1. business and industrial
2. agriculture and forestry
3. crops and seed

BIOLOGICAL NITROGEN FIXATION FOR SUSTAINABLE AGRICULTURE
Developments in Plant and Soil Sciences
VOLUME 49
The titles published in this series are listed at the end of this volume.
Biological Nitrogen Fixation
for Sustainable Agriculture
Extended versions of papers presented in the Symposium,
Role of Biological Nitrogen Fixation in Sustainable Agriculture
at the 13th Congress of Soil Science, Kyoto, Japan, 1990
Edited by
J. K. LADHA
International Rice Research Institute
Manila, Philippines
T. GEORGE and B. B. BOHLOOL
Nitrogen Fixation by Tropical
Agriculture Legumes
(NifTAL) Project
University of Hawaii,
Honolulu, Hawaii, USA
Reprinted from Plant andSoil, Volume 141 (1992)
Kluwer Academic Publishers
Dordrecht / Boston / London
in cooperation with the
International Rice Research Institute
Library of Congress Cataloging in Publication Data
ISBN 0-7923- 1774-2
Published by Kluwer Academic Publishers,
P.O. Box 17, 3300 AA Dordrecht, The Netherlands.
Kluwer Academic Publishers incorporates
the publishing programmes of Martinus Nijhoff,
Dr W. Junk, D. Reidel, and MTP Press.
Sold and distributed in the U.S.A. and Canada
by Kluwer Academic Publishers,
101 Philip Drive, Norwell, MA 02061, U.S.A.
In all other countries, sold and distributed
by Kluwer Academic Publishers Group,
P.O. Box 322,3300 AH Dordrecht, The Netherlands.
All Rights Reserved
© 1992 Kluwer Academic Publishers
No part of the material protected by this copyright notice
may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying,
recording, or by any information storage and retrieval
system, without written permission from the copyright
owner.
Printed in the Netherlands
Contents
Preface
vii
Dedication
xi
B. B. Bohlool, J. K. Ladha, D. P. Garrity and T. George, Biological nitrogen fixation for sustainable
agriculture: A perspective
1
M. B. Peoples and E. T. Craswell, Biological nitrogen fixation: Investments, expectations and actual
contributions to agriculture
13
P. A. Roger and J. K. Ladha, Biological N2 fixation in wetland rice fields: Estimation and contribution to
nitrogen balance
41
I. Watanabe and C. C. Liu, Improving nitrogen-fixing systems and integrating them into sustainable rice
farming
57
T. George, J. K. Ladha, R. J. Buresh and D. P. Garrity, Managing native and legume-fixed nitrogen in
lowland rice-based cropping systems
69
I. R. Kennedy and Y.T. Tchan, Biological nitrogen fixation in non-leguminous field crops: Recent
advances
93
H. H. Keyser and F. Li, Potential for increasing biological nitrogen fixation in soybean
119
S. F. Ledgard and K. W. Steele, Biological nitrogen fixation in mixed legume/grass pastures
137
K. Fujita, K. G. Ofosu-Budu and S. Ogata, Biological nitrogen fixation in mixed legume-cereal cropping
systems
155
S. K. A. Danso, G. D. Bowen and N. Sanginga, Biological nitrogen fixation in trees in agro-ecosystems
177
J. Ishizuka, Trends in biological nitrogen fixation research and application
197
Plant and Soil 141. vii, 1992.
Preface
Sustainability is defined as “the successful management of resources for agriculture to satisfy changing
human needs while maintaining or enhancing the quality of the environment and conserving resources”
(Technical Advisory Committee of Consultative Group on International Agricultural Research).
Economists view sustainability as the ratio of input to output, taking into account stock depletion. In
agriculture, stocks include soil, water, nonrenewable energy resources and environmental quality.
Industrially produced N fertilizer depletes nonrenewable energy resources and poses human and
environmental hazards. In developing countries, importing N or building fertilizer plants requires scarce
foreign exchange. Atmospheric n is virtually unlimited, but it must be converted to a plant-usable form
through biological or nonbiological processes.
Biological N fixation (BNF) is a vital component of agricultural sustainability. It can be a major N
source in symbiotic systems such as legumes, or a minor source supplementing inorganic N in
nonsymbiotic, nonlegumes such as cereals. Nonsymbiotic BNF is important in replenishing soil N
depleted by crop removal or lost through leaching, volatilization, and denitrification in nonfertilized or
moderately fertilized irrigated lowland rice where soil N is maintained naturally. Even heavily
N-fertilized systems marked by high cropping intensity may experience soil N depletion. In these soils
sustainability depends on symbiotic and nonsymbiotic BNF.
Tremendous progress in BNF biochemistry and genetics has been made in the last 30 years, yet we
are still hoping for breakthroughs in nif gene transfer from legumes to nonlegumes. Much more basic
work is required. In the meantime, we should ask whether basic research is being pursued at the
expense of applied research. Relevant basic research must continue. but of equal or greater importance
is applied farm-level research.
This book is a product of the symposium, Role of Biological Nitrogen Fixation in Sustainable
Agriculture, jointly organized by the International Society of Soil Science and the International Rice
Research Institute at the 13th International Congress of Soil Science, Kyoto, Japan. Papers on BNF in
major crops and cropping systems were augmented by discussions of integrated cropping systems
involving BNF, soil and N management, recycling of crop residues, and multipurpose legumes. Recent
work on promising BNF systems and their potential for exploitation were discussed. There was also a
presentation on donor perspectives on research priorities and funding. This present volume brings
together extended versions of all technical papers presented, and a perspective paper on BNF for Sustainable
Agriculture.
J.K. Ladha
Chairman of the Symposium
B. Ben Bohlool
1940-1991
Plant and Soil 141, xi, 1992.
B. Ben Bohlool
Dedication
The BNF community throughout the world was saddened by the death of Dr. B. Ben Bohlool,
University of Hawaii NifTAL Project Director, on 16 May 1991. This book is dedicated to his memory,
in honor of the many contributions he made to this field.
Ben was born in Tehran, Iran, in 1940. In 1971, he graduated with a Ph.D. in microbial ecology from
the University of Minnesota at Minneapolis under Dr. E.L. Schmidt. His doctoral work focused on the
use of immunofluorescence in the study of the ecology of Rhizobium japonicum in soils. Further
post-doctoral research with Dr. Schmidt resulted in their formulation of the lectin hypotheses. This
pioneering work suggested that lectins were a possible basis for specificity in the legume-rhizobia
symbiosis.
Ben dedicated his research career to understanding this nitrogen fixing symbiosis. In 1976, Ben joined
the faculty of the University of Hawaii. His research there encompassed several major areas including
the ecology of rhizobia in tropical soils, assessing the potential of fast-growing soybean rhizobia from
China, and studying the host regulation of the legume-rhizobia symbiosis. During his tenure at the
University of Hawaii, Ben was the recipient of the Alexander van Humboldt Fellowship, and spent six
months with the molecular genetics group at Max-Planck Institute in Cologne, Germany.
In 1985 Ben was appointed Director and Principal Investigator of the NifTAL (Nitrogen Fixation by
Tropical Agricultural Legumes) Project. As NifTAL’s Director, Ben fostered his vision of the Project
as both a dynamic research institute as well as a responsive resource for promoting BNF in agricultural
development. He led the Project in developing international outreach and training activities, and was
nominated for the World Food Prize for these efforts. Ben also remained active in research, and
supported an international network of researchers in rhizobia1 ecology to study environmental effects
on the occurrence, persistence, and performance of rhizobia in tropical soils.
Ben is remembered by his friends and colleagues as an excellent and creative scientist. He was blessed
with an insatiable curiosity, social gregariousness, and the ability to savor all aspects of life. Besides his
own scientific contributions, Ben’s legacy includes the continuing work of his graduate students, and
that of his colleagues promoting BNF for sustainable agricultural development.
Plant and Soil 141: 1-11, 1992
© 1992 Kluwer Academic Publishers. Printed in the Netherlands.
PLSO SA01
Biological nitrogen fixation for sustainable agriculture: A perspective
2
B. B. BOHLOOL 1 , J. K. LADHA 2, D. P. GARRITY and T. GEORGE 1
1
N i f T A L Project, University of Hawaii, 1000 Holomua Avenue, Paia, H I 96779, USA and
2
International Rice Research Institute ( I R R I ) , P.O. Box 933, Manila, Philippines
Key words: chemical fertilizer, crop production, developing countries, environment, inoculation,
legume, pollution
Table of contents
Abstract
Introduction
The issue of sustainability in agriculture
Economic considerations: The fossil-fuel outlook
Environmental considerations: Health and ecological impacts
Production sustainability: Management of internal resources
Constraints to utilization of nitrogen fixing systems
Environmental constraints
Biological constraints
Methodological constraints
Production-level constraints
Socio-cultural constraints
Conclusion
References
Abstract
The economic and environmental costs of the heavy use of chemical N fertilizers in agriculture are a
global concern. Sustainability considerations mandate that alternatives to N fertilizers must be urgently
sought. Biological nitrogen fixation (BNF), a microbiological process which converts atmospheric
nitrogen into a plant-usable form, offers this alternative. Nitrogen-fixing systems offer an economically
attractive and ecologically sound means of reducing external inputs and improving internal resources.
Symbiotic systems such as that of legumes and Rhizobium can be a major source of N in most cropping
systems and that of Azolla and Anabaena can be of particular value to flooded rice crop. Nitrogen
fixation by associative and free-living microorganisms can also be important. However, scientific and
socio-cultural constraints limit the utilization of BNF systems in agriculture. While several environmental factors that affect BNF have been studied, uncertainties still remain on how organisms respond to a
given situation. In the case of legumes, ecological models that predict the likelihood and the magnitude
of response to rhizobia1 inoculation are now becoming available. Molecular biology has made it possible
to introduce choice attributes into nitrogen-fixing organisms but limited knowledge on how they interact
with the environment makes it difficult to tailor organisms to order. The difficulty in detecting
introduced organisms in the field is still a major obstacle to assessing the success or failure of
2
Bohlool et al.
inoculation. Production-level problems and socio-cultural factors also limit the integration of BNF
systems into actual farming situations. Maximum benefit can be realized only through analysis and
resolution of major constraints to BNF performance in the field and adoption and use of the technology
by farmers.
Introduction
Chemical fertilizers have had a substantial impact on food production in the recent past, and
are today an indispensable part of modern agricultural practices. The Green Revolution of the
past half-century was fueled by technologies
heavily dependent on synthetic fertilizers. Russell et al. (1989) estimate that in 1985, the use of
38.8 million Mt of N fertilizer on cereals globally
resulted in increased world cereal production of
938 million Mt, more than half of the total cereal
production in that year. They defined a relationship between world N fertilizer use on cereals
(X) and mean world cereal yield (Y), between
1956 and 1985, which could be expressed by the
equation Y = 1202 + 13.3X, with R2 = 0.983 emphasizing the major driving force of N fertilizers
in food production today.
There are, on the other hand, vast areas of the
developing world where N fertilizers are neither
available nor affordable. Furthermore, in most
of these countries, removal of N fertilizer subsidies, due to balance of payment problems,
have resulted in higher prices and lower supplies.
Even in wealthier nations, economic and environmental considerations dictate that biological
alternatives which can augment, and in some
cases replace, N fertilizers must be sought. The
process of biological nitrogen fixation (BNF)
offers this alternative. In this paper, we outline
sustainability issues that dictate increased use of
BNF and the constraints to optimal use of BNF
in agriculture.
The issue of sustainability in agriculture
Sustainability is defined as ‘the successful management of resources to satisfy changing human
needs while maintaining or enhancing the quality
of the environment and conserving resources’
(TAC, CGIAR, 1978). Economists measure sustainability as the ratio of output to input, taking
into account stock depletion. Stocks in agriculture include soil, water, nonrenewable energy
resources, and environmental quality.
Modern agriculture is based on maximum output in the short term, with inadequate concern
for input efficiency or stock maintenance
(Odum, 1989). Nitrogen fertilizer ranks first
among the external inputs to maximize output in
agriculture. Input efficiency of N fertilizer is one
of the lowest among the plant nutrients and, in
turn, contributes substantially to environmental
pollution. The continued and unabated use of N
fertilizers would further accelerate depletion of
stocks of non-renewable energy resources used
in fertilizer production. The removal of large
quantities of crop produce from the land additionally depletes soil of its native N reserves. On
the other hand, BNF, a microbiological process
in the biosphere, converts atmospheric dinitrogen into a plant-usable form through the microbial enzyme nitrogenase. Nitrogen input through
BNF can help maintain soil N reserves as well as
substitute for N fertilizer to attain large crop
yields (Peoples and Craswell, 1992).
Another issue which needs urgent attention is
the decline in crop yields under continuous use
of N fertilizers. The trend of the 1970s of increasing crop yields from fertilizer N addition
has recently been slowed down both in the developed (Plucknett and Smith, 1986) and in the
developing countries (Barker and Chapman,
1988; Byrlee, 1987). Odum (1989) indicated that
a 4-fold increase in crop yield during the 1970s
and 80s occurred at an 11-fold increase in fertilizer N use in Georgia, USA. An analysis of
several years’ data on rice yields from experimental stations in Philippines, Indonesia and
Thailand by Pingali et al. (1990) shows trends of
stagnation or decline.
Economic considerations: The fossil-fuel
outlook
The oil crisis of the 1970s and the current Middle
BNF for sustainable agriculture
East problems are constant reminders of the
vulnerability of our fossil fuel-dependent agricultural and food production systems. As the price
of oil increases, so does that of agricultural
inputs, notably of N fertilizers.
Industrial processes for the manufacturing of
nitrogen fertilizer are heavily reliant on fossil
fuel, utilizing a vast proportion of our energy
supplies. Synthesis of compounds of nitrogen
from N2 gas (the Haber-Bosch process) requires
large quantities of hydrogen, usually obtained
from natural gas, in addition to substantial
amounts of energy to establish and maintain the
conditions of high temperature and pressure required for nitrogen to react with hydrogen to
produce ammonia. In addition to the cost of
production, a great deal of energy is also used
for transportation, storage, and application of N
fertilizer, all totalling to near 22,000 kilo calories
of energy per kilogram of fertilizer-N processed,
distributed, and applied (see Evans, 1975).
Capital costs for construction of a medium-size
N fertilizer factory alone exceed $100 million.
Considering that most of the hardware and
know-how for construction of the factory comes
from industrialized nations, it is difficult to promote expansion of the fertilizer manufacturing
industry in developing countries strapped for
hard currency.
Environmental quality considerations: Health
and ecological impacts
The external costs of environmental degradation
and human health far exceed economic concerns.
Nitrate in groundwater is a major health concern
in the corn belt of the U.S., and other intensively cultivated areas. Nitrogen in run-off and surface waters has led to extensive pollution and
eutrophication of rivers and lakes. The gaseous
oxides of nitrogen, derived from N fertilizers,
are highly reactive and pose a threat to the
stability of the ozone layer. Table 1, modified
from Keeney (1982), summarizes the potential
adverse impacts of excessive fertilizer N application.
Production sustainability : Management of
internal resources
Long-term sustainability of agricultural systems
must rely, as much as possible, on use and
effective management of internal resources. Nitrogen-fixing plants offer an economically attractive and ecologically sound means of reducing
external inputs and improving the quality and
quantity of internal resources. Biological nitrogen fixation can be a major source of N in
agriculture when symbiotic N2-fixing systems are
used; the amount of N input is reported to be as
high as 360 kg N ha -1 (Table 2). On the other
hand, the nitrogen contributions from nonsymbiotic (associative and free-living) microorganisms are relatively minor, thus requiring
fertilizer N supplementation.
Among symbiotic N,-fixing systems, nodulated
legumes have been used in cropping systems for
centuries. They can serve a multitude of pur-
Table 1. Potential adverse environmental and health impacts of excessive N use (Keeney, 1982)
Impact
Human health
Methemoglobinemia in infants
Cancer
Respiratory illness
Environmental health
Environment
Eutrophication
Materials and ecosystem damage
Plant toxicity
Excessive plant growth
Stratospheric ozone depletion
3
Causative agents
Excess NO3 and NO2 in waters and food
Nitrosamine illness from NO2, secondary amines
Peroxyacyl nitrates, alkyl nitrates, NO3 aerosols,
NO2, HNO3 vapor in urban atmospheres
Excess NO3 in feed and water
Inorganic and organic N in surface waters
HNO3 aerosols in rainfall
High levels of NO2 in soils
Excess available N
Nitrous oxide from nitrification, denitrification,
stack emissions
4
Bohlool et al.
Table 2. Estimates of dinitrogen fixed by different N2-fixing systems in agriculture
N2-fixing system
N2 fixed
(kgN ha -1 )
Free-living/associative
Rice-blue green algae
Rice-bacterial association
Sugarcane bacterial association
Symbiotic
Rice-Azolla
Legume-Rhizobium
Leucaena leucocephala
Glycine max
Trifolium repens
Sesbaniarostrata
Non-legume- Frankia Casuarina sp.
poses in sustainable agriculture. They are used as
primary sources of food, fuel, fiber and fertilizer,
or, secondarily, to enrich the soil, preserve moisture and prevent soil erosion. They can also be
used for windbreak, ground cover, trellis,
hedgerow and shade, or as a source of resins,
gums, dyes and oils. Some of the most ornamental flowering plants in the tropics are legumes
(see NAS 1979 for more detail).
Some of the nodulated nonlegumes, notably
species of Casuarina, are hardy nitrogen-fixing
plants, which produce high-quality fuelwood on
marginal lands, and have also been used for
stabilizing sand dunes and eroding hillsides, as
well as for reclaiming marshlands affected by
fluctuating brackish/fresh water. The trees are
useful for shade, windbreaks and hedges (for
more detail see NRC, 1984).
Of particular practical value in rice production
systems is the nitrogen-fixing symbiotic water
fern Azolla. Azolla in symbiosis with cyanobacterium Anabaena azollae can fix 2-4 kg N ha"
day-1 (Lumpkin and Plucknett, 1982). Recently
other benefits of Azolla have been recognized: a)
weed suppressor, b) K scavenger from floodwater, c) animal feed, d) fish feed, e) P
scavenger in sewage-treatment plants, and f)
suppressor of ammonia volatilization. (Watanabe
and Liu, 1992).
Associative and free-living microorganisms are
believed to contribute to production sustainability of flooded rice production systems (Roger and
Ladha, 1992). Approximately 50% of the N
requirement of a flooded rice crop is met from
References
10- 80 crop -1
10- 30 crop-1
20-160 crop -1
Roger and Ladha (1992)
Roger and Ladha (1992)
Urquiaga et al. (1989)
20-100 crop -1
Roger and Ladha (1992)
100-300yr -1
0-237crop -1
13-280 crop -1
320-360 crop-1
40- 60yr-1
Danso et al. (1992)
Keyser and Li (1992)
Ledgard and Steel (1992)
Ladha et al. (1990)
Gauthier et al. (1985)
the soil N pool (Bouldin, 1986) which is believed
to be maintained through BNF by associative
and free-living microorganisms (Koyama and
App, 1979). Contributions from non-symbiotic
N2 fixation in upland agriculture is generally not
substantial (Kennedy and Tohan, 1992), although N2 fixation to the order of 160 kg N ha-1
has been reported for sugarcane (Table 2).
Constraints to utilization of nitrogen-fixing
systems
There are still many unknowns in the scientific
understanding of N2 fixation. Research into the
basic mechanisms of the process is an important
goal for improving N2 fixation in the future
(Ishizuka, 1992), but few, if any, of these unknowns are constraining implementation of the
existing BNF technologies. So much of the existing knowledge is not being used, especially in
developing countries, that it is essential to devote major efforts towards adoption of what is
already known.
There are major technical, socio-economic and
human-resource obstacles to fuller implementation of BNF technologies in cropping systems.
The first can be addressed through comprehensive programs of basic and applied research. The
latter two through education, training and private-enterprise development.
Full utilization of, and maximal benefit from,
BNF systems can be realized only through analysis and resolution of major constraints to their
BNF for sustainable agriculture
optimal performance in the field, and their adoption and use by the farmers. These can be classified broadly into environmental, biological,
methodological, production-level and sociocultural constraints.
Environmental constraints
The revolution in modern molecular biology has
made it possible to alter the genetic composition
of living organisms and give them attributes desirable to man. However, our lack of understanding of how organisms interact with and
within their environments has made it difficult to
tailor organisms to order. The question is no
longer ‘can we genetically engineer beneficial
organisms?’, but ‘what should we engineer them
for?’. The challenge is to match the genetic
potential of the living systems to the controlling
parameters of the environment.
An essential consideration in establishment of
nitrogen-fixing systems is how they would respond to the adversities of the soil environment.
A thorough understanding of the ecology of the
various nitrogen-fixing systems is crucial to the
successful application and acceptance of BNF
technologies for sustainable productivity.
The ‘Law of the Minimum’ states that plant
productivity can be defined by a single limiting
factor in the system. Amelioration of the limiting
factor will increase productivity until another
essential growth factor in the system becomes
limiting. Therefore, no amount of biologically or
chemically fixed nitrogen can increase production as long as some other factor is limiting
growth. Several environmental factors that affect
the performance of legume/rhizobia symbiosis
(Alexander, 1985; Atkins, 1986) and actinorhizal
(Frankia) symbiosis (Tjepkema et al., 1986; Torrey, 1978) have been studied. Gibson (1977),
Munns and Francop (1982), Freire (1984) and
Ladha et al. (1992) have reviewed the soil constraints to symbiotic performance. Major considerations affecting either the microbe, the host, or
their symbiotic interaction include soil acidity
(Munns, 1977), other acid-related factors including aluminum and manganese toxicity, and calcium deficiency (Singer and Munns, 1987), phosphorus (Almendras and Bottomley 1988; Cassman et al., 1981; Helyar and Munns, 1975;
5
Keyser and Munns, 1979; Leung and Bottomley,
1987; Munns, 1979; Singleton et al., 1985), calcium (Beck and Munns, 1984, 1985), salinity
(Singleton and Bohlool, 1983) and flooding
(Ladha et al., 1992). Symbiotic activity within a
plant community is also conditioned by the
amount of N mineralized from organic sources
(George et al., 1988).
Watanabe and Liu (1992) have discussed the
diverse environmental constraints which limit the
optimum performance of azolla. They include P
deficiency, sensitivity to drought and high temperature.
One of the major obstacles to successful establishment and effective performance of introduced N2-fixing systems is competition from native organisms. This has been extensively demonstrated for legume inoculants, and is likely to
be a serious problem for other exotic organisms
in new environments.
The complexities involved in competition for
nodulation of legumes are numerous, as noted in
the recent review by Dowling and Broughton
(1986). Environmental influences such as soil
temperature (Weber and Miller, 1972), or additions of P and K (Almendras and Bottomley,
1988) can alter competition patterns.
The population size oc naturalized rhizobia
affects the likelihood of inoculant establishment
(Bohlool and Schmidt, 1973; Weaver and Frederick, 1974 a,b) and the magnitude of a legume
crop’s response to applied rhizobia (Singleton
and Tavares, 1986). The yield and nodulation of
a promiscuous host cowpea ( Vigna unguiculata
(L) Walp) can be increased with inoculation less
frequently due in part to the population size and
competitive ability of indigenous Bradyrhizobium sp. in tropical soils (Danso and
Owiredu, 1988). However, the belief that tropical legumes do not respond to rhizobial inoculation is a misconception based on faulty assumptions and very little data (Singleton et al., 1991).
Soybean, which has more specific rhizobial requirements, responds to inoculation with Bradyrhizobium rhizobia when first introduced into an
area (Thies et al., 1991a). As the introduced
rhizobia become naturalized, the yield response
to inoculation is no longer observed in subsequent crops (Dunigan et al., 1984; Ellis et al.,
1984). Thies et al. (1991b) have recently used
6
Bohlool et al.
indices of native rhizobial population size and
the nitrogen status of the soil to develop ecological models for predicting the likelihood, and the
magnitude, of legume response to rhizobial
inoculation.
Biological constraints
What ultimately determines the success of a BNF
practice is the genetic potential of the organisms
and how that interacts with components of the
environment. Figure 1 summarizes the data of
Thies et al. (1991a) with legume inoculation. It
compares the relative yield of uninoculated and
inoculated legumes with that of N treatment
(approximately 1000 kg of N per hectare, split
100 kg N per week), for a total of 29 combinations of 9 legume species at 5 diverse sites. The
N treatment is taken as the 100% full yield
potential for each species. The difference between uninoculated and inoculated plants represents opportunities for increasing yield of
legumes with existing knowledge and inoculation
technology. The gap between yields of inoculated and fertilized plants represents the window
of opportunity for improving symbiosis by gen-
etic manipulation, environmental management,
or both.
A variety of biological factors may influence
the expression of BNF in all nitrogen-fixing systems. In symbiotic associations, both partners
are subject to biological constraints, such as
disease and predation which can directly or indirectly affect the amount of N fixed, as well as
the quantity made available to other components
of the cropping system.
In general, and especially so in legumes, the
amount of nitrogen fixed is directly related to the
growth potential of the host in a particular system. When growth is limited, for example by
disease, nitrogen fixation will be reduced accordingly.
Methodological constraints
The difficulty in identifying the N2-fixing organisms introduced into the field is one of the major
obstacles to assessing the success or failure of the
technology. This is relatively easier with symbiotic legumes, since in most systems serological
methods can be used to successfully identify and
monitor the introduced rhizobia (for reviews, see
Fig. 1. Comparative benefits from rhizobia1 inoculation and N application. Results summarized from Thies et al. (1991a) with 9
legume species at five sites.
BNF for sustainable agriculture
Bohlool and Schmidt, 1980; Bohlool, 1987).
With the advent of DNA probes for genetic
analysis at the strain level (Holben et al., 1988),
it is likely that this barrier will be removed for
most BNF systems in the near future.
Inoculant technology is well-developed for
legume inoculants, but still in its infancy for
actinorhizal and other BNF systems (Torrey,
1978). Although there have been some successes
in growing Frankia in pure culture (Callaham et
al., 1978; Diem et al., 1982), it still remains a
problem to produce inoculum on a large scale.
Even for legumes, the scale of production, the
availability of suitable carrier material, preservation of germ plasm, and shelf-life of the finished
product are serious constraints to use of inoculants, especially in developing countries.
A major methodological constraint to BNF
use is the lack of reliable techniques for measuring nitrogen fixation in the field. This is particularly true for perennial pasture legumes and trees
(Danso et al., 1992). When non-fixing genetic
isolines are available, the N difference method
can be a reliable estimator of the contribution of
BNF. Unfortunately, however, only for soybeans, and to some extent for peas, have well
characterized nod-nod isolines been reported.
There are reports of nod-nod alfalfa, peanuts
and chick peas, but these have not been studied
as extensively. Several of the chapters in this
volume cover the state of the art in methodologies for BNF measurement and discuss the
strengths and limitations of various measurement
techniques.
Production-level constraints
The development of an efficient BNF system
carries no certainty that it will be successfully
incorporated into prevailing farming systems.
There are several field-level production constraints that limit the incorporation of BNF systems into farming systems.
Cereal crops dominate the agricultural landscape in both the temperate and tropical worlds.
Fitting legumes into these systems is a task of
challenging complexity, notwithstanding their
significant BNF benefits. More niches for
legumes must be found within the major cropping systems, and numerous agronomic limita-
7
tions overcome, in order to exploit BNF’s direct
and indirect contributions.
Wide-scale production of leguminous crops
has been particularly constrained in the humid
tropics (Khanna-Chopra and Sinha, 1989). High
precipitation and humidity during the rainy season induce insect and disease epidemics that
have so far proven difficult to control, thus creating yield instability. Grain legume pods are also
hygrophilic, and the seed deteriorates very rapidly in humid conditions (Shanmugasundaram,
1988). Special harvest precautions are needed
compared to the cereals. Also, most legume
cultivars have indeterminant-plant characteristics. They tend to have a low grain-to-biomass
ratio when produced in rainy, cloudy weather
patterns.
Due to these constraints, most legume production is undertaken in the wet-dry transition
period at the end of the monsoon season, or
during the dry-wet transition at the onset of the
rains. These seasons are subject to great rainfall
variability, resulting in alternating moisture
stress or excess (George et al., 1992). Crop
establishment is more difficult in the fluctuating
moisture conditions, and grain yield and BNF
contributions tend to be unreliable.
In the humid uplands, where Ultisols and Oxisols predominate on all three tropical continents, severe acidity and phosphorus deficiency
are characteristic problems (Sanchez, 1980). Unless they can be alleviated through soil amelioration with external nutrient sources, production of
conventional grain legumes is not favored. The
contribution of BNF on these soils is also hampered by the fact that nitrogen is often not the
primary limiting nutrient.
The approach of developing unconventional
BNF systems for the domain of strongly acid
soils is now receiving prominent attention. For
sustaining intensive production, particular emphasis is directed to BNF in alley-cropping systems using leguminous trees or forages (Kang,
1990). For fallow-rotation systems, which occupy
vast areas, managed fallows of leguminous cover
crops are being developed. Until recently, little
research attention had been devoted to these
practices.
In the wetland rice production systems of the
humid tropics, both leguminous green manures
8
Bohlool et al.
and grain legumes have traditionally played a
critically important BNF role (King, 1911). The
contribution of both declined in recent decades
as fertilizer-N use became widespread. Research
is developing new species to better exploit the
available cropping-system niches. One striking
example is the introduction of stem-nodulating
green manure species (Ladha et al., 1992) that
are better adapted to waterlogged conditions.
However, the typical agronomic constraints common to all green manures remain as real barriers
to adoption (Garrity, 1990). For most species,
sustainable seed production is primary among
them (Garrity et al., 1990), but successfully establishing the crop, and incorporating it into the
soil during land preparation for rice, must be
done with an absolutely minimal cost (Pradhan,
1988). Very little attention has been given to
these critical agronomic issues.
Although Azolla is a potential BNF system for
wetlands, the technology suffers from many
farm-level constraints which limit any widespread use by farmers. Major constraints are:
difficulty in maintaining and distributing inocula
round the year and susceptibility to insects and
diseases.
In the semi-arid tropics legumes occupy a
more important role than in the humid tropics
(Willey, 1979). They are commonly intercropped
with cereals (Fujita et al., 1992): pigeon peas
with sorghum in India, cowpeas with sorghum or
maize in West Africa. The superiority of legumecereal intercropping is well established for these
long-duration annual crop associations. However, the long-term viability of these systems is
still uncertain, unless they can be more effectively subjected to sustained improvements in agronomic efficiency.
In the temperate zone, legume oilseeds, particularly soybeans (Keyser and Li, 1992) and
forages (Ledgard and Steele, 1992) are of major
importance. Intensive agronomic research has
assured their products as valued commodities.
However, their indirect role as BNF suppliers to
subsequent crops in rotations, although readily
perceived by farmers, is discounted in practice.
This occurs because of the lack of dependable
quantitative measures of the actual N contribution, season by season, at the field level. Consequently, farmers usually ignore the contribution
of the prior legume N when they calculate N
fertilizer applications to the following cereal.
Unless better tools to estimate legume N contributions are perfected, legumes may be found to
unintentionally accelerate nitrate leaching into
the groundwater, thus creating negative environmental consequences.
Socio-cultural constraints
It is important to emphasize that the constraints
to fuller adoption of BNF technologies are not
solely scientific but include cultural, educational,
economic and political factors. A successful
BNF-based program, therefore, must involve, in
addition to scientific research, efforts in training,
education, outreach and technical assistance.
Evaluations of socio-economical constraints are
needed to publicize the benefits of BNF technology and provide advance warnings about potential difficulties, facilitating their removal or circumvention.
Most farmers in developing countries do not
know that legumes fix nitrogen in their root
nodules. Yet traditional and modern farming
systems of the tropics almost invariably include
legumes. The legume cultivation results from
recognition by farmers over many centuries that
legumes are valuable components in farming systems rather than from intentional exploitation of
biological nitrogen fixation per se.
The cost of inoculants is not usually a constraint to their use by farmers who outlay capital
for seed. Inoculant cost will seldom exceed 1%
of the seed cost. For subsistence farmers who do
not ordinarily purchase seed off the farm, the
capital outlay for inoculant, albeit small, may be
a disincentive to the use of inoculants. Cost
becomes a more important consideration with
granular forms of inoculant in which the rate of
application is much higher than with seedapplied inoculant.
BNF technology is a difficult technology to
deliver by normal extension mechanisms. Thus a
lack of illustrative and explanatory pamphlets
and other aids, both for extension agents and
farmers with whom they have contact, is also a
constraint to implementation of BNF technology
at the farm level.
Furthermore, few of the senior administrators
and decision makers who determine agricultural
policy in developing countries are fully aware of
the opportunities for legume-based BNF technol-
BNF for sustainable agriculture
ogy in the agriculture sector of their countries.
Most policy-makers are aware of some of the
attributes of legumes, but relatively few appreciate the role played by biological nitrogen
fixation in legumes. Among those few an even
smaller proportion recognizes that it may be
essential to employ specific technologies to ensure that nitrogen fixation occurs at all, let alone
at a maximal rate. Thus, there is a need for
educational material, especially developed for
this clientele group, bringing to their attention
the real need to adapt currently available technology to the particular circumstances in which it
is to be employed in their country.
The task of training and educating persons to
deliver BNF information and material has also
been made difficult by the critical shortage of
specialists with knowledge and interest in practical and applied aspects of the technology. The
mass exodus of microbiologists and plant biologists from applied BNF to molecular biology and
genetics has created a serious void in programs
addressing BNF needs of sustainable agriculture
systems in developing countries.
National and international subsidies for nitrogen fertilizers, in both developed and developing
countries, have been a disincentive for farmers
to use BNF products. However, as national
debts rise and subsidies are phased out, BNF
becomes more attractive and acceptable. Donor
agencies should do more to educate and train
in-country personnel to be able to respond to
increasing demands for BNF, as sustainability
and environmental issues demand biological alternatives to the use of industrial fertilizers.
Crop subsidies and support programs are, in
some countries, limited to cereal crops. The
United States is a major example. This situation
distorts the entire agricultural system. It provides
a de-facto incentive to abandon legume-cereal
crop rotations in favor of cereal mono-cropping.
In countries where virtually no differential crop
support is practiced, for example in Australia,
exploitation of cereal-legume systems is a dominant feature of agriculture.
9
become increasingly dependent on off-farm supplies, which requires cash and may not always be
available on time. The harmful effects on the
environment of heavy use of N fertilizer are
becoming more evident. Further, the fossil fuels
which are used in the production of N fertilizer
are becoming scarcer and more expensive. At
the same time, the demand for food is going up
as populations increase. Therefore, there is a
great need to search for all possible avenues to
improve biological nitrogen fixation and its use
by farmers. There is a need of sustainable farming which maintains soil fertility by using renewable resources easily and cheaply available on
the farm. Examples are: rotating cereal crops
with legumes, recycling manure and other organic wastes, and using chemical fertilizers moderately, but efficiently.
A quantitative understanding of the ecological
factors that control the fate and performance of
BNF systems in the field is essential for promotion and successful adoption of these technologies. We can no longer afford the time and
expense
of trial-and-error experimentation,
which has been the hallmark of experimental
agricultural sciences in the past. Future research
will have to be directed towards approaches that
generate information that is transferable from
one site to another. Scientists from biophysical
and socio-economical fields should work together
to identify and eliminate farmers’ limitations in
using BNF technology.
Recently, research efforts on basic aspects of
biochemistry and genetics of biological nitrogen
fixation have increased considerably. Yet, we are
still hoping for breakthroughs in nif gene transfer
from legumes to nonlegumes. Much more basic
work will still be required. However, in the
meantime, we should ask whether basic research
is pursued at the expense of applied research.
Relevant basic research must continue, but of
equal or greater importance is applied, farmlevel research.
References
Conclusion
Fertilizer nitrogen has become a major input in
crop production around the world. Farmers have
Alexander M 1985 Ecological constraints on nitrogen fixation
in agricultural ecosystems. Adv. Microb. Ecol. 8, 163-168.
Almendras A S and Bottomley P J 1988 Cation and phosphate influences on the nodulating characteristics of in-
10
Bohlool et al.
digenous serogroups of Rhizobium trifolii on soil grown
Trifolium subteraneum L. Soil Biol. Biochem. 20, 345-352.
Atkins C A 1986 The legume/Rhizobium symbiosis: Limitations to maximizing nitrogen fixation. Outl. Agric. 15,
128-134.
Barker R and Chapman D 1988 The Economics of Sustainable Agricultural Systems in developing Countries. Mimeo,
Cornell University, Ithaca, NY.
Beck D P and Munns D N 1984 Phosphate nutrition of
Rhizobium spp. Appl. Environ. Microbiol. 47, 278-282.
Beck D P and Munns D N 1985 Effect of calcium on the
phosphorus nutrition of Rhizobium meliloti. Soil Sci. Soc.
Am. J. 49, 334-337.
Bohlool B B 1987 Fluorescence methods for study of
Rhizobium in culture and in situ. In Symbiotic Nitrogen
Fixation Technology. Ed. G H Elkan. pp. 127-147. Marcel
Dekker, Inc., New York.
Bohlool B B and Schmidt E L 1973 Persistence and competitive aspects of Rhizobium japonicum observed in soil by
immunofluorescence microscopy. Soil Sci. Soc. Am. Proc.
37, 561-564.
Bohlool B B and Schmidt E L 1980 The immunofluorescence
approach in microbial ecology. In Advances in Ecology.
Ed. M Alexander. pp 203-235. Vol. 4. Plenum Press, New
York, London.
Bouldin D R 1986 The chemistry and biology of flooded soils
in relation to the nitrogen economy in rice fields. In
Nitrogen Economy of Flooded Rice Soils. Eds. S K De
Datta and W H Patrick, pp 1-14. Martinus Nijhoff Publishers, Dordrecht, The Netherlands.
Byrlee D 1987 Maintaining the momentum in post-green
revolution agriculture: A micro-level perspective from
Asia. MSU International Development Paper No. 10,
Michigan State University, East Lansing, MI.
Callaham D, Del Tredici P and Torrey J G 1978 Isolation
and cultivation in vitro of the actinomycete causing root
nodulation in Comptonia. Science 199, 899-902.
Cassman K G, Munns D N and Beck D P 1981 Growth of
Rhizobium strains at low concentrations of phosphate. Soil
Sci. SOC. Am. J. 45, 520-523.
Danso S K A and Owiredu J D 1988 Competitiveness of
introduced and indigenous cowpea Bradyrhizobium strains
for nodule formation on cowpeas Vigna unguiculatu (L.)
(Walp.) in three soils. Soil Biol. Biochem. 20, 305-310.
Danso S K, Bowen G D and Sanginga N 1992 Biological
nitrogen fixation in trees in agroecosystems. Plant and Soil
141, 177-196.
Diem H G, Gauthier D and Dommergues Y R 1982 Isolation
of Frankia from nodules of Casuarina equisetifolia. Can. J.
Microbiol. 28, 526-530.
Dowling D N and Broughton W J 1986 Competition for
nodulation of legumes. Annu. Rev. Microbiol. 40, 131157.
Dunigan E P, Bollick P K, Huchinson R L, Hicks P M,
Azubrecher F C, School S G and Mowers R P 1984
Introduction and survival of an inoculant strain of
Rhizobium japonicum in soil. Agron. J. 76, 463-466.
Ellis W R, Ham G E and Schmidt E L 1984 Persistence and
recovery of Rhizobium japonicum in a field. Agron. J. 76,
573-576.
Evans H J 1975 Enhancing Biological Nitrogen Fixation.
National Science Foundation, Dir. Biological and Medical
Sciences. Washington, DC 20550.
Freire J R 1984 Important limiting factors in soil for the
Rhizobium-legume symbiosis. In BNF Ecology, Technology and Physiology. Ed. M Alexander. pp 51-74. Plenum
Press, New York, London.
Fujita K, Ofosu-Budu K G and Ogata S 1992 Biological
nitrogen fixation in mixed legume-cereal cropping systems.
Plant and Soil 141, 155-176.
Garrity D P 1990 Agronomic .research on biofertilizers at
IRRI: Overcoming systems-level constraints. In Proc. of
the Nat. Symp. on Bio- and Organic Fertilizers. Univ. of
the Philippines, Los Baños, Philippines. pp 309-320.
Garrity, D P, Roberto T B, Crecensia C B, Pye Tin and Riaz
M 1990 Indigofera tinctoria: Farmer-proven green manure
for rainfed ricelands. In Proc. of the Nat. Symp. on Bioand Organic Fertilizers. Univ. of the Philippines, Los
Baños, Philippines. pp 261-284.
Gauthier D. Diem H G, Dommergues Y R and Ganry F
1985 Assessment of N2 fixation by Casuarina equisetifolia
inoculated with Frankia ORS021001 using 15N methods.
Soil Biol. Biochem. 17, 375-379.
George T, Ladha J K, Buresh R J and Garrity D P 1992
Managing native and legume-fixed nitrogen in lowland
rice-based cropping systems. Plant and Soil 141, 69-91.
George T, Singleton P W and Bohlool B B 1988 Yield, soil
nitrogen uptake, and nitrogen fixation by soybean from
four maturity groups grown at three elevations. Agron. J.
80, 63-567.
Gibson A H 1977 The influence of environmental and managerial practices in the legume-Rhizobium symbiosis. In A
Treatise on Dinitrogen Fixation. Section IV: Agronomy
and Ecology. Eds. R W F Hardy and A H Gibson. pp
393-450. Wiley, New York.
Helyar K R and Munns D N 1975 Phosphate fluxes in the
soil-plant system. Hilgardia 43, 103-130.
Holben W E, Jansson J K, Chelm B K and Tiedje J M 1988
DNA probe method for the detection of specific microorganisms in the soil bacterial community. Appl. Environ.
Microbiol. 54, 703-711.
Ishizuka J 1992 Trends in biological nitrogen fixation research and application. Plant and Soil 141, 197-209.
Kang B T 1990. Alley farming. Adv. Agron. 43, 315-359.
Keeney D 1982 Nitrogen management for maximum efficiency and minimum pollution. In Nitrogen in Agricultural
Soils. Ed. F J Stevenson. Agronomy Monograph 22, pp
605-649. ASA. Madison, WI.
Keyser H H and Munns D N 1979 Effects of calcium,
manganese, and aluminum on growth of rhizobia in acid
media. Soil Sci. SOC. Am. J. 43, 500-503.
Keyser H H and Li F 1992 Potential for increasing biological
nitrogen fixation in soybean. Plant and Soil 141, 119-135.
Khanna-Chopra R and Sinha R 1989 Impact of climate
variation on production of pulses. In Climate and Food
Security. pp 219-236. The Int. Rice Research Inst., Los
Baños, Philippines.
King F H 1911 Farmers of Forty Centuries. Rodale Press.
Inc., Emmaus, PA. 441 p.
Koyama T and App A 1979 Nitrogen balance in flooded rice
BNF for sustainable agriculture
soils. In Nitrogen and Rice, The Int. Rice Research Inst.
pp 95-104. Los Baños, Philippines.
Ladha J K, Pareek R P and Becker M 1992 Stem-nodulating
legume-Rhizobium-symbiosis and its agronomic use in lowland rice. Adv. Soil Sci. In press.
Ladha J K, Pareek R P, So R and Becker M 1990 Stem
nodule symbiosis and its unusual properties. In Nitrogen
Fixation: Achievements and Objectives. Eds. P M Gresshoff, L E Roth, G Stacey and W L Newton. pp 633-640.
Chapman and Hall. New York. London.
Ledgard S F and Steele K W 1992 Biological nitrogen fixation
in mixed legume/grass pastures. Plant and Soil 141, 137153.
Leung K and Bottomley P J 1987 Influence of phosphate on
the growth and nodulation characteristics of Rhizobium
trifolii. Appl. Environ. Microbiol. 53, 2098-2105.
Lumpkin T A and Plucknett D L 1982 Azolla as a green
manure: Use and management in crop production. Westview Tropical Agriculture Series No. 5, Westview Press,
Boulder, CO. 225 p.
Munns D N 1977 Soil acidity and related matters. In Exploiting the Legume-Rhizobium Symbiosis in Tropical Agriculture. Eds. J M Vincent, A S Whitney and J Bose. pp
211-236. Univ. Hawaii Coll. Trop. Agr. Misc. Publ. 145.
Munns D N and Franco A A 1982 Soil constraints to legume
production. In Biological Nitrogen Fixation Technology for
Tropical Agriculture. Eds. P H Graham and S C Harris. pp
133-152. CIAT Workshop 1981, Cali, Columbia.
National Academy of Sciences 1979 Tropical Legumes: Resources for the Future. Washington, DC. 331 p.
National Research Council (US) 1984 Casuarinas: Nitrogenfixing Trees for Adverse Sites. National Academy Press,
Washington, DC. 118 p.
Odum E P 1989 Input management of production systems.
Science 243, 177-182.
Peoples M B and Crasswell E T 1992 Biological nitrogen
fixation: Investments, expectations and actual contributions to agriculture. Plant and Soil 141, 13-39.
Pingali P L, Moya P F and Velasco L E 1990 The post-green
revolution blues in asian rice production - The diminished
gap between experiment station and farmer yields. IRRI
Social Science Division paper No. 90-01.
Plucknett D L and Smith N J H 1986 Sustaining agricultural
yields. Bioscience 36, 40-45.
Pradhan P 1988 The establishment requirements of Sesbania
rostrata as a pre-rice green manure. M.Sc. Thesis. Univ. of
the Philippines, Los Baños, Philippines. 189 p.
Roger P A and Ladha J K 1992 Biological N2 fixation in
wetland rice fields: Estimation and contribution to nitrogen
balance. Plant and Soil 141, 41-55.
Russell J P, Beech D F and Jone P N 1989 Grain legume
productivity in subsistence agriculture. Food Policy 14,
129-142.
Sanchez P A 1980 Management considerations for acid soils
with high phosphorus fixation capacity. In The Role of
Phosphorus in Agriculture. pp 471-514. American Society
of Agronomy, Madison. WI.
Shanmugasundaram S 1988 Seed production and manage-
11
ment of mungbean and soybean. In Sustainable Agriculture: Green Manure in Rice Farming. pp 359-376. The
Int. Rice Research Inst.. Los Baños, Philippines.
Singer M J and Munns D N 1987 Soils, an Introduction.
Macmillan, New York, 492 p.
Singleton P W. Abdel-Magid H M and Tavares J W 1985 The
effect of phosphorus on the effectiveness of strains of
Rhizobium japonicum. Soil Sci. Soc. Am. J. 49, 613-616.
Singleton P W and Bohlool B B 1983 Effect of salinity on the
functional components of the soybean- Rhizobium
juponicum symbiosis. Crop Sci. 23, 259-262.
Singleton P W, Bohlool B B and Nakao P 1992 Legume
response to rhizobial inoculation in the tropics: Myths and
realities. In Myths and Science of the Soils of the Tropics.
Eds. R Lal and P Sanchez. Soil Sci. Soc. Am. Spec. Publ.
No. 29. Madison. WI. In press.
Singleton P W and Tavares J W 1986 Inoculation response of
legumes in relation to the number and effectiveness of
indigenous Rhizobium populations. Appl. Environ. Microbiol. 51, 1013-1018.
TAC CGIAR 1988 Sustainable agricultural production: Implications for international agricultural research. CGIAR
(Consultative Group on International Agricultural Research) Meeting, Berlin. Germany.
Thies J E. Singleton P W and Bohlool B B 1991a Influence of
the size of indigenous rhizobial populations on establishment and symbiotic performance of introduced rhizobia on
field-grown legumes. Appl. Environ. Microbiol. 57, 19-28.
Thies J E, Singleton P W and Bohlool B B 1991b Modeling
symbiotic performance of introduced rhizobia in the field
by use of indices of indigenous population size and nitrogen status of the soil. Appl. Environ. Microbiol. 57, 29-37.
Tjepkema J D, Schwintzer C R and Bensen D R 1986
Physiology of actinorhizal nodules. Annu. Rev. Plant Physiol. 36, 209-232.
Torrey J G 1978 Nitrogen fixation by actinomycete-nodulated
angiosperms. Bioscience 28, 586-592.
Urquiaga S. Botteon P B L and Boddey R M 1989 Selection
of sugarcane cultivars for associated biological nitrogen
fixation using 15N labelled soil. In Nitrogen Fixation with
Non-Legumes. Eds. F A Skinner, R M Boddey and I
Fendrik. pp 311-319. Kluwer Academic Publisher, Dordrecht, The Netherlands.
Watanabe I and Liu C C 1992 Improving nitrogen-fixing
systems and integrating them into sustainable rice farming.
Plant and Soil 141, 57-67.
Weaver R W and Frederick L R 1974a Effect of inoculum
rate on competitive nodulation of Glycine max L. Merrill.
I. Greenhouse studies. Agron. J. 66, 229-232.
Weaver R W and Frederick L R 1974b Effect of inoculum
rate on competitive nodulation of Glycine max L. Merrill.
II. Field studies. Agron. J. 66, 233-236.
Weber D F and Miller V L 1972 Effect of soil temperature on
Rhizobium juponicum serogroup distribution in soybean
nodules. Agron. J. 64, 796-798.
Willey R W 1979 Intercropping - its importance and research
needs. Part 2: Agronomy and research approaches. Field
Crops Abstr. 32, 73-85.
Plant and Soil 141: 13-39, 1992.
© 1992 Kluwer Academic Publishers. Printed in the Netherlands.
PLSO SA02
Biological nitrogen fixation: Investments, expectations and actual
contributions to agriculture
MARK B. PEOPLES and ERIC T. CRASWELL
CSIRO Division of Plant Industry, GPO Box 1600 Canberra, ACT, Australia 2601 and TAC
Secretariat, FAO, Via delle Terme di Caracalla, I-0100 Rome, Italy. (Formerly: Australian Centre for
International Agricultural Research, Canberra, ACT, Australia)
Key words:
fixation
actinorhizal plants, associative N2 fixation, Azolla, legumes, rhizobia, symbiotic N2
Table of contents
Abstract
Introduction
The role of N2-fixing associations in agriculture
Role of legumes
Food legumes
Forage and green manures
Contribution of legume N to plant and animal production
Direct transfer
Contribution of legume residues
Contribution to livestock production
Role of non-legumes
Actinorhizal plants
Grasses and cereals
Forage grasses
Sugar cane
Rice
Research priority setting
Strategy to enhance N2 fixation
Crop and soil management
Tillage
Removal of plant or animal products
Rhizobia, inoculation and plant nodulation
Plant breeding and selection
Breeding for plant varieties which more successfully exploit strains of rhizobium used as
inoculants or already present in soil
Breeding for improved plant yield
Breeding for nitrate tolerance
Trends in research funding
Acknowledgements
References
14
Peoples and Craswell
Abstract
Inputs of biologically fixed N into agricultural systems may be derived from symbiotic relationships
involving legumes and Rhizobium spp., partnerships between plants and Frankia spp. or cyanobacteria,
or from non-symbiotic associations between free-living diazotrophs and plant roots. It is assumed that
these N2-fixing systems will satisfy a large portion of their own N requirements from atmospheric N2,
and that additional fixed N will be contributed to soil reserves for the benefit of other crops or forage
species. This paper reviews the actual levels of N2 fixation attained by legume and non-legume
associations and assesses their role as a source of N in tropical and sub-tropical ,agriculture. We discuss
factors influencing N, fixation and identify possible strategies for improving the amount of N, fixed.
Introduction
Funding of research is, like science itself, subject
to major shifts in thinking which wax, and sometimes wane, as our knowledge advances. In recent years, a major factor influencing the approach of donors to research funding has been
the widespread international concern about environmental pollution and the erosion of the
agricultural resource base. This concern was
crystallized in a report entitled ‘Our Common
Future” (WCED, 1987) which strongly advocated sustainable development i.e. ‘development
that meets the needs of the present without
compromising the ability of future generations to
meet their own needs’. Currently many governments of both developing and developed countries are embracing this policy. As a consequence, donor agencies are increasingly directing research and development funds towards
programs which specifically address the problems
of pollution and natural resource degradation. In
some developed countries, concern about environmental pollution and the effects of farm
chemical residues in food is affecting agricultural
trade and marketing.
Viewed in the context of this broadbrush policy picture, research on biological N, fixation
would appear to have much to contribute to the
sustainability of agriculture. The promise of
‘free’ nitrogen (N) from the atmosphere to enhance productivity, the virtual elimination of risk
of groundwater contamination by nitrate, the
supplementation of human and animal diets with
better-quality protein, and the provision of vegetative cover to reduce the susceptibility of soils
to erosion.
This chapter focuses on the role of N,-fixing
associations in tropical and sub-tropical agriculture, discusses the general framework within
which research priorities are set, and considers
some of the issues which influence the thinking
of donor agencies. The opinions are coloured by
the perception of an increased emphasis by
donors on accountability and cost/benefit analysis in research programming. To some extent this
emphasis reflects a disillusionment with science
amongst sectors of the public and a certain ‘aid
fatigue’ which is evident in some countries.
The role of N2-fixing associations in agriculture
The terrestrial flux of N from biological N2 fixation has been calculated to range from 1396
170 x 10 t N/yr (Burns and Hardy, 1975; Paul,
1988). While the accuracy of these numbers
might be questioned, they are nonetheless indicative of the importance of biological N, fixation in the context of the global N cycle. While
in absolute terms N, fixation is small compared
with total soil N reserves (105,000 x 10 6 t N), it
is at least several fold greater than inputs of N
from N fertilizer (65 x 106 t N/yr; Paul, 1988). It
would be expected that inputs of fixed N into an
ecosystem would be derived from various symbiotic systems involving for example:
- Legumes/ Rhizobium spp.,
- Actinorhizal associations (e.g. Casuarina/
Frankia) or
- cyanobacteria partnerships, (e.g. Azolla/
Anabaena, cycads/Nostoc, and lichens)
in addition to numerous non-symbiotic systems
where free-living diazotrophs fix N, while in
association with plant roots, during decomposition of crop residues, or as cyanobacteria crusts
Nitrogen fixation and agriculture
on soil. Considerable information is available on
N2 fixation inputs from legume sources (e.g.
Herridge and Bergersen, 1988; Ledgard and
Steele, 1991; Peoples and Herridge, 1990), but
although the biology of non-legume systems has
been extensively studied (see Dommergues and
Diem, 1982; Dart, 1986; Gibson et al., 1988;
Skinner et al., 1989; Peters and Meeks, 1989),
much less is known of the amounts of N, fixed in
these associations because of the methodological
difficulties in measuring it (Boddey, 1987; Chalk,
1991; Sougoufara et al., 1990). The relative contribution of symbiotic N, fixation, or associative
and free-living fixing systems to the global total
was assessed by Paul (1988) to be in the order of
70% symbiotic: 30% non-symbiotic. But in areas
of arable agriculture, symbiotic N2 fixation by
legumes is generally expected to be the dominant
source of N input (possibly greater than 80% of
the total; Burns and Hardy, 1975).
Legumes have played an important role in
traditional farming systems since ancient times,
but the Haber-Bosch process and the advent of
cheap N fertilizers led to the widescale abandonment of legume pasture leys and green manuring
in Europe and North America. In recent times
the use of Azolla and milk vetch ( Astragalus
sinicus ) in China and Japan has declined because
fertilizers are increasingly more widely available
and the opportunity cost for labour is high. Some
of the economic factors affecting farmer adoption of legume-ley farming systems in Australia
have been analysed by McCown et al. (1988).
They suggest that the recent decline in the use of
pasture legume leys reflects changes in the
economic circumstances of farmers. However, it
was concluded that a greater use of ley systems
could result if the relative prices and export
markets for wool and wheat change. The object
lesson is that farmers are motivated primarily by
economic gain and will readily adopt biological
N, fixation technology if it is economically advantageous to do so. In developed countries,
changes in macroeconomic and trade policy to
reduce overproduction and excessive use of N
fertilizers, in addition to the demand for ‘organically-grown’ foods may ultimately result in a
resurgence in use of N2-fixing associations in the
farming systems of north America and Europe.
Research funding agencies are monitoring these
15
trends and are increasingly taking into account
the economic and farming systems perspective of
technology options developed through research.
It is in the tropical and sub-tropical regions of
the World that the opportunities are greatest for
the successful exploitation of N2-fixing systems,
and it has been to these areas that a large
proportion of international donor funds has been
directed.
Soils in upland areas of the humid tropics and
sub-tropics are commonly leached and highly
weathered and, after clearing and cultivation
soon become low in both total and plant-available N. With improper management, land in
these regions is subject to high annual losses of
N and to rapid degradation because of overgrazing and deforestation. This trend is widespread
as cultivation intensifies in response to increasing
human populations (Lal, 1989). Crop yields are
often limited by N supply, and animal productivity and fecundity are restricted by poor-quality
forage. However, apart from high-return cash
crops, N fertilizers are used only to a limited
extent because of the low per-capita incomes and
limited credit facilities of most farmers, and
lack of effective infrastructures for fertilizer production and distribution (Wood and Myers,
1987).
When a N2-fixing association is used in agriculture it is presumed that it will satisfy all, or at
least part, of its own N requirements from atmospheric N2, and that fixed N surplus to its
needs will subsequently accrue in the soil and
benefit other crops. However, since the capacity
to fix N2 is dependent upon many physical,
environmental, nutritional and biological factors
(Chalk, 1991; Gibson et al., 1982; Watanabe,
1982), it can not be assumed that any N,-fixing
system will automatically make large contributions to the N cycle. In evaluating the role of
N,-fixing systems in the tropics and sub-tropics,
we direct most attention in this chapter to the
legume/Rhizobium symbiosis because of its importance to agriculture and because of the
wealth of information available. Non-legume associations are included where reliable estimates
of N, fixation have been reported. Readers are
referred to other chapters in this volume for
additional details on both symbiotic and nonsymbiotic systems.
16
Peoples and Craswell
Role of legumes
Table 1. Important food legumes produced in the tropics and
sub-tropics a
Legumes are widely used for food, fodder,
shade, fuel and timber, as cover crops and for
green manure. They are a feature of:
i) Cropping systems: where legumes are
grown in rotation or intercropped with
cereals either as crops in their own right or
as green manures.
ii) Grazing systems: including extensive grazing of natural vegetation in semi-arid regions, intensive pastoral type of agriculture and 'cut-and-carry' systems.
iii) Plantation systems: where legume cover
crops, food crops or shade trees are grown
in the inter-row space of tree crops such as
cocoa, coffee, tea, rubber and oil palm.
iv) Agroforestry systems: in which multipurpose tree and shrub legumes are utilized in
combination with crops and animals to
increase the productivity and the sustainability of the farming system.
Genus
Species
Common name
Arachis
Cajanus
Canavaliu
hypogaea
cajan
ensiformis
gladiata
arietinum
tetragonolba
max
geocarpa
purpureus
culinaris
uniflorum
pruriens
erosus
acutifolius
lunatus
vulgaris
tetragonolobus
aconitifolia
angularis
mungo
radiata
umbellata
Unguiculata
subterranea
Groundnut
Pigeonpea
Jack bean
Sword bean
Chickpea
Cluster bean
Soybean
Kersting's groundnut
Lablab bean
Lentil
Horse gram
Velvet bean
Yam bean
Tepary bean
Lima bean
Common bean, Kidney bean
Winged bean
Moth bean
Adzuki bean
Black gram
Green gram
Rice bean
Cowpea
Bambara groundnut
Food legumes
Cicer
Cyamopsis
Glycine
Kerstingiella
Lablab
Lens
Macrotyloma
Mucuna
Pachyrhizus
Phaseolus
Psophocarpus
Vigna
Voandzeia
a
Plant sources commonly provide 90% of the
calories and up to 90% of human dietary protein
in tropical regions. Commonly consumed legume
seeds contain 17-34% protein, and legume seeds
and vegetables may contribute between 15-25%
of the dietary protein intake. The wide variation
in consumption of the various legume sources by
humans is due to the broad spectrum of crops
grown, and the importance of animals in the
rural economy (Rachie and Roberts, 1974; Wijeratne and Nelson, 1987).
Many different species are utilized as components of lowland and upland cropping systems
(Table 1). Some species are grown widely as
oilseeds (groundnut and soybean) or pulses
(pigeonpea, chickpea, cowpea, green and black
gram, common bean), but often crops have localized use (e.g. lablab bean, rice bean, bambarra
groundnut, Kersting's groundnut). The distribution of crop legumes is generally determined by
their adaptation to particular climates and environments (Rachie and Roberts, 1974; Wood
and Myers, 1987).
Examples of the levels of N2 fixation by food
legumes are presented in Tables 2, 3, 7 and 11. It
After Peoples and Herridge (1990).
should be noted that these examples generally
come from experiments in which the conditions
for plant growth are quite favourable relative to
those in farmers' fields. The values recorded may
therefore be considered estimates of potential N2
fixation under the various agricultural systems.
Nonetheless, what becomes apparent from the
data is that the range of levels of N2 fixation is
large, and this appears to be unaffected by crop
species or by the country in which the investigation was undertaken. For individual species, the
proportion of plant N derived from N2 fixation
(P) range from 0.22-0.92 (groundnut); 0.100.88 (pigeonpea); 0.17-0.85 (chickpea); 0-0.95
(soybean); 0.10-0.71 (common bean) and 0.080.89 (cowpea). These values resulted in amounts
of N2 fixed ranging from 0-450 kg N ha-1. The
levels of fixation were found to be dependent
upon water supply, inoculation, crop rotation,
applications of fertilizer-N, and soil N fertility.
Clearly in those situations where there was low
fixation, legumes would be exploiting soil N
reserves rather than contributing to it. Soil N
fertility will be lost and the potential benefit of
Nitrogen fixation and agriculture
17
Table 2. Examples of estimates of the proportion ( P ) and amount of plant N derived from symbiotic N2 fixation for
commonly-grown food legumes in tropical and sub-tropical systems
Species
Treatment
variable
Total
crop N
(kg N ha -1 )
N2 fixed
P
Amount
(kgN ha -1 crop -1 )
Brazil
India
Water supply
Cultivar
Rotation
Inoculation
Cultivar
171-248
254-319
181-247
147-163
126-165
0.22-0.53
0.55-0.65
0.47-0.53
0.47-0.78
0.86-0.92
37-131
139-206
85-131
68-116
109-152
Pigeon pea
India
Season
Soybean
Brazil
Hawaii
Indonesia
Thailand
Site/season
Temperature
Rotation
Cultivar
Cultivar
Water supply
112-206
120-295
79-100
33-65
121-643
157-251
0.70-0.80
0.97-0.80
0.33
0.78-0.87
0.14-0.70
0-0.45
85-154
117-237
26-33
26-57
17-450
0-113
Boddey et al. (1990)
George et al. (1988)
Sisworo et al. (1990)
Sampet and Peoples”
Rennie et al. (1988)
Kucey et al. (1988)
Common bean
Brazil
Kenya
Cultivar
Phosphorus
18-71
128-183
0.16-0.71
0.16-0.32
3-32
17-57
Duque et al. (1985)
Ssali and Keya (1986)
Cowpea
Brazil
Indonesia
Kenya
Site/season
Rotation
Phosphorus
25-69
67-100
92-94
0.32-0.70
0.12-0.33
0.26-0.35
9-51
12-22
24-39
Boddey et al. (1990)
Sisworo et al. (1990)
Ssali and Keya (1984a)
Green gram
Thailand
Cultivar
71-74
0.89-0.90
64-66
Sampet and Peoples”
Black gram
Thailand
Cultivar
125-143
0.95-0.98
119-140
Sampet and Peoples”
Groundnut
a
Location
Australia
77-92
0.88
Reference
68-88
Bell and Peoples a
Boddey et al. (1990)
Giller et al. (1987)
Kumar Rao et al. (1987)
Unpublished data.
including a legume in a cropping sequence will
not be realised.
A strong inverse relationship between the
level of plant-available soil N (either present in
soil or supplied as N fertilizer) and P is apparent
in Tables 3 and 7. In almost all cases, P was
reduced in the presence of increased soil N,
resulting in generally lower amounts of N, fixed.
The exception occurred when an increase in N
yield with higher soil N was relatively greater
than the reduction in P (data of Yoneyama et al.,
1990a, Table 3). The negative relationship between plant-available soil N and P is arguably
the single most important limitation to N2 fixation.
Food legumes are not only grown in pure
stands; they are also interplanted with other
species in many parts of Africa, Asia and Latin
America (see Fujita et al., 1991). Growing two
or more species simultaneously in the same field
intensifies crop production and more effectively
exploits the environment. Combinations of crops
are determined by the length of growing season
and environmental adaptation, but usually earlyand late-maturing crops are combined to ensure
efficient utilization of the entire growing season
(Ofori and Stern, 1987). The quantity of N, fixed
by the legume (Table 4) in an intercrop depends
on the species, morphology, legume density in
the intercrop mixture, and crop management.
Differences in the competitive abilities of the
component crops for soil N can stimulate N,
fixation in the intercrop (Table 4). Legumes of
indeterminate growth with a climbing habit are
generally most successful.
Forages and green manures
Productivity in tropical and sub-tropical regions
depends not only on food crops, but also on
livestock production. Farmers prefer to feed
their animals at minimum or no cost except
labor, often using them to utilize weeds and crop
by-products (Little et al., 1989; Moong, 1986).
The impact of inadequate nutrition on ruminant
livestock under these conditions can be consider-
18
Peoples and Craswell
Table 3. Effect of soil mineral N and N fertilizers on crop N productivity and the proportion (P) and amount of crop N derived
from N2 fixation
Species
Location
Level
0
100
200
196
210
243
0.61
0.47
0.42
120
99
102
0
50
100
114
194
109
110
104
0.85
0.17
0.80
0.55
0.29
97
33
87
60
30
230
265
63
108
89
115
169
0.33
200
0.34
0.06
0.29
0.26
0.48
0.24
0.68
79
0.15
78
16
18
28
43
28
115
-
Chickpea
Australia
10 (to 120 cm)
326
India
Malaysia
Reference
Fertilizer
N
(kg N ha-1)
India
Australia
N2 fixed
Soil
mineral N
(kg N ha-1)
Groundnut
Soybean
Total
crop N
(kg N ha-1)
70 (to 120 cm)
260
-
0a
100
0b
100
40 at sowing
day 45 + further 20 as nitrate
or 20 as urea
P
Amount
(kg N ha-1
crop-1)
Yoneyama et al. (1990a)
Doughton, Saffigna
and Vallis c
Marcellos, Herridge
and Peoples c
Herridge et al. (1990)
Yoneyama et al. (1990a)
Norhayati et al. (1988)
30
Common
bean
Kenya
-
10
100
149
158
0.39
0.10
58
16
Ssali and Keya (1986)
Cowpea
Kenya
-
116
137
163
138
172
0.53
0.08
0.77
0.67
0.33
62
11
125
92
57
Ssali and Keya (1984b)
India
20
100
0
100
200
a
Yoneyama et al. (1990a)
Uninoculated. b Inoculated. c Unpublished data.
able. Deficiencies in forage protein contents
have been known to affect the rate of growth
and maturation of ruminants, and their effectiveness when used for draught-power. As a result, a
diverse range of herbaceous and tree legumes
are being utilized increasingly for forage in livestock systems (Blair et al., 1990; Peoples and
Herridge, 1990). Good-quality forage legumes
can be either grazed, or harvested and hand-fed
as a high protein supplement to grasses. Alternatively, legume leaves may be used to prepare
protein concentrates suitable for monogastric
animals (poultry and pigs), while the fibrous
residues can be used as leaf cake for ruminants.
Woody perennial legumes, however, can provide more than animal fodder. Tree and shrub
legumes are used to reclaim degraded wastelands, to retard erosion, or provide shade, fuelwood, mulch and green manure. The potential
for the integrated use of these species is well
illustrated in alley-farming systems. Alley cropping involves growing arable crops in the interspaces (alleys) between rows of planted trees or
woody shrubs that are periodically trimmed and
maintained as hedge rows. The leaf and twig
prunings are added to the soil as green manure
or mulch (Wilson et al., 1986). In the humid
tropics, hedge rows are often grown along contours on slopes to decrease water runoff velocity
and sediment transport capacity. Sediments trapped by the contour hedges facilitate the formation of natural terraces (Lal, 1989).
Cultivation of rubber ( Hevea brasiliensis ) and
oil palm ( Elaeis guineensis ) represent important
economic activities in many countries in the
tropics. During the initial immaturity period
(first 5-6 years after planting), the inter-row
space (70-80% of the land area), may be sub-
Nitrogen fixation and agriculture
19
Table 4. Effect of intercropping food legumes with cereals on the proportion (P) and amount of crop N derived from N2 fixation
Species
Rice bean
Location
Thailand
Intercrop
ratio
legume:cereal
100 : 0
75 : 25
50: 50
25 : 75
100:0
25 : 75
Cowpea
Australia
Hawaii
India
Nigeria
100 : 0
71 : 29
100 : 0
75 : 25
100:0
66: 33
100: 0
62 : 38
Level
Total
N
crop N
fertilizer
(kg N ha -1)
(kg N ha -1)
0
20
200
20
200
135
129
67
48
134
178
69
60
0
125
89
60
28
0
119
83
(60 to cereal rows)
25
150
100
136
122
25
100
103
jected to invasion by undesirable weeds that are
costly to control and which harbour diseases and
pests that can infest the developing saplings.
Establishment of twining, perennial legume
cover crops in the inter-row space is a common
practice to restrict weed growth, control leaching
and erosion in steep terrain, and as green manure to return organic matter and N to the soil.
Not only are the economic costs associated with
weed and disease control minimized, but the
trees also mature more quickly (Agamuthu and
Broughton, 1985; Peoples and Herridge, 1990).
The use of green-manure crops, however, is
not restricted to long-term alley and plantation
systems. Green manures are also widely used in
short-term annual rotation sequences with crops
such as rice (Buresh and De Datta, 1991; Ladha
et al., 1988a). Green manures in these instances
may involve both annual food legumes and
perennial species. Because of the requirement
for tolerance to flooding in rice-based systems,
there has been recent interest in the use of
Sesbania rostrata and species of Aeschynomene
and Neptunia as green manures since all of these
have the ability to produce nodules on both roots
and stems, and can grow and fix N2 under
N2 fixed
Reference
P
Amount
(kg N ha-1
0.39
0.63
0.84
0.86
0.72
0.32
0.90
0.51
49
81
56
41
97
57
62
32
Rerkasem et al. (1988)
0.69
0.66
0.30
0.34
0.54
0.58
87
59
18
10
64
48
Ofori et al. (1987)
0.79 118
0.81 110
0.59
72
0.64
66
crop-1)
Rerkasem and Rerkasem (1988)
van Kessel and Roskoski (1988)
Patra et al. (1986)
Eaglesham et al. (1981)
flooded and dry conditions (see Roger and
Ladha, 1991; Watanabe and Liu, 1991).
Despite the large number of legume genera
and species used for forage and green manure in
the tropics and sub-tropics (Blair et al., 1990;
Peoples and Herridge, 1990), there are relatively
few estimates of N, fixation by herbaceous
(Table 5) or woody species (Table 6) because of
methodological difficulties in accurately measuring amounts of N, fixed in perennials with a large
biomass (Danso et al., 1991; Peoples et al.,
1988). For this reason it is difficult to generalize
about levels of N, fixation which might be expected from forage and tree species. However, since
the amounts of N, fixed will be mediated through
P and N yield, the potential contributions by
fixation to the N economies of agricultural systems may be greater than those from crop
legumes (Tables 2-4) because levels of P can be
high (Tables 5 and 6) and the annual accumulation of N can be large. Particularly in tree
legumes (Blair et al., 1990; Danso et al., 1991),
levels of fixation are likely to be influenced by
plant genotype, environmental adaptation, and
management factors, such as frequency of cutting or grazing.
20
Peoples and Craswell
Table 5. Estimates of the proportion (P) and amount of legume N derived from N2 fixation for cover crops and herbaceous
forage legumes in mixed-grass swards
Species
Location
Centrosema pubescens
Malaysia
Pueraria phaseoloides
Desmodium intortum
Australia
Lotononis bainesii
Macroptilium atropurpureum
Stylosanthes guyanensis
Trifolium repens
Vigna luteola
Macroptilium atropurpureum India
Pueraria phaseoloides
(% legume in sward)
Stylosanthes spp.
a
Malaysia
0-40
41-60
61-80
81-100
Australia
Legume
N-yield
(kgN ha-1)
299
25-195 a
16-135 a
N2 fixed
P
Amount
(kg N ha-1)
0.50
150
0.94
24-183 a
(0.8-1.0 range)
0.92
15-124 a
(0.75-0.97 range)
76-95
28-48
0.78-0.83 59-79
0.84-0.87 23-39
12
24
44
30
2-86
2-34
3-102
0.75
0.92
0.86
0.76
0.83-0.92
0.60-0.88
0.73-0.90
9
22
38
23
2-75
2-20
3-84
Period of
measurement
Reference
annual estimate Agamuthu and Broughton (1985)
annual av.
over 2 years
yr 1
yr 2
3 months
Vallis et al. (1977)
Shivaram et al. (1988)
Zaharah et al. (1986a)
yr 1
Vallis and Gardener (1985)
yr 4
yr 6
(9-11 weeks regrowth)
Calculated from mean legume uptake of soil N and P value
Table 6. Estimates of the proportion (P) and amount of plant N derived from N2 fixation for tree and shrub legumes
Species
Acacia spp.
Location
Australia
Philippines
Senegal
Mexico
Aeschynomene afraspera Philippines
A. indica
Thailand
Albizia falcataria
Philippines
Australia
Cajanus cajan
Australia
Calliandra calothyrsus
Indonesia
Desmodium rensonii
Australia
Gliricidia sepium
Mexico
Philippines
Malaysia
Leucaena leucocephala
Nigeria
Tanzania
Thailand
Australia
Sesbania cannabina
Philippines
Indonesia
S. grandiflora
Senegal
S. rostrata
Philippines
Thailand
S. sesban
Indonesia
Senegal
a
Unpublished data.
Legume
N-yield
(kg N ha-1)
N2 fixed
10-20
205
-
0.52-0.66
0.30
0.70
0.75-0.94
0.55
0.65
0.14
0.68-0.84
0.75
0.60
0.58-0.78
0.34-0.39
0.59-1.00
0.70-0.93
0.93
0.79
0.36-0.51
0.88-0.91
0.93-1.00
0.84
0.13-0.18
79
132
296-313
288-344
136-202
127-199
214-230
157-312
54-100
P
Amount
(kg N ha-1)
12
3-6
34
143
11
99
13
182-231
98-134
110
126-141
119-188
83-109
140-286
7-18
Period of
measurement
Reference
annual estimate Langkamp et al. (1979)
Peoples, Almendras and Dart a
6.5 months
Cornet et al. (1985)
annual estimate Roskoski et al. (1982)
56 days
Becker et al. (1990)
Yoneyama et al. (l990b)
Peoples, Almendras and Dart a
90 days
Peoples and Palmer a
Peoples and Palmer a
90 days
Peoples and Nurhayati a
60 days
Peoples and Palmer a
90 days
annual estimate Roskoski et al. (1982)
Peoples and Ladha a
3 months
Zaharah et al. (1986b)
6 months
Sanginga et al. (1989)
annual estimate Hogberg and Kvarnstrom (1982)
Yoneyama et al. (1990b)
Chapman and Myers (1987)
seasonal av.
Pareek et al. (1990)
45-55 days
Peoples and Nurhayati a
2 months
Ndoye and Dreyfus (1988)
2 months
Pareek et al. (1990)
45-55 days
Yoneyama et al. (l990b)
Peoples and Nurhayati a
2 months
Ndoye and Dreyfus (1988)
2 months
Nitrogen fixation and agriculture
Contribution of legume N to plant and animal
production
Direct transfer
It is often assumed that a portion of the N2 fixed
by an intercropped legume is made available to
the associated non-legume during the growing
season, or that direct transfer of N from forage
legumes to companion grasses occurs in mixed
pasture swards (Fujita et al., 1991; Haque and
Jutzi, 198.5; Ofori and Stern, 1987). Decaying
roots and nodules are thought to be important in
this transfer of N. Although these organs generally contain only a fraction of total plant N (see
Buresh and De Datta, 1991), the proportion of
the root system that might be decomposing during growth has not been estimated. Furthermore,
the possibility of N exudation from living roots
should not be ignored (Poth et al., 1986).
Evidence of N transfer from legume to cereal
has been obtained in some intercropping studies
(Bandyopadhyay and De, 1986; Eaglesham et
al., 1981; Patra et al., 1986), although not confirmed in other investigations (Kumar Rao et al.,
1987; Ofori et al., 1987; Rerkasem and Rerkasem, 1988; van Kessel and Roskoski, 1988).
Similarly, there is little evidence of rapid N
transfer in mixed pasture swards (Vallis et al.,
1977; Haque and Jutzi, 1985). Direct transfer of
N from legume to non-legume therefore may not
occur under all conditions, or might only occur
slowly with time.
Contribution of legume residues
The total amount of N in a legume crop (Nl)
comes either from N, fixation (Nf), or by uptake
of mineral N from the soil. In food legumes,
total N is partitioned either into the harvested
seed (Nls) (proportion partitioned into seed described by the harvest index for N, NHI), or the
vegetative parts (leaves, stem and nodulated
roots), which generally remain as crop residues
(Wood and Myers, 1987). The net potential contribution of N, fixation to the N balance of a soil
following legume cropping may be considered as:
Net N-balance = Nf - Nls
(1)
where Nf
= (P x NI)
(2)
and Nls
= (NHI x NI)
(3)
21
Since leaf fall during crop legume development and the nodulated roots can each contain
up to 40 kg N ha-1 (Bergersen et al., 1989; Buresh and De Datta, 1991; Chapman and Myers,
1987; Kumar Rao and Dart, 1987), it is desirable
that they are both included when determining N1
to ensure an accurate estimate of the net N
balance. However, often only standing shoot N
data are available and so net N balances are
likely to be underestimated.
When the quantities of N involved in plant
growth, in N2 fixation, and in the seed are
calculated for food legume crops, it is apparent
that the net N balance can be low and in some
instances negative (Tables 7 and 11). Levels of
fixation achieved by many crops in the field may
be high, but are not always sufficient to offset
the N removed with the harvested seed. Clearly,
if food legumes are to contribute substantial
amounts of N to the soil, P must be considerably
greater than NHI. High-productivity crops such
as soybean would require P values greater than
80% to avoid a net loss of N from the system.
The potential net contribution, however, may be
even further diminished if crop residues are used
as fodder for animals. Despite the wide range of
net N balances summarized in Tables 7 and 11,
reported benefits of tropical crop legumes to
subsequent cereal crops are consistent and substantial and may persist for several seasons
(MacColl, 1989), regardless of whether the
legume was grown in monoculture or intercropped (Table 8). Improvements in cereal yield
following monocropped legumes lie mainly in the
0.5 to 3 t ha-1 range, representing around 30 to
3.50% increases over yields in cereal-cereal cropping sequences. Responses to previously intercropped legumes are more modest and yield
improvements of only 0.34 to 0.77 t ha-1 have
been measured. When the contribution of the
legume was quantified as fertilizer-N equivalence, as much as 68 kg N ha-1 was required in
the cereal-cereal sequence to achieve similar
yield improvements (Table 8).
If the benefits of crop legumes in rotations
cannot be explained solely in terms of residual
fixed N, then what are the sources of the benefits
demonstrated in Table 8? Most certainly a number of factors can operate, the relative importance of each dictated by site, season and crop
sequence. Extra yield from a rotation can result
22
Peoples and Craswell
Table 7. Net N balances for several legume crops following seed harvest
Species
Location
Net
Reference
Seed
Total
NHI c N2 fixed
a
b
crop N
N balance e
yield
kg N ha -1 kg N ha -1
P
Amount d
kg N ha -1 crop -1 kg N ha -1
Groundnut
Ghana
Thailand
54
116
Pigeon pea
India Early
Medium
Late
maturity
Soybean
Australia
India
Nigeria + 25 N g
+ 100N
128
245
0.42 0.79
0.47 0.61
72
120
134
0.50
0.41
0.21
262
202
204
195
329
232
255
244
Common bean Brazil
12
Green gram
Australia
Thailand
Cowpea
Australia
Ghana
Nigeria + 25 N g
a
b
c
d
e
f
g
101
150
+ 47
+ 34
Dakora et al. (1987)
Suwanarit et al. (1986)
0.10
0.46
0.51
7
55
69
- 32
+ 6
+ 41
Kumar Rao and Dart (1987)
0.80
0.87
0.80
0.80
0.95
0.55
0.89
0.65
312
128
227
159
+
+
-
Chapman and Myers (1987)
Chandel et al. (1989)
Eaglesham et al. (1982)
74
0.16
0.62
46
+ 34
Ruschel et al. (1982)
89
32
177
61
0.50
0.52
0.63
0.60
112
37
+ 23
+ 5
Chapman and Myers (1987)
Suwanarit et al. (1986)
80
65
48f
49
+ 100N 54 f
49
102 f
81
82
125
226
82
134
94
118
155
182
126
0.64
0.29
0.58
0.37
0.57
0.41
0.66
0.44
0.65
0.69 87
0.89 201
0.61 50
0.75 101
0.30 28
0.41 49
0.59
91
0.63 115
0.59 74
+ 7
+136
+ 2
+ 52
- 26
0
- 11
+34
+ 8
Ofori et al. (1987)
Dakora et al. (1987)
Eaglesham et ai. (1982)
39 f
49
28
50
74
23
36
Awonaike et al. (1990)
N removed in seed, Nls.
Total crop N at seed harvest, NI.
Nitrogen harvest index = Nls/NI.
Amount of N, fixed, Nf = N1 x P.
Net contribution of legume residue N to soil = Nf - Nls.
Cultivar comparison.
Rate of N-fertilizer application (kg N ha -1 ).
from:
i) Improvements in soil structure following
legumes (Hearne, 1986), or improvements
in soil water-holding and buffering capacity, and increased nutrient availability associated with incorporation of legume residues (Buresh and De Datta, 1991),
ii) Breaking of cycles of cereal pests and diseases, and phytotoxic and allelopathic effects of different crop residues (Sanford
and Hairston, 1984),
iii) Enhancement of soil microbial activity and
possibly heterotrophic N2 fixation following addition of legume residues (Ladha et
al., 1989), or
iv) More nitrate remaining in the soil following a legume than after a cereal or other
non-legume (Herridge and Bergersen,
1988). This might be due to less nitrate
being taken up by the legume, or result
from a stimulation of mineralization rates
under a legume crop.
However, much of the unexpected benefits
probably result from the use of a cereal-cereal
rotation as the 'bench-mark' comparison for the
performance of a legume-cereal sequence. Even
when most of the N is removed in legume seed
and there is little or no net gain of N by the soil
(Table 7), a much higher loss of N would be
expected from a non-leguminous/non-fixing crop
when its seed is harvested (Nnfs). This can be
illustrated by inserting the data of Ofori et al.
(1987) for monocrop cowpea and maize into the
equation (1) above:
Nitrogen fixation and agriculture
23
Table 8. Grain yield responses of cereals to previous legume crops relative to a cereal-cereal cropping sequence
Increase
in cereal
yield
(t ha-1)
Relative
increase
(%)
N-fertilizer
equivalence
(kg N ha-1)
Reference
Legume - maize
Pigeon-pea monocrop
intercrop a
Groundnut
Soybean
Green gram
Groundnut
Cowpea
Groundnut
Soybean
Lablab bean
Pigeon-pea
1.06
0.77
0.35
0.48
0.48
0.85
0.91
0.72
0.49
1.96
2.78
353
157
25
34
34
89
96
24
16
65
89
40
20+
Kumar Rao et al. (1983)
60
60
9b
7b
33b
67 c
Dakora et al. (1987)
Legume - rice
Soybean
Green gram
0.80
0.20
66
17
Legume - sorghum
Black gram
Green gram
3.68
2.82
79
61
Legume - wheat
Pigeon-pea
Black gram
Green gram
Cowpea
Groundnut monocrop
intercrop a
Green gram monocrop
intercrop a
Cowpea monocrop
intercrop a
0.27
1.26
0.65
0.74
0.75
0.34
1.60
0.48
1.00
0.38
21
98
51
58
23
10
49
15
30
11
Crop sequence
Suwanarit et al. (1986)
MacColl (1989)
Chapman and Myers (1987)
68
68
Doughton and MacKenzie (1984)
Singh and Verma (1985)
28
12
68
16
38
13
Bandyopadhyay and De (1986)
Intercropped with sorghum or maize in previous growing season.
Mean response over a 5-year period.
c Pigeon-pea grown for 2 years before incorporation.
a
b
Cowpea: Nf = 87 kg N ha-1 ;
Nls = 80 kg N ha-1.
Net N balance = 87 - 80
= +7 kg N ha-1
Maize: Nf = 0; Nnfs = 57 kg N ha-1.
Net N balance = 0 - 57
= -57 kg N ha-1
While the real benefit of the cowpea is only
7 kg N ha-1, the benefit to a following cereal,
assesed by comparing yields of maize-cereal vs
cowpea-cereal, would combine both the small
contribution by the legume (+7) and the N loss
through removal of maize cobs and seed (-57).
Therefore the apparent benefit is 64 kg N ha-1.
Care is also needed in the interpretation of N
fertilizer equivalence data because fertilizer-use
efficiencies can vary and fertilizer N may be
subject to volatilization losses. This would inevitably overestimate the beneficial effects of
legumes and make direct comparisons of data
difficult.
At the time of seed harvest and incorporation
of legume residues in the soil, the total N in the
plant is unevenly distributed between the different vegetative organs. These organs also differ
from one another in the extent to which they
release mineral N to the soil and ultimately, to
following crops. The N transformations occurring during breakdown of crop residues are in-
24
Peoples and Craswell
fluenced by temperature, residue management
and soil physical or chemical properties, and in
rice cropping systems, will be dependent upon
whether soils are flooded or remain aerobic
(Buresh and De Datta, 1991). Decomposition
may be modified by lignin and polyphenol content (Palm and Sanchez, 1991; Vallis and Jones,
1973), but it is tissue N concentration, C:N
ratio, and soil water status which are the primary
factors influencing the rate of mineralization and
availability of legume N to a subsequent crop
(Peoples and Herridge, 1990; Sisworo et al.,
1990). The potential for legume leaves to contribute N to a subsequent crop can be considerable since they represent the single largest source
of vegetative N remaining in the residue trash,
and because their high N content and low C: N
ratio favours mineralization. Field experiments
(cited by Peoples and Herridge, 1990) with 15Nlabelled soybean residues indicate that >30% of
the shoot-N of a cereal crop planted immediately
following incorporation of soybean trash can be
derived from mineralisation of soybean leaf material (3.2% N). Soybean stems (1.6% N), on
the other hand, contributed 6% of the cereal
N, despite the fact that the amount of stem-N in
the trash was one third of the N added as leaves.
A notable feature of mineralization of N in plant
residues is that after a short period of time the
rate of mineralization is quite slow, regardless of
the initial % N composition of the residues
(Henzell and Vallis, 1977). Therefore, N in crop
residues that is not mineralized during the first
season becomes available only very slowly thereafter (usually 4-10% per year) for successive
crops.
Combining the factors which influence the
contribution of legume N to cropping systems, it
is possible to derive a conceptual model to relate
the transfer of fixed N, from a leguminous crop
to the following crop (after Myers and Wood,
1987):
Contribution to a subsequent crop =
N1 x P(1-NHI) x Fm X E
(4)
Where N1, P, and NHI are as defined previously, Fm is the proportion of legume N mineralized, and E is the efficiency of utilization of
this mineral N.
To maximize the contribution of legume N to
a following crop, it is necessary to maximize N1,
P, Fm, and E and minimize NHI. There is little
control over the efficiency of use of the mineral
N by a subsequent crop, but the other factors in
the equation can be subjected to manipulation
through different management strategies. Unfortunately it is not always possible to optimize
these factors for food legumes grown in tropical
and subtropical farming systems (Myers and
Wood, 1987).
To maximize N yield and, therefore, N, fixation, the legume crop should be grown in the
most favourable growing period, i.e. the wet
season or under irrigation. In most Asian areas
the wet season, and irrigation in the dry season is
reserved for cereal crops, particularly rice. As a
result the food legume is commonly grown under
the less favourable conditions of the late wet or
dry season, and maximum yields are not
achieved. Another reason for low yields is that
farmers commonly apply much lower rates of
fertilizers to legumes that to other crops, and it
appears that in many countries food legumes
must rely largely on the natural soil fertility
(Craswell et al. , 1987). As a consequence,
legume production can be seriously limited by
nutrient deficiencies.
If the legume is grown for seed, it is desirable
to maximize yield and harvest index. This, in
turn, maximizes NHI. The higher the value of
NHI, the lower the amount of residual N to be
returned to the soil. Often the Fm will also be
low, because the quality of the residues is usually
reduced by the mobilization and re-allocation of
N to reproductive parts during seed filling and
senescence (Peoples and Dalling, 1988). When a
green manure crop is grown, however, the entire
crop is returned to the soil. The quantity of N
and the N concentration in the legume material
returned is likely to be higher than when seed is
removed from a food legume. As a consequence,
decomposition might also be expected to be
rapid and Fm high (Peoples and Herridge,
1990).
The N contents of legume green manures in
rice and alley cropping systems can range from
2-5% (Ladha et al., 1988a; Wilson et al., 1986)
and the inputs of N can be considerable (Table
9). Rates of N accumulation can be exceptionally
Nitrogen fixation and agriculture
25
Table 9. Increase in cereal grain yield by green manure crops relative to unfertilized fallow
Cereal
Legume used
as green manure
Rice
Aeschynomene spp.
Crotalaria juncea
Cyamopsis tetragonoloba
Glycine max
Sesbania aculeata
S. cannabina
S. rostra
S. sesban
Vigna unguiculata
Maize
Leucaena leucocephala
Gliricidia sepium
a
Duration of
green manure
crop
(days)
49
60
60
60
52
84
60
168
-
Legume
N input
(kg N ha-1)
Increase
in cereal
yield
(t ha)
Relative
increase
(%)
Reference
423-532
110
87
196
108
94
267
92
113
4.0-4.1
2.4
2.1
0.6
2.3
1.0
3.8
2.0
2.5
82-85
75
66
46
72
77
181
102
78
Alazard and Becker (1987)
Beri et al. (1989)
300
42 a
42 a
40
80
1.9
1.6
1.7
0.6
1.6
146
97
105
60
160
Sanginga et al. (1988)
Kang (1988)
Chapman and Myers (1987)
Beri et al. (1989)
Chapman and Myers (1987)
Rinaudo et al. (1983)
Palm et al. (1988)
Beri et al. (1989)
Haque and Jutzi et al. (1985)
N yield from hedgerow prunings during one season of alley-cropped maize.
rapid (from 0.4 to 10.8 kg N ha-1 day-1; Table
9). Studies of residue decomposition in alleycropping systems suggest that 50% of the added
legume N may be released within 1-9 weeks
depending on initial N content and prevailing
environmental conditions (Wilson et al., 1986).
It is perhaps not surprising, therefore, to find
substantial increases in yield of cereal crops following green manures (Table 9), equivalent to
50-120 kg fertilizer-N ha-1 (Beri et al., 1989;
Ladha et al., 1988a; Sanginga et al., 1988).
Clearly green manures can provide considerable
amounts of N to the soil that can greatly benefit
subsequent crops. However, if the green manure
is grown in the place of an alternative cash- or
food crop, the cost of this benefit will include a
lost-opportunity cost over and above the expense
of the labor and materials required for sowing
and incorporation of the leguminous material.
Green manure legumes (particularly flood-tolerant species) make their greatest contribution in
single-crop rainfed lowland systems where shortterm waterlogging occurs during the transition
between dry and wet season. Alternative crops
cannot be grown in these areas and farmers are
usually reluctant to apply inorganic fertilizers
because of the adverse climatic conditions and
lack of consistent fertilizer response.
The benefits of green manuring also exceed
the initial cost of legume establishment and
maintenance in cover-crop systems in plantation
estates. The return of N and organic matter from
legume covers to soil and reduction of leaching
losses of N in rubber and oil palm plantations
inevitably lead to better growth by the trees,
earlier commencement of harvesting and increased yield. The contribution of organic N in
leaf litter from perennial legume covers during
the 5-year immaturity period of rubber trees has
been measured to be 250-400 kg N ha-1, and
120 kg N ha-1 yr-1 has been measured in oil
palm (Agamuthu and Broughton, 1985; Peoples
and Herridge, 1990). Residual effects of cover
crops, however, appear to be very much larger.
Long-term trials have shown that trees would
require at least 900 kg of fertilizer N ha-1 over 13
years to achieve similar rubber yields to trees
under legumes. Though shading out of the
legume covers normally commences after about
4 years, the net residual effect of the covers on
rubber extends well into the 15th year of tapping, or up to the 20th year from planting. The
long-term beneficial effect of legumes seems to
result from improvements in soil physical properties and the increased rooting and exploitation
of the soil by trees, coupled with the slow decline
of the cover crops as the trees mature and the
progressive release of N and other nutrients from
legume leaf litter (Agamuthu and Broughton,
1985).
26
Peoples and Craswell
Of course legumes are also used as sources of
green manure in situations other than intensive
agriculture, or productive plantations. Native
legumes are used to restore soil fertility in bush
fallow/shifting cultivation systems, for instance,
and tree legumes play an important role in regenerating degraded or erosion-prone land. Leaf
and litterfall from stands of Leucaena in the dry
tropics can annually contribute 10 t of organic
matter to soil per ha, containing over 250 kg N
(Sandhu et al., 1990). Although decomposition
of this litter under generally harsh dryland conditions is likely to be slower than in the intensive
agricultural systems described above, rates of N
release to the soil will still largely be dependent
upon litter C: N ratio. Therefore it would be
anticipated that turnover rates for fruit litter and
leaf (C: N ratios of 17-18) would be faster than
for twigs or woody debris (C: N ratio >33).
Nonetheless, up to 80% of the total litterfall N
might be expected to be released to the soil
annually (Sandhu et al., 1990), representing a
major contribution to the cycling of N in degraded and infertile environments.
Contribution to livestock production
The role of forage legumes in animal production
is likely to be different under the contrasting
management protocols used in the extensive
grazing common in semi-arid Africa, the pastoral
systems of Australia and Latin America, and the
opportunistic browsing or cut-and-carry practices
employed in Asia. Therefore, only the main
quantitative aspects concerning the contribution
of forage legumes to livestock systems will be
discussed. Further details on the management of
particular forage legumes and their contribution
to the N economies of tropical and sub-tropical
environments are described elsewhere (e.g. Blair
et al., 1990).
Legumes can improve ruminant productivity
by increasing both potential stocking rates and
annual liveweight gains. Linear relationships
have commonly been found between animal performance and dietary intake of leguminous material or legume content in a pasture sward. Such
responses may be attributed to increased forage
production, and/or increased forage intake and/
or improved forage quality.
Annual yields of N from forage legumes can
be considerable (e.g. potential of >700
kgN ha-1 yr-1 in tree legumes; Blair et al.,
1990) and appreciable amounts of organic N
often accumulate in soils (Vallis, 1972). The
improvement in soil-N status by forage legumes
often leads to substantial increases in grass and
total herbage yields. The inclusion of legumes
can also help slow the progressive decline in
animal carrying capacity and plant productivity
which inevitably occurs as pastures age (Haque
and Jutzi, 1985).
Ultimately, ruminant production is dependent
upon dry matter digestibility and intake. Intake
is in turn controlled by the extent and rate of
digestion in the rumen. A minimum level of
protein (and energy) is required if microbial
growth in the rumen is not to be limited. Under
practical feeding conditions, a dietary intake of
fodder of 1.15% N (7.8% crude protein) is
necessary to provide maintenance ration and
1.74% N is considered adequate to meet the
protein requirements of beef cattle. Tropical
grass species are often low in N content (Little et
al., 1989). Although it is possible to maintain the
level of grass N above the critical minimum value
with N fertilizer, the rates of application required (e.g. >400 kgN ha-1 yr-1) make this an
unrealistic option. Rather, legumes whose
foliage generally contain 2-4% N (Little et al.,
1989) are used as a high-protein supplement.
Increasing the proportion of legume in the diet
will not only improve both protein content and
digestibility, but can also increase the voluntary
intake of the entire diet. Furthermore, inclusion
of legume in a pasture can raise the N content of
the associated grasses (Bryan and Velasquez,
1982).
Role of non-legumes
Actinorhizal plants
Apart from legumes (and plants of the genus
Parasponia; Trinick and Hadobas, 1990), which
are nodulated by Rhizobium spp., about 200
plant species covering 8 families and at least 17
genera in the tropics and sub-tropics are nodulated by N2-fixing actinomycetes (Frankia).
There are far fewer N2-fixing actinorhizal plants
Nitrogen fixation and agriculture
than N2-fixing legumes, but their numeric inferiority is offset by their capacity to regenerate
poor soils and stabilize eroding land surfaces,
produce timber and fuelwood, and to shelter
cattle and crops. Of all these actinorhizal associations, possibly the Casuarinas have the
greatest potential for agriculture (Dart, 1986).
There have only been a few measurements of
N2 fixation in Casuarina spp. Two field studies,
conducted in Senegal in containers (1 m3) of
N-deficient, fumigated soil (Gauthier et al.,
1985; Sougoufara et al., 1990), showed large
responses to Frankia inoculation, with estimates
of P ranging from 0.39 to 0.65 in 11- and 24month-old plants (Table 10). Extrapolation of
these data to a normal field density of 10,000
trees ha-1 gave estimates of N2 fixation of between
40-60 kg N ha -1 yr -1
and
400-440
kg N ha-1 yr-1. Another study in Thailand calculated values of P for mature Casuarinas either
growing on farms, or as part of natural communities to range from 0.65 to 0.90 under many
different environments and soil types (Table 10).
27
Azolla
Azolla is a free-floating aquatic fern with a global distribution which normally contains the
N2-fixing
cyanobacterium
heterocyst-forming
Anabaena azollae as a symbiont (Peters and
Meeks, 1989). The ability to the assimilate N2
enables Azolla to grow where N is limiting to
other aquatic plants. In the tropics, Azolla may
form thick mats on the surface of freshwater
ponds, marshes, drainage ditches, and rice paddies, and it has long been used both as a weed
suppressor and green manure in conjunction
with labor-intensive lowland rice cultivation in
Asia. Azolla may also be used as feed for cattle,
pigs, ducks and fish, or be composted for dryland
and vegetable crops (Watanabe, 1982; Watanabe
and Liu, 1991).
