Advances in agronomy volume 50

VOLUME
50

d:.


Advisory Board
Martin Alexander

Eugene J. Kamprath

Cornell University

North Carolina State University

Kenneth J. Frey

Larry P. Wilding

Iowa State University

Texas A&M University

Prepared in cooperation with the
American Society of Agronomy Monographs Committee
M. A. Tabatabai, Chairman
D. M. Kral
S. E. Lingle
R. J. Luxmoore
W. T. Frankenberger, Jr.

S. H. Anderson
P. S. Baenziger

G. A. Peterson

S. R. Yates


D V A N C E S IN

Agronomy
VOLUME
50
Edited by

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

ACADEMIC PRESS, INC.
A Division of Harcourt Brace & Company
San Diego New York Boston London Sydney Tokyo Toronto


This book is printed on acid-free paper. @

Copyright 0 1993 by ACADEMIC PRESS, INC.
All Rights Reserved.
No part of this publication may be reproduced or transmitted in any form or by

any means, electronic or mechanical, including photocopy, recording, or any
information storage and retrieval system, without permission in writing from the
publisher.

Academic Press, Inc.
1250 Sixth Avenue, San Diego. California 92101-431 1
United Kingdom Edition published by

Academic Press Limited
24-28 Oval Road, London NW 1 7DX

International Standard Serial Number: 0065-2 1 I3
International Standard Book Number: 0- 12-000750-9

PRINTED IN THE UNITED STATES OF AMERICA
93949596919X

BB

9

8

7

6

5

4

3

2

I


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

vii
ix

AGRONOMIC
IMPROVEMENT IN OILSEED
BRASSICAS

1.

I1 .

I11.

IV .
V.

R . K . Downey and S. R . Rirnmer
Introduction .............................................
Improving Yield ..........................................
Improving Resistance to Pests ...............................
Future Prospects..........................................
Summary and Conclusions ..................................
References ..............................................

1

10
24
39
49

so

POPULATION
DIVERSITY
GROUPINGS
OF
SOYBEANBRADYRHIZOBIA
Jeffry J . Fuhrrnann

1. Introduction .............................................
I1 . Genotypic Groupings .....................................
Ill . Phenotypic Groupings .....................................
IV . Summary of Phenotypic and Genotypic Relationships ............
V . Taxonomic Srarus of Bradyrhizobium japonirum .................
VI . Concluding Remarks ......................................

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

67
68
69
93
93
95
96

CROPRESPONSESTO CHLORIDE
Paul E . Fixen
I . Introduction .............................................
I1 . Chloride in Plants ........................................
Ill . Yield and Quality Responses to Chloride ......................
IV . Chloride Sources, Losses, and Application .....................
V Predicting Crop Response to Chloride ........................

VI . Summary and Future Research Needs .........................
References ..............................................
V

107
108
12.5
133

135
141
143

CONTENTS

vi

REDOXCHEMISTRY
OF SOILS
Richmond J . Bartlett and Bruce R .James
I . Introduction .............................................
I1 . Nature of the Electron .....................................
I11 . Derivation of Thermodynamic Relationships for Electron Activity
in Soils .................................................
IV . Kinetic Derivation of Thermodynamic Parameters for Redox ......
V . Uses of pe - pH Thermodynamic Information. . . . . . . . . . . . . . . . . . .
VI . Uses of pe - pH Diagrams ...................................
Reduction Status of Soils ............
VII . Measurement of Oxidation .
Free
Radicals
in
Redox
Processes
............................
VIII .
IX . Manganeseandlron .......................................
X . Soil Chromium Cycle ......................................
XI . Photochemical Redox Transformations in Soil and Water . . . . . . . . .
XI1. Humic Substances ........................................
XI11. Wetland and Paddy Properties and Processes ...................

XIV . Empirical Methods for Characterizing Soil Redox . . . . . . . . . . . . . . .
References ..............................................

152
153

155
158
160
165
172
176
178
187
188
190
195
198
205

PLANTNUTRIENT
SULFURIN THE TROPICS
AND SUBTROPICS

N. S. Pasricha and R. L . Fox
1. Introduction .............................................

I1. Extent of Sulfur Deficiency .................................
Ill . Forms of Sulfur in Soil .....................................
IV . Sulfur Cycling in the Tropics ...............................

V . Effects of Acid Rain .......................................
VI . Sulfur in Irrigation Waters .................................
VII . Sulfate Retention in Soil ...................................
VIII . Diagnosis of Sulfur Needs ..................................
IX . Critical soil Solution Concentration ..........................
X . Crop Responses ..........................................
XI . Sulfur Fertilization and Crop Quality .........................
XI1. Sulfur Interactions with Other Elements .......................
XI11. Summary and Conclusions ..................................
References ..............................................

210
211
215
217
223
226
227
237
241
246
252
256
257
260

INDEX

271

.................................................


Contributors
Numbers in parentheses indicare the pages on which the authors’ contriburions begin.

