ADVANCES IN
AGRONOMY
VOLUME 41
This Page Intentionally Left Blank
ADVANCES IN
AGRONOMY Prepared in Cooperation w...
26 downloads
1911 Views
23MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
ADVANCES IN
AGRONOMY
VOLUME 41
This Page Intentionally Left Blank
ADVANCES IN
AGRONOMY Prepared in Cooperation with the AMERICAN SOCIETY OF AGRONOMY
VOLUME 41 Edited by
N. C. BRADY
Science and Technology Agency for International Development Deportment of State Washington, DC
ADVISORY BOARD
T. M. STARLING G. H. HEICHEL H. J. GORZE. J. KAMPRATH R. J. KOHEL G. E. HAM
E. L. KLEPPER R. H. FOLLETT D. R. BUXTONE. S. HORNER
J. J. MORTVEM
ACADEMIC PRESS, INC. Harcourt Brace Jovanovich, Publishers
San Diego New York Berkeley Boston London Sydney Tokyo Toronto
COPYRIGHT 0
1987
BY ACADEMIC PRESS,
INC.
ALL RIGHTS RESERVED. NO PART OF THIS PUBLICATION MAY BE REPRODUCED OR TRANSMI'ITED 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.
ACADEMIC PRESS, INC. 1250 Sixth Avenue San Diego, California 92101
United Kingdom Edition published by ACADEMIC PRESS INC. (LONDON) LTD. 24-28 Oval Road, London NW1 7DX
LIBRARY OF CONGRESSCATALOGCARDNUMBER:50-5598
ISBN 0-12-OOO741-X (alk. paper)
PRINTED IN THE UNITED STATES OF AMERICA 87 88 89 90
9
8 7 6 5 4 3 2
I
CONTENTS PREFACE ......................................................
ix
SPECIATION. CYTOGENETICS. AND UTILIZATION OF Arachis SPECIES
.
H . T. Stalker and J . P Moss I . Introduction ............................................. I1. Botany and Taxonomy .................................... 111. Plant Collection and Maintenance .......................... I v. Centers of Origin ........................................ V. Cytogenetics of Aruchis Species ............................ v1. Interspecific Hybridization in Arachis ....................... VII . Germplasm Evaluation .................................... VIII . Utilization of Wild Aruchis Species ......................... IX . Successes and Potentials for Utilizing Arachis Germplasm . . . . . X . Conclusions ............................................. References ...............................................
1 3 4 7 8 10 14 22 32 34 35
CEREAL-LEGUME INTERCROPPING SYSTEMS
Francis Ofori and W. R . Stern I . Introduction ............................................. I1 . Background to Intercropping Systems ....................... 111. Competitive Relationships between Component Crops . . . . . . . . . I v. Some Agronomic Factors Influencing Productivity and Efficiency ................................ V. Nitrogen Economy of the System ........................... VI . Summary and Conclusions ................................ References ...............................................
41 42 52 61 72 83 85
GENOTYPIC VARIATION IN CROP PLANT ROOT SYSTEMS
J . C. O’Toole and W. L. Bland 1. Introduction ........................................... 11. Significance of Genotypic Variation in Root Systems ....... V
91 92
vi
CONTENTS Evidence of Genotypic Variation ........................... IV. Utilization in Research and Crop Improvement.. . . . . . . . . . . . . . V. Phenotypic Plasticity ..................................... V1. Conclusions and Research Needs ........................... References ...............................................
111.
94 120 133 139 140
APPLICATION OF CELL AND TISSUE CULTURE TECHNIQUES FOR THE GENETIC IMPROVEMENT OF SORGHUM, Sorghum bicolor (L.) Moench: PROGRESS AND POTENTIAL
S. Kresovich, R. E. McGee, L. Panella, A. A. Reilley, and F. R. Miller Introduction ............................................. Background ............................................. Goals in Breeding ........................................ 1V. Progress in Cell and Tissue Culture Research . . . . . . . . . . . . . . . . V. Potential Applications .................................... VI. Summary ............................................... References ............................................... I.
11. 111.
147 148 150 153 162 166 168
UPTAKE OF MINERAL NUTRIENTS AND CROP GROWTH: THE USE OF FLOWING NUTRIENT SOLUTIONS
A. Wild, L. H. P. Jones, and J. H. Macduff Introduction ............................................. Systems Employing Flowing Nutrient Solutions . . . . . . . . . . . . . . Nutrient Uptake ......................................... I v. Partitioning of Photosynthate between Shoots and Roots. . . . . . V. Conclusions and Summary ................................ References ...............................................
1. 11. 111.
171 172 177 21 1 214 216
MINERAL NUTRITION OF LINSEED AND FIBER FLAX
Peter J. Hocking, Peter J. Randall, and Andrew Pinkerton I. Introduction ............................................. Production, Crop Growth, and Agronomy. . . . . . . . . . . . . . . . . . . Effects of Mineral Nutrition on Yield.. ..................... I v. Nutrient Disorders. ....................................... V. Salinity ................................................. VI. Soil Acidity.. ............................................ VII. Interactions between Genotype and Mineral Nutrition. . . . . . . . . VIII. General Conclusions and Challenges ........................ References ............................................... 11. 111.
22 1 223 225 25 5 27 8 28 1 283 285 287
CONTENTS
vii
THE IMPACT OF SOIL AND FERTILIZER PHOSPHORUS ON THE ENVIRONMENT
A. N. Sharpley and R. G. Menzel I. Introduction ............................................. 11. Impact of Phosphorus on the Terrestrial Environment . . . . . . . . 111. 'Ransport of Phosphorus from the Terrestrial to Aquatic Environments ........................ IV. Impact of Phosphorus on the Aquatic Environment . . . . . . . . . . V. Conclusions ............................................. References ...............................................
297 298 301 311 317 319
BIOTECHNOLOGY APPLICATIONS IN WEED MANAGEMENT: NOW AND IN THE FUTURE
Kriton K. Hatzios Introduction ............................................. Production and Use of Biological Weed Control Agents.. . . . . . Naturally Occurring Herbicides ............................ Genetic Improvement of Crop Tolerance to Herbicides . . . . . . . . Additional Uses of Genetically Engineered Microorganisms in Weed Management ...................... VI. Conclusions and Future Prospects .......................... References ...............................................
I. 11. 111. IV. V.
325 326 332 337 357 364 367
RECLAMATION OF ACIDIC MINED LANDS IN HUMID AREAS
P. Sutton and W. A. Dick I. Introduction ............................................. 11. Properties of Spoil on Acidic Mine Lands.. . . . . . . . . . . . . . . . . . 111. Application of Various Amendments to Ameliorate Acidic Mine Spoil. ............................. IV. Seeding and Management of Amendment-Treated Spoil . . . . . . . V. Changes in Soil Properties as Affected by Addition of Amendments ................................. VI. Summary of Long-Term Results.. .......................... References ...............................................
371 379 383 389 394 40 1 40 1
viii
CONTENTS WATER AND QUALITY LOSS DURING FIELD DRYING OF HAY
Andy D . Macdonald and E . Ann Clark 1 . Introduction ............................................. I1 . Patterns of Water Loss .................................... I11. Factors Influencing Water Loss during Field Drying of Hay . . . IV. Losses during Forage Conservation. Storage, and Handling . . . . V. Preservation of Wet Hay .................................. VI . Conclusions ............................................. References ...............................................
INDEX .........................................................
407 407 409 417 430 432 435
439
PREFACE
The international status of Advances in Agronomy continues to prevail as illustrated by this volume. Authors from five countries have participated in preparing ten review articles. This participation confirms the wide-ranging interest in crop and soil science. There is a variety of topics covered this year, ranging from reviews of research to improve the yield and quality of two crops of worldwide importance, sorghum and groundnuts, to the reclamation of land around abandoned mines. The two crop improvement reviews show evidence of the use of modern biotechnology to alter genetically and to improve these important food crops. They are examples of what will likely become more and more common as new and improved methods of genetic improvement are applied to crop plants. Two articles are concerned with mineral nutrition. The first deals with one specific crop (flax) and focuses on the effects of nutrient stress on vegetative development, yield components, and yield quality. The second summarizes the advances made in understanding the requirement for nutrients through the use of flowing nutrient solutions which the authors have researched in depth. The effects of environmental factors on the genetic variation in root systems are also reviewed, with the literature covering a number of important field crop species. There appears to be considerable variation that could be used in crop improvement programs. Intraplant and whole-plant factors influencing water and quality loss of hay during drying are also considered. An excellent review is that covering the positive and negative effects of soil and fertilizer phosphorus on the terrestrial environment. Attention is given to potentially toxic heavy metals, which are commonly added with phosphate fertilizers, and to means of reducing phosphate levels in lakes. Cereal-legume intercropping systems that have become increasingly more common in the tropics are well reviewed by scientists who have had considerable experience in this research area. Such systems will likely receive more attention in the future, especially where low-input agriculture is being practiced. The management of weeds using biotechnological methods is the focus of another article. Increasing crop tolerance to herbicides promises to be one of the more exciting aspects of applied biotechnology. We will likely read more on this topic in the future. My thanks to the 24 contributors who prepared these articles. Their efforts should be appreciated by their fellow agronomists around the world.
N. C. BRADY ix
This Page Intentionally Left Blank
ADVANCES IN AGRONOMY, VOL. 41
SPECIATION, CYTOGENETICS, AND UTILIZATION OF Arachis SPECIES' H. T. Stalker2 and J. P. Moss3. *Department of Crop Science, North Carolina State University Raleigh, North Carolina 27695 'international Crops Research institute for the Semi-Arid Tropics (ICRISAT) Patancheru, P.O., Andhra Pradesh 502 324, India
I.
INTRODUCTION
Utilization of wild species for improvement of cultivated forms has been investigated since Faircloth made the first interspecific hybrids in 1717. Many wild species have been of value in crop improvement for a large number of traits (Harlan, 1976; Hawkes, 1977; Stalker, 1980a; Hadley and Openshaw, 1980). However, in leguminous oilseeds, utilization of species germplasm has proven difficult, in large part because of barriers to interspecific hybridization between species (Smartt, 1979). Further, sterility often restricts introgression from wild to cultivated accessions even when initial hybridization is possible. Interspecific hybridization is also difficult among the peanut species in the genus Arachis, but breeding populations derived from crossing A . hypogaea L. with related species are currently being evaluated for farmer use (Moss, 1985b). Four species of Arachis have been cultivated, including two diploids (2n = 2x = 20: A . villosulicarpa Hoehne and A . repens Handro) and two tetraploids (2n = 4x = 40:A . glabrata Benth. and A . hypogaea). Arachis villosulicarpa has only been cultivated by Indians in the northwestern part of the Brazilian state of Mato Grosso (Gregory et al., 1973). Arachis repens and A . glabrata have been grown in different parts of South America as forages or as ground covers in urban areas. Arachis glabrata has also been selected for forage qualities in Florida, where recent cultivar releases have been made (Prine et al., 1981). However, A . hypogaea is the only species which is cultivated extensively for commercial production of seeds and oil. 'This article is a contribution from paper No. 10726 of the Journal Series of the North Carolina Agricultural Research Service, Raleigh, North Carolina 27695, and Paper No. 635 of the Journal Series of the International Crops Research Institute for the Semi-Arid Tropics (ICRISAI), Patancheru, P.O., Andhra Pradesh 502 324, India. 1 Copyright 0 1987 by Academic Press. Inc. All rights of reproduction in any form reserved.
2
H. T. STALKER AND J. P. MOSS
The cultivated peanut, A . hypogaea, is a major crop in most tropical and subtropical areas of the world. As compared to other oilseeds, plants are relatively drought resistant, which makes them especially important in semiarid regions where precipitation exceeds evaporation for only 2-7 months per year (Bunting et al., 1985). Although the crop grows best in sandy, well-drained soils, peanuts are culfivated in a wide range of field conditions from clays to sands and from acidic to alkaline soils. In the United States, the seeds are used mainly for confectionary purposes or for peanut butter. However, approximately two-thirds of the total world peanut crop is used for oil, which cumulatively represents 20% of the world market (Woodroof, 1973). Most peanuts are used as a cash crop and even small farmers may sell their entire harvest. In addition to seeds being of high value, plant residues are extremely important as fodder for cattle in many regions of the world. Shells are also used for fuel, soil conditioners, fodder, chemicals, resin extenders, cork substitutes, and for hardboard (Gibbons, 1980). The peanut is becoming increasingly important as an income source in tree plantations, such as coconut, rubber, or banana, before tree crops mature. In Africa and Asia, many peanuts are intercropped between maize, sorghum, pulses or, in a few areas, between mature coconut trees. While A . hypogaea has the widest distribution of any Arachis species, it also has many pests and diseases which attack all parts of the plant. Disease controls are seldom practiced in less developed countries, and one of the major plant breeding objectives is to improve resistances to plant pests (Wynne and Gregory, 1981). In addition, increased stability of yield and adaptation of the peanut to many environments are important breeding objectives. Several Aruchis species have significantly higher levels of resistance for many important disease and insect pests than are found in cultivated accessions. As in the cases of other crop species, however, attempts to utilize wild species have limitations and obstacles. For example, sterility barriers and genomic incompatibilities restrict utilizing many potentially important genes even after initial interspecific hybids are obtained. However, significant progress has been made toward utilizing germplasm resources of Arachis. Collection, maintenance, and evaluation of peanut germplasm resources have occurred relatively late compared to many other crop species such as tobacco, wheat, or maize. Only during the past 20-25 years have concentrated efforts been undertaken to collect and evaluate the agronomic potential of germplasm resources in the genus. This chapter reviews the current status in the introgression of germplasm from wild to cultivated species of peanut. The successes have been due to a knowledge of the botany, taxonomy, cytogenetics, and genetics of the species related to A . hypogaea and to effective screening of the available germplasm. Therefore, summaries of
SPECIATION, CYTOGENETICS, AND UTILIZATION
3
species variability, genomic and species relationships, and resources of agronomically important characteristics are presented, followed by an account of the methods used and of achievements in Arachis germplasm utilization.
II.
BOTANY AND TAXONOMY
The genus Arachis belongs to the family Leguminosae, tribe Aeschynomeneae, subtribe Stylosanthenae. Species have alternately attached basal and dorsal anthers, flowers in terminal or axillary spikes or small heads, pinnate leaves, and few leaflets without stipules (Taubert, 1894). Most of the Arachis species have tetrafoliate leaves, but two species, A. tuberosa Benth. and A . guaranitica Chod. et Hassl., are trifoliate. Arachis species, along with Trifoliurnsubterraneum L. and Vigna subterraneu (L.) Verdc., flower above ground but produce fruit and seeds below the soil level. Species of peanut are self-pollinating, but outcrossing may also occur in up to 2.5% of the flowers (Norden, 1980). A structure called a peg grows geotropically after fertilization and carries the developing embryo into the soil, after which elongation ceases, the pod expands, and the embryo initiates a rapid growth phase. Both annual and perennial members of the genus are found in nature. The first peanut species described was A. hypogueu by Linnaeus in 1753. This species has two subspecies, each of which has two botanical varieties (Table I). Not until 1841 were wild species described, including A. villosu Benth., A . tuberosa, A. glabrata, and A . pusilla Benth. (Bentham, 1841). Taxonomic treatments of the genus were later completed by Chevalier
Table I Subspecific and Varietal Classificationof A . hypogaea Subspecies
Variety
Type
hypogueu
hypogaea
Virginia
hirsuta
Peruvian runner
fastigiuta
Valencia
vulgaris
Spanish
fastigiuta
Comments No floral axes on main stem; alternating pairs of floral and vegetative axes on branches; branches short, less hairy No floral axes on main stem; alternating pairs of floral and vegetative axes on branches; branches long, more hairy Floral axes on main stem; sequential floral axes on branches; little branched, curved branches Floral axes on main stem; sequential floral axes on branching; more branched, upright branches
H. T. STALKER AND J . P. MOSS
4
(1933, 1934, 1936), Hoehne (1940), and Hermann (1954). Currently, 22 species diagnoses have been published (excluding A . nambyquarae Hoehne, which is a cultivar of A . hypogaea, and A . batizogaea Krap. et Fern., which originated from a man-made hybrid), with another 12 names commonly used in the literature but not described (Table 11). The guidelines outlined by Resslar (1980) will be used in this paper even though proper diagnoses for many of the taxa have not been published. A complete revision of the taxonomy of the genus is greatly needed, and considering the vast number of new accessions collected within the past 10 years, perhaps 50 or more additional species will eventually be described. Compounding the problem of species names are those designations which have been incorrectly given to taxa in the literature. For example, A . nambyquarue Hoehne is not a separate species, but a cultivar of A . hypogaea. The name A . diogoi Hoehne (a species of section Arachis ) was incorrectly used for an unnamed section Erectoides species by Johansen and Smith (1956), and the names A . prostrata Benth. and A . marginata Gard. have been used incorrectly for many different accessions (Gregory et al., 1973). Since germplasm is widely distributed with collection numbers and names, and there is no written description of many species, misidentifications are easily made. Peanut researchers have associated collection numbers with taxa to circumvent partially the problem of nomenclature. Based on morphological comparisons, and to a lesser extent on crosscompatibility and pollen stainability of interspecific hybrids, Krapovickas (1969, 1973) and Gregory et al. (1973) proposed sectional classifications for the Arachis species. The classification suggested by Gregory et al. (1973) is more commonly adopted and will be followed throughout this review (Table 11). Although sectional names remain tentative because proper diagnoses have not been published, the groups are useful for defining general crossing relationships of taxa in the genus.
111.
PLANT COLLECTION AND MAINTENANCE
Taxa in the genus Arachis are widely distributed in South America from the Atlantic Ocean to the foothills of the Andes Mountains and from the mouth of the Amazon River in the North to approximately 34"s in Uruguay. Although the cultivated species usually prefers sandy, welldrained soil, wild species of Arachis are found in many habitats, including rocky areas, heavy soils, marshy areas, and even in running water (Valls et al., 1985). While many species proliferate in shaded areas, others prefer open and sunny environments. Plants are most often found in ecotypes such as rock outcroppings, broken forested areas, forest-grassland margins, or
Table I1 Taxonomic Subdivision of the Genus Arachif‘ Section Arachis nom. nud. Series Annuae Krap. et Greg. nom. nud. (2n = 2x = 20) A. batizocoi Krap. et Greg. A. duranensis Krap. et Greg. norn. nud. A . spegauinii Greg. et Greg. norn. nud. A . stenosperma Greg. et Greg. nom. nud. A . ipaensis Greg. et Greg. nom. nud. A . spinaclava
Series Perennes Krap. et Greg. norn. nud. (2n = 2x = 20) A. helodes Martius ex Krap. et Rig. A . villosa Benth. var. villosa A. villosa var. correntina Burkart [A. correntina (Burk.) Krap. et Greg. norn. nud.] A . diogoi Hoehne A. cardenasii Krap. et Greg. nom. nud. A . chacoense Krap. et Greg. norn. nud. Series Amphiploides Krap. et Greg. nom. nud. (2n = 4x = 40) A. hypogaea L. ( A . nambyquarae Horne) A . monticola Krap. et Rig. A . x batizogaea Krap. et Fern. (of experimental hybrid origin) Section Erectoides Krap. et Greg. norn. nud. (2n = 2x = 20) Series Trifoliolatae Krap. et Greg. norn. nud. A. guaranitica Chod. et Hassl. A . tuberosa Benth. Series Tetrafoliatae Krap. et Greg. norn. nud. A. benthamii Handro A. martii Handro A. paraguariensis Chod. et Hassl. A. oteroi Krap. et Greg. norn. nud. Series Procumbensae Krap. et Greg. norn. nud. A . rigonii Krap. et Greg. A. lignosa (Chod. et Hassl.) Krap. et Greg. norn. nud. Section Caulorhizae Krap. et Greg. nom. nud. (2n = 2x = 20) A . repens Handro A . pintoi Krap. et Greg. nom. nud. Section Rhizomatosae Krap. et Greg. nom. nud. Series Prorhizomatosae Krap. et Greg. nom. nud. (2n = 2x = 20) A. burkartii Handro Series Eurhizomatosae Krap. et Greg. norn. nud. (2n = 4x = 40) A . glabrata Benth. A. hagenbeckii Harms Section Extranervosae Krap. et Greg. nom. nud. (2n = 2x = 20) A . marginata Card. A . lutescens Krap. et Rig. A. villosulicarpa Hoehne A . macedoi Krap. et Greg. norn. nud. A. prostrata Benth. Section Ambinervosae Krap. et Greg. nom. nud. (2n = 2x = 20) (no species names, valid or invalid, have been given to forms in this section) Section Triseminalae Krap. et Greg. norn. nud. (2n = 2x = 20) A . pusilla Benth. Uncertain sectional affinity A . angustifolia (Chod. et Hassl.) Killip “After Gregory et al. (1973) and Resslar (1980).
6
H. T. STALKER AND J. P. MOSS
in disturbed habitats..They grow from sea level to approximately 1600 m in elevation. The largest number of taxa are found in the west central region of Brazil walls et al., 1985), with the second highest concentration found in Bolivia. Extensive genetic diversity exists in the genus for many traits of agronomic importance. Seeds of the cultivated peanut were among the earliest crops introduced to Europe from the New World and species have been periodically collected in South America since its first discovery. However, not until the late 1950s were concentrated efforts made systematically to collect and preserve variability in Arachis. This was largely due to the inaccessibility of many parts of South America and to the wide geographic distributions of peanut species. Twenty-four expeditions were organized between 1958 and 1983, and 639 wild species accessions plus 961 accessions of A . hypogaeu were collected walls et ul., 1985). Nearly a hundred wild species accessions have also been collected since 1983. Table I11 summarizes major germplasm collections of Arachis species. Priorities for future Arachis germplasm collection in South America for both cultivated and wild species of the genus have been established (Valls et al., 1985). The highest priority for collecting Arachis species is in the Brazilian states of Mato Grosso and Mato Grosso do Sul, and the second priority is for Paraguay. Although Bolivia also represents an important area for future collections, expeditions are currently not planned due to inaccessibility of some areas of the country. Germplasm resources of wild Arachis species are difficult to maintain due to specialized adaptations to many environments. For example, many species are adapted to arid climates, while others are found in wet habitats, and these extremes are difficult to duplicate. Many species accessions do not produce Table 111 Wild Arachis Accessions Collected and Conserved between 1936-1983O Conserved (1983) Section
Collected number
USA
Total
Ambinervosae Arachis Caulorhizae Erectoides Extranervosae Rhizomatosae Triseminalae Totals
20 171 17 345 99 209 9 -
8 150 11 77 45 93 5 -
159 17 82 72 110 9 -
870
389
464
“After Valls et at. (1985)
I5
SPECIATION, CYTOGENETICS, AND UTILIZATION
7
seeds when grown in the United States and, therefore, must be maintained as live plants. Many other accessions will produce seeds at one location and not another, so multiple germplasm storage facilities are required for seed increase and maintenance. Initiation of reproductive development in peanut species has not been adequately investigated, but many environmental factors probably influence pegging and pod development, such as photoperiod, heat, endogenous hormone levels, and plant stresses. A general trend in section Arachis species is profuse flowering in long-day photoperiods with a higher rate of peg formation in shortday photoperiods (Stalker and Wynne, 1983). However, several species [such as A. chacoense Krap. et Greg. nom. nud., A. correntina (Berk.) Krap. et Greg. nom. nud., and A. villosa] produce few to no flowers under shortday conditions. Investigations are urgently needed to find methods to induce seed set because of the expense associated with propagating germplasm collections as vegetative plants plus the required duplications at several locations to ensure long-term survival of accessions under cultivation.
IV.
CENTERS OF ORIGIN
The center of orgin for Aruchis species was most likely in central Brazil (Gregory et al., 1980). The geocarpic habit of the plant suggests that longdistance dispersal has been along water courses. Gregory et al. (1973) presented a theory that the most ancient species were found at high elevations and more recent speciation has occurred as seeds were washed down toward the sea and became isolated. To support this view, they noted that many species are adapted to highland conditions by having tuberoid roots, tuberiform hypocotyls, or rhizomes. Further, as seeds moved to lower elevations they became isolated in major river valleys and different sections of the genus evolved in parallel evolution. Although species in different sections of the genus were once believed to be isolated, considerable overlaps in distributions occur, especially for members of the sections Arachis, Erectoides, Extranervosae, and Rhizomatosae (Valls et al., 1985). Since the major sectional groups of the genus have widespread distributions, species most likely diverged early in the evolutionary history of the genus and subsequently distributed along watersheds. The cultivated species A. hypogaea probably originated from a wild allotetraploid species (Smartt and Gregory, 1967). Arachis monticola Krap. et Rig. is the only tetraploid known to be cross-compatible with A. hypogaea and the most likely direct progenitor. Since this species is found only in the southern Bolivia-northern Argentina region, this is the region of the presumed center of origin for the cultivated peanut (Krapovickas,
H. T. STALKER AND J. P. MOSS
8
1968). Although the tetraploid progenitor species is generally considered to be A. monticola, much speculation has centered around designating the diploid species which gave rise to the allotetraploid. Krapovickas et al. (1974) indicated that A. butizocoi Krap. et Greg. is one of the diploid progenitors and the species is now considered to be the donor of the B genome of A. hypogaea (Smartt et al., 1978a,b; Smartt and Stalker, 1982). The donor of the A genome is more elusive, however, and several species have been suggested, including A. villosu (Varisai Muhammad, 1973), A. duranensis Krap. et Greg. nom. nud. (Seetharam et al., 1973; Gregory and Gregory, 1976) and A. curdenusii Krap. et Greg. nom. nud. (Gregory and Gregory, 1976; Smartt et ul., 1978a). Because of distribution patterns and probable centers of origin of the cultivated peanut, diploid species of section Aruchis, now found far from the Bolivia-Argentina region, can most likely be eliminated as possible direct ancestors. However, as many unique taxa have been collected in Bolivia, and many more are probably still to be found, the donor of the A genome may await discovery. In addition to the primary center of origin, five secondary centers of variability exists for the cultivated species in South America (Gregory and Gregory, 1976; Wynne and Coffelt, 1982). Africa represents another center of diversity for the cultivated peanut (Gibbons et al., 1972).
V.
CYTOGENETICS OF Arachis SPECIES
The chromosome number of 2n = 40 was first reported by Kawakami (1930) for A. hypogaea. Husted (1931, 1933, 1936) confirmed the ploidy level and analyzed the meiotic and somatic chromosomes of seven cultivars. The meiotic chromosomes of A. hypogueu pair mostly as 20 bivalents, but a few multivalents have also been observed (Husted, 1936). Hybrids among subspecific accessions have mostly bivalents at metaphase I, but univalents also exist at a low frequency. Husted (1936), Raman (1976), and Stalker (1980b) concluded that chromosome structural differences exist between the subspecies hypogueu and fmtigiuta. Further, Gregory et al. (1980) observed reduced fertility in hybrids between subspecies, and genetic differences have been reported between the subspecies hypogaea and fustigiuta (Krapovickas, 1973; Wynne, 1974). The somatic chromosomes of A. hypogueu are small and most have a median centromere. Husted (1933, 1936) analyzed somatic chromosomes of several cultivars and distinguished a pair of small chromosomes, which he termed “A” chromosomes, and one pair with a secondary constriction, which he termed “B” chromosomes. Babu (1955) reported several types of secondary constrictions in A. hypogueu, and cultivars can be distinguished
SPECIATION, CYTOGENETICS, AND UTILIZATION
9
based on karyotypic differences (D’Cruz and Tankasale, 1961; Stalker and Dalmacio, 1986). At least I5 of the 20 chromosome pairs have been distinguished and, based on arm ratios and chromosome lengths, Stalker and Dalmacio (1986) were able to separate members of different botanical varieties based on somatic chromosome morphology. Analyses of somatic chromosomes support previous investigations with meiotic chromosomes of A . hypogaea, which illustrated cytological variation between subspecies. Aneuploidy was first observed in A . hypogaea by Husted (1936), who observed a plant with 41-chromosomes plus a chromosome fragment. Other naturally occurring aneuploids were observed by Spielman et al. (1979) and Stalker (198%) after observing somatic chromosomes of plants propagated from small seeds. Eight different trisomics or double trisomics (2n + 1 + 1) were cytologically verified by Stalker (198%). Chemical treatments (Ashri et al., 1977) or ionizing radiation (Patil and Bora, 1961; Patil, 1968; Madhava Menon et al., 1970) have also produced aneuploid plants. In addition, aneuploids are commonly observed after interspecific A . hypogaea hybrids are colchicine-treated (Smartt and Gregory, 1967; Spielman et al., 1979; Company et al., 1982). Davis and Simpson (1976) reported chromosome numbers ranging from 32 to 48 in derivatives of a 6x (A. hypogaea x A . cardenasii) hybrid. Wild species of Arachis were not analyzed cytologically until the late 1940s. A tetraploid (2n = 40)species, A . glabrata, was reported by Gregory (1946), and Mendes (1947) later observed four diploid species in the genus. Only 26 of 33 named species have chromosome numbers confirmed in the literature (Smartt and Stalker, 1982). Published information on the group is highly inadequate; however, judging from unpublished work in several laboratories and inferences from Gregory and Gregory (1979), most species in the genus are diploid (2n = 20). Polyploid (2n = 40) species are also found in sections Arachis and Rhizomatosae, and Smartt and Stalker (1982) concluded that polyploidy evolved independently in the two groups. Analyses of pollen mother cells (PMCs) indicate that chromosomes of diploid species pair mostly as bivalents (Raman, 1976; Resslar and Gregory, 1979; Smartt et al., 1978a,b; Stalker and Wynne, 1979; Singh and Moss, 1982), but quadrivalents have also been observed at a low frequency in the diploid species A . villosa and A . spegazzinii Greg. et Greg. nom. nud. (Singh and Moss, 1982). In polyploids of a section Rhizomatosae species, Raman (1976) reported up to four quadrivalents in PMCs. A second accession reported by Stalker (1985b) averaged 19.92 bivalents and only 0.04 quadrivalents per PMC. In addition to analyses of chromosome pairing at meiosis, Kirti et al., (1983) and Jahnavi and Murty (1985a,b) analyzed the pachytene chromosomes of species in sections Arachis, Erectoides, Extranervosae, Rhizomatosae, and Triseminafaeand distinguished chromosome pairs. Although chromosomes
H.T.STALKER AND J. P.MOSS
10
did not stain well, Jahnavi and Murty (1985b) concluded that six different chromosomes, three specialized chromosomes, and one nucleolus organizer chromosome are present in species of different groups. Most karyological analyses in Aruchis have used species in section Aruchis, in which nine species have been analyzed (Stalker and Dalmacio, 1981; Singh and Moss, 1982; Stalker, 1985a). Genomes of most species are symmetrical with median chromosomes. Although somatic chromosomes are small, ranging from 1.4 to 3.9 mm in length, species can be identified based on karyological differences. By using arm ratios as variables, species of section Aruchis can be divided into clusters in which species with an A genome group together and species with chromosomes typical of the species A. butizocoi (B genome) separate into a second cluster (Singh and Moss, 1982). Further, A. spinucluvu has a highly asymmetrical karyotype with subtelocentric chromosomes not found in other species of the group (Stalker, 1985a). Smartt (1964) reported that the distinctively small chromosome pair found in most section Aruchis species was not present in a section Erectoides species, A. puruguuriensis Chod. et Hassl., which implies that the karyotype of section Erectoides species may be differentiated from chromosomes of most section Aruchis species.
VI.