Under optimal cultural conditions, Azolla
doubles its biomass within 2-3 days. Its dry
matter contains 4-6% N, and one crop can
accumulate 30-100 kg N ha-1 . Under ideal conditions, this might represent an annual N produc-
Table 10. Estimates of the proportion ( P ) and amount of plant N derived from N 2 fixation by non-legume associations
Species
Location
Symbiotic associations
Casuarina spp.
Senegal
Azolla spp.
Thailand
Philippines
Non-symbiotic associations
Paspalum notatum
Brazil
Brachiaria decumbens
B. humidicola
B. ruzizensis
Panicum maximum
Kallar grass
Sugar cane
Rice
a
uninoculated.
Brazil
Brazil
Pakistan
Brazil
Philippines
Japan
Japan
b
N yield
(kg N ha-1)
N2 fixed
P
Amount
(kg N ha-1)
23-33
623
988
48
-
0.39-0.55
0.65
0.44
0.65-0.95
0.83
0.59-0.99
9-18
404
440
40
-
410
112
115
98
103
37-58
18-19
132
152-294
34
4
-
0.08-0.11
0.13
0.40
0.30
0.10
0.16-0.39
0.41 a -0.54b
0.22-0.26
0.44
0.37-0.56
0.32-0.35
0.11-0.19
0.10-0.31
inoculated treatment.
39
14
45
29
10
15-32
4-5
59
56-163
12
0.4-0.8
-
Period of
measurement
6.5 months
1st year
2nd year
Reference
Gauthier et al. (1985)
Sougoufara et al. (1990)
25 days
Yoneyama et al. (1990b)
Watanabe (1982)
Yoneyama et al. (1987)
22 months
13 months
13 months
Boddey et al. (1983)
Boddey and Victoria (1986)
Boddey and Victoria (1986)
26 weeks
7 weeks
60 days
12 months
250 days
2 crops
120 days
100 days
Miranda et al. (1990)
Malik et al. (1988)
de Freitas et al. (1984)
Lima et al. (1987)
Urquiaga et al. (1989)
Ventura and Watanabe (1983)
Yoo et al. (1986)
Fujii et al. (1987)
28
Peoples and Craswell
tion of 450-840 kg N ha-1 (Watanabe, 1982).
With values of P ranging from 0.59 to 0.99
(Table 10; see also Roger and Ladha, 1991), this
represents quite a potential contribution to the N
economies of rice-based agriculture. The growth
and N2-fixing capacity of Azolla, however, can
be affected by environmental variables, such as
temperature, light, and humidity, or be influenced by lack of mineral nutrients (particularly
phosphorus) , insect predation and pathogens, so
it is uncertain whether this potential is often
realised in farmers' paddies.
The decomposition of Azolla is rapid and recoveries of Azolla-N in subsequent rice crops can
range from 33-69%, depending upon method
and time of incorporation and the amount of
Azolla applied (Kumarasinghe et al., 1986; Swatdee et al. , 1988). These recoveries of N are
comparable to fertilizer efficiency rates achieved
with urea or ammonium sulphate in these
flooded rice systems.
Grasses and cereals
At least 6 genera of diazotrophic (N2-fixing)
bacteria have been isolated from the roots of
grasses and cereals of agricultural importance
(Boddey and Dobereiner, 1988). Yet despite
more than 15 years of intensive study, the role of
associative N2 fixation as a source of N in agriculture has remained a controversial issue (Michiels
et al., 1989), due in part to perceived limitations
in the methods used to quantify fixation. The
acetylene-reduction assay has long been used as
a means of demonstrating N2-fixing activity by
roots of such species as maize ( Zea mays ) and
sorghum (Sorghum bicolor) (Alexander and
Zuberer, 1988; Dart, 1986), but is has only been
in recent years that N balance and 15N techniques have provided convincing evidence that
some forage grasses, sugar cane ( Saccharum
spp.), and wet-land rice ( Oryza sativa ) can,
under certain conditions, derive considerable
amounts of N from associated N2-fixing bacteria
(Boddey, 1987; Chalk, 1991).
Forage grasses
Measurements of N, fixation associated with C4
grasses (Table 10) have largely been undertaken
in Brazil with several important forage species,
and in Pakistan with the salt-tolerant Kaller grass
( Leptochola fusca ). Estimates of P range from
0.08-0.13 for Paspalum notatum, to around 0.40
for Brachiaria decumbens, Panicum maximum
and Kallar grass (Table 10). Extrapolation from
small-plot data suggest that associative N, fixation may potentially contribute up to 3040 kg N ha" annually to pasture productivity
(Chalk, 1991).
Sugar cane
The experimental evidence for high inputs of N
from associative N, fixation with sugar cane
(Table 10), is consistent with field observations
that high productivity can be maintained despite
continuous cropping, minimal N-fertilizer inputs
and
periodic
removal
between
100-200
kg N ha-1 with each cane harvest (Lima et al.,
1987). The range of estimates of P from 0.220.56, and the calculated contributions of between
21 and 200 kg N ha-1 (Table 10) suggest that
there may be opportunity to improve sugar
cane's N2-fixing capacity through breeding (Urquiaga et al., 1989).
Rice
N-balance studies in pots (App et al., 1980) and
acetylene-reduction assays on field-grown plants
(Ladha et al., 1988b; 1989) have indicated measureable contributions of associative N2 fixation
in rice. Other studies have confirmed estimates
of P as high as 0.35 (Table 10), but the widespread significance of associative fixation with
rice in the field is difficult to evaluate since not
all investigations have revealed N2-fixing activity
(Chalk, 1991). There is, however, some evidence
of varietal differences in the ability of rice to
form effective N2-fixing associations (Ladha et
al., 1988b), which might be exploited in future
breeding programs.
There is now unequivocal experimental evidence that indicates that non-legumes can fix
agriculturally significant amounts of N, but often
it has only been demonstrated under carefully
controlled conditions, or been extrapolated from
pot studies. In many instances, the actual microorganisms responsible for N, fixation have not
been isolated or identified, and it might be
necessary for plants to be growing in impoverished, N-deficient soil or under special en-
Nitrogen fixation and agriculture
vironmental conditions before they derive any
net benefit from an association.
It is difficult to asign to role to N2-fixing
non-legumes in tropical and sub-tropical agriculture since there have been so few field measurements using reliable methodology. However, it
would appear that the potential for consistent
inputs from biological N2 fixation would be smaller for non-symbiotic systems than for symbiotic
systems involving actinorhizal plants or Azolla
because of the apparent tenuous nature of the
associative N2-fixing process in species such as
rice. Once the constraints to N2 fixation are
understood and these organisms are more widely
studied, the possibility exists to promote better
utilization of non-legume associations to improve
the capacity for N2 fixation in agricultural
systems.
Research priority setting
Amongst the many factors to be taken into
account in setting priorities, equity is a key
factor for development assistance agencies because the primary motivation for aid is the
humanitarian consideration. In this context research on N2 fixation should be attractive because it would seem to be especially appropriate
for subsistence farmers who cannot afford, or
have limited access to N fertilizers. On the other
hand, it is interesting to note the assertion (Hubbell, 1988) that most N2-fixation technology
transfer is directed to large-scale farmers who
may need it least. This dilemma arises from the
fact that transfer of N2-fixation technology to
poor upland farmers requires a great deal of
education and effort by extension agencies. Furthermore, the provision of inoculants and other
inputs may be difficult in upland areas because of
poor infrastructure.
Research priorities can be set in relation to the
target farming system or agroecological zone. As
discussed above, shrub legumes and trees provide poor farmers with more than simply a valuable source of N for infertile soils. They also
stabilize soils on hillsides against erosion and
supply feed for animals and fuelwood, thus easing pressure on native forests (Craswell and
Tangendjaja, 1985). Yet N2-fixing trees have
been a relatively neglected field for research
29
despite their multipurpose uses. Taking into account the socioeconomic environment of upland
farmers in the tropics, low input technologies
utilizing small inputs of phosphorus and/or
mycorrhizal plants should be given high priority.
In this regard, research should be designed on a
farming-systems basis to develop a package of
appropriate technology of which N2 fixation is
only a part.
In lowland systems with irrigation and good
infrastructure, N fertilizers rather than N2 fixation are likely to continue to be the main source
of N for high-yielding rice crops. Priority should,
however, be given to research which integrates
both sources of N in ecologically sound systems
which minimize pollution and maximize efficiency (Buresh and De Datta, 1991; De Datta and
Buresh, 1989). Examples of such integrated N
management might include the use of Azolla in
urea-fertilized plots to reduce volatilization losses of applied fertilizer (Eskew and Kumarasinghe, 1990), or combining Azolla and fish in
intensive rice-growing areas to increase productivity while reducing the need for fertilizer and
insecticide (Watanabe and Liu, 1991). Effective
development of these and other innovative technologies also requires multidisciplinary research
conducted with a farming-systems perspective.
Given the important considerations discussed
above, the ultimate goal of most research on
biological N2 fixation remains the enhancement
of the contribution of this process to agricultural
production and to human nutrition and welfare
(Anderson et al., 1987). It is apparent from the
data of Tables 2-10 that the measured benefits
from N2 fixation in the tropics and sub-tropics
are both variable and inconsistent. There remains, therefore, considerable opportunities for
improvement. Strategies which might be used to
enhance N2 fixation in legumes will be summarized in the following sections (see also Peoples
and Herridge, 1990). Similar principles are likely
to apply to other symbiotic systems and to nonsymbiotic associations (Chalk, 1991; Sougoufara
et al., 1990; Urquiaga et al., 1989).
Strategies to enhance N2 fixation
The level of N2 fixation (Nf) in any legume
system is determined by both plant reliance upon
30
Peoples and Craswell
N2 fixation for growth (P) and plant N yield (Nl;
see equation (2)). An increase in the amount of
N2 fixed should be achieved by:
i) Maximizing legume yield within the constraints imposed by agronomic, nutritional, managerial, or environmental factors,
ii) Reducing the amount of nitrate in the
rooting zone through grazing management, timing of cropping, or soil and tillage management,
iii) Optimizing the numbers and effectiveness
of rhizobia through strain selection and
inoculation techniques, or through plant
breeding for either promiscuous or selective nodulation, or
iv) Reducing the legume's sensitivity to the
inhibitory effects of soil nitrate on nodulation and N2 fixation through plant breeding and selection, or through enhancing
nodule initiation and development by increasing the populations of desirable
rhizobia in the rooting zone.
Crop and soil management
Both yield and P can be influenced by management strategies which manipulate levels of soil
mineral N. George et al. (1991) have described
management practices which regulate soil N during the wetting/drying cycles in rice/legume
cropping systems in the rainfed lowlands. Here
we examine two important practices:
(i) Tillage. Clean cultivation accelerates the
oxidation of organic matter in soils and generally
results in higher nitrate-N in the profile (e.g.
Thomas et al., 1973). Cultivation may also decrease the rates of denitrification, immobiliza-
tion, and leaching of nitrate (Rice and Smith,
1982; Thomas et al., 1973). For cereals under
no-tillage, additional fertilizer-N may be required to supplement the reduced soil nitrate-N,
but for legumes, the lowered soil nitrate levels
should result in enhanced N2 fixation. No-till
systems can also modify and improve soil structure to create more favourable soil moisture and
temperature regimes for plant growth (Lal,
1989).
In cropping systems research involving soybean in sub-tropical Australia, nodulation and
N, fixation were substantially improved under
no-tillage, compared with the cultivated system
(Table 11). The increased N, fixed resulted primarily from increased P, since yields were essentially unaffected by tillage practice. N balances
were positive for both systems, although substantially higher for no-tillage. Similar results have
been reported for several tillage trials in Thailand (Rennie et al., 1988).
(ii) Removal of plant or animal products. The
quantity of soil nitrate available to a legume can
be influenced by grazing management and recent
cropping. Legumes grown immediately after a
cereal crop can fix considerably more N than if
grown in previously fallowed soil (e.g. Bergersen
et al., 1989). This stimulation of N, fixation can
be such that the net N balance following seed
harvest in a soybean crop, for example, can be
increased from -44kg N ha-1 after fallow, to
+39 kg N ha-1 in a precropped area (Bergersen
et al., 1985). Thus with the proper choice of
rotation, cropping systems can be managed for
improved N2 fixation, although other factors
Table If. N budgets for soybean grown with conventional cultivation or zero-tillage
Tillage a
Nodule mass b
(mg/plant)
Crop N
Fixed N c
N balance
180
(0.73) c
232
(0.88)
+ 30
(kg N ha-1)
Cultivated
86
245
1.50
No-tillage
139
264
152
a
b
c
d
c
d
Seed N
Values shown are means of four crops grown over three years (Hughes and Herridge, 1989).
Sampled between 40 and 46 days after sowing.
Assessed using the ureide method (Herridge and Peoples, 1990).
Fixed N - Seed N.
Values of P shown in parentheses.
+ 80
Nitrogen fixation and agriculture
such as fallow storage of water in the soil and
nutritional aspects would also be needed to be
taken into consideration. In a different approach, intercropping maize and ricebean was
found to increase P above the levels attained by
the monocropped legume due to competition
with maize for soil N (Table 4). The net result
was a higher total N yield of the intercrop (maize
plus legume) relative to combined weighted N
yield of maize and ricebean monocrops (Rerkasem et al., 1988; Rerkasem and Rerkasem,
1988).
Similar strategies might also be envisaged for
forage systems. Low levels of soil nitrate could
be maintained and, hence, N2-fixing potential
increased by growing legumes in competition
with vigorous grasses rather than as pure pasture
swards, or through regular removal of leguminous N via intensive grazing management or
periodic ‘cut-and-carry’ practices.
Rhizobia, inoculation and plant nodulation
There are three major groups of legumes which
can be distinguished on the basis of compatibility
with a range of strains of Rhizobium (Peoples et
al., 1989). At one extreme is a group of legumes
which can form an effective symbiosis with a
wide range of strains. Members of this group are
nodulated by ‘cowpea-type’ rhizobia and these
Rhizobium spp. are so widespread in tropical
soils that such legumes seldom respond to inoculation. Yet even within this supposedly ‘promiscuous’ group there can be some host-strain
specificity in terms of the symbiotic effectiveness
of the associations formed (Gibson et al., 1982).
At the other extreme are legumes with very
specific rhizobia1 requirements. These specificities are most relevant when the legume is
introduced to new areas. Response to inoculation of these legumes is usually successful provided adequate numbers of rhizobia are applied
at sowing. The third and intermediate group of
legumes nodulate with many strains of
Rhizobium, but effectively fix N2 with only a
limited number of them. In this case inoculation
and nodulation failures are more frequent because the inoculum strain is unable to compete
with the ineffective, but established soil populations of rhizobia.
There are a number of conditions under which
31
soils may be devoid of Rhizobium capable of
forming an effective symbiosis with a legume,
and which may warrant inoculation (Peoples et
al., 1989; Roughley, 1988):
i) The absence of the same or a symbiotically
related legume in the immediate past
history.
ii) Poor nodulation when the same crop was
grown previously.
iii) When the legume follows a non-leguminous crop in a rotation.
iv) In land reclamation.
v) When environmental conditions are unfavourable for Rhizobium survival (e.g. in
very acidic or alkaline soils, under prolonged flooding, or very hot, dry conditions prior to planting).
Even when inoculation may be appropriate as
a farming practice, it remains the exception
rather than the rule. Exacting technology is essential for the production and distribution of
inoculants (Roughley, 1988), and the success of
inoculation in the field depends upon the procedure used, operator competence, the presence of
toxic agrochemical, or be influenced by the soil
environment (Brockwell et al., 1988; Date,
1988). Acid soil factors are often notable in
effecting survival of inoculant rhizobia. The
selection and use of superior acid-tolerant strains
can influence the establishment, persistence and
competition by the introduced Rhizobium for
nodule sites and ultimately increase legume yield
(Date, 1988; Howieson et al., 1992; Moawad and
Bohlool, 1984).
In Australia and the U.S.A., legume inoculation has played a fundamental role in the establishment of legume-based pasture and cropping
systems, but less use has been made of inoculants elsewhere. In Latin America, only two
countries use inoculants to any extent and even
in Brazil, the largest producer of seed legumes,
common beans are fertilized with N rather than
inoculated (Freire, 1982).
Inoculation responses in tropical soils appear
confined to newly introduced legume species and
to crops such as soybean which have specific
Rhizobium requirements (Ayanaba, 1977). Typically, responses to inoculation are substantial
when soil nitrate is low, although if populations
of infective rhizobia already exist in high num-
32
Peoples and Craswell
Table 12. Effect of inoculation on nodulation, N2 fixation and productivity of soybean grown in various backgrounds of soil
nitrate and Bradyrhizobium japonicum a
Pd
Crop dry matter
(t ha -1)
0
72
334
0.11
0.44
0.61
0
4
50
129
146
Nodule massc
Treatment
Crop N e
(kg N ha-1)
Seed yield
(t ha-a)
4.9
6.8
7.0
66
168
195
1.7
3.3
3.0
0
0
0.17
8.9
8.0
8.5
205
196
213
3.2
2.9
3.3
-
9.8
9.3
258
241
2.2
2.0
(mg/plant)
B. japonicum free soil
Low nitrate
uninoculated
normal inoculation
high inoculation
c
a
High nitrate
uninoculated
normal inoculation
high inoculation
High B. japonicum soil
Low nitrate
uninoculated
normal inoculation
b
Derived from Herridge and Brockwell (1988).
Derived from Herridge et al. (1987).
c Sampled either 62 or 70 days after planting.
d Calculated by the ureide method according to Herridge and Peoples (1990). Values represent the mean of 4 sap samplings
between days 70 and 109.
e Sampled when shoot dry matter and N were at a maxima.
a
b
bers in the soil, they present a formidable barrier
to the establishment of inoculant Rhizobium
strains and inoculation responses may not be
apparent (Table 12). In high-nitrate soils, nodulation of, and N2 fixation by, inoculated plants
can be expected to be severly suppressed (Table
12). However, the establishment of populations
of Rhizobium can be improved, and nodulation
and N2 fixation increased in the presence of high
nitrate by very high rates of inoculation (Table
12; see also Brockwell et al., 1989).
Multiple regression analysis of relationships
between N2 fixation, P and rhizobia1 numbers
and soil nitrate (Herridge and Brockwell, 1988)
2
indicate that they can be highly correlated (r =
0.80; Peoples and Herridge, 1990). This serves to
illustrate the predictable and quantitative nature
of N2 fixation and demonstrates the importance
of the interaction between soil nitrate and rhizobial numbers in its regulation.
Plant breeding and selection
The challenge to improve the N2-fixation capacity of the legumes through selection and breeding
is complex because there are two components to
consider: the rhizobia and the host plant. Although host genotype x strain interactions have
been shown to occur with a number of agriculturally important legumes (Graham and Temple,
1984), many of the programs aimed at enhancing
N2 fixation have chosen to ignore this complication and have focused mainly on manipulation of
the host. The objectives of breeding programs
fall within three broad categories:
(i) Breeding for plant varieties which more successfully exploit strains of Rhizobium used as
inoculants or already present in soil. In the
U.S. A., large populations of soybean Rhizobium
have become established in soil with cropping, so
that only a small proportion of nodules is formed
on soybean by the inoculant, and yield responses
are rare. Research programs (e.g. Cregan et al.,
1989) aim to produce soybean cultivars that bypass the resident rhizobia in the soil to be nodulated by superior inoculant strains. A similar
strategy was employed for groundnut by Nambiar et al. (1984) who exploited host x strain
specificity with cultivar Robut 33-1 and strain NC
92 to obtain consistent yield responses (mean of
16%) in soils containing high numbers of infective rhizobia and where all uninoculated plants
were well nodulated.
A contrasting strategy is to develop varieties
Nitrogen fixation and agriculture
which can be effectively nodulated by the resident soil rhizobia. Nangju (1980) observed that
soybean genotypes from South East Asia nodulated successfully with the indigenous rhizobia in
Nigeria, and, that U.S.-bred cultivars nodulated
poorly without inoculation. Hybridization of the
Asian and U.S. types has resulted in high-yielding lines capable of fixing large amounts of N
without inoculation (Kueneman et al., 1984).
(ii) Breeding for improved plant yield. Whilst it
is possible to identify cultivars of legumes with
increased plant yields and therefore increased
N2-fixing potential (e.g. Hardarson et al., 1984;
Kumar Rao and Dart, 1987), the real challenge
is to select cultivars which maintain higher levels
of N2 fixation and P at the same level of yield.
Agronomic and environmental considerations
may limit the size of individual plants and the
duration of the crop must be taken into account.
(iii) Breeding for nitrate tolerance. Nitrate is
one of the most potent inhibitors of N, fixation
(Table 3; Streeter, 1988). Development of symbioses where P is maintained at near-maximum
levels in the presence of high soil nitrate could
provide an important advance in the improvement of N2 fixation by legumes.
Plant mutagenesis has been used to generate
phenotypes exhibiting greater nodule mass under
field conditions, or forming high numbers of
nodules in the presence of nitrate (e.g. ‘supernodulating’ soybean; Carroll et al., 1985). Extreme super-nodulating mutants can form up to
ten-fold more nodules in both the absence and
presence of nitrate, but this trait results in substantial reductions in root and shoot growth, and
despite higher P values, the mutants are only
capable of fixing more N than the original wildtype parent under very high-nitrate conditions
(Hansen et al., 1989).
Another approach that has been considered is
the selection of mutants with a reduced ability to
utilize nitrate, i.e. lowered nitrate reductase activity (Carroll and Gresshoff, 1986; Nelson et al.,
1983). From biochemical characterizations of
such mutants it would appear that legumes can
contain several constitutive and nitrate-inducible
nitrate reductase isoenzymes (Streit and Harper,
1986), and a genotype completely lacking nitrate
33
reductase has yet to be reported. Nodulation and
N2 fixation is still apparently sensitive to nitrate
in those mutants already identified as having
lowered nitrate reductase activities.
Species differ considerably in their symbiotic
tolerance to mineral N and sufficient natural
variation may already exist among legume lines
and cultivars, so that it might not be necessary to
resort to mutagenesis procedures to induce further variation for breeding purposes (see Gibson
and Harper, 1985; Hardarson et al., 1984). High
N,-fixing soybean lines of Korean origin (Betts
and Herridge, 1987; Herridge and Betts, 1988)
have been used as donor parents in a breeding
program in Australia (Peoples and Herridge,
1990) and results to date indicate that nitrate
tolerance may be under quantitative genetic control with a broad-sense heritability of between
0.24 and 0.72 (Rose et al., 1989).
Trends in research funding
The importance of biological N2 fixation in agriculture and in nature generally continues to
justify considerable research on basic aspects of
the process. This is reflected in the fact that of
nearly 400 papers given at the 5th International
Symposium on N, fixation, 90% were devoted to
basic biochemical, physiological and ecological
studies (Kokke and Shaw, 1984). The question
of how this developing body of information can
be put to practical use has vexed some donors
and the balance and interaction between basic
and applied research continues to be discussed
(Henzell, 1988; Kokke and Shaw, 1984). Some
policy-makers believe that the gap between the
fundamental knowledge gained from basic research and its field application is widening and
that more effort should be put into applied field
research. Others feel that it is better to wait for
further progress before forcing the pace of application. Perhaps the most important need is to
ensure effective flow of information between
basic and applied researchers.
Molecular genetics holds much promise as a
tool for enhancing biological N, fixation, but
public-sector donors will carefully consider their
role in relation to private-sector investment in
what potentially could be profitable and patent-
34
Peoples and Craswell
able technologies before committing funds. New
technologies for biological N2 fixation based on
genetic engineering have, however, been slow to
appear, despite considerable investment. Beringer (1989) believes that still too little is known
about genes and their interrelationships to currently be able to manipulate them to produce
superior organisms. He advocates the wider and
better use of proven microbial inoculants before
attempting to use untested ones. In this regard
the widespread failure to effectively transfer the
relatively simple Rhizobium inoculant technology to many developing countries is disturbing,
because legume inoculation has been the most
important practical contribution from research
on N, fixation. The lack of availability of viable
inoculants of effective strains of Rhizobium in
the tropics continues to deny farmers a simple
but profitable technology. More effort is needed
to encourage Governments and private firms not
only to produce and distribute inoculants, but
also to institute efficient quality-control programs (Roughley, 1988).
Agricultural research priorities are currently
being set in a dynamic agricultural, political and
ecologically minded environment which researchers on biological N, fixation should take
into account when framing research objectives.
Above all research should be carefully focussed
on important problems (Henzell, 1988). Increasingly economic analysis of potential benefits will
be necessary to convince research donors to
further invest in N2-fixation research. Considerable economic returns have been demonstrated
with appropriate application of inoculant (Brockwell et al., 1989), but other studies suggest that
the direct economic impact of N, fixation research may be relatively small (Anderson et al.,
1987).
Much has been learnt about the important
process of biological N, fixation and much is
understood about its potential benefits, particularly in tropical and sub-tropical regions where
productivity is often constrained by N supply.
The challenge however, remains to uncover
more of its mysteries and to exploit more effectively what is already known so that N, fixation
can be managed for the future benefit of
humanity.
Acknowledgements
Much of the unpublished information presented
in this chapter has resulted from collaborative
experimentation between ACIAR Project 8800
and Projects 8363, 8419, and 8731. The support
of ACIAR is gratefully acknowledged. We also
wish to thank P M Chalk for providing invaluable information on non-legume systems.
References
Agamuthu P and Broughton W J 1985 Nutrient cycling within
the developing oil palm-legume ecosystem. Agric. Ecosys.
Environ. 13, 111-123.
Alazard D and Becker M 1987 Aeschynomene as green
manure for rice. Plant and Soil 101, 141-143.
Alexander D B and Zuberer D A 1989 Impact of soil
environmental factors on rates of N2 fixation associated
with roots of intact maize and sorghum plants. Plant and
Soil 110, 303-315.
Anderson J R, Herdt R W, Scobie G M, Pray C E and Janke
H E 1987 International agricultural research centres: A
study of achievements and potential. Agric. Econ. Bull.
No. 32. Dept. Agric. Econ. and Bus. Manag., Univ. New
England, Armidale, Australia.
App A A, Watanabe I, Alexander M, Ventura W, Daez C,
Santiago T and De Datta S K 1980 Non-symbiotic nitrogen
fixation associated with the rice plant in flooded soils. Soil
Sci. 130, 283-289.
Awonaike K 0, Kumarasinghe K S and Danso S K A 1990
Nitrogen fixation and yield of cowpea ( Vigna unguicalata )
as influenced by cultivar and Bradyrhizobium strain. Field
Crops Res. 24, 163-171.
Ayanaba A 1977 Toward better use of inoculants In Biological Nitrogen Fixation in Farming Systems of the Tropics.
Eds. A Ayanaba and P J Dart. pp 181-187. Wiley, Chichester, UK.
Bandyopadhyay S K and De R 1986 Nitrogen relationships
and residual effects of intercropping sorghum with
legumes. J. Agric. Sci., Camb. 107, 629-632.
Becker M, Ladha J K and Ottow J C G 1990 Growth and
nitrogen fixation of two stem-nodulating legumes and their
effect as green manure on lowland rice. Soil Biol. Biochem. 22, 1109-1119.
Bergersen F J, Brockwell J, Gault R R, Morthorpe L.
Peoples M B and Turner G L 1989 Effects of available soil
nitrogen and rates of inoculation on nitrogen fixation by
irrigated soybeans and evaluation of d 15 N methods for
measurement. Aust. J. Agric. Res. 40, 763-780.
Bergersen F J, Turner G L, Gault R R, Chase D L and
Brockwell J 1985 The natural abundance of 15 N in an
irrigated soybean crop and its use for the calculation of
nitrogen fixation. Aust. J. Agric. Res. 36, 411-423.
Beri V, Meelu O P and Khind C S 1989 Biomas production.
Nitrogen fixation and agriculture
N
accumulation, symbiotic effectiveness and mineralization
of green manures in relation to yield of wetland rice. Trop.
Agric. (Trinidad) 66, 11-16.
Beringer J E 1989 Biotechnology in relation to beneficial
microorganisms. In Biotechnology Study Project Papers.
Summaries of Commissioned Papers. pp 20-31, World
Bank/ISNAR/AIDAB/ACIAR.
Betts J H and Herridge D F 1987 Isolation of soyben capable
of nodulation and nitrogen fixation under high levels of
nitrate supply. Crop Sci. 27, 1156-1161.
Blair G, Catchpoole D and Horne P 1990 Forage tree
legumes: Their management and contribution to the nitrogen economy of wet and humid tropical environments.
Adv. Agron. 44, 27-54.
Boddey R M 1987 Methods for quantification of nitrogen
fixation associated with Granineae. CRC Crit. Rev. Plant
Sci. 6, 209-266.
Boddey R M and Dobereiner J 1988 Nitrogen fixation associated with grasses and cereals: Recent results and perspectives for future research. Plant and Soil 108, 53-65.
Boddey R M and Victoria R L 1986 Estimation of biological
nitrogen fixation associated with Brachiaria and Paspalum
grasses using 15N labelled organic matter and fertilizer.
Plant and Soil 90, 265-292.
Boddey R M, Chalk P M, Victoria R L, Matsui E and
Dobereiner J 1983 The use of the 15N isotope dilution
technique to estimate the contribution of associated
biological nitrogen fixation to the nitrogen nutrition of
Paspalum notatum cv. batatais. Can. J. Microbiol. 29,
1036-1045.
Boddey R M, Urquiaga S, Suhet A R, Peres J R and Neves
M C P 1990 Quantification of the contribution of N2
fixation to field-grown grain legumes: A strategy for the
practical application of the 15N isotope dilution technique.
Soil Biol. Biochem. 22, 649-655.
Brockwell J, Gault R R, Herridge D F, Morthorpe L J and
Roughley R J 1988 Studies on alternative means of legume
inoculation: Microbial and agronomic appraisals of commercial procedures for inoculating soybeans with Bradyrhizobium japonicum. Aust. J. Agric. Res. 39, 965-972.
Brockwell J. Gault R R, Morthorpe L J, Peoples M B.
Turner G L and Bergersen F J 1989 Effects of soil nitrogen
status and rate of inoculation on the establishment of
populations of Bradyrhizobium japonicum and on the
nodulation of soybeans. Aust. J. Agric. Res. 40, 753-762.
Bryan W B and Velasquez E R 1982 Legume establishment
in pangolagrass pastures in the humid tropics. In Biological
Nitrogen Fixation Technology for Tropical Agriculture.
Eds. P H Graham and S C Harris, pp 355-360. CIAT,
Cali, Colombia.
Burns R C and Hardy R W F 1975 Nitrogen Fixation in
Bacteria and Higher Plants. Springer-Verlag, Berlin, Germany. 189 p.
Buresh R J and De Datta S K 1991 Nitrogen dynamics and
management in rice-legume cropping systems. Adv.
Agron. 45, 1-59.
Carroll B J and Gresshoff P M 1986 Isolation and initial
characterization of constitutive nitrate reductase-deficient
35
mutants NR 328 and NR 345 of soybean (Glycine max).
Plant Physiol. 81, 572-576.
Carroll B J, McNeil D L and Gresshoff P M 1985 A
supernodulating and nitrate-tolerant symbiotic (nts) soybean mutant. Plant Physiol. 78, 34-40.
Chalk P M 1991 The contribution of associative and symbiotic nitrogen fixation to the nitrogen nutrition of nonlegumes. Plant and Soil 132. 29-39.
Chandel A S. Pandey K N and Saxena S C 1989 Symbiotic
nitrogen fixation and nitrogen benefits by nodulated soybean (Glycine max (L.) Merrill) to interplanted crops in
northern India. Trop. Agric. (Trinidad) 66, 73-77.
Chapman A L and Myers R J K 1987 Nitrogen contributed
by grain legumes to rice grown in rotation on the Cununurra soils of the Ord irrigation area, Western Australia.
Aust. J. Exp. Agric. 27. 155-163.
Cornet F, Otto C. Rinaudo G, Diem H G and Dommergues
Y 1985 Nitrogen fixation by Acacia holosericea grown in
field-simulating conditions. Acta Oecol. Oecol. Plant. 6,
211-218.
Craswell E T and Tangendjaja B 1985 Shrub Legume Research in Indonesia and Australia. ACIAR Proc. No. 3,
ACIAR. Canberra, Australia. 42 p.
Craswell E T, Loneragan J F and Keerati-Kasikorn P 1987
Mineral constraints to food legume crop production in
Asia. In Food Legume Improvement for Asian Farming
Systems. Eds. E S Wallis and D E Byth. pp 99-111.
ACIAR Proc. No. 18. ACIAR, Canberra. Australia.
Cregan P B. Keyser H H and Sadowsky M J 1989 Soybean
genotype restricting nodulation of a previously unrestricted
serocluster 123 bradyrhizobia. Crop Sci. 29, 307-312.
Dakora F D. Aboyinga R A, Mahama Y and Apaseku J 1987
Assessment of N2 fixation in groundnut ( Arachis hypogaea
L.) and cowpea ( Vigna unguicalata L. Walp) and their
relative N contribution to a succeeding maize crop in
northern Ghana. MIRCEN J. 3, 389-399.
Dart P J 1986 Nitrogen fixation associated with non-legumes
in agriculture. Plant and Soil 90. 303-334.
Date R A 1988 Problems in application of BNF in field crops.
In Biotechnology of Nitrogen Fixation in the Tropics. Eds.
Z H Shamsuddin. W M W Othman, M Marziah and J
Sundram. pp 283-300. Universiti Pertanian Malaysia, Serdang, Malaysia.
De Datta S K and Buresh R J 1989 Integrated nitrogen
management in irrigated rice. Adv. Soil Sci. 10, 143-169.
De Freitas J R, Victoria R L, Ruschel A P and Vose P B
1984 Estimation of N2 fixation by sugarcane. Saccharum
sp.. and soybean, Glycine max, grown in soil with 15Nlabelled organic matter. Plant and Soil 82, 257-261.
Doughton J A and MacKenzie J 1984 Comparative effects of
black and green gram (mung beans) and grain sorghum on
soil mineral nitrogen and subsequent grain sorghum yields
on the eastern Darling Downs. Aust. J. Exp. Agric. Anim.
Hush. 24. 244-249.
Dommergues Y R and Diem H G 1982 Microbiology of
Tropical Soils and Plant Productivity. Martinus Nijhoff/Dr
W Junk Publ., The Hague, The Netherlands. 328 p.
Duque F F, Neves M C P, Franco A A, Victoria R L and
36
Peoples and Craswell
Boddey R M 1985 The response of field grown Phaseolus
vulgaris to Rhizobium inoculation and the quantification of
N2 fixation using 15N. Plant and Soil 88, 333-343.
Eaglesham A R J, Ayanaba A, Ranga Rao V and Eskew D
L 1981 Improving the nitrogen nutrition of maize by
intercropping with cowpea. Soil Biol. Biochem. 13, 169171.
Eaglesham A R J, Ayanaba A, Rango Rao V and Eskew D
L 1982 Mineral N effects on cowpea and soybean crops in a
Nigerian soil. II. Amounts of N fixed and accrual to the
soil. Plant and Soil 68, 183-192.
Eskew D L and Kumarasinghe K S 1990 Azolla conserves
fertilizer nitrogen in rice production. Agron. Abstracts. p
248.
Freire J R 1982 Research into the Rhizobium/Leguminosae
symbiosis in Latin America. Plant and Soil 67, 227-239.
Fujii T, Huang Y D, Higashitani A, Nishimura Y, Iyama S,
Hirota Y, Yoneyama T and Dixon R A 1987 Effect of
inoculation with Klebsiella oxytoca and Enterobacter
cloacae on dinitrogen fixation by rice-bacteria asociations.
Plant and Soil 103, 221-226.
Gauthier D, Diem H G, Dommergues Y R and Ganry F
1985 Assessment of N2 fixation by Casuarina equisetifolia
inoculated with Frankia ORS021001 using 15N methods.
Soil Biol. Biochem. 17, 375-379.
Gibson A H and Harper J E 1985 Nitrate effect on nodulation of soybean by Bradyrhizobiurn japonicum. Crop Sci.
25, 497-501.
Gibson A H, Dreyfus B L and Dommergues Y R 1982
Nitrogen fixation by legumes in the tropics. In Microbiology of Tropical Soils and Plant Productivity. Eds. Y R
Dommergues and H G Diem pp 37-73. Martinus Nijhoff
Publ., The Hague, The Netherlands.
Gibson A H, Roper M M and Halsall D M 1988 Nitrogen
fixation not associated with legumes. In Advances in Nitrogen Cycling in Agricultural Ecosystems. Ed. J R Wilson.
pp 66-88. C A B International Wallingford, UK.
Giller K E, Nambiar P T C, Srinivasa Rao B, Dart P J and
Day J M 1987 A comparison of nitrogen fixation in
genotypes of groundnut ( Arachis hypogaea L.) using 15Nisotope dilution. Biol. Fertil. Soils 5, 23-25.
George T, Singleton P W and Bohlool B B 1988 Yield, soil
nitrogen uptake, and N2 fixation by soybean from four
maturity groups grown at three elevations. Agron. J. 80,
563-567.
Graham P H and Temple S R 1984 Selection for improved
nitrogen fixation in Glycine max (L.) Merr. and Phaseolus
vulgaris L. Plant and Soil 82, 315-328.
Hansen A P, Peoples M B, Gresshoff P M, Atkins C A, Pate
J S and Carroll B J 1989 Symbiotic performance of supernodulating soybean ( Glycine max (L.) Merrill) mutants
during development on different nitrogen regimes. J. Exp.
Bot. 40, 715-724.
Haque I and Jutzi S 1985 Potentials of and limitations to the
biological nitrogen contribution from forage legumes in
sub-saharan Africa. In Biological Nitrogen Fixation in
Africa. Eds H Ssali and S O Keya, pp 340-369. The
Nairobi Rhizobium MIRCEN, Kenya.
Hardarson G, Zapata F and Danso S K A 1984 Effect of
plant genotype and nitrogen fertilizer on symbiotic nitro-
gen fixation by soybean cultivars. Plant and Soil 82, 397405.
Hearne A B 1986 Effect of preceding crop on the nitrogen
requirements of irrigated cotton ( Gossypium hirsutum L.)
on a vertisol. Field Crops Res. 13, 159-175.
Henzell E F 1988 The role of biological nitrogen fixation
research in solving problems in tropical agriculture. Plant
and Soil 108, 15-21.
Henzell E F and Vallis I 1977 Transfer of nitrogen between
legumes and other crops. In Biological Nitrogen Fixation
in Farming Systems of the Tropics. Eds. A Ayanaba and P
J Dart. pp 73-88. Wiley, Chichester, UK.
Herridge D F and Bergersen F J 1988 Symbiotic nitrogen
fixation. In Advances in Nitrogen Cycling in Agricultural
Ecosystems. Ed. J R Wilson. pp 46-65. C A B International, Wallingford, UK.
Herridge D F and Betts J H 1988 Field evaluation of soybean
genotypes selected for enhanced capacity to nodulate and
fix nitrogen in the presence of nitrate. Plant and Soil 110,
129-135.
Herridge D F and Brockwell J 1988 Contributions of fixed
nitrogen and soil nitrate to the nitrogen economy of irrigated soybean. Soil Biol. Biochem. 20, 711-717.
Herridge D F and Peoples M B 1990 Ureide assay for
measuring nitrogen fixation by nodulated soybean calibrated by 15N methods. Plant Physiol. 93, 495-503.
Herridge D F, Bergersen F J and Peoples M B 1990 Measurement of nitrogen fixation by soybean in the field using
the ureide and natural abundance methods. Plant Physiol.
93, 708-716.
Herridge D F, Roughley R J and Brockwell J 1987 Low
survival of Rhizobium japonicum inoculant leads to reduced nodulation, nitrogen fixation and yield of soybean in
the current crop but not in the subsequent crop. Aust. J.
Agric. Res. 38, 75-82.
Hogberg P and Kvarnstrom M 1982 Nitrogen fixation by the
woody legume Leucaena leucocephala in Tanzania. Plant
and Soil 66, 21-28.
Howieson J G, Ewing M A, Thorn C W and Revell C K 1992
Increased yield in annual species of Medicago grown in
acidic soil in response to inoculation with acid tolerant
Rhizobium meliloti. Plant and Soil (In press).
Hubbell D H 1988 Extension/transfer of BNF technology In
Nitrogen Fixation by Legumes in Mediterranean Agriculture. Eds. D P Beck and L A Materon. pp 367-370.
Martinus Nijhoff Publ., Dordrecht, The Netherlands.
Hughes R M and Herridge D F 1989 Effect of tillage on
yield, nodulation and nitrogen fixation of soybean in far
north-coastal New South Wales. Aust. J. Exp. Agric. 29.
671-677.
Kang B T 1988 Nitrogen cycling in multiple cropping systems. In Advances in Nitrogen Cycling in Agricultural
Ecosystems. Ed. J R Wilson. pp 333-348. C A B International, Wallingford, UK.
Kokke R and Shaw W P 1984 Report from the policy
meeting on nitrogen fixation research. In Advances in
Nitrogen Fixation Research. Eds. C Veeger and W E
Newton. pp XI-XII. Martinus Nijhoff/ Dr W Junk Publ.,
The Hague, The Netherlands.
Kucey R M N, Chaiwanakupt P, Arayangkool T, Snitwongse
Nitrogen fixation and agriculture
P, Siripaibool C, Wadisirisuk P and Boonkerd N 1988
Nitrogen fixation (15N dilution) with soybeans under Thai
field conditions. II. Effect of herbicides and water application schedule. Plant and Soil 108, 87-92.
Kueneman E A, Root W R, Dashiell K E and Hohenberg J
1984 Breeding soybeans for the tropics capable of nodulating effectively with indigenous Rhizobium spp. Plant and
Soil 82, 387-396.
Kumar Rao J V D K and Dart P J 1987 Nodulation, nitrogen
fixation and nitrogen uptake in pigeonpea ( Cajanus cajan
(L.) Millsp) of different maturity groups. Plant and Soil 99,
255-266.
Kumar Rao J V D K, Dart P J and Sastry P V S S 1983
Residual effect of pigeonpea ( Cajanus cajan ) on yield and
nitrogen response of maize. Expl. Agric. 19, 131-141.
Kumar Rao J V D K, Thompson J A, Sastry P V S S. Giller
K E and J M Day 1987 Measurement of N2 fixation in
field-grown pigeonpea [ Cajanus cajan (L.) Millsp.] using
15N-labelled fertilizer. Plant and Soil 101, 107-113.
Kumarisinghe K S, Zapata F, Kovacs G, Eskew D L
and Danso S K A 1986 Evaluation of the availability of
Azolla -N and urea-N to rice using 15N. Plant and Soil 90,
293-299.
Ladha J K, Watanabe I and Saono S 1988a Nitrogen fixation
by leguminous green manure and practices for its enhancement in tropical lowland rice. In Sustainable Agriculture:
Green Manure in Rice Farming. pp 165-183. IRRI, Manila, Philippines.
Ladha J K, Tirol-Padre A, Punzalan G C, Watanabe I and
De Data S K 1988b Ability of wetland rice to stimulate
biological nitrogen fixation and utilize soil nitrogen. In
Nitrogen Fixation Hundred Years After. Eds. H Bothe. F
de Bruijn and W E Newton. pp 747-752. Gustav Fischer.
Stuttgart, Germany.
Ladha J K. Tirol-Padre A, Punzalan G C. Garcia M and
Watanabe I 1989 Effect of inorganic N and organic fertilizers on nitrogen-fixing (acetylene-reducing) activity associated with wetland rice plants In Nitrogen Fixation with
Non-Legumes. Eds. F A Skinner, R M Boddey and I
Fendrik. pp 263-272. Kluwer Academic Publ., Dordrecht,
The Netherlands.
Lal R 1989 Conservation tillage for sustainable agriculture:
Tropics versus temperature environments. Adv. Agron.
42, 85-197.
Langkamp P J, Swinden L B and Dalling M J 1979 Nitrogen
fixation (acetylene reduction) by Acacia pellita on areas
restored after mining at Groote Eylandt, Northern territory. Aust. J. Bot. 27, 353-361.
Lima E, Boddey R M and Dobereiner J 1987 Quantification
of biological nitrogen fixation associated with sugar cane
using a 15N aided nitrogen balance. Soil Biol. Biochem. 19,
165-170.
Little D A, Kompiang S and Petheram R J 1989 Mineral
composition of Indonesian ruminant forages. Trop. Agric.
(Trinidad) 66, 33-37.
MacColl D 1989 Studies on maize ( Zea mays ) at Bunda,
Malawi. II. Yield in short rotations with legumes. Exp.
Agric. 25, 367-374.
Malik K A, Bilal R, Azam F and Sajjad M I 1988 Quantification of N2 fixation and survival of inoculated diazotrophs
37
associated with the roots of kallar grass. Plant and Soil 108,
43-51.
McCown R L, Cogle A L, Ockwell A P and Reeves T G
1988 Nitrogen supply to cereals in legume ley systems
under pressure. In Advances in Nitrogen Cycling in Agricultural Ecosystems. Ed. J R Wilson. pp 292-314. C A B
International. Wallingford, UK.
Michiels K, Vanderleyden J and Van Goo1 A 1989 Azospirillurn-plant root associations: A review. Biol. Fertil.
Soils. 8. 356-368.
Miranda C H B. Urquiaga S and Boddye R M 1990 Selection
of ecotypes of Panicum maximum for associated biological
nitrogen fixation using the 15N isotope dilution technique.
Soil Biol. Biochem. 22, 657-663.
Moawad H and Bohlool B B 1984 Competition among
Rhizobium spp. for nodulation of Leucaena leucocephala
in two tropical soils. Appl. Environ. Microbiol. 48, 5-9.
Moong F A 1986 Forages in integrated food cropping systems. In Forages in Southeast Asian and South Pacific
Agriculture. Eds. G J Blair, D A Ivory and T R Evans. pp
152-156. ACIAR Proceedings No. 12, ACIAR, Canberra,
Australia.
Myers R J K and Wood I M 1987 Food legumes in the
nitrogen cycle of farming systems. In Food Legume Improvement for Asian Farming Systems. Eds. E S Wallis
and D E Byth. pp 46-52. ACIAR Proc. No. 18, ACIAR,
Canberra. Australia.
Nambiar P C T, Dart P J, Srinivasa Rao B and Ravishankar
H N 1984 Response of groundnut ( Arachis hypogaea L.) to
Rhizobium inoculation. Oleagineux 39, 150-154.
Nangju D 1980 Soybean response to indigenous Rhizobia as
influenced by cultivar origin. Agron. J. 72, 403-406.
Ndoye I and Dreyfus B 1988 N2 fixation by Sesbania rostrata
and Sesbania sesban estimated using 15N and total N
difference methods. Soil Biol. Biochem. 20, 209-213.
Nelson R S, Ryan S A and Harper J E 1983 Soybean mutants
lacking constitutive nitrate reductase. I. Selection and initial plant characterization. Plant Physiol. 72, 503-509.
Norhayati M, Mohd Noor S, Chong K, Faizah A W,
Herridge D F. Peoples M B and Bergersen F J 1988
Adaptation of methods for evaluating N2 fixation in food
legumes and legume cover crops. Plant and Soil 108,
143-150.
Ofori F and Stern W R 1987 Cereal-legume intercropping
systems. Adv. Agron. 41. 41-90.
Ofori F, Pate J S and Stern W R 1987 Evaluation of N2
fixation and nitrogen economy of a maize/cowpea intercrop system using 15N dilution methods. Plant and Soil 102,
149-160.
Palm C A and Sanchez P A 1991 Nitrogen release from the
leaves of some tropical legumes as affected by their lignin
and polyphenolic contents. Soil Biol. Biochem. 23, 83-88.
Palm 0, Weerakoon W L, De Silva M A P and Rosswall T
1988 Nitrogen mineralization of Sesbania sesban used as
green manure for lowland rice in Sri Lanka. Plant and Soil
108, 201-209.
Pareek R P, Ladha J K and Watanabe I 1990 Estimating
nitrogen fixation by Sesbania rostrata and S. cannabina
(stn. S. aculeata ) in lowland rice soil by the 15N dilution
method. Biol. Fert. Soils. 10, 77-78.
38
Peoples and Craswell
Patra D D, Sachdev M S and Subbiah B V 1986 15 N studies
on the transfer of legume-fixed nitrogen to associated
cereals in intercropping systems. Biol. Fertil. Soils 2, 165171.
Paul E A 1988 Towards the year 2000: directions for future
nitrogen research. In Advances in Nitrogen Cycling in
Agricultural Ecosystems. Ed. J R Wilson. pp 417-425. C
A B International, Wallingford, UK.
Peoples M B and Dalling M J 1988 The interplay between
proteolysis and amino acid metabolism during senescence
and nitrogen reallocation. In Senescence and Aging in
Plants. Eds. L D Nooden and A C Leopold. pp 181-217.
Academic Press, Orlando, FL. USA.
Peoples M B and Herridge D F 1990 Nitrogen fixation by
legumes in tropical and sub-tropical agriculture. Adv.
Agron. 44, 155-223.
Peoples M B, Faizah A W, Rerkasem B and Herridge D F
1989. Methods for Evaluating Nitrogen Fixation by Nodulated Legumes in the Field. ACIAR Monograph No. 11,
ACIAR, Canberra, Australia. 76 p.
Peoples M B, Herridge D F and Bergersen F J 1988 Measurement of nitrogen fixation in crop and shrub legumes.
In Sustainable Agriculture: Green Manure in Rice Farming. pp 223-237. IRRI, Manila, Philippines.
Peters G A and Meeks J C 1989 The Azolla-Anabaena
symbiosis: Basic biology. Annu. Rev. Plant Physiol. Plant
Molec. Biol. 40, 193-210.
Poth M, La Favre J S and Focht D D 1986 Quantification by
direct 15 N dilution of fixed N2 incorporated into soil by
Cajanus cajan (pigeon pea). Soil Biol. Biochem. 18, 125127.
Rachie K O and Roberts L M 1974 Grain legumes of the
lowland tropics. Adv. Agron. 26, 1-132.
Rennie R J, Rennie D A, Siripaibool C, Chaiwanakupt P,
Boonkerd N and Snitwongse P 1988 N2 fixation in Thai
soybeans: Effects of tillage and inoculation on 15 N-determined N2 fixation in recommended cultivars and advanced
breeding lines. Plant and Soil 112, 183-193.
Rerkasem B, Rerkasem K, Peoples M B, Herridge D F and
Bergersen F J 1988. Measurement of N2 fixation in maize
( Zea mays L.) - ricebean ( Vigna umbellata [Thunb.] Ohwi
and Ohashi) intercrops. Plant and Soil 108, 125-135.
Rerkasem K and Rerkasem B 1988 Yields and nitrogen
nutrition of inter-cropped maize and ricebean ( Vigna umbellata [Thunb.] Ohwi and Ohashi). Plant and Soil 108,
151-162.
Rice C W and Smith M S 1982 Denitrification in no-till and
plowed soils. Soil Sci. Soc. Am. J. 46, 1168-1173.
Rinaudo G, Dreyfus B and Dommergues Y 1983 Sesbania
rostrata green manure and the nitrogen content of rice crop
and soil. Soil Biol. Biochem. 15, 111-113.
Roskoski J P, Montano J, van Kessel C and Castilleja G 1982
Nitrogen fixation by tropical woody legumes: Potential
source of soil enrichment. In Biological Nitrogen Fixation
Technology for Tropical Agriculture. Eds. P H Graham
and S C Harris. pp 447-454. CIAT, Cali, Colombia.
Rose J A, Betts J H and Herridge D F 1989 Breeding for
improved nitrogen fixation in soybeans. In World Soybean
Research Conf. IV. Ed. A J Pascale. pp 1176-1181. Asoc.
Argentina De La Soja, Buenos Aires, Argentina.
Roughley R J 1988 Commercial applications of biological
dinitrogen fixation. In Biotechnology of Nitrogen Fixation
in the Tropics. Eds. Z H Shamsuddin, W M W Othman, M
Marziah and J Sundram. pp 147-154. Universiti Pertanian
Malaysia, Serdang, Malaysia.
Ruschel A P. Vose P B, Matsui E, Victoria R L and Saito S
M T 1982 Field evaluation of N, fixation and N-utilization
by Phaseolus bean varieties determined by 15 N isotope
dilution. Plant and Soil 65, 397-407.
Sandhu J, Sinha M and Ambasht R S 1990 Nitrogen release
from decomposing litter of Leucaena leucocephala in the
dry tropics. Soil Biol. Biochem. 22, 859-863.
Sanford J O and Hairston J E 1984 Effects of N fertilization
on yield, growth, and extraction of water by wheat following soybeans and grain sorghum. Agron. J. 76 623-627.
Sanginga N. Mulongoy K and Ayanaba A 1988 Nitrogen
contribution of Leucaena/Rhizobium symbiosis to soil and
a subsequent maize crop. Plant and Soil 112, 137-141.
Sanginga N, Mulongoy K and Ayanaba A 1989 Nitrogen
fixation of field-inoculated Leucaena leucocephala (Lam.)
de Wit estimated by the 15 N and the difference methods.
Plant and Soil 117, 269-274.
Shivaram S, Rai P V and Hegde S V 1988 Nitrogen fixation
and nitrogen balance studies in Rhizobium-VA mycorrhiza
inoculated legume-grass association by 15 N isotope dilution
technique under a field condition. Plant and Soil 111.
11-16.
Singh S B and Verma B S 1985. Effect of kharif grain
legumes on nitrogen economy in succeeding crop of wheat.
Indian J. Agron. 30, 397-400.
Sisworo W H, Mitrosuhardjo M M, Rasjid H and Myers R J
K 1990 The relative roles of N fixation, fertilizer, crop
residues and soil in supplying N in multiple cropping systems
in a humid, tropical upland cropping system. Plant and Soil
121, 73-82.
Skinner F A, Boddey R M and Fendrik I 1989 Nitrogen
Fixation with Non-Legumes. Kluwer Academic Publ.,
Dordrecht, The Netherlands. 336 p.
Sougoufara B, Danso S K A, Diem H G and Dommergues Y
R 1990 Estimating N2 fixation and N derived from soil
by Casuarina equistifolia using labelled 15 N fertilizer:
Some problems and solutions. Soil Biol. Biochem. 22.
695-701.
Ssali H and Keya S O 1984a Phosphorus and cultivar effects
on nodulation, growth, dinitrogen fixation and yield of
cowpea ( Vigna unguiculata ). E. Afr. Agric. For. J. 45.
188-193.
Ssali H and Keya S O 1984b Nitrogen level and cultivar
effects on nodulation, dinitrogen fixation and yield of grain
legumes. I. Cowpea cultivars. E. Afr. Agric. For. J. 45,
247-254.
Ssali H and Keya S O 1986 The effects of phosphorus and
nitrogen fertilizer level on nodulation, growth and dinitrogen fixation of three bean cultivars. Trop. Agric.
(Trinidad) 63, 105-109.
Streeter J 1988 Inhibition of legume nodule formation and N2
fixation by nitrate. CRC Critical Rev. Plant Sci. 7, 1-23.
Streit L and Harper J E 1986 Biochemical characterization of
soybean mutants lacking constitutive NADH: Nitrate reductase. Plant Physiol. 81, 593-596.
Nitrogen fixation and agriculture
Suwanarit A, Suwannarat C, and Chotechaungmanirat S
1986 Quantities of fixed N and effects of grain legumes on
following maize, and N and P status of soil as indicated by
isotopes. Plant and Soil 93, 249-258.
Swatdee P, Nammuang C, Chaiwanakupt P, Siripaibool C.
Chaiwatana B and and Luangsirorat S 1988 Isotopic studies
of nitrogen fixation, nitrogen cycling in Azolla and bluegreen algae. In Biotechnology of Nitrogen Fixation in the
Tropics. Eds. Z H Shamsuddin, W M W Othman, M
Marziah and J Sundram. pp 155-164. Universiti Pertanian
Malaysia, Serdang, Malaysia.
Thomas G W, Blevins R L, Phillips R E and McMahon M A
1973. Agron. J. 63, 736-739.
Trinick M J and Hadobas P A 1990 Symbiotic effectiveness
of Bradyrhizobium strains isolated from Parasponia and
tropical legumes on Parasponia host species. Plant and Soil
124, 117-126.
Urquiaga S. Botteon P B L and Boddey R M 1989 Selection
of sugar cane cultivars for associated biological nitrogen
fixation using 15N-labelled soil. In Nitrogen Fixation with
Non-Legumes. Eds. F A Skinner, R M Boddey and I
Fendrik. pp 311-319. Kluwer Academic Publ.. Dordrecht.
The Netherlands.
Vallis I 1972 Soil nitrogen changes under continously grazed
legume-grass pastures in subtropical coastal Queensland.
Aust. J. Exp. Agric. Anim. Husb. 12, 495-501.
Vallis I and Gardener C J 1985 Effect of pasture age on the
efficiency of nitrogen fixation by 10 accessions of Stylosanthes spp. Aust. J. Exp. Agric. 25, 70-75.
Vallis I and Jones R J 1973 Net mineralization of nitrogen in
leaves and leaf litter of Desmodium intortum and
Phaseolus atropurpureus mixed with soil. Soil Biol. Biochem. 5. 391-398.
Vallis I, Henzell E F and Evans T R 1977 Uptake of soil
nitrogen by legumes in mixed swards. Aust. J. Agric. Res.
28, 413-425.
Van Kessel C and Roskoski J P 1988 Row spacing effects on
N, fixation, N-yield and soil N uptake of intercropped
cowpea and maize. Plant and Soil 111, 17-23.
Ventura W and Watanabe I 1983 15'N dilution technique of
assessing the contribution of nitrogen fixation to rice plant.
Soil Sci. PI. Nutr. 29, 123-131.
Watanabe I 1982 Azolla-Anabaena symbiosis-its physiology
and use in tropical agriculture. In Microbiology of Tropical
39
Soils and Plant Productivity. Eds. Y R Dommergues and H
G Diem. pp 169-185. Martinus Nijhoff Publ., The Hague,
The Netherlands.
Watanabe I and Liu C 1992 Improving nitrogen-fixing systems and integrating them into sustainable rice farming.
Plant and Soil 141, 57-67.
World Commission on Environment and Development
(WCED) 1987 Our Common Future. Oxford University
Press. Oxford, UK.
Wijeratne W B and Nelson AI 1987 Utilisation of legumes as
food. In Food Legume Improvement for Asian Farming
Systems. Eds. E S Wallis and D E Byth. pp 183-192.
ACIAR Proc. No. 18. ACIAR, Canberra, Australia.
Wilson G F, Kang B T and Mulongoy K 1986 Alley cropping:
Trees as sources of green-manure and mulch in the tropics.
Biological Agric. Hortic. 3, 251-267.
Wood 1 M and Myers R J K 1987 Food legumes in farming
systems in the tropics and subtropics. In Food Legume
Improvement for Asian Farming Systems. Eds. E S Wallis
and D E Byth. pp 34-45. ACIAR Proc. No. 18. ACIAR,
Canberra. Australia.
Yoneyama T. Nambiar P T C, Lee K K. Srinivasa Rao B and
Williams J H 1990a Nitrogen accumulation in three
legumes and two cereals with emphasis on estimation of N2
fixation in the legumes by the natural 15N abundance
technique. Bid. Fertil. Soils.
Yoneyama T, Murakami T. Boonkerd N. Wadisirisuk P,
Siripin S and Kuono K 1990b Natural 15N abundance in
shrub and tree legumes, Casuarina, and non N2 fixing
plants in Thailand. Plant and Soil 128, 287-292.
Yoneyama T. Ladha J K and Watanabe I J 1987 Nodule
bacteroids and Anabaena: Natural 15N enrichment in the
legume-Rhozobium and Azolla-Anabaena symbiotic systems. J. Plant Physiol. 127, 251-259.
Yoo I D. Fujii T. Sano Y. Komagata K. Yoneyama T, Iyama
S and Hirota Y 1986 Dinitrogen fixation by rice-Klebsiella
associations. Crop Sci. 26, 297-301.
Zaharah A R, Sharifuddin H A H, Razley M N and Mohd
Saidi A K 1986a Measurement of nitrogen fixed by
Pueraria phaseoloides by 15N dilution technique. Pertanika
9, 45-39.
Zaharah A R. Sharifuddin A A H and Subramaniam R
1986b Nitrogen fixation by Leucaena leucocephala as measured by 15N dilution technique. Pertanika 9, 17-22.
Plant and Soil 141: 41-55, 1992.
© 1992 Kluwer Academic Publishers. Printed in the Netherlands.
PLSO SA03
Biological N2 Fixation in wetland rice fields: Estimation and contribution
to nitrogen balance
P.A. ROGER
1
and J.K. LADHA
International Rice Research Institute (IRRI), P.O. Box 933, Manila, Philippines. 1 Institut Français de
Recherche Scientifique pour le Développement en Coopération (ORSTOM), France (Visiting Scientist
at IRRI)
Key words: acetylene reducing activity, ARA, Azolla, cyanobacteria, green manure, heterotrophic
BNF, methods, N2 fixation, review, rice
Table of contents
Abstract
Introduction
Improvements and new methodological approaches
Acetylene-reducing activity measurements
Estimation of associative BNF
Estimation of photodependent BNF
Estimation of BNF by aquatic legumes
15
N2 incorporation
15
N dilution method
Difference in natural 15 N abundance ( d 15 N)
Estimation of BNF by various agents
Total heterotrophic BNF
Associative BNF and varietal differences
BNF associated with straw
Blue-green algae
Azolla
Leguminous green manure
Nitrogen balance and BNF contribution in wetland rice
Conclusions
Acknowledgements
References
Abstract
This paper 1) reviews improvements and new approaches in methodologies for estimating biological N2
fixation (BNF) in wetland soils, 2) summarizes earlier quantitative estimates and recent data, and 3)
discusses the contribution of BNF to N balance in wetland-rice culture.
Measuring acetylene reducing activity (ARA) is still the most popular method for assessing BNF in
42
Roger and Ladha
rice fields. Recent studies confirm that ARA measurements present a number of problems that may
render quantitative extrapolations questionable. On the other hand, few comparative measures show
ARA's potential as a quantitative estimate. Methods for measuring photodependent and associative
ARA in field studies have been standardized, and major progress has been made in sampling
procedures. Standardized ARA measurements have shown significant differences in associative N2
fixation among rice varieties.
The 15N dilution method is suitable for measuring the percentage of N derived from the atmosphere
(% Ndfa) in legumes and rice. In particular, the 15N dilution technique, using available soil N as
control, appears to be a promising method for screening rice varieties for ability to utilize biologically
fixed N. Attempts to adapt the 15N dilution method to aquatic N2 fixers (Azolla and blue-green algae
[BGA]) encountered difficulties due to the rapid change in 15N enrichment of the water.
Differences in natural 15N abundance have been used to show differences among plant organs and
species or varieties in rice and Azolla, and to estimate Ndfa by Azolla, but the method appears to be
semi-quantitative.
Recent pot experiments using stabilized 15N-labelled soil or balances in pots covered with black cloth
indicate a contribution of 10-30 kg N ha-1 crop-1 by heterotrophic BNF in flooded planted soil with no
or little N fertilizer used.
Associative BNF extrapolated from ARA and 15N incorporation range from 1 to 7 kg N ha-1 crop-1.
Straw application increases heterotrophic and photodependent BNF. Pot experiments show N gains of
2-4 mg N g-1 straw added at 10 tons ha-1.
N2 fixation by BGA has been almost exclusively estimated by ARA and biomass measurements.
Estimates by ARA range from a few to 80 kg N ha-1 crop-1 (average 27 kg). Recent extensive
measurements show extrapolated values of about 20 kg N ha-1 crop-1 in no-N plots, 8 kg in plots with
broadcast urea, and 12 kg in plots with deep-placed urea.
Most information on N2 fixed by Azolla and legume green manure comes from N accumulation
measurements and determination of % Ndfa. Recent trials in an international network show standing
crops of Azolla averaging 30-40 kg N ha-1 and the accumulation of 50-90kg N ha-1 for two crops of
Azolla grown before and after transplanting rice. Estimates of % Ndfa in Azolla by 15N dilution and
delta 15N methods range from 51 to 99%. Assuming 50-80% Ndfa in legume green manures, one crop
can provide 50-100kg N ha-1 in 50 days. Few balance studies in microplots or pots report extrapolated
N gains of 150-250 kg N ha-1 crop-1.
N balances in long-term fertility experiments range from 19 to 98 kg N ha-1 crop-1 (average 50 kg N)
in fields with no N fertilizer applied. The problems encountered with ARA and 15N methods have
revived interest in N balance studies in pots. Balances are usually highest in flooded planted pots
exposed to light and receiving no N fertilizer; extrapolated values range from 16 to 70 kg N ha-1 crop-1
(average 38 kg N). A compilation of balance experiments with rice soil shows an average balance of
about 30 kg N ha-1 crop-1 in soils where no inorganic fertilizer N was applied.
Biological N2 fixation by individual systems can be estimated more or less accurately, but total BNF
in a rice field has not yet been estimated by measuring simultaneously the activities of the various
components in situ. As a result, it is not clear if the activities of the different N2-fixing systems are
independent or related. A method to estimate in situ the contribution of N, fixed to rice nutrition is still
not available. Dynamics of BNF during the crop cycle is known for indigenous agents but the pattern of
fixed N availability to rice is known only for a few green manure crops.
Introduction
More than half of the world population is dependent on rice. The crop was planted to 14.5 million
hectares of land in 1988, producing 468 million
tons. About 75% of rice land are wetlands where
rice grows in flooded fields during part or all of
the cropping period.
From the point of view of yield sustainability,
traditional wetland rice cultivation has been ex-
Estimation of BNF in rice fields
tremely successful. Moderate but stable yield has
been maintained for thousands of years without
adverse effects on soil (Bray, 1986). This is
because flooding allows the establishment of environmental conditions that maintain soil N fertility. In particular, flooding leads to the differentiation of a wide range of macro- and
micro-environments that differ in their redox
potential, physical properties, light status, and
nutrient sources for the microflora. As a result,
all groups of N2-fixing microorganisms find environments suitable for their growth in wetland
rice fields. Those include photosynthetic bacteria
and blue-green algae (BGA) developing in the
photic zone (floodwater, soil-water interface,
and submerged plant biomass), heterotrophic
bacteria in the soil and associated with rice, and
Azolla and legumes used as green manure.
An additional 300 million tons of rice will be
needed in 2020 to meet the need of a fastgrowing human population. This requires a 65%
production increase within 30 years without
much expansion of actual cultivated area (IRRI,
1989a). However, increased rice production
should not be at the expense of future generations and should fulfill the concept of sustainability. A major issue is managing nutrients
in ways that reduce agrochemical use. Increased
use of inorganic fertilizer is inescapable, but, as
pointed out by Postgate (1990), a parallel return
to greater exploitation of BNF, still responsible
for providing 60-70% of the new N in the biosphere, is sensible. The development of economically feasible technologies to increase the contribution of BNF to N nutrition of rice in highly
productive systems is one major challenge faced
by rice scientists.
Biological N2 fixation in rice fields and its use
have been reviewed by Watanabe and Roger
(1984) and Roger and Watanabe (1986). Specific
reviews deal with BGA (Roger, 1991), heterotrophs (Yoshida and Rinaudo, 1982), BNF associated with straw (Ladha and Boonkerd, 1988),
rice genotypic differences in stimulating BNF
(Ladha et al., 1988b), Azolla (Watanabe, 1982),
and leguminous green manures (Ladha et al.,
1988c). The present paper reviews recent improvements and new methodological approaches
to estimate BNF in the wetlands, summarizes
earlier and recent quantitative estimates, and
discusses research needs.
43
Improvements and new methodological
approaches
There are currently three possible approaches in
estimating BNF during a crop cycle. The first
possibility is offered by N balance studies where
balance is defined as the difference between
easily measurable outputs (N in the exported
parts of the plant and soil N at the end of the
experiment) and inputs (N fertilizer applied and
soil N at the beginning of the experiment). In
such experiments, N losses by leaching, denitrification and volatilization, and atmospheric deposition are not recorded. Therefore such balance
values usually provide an underestimation of
BNF during the crop cycle. The second approach
involves making and integrating short-term measurements at regular intervals during the crop
cycle. This could be done with acetylene-reducing activity (ARA) and short-term 15 N incorporation measurements to determine BNF by
specific agents or groups of organisms. Currently, only ARA has been used in field studies at
the crop cycle level. The third approach is to
determine the maximum biomass and the percentage of N derived from the air (% Ndfa) of
the N2-fixing agents. This method is valid when
the organism being studied builds its maximum
biomass with little turnover. In case of organisms
with rapid turnover, such as BGA, the method
may lead to a marked underestimation of the N,
fixed. This method, therefore, has been used
only for estimating % Ndfa in rice and macrophytic green manures (Azolla and legumes).
Acetylene-reducing activity measurements
Many variations of the acetylene reduction method have been used for in-situ studies in rice fields
and the associated methodological problems
were reviewed by Watanabe and Cholitkul
(1979). These include a) variability of the conversion factor of acetylene reduced to N, fixed,
b) different diffusion rates and solubilities of
acetylene and N2, c) incomplete recovery of the
ethylene formed, d) disadvantage of short-term
assays in assessing overall activity, and e) disturbance of the environmental conditions for the
N2-fixing organism during assays.
Recent studies confirm limitations that may
make quantitative extrapolations risky. ARA
44
Roger and Ladha
was linear with time with aquatic legumes
(Ladha and Tirol-Padre, 1990) but not with associative (Barraquio et al., 1986) and algal BNF
(Roger et al., 1991). C2H2/N2 conversion factor
values varied from 1.6 to 7.9 with Azolla, depending on species, PN2, assay duration, and age
of culture (Eskew, 1987). With dense algal mats,
it varied from 3.9 to 30, depending mostly on PN,
used for incubation under 15 N2 (Roger et al.,
1991). But when PN2 was similar to that of air
and when all other factors were similar to those
used for C2H2 exposure, the C2H2/N2 ratio was
close to the value of 4.4 derived from regression
analysis by Peterson and Burris (1976). Most
incubations are done under 10% C2H2 in the air,
but ARA increases with up to 25% C2H2 in the
air with thick BGA blooms (Roger et al., 1991)
and associative BNF (Barraquio et al., 1986). In
situ, the greenhouse effect in enclosures used for
incubation reduced photodependent ARA of soil
(Roger et al., 1991) and Azolla (Li et al., 1987).
Despite its limitations, ARA was used in
about 2/3 of the 38 quantitative BNF studies
related to rice since 1985. To overcome the
limitations and to obtain reproducible values,
methods have been developed where composite
and/or standardized samples collected in situ are
incubated under controlled laboratory conditions
(Barraquio et al., 1986; Roger et al., 1991). In
field studies, emphasis has been on sampling
strategies.
Estimation of associative BNF
Several methods have been developed for in situ
measurement of ARA associated with the rice
plant (Lee et al., 1977). Laboratory methods,
such as water culture (Watanabe and Cabrera,
1979) have been developed to overcome problems encountered in field assays, especially those
dealing with gas transfer.
Earlier studies were mostly aimed at associative BNF quantification. Then, emphasis has
moved toward screening techniques based on
ARA to examine variations in associative N,
fixation among rice genotypes. Hirota et al.
(1978) and Sano et al. (1981) developed a
screening method that required assays under
anaerobic conditions and was only suitable for
pot-grown rice plants, but an immediate and
linear ARA with time was obtained. Barraquio
et al. (1986) developed a short-term laboratory
assay using the root biomass and the adhering
soil from cut plants incubated in the dark. The
method is destructive but relatively simple,
rapid, and reduces problems encountered in the
whole-plant and excised root assays.
In a subsequent study, Tirol-Padre et al.
(1988) developed a sampling strategy for
measuring differences among genotypes. A complete randomized block design with at least three
blocks was recommended. ARA measurements
were performed at heading for three consecutive
days, taking several plants per genotype daily. In
most trials, day-to-day variation in ARA for
three consecutive days of measurements was not
significant. There were also no significant differences in ARA among plants sampled at 0800,
1100, and 1600 h of the same day, but greater
variability among three consecutive days of measurements was observed when sampling was
done at 1100 and 1600 h. Sampling six plants
day-1 for three consecutive days allowed detection of a 40% difference (the least significant
difference was 1.3 µmol plant -1 6 h -1 for shortduration varieties and 2.1 pmol for long-duration varieties). Detecting a difference of 20%
required a sampling of 20 plants d -1 for three
consecutive days.
Estimation of photodependent BNF
To decrease variability among replicated measurements and to avoid the greenhouse effect
usually observed in situ in the transparent enclosures used for incubation under acetylene, recent
studies use composite samples of 7-13 soil cores
incubated under constant and moderate temperature (about 25°C) and light intensity (about
30 klux) (Roger et al., 1988; IRRI, 1989b).
From a study of 440 groups of replicated measurements in plots where a wide range of agronomic practices was used, Roger et al. (in
IRRI, 1989b) drew general conclusions on 1) the
distributional ecology of ARA, 2) the implication for sampling density within a plot, and 3)
the number of replicates needed for a given
accuracy. The study of correlation between
means and variances of the replicated measurements showed a slope of the regression curve
close to 2 (Fig. 1), indicating a log-normal distribution of the data (Roger et al., 1978). This
Estimation of BNF in rice fields
Fig. 1. Correlation between means and variances of 442
groups of ARA measurements in 5 replicated plots (4 x 4 m)
(Roger et al., unpubl.).
type of distribution was observed for 1) singlelocus samples collected in the same plot and 2)
single-locus and composite samples collected in
replicated plots.
In such a distribution, the accuracy of the
mean of n measurements (Pe), defined as half of
the confidence interval expressed as a fraction of
the mean, is calculated as
Pe =
1 (10
2
(tSy/
n )
- 10
-(tSy/
n )
)
(1)
where t is the statistic of Student-Fischer, n is the
number of replicates, and Sy is the standard
deviation of the logarithms of the data (Roger et
al. , 1978).
Figure 2 presents a graphic representation of
this function for values of Sy ranging from 0.2 to
1.0. When Sy estimates are available from previous measurements performed under similar con-
45
ditions, this graph can be used to determine the
number of replicated plots needed for a given
accuracy or the number of subsamples needed
for a given representativeness of a composite
sample within a plot. Figure 3 summarizes 496 Sy
values obtained from photodependent ARA
measurements under a wide range of conditions.
The median of 653 estimates of S, is 0.317.
Applying this value to Equation 1 shows that the
accuracy of averages of ARA measurements in
replicated plots is usually low. For example,
measurements in 10 replicated plots provide an
accuracy of 0.5, for which the confidence interval
is equal to the mean (Fig. 2).
The utilization of composite samples markedly
increases the representativeness of measurements within a plot. However, as shown in Figure 2, on the average (i.e. Sy = 0.3), 10 core
subsamples are needed for a representativeness
of 0.5. A representativeness of 0.3 requires 26
core subsamples and one of 0.2 requires 55 core
subsamples. These data show that the number of
subsamples collected from a plot is often determined by methodological limitations (i.e. the
maximum number of subsamples that can be
reasonably collected or handled) rather than a
chosen accuracy.
However, rather than the accuracy of a mean,
the major concern in field experiments is the
number of replicates needed to establish a significant difference between two means, X, and
Xb. Assuming that na = nb, and Sya = Syb, the
ratio Xa/Xb required to establish a significant
Counts
P=1/2 ((upper limit-lower limit)/mean)
Standard deviation of the log. of the data(Sy)
Number of replicates
Fig. 2. Accuracy of the mean of n replicated ARA measurements according to the number of replicates and the standard
error of the logarithms of the data (sy).
Fig. 3. Histogram of 496 estimates of the standard error of
the logarithms of replicated ARA measurements performed
in 4 or 5 replicated plots using composite samples of 8 to 16
soil cores, 2 cm in diameter.
46
Roger and Ladha
difference between Xa and Xb ( p = 0.05) is given
by
(2)
2
where n is the number of replicates, Sy is the
variance of transformed data, and t 2n-2 is the t
value of Student-Fischer statistic at (2n - 2) degrees of freedom (Roger et al., 1978). Figure 4
presents the calculation of Xa/Xb for 2 < n < 20
and 0 < Sy2 < 1. The values of Xa/Xb corresponding to the median value of Sy(0.3) show that, on
an average, 5 replicated plots will only permit to
find significant differences between values whose
ratio is about 3. Ten replicates would permit to
establish a significant difference between two
values whose ratio is two.
Most field experiments on rice usually have 3
or 4 replicates, which is adequate for normally
distributed data such as rice yield, but might
often be insufficient for estimating ARA with
good accuracy.
A coefficient of variation of 100% is observed
among single-locus measurements because of
their log-normal distribution. This coefficient of
variation rapidly decreases in composite sampling when the number of subsamples increases.
Therefore, the accuracy of the mean of measurements on composite samples in replicated plots
depends mainly on the number of plots. In particular, it does not improve when the value of
the accuracy of the measurements in each plot
decreases beyond the value of the accuracy that
X2/X1
Number of replicates (n)
Fig. 4. Ratio (X1/X2) between two means of n replicated
field ARA measurements ensuring a significant difference at
p = 0.05 according to the number of replicates and the
standard error of the logarithms of the data (sy).
can be expected from the number of replicated
plots. Therefore, the determination of the optimal number of subsamples to be collected in a
plot should take into account accuracy or representativeness for the lower limit and the number
of replicated plots for the upper limit.
Interplot variability of daily ARA measurements cannot be reduced, but ARA values integrated for the crop cycle are less variable than
daily values.
These data show the importance of sampling
strategies in performing field measurements of
ARA. Such measurements are extremely tedious
and an adequate sampling and measurement
strategy can avoid either superfluous sample collection or measurements (that will not improve
significantly the accuracy of the data), or measurements that have little chance of producing
conclusive results.
Estimation of BNF by aquatic legumes
The ARA method has been widely used for the
measurement of N2 fixation by root-nodulated
upland legumes. The methodological aspects
were discussed by Hardy et al. (1973) and Turner and Gibson (1980). This method was recently applied for aquatic legumes used as green
manure in lowland rice, such as Sesbania rostrata
(Ladha and Tirol-Padre, 1990). ARA measurement for such legumes may require special consideration because of their growth habit under
submerged condition and their ability to produce
root and stem nodules. Ladha and Tirol-Padre
(1990) showed that when cut plants (aerial parts)
were incubated in a plastic bag, ARA was linear
from 0 to 2.5 h of incubation and did not significantly differ, regardless of whether incubation
was under light or in the dark. ARA was highest
in plants harvested between 1300 and 1600 h,
and the optimum acetylene concentration was
10-15%. A critical factor in assaying large numbers of field-grown plants is the prolonged time
interval between sampling and the assay. It was
found that a 4-h interval between sampling and
assay, compared with a 1-h interval, significantly
reduced ARA.
15
N2 incorporation
15
N2 incorporation was used in short-term studies
Estimation of BNF in rice fields
to assess BNF by various agents, to identify
active sites in the soil or rice plant, and to
establish the C2H2/N2 conversion factor in
BGA, Azolla (Eskew, 1987 Watanabe and
Roger, 1982), and stem-nodulating legumes
(Becker et al., 1990).
15 N
dilution method
This method is attractive because a single measurement can provide an estimate of BNF integrated over time. It is used to estimate BNF in
plants but not in soil. The major assumption is
that the 15 N/ 14 N ratio of N absorbed from the
soil and water in flooded conditions, is the same
for the N2-fixing and the non-N2-fixing plant.
This assumption is met when 15 N enrichment of
soil N available to the N2-fixing system is constant during the experiment (this implies that
15 N added is equilibrated with soil N and labeling is constant throughout the soil) or when
the N2-fixing and the non-N2-fixing plant have
similar N uptake patterns (this implies that the
ratio of soil N [S] to that of fertilizer N [F]
assimilated by N2-fixing and non-N2-fixing plants
are equal).
The approach that ensures a constant 15 N/ 14 N
ratio is promising in wetland rice fields in as
much as it is easier to label soil and stabilize 15 N
here than in upland soils; this is because the
plow layer delineates an Ap horizon in which
most roots are located and which is relatively
easy to homogenize. Either a non-N2-fixing plant
or available soil N (Zhu et al., 1984) can be used
as control, but the validity of the control depends on the level of % Ndfa.
To ensure a similar S/F uptake in the non-N2fixing and N2-fixing plants, the approach requires
a method to test S/F values in both plants.
Wagner and Zapata (1982) attempted to indirectly determine S/F values by comparing the
uptake of 35 S-labeled sulphate and indigenous
soil sulphate. However, this method proved unsatisfactory because plants differ in their relative
uptake of N to sulphate. Ledgard et al. (1985)
estimated S/F in N2-fixing and non-N2-fixing
plants by growing them in soils which received
different levels of 15 N while keeping the amount
of N constant. The proportions of S and F were
47
determined from the intercept and slope of regression between 15 N enrichment of plant N and
added N.
Pareek et al. (1990) quantified BNF by Sesbania rostrata and S. aculeata in two subsequent
crops. The first crop was grown in soil having a
fast decline of 15 N in available soil N and the
second crop was grown in soil having a slow
decline of 15 N. Estimates were computed using
1) three non-N2-fixing reference species, 2) 15 N
enrichment of available soil N as reference, and
3) the varying 15 N level technique. The study
concluded that accurate estimates can be obtained after about 100 days of stabilization of 15 N
applied to the soil under flooded, preferably
planted conditions. The need for a control having similar S/F as that of the N2-fixing plant
becomes less critical, if soil with a fairly good
stabilization of 15 N is used.
The accuracy of the 15 N dilution method is
correlated with the proportion of N derived from
BNF. This was shown by modelling the %Ndfa
in relation to 15 N enrichment of N2-fixing and
non-N2-fixing crops (Hardarson et al., 1988) or
the S/F of N2-fixing and non-N2-fixing crops
(Pareek et al., 1990). In their model, Hardarson
et al. (1988) used an arbitrary value of 0.2 atom
% 15 N excess in the N2-fixing crop and various
assumed values of atom % 15 N excess in the
non-N2-fixing crop with an error of ±10% for
each atom % 15 N excess value. The model
showed that when %Ndfa is low, small differences in 15 N enrichment of the non-N,-fixing
crop result in large changes in %Ndfa estimates.
On the other hand, when %Ndfa is high, the
error is low (Fig. 5).
Whereas the 15 N dilution method was found
suitable for estimating Ndfa in leguminous green
manure and rice, trials with aquatic N2 fixers,
such as Azolla and BGA, have shown that fast
changes in 15 N enrichment in the water over time
(Eskew, 1987), due in particular to losses by
ammonia volatilization, result in large errors in
%Ndfa estimation (Witty, 1983). These problems can be solved, in the case of Azolla, by the
sequential addition of 15 N in water (Kulasooriya
et al., 1988). But, with BGA the N level in water
sufficient for growth of the non-N2-fixing control
algae may inhibit BGA growth directly or
through competition (Roger, 1991).
48
Roger and Ladha
black cloth averaged 36 mg N crop-1 pot-1 or
7 kg N ha-1 (App et al., 1986). In similar trials,
Trolldenier (1987) found balances negatively
correlated with the amount of N applied, ranging
from -440 to +418 mg N crop-1 pot-1. Extrapolated values averaged 19 kg N ha-1 crop-1 with
65 kg N ha-1, -0.3 with 112 kg N, and -14 with
146 kg N. Using available N of a stabilized 15Nlabelled soil as control, Zhu et al. (1984) estimated that when no N fertilizer was applied and
photodependent BNF was controlled, heterotrophic BNF contributed 16-21% of rice N, or
11-16 kg N ha-1 crop-1.
% nitrogen fixed
Associative BNF and varietal differences
Atom % 15N excess (nfc)
Fig. 5. Estimated values of % N2 fixed in a plant containing
0.2 atom % excess 20.02 SD and a range of assumed atom %
15
N excess ±10% SD in nonfixing reference crop (after
Hardarson et al., 1988). The bars represent the upper and
lower limits of standard deviation of the estimates.
Difference in natural 15N abundance ( d 15N)
The d 15N method uses the fact that soil has a
higher d 15N than air and can serve as a naturally
labelled medium. The method is advantageous
because of the stable isotopic composition of N
sources, but in situ a 15N gradient observed with
soil depth can be a serious source of error.
Growing plants in pots with well-mixed soil avoids this problem (Watanabe et al., 1987). This
method was used to test differences among rice
plant organs and varieties, and to estimate Ndfa
in Azolla (Watanabe et al., 1987, 1990;
Yoneyama et al., 1987). It is currently supposed
to be suitable only for semiquantitative estimations, but its potential for quantitative measurement has not been fully explored (Peoples and
Herridge, 1990).
Estimation of BNF by various agents
Total heterotrophic BNF
Total heterotrophic BNF estimated from the N
balance in unfertilized planted pots covered with
ARA associated with rice is usually measured at
heading because it is highest at that stage. Data
summarized by Roger and Watanabe (1986)
range from 0.3 µmol to 2 µmol C2H4 hill-1 h-1.
Ladha et al. (1987) screened 16 varieties and
found activities ranging from 0.4 to 2 µmol C2H2
hill-1 h-1. Assuming 1) that ARA measured at
heading lasts 50 days, 2) a C2H2/N2 ratio of
four, and 3) a plant density of 25 m-2, N2-fixing
rate would be 1-5 kg N ha-1 crop-1. Extrapolations from 15N incorporation experiments range
from 1.3 to 7.2 kg ha-1 crop-1 (Roger and
Watanabe, 1986). Differences in the ability of
rice varieties to support associative BNF were
suspected from N balance experiments in pots
with soil exposed to light (App et al., 1986).
Varietal differences in associative BNF were
shown by ARA assays (Ladha et al., 1987,
1988b) but little is known about their physiological basis. The idea of breeding varieties higher in
associative BNF is attractive, because such varieties would enhance BNF without requiring additional cultural practices. However, a rapid
screening technique is needed.
BNF associated with straw
Early estimates of BNF after straw incorporation
range from 0.1 to 7 mg N fixed g-1 straw added
(mean 2.1) in 30 d (Roger and Watanabe, 1986).
Most data originate from laboratory incubations
of soil in the dark with 1-100% straw added
(average 22%). Such experiments simulate composting rather than the field situation where
Estimation of BNF in rice fields
straw left is always less than 1% of soil dry
weight. Moreover, dark incubation allows
heterotrophic BNF only, whereas straw incorporation also favors photodependent BNF (Ladha
and Boonkerd, 1988). Recent greenhouse and
field experiments show that straw application
may significantly increase populations and N2fixing activity of photosynthetic bacteria and
BGA (Ladha and Boonkerd, 1988). However,
pot experiments by Santiago-Ventura et al.
(1986) showed N gains of 2-4mg N g-1 straw
added with no difference when soil was exposed
to light or kept in darkness. Quantitative estimates of BNF in field experiments with straw are
not available, but a few semiquantitative data
and laboratory data suggest that straw might
increase BNF by 2-4 kg N t-1 applied.
Blue-green algae
N2 fixation by BGA has been almost exclusively
estimated from ARA. Data published before
1980 vary from a few to 80 kg N ha-1 crop-1
(mean 27 kg) (Roger and Kulasooriya, 1980).
About 200 crop cycle measurements in experimental plots at IRRI (Roger et al., 1988)
show activities of the same order: 0-1200 µmol
C2H2 m-2 h-1 for daily values and 20-500 µmol
C2H2 m-2 h-l for average ARA during a crop
cycle. Extrapolated values (assuming C2H2/
N, = 4) ranged from 0.2 to 50 kg N ha-1 crop-1
and averaged 20 kg in no-N control plots, 8 kg in
plots with ,broadcast urea, and 12 kg in plots
where N was deep-placed. ARA was negligible
in 75% of the 60 plots where urea was broadcast
49
(Roger et al., 1988). As N2-fixing BGA usually
bloom only when the photic zone is depleted of
inorganic N, most of their N can be assumed to
originate from BNF. Inubushi and Watanabe
(1986) reported that BGA in 15N-labeled plots
had about 90% Ndfa. A BGA bloom usually
corresponds to less than 10 kg N ha-1, a dense
bloom may contain 10-20 kg N ha-1 (Roger,
1991). However, estimates of BGA biomass do
not allow the N2 fixed to be quantified, because
no data are available on the turnover of the algal
biomass.
Azolla
BNF by Azolla has usually been estimated from
biomass measurement and the assumption that
most of Azolla N originates from BNF. The N
potential of Azolla was summarized by Roger
and Watanabe (1986) from data obtained mostly
in experimental plots. The N content in maximum standing crops ranged from 20 to
146 kg ha-1 and averaged 70 kg ha-1 (n = 17;
C.V. = 58%). N2-fixing rate ranged from 0.4 to
3.6 kg N ha-1 d-1 and averaged 2 kg N ha-1 d-1
(n = 15, C.V. = 47%). However, data from a 4year field trial at 37 sites in 10 countries show
lower productivities in full-scale trials than in
experimental plots (Watanabe, 1987). Biomass
was 5-25 t fresh weight ha-1 (10-50 kg N ha-1)
for Azolla grown before or after transplanting
(average 15 t ha-1 or 30 kg N).
Recent experiments focus on Ndfa determination (Table 1). Using the 15N dilution method
with Lemna and Salvinia as non-N2-fixing con-
Table 1. Ndfa in Azolla estimated by 15N dilution a
Authors
Reference plant
Ndfa %
Kumarasinghe et al. (1985)
Salvinia
Salvinia
Salvinia
Lemna/Salvinia
Lemna/Salvinia
Lemna
Lemna
Salvinia/Lemna
Salvinia/Lemna
Lermna/Spirogyra
Lemna
92
79
76
81-83
82-79
40-59
63-71
52-55
58-64
86-93
80-81
You C.B. et al. (1987)
Kulasooriya et al. (1988)
Watanabe et al. (1990)
a
After Watanabe, 1990.
Remarks
+ 15 mg N/tray
+75 mg N/tray
+150 mg N/tray
with rice
without rice
labelled Azolla
without rice, 40 ppm N
with rice, 40 ppm N
3 crops of Azolla
after a rice crop
50
Roger and Ladha
trols, Watanabe and Talukdar, and Kumarasinghe (unpublished data cited by Eskew, 1987)
estimated that 80-85% Azolla N was Ndfa.
Using similar controls and applying 15 N-labelled
urea at 3-day intervals for 14 days, Kulasooriya
et al. (1988) found 51-61% Ndfa and BNF of
10-14 kgN ha -1 in 14 days. Using the d 15 N
method, Yoneyama et al. (1987) estimated that
59-99% of N of the strains tested was Ndfa.
Anabaena-free A. filiculoides and the Lemna
control had a similar d 15 N which was not influenced by the water level (flooded or saturated
soil).
Leguminous green manure
BNF by leguminous green manure (LGM) used
for rice has usually been estimated from total N
measurement and the assumption that 50-80%
of accumulated N is Ndfa. Values of N accumulated in a root-nodulating pre-rice LGM crop
summarized by Roger and Watanabe (1986) average 114 kg N ha -1 . Highest values (146267 kg N ha -1 in 52 d) were observed for Sesbania. Values published after 1985, which include several stem-nodulating legumes, average
133 kg N ha -1 (Ladha et al., 1988c). Ranges in
kgN ha -1 are 40-225 for aquatic LGM, and
33-115 for grain legumes. Assuming 50-80%
Ndfa, one LGM crop can fix an average 1.01.6 kgN ha -1 d -1 or 60-100 kgN ha -1 in 5060 d.
Using 15 N dilution, Pareek et al. (1990) quantified the contribution of BNF to total N of S.
rostrata and S. cannabina grown up to 65 days in
the dry season (short days) and the wet season
(long days). At 25 days, the percentage of total
N obtained from the atmosphere by both species
was 50% during the dry season and 75% during
the wet season. The contribution by BNF increased with age of the plant and ranged from 70
to 95% at 45-55 days. Although the %Ndfa in
the two species were similar, S. rostrata showed
substantially higher fixation than S. cannabina,
and the difference was more pronounced in the
wet season. This was attributed to the difference
in total N uptake by the two species. The estimates of BNF per plant were converted to N,
fixed per hectare by assuming a plant density of
500,000 ha -1 . At 45 days, the dry-season crop of
S. rostrata had fixed 100 kg N ha -1 and the wetseason crop 140 kg ha -1 , while the values for S.
cannabina were 78 and 119 kg ha -1 , respectively.
The daily N gain was 17 kg ha -1 between 45 and
55 days in the wet-season crop. Almost similar
% estimates of Ndfa by S. rostrata were reported
by Yoneyama et al. (1990) using natural abundance 15 N dilution. Based on 15 N2-calibrated
ARA, Becker et al. (1990) reported %Ndfa of
80 and 70% by 8-week-old S. rostrata and Aeschynomene afraspera, respectively. The %Ndfa
did not differ in both species when grown at
different times of the year.
On the other hand, Rinaudo et al. (1988) and
Ndoye and Dreyfus (1988) reported a much
lower contribution from N2 fixation, ranging
from 30 to 50% by 53- to 63-day-old S. rostrata.
A possible reason for the lower estimates is that
they used non-inoculated S. rostrata as a reference plant.
A few estimates of BNF by S. rostrata as a
pre-rice LGM are available from small-scale balance studies. Rinaudo et al. (1988) reported a
gain of 267 kg N ha -1 after incorporating a 52-d
crop. In a 45-d Sesbania-rice (WS)/55-d Sesbania-rice (DS) sequence, Ladha et al. (1988a)
estimated that Sesbania fixed 303 kg N ha -1
year -1 when uninoculated, and 383 kg N when
inoculated with Azorhizobium.
Nitrogen balance and BNF contribution in
wetland rice
Nitrogen balances in long-term fertility experiments listed by Greenland and Watanabe
(1982) ranged from 19 to 98 kg N ha -1 crop -1
(average 51 kg) in 9 fields with no N fertilizer.
In 4 fields with N fertilizer, the average was
-1.5 kg N ha -1 crop -1 . With rice grown alternately with an upland crop, the average was
44 kg N ha -1 year -1 in 3 fields with no N fertilizer and -29kg N in 2 fields with applied urea.
At two Philippine sites, App et al. (1984) found
no decrease in total soil N after 24 and 17 crops,
respectively. Calculations based on yields and
known inputs suggested that two crops year -1
resulted in balances of 79 and 103 kg N ha -1
year -1.
Compared with pot experiments, balance
Estimation of BNF in rice fields
51
Table 2. Effect of various factors on N balance a
Factor
No. of
observations
Mean
(kg N ha-1 crop 1)
Standard
deviation
Level of significance
of the difference
No inorganic N
With inorganic N
Planted
Unplanted
Light
Dark
No N, Light
No N, Dark
166
45
193
18
197
14
152
14
29.7
4.0
26.5
- 0.5
25.0
13.2
31.2
13.2
25.4
47.6
30.7
46.2
33.0
13.8
25.7
13.8
0.000**
0.001**
0.108ns
0.011*
Coefficients of correlation between balance and N Applied (n = 211):
Inorganic N
Organic N
r = -0.320**
r = -0.157*
Total N
r = -0.365**
~~
Analysis of data compiled from: Ando (1975); App et al., (1980, 1084, 1986); De and Sulaiman (1950); Firth et al. (1973);
Greenland and Watanabe (1982); lnatsu and Watanabe (1969); Konishi and Seino (1961); Koyama and App (1979);
Santiago-Ventura et al. (1986); Singh and Singh (1987); Trolldenier (1987); and Willis and Green (1948).
177 values obtained in pots were extrapolated to kg N ha -1 crop -1 on an area basis.
a
studies in the field encounter additional difficulties because of sampling errors, unaccounted
subsoil contribution, and losses by leaching.
Therefore, there is renewed interest in pot
studies (App et al., 1986; Santiago-Ventura et
al., 1986; Singh and Singh, 1987; Trolldenier,
1987). In a 4-crop experiment comparing organic
N (Azolla and BGA) and urea, Singh and Singh
(1987) found positive N balances ranging from
13 to 163 mg N crop-1 pot-1. Balances were highest (133-163 mg Ncrop-1 pot-1) in pots that received 60 kg organic N ha-1, and lowest (1329 mg) in pots that received 30-60 kg N ha-1 as
urea. Balance in the control was 51 mg N crop-'
pot -1.
In a second experiment comparing the effects
of soil exposure to light, presence of rice, and
flooding in nonfertilized plots, N balance ranged
from 78 to 103 mg crop-1 pot-1 in fallow pots not
exposed
to
light
and
from
243
to
277 mg crop-1 pot-1 in planted pots exposed to
light. The N balances reported by Singh and
Singh (1987) in flooded pots exposed to light (51
and 277 mg N crop-1 pot-1) cover a similar
range than 89 values reported by App et al.
average
(1986)
(70-260mg N crop -1 pot -1,
153 mg). Extrapolating values of App et al.
(1986) shows N balances ranging from 16 to
70 kg N ha-1 crop-1 (average 38 kg) in unfertilized planted pots exposed to light. SantiagoVentura et al. (1986) reported balances of about
100 mg N crop-1 pot-1 in pots exposed to light
and receiving no or low levels of N fertilizer.
With high levels of N fertilizer, the balance was
not significant.
Figure 6 and Table 2 summarize the balance
measurements in rice soils. Data are from 14
reports, including both field and pot studies. To
allow for comparisons, data from pot experi-
Count
Balance (kg N ha-1 crop-1
Fig. 6. Histogram of 211 N balance values from field and pot
experiments dealing with rice. Data compiled from: Ando
(1975); App et al. (1980, 1984, 1986); De and Sulaiman
(1950); Firth et al. (1973); Greenland and Watanabe (1982);
Inatsu and Watanabe (1969); Konishi and Seino (1961);
Koyama and App (1979); Santiago-Ventura et al. (1986);
Singh and Singh (1987); Trolldenier (1987); and Willis and
Green (1948).
177 values obtained in pots were extrapolated to kg N ha-1
crop-1 on an area basis.
