is the only significant variable when considering plant phytotoxicity for inorganic As. It was reported that inorganic As was five times more toxic to plants in sands (mean = 40 mg As k g ’ ) than in clay (mean = 200 mg As kg-’) soils. Arsenic phytotoxicity is expected to be greater in sandy soils than in other soil types, since sandy soils generally contain low amounts of Fe and A1 oxides and clays. These soil constituents have been implicated in the adsorption of As from solution in soils.

182

E. SMITH ETAL.

The nature of As species in the soil solution may also determine the phytotoxicity. Although As is primarily present as AsV or As"' in the soil-water environment (Bohn, 1976), MMAA and DMAA compounds may also be present. Marin et af. (1992) reported that both As']' and MMAA were found to be phytotoxic to rice plants grown in nutrient solution, whereas AsV did not affect plant growth at the same concentration (0.8 mg As liter-l). The degree of As uptake by rice followed the trend As"' > MMAA > AsV > DMAA. However, these observations were made after nutrient solutions were amended with different As compounds to produce concentrations (0.05,0.20, and 0.80 mg As literp1)that were much higher than those encountered under normal field conditions. Furthermore, the effect of soil As on plant growth may be complicated by the presence of competing anions, notably P. Phosphate and As exhibit similar physiochemical behavior in the soil, and both synergistic and antagonistic effects of P on uptake of As by plants have been reported. Davenport and Peryea (1991) have reported a reduction of As uptake by plants after the application of P. In other studies, P has been found to increase As availability and therefore to increase the As concentration in plants (Woolson, 1973). The increased availability of As is particularly noticeable in sandy soils, where fewer adsorption sites are present and added Pmay displace some of the bound As ions into soil solution (O'Neill, 1995). Apart from P, N fertilizers, lime, and SO:- have also been observed to alleviate or depress the availability of As to plants (Thomas, 1977; Merry et al., 1986).

VIII. SOIL As AND MICROORGANISMS The effects of metal pollution on biotic communities have been extensively studied, and many of these studies have focused on the adaptation of bacterial communities through the development of resistance or tolerance mechanisms. Biological transformations are important in redistributing As in soils (Fig. 7). Arsenic may have a direct influence on the microbial populations present in the soil. Decreases have been reported in microbial populations in soils that have been polluted with As compounds (Malone, 1971; Bisessar, 1982; Maliszewska et al., 1985). Bisessar (1982) reported that the decrease in population counts of bacteria, actinomycetes, fungi, and nematodes was significantly correlated (p 5 0.05) with concentrations of As in soil collected from several sites near a secondary Pb smelter (Table VIII). Although As may have influenced a decline in the soil microflora population, the decline of the soil biological population was probably due to the combined effects of all metals present. Speir et af. (1992), in contrast, reported that very few negative effects were attributable to Cu, Cr, and As in a pot trial conducted to assess the feasibility of using CCA and boric-treated sawdust as a soil amendment. The total concentrations of elements in the pot trial (45, 136,

183

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

Industrial Sources

(CH&&O(OH) (CH&AsO

Arsenical Herbicides and Pesticldes

T

Local

Dust

I I Consumptton iH

Oxygen present HAs02

1

As406

I

Oxygen absent H2A~04-

H2AS04-

HAs02

Leaching Figure 7 Soil-air cycle (WHO, 198 1; reprinted by permission of the publisher).

63, and 32 mg kg- for Cu, Cr, As, and B, respectively) were considerably lower than those present in Bisessar's (1982) work. Considerable variation in tolerance to As compounds applied to the soil has been shown by the soil microflora. Maliszewska et al. (1985) reported that As"' compounds were more toxic than As" compounds to microorganisms important to maintaining soil fertility.Arsenite compounds were applied at 500, 1500,and 3000 mg As kg- I and As" compounds at 1000,5000, and 10,000 mg As kg- to sandy and alluvial soils. The influence of As"' and As" compounds on the soil microflora differed, depending on the microflora and the soil type. In sandy soil, As"' depressed the growth of bacteria, whereas As" stimulated their proliferation. Overall, both As compounds had little effect on the development of actinomycetes and fungi flora and suppressed the growth of Azotobacter sp. Maliszewska et al. (1985) also measured a decrease of approximately 30% in dehydrogenase activity in the soils. The decrease in activity was greatest in the sandy soil and with the application of As"'. Dehydrogenase is an unspecific enzyme for assessing the effect of As compounds on the soil microflora; other authors have found that enzymes involved in more specific biochemical reactions are inhibited by the addition of As to the soil environment. Tabatabai (1977) reported that As"' greatly inhibited the urease activity of some soils at a concentration of 375 mg As"' kg- I , but the addition of As" at a similar concentration to the soil had no effect on activity. Bardgett er al. (1 994) investigated microbial properties

Table VIII Total As, Pb, Cd, and Cu Concentration and Counts of Sod Microflora Near a Secondary Pb SmelteP Total heavy metal concentration (mg kg-I) Location (eastofsource)

Pb

As

Cd

Total no. of colonies Cu

Bacteria

Actinomyces

~ _ _ _ _ _ _

15 m 90 m 150 m 180 m

Control (1000 m south)

28,000 8333 4800

3564 703

972 554 230 163 57

151

599 398 287

102 33 26

333

5

73

10.5 12 12.6 15.1 17.2

Fungi

in log soil Nematodes

_ _ _ _ _ ~

1

1.4 1.6 2.1 2.4

“Bisessar, 1982; reprinted with kind permission from Kluwer Academic Publishers.

2.5 3.2 3.5 4.3 4.6

Earthworms

~

16 30.9 26.6 58 98

0

0 0 1.3 2.3

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

185

of the surface (0-50 mm) of a pasture soil contaminated by runoff by preserving liquor runoff from an adjacent timber-treatment plant. Total concentrations of Cr, Cu, and As ranged from 86 to 1260,70 to 1233, and 79 to 1265 mg kg-', respectively. They found no significant relationship between soil As, Cu, and Cr concentrations and urease activity, whereas a significant relationship was observed between As, Cu, and Cr (? values = As, 0.923; Cu, 0.933; Cr, 0.872) and a decline in sulphatase activity. However, many factors may influence the inhibitory effect of metals on the soil microflora, with soil type being an important factor in determining the bioavailability of the metals (Maliszewska et al., 1985).

A. BIOTRANSFORMATION OF As INSOILS The biotransformation of As in soils has been recognized for many years. Two microbial processes, oxidation and reduction, are of current interest because of their possible application for bioremediation of contaminated soil. A number of review articles have discussed As and its transformation in various environments (Cullen and Reimer, 1989; Ehrlich, 1996), and these articles provide a more in depth review of the subject than is provided here. Bacterial oxidation was first identified by Green in 1918 wh en a bacterium from a cattle-dipping fluid was isolated. Many other heterotrophic As"' oxidizing bacteria have since been characterized, with many bacterium classified as Bacillus or Pseudomonas spp. Studies by Osborne and Ehrlich (1976) reported that Alcaligenesfaecalis was able to oxidize As"' to As", although it was not clear if their organism derived energy from this process. The oxidation of reduced As species has been less widely studied than the processes of microbial reduction of As. Various bacteria, fungi, and algae organisms that are able to reduce As compounds have been identified. The reduction of AsV to As"' has been reported to be carried out by Pseudomonasjuorescens under anaerobic conditions, wine yeast, rumen bacteria, and cyanobacteria (Cullen and Reimer, 1989). Cheng and Focht ( 1979) have also identified that Pseudornonas and Alcaligenes were able to produce arsine gas directly from As" solutions in glucose and urea-enriched soils under anaerobic conditions. The ability of organisms to reduce inorganic As species directly has only been reported so far for bacteria. The bacterial methylation of inorganic As has been extensively studied in methanogenic bacteria. Methanogenic bacteria are a morphologically diverse group that produce methane as their primary metabolic end product under anaerobic conditions (Tamaki and Frankenberger, 1992). McBride and Wolfe (1971) reported that cell extracts of whole cells of the Methanobacterium strain MOH, growing anaerobically, reduced and methylated As" to dimethylarsine. Anaerobic biomethylation of As only proceeds to dimethylarsine, which is stable in the absence of 0, but is rapidly oxidized under aerobic conditions (Cullen and Reimer, 1989). The pathway of As" methylation

186

E. SMITH ETAL.

initially involves the reduction of AsVto As"', with the subsequent methylation of As"' to dimethylarsine by coenzyme M (Frankenberger and Losi, 1995). In addition to bacteria, several fungi species have the ability to reduce As species. It is well known that fungi and algae are able to methylate As. Toxic trimethylarsine gas is volatilized, and the liberalization of this gas by molds growing on wallpaper decorated with As pigments led to a number of poisoning incidents in the 1800s in England and Germany (Challenger, 1945). Cox and Alexander (1973a) reported that three fungal species, Candida humicola, Gliocladium roseurn, and Penicilliurn sp. were capable of transforming methylarsonic and dimethylarsonic acids to trimethylarsine. Further work by Cox and Alexander (1973b) showed that other anions, notably phosphate, inhibited the formation of trimethylarsine from inorganic As and methylarsonic acid. The extent to which microbial activity is involved in the transformation and movement of As in the soil is difficult to quantify. Woolson (1977) detected the generation of alkyarsines, notably dimethyarsine and trimethylarsine, in both anaerobic and aerobic conditions in the laboratory. Furthermore, the rate of formation of volatile compounds from the three As compounds applied to the soil (74As-sodium arsenate, I4C-MSMA, and 14C-cacodylicacid) was fastest under aerobic conditions, with the nature of the As compounds influencing the rate of formation. Hassler el al. (1984) suggested that, in addition to methylating As compounds, microbiological mobilization of As may occur under certain soil conditions. Woolson and Kearney (1973) found that the loss of 14C-cacodylicacid, applied to a range of soils and incubated over several months, was influenced by soil type, concentration of cacodylic acid applied, and soil moisture levels. Losses were attributed to the production of methylarsines and were greatest at the higher rates of cacodylic applied (100 mg As kg-I) in anaerobic conditions and in sandy soils. Few studies have investigated the long-term effects of microbial transformations of As species. In many instances, microbial influences on As sorption and desorption processes have been largely ignored in laboratory studies. Under shortterm laboratory studies and with previously uncontaminated soils, the microorganisms may have little influence on sorption processes in soils. However, there may be important implications for long-term studies, or where the soil microflora and microfauna have been predisposed to As when studying As sorption-desorption processes in laboratory situations.

M.CONCLUSIONS Arsenic is widely distributed in nature, with traces of As in the soil almost universal. The behavior of As in the soil is dominated by As speciation. In soil solutions the inorganic As species predominate, either as AsV or As"'. Under moderately oxidizing conditions (> + 100 mV) AsV will predominate, whereas under

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

187

moderately reducing conditions As111will predominate. The conversion of As from one oxidation state to another in soils is also affected by other soil parameters, including pH and microbial activity. The importance of microbial activity in controlling As species present in soil environments is well acknowledged, but the pathways and kinetics of the microbial processes are not well understood. An understanding of the physiochemical behavior of As in the soil is important for quantifying the persistence and bioavailability of As in the environment. The sorption of As onto soil surfaces plays an important role in mediating the availability of As in the environment. Iron, Al, and to a lesser extent Mn oxides are important soil constituents in controlling soil solution concentrations of As. Soil pH has a major influence on the availability of As. Arsenic is apolyprotic acid, and pH has a major influence on the valence charge of the As ion in soil solution and hence on the As adsorbed. In general, bioaccumulation of As to hazardous levels for human and animal consumption in the edible portions of plants seldom occurs because of the phytotoxic effects before such hazardous levels are reached. Plants accumulate the highest concentrations of As in plant roots. Unlike other elements such as P, As is not generally translocated to other parts of the plant. However, some plant species have been found to translocate As to a greater extent than others. Future research of As in soils is needed to understand factors controlling the nature of As species in the soil solution, as well as the role of microbes in controlling As speciation. Although adsorption processes have been well studied, further work is needed in understanding desorption processes and the factors that influence the kinetics of these processes. Scope also exists for studies on both plant and microbial uptake of As and the possible use of plants as low-cost, long-term means of remediation of As-contaminated sites.

ACKNOWLEDGMENTS We would like to thank Primary Industries for South Australia and the Co-operative Research Centre for Soil and Land Management for their support.

REFERENCES Adriano, D. C.. Page, A. L.. E1seewi.A. A., Chang. A. C.. and Straughan, I. (1980). Utilization and disposal of fly ash and other residues in terristrial ecosystems: Areview. J. Environ. Quol. 9,333-344. Ajwa, H. A,, and Tabatabai, M. A. (1995). Metal-induced sulfate adsorption by soils: I. Effect of pH and ionic strength. Soil Sci. 159,3242. Anastasia, E B., and Kender, W. J. (1973). The influence of soil arsenic on the growth of lowbeny bush. J. Environ. Quol. 2,335-337. Anderson, M. A,, and Malotky, D. T. (1979). The adsorption of protolizable anions on hydrous oxides at the isoelectric pH. J. Colloid Interface Sci. 72,413427.

188

E. SMITH ETAL.

Anderson, M. A., Ferguson, J. F., and Gavis. J. (1976). Arsenate adsorption on amorphous aluminium hydroxide. J. Colloid Inferface Sci. 54,391-399. Australian Bureau of Statistics. (1995). “Imports of arsenic by all countries and states.” Adelaide, Australia. Australian and New Zealand Environment and Conservation Council (ANZECC). (1992). “Australian and New Zealand Guidelines for the Assessment and Management of Contaminated Sites.”Aust. Govt. Pub. Service, Canberra. Bardgett, R. D., Speir, T. W., Ross, D. J., Yeates, G. W., and Kettles, H. A. (1994). Impact of pasture contamination by copper, chromium, and arsenic timber preservative on soil microbial properties and nematodes. Biol. Fen. Soils. 18,7 1-79. Barrow, N. J. (1974). On the displacement of adsorbed anions from soil: 2. Displacement of phosphate by arsenate. Soil Sci. 117,28-33. Barrow, N. J. (1983). A mechanistic model for describing the sorption and desorption of phosphate by soil. J. Soil Sci. 34,733-750. Barrow, N. J. (1984). Modelling the effects of pH on phosphate sorption by soils. J . Soil Sci. 35, 283-297. Barrow, J., Bowden, J. W., Posner. A. M., and Quirk, J. P. (1980). Describing the effects of electrolyte on adsorption of phosphate by a variable charge surface. Aust. J. Soil Res. 18,395-404. Barzi, F., Naidu, R., and McLaughlin, M. J. (1996). Contaminants and the Australian soil environment. In “Contaminants and the Australian Soil Environment in the Australasia-Pacific Region” (Naidu et al., eds.), pp. 45 1 4 8 4 . Kluwer Academic Publishers, Dordrecht. Belzile, N., and Tessier, A. (1990). Interactions between arsenic and iron oxyhydroxides in lacustrine sediments. Geochim. Cosmochim. Acfa 54, 103-109. Bencko, V., and Symon, K. (1977). Health aspects of burning coal with a high arsenic content. Environ. Res. 13,378-385. Beretka, J., and Nelson, P. (1994). The current state of utilisation of fly ash in Australia. In “Ash-A Valuable Resource,” Vol. I , pp. 51-63. South African Coal Ash Association. Bernard, D. W. (1983). Arsenic leachate from an abandoned smelter, Delero. Ontario, Canada. In “Papers of the International Conference on Groundwater and Man,” Vol. 2, pp. 29-41. Aust. Govt. Pub. Service, Canberra. Bhumbla, D. K., and Keefer. R. F. (1994). Arsenic mobilization and bioavailability in soils. In “Arsenic in the Environment, Part I: Cycling and Characterization.” (J. 0. Nriagu, ed.), pp. 51-82. Wiley & Sons, New York. Bisessar, S. (1982). Effects of heavy metals on microorganisms in soils near a secondary lead smelter. Waf.Air Soil Poll. 17,305-308. Bishop, R. F., and Chisholm, D. (1961). Arsenic accumulation in Annapolis Valley orchards. Can. J. Soil Sci. 42,77-80. Bohn, H. L. (1976). Arsenic Eh-pH diagram and comparisons to soil chemistry of phosphorus. Soil Sci. 121, 125-127. Bolan, N. S., Syers, J. K., and Tillman, R. W. (1986). Ionic strength effects on surface charge and adsorption of phosphate and sulphate by soils. J . Soil Sci. 37,379-388. Braman, R. S., and Foreback, C. C. (1973). Methylated forms of arsenic in the environment. Science 182,1247-1 249. Brannon, J. M., and Patrick, Jr., W. H. (1987). Fixation, transformation, and mobilization of arsenic in sediments. Environ. Sci. Technol. 21,450459. Carbonell Barrachiina, A., Burl6 Carbonell, F., and Mataix Beneyto, J. (1996). Kinetics of arsenite desorption in Spanish soils. Commun. Soil Sci. Plant Anal. 27, 3101-31 17. Challenger, F. (1945). Biological methylation. Chem. Rev. 36,315-361. Cired in Hassler, R.A,, Klein, D. A., and Meglen, R. R. (1984). Microbial contributions to soluble and volatile arsenic dynamics in retorted oil shale. J. Environ. Qual. 13,466470,

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

189

Charter, R. A., Tdbatabai, M. A., and Schafer, J. W. (1995).Arsenic, molybdenum, selenium, and tungsten contents of fertilizers and phosphate rocks. Commun. Soil Sci. Planr Anal. 26, 30513062. Chen, Jen Chen, and Lin. Li-Ju. (1994). Human carcinogenicity and atherogenicity induced by chronic exposure to inorganic arsenic. In “Arsenic in the Environment: Part 11. Human Health and Ecosystems Effects.” (J. 0. Nriagu, ed.), pp. 109-157. Wiley & Sons, New York. Cheng, C. N., and Focht. D. D. (1979). Production of arsine and methylarsines in soil and in culture. Appl. Environ. Microbiol. 38,494-498. Chilvers, D. C.. and Peterson, P. J. (1987). Global cycling of arsenic. In “Lead, Mercury, Cadmium, and Arsenic in the Environment” (T. C. Hutchinson and K. M. Meema. eds.), pp. 279-301. Wiley & Sons, New York. COX.D. P.. and Alexander, M. ( 1973a). Production of trimethylarsine gas from various arsenic compounds by three sewage fungi. Bull. Environ. Contam. Toxicol. 9 , 8 6 8 8 . COX,D. P., and Alexander, M. (1973b). Effect of phosphate and other anions on trimethylarsine formation by Candida humicola. Appl. Microbiol. 25,408-413. Crecelius, E. A., Johnson, C. J., and Hoffer, G. C. (1974).Contamination of soils near a copper smelter by arsenic, antimony, and lead. Wat. Air Soil Poll. 3,337-342. Cullen, W. R., and Reimer, K. J. (1989). Arsenic speciation in the environment. Chem. Rev. 89, 7 13-764. Curtin, D., Syers, J. K., and Bolan, N. S. (1992). Phosphate sorption by soil in relation to exchangeable cation composition and pH. Ausr. J. Soil Rex 31, 137-149. Das, D., Samanta, G., Mandal, B. K.. Chowdhury, T. R., Chanda, C. R., Chowdhury, P. R., Basu, G. K.,and Chakraborti, D. (1996). Arsenic in groundwater in six districts of West Bengal, India. Environ. Geochem. and Health. 18,5-15. Davenport, I. R., and Peryea, F. J. (1991). Phosphate fertilizers influence leaching of lead and arsenic in soil contaminated with lead arsenate. War. Air Soil Poll. 57-58, 101-110. Department of Mines. Perth, Western Australia. (1980). “Mineral Resources of Western Australia,” p. 21. Dept. of Mines, Perth, Western Australia. Deuel, L. E., and Swoboda, A. R. (1972). Arsenic solubility in a reduced environment. Soil Sci. Am. Proc. 36,276-278. DiGiano. F. A.. Baldauf, G., Frick, B., and Sontheimer, H. (1978). A simplified competitive equilibrium adsorption model. Chem. Eng. Sci. 33,1667-1673. DIPMAC Report. (1992). “Report on the Management of Contaminated Waste at the Cattle Dip Sites in North Eastern New South Wales.” NSW Gov.. Sydney, Australia. Doelman, P., and Haanstra, L. (1984). Short-term and long-term effects of cadmium, chromium, copper, nickel, lead, and zinc on soil microbial respiration in relation to abiotic soil facetors. Plant and Soil 79,3 17-327. Dowdy, R. H., and Volk, V. V. (1984). Movement of heavy metals in soils. In Chemical Mobility and Reactivity in Soil Systems (0.W. Nelson et al., eds.), pp. 229-240. SSSA Special Publication No. 1 1. ASA and SSSA, Madison, WI. Dudas, M. J. (1987).Accumulation of native arsenic in acid sulphate soils in Alberta. Can. J. Soil Sci. 67,317-331. Dudas, M. J., and Pawluk, S. (1980). Natural abundances and mineralogical partitioning of trace elements in selected Alberta soils. Can. J. Soil Sci. 60,763-771. Ehrlich, H. L. (1996). Geomicrobiology, 3rd ed., pp. 276-286. Marcel Dekker, New York. Elkhatib, E. A,, Bennett, 0. L.. and Wright, R. J. (1984). Arsenite sorption and desorption in soils. Soil Sci. Soc. Am. J. 48, 1025-1 030. Fendorf, S., Eick, M. J., Grossl, P., and Sparks, D. L. (1997). Arsenate and chromate retention mechanisms on goethite. I. Surface structure. Environ. Sci. Technol. 31,315-320. Fergus, I. F. (1955). A note on toxicity in some Queensland soils. Queensland J. Ag. Sci. 12,95-100.

190

E. SMITH ETAL.

Fordham, A. W., and Norrish, K. ( 1979).Arsenate-73 uptake by components of several acidic soils and its implications for phosphate retention. Aust. J. Soil Res. 17,307-316. Fordham, A. W., and Norrish, K. (1983). The nature of soil particles particularly those reacting with arsenate in a series of chemically treated samples. Ausr. J. Soil Rex 21,455477. Mass, M. J., McKinney, J. D., North, D. W., Ohanian, E. V., and Fowle 111, J. R., Abemathy, C. 0.. Uthus, E. (1991). Arsenic research needs. Trace Subst. Environ. Healrh 25,257-271. Fowler, B. A. (1977). Toxicology of Environmental Arsenic. In “Toxicology of Trace Elements.” (R. A. Coyer and M. A. Mehlman, eds.), pp. 79-122. Wiley & Sons, New York. Frank, R., Braun, H. E., Ishida, K., and Suda, P. (1976). Persistent organic and inorganic pesticide residues in orchard soils and vineyards of Southern Ontario. Can. J. Soil Sci. 56, 463484. Frankenberger, W. T., and Losi, M. E. (1995). Applications of bioremediation in the cleanup of heavy metals and metalloids. In “Bioremediation,” (H. D. Skipper and R. F. Turco, eds.), pp. 173-210. SSSA Special Publication 43 Soil Sci. SOC.Am ., Madison, WI. Freeman, J. S., and Rowell, D. L. (1981). The adsorption and precipitation of phosphate onto calcite. J. Soil Sci. 3 2 , 7 5 4 4 . Frost, R. R., and Griffin, R. A. (1977). Effect of pH on adsorption of arsenic and selenium from landfill leachate by clay minerals. Soil Sci. SOC.Am. J. 41,53-57. Garcia-Rodeja, I., and Gil-Sortes, F. (1995). Laboratory study of phosphate desorption kinetics in soils of Galicia (N.W. Spain). Commun.Soil Sci. Plant Anal. 26,2023-2040. Gilmor, J. T., and Wells, B. R. (1980). Residual effects of MSMA on sterility in rice cultivars. Agron. J. 72, 10661067. Goldberg, S. (1986). Chemical modelling of arsenate adsorption on aluminium and iron oxide minerals. Soil Sci. SOC.Am. J . 50, 1154-1 157. Goldberg, S. (1992). Use of surface complexation models in soil chemical systems. Adv. Agron. 47, 233-330. Goldberg, S., and Glaubig, R. A. (1988). Anion sorption on a calcareous montmorillonitic soil-Arsenic. Soil Sci. Am. J. 52, 1297-1300. Goodroad, L. L., and Caldwell, A. C. (1979). Effects of phosphorus fertilizer and lime on the As. Cr, Pb, and V content of soils and plants. J. Environ. Quul. 8,493496. Green, H. H. (1918). Rep. Ver. Res. S.Afr 516 (1918), 595 Cired in Cullen, W. R., and Reimer, K. J. (1989). Arsenic speciation in the environment. Chem. Rev. 89,713-764. Grossl, P. R.,Eick, M. Sparks, D. L., Goldberg, S., and Ainsworth, C. C. (1997). Arsenate and chromate retention mechanisms on goethite. 2. Kinetic evaluation using a pressure-jump relaxation technique. Environ. Sci. Technol. 31,32 1-326. Gustafsson, J. P., and Jacks. G. (1995). Arsenic geochemistry in forest soil profiles as revealed by solid-phase studies. Appl. Geochem. 10,307-3 15. Harrison, J. B., and Berkheiser, V. E. (1982). Anion interactions with freshly prepared hydrous iron oxides. Cluys Clay Mine,: 30,97-102. Harter, R. D., and Naidu, R. (1995). Role of metal-organic complexation in metal sorption by soils. Adv. Agmn. 55,219-264. Hassler, R. A., Klein, D. A., and Meglen, R. R. (1984). Microbial contributions to soluble and volatile arsenic dynamics in retorted oil shale. J. Environ. Qual. 13,466-470. Hess, R. E., and Blanchar. R. W. (1976). Arsenic stability in contaminated soils. Soil Sci.SOC.Am. J. 40,847-852. Heylar, K. R., Munns, D. N., and Burau, R. G. (1976). Adsorption of phosphate by gibbsite. II. Formation of a surface complex involving divalent cations. J. Soil Sci. 27,35 1-323. Hillier, G. (1980). Arsenic. In “Australian Mineral Industry Annual Review for 1978,” p. 50. Aust. Govt. Publishing Service, Canberra. Hiltbold, A. E., Hajek, B. F.. and Buchanan, G. A. (1974). Distribution of arsenic in soil profiles after repeated applications of MSMA. Weed Sci. 22,272-275.

ARSENIC IN THE SOIL ENVIRONMENT:A REVIEW

191

Hingston, F. J., Atkinson, R. J., Posner. A. M., and Quick, J. P. (1967). Specific adsorption of anions. Narure 215, 1459-1461. Hingston, F. J., Posner. A. M., and Quirk, J. P. (1971). Competitive adsorption of negatively charged ligands on oxide surfaces. Discuss. Faraday Soc. 52,334-342. Hingston. F. J., Posner, A. M., and Quirk, J. P. (1972).Anion adsorption by goethite and gibbsite. I: The role of the proton in determining adsorption envelopes. J. Soil Sci. 23, 177-191. Jacobs, L. W., Syere, J. K., and Keeney. D.R. (1970).Arsenic sorption by soils. Soil Sci. SOC.Am. Proc. 34,750-754. Jiang, Q. Q.. and Singh, B. R. (1994).Effect of different forms and sources of arsenic on crop yield and arsenic concentration. War. Air Soil Poll. 74,32 1-343. Johnson, L. R., and Hiltbold, A. E. (1969). Arsenic content of soil and crops following use of methanearsonate herbicides. Soil Sci. Soc. Am. Proc. 33,279-282. Kabata-Pendias, A., and Pendias, H. (1992). “Trace Elements in Soils and Plants,” 2nd ed. CRC Press, Boca Raton, FL. Kiss, A. M., Oncsik, M., Dombovari, J. Veres. S., and Acs, G. (1992). Dangers of arsenic drinking and irrigation water to plants and humans. Antagonism of arsenic and magnesium. Acra Agron. Hung. 41,3-9. Leblanc, M., Achard, B., Ben Othman. D.,and Luck, J. M. ( I 996). Accumuulation of arsenic fron acidic mine waters by ferruginous bacterial accretions (stromatolites).Appl. Geochem. 11,541-554. Leonard, A. (1991).Arsenic. In “Metals and Their Compounds in the Environment. Occurrence, Analysis, and Biological Relevance,” 2nd ed. (E. Merian, in cooperation with Clarkson et al., eds.), pp. 751-773. Weinheim, VCH. Li, X., and Thornton, I. (1993).Arsenic, antimony, and bismuth in soil and pasture herbage in some old metalliferous mining areas in England. Environ. Geochem. Health 15, 135-144. Livesey, N. T., and Huang, P. M. (1981).Adsorption of arsenate by soils and its relation to selected chemical properties and anions. Soil Sci. 131,88-94. Loebenstein, J. R. (1993).Arsenic. In “Minerals Yearbook 1990,” pp. 167-170. US.Govt. Printing Oftice, Washington, D.C. Lookman, R., Freese, D.. Merckx, R., Vlassak, K., and Van Riemsdijk, W. H. (1995). Long-term kinetics of phosphate release from soil. Environ. Sci. Technol. 29, 1569-1575. Lu, F, J. (1990). Blackfoot disease: Arsenic or humic acid? Lancer 336, 115-1 16. Lumsdon, D. G., Fraser, A. R., Russell, J. D., and Livesey, N. T. (1984). New infrared band assignments for the arsenate ion adsorbed on synthetic goethite (a-FeOOH). J. Soil Sci. 35, 381-386. Lund, U., and Fobian, A. (1991). Pollution of two soils by arsenic, chromium, and copper, Denmark. Geodema 49,83-103. MacLean, K. S., and Langille, W. M. (1981).Arsenic in orchard and potato soils and plant tissue. Plant and Soil 61,4 1 3 4 1 8. Maliszewska, W., Dec, S.. Wierzbicka, H., and Wozniakowska, A. (1985). The influence of various heavy metal compounds on the development and activity of soil micro-organisms. Environ. Polhi.9,A. :37, 195-2 15. Malone, C. R . (1971). Responses of soil microorganisms to non-selective vegetation control in a fescue meadow. Soil B i d . Biochem. 3, 127-1 3 I . Manful, G. A,, Verloo, M., and De Spiegeleer, F. (1989). Arsenate sorption by soils in relation to pH and selected anions. Pedologie 39,5548. Manning, B. A,, and Goldberg, S. (1996). Modeling competitive adsorption of arsenate with phosphate and molybdate on oxide minerals. Soil Sci, Soc. Am. J. 60,121-131. Marin, A. R., Masscheleyn, P. H., and Patrick Jr., W. H. (1992). The influence of chemical form and concentration of arsenic on rice growth and tissue arsenic concentration. Plant and Soil 139, 175-183. Marin, A. R., Masscheleyn, P. H., and Patrick Jr., W. H. (1993). Soil redox-pH stability of arsenic specis and its influence on arsenic uptake by rice. Plant and Soil 152,245-253.