R I C H M O N D J. B A R T L E T T (1 5 l ) , Department of Plant and Soil Science,
University of Vermont, Burlington, Vermont 05405
R. K . D O W N E Y (l), Agriculture Canada Research Station, Saskatoon, Saskatchewan, Canada S 7 N OX2
P A U L E. FIXEN (107), Potash 6 Phosphate Institute, Brookings, South Dakota 57006
R. L. F O X (209), Department of Agronomy and Soil Science, University of
Hawaii at Manoa, Honolulu, Hawaii 96822
JEFFRY J. F U I I R M A N N (67), Department of Plant and Soil Sciences, University of Delaware, Newark, Delaware 1971 7
B R U C E R. JAMES (1 5 l ) , Department OfAgronomy, Univerdy of Maryland,
College Park, Maryland 20742
N. S. PASRICHA (209), Department of Soils, Punjab Agricultural University,
Ludhiana, India
S. R. RIMMER ( l ) , Department ofplant Science, University o f Manitoba,
Winnipeg, Manitoba, Canada R 3T 2 N 2

vii


This Page Intentionally Left Blank


Preface
Volume 50 includes state-of-the-art reviews written by recognized experts
on several topics of interest to crop and soil scientists. The first chapter
discusses advances in agronomic improvement in oilseed brassicas. These

cruciferous crops are cultivated throughout the world as vegetable crops for
human consumption, as condiments and spices for improved flavor of
human diets, and as fodder crops for livestock feeding. However, the
largest cultivation of these crops is for edible vegetable oil production. The
first chapter also reviews the world socioeconomic importance of the
oilseed brassicas, ways to improve yields and resistance to pests, and future
improvement of oilseed brassicas through molecular genetics and other
biotechnological means.
The second chapter presents a comprehensive overview of population
groupings of soybean bradyrhizobia, including discussions on genotypic
groupings; phenotypic groupings include serology, intrinsic antibiotic resistance, uptake hydrogenase, dissimilatory nitrate reduction, rhizobitoxine, surface polysaccharides, protein profiles, rhizobiophage typing, plant
growth regulating substances, and other phenotypes; a summary of phenotypic and genotypic relationships, and the taxonomic status of Bradyrhizo-

bium japonicum.
The third chapter is a comprehensive review of crop responses to chloride. Topics covered are aspects of chloride in crops, including biochemical
functions, osmoregulatory functions, disease suppression, crop development, and interaction with other nutrients, plus yield and quality responses
of various crops to chloride, chloride sources, losses, and application, and
ways to predict crop response to chloride.
The fourth chapter presents a thorough treatment of redox chemistry in
soils, a topic of immense interest to soil and environmental scientists, and
which Advances in Agronomy has not reviewed in many years. Discussions
on the nature of the electron, derivation of thermodynamic parameters for
redox, use of pe- pH diagrams, measurement of oxidation - reduction
status of soils, free radicals in redox processes, manganese and iron, the soil
chromium cycle, photochemical redox transformations in soils and waters,
humic substances, wetland and paddy properties and processes, and empirical methods for characterizing soil redox are included in this review.
The fifth chapter is concerned with plant nutrient sulfur in the tropics
and subtropics. Topics reviewed include the extent of sulfur deficiency in
these areas, sulfur cycling in the tropics, effects of acid rain, sulfur in
ix

X

PREFACE

irrigation waters, sulfate retention in soil, diagnosis of sulfur needs, critical
soil solution concentrations, crop responses, sulfur fertilization and crop
quality, and interactions of sulfur with other elements.
Many thanks to the authors for their excellent chapters.
DONALD
L. SPARKS


AGRONOMIC
IMPROVEMENTIN OILSEED
BRASSICAS
R. K. Downey' and S. R. Rimmer2
'Agriculture Canada Research Station,
Saskatoon, Saskatchewan, Canada S7N OX2
'Depanmenr of Plant Science,
Universiry of Manitoba,
Winnipeg, Manitoba, Canada R 3 T 2N2

1. Introduction

I!.
111.

IV.

V.

A. World Socioeconomic lmportance of the Oilseed Brassicas
B. Brossira Oilseed Species
Improving Yield
A. Seed Yield
B. Oil and Protein Yield
Improving Resistance to Pests
A. Diseases
B. Development of Herbicide-Tolerant Cultivars
Future Prospects
A. Interspecific Hybridization
B. Improvements Based on Biotechnologies
C. Uses of DNA Markers
Summary and Conclusions
References

I. INTRODUCTION
Brassica and other closely related cruciferous crops are widely cultivated
throughout the world as vegetable crops for human consumption, as condiments and spices for improved flavor of human diets, and as fodder
crops for livestock feeding. However, the largest cultivation of these crops
is for edible vegetable oil production. In recent years, a number of monographs and reviews (Tsunoda et al., 1980; Downey, 1983; Stefansson,
Adwnrtr rn A ~ w n a n y Yo/
,
$0
Copyrighi 0 1993 by Academic Press, Inc. All nghts of reproducuon in MY form reserved.

