BIOFERTILISERS IN ACTION

A report for the Rural Industries Research and Development Corporation Edited by Ivan R Kennedy and Abu T M A Choudhury

July 2002 RIRDC Publication No 02/086 RIRDC Project No WS990-23

ii

© 2002 Rural Industries Research and Development Corporation. All rights reserved. ISBN 0 642 58485 0 ISSN 1440-6845 BIOFERTILISERS IN ACTION Publication No. 02/086 Project No. WS990-23 The views expressed and the conclusions reached in this publication are those of the author and not necessarily those of persons consulted. RIRDC shall not be responsible in any way whatsoever to any person who relies in whole or in part on the contents of this report. This publication is copyright. However, RIRDC encourages wide dissemination of its research, providing the Corporation is clearly acknowledged. For any other enquiries concerning reproduction, contact the Publications Manager on phone 02 6272 3186. In submitting this report, the researcher has agreed to RIRDC publishing this material in its edited form. Editors’ Contact Details Ivan R Kennedy Director, SUNFix Centre for Nitrogen Fixation Ross Street Building A03 The University of Sydney, NSW 2006, Australia Phone: 61-2-9351 3546 Fax: 61-2-9351 5108 Email:i.kennedy@acss.usyd.edu.au

Abu T M A Choudhury SUNFix Centre for Nitrogen Fixation Ross Street Building A03 The University of Sydney, NSW 2006, Australia Phone: 61-2-9351 2379 Fax: 61-2-9351 5108 Email:a.choudhury@acss.usyd.edu.au

RIRDC Contact Details Rural Industries Research and Development Corporation Level 1, AMA House 42 Macquarie Street BARTON ACT 2600 PO Box 4776 KINGSTON ACT 2604 Phone: 02 6272 4539 Fax: 02 6272 5877 Email: rirdc@rirdc.gov.au. Website: http://www.rirdc.gov.au Published in July 2002 Printed on environmentally friendly paper by Canprint

iii

Foreword This publication is a product of the 8th International Symposium on Nitrogen Fixation with Non-Legumes, held at the University of Sydney in December 2000. The theme of Biofertilisers in Action was adopted for the Symposium and this book comprises a set of invited papers selected as most relevant to this theme. This is a forward-looking book related to the future application of the plant growth-promoting rhizobacteria (PGPR) to help provide a more sustainable organically-based agriculture. This book is a genuine attempt to set the stage for this development. If justified by subsequent field trials and economic and environmental advantages, the area has a major potential to influence agriculture. The project was funded from RIRDC Core Funds which are provided by the Federal Government of Australia and is an addition to RIRDC’s diverse range of over 800 research publications. It is part of our Resilient Agricultural Systems sub-program which aims to enable agricultural production systems that have sufficient diversity, flexibility and robustness to be resilient and respond to challenges and opportunities. The report includes information mainly on non-symbiotic nitrogen fixation. It contains useful information on biological nitrogen fixation (BNF) associated with a range of different crops including rice, wheat, maize and even bananas. The use of BNF technology in agriculture can reduce environmental problems like production of greenhouse gases in the atmosphere as well as nitrate toxicity in the ground water caused by chemical fertilizers. How this might be achieved is thoroughly discussed in this book. The information published in the book can be used in agricultural industries everywhere, including Australia. Most of our publications mainly related to innovations in agriculture are available for viewing, downloading or purchasing online through our website: • downloads at www.rirdc.gov.au/reports/Index.htm • purchases at www.rirdc.gov.au/eshop Simon Hearn Managing Director Rural Industries Research and Development Corporation

iv

Acknowledgments The authors of this book would like to acknowledge the support of the Rural Industries Research and Development Corporation, the Grains Research and Development Corporation, the Sugar Research and Development Corporation, Australian Agency for International Development, Australian Centre for International Agricultural Research, The Institution of Engineers Australia, Bio-Care Technology, United National Educational, Scientific and Cultural Organization (UNESCO Pacific Region, Samoa), the Australian Academy of Science and the Australian Journal of Plant Physiology. Without their support, the important event of the 8th International Symposium on Nitrogen Fixation with Non-legumes with its focus on the development of biofertilisers would not have been possible. Their support was clearly given in the planetary interest. Ivan R Kennedy Professor in Agricultural & Environmental Chemistry Editor

v

Contents Foreword ............................................................................................................................................... iii Acknowledgments................................................................................................................................. iv Executive Summary ........................................................................................................................... viii 1. Concerted action for cereal and other non-legume crop nitrogen fixation, enhanced growth and yield, E. C. Cocking ................................................................................. 1

1.1 The need for action ................................................................................................................ 1 1.2 A research plan of concerted action....................................................................................... 2 1.3 The proposed international concerted action programme...................................................... 2 1.4 References.............................................................................................................................. 3

2. The inoculant biofertiliser phenomenon and its potential to increase yield and reduce costs of crop production: The need for quality control, I. R. Kennedy and R.J. Roughley ......................................................................................................................... 4

2.1 Abstract .................................................................................................................................. 4 2.2 Introduction............................................................................................................................ 4 2.3 Inoculant biofertilisers ........................................................................................................... 5 2.4 Quality control ....................................................................................................................... 7 2.5 Conclusion ............................................................................................................................. 9 2.6 References.............................................................................................................................. 9

3. Removing nutritional limits to maize and wheat production: A developing country perspective, T. G. Reeves, S. R. Waddington, I. Ortiz-Monasterio, M. Bänziger, and K. Cassaday .................................................................................................. 11

3.1 Abstract ................................................................................................................................ 11 3.2 Introduction.......................................................................................................................... 11 3.3 Maize and wheat production systems in the developing world: are they being stretched to the limit? ................................................................................................. 11 3.4 The need for sustainable solutions to the food production challenge .................................. 13 3.5 Soil fertility strategies to sustain small-scale maize farmers in Africa ................................ 14 3.6 Improving efficiency and reducing environmental consequences of N use in irrigated spring wheat ...................................................................................................... 26 3.8 References............................................................................................................................ 31

4. The response of field-grown rice to inoculation with a multi-strain biofertiliser in the Hanoi district, Vietnam, Nguyen Thanh Hien, I. R. Kennedy and R. J. Roughley .... 37

4.1 Abstract ................................................................................................................................ 37 4.2 Introduction.......................................................................................................................... 37 4.3 Materials and Methods......................................................................................................... 38 4.4 Results.................................................................................................................................. 40 4.5 Discussion ............................................................................................................................ 43 4.6 Conclusion ........................................................................................................................... 44 4.7 Acknowledgments................................................................................................................ 44 4.8 References............................................................................................................................ 44

vi

5. Azobiofer: A technology of production and use of Azolla as biofertiliser for irrigated rice and fish cultivation, M. H. Mian................................................................... 45

5.1 Abstract ................................................................................................................................ 45 5.2 Introduction.......................................................................................................................... 45 5.3 Phase I. Quantification of N transfer from Azolla to rice plants .......................................... 46 5.4 Phase II. Establishment of Azolla as the source of N at field level...................................... 46 5.5 Phase III. Field trials on farms to test Azolla as nitrogenous biofertiliser for lowland rice . 47 5.6 Phase IV. Modelling of the technology ............................................................................... 49 5.7 Phase V. Finalization of the technology model ................................................................... 50 5.8 Acknowledgements.............................................................................................................. 53 5.9 References............................................................................................................................ 53

6. The spermosphere model to select for plant growth promoting rhizobacteria, J. Balandreau ...................................................................................................... 55

6.1 Abstract ................................................................................................................................ 55 6.2 Introduction.......................................................................................................................... 55 6.3 The spermosphere model ..................................................................................................... 56 6.4 Isolation of bacteria.............................................................................................................. 58 6.5 Comparison of strains .......................................................................................................... 58 6.6 Mathematical procedure used to compare strains ................................................................ 59 6.7 Fate of selected strains ......................................................................................................... 60 6.8 Conclusion ........................................................................................................................... 62 6.9 Acknowledgments................................................................................................................ 62 6.10 References............................................................................................................................ 62

7. Root stimulation and nutrient accumulation of hydroponically-grown tissue-cultured banana plantlets inoculated with rhizobacteria at lower level of nitrogen fertilisation, M. A. B. Mia, Z. H. Shamsuddin, W. Zakaria and M. Marziah ........................................................................................................................... 64

7.1 Abstract ................................................................................................................................ 64 7.2 Introduction.......................................................................................................................... 64 7.3 Materials and methods ......................................................................................................... 65 7.4 Results.................................................................................................................................. 65 7.5 Discussion ............................................................................................................................ 68 7.6 Conclusion ........................................................................................................................... 70 7.7 Acknowledgments................................................................................................................ 70 7.8 References............................................................................................................................ 71

8. The role of plant-associated beneficial bacteria in rice-wheat cropping system, K. A. Malik, M. S. Mirza, U. Hassan, S. Mehnaz, G. Rasul, J. Haurat, R. Bally and P. Normand............................................................................................................ 73

8.1 Abstract ................................................................................................................................ 73 8.2 Introduction.......................................................................................................................... 73 8.3 Materials and methods ......................................................................................................... 74 8.4Results.................................................................................................................................... 76 8.5 Discussion ............................................................................................................................ 78 8.6 Acknowledgments................................................................................................................ 80 8.7 References............................................................................................................................ 81

vii

9. Facilitating a N2-fixing symbiosis between diazotrophs and wheat, N. Islam, C. V. S. Rao and I. R. Kennedy ..................................................................... 84

9.1 Abstract ................................................................................................................................ 84 9.2 Introduction.......................................................................................................................... 84 9.2 Materials and methods ......................................................................................................... 85 9.4 Results.................................................................................................................................. 86 9.5 Discussion ............................................................................................................................ 89 9.6 Acknowledgements.............................................................................................................. 91 9.7 References............................................................................................................................ 92

10. Sesbania: A potential nitrogen source for sustainable rice production, A. T. M. A. Choudhury, S. K. Zaman and N. I. Bhuiyan ................................... 94

10.1 Abstract ................................................................................................................................ 94 10.2 Introduction.......................................................................................................................... 94 10.3 Potentials of Sesbania to supply plant nutrients .................................................................. 95 10.4 Establishment of Sesbania in the dry season-fallow-rainy season cropping pattern ........... 96 10.5 Establishment of Sesbania in the summer season-rainy season cropping pattern ............... 98 10.6 Conclusions........................................................................................................................ 100 10.7 Acknowledgements............................................................................................................ 100 10.8 References.......................................................................................................................... 100

11. An economic analysis of inoculant biofertiliser production and use in Vietnam, G. Barrett and S. Marsh................................................................................ 102

11.1 Abstract .............................................................................................................................. 102 11.2 Introduction........................................................................................................................ 102 11.3 Background........................................................................................................................ 102 11.4 The cost of biofertiliser production at Ba Vi biofertiliser factory ................................... 103 11.5 Expected economic benefits for farmers from the use of biofertiliser technology ............ 105 11.6 Perceptions of environmental and social benefits.............................................................. 108 11.7 Areas for further study ....................................................................................................... 109 11.8 Conclusions........................................................................................................................ 110 11.9 Acknowledgements............................................................................................................ 111 11.10 References........................................................................................................................ 111

12. A model for testing the effectiveness of biofertiliser for Australian rice production, R.L. Williams and I.R. Kennedy ......................................................................... 112

12.1 Abstract .............................................................................................................................. 112 12.2 Introduction........................................................................................................................ 112 12.3 Methods ............................................................................................................................. 112 12.4 Conclusion ......................................................................................................................... 114 12.5 References.......................................................................................................................... 114

13. Author Affiliations .................................................................................................................... 115