52
Roger and Ladha
ments were extrapolated on an area basis and
expressed in kg N ha-1 crop-1. Figure 6 presents
the histogram of the 211 balance values,
with an average change of 24.2 kg N ha-1 crop-1
(median: 27 kg). Ninety-five percent of the
balance
values
are
between
-60 and
+90 kg N ha-1 crop-1. Extreme values are from
balance experiments in pots performed over one
crop cycle only (Willis and Green, 1948); other
pot experiments were conducted for three to six
successive crops.
A comparison of the average values of data
grouped according to various criteria by the Student-Fischer t test (Table 2) shows a larger balance in the absence of inorganic N
(29.7 kg N ha-1 crop-1) than when inorganic N is
applied (4.0 kgN ha-1 crop-1). A study of the
correlation between N balance and the level of N
fertilizer applied shows highly significant ( p =
0.01) negative values for inorganic N and total N
(inorganic + organic) and a negative value ( p =
0.05) for organic N (Table 2). This supports the
general concept that inorganic N is more susceptible to losses than organic N.
Data in Table 2 also show a larger balance in
planted soil (26.5 kg N ha-1 crop-1) than in unplanted soil (-0.5 kg). The difference between
balance in pots where soil was either exposed to
light or kept in the dark is not statistically significant when considering all data. It becomes highly
significant when only experiments with no inorganic N fertilizer applied are considered. The
data indicate that under such conditions, photodependent BNF might contribute roughly twice
more N than heterotrophic BNF.
Conclusions
Recent methodological progress in measuring
BNF in rice fields includes improved strategies
for sampling and a better understanding of the
potential of the 15N dilution methods (labeled
substrate and natural abundance). 15N dilution,
using available soil N as control, is a promising
method for screening rice varieties for ability to
utilize biologically fixed N.
Ranges of reported estimates of BNF by in-
Table 3. Range of estimates of N2 fixed by various agents in wetland rice fields (kg N ha-1 crop-1) and theoretical maximum
potential (value and assumptions)
Component
Reported range of
estimates
Theoretical maximum potential
and assumptions
BNF associated with
rice rhizosphere
1-7 kgN ha-1 crop-1
BNF associated with
straw
2-4 kg N t-1 straw
40 kgN ha-1 crop-1
• All rhizospheric bacteria are N2-fixers,
• C flow through rhizosphere is 1 t ha-1 crop-1, and
• 40 mg N is fixed g C-1.
35kg N ha-1 crop-1
• 5 t of straw is applied, and
• 7 mg N is fixed g-1 of straw.
Total heterotrophic
BNF
1-31 kgN ha-1 crop-1
60 kg N ha-1 crop-1
• All C input (2 t crop-1) is used by N2-fixers.
Blue-green algae
0-80 kg N ha-1 crop-1
70 kg N ha-1 crop-1
• Photosynthetic aquatic biomass is
composed exclusively of N2-fixing BGA (C/N = 7), and
• primary production is 0.5 t C ha-1 crop-1.
Azolla
20-150 kgN ha-1 crop-1
(experimental plots)
10-50 kg N ha-1 crop-1
(field trials)
224 kg N ha-1 crop-1
• one Azolla standing crop is 140 kg N ha-1, and
• two Azolla crops are grown per rice crop,
• Ndfa is 80%.
Legume
green manures
20-260 kg N ha-1 crop-1
260 kg N ha-1 in 55 days
• Sesbania rostrata is used as green manure
• 290 kg N ha-1 is accumulated in 50-60 d, and
• Ndfa is 90%.
Estimation of BNF in rice fields
dividual systems are presented in Table 3. Estimates for green manure (Azolla and legumes)
are based on biomass measurements combined
with Ndfa determination and are probably more
reliable than estimates for indigenous fixers
based mostly on indirect methods (ARA) or
balance methods in small-scale trials. However,
total BNF in a rice field has not yet been estimated by measuring simultaneously the activities
of the various components in situ. As a result,
the relation between the different N2-fixing systems, especially indigenous ones, are not fully
understood and it is not clear if their activities
are independent or related. A method to estimate in situ the contribution of fixed N2 to rice
nutrition is still not available. The dynamics of
BNF during the crop cycle are known for indigenous agents, but the pattern of fixed N availability to rice is known only for a few green
manure crops. As a result, BNF in models of N
cycling in wetlands is either not taken into account or considered as a nondynamic input.
Because of technological and socio-economical
limiting factors, the agronomic potential of BNF
is still largely underutilized in rice cultivation
(Roger, 1991). Methodological progress and
comprehensive evaluations of BNF in rice fields
are still needed to develop and test agricultural
practices that take advantage of BNF in this
important agroecosystem.
Acknowledgements
The authors thank Drs. I Watanabe (IRRI) and
R Buresh (International Fertilizer Development
Center) for their very useful comments.
This work was conducted under a scientific
agreement between IRRI and ORSTOM
(France).
References
Ando S 1975 Effect of crude organic matter applied to the
excessively draining low-yield paddy field. Bull. Kagawa
Agric. Expt. Stn. 27, 1-81. ( In Japanese, English
summary ).
App A A, Watanabe I. Alexander M, Ventura W, Daez C,
Santiago T and De Datta S K 1980 Nonsymbiotic nitrogen
fixation associated with the rice plant in flooded soils. Soil
Sci. 130, 283-289.
53
App A A. Watanabe I, Santiago-Ventura T, Bravo M and
Daez-Jurey C 1986 The effect of cultivated and wild rice
varieties on the nitrogen balance of flooded soil. Soil Sci.
141, 448-452.
App A. Santiago T, Daez C, Menguito C, Ventura W, Tirol
A. Po J, Watanabe I, De Datta S K and Roger P 1984
Estimation of the nitrogen balance for irrigated rice and
the contribution of phototrophic nitrogen fixation. Field
Crops Res. 9, 17-27,
Barraquio W L, Daroy M L G. Tirol A C, Ladha J K and
Watanabe I 1986 Laboratory acetylene reduction assay for
relative measurement of N2-fixing activities associated with
field-grown wetland rice plants. Plant and Soil 90, 359-372.
Becker M, Ladha J K and Ottow J C G 1990 Growth and
N2-fixation of two stem-nodulating legumes and their effect
as green manure on lowland rice. Soil Biol. Biochem. 22,
1109-1119.
Bray F 1986 The Rice Economies, Technology and Development in Asian Societies. Basil Blackwell. 254 p.
De P K and Sulaiman M 1950 The influence of algal growth
in the rice fields on the yield of crops. Indian J. Agric. Sci.
20, 327-342.
Eskew D L 1987 Use of 15 N in N2-fixation and N cycling
studies of Azolla. In Azolla Utilization. pp 233-239. The
International Rice Research Institute, Los Baños, Philippines.
Firth P, Thitipoca H, Suthipradit S, Wetselaar R and Beech
D F 1973 Nitrogen balance studies in the central plains of
Thailand. Soil Biol. Biochem. 5, 41-46.
Greenland D J and Watanabe I 1982 The continuing nitrogen
enigma. Trans. Int. Congr. Soil Sci, New Delhi 5, 123-137.
Hardarson G, Danso S K A and Zapata F 1988 Dinitrogen
fixation measurements in alfalfa-ryegrass swards using 15 N
and influence of the reference crop. Crop Sci. 28. 101-105.
Hardy R W F, Burns R C and Holsten R D 1973 Application
of the acetylene-ethylene assay for measurement of nitrogen fixation. Soil Biol. Biochem. 5, 47-81.
Hirota Y, Fujii T, Sano Y and Iyama S 1978 N2-fixation in
the rhizosphere of rice. Nature, London 276, 416-417.
Inatsu O and Watanabe K 1969 Changes of fertility in peaty
paddy soils in Sorachi district. Hokuno 36. 31-38. (cited by
Greenland and Watanabe, 1982). ( In Japanese ).
Inubushi K and Watanabe I 1986 Dynamics of available
nitrogen in paddy soils. II. Mineralized N of chloroformfumigated soil as a nutrient source for rice. Soil Sci. Plant
Nutr. 32, 561-577.
IRRI 1989a IRRI Toward 2000 and Beyond. Manila, Philippines. 72p.
IRRI 1989b Photodependent nitrogen fixation and photosynthetic aquatic biomass. In Annual Report for 1988. pp
347-353. International Rice Research Institute, Los
Baiios, Philippines.
Konishi C and Seino K 1961 Studies on the maintenancemechanism of paddy soil fertility in nature. Bull. Hokuriku
Agric. Exp. Stn. 2 Niigata Prefecture. 41 p. ( In Japanese,
English summary ).
Koyama T and App A A 1979 Nitrogen balance in flooded
rice field. In Nitrogen and Rice. pp 95-104. The International Rice Research Institute, Los Baiios, Philippines.
Kulasooriya S A, Senviratne P R G, De Silva W S A G,
Abeysekera S W, Wijesundara C and De Silva A P 1988
54
Roger and Ladha
Isotopic studies on N2-fixation in Azolla and the availability of its nitrogen to rice. Symbiosis 6, 151-166.
Ladha J K and Boonkerd N 1988 Biological nitrogen fixation
by heterotrophic and phototrophic bacteria in association
with straw. In Proceedings of the First International Symposium on Paddy Soil Fertility, Chiangmai, Thailand, 6-13
December 1988. pp 173-187. ISSS, Paddy Soil Fertility
Group.
Ladha J K and Tirol-Padre A 1990 Standardization of
acetylene reduction assay with field grown aquatic legume.
Int. Rice Res. Newsl. 15, 16-17.
Ladha J K, Miyan S and Garcia M 1988a Sesbania rostrata as
green manure for lowland rice: Growth, N2-fixation, Azorhizobium sp. inoculation, and effects on succeeding crop
yields and nitrogen balance. Biol. Fert. Soils 7, 191-197.
Ladha J K, Padre A T, Punzalan G C, Watanabe I and De
Datta S K 1988b Ability of wetland rice to stimulate
biological nitrogen fixation and utilize soil nitrogen. In
Nitrogen Fixation: Hundred Years After. Eds. H Bothe, F
J de Bruijn and W E Newton, pp 747-752. Gustav Fischer,
Stuttgart, New York.
Ladha J K, Tirol-Padre A, Punzalan G C and Watanabe I
1987 Nitrogen-fixing (C2H2-reducing) activity and plant
growth characters of 16 wetland rice varieties. Soil Sci.
Plant Nutr. 33, 187-200.
Ladha J K, Watanabe I and Saono S 1988c Nitrogen fixation
by leguminous green manure and practices for its enhancement in tropical lowland rice. In Sustainable Agriculture:
Green Manure in Rice Farming. pp 165-183. The International Rice Research Institute, Los Baños, Philippines.
Ledgard S F, Simpson J R, Freney J R and Bergersen F J
1985 Assessment of the relative uptake of added and
indigenous soil nitrogen by nodulated legumes and reference plants in the 15N dilution measurement of N2 fixation:
Glasshouse application of method. Soil Biol. Biochem. 17,
323-328.
Lee K K, Alimagno B V and Yoshida T 1977 Field technique
using the acetylene reduction method to assay nitrogenase
activity and its association with rice rhizosphere. Plant and
Soil 47, 519-526.
Li Zhuo-Xin, Zu Shou-Xian, Mao Mei-Fei, Wang Fu-Lai and
Zhao Bing-Bo 1987 Determination of amount of N2-fixation and change in N2-fixing activity of Azolla in natural
environment. In Azolla Utilization. pp 223-231. The International Rice Research Institute, Los Baños, Philippines.
Ndoye I and Dreyfus B 1988 N2-fixation by Sesbania rostrata
and Sesbania sesban estimated using 15N and total N
difference methods. Soil Biol. Biochem. 20, 209-213.
Pareek R P, Ladha J K and Watanabe I 1990 Estimation of
N2-fixation by Sesbania rostrata and S. cannabina in lowland rice soil by 15N dilution method. Biol. Fert. Soils 10,
77-88.
Peoples M B and Herridge D F 1990 Nitrogen fixation by
legumes in tropical and sub-tropical agriculture. Adv.
Agron. 44, 155-223.
Peterson R B and Burris R H 1976 Conversion of acetylene
reduction rate to nitrogen fixation rates in natural populations of blue-green algae. Anal. Biochem. 73, 404-410.
Postgate J 1990 Fixing the nitrogen fixers. New Scientist, 3
February 1990, 57-61.
Rinaudo G, Alazard D and Moudiongui A 1988 Stemnodulating legumes as green manure for rice in West
Africa. In Green Manure in Rice Farming. pp 97-109. The
International Rice Research Institute, Los Baños, Philippines.
Roger P A 1991 Blue-green algae in agriculture. In Microorganisms that Promote Plant Productivity. Eds. J 0
Dawson and P J Dart. Kluwer Academic Publishers, Dordrecht, The Netherlands.
Roger P A and Kulasooriya S A 1980 Blue-green algae and
rice. The International Rice Research Institute, Los
Baños, Philippines. 112 p.
Roger P A and Watanabe I 1986 Technologies for utilizing
biological nitrogen fixation in wetland rice: Potentialities,
current usage, and limiting factors. Fert. Res. 9. 39-77.
Roger P A, Jimenez R and Santiago-Ardales S 1991 Methods
for studying blue-green in ricefields: distributional ecology.
sampling strategies, and estimation of abundance. IRRI
Research Paper Series no. 150. The International Rice
Research Institute, Los Baños, Philippines.
Roger P A, Reddy P M and Remulla-Jimenez R 1988
Photodependent acetylene reducing activity (ARA) in
ricefields under various fertilizer and biofertilizer management. In Nitrogen Fixation: Hundred Years After. Eds. H
Bothe, F J de Bruijn and W E Newton. p 827. Gustav
Fischer, Stuttgart. New York.
Roger P A, Reynaud P A and Monniaux G 1978 Normalisation des donnees et calcul de la precision des mesures en
microbiologie du sol (Normalization of data and calculation of the accuracy of measurements in soil microbiology).
Cah. ORSTOM, Ser. Biol. 13, 171-180.
Sano Y, Fujii T, Iyama S, Hirota Y and Komagata K 1981
Nitrogen fixation in the rhizosphere of cultivated and wild
rice strains. Crop Sci. 21, 758-761.
Santiago-Ventura T, Bravo M, Daez C, Ventura W,
Watanabe I and App A A 1986 Effects of N fertilizers,
straw, and dry fallow on the nitrogen balance of a flooded
soil planted with rice. Plant and Soil 93, 405-411.
Singh A L and Singh P K 1987 Nitrogen fixation and balance
studies of rice soil. Biol. Fert. Soils 4, 15-19.
Tirol-Padre A, Ladha J K, Punzalan G C and Watanabe I
1988 A plant sampling procedure for acetylene reduction
assay to detect rice varietal differences in ability to stimulate N2-fixation. Soil Biol. Biochem. 20, 175-183.
Trolldenier G 1987 Estimation of associative nitrogen fixation
in relation to water regime and plant nutrition in a longterm pot experiment with rice. Biol. Fert. Soils 5, 133-140.
Turner G L and Gibson A H 1980 Measurement of dinitrogen fixation by direct means. In Methods for Evaluating Biological Nitrogen Fixation. Ed. F J Bergersen. pp
111-138. Wiley, New York.
Wagner G H and Zapata F 1982 Field evaluation of reference
crops in the study of nitrogen fixation by legumes using
isotope techniques. Agron. J. 74, 607-612.
Watanabe I 1982 Azolla-Anabaena symbiosis - its physiology
and use in tropical agriculture. In Microbiology of Tropical
Soil and Plant Productivity. Eds. Y Dommergues and H S
Diem. pp 169-185. Martinus Nijhoff Publishers. The
Hague, The Netherlands.
Watanabe I 1987 Summary report of the Azolla program of
Estimation of BNF in rice fields
the International Network on Soil Fertility and Fertilizer
Evaluation for Rice. In Azolla Utilization. pp 197-205.
The International Rice Research Institute, Los Baños,
Philippines.
Watanabe I and Cabrera D R 1979 Nitrogen fixation associated with the rice plant grown in water culture. Appl.
Environ. Microbiol. 37, 373-378.
Watanabe I and Cholitkul W 1979 Field studies on nitrogen
fixation in paddy soils. In Nitrogen and Rice. pp 223-239.
The International Rice Research Institute, Los Baños,
Philippines.
Watanabe I and Roger P A 1982 Use of 15N in the study of
biological nitrogen fixation in paddy soils at the International Rice Research Institute. In Proceedings of the Consultants Meeting, FAO/IAEA Joint Project, Vienna. pp
81-98.
Watanabe I and Roger P A 1984 Nitrogen fixation in wetland
rice fields. In Current Development in Biological Nitrogen
Fixation. Ed. N S Subba Rao. pp 237-276. Oxford and
IBH Pub Co.. New Delhi.
Watanabe I, Yoneyama T, Padre B and Ladha J K 1987
Difference in natural abundance of 15N in several rice
( Oryza sativa L.) varieties: Application for evaluating N2fixation. Soil Sci. Plant Nutr. 33, 407-415.
Watanabe I, Yoneyama T, Talukdar H and Ventura W 1992
The contribution of atmospheric N2 to Azolla spp. grown
in flooded soils. Soil Sci. Plant Nutr. 37. 101-109.
55
Willis W H and Green V E 1948 Movement of nitrogen in
flooded soils planted to rice. Soil Sci. Soc. Am. Proc. 13,
229-237.
Witty J F 1983 Estimating N2-fixation in the field using
15N-labelled fertilizer: Some problems and solutions. Soil
Biol. Biochem. 15, 631-639.
Yoneyama T 1990 Use of natural 15N abundance for evaluation of N2-fixation by legumes and Azolla in remote fields.
Vol IV. pp 126-130. Proc. 14th International Congress of
Soil Science, Kyoto. Japan.
Yoneyama T. Ladha J K and Watanabe I 1987 Nodule
bacteroids and Anabaena: Natural 15N enrichment in the
legume-rhizobium and Azolla-Anabaena symbiotic systems. J. Plant Physiol. 127, 251-259.
Yoshida T and Rinaudo G 1982 Heterotrophic N2-fixation in
paddy soils. In Microbiology of Tropical Soils and Plant
Productivity. Eds. Y R Dommergues and H G Diem. pp
75-107. Martinus Nijhoff/Dr W Junk, Publishers, The
Hague, The Netherlands.
Zhu Zhao-liang. Liu Chong-qun and Jiang Bai-fan 1984
Mineralization of organic nitrogen. phosphorus, and sulfur
in some paddy soils of China. In Organic Matter and Rice.
pp 259-272. The International Rice Research Institute.
Los Baños, Philippines.
Plant and Soil 141: 57-67, 1992.
© 1992 KIuwer Academic Publishers. Printed in the Netherlands.
PLSO SA04
Improving nitrogen-fixing systems and integrating them into sustainable
rice farming
I. WATANABE and C.C. LIU
The International Rice Research Institute (IRRI), P.O. Box 933, Manila, The Philippines and Fujian
Academy of Agricultural Sciences, Fuzhou, Fujian, People's Republic of China
Key words:
aquatic legumes, Azolla, Cyanobacteria, flooded rice soils, nitrogen, N2 fixation, tropics
Table of contents
Abstract
Introduction
Nitrogen fixation by cyanobacteria and its possible use
Nitrogen fixation by Azolla-Anabaena symbiosis
Factors restricting wide use by farmers
Recent development in genetic enhancement
Multiple uses
Rice-Azolla-fish culture
Effect of Azolla on the fate of applied inorganic N fertilizer
Nitrogen-fixing aquatic legumes
Exploration of fast-growing aquatic legumes as green manure crop
Areas with potential for aquatic leguminous green manures
Concluding remarks
References
Abstract
This paper summarizes recent achievements in exploiting new biological nitrogen fixation (BNF)
systems in rice fields, improving their management, and integrating them into rice farming systems. The
inoculation of cyanobacteria has been long recommended, but its effect is erratic and unpredictable.
Azolla has a long history of use as a green manure, but a number of biological constraints limited its use
in tropical Asia. To overcome these constraints, the Azolla-Anabaena system as well as the growing
methods were improved. Hybrids between A. microphylla and A. filiculoides (male) produced higher
annual biomass than either parent. When Anabaena from high temperature-tolerant A. microphylla
was transferred to Anabaena-free A. filiculoides, A. filiculoides became tolerant of high temperature.
Azolla can have multiple purposes in addition to being a N source. An integrated Azolla-fish-rice
system developed in Fujian, China, could increase farmers' income, reduce expenses, and increase
ecological stability. A study using Azolla labeled with 15 N showed the reduction of N losses by fish
uptake of N. The Azolla mat could also reduce losses of urea N by lowering floodwater-pH and storing
a part of applied N in Azolla. Agronomically useful aquatic legumes have been explored within
58
Watanabe and Liu
Sesbania and Aeschynomene. S. rostrata can accumulate more than 100 kg N ha-1 in 45 d. Its N2
fixation by stem nodules is more tolerant of mineral N than that by root nodules, but the flowering of S.
rostrata is sensitive to photoperiod. Aquatic legumes can be used in rainfed rice fields as N scavengers
and N2 fixers. The general principle of integrated uses of BNF in rice-farming systems is shown.
Introduction
Rice is the most suitable crop in wet monsoon
Asian areas. Heavy rain floods the land during
most or part of the rice-growing period. Flooded
culture is effective in reducing soil erosion and
avoiding acidification, which often occur under
dryland conditions. The nitrogen fertility of soil
is sustained better under flooded conditions than
under dryland conditions (Watanabe and Roger,
1984; Watanabe, 1986). Favorable conditions for
biological nitrogen fixation (BNF) is one of the
reasons for the relatively stable yield of rice
under flooded condition. After soil flooding, differentiation of an oxidized layer and a reduced
layer and development of a photodependent
biomass in floodwater and on the soil surface
occur. The reduced soil conditions favor heterotrophic nitrogen fixation. The floodwater is the
site for photodependent nitrogen fixation by
free-living cyanobacteria and photosynthetic bacteria, and symbiotic cyanobacteria with Azolla.
The rhizosphere of rice also provides favorable
conditions for microaerophilic bacterial N2 fixation.
The centuries-long practice of using green manures was overtaken by the use of chemical
fertilizers in the 1960s. Increasing petroleum
prices and environmental issues rekindled interest in and attention to the importance of BNF
and its use in rice production. The potentialities
and practical values differ among BNF systems.
Four nitrogen-fixing systems can function in the
wetland rice-farming systems. The approximate
amount of N gains, the technological availability
of these systems, and possible constraints to
technological dissemination are shown in Table 1
(modified from Roger and Watanabe, 1986). The
quantity of nitrogen fixed by indigenous nitrogen-fixing organisms (bacteria and cyanobacteria) is not high, but this type of nitrogen
fixation operates everywhere. Because of competition with native flora, replacement of indigenous strains with better strains is difficult. Cultural
techniques to stimulate the growth and activity
of indigenous organisms are required. On the
other hand, the amount of nitrogen fixed by
inoculated organisms (Azolla and legumes) is
high, but high cash-and labor inputs are required
to make the best use of their nitrogen-fixing
potentials.
The management and manipulation of BNF
systems in flooded rice systems were reviewed
earlier (Roger and Watanabe, 1986; Watanabe,
1986). Since these reviews, progress has been
made in understanding cyanobacteria inoculation, genetic enhancement of Azolla-Anabaena
symbiosis, integrated uses of Azolla, exploration
of fast-growing stem-nodulating legumes and
their use in wetland rice-farming systems.
Table 1. N gains, technological availability, and constraints in use of various BNF systems in wetland rice farming system
System
Non-symbiotic
Cyanobacteria
Associative bacteria
N gains a
(kg N ha-1 per crop)
10-80
< 10
Technological availability
High but questionable
Not available
P deficiency, grazer pressure
Poor water control, insect
damage, P deficiency, etc.
Seed production,
photoperiod sensitivity
Symbiotic
Azolla-Anabaena
20-100
High
Aquatic legumes
30-230
High
a
Constraints
N gains are per crop in a symbiotic system, and per crop of rice in a non-symbiotic system.
Improving N2-fixing systems in rice farming
In this paper, recent developments in the modification of nitrogen-fixing systems and their use
in rice-farming systems are presented.
Nitrogen fixation by cyanobacteria and its
possible use
Since the discovery of the importance of
cyanobacteria in N gains under flooded conditions, many inoculation trials of cultured
cyanobacteria have been made. From intensive
surveys of the results of inoculation and from his
own experiments, Roger (1990) presented a review of the inoculation of cyanobacteria. The
inoculation increased rice grain yields by an average of 350 kg ha -1 . When successful, the inoculation is low-cost technology, but its effect is
erratic and unpredictable. No reliable method is
available to assure the success of inoculation
with 'efficient' cyanobacterial strains. Recent
studies showed (Reddy and Roger, 1988; Roger
et al., 1987; Roger, 1990) that N2-fixing
cyanobacteria are present in rice fields at a much
higher rate than was previously thought and that
non-indigenous strains rarely establish themselves. A shortage of P, the presence of combined N in floodwater, and grazer populations
often limit the growth and activities of cyanobacteria. Alleviation of those conditions could
stimulate the growth and N2 fixation of indigenous cyanobacteria, but the economic viability
and efficiency of applying P and pesticides to kill
grazers have to be determined. Efficiency of
P application to cyanobacterial nitrogen fixation
in floodwater is lower than that to Azolla
(Watanabe and Cholitkul, 1990). The applications of inorganic N fertilizers, and probably
herbicides, inhibit cyanobacterial nitrogen fixation. Although NH4-resistant mutants were reported (Latorre et al., 1986), competitiveness of
the strains against the indigenous strain was
weak (Roger, ORSTOM, personal communication). A herbicide-resistant strain was isolated
(Vaishampayan, 1984). Because the level of herbicide for selection of resistant strains was much
higher than the recommended dose of herbicides, the ecological significance of resistant
strains is questionable.
59
Nitrogen fixation by Azolla-Anabaena symbiosis
For many centuries, the aquatic fern Azolla has
been used in southern China and northern Vietnam as green manure for rice (NierzwickiBauer, 1990; Shi and Hall, 1988), but in Southeast Asia Azolla was not popular. In the 1980s,
Azolla was first used in rice fields in southern
Mindanao in the Philippines. Because soils in
this area were rich in available P, the growth of
Azolla was abundant (Watanabe and Ramirez,
1990). Economical analysis showed that in this
area, the use of Azolla before rice transplanting
could save inputs of at least US$10 at 1981 prices
(Kikuchi et al., 1984). Azolla pinnata var. imbricata is indigenous in Asia, but recently species
belonging to the Euazolla section ( A.
filiculoides, A. microphylla, A. caroliniana, and
A. mexicana ) have been introduced for agricultural use, because their potential biomass production is higher than that of A. pinnata. One
crop of Azolla at full cover can accumulate
20-100 kg N ha-1(Watanabe and Roger, 1984).
Two crops of Azolla (about 30 days each) can
supply N for a crop of rice. At least 70% of N
accumulated by Azolla comes from atmospheric
N (see Roger and Ladha, 1992).
Factors restricting wide use by farmers
Despite the popularity of Azolla among scientists
in Asia, actual use by farmers in Asia is still
limited. In China and Vietnam, its use has diminished because of the availability of chemical
N fertilizer and the cultivation of more economical crops. Many technical constraints inhibit wide
use by farmers in South and Southeast Asia.
These are a) the difficulties of maintaining Azolla inoculum throughout the year, particularly in
the dry season, b) P deficiency, c) low tolerance
for high temperature, d) damage by insect and
fungi, and e) poor water control. To overcome
these constraints, technological improvement
and selection of Azolla strains suitable for various conditions have been sought. To overcome
the maintenance of Azolla in winter or summer,
the technologies of harvesting and seeding Azolla sporocarps - which have tolerance for adverse
conditions were developed in China (Lu, 1987).
Scientists in Zhejiang, Guandong, and Hunan
60
Watanabe and Liu
Provinces have studied this technique, but the
slow growth in seedbeds and sporulation in only
a limited number of strains limit the wide use of
this technology. Some strains of Azolla like A.
caroliniana (IRRI ac. No 3001) are tolerant of
adverse conditions and recover quickly after a
dry or cool period, despite its inability to sporulate. This A. caroliniana strain grows abundantly
in South China (Liu and Zheng, 1989).
Surveys of Azolla plants grown in ponds and
rice fields in the Philippines showed that at least
half of the samples were deficient in P
(Watanabe and Ramirez, 1990). The average
available P content of soils where Azolla plants
grew was 25 ppm - higher than the average for
Philippine soils (probably below 10 ppm) suggesting that Azolla could grow in soils with
higher available P content. Averages of plant P
content in A. microphylla and available P content of soils where this species grew, were higher
than those in A. pinnata, suggesting that the
requirement for available P was higher in A.
microphylla than in A. pinnata. Water-culture
experiments also showed that A. pinnata grows
better under low-P condition than A. microphylla and A. mexicana (Kushari and Watanabe,
1991). Split application of P to nursery beds of
Azolla can greatly increase the efficiency of P in
N production by Azolla (Watanabe et al., 1988).
After inoculation to the field, Azolla enriched
with P can multiply 7-10 times until it becomes
P-deficient .
A. filiculoides is least tolerant of high temperature. Introduction of A. filiculoides in China
in the early 1970s expanded the use of Azolla in
the North of China (Liu and Zheng, 1989).
Although A. microphylla is tolerant of high temperature, growth is greatly reduced at water
temperature exceeding 40°C. In the fields, damage by heat is confounded with insect damage
because, as temperature increases, insect growth
increases and growth of Azolla decreases. Major
insect pests are Pyrlalidae ( Ephestiopsis and
Elophila ) (Mochida et al., 1987) in the humid
tropics, and Chironomidae and snails in addition
to Pyrlalidae in China (Liu and Zheng, 1989).
Use of chemical pesticides in the fields is not
economically feasible (Kikuchi et al., 1984).
Biological control should be sought. Moist soil
culture prevents damage by insect and high temperature, because the absence of floodwater inhibits the movement of aquatic insect larvae (Liu
and Zheng, 1989). No strong tolerance for
pyralid pests was found among Azolla strains
(Liu and Zheng, 1989).
Recent development in genetic enhancement
Although selection of Azolla species or strains
and improvement of cultural practices were partly successful in overcoming constraints and in
increasing N2-fixing activity, genetic enhancement of the Azolla-Anabaena system is needed.
This has recently been accomplished by sexual
hybridization and algal inoculation.
An easily sporulating strain of A. filiculoides,
introduced in China in the 1970s, gave material
for studying the sexual life cycle of Azolla. In
1979, scientists in Jiangsu and Fubei Provinces
reported the hybridization of Azolla. Wei et al.
(1986) reported the hybridization between A.
filiculoides and A. microphylla. When A.
filiculoides was a female, a higher number of
albinos appeared in the F1. Hybridization was
confirmed by isozyme pattern of esterase. Wei et
al. (1986) also reported crosses between A. mexicana and A. filiculoides. In all of the F1, no
megasporocarps (female organ) were formed,
and microspores were often aborted and became
non-circular during their development. Improved
tolerance for heat and snail attack was observed
in hybrids. The annual biomass production of a
hybrid ( A. microphylla x A. filiculoides ) was
better than that of the parents in Fuzhou (Table
2) (Liu and Zheng, 1989). In spring, low temTable 2. Biomass production of hybrid Azolla and its parents
in Fuzhou. China (FAO, 1988)
Species
Fresh weight (t ha-1)
Apr.-Jul.
Hybrid
A. microphylla (female)
A. filiculoides (male)
A. caroliniana
a
a
118
79
101
93
Aug.-Nov.
43
50
25
44
Total/year
161
129
126
137
A. microphylla X A. filiculoides, Strain name-Rong-Pin 1-4.
Improving N2-fixing systems in rice farming
perature-tolerant A. filiculoides grew better than
high temperature-tolerant A. microphylla. In
summer and autumn, the opposite was found. In
spring, the hybrid produced biomass comparable
to that of A. filiculoides; in summer and autumn
it was comparable to A. microphylla. Consequently, annual production of the hybrid in
Fuzhou was higher than that of the parents
(Table 2) (Liu and Zheng, 1989). Do van Cat et
al. (1989) also reported hybrids between A. microphylla and A. filiculoides (IRRI ac. No. 4028,
and 4030). Hybrids were confirmed by zymograms of three enzymes. Megasporocarps were
not formed and microspores were also abnormal.
The N content of hybrids grown in IRRI’s field
or laboratory was higher than that of the parent
A. microphylla. Biomass production of hybrids
in IRRI fields was higher than that of A. microphylla except in May, the hottest month.
These results indicate that sexual hybridization
could change Azolla’s reactions to environments.
Scientists in University of Hanoi (Do van Cat,
person. comm.) and the University of Philippines
at Los Baños (P.C. Payawal, personal comm.)
reported the hybridization between A. microphylla and A. mexicana. Although cyanobacteria were isolated from the fern, their immuno-chemical and DNA properties were not
identical with those of the endosymbiont (Zimmerman et al., 1989). Lin et al. (1989) showed
that the inoculation of one of such isolated
Anabaena could not establish complete symbiosis. It is unlikely that genuine cyanobionts
have been isolated. Anabaena cells are present
beneath the indusium of the megasporocarp and
transfer to a new generation of sporophytes. By
transferring the indusium to Anabaena-free
megasporocarps of another species, Lin et al.
(1989) succeeded in transferring Anabaena of A.
microphylla to A. filiculoides and vice versa (Fig.
1). A. filiculoides having a cyanobiont from A.
microphylla became more tolerant of heat (37°Cday/29°C-night) than A. filiculoides, but less tolerant than A. microphylla, suggesting that heat
tolerance is partly controlled by Anabaena
(Watanabe et al., 1989).
Although these achievements in sexual hybridization and the exchange of cyanobionts were
preliminary, the methods opened a way to
genetic enhancement and study of the mecha-
61
Fig. 1. Diagram for exchanging symbiotic Anabaena in
megasporocarps.
nisms of Azolla-Anabaena symbiosis. We expect
that some of the unfavorable characteristics of
various Azolla species - high P requirement, low
nutrient content and low digestibility to animals,
susceptibility to high temperature, and sensitivity
to insect attack - may be improved by hybridization or by changing the symbiotic cyanobacteria
or even by introducing foreign DNA-like insect
resistance genes to either the fern or the
cyanobiot.
For genetic enhancement, knowledge of the
genetic or strain or species variabilities is required. Table 3 summarizes known variabilities
in various traits. The gene pool for these widely
variable characteristics is, however, still virtually
unanalyzed.
Multiple uses
Recently, multiple benefits from Azolla besides
being an organic N fertilizer have been recognized. These are as (a) weed suppressor (Janiya
and Moody, 1981), (b) K scavenger from floodwater (Liu, 1987), (c) animal feed, (d) fish feed,
62
Watanabe and Liu
Table 3. Genetic variabilities of Azolla in various traits
Traits
Species
Among
strains
Mutants
or hybrids
Details"
Sporulation
Yes
Yes
Yes
In hybrids, no female spores
Heat tolerance
Yes
Yes
Yes
Cold tolerance
P deficiency
Yes
Yes
?
Yes
?
Yes
Amic, Amex > Api > Af
Latin America Af > Europe af
Af > others
Ap > Af, Amic, Amex
A mutant 4138
NaCl tolerance
Yes
Yes
Yes
AI tolerance
Yes
Yes
Yes
Appetency to fish
Inhibition of N2 fixation
by NH 4
Insect tolerance
Yes
Yes
Yes
Yes
Yes
Yes
References
?
?
Amic, Amex > Ap
A mutant 8028
Ap > other
A mutant 413
Amic > Api
App > others
Wei et al. (1986)
Payawal and Paderson (1987)
Watanabe, unpublished
Watanabe, unpublished
Li et al. (1982)
Kushari and Watanabe (1991)
Liu and Watanabe, unpublished
Desmadyl et al. (1988)
Liu and Zheng (1989)
H. Brunner, IAEA, pers. comm.
Wauthelet et al. (1988)
Liu, unpublished
Antonie et al. (1986)
Okoronko et al. (1989)
?
App >others
Mochida, IRRI, pers. comm.
a Ap: A. pinnata, Api: A. pinnata var. imbricata, App: A. pinnata var. pinnata, Af: A. filiculoides, Amic: A. microphylla, Amex:
A. mexicana. Numbers are accession number of IRRI Azolla Collection.
(e) P scavenger in sewage treatment (Shiomi and
Kitoh, 1987), and (f) suppressor of ammonium
volatilization (Villegas and San Valentin, 1989).
The use as fish feed and effect on losses of
inorganic N fertilizer are described here.
Rice-Azolla-fish culture
Asian farmers have been rearing fish with
flooded rice culture. Fish production is limited,
however, by aquatic biomass production in
flooded rice fields. The Fujian Academy of Agriculture Science (FAAS) developed the riceAzolla-fish system to increase farmers' income
by the higher production of fish and reduced
consumption of fertilizers and pesticides (FAO,
1988; Liu and Zheng, 1989).
The essence of this technique lies in (a) the
mixture and proper ratio of fish with different
eating and rearing habits, (b) proper ratio of the
area for pits and ditches to rice-planted area in
one plot of the field, (c) double narrow row
planting of rice, (d) mixture of Azolla species,
and (e) supplement of fish feed, if necessary.
Three or more kinds of fish - grass carp (herbivorous), Tilapia nilotica (omnivorous), and
other omnivorous fish - are mixed. Rice spacing
is such as to allow sufficient space to fish without
reducing rice yield. Pits and ditches are dug in
the rice field, and rice is often grown on the
ridges.
Several kinds of Azolla are mixed to give
stable biomass production in different seasons.
The mixture of fish having different eating habits
may accelerate N transformation among Azollafish-fish faeces-rice. Experiments with 15 N-labelled Azolla showed Azolla N to be incorporated
into fish protein and the rice plant. The recoveries of Azolla N labelled with 15 N were 26%
to rice and 35% in soil and floodwater (39%
loss) in the rice-Azolla system; in the riceAzolla-fish system, 27% was recovered in fish,
23% in rice, and 35% in soil and floodwater
(15% loss). These data show a more efficient use
of biologically fixed N in the rice-Azolla-fish
system. Potassium is also recycled in the Azollafish-fish faeces system. Consequently, this system
could reduce substantially the use of inorganic
fertilizers. The soil surface in rice fields and fish
pits was enriched with N, P, and K. The incidence of rice disease and insect pest also decreased. This system, therefore, not only protects the environment, but also increases farmers' income. The experiments to reduce the use
of chemical fertilizer and pesticides confirmed
that this system increases farmers' income by
producing fish without loss of rice yield and by
reducing inputs (Table 4). This system could give
an additional income of about $2000/ha per
year, compared to the conventional two rice
crops (FAO, 1988; Liu and Zheng, 1989).
Improving N2-fixing systems in rice farming
63
Table 4. Experiments in new farming (rice-Azolla-fish) system at Fujian Acad. Agric. Sci. (1987-1988)
Azolla applied
(t ha -1)
Fertilizers used
(kg ha-1 yr-1)
N
SP a
K2O
Fish yield
(t ha -1)
Rice yield
(t ha -1)
Early
Late
Total
1987
Rice-rice system
Rice-Azolla-fish
207.8
37
194.3
42.25
127.4
42.25
/
30
6.0
5.8
6.4
4.4
13.0
10.2
/
4.0
1988
Rice-rice system
Rice-Azolla-fish
274.8
67.3
180.4
69.4
152.9
50.9
/
45
4.7
4.4
4.8
4.5
9.5
8.9
/
4.3
Superphosphate.
a
Agency) / FAO (Food Agriculture Organization
of the United Nations)/SIDA (Swedish International Aids Agency) project on the use of 15N in
N2 fixation of Azolla and blue-green algae confirmed the beneficial effects of Azolla cover on
reducing ammonia volatilization (Table 5).
Effect of Azolla on the fate of applied inorganic
N fertilizer
The rise in floodwater-pH because of the photosynthetic activity of the aquatic community
stimulates N volatilization from surface-applied
urea. Under the Azolla cover, floodwater-pH
does not rise because of light interception by
Azolla (Kroeck et al., 1988). This may decrease
ammonia volatilization. Furthermore, the applied N is absorbed by Azolla and used by rice.
A small-scale experiment in the laboratory
showed ammonia volatilization to be 7-40% of
applied ammonium, depending on pH, temperature, and wind speed. The presence of an Azolla
cover reduced ammonia volatilization by 2050% of that without Azolla (Villegas and San
Velentin, 1989). The experiment in FAAS, using
15N-labeled urea, showed that 50% of urea N
was lost when urea was applied to floodwater 2
weeks after transplanting rice. When N fertilizer
was applied to the Azolla mat. the loss was 42%.
When Azolla was further grown 2 weeks after
applying urea and incorporated in the soil, recovery of 15N in rice further increased, and loss
was reduced to 25%. The experiments coordinated by IAEA (International Atomic Energy
Nitrogen-fixing aquatic legumes
Exploration of fast-growing aquatic legumes as
green manure crops
Leguminous green manure crops have been used
as green manures for rice. The major ones are
Astragalus sinicus in the temperate region and
Sesbania cannabina (syn. aculeata ) in the tropics.
Although green manure use generally decreased,
it has continued in some Asian regions like
Bangladesh and some parts of India. The interest
in green manure legumes was revived recently
(Ladha et al., 1988; Meelu and Morris, 1988).
Under the conditions of intensive agriculture, in
which land and time are major constraints, green
manure crops can be inserted in the cropping
calender when the growth interval is short, or
Table 5. Effects of Azolla cover and incorporation on rice grain yields and
the use of 15N in Azolla and blue-green algae in rice fields. 1989) a
Treatment
Indonesia
Yield
(t ha-1)
Urea
Azolla + urea
Incorporated
a
4.2
4.2
4.4
Sri Lanka
N
%
recovery
15
59(76)
62(86)
59(90)
Yield
(t ha-1)
2.9
3.2
3.8
N
%
recovery
15
27
35
42
N recovery in plants and in soil and plant in parentheses.
15
N recoveries (IAEA/FAO/SIDA joint project on
15
Beijing, China
Fuzhou, China
Philip.
Thailand
Yield
(t ha-1)
N
%
recovery
Yield
(t ha-1)
N
%
recovery
Yield
(t ha-1)
Yield
%
recovery
16
30
4.1
4.2
4.3
25(50)
29(58)
35(75)
4.7
6
6.7
3.5
4.0
4.7
10.1
10.4
15
15
Watanabe and Liu
can be grown as an intercrop. In particular, the
discovery of the potential as green manure of
stem-nodulating
and
fast-growing
aquatic
legumes - Sesbania rostrata (Dreyfus and Dommergues, 1981) and various Aeschynomene
species (Alazard and Becker, 1987) - led to new
ways in using green manure legumes in ricebased farming systems. Many species of Sesbania
and Aeschynomene have nodules on the stems
(Ladha et al., 1991). There are three patterns of
stem nodulation.
Type A: Profuse nodulation all over the aerial
stem. S. rostrata, A. afraspera, A.
nilotica
Type B: Scarce nodulation on the aerial stem,
but good nodulation on the stem portion submerged in water. A. aspera,
A. cilita, A. denticulata, A. evenia,
A. indica, A. pratensis, A. rudis, A.
schimperi, A. scabra, A. sensitiva, A.
tambacoundensis.
Type C: No nodulation on the aerial stem and
scarce nodulation on the submerged
stem. A. crassicaulis, A. pfundii, A.
elaphroxylon, S. sesban, S. punctata,
S. speciosa, S. javanica, Neptunia
oleracea
Because of stem nodules, the plant can continue to fix N2 under submerged conditions and
can be grown before wet-season rice, when
flooding frequently occurs. N2 fixation by stem
nodules is more tolerant of combined N than is
that by root nodule. Field-and pot experiments
showed that N2 fixation by S. rostrata was stimulated at 30 kg N ha -1 , and was not inhibited at
60 kg N ha -1 (Becker et al., 1988; 1990). Among
them, S. rostrata and A. afraspera, which originated from Africa, have been studied most (Becker et al., 1988; 1990; Ladha et al., 1988; Meelu
and Morris, 1988). Experiments in the Philippines showed that within 45 days S. rostrata
could accumulate 40-110 kg N ha -1 and A. afraspera 80-90kg N ha -1 . N accumulation varied
with variation in period to flowering. S. rostrata
is more sensitive to photoperiodism than is A.
afraspera. Days to accumulate 100 kg N ha -1 in
the Philippines ranged from 41 (May planting) to
61 (December planting). In S. rostrata, the val-
Rice yield (t/ha)
64
season
Fig. 2. Rice grain yields, as affected by various sources of N
(1985-1989at IRRI).
ues ranged between 45 and 50 d in A. afraspera
(Becker et al., 1988; 1990). The difference in
day length is only 1.5 h in a year. S. rostrata is,
therefore, suitable for the long-day season. A.
afraspera has lower C/N ratios (13-16 at 7-week
growth) than S. rostrata (22-24). Because the
bulky biomass of these green manure crops requires additional labor for incorporation, a green
manure crop with higher N content or less
biomass at the same rate of applied N is preferable (Becker et al., 1988; 1990).
Urea, Azolla, S. rostrata, and A. afraspera
have been compared as sources of N in IRRI
fields. Azolla was incorporated twice before
transplanting rice and once after transplanting.
Legumes were grown for 40 to 50 d before incorporation. The average amount of N incorporated
per hectare over 4.5 yr (9 crop cycles) was 57 kg
urea N, 84 kg N from Azolla, and 73 kg N from
legume green manures. Azolla produced more
biomass N than legumes during the dry season,
but not during the wet season. Average rice
yields were 4.5 t ha -1 in the control, 6.0 t ha -1
with urea. 6.6 t ha -1 with Azolla, and 6.1 t ha -1
with legume green manures (Fig. 2).
Areas with potential for aquatic leguminous green
manures
Because rainfed fields afford few opportunities
for using chemical fertilizers and have uncontrolled water regimes, aquatic legume green manure would be more suitable there than in irrigated rice fields. Tolerance of aquatic legumes
for flooding allows their continuous growth after
flooding, which often occurs at the beginning of
the wet season. The usefulness of Sesbania in
Improving N2-fixing systems in rice farming
infertile rainfed rice areas is being studied
(Arunin et al., 1988). S. rostrata is tolerant of
salt, and is grown in salt-affected areas as a
source of mineral nutrients and organic matter
(Arunin et al., 1988). Addition of organic matter
is effective in preventing salt accumulation and
in accelerating leaching of salt (H. Wada, Agricultural Development Research Center of
Northeast Thailand, personal communication).
Legumes planted before the wet season save soil
N, which is otherwise lost by leaching or denitrification by heavy rain and flooding. This effect is
an additional benefit of aquatic legumes, discussed by George et al. (1992).
Constraints of fast-growing, stem-nodulating
leguminous plants are low seed production,
labor-requiring incorporation of bulky biomass
into the soil, S. rostrata’s high sensitivity to
photoperiod, and insect incidence.
To overcome these constraints and to find
green manure crops suitable for different conditions, exploration of various aquatic legumes,
particularly Aeschynomene species, has been
started.
Genetic improvement of recently explored
aquatic legumes is not yet in progress. As in
Azolla, the means of increasing income by growing leguminous plants should be considered.
Concluding remarks
As discussed above, nitrogen-fixing organisms
give many benefits additional to accumulating N
from the atmosphere. These benefits should be
fully used. Nitrogen-fixing agents like Azolla and
legumes contain not only N, but also useful
substances like carbohydrates, amino acids, and
vitamins. When these plants are incorporated
into the soil, the organic components are decomposed. Mineralized N accumulated in soil may be
lost, or may contaminate the environment before
its absorption by the rice plant. Integrated use of
BNF aims at multi-purpose uses and recycling of
nutrients among animal, plant, and soil. The
proteins in the primary producer (BNF agents)
are converted into animal-or fish proteins. Excreta of an animal or fish are decomposed to
provide mineral nutrients to BNF agents and rice
or to provide energy and carbon for biogas pro-
65
Integrated system
Fig. 3. Integrated use of BNF.
duction. The effluent of a biogas digester provides mineral nutrients to BNF agents and rice in
the rice fields (Fig. 3). A Philippine farmer
practices integrated use of Azolla on his small
(0.7 ha) farm to support his family (Johnson,
1988). Azolla is used as animal feed and fertilizer
to rice and upland crops. Excreta of animals and
unconsumed feed are used for biogas and the
effluent of the digester is returned to the rice
field. No P fertilizer is applied to Azolla, because
the effluent from the biogas digester can provide
that P. This is an example of sustainable agriculture with reduced dependence on external
sources.
Genetic enhancement of BNF agents used in
rice fields is far less developed than is that of
grain legumes. The improvement of BNF systems by traditional breeding and modern
biotechnology is absolutely necessary. For Azolla, this effort is in progress. For aquatic leguminous plants, the exploration of genetic variabilities
among species and population has just begun.
References
Alazard D and Becker M 1987 Aeschynomene as green
manure for rice. Plant and Soil 101, 141-143.
Antonie T, Carro S. Mecha J C and Van Hove C 1986
Comparative appetency for Azolla of Cichlasoma and
Oreochromis tilapia. Aquaculture 53, 95-99.
Arunin S. Dissataporn C, Anuluxtipan Y and Nana D 1988
Potential of Sesbania as green manure in saline soils in
Thailand. In Sustainable Agriculture. Green Manure in
Rice Farming. pp 83-95. International Rice Research Institute. Los Baños, Philippines.
66
Watanabe and Liu
Becker M, Ladha J K and Ottow J C G 1988 Stem-nodulating legumes as green manure for lowland rice. Philip. J.
Crop Sci. 13, 121-127.
Becker M, Ladha J K and Ottow J C G 1990 Growth and
N2-fixation of two stem-nodulating legumes and their effect
as green manure in lowland rice. Soil Biol. Biochem. 22,
1109-1119.
Desmadyl D, Godard P, Wauthelet M and Van Hove C 1988
Phosphorus needs and accumulation potential of various
Azolla species and strains. Proc. 3rd Meeting of African
Association for Biological Nitrogen Fixation 7-12 Nov.
1988, Dakar, Senegal.
Do V C, Watanabe I, Zimmerman W J, Lumpkin T A and
De Waha Baillonville T 1989 Sexual hybridization among
Azolla species. Can. J. Bot. 67, 3482-3485.
Dreyfus B L and Dommergues Y R 1981 Nitrogen fixing
nodules induced by Rhizobium on the stems of the tropical
legumes Sesbania rostrata. FEMS Microbiol. Lett. 10, 313317.
FAO 1988 The ‘Rice-Azolla-Fish’ System. RAPA Bulletin
Vol. 4.
George T, Ladha J K, Buresh R J and Garrity D P 1992
Managing native and legume-fixed nitrogen in lowland
rice-based cropping systems. Plant and Soil 141, 69-91.
Janiya J D and Moody K 1981 Weed suppression in transplanted rice with Azolla pinnata R. International Pest
Control 5, 136-137.
Johnson A 1988 Azolla-the all-purpose crop. Asian Agribusiness 5, 14-15.
Kikuchi M, Watanabe I and Haws L D 1984 Economic
evaluation of Azolla use in rice production. In Organic
Matter and Rice. pp 569-592. The International Rice
Research Institute. Los Baños, Philippines.
Kroeck T, Alkaemper J and Watanabe I 1988 Effect of an
Azolla cover on the conditions in floodwater. J. Agron.
Crop Sci. 161, 185-189.
Kushari D P and Watanabe I 1991 Differential responses of
Azolla to phosphorus deficiency. Part I. Screening methods
in quantity-controlled condition. Soil Sci. Plant Nutr. 37,
271-282.
Ladha J K, Pareek R P, So R and Becker M 1991 Stem
nodule symbiosis and its unusual properties In Nitrogen
Fixation: Achievements and Objectives. Eds. PM Gresshoff, LE Roth, G Stacey and W E Newton. pp 633-640.
Chapman and Hall, New York; London.
Ladha J K, Watanabe I and Saono S 1988 Nitrogen fixation
by leguminous green manure and practices for its enhancement in tropical lowland rice. In Sustainable Agriculture:
Green Manure in Rice Farming. pp 165-183. International
Rice Research Institute, Los Baños, Philippines.
Latorre C, Lee J H, Spiller H and Shanmugam K T 1986
Ammonium ion-excreting cyanobacterial mutants as a
source of nitrogen for growth of rice: A feasibility study.
Biotechnol. Letters 8, 507-512.
Li Z X, Zhu S X, Mao M F and Lumpkin T A 1982 Study on
the utilization of 8 Azolla species in agriculture. Zhungguo
Nongye Kexue 1, 19-27.
Lin C, Liu Z Z, Zheng D Y, Tang L F and Watanabe I 1989
Re-establishment of symbiosis to Anabaena-free Azolla.
Science in China 32B, 551-559.
Liu C C 1987 Reevaluation of Azolla utilization in agricultural production. In Azolla Utilization. pp 67-76. International Rice Research Institute, Los Baños, Philippines.
Liu Z Z (Liu C C) and Zheng W W 1989 Azolla in China.
Agric. Publishing House, Beijing. 427 p.
Lu S Y 1987 Methods for using Azolla filiculoides sporocarps
to culture sporophytes in the field. In Azolla Utilization.
pp 27-32. International Rice Research Institute, Los
Baños, Philippines.
Meelu O and Morris R 1988 Green manure management in
rice based cropping systems. In Sustainable Agriculture:
Green Manure in Rice Farming. pp 209-222. International
Rice Research Institute, Los Baños, Philippines.
Mochida O, Yoshiyasu Y and Dimaano D 1987 Insect pests
of Azolla in the Philippines. In Azolla Utilization. pp
207-222. International Rice Research Institute, Los
Baños, Philippines.
Nierzwicki-Bauer S A 1990 Azolla-Anabaena symbiosis: Use
in agriculture. In Handbook of Symbiotic Cyanobacteria.
Ed. R Haeselkorn. pp 119-136. CRC Press, Boca Raton,
FL.
Okoronko N, Van Hove C and Eskew D L 1989 Evaluation
of nitrogen fixation by different strains of the AzollaAnabaena symbiosis in the presence of a high level of
ammonium. Biol. Fert. Soils 7, 275-278.
Payawal P C and Paderson E M 1987 Sporocarps formation
and spore germination of Azolla strains in the Philippines.
Phil. Agric. 69, 633-643.
Reddy P M and Roger P A 1988 Dynamics of algal populations and acetylene-reducing activity in five rice soils inoculated with blue-green algae. Biol. Fertil. Soils 6, 14-21.
Roger P A, Grant I F, Reddy P M and Watanabe I 1987 The
photosynthetic aquatic biomass in wetland rice fields and
its effect on nitrogen dynamics In Efficiency of Nitrogen
Fertilizers for Rice. pp 46-68. International Rice Research
Institute, Los Baños, Philippines.
Roger P A and Watanabe I 1986 Technologies for utilizing
biological nitrogen fixation in wetland rice: Potentialities,
current usage, and limiting factors. Fertil. Res. 9, 39-77.
Roger P A 1991 Reconsidering the utilization of blue-green
algae in wetland rice cultivation. In Biological N2-Fixation
Associated With Rice Production. pp 119-141. Oxford and
IBH Publ.. New Delhi, India.
Roger P A and Ladha J K 1992 Biological N2 fixation in
wetland rice fields: Estimation and contribution to nitrogen
balance. Plant and Soil 141, 41-55.
Shi D J and Hall D O 1988 The Azolla-Anabaena association: Historical perspective. symbiosis and energy metabolism Bot. Rev. 54, 353-386.
Shiomi N and Kitoh S 1987 Use of Azolla as a decontaminant
in sewage treatment. In Azolla Utilization. pp 169-176.
International Rice Research Institute, Los Baños, Philippines.
Vaishampayan A 1984 Biological effects of a herbicide on a
nitrogen-fixing cyanobacteria (blue-green alga): An attempt for introducing herbicide-resistance. New Phytol. 96,
7-11.
Villegas G G and San Valentin G O 1989 Effect of Azolla
cover on nitrogen and rice in flooded Mahaas clay. In
Azolla: Its Culture, Management and Utilization in the
Improving N2-fixing systems in rice farming
Philippines. pp 65-90. National Azolla Action Program,
The Univ. of the Philippines, Los Baños, Philippines.
Watanabe I 1986 Problems of using nitrogen fixation in the
tropics. In Nitrogen Fixation with Non-Legumes. Eds. F A
Skinner and P Uomala. pp 343-357. Martinus Nijhoff
Publishers, The Hague, The Netherlands.
Watanabe I and Cholitkul W 1990 Phosphorus as a factor
limiting nitrogen fixation in flooded rice soils. In Phosphorus Requirement for Sustainable Agriculture in Asia and
Oceania. pp 281-294. International Rice Research Institute, Los Baños, Philippines.
Watanabe I, Lapis T M, Oliveros R and Ventura W 1988
Improvement of phosphate fertilizer application to Azolla.
Soil Sci. Plant Nutr. 34, 557-569.
Watanabe I, Lin C and Ventura T 1989 Responses to high
temperature of the Azolla-Anabaena association. determined in both the fern and in the cyanobacterium. New
Phytol. 111, 625-630.
Watanabe I and Ramirez C 1990 Phosphorus and nitrogen
67
contents of Azolla grown in the Philippines. Soil Sci. Plant
Nutr. 36. 319-331.
Watanabe I and Roger P A 1984 Nitrogen fixation in wetland
rice field. In Current Developments in Biological Nitrogen
Fixation. Ed. N S Subba Rao. pp 237-276. Oxford & IBH,
New Delhi.
Wduthelet M. Desmadyl D, Godard P and Van Hove C 1988
Sensitivity to aluminum of various Azolla species and
strains. Proc. 3rd Meeting of African Association for
Biological Nitrogen Fixation 7-12 Nov. 1988, Dakar,
Senegal.
Wei W X. Jin G Y and Zhang N 1986 Preliminary report on
Azolla hybridization studies. Bull. Fujian Acad. Agric. Sci.
1, 73-79. (In Chinese).
Zimmerman W J. Rosen B H and Lumpkin T A 1989
Enzymatic, lectin, and morphological characterization and
classification of presumptive cyanohionts from Azolla Lam.
New Phytol. 113. 497-503.
Plant and Soil 141: 69-91, 1992.
© 1992 Kluwer Academic Publishers. Printed in the Netherlands.
PLSO SA05
Managing native and legume-fixed nitrogen in lowland rice-based
cropping systems*
T. GEORGE 1,2, , J.K. LADHA 1 , R.J. BURESH 1,3 and D.P. GARRITY 1
1 International Rice Research Institute (IRRI), P.O. Box 933, Manila, Philippines,
2
Nitrogen Fixation by Tropical Agricultural Legumes (NifTAL) Project, University of Hawaii,
1000 Holomua Avenue, Paia, HI 96779, USA and
3
International Fertilizer Developement Center (IFDC), P.O. Box 2040, Muscle Shoals, AL 35662,
USA
Key words: biological nitrogen fixation, denitrification, fallow, flooded soil, leaching, legume, nitrate,
nitrogen balance, nitrogen loss, Oryza sativa, rice, weeds
Table of contents
Abstract
Introduction
Major lowland rice-based cropping systems
Year-round lowland rice culture
Lowland rice culture with short non-flooded season
Lowland rice culture with long non-flooded season
Soil nitrogen dynamics in rice lowlands
Soil nitrate accumulation
Loss of nitrate from soil
Nitrogen dynamics of major lowland rice-based cropping systems
Impacts of nitrate loss from rice lowlands
Legumes in lowland rice-based cropping systems
Biological nitrogen fixation by legumes in rice lowlands
Rhizobial determinants
Plant determinants
Environmental determinants
Nitrogen contribution from legumes in rice lowlands
Food and forage legumes
Stem-nodulating aquatic green manure legumes
Legumes in the post-rice as opposed to the pre-rice niche
Maximizing total nitrogen capture by legumes in rice lowlands
Managing soil nitrate supply
Allowing for increased nitrate uptake
Minimizing soil supply of nitrate
Managing inefficient nitrate uptake by legumes
* IRRI-NifTAL-IFDC joint contribution.
70
George et al.
Use of legume nitrogen for lowland rice
Belowground residues
Aboveground residues
Nitrogen release from legume residues
Conclusions and research directions
References
Abstract
Lowlands comprise 87% of the 145 M ha of world rice area. Lowland rice-based cropping systems are
characterized by soil flooding during most of the rice growing season. Rainfall distribution, availability
of irrigation water and prevailing temperatures determine when rice or other crops are grown. Nitrogen
is the most required nutrient in lowland rice-based cropping systems. Reducing fertilizer N use in these
cropping systems, while maintaining or enhancing crop output, is desirable from both environmental
and economic perspectives. This may be possible by producing N on the land through legume biological
nitrogen fixation (BNF), minimizing soil N losses, and by improved recycling of N through plant
residues. At the end of a flooded rice crop, organic- and NH4-N dominate in the soil, with negligible
amounts of NO,. Subsequent drying of the soil favors aerobic N transformations. Organic N mineralizes
to NH4, which is rapidly nitrified into NO,. As a result, NO, accumulates in soil during the aerobic
phase. Recent evidence indicates that large amounts of accumulated soil NO, may be lost from rice
lowlands upon the flooding of aerobic soil for rice production. Plant uptake during the aerobic phase
can conserve soil NO, from potential loss. Legumes grown during the aerobic phase additionally
capture atmospheric N through BNF. The length of the nonflooded season, water availability, soil
properties, and prevailing temperatures determine when and where legumes are, or can be, grown. The
amount of N derived by legumes through BNF depends on the interaction of microbial, plant, and
environmental determinants. Suitable legumes for lowland rice soils are those that can deplete soil NO,
while deriving large amounts of N through BNF. Reducing soil N supply to the legume by suitable soil
and crop management can increase BNF. Much of the N in legume biomass might be removed from the
land in an economic crop produce. As biomass is removed, the likelihood of obtaining a positive soil N
balance diminishes. Nonetheless, use of legumes rather than non-legumes is likely to contribute higher
quantities of N to a subsequent rice crop. A whole-system approach to N management will be necessary
to capture and effectively use soil and atmospheric sources of N in the lowland rice ecosystem.
Introduction
Rice ( Oryza sativa ) is the major staple food crop
of the tropical world and rice-based cropping
patterns are prevalent in much of the tropics.
Rice is grown on 145 M ha, 87% of which is in
wetland (lowland) culture, with the remaining
grown on dry lands. The annual cultivation
period for rice is determined by the distribution
of rainfall, availability of irrigation water, and
temperatures. Upland crops are typically grown
during the nonflooded period, particularly in the
transition seasons either at the end of the rains
or before the onset of heavy rainfall. Coolseason crops, such as wheat and winter legumes,
are grown during the nonflooded season in areas
where cool temperatures limit rice production.
Lowland rice is grown in bunded fields and is
surface-flooded during most of the growing season. Rice soils are typically flooded and puddled,
and rice seedlings are then transplanted. Alternatively, germinated seeds are broadcast onto
flooded soil that is either puddled or nonpuddled. Dry seeding, where seeds are dibbled
or broadcast onto aerobic soil, is also practiced.
The soil is flooded when there is sufficient water
to be impounded.
Nitrogen is the input required in largest quantities for lowland rice production. Soil N and
biological nitrogen fixation (BNF) by associated
Managing native and fixed N in rice-based cropping system
organisms are major sources of N for lowland
rice. More than 50% of the N used by flooded
rice receiving fertilizer N is derived from the
combination of mineralization of soil organic N
(Bouldin, 1986; Broadbent, 1984) and BNF by
free-living and rice plant-associated bacteria
(Roger and Ladha, 1992). The remaining N
requirement is normally met with fertilizer.
However, the efficiency of fertilizer N uptake by
rice is low due to large N losses from flooded soil
(De Datta and Buresh, 1989; De Datta and
Patrick, 1986; Savant et al., 1982). Further, declining or stagnating yields have recently been
noticed with flooded rice culture receiving fertilizer N (Flinn and De Datta, 1984). An additional concern is that the capacity of soil to supply N
may decline with continuous intensive rice cropping necessitating the application of successively
larger quantities of fertilizer N to maintain yields
(Bacon, 1990). On the other hand, large
amounts of soil NO, may be lost from rice
lowlands upon the flooding of aerobic soil before
the establishment of rice (Buresh et al., 1989).
Reducing N fertilizer use while maintaining
the native soil N resource and enhancing crop N
output is desirable from both environmental and
economic perspectives. This may be possible by
obtaining more N on the land through biological
nitrogen fixation (BNF), reducing losses of N,
and by recycling the N captured in vegetation
during the off-season. Buresh and De Datta
(1991) recently reviewed the dynamics and management of N in rice-legume cropping systems.
Their review indicates that soil N losses may be
reduced by fallow management, and legumes can
partially contribute to the N requirement of a
succeeding rice crop. Thus, the management of
indigenous soil N and N derived in situ through
legume BNF poses potentials for enhancing the
N nutrition and N use efficiency of crops and the
total N output from a lowland rice-based cropping system. The integrated application of possible N management tactics will require detailed
knowledge of the N dynamics of the cropping
system as influenced by variation in hydrological
characteristics and diversity of cropping practices.
In this paper, we address the N dynamics for a
range of major tropical rice-growing ecosystems.
We attempt to define the major lowland rice-
71
based cropping systems, and to document their
hydrological, climatic and management effects
on soil N dynamics. We then discuss the use of
legumes in the capture and use of indigenous soil
nitrogen and fixed atmospheric nitrogen in major
rice-based cropping systems.
Major lowland rice-based cropping systems
Rice-based cropping patterns are prevalent in
those tropical areas with 1500 mm or more annual rainfall with at least three consecutive months
of 200 mm per month. The amount and duration
of rainfall normally determine the number of
crops possible. The rates of onset and decline of
rainfall determine the type of rice culture, other
crops in the pattern, and difficulty in tillage
(Harwood and Price, 1976). Growing upland
crops after paddy rice requires soil conversion
from an anaerobic puddled state to an aggregated aerobic condition. This is often difficult in
soils high in montmorillonite clays. These soils
are often left fallow during the dry months even
when irrigation water is available (Palaniappan,
1985).
The number of rice crops grown in lowlands
varies with supplemental water availability and
prevailing temperatures. The type of upland
crops that can be grown are influenced by the
durations and timing of non-flooded periods. For
purposes of this discussion, we distinguish three
broad classes of tropical lowland rice culture in
terms of the annual period in which the soil is
not flooded: (1) continuously flooded or saturated year-round (Fig. 1A), (2) short nonflooded (Fig. lB), and (3) long non-flooded
(Fig. 1C). The three classes may be further
subdivided on the basis of other important determinants of N dynamics and upland crop adaptation, such as temperature and water availability.
Year-round lowland rice culture
Continuous rice production systems were rare
prior to the introduction of photoperiod-insensitive rice cultivars. However, they have become
increasingly common. At the most intensive
limit, three rice crops are successively cultivated
72
George et al.
Fig. 1. Major lowland rice-based cropping systems from a soil nitrogen dynamics perspective. (A) permanent flooding (or soil
saturation), (B) short, non-flood period, (C) long, non-flood period; (i) warm tropics and (ii) cool tropics/subtropics.
per annum under full irrigation, and the fields
remain constantly saturated or flooded. The total
area is not large, but this system is locally prominent in such areas as north Java and Bali, Indonesia.
Much more important are the irrigated
double-crop ricelands, which are very common
in countries at latitudes within 14º of the
equator. Huke (1982) estimated double-cropped
ricelands at 8.8 M ha or 13% of the total irrigated rice area worldwide. Much of the double
cropped land remains flooded or saturated virtu-
Managing native and fixed N in rice-based cropping system
ally year-round. An example is the double-cropped pump irrigated areas in Central Luzon,
Philippines. Some very poorly drained rainfed
and deepwater ricelands also experience yearround waterlogging.
Lowland rice culture with short non-flooded
season
In much of Asia's irrigated double-cropped riceland there is a short dry period of one to two
months between rice crops. This generally occurs
in the late wet season before the dry-season crop
is established, and/or after the dry-season crop is
harvested.
Some rainfed areas that experience long wet
seasons support two successive rice crops, leaving a short dry soil period of two to three
months. The fields are usually left fallow due to
lack of water or difficulty in tillage. In some
areas legumes are produced on residual soil
moisture, or established as relay crops on supplemental irrigation. Examples of this cropping system are rice-rice-grain legumes in southern India
and Philippines, rice-rice-soybeans in Indonesia
and rice - rice - Astragalus sp. in China.
Lowland rice culture with long non-flooded
season
Single-crop rice production is the dominant rice
cultural system. The soil remains non-flooded
from four to eight months of the year. This is the
dominant case in the tropics because most riceland is rainfed, or can be irrigated only in the
wet season (Garrity et al., 1986). It is also
dominant in the subtropics because although the
land is fully irrigated, rice is constrained by cool
temperatures during the winter season. Temperature also affects the soil N dynamics. Therefore, we recognize two subcategories of singlecrop lands, those in the warm tropics, and those
with cool winters.
In single-crop ricelands, the rice crop may
occupy the field from 90 to 210 days. There is
great diversity in the rainfed rice ecosystems.
They range from favorable rainfed conditions, to
drought or submergence-prone shallow (0 to
25 cm), medium deep (25 to 50 cm), or deep
water (>50 cm) flooding regimes. The highly
73
variable surface hydrology has a profound influence on the nature of the N dynamics. Particularly significant is the unpredictable occurrence
of aerobic or dry soil surface conditions during
the growing season, alternating with surface
flooding.
During the off-season, rainfed ricelands are
typically fallowed (De Datta, 1981). The straw
and fallow weed vegetation are subject to grazing by farm livestock. In a minor fraction of the
area with conducive residual soil water-holding
capacity, and/or a high groundwater table, upland crops, including legumes, are grown in the
post-rice season. This practice is most common
where the soil textures are loamy and exhibit
superior tillage properties.
In well-drained ricelands with a gradual onset
of rainfall in the dry-to-wet transition period,
upland crops are grown prior to rice. Very early
duration is advantageous, to permit maturity before the soil waterlogs. Mungbean is a very
common grain legume in the pre-rice niche.
At latitudes above 14°N rice may experience
significant cool-temperature stress during the
coolest periods of the year. The establishment of
dry-season rice crops is deliberately delayed to
avoid this damage. Above 20° N dry-season rice
cultivation is generally not feasible due to cold
stress. Winter cereals or legumes are often
grown in rotation with rice. The rice-wheat cropping pattern is dominant, occupying some
18 M ha in South Asia and China. Rice-legume
patterns are also common. These often include
chickpea, lentil, or soybean.
Soil nitrogen dynamics in rice lowlands
Fertilizer N transformations in flooded rice fields
have been researched extensively (De Datta and
Buresh, 1989; De Datta and Patrick, 1986).
However, little attention has been paid to N
transformations during the dry season when the
soil is usually aerated (Buresh and De Datta,
1991). Knowledge is particularly lacking on the
dynamics of soil N during the transition periods
between dry and wet seasons.
A conceptualization of the dynamics of soil
mineral N in a lowland, rice-growing soil undergoing a flooded-nonflooded sequence is pre-
74
George et al.
Fig. 2. Dynamics of soil mineral nitrogen in a lowland rice
soil.
sented in Figure 2. Under saturated or flooded
conditions, nitrification is restricted by a limited
O2 supply and NH4 accumulates in the soil. At
the end of a flooded rice crop, organic N and
NH4-N dominate and NO, present is typically
negligible. Subsequent drying of the soil favors
aerobic N transformations. Organic N mineralizes to NH4, which is rapidly nitrified into
NO,. As a result, NO, accumulates in the soil
during the aerobic phase, but is rapidly lost upon
soil flooding.
Plant growth can influence the magnitudes and
forms of soil mineral N (Fig. 3). Nitrate, the
Fig. 3. Conceptual role of plants in conserving soil mineral
and atmospheric nitrogen in a lowland rice-based cropping
system.
predominant form of soil mineral N during the
nonflooded season, will be assimilated by plants
growing during the dry season and dry-to-wet
transition periods. Buresh et al. (1989) found
that soil NO, during the dry season in a lowland
rice soil correlated inversely with the amount of
N accumulated in weeds. Legumes grown during
this season additionally accumulate atmospheric
N through BNF. Part or all of the N in the plant
biomass is removed from the land through the
harvestable products of an economic crop. The
remaining plant N can be returned to the soil as
residues.
Decomposition of plowed-under dry-season
plant residues releases NH4 into the flooded soil.
Rice uptake decreases NH4, the predominant N
form in the flooded soil. Declining NH4 level in
flooded soil during the later part of rice season is
partly replenished through BNF in flooded soil
(Eskew et al., 1981; Yoshida and Yoneyama,
1980) and from decomposition of rice roots and
stubbles in addition to mineralization of soil N.
Soil nitrate accumulation
Nitrate-N accumulates when lowland rice soils
become dry and aerated (Patrick and Wyatt,
1964, Ventura and Watanabe, 1978). Nitrate-N
levels prior to flooding for wet season rice in
several Philippine lowland rice soils ranged from
5 to 39 mg kg-1 soil (Ponnamperuma, 1985).
Buresh et al. (1989) measured 39 to 91 kg soil
NO3-N ha-1 in the top 60 cm layer at the end of
the dry season at three Philippine sites.
The production of NO3 depends on the
amount of organic N available for mineralization. Organic N varies considerably in lowland
rice soils, depending on the level of native organic matter, and the history of cropping and
organic amendments. The amount of N mineralized is increased by higher soil organic N levels
(Broadbent, 1984; Hadas et al., 1986; Sahrawat,
1983). Bacon (1990) reported that successive soil
incorporation of rice stubble gradually increased
the N mineralization potential.
Residual soil water from the preceding wetseason rice crop, intermittent rains, and irrigation cause fluctuating soil water conditions. Soil
water fluctuation generally results in a large production of NO, (Birch, 1960; Herlihy, 1979;
Managing native and fixed N in rice-based cropping system
Sahrawat, 1980; Seneviratne and Wild, 1985). In
a greenhouse study, continuous soil flooding during the dry season prevented NO, accumulation
(Ventura and Watanabe, 1978). Drying, as well
as alternate drying and wetting treatments resulted in the accumulation of soil NO,.
Dry-season fallows are subjected to tillage for
upland crop establishment or for weed control.
Tillage increases soil aeration and tends to enhance the nitrification of mineralized NH4 into
NO, (Dowdell et al., 1983). Herridge (1986)
reported more NO, in a cultivated fallow compared with a non-tilled fallow in a rainfed environment in Australia.
Loss of nitrate from soil
Flooding of aerated soil has generally been accepted as a major factor of NO, loss from rice
soils (Bacon et al., 1986; Buresh et al., 1989;
Janssen and Metzger, 1928; Patrick and Wyatt,
1964; Strickland, 1969; Wallihan, 1938; Yamane,
1957). With the use of 15N microplots, Bacon et
al. (1986) reported 90% reduction in NO3 concentration after flooding of a rice soil. Loss of up
to 91 kg ha"of native NO3-N after flooding of
fallow rice fields in the Philippines has been
reported by Buresh et al. (1989).
Soil flooding can promote leaching losses of
NO,. Downward water flow carries NO, to
lower soil layers (White, 1988). Two direct determinants of NO, leaching are the net downward flux of water and the concentration of NO,
in the leaching water. Measuring these variables
presents considerable problems because of difficulty in sampling and spatial and temporal variability in water flux and NO, concentration
(Keeney, 1986; White, 1988). Large NO, losses
through leaching occur generally from aggregated soils and from soils under bare fallow
(Legg and Meisinger, 1982; White, 1988). These
conditions exist in lowland soils during the
aerobic phase before puddling for rice culture.
Soil saturation depletes soil of its 0,, thus
creating favorable conditions for denitrifying organisms to reduce NO,. Denitrification results in
the production of N2 and N2O (Knowles, 1982).
The rates of N2 and N2O evolution from soils
will be higher under conditions of fluctuating
redox potentials (wetting and drying cycles) than
75
when the redox potential is continuously high
(aerobic
condition)
or continuously
low
(anaerobic condition) (Letey et al., 1981; Reddy
and Patrick, 1975). Mosier et al. (1986) measured large fluxes of N2 and N2O from both
fallow and cropped aerobic soils following water
applications. Conditions favorable for N2 and
N2O emissions may occur during the dry-to-wet
transition period in lowland rice-based cropping
systems.
Bacon et al. (1986) attributed rice soil NO3
loss primarily to denitrification. An incubation
study with the use of 15N-labeled NO3 by Buresh
et al. (1989) indicated that NO3 completely disappeared after 9 days of flooding, and 5% or less
of the added NO3 remained in the soil as either
NH4 or organic N. Since conversion of NO3 to
NH4 was negligible, either denitrification or
leaching appeared to be the major loss mechanism.
Nitrogen dynamics of major lowland rice-based
cropping systems
Soil N is essentially conserved under year-round
flooded rice culture. Maintaining flooding between and during rice crops increases the soil N
supply (Ponnamperuma, 1985; Santiago-Ventura
et al., 1986; Ventura and Watanabe, 1978). Nitrate does not accumulate in the soil since soil
drying between rice crops is minimal. Therefore,
soil NO3 losses associated with flooding events
are not experienced. Rice field-associated BNF
contributes to the maintenance of available soil
N in these soils (Koyama and App, 1979).
In warm, tropical single-crop rice lowlands,
large quantities of NO3 might accumulate in soil
over the prolonged aerated soil phase, then be
lost during the dry-to-wet transition period. The
quantity of NO3 accumulated will depend on
how the fallow is managed. Tillage and occasional rains might increase soil NO3.
In intensively cultivated rice-wheat lowlands,
fertilizer N application in excess of wheat crop
requirements may lead to large quantities of
residual mineral N after wheat harvest. Chaney
(1990) observed significant increases in residual
soil NO3 levels after winter wheat at several sites
in England when the N applications rates ex-
76
George et al.
ceeded the optimum. Soil NO3 after wheat in a
sandy loam soil in India averaged 82 kg N ha-1
(Sharma et al., 1985), but the portion originated
from fertilizer might have been low (Macdonald
et al. , 1989). Sharma et al. (1985) also found
that NO3 accumulation during a mungbean crop
following wheat was substantial. Prior to the
start of the wet season, the soil contained 223 kg
NO3-N ha-1.
Impacts of nitrate loss from rice lowlands
It is well understood that the loss of NO3 via
leaching and denitrification has undesirable environmental consequences. Leaching leads to
NO, pollution of ground-water and N2O emission from denitrification poses a threat to the
stability of the ozone layer (Pretty and Conway,
1989). The significance, if any, of NO, loss from
rice lowlands on the environment, and the negative impacts of NO, loss on the long-term N
fertility of lowland rice soil, have thus far been
largely ignored by researchers.
Available reports suggest possible high losses
of NO, during soil flooding for rice production,
but do not account for the cumulative loss that
can occur during the entire dry and dry-to-wet
transition periods. Lowland rice soils undergo
short-term wetting or flooding, especially from
intermittent early rain near the end of the dry
season. During short-term wetting of the soil,
NO, can be lost through denitrification (Mosier
et al., 1986). Heavy rain would cause substantial
downward water flow, especially in sandy soils,
leading to NO, loss through leaching (White,
1988). It is likely that the total NO, loss is
greater than the reported disappearances of NO,
after soil flooding for rice production.
Legumes in lowland rice-based cropping systems
Legumes are grown in rice lowlands primarily for
food, and to a lesser extent for green manure
and forage. The importance of legumes in rice
lowlands is enhanced by their ability to meet part
of their N requirements from the atmosphere
through BNF and their potential to enhance soil
N fertility. The major food legumes grown include cowpea ( Vigna unguiculata ), mungbean
( Vigna radiata ), peanut ( Arachis hypogaea ),
pigeon pea ( Cajanus cajan ), chick pea ( Cicer
arietinum ), and soybean ( Glycine max ). Major
green manure legumes include milk vetch ( Astragalus sinicus ) and Sesbania spp. in China,
Sesbania spp., sunnhemp ( Crotalaria juncea ) and
berseem clover ( Trifolium alexandrinum ) in
India;
and indigo ( Indigofera tinctoria )
in the Philippines (Garrity and Flinn, 1988;
Singh et al., 1991; Yost and Evans, 1988). The
most promising forage legumes include siratro
( Macroptilium atropurpureum ), lablab bean
( Lablab purpureus ), sunnhemp, and clitoria
( Clitoria ternatea ) (Carangal et al. , 1988). Several of the green manure legumes such as berseem
clover ( T. alexandrinum ) are also suitable as
forages.
The length of the non-flooded season, water
availability, soil properties and climate primarily
determine when and where legumes can be
grown. These together with socio-economic factors determine whether food, forage, or green
manure legumes are grown on ricelands. Under
continuous rice culture, the periods between rice
crops are too short to include legumes. In lowlands where two rice crops are taken, the nonflooded fallow may be long enough to grow a
short-duration legume. Green manure legumes
could be planted during a 40-to-60 day transition
period between rice crops (Garrity and Flinn,
1988). Relay cropping of pulse crops such as
mungbean after the second rice crop is common
in parts of India (Palaniappan, 1985).
In single rice crop lowlands of the warm
tropics with a long nonflooded season, there are
two niches for short-duration legumes. They are
the post-rice dry season for grain or forage types
and the pre-rice dry-wet transition for grain or
green manure types (Garrity and Flinn, 1988).
Long-duration forage, green manure and cover
crop legumes are grown sometimes as well. They
may occupy most of the non-flooded season. In
single rice crop lowlands of the cool tropics/
subtropics, short- and long-duration winter
legumes are grown for food, forage or green
manure.
Managing native and fixed N in rice-based cropping system
Biological nitrogen fixation by legumes in rice
lowlands
The determinants of BNF by legumes have been
dealt with in detail in other recent papers
(Bohlool et al., 1992; Peoples and Craswell,
1992; Peoples and Herridge, 1990). The coverage here will be limited to general principles and
to specific determinants that are applicable to
legume production in lowland rice-based cropping systems. In general, BNF can supply N to a
legume as long as a factor other than N is not
limiting plant growth and an effective Rhizobium
symbiosis is ensured (Bohlool et al., 1992).
Abundant mineral N supply, however, inhibits
or replaces BNF as a N source for legumes
(Munns, 1977; Streeter, 1988). The interplay of
Rhizobium, plant, and environment determines
the proportions of legume N derived from atmospheric and soil sources.
Rhizobia1 determinants
Indigenous soil rhizobia are often ineffective or
partially effective in BNF, but are quite successful in competing for nodule sites with introduced
elite strains. This frequently results in no observed increase in BNF from rhizobial inoculation (Dowling and Broughton, 1986; Ham, 1980;
Meade et al., 1985; Weaver and Frederick,
1974). Nonetheless, a response to inoculation
may not be obtained if the number of native
effective soil rhizobia exceeds 20 to 50 cells g -1
soil (Singleton and Tavares, 1986; Thies et al.,
1991a).
Soil flooding during the rice season decreases
rhizobial numbers, but strains differ in survival
ability (Boonkerd and Weaver, 1982), and often
sufficient numbers of rhizobia for effective nodulation thrive (Weaver et al., 1987). Ladha et al.
(1989a) reported survival in high numbers of
rhizobia of the aquatic legume Sesbania rostrata
in flooded rice rhizosphere. Rain splash and
flooding often promote stem nodulation by S.
rostrata; however, stem inoculation is probably
beneficial under dry conditions (Ladha et al.,
1991). Upland legumes following prolonged soil
flooding may require inoculation to ensure effective symbiosis establishment. With proper
77
Increase in economic yield due to inoculation (%)
Log (1+number of indigenous rhizobia)/g soil
Fig. 4. Model describing legume inoculation response.
Model inputs are numbers of indigenous rhizobia and soil N
mineralization potential. Adapted from Thies et al. (1991b).
characterization of native soil rhizobia and soil N
supplying capacity, the inoculation requirements
of legumes may be predicted for various environments using the models developed by Thies et al.
(1991b). One model which incorporates the
numbers of indigenous rhizobia and soil N mineralization potential is presented in Figure 4.
The model input variables can be obtained
through soil analysis prior to legume planting.
Plant determinants
The potential BNF capacity may be defined as
the aggregate of the per-day deficits in mineral N
uptake during the legume growth cycle (George
and Singleton, 1992). Therefore, the higher the
N yield potential of a legume for a given growth
duration and soil N supply, the higher would be
the proportion and amount of plant N derived
through BNF (Fig. 5). For the sake of simplicity,
it is assumed that the soil N uptake rate remains
constant throughout the growth period which is
unlikely in field environments.
The amount of N derived from BNF is also
altered by the pattern of plant N assimilation
(George and Singleton, 1992). The legumeRhizobium symbiosis becomes functional in BNF
only after the early rhizobial infection and
nodule formation (Hardy et al., 1971; Vincent,
78
George et al.
(1992) determined in the field that common bean
scavenges soil N much more efficiently than soybean, resulting in a lower proportion of N derived from BNF (Table l). In their study, soybean derived 70% of its N from BNF compared
to 23% by common bean at a similar flowering
date, but the mineral N uptake efficiency of
soybean was less than half that of common bean
as determined by 15 N. Consequently, KClextracted mineral N in the soil under soybean
was almost twice as high as that under common
bean.
Environmental determinants
Fig. 5. Conceptual relationship between nitrogen yield
potential and nitrogen derived from BNF under a constant
soil nitrogen supply. It is assumed arbitrarily that the increasing nitrogen yield potential has no influence on the amount of
nitrogen derived from soil.
1980). Therefore, during early legume growth,
soil N is the major source of plant N. It is likely
that a legume capable of meeting its early N
requirement from the soil, but having a subsequent N requirement exceeding the soil N
supply, will benefit from BNF. Consequently,
soybean, which has a high N requirement during
the podfill period, also has a high N2 fixation rate
during the same period (Afza et al., 1987;
George and Singleton, 1992; Imsande, 1988,
1989; Zapata et al., 1987).
The efficiency of mineral N uptake by legumes
further interacts with the N yield potential and
the pattern of N assimilation in determining the
amount of N derived from BNF. Under similar
soil mineral N levels, legumes that are more
efficient in extracting soil mineral N will benefit
less from BNF. Using 15 N, George and Singleton
Major environmental determinants of optimal
BNF performance are soil-related constraints
and climatic stresses. The soil mineral N supply
is of dominant importance. Increased soil mineral N supply invariably decreases the proportion
of N derived from BNF (Deibert et al., 1979;
Eaglesham et al., 1983; George et al., 1988;
George and Singleton, 1992; Gibson and Harper, 1985; Herridge and Brockwell, 1988; Thies
et al., 1991b). However, soil mineral N and BNF
can be complementary in meeting the N requirement of the legume (Bergersen et al., 1985;
Deibert et al., 1979; Eaglesham et al., 1983;
Herridge and Brockwell, 1988; Herridge et al.,
1984).
On the other hand, the supply of mineral N in
early legume growth can enhance BNF, a phenomenon often referred to as the 'starter N'
effect. Insufficient soil N can restrict initial plant
growth which limits subsequent plant growth and
N demand, resulting in reduced need for BNFsupplied N. Alleviation of early N stress increases the amount of N derived from BNF
(Table 2). Soybean plants receiving small
amounts of mineral N formed a higher number
Table 1. Plant N, N derived from BNF, N uptake efficiency a and KCl-extractable soil N at flowering of field grown soybean and
common bean (George and Singleton, 1992)
Legume
Plant N
(kg N ha -1 )
N from BNF
(%)
15 N uptake
(%)
KCl-extr. N
(mgN kg -1 soil)
Soybean
Common bean
58
80
70
23
5
12
17
10
a
N uptake efficiency is determined by using
assimilated by the plant.
15
N isotope as tracer and is the percent of applied
15
N fertilizer (3 kgN ha -1 )
Managing native and fixed N in rice-based cropping system
79
Table 2. Mineral N effects on nodulation, total N, N derived through BNF and trifoliate leaf area by soybeans grown for 33 days
in pots amended with 0.8% bagasse (Singleton P W, 1991; pers. commun.)
Nitrogen
applied
(kg N g 1 soil)
Nodule
(no. pot
0
50
100
200
400
148
173
144
87
66
-1
)
Nodule
weight
(mg pot -1 )
440
651
667
459
147
Total
N
189
316
343
541
593
N from BNF
a
Amount
% of total
168
280
251
219
0
99
95
78
42
0
Area of
trifoliate b
(cm2)
30
41
53
51
42
a
Determined by the N difference method.
b Area of trifoliate at 16 days of growth.
and weight of nodules and derived larger
amounts of N from BNF than plants receiving
either no N or large amounts of mineral N. The
'starter N' effect was due to increased plant
growth as reflected by the increased area of the
trifoliate at 16 days.
From the data presented in Table 2 and in
other reports (Deibert et al., 1979; Eaglesham et
al., 1983; George and Singleton, 1992; Herridge
and Brockwell, 1988; Thies et al., 1991a,b), a
generalized conceptual relationship between soil
N supply and N derived from BNF by legumes
can be illustrated (Fig. 6). Initially, a 'starter N'
effect tends to increase BNF but N from BNF
approaches zero at the highest level of soil N
supply. Increasing N supply, however, tends to
increase total legume N.
Fig. 6. A generalized conceptual relationship between soil N
supply and nitrogen derived from soil or BNF. Data from
Table 2 and data reported by Deibert et al. (1979); Eaglesham et al. (1983); George and Singleton (1992); Herridge
and Brockwell (1988); Thies et al. (1991a,b) are considered
in the proposed relationship.
Other soil constraints include acidity, salinity,
nutrient toxicities, and nutrient deficiencies
(Craswell et al., 1987; Munns, 1977; O'Hara et
al., 1988; Singleton and Bohlool, 1983; Singleton
et al., 1985). Application of P in soils of low P
availability (Cassman et al., 1981; Ssali and
Keya, 1986) and liming acidic soils (Torres et al.,
1988) increases total N accumulation by legumes.
Further, N2-fixing legumes may have differing
requirements of mineral nutrients compared to
mineral N-dependent plants. Insufficient supply
of mineral elements such as Ca, Mo, B and Fe
can particularly limit N2 fixation (O'Hara et al.,
1988; Robson, 1978). External P requirements of
N,-fixing legumes are higher than for mineral
N-dependent legumes because of limited proliferation of the root system (Cassman et al., 1980,
1981).
Likewise, waterlogging or water deficit negatively influences either plant growth or the symbiosis (Bennett and Albrecht, 1984; Chapman
and Muchow, 1985; De Vries et al., 1989a,b,
Kucey et al., 1988; Minchin and Summerfield,
1976; Muchow, 1985; Sinclair et al., 1987; Smith
et al., 1988; Timsina, 1989). Water stress reduces
total N accumulation (Chapman and Muchow,
1985) and, especially during late growth stages,
limits pod addition and subsequent seed N demand in grain legumes (De Vries et al., 1989b),
thereby reducing N derived from BNF.
Additionally, climatic stresses such as low temperature, low solar radiation and unfavorable
photoperiod can deleteriously affect plant and
symbiotic performance (Becker et al., 1990;
George et al., 1988, 1990; Hodges and French,
1985; Lawn and Williams, 1987; Muchow, 1985).
Becker et al. (1990) reported that the number of
80
George et al.
days required to produce 100 kg N by S. rostrata
was higher when the plants were grown during
short days than during long days.
Nitrogen contribution from legumes in rice
lowlands
Inclusion of legumes in the cropping pattern with
rice can increase the total N output from the
cropping system but does not necessarily result
in a positive soil N contribution. The amount of
N derived from BNF, the net effect of the
legume on N losses and the amount of N removed with harvested plant material will determine whether there is a positive net soil N
balance.
Food and forage legumes
The amount of N removed from the land and the
amount available for recycling will vary with the
N harvest index (NHI), which is the fraction of
total N removed in the harvested produce. To
achieve a positive soil N balance after legume
harvest, it is essential that the amount of N
derived from BNF exceed the amount of N
removed in the legume produce, i.e. the proportion of total N derived from BNF should exceed
NHI. Nitrogen removal through grain or forage
may deplete soil N. Of the 23 cases of grain
legumes compiled by Peoples and Craswell
(1992), N balances were positive in 17 cases and
negative or unchanged in six cases.
Examples of soil N gain or depletion by grain
legumes under varying N availability and yield
potentials are presented in Table 3. The data
demonstrate that even with a large N input from
BNF, the net soil N balance may be negative.
Despite a high soybean crop N value (404 kg N
ha-1) and high BNF input (290 kgN ha-1) the
net soil N balance was negative because NHI
exceeded the proportion of N derived from BNF
(Table 3, Chapman and Myers, 1987). In the
above example, a slight reduction in NHI or a
slight increase in N derived from BNF would
have resulted in a large soil N gain since the N
yield was large. Since NHI is an attribute less
amenable to field management, obtaining a posi-
tive soil N gain may primarily depend on
strategies to increase BNF.
As Peoples and Craswell (1992) have also
argued, even when there is little or no net soil N
gain, the depletion of soil N is still much less
after a grain legume than a nonlegume. Nevertheless, grain legumes increase the N output of
the cropping system that is economically usable.
Management approaches to limit the amount of
N removed in an economic produce to the
amount of N derived from BNF will at least
prevent depletion of native soil N. One approach
which meets the purposes of both grain N harvest and residue N production is the use of
dual-purpose food legumes (Buresh and De
Datta, 1991; Kulkarni and Pandey, 1988).
The effect of cutting and removing plant residues as fodder for farm animals or of allowing
farm animals to graze on legume cover-cropped
fallows has a similar effect as exporting N
through harvested grains (Myers and Wood,
1987). The amount of N returned to the land
depends on how the animal residues are managed. Nevertheless, grazing by animals is likely
to remove less N from land than cutting and
removing plant residues because part of the forage ingested by grazing animals is returned to
the land via urine and dung.
Stem-nodulating aquatic green manure legumes
Stem-nodulating aquatic green manure legumes
are fast-growing and can accumulate large
amounts of N in 45 to 60 days of growth (Ladha
et al., 1992). Data on the most frequently
studied stem-nodulating aquatic green manure
legume, S. rostrata, show that it derives larger
amounts of N from BNF than do root-nodulating
waterlogging-tolerant legumes, under either
flooded or non-flooded conditions and it accumulates as much or more N under flooded conditions as do root-nodulating legumes such as S.
cannabina and S. sesban grown under drained
conditions (Morris et al., 1989; Ndoye and
Dreyfus, 1988; Pareek et al., 1990). These features make stem-nodulating legumes uniquely
suitable for areas prone to flooding in the prerice season.
The reported tolerance of BNF by stemnodulating aquatic legumes to mineral N under
Table 3. Net soil nitrogen gains due to BNF following cropping of legumes at varying yield-potential and soil N availability levelsa
Legume
N status at planting
Total
N
(%)
Cowpea
Mungbean
Peanut
Soybean
0.12
0.04
0.16
0.16
0.16
0.16
0.04
0.12
0.16
0.16
0.30
0.13
0.13
Miner.
N
(µg g-1)
-b
81
81
3.5
3.5
Low
38
19
Total
crop
N
Fert.
N
(kg N ha-1)
30
0
25
25
100
0
20
0
20
30
0
20
0
33
0
0
% total
crop N in
produce
removed
(%)
84
125
153
100
98
172
60
62
241
250
404
90
186
174
246
222
347
267
b
% total
crop N
derived
from BNF
19
70
54
70
36
65
67
56
60
62
72
33
65
68
68
46
9
56
Net soil
N gain
due to
BNF
Reference
kg N ha-1
67
64
64
49
53
52
52
53
48
47
73
62
9
9
62
73
47
68
- 40
+ 7
- 16
+ 21
- 17
+ 23
+ 9
+ 2
+ 29
+ 38
-6
- 26
+ 104
+ 103
+ 14
- 60
- 132
- 32
Sisworo et al. (1990) c
Ofori et al. (1987)
Ofori et al. (1987)
Eaglesham et al. (1982) d
Eaglesham et al. (1982)
Chapman and Myers (1987)
Suwanarit et al. (1986)
Suwanarit et al. (1986)
Suwanarit et al. (1986)
Suwanarit et al. (1986)
Chapman and Myers (1987)
Sisworo et al. (1990)'
Suwanarit et al. (1986)
Suwanarit et al. (1986)
Thurlow and Hiltbold (1985) f
Zapata et al. (1987)
Bergersen et al. (1989) g
Bergersen et al. (1989) b
Values derived from available data when originally not given. Data not available. c Average for four crops. d Average for four cultivars. 'Average for two crops. 'Average for
'Lee' soybean at six sites. g Pre-fallowed soil; total N includes roots, nodules and fallen leaves. h Pre-cropped (oats) soil; total N includes roots, nodules and fallen leaves.
a
82
George et al.
flooded conditions (Becker et al., 1986, 1990;
Dreyfus and Dommergues, 1980; Eaglesham and
Szalay, 1983; Ladha et al., 1992) may suggest
that these legumes prefer N from BNF over soil
N. Whether this characteristic leads to soil NO,
accumulation in aerobic soil is yet to be determined. However, similar total N accumulation
and similar or decreased N2 fixation by S. rostrata in non-flooded compared to flooded soil
(Morris et al., 1989; Ndoye and Dreyfus, 1988)
may indicate a significant capacity for soil N
uptake in aerobic soil.
Legumes in the post-rice as opposed to the
pre-rice niche
Post-rice legumes with subsequent fallow periods
before wet-season rice, and pre-rice legumes
grown immediately before the wet-season rice,
may have varied effects on native soil N and on
N contribution to rice. It appears that soil NO3
accumulation in fallows following legumes could
be significant. Strong et al. (1986) presented data
on soil mineral N (NO3 and NH4) accumulation
during 6 months of fallow following several
legumes, oil seeds, and cereals. Following harvest, crop residues were left in the field, and the
soil was tilled to a depth of 8 cm. Soil mineral N
accumulation during the fallow period was generally higher following legumes than following
oilseeds or cereals. They attributed differences in
soil mineral N to differences in N content and
C:N ratios of the crop residues.
High quantities of mineral N following the
legumes are not advantageous to rice. Nitrogen
mineralization during the post-legume fallow,
and subsequent N loss upon soil flooding will
remove the mineral N from the system. There is
evidence that, regardless of post-rice treatments,
soil NO, may rapidly drop to uniformly low
levels by the start of the wet-season rice crop.