INTERSPECIFIC HYBRIDIZATION IN Arachis
Interspecific hybrids in the genus were first attempted by Hull and Carver (1938) when they tried to cross A. hypogueu and A . glubrutu, a species now known to be distantly related to the cultivated peanut. The first successful hybrid reported in the genus was between A. hypogueu and the diploid species A. villosu var. correntinu in 1951 (Krapovickas and Rigoni, 1951). The cultivated species has since been hybridized with at least 12 and possibly as many as 18 species of the genus (Kumar et ul., 1957; Smartt and Gregory, 1967; Gregory and Gregory, 1979; Singh, 1985; Singh and Moss, 1984b; Pompeu, 1977; Stalker, unpublished data). To date, the only hybrids between the cultivated peanut and wild species have been with members of section Aruchis. Since many accessions have been introduced recently from South America, this conclusion must be verified using a wide range of variability in the genus, but results thus far have generally coincided with the crossing relationships established by Gregory and Gregory (1979) for sectional groups. Although there are reports of crosses between A. hypogueu and members of other sections, these crosses have not been repeated and thus will not be discussed further in this chapter. Raman and Kesavan (1962) reported the first hybrids among wild species in the genus between A. durunensis and A . villosu var. correntinu. Hybrids
I
SPECIATION, CYTOGENETICS, AND UTILIZATION
11
were fertile and had regular chromosome pairing during meiosis. Since the first hybrid was reported, hundreds of interspecific crosses have been produced to determine the biosystematic relationships among species or to introgress germplasm to cultivated peanut. The most extensive single hybridization program conducted thus far was by Gregory and Gregory (1979), who reported cross-compatibility relationships among 91 accessions of Aruchis species. They showed that intrasectional hybrids are much easier to produce than intersectional ones, but low frequencies of success are still observed for many hybrid combinations within groups. Gregory and Gregory (1979) determined relationships among taxa based on both crossability and pollen stainability data. They found that pollen stainability of intrasectional hybrids of section Arachis averaged 30.2% when crosses were made among species at the same ploidy level. Intrasectional hybrids among species within other groups ranged from a low of 0.2% in section Extrunervosue to a high of 86.8% in section Cuulorhizae. All intersectional hybrids were completely female-sterile and averaged only 1.9% pollen stainability (Gregory and Gregory, 1979). Since A . hypogaeu belongs to section Aruchis, researchers have concentrated efforts within this section. Most interspecific hybrids between species with an A genome have 10 bivalents during meiosis (Resslar and Gregory, 1979; Smartt et a/., 1978a,b; Stalker and Wynne, 1979; Singh and Moss, 1984a). Perennial species of the group generally hybridize more easily as male rather than as female parents. Although meiosis is regular, pollen stainability ranges between 20 and 85% and seed production is limited for several hybrid combinations. In contrast to hybrids between A genome species, when crosses are made between A. batizocoi (B genome) and other members of section Aruchis, all hybrids are sterile and have irregular meioses with a range of 4.6-8.6 bivalents per PMC (Gibbons and Turley, 1967; Smartt et a/., 1978a,b; Stalker and Wynne, 1979; Singh and Moss, 1984a). When the species A . spinucluva (D genome) is hybridized with either A or B genome species, all hybrids are sterile and meiotically irregular (Stalker, 1985a). Many other recently collected taxa must also be analyzed cytologically and, based of fertility data of F, hybrids, additional unique genomes may be found in the group. Because of high levels of sterility in intersectional hybrids between diploid species, crosses have been attempted after raising the ploidy level of species or their hybrids. All attempted crosses between amphidiploids of section Arachis species and amphidiploids or natural tetraploids of species in other sections (Erectoides or Rhizomatosae) have failed. Hybridization at the tetraploid level is more difficult than between diploids and tetraploids for at least some groups of the genus. For example, the two diploid (2n = 20) section Arachis species A . durunensis and A. stenosperma have been hybridized with the 40-chromosome amphidiploids (A. rigonii x A . sp. coll. GKP
12
H.T.STALKER AND J. P. MOSS
9841, PI 262278) of the section Erectoides (Stalker, 1981). A high frequency of bivalents was observed and Stalker concluded that chromosome homologies exist among members of sections Arachis and Erectoides. Complex hybrids between sections Erectoides and Rhizomatosae have also been cytologically analyzed and chromosome homologies reported for at least one hybrid combination (Stalker, 1985b). Also, plants of one 40-chromosome intersectional Erectoides x Rhizomatosae hybrid combination were male-fertile and produced selfed seeds. Several triploid hybrids between section Arachis (2x) and (Erectoides x Rhizomatosae)(4x) have also been made but all hybrids failed to flower even though they had been propagated for several years (Stalker, 1985b). Based on the cumulative cross-compatibility data of interspecific hybrids by many investigators, a series of genomes for Arachis species were proposed by Smartt and Stalker (1982) and Stalker (1985b) as follows: A: B: D: Am: C: E: Ex: T: R:
section Arachis, perennials and most annuals section Arachis (A. batizocor) section Arachis (A. spinaclava) section Ambinervosae section Caulorhizae section Erectoides section Extranervosae section Triseminalae section Rhizomatosae, series Prorhizomatosae
Arachis hypogaea and A . monticola have an AB genome, while the genomes of tetraploid species in section Rhizomatosae may be similar to the A genome of section Arachis and the E genome of section Erectoides. Only the A, By and D genome of section Arachis have been studied intensively, and other genomic designations in the genus remain to be verified cytologically. However, even in section Arachis there are unanswered questions, such as the real differentiation between designated A and B genomes. Based on cytological analyses, only two to four chromosome pairs are differentiated between A . batizocoi and A genome species (Stalker and Wynne, 1979; Singh and Moss, 1984a). Triploid hybrids between A . hypogaea and diploid species have a few trivalents, but hexaploids obtained after colchicine treatment average six or more univalents and may have as many as 20 unpaired chromosomes (Company et al., 1982; Singh, 1985). Pairing mechanisms, or lack thereof, are apparently under genetic control. Further, after backcrossing hexaploids with A . hypogaea, pentaploids are produced with the expected 20 bivalent plus 10 univalents, but after one generation of self-pollination, 25 bivalents have been observed in some progenies (Stalker, unpublished data). This indicates that considerable homology exists among the A and B genomes.
SPECIATION, CYTOGENETICS, AND UTILIZATION
13
Genomic designations outside of section Aruchis are based mostly on cross-compatibilities. Since incompatibilities may result from single genes, cytoplasmic effects, or other factors, there may be considerable homology among the genomes which have been designated as unique. Several problems also remain unanswered for groups of species. For example, why will diploid Prorhizomutosue not hybridize with tetraploid members of the same section when taxa from other sections will hybridize with the tetraploid rhizomatous species? A D genome has been designated for a species which is morphologically identified with members of section Aruchis, but the taxa may be genomically more similar to species in other sections. Regardless, sectional names are useful for communication concerning groups of species and, from present knowledge, potentials for utilizing species in the genus can be determined. Germplasm pools can be also designated for establishing potentials for introgression to A . hypogueu. The primary gene pool comprises A. hypogueu accessions and genetic stocks plus the closely related tetraploid species A . monticolu. Large collections of the cultivated species exist in the United States (cu. 4000 accessions) and at the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), which has more than 8000 lines (Wynne and Coffelt, 1982). Although several accessions of A . monticolu have been cataloged in germplasm lists, they all represent collections from at most two sites in South America. Aruchis hypogueu will hybridize with A . monticolu and produce fertile hybrids which have normal meiosis (Krapovickas and Rigoni, 1957; Raman, 1958). Analyses of somatic chromosomes have confirmed the close relationship between the species, indicating that they belong to the same biological species. The secondary gene pool is represented by diploid members of section Aruchis which have an A or B genome. Hybrids between the diploid and tetraploid species of the group are sterile, but fertility can be restored by manipulating ploidy levels and good evidence exists for homology between the chromosomes of wild and cultivated species (Singh, 1985; Stalker, 1985b). Although A . butizocoi is the most likely representative donor of the B genome in A . hypogueu, identification of the A genome donor species has not been made. However, enough similarities exist between A . hypogueu and most of the A genome species of section Aruchis that gene transfer can occur from wild taxa to the cultivated peanut. The tertiary gene pool includes all taxa outside section Aruchis plus species of section Arachis which do not have an A or B genome (for example, A. spinucluvu). Hybrids between A . hypogaeu and these species have not been produced and specialized techniques will be required to produce hybrids. However, F, generation plants are expected to be completely sterile and methods to introgress small chromosome segments will be necessary for utilization of these germplasm resources.
H. T. STALKER AND J. P. MOSS
14
VII.
GERMPLASM EVALUATION
Identification of desirable traits, especially for disease and insect resistances, in Arachis species must precede utilization of germplasm resources. Disease and insect resistances have had the highest rates for successful introgression from wild species to many crop plants (Watson, 1970; Knott and Dvorak, 1976). Likewise, the most commonly investigated agronomic traits in wild species are pest resistances. The peanut is plagued by a large number of pests, many of which are now worldwide in distribution. Because of the agronomic importance and impact of diseases and pests on yield and quality, introgression of disease resistance from wild species to cultivars has been a high priority in many breeding programs. A.
DISEASE RESISTANCES
The three most important diseases of A . hypogaea worldwide are Cercospora arachidicola Hori (early leafspot), Cercosporidium personatum (Berk. et Curt.) Deighton (late leafspot) and Puccinia arachidis Speg. (peanut rust). Subrahmanyam et al. (1984) estimated that in India, which is one of the largest producers of peanuts, yield losses due to rust and leafspots are approximately 70% annually. Gibbons (1980) estimated that production in locations where fungicides are not used, largely because of high chemical costs, have yield decreases of approximately 50%, and even when chemicals are applied yield may be decreased by 10% (Jackson and Bell, 1969). In addition to actual production losses directly due to the diseases are costs of chemicals, application expenses, and plant damage incurred during applications. Although only one of the leafspots may be common at a particular location during the year, the disease populations may change over years as cultivars are replaced (Smith and Littrell, 1980). Many Arachis species have been evaluated for resistance to the C. arachidicola pathogen (Table IV). The three species A . glahratu, A. hagenheckii, and A . repens have high levels of resistance to this pathogen (Gibbons and Bailey, 1967). Abdou et al. (1974) screened 94 species accessionsin the greenhouse and found members of sections Arachis (A. chacoense GKP 10602), Caulorhizae(A. repens GKP 10538), Extranervosae (A. villosulicarpa, three accessions), and Rhizomatosae (A. sp. GKP 10596) to be immune to C. arachidicola. Melouk and Banks (1978) confirmed the immune reaction of A . chacoense, but Foster et al. (1981) and Company et al. (1982) observed small lesions on leaves of field-grown plants. Kolawole (1976) reported high levels of resistance in a second section Arachis species which Sharief et al. (1978) concluded was the A . stenosperma Greg. et Greg. nom. nud., collection HLK410. Because evaluations of both cultivated and wild species at different locations
SPECIATION, CYTOGENETICS, AND UTILIZATION
15
had been done with different techniques and disease pressures, Foster et a/. (1981) compared 9 section Aruchis species with 14 reportedly resistant cultivated genotypes. They confirmed a high level of resistance in A . stenospermu (HLK 410) and found that A . chucoense had significantly fewer lesions per leaf than any other speciestested. Sharief et a/. (1978)reportedthat resistance in Aruchis species is multigenic, while Company et a/. (1982) found triploid interspecific hybrids between A . hypogueu and the Aruchis speciesA . curdenusii and A . chucoense to be resistant to early leafspot in the field. This indicated at least partial dominance for resistance. Cercosporu uruchidicoluresistance appears to have been introgressed from the wild diploid speciesA . curdenusiito A . hypogueu (Stalker, 1984). Late leafspot (C. personutum) is the most severe peanut disease in many production areas. Abdou et a/. (1974) screened the same 94 accessions mentioned previously for this pathogen in the greenhouse. They reported high levels of resistancein several taxa of sectionsAruchis, Cuulorhizue,Extranervosue, and Rhizomutosue. Aruchis curdenusii was the only speciesin the accessions which is both crosscompatible with A. hypogueu and immune; Subrahmanyam eta/. (1980) confirmed the reactions in separate experiments. Kolawole (1976) reported high levels of resistance in a second species, now believed to be A . stenospermu. Resistance has recently been selected in 40-chromosome interspecific hybrid derivativesof A . hypogueu x sectionAruchis species (Moss, 1985b).
Because populations of leafspotscan change, breeding for only one pathogen and not the other may be futile. Fortunately, high levels of resistance have been found in several genotypes to both early and late leafspots (Abdou et a/., 1974; Kolawole, 1976). In addition, A . curdenasii (which is reported as resistant to C. personutum but susceptible to C. uruchidicolu) has at least moderate levels of resistance to the early leafspot pathogen for lesion number and lesion size, but not for genes conditioning defoliation (Foster eta/., 1981). In addition to fungal pathogens, peanuts are invaded by many viruses. Several of these cause severe damage and yield loss such as bud necrosis and peanut stunt, while others expressonly mild symptomswhich may have little effect on yield. A comprehensivereview of peanut diseases was made by Porter et a/. (1982), who discussed descriptions of causal organisms, symptoms, disease cycles, and controls. This chapter will thus be restricted to reports directly related to wild Aruchis species. Groundnut rosette virus is restricted to Africa south of the Sahara and may be the most destructive virus disease of peanut (Porter et a/., 1982). While resistancehas been found in the AruchisspeciesA . glubrutu and A . repens (Gibbons, 1969), the species A . glubrutu and A . prostrutu Benth. (identified as A . hugenheckii and A . repens by Gibbons, 1969) were reported as being symptomless carriers of the virus by Klesser (1967). Because resistance was found only in species which will not hybridize with A . hypogueu, and at about the same time adequate levels of resistance were also found in the cultivated accessions,
Table IV Pest Resistnnce in Wild Arachh Species" _
_
Collection
e a\
_
~
COllectOP
Section Ambinervosae 12943 GK 129445 GK Section Arachb 408 HLK 410 HLK 7264 K 7830 K 7897 K 7988 K 9484 K 9530 GKP 953 1 GKP 9548 GKP 9901 10017 10038 10602 22585 30006 3001 1 3003 1 30035 30063 30085
A . correntina Manfredi #5 Manfredi #8 Manfredi #36 A. villosa
GKP
GKP GKP GKP Bu GK GK GK GK GKBSPSc GKBSPScZ
-
Species
PI 338452 338454
A . sp. A . sp.
338279 338280 219824 261871 262873
A . stenosperma A . stenosperma A . monticola A . correntina A . correntina A . duranensb A . batizocoi A . correntina A . correntina A . correntina A . sp. A . cardenasii A . spegauinii A . chacoense A , villosa A . sp. A . sp. A . helodes A . sp. A . sp. A . sp. A . correntina A . correntina-villosa A . correntina-villosa A . correntina-villosa A . villosa
-
298639 262808 262809 262839 262270 262141 262133 276235 298636 468 150 468 154
-
468168 468199 468331
-
-
C.a
C.p
R
PSV
Th
PHL
CEW
SM
LCSB
HR HR
I HR
HR HR
S
S S
HR HR
HR HR
S S S S
MR MR
-
-
HR
HR
HR HR
-
R HR
-
-
R R I R R
-
S
-
s
-
-
MR
MR
MR S
S
s
HR MR HR R
HR
s
I I I I
-
-
I I I
-
HR R -
-
-
R
s
-
S
R S
-
I
I I HR HR
-
s
-
I I
R I R R S
-
I
S
I HR
-
I HR HR
-
-
MR
HR HR HR
-
-
-
HR R HR R
HR HR HR HR
HR HR HR HR
S
S
MR
S S S
S
MR MR
MR MR R R
Section Caulorhizae 10538 12787
Section Erectoides
e
4
565-66 9646 9764 9769 9788 9795 9812 9820 9825 9835 9837 9841 9990 9993 10002 10034 10541 10543 10573 10574 10576 10580 10582 10585 10588 11462 11488 14444
GKP GK
GKP GKP GKP GKP GKP GKP GKP GKP GKP
GKP GKP GKP
GKP GKP GKP GKP GK GK GKP GK GK GK GK GK KFC
276199 338447 338297 262842 262859 262862 262790 262863 26279 1 263 105 262278 261877 261878
-
262142 276208 276209 276225 276226 276228 276229 276230 27623 1 276232
KC KHe
338320
A . repens A . pintoi
s -
s -
-
I
I
s
-
HR
-
HR
-
-
S
-
A. paraguariensis A . benthamii
A . benthamii A . sp. A. sp. A. sp. A. sp. A. A. A. A.
sp. sp. sp. sp. A . sp.
A . sp. A. sp. A. rigonii A. oteroi A . sp. A. sp.
A. sp. A. sp. A. sp. A. sp. A. paraguariensis A. sp. A . paraguariensk A. paraguariensis A. sp.
R R MR MR H R R MR MR
-
I I
-
-
s s
s s
-
R R
MR R
-
s
-
-
-
HR HR
S S
R R
HR HR I
HR HR HR
S S
-
-
-
-
R R
I
MR MR
-
MR s
HR R
-
I H R - HR - HR S HR
-
-
-
-
I R S I
-
- MR s - - s R s
-
-
-
-
-
-
I
-
-
R
S
-
R
S
-
-
I
HR
HR
I
I
HR
MR -
-
-
-
HR HR
-
-
-
R
-
-
-
-
I
HR
I
S
-
-
MR
-
S
S
MR
R
R
-
(continued)
Table IV (Continued) Collection
L
00
Section Extranervosae 9906 10127 10406 A. villosulicarpa 1960 #3 1968 #lo0 9553 9562 9564 9566A 9566B 9567 9568 9569 9570 957 1 9572 9574 9575 9576 9578 9580 9587 9591 9592 9610 %1OB %I8 %29 %34 %42
CoUector”
PI
GKP GKP GKP
262272 276203 276198 336985
GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP
-
262801 26281 1 262812 262813 262814 2628 15 2628 16 262817 2628 18 262819 262820 262821 262822 262824 262825 262826 262827 262828 262832 262832
-
262834 262836 262839
A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A.
Species
C.a
C.p
R
PSV
Th
lutescens macedoi margenata villosulicarpa
R R
HR
-
-
-
S
S
-
-
-
sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. A. sp. A. sp. A. sp.
HR -
-
-
HR -
HR R
MR HR
S S S
S S S
HR HR MR HR MR HR
HR HR MR R MR MR
S
S
HR R
R HR
S S
S S
R
MR
MR HR HR
MR MR MR
S
S MR R
-
-
MR MR
-
I
MR -
I
-
HR
-
I I I
-
-
I
-
-
I I
-
-
I
-
I
-
I
I
I -
-
I I I I I
I I
-
PHL
R HR I
s I
R I I I I I I
s
I
HR
-
-
-
R
-
HR HR -
HR -
-
-
HR HR
-
-
CEW
SM
LCSB
L
rD
9644 9645 9649 9664 9667 9797 9806 9813 9815 9822 9827 9830 9834 9882 9893 992 1 9922 9925 9935
9%6 10105 10120 10550 10559 10566 105% A. glabrata-B,
GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GKP GK GKP GK GK
-
262840 262841 262844 262847 262848 262807 262792 262793 262794 262795 2627% 262797 262798 262286 262287 2622% 262297 262299 262301 262306 276200 276202
-
276217 276223 276233
-
sp. sp. sp. sp. sp. sp. sp. sp. sp. sp. glabrata glabrata sp. sp. sp. sp. sp. A. sp. A . sp. A. sp. A . sp. A. sp. A. sp. A. sp. A. sp. A. sp. A. glabrata A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A. A.
Section Trbeminalae 12922
GKP
338449
A. pusilla
HR HR
MR MR R
S S
MR
-
-
S
MR HR S S S
s
I MR I
-
-
HR HR
HR HR
I I I I
R
HR
I
I
1
-
-
I I
I I R
I I
-
-
MR
s
-
HR R HR HR HR MR MR
S S HR R R MR MR R MR
-
S
s
-
MR
MR
-
-
S S
s
-
-
HR MR
-
-
I I I I
S
R S
-
1
I I I I I I
-
-
I I I
-
I
-
-
-
HR -
I I
I I
HR HR
I HR HR HR
-
-
HR I HR HR
HR HR HR HR
-
-
HR HR HR
HR I
R HR
HR
HR R HR
-
-
-
I
I
S HR MR
-
MR HR
-
I I I
-
-
-
-
HR HR
-
I I -
I
I
-
R R R R R MR R MR
-
HR
MR
MR
%sect or disease: C.a, Cercospora arachidicola; C.p, Cercosporidiumpersonatum; R, rust (Pucciniaarachidis);PSV, peanut stunt virus; Th, thrips (Frankliniellafusca);PHL,potato leafhopper (Empoascafabae); CEW, corn earworm (Heliothb zea); SM. spider mite (Tetranychus urticae); LCSB, lesser cornstalk borer (Elasmopalpuslignosellus).Rating: S , susceptible,MR, moderately resistant; R, resistant, HR, highly resistant; I, immunity (based on authors’ interpretation of literature cited in text). bCollectors: B, Banks; Bu, Burkart; C, Cristobal; F, Fugarazzo; G, Gregory; H, Hammons; He, Hemsy; K, Krapovickas; L, Langford; P, Pietrarelli; S, Simpson; Sc, Schinini; Z. Zurita.
20
H.T. STALKER AND J. P. MOSS
little screening or attempted utilization of wild species has occurred. Highyielding rosette-resistant cultivars have been released for grower use in Africa (Gillier, 1978). Tomato spotted wilt virus, the causal organism for bud necrosis disease, is widespread in many peanut production areas and may cause up to 90% yield loss (Saint-Smith et al., 1972). Ghanekar (1980) screened approximately 7000 A . hypogaea accessions and did not find field resistance to the disease. Subrahmanyam et al. (1985) inoculated 42 Arachis species in the greenhouse and field. The species A . pusilla GK 12922, A . correntina GKP 9530, and A . cardenasii GKP 10017 became infected in the greenhouse but expressed no symptoms in the field. Arachis chacoense GKP 10602 showed infection neither after mechanical injection of the virus nor after infection by thrips (although virus was detected after grafting), and this species may represent the best source of resistance to bud necrosis. Since A . chacoense has been hybridized with A. hypogaea, it should serve as a usable source for resistance in a peanut breeding program. Peanut stunt virus was reported in Virginia during 1964 (Miller and Troutman, 1966), and epidemics occurred in the following years. Since then, stunt virus has been found in other production regions in the United States, Japan, and Africa (Porter et al., 1982). Hebert and Stalker (1981) screened approximately 4OOO cultivated accessions and all were susceptible to the virus. However, after evaluating 90 Arachis species accesions, they found high levels of resistance in species of sections Arachis, Caulorhizae, Erectoides, and Rhizomatosae (Table IV). Again, the most accessible species were the highly resistant sources A . duranensis K 7988, A . villosa B 22585, and an A . correntina-villosa genotype (Manfredi #8), and several others with high tolerance levels [A. correntina K 7897 and GKP 9548 and two A . correntina-villosagenotypes (Manfredi #5 and #36)] found in section Arachis. Hebert and Stalker (1981) also reported that resistance is not conditioned by a single dominant gene, so introgression may be difficult from wild to cultivated species. Because peanut stunt virus has effectively been controlled through cultural practices and incidence has been insignificant during the past 10 years, no concentrated efforts have been made to transfer genes conferring stunt virus resistance from the wild to cultivated species. Peanut mottle virus is found worldwide and can infect almost every plant in a peanut field. Demski and Sowell (1981) concluded that economic losses due to the virus are second only to leafspots in the southeastern United States. Kuhn et al. (1968) screened more than 450 cultivated accessions but did not find usable levels of resistance. Later, Kuhn et al. (1978) identified two tolerant A . hypogaea accessions. Demski and Sowell (1981) evaluated seven species accessions of section Rhizornatosae and six of these were immune to peanut mottle virus. Subrahmanyam et al. (1985) screened an additional 50 Arachis accessions and found no infection after mechanical or airbrush inoculations in the species A . pusilla GK 12922, A . chacoense GKP 10602, A .
SPECIATION, CYTOGENETICS, AND UTILIZATION
21
cardenasii GKP 10017, and A . correntina GKP 9530. In addition, A . pusilla and A . chacoense were not infected after grafting infected scions onto their stems; therefore, these two species accessions may have true immunity. Fitzner et al. (1985) evaluated 14 species of section Arachis for the soilborne disease Cylindrocladium black rot, caused by Cylindrocladium crotalariae (Loos) Bell and Sobers. Resistance was reported only for the species A . monticola GKBSPSc 30062. They indicated that a valuable source of resistance for developing cultivars may have been found because A . monticola produces fertile hybrids with A . hypogaea. Another soil-borne problem for peanut production in many areas is nematode infestations. In some regions of the world, peanuts cannot be grown without nematode controls (Porter et al., 1982). Nematodes may also be associated with high levels of aflatoxin and with other soil-borne diseases (see Porter et al., 1982). Meloidogyne hapla Chitwood (northern root knot nematodes) are the most important species which attack peanut. Of 33 Aruchis accessions evaluated by Banks (1969), only accession PI 262286 (of section Rhizomatosae) had moderate levels of resistance. Castillo et al. (1973) evaluated 12 wild species accessions and, in addition to confirming resistance in PI 262286, reported three additional PIS, 262841, 262814, and 262844, as being more resistant than the cultivated controls. B.
INSECTRESISTANCES
Insects can cause severe yield losses in peanut by feeding on all plant parts. In addition, insect specieshave been shown to bevectors for viruses (see Smith and Barfield, 1982). Surveys have not been taken to establish the economically important pests in many production areas, but Smith and Barfield (1982) listed 360 insect species which attack peanuts. The lesser cornstalk borer [Elasmopalpus lignosellus (Zener)] is the most severe subterranean insect pest in the United States, and the southern corn rootworm (Diabrotica undecimpunctata howardii Barber) also has a wide distribution. Aboveground foliage insects-including tobacco thrips (Frankliniellafusca Hinds), potato leafhoppers, (Empoasca fabae Harris), corn earworms (Heliothis zea Bodie), and fall armyworms (Spedopterafragiperda J. E. Smith) are among the most severe insect pests of peanuts in the United States. Many insects move into peanut fields in successive waves. Thrips arrive early in the growing season, followed by corn earworm invasion at peak bloom and then by potato leafhopper migrations (Campbelland Wynne, 1980). In the semiarid tropics, the predominant species of insect pests include the groundnut aphid (Aphis craccivora Koch), thrips [Scirtothripsdorsalis Hood, Caliothrips indicus Bagnall, Frankliniella schultzei (Trybom), F. fusca, and Ennoethripsflavens Moulton], jassids (Empoascasp.), armyworms (Spodoptera sp.), and termites (Microtermes sp. and Odontotermes sp.) (Amin, 1985).
22
H. T. STALKER AND J . P. MOSS
A large number of Aruchis accessions have been screened for insect resistances including resistance to thrips, leafhoppers, and corn earworms (Stalker and Campbell, 1983);armyworms (Lynch et ul., 1981);lesser cornstalk borers (Stalker et ul., 1984); and spider mites (Leuck and Hammons, 1968;Johnson et ul., 1977). High levels of resistance have been found for the first four insect pests listed above, whereas lower (but still high) levels of resistance have been reported for the lesser cornstalk borer and spider mites (Table IV). Most importantly, resistances are found in species which will hybridize with A. hypogueu. Stalker and Campbell (1983)indicated that the mechanism of resistance to H. zeu in Aruchh species was antibiosis. In attempts to utilize the resistances for insect pests, 40-chromosome interspecific hybrid derivatives with A. hypogueu were evaluated and high levels of resistance for several insect pests were found (Stalker and Campbell, 1983;ICRISAT, 1985).
C. OTHERTRAITS Aruchis species are potentially valuable germplasm resources for traits other than disease and insect resistances. For example, peanut seed protein is usually low in the sulfercontaining amino-acids and tryptophan. Aruchis villosulicurpu is high in tryptophan (range of 1.44-1.66Vo)as compared to the highest A. hypogueu line (1.41%) tested by Amaya et al. (1977). Many of the Aruchis species are extremely drought resistant and research in the area of physiological traits needs to be conducted. In addition to seeds and vegetative plants collected on exploration trips in South America since 1976,nodules have also been obtained from many accessions. Brudyrhizobium have been isolated and maintained from both cultivated and wild Aruchis species (Elkan et ul., 1981). Many strains have been evaluated for plant-Rrudyrhizobium interactions; and one strain, NC 92, increased peanut yields under field conditions when applied to the cultivar Robut 33-1 in India (ICRISAT, 1983).
VIII.
UTILIZATION OF WILD Arachis SPECIES
Wild species of Aruchis are a valuable source of desirable characters for cultivar improvement. Especially when the wild species are the only source of resistance to diseases (for example, to peanut clump virus), attempts need to be made to combine the genomes of cultivated and wild species to develop genotypes with stable resistance. Results from the few studies so far conducted indicate that genes for resistance in wild peanut species may be different from those in cultivated peanut (for example, peanut rust) (Singh et ul., 1984). However, difficulties of gene transfer from wild species have precluded widespread use of potential germplasm and have also limited our knowledge of the nature and genetics of resistance in the wild species.
SPECIATION, CYTOGENETICS, AND UTILIZATION
A.
23
INCOMPATIBILITIES RESTRICTING GENE TRANSFER
The major restriction for study and use of wild species in Arachis is crossincompatibilities between most species and the cultivated peanut. With the exception of A. monticofa, all wild species in section Arachis that can be hybridized with the cultivated peanut are diploids. Thus, hybrids are sterile. Even among these cross-compatible species the success rate for interspecific hybrid production may be low. Reasons for this include incompatibilities among species, especially when the wild species is used as a female parent; hybrid sterility due to the polyploidy; genomic differences among species; irregular meiosis in colchicine-treated hybrids; and difficulties encountered during backcross generations when sterile aneuploid or pentaploid plants are obtained. Even when hybrids can be obtained, the problem of eliminating undesirable wild species characters still exists. When genomes are common to both cultivated and wild species, this may not be a problem. If there are no chromosome homologies, it may be necessary to induce translocations for gene introgression. Even where genomes are homologous or homoeologous and pairing occurs, linkage may restrict recombination between desired and undesirable genes and prevent the production of A. hypogaeu-like lines with the desired agronomic characters (Stalker et af., 1979). Before a program can be designed to circumvent interspecific hybridization barriers between A. hypogaea and other species, reproductive ontogeny and isolation barriers must be understood. Approximately 12 hr elapse between pollination and fertilization (Smith, 1956). About a week after fertilization, an intercalary meristem of the peg located in the ovary proximal to the ovules begins a rapid geotropic elongation (Jacobs, 1947). Gibberellic acid has a significant stimulatory effect on peg elongation (Amir, 1969), while auxin inhibits peg elongation and is associated with fruit enlargement (Jacobs, 195 1). Several days after elongation initiates, the peg penetrates the soil, ceases to grow, and expands into a pod (Smith, 1950, 1956; Yasuda, 1943). Ziv (1981) reported that light is necessary for peg elongation. When pegs fail to reach the soil, they remain viable for several days and then wither. Moisture in the absence of light appears to be the most important factor governing pod development (Yasuda, 1943). Although the peg follows a sigmoidal growth pattern, the embryo and endosperm initially divide and then become quiescent; the embryo contains between 5 and 27 cells at the time the peg penetrates the soil (Schenk, 1961). After the peg is underground, the embryo then maintains a rapid growth phase. Both the root apex and cotyledons are initiated in the globular embryo stage, and by the heart-shaped embryo stage the cotyledons appear as a projection (Pallai and Raju, 1975). Schenk (1961) reported physiological changes associated with peg development, and Brennan (1969), Gregory et af. (1973), and Periasamy and Sampoornam (1984) reviewed in more detail the reproductive development of the cultivated peanut. Halward and Stalker (1985, 1987a) reported differences in development of reproductive tissues of wild and cultivated peanuts.
24
H. T. STALKER AND J. P. MOSS
Although many hybridization failures of interspecific crosses have been attributed to embryo abortion, only a few investigations have been reported detailing mechanisms of incompatibility. Johanson and Smith (1956) attributed the failure of A . hypogueu x A . diogoi (not true diogoi vide Gregory and Gregory, 1979) to slow growth and degeneration of the embryo accompanied by hypertrophy of integuments. In A . hypogueu x A . glubrutu (section Rhizomutosae) crosses, Murty et ul. (1980) observed up to a 48-hr delay in fertilization and early embryo abortion. In A . monticolu X A . sp. (section Rhizomutosae) crosses, Sastri and Moss (1982) observed large callus plugs along pollen tubes; however, a few pollen tubes were observed in the ovary. They further reported that gibberellic acid and kinetin treatments stimulated peg production in incompatible crosses. In the more closely related species of section Arachis incompatibility among species can be caused by the failure of pollen to germinate on the stigma, restriction in fertilization, and/or embryo abortion (Halward and Stalker, 1984). Further, interspecific hybrids can abort a early as 6 days after pollination (for example, in diploid x hexaploid hybrids) or remain viable but undeveloped until the time of normal maturity (for example, hexaploid x diploid hybrids) (Halward and Stalker, 1985, 1987b). Future recovery of interspecific hybrids will thus depend on developing techniques to initiate peg development, promote embryo growth on the plant, and to recover viable, but small, embryos in vitro. Promoting embryo development can either be done through ovule culture or, in the case of peanuts, by applying hormones to developing tissues in vivo (Mallikarjuna and Sastri, 1985a). Application of embryo culture techniques will then be necessary to recover small reproductive tissues after embryos have reached the heart-shaped embryo stage of development.
B. DIRECT HYBRIDIZATION WITHIN SECTION Aruchis I . Hybrids between Tetruploid Species
Aruchis monticolu is a tetraploid in series Amphiploides of section Aruchis and is the only wild Aruchis taxon which can be readily crossed with A . hypogueu to produce fertile progeny. The species has been given specific status, but it is a member of the same biological species as A . hypogueu and logically could be considered as a subspecies of the cultivated peanut. Aruchis rnonticolu was used by Hammons (1970) in the pedigree of Spancross and probably by Simpson and Smith (1974) to develop Tamnut 74. For all practical purposes, A. rnonticolu can be considered a wild form of peanut which does not need species manipulations for its utilization in breeding programs.
SPECIATION, CYTOGENETICS, AND UTILIZATION
25
Table V Number of Pods Produced per 100 Pollinations in Successive Backcrosses of Hexaploids to A. hypogaetf Backcross Wild species used in production of hexaploid cardenasii chacoense stenosperma correntina villosa batizocoi Mean
A. A. A. A. A. A.
~
~~~
BC,
BC,
BC,
BC,
BC,
9 6
18 10 16
12 11 8
26
-
I
16
14 25
~~
~
7 10
-
-
9
16
11
5
-
-
-
-
10
16
4
Number of fertile derivatives selected 9 9
2
-
~
“From ICRISAT (1981-1982).