14. Selection of pictures presented at the conference ................................................................... 118

viii

Executive Summary This book is published with the funding from the Rural Industries Research and Development Corporation (RIRDC Grant Number WS 990.23). It includes some selected papers presented in the 8th International Symposium on Nitrogen Fixation with Non-Legumes held on December 3-7, 2000, at the University of Sydney, using the theme Biofertilisers in Action. The symposium involved an international meeting of about 150 scientists seeking ways to reduce the need for chemical fertilisers, such as ammonia and urea, for crops as well as plant-growth promoting rhizosphere (PGPR) effects caused by microbes based on phytohormonal effects and the mobilisation of nutrients such as phosphorus, nitrogen and other elements. An emphasis was placed on interaction with the farming community at the symposium and a group of about 15 farmers or farming industry representatives also attended the symposium, particularly on day 1, which had a stronger emphasis on farm-related issues. It was obvious from the papers and discussion at the symposium that this area of agricultural science is coming of age and that there is now sufficient data available to begin defining the scientific principles that will be involved in the successful application of this new technology on farms. On subsequent days, the symposium heard ‘success stories’ described for microbial biofertilisers. One example was for the sugar industry in Brazil and Cuba, and a proposal to attempt to transfer this system to Mauritius; this system is based on the colonisation of the roots and stems of sugar plants with nitrogen-fixing organisms. It is of interest that Mauritius is the country from which all the current cultivars of commercial sugar production were derived over the past 300 years, since sugar was established as a crop there by the Portuguese. Brazil grows its sugar crop using about half the nitrogen fertiliser application rate per unit of product as other countries such as Australia and Mauritius. Research data was also given for successful trials involving the application of inoculants to maize in the United States and Mexico. Eighteen invited papers from the Symposium have been published in a special issue of the Australian Journal of Plant Physiology (volume 28, number 9, 2001). Another set of invited papers are reproduced in this volume. The successful application of microbial inoculants under farming conditions for the growth of rice in Vietnam, Pakistan and Bangladesh is described. These inoculant biofertilisers involve the application of mixed cultures of microorganisms which have been selected for their demonstrated capacity to stimulate the vegetative and grain yield of rice. Other papers at the symposium, such as one given by the director of CIMMYT included in this volume, dealt with more general conditions required for the successful growth of crops in developing countries with special reference to needs for nitrogen. Often, there are social and political problems as well as scientific obstacles to growing sufficient cereals for local needs. But the application of inoculant biofertilisers may be very beneficial for poor farmers in such circumstances, as is examined by an economic analysis of inoculant biofertiliser production in Vietnam included in this report. The total operating budget for the symposium was around $180,000 (see attachment). This involved about $90,000 of special funding contributed by a range of agencies including GRDC, RIRDC, AusAID (ISSS), SRDC, ACIAR, UNESCO, the Crawford Fund and almost $10,000 of donations from industry and SUNFix. Other funding was from about $90,000 of registration subscriptions. RIRDC funding of $10,000 was received in 2000 (included in the above budget), with a further $5,000 to be paid to the University of Sydney on submission of camera-ready copy of Biofertilisers in Action. The received RIRDC funding was applied to subsidise post-graduate students, accommodation of some scientists from Asia and Europe, and partial payment of the media consultant. There was extensive media publicity during the symposium. In addition to the selected published papers in the Australian Journal of Plant Physiology and in this book, the abstracts of all the presented papers are included in the Book of Abstracts published by the SUNFix Centre for Nitrogen Fixation, the University of Sydney prior to the commencement of the

ix

symposium. The published materials in this book are informative, and should be useful to the scientists, administrators and policy makers working on environmental and agricultural sectors elsewhere in the world and in Australia. Strong appreciation is expressed to RIRDC for the funding supplied for this symposium and publication of Biofertilisers in Action. The organisers and participants consider that this will be considered a landmark conference, enabling the establishment of new agricultural technology with strong economic and environmental advantages.

1

1. Concerted action for cereal and other non-legume crop nitrogen fixation, enhanced growth and yield E. C. Cocking 1.1 The need for action

More than 20 years ago the Bonn Conference on Agricultural Production highlighted the fact that most of the staple cereal crops, with little or no access to biological fixed nitrogen, require generous applications of man-made nitrogen fertiliser to express their full yield potential (Lampe et al. 1979). But that in India and Brazil, however, associations with nitrogen-fixing bacteria had been found among some cereal grains, which make up as much as 90 percent of staple foods in developing countries. At that time the potential economic values of these associations was still unknown. In the case of rice, data suggested that bacterially fixed nitrogen of the order of 10 to 30 kg ha-1 per crop could be supplemented with additional nitrogen fixed by algae in the rice-land waters. The Conference recommended that the possibility that commercial nitrogen applications could be eliminated or reduced on non-leguminous crops by such plant-microbe relationships should be thoroughly examined. Indeed Fritz Haber, inventor of industrial ammonia synthesis, when concluding his Nobel Prize acceptance speech in 1920 noted: 'it may be that this solution is not the final one. Nitrogen bacteria teach us that Nature, with her sophisticated forms of the chemistry of living matter, still understands and utilises methods which we do not as yet know how to imitate' (Smil 2001). The Discussion Meeting at the Rockefeller Foundation Bellagio Conference Centre, Lake Como, Italy in 1997 on Biological Nitrogen Fixation : The Global Challenge and Future Needs, highlighted that substantial extra demands for fixed nitrogen in the coming period to 2020 are certain, as result of the Earth's increasing human population (Kennedy and Cocking 1997). In discussing the choice between biological nitrogen fixation (BNF) and industrial (chemical) nitrogen fixation it was clear that, whilst both sources of fixed nitrogen will be needed to meet the demand for food production, BNF has the advantages of lower cost, reduced production of greenhouse gasses, such as oxides of nitrogen and carbon dioxide, and less nitrate contamination of ground water. Furthermore BNF is more consistent with the development of sustainable farming, including organic farming.

In a more recent Round Table Discussion it was pointed out that the striking rise in cereal grain yields in Developed Countries since 1950 is directly attributable to a 10-fold increase in nitrogenous fertiliser use (Rolfe et al. 1998). The 'Green Revolution' in agriculture in the developing world, which resulted in large increases in cereal grain production since the 1960s, has been the result of the development of plant genotypes of rice and wheat which are highly responsive to chemical fertilisers, particularly nitrogenous fertilisers. However, there have arisen also a series of concomitant environmental problems along with this increased use of nitrogenous fertiliser. Indeed this global nitrogen overload now grows critical and it is clear that the biosphere is becoming glutted with nitrogen compounds (Moffatt 1998).

As we move into the 21st Century it is increasingly pressing for an action plan to enhance biological nitrogen fixation, especially in the main cereals of the world, rice, wheat and maize. "For the immediate future there is a pressing need, in all our important food crops, for improved resistance to viruses and insect pests, for tolerance to salt, drought and heat, for higher-quality grain and other products and for the most demanding goal of all, improved systems of nitrogen fixation" (Conway 1997). Overall we need an ‘Evergreen Revolution', rooted in the principles of ecology, equity, economics, employment generation and energy use efficiency (Swaminathan 1996).

2

1.2 A research plan of concerted action

The general background It is clear that the extension of BNF to the World's major cereal crops would be of enormous economical and environmental value. One of the major strategies envisaged to achieve this objective is the establishment of an effective association between cereals and diazotrophic (nitrogen-fixing) bacteria. Only those systems in which the diazotrophic bacteria grow and fix nitrogen within the plant can be effective for the establishment of symbiotic nitrogen fixation. In such endophytic associations, there can be intimate metabolic exchange with the plant and bacteria are protected from competition with other microorganisms of the root environment. An alternative plan to genetically engineer crops to fix their own nitrogen, without the use of diazotrophic bacteria, requiring the co-ordinated and regulated expression of sixteen nitrogen fixing genes in an appropriate plant cellular location is an extremely complex task and is not likely to be achieved at the earliest before the middle of this Century. The most efficient systems of endophytic biological nitrogen fixation are the rhizobia-legumes and Frankia-woody plant non-legume symbioses, in which the bacteria fix nitrogen within specialised plant organs, called nodules, resulting in considerable assimilation of fixed nitrogen by the host plants. In comparison certain diazotrophs (called associative diazotrophs) colonize mainly the surfaces of plant roots, where they are in competition with other rhizosphere, root inhabiting microorganisms, and very little of the biologically fixed nitrogen benefits the plant. Associative diazotrophs have been found and characterised on roots of several of the most important non-legume crops, such as rice, wheat and maize. However, because of the inefficiency of such associations, it is now a primary aim to establish endophytic nitrogen-fixing associations in these cereals in which diazotrophic bacterial grow and fix nitrogen within the plant. Encouragingly, such an endophytic nitrogen-fixing association, without nodulation, has recently been described in certain Brazilian varieties of sugarcane, which have been growing for many years without added nitrogenous fertiliser. Sugarcane is a member of the Gramineae, the grass family which also includes the cereals, the large-seeded annual grasses. Large populations of endophytic diazotrophs have been found in sugarcane. These bacteria intercellularly colonize the plant, including the xylem, and seem likely to be major contributors in sugarcane to the observed high levels of biological nitrogen fixation (Boddey 1995). This suggests that endophytic nitrogen fixation by diazotrophs, without nodule formation, is likely to be possible in other members of the Gramineae such as rice, wheat and maize, and in other non-legume crops. 1.3 The proposed international concerted action programme

This Concerted Action will involve an integrated approach utilising facilities and expertise arising from the spectrum of studies on the interaction of diazotrophs with non-legumes already being undertaken in individual laboratories. This will involve a Multi-National Group representing a broad range of current skills available in the fields of the plant sciences, microbiology, biochemistry, molecular genetics and molecular biology. The present Management Plan (which can be amended as necessary) will involve a Co-ordinator (UK) and Partners in Australia, USA, Germany, Switzerland, Egypt, India and the Philippines who will internationally co-ordinate (utilising e-mail, web sites, seminars, site visits and exchanges of staff) research integrated to focus on particular research objectives. The research objectives will be progressive, ranging from basic studies on the factors influencing the interaction of various diazotrophs primarily with rice and wheat, under controlled laboratory conditions, to establish nitrogen fixing endophytic colonisation, to assessment of such colonisation on plant growth, yield, nitrogen fixation and fertiliser requirements under field conditions. These studies will provide a foundation for the production of seed or soil inoculants with different sets of bacterial strains, and interaction stimulants such as flavonoids, to cover various soil and environmental conditions.

3

1.4 References

Boddey RM (1995) Biological nitrogen fixation in sugarcane: a key to energetically viable biofuel production. Critical Reviews in Plant Sciences 14, 263-279.

Conway G (1997) The Doubly Green Revolution: Food for All in the Twenty-first Century. (Penguin Books Ltd, Hamonsworth, Middlesex, England, United Kingdom).

Kennedy IR, Cocking EC (1997) Biological Nitrogen Fixation: The Global Challenge & Future Needs. Position Paper, The Rockefeller Foundation Bellagio Conference Centre, Italy, April 8-12, 1997. ISBN: 1-86451-364-7. (SUNFix Press, The University of Sydney, Australia).

Lampe KJ, Kruesken E, Treitz W, Pino JA (1979). Agricultural Production: Research Development Strategies for the 1980s: Conclusions or Recommendations of the Bonn Conference, 8-12 Oct, 1979. (German Foundation for International Development, German Agency for Technical Cooperation, Federal Ministry of Economic Cooperation and The Rockefeller Foundation, United States of America).

Moffatt AS (1998) Global nitrogen overload: Problem grows critical. Science 279, 988-989. Rolfe BG, Verma DPS, Potrykus I, Dixon R, McCully M, Sautter C, Dénarié J, Sprent J, Reinhold-

Hurek B, Vanderleyden J, Ladha JK, Dazzo FB, Kennedy IR, Cocking EC (1998) Round Table: Agriculture 2020: 8 Billion People. In ‘Current Plant Science and Biotechnology in Agriculture: Biological Nitrogen Fixation for the 21st Century’. (Eds. C. Elmerich, A. Kondorosi, and WE Newton) pp. 685-692. (Kluwer Academic Publishers, The Netherlands).

Smil V (2001) Enriching the Earth: Fritz Haber, Carl Bosch and the Transformation of World Food Production. 338 pp. (The MIT Press, Cambridge, Massachusetts, United States of America).

Swaminathan MS (1996) Sustainable Agriculture; Towards an Evergreen Revolution. (Konark Publishers PVT Ltd, Delhi 110092, India).

4

2. The inoculant biofertiliser phenomenon and its potential to increase yield and reduce costs of crop production: The need for quality control I. R. Kennedy and R.J. Roughley 2.1 Abstract

An assessment of the inoculant biofertiliser phenomenon indicates that significantly increased yields of cereal crops such as rice and maize can now be reliably obtained. Not only are chemical fertiliser input costs reduced, but the increased crop yield possible provides farmers with an even more attractive increase in income, shown by an economic analysis of a farming system involving biofertiliser application. Such effects on yield apparently result from a complex of cooperative positive effects exerted by microorganisms associated with plant tissues, particularly in the root zone. This paper assesses the evidence for positive biofertiliser effects in greenhouse and field experiments and the positive effects on grain yield. The papers presented in the book and elsewhere support its validity and illustrate the key experimental methods that have been used for this demonstration. However, the possibility of variable responses from biofertiliser products demands that obtaining benefits consistently will require significant attention to quality control of inoculants and of their beneficial effects. In the conclusion to this paper, quality control methods ensuring adequate inoculum potential of the strains shown to improve crop yields on farms are discussed.