The higher soil NO, level following post-rice
legumes compared to weeds reported by Buresh
and De Datta (1991) in a Philippine soil persisted until the next sampling, 3 weeks later. However, NO, in all treatments declined to a low
level before the start of the wet-season rice
culture (Table 4, IRRI-NiffAL-IFDC collaborative research). Weed growth during the fallow
period after the legume harvest probably assimilated some of the soil NO,, but NO, loss was
likely to have been large in the treatments.
Results of Hamid et al. (1984) show no difference in rice yields when rice is grown six weeks
after legume harvest or after approximately 18
weeks of continuous fallow. However, when sorghum was the post-rice crop the rice yield was
severely depressed. In a study by John et al.
(1989), yield of rice immediately after cowpea
cropping with all aboveground residues removed
or after a weedy fallow were similar. The addition of pre-rice cowpea residues, however, did
increase rice yield.
Maximizing total nitrogen capture by legumes
in rice lowlands
Maximizing N gains through legumes in rice
lowlands entails approaches that maximize BNF
Table 4. Soil-nitrate accumulation during the fallow period following post-rice legumes and weedy fallow in a lowland rice-based
cropping system
Post-rice
treatment
Sampling time for nitrate
Legume
harvest a
(0-80 cm)
3 wk after
harvest
(0-60 cm)
6 wk after
harvest b
(0-60 cm)
47
52
SO
40
28
8
9
8
8
7
(kgNO3 - N ha-1)
Cowpea
Mungbean
Nodulating soybean
Nonnodulating soybean
Weedv fallow
a
b
Data from Buresh and De Datta (1991).
Start of wet-season rice culture.
53
53
54
47
32
Managing native and fixed N in rice-based cropping system
while minimizing soil N losses. Soil and crop
management should ensure establishment of an
effective symbiosis, and prevent excessive accumulation and subsequent loss of soil NO,. The
quantity of mineral N present at planting and
mineralized during the legume growing season
will influence the amount of N derived from
BNF.
Managing soil nitrate supply
Soil nitrate associated with the aerobic phase in
ricelands can be a source of N for both legume
and rice crop, if it can be managed effectively. A
lack of response of legumes to external N application indicates that the initial N needs are generally met by native soil NO,. Moreover, legume
N needs may be limited by a reduced yield
potential due to inadequate supplies of water
and nutrients other than N. Nitrate-N assimilated by the legume is conserved in the system
against possible loss due to subsequent soil flooding. The management objective then is to utilize
soil NO, that might otherwise be lost, and to
maximize BNF.
Allowing for increased nitrate uptake
The total amount of N that can be accumulated
by a legume crop in a given environment is fixed
by physiological and environmental determinants. Legume management aims to meet this
potential by maximizing the plant demand for N
and by optimizing the soil mineral N supply to
maximize aggregate plant N accumulation
through soil uptake and BNF.
Soil water deficits often limit legume growth
and plant N demand in lowland rice-based cropping systems. Establishing a post-rice legume
crop prior to or immediately after the rice harvest is crucial in maximizing the use of the entire
residual soil water reservoir. Water deficits are
likely during the late growth stages of a post-rice
legume because of a receding water table (Senthong and Pandey , 1989; Timsina, 1989).
In areas with limited soil-water availability,
drought-tolerant legumes such as pigeon pea and
indigo are suitable post-rice legumes that accumulate N throughout the dry season. Peanut is
reported to withstand relatively larger soil water
deficits than most grain legumes because of a
83
higher root density in the lower soil depths (De
Vries et al., 1989a; Senthong and Pandey, 1989).
Early rains facilitate establishment of pre-rice
legumes. However, intermittent waterlogging
during this season is deleterious to crop growth
and yield formation (Alam, 1989; Minchin et al.,
1978; Timsina, 1989), thus limiting N derived
from BNF. Alam (1989) observed that cowpea
and mungbean are unsuitable in poorly drained
soils where short-term flooding occurs during the
pre-rice dry-to-wet transition period. Aquatic
legumes can increase N input in these soils.
Green manure legumes such as S. rostrata and
Aeschynomene afraspera are adapted to flooded
soil conditions (Ladha et al., 1992). Aquatic
legumes can possibly conserve soil NO3 during
growth in nonflooded soil. Stem nodule BNF can
additionally provide N to the plant when the soil
is flooded.
Increased soil water availability not only enhances legume growth, but also promotes soil
NO3 production. Ideally, the previously accumulated NO3 should be depleted during early
legume growth, and NO3 supply thereafter
should be less than plant N demand. A reduced
soil NO3 supply combined with sufficient water
for legume growth can result in enhanced BNF
as the legume enters the reproductive phase.
Minimizing soil supply of nitrate
Management may play a role in influencing the
amount of NO3-N formed in aerobic soil and its
subsequent assimilation by plants or loss from
the soil-plant system. Planting legumes with notillage is a strategy to decrease soil NO3, since
tillage can increase soil NO3 levels. Peoples and
Craswell (1992) have presented data showing
improved N2 fixation by soybean under notillage. Therefore, relay seeding of legumes before harvest of the preceding rice crop might be
suitable for maximizing residual water use and
decreasing NO3 accumulation.
Rice stubble management can influence the
amount of NO3 produced in soil. Incorporating
the stubble rather than leaving it on the soil
surface or burning it following harvest, significantly reduces soil NO3 (Bacon, 1987, 1990).
Bacon (1987) reported reduced soil NO3 levels
for up to 12 weeks, and reduced N uptake by the
following wheat crop, when rice stubble
84
George et al.
(15 t ha -1 ) was incorporated soon after harvest.
Nitrogen depletion by crops preceding the
legume or competition for N in a legumenonlegume intercropping system, influences the
amount of NO3 available to the legume crop.
Soil NO3 is low immediately after a flooded rice
crop. Planting a legume immediately after the
rice crop might maximize BNF provided soil
water is adequate. Total N derived from BNF by
several legumes increased to 93-95% when
grown immediately after a dry-season rice crop,
compared to 65-72% when grown after a prolonged fallow (Chapman and Myers, 1987).
Legumes following cereals usually fix higher
amounts of N than legumes following a fallowed
soil (Bergersen et al., 1985, 1989) because of
decreased soil mineral N levels. Likewise, cereallegume intercropping can enhance N derived
from BNF by the legume (Fujita et al., 1992;
Rerkasem et al., 1988; Rerkasem and Rerkasem,
1988) due to mineral N depletion by the cereal.
Managing inefficient nitrate uptake by legumes
Legumes that are inefficient in mineral N uptake
might meet their N requirement through BNF
because of a complementary relationship between soil N uptake and BNF. This might cause
no substantial change in the amount of legume N
accumulated but might lead to larger net NO3
loss from lowland rice soils. While some reports
indicate inefficient NO3 uptake during the
legume reproductive phase (Imsande, 1986; Imsande and Edwards, 1988) others indicate large
N uptake from aerobic soil throughout the
legume reproductive phase (Afza et al., 1987;
George and Singleton, 1992). Herridge and
Bergersen (1988) described the "sparing" of soil
NO3 by N2-fixing legumes, where more NO3
remains in the soil after a legume compared to a
nonlegume. Nonetheless, excess NO3 may remain in a lowland soil after the growth of a
legume. This NO3 can be lost after soil flooding.
Larger amounts of residual soil mineral N
have been reported after cropping of legumes
than of other crops (Doughton and MacKenzie,
1984; McEwen et al., 1989; Stong et al., 1986).
Reports on post-legume soil NO3 levels in lowland rice soils are few. Buresh et al. (1989)
reported 25 kg NO3-N ha -1 in the top 60 cm soil
layer at harvest of mungbean grown after lowland rice. Table 4 shows identical soil NO3 levels
(52 kg NO3-N ha -1 ) in the top 80 cm layer at
harvest of soybean, cowpea, mungbean and a
non-nodulating soybean isoline in a lowland rice
soil, but a reduced NO3 level (32 kg NO3N ha -1 ) under a weedy fallow. The corresponding NO, levels in the top 20 cm soil layer were
30 kg NO3-N ha -1 for legumes and 12 kg NO3N ha -1 for a weedy fallow, indicating that the
weeds were more effective than legumes either
in depleting soil NO3 or in maintaining low NO3
levels by other mechanisms. Nitrogen fixation by
the legumes, however, had not influenced the
soil NO3 levels, as the non-nodulating soybean
and the N2-fixing legumes maintained similar soil
NO3 levels.
Weed N accumulation during the dry season in
lowland rice soils reported by Buresh and De
Datta (1991) ranged from 11 to 29 kg N ha -1 . It
is likely that most legumes that derived at least
50% of their N from BNF would have equal or
larger amounts of N removed from soil compared to weeds; there might be other effects of
weeds on soil NO3 than uptake. Soil NO3 levels
are usually lower in grass and hay crop fields
than in fields of row-planted crops such as maize
(Magdoff, 1982).
It follows that strategies to develop or select
legume-Rhizobium symbioses tolerant to soil
NO3 (Betts and Herridge, 1987; Carroll et al.,
1985; Hansen et al., 1989; Herridge and Betts,
1988) or legumes with reduced ability to use NO3
(Carroll and Gresshoff, 1986; Nelson et al.,
1983) might not be advantageous in lowland
rice-based cropping systems unless methods
other than plant uptake are found to prevent the
buildup and subsequent loss of NO3. Suitable
legume genotypes for lowland rice soils might be
those that can deplete soil NO3 while still deriving large amounts of N from BNF. Part of the
NO3 could possibly also be assimilated by weeds
either during a post-legume fallow period before
rice or during the legume crop.
Use of legume nitrogen for lowland rice
Use of legume green manures and grain legume
residues as a N source to rice has been well
Managing native and fixed N in rice-based cropping system
reviewed (Buresh and De Datta, 1991; Singh et
al., 1991). Soil incorporation of legume green
manures (Ladha et al., 1989b; Morris et al.,
1986a, 1989; Ventura et al., 1987) and of legume
harvest residues (Alam, 1989; John et al., 1989;
Norman et al., 1990) in lowland rice culture
generally results in increased N uptake and yield
of rice. However, the efficiency of residue N use
by rice might be exaggerated, because most reports attribute the positive influence exclusively
to a N effect.
Belowground residues
Data compiled by Buresh and De Datta (1991)
showed legume root N accumulation ranging
from 3 to 50 kg N ha-1 .They assessed the contribution from belowground legume residues as a N
source for rice to be negligible in most cases.
This would be especially true with grain legumes
grown to maturity because of the remobilization
of N from roots and nodules to the grain. The
belowground N contribution may be significant if
legumes are incorporated as green manure during the flowering and early reproductive stages
when root nodulation could be abundant.
Nodule biomass at flowering of several grain
legumes ranged approximately from 4 to
118 kg ha-1 (Table 5, Abaidoo et al., 1990;
George et al., 1987; Thies et al., 1991a). Assuming a maximum 7% N content, this nodule
biomass
contained
approximately
0.3
to
8 kg N ha-1. Adding the amount of N present in
Table 5. Nodule biomass at flowering of legumes in tropical
soils
Legume
Nodule biomass
(kg ha-1)
Common bean
4- 68
8- 65
38-105
13- 67
16- 75
24- 87
26-118
Cowpea
Lima bean
Soybean
Reference
Abaidoo et al., (1990) a
Thies et al., (1991a) b
Thies et al., (1991a)
Thies et al., (1991a)
Abaidoo et al., (1990) a
Thies et al., (1991a) b
George et al., (1987) c
a Across two sites of differing mean temperature and three N
application regimes.
b Across several sites of varying soil N status and yield
potential.
c Across three sites of varying soil N status and mean temperature.
85
the roots, total belowground N could become
significant in certain cases. However, the value
of this N to lowland rice is questionable because
N-rich nodule tissue would decompose and release N very rapidly. Morris et al. (1989) found
no difference in rice N-gain between legume
green manure grown in situ or that transported
from a nearby field, suggesting no N contribution
from belowground biomass. Nevertheless, quantitative data on belowground residue N are required to understand N cycling in rice-based
cropping systems.
A boveground residues
The amount of N that can be incorporated
through some legume green manure crops
(Ladha et al., 1992) is larger than the N needs of
rice. Buresh and De Datta (1991) compiled data
on amounts of N incorporated with aboveground
green manure and grain legume residues in lowland rice-based cropping systems. Green manure
N ranged from 21 to 267 kg N ha-1; N in aboveground residues of grain legumes from 17 to
101 kg N ha-1. Nitrogen fertilizer substitution for
lowland rice by incorporated green manures
ranged from 24 to 137 kg N ha-1, and by grain
legume residues from 37 to 100 kg N ha-1.Similar data are presented by Singh et al. (1991).
Furoc and Morris (1989) suggested that the addition of N in S. rostrata green manure in excess of
100 kg ha-1 has limited value for increasing rice
yield. Moreover, large application of N-rich
legume residues has the potential to supply excessive N to rice, leading to crop lodging and
yield reduction (Alam, 1989).
Nitrogen release from legume residues
Many studies demonstrate a rapid release of N,
and an increase in soil NH4, followed by a
plateau or decline in NH4 levels after green
manure incorporation (Khind et al., 1985;
Nagarajah et al., 1989; Palm et al., 1988). The
release of NH4 from green manure is, therefore,
not matched by the N uptake by rice. Thus, NH4
released may be subject to immobilization or
loss. Morris et al. (1986b) reported that recovery
of N from green manure was 33%, similar to that
from N fertilizer. In contrast, Ventura et al.
86
George et al.
(1987) reported up to 75% recovery of S. rostrata green manure N by rice. The differences in
green manure N recoveries reported may be
partly due to differences in the amount applied,
in C:N ratio, and in timing and method of application. Recent data show that NH4-N loss from
green manure was lower than that with equivalent amounts of fertilizer N (Biswas, 1988). Rice
recovery of green manure N is larger than from
urea, and green manure increases the recovery
of urea-N (Diekmann, 1991). Supplementing
moderate amounts of basally applied green manure N with topdressed fertilizer N may maximize both green manure and fertilizer N recoveries
(Morris et al., 1986b).
Conclusions and research directions
The large amounts of NO3-N, which accumulate
in aerobic soil during the non-flooded season,
can be lost for lowland rice upon flooding. Prevention of this loss might make a significant N
contribution to crops in the dry-to-wet transition
period or to a subsequent rice crop. In addition,
the inclusion of legumes for grain, forage, and
green manure production might increase the
total N output from the cropping system and
further add to soil N through BNF. The N
derived from BNF by legumes can be increased
by maximizing plant N demand and by minimizing soil NO3 supply. A post-rice nonlegume crop
followed by a pre-rice green manure legume or
intercropping legumes with nonlegumes during
both the post-rice and pre-rice seasons might
maximize the conservation and effective use of
both soil NO3 and atmospheric N in lowland
rice-based cropping systems. Legumes would be
ideal crops for soils low in native N, where other
upland crops may grow poorly without external
N supply. Where soils are left fallow during the
non-flooded season due to production constraints, increasing the legume component in the
weed flora may be suitable for capturing both
soil NO3 and atmospheric N.
Our knowledge of the processes and factors
regulating NO3 accumulation and loss during the
nonflooded season and the dry-to-wet season
transition period and how legume BNF fits into
this complex are still very limited. Research is
needed to understand the mechanisms of N gains
and losses and to identify and refine appropriate
soil and crop management practices to maximize
soil N and legume BNF use in lowland rice-based
cropping systems.
So far, research on N dynamics and balances
in legume-lowland rice sequences has been primarily with pre-rice legumes grown just prior to
wet season rice. Because dry-season legumes are
also an important post-rice crop in lowland ricebased cropping system, there is a need to ascertain the effects of post-rice legumes with subsequent fallow periods before wet-season rice on
native soil N dynamics and N contribution to
rice.
Maintaining soil N levels in the rice-based
farming system is not the dominant objective of
the farmer in the choice of cropping pattern and
management practices. Overall profitability shall
continue to dominate as the basic farmer objective. However, there is a major potential for
directing research toward identification of management practices aimed at maintaining longterm soil fertility while providing profitability in
the short term. This will depend upon a more
solid knowledge base of system-level N
dynamics.
References
Abaidoo R C, George T, Bohlool B B and Singleton P W
1990 Influence of elevation and applied nitrogen on rhizosphere colonization and competition for nodule occupancy
by different rhizobial strains on field-grown soybean and
common bean. Can. J. Microbiol. 36, 92-96.
Afza R G, Zapata H H and Danso S K A 1987 Effect of
delayed soil and foliar N fertilization on yield and N2fixation of soybean. Plant and Soil 97, 361-368.
Alam M 1989 Stability and comparative nitrogen contribution
of pre-rice legumes: Grain legume vs. green manure.
M.Sc. thesis. University of the Philippines, Los Baños,
Philippines.
Bacon P E 1987 Effect of nitrogen fertilization and rice
stubble management techniques on soil moisture content,
soil nitrogen status, and nitrogen uptake by wheat. Field
Crops Res. 17. 75-90.
Bacon P E 1990 Effects of stubble and N fertilization management on N availability and uptake under successive rice
( Oryza sativa L.) crops. Plant and Soil 121, 11-19.
Bacon P E, McGarity J W, Hoult E H and Alter D 1986 Soil
mineral nitrogen concentration within cycles of flood irrigation: Effect of rice stubble and fertilization management.
Soil Biol. Biochem. 18, 173-178.
Managing native and fixed N in rice-based cropping system
Becker M, Alazard D and Ottow J C G 1986 Mineral
nitrogen effect on nodulation and nitrogen fixation of the
stem-nodulating legume Aeschynomene afraspera. Z.
Pflanzenernaehr. Bodenkd. 149, 485-491.
Becker M, Ladha J K and Ottow J C G 1990 Growth and N2
fixation of two stem-nodulating legumes and their effect as
green manure on lowland rice. Soil Biol. Biochem. 22,
1109-1119.
Bennett J M and Albrecht S L 1984 Drought and flooding
effects on N fixation, water relations and diffusive resistance of soybean. Agron. J. 76, 735-740.
Bergersen F J. Turner G L, Chase D L, Gault R R and
Brockwell J 1985 The natural abundance of 15N in an
irrigated soybean crop and its use for the calculation of
nitrogen fixation. Aust. J. Agric. Res. 36. 411-423.
Bergersen F J, Brockwell J, Gault R R. Morthorpe L,
Peoples M B and Turner G L 1989 Effects of available soil
nitrogen and rates of inoculation on nitrogen fixation by
irrigated soybeans and evaluation of d 15N methods for
measurement. Aust. J. Agric. Res. 40. 763-780.
Betts J H and Herridge D F 1987 Isolation of soybean
capable of nodulation and nitrogen fixation under high
levels of nitrate supply. Crop Sci. 27. 1156-1161.
Birch H F 1960 Nitrification in soils after different periods of
dryness. Plant and Soil 12, 81-96.
Biswas T K 1988 Nitrogen dynamics and nitrogen-15 balance
in lowland rice as affected by green manure and urea
application. Ph.D. thesis, Indian Agric. Res. Inst.. New
Delhi, India.
Bohlool B B, Ladha J K, Garrity D P and George T 1992
Biological nitrogen fixation for sustainable agriculture: A
perspective. Plant and Soil 141. 1-11.
Boonkerd N and Weaver R W 1982 Survival of cowpea
rhizobia in soil by soil temperature and soil moisture.
Appl. Environ. Microbiol. 43, 585-589.
Bouldin D R 1986 The chemistry and biology of flooded soils
in relation to the nitrogen economy in rice fields. In
Nitrogen Economy of Flooded Rice Soils. Eds. S K De
Datta and W H Patrick Jr. pp 1-14. Martinus Nijhoff
Publishers, Dordrecht, The Netherlands.
Broadbent F E 1984 Plant use of soil nitrogen. In Nitrogen in
Crop Production. Ed. R D Hauck. pp 171-182. ASACSSA-SSSA, Madison, WI.
Buresh R J and De Datta S K 1991 Nitrogen dynamics and
management in rice-legume cropping systems. Adv.
Agron. 45, 1-59.
Buresh R J, Woodhead T, Shepherd K D, Flordelis E and
Cabangon R C 1989 Nitrate accumulation and loss in a
mungbean/lowland rice cropping system. Soil Sci. Soc.
Am. J. 53, 477-482.
Carangal V R, Tengco P L, Miah N I and Topar-Ngarm A
1988 Food and forage crop intercropping under rainfed
lowland conditions. In Proceedings of the Crop-Animal
Systems Research Workshop. pp 435-451. Serdang,
Malaysia.
Carroll B J and Gresshoff P M 1986 Isolation and initial
characterization of constitutive nitrate reductase-deficient
mutants NR328 and NR345 of soybean ( Glycine max ) .
Plant Physiol. 81. 572-576.
Carroll B J, McNeil D L and Gresshoff P M 1985 A
87
supernodulating and nitrate tolerant symbiotic (nts) soybean mutant. Plant Physiol. 78, 34-40.
Cassman K G. Whitney A S and Stockinger K R 1980 Root
growth and dry matter distribution of soybean as affected
by phosphorus stress, nodulation. and N source. Crop Sci.
20, 239-244.
Cassman K G, Whitney A S and Fox R L 1981 Phosphorus
requirements of soybean and cowpea as affected by mode
of N nutrition. Agron. J. 73, 17-22.
Chaney K 1990 Effect of nitrogen fertilizer rate on soil
nitrate nitrogen content after harvesting winter wheat. J.
Agric. Sci. Cambridge 114, 171-176.
Chapman A L and Muchow R C 1985 Nitrogen accumulated
and partitioned at maturity by grain legumes grown under
different water regimes in a semi-arid tropical environment. Field Crops Res. 11, 69-79.
Chapman A L and Myers R J K 1987 Nitrogen contributed
by grain legumes in rotation with rice on the Cununurra
soils of the Ord Irrigation Area, Western Australia. Aust.
J. Exp. Agric. 27. 155-163.
Craswell E T. Loneragan J F and Keerati-Kasikorn P 1987
Mineral constraints to food legume crop production in
Asia. In Food Legume Improvement for Asian Farming
Systems: Proceedings of an International Workshop. Khon
Kaen, Thailand. September 1986. ACIAR Proc. Ser. No.
18. Eds. E S Wallis and D E Byth. pp. 99-111.
Deibert E J, Bijcriego M and Olson R A 1979 Utllization of
15
N fertilizer by nodulating and nonnodulatmg soybean
isolines. Agron. J. 71. 717-722.
De Datta S K 1981 Principles and Practices of Rice Production. Wiley. New York. 618 p.
De Datta S K and Buresh R J 1989 Integrated nitrogen
management in irrigated rice. Adv. Soil Sci. 10, 143-169.
De Datta S K and Patrick W H 1986 Nitrogen Economy of
Flooded Rice Soils. Martinus Nijhoff Publishers, Dordrecht. The Netherlands. 186 p.
De Vries J D. Bennett J M. Albrecht S L and Boote K J
1989a Water relations. nitrogenase activity, and root development of three grain legumes in response to soil water
deficits. Field Crops Res. 21, 215-226.
De Vries J D. Bennett J M. Boote K J. Albrecht S L and
Maliro C E 1989b Nitrogen accumulation and partitioning
by three grain legumes in response to soil water deficits.
Field Crops Res. 22, 33-44.
Diekmann K H 1991 Nitrogen transformation processes involving 15N labeled Sesbania and Aeschynomene afraspera
green manure in lowland rice ( Oryza sativa L.). Ph. D.
Thesis. Justus-Liebig University. Giessen, Germany.
Dowdell R J, Crees R and Cannell R Q 1983 A field study of
contrasting methods of cultivation on soil nitrate content
during autumn, winter and spring. J. Soil Sci. 34. 367-379.
Dowling D N and Broughton W J 1986 Competition for
nodulation of legume. Annu. Rev. Microbiol. 40. 131-157.
Doughton J A and MacKenzie J 1984 Comparative effects of
black and green gram (mung beans) and grain sorghum on
soil mineral nitrogen and subsequent grain sorghum yields
on the Eastern Darling Downs. Aust. J. Exp. Agric.
Anim. Husb. 24. 244-249.
Dreyfus B and Dommergues Y 1980 Non-inhibition de la
fixation d'azote atmospherique par I'azote combine chez
88
George et al.
une legumineuse a nodules caulinaires, Sesbania rostrata.
C.R. Acad. Sc. Paris 291, 767-770.
Eaglesham A R J and Szalay A A 1983 Aerial stem nodules
on Aeschynomene pp. Plant Sci. Lett. 29, 265-272.
Eaglesham A R J, Ayanaba A, Rao V R and Eskew D L
1982 Mineral N effects on cowpea and soybean crops in a
Nigerian soil. II. Amounts of N fixed and accrual to the
soil. Plant and Soil 68, 183-192.
Eaglesham A R J, Hassouna S and Seegers R 1983 FertilizerN effects on N2 fixation by cowpea and soybean. Agron. J.
75, 61-66.
Eskew D L, Eaglesham A R J and App A 1981 Heterotrophic N2 fixation and distribution of newly fixed nitrogen
in a rice-flooded soil system. Plant Physiol. 68, 48-52.
Flinn J C and De Datta S K 1984 Trends in irrigated
rice-yields under intensive cropping at Philippine Research
Stations. Field Crops Res. 9, 1-15.
Fujita K and Ofosu-Budu Ogata S 1992 Biological nitrogen
fixation in mixed legume-cereal systems. Plant and Soil
141, 155-176.
Furoc R E and Morris R A 1989 Apparent recovery and
physiological efficiency of nitrogen in Sesbania incorporated before rice. Agron. J. 81, 797-802.
Garrity D P and Flinn J C 1988 Farm-level management
systems for green manure crops in Asian rice environments. In Sustainable Agriculture: Green Manure in Rice
Farming. pp 111-130. International Rice Research Institute, Los Baños, Philippines.
Garrity D P, Oldeman L R, Morris R A and Lenka D 1986
Rainfed lowland rice ecosystems: Characterization and
distribution. In Progress in Rainfed Lowland Rice. pp
3-23. International Rice Research Institute, Los Baiios,
Philippines.
George T and Singleton P W 1992 Relationship between
nitrogen assimilation traits and dinitrogen fixation in soybean and common bean. Agron. J. ( In press ).
George T, Bohlool B B and Singleton P W 1987 Bradyrhizobium japonicum - environment interactions: Nodulation and interstrain competition in soils along an elevational transect. Appl. Environ. Microbiol. 53, 1113-1117.
George T, Singleton P W and Bohlool B B 1988 Yield soil
nitrogen uptake and nitrogen fixation by soybeans from
four maturity groups grown at three elevations. Agron. J.
80, 563-567.
George T, Bartholomew D P and Singleton P W 1990 Effect
of temperature and maturity group on phenology of field
grown nodulating and nonnodulating soybean isolines.
Biotronics 19, 49-59.
Gibson A H and Harper J E 1985 Nitrate effect on nodulation of soybean by Bradyrhizobium japonicum. Crop Sci.
25, 497-501.
Hadas A, Feigenbaum S, Feigin A and Portnoy R 1986
Nitrogen mineralization in profiles of differently managed
soil types. Soil Sci. Soc. Am. J. 50, 314-319.
Ham G E 1980 Inoculation of legumes with Rhizobium in
competition with naturalized strains. In Nitrogen Fixation.
Vol. II. Eds. W E Newton and W H Orme-Johnson. pp
131-138. University Park Press, Baltimore, MD.
Hamid A, Paulsen G M and Zandstra H G 1984 Performance
of rice grown after upland crops and fallow in the humid
tropics. Trop. Agric. (Trinidad) 61, 305-310.
Hansen A P, Peoples M B, Gresshoff P M, Atkins C A, Pate
J S and Carroll B J 1989 Symbiotic performance of supernodulating soybean ( Glycine max (L.) Merrill) mutants
during development on different nitrogen regimes. J. Exp.
Bot. 40, 715-724.
Hardy R W F, Burns R C, Herbert R R, Hosten R D and
Jackson E K 1971 Biological nitrogen fixation: A key to
world protein. Plant and Soil Spec. Vol. 561-590.
Harwood R R and Price E C 1976 Multiple cropping in
tropical Asia. In Multiple Cropping. pp 11-40. ASA special publication number 27, American Society of Agronomy, Crop Science Society of America, Soil Science
Society of America, Madison, WI.
Herlihy M 1979 Nitrogen mineralization in soils of varying
texture, moisture and organic matter. I. Potential and
experimental values in fallow soils. Plant and Soil 53,
255-267.
Herridge D F 1986 Nitrogen fixation dynamics by rainfed
grain legume crops: Potential for improvement. In Transactions of the XIII Congress of International Society of
Soil Science. pp 794-804. Hamburg, Germany.
Herridge D F, Roughly R J and Brockwell J 1984 Effects of
rhizobia and soil nitrate on the establishment and functioning of the soybean symbiosis in the field. Aust. J. Agric.
Res. 35, 149-161.
Herridge D F and Bergersen F J 1988 Symbiotic nitrogen
fixation. In Advances in Nitrogen Cycling in Agricultural
Ecosystems. Ed. J R Wilson. pp 46-65. CAB International, Wallingford, UK.
Herridge D F and Betts J H 1988 Field evaluation of soybean
genotypes selected for enhanced capacity to nodulate and
fix nitrogen in the presence of nitrate. Plant and Soil 110,
129-135.
Herridge D F and Brockwell J 1988 Contribution of fixed
nitrogen and soil nitrate to the nitrogen economy of irrigated soybean. Soil Biol Biochem. 20, 711-717.
Hodges T and French V 1985 Soyphen: Soybean growth
stages modeled from temperature, daylength, and water
availability. Agron. J. 77, 500-505.
Huke R E 1982 Rice area by type of culture: South, Southeast, and East Asia. International Rice Research Institute.
Los Baiios. Philippines. 32 p.
Imsande J 1986 Ineffective utilization of nitrate by soybean
during pod fill. Physiol. Plant. 68, 689-694.
Imsande J and Edwards D G 1988 Decreased rates of nitrate
uptake during pod fill by cowpea, green gram, and soybean. Agron. J. 80, 789-793.
Imsande J 1988 Interrelationship between plant development
stage. plant growth rate, nitrate utilization and nitrogen
fixation in hydroponically grown soybean. J. Exp. Bot. 39.
775-785.
Imsande J 1989 Rapid dinitrogen fixation during soybean pod
fill enhances net photosynthetic output and seed yield: A
new perspective. Agron. J. 81, 549-556.
Janssen G and Metzger W H 1928 Transformation of nitrogen in rice soil. J. Am. SOC. Agron. 20, 459-476.
John P S, Pandey R K, Buresh R J and Prasad R 1989
Lowland rice response to urea following three cowpea
cropping systems. Agron. J. 81, 853-857.
Keeney D 1986 Sources of nitrate to groundwater. CRC
Critical Reviews in Environ. Control 16, 257-304.
Managing native and fixed N in rice-based cropping system
Khind C S. Josan A S and Beri V 1985 Nitrogen release from
sesbania green manure and effect of time of application of
N fertilizer on lowland rice. Int. Rice Res. Newsl. 10.
26-27.
Knowles 1982 Denitrification. Microbiol. Rev. 46, 43-70.
Koyama T and App A 1979 Nitrogen balance in flooded rice
soils. In Nitrogen and Rice. pp 95-104. International Rice
Research Institute, Los Baños, Philippines.
Kucey R M N, Chaiwanakupt P, Arayangkool T, Snitwongse
P, Siripaibool C, Wadisirisuk P and Boonkerd N 1988
Nitrogen fixation (15N dilution) with soybeans under Thai
field conditions. II. Effect of herbicides and water application schedule. Plant and Soil 108, 87-02.
Kulkarni K R and Pandey R K 1988 Annual legumes for food
and as green manure in a rice-based cropping system. In
Sustainable Agriculture: Green Manuring in Rice Farming.
pp 289-299. International Rice Research Institute, Los
Baños, Philippines.
Ladha J K, Garcia M, Miyan S, Padre A T and Watanabe I
1989a Survival of Azorhizobium caulinodans in the soil and
rhizosphere of wetland rice under Sesbania rostrata -rice
rotation. Appl. Environ. Microbiol. 55, 454-460.
Ladha J K, Miyan S and Garcia M 1989b Sesbania rostrata as a green manure for lowland rice: Growth, N2
fixation, Azorhizobium sp. Inoculation, and effects on succeeding crop yields and N balance. Biol. Fert. Soils 7,
191-197.
Ladha J K, Pareek R P and Becker M 1992 Stem-nodulating
legume-Rhizobium symbiosis and its agronomic use in
lowland rice. Adv. Soil Sci. (In press).
Lawn R J and Williams J H 1987 Limits imposed by
climatological factors. In Food Legume Improvement for
Asian Farming Systems: Proceedings of an International
Workshop, Khon Kaen, Thailand, September 1986
ACIAR Proc. Ser. No. 18. Eds. E S Wallis and D E Byth.
pp. 83-98.
Legg J 0 and Meisinger J J 1982 Soil nitrogen budgets. In
Nitrogen in Agricultural Soils. Ed. F J Stevenson. Agron.
Monograph No. 22 pp 503-566. Am. Soc. Agron.
Madison, WI.
Letey J, Valoras N, Focht D D and Ryden J C 1981 Nitrous
oxide production and reduction during denitrification as
affected by redox potential. Soil Sci. Soc. Am. J. 45,
727-730.
Macdonald A J, Powlson D S, Poulton P R and Jenkinson
D S 1989 Unused fertiliser nitrogen in arable soils - Its
contribution to nitrate leaching. J. Sci. Food Agric. 46,
407-419.
Magdoff F R 1982 Soil nitrogen availability under grassland
and cultivated corn. Plant and Soil 68, 395-398.
McEwen J, Darby R J, Hewitt M V and Yeoman D P 1989
Effects of field beans, fallow, lupins, oats, oilseed, rape,
peas, ryegrass, sunflowers and wheat on nitrogen residues
in the soil and on the growth of a subsequent wheat crop.
J. Agric. Sci. Camb. 115, 209-219.
Meade J, Higgins P and O'Gara F 1985 Studies on the
inoculation and competitiveness of a Rhizobium
leguminosarum strain in soils containing indigenous
rhizobia. Appl. Environ. Microbiol. 49, 899-903.
Minchin F R and Summerfield R J 1976 Symbiotic nitrogen
fixation and vegetative growth of cowpea ( Vigna un-
89
guiculata (L.) Walp.) in waterlogged conditions. Plant and
Soil 45, 113-127.
Minchin F R, Summerfield R J, Eaglesham A R J and
Stewart K A 1978 Effects of short-term waterlogging on
growth and yield of cowpea. J Agric. Sci. Camb. 90,
335-336.
Morris R A. Furoc R E and Dizon M A 1986a Rice
responses to a short duration green manure. I. Grain yield.
Agron. J. 78, 409-412.
Morris R A, Furoc R E and Dizon M A 1986b Rice
responses to a short duration green manure. II. N recovery
and utilization. Agron. J. 78, 413-416.
Morris R A. Furoc R E. Rajbhandari N K, Marqueses E P
and Dizon M A 1989 Rice response to waterlog-tolerant
green manures. Agron. J. 81, 803-809.
Mosier A R. Guenzi W D and Schweizer E E 1986 Soil losses
of dinitrogen and nitrous oxide from irrigated crops in
Northeastern Colorado. Soil Sci. Soc. Am. J. 50, 344-348.
Muchow R C 1985 Phenology, seed yield. and water use of
grain legumes grown under different soil water regimes in a
semi-arid tropical environment. Field Crops Res. 11, 8197.
Munns D N 1977 Mineral nutrition of the legume symbiosis.
In A Treatise on Dinitrogen Fixation. IV. Agronomy and
Ecology. Eds. R W F Hardy and A H Gibson. pp 353-391.
Wiley, New York.
Myers R J K and Wood I M 1987 Food legumes in the
nitrogen cycle of farming systems. In Food Legume Improvements for Asian Farming Systems: Proceedings of an
International Workshop, Khon Kaen, Thailand, September
1986. ACIAR Proc. Ser. No. 18. Eds. E S Wallis and D E
Byth. pp 46-52.
Nagdrajah S. Neue H U and Alberta M C R 1989 Effect of
Sesbania. Azolla and rice straw incorporation on the
kinetics of NH4, K. Fe, Mn, Zn and P in some flooded rice
soils. Plant and Soil 116, 37-48.
Ndoye I and Dreyfus B 1988 N2 fixation by Sesbania rostrata
and Sesbania sesban estimated using 15N and total N
difference methods. Soil Biol. Biochem. 20, 209-213.
Nelson R S. Ryan S A and Harper J E 1983 Soybean mutants
lacking constitutive nitrate reductase. I. Selection and
initial plant characterization. Plant Physiol. 72, 503-509.
Norman R J, Gilmour J T and Wells B R 1990 Mineralization
of nitrogen from nitrogen-15 labeled crop residues and
utilization by rice. Soil Sci. Soc. Am. J. 54, 1351-1356.
Ofori F. Pate J S and Stern W R 1987 Evaluation of N,
fixation and nitrogen economy of a maize/cowpea intercrop system using 15N dilution methods. Plant and Soil 102,
149-160.
O'Hara G W. Boonkerd N and Dilworth M J 1988 Mineral
constraints to nitrogen fixation. Plant and Soil 108, 93-110.
Palaniappan S P 1985 Cropping Systems in the Tropics.
Wiley Eastern Ltd., New Delhi, India. 215 p.
Palm 0, Weerakoon W L, De Silva M A P and Rosswall T
1988 Nitrogen mineralization of Sesbania sesban used as
green manure for lowland rice in Sri Lanka. Plant and Soil
108, 201-209.
Pareek R P. Ladha J K and Watanabe I 1990 Estimating N2
fixation by Sesbania rostrata and S. cannabina (syn. S.
acuvleata ) in lowland rice soil by the 15N dilution method,
Bid. Fertil. Soils 10, 77-88.
90
George et al.
Patrick Jr W H and Wyatt R 1964 Soil nitrogen loss as a
result of alternate submergence and drying. Soil Sci. Soc.
Am. Proc. 28, 647-653.
Peoples M B and Craswell E T 1992 Biological nitrogen
fixation: Investments, expectations and actual contributions to agriculture. Plant and Soil 141, 13-39.
Peoples M B and Herridge D F 1990 Nitrogen fixation by
legumes in tropical and subtropical agriculture. Adv.
Agron. 44, 156-223.
Ponnamperuma F N 1985 Chemical kinetics of wetland rice
soils relative to soil fertility. In Wetland Soils Characterization, Classification, and Utilization. pp 71-80. International Rice Research Institute, Los Baños, Philippines.
Pretty J N and Conway G R 1989 Agriculture as a global
polluter. Gatekeeper Series No. SA11, International Institure for Environment and Development. London, UK.
15 p.
Reddy K R and Patrick W H Jr 197.5 Effect of alternate
aerobic and anaerobic conditions on redox potential, organic matter decomposition and nitrogen loss in a flooded
soil. Soil Bid. Biochem. 7, 87-94.
Rerkasem K and Rerkasem B 1988 Yields and nitrogen
nutrition of intercropped maize and ricebean ( Vigna umbellata [Thunb.] Ohwi and Ohashi). Plant and Soil 108,
151-162.
Rerkasem B, Rerkasem K, Peoples M B, Herridge D F and
Bergersen F J 1988 Measurement of N2 fixation in maize
( Zea mays L.)-ricebean ( Vigna umbellata (Thunb.] Ohwi
and Ohashi) intercrops. Plant and Soil 108, 125-135.
Robson A D 1978 Mineral nutrients limiting nitrogen fixation
in legumes. In Mineral Nutrition of Legumes in Tropical
and Subtropical Soils. Eds. C S Andrew and E J Kamprath. pp 277-293. CSIRO, Melbourne, Australia.
Roger P A and Ladha J K 1992 Biological N2 fixation in
wetland rice fields: Estimation and contribution to nitrogen
balance. Plant and Soil 141, 41-55.
Sahrawat K L 1980 Soil and fertilizer nitrogen transformations under alternate flooding and drying moisture regimes.
Plant and Soil 55, 225-233.
Sahrawat K L 1983 Mineralization of soil organic nitrogen
under waterlogged conditions in relation to other properties of tropical rice soils, Aust. J. Soil. Res. 21, 133-138.
Santiago-Ventura T, Bravo M, Daez C. Ventura V, Watanabe
I and App A A 1986 Effects of N-fertilizers, straw, and dry
fallow on the nitrogen balance of a flooded soil planted
with rice. Plant and Soil 93, 405-411.
Savant N K, De Datta S K and Craswell E T 1982 Distribution patterns of ammonium nitrogen and N uptake by
rice after deep placement of urea supergranules in wet
land soil. Soil Sci. Soc. Am. J. 46, 567-573.
Seneviratne R and Wild A 1985 Effect of mild drying on the
mineralization of soil nitrogen. Plant and Soil 84, 175-179.
Senthong C and Pandey R K 1989 Response of five food
legume crops to an irrigation gradient imposed during
reproductive growth. Agron. J. 81, 680-686.
Sharma K N, Bhandari A L, Kapur M L and Rana D S 1985
Influence of growing various crops in five different fixed
rotations on the changes in nitrate and total nitrogen
content of soils. J Agric. Sci. Camb. 104, 609-613.
Sinclair T R, Muchow R C, Bennett J M and Hammond L C
1987 Relative sensitivity of nitrogen and biomass accumu-
lation to drought in field-grown soybean. Agron. J. 79,
986-991.
Singh Y, Khind C S and Singh B 1991 Efficient management
of leguminous green manures in wetland rice. Adv. Agron.
45, 135-189.
Singleton P W and Bohlool B B 1983 The effect of salinity on
the functional components of the soybean- Rhizobium
japonicum symbiosis. Crop Sci. 23, 815-818.
Singleton P W, AbdelMagid H M and Tavares J W 1985
Effect of phosphorus on the effectiveness of strains of
Rhizobium japonicum. Soil Sci. Soc. Am. J. 49, 613-616.
Singleton P W and Tavares J W 1986 Inoculation response of
legumes in relation to the number and effectiveness of
indigenous rhizobium populations. Appl. Environ. Microbiol. 51, 1013-1018.
Sisworo W H, Mitrosuhardjo M M, Rasjid H and Myers
R J K 1990 The relative roles of N fixation, fertilizer, crop
residues and soil in supplying N in multiple cropping
systems in a humid, tropical upland system. Plant and Soil
121, 73-82.
Smith D L, Dijak M and Hume D J 1988 The effect of water
deficit on nitrogen fixation (acetylene reduction) by white
bean and soybean. Can. J. Plant Sci. 68, 957-968.
Ssali H and Keya S O 1986 The effects of phosphorus and
nitrogen fertilizer level on nodulation, growth and dinitrogen fixation of three bean cultivars. Trop. Agric.
(Trinidad) 63, 105-109.
Streeter J 1988 Inhibition of legume nodule formation and N2
fixation by nitrate. CRC Critical Rev. Plant Sci. 7, 1-23.
Strickland R W 1969 Fluctuations in available mineral nitrogen in a flooded rice soil on the sub-coastal plains of the
Adelaide River, N.T. Aust. J. Exp. Agric. Anim. Husb. 9,
532-540.
Strong W M, Harbison J, Nielson R G H, Hall B D and Best
E K 1986 Nitrogen availability in a Darling Downs soil
following some cereal, oilseed and grain legume crops. 1.
Soil nitrogen accumulation. Aust. J. Exp. Agric. 26, 347351.
Suwanarit A, Suwanarit C and Chotechaungmanirat S 1986
Quantities of fixed N and effects of grain legumes on
following maize, and N and P status of soil as indicated by
isotopes. Plant and Soil 93, 249-258.
Thies J E, Singleton P W and Bohlool B B 1991a Influence of
the size of indigenous rhizobia1 populations on establishment and symbiotic performance of introduced rhizobia on
field-grown legumes. Appl. Environ. Microbiol. 57, 19-28.
Thies J E. Singleton P W and Bohlool B B 1991b Modeling
symbiotic performance of introduced rhizobia in the field
by use of indices of indigenous population size and nitrogen status of the soil. Appl. Environ. Microbiol. 57, 29-37.
Thurlow D L and Hiltbold A E 1985 Dinitrogen fixation by
soybeans in Alabama. Agron. J. 77, 432-436.
Timsina J 1989 Adaptation of diverse cowpea ( Vigna unguiculata (L.) Walp.) cultivars to excess moisture and
drought in riceland. Ph.D. Thesis, University of the Philippines, Los Baños, Philippines.
Torres R O, Magbanua R D and Garrity D P 1988 Evaluation of cowpeas as legume component in acid upland
rice-based cropping systems. Philipp. J. Crop Sci. 13,
91-98.
Ventura W and Watanabe I 1978 Dry season soil conditions
Managing native and fixed N in rice-based cropping system
and soil nitrogen availability to wet season wetland rice.
Soil Sci. Plant Nutr. 24, 535-545.
Ventura W, Mascarina G B, Furoc R E and Watanabe I 1987
Azolla and Sesbania as biofertilizers for lowland rice.
Philipp. J. Crop Sci. 12. 61-69.
Vincent J M 1980 Factors controlling the legume-Rhizobium
symbiosis. In Nitrogen Fixation. II. Eds. WE Newton and
W H Orme-Johnson. pp 103-120. University Park Press,
Baltimore, MD.
Weaver R W and Frederick L R 1974 Effect of inoculum rate
on competitive nodulation of Glycine max L. Merrill. II.
Field studies. Agron. J. 66, 233-236.
Weaver R W, Morris D R, Boonkerd N and Sij J 1987
Populations of Bradyrhizobium japonicum in fields cropped with soybean rice rotation. Soil Sci. Soc. Am. J. 51,
90-92.
White R E 1988 Leaching. In Advances in Nitrogen Cycling
in Agricultural Ecosystems. Ed. J R Wilson. pp 193-211.
C.A.B. International, UK.
91
Wallihan E F 1938 The effect of water logging on the rate of
nitrification in soils subsequent to drainage. Soil Sci. SOC.
Am. Proc. 2, 259-262.
Yamane 1 1957 Nitrate reduction and denitrification in
flooded soils. Soil Plant Food 3, 100-103.
Yoshida T and Yoneyama T 1980 Atmospheric dinitrogen
fixation in the flooded rice rhizosphere determined by the
N-15 isotope technique. Soil Sci. Plant Nutr. 26, 551-559.
Yost R and Evans D 1988 Green manure and legume covers
in the tropics. Hawaii Institute of Tropical Agriculture and
Human Resources Research Series 055. College of Tropical Agriculture and Human Resources, University of
Hawaii. 44 p.
Zapata F. Danso S K A and Fried M 1987 Time course of
nitrogen fixation in field grown soybean using nitrogen-15
methodology. Agron. J. 79, 172-176.
Plant and Soil 141: 93-118, 1992.
© 1992 Klsuwer Academic Publishers. Printed in the Netherlands.
PLSO SA06
Biological nitrogen fixation in non-leguminous field crops: Recent advances
IVAN R. KENNEDY and YAO-TSENG TCHAN
Department of Agricultural Chemistry and Department of Chemical Engineering, University of Sydney,
N.S.W. 2006, Australia
Key words:
nodules
Azospirillum, bacteria-plant association, non-legumes, para-nodules, Rhizobium, root
Table of contents
Abstract
Introduction
Past and future
Nitrogen fixation and substainable agriculture
Key features in achieving N2-fixing activity
The nitrogenase reaction
The energy requirement
Sensitivity to oxygen
Non-leguminous N2-fixing systems
Nitrogen fixation in non-leguminous symbiosis
Free-living N-fixing organisms
Azotobacter
Beijerinckia
Derxia
Clostridium
Photosynthetic N,-fixing bacteria and cyanobacteria
Associative diazotrophic systems
Azospirillum
Acetobacter
Pseudomonas
Alcaligenes
Klebsiella
Miscellaneous species of significance
Azorhizobium
Effectiveness of Azospirillum and other plant associations/symbioses
Associative organisms and wheat
Rice
Maize
Millet
Sugarcane
Sorghum
New nodulation technology for non-leguminous field crops
94
Kennedy and Tchan
Induction with hydrolytic enzymes
Nodulation of oilseed rape by Parasponia-infecting rhizobia
Non-genetical approaches to induced nodulation of non-legumes
Genetical approaches to induced nodulation of non-legumes
Recent studies on 2,4-D-induced para-nodules
Experiments with wheat plants
Inoculation with rhizobia
Inoculation with azospirilla
Effects of plant growth regulators and colchicine
The rate of N2 fixation in para-nodules, other induced and associative systems
Comparisons if legume nodules with para-nodules and other induced nodules on non-legumes
Para-nodules
Nodules on oilseed rape
General
Abolition of non-nodulating genetical character by 2,4-D
The future
Areas of inadequate knowledge
General techniques
Applications in agriculture
Acknowledgements
References
Abstract
There is strong evidence that non-leguminous field crops sometimes benefit from associations with
diazotrophs. Significantly, the potential benefit from N2 fixation is usually gained from spontaneous
associations that can rarely be managed as part of agricultural practice. Particularly for dryland systems,
these associations appear to be very unreliable as a means of raising the nitrogen status of plants.
However, recent technical advances involving the induction of nodular structures on the roots of cereal
crops, such as wheat and rice, offer the prospect that dependable symbioses with free-living diazotrophs, such as the azospirilla, or with rhizobia may eventually be achieved.
Introduction
Past and future
It is now well over 100 years since the existence
of microorganisms capable of biological fixation
of atmospheric N2 was experimentally proven
(Tchan, 1988). Barely 100 years ago the N2fixing capacity of the legume-rhizobial symbiosis
was firmly established. Since then, this symbiotic
system has become well understood and exTo avoid a too voluminous bibliography, where possible the
authors have used review papers to provide background
information. Earlier important papers were not overlooked,
but are covered in the relevant, more recent, reviews.
ploited as an effective means of raising the nitrogen status of soils, providing nitrogen for crops
and pastures (Vincent, 1984). Optimism that
non-leguminous crops could be similarly benefited was fueled in the 1970s and 1980s by the
discovery of several N2-fixing organisms (diazotrophs) forming apparently specific associations
with non-legumes. On the whole, however, this
optimism has not be rewarded.
Recently, several approaches using techniques
developed in the area of biotechnology have
raised new hopes that success in this secondary
objective may yet be realized. It is the authors'
opinion that there are now sound reasons to
anticipate that at least some non-leguminous
Induced nodulation
field crops may also become independent of soil
nitrogen. We intend to explain the reasons for
this renewed optimism, against the background
of knowledge accumulated in the past century
that will be relevant to any ultimate success in
exploiting these new approaches.
Nitrogen fixation and sustainable agriculture
A theme of this book is the relationship between
biological N2 fixation and sustainable agriculture.
Almost by definition, biological N2 fixation is
synonymous with sustainability. Systems capable
of fixing their own nitrogen exploit their own
environment less and may even provide a positive contribution.
Nitrogen applied in fertilizers usually provides
a benefit to plants. But, if applied inefficiently, it
can also have serious disadvantages in causing
pollution. It is difficult to match nitrogen supply
to actual requirements of a crop at a given
ecosite and any excess may damage this or other
ecosites. Excess reduced nitrogen (ammonium)
in agricultural or forest ecosystems may lead to
their acidification through the process of nitrification, if significant leaching of nitrate occurs
(Kennedy, 1986a; 1991). In agricultural ecosystems, however, biological N2 fixation may usually be expected not to exceed the actual nitrogen
requirements of an ecosystem, thus being less
likely than fertilizers to cause pollution. This
conclusion is based on the active repression of
the N2-fixing apparatus by a condition of ample
fixed nitrogen and the poorer competitive ability
of N2-fixing organisms when nitrogen supply is
adequate.
Key features in achieving N2-fixing activity
The nitrogenase reaction
Chemically, biological N2 fixation is essentially
the conversion of dinitrogen (N2) to ammonia,
catalyzed by the enzyme nitrogenase according
to the reaction (Fd = ferredoxin):
N2 + 8H + + 8Fd - + 16MgATP 2- + 18H2O
+
Ò 2NH4 + 2OH - + H2 + 8Fd + 16MgADP -
+ 16H2PO4
95
Of course, the process of N, fixation is much
more complex than this, but the biochemical
reaction contains the key ingredients and points
to the major requirements. It indicates the dual
needs for reducing potential and the substantial
energy requirement in the form of ATP. It also
suggests the need for a mechanism for ammonia
utilization that will simultaneously neutralize the
alkalinity generated (Kennedy, 1986a, 1991). In
order to fully appreciate the challenges of extension of the range of N,-fixing systems, we must
examine these requirements in greater detail.
The energy requirement
Estimates of the energy requirement (Schubert,
1982) have been based on theoretical as well as
many experimental results. However, the energy
requirement for N2 fixation is actually almost
identical to that required for nitrate assimilation
(Gibson, 1966; Kennedy, 1988), the other main
source of nitrogen for most field crops, excepting
rice.
A number of studies have attempted to estimate energy requirements for biological N2 fixation by comparing the rate of growth of legumes
growing on nitrate with their growth on dinitrogen. Measurement of dry matter accumulation has generally indicated little or no difference
in growth between the two nitrogen sources provided in ample quantity (except for the initial
period, when nodules are being established),
suggesting little difference in the need for photosynthate, in agreement with the theoretical predictions based on metabolic pathways.
The other main experimental approach has
been the use of efflux of carbon dioxide from
legumes growing with either nitrogen source
(e.g. Silsbury, 1977). In general, a greater carbon dioxide efflux has been observed on plants
growing with dinitrogen than with nitrate (see
Schubert, 1982), suggesting an increased energy
demand as extra carbohydrate required. Unfortunately, none of these studies considered the
fact that nitrate-grown plants produce bicarbonate rather than carbon dioxide in assimilating
nitrate (Kennedy, 1986b). When this difference
in the metabolism of the two nitrogen sources is
taken into account, the apparent difference in
energy consumption could disappear.
96
Kennedy and Tchan
Nevertheless, for diazotrophs with only indirect access to energy compounds, the criticism
regarding adequacy of energy supply is probably
valid. Calculations using models of the diffusive
flux of ‘energy’ compounds from the roots of
plants indicate that adequate growth of associative organisms in the rhizosphere with biological
N2 fixation significant to the plant is not possible
(Whipps and Lynch, 1983). Competition from
non-diazotrophs would make this benefit even
less likely.
On the other hand, if the diazotrophic organism in an engineered system could have direct
access to an ample energy supply from the plant,
such negative opinions about proposals to engineer cereals to fix dinitrogen based on energy
requirement could be questioned. Substantial infection of the root cortex by organisms like
Azospirillum would be a case in point. For such
pessimism to be sustained, one would need to
dismiss the positive role of leguminous nodules
in fixing N2, which is absurd. To be effective,
however, not only penetration but also adequate
colonization with the diazotrophs and some kind
of symbiosis would be required.
Sensitivity to oxygen
Another potentially negative factor in extending
the range of N2-fixing systems is the well-known
sensitivity to molecular oxygen. This sensitivity
results from the great instability of nitrogenase
when exposed to oxygen. Most diazotrophs regulate this sensitivity by using one of several
protective mechanisms (Kennedy, 1979; Postgate, 1974). These protective mechanisms include excessive respiratory consumption of oxygen, the cellular location of nitrogenase, the
presence of protective proteins associated with
nitrogenase, oxygen buffering as by leghaemoglobin in legume nodules and the less direct
means of hydrogenase, catalase and superoxide
dismutase activities (Clara and Knowles, 1985).
For example, without the protection of
leghaemoglobin, isolated Rhizobium bacteroids
will fix N2 only when external oxygen concentration is reduced to about 0.01 atm (Bergersen
and Turner, 1967). At the same time, excepting
obligately anaerobic clostridia or some facultative organisms, molecular oxygen is an absolute
requirement for significant N2 fixation, since it is
the terminal electron acceptor and it is needed to
allow formation of ATP by oxidative phosphorylation.
The actual pO2 at the root surface in soil has
rarely been determined. It might seem that, in a
well-aerated soil, the oxygen pressure would be
too high for sustained N2 fixation by many
diazotrophs. On the other hand, in a waterlogged soil, oxygen pressure would probably be
too low. The bulk concentration of oxygen in the
soil atmosphere is commonly about 10% of gas
content, half that of the free atmosphere and
oxygen pressure in soil obviously varies in the
range 0-0.20atm (Nye and Tinker, 1977). Root
growth has been observed to be retarded below
0.10 atm, although root metabolism is not seriously affected until oxygen pressure is reduced to
0.02-0.05 atm. Furthermore, an idealized model
suggests that, as a result of oxygen consumption
by microbes and healthy plant roots, a pO2 of
0.02 atm at the root surface 0.5 metres below the
surface is anticipated for a wheat crop (Kirkham,
1990) - one-tenth that of the free atmosphere,
even in an aerated soil at field moisture capacity.
Obviously, more experimental data on this point
is essential. (However, it can be noted now that
0.02 atm oxygen coincides with the value that we
use in our standard assay system for N2 fixation
by para-nodulated wheat, to be discussed at
length in the second half of this chapter).
Non-leguminous N2-fixing systems
Nitrogen fixation in non-leguminous symbioses
A number of non-leguminous plants are recognized to be well-endowed with N2-fixing capacity. Those capable of forming specialized symbioses with Frankia provide many such examples
(Millet et al., 1989; Quispel, 1974; RodriquezBarrueco et al., 1989). The symbioses between
cyanobacteria and various plants (Millbank,
1974) including Azolla (see Skinner et al, 1989)
are also significant N2-fixers. It is not intended to
discuss these highly developed symbioses in this
chapter, except rarely where comparisons are to
be made. These functional non-leguminous symbioses are well described elsewhere.
Induced nodulation
Free-living N 2-fixing organisms
The free-living diazotrophs were the first to be
recognized. It is now clearly understood that a
diverse range of microbial species are capable of
biological N2 fixation (Mulder and Brotonegoro,
1974; Postgate, 1972). At the very least, this
capacity allows these organisms to pioneer the
colonization of nitrogen-poor niches in ecosystems because of the selective advantage they
enjoy. It is also recognized that these organisms
provide a positive contribution to ecosystems in
the nitrogen they fix, perhaps indirectly when
they decompose.
As pointed out by Dobereiner (1989), the use
of selective media that simulate the natural environment of diazotrophs is still leading to the
discovery of previously undescribed N2-fixing
bacteria. This expands the genetic diversity available in the known library of diazotrophs and
should be helpful in new technological developments. In this paper, the possible value of these
organisms is viewed less from the effect of their
growth in situ, but for their potential to be
exploited in new technology.
Azotobacter
The hope of increased crop yields from soil
inoculation with N2 fixing bacteria can be traced
back almost a century to Caron in Germany.
Azotobacter was one of the organisms involved
in early research, with favourable results reported from Russia by Kostychev et al. (1926).
though the findings in Russia and elsewhere were
generally inconsistent and variable (Mishustin
and Shil’Nikova, 1971). The problem of inconsistency was not overcome by using well authenticated strains of inoculum. One of the problems
with Azotobacter is its poor performance in colonizing the rhizosphere with the exception of the
Azotobacter paspali - Paspalum notatum association (Dobereiner, 1974). Obviously, A. paspali
forms a very close and specialized association
with a cultivar of paspalum and can be considered as an exceptional associative diazotroph
although it does not penetrate the root as does
Azospirillum.
Much of the early work with Azotobacter
inoculation was carried out in Russia, Eastern
Europe and India. Only about a third of field
97
trials showed any positive effects, which ranged
up to 39%. Several processes other than N2
fixation could account for these positive effects,
including production of growth regulators, protection from root pathogens and modification of
nutrient uptake by the plant (Tchan, 1988).
A report that Azotobacter had been inserted
into a fungal mycorrhiza associated with a woody
plant (Pinus) (Giles and Whitehead, 1977) - thus
providing the first case of a biologically engineered diazotrophic symbiosis - is in serious
doubt. Apart from a controversy about a potential threat to pine plantations from associated
disease, the validity of the introduction has been
strongly disputed (Terzachi and Christensen,
1986).
Beijerinckia
In contrast to the acid-sensitive Azotobacter,
Beijerinckia is indifferent to soil pH. This bacterium is adapted to sub-tropical or tropical soils
(which are often acid). However, it is not regarded as a rhizosphere organism.
Derxia
Derxia, discussed by Postgate (1972), is unable
to commence N2 fixation and most colonies fail
to develop unless oxygen pressure is reduced
(Ravishankar et al., 1986). Apparently, Derxia
has a relatively inefficient respiratory protection
compared to Azotobacter.
Clostridium
The clostridial species such as Clostridium pasteurianum and Cl. butyricum are strict
anaerobes, and therefore fix N2 only in the condition of the absence of oxygen. Provided sufficient soluble energy substrates are available they
grow readily and are quite easy to isolate in
nitrogen-deficient media using soil as inoculum.
Photosynthetic N 2-fixing bacteria and
cyanobacteria
These organisms grow well in rice fields well
supplied with water and could contribute to the
supply of nitrogen to rice crops.
Associative diazotrophic systems
Attempts to discover beneficial associations be-
98
Kennedy and Tchan
tween plants and N2-fixing bacteria have a long
history. Obviously, a large proportion of the
total microbial activity in soil is likely to be
associated with the growth of plants, or their
subsequent breakdown, simply because this is
where growth substrates occur in greatest abundance. For N2-fixers, Dobereiner (1989) has stressed that too much may be assumed about the
degree of an associative relationship simply because bacteria are located in the rhizosphere
near the root surface. It may also be remembered that as long as the N2 fixation carried out
by such diazotrophs remains bound to the
growth of the bacterium, the possibility of a
significant effect on the plant remains slight.
Nevertheless, certain bacterial species (in particular the genus Azospirillum, are thought to
have a close association with plants, to a point
bordering on symbiosis. Dart (1986) has prepared a comprehensive review of N2-fixing organisms associated with nonlegumes (other than
rice) in agriculture.
Azospirillum
The genus Azospirillum is extensively discussed
in the book edited by Skinner et al. (1989). This
organism is known to penetrate the root of
gramineous plant species and to grow intercellularly to a degree, as well as growing in the
rhizosphere (Reinhold and Hurek, 1989).
Physiological aspects of the growth and N2
fixation of various Azospirillum species and related bacteria were discussed by Hartmann
(1989). The nitrogenase activity of Azospirillum
is strongly influenced by both ammonium (Gallani and Bazzicapulo, 1985; Hartmann et al.,
1986) and oxygen. When ammonium is added, or
anaerobic conditions are imposed, the nitrogenase of A. brasilense and A. lipoferum is
'switched-off' by ADP-ribosylation (Hartmann et
al., 1985, 1986; Hartmann and Burris 1987;
Kanemoto and Ludden, 1984) of nitrogenase
reductase (iron-sulphur protein). High oxygen
levels, inhibitory to nitrogenase activity, do not
cause this covalent modification to occur (Hartmann and Burris, 1987).
The release of nitrogen compounds by Azospirillum has been examined on only a few occasions. Release of nitrogen compounds in batch
cultures has been observed for A. lipoferum
(Volpon et al., 1981), A. amazonense and A.
halopraeferens (Hurek et al., 1987). Energystarved cultures of Azospirillum (Hartmann et
al., 1985) and at low levels of glutamate or
aspartate (Hartmann et al., 1986) release low
amounts of ammonium. Glutamine auxotrophic
mutants of A. brasilense, defective in glutamine
synthetase (Gauthier and Elmerich, 1977) or
glutamate synthase (Bani et al., 1980) release
ammonium under some conditions (Hartmann,
1989). Of course, the rapid and prolonged release of ammonia is a prominent feature of
Rhizobium bacteroids in legume nodules, but the
potential for efficient release of nitrogen compounds to the host plant by Azospirillum is
poorly understood at this stage. No clear evidence that azospirilla actively excrete newlyfixed nitrogen compounds into plant tissues is
available. If it is true that N2 fixation by azospirilla is bound to growing cells of the microorganism in plant associations, this would be
a possible reason for poor performance of the
system.
The oxygen sensitivity of the nitrogenase activity of Azospirillum has been discussed in detail (see Fig. l). In the figure redrawn from
Hartmann (1989), the complete abolition of nitrogenase at 1.0 kPa oxygen concentration is
shown, with a strain of A. amazonense being
more tolerant of oxygen.
The natural occurrence, distribution and survival of Azospirillum spp as a crop inoculant in
Belgian agricultural soils was discussed by De
Coninck et al. (1989). They point out that the
three modes of entry by N2-fixing bacteria into
plant roots are through wounds, especially at the
junction with lateral roots, through root hairs
and between undamaged epidermal cells;
rhizobia are known to employ all three methods,
and Frankia the second two. But the mode of
entry of Azospirillum remains unknown.
Acetobacter
Acetobacter
diazotrophicus,
an
associative
diazotroph, provides an outstanding example of
the bacterial ecological requirement. This organism needs high sugar concentration in the
medium to grow (Dobereiner, 1989), which need
is apparently satisfied in association with sugarcane (Li and MacRae, 1991).
Induced nodulation
99
prominent N2-fixing inhabitant of the root system of sugarcane (Li and MacRae, 1991).
Acetylene
reduction
rate (%)
Miscellaneous species of significance
pO2 (kPa)
Fig. 1. Oxygen tolerance of nitrogen fixation (C2H2 reduction) in azospirilla. Redrawn from Hartmann (1989).
Pseudomonas
Wetland rice may have large populations of N2fixing pseudomonas diazotrophicus, sometimes
amounting to about 80% of the total bacterial
population associated with rice roots. The ability
of this organism to grow autotrophically with H2
and CO2 may be of great significance under
flooded rice conditions where H2 is often evolved
(Barraquio et al., 1983; Watanabe et al., 1987a).
Pseudomonads have also been observed in the
root systems of sugarcane (Li and MacRae,
1991).
Alcaligenes
The diazotrophic Alcaligenes faecalis is capable
of entering rice roots, as shown by You and
Zhou (1989).
Klebsiella
Klebsiella pneumoniae has been identified as a
Azorhizobium
The
diazotrophic
organism,
Azorhizobium
caulinodans, capable of free-living N2 fixation, is
the microsymbiont in the stem nodules of Sesbania (discussed by Ladha et al., 1990). Stem
nodules are regarded as an adaptation to flooding to achieve adequate oxygenation. The rates
of nitrogen fixation achieved in such nodules can
be as high as 17 kg N per ha per day. The
symbiosis between A. caulinodans and S. rostrata
is highly specific and the microsymbiont is unable
to produce stem or root nodules in any of the
stem-nodulating Aeschynomene species. A.
caulinodans is also exceptional in being able to
grow on N2 under free-living conditions without
added fixed nitrogen (Dreyfus et ai., 1983). Recently, Ladha et al (1989) reported the presence
of A. caulinodans in the rhizosphere of rice.
Further, rice seedlings when inoculated with this
organism showed significant nitrogenase activity.
There are also recent reports (Ladha et al.,
1990) that this organism can contain bacteriochlorophyll and photosynthetic reaction centres.
If the organism can carry out photosynthesis
growing at up to 0.03 atm O2, this is most exceptional for a photosynthetic bacterium and this
proposal requires careful examination. Such a
capability would be of obvious benefit to a system biologically fixing nitrogen.
Effectiveness of Azospirillum and other plant
associations/symbioses
Since the excitement of the initial findings about
Azospirillum by Dobereiner and her colleagues
(Dobereiner, 1983), expectations regarding the
practical application of this organism in agriculture have been largely unfulfilled. No significant
agricultural systems have been established that
involve extensive inoculation by farmers of field
crops with organisms such as Azospirillum, a fact
that highlights the lack of its successful application. Despite this, a considerable amount of
100
Kennedy and Tchan
information exists with several field crops that
needs to be considered.
Associative organisms and wheat
Experimental work in countries other than Australia has shown that inoculation of wheat with
A. brasilense and A. lipoferum can increase significantly yields of foliage and grain (Hegazi and
Saleh, 1985; Kapulnik et al., 1983; Mertens and
Hess, 1984; Rai and Gaur, 1982). The cause of
such increases has not been determined, but
improved nitrogen nutrition from N2 fixation by
azospirilla cannot be excluded, although this has
been questioned (Jagnow, 1990). Positive results
from work in Israel (Millet and Feldman, 1984)
has even led to some field inoculations as an
agronomic practice (Okon et al., 1988), although
its success under farming conditions still needs
verification.
From a survey of wheat plants grown in the
field throughout the eastern Australian wheat
belt (New and Kennedy, 1989) a culture collection of more than 20 strains of Azospirillum
isolated from the root systems has been established. A pattern of distribution has emerged in
which only neutral soils contain a significant
population of these diazotrophs (New and Kennedy, 1989). Soils below pH 5.0 contained few if
any native azospirilla.
Although our greenhouse trials have failed to
show significant effects from inoculation of
wheat with azospirilla, apart from some stimulation of early vegetative yield, tests in laboratory
model systems have shown that particular strains
of azospirilla are more effective than others in
giving significant C2H2 reduction with particular
cultivars of wheat. The best of these was a A.
brasilense (Sp 107)/Sunelg, University of Sydney) combination (Kennedy et al., 1990), where
3-week-old wheat plants produced 48 nmoles of
ethylene per h per plant (30°C). Assuming that
this nitrogenase activity could be sustained and
would increase proportional to root weight, a
substantial contribution to the nitrogen needs of
the wheat plant would be made. More typically,
C2H2 reduction activity in these model associative systems lies between just above zero activity
in more than half our tests to about 10 nmoles
ethylene produced per hr per plant at this plant
age. No obvious reason for this variability with
wheat seedlings, documented extensively by Jagnow ( 1990), could be found. Furthermore, even
those plants with a high rate of C2H2 reduction
did not necessarily contain the greatest number
of viable azospirilla and the two parameters number of Azospirillum and rate of N2 fixationwere poorly correlated.
A major problem in achieving significant and
reliable N2 fixation in such associations could
result from the poor internal colonization of the
roots of wheat by the azospirilla. For example, a
factor of l04 fewer azospirilla ( A. brasilense Sp7)
were found in the root interior (endorhizosphere) compared to the root exterior (New et
al., 1991). Even if the endorhizosphere population is underestimated by a factor of 100 because
of the procedure employed to establish most
probable numbers of bacteria in this location, it
would appear that only a low proportion of the
azospirilla associated with wheat roots has direct
access to energy substrates. Whether the occasional seedlings that display strong C2H2 reduction have more internal colonization, or show
strong excretion of energy substrates or are affected otherwise, is not known.
In view of the difficulty in satisfying all the
demands for a satisfactory symbiosis, it is not
surprising that positive effects are inconsistently
observed. One must question whether sufficient
numbers of azospirilla with sufficient carbon substrates are present in most cases to make a
significant contribution to the nitrogen needs of
the plant.
Rice
Published evidence on experiments with rice
present an overall picture that is, on the whole,
more positive for associative N2 fixation than for
wheat.
Several reports of N2 fixation associated with
the rhizosphere of rice have been made. These
are summarised by Yoo et al. (1986), who indicate a number of associated organisms (Azospirillum,
Clostridium,
Enterobacter,
Alcaligenes, Pseudomonas) in a report in which
they described the contribution of Klebsiella oxytoca to N2 fixation benefiting rice plants. A 6%
increase in plant + soil nitro en content and a
significant incorporation of 15N2 was observed.
Up to 30-40 kg nitrogen ha-1 crop-1 are attribu-
Induced nodulation
ted to association with unidentified diazotrophs
(App et al., 1984). Fujii et al. (1987) have
described the benefits of use of strains of Klebsiella oxytoca and Enterobacter cloacae, able to
fix N2 in the presence of ammonium, for inoculation of rice.
A substantial number of studies have been
conducted at the International Rice Research
Institute in the Philippines (Ladha et al., 1986;
1987; 1989; Roger and Ladha, 1990) using rice.
On the whole, these results suggest that significant nutrition of rice supplying perhaps 20-25%
of the total needs of rice from associative fixation
is occurring (App et al., 1980; Watanabe et al.,
1987b; Zhu, 1989). Critical evidence of the likelihood of actual yield increases by inoculation of
rice is lacking.
Maize
Diem and Dommergues (1979) reviewed results
from some field trials using inoculation with
Azospirillum in Wisconsin, where the soils were
practically devoid of the organism before treatment. Of four experiments with maize only two
showed significant positive results.
An evaluation of the effect of oxygen pressure
and available carbon substrates (Alexander and
Zuberer, 1989) on C2H2 reduction by maize
showed that both these factors were significant.
C2H2 reduction activity was greatest with a pO2
between 1.3-2.1 kPa and amendment with several organic substrates greatly increased activity if
it was initially low. Inorganic nitrogen (4-20mg
L-1) abolished C2H2 reduction within about 2 h.
Rates of activity with individual plants varied
greatly but the maximum rates observed were of
the order of 300-800nmoles C2H4 per h per g
dry weight of roots. There was no effect of
inoculation with azospirilla on dry weight of
plants grown in soil in a series of 15 experiments,
and C2H2 reduction was depressed compared to
uninoculated controls with an indigenous population of bacteria. Thus, either the indigenous
bacteria are more effective or inoculation is
counterproductive.
Millet
Smith et a1. (1976) found that pearl millet
showed a significant increase in dry matter, associated with increased harvest of nitrogen per unit
101
area, but subsequent work did not prove that
these gains occurred from N2 fixation by the
plant (Smith et al., 1978, 1984). Instead, as Tien
et al. (1979) and Harari et al. (1989) have
indicated, growth substances produced by Azospirillum may be the source of these gains by
plants.
Occasional trials conducted at ICRISAT with
pearl millet have produced increased grain yield
up to 33% over uninoculated controls. More
typically, no significant increase was observed
(Wani et al., 1989) from inoculation with A.
lipoferum. A similarly variable result was obtained from inoculation with Azotobacter
chroococcum. Although an increased nitrogen
uptake by plants was measured in this series of
experiments, equivalent to about 3-19 kg per ha
over three seasons, no evidence was obtained
that this was the result of N2 fixation.
Sugarcane
A substantial input of nitrogen from associative
fixation is indicated in trials using 15N (Urquiaga
et al., 1989), with from 20-55% of the plant
nitrogen indicated from this source. The range of
organisms involved under Australian conditions
includes pseudomonads, enterobacters, Acetobacter, Beijerinckia and Klebsiella pneurnoniae,
but did not include Azospirillum or Azotobacter (Li and MacRae, 1991).
Sorghum
Recent studies on the effect of inoculation with
azospirilla (Pacovsky, 1989; Pereira et al., 1989)
do not provide convincing evidence that grain
yield can be increased from the association with
this diazotroph.
New nodulation technology for non-leguminous
field crops
The discovery of the first known rhizobial symbiosis with a non-legume (Parasponia-Rhizobium) by Trinick (1973, 1988) provided an important object lesson for scientists with interests
in biological N2 fixation. Sprent and de Faira
(1989) have recently emphasised that many of
the widely accepted dogmas for 'normal' symbioses, for example root hair infection and the
102
Kennedy and Tchan
necessity for bacteria to be released from infection threads before they differentiate into N2fixing forms, are not universal. They suggest that
a comparative study of a range of systems may
help exploit the possibilities for increasing the
range of N2-fixing systems. Infection through
wounds is now well established as a normal part
of nodule initiation in some legumes (Sprent and
de Faria, 1989), so that the idea of a fixed,
obligatory procedure cannot be sustained although such schemes provide helpful conceptual
models for study.
The recognition that Bradyrhizobium cells initially enter the roots of Parasponia at localized
sites of plant cell division having disrupted the
epidermis (Bender et al., 1987) also emphasises
that various developmental paths may lead to a
functional nodule. Only subsequently do infection threads develop from within the cortex and
participate in the nodule development. The absence of infection threads in the formation of
lupin nodules contrasts with their occurrence in
nodule formation in serradella roots, even
though Rhizobium lupini is the microsymbiont in
each case (Haack, 1961). Despite genetical
studies lasting more than ten years, no evidence
has been found that the legume-Rhizobium symbiosis involves unique plant gene products that
are necessarily restricted to the plant species that
so far successfully nodulate and fix N2. Neither
do any of the proposed steps in Rhizobiumlegume symbiosis (e.&. coded as hac (hair curling), noi (nodule initiation), etc.) infer a uniqueness that would absolutely exclude different
plant species from also allowing establishment of
persistent symbioses with diazotrophs.
The concept of establishment of artificially
induced symbioses between plants and so-called
L-forms of bacteria was experimentally explored
by Aloysius and Paton (1984). Protoplasts of
bacteria (L-forms) were suggested to have the
ability to penetrate the cell wall and membrane
structures of living plant cells and to colonize
plant tissue. L-forms of Azotobacter, Pseudomonas syringae, Bacillus polymyxa and Beijerinckia indica were all considered as capable of
penetration of plant tissue and some evidence of
expression of bacterial metabolic activity in plant
tissue was obtained. No tests for nitrogenase
activity were performed. However, approaches
such as this offer possible means of allowing
non-legumes such as cereals to fix their own N2.
Further investigations are needed.
Here, we will review recent developments suggesting that the occurrence of N2-fixing symbioses may eventually comprise a much more
diverse range of systems than has previously
been imagined.
Induction with hydrolytic enzymes
Experiments performed at Nottingham in the
U.K. (Cocking et al., 1990) have developed new
associations between Rhizobium spp. and a
range of non-leguminous field crops. One of the
areas of investigation has been the removal of
barriers in the plant to entry by the bacterium,
using enzymatic degradation of the plant cell
walls and of the root hairs. Some progress has
been made with a range of legumes and nonlegumes, including cereals. Their technique involves immersion of rapidly growing roots in cell
wall-degrading enzymes in hormone-free culture
medium with sucrose. Rapid degradation of the
plant cell wall at the apices of root hairs has been
observed. The use of a cellulase-pectolyase enzyme solution has been extended from white
clover seedlings (where the specificity of rhizobial infection was broken down) to rice. Up to
ten per cent of rice seedlings developed nodular
structures after treatment with enzymes and
inoculation with various rhizobia, both singly
and in combination. Omission of the enzyme
treatment resulted in no such structures being
formed.
Al-Mallah et al. (1989) investigated the formation of nodular structures by rhizobia on rice
seedlings. They suggested that nodulation of
non-legumes may be promoted by enzymatic cell
wall degradation coupled with polyethylene
glycol treatment. This apparently assists the
entry of rhizobia, though nitrogenase activity in
the resulting nodules was barely detectable. Bacteria in the structures were mostly located between cell layers, with some infection of cells
that usually appear to lack well-defined cytoplasm.
Trials with some varieties of wheat ( Triticum
aestivum ) also produced nodules on the roots,
following enzyme treatment and inoculation with
Induced nodulation
Rhizobium in the presence of polyethylene
glycol (PEG). Signs of N2 fixation were barely
detectable, possibly due to the limited extent of
infection of host cells.
Some varieties of Rhizobium induce nodules
on the roots of oilseed rape, inoculated subsequent to an enzyme treatment of the seedlings.
A synergistic interaction between R. loti and
Bradyrhizobium produced the highest rate of
nodulation on all plant varieties tested. Invasion
occurred both within and between the cells of
the nodules. Nodules which developed on
oilseed rape reduced C2H2 per unit fresh weight
very feebly, at less than 0.1% of that seen typically in legume nodules.
Nodulation of oilseed rape by Parasponia-infecting rhizobia
Bradyrhizobium parasponium is capable of infecting the roots of oilseed rape ( Brassica napus )
without need of enzyme treatment (Cocking et
al., 1990). A variety of nodule types have been
observed after infection with B. parasponium;
some contain red/pink pigmentation, suggesting
but not proving the production of leghaemoglobin. Rhizobia are present both within and between the cells of the host tissue, and some were
also associated with the surface of the nodules.
Non-genetical approaches to induced nodulation
of non-legumes
From the early 1980s, Nie and his colleagues at
Shandong University in China have published a
series of papers in Chinese journals, reporting
the nodule-inducing effect of 2,4-dichlorophenoxyacetate (2,4-D) on the roots of a large
number of plant species, including wheat (see
Tchan and Kennedy, 1989, for review). This
approach resulted from an initial observation by
Nie, made when using a plant tissue culture
medium containing 2,4-D.
Even earlier observations were made of the
development of such structures in both legumes
(Allen et al., 1953; Vincent, pers. commun.;
Wilde, 1951) and non-legumes (Nutman, 1944,
pers. commun.) on treating plants with hormone
herbicides. However, no inference was made in
earlier work that these structures might contain
103
bacteria or that the structures might be beneficial. To its credit, Nie’s group recognized this,
although it was also understood that the structures were primarily a response to 2,4-D treatment, since they were formed irrespective of
whether the roots were inoculated with a
Rhizobium species or not. The Chinese researchers have continued this work in the hope of
eventually emulating leguminous N2-fixing
nodules, but without obtaining acceptable evidence of positive N2 fixation (see Tchan and
Kennedy, 1989).
Other research centres in China have subsequently carried out work on inducing nodular
structures on non-legumes not normally nodulated. Chen and his colleagues in Beijing confirmed the results of Nie’s group. They have also
reported stimulation in the growth of wheat
when treated with 2,4-D, but this effect was not
dependent on inoculation with rhizobia. Chen
(personal communication, 1990) has also reported positive but low rates of N2 fixation in
2,4-D-induced nodulated roots of wheat using
Azorhizobium caulinodans as microsymbiont.
Jing et al. (1990a) also contributed in China to
this activity regarding an extended potential for
diazotrophs. Their findings, published in a European journal, suggest that magnetic effects are
capable in some unexplained way of inducing
nodular structures on barley roots. These structures apparently contain inoculated bacteria
( Rhizobium astragali ) and formation of the
‘pseudonodules’ was accompanied by the presence of infection threads and formation of bacteroids enclosed within a peribacteroid membrane. No C2H2-reducing activity was observed.
Unfortunately, it is impossible to assess their
published results critically, since no quantitative
data were given on the relative rates of nodulation in test plants versus control plants. The
paper does not explicitly state that the controls
were also negative. These experiments will need
to be supported by suitable quantitative data
before their relevance can be assessed.
Genetical approaches to induced nodulation of
non-legumes
One approach to forming nodules on plants not
normally nodulated is the purely genetical one.
104
Kennedy and Tchan
For example, a successful transfer of genes associated with root hair curling (hac) was made
from Rhizobium trifolii into a derivative
Rhizobium trifolii on the plasmids pKT230 and
pRK290 (Plazinski et al., 1985). The root haircurling character could then be expressed on
infection of rice seedlings. A similar transfer of
the genes from Rhizobium meliloti into Azospirillum brasilense on the plasmid pVKl00 was
successfully achieved. Subsequent inoculation of
maize seedlings also resulted in root hair curling
(Piana et al., 1988).
Rolfe and Bender (1990) recently described
further developments in their attempts to evolve
rhizobia for non-legume nodulation. A strain of
Rhizobium reconstructed from one capable of
nodulating Parasponia with a nodD allele inserted into its genome inoculated onto rice seedlings occasionally resulted in nodule-like structures on the roots (see Fig. 2). These contained
vascular bundles and membrane-encapsulated
bacteria, though there was no sign of an infection thread and no detectable N2 fixation.
In other work with rice, Jing et al. (1990b)
have described formation of root nodules using a
mutant of Rhizobium sesbania (Azorhizobium
caulinodans). The structure of these nodules and
their formation was claimed to be similar to
those of legume nodules such as soybean, including the presence of leghaemoglobin. A rate of
C2H2 reduction per g of nodule tissue about
1.5% of that shown by legume nodules was
observed, although it is not clear that adequate
controls were performed.
Recent studies on 2,4-D-induced para -nodules
Tchan and Kennedy (1989) reviewed work on
2,4-D induction from original Chinese sources
(supplied by Nie in late 1987) and reported their
own confirmation of the effect of 2,4-D in inducing nodular structures on wheat. Deliberately
seeking an alternative term for these structures
that emphasises their ‘apart-ness’ from legume
nodules, Tchan and Kennedy have named them
para-nodules (Kennedy et al., 1990; e.g. para =
beyond). Since our review and confirmation of
the results obtained by Nie in China from 1980
(Tchan and Kennedy, 1989), we have consistently regarded it as counterproductive to consider
para -nodules simply as models of legume
nodules. We have the firm opinion that progress
is more likely if the dogma of the process of
legume nodulation is not used in their future
development, although the reasons for success of
legume nodules obviously need to be considered.
Bender et al. (1990) recently also examined
the effect of 2,4-D on wheat seedlings. Within a
narrow concentration range, some lateral roots
became modified to form structures resembling
the nodules formed on legumes, while other
lateral roots and top growth were normal. Plants
exposed to higher concentrations (5 x 10-6 M
2,4-D and above) had few or no lateral roots,
while the top growth was yellow and reduced in
size. As previously observed (see also Tchan and
Kennedy, 1989) all inoculated wheat plants
formed nodules (unlike nodules on rice as referred to above), but these were not observed to
be highly organized internally. No evidence of
N2 fixation has been reported for these nodules,
using rhizobia as the infective bacteria.
Experiments with wheat plants
Inoculation with rhizobia
Under aseptical conditions modelled on an approach pioneered by Nie and his colleagues at
Shandong University, para-nodules are readily
formed on wheat (see Fig. 3) using 2,4-D (0.51 ppm in the nitrogen-free hydroponic medium);
these may occasionally be infected with inoculated rhizobia. In experiments conducted in Sydney (Kennedy et al., 1991), there is no clear-cut
evidence of intra-cellular infections with bacteroids surrounded by peribacteroid membranes
as in legume nodules (although some earlier
electron micrographs obtained by Nie in China
are suggestive of these). Significant infection by
rhizobia occurs in only about 10-20% of cases
and the para -nodules are disorganized internally
with most bacteria being located distant from
rudimentary central vascular tissue in extracellular pockets (Fig. 4) or in cells without normal
cytoplasmic contents. Similar results were obtained by Yu (1988).
Furthermore, no N2 fixation (C2H2 reduction)
is observed when rhizobia are used for inoculation (Kennedy et al., 1990). In one recent experiment using 15 diverse strains of rhizobia
105
Fig. 2. Small, spherical nodule-like structures formed on rice seedlings (Calrose) inoculated with ANU536. (a) and (b) light
micrograph sections showing infected cells and vascular bundles (vb); (c), (d) and (e) electron micrographs of infected plant cells
showing membrane encapsulated bacteria. Bar = 1 µm. Reproduced from Rolfe and Bender (1990). These micrographs were
kindly provided by Professor B. Rolfe.
106
Kennedy and Tchan
Fig. 3. Scanning electron micrograph showing a para-nodule of wheat, inoculated with Rhizobium leguminosarum. Bar =
100 pm.
representing the range of commercial strains
used in Australia (Roughley, 1988), provided by
the Australian Inoculants Research and Control
Service, no C2H2-reducing activity was observed
for each association. In these experiments, oxygen pressure was varied during a three-day exposure to C2H2 over the range 0.02 atm to 0.21 atm
and still no ethylene was formed.
Inoculation with azospirilla
Our philosophical viewpoint that para-nodules
should not be regarded as models of the legumeRhizobium symbiosis immediately led to a proposal to employ a free-living diazotroph such as
Azospirillum as a potential N2-fixing microsymbiont. This decision was also based on the wellknown fact that azospirilla are associative diazo-
Induced nodulation
107
Fig. 4. Electron micrograph of wheat pura-nodule infected with Rhizobium astragali. The cell wall (cw) and rhizobia (Rh) are
shown, without evidence of infection of plant cell cytoplasm ( PC ). A, Bur = 1 µm; B, C, Bar = 0.5 µm.
trophs of grasses, such as wheat. Encouraging
results have been obtained since adopting this
viewpoint, in testing the hypothesis that these
para-nodules based on Azospirillum as the infec-
tive agent provide a new model of a N2-fixing
symbiosis in non-legumes (Zeman et al., 1991).
In contrast to experiments with rhizobia, we
now consistently observe substantial rates of
108
Kennedy and Tchan
ethylene production when para - nodulated wheat
seedlings have been inoculated with Azospirillum. The ethylene production under a range of
treatments is dependent on both inoculation with
azospirilla and on the presence of C2H2. There is
no spontaneous measurable ethylene production
by the plant seedlings treated with 2,4-D unless
both C2H2 and azospirilla are present.
Seedlings inoculated with Azospirillum, but
with no 2,4-D treatment, often reduce C2H2 as
well, but only in the range of 0-20% the rate of
the para-nodulated plants within individual experiments. Another characteristic of 2,4-D-influenced C2H2 reduction by azospirilla is a much
higher degree of reproducibility with replicated
plants compared to seedlings inoculated with
Azospirillum alone.
In order to be certain that the nitrogenase
activity is not that of free-living azospirilla, we
routinely flood the roots with Winogradsky’s
mineral nitrogen-free medium and shake the
flasks using an atmosphere of 0.02 atm O2 (Kennedy et al., 1990; Tchan et al., 1990). We have
verified (Tchan et al., 1991) that, under these
conditions, the nitrogenase activity of free-living
bacteria and of azospirilla added to sterile wheat
seedlings is abolished by oxygen inhibition at
0.02 atm oxygen, as would be predicted from the
known sensitivity of Azospirillum shown in Figure 1. Apparently, the agitation in Winogradsky’s medium is sufficient to disperse microbes
that develop in colonies on the root surface.
Consequently, if exposures are not too prolonged, the activity measured should be of azospirilla infecting the para-nodules or colonizing
the interior of root tissues. However, this situation is complex and more critical tests will be
needed before strong assertions about the exact
location of the 2,4-D-dependent activity can be
made.
The rate of C2H2 reduction by Azospirillum
brasilense Sp7 in associative and para - nodulated
wheat seedlings in the range 0-0.08 atm pO2 is
shown in Figure 5. Note that in the standard
assay system with Winogradsky’s medium added,
no C2H2 reduction occurs in the associative system with pO2 higher than 0.01 atm, a result
similar to that obtained when Azospirillum in
liquid culture is added to uninoculated wheat
seedlings at the commencement of the nitrogen-
Fig. 5. Effect of pO2 on the rate of C2H2 reduction by
associative azospirilla and in para-nodules. Seedlings with
and without 2,4-D treatment were inoculated with A.
brasilense Sp7 at one week of age and grown under continuous light for two more weeks. C2H2-reducing activity was
measured at 30°C, with Winogradsky’s N-free medium (WM)
sometimes immersing the roots.
ase assay. This suggests that nearly all active
azospirilla in the associative system without
added 2,4-D reside near the surface of the roots,
within easy reach of external oxygen.
Inspection by phase microscopy of individual
plants previously tested as positive for N2-fixing
activity suggests that the azospirilla colonize the
para-nodules and their vicinity strongly (Kennedy et al., 1991a). INT-stained root systems
from the most active plants invariably show large
numbers of stained bacteria, particularly near
the base of the para - nodules. While not direct
proof that these bacteria in para - nodules are
reducing C2H2, this qualitative association is suggestive of this.
Induced nodulation
109
The findings of dependence on inoculation
with azospirilla, the presence of C2H2, the inhibition by oxygen and a positive correlation between high rates of activity and the obvious
presence of azospirilla in the basal para-nodule
cells suggest strongly that the C2H4 production
observed results from nitrogenase activity. Furthermore, or preliminary experiments with 15N2
indicate significant enrichment of total Kjeldahl
nitrogen (0.040 atom per cent excess 15N) in
para-nodulated root systems. In view of these
results, it is unlikely that the ethylene production
observed is other than nitrogenase-related.
Effects of plant growth regulators and colchicine
We have tested the effects of indoleacetic acid
(IAA), naphthalene acetic acid (NAA) and colchicine on the rate of development and C2H2reducing activity of para-nodules (Kennedy et
al., 1991a). For para-nodules to play a significant
role in the nutrition of field crops such as wheat,
they will need to remain in function throughout
the vegetative growth phase of the plants or be
continuously initiated on the root system. Therefore, information on the effects of growth regulators or other substances on their rate of initiation and development should be of value.
Initially, the effect of colchicine was tested in
case, by increasing the frequency of polyploid
cells, it might increase the frequency of paranodulation, an argument based on the observation that the initiation of legume nodules involves production of polyploid cells (Vincent,
1980). It was found that no such increase
occurred but that a marked increase in the size
of para-nodules was obtained with a treatment
for 24 h of 0.005% colchicine (see Fig. 6). This
occurred below the polyploidy threshold of
0.007% for wheat seedlings (Darvey, 1972) and
the effect was abolished by more than 0.010%
colchicine. With Rhizobium meliloti as inoculant,
no C2H2 reduction was observed with these colchicine-affected para-nodules.
In contrast, with Azospirillum brasilense as
inoculant, strong C2H2 reduction is obtained
with 2,4-D-treated seedlings and in a limited
number of experiments a substantial increase in
the rate of C2H2 reduction with colchicine treatment as well has been observed, although this
result is rather erratic. Colchicine did not reli-
Fig. 6. Effect of colchicine treatment on the volume of
para-nodules of wheat infected with R. meliloti Wheat ( Triticum aestivum L.) seedlings were germinated on agar and
grown hydroponically until one week old before pre-treatment with colchicine for 24 h just prior to para-nodulation
(Kennedy et al., 1991). Nodule dimensions were determined
two weeks later using root systems stained with iodonitrotetrazolium (0.2% in water). using a micrometer eyepiece on an
Olympus dissecting microscope. Treatments: 1 =2,4-D;
2 = 2,4-D + R. meliloti; 3 = 2 + 0.005% colch.: 4 = 3 with
continuous colchicine.
ably increase the rate of C2H2 reduction of
wheat seedlings inoculated with azospirilla alone
without the formation of para-nodules (Kennedy
et al., 1991a). Colchicine has been reported to
affect the initiation of lateral root primordia
(Foard et al., 1965).
With NAA plus 2,4-D, a strong stimulation of
C2H2 reduction was obtained (Table 1) accompanied by some modification of the development
of the para-nodules. Perhaps significantly, wheat
seedlings without 2,4-D treatment but with IAA
110
Kennedy and Tchan
Table 1. C2H2 reduction in paru-nodules and associative systems
Para-nodules
nmoles h -l plant -1 (ca. 100 mg f.w. roots)
Azos.
Exp. 1 (n = 5)
Exp. 2 (n = 9)
Exp. 3 (n = 5)
a
c
0.4(0.4) b
0.0(0)
0.2(0.4)
2,4-D + Azos.
NAA + Azos.
NAA/2,4-D/ Azos
2.7(3.1)
7.9(19.1)
-
34.1(35.0)
7.0(5.4)
18.9(18.3)
79.2(86.8)
38.3(21.8)
-
x40 converts these values to per g dry weight roots (approx.)
Oil-seed rape nodules
Rice nodules
Para-nodulated seedlings
nmoles h-1 g-1 f.w. nodules
ref.
10
204
24,750
Cocking et al. (1990)
Jing et al. (1990b)
d
Mean value, data above
n = number of replicate seedlings.
The standard error is shown in parentheses.
c Azos = Azospirillum brasilense Sp7.
d not neccessarily restricted to pura-nodules.
a
b
(5 ppm) and NAA (1 ppm) treatment also sometimes reduced C2H2 much more rapidly than did
control plants without any hormone treatment
(Kennedy et al., 1991a). However, both NAA
and IAA also induced strong modification of
lateral roots, which were thickened and prolific
in root hairs. Inspection by microscopy indicated
that azospirilla colonized in a similar manner as
with 2,4-D-treated plants, at the basal region
close to the root stele of the main root. Although
these auxin-affected structures lack the wellrounded appearance of 2,4-D induced paranodules, this process may be akin to para -nodulation. Indeed, when the NAA concentration
was raised to 10 ppm, the well-rounded shape of
2,4-D-induced nodules was also obtained.
The rate of N2 fixation in para-nodules, other
induced nodules and associative systems
In Table 1 are shown the approximate C2H4forming activities in several of our experiments.
A control of wheat seedlings inoculated with
Azospirillum, but without 2,4-D added, is
routinely included. Under the conditions of
assay, azospirilla in the rhizosphere are considered as limited in their ability to express nitrogenase activity because of oxygen toxicity (see
Fig. 5).
It is difficult to compare these rates with those
of associative systems, since data for young seedlings are rarely provided. However, between
100-800 nmoles of C2H2 reduction per g dry
weight roots per hour have been reported for
systems not necessarily involving Azospirillum
(Alexander and Zuberer; Diem and Dommergues, 1979; Ladha et al. 1986). A pure culture of
Azospirillum brasilense Sp7 is capable of producing 240 nmoles C2H4 per h per mL of culture
containing about 108 cells in the log phase of
growth (Hartmann and Burris, 1987).
The apparent rate of C2H2 reduction in 2-3week-old para-nodules is already clearly significant. Although the variation in nitrogenase activity is high, with standard deviations of the
same order as the value of the mean, this is not
unexpected. Observations have shown that activities are qualitatively correlated with the extent of colonization of para -nodules - a process
that is likely to follow the Poisson distribution
rather than a normal distribution.
However, now that an effective model system
(viz. the para -nodule, Zeman et al., 1991) is
available for testing, further improvements in the
consistency of the rate of C2H2 reduction can be
anticipated. We are currently testing about 20
collected strains of azospirilla, isolated (New and
Kennedy, 1989) from the roots of wheat plants
growing in the eastern Australian wheat zone.
Comparisons of legume nodules with
para-nodules and other induced nodules on
non-legumes
The concept of coupling mechanisms has been
advanced to allow a more systematic analysis of
Induced nodulation
the ecological success of different entitites in
ecosystems (Kennedy 1983, 1986a; Tchan and
Kennedy, 1988). A coupling mechanism is a
physical, chemical or biological apparatus
catalysing a thermodynamic work process (an
action). Such an action process (e.g. the growth
to maturity of a legume such as soybean) occurs
with greatest efficiency when factors affecting the
coupling mechanism(s) are optimised. In the case
of N2 fixation by legumes, each of the actions,
viz.
i) infection and colonization of root nodules by
rhizobia (Dart, 1988)
ii) the delivery of an energy supply (as photosynthate)
iii) the supply of oxygen (via leghaemoglobin)
iv) an efficient assimilation of ammonia
(Schubert, 1986), as favoured by a high
carbon-to-nitrogen ratio is optimised.
All the coupling mechanisms involved are subject to a complex genotype(s) X environment that
determines the actual success of each. In
legumes, these mechanisms have been refined
through evolution to the point where the operation of each appears to integrate almost perfectly with the others. Indeed, the specificity between particular bacterial strains and legume
species may be seen as an evolutionary refinement to ensure that greatest optimisation is
achieved by excluding less efficient symbioses.