1

2

R. K. D O W N E Y A N D S. R. RIMMER

1983; Scarisbrick and Daniels, 1986; Downey and Robbelen, 1989) have
dealt in detail with many aspects of oilseed Brussicu improvement, especially those which relate to improvements in fatty acid composition and
the reduction in levels of glucosinolates in the residual meal. Substantive
changes in the quality of seed oil and meal composition have resulted in
dramatic increases in areas of production in Canada and western Europe
(see below). Unfortunately, this has also resulted in a rather narrow germ
plasm base of cultivated oilseed brassicas, especially in Brussicu nupus L.
Emphasis in plant breeding has consequently shifted from quality improvement toward increasing seed yield, incorporating resistance to diseases and pests, and improving tolerance to stress. This review focuses on
recent developments and current and future trends for agronomic improvements in oilseed Brussicu crops.

A. WORLD
SOCIOECONOMIC
IMPORTANCEOF THE OILSEED
BRASSICAS
Historically, human consumption of vegetable oil obtained from Brussicu spp. was primarily concentrated in asiatic countries, predominantly in
the northern Indian subcontinent and in China. The cultivation in these
countries of oilseed types of Brussicu rupu L. (syn. Brussicu cumpestris L.)
and Brussicu junceu (L.) Czern. dates back to approximately 1500 BC
(Prakash, 1980), and these areas today are still major producers and consumers of Brussicu vegetable oils.
Since the second world war, a dramatic increase in Brussicu oilseed
production has occurred worldwide. In Canada and in Europe this was
associated with seed quality improvements through plant breeding involving the modification of the fatty acid composition (elimination of erucic
acid) and the reduction of glucosinolate content in the residual meal. The
large production increase in Europe was also related to economic support
from the Common Agricultural Policy of the European Economic Community (EEC). Thus, in the 1948- 1952 period, 70% of a world total

oilseed Brussicu production of 2.8 million tonnes was produced in Asia,
but by 1984, Canada (20%) and Europe (35%) produced more than half the
total world production of 15.9 million tonnes, with the Indian subcontinent ( 18%) and China (25%) producing the balance (Bunting, 1986). Total
world production values of oilseeds, edible vegetable oils, and residual
protein meals for the years 1985-1989 are given in Table I. Oilseed
brassicas account for approximately 10% of total world oilseed production
and 14- 15% of the total edible vegetable oil production. Production by the
primary producing regions of oilseed brassicas is shown in Table 11. Total
world production is now in excess of 20 million tonnes annually.


3

AGRONOMIC IMPROVEMENT IN OILSEED BRASSICAS
Table I
World Production of Oilseeds, Edible Vegetnble Oils, and
Derived Protein Meals, 1985 - 1989a

Millions of t o n n e produced by years
Commodity

1985- I986

1986- 1987

1975- 19Mb

1988- 1989'

Oilseeds

Soybean
Cottonseed
Peanuts
Sunflowerseed
Rapeseed
Flaxseed
Coconut
Palm kernel

97.03
30.63
19.94
19.56
18.57
2.36
5.32
2.56

97.92
27.13
20.44
19.25
19.46
2.69
4.72
2.60

103.17
3 1.05
19.72

20.5 1
22.97
2.28
4.24
2.67

93.13
32.24
21.56
20.96
21.72
1.75
4.56
2.89

195.57

194.21

206.61

198.81

13.85
3.47
2.94
6.65
6.19
1.63
3.3 I

1.11
8.17

15.19
3.05
3.10
6.57
6.83
I .56
2.95
I .09
8.09

15.20
3.46
2.85
7.20
7.65
I .90
2.59
8.53

14.85
3.62
3.35
7.57
7.27
1.41
2.76
I .26

9.36

47.32

48.43

50.56

5 1.45

6 1.07

1.89
1.33

67.12
9.83
4.42
7.54
11.09
I .20
1.72
I .32

67.37
11.17
4.01
8.13
12.51
1.14

1S O
1.41

65.54
11.62
4.77
8.59
11.80
0.99
1.60
I .49

98.6 1

104.24

107.30

106.40

Total
Edible vegetable oils
Soybean
Cottonseed
Peanuts
Sunflowerseed
Rapeseed
Olive
Coconut
Palm kernel

Palm
Total
Protein meals
Soybean
Cottonseed
Peanuts
Sunflowerseed
Rapeseed
Flaxseed
Coconut
Palm kernel
Total

11.10

4.22
7.66
10.19
1.15

From United States Department of Agriculture, 1988.
Preliminary estimates.
'Forecast estimates.
a

1.18


R. K. DOWNEY AND S. R. RIMMER

4

Table 11
Production of Oilseed Brassicas by Main Producing Countries/Regions, 1982- 1989'
Millions of tonnes by years
Average
Country or region

1982/1983- 1986/1987

1987- 1988'

1988- 1989'

India
China
Canada
EEC
Europe (excluding EEC)
Other

2.64
5.13
3.1 I
3.18
I .70
1.06

3.10
6.61

3.85
5.95
2.16
1.31

3.50
5.04
4.24
5.3 I
2.18
1.45

16.82

22.98

21.72

Total

From United States Department of Agriculture, 1988.