2.2 Introduction

Most of the papers in this book were submitted by speakers at the 8th International Symposium on Nitrogen Fixation with Non-Legumes, held under the auspices of the SUNFix Centre for Nitrogen Fixation at the University of Sydney, 3-7 December 2000. These symposia owe their commencement to the pioneering work of the late Dr Johanna Döbereiner of Brazil for her work in the 1960s with N2-fixing associative bacteria such as Azospirillum for grasses and cereals; they also recognise that biological nitrogen fixation by symbiotic non-legumes by microorganisms such as Frankia in plant species like Casuarina and the cyanobacteria in plant species like Azolla are also very important in the nitrogen cycle of natural ecosystems. James M. Vincent, who like Johanna Döbereiner died in the year 2000 just a few weeks before the symposium was held, also made a major contribution to the planet’s agricultural microbiology important for biological nitrogen fixation with his outstanding work on the quality control of rhizobial inoculants, ensuring that the symbiotic genomes of plants and microbes could be properly matched (Murrell and Kennedy 1988). A similar challenge is now being faced by those working with many other microbial species that have now been recognised as the beneficial root-zone or rhizo-bacteria. Provided that the same rigorous quality control can be applied to inoculant biofertilisers as was applied by Vincent to rhizobia, and that these systems will be studied in the field by agronomists using microbes rather than chemical fertilisers, as Jim always demanded, the prospects for success in application of plant-growth promoting rhizobacteria (PGPR) in agriculture will be much more secure. The theme of the symposium was the same as the title of this book - Biofertilisers in Action. This theme recognised that the concept of cooperative action between microbes and plants is a significant element in the successful functioning of ecosystems. However, it is important to understand that the phenomenon of microbial fertilisation of plants embraces far more than the process of biological

5

nitrogen fixation. Both the symbiotic and associative or endophytic microbe systems also help mobilise phosphorus and other soil nutrients (e.g. K, Mg, Ca, Zn, Na and Mo) that plants might otherwise have difficulty in accessing at an adequate rate. The synergistic effect of N2-fixing strains and P-solubilisers in the rhizosphere of crop plants and of mangroves (Rojas et al. 2001) has been clearly demonstrated. In addition, the rhizobacteria can modify the morphological development of plant roots in such an interactive way that greater crop productivity is apparently possible. 2.3 Inoculant biofertilisers

A large number of specially prepared inoculant biofertilisers using isolated strains of microorganisms have now been produced in countries such as China, India, Pakistan Egypt, Vietnam and Indonesia (Kennedy and Cocking 1997). These have been subjected to laboratory and field testing to variable extents and are thus the data are of variable reliability. The authors have personally observed in several countries that these products are usually capable of stimulating early growth of cereal crops in trials on field stations and there is a significant number of literature reports of such effects on growth. However, reliable statistically valid data showing significant increases in grain yield on farms are less evident. The data given by Nguyen et al. in this volume is now one such verified case. Here, a field trial in which the biofertiliser effect was verified at a 0.1% level of statistical significance is reported, leading to an increase in grain yield of significant benefit to farmers. This is verified by even greater average increases in yield of rice by more than 60 farmer trials where half the farm is treated with a full suite of chemical fertilisers and the other half receives biofertiliser with a recommended reduction in the use of fertilisers imported into the farm, particularly of urea (see Barrett and Marsh, this volume). This result in substantial benefits to farmers, reducing input costs but more importantly, generating more income by enhanced yields. With more confidence in the biofertiliser system and knowledge of soil fertility, it might be possible to obtain an even greater reduction in use of industrial fertilisers but the benefits from higher yields are also needed to justify this approach.

The need for a sustainable system based on use of biofertilisers Reeves et al. writing from CIMMYT’s perspective in this volume indicate the magnitude of the challenge required to lift the productivity of crops like maize and wheat by farmers everywhere in an affordable way, particularly in developing countries, emphasising that obtaining sufficient N fertiliser will be the key factor. Although no direct role for inoculant fertilisers was suggested by these authors in achieving this outcome, the paper nonetheless describes the context in which this could be achieved, pointing to the need for integrated approaches. The application of biofertiliser products that can reduce the need for costly N fertiliser inputs, whilst maintaining or even increasing grain yields as discussed by Barrett and Marsh is proposed to be part of the solution to this problem and should be added to the ‘best bet’ technologies referred to in Table 2 of Reeves et al. However, implementing this application will require proper organisation with adequate attention paid to optimising the biological factors involved. Unfortunately, planners attempting to raise productivity will continue to recommend inputs of chemical fertilisers, for which cost and yield estimates may more readily be made, rather than recommend a biological approach while uncertainty regarding effectiveness exists. However, a systems approach to agriculture and food production where farming is recognised as part of a community activity would more readily accommodate such an integrated approach. The result may be a system that generates more wealth locally with a greater range of participants if the infrastructure required can be provided.

Cooperating plant-microbe-microbe systems Although there has been a tendency in the past to apply a single biofertiliser strain to crops similar to the use of Rhizobium on legumes, there is now a shift towards regarding a suite of organisms with three or more strains applied to crops simultaneously as much more likely to prove successful. This is based on the concept that plants respond to a range of fertilisers with quite different responses and

6

that the PGPR effect may involve several phytohormonal responses all combining to improve yield. The use of multi-strain inoculants may also provide more reliability where similar capabilities such as the ability to fix N2 or mobilise organic nitrogen or insoluble phosphorus are shared by different strains. Such redundancy in growth-promoting properties could be very important in providing the full range of responses needed to increase crop yields in a wide range of environmental conditions. Several of the papers in this book provide evidence on the response to individual biofertiliser or PGPR strains (Islam et al.; Malik et al., Mian; Mia et al.), including Azospirillum, Bacillus, Pseudomonas, Enterobacter, Citrobacter, Zoogloea, Burkholderia, and others. These papers also provide some evidence of the extra benefits possible from inoculation with several strains at once. A difficulty in demonstrating the effectiveness of applying multiple strains is the extra complexity of such trials and of showing conclusively the combined effect as significant. However, it is entirely reasonable that such combinations of strains will prove more reliable. Clearly, there is a need to develop the methods needed to study such systems.

Traditional and new biofertilisers Although the emphasis in this book is on the application of microbial strains as inoculants, the term biofertilisers is a nomenclature by no means exclusive to these in the literature. The papers by Mian and Choudhury and their colleagues illustrate the continuing role for the symbiotic fern Azolla and for the rapidly-growing legume Sesbania as means of fertilising cereal crops grown either simultaneously or in rotation, partly as green manures. Studying these systems raises many of the issues that emerge when considering the effectiveness of inoculant biofertilisers. Hundreds of kilograms per ha of nitrogen for each crop can be input into farms using such approaches. These positive effects from biofertilisers are also well illustrated in a complementary set of papers published in a special issue of the Australian Journal of Plant Physiology, also generated from the international symposium. For example, maize-endophyte associations with diazotrophic microbes such as Klebsiella and Acetobacter or Herbaspirillum species clearly result in greater grain yield (Riggs et al. 2001), although this is recognised not as a result of biological N2 fixation but of other components of the PGPR effect. The challenge is to extend this effect to include biological nitrogen fixation for which we have the example of other plant species which do fix N2. The application of 15N analysis by mass spectrometry has confirmed that other tropical pasture species (Reis et al. 2001) and sugarcane (Boddey et al. 2001) do benefit in their nitrogen nutrition from their cooperative action with various genera of rhizobacteria. The azospirilla continue to occupy an important role for agronomically important crops as discussed by Dobbelaere et al. (2001) and even Rhizobium leguminosarum biovar trifolii may have stimulated the yield of rice for many hundreds of years (Yanni et al. 2001), although not from N2 fixation. The role of diazotrophic cyanobacteria or blue-green algae, who are often potentially beneficial rather than undesirable because of animal toxins, have been assessed for Bangla Deshi rice paddies by Hashem (2001). For those who have made their careers studying these symbiotic biological systems, the thought is quite an easy one that Darwinian evolution is only partly about survival under competition and more often about natural selection of cooperative systems. This concept of cooperative evolution as a result of natural selection was given a possible causal basis in a recent book by the organiser of this symposium (Kennedy 2001), who promoted the idea that competition provides an efficient cause for purposive evolution of cooperative systems, so that different species will usually occupy complementary niches in ecosystems even if they are not obviously symbiotic. One example is the coupled system of nitrification followed by denitrification (see Kloos et al. 2001) and nitrogen fixation all carried out by different species of bacteria. In this hypothesis, natural selection operates as a direct result of thermodynamic forces, expressed as specific impulses from the momentum of the quanta of field energy. These oscillations in enthalpy selectively generate new forms of biological action that can be recorded for transmission to progeny in the DNA base sequences of genes. This proposal represents an extension of the ideas put forward by Schrödinger (1944) who claimed that

7

the quantum theory, perhaps in some newer guise, would eventually prove to be the fundamental physical basis for life. If so, the main thrust of this symposium of attempting to do the basic science needed to generate new effective symbiotic associations between cereals and the biofertiliser strains has nature on its side. Other papers in this volume address some of the ways and means of exerting the selective pressure needed to evolve these systems (Richardson 2001; Perrine et al. 2001; Bauer and Teplitski 2001) including control of the process of ammonium transport (Van Dommelen et al. 2001; Wood et al. 2001) and the use of the genetically accessible Arabidopsis thaliana as a possible plant model of these associations (O’Callaghan et al. 2001). Fears of the possible reduction of the nitrogenase capacity of heat-treating sugar cane setts for pathogen control has been allayed (Ortega et al. 2001).

2.4 Quality control

One hundred years ago, rhizobial inoculants were often of poor quality and their usefulness was widely doubted as a result. Even as late as 1932, almost 50 years after Rhizobium had been established as causing N2-fixing nodulation in legumes, Fred, Baldwin and McCoy (1932) could still claim in their classical book that “The writers cannot but feel that the exaggerated and sometimes totally unwarranted claims of some of the advertisements, as well as the lack of information concerning the proper method of handling did much to discredit the use of cultures. No doubt the opinion of scientists from other countries also contributed to the sceptical attitude of the public regarding pure culture inoculation”. Successful application of rhizobia in Australia was only achieved when the following conditions were met: • Quality control for standards of inoculants was voluntarily accepted by manufacturers. • Quality control was also applied to primary (mother) cultures used by manufacturers and their

commercial products by an independent body (UDALS/AIRCS/ALIRU). • Distributors and users were educated in the proper use of the peat inoculants. Although bacterial cultures in peat is the standard carrier adopted in Australia, because of greater reliability, there is no reason why other delivery systems might not also be approved in a quality control procedure if the required standards are met. As with Rhizobium, the main requirement in the potential application of biofertilisers is the availability of high quality inoculant products. Because of the more diverse mechanisms involved in the function of biofertilisers and the larger range of systems it is even more difficult to be prescriptive about quality control. Nonetheless, there are three main criteria that must be applied if the possible benefits that have been shown from inoculant biofertilisers are to be achieved on farms. These are: • The biofertiliser product must be confirmed to contain beneficial strains of microorganisms.

Strain selection is the first step of inoculation technology, requiring much care and time. According to Nguyen Thanh Hien (see Nguyen et al. 2001) criteria for successful selection of biofertiliser strains for rice are (i) the strain should be the most abundant of its kind in the soil (ii) it should have high activity (e.g. N2 fixation or phosphate solubilising activity) (iii) the strain should be as fast-growing as possible, improving the success rate if non-sterile carrier media must be used (iv) the strain must be shown not to cause root disease and finally (v) the strains incorporated should be continuously reselected to maintain their effectiveness. Confirmation of the correct identity of biofertiliser strains used to prepare biofertiliser inoculants can be based on cultural characteristics, the use of PCR methods, immunodiagnostic tests or immunoblots, DNA hybridisation techniques, etc.

• The biofertiliser must be shown to contain adequate numbers of microbes for the purpose. For this reason, starter cultures should be of good quality and adequate to allow rapid growth in the

8

media selected. The different strains should be grown separately as far as possible and the strains mixed just before preparation of the final product for application to plants to reduce the effects of competition on growth. Carrier media (peat, soil with high in organic matter, clay) should be carefully selected and verified so that it allows growth of sufficient numbers of each strain in the biofertiliser product. Confirmation of approximate numbers of each strain in the biofertiliser products can be achieved by dilution counts, immunodiagnostic tests, etc.