192

E. SMITH ETAL.

Masscheleyn, P. H., Delaune, R. D., and Patrick Jr., W. H. (1991). Effect of redox potential and pH on arsenic speciation and solubuility in a contaminated soil. Environ. Sci. Technol. 25, 1414-1418. McBride, B. C., and Wolfe, R. S. (1971). Biosynthesis of dimethylarsine by Methnnobacrerium. Biochem. 10,43124317. McLaren, R. G.. Carey, P. L., Cameron, K. C., Adams, J. A,, and Sedcole. J. R. (1994). Effect of soil properties and contact period on the leaching of copper, chromium, and arsenic through undisturbed soils. In “15th World Cong. Soil Sci.,” Vol. 3a, pp. 156-169. Intl. Soc. Soil Sci. and The Mexican Soc. of Soil Sci., Acapulco, Mexico. McMillan, M. G. (1988). Weed control in cotton: Agdex 151/640. NSW Agriculture and Fisheries. Melamed, R., Jurinak, J. J., and Dudley, L. M. (1995). Effect of adsorbed phosphate on transport of arsenate through an oxisol. Soil Sci. Am. J. 59, 1289-1294. Merry, R. H., Tiller, K. G., and Alston, A. M. (1983).Accumulation of copper, lead, and arsenic in some Australian orchard soils. Aust. J. Soil Res. 21,549-561. Merry, R. H., Tiller, K. G., and Alston, A. M. (1986). The effects of contamination of soil with copper, lead, and arsenic on the growth and composition of plants. I. Effects of season, genotype, soil temperature, and fertilizers. Plant and Soil 91, 115-128. Naidu. R., and Rengasamy, P. (1993). Ion interactions and constraints to plant nutrition in Australian sodic soils. Aust. J. Soil Res., 31,801-819. Naidu, R., Syers, J. K., Tillman, R. W., and Kirkman, J. H. (1990). Effect of liming and added phosphate on charge characteristics of acid soils. J. Soil Sci. 41, 157-164. Naidu, R., Merry, R. H., Churchman, G. J., Wright, M. J., Murray, R. S., Fitzpatrick, R. W., and Zarcinas, B. A. (1993). Sodicity in South Australia-A review. Aust. J . Soil Res. 31,911-929. Naidu. R., Bolan, N. S., Kookana, R. S., and Tiller, K. G. (1994). Ionic-strength and pH effects on the sorption of cadmium and the surface charge of soils. Eur J. Soil Sci. 45,419429. Naidu, R., Smith, J., Smith, L. H., Tiller, K. G., and McDougall, K. W. (1995). ArsenicDDT residues at cattle tick contaminated sites: Preliminary investigation. CSIRO Division of Soils Technical Report No. 60. Adelaide, Australia. National Academy of Sciences. ( I 977). “Medical and Biologic Effects of Environmental Pollutants: Arsenic.” National Research Council, Washington, D.C. Cited in Leonard, A. (1991).Arsenic. In “Metals and Their Compounds in the Environment. Occurrence, Analysis. and Biological Relevance” 2nd ed. (E. Merian, in cooperation with Clarkson er al., eds.), pp. 751-773. Weinheim, VCH. National Food Authority. (1993). “Australian Food Standards Code: March, 1993.” Australian Govt. Pub. Service, Canberra. National Research Council of Canada. (1978). “Effects of Arsenic in the Canadian Environment.” NRCC no. I539 1. Ottawa, Canada. Natush, D. F. S., Bauer, C. F.. Matusiewicz, H., Evans, C. A., Baker, J., Loh, A., Linton, R. W., and Hopke, P. K. (1975). Characterization of trace metals in fly ash. In “International Conference on Heavy Metals in the Environment.” Vol. 11. pp. 553-575. Toronto. Nriagu, J. 0. (1994).Arsenic: Historical perspectives. I n “Arsenic in the Environment: Part 1. Cycling and Characterization” (J. 0. Nriagu, ed.), pp. 3-15. Wiley & Sons, New York. Nriagu, J. 0..and Pacyna, J. M. (1988). Quantitive assessment of worldwide contamination of air, water, and soils by trace metals. Nature 333, 134-139. O’Neill, P. (1990).Arsenic. In “Heavy Metals in Soils” (B. J. Alloway, ed.), pp. 83-99. Blackie. London. O’Neill, P. (1995). Arsenic. In “Heavy Metals in Soils,’’ 2nd ed. (B. J. Alloway, ed.), pp. 105-121. Blackie, London. Onken, B. M., and Hossner, L. R. (1995). Plant uptake and determination of arsenic species in soil solution under flooded conditions. J. Environ. Quaf. 24,373-38 l . Osborne, F.H., and Ehrlich, H. L. (1976). Oxidation of arsenite by a soil isolate of alcaligenes. J . Appl. Bacreriol. 41.295-305.

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

193

Oscarson. D. W., Huang, P. M., Defosse, C., and Herbillon. (1981). Oxidative power of Mn(IV) and Fe(II1) oxides with respect to As(1II) in terristrial and aquatic environments. Nature 291, 50-51. Oscatson, D. W.. Huang, P. M., Liaw, W. K., and Hammer, U. T. (1983a). Kinetics of oxidation of arsenite by various manganese dioxides. Soil Sci. SOC.Am. J. 46,644-648. Oscarson, D. W., Huang. P. M., and Hammer, U. T. (1983b).Oxidation and sorption of arsenite by manganese dioxide as influenced by surface coatings of iron and aluminium oxides and calcium carbonate. War.Air Soil Poll. 20,233-244. Peryea. F. J. (1991). Phosphate induced release of arsenic from soils contaminated with lead arsenate. Soil Sci. SOC.Am. J. 55, 1301-1306. Peryea, F. J., and Creger, T. L. (1994). Vertical distribution of lead and arsenic in soils contaminated with lead arsenate pesticide residues. War. Air Soil Poll. 78, 297-306. Peterson, P. J., Benson, L. M.,and Porter, E. K. (1979). Biogeochemistry of arsenic on polluted sites in SW England. In “Heavy Metals in the Environment, London, September 1979,”pp. 198-201, CEP Consultants Ltd., Edinburgh. Petito, C. T., and Beck, B. D. (1990). Evaluation of evidence of nonlinearities in the dose-response curve for arsenic carcinogenesis. Trace Subsr. Environ. Health 24, 143-176. Pierce, M. L., and Moore, C. B. (1980). Adsorption of arsenite and arsenate on amorphous iron hydroxide from dilute aqueous solutions. Environ. Sci. Technol. 14,214-216. Pierzynski, G. M., Schnoor, J. L., Banks, M. K., Tracy, J. C., Licht, L. A., and Erickson, L. E. (1994). Vegetative remediation at superfund sites. In “Mining and its Environmental Impact” (R.E. Hester and R. M.Harrison, eds.), pp. 49-69. Royal SOC.of Chem., Cambridge. Pignatello, J. I.,and Xing, B. (1996). Mechanisms of slow sorption of organic chemicals to natural particles. Environ. Sci. Tech. 30, 1-11. Ragaini, R. C.. Ralston, H.R., and Robert, N. (1977). Environmental trace metal contamination in Kellog, Idaho, near a lead smelting complex. Environ. Sci. Tech. 11,773-781. Rimstidt, J. D., Chermak, J. A,, and Gagen, P. M. (1994). Rates of reaction of galena, sphalerite, chalcopyrite, and arsenopyrite with Fe(1n) in acidic solutions. In “Environmental Geochemistry of Sulfide Oxidation” (C. N. Alpers and D. W. Blowes, eds.), pp. 2-13. Am. Chem. Soc., Washington, D.C. Rittle, K., Dreiver, J., and Colberg, P. (1995). Precipitation of arsenic during bacterial sulfate reduction. Geomicrobiol. J. 13, 1-1 1 . Roy, W. R., Hassett, J. J., and Griffin, R. A. (1986).Competitive interactions of phosphate and molybdate on arsenate adsorption. Soil Sci. 142,203-210. Sachs, R. M., and Michael, J. L. (1971).Comparative phytotoxicity among four arsenical herbicides. Weed Sci. 19,558-564. Sadiq, M. (1986). Solubility relationships of arsenic in calcareous soils and its uptake by corn. Plant and Soil 91,241-248. Sadiq, M., Zaidi, T. H.. and Mian, A. A. (1983).Environmental behaviour of arsenic in soils: Theoretical. Wat. Air Soil Poll. 20,369-377. Sadler. R. Olszowy, H.. Shaw, G.. Biltoft, R., and Connell, D. (1994). Soil and water contamination by arsenic from tannery waste. War. Air Soil Poll. 78, 189-198. Schuthess, C. P., and Huang, C. P. (1990). Adsorption of heavy metals by silicon and aluminium surfaces on clay minerals. Soil Sci. SOC.Am. J . 54,679-688. Scott, M. J., and Morgan, J. J. (1995). Reactions of oxide surfaces. 1. Oxidation of As(II1) by synthetic birnessite. Environ. Sci. Technol. 29, 1898-1905. Sheppard, S. C. (1992).Summary of phytotoxic levels of soil arsenic. War. AirSoil Poll. 64,539-550. Shiendorf, C., Rebhun, M., and Sheintuch, M. (1981). A Freundlich-type multicomponent isotherm. J. Colloid Interface Sci. 79, 136-142. Shuman, L. M. (1976). Zinc adsorption isotherms for soil clays with and without iron oxides removed. Soil Sci. SOC.Am. J. 40,349-352. Speir. T. W., August, J. A., and Feltham, C. W. (1992b). Assessment of the feasibility of using CCA

194

E. SMITH ETAL.

(copper, chromium, and arsenic)-treated and boric acid-treated sawdust as soil admendments. 11. Soil biochemical and biological properties. Plant and Soil 142,249-258. Steevens, D. R., Walsh. L. M., and Keeney, D. R. (1972). Arsenic phytotoxicity on a Plainfield sand as affected by ferric sulfate or aluminium sulfate. J . Environ. Qual. 1,301-303. Tabatabdi, M. A. (1977). Effects of trace elements on urease activity in soils. Soil Biol. Biochem. 9, 9-13. Takamatsu. T., Aoki, H., and Yoshida, T. (1982). Determination of arsenate, arsenite, monomethylarsonate, and dimethylarsinate in soil polluted with arsenic. Soil Sci. 133,239-246. Tamaki, S., and Frankenberger Jr., W. T. (1992). Environmental Biochemistry of Arsenic. In “Reviews of Environmental Contamination and Toxicology,” Vol. 124, pp. 79-1 10. Springer-Verlag. New York. Tammes, P. M., and de Lint, M. M. (1969). Leaching of arsenic from soil. Neth. J. Agric. Sci. 17, 128-132. Thanabalasingam, P., and Pickering, W. F. (1986). Arsenic sorption by humic acids. Environ. Pollur. Ser: B 12,233-246. Thomas. R. (1977). Arsenic pollution arising from mining activities in south-west England. In “Inorganic Pollution and Agriculture, April 1977,” pp. 126-141. Ministry of Agriculture, Fisheries and Food, London. Tiller, K. G. (1992). Urban soil contamination in Australia. A m . J . Soil Res. 30,937-957. Turner, R. R. (1981). Oxidation state of arsenic in coal ash leachate. Environ. Sci. Technol. 15, 1062- 1066. Vaughan, G . T. (1993). Investigation Report CETLHIR148: The environmental chemistry and fate of arsenical pesticides in cattle tick dip sites and banana plantations. CSIRO, Division of Coal and Energy Technology, Centre for Advanced Analytical Chemistry, Sydney, Australia. Walsh, L. M., and Keeney, D. R. (1975). Behaviour and Phytotoxicity of Inorganic Arsenicals in Soils. In “Arsenical Pesticides” (E. A. Woolson, ed.), pp. 35-53. Am. Chem. SOC.,Washington, D.C. Walsh, L. M., Sumner, M. E., and Keeney, D. R. (1977). Occurrence and distribution of arsenic in soils and plants. Environ. Healrh Perspect. 19,67-71. Wauchope, R. D. (1975). Fixation of arsenical herbicides, phosphate, and arsenate in alluvial soils. J. Environ. Qual. 4, 355-358. Waychunas, G . A., Rea, B. A., Fuller, C. C., and Davis, J. A. (1993). Surface chemistry of ferrihydrite: Part 1. EXAFS studies of the geometry of coprecipitated and adsorbed arsenate. Geochim. et Cosmochim. Acta 51,225 1-2269. Waychunas, G . A., Fuller, C. C., Rea, B. A., and Davis, J. A. (1996). Wide angle X-ray scattering (WAXS) study of “two-line” ferrihydrite structure: Effect of arsenate sorption and counterion variation and comparison with EXAFS results. Geochim. er Cosmochim. Acta 60,1765-1781. Wells, B. R., and Gilmor, J. T. (1977). Sterility in rice cultivars as influenced by MSMA rate and water management. Agron. J. 69,45 1-454. Wild, A. (1993). Soils and the Environment: An Introduction. Cambridge University Press, Cambridge. Wild, H. ( I 973-74). Arsenic tolerant plants species established on arsenical mine dumps in Rhodesia. Kirkia 9,265-277. Woolson, E. A. (1973).Arsenic phytotoxicity and uptake in six vegetable crops. WeedSci. 21,524-527. Woolson, E. A. (1977). Generation of alkylarsines from soil. Weed Sci. 25,412416. Woolson, E. A., and Kearney, P. C. (1973). Persistence and reactions of 14C-cacodylicacid in soils. Environ. Sci. Technol. 1,47-50. Woolson, E. A., Axely, J. H., and Kearney, P. C. (1971a). Correlation methods between available soil arsenic, estimated by six methods, and response to corn (Zm mays L.). Soil Sci. Am . Proc. 35, 101-1 05. Woolson, E. A., Axely, J. H., and Kearney, P. C. (1971b). The chemistry and phytotoxicity of arsenic in soils. 1. Contaminated field soils. Soil Sci. Am. Proc. 35,938-943.

ARSENIC IN THE SOIL ENVIRONMENT: A REVIEW

195

Woolson, E. A., Axely, J. H., and Kearney, P. C. (1973).The chemistry and phytotoxicity of arsenic in soils: Effects of time and phosphorus. Soil Sci. Am. Proc. 37,254-259. World Health Organization. (198 I). “Environmental Health Criteria 18; Arsenic.” World Health Organization, Geneva. Xu. H., Allard, B., and Grimvall, A. (1988). Influence of pH and organic substance on the adsorption of As(V) on geologic materials. Waf. Air Soil Poll. 40,293-305. Xu, H., Allard, B.. and Grimvall, A. (1991). Effects of acidification and natural organic materials on the mobility of arsenic in the environment. Waf.Air Soil Poll. 57-58,269-278. Ydn-Chu, H. (1994).Arsenic distribution in soils. In “Arsenic in the Environment, Part 1: Cycling and Characterization” (J. 0. Nriagu, ed.), pp. 17-49. Wiley & Sons, New York.

This Page Intentionally Left Blank

DRYLAND CROPPING A FUNDAMENTAL INTENSIFICATION: SOLUTIONTO EFFICIENT USE OF PRECIPITATION H. J. Farahani,'* G. A. Peterson,2 D. G. Westfal12 'USDA-Agricultural Research Service Great Plains Systems Research Fort Collins, Colorado 80521 2Depamnent of Soil and Crop Sciences Colorado State University Fort Collins, Colorado 80523

I. Introduction A. The Paradox of Summer Fallow B. Objective 11. Summer Fallow: A Second Look A. Spring Wheat-Fallow System B. Winter Wheat-Fallow System 111. Dryland Cropping Intensification A. Modern Dryland-No-Till Cropping Systems B. Summer Cropping: Key to Efficient Use of Precipitation Tv. A Systems Approach to Intensification A. Systems Analysis B. Systems Evaluation: Quantitative Indices V. Conclusion References

I. INTRODUCTION Pioneering farmers in the 19th century, frustrated with the erratic yields associated with annual cropping of small grains, began alternating crop with fallow to *Present Address: Department of Soil and Crop Sciences, Colorado State University, Fort Collins, Colorado 80523. 197 Advances in A p n w q . Volunir 61 Copyright 0 1998 by Academic Press. All righrs of reproduction in any form reserved 0065-21 13/98 SZS.00

198

H. J. FARAHANI ETAL.

improve yields (on a per harvest basis) and reduce total crop failure and labor. A dryland farming practice known as “summer fallow” soon dominated the North American Great Plains in regions that receive annual precipitation of less than 500 mm. With this practice, no crop is grown during the fallow, and weeds are controlled by cultivation or chemicals to enhance soil-water storage and nutrient availability for the subsequent crop. Both winter and spring wheat-fallow systems are practiced in the Great Plains. In the central and southern Great Plains, hard red winter wheat (triticum aestivum L.) is the dominant dryland crop primarily because of its high-yielding potential (Greb et al., 1979) with limited crop substitutions (Johnson, 1977). In the northem plains, hard red spring wheat is dominant. For winter wheat, the fallow period is approximately 14 months, running from harvest in July to planting in September of the next year. The fallow period for spring wheat is about 21 months, extending from early August harvest to planting in the second spring. To conserve soil water during fallow, in the early days of dryland farming, weeds were controlled by multiple tillage (plow, harrow, one-way disk) operations. As the acreage of clean-fallowed land increased, the hazards of water and wind erosion multiplied, resulting in the Dust Bowl in the 1930s. Scientists and farmers in the plains then turned to stubble-mulch (a V-shaped sweep or blade pulled at shallow depth) or subsurface tillage to control erosion (Duley and Russel, 1939). Stubble-mulch tillage is currently the dominant method of summer fallow in the plains. It is well documented that fallow increases the probability of having adequate soil water at planting to maximize initial wheat stand establishment and development, and therefore we do not dwell on this. In this chapter we further investigate “the paradox of summer fallow” first noted by Haas et al. ( 1974).

A. THEPARADOX OF SUMMER FALLOW Most farmers in the Great Plains agree that water is the primary limiting factor controlling dryland production. Yet only a small portion of the precipitation received is stored during fallow, and soil evaporation far exceeds other losses by weeds, volunteer plants, runoff, deep seepage, and snow blowoff. In a classic USDA Conservation Research Report, Haas et al. (1974) state, “It seems paradoxical that water should be proclaimed the primary factor limiting crop production in the northern Great Plains, when more than 1 year’s precipitation is lost during the fallow period for spring wheat.” For no-till winter wheat-fallow in the west-central Great Plains, Farahani et al. (1998) found that on average, only 20% of the precipitation received during the fallow was stored in the soil profile. For the region, average (1948-1995) precipitation for the 14-month fallow is 552 mm, resulting in 442 mm of lost precipitation. That is indeed more than an average year’s precipitation of 410 mm.

DRYLAND CROPPING INTENSIFICATION

199

Mathews and Army ( 1 960) summarized soil-water and precipitation data for 25 stations representing over 450 wheat-fallow years on well-managed fallow lands in the Great Plains. The average soil-water storage during the fallow (for both winter and spring wheat-fallow systems) was 100 mm or 16% of the precipitation (617 mm), corresponding to a 84% loss of precipitation. They attributed this loss to evaporation from the soil, since runoff and deep percolation losses were known to be very low. Although significant progress in fallow tillage and management has been made since then, investigators still report unacceptably low fallow water storage efficiencies, even under modem conservation practices of reduced- and no-till (Unger, 1984; Stewart and Steiner, 1990; Norwood, 1994; Jones and Popham, 1997; McGee ef al., 1997). In a recent review, Peterson et al. (1996) examined the effects of tillage and residue management on fallow soil-water storage from Canada to Texas. Water storage efficiencies using no-till summer fallow in the Great Plains were reported as 10% in Texas, 22% in eastern Colorado, and 25-30% in western Kansas for the 14-month winter wheat-fallow system; and from 18 to 37% in the northern plains for the 21-month fallow of spring wheat. From their summary, an average efficiency of 25% was found for water storage during fallow (both winter and spring wheat) in the Great Plains. Comparing this with the earlier findings of Mathews and Army ( 1 960), one may conclude that from the dust mulch days in the early 1900s to the present era, fallow efficiency has only improved from 16% storage to 25% storage with no-till fallow. A huge loss, 75% of the fallow precipitation, still remains a reality, even with our best known soil and water conservation practices. Summer rainfall prevails in the Great Plains, with nearly 75% of the annual precipitation occurring from April to September. Ironically, precipitation-storage efficiency during fallow is lowest, even negative at times, during summer periods when precipitation is greatest. Paradoxically,fallow is not only inefficient but most inefficientduring the periods when precipitation is most substantial (i.e., summer). There appears to be little possibility of further reducing evaporation by use of surface residue, particularly since residue production in Great Plains dryland agriculture is limited for efficient water storage (Peterson et al., 1996). Existing soil and water conservation practices, very important to erosion and soil productivity, are at or near their practical limits. A different approach to water conservation and efficient use of precipitation is obviously needed. Enhancing the efficient use of precipitation is the primary key to a sustainable dryland agriculture (Peterson et al., 1996). It appears that the most direct and practical solution to improving efficient use of precipitation is the inclusion of a summer crop (i.e., corn [Zea mays L.], sorghum [Sorghumbicolor L. Moench], millet [Panicurn miliaceum L.], or sunflower [Helianrhus annuus L.]) in the year following the wheat crop that would utilize the summer precipitation. Peterson and Westfall (1 996) stated, “Planting a spring crop that can utilize both the stored water and the summer precipitation is the key; . . . the summer precipitation is used

H. J. FARAHANI ETAL.

by the crop instead of being lost to evaporation during the second summer of fallow.” The 2-year wheat-fallow system is replaced by a 3-year wheat-corn (-sorghum, millet, or -sunflower) fallow rotation. The former produces one crop every 2 years; or a 0.5 cropping (and 0.5 summer fallow) intensity per year. In the 3-year system, cropping intensity increases to 0.67 (two crops every 3 years), and summer fallow intensity decreases to 0.33 (one summer fallow every 3 years). The term “cropping intensification” is used as an umbrella term, defining dryland systems with more crops and less summer fallow per unit time. In the Great Plains, dryland-cropping intensification has shown pronounced increases in annualized grain yield and biomass production (Peterson et al., 1993, 1996; Halvorson et al., 1994; Norwood, 1994; Jones and Popham, 1997). Even soil-surface organic matter has increased in some instances (Wood et al., 1991). Dhuyvetter et al. (1996) summarized economic studies from across the Great Plains and concluded that more intensive systems also yielded greater net returns. What principles govern the efficient use of precipitation in intensified systems? The underlying concepts that favor cropping intensification as a solution to inefficient fallow are not entirely evident from the literature. The question is, How does intensification provide the potential for growing more crops (per unit time) in a given precipitation regime that traditionally produced only one wheat crop every 2 years?

B. OBJECTIVE Our objective in this article is two-fold: ( I ) to explore the concept of drylandcropping intensification as a fundamental and practical solution to improved use of precipitation, and (2) to propose a systems approach for analyzing, evaluating, and comparing intensified dryland-cropping systems. In this quest, we first present a review (Section 11) of research on precipitation storage and efficiency during different parts of the fallow period in the Great Plains. The review is not intended to be exhaustive, but it examines significant findings in winter and spring wheat-fallow systems. The emphasis in this chapter is mainly on systems involving winter wheat, but the concepts discussed are equally relevant to spring wheat. We then provide (Section 111) a more in-depth examination of the various periods of fallow using data from a long-term dryland-no-till cropping systems field study. The number of crop and noncrop periods in an intensified cropping system depends on the degree of intensification. Evaluation and comparison of intensified systems are made difficult because the duration and frequency of crop and noncrop periods vary, and their time-of-year precipitation characteristics vary among systems with differing crop choice and sequence. Quantitative measures and indices are needed to evaluate intensified rotations on a system basis. In Section IV, we propose a systems approach to intensification and present a collection of single-

DRYLAND CROPPING INTENSIFICATION

201

value system indicators that allow comparison of cropping systems on an equal basis, i.e., irrespective of the cropping intensity. Our goal is to simplify cropping systems analysis for the purposes of research and application.

II. SUMMER FALLOW: A SECOND LOOK The focal point of previous fallow research has been enhanced soil-water storage through improved tillage equipment, reduced number of tillage operations, and increased surface residue cover. Less tillage coupled with more surface residue coverage has provided the most practical means of minimizing erosion, enhancing infiltration, and retarding runoff and evaporation. Most previous research, however, concentrated on evaluating fallow as a whole. Literature on precipitation storage and efficiency during different parts of the fallow period is limited. A summary of significant research is presented in Tables I (spring wheat) and I1 (winter wheat).

A. SPRINGWHEAT-FALLOW SYSTEM Haas and Willis (1962) summarized data collected over 40 years for the alternate spring wheat-fallow system and reported that 54% of the total 21 -month fallow storage of 1 l 1 mm was stored from August harvest to spring (Table I), and 84% was stored by July 1 (not shown in Table I). Of the 300-mm precipitation received from spring to fall, only 17% (5 1 mm) was stored in the soil profile. On the average, no precipitation was stored during the second winter of fallow. These inefficient periods of fallow reduced the 40-year mean efficiency for the entire 21month fallow to only 19%. These results were reconfirmed in a study conducted at Sidney, Montana, by Black and Power (1965). As summarized in Table I, fallow storage efficiency was the highest from harvest to spring (60%) and lowest for the summer of fallow from spring to fall (5%). On average, of the 109-mm total fallow storage, 76% was stored the first winter, 9% during the summer of fallow, and 15% the second winter. Both sets of investigators regarded runoff as insignificant on their sites, suggesting that the evaporation was the major cause of low efficiency. For stubble-mulch and no-till fallow in spring wheat, Tanaka and Aase (1987) reported that over 60% of water storage occurred from harvest to spring when the land was in stubble, a lesser amount from the following spring until fall, and still less during the second winter. For the northern plains data (Table I), mean precipitation was greatest during the summer of fallow (2 14 mm), of which 84% (1 80 mm) was lost. Note that the mean precipitation storage of 74 mm from harvest to spring plus the average 2 14

Table I Soil-Water Storage (SWS)and Precipitation Storage Efficiency (PSE) during Specific Periods of the 21-Month Fallow in a Spring Wheat-Fallow System Specific periods of the 21-month fallow" Harvest to spring

Spring to fall

SWS (mm)

PSE (70)

of total SWS

SWS

PSE

gr, of total

(mm)

(%)

60

33

54

51

83

60

76

72 80

51

56

65 62

o/c

Reference

Haas and Willis, 1962 Plow (Mandan, ND) Black and Power, 1965 Minimum- and no-tiU (Sidney, h4T) Tanaka and Aase, 1987 Stubble-mulch (Sidney, MT) No-till

Yearsofdata

8 of total SWS

PSE

SWS

SWS (mm)

17

46

0

0

0

111

19

10

5

9

16

19

15

109

27

34 41

19 23

31 31

4 9

9 19

4 7

110 130

29 35

(%)

SWS (mm)

PSE

(a)

1915-1954 1956-1964 1981-1 985

Mean precipitation (mm) Soil water storage (mm) Precipitation lost (mm) "SWS (for a given fallow period)

Entk21-mth fallow

Fall to seeding

15 1

74 77 = profile

214 34 180

64 7 56

429 I15 3 14

soil water at the end minus the profile soil water at the beginning of the fallow period. PSE (for a given fallow period)

= (SWSdivided by precipitation during that fallow period) X 100. Percentage of total SWS = SWS during a given period of fallow divided by total stored water during the entire fallow) X 100.

DRYLAND CROPPING INTENSIFICATION

203

mm of summer precipitation provides the potential for 288 mm of available water for possible inclusion of a summer crop in the rotation.