'Preliminary estimates.
Forecast estimates.

B. Brcusica OILSEED
SPECIES
Four species of Brassica have been widely cultivated as oilseed crops,
Brassica carinata Braun, B. rapa, B. juncea, and B. napus. Where conditions are appropriate, namely cool temperate climates with good moisture
availability, winter forms of B. napus are preferred and are the most

productive. Most of the land area cultivated to oilseed brassicas in Europe
and China is sown to winter oilseed rape. However, as latitude or altitude
increases, the winter form of B. nupus is supplanted by the summer form of
B. napus or the winter or summer form of B. rapa. In Canada, cultivation
consists of approximately equal amounts of the summer types of these two
species.
Brassica juncea is well-adapted to drier conditions and is relatively fast
maturing. On the Indian subcontinent B. juncea is the dominant species
grown, although large areas are also sown to B. rapa types (toria and
sarsons) (Prakash, 1980). In these climates, with hot dry summers, nonvernalization types of oilseed brassicas are cultivated in the cool moist winter
season. Brassica juncea is also grown in many parts of China outside of the
Yangtse/Yellow river flood plains (Stinson et a!., 1982). In western Canada
B. juncea is grown as a crop for condiment on some 8 1,000 ha but has
strong potential as an oilseed crop for this region (Woods et al., 1991).
Brassica carinata may perform well under long season growing conditions.


AGRONOMIC IMPROVEMENT IN OILSEED BRASSICAS

5

Its distribution presently is largely confined to North East Africa, principally Ethiopia. Clearly, these oilseed crops are well adapted to many
different parts of the world.
1. Genomic Relationships

The genomic relationships among the four oilseed Brassica species are
well known [see Mizushima ( 1980), Olsson and Ellerstrom ( I 980), and
Downey and Robbelen (1989)l. Our modem understanding of these relationships was initiated by Morinaga and co-workers (Morinaga, 1934),
who provided cytological evidence to show that Brassicu nigra (n = 8; B),
Brassica oleracea (n = 9; C), and B. rapa (n = 10; A) are primary species

and that 8. carinuta ( n = 17; BC), B. junceu (n = 18; AB), and B. napus
(n = 19; AC) are amphidiploids resulting from crosses between corresponding pairs of the primary species. These relationships were later confirmed by U (1935), who succeeded in the artificial synthesis of B. napus
from crosses between the diploid species B. rapa and B. oleracea. Synthesis
of B. juncea and B. curinata has subsequently been accomplished by the
interspecific hybridization between B. nigra and B. rapa or B. oleracea [see
Downey et al. (1975) and Olsson and Ellerstrom (198O)J.
Understanding the relationship among these Brassicu species has enabled plant breeders to create synthetic amphidiploids and to transfer
useful agronomic characteristics from species to species through interspecific hybridization. No cultivars have as yet been released as a direct result
of artificial reconstitution of a species through interspecific crosses, although some desirable characteristics have been successfully transferred
from one species to another through artificially synthesized amphidiploids
that function as a “bridge.” For instance, the first double-low (low erucic
acid content in the oil and low glucosinolate content in the meal) strains of
turnip rape were developed from interspecific crosses among turnip rape
(B. rupa), rape (B. napus), and oriental mustard (B. junceu) (Downey et al.,
1975). Similarly, the development of low-glucosinolate B. juncea involved
interspecific hybridization of B. rapa and B. junceu (Love et al., 1990).The
transfer of resistance to blackleg disease from B. junceu to B. nupus (Roy,
1984) is another example.
In China and Japan, interspecific crosses between B. rupa and B. napus
have often been used to transfer characteristics such as early maturity,
cytoplasmic male sterility, self-incompatibility, and yellow seed coat, from
the former to the latter, and to broaden the genetic basis of B. nupus
through genome substitution (Liu, 1985).


6

R. K. DOWNEY AND S. R. R I M M E R

2. Plant and Seed Description

Brassica rapa (AA, 2n = 20) is one of the primary diploid species and
occurs wild in the high plateaus of the Irano-Turanian regon (Hedge,
1976), where it is well adapted to the cool, short season environment of this
area. This species has a high relative growth rate under cool temperatures
and can produce abundant seed. Both spring and winter forms are cultivated and the most cold-hardy cultivars of the oilseed brassicas occur
within this species. This species is considered to be of the seed vernalization
type. Full clasping of the upper leaves around the stem, the positioning of
the terminal buds below newly opened flowers, and a high ratio of beak to
pod length are characteristic of this species. Both dark- and yellow-seeded
types occur.
Brassica carinata (BBCC, 2n = 34) is the amphidiploid between B.
oleracea and B. nigra. It shows a slow steady growth, probably derived
from the B. oleracea genome. Leaves, which are generally waxy and light
green in color, are attached to the stem with a true petiole. Though seeds
are predominantly dark, some types have yellow seed. Cultivation is limited to the Ethiopian plateau and adjacent areas of east Africa. It is
currently under evaluation and shows promise agronomically in many
other parts of the world.
Brassica juncea (AABB, 2n = 36) is the amphidiploid of B. rapa and B.
nigra. It has a high leaf area ratio and a high relative growth rate, comparable to B. rapa (Sasahara and Tsunoda, 197 1 ). Asia, especially China, is rich
in variations of cultivated forms of this species. It is grown widely for oil in
the north Indian subcontinent and in various regions of China (Xinjiang
Autonomous Region, Szechuan). This species is also characterized by
having leaves with true petioles. Leaves vary considerably in shape but are
generally of a dark green coloration. Seeds may be dark or yellow and the
“bold” types from India have a large seed size. It has considerable potential
as an oilseed crop in many other parts of the world.
Brassica napus (AACC, 2n = 38) is the amphidiploid of B. rapa and B.
oleracea. The existence of a wild form of B. napus is uncertain; if it does
exist it will probably be found in the European-Mediterranean region