• The inoculant biofertiliser must be clearly demonstrated as effective in improving crop yield and in reducing the need for chemical fertilisers on farms. While experimental field trials may be useful in demonstrating effectiveness of crop response, the recommended means of demonstrating effectiveness are farmer trials with full-fertiliser treatments as controls, as employed in Nguyen et al. in this book.

Achieving these standards must be an evolutionary process. Field data from well characterised field sites must be accumulated in the beginning phase. Obviously, inoculants must be well characterised, shown to be pure and numbers of colony forming units (cfus) measured. This will allow definition of the conditions under which positive responses as extra crop production can be obtained and the inoculum potential required. Any other factors associated with effectiveness and failure must also be recorded. Only when all these requirements have been met will it be possible to begin to formulate quality control criteria. Confirmation that biofertiliser strains have survived after inoculation and effectively colonised the root zone of crops using genetic markers or immunodiagnostic or other tests is also desirable. The more difficult tests mentioned here should be restricted to quality control for stater cultures used to produce biofertiliser products in small factories. Testing at the field stage of application requires the simplest tests possible. As with standards for rhizobia, it may be prudent to adopt minimum standards when quality control is first applied, raising quality standards such as numbers of organisms per g as methods improve. For example, a decision could be taken to fail the poorest 20% of product samples. In this way, the quality control system provides a guiding influence to improvement of standards. It is obvious that success in the application of biofertiliser products will also require a significant infrastructure. This would not be dissimilar to methods of quality control imposed for rhizobia, such as the Australian inoculants research and control service (UDALS/AIRCS/ALIRU) maintained by manufacturers of Rhizobium inoculants near Sydney since the 1950s. An infrastructure closely linked to the biofertiliser production industry allowing ongoing research to improve inoculant quality as well as providing quality control of current production and of stored commercial products is desirable. Professor Cocking in this volume has called for concerted action to encourage biofertiliser production to become a more significant feature of planning in world agriculture. This would help to overcome chronic problems such as that of low farm productivity and poor returns on labour referred to by Reeves in this volume. It is the hope of the authors of this book that international and national agencies will take note of this need. The paper of Barrett and Marsh in this book indicates that the biofertiliser industry may to some extent be self-sustaining for rice in Vietnam. But international agencies with more significant resources must invest in the development of the technical and educational infrastructure producing quality control for this industry, sensitive to the economic and environmental options for international agriculture in developing countries in the 21st century outlined by Reeves et al.

9

2.5 Conclusion

It now appears that the prospects for effective microbial biofertilisers for cereal crops like rice, maize and (eventually) wheat becoming available soon are bright. However, it is also apparent that much work still needs to be done to achieve this beneficial outcome for the planet’s farmers. That this work is an imperative for the beginning decade of the third millennium has already been signalled by evidence of the declining availability of fossil fuels, with the failure rate for new oil wells now increasing (see recent articles in Scientific American and elsewhere). Should the price for chemical fertilisers increase, it will be essential for farmers in developing countries to have greater access to cheaper biofertiliser technology. It is also evident that chemical fertilisers generate much more greenhouse gases such as N2O because of their inefficient utilisation by crops. So inoculant biofertilisers may be more environmentally sound and their use could help mitigate the onset of global warming as well as reduce the fertiliser input costs of farmers. 2.6 References

Bauer WD, Teplitski M (2001) Can plants manipulate bacterial quorum sensing? Australian Journal of Plant Physiology 28, 913-921.

Boddey RM, Polidoro JC, Resende AS, Alves BJR, Urquiaga S (2001) Use of the 15N natural abundance technique for the quantification of the contribution of N2 fixation to sugar cane and other grasses. Australian Journal of Plant Physiology 28, 889-895.

Dobbelaere S, Croonenborghs A, Thys A, Ptacek D, Vanderleyden J, Dutto P, Labandera-Gonzalez C, Caballero-Mellado J, Aguirre JF, Kapulnik Y, Brener S, Burdman S, Kadouri D, Sarig S, Okon Y (2001) Responses of agronomically important crops to inoculation with Azospirillum. Australian Journal of Plant Physiology 28, 871-879.

Fred EB, Baldwin IL, McCoy E (1932) Root Nodule Bacteria and Leguminous Plants. University of Wisconsin Studies in Science No. 5. (University of Wisconsin Press, Madison, Wisconsin, United States of America).

Hashem MA (2001) Problems and prospects of cyanobacterial biofertilizer for rice cultivation. Australian Journal of Plant Physiology 28, 881-888.

Kennedy IR (2001) Action in Ecosystems: Biothermodynamics for Sustainability, (Research Studies Press, Baldock, United Kingdom).

Kennedy IR, Cocking EC (1997) Biological Nitrogen Fixation: The Global Challenge & Future Needs. Position Paper, The Rockefeller Foundation Bellagio Conference Centre, Italy, April 8-12, 1997. ISBN 1-86451-364-7. (SUNFix Press, The University of Sydney, Australia). 83 pp.

Kloos K, Mergel A, Rösch C, Bothe H (2001) Denitrification within the genus Azospirillum and other associative bacteria. Australian Journal of Plant Physiology 28, 991-998.

Murrell WJ, Kennedy IR (1988) Microbiology in Action. (Research Studies Press, Letchworth, United Kingdom).

Nguyen, Thanh Hien, Kennedy IR, Roughley RJ, Deaker R (2001) Quality Control Protocols for Inoculant Biofertiliser Production for Rice Crops. AusAID CARD Project Workshop, Hanoi June 2001. (SUNFix Press, The University of Sydney, Australia).

O’Callaghan KJ, Dixon RA, Cocking EC (2001) Arabidopsis thaliana: a model for studies of colonization by non-pathogenic and plant-growth-promoting rhizobacteria. Australian Journal of Plant Physiology 28, 975-982.

Ortega E, Rodés R, de la Fuente E, Fernández L (2001) Does the routine heat treatment of sugarcane stem pieces for xylem pathogen control affect the nitrogen activity of an N2-fixing endophyte in the cane? Australian Journal of Plant Physiology 28, 907-912.

Perrine FM, Prayitno J, Weinman JJ, Dazzo FB, Rolfe BG (2001) Rhizobium plasmids are involved in the inhibition or stimulation of rice growth and development. Australian Journal of Plant Physiology 28, 923-937.

10

Reis VM, dos Reis Jr FB, Quesada DM, de Oliveira OCA, Alves BJR, Uquiaga S, Boddey RM (2001) Biological nitrogen fixation with tropical pasture grasses. Australian Journal of Plant Physiology 28, 837-844.

Richardson AE (2001) Prospects for using soil microorganisms to improve the acquisition of phosphorus by plants. Australian Journal of Plant Physiology 28, 897-906.

Riggs PJ, Chelius MK, Iniguez AL, Kaeppler SM, Triplett EW (2001) Enhanced maize productivity by inoculation with diazotrophic bacteria. Australian Journal of Plant Physiology 28, 829-836.

Rojas A, Holguin G, Glick BR and Bashan Y (2001) Synergism between Phyllobacterium sp. (N2-fixer) and Bacillus licheniformis (P-solubilizer) both from a semi-arid mangrove rhizosphere. FEMS Microbiology Ecology 35, 181-187.

Schrödinger E (1944) What is Life? (Cambridge University Press, Cambridge, United Kingdom). Van Dommelen A, De Mot R, Vanderleyden J (2001) Ammonium transport: unifying concepts and

unique aspects. Australian Journal of Plant Physiology 28, 959-967. Wood CC, Islam N, Ritchie RJ, Kennedy IR (2001) A simplified model for assessing critical

parameters during associative 15N2 fixation between Azospirillum and wheat. Australian Journal of Plant Physiology 28, 969-974.

Yanni YG, Rizk RY, El-Fattah FKA, Sqaurtini A, Corich V, Giacomini A, de Bruijn F, Rademaker J, Maya-Flores J, Ostrom P, Vega-Hernandez M, Hollingsworth RI, Martinez-Molina E, Mateos P, Velásquez E, Wopereis J, Triplett E, Umali-Garcia M, Anarna JA, Rolfe BG, Ladha JK, Hill J, Mujoo R, Ng PK, Dazzo FB (2001) The beneficial growth-promoting association of Rhizobium leguminosarum bv. trifolii with rice roots. Australian Journal of Plant Physiology 28, 845-870.

11

3. Removing nutritional limits to maize and wheat production: A developing country perspective

T. G. Reeves, S. R. Waddington, I. Ortiz-Monasterio, M. Bänziger, and K. Cassaday 3.1 Abstract

This paper outlines soil fertility challenges in developing countries and discusses approaches to those challenges. The authors begin by reviewing indicators of human welfare in developing countries and describing the projected supply and demand for maize and wheat. As developing countries struggle to match food supply to demand, the use of N fertiliser worldwide is projected to change. Developing countries are estimated to apply 40 million metric tonnes of N fertiliser, half of all N fertiliser used in agriculture worldwide. By 2025 consumption of N fertiliser may increase 60-90%, with two-thirds of the increase occurring in poor countries, although some of the poorest countries may actually register declines in fertiliser use. Research related to improving N availability in two contrasting settings is described: rainfed maize production systems of southern Africa, where low soil fertility has been identified as the single greatest constraint to food security, and the irrigated, highly productive environments where the bulk of the developing world’s spring wheat is produced. These examples indicate that no single approach to improving plant nutrition and crop yields will ease the growing demand for food in developing countries. Instead, integrated approaches will be needed. Various options are discussed, highlighting both the challenges and the opportunities for fostering sustainable food security in the developing world through improved plant nutrition. Keywords: Nutritional limits, maize, wheat, developing country. 3.2 Introduction

The International Maize and Wheat Improvement Centre (CIMMYT) works with research partners worldwide to devise strategies that enable farmers in developing countries to raise the productivity, profitability, and sustainability of maize and wheat cropping systems. Many farmers in developing countries may be unaware of issues related to nitrogen (N) fixation, but it is likely that they are even more preoccupied than their counterparts in high-income countries with their importance in sustaining and increasing crop production. This paper seeks to bring a developing country perspective to the issues raised at this international symposium. The objective of this paper is to outline soil fertility challenges in developing countrties and provide the context for subsequent discussion of novel approaches to those challenges. It begins by reviewing some indicators of human welfare in developing countries and next examines projected supply and demand for maize and wheat. The crucial role of sustainable agriculture in altering the landscape of food production in the developing world is described, with an emphasis on CIMMYT’s research related to improving N availability in two contrasting settings of the developing world. 3.3 Maize and wheat production systems in the developing world: are they being stretched to the limit?

At least one-quarter of the food calories consumed in the developing world come from maize and wheat. These crops are crucial to food security for hundreds of millions of rural and urban poor. It is evident that the world produces enough food to feed everyone, but the fact remains that many people cannot grow or buy enough food. It is also evident that the numbers of hungry, impoverished people are growing. The world will gain an additional 2 billion people in the next 25 years, and 97% of them