2. Hybrids between A. hypogae and Diploid Species
a. Triploids All attempts at crossing A. hypogaea with diploid species with an A or B genome have produced hybrids, although the crossing success varies depending on the species used and the direction of the cross (Table V). The interspecific hybrids are triploids, usually vigorous, flower profusely, and are mostly sterile (Smartt and Gregory, 1967; Raman, 1976; Gregory and Gregory, 1979; Seetharam et al., 1973). There were an average of 8.8 univalents, 9.1 bivalents, and 1.0 trivalents per PMC in the triploids between A. hypogaea and eight diploid species (Singh and Moss, 1984b). Segregation was irregular and the percentage of stainable pollen grains varied in size. Sterility has been overcome by colchicine treatment to produce hexaploids for many hybrid combinations (Smartt and Gregory, 1967; Spielman et al., 1979; Spielman and Moss, 1976; Company et al., 1982). Although tripolids are usually sterile, seeds were produced on several hybrids of different cross-combinations of interspecific A. hypogaea hybrids (Simpson and Davis, 1983; Singh and Moss, 1984b). Eighty-two percent of the progeny derived from selfing triploids were hexaploid, indicating the formation of competent unreduced gametes (Singh and Moss, 1984b). Progenies other than hexaploids had chromosome numbers ranging from 2n = 20 to 59, indicating that gametes with fewer than 30 chromosomes can be functional. The hexaploids are of special interest as they have been obtained without the need for colchicine treatment, and, unlike colchiploids they have arisen from postmeiotic cells and thus pairing between wild and cultivated chromosomes has occurred. This is also true for the tetraploids; although produced at lower frequencies (8%) than the hexaploids (82Oro), they have the advantage of being at the same ploidy level as
26
H. T. STALKER AND J. P. MOSS
the cultivated peanut. Whereas colchiploids have identical homologies and chromosomes will not segregate (except as a result of homologous pairing), the progenies arising from selfed triploids will be unique because chromosome segregation has occurred. b. Hexaploids. Hexaploids, whether produced by colchicine treatment or from selfing of partially fertile triploids, have many undesirable characters associated with wild species, and none have been seriously considered suitable as the basis for developing the hexaploid peanut as a crop. Hexaploids are, therefore, an intermediate stage in a hybridization and selection process. The chromosome number must be reduced to the teraploid level and then undesireable wild characters eliminated. Tetraploidy can theoretically be achieved in one step by crossing hexaploids with diploid wild species. Although this will achieve the desired ploidy level and provide an opportunity to incorporate additional characters to A. hypogaea if a nonparental wild species is used, it also reduces the proportion of cultivated chromosomes in tetraploid derivatives. Hybrids between hexaploids and diploids have been difficult to produce due to embryo abortion (Halward and Stalker, 1984), and no tetraploid populations have been developed from these crosses. Hexaploids vary in meiotic regularity and in fertility but can be crossed with A. hypogaea. Pods per 100 pollinations of these crosses varies from 7 to 25 (Table VI). From three different hexaploids backcrossed to BCI or to BC, generation, 6775 pollinations produced 894 pods (13010), but only 20 fertile plants with regular meiosis were selected from pregenies (ICRISAT, 1982). c. Alteration of Ploidy Levels. Other ploidy manipulations are available to bypass the sterility of triploids and difficulties of backcrossing hexaploids. These all involve producing tetraploid derivatives of the wild species which can then be crossed with A . hypogaea. Further, the crosses Table V1 Number of Pods Produced per 100 Pollinations in Successive Backcrosses of A. sp. Autotetraploids to A. hypogae& Autotetraploid
BC,
BC,
batizocoi villosa correntina stenosperma spegazzinii Mean
6 5
7
13
6
-
2 3 11
20 14
-
-
-
1
5
8
12
6 (Total)
A. A. A. A. A.
"From ICRISAT (1981-1982).
BC,
Fertile stable derivatives 3 1 1
4
SPECIATION, CYTOGENETICS, AND UTILIZATION
27
have the advantage of producing wild x cultivated hybrids with a range of genome formulas (AABB, AAAB, or ABBB) which encourage intergenomic AB pairing, which Smartt et al. (1978a,b) predicted would be a problem in transferring wild species characters into A . hypogaea. The known genomes in section Arachis are the A and B of A . hypogaea: the B genome in wild A . batizocoi and the A genome of all other wild species except A. spinuclava, which has a D genome (Stalker, 1985a). There are differences in chromosome morphology and degree of pairing in hybrids between the wild A genome species (Singh and Moss, 1984a). Thus, the possible ploidy manipulations are to produce autotetraploids of the A, B, and D genomes and AB, AD, and BD amphiploids. In addition, amphiploids can be produced by crossing two different A genome species and doubling the chromosome number of the hybrid (ICRISAT, 1981; Gardner and Stalker, 1983). A wild range of crosses have been made and tetraploid derivatives produced. i. Autotetraploids. Autotetraploids of A . villosa, A correntina, A . stenosperma, A . duranensis,A . spegazzinii, A . chacoense,A . cardenasii, and A . batizocoi have been produced (Singh, 1986a). Multivalents are frequent in autotetraploids; mean quadrivalent number in different autotetraploids ranges from 2.4 to 4.8 per PMC (Singh, 1986a), and pollen stainability ranges from 8 to l6%, except for autotetraploid A . batizocoi, which has 37% stainable pollen. Although vegetatively vigorous, autotetraploids are difficult to maintain due to reduced seed fertility; when crossed with A . hypogaea, from 2 to 11 pods per hundred pollinations (mean 5%) were produced, but with successive backcrosses to A . hypogaea the crossability increased as hybrids became more fertile (Table VI). The first cross of an autotetraploid to A . hypogaea results in plants with genomic constitution AAAB or ABBB, and the absence of homologs for one genome could lead to increased homoeologous (A-B) pairing, which would increase the range of recombinants produced. Forty-chromosome, meiotically regular, and fertile plants can be obtained from autotetraploids within three generations of crossing to A . hypogaea (ICRISAT, 1982). For example, of five autotetraploids backcrossed to A . hypogaea, 3368 pollinations produced 249 pods (7.4070), from which six fertile derivatives were obtained. Fertile, stable derivatives can be obtained after one cross. Thus, although autotetraploids provide opportunity for homoeologous pairing when backcrossed with A . hypogaea, the number of resulting desirable recombinants is low. ii. Amphidiploids. Presently there are 12 named diploid species in section Arachis, although as recent collections are studied this number will likely increase to 15-20 species in the group. Currently, there are only three genomes known, so only the genome combinations AABB, AADD, and BBDD can be produced. A sum of 10 AABB, 10 AADD, and 1 BBDD
H.T. STALKER AND J. P. MOSS
28
species combinations are thus possible among named species. However, amphiploids can also be produced from two A genome species, so the total number of amphiploids which can be produced among all described taxa is 132; this number includes reciprocal hybrids to account for possible differences in cytoplasmic effects. The amphidiploids represent an important gene pool for peanut improvement , and each one potentially incorporates genes from two species at the same ploidy level as cultivated peanuts. Further, each amphidiploid has at least one genome in common with A. hypogaea, which should promote gene transfer. Many diploid interspecific hybrids have been produced between section Arachis species. Pairing and fertility in AA hybrids are reasonably good, with 8.96-9.8 bivalents per PMC and 40-85'70 pollen stainability. However, in AB, AD, and BD hybrids, paring and fertility are reduced to a mean number of 4.7-8.6 bivalents per PMC and pollen stainability ranging from 3 to 7% for AB, AD, and BD hybrids (Stalker and Wynne, 1979; Singh and Moss, 1982, 1984a; Stalker, 1985a). Comparisons of AA and AB hybrids with the corresponding amphiploids show that pairing is more regular and pollen stainability higher in the AABB amphiploids, but the reverse is true for the AAAA amphiploids (Table VII). When A. hypogaea was hybridized with the four amphiploid genotypes, little difference in frequencies of trivalents (mean = 0.34) or quadrivalents (mean = 0.07) per PMC are observed (Gardner and Stalker, 1983). AABB (A. hypogaea x amphidiploid) hybrids form fewer bivalents (mean = 14) and have more univalents (mean = 6) than would be expected from their genomic formula (ICRISAT, 1982; Singh, 1986b). Ten univalents in the AAAB hybrids were also observed which could easily be assigned to the B genome, but this assumption is not justified, due to the high univalent frequency in AABB hybrids. This indicates that in A. hypogaea x amphidiploid hybrids, both AB and A (wild)-A (Cultivated) pairing is highly probable. Pollen stainability and plant fertility in amphiploids and their hybrids with A. hypogaea is high enough to conduct a backcrossing program. Three fertile and stable derivatives were obtained from a total of 321 pods obtained after backcrossing A. hypogaea with AABB wild species amphiploids, and seven were obtained from 527 pods for AAAA amphiploids (ICRISAT, 1982). These frequencies of about 1% show that both types of amphiploids are a practical means of introgressing genes from wild species to A. hypogaea.
c.
INTERSECTIONAL HYBRIDIZATION FOR GENEINTROGRESSION
Successful intersectional hybridization is rare in Arachis. Of 42 possible combinations, including reciprocals, attempted by Gregory and Gregory (1979), only eight hybrids were produced and none of these involved A.
P
Table VII Frequency of Chromosome Pairing and Pollen Stainability in Diploid Hybrids, Amphiploids, and A . h y p o g w I
I1
Mean
Range
AB hybrid
9.1
7.2-10.6
AABB amphiploid AABB A. hypogaea x amphiploid
3.7
1.5-6.3
6.2
4.5-7.9
AA hybrid AAAA
0.3 2.4 10.6
Mean
IV
111
Range
Mean
5.2
4.7-5.5
0.1
0-0.4
16.6
14.9-18.2
0.4
0.3-0.5
0.4
0.2-0.7
14.1
13.3-14.9
1.1
0.7-1.5
0.6
0.3-0.9
0-0.8 0.9-3.5
9.6 13.6
9.4-9.1 13.0-14.5
0 0.4
0.2-0.8
0.1 2.2
9.7-11.8
11.1
7.2-13.0
1.1
0-2.4
1.0
0
Amphiploids"
Pollen stainabilityb
Range
-
Mean
X
Range
Mean
Range
nc
4
3 4
4
(32)
8-60
4
(63)
-
2
0-0.2 1.4-3.0
77 (41)
74-81
4
-
4
0.3-1.4
(41)
35-51
7
-
amphiploid AAAB A. hypogaea x amphiploid
%om ICRISAT (1981). bFigures in parentheses indicate that data were not available for all hybrids or amphiploids. 'AB-AABB and AA-AAAA comparisons are for the same species combinations, but comparisons of amphiploids and hybrids with A. hypogaea involve different species combinations.
30
H.T. STALKER AND J . P. MOSS
hypogaea. Direct intersectional hybridization with A .hypogaea is currently not a possible means of introgression from wild to the cultivated species. Either hormone treatment and/or embryo rescue will be necessary to produce hybrids, or highly crossable genotypes of Arachis must be found before germplasm from most species of the genus can be transferred to A . hypogaea. Hybrids have been produced between annual diploids (with either an A or B genome) of section Aruchis and a number of section Rhizomatosae accessions, and between section Arachis species and two accessions of section Erectoides (Gregory and Gregory, 1979). This suggests that diploid section Arachis species may be useful for introgressing genes from species distantly related to the cultivated peanut. Gregory and Gregory (1979) suggested elements in common between sections on the basis of crossability, even though the hybrids were all highly sterile. Stalker (1985b) studied a number of intersectional crosses including a 40-chromosome (Erectoides x Erectoides) x (Erectoides x Rhizomatosae) hybrid. The mean number of univalents was less than three, indicating a considerable degree of pairing between members of sections Erectoides and Rhizomatosae chromosomes. A common genome to taxa of both these sections is likely to exist. The intersectional hybrids have potential for making additional crosses and for ploidy manipulation to introgress genes into A . hypogaea, but there have not been concentrated efforts to do this. FOR INTERSPECIFIC HYBRIDIZATION D. In Vitro TECHNIQUES
I.
Cell and Protoplast Culture
Callus is generally easy to produce in Arachis but plant regeneration is difficult. Peanut cotyledons have been used to define biochemical parameters for callus growth in cultivars (Verma and van Huystee, 1971; Verma and Marcus, 1974; Russo and Varnell, 1978; Guy et al., 1978). However, regeneration of plants from cotyledon callus of peanuts has not been obtained. Plants were recovered from tissue segments obtained by freeze-shattering cotyledons and growing the segments on moist filter paper (Illingworth, 1968). Mroginski et al. (1981) and Pittman et al. (1983) were able to regenerate plants in vitro from 3- to 5-day-old immature leaves. Isolation of protoplasts in peanuts was first obtained by Jullian (1970), who mechanically broke the cells. Verma and van Huystee (1971) observed that the cell suspension cultures were heterogenous in size and balanced growth was unlikely. They then developed a method to obtain large numbers of uniform cells in suspension. By using mannitol solutions of different molarity, Holden and Hildebrandt (1972) were able to obtain protoplasts without mechanically injuring the cells. Callus and roots have
SPECIATION, CYTOGENETICS, AND UTILIZATION
31
subsequently been produced from single cells of peanuts (Yung-ru and YuHung, 1978). In summary, callus and protoplast culture techniques could greatly facilitate introgression to A . hypogueu for producing both initial F, hybrids and for manipulating polyploid levels to recover 40-chromosome hybrid derivatives. However, in Aruchis the basic work to regenerate plants from single cells or callus is needed before techniques will be useful as a plant breeding tool. 2. Ovule and Embryo Culture
In vitro culture of ovules or embryos has successfully been used to produce interspecific hybrids is many genera (Narayanaswami and Norstog, 1964; Raghavan, 1980; Collins et ul., 1984). Numerous reviews have also been published describing media requirements and technical aspects of tissue preparation (North, 1976; Raghavan, 1977,1980;Williams et a/., 1982; Collins et ul., 1984). Embryo rescue in peanuts has had a long but sporadic history. Harvey and Schulz (1943) and then Nuchowiz (1955) initiated studies on regenerating peanut embryos. Martin (1970) regenerated peanut ovules only 0.3 mm in length to produce viable plants. However, in an attempt to duplicate Martin’s results, Sastri et ul. (1980), using the same media, could only produce callus, which became necrotic. Further, Johnson (198 1) concluded that in vitro culture of small embryos required a two-step process in which ovules could be cultured until they became large enough to dissect embryos, and then embryos could be cultured to generate plants. Embryo culture in peanuts is difficult for some genotypes, while easy for others. Ziv and Zamski (1975) produced callus and mature seeds from pegs grown on the plant which were allowed to elongate into media until the pod enlarged and embryos reached the heart stage. They then cultured the heart stage embryo in vitro. Ziv and Sager (1984)further found that embryos would develop into young seedlings under red, blue, or far-red subsaturated flux densities, but pod formation was inhibited. Moss et a/. (1985) cultured 1- to 4-day-old peg tips of A . hypogueu and reported ovule growth in many of the cultures. Reports of rescuing interspecific hybrid embryos are less frequent than reports of in vitro culture of A . hypogueu. Bajaj et al. (1982) cultured 30-dayold F, embryos of A . hypogueu x A . villosu. However, this is a hybrid combination which can relatively easily be obtained without the aid of in vitro techniques. Bajaj (1984) reviewed the tissue culture literature and concluded that peanut self and hybrid embryos can be cultured in vitro, but application of techniques are yet to be realized. A series of media varying in auxins, cytokinins, and gibberellic acid produced no significant breakthroughs in ovule culture (Sastri et ul., 1982). Mallikarjuna and Sastri (1985b) and Stalker et ul.
H. T. STALKER AND J. P. MOSS
32
(1985) found genotypic differences for responses to gibberellic acid treatments to stimulate peg elongation in the interspecific crosses. Pod and seed development have further been stimulated by applying hormones to peg tissues 10 days after pollination (Sastri and Moss, 1982; Mallikarjuna and Sastri, 1985a). Although hybrid ovules have been stimulated to expand and shoots and roots to develop, fewer than five intersectional hybrids have been recovered (Sastri and Moss, 1982). A major obstacle to hybrid recovery is establishing cultured plants in the greenhouse.
IX.
SUCCESSES AND POTENTIALS FOR UTILIZING Arachis GERMPLASM
Thousands of interspecifichybrids have been produced in many genera between wild species or between wild species and cultivars, but introgression of genes from wild species to genotypes with agronomic potential and subsequent cultivar release are rare. The major constraint to efficient utilization of interspecific hybrids is low fertility, which results in small populations and the inability to select desirable recombinants. Not only must the desirable genetic material be transferred into the cultivated genome, but yield and quality standards must also be met before wild species can be utilized by the grower. Arachis monticola has been used for two cultivar releases. This situation is analagous to the use OfAvenasteralisL. for improvementofA. sativaL. (Frey, 1976) or Hordeum spontaneum C. Kosch for improving H. vulgare L. (Rodgers, 1982), in which cases wild species which are completely crosscompatible with the cultigen were hybridized and selected. Reports of utilization of species germplasm from taxa in which the first generation hybrids are sterile are much more difficult to find in the literature (Stalker, 198Oa). Whole genomes may be added to a cultivar (such as in the case of sugarcane), or, when haploidy techniques are available (such as in the Solanaceae genera Solanum), diploids can readily be produced to select progenies with desired traits and tetraploids then be regenerated. However, few wild species in legume genera have been utilized for crop improvement. In peanut, Stalker et al. (1979) described a highly variable interspecifichybrid population derived from an A. hypogaea x A. cardenasiicrosswhich was selfed for at least eight generations and hybrid derivatives having 40chromosomes recovered. Selections were made from the population for C. arachidicola (Stalker, 1984), C. personatum (ICRISAT, 1984), and P.arachidis resistances (ICRISAT, 1984) and for resistance to several insects (Stalker and Campbell, 1983). Several interspecific hybrid populations are in the final stages of testing for release as general breeding materials, and some are already being used in several breeding programs. Making selections for high yields is as important as developingcultivarswith increased diseaseresistances,and a recurrent selection
SPECIATION, CYTOGENETICS, AND UTILIZATION
33
program has resulted in high-yielding lines from the same A. hypogueu x A. curdenusii population which have potential for cultivar release (Guok et ul., 1986). Several of the lines are now being tested in advanced yield trials in North Carolina and Virginia. Lines with high yields and C. personuturn and P. armhidis resistances are also being tested in the national Indian yield trial. Not only are seed yields high in these lines, but there is little defoliation making the vegetative material an extremely valuable source for cattle feed. Efforts to utilize species other than A. curdenusii are also being made to improve peanut production. Many fertile interspecific hybrid populations resulting from triploids or amphidiploids of Aruchis species are currently being evaluated in several peanut breeding programs. Several of these hybrid derivatives have potential as cultivar releases. In most crop species many cultivars have been released which have disease and insect resistances. However, in peanut only the cultivar NC 6 has been released for insect (southern corn rootworm) resistance (Wynne et al., 1977). In addition, several cultivar releases have been made for disease resistances, including NC 8C, selected for Cylindrocladium black rot resistance (Wynne and Beute, 1983); Va 81B, selected for resistance to Sclerotina blight (Coffelt et al., 1984); Southern Runner, resistant to late leafspot (Gorbet et ul., 1986); and several rosette-resistant cultivars in Africa (Gillier, 1978). Obviously missing from this list are cultivars resistant to the important diseases caused by early leafspot (C. urachidicolu) and peanut rust (P. armhidis). Further, cultivars used in the United States have been selected for rather narrow environments and only one genotype for each disease has been released. This is the result of several interrelated factors. First, during the early days of peanut production relatively few diseases were economically important. In the United States, diseases have been controlled with chemicals, and there has been the common belief until recently that no variability was available in cultivated accessions. This last point concerning lack of variability resulted in large efforts by botanists, cytogeneticists, and plant breeders in cooperation with pathologists and entomologists to collect, evaluate, and attempt to utilize the wild species of the genus. In many other crop species the plant breeder has been responsible for acquiring, evaluating, and utilizing germplasm resources of related wild species along with those of the crop species. In Aruchis, efforts have been more diversified, with teams of cytogeneticists working with plant breeders to identify and manipulate useful germplasm. Fortunately, many of the most important traits are found in species which are cross-compatible with the cultivated peanut. In the future, selection for leafspot and rust resistances have a high probability of making a significant contribution to peanut production, especially in tropical and semitropical areas where chemicals are infrequently used by the grower. Significant progress toward exploiting the variability in the genus Aruchis is being made. However, the most productive pathway to obtain 40-chromosome hybrid derivatives has not been determined, and perhaps
34
H.T. STALKER AND J. P. MOSS
no one method will prove to be the best for all traits. Interspecific hybrid derivatives which are cross-compatible with A. hypogaea have been selected for fertility. Most desired traits in Arachis are multigenic and large populations must be derived to utilize the germplasm resources. Like other crop species, hybridization and fertility restoration processes are not easily accomplished, but significant progress is being made toward utilizing wild species of Arachis.
X.
CONCLUSIONS
Utilization of wild species requires a collective effort by botanists, geneticists, cytogeneticists, and plant breeders. In Armhis there exists a large reservoir of germplasm with agronomic potential in numerous taxa. A taxonomic treatment of the genus is urgently needed to avoid confusion and to open communication channels among scientists. However, efforts have been made to evaluate available accessions for the economically important disease and insect pests of peanuts. Arachis hypogaea is an allotetraploid species and will only hybridize with members of section Arachis. In this group are 15-20 species which have genomic similarities to the cultivated peanut. In addition, species accessions have been identified which are highly resistant to the most severe pests of cultivated peanuts. Utilization of these germplasm resources is hindered by sterility of interspecific hybrids due to ploidy and, to a lesser extent, genomic incompatibilities. Attempts to introgress genes conferring resistances to different diseases and insects have proceeded through direct hybridization and through ploidy manipulations. Traits can be transferred to the A. hypogaea genome using either pathway, but eliminating undesirable traits is a major problem which must be solved before growers will actually use the available germplasm. The methodology to utilize species outside section Arachis is proceeding in the area of embryo rescue and recovery of hybrids via growth regulator applications to reproductive structures. However, the cytogenetic similarities among genomes of taxa in different sections remain unclear. Because hybrid embryos can be obtained for wide crosses in the genus, specialized techniques such as protoplast fusion may be unnecessary to utilize the germplasm resources. Regardless, efforts to utilize distantly related species should only be conducted when desired traits are not found in the cultivated species or in species closely related to A . hypogaea. Wild peanut species have been used to create highly variable plant populations which contain early and late leafspot, rust, and insect resistances. Further, transgressive segregates for yield have resulted from
SPECIATION, CYTOGENETICS, AND UTILIZATION
35
crossing an unadapted A. hypogueu accession with a small-seeded diploid species. High yields were unexpected, but results of introgression greatly enhanced the potential exploitation of Aruchis species germplasm. Perhaps the most valuable use of these hybrid derivatives will be in tropical and semitropical areas where chemicals are infrequently used, but even small increases in disease resistance will result in significant yield increases. In addition, fodder production is an important consideration in many production areas outside the United States. Yield and disease trials are now being conducted with advanced breeding lines derived from wild diploid species of A ruchis.
REFERENCES Abdou, Y. A. M., Gregory, W. C., and Cooper, W. E. 1974. Peanut Sci. 1, 6-11. Amaya, J. F., Young, C. T., and Hammons, R. 0. 1977. Oleaginewr 32, 225-229. Amin, P. W. 1985. In “Proceedings of an International Workshop on Cytogenetics of Arachis,” pp. 57-60. ICRISAT Center, Patancheru, A.P., India. Amir, J. 1969. Ann. Bot. 33, 33-338. Ashri, A. Offenbach, R., Cahaner, A., and Levy, A. 1977. 2. PfTanzenzuecht. 79, 220-228. Babu, C. N. 1955. Indian J. Agric. Sci. 25,41-46. Bajaj, Y. A. S. 1984. In “Handbook of Plant Cell Culture” (P.V. Ammirato, D. A. Evans, W. R. Sharp, and Y. Yamada, eds.), Vol. 3, pp. 193-225. Macmillan, New York. Bajaj, Y. A. S., Kumar, P., Singh, M.M., and Labana, K. S. 1982. Euphytica 31, 370-375. Banks, D. J. 1969. J. Am. Peanut Res. Educ. Assoc. 1, 23-28. Bentham, G. 1841. Trans. Linn. SOC. (London) 18, 155-162. Brennan, J. R. 1969. Biologists 51, 71-82. Bunting, A. H., Gibbons, R. W., and Wynne, J. C. 1985. In “Grain Legume Crops” (R. J. Summerfield and E. H. Roberts, eds.), pp. 747-800. Collins, London. Campbell, W. V., and Wynne, J. C. 1980. In “Proceedings of the International Workshop on Groundnuts,” pp. 149-157. ICRISAT, Patancheru, A.P., India. Castillo, M. B., Morrison, T. S., Russel, C. C., and Banks, D. J. 1973. J. Nematol. 5,281-285. Chevalier, A. 1933. Rev. Bot. Appl. Agric. Trop. 13, 689-789. Chevalier, A. 1934. Rev. Bot. Appl. Agric. Trop. 14, 565-632, 709-755, 833-864. Chevalier, A. 1936. Rev. Bot. Appl. Agric. Trop. 16, 673-871. Coffelt, T. A., Mozingo, R. W., Porter, D. M., and Allison, A. H. 1984. Va. Agric. Exp. Sta. Bull. 84-1. Collins, G. B., Taylor, N. L., and De Verna, .IW. . 1984. In “Gene Manipulation in Plant Improvement” (J. P. Gustafson, ed.), pp. 323-383. Plenum, New York. Company, M., Stalker, H. T., and Wynne, J. C. 1982. Euphytica 31, 885-893. Davis, K. S., and Simpson, C. E. 1976. Proc. Am. Peanut Res. Educ. SOC.8, 93 (Abstr.). D’Cruz, R., and Tankasale, M. P. 1961. Indian Oilseeds J. 5 , 58-59. Demski, J. W., and Sowell, G., Jr. 1981. Peanut Sci. 8, 43-44. Elkan, G. H., Wynne, J. C., and Schneeweis, T. J. 1981. Trop. Agric. (Trinidad) 58, 297-305. Fitzner, M. S., Alderman, S. C., and Stalker, H. T. 1985. Proc. Am. Peanut Res. Educ. SOC. 17, 28 (Abstr.). Foster, D. J., Stalker, H. T., Wynne, J. C., and Beute, M. K. 1981. Oleagineux 36, 139-143. Frey, K. J. 1976. Egypt. J. Genet. Cytol. 5 , 4 6 0 4 2 .
36
H. T. STALKER AND J. P. MOSS
Gardner, M. E. B., and Stalker, H. T. 1983.Crop Sci. 23, 1069-1074. Ghanekar, A. M. 1980. In “Proc. International Workshop on Groundnuts,” pp. 211-216. ICRISAT, Patancheru, A.P., India. Gibbons, R. W. 1969. In “Third East African Cereals Conf.,” Zambia and Malawi. Mimeo report. Gibbons, R. W. 1980.In “Advances in Legume Science” (R. J. Summerfield and A. H. Bunting, eds.), pp. 483-493. Royal Botanic Gardens, Kew. Gibbons, R. W., and Bailey, B. E. 1967.J. Agric. Res. 5, 57-59. Gibbons, R. W., and Turley, A. C. 1967. Annu. Rep. Agric. Res. Council Cent. Afr. pp. 66-90. Gibbons, R. W., Bunting, A. H., and Smartt, J. 1972.Euphytica 21, 78-85. Gillier, P. 1978. Oleagineux 33, 25. Gorbet, D.W., Norden, A. J., Scokes, F. M., and Knauft, D. A. 1986.Flu. Agric. Exp. Circue. s-324,13. Gregory, M. P., and Gregory, W. C. 1979. J. Hered. 70, 185-193. Gregory, W. C. 1946. Res. Farm., 69th Annu. Rep., N.C. Agric. Exp. Sta. pp. 42-44. Gregory, W. C., and Gregory, M. P. 1976. In “Evolution of Crop Plants” (N. W. Simmonds, ed.), pp. 151-154.Longman, London. Gregory, W. C., and Gregory, M. P., Krapovickas, A., Smith, B. W., and Yarbrough, J. A. 1973. In “Peanuts: Culture and Uses,” Chap. 3. Am. Peanut Res. Educ. Assoc., Stillwater, OK. Gregory, W. C., Krapovickas, A., and Gregory, M. P. 1980.In “Advances in Legume Science” (R. J. Summerfield and A. H. Bunting, eds.), Vol. 1. International Legume Conference, Royal Botanic Gardens, Kew. Guok, H. P., Wynne, J. C., and Stalker, H. T. 1986.Crop Sci. 26, 249-253. Guy, A. L., Heinis, J. L., and Pancholy, S. K. 1978.Peanut Sci. 5, 78-82. Hadley, H. H., and Openshaw, S. J. 1980.In “Hybridization of Crop Plants” (W. R. Fehr and H. H. Hadley, eds.), pp. 133-159.Am. SOC.Agron., Madison, WI. Halward, T. M., and Stalker, H. T. 1984. Agron. Abstr. 69 (Abstr.). Halward, T. M., and Stalker, H. T. 1985. Agron. Abstr. 56 (Abstr.). Halward, T. M., and Stalker, H. T. 1987a.Ann. Bot., 59, 9-14. Halward, T. M., and Stalker, H. T. 1987b. Crop Sci., 27, 456-460. Hammons, R. 0.1970. Crop Sci. 10,459. Harlan, J. R. 1976. Crop Sci. 16,329-333. Harvey, R. H., and Schulz, E. F. 1943. J. A m . SOC.Agron. 35, 631. Hawkes, J. G. 1977. Euphytica 26, 615-621. Hebert, T. T., and Stalker, H. T. 1981. Peanut Sci. 8, 45-47. Hermann. F. J. 1954.Agr. Monogr. U.S.D.A. (19). Hoehne, F. C. 1940. Flora Bras. 25(lI) 122, 1-20. Holden, D. J., and Hildebrandt, A. C. 1972 Phyton 29,47-54. Hull, F. H., and Carver, W. A. 1938. Flu. Agric. Exp. Sta. Annu. Rep. pp. 39-40. Husted, L. 1931.Am. Nut. 65,476-477. Husted, L. 1933. Cytologia 5, 109-117. Husted, L. 1936. Cytologia 7, 396-423. ICRISAT (International Crops Research Institute for the Semi-Arid Tropics). 1981-1985. “Annual Reports.” Patancheru, A.P., India. Illingworth, J. E. 1968. Hortic. Sci. 3, 274-276. Jackson, C. R., and Bell, D. K. 1%9. Go. Agric. Exp. Sta. Res. Bull. 56. Jacobs, W. P. 1947.Am. J. Bot. 34, 361-370. Jacobs, W.P. 1951. Am. J. Bot. 38, 307-310. Jahnavi, M. R., and Murty, U. R. 1985a. Cytologia 50, 747-758.
SPECIATION, CYTOGENETICS, AND UTlLIZATION
37
Jahnavi, M. R., and Murty, U. R. 1985b. Theor. Appl. Genet. 70, 157-165. Johansen, E. L., and Smith, B. W. 1956. Am. J. Bot. 43, 250-258. Johnson, B. B. 1981. Environ. Exp. Bot. 21, 439 (Abstr.). Johnson, D. R., Wynne, J. C., and Campbell, W. V. 1977. Peanut Sci. 4,9-11. Jullian, M. 1970. C.R. Hebd. Acad. Sci, Paris 270, 3051-3054. Kawakami, J. 1930. Bot. Mag. (Tokyo) 44. 319-328. Kirti, P. B., Bharathi, M., Murty, U. R., and Rao, N. G. P. 1983. Cytologia 48, 139-151. Klesser, P. J. 1967. S. Afr. J. Agric. Sci. 10, 919-927. Knott, D. R., and Dvorak, J. 1976. Annu. Rev. Phytopathol. 14,211-235. Kolawole, K. B. 1976. Samaru Agric. Newslett. 18, 40-43. Krapovickas, A. 1%8. In “The Domestication and Exploitation of Plants and Animals” (P. J. Ucko and 1. S. Falk, eds.), pp. 424-441. Duckworth, London. Krapovickas, A. 1969. Semin. Adv. Genet. Agric. Am. Lat., Maracay, Venezuela SAGA/B(d), 1-4. Krapovickas, A. 1973. In “Agriculture Genetics. Selected Topics” (R. Moav, ed.), pp. 135-151. Wiley, New York. Krapovickas, A., and Rigoni, V. A. 1951. Rev. Invest. Agric. (Buenos Aires) 5,289-293. Krapovickas, A., and Rigoni, V. A. 1957. Darwiniana 11,431-455. Krapovickas, A., Fernandez, A., and Seeligman, P. 1974. Bonplandia 3, 129-142. Kuhn, C. W., Sowell, G., Jr., Chalkley, J. H., and Stubbs, H. F. 1968. Plant D k . Rep. 52, 467-468. Kuhn, C. W., Paquio, 0. R., and Adams, D. B. 1978. Plant Dis. Rep. 62, 365-368. Kumar, L. S. S., D’Cruz, R., and Oke, J. G. 1957. Curr. Sci. 4, 121-122. Leuck, D. B., and Hammons, R. 0. 1968. J. Econ. Entomol. 61,687-688. Lynch, R. E., Branch, W. D., and Garner, J. W. 1981. Peanut Sci. 8, 106-109. Madhava, Menon, P., Raman, V. S., and Krishnaswami, S. 1970. Madras Agric. J. 57, 80-82. Mallikarjuna, N., and Sastri, D. C. 1985a. In “Proceedings of an International Workshop on Cytogenetics of Arachis,” pp. 147-151. ICRISAT, Patancheru, A.P., India. Mallikarjuna, N., and Sastri, D. C. 1985b. In “Proceedings of an International Workshop on Cytogenetics of Arachk ,” pp. 153-158. ICRISAT, Patancheru, A.P., India. Martin, J. P. 1970. Oleagineux 25, 155-156. Melouk, H. A., and Banks, D. J. 1978. Peanut Sci. 5, 112-114. Mendes, A. J. T. 1947. Bragantia 7, 257-267. Miller, L. I., and Troutman, J. L. 1966. Plant Dis. Rep. 50, 139-143. Moss, J. P. 1985a. In “Proceedings of an International Center Seminar on IARC’s and Biotechnology.” lnt. Rice Res. Inst., Los Banos, Laguna, Philippines. Moss, J. P. 1985b. In “Proceedings of an International Workshop on Cytogenetics of Arachk,” pp. 93-99. ICRISAT, Patancheru, A.P., India. Moss, J. P., Stalker, H. T., and Pattee, H. E. 1985. Proc. Am. Peanut Res. Educ. SOC.17, 22 (Abstr). Mroginski, L. A., Kartha, K. K., and Shylok, J. P. 1981. Can. J. Bot. 59, 826-830. Murty, U. R., Rao, N. G. P., Kirti, P. B., and Bharathi, M. 1980. T A R 2 Reg. Sta. Rep., Hyderbad, India. Narayanswami, S . , and Norstog, K. 1964. Bot. Rev. 30, 587-628. Norden, A. J. 1980. In “Hybridization of Crop Plants,” (W. R. Fehr and H. H. Hadley, eds.), pp. 443-456. Am. Soc. Agron., Crop Sci. SOC.Am., Madison, WI. North, C. 1976. Acta Hortic. 63.67-74. Nuchowicz, A. 1955. L. Agric. Louvain 3, 3-37.