B. WINTERWHEAT-FALLOW SYSTEM Black et ul. (1974) summarized 14 years of winter wheat-fallow data from Sidney, Montana, and divided the fallow into two periods: (1) harvest to spring, and (2) summer of fallow. For stubble-mulch fallow, an astonishing 84% of the total fallow storage was saved from harvest to spring, a period in which only 36% of total fallow precipitation was received. The remaining 64% of precipitation (216 mm) during the summer of fallow only contributed 16% (15 mm) to storage and the rest (20 I mm) to evaporation. Greb et ul. ( 1967) studied the effects of mulch loading rates on fallow storage in a winter wheat-fallow system and reported that water storage from harvest to late spring represented over 90% of the total fallow storage (determined from Table I1 by summing storage from harvest to fall and from fall to late spring). As shown in Table 11, a large portion of precipitation storage occurred during winter of fallow. During the summer of fallow (from late spring to seeding), 7 out of 10 experiments yielded a negative water storage, even under residue amounts as high as 10 t ha-'. Over the entire fallow period, fallow storage increased with increasing residue loading rate. Examining a range of tillage and residue management methods, the work of Smika and Wicks (1968) and Tanaka and Aase (1987) confirmed previous findings (Table 11). The most intriguing observation from Table I1 is that across the Great Plains, from 68 to 148 percent of total precipitation storage in the entire 14-month fallow period was achieved from harvest to spring. The overwinter period was by far the most efficient, and the summer of fallow the least efficient period of fallow. Precipitation during the latter period was almost entirely lost to evaporation. The problem of low precipitation-storage efficiency has been only partially improved by modem tillage and residue management practices. Fallow storage efficiency has increased from the 10% range under intense tillage operation at the turn of this century to the 20-30% range under the modern no-till and residue management techniques. It is evident that even under modem conservation practices, the original criticism of fallow still remains, and fallow precipitation-storage efficiency remains low.

In. DRYLAND CROPPING INTENSIFICATION Enhanced soil and water conservation is essential to the sustainability of dryland agriculture in the Great Plains. Fallowing is highly inefficient, as shown in

Table I1 Soil Water Storage (SWS)and Precipitation Storage Efficiency (PSE) during Specific Periods of the 14-Month Fallow in a Winter Wheat-Fallow System Specific periods of the 14-month fallow”

Harvest to late fall

Reference Greb et al., 1967 Sidney, MT

Akron, CO

North Platte, NE

Smika and Wicks, 1968 Plow (North Platte, NE) Stubble-mulch Reduced-till No-till Tanaka and Aase, 1987 Stubble-mulch (Sidney,MT) No-till

Residue level (ths-’)

Late fall to late spring % of

Yearsofdata

SWS (mm)

PSE

Entire 14-month fallow

% of

SWS

PSE

(mm)

(%)

-

-

56 67

-22 -21 -7 15 18 20 -7 -3 -17

12

178 184 220 258

64 66 78 92

122 91 97 94

37 35

41 38

52 48

41 33

total SWS

SWS (mm)

PSE

9 5 5 28 19 24 48 45 40

78 87 86 87 114 119 105 114 150

-

-30 2

34

-23 2 6 18

37 40

43 46

(%)

Late spring to seeding

(%)

total SWS

Precip. (mm)

SWS (mm)

PSE

-

355 355 355 549 549 549 648 648 648

56 66 79 142 163 183 190 203 223

16 19 22 26 30 33 29 31 34

12 14 -6 -18

7 9 -4 -11

640 640 640 640

146 203 226 214

23 32 35 43

21 36

16 27

299 299

99 114

33 38

(%)

1962- 1965 0 1.7 3.4 1.7 3.4 6.7 3.4 6.7 10.1

5

3 4 40 31

-

-

4 4 92 92 90 -

-

-

139 132 109

61 70 65 55

-

-

-

1963-1966 -44 4 11

5

1981-1984

“SWS (for a given fallow period) = profile soil water at the end minus the profile soil water at the beginning of the fallow period. PSE (for a given fallow period) = (SWSdivided by precipitation during that fallow period) X 100. Percentage of total SWS = (SWS during a given period of fallow divided by total stored water during the entire fallow) X 100.

DRYLAND CROPPING INTENSIFICATION

205

the previous section. The soil-water storage data suggest that enhanced efficient use of precipitation may be possible if summer crops are inserted in periods that have low water-storage efficiency. In the remainder of this chapter, we use data from the Sustainable Dryland Agroecosystem Management Project (Peterson et al., 1993) as a case study to develop a better understanding of the concept of intensification and its influence on precipitation storage and use. That project was established in 1985 to address precipitation use efficiency under dryland-no-till cropping systems at three locations in the west-central Great Plains region. The experimental locations, with long-term precipitation ranging from 400 to 450 mm year-', represent nearly a two-fold increase in pan evaporation from north (Sterling, Colorado) to south (Walsh, Colorado). The crop-management systems imposed in each location are a continuum with increasing cropping intensity and fewer summer fallow periods per unit time (Table 111).All systems are managed with no-till techniques. The benchmark cropping system is the winter wheat-fallow (WF). Cropping intensity increases for the 3-year rotations of winter wheat-corn-fallow (WCF) and winter wheat-sorghumfallow (WSF), and the 4-year rotations of winter wheat-corn-millet-fallow (WCMF) and winter wheat-sorghum-sorghum-fallow (WSSF). Herafter we refer to sorghum in WSF and the first-year sorghum in WSSF as sorghum- 1 and the second-year sorghum in WSSF as sorghum-2. In part A of this section, we provide a summary of results tr7nn-rkepreceding case study along with other modern no-till cropping studies from the central and southern Great Plains region. Our intention is to reiterate the state-of-the-art research findings regarding the potential of intensifying cropping systems in the Great P1ai n s .

A. MODERNDRYLAND-NO-TILL CROPPING SYSTEMS Table IV provides a summary comparison of modern no-till winter WF and more intense 3- and 4-year cropping systems from the Great Plains. The longest fallow period in a dryland cropping system always precedes the winter wheat crop and varies in duration from approximately 14 months in WF to 10-13 months in the 3- and 4-year systems. Length and time of fallow influence the amount of precipitation received during fallow, with the 14-month fallow in WF having the largest mean precipitation of 657 mm for all locations. Two of the most significant observations from Table IV are as follows. First, available soil water at wheat planting in all systems at a given location was similar, in spite of the fact that precipitation received during the 14-month fallow in WF was 140-250 mm greater than precipitation during the fallow preceding wheat in the 3- and 4-year systems. We can conclude that available soil water at wheat planting is not a function of the intensity of the cropping system as long as the

Table III Yearly Crop and Fallow Sequence in the Dryland Cropping Systems Case Study at the Sterling and Stratton [wheat-fallow (WF), wheat-corn-fallow (WCF), and wheat-corn-millet-fallow(WCMF)] and Walsh [wheat-fallow (WF), wheat-sorghum-fallow (WSF), and wheat-sorghum-sorghum-fallow(WSSF)] Experimental Locationsin Eastern Colorada" Cropping system WF WF WCF, WSF WCF, WSF WCF, WSF WCMF, WSSF WCMF, WSSF WCMF, WSSF WCMF, WSSF

Yearly crop and fallow sequence 1988-1989

1989-1 990

1990-1991

1991-1 992

1992-1993

1993-1994

W F W

F W

W F F W

F W W

W F

F W

-C(S)

- C(S) F

F W -C(SI) -M(S2) F

W -C(Sl) -M (S2) F W

- C(S) - M(S2)

F W -C(Sl)

- C(S)

-C(S) F

F F W -C(Sl) -M(S2)

W W -C(Sl) -M(S2) F

W F W

F W -C(S) -C(Sl) -M(S2) F W ~~

1994-1 995

- C(S) F -M(S2) F W -C(Sl)

~~

"W, winter wheat; F. fallow preceding wheat; C(S), corn (sorghum); C(S 1). corn (first sorghum following wheat); M(S2), millet (second sorghum following wheat). At Walsh, proso millet (1989-1990) and forage sorghum (1 991-1992) were planted in place of second sorghum in the WSSF system. All phases of each cropping system are present every year.

DRYLAND CROPPING INTENSIFICATION

207

wheat is preceded by a lengthy fallow. One may additionally conclude that from a precipitation management perspective, the similarity of soil water at wheat planting favors the intensified systems over WF, since on the average 200 mm less precipitation was required in the intensified systems to store nearly the same amount of soil water. Even under the most intense conservation practice of no-till, the amount of precipitation stored in the soil profile during the lengthy fallow that precedes winter wheat is extremely low, ranging from a minimum of 11% at Bushland, Texas, to a maximum of 27% at Stratton, Colorado. The noncrop (fallow) period preceding corn (sorghum- 1) plantings averages 11 months. As shown in Table IV (and previously by McGee et al. [ 1997]), both precipitation storage and efficiency during the noncrop period preceding corn (sorghum-1) were much greater than for fallow preceding wheat in WF, even though precipitation was on average 232 mm less in the former than in the latter case. For instance, mean precipitation storage and efficiency during the noncrop period preceding corn (sorghum-1) were 137 mm and 35% storage, respectively, compared with 119 mm and 18% storage for WF. The second important observation (Table IV) is that at each location, plantavailable soil water at corn (in WCF and WCMF) and sorghum-1 (in WSF and WSSF) planting was very similar to the corresponding soil water at wheat planting, in spite of the fact that corn (sorghum-1) was planted in May (June) while wheat was planted 5 (4) months later in September. This was true even though nearly 6 0 4 5 % of annual precipitation occurred during the 4-5 month period between corn (sorghum-1) and wheat planting. This means that most precipitation received during the summer of fallow just preceding wheat planting in the WF system was lost. That precipitation, however, was efficiently used in the intensified systems to grow an additional crop of corn or sorghum. Having equal amounts of soil water at wheat and corn (sorghum-1) planting reflects the underlying basis for the practice of opportunity cropping, in which spring soil-water content is used to determine if conditions are favorable for summer cropping. According to soil-water profile values at corn (sorghum- 1) planting (Table IV), spring conditions for summer cropping were on average favorable at all locations.

B. SUMMERCROPPING: KEYTO EFFICIENT USE OF PRECIPITATION In the preceding section, we document the potential for intensifying the 2-year WF system. Here we investigate the key elements influencing the efficient use of precipitation by cropping intensification and begin by examining precipitation storage and efficiency during different periods of the long fallow preceding wheat. Figure 1 illustrates a side-by-side comparison of the WF and WCF systems. The long fallow in WF can be divided into three distinct periods: (1) early period (from

Table IV Summary of Plant-Available Soil Water (PASW) at Crop Plantingand Precipitation (P), Soil-WaterStorage (SWS), and Precipitation Storage Efficiency (PSE)during the Noncmp (Fallow)Period Just Preceding That Cmf

Unger. 1994

1984-1991

Bushland. TX (notlll)

Nowood, 1994

WSF

226

463

74

16

228

598 455

I37

23

-

-

-

146

32

215

38 I

175

517

I45

1987-1992

Ga&n City. KS noti ill)

Jones and Popham. 19Y7

WF

212

WSF

181

WF

212

730

80

II

-

-

-

WSF

205

477

81

17

214

480

101

1984- I993

Bushland. TX (no-1111)

28

-

-

-

-

Farahani el 0 1 . 1998 Sterling. CO (no-till)

Stranon. CO (no till)

Walah. CO (no-ull)

Mean

198&199S

WF

I a7

610

99

16

-

-

-

-

WCF

I76

420

82

19

I97

362

131

37

WCMF

200

449

51

II

286

644

I75

27

361 -

127

WF

200 -

-

36 -

WCF

27 I

456

I I2

2s

286

338

161

54

WCMF

219

485

63

12

203

703

I03

IS

338 -

161

WF

278 -

-

48 -

WSF

216

47 I

I17

26

215

474

I08

?I

WSSF

20s

454

77

17

202

474

98

19

2-year

220

657

1 I9

18

-

-

-

-

-

213

451

I02

23

226

425

I37

3s

-

-

3-year

-

-

-

&year

228

463

63

13

227

39 1

I29

34

193

210

85

43

"Values are means across years. PASW = total soil-water profile (average depth of 1.5 m) minus soil-water profile at 15-bar water content. SWS (for a given fallow) = soil-water profile (average depth of 1.5 m) at the end of fallow minus soil water at the beginning of fallow. PSE (for a given fallow) = (soil-water storage divided by precipitation during that fallow period) X 100. %orghum-l denotes sorghum in WSF and the first sorghum following wheat in WSSF. 'Sorghum-2 denotes the second sorghum following wheat in WSSF.

H. J. FAEUHANI ETAL.

2 10

I +--

Fallow

-+1

102 n

B B 76 w

WCF

I

I

. .

*&

3

.a

a

*ij

t

51

PI

Jan.

DW. Year 1

Jan. Year 2

Dec. Year 3

Year 4

Figure 1 A time-scaled representation of the winter wheat-fallow (WF) and winter wheat-cornfallow (WCF) systems marking the beginning and ending of all crop and noncrop periods. Average (1948-1995) monthly precipitation amounts are also shown for the Stratton experimental location. (Numbers above bars represent percentage of yearly precipitation occurring in that month.)

wheat maturity in July to mid-September), (2) overwinter period (from fall to early May), and (3) late period (from spring to wheat planting in mid-September). Note that the various crop and noncrop phases in the WCF system fit nicely within these periods. The noncrop period preceding corn is represented by the sum of the early and overwinter periods, and the fallow after corn harvest corresponds to the sum of the overwinter and late periods in WF. The partitioning of fallow into these three periods was not arbitrary. Each phase bears a distinct identity in regard to soil-water status, climate, precipitation, and duration conditions (Black and Bauer, 1988),except that the climate is similar during the early and late fallow periods. The early period represents the highest residue level in the WF cycle (and even higher in the WCF system due to remaining corn residue), the driest soil profile in the cycle, a short duration of about 3 months, and an average (1988-1995) precipitation of about 200 mm. In this period, the high residue levels coupled with dry soil profiles are ideal for enhanced in-

DRYLAND CROPPING INTENSIFICATION

211

filtration, even though evaporation potential is high. The overwinter period has the lowest potential evaporation rates, low to medium soil-water profiles, and a long duration of about 6 months with a mean precipitation of 186 mm-conditions favorable for potentially high storage of precipitation. The late fallow period, on the other hand, represents the lowest residue levels in the cycle, the highest potential evaporation rates, medium to wet soil-water profiles, and a duration of about 4 months with a mean precipitation of 261 mm. These conditions favor evaporation and runoff. These qualitative descriptions of the three periods of fallow in WF are represented quantitatively in Table V. This table was constructed by using the 1988-1995 data from WF and WCF (WSF at Walsh) systems in our case study. Soil-water storage values were first calculated for the early, overwinter, and late periods. For each fallow period in our systems, mean rates of evaporation (mm day- I ) were determine as the ratio of fallow precipitation minus the storage to fallow duration in days. The three periods of fallow are distinctly different (Table V). Ranked in order of fallow efficiency, overwinter period was the most efficient (61 %), having the lowest mean rate of evaporation per day (0.56 mm day-') and the greatest amount of storage ( 1 1 1 mm) even though precipitation was at its lowest (at Sterling and Stratton) during this period. The early period ranked second in terms of storage (22 mm), efficiency ( 12%),and evaporation rate (1.86 mm day-'). The late (or the summer of fallow) period was by far the most inefficient (-4% storage), even though the greatest amount of precipitation (261 mm) is received during this time. It had evaporation rates of about 2.2 mm day-'. To simplify discussion of results, we assigned colors (zones) to each fallow period based on the intensity-of-evaporation rates during the period-the orange zone refers to the early period, the blue zone to the overwinter period, and the red zone to the late period. On average (see Table V), 11 1 mm (or 89%) of the total 125-mm fallow storage occurred during the blue zone. During the red zone, no water was conserved in this no-till fallow. As shown in Fig. 1, the red zone fallow (i.e., the primary zone of inefficiency in the WF system) is precisely the period that is replaced by corn or sorghum in the more intensified 3-year systems. It appears that if no plants are present to use the soil-water reservoir during the red zone fallow, the atmosphere will consume it through evaporation. The red zone or the summer of fallow can be eliminated only by abandoning winter wheat. That solution is unrealistic, since winter wheat is the comer stone of dryland agriculture in the Great Plains. Thus, the only plausible and practical solution to the unavoidable red zone fallow is to reduce its frequency of occurrence by intensification or summer cropping. Inclusion of one summer crop in the WF system reduces the frequency of occurrence of the red zone fallow by 33%-from one in every 2 years to one in every 3 years.

Table V Mean Precipitation (P), Soil-Water Storage (SWS),Precipitation Storage Efficiency (PSE),Duration (D),and Daily Evaporation Rate (Er) for Three Specific Periods of the 16Month Fallow in the Wheat-Fallow System as Affected by Location (Climate) ’hahnenW Entire 14-month fallow in wheat-fallow system Early period (orange zone)

Location

D (days)

P (mm)

SWS (mm)

PSE

90

17

64

(%)

Sterling Stratton Walsh

87 111

197 189 232

- 14

9 34 -6

Mean

96

204

22

12

“Values are means for 1988-1995.

Overwinter period (blue zone)

Late period (red zone)

Er (mmday-I)

D (days)

P (mm)

SWS (mm)

PSE (%)

Er (nunday-’)

D (days)

P (mm)

SWS

PSE

(mm)

(%)

Er (mmday-’)

1.92 1.44 2.22

180 178 252

169 146 243

114 97 123

66 66 51

0.63 0.55 0.49

129 136 102

251 310 229

- 32 14 -6

13 5 -2

2.17 2.16 2.30

1.86

203

188

111

61

0.56

122

261

-8

-4

2.21

-

DRYLAND CROPPING INTENSIFICATION

213

W. A SYSTEMS APPROACH TO INTENSIFICATION

A. SYSTEMS ANALYSIS The importance of evaluating field agronomic problems from a systems perspective was emphasized by Peterson et al. (1993). The complexity and the highly interrelated processes in the natural environment require a systems approach. The element of interest herein is the influence of varying crop and noncrop periods on systems behavior. We suggest an analytical approach that simplifies the complexity of the interactions among the many possible crop and noncrop phases and provides a tool for preliminary testing of newly proposed systems. In our study, the system is defined as a complete cycle of a dryland rotational cropping sequence. A dryland system includes all crop and noncrop periods. For convenience, the system is assumed to have a similar day-of-year beginning and ending. For instance, in 2-, 3-, and 4-year systems of WF, WCF, and WCMF, respectively, wheat planting to wheat planting defines system duration. The benchmark cropping system is the winter WF with one crop every 2 years; or a 0.5 cropping (and 0.5 summer fallow) intensity per year. Cropping intensity increases to 0.67 (0.33 summer fallow intensity) for 3-year rotations that include a fallow preceding wheat (i.e., WCF and WSF), and to 0.75 (0.25 summer fallow intensity) for 4-year rotations that include a fallow preceding wheat (i.e., WCMF and WSSF). A cropping intensity of unity is attained for continuous (not necessarily monoculture) cropping, i.e., winter wheat-corn-millet (WCM). We used the long-term (19484995) precipitation data in Table V to conduct an analysis of cropping systems. The systems evaluated were WF, WCF (WSF), and WCMF (WSSF), as in the case study presented earlier. Hypothetical systems of even greater intensity, such as WCM, winter wheat-millet (WM), and winter wheat-corn (WC), also were evaluated. Figure 2 is a time-scaled representation of the systems we analyzed. A 12-year period was selected for the analysis because it marks the first simultaneous closure of all systems. This choice, however, has no bearing on the interpretations. For the analysis, planting and harvest dates of each crop were set as constants for every year in the 12-year period and corresponded to average dates obtained from our previous field experiments. From the 47-year ( 1948-1995) precipitation record for the three experimental locations, average precipitation was determined for the orange, blue, and red zones, along with annual and growing-season precipitation for all crops. For convenience, the length of the orange, blue, and red zones was set at 2.5 months (July 1 to mid-September), 7.5 months (mid-September to the end of April), and 4.5 months (May 1 to midSeptember), respectively. We then used the efficiencies computed previously for each of the three periods of fallow (Table V) to construct Table VI. From Table VI, we can see that as cropping intensity increased from 0.5 for WF

H. J. FARAHANI ETAL.

2 14 Plantinr!

Harvest

I

I

WF

7n Wheat (9.5)

Wheat

01 I

B

Corn (4.5)

B

' Wheat

Corn

WCF

I

I

B

511 R9n5

I

{

WCMF

(3)

wc

Figure 2 A time-scaled representation of the wheat-fallow (WF), wheat-corn-fallow (WCF), and wheat-corn-millet-fallow (WCMF) systems of the case study plus the hypothetical cropping system of wheat-corn (WC). 0, orange zone; B, blue zone; R, red zone; (#), period in months.

to 0.75 for the 4-year systems, the total months in noncrop (fallow) for the 12-year ( 144 months) period actually increased. Cropping intensification did not reduce months of noncrop (fallow), but the duration of the fallow period preceding wheat planting was reduced. The greatest reduction in noncrop (fallow) months occurred for the hypothetical continuous cropping systems of WCM, WM, and WC. It is important to note that intensification, moving from WF to 3- and 4-year systems, increased the amount of noncrop time in the blue zone and reduced the time in the red zone. By reducing the noncrop time spent in the red zone, the amount of precipitation lost to evaporation was decreased. In comparing WF with the 3- and 4year systems, total noncrop (fallow) time was reallocated to the more efficient blue zone from the efficient orange and red zones. This decreased the EIET ratio (the ratio of total system fallow evaporation, E, to total system evapotranspiration, ET) from 1.13 (1.7 I at Walsh) for WF to 0.65 for WCF (1.09 at Walsh for WSF) and 0.6 for WCMF (1.OO at Walsh for WSSF). Note that the sum of E and ET is equal to total system precipitation (assuming no precipitation losses other than evaporation). Thus, the ratio EIET quantifies the

Table VI Twelve-Year (144 months) Analysis of Dryland Cropping Systems in the West-Central Great Plains" Total I2-year duration (D).precipitation (PI. and soil-water storage (SWSI Total 12-yeilrr

Orange zone (early)"

Blue zone (ovenv~nter)~

Total 12-yearpericd

Red zone (late)h

Noncrop

Crop

Cropping

Cropping

Noncrop

Crop'

D

P

sws

D

P

SWS

D

P

SWS

P

I?

SWS

P

ETF

UET

*)stem

~ntensity

(months!

(months)

(months)

(mm)

(mm)

(months)

(nun)

imm)

(months)

(mml

(mm)

(nun)

(mm)

(mm)

(mm)

(mml

(mdmm)

Systems at Walsh. Colorado

WF

0.5

87

s7

15

869

- 52

45

747

18I

27

1649

- 33

3264

2968

2%

1440

1736

1.71

WSF

0.67

90

54

10

579

- 35

58

962

49 1

22

1344

-27

2885

2456

429

1819

2248

1.09

WSSF

0.75

92

52

10

s79

-

35

65

1070

546

20

1191

-224

2840

23S3

487

I864

23S1

1.00

Systems at Sterling and Stratton. Colorado WF

0.5

87

57

15

854

184

45

873

576

27

1698

-68

3425

2733

692

1723

2415

1.13

WCF

0.67

88

56

10

569

122

60

I164

768

18

1132

-45

2865

2020

845

2283

3128

0.65

WCMF

0.75

93

51

9

512

110

68

1310

864

17

1038

- 42

2859

1927

933

2289

3222

0.60

WCM

1

76

68

I2

683

147

60

1164

768

4

252

- I0

2098

I193

905

3050

3955

0.30

WM

1

63

81

15

854

184

45

873

576

9

566

-23

2293

I556

137

2855

3592

0.43

WC

I

60

84

15

854

184

45

873

576

0

0

1727

%7

760

3422

4182

0.23

0

"Mean long-term (1948-1995) annual (Jan-Dec.) precipitation for the three locations analyzed are 429 mm at Sterling and Stratton and 392 mm at Walsh. "Mean long-term (1948-1995) precipitation for the non-crop (fallow) periods are 571 mm (Sterling and Stratton) and 534 m m (Walsh) for the 14.5-month fallow (July I-Sept. 15). 142 m m (Sterling and Stratton) and 145 m m (Walsh) for the 2.5-month orange zone (July I-Sept. 15). 146 nun (Sterling and Stratton) and 124 mm (Walsh) for the 7.5-month blue zone (Sept. I&April30), and 283 m m (Sterling and Stratton) and 275 m m (Walsh) for the 4.5-month red zone (May I-Sept. 15). 'Beginning and ending crop growing seasons used to construct the table are wheat (Sept. 16-June 30),corn (May I-Sept. 15). millet (June I-Aug. 30), sorghum (June I-Sept 30). dE = total 12-year evaporation from the soil during all noncrop (fallow) periods. 'ET = total I '-year evapotranspiration during all crop periods.

2 16

H. J. FARAHANI ETAL.

relative allocation of system precipitation to noncrop (fallow) evaporation and crop E’I: defined herein as “system precipitation allocation index (SPAI).” As shown in Table VI, the EIET ratio for the WF system was about double that for the 3- and 4-year systems. In a relative sense, the lower the ratio, the more precipitation efficient the system. Note that for WF, the ratio WET was above unity (particularly at Walsh), implying that the loss of system precipitation to fallow evaporation exceeded the water allocated for crop production (or E T ) by 13% (Sterling and Stratton) and 7 1% (Walsh). It is interesting that the inclusion of a summer crop in the WF system (i.e., the 3-year WCF system) caused a significant reduction in the EIET ratio as a result of reducing E by 26% and increasing ET by 30%. Continuous cropping systems like our hypothetical WCM, WM, and WC substantially decreased the EIET ratio. In the WC system, the E/ET ratio was decreased five-fold compared with WF, as a result of a 65% reduction in E and a 73% increase in Ei? The red zone fallow period was reduced to nil. Note that the total 12-year soil-water storage during all noncrop (fallow) periods in WC was about 70 mm (or 10%)more than the soil-water storage during fallow in WF, while precipitation received in the latter fallow was about 1700 mm greater than in the former fallow. In other words, it was not the amount of precipitation storage between crops that made the difference, but the strategically placed summer corn crop in the red zone that utilized the 1700 mm of precipitation to produce biomass in the WC system as opposed to being lost to evaporation in the WF system. Some have credited the increased surface residue mass (cover) as being the major factor contributing to the improved efficient use of precipitation in intensified cropping systems. In contrast, our data indicate that residue is not the key concept but only a single component of the system. The gains in efficiency with cropping intensification are due not to an enhanced water conservation but to a reallocation of water from evaporation from the soil during the summer of fallow into the transpiration stream of a plant. Thus, the underlying basis for intensification is a partial replacement of soil evaporation with crop transpiration. Figure 3 shows the systems ranked according to their EIET ratio. The systems with an intensity of unity (i.e., WCM, WM, or WC) are predicted to be much superior to WF. In these hypothetical systems, it is quite obvious that the wheat crop may have a yield reduction due to less stored water at planting. The possible superiority of the continuous systems in efficient utilization of precipitation still may not be profitable. The research question at hand is, is a cropping intensity of unity economically sustainable? That, of course, remains to be determined.