12

will be born in developing countries (World Bank 2000). Already, nearly half of the world’s people (2.8 billion) live on less than US$2 per day, and 1.2 billion survive on less than US$1 per day. As many as half of all children in the poorest countries are malnourished, and numbers of poor people are rising in Latin America, South and Central Asia, and Sub-Saharan Africa (World Bank 2000). Many of these people still live in rural areas and depend on agriculture to survive. Twenty years from now, the world’s farmers will have to produce 40% more grain to meet demand for cereals, including wheat and maize (Pinstrup-Andersen et al. 1999). In developing countries, the demand for wheat and maize will rise faster than demand for rice, the other major food staple. Demand for wheat will grow by 1.58% per year; demand for maize will grow by 2.35% per year. In two decades, 67% of the world’s wheat consumption and 57% of the world’s maize consumption will occur in developing countries. Even with projected production increases, by 2020 wheat will constitute more than 50% of the developing world’s net cereal imports. Maize will constitute 33% (Rosegrant et al. 1997). The potential for meeting food demand in the poorest countries with imports is doubtful because many developing nations cannot generate enough foreign exchange to buy the grain they need, and even if they could, demand is unlikely to be met through global trade. McCalla (2000) has observed that although world grain trade has more than doubled since 1960, “the share of world grain consumption that is traded has remained constant at about 10%.” Furthermore, he has noted that “if grain demand over the next 25 or 30 years increases 50-60%, and if trade increases only proportionately, to say 300 million tons, then it is clear that most of the increase in food production must come from production systems in the countries where the additional people will live.” As developing countries struggle to match food supply to demand, the use of N fertiliser worldwide is projected to change. Globally, fertiliser N applications are approximately 80 million metric tonnes; half of this amount is applied in developing countries (FAO 1990). Galloway et al. (1995) has estimated that by 2025 the consumption of N fertiliser will increase 60-90%, with two-thirds of the increase occurring in the developing world. Even though fertiliser use is projected to increase in some developing countries, it must not be forgotten that in some of the world’s poorest countries fertiliser use on food crops is actually declining. In many countries of sub-Saharan Africa, such as Kenya, Malawi, Nigeria, Zambia, and Zimbabwe, increasing poverty and the effects of structural adjustment programs have made it next to impossible for a very large number of farmers to use fertiliser, and their interest and reliance on alternatives, such as N-fixing legumes and biofertiliser inoculants such as those discussed elsewhere in this volume is growing. Wherever they occur, low soil fertility, declining fertiliser use, and growing populations are a potentially lethal combination for people and for civil society, because the result is less food and income per capita. Hunger and poverty are known to provoke conflicts over land, ignite political and ethnic unrest, and otherwise undermine the fragile ties that hold societies together. The urgency of devising better crop nutrition strategies to increase food production in the developing world can be seen from some calculations made by Smil (1999), which reveal the vital role of N fertilisers in current food production. Smil noted that in 1900, when agriculture did not rely on chemical N fertiliser, agriculture supported 1.625 billion people “by a combination of extensive cultivation and organic farming” on 850 million hectares. If the world used those same practices on today’s 1.5 billion hectares of farm land, no more than 53% of the world’s people could be fed a fairly meagre diet. If the world tried to feed its people at the current level of per capita food consumption using the agricultural practices in place in 1900, only 40% of the world’s people could be fed. He further estimated that nitrogen fertiliser currently provides the nutrients that feed the crops on which 2.2 billion people in low-income countries subsist. Instructive as they are, Smil’s calculations should be seen in an even wider context, because no single approach to improving plant nutrition and crop yields will ease the growing demand for food in developing countries. Instead, integrated approaches will be needed (not just better varieties, or more N-efficient varieties, alone; not just more inorganic nutrients, or organic nutrients, alone; but a

13

combination of approaches). Figure 1 provides an extremely simple but telling illustration of why complementary strategies are essential. The figure demonstrates the importance of improved varieties of food, feed, and fiber crops in combination with better plant nutrition. If the US had retained the same crop varieties used in 1938-40 and used no additional nutrients, and if farmers wanted to match contemporary levels of production, they would have to put an additional 200 million hectares under cultivation—a massive area that has been spared for natural habitats and other uses. It seems only reasonable to assume that the potential contribution of integrated crop and nutrient management strategies in other parts of the world could be correspondingly large. 3.4 The need for sustainable solutions to the food production

challenge

The world’s farmers need alternative crop and resource management strategies to sustain the productivity and profitability of cropping and also to sustain the natural resource base. Achieving sustainable food security implies achieving sustainable, high output agricultural systems. These systems must be economically viable, environmentally sound, socially acceptable, and politically supportable (Reeves 1999): • Sustainable farming systems must be economically viable at the farm and national levels. Poor

farmers cannot invest in systems that will not produce reasonable yields and also cash income, now and in the future. At the national level, agriculture must also contribute significantly to GDP and export earnings. The reality in most countries, especially developing countries, is that economic well-being and development are almost invariably based on productive and profitable agriculture.

• Sustainable farming systems must be environmentally sound as well. An economically viable agricultural sector cannot exist at the expense of our soils, air, water, landscapes, and indigenous flora and fauna, or the costs of temporary economic success will be heavy in the long term.

• Sustainable farming systems must be socially acceptable, appropriate to the people who, relying on their own meager resources, are responsible for implementing and managing them. The need for socially acceptable systems implies the need for a better understanding of farmer and community needs and values, as well as better targeting of technology to meet local conditions.

• Finally, sustainable farming systems must be politically supportable. Political support depends largely on successfully meeting the first three requirements of sustainability. If economic growth is catalysed by agriculture within an environmentally sound, socially acceptable framework, politicians will continue to support agriculture.

14

All four components combine to form the whole: sustainable agriculture. If one is neglected, it can seriously reduce the rate and extent of progress towards sustainability and food security. To attain sustainable food security we must change the way we plan, conduct, and communicate about research. At CIMMYT, we pursue an integrative research paradigm that assembles the best genotypes (G), in the right environments (E), under appropriate crop management (M), generating appropriate outcomes for people (P). Partnerships and consortia that assemble the best teams to execute the GXEXMXP approach are crucial to the timely and successful achievement of sustainable farming systems and future food security. Nowhere is this more important than in our research related to removing barriers to better plant nutrition and yields at the farm level. The following sections of this paper provide an overview of research conducted for two contrasting but extremely important settings in developing countries: 1) the rainfed maize production systems of southern Africa, where—apart from recurrent drought—low soil fertility has been identified as the single greatest constraint to food security, and 2) the irrigated, highly productive environments where the bulk of the developing world’s spring wheat is produced. Research in both of these settings addresses critical regional food security objectives, but the differences between the challenges and research opportunities in both settings serve to highlight the range of scientific disciplines, innovation, and partnerships upon which CIMMYT relies to fulfill its mission on behalf of developing countries. 3.5 Soil fertility strategies to sustain small-scale maize farmers in

Africa

The research setting Sub-Saharan Africa is the region of the developing world where the population of poor and undernourished people is growing most rapidly. In eastern and southern Africa, where CIMMYT concentrates its maize research, about one-quarter of a billion people get their subsistence and income directly from agriculture, and maize is their preferred staple. Per capita maize consumption surpasses 100 kg in several countries in the region (CIMMYT 1999). The vast majority of maize producers in sub-Saharan Africa have the farm size of 0.5-3.0 ha (Byerlee and Heisey 1997). Only 5% farmers grow maize commercially on holdings those exceed 50 ha. Most of these smallholders have no choice but to rely on extremely low-input, low-risk cropping systems. The urgency of producing enough maize to feed the family, even in drought years, means that scarce land and labour are devoted to maize production first and foremost, with little left over for other crops that could improve human nutrition and soil fertility. Throughout eastern and southern Africa, annual legumes are used as sole crops in rotation with maize, are intercropped, or are occasionally used as green manures. Maize is often grown in loose rotations with such crops as groundnut (Arachis hypogaea), bambara nut (Voandzeia subterranea), and occasionally soybean (Glycine max). Increasingly, however, maize is grown on the same land year after year as a sole crop or sparsely intercropped with beans, groundnuts, or cowpeas. Highly erratic rainfall and drought lead to tremendous variability in maize production across the region (Figure 2) and at the individual farm level. Infertile soils exacerbate the effects of insufficient rainfall. At the regional and national level, below-average maize production and associated cash constraints of farmers discourage the development of the agricultural input sector, particularly in more remote areas, and in some cases have made the credit systems of entire countries collapse. At the farm level, below-average maize production means a reduction in income, with concomitant negative effects on investment in fertilisers and other inputs in the next crop cycle (Zambezi and Mwambula 1997).

15

Crop improvement research for smallholder maize production Given the highly erratic rainfall pattern in southern and eastern Africa, smallholder farmers grow a disproportionally large part of their farm to their preferred staple, maize, to maintain houshold food security even in years of below average rainfall. Additionally, they are hesitant to invest in expensive fertilisers as dry spells may prevent an economic return on monies invested in fertiliser, or may even result in a net loss. CIMMYT considers maize improvement for drought and low soil fertility conditions to be one avenue for improving nutritional limits to maize and production in sub-Saharan Africa. By producing maize varieties with higher and more stable yields under conditions typical for resource-poor farmers (in other words, at the 1.2 t ha-1 yield level), smallholder farmers will have more monies available to invest in fertilisers. And, as household food security can be maintained on a smaller amount of land area, crop production is expected to become more diverse and include a larger proportion of legumes and cash crops. Compared to conventional breeding attempts that usually focused on raising yields under optimal, agronomically well-managed conditions, CIMMYT introduced in 1996 a very different breeding approach to sub-Saharan Africa in which maize genotypes are evaluated under carefully managed drought and N stress and the best genotypes selected, provided they respond similarly well to optimal conditions (Bänziger and Cooper 2001). This breeding approach is based on more than 12 years of strategic research at CIMMYT. Its principles, derived from the work of many researchers, are summarised in Bänziger et al. (2000a).

16

After four years, these breeding activities are delivering results. Among cultivars that had already been released and were available on the market, several cultivars of potential value to smallholders were identified. These cultivars yielded significantly better than others under drought and/or N stress, although this advantage, because it was unknown, had not been considered when the cultivars were originally released to farmers. In addition, researchers identified experimental maize germplasm that yielded 50-75% additional grain under drought and N stress, at a yield level of about 1-2 t ha-1 (Bänziger et al. 1999). One variety that particularly attracted the attention of breeders, agronomists, extension staff, non-governmental organisations (NGOs), and farmers was ZM521, an open-pollinated variety developed by the project. In trials from Ethiopia to South Africa, ZM521 out yielded the current releases on average by 34%, and showed impressive yield stability (Figure 3). In trials averaging 1-2 t ha-1, ZM521 yielded 2.2 t ha-1 compared to the local check cultivars, which yielded 1.4 t ha-1 on average, surpassing them by 50% at yield levels typical for smallholder farmers (Bänziger et al. 2000b). ZM521 is planted in on-farm trials across the SADC region, and commercial seed production has begun. Several studies have shown that these drought- and N-stress tolerant varieties and hybrids do not take up more water or nutrients (Bänziger et al. 1999; Bolaños and Edmeades 1993a, 1993b; Bolaños et al. 1993; Lafitte and Edmeades 1994a, 1994b, 1994c). They use water and nutrients more efficiently for grain production because they have a higher harvest index under conditions that usually reduce the harvest index to less than 0.2-0.3. For this reason, the new cultivars meet all the requirements for sustainably increasing maize yields and improving food and income security. They give smallholder farmers, for the first time, incentives to use improved management practices and diversify crop production. Some of these improved practices are discussed in the sections that follow.

17

Crop management research to overcome soil fertility constraints in Africa’s smallholder maize systems In the 1990s, soil infertility was widely identified as the most severe and widespread biophysical constraint to smallholder crop productivity and to long-term food security in eastern and southern Africa (Blackie 1994; Blake 1995; Kumwenda et al. 1996, 1997; Waddington and Heisey 1997; Sanchez et al. 1997). Soil infertility is a major cause of inefficiency in the returns to other inputs and management committed to smallholder farms, including N fertiliser (Mushayi et al. 1999).

Factors contributing to low soil fertility in Southern Africa The old and already highly leached soils of humid and sub-humid Africa have inherently low nutrient levels. In southern Africa, sandy and sandy loam soils derived from granite, with very low organic matter (<0.5%) and cation exchange capacity, are widespread. Deficiency of N is ubiquitous on these soil types, and deficiencies of phosphorus (P), sulfur (S), magnesium (Mg), and zinc (Zn) are common. Zambia and Mozambique have large areas of acidic soils with free aluminum (Al3+). Nutrient balance studies conducted from single farm to national scales have shown that nutrient depletion rates far exceed replenishment (Figure 4). The estimated annual net nutrient depletion exceeding 30 kg N ha-1 and 20 kg potassium (K) ha-1 of arable land in Ethiopia, Kenya, Malawi, Nigeria, Rwanda, and Zimbabwe (Stoorvogel and Smaling 1990; Smaling 1993; Stoorvogel, et al. 1993). Traditional African agricultural systems were largely based around extended fallows and the harvesting of nutrients stored in woody plants. In most arable areas fallowing has almost disappeared from the now-sedentary agricultural system, while in more sparsely populated areas the length of fallows continues to decline. In some areas, such as the wetter communal lands of northern Zimbabwe, soils hold so few nutrients that maize will yield virtually no grain without fertiliser. Soil organic matter (SOM) maintenance and management are central to the sustainability of soil fertility on smallholder farms in the tropics (Swift and Woomer 1993; Woomer et al. 1994). Soil organic matter helps retain mineral nutrients (N, S and micronutrients) in the soil and make them available to plants in small amounts over many years as SOM is mineralised. Soil organic matter increases soil flora and fauna (associated with soil aggregation, improved infiltration of water, and reduced soil erosion), complexes toxic aluminum (Al) and manganese (Mn) ions (leading to better rooting), increases the buffering capacity on low-activity clay soils, and increases water-holding capacity (Woomer et al. 1994). Current SOM inputs (for example, from fallows, tree leaf litter, cereal and legume crop residues, and animal manures) are insufficient to maintain SOM levels in most smallholders’ soils. In low-rainfall areas it is impossible to grow enough biomass to maintain SOM.