( McNaughton, 1976). Olsson (1 960) suggested that the amphidiploid B.
napus (genome AACC) might have arisen at different locations by hybridization of various forms of B. oleracea (genome CC) and B. rapa (genome
AA). Leaves of this species lack a true petiole as does B. rapa, but only
partial clasping of the stem occurs. Seeds are dark, generally larger than
those of B. rapa, and no natural yellow-seeded types are known. Development of a yellow seed form that is known to be associated with a thinner
seed coat (and thus reduced fiber content in the meal) is one of the current


AGRONOMIC IMPROVEMENT IN OILSEED BRASSICAS

7

objectives in many breeding programs. Production of oil in Europe from B.
nupus occurred as early as the thirteenth century, when it was used primarily as lamp oil (Appelqvist, 1972).
3. Mode of Pollination

Brussicu rupu is primarily a self-incompatible species, as are the other
diploid brassicas, although some types of B. rupu, e.g., yellow sarson, are
self-compatible. The self-incompatibility (SI) in cruciferous species is of
the homomorphic sporophytic type determined by a single S locus. About
50-60 alleles are known at the S locus in B. oleruceu (Nasrallah and
Nasrallah, 1989). The allelic interactions at the S locus are dominant,
codominant, or recessive depending on the alleles involved. This system
ensures that B. rupu is normally 100% outbreeding and consequently
breeding methodologies for this species are designed to take advantage of
this natural heterozygosity.
The amphidiploids, B. nupus, B. junceu, and B. curinutu, are normally
self-compatible species, though S alleles from B. rupu have been introduced into some genotypes of B. nupus in order to develop SI-based F,
hybrids. Such hybrids have recently been registered for commercial production in Canada. Generally, self-pollination occurs readily in the amphidiploid species and selfed seed may easily be obtained by enclosing the
flowering racemes in bags. Under field conditions, outcrossing, from pollination due to insects and wind, has been estimated to range from 5 to

15% (Huhn and Rakow, 1979)to about 27 to 35% (Olsson, 1952; Persson,
1956) in winter rape, 22 to 36% in summer rape (Persson, 1956; Rakow
and Woods, 1987), and 19% for B. junceu (Rakow and Woods, 1987).
4. Oilseed Quality Improvements

At present, cultivars of two species (B. nupus and B. rupu) have been
developed with both low-erucic and low-glucosinolate (double low, or
canola) quality, and these are now widely grown commercially. In North
America the term “canola” has been coined to describe cultivars that meet
specific requirements for erucic acid in the extracted seed oil (less than 2%
erucic acid as a percentage of total fatty acids) and aliphatic glucosinolate
content in the residual meal (less than 30 pmol g-I). [For a discussion of
the development of low-erucic acid cultivars of B. nupus and B. rupu and
the genetics of the inheritance of erucic acid in these species, see Stefansson
( 1983).] It is likely that canolaquality cultivars of B. junceu and perhaps B.
curinutu will be developed in the near future, and, if this occurs, it will
significantly influence the choice of oilseed Brussicu species in some areas.


Table I11
Fatty Acid Composition of Oilseed Brassicu Crops and Other Common Vegetable Oils

Fatty acid composition (%)"
CY2

Species, crop,
cultivar, and type
Brussicu nupus (rape)
Victor winter
Jet Neuf winter

Hero summer
Westar summer
Stellar summer
Brussicu rupu (turnip rape)
Duro winter
Yellow sarson
Echo summer
Tobin summer
Brussicu junceu (mustard)
Indian origin
Cutlass

Ref!