Para-nodules
Even by comparisons with legume nodules, the
para-nodulated wheat roots have creditable rates
of C2H2 reduction. If all the C2H2 reduction was
taking place in the para-nodules, these would
have an activity several times greater than that of
legume nodules (over 1000 nmoles min-1 g-1
f.w. versus ca. 200 nmo1es min-l g-1 f.w.1. It
should be pointed out that the weight of the
wheat nodules (ca. 1-4mg) as a proportion of
the root system (ca. 100 mg f.w.) is several times
lower than in the case of legume nodules. Also,
it has not been established that the C2H2 reduction is restricted to azospirilla located in the
para-nodules.
Further, it must be emphasised that these
para-nodulated seedlings have only been tested
when around 2-3 weeks of age. Experiments to
test the persistence of the para-nodules and the
11 1
capacity to continue these rates of C2H2 reduction and to contribute to the nitrogen nutrition
of the wheat plants are now in progress.
We have also found a possible direct link
between photosynthesis and C2H2 reduction in
para-nodules on wheat, similar to that known for
legume nodules. When seedlings had exhausted
their seed reserves of carbohydrates, the C2H2
reduction continued provided the seedlings were
exposed to light (Tchan et al., 1991). In preliminary experiments, the application of diuron
(DCMU), an inhibitor of photosystem II in chloroplasts, abolished C2H2 reduction (Zeman et
al., 1991). Thus, one of the major functional
characteristics expected in a true symbiosis will
be satisfied by para-nodules, if this effect of
diuron can be confirmed as direct result of preventing supply of fresh photosynthate to Azospirillum and photosynthate is needed to allow
nitrogenase activity to proceed. This would then
indicate that the elevated activity of Azospirillum in para-nodules compared to that in associative systems (Table 1) is not a result of extensive
parasitism of wheat seedlings imposed by a degenerative condition caused by 2,4-D treatment,
as suggested by some scientists. On the contrary,
their N2 fixation may be viably linked to the
transport of photosynthate from the leaves.
However, the other key characteristic of symbiosis, the ability of a microsymbiont to benefit
the plant, has not been established. Thus, it is
too early to say whether para-nodules possess a
sufficient number of features described above
allowing legume nodules to be so successful.
Some of these features of legume nodules are
compared below to those of para-nodules, possibly throwing some light on the prospects for
eventual success of the new technology.
Nodules on rice
Nodules induced on rice, by the use of genetically modified rhizobia (Rolfe and Bender, 1990),
are characterized by internal structures rather
similar to legume nodules, with up to 20% of the
plant cells infected and containing bacteroids
encapsulated by peribacteroid membranes. Furthermore, these nodules contain vascular bundles surrounding the infected zone, similar to
legume nodules, although only about one in
400-1000 rice plants forms nodules. Better re-
112
Kennedy and Tchan
sults have been claimed by Jing, using a
mutagenized strain of Rhizobium sesbania ( A
caulinodans ), with a much greater frequency of
nodulation. If the possibility of successful N2
fixation depends on achieving a legume nodulelike structure, then this is clearly a significant
result. In neither case is significant nitrogenase
activity found in these nodules.
Rice nodules formed by the use of enzymes
resemble legume nodules less. No evidence of
peribacteroid membranes has been reported
(Cocking et al., 1990) and infected plant cells
appear to lack normal cell cytoplasm (Al-Mallah
et al., 1989). Neither do these nodules display
nitrogenase activity in the C2H2 reduction test.
Nodules on oilseed rape
Nodules formed on oilseed rape roots (Cocking
et al., 1990) have an internal structure similar to
legume nodules. They also show slight rates of
C2H2 reduction, although rates reported so far
would be too low to significantly benefit the
plant.
General
Legume symbioses are thought to have evolved
from loose associations between free-living bacteria
(Rhizobium’s
ancestors)
and
the
Leguminoseae family (Dilworth and Parker,
1969). The question of why (with the exception
of Parasponia) this development was limited to
the legumes has not been answered. To what
extent para-nodules already satisfy the requirements for success outlined above for legume
nodules must still be evaluated. Whether a process of evolution possibly (but perhaps, not
necessarily) taking an extremely lengthy period
to optimise can be achieved artificially in a few
years is also a controversial question that will
only be answered in the future.
On the basis of their relative frequencies on
each root, para-nodules on wheat appear to be
modified lateral roots. Their frequency closely
matches that of actual laterals, although the
depressive effect of 2,4-D on root extension
usually shortens their separation. With 2,4-D,
most lateral roots that might have developed
during the treatment become para-nodules.
According to Nutman’s ‘focus hypothesis’,
legume nodules are considered to be modified
lateral roots (Libbenga and Bogers, 1974). These
authors appraised the evidence for and against
the focus hypothesis suggested by Nutman. This
hypothesis states that sites of latteral root initiation are the focal points for infection and initiation of nodules. Libbenga and Bogers concluded
that the lateral root are so fundamentally different from nodules in structure and development
that the idea of a common origin cannot be
sustained. They also suggested that one of the
most persuasive pieces of evidence in favour of
the focus hypothesis, viz. the approximately
similar numbers of laterals plus nodules produced with effective and ineffective inoculation,
might be mere coincidence. The phenomenon of
‘supernodulation’ in soybeans also seems to refute this idea. Despite the controversy concerning the focus hypothesis, the possible link between lateral root initiation and nodule development in legumes adds credence to the possibility
that these new nodular structures in nonleguminous crop plants may also eventually be
functional.
Another possible viewpoint would be that vascularization of legume nodules should be a key
factor in deciding whether or not they are valid
nodules. On this basis, the tendency of paranodules to possess a central vascular bundle with
infecting bacteria occurring in surrounding cortical tissue would rule them invalid. However, the
nodules of Parasponia also have this arrangement (Trinick, 1988), with a central vascular
bundle and they are fully functional, providing
rates of N, fixation per hectare exceeding most
legumes including soybean. Also, para-nodules
infected with Azospirillum are clearly quite different in character from legume nodules and a
criterion such as this need not be considered as
significant.
Abolition of non-nodulating genetical character
by 2,4-D
Very recently, Akao et al. (1991) have demonstrated that a non-nodulating line of soybean
(T201) could be effectively nodulated by treatment of soybean roots with 2,4-D at the same
time as Bradyrhizobium was added. The shape
of these 2,4-D-induced nodules (formed irrespective of whether rhizobia were added or not, as in
Induced nodulation
the case of para - nodules on wheat) was quite
different from the normal nodules that appear on
nodulating soybean lines only when Bradyrhizobium was present. The 2,4-D-induced
nodules were usually gourd-shaped or in a doublet (‘bar-bell’) form, shown in Figure 7, and
these were inducible on a nodulating cultivar
(T202) as well. When infected with Bradyrhizobium japonicum, these induced nodules
were effective in fixing N2, as shown by the
greening of the soybean leaves and by the C2H2
reduction test.
A result such as this indicates that the effect of
2,4-D observed in wheat seedlings can also be
compared to the normal nodulation process in
leguminous plants, such as soybean. Obviously,
in both non-nodulating and nodulating soybean
(lacking rhizobia), 2,4-D initiates a process quite
similar to normal nodule formation induced by
rhizobia. The fact that the soybean nodules
initiated under the influence of 2,4-D have unusual morphology is not surprising. In the normal process, any auxin production associated
with the penetration and colonization by rhizobial bacteria is spatially directed (i.e. vectorial),
whereas 2,4-D in solution is not.
113
We have observed a similar 2,4-D-induced
nodulation of lucerne plants ( Medicago sativa ) in
the absence of Rhizobium meliloti (Yu, pers.
comm.). As mentioned above, Allen et al.
(1953) have also described pseudo-nodulation of
legumes induced with hormone herbicides, although they failed to report microbial infection
of these structures.
The future
Areas of inadequate knowledge
Although we have successfully attempted to construct an in vitro cellulo-biochemical model of
N2-fixing symbiosis between wheat seedlings and
Azospirillum, there are many challenges to be
negotiated before this model system can be considered as a true symbiosis. At this stage the
main questions that need to be asked about the
model are:
i) is there direct transfer of fixed nitrogen to
the host plant, or is any N2 fixation simply
bound to the growth of Azospirillum? If the
latter is true, it is unlikely that sufficient
Fig. 7. Barbell-shaped nodules formed on non-nodulating soybean (cv. T201). treated with 2,4-D. Scale divisions = 1 cm. We are
grateful to Dr. S. Akao for providing this photograph.
114
Kennedy and Tchan
nitrogen could be transferred to the host
plant. It is surprising, considering the potential importance that has been attributed to
Azospirillum, that little is known of any
mechanism of transfer of fixed nitrogen to a
host plant. Indeed, no evidence exists that a
direct transfer occurs at all (Zimmer and
Bothe, 1989).
ii) how long a period are the para - nodules likely to remain as N2-fixing organs? It is possible that the infected tissue might be unstable, or that nodulation might be restricted
to an insufficient number of sites and not be
able to provide sufficient N2-fixing tissue to
satisfy the needs of mature plants. The effect
of 2,4-D or NAA is to induce many lateral
root initials simultaneously, possibly stressing the plant by causing excessive demand
for photosynthate (C. Atkins, pers. commun.). However, Nie (see Tchan and Kennedy, 1989) has pointed out that a nodular
structure, when 2,4-D is removed, may revert to a lateral root, so that the process of
para-nodulation may be reversible.
iii) is the oxygen requirement for N2 fixation
likely to be satisfied in the para - nodules?
Without an effective means of regulating
oxygen activity around the bacteria, the rate
of fixation may be far from maximal.
The reasons for the success of legumes discussed above are far from verified in this model
system in wheat. Quite plainly, there is a need to
determine the actual effect of para-nodulation
with diazotrophs on the growth of wheat. The
negative effect of 2,4-D on root growth may
increase the difficulty of demonstrating a positive
effect on wheat growth, and alternatives to 2,4-D
may need to be found. There is also a potential
environmental objection to the use of 2,4-D for
field crops, even for localized applications on
roots. At this stage, our attitude is to continue
the use of 2,4-D as an experimental tool, but to
also seek better alternatives so that these will be
available in case they are required. It should be
remembered, however, that 2,4-D is still extensively used as a herbicide for broad-leaved plants
and it is rapidly broken down in soil (Jackson,
1985). Fears of environmental damage may be
exaggerated when balanced against the potential
environment benefits of wheat fixing its own N2.
The problem of the longevity of effective paranodules might prove even more intractable, but
it is too early to predict how difficult this will be,
or to predict that other serious problems will not
emerge.
Genetical techniques
It is significant to observe that the degree of
success achieved with para - nodules so far owes
nothing to genetic engineering, but it should also
be clear that further progress may depend on
genetic manipulation of either microsymbiont
(Elmerich, 1984), host plant or both. For example, the problem of nitrogen transfer might need
to be overcome by producing ammonia-'leaky'
mutants. Indeed, mutants of Azospirillum that
might enhance the process have been described
(Pedrosa et al., 1989), capable of excreting ammonia to an external concentration of at least
5.1 mM. A recent report that Azospirillum
brasilense Sp7 possesses plasmid genes homologous to nodulation genes of Rhizobium meliloti
(Vieille and Elmerich, 1990) may also be of
relevance, although we would emphasize the uniqueness of the para - nodule.
Applications in agriculture
There is no doubt that if this in vitro model can
be shown to be a working system in the field, this
would lead to widespread demand for the application of para-nodulation in agriculture. Even
then, many challenging tasks would need to be
faced in providing a viable system for the use of
farmers. The process of para - nodulation under
field conditions could require extensive redesign
before a reliable procedure could be available.
The effective delivery of azospirilla to seedlings
would also represent a challenge, although extensive experience with legume nodulation must
provide valuable lessons here. Possibly, initial
applications of the para - nodule model will need
to be conducted for horticultural crops that can
be grown hydroponically.
The various attempts to achieve effective
nodulation of non-leguminous field crops using
Rhizobium may also yet prove beneficial. This
will require suitable measures to be taken to
allow expression of nitrogenase activity. Similar
Induced nodulation
problems for persistence of nodules and continued activity would still remain.
Despite these severe challenges, we are convinced that the eventual application of this developing technology in one form or other is now
more likely than not. By her brilliant examples
represented in the legume-Rhizobium and the
plant-Frankia symbioses, ‘Nature’ has provided
the models that must be matched, at least in
overall function. However, these new technologies with very different species of plants may
not necessarily mimic these naturally evolved
symbioses. The next ten years will tell whether
this confidence is justified. It is also conceivable
that the technique of para -nodulation with desirable bacteria may find new applications outside
the N2-fixing area.
Acknowledgements
This work was supported by grants from the
Wheat Research and Development Corporation
and the Reserve Bank of Australia.
References
Akao S. Nakata S and Yoneyama T 1991 Formation of
nodules on non-nodulating soybean T201 after treatment
with 2,4-dichlorophenoxyacetate. Plant and Soil 138. 207212.
Alexander D B and Zuberer D A 1989 Impact of soil
environmental factors on rates of N2-fixation associated
with roots of intact maize and sorghum plants. In Nitrogen
Fixation with Non-Legumes. Eds. F A Skinner, R M
Boddey and 1 Fendrik. pp 273-285. Kluwer Academic
Publishers. Dordrecht, The Netherlands.
Allen K A, Allen 0 N and Newman A S 1953 Pseudonodulation of leguminous plants induced by 2-bromo-3,5dichlorobenzoic acid. Am. J. Bot. 40, 429-435.
AI-Mallah MK, Davey MR and Cocking EC 1989 Formation
of nodular structures on rice seedlings by rhizobia. J. Exp.
Bot. 40. 473-478.
Aloysius S K D and Paton A M 1984 Artificially induced
symbiotic associations of L-form bacteria and plants. J.
Appl. Bacterid. 56, 465-477.
App A A, Watanabe I, Alexander M. Ventura W. Daez C.
Santiago T and De Datta S K 1980 Nonsymbiotic nitrogen
fixation associated with the rice plant in flooded soils. Soil
Sci. 130, 283-289.
App A A, Santiago T, Daez C. Menguito C, Ventura W,
Tirol A, Po J, Watanabe I, De Datta S K and Roger P
1984 Estimation of the nitrogen balance for irrigated rice
115
and the contribution of phototrophs. Field Crops Res. 9,
17-27.
Bani D C. Barberio M, Faviti F. Gallari E and Polsinelli M
1980 Isolation and characterization of glutamate synthetase
mutants of Azospirillum brasilense. J. Gen. Microbiol.
118-9, 239-244.
Barraquio W L. Ladha J K and Watanabe I 1983 Isolation
and identitication of N2-fixing Pseudomonas associated
with wetland rice. Can. J. Microbiol. 29, 867-873.
Bender G L, Goydych W. Rolfe B G and Nayudu M 1987
The role of Rhizobium conserved and host specific nodulation genes in the infection of the non-legume Parasponia
andersonii Mol. Gen. Genet. 210, 299-306.
Bender G L, Preston L, Barnard D and Rolfe B G 1990
Formation of nodule-like structures on the roots of the
non-legumes rice and wheat. In Nitrogen Fixation:
Achievements and Objectives. Eds. P M Gresshoff, L E
Roth, G Stacey and W L Newton. Chapman and Hall,
New York.
Bergersen F J and Turner G L 1967 Nitrogen fixation by the
bacteroid fraction of breis of soybean root nodules Biochim. Biophys. Acta 141, 507-520.
Clara R W and Knowles R 1985 Superoxide dismutase,
catalase and peroxidase in ammonium-grown and nitrogenfixing Azospirillum brasilense. Can. J. Microbiol. 30, 12221228.
Cocking E C. AI-Mallah M K. Benson E and Davey M R
1990 Nodulation of non-legumes by rhizobia. In Nitrogen
Fixation: Achievements and Objectives. Eds. P M Gresshoff. L E Roth. G Stacey and W L Newton. Chapman
and Hall. New York.
Dart P J 1986 Nitrogen fixation associated with non-legumes
in agriculture. Plant and Soil 90, 303-334.
Dart P J 1988 Infection of legumes by Rhizobium. In Microbiology in Action. Eds. W G Murrell and I R Kennedy.
pp 35-52. Research Studies Pres Wiley. Chichester, UK.
Darvey N 1972 Ph.D. thesis (A cytogenetic study of wheat),
The University of New South Wales. Sydney, Australia.
De Coninck K. Horemans S, Randombage S and Vlassak K
1989 Occurrence and survival of Azospirillum spp in temperate regions. In Nitrogen Fixation With Non-Legumes.
Eds. F A Skinner. R M Boddey and I Fendrik, pp
109-114. Kluwer Academic Publishers, Dordrecht, The
Netherlands.
Diem H G and Dommergues G M 1979. Significance and
improvement of rhizospheric nitrogen fixation. In Recent
Advances in Biological Nitrogen Fixation. Ed. N S Subba
Rao. pp 190-226. Oxford Publishing, New Delhi.
Dilworth M J and Parker C A 1969 Development of the
nitrogen-fixing system in legumes. J. Theor. Biol. 25.
208-218.
Dobereiner 1974 Nitrogen fixing bacteria in the rhizosphere.
In The Biology of Nitrogen Fixation. Ed. A Quispel. pp
86-120. North-Holland, Amsterdam.
Dobereiner J 1983 Ten years Azospirillum. In Azospirillum.
II. Genetics. Physiology and Ecology. Ed. W. Klingmuller.
pp 9-23, Birkhauser-Verlag, Basel.
Dobereiner J 1989 Isolation and identification of root associated diazotrophs In Nitrogen Fixation With Non-Legumes.
Eds. F A Skinner, R M Boddey and I Fendrik. pp
116
Kennedy and Tchan
103-108. Kluwer Academic Publishers, Dordrecht, The
Netherlands.
Dreyfus B, Elmerich C and Dommergues Y R 1983 Freeliving Rhizobium strains able to grow on N2 as the soil
nitrogen source. Appl. Environ. Microbiol. 45, 711-713.
Elmerich C 1984 Molecular biology and ecology of diazotrophs associated with non-leguminous plants. Bio. Technology 2, 967-978.
Foard D E, Haber A H and Fishman T N 1965 Initiation of
lateral root primordia without completion of mitosis and
without cytokinesis in uniserate pericycle. Am. J. Bot. 52,
580-590.
Fujii T, Huang Y-D, Higashitani A, Nishimura Y, Iyama S,
Hirota Y, Yoneyama T and Dixon R A 1987 Effect of
inoculation with Klebsiella oxytoca and Enterobacter
cloacae on dinitrogen fixation by rice-bacteria associations.
Plant and Soil 103, 221-226.
Gallani E and Bazzicapulo 1985 Effect of nitrogen compounds on nitrogenase activity in Azospirillum brasilense.
FEMS Microbiol. Lett. 28, 35-38.
Gauthier D and Elmerich C 1977 Relationship between
glutamine synthetase and nitrogenase in Spirillum
lipoferum. FEMS Microbiol. Lett. 2, 101-104.
Gibson A H 1966 The carbohydrate requirements for symbiotic nitrogen fixation; A ‘whole plant’ growth analysis
approach. Aust. J. Biol. Sci. 19, 499-515.
Giles K L and Whitehead H C M 1977 Re-association of a
modified mycorrhiza with the host plant roots ( Pinus
radiata ) and the transfer of acetylene reduction activity.
Plant and Soil 48, 143-152.
Haack A 1961 Uber den Ursprung der Wurzelknöllchen von
Ornithopus sativus L. und Lupinus albus L. Z. Bakteriol.
114, 575-589.
Harari A, Kigel J and Okon Y 1989 Involvement of IAA in
the interactions between Azospirillum brasilense and
Panicum miliaceum In Nitrogen Fixation With NonLegumes. Eds. F A Skinner, R M Boddey and I Fendrik.
pp 227-234. Kluwer Academic Publishers, Dordrecht, The
Netherlands.
Hartmann A 1989 Ecophysiological aspects of growth and
nitrogen fixation in Azospirillum spp. In Nitrogen Fixation
With Non-Legumes. Eds. F A Skinner, R M Boddey and I
Fendrik. pp 123-136. Kluwer Academic Publishers, Dordrecht, The Netherlands.
Hartmann A, Fu H, Song S-D and Burris R H 1985 Comparison of nitrogenase regulation in A. brasilense, A.
lipoferum and A. amazonense. In Azospirillum III: Genetics, Physiology, Ecology. Ed. W Klingmiiller. pp 116126, Springer-Verlag, Berlin.
Hartmann A, Fu H and Burris R H 1986 Regulation of
nitrogenase activity by ammonium chloride. J. Bacteriol.
165, 864-870.
Hartmann A and Burris R H 1987 Regulation of nitrogenase
activity by oxygen in Azospirillum brasilense and Azospirillum lipoferum. J. Bacteriol. 169, 944-948.
Hegazi N A and Saleh H 1985 Possible contribution of
Azospirillum spp. to the nutritional status of wheat plants
grown in sandy soil. In Azospirillum. III. Genetics, Physiology, Ecology. Ed. W Klingmüller. pp 189-202. SpringerVerlag, Berlin.
Hurek T, Reinhold B and Niemann E-G 1987 Effect of
oxygen on NH+4-grown continuous cultures of Azospirillum
spp. and diazotrophic rods closely associated with Kallar
grass. Can. J. Microbiol. 33, 919-922.
Jackson R B 1985 Pesticide residues in soils. In Soils: An
Australian Viewpoint. CSIRO publication, Division of
Soils. pp 825-842. Canberra, Australia.
Jagnow G 1990 Differences between cereal crop cultivars in
root-associated nitrogen fixation, possible causes of variable yield responses to seed inoculation. Plant and Soil
123, 255-259.
Jing Y, Zhang B T and Shan X Q 1990a Pseudonodules
formation on barley roots induced by Rhizobium astragali.
FEMS Microbiol. Lett. 69, 123-127.
Jing Y X, Li G S, Shan X Q and Li J G 1990b Rice root
nodules with nitrogenase and hemoglobin. Proc. Int.
Symp. Nitrogen Fixation, H16. Knoxville, TN.
Kanemoto R H and Ludden P W 1984 Effect of ammonia,
darkness and phenazine methosulfate on whole-cell nitrogenase activity and Fe protein modification in Rhodospirillum rubrum. J. Bacteriol. 158, 713-720.
Kapulnik Y, Sarig S, Nur I and Okon Y 1983 Effect of
Azospirillum inoculum on yield of field-grown wheat. Can.
J. Microbiol. 29, 895-899.
Kennedy I R 1979 Integration of nitrogenase in cellular
metabolism. In A Treatise of Dinitrogen Fixation: Inorganic and Physical Chemistry and Biochemistry. Eds. R W
F Hardy, R Bottomley and R C Burns. pp 653-90. Wiley.
New York.
Kennedy I R 1983 Action and entropy in a neurological
disorder. In Molecular Aspects of Neurological Disorders.
Eds. P L Jeffrey and L Austin. pp 147-50. Academic
Press, Sydney.
Kennedy I R 1986a Acid Soil and Acid Rain, Research
Studies Press/Wiley, Chichester, UK.
Kennedy I R 1986b Relative energy costs of dinitrogen and
nitrate assimilation: Influence on cytoplasmic acidity. 8th
Aust. Nitrogen Fixation Conf., Eds. W Wallace and S E
Smith, AIAS Occasional Publn. 25, 121-122.
Kennedy IR 1988 The molecular basis of symbiotic nitrogen
fixation. In Microbiology in Action. Eds. W G Murrell and
I R Kennedy. pp 143-155. Research Studies PressiWiley,
Chichester, UK.
Kennedy I R 1091 Acid Soil and Acid Rain. Research
Studies PressiWiley, Chichester, UK. 2nd edn. In press.
Kennedy I R, Zeman A, Tchan Y T, New P B, Sriskandarajah, S and Nie Y F 1990 Biological nitrogen fixation and
prospects for yield increases in wheat. Trans. 14th Int.
Contr. Soil Sci. III, 146-151.
Kennedy I R, Sriskandarajah S, Yu D, Nie Y F and Tchan Y
T 1991 C2H2 reducing para - nodules in wheat: Effects of
growth regulators and colchicine. Proc. 9th Aust. Nitrogen
Fixn. Conf. 9, 78-79.
Kirkham M B 1990 Transit time for oxygen flow along
streamlines in vertical and horizontal tillage: A model
study. Trans. 14th Int. Congr. Soil Sci. I, 298-301.
Kostychev S, Sheloumova A and Shul’gina 1926 Nitrogen
content of soils of the southern coast of Crimea ( Russian ).
Sonetskii Agron. 1, 1926.
Ladha J K, Tirol A C. Laroy L G, Caldo G, Ventura W and
Induced nodulation
Watanabe I 1986 N2 fixation (C2H2 reduction) by five rice
varieties, and relationship with plant growth characters as
affected by straw incorporation. Soil Sci. Plant Nutr. 32.
91-106.
Ladha J K, Tirol-Padre A, Punzulan G C and Watanabe I
1987 Nitrogen-fixing (C2H2-reducing) activity and plant
growth characters of 16 wetland rice varieties. Soil Sci.
Plant Nutr. 33, 187-200.
Ladha J K, Garcia M, Miyan S, Padre A T and Watanabe I
1989 Survival of Azorhizobium caulinodans in the soil and
rhizosphere of wetland rice under Sesbania rostrata -rice
rotation. Appl. Environ. Microbiol. 55, 454-460.
Ladha J K, Pareek R P, So R and Becker M 1990 Stem
nodules symbiosis and its unusual properties. In Nitrogen
Fixation: Achievements and Objectives. Eds. P M Gresshoff, L E Roth, G Stacy and W L Newton. Chapman and
Hall, New York.
Li R P and MacRae I C 1991 Nitrogen fixing bacteria
associated with sugarcane in Queensland. Proc. 9th Aust.
Nitrogen Fixation Conf. 9, 9-10.
Libbenga K R and Bogers R J 1974 Root-nodule morphogenesis. In The Biology of Nitrogen Fixation. Ed. A
Quispel. pp 430-472. North-Holland. Amsterdam.
Mertens T and Hess D 1984 Yield increases in spring wheat
( Triticum aestivum L.) inoculated with Azospirillum
lipoferum under greenhouse and field conditions of a
temperate region. Plant and Soil 82. 87-99.
Millbank J W 1974 Associations with blue-gree algae. In The
Biology of Nitrogen Fixation. Ed. A Quispel. pp 238-264.
North-Holland, Amsterdam.
Millet E and Feldman M 1984 Yield response of a common
spring wheat cultivar to inoculation with Azospirillum
brasilense at various levels of nitrogen fertilization. Plant
and Soil 80, 255-259.
Millet E, Longeri L and Arbarzua M 1989 Ultrastructure of
Frankia isolated from three Chilean shrubs (Rhamnaceae).
In Nitrogen Fixation With Non-Legumes. Eds. F A Skinner, R M Boddey and I Fendrik. pp 47-56. Kluwer
Academic Publishers, Dordrecht. The Netherlands.
Mishustin E N and Shil’Nikova V K 1971 Biological Fixation
of Atmospheric Nitrogen. MacMillan, London.
Mulder E G and Brotonegoro S 1974 Free-living heterotrophic nitrogen-fixing bacteria. In The Biology of Nitrogen
Fixation. Ed. A Quispel. pp 37-85, North-Holland Amsterdam.
New P B and Kennedy I R 1989 Regional distribution and
pH sensitivity of Azospirillum associated with wheat roots
in eastern Australia. Microbial Ecol. 17, 299-309.
New P B, Sukiman H I and Kennedy I R 1991 Host
plant-bacterial strain specificity in Azospirillum. Proc. 9th
Aust. Nitrogen Fixation Conf. 9, 76-77.
Nye P H and Tinker P B 1977 Solute Movement in the
Soil-root system. p 207. Blackwell, Oxford.
Okon Y, Kapulnik Y and Sarig S 1988 Field inoculation
studies with Azospirillum in Israel. In Biological Nitrogen
Fixation. Ed. N S Subba Rao. pp 175-196. Oxford and
IBH Publishing, New Delhi.
Pacovsky R S 1989 Influence of inoculation with Azospirillum
brasilense and Glomus fasciculatum on sorghum nutrition.
In Nitrogen Fixation With Non-Legumes. Eds. F A Skin-
117
ner, R M Boddey and I Fendrik. pp 235-240. Kluwer
Academic Publishers, Dordrecht, The Netherlands.
Pedrosa F 0, De Souza E M, Machado H B, Rigo L U and
Funayama S (1989) Regulation of nif genes expression in
Azospirillum brasilense and Herbaspirillum seropedicae. In
Nitrogen Fixation With Non-Legumes. Eds. F A Skinner,
R M Boddey and I Fendrik. pp 155-163. Kluwer Academic Publishers, Dordrecht, the Netherlands.
Pereira J A R. Cavalcante V A, Baldini V I and Dobereiner
J 1989 Field inoculation of sorghum and rice with Azospirillum spp. and Herbaspirillum seropedicae In Nitrogen
Fixation With Non-legumes. Eds. F A Skinner, R M
Boddey and I Fendrik. pp 219-226. Kluwer Academic
Publishers, Dordrecht, The Netherlands.
Piana L, Delledone M, Antonelli M N and Fogher C 1988
Root hair deformation induced on maize and medicago by
an Azospirillum transconjugant containing a Rhizobium
meliloti nodulation region. In Azospirillum IV. Genetics,
Physiology, Ecology. Ed. W Klingmiiller. pp 83-91.
Springer Verlag, Berlin.
Plazinski J. Innes R W and Rolfe B G 1985 Expression of
Rhizobium trifolii early nodulation genes on maize and rice
plants. J. Bacteriol. 163, 812-815.
Postgate J 1972 Biological Nitrogen Fixation. Merrow Publn.
Co. Watford.
Postgate J R 1974 Prerequisites for biological nitrogen fixation in free-living heterotrophic bacteria. In The Biology of
Nitrogen Fixation. Ed. A Quispel. pp 663-686. NorthHolland. Amsterdam.
Quispel A 1974 The endophytes of the root nodules in
non-leguminous plants. In The Biology of Nitrogen Fixation. Ed. A Quispel. pp 499-520. North-Holland Amsterdam.
Rai S N and Gaur A C 1982 Nitrogen fixation by Azospirillum spp. and effect of Azospirillum lipoferum on the yield
and N-uptake of wheat crop. Plant and Soil 69, 233-238.
Ravishankar H N. Kennedy I R and New P B 1986 Autotrophic growth and nitrogen fixation in Derxia gummosa. J.
Gen. Microbiol. 132, 1797-1403.
Reinhold B and Hurek T 1989 Location of diazotrophs in the
root interior with special attention to the kallar grass
association. In Nitrogen Fixation With Non-Legumes. Eds.
F A Skinner, R M Boddey and I Fendrik. pp 199-208.
Kluwer Academic Publishers, Dordrecht, The Netherlands.
Rodriquez-Barrueco C, Sevillano F and Nebrada T M 1989
Effective nodulation of Coriaria myrtifolia L. induced by
Frankia strains from Alnus glutinosa. In Nitrogen Fixation
With Non-Legumes. Eds. F A Skinner, R M Boddey and I
Fendrik. pp 15-24. Kluwer Academic Publishers, Dordrecht. The Netherlands.
Roger P A and Ladha J K 1990 Estimation of biological N2
fixation and its contibution to nitrogen balance in wetland
rice field. Trans. 14th Int. Congr. Soil Sci. III, 128-133.
Rolfe B G and Bender G L 1990 Evolving a Rhizobium for
non-legume nodulation. In Nitrogen Fixation: Achievements and Objectives. Eds. P M Gresshoff, L E Roth, G
Stacey and W L Newton. Chapman and Hall, New York.
Roughley R J 1988 Principles of inoculant strain selection,
inoculant production and quality control. In Microbiology
118
Induced nodulation
in Action. Eds. W J Murrell and I R Kennedy. pp 53-66.
Research Studies Press/Wiley, Chichester, UK.
Schubert K R 1982 The energetics of biological nitrogen
fixation, Workshop Summaries, I, 1, Am. Soc. Plant Physiol. pp 4-20.
Schubert K R 1986 Products of biological nitrogen fixation in
higher plants: Synthesis, transport and metabolism. Annu.
Rev. Plant Physiol. 37, 539-556.
Silsbury J H 1977 Energy requirement for symbiotic nitrogen
fixation. Nature, London 267, 149-150.
Skinner F A, Boddey R M and Fendrik I (Eds.) 1989
Nitrogen Fixation with Non-Legumes. Kluwer Academic
Publishers, Dordrecht. The Netherlands.
Smith R L, Bouton J H, Schank S C, Quesenberry K H,
Tyler M E, Milam J R, Gaskins M H and Litell R C 1976
Nitrogen fixation in grasses inoculated with Spirillum
lipoferum. Science 193, 1003-1005.
Smith R L, Schank S C, Bouton J H and Quesenberry K H
1978 Yield increases of tropical grasses after inoculation
with Spirillum lipoferum. Ecol. Bull. Stockholm 26, 38035.
Smith R L, Schank S C, Milam J R and Baltensperger A
1984 Response of Sorghum and Pennisetum species to the
N2-fixing bacterium Azospirillum brasilense. Appl. Environ. Microbiol. 47, 1331-1336.
Sprent J and de Faria M 1989 Mechanisms of infection of
plants by nitrogen fixing organisms In Nitrogen Fixation
With Non-Legumes. Eds. F A Skinner, R M Boddey and 1
Fendrik. pp 3-14. Kluwer Academic Publishers, Dordrecht, The Netherlands.
Tchan Y T 1988 Some aspects of non-rhizobia1 diazotrophs:
their past and their future. In Microbiology in Action. Eds.
W G Murrell and I R Kennedy. pp 193-207. Research
Studies Press/Wiley, Chichester, UK.
Tchan Y T and Kennedy I R 1988 Asymbiotic nitrogen
fixation. In Nitrogen Cycling in Temperate Agricultural
Systems. Eds. P E Bacon, J Evans, R R Storrier and A C
Taylor. pp 120-142. Riverina Branch: Australian Society
of Soil Science.
Tchan Y T and Kennedy I R 1989 Possible N2-fixing root
nodules induced in non-legumes. Agric. Sci. (AIAS, Melbourne) 2, 57-59.
Tchan Y T, Zeman A and Kennedy I R 1991 Nitrogen
fixation in para - nodules of wheat roots by introduced
free-living diazotrophs. Plant and Soil 137, 43-47.
Terzachi B E and Christensen M J 1986 Should the transfer
of nitrogen-fixing ability to an eukaryotic cell be reconsidered? J. Plant Physiol. 122, 275-283.
Tien T M, Gaskins M H, and Hubbell D H 1979 Plant
growth substances produced by Azospirillum brasilense and
their effect on the growth of pearl millet ( Pennisetum
americanum L.), Appl. Environ. Microbiol. 37, 10161024.
Trinick M J 1973 Symbiosis between Rhizobium and the
non-legume Trema aspera. Nature, London 144, 459.
Trinick M J 1988 Bradyrhizobium of the non-legume Parasponia. In Microbiology in Action. Eds. W G Murrell and I
R Kennedy. pp 107-118. Research Studies PressiWiley,
Chichester, UK.
Urquiaga S, Botteon P B L and Boddey R M 1989 Selection
of sugarcane cultures for associated biological nitrogen
fixation using 15 N soil. In Nitrogen Fixation with Nonlegumes. Eds. F A Skinner, R M Boddey and I. Fendrik
pp 311-319. Kluwer Academic Publishers, Dordrecht, The
Netherlands.
Vieille C and Elmerich C 1990 Characterization of two
Azospirillum brasilense Sp7 plasmid genes homologous to
Rhizobium meliloti Nodpq. Mol. Plant-Microbe Inter. 3,
389-400.
Vincent J M 1980 Factors controlling the legume-Rhizobium
symbiosis. In Nitrogen Fixation, Volume II. Eds. W E
Newton and W H Orme-Johnson. pp 103-129. University
Park Press, Baltimore, MD.
Vincent J M 1984 Potential for enhancing biological nitrogen
fixation. In Crop Breeding: A Contemporay Basis. Eds. P
B Jose and S C Blixt. pp 185-215. Pergamon Press,
Oxford.
Volpon A G T, De-Polli H and Podenier J 1981 Physiology of
nitrogen fixation in Azospirillum lipoferum BR17. Arch.
Microbiol. 128, 371-375.
Wani S P, Chandrapalaiah, Zanbre M A, and Lee K K 1989
Association between N2-fixing bacteria and pearl millet
plants: Responses, mechanisms and persistence. In Nitrogen Fixation with Non-Legumes. Eds. F A Skinner, R M
Boddey and I Fendrik. pp 249-262. Kluwer Academic
Publishers, Dordrecht, The Netherlands.
Watanahe I, Yoneyama T, Padre B and Ladha J K 1987a
Difference in natural abundance of 15 N in several rice
( Oryza sativa L.) varieties: Application for evaluating N2
fixation. Soil Sci. Plant Nutr. 33, 407-415.
Wdtanabe I, So R, Ladha J K, Katayama-Fujimura Y and
Kuraishi H 1987b A new nitrogen-fixing species of pseudomonad: Pseudomonas diazotrophicus sp. nov. isolated
from the root of wetland rice. Can. J. Microbiol. 33,
670-678.
Whipps J M and Lynch J M 1983 Substrate flow and utilization in the rhizosphere of cereals. New Phytol. 95, 605623.
Wilde M H 1951 Anatomical modifications of bean roots
following treatment with 2,4-D. Am. J. Bot. 38, 79-91.
Yoo I D, Fujii T, Sano Y, Komagata K, Yoneyama T, Iyama,
S and Hirota Y 1986 Dinitrogen fixation of rice-Klebsiella
associations. Crop Sci. 26, 297-301.
You C and Zhou F 19, 89 Non-nodular endorhizospheric
nitrogen fixation in wetland rice. Can. J. Microbiol. 35,
403-408.
Yu D G 1988 Studies on 2,4-D induction of wheat nodules
and the introduction of Rhizobium into nodules. M.Sc.
thesis supervised by T W Chen (CAAS, Beijing).
Zeman A. Tchan Y T and Kennedy I R 1991 The paranodule on wheat as a model of a N2-fixing symbiosis on a
non-legume. Proc. 9th Aust. Nitrogen Fixation Conf. 9,
8o-81.
Zhu Z 1989 Dynamics of soil nitrogen and its management.
In Progress in Irrigated Rice Research. pp 151-164. International Rice Research Conf., IRRI, Manila.
Zimmer W and Bothe H 1989 The phytohormonal interactions between Azospirillum and wheat. In Nitrogen Fixation With Non-Legumes. Eds. F A Skinner, R M Boddey
and I Fendrik. pp 137-146. Kluwer Academic Publishers,
Dordrecht, The Netherlands.
Plant and Soil 141: 119-135, 1992.
© 1992 Kluwer Academic Publishers. Printed in the Netherlands.
PLSO SA07
Potential for increasing biological nitrogen fixation in soybean
HAROLD H. KEYSER and FUDI LI
NifTAL Project, University of Hawaii, 1000 Holomua Ave., Paia, HI 96779, USA and Laboratory
for Biological Nitrogen Fixation, Huazhong (Central China) Agricultural University, Wuhan, Hubei
430070, People's Republic of China
Key words: biological nitrogen fixation, Bradyrhizobium japonicum, competition, genetics, Glycine
max, inoculation, limiting factors, Sinorhizobium
Table of contents
Abstract
Introduction
Importance of soybean in world agriculture
Recent soybean statistics
The need to increase soybean productivity within sustainable systems
Biological nitrogen fixation in the soybean-bradyrhizobia symbiosis
Amounts of N2 fixed
Assimilation of nitrogen
Host-strain compatibility
Competition for nodule occupancy
Research strategies to increase BNF in soybean
Selection and engineering of bradyrhizobia
Selection and breeding of soybean genotypes
Improved inoculation techniques
Production strategies to increase BNF in soybean
Inoculant production, quality control and training
Matching soybean genotypes to the environment
Management of other inputs
Concluding
remarks
References
Abstract
The importance of soybean as a source of oil and protein, and its ability to grow symbiotically on low-N
soils, point to its continued status as the most valuable grain legume in the world. With limited new
land on which to expand, and emphasis on sustainable systems, increases in soybean production will
come mostly from increased yield per unit area. Improvements in biological nitrogen fixation can help
achieve increased soybean production, and this chapter discusses research and production strategies for
such improvement.
The soybean-Bradyrhizobium symbiosis can fix about 300 kg N ha -1 under good conditions. The
120
Keyser and Li
factors which control the amount of N fixed include available soil N, genetic determinants of
compatibility in both symbiotic partners and lack of other yield-limiting factors. Response to inoculation is controlled by the level of indigenous, competing bradyrhizobia, the N demand and yield
potential of the host, and N availability in the soil.
Research efforts to improve BNF are being applied to both microbe and soybean. While selection
continues for effective, naturally occurring bradyrhizobia for inoculants and the use of improved
inoculation techniques, genetic research on bradyrhizobia to improve effectiveness and competitiveness
is advancing. Selection, mutagenesis and breeding of the host have focused on supernodulation,
restricted nodulation of indigenous B. japonicum, and promiscuous nodulation with strains of bradyrhizobia from the ‘cowpea’ cross-inoculation group. The research from the host side appears closer to
being ready for practical use in the field.
Existing knowledge and technology still has much to offer in improving biological nitrogen fixation in
soybean. The use of high-quality inoculants, and education about their benefits and use can still make a
significant contribution in many countries. The importance of using the best adapted soybean genotype
with a fully compatible inoculant cannot be overlooked, and we need to address other crop management factors which influence yield potential and N demand, indirectly influencing nitrogen fixation. The
implementation of proven approaches for improving nitrogen fixation in existing soybean production
demands equal attention as received by research endeavours to make future improvements.
Introduction
Soybean ( Glycine max L. Merrill) is one of
man’s most important sources of food and feed.
It is one of nature’s most versatile plants, and
produces an abundant supply of protein and oil
in both temperate and tropical environments.
Since its domestication around the 11th century
B.C. in northeast China, soybean has been a
staple food in eastern Asia (Hymowitz, 1970).
Its significance as a world crop came after its
introduction and highly successful adoption in
the U.S.A. (Shanmugasundaram, 1989). From
1941 to present, soybeans ascended from an
insignificant forage crop to become the second
most valuable U.S. crop, surpassed only by
maize (Anon., 1984). Brazil and Argentina have
also seen large increases in the production of
soybeans in the past two decades, leading the
way in increased global production of soybean.
As a nodulating legume, soybean forms a nitrogen-fixing symbiosis with Bradyrhizobium
japonicum and Sinorhizobium species. This attribute, in addition to its valuable oil and protein
and wide environmental adaptation, ensures that
it will continue to be an important world food
crop in an era of increasing food demand and
concerns for the sustainability of agricultural
production systems. A great challenge will be
meeting the food requirements of an increasing
world population through comparable increases
in productivity in sustainable systems (Cassman,
1990).
Several good reviews of biological nitrogen
fixation (BNF) focusing on soybean have been
written (Hardy and Havelka, 1976; Harper,
1987; Vest et al., 1973). The present paper is
intended to serve as an update in this area, with
emphasis on the established principles and new
findings that have practical application, or a
reasonable potential to be applied, toward increasing BNF in soybean.
Importance of soybean in world agriculture
Soybean is the world’s premier oilseed crop. Of
the eight major oilseeds traded in world markets
(soybean,
cottonseed,
peanut,
sunflower,
rapeseed, flaxseed, copra and palm kernel), soybean’s production has been twice that of any
other oilseed since 1970 (Smith and Huyser,
1987).
The approximate composition of soybean is
40% protein, 21% oil, 34% carbohydrate and
5% ash (Scott and Aldrich, 1983). Soybean meal
accounts for 60-70% of the value of the soybean, with the balance from oil. Soybean supplies one-fourth of the world’s fats and oils,
about two-thirds of the world’s protein-concentrate animal feeds, and three-fourths of the
BNF in soybean
world trade in high-protein meals (Anon., 1984).
Major importers of soybean include countries
of the European Economic Community, Japan,
and Eastern Europe. The demand for soybean is
mainly for oil and meal products, rather than
whole beans, In accounting for utilization of
soybean, some 39 products have been identified
ranging from livestock feed, salad oils and baby
foods to industrial adhesives, putty and use in
pharmaceuticals (Smith and Huyser , 1987).
By weight, the protein yield of soybean is
about twice that of meat and of most beans and
nuts, four times that of eggs and cereals, and
twelve times that of milk (Anon., 1984). In
countries with rapidly increasing populations,
soybean is viewed as a crop that enhances nutritive value of the local diets and lessens national
shortages of vegetable oil (Hume et al., 1985).
With the diet of many people in the world
deficient in protein and calories, soybean seems
destined to remain an important commodity.
Recent soybean statistics
Over the past two decades, world soybean production has increased nearly two and one-half
fold (Table 1). This can be attributed to nearly
equal increases in area and yield.
While the U.S. is still the largest producer of
soybeans, over the past two decades Brazil and
Argentina have made large gains. Most likely
Argentina will surpass China in this decade.
These four countries dominate the world soybean market and will continue to do so for some
time. The world production reached in 1987 will
have to almost double to meet the projected
121
production and demand of 190 million tonnes in
the year 2002 (Smith and Huyser, 1987).
The need to increase soybean productivity
within sustainable systems
A formidable challenge will be meeting future
global food demands when faced with declining
productivity in areas where soil and water resources are being depleted, combined with the
diminishing supply of potentially arable land on
which to expand agricultural production (Cassman, 1990; Larson, 1986; Russell et al., 1989).
This means that sustainable agricultural systems
will have to intensify production (yield) per unit
area.
Yield increases in soybean have typically been
gradual. The U.S. national average yield has
risen from 1.579 kg ha-1 in 1960 to about
2150 kg ha-1 in 1982 (Anon., 1984). Studies on
the rate of yield increase reveal that the average
rate was 15.1 kg ha-1 yr-1, a 0.6% average yield
increase per year (Fehr, 1987). The world as a
whole has also made steady yield improvement
(Table
1).
Experimental
yields
up
to
8,500 kg ha-1 indicate that the genetic potential
is far above current production levels (Lawn and
Byth, 1989; Troedson, 1988). However, the rate
of soybean yield increase has been well below
that of cereals (Russell et al., 1989). With advantages in yield, rate of yield increase and preference as a staple, the importance of cereal crops
will remain or even increase, further limiting
area for expansion of legume production (Cassman, 1990). It therefore appears likely that the
bulk of future increased soybean production will
Table 1. World trend in soybean production and yield per ha during the past two decades
Yield (kg ha-1)
Region or
country
Area (1000 ha)
Production (1000 t)
1967
1977
1987
1967
1977
1987
1967
1977
1987
World
Asia
Africa
Europe
Argentina
Brazil
China
USA
33,556
1,595
48
56
17
612
14,050
16,093
49,243
15,986
250
306
660
7,070
14,236
23,314
51,567
12,002
398
1.023
3,510
9,161
8.404
22,839
40,638
1,186
34
50
21
716
11,100
26,564
79,206
14.647
145
399
1,400
12,513
12.955
47,948
98,288
15,437
414
2.454
7.000
16,876
11,816
51,839
1,210
740
710
890
1,190
1,170
790
1.650
1.608
916
580
1,305
2,121
1,770
910
2,057
1.906
1,286
1,039
1,299
1,994
1,842
1,406
2,270
Source: FAO data summarized by Shanrnugasundaram, 1989.
122
Keyser and Li
have to come from yield improvements rather
than increased planting area. Improving the BNF
component for soybean has been identified as
part of the overall strategy for increasing productivity (Anon., 1984; Olsen, 1982; Russell et al.,
1989; Scott and Aldrich, 1983).
Biological nitrogen fixation in the soybeanbradyrhizobia symbiosis
The symbiosis involving soybean and bradyrhizobia (used in this chapter to refer to both
Bradyrhizobium and Sinorhizobium, unless
otherwise specified) is a well-organized system
and goes through many steps, beginning at the
root surface and resulting in a N2-fixing nodule
(Vincent, 1980). The host plant provides carbon
substrate as a source of energy, and the bacteria
reduce atmospheric N2 to NH3 which is exported
to plant tissues for eventual protein synthesis.
The efficiency of symbiotic BNF is markedly
dependent on the mutual compatibility of both
partners, and is influenced by a number of environmental factors (Sprent and Minchin, 1983;
Vincent, 1980). Major factors that affect the
symbiotic system are summarized in Table 2.
The following discussion focuses on those significant principles of BNF in the soybean-brady-
rhizobia symbiosis which most influence our ability to manage and manipulate it at the field level.
Amounts of N2 fixed
Soybean, like other nodulated legumes, utilizes
two sources of N for its growth-mineral N in the
+
soil (in the form of NO3 and NH4) and atmospheric N2 fixed in nodules. The soybean has been
characterized as being rather non-responsive to
the application of fertilizer N (Mengel et al.,
1987; Scott and Aldrich, 1983). If abundantly
nodulated, soybean is capable of fixing substantial amounts of its required N from BNF. The N
requirement of soybean is the highest among
agronomic crops (Sinclair and de Wit, 1975).
Each tonne of soybean seed requires the crop to
assimilate approximately 100 kg N.
Different techniques have been used to
evaluate N, fixation by soybean. They are reviewed elsewhere (Guffy et al., 1989; Peoples et
al., 1989). The proportion of N derived from
fixation varies substantially from zero to as high
as 97%. As shown in Table 3, most estimates fall
between 25% to 75%. LaRue and Patterson
(1981) reported an average estimate of N, fixation in soybean to be 75 kg N ha -1 , using average
commercial yields and assuming that 50% of the
N was from fixation. Yet, field studies by Bez-
Table 2. Factors that affect symbiotic nitrogen fixation
Factor
Effect
Macrosymbiont
Variety
Nodulin (e.g. Leghemoglobin)
Photosynthate availability
Tolerance of stress
Nodulation and nitrogen fixation
Nodule function
Nitrogen-fixing efficiency
Establishment of symbiosis
Microsymbiont
Infectiveness
Effectiveness
Competitive ability
Saprophytic competence
Nodule formation
Nitrogen fixation
Nodule occupancy
Persistence of rhizobia in soil
Environment
Combined nitrogen
Light
Temperature
Water and aeration
Salinity
Biotic agent
Nodulation and nitrogen fixation
Nitrogen-fixing efficiency
Soybean growth and nitrogen fixation
Nitrogenase activity
Reduce nitrogenase activity
Rhizobium viability and infection
BNF in soybean
123
Table 3. Estimates of symbiotic nitrogen fixation by soybean
Location
N2 fixed
Delaware. USA
Nigeria
Alberta. Canada
Jilin, China
Jilin, China
Australia
Hawaii, USA
USA
Washington. USA
a
Reference
kgN ha
per year
Ndfa
80-120
125-188
33-151
57-151
45-100
0-236
166-237
41 -237
0-311
25-30
84-87
14-62
43-85
50-70
0-71
66-97
36-48
0-80
-1
a
Method
C2H2 reduction
N difference
N15 isotope dilution
C2H2 reduction
N15 isotope dilution
Urelde
N difference
N15 isotope dilution
N difference
Hardy et al. (1973)
Eaglesham et al. (1982)
Rennie (1984)
Zhang et al. (1986)
Gao et al. (1987)
Herridge and Holland (1987)
George et al. (1988)
Guffy et al. (1989)
Bezdicek et al. (1978)
Nitrogen derived from the atmosphere. percent of total N.
dicek et al. (1978) show that soybeans are capable of fixing over 300 kg N ha-1 when the soil
is low in available N and effective strains of
bradyrhizobia are supplied in high number.
Available soil N has a large influence on BNF.
George et al. (1988) found that soil-N availability at different sites determined the relative contribution of symbiotic N2 fixation, regardless of
crop duration and total N accumulation by different varieties. The amounts and proportion of
N derived from the atmosphere is also influenced
by several other factors, including temperature,
soybean cultivar, bradyrhizobial strain, root
nodule position, quantity and form of fertilizer N
and management practices (Buttery and Dirks,
1987; Danso et al., 1987; Eaglesham et al., 1983;
George et al., 1988; Hardarson et al., 1989;
Herridge and Holland, 1987).
It has been suggested that increasing the
amounts of N, fixed in soybean, and the portion
of total plant N derived from fixation, may only
be achieved with concomitant yield increases
(Herridge and Bergersen, 1988). Experimentally, it is not easy to separate N2 fixation from
yield. This is apparent from studies comparing
the two parameters in soybeans of differing
maturities; the late maturing cultivars fix more
N2 and yield more than earlier types due to a
longer reproductive phase, when rates of N2
fixation and seed biomass accumulation are high
(Patterson and LaRue, 1983; George et al.,
1988). However, it appears that the portion of
total N derived from fixation remains fairly constant for cultivars of different maturity at a given
site (George et al., 1988), as summarized below
in Table 4.
Assimilation of nitrogen
The N requirement of soybean can be met by
both mineral N assimilation and symbiotic N,
fixation. Although each N input system has independent pathways and control points, the soybean plant under almost all field conditions will
use both systems, and these systems are interdependent (Harper, 1987).
Nitrogen fixation generally reaches a peak at
early podfill and declines during late reproductive phases (Imsande, 1989; Latimore et al.,
1977; Lawn and Brun, 1974; Thibodeau and
Jaworski, 1975). The plant mobilizes a large
quantity of N from vegetative tissue to meet the
demand for seed N, whereas the net rate of NO3
uptake gradually declines throughout pod fill.
On the other hand, nodules produced with the
first infections on the primary root of soybean
only have an average duration of 65 days, and
undergo rapid aging just after flowering (Bergersen, 1958). Because of the time lag (approximately 4 weeks) between infection and rapid N2
fixation, infection must occur at approximately
the R2 stage (see Fehr et al., 1971), if rapid N,
fixation is to occur during R5 when pod fill
proceeds at an appreciable rate (Imsande, 1989).
Therefore nodulation on lateral and deep roots
may be essential for maximum N, fixation, in
order to match the high N demand during pod
fill (Imsande, 1989; Zapata et al., 1987).
124
Keyser and Li
Table 4. Total N accumulation and percent of total N derived from symbiotic fixation by five soybean varieties grown at three
different elevations
Site
(elevation, m)
Varieties
Clay
00a
Clark
IV
D68-0099
VI
Kuiaha (320)
Haleakala (660)
Olinda (1050)
Total N accumulation (kg ha-1)
316
236
317
199
246
271
44
99
128
Kuiaha (320)
Haleakala (660)
Olinda (1050)
Percent N from
85
62
95
fixation
82
71
98
85
65
98
N77-4262
VII
Hardee
VIII
Mean
260
258
186
349
276
144
295Ah
250B
120C
70
66
98
80
68
96
80B
66C
97A
Indicates maturity group designation of the variety.
Data followed by the same letter within the column are not significantly different ( p = 0.05) by Duncan’s Multiple Range test.
Source: George et al., 1988.
a
b
It is well established that increasing levels of
mineral N in the rhizosphere inhibit soybean
nodule formation and functioning. Concentration of available N above 2 mM generally decreases N2 fixation in symbioses (Phillips and
DeJong, 1984). Under controlled conditions,
15 mM N has been found to decrease nodule
numbers 2.5-fold (Malik et al., 1987). Imsande
(1986) reported that short-term (3-6 days) exposure to 4 mM N only temporarily delayed nodule
formation, and the late steps of nodule development were reversibly inhibited. Extended growth
in the presence of 4 mM N blocked both early
and late steps of nodule development. Split-root
experiments have shown that N inhibited soybean nodule formation through localized effects
on the root system rather than as a function of
whole plant nutrition (Eaglesham, 1989b; Hinson, 1975).
Symbiotic N2 fixation may not meet the soybean N requirement during the early and late
phases of growth. Small amounts of available N
(0-2 m M ), supplied early, often promote growth
and N2 fixation in legumes, indicating N-limited
early growth (Phillips and DeJong, 1984). Increasing the N supply to soybean during flowering and pod fill have increased total N and seed
yield in the field (Brevedan et al., 1978; George
et al., 1988; Thies et al., 1991). Field response of
soybean to fertilizer N may be related to the
amount of NO; in the root zone. When this
amount was low, the use of N fertilizer significantly increased soybean seed yield at several
soil moisture levels (Al-Ithawi et al., 1980). In
the absence of indigenous rhizobia, Thies et al.
(1991) found that inoculation response was directly proportional to the availability of mineral
N in the soil.
Nitrogen fixation requires about 10 kg of carbohydrates/kg of N2 fixed, and the equivalent of
25-28 molecules of ATP for each molecule of N,
fixed (Havelka et al., 1982). Soybean has the
relatively inefficient C3 photosynthetic system.
Soybean growth rate is thought to be photosynthate source-limited rather than sink-limited.
Due to large energy requirements, BNF is
thought to be closely coupled to photosynthate
production, particularly during the reproductive
stages (Hardy and Havelka, 1976). A recent
proposal based on experimental and theoretical
evidence indicates that carbon and nitrogen are
simultaneously limiting soybean yield increases
(Sinclair, 1989). Related proposals by others
(Imsande, 1989; Millhollon and Williams, 1986)
indicate that this area of research is quite complicated, but future findings should be valuable in
guiding strategy to increase soybean yields.
Host-strain compatibility
The establishment and functioning of an effective symbiosis is dependent on genetic determin-
BNF in soybean
ants in both plant and bacteria. The fully compatible symbiosis proceeds from recognition,
penetration, stimulation of host-cell division, differentiation
of
rhizobia
into
bacteroids,
leghemoglobin synthesis, nitrogenase synthesis
and its activity. Host-strain compatibilities for
nodulation, effectiveness and efficiency of N2
fixation, and competitiveness for nodule occupancy have been studied intensively in the
soybean-bradyrhizobia system, and some serve
as classic models for other symbioses (Cregan et
al., 1989a; Evans et al., 1980; Vest et al., 1973).
To date, some 45 genes across eight legume
species have been identified as affecting nodulation and N2 fixation, including at least eight
genes in soybean (Vance et al., 1988). Five of
these genes control restricted nodulation and are
summarized in Table 5. These host genes control
nodulation at the species, serogroup and strain
level within B. japonicum and Sinorhizobium
spp. The rj1 gene that conditions non-nodulation
has been transferred through conventional
breeding to several soybean cultivars for use in
estimating response to inoculation and the
amount of N2 fixed.
A unique incompatibility expressed in many
soybean genotypes when nodulated by some
strains of certain serogroups is the production of
rhizobitoxine (Johnson and Clark, 1958; La
Favre and Eaglesham, 1986). The toxic compound is produced in the roots and translocated
to the newly developing leaves, where it produces a chlorosis. The effect can be severe
enough in contained systems (such as small plant
growth assemblies) to kill the plant. In the field,
expressions of the chlorotic symptoms are tran-
125
sient and have been associated with high rates of
nodule occupancy by these same serogroups
(D.F. Weber, unpublished data). Many of these
same strains are also capable of nodulating the
non-nodulating (rj1rj1) soybean under controlled
conditions (Clark, 1957; Devine and Weber,
1977).
Reports have shown that the number of
nodules and their distribution patterns on soybean roots are largely dependent on host influence (Carroll et al., 1985; Kosslak and
Bohlool, 1984). Singleton and Stockinger (1983)
demonstrated that the soybean will compensate
for ineffective nodulation by producing more
mass in those nodules containing effective
strains. Soybean mutants have been developed
which lack normal regulation of amount of nodulation, allowing ‘supernodulation’ to occur (Carroll et al., 1985).
Competition for nodule occupancy
One of the major problems in inoculation technology with soybean is the establishment of an
introduced inoculant strain of B. japonicum in
the nodules of soybean grown in soils which
contain indigenous populations of bradyrhizobia
(Ham, 1976, 1980; Tang, 1979; Vest et al.,
1973). Previous inoculation and continued cropping of soybean confer a formidable advantage
in numbers and environmental adaptation to the
indigenous population in competition with the
introduced strains. Thus, when the indigenous
strains dominate the nodules, response to inoculation is not observed (Ge and Xu, 1982; Kapusta and Rouwenhorst, 1973; Kvien et al., 1981).
Table 5. Host genetic control of nodulation in soybean
Allele
Phenotype
Reference*
rj1
Rj2
Non-nodulating
Cortical proliferations or ineffective nodules
formed by strains in serogroup 6(c1) and 122
with soybean cv. Hardee
Small. nodule-like structures with white
interior by strain USDA 33 with cv. Hardee
Very few cortical proliferations by strain
USDA 61 with cv. Hill
Swellings or rudimentary nodules by
Sinorhizobium fredii strain USDA 205
with cv. Kent.
1
Rj3
Rj4
Dominant
undesignated
a
1) Williams and Lynch, 1954; 2) Caldwell, 1966; 3) Vest, 1970; 4) Vest and Caldwell, 1972; 5) Devine, 1984.
2
3
4
5
126
Keyser and Li
The soil population density of indigenous
bradyrhizobia is a major factor determining competition for nodule occupancy and response to
inoculation. Using the estimate of a hectare of
6
plow-layer soil to weigh 2.24 x 10 kg, and assuming an indigenous bradyrhizobial population
of 104 cells g-1 soil and an inoculation rate of
10 g peat kg-1 seed (assuming an inoculum density of 108 cells g-1 peat), calculation shows that
at normal seeding rates the indigenous population has about a three hundred-fold advantage in
population density over the inoculum. Field trials have demonstrated that to achieve nodule
occupancy of greater than 50%, inoculant
bradyrhizobia must be applied at a rate at least
1,000 times greater than the estimated number of
indigenous bradyrhizobia (Weaver and Frederick, 1974). Use of massive inoculation rates
can overcome competition from indigenous
strains (Kapusta and Rouwenhorst, 1973), but
such a delivery system is not yet practical or
economical. Recent experiments from multi-site,
standardized field inoculation trials with several
legumes revealed that 59% of the variation in
inoculation responses could be accounted for by
the relationship of inoculation responses to numbers of indigenous rhizobia (Thies et al., 1991).
Despite the strong influence indigenous population density has on competition, it is a complex
problem with interactions involving the bacterial
genomes, host genomes and the environment
(Dowling and Broughton, 1986; Triplett, 1990a).
As yet, mechanisms responsible for competition
per se have not been identified. For instance, it
has been demonstrated that some factor(s) other
than comparative numbers in the rhizosphere is
(are) determining the outcome of competition
(Ellis et al., 1984; Moawad et al., 1984). Indeed,
the mechanism(s) may be so multifaceted that
their identification will be slow and piecemeal.
Some factors determined by genomes of both
partners have been identified, such as host range
(Cregan et al., 1989a; Sadowsky et al., 1987),
mobility of rhizobia (Catlow et al., 1990; Mellor
et al., 1987), and bacteriocin production (Hodgson et al., 1985; Schwinghamer and Brockwell,
1978; Triplett, 1990a). In the absence of indigenous bradyrhizobia, the pattern of competition
between inoculum strains was found to be a
stable (and therefore selectable) characteristic,
independent of rhizosphere population size, nitrogen application, elevation (temperature) or
soil type (Abaidoo et al., 1990; George et al.,
1987).
Symbiotic effectiveness seems to have no direct relationship with competitiveness (Triplett,
1990b). Cregan and Keyser (1988) showed that
host genotypes of Glycine max and G. soja have
great influence on competition and effectiveness,
both with B. japonicum and strains of Sinorhizobium (formerly R. fredii ). This host control
was exploited to select genotypes that would
restrict nodulation of the commonly occurring,
very competitive and yet relatively ineffective B.
japonicum serogroup 123, found in the midwestern U.S. soybean growing region. Such
genotypes were found which drastically altered
competition (compared to commercial cultivars)
in favor of inoculant strains over strains in the
123 serogroup (Cregan and Keyser, 1986; Keyser
and Cregan, 1987).
Research strategies to increase BNF in soybean
The need to devote research efforts towards
improvement of BNF in legumes in general and
soybean in particular has been recognized frequently (Anon., 1984; Olsen, 1982; Russell et
al., 1989; Scott and Aldrich, 1983). Areas of
research with good prospects for contributing to
this objective are discussed below.
Selection and engineering of bradyrhizobia
Soil microbiologists have extensive experience in
the selection of rhizobia and bradyrhizobia for
symbiotic effectiveness with legumes (Brockwell
et al., 1982; Burton, 1980; Date, 1976). To date,
the provision of superior strains for use in inoculants remains the primary applied contribution
from the field of BNF. In addition to selecting
strains which are highly effective and efficient at
nitrogen fixation with a given legume genotype,
Brockwell et al. (1982) list 10 other important
attributes; 1) competitive ability, 2) N2-fixing
ability over a range of environmental conditions,
3) nodulation and N2 fixation in the presence of
soil nitrogen, 4) ability to multiply in broth and
survive in inoculant carriers, 5) ability to survive
BNF in soybean
when incorporated in seed pellets, 6) persistence
in soil, 7) ability to migrate from initial site of
inoculation, 8) ability to colonize soil away from
influence of host roots, 9) ability to survive
adverse physical conditions such as desiccation,
heat, or freezing, and 10) strain-stability during
storage and growth. Selection of naturally occurring strains for these attributes continues to be
an important research activity in soybean programs worldwide, especially as production expands on to less desirable soils.
Competition for nodule occupancy is a complex problem which is of practical concern. In
response to this there have been methods developed to screen inoculant strains for their competitive ability. The most straightforward method is to assess the response of the legume to an
effective inoculant in the presence of ineffective
(unable to fix N2) competitors (Amarger, 1981;
Jones et al., 1978). The value of this method was
demonstrated in field trials of inoculant strains
competitive
against ineffective,
indigenous
rhizobia nodulating subclover ( Trifolium subterraneum L.). The inoculants produced large
responses by subterranean clover (Jones et al.,
1978). A more complicated method is that proposed by Paau (1989) that involves isolation of
indigenous strains, which are assumed to be well
adapted and competitive in their specific environment, subjecting them to mutagenesis, and
selecting mutants for increased N2-fixing ability
for use as inoculant. Multiple-site field tests of
these mutants in soils with indigenous B.
japonicum over a four-year period showed an
average soybean yield increase of 169 kg ha-1, or
6.7% (Paau. 1989).
Significant progress has been achieved in the
studies of genetics of rhizobia and bradyrhizobia
since the mid 1970s (Johnston et al., 1987).
Bradyrhizobium japonicum is becoming a genetically well-studied bacterium as far as nodulation
and symbiotic nitrogen fixation genes are concerned (Hennecke et al., 1988). Some of the first
attempts at genetic manipulation for practical
purposes focused on increasing the nitrogen fixation rate of the bacteria through mutagenesis
(Maier and Brill, 1978; Williams and Phillips,
1983). The mutant used by Williams and Phillips
produced a significant soybean yield increase
above that of the parent strain USDA 110 in
127
field trials. Introducing the hydrogen uptake
(hup) genes into bradyrhizobial strains has been
proposed as a manipulation to increase their
energy efficiency (Lim et al., 1980). The Eisbrenner and Evans (1983) review of this topic
showed that the majority of reports on the comparison of Hup+ and Hup- strains found a
significant response in total N with Hup+
strains. Evans et al. (1985) reported that soybean yields were higher when plants were nodulated by a strain of B. japonicum that was Hup+
as compared to plants nodulated by an isogenic
Hup- strain. However, there are Hup- strains
of B. japonicum that are equal or superior to
Hup+ strains in symbiotic performance in the
field (Hume and Shelp, 1990).
Genetic manipulation to alter competitive ability in the soybean bacteria has not been successful as yet. However, there is still a need to
pursue this approach. In a review of this subject,
Triplett (1990b) identifies bacterial phenotypes
found to play a role in competition, including
motility, cell-surface characteristics, speed of infection, and bacteriocin production. Given the
rapid rate of advances made in the field of
molecular genetics of rhizobia and bradyrhizobia, it seems quite likely that it will contribute to our understanding, and ability to manage,
competition for nodule occupancy.
Selection and breeding of soybean genotypes
On the host side, improving BNF has been approached by selection for improved nitrogen fixation per se (if indeed this can be separated from
yield), selection for ability to nodulate and fix
nitrogen in the presence of high soil-N levels, the
development of soybeans with the ability to restrict nodulation by selected indigenous populations and still nodulate with effective inoculant
strains, and the development of soybeans that
nodulate promiscuously with indigenous Bradyrhizobium spp. Several methods are available to
undertake these objectives, including selection of
existing germplasm, conventional plant breeding,
mutagenesis, and gene transfer (Dreyfus et al.,
1988).
Traditional legume breeding programs have
not included enhancement of BNF as a direct
objective (Phillips and DeJong, 1984). A review
128
Keyser and Li
of the reports on soybean, examining the relationship between total N fixed and yield, shows
them to be positively and highly correlated (Leffel, 1989). It appears that direct selection for
yield improvement in soybean has indirectly included improved capacity to fix N2. Further, the
potential for yield increase due directly to increased BNF exists (Burias and Planchon, 1990;
Imsande 1989). The one preliminary report that
shows differences between BNF in soybean
genotypes of equal yield and maturity is that of
Cregan et al. (1989b). They found that selected
high seed-protein genotypes attained higher total
N from higher N2 fixation compared to normal
genotypes.
Soybean genotypes have recently been identified which have greater numbers of nodules in
the presence of high mineral N (Betts and Herridge, 1987; Herridge et al., 1988). Similar
studies have been conducted on a smaller scale
(Danso et al., 1987; Gibson and Harper, 1985).
The benefit of this approach is that soybean
would fix more at a given level of mineral N,
thus saving or ‘sparing’ that soil N for subsequent crops. Mutagenesis of soybean has produced supernodulating mutants which are also
more nitrate-tolerant (Carroll et al., 1985; Gremaud and Harper, 1989). A mutant characterized as a moderate supernodulator has produced
increased yields in the field under some conditions (Boerma and Ashley, 1988; Carroll et al.,
1988), indicating the potential of this research to
increase yields through manipulation of a trait
related to BNF.
As mentioned above in the section on competition, soybean genotypes have been selected
which restrict nodulation with serogroup 123, an
indigenous,
heterogeneous
group
of
B.
japonicurn present in much of the northern midwest U.S.A. (Cregan and Keyser, 1986). The
objective is to permit a higher portion of the
nodules to be occupied by highly effective inoculant strains. This could lead to higher levels of
N, being fixed, or a greater portion of total N
from BNF, thereby sparing soil N. Generations
of backcrossing to productive, commercial cultivars will be necessary to adequately evaluate
the success of this approach in production fields.
However, preliminary data, summarized below
in Table 6, indicate that the restriction of the
Table 6. Nodule occupancy in selected soybean genotypes
a
Soybean
genotype
Percent of nodules occupied by
USDA 123
USDA 122 or
USDA 138
Other
Williams
PI 371607 b
PI 377578 b
75.7
3.0
5.0
20.5
88.7
91.5
3.8
8.3
3.5
Average of 2 years, field experiments.
Genotypes identified as restricting USDA 123 in greenhouse trials.
Source: Keyser, Cregan and El-Maksoud, unpublished data.
a
b
serogroup-type strain, USDA 123, is expressed
under field conditions where 123 was mixed into
the soil and inoculant was applied on the planted
seed.
Soybean genotypes have also been selected for
ability to nodulate with indigenous strains of B.
japonicum and for the ability to nodulate preferentially with the effective inoculant strain
USDA 110 (Kvien et al., 1981). Further development of this approach indicates that while
selection for preferential recovery of USDA 110
will be difficult, a positive relationship of nodule
mass to seed yield was found indicating that
further selection for increased nodulation with
native B. japonicum may be warranted (Greder
et al., 1986).
Selection and breeding of soybeans for a lack
of dependence on B. juponicurn has been carried
out by workers at the International Institute of
Tropical Agriculture (IITA) in Nigeria. Because
the production and distribution of inoculant in
parts of Africa is difficult, the researchers at
IITA have sought soybean lines which are promiscuous in their nodulation. These lines nodulate
with indigenous bacteria, presumably strains of
Bradyrhizobium spp. , which nodulate legumes in
the ‘cowpea’ cross-inoculation group, thus obviating the need to inoculate with B. juponicurn
(Bromfield and Roughley, 1980; Nangju, 1980;
Pulver et al., 1982; 1985). Selected promiscuous
lines bred to lines with superior agronomic traits
and compared for yield and nodulation with
(B. juponicurn requiring)
non-promiscuous
genotypes show that progress has been made
(Dashiell et al., 1985). However, there is some
concern over the potential of this system. Eaglesham (1985) indicated that many of the brady-
BNF in soybean
rhizobia which do nodulate these lines are relatively ineffective, and may be a subset of the
general population of Bradyrhizobium spp.
which could be quite variable in its representation throughout African soils. It is clear that
further yield increase through better BNF is
possible in some promiscuous lines (Pal, 1989).
Also, it needs to be clearly demonstrated that
there are not indigenous, true B. japonicum,
which have persisted from earlier studies, at sites
where these lines are evaluated. Such characterization of the microsymbiont population structure
in the soils where these lines are evaluated would
seem a necessary component to assist the evaluation and progress of this approach.
Improved inoculation techniques
Application of peat-based inoculant to the seed
just prior to planting is the most common form
of inoculation. This technique does not always
give abundant nodulation of soybean (Li et al.,
1986; Wadisirisuk et al., 1989). Evidence is accumulating that other techniques of inoculation
can provide better nodulation and plant growth
(Danso and Bowen, 1989; Hardarson et al.,
1989; Kamicker and Brill, 1987; McDermott and
Graham, 1989). Under conditions where soybean is a new crop, or where stress such as high
temperature is encountered, soybean may respond to the higher numbers of bacteria provided in granular or spray inoculation as compared to seed applied inocula (Bezdicek et al.,
1978; Scudder, 1974).
Hardarson et al. (1989) found that when inoculant was distributed throughout the soil it gave
profuse nodulation throughout the root system,
and nodules formed in the bottom part where
roots were younger and contributed large
amounts of fixed N to the soybean during seed
formation. Inoculation through irrigation water
at the third-node stage (V3) produced nodules
that were very active in N2 fixation during the
reproductive stage, resulting in an increase in
seed yield and seed protein (Ciafardini and Barbieri, 1987).
Burton (1980) reviewed the need for improved
inoculant-delivery systems. The importance of
delivering large numbers of bradyrhizobia is a
challenge, and the best systems identified to date
129
are the soil-applied granular and seed bedsprayed inoculants. Higher numbers of inoculant
rhizobia will be required if soybean production
moves onto marginal soil that is acidic, has high
temperatures at the time of planting, or has
other stresses which affect inoculant viability.
Also, as the levels of agricultural inputs increase
in developing-country agroecosystems, inoculant
compatibility with pesticides and insecticides will
need to be addressed.
Production strategies to increase BNF in
soybean
A century of research on BNF has brought tremendous progress in both our basic understanding of the process and in its application for
improving legume growth (Nutman, 1987; Quispel, 1988). In recent years the majority of research on BNF has focused on bacterial genetics
and biochemistry, with elaborate and sophisticated knowledge accumulating at an impressive
rate. Gene transfer systems for plants are also
being developed. Developments in these areas
may be applicable in the future for enhancing
BNF in soybean. Until then, such improvement
will probably come from those strategies that
were established many years ago, and while they
are proven approaches, their use is far from
ubiquitous and they probably provide the best
potential for improvements at the field level.
Inoculant production, quality control and
training
The greatest impacts on world agriculture from
inoculation have been made with soybean. While
there may not be large new areas planted to
soybean on a global scale, there are still going to
be regional and local needs for inoculation that
will produce economic returns and improve
BNF. In a global survey of inoculant use and
availability, Eaglesham (1989a) found that inoculant is still perceived as being needed in many
cases, especially where it is not available. Concerted efforts are needed to deliver and adapt
existing inoculant technology to local conditions,
especially in developing countries.
Many countries have no mandatory quality-
130
Keyser and Li
control standards for legume inoculant. The result can be products that are highly variable in
quality as determined by symbiotic effectiveness,
population density and shelf-life. The production
of high-quality inoculant and proper quality control cannot be addressed here (see Roughley,
1976; Roughley and Pulsford, 1982; Thompson,
1984), but their importance is critical to the
improvement of BNF. An essential part of the
adoption of this technology is its adaptation to
the needs of individual or highly localized situations (Hubbell, 1988). In many developing countries this means improvising with the non-traditional materials and equipment available to
achieve economic production of a satisfactory
product.
Education about the benefits and use of
legume inoculants can still play a significant role
in improving BNF in many countries. Many farmers may not know of the existence of inoculant,
especially if the crop is new, as soybean often is,
and the agricultural extension agents may only
have a limited understanding of the technology.
Having perceived the need for such practical
training, NifTAL (University of Hawaii U.S.A.I.D.) developed a BNF training course
for extension specialists. This course was presented to Indonesian and Ugandan government
and private-agency extension personnel, and was
met with tremendous enthusiasm. The course
was designed to give the extension specialists
sufficient applied experience and theoretical
knowledge to set up mini-courses on BNF and
inoculation for their district farmers. Such training should be viewed as part of the overall
inoculant technology transfer to other countries.
Matching soybean genotypes to the environment
There is tremendous germplasm diversity in soybean, and using a genotype well adapted to a
given site is probably one of the best and simplest strategies for improving BNF, through improving yield. This of course assumes that the
soybean is well nodulated with effective bradyrhizobia. Some excellent data are found in
INTSOY’s International Soybean Variety Experiment (ISVEX) verifying that the varieties
may differ in their yield performance in a given
environment, and even at different sites in the
same environmental zone (Jackobs et al., 1985;
Judy and Whigham, 1978). This world-wide variety evaluation program provides a wealth of
information concerning environmental influences
on several agronomic traits of soybean, and identifies important relationships which can assist
plant breeders. A standard granular inoculant
was provided to network cooperators for use in
ISVEX. Any breeding program should include
cooperation with a microbiologist to ensure that
strains of bradyrhizobia are identified which are
compatible with a given soybean genotype in a
particular environment.
Management of other inputs
As with selecting the best genotype for yield,
other management variables that increase yield
should also increase the amount of N2 fixed. As
emphasized by Eaglesham (1989a) N is not always the primary limiting factor, and when it is
not there will not be a response to inoculation.
Other factors which limit soybean yield will then
by definition also limit inoculation and N response.
Phosphorus is also a common limiting nutrient
in many soils, and its management is important
for attaining high yields of soybean. Cassman et
al. (1981) found that field-grown soybean has a
higher P requirement when it is dependent on
BNF for its N supply as compared to mineral N
dependency. In this study, soybean dependent
on BNF but not supplied with P attained only
28% of the maximum yield obtained at optimum
P levels.
Further evidence of the importance of other
management factors such as P is supplied below
in Table 7. These are results from 140 on-farm
demonstration plots established in four districts
of Uganda during 1989. The addition of P and
inoculation improved yields. Averaged over all
trials, each input gave approximately 300 kg
yield increase, and applied together a 600 kg
increase. The return on investment from the
combined inputs was about 12-fold.
Supplying an adequate symbiosis to the soybean provides it only with a biological source of
N. It is not a panacea against poor growth
resulting from limitations of water, lime, nutrients, disease or any other factors. This per-
131
BNF in soybean
Table 7. Response of soybeans in Uganda, 1989 on-farm demonstration trials
District
Average yield of soybeans (kg ha-1)
Local variety
in rows
Masindi
Mubende
Luwero
Kasese
920
738
1,070
843
LSD 5%
Improved variety
in hills
in hills + P a
in hills + P a
+ inoculum
750
588
1,050
772
1,080
899
1,450
1,021
1,290
1.260
1.790
1.309
107
102
253
357
45 kg P2O5 ha -1.
Source: Simkins, Kalule and Baguma, unpublished data.
a
spective is necessary for evaluating the potential
for improving BNF in soybean in a given
situation.
Concluding remarks
The world soybean production is continuing to
increase, and the prospects for future increases
in sustainable systems will likely come mostly
from improvements in yield per area. Research
on BNF in the soybean-bradyrhizobia symbiosis
will have a role in contributing to such improvements. Manipulations of both the bacteria and
the host plant for attributes related to BNF
continue to be impressive. Still, there is much of
the proven technology of BNF that has not yet
been applied in cases where it could make immediate improvements in soybean yields. Research
agencies, national planning bodies and international development agencies need to place proper emphasis on both research and application.
References
Abaidoo R C, George T, Bohlool B B and Singleton P W
1990 Influence of elevation and applied nitrogen on rhizosphere colonization and competition for nodule occupancy
by different rhizobia1 strains on field-grown soybean and
common bean. Can. J. Microbiol. 36, 93-96.
Al-Ithawi B, Deibert E J and Olson R A 1980 Applied N and
moisture level effects on yield depth of root activity and
nutrient uptake by soybeans. Agron. J. 72, 827-832.
Anonymous 1984 U.S. soybean production and utilization
research. A report to the Senate Committee on Agriculture, Nutrition and Forestry and House Committee on
Agriculture. Soybean Research Advisory Institute,
Washington, DC. 69 p.
Amarger N 1981 Selection of Rhizobium strains on their
competitive ability for nodulation. Soil Biol. Biochem. 13,
481-486.
Bergersen F S 1958 The bacterial component of soybean root
nodule: Changes in respiratory activity, cell dry weight,
and nucleic acid content with increasing nodule age. J.
Gen. Microbiol. 19, 312-323.
Betts J H and Herridge D F 1987 Isolation of soybean lines
capable of nodulation and nitrogen fixation under high
levels of nitrate supply. Crop Sci. 27. 1156-1161.
Bezdicek D F. Evans D W. Adebe B and Witters R E 1978
Evaluation of peat and granular inoculum for soybean
yield and N fixation under irrigation. Agron. J. 70, 865868.
Boerma H R and Ashley D A 1988 Performance of three
soybean mutants with increased nodulation. Agron. Abstracts. p 104.
Brevedan R E, Egli D B and Leggett J E 1978 Influence of N
nutrition on flower and pod abortion and yield of soybeans. Agron. J. 70. 81-84.
Brockwell J. Diatloff A. Roughley R J and Date R A 1982
Selection of rhizobia for inoculants. In Nitrogen Fixation in
Legumes. Ed. J M Vincent. pp 173-191. Academic Press,
Sydney.
Bromfield E S P and Roughley R J 1980 Characterization of
rhizobia isolated from nodules on locally-adapted Glycine
max grown in Nigeria. Ann. Appl. Bid. 95, 185-190.
Burias N and Planchon C 1990 Increasing soybean productivity through selection for nitrogen fixation. Agron. J. 82,
1031-1034.
Burton J C 1980 Rhizobium inoculation and soybean production. In World Soybean Research Conference. part II:
Proceedings. Ed. F T Corbin. pp 89-100. Westview Press.
Boulder. CO.
Buttery B R and Dirks V A 1987 The effects of soybean
cultivar. rhizobium strain and nitrate on plant growth,
nodule mass and acetylene reduction rate. Plant and Soil
98, 285-293.
Caldwell B E 1966 Inheritance of a strain-specific ineffective
nodulation in soybeans. Crop Sci. 6, 427-428.
Carroll B J. McNeil D L and Gresshoff P M 1985 A
supernodulation and nitrate-tolerant symbiotic (nts) soybean mutant. Plant Physiol. 78, 34-40.
Carroll B J, Hartley R, Ashley D A, Boerma H R, Lawn R,
132
Keyser and Li
Byth D and Greshoff P 1988 Field evaluation of supernodulating soybean mutants. In Nitrogen Fixation: Hundred Years After. Eds. H Bothe, F J de Bruijn and W E
Newton. p 808. Gustav Fischer, Stuttgart, Germany.
Cassman K G, Whitney A S and Fox R L 1981 Phosphorus
requirements of soybean and cowpea as affected by mode
of N nutrition. Agron. J. 73, 17-22.
Cassman K C 1990 The role of soil fertility research in
developing sustainable food production systems. Better
Crops With Plant Food 74, 16-19.
Catlow H Y, Glenn A R and Dilworth M J 1990 Does
rhizobia1 motility affect its ability to colonize along the
legume root? Soil Biol. Biochem. 22 573-575.
Ciafardini G and Barbieri C 1987 Effects of cover inoculation
of soybean on nodulation, nitrogen fixation, and yield.
Agron. J. 79, 645-648.
Clark F E 1957 Nodulation responses of two near-isogenic
lines of the soybean. Can J. Microbiol. 3, 113-123.
Cregan P B and Keyser H H 1986 Host restriction of
nodulation by Bradyrhizobium japonicum strain USDA
123 in soybean. Crop Sci. 26, 911-916.
Cregan P B and Keyser H H 1988 Influence of Glycine spp.
on competitiveness of Bradyrhizobium japonicum and
Rhizobium fredii. Appl. Environ. Microbiol. 54, 803-808.
Cregan P B, Keyser H H and Sadowsky M J 1989a Soybean
genotype restricting nodulation of a previously unrestricted
serocluster 123 bradyrhizobia. Crop Sci. 29, 307-312.
Cregan P B, Leffel R C and A P Bolgiano 1989b Nitrogen
metabolism of high and low seed protein soybean
genotypes. Agron. Abstracts. p 110.
Danso S K A, Hera C and Douka C 1987 Nitrogen fixation in
soybean as influenced by cultivar and Rhizobium strain.
Plant and Soil 99, 163-174.
Danso S K A and Bowen G D 1989 Methods of inoculation
and how they influence nodulation pattern and nitrogen
fixation using two contrasting strains of Bradyrhizobium
japonicum. Soil Biol. Biochem. 21, 1053-1058.
Dashiell K E, Kueneman E A, Root W R and Singh S R
1985 Breeding tropical soybean for superior seed longevity
and for nodulation with indigenous rhizobia. In Soybean in
Tropical and Subtropical Cropping Systems. Eds. S Shanmugasundaram and E W Sulzberger. pp 133-139. Asian
Vegetable Research and Development Center, Shanhua,
Taiwan.
Date R A 1976 Principles of Rhizobium strain selection. In
Symbiotic Nitrogen Fixation in Plants, International
Biological Programme 7. Ed. P S Nutman. pp 137-150.
Cambridge University Press, Cambridge.
Devine T E and Weber D F 1977 Genetic specificity of
nodulation. Euphytica 26, 527-535.
Devine T E 1984 Inheritance of soybean nodulation response
with a fast-growing strain of rhizobia. J. Heredity 75,
359-361.
Dowling D N and Broughton W J 1986 Competition for
nodulation of legumes. Annu. Rev. Microbiol. 40, 131157.
Dreyfus B L, Diem H G and Dommergues Y R 1988 Future
directions for biological nitrogen fixation research. Plant
and Soil 108, 191-199.
Eaglesham A R J, Ayanaba A, Rao V and Eskew D L 1982
Mineral N effects on cowpea and soybean crop in a
Nigerian Soil. II. Amount of N fixed and accrued in the
soil. Plant and Soil 68, 183-192.
Eaglesham A R J, Hassouna S and Seeger R 1983 FertilizerN effects on N2 fixation by cowpea and soybean. Agron. J.
75, 61-66.
Eaglesham A R J 1985 Comparison of nodulation promiscuity of US- and Asian-type soya beans. Trop. Agric. (Trinidad) 62, 105-109.
Eaglesham A R J 1989a Global importance of Rhizobium as
an inoculant. In Microbial Inoculation of Crop Plants. Eds.
R Cambell and R M MacDonald. pp 29-48. Oxford University Press, Oxford.
Eaglesham A R J 1989b Nitrate inhibition of root-nodule
symbiosis in doubly rooted soybean plants. Crop Sci. 29,
115-119.
Eisbreinner G and Evans H J 1983 Aspects of hydrogenase
metabolism in nitrogen-fixing legumes and other plantmicrobe associations. Annu. Rev. Plant Physiol. 34, 103136.
Ellis W R, Ham G E and Schmidt E L 1984 Persistence and
recovery of Rhizobium japonicum inoculum in a field soil.
Agron. J. 76, 573-576.
Evans H J, Emerich D W, Ruiz-Argueso T, Maier R J and
Albrecht S L 1980 Hydrogen metabolism in the legumeRhizobium symbiosis. In Nitrogen Fixation, Vol. 2. Eds. W
H Orme-Johnson and W E Newton. pp 67-86. University
Park Press, Baltimore, MD.
Evans H J, Hanus F J, Haugland R A, Cantrell M A, Xu L,
Russell S A, Lamber G A and Harker A R 1985 Hydrogen
recycling in nodules affects nitrogen fixation and growth of
soybeans. In World Soybean Research Conference III:
Proceedings. Ed. R Schibles. pp 935-942. Westview Press
Boulder, CO.
Fehr W R, Caviness C E, Burmood D T and Pennington J S
1971 Stages of development descriptions for soybean,
Glycine max (L.) Merr. Crop Sci. 11, 929-930.
Fehr W R 1987 Breeding methods for cultivar development.
In Soybeans: Improvement, Production and Uses. Second
Edition. Ed. J R Wilcox. pp 249-293. American Society of
Agronomy. Madison, WI.
Gao J, Wang Q. Hao Z, Zhang H, Zhao G, Zhang G, Wang
X and Xue B 1987 Study on the symbiotic nitrogen fixation
of soybean by 15N. Soybean Sci. 6, 55-61.
Ge C and Xu L 1982 Observation of the infection behaviour
of indigenous Rhizobium japonicum and its distribution in
different varieties of soybean in the fields. Chinese Oil
Crops 3. 56-58 (In Chinese).
George T. Bohlool B B and Singleton P W 1987 Bradyrhizobium japonicum - environment interactions: Nodulation and interstrain competition in soils along an elevational transect. Appl. Environ. Microbiol. 53, 1113-1117.
George T, Singleton P W and Bohlool B B 1988 Yield, soil
nitrogen uptake, and nitrogen fixation by soybean from
four maturity groups grown at three elevations. Agron. J.
80, 563-567.
Gibson A H and Harper J E 1985 Nitrate effect on nodulation of soybean by Bradyrhizobium japonicum. Crop Sci.
25, 497-501.
Greder R R, Orf J H and Lambert J W 1986 Heritabilities
BNF in soybean
and associations of nodule mass and recovery of Bradyrhizobium japonicum serogroup USDA 110 in soybean.
Crop Sci. 26. 33-37.
Gremaud M F and Harper J E 1989 Selection and initial
characterization of partially nitrate tolerant nodulation mutants of soybean. Plant Physiol. 89, 169-173.
Guffy R D, Heuvel R M V, Vasilas B L, Nelson R L, Frobish
M A and Hesketh J D 1989 Evaluation of the N2 fixation
capacity of four genotypes by several methods. Soil Biol.
Biochem. 21, 339-342.
Ham G E 1976 Competition among strains of rhizobia. In
World Soybean Research. Ed. L D Hill. pp 144-150.
lnterstate Printers and Publishers. Danville, IL.
Ham G E 1980 Inoculation of legumes with Rhizobium in
competition with naturalized strains. In Nitrogen Fixation,
Vol. 2. Eds. W E Newton and W H Orme-Johnson. pp
131-138. University Park Press, Baltimore, MD.
Hardarson G, Golbs M and Danso S K A 1989 Nitrogen
fixation in soybean ( Glycine max L. Merrill) as affected by
nodulation patterns. Soil Biol. Biochem. 21, 783-787.
Hardy R W F, Burns R C and Holsten R D 1973 Applications of the acetylene ethylene assay for measurement of
nitrogen fixation. Soil Biol. Biochem. 5. 47-81.
Hardy R W F, and Havelka U D 1976 Phytosynthate as a
major factor limiting nitrogen fixation by field-grown
legumes with emphasis on soybean. In Symbiotic Nitrogen
Fixation In Plants. Ed. P S Nutman. pp 421-439. Cambridge University Press, Cambridge.
Harper J E 1987 Nitrogen metabolism. In Soybeans: Improvement, Production, and Uses. 2nd edition. Ed. J R
Wilcox. pp 497-533. Am. Soc. Agron.. Madison, WI.
Havelka U D, Boyle M G and Hardy R W F 1982 Biological
nitrogen fixation. In Nitrogen in Agricultural Soils. Ed. J
Stevenson. pp 365-422. American Society of Agronomy,
Madison, WI.
Hennecke H, Meyer L, Gottfert M and Fischer H 1988
Genetics of the Bradyrhizobium japonicum - soybean symbiosis: Recent developments on genes for nodulation. bacteroid respiration, and regulation of nitrogen fixation. In
Molecular Genetics of Plant-Microbe Interactions. Eds. R
Palacios and D P S Verma. pp 118-123. American
Phytopathological Society, St. Paul. MN.
Herridge D F and Holland J F 1987 Effects of tillage on plant
available nitrogen and N2 fixation by soybean. In Nitrogen
Cycling in Agricultural Systems of Temperate Australia.
Eds. P E Bacon, J Evans, P R Storrier and A R Taylor. pp
390-396. Aust. Soc. Soil Sci. Inc., Wagga Wagga, Australia.
Herridge D F and Bergersen F J 1988 Symbiotic nitrogen
fixation. In Advances in Nitrogen Cycling in Agricultural
Ecosystems. Ed. J R Wilson. pp 46-65. C.A.B. International, Wallingford, UK.
Herridge D F, Betts J H and Rose I A 1988 Breeding for
improved nodulation and nitrogen fixation by soybean. In
Proc. 5th Austr. Soybean Research Workshop. pp 75-77.
NSW Agriculture & Fisheries, Agricultural Research Centre, Tamworth, Australia.
Hinson K 1975 Nodulation responses from nitrogen applied
to soybean half-root systems. Agron. J. 67, 799-804.
Hodgson A L M, Roberts W P and Waid J S 1985 Regulated
133
nodulation of Trifolium subterraneum inoculated with bacteriocin producing strains of Rhizobium trifolii. Soil Biol.
Biochem. 17, 475-478.
Hubbell D H 1988 Extension/transfer of BNF technology. In
Nitrogen Fixation by Legumes in Mediterranean Agriculture. Eds. D P Beck and L A Materon. pp 367-370.
Kluwer Academic Publishers, Dordrecht, The Netherlands.
Hume D J. Shanmugasundaram S and Beversdorf W D 1985
Soybean ( Glycine max L. Merrill). In Grain Legume
Crops. Eds. R J Summerfield and E H Roberts. pp 391432. Collins Professional and Technical Books, London.
Hume D J and Shelp B J 1990 Superior performance of the
Hup- Bradyrhizobium japonicum strain 532C in Ontario
soybean field trials. Can. J. Plant Sci. 70, 661-666.
Hymowitz T 1970 On the domestication of the soybean.
Econ. Bot. 24. 408-421.
Imsande J 1986 Inhibition of nodule development in soybean
by nitrate or reduced nitrogen. J. Exp. Bot. 37, 348-355.
Imsande J 1989 Rapid dinitrogen fixation during soybean pod
fill enhances net photosynthetic output and seed yield: A
new perspective. Agron. J. 81, 549-556.
Jackobs J A, Smyth C A and Erickson D 1985 International
Soybean Variety Experiment. Tenth Report of Results
1983. INTSOY Series Number 16, Univ. of Illinois, Urbana-Champaign. 113 p.
Johnson H W and Clark F E 1958 Role of the root nodule in
the bacterial-induced chlorosis of soybeans. Soil Sci. Soc.
Am. Proc. 22. 527-528.
Johnston A W B, Downie J A, Rossed L, Shearman C A,
Firmin J L. Borthakur D, Wood E A, Bradley D and
Brewin N J 1987 Molecular analysis of the Rhizobium
genes involved in the induction of nitrogen-fixing nodules
on legumes. Phil. Trans. R. Soc. Lond. B317, 193-207.
Jones M B, Burton J C and Vaughn C E 1978 Role of
inoculation in establishing subclover on California annual
grasslands. Agron. J. 70, 1081-1085.
Judy W H and D K Whigham 1978 International Soybean
Variety Experiment. Fourth Report of Results 1976.
INTSOY Series Number 16, Univ. of Illinois, UrbanaChampaign, 401 p.
Kamicker B J and Brill W J 1987 Methods to alter the
recovery and nodule location of Bradyrhizobium
japonicum inoculant strains on field-grown soybeans. Appl.
Environ. Microbiol. 53, 1737-1742.
Kapusta G and Rouwenhorst D L 1973 Influence of inoculum
size on Rhizobium japonicum serogroup distribution frequency in soybean nodules. Agron. J. 65, 916-919.
Keyser H H and Cregan P B 1987 Nodulation and competition for nodulation of selected soybean genotypes among
Bradyrhizobium japonicum serogroup 123 isolates. Appl.
Environ. Microbiol. 53, 2631-2635.
Kosslak R M and Bohlool B B 1984 Suppressions of nodule
development of one side of a split-root system of soybeans
caused by prior inoculation of the other side. Plant Physiol.
75, 125-130.
Kvien C S, Ham G E and Lambert J W 1981 Recovery of
introduced Rhizobium japonicum strains by soybean
genotypes. Agron. J. 73, 900-905.
La Favre J S and Eaglesham A R J 1986 Rhizobitoxine: A
134
Keyser and Li
phytotoxin of unknown function which is commonly produced by bradyrhizobia. Plant and Soil 92, 443-452.
Larson W E 1986 Presidential Address: The adequacy of
world soil resources. Agron. J. 78, 221-225.
LaRue T A and Patterson T G 1981 How much nitrogen do
legumes fix? Adv. Agron. 34, 15-38.
Latimore M J, Giddens J and Ashley D A 1977 Effect of
ammonium and nitrate nitrogen upon photosynthate supply and nitrogen fixation by soybeans. Crop Sci. 17, 399404.
Lawn R J and Burn W A 1974 Symbiotic nitrogen fixation in
soybeans. I. Effect of photosynthetic source-sink manipulations. Crop Sci. 14, 11-16.
Lawn R J and Byth D E 1989 Saturated soil culture - a
technology to expand the adaptation of soybean. In Proceedings World Soybean Res. Conf. IV. Ed. A J Pascale.
pp 576-581. Asociacion Argentina de la Soja, Buenos
Aires, Argentina.
Leffel R C 1989 Breeding soybean for enhanced nitrogen
metabolism. In Proceedings World Soybean Res. Conf. IV.
Ed. A J Pascale. pp 1125-1130. Asociacion Argentina de
la Soja, Buenos Aires, Argentina.