38
H.T. STALKER AND J. P. MOSS
Pallai, S. K., and Raju, E. C. 1975. Flora Ed. 1645, 487-496. Patil, S. H. 1968. Can. J . Gen. Cytol. 10, 545-550. Patil, S. H., and Bora, K. C. 1961. Indian J . Genet. 21, 59-74. Periasamy, K., and Sampoornam, C. 1984. Ann. Bol. 53, 399-411. Pittman, R. N., Banks, D. J., Kirby, J. S., Mitchell, E. D., and Richardson, P. E. 1983. Peanut Sci. 10, 21-25. Pompeu, A. S. 1977. Ci. Cult. 29, 319-321. Porter, D. M., Smith, D. H., and Rodriguez-Kabana, R. 1982. In “Peanut Science and Technology” (H. E. Pattee and C. T. Young, eds.), pp. 326-410. Am. Peanut Res. Educ. SOC.,Yoakum, TX. Prine, G. M., Dunavin, L. S., Moore, J. E., and Roush, R. D. 1981. Flu. Agric. Exp. Sta. Bull. S-275. Raghavan, V. 1977. In “Plant Cell, Tissue and Organ Culture” (J. Reinert and Y. P. S. Bajaj, eds.), pp. 375-397. Springer-Verlag, Berlin. Raghavan, V. 1980. Inf. Rev. Cytol. Suppl. 11B, 209-240. Raman, V. S. 1958. Indian Oilseeds J. 2, 20-23. Raman, V. S. 1976. “Cytogenetics and Breeding in Arachis.” Today and Tomorrow’s Printers and Publishers, New Delhi. Raman, V. S., and Kesavan, P. C. 1962. Nucleus (Calcutta) 5, 123-126. Resslar, P. M., 1980. Euphytica 29, 813-817. Resslar, P. M., and Gregory, W. C. 1979. J. Hered. 70, 13-33 16. Rodgers, D. M. 1982. Diss. Absfr. Int. 43, 4, 9378-938B. Russo, S. L., and Varnell, R. J. 1978. Soil Crop Sci. SOC.Flu. Proc. 37, 34-36. Saint-Smith, J. H . , McCarthy, G. J. P., Rawson, J. E., Langford, S., and Colbran, R. C. 1972. “Peanut Growers Advisory, Leaflet No. 1178,” p. 18. Div. Plant Industry, Dept. Plant Industry, Dept. of Primary Industries, Australia. Sastri, D. C., and Moss, J. P. 1982. J. Exp. Eot. 33, 1293-1301. Sastri, D. C., Nalini, M. S., and Moss, J. P. 1980. Proc. Symp. Plant Tissue Cult., Genet. Manip. Somat. Hybrid. pp. 366-373. Bhabha At. Res. Center, Trombay, India. Sastri, D. C., Nalini, M. S., and Moss, J. P. 1982. In “Tissue Culture of Economically Important Plants: Proceedings of an International Syposium”(A. N. Rao, ed.), pp. 42-57. Committee on Science and Technology in Developing Countries and Asian Network for Biological Sciences, National University of Singapore, Singapore. Schenk, R. U. 1961. Ga. Agric. Exp. Sta. Tech. Bull. 22, 5-53. setharam, A., Nayar, K. M. D., Sreekantaradhya, R., and Achar, D. K. T. 1973. Cytologia 38, 277-280. Sharief, Y., Rawlings, J. O., and Gregory, W. C. 1978. Euphytica 27, 741-751. Simpson, C. E., and Davis, K. S. 1983. Crop Sci. 25, 581-584. Simpson, C. E., and Smith, 0. D. 1974. Tex. Agric. Exp. Sta. Bull. L-1348. Singh, A. K. 1985. In “Proceedings of an International Workshop on Cytogenetics of Arachis,” pp. 107-1 17. ICRISAT, Patancheru, A.P., India. Singh, A. K. 1986a. Theor. Appl. Genet. 72, 164-169. Singh, A. K. 1986b. Theor. Appl. Genet., 72, 433-439. Singh, A. K., and Moss, J. P. 1982. Theor. Appl. Genet. 61, 305-314. Singh, A. K., and Moss, J. P. 1984a. Theor. Appl. Genet. 68, 355-365. Singh, A. K., and Moss, J. P. 1984b. Peanut Sci. 11, 17-21. Singh, A. K., Subrahmanyam, P., and Moss, J. P. 1984. Oleagineux 39, 535-537. Smartt, J. 1964. Proc. Nat. Peanut Res. Conf. 3rd. Auburn, AL. pp. 53-56. Smartt, J. 1979. Econ. Bol. 33, 329-337. Smartt, J., and Gregory, W. C. 1967. Oleagineux 22,455-459. Smartt, J., and Stalker, H. T. 1982. In “Peanut Science and Technology” (H. E. Pattee and C. T. Young, eds.), pp. 21-49. Am. Peanut Res. Educ. SOC.,Yoakum, TX.
SPECIATION, CYTOGENETICS, AND UTILIZATION
39
Smartt, J., Gregory, W. C., and Gregory, M. P. 1978a. Euphyticu 27, 665-675. Smartt, J., Gregory, W. C., and Gregory, M. P. 1978b. Euphyticu 27, 677-680. Smith, 9. W. 1950. Am. J. Eot. 37, 802-815. Smith, B. W. 1956. Am. J. Bot. 43, 81-89, 233-240. Smith, D. H . , and Littrell, R. H. 1980. PIanr Dis. Rep. 64, 356-361. Smith, J. W., Jr., and Barfield, C. S. 1982. In “Peanut Science and Technology” (H. E. Pattee and C. T. Young, eds.), pp. 250-325. Am. Peanut Res. Educ. SOC.,Yoakum, TX. Spielman, I. V., and Moss, J. P. 1976. Oleugineux 31, 491-494. Spielman, I. V.. Burge, A. P., and Moss, J. P . 1979. 2. Pfunzenzuecht. 83, 236-250. Stalker, H. T. 1980a. Adv. Agron. 33, 111-147. Stalker, H. T. 1980b. In “Proceedings of an International Workshop on Groundnuts,” pp. 73-81. ICRISAT, Patancheru, A.P., India. Stalker, H. T. 1981. Crop Sci. 21, 359-362. Stalker, H. T. 1984. Euphyticu 33, 529-538. Stalker, H. T. 1985a. Proc. Am. Peanut Res. Educ. Soc. 17, 23 (Abstr.). Stalker, H. T. 1985b. In “Proceedings of an International Workshop on Cytogenetics of Aruchis,” pp. 65-79. ICRISAT, Patancheru, A.P., India. Stalker, H. T. 1985~.In “Proceedings of an International Workshop on Cytogenetics of Aruchis,” pp. 119-123. ICRISAT, Patancheru, A.P., India. Stalker, H. T., and Campbell, W. V. 1983. Peanut Sci. 10, 30-33. Stalker, H. T., and Dalmacio, R. D. 1981. J. Hered. 72, 403-408. Stalker, H. T., and Dalmacio, R. D. 1986. Curyologia, 51, 617-629. Stalker, H. T., and Wynne, J. C. 1979. Peanut Sci. 6, 110-114. Stalker, H. T., and Wynne, J. C. 1983. Peanut Sci. 10, 59-62. Stalker, H . T., Wynne, J. C., and Company, M. 1979. Euphyticu 28, 675-684. Stalker, H. T., Campbell, W. V., and Wynne, J. C. 1984. J. Econ. Ent.. 7, 53-57. Stalker, H. T., Moss, J. P., and Seitz, M. H. 1985. Agron. Abstr. 1985, 72, (Abstr.). Subrahmanyam, P., Mehan, V. K., Nevill, D. J., and McDonald, D. 1980. In “Proceedings of an International Workshop on Groundnuts,” pp. 193-198. ICRISAT, Patancheru, A.P., India. Subrahmanyam, P., Williams, J. H., McDonald, D., and Gibbons, R. W. 1984. Ann. Appl. Biol. 104,467-476. Sunrahmanyam, P., Ghanekar, A. M., Knolt, 9. L., Reddy, C. V. R., and McDonald, D. 1985. In “Proceedings of an International Workshop on Cytogenetics of Aruchis,” pp. 49-55. ICRISAT, Patancheru, A.P., India. Taubert, P. 1894. In “Die natuerlichen Pflanzenfamilien 111 Tiel” (A. Engler and K. Prantl, eds.), pp. 322, 324-325, Part 3. Valls, J. F. M., Rao, V. R., Simpson, C. E., and Krapovickas, A. 1985. In “Proceedings of an International Workshop on Cytogenetics of Aruchis,” pp. 15-33. ICRISAT, Patancheru, A.P., India. Varisai Muhammad, S. 1973. Madras Agric. J. 60, 1428-1432. Verma, D. P. S., and Marcus, A. 1974. J. CeN Sci. 41, 331-337. Verma, D. P. S., and van Huystee, R. 9. 1971. Exp. Cell Res. 6 9 , 4 0 2 4 8 . Watson, I. A. 1970. In “Genetic Resources in Plants-Their Exploration and Conservation” (0.H. Frankel and E. Bennett, eds.), pp. 441-457. Davis, Philadelphia. Williams, E. G.,Verry, I. M.,and Williams, W. M. 1982. In “Plant Improvement and Somatic Cell Genetics” (I. K. Vasil, W. R. Scowcroft, and K. J. Frey, eds.), pp. 119-128. Academic Press, New York. Woodroof, J. G. 1973. In “Peanuts-Production, Processing, Products,” 2nd Ed. Westport, CT . Wynne, J. C. 1974. Ph.D. thesis, N.C. State Univ., Raleigh. Wynne, J. C., and Beute, M.K. 1983. Crop Sci. 23, 184.
40
H.T. STALKER AND J. P. MOSS
Wynne, J. C., and Coffelt, T. A. 1982. I n “Peanut Science and Technology” (H. E. Pattee and C. T. Young, eds.), pp. 50-94. Am. Peanut Res. Educ. SOC.,Yoakum, TX. Wynne, J. C., and Gregory, W. C. 1981. Adv. Agron. 34, 39-72. Wynne, J. C., Campbell, W. V., Emery, D. A,, and Mozingo, R. W. 1977. N.C. Agric. Exp. Sta. Bull. 458. Yasuda, S . 1943. Jpn. J. Bot. 13, 243-253. Yung-ru, S., and Yu-hung, L. 1978. Proc. Symp. Plant Tissue Cult., Peking p. 529. Ziv, M. 1981. Ann. Bof. 48,353-359. Ziv, M., and Sagar, J. C. 1984. Plant Sci. Lett. 34, 211-218. Ziv, M., and Zamski, E. 1975. Ann. Bot. 39, 579-583.
ADVANCES IN AGRONOMY, VOL. 41
CEREAL-LEGUME INTERCROPPING SYSTEMS Francis Oforili2and W. R. Stern2 'Plantations Limited, Ghana Cocoa Board, Private Post Bag Accra North, Ghana, West Africa 'Agronomy Group, School of Agriculture The University of Western Australia Nedlands, Western Australia, Australia 6009
1.
INTRODUCTION
In terms of land use, growing crops in mixed stands is regarded as more productive than growing them separately (Andrew and Kassam, 1976; Willey, 1979). Mixed cropping is practiced traditionally in many parts of Africa, Asia, and Latin America (Ahmed et al., 1979) and interest in cereal-legume intercropping is developing in some temperate regions with warm climates such as Australia and the United States (Searle et al., 1981; McCollum, 1982; Allen and Obura, 1983; Chui and Shibles, 1984). This may be due to some of the established and speculated advantages for intercropping systems such as higher grain yields, greater land use efficiency per unit land area, and improvement of soil fertility through the addition of nitrogen by fixation and excretion from the component legume (Agboola and Fayemi, 1972; Willey, 1979; Eaglesham et al., 1981). It seems worthwhile to develop cropping systems that have the capacity to maximize crop yields per unit land area while keeping the fertilizer nitrogen requirement to a minimum. The intercropping of legumes with cereals offers scope for developing energy-efficient and sustainable agriculture (Papendick et al., 1976; IAEA, 1980). Since the reviews by Willey (1979) on intercropping and by Trenbath (1974) on mixed cropping, many papers have been published from both temperate and tropical areas that emphasize general agronomic principles and research needs of intercropping systems. The recent book on multiple cropping systems (Francis 1986) summarizes some of that work and also develops some new ideas. The major emphasis in this chapter is on cereal-legume intercrop systems and we review the available information on various crop combinations. We examine cultural factors such as choice of compatible component crops with diverse morphology, crop geometry and 41 Copyright 0 1987 by Academic Press, Inc. All rights of reproduction in any form reserved.
42
FRANCIS OFORI AND W. R. STERN
density, relative time of sowing of component crops, and competitive relations, as well as nitrogen fertilizer application and nitrogen economy.
II. BACKGROUND TO INTERCROPPING SYSTEMS Intercropping is the growing of two or more crop species simutaneously in the same field during a growing season. Other suggested definitions are (a) any form of cropping system in which there is a significant amount of intercrop competition (Willey, 1979); and (b) the growing of two or more crops in different but proximate rows (Ruthenberg, 1980). Common characteristics of different forms of intercropping are that they have the effect of intensifying crop production and exploiting more efficiently environments with limiting or potentially limiting growth resources (Papendick et al., 1976; Trenbath, 1982). Andrews and Kassam (1976) identified four main types of intercropping: 1. Mixed intercropping: growing component crops simultaneously with no distinct row arrangement; this is commonly used in labor-intensive subsistence farming. 2. Row intercropping: growing component crops simultaneously in different rows; this is used in mechanized agriculture and in present research on intercropping, because it permits crop-specific operations. 3. Strip intercropping: growing component crops simultaneously in different strips to permit the independent cultivation of each crop. 4. Relay intercropping: growing component crops in relay, so that growth cycles overlap.
A. CROP COMBINATIONSAND GLOBALDISTRIBUTION Crop combinations differ with geographical location; there may be intercropping of tree crops, intercropping of tree and field crops, or intercropping of field crops (Trenbath, 1982). On the basis of morphology and growth duration, Herrera and Harwood (1973) distinguished the following crop combinations: crops of similar heights and growth durations such as barley and oats; crops of similar morphology but different growth durations, e.g., 6-month sorghum and 3-month millet; annual or biennial crops with those of longer growth durations such as millet and cassava or soybean and sugar cane; annual crops of cereals; and legumes such as sorghum and pigeonpea, maize and cowpea. Combinations of crops are determined primarily by the length of the growing season and the adaptation of crops to particular environments. In areas with annual rainfall of less than 600 mm and a short growing season, such as northern Nigeria, early-maturing and drought-tolerant crops such
CEREAL-LEGUME INTERCROPPING SYSTEMS
43
as millet and sorghum dominate (Andrews, 1972; Baker, 1979). In areas with annual rainfall greater than 600 mm, cereals and legumes of varying maturities are used. In the tropical and subtropical regions, the cereal component is usually maize, sorghum, millet, or, to a lesser extent, rice, and the legume is usually cowpea, groundnut, soybean, chickpea, bean, or pigeonpea. Both early- and slow-maturing crops are combined to ensure efficient utilization of the whole growing season (Baker, 1979). The millet or sorghum and pigeonpea combinations of the Hyderabad area of India are a typical example (Ruthenberg, 1980). Maize seems to dominate as the cereal component and it is combined with many different legumes. In India, short-duration sorghum and millet (90days) are intercropped with pigeonpea that matures 90days later than the cereal (Willey, 1979). In high rainfall areas of West Africa, a common crop combination is maize and cowpea (Okigbo and Greenland, 1976), whereas in South and Central America, maize and different types of beans dominate (Francis et al., 1976). Combinations of rice and other cereals or legumes are also found in high rainfall areas with a single extended rainy season such as Southeast Asia (Ruthenberg, 1980). In some temperate regions with warm climates, intercrop systems consist of wheat, oats, or barley as the cereal component and field bean, vetch, lupin, or soybean as the legume component (Table I). Geographical patterns for different crop combinations are discernable (Table I), and these generally follow general climatic classificationssuch as the one of Koppen (Trewartha, 1954). B.
INDICES FOR EVALUATING PRODUCTIVITY AND EFFICIENCY
Different indices have been suggested for evaluating productivity and efficiency per unit area of land of cereal-legume intercrop systems (Willey, 1979, 1985; Beets, 1982). These include comparisons of absolute yields, protein yields, caloric equivalent, and in economic terms, gross returns from intercrops and sole crops. Evaluation in economic terms is considered inappropriate due to seasonal price fluctuations of inputs and the lack of cash economy in most areas where intercropping is practiced (Beets, 1982). Yields and prevailing prices of crops tend to fluctuate, composition and quality of crop products will vary, and energy contents and growth durations of the component crops differ, so combined yields are of little value (de Wit et al., 1966; Beets, 1982). Van den Bergh (1968) suggested relative yields for comparing performance of crops in intercropping mixtures. Willey (1 979) also suggested standardizing component crop yields in relation to sole crop yields in order to evaluate intercropping efficiency.
I . Relative Yield Total The mixture yield of a component crop expressed as a proportion of its yield as a sole crop from the same replacement series is the relative yield of the
Table I The Distribution of Various Intercrop Systems in Different Climatic Types
Crop combinations Maize intercropped with Beans
Climatic type after Koppen"
Irrigated or rain fedb
Country and location of study
Reference
Af Am Am Aw Aw Aw
R R R R R R
Costa Rica, Turrialba Sri Lanka, Peradeniya Colombia, Palmira Uganda, Kampala Tanzania, Morogoro Brazil, Pernambuco
9.56"N, 7.16"N. 3.33"N, 0.19"N, 6.49"N, 8.14"S,
83.48"W 80.37"E 76.17"W 32.53"E 37.40"E 38.00"W
Hart (1975)
cowpea
Af Af Am Am Aw Aw BS
R R R R R R
5.54"S, 9.56"N, 7.23"N, 7.16'N, 6.49"N, 8.14"N, 5.33"N,
76.07 "W 83.48"W 3.56"W 80.37"E 37.40"E 38.00"W 0.1S"W
Wade and Sanchez (1984) Chang and Shibles (1985a) Remison (1978) Gunasena et al. (1979) Enyi (1973) Mafra el of. (1979) Haizel (1974)
Groundnut
Am Aw Aw Ca Am Am Aw Ca Da Da Da
R R R R R R R
Peru, Yurimaguas Costa Rica, Turrialba Nigeria, Ibadan Sri Lanka, Peradeniya Tanzania, Morogoro Brazil, Pernambuco Ghana, Accra Sri Lanka, Peradeniya Tanzania, Morogoro India, Hyderabad India, Meerut Trinidad, St Augustine Sri Lanka, Peradeniya Zimbabwe, Harare Australia, Camden U.S.A., Minnesota U.S.A., Ames U.S.A., Amherst
7.16 ON, 80.37 "E 6.49"N, 37.40"E 17.22"N, 78.26"E 29.00"N, 77.42"E 3.38"N, 61.31"W 7.16"N, 80.37"E 17.43"S, 31.OS"E 34.04"S, 150.40"E 40.06"N. 91.46"W 42.02"N. 93.39"W 42.23"N, 72.31"W
Gunasena et af. (1979) Evans (1960) Nambiar et al. (1983) Gangwar and Karla (1982) Dallal (1974) Gunasena et a/. (1979) Beets (1977) Searle et al. (1981) Crookston and Hill (1979) Chui and Shibles (1984) Putnam et a/. (1985)
%
Soybean
1, R
1, R
I I I
Gunasena et al. (1979) Francis et al. (1982a) Willey and Osiru (1972) Enyi (1973) Mafra et of. (1979)
Blackgram Greengram
Ca Am
R
India, Meerut Nigeria, Ibadan
29.00"N, 7.23"N.
Clover Rice
Ca Af
I R
U.S.A., Urbana Phillipines, Manila
40.07'N, 88.12W 14.37"N, 120.58"E
Millet
Aw AW
R
Sorghum
R
Nigeria, Samaru Nigeria, Samaru
11.1l0N, 11.1l0N,
7.38"E 7.38"E
Cassava
Af
R
Costa Rica, Turrialba
9.56"N,
83.48"W
Aw Aw Aw
R R R
Uganda, Kampala Tanzania, Morogoro Brazil, Pernambuco
Cowpea
Aw Aw Ca BS
R R R I
Tanzania, Morogoro Brazil, Pernambuco India, New Delhi U.S. A., Riverside
0.19"N, 32.35"E 6.49"S, 37.40"E 8.149, 38.00"W 6.49"s. 37.40"E 8.14"S, 38.00"W 28.37"N, 77.13"E 33.59"N, 117.22"W
Groundnut
Aw Ca
R R
Tanzania, Morogoro India, New Delhi
6.49"S, 28.37"N,
37.40"E 77.13"E
Soybean
Aw Ca Da
R R I
Blackgram
Ca Aw
R R
Puerto Rico, Isabela India, New Delhi U.S.A., Urbana India, New Delhi Tanzania, Morogoro
18.30°N, 28.37"N, 40.07ON, 28.37"N, 6.49"S,
67.20"W 77.13"E 88.12OW 77.13"E 37.40"E
Ca
R
India, New Delhi
28.37"N.
77.13"E
Singh (1981)
Aw Aw Ca
R R R
Tanzania, Morogoro Brazil, Pernambuco India, New Delhi
6.49"N. 8.14"S, 28.37"N,
37.40"E 38.00"W 77.13"E
Enyi (1973) Mafra et at. (1979) Singh (1981)
R
77.42"E 3.56"E
Gangwar and Karla (1982) Agboola and Fayemi (1972) Kurtz et a/. (1952) IRRI (1974) Andrews (1972) Baker (1979)
Hart (1975)
Sorghum intercropped with Beans
111 P
Chickpea Greengram Pigeonpea
Osiru and Willey (1972) Enyi (1973) Mafra et 01. (1979) Enyi (1973) Mafra et 01. (1979) Singh (1981) Shackel and Hall (1984) Evans (1960) Wagmar and Singh (1984) Elmore and Jacobs (1984) Singh and Jain (1984) Wahua and Miller (1978) Singh (1981) Chowhury and Misangu (1981)
(Continued)
Table I (Continued)
Crop combinations
Climatic type after KoppenO
Irrigated or rain fedb
Country and location of study
Reference
Sorghum intercropped with (cont.) Cotton
Aw
11.1l0N,
7.38"W
Baker (1979)
Aw Aw
R R R
Nigeria, Samaru
Sorghum Millet
Nigeria, Samaru Nigeria, Samaru
11.11"N, 11.11"N,
7.38"W 7.38"W
Norman (1974) Andrews (1974)
Cowpea
Aw
R
Nigeria, Samaru
11.11 ON,
7.38"W
Andrews (1974)
Groundnut
Aw
R
India, Hyderabad
17.22"N,
78.26"E
Reddy and Willey (1981)
Greengram
Aw
R
Tanzania, Morogoro
6.49"S,
37.40"E
May (1982)
Cotton
Aw
R
Nigeria, Samaru
11.1l0N,
7.38"W
Cowpea
Af
R
Phillipines, Manila
14.37"N,
120.58"E
Harwood and Price (1976)
Mungbean
Af
R
Phillipines, Manila
14.37"N, 120.58"E
Harwood and Price (1976)
Peas
BS
R
Cyprus, Prastio
35.10"N.
33.45 "E
Hadjichristodoulou (1973)
Vetch Soybean
BS Ca
R R
Cyprus, Prastio U.S.A., Oklahoma
35.10°N, 35.28"N,
33.45 "E 97.33"W
Hadjichristodoulou (1973) Crabtree and Rupp (1980)
Lupins
Da
R
Greece, Thessaloniki
40.38"N,
22.58"E
Field beans
Da
R
U.K., Reading
51.28"N,
0.59"W
Millet intercropped with
Baker (1979)
Rice intercropped with
Wheat intercropped with
Barley intercropped with
'Koppen's classification as in Trewartha (1954). bR, rain-fed, I, irrigated.
Papadakis (1941) Martin and Snaydon (1982)
CEREAL-LEGUME INTERCROPPING SYSTEMS
41
crop (de Wit and van den Bergh, 1965; van den Bergh, 1968). The sum of the relative yields of component crops is called the Relative Yield Total and is denoted by RYT (de Wit and van den Bergh, 1965; Harper, 1977). When the RYT is equal to or less than 1, there is no advantage to intercropping. Although the calculation of RYT was originally based on the replacement series in competition studies, where proportions of the components in binary mixtures are varied but the overall crop densities remain constant (de Wit, 1960; de Wit and van den Bergh, 1965), the calculation can in fact be applied to any density situation in intercropping. However, Mead and Riley (1981) have argued against its use because the objectives of intercropping are essentially agronomic, that is to find the best ways of growing crops together. 2. Land Equivalent Ratio Willey and Osiru (1972) proposed the concept of the Land Equivalent Ratio (LER) as an index of combined yield for evaluating the effectiveness of all forms of intercropping. LER is defined as the total land area required under sole cropping to give the yields obtained in the intercropping mixture. It is expressed as: LER =
(Yjj/Y,i)
+ (Y,i/qj)
where Y is the yield per unit area, Yi and Yj are sole crop yields of the component crops i and j, and yij and Tiare intercrop yields (Mead and Willey, 1980). The partial LER values, Li and L j , represent the ratios of the yields of crops i and j when grown as intercrops, relative to sole crops. Thus, Li
=
(xj/Yi) and Lj = (Yi/qj)
LER is the sum of the two partial land equivalent ratios so that LER
=
L,
+ Lj
When LER = 1, there is no advantage to intercropping in comparison with sole cropping. When LER > 1, a larger area of land is needed to produce the same yield of sole crop of each component than with an intercropping mixture. For example, when LER = 1.25, 25% more land is needed to produce the same yield from the components as sole crops.
3. Area Time Equivalent Ratio The Area Time Equivalent Ratio (ATER) was proposed by Hiebsch (1980)
and McCollum (1982) as a modification of the LER. This takes into account
48
FRANCIS OFORI AND W. R. STERN
the duration of the crop, i.e., the time it occupies from planting to harvesting; it also permits an evaluation of crops on a yield-per-day basis (Hiebsch and McCollum, 1987). It does not appear to have been adopted widely. It is calculated as: ATER = (Liti
+ Ljtj)/T
where Li and Lj are relative yields or partial LERs of component crops i and j, ti and tj are the durations (days) for crops i and j , and T is the duration (days) of the whole intercrop system. 4. Staple Land Equivalent Ratio
In situations where the primary objective is to produce a fixed yield of one component (staple) crop, usually the cereal, and some yield of the legume, Reddy and Chetty (1984) proposed the concept of the Staple Land Equivalent Ratio (SLER) as an extension of the LER. It is based on the assumption of a basic requirement for minimum supply of a major staple crop such as the cereal. It is estimated as: SLER =
(q/qi) + Pij ( Y j i / Y j j )
where &/yiis “the desired standardized yield” of staple i, Pijis the proportion of land devoted to intercropping, and Yji/yjj is the relative yield of crop j. This index is peculiar to India and does not appear to have been used widely there either.
5.
Comparison of RYT, LER, ATER, and SLER
Examples are given in Table I1 of some RYT, LER, ATER, and SLER values calculated from two experiments that examined maize intercropped with either pigeonpea or soybean in replacement series (Dallal, 1974, 1977). For both intercrop systems, all four functions indicate higher biological efficiency in the maize-pigeonpea combination because components are of contrasting maturities. The lower efficiency of the maize-soybean combination, however, could be due to competitive factors because with similar maturities, the components would be exploiting the same environment. The values of RYT and LER were similar in both intercrop systems but as has already been mentioned, RYT is generally based on the replacement series (de Wit, 1960; de Wit and van den Bergh, 1965). Besides the RYT, both ATER and SLER are restricted to specific intercrop situations. ATER
Table I1 Comparison of Calculated Values of Four Indices for Evaluating Cereal-Legume Intercropping Efficiene Maize-pigeonpea (Dallal, 1974)
Crop duration (days)b Sole crop yield (kg/ha) Intercrop yield (kg/ha) Relative yield total (RYT) Land equivalent ratio (LER) Area time equivalent ratio (ATER) Staple land equivalent ratio (SLER)
Maize-soybean (Dallal, 1977)
Maize
Pigeonpea
Maize
Soybean
110 3130 2979
170 1871 1481
110 5353 5118
110 1634 517
1.74 1.74 1.41 1.35
‘Derived from the data of Dallal (1974, 1977). %ese values have been rounded off to the nearest 10.
1.28 1.28 1.28 1.28
50
FRANCIS OFORI AND W.
R. STERN
is only appropriate in systems with component crops of contrasting maturities such as 110-daymaize and 170-daypigeonpea (Dallal, 1974);when components are of similar growth durations, ATER values are similar to RYT and LER (Table 11) (Hiebsch and McCollum, 1987). SLER is only applicable when it is desired to attain a specificyield of a staple cereal crop and yield from the legume component is a bonus (Willey, 1979). SLER therefore cannot be used to evaluate intercropping efficiency in situations where it is desired to produce yield from equally acceptable component crops, and this is the most common goal of cereal-legume intercrop systems (Willey, 1979). From this, LER is considered to be the most appropriate general function to determine the efficiency of cereal-legume intercrop systems and could be applied to any form of intercropping. However, when the difference between growth durations of component crops is substantial, time becomes an important element and ATER is considered to be a more appropriate index of efficiency of the system. In this chapter, except where otherwise stated, intercropping efficiency or yield advantage due to intercropping will be estimated by the LER.
c. THELANDEQUIVALENT RATIO: ITSADVANTAGES AND DISADVANTAGES The Land Equivalent Ratio (LER) is the most frequently used index to determine the effectiveness of intercropping relative to growing crops separately (Willey, 1985). Generally, the value of LER is determined by several factors including density and competitive abilities of the component crops in the mixture, crop morphology and duration, and management variables that affect individual crop species (Enyi, 1973; Natarajan and Willey, 1980a; Fawusi et al., 1982). It has been suggested that in density studies of cereal-legume intercrop systems, the sole crop yields used as standardization factors for estimating LER should be at the optimum densities of the crops (IRRI, 1974; Huxley and Maingu, 1978). This avoids the confounding of beneficial interactions between components with a response to change in density (Trenbath, 1976). The values of LER follow the density of the legume component rather than that of the cereal (Ofori and Stern, 1987). Differences in competitive ability affect the relative performance of component crops and thus the LER values of different cereal-legume intercrop systems. In maize-bean combinations, Davis and Garcia (1983) found that yield reductions were related to plant height. Maize yields were reduced by 17% when intercropped with the most competitive bean cultivars (Group IVB, the aggressive climbers), by 14% with the intermediate cultivars (Group IVA), and by 8% with the weak climbing cultivars (Group IIIB). The LER values calculated using the mean yield of 114 cultivars of beans as a standardization factor were 1.72 for maize cultivar ICA intercropped with bean Group IVA,
CEREAL-LEGUME INTERCROPPING SYSTEMS
51
and 1.30 with bean Group IIIB. In another study, which compared three contrasting maize cultivars intercropped with beans, LER values for the short maize cultivar, La Posta, and the medium height cultivers, Suwan-1, were 1.04, and with the taller cultivar, ICA H-210, it was 0.94. Using sorghum and soybean, Wahua and Miller (1978a) obtained an LER value of 1.11 with tall and 1.09 with short sorghum, not a significant difference. Differences in growth durations of component crops affect the magnitude of the LER. The LER values in crops with similar maturities are usually less in crop combinations with contrasting maturities (Table 11) (Trenbath, 1976; Willey, 1979). Enyi (1973) studied maize or sorghum intercropped with either cowpea or beans of similar growth durations and found that productivity was less when compared to intercropping of these cereals with 240-day pigeonpea. The estimated partial LER of maize was 0.72 with pigeonpea, 0.64 with beans, and 0.50 with cowpea. With sorghum, a partial LER of 1.65 was obtained with pigeonpea, 1.01 with beans, and 0.93 with cowpea. The almost similar and lower partial LER values for the associations with beans and cowpea may be due to competition for growth-limiting factors, because peak demands on the environment by these crops might have coincided with those of the cereals. The higher partial LERs obtained with sorghum compared to maize indicates that it was offered less competition from the associated legumes and therefore was able to maintain higher plant yields (Andrews, 1974). The availability of water also appears to influence the LER. In maizebean (Fisher, 1977) and sorghum-cowpea intercrop systems (Mafra et al., 1981; Rees, 1986), LER values increased with the availability of water and diminished when water was limiting. However, Natarajan and Willey (1980b) found that LER increased under limited water situations. 1. Advantages for LER
As an index of combined yield, LER provides a quantitative evaluation of the yield advantage due to intercropping (Willey, 1979). Although component crops may give greatly different yields, the estimate of relative yields with sole crops at optimum or recommended densities as references gives comparable scales for both components, permitting comparisons of various crop combinations. LER could be used either as an index of biological efficiency to evaluate the effects of various agronomic variables (e.g., fertility levels, density and spacing, comparison of cultivar performance, relative time of sowing, and crop combinations) on an intercrop system in a locality or as an index of productivity across geographical locations to compare a variety of intercrop systems (Chetty and Reddy, 1984).