14:O

16:O

16:l

18:O

18:l

18:2

18:3

20:O

20:l

22:O

22:l

24:O

0.3
0.4
0.2
0. I
tr

0.8
1.4

9.9
56.4
12.9
57.7
59.1

13.5
24.2
12.2
20.8
28.9

9.8

10.5
9.0

0.6
0.7
0.8
0.6
0.5

6.8
1.2
7.5
1.4
1.4

0.7
0.3
0.8
0.3
0.4

53.6
0.0
50.2
0.5
0.1

0.0
0.0
0.3

0.3
0.2

13.4
12.0
18.8
24.0

9.1
8.2
8.9
10.3

0.7
0.9

9.6
6.2
12.0
1.0

0.2
0.0
0.0
0.1

49.8
55.5
23.5
0.3

0.0
0.0

1.2

12.9
13.1
32.5
58.6

1.2
1.2

8.0
17.2

16.4
21.4

11.4
14.1

6.4
11.4

1.2
0.4

46.2

25.8

1

0.0

2
3
4
4

0.0
0.0
0.0
0.0

3.0
4.9
2.8
3.6
4.1

I
2
2
2

0.0
0.0
0.0

0.0

2.0
1.8
2.5
3.8

0.2
0.2
0.2
0.1

5
6

0.0
tr

2.5
3.3

0.3
0.3

1.O

1.6
I .4
I .o

0.9

1.O

11.5

3.3

0.6
0.6
1.2
0.1

0.0
0.0
0.1

0.2

24:l
1 .o

0.0
1.2
0.0
0.0
1.1

1.2
0.0

0.0
I .9
1.7


2km 1

2

tr

3.6

0.4

2.0

45.0

33.9

11.8

0.7

1.5

0.3

0.1

0.2

0.5

6

tr

3.2

0.2

0.9

9.8

16.2

13.9

0.7

7.5

0.7

41.6

0.6

2.0

7

0.0

15.3

0.0

4.2

23.6

48.2

8.7

0.0

0.0

0.0

0.0

0.0

0.0

8

0.1

5.8

0.1

5.2

16.0

71.5

0.2

0.2

0.1

0.7

0.0

0.1

0.0

9

9

9.2
6.7

0.0
0.0

0.0
0.0

3.1
4.3

57.2
71.4

23.4

0.0
0.0

1.4
1.6

1.4
1.0

2.6
2.7

0.0

11.1

0.0

1.8
1.3

0.0
0.0

10

tr

11.5

0.0

2.2

26.6

58.7

0.8

0.2

0.0

0.0

0.0

0.0

0.0

11

0.0
1.0

7.6
23.4

0.0
0.8

2.0
2.5

10.8
17.9

79.6
54.2

0.0
0.0

0.0
0.0

0.0
0.0

0.0
0.0

0.0
0.0

0.0
0.0

0.0
0.0

Brassicu carinafa
Ethiopian mustard
Glycine may (soybean)
Group 1 variety
Helianrhus annuus (sunflower)
Peredovik
Arachis hypoguea (peanut)
Virginia Bunch

Cook Jumbo
Zea mays (corn)
United States sources
Curlhamus finctorius (safflower)

us10
Gossypium hirsutum (conon)

12

Fatty acids represented by carbon chain length and number of double bonds; tr, trace amounts.
References: (1) Appelqvist (1969), (2) Downey (1983), (3) R.Scarth (unpublished), (4) Scarth ef al. (l988), (5) Appelqvist (1970). (6) R. K. Downey
(unpublished), (7) Hymowitz ef al. (1972), (8) Earle ef ul.(1968). (9) Worthington and Hammons (1971), (10) Beadle ef al. (1965), ( 1 1) Knowles (1968).
and ( 1 2) Anderson and Worthington (197 1).
a


10

R. K. DOWNEY AND S. R. RIMMER

In developed countries, the production of edible oil from oilseed brassicas
is now obtained exclusively from low-erucic acid cultivars and this trend is
expected to continue for production in developing countries.
Low-erucic acid strains of B. juncea have been recently obtained. These
strains were obtained by crossing plants from an accession with intermediate levels of erucic acid content and screening for low erucic acid in the
F, progeny using the half-seed technique (Kirk and Oram, 1981). The
development of low-glucosinolate B. junceu required interspecific hybridization of B. rupu and B. juncea (Love el ul., 1990) in order to transfer the
Bronowski block for aliphatic glucosinolate synthesis from a B. rupu line
producing low glucosinolate to a strain of B. juncea that produced 3-butenyl glucosinolate but no 2-propenyl (allyl) glucosinolate.

Continued improvement for oilseed quality includes development of
strains with modified fatty acid composition. These include development
of strains with lower levels of linolenic acid, higher levels of linoleic acid,
high oleic acid levels, and other modifications (see Table 111 for a comparison of the fatty acid compositions of oilseed brassicas and other vegetable
oilseed crops). A cultivar with low levels of linolenic acid (<4%) has been
developed and registered in Canada (Scarth el ul., 1988). The oil with
reduced levels of linolenic acid from this cultivar has been shown to have a
prolonged cooking and shelf life (Kay, 1988). Oilseed brassicas with high
(>60%) levels of erucic acid would also be desirable for industrial purposes. Breeders have found it extremely difficult to achieve levels of erucic
acid higher than 55% of the total oil content. Because of the inability of
acyl transferases to insert erucoyl moieties in the 2-position of the triglyceride there may be a natural upper limit of 66% erucic acid obtainable in
Brussicu spp. (Taylor el ul., 1992).