B. SYSTEMSEVALUATION QUANTITATIVE INDICES Information regarding individual crop and noncrop (fallow) periods are not by themselves a sufficient measure of the effectiveness of the entire system. For the

DRYLAND CROPPING INTENSIFICATION

217

'1 W

0

1

1

Dryland Cropping Systems Figure 3 Ranking of dryland cropping systems in order of increasing system precipitation use, given by the ratio E / E T ( E = total system fallow evaporation, ET = total system crop evapotranspiration). Systems denoted by * are based on precipitation and storage data from Walsh, Colorado; all others are based on mean precipitation and storage data from Sterling and Stratton, Colorado.

purposes of system design, evaluation, comparison, and management, quantitative indicators are needed to measure the systems individually and to weigh them against each other. In terms of system design, there are many elements associated with intensified systems. In the Great Plains, however, the most important system element is the fate of incident precipitation. Other important elements are weeds, fertility, pests, and equipment. Obviously, system adaptation by farmers would require additional information about system economics and practical feasibility, which are not discussed in this chapter. Our discussion concerns quantifying the effectiveness of an intensified system to utilize precipitation. En route, three questions are of particular importance: (1) how efficiently precipitation received during the noncrop periods is stored in the soil, (2) how effectively system precipitation is allocated between crop and noncrop periods, and (3) how efficiently the stored water is utilized to produce biomass. Farahani et al. (1998) calculated system indices to address the first two questions. These indices are the system precipitation storage index (SPSI), a measure of how efficiently the incident precipitation during all noncrop (fallow) periods is collectively stored in the soil, given by Ef System precipitation storage index (SPSI) = 1 - Pf

(1)

218

H. J. FARAHANI ETAL.

and system precipitation use index (SPUI), a measure of how the system as a whole allocates total incident precipitation to crop production, given by Ef System precipitation use index (SPUI) = 1 - -

(2)

PS

where E, is the sum of all noncrop (fallow) precipitation losses (assumed to be equal to evaporation from the soil), P , is the sum of all noncrop (fallow) precipitation, and P, is the total precipitation during a complete cycle of the system (i.e., from wheat planting to wheat planting). The difference between P , and P , is the total amount of incident precipitation during all crop periods (Pc).Both indices have upper limits of unity recognized as Ef + zero. The SPSI has a lower limit of zero as fallow evaporation (or losses) approaches P,. The lower limit for SPUI varies among systems; however, it is equal to PJP, as E f + P,. The SPSI quantifies the unit fraction of noncrop (fallow) precipitation allocated to soil-water storage (S,), and thus may be written as S,/P,. This is the equivalent of the storage efficiency for a single fallow but is written for the whole system. The SPUI quantifies the unit fraction of system precipitation (P,) allocated to crop season (i.e., evapotranspiration, ET,), and thus may be written as ETJP,. The advantages of these indices over storage efficiency for individual noncrop (fallow) periods are that they synthesize the behavior of all phases of the system into single-value indicators, allowing system comparison on an equal basis (i.e., irrespective of the intensity of the cropping system). The goal is to devise systems that increase both SPSI and SPUI toward unity within the bounds of commercial feasibility. By examining Eqs. (1) and (2), the most obvious solution to enhancing both indices is reducing noncrop (fallow) evaporation E,. Our third question concerns system production and productivity and its relation to the enhanced use of precipitation. Water-use efficiency ( W E ) , defined as the ratio of dry matter produced per unit of water used, has been used extensively in the past to quantify productivity on a seasonal basis. Peterson et al. (1996) considered WUE an equally important parameter for evaluating intensified systems, serving as a diagnostic tool that provides a single quantitative measure combining production and water use. Based on a literature review from the Great Plains, Peterson et al. (1996) concluded that with modern no-till techniques, WUE for WF has not increased significantly since the 1970s-a direct consequence of the corresponding stagnant fallow storage efficiencies. They stated, “Cropping systems intensification has allowed us to make the next step in improving WUE in the Great Plains.” Many investigators have discussed means of improving individual crop WUE (Tanner and Sinclair, 1983). Our interest is in WUE on a system basis, defined by WUEs (Peterson et al., 1996) V

DRYLAND CROPPING INTENSIFICATION

219

where Y, is the system yield (i.e., sum of harvest grain yields from all crops) (kg ha-') and ETs is the system growing season ET(i.e., sum of growing-season cropwater use from all crops) (mm). The ET for each crop was estimated as seasonal soil-water depletion plus seasonal precipitation. The advantage of WUEs over WUE for single crops is that it synthesizes the productivity of all crops in the system into a single-value indicator, allowing system comparison on an equal basis. A general rule to ensure that WUEs is increased by moving from the 2-year WF system to a 3-year intensified system is that the added crop must have a WUE value greater than that of wheat. Two examples from the literature are sorghum at Garden City, Kansas, with a WUE value of 12.6 as compared with 7.1 kg ha-' mmof ETfor wheat (Norwood, 1994); or corn at Sterling, Colorado, with a WUE value of 9.3 as compared with 6.0 kg ha-' mm-' of ET for wheat (Peterson et al., 1996). Fortunately, most adapted summer crops in the Great Plains have WUE values greater than that of winter wheat. This is why nearly all WUEs values in every climate regime from Texas, Kansas, and Colorado were found to be greater than the corresponding values for the 2-year wheat-fallow system (Peterson et al., 1996), averaging 8.5 kg ha-' mm-' in the 3-year system as compared with 6.1 kg ha-' mm-' for WF. By the same argument, improving on the WUEs of a 3-year system by moving to a 4-year system will be ensured by adding a crop with a WUE greater than the WUEs of the 3-year system. For instance, to improve on the 8.5 kg ha- I mm-' in the 3-year systems reported by Peterson et al. (1996), we need to include a crop with a WUE value greater than 8.5 kg ha-' mm-' (e.g., proso millet). The preceding procedure may be used to devise intensified systems that tend to increase WUEs. For our dryland-no-till case study, mean values for system indices and indicators of SPAI, SPSI, SPUI, and WUEI and annualized grain yields are reported in Table VII. The annualized grain yield values are single-value measures of system production. According to Table VII, WUEs averaged 5.4, 6.9, and 7.4 kg ha-' mm- I for the 2-, 3-, and 4-year systems, corresponding to annualized grain yields of 1030, 1770, and 1950 kg ha-', respectively. Although differences between the 3- and 4-year systems are small, intensifying beyond the 2-year wheat-fallow system increased productivity (i.e., WUEJ by 29 and 39% and production (annualized yield) by an astonishing 72 and 90% per year in the 3- and 4-year systems, respectively. According to SPSI results (Table VII), for every unit of incident precipitation during the noncrop (fallow) periods, 0.19, 0.28, and 0.26 units were stored in the 2-, 3-, and 4-year rotations, respectively. This means that the noncrop (fallow) periods in the 3- and 4-year rotations were collectively 47 and 37%, respectively, more efficient in storing precipitation than fallow in WE According to SPUI results, for every unit of precipitation in the WF system, only 0.36 (Walsh), 0.44 (Sterling), and 0.49 (Stratton) units are made available for crop production, with the remainder being lost. As the intensity of the cropping system increased, so did SPUI. However, the 3- and 4-year systems were not significantly different.

'

220

H. J. EARAHANI ETAL. Table VII

Summary of System Precipitation Storage Index (SPSI), System Precipitation Use Index (SPUI), System PrecipitationAllocation Index (SPAI), System WUE (WUE. ), and Annualized Grain Yield Values for the 2-, 3-, and 4-Year Cropping Systems at Three Experimental Locations in the West-CentralGreat Plain@ Single-value system indices and indicators''

SPSI Location Sterling

Stratton

Walsh

Mean

Cropping system

SPUI

(S,./P,) (ETJPJ (mm m-I)(mm mm-'1

SPAI

(E,IETJ (YJET,) (mm m-') (kg ha-' m m - ' )

Annualized grain yield (kg ha-')

WF WCF WCMF WF WCF WCMF WF WSF WSSF

0.16 0.27 0.26 0.27 0.34 0.3 1 0.15 0.24 0.22

0.44 0.57 0.58 0.49 0.62 0.61 0.36 0.47 0.51

1.30 0.69 0.68 1.16 0.6 1 0.63 1.I2 1.03 0.93

4.8 6.4 7.2 5.9 7.3 7.7 5.3 7.0 7.5

930 1770 I960 1250 1960 2110 910 1590 I790

2-year 3-year 4-year

0.19 0.28 0.26

0.43 0.56 0.57

I .40 0.78 0.75

5.4 6.9 7.4

1030 1770 1950

"Values are means for 1988-1995. 'The following variables are defined for a complete cycle of the cropping system (i.e,, from wheat planting to wheat planting): S, = sum of precipitation storage during all noncrop (fallow) periods in the system; E,. = sum of precipitationlosses (assumed to be equal to evaporation from the soil) during all noncrop (fallow) periods in the system; P, = sum of incident precipitation during all noncrop (fallow) periods in the system; P, = total incident precipitationduring a complete cycle of the system ( i c , from wheat planting to wheat planting); ETs = sum of growing-seasoncrop ETfor all crops in the system; Ys = sum of all harvest grain yields from all crops in the system.

Comparing the locations, Walsh, the site with the highest potential ET was the least-efficient utilizer of precipitation, with a SPUI ranging from 0.36 to 0.5 1. A timely placed summer crop, such as corn or sorghum, increased the unit fraction of precipitation allocated to crop production (i.e., SPUI)from 0.43 in WF to 0.56 (i.e., an increase of 30%) in 3-year systems. System indices and indicators of SPAI,SPSI,SPUI,WUEs and annualized yield (Table VII) collectively suggest that intensification can substantially improve on the WF system by enhancing precipitation use, production, and productivity. The gains by intensification result from using water that would be lost by evaporation from the soil during fallow in the transpiration stream of a plant and the associated increase in biomass production. Note that for the experimental period in our

DRYLAND CROPPING INTENSIFICATION

22 1

case study, annual precipitation was at or greater than normal. The potential of intensification to enhance efficient use of precipitation during dry years, with precipitation amounts of less than 300 mm, is not known.

V. CONCLUSION Research before the 1980s focused on improving the fallow practice, although Haas et al. ( 1974)and others questioned the wisdom of fallowing. Perspectives on fallowing began to change in the 198Os, and the underlying objective has been broadened to enhancing the efficient use of precipitation rather than just improving summer fallow efficiency. Particularly since research on the winter wheat-fallow system shows that in the Great Plains the amount of soil water accumulated by the late spring of the lengthy fallow preceding wheat is not significantly different from soil water accumulated 5 month later at wheat planting. This is in spite of the fact that nearly 65% of annual precipitation occurs during this latter 5-month period; meaning that on average most precipitation received during the last summer of fallow is lost unless a summer crop is planted. Cropping diversification is an integral part of intensification.For instance, annual cropping of winter wheat is cropping intensification as compar-d with alternating wheat with fallow, but the former may or may not be a feasible alternative. Furthermore, in moving from the 2-year WF to 3- and 4-year rotations, cropping intensity per year increases from 0.5 to 0.67 and 0.75, respectively. However, neither the annualized noncrop (fallow) duration (0.6 for WF, 0.61 for WCF, and 0.65 for WCMF) changes (actually, it increases slightly) nor the time-in-crop per unit time increases with cropping intensification. Intensification does decrease the summer fallow intensity per year, from 0.5 in WF to 0.33 in WCF (WSF) and 0.25 in WCMF (WSSF). A new era of dryland farming, characterized by cropping intensification and diversification, is emerging on the Great Plains and may someday dominate as summer fallow has in the past. Perhaps an even more stimulating thought is the hypothesis by Peterson and Westfall ( 1997) that “zero tilling, coupled with intensified crop rotations, is a movement toward an agroecosystem that mimics the Great Plains prairie ecosystem before cultivation began.”

ACKNOWLEDGMENTS Appreciation is extended to Lucretia L. Sherrod, USDA-ARS. M A , GPSR, Fort Collins, Colorado, for her generous help with data compilation and analysis. We also thank Ordie R. Jones, USDA-ARS, SPA, CPRL, Bushland. Texas, and John F. Shanahan, Dept. of Soil and Crop Sciences, Colorado State

222

H. J. FARAHANI ETAL.

University, Fort Collins, for their feedback on the first draft of the manuscript. In particular, we are indebted to James F. Power, formerly with USDA-ARS, NPA, SWCR, Lincoln, Nebraska, for his thorough review and invaluable suggestions. H. Farahani also thanks Laj R. Ahuja, USDA-ARS, GPSR, Fort Collins, Colorado, for his support during the analysis of the case study data.

REFERENCES Black, A. L., and Bauer, A. (1988). Strategies for storing and conserving soil water in the Northern Great Plains. In “Challenges in Dryland Agriculture: A Global Perspective” (P. W. Unger et al., eds.), pp. 137-139. Texas Agric. Exp. Sta., Amarillo, TX. Black, A. L., and Power, J. F. (1965). Effect of chemical and mechanical fallow methods on moisture storage, wheat yields, and soil erodibility. Soil Sci. SOC.Am. Proc. 29,465468. Black, A. L., Siddoway, F. H.. and Brown, P. L. (1974).Chapter 3. Summer fallow in the northern Great Plains (winter wheat). In “Summer Fallow in the Western United States.” USDA Conserv. Res. Report 17, pp. 36-50. U.S. Gov. Printing Office, Washington, DC. Dhuyvetter, K. C., Thompson, C. R., Norwood, C. A., and Halvorson, A. D. (1996).Economics of dryland cropping systems in the Great Plains: Areview. J. Prod. Agric. 9,216-222. Duley, F. L., and Russel, J. C. (1939). The use of crop residues for soil and moisture conservation. Agron. J. 31,703-709. Farahani, H. J., Peterson, G. A,, Westfall, D. G., Sherrod, L. A,, and Ahuja, L. R. (1998).Fallow water storage in drylandno-till cropping systems. Soil Sci. SOC.Am. J. (In press). Greb, B. W. (1979). “Reducing Drought Effects on Croplands in the West-Central Great Plains.” USDA Info. Bull. 420. U S . Gov. Printing Office, Washington, D.C. Greb, B. W., Smika, D. E., and Black, A. L. (1967). Effects of straw-mulch rates on soil water storage during summer fallow in the Great Plains. Soil Sci. SOC.Am. Proc. 31,556-559. Greb, B. W., Smika, D. E., and Welsh, J. R. (1979). Technology and wheat yields in the central Great Plains. J. Soil and WaferCons. 34,264268. Haas, H. J., and Willis, W. 0. (1962).Moisture storage and use by dryland spring wheat cropping systems. Soil Sci. SOC.Am. Proc. 26,506509. Haas, H. J., Willis, W. O., and Bond, J. J. (1974).Chapter2. Summer fallow in the northern Great Plains (spring wheat). In “Summer Fallow in the Western United States,” pp. 12-35. USDA Conserv. Res. Report 17. U.S. Gov. Printing Office, Washington, D.C. Halvorson, A. D., Anderson, R. I., Hinkle, S. E., Nielson, D. C., Bowman, R. A,, and Vigil, M. F. (1994). Alternative crop rotations to winter wheat-fallow. In “Proc. Great Plains Soil Fertility Conference,” pp. 6-1 I . Denver, Colorado. Johnson, V. A. (1977). “The Role of Wheat in America’s Future.” Sec. Pub. 30, pp. 3 7 4 4 . Am. SOC. Agron., Madison, W1. Jones, 0.R.,and Popham, T. W. (1997). Cropping and tillage systems for dryland grain production in the southern High Plains. Agron. J. 89,222-232. Mathews, 0.R., and Army, T. J. (1960). Moisture storage on fallowed wheatland in the Great Plains. Soil Sci. SOC.Am. Proc. 24,414418 . McGee, E. A., Peterson, G. A., and Westfall, D. G . (1997).Water storage efficiency in no-till dryland cropping systems. J. Soil and Water Cons. 52, I3 1-1 36. Norwood, C. A. (1994). Profile water distribution and grain yield as affected by cropping system and tillage. Agron. J. 86,558-563. Peterson, G. A., and Westfall, D. G. (1996). Maximum water conservation after wheat harvest. Nail. Cons. Tillage Dig. 3545). 9. Peterson, G. A., and Westfall, D. G. (1997). Benefits of zero till and rotations in the North American

DRYLAND CROPPING INTENSIFICATION

223

Great Plains. In “Proc. 19th Ann. Manitoba-North Dakota Zero Tillage Workshop,” pp. 5-16. Brandon, Manitoba. Canada. Peterson, G . A,, Westfall. D. G., and Cole, C. V. (1993). Agroecosystem approach to soil and crop management research. Soil Sci. Soc. Am. J. 57, 13541360. Peterson, G. A,, Schlegel, A. J., Tanaka, D. L., and Jones, 0. R. (1996). Precipitation use efficiency as affected by cropping and tillage systems. J. Prod. Ag. 9, 180-186. Smika, D. E., and Wicks, G. A. (1968). Soil water storage during fallow in the Central Great Plains as influenced by tillage and herbicides treatments. Soil Sci. Soc. J. 32,591-595. Stewart, B. A,, and Steiner, J. L. (1990). Water-use efficiency. Adv. Soil Sci. 13, 151-173. Tanaka, D. I., and Aase, J . K. (1987). Fallow method influences on soil water and precipitation storage efficiency. Soil and Tillage Res. 9,307-3 16. Tanner, C. B., and Sinclair, T. R. (1983). Efficient water use in crop production: Research or re-search? In “Limitations to Efficient Water Use in Crop Production” (H.M. Taylor et a/., eds.), pp. 1-27. American Society of Agronomy, Madison, WI. Unger, P. W. (1984). Tillage and residue effects on wheat, sorghum, and sunflower grown in rotation. Soil Sci. Soc. Am. J. 48,885-89 I. Unger, P. W. (1994). Tillage effects on dryland wheat and sorghum production in the southern Great Plains. Agron. J. 86, 310-314. Wood, C. W., Westfall, D. G., and Peterson, G. A. (1991). Soil carbon and nitrogen changes upon initiation of no-till cropping systems in the west central Great Plains. Soil Sci. Soc. Am. J . 55, 47M76.

This Page Intentionally Left Blank

How Do PLANTROOTSACQUIRE MINERALNUTRIENTS? CHEMICAL PROCESSES INVOLVED IN THE RHIZOSPHERE P. Hinsinger* Faculty of Agriculture University of Western Australia Nedlands, Western Australia 6907 Australia

I. Introduction 11. Definition of the Rhizosphere 111. Root-Induced Changes of Ionic Concentrations in the Rhizosphere N. Root-Induced Changes of Rhizosphere p H V. Root-Induced Changes of Redox Conditions in the Rhizosphere VI. Root-Induced Complexation of Metals in the Rhizosphere VII. Other Interactions Involving Root Exudates VIII. Conclusion References

I. INTRODUCTION The acquisition of mineral nutrients by plant roots is a broader concept than the uptake, which is the transfer of nutrients from the outer medium into the root. Indeed, the acquisition of nutrients also encompasses the many processes occurring in the immediate environment of the root as a consequence of its activity and influencing the dynamics of nutrients in the so-called rhizosphere prior to being ultimately taken up by the plant. Following Clarkson (1985), I should still “suggest that many of the intriguing processes that occur in the root-soil interface merit a more purposeful and integrated investigation, especially by plant physiologists, than they have received,” and I would add “by soil scientists too.” *Present address: Institut National de la Recherche Agronomique, UFRA de Science du Sol, F-34060 Montpellier Cedex 2, France 22s Advnnrcr in Agronomy,Volume 64 Copyright 0 1998 by Academic Press. All rights of reproduction in any form reserved 0065-2113/98 $25.00

226

P. HINSINGER

Most plant physiologists have addressed plant nutrition exclusively for hydroponic conditions. Consequently, they have concentrated their research on the sole uptake processes. Similarly, most soil scientists have studied the dynamics of mineral nutrients in soils in the absence of plants. Even though these reductionist approaches have increased our knowledge of the mechanisms involved in nutrient dynamics, their major limitation has been to ignore, or at least underestimate, the importance of the interactions between plant roots and soil constituents for mineral nutrition. Nevertheless, evidence of the profound chemical changes that occur in the rhizosphere has accumulated, as reviewed by Marschner et af. (1986), Darrah (1993), and Hinsinger (1994). Chemical conditions in the rhizosphere can thus be drastically different from those in bulk soil. This should alter the quality of the diagnosis of nutrient bioavailability, which is commonly deduced from the analysis of bulk samples of soil. A better prediction of nutrient bioavailability and mineral nutrition would thus take into account the chemical state and physical extent of the rhizosphere. These factors are either a direct effect of root activity itself or an indirect effect of the roots, i.e., the effect of the root-stimulated rhizosphere microflora. Even though the latter effect may be of prime importance, especially for major nutrients such as N and P,it will not be considered in this review, which concentrates on the direct effects of plant roots themselves. Some major groups of microorganisms occurring in the rhizosphere are root symbionts, such as N-fixing bacteria and mycorrhizal fungi (Curl and Truelove, 1986). Their role in supplying especially N and P to plants has been extensively reviewed elsewhere (e.g., Quispel, 1983; Bolan, 1991; Marschner and Dell, 1994). The present review aims at highlighting the processes involved in the acquisition of mineral nutrients from soil by plant roots, with a particular emphasis on the chemical interactions occurring in the rhizosphere as a direct result of root activity.

II. DEFINITION OF THE RHIZOSPHERE Since the early work of Hiltner (1904), who first used the term rhizosphere to describe the stimulation of microbial biomass and activity in the soil surrounding plant roots, the concept behind the word has received several more-or-less loose definitions (Curl and Truelove, 1986). It is now best defined as the volume of soil influenced by root activity. According to this broad definition, the size of the rhizosphere may extend well beyond the first millimeters of soil adhering to plant roots when one considers the uptake of water and mobile nutrients, notably nitrate, or the release of volatile compounds (Darrah, 1993). For the poorly mobile nutrients such as phosphate, the extension of the rhizosphere is often limited to less than 1 mm (Hubel and Beck, 1993). In addition, the spatial extension of the rhizosphere varies widely for any single nutrient, as shown in Table I for phosphorus.

Table I

Spatial Extension of Phosphous Depletion in the Rhizosphere of Various Species" Plant properties

Species

Age (days)

Rape

5 6 5 10 15 10 10 23

Onion

7 49

Maize

Clover Wheat Ryegrass

Bulk soil properties Sand

(mgg-') 770 100

850 170 170 680

Rhizosphere properties

Clay (mg g - ' )

OrganicC (mg g-')

pH

80 210 65 210 210 90 500 150

20 22 46 46 23 5 19

5.7 5.6 5.8 7.4 7.4 7.4 5.7 6.7

sandy loam silty clay loam

6.9 7.3

P (mg kg-')

158" 6' 42c 36od 360" 347' 400' 28Y 398 20h 27oh 2oJ

Analytical method A' A'

AJ Organic pd Organic pd

HCI' HCI' H,SO,f NaOH8 KHC0,h A'

A'

"Modified from Hiibel and Beck, 1993, with kind permission from Kluwer Academic Publishers. "Exchangeable P. 'Water-soluble P (Paauw, 1971). dOrganic P (Tarafdar and Jungk, 1987). '4M HC1-extractable P. f6M H,SO,-extractable P in a sequential procedure (Gahoonia and Nielsen, 1992). g0.lM NaOH-extractable P in a sequential procedure (Gahoonia and Nielsen, 1992). h0.5M KHC0,-extractable P in a sequential procedure (Gahoonia and Nielsen, 1992). 'Determined by 33P-autoradiography. 'NaHC0,-extractable P.

P-depletion zone (mm)

0.4 0.7 1.9 0.8 1.5 1.0-3.5 1.8-2.3 I .6-3.0 1.7-2.4 4.0 >4.0 2.0

Source Kraus et al., 1987 Hubel and Beck, 1993 Hendriks er al., 1981 Tarafdar and Jungk, 1987 Tarafdar and Jungk. 1987

Gahoonia et a[., 1992a Gahoonia et al., 1992a Gahoonia and Nielsen, 1992 Gahoonia and Nielsen, 1992 Gahoonia and Nielsen. 1992 Bhat and Nye, 1973 Owusu-Bennoah and Wild, 1979

228

P. HINSINGER

The variations in the extension of the rhizosphere largely depend on the physical properties of soil that influence the transfer of ionic and molecular compounds and the resulting shape of their concentration gradients in the soil. Particle size, soil structure, and water content, among others, are critical physical properties of soil that determine the geometry of the rhizosphere (Kuchenbuch and Jungk, 1984; Nye, 1986). Plant roots can alter some of these properties-for example, roots alter soil structure by decreasing soil porosity in the rhizosphere as a consequence of their radial growth (Dexter, 1987; Bruand ef al., 1996). They can also dramatically decrease the water content in the rhizosphere as a direct consequence of water uptake (Hamza and Aylmore, 1992). They may thereby directly influence the extension of the rhizosphere and the dynamics of all nutrients. The spatial extension of the rhizosphere also varies with plant species (Table I). This partly relates to plant species varying in their strategies to further extend their rhizosphere through the development of root hairs (Drew and Nye, 1969) and indirectly through mycorrhizal symbiosis (Clarkson, 1985; Bolan, 1991). Li et al. (1991) showed that P, which supposedly is taken up within several millimetres from the root surface (Table I), can be acquired from a distance of up to 12 cm by the hyphae of mycorrhizal fungi. This distance corresponds to about the maximum spread of external hyphae of a range of vesicular-arbuscular mycorrhizal fungi, as shown by Jakobsen et al. (1992). Plant species also differ in their rooting patterns. Some species develop original features such as proteoid roots, which result in large variations in the surface area of the soil-root interface and in the total volume of rhizosphere soil used to acquire nutrients (Bowen, 1980; Lamont, 1982; Clarkson, 1985; Dinkelaker el al., 1995). Even though such considerations on the geometry of the rhizosphere are critical for a quantitative approach to plant nutrition at the rhizosphere level (Darrah, 1993) and even more so at the whole-plant level (Clarkson, 1985;Barber, 1995),they will not be further discussed here. This chapter concentrates on the qualitative aspects of the chemical processes involved in the acquisition of nutrients from rhizosphere soil.

III. ROOT-INDUCED CHANGES OF IONIC CONCENTRATIONS IN THE RHIZOSPHERE The uptake of water and nutrients, which is the major function of plant roots, results in either the accumulation or depletion of all the ions contained in the soil solution in the rhizosphere. This process occurs both for mineral nutrients and for other, nonessential elements (e.g., Si) and possibly for toxic elements. The nature and intensity of the changes in ionic concentrations depend on the correspondence of the requirements of the plant and the supply by the soil. Nutrients such as Ca or Mg, which usually occur at large concentrations in the soil solution (Fried and

HOW DO PLANT ROOTS ACQUIRE MINERAL NUTRIENTS? 229 Table I1 Estimated Contribution of Root Interception, Mass-Flow, and Diffusion to the Mineral Nutrition of a Field-GrownMaize Yielding 9500-kg Grain per Hectare‘ Process (kg ha-’) Mineral nutrient

Uptake

Interception”

K

40 45 I95

60 15 4

Anions S N P

22 190 40

2

Cations Ca Mg

1

I

Mass flow

I50

Diffusiond

35

0 0 156

65 I50 2

0 38 31

100

“From S. A. Barber, 1995,“Soil Nutrient Bioavailability: AMechanistic Approach,” 2nd. ed. Copyright 1995 by John Wiley and Sons, Inc.; adapted by permission of John Wiley and Sons, Inc. ”The amount of nutrients supplied by root interception was calculated assuming that growing roots would intercept 1% of the available nutrients in the soil. T h e amount of nutrients supplied by mass flow was calculated from measured average soil solution concentrations of nutrients in the bulk soil multiplied by an estimated water consumption of 2500 m3 per hectare. dThe contribution of diffusion to plant uptake was obtained by difference.

Shapiro, 1961),may be transferred by mass flow to the root-soil solution interface at a greater flux than required by the root (Table 11). It is the same for sulphate, among major nutrients that are taken up as anions. These ions may thus accumulate in the rhizosphere, as shown for Ca and Mg by Youssef and Chino (1987) and Lorenz et al. (1994). In calcareous soils the accumulation of Ca can generate calcium carbonate precipitates around roots (Jaillard, 1985), a process that may impair plant growth through lime-induced chlorosis. Similarly, when Ca and sulphate accumulate concurrently in the rhizosphere, precipitation of calcium sulphate (gypsum) can occur (Malzer and Barber, 1975; Jungk, 1996). In saline soils, Na and CI that occur in large concentrations in the soil solution can accumulate in the rhizosphere and reach much greater levels of concentration due to water intake by plant roots (Sinha and Singh, 1974,1976; Marschner, 1995). Such accumulation of Na and CI near the roots can severely impair plant growth and mineral nutrition, even for soils in which salinity would not be considered excessive due to its acceptable level in bulk soil conditions as derived from soil paste measurements. In addition, the resulting increase in osmotic potential (in absolute value) in the rhizosphere (Hamza and Aylmore, 1992) can restrict water uptake by roots, causing species that are not adapted to saline conditions to rapidly wilt and

2 30

P. HINSINGER

ultimately die. Hamza and Aylmore (1992) showed that for high soil-water content, the accumulation of salt does not increase exponentially near plant roots as one might predict because of an important back-diffusion of the solutes to the bulk soil. The processes of accumulating ions and salts in the rhizosphere are expected to be enhanced by an increased water uptake rate, as shown by Sinha and Singh (1974, 1976) for Na and C1. The effect on soil osmotic potential will in turn depress the water transpiration rate so that complex interrelationships between solute accumulation and water uptake will be established (Hamza and Aylmore, 1992; Stirzaker and Passioura, 1996). Conversely, nutrients commonly occurring at low concentrations in the soil solution, such as K and even more so P (Fried and Shapiro, 1961), are transferred by the mass-flow process in amounts insufficient to meet the requirements of the plant (Table 11). Their uptake thus results in a decrease in their concentration in the soil solution near plant roots; this depletion then generates a concentration gradient and diffusion of ions toward the roots (e.g., see Lewis and Quirk, 1967; Farr et al., 1969; Kraus et al., 1987). Claassen and Jungk (1982) have estimated that the K concentration in the soil solution in the rhizosphere of maize may thereby be decreased from several hundred to only 2-3 kmol per litre (Fig. 1). Such a severe decrease in K concentration results not only in the diffusion of K toward the root but also in profound consequences for K dynamics. According to the mass-action law applied to ion exchange, it shifts the equilibrium of adsorption-desorption of K toward an enhanced desorption, leading to a depletion of exchangeable K (Claassen and Jungk, 1982; Kuchenbuch and Jungk, 1982; K in the soil solution (PM) 800 600

400 200

-

0

- - - - - .

0

I

I

- I

2 4 6 Distance from the roots (mm)

i

8

Figure 1 Profile of K concentration in the soil solution as a function of distance from the root surface of 2.5-day-old maize seedlings grown in two different soils. Concentrations of K near roots were decreased to as low as 2-3 )LM(modified from Claassen and Jungk, 1982, with kind permission from Wiley-VCH Verlag GmbH and Professor N. Claassen).

HOW DO PLANT ROOTS ACQUIRE MINERAL NUTRIENTS?

23 1

AK (cmol kg'soil)

nonexch. K

exch. K

I. 0.4 4 0

I

I

I

I

2

4

6

8

1

10

Distance from the roots (mm) Figure 2 Profile of change in exchangeable and nonexchangeable K in the rhizosphere of rape grown for 7 days o n a loess-derived soil. Exchangeable K was depleted up to about 8 mm from the roots. Nonexchangeable K as measured by substracting exchangeable K to HCI-extractable K was mostly depleted within less than 2 mm from the root surface. By comparing the amount ofexchangeable K depleted from the rhizosphere with the amount of K taken up by rape plants, it was shown that total nonexchangeable K contributed a major proportion (71%) of K uptake (modified from Kuchenbuch and Jungk, 1984, with kind permission from Wiley-VCH Verlag GmbH and Professor A. Jungk).