Li F, Chen H and Wang F 1986 Rhizobium numbers and
their distribution in soil in relation to nodule occupancy. In
Trans. 13th. Congr. ISSS, Vol. 2. pp 602-603.
Lim S T, Andersen K, Tait R and Valentine R C 1980
Genetic engineering in agriculture: Hydrogen uptake (hup)
genes. Trends Biochem. Sci. 5, 167-170.
Maier R J and Brill W J 1978 Mutant strains of Rhizobium
japonicum with increased ability to fix nitrogen for soybean. Science 201, 448-450.
Malik N S A, Calvert H E and Bauer W D 1987 Nitrate
induced regulation of nodule formation in soybean. Plant
Physiol. 84, 266-271.
McDermott T R and Graham P H 1989 Bradyrhizobium
japonicum inoculant mobility, nodule occupancy, and
acetylene reduction in the soybean root system. Appl.
Environ. Microbiol. 55, 2493-2498.
Mellor H Y, Glenn A R, Arwas R and Dilworth M 1987
Symbiotic and competitive properties of motility mutants
of Rhizobium trifolii TA1. Arch. Microbiol. 148, 34-39.
Mengel D B, Segars W, and Rehm G W 1987 Soil fertility
and liming. In Soybeans: Improvement, Production and
Uses. Second Edition. Ed. J R Wilcox. pp 461-496.
American Society of Agronomy, Madison, WI.
Millhollon E P and Williams L E 1986 Carbohydrate partitioning and the capacity of apparent nitrogen fixation of
soybean plants grown outdoor. Plant Physiol. 81,280-284.
Moawad H A, Ellis W R and Schmidt E L 1984 Rhizosphere
response as a factor in competition among three serogroups of indigenous Rhizobium japonicum for nodulation
of field-grown soybeans. Appl. Environ. Microbiol. 47,
607-612.
Nangju D 1980 Soybean response to indigenous rhizobia as
influenced by cultivar origin. Agron. J. 72, 403-406.
Nutman P S 1987 Centenary lecture. Phil. Trans. R. Soc.
Lond. B317, 69-106.
Olsen S R 1982 Presidential Address: Removing barriers to
crop productivity. Agron. J. 74, 1-4.
Paau A S 1989 Improvement of Rhizobium inoculants. Appl.
Environ. Microbiol. 55, 762-865.
Pal U R 1989 Comparative contributions of native rhizobia
VS. strains of Rhizobium japonicum to nitrogen uptake of
promiscuous soybeans in Nigerian savanna. In Proceedings
World Soybean Res. Conf. IV. Ed. A J Pascale. pp 494499. Asociacion Argentina de la Soja, Buenos Aires,
Argentina.
Patterson T G and LaRue T A 1983 N2 fixation (C2H2) and
ureide content of soybeans: Environmental effects and
source-sink manipulations. Crop Sci. 23, 819-824.
Peoples M B, Faizah A W, Rerkasem B and Herridge D F
1989 Methods for Evaluating Nitrogen Fixation by Nodulated Legumes in the Field. Australian Centre for International Agricultural Research, Canberra. 76 p.
Phillips D A and DeJong T M 1984 Dinitrogen fixation in
leguminous crop plants. In Nitrogen in Crop Production.
Ed. R D Hauck. pp 121-132. American Society of Agronomy, Madison, WI.
Pulver E L, Brockman F and Wien H C 1982 Nodulation of
soybean cultivars with Rhizobium spp. and their response
to inoculation with R. japonicum. Crop Sci. 22, 1065-1070.
Pulver E L, Kueneman E A and Ranga-Rao V 1985 Identification of promiscuous nodulating soybean efficient in N2
fixation. Crop Sci. 25, 660-663.
Quispel A 1988 Hellriegel and Wilfarth’s discovery of (symbiotic) nitrogen fixation hundred years ago. In Nitrogen
Fixation: Hundred Years After. Eds. H Bothe F J de
Bruijn and W E Newton. pp 3-12. Gustav Fisher, Stuttgart, Germany.
Rennie R J 1984 Comparisons of N balance and 15N isotope
dilution to quantify N2 fixation in field-grown legumes.
Agron. J. 76, 785-790.
Roughley R J 1976 The production of high quality inoculants
and their contribution to legume yield. In Symbiotic Nitrogen Fixation in Plants, International Biological Programme
7. Ed. P S Nutman. pp 125-136. Cambridge University
Press, Cambridge.
Roughley R J and Pulsford D J 1982 Production and control
of legume inoculants. In Nitrogen Fixation in Legumes.
Ed. J M Vincent. pp 193-209. Academic Press, Sydney.
Russell J S, Beech D F and Jones P N 1989 Grain legume
productivity in subsistence agriculture. Food Policy May,
1989. 129-141.
Sadowsky M J. Tully R E, Cregan P B and Keyser H H 1987
Genetic diversity in Bradyrhizobium japonicum serogroup
123 and its relation to genotype-specific nodulation of
soybean. Appl. Environ. Microbiol. 53, 2624-2630.
Schwinghamer E A and Brockwell J 1978 Competitive advantage of bacteriocin- and phage-producing strains of
Rhizobium trifolii in mixed culture. Soil Biol. Biochem. 10,
383-387.
Scott W 0 and Aldrich S R 1983 Modern Soybean Production. S & A Publication, Champaign, IL. 209 p.
Scudder W T 1974 Rhizobium inoculation of soybeans for
sub-tropical and tropical soils. I. Initial field trials. Soil
Crop Sci. SOC. Fla. Proc. 34, 79-82.
Shanmugasundaram S 1989 Global cooperation for the improvement of soybean research and development. In Proceedings World Soybean Res. Conf. IV. Ed. A J Pascale.
pp 1939-1947. Asociacion Argentina de la Soja, Buenos
Aires, Argentina.
Sinclair T R and Wit C T de 1975 Photosynthate and nitrogen
BNF in soybean
requirements for seed production by various crops. Science
189, 565-567.
Sinclair T R 1989 Simultaneous limitation to soybean yield
increase by carbon and nitrogen. In Proceedings World
Soybean Res. Conf. IV. Ed. A J Pascale. pp 183-188.
Asociacion Argentina de la Soja. Buenos Aires.
Argentina.
Singleton P W and Stockinger K R 1983 Compensation
against ineffective nodulation in soybean. Crop Sci. 23.
69-12.
Smith K J and Huyser W 1987 World distribution and
significance of soybean. In Soybean: Improvement. Production and Use. 2nd edition. Ed. J R Wilcox. pp 1-22.
Am. Soc. Agron., Madison, WI.
Sprent J I and Minchin F R 1983 Environmental factors on
the physiology of nodulation and nitrogen fixation. In
Temperate Legumes: Physiology, Genetics and Nodulation. Eds. D G Jones and D R Davies. pp 269-317. Pitman
Advance Publishing Program, Boston.
Tang S 1979 Study on the nodulation and nitrogen fixation of
soybean in lessive soils. Acta Pediologica Sinica 16, 9-16.
Thibodeau P S and Jaworski E G 1975 Patterns of nitrogen
utilization in the soybean. Planta 127, 133-147.
Thies J E, Singleton P W and Bohlool B B 1991 Modeling
symbiotic performance of introduced rhizobia in the field
by use of indices of indigenous population size and nitrogen status of the soil. Appl. Environ. Microbiol. 57. 29-37.
Thompson J A 1984 Production and Quality Control of
Carrier-based Legume Inoculants. ICRISAT. Andhra
Pradesh, India. 37 p.
Triplett E W 1990a Construction of a symbiotically effective
strain of Rhizobium leguminosarum bv. trifolii with increased nodulation competitiveness. Appl. Environ. Microbiol. 56, 98-103.
Triplett E W 1990b The molecular genetics of nodulation
competitiveness in Rhizobium and Bradyrhizobium.
Molecular Plant-Microbe Interactions 3, 199-206.
Troedson R J 1988 Physiological aspects of the acclimation
135
and growth of soybean ( Glycine max (L.) Merrill) in
saturated soil culture. PhD thesis, Univ. of Queensland.
Vance C P. Egli M A, Griffith S M and Miller S S 1988 Plant
regulated aspects of nodulation and N2 fixation. Plant, Cell
Environ. 11, 413-427.
Vest G 1970 Rj3 - A gene conditioning ineffective nodulation
in soybean. Crop Sci. 10. 34-35.
Vest G and Caldwell B E 1972 Rj4 - A gene conditioning
ineffective nodulation in soybean. Crop Sci. 12, 692-693.
Vest G, Weber D F and Sloger C 1973 Nodulation and
nitrogen fixation. In Soybean: Improvement, Production,
and Uses. Eds. B E Caldwell. R W Howell, R W Judd and
H W Johnson. pp 353-390. American Society of Agronomy, Madison. WI.
Vincent J M 1980 Factors controlling the legume- Rhizobium
symbiosis. In Nitrogen Fixation. Vol. 2. Eds. W E Newton
and W H Orme-Johnson. pp 103-129. University Park
Press, Baltimore, MD.
Wadisirisuk P, Danso S K A, Hardarson G and Bowen G D
1989 Influence of Bradyrhizobium japonicum location and
movement on nodulation and nitrogen fixation in soybeans. Appl. Environ. Microbiol. 55, 1711-1716.
Weaver R W and Frederick L R 1974 Effect of inoculum rate
on competitive nodulation of Glycine max L. Merrill. II
Field studies. Agron. J. 66, 233-236.
Williams L E and Phillips D A 1983 Increased soybean
productivity with a Rhizobium japonicum mutant. Crop
Sci. 23, 246-250.
Williams L F and Lynch D L 1954 Inheritance of a nonnodulating character in the soybean. Agron. J. 46, 28-29.
Zapata F. Danso S K A, Hardarson G and Fried M 1987
Time course of nitrogen fixation in field-grown soybean
using nitrogen-15 methodology. Agron. J. 79, 172-176.
Zhang H, Zhang G, Zhao G. Wang X, Xu B and Zhao F
1986 Nitrogenase activity, nodulation and the N2 fixation
of the indigenous Rhizobium japonicum. Soybean Sci. 5,
47-56.
Plant and Soil 141: 137-153, 1992.
© 1992 Kluwer Academic Publishers. Printed in the Netherlands.
PLSO SA08
Biological nitrogen fixation in mixed legume /grass pastures
S.F. LEDGARD and K.W. STEELE
Ruakura Agricultural Centre, MAF Technology, Private Bag, Hamilton, New Zealand and MAF
Technology, Private Bag, Wellington, New Zealand
Key words: competition with grass, legume, methodology, mixed pasture, nitrogen fixation,
nitrogen transfer, review, soil nitrogen
Table of contents
Abstract
Introduction
The legume/Rhizobium symbiosis
Nodule development
Inoculation
Host plant/Rhizobium strain interaction
Rhizobial denitrification
Methods for measuring BNF in the field
Acetylene reduction assay (ARA)
Total legume nitrogen
15
N isotope dilution
15
N natural abundance
Factors affecting BNF
Effect of soil nitrogen status
Changes with pasture development
Fluctuations during the year
Spatial variability due to animal excreta
Effect of legume persistence and production
Soil moisture status
Soil acidity
Nutrition
Pests and disease
Competition with the associated grass
Associated grass species or cultivar
Grazing animal
Grazing or cutting management
Relationship between legume production and BNF
Transfer of fixed nitrogen to associated grasses
Conclusions
References
138
Ledgard and Steele
Abstract
Biological nitrogen fixation (BNF) in mixed legume/grass pastures is reviewed along with the
importance of transfer of fixed nitrogen (N) to associated grasses. Estimates of BNF depend on the
method of measurement and some of the advantages and limitations of the main methods are outlined.
The amounts of N fixed from atmospheric N2 in legume/grass pastures throughout the world is
summarised and range from 13 to 682 kg N ha-1 yr-1. The corresponding range for grazed pastures,
which have been assessed for white clover pastures only, is 55 to 296 kg N ha-1 yr-1.
Biological nitrogen fixation by legumes in mixed pastures is influenced by three primary factors;
legume persistence and production, soil N status, and competition with the associated grass(es). These
factors and the interactions between them are discussed. Legume persistence, production and BNF is
also influenced by many factors and this review centres on the important effects of soil moisture status,
soil acidity, nutrition, and pests and disease.
Soil N status interacts directly with BNF in the short and long term. In the short-term, increases in
soil inorganic N occurs during dry conditions and where N fertiliser is used, and these will reduce BNF.
In the long-term, BNF leads to accumulation of soil N, grass dominance, and reduced BNF. However,
cyclical patterns of legume and grass dominance can occur due, at least in part, to temporal changes in
plant-available N levels in soil. Thus, there is a dynamic relationship between legumes and grasses
whereby uptake of soil N by grass reduces the inhibitory effect of soil N on BNF and competition by
grasses reduces legume production and BNF. Factors affecting the competition between legumes and
grasses are considered including grass species, grazing animals, and grazing or cutting management.
Some fixed N is transferred from legumes to associated grasses. The amount of N transferred
'below-ground', predominantly through decomposition of legume roots and nodules, has been estimated at 3 to 102 kg N ha-1 yr-1 or 2 to 26% of BNF. In grazed pasture, N is also transferred
'above-ground' via return in animal excreta and this can be of a similar magnitude to 'below-ground'
transfer.
Increased BNF in mixed legume/grass pastures is being obtained through selection or breeding of
legumes for increased productivity and/or to minimise effects of nutrient limitations, low soil moisture,
soil acidity, and pests and disease. Ultimately, this will reduce the need to modify the pasture
environment and increase the role of legumes in low-input, sustainable agriculture.
Introduction
The importance of pasture legumes for improving the nitrogen (N) status of soils and for
maintaining a high level of total sward production without N fertiliser has long been recognised. However, during the past few decades,
agricultural production in Europe has relied
heavily on frequent applications of N fertiliser to
grass-only pastures. With the increasing interest
in low-input sustainable agriculture throughout
the world and the concern about possible environmental problems associated with high N
fertiliser use, interest has rekindled in using pasture legumes in Europe and USA as a source of
biologically-fixed N. In contrast, Australia and
New Zealand have always relied on legumes in
extensive and intensive pastoral agriculture to
maintain a low-cost farming system.
The level of biological nitrogen fixation (BNF)
by pasture legumes can vary greatly (Table l),
being site-dependent and influenced by a wide
range of factors including soil properties, environmental conditions, pests and disease, and
management. In this overview, we will centre on
some major factors that influence legume persistence, production and BNF. This will be preceded by an introduction to the legume/
Rhizobium symbiosis and the methods of measurement of BNF. The ultimate effects of BNF
on total pasture production is greatly influenced
by the transfer of fixed N to associated grasses.
Nitrogen fixation in legurnelgrass pastures
139
Table 1. Summary of BNF in mown and grazed legume/grass pastures (further estimates are given in Table 6)
Country
Mowing trials
UK (hills)
UK
Ireland
Switzerland
Sweden
Austria
Canada
USA
USA
Uruguay
Australia
Australia
New Zealand
New Zealand
Grazed trials
Ireland
Wales (hills)
New Zealand
New Zealand
(hills)
New Zealand
a
total N yield;
b
Legume
Method
BNF
(kgNha-1yr-1)
White clover
White clover
White clover
White clover
Red clover
Lucerne
Lucerne
Alsike clover
Red clover
Lucerne
Sub. clover
Crimson clover
White clover
Various
Tropical sp.
White clover
White clover
TN a, AR b
TN
AR
15N
15N
15N
15N
AR
AR
15N
15N
15N
15N
TN
TN
15N
AR
47-125
74-280
268
83-283
49-373
242-319
104-108
21-143
15-77
114-224
104-206
124-185
13-109
20-220
51-137
45-142
76-105
Labandera et al. (1988)
Vallis and Gardener (1984)
Johansen and Kerridge (1979)
Edmeades and Goh (1978)
Crush et al. (1983)
White clover
White clover
White clover
AR
TN
AR
83-296
90-98
85-342
Masterton and Murphy (1976)
Munro and Davies (1974)
Hoglund et al. (1979)
White clover
White clover
15
N
N
55-85
82-291
Ledgard et al. (1987)
Ledgard et al. (1900)
15
Reference
Newbould (1982)
Cowling (1982)
Halliday and Pate (1976)
Boller and Nosberger (1987)
Wivstad et al. (1987)
Danso et al. (1988)
Rice (1980)
Heichel et al. (1984)
Brink (1990)
acetylene reduction.
Thus, the importance of N transfer will also be
discussed along with the consequences to longerterm growth and BNF by the legume.
The legume/Rhizobium symbiosis
Biological nitrogen fixation is dependent on establishment of a symbiotic relationship between
the legume and an effective Rhizobium strain.
Reviews of the important processes that lead to
development of the symbiotic association of
rhizobia within the root nodules of legumes have
been given by Bauer (1981), Dowling and
Broughton (1986), Heichel and Brophy (1985)
and Roughley (1985).
Nodule development
Nodule development involves a complex series
of processes:
(1) Successful colonisation and survival of effec-
tive Rhizobium strains in the soil. This is the
first requirement and depends on satisfactory
soil and environmental conditions.
(2) Infection of the host plant. This begins with
recognition and proliferation of rhizobia
around the legume roots. The rhizobia then
induce root-hair deformation (e.g. curling)
and penetrate the deformed root by localised
digestion of the root-hair wall. An infection
thread is then initiated and this remains
extracellular within the root hair. These
early stages of infection can be greatly affected by environmental conditions and by
factors such as low soil pH. The infection
process is also sensitive to nitrate resulting in
reduced root-hair deformation and infection.
(3) Nodule initiation and development. The infection thread passes into root cortical cells
and this stimulates cell division and nodule
initiation, thereby forming the root nodule.
The infection thread releases bacteria which
multiply and differentiate into bacteroids,
140
Ledgard and Steele
where fixation of atmospheric N2 occurs.
The biochemistry of BNF in the bacteroids
has been described by Bergersen (1971).
Inoculation
When legumes are first introduced into soils,
root nodules may fail to develop because of a
low population of compatible and effective
strains of Rhizobium in the soil. Thus, addition
of appropriate strains of Rhizobium by inoculation is generally essential where legumes have
not been grown before and where there are no
naturalised rhizobia. The aim of inoculation of
legume seed is to coat the seed with a sufficiently
large number of viable rhizobia of the correct
strain to give rapid and effective nodulation of
that legume in the field.
Legume inoculation has been reviewed by
Date (1970) and Roughley (1976). A Rhizobium
culture (often in a peat medium) can be mixed
with seed prior to sowing. Alternatively, inoculum is often pelleted onto legume seed and
covered with a protective coating material such
as lime. Lime-coating can also enhance nodulation in low-pH soils giving an effect similar to
that from large surface applications of lime to
soil (Lowther, 1975).
The subsequent survival of rhizobia in soil is
influenced by environmental and soil factors, and
this has been reviewed by Alexander (1985).
Host plant/Rhizobium strain interaction
The amount of N2 fixed is influenced genetically
by both Rhizobium and host plant characteristics
(Mytton, 1983). Improved strains of Rhizobium
have been selected for improved BNF in certain
conditions (Roughley , 1985). However, improved strains must be capable of successful
establishment in the field and this depends on a
number of factors including their ability to compete with rhizobia already established in the soil.
Also, strains vary in competitive ability and effectiveness. Thus, it may be difficult to capitalise
on highly effective strains because of poor competitiveness with resident strains.
One approach is the breeding of a high degree
of host-strain specificity for nodulation and BNF.
The host plant can influence which strains in a
mixture form nodules and can control functionality of nodules (Heichel and Brophy, 1985).
Mytton (1983) reviewed the subject of selection
for improved symbiotic efficiency and concluded
that specificity should be utilised by breeding
plants for increased yield and simultaneously
selecting for improved Rhizobium strains. This
approach is being used in studies on the legume/
Rhizobium symbiosis in acid soils (SylvesterBradley et al., 1988).
Rhizobia1 denitrification
As well as fixing atmospheric N2 in association
with legumes, rhizobia are capable of utilising
soil nitrate and converting it to gaseous oxides of
N and N2 by denitrification. Rhizobial denitrification in temperate grassland soils has been reviewed by Steele and Bonish (1985).
Denitrification can occur in various strains of
Rhizobium and Bradyrhizobium. In some New
Zealand soils, the populations of denitrifying
rhizobia equal or exceed those of other denitrifying bacteria.
O'Hara et al. (1984) examined the effect of
Bradyrhizobium sp. (lotus) on rhizobial denitrification and measured potential losses of up to
2.3 kg N2 O-N ha-1 day-1 with a soil population
of 2 X 108 cm-2 soil. However, subsequent field
studies (Steele and Bonish, 1985) at the more
commonly found rhizobial populations of about
104 g-1 soil showed low levels of N2O loss from
grassland soils (<30 g N2O-N ha-1 day-1).
Methods for measuring BNF in the field
Estimates of BNF may be influenced, in part, by
the method of measurement used (Table 2).
Some of the advantages and limitations of the
main methods will be discussed. Greater detail
on methodology of measurement of BNF is given
in reviews on this topic (e.g. Knowles, 1981;
Ledgard and Peoples, 1988).
Acetylene reduction assay (ARA)
This technique was recently reviewed by Witty
and Minchin (1988) and commonly involves the
incubation of whole plants or plant parts in a
Nitrogen fixation in legume/grass pastures
141
Table 2. Effect of method of measurement on BNF (kg N ha -1 ) by field-grown pasture legumes
Legume
Reference
plant
Sub.' clover
Bromus mollis
White clover
White clover
Ryegrass
Ryegrass
Sub. clover
Ryegrass
Sub. clover
Sub. clover
Lucerne
Lucerne
White clover
Ryegrass
Phalaris
Ryegrass
Phalaris
Ryegrass
White clover
Ryegrass
Method
ARA a
Reference
N diff. b
57
89
6.4
15 N
dilution
15 N natural
abundance
103
183
4.9
1.24
1.83
-0.59
-0.16
-0.35
0.05
113
(85) d
(91)
21
47
1.70
1.30
0.88
0.56
0.92
1.24
269
(45)
(62)
14
48
2.05
2.24
0.81
0.51
Phillips and
Bennett (1978)
Goh et al. (1978)
Edmeades and
Goh (1979)
Bergersen and
Turner (1983)
Ledgard
et al. (1985c)
Eltilib and
Ledgard (1988)
Ledgard (1991)
a
Acetylene reduction assay.
Nitrogen difference.
c Subterranean.
d Values in parentheses are percentage of legume N fixed.
b
closed vessel containing about 10% acetylene for
a time period of approximately 0.5-2 hours. The
reduction of N2 by the nitrogenase enzyme is
inhibited and instead the enzyme reduces
acetylene to ethylene. These gases are measured
by gas chromotography .
The ARA is simple, rapid, sensitive and relatively low-cost. However, there are a number of
limitations of the technique and the main ones
are:
(1) Short-term measurement. A series of measurements to cover diurnal, daily and seasonal variations are required for long-term
estimates.
(2) Need for calibration of ethylene production
with BNF. Measured ratios of acetylene/N2
conversion have ranged between 1.5:1 and
25 : 1, and may vary even between treatments within an experiment.
(3) Influence of assay on nitrogenase activity. In
some legumes there can be a substantial
decline in nitrogenase activity within several
minutes of commencing the assay and this
can lead to an underestimation of BNF by
up to 50%.
The ARA was widely used in field studies on
pasture legumes in the 1970s, but has now largely been replaced by 15 N techniques, particularly
where the aim is to quantify the amount of N
fixed.
Total legume nitrogen
This is one of the simplest methods for estimating BNF. However, it assumes that all of the
legume N has been fixed and this is rarely true in
the field because legumes also utilise plant-available soil N. Thus, it overestimates BNF. Nevertheless, on soils of low N availability the proportion of legume N fixed from atmospheric N2 ( P N )
is often near 90% (e.g. Bergersen and Turner,
1983; Ledgard et al., 1987).
This method, and the subsequent methods described, may significantly underestimate total
BNF when only above-ground legume herbage is
analysed. The root, nodule and stolon/rhizome
(where present) components can contain a relatively large amount of fixed N.
Nitrogen difference
With this method, BNF is estimated from the
difference in N yield between the legume and a
non-N2-fixing reference plant grown in the same
soil. It is assumed that the legume and reference
plant absorb the same amount of soil N.
142
Ledgard and Steele
Hardy and Holsten (1977) concluded that this
method generally underestimates BNF because
the legume utilises less soil N than the reference
plant. Indeed, with mixed legume/grass pastures
this method can result in negative estimates of
BNF (Eltilib and Ledgard, 1988; Ledgard et al.,
198%).
15
N isotope dilution
The term 15N isotope dilution' is commonly
used to describe a method which utilises differences in 15N enrichment of atmospheric N2 and
soil N, where the soil N is labelled by addition of
15N-enriched
(or occasionally 15N-depleted) ma15
terial. The N isotope dilution method was reviewed by Chalk (1985). It enables estimation of
the proportion of legume N fixed from atmospheric N2(PN):
PN =
(atom %
Nref - atom %
15
(atom %
Nref - B)
Nlegume)
15
15
(1)
where ref is a non-N2-fixing reference plant
growing in the same soil as the legume and B is
the atom %15N of N derived from atmospheric
N2. The latter is commonly assumed to be 0.3663
which is the 15N concentration of atmospheric
N2. The amount of N fixed is estimated from
PN X legume N yield.
The main advantage of the method is that it
can be used to obtain a 'time-averaged' estimate
of PN and BNF. A major assumption is that
legume and reference plant have the same ratio
(R) of N assimilated from the added 15N material
to N assimilated from the indigenous soil N.
The main potential limitations of the method
are as follows, with the first being by far the
most significant:
(1) The N uptake characteristics of the legume
and reference plant may differ. Differences
in the pattern of N uptake over time, or with
soil depth, result in different R values for the
legume and reference plant (e.g. Ledgard et
al., 198Sd). This is due to associated temporal and spatial changes in the 15N enrichment
of soil. Thus, any procedures which minimise these changes in enrichment reduce the
potential error due to mis-matching of the
legume and reference plant. Such a procedure is the use of slow-release forms of 15N
(Witty, 1983).
Use of several reference plants can assist
in examining the potential error. However,
this is not usually possible in a mixed
legume/grass pasture where the existing
grass(es) must be used as the reference
plant(s). Ledgard et al. (1985b) described a
technique whereby R can be estimated and
the suitability of the reference plant examined.
(2) Direct transfer of fixed N from legume to
reference plant. This is a potential error in
mixed pastures where the legume and reference plant (the associated grass) are grown
together. In practice, N transfer appears to
be relatively slow and the effects on estimating BNF can be minimised by using new 15N
microplots at regular time intervals (e.g. 2-4
months). Alternatively, N transfer can be
adjusted for by measuring it using the techniques of Vallis et al. (1967) or Ledgard et
al. (1985a).
N natural abundance
15
This method was reviewed by Shearer and Kohl
(1986) and is essentially the same as the 15N
isotope dilution method except that no 15N material is added. It utilises the small, natural enrichment of 15N present in most soils. This natural 15N enrichment is relatively uniform over time
and with soil depth (e.g. Bergersen et al., 1985;
Ledgard et al., 1984, 1985c). Thus, the major
limitation of the 15N isotope dilution method
which requires the legume and reference plant to
have similar N uptake characteristics, is of relatively little importance with the 15N natural
abundance method.
The main potential limitations of the method
are:
(1) Analytical sensitivity and errors. An isotope
ratio mass spectrometer capable of accurately measuring differences of about ±0.00004
atom % 15N is required. Great care is required during sample preparation and analysis to avoid isotopic fractionation and contamination from 15N-enriched material.
Nitrogen fixation in legurnelgrass pastures
(2) Low natural 15N enrichment of soil N. The
error in estimating PN increases considerably
as the 15N concentration of plant-available
soil N nears that of fixed N, (Ledgard and
Peoples, 1988). Steele (1983) reported low
natural 15N enrichments of soil N in intensively grazed pastures in New Zealand,
whereas relatively high enrichments were
measured in Australian studies (Bergersen
and Turner, 1983; Ledgard et al.. 1985c).
(3) Isotopic fractionation during N2 fixation.
This results in the value for B in equation 1
being different from the 15N concentration of
atmospheric N2. However, B appears to be
constant so that it can be measured by growing the legume in a N-free medium. Nevertheless, B should be determined in conditions relevant tu the field because it can be
influenced by factors such as nutrition, moisture and rhizobia1 strain (Ledgard, 1989).
(4) Isotopic fractionation during uptake of soil
N. Mariotti et al. (1980) showed this factor
to be insignificant for a range of legumes and
non-legumes in solution culture studies.
However, Hogberg (1990) recently found
that the association of plant roots with ectomycorrhiza or vesicular-arbuscular mycorrhiza can alter the 15N concentration of N
derived from soil. Thus, it is preferable that
the legume and reference plant have similar
mycorrhizal associations.
143
Comparisons of the 15N natural abundance
method with the 15N isotope dilution method
have often resulted in similar estimates (Table 2)
with similar precision. All methods have some
potential limitations and ideally two (or more)
methods should be used with results related to
associated measurements of plant nodulation and
soil N status.
Factors affecting BNF
Biological nitrogen fixation by legumes in mixed
pastures is influenced by three primary factors;
soil N status, legume persistence and production,
and competition with the associated grass. They
interact as shown in Figure 1. These primary
factors are themselves influenced by a wide
range of soil, environmental and biotic factors.
Effect of soil nitrogen status
The inhibition of BNF by increasing levels of
inorganic N in soil is well known. Thus, fluctuations in inorganic soil N can influence legume
production and BNF in the long and short term.
Changes with pasture development
When legume/grass pastures are sown into undeveloped land or into cropped and newly cultivated land, there is a period with several years of
Fig. I . Diagrammatic representation of the major factors which interact to determine the level of biological nitrogen fixation.
144
Ledgard and Steele
legume dominance. This is primarily due to low
levels of inorganic and mineralisable N in soil
and can result in extremely high levels of BNF
(e.g. 682kg N ha-1 yr-1, Sears et al., 1965).
Over time there is an increase in soil N and a
succession to grass dominance. However, even in
established pastures this successional pattern
may reoccur at regular intervals (e.g. 4-year
cycles, Steele, 1983), presumably due in part to
variations in soil N status. It may also occur
spatially within a grazed pasture. For example,
areas affected by dung deposition can result in
death of plants, colonisation by white clover
( Trifolium repens L.) and dominance for up to
1.5 years (Weeda, 1967).
Fluctuations during the year
In pastoral soils, the level of inorganic soil N
fluctuates with environmental changes and is
often high during dry summer conditions. This
can lead to a large decline in the proportion of
legume N fixed from atmospheric N2 (PN) (Ledgard et al., 1987). In contrast, the application of
moderate rates of N fertiliser (up to 50 kg
N ha-1) may have relatively little effect on PN
(Boller and Nosberger, 1987; Steele and Percival, 1984).
The levels of inorganic soil N and of BNF are
influenced by differences in the seasonal pattern
of growth between legumes and grasses. For
example, the high potential growth rate of ryegrass ( Lolium perenne L.) in late winter/early
spring depletes soil N thereby enhancing legume
growth and BNF in late spring/summer when
temperature is favourable for legumes (Harris,
1987).
Spatial variability due to animal excreta
Seventy-five to 95% of N ingested by grazing
animals is returned in excreta to localised areas
at rates equivalent to up to 1200 kg N ha-1 (Henzell and Ross, 1973; Steele, 1982). Approximately 50-80% of excreta N is in urine, being highest
from high N-containing pastures.
Biological nitrogen fixation by legumes within
urine-affected areas may decline by up to 90%
(Ball et al., 1979; Ledgard et al., 1982). The
BNF per unit of legume growth recovers after
approximately 30-60 days, when levels of inorganic N in soil fall to 'back-ground' levels. HOW-
ever, urinary N induces an increase in associated
grass growth and competition with the legume.
The net effect of this is that BNF per unit area is
reduced for relatively longer periods (e.g. up to 3
months). Thus, in an intensive dairy farm, where
up to 40% of the area may be affected by excreta
(Saunders, 1984) and BNF within these areas
may decline by an average of about 60% (Ledgard et al., 1982), total BNF may be reduced by
up to 24% on an annual basis.
Legume tolerance to high soil nitrogen
Recent research on crop legumes has examined
the potential for increasing BNF by breeding or
selecting legumes that fix atmospheric N, under
high levels of inorganic soil N (e.g. Herridge and
Bergersen, 1988). Rys and Mytton (1985) have
also shown differences between white clover
selections in tolerance to inorganic N.
Herridge and Bergersen (1988) stated that
'development of symbioses where PN is maintained at near-maximum levels in the presence of
high soil nitrate could provide the biggest single
advance in the improvement of BNF by
legumes'. While this may be true for crop
legumes, it is likely to be of much less value in
legume/grass pastures, except perhaps in European pastures where high rates of N fertiliser are
regularly used. Even then it is likely that greater
advances in BNF could be made by improving
the production and competitiveness of legumes
in a mixed sward.
Rapid uptake of soil N by grasses in mixed
perennial pastures acts to maintain inorganic soil
N at a low level. Consequently, PN often exceeds
80% in mowing studies (e.g. Edmeades and
Goh, 1978; Hardarson et al. 1988), although this
may decline to an average of about 70% under
grazing (e.g. Ledgard et al., 1990; Ledgard et
al., 1987). Urinary-N can cause a relatively large
short-term decline in P, in pasture legumes but
the main decrease in BNF from urinary-N is
associated with a decline in legume production
because of enhanced competition from the associated grass (Ledgard and Saunders, 1982).
Thus, improvement in the N tolerance of
legumes in perennial pastures may provide only
a minor improvement in BNF compared to that
achievable in crop legumes.
Nitrogen fixation in legume/grass pastures
Effect of legume persistence and production
There are a large number of factors that influence the persistence and production of pasture
legumes. For example, environmental conditions
can have a major effect on legume persistence
and production (Hochman and Helyar, 1989)
and on BNF (Gibson, 1976).
We have restricted this section to the important effects of soil moisture status, soil acidity, nutrition, and pests and disease on legume
persistence, production and BNF. Norman
(1982) noted that ‘the two most important
ecological limitations, on a global scale, are acid
soils and long dry seasons’.
Soil moisture status
The amount and distribution of rainfall has a
major effect on the persistence of pasture
legumes, particularly shallow-rooting perennial
legumes such as white clover. Thus, in drier
areas of Australia and USA there is a predominance of deep-rooting perennials such as lucerne
( Medicago sativa L.) and red clover ( Trifolium
pratense L.) (Hochman and Helyar, 1989;
Matches, 1989).
BNF is highly sensitive to moisture stress. In
temperate hill country pastures, Ledgard et al.
(1987) measured a large decline in the total N
concentration of legumes due to dry summer
conditions. This was also associated with a decrease in P N from about 85 to 70%. Wery et al.
(1986) observed that drought-stressed lucerne
showed a larger decline in BNF than in nitrate
assimilation. These studies indicate that the intensity of stress from dry soil conditions on plant
parameters is of the order:
BNF > uptake of soil N > dry matter
accumulation (or photosynthesis)
Dry soil conditions lead to accumulation of
inorganic N in soil and therefore to prolonged
effects on BNF (e.g. Hoglund and Brock, 1978).
Consequently, uptake of soil N may be more
important than BNF during recovery by legumes
after drought stress.
The occurrence of droughts can have a major
effect on the total amount of N fixed. For example, Crouchley (1979) measured BNF in white
145
clover/ryegrass pasture at 90 and 240 kg N
ha -1 yr -1 during years with dry and wet summers, respectively. Robin et al. (1989) showed
that white clover cultivars can differ in the immediate effects of moisture stress on BNF. However, this may have only a small effect on the
annual amount of N fixed.
Soil acidity
Soil acidity is a major factor influencing legume
persistence, production and BNF in tropical regions throughout the world and in humid areas
of USA (Matches, 1989). It also affects about 10
million ha of pastoral soils in Australia (Gramshaw et al.. 1989).
In many cases the effects of soil acidity are
manifested through aluminium (AI) and/or manganese (Mn) toxicity. Aluminium toxicity generally decreases legume growth more when it is
dependent on BNF than when dependent on
uptake of soil N. Wood et al. (1984) found that
effects of AI on rhizobia1 growth were greater
than those on root growth of white clover. There
are large differences between legume species in
their tolerance to Al and therefore the choice of
legume for sowing in high-Al soils is important.
Selection of rhizobia and/or legume genotypes
can also reduce the effects of AI toxicity on BNF
(Jarvis and Hatch, 1987). Mutants of soybean
selected for nodulation and BNF under high
nitrate also nodulated under high AI conditions
(Aha et al., 1988), indicating the potential for
breeding for AI tolerance.
Effects of Mn toxicity are also greatest on
legumes deriving their N from BNF, and
genotype selection can improve the tolerance to
Mn (Evans et al., 1987).
Unfortunately, BNF also contributes to the
problem of soil acidification. Inputs of biologically fixed N to soil and the subsequent leaching of
nitrate from soils represents an important cause
of acidification of soils (Hochman and Helyar,
1989). In many situations it is uneconomic to
counter acidification by applying lime to soils.
Thus. the selection or breeding of tolerant
legumes and rhizobia is an area into which much
research is being directed. However, it may not
be a long-term solution if soil acidification continues to occur.
146
Ledgard and Steele
Nutrition
Nutrient deficiencies or toxicities can severely
limit legume persistence, production and BNF.
This subject has been reviewed by Dunlop and
Hart (1987), O’Hara et al. (1988) and Robson
(1978).
Evans (1977) showed that the commonly used
temperate pasture legumes of white, red and
subterranean clover have shorter, thicker lessbranched roots and shorter, less numerous root
hairs than the grasses with which they are usually
associated. Thus, these clovers are disadvantaged in mixed swards in terms of uptake of
nutrients. This has been established in plant
competition studies between clovers and grasses
for limiting nutrients such as phosphorus (P) and
potassium (K) (Jackman and Mouat, 1972; Robson and Loneragan, 1978). Thus, fertiliser application to mixed legume/grass pasture must be
sufficient to meet the requirements of the legume
component in a competitive sward, and a decline
in legume production can occur when fertiliser is
withheld (e.g. O’Connor et al., 1990).
Phosphorus is the main limiting nutrient (apart
from N) for pasture production in tropical soils
and in many temperate soils (e.g. Skerman et al.,
1988). Legumes can vary greatly in their tolerance of low soil P levels. The tropical legume
Stylosanthes humilis is very efficient in P uptake
from low-P soils and may have a relatively low P
requirement for maximum growth (Skerman et
al., 1988). Cadisch et al. (1989) measured differences between eight tropical legumes in their
response to P and K addition, with increases in
BNF ranging between 10 and 66 kg N ha-1 .
The requirement for P is primarily for growth
of the host legume plant. Even when a legume is
P-deficient, nodules contain adequate P and
nodule function is not impaired (O’Hara et al.,
1988). The specific nutrient constraints to BNF,
in terms of the components of the legume/
Rhizobium symbiosis, are summarised in Table
3.
Pests and disease
The most important cause of lack of persistence
of the important perennial legumes in USA is
the pest-disease complex (Matches, 1989). Costs
of chemical control can be high and therefore a
significant research effort has gone into breeding
Table 3. Summary of nutrients for which there is a relatively
high requirement for specific components of BNF in legumes
(from O’Hara et al., 1988 and Robson, 1978)
Components of BNF
Nutrients required in a relatively
high concentration
Nodule initiation
Nodule development
Nodule function
Host plant growth
Co
B, Ca
Mo, Fe, Co, Ca, Cu
P. S. K. etc.
legumes resistant to a number of pests and disease. Cultivars of lucerne are now available with
resistance to many pests and diseases, although
resistance to the Fusarium root rot complex remains elusive (Watson et al., 1989).
Stylosanthes humilis and S. guianensis were
once the most important pasture legume species
in Northern Australia but the development of
anthracnose diseases has caused almost complete
annihilation of them (Irwin, 1989). Current
breeding programmes are examining utilisation
of a range of resistant and partial-resistant accessions.
In New Zealand, insect pests have resulted in
losses of pasture plants of up to 40% (East and
Pottinger, 1984). Subclinical effects of pests and
disease on legume growth and BNF can also be
large. Steele et al. (1985) applied insecticide
treatments to 16 different white clover/ryegrass
pastures that had no visual signs of pests or
disease. They obtained average increases in total
pasture production, clover production and BNF
of 13, 28 and 57%, respectively, and it appeared
that a major reason for this was the control of
soil nematodes. These results clearly indicate
that large increases in production and BNF may
be achieved by breeding legumes resistant to
pests. A new cultivar of white clover, ‘Grasslands Kopu’, is resistant to stem nematode and
selection programmes are underway to improve
tolerance or resistance to the more important
soil nematodes (Watson et al., 1989). Molecular
biology techniques may widen the scope for
breeding pest-resistant legumes.
Competition with the associated grass
Legume production and BNF is influenced by
competition between the legume and the associated grass(es) (Fig. 1). Factors affecting this
147
Nitrogen fixation in legume/grass pastures
the order: sheep < cattle < goats (Sheath and
Hodgson, 1989).
competition in white clover pastures were reviewed by Harris (1987). In general, any factors
that favour an increase in legume growth cause a
decrease in yield of the associated grass. However, any stress that has a greater effect on the
legume component (e.g. selective pests or diseases) can enhance the competitiveness of the
grass and further reduce legume production and
BNF.
Grazing or cutting management
This can be used to favour high legume production (and BNF) relative to the grass production.
For example, intensive grazing (especially frequent defoliation during spring) of white clover
swards produced annual increases in BNF of 33
and 10% in Argentinian (Refi et al., 1989) and
New Zealand (Brock et al., 1983) studies, respectively.
Associated grass species or cultivar
The effect of associated grass species or cultivar
on BNF can be manifested through differences in
uptake of plant-available soil N. Thus, at some
stages of pasture growth (e.g. during establishment) grasses more active in uptake of soil N
reduce the inhibitory effect of soil N on BNF.
For example, Ledgard et al. (1985d) measured
20% more N fixed by establishing subterranean
clover when grown with annual ryegrass compared to when grown with the slow-establishing
phalaris, despite a 15% lower total N yield.
However, in most situations with established
pasture, more aggressive grass species or cultivars result in less legume growth and lower
BNF. Harris and Hoglund (1977) measured a
65% decrease in BNF (acetylene reduction)
when white and red clovers were grown with a
more aggressive and productive ryegrass cultivar
than with a less aggressive one.
Relationship between legume production and
BNF
Biological nitrogen fixation is a product of
legume production, total N concentration and
PN. Several studies have shown close correlations between BNF and legume production, particularly in mown swards or in grazed swards on
low-fertility soils where PN exceeds 80% (e.g.
Ledgard et al., 1987). This indicates that increases in legume production through selection
or breeding would lead to similar increases in
BNF. However, this will not always be the case.
Increases in legume growth may lead to a
proportionally larger increase in BNF. This may
occur because increases in legume growth generally have little effect on uptake of soil N by the
legume (e.g. Tables 4 and 5). Thus, increases in
legume growth and N assimilation may lead to a
proportionally larger increase in BNF, because
PN also increases. For example, Eltilib and Ledgard (1988) compared two white clover cultivars
in mixed swards under sheep grazing and found
that Kopu produced 14% more dry matter than
Huia, but fixed 35% more N (Table 4).
Greater legume production can also be associated with a lower total N concentration in
Grazing animal
Competition between legumes and grasses can
be influenced by the grazing animal. Clark et al.
(1982) found marked differences between sheep
and goats in selection for clover during grazing.
Goats selectively grazed browse species, grasses
and weeds, and this led to a large increase in
white clover production. The clover content of
swards grazed by different animals is generally in
Table 4. Effect of cultivar of white clover on growth and BNF during a period of rapid spring growth in a mixed sward under
sheep grazing (Eltilib and Ledgard, 1988)
Cultivar
Clover growth (kg DM ha day )
PN(%)
Clover N fixed (kg N ha-1 day-1)
Clover N from soil (kg N ha-1 day-1)
-1
-1
LSD(5%)
Huia
Kopu
25.5
19.6
0.92
0.25
29.1
84.0
1.24
0.25
4.4
3.0
0.14
0.05
148
Ledgard and Steele
Table 5. Effect of cultivar on growth, N concentration, BNF and soil N uptake by white clover in mixed swards under dairy cow
grazing (Ledgard et al., 1990)
Clover
cultivar
Clover DM
(kg ha -1 yr -1 )
Aran
Kopu
Resident
LSD (5%)
8260
7690
7210
790
%N
PN
(%)
4.54
4.95
5.26
0.42
74
75
70
19
legume herbage, resulting in no increase in BNF
(Table 5).
These different associations between legume
production and BNF illustrate that breeding or
selection of legumes for improved BNF should
not be based solely on legume production, but
should include measurements of total N concentration and P N.
Transfer of fixed nitrogen to associated grasses
An important consequence of BNF in mixed
pastures is the transfer of fixed N to associated
grasses. Within a grazed pasture, this can occur:
'Above-ground' via the grazing animal where
fixed N is returned in dung and urine; and
'Below-ground' via direct root excretion or via
microbial decomposition of legume material.
Above-ground transfer of fixed N has not been
researched, presumably because of difficulties of
measurement. Nevertheless, Ledgard (1991)
used an indirect method to assess above-ground
transfer in a white clover/ryegrass sward at
60 kgN ha -1 yr -1 (22% of BNF), and this was
similar to the estimate of below-ground transfer
(70 kg N ha -1 yr -1 ). Cycling of N via the grazing
animal is considered inefficient because of the
localised return in excreta at high N rates which
Soil N uptake
(kgN ha -1 yr -1 )
BNF
(kgN ha -1 yr -1 )
278
285
266
33
97
96
113
42
can lead to large N losses (e.g. Steele, 1982).
Studies by Hoglund (1985) showed that an increase in grazing intensity caused an increase in
N consumption by grazing animals and less accumulation of N in soil compared with that from
lax grazing where more fixed N was returned to
soil in senesced herbage.
Below-ground transfer of fixed N to grasses
has received renewed interest and during the last
few years a 15 N dilution method (Vallis et al.,
1967) has been used to obtain estimates of 3 to
70 kg N ha -1 yr -1 (Table 6). Overall this accounted for 2-26% of BNF and it contributed 8-39%
of grass N yields.
The grazing study of Ledgard (1991) clearly
showed that the transfer of fixed N (above and
below ground) was important for maintaining a
high N-demanding ryegrass component in the
sward and avoiding reversion to lower Ndemanding and lower producing grasses and
weeds. Thus, the level of N transfer affects the
competitive interaction between the grass and
legume in a mixed sward. This has been described mathematically and was reviewed by
Harris (1987). Large N transfer leads to increased grass growth and decreased legume
growth and BNF. Eventually, available soil N
reserves decrease, grass growth slows and
Table 6. Below-ground transfer of fixed N from legumes to associated grasses in mixed pastures estimated using
Legume
Lucerne
Lucerne
Lucerne
White clover
Red clover
White clover
a
N transfer
(kg ha yr )
AS % of
BNF
93-258
51-172
114-282
83-283
49-373
224-291
3-27
7-13
3-5
11-52
14-42
54-102
(+60) a
7
10
2
21
23
26
(+22)
Values in parentheses are above-ground N transfer.
N
Reference
BNF
(kg ha -1 yr -1 )
-1
15
-1
AS % of
grass N
37
22
8
38
39
27
(+23)
Burity et al. (1989)
Ta and Faris (1987)
Hardarson et al. (1988)
Boller and Nosberger (1987)
Boller and Nosberger (1987)
Ledgard (1991)
Nitrogen fixation in legurnelgrass pastures
149
Table 7. Estimates of the flow of N (kg ha-1) from clover/grass (CIG) swards grazed at 3, 6 or 9 cm and a grass-only (G-only)
sward (grazed at 6 cm) receiving 420 kg ha-1 fertiliser N over a 5 month period in the UK (Parsons et al., 1990)
Stocking rate (ewes/ ha)
N inputs - BNF
- Fertiliser
N cycling - excreta
N outputs
- wool/meat
- denitrification
- volatilisation
CIG
3 cm
CIG
6 cm
C/ G
9 cm
21
35
15
24
9
39
348
238
165
420
396
27
8
8
24
6
6
17
1
-
29
51
16
legume growth and BNF increase. Thus, in the
longer term, the level of BNF should approximately match losses of N from the pastoral
system.
Parsons et al. (1990) examined N cycling within sheep-grazed white clover/ryegrass pastures
and a ryegrass-only sward receiving 420 kg ha-1
of fertiliser N. They showed that losses of N
from the legume pastures were small relative to
those from the N-fertilised swards, although the
losses increased with increasing stocking rate
(Table 7). While leaching was not quantified,
G-only
420 kg N ha-1
19
levels of potentially leached nitrate in soil (01.8 m depth) in autumn were 4-fold higher under
the N-fertilised sward. The estimated rate of
BNF (ca. 40 kg N ha-1 using acetylene reduction)
was low, in contrast to the amount of N cycling
through the grazing animals (ca.
200400 kg N ha-1). This indicates that the legume
pasture was relatively efficient in cycling of N
and minimising environmental losses of N.
Nevertheless, other measurements near this site
indicate that at lower inputs of fertiliser N (e.g.
210 kg N ha-1 ) to grass-only swards, losses of N
Fig. 2. Diagrammatic representation of the major factors which interact to determine the level of biological nitrogen fixation, and
the secondary effects of stresses on these factors.
150
Ledgard and Steele
by volatilisation were greatly reduced and near
those from legume/grass pastures (Jarvis et al.,
1989). This suggests that there may be little
difference in efficiency of N cycling where similar
N inputs are derived from BNF or N fertiliser.
However, the natural feedback mechanism between BNF and soil N (described earlier; Fig. l)
may mean greater efficiency of N flow and less
potential for loss. Similarly, BNF is better
matched to spatial variability. For example, BNF
decreases greatly in areas affected by high levels
of plant-available N from excreta and increases
when available N declines.
Conclusions
A mixed legume/grass pasture is a dynamic system whereby BNF varies with differences in
legume growth potential, inorganic soil N level
and grass competitiveness. Superimposed over
this are fluctuations induced by various stresses
(e.g. environmental, soil, pests and diseases; Fig.
2). Research aimed at improving BNF by pasture
legumes is increasingly being directed at selecting or breeding legumes to minimise the effects
of such stresses. Ultimately this will mean less
need to modify the environment in which pasture
legumes are grown (e.g. stopping use of pesticides) and the greater use of legumes in lowinput, sustainable pastoral systems.
References
Alexander M 1985 Ecological constraints on nitrogen fixation
in agricultural ecosystems. Adv. Microbial Ecol. 8, 163183.
Aha A K, Edwards D G. Carroll B J, Asher C J and
Gresshoff P M 1988 Nodulation and early growth of soybean mutants with increased nodulation capacity under
acid soil infertility factors. Agron. J. 80, 836-841.
Ball R, Keeney D R, Theobald P W and Nes P 1979
Nitrogen balance in urine-affected areas of a New Zealand
pasture. Agron. J. 71, 309-314.
Bauer W D 1981 Infection of legumes by rhizobia. Annu.
Rev. Plant Physiol. 32, 407-449.
Bergersen F J 1971 Biochemistry of symbiotic nitrogen fixation in legumes. Annu. Rev. Plant Physiol. 22, 121-140.
Bergersen F J and Turner G L 1983 An evaluation of 15N
methods for estimating nitrogen fixation in a subterranean
clover-perennial ryegrass sward. Aust. J. Agric. Res. 34,
391-401.
Bergersen F J, Turner G L, Gault R R, Chase D L and
Brockwell J 1985 The natural abundance of 15N in an
irrigated soybean crop and its use for the calculation of
nitrogen fixation. Aust. J. Agric. Res. 36, 411-423.
Boller B C and Nosberger J 1987 Symbiotically fixed nitrogen
from field-grown white and red clover mixed with ryegrasses at low levels of 15N-fertilization. Plant and Soil 104,
219-226.
Brink G E 1990 Seasonal dry matter, nitrogen and dinitrogen
fixation patterns of crimson and subterranean clovers.
Crop Sci. 30, 1115-1118.
Brock J L, Hoglund J H and Fletcher R H 1983 Effects of
grazing management on seasonal variation in nitrogen
fixation. Proc. XIV Int. Grassl. Cong. 339-341.
Burity H A, Ta T C, Faris M A and Coulman B E 1989
Estimation of nitrogen fixation and transfer from alfalfa to
associated grasses in mixed swards under field conditions.
Plant and Soil 114, 249-255.
Cadisch G, Sylvester-Bradley R and Nosberger J 1989 15Nbased estimation of nitrogen fixation by eight tropical
forage legumes at two levels of P:K supply. Field Crops
Res. 22, 181-194.
Chalk P M 1985 Estimation of N2 fixation by isotope dilution:
An appraisal of techniques involving 15N enrichment and
their application. Soil Biol. Biochem. 17, 389-410.
Clark D A, Lambert M G, Rolston M P and Dymock N 1982
Diet selection by goats and sheep on hill country. Proc.
N.Z. Soc. Anim. Prodn. 42, 155-157.
Cowling D W 1982 Biological nitrogen fixation and grassland
production in the United Kingdom. Phil. Trans. R. Soc.
Lond. B 296, 397-404.
Crouchley G 1979 Nitrogen fixation in pasture. VIII.
Wairarapa plains dryland, Masterton. N.Z.J. Exp. Agric.
7,31-33.
Crush J R, Cosgrove G P and Brougham R W 1983 Nitrogen
fixation during 1979-81 in 2 pastures on the Manawatu
Plains. N.Z.J. Exp. Agric. 11, 17-20.
Danso S K A. Hardarson G and Zapata F 1988 Dinitrogen
fixation estimates in alfalfa-ryegrass swards using different
Nitrogen-15 labeling methods. Crop Sci. 28, 106-110.
Date R A 1970 Microbiological problems in the inoculation
and nodulation of legumes. Plant and Soil 32, 703-725.
Dowling D N and Broughton W J 1986 Competition for
nodulation of legumes. Annu. Rev. Microbiol. 40, 131157.
Dunlop J and Hart A L 1987 Mineral nutrition. In White
Clover. Eds. M J Baker and W M Williams. pp 153-184.
C.A.B. International, Wallingford, UK.
East R and Pottinger R P 1984 The cost of pasture pests.
N.Z. Agric. Sci. 18, 136-140.
Edmeades D C and Goh K M 1978 Symbiotic nitrogen
fixation in a sequence of pastures of increasing age measured by a 15N dilution technique. N.Z.J. Agric. Res. 21.
623-628.
Eltilib A M and Ledgard S F 1988 Production and nitrogen
fixation by 'Grasslands Kopu' and 'Grasslands Huia' white
clovers under different nitrogen regimes. N.Z.J. Agric.
Res. 31, 325-330.
Evans J, Scott B J and Lill W J 1987 Manganese tolerance in
subterranean clover ( Trifolium subterraneum L.) genotypes
Nitrogen fixation in legurne/grass pastures
grown with ammonium nitrate or symbiotic nitrogen. Plant
and Soil 97, 207-215.
Evans P S 1977 Comparative root morphology of some pasture grasses and clovers. N.Z.J. Agric. Res. 20, 331-335.
Gibson A H 1976 Recovery and compensation by nodulated
legumes to environmental stress. In Symbiotic Nitrogen
Fixation in Plants. Ed. P S Nutman. pp 385-403. Cambridge University Press, Cambridge.
Goh K M, Edmeades D C and Robinson B W 1978 Field
measurements of symbiotic nitrogen fixation in an established pasture using acetylene reduction and a 15N method.
Soil Biol. Biochem. 10, 13-20.
Gramshaw D, Read J W, Collins W J and Carter E D 1989
Sown pasture and legume persistence: An Australian overview. In Persistence of Forage Legumes. Eds. G C Marten
et al. pp 1-22. American Society of Agronomy. Madison,
WI.
Halliday J and Pate J 1976 The acetylene reduction assay as a
means of studying nitrogen fixation in white clover under
sward and laboratory conditions. J. Brit. Grassl. Soc. 31.
29-35.
Hardarson G. Danso S K A and Zapata F 1988 Dinitrogen
fixation measurements in alfalfa-ryegrass swards using nitrogen-15 and influence of the reference crop. Crop Sci.
28, 101-105.
Hardy R W F and Holsten R D 1977 Methods for measurement of dinitrogen fixation. In A Treatise on Dinitrogen
Fixation. IV. Agronomy and Ecology. Eds. R W F Hardy
and A H Gibson. pp 451-486. Wiley. New York.
Harris W 1987 Population dynamics and competition. In
White Clover. Eds. M J Baker and W M Williams. pp
203-297. C.A.B. International, Wallingford, UK.
Harris W and Hoglund J H 1977 Influences of seasonal
growth periodicity and N-fixation on competitive combining abilities of grasses and legumes. Proc. XIII. Int. Grassl.
Cong. 239-243.
Heichel G H and Brophy L S 1985 Role of the legume in the
legume-Rhizobium symbiosis. In Forage Legumes for
Energy-efficient Animal Production. Eds. R F Barnes et
al. pp 146-154. USDA, Washington, DC.
Heichel G H, Barnes D K, Vance C P and Henjum K I 1984
N2 fixation, and N and dry matter partitioning during a
4-year alfalfa stand. Crop Sci. 24, 811-815.
Henzell E F and Ross P J 1973 The nitrogen cycle of pasture
ecosystems. In Chemistry and Biochemistry of Herbage 2.
Eds. G W Butler and R W Bailey. pp 227-245. Academic
Press, London.
Herridge D F and Bergersen F J 1988 Symbiotic nitrogen
fixation. In Advances in Nitrogen Cycling in Agricultural
Ecosystems. Ed. J R Wilson. pp 46-65. The Cambrian
News Ltd., Aberystwyth.
Hochman Z and Helyar K R 1989 Climatic and edaphic
constraints to the persistence of legumes in pastures. In
Persistence of Forage Legumes. Eds. G C Marten et al. pp
177-203. American Society of Agronomy, Madison, WI.
Hogberg P 1990 15N natural abundance as a possible marker
of the ectomycorrhizal habit of trees in mixed African
woodlands. New Phytol. 115, 483-486.
Hoglund J H 1985 Grazing intensity and soil nitrogen accumulation. Proc. N.Z. Grassl. Assn. 46, 65-69.
151
Hoglund J H and Brock J L 1978 Regulation of nitrogen
fixation in a grazed pasture. N.Z.J. Agric. Res. 21, 73-82.
Hoglund J H. Crush J R, Brock J L, Ball R and Carran R A
1979 Nitrogen fixation in pasture. XII. General discussion.
N.Z.J. Exp. Agric. 7, 45-51,
Irwin J A G 1989 Diseases of pasture legumes in Australia.
In Persistence of Forage Legumes. Eds. G C Marten et al.
pp 399-418. American Society of Agronomy, Madison
WI.
Jackman R H and Mouat M C H 1972 Competition between
grass and clover for phosphate. 1. Effect of browntop
( Agrostis renuis Sibth.) on white clover ( Trifolium repens
L.) growth and nitrogen fixation. N.Z.J. Agric. Res. 15,
653-666.
Jarvis S C and Hatch D J 1987 Differential effects of low
concentrations of aluminium on the growth of four
genotypes of white clover. Plant and Soil 99, 241-253.
Jarvis S C. Hatch D J and Roberts D H 1989 The effects of
grassland management on nitrogen loss from grazed swards
through ammonia volatilisation: The relationship to excreta N returns from cattle. J. Agric. Sci., camb. 112,
205-216.
Johansen C and Kerridge P C 1979 Nitrogen fixation and
transfer in tropical legume-grass swards in south-eastern
Queensland. Trop. Grassl. 13, 165-170.
Knowles R 1981 The measurement of nitrogen fixation. In
Current Perspectives in Nitrogen Fixation. Eds. A H Gibson and W E Newton. pp 327-337. Australian Academy of
Science. Canberra, Australia.
Labandera C. Danso S K A. Pastorini D, Curbelo S and
Martin V 1988 Nitrogen fixation in a white clover-fescue
pasture using three methods of nitrogen-15 application and
residual nitrogen-15 uptake. Agron. J. 80, 265-268.
Ledgard S F 1989 Nutrition, moisture and rhizobia1 strain
influence isotopic fractionation during N2 fixation in pasture legumes. Soil Biol. Biochem. 21, 65-68.
Ledgard S F 1991 Transfer of fixed nitrogen from white
clover to associated grasses estimated using 15N methods in
swards grazed by dairy cows. Plant and Soil 131, 215-223.
Ledgard S F and Peoples M B 1988 Measurement of nitrogen
fixation in the field. In Advances in Nitrogen Cycling in
Agricultural Ecosystems. Ed. J R Wilson. pp 351-367.
C.A.B International. Wallingford. UK.
Ledgard S F and Saunders W M H 1982 Effects of nitrogen
fertiliser and urine on pasture performance and the influence of phosphorus and potassium status. N.Z.J. Agric.
Res. 25. 541-547.
Ledgard S F, Steele K W and Saunders W M H 1982 Effects
of cow urine and its major constituents on pasture properties. N.Z.J. Agric. Res. 25, 61-68.
Ledgard S F, Freney J R and Simpson J R 1984 Variations in
natural abundance of 15N in the profiles of some Australian
pasture soils. Aust. J. Soil Res. 22, 155-164.
Ledgard S F, Freney J R and Simpson J R 1985a Assessing
nitrogen transfer from legumes to associated grasses. Soil
Biol. Biochem. 17, 575-577.
Ledgard S F, Morton R, Freney J R, Bergersen F J and
Simpson J R 1985b Assessment of the relative uptake of
added and indigenous soil nitrogen by nodulated legumes
and reference plants in the 15N dilution measurement of N2
152
Ledgard and Steele
fixation. 1. Derivation of method. Soil Biol. Biochem. 17,
317-321.
Ledgard S F. Simpson J R, Freney J R and Bergersen F J
1985c Field evaluation of 15N techniques for estimating
nitrogen fixation in legume-grass associations, Aust. J.
Agric. Res. 36, 247-258.
Ledgard S F, Simpson J R, Freney J R and Bergersen F J
1985d Effect of reference plant on estimation of nitrogen
fixation by subterranean clover using 15N methods. Aust. J.
Agric. Res. 36, 663-676.
Ledgard S F, Brier G J and Littler R A 1987 Legume
production and nitrogen fixation in hill pasture communities. N.Z.J. Agric. Res. 30, 413-421.
Ledgard S F, Brier G J and Upsdell M P 1990 Effect of
clover cultivar on production and nitrogen fixation in
clover-ryegrass swards under dairy cow grazing. N.Z.J.
Agric. Res. 33, 243-249.
Lowther W L 1975 Interaction of lime and seed pelleting on
the nodulation and growth of white clover. II. Oversown
trials. N.Z.J. Agric. Res. 18, 357-360.
Mariotti A, Mariotti F, Amarger N, Pizelle G, Ngambi J M.
Champigny M L and Moyse A 1980 Fractionnements
isotopiques de l’azote lors des processus d’absorption des
nitrates de fixation de I’azote atmospherique par les plants.
Physiol. Veg. 18, 161-181.
Masterton C L and Murphy P M 1976 Application of the
acetylene reduction technique to the study of nitrogen
fixation by white clover in the field. In Symbiotic Nitrogen
Fixation in Plants. Ed. P S Nutman. pp 299-316. Cambridge University Press, Cambridge, UK.
Matches A G 1989 A survey of legume production and
persistence in the United States. In Persistence of Forage
Legumes. Eds. G C Marten et al. pp 37-44. American
Society of Agronomy, Madison, WI.
Munro J M M and Davies D A 1974 Potential pasture
production in the uplands of Wales. V. The nitrogen
contribution of white clover. J. Brit. Grassl. Soc. 29,
213-223.
Mytton L R 1983 Host plant selection and breeding for
improved symbiotic efficiency. In Temperate Legumes:
Physiology, Genetics and Nodulation. Eds. D G Jones and
D R Davies. pp 373-393. Pitman Advanced Publishing
Programmes, London.
Newbould P 1982 Biological nitrogen fixation in upland and
marginal areas of the U.K. Phil. Trans. R. Soc. Lond. B
296, 405-417.
Norman M J T 1982 A role for legumes in tropical agriculture. In Biological Nitrogen Fixation Technology for
Tropical Agriculture. Eds. P H Graham and S C Harris. pp
9-26. CIAT, Cali, Colombia.
O’Connor M B, Smart C E and Ledgard S F 1990 Long-term
effect of withholding phosphate application on North IsLand hill country: Te Kuiti. Proc. N.Z. Grassl. Assn. 51,
21-24.
O’Hara G W, Daniel R M, Steele K W and Bonish P M 1984
Nitrogen losses from soils caused by Rhizobium-dependent
denitrification. Soil Biol. Biochem. 16, 429-431.
O’Hara G W, Nantakorn B and Dilworth M J 1988 Mineral
constraints to nitrogen fixation. Plant and Soil 108, 93-110.
Parsons A J, Orr R J, Penning P D, Lockyer D R and Ryden
J C 1990 Uptake, cycling and fate of nitrogen in grass-
clover swards continuously grazed by sheep. J. Agric. Sci.
116, 47-61.
Phillips D A and Bennett J P 1978 Measuring symbiotic
nitrogen fixation in rangeland plots of Trifolium subterraneum L. and Bromus mollis L. Agron. J. 70, 671-674.
Refi R O, Navarro C A and Escuder C J 1989 Nitrogen
fixation in mixed pastures under two grazing intensities.
Proc. XVI Int. Grassl. Cong. 181-182.
Rice W A 1980 Seasonal patterns of nitrogen fixation and dry
matter production by clovers grown in the Peace River
region. Can. J. Plant Sci. 60, 847-858.
Robin C, Shamsun-noor L, Schontz D and Guckert A 1989
Etude des relations entre le deficit hydrique et la fixation
d’azote chez le trefle blanc. Proc. XVI Int. Grassl. Cong.
129-130.
Robson A D 1978 Mineral nutrients limiting nitrogen fixation
in legumes. In The Mineral Nutrition of Legumes in
Tropical and Subtropical Soils. Eds. C S Andrew and E J
Kamprath. pp 277-293. CSIRO, Melbourne, Australia.
Robson A D and Loneragan J F 1978 Responses of pasture
plants to soil chemical factors other than nitrogen and
phosphorus, with particular emphasis on the legume symbiosis. In Plant Relations in Pastures. Ed. J R Wilson. pp
128-142. C.S.I.R.O., Melbourne, Australia.
Roughley R J 1976 The production of high quality inoculants
and their contribution to legume yield. In Symbiotic Nitrogen Fixation in Plants. Ed. P S Nutman. pp 125-136. IBP
Vol. 7. Cambridge University Press, Cambridge, UK.
Roughley R J 1985 Nitrogen fixation by the legumeRhizobium symbiosis: The role of the bacteria. In Forage
Legumes for Energy-efficient Animal Production. Eds. R F
Barnes et al. pp 140-145. USDA, Washington DC.
Rys G J and Mytton L R 1985 The potential for breeding
white clover ( Trifolium repens L.) with improved nodulation and nitrogen fixation when grown with combined
nitrogen. 1. The effects of different amounts of nitrate
nitrogen on phenotypic variation. Plant and Soil 88. 181195.
Saunders W M H 1984 Mineral composition of soil and
pasture from areas of grazed paddocks, affected and unaffected by dung and urine. N.Z.J. Agric. Res. 27, 405-412.
Sears P D. Goodall V C, Jackman R H and Robinson G S
1965 Pasture growth and soil fertility. VIII. N.Z.J. Agric.
Res. 8. 270-283.
Shearer G and Kohl D H 1986 Review: N2-fixation in field
settings: Estimations based on natural 15N abundance.
Aust. J. Plant Physiol. 13, 699-756.
Sheath G W and Hodgson J 1989 Plant-animal factors influencing legume persistence. In Persistence of Forage
Legumes. Eds. G C Marten et al. pp 361-374. American
Society of Agronomy, Madison, WI.
Skerman P J, Cameron D G and Riveros F 1988 Tropical
Forage Legumes. F.A.O.. Rome, Italy. 692 p.
Steele K W 1982 Nitrogen in grassland soils. In Nitrogen
Fertilisers in New Zealand Agriculture. Ed. P B Lynch. pp
29-44. Roy Richards, Auckland, New Zealand.
Steele K W 1983 Quantitative measurements of nitrogen
turnover in pasture systems with particular reference to the
role of 15N. In Nuclear Techniques in Improving Pasture
Management. pp 17-35. I.A.E.A., Vienna, Austria.
Steele K W and Bonish P M 1985 Rhizobia1 denitrification in
Nitrogen fixation in legume/grass pastures
temperate grassland soils. In Nitrogen Cycling in Temperate Agricultural Systems. Eds. P E Bacon, J Evans, R R
Storrier and A C Taylor. pp 346-352. Australian Society of
Soil Science Inc., Riverina Branch, Australia.
Steele K W and Percival N S 1984 Nitrogen fertiliser application to pastures under Pinus radiata. N.Z.J. Agric. Res.
27, 49-55.
Steele K W, Watson R N, Bonish P M. Littler R A and
Yeates G W 1985 Effect of invertebrates on nitrogen
fixation in temperate pastures. Proc. XV Int. Grassl. Cong.
450-451.
Sylvester-Bradley R, Mosquera D and Mendez J E 1988
Selection of rhizobia for inoculation of forage legumes in
savanna and rainforest soils of tropical America. In Nitrogen Fixation by Legumes in Mediterranean Agriculture.
Eds. D P Beck and L A Materon. pp 225-234. Kluwer
Academic Publishers, Dordrecht, The Netherlands.
Ta T C and Faris M A 1987 Effects of alfalfa proportions and
clipping frequencies on timothy-alfalfa mixtures. II. Nitrogen fixation and transfer. Agron. J. 79, 820-824.
Vallis I and Gardener C J 1984 Nitrogen inputs into agricultural systems by Stylosanthes. In Agronomy of Stylosanthes. Eds. H M Stace and L A Edye. pp 359-380. Academic Press, Sydney, Australia.
Vallis I, Haydock K P, Ross P J and Henzell E F 1967
Isotope studies on the uptake of nitrogen by pasture
plants. III. The uptake of small additions of 15 N-labelled
fertiliser by Rhodes grass and Townsville lucerne. Aust. J.
Agric. Res. 18, 865-877.
153
Watson R N, Skipp R A and Barratt B I P 1989 Initiatives in
pest and disease control in New Zealand towards improving legume production and persistence. In Persistence of
Forage Legumes. Eds. G C Marten et al. pp 441-464.
American Society of Agronomy. Madison, WI.
Weeda W C 1967 The effect of cattle dung patches on pasture
growth. botanical composition and pasture utilisation.
N.Z. Agric. Res. 10. 150-159.
Wery J, Turc O and Salsac L 1986 Relationship between
growth, nitrogen fixation and assimilation in a legume
( Medicago sativa L.). Plant and Soil 96, 17-29.
Witty J F 1983 Estimating N2 fixation in the field using
15 N-labelled fertiliser: Some problems and solutions. Soil
Biol. Biochem. 15, 631-639.
Witty J F and Minchin F R 1988 Measurement of nitrogen
fixation by the acetylene reduction assay: myths and mysteries. In Nitrogen Fixation by Legumes in Mediterranean
Agriculture. Eds. D P Beck and L A Materon. pp 331344. Kluwer Academic Publishers. Dordrecht, The Netherlands.
Wivstad M, Martensson A M and Ljunggren H D 1987 Field
measurement of symbiotic nitrogen fixation in an established lucerne ley using 15 N and an acetylene reduction
method. Plant and Soil 97, 93-104.
Wood M, Cooper J E and Holding A J 1984 Aluminium
toxicity and nodulation of Trifolium repens. Plant and Soil
78, 381-391.
Plant and Soil 141: 155-175, 1992.
© 1992 Kluwer Academic Publishers. Printed in the Netherlands.
PLSO SA09
Biological nitrogen fixation in mixed legume-cereal cropping systems
K. FUJITA, K.G. OFOSU-BUDU and S. OGATA
Faculty of Applied Biological Science, Hiroshima University, Higashihiroshima City, 724 Japan
Key words: biological nitrogen fixation (BNF), cereal, dry matter production, legume, mixed cropping,
N transfer, sustainable agriculture
Table of contents
Abstract
Introduction
Common mixed-cropping systems
BNF in legume-cereal mixed cropping systems
Factors affecting BNF in cereal-legume systems
Effect of plant type
Combined N effect
Effect of light
Nitrogen transfer from legume to cereal
Factors affecting N transfer
Nitrogen balances in legume-cereal mixed cropping systems
Residual effects of legume-cereal mixed cropping
Improving productivity of legume-cereal mixed cropping systems
Selection of component crops: Important characteristics of cereals and legumes
Culture management
Row arrangement and plant spacing
Time of sowing
Fertility management
Nitrogen application
Phosphorus
Water management
Sustainability issue
Nutrient management
Role of VAM fungi
Use of tolerant species/cultivars
Pest and weed management
Economics
Future directions
References
156
Fujita et al.
Abstract
Cereal/legume intercropping increases dry matter production and grain yield more than their monocultures. When fertilizer N is limited, biological nitrogen fixation (BNF) is the major source of N in
legume-cereal mixed cropping systems. The soil N use patterns of component crops depend on the N
source and legume species. Nitrogen transfer from legume to cereal increases the cropping system’s
yield and efficiency of N use. The use of nitrate-tolerant legumes, whose BNF is thought to be little
affected by application of combined N, may increase the quantity of N available for the cereal
component. The distance between the cereal and legume root systems is important because N is
transferred through the intermingling of root systems. Consequently, the most effective planting
distance varies with type of legume and cereal. Mutual shading by component crops, especially the
taller cereals, reduces BNF and yield of the associated legume. Light interception by the legume can be
improved by selecting a suitable plant type and architecture. Planting pattern and population at which
maximum yield is achieved also vary among component species and environments. Crops can be mixed
in different proportions from additive to replacement or substitution mixtures. At an ideal population
ratio a semi-additive mixture may produce higher gross returns.
Introduction
Mixed cropping is an age-old, widespread practice in the warmer climates of the world (Agboola and Fayemi, 1972; Searle et al., 1981),
especially the tropics (Willey, 1979a). The method allows maximum benefit to be made of natural resources available for production. Total grain
and plant-N yields can often be increased by
intercropping legumes with non-legumes (Barker
and Blarney, 1985; Singh et al., 1986). Most
farmers in developing countries who have adopted this low-input system have done so principally for climatic and socio-economic reasons
(Okigbo and Greenland, 1976). Growing interest
in mixed cropping in developed countries (Ofori
and Stern, 1987) stems from an increasing awareness of environmental degradation arising from
high chemical inputs (Nielson and Mackenzie,
1977) and gives rise to a search for ways to
reduce modern agriculture’s overdependence on
fertilizers, manufactured mainly with use made
of fossil energy.
Crop varieties grown vary by region, depending on several factors including rainfall, and
edaphic and socio-economic factors. Crop mixtures may be legume/legume (Rao and Mittra,
1989) or legume/non-legume (Maldal et al.,
1990). However, mixed cropping of cereals and
legumes is widespread (Ofori and Stern, 1987)
because legumes used in crop production have
traditionally enabled farmers to cope with soil
erosion and with declining levels of soil organic
matter and available N (Scott et al., 1987).
Biological N fixation (BNF), which enables
legumes to use atmospheric N, is important in
legume-based cropping systems when fertilizer N
is limited. Biological nitrogen fixation contributes N for legume growth and grain production
under different environmental and soil conditions. In addition, the soil may be replenished
with N through decomposition of legume residues when BNF contributes more N than the
seeds require. Evidence also suggests that associated cereals may benefit through N transfer from
legumes (Fujita et al., 1990a).
Yield advantages from intercropping as compared to sole cropping are often attributed to
mutual complementary effects of component
crops, such as better total use of available resources. Generally, monocropped legumes have
higher yields than in intercropping systems
(Table 1). However, in most cases, land productivity measured by Land Equivalent Ratio and
monetary gain clearly show the advantages of
mixed cropping of cereals and legumes (Mandal
et al., 1990; Yunusa, 1989).
Depending on the component crops, yield advantages may vary considerably due to several
factors, including differences in plant architecture, rooting patterns, competitive advantages
and potential nitrogen fixing capacity of the
legume. These, in turn, determine the optimum
plant population density, time of sowing and
BNF in mixed legume-cereal systems
157
Table 1. Effects of spacing on amount of N derived from soil. dinitrogen fixation, and transfer by soybean and sorghum 138 days
after planting in mono- and mixed cropping
Cropping
system
Mono
Spacing
(cm)
50 x 50
25 x 25
17.7 x 17.7
12.5 x 12.5
Mixed
50 x 50
25 x 25
17.7 x 17.7
12.5 x 12.5
a
b
Soil
Dinitrogen
fixation
Transfer
Total
amount
of N
(g m-2)
Sorghum
Soybean
Sorghum
Soybean
Sorghum
Soybean
Sorghum
Soybean
3.75
3.75
3.41
3.41
3.17
3.17
3.01
3.01
0
20.28
0
29.29
0
22.85
0
28.67
-
3.75
24.03
3.41
32.50
3.17
26.02
3.01
31.68
Sorghum
Soybean
Sorghum
Soybean
Sorghum
Soybean
Sorghum
Soybean
1.88
1.88
1.71
1.71
1.59
1.59
1.51
1.51
0
14.06
0
11.17
0
10.35
0
8.67
0.89
1.77
1.86
-
2.77
15.94
3.47
12.88
3.45
11.94
3.56
10.18
Crop
Amount of N (g m-2)
derived from
2.05
-
Proportion of
N transferred
to total in
soybean (%) a
Proportion of
N transferred
to total in
sorghum (%) b
-
-
-
-
5.6
13.7
15.6
20.1
32.1
51.0
53.9
57.6
(Amount of N transferred from soybean to sorghum/total amount of N in soybean) x 100.
(Amount of N transferred from soybean to sorghum/total amount of N in sorghum) x 100. (Fujita et al.. 1990a).
amount of fertilizer N. Using five different
legumes intercropped with rice at various population densities and times of sowing, Mandal et
al. (1990) observed significant yield differences.
Tanaka and Fujita (1979) also showed that soybean responded to fertilizer N less than field
bean, perhaps due to differences in potential to
fix N. Legumes that responded well to fertilizer
N had lower N fixation activity relative to their N
demand than those that responded marginally.
Several workers have studied the advantages
of mixed cropping of cereals and legumes (Agboola and Fayemi, 1972; Ofori and Stern, 1987).
Other key papers dealt with the nature of crop
competition and resource use (e.g. de Wit,
1960). Willey (1979a, b) and Francis (1989) have
provided insight in the biological interactions of
mixed cropping systems in general, while Ofori
and Stern (1987) dealt with cereal/legume intercropping systems in particular. More recently, a
number of papers have been published on dinitrogen fixation and N release and, more importantly, the need to develop legume-based
technologies to sustain agricultural production.
In the present paper, mixed cropping of cereal
and legumes will be emphasized, and recent
information and its implications with respect to
the full use of BNF will be reviewed. The physiological relationship between dinitrogen fixation
and N release and the N economy of the total
system are considered. This review is by no
means complete, but is an attempt to develop
new ideas to combat environmental degradation
and food production problems.
Common mixed-cropping systems
Growing legumes and cereals together for food
production is not only popular among peasant
farmers in the tropics, who produce the bulk of
food in developing countries, but is also expanding to warmer regions in the subtropics. This is
illustrated by the vast areas of the world stretching from South Western Japan, 34° 40' N', (Fujita et ai., 1990a, b) to Western Australia,
32°51'S, (Ofori and Stern, 1986), where mixed
cropping of legumes and cereals is either practiced or has been studied. Mixed cropping is
practiced under different climatic regions such as
the humid tropics (Agboola and Fayemi, 1971);
semi-arid regions (Faris et al., 1983, 1990)
158
Fujita et al.
mediterranean regions (Ofori and Stern, 19861,
and in temperate climates (Fujita et al., 1990a).
The system's benefits are realized in areas where
the rainy season is long and favorable enough to
grow more than one crop of different maturation
periods simultaneously or successively (Okigbo
and Greenland, 1976), or where irrigation is
available (Mandal et al., 1990). Recent reports
indicate benefits even under dry and unpredictable rainfall conditions (Ntare, 1989; Papastylianou, 1990).
Legumes are a major component in cropping
systems of developing countries. Several species
are used throughout the wet and dry tropics,
both in sole-cropping and mixed-cropping practices (Rachie and Roberts, 1974). It has been
estimated that high proportions of basic cereals
and pulses are produced in various legume-based
mixed cropping systems in different parts of the
world. Included are 76% of maize, 90% of millet, 95% of peanut and 99% of cowpea in
Nigeria (Okigbo and Greenland, 1976), and 84%
of maize, 56% of peanut, 81% of beans, and
76% of pigeon peas in Uganda (Jameson, 1970).
In the Latin American tropics, between 80 and
90% of beans produced are in mixed cropping
systems (Francis et al., 1976). In the Indian
subcontinent, pigeon pea is in almost every cropping system (Patra and Chatterjee, 1986). Intercropping maize with soybean for silage production is a common practice to increase on-farm
protein production without severely reducing
total forage dry matter yield, and all intercrop
patterns produce more dry matter than monocultures (Herbert et al., 1984). Protein concentration was increased from 69-81 g kg-1 for
maize monoculture to 88-108 g kg-1 for the various intercropping patterns. Amount and distribution of rainfall, soil fertility, socio-economic
and other cultural factors influence the composition of the component crops. In low and unpredictable rainfall areas where irrigation is unavailable, early maturing and/or water stress-tolerant
legumes and small-grained cereals are cultivated.
Examples are finger millet and green gram in
India (Kaushik and Gautam, 1987), oats and
vetch in Cyprus (Droushiotis, 1989), cowpea and
sorghum in the sub-Saharan West African subregion (Ntare, 1989). In seasons or areas where
rainfall is abundant, maize and rice are often
planted with such legumes as field bean, cowpea
and soybean (Ezumah et al., 1987).
The first of the two main methods of mixedcropping cereals and legumes is superimposition
of one crop on another as described for maize/
beans (Fischer, 1977) where the total plant
population per unit area is higher than the optimum population in the monoculture. Second is
replacement type, where various proportions of
one component crop replace the same proportions of the other component crop in the mixture
(Yunusa, 1989). The second system is receiving
increased attention of scientists because of reduced intercrop competition and increased yield.
BNF in legume-cereal mixed cropping systems
Various direct and indirect methods have been
used to evaluate how much N the legumeRhizobium symbiosis contributes to legumebased ecosystems (Ofori and Stern, 1987; Peoples and Herridge, 1990). Advantages and disadvantages of the various BNF assessment methods are discussed by Peoples and Herridge
(1990). Briefly, the current methods used to
assess BNF are: 15N-isotopic techniques, Ndifference method, ureide method, N balance,
acetylene reduction assay (ARA), N fertilizer
equivalence, and nodule evaluation.
Factors affecting BNF in cereal-legume systems
The amount of N fixed by the legume component
in legume-cereal intercropping systems depends
on several factors, including species, plant morphology, density of component crops, type of
management, and competitive abilities of the
component crops (Ofori and Stern, 1987). Variation in BNF activities of various legumes have
been reported with both mono- and mixed cropping systems.
Effect of plant type
Indeterminate legumes fix more N than determinate types in intercropping. Graham and Rosas
(1978) and Francis (1986) observed that BNF in
climbing bean was unaffected by intercropping
with maize. BNF (ARA) of a determinate soybean type decreased when intercropped with sor-
BNF in mixed legume-cereal systems
159
the N economy of a maize-cowpea intercrop
system using both 15 N natural abundance and
15 N-dilution methods and reported that cowpea
could fix atmospheric nitrogen when intercropped with maize, but that N fertilizer applications
reduced N fixation.
The role of soil N (organic matter) in the N
nutrition of a sorghum/soybean intercropping
system has recently been studied in soils differing
in N amounts (Ofosu-Budu et al., unpublished
data) using the 15 N-dilution method. Evidence
suggests that on soil with a relatively high N
content (high organic matter) the mixed cropping yield increased by 25% due to enhanced soil
N uptake by the sorghum component, while the
soybean component depended mostly on BNF.
Dinitrogen fixation and N transfer were about
35% higher in the high-N soil.
ghum (Fujita et al., 1990a). However, intercropping siratro, a climbing legume, with sorghum
had little effect on BNF (Ogata et al., 1986).
Field-grown cowpea was estimated to receive
53-69% of its total N from BNF, which was
changed little by intercropping (Ofori et al.,
1987; van Kessel and Roskoski, 1988) (Table 2).
Higher BNF activity of ricebean in the intercrop
than in the monocrop was mainly due to its
vigorous climbing habit. The contribution of N2
fixation to ricebean N yields was greater in intercrop than in monocrop (see Fig. 2) even at the
lowest maize-ricebean ratio (Rerkasem et al.,
1988). Comparable fixation by sole cowpea was
higher, but this advantage was outweighed by
greater land use efficiency of intercrops than sole
crops. Shading by associated cereals reduces
BNF by component legumes (Wahua and Miller,
1987b). In fact, partial defoliation of component
sorghum increased sunlight availability and increased the BNF of the component groundnut
crop (Nambiar et al., 1983).
Effect of light
Because dinitrogen fixation is energy-dependent,
reduction in the photosynthate supply to the
nodules is detrimental. If the non-legume is taller than the legume, shading occurs and results
in reduced photosynthesis and dinitrogen fixation
(Trang and Giddens, 1980; Wahua and Miller,
1978). Total N fixed in a cowpea/maize system
at different spacing was more dependent on the
type of cropping system (generally lower in
mixed than in monocrop cowpea) than on the
crop spacing. When planted with maize, cowpea
fixed about 50% of the nitrogen fixed by the
Combined N effect
With no applied N, shading did not affect N2
fixation by the component groundnut crop although incoming light reaching the legume was
reduced 33% (Nambiar et al., 1983). When
50 kg N ha -1 was applied, BNF was reduced
55%, although light reaching the groundnut was
54% of incoming radiation. This suggests that
heavy application of combined N significantly
reduces BNF. Ofori and Stern (1987) evaluated
Table 2. Percentage of N and kg N derived from N2 fixation by cowpea grown under different cropping systems
kg Ndfa ha -1
Cropping
system
% Ndfa
50 DAP a
no DAP
SO DAP
80 DAP
Mb (40)
M (50)
M (60)
M (80)
M ( 100)
M (120)
I (40)
I (SO)
I (60)
41.4abcdc
23.6e
40.3abcd
36.6bcde
31.4cde
25.3de
48.5ab
51.7a
47.9abc
47.2a
5.8a
28.2a
42.7a
34.8a
30.h
42.6a
44.2a
34.4a
34.4aa
18.8b
33.9a
20.7b
15.8h
11.5b
14.1b
16.3b
1S.3b
42.22a
40.0a
21.8bc
29.6ab
19.5bc
18.1bc
12.3c
12.2c
10.0c
a
DAP = days after transplanting.
M = monocropped, I = intercropped at different row spacing. Values between parentheses indicate row spacing in cm.
c Means with different letters within a column are significantly different at the p = 0.05 level based on Duncan's Multiple Range
Test (Van Kessel and Roskoski. 1988).
b
160
Fujita et al.
monocrop (Ofori and Stern, 1987; van Kessel
and Roskoski, 1988).
Plant density has also been reported to influence dinitrogen fixation, but total nitrogenfixing activity on an area basis appeared less
15
variable. Using the
N-dilution method, van
Kessel and Roskoski (1988) reported that the
percentage of total N derived from N2 fixation in
cowpea was largely independent of spacing and,
overall, cowpea derived from 30 to 50% of its N
from BNF. Both reports indicate plant density
has little effect on quantity of N derived from
dinitrogen fixation. More importantly, the BNF
of the legume is not always reduced, but is
dependent on the legume's ability to intercept
light.
Nitrogen transfer from legume to cereal
Symbiotically fixed N has been considered a
useful source of N to non-fixing plants (Virtanen
et al., 1937) in mixed cropping systems. This N
transfer is considered to occur through root excretion, N leached from leaves, leaf fall, and
animal excreta if present in the system. Evidence
suggests that N2 fixed by a legume component
may be available to the associated cereal in the
current growing season (Brophy and Heichel,
1989; Eaglesham et al., 1981; Ta et al., 1989) or
as residual N for a subsequent cereal crop (Searle et al., 1981; Singh, 1983). Both current and
residual N transfer are important and could improve the N economy of legume-based intercropping systems. Other researchers have reported
little or no current N transfer in legume/cereal
mixed cropping (Ofori and Stern, 1987; Rerkasem and Rerkasem, 1988; van Kessel and
Roskoski, 1988). This suggests that N transfer
may occur only under certain conditions.
Eaglesham et al. (1981), using the 15N-labelled
fertilizer method, observed an increase in N
concentrations in maize when intercropped with
cowpea by the transfer of N from cowpea to
maize as indicated by significant dilution of 15N
in the component maize crop. Substantial N
transfer from component legume to associated
cereal was observed in a wheat/gram and maize/
cowpea mixed cropping system (Patra et al.,
1986). This observation was not confirmed by
Ofori et al. (1987) in an 15N enrichment study in
a cowpea/maize mixed cropping system or by
van Kessel and Roskoski (1988) at different
planting distances.
The idea of N transfer in the current season
from legumes to component cereals originated
from earlier studies on legume/nonlegume mixtures in pots under greenhouse conditions (Virtanen et al., 1937). Cereals grown with legumes
receive some of their N from the legumes
(Broadbent et al., 1982), so that less soil N is
removed than under cereal monoculture. Agboola and Fayemi (1972) observed that earlymaturing legumes, such as green gram, improved
yields and N nutrition of associated maize in the
current season, while such benefits were not
observed with late-maturing plants such as cowpea ( Vigna unguiculata L.) and calapogonium
( Calapogonium mucunoides Desv.).
Remison (1978) found a 72% increase in intercrop maize grain yield over that of sole maize in
a maize-cowpea combination, and Waghamare et
al. (1982) observed an increase in grain yield and
grain protein of intercrop sorghum when grown
with green gram, groundnut, soybean, or cowpea. Agboola ad Fayemi (1972) found that 3%
of the N fixed by green gram (Vigna radiata L.)
was released into the root zone, and Eaglesham
et al. (1981) showed that 24.9% of N fixed by
cowpea was transferred to maize. Brophy and
Heichel (1989) observed a release of 10.4% of
symbiotically fixed N in soybean ( Glycine max
(L.) Merr. Cv. Fiskeby). Release of about 30%
of fixed N by the soybean root system into the
nutrient culture medium has been observed
(Ofosu-Budu et al., 1990).
Nitrogenous compounds such as amino acids,
proteins, and peptides were identified in leachates from root zones of legume seedlings grown
under sterile sand conditions (d'Arcy, 1982;
Wacquant et al., 1989). Ofosu-Budu et al. (1990)
found that about 10% of the N released was in
ureide form at the different growing stages
studied. They suggested that some of the recently fixed N is released, although no direct relationship between ureide excretion and dinitrogen
fixation was observed. Furthermore, no significant difference in N compounds release was
observed in soybean cultivars Bragg and its mutants, nts 1007 and nts 1116, which differed sig-
BNF in mixed legume-cereal systems
nificantly in dinitrogen fixation (Ofosu-Budu et
al. , unpublished data).
In general, the root zone immediately behind
the root tip is considered the major site of
exudation (Pearson and Parkinson, 1961), but
different sites have been reported for different
plant species (Schroft and Snyder, 1962). In
soybean, Ofosu-Budu et al. (1990) found that
the major N release is from roots, with negligible
release from nodules. Root caps of many plants,
including legumes, shed cells that become part of
a mucilaginous sheath that surrounds the growing root (Rougier, 1981). These shed cells, products of root cap turnover, have traditionally
been referred to as 'sloughed' root cap cells
because it was believed they were dead (Paull
and Jones, 1976). However, recent reports suggest that the shed cells exhibit 90 to 100% viability, and in pea ( Pisum sativum L.) on average
3,400 cells per root in water-culture conditions
were observed (Hawes and Lin, 1990). Assuming that these cells release their N content, the N
release by legumes to cereals could be substantial.
Factors affecting N transfer
N release by legume root systems is not well
understood, but there is indication that recently
fixed N is released. Recent reports relating N
release with dinitrogen fixation are conflicting.
Using 15N tracer, Ta et al. (1986) reported that
recently fixed N was the major source of N
release. However, Ofosu-Budu et al. (1990) reported that ureide released by soybean formed
only 10% of the total N released. Ureide has
also been found to be released in other ureidetransporting legumes such as siratro ( Macroptilium atropurpureum DC. cv. siratro) and desmodium ( Desmodium intortum. (Mill) cv.
Greenleaf), but not in sesbania ( Sesbania
capitata Pers.), suggesting that recently fixed N is
released into the environment (Ofosu-Budu et
al., unpublished data). Soluble proteins made up
a larger portion of the excreted N, confirming an
earlier report with alfalfa (Brophy and Heichel,
1989).
During subsequent experiments, Fujita et al.
(1990b) found that out of three treatments that
161
decreased BNF in soybean (pod removal, defoliation, and stem girdling), only stem girdling
promoted N release. They suggested that a decrease in root and nodule sugar concentration
alone, common to defoliated and stem-girdled
treatments, is not the sole factor that stimulated
N release, but could be coupled with a high
concentration of nitrogen fixation compounds. In
an investigation of N transfer to associated grass,
using 15N tracer in a split-root system, Catchpole
and Blair (1990) reported that in Leucaena spp.
severe defoliation did not promote N transfer.
Variation in N release among legumes (OfosuBudu et al., unpubl. data; Ta and Faris, 1989)
and their release patterns have been reported
(Agboola and Fayerni, 1972). Benefits to associated cereals in mixed cropping systems have
been suggested to be due to factors such as
component crop densities, which determine the
closeness of legume and non-legume crops. Increased N transfer was reported for sorghum that
was closer to associated soybean. The benefit
decreased as the distance increased (Fujita et al.,
1990a). In a soybean/sorghum intercrop system,
N transfer increased from 0.89 g N m-2 at 5 x
50 cm spacing to 2.05 g N m-2 at 12.5 X 12.5 cm
spacing. Nitrogen transfer was 5.6 and 20.1% of
soybean N at 50 x 50cm and 12.5 X 12.5 cm
spacings, respectively.
Legume growth stages may influence the N
release rate. For unharvested alfalfa, the peak in
ninhydrin-N release was near midgrowth cycle
(Richter et al., 1968). Brophy and Heichel
(1989) showed an increase in amino acids released by sterile alfalfa for up to 7 or 8 weeks of
growth. Ofosu-Budu et al. (1990) also reported
that N release by the soybean root system was
higher during the pod-filling stage. They attributed this to the relative increase in root size, but
the physiological basis could not be established.
Shoot harvest could cause passive leakage of
recently fixed N because it removes the aerial
sink for N. Whitney and Kaneshiro (1967) observed considerable N loss at the first harvest,
followed by smaller responses at later harvests in
the tropical pasture species desmodium and centrosema. A similar pattern of ninhydrin-N loss
from harvested plants was observed in alfalfa
(Brophy and Heichel, 1989).
Although N release mechanisms are unclear,
162
Fujita et al.
some physiological and environmental factors
appear to stimulate N loss from roots. The possible role of pectolytic enzymes in living root cell
release has been proposed (Hawes and Lin,
1990).
Furthermore, factors that promote N release
may be specific. High temperature (35°C) led to
an increase in N released from soybean, sesbania, and Chinese milk vetch roots (Ofosu-Budu
et al., unpublished data). However, lower temperatures (25° and 15°C) had no effect. Vancura
and Stanek (1975) and Brophy and Heichel
(1989) reported an increase in the release of
various materials from roots after water stress.
Possible reduction in transpiration resulting from
water deficit and subsequent build-up of recently
assimilated N in roots and nodules may result in
passive loss of soluble N. Water stress is said to
affect root cell membrane permeability (Hale et
al., 1978). Iron (Whitney and Kaneshiro, 1967)
and P stresses have also been thought to promote N release. It has also been suggested that
release of some substances from the cereal component could stimulate N release by legumes (Ta
and Faris, 1987; Wacquant et al., 1989).
Nitrogen balances in legume-cereal mixed
cropping systems
The main nitrogen sources in cereal/legume systems are N fixed through BNF by the legume
component, fertilizer N and soil N. The only
published data illustrating N budgeting are on
studies conducted with maize and cowpea by
Eaglesham et al. (1981) in Western Nigeria
(tropics) and Ofori et al. (1987) in Western
Australia (mediterranean). Using the equation
suggested by Rennie et al. (1982) to calculate N
from fixation, fertilizer, and soil, Eaglesham et
al. (1981) prepared a N balance sheet for the
system. The N contribution by seeds of maize
and cowpea at sowing was less than 2 kg ha-1,
fixed N by component cowpea was about
41 kg ha-1, N from fertilizer was 3 kg ha-1 and
soil N was 53 kg ha-1, with total N in the crop at
about 99 kg ha-1. 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-1 (28 kg ha-1 from maize and
24 kg ha-1
from
cowpea)
leaving
about
46 kg ha-1 in the residues.
The resulting net changes in soil N after the
grain harvests and the return of residues are
calculated as N = N (residues) - N (uptake from
soil). The maize-cowpea intercrop would result
in a loss of 14 kg N ha-1 to the soil, compared to
a 21 kg N ha-1 loss after sole cropping of maize
and a 36 kg N ha-1 gain after sole cowpea.
The data of Eaglesham et al. (1981) indicated
that, compared to cowpea monocropping, intercropping maize and cowpea does not excessively
deplete soil N. Cowpea monocropping might
enhance soil nitrogen status and could benefit a
subsequent cereal in a legume/cereal crop rotation, provided the high N content stover is returned to the soil, while sole cropping maize
depletes the soil nitrogen. Similar observations
were reported by Ofori et al. (1987).
Residual effects of legume-cereal mixed
cropping
Beneficial effects of mono-and intercropped
legumes on subsequent cereal crops are well
documented (Papastylianou, 1988). Various factors, such as increase in organic matter, improved soil structure, and, most importantly,
increase in soil N, might account for this phenomenon. Cereals, on the other hand, cannot
replenish the soil with N because it requires an
external N source for optimum growth.