11. IMPROVING YIELD

A. SEEDYIELD
1. Yield Components and Breeding Methods

Although improved nutritional quality of the oil and meal has been a
major breeding objective of Brussicu oilseed breeders, yield of seed, oil, and
protein must all be maintained and improved if these crops are to remain
competitive. Because seed yield is probably the most difficult and costly
trait to measure accurately, numerous attempts have been made to identify
the most important yield component(s). Positive relationships have fre-


AGRONOMIC IMPROVEMENT IN OILSEED BRASSICAS

11

quently been cited between the seed yield and the numbers of pods per
plant and per main raceme, as well as the numbers of seeds per pod and
seed weight per pod (Thompson, 1983; Shabana et al., 1990). In examining
yield and yield components of 10 European winter rape cultivars over a
3-year period, Grosse et al. (1992) concluded that high yields could be
attained from different combinations of three yield components- seeds
per pod, number of pods, and individual seed weight. However, as noted
by Thurling (1974b) and others, compensation among the various yield
components in response to environment occurs to such an extent in oilseed
brassicas that few breeders practice selection for one or even a few yield
components.
Observations on the contribution of various yield components to the
observed heterosis in hand-crossed hybrids have confirmed earlier findings.
Heterosis effects varied for each yield component depending on the environmental and/or genotypic effect when number of pods per plant, number of seeds per pod, single seed weight, and plant density were considered
(Lefort-Buson and Dattke, 1982; Schuster et al., 1985; Uon, 1989; Schuler
et al., 1992).
Given the importance of the oilseed brassicas, very few studies have been
undertaken to determine the physiological basis for increased yield. Thurling ( 1974a), who studied three Australian cultivars, found that correlations of total dry matter and yield were positive and highly significant
(r = 0.70). Allen and Morgan (1975) reported that leaf area index at first
flower was correlated to yield and concluded that a greater photosynthetic
source at flowering and after first flower would result in higher yield.
Campbell and Kondra (1978), studying single plants of three B. napus
cultivars, found that seed yield was significantly correlated with total dry
matter production (r = 0.2 1 to 0.52) per plant. Thurling (199 1) concluded
from a series of experiments with cultivars and breeding lines that early
flowering and maximum light penetration of the crop canopy are required
to maximize seed yield. The importance of light penetration of the crop
canopy is supported by the findings of Mendham et al. ( 1991). Comparing
the seed yield of an apetalous strain to a closely related petalous variety,
they attributed the higher yield of the apetalous strain to the 30% greater

solar radiation transmitted through the apetalous canopy. Although these
studies provide the breeder with some insight into the plant type that may
be highly productive, the measurement of such parameters is normally not
as efficient or effective in oilseed brassicas as the total measurement of
yield.
In conventional B. napus and B. juncea breeding programs for yield,
various forms of the pedigree system are employed [see Thompson (1983),
Downey and Rakow ( 1987),and Downey and Robbelen ( I989)]. However,


R. K. DOWNEY AND S. R. RIMMER

12

Table IV
Average Relative Yield of Winter Rape Parental Lines and Cultivars Compared to
Performanceof Syn-1 Synthetics and Seed Mixtures of P a r e d
seed yield as percentage of parents
Average of
parents

Syn- I
synthetic

seed
mixtures

Reference

100

100
100
100

I04
I14
I06
108

97
I05
106
104

Grabiec and Krzymanski ( 1 984)
Schuster and Friedt ( 1985)
E o n ( 1987)
Lkon and Diepenbock (1987)

'After Becker (1988).
the parameters of these systems differ from pedigree cereal programs in two
important respects. First, the oilseed brassica crops have a high multiplication rate per generation (- 1000: I), and second, the plant-to-plant outcrossing rate is much higher, ranging from 5 to 36% (see Section I,B,3).
Thus replicated progeny testing can begin as early as the F, and a certain
level of heterosis from the initial cross can be captured and retained in
subsequent generations. In a comparison of winter rape selection techniques, Sauermann (1989) found that in winter B. napus visual selection in
the F, for yield was superior to a random line selection, but the highest
yielding lines were identified by measuring yields of single-row F, progenies in a three-replicate test at one location or by testing sublines in the F4
with one replicate at each of three locations.
Because of the potential for significant levels of heterosis for yield in the
oilseed brassica species, the degree of natural interplant crossing, and the

absence of a highly efficient system of pollen control, synthetics have been
suggested as a means of capturing part of the available heterosis. Becker
( 1988) compared the performance of experimental synthetics of winter B.
napus to their parent lines or cultivars and noted that the synthetics yielded
some 4 to 14% more seed (Table IV). He postulated that even higher levels
of heterosis could be captured if the parents were selected on the basis of
their combining ability. On the other hand, in three of the four experiments, sowing mixed seed of the parents in the same drill run also resulted
in yield increases, which in some instances approached or equaled the yield
of the synthetics (Table IV). U o n (199 I ) found that cultivar mixtures and
Syn- 1s displayed greater yield stability than their corresponding F, hybrids
or any of the individual cultivars, suggesting that heterozygosity and het-