Wehrmann and Coldewey-Zum Eschenhoff, 1986; Niebes et al., 1993) and eventually to a depletion of nonexchangeable K (Kuchenbuch and Jungk, 1984;Jungk and Claassen, 1986), as shown in Fig. 2. This root-induced release of nonexchangeable K contributed up to 80% of the uptake of the plants in soils where the release of nonexchangeable K would have been expected to be negligible when considering the concentration of K in the bulk soil solution (Kuchenbuch and Jungk, 1984; Niebes et al., 1993). Indeed, the release of nonexchangeable K from interlayer sites in phyllosicates, which contribute a large proportion of soil K in many soils, requires the occurrence of very small concentrations of K in the soil solution (Sparks, 1987; Fanning et al., 1989)due to the high affinity of these sites for K. Concentrationsof K as measured in the bulk soil solution of many soils and especially in intensively cultivated soils are usually much above these critical concentrations. Thus, only negligeable amounts of nonexchangeable K are generally expected to be released, even though many field experiments have revealed that nonexchangeable K can contribute to a significant and sometimes major proportion of K removal by crops (Bertsch and Thomas, 1985; QuCmener, 1986; Bosc, 1988). But when taking into account the peculiar conditions occurring in the rhizosphere such as the extremely small K concentration reported by Claassen and

P. HINSINGER

232 Ryegrass

I

0

Rape

.

I 2

ij

Figure 3 X-ray diffraction spectra of mineral material obtained from the rhizosphere of Italian ryegrass and rape as a function of cropping duration. The only source of K for plants was interlayer, nonexchangeable K supplied as phlogopite, which exhibits a typical peak at 1.0 nm. No mineralogical transformation of phlogopite was detectable in the control without plants (data not shown) and in the rhizosphere of ryegrass after 2 days. Conversely, the appearance of a second peak at 1.4 nm after 2-3 days for both species indicated the transformation of phlogopite into a vermiculite clay mineral, which accompanied the release of interlayer K due to K uptake by plant roots. After 16-32 days of cropping, the vermiculitization of phlogopite was almost complete, suggesting that plant roots can induce severe weathering of soil minerals due to chemical interactions occurring in the rhizosphere (modified from Hinsinger and Jaillard, 1993, and Hinsinger etal., 1993, with kind permission from Blackwell Science Ltd.).

Jungk (1982, fig. l), one can explain this apparent contradiction. By decreasing K concentration below the critical value required for K release to occur, plant roots can mobilize significant amounts of nonexchangeable K. This has been clearly established when supplying ryegrass with a K-bearing phyllosilicate as the sole

HOW DO PLANT ROOTS ACQUIRE MINERAL NUTRIENTS? 2 3 3

source of K, exclusively as nonexchangeable, interlayer K, and measuring the uptake of K by ryegrass and the concurrent depletion of solution K in the rhizosphere (Hinsinger and Jaillard, 1993). This root-induced release of interlayer K resulted in a concomitant mineralogical transformation of the K-bearing phyllosilicate in the rhizosphere (Fig. 3) within only a few days of cropping (Hinsinger etal., 1992; Hinsinger and Jaillard, 1993). Similar mineralogical transformations were reported from field plots by comparing bulk soil mineralogy with the mineralogical composition of soil sampled in the rhizosphere of maize roots (Kodama et al., 1994). Plant roots are thus able to induce a substantial weathering of soil minerals as a direct consequenceof the high flux of K uptake and of the subsequent, steep decrease in K concentration occurring in their rhizosphere (Hinsinger and Jaillard, 1993). Similar depletion of exchangeable and nonexchangeable ammonium is very likely to occur given that its chemical behavior is close to that of K. Wehrmann and Coldewey-Zum Eschenhoff (1 986) and Trofymow et al. (1987) evidenced a severe depletion of exchangeable ammonium in the rhizosphere. Mengel el al. (1990) showed the ability of plants to promote the release of fixed, nonexchangeable ammonium in their rhizosphere. Species such as Italian ryegrass and oilseed rape were particularly efficient at depleting nonexchangeable ammonium in the first 2-3 mm of the rhizosphere (Scherer and Ahrens, 1996),which agrees with the peculiar ability of these species to deplete nonexchangeable K (Kuchenbuch and Jungk, 1984; Jungk and Claassen, 1986; Hinsinger et al., 1992; Hinsinger et al., 1993).The sink effect of the absorbing roots, and the consequent shift of the cation exchange equilibria, is the driving force of the root-induced release of nonexchangeable ammonium, as it is for K (Hinsinger and Jaillard, 1993). Fixed ammonium may thereby contribute a major proportion of N nutrition to crops (Mengel and Scherer 1981 ;Keerthisinghe et d., 1985; Baethgen and Alley, 1987).Once again, the peculiar chemical conditions of the rhizosphere lead to reassessing the relative importance of the various processes involved in nutrient dynamics. The depletion of P in the rhizosphere of various species has also been shown in different soils by numerous authors (Lewis and Quirk, 1967; Bhat et al., 1976; Kraus et al., 1987;Steffens, 1987;Gahoonia er al., 1992a;Kirk and Saleque, 1995; Saleque and Kirk, 1995; Hinsinger and Gilkes, 1996) (Fig. 4A). As in soil K, a severe decrease in soil P in the rhizosphere may cause a shift in the adsorption-desorption and dissolution-precipitation equilibria involved in the dynamics of soil P. However, due to the poor reversibility of P sorption onto soil constituents (Barrow, 1983; Parfitt, 1978) and to the low solubility of the various phosphate minerals occurring in soils (Lindsay et d.,1989),very small solution P concentrations must be reached for these phenomena to proceed to a significant extent. Such critical P concentrations might then be too low for sustaining adequate growth of plants. For some species, the external P requirement, i.e., solution P concentration required for near-maximum plant growth, can be as low as 1-5 pM (Asher and Loneragan, 1967;Fohse et al., 1988) or even 10 times lower for P-efficient species

P. HINSINGER

2 34

NaOH-P (pg P g-' soil) 400

A

r m

loo

t

0 '

40

Q

1

I

B

r

30 20 10

0 0

5

10

Distance from roots (mm) Figure 4 Profile of NaOH extractable P(Na0H-P) in the rhizosphere of ryegrass grown on an alumina sand with P supplied (A) as P sorbed onto alumina or (B) as phosphate rock. The profiles obtained in the rhizosphere of ryegrass grown for 14 days with either alumina P or phosphate rock P as the sole source of P are compared with profiles obtained in the absence of plants (control). A distinct depletion of NaOH-P is detectable up to 2 mm from the root surface when P is supplied as P sorbed onto alumina (A). Conversely, when Pis supplied as phosphate rock (B),an accumulation of NaOH-P occurs in the rhizosphere of ryegrass, which is maximal at about 1-1.5 mrn from the roots. This increase of NaOH-P indicates a root-induced dissolution of phosphate rock occurring at a rate faster than that of P uptake (modified from Hinsinger and Gilkes, 1996, with kind permission from Blackwell Science Ltd.).

such as perennial ryegrass (Breeze ef al., 1984).Other species, including such vegetable crops as lettuce, tomato, and potato, require much higher P concentrations for achieving maximum growth (Asher and Loneragan, 1967; Fox, 1981). Thus,

HOW DO PLANT ROOTS ACQUIRE MINERAL NUTRIENTS?

23 5

plant species vary widely in their ability to cope with low soil-Pconcentrations. In addition to such considerations, the rate of desorption of soil P or the rate of dissolution of P-bearing soil constituents would need to be larger than fluxes of P uptake to prevent any growth restriction. As pointed out by Darrah (1993), we still lack knowledge about the kinetics of those reactions involved in P dynamics, so the amount of P mobilized by plant roots from the rhizosphere is hardly predictable. Surprisingly, though, some studies suggest that P may sometimes accumulate in the rhizosphere rather than be depleted (Brewster et al., 1976; Hiibel and Beck, 1993; Hinsinger and Gilkes, 1995; Saleque and Kirk, 1995; Hinsinger and Gilkes, 1996, 1997) (Fig. 4B). Ruiz ( I 992) and Ruiz and Arvieu ( 1 992) calculated a fourfold increase in solution P concentration at the root-mineral interface of oilseed rape seedlings supplied with hydroxyapatite as the sole source of P. They concluded that the roots induced a dissolution of hydroxyapatite at a faster rate than that of P uptake, leading to the relative accumulation of P in the rhizosphere. They suggested that roots acted either through proton excretion or Ca uptake, as can be deduced from the following dissolution equation of hydroxyapatite: Ca,(PO,),OH

+ 7 H,Of

-

5 Ca2+

+ 3 H,PO, + 8 H,O

(1)

Similar conclusions were drawn by Hinsinger and Gilkes (1995, 1996, 1997) to explain the increase in NaOH-extractable P and concomitant dissolution of phosphate rock (carbonate apatite) that occurred in the rhizosphere of various species (Fig. 4B). They showed for ryegrass that proton excretion was likely to be the driving force of this root-induced dissolution of phosphate rock occumng at a faster rate than that of P uptake. Kirk and Saleque (1995) calculated that, whereas total soil P was depleted to about 50% of its initial value in the vicinity of rice roots growing in a flooded soil at various rates of P fertilization, a 5-20-fold increase in solution P concentration occurred at 2 4 mm from the root surface. These results agree with a model proposed by Nye (1983) for the diffusion of two interacting solutes in the rhizosphere. In this model-for the diffusion of protons away from the roots and the diffusion of P dissolved from a Ca-phosphate source toward the roots (due to differential coefficients of diffusion of the two solutes)-Pcan be expected to be depleted in the immediate vicinity of plant roots and to accumulate away from the roots, This would occur in the soil zone where dissolution due to proton excretion is faster than the diffusion of P toward the roots due to P uptake. It would result in some dissolved P diffusing away from the roots rather than toward the roots (Nye, 1983; Kirk and Saleque, 1995). The profile of isotopically exchangeable P obtained by Hiibel and Beck (1993) in the rhizosphere of maize also agrees with this model (Fig. 5). Hubel and Beck attributed the accumulation of P peaking at about 0.5 mm away from the roots to the mobilization of some organic soil P due to phosphatase excreted by plant roots or to an artifact (P contained in the root hairs). For mineral nutrients of which dynamics are not simply the re-

P. HINSINGER

236

lsotoplcally exch. P (mg dm") 80 60

1

40 20 0

1 0

1

2

3

Distance from the root (mm) Figure 5 Profile of isotopically exchangeable P around a primary root of maize grown for 4 days in 33P-phosphate-labeled soil. Isotopically exchangeable P peaks at about 0.5 m m from the roots, indicating that both depletion near plant roots and accumulation further away from the roots combine in the rhizosphere. The accumulation of isotopically exchangeable P can be interpreted either as an artifact (P included in root hairs, which extend up to about 0.5 mm) or as the result of root-induced mobilization of poorly mobile P, such as organic P due to phosphatase excretion (modified from Hiibel and Beck, 1993, with kind permission tiom Kluwer Academic Publishers).

sult of an ion exchange reaction between soil solution and one single pool of soil solid constituent and subsequent diffusion toward the root, multiple processes can interact and render difficult the task of predicting rhizosphere conditions and consequent plant uptake. Nevertheless, these studies suggest that the uptake of mineral nutrients not only is the ultimate stage of the acquisition process at the root-soil interface but also results itself in severe changes in ionic concentrations that can then shift the equilibria of adsorption-desorption or dissolution-precipitation involved in the dynamics of nutrients in the soil. The uptake may thus be regarded as a crucial and active (in the broad sense) stage of the acquisition process. However, when considering the depletion of a nutrient in the rhizosphere, estimating its actual benefit for the plant is still difficult. Although nutrient depletion may result in the mobilization of less spatially or chemically accessible forms of the depleted nutrient as shown for K, it should result as well in a decrease in the uptake rate according to the Michaelis-Menten relationship between uptake flux and external concentration (Epstein and Hagen, 1952). This holds true unless concentration levels reached in the rhizosphere are maintained above the level required for maximum growth (external requirement). Because of the lack of quantification of the different processes involved here, it is difficult to state if these phenomena may be re-

HOW DO PLANT ROOTS ACQUIRE MINERAL. NUTRIENTS?

237

garded as efficient strategies of mineral nutrition (Darrah, 1993), although some work has clearly shown for P that the plant species that exhibit the lower external Prequirements are among the most efficient in acquiring soil P (Fohse e l al., 1988). Undoubtfully these root-induced changes of ionic concentrations deserve further consideration as major components of the whole process of nutrient acquisition by higher plants.

N.ROOT-INDUCED CHANGES OF RHIZOSPHERE pH The uptake of nutrients can also affect soil pH. Indeed, as nutrients are taken up as ions, the differential rates of uptake of cations and anions result in an imbalance of positive and negative charges entering the root cells. When an excess of cations over anions are taken up, the plant root compensates by releasing excess positive charges as protons, thereby resulting in acidification of the rhizosphere (Nye, 1981; Romheld, 1986; Haynes, 1990). Conversely, when an excess of anions over cations are taken up by the root, the excess negative charges are released as hydroxyls or bicarbonate ions, leading to alkalinization of the rhizosphere. The corresponding changes in pH are the most documented of the root-mediated chemical changes occurring in the rhizosphere. This is partly due to the development of simple methods for measuring rhizosphere pH, notably, the dye indicator-agar technique (Marschner et al., 1982). Since N is the most demanded mineral nutrient for numerous plants (Mengel and Kirkby, 1987; Marschner, 1995) and since it can be taken up by the plant as an anion (as NO;), as a cation (as NH:), or as a molecule (as NJ, it has often been reported to play a major role in the overall cation-anion balance of the plant (Nye, 1981; Romheld, 1986). Indeed, many studies have shown that plants supplied with ammonium acidify their rhizosphere, whereas they alkalinize it when supplied with nitrate (Riley and Barber, 1971; Jarvis and Robson, 1983; Marschner and Romheld, 1983; Weinberger and Yee, 1984; Romheld, 1986; Gahoonia et al., 1992a). In addition, legumes relying on symbiotic N, fixation have been shown to acidify their rhizosphere to compensate for an excess of cations being taken up (Jarvis and Robson, 1983; Romheld, 1986). This is, however, an oversimplified picture of the process, since nitrate-fed plants such as oilseed rape, chickpea, or lupins have been repeatedly shown to excrete protons and acidify their rhizosphere (Grinsted et al., 1982; Marschner and Romheld, 1983; Loss et al., 1993; Hinsinger and Gilkes, 1995, 1996). Some studies have also shown that the various parts of a single root exposed to identical, external conditions can behave differently, leading to localized processes of acidification and alkalinization (Weisenseel er al., 1979; Marschner et al., 1982; Marschner and Rornheld 1983; Haussling et al., 1985; Jaillard et al., 1996) (Fig. 6).

238

P. HINSINGER

Figure 6 Map of pH values as obtained around the roots of a 7-day-old seedling of maize according to the videodensitometry technique of Jaillard er al. (1996).The image was obtained 120 min after embedding the roots in an agarose sheet containing bromocresol green as a pH dye indicator and KNO, I mM, which had been adjusted at an initial pH of 4.60 by adding HCI. The image was acquired with a scanning video camera and was then computed using image-analysis software. The pH map shows that various parts of the roots behave differently. While the apical region is excreting hydroxyl equivalents, resulting in an increased rhizosphere pH, the basal parts of the roots are excreting protons, especially in the zones of emergence and elongation of laterals, at 2&30 and 80-100 mm respectively, from the root tip (modified from Jaillard er aL, 1996. with kind permission from Kluwer Academic Publishers).

It is now largely accepted that pH changes in the rhizosphere essentially originate from the imbalance of anions and cations taken up by plants (Nye, 1981; Haynes, 1990). Compared with the corresponding excretion of protons, the contribution to rhizosphere acidification of other processes, such as root respiration and organic acid exudation, has not been much studied. According to Nye (1986), the respired CO, may contribute a significant proportion of rhizosphere acidification only in alkaline and calcareous soil conditions and/or when its diffusion is impaired (as in waterlogging, for instance). The abundance of calcareous soils in temperate regions of the world suggests that the contribution of this phenomenon to rhizosphere acidification would require more thorough investigations. The exudation of organic acids has occasionally been reported to contribute to pH changes around roots of P-deficient seedlings of oilseed rape (e.g., Hoffland, 1992). Petersen and Bottger (1991) estimated that organic acids excreted by maize roots contributed to less than 0.3% of rhizosphere acidification. Bearing in mind that the common organic acids that can be excreted in the rhizosphere are dissociated in the pH conditions of the cytoplasm (Hedley et al., 1982a; Nye, 1986; Haynes, 1990; Jones and Darrah, 1994), they should thus be released as organic anions and not be regarded as responsible per se for an acidification of the rhizosphere. Nevertheless, their release should be taken into account in the overall balance of cations and anions crossing the plasmalemma (e.g., Dinkelaker er af.,1989), which finally determines the net excretion of protons or hydroxyl equivalents. Whatever

HOW DO PLANT ROOTS ACQUIRE MNERAL NUTRIENTS?

239

the origin of pH changes, modifications of up to 1-2 pH units have been commonly reported in the rhizosphere of diverse species ( e g , see Riley and Barber, 1971; Marschner and Romheld, 1983). Soil pH is known to be a critical factor influencing many chemical reactions in the soil environment (Mengel and Kirkby, 1987). For instance, the dynamics of various forms of inorganic Pare strongly pH-dependent, with dissolution of P from crystalline and sorption complexes and speciation of P in solution being strongly dependent on the pH of soil solution (Murrmann and Peech, 1969; Barrow, 1984; Lindsay et al., 1989). In soils of moderate to high pH, some P reacts with Ca ions to form various sparingly soluble calcium phosphates such as octocalcium phosphate or hydroxyapatite (Arvieu, 1980; Freeman and Rowell, 198l), which require a supply of protons to dissolve and release P (see Eq. 1). According to the mass action law, the excretion of protons by plant roots should shift this equilibrium reaction to the right, thereby enhancing the dissolution of hydroxyapatite (Khasawneh and Doll, 1978; Kirk and Nye, 1986). Indeed, the ability of some species, such as buckwheat, oilseed rape, and various legumes, to utilize P when supplied as a phosphate rock (i.e., a carbonate apatite that also requires protons to dissolve) is related to their capacity to excrete protons (Aguilar and van Diest, 1981; Bekele et al., 1983; Ruiz, 1992; Hinsinger and Gilkes, 1995 and 1997). Riley and Barber (1971) and Gahoonia et al. (1992a,b) showed that soybean and ryegrass fed with ammonium were more efficient for mobilizing soil P in some soils than when fed with nitrate. Similar conclusions were drawn by Hinsinger and Gilkes (1996) for ryegrass supplied with a phosphate rock as the sole source of P. These studies thus suggest that proton excretion occurring especially when N is supplied as ammonium might improve Pnutrition by enhancing the dissolution of some forms of inorganic P, most likely Ca-bound P, in the rhizosphere. In flooded soils where adapted plants, such as lowland rice, are expected to rely solely on ammonium (because nitrate is reduced to ammonium as a result of the ambient reducing conditions), root-induced solubilization of acid-soluble soil phosphates has been shown to contribute a substantial proportion of P uptake (Kirk and Saleque, 1995; Saleque and Kirk, 1995). These authors showed, however, that in this particular case the rootinduced dissolution of soil P was only partly due to proton excretion by plant roots (see Section V). In addition, a stimulation of proton excretion has been reported for P-deficient species such as oilseed rape (Grinsted er al., 1982; Moorby et al., 1988; Ruiz, 1992). In this respect proton excretion by plant roots may thus be regarded as an adaptative strategy for P acquisition. As for P, the concentration of Fe in soil solution is severely decreased when pH increases, reaching a minimum for pH ranging from 7.4 to 8.5, due to the pH-dependent solubility of iron oxyhydroxides (Lindsay, 1974; Lindsay, 1979; Schwertmann, 1991). Considering the solubility diagram of these Fe-bearing minerals in oxidizing conditions (Lindsay, 1974), the activity of total soluble Fe in soil solution decreases from lO-'Mat pH 3.5 down to lo-" Mfor pH 8.5.Aroot-induced

240

P. HINSINGER

decrease in rhizosphere pH would thus increase the activity of Fe in the soil solution by up to several orders of magnitude. Indeed, Oertli and Opoku (1974) showed that an enhanced proton excretion by maize roots, as obtained in response to a large K supply and consequent excess of uptake of cations over anions, resulted in an improved mobilization of Fe from a synthetic ferric hydroxide. In addition, many species have been shown to respond to Fe-deficiency by acidifying their rhizosphere (Romheld et al., 1984; Marschner et al., 1986, 1989). Nevertheless, in the pH range commonly found in soils, Fe activity is always below M (see preceding discussion), which is the value required for many plants to meet their Fe requirements. Such a value is attained only for soil pH of about 3 (Lindsay, 1974). The extent to which proton excretion is capable of supplying a sufficient amount of Fe to roots for adequate plant growth is thus questionable unless very high fluxes of proton excretion occur at the soil-root interface. Considerable proton effluxes have indeed been encountered in the rhizosphere. Romheld et al. ( 1 984) reported that the roots of Fe-deficient sunflower excreted an average of about 5.6 pmol H+ hour-' g-' fresh weight of root, whereas Fe-adequate plants released small amounts of hydroxyl equivalents. They found that locally the proton efflux could be as high as 28 pmol H+ hour-' g-' fresh weight of root, especially near apical root zones, which were the preferred sites of excretion. Proton effluxes of the same order of magnitude have been reported for roots of oilseed rape (Jaillard, 1987; Ruiz, 1992). Jaillard (1985, 1987) has shown that plant roots were able to grow in compacted, highly calcareous soils by dissolving the surrounding calcium carbonate due to a large flux of Ca uptake and to a consequent, large proton efflux. The dissolved calcium carbonate (calcite) was shown to reprecipitate subsequently into the vacuole of root cells (Jaillard et af., 1991) to form calcified roots, which can constitute up to 25%of the total calcium carbonate present in some calcareous soils under natural grasslands (Jaillard, 1984). Jaillard (1987) showed in short-term experiments in controlled conditions that living roots of oilseed rape were able to precipitate calcite into their cells within only a few hours. Hinsinger et af. (1993) have shown that oilseed rape was also able to mobilize nonexchangeable Mg by dissolving a Mg-bearing phyllosilicate as a result of the severe pH decrease that its roots induced in the rhizosphere (Table 111). Conversely, ryegrass grown in identical conditions proved unable to mobilize any significant amount of Mg (Hinsinger and Jaillard, 1993), as a consequence of the high pH that it maintained in its rhizosphere (Table 111). The preceding examples address the case of acidification of the rhizosphere as a profitable strategy for acquiring mineral nutrients. However, alkalinization of the rhizosphere is likely to be as widespread as acidification, or even more so, as inferred by Nye ( I 986), considering that nitrate is the prominent source of N for nonlegume plants in most field conditions. In addition, the excretion of protons should not be regarded as a universal solution to the problems encountered by plants while acquiring nutrients. Rhizosphere acidification can have detrimental effects on root

HOW DO PLANT ROOTS ACQUIRE MINERAL NUTRIENTS?

241

Table 111 Rhizosphere pH and Mobilizationof Mg (expressed as g kg-' of applied Mg) as a Function of Time in the Cropping ExperimenLa ~

Time (days) Rhizosphere pH Amounts of Mg mobilized

0 rape ryegrass rape ryegrass

7.03 7.03 -

-

4

6.56 7.22 -1 -1

8

16

32

5.55

4.22 7.48 15 2

4.27 7.69 21 3

8.09 3 0

UMgwas supplied as a Mg-bearing phyllosilicate (phlogopite). The Mg cations contained in this mineral constitute the silicate framework and are thus nonexchangeable. Their release requires a dissolution of the phyllosilicate, which can occur in acid conditions such as those encountered in the rhizosphere of rape after 1 6 3 2 days of cropping.

growth and mineral nutrition when soil pH is eventually decreased to very low values. Gahoonia ( 1993) showed, for instance, that ryegrass fed with ammonium decreased its rhizosphere pH to 4.4, which is a pH value prone to aluminium toxicity (Kinraide, 1991). Indeed, Gahoonia (1993) measured a concurrent increase in extractable Al in the rhizosphere of ryegrass. Thus, it is clear that plant roots should not always acidify their rhizosphere and particularly not when growing in already acid soils so as to prevent increased risks of aluminium, manganese, or even proton toxicity (Marschner, 1995). Some researchers have shown that species growing naturally in very acid soils, such as Norway spruce for instance, rather alkalinize their rhizosphere (Marschner et al., 1991). Youssef and Chino (1989) have shown that some plant species can increase rhizosphere pH in acid conditions and decrease it in neutral or alkaline conditions, revealing the capacity of plants to adapt to adverse soil conditions. Another major limitation of acid soils for plant growth is related to deficiencies in various mineral nutrients, especially P deficiency (Marschner, 1995). In acid soils, phosphate ions are indeed strongly sorbed on various soil minerals and especially Fe- and Al-oxyhydroxides; the charge of these minerals being pH dependent. Since the desorption of phosphate ions from these minerals involves ligand exchange (Parfitt, 1978), hydroxyls or bicarbonate ions excreted by plant roots may desorb some phosphate and render it available to the plant, as suggested by Gahoonia et al. (1 992a). Gahoonia et al. (1992a) showed that when increasing their rhizosphere pH, roots of ryegrass fed with nitrate were more efficient at desorbing P in an Fe-oxyhydroxide-rich soil (oxisol) than when fed with ammonium. Improved P nutrition can thus result from either root-induced decrease or an increase in rhizosphere pH, depending on the dominant forms of inorganic P present in the soil (Gahoonia et al., 1992a). Whether root-mediated pH changes of the rhizosphere should be regarded as an adaptative strategy of nutrient acquisition or not, there is no doubt that the actual

2 42

P. HINSINGER

pH in the rhizosphere should be taken into account rather than the pH of the bulk soil when considering nutrient dynamics. Furthermore, the actual pH may be a poor indicator of the real effect exerted by the root on its environment. Considering that the activity of protons and hydroxyls is influenced by the total solute content of the rhizosphere solution and that a large part of protons or hydroxyl equivalents produced by the roots may be consumed in diverse adsorption-desorption or dissolution-precipitation reactions with soil minerals, the consequences of the excretion activity of the roots may extend much beyond what is indicated by a direct measurement of the resulting pH in the rhizosphere. For instance, for nitratefed ryegrass and subterranean clover grown in an artificial soil with phosphate rock (carbonate apatite) as the sole source of P,Hinsinger and Gilkes (1996) found that up to 20-25% of the applied phosphate rock dissolved in the rhizosphere, whereas the pH decreased only minimally (
V. ROOT-INDUCED CHANGES OF REDOX CONDITIONS IN THE RHIZOSPHERE The redox conditions of a soil, which can be described by either the redox potential (Eh) or the pe (negative log of the activity of electrons; with Eh (mV) = 59.2 pe), is an environmental parameter of critical importance for the dynamics of those elements that can occur at different oxidation states in soils. Among mineral nutrients, this is particularly the case for Fe and Mn. As previously pointed out, because of the low solubility of Fe-bearing minerals such as iron oxyhydroxides and iron oxides, the activity of soluble Fe species in soils is commonly much below the activity required for adequate plant growth (Lindsay, 1974). Besides its strong dependency on pH, the solubility of iron oxyhydroxides and iron oxides is very much dependent on redox conditions. When oxidant conditions are prevailing, which is the case in most soils as long as the

HOW DO PLANT ROOTS ACQUIRE MINERAL NUTRIENTS?