52
FRANCIS OFORI AND W. R. STERN
LER is identical to the RYT of de Wit and van den Bergh (1965), and can be used for any set of intercropping treatments (Willey, 1979). The partial LER values give an indication of the relative competitive abilities of the components of intercrop systems. In the intercropping mixture, the species with higher partial LER is considered to be more competitive for growth-limiting factors than the species with lower partial LER (Willey, 1979). Partial LER is more applicable to intercropping experiments than the relative crowding coefficient, k, used in measuring competitive ability in competition studies (de Wit, 1960; Hall, 1974a,b). 2. Disadvantages of LER
LER is based on land area only and does not take the duration of component crops into consideration. However, crop production is a function of both crop duration (time) and land area because land occupancy by a given intercrop system is frequently of longer duration than for sole crops. In these situations, the concept of Area Time Equivalent Ratio (ATER) developed by Hiebsch (1980) is appropriate; a detailed analysis of its applicability is given by Hiebsch and McCollum (1987). Several methods have been suggested in the literature for calculating LER using different sole crop values as standardization factors. These include averaging all the sole yields in each block or replication (Fisher, 1977); using the average sole crop yields in the entire experiment (Mead and Stern, 1980; Oyejola and Mead, 1982); using the average sole crop yield at each treatment level in studies that involve graded levels of a factor such as fertilizer or herbicide (Mead and Willey, 1980); and using the yield of the best sole crop treatment of each crop (Huxley and Maingu, 1978; Mead and Willey, 1980). The choice of sole crop yield for standardizing mixture yield in the estimation of LER is not clear and a generalization is not possible. The method used will depend on the aim of the experiment. As an index of biological efficiency, LER is based on harvested products and not on desired yield proportions of the component crops predetermined at sowing (Mead and Stern, 1980; Mead and Willey, 1980; Riley, 1984). This is overcome by the “effective LER” proposed by Mead and Stern (1980), and again later by Mead and Willey (1980) as an alternative to LER for evaluating the biological efficiency of a given required proportion of component crops in an intercrop system.
111. COMPETITIVE RELATIONSHIPS BETWEEN COMPONENT CROPS In plant populations, competition is defined as the situation in which each of two or more plants growing together in the same area seek the same
CEREAL-LEGUME INTERCROPPING SYSTEMS
53
growth factor, which is below their combined demands (Clements et al., 1929; Donald, 1963). Willey (1979) pointed out that the efficiency of production in cereal-legume intercrop systems could be improved by minimizing interspecific competition between the component crops for growth-limiting factors. Growing component crops with contrasting maturities so that they complement rather than compete for the same resources at the same time is one way of achieving this. It is substantiated by the yield advantages reported from various studies: 85-day bean and 120-day sorghum gave a 55% total yield increase (Willey and Osiru, 1972; Reddy et al., 1980) obtained a 31% yield advantage with 82-day millet and 105-day groundnut; and Natarajar) and Willey (1980a) found a 62% yield advantage with 82-day sorghum and 173-day pigeonpea. In contrast, lesser advantages have been reported in crop combinations in which interspecific competition is evident due to similar or almost overlapping growth durations. In this category, Wahua and Miller (1978a) obtained an 11Vo advantage with a sorghum-soybean combination; 11Yo was obtained with 112-day maize and 116-day soybean (Dallal, 1977); and 8% with a maize-cowpea system (Wanki et al., 1982). No yield advantages were found in maize-cowpea (Haizel, 1974) and sorghum-cowpea (Andrews, 1972; Rees, 1986) intercrop systems in which components were of similar growth durations. Competition between component crops for growth-limiting factors is regulated by basic morpho-physiological differences and agronomic factors such as the proportion of crops in the mixture, fertilizer applications, and relative time of sowing (Harper, 1961; Trenbath, 1976). Where component crops are arranged in defined rows, the degree of competition is determined by the relative growth rates, growth durations, and proximity of roots of the different crops. The cereal component, with relatively higher growth rate, height advantage, and a more extensive rooting system, is favored in the competition with the associated legume. The cereal is described as the dominant component and the legume as the dominated component (Huxley and Maingu, 1978). In considering the relative yields of cereals and legumes in intercropping systems, a survey of 40 published papers shows that the yield of the legume component declined, on average, by about 52% of the sole crop yield, whereas the cereal yield was reduced by only 11% (Table 111). Thus the general observation is that yields of legume components are significantly depressed by cereal components in intercropping. A. THELIGHTFACTOR The rate of dry matter production in crops depends on the efficiency of the interception of photosynthetically active radiation (PAR) (Biscoe and Gallagher, 1977; Monteith, 1977).
54
FRANCIS OFORl AND W. R. STERN
Table I11 Yields of Component Crops and Relative Yield Loss Due to Intercropping in Various Cereal-Legume Intercrop Systems Sole crop yield (kglha) Crop combination Maize-beans
Maize-cowpea
Maize-soybean
Maize-groundnut Maize-calopo Maize-greengram Maize-pigeonpea Sorghum-cowpea
Sorghum-soybean Sorghum-groundnut Sorghum-greengram Sorghum-pigeonpea
Sorghum-blackgram Sorghum-chickpea Millet-groundnut Millet-pigeonpea
Cereal
Legume
7320 7290 4126 5591 2080 1004 3860 6500 2678 7408 8455 9800 5353 8189 41 14 3467 7200 1987 8189 2080 2080 3170 2891 2579 3568 4860 3340 3680 4670 2794 8255 4467 3208 2853 3603 4325 2302 2407 2795 2354
1620 1958 1493 2986 1159 130
1176 2035 584 1500 3430 2100 1634 2677 1824 2290 3278 2441 1742 1159 I210 1195 997 584 676 2741 986 889 1430 704 2304 1017
Cereal
7 10
8 30 +I1 31 + 45 33 25 18 12 46 9 13
14 4 2
Legume
Fisher (1977) Cordero and McCollum (1979) Francis et a/. (1982b) Davis and Garcia (1983)
40
Agboola and Fayemi (1971) Haizel (1974) Remison (1978) Wanki et a/. (1982) Faris et a/. (1983) Ofori and Stern (1986)
83 12 49 41 46 59 44 80 80 46 72 87
1
18
74
3
16 +5
17 6
1446
11
14 9 0 +2
52 48 7
Reference
60 69 39 74
3 +4 +9 4 31 35 +2 47 4
1380 1084
852 849 2463 1235 1244
“Percentage of sole crop yield.
Yield loss due to intercropping ( V O ) ~
43 48 68 72 33 40
Betse (1976) Beets (1977) Dallal (1977) Searle et a/. (1981) Ahmed and Rao (1982) Chetty and Reddy (1984) Chui and Shibles (1984) Baker (1978) Searle el a/. (1981) Agboola and Fayemi (1971) Agboola and Fayemi (1971) Yadav (1982) Andrews (1972) Faris et a/. (1983) Singh and Jain (1984)
42 47 65 77 44 39 32 43 40 36
Wahua and Miller (1978a) Singh and Jain (1984)
60 87 17 73 22
Singh and Jain (1984)
Singh and Jain (1984) Chetty and Reddy (1984) Singh and Jain (1984) Freyman and Venkateswarlu (1977) Natarajan and Willey (1980a) Rao and Willey (1980) Rego (1981) Singh and Jain (1984) Chowdhury and Misangu (1981) Reddy et a/. (1980) Willey and Reddy (1981) Rao and Willey (1983)
CEREAL-LEGUME INTERCROPPING SYSTEMS
55
1. Light Interception
The amount of light intercepted by the component crops in an intercrop system depends on the geometry of the crops and foliage architecture (Trenbath, 1982; Tsay, 1985). The generally taller cereal shades the legume, and at high densities causes reduced growth and yield of the companion legume. Gardiner and Cracker (1981), maintaining a constant bean density of 220,000 plantdha, found that varying maize density from 18,000 to 55,000 plantdha progressively reduced the amount of light available to the beans. At the low maize density (18,000 plantdha), bean received 50% of the incident light, compared to 20% at the highest maize density (55,000 plantdha). At the highest maize density, yield of the intercrop bean was only 30% that of the sole bean. Several studies at the International Crops Research Institute for the SemiArid Tropics (ICRISAT), India, have measured percentage light intercepted using short-duration cereal (e.g., maize, sorghum, or millet) and longduration legume (e.g., pigeonpea or groundnut) both as sole crops and as intercropping mixtures (Reddy and Willey, 1981; Sivakumar and Virmani, 1980, 1984; Marshall and Willey, 1983). In a sorghum-pigeonpea combination, the amount of light intercepted relative to incoming incident radiation at 55 days was 84% in sole sorghum, 65% in sole pigeonpea, and 80% in the intercropping mixture (Natarajan and Willey, 1980b). The data of Sivakumar and Virmani (1980) from a maize-pigeonpea intercrop system shows that light interception was low with the initial slow increase in leaf area index (LAI), and above 80% when LA1 reached about 3. Although light interception in the sole crops and the intercrops were almost similar, the foliage canopy of the intercrop was more effective in capturing the light. The intercrop system attained an LA1 of 3 in 45 days, compared to 50 days in the sole maize and 115 days in the sole pigeonpea. Although the ICRISAT studies do not report the quantity of light reaching the top of the associated legume canopy, the light intercepted by the similarly spaced sole cereal crops suggests significant reductions in light available to the companion legume. Assuming a peak LA1 of 2 for the intercrop cereal in these studies, the amount of light incident on the companion legume canopy would be 30% or less of the total incoming radiation. The reduction in light reaching the legume canopy when intercropped with a taller component crop has been clearly demonstrated in a cassava-soybean intercrop system in southern Queensland by Tsay (1985). He found that PAR transmission was closely related to the distribution of leaf area in the cassava canopy, and isopleths of PAR relative to full sunlight decreased towards the top of the legume canopy (Fig. 1). When intercropped soybean rows were 45-90 cm from cassava rows with an LA1 of about
56
FRANCIS OFORI AND W. R. STERN
3, the amount of light reaching the soybean canopy was about 25% of the total incoming radiation. 2. Efficiency of Conversion of PAR into Dry Matter
The higher productivity of intercrop systems compared to the sole crops may be attributed to better light utilization by a crop canopy composed of plants with different foliage distributions (Willey and Roberts, 1976; Willey, 1979). In a study of maize and pigeonpea, Sivakumar and Virmani (1980) found that dry matter production per unit of PAR absorbed was higher in the mixture than in the sole crops (Table IV). The higher PAR conversion efficiencies of these systems relative to the sole crops may be due to spread of light over greater leaf area, and more efficient distribution of light in the intercrop canopies during early stages of growth. For example, using a pearl millet-groundnut system, Reddy and Willey (1981) obtained energy conversion efficiencies at 68 days of 1.70 g/MJ in sole millet, 1.07 g/MJ in sole groundnut, and 1.95 g/MJ in the mixture. At maximum green leaf area index (61 days after sowing), Marshall and Willey (1983) observed energy conversion efficiencies of 4.1 g/MJ in sole millet, 2.5 g/MJ in sole groundnut, and 4.3 g/MJ in the intercropping mixture. The similarity in efficiency of dry matter production in the sole millet and the mixture was attributed to the greater proportion (60%) of intercrop millet in the total intercrop. 90cm x 90cm
180cm x 45cm
. = 0.15
LA1
LA1 = 3.69
LA1 = 2.8 1
3535d . . ... ..
I: - I.:
.-
.:
I.:.
I
I . . .
..
:.
,
I
-
I"
0
45
0
1
0
I
I
1
45
90
45
J
0
Distance from cassava row (cm)
FIG. 1. Isopleths of relative percentages of available radiation (75%, m-m; A-A;
50%,
of full sunlight) within intercrop cassava canopies in the density of 1.23 plants/m* with interrow spacings of 90 and 180 cm at 98 days after cassava planting. (From Tsay, 1985.) 25%, 0-
0
57
CEREAL-LEGUME INTERCROPPING SYSTEMS
Table IV Total Dry Matter, Grain Yield, Leaf Area Index, Light Interception (@lo), and Efficiency of Dry Matter Production of a Maize-Pigeonpea Intercrop System"
Cropping system
Total dry matter yield (kg/ha)
Grain yield (kglha)
Peak LA1
Sole maize Sole pigeonpea Intercrop total
8,130 7,870 15,290
3,500 1,833 5,038
3.5 3.2 6.1
Light interception at peak LA1
VO)
Dry matter production efficiency WMJ)
90 85 83
3.1 1.2 4.3
%om data of Sivakumar and Virmani (1980).
B. THESIGNIFICANCE O F WATER Water is a most important soil factor in semiarid and subtropical regions, where intercropping is extensively practiced in dryland farming systems and inadequate rainfall may frequently limit crop production (Baker and Norman, 1975). The differences in root systems, depth of rooting, lateral root spread, and root density are factors that affect competition for water between component crops (Babalola, 1980; Haynes, 1980). The use of different parts of the soil profile by root systems of different crop species minimizes the degree of competition for water (Haynes, 1980). When component crops compete for available water, the cereal, with its higher growth rate and more extensive root system, is generally favored. In India, Natarajan and Willey (1980b) found that the amount of water transpired in a sorghum-pigeonpea intercrop system depended on growth durations of the crops. The total water use by sole pigeonpea at the end of the growing period (173 days after sowing) was 584 mm and in the mixture 585 mm; in sole sorghum at the final harvest (82 days after sowing) it was 434 mm. Using 82-day millet intercropped with 105-day groundnut, Reddy and Willey (1981) obtained a total water use of 406 mm in the mixture, compared to 303 mm in sole millet and 368 mm in sole groundnut. Shackel and Hall (1984) in the United States, studying plant water deficits in sorghum intercropped with cowpea in terms of xylem pressure potential, and osmotic potential under irrigated and water-limited conditions, found that they were not substantially affected by intercropping. However, shading by the associated sorghum slightly increased midday xylem pressure potential and osmotic potential of intercropped cowpea leaves. The patterns of cumulative water depletion from the top 195 cm of soil in the intercropped treatments were similar to values obtained in the sole crops. Sorghum
58
FRANCIS OFORI AND W. R. STERN
and cowpea appeared to be competing equally for soil water as evidenced by absence of substantial effects of intercropping on the water relation of these crops. From the humid tropics of Nigeria, Hulugalle and Lal (1986) reported that water use efficiency (WUE) in maize-cowpea intercrop was higher than in the sole crops when soil water was not limiting; however, under drought conditions, WUE in the intercrop was lower compared to the sole maize. For the favorable moisture regimes, WUE (kg grain per mm per ha) of the intercrop (alternate row arrangement) was 3.6 compared to 2.1 in either of the sole crops, and for the droughty conditions, 1.6 for the intercrop, 2.2 for the sole maize, and 0.5 for the sole cowpea. From the above, it may be concluded that cereal and legume intercrops use water equally, and that competition for water may not be an important factor in determining the efficiency of intercrop systems, except under favorable soil moisture conditions.
c.
THEROLEOF SOIL NUTRIENTS
The major soil nutrients for which component crops compete when in limited supply are nitrogen, phosphorus, and potassium. The cereal component, with a faster-growing or more extensive root system, generally has a competitive advantage over the associated legume (Trenbath, 1976). The inability of the legume to compete for these nutrients are attributed to lesser ramification of their root systems (Rabotnov, 1977). Competition for nutrients is important and could begin early in the growth of the component crops in cereal-legume intercropping systems (Wahua, 1983).
I. Nitrogen The mobility and high demand for nitrogen by most crops, particularly nonlegume, leads to severe competition for it in nonlegume-legume associations (Allison, 1973; Beets, 1982). In these associations, Evans (1977) pointed out that the absorption of nitrogen is controlled by the roots of component crops. In cereal-legume intercropping, the legume component is capable of fixing atmospheric N,under favorable conditions and this is thought to reduce competition for N with the cereal component (Trenbath, 1976). In the absence of an effective N-fixing system, both cereal and intercrop legume compete for available soil N (Ofori et al., 1987). Ibrahim and Kabesh (1971) found that horsebean reduced N uptake of wheat by 26% whereas wheat reduced that of horsebean by 44%. It is apparent that dry matter yield and N uptake of the component crops were
CEREAL-LEGUME INTERCROPPING SYSTEMS
59
severely affected in the mixed stands, as shown by the partial LER values:
Dry matter yield N uptake
Wheat
Horsebean
0.63 0.74
0.63 0.56
In a maize-cowpea intercrop system, Wahua (1983) found that at 105 kg N/ha, the crops were in competition for N and that this occurred before anthesis or flowering. The severity of the competition for N was greater for cowpea and occurred at 40 days and was evident in the associated maize 10 days later. Nitrogen uptake by intercrop cowpea was 64 kg/ha compared to 88 kg/ha in the sole cowpea. Nitrogen uptake of intercrop maize was reduced by 17% compared to sole maize. Without applied N, Chang and Shibles (1985a) and Ofori and Stern (1986) reported strong competition for soil N by intercrop maize and cowpea. This was particularly evident between 49 and 63 days when both crop species were at the reproductive stage and required substantial amounts of N. 2. Phosphorus Phosphorus is a major nutrient that determines the production potential of most grain legumes usually intercropped with cereals (Williams, 1936). Legumes are poorer competitors for P when intercropped with grasses or cereals, this being attributed to differences in root morphology (Donald, 1963; Jackman and Mouat, 1972; Evans, 1977). Lai and Lawton (1962) evaluated root competition for P between corn and intercrop field because using 32P-labeledfertilizer placed at different depths. They found that corn was more vigorous in the uptake of P than beans as a result of its more extensive roots. Using a replacement series design, Dallal (1974) observed that intercropped maize and pigeonpea were competing for P from 28 days onward right up to the maize final harvest (112 days after sowing) and that maize was more competitive. At 112 days, intercropping reduced dry matter yield of maize by 32% and pigeonpea by 66% compared to the sole crops; total P uptake was reduced by 25% in maize and 70% in pigeonpea. Wahua (1983) found that maize and cowpea were competing for P and this was evident at anthesis or flowering (Table V). In the absence of applied P, maize was more competitive than cowpea in the initial stages. However, at high rates of applied P, P uptake of intercrop maize was reduced by 30%, indicating competition for P from cowpea. Competition was clearly expressed in the observation that intercrop cowpea took up only 50% of the sole cowpea P uptake in the absence of applied P, while at a high level of P,
60
FRANCIS OFORI AND W. R. STERN Table V Yield and P Uptake of Maize Intercropped with Pigeonpea or Cowpea ~~~~
P uptake (kg/ha)
Dry matter yield (kg/ha) Source of reference and cropping system
28 DAY
112 DAS"
28 DAY
112 DAY ~
Dallal (1974) Sole maize Intercrop maize Sole pigeonpea Intercrop pigeonpea
880 449 197 44
6408 4361 822 28 1
3.6 2.0 1.1 0.2
13.2 10.0 1 .o 0.3
Grain yield (kg/ha) P rate (kg/ha)
Remision (1978) Sole maize Intercrop maize Sole cowpea Intercrop cowpea
3243 5591 1254 1060
3544 5042 1206 1138
P uptake at 50 DAS P rate (kg/ha)
Wahua (1983) Sole maize Intercrop maize Sole cowpea Intercrop cowpea
3.5 4.6 1.4 0.7
5 .O 3.5 1.7 1.1
uDAS, Days after sowing.
65% was taken up. Remison (1978) concluded, however, from grain yield data that intercropped maize and cowpea grown at two levels of P did not compete for P, because there were no significant differences in yields of the sole crops and the intercrops (Table V).
3. Potassium Studies by Drake et al. (1951) showed that cation exchange capacities (CECs) of roots of legumes are approximately double those of cereals. The relatively high CEC of legumes indicates that on soils with low levels
CEREAL-LEGUME INTERCROPPING SYSTEMS
61
of exchangeable K, the legume would be deficient in K because larger amounts of divalent cations would be adsorbed by the roots. The level of K in many soils decreases as the growing season progresses; consequently K uptake in competition with cereal becomes increasingly difficult for the legume (Drake et al., 1951). Dallal (1 974) found that intercropped maize and pigeonpea were competing for K at different stages of growth: Reduction of K uptake due to companion crop (Olo) 42 days
Maize Pigeonpea
112 days
33
16
52
63
In an 82-day sorghum and 173-day pigeonpea intercrop system, Natarajan and Willey (1980b) found sorghum to be more aggressive for K than pigeonpea, and this severely affected the early growth of pigeonpea. At 74 days, intercropping markedly affected K uptake of sorghum; the sole sorghum absorbed 45 kg/ha of K and the intercrop 165 kg/ha. In pigeonpea, K uptake was 28.6 kg/ha in the sole crop and 3 kg/ha in the intercrop, a reduction of 87.5070. In a maize-cowpea combination, Wahua (1983) found maize to be more competitive for K than cowpea, particularly when N was high. At 50 days, application of 115 kg/ha of N caused reductions of 3 1070 in uptake of K in the intercrop maize, and 50% in the intercrop cowpea, compared to the respective sole crops.
IV.
SOME AGRONOMIC FACTORS INFLUENCING PRODUCTIVITY AND EFFICIENCY
The productivity and efficiency of cereal-legume intercrop systems are affected by various agronomic variables that affect crop yields. In this section, the influence of variables such as component crop density, plant spacing and arrangement, relative time of sowing of component crops, and the effect of applied nitrogen will be discussed. A.
COMPONENT CROP DENSITY
The overall mixture densities and the relative proportions of component crops are important in determining yields and production efficiency of cereal-legume intercrop systems (Willey and Osiru, 1972; Lakhani, 1976). When the components are present in approximately equal numbers, productivity and efficiency appear to be determined by the more aggressive
62
FRANCIS OFORI AND W. R. STERN
crop, usually the cereal (Osiru and Willey, 1972;Lakhani, 1976). This can be illustrated with 5O:SO mixture data from a replacement series experiment in a sorghum-bean intercrop (Osiru and Willey, 1972). The optimal density of either crop of 200,000plants/ha gave an LER of 1.41;this was due to increase in intercrop sorghum yield, because the intercrop bean yield remained unchanged. In a study of maize intercropped with cowpea at densities ranging between 10,000and 40,000plants/ha for either crop and planted in the same hill, maize (cv. TZPB composite) was more competitive than cowpea (cv. TVu-1209)(Fawusi et af., 1982). The response of intercrop maize to increasing component density was similar to that of sole maize. At the lowest mixture density, the intercrop maize yield was 2300 kg/ha, 15% less than the sole maize, and increased to 4600 kg/ha (8% less than sole maize) at 40,000 plants/ha. Pod yield of intercrop cowpea with the lowest density of maize was 941 kg/ha, a reduction of 41(70 of sole cowpea yield at optimum density. At the highest overall density, intercrop cowpea yield was 700 kg/ha, i.e., a yield reduction of 66%. When sole crop yields of maize and cowpea in the optimum density treatment were used to standardize mixture yields, LER values rose with increasing mixture density. From the lowest to the highest density, the LER values were 0.91, 1.14, 1.20,and 1.26. The growth and yield of legume component is reduced markedly when intercropped with high densities of the cereal component. In a maize-bean intercrop system, increasing maize density three-fold, from 18,000to 55,000 plant/ha, caused reductions of 24% in leaf area index and 70% in seed yield of the associated bean (Gardiner and Craker, 1981). Using repacement series designs in a maize-cowpea intercrop system, Chang and Shibles (1985b)showed that the level of the maize population generally imposed a limit on the yield of the intercrop cowpea, and that there was no effect of increasing cowpea density. Even though the cereal component usually contributes a greater proportion of the mixture yield, the magnitude of intercropping advantage or efficiency seems to be determined by the legume component (Ofori and Stern, 1986, 1987). Fisher (1977)studied maize-bean intercrop systems at varying densities which at harvesting were 13,700,27,000,and 47,700plants/ha of maize combined with 23,300, 56,300, and 121,000 plants/ha of beans, respectively, designated as low, medium, and high densities. At each density, the yields of intercrop maize did not differ from those of the sole maize. However, intercrop bean yield significantly increased with a rise in bean density. The seed yields of beans were 320 kg/ha, 650 kg/ha, and 940 kg/ha from the lowest to the highest density. Although maize contributed more than 80% of the mixture yield at each density, the LER values followed the trends in intercrop bean yields: 0.57,0.92,and 1.39 from the lowest to the highest density. These observations are consistent with the data of Francis
CEREAL-LEGUME INTERCROPPING SYSTEMS
63
et al. (1982a,b) on maize intercropped with bush or climbing beans at different component crop densities. Results from two separate studies using sorghum and pigeonpea also show that the cereal contributes a greater proportion to mixture yield, but that the efficiency of the system measured in terms of LER follows the trends in the intercrop legume yields (Freyman and Venkateswarlu, 1977; Natarajan and Willey, 1980a). When intercrop sorghum densities were varied from 55,000 to 220,000 plants/ha and combined with constant pigeonpea density of 37,000 plantdha, the intercrop sorghum yield reponse was linear. In contrast, the intercrop pigeonpea yield decreased with rising sorghum density (Freyman and Venkateswarlu, 1977). The highest LER value was obtained at the lowest sorghum density and decreased with rising sorghum density (Table VI). It seems that density of the cereal component determines the level of combined mixture yield, but that the efficiency of cereal-legume intercropping systems, measured in terms of LER, follows the trend of the legume component. B. PLANT ARRANGEMENT AND SPACING Row arrangements, in contrast to arrangements of component crops within rows, improve the amount of light transmitted to the lower legume canopy. Such arrangements can enhance legume yields and efficiency in cereal-legume intercrop systems (Mohta and De, 1980). In a maize-groundnut intercrop system, Evans (1960) obtained LERs of 1.09 in the same row arrangement compared to 1.30 in alternate rows and 1.38 in a hill arrangement (Table VI). In two separate studies involving pigeonpea and soybean, maize yield was not significantly reduced when planted in alternate rows with the legume in comparison to planting in the same rows (Dallal, 1974, 1977). From Table VI, it may be seen that biological efficiency measured as LER was higher when component crops were sown in alternate rows rather than in the same row. In the maize-pigeonpea system, maize yield was not affected in the alternate row arrangement, but this was reduced by 20% when the pigeonpea was in the same row (Dallal, 1974). From these studies, it was found that arrangement of component crops in alternate rows is more beneficial than in the same rows. In contrast to these observations, Agboola and Fayemi (1971) did not observe any difference whether maize and cowpea were planted in the same or alternate rows (Table VI). The use of double rather than single alternate row arrangements of component crops improve the yield and light penetration to the canopy of the
Table VI Effects of Plant Arrangement and Spacing of Component Crops on Productivity and Efficiency of Various Cereal-Legume Intercrop Systems Yield’ (kg/ha) Crop combination and treatment Maize with Groundnut Same row Alternate row In hills (Sole crop) Cowpea Same row Alternate row (Sole crop) Pigeonpea Same row Alternate row (Sole crop) Soybean Same row Alternate row Alternate 2-rows (Sole crop) Soybean Alternate rows Alternate 2-rows (Sole crop) Soybeanb (alternate rows) 70 x 30cm 70 x 60cm 70 x 90cm (Sole crop) Sorghum with Pigeonpea Same row (60 cm)‘ Alternate 2-rows (90 Alternate 2-rows (135 cm)‘ (Sole crop) Soybean Alternate rows Alternate 2-rows (Sole crop) Groundnut Alternate 2-rows (60 Alternate 2-rows (90 Cowpea Alternate 2-rows (60 Alternate 2-rows (90
Intercrop cereal
Intercrop legume
2,851 2,554 2,580 (3,330)
74 1 744 851 (1,422)
LER
Reference
1.09 1.30 1.38
Evans (1960)
593
1.54
Agboola and Fayemi (1971)
590
1.55
1,915 1,950 (L990)
(1.033)
2,4% 2,979 (3,130)
1,239 1,481 (1,871)
1.46
4,480 5,118 5,051 (5,353)
170 517 303 (1,634)
0.94 I .27 1.13
Dallal (1977)
3,530 3,700 (3,344)
550 720 (2,783)
1.25 1.37
Mohta and De (1980)
6,980 7,095 6,910 (7Jm
460 650 820 (3,405)
1.05 1.12 1.14
Chui and Shibles (1984)
8,380
1,040
1.24
Freyman and Venkateswarlu (1977)
8,840
1,070
1.29
9,110
1,150
1.35
(10,483)
(1,634)
1,540 1,500 (1,760)
1,247 (2,783)
990
Dallal (1974)
1.74
1.23 1.30
Mohta and De (1980)
Singh (1981)
3,055
323
I .47
3,185
386
1.52
3,605
418
1.87
3,635
716
2.19
Singh (1981)
‘Figures in parentheses are sole crop yields for calculating LER. bData of 1981 experiment. ‘Denotes cereal row spacing.
CEREAL-LEGUME INTERCROPPING SYSTEMS
65
legume component. In maize-soybean and sorghum-soybean studies conducted by Mohta and De (1980), the yields of the cereals were not affected by intercropping with soybeans when arranged in either single or double alternate rows. In the maize-soybean combination, there was a 3 1070yield increase in the intercropped soybeans when components were arranged in double alternate rows relative to single alternate rows. The LER was 1.37 in the double alternate row and 1.25 in the single alternate row arrangement. With sorghum, intercrop soybean yield in the double alternate rows was 1247 kg/ha, an increase of 26% over the single alternate row arrangement. The LER was 1.30 in the double alternate row arrangement and 1.23 in the single alternate row (Table VI). When alternating pairs of sorghum rows 90 cm apart with two rows of an associated legume, Singh (1981) found that LER was greater compared to sorghum at 60 cm between rows with two rows of the legume in between (Table VI). In a parallel study, Wagmare et al. (1982) found that light penetration was markedly increased in the wider row arrangement. At 60 days, light incident on the intercrop legume canopy with sorghum rows spaced at 60 cm was 52% of the incoming radiation, and 70% when sorghum was spaced at 90 cm. Widening interrow spacing of the cereal component to accommodate more rows of the legume component improves legume yield and efficiency of the intercrop system (Table VI). In a study of sorghum at 220,000 plantslha intercropped with pigeonpea at 37,000 plants/ha, 90-cm interrow spacing of sorghum improved the yield of intercrop sorghum by 5% and that of associated pigeonpea by 3% when compared to 60cm rows (Freyman and Venkateswarlu, 1977). Further widening of sorghum interrows to 135 cm gave a 9% increase in sorghum yield over the 60-cm spacing, and pigeonpea yield was increased by 11070. The LER values for the three spacingswere 1.20 for 60cm, 1.29 for 90cm, and 1.35 for 135 cm. The yield of intercrop soybean increased by 78% when spacing between hills with associated maize was adjusted from 30 to 90cm; the LER for the 30-cm spacing was 1.05 and 1.14 for the 90 cm (Chui and Shibles, 1984).
From several studies, it would appear that the yield of the cereal component is usually less affected by component densities and manipulation of spacing between component crops. Intercrop legume yield usually is reduced significantly, however, depending on the proximity of the cereal component. This could be due to the intensity of shading at the top of the legume canopy, as observed recently in a cassava-soybean combination by Tsay (1985) (see Fig. 1).
c.
RELATIVE TIMEOF SOWING OF COMPONENT CROPS
The relative time of sowing of component crops is an important management variable manipulated in cereal-legume intercrop systems but has not been extensively studied. Andrews (1972) pointed out that differential sowing improves productivity and minimizes competition for growth-limiting
66
FRANCIS OFORI AND W. R. STERN
factors in intercropping. Willey (1979)also pointed out that sowing component crops at different times ensures full utilization of growth factors because crops occupy the land throughout the growing season. Francis et al. (1976)found that sowing maize and beans 5-15 days apart reduced yields of the intercrops compared to sole crops (Table VII). In contrast to simultaneous sowing, maize sown 5-15 days earlier than beans increased maize yields by 13-43'70,and the associated bean yields were rcduced by 20-27%. On average, intercropping efficiency measured as LER was 39% higher when beans were sown 5-15 days before maize. In studies on maize intercropped with four contrasting bean cultivars sown 5-10 days apart, results suggest that near-simultaneous sowing of component crops is optimal to attain the highest combined yields and intercropping efficiencies (Francis, 1978;Francis et al., 1982a). In Columbia, Francis et al. (1982b)varied dates of sowing maize and indeterminate beans (types I1 and 111) and found maize to be more competitive than beans at all sowing dates, except when beans were sown 10 days earlier. Simultaneous sowing resulted in a bean yield reduction of 51 070 and maize yield reduction of 3 1 Yo. Sowing maize 10 days before beans reduced bean yield by 69% and maize by only 7%. Beans sown 10 days earlier reduced maize yield by 53% and bean yield by 21 %. In terms of LER, similar results were obtained by staggered and simultaneous sowings. However, the LER followed the trends in bean yields rather than the maize yield, because sowing maize 5 and 10 days before beans gave relatively lower LER values
Table VII Yield and LER of Component Cereal and Legume Sown Less Than 1 Week Apart Yield (kg/ha) Sowing dates
Cereal
Legume
LER
Reference
Maize-bean combination Sole crop Maize, 5-15 days before beans Simultaneous sowing Beans, 5-15 days before maize
7270 5040 5710 69106
939 394 500 483
-
Francis et at. (1976)O
1.11 1.32 1.46
Millet-greengram combination Sole crop Millet, 7 days before greengram Simultaneous sowing Greengram, 7 days before millet
2288 2115 1825 1528
349 26 87 229
1 .o 1.05 1.32
-
May (1982)
OBean variety Pijoo planted at 300,000 plants/ha and maize H-207 at 40,OOOplantslha. *Seems high but verified against the original data.