AGRONOMIC IMPROVEMENT IN OILSEED BRASSICAS

13

erogeneity of both cultivar mixtures and synthetics have a positive effect
on yield stability.
In B. rupu, where self-incompatibility ensures a high degree of heterogeneity, recurrent selection has been the most effective method for increasing
seed yield as well as oil content (Downey and Rakow, 1987). However,
with the numerous agronomic and quality traits that must now be incorporated into any new B. rupa canola cultivar, more than one specialized
recurrent selection program, to be run in parallel with the main recurrent
selection program, may be required. After sufficient progress is made
within each specialized composite, two or more may be combined to create
a new composite cultivar source.
The presence of the natural self-incompatibility (SI) system in most B.
r a p cultivars suggests that synthetic cultivars would be attractive alternatives to F, populations derived by recurrent selection. Falk (1991), using
four parental cultivars of B. rupu, compared the seed yield of the parent
cultivars and all possible F,s with their various two-, three- and four-component Syn- Is and Syn-2s. It was found that the agronomic performance of

the synthetics could rival the single-cross F,s. On average, hybrids yielded
15 to 30% more seed than their parent cultivars over 2 years of testing.
However, the Syn-1 populations averaged only 1% less seed than the F,s in
each testing year, while the average of the Syn-2s in the last year of testing
yielded only 3 and 2 percentage points less than the F,s and Syn-ls,
respectively.
2. Heterosis and F, Hybrids

Although B. nupus is usually classified as a largely self-pollinating species, significant levels of heterosis for yield have been obtained in F,
hybrids of both the spring and the winter forms. Based on the results from
reciprocal top crosses of seven summer rape cultivars with the Canadian
cultivar Regent, Sernyk and Stefansson ( 1983) concluded that it should be
possible to develop hybrid cultivars of summer rape with a commercial
heterosis for yield of about 40%. Also, in spring rape Grant and Beversdorf
( 1985)found high-parent heterosis for seed yield of up to 72%, with specific
combining ability being more important than general combining ability. In
winter rape, several researchers have documented the potential for hybrids,
reporting heterosis for seed yield up to 60-70% (Schuster and Michael,
1976; Lefort-Buson and Dattke, 1982, 1985). Lefort-Buson et al. (1987)
related heterosis to genetic distance in crosses between and within groups
of European and Asiatic B. nupus cultivars and strains. As might be
expected, hybrids between distant groups showed greater heterosis than


14

R. K. DOWNEY AND S. R. RIMMER

within-group hybrids. Additive and dominant genetic variance was more
important for the within-group than the between-group hybrids.

In the self-incompatible B. rapa species, Arunachalam and Bandyopadhyay ( 1984)were also able to relate the magnitude of heterosis exhibited in
the F, to the genetic divergence of the parental phenotype. Singh and
Gupta (1985) outlined a procedure to identify diverse genotypes using 3 1
B. rapa strains tested in 12 different environments. Indian researchers have
reported varying degrees of heterosis for yield within and among Indian
oilseed types of B. rapa (i.e., Toria and brown- and yellow-seeded sarson).
In general, such hybrids have shown positive high parent heterosis for seed
yield with a high proportion showing commercial heterosis. Unfortunately,
many of these observations are based on single plant yields (Prasad and
Singh, 1985; Yadav and Yadava, 1985; Singh and Gupta, 1985) or have
been taken from space-planted, single-row plots (Devarathinam ef al.,
1976; Labana et al.. 1978), and the results reported are often based on only
I year and a single location. In most of the studies, combining-ability
analysis has indicated yield to be primarily under the control of nonadditive gene action. High parent heterosis reported from Indian studies has
been as great as 63% on single plant yields in yellow sarson (Labana et al.,
1978).
In Canada, a natural top cross of a canola breeding line onto the yellow
sarson rapeseed cultivar, R500, yielded 46% more seed than the commercial B. rapa canola cultivar Candle (Hutcheson ef al., 1981). Further
studies over a 2-year period, using R500 as the female parent in crosses
with three Canadian oilseed cultivars and strains, showed high parent
heterosis for seed yield between I6 and 37% (Hutcheson, 1984). Schuler
(1989) tested hand-crossed hybrids between the B. rapa canola cultivar
Tobin and 19 European and Canadian parental strains and cultivars at
four locations over a 2-year period. He reported the highest average commercial heterosis for seed yield to be 64%, the same range as that found by
Labana et al. ( 1978). However, the best canola-quality hybrid yielded only
22% over the commercial cultivar Tobin. Falk ( I99 1) reported that 4 of 12
hybrids from diallel crosses between four B. rapa Canadian and European
cultivars exhibited high parent heterosis for seed yield. The maximum high
parent heterosis found in the 2-year multilocation study was 26%, whereas
the best commercial heterosis for seed yield, between canolaquality parents, was 22%.

Studies of heterosis in B. junceu have for the most part been camed out
on the Indian subcontinent and, as in the B. rapa studies, are largely based
on single plant yields in space-planted plots. Singh (1 973) reported that six
B. juncea hybrids showed an average high parent heterosis of 49% over 2
years of testing. Banga and Labana (1 984) found the range of seed yield