243

transfer of gases is not impaired and PO, is about atmospheric PO,, the activity of Fe2+ and other Fe" species is less than the activity of Fe3+and other Fe"' species at the usual pH values found in the soils, except for alkaline soils (Lindsay, 1979). However, when reduction occurs and pe decreases, Fe" species rapidly become dominant over Fe"' species; the ratio Fe2+-Fe3+ increases 10-fold for a decrease in pe by one unit (Lindsay, 1979). The reduction of iron-hydroxide can be described by the following equation (Lindsay, 1979): Fe(OH),

+ 3 H30+ + e-

-

Fe2+

+ 6 H,O,

(2)

from which the following can be deduced: log(Fe2+) = 15.74 - (pe + pH) - 2 pH

(3)

In oxidant conditions, when PO, equals atmospheric PO,, pe + pH = 20.61; whereas it can decrease to as low as 2 when reducing conditions are prevailing (Linday, 1979). Under such circumstances the activity of Fe" species can increase by several orders of magnitude. Reduction processes can thus be very efficient for enabling higher plants to meet their Fe nutritional demand. Root respiration consumes 0, and may thus decrease PO, and pe, thereby increasing the activity of Fe in the soil solution. However, PO, must be decreased to very low values before a significant increase occurs in the solubility of iron oxyhydroxides and iron oxides. This can occur only when soil physical conditions are impairing the transfer of gasses to a great extent, i.e., in anoxic conditions that are altogether unfavorable for adequate plant growth (discussed later). The reducing activity of plant roots can also be attributed to the release of reducing compounds such as phenolics (caffeic acid) or aliphatic acids (malic acid) in the rhizosphere (Brown and Ambler, 1973; Romheld and Marschner, 1983). Although this process has been suspected to be of ecological importance, it has been shown to be too restricted to account for the reducing capacity of most plant species (Romheld and Marschner, 1983; Bienfait et al., 1983). For bean, for instance, Bienfait et al. (1983) found that the release of reductant in the rhizosphere contributed about 14% of the total amount of Fe reduced by roots of Fe-sufficient plants and less than 2% for Fe-deficient plants. They also showed that the release of reductant less than doubled for Fe-deficient compared with Fe-sufficient bean plants, whereas the reducing activity of the roots increased 15-fold. This reducing activity, which was found to be located at the root surface rather than being released into the rhizosphere, was attributed to a plasmalemma-bound reductase sytem (Chaney et al., 1972; Bienfait et al., 1983; Romheld and Marschner, 1983). This suggests that the corresponding root-induced change in redox potential is likely to be spatially restricted to the root-soil interface and not to extend into the rhizosphere. The efficiency of such an enzymatic process for Fe reduction in calcareous soils can, however, be questionned: it is certainly of little significance if not accompanied by a concomitant acidification of the rhizosphere, since this re-

P. HINSINGER

244

a b l e IV Effect of Fe Deficiency on Rhizwphere pH and on the Amount of Fe Reduction and Fe Uptake by a Chlorosis-ResistantCultivar of Chickpep

Fe status

Rhizosphere pH

Fe reduction (nmol Fe g-’ root fresh wt. hour-’)

Fe-sufficient Fe-deficient

6.0 3.9

200 1100

Fe uptake (nmol Fe g - ’ root fresh wt. hour-’) 0.39 7.58

“Modified from Marschner, 1990.

ductase system has been found to be inhibited at pH above 7 (Marschner ef al., 1989). Nevertheless, numerous species respond to Fe deficiency both by an enhanced proton excretion and by an increase in the reducing capabilities of their roots, showing a remarkable example of “cooperative strategy” (Romheld and Marschner, 1983). This is illustrated for chickpea in Table IV. The stimulation by Fe deficiency of the reducing activity andor proton excretion by plant roots has been described as “strategy I” for Fe acquisition by nongraminaceous species (Marschner et al., 1986). The reduction of Fell1to Fe” as related to the reducing activity of the roots is thus a major adaptative strategy of dicotyledonous species and nongrass monocotyledonous species for coping with the poor solubility of iron oxyhydroxides and iron oxides in soils (Chaney ef al., 1972; Marschner et al., 1989). The reducing activity of roots is likely to play an equally important role in Mn nutrition: The reduction of MnO, has been observed in the rhizosphere of different species (Godo and Reisenauer, 1980; Gardner et al., 1982). Uren (1981) has clearly established with axenic-grown sunflower plants that the roots were themselves directly responsible for the reduction of manganic Mn, even though the mechanism has not been fully elucidated. When anoxic, reducing ambient conditions are prevailing in the bulk soil, as in waterlogged soils, the activity of Fe” species can dramatically increase, as deduced from Eq. (2). Ferrous Fe, and also manganous Mn and sulfides, can thereby reach phytotoxic levels that, together with the lack of 0,, may preclude plant growth (Drew, 1988). To cope with such conditions, the roots of rice and other plants naturally growing in submerged areas have developed an anatomical adaptation in their cortical tissue called aerenchyma: The translocation of 0, from the shoots to the roots through the aerenchyma not only allows root cells to respire but also results in a leakage of excess 0, in the rhizosphere (Armstrong, 1967; Ando et al., 1983). This excess of 0, released by rice roots then diffuses in the rhizosphere, leading to a significant increase in Eh (Fig. 7) up to a few millimetres away from the roots (Trolldenier, 1988; Flessa and Fischer, 1992). The diffusion of 0,

HOW DO PLANT ROOTS ACQUIRE MINERAL, NUTRIENTS?

245

Eh (mV)

Submerged soil

-200 -400

1 0

-iment I

I

5

10

Distance from the root (mm) Figure 7 Profile of redox potential in the rhizosphere of 5-week-old rice grown in a submerged soil and in river sediment. Redox potential as measured by redox microelectrodes decreased significantly due to oxygen leakage from the primary root of rice, leading to a root-induced oxidation of the rhizosphere detectable up to 1-3 mm from the roots (modified from Flessa and Fischer, 1992, with kind permission from KIuwer Academic Publishers).

in the rhizosphere of roots of lowland rice thereafter results in an oxidation of ferrous Fe, as revealed by the decrease of Fe" concentrationand concomitant increase of Fe"' concentration measured within a few millimeters from the root surface (Fig. 8) in a flooded soil (Begg et al., 1994; Saleque and Kirk, 1995). A further consequence of this oxidizing effect of rice roots is the precipitation of ferric Fe as iron oxyhydroxides (Chen er al., 1980), which contributes the redbrownish discolorationthat is clearly visible at the surface of rice roots after flooding of the soil. Chen et al. (1980) identified iron coatings around rice roots as being mainly formed of goethite and lepidocrocite. Iron coatings have been reported around roots of other plant species growing in submerged soils, e g , slash pine (Fisher and Stone, 1991). The deposition of iron oxyhydroxide precipitates has also been described to occur inside the root cortex in cell walls and intercellular spaces of rice (Green and Etherington, 1977). The increase in redox potential in the rhizosphere and subsequent immobilizationof ferrous Fe by oxidation and precipitation as iron oxyhydroxide by plant roots can prevent the uptake of toxic amounts of ferrous Fe (Green and Etherington, 1977). In addition, the oxidation of Fe" with 0, produces 2 mol of protons per mole of Fe, according to Eq. (3) and the following equation (Ahmad and Nye, 1990): 4 F e 2 + + 0 , + 18H20*4Fe(OH),+8H,0+

(4)

The root-induced oxidation of ferrous Fe is thus expected to result in an acidification of the rhizosphere. This has been shown by Begg et al. (1994) and Saleque and Kirk (1995) for rice grown in a flooded soil (Fig. 8). These researchers showed

I? HINSINGER

2 46

Fe (mmol kg-' soil) 250

PH

r7

.* * *

t6

200 150

0

0

5 10 15 Distance from the roots (mm)

t2 20

Figure 8 Profile of Fell and Fell' concentration and of pH in the rhizosphere of 2 I -day-old lowland rice grown in a flooded soil for 10 days. The Fe" concentration decreased up to 4 mm from the roots, whereas Fe"l concentration steeply increased near rice roots, indicating that root-induced oxidation of Fe occurred due to oxygen leakage into the rhizosphere. The larger extent of the depletion zone of Fe" relative to the spread of accumulation of Fe"' is related to the larger mobility of Fe" compared with Fe"'. Simultaneously, a steep decrease in soil pH was encountered, which is partly attrihuted to soil acidification resulting from the oxidation of Fe" according to Eq. (4) and to protons excreted by rice roots to compensate for an excess of cation uptake (modified from Begg ef al., 1994).

that part of the strong decrease in pH that they measured in the rhizosphere of rice was also due to proton excretion by rice roots that were relying exclusively on ammonium as the soil-N source considering the reducing conditions of the bulk soil. Kirk and Saleque (1995) and Saleque and Kirk (1995) also showed that the acidification of the rhizosphere of lowland rice was responsible for the solubilization of substantial amounts of soil P. In other words, root-induced changes in redox potential not only influence the dynamics of elements with varying states of oxidation such as Fe and Mn, but due to the concomitant change of pH that they impose on the rhizosphere they can alter the availability of other nutrients, as evidenced here for P (Kirk and Saleque, 1995; Saleque and Kirk, 1995). Nevertheless, the major benefit of root-induced oxidation to the plant is the detoxification of the root environment through a decrease in concentration of ferrous Fe and possibly manganous Mn as well (Marschner, 1995). The increase in Eh occurring in the rhizosphere of lowland rice in waterlogged conditions and the consequent precipitation of iron oxyhydroxides around the roots is probably the best-known evidence of an extensive change in redox condi-

HOW DO PLANT ROOTS ACQUIRE MINERAL NUTRIENTS?

247

tions as induced by plant roots. In a similar way manganese oxides have been reported to sometimes precipitate in the rhizosphere of lowland rice (Bacha and Hossner, 1977).

VI. ROOT-INDUCED COMPLEXATION OF METALS IN THE RHIZOSPHERE The release of organic compounds by plant roots is a well-known phenomenon that is at the origin of the stimulation of soil microflora in the immediate environment of the roots; i.e., the “rhizosphere effect” that was first defined as such by Hiltner in 1904. It is now largely accepted that up to 20-30% of the total C assimilated by higher plants is released in the rhizosphere as diverse exudates, including respired CO, (Merckx et al., 1986a; Helal and Sauerbeck, 1989). These exudates, however, remain difficult to quantify because of their rapid microbial degradation (Uren and Reisenauer, 1988), and because similar metabolites are released by soil microorganisms. Nevertheless, some root exudates exhibit complexing or chelating properties with respect to metallic ions. Among those, numerous organic anions, which are the conjugated bases of organic acids, have been reported to play a role in the mobilization or immobilization of mineral nutrients or undesirable elements, as related to their complexing properties (Jones and Darrah, 1994). In addition, researchers have shown that the exudation of organic acids by plant roots increases as a response to nutrient deficiencies (Kraffczyk et al., 1984; Lipton et al., 1987; Jones and Darrah, 1995), which might suggest their possible implication in the acquisition of mineral nutrients from the rhizosphere. Citrate, for instance, has been shown to be released in considerable amounts by proteoid roots of white lupin (Gardner et al., 1982; Dinkelaker et al.,1989; Gerke et al., 1994) and Proteaceae of the Banksia genus (Grierson, 1992; Dinkelaker et al., 1995). Gardner et al. ( 1983)proposed that citrate excreted by the proteoid roots of white lupin may play a major role in the dissolution of iron phosphates, through the formation of a ferric hydroxy-phosphate complex. These researchers suggested that this complex might be of prime importance for the acquisition of P and Fe by white lupin roots. This mechanism may partly explain the peculiar ability of white lupin and presumably of most native Proteaceae species to cope with the very poor P status of many Australian soils (Bowen, 1980). Furthermore, the enhanced formation of proteoid roots in P-deficient lupin (Marschner et d., 1986) and the enhanced exudation of citrate in the rhizosphere of alfalfa and rape as a response to P deficiency (Lipton et al., 1987; Hoffland et al., 1989, 1992) suggest that they may be considered as adaptative strategies for the acquisition of nutrients such as P. Indeed, the solubility of soil P has been shown to be influenced by

2 48

P. HINSINGER

organic anions such as citrate (Jones and Darrah, 1994). Staunton and Leprince ( 1996) showed that, compared with acetate, tartrate, salicylate, and oxalate, citrate was the most efficient organic anion for increasing the proportion of phosphate in soil solution. They showed that solution phosphate increased by a factor of two to three for citrate concentrations ranging from 0.1 to 1 mM (Staunton and Leprince, 1994). Gerke (1994) reported a 20-fold increase in phosphate desorption from a soil on addition of 50 pnol citrate g-I soil, which corresponds to the tremendous concentration of citrate that Dinkelaker et al. (1989) found in the rhizosphere of proteoid roots of white lupin. While Parfitt (1979) and Grimal et al. (1995) privileged the hypothesis that organic anions excreted by ryegrass and maize may be involved in ligand exchange reaction with phosphate ions sorbed onto iron oxyhydroxide (goethite) surfaces, Bolan et al. (1994) proposed that the major effect of organic acids involved in the release of soil P was related to A1 being complexed and to the subsequent solubilization of P-A1 compounds. Indeed, they showed that addition to soil of various organic acids commonly found in the rhizosphere resulted in decreased P sorption and that organic acids extracted more soil P according to their ability to form stable complexes with Al (log K,,). Among the range of organic acids investigated, oxalic and citric acids had the highest log K,, and had the largest effect on P uptake and plant growth of ryegrass (Bolan et al., 1994). Ae et al. (1990) proposed a similar mechanism to explain the peculiar ability of pigeon pea to take up P in a P-deficient alfisol from India for which a major proportion of soil P was Fe-bound P. They suggested that the roots of pigeon pea were more efficient than the roots of other species due to excretion of organic acids that complexed Fe and resulted in releasing Fe-bound P. Nevertheless, they found that pigeon pea had less citrate, malate, malonate, and succinate in its root exudates than other less P-efficient species such as soybean. Piscidic acid and its derivatives were identified as the peculiar root exudates of pigeon pea that explained its ability to use Fe-bound P, which was found to be almost unavailable to the other crops studied (Ae et al., 1990). The complexation of diverse micronutrients such as Co, Cu, Mn, and Zn (Merckx et al., 1986b; Mench et al., 1987) and undesirable heavy metals such as Cd (Mench and Martin, 1991) and Pb (Mench et al., 1987) has been shown to occur in the rhizosphere as a consequence of root exudation. Even tnough the root exudates directly responsible for the complexing of these metals have not been identified, speculations about the role of simple organic acids are supported by the results of Mench and Martin (199 1). In addition, Gardner et al. ( 1982) and Dinkelaker et al. ( 1 989) found an increase in amounts of available micronutrients such as Fe, Mn, and Zn in the rhizosphere sampled near proteoid roots of white lupin, which were also evidenced as root zones responsible for intense excretion of citrate. Dinkelaker et al. (1989) estimated that a considerable amount of citrate was excreted per plant, i.e., about 5.5 mmol per plant, which represented about 23% of

HOW DO PLANT ROOTS ACQUIRE MINERAL NUTRIENTS? 249 Table V Amounts of Citrate Excreted and of DTPA-ExtractableFe, Mn, and Zn in the Rhizosphere of White Lupin Relative to Bulk Soil (mean value % standard deviationpb

Bulk soil Rhizosphere soil

Citrate (kmol g-I soil)

Fe (pmol kg-’ soil)

Mn (pmol kg-I soil)

Zn (pmol kg-’ soil)

47.7 2 7.2

34 ? 6 251 ? 43

4428 222 ? 23

2.8 ? 0.4 16.8 -t 2.4

“Modified from Dinkelaker et a/.. 1989, with kind permission from Blackwell Science Ltd. hRhizosphere soil was sampled around proteoid rootlets of white lupin grown in a calcareous soil for 13 weeks. DTPA (diethylen-triamin-penta acetic acid) extraction gives an estimate of available Fe, Mn, and Zn. “Not detectable (i.e. < 0.05 pmol g - ’ soil).

the dry weight of the whole plant. Citrate was excreted mostly around proteoid rootlets where DTPA-extractable Fe, Mn, and Zn were found to increase-about seven-, five- and six-fold, respectively, relative to bulk soil (Table V). Since a decrease in soil pH from 7.5 to 4.8 was also recorded in the rhizosphere of proteoid roots, and since some reduction was likely to occur as well (Gardner et al., 1982), it was not possible to definitely state that the exudation of citrate by proteoid roots of white lupins contributed a major proportion of the mobilization of soil Fe, Mn, and Zn (Dinkelaker et al., 1989). Jones et al. (1996a) recently demonstrated that, except for alkaline pH conditions, citrate and malate might be responsible for a substantial dissolution of iron hydroxide. They calculated that at rates of root exudation reported in the literature, citrate and malate may thereby satisfy a significant proportion of the Fe demand of plants. Considering, however, that simple organates such as citrate, oxalate, and malate, which are commonly reported as important exudates in axenic-grown plants, might be rapidly metabolized by rhizosphere microflora, their effective role in complexation processes in natural environments still remains questionnable (Darrah, 1991 ; Jones et d . , 1996b). Treeby et al. ( 1989) reported that other root exudates such as the so-called phytosiderophores can complex micronutrients such as Cu, Mn, and Zn. Mench and Fargues (1995) reported that phytosiderophores produced by roots of an Fe-efficient oat cultivar might also be involved in the mobilization of undesirable heavy metals such as Cd and Ni from sludge-contaminated soils. Phytosiderophores have been defined by Takagi et al. (1984) as a group of root exudates exhibiting strong complexing properties with respect to ferric Fe (Takagi, 1976) and identified as nonproteinogenic amino acids, such as mugineic acid and its derivatives. In this respect, they are analogues of microbial siderophores, which are literaly “iron bearers.” Literature on this topic has been extensively reviewed by Romheld and Marschner (1986a), Marschner ef al. (1989), and

250

P. HINSINGER Phytosiderophore release (pino1 g-l root d1

w

1

lo 86420 -

Barley

Wheat

Oat

Maize

Sorghum

Figure 9 Amount of phytosiderophores(PS) released by roots of Fe-sufficient ( f Fe) and Fe-deficient (- Fe) seedlings of Graminaceae species differing in their tolerance to lime-induced chlorosis. The most chlorosis-resistant species, such as barley and wheat, exhibited the highest rates of PS release and largest response to Fe-deficiency (modified from Romheld and Marschner. 1990, with kind permission from Kluwer Academic Publishers).

Romheld (1991). The synthesis and release of phytosiderophores in the rhizosphere are stimulated by Fe deficiency (Romheld, 1991) and have been described as “strategy 11” for Fe acquisition, as developed exclusively by graminaceous species (Marschner et al., 1986). Among Graminaceae, species differ widely in their ability to produce phytosiderophores, both quantitatively (Fig. 9) and qualitatively. Most remarkably, among the range of graminaceous species studied by Marschner et al. (1 989), the enhancement of the release of phytosiderophores by Fe deficiency was reported to increase accordingly to the resistance of the species to lime-induced chlorosis (Fig. 9). The efficiency of phytosiderophores for complexing Fe as a function of the competition with other metals or other complexing substances has been discussed by Romheld (1991). Romheld and Marschner (1986b, 1990) showed that, compared with Fe supplied as a microbial siderophore-Fe complex, a much larger uptake of Fe was achieved by Fe-deficient barley plants when supplied with a phytosiderophore-Fe complex, although both phyto- and microbial siderophores were responsible for a similar amount of soil Fe being mobilized (Table VI). This preferential uptake of Fe-phytosiderophore relative to other complexes (more than three orders of magnitude for uptake of Fe-HMA versus Fe-DFOB in Table VI) is related to the occurrence of a specific system of uptake of the undissociated Fe-phytosiderophore complexes (Romheld and Marschner, 1986b; W i r h ef al., 1994). This partly explains the efficiency of strategy I1 and justifies the terminology of

HOW DO PLANT ROOTS ACQUIRE MINERAL NUTRIENTS?

251

Table VI Rates of Fe Mobilization from a Calcareous Soil for a Phytosiderophoreand a Microbial Siderophore and Rates of Uptake of Fe by Fe-Deficient Barley Plants When Supplied with Fe as Fe-HMA or Fe-DFOB“.” Fe mobilization (nmol Fe g soil hour-’)

Fe uptake (nmol Fe g-’ root dry weight hour-’)

I .91 1.60

864.0 0.3

’

Phytosiderophore (HMA) Microbial siderophore (DFOB)

HMA, hydroxymugineic acid; DFOB, fenioxamine B, or desfeml. “Modified from Romheld and Mmchner, 1990, with kind permission from KluwerAcademic Publishers. *he rates of Fe mobilization were obtained with applied concentrations of HMA and DFOB of lo-’ M.

phytosiderophore, in spite of the ability of these substances to complex metals other than Fe (Treeby er al., 1989). Again, a major limitation of the efficiency of phytosiderophores is their likelihood to be degraded by rhizosphere microbes (Takagi et al., 1988; W i r h er al., 1993). However, as inferred by Romheld (1991), it seems that the production of phytosiderophores is sufficient to meet Fe requirements, at least in chlorosis-resistant species such as barley, because phytosiderophores are released at high rates that are both spatially and temporally confined. Indeed, Takagi et al. (1984) reported that the release of phytosiderophores is a rythmic phenomenon restricted to a period of 2-8 hours after the onset of daylight. In addition, Marschner et al. ( 1987) found for Fe-deficient barley that the release of phytosiderophores was maximal immediately behind the root apex (Fig. 10).These characteristics of the excretion of phytosiderophores are certainly beneficial to the acquisition of Fe, according to the results deduced from the model put forth by Darrah (199 1). In this model, Darrah (1991) predicted that short-term exudation at a high concentration of a chelating compound such as phytosiderophores, which would be localized behind the root tip, would lead to a more efficient acquisition of metal nutrients such as Fe than a persistent exudation at a lower concentration, which would be uniformly distributed along the root. This holds true particularly when assuming that rhizosphere microbial biomass is minimal behind the root tip, as evidenced by Uren and Reisenauer ( 1988) and W i r h er al. ( 1993); in this case the site for maximal exudation coincides with the site for minimal potential degradation of exudates by rhizosphere microflora. The biosynthesis and excretion of complexing substances such as phytosiderophores thus appear to be a “sophisticated” strategy developed by graminaceous species for coping with the low solubility of naturally occurring, Fe-bear-

252

P. HINSINGER Phytosiderophorerelease (pmol g" root DW per 4h) 100 1 80

-

60

-

40

-

20

-

-- -

0

I

0 1

0

'c

I

1

1

5

10

15

Distance from root apex (cm) Figure 10 Flux of phytosiderophore (PS) release as a function of the distance from root apex (along the root) for Fe-sufficient (+ Fe) and Fe-deficient (- Fe) 15-day-old barley plants. Phytosiderophores were collected over a 4-hour period starting 2 hours after the onset of the light period. Vertical bars indicate standard deviations (modified from Marschner et al., 1987).

ing secondary minerals (iron oxides) and for acquiring soil Fe. Studying the experimental weathering of a basalt rock containing Fe essentially as primary minerals (olivine and pyriboles), Femandes Barros and Hinsinger (1994) showed that only minor amounts of Fe were released into the leaching solution in the absence of plants due to the low solubility of Fe-bearing minerals in oxidizing conditions. Conversely, in the presence of plants, they reported considerable amounts of Fe released by the basalt and taken up by the seedlings of the various species studied and above all by maize. Although the release of phytosiderophores was not measured in this experiment, it is likely that the mobilization of Fe from the primary Fe-bearing minerals contained in the basalt was related to phytosiderophores excreted by this graminaceous species. These results suggest that in addition to being of prime importance for plant nutrition, root exudates such as phytosiderophores may play a significant role in mineral weathering and pedogenesis (Femandes Barros and Hinsinger, 1994) as already shown for numerous organic substances excreted by soil microorganisms (Robert and Berthelin, 1986). As previously mentioned, most root exudates act through an enhanced dissolution of the metal-bearing compound, increased release of the metal, and subsequent increase in its mobility toward the root. In some instances, however, the

HOW DO PLANT ROOTS ACQUIRE MINERAL NUTRIENTS?

253

metalkxudate complexes can precipitate in the rhizosphere, resulting in an immobilization of metals, as can occur for Ca in calcareous soils (Dinkelaker et al., 1989) or A1 in acid soils (Dinkelaker etal., 1993; Jones and Darrah, 1994). As suggested by Dinkelaker et al. (1993), the root-induced complexation of A1 due to the release of complexing exudates into the rhizosphere of Norway spruce may partly explain the tolerance of this species to A1 toxicity. Hue et al. (1986) clearly showed that organic acids can detoxify A1 in acid soils. In their experiments, the addition of organic acids alleviated the alteration of root elongation of cotton seedlings, which normally occurs at concentrations of monomeric A1 above 1 IJ.M in the bulk soil solution. Hue et al. (1986) showed that the most efficient A1 detoxifying organic acids were those that formed the most stable complexes with Al, notably citric and oxalic acids, and to a lesser extent tartaric, malic, and malonic acids. This work and that of others (e.g., Pellet et al., 1995) suggests that these root exudates, which are commonly found in the rhizosphere, may contribute to the adaptation of plants to A1 toxicity in acid soils.

Vn. OTHER INTERACTIONS INVOLVING ROOT EXUDATES Other very specifically oriented substances are produced by plant roots, such as phosphatase and phytase ectoenzymes, that may play a major role in the catalytic hydrolysis of organic P and in the subsequent acquisition of soil P (Tarafdar and Jungk, 1987; Findenegg and Nelemans, 1993). This phenomenon might be critical, since 2040% of total soil Pis present as organic P (Mengel and Kirkby, 1987). Tarafdar and Jungk ( I 987) showed a 3- to 10-fold increase in acid phosphatase activity in the rhizosphere of onion, oilseed rape, clover, and wheat relative to bulk soil and up to about a 3-fold increase in activity for alkaline phosphatase. An increase in phosphatase activity in rhizosphere relative to bulk soil was also reported for oilseed rape (Hedley er al., 1982b), maize (Dinkelaker and Marschner, 1992), lupins (Adams and Pate, 1992), and forest tree species such as Norway spruce (Haussling and Marschner, 1989). In the work of Tarafdar and Jungk (1987) the diverse phosphatases excreted in the vicinity of plant roots were likely to be responsible for the concomitant, significant mobilization of soil organic P that was indicated by the depletion of organic P in the rhizosphere (see Table I). Because they concurrently measured an increase in both fungal and bacterial biomasses in the rhizosphere relative to bulk soil, they could not definitively conclude whether these diverse rhizosphere phosphatases were plant-borne or of microbial origin. Experiments with axenic plants, however, have established that plant-borne phosphatases are excreted into the rhizosphere in the absence of microorganisms (Amann and Amberger, 1989; Grimal et al., 1992) and may account for a sub-

254

P. HIN'SINGER

stantial part of the increase in phosphatase activity near plant roots (McLachlan, 1980). In addition, the excretion of phosphatase by plant roots has been shown to be stimulated by P deficiency (McLachlan, 1980; Amann and Amberger, 1989; Grimal et al., 1992), suggesting that the release of these enzymes should be considered as an adaptative response of the P-deficient plant. However, in addition to their possible degradation by microbes, the competition of these enzymes with microbial phosphatases and their possible inactivation by adsorption onto soil reactive components such as clay minerals (Quiquampoix et al., 1995; Leprince and Quiquampoix, 1996) bring into question their effective role in soil environments. Further work is needed in this area, since in many soils organic Pcontributes a major proportion of soil P. Major root exudates are the so-called mucilage-a gelatinous material made of high-molecular-weight polysaccharides (Curl and Truelove, 1986). Polyuronic acids that are well known for their important role in the cation exchange capacity of the root cell walls account for a large proportion of this mucilagenous exudate. The consequent exchange properties of mucilage explain their ability to bind heavy metals such as Pb and Cd or micronutrients such as Cu and Zn (Morel et al., 1986; Mench et al., 1987). In acid soils, Al can similarly be detoxified by a massive adsorption on mucilage (Horst et al., 1982). In addition to these binding properties of polyuronic sites in mucilage with respect to metal cations, polyuronate ions may help desorb some anions, such as phosphate ions sorbed on soil minerals as shown for polygalacturonate by Nagarajah et al. (1970). Such a process agrees with the findings of Grimal er al. (1995). They showed that mucilage excreted by axenic-grown plants was sorbed on goethite, whereas phosphate was desorbed from goethite in the rhizosphere of maize. Many other benefits have been attributed to mucilage, including their role in establishing a better contact between the roots and the porous soil matrix (Uren and Reisenauer, 1988), thereby improving the transfer of water and mineral nutrients to the roots.

VIII. CONCLUSION The rhizosphere, i.e., the volume of soil that is influenced by root activities, can exhibit drastically different conditions compared with the bulk soil. Since the rhizosphere conditions are those that are encountered by plant roots, understanding them is critical to improving our knowledge of root functioning and plant nutrition. The rhizosphere was once recognized only for its singular microbiology. However, over the last two or three decades, evidence has accumulated that severe changes in chemical conditions relative to the bulk soil are a major trait of the rhizosphere. This review has concentrated on those peculiar modifications of chemical conditions that are occurring in the rhizosphere as a direct consequence of the

HOW DO PLANT ROOTS ACQUIRE MINERAL NUTRIENTS?