Nair et al. (1979) found wheat yield to increase 30% after a maize/soybean intercrop and
34% after maize/cowpea compared to wheat
planted after sole maize. De (1980) evaluated
residual N of various legume-based intercrop
systems and found that in all N treatments black
gram intercropped with either maize or sorghum
improved succeeding wheat yields.
Higher N uptake by wheat was observed when
it followed maize /groundnut or maize/soybean
intercrop systems than after maize alone. The N
uptake without fertilizer N application by a subsequent wheat crop after cropping maize was
12 kg N ha-1, after maize/soybean 19 kg N ha-1,
after maize/peanut 46 kg N ha-1, and after
peanut 54 kg N ha-1 (Searle et al., 1981). This
shows that a subsequent crop could benefit as
BNF in mixed legume-cereal systems
much from following one of the maize/legume
intercropping systems with no fertilizer N applied as from planting after a sole-maize crop
applied with 100 kg N. Singh (1983) estimated N
benefits to wheat derived from various preceding
legume intercrops. Comparing wheat after sole
sorghum with wheat after intercrop, he obtained
N fertilizer equivalents of 3 kg ha-1 with soybean, 31 kg ha-1 with green gram, 46 kg ha-1
each with grain cowpea and groundnut,
54 kg ha-1 with fodder cowpea. N uptake by a
succeeding crop, when 100 kg N ha-1 was applied
to the preceding crop, was always higher following sole or intercropped cowpea. This could be
due to less immobilization of the freshly applied
fertilizer N by legume crop residues rich in N
(Patra et al., 1989).
In Cyprus, a semi-arid region, Papastylianou
(1990) examined the response of a mixture of
oats ( Avena sativa L.) and two legumes, vetch
( Vicia sativa L.) and peas ( Pisum sativum L.), to
N fertilization and the residual effect on subsequent barley ( Hordeum vulgare L.). He reported that grain and N yields of the barley crop
were higher after the legumes than after oats,
with intermediate yields after the mixtures. Furthermore, the residual effect of N fertilizer on
subsequent cereals also depended on rainfall.
With very low rainfall (234 mm), a residual effect
of the fertilizer on subsequent barley yield was
observed. However, very little or no residual N
effect was observed when high rainfall occurred.
It was concluded that a low or no residual fertilizer effect in high-rainfall periods was due to
leaching losses.
Beneficial effects of mixed cropping cereals
and legumes on the N nutrition and yield of the
subsequent crop has also been studied in India.
The preceding intercrop of pearl millet with
several legumes increased the yield of succeeding
wheat ( Triticum aestivum L.) from 7 to 17%
compared to sole pearl millet (Patil and Mahendra Pal, 1988). They claimed 80 kg N ha-1was
saved for the succeeding bread wheat by the
preceding intercrop of pearl millet with black
gram or cowpea.
A six-growing season study in Thailand
showed improvement of soil organic matter
under mixed cropping. Soil organic matter content in a maize/ricebean intercrop study doubled
163
to 1.64% over this period, while it changed only
slightly, from 0.83 to 1.10%, under mungbean
after a maize cropping system (Phetchawee et
al., 1986).
Nitrogen in crop residues can contribute a
significant amount to the next crop and, compared to cereals, residues from leguminous crops
often contribute substantial amounts of N
(Power and Doran, 1988). Data on N availability
from 15N-labelled rice, soybean and wheat to
subsequent rice showed the minimum estimate of
residue-N mineralized from the time of residue
incorporation until harvest to be 9% of rice,
52% of soybean. and 33% of wheat (Normal et
al., 1990). Residue-N recovered in the subsequent rice crop was 3% of the rice, 11% of the
soybean and 37% of the wheat residue. The
higher the C/N ratio and the amount of N in the
residue, the higher the amount of residue-N
mineralized. Loss of easily decomposed N fractions due to the state of the soybean residue and
the high N concentration of the wheat residue
might explain why wheat residue contributed
more N than soybean residue.
Improving productivity of legume cereal mixed
cropping systems
Selection of component crops : Important
characteristics of cereals and legumes
When grown in association, basic physiological
and morphological differences between cereals
and legumes affect their mutual relationship.
Cereals are taller. have a larger mass of fine
roots, and are adaptable to a wide range of
environmental conditions, including low soil fertility. In mixed cropping, the component with its
leaves higher in the canopy structure is at an
advantage, particularly if the leaves are broad
and horizontal (Trenbath, 1976). Optimum conditions must be attained to realize good growth
for legumes, especially if they will derive most of
their N from BNF.
The role of root cation exchange capacity
(CEC) in ion accumulation by plants is controversial (Nye and Tinker, 1977), although the
concept has been used to explain the low competitive ability of legumes in uptake of ions such
164
Fujita et al.
as P, K and S as compared to cereals. The root
CEC of legumes is about twice that of cereals
and legume roots have been suggested to absorb
more divalent cations (such as Ca) than those of
cereals with low CEC (Caradus, 1990). Rabotnov (1977) attributes the legume inability to
compete for P, K and S to their root system,
which is less ramified that that of cereals.
Soil N uptake by ricebean, being 1.3 gNm -2
in intercropping against 6.4 gN m -2 in monoculture, (Rerkasem et al., 1988), and by cowpea
(van Kessel and Roskoski, 1988) was markedly
suppressed when intercropped. When monocropped at the same row spacing, both cowpea
and maize took up equal amounts of soil and
fertilizer N (van Kessel and Roskoski, 1988).
The uptake of fertilizer N is dependent on cropping system and row spacing. Highest N uptake
occurred in those cropping systems with the
highest plant population. No significant differences were found between monocropped maize
and cowpea and the sum of intercropped maize/
cowpea at the same row spacing. Component
maize at the wider row spacing took up more
fertilizer-N than maize at the narrowest spacing.
Soil N uptake was also a function of row spacing
rather than of crop combination (van Kessel and
Roskoski, 1988). Patra et al. (1989) examined
effects of intercropping on the N use efficiency in
India. They reported that applications of
100 kg N ha -1 as urea was used more efficiently
by intercropped maize/cowpea than by monocropped maize.
The rate of crop dry matter production depends mainly on efficiency of photosynthetically
active radiation, PAR, (Biscoe and Gallagher,
1977). Amount of light intercepted by the component crop in an intercrop system depends on
the geometry of the crop and foliage architecture
(Tsay, 1985). Generally the taller cereals shade
the legume and cause reduced growth and yield
of the legume at high densities. Component
crops increase in height in mixed cropping systems, probably competing for light. Differential
increase in plant height of component crops as a
result of mixed cropping increases with increasing density (Fujita et al., 1990a). The magnitude
of the increase (plastic response) was more pronounced in sorghum than in soybean, suggesting
that sorghum is better equipped to compete for
light by elongating the stem and developing
higher-positioned leaves (Fig. 1). When consid-
Fig. 1. Dry matter production and light transmission ratio at different heights at 17.7 x 17.7 cm in mono- and mixed cropping on
116 days after planting. a: leaf blades (major photosynthetic part); b: stems + petioles a (leaf sheath) + reproductive parts
(non-photosynthetic part), ; light transmission ratio (LTR). a In soybean. A: soybean monocrop; B: mixed cropping soybean and
sorghum, C: sorghum monocrop. soybean
;sorghum £ ; (Fujita et al., 1990a).
BNF in mixed legume-cereal systems
competition for light, this plant characteristic
may be important in determining optimum yield
densities for various cereal/legume systems. In
addition, most of the cereal components in intercrops, such as maize and sorghum, are C4 plants,
and the legumes are C4 plants. Since cereals and
legumes differ considerably in their responses to
light, such factors should be taken into consideration, when component crops and their population densities are selected.
Yield change in a maize-cowpea intercropping
system in response to fertilizer N application was
a reflection of maize and cowpea architecture
and growth habits (Ezumah et al., 1987). Maize
grain yields increased 62% with N rates from 0
to 120 kg N ha-1, while average cowpea yield
decreased 27%. The early maturing, determinate, semi-erect TV x 3236 cowpea cultivar did
not respond to applied N, while the indeterminate, photoperiod-sensitive, spreading Vita 5
yield decreased with increasing N.
Cultural management
The influence on yield and production efficiency
of component crop densities and manipulations
of spacing between component crops, such as
row arrangement and inter-row spacing has been
evaluated by Ofori and Stern (1987). They suggested that the cereal component was usually
little affected by these manipulations, while the
legume yield usually decreased significantly depending on the proximity of the cereal, perhaps
due to the top of the legume canopy being
shaded. However, these trials were mostly carried out using high levels of fertilizer N promoting heavy leaf production. Ofori and Stern
(1987) concluded that, although the cereal usually contributes a larger portion of total yield, the
legume seems to determine the magnitude of the
intercropping advantage or efficiency.
Row arrangement and plant spacing
Overall mixture densities and proportions of
component crops determine yields and production efficiency of cereal/legume intercrop systems (Lakhani, 1976; Willey and Osiru, 1972).
In systems with equal numbers of component
crops, the more aggressive crop, usually the
cereal, appears to determine productivity ef-
165
ficiency (Lakhani, 1976; Osiru and Willey,
1972). When maize and cowpea were planted
alternatively in the same row rather than in
alternate rows, grain yield and WUE were significantly higher. However, no difference in
evapotranspiration was observed (Hulugalle and
Lal, 1986). Planting maize and cowpea in alternate rows did not affect maize yield, but planting
cowpea in the same row with maize increased
maize yield by 2.23%. Compared with seeding
sole maize or cowpea, water use efficiency
(WUE) increased 100% when seeding maize and
cowpea alternatively in the same row and 71.4%
when seeding in alternate rows. This suggests
that plant arrangement determines how effectively available resources are used, especially soil
water.
Mohta and De (1980), from maize/soybean
and sorghum/soybean intercrop trials, reported
that cereal yields were little affected by intercropping with soybean, when arranged in either
single or double alternate rows. The intercropped soybean yield increased 31% when component crops were arranged in double alternate
rows rather than single rows. Waghamare et al.
(1982) found that light penetration was greatly
increased in the wider row arrangement. After
60 days, light incident on the intercrop legume
canopy was 52% of the incoming radiation with
sorghum rows spaced at 60cm, and 70% when
sorghum was spaced at 90cm.
In maize/soybean mixed cropping at different
proportions and arrangements, Yunusa (1989)
reported that reducing the maize proportion in
the mixtures improved the soybean yield. Increasing the proportion of any component crop
in the mixtures, up to 100% for maize and 67%
for soybean, increased total yield. Yunusa (1989)
concluded that in a' mixed cropping system of
maize and soybean, the crops should be sown in
alternate ridges at 67% of the recommended sole
crop densities of 5 plants m-2 for maize and 20
plants m-2 for soybean.
In a recent study (Fujita et al., 1990a), the
proportion of dry matter contribution of
sorghum was lower at 50 x 50cm (40,000
plants ha-1), but it increased with population
density and surpassed the soybean contribution
at spacings closer than 25 X 25 cm (160,000
plants ha-1). Intercrop efficiency increases with
166
Fujita et al.
high population densities, as found in a
sorghum / bean intercrop (Osiru and Willey ,
1972) and in a maize/cowpea intercrop (Fawusi
et al., 1982)
Increasing population density reduced the
amount of light available to the soybean component at different plant spacings, from 50 X 50 cm
to 12.5 x 12.5 cm, equivalent to 40,000 plants
ha"to 640,000 plants ha-1 (Fujita et al., 1991a).
When competition for light is intense in the
cereal/legume intercrop system, the taller cereal
is at an advantage over the legume. For example, at 50 X 50 cm competition for light between
sorghum and soybean was slight, and about half
of the sorghum leaves were above the soybean
leaves. At a 17 X 17 cm spacing, when competition for light became more intense, both plants
became taller and developed their leaves at higher positions. At a low density (40,000 plants
ha-1), the light transmission ratio (LTR) at
ground level was 3% for the intercrop, 45% for
sole-crop sorghum and 20% for sole-crop soybean. Their leaf area indices (LAI) were 2.0 and
1.7 for intercropped soybean and sorghum, 3.8
for sole crop soybean, and 1.3 for sole crop
sorghum. No difference between intercrop and
sole crop soybean was observed. At the high
density (320,000plants ha-1), the LTR was 2%
for intercrop, 6% for sole sorghum, and 10% for
sole soybean. LAI were 5.0 and 6.7 for intercrop
soybean and sorghum, 5.1 for sole soybean, and
3.3 for sole sorghum.
Efficiency in cereal/legume mixed cropping
systems could be improved by reducing interspecific competition between component crops
for limiting growth factors (Willey, 1979a). The
cereal is the dominant component and the
legume the dominated component (Huxley and
Mainger, 1978). The legume yield is generally
reduced in intercropping, but at varying levels
depending on the species and plant type. Ofori
and Stern (1987) reported that the average
legume component yield declined to 52% of the
sole crop, while the cereal component yield decreased by only 11%. The LER was 1.91 without
applied N and 1.43 at 120 kg N ha-1.
Time of sowing
Relative yield of the legume component has
been reported to increase if planted before the
cereal component. According to Francis (1986),
bean yield was reduced more than 50% when
maize was planted before the beans. However,
when the legume component was planted 15 days
earlier than the cereal, a higher yield was obtained. At the same density the intercrop soybean's total dry matter yield was 51% of that of
sole soybean.
The relative time of sowing component crops
is often manipulated in cereal/legume intercrop
systems. A 15-day advantage for the bean component gave higher yields than when maize and
beans were planted together (Francis, 1986).
Bean yield was reduced to 50% of that of the
monocrop when planted at the same time. Ofori
and Stern (1987), however, concluded that variation in time of sowing on intercrop yields has no
advantage over simultaneous sowing. In staggered sowing, the earlier planted component has
an initial advantage over the later planted component. At maturity, yield loss due to later
sowing of the component crop could not be fully
recovered.
Mandal et al. (1990) observed yield differences when various legumes and rice were intercropped simultaneously and when rice planting
was deferred 30 days after the legume component. They concluded that peanut, soybean,
mungbean, and blackgram were profitable when
intercropped simultaneously with rice, but rice
with blackgram in deferred planting was more
remunerative under West Bengal conditions.
Yield differences among the various combinations were attributed to differences in growth
habits, acquisition of nutrients, etc. It is apparent that each species, especially the legume's
growth habits and plant architecture, must be
considered when deciding to defer planting of
any component crop.
Fertility management
In legume/cereal mixed cropping systems,
edaphic factors such as nutrients and water frequently determine productivity. Under rainfed
or irrigated conditions, water supplies may be
inadequate and/or poorly distributed. Moisture
availability is linked to mineral nutrition. Mixed
cropping is often more productive when component crop growth is not limited by the same
BNF in mixed legume-cereal systems
167
Ofori and Stern (1987) reviewed the influence
of applied N on various intercropping systems.
They found that intercrop cereal yields increased
progressively with N application, while seed
yield of the legume either decreased or responded less. They concluded that N application
did not improve the land equivalent ratio and,
thus, the efficiency of the cereal/legume intercrop system. Cowpea in mixtures was unable to
derive all its N from N2 fixation. Chang and
Shibles (1985) attributed this to a shading effect.
They observed that under high fertilization
(N, P) mono- and mixed cultures obtained the
same yields. However, under low fertilization
more land planted in monocultures would be
needed to produce the same yield as a hectare of
mixed cropping.
Searle et al. (1981) in a maize/groundnut intercrop with N application up to 100 kg N ha-1,
found that grain yield of intercropped maize
increased progressively with increasing N application, while groundnut seed yields decreased.
Intercropping efficiency measured by LER was
1.36 at no fertilizer application and 1.24 at
100 kg N ha-1. When N was applied in a sorghumipigeon pea system, Rego (1981) reported
a significant increase in intercropped sorghum
yield, with no significant effect on pigeon pea.
resources. In cereal/legume mixed cultures generally the relative competitiveness of component
species changes with management practices such
as fertilizer application and plant population,
that are mostly associated with variations in the
ability to compete for light (Rhodes and Stern,
1978). Even when light competition is not intense, heavy fertilization may inhibit the way the
component crops complement each other's resource use (Hall, 1978).
Nitrogen application
Intercropped maize ( Zea mays L.) and ricebean
( Vigna umbellata Thumb [Ohwi and Ohashi])
under a constant planting density of 8 maize and
16 ricebean per m-2 and varied levels of combined N application under rainfed conditions was
evaluated in Northern Thailand (Rerkasem and
Rerkasem, 1988). Combined N application ranging from 0 to 200 kg N resulted in significantly
higher dry matter, grain, and N yield of intercropped maize and ricebean as compared to their
monocrop yields (relative yield). The intercrop's
advantage was speculated to be associated with
N nutrition, enhancement of dinitrogen fixation,
and efficient use by the maize crop of mineral N
(Rerkasem and Rerkasem, 1988; Rerkasem et
al., 1988) (Table 3).
Table 3. Sources of nitrogen (in kg N ha-1) in intercropped and monocultured maize and ricebean
kg N ha-1
added
M: R
a
Maize
F-N b
Ricebean
S-N
F-N
Cropping system total
S-N
-
Fixed N
-
F-N
S-N
Fixed N
T-N
68.32
64.82
66.38
35.07
52.19
97.90
96.79
74.34
132.69
169.06
134.40
20
100: 0
75:25
25:75
0: 100
6.02
5.14
2.88
-
68.32
58.58
39.63
-
0.54
1.90
2.54
200
100:0
75 : 25
75 : 75
0: 100
105:92
97.08
46.24
-
123.56
85.65
46.22
-
14.63
46.53
56.65
13.24
46.91
65. 50
31.78
60.32
55.45
105.92
111.71
92.77
56.65
123.56
98.89
93.13
65.50
31.78
60.32
56.65
229.48
242.38
246.22
177.60
10.10
16.79
7.04
13.74
9.10
8.70
15.19
9.10
11.37
p <0.001
p <0.001
p <0.001
p < 0.05
LSD (p < 0.05)
Significant
effects
Intercrop
ratio x N
Nitrogen x
Intercrop ratio
a
b
6.24
26.75
35.07
62.19
97.90
96.79
6.02
5.68
4.78
2.54
NS
p < 0.001
p < 0.001
Maize: ricebean ratio.
F-N = fertilizer-N; S-N = soil-N; T-N = total N (Rerkasem and Rerkasem, 1988)
168
Fujita et al.
Ofori and Stern (1986) in a maize/cowpea
intercrop system, observed that the LER decreased with N application increasing from
0 to 100 kgN ha -1 , independent of intercrop
maize.
Changes in crop yield and production efficiency caused by intercropping and combined N
application can be explained in terms of competition for N and light. Generally, the cereal component is taller and with its more extensive root
system has a competitive advantage over the
legume. The legume capable of fixing N2 is
thought to compete less with the cereal component for soil N (Trenbath, 1976). Chang and
Shibles (1985) and Ofori and Stern (1986) reported strong competition between maize and
cowpea for soil N when no N fertilizer was
added.
Strong interspecific competition between
maize and cowpea was demonstrated by Ezumah
et al. (1987). They suggested that inconsistent
yields in mixed cropping may be attributed to
varying growth habits and plant architectures
(Koli, 1975) of component crops or to fertilizer
application management (Haizel, 1974). This
was most evident between 49 and 63 days, when
both crops were at the reproductive stage and
required substantial N. Similarly, Ofori and
Stern (1986), in a maize/cowpea intercrop trial,
showed that without applied N, the intercrops
gave larger LAI and total dry matter yield than
the sole crops. The LAI was 2.7 in intercrop with
short maize, 3.5 in intercrop with tall maize, 1.8
in sole short maize, 2.1 in sole tall maize, and 2.3
in sole cowpea.
Large N applications cause excessive vegetative growth of the cereal, causing it to shade and
suppress the legume's yield. Ofori and Stern
(1986), in a maize/cowpea combination, observed a similar phenomenon when intercrop
seed yield was significantly reduced by
25 kg N ha -1 . In a sorghum/soybean intercrop
system, Ogata et al. (1986) found that at
300 kg N ha -1 intercrop soybean dry matter yield
was significantly reduced, and that total dry matter yield in the intercrop was equal to that of sole
crop sorghum. It appears that N application
intensifies competition for light between component crops and suppresses growth of the legume.
Phosphorus
Phosphorus fertilizer application at 40 kg P2O5/
ha increased intercrop grain yield as compared to
sole crop pearl millet (Kaushik and Gautam,
1987). The response to P was observed only on
soils initially low in P.
Water management
Availability of water is one of the most important factors determining productivity in
legume / cereal mixed cropping systems. Mixed
cropping is usually practiced by farmers in the
semi-tropical and tropical regions under rainfed
conditions. Shackel and Hall (1984) found no
differences in either total water absorbed or in
uptake patterns when cowpea and sorghum sole
crops were compared to their intercrops. Ofori
and Stern (1987) concluded that cereals and
legumes use water equally and that competition
for water may not be important in determining
intercrop efficiency, except under unfavorable
conditions.
Water use by intercrops has mostly been
studied in terms of water use efficiency (WUE).
An intercrop of two crop species such as legumes
and cereals may use water more efficiently than a
monocrop of either species (Willey, 1979a) exploring a larger total soil volume for water,
especially if the component crops have different
rooting patterns. Natarajan and Willey (1980)
reported higher WUE by sorghum/pigeon pea
when they were intercropped than when monocropped.
Advantages in intercrops can be explained by
the concepts of Snaydon and Harris (1979), in
which component crops compete for water at
partially different times (i.e. growth and development were not concurrent) or from partially different zones (i.e. extracting from different
depths). Where water is most limiting, intercrops
may offer a temporal and spatial advantage in
water use (Baker and Norman, 1975), although
Hulugalle and La1 (1986) could not confirm this.
They reported that WUE in a 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 than in the sole maize. Greater WUE was
found in two patterns of sole-cropped sorghum.
BNF in mixed legume-cereal systems
In the Philippines, Morris et al. (1980)
evaluated a sorghum/cowpea intercrop system
for efficiency of use of residual water and irregular rainfall during the dry season after rice cultivation. Average WUE by monocropped cowpea
was 11.3 kg glucose ha-1 mm -1, by monocropped
sorghum was 12.4, and by the intercrop was
16.5. Monocropped cowpea and component cowpea each used 50% more water than was lost in
fallow. However, the monocropped cowpea yield
was more stable. A higher WUE by a C, species
(sorghum) than a C, species (cowpea) (Angus et
al., 1983) may contribute to the apparent differences between monocropped and intercropped
systems. More data are needed on WUE under
irregular rainfall distributions; this may be a
factor in determining high dry matter yield.
Both N and P application increased total water
use efficiency of the intercrops. However, the
increase in water use with N application was
more apparent (Narindar Singh et al., 1985).
They speculated that N and P fertilization of
wheat and chickpea mixed cropping induced
deeper rooting, leading to better exploitation of
available water at deeper layers.
Sustainability issue
A need to increase food production to keep pace
with increasing population growth while ensuring
minimum environmental degradation has recently gained more attention. Modern crop production technology is based on major inputs, such as
quality seeds, fertilizers, pesticides, machinery,
and irrigation. Use of inputs frequently has been
abused, causing severe environmental deterioration. Agricultural production can only be sustained if productivity of land, labor, and water
resources upon which agriculture is based are not
allowed to deteriorate further. Farmers must
either be adequately paid or must adopt alternative production methods that will make farming
profitable.
Nutrient management
It is better to include legumes that do not respond to N application in mixed cropping system
169
because the cereal component must be fertilized
to achieve optimum yield and high economic
returns. The balance, therefore, must be the
correct level of N application where N fixation of
the legume is not sacrificed and sustainable yield
is maintained. This was shown in a ricebean/
maize cropping system in Thailand (Rerkasem
and Rerkasem, 1988), where the mixed-cropping
advantage in terms of dry matter, seed and N
yield was related to efficient mineral nitrogen use
by the maize and to dependence on fixation by
the ricebean (Fig. 2). However, this efficient
resource use by component crops was not confirmed in a cowpea/maize cropping system (van
Kessel and Roskoski, 1988). For efficient resource management in mixed cropping systems,
it is of utmost importance to find suitable
species, growth habits and plant architecture of
component crops.
Harvesting and transporting agricultural products away from their production sites causes a
depletion of soil minerals. The key to managing
plant nutrients in cropping systems to sustain
high yields is to balance the nutrient output by
recycling the crop residue nutrients efficiently
and to apply a minimum of fertilizer (Tanaka,
1986). Therefore, incorporation of residues into
soil and/or use of farmyard manure will provide
considerable mineral nutrients for subsequent
crops. Ofosu-Budu et al. (unpublished data) used
farmyard manure as a N source and obtained
high dry matter yield in a sorghum-soybean
mixed cropping system. The enhanced N availability to component sorghum was due to the high
dependence of the component soybean on BNF.
Underground water pollution through nutrient
leaching can be reduced when absorption of the
nutrients by component crops is maximized. If
component crops have different rooting systems
and uptake patterns, a more efficient use of
available nutrients may occur in intercropping
systems than in their monocropping counterparts. This has been demonstrated for K, Ca, Mg
(Dalal, 1974), and N (van Kessel and Roskoski,
1988).
Role of VAM fungi
Although they have received little attention, soil
170
Fujitu et al.
(a) FROM SOIL AND ATMOSPHERE
(b) FROM SOIL ONLY
Maize : ricebean ratio
Fig. 2. The effect of variation in the maize-ricebean ratio on the quantity of N derived from (a) soil and atmosphere and (b) soil
only in the tops of ricebean (
, v) and maize(
, z) in intercrop and monoculture (Rerkasem et al., 1988)
microbial species may affect yield and production efficiency in mixed legume/cereal systems.
Reports (Bethlenfalvay and Ferrera-Cerrato,
1990; van Kessel et al., 1985) show that VAM
fungi transport inorganic elements, such as N
and P.
When inoculated with VAM fungi, significant
two-way nutrient transport between soybean and
maize has been reported (Bethlenfalvay and Ferrera-Cerrato, 1990). N transfer from legume to
cereal, and P transfer from cereal to legume in a
source-sink relationship has been suggested.
When associated with nodulated soybean, maize
declined 16% in P content and increased 22% in
N content. If such fluxes between plants are
controlled by source-sink effects, high N concentrations in soybean could account for the N
fluxes to maize, and high nodule P requirement
for the reverse flux of P. Other growth-promoting roles, such as increased absorption capacity
and tolerance to mild water stress in legume and
cereal crops have been suggested, but little is
known about effectiveness in mixed cropping
under field conditions.
Dual inoculation of Rhizobium spp. and
Glomus fasciculatum enhances nodulation, yield
and nitrogen fixation in chickpea ( Cicer
arietinum L.). Synergistic effects of inoculation
of VAM fungi with rhizobia in soybean (Azimi et
al., 1980) and cowpea (Islam et al., 1980) have
been reported. Although the actual mechanism
is not clear, it has been speculated that a larger
root surface for nutrient absorption, particularly
P by the host plant (Tinker, 1982), stimulates
nitrogen fixation. Raju et al. (1989) also indicated VAM strain-specificity in nutrient absorption.
Use of tolerant species/cultivars
Intercropping has frequently been practiced
under adverse conditions, such as acidic and
water stress conditions. Water-use efficient
legumes and cereals should be selected for areas
with low and unpredictable rainfall patterns
(Yunusa, 1989). Most farmers do not have access
to irrigation facilities in areas where mixed cropping is practiced.
Considerable numbers of plant species of
economic importance are generally regarded as
tolerant of acidic conditions in the tropics. To
reduce excessive application of soil amendments,
acid-tolerant cultivars must be introduced. Little
is known about the effect of acidic stress conditions on mixed cropping. Mixed cropping of
pigeon pea ( Cajanus cajan L.), tolerant of low
BNF in mixed legume-cereal systems
soil P (Adu-Gyamfi et al., 1989; Sanchez and
Salinas, 1981) is popular in tropical areas such as
India. Cultivars that are tolerant of harsh environmental conditions, such as early maturing
cultivars of cowpea (Natarajan and Willey, 1980)
and pigeon pea for drier areas (Mohta and De,
1980), have been successfully developed.
Pest and weed management
One advantage of mixed cropping is a decrease
in losses caused by insects and diseases (Perrin,
1978). Crop mixtures may create a less favorable
environment for certain pests of one or both of
the component crops (Gerard, 1976). Insects can
have difficulty finding their host among the nonhost plants because of changes in color or texture
of the total background, masking of chemical
attractants or presence of chemical repellents
(Hasse and Litsinger, 1981). This advantage is
far from universal, but most reports suggest reductions in pest populations. Intercropping
groundnuts and maize reduced infestation by the
maize stalkborer, Eldana saccharina Wlk
(Anon., 1960). A less severe insect attack on
cowpea intercropped with sorghum was reported
in Northern Nigeria (Raheja, 1973).
Leguminous cover crops placed between rows
of taller, more desirable crop species such as
maize or sorghum (Francis, 1989) can help with
weed control. Legumes (indeterminate and
climbing types) are often used as cover crops
because they are less competitive with cereals,
have the capacity to produce N, and can entwine
and kills weeds.
Economics
Net income to farmers from crop production is a
function of yields, prices, and input costs. The
major difference between mixed cropping and
sole cropping is the introduction of several new
factors, such as relative prices of two or more
inputs and production efficiency, which result in
lower production costs. Earlier studies on mixed
cropping systems compared yield advantages
(Natarajan and Willey, 1980) rather than real
monetary gains. Recent publications (Mandal et
al., 1990; Ntare, 1989; Yunusa, 1989) have dem-
171
onstrated the economic advantages of mixed
cropping in monetary terms.
Ntare (1989) demonstrated variable income
for intercropping cowpeas in Niger because of
the high demand for the haulms (harvested stems
and tops), which serve as livestock fodder. Fodder demand far exceeds production, and a ready
market exists for cowpea haulms, which command nearly the same price as the cowpea grain.
A medium-maturing, spreading and indeterminate type (TN 88-63, TN 58-57) returned more
than twice the money to the farmer than an early
maturing, erect and intermediate (IT82E-60,
IT82D-716) type. This was mainly due to higher
cowpea grain and fodder yield. In Thailand,
ricebean, mostly intercropped with maize, is
grown over about 30,000 ha and according to
Rerkasem and Rerkasem (1988) 21,000 tons of
beans were exported with a market value of
US$7.9 million.
Future directions
Maximizing production efficiency in legume/
cereal mixed cropping system by enhancing BNF
by the legume component and legume N transfer
to the cereal component might happen through
improved cultural practices and genetic manipulation of the legume. Steps toward this goal are
discussed below:
1. Identify the physiological characteristics of
ideal component crops that will increase productivity.
a) Shading of the component legume most
often results in decreased BNF. However,
in some leguminous plants, BNF is either
slightly or not at all affected by shading, so
that BNF can continue in spite of reduced
light availability. This trait is important for
a component legume. Such traits need to
be incorporated when breeding plant types
suitable for mixed cropping.
b) Increased BNF activity through use of effective Rhizobium strains and optimum
management of mineral nutrition, in addition to tolerance of the legume component
to NO3, is important for increasing the
system’s N economy. Legumes with high
BNF activities and/or with high nitrate
tolerance (Caroll et al., 1985) should be
172
Fujita et al.
evaluated in terms of yield and production
efficiency in mixed cropping.
c) The controlling factors and mechanisms of
N release and transfer from legume to
associated cereal should be clarified in relation to legume BNF, environmental factors (i.e. light, water, and nutrients), and
other physiological factors (i.e. N pool,
root membrane stability, and enzymatic
effects). Legume and cereal compatibility
should also be considered, because allelochemical effects suppressing legume
BNF might exist.
2. Qualitative and quantitative data on light distribution in the production structure and dry
matter production is essential. The BNF activity and N release by shaded legumes should
be evaluated in terms of varying plant production densities and arrangements.
3. Cultural management and nutrient management are important for increasing yield and
production efficiency of intercrops.
a) The significance of crop residues and
farmyard manure in maintaining nutrient
sustainability in intercropping systems,
with particular emphasis on recycling N, P
and K under different crop combinations
and different climatic zones, should be
evaluated.
b) The role of VAM fungi in N and P transfer
and the overall dry matter production
should be assessed.
c) Mixed cropping is mostly conducted under
adverse conditions such as acidic soils
and unpredictable soil moisture regimes.
The major limiting factor in yield production for specific localities should be
examined.
4. Capital input for system sustainability cannot
be ignored. For example, some form of fertilizer or improved seed may be essential for
increased yield and productivity. Consequently, the economics of labor cost, agricultural
inputs, and gross returns to the farmer under
different legume/cereal combinations and
varying socio-economic conditions must be
addressed to identify a profitable combination
for specified areas.
References
Adu-Gyamfi J J, Fujita K and Ogata S 1989 Phosphorus
absorption and utilization efficiency of pigeon pea ( Cajanus cajan L. Millsp.) in relation to dry matter production
and dinitrogen fixation. Plant and Soil 119, 315-324.
Agboola A A and Fayemi A A 1971 Preliminary trials on
intercropping of maize with different tropical legumes in
Western Nigeria. J. Agric. Sci. 77, 219-255.
Agboola A A and Fayemi A A 1972 Fixation and excretion
of nitrogen by tropical legumes. Agron. J. 64, 409-412.
Angus J S, Hasegawa S, Hsiao T C, Libbon S P and Zandstra
H G 1983 The water balance of post-monsoonal dryland
crops. J. Agric. Sci. (Camb.) 101, 699-710.
Anon. 1960 Annual report of Entomologist for the year 1990.
Department of Agriculture, Tanganyika.
Azimi S, Giaininazzi P V and Giaininazzi S 1980 Influence of
increasing soil phosphorus levels on interactions between
vesicular-arbuscular mycorrhiza and rhizobium in soybeans. Can. J. Bot. 58, 2200-2205.
Baker E F I and Norman D W 1975 Cropping systems in
Northern Nigeria. In Proc. of Cropping Systems Workshop, IRRI, Los Baños, Philippines. pp 334-361.
Barker C M and Blarney F P C 1985 Nitrogen fertilizer
effects on yield and nitrogen uptake of sorghum and
soybean growth in sole and intercropping systems. Field
Crops Res. 12, 233-240.
Bethlenfalvay G J and Ferrera-Cerrato R 1990 Mycorrhizae
in intercrops: Significant N transfer from soybean to corn.
Agron. Abstr. 1990 Annual Meeting. American Society of
Agron. p. 53 San Antonio, TX.
Biscoe P V and Gallagher J W 1977 In Environmental Effects
of Crop Physiology. Eds. J J Landsberg and C V Cutting.
pp. 75-100. Academic Press, New York.
Broadbent F E, Nakashima T and Chang G Y 1982 Estimation of nitrogen fixation by isotope dilution in field and
greenhouse experiments. Agron. J. 74, 625-628.
Brophy L S, Heichel G H and Russelle M P 1987 Nitrogen
transfer from forage legumes to grass in a systematic
planting design. Crop Sci. 27, 753-758.
Brophy L S and Heichel G H 1989 Nitrogen release from
root of alfalfa and soybean grown in sand culture. Plant
and Soil 116, 77-84.
Caradus J R 1990 The structure and function of white clover
root systems. Adv. Agron. 43, 2-37.
Carol1 B J, McNeil D L and Gresshoff P M 1985 A supernodulation and nitrate-tolerant symbiotic (nts) soybean
mutant. Plant Physiol. 78, 34-40.
Catchpoole D W and Graeme Blair J 1990 Forage tree
legumes. II. Investigation of nitrogen transfer to associated
grass using a split-root technique. Aust. J. Agric. Res. 41,
531-538.
Chang J F and Shibles R 1985 Analysis of competition
between intercropped cowpea and maize. 1. Soil N and P
levels and their relationships with dry matter and seed
production. Field Crops Res. 12, 133-143.
Dalal R C 1977 Effects of intercropping maize with pigeon
BNF in mixed legume-cereal systems
peas on grain and nutrient uptake. Exp. Agric. 10, 219224.
d’Arcy Lameta A 1982 Etude des exsudats racinaires de soja
et de lentille. 1. Cinetique d’exsudation des composes
phenoliques, des amino acides et des sucres, au cours des
premiers jours de la vie des plantules. Plant and Soil 68,
399-403.
De Ru 1980 In Nuclear Techniques in the Development of
Management Practices for Multiple Cropping Systems. pp
73-84, IAEA-TECDOC, 235. Vienna.
De Wit C T 1960 On competition. Verslagen Landbouwkundige Onderzoekingen No. 66 (8), 1-82.
Droushiotis D N 1989 Mixture of annual legumes and smallgrained cereals for forage production under low rainfall. J.
Agric. Sci. (Camb.) 113, 249-253.
Eaglesham A R J, Ayanaba A, Ranga Rao V and Eskew D
L 1981 Improving the nitrogen nutrition of maize by
intercropping with cowpea. Soil Biol. Biochem. 13, 169171.
Ezuma HC, Nguyen Ky Nam and Walker P 1987 Maizecowpea intercropping as affected by nitrogen fertilization.
Agron. J. 79, 275-280.
Faris M A, Burity H A, Dos Reis O V and Mafra R C 1983
Intercropping of sorghum or maize with cowpeas or common beans under two fertility regimes in Northeastern
Brazil. Exp. Agric. 19, 251-261.
Fawusi M O A, Wanki S B C and Nangju D 1982 Plant
density effects on growth, yield. leaf area index and light
transmission on intercropped maize and Vigna unguiculata
(L.) Walp in Nigeria. J. Agric. Sci. (Camb.) 99, 19-23.
Fisher N M 1977 Studies in mixed cropping. II. Population
pressures in maize-bean mixtures. Exp. Agric. 13, 185191.
Francis C A, Flor C A and Temple S R 1976 Adapting varieties for intercropped systems in the tropics. In Multiple
Cropping. ASA Spec. Publ. Agron. Eds. R I Papendick,
P A Sanchez and G B Triplett. pp 235-253. Madison, WJ.
Francis C A 1986 In Multiple Cropping Systems. MacMillan,
New York. 383 p.
Francis C A 1989 Biology efficiencies. In Multiple Cropping
Systems. Adv. Agron. 42, 1-42.
Fujita K, Ogata S, Matsumoto K, Masuda T, Ofosu-Budu K
G and Kuwata K 1990a Nitrogen transfer and dry matter
production in soybean and sorghum mixed cropping systems at different population densities. Soil Sci. Plant Nutr.
36, 233-241.
Fujita K, Ogata S, Ofosu-Budu K G and Matsumoto K 1990b
Significance of mixed cropping of legumes and grass
species on dry matter production and nitrogen balance.
Symposium III-4. pp 164-169. 14th Int. Congress of Soil
Sci., Kyoto, Japan 12-18 August 1990.
Gerard B M 1976 The effect of plant density on maize and
cowpea pests. Paper presented at the symposium on intercropping in semi-arid areas. Morogoro, Tanzania, May
1976.
Graham P H and Rosas J C 1978 Plant and nodule development and nitrogen fixation in climbing cultivars of
173
Phaseolus vulgaris L. grown in monoculture or associated
with Zea mays L. J. Agric. Sci. (Camb.) 90, 311-317.
Haizel K A 1974 The significance of mixed cropping. I. Maize
interplanted with cowpea. Ghana J. Agric. Sci. 7, 169-178.
Hale M G, Moore L D and Griffin G J 1978 Root exudates
and exudation. In Interactions Between Non-pathogenic
Soil Micro-organisms and Plants. Eds. Y R Dommergues
and S V Krupa. pp 163-203. Elsevier, Amsterdam, The
Netherlands.
Hall R L 1978 The analysis and significance of competitive
and non-competitive interference between species. In Plant
Relations in Pastures. Ed. Wilson. pp 163-174. CSIRO,
Melbourne.
Hasse V and Litsinger J A 1981 IRRI Saturday Seminar,
Entomology Dept., IRRI, Los Baños, Philippines.
Hawes M C and Hao-Jin Lin 1990 Correlation of pectolytic
enzyme activity with the programmed release of cells from
root caps of pea ( Pisum sativum ). Plant Physiol. 94, 18551859.
Herbert S J, Putnam D H, Poos-Floyd M I. Vargas A and
Creighton J F 1984 Forage yield of intercropped corn and
soybean in various planting patterns. Agron. J. 76, 507515.
Hulugalle N R and Lal R 1986 Soil water balance in intercropped maize and cowpea grown in a typical hydromorphic soil in Western Nigeria. Agron. J. 77, 86-90.
Huxley P A and Mainger Z 1978 Use of systematic spacing
design as an aid to the study of intercropping: Some
general considerations. Exp. Agric. 14, 49-56.
Islam R. Ayanaba A and Sanders F E 1980 Response of
cowpea ( Vigna unguiculata ) to inoculation with VA-mycorrhizal fungi and to rock phosphate fertilization in some
unsterilized Nigerian soils. Plant and Soil 54, 107-117.
Jameson J E (Ed.) 1970 Agriculture in Uganda, Uganda
Government. Ministry of Agriculture and Industry, Oxford
University Press, London.
Kaushik S K and Gautam R C 1987 Effect of nitrogen and
phosphorus on the production potential of pearl-millet/
cowpea or green gram intercropping systems under rainfed
conditions. J. Agric. Sci. 108, 361-364.
Koli E S 1975 Pure cropping and mixed cropping of maize
and groundnuts in Ghana. Ghana J. Agric. Sci. 8, 23-30.
Lakhani D A 1976 A physiological study of mixtures of
sunflower and fodder radish, Ph.D. Thesis, Univ. of
Reading.
Mandal B J and Mahapatra S K 1990 Barley, lentil and flax
under different intercropping systems. Agron. J. 82. 10661068.
Mandal B J, Dhara M C, Mandal B B, Das S K and Nandy R
1990 Rice. mungbean, soybean, peanut, ricebean and
black gram yields under different intercropping systems.
Agron. J. 82, 1063-1066.
Mohta N K and De R 1980 Intercropping maize and sorghum
with soybeans J. Agric. Sci. (Camb.) 95, 117-122.
Morris R A, Villegas A N, Polthanee A and Centeno H S
1990 Water use by monocropped and intercropped cowpea
and sorghum grown after rice. Agron. J. 82, 664-669.
174
Fujita et al.
Nair K P P, Patel U K, Singh R P and Kaushik M K 1979
Evaluation of legume intercropping in conservation of
fertilizer nitrogen in maize culture. J. Agric. Sci. (Camb.)
93, 189-194.
Nambiar P C T. Rao M R, Reddy M S, Floyd C, Dart P J
and Willey R W 1983 Effect of intercropping on nodulation
and N2 fixation by groundnut. Exp. Agric. 19, 79-86.
Narindar Singh, Nathu S, Ranjodh S, Tejinder S and Mukhtar S 1985 Yield and water use efficiency of rainfed wheat
+ chickpea mixture as affected by N and P applications in
clay-loam, sandy-loam and loamy-sand soils. Indian J.
Agric. Sci. 55, 13-17.
Natarajan M and Willey R W 1980 Sorghum-pigeon pea
intercropping and the effects of plant population density. J.
Agric. Sci. 95, 59-65.
Nielson G N and Mackenzie A F 1977 Soluble and sediment
nitrogen losses as related to land use and type of soil in
Eastern Canada. J. Environ. Qual. 6, 318-321.
Normal R J, Gilmour J T and Wells B R 1990 Mineralization
of nitrogen from nitrogen-15 labeled crop residues and
utilization by rice. Soil Sci. Soc. Am. J. 54, 1351-1356.
Ntare B R 1989 West Africa's dry savanna: Intercropping
cowpeas with millet. IITA Research Briefs, Dec. 1989.
Nye P H and Tinker P B 1977 Solute Movement in the
Soil-Root System. Blackwell, Oxford.
Ofori F and Stern W R 1986 Maize/cowpea intercrop system:
Effect of nitrogen fertilizer on productivity and efficiency.
Field Crops Res. 14. 247-261.
Ofori F, Pate J S and Stern W R 1987 Evaluation of N2
fixation and nitrogen economy of a maize/cowpea intercrop system using 15N dilution methods. Plant and Soil 102,
149-160.
Ofori F and Stern W R 1987 Cereal-legume intercropping
systems. Adv. Agron. 41, 41-90.
Ofosu-Budu K G, Fujita K and Ogata S 1990 Excretion of
ureide and other nitrogenous compounds by the root system of soybean at different growth stages. Plant and Soil
128, 135-142.
Ogata S, Fujita K, Matsumoto K and Saneoka H 1986
Studies on mixed cropping of grasses and legumes. I. Dry
matter production and behavior of fixed N in the system of
mixed cropping of sorghum, soybean and siratro. J. Jpn.
Grassl. Sci. 32, 36-43. (In Japanese with English
summary.)
Okigbo B N and Greenland D J 1976 Intercropping systems
in tropical Africa. In Multiple Cropping. Eds. Papendick et
al. pp 63-101. Spec. Publ. No. 27. Am. Soc. Agron.
Madison, WI.
Osiru D S and Willey R W 1972 Studies on mixtures of dwarf
sorghum and beans ( Phaseolus vulgaris ) with particular
reference to plant population. J. Agric. Sci. (Camb.) 79,
531-540.
Papastylianou I 1988 The 15N methodology in estimating N2
fixation by vetch and pea grown in pure stand or in
mixtures with oats. Plant and Soil 107, 183-188.
Papastylianou I 1990 Response of pure stands and mixtures
of cereals and legumes to nitrogen fertilization and residual
effect on subsequent barley. J. Agric. Sci. (Camb.) 115,
15-22.
Patil B P and Mahendra Pal 1988 Associative effect of
intercropped legumes on pearl millet ( Pennisetum
glaucum ) and their residual effect on nitrogen nutrition of
succeeding bread wheat ( Triticum aestivum ) Indian J.
Agric. Sci. 58. 429-432.
Patra A P and Chatterjee 1986 Intercropping of soybean with
rice, maize and pigeon pea in the plains of West Bengal.
Indian J. Agric. Sci. 56, 413-417.
Patra D D, Sachdev M S and Subbiah B V 1986 15N studies
on the transfer of legume-fixed nitrogen to associated
cereals in intercropping systems. Biol. Fert. Soils 2, 165171.
Patra D D, Sachdev M S and Subbiah B V 1989 Residual
value of 15N-labeled fertilizer applied to a maize-cowpea
intercropping system. Biol. Fert. Soils 8, 183-188.
Paull R E and Jones R L 1976 Studies on the secretion of
maize root cap slime. V. The cell wall as a barrier to
secretion. 2. Pflanzenphysiol. 79, 154-164.
Pearson R and Parkinson D 1961 The sites of excretion of
ninhydrin-positive substances by broad bean seedlings.
Plant and Soil 57, 69-83.
Peoples M B and Herridge D F 1990 Nitrogen fixation by
legumes in tropical and subtropical agriculture. Adv.
Agron. 44, 156-216.
Perrin R M 1978 Some effects of mixed cropping on the
population dynamics of insect pests. Ent. Exp. Applic. 24.
386-393.
Phetchawee S, Vibulsukh N, Theppoolpon M and
Masarngsan W 1986 Long term effect of mulching with
fertilization under cropping corn-legumes on crop yield and
improvement of soil chemical-physical properties. Proc.
Int. Sem. on Yield Maximization of Feed Grains Through
Soil and Fertilizer Management.
Power J F and Doran J W 1988 Role of crop residue
management in nitrogen cycling and use. In Cropping
Strategies for Efficient Use of Water and Nitrogen. Ed.
Hargrove et al., pp 101-113. ASA Spec. Publ. 51, ASA
CSA and SSSA, Madison, WI.
Rabotnov T A 1977 In Application of Vegetation Science to
Grassland Husbandry. Ed. W Kroese. pp 459-497. Dr W.
Junk, The Hague, The Netherlands.
Rachie K O and Roberts L M 1974 Grain legumes of lowland
tropics. Adv. Agron. 26, 2-118.
Raju P S, Chalk R B, Ellis J R and Maranville J W 1989
Effects of species of VA-mycorrhizal fungi on growth and
mineral uptake of sorghum at different temperatures. Plant
and Soil 121, 165-170.
Rao L J and Mittra B N 1989 It pays to intercrop groundnut
and arhar during winter in West Bengal. Indian Farming
Sept. 1989. p 25.
Rego T J 1981 Nitrogen response studies of intercropped
sorghum with pigeon pea. Proc. Workshop Intercrop. (ICRISAT). 10-13 January 1979. Hyderabad, pp 210-216.
Remison S U 1978 Neighbor effects between maize and
cowpea at various levels of N and P. Exp. Agric. 14.
205-212.
Rennie R J, Dubetz S, Bole J B and Muendel H-H 1982
Dinitrogen fixation measured by 15N isotope dilution in
two Canadian soybean cultivars. Agron. J. 74, 725-730.
Rerkasem K and Rerkasem B 1988 Yields and nitrogen
nutrition of intercropped maize and ricebean ( Vigna um-
BNF in mixed legume-cereal systems
bellara [Thumb] Ohwi and Ohashi). Plant and Soil 108.
151-162.
Rerkasem B, Rerkasem K, Peoples M B, Herridge D F and
Bergersen F J 1988 Measurement of N2 fixation in maize
( Zea mays L.) - ricebean ( Vigna umbellam [Thumb] Ohwi
and Ohashi) intercrops. Plant and Soil 108, 125-135.
Rhodes I and Stern W R 1978 Competition for light. In Plant
Relations in Pastures. Ed. J R Wilson. pp 175-189.
CSIRO. Melbourne, Vic.
Ritcher M. Wilms W and Scheffer F 1968 Determination of
root exudates in a sterile continuous How culture. II.
Short-term and long-term variations of exudation intensity.
Plant Physiol. 43, 1747-1754.
Roughier M 1981 Secretory activity of the root cap. In
Encyclopedia of Plant Physiologist. New Series. Plant Carbohydrates II. Vol. 13B. Eds. W Tanner and F Loewus. pp
542-574. Springer-Verlag, Berlin.
Sanchez P A. Salinas J G 1981 Low-input technology for
managing oxisols and ultisols in tropical America. Adv.
Agron. 34, 280-398.
Scott T W, Mt Pleasant J, Burt R F and Otis D J 1987
Contributions of groundcover, dry matter and nitrogen
from intercrops and cover crops in a corn polyculture
system. Agron. J. 79, 792-798.
Scroft M N and Snyder W C 1962 Exudation patterns from
bean seeds and hypocotyls and their effects on Fusarium
solani f phaseoli. Phytopathology 52. 751.
Searle P G E, Comudom Y, Sheddon D C and Nance R A
1981 Effect of maize + legume intercropping systems and
fertilizer nitrogen on crop yields and residual nitrogen.
Field Crops Res. 4. 133-145.
Shackel K A and Hall A E 1984 Effect of intercropping on
the water relations of sorghum and cowpea. Field Crops
Res. 8. 381-387.
Singh S P 1983 Summer legume intercrop effects on yield and
nitrogen economy of wheat in the succeeding season. J.
Agric. Sci. (Camb.) 101, 401-405.
Singh N B, Singh P P and Nair K P P 1986 Effect of legume
intercropping on enrichment of soil nitrogen, bacterial
activity and productivity of associated maize crops. Expl.
Agric. 22. 339-344.
Snaydon R W and Harris R M 1979 Nutrient and rooting
pattern interaction below ground - the use of nutrients and
water. In Proc. Int. Workshop on Intercropping. pp 188201. Hyderabad, India Jan. 1979.
Ta T C, MacDowall F D H and Faris M A 1986 Excretion of
nitrogen assimilated from N2 by nodulated roots of alfalfa
(Medicago sativa). Can. J. Bot. 64, 2063-2067.
Ta T C and Faris M A 1987 Species variations in the fixation
and transfer of nitrogen from legumes to associated grasses. Plant and Soil 98, 265-274.
Ta T C. Faris M A and MacDowall F D H 1989 Evaluation
of 15N methods to measure nitrogen transfer from alfalfa to
companion timothy. Plant and Soil 114. 243-247.
Tanaka A 1986 Potential yield of upland cropping systems
and approaches to its expression. Proc. of Inter. Conf. on
the Management and Fertilization of Upland Soils in the
Tropics and Sub-Tropics. Nanjing, China. p 11-21.
175
Tanaka A and Fujita K 1979 Growth. photosynthesis and
yield components in relation to grain yield of the field
bean. J. Faculty of Agriculture. Hokkaido Univ. 59, 145238.
Tinker P B 1982 Mycorrhizas. In Trans. 12th Inter. Congr.
Soil Sci. pp 155-162. New Delhi, India.
Trang K M and Giddens J 1980 Shading and temperature as
environmental factors affecting growth. nodulation and
symbiotic N2 fixation by soybeans. Agron. J. 72. 305-308.
Trenbath B P 1976 Plant interaction in mixed communities.
In Multiple Cropping. Proc. of a Symposium. Eds. Papendick et al. pp 76-102.ASA. CSSA and SSSA. Madison. WI.
Tsay J 1985 Physiological studies of cassava-soybean intercropping. Ph.D. thesis. Univ. of Queensland, Australia.
Vancura V and Stanek M 1975 Root exudates of plants. V.
Kinetics of exudates from bean roots as related to the
presence of reserve compounds in cotyledons. Plant and
Soil 43. 547-559.
Van Kessel C. Singleton P W and Hoben H J 1985 Enhanced
N-transfer from a soybean to maize by vesicular-arbuscular
mycorrhizal fungi. Plant Physiol. 79. 562-563.
Van Kessel C and Roskoski J P 1988 Row spacing effects on
N, fixation. N-yield and soil N uptake of intercropped
cowpea and maize. Plant and Soil 111. 17-23.
Virtanen A 1. Von Hausen S and Laine T 1937 Investigation
on the root nodule bacteria of leguminous plants. XX.
Excretion of nitrogen in associated cultures of legumes and
non-legumes. J. Agric. Sci. 27. 610.
Wacquant J P. Ouknider M and Jacquard P 1989 Evidence of
periodic excretion by roots of grass-legume associations.
Plant and Soil 116. 57-68.
Waghamarc A B. Krishnan T K and Singh S P 1982 Crop
compatibility and spatial arrangement in sorghum-based
intercropping systems. J. Agric. Sci. (Camb.) 99. 621-629.
Wahua T A T and Miller D A 1978a Effects of intercropping
on soybean N, fixation and plant composition on associated sorghum and soybeans. Agron. J. 70, 292-295.
Wahua T A T and Miller D A 1978 Effects of shading on the
N2 fixation. yield and plant composition of field-grown
soybeans. Agron. J. 70. 387-392.
Whitney A S and Kaneshiro Y 1967 Pathway of nitrogen
transfer in some tropical legume-grass associations. Agron.
J. 59, 585-588.
Willcy R W 1979a Intercropping-its importance and research needs. Part 1: Competition and yield advantages.
Field Crops Abstr. 32. 1-10.
Willey R W 1979b Intercropping- its importance and research needs. Part 2: Agronomy and research approaches.
Field Crops Abstr. 32. 73-85.
Willey K W and Osiru D S O 1972 Studies on mixtures of
maize and beans ( Phaseolus vulgaris ) with particular reference to plant population. J. Agric. Sci. (Camb.) 79. 517529.
Yunusa I A M 1989 Effects of planting density and plant
arrangement pattern on growth and yields of maize ( Zea
mays L.) and soybean ( Glycine max L. Merr.) grown in
mixtures. J. Agric. Sci. (Camb.) 112, 1-8.
Plant and Soil 141: 177-196, 1992.
© 1992 Kluwer Academic Publishers. Printed in the Netherlands.
PLSO SA10
Biological nitrogen fixation in trees in agro-ecosystems
S.K.A. DANSO , G.D. BOWEN and N. SANGINGA
1
Joint F A O / I A E A Division, P.O. Box 100, Wagramerstrasse 5 A-1400 Vienna, Austria. Present
addresses: 2 Division of Soils, CSIRO, Glen Osmond 5064, Australia and 31nternational Institute of
Tropical Agriculture, P M B 5320, lbadan, Nigeria
1
2
3
Key words: A-value, acetylene reduction assay, agroforestry, isotope dilution, 15N, nitrogen fixation,
nodules, trees, ureide technique
Table of contents
Abstract
Introduction
Measurement of BNF in trees
The total nitrogen difference (TND) method
The acetylene reduction (ARA) assay
The 15N isotope methodology
The natural 15N abundance or 6 15N method
Nitrogen-15 fertilizer enrichment methods
Selection of reference plant
Rate, formulation, time and frequency of 15N fertilizer addition
Sampling of plants
Effect of remobilized N
Effect of litter turnover
Ureides as a measure of BNF
Nitrogen fixation potential of trees
The occurrence of nodulation
Genotypic variation in BNF
Plant and Rhizobium factors
The age factor
The Rhizobium component
Specificity in nodulation
Effectiveness
Management practices and environmental effects on BNF
Effects of management
Environmental factors
Season
Soil moisture
Temperature
Soil nitrogen
Soil toxicity
178
Danso et al.
Nutrition
Mycorrhizas
Conclusion and future research needs
References
Abstract
The integration of trees, especially nitrogen fixing trees (NFTs), into agroforestry and silvo-pastoral
systems can make a major contribution to sustainable agriculture by restoring and maintaining soil
fertility, and in combating erosion and desertification as well as providing fuelwood. The particular
advantage of NFTs is their biological nitrogen fixation (BNF), their ability to establish in nitrogendeficient soils and the benefits of the nitrogen fixed (and extra organic matter) to succeeding or
associated crops.
The importance of NFTs leads to the question of how we can maximise or optimize their effects and
how we can manage BNF and the transfer of nitrogen to associated or succeeding plantings. To be able
to achieve these goals, suitable methods of measuring BNF in trees are necessary. The total nitrogen
difference (TND) method is simple, but is better suited for low than high soil N conditions. The
acetylene reduction assay (ARA), although sensitive and simple, has many technical limitations
especially for NFTs, and the estimates of BNF have generally been very low, compared to other
methods. For NFTs, the 15N techniques are still under development, but have already given some
promising results (e.g., has been used to measure large genetic variability in BNF within different
NFTs).
Various factors affect BNF in trees. They include the age of trees, the microbial component, soil
moisture, temperature, salinity, pH, soil N level and plant nutrient deficiencies. Some of the factors,
e.g. temperature, affect the symbiosis more than plant growth, and differences in the effects of these
factors on BNF in different NFT genotypes have been reported. These factors and research needs for
improving BNF in trees are discussed.
Introduction
The use of trees, especially nitrogen fixing trees
(NFTs), in agroforestry and silvopastoral systems, is receiving considerable attention for several reasons. Population increases are outstripping food production in many parts of the world
and more forest land is being brought into agricultural production - several million hectares of
forest are being lost every year. Most tropical
soils are extremely fragile and give very poor
crop yields after only a few years of cultivation
without expensive fertilizer inputs. The decline
in soil fertility is followed in many cases by soil
erosion and further decline in soil nutrients. The
use of nitrogen-fixing trees in agroforestry systems in the humid/sub-humid tropics is therefore
attractive as a source of N and organic matter
needed in the rehabilitation of damaged soils or
sustainability in undamaged soils, as protection
against erosion, and for fuelwood - the major
source of energy for over half the world and a
commodity in extremely short supply in many
countries. Sanginga et al. (1986) found that the
additions of prunings of 6-month-old inoculated
Leucaena leucocephala increased subsequent
maize yield from 2.2 t ha-1 to 4.6 t ha-1. La1
(1989) reported that L. leucocephala planted at
2-m intervals reduced erosion from interrow
maize plantings from 4.3 t ha-1 (plow-till) to
0.1 t ha-1, i.e. the same as no-till. In a different
context, Ladha et al. (1989) found that green
manure from the nitrogen-fixing Sesbania rostrata (a stem- and root-nodulating legume) was
equivalent to 60 to 120 kg N ha-1 as urea in a
rice ecosystem.
Similar problems as in the humid/sub-humid
areas also occur in the vast semi-arid/arid areas
Nitrogen fixation in trees
of the world. Here, wind erosion and desertification also become major problems and NFTspreferably deep rooted ones with access to
groundwater - will probably have an important
role. Analysis of soils from under Acacia albida
( syn. Faidherbia albida ) in Senegal (Charreau
and Vidal, 1965) indicated a remarkable fertility
gradient from bare soil to soil under the foliage;
protein yields of millet near trees were increased
3- to 4-fold. Leguminous trees such as Acacia
cyanophylla are commencing to have an important agroforestry role in areas of North Africa with 250-400mm rainfall. Significant accumulations of organic C, N, Ca, Mg, P and K
have been measured in the surface soil beneath
Prosopis canopies in natural ecosystems (Virginia, 1986), due partly to transport of these
elements from lower depths and to BNF.
Casuarina equisetifolia (symbiotic with the actinomycete Frankia ) has found extensive use to
consolidate sandy coastal soils.
While there is recognition of the importance of
NFTs in maintaining fertility, in soil conservation
and in providing wood, there have been few
studies on the nitrogen-fixing symbioses, their
potential and constraints, and management
strategies to enhance their effects. In this paper
we address the problem of measuring biological
nitrogen fixation in trees, indicate some of the
variables affecting fixation and some of the opportunities for management. We have focussed
particularly on leguminous trees, with some reference to the actinorrhizal trees, especially
casuarinas. While casuarinas have roles in soil
conservation and reclamation of saline sites, they
have not yet found as wide an acceptance in
agroforestry as the leguminous trees.
Measurement of BNF in trees
The problems associated with BNF measurements are much more complex with NTFs than
with annuals due largely to the size of trees and
their perennial nature - hence, only few studies
have been conducted on BNF in trees. The
practical difficulties in harvesting whole trees or
of realistic non-destructive sampling, and in obtaining suitable controls valid over seasons or
years, are not easy to overcome, and for this
179
reason many researchers have estimated BNF in
young trees over short periods. However, questions such as - For how long do trees continue to
fix N2? What are the effects of root and litter
turnover on N2 fixation and soil fertility? cannot be resolved through studies on N, fixation
in young trees. Another problem is the number
of trees to be used per study consistent with the
larger area required by tree experiments, and
with available resources. For a given isoline or
provenance the variation in a given trait can be
large. Sanginga et al. (1989a) observed very high
variability in nodulation by pot-grown Allocasuarina spp.; although all plants were inoculated, some plants in the same pot nodulated,
but others did not. For practical reasons,
economic reasons and also to reduce the degree
of soil variability in studies with trees, single or
few tree-plots have however had to be used.
The commonly used methods for measuring
BNF are: total nitrogen difference (TND),
acetylene reduction assay (ARA), the natural
15N abundance method, and procedures with 15N
fertilizer addition to soil (isotope dilution (ID)
and A-value (Av), all of which have been extensively reviewed (Danso, 1985; Knowles, 1980;
Rennie and Rennie, 1983; Shearer and Kohl,
1986).
The total nitrogen difference ( T N D ) method
The TND method is simple, and has provided
several estimates of BNF in trees, with examples
in Table 1. The TND method is based on estimating how much of the total (N.,) in the fixing
plant was accumulated from soil (Ns); the remainder is then attributed to fixed N. The total
N in a control non-N2-fixing plant is assumed to
reflect the N, in the N2-fixer.
Unlike some grain legumes, non-nodulated
NFT isolines are not yet available, and uninoculated NFT controls have been used in soils without indigenous rhizobia (Ndoye and Dreyfus,
1988; Sougoufara et al., 1990), while completely
different species have also served as controls
(Pareek et al., 1990). Serious errors in the estimates of N, will therefore affect the accuracy of
the BNF estimates (Danso, 1985; Rennie and
Rennie, 1983). The chances for errors are small
180
Danso et al.
Table 1. Estimates of BNF in nitrogen-fixing trees by the total nitrogen difference (TND) method
Species
Leucaena leucocephala
L. leucocephala
L. leucocephala
Casuarina equisetifolia
C. equisetifolia
C. equisetifolia
Sesbania rostrata
S. sesban
Gliricidia sepium
Albizia lebbeck
Estimated BNF
Reference
%
kgN ha -1 yr -1
52-64
65
45 -63
49
55 -76
39-55
35 -45
11-18
72
60
448-548
304
in low-N soils because of the small contribution
from NS relative to fixed N2 (which would be the
main source of NT). With increasing soil N, NS
may contribute significantly to plant growth and
errors in estimating N, may then have a significant impact on the estimate of BNF. The simple
TND method is therefore most recommended
for measuring BNF in sandy or low-N soils and
then shows little difference from other methods
(e.g. Gauthier et al., 1985; Ndoye and Dreyfus,
1988). Even so, the gradual build-up of N under
NFTs should be considered for long-term
studies.
For trees more than a year old, distinguishing
the N reabsorbed from the decomposition of
fallen leaves and senesced roots and nodules
from native soil N and fixed N2 is an additional
problem. These factors might account for several
cases of poor agreement between TND estimates
of BNF and other techniques (Pareek et al.,
1990; Sanginga et al., 1990a, d; Stewart and
Pearson, 1967).
The acetylene-reduction assay (ARA)
The ARA technique is simple, inexpensive, and
the most sensitive method for detecting BNF,
and accounts for most of the estimates of BNF in
trees. The amount of ethylene produced by incubating excised nodules, decapitated roots or
whole plants in an atmosphere containing
acetylene is converted into total N2 fixed by
multiplying with a conversion factor which, on
theoretical grounds, was fixed at 3 (Hardy et al.,
1968).
Several limitations of the ARA technique have
Sanginga et al. (1985)
Liya et al. ( 1990)
Sanginga et al. (1990d)
Duhoux and Dommergues (1985)
Sougoufara et al. (1990)
Gauthier et al. (1985)
Ndoye and Dreyfus (1988)
Ndoye and Dreyfus (1988)
Liya et al. (1990)
Liya et al. (1990)
53
43-60
505-581
43-102
108
94
been reported. The two major ones are (i) it is
an instantaneous assay and may not truly reflect
BNF over long durations (Fried et al., 1983) and
(ii) the conversion ratio of 3 does not apply in all
cases (Rennie et al., 1978; Hansen et al., 1987).
The accurate estimation of BNF in woody perennials would therefore require a burdensome
number of measurements and calibrations. Even
then, (iii) large errors are likely to arise because
it is very difficult to recover all nodules, under
most circumstances. Nodules located at great
depths, up to 10 m, for example (Jenkins et al.,
1988), can be difficult to recover.
Some reported ARA estimates of BNF in
NFTs are presented in Table 2. The values are
very low, including even the 110 kg N ha -1 yr -1
reported for Leucaena leucocephala when compared to values obtained using other methods.
(e.g. Sanginga et al., 1985; Zaharah et al., 1986).
Where detached nodules were used. the values
are most likely to be underestimates, given the
drastic declines in ARA that commonly result
from detaching nodules from plants (Langkamp
et al., 1979). We know of no in situ ARA
determinations on field-grown NFTs.
The
15
N isotope methodology
The major strengths of BNF measurements using
either added 15 N fertilizers or the natural 15 N
abundance in soil are: they can estimate the
separate contributions of N from soil, fertilizer
and BNF, and measure the integrated amounts
and proportions of N2 fixed over desired periods.
They are therefore useful for assessing BNF
differences among genotypes belonging to the
Nitrogen fixation in trees
181
Table 2. Some estimates of BNF in nitrogen-fixing trees by the acetylene reduction assay (ARA)
Species
Estimated BNF
%
Acacia pulchella
A. pulchella
A. alata
A. alata
A. pellita
A. pennatula
A. extensa
Brassiaeaaquilifolim
Gliricidiasepium
Inga jinicuil
Leucaenaleucocephala
Purshia tridenta
13.0
3.7
same species or different species (Sanginga et
al., 1990a,b) and effects such as soil N on BNF
(Sanginga et al., 1987). In pot studies Sanginga
et al. (1990b) found that 5% Ndfa varied from 37
to 72% (mean, 65%) with L. Ieucocephala isolines and 6 to 36% (mean, 20%) with A. albida
provenances. Ndoye and Dreyfus (1988) found
that stem-nodulated Sesbania rostrata fixed much
more N2 in 60 days (83-109 kg N ha-1) than the
root-nodulated S. sesban (7-18 kg N ha-1). With
the 15N techniques, it should be possible to
identify NFT genotypes in which BNF is less
affected by soil N.
The 15N methods are however better developed for herbaceous legume systems than for
woody perennials. Although most measurements
of BNF in trees so far are on greenhouse-grown
plants there appears to be much promise in the
use of the 15N techniques to measure BNF in
woody perennials (e.g. large genetic variability in
BNF within isolines and provenances of NFTs
were reported by Sanginga et al., 1990a, b). The
15N soil appoaches are the 15N natural abundance, the 15N isotope dilution (ID) (Fried and
Middelboe, 1977; McAuliffe et al., 1958) and the
A-value (Av) approaches (Fried and Broeshart,
1975).
The natural 15N abundance or d 15N method
Many soils have N of slightly higher 15N abundance than that of atmospheric N2, and thus
relative to atmospheric N2, most soils are slightly
enriched in 15N. The extent to which the 15N (or
Reference
kg N ha-1 yr-1
0.49
1.6
12.0
34.0
0.1
0.19
13.0
35.0
110.0
0.06
Hansen and Pate (1987b)
Hansen et al. (1987)
Hansen and Pate (1987b)
Hansen et al. (1987)
Langkamp et al. (1979)
Roskoski et al. (1982)
Hansen et al. (1987)
Hansen et al. (1987)
Roskoski et al. (1982)
Roskoski (1981)
Hogberg and Kvarnström (1982)
Dalton and Zobel (1977)
d 15N) accumulated from soil is diluted by N,
fixed in the fixing plant is then used to estimate
BNF (Delwiche and Steyn, 1970; Mariotti, 1983;
Shearer et al., 1974).
The advantages of the d 15N methodology include: (i) A fairly stable d 15N with time, resulting in less errors in the determination of BNF;
(ii) Does not require addition of costly 15N
fertilizers. For trees, this could involve substantial savings; (iii) the d 15N method does not
involve any disturbance of the soil as occurs with
incorporation of 15N fertilizer and therefore is
more appropriate to natural ecosystems and established plantations.
Shearer et al. (1983) observed significantly
lower d 15N in Prosopis, the dominant tree
species in a Sonoran desert ecosystem, than in
either soil N or presumed non-N2-fixing plants
deriving their N entirely from soil. Shearer and
Kohl (1986) estimated that P. glandulosa derived
between 43 and 65% of its N from BNF, and
also measured differences in BNF in P. glundulosa at different sites. With Alnus, depending on
age, the percentage of N derived from fixation
(% Ndfa) in A. incana was shown by the d 15N
method to be between 85 and 100%, while %
Ndfa differed between 40 and 80% in seven A.
glutinosa
genotypes
of
different
origin
(Domenach et al., 1989). The method was able
to detect legumes that did not fix N2 (Shearer et
al., 1983), and shows much promise for exploratory studies in natural ecosystems for potential
NFTs.
A major drawback with the d 15N method is
182
Danso et al.
that unlike the 15 N enrichment methods, small
isotope fractionations or variability (site or
among plant parts) can cause significant errors
(Shearer and Kohl, 1986). Also, the 15 N natural
abundance method may be unsuitable for BNF
studies in some NFTs in natural communities,
because the d 15 N in these soils is usually lower
than in agricultural soils (Peoples et al., 1991).
Although some reports have suggested high variability across sites (Broadbent et al., 1980; Selles
et al., 1986), Shearer et al. (1978) suggested that
the spatial variability of d 15 N in many soils may
well be within acceptable limits.
Because of their big sizes and difficulty in
whole-tree sampling, the idea of sampling representative plant parts to estimate BNF for whole
trees is appealing. However, large differences in
d 15 N have been observed among different plant
parts (Shearer et al., 1983), thus questioning the
validity of using the d 15 N in a particular organ to
estimate BNF on a whole-tree basis. More research is needed.
Nitrogen-15 fertilizer enrichment methods
The addition of 15 N-enriched fertilizers to soil
significantly increases the differences in the 15 N
enrichments between soil N and atmospheric N2,
and increases the ease of 15 N detection in plants
and the chances of obtaining more precise and
possibly more accurate estimates of BNF.
Duhoux and Dommergues (1985) suggested that
the 15 N enrichment techniques are probably the
most reliable for measuring BNF in trees. Despite its advantages, 15 N enrichment of soil has
not been used routinely for measuring BNF in
trees, and may relate to the difficulty of adapting
the techniques to the massive sizes of trees, and
their long duration of growth.
Selection of reference plant
The 15 NN/ 14 N ratio in the available soil N needs
to be accurately determined to obtain accurate
estimates of BNF. The reference plant used for
assessing the integrated 15 N enrichment of the N
absorbed by the NFT from soil is therefore very
critical for the 15 N methods. Sanginga et al.
(1990c) and Pareek et al. (1990) are among the
few to have examined the selection of reference
plants for NFTs and clearly showed that highly
erroneous values of BNF can be obtained with
unsuitable reference plants.
Errors due to the use of unsuitable reference
plants are however less critical when BNF is high
(Hardarson et al., 1988); Sanginga et al. (1990c)
showed that the errors in the estimates of BNF
were far greater in the poorer fixer A. albida
than in L. leucocephala. Danso et al. (1986) and
Hardarson et al. (1988) suggested that errors in
BNF estimates due to the reference plant are
quite insignificant when % Ndfa is in the 80%
range, but the % Ndfa in many NFTs fall significantly below 80% (e.g. Gauthier et al., 1985;
Hansen and Pate, 1987a; Sanginga et al., 1990b).
Also, when the 15 N enrichment in soil does not
decline rapidly, the chances for accurate BNF
estimates are greatly improved (Witty and Ritz,
1984). However, there are no reported studies of
BNF in NFTs using slow-release 15 N formulations or 15 N-labelled organic matter. The practical difficulty in incorporating the 15 N-labelled
organic matter into soil in an established tree
stand, limits the applicability of this labelling
method in NFTs. Many studies don't however
need to quantify the N2 fixed accurately while for
some, there may be little need for a reference
plant. For example, the ranking of Rhizobium
strains or NFT genotypes for BNF capacity in a
soil enriched with 15 N should be the same with or
without a reference plant. Under such circumstances, not much advantage is gained by including a reference plant.
The criteria for selecting a reference crop for
grain legumes (Fried et al., 1983) apply to NFTs
as well. The reference plant should be a non-N2fixer, with similar growth and time course of N
uptake as the NFT, unless the 15 N/ 14 N ratio of
soil is stable, and both plants should obtain their
N from a similar soil N pool.
Absence of nitrogenase activity by ARA can
be used to ascertain if a putative reference plant
fixes N2 (Fried et al., 1983). Non-nodulating
isolines have not yet been identified for NFTs. A
closer examination of some of the poorly or
irregularly nodulated NFTs, e.g. Allocasuarina
spp. (Reddell et al., 1986b; Sanginga et al.,
1989a) might reveal some non-nodulating isolines. Allocasuarina, which is poorly nodulated
in nature, may be a good reference plant for
naturally nodulating Casuarina spp. as they may
Nitrogen fixation in trees
have similar rooting habits. Uninoculated NFTs
have sometimes been used as references plants
(Sanginga et al.. 1990e; Gauthier et al., 1985;
Ndoye and Dreyfus, 1988), with the advantage
that BNF can be estimated by two independent
methods, TND and ID. In practice, however, it
has often been difficult to avoid cross-contamination of the uninoculated reference plant, particularly in the field (e.g. Gauthier et al., 1985;
Sanginga et al., 1988).
The selection of fixing and non-N2-fixing
plants with similar growth and N uptake patterns
(Witty, 1983) is easier in annuals than in trees in
which growth progresses through several, often
contrasting seasons which could affect growth of
the reference and N2-fixing plants differently.
However, because the decline in 15 N enrichment
in soil into which a 15 N fertilizer has been added,
decreases with time (Pareek et al., 1990), the
subsequent greater stabilization of 15 N enrichment should minimize potential errors in the
long-term estimations of BNF.
The requirement for the fixing and non-N2fixing plants to obtain their N from the same N
pool does not necessarily mean that plants
should have equal rooting depths. In the hypothetical case in Figure 1, the reference plant No.
2 is for all purposes ideal, and perhaps as good a
reference as is plant No. 4, which although has
shallower roots, but is absorbing its N from the
Fig. 1. The effect of root depth of reference plants on
N/ 14 N uptake. The fixing plant is represented by 1, while 2,
3 and 4 represent potential reference plants. A and B represent soil horizons.
15
183
same zone A, within which the 15 N was incorporated. This therefore allows for some flexibility
in selecting for reference crops, even if their
rooting habits differ slightly from those of the
putative N2-fixer (Fried et al., 1983). For NFTs,
this is very important. Many trees may be absorbing their nutrients predominantly from the
topsoil even though roots may penetrate far
deeper (Anonymous, 1975). Furthermore, plantavailable N in most soils is mainly in the topsoil,
in which case any dilution of 15 N by unlabelled N
from the subsoil (e.g. plant No. 3 in Fig. 1) could
be small. Is it then justified to use non-tree
species, such as grasses and cereals, with different root systems as reference crops for trees?
Pareek et al. (1990), found the 15 N enrichment
of soil extract was similar to that assessed by four
non-fixing weeds and cereal crops, and BNF
estimates were also similar using these as references. Unfortunately, no trees were used as
references. Also, because BNF was very high,
and the influence of reference crop errors is
minimal under such situations, it is difficult to
make a generalization from this study.
A special problem with trees could be the
horizontal spread of roots; they are likely to
absorb N from well outside the zones within
which 15 N was incorporated, particularly if 15 N
was applied as a band. For this reason, Baker et
al. (1990) recommend the trenching of the
perimeter of each replication block and installing
a multi-layered plastic film barrier to a depth of
1.5 m to contain the roots within the zone in
which 15 N was applied.
Rate, formulation, time and frequency of 15 N
fertilizer addition
To enrich a soil with 15 N for BNF measurement,
the atom % 15 N excess, rate, frequency and
method of application of the 15 N fertilizer will all
influence the precision and the accuracy of the
estimate.
Essentially, the amount of N applied should be
small enough not to suppress BNF significantly,
while that for the reference plant should be large
enough to allow good but not necessarily maximum growth. Where the native soil N would
adequately support the growth of the reference
non-N2-fixing plant, the ID method is preferred,
184
Danso et al.
because of the greater simplicity of the ID procedure (only one N rate is used) and calculations.
The AV method is preferred where soil N is low,
and where it may be necessary to eliminate
traces of N2 fixation in the reference plant, or
where nodules from cross-contamination are
likely to form on the reference crop at a lower N
rate. Sanginga et al. (1990c) used both the ID
and Av approaches, to measure BNF in L.
leucocephala and A. albida, with the corresponding uninoculated species included with two other
non-NFTs as reference plants. They found that
at the 10mg N kg -1 soil rate (ID method), the
uninoculated legumes were nodulated, and large
differences in the estimates of BNF by the different reference trees were obtained, including
some negative values for the poorer fixer A.
albida. No nodules were, however, found when
the N applied to the reference plants was raised
to 100 mg N kg soil -1 (AV method), and the
estimates of BNF with all reference plants were
quite close for each NFT. Sanginga et al. (1990c)
observed that the pattern of N uptake by the
reference and fixing plant was closer (a requirement for a satisfactory reference plant) when 100
rather than 10 mg N kg soil -1 was applied.
For perennial pastures, the frequent, small
additions of 15 N to soil is a suitable approach for
measuring BNF (Vallis et al. 1967; 1977) as the
application of N as small doses prevents the
suppressive effects of applied N on BNF, and
also results in a more stable 15 N/ 14 N ratio in soil
(Hardarson et al., 1988; Labandera et al., 1988).
For trees, the use of multiple 15 N fertilizer applications (2-4 a year) probably before periods of
high root activity appears to be an attractive
method, but has been little examined. Preliminary study by N. Sanginga, F. Zapata, S.K.A.
Danso and G.D. Bowen (unpublished) compared the effect of different amounts of 15 Nlabelled fertilizer and the frequency of application on estimates of N, fixed in L. leucocephala
and G. sepium over 9 months. They found that a
single application of 20 mg kg soil -1 decreased
BNF, and that applying this amount in three
splits gave different results, depending on
whether it was repeatedly applied to the same
plot or each split application was on a previously
unlabelled plot.
Sampling of plants
The practical difficulty in harvesting whole trees
increases with age, and it is difficult and labourintensive to attempt to recover all roots. A representative organ would be ideal for BNF estimation, provided the 15 N enrichment is representative for the whole tree. The 15 N enrichments in different organs of trees can however
vary widely (Sanginga et al., 1990e), and raises
the question as to which plant part is most
representative. Aboveground organs might be
preferred for ease of sampling, and the leaves
which contain most of the aboveground N seem
the most appropriate choice. Even where only
leaves are to be sampled, it could be laborious to
sample all leaves. For this reason, Baker et al.
(1990) tested different sampling strategies, and
suggested that the simple procedure of collecting
small numbers of leaves (20 to 60) at random
from the trees was sufficient to estimate % Ndfa
in Leucaena leucocephala. However, measuring
the total BNF should be more problematic
While the % Ndfa in the leaves of soybean
was found to be close to that of the whole plant,
total N fixed differed (Danso and Kumarasinghe,
1990). Studies by Sanginga et al. (1990e) showed
that L. leucocephala roots alone contained as
much as 60% of the N fixed in the whole plant.
This could still well be an underestimate because
of fine-root turnover which can be extremely
large. Thus, any estimates of BNF in trees that
exclude roots should be regarded as underestimates. Allometric relations, worked out with
forest trees (Felker et al., 1982), may be of some
use in estimating total biomass, but even these
are only approximate. The costs and practicalities of doing tree experiments dictate that
the sampling problem needs urgent attention,
especially if one is trying to quantify both aboveground and below-ground nitrogen which will
influence the N available to associated or succeeding plants.
Effect of remobilized N
Perennial plants go through periods of N mobilization and remobilization. Domenach and Kurdali (1989) estimated that retranslocated N is
important for regrowth, with 10% of the N in A.
glutinosa at the end of the growing period being
Nitrogen fixation in trees
traced to retranslocated N. Retranslocation of N,
however, introduces methodological limitations
because the 15N enrichment of the translocated
N in the NFT may be significantly different
(lower) from that of the reference plant. These
differences will therefore be compounded in the
differences in the final 15N enrichments in the
fixing and reference plant, and unless corrected
for, will result in erroneous BNF estimates. The
following equation is used to correct for retranslocated N and also for differences in amounts of
stored N retranslocated from a preceding harvest
(e.g. in sequentially harvested NFTs where 15N is
applied always to a previously unlabelled tree)
on the 15N enrichment at final harvest:
Corrected atom %
N excess (CAE) in plant
15
Where T2 is the time of the final (or later)
harvest and T1 is the harvest immediately preceding regrowth and 15N application, and N is
the total N in plant at the indicated times.
The relative contribution or importance of retranslocated N, however, diminishes with time
(Domenach and Kurdali, 1989) and therefore
the potential error in BNF estimates is reduced
with time. Much of the errors will also depend
on the frequency and severity of pruning.
Domenach and Kurdali (1989) suggested that
harvesting the most recently formed leaves at the
end of the growing season significantly reduced
the influence of nitrogenous reserves on measurements of BNF. Also, the frequent addition
of 15N fertilizers compared to reliance on residual 15N in soil seems to have the advantage, that
plants (both fixing and non-N2-fixing) would be
absorbing higher enriched 15N than already present in the plants, resulting quickly in a narrowing
of the existing 15N enrichment differences in the
plants (Danso et al., 1988).
Effect of litter turnover
Nutrient cycling (largely from litter turnover) is a
major source of nutrients in tree growth (Mahendrappa et al., 1986), and also influences the 15N
enrichment in the soil. The ID approach assumes
185
that both fixing and non-N2-fixing plants are
growing on soil of the same 15N/ 14N ratio. However, with the litter of the fixing plant being
lower in 15N enrichment than that of the non-N2fixing plant, the soil under a NTF may have a
lower 15N enrichment than that under the nonN2-fixing plant. Thus, a crucial assumption of the
methodology, that both the fixing and non-N2fixing plants are sampling N of the same 15N
enrichment is invalidated. Again, this is where
frequent addition of 15N has special advantages.
By superimposing fresh 15N fertilizer on an already low level of 15N enrichment, it should be
possible to override such 15N/ 14N differences in
the two root zones (Danso et al., 1988).
Ureides as a measure of BNF
Another method which has shown much promise
in some grain legumes is the measurement of
BNF based on the composition of nitrogen compounds in the xylem sap. Where ureides are
uniquely produced by BNF, their relative abundance in the total nitrogen exported from the
nodules in the xylem stream to the leaves serves
as an indication of the proportion of N fixed
(Herridge, 1982; McClure et al., 1980). This
approach therefore overcomes problems of
quantifying N uptake from soil separately. However, of 35 NFTs studied by van Kessel et al.
(1988), only two showed a high abundance of
ureides in the xylem sap. Sanginga et al. (1988)
could not find significant correlation between
ureide production and BNF in Leucaena, while
the study by Hansen and Pate (1987a) indicated
that xylem sap analysis was not suitable for
measuring BNF in three Acacia species. Results
to date do not therefore indicate that this method has much potential for measuring BNF in
trees.
Nitrogen fixation potential of trees
Dommergues (1987) defined two parameters:
(i) The nitrogen fixing potential (NFP) of a
species, i.e. the nitrogen fixed with all environmental constraints removed, including
the possible inhibitory effect of soil nitrogen.
186
Danso et al.
However, almost without exception the field
data reported are subject to some environmental constraints and the concept of NFP is
a qualified one, and
(ii) the actual nitrogen fixed (ANF), which is the
resultant of NFP, modified by environmental
constraints.
Dommergues (1987) identified high and low
NFP species. The former included such species
as L. leucocephala and A. mangium for which
records occur of 100 to 300 (sometimes 500) kg
N fixed ha -1 yr -1 ; the latter include such species
as A. albida, A. senegal, and A. pellita (Langkamp et al., 1979) with which fixation has been
reported as less than 20 kg N ha -1 yr -1 . It is
likely that when more species are studied there
will be more of a continuum. The major factor in
high NFP, a high percentage of nitrogen derived
from the atmosphere (% Ndfa), generally appear
to be less affected by environmental conditions
than total N fixed (Danso, 1986; Materon and
Danso, 1991).
Fast growth rate alone is not the sole indicator
of high nitrogen fixation. Some species like the
non-fixing legume Cassia siamea achieve extremely high growth rates because of an extensive root system which, at least in the first year,
explores a considerable volume of soil (S.
Hauser, pers. comm.). Rapid growth rates may
also be achieved in poorly fixing species by efficient retranslocation of absorbed or fixed N.
There are few studies of genotype differences in
the physiological efficiency of use of nitrogen in
legumes, whether it be absorbed from soil or
fixed from the atmosphere.
The occurrence of nodulation
Although the vast majority of the Leguminosae
form nodules, a number do not. In the subfamily Caesalpinioideae, nodulation is largely restricted to the tribe Caesalpinieae and the genus
Chamaecrista from the Cassieae. In the
Mimosoideae, nodulation is general, except for 4
groups within the tribe Mimoseae, and a few
species of Acacia. The only tribe from the
Papilionoideae which appears not to nodulate is
the Dipterygeae. A number of genera in the
Swartzieae do not nodulate (de Faria et al.,
1989).
Cassia siamea, often used in agroforestry systems, does not nodulate and its vigorous growth
is due to a highly effective rooting system. Other
prominent woody legumes lacking nodulation
are the carob ( Ceratonia siliqua ) Bauhinia spp.
(Sprent et al., 1988), Parkia biglobosa (Dommergues, 1987) and Gleditsia spp. (Allen and
Allen, 1981).
Attention is drawn to stem nodulation of
legumes - usually shrubs occurring in wet (or
swampy) sites - and their potential for green manure in rice ecosystems. Stem nodulation, by
Azorhizobium and some strains of Rhizobium
(Alazard et al., 1988) occur on 17 species of
Aeschynomene, 3 species of Sesbania and 1
species of Neptunia (Becker et al., 1988).
The nodulation phenomenon in actinorrhizal
genera (symbiotic with Frankia spp.) needs more
examination. For example, in the Casuarinaeae,
less than half of the 70-80 species have been
examined. In the genus Casuarina, nodulation
appears to be high and nodules at a site are
usually well developed and uniform. By contrast,
for many species of the large genus Allocasuarina, nodules are often absent in the field
(or extremely difficult to find), and frequently
only a few of the individual plants are nodulated
(Reddell et al., 1986b).
Genotypic variation in BNF
Pot studies, and limited field studies, have shown
large differences between genotypes within a
species in their nitrogen fixation. Pot studies give
indications of differences in nitrogen-fixing
potential, but have their shortcomings. Figure 2
indicates not only significant differences in nitrogen fixation between Rhizobium strains, but also
large plant genotype differences within L.
leucocephala which disappeared after some 24
weeks. Herein lies a problem with pot studies:
are the marked genotype differences for the first
24 weeks only a reflection of speed of infection/
response or is the equality of growth in the third
period merely a reflection of root/plant growth
constraints imposed by pot studies? While
screening for genotypic differences by pot/glasshouse studies might be useful, the major findings
must be field-tested. The marked genotypic differences (2.5-fold over 7 months) observed in
Nitrogen fixation in trees
187
In a mixed cropping system like alley cropping, one may not always be looking for the
plant genotype with the highest BNF. Such a
plant may be too vigorous a competitor for
associated crops. In such cases selecting for maximum BNF needs to be accompanied by selection to reduce competition, e.g. selection for
particular types of root systems. The NFT needs
to be considered in the context of the ecosystem,
not its BNF alone.
Plant and Rhizobium factors
Fig. 2. Total nitrogen (mg plant -1 ) of Leucaenu leucocephala
genotypes K636 and K28 inoculated with different
Rhizobium strains during successive cuttings at 12-week intervals. From Sanginga et al. (1990d).
nitrogen fixed by C. equisetifolia in pot studies
by Sougoufara et al. (1987) were also substantiated in the field (Dreyfus et al., 1988).
Sanginga et al. (1990a,b) have shown 2-3-fold
differences in nitrogen fixed between provenances/genotypes of A.
albida and
L.
leucocephala after 12 weeks and 70% differences
in pots after 36 weeks; 3-5-fold differences were
found between 11 provenances of C. equisetifolia
and of C. cunninghamiana. The potential application of such findings is obvious, especially as
many tree species are readily propagated vegetatively. However, selection for high BNF can not
be the only consideration in mixed ecosystems;
for example, it is necessary to select for tolerance of stress conditions as well.
In the presence of appreciable soil nitrogen
some genotypes may fix more nitrogen and absorb less soil nitrogen than others. Genotypic
differences in the effect of soil N on inhibition of
BNF may have important implications for the
nitrogen fixed in second and subsequent seasons
by different genotypes.
The age factor
To date, there are few studies on BNF by trees
in the second and subsequent seasons after planting. Based on acetylene reduction activity and on
nodule mass data, Hansen and Pate (1987b)
concluded that % Ndfa in two species of Acacia
in the first year were 37% and 29%, compared to
only 9% and 2% in the second year. In one
species, the calculated amount of N fixed in the
second year was similar to that for the first year
and for the second species it was 50% greater in
the second year. During the first season, when
plant growth is exponential, BNF will usually
increase with time, e.g. data of Sanginga et al.
(1990d) indicate 55 mg, 120 mg and more than
250mg N fixed per plant of L. leucocephala
grown in pots over successive 12-week periods.
There is a strong case for more detailed
studies of nitrogen fixed in second and subsequent seasons: If as with L. leucocephala,
some 300 kg N or more can be fixed aboveground in a year and if, as data indicate, an
equal amount occurs belowground (Sanginga et
al., 1990e) then soil N levels from litter fall and
sloughed root material may well be sufficient to
suppress BNF. Sanginga et al. (1987) showed
that 40 kg N ha -1 of applied fertilizer reduced
BNF by L. leucocephala by some 50%. A marked reduction in BNF in second or subsequent
seasons may affect management options. Furthermore, as the tree ages, N stored in roots and
stem may reduce BNF by a feedback mechanism
but this needs further study also. If future studies
do show a reduction of BNF in the second or
later seasons, the two postulated mechanisms
above could be discerned by pot studies compar-
188
Danso et al.
ing BNF of seedlings planted to soil from under
trees and adjacent soil outside their influence.
The Rhizobium component
Specificity in nodulation. Some leguminous trees
appear to nodulate only with fast-growing strains
of Rhizobium, some only with the slow-growing
Bradyrhizobium and some with both Rhizobium
and Bradyrhizobium (e.g. Dreyfus and Dommergues, 1981). However, even in each of these 3
groups considerable specificity can occur. For
example, the data of Dreyfus and Dommergues
(1981) show that two of five Rhizobium isolates,
and one of five Bradyrhizobium isolates tested
did not infect A. sieberiana. All five Rhizobium
isolates infected A. farnesiana, but only two of
the 5 Bradyrhizobium infected it. In studies by
Habish and Khairi (1970), A. albida nodulated
with only one isolate from A. albida and 1 of
another 9 isolates from various Acacia species.
The nodulation affinities of some tree legumes
with potential in agroforestry are indicated in
Table 3. At this stage no consistent conclusions
can be made on any relationship between
Bradyrhizobium/Rhizobium specificities and systematic plant classification. However, some patterns may emerge as revisions of larger genera,
such as Acacia, are made.
Effectiveness. Table 4, from Dreyfus and Dommergues (1981), shows clearly that considerable
differences can occur between rhizobia in their
effectiveness on particular hosts, a finding collaborated by other workers, e.g. Olivares et al.
(1988) on A. cyanophylla, A. melanoxylon and
Prosopis chilensis. These data also indicate cases
where rhizobia isolated from a species is either
ineffective or poorly effective on the same
species. It follows that any assumption that indigenous rhizobia are effective and thus no response will be achieved by inoculation with selected effective rhizobia for that species, is
premature.
Thus,
Rhizobium/Bradyrhizobium
strain selection is a critical part of legume tree
Table 3. Rhizobial requirements of some potentially useful leguminous trees
Species
Nodulation by
B and/or R*
Acacia albida
A. constricta
A. holosericea
A. longifolia var. sophorae
A. mearnsii
A. melanoxylon
A. nilotica
A. raddiana
A. saligna
(syn. A. cyanophylla )
A. senegal
A. seyal
B
R
B
R
B
B
R
R
R, B
A. tumida
Albizzia lebbek
Calliandra calothyrus
Erythrina poeppigiana
Gliricidia sepium
R, B
B
R
B
R, B
Leucaena leucocephala
Mimosa scabrella
Prosopis alba
P. chilensis
P. glandulosa
Sesbania grandiflora
R, B
R, B
R
R, B
R, B
R
* B = Bradyrhizobium.
R = Rhizobium.
R
R, B
Reference
Dommergues (1987)
Waldon et al. (1989)
Dreyfus and Dommergues (1981)
Lawrie (1983)
Lawrie (1983); Dommergues (1987)
Lawrie (1983)
Dommergues (1987)
Dommergues (1987)
Barnet et al. (1985)
Olivares et al. 1988)
Dommergues (1987)
Dommergues (1987)
Habish and Khairi (1970)
Dreyfus and Dommergues (1981)
Ribero et al. (1987)
Dommergues (1987)
Dommergues (1987)
Dommergues (1987)
K.O. Awonaike, pers. corn.
Dommergues (1987)
Dommergues (1987)
Torres (1985)
Olivares et al. (1988)
Jenkins et al. (1989)
Dommergues (1987)
Nitrogen fixation in trees
189
Table 4. Nodulation of Acacia species by Rhizobium and Bradyrhizobium (from data of Dreyfus and Dommergues, 1981)
Nodulation
~
Rhizobium
Bradyrhizobium
Host of isolation
A.se.*
A.bi.
A.f.
Host of isolation
A.h.
A.si.
A.bi.
L.1.
Native African species
Acacia albida
A. senegal
A. seyal
A. sieberiana
0
E
e
I
0
E
E
e
0
E
E
0
E
0
E
E
E
0
e
e
E
0
e
E
E
0
E
e
Introduced species
A. bivenosa
A. farnesiana
A. holosericea
A. tumida
I
E
0
0
E
E
0
I
I
E
0
I
E
0
E
e
e
I
e
e
E
0
e
I
I
0
e
0
E = effective nodulation; e = partially effective; I = ineffective;
0 = no nodulation.
*A.se = Acacia senegal; A.bi. = A. bivenosa; A.f. = A. farnesiana; A.h. = A. holosericea; A.si. = A. sieberiana; L.1. = Leucaena
leucocephala.
species evaluation, for both introduced and indigenous species. It may well be, for example,
that Australian species of Acacia may perform
poorly in Latin America or Africa, where indigenous rhizobia may be effective on indigenous
species. but ineffective on introduced species.
The problem of competition between introduced effective strains of rhizobia and poorly
effective indigenous strains (thus reducing the
inoculation response) has not been examined
with tree species. Bowen (1978) pointed out that
species of legumes which nodulate at least partially effectively with a range of rhizobia may
establish and grow reasonably well over large
areas without inoculation and indeed this may be
a major factor in the wide success of many
agricultural legumes. However, inoculation of
NFTs with highly effective rhizobia may not
achieve the desired productivity due to competition by less effective indigenous strains. Specificity in nodulation requirements then becomes an
advantage, allowing the full expression of the
potential
of
the
introduced
Rhizobium/
Bradyrhizobium. Many soils, for example, lack
rhizobia for specific tree legumes, e.g. L.
leucocephala (Sanginga et al., 1989b), and this
would allow the full response to inoculation with
effective rhizobia.
Management practices and environmental effects
on BNF
Effects of management
We have indicated above that some 50% or more
of the tree’s nitrogen may be below ground. This
may well be an underestimate because it neglects
root turnover which can be extremely high, e.g.
40-92% of the standing crop in a forest ecosystem (Fogel, 1985). Some estimates on temperate
forest species indicate the N allocated to fine
roots can be up to 48% (Nadelhoffer et al.,
1985) and it is highly likely that nitrogen does
not retranslocate from senescing fine roots
(Nambiar, 1987). As with severe pruning of
herbaceous legumes (Bowen and Kennedy,
1959), after which a majority of roots and
nodules senesce, severe pruning of L.
leucocephala resulted in the death of approximately half of the nodules in the subsequent 3
weeks (Sanginga et al., 1990d) (root biomass was
not studied). Thus, timing and severity of pruning may allow for some management of underground transfer of fixed nitrogen to associated
crops, as well as regulating root competition
between established hedgerows of NFT’s and
recently planted inter-row annuals. Reduction in
190
Danso et al.
competition between the tree and the annual
species and/or sustained transfer of N from the
tree to the alley crop may be the reasons for the
observation of Duguma et al. (1988) of increased
maize and cowpea yields with increased pruning
frequency and decreased pruning height of L.
leucocephala, G. sepium and S. grandgora.