18

Heisey and Mwangi (1996), in reviewing the use of inorganic fertiliser in sub-Saharan Africa, found that farmers applied fertiliser nutrients less than 10 kg ha-1 year-1 of arable land while the corresponding rate for the developing world as a whole is 83 kg. They compared the years that aggregate fertiliser applications reached 10 kg ha-1 in various countries, and African countries generally reached this level later (Table 1). Fertiliser is more expensive in southern Africa than in most other parts of the developing world (Waddington and Heisey 1997). In fact, fertiliser is the most costly cash input used by the typical small-scale farmer in southern Africa, and few farmers can afford it. Fertiliser subsidies have long been a thing of the past, farmers have a very hard time obtaining credit to purchase fertiliser or other inputs, and as greater numbers of rural people join the ranks of the impoverished, their ability to purchase inputs diminishes accordingly. Because rainfall is so unreliable, even farmers who do manage to obtain fertiliser cannot be sure when to apply it and whether they will get the benefit for their maize crop at all. A little over one-third of the regional maize crop receives some chemical fertiliser (Heisey and Mwangi 1996). This is well below crop and soil maintenance requirements and is likely to remain so. The most profitable fertiliser recommendation for commercial maize production in most parts of Malawi in 1995-98 was to apply no fertiliser at all or just 35 kg N ha-1 (Benson 1998). Part of the reason for low profitability is that the efficiency of fertiliser use on-farm, as measured by the grain yield response to the addition of chemical N and P fertilisers, is often poor (Jones and Wendt 1995). For example, in Malawi and Zambia, farmers using current fertiliser practices can expect just 9.5-16 kg of maize grain per kilogram of N applied when growing local traditional maize, and 17-19 kg of maize grain per kilogram of N applied when using hybrids. In many cases, other nutrient (Zn, S, Mg, and B) deficiencies reduce the response to N. The situation may be worse on shallow, sandy soils in Zimbabwe that are prone to waterlogging. On these soils, N-use efficiencies are below 10 kg of maize grain per kilogram of N applied to many smallholder maize crops (Mushayi et al. 1999)

19

Integrated soil fertility improvement How then can we maintain (and perhaps in some cases build up) soil fertility under the income and, increasingly, land and labour constraints faced by smallholders? The soil fertility challenge in the region is now so large, the solutions so scarce, and individual solutions largely ineffective by themselves, that farmers need to integrate several soil fertility options, ranging from sources on farm to external inputs, on different parts of their farms (Kumwenda et al. 1996; Waddington and Heisey 1997). These options include modest chemical fertiliser inputs, organic inputs (particularly organic matter and biological N fixation from legumes), and N-use efficient maize cultivars (described earlier). Over the last decade, researchers have identified a number of soil fertility management technologies that smallholders can use to maintain, and in some cases build up, soil fertility, despite income, land, and labour constraints. In this paper, we focus on options developed through a major initiative in southern Africa, called the Soil Fertility Research and Extension Network (“Soil Fert Net”) for Maize-based farming Systems in Malawi and Zimbabwe. Soil Fert Net is coordinated from CIMMYT’s regional office in Zimbabwe. Table 1. The year when aggregate fertiliser application rate reached 10 kg ha-1 (NPK) in various countries

Country Year Country Year Argentina 1993 Mexico 1964 Brazil 1967 Nepal 1983 China 1958? Nigeria 1969 Colombia before 1961 Pakistan 1968 Côte d’Ivoire 1972 Paraguay 1993 Ecuador 1967 Peru before 1961 Ethiopia 1993 Philippines before 1961 Ghana Not reached by 1996 South Africa before 1961 Guatemala before 1961 Tanzania 1974 Honduras 1965 Thailand 1972 India 1968 USA 19 45 Indonesia 1968 Venezuela 1968 Japan before 1880 Vietnam before 1961 Kenya 1969 Zambia 1971 Malawi 1971 Zimbabwe before 1961

Source: Heisey and Mwangi (1996). Hectarage is calculated as total of “arable land and permanent crops.”

“Best Bet” soil fertility technologies One of Soil Fert Net’s chief aims has been to develop a range of organic and inorganic soil fertility management technology options or “best bets” for smallholders (Waddington et al. 1998) through widespread participatory research and testing with farmers. A “best bet” technology makes a longer-term contribution in raising soil fertility and crop yields, and generating profit in the short term. It is appropriate for many farmers across important agroecologies; compatible with other components of the farming system; has small additional cash and/or labour requirements; involves only a small reduction in maize yields or substitution by production of another crop; and, where possible, involves little competition for arable land. Most technologies that meet these criteria provide some short-term soil fertility and crop productivity benefits and have several end uses, all of which make them attractive to farmers. They are compatible with farmers’ circumstances and effective within farmers’ resource constraints. For these reasons, the technologies that meet the criteria listed above offer farmers the “best bets” for improved productivity, sustainability, useful products, and income.

20

An assessment of Soil Fert Net’s “best bet” soil fertility technologies is provided in Table 2. By the end of 1999, 18 technologies were considered ready for promotion (Table 2). Some of these technologies are described briefly in the sections that follow. Table 2a. Adoption potential of Soil Fert Net’s “best bet” soil fertility technologies for smallholder maize-based farming systems in Malawi Technology Farm type Adoptiona Adoptionb

(number of farmers)

Ease of 2000 Potential

Area-specific fertiliser recommendation for maize

All areas by soil type and richer farmers’ market or home use

++ 60,000 900,000

Pigeonpea + maize intercropping

South and Central Malawi smaller holdings

++++ 500,000

1,000,000

“Magoye” promiscuous soybean

All mid-eleven areas richer cash croppers

++ 25,000 300,000

Groundnut in rotation with maize holdings and pigeonpea intercropped with other grain legumes

All mid-eleven areas (medium to large)

+++ 60,000

400,000

Tephrosia undersowing of maize

Mid-eleven and medium to large holdings in lakeshore areas

++ (?)200 400,000

Macuna + maize rotations Most of Malawi; poorer soils, medium to larger holdings

+ (?)100

200,000

Faidherbia albida trees in crop land

(500-1000 masl) ++ (?)10,000

500,000

Sesbania undersowing Mid-eleven areas, larger holdings

+ (?)1,000 100,000

Optimum combinations of organic and mineral fertilisers

Most of Malawai ++ ? 1,000,000

Soil fertility x Striga interactions

Striga-affected areas +++ ? 150,000

a+ = low, ++ = moderate, +++ = extremely high. bFrom key informants. Estimated by the Soil Fert Net coordinator, using information from Soil Fert Net members, with amendments from K. Giller. Late-maturing pigeonpea intercropped with maize (Malawi) In southern Malawi, where land is scarce and human population is high, many farmers intercrop grain legumes with maize to produce more food and to maintain soil fertility. Late-maturing pigeonpea is especially promising, because pigeonpea competes very little with the maize crop and matures on residual moisture after maize is harvested. Late-maturing pigeonpea intercropped with maize can often produce a dry matter yield of 3 t ha-1 from leaf litter and flowers. Even if the seed is harvested for food, the leaf fall is sufficient for N accumulation. Malawi’s Maize Commodity Team has identified ICP9145 pigeonpea, a long-duration variety that resists Fusarium wilt and yields better than local varieties, as the best variety, and has found that the most economic planting pattern consists of placing one pigeonpea station (with three seeds hill-1) between maize stations (37,000 plants ha-1). One disadvantage with this technology is that pigeonpea is highly attractive to livestock, which can pose problems in communities where livestock are free ranging.

21

Area-specific fertiliser recommendations (Malawi) These recommendations have greatly improved the regional efficiency of fertiliser use on maize within Malawi. Until the late 1990s, Malawi had one blanket—and generally noneconomic—fertiliser recommendation for maize, but in a major effort (principally by the Maize Commodity Team), area-specific fertiliser recommendations were developed. Missing nutrient trials and widespread chemical analyses of soil showed regional deficiencies of Zn, S, B, and K. In deficient regions, average yields improved by 40% over the existing N and P application when the deficiencies were satisfied. New basal fertiliser blends with these nutrients were developed with fertiliser suppliers. Area-specific fertiliser recommendations were verified at well over 2,000 on-farm sites throughout Malawi with the extension service and farmers. Through economic analysis, GIS maps, and decision trees, economic area-specific fertiliser recommendations were developed based on soil texture and farmers’ production goals. The new recommendations are usually 35 kg N ha-1 if the maize was for market sale and either 69 or 92 kg N ha-1 if for home consumption . Sole crop green manure (Malawi and Zimbabwe) Compared to other green manure crops, in Malawi and Zimbabwe velvet bean (Mucuna pruriens) proved to be the most consistent producer of biomass and provider of N, even on some of the most depleted soils where its contribution is most needed. Additionally the grain is used as a food crop of last resort in southern Malawi and parts of northern Mozambique. In Malawi, velvet bean averaged over 7 t ha-1 of aboveground biomass when grown with P at some relatively depleted sites (Kumwenda and Gilbert 1998). On more depleted sandy soils in Zimbabwe, velvet bean routinely produced over 2 t ha-1 of above ground biomass and even over 5 t ha-1 in many cases, performing far better than other green manures (Hikwa et al. 1998). When velvet bean residues are incorporated early, at flowering, maize grain yields are higher than when residues are incorporated late, after seed harvest. Because farmers may find it difficult to adopt a sole crop as green manure, researchers are looking out for other uses of velvet bean to improve its attractiveness to farmers. In some areas of Zimbabwe, for example, farmers are as interested in velvet bean for feeding cattle and goats as for feeding the soil (Bellon et al. 2000). If plant breeders successfully develop velvet bean cultivars with lower concentrations of L-Dopa in the seed, this legume could be converted into a new, high-grain-yield legume that would be robust under the poor biophysical conditions common on smallholders’ farms. Biomass transfer: the example of Tithonia (Malawi, Zimbabwe and Zambia) Farmers have many options with biomass transfer, ranging from the collection and application of leaf litter from indigenous trees found in Miombo woodland through to the deliberate planting and pruning of leguminous shrubs for transfer to crops. One relatively new example of biomass transfer that has recently attracted attention from researchers and farmers is the use of Tithonia shrubs found on field boundaries, on the edges of dambos, and as a weed in cropland. Tithonia is not a legume, but its shoot biomass has a high nutrient content (4% N and K; 0.5% P). When applied to growing maize crops, it mineralises quickly and increases maize yields. In Malawi 1.5 t ha-1 of Tithonia leaves more than doubled maize grain yield (Ganunga et al. 1998). In Zimbabwe, Jiri and Waddington (1998) harvested large amounts of fresh prunings of T. diversifolia and T. rotundifolia sufficiently early in the crop season to apply to maize crops, allow the residues to mineralise, and raise maize yield. Four tonnes per hectare of leaf prunings applied four weeks after maize emergence (T1) almost tripled maize yields from 0.86 t ha-1 to over 2.3 t ha-1, but labour costs for pruning and application were high. Additionally, combining rock phosphate with Tithonia can help make P available to crops and improve the yield of maize on acidic soils in northern Zambia (Malama 2000). Thus Tithonia can be a useful additional organic fertiliser for farmers where it is available. It propagates easily by cuttings or seed and is now being recommended to farmers to plant on niches such as field boundaries. More economic fertiliser use (Zimbabwe) The erratic and uneven distribution of rainfall in Zimbabwe makes fertiliser use by smallholder farmers highly risky. Over the last decade, research has focused on developing practical methods of applying split doses of N fertiliser, depending on prevailing rainfall, to optimize the economic

22

efficiency of fertiliser use. Earlier works indicated that the profitability of fertiliser was significantly increased by adding P, K, and S as a basal dressing and adjusting N fertiliser top-dressings to the evolving rainfall pattern in any given season. Trials over five years on farmers’ fields gave 25-42% more yield and 21-41% more profit than current fertiliser recommendations, which were also found to be too risky for lower rainfall areas and too high to be profitable. Intensive farmer training and an input loan scheme have been used to promote the new fertiliser recommendations for maize. In both subhumid and semiarid areas, the conditional fertiliser recommendation gave more than 100% additional yield and profit in many cases, and farmers’ loan repayment rate was above 90%. Zimbabwe is building on the success of this package and expanding to more farmer groups in collaboration with AGRITEX, the Zimbabwe Farmers Union, and a smallholder credit union. A major element of this activity involves financial management and marketing. Table 2b. Adoption potential of Soil Fert Net’s “best bet” soil fertility technologies for smallholder maize-based farming systems in Zimbabwe Technology Farm type Adoptiona Adoptionb

(number of farmers)

Ease of 2000 Potential

Fertiliser management package for maize areas (conditional on rainfall) and grain legumes

Subhumid and semiarid, all except poorest farms in driest areas

+++ 1,000 1,000,000

Liming on acidic sandy soils areas

Acidic soils in subhumid, higher input farms

++ 20,000 300,000

Soybean (inoculated and promiscuous) in rotation with maize

Subhumid areas, on cash crop farmers’ better soils

+++ 10,000 300,000

Other grain legume rotations

Subhumid and wetter semi-arid areas

+++ (?)80,000 700,000

Improved cattle management including anaerobic composting

All except driest areas. Farmers with cattle where they are reluctant to use manure

+ (?)500 250,000

Pigeonpea rotations and intercropping

Subhumid areas ++ ? 150,000

Optimum combinations of organic and mineral fertilisers

Subhumid and wetter semiarid areas

++ ? 6000,000

Mucuna + maize rotations

Subhumid areas + ?100 100,000

a+ = low, ++ = moderate, +++ = extremely high. bFrom key informants. Estimated by the Soil Fert Net coordinator, using information from Soil Fert Net members, with amendments from K. Giller. Smallholder soybean systems (Zimbabwe) Some of the more widely used grain legumes in Zimbabwe, such as groundnut, performed poorly on smallholder farms, prompting researchers and farmers to look for alternatives. Soybean is a versatile, multi-use grain legume, but few smallholders had experience growing it, and inoculation with Rhizobium was needed for N fixation and high yields. Inoculants were available and cheap but complicated the management of the crop.