255

activity of plant roots. Obviously, some of these, such as many changes in ionic concentrations and pH, are due simply to the uptake activity of the root. Although changes in pH have received considerable attention, modifications of ionic concentrations certainly have to be considered as equally important and universally widespread features of rhizosphere chemistry. Indeed, ionic concentrations and pH are critical parameters that control many chemical reactions occumng at the root-soil interface. Root-induced changes of these conditions will therefore influence the dynamics of many nutrients in the rhizosphere and ultimately their acquisition by plant roots. The uptake of nutrients thus operates as a major driving force in nutrient acquisition. It should, however, no longer be solely considered as the ultimate mechanism involved in plant nutrition. One has to bear in mind that nutrient uptake has a major effect on the chemical conditions occumng in the rhizosphere, which will reciprocally determine the uptake activity of the root. As evidenced mostly over the last decade, in addition to these interactions, other chemical processes, such as redox reactions, complexation, or enzymatic catalyses, can take place in the rhizosphere as a direct consequence of the exudation of more or less specifically oriented metabolites produced by plant roots. The exudation of organic acids and enzymes, for instance, can contribute a significant proportion of the supply of major nutrients such as P to plant roots. Similarly, the exudation of phytosiderophores by roots of grass species plays a major role in the acquisition of poorly mobile micronutrients such as Fe and many other metals. A better understanding of these peculiar chemical processes occurring at the root-soil interface is thus a prerequisite for more accurately predicting the nutrition needs of plants and the risks of undesirable micropollutants such as heavy metals entering the food chain. Many of the aforementioned processes seem to be induced or stimulated in response to nutrient deficiencies. This suggests that they may be regarded as strategies of plant nutrition that evolved among higher plants to overcome adverse soil chemical conditions (Marschner, 1995). Whether these processes can be considered as such, it should be borne in mind that the acquisition of mineral nutrients not only relies on these diverse chemical processes but is largely influenced by (1) the colonization of the soil by the root system and (2) the physical properties of the intimate contact between the roots and the solid, liquid, and gaseous phases of the soil. Considerable progress has been made in improving our knowledge of root growth and rooting patterns (architecture of the root systems). In comparison, only a limited amount of scientific data is relevant to the physical dimension of root-soil interactions occurring in the rhizosphere. Thus, further investigations are needed in this area. In this chapter, chemical processes that occur in the rhizosphere as a direct consequence of root activity were addressed. However, other processes that significantly contribute to plant nutrition occur as a result of rhizosphere microflora. This phenomenon can be regarded as an indirect effect of plant roots, since the activi-

256

P. HINSINGER

ty of microorganisms in the rhizosphere is largely supported by root exudation of C compounds. Although this “rhizosphere effect” has been studied over almost a century, many questions remain, especially regarding its actual benefit for plant nutrition and plant growth. For instance, rhizosphere microorganisms are likely to rapidly degrade those exudates-such as organic anions, phytosiderophores, and enzymes-that are supposed to assist the plant in acquiring some mineral nutrients. They also compete with plant roots for mineral nutrients. Rhizosphere microflora can thus have a detrimental effect on plant nutrition. The energetic cost of rhizosphere microflora has been particularly addressed in the case of symbiotic rhizosphere microorganisms such as N,-fixing bacteria and mycorrhizal fungi. Nevertheless, mycorrhized plants often have a better P status than do nonmycorrhized plants, and over 95% of plant species are indeed mycorrhized. More interestingly, some species that are never mycorrhized, such as oilseed rape and white lupin, among crops, and many members of the Proteaceae family, among wild species (Harley and Harley, 1987; Brundrett and Abbott, 1991), have been reported as being some of the most efficient species for mobilizing soil P. This is attributed to their peculiar ability to excrete considerable amounts of protons and/or organic anions, such as citrate in particular. One may thus question whether these root-induced chemical processes evolved in these species to compensate for the lack of mycorrhizal support in P acquisition.Whatever the answer, the occurrence of such plant species suggests that some rhizosphere characteristics may be worth taking into account in plant breeding programs. In today’s world, where conventional, intensive agricultural practices are being challenged for both economic and environmental reasons, we should no longer breed crops and pasture species that give a maximum yield under optimal growing conditions.From the plant-nutritionview point, this practice assumes that such optimal conditions can be achieved with an adequate, and most often massive, use of fertilizers. Sustainableagriculture, however, requires moderate consumption of fertilizers. In this perspective, we should aim instead at selecting those species and varieties that can most efficiently cope with a range of nonoptimal soil conditions. A prerequisite to incorporating such considerations into our breeding programs is a better understanding of the actual, combined effect on nutrient acquisition of the various processes that occur in the rhizosphere. New experimental tools derived from molecular biology, such as using mutants and genetic manipulations,will certainly help in ascertaining the relative contribution of the numerous mechanisms that are involved. Moreover, using mathematical models of the combined phenomena involved in the process of mineral nutrient acquisition will also help us to improve our understanding of plant nutrition. For this purpose, as pointed out by Darrah (1993), a more integrative and quantitative approach of rhizosphere processes is indeed required. This is a fundamentalprerequisite to managing plant nutrition in agricultural, forested, and natural environments.

HOW DO PLANT ROOTS ACQUIRE MINERAL NUTRIENTS?

257

ACKNOWLEDGMENTS This chapter is dedicated to the memory of Professor Horst Marschner, who contributed much to our understanding of the chemical processes involved in the rhizosphere. I also thank Professor R. J. Gilkes, Dr. J. C. Arvieu, and Dr. B. Jaillard for their comments on an earlier version of this chapter.

REFERENCES Adams, M. A., and Pate, J. S. (1992). Availability of organic and inorganic forms of phosphorus to lupins (Lupinus spp.). PIanr Soil 145, 107-1 13. Ae, N., Arihara, J.. Okada, K..Yoshihara. T., and Johansen, C. (1990). Phosphorus uptake by pigeon pea and its role in cropping systems of the Indian subcontinent. Science 248,477480. Aguilar, S. A., and van Diest, A. (1981). Rock phosphate mobilization induced by the alkaline pattern of legumes utilizing symbiotically fixed nitrogen. Pfant Soil 61,2742. Ahmad, A. R., and Nye, P. H. (1990). Coupled diffusion and oxidation of ferrous iron in soils. 1. Kinetics of oxygenation of ferrous iron in soil suspension. J. Soil Sci. 41,395409. Amann, C., and Amberger, A. (1989). Phosphorus efficiency of buckwheat (Fagopyrum esculentum). 2. Pjlanzenem. Bodenk. 152, 181-189. Ando, T., Yoshida, S., and Nishiyama, I. (1983). Nature of oxidizing power of rice roots. Plant Soil 72, 57-7 1. Armstrong. W. (1967). The use of polarography in the assay of oxygen diffusing from roots in anaerobic media. Physiol. Plant. 20,540-553. Arvieu, J. C. (1980). RCactions des phosphates minkraux en milieu calcaire; constquence sur I’Ctat et la solubilitk du phosphore. Sci. Sol 3, 179-190. 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. Bacha. R. E., and Hossner, L. R. (1977). Characteristics of coating formed on rice roots as affected by iron and manganese additions. Soil Sci. Soc. Am. J. 41,931-935. Baethgen, W. E.. and Alley, M. M. (1987). Nonexchangeable ammonium nitrogen contribution to plant available nitrogen. Soil Sci. SOC.Am. J. 51, 110-1 15. Barber, S. A. (1995). “Soil Nutrient Bioavailability: A Mechanistic Approach,” 2nd ed. John Wiley, New York. Barrow, N. J. (1983). On the reversibility of phosphate sorption by soils. J. Soil Sci. 34,751-758. Barrow, N. J. (1984). Modelling the effect of pH on phosphate sorption by soils. J. Soil Sci. 35, 283-297. Begg, C. B. M., Kirk, G. 1. D., MacKenzie, A. E, and Neue, H.-U. (1994). Root-induced iron oxidation and pH changes in the lowland rice rhizosphere. New Phyfol. 128,469477. Bekele, T., Cino, B. J., Ehlert, P. A. I., van der Maas, A. A,, and van Diest, A. (1983). An evaluation of plant-borne factors promoting the solubilization of alkaline rock phosphates. Planr Soil 75, 361-78. Bertsch, P. M., and Thomas, G. W. (1985). Potassium status of temperate region soils. In “Potassium in Agriculture” (R. D. Munson, ed.), pp. 131-162. American Soc. of Agron., Crop Sci. Soc.Am. and Soil Sci. Soc.Am., Madison, W. Bhat, K. K. S., and Nye. P. H. (1973). Diffusion of phosphate to plant roots in soil. I. Quantitative autoradiography of the depletion zone. Plunr Soil 38, 161-175. Bhat, K. K. S., Nye, P. H., and Baldwin, J. P. (1976). Diffusion of phosphate to plant roots in soil. IV.

258

P. HINSINGER

The concentration distance profile in the rhizosphere of roots with root hairs in a low-P soil. Planr Soil 44,63-72. Bienfait, H. F., Bino, R. J., van der Bliek, A. M., Duivenvoorden, J. F., and Fontdine, J. M. (1983). Characterization of ferric reducing activity in roots of Fe-deficient Phaseolus vulgaris. Physiol. Plant. 59, 196202. Bolan, N. S. (1991). A critical review on the role of mycorrhizal fungi in the uptake of phosphorus by plants. Plant Soil 134, 189-207. Bolan, N. S., Naidu, R., Mahimairaja, S., and Baskaran, S. (1994). Influence of low-molecular-weight organic acids on the solubilization of phosphates. Biol. Fer/. Soils 18,311-319. Bosc, M. (1988). Enseignements fournis par des essais de longue durke sur la fumure phosphatke et potassique. 3. Essai sur la fumure potassique. In “Phosphore et potassium dans les relations solplante: conskquences sur la fertilisation” (L. Gachon, ed.), pp. 403-466. Institut National de la Recherche Agronomique, Paris. Bowen, G. D. (1980). Coping with low nutrients. In “The Biology of Australian Plants” (J. S. Pate and A. J. McComb, eds.), pp. 33-64. Univ. of Western Australia Press, Nedlands. 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 Lolium perenne L. at constant phosphate concentrations. J. Exp. Bor. 35, 1210-1221. Brewster, J. L., Bhat, K. K. S., and Nye, P. H. (1976). The possibility of predicting solute uptake and plant growth response from independently measured soil and plant characteristics. V. The growth and phosphorus uptake of rape in soil at a range of phosphorus concentrations and a comparison of results with the predictions of a simulation model. Plant Soil 44,295-328. Brown, J. C., and Ambler, J. E. (1973).“Reductants” released by roots of Fe-deficient soybeans. Agran. J. 65,311-314. Brundrett, M. C., and Abbott, L. K. (1991). Roots of jarrah forest plants. I. Mycorrhizal associations of shrubs and herbaceous plants. Ausr. J. Bot. 39,445-457. Bruand, A,, Cousin, I.. Nicoullaud, B., Duval, O., and BCgon, J. C. (1996). Backscattered electron scanning images of soil porosity for analysing soil compaction around roots. Soil Sci. SOC.Am. J . 60, 895-901. Chaney, R. L., Brown, J. C., andTiffin, L. 0. (1972). Obligatory reduction of ferric chelates in iron uptake by soybeans. Plant Physiol. 50,208-213. Chen, C. C., Dixon, J. B., and Turner, F. T.(1980). Iron coatings on rice roots: Mineralogy and quantity influencing factors. Soil Sci. Soc. Am. J. 44,635-639. Claassen, N., and Jungk, A. (1982). Kaliumdynamik im wurzelnahen Boden in Beziehung zur Kaliumaufnahme von Maispflanzen. 2. Pfanzenern. Bodenk. 145,s 13-525. Clarkson, D. T. (1985). Factors affecting mineral nutrient acquisition by plants. Ann. Rev. Planr Physiol. 36,7771 15. Curl, E. A., and Truelove, B. (1 986). “The Rhizosphere.” Springer-Verlag. Berlin-Heidelberg. Darrah, P. R. (1991). Models of the rhizosphere. 1. Microbial population dynamics around a root releasing soluble and insoluble carbon. Planr Soil 133, 187-199. Darrah,P. R. (1993). The rhizosphere and plant nutrition: Aquantitative approach. Plant Soil 159156, 1-20. Dexter, A. R. (1987). Compression of soil around roots. Plant Soil 97,401406. Dinkelaker, B., and Marschner. H. (1992). In vivo demonstration of acid phosphatase activity in the rhizosphere of soil-grown plants. Plunt Soil 144, 199-205. Dinkelaker, B., Romheld, V., and Marschner, H. (1989). Citric acid excretion and precipitation of calcium citrate in the rhizosphere of white lupin (Lupinus albus L.). PIanr Cell Envi,: 12,285-292. Dinkelaker, B., Hahn. G., and Marschner, H. (1993).Non-destructive methods for demonstrating chemical changes in the rhizosphere. LI. Application of methods. Plant Soil 155/156,73-76. Dinkelaker, B., Hengeler, C., and Marschner, H. (1995). Distribution and function of proteoid roots and other root clusters. Bor. Acra 108, 183-200.

HOW DO PLANT ROOTS ACQUIRE MINERAL NUTRIENTS?

259

Drew. M. C. (1988). Effect of flooding and oxygen deficiency on plant mineral nutrition. Adv. Plant Nutr: 3, 115-159. Drew, M. C., and Nye, P. H. (1969). The supply of nutrient ions by diffusion to plant roots in soil. 11. The effect of root hairs on the uptake of potassium by roots of ryegrass (Lolium multiflorum). Plant Soil 31,407424. Epstein, E., and Hagen, C. E. (1952).A kinetic study of the absorption of alkali cations by barley roots. Plant Physiol. 27,457474. Fanning. D. S.. Keramidas. V. 2..and El-Desoky, M. A. (1989). Micas. In ”Minerals in Soil Environment,” 2nd ed. (J. B. Dixon and S. B. Weed, eds.), pp. 55 1-634. Soil Sci. Soc.Am., Madison, WI. Fan; E., Vaidyanathan, L. V., and Nye, P. H. (1969). Measurement of ionic concentration gradients in soil near roots. Soil Sci. 107, 385-9 1. Fernandes Barros, 0. N., and Hinsinger, P. (1994). Basalt weathering as induced by the roots of higher plants. In “Proc. 15th ISSS Cong.” Acapulco. Findenegg, G. R., and Nelemans, J. A. (1993). The effect of phytase on the availability of P from myoinositol hexaphosphate (phytate) for maize roots. Plant Soil 154, 189-196. Fisher, H. M., and Stone, E. L. (1991). Iron oxidation at the surfaces of slash pine roots from saturated soil. Soil Sci. Soc. Am. J. 55, I 123- I 129. Flessa, H., and Fischer, W. R. (1992). Plant-induced changes in the redox potentials of rice rhizospheres. Plant Soil 143,55-60. Fohse, D., Claassen. N.. and Jungk, A. (1988). Phosphorus efficiency of plants. I. External and internal requirement and P uptake efficiency of different plant species. Plant Soil 110, 101-109. Fox, R. L. ( I 98 I). External phosphorus requirements ofcrops. In “Chemistry in the Soil Environment.” pp. 223-239. Am. SOC.Agron., Soil Sci. Soc.Am., Madison, WI. Freeman. J. S.. and Rowell, D. L. (1981). The adsorption and precipitation of phosphate onto calcite. J. Soil Sci. 32,75584. Fried, M., and Shapiro, R.E. (1961). Soil plant relationships in ion uptake. Ann. Rev. Plant Physiol. 12,91-112. Gahoonia, T. S. (1993). Influence of root-induced pH on the solubility of soil aluminium in the rhizosphere. Plant Soil 149,289-291. Gahoonia, T. S., and Nielsen, N. E. (1992). Control of pH at the soil-root interface. Planf Soil 140, 49-54. Gahoonia, T. S., Claassen, N., and Jungk, A. (1992a). Mobilization of phosphate in different soils by ryegrass supplied with ammonium or nitrate. Plant Soil 140,241-248. Gahoonia, T. S., Claassen, N., and Jungk, A. (1992b). Mobilization of residual phosphate of different phosphate fertilizers in relation to pH in the rhizosphere of ryegrass. Ferr. Res. 33,229-237. Gardner, W. K., Parberry, D. G., and Barber, D. A. ( I 982). The acquisition of phosphorus by Lupinus albus L. I. Some characteristics of the soil/root interface. Plant Soil 68, 19-32. Gardner, W. K., Barber, D. A,, and Parbeny, D. G. (1983). The acquisition of phosphorus by Lupinus albus L. 111. The probable mechanism by which phosphorus movement in the soil/root interface is enhanced. Plant Soil 70,107-124. Gerke, J. (1994). Kinetics of soils phosphate desorption ffected by citric acid. Z. Pjanzenern. Bodenk. 157, 17-22. Gerke, J., Romer. W., and Jungk, A. (1994). The excretion of citric and malic acid by proteoid roots of Lupinus albus L.: Effects on solubility of phosphate, iron, and aluminum in the proteoid rhizosphere in samples of an oxisol and a luvisol. 2. Pjanzenern. Bodenk. 157,289-294. Godo, G. H.. and Reisenauer. H. M. (1980). Plant effects on soil manganese availability. Soil Sci. Soc. Am. J. 44,993-995. Green, M. S., and Etherington, J. R. (1977). Oxidation of ferrous iron by rice (Uryza sativa L.) roots: A mechanism for waterlogging tolerance? J. Exp. Bot. 28,678-690. Grierson, P. E (1992). Organic acids in the rhizosphere of Banksia integrifolia L. f. Plant Soil 144, 259-265.

260

P. HNSINGER

Grimal, J. Y., Delauney, A,, Frossard, E., and Morel, J. L. (1992). Utilisation de phosphore organique par un mays cultivt en conditions stbriles. In “Proc. 4th Intl. IMPHOS Conf., pp. 629-63 I . Ghent, Belgium. Grimal, J. Y., Frossard, E., and Morel, J. L. (1995). Acquisition of phosphorus sorbed on goethite by Zea mays. In “Soil Management in Sustainable Agriculture” (H. Cook and H. Lee, eds.). Wye College Press, London. Grinsted, M. J., Hedley, M. J., White, R. E., and Nye, P. H. (1982). Plant-induced changes in the rhizosphere of rape (Brassica napus var. Emerald). 1. pH changes and the increase in P concentration in the soil solution. New Phyfol. 91, 19-29. Hamza, M. A., and Aylmore, L. A. G. (1992). Soil solute concentration and water uptake by single lupin and radish plant roots. I. Water extraction and solute accumulation. Plant Soil 145, 187-196. Harley, J. L., and Harley, E. L. (1987).A check-list of mycorrhiza in the British flora. New Phyrol. 105, 1-102.

Haussling, M.. and Marschner, H. (1989). Organic and inorganic soil phosphates and acid phosphatase activity in the rhizosphere of 80-year-old Norway spruce (Picea abies (L.) Karst) trees. Biol. Fen. Soils 8, 128-133. Haussling, M., Leisen, E., Marschner, H., and Romheld, V. (1985). An improved method for non-destructive measurements of the pH at the root-soil interface (rhizosphere). J. Plant Physiol. 117, 371-375. Haynes. R. J.. (1990).Active ion uptake and maintenance of cation-anion balance: Acritical examination of their role in regulating rhizosphere pH. Plant Soil 126,247-264. Hedley, M. J., Nye, P. H.,and White, R. E. (1982a). Plant-induced changes in the rhizosphere of rape (Brassica napus var. Emerald) seedlings. II. Origin of the pH change. New Phyrol. 91,3144. Hedley, M. J., White, R. E., and Nye, P. H. (1982b). Plant-induced changes in the rhizosphere of rape (Brassica napus var. Emerald) seedlings. IU.Changes in L value, soil phosphate fractions, and phosphatase activity. New Phyrol. 91,45-56. Helal, H. M., and Sauerbeck, D. (1989). Carbon turnover in the rhizosphere. Z. Pjanzenem. Bodenk. 152,211-216. Hendriks, L., Claassen, N., and Jungk, A. (1981). Phosphatverarmung des wurzelnahen Bodens und Phosphataufnahme von Mais und Raps. Z. Pflanzenern. Bodenk. 144,486-499. Hiltner, L. (1904). Uber neuere Ehrfahrungen und Problem auf dem Gebiet der Bodenbakteriologie unter besonderer Beriicksichtigung der Grundiingung und Brache. Arb. Drsch. Landwirr. Ges. 98, 59-78. Hinsinger, P. (1994). The acquisition of mineral nutrients by roots: Rhizosphere processes. In “Proc. 3rd Cong. European Soc. for Agron.” (M. Bonn and M. Sattin, eds.), pp. 428437. European SOC. for Agron., Colmar, France. Hinsinger, P., and Gilkes, R. J. (1995). Root-induced dissolution of phosphate rock in the rhizosphere of lupins grown in alkaline soil. Ausr. J. Soil Res. 33,477489. Hinsinger, P., and Gilkes, R. J. (1996). Mobilization of phosphate from phosphate rock and aluminasorbed phosphate by the roots of ryegrass and clover as related to rhizosphere pH. Eur: J. Soil Sci. 47,533-544. Hinsinger, P., and Gilkes, R. J. (1997). Dissolution of phosphate rock in the rhizosphere of five plant species grown in an acid, P-fixing mineral substrate. Geoderma 75,231-249. Hinsinger, P., and Jaillard, B. (1993). Root-induced release of interlayer potassium and vermiculitization of phlogopite as related to potassium depletion in the rhizosphere of ryegrass. J. Soil Sci. 44, 525-534. Hinsinger, P., Jaillard, B., and Dufey, J. E. (1992). Rapid weathering of a trioctahedral mica by the roots of ryegrass. Soil Sci. SOC.Am. J. 56,977-982. Hinsinger, P.,Elsass. F.,Jaillard, B., and Robert, M. (1993). Root-induced irreversible transformation of a trioctahedral mica in the rhizosphere of rape. J. Soil Sci. 44,535-545.

HOW DO PLANT ROOTS ACQUIRE MINERAL, NUTRIENTS?

261

Hoffland, E. (1992). Quantitative evaluation of the role of organic acid exudation in the mobilization of rock phosphate by rape. Planr Soil 140,279-289. Hoffland, E., Findenegg, G. R., and Nelemans, J. A. (1989). Solubilization of rock phosphate by rape. 11. Local root exudation of organic acids as a response to P-starvation. Plant Soil 113, 161-165. Hoffland, E., Van den Boogaard, R.. Nelemans, J. A., and Findenegg, C. R. (1992). Biosynthesis and root exudation of citric and malic acids in phosphate-starved rape plants. New Phyrol. 122, 675-680. Horst, W. J., Wagner, A., and Marschner, H. (1982). Mucilage protects root meristems from aluminium injury. 2.Pfanzenern. Bodenk. 105,435-444. Hiibel, F., and Beck, E. (1993). In-situ determination of the P-relations around the primary root of maize with respect to inorganic and phytate-P. Plant Soil 157, 1-9. Hue, N. V., Craddock. G. R.. and Adams, F. (1986). Effect of organic acids on aluminum toxicity in subsoils. Soil Sci. SOC.Am. J. 50,28-34. Jaillard, B. (1984). Mise en evidence de la neogentse de sables calcaires sous I’influence des racines: Incidence sur la granulometrie du sol. Agronomie 4,91-100. Jaillard, B. (1985). Activitt! racinaire et rhizostructures en milieu carbonate. Pidologie 35, 297-3 13. Jaillard, B. (1987). Les structures rhizomorphes calcaires: modtle de reorganisation des minkraux du sol par les racines. Ph.D. thesis, Universitt! des Sciences et Techniques du Languedoc, Montpellier, France. Jaillard, B., Guyon, A., and Maurin. A. F. (1991).Structure and composition of calcified roots, and their identification in calcareous soils. Geodema 50, 197-210. Jaillard. B., Ruiz, L., and Arvieu, J. C. (1996). pH mapping in transparent gel using color indicator videodensitometry. Plant Soil 183,85-95. Jakobsen, I., Abbott, L. K., and Robson, A. D. (1992). External hyphae of vesicular-arbuscular mycorrhizal fungi associated with Trifolium subrerruneum L. I . Spread of hyphae and phosphorus inflow into roots. New Phytol. 120,371-380. Jarvis, S . C., and Robson, A. D. (1983).The effects of nitrogen nutrition of plants on the development of acidity in Western Australian soils. 1. Effects with subterranean clover grown under leaching conditions. Aust. J. Agric. Res. 34,341-353. Jones, D. L., and Darrah, P.R. (1994). Role of root derived organic acids in the mobilization of nutrients from the rhizosphere. Planr Soil 166,247-257. Jones, D. L.. and Ddrrah, P.R. (1995). Influx and efflux of organic acids across the soil-root interface of Zea mays L. and its implications in rhizosphere C flow. Plant Soil 173, 103-109. Jones, D. L., Darrah, P. R.,and Kochian, L. V. (1996a). Critical evaluation of organic acid mediated dissolution in the rhizosphere and its potential role in root iron uptake. Plunt Soil 180,5746. Jones, D. L., Prabowo, A. M., and Kochian, L. V. (1996b).Kinetics of malate transport and decomposition in acid soils and isolated bacterial populations: The effect of microorganisms on root exudation of malate under Al stress. Plant Soil 182,239-247. Jungk, A. (1996). Dynamics of nutrient movement at the soil-root interface. In “Plant Roots. The Hidden Half” 2nd ed. (Y.Waisel, A. Eshel, and U. Kafkafi, eds.), pp. 529-556. Marcel Dekker, New York. Jungk, A,. and Claassen, N. (1986). Availability of phosphate and potassium as the result of interactions between root and soil in the rhizosphere. Z. Pfunzenern. Bodenk. 149,411-427. Keerthisinghe, G., De Datta, S. K., and Mengel, K. (1985). Importance of exchangeable and nonexchangeable NH; in nitrogen nutrition of lowland rice. Soil Sci. 140, 194-201. Khasawneh, F. E., and Doll, E. C. (1978).The use of phosphate rock for direct application to soils. Adv. Agron. 30, 159-206. Kinraide, T. B. (1991). Identity of the rhizotoxic aluminium species. Plant Soil 134, 167-178. Kirk, G. J. D., and Nye, P. H. (1986). A simple model for predicting the rates of dissolution of sparingly soluble calcium phosphates in soil. I. The basic model. J. Soil Sci. 37,529-540.

262

P. HINSINGER

Kirk, G. J. D., and Saleque, M. A. (1995). Solubilization of phosphate by rice plants growing in reduced soil: Prediction of the amount solubilized and the resultant increase in uptake. European J. Soil Sci. 46,247-255. Kodama, H., Nelson, S., Yang. F., and Kohyama, N. (1994).Mineralogy of rhizospheric and non-rhizospheric soils in corn fields. Clays Clay Min. 42,755-763. 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. Kraus, M., Fusseder. A., and Beck, E. (1987). In siru determination of the phosphate-gradient around a root by radioautography of frozen soil sections. Plant Soil 97,407418, Kuchenbuch, R., and Jungk. A. (1982).A method for determining concentration profiles at the soil-root interface by thin slicing rhizospheric soil. Plant Soil 68,3914. Kuchenbuch, R., and Jungk, A. (1984).Wirkung der Kaliumdiingung auf die Kaliumverfiigbarkeit in der Rhizosph~evon Raps. Z. Pjanzenern. Bodenk. 147,435-448. Lamont, B. (1982). Mechanisms for enhancing nutrient uptake in plants with particular reference to mediterranean South Africa and Western Australia. Bot. Rev. 48,597-689. Leprince, F., and Quiquampoix, H. (1996). Extracellularenzyme activity in soil: Effect of pH and ionic strength on the interaction with montmorillonite of two acid phosphatases secreted by the ectomycorrhizal fungus Heheloma cylindrosporum. Eur: J. Soil Sci. 47,5 1 1-522. Lewis, D. G., and Quirk, J. P. (1967).Phosphate diffusion in soil and uptake by plants. 111. P”’-movement and uptake by plants as indicated by P’2-autoradiogrdphy. Plant Soil 27,445453, Li, X.-L., George, E., and Marschner, H. (1991). Extension of the phosphorus depletion zone in VAmycorrhizal white clover in a calcareous soil. Plant Soil 136,4148. Lindsay, W. L. (1954). Role of chelation in micronutrient availability. In “The Plant Root and Its Environment” (E. W. Carson, ed.), pp. 507-524. University Press of Virginia. Lindsay, W. L. (1979). “Chemical Equilibria in Soils.” John Wiley and Sons, New York. Lindsay, W. L., Vlek, P. L. G., and Chien, S. H. (1989). Phosphate minerals. In “Minerals in Soil Environment,” 2nd ed. (J. B. Dixon and S. B. Weed, eds.), pp. 1089-1130. Soil Sci. Soc. of Am., Madison, WI. Lipton, D. S., Blanchar, R. W., and Blevins, D. G. (1987). Citrate, malate, and succinate concentration in exudates from P-sufficient and P-stressed Medicago sariva L. seedlings. Plant Physiol. 85, 3 15-3 1 7. Lorenz, S. E., Hamon, R. E.. and McGrath, S. P. (1994). Differences between soil solutions obtained from rhizosphere and non-rhizosphere soils by water displacement and soil centrifugation. Eua J. Soil Sci. 4 5 4 3 1-438. Loss, S. P., Robson, A. D.. and Ritchie, G. S. P. (1993). H+/OH- excretion and nutrient uptake in upper and lower parts of lupin (Lupinus angustijolius L.) root systems. Annals Bor. 72,3 15-20. Malzer, G. L., and Barber, S. A. (1975). Precipitation of calcium and strontium sulfates around plant roots and its evaluation. Soil Sci. SOC.Amer: Pmc. 39,492495. Mdrschner, H. (1990). Mobilization of mineral nutrients in the rhizosphere In “Proc. 1st Cong. European Soc. Agron.” (A. Scaife, ed.), Session 3, 0 00, pp. 1-11. European Soc. Agron., Colmar, France. Marschner, H. (1995). “Mineral Nutrition of Higher Plants, 2nd ed. Academic Press. London. Mdrschner, H., and Dell, B. (1994). Nutrient uptake in rnycorrhizal symbiosis. Plant Soil 159,89-102. Marschner, H., and Romheld, V. (1983). In vivo measurement of root-induced pH changes at the soil-root interface: Effect of plant species and nitrogen source. Z. Ppbnzenphysiol. 111,241-251. Marschner, H.. Romheld, V., and Ossenberg-Neuhaus, H. (1982). Rapid method for measuring changes in pH and reducing processes along roots of intact plants. 2.Pflanzenphysiol. 105,407416. Marschner, H., Romheld, V., Horst, W. J., and Martin, P. (1986). Root-induced changes in the rhizosphere: Importance for the mineral nutrition of plants. Z. Pfanzenern. Bodenk. 149,441456. Marschner, H., Romheld, V., and Kissel, M. (1987). Localization of phytosiderophore release and iron uptake along intact barley roots. Physiol. Plant. 71, 157-162.

HOW DO PLANT ROOTS ACQUIRE MINERAL NUTRIENTS?