As indicated above, the accession of nitrogen
to soil by root sloughing, and addition of aboveground litter may decrease BNF in later seasons.
Environmental factors
The actual nitrogen fixed is conditioned by the
plant's nitrogen fixing potential (in association
with
a
highly
effective
Rhizobium/
Bradyrhizobium strain) and by environmental
factors. These may operate in two general ways:
(i) by reducing plant growth and (ii) by direct
effects on the symbiosis, e.g. by nodule physical
conditions by affecting oxygen availability to
bacteriods (Sprent et al., 1988) or by affecting
the biochemistry of fixation, through e.g. molybdenum deficiency. Below, we indicate some of
these factors.
Season
As with herbaceous legumes (e.g. Sprent, 1979),
nitrogen fixation varies considerably within and
between seasons due to plant age and environmental effects on plant growth.
Soil moisture
Soil moisture plays a dominant role in seasonality of nodulation and of nitrogenase activity.
Some tree species have perennial nodules but
many, if not most, shed their nodules and many
fine roots when the soil becomes dry. For example, in the Mediterranean climate of Western
Australia, Hansen et al. (1987) found nitrogenase activity per plant of two species of Acacia to
be greatest in August/September (late winter/
spring) due to a combination of increased
specific acetylene reduction activity of the
nodules and peak numbers of nodules. There
was virtually no nitrogenase activity after
November/ December because of nodule senescence (decreasing soil moisture) through to the
next March/ApriI (arrival of rains), when nodulation recommenced. Habish (1970) recorded increased nodulation of A. mellifera in sand with
increase in soil moisture to an optimum of 15%
soil moisture.
Soil moisture may affect BNF indirectly
through plant growth, and direct effects on infection (Sprent , 1979) and nodule characteristics,
e.g. nodules with a great diffusion barrier to
oxygen (Witty and Minchin, 1988).
Not much is known about nodulation of deeprooted trees in arid and semi-arid areas, in which
the surface layers of soil are dry for considerable
periods, sometimes for years. Prosopis glandulosa has been recorded to nodulate at 10-m soil
depth (Jenkins et al., 1988). Jenkins et al. (1989)
found that Rhizobium dominated in the surface
several metres but that at depth, Bradyrhizobium dominated. The nitrogen fixed by
nodules at depth when the surface horizon is dry
needs more quantitative study. Such nodules
need only fix small amounts of N to be of
significance in nitrogen-deficient soils. Rundel et
al. (1982) suggested some 25-30kg N ha -1 yr -1
may be fixed by P. glandulosa, largely by nodules
at the capillary fringe of the groundwater table,
as they could find no nodules in the surface
layers of the dry soil.
Temperature
Various stages of nodulation and nitrogen fixation of herbaceous legumes are affected markedly by soil temperatures (Sprent, 1979) and this
should hold for tree legumes. High soil temperatures may also affect survival of inoculated
rhizobium in tropical soils (Bowen and Kennedy,
1959). Casuarina cunninghamiana plants supplied with nitrogen fertilizer grew moderately
well at 15°C soil temperature, but plants dependent on Frankia showed no nitrogen fixation and
extremely poor growth at that temperature
(Reddell et al., 1985; Z. Zhang, G.D. Bowen
and F. Zapata, pers. comm.). Nodulation and
nitrogen fixation occurred at 20°C and above.
The optimum soil temperature was 25°C for the
symbiotic plant and 20°C for nitrogen fertilized
plants. The much lower optimum temperature
for symbiotic plants (25°C) than that reported
for nitrogenase activity of detached nodules
(40°C, Bond and MacIntosh, 1975a) suggests
that the major determinant in the intact symbiotic system was the temperature effect on plant
growth. For the two Frankia strains studied, the
Nitrogen fixation in trees
nitrogen fixed g -1 nodule tissue was similar for 5
of the 6 temperature x Frankia combinations at
20°, 25° and 30°C, suggesting a major effect of
assimilate distribution on nodule growth at the
different soil temperatures. Reddell et al. (1985)
suggested that such temperature requirements
for significant nodulation and nitrogen fixation
affects the natural distribution of casuarina
species. Casuarina spp. occur in soils with low N
and seem to be dependent almost entirely on
BNF.
Habish (1970) recorded nodulation and nitrogen fixation by A. mellifera up to 35°C soil
temperature.
Soil nitrogen
Inorganic N inhibits BNF, and appears important in NFTs. Sanginga et al. (1987) reported
that 40 kg N ha -1 of applied fertilizer reduced
BNF by L. leucocephala by some 50%. The
suppressive effect of inorganic N on BNF has
been shown in Alnus spp. and Myrica spp.
(Stewart and Bond, 1961), Coriaria spp. and
Hippophäe spp. (Bond and Mackintosh, 1975b),
C. equisetifolia (Sougoufara et al., 1990), L.
leucocephala (Sanginga et al., 1987, 1988) and
Acacia spp. (Hansen and Pate, 1987a). Few
studies have compared the relative abilities of
different NFT species, provenances, isolines or
clones to fix N2 at different soil N levels. If large
differences in BNF exist between genotypes
within species in response to combined N, this
may reflect on differences in BNF inhibition in
second and subsequent seasons due to the buildup in soil N. This genotypic variation in sensitivity to soil nitrogen has been shown to be important in grain legumes (e.g. Hardarson et al.,
1984). Domenach et al. (1989) reported that
Alnus incana is capable of fixing N2 in the presence of high soil N. Stewart and Bond (1961)
reported that in Alnus spp., but not in Myrica
spp., BNF was considerably enhanced by a low
level of ammonium nitrogen, due to greater
nodule development. Hansen and Pate (1987a)
however observed that BNF in A. pulchella, A.
alata and A. extensa was severely inhibited by
inorganic N.
Soil toxicities
The effect of salinity on nodulation and nitrogen
191
fixation by herbaceous legumes (Sprent, 1979)
may also hold for tree species, many of which
have potential for use in saline soils. For example, many species of Prosopis can withstand
salinities generally considered too brackish for
agricultural crops (Felker et al., 1981). Indeed,
some can grow at salinity levels higher than that
of sea water. Reddell et al. (1986a) found differences between two Frankia strains in their symbiosis with the extremely salt-tolerant species
Casuarina obesa.
There is scope to combat many deleterious soil
conditions (and to increase BNF) by selection of
host genotypes and possibly by Rhizobium/
Bradyrhizobium selection. Selection for tolerance of acidity may be especially relevant to the
survival and persistence of inoculum, but the
tolerance of the plant genome may play the
dominant overall role in BNF. Habish (1970)
recorded growth of A. mellifera with applied
nitrogen at pH 3.8-4.2 but no nodulation and
little plant growth at pH 5-5.5 in the absence of
added nitrogen. More studies are needed to
identify pH-related constraints, such as effects
on rhizosphere growth of the bacterium or effects on infection and nodule development. Halliday and Somasegaran (1982) indicated that the
effect of acidity on Leucaena itself, rather than
on the Rhizobium, is the main factor in acidity
tolerance of genotypes of L. leucocephala. However, many more studies are needed of the sensitivity of the symbiotic partners to soil toxicities
(and other constraints) and the scope for genetic
manipulation.
Nutrition
Low soil phosphate is an even more widespread
constraint to plant growth than acidity. Sanginga
et al. (1991) examined 23 provenances of
Gliricidia sepium and 11 isolines of L.
leucocephala for growth in low-phosphate soils;
2.30-fold and 2.10-fold differences in growth occurred between L. leucocephala and G. sepium
genotypes, respectively, in a low-phosphate soil.
Additions of phosphate had little effect on rhizosphere growth of Rhizobium, but reduced the
time for nodulation and increased nodule numbers and nodule dry weights by 4- to 5-fold,
while having little effect on % Ndfa. The constancy of % Ndfa indicates that effects on BNF
192
Danso et al.
were probably via effects on plant growth (and
assimilate available for BNF) rather than an
effect specifically on the fixation process. The
selection of genotypes of NFTs for high efficiency in uptake and use of phosphate should result
in increased BNF.
Mycorrhizas
Mycorrhizal fungi enhance the uptake of phosphorus and many other nutrients which also
enhance biological nitrogen fixation, e.g. Zn,
Mo, Cu, etc. (Bowen, 1980). Tree growth is
stimulated by the ‘tripartite’ symbiosis of plantnitrogen-fixing microorganism-mycorrhiza. For
example, inoculation of L. leucocephala by
species of Glomus doubled plant growth, increased nodule fresh weight and specific
acetylene reduction (nitrogenase) activity each
by 50% (Purcino, et al., 1986). In a phosphatedeficient soil, De la Cruz et al. (1988) obtained
little increase in the N per plant of A. auriculiformis, A. mungium and Albizzia falcata
from the inoculation with Rhizobium alone, but
8- to 25-fold increases when inoculated both with
Rhizobium and selected vesicular arbuscular
mycorrhizal (VAM) fungi. These were paralleled
by acetylene reduction increases. Mycorrhizal
fungi differed in their effectiveness, and there is
considerable scope for selection of highly efficient mycorrhizal fungi.
Vesicular arbuscular mycorrhizas are by far the
most dominant form of mycorrhiza, especially on
leguminous trees. Some trees, e.g. Allocasuarina
and Casuarina are infected by both ectomycorrhiza and VAM (Reddell et al., 1986b),
ectomycorrhiza increasing in occurrence in extremely low-phosphate soils in both casuarinas
and in leguminous trees (Hogberg, 1986).
Technology of inoculation with ectomycorrhizal fungi is now fairly well established (Marx
et al., 1984) mainly because they are easy to
grow in artificial media. The inability to grow
VAM fungi in artificial media makes inoculation
on a broad scale difficult because inoculum must
be raised on living plant roots. However, inoculation with appropriately raised root inoculum is
quite feasible in nursery situations. Thus inoculation of the tree species with selected mycorrhizal
fungi before outplanting is quite feasible. Additionally, such outplanting of infected tree stock
will supply an inoculum source for interplanted
legumes and cereals, although more research is
needed on rates of spread of the fungus to such
crops. One may, however, not always get a
mycorrhizal response as many soils already contain populations of effective mycorrhizal fungi.
Identifying the soils/plant species which will respond to inoculation is a high research priority.
Conclusion and future research needs
The rising demand for increased food and fuelwood production by resource-poor farmers and
concern for environmental degradation under intensive agriculture have increased our interest in
agroforestry and nitrogen-fixing trees (NFTs).
Large differences in the biological nitrogenfixing (BNF) capacities of different NFTs have
been noted, and have been classified broadly
into high- and low-NFT species. While species
like Leucaena leucocephala and Sesbania rostrata
can accumulate up to 500 (or more) kg N ha -1
yr -1 of N2 fixed in their biomass, others like
many Acacia species and S. sesban derive substantially less (often <50 kg N ha-1 yr-1). Also,
within each species of NFT examined, large differences in BNF capacities exist. Such genotypic
differences should be exploited for enhancing the
BNF contribution in agroforestry. The 15N isotope labelling techniques are useful for assessing
genotypic differences in BNF.
Inorganic N inhibits BNF in NFTs. This raises
some important questions, such as (i) does the
rapid litter and N turnover commonly associated
with tree growth limit the period of N2 fixation in
NFTs to their first few years of growth and (ii)
are there substantial differences in the tolerance
of BNF to inorganic N in different NFTs? These
important questions have received scanty attention and should be examined. Differences in the
tolerance of different NFT genotypes to inorganic N could be exploited to extend the duration of BNF contribution to soil and plants.
Management practices like pruning intensity
and frequency will not only affect BNF in trees,
but may increase the rate of N release from
senescing roots and nodules, and thus the soil N
status and N transfer to companion crops. These
management aspects have been little studied.
Nitrogen fixation
The 15 N isotope labelling technique will be very
useful in such studies.
It is necessary to examine soils for the presence of effective Rhizobium and Frankia strains
for NFTs to be used in agroforestry, particularly
when these trees are being newly introduced.
The few available data suggest that in many soils
Rhizobium or Frankia inoculation would increase BNF and growth of many NFTs. Inoculation with highly effective strains at the nursery
stage appears to be an effective way of enhancing nodulation success prior to field establishment, and is highly recommended. This can be
complemented by mycorrhizal inoculation to increase NFT growth and BNF in low-P soils.
References
Alarard D. Ndoye I and Dreyfus B 1988 Sesbania rostrata
and other stem nodulated legumes. In Nitrogen Fixation:
A Hundred Years. Eds. H Bothe, F J de Bruijn and W E
Newton. pp 765-769. Gustav Fischer, Stuttgart, Germany.
Allen O N and Allan E K 1981 The Leguminosae: A Source
Book of Characteristics. Uses and Nodulation. University
of Wisconsin Press. Madison. 812p.
Anonymous 1975 Root activity patterns of some tree crops:
Results of a Five-Year Coordinated Research Programme
of the Joint FAO/IAEA Division of Atomic Energy in
Food and Agriculture. STI/DOC/10/170, IAEA, Vienna,
Austria.
Baker D D. Wheeler K A, and Fried M 1990 Estimation of
biological dinitrogen fixation by 15 N-dilution in tree plantations. I. Sampling strategies and first harvest results for
three legume species. Paper presented at fourth African
Association for Biological Nitrogen Fixation. International
Institute of Tropical Agriculture, Ibadan, Nigeria, 24-28
September 1990.
Barnet Y M, Catt P C and Hearne D H 1985 Biological
nitrogen fixation and root-nodule bacteria ( Rhizobium sp.
and Bradyrhizobium sp.) in two rehabilitating sand dune
areas planted with Acacia spp. Aust. J. Bot. 33, 595-610.
Becker M, Ladha J K and Ottow J C G 1988 Stem nodulating
legumes as green manure for lowland rice. Phil. J. Crop
Sci. 13: 121-127.
Bond G and Mackintosh A H 1975a Diurnal changes in
nitrogen fixation in the root nodules of Casuarina, Proc.
Roy. Soc. London. B 192, 1-12.
Bond G and Mackintosh A H 1975b Effect of nitratenitrogen on the nodule symbioses of Coriaria and Hippophaeë. Proc. Roy. Soc. Lond. B. 190, 109-209.
Bowen G D 1978 Dysfunction and shortfalls in symbiotic
response. In Plant Disease - An Advanced Treatise. Vol.
III. Eds. J G Horsfall and E B Cowling. pp 231-256.
Academic Press, New York.
Bowen G D 1980 Mycorrhizal roles in tropical plants and
in
trees
193
ecosystems. In Tropical Mycorrhiza Research, Ed. P
Mikola. pp 165-190. Oxford University Press. Oxford,
UK.
Bowen G D and Kennedy M M 1959 Effect of high soil
temperature on Rhizobium spp. Old. J. Agric. Sci. 16.
17-197.
Broadbent F E, Rauschkolb R S. Lewis K A and Chang G Y
1980 Spatial variability in nitrogen-15 and total nitrogcn in
some virgin and cultivated soils. Soil Sci. Soc. Am. J. 44.
524-527.
Charreau C and Vidal P 1965 Influence de l'Acacia albida
Del sur le sol, nutrition minérale et rendements des mils
Pennisetum au Sénégal. Agron. Trop. 20. 600-626.
Dalton D A and Zobel D B 1977 Ecological aspects of
nitrogen fixation by Purshia tridentata. Plant and Soil 48.
57-80.
Danso S K A 1985 Methods for estimating biological nitrogcn
fixation. In Biological Nitrogen Fixation in Africa. Eds. H
Ssali and S O Keya. pp 213-243. Nairobi MIRCENS.
Nairobi. Kenya.
Danso S K A, Hardarson G and Zapata F 1986 Assessment
of dinitrogen fixation potentials of forage legumes with 15 N
techniques: In Proceedings of a Workshop on Potentials of
Forage Legumes in Farming Systems of Sub-Saharan Africa. Eds. I Haque, S Jutzi and P J H Neate. pp 26-57.
ILCA, Addis Ahaba.
Danso S K A 1986 Review: Estimation of N2-fixation by
isotope dilution: An appraisal of techniques involving 15 N
enrichment and their application - comments. Soil Biol.
Biochem. 18, 243-254.
Danso S K A, Hardarson G and Zapata F 1988 Dinitrogen
fixation estimates in alfalfa-ryegrass swards using different
nitrogen-15 labeling methods. Crop Sci. 28, 106-110.
Danso S K A and Kumarasinghe K S 1990 Assessment of
potential sources of error in nitrogen fixation measurements by the nitrogen-15 isotope dilution technique. Plant
and Soil 12.5, 87-93.
Dela Cruz R E, Manalo M O, Aggangan N S and Tambalo J
D 1988 Growth of three legunme trees inoculated with VA
mycorrhizal fungi and Rhizobium Plant and Soil 108,
111-115.
Delwiche C C and Steyn P L 1970 Nitrogen isotope fractionation in soils and microbial reactions. Environ. Sci. Tech. 4,
927-935.
Domenach A M and Kurdali F 1989 Influence des réserves
azotées sur la formation des feuilles d' Alnus glutinos et
ses conséquences dans I'estimation de la fixation d'azote.
Can. J. Bot. 67. 865-871.
Domenach A M. Kurdali F and Bardin R 1989 Estimation of
symbiotic dinitrogen fixation in alder forest by thc method
based on natural 15 N abundance. Plant and Soil 118,
5 1-59.
Dommergues, Y R 1987 The role of biological nitrogen
fixation in agroforestry. In Agroforestry, a Decade of
Development. Eds. H A Steppler and P K K Nair. pp
245-271. ICRAF, Nariobi, Kenya.
Dreyfus B L and Dommergues Y R 1981 Nodulation of
Acacia species by fast- and slow-growing tropical strains of
Rhizobium. Appl. Environ. Microb. 41, 97-99.
Dreyfus B L, Diem H G and Dommergues Y R 1988 Future
194
Danso et al.
directions for biological nitrogen fixation research. Plant
and Soil 108, 191-199.
Dugama B, Kang B T and Okali D U U 1988 Effect of
pruning intensities of three woody legumes species grown
in alley cropping with maize and cowpea on an alfisol.
Agrof. Systems 6, 19-35.
Duhoux E and Dommergues 1985 The use of nitrogen fixing
trees in forestry and soil restoration in the tropics. In
Biological Nitrogen Fixation in Africa. Eds. H Ssali and S
0 Keya. pp 384-400. Nairobi MIRCEN, Nairobi. Kenya.
Faria de S M, Lewis G P, Sprent J I and Sutherland J M 1989
Occurrence of nodulation in the Leguminosae. New
Phytol. III, 607-619.
Felker P, Clark P R, Lang A E and Pratt P F 1981 Salinity
tolerance of the tree legumes mesquite ( Prosopis glandulosa var. Torrey, P. vektina and P. articulata) algarrobo ( P.
chilensis), Kiawe ( P. palliba ) and tamarugo ( P. tamarugo)
grown in sand culture on nitrogen-free media. Plant and
Soil 61, 311-317.
Felker P, Clark P R, Osborn J F and Cannel G H 1982
Biomass estimation in a young stand of mesquite ( Prosopis
spp.), Ironwood (Olneya tesota), Palo Verde ( Cercidium
floridium, and Parkinsonia aculeata), and Leucaena
(Leucaena leucocephala ). J. Range Manag. 35, 87-89.
Fogel R 1985 Roots as primary producers in below-ground
ecosystems. In Ecological Interactions in Soil. Eds. A H
Fitter, D Atkinson, D J Read and M Usher. pp 23-26.
Brit. Ecol. Soc. Spec. Publ. 4. Blackwell Scientific.
Fried M and Broeshart H 1975 An independent measurement
of the amount of nitrogen fixed by legume crops. Plant and
Soil 43, 707-711.
Fried M, Danso S K A and F Zapata 1983 The methodology
of measurement of N2 fixation by nonlegumes as inferred
from field experiments with legumes. Can. J. Microbiol.
29, 1053-1062.
Fried M and Middleboe V 1977 Measurement of amount of
nitrogen fixed by a legume crop. Plant and Soil 47, 713715.
Gauthier D. Diem H G, Dommergues Y R and Ganry F
1985 Assessment of N2 fixation by Casuarina equisetifolia
inoculated with Frankia ORS021001 using 15N methods.
Soil Biol. Biochem. 17, 375-379.
Habish H A 1970 Effect of certain soil conditions on nodulation of Acacia spp. Plant and Soil 33, 1-6.
Habish H A and Khairi S L M 1970 Nodulation of legumes in
the Sudan. II. Rhizobium strains and cross inoculations of
Acacia spp. Expt. Agric. 6, 171-176.
Halliday J, Somasegaran P 1982 Nodulation, nitrogen fixation and Rhizobium strain affinities in the genus Leucaena.
In Leucaena Research in the Asian-Pacific Region. pp
33-37. IDRC. Ottawa, Ontario, Canada.
Hansen A P and Pate J S 1987a Evaluation of the 15N natural
abundance method and xylem sap analysis for assessing N2
fixation of understorey legumes in Jarrah ( Eucalyptus marginata Donn ex Sm.) forest in S.W. Australia. J. Expt. Bot.
38, 1446-1458.
Hansen A P, Pate J S 1987b Comparative growth and symbiotic performance of seedlings of Acacia spp. in defined
pot culture or as natural understorey components of a
eucalypt forest ecosystem in S.W. Australia. J. Exp. Bot.
38, 13-35.
Hansen A P, Pate J S, Hansen A and Bell D T 1987 Nitrogen
economy of post-fire stands of shrub legumes in Jarrah
(Euculyptus murginata) forests of S.W. Australia. J. Exp.
Bot. 38, 26-41.
Hardarson G, Zapata F and Danso S K A 1984 Effect of
plant genotype and nitrogen fertilizer on symbiotic nitrogen fixation by soybean cultivars. Plant and Soil 82, 397405.
Hardarson G, Danso S K A and Zapata F 1988 Dinitrogen
fixation measurements in alfalfa-ryegrass swards using nitrogen-I5 and influence of the reference crop. Crop Sci.
28, 101-105.
Hardy R W F, Holsten R D, Jackson E K and Burns R C
1968 The acetylene-ethylene assay for N2 fixation: laboratory and field evaluation. Plant Physiol. 43, 47-81.
Herridge D F 1982 Relative abundance of ureides and nitrate
in plant tissues of soybean as a quantitative assay of
nitrogen fixation. Plant Physiol. 70, 1-6.
Hogberg P 1986 Soil nutrient availability, root symbioses and
tree species composition in tropical Africa: A review. J.
Trop. Ecol. 2, 359-372.
Hogberg P and Kvarström M 1982 Nitrogen fixation by the
woody legume Leucaena leucocephala in Tanzania. Plant
and Soil 66, 21-28.
Jenkins M B, Virginia R A and Jarrel W M 1988 Depth
distribution and seasonal population of mesquite-nodulating rhizobia in warm desert ecosystems. Soil Sci. Soc. Am.
J. 52, 1644-1650.
Jenkins M B, Virginia R A and Jarrel W M 1989 Ecology of
fast-growing and slow-growing mesquite-nodulating
rhizobia in Chihuahuan and Sonoran Desert ecosystems.
Soil Sci. Soc. Am. J. 53, 543-549.
Kessel C van, Nakao P, Roskoski J P and Kevin K 1988
Ureide production by N2-fixing leguminous trees. Soil
Biol. Biochem. 20, 891-897.
Knowles R 1980 Nitrogen fixation in natural plant communities and soils. In Methods for Evaluating Biological
Nitrogen Fixation. Ed. F J Bergersen. pp 557-582. Wiley,
New York.
Labandera C. Danso S K A, Pastorini D, Curbelo S and
Martin V 1988 Nitrogen fixation in a white clover-fescue
pasture using three methods of nitrogen-15 application and
residual nitrogen-15 uptake. Agron. J. 80, 265-268.
Lal R 1989 Agroforestry systems and soil surface management of a tropical alfisol. II. Water runoff, soil erosion and
nutrient loss. Agrof. Systems 8, 97-111.
Langkamp P J, Swinden L B and Dalling M J 1979 Nitrogen
fixation (acetylene reduction) by Acacia pellita on areas
restored after mining at Groote Eylandt, Northern Territory. Aust. J. Bot. 27, 353-361.
Lawrie A C 1983 Relationships among rhizobia from native
Australian legumes. Appl. Environ. Microb. 45, 18221828.
Liya S M, Odu C T I, Agboola A A and Mulongoy K 1990
Estimation of N2 fixation by nitrogen fixing trees in the
subhumid tropics using 15-N dilution and difference methods. Paper presented at fourth African Association for
Nitrogen fixation in trees
Biological Nitrogen Fixation, IITA, Ibadan, 24-28 September 1990.
Mahendrappa M K, Foster N W, Weetman G F and Krause
H H 1986 Nutrient cycling and availability in forest soils.
Can. J. Soil Sci. 66, 547-572.
Mariotti A 1983 Atmospheric nitrogen is a reliable standard
for natural 15 N abundance measurements. Nature 303,
685-687.
Marx D H, Cordell C E, Kenney D S. Mexal J G, Artmann J
D, Riffle J W and Molina R J 1984 Commercial vegetative
inoculum of Pisolithus tinctorius and inoculation techniques
for development of ectomycorrhizae on bare-root tree
seedlings. For. Sci. Monogr. 25, 1-101.
Materon L A and Danso S K A 1991 Nitrogen fixation in two
annual medic legumes as affected by inoculation and seed
density. Field Crops Res. ( In press ).
McAuliffe C, Chamblee D S. Urige-Arango H and
Woodhouse Jr W W 1958 Influence of inorganic nitrogen
on nitrogen fixation by legumes as revealed by 15 N. Agron.
J. 50, 334-337.
McClure P R, Israel D W and Volk R J 1980 Evaluation
of the relative ureide content of xylem sap as an indicator of N2 fixation in soybeans. Plant Physiol. 66, 720725.
Nadelhoffer K J, Aber J D and Melillo J M 1985 Fine roots,
net primary production and soil nitrogen availability: A
new hypothesis. Ecology 66, 1377-1390.
Nambiar E K S 1987 Do nutrients retranslocate from fine
roots? Can. J. For. Res. 17, 913-918.
Ndoye I and Dreyfus B 1988 N2 fixation by Sesbania rostrata
and Sesbania sesban estimated using 15 N and total N
difference methods. Soil Biol. Biochem. 20, 209-213.
Olivares J, Herrera M A and Bedmar E J 1988 Woody
legumes in arid and semi-arid zones. The Rhizobium Prosopis chilensis symbiosis. In Nitrogen Fixation by
Legumes in Mediterranean Agriculture. Eds. D P Beck
and L A Materon. pp 65-71. ICARDA. Kluwer Academic
Publishers, Dordrecht, The Netherlands.
Pareek R P, Ladha J K and Watanabe I 1990 Estimating N2
fixation by Sesbania rostrata and S. cannabina (Syn. S.
aculeata ) in lowland rice soil by 15 N dilution method. Biol.
Fertil. Soils. ( In Press ).
Peoples M B, Bergersen F J, Turner G L, Sampet C.
Rerkasem B. Bhromsiri A. Nurhayati D P, Faisah A W.
Sudin M N, Norhayati M and Herridge D F 1991 Use of
natural enrichment of 15 N in soil mineral N for the measurement of symbiotic N2 fixation. In Symposium on the
Use of Stable I1sotopes in Plant Nutrition, Soil Fertility and
Environmental Studies IAEA, Vienna, 1-5 October 1990.
( In Press).
Purcino A A C, Lurlarb C, Lynd J Q 1986 Mycorrhiza and
soil fertility effects with growth, nodulation and nitrogen
fixation of Leucaena grown on a typic eutrostox. Comm.
Soil Sci. Plt. Anal. 17, 473-489.
Reddell P, Bowen G D and Robson A D 1985 The effects of
soil temperature on plant growth. nodulation and nitrogen
fixation in Casuarina cunninghamiana. New Phytol. 101,
441-450.
Reddell P, Foster R C and Bowen G D 1986a Effects of
195
salinity on nodulation and nitrogen fixation by
Casuarinaceae. New Phytol. 102. 397-408.
Reddell P, Robson A D and Bowen G D 1986b Nodulation
of Casuarinaceae in relation to host species and soil properties. Aust. J. Bot. 34, 435-444.
Rennie R J, Rennie A D and Fried M 1978 Concepts of "N
usage in dinitrogen fixation studies. In Isotopes in Biological Dinitrogen Fixation. pp 107-133. STI/PUB/478. International Atomic Energy Agency, Vienna, Austria.
Rennie K J and Rennie A D 1983 Techniques for quantifying
N2 fixation in association with non-legumes under field and
greenhouse conditions. Can. J. Microbiol. 29. 1022-1035.
Ribeiro W Q Jr, Lopes E S and Franio A A 1987 Efficiencia
de estribes de Bradyrhizobium spp. para quatro
leguminosas arbóreas e competitividade das estripes em
Albizia lebbek em latossolo ácido. R. bras. Ci. Solo. 11,
275-282.
Roskoski J P 1981 Nodulation and N2-fixation by Inga
jinicuil, a woody legume in coffee plantations. I. Measurements of nodule biomass and field C2H2 reduction rates.
Plant and Soil 59, 201-206.
Roskoski J P, Montano J, van Kessel C and Castillejo G 1982
Nitrogen fixation by tropical woody legumes: Potential
source of soil enrichment. In BNF Technology for Tropical
Agriculture. Eds. P H Graham and S C Harris. pp 447454. CIAT, Cali,
Rundel P W, Nilsen E T, Sharrh M R. Virginia R A, Jarrell
W M, Kohl D H and Shearer G B 1982 Seasonal dynamics
of nitrogen cycling for a Prosopis woodland in the Sonoran
Desert. Plant and Soil 67, 343-353.
Sanginga N, Mulongoy K and Ayanaba A 1985 Effect of
inoculation and mineral nutrients on nodulation and
growth of Leucaena leucocephala. In Biological Nitrogen
Fixation in Africa. Eds. H Ssali and S O Keya. pp 419427. Nairobi MIRCEN, Nairobi, Kenya.
Sanginga N, Mulongoy K and Ayanaba A 1986 Inoculation
of Leucaena leucorephala (Lam.) de Wit with Rhizobium
and its nitrogen contribution to a subsequent maize crop.
Biol. Agric. Hortic. 3, 341-347.
Sanginga N, Mulongoy K and Ayanaba A 1987 Evaluation of
indigenous strains of Rhizobium for Leucaenca de Wit in
Nigeria. In Les Arbres Fixateurs d'Azote et L'Amélioration Biologique de la Fertilité du Sol. Colloques et
Séminaires. pp 416-436. ORSTOM, Paris.
Sanginga N, Mulongoy K and Ayanaha A 1988 Nodulation
and growth of Leucaena leucocephala (Lam.) de Wit as
affected by inoculation and N fertilizer. Plant and Soil 112,
129-135.
Sanginga N. Danso S K A and Bowen G D 1989a Nodulation
and growth response of Allocasuarina and Casuarina
species to phosphorus fertilization. Plant and Soil 118,
125-132.
Sanginga N, Mulongoy K and Ayanaba A 1989b Effectivity
of indigenous rhizobia for nodulation and early nitrogen
fixation with Leucaena leucocephala grown in Nigerian
Soils. Soil Biol. Biochem. 21, 231-235.
Sanginga N, Bowen G D and Danso S K A 1990a Genetic
variability in symbiotic nitrogen fixation within and between provenances of two Casuarina species using the
196
Nitrogen fixation in trees
15N-labelling methods. Soil Biol. Biochem. 22. 539-547.
Sanginga N, Bowen G D and Danso S K A 1990b Assessment of genetic variability for N2 fixation between and
within provenances of Leucaena leucocephala and Acacia
albida estimated by "Nlabelling techniques. Plant and Soil
127, 169-178.
Sanginga N, Danso S K A, Zapata F and Bowen G D 1990c
Influence of reference trees on N2-fixation estimates in
Leucaena leucocephala and Acacia albida using 15N-labelling techniques. Biol. Fertil. Soils 9, 37-42.
Sanginga N, Zapata F, Danso S K A and Bowen G D 1990d
Effect of successive cutting on nodulation and nitrogen
fixation of Leucaena leucocephala using 15N dilution and
the difference methods. In Plant Nutrition-Physiology and
Applications. Ed. M L van Beusichem. pp 667-674.
Kluwer Academic Publishers, Dordrecht, The Netherlands.
Sanginga N, Zapata F, Danso S K A and Bowen G D 1990e
Effect of successive cuttings on uptake and partitioning of
15N among plant parts of Leucaena leucocephala. Biol.
Fertil. Soils 9, 37-42.
Sanginga N. Bowen G D and Danso S K A 1991 Intraspecific variation in growth and P accumulation of
Leucaena leucocephala and Gliricidia sepium as influenced
by soil and phosphate status. Plant and Soil 133, 201-208.
Selles F, Karamanos R E and Kachangski R G 1986 The
spatial variability of nitrogen-IS and its relation to the
variability of other soil properties. Soil Sci. Soc. Am. J. 50,
105-110.
Shearer G, Kohl D H and Commoner B 1974 A steady state
model of isotopic fractionation accompanying nitrogen
transformations in soil. Soil Sci. Soc. Am. Proc. 38, 315321.
Shearer G, Kohl D H and Chien S H 1978 The nitrogen-15
abundance in a wide variety of soils. Soil Sci. Soc. Am. J.
42, 899-902.
Shearer G, Kohl D H, Virginia R A, Brjan B A, Skeeters J
L, Nilsen E T, Sharifi M R and Rundel P W 1983
Estimates of N2 fixation from variation in the natural
abundance of 15N in Sonoran Desert ecosystems.
Oecologia 56, 365-373.
Shearer G and Kohl H 1986 N2-fixation in field settings:
Estimations based on natural 15N abundance. Aust. J.
Plant Physiol. 13, 699-757.
Shearer G and Kohl D 1988 Estimates of N, fixation in
ecosystems: The need for and basis of the 15N natural
abundance method. In Stable Isotopes in Ecological Research. Eds. P W Rundel, Jr Ehleringer and K A Nagy pp
342-374. Springer-Verlag, New York.
Sougoufara B, Dohoux E, Corbasson M and Dommergues Y
R 1987 Improvement of nitrogen fixation by Casuarina
equisetifolia through clonal selection: A research note:
Arid Soil Res. Rehab. 1, 129-132.
Sougoufara B, Danso S K A, Diem H G and Dommergues Y
R 1990 Estimating N2 fixation and N derived from soil by
Casuarina equisetifolia using labelled 15N fertilizer: Some
problems and solutions. Soil Biol. Biochem. 22, 695-701.
Sprent J I 1979 The Biology of Nitrogen Fixing Organisms.
McGraw-Hill, London. 196 p.
Sprent J I, Becana M and Sutherland J M 1988 Optimising
nitrogen fixation in legume crops and trees. In Nitrogen
Fixation: A Hundred Years After. Eds. H Bothe, F J de
Bruijn and W E Newton. pp 725-733. Gustav Fischer,
Stuttgart, Germany.
Stewart W D P and Bond G 1961 The effect of ammonium
nitrogen on fixation of elemental nitrogen in Alnus and
Myrica. Plant and Soil 16, 347-359.
Stewart W D P and Pearson M C 1967 Nodulation and
nitrogen-fixation by Hippophaë rhamnoides L. in the field.
Plant and Soil 26, 348-360.
Torres M E 1985. Biological nitrogen fixation in Prosopis
tamarugo and P. alba. In The Current State of Knowledge
on Prosopis tamarugo. Ed. M A Wabit. pp 413-420. FAO
Regional Office, Latin America.
Vallis I, Haydock K P, Ross P J and Henzell E F 1967
Isotopic studies on the uptake of nitrogen by pasture
plants. III. The uptake of small additions of 15N-labelled
fertilizer by Rhodes grass and Townsville lucerne. Aust. J.
Agric. Res. 18, 865-877.
Vallis I, Henzell E F and Evans T R 1977 Uptake of soil
nitrogen by legumes in mixed swards. Aust. J. Agric. Res.
28, 413-425.
Virginia R A 1986 Soil development under legume tree
canopies. For. Ecol. Manag. 16, 69-79.
Waldon H B, Jenkins M B, Virginia R A and Harding E E
1989. Characteristics of woodland rhizobia1 populations
from surface and deep soil environments of the Sonoran
Desert. Appl. Environ. Microb. 55, 3058-3064.
Witty J F 1983 Estimating N2-fixation in the field using 15N
labelled fertilizer: Some problems and solutions. Soil Biol.
Biochem. 15, 631-639.
Witty J F and Minchin F R 1988 Measurement of nitrogen
fixation by the acetylene reduction assay: Myths and Mysteries. In Nitrogen Fixation by Legumes in Mediterranean
Agriculture. Eds. D P Beck and L A Materon. Kluwer
Academic Publishers, Dordrecht, The Netherlands.
Zaharah A R. Sharifuddin H A H and Subrarnaniam R 1986
Nitrogen fixation by Leucaena leucocephala as measured
by N-15 dilution technique. Pertanika 9, 17-22.
Plant and Soil 141: 197-209, 1992
© 1992 Kluwer Academic Publishers. Printed in the Netherlands.
PLSO SA11
Trends in biological nitrogen fixation research and application
JUNJI ISHIZUKA
Faculty of Agriculture, Kyushu University, Fukuoka 812, Japan
Key words: actinorhizal tree, Azolla, biological nitrogen fixation (BNF), Blue-green algae, Frankia,
legume, Rhizobium
Table of contents
Abstract
Introduction
Rice paddies
Tree ecosystems
Upland crop yields
Associative N2-fixing symbioses
Legume-rhizobium symbiosis
Conclusion
References
Abstract
In the world each year 17.2 X 107 tons of N are biologically fixed. Biological nitrogen fixation (BNF)
contributes to plant production in arable lands and in natural ecosystems. Research to improve BNF is
progressing through the breeding of efficient N-fixing organisms and host plants, selection of the best
combinations of host plant and microsymbiont, and by the improvement of inoculation techniques and
field management. Biotechnology is useful for the creation of promising N2-fixing organisms. However,
to increase plant production through enhanced BNF the constraints in establishing effective N2-fixing
systems in the field should be understood and eliminated.
Introduction
Since World War II, N fertilizer use has been
rising to increase crop yields. Expansion of industrial fertilizer production, at lower cost, encouraged their use, especially in developed countries. Therefore, enthusiasm for the agronomic
use of BNF has tended to decline in those countries. At the same time the current recognition
that increased industrial consumption of energy
produced by burning fossil-fuel has led to world-
wide increases in atmospheric CO2 and global
warming, causing climatic changes. Reduction of
energy produced by fossil-fuel consumption is
becoming increasingly important, and the effective use of BNF in agriculture would contribute
to this aim.
Total world biological N2 fixation is 17.2 x 10 7
tons per year. This figure, three times that of
industrially fixed N, demonstrates the significance of biological N2 fixation in agricultural and
natural N cycles. Organisms capable of fixing
198
Ishizuka
molecular N are all procaryots, as shown in
Table 1. Part of these species are free-living N2
fixers. They live all over the earth and play an
important role in pastures, forestry, arable land
and in natural N cycles. However, amounts of N2
fixed by free-living bacteria are generally low.
Other procaryots that live symbiotically include blue-green algae (BGA), Frankia, and
Rhizobia. They associate with an aquatic fern,
with actinorhizal plants, and with legumes, respectively, fix larger amounts of N2 than do
free-living bacteria and are important in agriculture. Enhancement of the symbiotic N2-fixing
capacity of legumes has been pursued since
legume cropping was found to increase soil fertility. In tropical and subtropical regions, the
Azolla-Anabaena association operative in paddy
fields, has been extensively investigated. Research on Frankia symbiosis is particularly relevant in deforestation and afforestation programs. This paper deals with trends in research
and the application of BNF in various
ecosystems.
Rice paddies
Before industrial production of chemical fertilizers began, the overall nutrient balance was
slightly negative, and soil fertility decreased, especially in upland soils, because each year nutrients were removed in produce. However, it is
reported that soil fertility in paddy fields is more
sustainable as compared with upland fields, and
that certain levels of rice yield are maintained for
long periods without fertilizer application
(Konishi and Seino, 1961).
The sustaining of fertility in paddy fields, despite annual removal by plants, is largely due to
BNF. From the results of a long-term field experiment, Japanese scientists (Konishi and
Seino, 1961) estimated that at least 20 kg N ha -1
per year was fixed in paddy fields where no N
fertilizer was applied. Alimagno and Yoshida
(1977) showed that BNF in Philippine rice fields
ranged from 2.3 to 33.3 kg ha -1 .
The BNF by free-living bacteria contributes
significantly to rice production in paddy fields
Table 1. List of nitrogen fixing organisms (Bothe et al., 1983; modified by the author)
Free-living nitrogen-fixers
Genus
Species
Obligate aerobes
Azotobacter
azospirillum
vinelandii, beijerinckii
brasilense, lipoferum
Facultative anaerobic bacteria
Klebsiella
Bacillus
pneumoniae
polymyxa, macerans
Obligate anaerobes
Clostridium
Desulfovibrio
pasteuriunum
vulgaris, desulfuricans
Phototrophic bacteria
Rhodospirillum
rubrum, photometricum
Blue-green algae
Nostoc
Anabaena
muscorum
cylindrica, variabilis
Symbiotic nitrogen-fixing systems
Host genus
Microsymbiont
Legume
Glycine
Vigna
Bradyrhizobium
Rhizobium
Non-legume
Alnus, Myrica
Actinomycetes 'Frankia'
Lichens
Collema
Nostoc
Waterfern
Azolla
Anabaena
Research trends in BNF
where insufficient N fertilizer is applied. The
major free-living N2-fixers in paddy fields are
heterotrophic and phototrophic bacteria and
BGA. Inoculation of rice plants with heterotrophic bacteria was conducted in India (Dewan
and Rao, 1979; Rao et al., 1983). Beneficial
effects on yield and N uptake were shown in
some fields with low rates of N application.
In pot experiments in the Philippines, tillering
and grain filling rates of rice plants were enhanced by inoculation with Azospirillum lipoferum, but effects of inoculation on total straw
dry weight and BNF were not observed at harvest time (Watanabe and Lin, 1984). Thus, benefits of inoculation with these species, on yield
and N2 fixation, are not manifest, especially not
in fertilized soils or in the presence of indigenous
N2-fixing bacteria.
Recently, Fujii et al. (1987) produced Klebsiella oxytoca NG13/pMC71A and Enterobacter
cloacae B26/pMC71A by inserting a nif A containing K. pneumoniae plasmid (pMC71A) into
K. oxytoca NG13 and E. cloacae E26, respectively. Inoculation effects of the mutants were
tested on two rice varieties, Oryza sativa L.
C5444 (Indica type) and T65 (Japonica type),
199
and growth and N increments were enhanced,
particularly in T65, as indicated in Table 2. It
was concluded that inoculation with genetically
improved bacteria might enhance N2 fixation
when associated with rice.
Incorporation of organic matter stimulates
BNF of heterotrophic N2-fixers in submerged
soil. Yoneyama et al. (1977) and Ladha et al.
(1986) indicated the effect of rice-straw amendment on BNF in submerged Philippine soils.
Incorporating crop residues or compost into
paddy fields improves fertility (Hashimoto, 1977;
Shiga, 1984; Veal and Lynch, 1984). In Japan,
compost was applied at a high rate to increase
the yield and the quality of rice until 1960. After
1960, N fertilizer use increased with a corresponding decrease in compost use. Sizeable
amounts of rice straw and hull have been burned
in fields. Recently, however, leaving rice residues in the field is increasingly practiced and is
reported to maintain soil fertility and improve
the soil ecosystem, and is an efficient use of
available resources (Hashimoto, 1977).
Shiga (1984) used a newly devised method to
examine the decomposition of organic materials
in paddy soil. In this method, a mixture of soil
Table 2. Dry weight, N content. 15N concentration, and acetylene reducing activity in rice plants grown in pots. Oryza sativa L.
vars. CS444 (Indica type) and T65 (Japonica type) at 100 days after inoculation with Klebsiella oxytoca NG13, Enterobacter
cloacae E26 and their mutants (Fujii et al., 1987)
Rice
variety
Bacteria
inoculated
Plant dry
weight
(g/plant)
Plant N
content
(mg/plant)
N in
plant
(atom %
excess)
Acetylene
reduction
(C2H4
formed
nmol/h/
plant)
C5444
NG1389 a
NG13
NG13/pMC71A
E26
E26/pMC71A
1.99ab
2.14a
2.23a
2.08a
2.16a
9.83a
10.55a
1.68a
10.30a
10.76a
4.662a
1.155b
3.226c
4.064b
3.633
154a
367a
483a
331a
805b
T65
NG1389
NG13
NG13/pMC71A
E26
E26/pMC71A
1.65a
1.88b
1.93bc
1.87b
2.00c
8.91a
9.95b
10.36b
10.32b
11.01b
1.708a
4.018b
3.722b
3.968b
3.745b
213a
545bc
462b
438b
722c
15
a NG1389, NG13/pMC71A and E26/pMC71A indicate a nif mutant from K. cloucae and mutants produced by inserting a nif A
containing K. pneurnoniae plasmid (pMC71A) into K. oxytoca NG13 and E. cloucae E26. respectively. Each data is the average
of ten samples.
b
Numbers followed by the same letters in each rice variety do not differ significantly at p = 0.05 using Duncan's multiple range
test.
200
Ishizuka
and organic material was transferred to a tube
made of glass filter fiber, then buried in the plow
layer of the field. It was then analyzed for
changes in the mixture's C and N content. The C
decreased gradually during incubation regardless
of the chemical composition of added materials.
The N content of materials with C:N ratio >30
increased until the ratio became 30. In the mixture of wheat straw and soil, N content was 2.2
times larger than the original content. The material with C : N ratio <30 did not show a similar
N increase. The N increment was considered to
be either caused by BNF or by immobilization of
diffused mineral N.
In an incubation study, Ono (1990) investigated effects of incorporated organic matter on
N2 fixation in submerged soils. He showed that
the effects of slightly soluble carbohydrates such
as starch and glycogen were larger than the
effects of soluble and low molecular weight substances such as glucose and organic acids, probably due to their providing a longer lasting energy source for BNF. He reported the extrapolated
value of N2 fixed per hectare to be 32-37 kg in a
paddy field when four tons of rice straw were
applied.
Hamada (1987) estimated BNF in a paddy
field by tracer technique, using irrigation water
saturated with a mixture of 15 N2 and O2 gases at
an 80: 20 ratio. A 23-cm-tall, stainless steel cylinder with a 20-cm diameter was inserted into
paddy soil, up to plow-sole depth, around each
rice plant; 5 cm of the cylinder remained above
the soil surface. An acrylic cover, equipped with
a frame for supporting rice plants and inlets for
supplying and sampling 15 N water, was fitted to
the cylinder. The rice plant was supported by the
frame and the chamber was made air-tight with
Apieson Q (Apieson Co. England). The cylinder
was filled with 15 N2 for 23 days from September
1. Immediately after the end of the 15 N feeding
(flowering), and again at harvest, 15 N distributions in the floodwater, the soil within the cylinder and the plant, were examined. The values in
Table 3 were converted from the original data
into kg ha -1 of N biologically fixed for 23 days.
They coincided with the values estimated from
acetylene-reducing activities (ARA) of soils adjacent to the cylinder, using 3.6 as the conversion
factor. When 4 t/ha of barley straw was incorpo-
Table 3. Distribution of biologically fixed 15 N in kg ha -1 at
flowering (Sept. 23) and harvest of rice a (Hamada, 1987)
Flowering
Harvest
Flood water
Upper plow layer
(0-1.0 cm)
0.00
0.70
(10.3)
0.79
(11.6)
Middle plow layer
(1.0-6.5)
2.08
(32.3)
2.00
(29.3)
1.79
(27.8)
2.03
(29.9)
1.34
(20.8)
1.32
(19.3)
0.53
6.44
(8.3)
(100)
0.67
6.81
(9.9)
(100)
Lower plow layer
(6.5-12)
Plow sole
(12-18)
Rice plant
Total
(0.0)
b
a
Irrigated with water saturated with 15 N-N2 for 23 days from
September 1-23. b Data in parentheses show percent of 15 N
content in each part to total fixed 15 N.
rated into paddy soil, BNF increased 25%. Application of N fertilizer stimulated BNF, although during early rice growth BNF was repressed. However, only 10% of the fixed N was
mineralized and absorbed by plants in one cropping season; the rest was deposited in the soil.
The major free-living N2 fixer in surface soil is
BGA. Ono (1990), using submerged paddy soil
in bottles, estimated BNF potential in surface
soil in Kyushu at 23-29kg N ha -1 per crop.
Roger and Watanabe (1986) reviewed BNF
potential of BGA and its use in rice cultivation.
The BNF potentials of BGA was estimated at
10-20kg N ha -1 based on a maximum standing
biomass in a rice field at blooming time, and
37-150 kg N ha -1 per crop based on carbon
input in the flood water and surface soil. Estimation by acetylene reduction assay frequently provides erroneous results, due to ARA variations
during the day and depending on the factor used
to convert ARA values to fixed N2 values. The
average value calculated from reported ARA
data was 27 kg N ha -1 per crop. Roger and
Watanabe (1986) concluded that BGA have less
potential for fixing N, than legumes or Azolla.
They concluded that the technological problems
governing inoculum quality and establishment
and the generally low yield increases caused by
algal inoculation limit their usage.
As reviewed by Roger and Watanabe (1986),
the research on BGA inoculation is conducted
intensively in India, where algal inoculum is
produced by open-air soil culture. The inoculum
Research trends in BNF
is multiplied in shallow trays or tanks with about
4 kg soil m-2 and 5-15 cm standing water. Triple
superphosphate is applied at 100 g m-2, and insecticide is used. After a 1-3 week culture, water
in the trays is allowed to evaporate. Algal flakes
are scraped off and used as inoculum.
Agronomic use of BNF in Azolla-Anabaena
symbiosis has been intensively studied in tropical
and subtropical countries. The nitrogen-fixing
activity of Azolla-Anabaena symbiosis is high
under optimum conditions. Watanabe et al.
(1980) obtained 22 harvests of Azolla with a
total amount of 465 kg N ha-1 per year. However, substantial N2-fixation by Azolla in rice
fields is restricted by environmental factors such
as pH, temperature, light intensity, insect damage, and low supply of mineral nutrients, especially phosphorus. Incorporation of Azolla into
soil increased grain yield 20% in field experiments at 12 sites in south Asian countries
(Watanabe, 1982).
Asakawa et al. (1988) grew Azolla for 38 days
in a flooded rice field before transplanting. In
1985 and 1986 they incorporated 60 and 63 t ha-1
of Azolla, equivalent to 60 and 110 kg N ha-1
into soils. With 60 t ha-1 Azolla, nitrogen uptake
of rice plants increased by 49% and 42 kg N ha-1
per cropping season. With 63 tons ha-1 of Azolla, uptake of plants increased by 42% and 33 kg
N ha-1 per cropping season. It was also important to get a large amount of healthy Azolla
inoculum under low mean temperatures (1720°C) and to flood the plots one month earlier
than the surrounding fields. Practically, AzollaAnabaena symbiosis as green manure has limited
use in intensive agriculture in the temperate
zone.
Tree ecosystems
Tree ecosystems depend completely on inputs of
natural N. The major inputs are from BNF by
legumes, actinorhizal plants, and free-living bacteria. Leguminous trees and shrubs contribute to
the nitrogen economy of tropical regions (Dawson, 1983), especially in the saline and salinealkali soils that are usually barren (Bala et al.,
1990; Felker et al., 1981).
Actinorhizal trees, especially alder, were wide-
201
ly used to produce constructional timber, furniture, firewood, and charcoal in the temperate
region. Casuarina, an actinorhizal plant, is important in tropical and subtropical countries, especially as a firewood source. Their symbiotic
nitrogen fixation helps to improve the nitrogen
fertility in forests. Furthermore, actinorhizal
plants have been used in rehabilitating mine
spoils and in stabilizing recent flood deposits and
landslide areas. This is because the plants grow
well in low-N soils where other plants are hard to
grow (Dawson, 1983).
Though precise estimation of BNF in tree
ecosystems is difficult, Bothe et al. (1983) estimated the gross contribution of Frankia with
nonleguminous trees in fields at 27-150kg N
ha-1 per year. However, this contribution is
seldom sufficient for optimum growth of trees
that do not have N2-fixing symbionts (Dawson,
1983). Application of N fertilizer may enhance
productivity in forests, but high fertilizer costs
restrict its use. Therefore, foresters are interested in improvement of a silvicultural system
using actinorhizal plants. Hansen and Dawson
(1982) examined the effect of interplanted Alnus
glutinosa on the growth of Populus. They found
that Populus growth, when interplanted with a
higher percentage of Alnus, was comparable to
its growth with treatments of optimal rates of
applied ammonium nitrate. They also found that
soil N accretion in the upper 4 cm of soil, adjacent to and at a distance of 15 cm from each
alder stem, was 222 and 158 mg kg-1 of air-dry
soil every 2 years, respectively.
Houwers and Akkermans (1981) showed that
inoculation with homogenates of A. glutinosa
nodules enhanced nodulation and growth of A.
glutinosa in soils with very low numbers of Frankia, in proportion to the amount of inoculum
applied. Actinorhizal plants are already wellnodulated in nature, however, because Frankia
is widely distributed in soils and can survive for
long periods (Smolander and Sundman, 1987).
To increase BNF and plant production with improved inoculation techniques, comparative
studies of the relative efficiency of different
Frankia strains with selection of efficient combinations of host plant and microsymbiont are
required.
Studies on the interaction between Frankia
202
Ohizuka
and Alnus have had mixed results. Simon et al.
(1985) reported that significant growth differences were observed when clones of A. glutinosa
were inoculated with four strains of Frankia.
Prat (1989) inoculated plants of eight Alnus
species with mixtures of pure strains isolated
from different Alnus species and could not detect
significant interaction between Frankia strains
and the Alnus species. However, some species
showed higher growth when they were inoculated with the strain isolated from the same
species. This may indicate interaction between
the Frankia strain and the host Alnus. A strong
indication of preferential association between
Frankia strains and Alnus species in Finland was
shown by Weber (1986) and Weber et al. (1987).
They isolated Frankia from A. incana and A.
glutinosa in different ecosystems and found that
the majority of strains from A. incana were
spore-positive and those from A. glutinosa were
spore-negative. The spore-positive strains from
A. incana were ineffective in A. glutinosa, while
the rare spore-positive strains from A. glutinosa
were effective in both species. On the other
hand, Houwers and Akkermans (1981) reported
that A. glutinosa in the Netherlands was nodulated sufficiently, with very small amounts of
crushed nodules of spore-positive strains compared with those of spore-negative strains. Distribution of spore-positive and spore-negative
isolates was related to host plant species and
their origins (Normand and Lalonde, 1982).
Prat (1989) showed that a mixture of Frankia
strains generally provided higher Alnus growth
than an individual strain. Hahn et al. (1990)
reported enhanced growth of Alnus using dual
inoculation with effective and ineffective Frankia
strains. The enhanced effect of the ineffective
strain was independent of the plant clone's susceptibility to the ineffective strain and of the
nodulation condition. It was suggested that a
positive interaction may occur between Frankia
strains. If a positive interaction between Frankia
strains can occur with mixed strains, it may also
happen between indigenous and introduced
strains in the field (Prat, 1989). Lack of methodology for the isolation and identification of
Frankia strains in natural systems has inhibited
clarification of the positive interaction between
strains (Hahn, 1990). Utilization of the inter-
action in silvicultural systems, however, should
improve plant productivity.
The contribution of free living N2-fixers to N
balance in forests was considered to be smaller
than that of symbiotic N2-fixers. N2-fixation rates
were 0.3-38.0kg ha -1 per year, with wide variation among the ecosystems (Granholl and Lindberg, 1978). Heterotrophic N2-fixation, mainly
due to facultative anaerobic bacteria, was found
in decomposing woody litter and plant rhizospheres. N2-fixing activity decreases with increasing soil depth, acidity, and O2 tension, but with
decreasing soil temperature and organic matter
content. Photo-dependent N2-fixation by cyanobacteria at the soil surface is also reported (reviewed by Dawson, 1983). Phyllospheric N fixation by free-living bacteria in forest trees has
been reported (Granholl and Lindberg, 1978).
Favilli and Messini (1990) estimated N2 fixation
in leaves of coniferous plants in a 3-year investigation where the highest nitrogenase activity,
8.1 nmole of N, fixed h -1 g -1 , was detected in
the spring.
Upland crop fields
Nitrogen-fixing organisms in upland fields fix N,
in the states of free-living, Rhizobium, or associative symbiosis. Bothe et al. (1983) estimated N fixation in various systems from published data. According to their estimate, N2fixation by rhizobium symbiosis was 40-460kg
ha -1 per year; by Actinomycetes, 27-150kg
ha -1 per year; by blue-green algae, 5-8kg ha -1
per year; by associative symbiosis, 3-10 kg ha -1
per year; and by free-living bacteria, less than
1 kg ha -1 per year as shown in Table 4. Therefore, BNF by free-living bacteria does not seem
to contribute to the N balance of upland
ecosystems.
Associative N2-fixing symbioses
The amount of N2 fixed in associative symbiosis
with non-leguminous plants is also low. However, Dobereiner (1983) reported 60-90 kg ha -1
per cropping season of N2 fixed by maize under
tropical conditions. Interrelationships between
host plants and bacteria in associative symbiosis
Research trends in BNF
203
Table 4. Estimates of the performance of nitrogen-fixing microorganisms in the field (Bothe et al., 1983)
Plant
Associative organism
Fixation rate
(kg ha -1 year) -1
Rhizobium symbiosis
Soybean
Lucerne
Bradyrhizobium japonicum
Rhizobium meliloti
40-200
50-460
Actinomycete symbiosis
Alder
Bog myrtle
Frankia
Frankia
50-150
27
Blue-green algal symbiosis
Azolla
Lichen
Anabaena azollae
Nostoc
50-80
5-10
Associative symbiosis
Grass
Sugar cane
Azotobacter paspali
Azospirillum, Beijerinckia
10
3
Free-living organism
Blue-green algae
Azotobacter, Clostridium, Klebsiella
were very specific. Jagnow (1990) found large
differences in N2 fixing activities between 14 cereal cultivars inoculated with a strain of Azospirillum, concluding that genotypic variability existed
between cereal cultivars. Hurek et al. (1988)
described a habitat-specific chemotaxis of Azospirillum spp. originating from different sites. C4
plants, such as maize, associate mainly with
Azospirillum lipoferum, while C, plants associate
mainly with A. brasilense (Dobereiner, 1983).
Inoculations of grasses and cereal crops with
Azospirillum have been investigated extensively.
Many reports showed beneficial effects on plant
growth and N2 fixation (Nur et al., 1980; Warenbourg et al., 1987). However, these effects were
influenced by the combination of host plant and
inoculant strain, and by environmental factors
such as soil fertility and moisture. Baltensperger
et al. (1978) inoculated eight genotypes of Bermuda grass with Azospirillum and Azobacter and
showed that the top growth increased by an
average 17% at low N fertility, but not at high N
fertility. Plausible association mechanisms between host plant and microsymbiont have been
reported as production of growth-regulating substances that promote the growth of roots and
root hairs, and as enhanced nutrient uptake
(Zimmer and Bothe, 1988). Tien et al. (1979)
found that A. brasilense produced indole acetic
80- 120
<1
acid (IAA), gibberellin, and cytokinin-like substances, and that combination of those substances produced morphological changes in pearl
millet roots similar to changes produced by A.
brasilense. Kapulnik et al. (1987) inoculated five
wheat cultivars with A. brasilense and showed
that inoculation increased grain yield in only one
of them, cv Mirian. However, they suggested
that the positive response was due to relief from
water stress at the end of the growth season.
Dobereiner et al. (1988) inoculated fieldgrown wheat with four strains of Azospirillum
spp labeled with antibiotic resistance and showed
that total N and grain yield were increased significantly by inoculation with A. brasilense sp 245
which accounted for 100% of isolates from the
washed roots. Ferreira et al. (1987) considered
that the effects of A. brasilense sp 245 inoculation were due more to efficient assimilation of
NO3 than to an increase in N2 fixation. Some
strains of Azospirillum can use nitrate to respire
and denitrify (Dobereiner, 1983); therefore, they
are also consumers of fixed nitrogen.
As stated above, the inoculation effects in
associative symbiosis vary with combinations of
host plant and microsymbiont. The association
mechanisms cause not only BNF but also regulation of plant growth due to a supply of growthregulating substances from the microsymbiont.
204
Ishizuka
Initially, strain selection is important for developing inoculation practices for associative
symbiosis (Baldani et al., 1987). However, even
when an effective combination is established, the
plants are less developed than those fertilized
with N, and Nur et al. (1980) pointed out that
exchanging Azospirillum for N fertilizer in modern agriculture remains a remote goal.
Legume-rhizobium
symbiosis
BNF by Rhizobium-legume symbiosis has been
widely used and intensively studied in agriculture. Investigation of the genetic traits related to
the BNF by hosts has been progressing, and
selections of cultivars with high N-fixing abilities
have succeeded with alfalfa (Duhigg et al.,
1978), red clover (Nutman and Riley, 1981),
crimson clover (Smith, 1982), and pea (Hobbs
and Mahon, 1982).
Nitrate tolerance is a useful trait for BNF
enhancement in N-rich soils. Nitrate tolerant
soybean mutants have been selected (Carroll et
al., 1985; Gremaud and Harper, 1989). Some of
the mutants nodulate superabundantly, compared with their parent cultivars, not only in the
presence of nitrate, but also in its absence. This
is because they were defective in a feedback
mechanism that regulated nodulation (CaetanoAnolles and Gresshoff, 1990). The mutants generally did not yield as well as the parent cultivars, probably due to additional mutations unrelated to supernodulation. Thus, this genetic
trait will be used in breeding programs designed
to improve the agronomic potential of soybean
(Carroll et al., 1988).
Microsymbiont
improvement
is
possible
through the selection of highly effective or efficient strains. Hydrogen evolution from nodules
was equivalent to 40-60% of energy loss and
results in reduction of N2-fixation efficiency
(Schubert and Evans, 1976). Hup + Rhizobium
strains have uptake-hydrogenase capable of preventing hydrogen evolution and reducing energy
losses. Because Hup' strains have efficient
nodule production (Evans et al., 1987), they
have been examined for use as an inoculant.
Evans et al. (1987) compared Hup + strain as an
inoculant with the performance of Hup - strain,
in a growth trial with soybean. They showed that
total N content of plants inoculated with Hup +
strain had increased 8.6% in the seeds, 27% in
the leaves, and 11% in the total plants after a
141 day-growth period.
Mutants of B. japonicum, which nodulates
earlier than its parent strain, were isolated
(Maier and Brill, 1978). Biological nitrogen fixation and growth of soybean nodulated with the
mutants were distinctly increased at the early
growth stage, but not thereafter. Hunter and
Kuykendall (1990) isolated a prototrophic revertant (TA11 NO + ) of a nodulation defective
tryptophan auxotroph of B. japonicum. They
showed that soybeans inoculated with the revertant and grown hydroponically for 35 days,
had 33% more dry weight and 52% more N than
the wild-type strain. These results demonstrate
that highly effective rhizobium strains can be
obtained by mutagenesis. However, many problems must be solved before beneficial effects can
be realized and maintained in the presence of
highly competitive indigenous strains (Evans et
al. , 1987).
Some strains of B. japonicum reduce acetylene
to ethylene in free-living culture. A mutant of B.
japonicum strain, USDA 110, with approximately 100% greater C2H2-reducing activity and deficient in nitrate reductase in free-living culture,
was selected (Williams and Phillips, 1983). The
plants inoculated with the mutant increased in
dry weight and in N content by about 40%
compared with plants inoculated with the parent
strain under controlled environmental conditions. In a field trial, seed yield was increased by
about 12%, and more than 80% of the nodules
of plants inoculated were occupied by the
mutant.
Recently, Maier and Graham (1990) obtained
promising mutant strains of B. japonicum, JH359
and JH310, after transposing Tn5 mutagenesis of
wild-type strain JH. The mutants required higher
levels of Mo than the parent strain to express
nitrate reductase in free-living culture, while
their nitrogenase activities were influenced little
by Mo supplementation. Soybeans inoculated
with strain JH359 showed higher BNF activity
and greater fresh weight compared with the parent strain inoculated plants.
Selection of a superior B. japonicum mutant is
possible in free-living culture. However, benefits
Research trends in BNF
in the field would happen only with low N availability and small populations of competitive indigenous rhizobia. Even though a highly efficient
association can be formed by selection or breeding of both symbionts in the laboratory, it does
not always result in high N2-fixation or high yield
in the field. This is mainly due to the restriction
of nodulation with inoculant strains by exclusive
invasion of less effective indigenous rhizobia. It
has been shown by SDS-polyacrylamide gel electrophoresis of proteins of nodule isolates, that
inoculant strains scarcely occupied nodules at
ordinary inoculation rates of 10 5 -10 6 cells/grain
on southeastern Wisconsin soybean farms
(Kamicker and Brill, 1986). In order to increase
BNF and yield of leguminous crops by inoculation with efficient rhizobia, the proportion of
nodules occupied by inoculant strains must be
sufficiently increased. Weaver and Frederick
(1974) predicted that, if the inoculation of a
useful strain is to be beneficial, the ratio of
nodules occupied by inoculated rhizobia must be
increased to more than 50%.
Heavy inoculation is a way of increasing
nodule formation by inoculated strains. Weaver
and Frederick (1974) indicated that if inoculated
rhizobia were to form 50% or more of the
nodules, an inoculum rate of at least 1000 times
the soil population must be used. Takahashi et
al. (1986) increased the ratios of nodules occupied by inoculated B. japonicum Hup + strain
to more than 60% by heavy inoculation at a rate
of 10 11 cells/seed.
The methods of inoculation affect the vertical
distribution of nodules occupied by inoculant
strains. Inoculum added to the seed furrow and
tilled into the soil produces nodules mainly in the
top and bottom sections of the soybean roots.
The nodules produced in the bottom part were
younger at seed formation and contributed to
seed N (Kamicker and Brill, 1987).
Nodule formation by inoculation also depends
on the inoculant carrier. Clay and peat powder
have been used as carriers for maintenance of
infectivity and simplification of inoculation practice. Kremer and Peterson (1983) showed that a
suspension of lyophilized rhizobia cells in
peanut-oil was more effective in comparison with
a peat-base inoculant. Recently, seeds inoculated
by vacuum-infiltration in the factory, 'Noculid
205
seeds', have been used by Japanese farmers.
Noculid seeds maintain infectivity for four weeks
under low temperature and optimum humidity,
so farmers are spared the complicated labor for
inoculation and maintenance of the inoculant
(Takahashi, personal communication).
Heavy inoculation is neither realistic, nor
economical in agricultural practice. Thus, the
reasons why such heavy inoculation is needed for
effective nodulation must be found. TO do this,
the following must be initially investigated: behavior of inoculated rhizobia in soils, mechanisms of host recognition by microsymbionts,
preference of host plant for microsymbionts, infection and nodule formation, and regulation of
BNF.
The first event in infection of legumes by
rhizobia is the release of a flavonoid compound,
by host roots, into the rhizosphere (Rolfe and
Gresshoff, 1988). Rhizobia then interact with the
compound and produce factors that stimulate
cell division. The cortex cells of the roots start to
divide. and form pseudo-infection sites. The next
steps are: formation of the primary nodule
meristem. contact or binding of rhizobia with the
root hair emergence sites. and invasion of
rhizobia into the root hair. The binding of
rhizobia with root hairs is thought to be
mediated by a specific protein, lectin (Dazzo and
Hubbell, 1975). The infected root hair curls and
forms infection threads through which bacteria
are transported to the cortex to proliferate.
When any one of the successive reactions is
blocked, nodules are not effectively formed. For
instance, the binding of rhizobia with root hairs
through mediation of lectin is considered essential for mutual recognition between host and
appropriate rhizobia.
In soybean, the binding ability of the rhizobia1
cell with lectin decreased rapidly with culture age
and was completely lost at the late logarithmic
phase (Mort and Bauer, 1980). The development
of specific lectin receptors on the cell surface of
B. japonicum was conditioned by the environment provided by the host plant roots. It was not
detected in some strains of B. japonicum cultured in a synthetic salt medium (Bhuvaneswari
and Bauer, 1978). There is sufficient justification
for improvement of nodule formation by inoculum preparation instead of extremely heavy
206
Ishizuka
inoculation, because the lectin-binding receptors
on B. japonicum are transient (Bhuvaneswari et
al., 1977; Mort and Bauer, 1980) and their appearance might be dependent on the bacterial
growth environment (Bhuvaneswari and Bauer,
1978).
Another strategy to increase nodule formation
by the inoculant strain, is to exclude nodulation
by less efficient indigenous rhizobium strains.
For this, physiological and genetical mechanisms, that determine host plant preference for
rhizobial strain or compatibility of rhizobia with
the legume cultivar, should be investigated. Ellis
et al. (1984) investigated serotypes of isolates
from soybeans grown in the midwestern USA,
showing that 95% of the nodules were occupied
by rhizobia belonging to serogroup USDA 123.
According to Kvien et al. (1981), rhizobia of
serogroup USDA 123 were less efficient.
Cregan and Keyser (1986) selected 13 of 1278
soybean genotypes that barely nodulated with
USDA 123. Two of the genotypes, PI 371607
and PI 377578, and cv. Williams were inoculated
with mixed inoculum of USDA 123 and either
USDA 110, 122, or 138. They were analyzed
serologically for bacteroids inhabited in nodules.
More than 75% of the nodules of cv. Williams
were occupied by USDA 123, whereas in all
other cases less than 10% of nodules of the PI
genotype were occupied by USDA 123. Table 5
shows part of the results. Other genotypes of
soybean, Rj2, Rj3, and Rj4, that restrict nodulation with appropriate strains also have been reported (Caldwell, 1966; Caldwell et al., 1966;
Vest, 1970; Vest and Caldwell, 1972).
We isolated rhizobia from various Rjgenotypes of soybeans grown in the field and
examined the compatibilities of isolates with
each soybean Rj-genotype (Ishizuka et al.,
Table 5. Strains recovered from the nodules of PI371607,
PI377578 and Williams soybeans in nodulation competition
experiment (Cregan and Keyser, 1986)
Soybean
genotype
Recovery from nodules (%)
USDA 123
USDA 110
PI 371607
PI 377578
Williams
0
0
83
100
98
25
Other
2
6
2
Table 6. Preferences of Rj-soybean cultivars, grown in the
field, for indigenous Bradyrhizobium japonicum. The isolates
from nodules were classified into 3 types by compatibilities
with Rj-cultivars (Ishizuka et al., 1991)
Type of isolate
Percentage of each type
of isolate
Genotype of cultivar from
which rhizobia were isolated
Compatible with Rj2- and
Rj4-cv
Incompatible with Rj2-cv.
and compatible with Rj4-cv.
Compatible with Rj2-cv.
and incompatible with Rj4-cv.
Non-Rj
Rj2
Rj4
75
38
39
13
0
61
13
62
0
1991). As shown in Table 6, Rj2- and Rj4genotypes of soybeans preferred incompatible
rhizobia, with Rj, and Rj, genotypes of soybeans, respectively. Such preferences were not
observed in the cultivars that do not have the Rj
genes. The Rj-genotypes of soybeans not only
restricted nodulation with the appropriate
rhizobia, but preferred the rhizobia harboring a
specific trait. Devine et al. (1990) tested for the
nodulation response of 119 isolates to the Rj,
allele. They showed that a majority of isolates
interdicted by the Rj, allele were chlorosisinducing bacteria and suggested that the Rj,
allele benefitted the host plant by shielding it
from nodulation by certain bradyrhizobial groups
with impaired efficiency of N2 fixation with
soybean.
The evidence indicates that some leguminous
crop cultivars have nodulating-conditioning
genes, restrict nodulation with certain rhizobia,
and prefer other rhizobia. The establishment of
efficient combinations of genotypes of host plant
and inoculant rhizobial strains, excluding nodulation with less efficient indigenous rhizobia may
enhance BNF. Plant breeders may be able to
construct such systems for agronomic use (Devine et al., 1990).
Conclusions
Genetic improvement of BNF systems and their
management for agronomic use have been inves-
Research trends in BNF
tigated intensively. Examples are breeding and
selection of high-yielding cultivars (Duhigg et
al., 1978; Hobbs and Mahon, 1982; Nutman and
Riley, 1981; Smith et al., 1982) and isolation of
efficient N2-fixing bacteria (Maier and Brill,
1978; Maier and Graham, 1990). These studies
have contributed to the elucidation of mechanisms of symbiosis and N2-fixation (Rolfe and
Gresshoff, 1988).
On the other hand, many problems of enhancement of BNF and productivity in agriculture remain unsolved. Recently, soybean mutants that nodulated abundantly in the presence
of nitrate were isolated, but their yields were not
high compared with current cultivars (Carrolle et
al., 1988). Thus, abundant nodule formation
does not necessarily result in high N2 fixation and
increased yield. Various promising N2-fixing bacteria selected and produced by mutagenesis have
proven to have beneficial effects on plant production under aseptic conditions or in the absence of competitive indigenous N2-fixers but,
unequivocal effect in field trails has not been
found.
Therefore, in addition to analyzing individual
functions of both symbionts related to BNF, one
must investigate the overall BNF system and
factors affecting it. Then one must clarify the
restricting factors of symbiotic association and
productivity. The most important targets of
studies on legume-rhizobia symbiosis are to increase occupancy of inoculant rhizobia and to
improve photosynthetic plant characteristics as
an energy supply for BNF.
BNF by free-living N2-fixers and associative
symbiosis can not yet replace N fertilizers in
modern agriculture (Nur et al., 1980). However,
there is potential for increasing productivity
through breeding of N2-fixing bacteria (Fujii et
al., 1987) and through improvement of combinations with crops, inoculation methods, and cultural practices.
Biotechnology is useful for the creation of
promising N2-fixing organisms. An example is
the transfer of the bacterial nif gene into a
non-legume. However, its feasibility is not well
defined, although it is not impossible. Even
though prokariotic genes can be transferred to
eukariotic cells, gene expression will be difficult,
because organisms capable of N2-fixation are
207
only procaryots. Other genes concerning BNF,
such as bacterial infection, nodule formation,
and regulation of BNF activity, must be simultaneously transferred. Furthermore, there are
many problems that should be solved, because
biological N2-fixation depends on many interrelated factors. When these problems are solved in
the future, the most important problem will be
what kind of organism should be created to
enhance practical use of N2-fixation in agriculture and how to use that organism.
References
Alimagno B V and Yoshida T 1977 In situ acetylene-ethylene
assay of biological nitrogen fixation in lowland rice soils.
Plant and Soil 47. 239-244.
Asakawa S, Araragi M, Hayano K and Takabayashi M 1988
Effects of Azolla application on rice yield. amount of
nitrogen uptake and soil nitrogen. Jap. J. Soil Sci. Plant
Nutr. 59, 397-402( In Japanese ).
Bala N, Sharma P K and Lakshminarayana K 1990 Nodulation and nitrogen fixation by salinity-tolerant rhizobia in
symbiosis with tree legumes. Agric. Ecosys. Environ. 33,
33-46.
Baldani V L D. Baldani J J and Dobereiner J 1987 Inoculation of field-grown wheat ( Triticum aestivum ) with Azospirillum spp. in Brasil. Biol. Fert. Soils 4. 37-40.
Baltensperger A A, Schank S C, Smith R L, Littell R C,
Bouton J H and Dudeck A C 1978 Effect of inoculation
with Azospirillum and Azotobacter on turf-type Bermuda
genotypes. Crop Sci. 18. 1034-1045.
Bhuvaneswari T V, Pueppke S G and Bauer W D 1977 Role
of lectins in plant-microorganism interactions. I. Binding of
soybean lectin to rhizobia. Plant Physiol. 60. 486-491.
Bhuvaneswari T V and Bauer W D 1978 Role of lectins in
plant-microorganism interactions. III. Influence of rhizosphere/rhizoplane culture condition on the soybean lectinbinding properties of rhizobia. Plant Physiol. 62, 71-74.
Bothe H, Yates M G and Cannon F C 1983 The nitrogenfixing organism and the nitrogenase reaction. In Encyclop.
Plant Physiol. New Series Vol. 15A. Eds. A Lauchli and R
L Bieleski. pp 241-285. Springer-Verlag, Berlin.
Caetano-Anolles G and Gresshoff P M 1990 Early induction
of feedback regulatory responses governing nodulation in
soybean. Plant Sci. 71, 69-81.
Caldwell B E 1966 Inheritance of a strain-specific ineffective
nodulation in soybean. Crop Sci. 6, 427-428.
Caldwell B E, Hinson K and Johnson H W 1966 A strainspecific ineffective nodualtion reaction in the soybean
Glycine max L. Merrill. Crop Sci. 6, 495-496.
Carroll B J, McNeil D L and Gresshoff P M 1985 A
supernodualtion and nitrate-tolerant symbiotic (nts) soybean mutant. Plant Physiol. 78, 34-40.
Carroll B J, Haltley R, Ashley D A, Boerma H R, Lawn R,
Byth D and Gresshoff P M 1988 Field evaluation of
208
Ishizuka
supernodulating soybean mutant. In Nitrogen Fixation:
Hundred Years After. Eds H Bothe, F J de Bruyn and W
E Newton. Gustav Fischer, Stuttgart, New York. p 808.
Cregan P B and Keyser H H 1986 Host restriction of
nodulation by Bradyrhizobium japonicum strain USDA
123 in soybean. Crop Sci. 26, 911-916.
Dawson J 0 1983 Dinitrogen fixation in forest ecosystems.
Can. J. Microbiol. 29, 979-992.
Dazzo F B and Hubbell D H 1975 Cross reactive antigens
and lectin as determinants of symbiotic specificity in the
Rhizobium-clover association. Appl. Microbiol. 30, 10171033.
Devine T E, Kuykendall L D and O’Neill J J 1990 The Rj4
allele in soybean represses nodulation by chlorosis-inducing bradyrhizobia classified as DNA homology group I1 by
antibiotic resistance profiles. Theor. Appl. Genet. 80, 3337.
Dewan G I and Rao N S S 1979 Seed inoculation with
Azospirillum brasilense and Azotobacter chroococcum and
root biomass of rice (Oryza sativa L.). Plant and Soil 53,
295-302.
Döbereiner J 1983 Dinitrogen fixation in rhizosphere and
phyllosphere association. In Encyclop. Plant Physiol. New
Series, Vol. 15A. Eds. A Lauchli and R L Bielski. pp
332-350. Springer-Verlag, Berlin.
Dobereiner J, Reis V M and Lazarini A C 1988 New N2
fixing bacteria in association with cereals and sugar cane.
In Nitrogen Fixation: Hundred Years After. Eds. H.
Bothe, F J de Bruyn and W E Newton pp 717-722. Gustav
Fischer, Stuttgart, New York.
Duhigg P, Melton B and Baltensperger A 1978 Selection for
acetylene reduction rate in Misilla alfalfa. Crop Sci. 18,
813-816.
Ellis W R, Ham G E and Schmidt E L 1984 Persistence and
recovery of Rhizobium japonicum inoculum in a field soil.
Agron. J. 16, 573-576.
Evans H J, Harker A R, Papen H, Russell S A, Hanus F J
and Zuber M 1987 Physiology, biochemistry, and genetics
of the uptake hydrogenase in Rhizobia. Annu. Rev. Microbiol. 41, 335-361.
Favilli F and Messini A 1990 Nitrogen fixation at phyllospheric level in coniferous plants in Italy. Plant and Soil
128, 0195.
Felker P, Clark P R, Laag A E and Pratt P F 1981 Salinity
tolerance of the tree legumes: Mesquite (Prosopis glandulosa var. Torreyana. P. velutina and P. articulata), algarrobo (P. chilensis), kiawe ( P. pallida ) and tamarugo ( P.
tamarugo) grown in sand culture on nitrogen-free media.
Plant and Soil 61, 311-317.
Ferreira M C B, Fernandes M S and Dobereiner J 1987 Role
of Azospirillum brasilense nitrate reductase in nitrate assimilation by wheat plants. Biol. Fertil. Soils 4, 47-53.
Fujii T, Huang Yi-De, Higashitani A, Nishimura Y, Iyama S,
Hirota Y, Yoneyama T and Dixon R A 1987 Effect of
inoculation with Klebsiella oxytoca and Enterobacter
cloacae on dinitrogen fixation by rice-bacteria associations.
Plant and Soil 103, 221-226.
Granholl U and Lindberg T 1978 Nitrogen fixation in some
coniferous forest ecosystems. Ecol. Bull. 26, 178-192.
Gremaud M F and Harper J E 1989 Selection and initial
characterization of partially nitrate tolerant nodulation mutants of soybean. Plant Physiol. 89, 169-173.
Hamada E 1987 Studies on biological nitrogen fixation and
movement of fixed nitrogen in paddy field. Ph.D. Dissertation. Kyushu University, Japan ( In Japanese ).
Hahn D, Starrenburg M G C and Akkermans A D L 1990
Growth increment of Alnus glutinosa upon dual inoculation with effective and ineffective Frankia strains. Plant
and Soil 122, 121-127.
Hansen E A and Dawson J 0 1982 Effect of Alus glutinosa
on hybrid Populus height grown in a short-rotation intensively cultured plantation. Forest Sci. 28, 49-59.
Hashimoto H 1977 Theory and Practice of Organic Matter
Application in Agriculture. Noubunkyou, Tokyo. 13 p.
Hobbs S L A and Mahon J D 1982 Effects of pea ( Pisum
sutivum ) genotypes and Rhizobium leguminosarum strains
on N2 (C2H2) fixation and growth. Can. J. Bot. 60,
2594-2600.
Houwers A and Akkermans A D L 1981 Influence of inoculation on yield of Alnus glutinosa in the Netherlands. Plant
and Soil 61, 189-202.
Hunter W J and Kuykendall L D 1990 Enhanced nodulation
and nitrogen fixation by a revertant of a nodulationdefective Bradyrhizobium japonicum tryptophan auxotroph. Appl. Environ. Microbiol. 56, 2399-2403.
Hurek T, Reinhold B and Niemann E G 1988 Habitatspecific chemotaxis of Azospirillum spp. In Nitrogen Fixation: Hundred Years After. Eds. H. Bothe, F J de Bruyn
and WE Newton. Gustav Fischer, Stuttgart, New York. p
773.
Ishizuka J, Suemasu Y and Mizogami K 1991 Preference of
Rj-soybean cultivars for Bradyrhizobium japonicum for
nodulation. Soil Sci. Plant Nutr. 37, 15-21.
Jagnow G 1990 Differences between cereal crop cultivars in
root-associated nitrogen fixation, possible causes of variable yield response to seed inoculation. Plant and Soil 123,
255-259.
Kamicker B J and Brill W J 1986 Identification of Bradyrhizobium japonicum nodule isolates from Wisconsin soybean farms. Appl Environ. Microbiol. 51, 487-492.
Kamicker B J and Brill W J 1987 Methods to alter the
recovery and nodule location of Bradyrhizobium
japonicum inoculant strains on field-grown soybeans. Appl.
Environ. Microbiol. 54, 1787-1742.
Kapulnik Y. Okon Y and Henis Y 1987 Yield response of
spring wheat cultivars ( Triticum aestivum and T. turgidum )
to inoculation with Azospirillum brasilense under field
conditions. Bid Fert. Soils 4, 27-35.
Konishi C and Seino K 1961 Studies on the maintenance
mechanism of paddy soil fertility in nature. Bull. Hokuriku
Agr. Exp. Stn. 2, 41-136 (In Japanese ).
Kremer R J and Peterson H L 1983 Field evaluation of
selected Rhizobium in an improved legume inoculant.
Agron. J. 75, 139-143.
Kvien C S, Ham G E and Lambert J W 1981 Recovery of
Rhizobium japonicum strain by soybean genotypes. Agron.
J . 73, 900-905.
Ladha J K, Tirol A C, Daroy M L G, Caldo G, Ventura W
and Watanabe I 1986 Plant-associated N2 fixation (C2H2reduction) by five rice varieties, and relationship with plant
Research trends in BNF
growth characters as affected by straw incorporation. Soil
Sci. Plant Nutr. 32, 91-106.
Maier R J and Brill W J 1978 Mutant strains of Rhizobium
japonicum with increased ability to fix nitrogen for soybean. Science 201, 448-450.
Maier R J and Graham L 1990 Mutrant strain of Bradyrhizobium japonicum with increased symbiotic N2-fixation
rates and altered Mo metabolism properties. Appl. Environ. Microbiol. 56, 2341-2346.
Mort A J and Bauer W D 1980 Composition of the capsular
and extracellular polysaccharides of Rhizobium japonicum.
Plant Physiol. 66, 158-163.
Normand P and Lalonde M 1982 Evaluation of Frankia
strains isolated from provenances of two Alnus species.
Can. J. Microbiol. 28, 1133-1142.
Nur I, Okon Y and Henis Y 1980 An increase in nitrogen
content of Setaria italica and Zea may inoculated with
Azospirillum. Can. J. Microbiol. 26, 482-485.
Nutman P S and Riley J 1981 Breeding of nodulated red
clover ( Trifolium pratense ) for high yield. Ann. Appl.
Biol. 98, 319-331.
Ono S 1990 Studies on the mechanism of sustainability of soil
nitrogen in wetland rice field. Ph.D. Diss. Kyushu University, Japan (in Japanese).
Prat D 1989 Effects of some pure and mixed Frankia strains
on seedling growth in different Alnus species. Plant and
Soil 113. 31-38.
Rao V R, Nayok DN, Charyulu P B B N and Adhya T K.
1983 Yield response of rice to root inoculation with Azospirillum. J. Agric. Sci. UK, 100, 689-691.
Roger P A and Watanabe I 1986 Technologies for utilizing
biological nitrogen fixation in wetland rice: Potentialities.
current usage, and limiting factors. Fert. Res. 9, 39-77.
Rolfe B G and Gresshoff P M 1988 Genetic analysis of
legume nodule infection. Annu. Rev. Plant Physiol. Plant
Mol. Biol. 39, 297-319.
Schubert K R and Evans H J 1976 Hydrogen evolution: A
major factor affecting the efficiency of nitrogen fixation in
nodulated symbionts. Proc. Nat. Acad. Sci. USA 73,
1207-1211.
Shiga H 1984 Criteria for application of organic materials in
paddy soil. Jap. J. Soil Sci. Plant Nutr. 55, 374-380 (In
Japanese).
Simon L, Stein A and Lalonde 1985 Performance of in vitro
propagated Alnus glutinosa (L.) Gaertn. clones inoculated
with Frankia. Plant and Soil 87, 125-133.
Smith G R, Knight W E and Peterson H L 1982 The
inheritance of N2 fixation efficiency in crimson clover.
Crop Sci. 22, 1091-1094.
Smolander A and Sundman V 1987 Frankia in acid soils of
forests devoid of actinorhizal plants. Physiol. Plant. 70.
297-303.
209
Takahashi T, Ito A and Suzuki H 1986 Interaction between
effective bacteria and host plant. In Studies on Nodule
Formation and Nitrogen Fixation in Legume Crops. Ed. J.
Ishizuka, pp 92-107. Ministry of Agriculture, Forestry and
Fisheries, Tokyo, Japan ( In Japanese ).
Tien T M, Gaskins M H and Hubbell D H 1979 Plant growth
substances produced by Azospirillum brasilense and their
effect on the growth of pearl millet ( Pennisetum
americanum L.). Appl. Environ. Microbiol. 37, 10161024.
Veal D A and Lynch J M 1984 Associative cellolysis and
dinitrogen fixation by co-cultures of Trichoderma harzianum and Clostridium butyricum. Nature 310, 659-697.
Vest G 1970 Rj3-A gene conditioning ineffective nodulation
in soybean. Crop Sci. 10, 34-35.
Vest G and Caldwell B E 1972 Rj4-A gene conditioning
ineffective nodulation in soybean. Crop Sci. 12, 692-693.
Warembourg E R, Dreeseen R, Vlassak K and Lafont F 1987
Peculiar effect of Azospirillum inoculation on growth and
nitrogen balance of winter wheat ( Triticum aestivum ).
Plant and Soil 4, 55-59.
Watanabe I, Berja N S and Rosario D C D 1980 Growth of
Azolla in paddy fields as affected by phosphorus fertilizer.
Soil Sci. Plant Nutr. 26, 301-307.
Watanabe I 1982 Azolla-Anabaena symbiosis - Its physiology
and use in tropical agriculture. In Microbiology of Tropical
Soils and Plant Productivity. Eds. Y R Dommergues and H
G Diem. pp 169-186. Martinus Nijhoff/Dr. W Junk Publishers, The Hague.
Watanabe I and Lin C 1984 Response of wetland rice to
inoculation with Azospirillum lipoferum and Pseudomonas
sp. Soil Sci. Plant Nutr. 30, 117-124.
Weaver R W and Frederick L R 1974 Effect of inoculum rate
on competitive nodulation of Glycine max L. Merrill. II.
Field studies. Agron. J. 66. 233-236.
Weber A 1986 Distribution of spore-positive and sporenegative nodule in stands of Alnus glutinosa and Alnus
incana in Finland. Plant and Soil 96, 205-213.
Weber A, Nurminaho-Lassila E L and Sundman V 1987
Features of the intrageneric Alnus-Frankia specificity. Physiol. Plant. 70, 289-296.
Williams L E and Phillips D A 1983 Increased soybean
productivity with a Rhizobium japonicum mutant. Crop
Sci. 23, 246-250.
Yoneyama T, Lee KK and Yoshida T 1977 Decomposition of
rice residues in tropical soils. IV. The effect of rice straw
on nitrogen fixation by heterotrophic bacteria in some
Philippine soils. Soil Sci. Plant Nutr. 23, 287-205.
Zimmer W and Bothe H 1988 The phytohormonal interactions between Azospirillum and wheat. Plant and Soil 110,
239-247.
Developments in Plant and Soil Sciences
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
J. Monteith and C. Webb (eds.): Soil Water and Nitrogen in Mediterranean-type Environments. 1981
ISBN 90-247-2406-6
J. C. Brogan (ed.): Nitrogen Losses and Surface Run-off from Landspreading of Manures. 1981
ISBN 90-247-2471-6
J. D. Bewley (ed.): Nitrogen and Carbon Metabolism. 1981
ISBN 90-247-2472-4
R. Brouwer, I. Gašparíková, J. Kolek and B. C. Loughman (eds.): Structure and Function of Plant Roots.
1981
ISBN 90-247-25 10-0
Y. R. Dommergues and H. G. Diem (eds.): Microbiology of Tropical Soils and Plant Productivity. 1982
ISBN 90-247-2624-7
G. P. Robertson, R. Herrara and T. Rosswall (eds.): Nitrogen Cycling in Ecosystems of Latin America and
the Caribbean. 1982
ISBN 90-247-27 19-7
D. Atkinson, K. K. S. Bhat, M. P. Coutts, P. A. Mason and D. J. Read (eds.): Tree Root Systems and Their
Mycorrhizas. 1983
ISBN 90-247-2821-5
M. R. Sari and B. C. Loughman (eds.): Genetic Aspects of Plant Nutrition. 1983
ISBN 90-247-2822-3
J. R. Freney and J. R. Simpson (eds.): Gaseous Loss of Nitrogen from Plant-Soil Systems. 1983
ISBN 90-247-2820-7
United Nations Economic Commission for Europe (ed.): Efficient Use of Fertilizers in Agriculture. 1983
ISBN 90-247-2866-5
J. Tinsley and J. F. Darbyshire (eds.): Biological Processes and Soil Fertility. 1984
ISBN 90-247-2902-5
A. D. L. Akkermans, D. Baker, K. Huss-Danell and J. D. Tjepkema (eds.): Frankia Symbioses. 1984
ISBN 90-247-2967-X
W. S. Silver and E. C. Schroder (eds.): Practical Application of Azolla for Rice Production. 1984
ISBN 90-247-3068-6
P. G. L. Vlek (ed.): Micronutrients in Tropical Food Crop Production. 1985
ISBN 90-247-3085-6
T. P. Hignett (ed.): Fertilizer Manual. 1985
ISBN 90-247-3 122-4
D. Vaughan and R. E. Malcolm (eds.): Soil Organic Matter and Biological Activity. 1985
ISBN 90-247-3154-2
D. Pastemak and A. San Pietro (eds.): Biosalinity in Action. Bioproduction with Saline Water. 1985
ISBN 90-247-3 159-3
M. Lalonde, C. Camiré and J. O. Dawson (eds.): Frankia and Actinorhizal Plants. 1985
ISBN 90-247-3214-X
H. Lambers, J. J. Neeteson and I. Stulen (4s.): Fundamental, Ecological and Agricultural Aspects of
Nitrogen Metabolism in Higher Plants. 1986
ISBN 90-247-3258-1
M. B. Jackson (ed.): New Root Formation in Plants and Cuttings. 1986
ISBN 90-247-3260-3
F. A. Skinner and P. Uomala (eds.): Nitrogen Fixation with Non-Legumes (Proceedings of the 3rd Symposium, Helsinki, 1984). 1986
ISBN 90-247-3283-2
A. Alexander (ed.): Foliar Fertilization. 1986
ISBN 90-247-3288-3
H. G. v.d. Meer, J. C. Ryden and G. C. Ennik (eds.): Nitrogen Fluxes in Intensive Grassland Systems. 1986
ISBN 90-247-3309-X
A. U. Mokwunye and P. L. G. Vlek (eds.): Management of Nitrogen and Phosphorus Fertilizers in SubSaharan Africa. 1986
ISBN 90-247-33 12-X
Y. Chen and Y. Avnimelech (eds.): The Role of Organic Matter in Modern Agriculture. 1986
ISBN 90-247-3360-X
S. K. De Datta and W. H. Patrick Jr. (eds.): Nitrogen Economy of Flooded Rice Soils. 1986
ISBN 90-247-336 1-8
W. H. Gabelman and B. C. Loughman (eds.): Genetic Aspects of Plant Mineral Nutrition. 1987
ISBN 90-247-3494-0
A. van Diest (ed.): Plant and Soil: Interfaces and Interactions. 1987
ISBN 90-247-3535-1
Developments in Plant and Soil Sciences
29. United Nations Economic Commission for Europe and FAO (eds.): The Utilization of Secondary and Trace
Elements in Agriculture. 1987
ISBN 90-247-3546-7
30. H. G. v.d. Meer, R. J. Unwin, T. A. van Dijk and G. C. Ennik (eds.): Animal Manure on Grassland and
Fodder Crops. Fertilizer or Waste? 1987
ISBN 90-247-3568-8
31. N. J. Barrow: Reactions with Variable-Charge Soils. 1987
ISBN 90-247-3589-0
32. D. P. Beck and L. A. Materon (eds.): Nitrogen Fixation by Legumes in Mediterranean Agriculture. 1988
ISBN 90-247-3624-2
33. R. D. Graham, R. J. Hannam and N. C. Uren (eds.): Manganese in Soils and Plants. 1988
ISBN 90-247-3758-3
34. J. G. Torrey and J. L. Winship (eds.): Applications of Continuous and Steady-State Methods to Root Biology.
1989
ISBN 0-7923-0024-6
35. F. A. Skinner, R. M. Boddey and I. Fendrik (eds.): Nitrogen Fixation with Nan-Legumes (Proceedings of the
4th Symposium, Rio de Janeiro, 1987). 1989
ISBN 0-7923-0059-9
36. B. C. Loughman, O. Gasparikova and J. Kolek (eds.): Structural and Functional Aspects of Transport in
Roots. 1989
ISBN 0-7923-0060-2; Pb 0-7923-0061-0
37. P. Plancquaert and R. Haggar (eds.): Legumes in Farming Systems. 1990
ISBN 0-7923-0134-X
38. A. E. Osman, M. M. Ibrahim and M. A. Jones (eds.): The Role of Legumes in the Farming Systems of the
Mediterranean Areas. 1990
ISBN 0-7923-0419-5
39. M. Clarholm and L. Bergstrom (eds.): Ecology of Arable Land - Perspectives and Challenges. 1989
ISBN 0-7923-0424-1
40. J. Vos, C. D. van Loon and G. J. Bollen (eds.): Effects of Crop Rotation on Potato Production in the
Temperate Zones. 1989
ISBN 0-7923-0495-0
ISBN 0-7923-0740-2
41. M. L. van Beusichem (ed.): Plant Nutrition - Physiology and Applications. 1990
42. N. El Bassam, M. Dambroth and B.C. Loughman (eds.): Generic Aspects of Plant Mineral Nutrition. 1990
ISBN 0-7923-0785-2
43. Y. Chen and Y. Hadar (eds.): Iron Nutrition and Interactions in Plants. 1991
ISBN 0-7923-1095-0
44. J. J. R. Groot, P. de Willigen and E. L. J. Verbeme (eds.): Nitrogen Turnover in the Soil-Crop System. 199 1
ISBN 0-7923-1 107-8
45. R. J. Wright, V.C. Baligar and R. P. Munmann (eds.): Plant-Soil Interactions at Low pH. 1991
ISBN 0-7923-1 105-1
46. J. Kolek and V. Kozinka (eds.): Physiology of the Plant Root System. 1992
ISBN 0-7923-1205-8
47. A. U. Mokwunye (ed.): Alleviating Soil Fertility Constraints to Increased Crop Production in West Africa.
1991
ISBN 0-7923-1221-X; Pb 0-7923-1222-8
48. M. Polsinelli, R. Materassi and M. Vincenzini (eds.): Nitrogen Fixation (Proceedings of the 5th Symposium,
Florence, 1990). 1991
ISBN 0-7923- 14 10-7
49. J. K. Ladha, T. George and B. B. Bohlool (eds.): Biological Nitrogen Fixation for Sustainable Agriculture.
1992
ISBN 0-7923-1774-2
Kluwer Academic Publishers - Dordrecht / Boston / London