67
CEREAL-LEGUME INTERCROPPING SYSTEMS
of 1.30 and 1.23 whereas earlier sowing of beans gave LER values of 1.36 and 1.27, respectively. In Nigeria, Remison (1982) did not find any advantage of staggered over simultaneous sowing of maize and cowpea. Intercropping reduced cowpea yield by 57% and maize yield by 35% when sown simultaneously. Sowing cowpea 14 days earlier than maize gave a LER value of 0.88 whereas maize sown 14 days earlier gave a LER value of 1.20. LER values were higher when maize was sown earlier than cowpea but results of this study do not demonstrate consistent LER trends relative to yields of either of the component crops. In Western Australia, Ofori and Stern (1987) concluded from a maize-cowpea intercrop system that staggered sowing of component crops at intervals of 10 or 21 days were of no advantage over sowing them simultaneously. However, they found that the LER followed the trends in cowpea yield rather than maize (Fig. 2) Cowpea sowing relative to maize (days)
'2 1 10
-
w
I,.,
c
ooc -
'10 I
I
0
-10
I
I
-2 1 I
Sole maize
-0-
Intercrop maize
eooa
Y Q) v
.-*
'
600C
D 0)
4000
C
2000
-A-
I
.
J 1.0
Sole cowpea
'
-/intercrop cowpea -2 1
-10
0
'10
+21
Maize sowing relative to cowpea (days]
FIG. 2. Seed yields of maize and cowpea and LER of the intercrop system when component crops are sown differentially. Sole crops were established only at simultaneous sowing, i.e., 0 days. I, Least significant difference (LSD) @ < 0.05) to compare sole crops and intercrops of maize (M) and cowpea (C). (From Ofori and Stern, 1987.)
68
FRANCIS OFORI AND W.R. STERN
Manipulating the time intervals between growth durations of component crops influences efficiency of cereal-legume intercrop systems. In an 85-day bean and 120-day maize combination, a yield advantage of 20% was removed by sowing beans 28 days after maize (Osiru and Willey, 1976). In another study, May (1982) found that a yield advantage of 32% completely disappeared when greengram was sown one week after bulrush millet (Table VII). From these studies, it may be concluded that staggered or differential sowing of component crops is of no advantage over simultaneous sowing. In staggered sowing, the earlier sown component has an initial advantage over the later sown component. Component crops are unable to compensate fully for yield loss due to earlier or later association with the other component. However, the efficiencies of cereal-legume intercrop systems measured in terms of LER follow trends in the legume yields. It also appears that light is an important factor in determining the magnitude of the LER. D. EFFECTOF APPLIED NITROGEN The responses to applied nitrogen of component crops differ in different cereal-legume combinations 'thus affecting the efficiencies of various intercrop systems (Ahmed and Rao, 1982; Boucher, 1986). I . Maize as the Cereal Component Most studies on the effects of applied N on intercropping systems are based on maize, and these indicate similar grain yield responses to applied N for sole and intercrop maize (Dallal, 1977; Rao et al., 1979; Searle et al., 1981; Chui and Shibles, 1984; Ofori and Stern, 1986). In a maize-soybean intercrop system, Dallal(l977) reported that intercropping drastically reduced the seed yield of soybean, and seed yield showed no response to applied nitrogen (Table VIII). The yields of intercrop soybeans relative to the sole crops were reduced by 80% either with or without applied nitrogen. The efficiency of intercropping measured by the LER was 1.15 without applied nitrogen and 1.09 with 100 kg N/ha. In a maize-groundnut combination with N applied at 0, 25, 50, and 100 kg/ha, Searle et al. (1981) found significant yield reductions ranging from 70 to 88% for intercrop groundnut compared to the sole groundnut. The LER values were 1.36 at nil N, 1.20 at 50 kg N/ha, and 1.24 at 100 kg N/ha (Table VIII). From the United States, Chui and Shibles (1984) reported that maize grain yield responded to applied nitrogen at 135 kg N/ha both as a sole crop and as an intercrop of soybean. Seed yield of the sole soybean at 135 kg N/ha was
69
CEREAL-LEGUME INTERCROPPING SYSTEMS Table VIII Yields (kg/ha) and LER of Various Cereal-Legume Intercrop Systems as Influenced by Applied Nitrogen N rate (To of maximum applied) Source of data and treatments Dallal (1977) Maize Sole Intercrop Soybean Sole Intercrop LER Searle et al. (1981) Maize Sole Intercrop Groundnut Sole Intercrop LER Rego (1981) Sorghum Sole Intercrop Pigeonpea Sole Intercrop LER Chui and Shibles (1984)" Maize Sole Intercrop Soybean Sole Intercrop LER
0
50
5082 4666
-
1478 334 1.15
-
6680 7080 1741 530 1.36
930 1240 1390 800 1.91
6160 6120 3400 570 1.16
-
-
8604 8237
418 1.20
2890 2410
820 1.42
-
100
5623 5099 1789 326 1.09
8941 9178
378 1.24
4590 3710
870 1.43
-
9160 8550
-
3410 420 1.05
"Data for 1981 experiment.
13% more than the yield obtained without nitrogen, but intercrop soybean yield was 585 kg/ha compared to 730 kg/ha without N (Table VIII). The LER value at nil N was 1.28 and 1.13 at 135 kg N/ha. A multilocation appraisal of various maize-legume intercrop systems was conducted under the 5-year INPUTS (Increased Productivity Under Tight
70
FRANCIS OFORI AND W.R. STERN
Supplies) program coordinated in Hawaii, with locations in the tropics and temperate regions (Ahmed et al., 1979). Maize-cowpea and maize-groundnut systems were studied at three locations, maize-mungbean in four locations, and maize-soybeans in seven locations. Levels of applied N in the experimental treatments at each location were specified as 0,25,50, and 100% of the local recommended rate. Table IX shows fitted regression models for sole and intercrop maize grain yields and the intercrop legume seed yields, averaged over all locations for each crop combination in response to applied N. For the maize-groundnut system, only the data from Australia are shown because data from the other locations were incomplete. From these studies, maize grain yield responses to applied nitrogen were positive and behaved similarly to those of cowpea, groundnut, mungbean, or soybean (Fig. 3). However, in the sole maize, there was a lack of response to applied N in the
Table IX Fitted Regression Equations for Component Crop Yields in Response to Applied Nitrogen in Various Maize-Legume Intercrop Systems' Cropping system
Number of locations
Maize-cowpea Sole maize Intercrop maize Intercrop cowpea
3
Maize-groundnutc Sole maize Intercrop maize Intercrop groundnut
1
Maize-mungbean Sole maize Intercrop maize Intercrop mungbean
4
Maize-soybean Sole maize Intercrop maize Intercrop soybean
7
Equationb
R'
y = 2066 y = 1605 y = 1499
+ 29.6~
y = 6830 y = 6964 y = 526
+ 6 2 . 4 ~- 0 . 4 2 ~ ' + 22.3~ + 3 . 2 ~+ 0.02~'
0.91 0.96 0.98
y
= 2257 y = 2044 y = 418
+ + +
19.2~ 15.2~ 4 . 7 ~- 0.04~'
0.95 0.83 0.99
y = 3019 y = 2835 y = 954
+ + +
1806X 11.7~ 5.3x - 0.04xI
0.88 0.96 0.90
+ 22.9~
+
6 . b - 0.04~~
'From the data of Rao eta/. (1979), data averaged over all locations for each system. by is grain or seed yield (kglha); x is N as percentage of recommended dose in kg/ha. 'Data from Australia only.
0.91 0.89 0.99
71
CEREAL-LEGUME INTERCROPPING SYSTEMS
L 0
I
25
I 50
I 100
0
I
I
25
50
I 100
d
------.)--
It--
.
‘.
J1.40
looo
t
RG.3. Mean responses of component crops and land equivalent ratio to applied nitrogen in various maize-legume intercrop systems. ( 0 ), Intercrop maize; (m), intercrop legume; (A), LER estimated from the fitted yield curves. (Drawn from the data of Rao et al., 1979.)
maize-groundnut system that was attributed to high soil N status from previous cropping (Searle et al., 1981). The LER values estimated from the regression models at each nitrogen level ranged from 1.30 to 1.36 in the maize-cowpea system, 1.43-1.55 in the maize-mungbean system, 1.21-1.25 in the maize-groundnut system, and 1.42-1.47 in the maize-soybean system. Intercrop maize increased progressively in response to applied N, heavy applications of N reducing yields of the associated legumes. In a recent study of maize-cowpea intercrop systems, Ofori and Stern (1986) also found declining trends in LER in response to applied nitrogen. They used two maize cultivars of contrasting heights; LER values declined from 1.67 to 1.42 with the short maize XL66 and from 1.40 to 1.32 with the tall maize SR99 as applied N was varied from 0 to 100 kg N/ha.
72
FRANCIS OFORI AND W. R. STERN
The results of these studies show that LER varies with nitrogen status of the soil and in general, where the inherent fertility of the soil is low, there appears to be a greater advantage of intercropping maize with legumes. 2.
Other Cereal Components
Using a 45-cm alternate row arrangement in a sorghum-pigeonpea system with N applied at 0, 60,and 120 kg N/ha at ICRISAT, India, Rego (1981) obtained a progressive increase in grain yields of both sole and intercrop sorghum (Table VIII). At nil N, intercrop sorghum yield was 33% higher than sole sorghum. At 60 kg N/ha, intercropping reduced sorghum yields by 17% and at 120 kg N/ha, the yield reduction was 20%. Applied N did not significantly affect the yield of intercrop pigeonpea, and yield reductions from intercropping ranged from 37 to 42% compared to sole pigeonpea. The LER value without applied N was 1.91 and declined to 1.42 at 120 kg N/ha. These studies show that the intercrop cereal grain yields increase progressively with applied N, while seed yields of companion legumes decrease or are less affected. It would appear that applying nitrogen does not improve the land equivalent ratio and therefore the efficiency of cereal-legume intercrop systems.
V.
NITROGEN ECONOMY OF THE SYSTEM
In fixing atmospheric N,, legumes contribute to the N content of soil either as sole crops in rotation, or as intercrops (LaRue and Patterson, 1981; Heichel and Vance, 1984). In such systems, legumes may either increase the soil N status through fixation and excretion, or in the absence of an effective N-fixing system, compete for N (Trenbath, 1976). In the literature on cereal-legume intercropping, there are surprisingly few data concerning different parts of the system. In the sections following, the existing literature is reviewed on N, fixation, N transfer, losses, and budgeting. Many of the examples cited are derived from agricultural systems other than cereal-legume intercrop systems because the necessary measurements to understand the processes involved are not always available from intercropping studies. A. NITROGEN FIXATION BY THE LEGUMECOMPONENT
The quantity of N, fixed by the legume component in cereal-legume intercropping depends on the species, morphology, density of legume in the
73
CEREAL-LEGUME INTERCROPPING SYSTEMS
mixture, the type of management, and the competitive abilities of the component crops. Legumes of indeterminate growth are more efficient, in terms of N, fixation, than determinate types. Eaglesham et af. (1982) found that in a growing season, soybean fixed more nitrogen than cowpea, but soybean used a greater amount of the N2 fixed to produce seed. While cowpea fixed less N, however, because it had a lower seed N harvest index it thus accrued more N to soil from residues. They also found that this applied to determinate (det) and indeterminate (ind) cowpea types, but there was a difference in the amount accrued: kg N/ha Cowpea cultivar ER-1 (det) TVu 1190 (ind)
N N uptake fixed ~ 32 50 33 101
Seed N
48 49
Residue N balance N 34 + 2 85 + 52
The literature contains only semiquantitative data on N, fixation of intercrop legume. Graham and Rosas (1978) found that N2 (C,H,) fixation by climbing bean (cv. P590)was essentially unaffected by intercropping with maize. N2 fixation reached a maximum of 20.6 moles/plant/hr in the sole bean at 68 days and was reduced by only 10% when intercropped with maize. Shading by the cereal reduces both the seed yield and the N, fixation potential of the companion legume (Wahua and Miller, 1978a,b). In a sorghum-soybean intercrop system, a tall variety of sorghum reduced soybean yield by 75% and N, fixation at the early pod-fill stage by 99% (Wahua and Miller, 1978b).
Soybean intercropped with Tall sorghum Short sorghum
N fixation (mole C,H,/plant/hr) 0.012 3.17
Seed yield (kdha)
688 2275
With the short sorghum, soybean received more than 90% of the incoming radiation, compared to less than 50% with the tall sorghum. In a sorghum-groundnut intercrop system, partial defoliation of sorghum increased the amount of light for the associated legume and enhanced N,(C,H,) fixation (Nambiar et af., 1983). In the absence of applied N, N,(C,H,) fixation of groundnut intercropped with maize was not influenced by shading, although light reaching the legume was reduced by 33% (Nambiar et af., 1983). When 50% kg/ha of N
74
FRANCIS OFORI AND W. R. STERN
was applied, fixation was reduced by 55% and light reaching the groundnut canopy was 54% of daylight. Heavy applications of N fertilizer (100 and 150 kg N/ha drastically reduced fvtation: 7.0 moles/plant/hr at 100 kg N/ha and 3.5 moles/plant/hr at 150 kg N/ha. Using the "N fertilizer dilution method, estimates of N derived from atmospheric N, by cowpea intercropped with maize were about 60%, corresponding to fixation of 81 kg N/ha, and in the sole cowpea, 56%, corresponding to 83 kg N/ha (Eaglesham et al., 1981). Ofori et al., (1987) evaluated the N economy of a maize-cowpea intercrop system using both "N natural abundance and "N-labeled fertilizer methods. They found that cowpea maintained its ability to fix atmospheric N, when intercropped with maize, but that N, fixation was reduced by N fertilizer applications (Table X). The comparable P values (percentage of N derived from atmospheric N,) of the intercropped cowpea with or without applied N was attributed to greater N uptake by the associated maize, which induced the companion cowpea to be more symbiotic.
Methods of Estimating N, Fixation Both direct and indirect methods are used to estimate N, fixation by legumes in the field. The indirect methods are the ureide and acetylene reduction assays; these provide qualitative and semiquantitative information only on N, fixation (LaRue and Patterson, 1981). The ureide method is limited to legumes that produce and translocate N as ureides, e.g., soybean and cowpea (Atkins, 1982), and will not be discussed further. The three principal ways of measuring N2fixation in cereal-legume intercropping are 1. the acetylene reduction method 2. the total N-difference method 3. the I5N dilution methods a. Acetylene Reduction Method. This method evolved from the observations that the enzyme nitrogenase reduced acetylene (C,H,) to ethylene (C,€&) (Dilworth, 1966, Schollhorn and Bums, 1966). Techniques for field assay include excavation and incubation of freshly excised nodulated roots in a chamber with 1-209'0 C,H, for 30 to 120 min, followed by analysis of the ethylene produced with gas chromatography. The ratio of acetylene reduced to N, fixed is assumed to be 3:1, because the reduction of N, to ammonia uses six electrons while the production of ethylene requires two (Hardy et al., 1973). The advantages of the method are sensitivity, speed, and economy. It has been used to evaluate N, fixation in cereal-legume intercropping (Graham and Rosas, 1978; Nambiar et al., 1983). It gives an instantaneous, semiquantitative measure of fixation rather than permitting the measurement of fixation over a growing season by a crop (LaRue and Patterson, 1981), something that most agronomists would be interested in.
75
CEREAL-LEGUME INTERCROPPING SYSTEMS
Table X Proportional Dependence on N, Fixation (P) and N, Fixed by Sole and Intercrop Cowpea Estimated by "N Natural Abundance (NA) and '"-Labeled Fertilizer (NL) Methods under Field Conditionsa Harvest
Cropping system
P
(070)
N, fixed
No added fertilizer N (NA experiment) kg N/ha 150 days
(straw) 150 days
(seed) 150 days
(seed
+ straw)
Sole crop Intercrop
72.0 61.3
32.4 18.0
Sole crop Intercrop
68.0 67.8
54.7 40.8
Sole crop Intercrop
69.4 65.7
87.1 58.8
Fertilizer N added (25 kg N/ha) (NL experiment) kg N/ha 150 days
(straw) 150 days (seed) 150 days
(seed
+ straw)
Sole crop Intercrop Sole crop Intercrop
47.6 64.9
26.2 25.9
57.1 69.3
55.7 47.1
Sole crop Intercrop
53.4 67.7
81.9 73.0
aFrom Ofori el al. (1987).
b. Total N-Difference Method. This estimates N, fixed by a legume as the difference in total N uptake by legume and a nonfixing control plant such as grass or cereal, expressed as:
N2 fixed by legume = Total NIegume - Total Nnonfixing plant The total N of the nonfixing plant is derived solely from soil, and the difference in values of N uptake between the legume and nonlegume is assumed to be the quantity derived by fixation (Williams et al., 1977). The basic assumption of the method is that both the legume and the nonlegume remove soil N from the same N pool and in the same proportion as the amount of N available (LaRue and Patterson, 1981). In addition to nonlegumes, nonnodulating and uninoculated legume genotypes have been used (Nutman, 1976). Bezdicek et al. (1978) compared inoculated and uninoculated soybeans under field conditions in a 2-year study. They found that uninoculated
76
FRANCIS OFORl AND W. R. STERN
soybeans accumulated an average of 90 kg N/ha, while nodulated soybeans yielded 378 kg N/ha. By difference, the net N, fixation was calculated to be 288 kg N/ha. c. I5NDilution Methods. The use of the stable isotope "N for estimating N2fixation is a direct and definitive method extensively used for checking the validity of other methods (Burris, 1974). The technique is based on the fact that 15Noccurs naturally in an almost constant ratio with I4N. The ratio of I4N: ''N is approximately 272:1, so atmospheric Nz contains 0.3663 at.% "N or 3663 ppm "N. The only exceptions to this general rule are the slight variations in Is N enrichment of natural nitrogenous substances such as soils, rocks, and coal (Hauck and Bremner, 1976). The changes in isotopic composition of samples relative to the atmosphere permit calculation of enrichment of "N in samples for the study of fixation, transfer, and transformations of N in systems. The main factors limiting the widespread use of "N techniques have been the high cost of N compounds and the need for mass spectrometers to analyze for "N abundance (Hauck and Bremner, 1976). With l5 N methods, there are two approaches, namely, "N-labeled fertilizer and "N natural abundance. Both methods are based on the assumptions that the N,-fixing and nonfixing plants have similar rooting patterns and obtain equal proportions of N from the soil, and that there is no transfer of N during the measuring period (Chalk, 1985). Eaglesham et al. (1981) used the "N-labeled fertilizer method to demonstrate N transfer in a maize-cowpea intercrop system, and Ofori et af. (1987) used both I5N natural abundance and "N-labeled fertlizer to quantify N2 fixation and evaluate the N economy of a similar intercrop system (Table X). B. NITROGEN TRANSFER Evidence in the literature suggests that the N, fixed by the intercrop legume may be available to the associated cereal in the current growing season (Agboola and Fayemi, 1972; Remison, 1978; Eaglesham et al., 1981; Pandey and Pendleton, 1986) or as a residual N for the benefit of a succeeding cereal crop (Nair et al., 1979; Searle et af., 1981; Singh, 1983). Both forms of N transfer are considered to be important and could improve the N economy of various legume-based intercrop systems. This has led to the suggestion that both current and residual N benefits should be evaluated in intercrop systems in which legumes are a component (Willey, 1979). Roots and nodules of legumes are thought to be the important sources of N transfer because of their high N contents (Butler and Bathurst, 1956). In cowpea, Minchin et al. (1978) found N from these sources to be only 6% of the total plant N; this may be inadequate to produce any substantial N
CEREAL-LEGUME INTERCROPPING SYSTEMS
77
benefit for a subsequent crop. From pot studies, Peoples et al. (1983) report that N from roots and nodules of cowpea are 13% of the total plant N. The degree to which N from intercrop legume may benefit a cereal crop depends on the quantity and concentration of the legume N, microbial degradation (mineralization) of the legume residues, utilization of these residues, and the amount of N2 fixed by the legume (Henzell and Vallis, 1977; Herridge, 1982). The N in legume residues may be tied up in the soil organic N pool and may not be readily available to the cereal crop (Ladd et al., 1983). The rate of mineralization of organic N, determined by microbial activity, is primarily influenced by the prevailing moisture and temperature regimes (Ladd et al., 1984). Henzell and Vallis (1977) estimated that under tropical conditions, 30% of the N in legume residues could be mineralized and taken up by grass after 24 weeks.
1. Current N Transfer
The idea of possible N tranfer in the current season from legumes to intercropped cereals originated from early studies on legume-nonlegume mixtures in pots under greenhouse conditions (Nicol, 1935; Virtanen et al., 1937). In a greenhouse experiment, Agboola and Fayemi (1972) observed that early-maturing legumes could possibly improve yields and N nutrition of associated maize in the current season, while late-maturing types could have this effect in the following season. They found that greengram fixed 224 kg/ha of N2in 49 days and improved intercropped maize yield by 72%. Although cowpea fixed 450 kg/ha of N, in 98 days and calopo fixed 354 kg/ha of N, in 84 days, the N2 fixed by these legumes did not influence the yields of the associated maize. Eaglesham et al. (1981) presented evidence from the field of transfer of N from legume to an intercrop cereal, using the "N-labeled fertilizer method. It was evident that at the low N rates, i.e., No and Nzs, N concentrations (percentages) of the intercrop maize were higher in the presence of cowpea than in the sole maize. The transfer of N was confirmed by the significant dilution of IsN in the intercrop maize compared to sole maize at NZs.This was because fixed N would have an enrichment close to natural abundance and its transfer would result in the dilution of "N enrichment in the intercrop maize derived from the fertilizer. Using replacement series designs, Patra et al. (1986) have reported substantial transfer of N from the legume component to the associated cereal in wheat-gram and maize-cowpea intercrop systems in both greenhouse and field studies. However, they did not provide "N data to demonstrate transfer of legume N that would have obviously resulted in
78
FRANCIS OFORI AND W. R. STERN
lower 15Nenrichment of the intercrop cereal compared to sole crop. In a maize-soybean intercrop system, Pandey and Pendleton (1986) reported that soybean green manure provided 28 kg/ha of N to maize when no N fertilizer was applied. The transfer of N from cowpea to the associated maize was not evident from either the field or the greenhouse pot studies by Ofori et al. (1987), because similar I5Nenrichments were obtained in the sole and intercropped maize. They concluded that cowpea and associated maize were competing for applied N and that the N, fixed by cowpea ended up in the seed and was harvested from the system. These findings are consistent with those reported by Danso et al. (1987)using fava beans and barley. In a maize-cowpea combination planted without N fertilizer, Remison (1978)attributed a 72% increase in intercrop maize grain yield over that of sole maize to the transfer of N from cowpea to the associated maize. Unfortunately, no crop N analysis data were provided to justify the conclusion of current N transfer claimed by the author. Based on yield and N contents, Waghmare et al. (1982) found that sorghum benefited from greengram, groundnut, soybean, and fodder and grain cowpea intercrops. Intercrop sorghum grain yield increases ranged from 3 to 16%, grain protein from 10 to 33V0, and N uptake from 16 to 49%. 2. Residual N Transfer The intercrop legume may accrue to N the soil and this may not become available until after the growing season, improving soil fertility to benefit a subsequent cereal crop. Table XI shows results from various studies that used wheat to measure residual fertility of various cereal-legume intercrop systems. Nair et al. (1979)found a mean wheat yield increase of about 30% after a maize-soybean intercrop, and after maize-cowpea the yield increase was 34% when compared to wheat after sole maize. De (1980)evaluated the residual N value of various legume-based intercrop systems and found that blackgram intercropped with either maize or sorghum improved succeeding wheat yield in all N treatments. Searle et al. (1981)found N uptake of wheat following maize-groundnut and maize-soybean intercrop systems to be higher than after maize alone. At 60 days, N uptake by wheat was about 18 kg/ha when preceded by either intercrop systems, and this was similar to N uptake of wheat supplied with 100 kg/ha. Singh (1983) estimated the N benefits to wheat of various preceding legume intercrops. When comparing wheat after sole sorghum with wheat
79
CEREAL-LEGUME INTERCROPPING SYSTEMS
Table XI Effects of Preceding Cereal-Legume Intercrop System and Applied Nitrogen on Grain Yield and Nitrogen Uptake of Succeeding Wheat N applied (kg N/ha) Parameter and treatment
0
50
100
Reference
Grain yield (kg/ha) After sole maize After maize-soybean After maize-cowpea After sole maize After maize-blackgram After sole sorghum After sorghum-blackgram
940 1240 1260 2520 2900 1610 2280
920 1150 1250 4270 5080 41 10 4540
900
Nair et al. (1979)"
1120 1190 5450 5690 5250 5910
Nitrogen uptake (kg N/ha) After sole maize After maize-soybean After maize-groundnut
11.6 17.5 17.8
15.6 21.8 21.2
17.5 21.8 21.2
De (1980)b
Searle et al. (1981)"
"Derived data; N was applied only to the preceding treatment. bN applied to the wheat crop.
after intercrops, he obtained N fertilizer equivalents of 3 kg/ha with soybean, 31 kg/ha with greengram, 46 kg/ha with grain cowpea and with groundnut, and 54 kg/ha with fodder cowpea. It seems that indeterminate legumes with lower seed yield potentials benefit the associated cereals in terms of N in the current season or as residual N for succeeding crops.
C. NITROGENLOSSES Nitrogen is lost from cropping systems through harvested crop products, gaseous N, and leaching of nitrates from soil beyond the root zone (Greenland, 1977). The complexity of the processes involved in gaseous N losses, namely, denitrification and volatilization, and the paucity of direct measurements that vary widely with environmental conditions, especially temperature and water content, has generally limited an accurate assessment of the magnitude of N losses by these means under field conditions (Vlek et al., 1981). Efforts to quantify N gaseous losses are a recent development (Simpson and Steele, 1983), and earlier estimates were made by inference from
80
FRANCIS OFORI AND W. R. STERN
measurements in nitrogen balance studies (Hauck, 1971). The losses of N in cereal-legume intercropping through denitrification and volatilization have not been studied and the examples cited in this section are mainly from other cropping systems. 1. N Harvested in Crop Products
Nitrogen harvested from crops as seed is the largest source of N loss from any cropping system. Assuming N concentration of 1-3% in cereal grain and 3-6% in legume seed, a cereal yield of 3000 kg/ha of grain removes 30-90 kg/ha of N from the soil, and 800 kg/ha of legume seed removes 24-48 kg/ha of N. 2. Gaseous Losses
The important pathways of gaseous N losses from cropping systems are through denitrification, the reduction of NO, to N 2 0 and N, by microorganisms, and volatilization of NH,. a. Denitrifeation. Denitrification results in an array of gaseous N products that renders it difficult to quantify satisfactorily the N loss via this pathway. Hauck (1971) reviewed the literature pertaining to N balances in plant-soil systems and concluded that N losses via denitrification could be of the order of 10-30070 of the N applied, and that this commonly occurs in soils wet for prolonged periods, with low CO, concentrations (Freney et al., 1978).
b. NH, Volatilization. The review by Chalk and Smith (1983) shows that N losses through NH, volatilization are usually small and that these are generally less than 2% of the total N applied. However, on a calcareous soil, Smith and Chalk (1980) measured significant losses of N, (10% of applied NH, N) and N,O (6% of applied NH, N). Simpson (1968) showed up to 60% loss of urea N applied as topdressings between 67 and 112 kg N/ha and considered NH, volatilization to be important. NH, losses from applied N depend on the ammonium source of the fertilizer and on the soil pH (Zamyatina et al., 1968). The rate loss from urea is more than from either ammonium sulfate or ammonium nitrate. Vlek et al. (1981) presented data from various studies in which different forms of N had been applied ranging between 100 and 500 kg N/ha. Assuming mean N applied to be 300 kg N/ha, the N loss via volatilization was 16.7% of N applied from urea, 1 1.4% from ammonium sulfate, and 5% from ammonium nitrate. A review on NH, volatilization by Freney et al. (1983) has shown that N losses increase with rising pH. This effect was demonstrated by Jewitt
81
CEREAL-LEGUME INTERCROPPING SYSTEMS
(1942) who found NH, losses equivalent to 0, 13, and 87% of N when ammonium sulfate was applied to soils of pH 7.0, 8.6, and 10.5, respectively.
3. Leaching Wetselaar (1962) demonstrated that rainfall is the most important factor affecting leaching of nitrates in soils. He estimated that for each millimeter of rain, the mean movement of nitrates down free-drained soils was 1.075 mm. Although severe leaching of soils of the humid tropics is regarded as a major constraint to their productivity, actual quantitative data on removal of nitrogen from soils remain sparse (Greenland, 1977). The only documentation we could find regarding N loss in cereal-legume intercrop systems was concerned with leaching of soil nitrates. Singh et af. (1978) found in a 180-cm soil profile that maize intercropped with mungbean reduced NO, N loss by 60% and intercropped with blackgram, by 41% when compared with sole maize (Table XII). The total N uptake of the maize-mungbean and maize-blackgram intercrop systems were 36.8 and 23.3 kg N/ha, respectively, over and above the N uptake by sole maize. In an 83-day maize and 155-day pigeonpea intercrop system, Yadav (1982) found that the soil profile (0-105 cm) contained larger amounts of NO, N in either the sole pigeonpea or the sole maize systems than in the mixture. The NO, N content after the mixture was 240 kg/ha, compared to 619 kg/ha after the sole maize and 419 kg/ha after the sole pigeonpea. The higher soil nitrate N in sole crops may be due to insufficient early spread of roots to absorb nitrates from all the interrow spaces, resulting in nitrate N leaching down and so enriching the lower soil profiles. From what has been said so far, it is apparent that a lack of quantitative data does not permit a satisfactory partitioning of N losses via the three main pathways, viz., denitrification, leaching, and volatilization. It can be surmised that N losses by each of these pathways could limit the productivity and
Table XI1 Proportion of Applied NO, N in 1804111Soil Profiles and Total N Uptake of Maize, Maize-Mungbean, and Maize-Blackgram Intercrop System@ NO, N loss as @lo of quantity applied
Total N uptake *g/ha)
N recovery
Treatment Sole maize Maize-mungbean Maize-blackgram
56.8 22.7 33.6
61.9 98.7 85.2
48.3
“Derived from Singh et at. (1978).
(@lo)
78.2 67.5
82
FRANCIS OFORI AND W. R. STERN
efficiency of cereal-legume intercrop systems. However, it appears that leaching losses of nitrates in intercropping systems may be less than in sole crops.
D. NITROGEN BUDGETING There are three main sources of nitrogen in cereal-legume intercrop systems: these are N fixed by the legume component from the atmosphere, from fertilizer, and from soil. The only published data that offer some scope for illustrating N budgeting are those of Eaglesham et al. (1981) in Western Nigeria and Ofori et al. (1987) in Western Australia, both studies with maize and cowpea. Using equations suggested by Rennie et al. (1982) to calculate N from fixation, from fertilizer, and from soil, a N balance sheet was constructed for such a system with the data of Eaglesham el al. (1981) (Table XIII).The densities of component crops as sole crops were 60,OOOplantdha of maize and 110,OOO plantslha of cowpea; the intercrop density was half of each sole crop density.
Table XIII N Balances (kg N/ha) of Sole Crops and Intercrops of Maize and Cowpea Determined by "N-Labeled Fertilizer Method with N Applied at 25 kg/hau Cropping system Sole crop Components Sources of N Soil (available) Seed at sowing Fertilizer Atmosphere Total Crop N Harvested seed Residues Total N lossesb Soil N balance after return of residues'
Maize
Intercrop
Cowpea
23.9
60.4
0.5
2.0
4.0 28.4
4.0 83.2 149.6
25.6 2.8 28.4
53.6 96.0 149.6
-
21.0 -21.1
Maize 28.7 0.3 1.6
Cowpea 24.6 1.O
Total 53.3 1.3 3.3
-
1.7 40.7
40.7
30.6
68.0
98.6
28.0 2.6 30.6
24.5 43.5 68.0
52.4 46.1 98.6
21 .o
4.7
4.6
9.3
+ 35.9
- 26.1
+ 18.9
-1.2
'Derived from data of Eaglesham el at. (1981). bLosses of applied N through unknown mechanisms. 1v in residues minus soil N available to crop.