23

Promiscuous soybean varieties such as Magoye, developed in Zambia, fix N with a wider range of Rhizobia often found in smallholders’ fields. Naturally nodulating types have a large aboveground biomass, a lower grain and N harvest index, and a more indeterminate development pattern than most specifically nodulating types. Experiments on smallholder farms at many sites in Zimbabwe confirmed that naturally nodulating types nodulate abundantly while maintaining substantial grain yields (about 1 t ha-1) similar to those of specifically nodulating types. Their larger biomass results in greater residual soil fertility benefits. Initial results indicate that Magoye is a net contributor of N to the soil in Zimbabwe. Packages with fertiliser, lime, and rhizobia can be managed by farmers and are very economic. Since around 1996, a major soybean promotion drive among smallholders has involved a task force from AGRITEX, the University of Zimbabwe, the Zimbabwe Farmers’ Union, and private input suppliers and grain processors (Pompi et al. 1998). Beginning with 55 farmers in 1996-97, in 1999-2000 this effort included over 10,000 smallholders growing soybean on 4,000 ha and marketing 4,000 t. Even at average grain yields of 800 kg ha-1 in smallholders’ fields, soybean is still economic, offering better returns to cash and labour than maize. Liming (Zimbabwe) About 70% of the arable land in Zimbabwe’s communal areas, where most small-scale farmers are located, have been shown to have a pH of less than 4.5. In many of these soils, the Al saturation exceeds 20% of the cation exchange capacity. Soil acidity is a major constraint in smallholder areas, reducing the growth and yield of legumes and reducing the response of maize to N fertiliser. Declines in cattle manure inputs, removal of crop residues and increased use of chemical N fertiliser in the 1980s have worsened the problem (Dhliwayo et al. 1999). Yields are diminished by toxicities of Al and Mn and by deficiencies of Mg and occasionally Ca. There is good evidence that liming to raise pH is an important “priming” input, making it possible for farmers to benefit fully from other soil fertility inputs and helping to increase the adoption of legumes. Agricultural lime is recommended to be applied based on pH measurement (rather than Al titration) to raise soil pH to between 5.0 and 5.2, because over-liming can lead to micronutrient deficiencies and lower yields. In widespread on-farm experiments, lime raised maize grain yields by between 0.6 and 2.6 t ha-1. Liming in combination with applications of chemical fertiliser and cattle manure is especially effective in raising crop yields and raising fertiliser-use efficiencies.

Adoption issues The potential impact of “best bet” technologies is large (Table 2), but much work remains to be done to foster their adoption and enable farmers to capture their considerable benefits. Many farmers face a conflict between the short-term need to ensure today’s food supply and the long-term need to ensure that soil fertility is adequate to meet tomorrow’s food requirements. Farmers discount the value of benefits that will be realised some time after the initial investment in a soil fertility technology is made. Legume systems that are the best soil improvers (e.g., hedgerow intercrops, green manures, and improved fallows) tend to have few other uses and to occupy land, with the result that they are less likely to be adopted (Figure 5) without significant support. Broadly speaking, the larger the potential soil benefit from a legume technology is likely to be, the larger the initial investment required in labour and land, and the fewer short-term benefits it is likely to have. To date farmers have tended to adopt current (rather than entirely new) practices that are being more widely promoted or improved upon, notably pigeonpea + maize intercrops in southern Malawi and groundnut + maize rotations in northern Zimbabwe. The sections that follow discuss strategies for further encouraging the adoption of soil fertility technologies under development in the region. Promoting adoption through partnerships that support farmers The “best bet” technologies are being widely promoted through a range of partnerships with government extension services, farmer groups, and NGOs (Non Government Organisations) in Malawi and Zimbabwe. Working with these partners allows widespread coverage and can improve the

24

efficiency with which technologies are promoted. Many partners use farmer participatory approaches to test and modify the technologies. In Chihota Communal Area of Zimbabwe, for example, extensionists work with farmer groups to expose about 4,000 farmers to liming, grain legume + maize rotations, and green manure technologies through hundreds of participatory on-farm demonstrations and experiments (1998-2001). After group and baseline surveys were done, extension workers and farmer groups received training on the technologies and on conducting the demonstrations. Some 60 field days and mid-season and end-of-season evaluations have given farmers, extensionists, and researchers opportunities to assess the demonstrated technologies. Farmers have tested these technologies, incorporated them into their cropping systems, and provided feedback through group and individual interviews. Aside from group experimentation and learning, the extension approach emphasises farmer-to-farmer transfer of information, using songs and drama to distribute messages. End-of-season workshops enabled everyone to discuss results of the demonstrations, especially farmers’ thoughts about the various technologies, their adoption, and modification (Gambara et al. 2000).

The experience with improved two-year fallows using the leguminous trees Sesbania sesban and Tephrosia vogelii in Zambia’s Eastern Province provides another example of how farmers will take up soil fertility technologies if given sufficient support. These have been the subject of a major research and promotion drive led by one of CIMMYT’s sister institutes, the International Centre for Research in Agro-Forestry (ICRAF), with Zambian Government partners and several NGOs. The improved fallows tripled grain yield in subsequent maize crops (Kwesiga and Coe 1994) and had economically attractive returns to labour (Kwesiga 1998), but they appeared to have characteristics that might discourage adoption: farmers had to forgo a maize crop for one to two years and often had to nursery-rear, plant, and weed the trees (a non-food crop). The number of farmers testing the technology rose to over 4,000 during 1994-97, however, and has expanded since then. Keswiga (1998) has attributed this success to a combination of factors, including careful agroecological targeting to relatively land- and rainfall-abundant areas with N-deficient and -responsive soils and a history of bush fallows; major international and national research input; identifying and addressing constraints to farmer use; and training and

25

mobilization of extension and NGOs to demonstrate the technology to farmers, provide information and inputs, and encourage farmer experimenation, modification, and farmer-to-farmer transfer.

Promoting adoption through national task forces Larger scale national commodity task forces (on maize in Malawi and soybean in Zimbabwe) have proven highly effective in focusing awareness, resources, and partnerships onto national initiatives to address soil fertility issues and to disseminate some of the “best bets.” In the late 1990s, the research and extension services within the Malawi Maize Productivity Task Force mounted thousands of on-farm demonstrations and provided thousands of brochures with area-specific fertiliser recommendations for maize and the use of legumes throughout the country (Kumwenda 1998). The task force helped the extension service to accept the area-specific fertiliser recommendations in 1997 and encouraged the policy implications to be assessed with the government. These more flexible recommendations are now promoted nationwide. Soil Fert Net members within the Maize Task Force provided technical input on expected benefits from the technologies and helped develop input support strategies for a nationwide initiative to give fertiliser, maize, and legume seed starter packs (Mann 1998) to all 1.8 million smallholder households in Malawi during the 1998-99 and 1999-2000 cropping seasons. Collectively the Government of Malawi, Department for International Development (UK), European Union, and the World Bank provided over US$23 million to this program in 1998-99. It has had a major impact on human nutrition and household food security in Malawi and is an excellent example of where technical scientists have influenced Government and donor policy. Soil Fert Net has sought to duplicate that success in Zimbabwe by helping to establish the Agro-Natural Resources (Soil Fertility) Council in 1999 as part of the new Zimbabwe Agricultural Productivity and Food Security Task Force. As noted, Soil Fert Net members have been the leaders in a smallholder Soybean Promotion Task Force in Zimbabwe that over four years has led 10,000 smallholder farmers to adopt soybean. The development and support of input and output markets and home utilisation through the Soybean Promotion Task Force has been key to its success. The private sector (including seed and fertiliser companies) has been active in supplying soybean seed, fertiliser, and lime to smallholder areas. Produce markets were assured through Olivine Industries, a major oil processor, which agreed to take a quota from smallholder areas and helped collect the grain through local traders. Farmers obtained a better price than through the normal sales channel (the Grain Marketing Board). A recent study has concluded, however, that rural distribution and assembly costs need to be reduced, communication between partners needs to be improved, and local traders’ access to capital needs to be strengthened (Rusike et al. 2000).

Policy and Economics Support As the successful promotion of soil fertility technologies indicates, researchers and their partners need to provide sound technical information to guide policy decisions related to smallholders’ adoption of soil fertility technologies, including such inputs as lime, fertiliser, and legume seed. Soil Fert Net is developing information and advocacy strategies to further support adoption. A new Economics and Policy Working Group will provide: • A framework for closer interaction among soil fertility experts, economists, extensionists, and

policymakers on methods of solving soil fertility problems through increased participation of all stakeholders, including farmers.

• Objective economic evaluations of “best bet” technologies. • Priority setting and targeting of potential “best bets” for smallholder farmers. • Policy research and advocacy that help create a policy environment that enables farmers to use

improved soil fertility management technologies. • Strategic and relevant partnerships to scale up the work of Soil Fert Net, enabling technologies to

reach larger numbers of smallholders.

26

Several multidisciplinary studies are under development to learn more about off- and on-farm views and constraints to adoption of soil fertility technologies and help make better preparations for adoption. 3.6 Improving efficiency and reducing environmental

consequences of N use in irrigated spring wheat

The research setting Wheat is the most widely consumed staple food in the world, and its role in maintaining food security in developing countries cannot be understated. Compared to other wheat-producing regions in the developing world, in the irrigated spring bread wheat areas wheat yield potential and levels of input use are high. For several reasons, CIMMYT gives high priority to these irrigated wheat production environments, which include the Indo-Gangetic Plains of South Asia, the Nile Valley in Egypt, and the Yaqui Valley in northwestern Mexico. First, these areas are estimated to produce 42% of the developing world wheat crop (Rajaram et al. 1995). Second, South Asia, where a great deal of the world’s spring wheat is produced, has 73% of the world’s poor people and will continue to be the home of half of the world’s poor well into the new millennium (World Bank 1997). Third, malnutrition and hunger are still extremely serious problems for a large segment of society in irrigated wheat-producing areas, and population continues to grow. Finally, if these important production environments fail to keep pace with growing food demand, the implications for many countries and for hundreds of millions of poor people will be serious. Pingali and Rajaram (1999), in a review of trends in irrigated wheat production environments, have noted growing concern over the future of wheat production in these areas. Newly released varieties still offer higher yields than older releases, but yield potential has grown at a slower rate in the most recent decade (Sayre 1996). The economically exploitable gap between potential yields and yields achieved on farmers’ fields has diminished, so the cost of marginal increments in yield, given existing technologies and policies, may in some cases exceed the incremental gain. The cost is high in terms of increased input use (e.g., fertiliser, fuel, water) and in terms of increased management time to improve the efficiency of input use (including fertiliser) (Pingali and Heisey 1996; Byerlee 1996). These concerns pose a major challenge for research, especially research on N-use efficiency. Wheat production in many irrigated spring wheat areas is presently dependent on synthetic N fertiliser for a number of reasons. First, the traditional use of animal manure to improve soil fertility is increasingly limited. Manure may be applied to other, higher value crops before it is applied to wheat, or it may not be applied to any crops at all because it is in such demand for use as fuel. Animal numbers have declined in some areas as tractor use and the expense of keeping animals has risen (Hobbs et al. 1998; Fujisaka et al. 1994). Second, many of the soils in irrigated wheat areas have naturally low levels of SOM, and there is evidence that SOM continues to decrease as a result of increased tillage and the burning and/or removal of crop residues for animals (Meisner et al. 1992; Hobbs and Morris 1996). Finally, few legumes that can supply symbiotically fixed N are present in the main wheat rotations (rice-wheat, cotton-wheat, maize-wheat, soybean-wheat, and sorghum-wheat) in irrigated wheat areas. In the developing world’s irrigated wheat production systems, N deficiency is the most widespread nutritional problem, and N fertiliser recovery tends to be low. For example, in Pakistan N recovery was estimated at about 30% (Byerlee and Siddiq 1994), and in Mexico the estimates for the Yaqui Valley are less than 50% under N management similar to farmers’ management (Ortiz-Monasterio et al. 1994). Under current agronomic practices, further intensification could lead to higher inefficiencies and therefore higher N losses. It is important to emphasise that the N that is lost in these systems, in addition to being an expense to farmers and consumers, also has an environmental cost. Land conversion and intensification alter the biotic interaction and patterns of resource availability in ecosystems and can have serious local, regional, and global environmental

27

consequences (Matson et al. 1997). Therefore it is extremely important to identify N management practices and N-use efficient cultivars that permit us to meet the increasing demand for food and fiber, minimise negative environmental impacts, and are economically attractive to farmers.