263

Marschner, H.,Treeby, M., and Romheld, V. (1989). Role of root-induced changes in the rhizosphere for iron acquisition in higher plants. Z. P’anzenern. Bodenk. 152, 197-204. Marschner, H., Hiussling, M.. and George, E. (1991). Ammonium and nitrate uptake rates and rhizosphere-pH in non-mycorrhizal roots of Norway spruce (Picea abies (L). Karst). Trees Strucr. Funct. 5, 14-2 I . McLachlan, K. D. (1980).Acid phosphatase activity of intact roots and phosphorus nutrition in plants. I. Assay conditions and phosphatase activity. Ausr. J. Agric. Res. 3 1 , 4 2 9 4 0 . Mench. M., and Fargues, S. (1995). Metal uptake by iron-efficient and inefficient oats. In “Iron Nutrition in Soils and Plants,” (J. Abadia, ed.), pp. 217-223. Kluwer Academic Publishers, Dordrecht, the Netherlands. Mench, M., and Martin, E. (1991). Mobilization of cadmium and other metals from two soils by root exudates ofZeu m a w L., Nicotiana tahacum L., and Nicotiana rustica L. Plant Soil 132,187-196. Mench, M., Morel, J. L.. and Guckert, A. (1987). Metal binding properties of high molecular weight soluble exudates from maize (Zea mays L.) roots. Biol. Fen. Soils 3, 165-169. Mengel. K., and Kirkby, E. A. (1987). “Principles of Plant Nutrition,” 4th ed. Intl. Potash Institute, Bern, Switzerland. Mengel, K., and Scherer, H.W. (198 I). Release of nonexchangeable (tixed) soil ammonium under field conditions during the growing season. Soil Sci. 131.226232. Mengel. K., Horn, D., and Tributh, H. (1990). Availability of interlayer ammonium as related to root vicinity and mineral type. SoilSci. 149, 131-137. Merckx, R., van Ginkel, J. H., Sinnaeve, J., and Cremers, A. (1986a). Plant-induced changes in the rhizosphere of maize and wheat. I. Production and turnover of root-derived material in the rhizosphere of maize and wheat. Plant Soil 96,85-93. Merckx, R., van Ginkel, J. H.,Sinnaeve. J.. and Cremers, A. (1986b). Plant-induced changes in the rhizosphere of maize and wheat. 11. Complexation of cobalt, zinc, and manganese in the rhizosphere of maize and wheal. Plant Soil 96,95-107. Moorby, H., White, R. E., and Nye, P. H. (1988). The influence of phosphate nutrition on H ion efflux from the roots of young rape plants. Plant Soil 105,247-56. Morel, J. L., Mench, M., and Guckert,A. (1986).Measurement of PbZ+,Cu2+,and Cd2+ binding with mucilage exudates from maize (&a mays L.) roots. B i d . Fert. Soils 2,29-34. Murrrnann. R. P., and Peech. M. (1969). Effect of pH on labile and soluble phosphate in soils. Soil Sci. Soc. Am. Proc. 33, 205-2 10. Nagarajah, S., Posner. A. M., and Quirk, J. P. (1970). Competitive adsorptions of phosphate with polygalacturonate and other organic anions on kaolinite and oxide surfaces. Nature 228, 83-84. Niebes, J.-F., Dufey, J. E., Jaillard, B., and Hinsinger, P. ( I 993). Release of nonexchangeable potassium from different size fractions of two highly K-fertilized soils in the rhizosphere of rape (Brassicu napus cv. Drakkar). Plant Soil 155/156,403406. Nye, P. H.(1981). Changes of pH across the rhizosphere induced by roots. Plant Soil 61,7-26. Nye, P. H.(1983). The diffusion of two interacting solutes in soil. J. Soil Sci. 34,677-691. Nye, P. H.(1986). Acid-base changes in the rhizosphere. Adv. Plant NutK 2, 129-153. Oertli, J. J., and Opoku, A. A. (1974). Interaction of potassium in the availability and uptake of iron from ferric hydroxide. Soil Sci. Soc. Am. Proc. 38,451454. Owusu-Bennoah, E., and Wild, A. (1979). Autoradiography of the depletion zone of phosphate around onion roots in the presence of vesicular-arbuscular mycorrhiza. New Phytol. 82, 133-140. Paauw, E van der. ( I 97 I ). An effective water extraction method for the determination of plant available soil phosphorus. Plant Soil 3 4 , 4 6 7 4 8 I . Parfitt, R. L. (1978). Anion adsorption by soils and soil materials. Adv. Agron. 30, 1-50. Parlitt, R. L. ( 1979).The availability of Pfrom phosphate-goethite bridging complexes. Desorption and uptake by ryegrass. Plant Soil 5 3 , 5 5 4 5 . Pellet, D. M., Grunes, D. L., and Kochian, L. V. (1995). Organic acid exudation as an aluminum-tolerance mechanism in maize (Zeu movs L.). Plunta 196,788-795.

264

P.HINSINGER

Petersen, W.. and Bottger, M. (1991). Contribution of organic acids to the acidification of the rhizosphere of maize seedlings. Plant Soil 132, 159-163. Qutmener. J. (1986). Important factors in potassium balance sheets. In “Nutrient Balances and the Need for Potassium. Proc. 13th IPI Cong.,” pp. 41-72. Intl. Potash Institute, Bern, Switzerland. Quiquampoix, H.. Abadie, J., Baron, M. H.,Leprince, F.. Matumoto-Pintro, P. T., Ratcliffe, R. G., and Staunton, S. (1995). Mechanisms and consequences of protein adsorption on soil mineral surfaces. In “Proteins at Interfaces 11. ACS Symposium Series 602” (T. A. Horbett and J. L. Brash, eds.), pp. 321-333. Am. Chem. Soc.,Washington, D.C. Quispel, A. (1983). Dinitrogen-fixing symbioses with legumes, non-legumes angiosperms, and associative symbioses. In “Inorganic Plant Nutrition. Encyclopedia of Plant Physiology.” (A. Lauchli and R. L. Bieleski, eds.), New Series, Vol. 15A, pp. 286-329. Springer-Verlag. Berlin. Riley, D., and Barber, S. A. (I97 I). Effect of ammonium and nitrate fertilization on phosphorus uptake as related to root-induced pH changes at the root-soil interface. Soil Sci. SOC. Am. froc. 35, 301-306. Robert, M., and Berthelin. J. (1986). Role of biological and biochemical factors in soil mineral weathering. In “Interactions of Soil Minerals with Natural Organics and Microbes,” (P.M. Huang and M. Schnitzer, eds.), Special publication 17, pp. 453-495. Soil Sci. Soc.Am., Madison, W1. Romheld, V. (1986). pH-Veranderungen in der Rhizosphiire verschiedener Kulturpflanzenarten in Abhangigkeit vom Nhstoffangebot. Potash Rev. 55, 1-8. Romheld, V. (1991). The role of phytosiderophores in acquisition of iron and other micronutrients in graminaceous species: An ecological approach. Plant Soil 130, 127-1 34. Romheld, V.. and Marschner, H. (1983). Mechanism of iron uptake by peanut plants. I. Fe(II1) reduction, chelate splitting, and release of phenolics. Plant Physiol. 71,949-954. Romheld, V., and Marschner. H.(1986a). Mobilization of iron in the rhizosphere of different plant species. Adv. Plant Nutr: 2, 155-204. Romheld, V.. and Marschner, H. (1986b). Evidence for a specific uptake system for iron phytosiderophores in roots of grasses. Plant Physiol. 80, 175-180. Romheld, V., and Marschner, H. (1990). Genotypical differences among graminaceous species in release of phytosiderophores and uptake of iron phytosiderophores. Plant Soil 123, 147-153. Romheld, V., Miiller, C., and Marschner, H. (1984). Localization and capacity of proton pumps in roots of intact sunflower plants. Plant Physiol. 76,603-606. Ruiz, L. (1992). Mobilisation du phosphore des apatites dans la rhizosphbre. Rble de I’excrttion de protons par le racines. Ph.D. thesis, Universitt Montpellier II, France. Ruiz, L.. and Arvieu, J. C. (1992). Solubilization of phosphate minerals by plant roots. In “Proc. 4th Intl. IMPHOS Conf.” Ghent, Belgium. Saleque, M. A., and Kirk, G. J. D. (1995). Root-induced solubilization of phosphate in the rhizosphere of lowland rice. New Phytol. 129,325-336. Scherer, H. W., and Ahrens, G. (1996). Depletion of non-exchangeable NH4-N in the soil-root interface in relation to clay mineral composition and plant species. Eur: J. Agron. 5, 1-7. Schwertmann, U. (1991). Solubility and dissolution of iron oxides. Plant Soil 130, 1-25. Sinha, B. K., and Singh, N. T. (1974). Effect of transpiration rate on salt accumulation around corn roots in a saline soil. Agron. J. 66,557-560. Sinha, B. K., and Singh, N. T. (1976). Chloride accumulation near corn root under different transpiration, soil moisture, and soil salinity regimes. Agron. J. 68,346348. Sparks, D. L. (1987). Potassium dynamics in soils. Adv. Soil Sci. 6,1-63. Staunton, S., and Leprince, F. (1996). Effect of pH and some organic anions on the solubility of soil phosphate: Implications for P bioavailability. Eur: J. Soil Sci. 47,231-239. Steffens, D. ( 1987). Einfluss langjahriger Diingung mit verschiedenen Phosphatdungerformen aud die Phosphatverfiigbarkeit in der Rhizosphiire von Raps. Z. fjanzenern. Bodenk. 150,75-80. Stirzaker, R. J., and Passioura, J. B. (1996). The water relations of the root-soil interface. Plant Cell Envi,: 19,201-208.

HOW DO PLANT ROOTS ACQUIRE MINERAL NUTRIENTS?

265

Takagi. S. (1976) Naturally occurring iron-chelating compounds in oat- and rice-root washings. I. Activity measurement and preliminary characterization. Soil Sci. Plant Nut% 22,423433. Takagi, S., Nomoto, K., and Takemoto, T. (1984). Physiological aspects of mugineic acid, a possible phytosiderophore of graminaceous plants. J. Plant Nut% 7,469477. Takagi, S., Kamei, S.. and Yu, M.-H. (1988). Efficiency of iron extraction from soil by mugineic acid family phytosiderophores. J. Plant Nurr 11,643-65 I. Tarafdar, J. C., and Jungk, A. (1987). Phosphatase activity in the rhizosphere and its relation to the depletion of soil organic phosphoms. Eiol. Fert. Soils 3, 199-204. 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. Trofymow, J. A., Coleman, D. C., and Cambardella, C. (1987).Rates of rhizodeposition and ammonium depletion in the rhizosphere of axenic oat roots. Plant Soil 97,333-344. Trolldenier, G. (1988). Visualization of oxidizing power of rice roots and of possible participation of bacteria in iron deposition. Z. Pfanzenern. Eodenk. 151, 117-121. Uren, N. C. (1981). Chemical reduction of an insoluble higher oxide of manganese by plant roots. J . Plant Nutr: 4,65-7 I . Uren, N. C., and Reisenauer, H. M. (1988). The role of root exudates in nutrient acquisition. Adv. Plant Nut% 3,79- I 14. Wehrmann, J., and Coldewey-Zum Eschenhoff, H. (1986). Distribution of nitrate, exchangeable and non-exchangeable ammonium in the soil-root interface. Plant Soil 91,42 1424. Weinberger, P., and Yee. D. (1984).The influence of nitrogen sources on root-mediated changes in substrate pH. Can. J. Eof.62, 161-162. Weisenseel, M. H., Dorn, A,, and Jaffe, L. F. (1979). Natural H + currents traverse growing roots and root hairs of Barley (Hordeum vulgare L.). Planr Physiol. 6 4 , 5 12-518. Wirkn, N. von, Romheld. V.. Morel, J. L., Gucken, A,, and Marschner, H. (1993). Influence of microorganisms on iron acquisition in maize. Soil Eiol. Eiochem. 25,371-376. Wirkn, N. von, Mori, S., Marschner, H., and Romheld, V. (1994). Iron inefficiency in maize mutant ysl (Zea mays L. cv. Yellow-Stripe) is caused by a defect in uptake of iron phytosiderophores. Planr Phv.~iol.106,71-77. Youssef, R. A,, and Chino, M. (1987). Studies on the behavior of nutrients in the rhizosphere. I. Establishment of a new rhizobox system to study nutrient status in the rhizosphere. J. Plant Nutr: 10, 1185-1195. Youssef, R. A,, and Chino, M. (1989).Root-induced changes in the rhizosphere of plants. I. pH changes in relation to the bulk soil. Soil Sci. Plant Nut% 35,461468.

This Page Intentionally Left Blank

Index A Adsorption4esorption processes of arsenic description of, 169-170 effect of phosphate on, 170 factors that affect clay minerals, 17 I competing ions, 174 -176 index cations, 176 ionic strength. 176 magnesium oxides, 172 pH, 173-174 soil properties, 171-173 kinetics of, 176-1 79 mechanisms involved in, 169-170 of arsenic acid fulvic acid effects, 175-176 kinetics of, 177-178 magnesium oxide effects, 172 pH effects on, 173-174 of arsenous acid description of, 170 kinetics of, 178-179 magnesium oxide effects, 172 Agricultural operations, arsenic contamination caused by cattle dips, 160-161 herbicides, 161-1 62 pesticide sources, 159-161 Aliphatic acids, 243 Alkyarsines, I86 Apomixis, of pearl millet definition of, 16 genetic basis for, 17 for heterosis, 17-18 prevalence of, 16-17 Apospory, 16 Arsenic anthropogenic sources cattle dips. 160-161 coal, 157-158 description of, I53 fertilizers, 162-1 63

forestry operations, 158-159 herbicides, 161-162 industry, 153-154 mill tailings, 156 mining, 154-157 pesticides, 159-161 tannery wastes, 158 atmospheric deposition description of, 151 from mining operations, 156 bioavailability of, 181 in biota, 163-164 biotransformation of, by microorganisms, I 82- I 85 contamination sites, 150 human exposure to, 164 - 165 hybride-mercury vapor generator determination of trace levels, 97-99 overview of, 150 in periodic table, I50 - I5 I in plants and plant products phytotoxicity, 180. 187 tolerance to arsenic, 180 -182 toxicity, 163-164 uptake, 180 poisoning levels of, 150 methods of, 150 in rocks, 151 in soils description of, 151-153 effect of soil type on phytotoxicity, 181 microorganism effects biotransformdtion by bacteria, 185-1 86 by fungi, 186 description of, 182-185 tolerance mechanisms, 182-1 83 physicochemical behavior adsorption-desorption processes, see Adsorption4esorption processes inorganic compounds, 166-1 67 organic compounds, 167-168 overview, 165, 187

2 68

INDEX

Arsenic (continued) solubility, 168-169 regional variations, 152-153 summary of, 187 toxicity, 163-165 Arsenic acid adsorption-desorption processes fulvic acid effects, 175-176 kinetics of, 177-1 78 magnesium oxide effects, 172 pH effects on, 173-174 chemical formula, 166 Arsenic trioxide, 153 Arsenous acid adsorption4esorption processes description of, 170 kinetics of, 178-179 magnesium oxide effects, 172 chemical formula, 167 As"', see Arsenous acid As", see Arsenic acid Atomic emission spectrometry, 42-43, see also Inductively coupled plasma-atomic emission spectrometry

C Calcium in plant tissue digests, inductively coupled plasma-mass spectrometry detection of, 100 -101 in rhizosphere, root-induced changes in, 228-230 Cation exchange capacity, 129 Cattle dips, arsenic contamination caused by, 160-161 CEC. see Cation exchange capacity Chloride, in rhizosphere, root-induced changes in, 229-230 Chlorophyll definition of, 138 nitrogen status evaluations using, 138 Chlorosis, 119 Coal, arsenic concentrations in, 157-158 Cotton bolls load, foliar nitrogen fertilization and, 133. I42 nitrogen requirements, 122-124 yield and, 122

growth habit description of, 118-119 nitrogen and, interactions between, 121-1 25 leaf and fruit appearance of, 118-1 19 cessation of, I19 nitrogen assays, 137 macronutrients. 1 I6 nitrogen supply crop maturity effects, 125 deficiencies in, 119-121 effect on growth habit, 121-125 fertilization patterns, 116-1 18 foliar fertilization absorption of nitrogen, 132 description of, 132 reasons for, 132 uptake patterns, factors that affect, 133 urea, 132 management methods computer models, 138-139, 142 crop-water use, 139-141 nitrate petiole analysis of, 135-1 36 as principal source, 119 in soil. 129 storage of, 12 1 requirements, 126-128 soil availability crop rotation methods, 131 description of, 128 fertilizers to compensate, 131-132 inorganic sources, 129- I30 organic sources, 130 -13 1 soil-nitrogen cycle, 128-129, 142 status of, methods for monitoring chlorophyll determination, 138 fertilizer tests, 134 nitrate reductase activity, 137-138 overview, 133-134 petiole nitrate analysis, 135-1 36 plant-tissue analyses, 135 soil testing, 134 -135 total leaf nitrogen assay, 137 uptake, factors that influence, 126 Cropping intensification case study of, 205 description of, 199-200

INDEX in Great Plains area, 200 modern dryland-no-till cropping systems, 205-207 overview of, 203, 205 summer cropping, 207-2 12 systems approach to quantitative indices, 2 16221 systems analysis, 213-216

D DDT, see Dichlorodiphenyltrichloroethane Denitrification, 130 Detection limits definition of, 45, 56 of inductively coupled plasma-atomic emission spectrometry, 45,56,75-77 of inductively coupled plasma-mass spectrometry, 75, 77-78 Dichlorodiphenyltrichloroethane, 159 Dimethyarsine, 186 Dimethylarsinic acid, 167 DIN, see Direct injection nebulizers Direct injection nebulizers, 41 Disodium methanearsonate, 161 DMA, see Dimethylarsinic acid Dryland cropping intensification case study of, 205 description of, 199-200 in Great Plains area, 200 modern dryland-no-till cropping systems, 205-207 overview of, 203,205 summer cropping, 207-212 systems approach to quantitative indices, 2 1 6 2 2 1 systems analysis, 2 13-216 DSMA. see Disodium methanearsonate

E Exudates of plant roots effect on metals complexation, 248-249 phosphate hydrolysis, 253 types of mucilage, 254 phosphatase. 253 phytase ectoenzymes, 253 phytosiderophore. 249-252

269 F

Fallow systems, see Summer fallow; Wheatfallow systems Fertilizers arsenic contamination, 162-163 for providing nitrogen supply in soil for cotton growth, I3 1-1 32, 134 Fly ash, 157-158 Fulvic acid, effect on adsorption4esorption of arsenic in soils, 175-176

G Gahi I , 14-15 Generator systems for inductively coupled plasma-atomic emission spectrometry design of, 3 1-32 initial radiation zone, 32 normal analytical zone, 33 properties of, 32-36 for inductively coupled plasma-mass spectrometry design of, 3 1-32 initial radiation zone, 35-36 normal analytical zone, 36 properties of, 34 -36

H Heterosis, of pearl millet description of, I3 for forage yield, 14 -15 germplasm, 15 for grain yield, 13-14 hybrid types, 15-16 Hybride-mercury vapor generator applications arsenic trace levels, 97-99 selenium trace levels, 97-99 description of, 39 Hybridization, of pearl millet with barley, 1I description of, I 1 fodder use of hybrids, 11-1 2 intergenetic hybrids, 13 interspecific hybrids, 12-13 with Napier grass, 12-13

2 70

INDEX

Hybridization, of pearl millet (continued) with oat, 11 seed yields, 15-16

I ICP-AES, see Inductively coupled plasma-atomic emission spectrometry ICP-MS, see Inductively coupled plasma-mass spectrometry Inductively coupled plasma-atomic emission spectrometry accessories for, 43-44 analytical capabilities detection limits, 45, 56, 75-77 wavelength selection, 44 4 5 applications grinding of soil samples, 91 organic matter digestion, 92-97 silicate dissolution, 92-97 soil digests, 93-94 soil extracts, 91-94 argon gas use, 3 1-32 generator systems design of, 3 1-32 initial radiation zone, 32 normal analytical zone, 33 properties of, 32-36 inductively coupled plasma-mass spectrometry and, comparison, 29-31 interferences correction for, 79-83 ionization, 78 solute vaporization, 78 unwanted radiation, 78-79 isotope selection, description of, 45-56 new developments in, 28 quality control methods, 99 spectrometers, 4 2 4 3 Inductively coupled plasma-mass spectrometry accessories for, 43-44 analytical capabilities detection limits, 75, 77-78 wavelength selection, 44 4 5 applications grinding of soil samples, 91 organic matter digestion, 92-97 silicate dissolution, 92-97 soil digests, 93-94

soil extracts, 91-94 costs associated with, 30 -31 development of, 28-29 generator systems design of, 31-32 initial radiation zone, 35-36 normal analytical zone, 36 properties of, 34 -36 interferences correction methods, 87-91 description of, 83 isobaric correction for, 88 description of, 83 mass discrimination, 83.86-87 nonspectroscopic causes of, 30 correction for, 89 description of, 84 -86 solids deposition on sampler and skimmer cones, 83-84 spectral, 30 unwanted ions, 87 isotope monitoring, 29 isotope selection description of, 45-56 for determining trace elements in plant tissue digests calcium, 100-101 iron, 102-103 lead, 105-106 nickel, 103 zinc, 104-105 problems associated with, 29-3 1 quality control methods, 99 sample introduction systems direct injection nebulizers, 41 hybride-mercury vapor generator, 39 laser sampling of solids, 3 9 4 1 nebulizers, 36-38 types of, 39 spectrometers, 43-44 studies of, 28-29 Interferences inductively coupled plasma-atomic emission spectrometry correction for, 79-83 ionization, 78 solute vaporization, 78 unwanted radiation, 78-79

INDEX inductively coupled plasma-mass spectrometry correction methods, 87-91 description of, 83 isobaric correction for. 88 description of, 83 mass discrimination, 83, 86-87 nonspectroscopic causes of, 30 correction for, 89 description of, 84 -86 solids deposition on sampler and skimmer cones, 83-84 spectral, 30 unwanted ions, 87 Ions, in rhizosphere, root-induced changes calcium, 228-230 chloride, 229-230 magnesium, 228-230 phosphate, 230 -235 potassium, 230 -23 I sodium, 229-230 Iron oxidation of, root-induced, 245-246 oxyhydroxides, 242-243 phytosiderophore-inducedcomplexation of, 250 -25 I in plant tissue digests, inductively coupled plasma-mass spectrometry detection of, 102- I03 in soil, effect of pH level increases, 239-240, 244 Isotope selection, for spectrometric analysis, 45-56

L Lead, in plant tissue digests, inductively coupled plasma-mass spectrometry detection of, 105- 106 Legumes, as nitrogen source for cotton, 13 1

M Magnesium, in rhizosphere, root-induced changes in, 229-230 Metals inductively coupled plasma-mass spectrometric detection of, 41 root-induced complexation of, 247-253

27 1

Methanearsonates, 162 Millet, see Pearl millet Mineralization immobilization turnover, 130 Mining, arsenic contamination caused by, 154 -157 MIT, see Mineralization immobilization turnover Monomethylarsonic acid, 167 Monosodium methanearsonate, 16 1 MSMA, see Monosodium methanearsonate Mucilage, 254

N Napier grass chromosomes comparison with pearl millet chromosomes, 8 description of, 7 hybridization with pearl millet, 12-13 Nebulizers Babington, 38 description of, 36 direct injection, 41 pneumatic, 37 ultrasonic, 38-39 Neodymium-yttrium-aluminum-garnet laser, 40 Nickel, in plant tissue digests, inductively coupled plasma-mass spectrometry detection of, 103 Nitrates petiole analysis of, 135-136 as principal source, 119 in soil, 129 Storage Of, 12 1 Nitrification, I29 Nitrogen supply, for cotton growth crop maturity effects, 125 deficiencies in, 119-121 effect on growth habit, 121-125 fertilization patterns, 116-1 18 foliar fertilization absorption of nitrogen, 132 description of, 132 determination of need for, 142 reasons for, 132 uptake patterns, factors that affect, 133 urea, 132 management methods computer models, 138-139, 142

272

INDEX

Nitrogen supply, for cotton growth (continued) crop-water use, 139-141 nitrate petiole analysis of, 135-136 as principal source, 119 in soil, 129 storage of, 121 requirements, 126-128 soil availability crop rotation methods, 13I description of. 128 fertilizers to compensate, 131-132 inorganic sources, 129-130 organic sources, I30 - I3 1 soil-nitrogen cycle, 128-129, 142 status of, methods to monitor chlorophyll determination, 138 fertilizer tests, 134 nitrate reductase activity, 137-1 38 overview, 133-1 34 petiole nitrate analysis, 135-136 plant-tissue analyses, 135 soil testing, 134 -135 total leaf nitrogen assay, 137 uptake, factors that influence, 126 Nonspectroscopic interferences. 84 -86

P Pearl millet aneuploids of. 10 apomixis definition of, 16 genetic basis for, 17 for heterosis, 17-18 prevalence of, 16-17 characteristics of, 2-3 chromosome complement, 7 chromosome number and size. 6 classification of, 4 consumer uses of, 2 cultivation of, 2, 20 diversity of, 5 effect of sex on recombination in, 1 1 gene mapping of, 10 -1 1 genome relationships, 8-10 heterosis of description of, I3 for forage yield, I I , 14 -15 germplasm. 15

for grain yield, 13-14 hybrid types, 15-16 hybridization of with barley, 11 description of, 11 fodder use of hybrids, 11-12 intergenetic hybrids, 13 interspecific hybrids, 12-1 3 with Napier grass, 12-1 3 with oat, 1I seed yields. 15-16 nutritional quality of, 20 -2 I origin of, 3-4 perennial relatives of, 6 qualitative traits of, 18-1 9 quantitative traits of, I9 Sclerosporu graminicolu effects on, 20 taxonomic placement of, 4 -5 wild annual relatives of description of, 5 reproductive barrier with cultivated types, 5 Penning ionization, 85 Penniseium spp. apomixis in, 16-17 chromosomes of. 6 description of, 2 t? mollissimum, 7 F1 orientale, 10 t? purpureum, 6 t? violuceum, 7 pH, in rhizosphere, root-induced changes, 237-242 Phenolics, 243 Phosphate effect on arsenic levels adsorption-desorption in soils, 174 in plants, I82 hydrolysis of, by root exudates, 253 rhizosphere concentrations of, root-induced changes, 230 -235 soil pH effects on, 239 Ph y tosiderophores definition of, 249-250 metals complexation, 250 -252 Phytotoxicity, from arsenic effect of soil type on, I8 I signs of, 180 Plant roots description of, 225-226 exudates

INDEX mucilage, 254 phosphatase, 253 phytase ectoenzymes, 253 phytosiderophore, 249-252 rhizosphere changes related to ionic concentrations calcium, 228-230 chloride, 229-230 magnesium, 228-230 phosphate, 230 -235 potassium, 230 -23 1 sodium, 229-230 metal complexation. 247-253 nutrient deficiency-related etiology, 255 pH. 237-242 proton effluxes, 240 redox conditions, 242-247 Polyuronic acids, 254 Potassium, root-induced changes in rhizosphere concentrations. 230 -23 I Precipitation, dryland agriculture and, 199-200

R Rhizosphere acidification of, 240 -24 I alkalinization of, 240 -24 I chemical conditions in, 226 definition of, 226-228 plant species and. 228 root-induced changes ionic concentrations calcium, 228-230 chloride, 229-230 magnesium, 228-230 phosphate, 230 -235 potassium. 230 -23 I sodium, 229-230 metal complexation. 247-253 nutrient deficiency-related etiology, 255 pH. 237-242 proton effluxes. 240 redox conditions. 242-247 root symbionts in, 226 Rhizosphere effect, 247 Roots, see Plant roots Root symbionts, 226

273 S

Scanning monochromator spectrometer, 4 2 4 3 Sclerospora graminicola, 20 Selenium, hybride-mercury vapor generator determination of trace levels, 97-99 Sodium, in rhizosphere, root-induced changes in, 229-230 Soil arsenic in description of, 151-153 effect of soil type on phytotoxicity, 18 I microorganism effects biotransformation by bacteria, 185-186 by fungi, 186 description of, 182-185 tolerance mechanisms, 182-1 83 physicochemical behavior adsorption4esorption processes, see Adsorption-desorption processes inorganic compounds, 166-167 organic compounds, 167-168 overview, 165, 187 solubility, 168-169 regional variations, 152-153 ionic concentrations, root-induced changes in calcium, 228-230 chloride, 229-230 magnesium, 228-230 phosphate, 230 -235 potassium, 230 -23 1 sodium, 229-230 nitrogen supply in, for cotton growth crop rotation methods, I3 I description of, I28 fertilizers to compensate, 131-132 inorganic sources, 129-1 30 organic sources. I30 -1 3 1 soil-nitrogen cycle, 128-129, 142 redox conditions, root-induced changes of, 242-247 Spectrometers analytical capabilities ICP-AES detection limits, 45, 56, 75-77 ICP-MS detection limits, 75, 77-78 isotope selection, 45-56 wavelength selection, 44 4 5 atomic emission, 4 2 4 3 direct readers, 42

2 74

INDEX

Spectrometers (continued) ICP-AES, see Inductively coupled plasma-atomic emission spectrometry ICP-MS, 43-44, see also Inductively coupled plasma-mass spectrometry scanning monochromators, 42-43 SPSI. see System precipitation storage index Summer fallow history of, 197-198 paradox of, 198-200 spring wheat-fallow system, 201-203 winter wheat-fallow system, 203 System precipitation storage index, 217218

T Trace elements, in plant tissue digests, isotope selection for ICP-MS determination of calcium, 100-101, see also Calcium iron, 102-103, see also Iron lead, 105-106 nickel, 103 zinc, 104 -105

Trimethylarsine, 186 Triploid hybrid, 12-13

W Water conservation, using summer fallow history of, 197-198 paradox of, 198-200 spring wheat-fallow system, 201-203 winter wheat-fallow system, 203 Wheat-fallow systems description of, 198 soil-water storage average amount, 199 tillage and residue management effects, I99 spring, 201-203 winter, 203

Z Zinc, in plant tissue digests, inductively coupled plasma-mass spectrometry detection of, 104-105

This Page Intentionally Left Blank