CEREAL-LEGUME INTERCROPPING SYSTEMS
83
The N contributed by seeds of maize and cowpea at sowing was less than 2 kg/ha, fixed N, by intercrop cowpea was about 41 kg/ha, and N from fertilizer N was 3 kg/ha. The total N in the crops was about 99 kg/ha, consisting of N from seeds, fertilizer, N, fixation, and 53 kg/ha from the soil. Assuming a seed N harvest index of 36% for cowpea and 90% for maize, the quantity of N removed in the intercrop system was about 52 kg/ha, 28
kg/ha from maize and 24 kg/ha from cowpea. The N remaining in residues was 46 kg/ha. If the N losses in the study were due to denitrification, leaching, and volatilization, these accounted for 74% of the N applied. NH, volatilization could account for a large proportion of the N loss in the maize-cowpea study if the N fertilizer was not well incorporated in the soil. Denitrification and leaching could also be important; however, the use of ammonium fertilizer suggests that loss through leaching may be low because of the possible sorption of ammonium ions on clay minerals and organic matter (Freney et al., 1983). The resultant net change in soil N after the grain harvests and the return of residues may be calculated as N = N (residues) - N (uptake from soil)
The maize-cowpea intercrop would result in a loss of 14 kg N/ha to the soil, compared to a loss of 21 kg N/ha after sole cropping of maize and a gain of 36 kg N/ha after sole cowpea. The data of Eaglesham et al. (1981) indicate that growing maize and cowpea together does not deplete N excessively from the soil. Sole cropping of cowpea may enhance soil nitrogen fertility and could be beneficial in cereal and legume rotations if stover of high N concentrations are returned to the soil, whereas sole cropping of maize depletes soil nitrogen. The findings of Ofori et al. (1987) are consistent with these conclusions.
VI.
SUMMARY AND CONCLUSIONS A. AGRONOMIC CONSIDERATIONS
It is apparent that light and nitrogen are the main factors influencing the production efficiency of cereal-legume intercropping systems, as determined by the land equivalent ratio (LER). Various studies conducted at ICRISAT, India, and evidence from the cassava-soybean intercrop study by Tsay (1985) show that the taller component suppresses the companion legume through shading, and this is accentuated by application of nitrogen fertilizer.
84
FRANCIS OFORI AND W.R. STERN
The cereal determines the levels of dry matter, grain, and seed protein yields of the intercropping mixture. The trends in the LER, an index of intercropping efficiency, are, however, associated with the yields of legume rather than that of the cereal. In mixed swards, Harper (1977) proposed that when the components compete for growth-limiting factors, the weaker competitor determines the yield efficiency of the mixture. This theory seems to hold in maize-cowpea intercropping, because cowpea, the weaker competitor for N in the various associations with maize, accounts for more of the variations in the LER than maize (Ofori and Stern, 1986, 1987). IN THE SYSTEM B. THEROLEOF NITROGEN
The observations that the efficiency of cereal-legume intercropping relative to growing crops separately is greatest at low levels of N, suggest that at low N, the intercrop legume has a greater dependence on atmospheric N, and, presumably, competition from the associated cereal is minimized. "N dilution methods, that is, 15Nnatural abundance and "N-labeled fertilizer, provide reliable information on the amounts of N, fixed from the atmosphere by the legume as a sole crop or as an intercrop with cereals. Because in the intercropping situation, the cereal is more effective than the legume in taking up nitrogen, the legume's dependence on the atmosphere for N is enhanced and it fixes N, more efficiently than as a sole crop (Danso et al., 1987). With the increasing availability of N from other sources, such as soil and fertilizer, this dependence on atmospheric N, by the legume decreases (Ofori et al., 1987). Although 15Nmethods provide quantitative data showing the partitioning of N in intercrop systems and thus help to interpret mechanisms of nitrogen movements and transfer, the high cost involved in the analyses using a mass spectrometer precludes a much wider adoption of these methods. With the natural abundance method, the low "N abundance measured as parts per thousand units requires special analytical procedures (Bergersen, 1980). In addition, the "N-labeled fertilizer method also suffers from nonuniform labeling and a decline with time of the enrichment of 15Nin the soil N pool (Witty, 1983). To conclude, it appears that the production efficiency of cereal-legume intercropping could be enhanced through the use of more effective strains of rhizobia and low rates of fertilizer N so as to maximize N, fixation of the intercrop legume. This will partially eliminate competition for N between cereals and legumes as intercrops. To meet the high N requirements of the intercrop cereal while at the same time promoting N, fixation of the companion legume, slow release fertilizers might be used. This might render the N available to the cereal at about the peak vegetative stage, presumably after the N-fixing system of the legume has become well established.
CEREAL-LEGUME INTERCROPPING SYSTEMS
c.
SUGGESTIONS FOR
85
FUTURERESEARCH
Considerable efforts have been devoted to agronomic factors affecting the efficiencies of various cereal-legume intercropping systems. However, there do not seem to be any general proposals for maximizing the production efficiencies of different systems. Some suggestions for future research follow. Research is needed on: 1. The pathways of N losses from cereal-legume intercrop systems to maximize the utilization of N fertilizers and so reduce wastage of an expensiveinput. 2. The application of low rates of N fertilizer early in order to encourage N2 fixation of the intercrop legume, together with later application of N during the peak vegetative stage of the cereal, in order to minimize competition for N. 3. The effects of applied N on N2 fixation of the intercrop legume. 4. The effectiveness of slow-release N fertilizers to establish whether they minimize the losses of applied N. 5 . The amounts of fixed N in belowground parts of component crops to permit accurate estimates of N balances in cereal-legume intercropping systems. 6. The component density combinations that give maximum yields and efficiencies in various cereal-legume intercrop systems. With different crop combinations, the responses may vary with the optimal densities of the different components crops. 7. Finding ways of improving the production efficiency of cereal-legume intercropping systems by giving high priority to maintaining the legume component. This may be achieved through manipulating the interrow spaces of the taller associated cereal, which would also minimize interspecific competition for growth-limiting factors such as light. 8. The efficiencies of specific crop combinations of cereal-legume intercropping systems in different locations. The definitive but expensive ''N methods of quantifying N2 fixation might be used as standards for less expensive indirect approaches such as the total N-difference and the acetylene reduction methods. This requires the cooperative efforts of funding agencies in supporting costs of 15Nfertilizers and the analyses of samples from areas where intercropping remains the most common cropping system. REFERENCES Agboola, A. A., and Fayemi. A. A. 1971. J. Agric. Sci. 219-225. Agboola, A. A., and Fayemi, A. A. 1972 Agron . I 64,409412. . Ahmed, S., and Rao, M. R. 1982. Field Crops Res. 15, 147-161. Ahmed, S., Gunasena, H. P. M., and Yang, Y. H. 1979. Proc. Final INPUTS Rev. Meet. 20-24 Aug., Honolulu, p, 251. Allen, J. R., and Obura, R. K. 1983. Agron. J. 75, 1005-1009. Allison, F. E. 1973. Dev. SoiiSci. 3, 639.
86
FRANCIS OFORI AND W. R. STERN
Andrews, D. J. 1972. Exp. Agric. 8, 139-150. Andrews, D. J. 1974. Exp. Agric. 10, 57-63. Andrews. D. J., and Kassam, A. H. 1976.I n “Multiple Cropping” (R.1. Papendick, P. A. Sanchez, and G. B. Triplett, eds.), pp. 1-10. Spec. Pub. No. 27 Am. Soc. of Agron., Madison, Wisconsin. Atkins, C. A. 1982. I n “Advances in Agricultural Microbiology” (N. S. Subba Rao, ed.), pp. 25-51. Oxford and IBH Publ., New Delhi. Babalola, 0. 1980. Plant Soil 56, 59-69. Baker, E. F. I. 1978. Exp. Agric. 14,293-298. Baker, E. F. I. 1979. Exp. Agric. 15,4148. Baker, E. F. I., and Norman, D. W. 1975.Proc. Cropping Sys. Workshop, pp. 334-361. IRRI, Los Banos, Phillipines. Beets, W. C. 1977. Neth. J. Agric. Sci. 25, 95-102. Beets, W. C. 1982. “Multiple Cropping and Tropical Farming Systems,” p.156. Westview, Boulder, Colorado. Bergersen, F. J. 1980. In “Methods for Evaluating Biological Nitrogen Fixation” (F. J. Bergersen, ed.), pp. 65-110. Wiley, Chichester. Betse, C. E. 1976.Hortscience 11, 236-238. Bezdicek, D. F., Evans, D. W., Adebe, B., and Witters, R. E. (1978).Agron. J. 70, 865-898. Biscoe, P. V., and Gallagher, J. N. 1977.I n “Environmental Effects on Crop Physiology” (J.J. Landsberg and C. V. Cutting, eds.), pp. 75-100. Academic Press, New York. Boucher, D. H. 1986. Field Crops Res. 14, 105-115. Burris, R. H. 1974. I n “The Biology of Nitrogen Fixation (A. Quispe, ed.), pp. 9-31.NorthHolland Publ., Amsterdam. Butler, G.W., and Bathurst, N. 0. 1956. Proc. Int. Grassl. Congr. Palmeston North 7th, pp. 168-178. Chalk, P. M. 1985. Soil Biol. Biochem. 17, 389-410. Chalk, P. M., and Smith, C. J. 1983. Dev. Plant Soil Sci. 9,65-89. Chang, J. F., and Shibles, R. M. 1985a.Field Crops Res. 12, 133-143. Chang, J. F., and Shibles, R. M. 1985b.Field Crops Res. 12, 145-152. Chetty, C. R. K., and Reddy, M. N. 1984. Exp. Agric. 20,31-40. Chowdhury, M. S. and Misangu, R. N. 1981.Proc. Int. Workshop Intercropping (ICRISAT), 10-13 Jan. 1979, Hyderahad, pp. 217-221. Chui, J. A. N., and Shibles, R. 1984. Field Crops Res. 8, 187-198. Clements, F. E., Weaver, J. E., and Hanson, H. C. 1929.Carnegie Inst. WashingronPubl. 398, 202-233. Cordero, A., and McCollum, R.E. 1979.Agron. J. 71, 831-842. Crabtree, R. J., and Rupp, R. N. 1980. Agron. J. 72,445-448. Crookston, R. K., and Hill, D. S. 1979.Agron. J. 71,414. Dallal, R. C. 1974.Exp. Agric. 10,219-224. Dallal, R. C. 1977. Trop. Agric. (Trinidad) 54, 189-191. Danso, S. K. A., Zapata, F., Hardarson, G., and Fried, M. 1987. Soil Biol. Biochem. (submitted). Davis, J. H. C., and Garcia, S. 1983. Field Crops Res. 6, 59-75. De, R. 1980. I n “Nuclear Techniques in the Development of Management Practices for Multiple Cropping Systems,” pp. 73-84. IAEA-TECDOC-235, Vienna. De Wit, C. T. 1960. Versl. Landbouwkd. Onderz. 66, 1-82. De Wit, C. T., and van den Bergh, J. P. 1965. Neth. J. Agric. Sci. 13,212-221. De Wit, C. T., Tow, P. G., and Ennik, G. C. 1966. Versl. Landbouwkd.0nderz. 68, 1-30. Dilworth, M. J. 1966. Biochim. Biophys. Acta 127, 285-294. Donald, C. M. 1%3. Adv. Agron. 15, 1-118. Drake, M., Vengris, J., and Colby, W. C. 1951. Soil Sci. 72, 139-147.
CEREAL-LEGUME INTERCROPPING SYSTEMS
87
Eaglesham, A. R. J., Ayanaba, A., Rao, V. R., and Eskew, D. L. 1981. Soil Biol. Biochem. 13, 169- 17 1. Eaglesham, A. R. J., Ayanaba, A., Rao, V. R., and Eskew, D. L. 1982. Plant Soil 68, 183-192. Elmore, R. W., and Jacobs, J. A. 1984. Agron. J. 76, 561-564. Enyi, B. A. C. 1973. Exp. Agric. 9, 83-90. Evans, A. C. 1960. East Afr. Agric. For. J. 26, 1-10, Evans, P. S. 1977. N.Z. J. Agric. Res. 20, 331-335. Faris, M. A., Burity, H. A., Dos Reis. 0. V., and Mafra, R. C. 1983. Exp. Agric. 9, 251-261. Fawusi, M. 0. A., Wanki, S. B. C., and Nangju, D. 1982. J. Agric. Sci. 99, 19-23. Fisher, N. M. 1977. Exp. Agric. 13, 185-191. Francis, C. A. 1978. Hortic. Sci. 113, 12-17. Francis C. A. (ed). 1986. “Multiple Cropping Systems,” p. 383 Macmillan, New York. Francis, C. A., Flor, C. A., and Temple, S. R. 1976. In “Multiple Cropping” (R. I. Papendick, P. A. Sanchez, and G. B. Triplett, eds.), pp. 235-253. Spec. Pub. No. 27. Am. SOC.of Agron., Madison, Wisconsin. Francis, C. A., Prager, M., and Tejada, G. 1982a. Field Crops Res. 5,45-54. Francis, C. A., Prager, M., and Tajada, G. 1982b. Field Crops Res. 5,253-264. Freney, J. R., Denmead, 0. T., and Simpson, J. R. 1978. Nature (London) 273, 530-532. Freney, J. R., Simpson, J. R., and Denmead, 0. T. 1983. Dev. Plant Soil Sci. 9, 1-32. Freyman, S.,and Venketeswarlu, J. 1977. Can. J. Plant Sci. 57, 697-705. Gangwar, B., and Karla, G. S. 1982. Indian J. Agric. Sci. 51, 113-1 16. Gardiner, T. R., and Craker, L. E. 1981. Field Crops Res. 4, 313-320. Graham, P. H., and Rosas, J. C. 1978. J. Agric. Sci. 90, 311-317. Greenland, D. J. 1977. In “Biological Nitrogen Fixation in Farming Systems in the Tropics” (A. Ayanaba and P. J. Dart, eds.), pp. 13-25. Wiley, Chichester. Gunasena, H. P. M., Sangakkara, R., and Wickremasinghe, P. 1979. J. Natl. Sci. Counc. Sri Lanka 7, 85-93. Hadjichristodoulou, A. 1973. Cyprus Agric. Res. Inst. Tech. Bull. 14, 3-18. Haizel, K. A. 1974. Ghana J. Agric. Sci. 7, 169-178. Hall, R. L. 1974a. Aust. J. Agric. Res. 25, 739-747. Hall, R. L. 1974b. Aust. J. Agric. Res. 25, 749-756. Hardy, R. W. F., Burns, R. C., and Holsten, R. D. 1973. Soil Biol. Biochem. 5,47-81. Harper, J. L. 1961. Symp. SOC.Exp. Biol. 15, 1-39. Harper, J. L. 1977. “Population Biolgy of Plants.” Academic Press, London. Hart, R. D. 1975. Turrialba 25,294-301. Harwood, R. R., and Price, E. C. 1976. In “Multiple Cropping” (R. 1. Papendick, P. A. Sanchez, and G. B. Triplett, eds.), pp. 11-40, Spec. Pub. No. 27. Am. SOC.of Agron., Madison, Wisconsin. Hauck, R. D. 1971. In “Nitrogen-15 in Soil-Plant Studies,” pp. 65-74. IAEA-PI-341/6, Vienna. Hauck, R. D., and Bremner, J. M. 1976. Adv. Agron. 28, 219-266. Haynes, R. J. 1980. Adv. Agron. 33, 227-261. Heichel, G. H. 1985. In “Forage Legumes for Energy-Efficient Animal Production” (R. F. Barnes, P. R. Ball, R. W. Brougham, G . C. Martin, and D. J. Minson, eds.), pp. 183-190. Proc. Trilateral Workshop, 30 April-4 May, 1984, Palmeston North, New Zealand. USDA. Henzell, E. F., and Vallis, I. 1977. In “Biological Nitrogen Fixation in Farming Systems of the Tropics” (A. Avanaba and P. J. Dart, eds.), pp. 73-88. Wiley, Chichester. Herrera, W. T., and Harwood, R. R. 1973. “Crop Interrelationship in Intensive Cropping Systems” (mimeo.), IRRI, Los Banos, Phillipines.
88
FRANCIS OFORl AND W. R. STERN
Herridge, D. F. 1982. I n “Biological Nitrogen Fixation Technology for Tropical Agriculture” (P. H. Graham and S. C. Harris, eds.), pp. 593-608. CIAT, Cali, Columbia. Hiebsch, C. K. 1980. Principles of intercropping: Effects of nitrogen fertilization, plant population and crop duration on equivalency ratios in intercrop versus monoculture comparisons. PhD thesis, North Carolina State Univ., Raleigh. Hiebsch, C. K., and McCollum, R. E. 1987. Agron. J. 79, 15-22. Hulugalle, N. R., and Lal, R. 1986. Agron. J. 77, 86-90. Huxley, P. A., and Maingu, Z. 1978. Exp. Agric. 14.49-56. IAEA 1980. “Nuclear Techniques in the Development of Management Practices for Multiple Cropping Systems,” p. 154. IAEA-TECDOC-235. Vienna. Ibrahim, N. E., and Kabesh, M. 0. 1971. U.A.R. J. SoilSci. 11, 271-283. IRRI 1974. Annual Report for 1974, pp. 323-347. Los Banos, Phillipines. Jackman, R. H., and Mouat, C. H. 1972. N.Z. J. Agric. Res. 15,667-675. Jewitt, T. N. 1942. Soil Sci. 54, 401-409. Kurtz, T., Melsted, S. W., and Bray, R. H. 1952. Agron. J. 44, 14-17. Ladd, J. N., and Amato, M. 1984. Proc. Int. Syrnp. Nitrogen Manag. Farm Syst. Humid Trop. IITA, Ibadan, Nigeria. Ladd, J. N., Amato, M., Jackson, R. B., and Butler, J. H. A. 1983. Soil Biol. Biochem. 15, 23 1-238.
Ladd, J. N., Butler, J. H. A., and Amato, M. 1984. “The role of microorganisms in a sustainable agriculture.” Proc. Int. Meet. pp. 269-286. Wye, England. Lai, T. M., and Lawton, K. 1962. Soil Sci. SOC.Am. Proc. 26, 58-62. Lakhani, D. A. 1976. A crop physiological study of mixtures of sunflower and fodder radish. PhD. thesis, Univ. of Reading. LaRue, T. A., and Patterson, T. G. 1981. Adv. Agron. J. 34, 15-38. McCollum, R. E. 1982. Expert Consult. Fertilizer Use Multiple Crop. Syst. FAO, 1-6 Feb., New Delhi. Mafra, R. C., de A. Lira, M., Arcoverde, A. S. S., Robeiro, G., and Faris, M. A. 1981. Proc. Int. Workshop Intercropping (ICRISA T), 10-13 Jan., 1979, Hyderabad, pp. 46-51. Marshall, B., and Willey, R . W. 1983. Field Crops Res. 7 , 141-160. Martin, M. P. L. D., and Snaydon, R. W. 1982. Exp. Agric. 18, 139-148. May, K. W. 1982. Exp;. Agric. 18, 149-156. Mead, R., and Riley, J. 1981. J. R. Statist. SOC. Ser. A 144, 462-509. Mead, R., and Stern, R. D. 1980. Exp. Agric. 16,329-342. Mead, R., and Willey, R. W. 1980 Exp. Agric. 16, 217-228. Minchin, F. R., Summerfield, R. J., and Eaglesham, A. R. J. 1978. Trop. Agric. (Trinidad) 55, 107-115.
Mohta, N. K., and De, R. 1980. J. Agric. Sci. 95, 117-122. Monteith, J. L. 1977. Philos. Trans. R. SOC.London Ser. B. 281, 277-294. Nair, K. P. P., Patel, U. K., Singh, R. P., and Kaushik, M. K. 1979. J. Agric. Sci. 93, 189-194.
Nambiar, P. T. C., Rao, M. R., Reddy, M. S., Floyd, C. N., Dart, P. J., and Willey, R. W. 1983. Expl. Agric. 19, 77-86. Natarajan, M., and Willey, R. W. 1980a. J. Agric. Sci. 95, 51-58. Natarajan, M., and Willey, R. W. 1980b. J. Agric. Sci. 95, 59-65. Nicol, H., 1935. Emp. J. Exp. Agric. 3, 189-195. Norman, D. W. 1974. J. Dev. Stud. 11, 3-21. Nutman, P. S. 1976. I n “Symbiotic Nitrogen Fixation in Plants” (P. S. Nutman, ed.), pp. 21 1-237. Cambridge Univ. Press, London. Ofori, F., and Stern, W. R. 1986. Field Crops Res. 14, 247-261. Ofori, F., and Stern, W. R. 1987. Exp. Agric. 23,41-52.
CEREAL-LEGUME INTERCROPPING SYSTEMS
89
Ofori, F., Pate, J. S., and Stern, W. R. 1987. Plant Soil (in press). Okigbo, B. N., and Greenland, D. J. 1976. I n “Multiple Cropping” (R. I. Papendick, P. A. Sanchez, and G. B. Triplett, eds.), pp. 63-101. Spec. Pub. No. 27. Am. SOC. of Agron., Madison, Wisconsin. Osiru, D. S. O., and Willey, R. W. 1972. J. Agric. Sci. 79, 531-540. Osiru, D. S. 0.. and Willey, R. W. 1976. Rep. Symp. Intercrop. Semi-arid Areas, 10-12 May, Morogoro, Tanzania p. 71. Oyejola, B. A., and Mead, R. 1982. Exp. Agric. 18, 125-138. Pandey, R. K., and Pendleton, J. W. 1986. Exp. Agric. 22, 179-185. Papadakis, J. S. 1941. J. Am. Soc. Agron. 33, 504-511. Papendick, R. I., P. A. Sanchez, and G. B. Triplett 1976. “Multiple Cropping,” p. 378. Spec. Pub. No. 27. Am. SOC.of Agron., Madison, Wisconsin. Patra, D. D., Sachdev, M. S., and Subbiah, B. V. 1986. Biol Fertil. Soils 2, 165-171. Peoples, M. B., Pate, J. S., and Atkins, C. A. 1983. J. Exp. Bot. 34, 563-578. Putnam, D. H., Herbert, S. J., and Vargas, A. 1985. Exp. Agric. 21,4141. Rabotnov, T. A. 1977. I n “Application of Vegetation Science to Grassland Husbandry” (W. Krause, ed.), pp. 459-497. Jung. The Hague. Rao, M. R., and Willey, R. W. 1980. Exp. Agric. 16, 29-39. Rao, M. R., and Willey, R. W. 1983. Exp. Agric. 19, 67-78. Rao, M. R., Ahmed, S., Gunasena, H. P. M., and Alcantra, A. P. 1979. h o c . Final INPUTS Rev. Meet. 20-24 Aug., Honolulu pp. 123-160. Reddy, K. A., Reddy, K. R., and Reddy, M. D. 1980. Exp. Agric. 16, 179-184. Reddy, M. N., and Chetty, C. K. R. 1984. Exp. Agric. 20, 171-177. Reddy, M. S., and Willey, R. W. 1981. Field Crops Res. 4, 13-24. Rees, D. J. 1986. Exp. Agric. 22, 169-177. Rego, T. J. 1981. Proc. Int. Workshop Intercrop. (ICRISAT), 10-13 Jan. 1979, Hyderahad, pp. 210-2 16.
Remison, S. U. 1978. Exp. Agric. 14, 205-212. Remison, S. U. 1982. Indian J . Agric. Sci. 52, 500-505. Rennie, R. J., Dubetz, S., Bole, J. B., and Muendel, H.-H. 1982. Agron. J. 74, 725-730. Riley, J. 1984. Exp. Agric. 20, 19-29. Ruthenberg, H. 1980. “Farming Systems in the Tropics,” 3rd Ed., p. 424. Clarendon, New York. Schollhorn, R., and Burris, R. H. 1966. Proc. Natl. Acad. Sci. U.S. A . 38,213-216. Searle, P. G. E., Comudom, Y., Shedden, D. C., and Nance, R. A. 1981. Field Crops Res. 4, 133-145.
Shackel. K. A., and Hall, A.E. 1984. Field Crops Res. 8, 381-387. Simpson, J. R. 1968. Trans. Int. Congr. Soil Sci., 9th, Alelaide, pp. 459-466. Simpson, J. R., and Steele, K. W. 1983. Dev. Plant Soil9, 215-236. Singh, B., Rana, D. S., and Sekhon, G. S. 1978. Plant Soil 49,633-639. Singh, S. P. 1981. J. Agric. Sci. 97, 655-661. Singh, S. P. 1983. J. Agric. Sci. 101,401-405. Singh, S . P., and Jain, 0. P. 1984. J. Agron. Crop Sci. 153,4041. Sivakumar, M. V. K., and Virmani, S. M. 1980. Exp. Agric. 16, 377-386. Sivakumar, M. V. K., and Virmani, S. M. 1984. Agric. For. Meteorol. 31, 131-141. Smith, C. J., and Chalk, P. M. 1980. SoilSci. Soc. Am. J. 44,277-282. Spitters. C. J. T. 1983. Neth. J. Agric. Sci. 31, 143-155. Trenbath, B. R. 1974. Adv. Agron. 26, 177-204. Trenbath, B. R. 1976. In “Multiple Cropping” (R. I. Papendick, P. A. Sanchez, and G. B. Triplett, eds.), pp. 129-169. Spec.Pub. No. 27. Am. SOC.of Agron., Madison. Wisconsin. Trenbath, B. R., 1982. Proc. Front. Res. Agric. Indian Statist. Inst. Golden Jubilee Int. Conf.. Calcutta, pp. 265-286. Trewartha, 0. H. 1954. “An Introduction to Climate.” McGraw-Hill, London.
90
FRANCIS OFORI AND W. R. STERN
Tsay, J. 1985. Physiological studies of cassava-soybean intercropping. PhD thesis, Univ. of Queensland, Australia. Van den Bergh, J. P. 1968. Versl. Landbouwkd. Onderz. 714, 1-71. Virtanen, A. I., van Hausen, S., and Laine, T. 1937. J. Agric. Sci. 27, 323-349. Vlek, P. L. G., Fillery, I. R. P., and Burford, J. R. 1981. Plant Soil 58, 133-175. Wade, M. K., and Sanchez, P. A. 1984. Field Crops Res. 9, 253-263. Wagmare, A. B., and Singh, S. P. 1982. Indian J. Agron. 27, 423428. Wagmare, A. B., and Singh, S. P. 1984. Exp. Agric. 20, 261-265. Wagmare, A. B., Krishnan, T. K., and Singh, S. P. 1982. J. Agric. Sci. 99, 621-629. Wahua, T. A. T. 1983. Exp. Agric. 19, 263-275. Wahua, T. A. T., and Miller, D. A. 1978a. Agron. J. 70. 287-291. Wahua, T. A. T., and Miller, D. A. 1978b. Agron. J. 70, 292-295. Wanki, S. B. C., Fawusi, M. 0. A., and Nangju, D. 1982. J. Agric. Sci. 99, 13-17. Wetselaar, R. 1962. Plant Soil 16, 19-31. Willey, R. W. 1979. Field Crops Absfr. 32, 1-10, 73-85. Willey, R. W. 1985. Exp. Agric. 21, 119-133. Willey, R. W., and Osiru, D. S. 0. 1972. J. Agric. Sci. 79, 517-529. Willey, R. W., and Reddy, M. S. 1981. Exp. Agric. 17, 257-264. Willey, R. W., and Roberts, E. H. 1976. Proc. Inf. Solar Energy Soc. Conf., Univ. Reading. Williams, R. 0 . 1936. Am. J. Exp. Biol. Med. Sci. 14, 165-185. Williams, W. A., Jones, M. B., and Delwiche, C. C. 1977. Agron. J. 69, 1023-1024. Witty, J. F. 1983. Soil Biol. Biochem. 15, 631-639. Yadav, R. L. 1982. Exp. Agric. 18, 37-42. Zamyatina, V. B., Borisova, N. I., Varyushkina, N. M., Burtzeva, S. V., and Kirpaneva, L. I. 1968. Trans. Int. Congr. Soil Sci., 9th. Adelaide, pp. 513-522.
ADVANCES IN AGRONOMY. VOL. 41
GENOTYPIC VARIATION IN CROP PLANT ROOT SYSTEMS J. C. O’Toole and W. L. Bland Texas Agricultural Experiment Station, Texas A&M University Blackland Research Center, Temple, Texas 76503
I.
INTRODUCTION
Improvement of crop root systems has lagged behind that of aboveground plant characteristics. This disparity may be attributed to the root system’s concealment in the soil and variable nature, both of which enormously complicate observation and experimentation. The root system has traditionally been associated with water and nutrient absorption and anchorage of the plant. More recently increased interest in root biosynthesis of compounds involved in regulating development of the entire plant, interaction of roots with mycorrhizal fungi, and the increasing importance of symbiotic nitrogen fixation have modified the importance of studying plant root systems. The increase in root studies during the 1970s was evidenced by publication of symposia proceedings and authored books covering all aspects of root science (Whittington, 1969; Carson, 1974; Torrey and Clarkson, 1975; Russell, 1977; Harley and Russell, 1979). With the growing interest in root systems came an increased appreciation for interspecies variation in root form. Genetic improvement of crop species, however , requires knowledge of intraspecies variability in significant root parameters and its genetic control. Our review of this subject will emphasize anatomical and morphological characteristics relevant to the spatial and temporal expansion of the root system of a growing crop plant. The nature of this literature ranges from studies in which serendipitous findings revealed differences in dry weight or gross morphology to a few definitive inheritance studies and the even rarer report on genotypic variation of root physiological processes. We will not emphasize genotypic variation in root system response to nutrient concentration and related reports on selective ion transport (see Epstein and Jefferies, 1964; Barley, 1970; Lafever, 1981; Barber and Silberbush, 1984; Gabelman and Gerloff, 1986) but will include relevant morphological findings from that literature. In addition, we address the genetic bases of intraspecies variation in root parameters, although the literature is limited. 91 Copyright 0 1987 by Academic Press, Inc. All rights of reproduction in any form reserved.
92
J . C. O’TOOLE AND W. L. BLAND
Finally, we discuss variation in root system phenotypes from the viewpoint that such flexibility in root system development is a valuable trait rather than a problematic “genotype x environment” interaction in the anaylsis of variance of experimental results. The last review of genotypic variation in root system traits that we are aware of was presented to a symposium in 1968 (Troughton and Whittington, 1969). Since then a number of limited reviews have been published which treat genotype variation in root parameters from the perspective of the author’s discipline or that of specific crops: tomato and bean (Zobel, 1975); cereals (Hurd, 1974, 1976; Hurd and Spratt, 1975); general crop species (Russell, 1977); plant breeding (Mackey, 1980; Quisenberry, 1982); crop physiology (Sullivan, 1983); C, grasses (Jones, 1985); and soil science (Hamblin, 1985; Taylor and Nguyen, 1987).
II. SIGNIFICANCE OF GENOTYPIC VARIATION IN ROOT SYSTEMS A. ADAPTATION TO DIVERSE ENVIRONMENTS
The soil component of the plant’s environment is more complex and heterogeneous than the aerial component. Soil chemical and physical properties can change severalfold in the space covered by 2 or 3 days’ growth of a single root tip (cu. 10 cm) or in the time it takes for rainwater to infiltrate to a profile location. Ecological studies provide many examples of plant root systems adapted to cope with variability in major edaphic environmental factors such as temperature and water (Ludwig, 1977; Chapin, 1977; Larcher, 1980). For example, Sydes and Grime (1984) demonstrated how four perennial grassland species were adapted to particular edaphic niches in a shallow limestone soil profile. Vertical distribution of root dry matter illustrated that some species were better adapted to rooting in the shallow soil near the surface while others exploited crevices in the underlying limestone. The study also illustrated intrinsic differences in root development among the species when nutrient and surface drought treatments were applied. Collectively the four species demonstrated adaptive traits which optimized utilization of available soil resources. The concept of adaptation to and optimization of edaphic niches appears applicable to agriculture as well. B. EDAPHIC NICHES IN AGRICULTURE
Great diversity exists among the soils of the earth’s tropical and temperate agricultural zones. The ecological analog predicts that root
GENOTYPIC VARIATION IN PLANT ROOT SYSTEMS
93
systems optimally adapted to them must also exhibit great diversity. However, unlike adaption of wild species, the success of agricultural species is based on the magnitude and stability of the economic product. We now turn to the discussion of edaphic niches in irrigated and rain-fed agriculture.
I . Irrigated High-Input Systems In modern irrigated high-input agricultural systems every effort has been made to eliminate soil physical and chemical constraints to high yield. In such a system high levels of chemical fertilizers, supplied in the root zone several times throughout the crop season, assure luxury levels of nutrient availability in the soil solution. In combination with irrigation, appropriate root zone chemical and physical modifications can potentially result in a growth environment with low levels of root zone stresses (Arkin and Taylor, 1981; Barber, 1984). In this extreme case of high-level technology there may be an opportunity to favorably alter the root morphology of crop species by reducing the dry matter allocation to the root system (Passioura, 1983; Smucker, 1984). For example, a modern semidwarf rice cultivar growing in irrigated and puddled soil with slow release fertilizer placed into the root zone during transplanting may require only a 15-cm rooting depth for its small fibrous root system. (Savant et al., 1982; Yoshida and Hasegawa, 1982). In contrast many landraces of rice adapted to rain-fed culture exhibit substantially larger root systems with some large-diameter roots extending into the plow pan and deeper soil levels (IRRI, 1984; Ekanayake et al., 1985). 2. Rain-Fed Variable-Input Systems
In spite of the promise of irrigated agriculture, most of the world’s cropland is rain-fed and will remain so in the foreseeable future (Boyer, 1982). Thus, the root environment will often be suboptimal. By comparison the root system suited to irrigated high-technology culture may be simpler in form and function than a root system able to cope with temporal and spatial variability in soil characters associated with uncertain soil water status. The amount and frequency of precipitation during the growing season, soil temperature regime, and level of native soil fertility greatly impact the soil environment and root growth and function. Rain-fed agricultural ecosystems range from no-input mixed cropping of shifting cultivators on marginal soils to complex soil, water, and fertility management systems rivaling irrigated areas in production. One avenue of further crop improvement for these diverse rain-fed agricultural complexes lies in optimizing root systems to decrease soilrelated stresses. Success in this endeavor depends on our ability to define
94
J. C. O’TOOLE AND W. L. BLAND
and understand the edaphic environment while exploiting the range of genetically based adaptability in crop species.
111.
EVIDENCE OF GENOTYPIC VARIATION
Our review of genotypic variation is presented by species, addressing first the monocotyledons and then the dicotyledons. Utilization of genotypic variation in improving crop plants requires knowledge of heritability and genetic control. These topics are also reviewed. We begin with a brief review of relevant aspects of root morphology and anatomy. A.
MORPHO-PHYSIOLOGICALROOTPARAMETERS
I . Monocotyledons
The root morphology and anatomy of monocotyledons differs from that of the dicotyledons in several fundamental respects (Esau, 1977). For the purpose of our discussion the majority of the monocotyledon root system is composed of adventitious or nodal roots arising from a meristem at the base of the stem. Roots initiated in the embryo on the hypocotyl or higher on the axis are termed seminal roots and may be numerous in addition to the embryonic radicle. Some or all seminal roots may die after the adventitious root system begins to develop. The formation of adventitious roots in the Graminae is often associated with tiller development, creating distinct root-tiller relationships. Monocotyledon root anatomy is characterized by a cortex of variable thickness surrounding a central stele with vascular tissues arranged around a central pith. Secondary root growth does not take place in monocotyledons; thus the hydraulic characteristics of the root are established early in its life. In both monocotyledons and dicotyledons lateral roots are initiated in the pericycle and grow through the cortex to emerge at the surface of the parent root. 2. Dicotyledons
Dicotyledon root system morphology varies widely. Most possess a taproot from which lateral roots arise. The taproot produces lateral or branch roots with the youngest located nearest the root apical meristem and the oldest near the base of the plant. Typically the taproot is referred to as the primary root with lateral roots and their branches referred to as secondary and tertiary roots, etc., up to fourth or fifth order branches.
GENOTYPIC VARIATION IN PLANT ROOT SYSTEMS
95
Lateral roots originate from the pericycle of taproots or other lateral roots. In addition, adventitious roots may arise from nonroot tissues or nonpericycle tissues of older roots. Dicotyledon roots, unlike monocotyledons, are capable of secondary growth arising from the vascular cambium. The anatomy and vascular arrangement in the primary tissue of dicot roots is characterized by the number of xylem poles or vascular bundles arranged around a central point within the stele, i.e., diarch, triarch, tetrarch, or polyarch. As in monocotyledons elongation of individual root axes is by celluar division at the apical meristem and elongation in the zone behind the apex. 3. Pathway of Water Flow
The flux of water through the plant root is driven by gradients in the chemical potential of water from the soil to the leaf mesophyll cell surface, where evaporation takes place. Water entering the root system flows from the soil matrix to the outer wall of the epidermis (includingany root hairs), through the cells of the epidermis, cortex, endodermis, and stelar tissues to the lumen of a xylem vessel. The relative importance of flow along cell walls versus flow through the cells remains a topic of debate (Boyer, 1985). Water movement up to the shoot proceeds via the xylem vessels. The major resistances to water flow in the root can be partitioned into radial (inter- and intracellular from the epidermis to the stelar tissues) and axial resistance (within the xylem vessels) (Taylor and Klepper, 1978). The axial resistance can be viewed primarily as physical and anatomical in nature, dependent on the diameter and number of xylem vessels. However, the radial resistance also exhibits a physiologicalcomponent due to the contribution of cellular membranes (Fiscus, 1983). The pathways, physiology, and kinetics of ion uptake are beyond the scope of our discussion and have been reviewed elsewhere (Epstein, 1972; Barber, 1984). B.
MONOCOTYLEDONS
Genotypic variation has been documented for a wide variety of root parameters in monocotyledonouscrop species. Table I provides a summary of recent literature including a brief indication of the root parameters studied, the proposed value or relevance, and the authors. Not all studies cited in Table I are included in the following discussion due to space limitations. 1. Barley
Although we do not intend to review genotypic variation in ion uptake or transport, variation in root morphology has been revealed by such studies.
Table I
A Summary of Recent Reports on Genotypic Variation in Crop Plant Root Systems of Monocotyledon S p e c i d Crop
Root system parameter
Purpose or value
Barley (Hordeum vulgare L.)
Dry weight, number, and surface area of root members Root diameter, angle Root length and branching
Root system' response to nutrient deficiency Lodging resistance Morphogenetic effects of nitrogen and temperature Drought resistance Optimize water use efficiency
Root depth, number, length Root weight and length or depth Corn (Zea mays L.)
Oats (Avena sp. L.)
Orchardgrass (Dactylisglomerata L.)
Root number, length, dry weight and shootxoot ratio Root weight, pulling resistance Root weight, length per plant Shoot:root ratio, senescence Root pulling resistance, system lateral spread Root length and branching Root elongation rate Number, diameter and angles of nodal root growth Seedling root length Root distribution and fineness Root system length and area Rate of root volume and dry weight increase Root length, dry weight, and volume Number of crown roots, shoot: root ratio Root length to leaf area ratios Root dry weight
Developmental morphology, response to root pruning Lodging resistance Phosphorous uptake Root morphology Tolerance to corn root worms, root rot. and lodging Morphogenetic effects of nitrogen and temperature Nitrogen uptake and translocation Adaptedness to cultural systems
Reference Hackett (1%8) Jezowski (1978) Geisler and Krutzfeldt (1983) Gorny and Patyna (1984) Brown et al. (1987); Cooper et al. (1987) Andrew and Solanki (1966) Nass and Zuber (1971) Nielsen and Barber (1978) Chakraviuty and Karmakar (1980) Jenison et al. (1981); Penny (1981) Geisler and Krutzfeldt (1983) Pan et 01. (1985) Irwin et al. (1985)
Seedling drought tolerance Drought tolerance Nitrogen effects Assessment of genotypic variation
Cutforth et 01. (1986) Aina and Fapohunda (1986) Mackay and Barber (1986) Carrigan and Frey (1980)
Establish range of genetic variation Assessment of genotypic variation Drought resistance Adaptability to soil temperature and moisture
Murphy and Long (1979); Murphy et al. (1982) Heen (1980) Larsson (1982, 1986) Finn and Mack (1964)
Pensacola bahiagrass (Paspalum notatum Fh?e) Rice (Oryza sativa L.)
Rye (Secale cereale L.) Ryegrass (Lolium perenne L.) Ryegrass (Lolium sp.) Sorghum (Sorghum bicolor L. Moench.)
Root diameter
Penetrate compacted soil layers
Elkins and Sickle (1984)
Root diameter, branching, water extraction pattern Root system morphology Root activity at depth Root pulling force Deep root:shoot ratio Rooting depth, density, angle of penetration and general pattern Length, thickness, number, weight Maximum length, thickness Root pulling force Growth rate, cross-sectional area
Drought resistance, adaptation to upland culture Drought resistance Increased soil water use Drought resistance Exploitation of soil water Higher plant water status and yield Drought resistance Drought resistance Drought tolerance Seedling vigor
Nicou et al. (1970)
Armenta-Soto et al. (1983) Ekanayake et al. (1985a) Ekanayake et al. (1985b) Sullivan and Pfahler (1986)
Rate of post-water-stress new root axes development Root weight
Recovery from drought
Troughton (1980)
Relation to nitrate uptake
Goodman (1983)
Rooting depth and morphology
Vegetative drought tolerance
Bhan et al. (1973)
Growth rate, lateral branching Root weight, shoot:root ratio Root activity Weight, length, volume Shoot:root ratios, vertical root length distribution
Morphology, maturity genes Adaptation to semiarid zones Water and nutrient uptake Drought resistance Assessment of genotypic variation, drought resistance
Root length density and root dry weight Root length density and distribution
Increased water extraction at depth during drought Drought resistance, root system response to drought
Blum et al. (1977a) Saint-Clair (1977) Damodar et al. (1978) Nour and Weibel (1978) Jordan et al. (1979); Jordan and Miller (1980); Jordan and Monk (1980) Wright (1981); Wright and Smith (1983) Retta et al. (1982, 1983, 1984)
Chang et al. (1972) Reyniers and Binh (1978) O’Toole and Soemartono (1981) Yoshida and Hasegawa (1982) Mambani and Lal (1983a,b)
(continued)
Table I (Continued) Crop Sugarcane (Saccharurn sp.) Tall fescue (Festuca arundinacea Schreb.) Timothy (Phleumpratense L.) wheat (Triticurn aestivurn L.)
wheatgrass, bluebunch [Agropyron spicatum (Pursh) Scribn., and Smith]
Root system parameter
Purpose or value
Reference
Degree of branching: vertical vs. horizontal orientation Root diameter
Adaptation to soil texture and lodging resistance Root penetration of plowpans
Cited by Jones (1985)
Root diameter or thickness Root weight
Survival and ground cover Response to soil moisture, temperature, and fertility Drought resistance
Williams et al. (1981, 1982) Mack and Finn (1970)
Lodging resistance Water extraction at depth
Pinthus (1%7) Derera et ul. (1%9)
Variety-specific agronomy Descriptive morphology
Katyal and Subbiah (1971) MacKey (1973, 1980)
Effect of semidwarf genes on root system expression Increased depth of soil water extraction Adaptation to rain-fed conditions
Welbank et al. (1973); Lupton et al. (1974) Campbell and Lafever (1977)
Axial root conductance
Cornish (1981) Richards and Passioura (1981a.b); Richards (1987) Sadhu and Bhaduri (1984) Harris and Gwbel(l976)
Vertical growth rate, proportion of root system below 30 cm Spreading angle of root system Maximum depth, number, and weight Horizontal and vertical spread Number, depth, and vertical proportioning of root weight Distribution of root dry weight, length, and activity Seminal root length and growth rate Seminal and nodal root length and extension rate Xylem vessel number and radii Seminal xylem vessel diameter Root length and number Growth at low (2 T)temperature
Traps are listed in alphabetical order by common name.
Increased axial resistance to water
flow Genotypic variation Establishment and adaptedness
Elkins et al. (1977, 1979)
Hurd (1964,1%8)
Sirohi et al. (1978)
GENOTYPIC VARIATION IN PLANT ROOT SYSTEMS
99
The work of Hackett (1968) demonstrated the effect of nutrient level on root morphology. He subjected two barley (Hordeum vulgare L.) cultivars to phosphorus- and potassium-deficient nutrient supplies. In the control plants, the cultivars differed significantly in number and length of seminal, nodal, and lateral roots. Nutrient deficiency dramatically reduced the measured parameters in both cultivars, although to different degrees. Observations of this aspect of genotypic variation, i.e. , differential sensitivity to environmental parameters, are significant and may easily be overlooked. Few studies have compared root growth of a modern cereal cultivar with that of an endemic landrace over the entire crop duration. Brown et al. (1987) monitored growth and distribution of roots of two barley cultivars: Beecher, from California, and Arabic Abiad, a local landrace of Aleppo province, Syria. The study was conducted at two field sites varying in soil type (chromoxeret and calciorthid) and located in different rainfall zones. The landrace exhibited greater total root weight and length below 15-cm soil depth at all growth stages sampled. The vigorous early root and shoot growth and greater root length density at depth exhibited by the landrace resulted in higher evapotranspiration rates throughout the crop season (Cooper et al., 1987). These comprehensive studies illustrated both the temporal and spatial adaptation of the landrace root system. 2. Corn Regardless of the research objective, studies have generally found abundant genotypic variation in root systems among inbred lines of corn (Zea mays L.). Andrew and Solanki (1966) carried out greenhouse observational studies with hydroponically cultured plants. They found significant variation among 14 inbred lines for a number of root morphological parameters. When root parameters were related to field performance, the proportion of seminal roots was highly correlated with early plant height growth and vigor in the field. The role of seminal roots in the latter case may be indirect due to rapid seminal root establishment and the consequent early generation and development of the permanent nodal roots. In addition, when root pruning was performed in the field to simulate stress or mechanical damage to nodal roots, the inbreds with early vigorous seminal root development showed less reduction in height and ear weight. Cutforth et al. (1986) demonstrated variation among corn hybrids for early root system growth. No difference was observed among three hybrids for growth rate as a function of temperature. However, mean root length proved much less sensitive to lowered soil water contents in one of the three hybrids than in the others.
100
J. C. O’TOOLE AND W. L. BLAND
Other studies have sampled mature field-grown plants in an effort to describe gross morphological structure of the entire root system. Chakravarty and Karmakar (1980) excavated and detailed the threedimensional distribution of root dry weight in four composite cultivars and a local landrace. The horizontal and vertical distribution at 45, 75, and 105 days after sowing revealed that the landrace had the smallest root system and the highest shoot:root ratio. The composite cultivars also varied significantly in total root system dry matter. Two of the composite cultivars continued to show increased root weights up to maturity, contrary to the common perception that cereals suspend root growth after flowering in favor of the ear and developing grains. Irwin et al. (1985) found two distinct phenotypes for nodal root angle in a survey of 11 hybrids. The hybrids had either horizontally spreading or more vertically oriented nodal root systems. Their observational study did not relate the structural differences to function in the plant or crop. Plant breeders have developed a vertical-pull-resistance technique to test entire corn root systems for tolerance to the rootworm complex. Jenison et al. (1981) studied the gross characteristics of root dry weight and nodal root spread in 44 inbreds. Vertical-pull resistance was significantly correlated with both variables and genotypic variation exhibited for all three measurements. In contrast to more detailed root system parameters, these gross characteristics were found to be relatively stable over environments (Jenison et al., 1981; Penny, 1981). Genotypic variation in root morphology has often been exhibited in studies of nutrient uptake. Nielsen and Barber (1978) measured root weights and lengths of 12 inbreds grown in solution culture. They described two- to threefold differences in root weight of seedlings and the relationship of root length to root or shoot dry weight. Although the extension of this work to a field situation did not involve root measurements, shoot uptake of P provided indirect evidence that the variation expressed in the controlled environment of solution culture could be extrapolated to field-grown corn plants. Pan et al. (1985) investigated the differential ability of corn genoppes to take up and translocate nitrate. Lateral root proliferation and specific root length in rapidly growing 5- to 8-day-old seedlings was related to nitrate translocation. Genotypes varied in these morphological parameters through time, leading the authors to conclude that future research on nitrate uptake should also address genotypic variation in root developmental morphology. 3. Forage Grasses
Studies involving the root systems of forage grass species cover a wide range of objectives from perforating plowpans to low soil temperature
GENOTYPIC VARIATION IN PLANT ROOT SYSTEMS
101
tolerance. A number of papers have discussed the use of tall fescue (Festuca arundinacea Schreb.) and Pensacola Bahiagrass (Paspalum notatum Flugge) as means of creating macropores in plowpans. Large root diameter was associated with the desired penetration ability. Succeeding crops are then able to root through the plowpan to exploit additional soil volume for water and nutrients (Elkins and Sickle, 1984; Elkins, 1985). Published accounts of the actual screening are vague, but it appears that there is genotypic variation in root diameter among lines of both Bahiagrass and tall fescue. Although no data are available for Bahiagrass, tall fescue lines exhibited a range of root diameters from 0.5 mm to 3.6 mm (Elkins et a/., 1977). A large-rooted tall fescue line was associated with better stand persistance, percentage plant survival, and sward cover scores. The superior establishment and forage yield of the large-rooted line is presumed to be related to penetration of the compacted plowpan, although measurements of soil water content did not substantiate that presumption (Williams et al., 1981). The combined effects of genotype and environment on root weight of orchardgrass (Dactylis glomerata L.) was investigated by Finn and Mack (1964) and on timothy (Phleum pratense L.) by Mack and Finn (1970). Four cultivars of orchardgrass cultured in a greenhouse showed differential response (changed rank) in total root weight across temperature, moisture, and fertility treatments. Cool soil conditions (10°C) yielded higher root weight than warm (20 "C) temperatures and more important, the cool temperature treatment increased the ability to discriminate genotypes. One entry in the timothy study of Mack and Finn (1970) showed a 25% greater total root dry weight than the other five genotypes when exposed to cool soil temperature (10 "C), whereas at higher soil temperatures (27 and 20 "C) varietal differences were not apparent. These studies illustrate well the complexity of response to be expected when such major factors as water, temperature, and fertilization are varied. They also underline the need to identify the most limiting environmental factor(s) for a particular agroecological region if screening or selection for genotypic differences is to be effective. Striking genotypic variation in growth of roots at low soil temperature was demonstrated by Harris and Goebel (1976) among ecotypes of bluebunch wheatgrass [Agropyron spicatum (Pursh) Scribn. and Smith]. They collected seed at 45 sites from Arizona to Alaska. Figure 1 illustrates the ecotypic variation found for root growth over a 21-day period a 2 "C, ranging from 1.4 m to 11.9 m in length. 4.
Oat'
Carrigan and Frey (1980) compared root and shoot growth of two varieties of cultivated oat (Avena sativa L.) and four lines of a weed species 'See also Section IV.
102
J. C. O'TOOLE AND W. L. BLAND
FIG. 1. Root growth of 45 bluebunch wheatgrass genotypes after 21 days at 2°C. (Adapted from Harris and Goebel, 1976.)
(A. sterilis L.) which originated from arid zones around the Mediterranean Sea. The genotypes were grown to maturity in hydroponic culture. Periodically the root volume was estimated by water displacement and dry weight of the entire root system was measured at final harvest. Genotypic variation was found for both root volume and dry weight. One A. sterilis L. line, however, not only produced three to nine times more final root dry weight than the other five genotypes, but continued to expand its root system until maturity. Root volume in the other five genotypes stabilized or increased only slightly after flowering. The exceptional line flowered within 6 days earlier or later than the others and doubled its root volume from flowering to maturity. The authors concluded that the sustained and vigorous root growth of the unusual A. sterilis L. line may provide unique genes for future oat cultivars. 5. RiceZ
The root systems of rice (Oryza sutiva L.) ecotypes or landraces adapted to nonflooded upland conditions have frequently been compared with lowland rices adapted to flooded culture. Hasegawa (1963) reported that Japanese upland cultivars rooted deeper than the lowland types. French workers in West Africa (Nicou et al., 1970) found that Asian lowland * See also Section IV.
GENOTYPIC VARIATION IN PLANT ROOT SYSTEMS
103
cultivars had root systems with small-diameter, highly branched roots. In contrast, upland cultivars from Africa and South America had root systems of greater weight with less branching and larger-diameter main axes compared to those of lowland cultivars. As a part of their drought resistance program in the Ivory Coast, Reyniers and Binh (1977) screened 50 cultivars representing a broad range in geographic and hydrologic orgin. Deeprootedness and root density were ascertained from the uptake of the radioisotope 32P, placed at the 95-cm soil depth. Four well-adapted upland rices from West Africa and South America had activity considerably higher than any others. The authors felt that selection of well-adapted parental lines could be accomplished by this method. Mambani and Lal (1983a), working in Nigeria, conducted field experiments in which they analyzed varietal differences in the vertical and lateral distribution of root length density. Three cultivars, an African upland and two Asian rices, were grown under rain-fed and irrigated conditions. The African cultivar rooted to the depth of measurement, 25 cm, and had greater root mass at any depth or lateral distance from the plant row than did the more shallowly rooted Asian lines. During a period of water stress, soil water matric potential patterns resembled rooting patterns and the African upland rice maintained a higher (wetter) leaf water potential than the Asian cultivars. Relative yield (rain-fed versus irrigated) was also greatest for the African genotype. In a related study Mambani and La1 (1983b) investigated root development, plant water status, and grain yield of 10 rice cultivars. The rices were grown at three locations on a “toposequence” where depth to the water table varied from 100 to 15 cm. The soil water regime did not affect total root dry weight, but depth of root system penetration varied across the water table depth gradient. Genotypes differed in their ability to maintain yield under the more adverse water regimes. As in the previous study, plant water status and relative yield varied greatly among the 10 genotypes and was directly related to rooting depth and density measurements. Due to the fluid nature of flooded and puddled rice soils, traditional soil coring techniques are not always applicable. As an alternative, O’Toole and Soemartono (1981) investigated the association between root parameters and the force required to pull a single rice plant vertically from a flooded paddy. They used correlation and path coefficient analysis to illustrate the relationship of root pulling force and root number, length, weight, branching, root:shoot ratio, and thick root number. Root pulling force is now routinely used in the drought resistance screening program at the International Rice Research Institute (IRRI), Philippines. A significant amount of research, primarily concerned with genotypic variation in rice root systems, has been conducted over the past 19 years at the IRRI. It is beyond the scope of this discussion to adequately review that
104
J. C. O’TOOLE AND W. L. BLAND
work. However, the reader is referred to Section IV and the annual reports of the IRRI beginning in 1967. 6. Rye
Because rye (Secale cereale L.) is sown in the fall, it must establish itself when seasonal air and soil temperature are declining. Sullivan and Pfahler (1986) measured primary root growth rate (millimeters/day) at 12, 20, and 28”C, as well as a root “bluntness ratio” of several rye genotypes at the seedling stage. The bluntness ratio was a measure of the shape of the primary root tip, which the authors thought may be related to root system ability to penetrate compacted soil. The genotypes were diploid populations and colchicine-derived tetraploids from five adapted cultivars. For root growth rate, they demonstrated significant effects of ploidy level, genotypic background (original cultivar), and temperature as well as numerous significant interactions. The tetraploid group exhibited a higher root elongation rate than the diploid group at all three temperatures. However, it is significant in light of the study objective that the greatest difference between diploid and tetraploid growth rate was at the lowest temperature (12 “C). In addition, the difference between diploids and tetraploids was significantly influenced by the genotypic background. The tetraploids had the lowest bluntness ratios, indicative of a blunter, less tapered root tip. The results indicated the existence of substantial genotypic variation for rye seedling root growth rate over a wide range of temperatures, both within and among cultivars at both ploidy levels.
7. Sorghum As sorghum (Sorghum bicolor L. Moench.) improvement has progressed, many studies of the root system have been undertaken to determine which root parameters are responsible for the crop’s ecological adaptation and drought-resistant character. Due to a naturally occurring drought beginning 17 days after sowing Bhan et al. (1973) observed four resistant and four susceptible sorghum cultivars among entries in a field trial. The cultivars varied greatly in degree of leaf wilting at noon after a 23-day rainless period. Excavation of the root systems revealed that drought-resistant genotypes had more vertically oriented roots, averaging about 15 cm deeper than the susceptible group. The resistant genotypes also had a higher root number, weights, and root:shoot dry weight ratios. Nour and Weibel(l978) compared the root characteristics of 10 sorghum cultivars previously ranked for drought resistance. Three-week-old seedlings were sampled for root weight, length, volume, and dry weight
GENOTYPIC VARIATION IN PLANT ROOT SYSTEMS
105
root:shoot ratio. In general, these root parameters were positively correlated with drought resistance. Damodar et al. (1978) described the rooting pattern at the vegetative and flowering stages for 11 cultivars of field-grown sorghum in India. They used 32Puptake to assess the vertical and horizontal activity of the root systems to the 40-cm depth. The vertical distribution of relative root activity in the vegetative stage varied among genotypes and was related to field performance during previous drought years. In the United States, scientists in Texas and Nebraska have been especially active in research on sorghum root systems. Because these studies were often components of programs concerned with sorghum improvement for drought resistance, diverse genotypes were frequently included in the experiments. In a notably detailed study, Blum et al. (1977a) investigated the effect of maturity genes on the rate of root appearance and growth with isogenic lines. In addition, Blum et al. (1977b) studied the manifestation of heterosis and kinetics of root system development. Hybrids exhibited greater seminal root length, root volume, and rate of adventitious root growth than did their parents. In a succeeding study Jordan et al. (1979) used hydroponic culture to assess the degree of genotypic variation for root characteristics and shoot:root relationships. Their experiment involved 30 sorghum genotypes including several standard parental lines used for hybrid seed production and exotic lines from the USDA-Texas Agricultural Experiment Station Sorghum Conversion Program. The standard lines differed only slightly in shoot:root relations, while lines from the Sorghum Conversion Program exhibited significant variation. Ratios of shoot:root dry weight, root length of nodal axes:root volume, and leaf area:root length of nodal axes ranged from 0.5 to 1.7 among the 30 genotypes. Jordan and Miller (1980) presented frequency distributions of the same ratios for additional sorghum genotypes. They also reported the extension of the hydroponic studies to field-level investigation of 10 genotypes using the minirhizotron technique during the 1977 crop season. Observation of shallow-rooted cultivars in hydroponics coincided with field results. However, hydroponics failed to identify cultivars which rooted deeply in the field in response to drought (Jordan and Monk, 1980). Similar observation of stress-induced root growth patterns were obtained in Nebraska by Retta et al. (1982, 1983) in field studies of root length and soil water extraction patterns among sorghum hybrid genotypes underirrigated and stressed soil water treatments. Genotypes varied in their ability to express stress-induced deep root growth patterns. Wright and Smith (1983) conducted field experiments under rain-fed conditions to determine the morphological and physiological mechanisms responsible for yield differences under water stress conditions in two sorghum genotypes. The higher-yielding genotype E-57 had higher root
106
J. C. O’TOOLE AND W. L. BLAND
length densities than cultivar TX-671, especially below 80 cm, yet used less water than TX-671 prior to booting, due to differences in timing of leaf area development. After the booting stage, when the upper soil profile (0-60cm) water supply was nearly exhausted, water use was reversed, with E-57 extracting more water from below the 80-cm depth. Blum (1972) and Passioura (1972) have shown for sorghum and wheat, respectively, that, given a limited water supply, water use efficiency is increased when the proportion of the total soil water extraction which occurred preanthesis was reduced. In the study of Wright and Smith (1983), E-57 maintained higher water use rates than TX-671 during the reproductive period, resulting in less floral abortion and greater yield in E-57.
8.
Wheat3
Genotypic variation in the root system of wheat (Triticum aestivum L.) has been the subject of numerous studies. Objectives have varied from investigations of genotypic effects on coordinated growth pattern of root and shoot to lodging resistance and the resistance to water flow in the root vascular tissue. Genotypic variation for the degree of horizontal versus vertical orientation of nodal roots has been identified in corn (Irwin et al., 1985) and sorghum (Bhan et al., 1973) and associated with adaptation to cultural practices and drought resistance, respectively. Pinthus (1967) screened wheat cultivars for this gross root system trait. Comparison of lodging scores of 10 wheat genotypes with the spreading angle of the root system indicated that the roots of lodging-resistant varieties had greater horizontal spread than did susceptible genotypes. In Australia Derera et al. (1969) grew diverse wheat cultivars in soil-filled polyethylene tubes ‘in a field environment. Significant genotypic variation was found for maximum depth and rate of penetration, total number of seminal and nodal roots, root weight, and length per unit root weight. Early maturity was the major factor in drought tolerance in their study. In spite of the differences in root development observed, the authors suggested that there was little room for improvement in the exploitation of the soil volume. As in other cereals, the movement of injected radioisotopes has been used to evaluate the rooting pattern of wheat. Katyal and Subbiah (1971) injected 32Pinto wheat plants and after 5 days measured the resultant activity in roots recovered from soil cores. Genotypes varied significantly in vertical and lateral distribution of root activity. The authors speculated that more deeply rooted cultivars would have greater drought resistance and lines with less lateral proliferation could be planted at higher densities.
’ See also Section IV.
GENOTYPIC VARIATION IN PLANT ROOT SYSTEMS
107
A comprehensive root study program was started in 1965 at the Rothamsted Experiment Station, England, to evaluate root systems of modern, semidrawf wheats (Welbank et al., 1973; Lupton et al., 1974). Differences in rooting between the semidrawfs and traditional taller varieties were small; the modern lines may have been more extensively rooted deep in the profile. Welbank et al. (1973) provided a thorough account of methods used. Mackey (1973) reviewed the developmental nature of the seminal and nodal root systems of wheat and presented detailed analyses of shoot and root parameters from spring and winter wheat genotypes and monosomics of a spring wheat. Mackey (1973) argued that a higher number of seminal roots may be a desirable configurations as seminal roots are low in weight per unit length and should penetrate deeply in the soil. This hypothesis led to extensive screening for seed size and associated seminal root number (Mackey, 1980). A positive correlation was found between seed weight and seminal root number in Triticum and Aegilops spp. One wheat cultivar, however, departed appreciably from the others, indicating that it may be possible to combine small seed size with a relatively high (three to five) number of seminal roots. Variation in seminal root xylem vessel diameter in wheat was investigated by Cornish (1981) because of its implication for resistance to water flow (axial root resistance). The five wheat genotypes studied differed significantly in effective xylem radius of the largest xylem vessel, ranging from 34 to 45 pm.
C. DICOTYLEDONS Recent studies of genotypic variation in root parameters of dicotyledonous crop species are summarized in Table 11. The table briefly indicates the type of root parameter(s) studied, summarizes proposed value in crop improvement, and identifies the authors. Due to space limitations, not all studies listed in Table I1 are discussed in the following section. 1. AIfalfa4
Studies of genotypic variation in root parameters of alfalfa (Medicago sativa L.) have addressed tolerance to acid soil, response to soil moisture, and variant anatomy and morphology. Simpson et al. (1979) exposed four alfalfa cutivars to varying treatments of lime application to an acidic subsoil and a range of soil water conditions. the genotypes differed in root fineness (m/g), root length at the 60-85-cm depth, and dry weight shoot:root ratios. ‘See also Section IV
Table Il
A Summary of Recent Reports on Genotypic Variation in Crop Plant Root Systems of Dicotyledon Specid Crop Alfalfa (Medicago safiva L.)
Bean (Phmeolus vulgaris L.)
Root system parameter
Adaptation to cultural practices
Simpson ef al. (1979)
Fibrous vs. taproot habit Root bark development Root length Root system morphology
Winter hardiness Nitrogen and carbohydrate storage Drought avoidance Exploitation of soil volume Lodging resistance
McIntosh and Miller (1980) Brick and Barnes (1982) Carter ef ul. (1982) Barnes (1983) Stoffella ef al. (1979a)
Improved water and nutrient uptake, lodging resistance Vigorous root system Morphogenetic effects of nitrogen and temperature Tolerance to flooding
Stoffella ef al. (1979b, 1981)
Fibrous vs. taproot habit
meter, uprooting resistance Dry weight Root length and branching
Cotton
(Gossypiurn hirsufurnL.)
cowpea [ Vigna unguiculafa (L.1 WalP.1
Reference
Root fineness, length, and weight
Basal and taproot weight and dia-
Chickpea (Cicer ariefinurn L.)
Purpose or value
Several root system components Dry weight, total root length, rates of extension Root:shoot ratios, taproot length, and lateral root number Root length and weight Vascular bundle arrangement, lateral root number and length Total and taproot dry weight, location of laterals Taproot and lateral root dimensions Vertical growth rate
Fawole eta/. (1982) Geisler and Krutzfeldt (1983)
Kahn ef ul. (1985)
Soil water extraction, drought resistance
Vincent and Gregory (1987)
Drought resistance and water use efficiency seedling vigor Decreased axial resistance, drought resistance Drought resistance
Quisenberry ef al. (1981)
Adaptedness to specific soil conditions Exploit soil water at depth
Eissa ef al. (1983) McMichael ef ul. (1985) Cook (1985) Sawyer ef al. (1984) Robertson ef al. (1985)
Measurement of taproot, lateral, and adventitious roots
Descriptive morphology
Kahn and Stoffella (1987)
Root:shoot ratios, root length, and dry weights
Descriptive morphology
Ali-Khan and Snoad (1977)
Primary and lateral root length, volume, and dry weight
Descriptive morphology, effect of growth media
Veitenheimer (1981); Veitenheimer and Gritton
Root weight and length
Establish genetic variation
Jensen (1985)
Peanut (Arachb hypogaea L.)
Taproot length and root number at 1 m depth
Ketring et 01. (1982)
Pepper (Capsicum annuum L.)
Shoot:root dry weight ratio
Water and nutrient uptake deep in the soil profde Lodging resistance
safflower (Carthamus tinctorius L.)
Root length density, taproot vs. fibrous system Root distribution and activity
Breaking up and de-watering heavy soils Variety specific agronomy
Harrigan and Barrs (1984)
Root weight, surface area, length
Water-nutrient uptake
Development rate, dry weight, and pattern Root length density Taproot elongation rate Taproot elongation rate Spatial and temporal development of root length density Vertical root density Angle of lateral root growth Nodule fresh weight, nitrogen content, response to temperature
Adaptation of cultivars to soil environment Irrigation protocol Increased water extraction Sod water extraction Comparison of determinant and indeterminant genotypes Improved crop adaptation Root system configuration Cultivar x Rhizobium strain interactions
Pea (Pisum sativum L. and P.fulvum)
(1984)
Soybean (Glycine may L.)
Kahn (1985)
Sathe et al. (1984) Raper and Barber (1970) Mitchell and Russell (1971) Burch et al. (1978) Taylor ef al. (1978) Kaspar et al. (1978, 1984) Sanders and Brown (1979) Boyer et al. (1980) Kaspar el al. (1981) Munevar and Wollum (1982) (continued)
Table n (Continued) Crop
Root system parameter
Root system permeability
Stone and Taylor (1983) Smucker (1985) Stone (1985) Hernandez and Orioli (1985)
Root fresh and dry weight
Lodging resistance
Aycock and McKee (197