Improved N-use efficiency through plant breeding When semidwarf spring bread wheats were first adopted in developing countries, starting in irrigated areas in the mid-1960s (Byerlee and Moya 1993; Byerlee 1996), some early observers of the adoption process contended that the new wheats could not perform well in the absence of N fertiliser (Simmonds 1979) and that farmers would be better off growing their old tall cultivars if no fertiliser was available. Since then it has become abundantly clear that this is not the case. In Mexico, Ortiz-Monasterio et al. (1997) found that semidwarf spring wheats developed by CIMMYT yielded better than old tall cultivars under high or low N conditions. Other studies in other countries have confirmed that semidwarf wheats either yielded the same or more than old tall cultivars under low N fertility conditions (Jain et al. 1975; Wall et al. 1984; Entz and Fowler 1989; Austin et al. 1993). Semidwarf wheat cultivars do not “require” more N; in fact, they often need less N to produce the same yield per unit of available N than old tall cultivars. The misconception may have evolved because semidwarf wheats respond better to N and therefore have a higher optimum economic rate (Figure 6). Moll et al. (1982) have subdivided N-use efficiency (grain yield/N supplied) in cultivar development into two components: 1) uptake efficiency (plant total N/N supplied), or the plant’s ability to extract N from the soil, and 2) utilisation efficiency (grain yield/plant total N), or the plant’s capacity to convert the N it has

absorbed into grain yield. An important condition for breeding N-use efficient wheats, of course, is the presence of genetic diversity for that trait, and such diversity has been reported (Dhugga and Waines 1989; van Sanford and MacKown 1986; Ortiz-Monasterio et al. 1997). Breeding methodology also plays an extremely important role. The performance of CIMMYT spring wheats developed from 1950 to 1985 (under medium to high N fertility) improved when these same cultivars were grown under high and low N fertility. Increased grain yield appears to be associated with gains in uptake and utilisation efficiency at medium to high N fertility levels and with uptake efficiency alone under low N fertility levels (Ortiz-Monasterio et al. 1997). These findings suggest that the level of soil N plays a very important role in the genetic expression of uptake and utilisation efficiency in wheat (at low N levels, there is a better expression of uptake efficiency, while at high N levels, utilisation is better expressed). In theory, the N level in the soil could be manipulated together with the genetic diversity of the crop to breed wheats with improved uptake and/or utilisation efficiency.

28

Breeding for high yield potential in environments where N is not limiting has produced germplasm that demonstrates improved N-use efficiency under low or high N fertility in irrigated areas. It remains to be seen whether other breeding strategies (considering low N, or alternating low- and high-N environments during the selection of segregating populations and grain yield evaluations) could result in germplasm with higher N-use efficiency. Improved lodging tolerance was a primary characteristic of the first semidwarf wheat cultivars. In some irrigated areas, however, grain yields have risen to the point at which lodging may be keeping the levels of N fertiliser below the agronomic optimum (Hobbs et al. 1998). Further progress in lodging tolerance could be fundamental for breeding new cultivars with higher N-use efficiency under high fertility conditions.

Improved N-use efficiency through crop management research Doerge et al. (1991) have shown that N uptake in irrigated spring wheat proceeds slowly until tillering and that the N flux (kg N ha-1 day-1) increases to reach a maximum during the jointing stage. These findings point to the beginning of stem elongation—Zadoks 31 (Zadoks et al. 1974) or Feekes 6 (Large 1954)—as the initiation of rapid N uptake by the wheat crop. Researchers at CIMMYT and other institutions have evaluated several strategies for ensuring that high N demand is matched to N availability (which occurs several weeks after planting). These strategies are discussed in the sections that follow. Before moving on to that discussion, however, it is important to emphasise that decisions on the timing of N applications must also consider the planting method, equipment available to the farmer, and irrigation management. For example, N applications should coincide with irrigation to insure incorporation and availability of N to the plant. In the bed planting systems under development by CIMMYT and its partners, field access is easy around Z31, allowing N applications. Cultivation for weed control can be done at the same time to reduce the number of field operations. Finally, on small farms such as those in some of the irrigated areas of the Asian Subcontinent, farmers can walk into the field and broadcast fertiliser at the critical stages.

Delayed N applications Delayed N applications (in which all the N is applied close to Z31, rather than at planting, which is the common farmer practice) have produced very interesting results. Even when the wheat crop has been severely N stressed early in the crop cycle, breaking the stress with a delayed N application by Z31 results in higher N recoveries, which often translate into higher yields and consistently produce higher protein concentration in the grain (Fischer et al. 1993; Ortiz-Monasterio et al. 1994), less

29

lodging (Hobbs et al. 1998), and reduced incidence of Karnal bunt disease (Neovossia indica) (Ortiz-Monasterio et al. 1993). A recent worldwide study of ten countries by the International Atomic Energy Agency (IAEA) in collabouration with CIMMYT and International Fertiliser Development Centre (IFDC) evaluated the effect of N timing application over five years using 15N tracer techniques (IAEA 1999). The study compared Z31 N applications in irrigated wheat systems with applications at planting and concluded that in nine of the ten countries N recoveries were higher at Z31 stage than at planting.

Split N applications Split applications in which some N is applied at planting and most is applied at Z31 have generally resulted in higher yields than applications of all N at planting or Z31. Although N applications are more efficient at Z31 than at planting, it appears that some N needs to be supplied early in the crop cycle, particularly in soils that are highly N deficient. A study over four environments compared N application at planting, delayed application at Z31 (all N was applied at that stage), and a split application (one-third of the total rate was applied at planting and the remainder at Z31). The split application resulted in higher grain yields and higher apparent N recovery than the other two treatments. Furthermore, a three-way split application, in which one-third of total N was applied at planting, another third at Z31, and another third at Z37 (flag leaf just visible, 54 days after emergence in the Yaqui Valley) delivered the same grain yield as the treatment in which one-third of the N application was provided at planting and two-thirds at Z31 (Ortiz-Monasterio et al. 1994). These experiments were done in heavy clay soils; it is expected that in lighter textured soils, with potentially higher leaching problems, the three- or four-way split could be more efficient than the two-way split application.

Environmental aspects of N use in irrigated wheat production in developing countries Globally, agricultural activities have had a major impact on the N cycle. Production of N fertiliser and planting of leguminous crops fix more N globally than all natural ecosystems (Vitousek and Matson 1993). The use of N fertiliser in irrigated cropping systems can result in nitrate leaching and reduced water quality (CAST 1985; Keeney 1982). High nitrate concentrations in drinking water represent a human health concern, causing methamoglobinemia (Aldrich 1980). Nitrate also influences the health of natural systems. Eutrophication of estuaries and other coastal marine environments can cause low or no oxygen conditions in stratified waters, leading to losses of fish and shellfish and to blooms of nuisance algae and organisms that are toxic to fish (Howarth et al. 1996). The use of N fertiliser also results in the emission of N gases to the atmosphere, including nitrous oxide (N2O), nitric oxide (NO), ammonia (NH3), and diN (N2). The first three gases have a negative impact on the environment. Nitrous oxide has a global impact, absorbing infrared radiation and contributing to greenhouse warming and the depletion of the stratospheric ozone layer (Granli and Bockman 1994). The effects of nitric oxide occur on a more regional scale. This gas reacts to form tropospheric ozone, a major atmospheric pollutant that affects human health, agricultural crops, and natural ecosystems. Chameides et al. (1994) have suggested that as much as 35% of cereal crops worldwide may be exposed to damaging levels of ozone. Nitric oxide is a precursor to nitric acid, a principal component of acid deposition. Together with ammonia, which is also emitted from agricultural systems, nitric oxide may be transported and deposited in gaseous form or in solution to terrestrial and aquatic ecosystems down-wind (Matson et al. 1997). This deposition constitutes, in effect, inadvertent fertilisation, and can lead to acidification, eutrophication, shifts in species diversity, and effects on predator and parasite systems (Vitousek et al. 1997). Recent research has shown that farmers’ typical, intensively managed, irrigated wheat production systems in the Yaqui Valley have led to extremely high fluxes of nitrous oxide and nitric oxide

30

compared to practices in other agricultural areas. Alternative practices that changed the rate and timing of N application reduced emissions of these two gases by more than 50% without reducing grain yield or quality (Matson et al. 1998; Ortiz-Monasterio et al. 1996).

Some final thoughts The case of nitrogen management in the Yaqui Valley, described above, is instructive because trends in the Yaqui Valley’s intensive wheat production systems have often heralded similar trends in other areas of the developing world. Further research is essential if developing countries are to attain the means of reconciling the twin goals of increased food production and greater protection of the environment (Matson et al. 1997). Given that most irrigated wheat production systems in developing countries are highly dependent on chemical fertilisers, it is absolutely vital to explore more integrated approaches to nutrient management. Some management strategies for improving N-use efficiency in irrigated wheat areas can be transferred readily to different regions of the world, such as practices that permit better synchronisation of N supply and demand. Other practices will require local calibration. The fact that blanket N fertiliser recommendations are still customary in many wheat production systems suggests that there are good opportunities for improving N-use efficiency by identifying proper N rates at the individual field level. Refinements of blanket recommendations would take into account previous management, soil types and crop variety, so that instead of having one recommendation for one region, three to five recommendations could be developed to take soil type, crop variety and/or rotation into account. A further level of refinement could be achieved through the use of appropriate soil and plant tissue diagnostic tests at critical stages of plant development. Information from these tests would allow the N recommendations to be developed for site- and season-specific conditions. It would also significantly improve N-use efficiency if used in conjunction with information on appropriate timing, N source, and fertiliser placement. Finally, it is important to remember the role of policy in providing incentives or disincentives for efficient production practices, as noted by Pingali and Rajaram (1999). Farmers’ adoption of more efficient N application practices for wheat will depend on the ratio of the wheat price to the N price. When this ratio becomes smaller because wheat prices fall or N prices rise, the incentive for adopting efficiency-enhancing technologies increases. Some analysts believe that liberalisation of the agricultural sector in developing countries could increase the adoption of N efficiency technologies. In any case, communication between research and policymakers is essential to reach the goals of nutrient use efficiency, food security, and environmental security. 3.7 Conclusions

Given that a large majority of developing nations will continue to depend on domestic maize and wheat production to meet their growing demand for food and feed in the foreseeable future, the kinds of research described in this paper can be expected to grow rather than diminish in importance. Considerable ingenuity will be required to meet the challenge of devising more sustainable and productive cropping systems with farmers. At present, very little information is available about the consequences of N use in developing countries, even though it is projected that some developing countries will use more N fertiliser in the future. For this reason, researchers need to be particularly vigilant about the potential environmental consequences of practices being developed to increase productivity. This paper has emphasised very clearly that research for developing country settings must never lose sight of the fact that farmers are working with extremely limited resources, under burdensome constraints that are not seen in high-income countries. The days when development specialists naively believed that technology could be imported wholesale into developing countries and fit

31

smoothly into local conditions are over. Our work in Africa, for example, emphasises that researchers must work carefully with a diverse range of partners—including farmers, NGOs, local schools, private companies, policymakers, and others—to devise locally appropriate solutions. Here it is important to note that “locally appropriate” should never be interpreted as “second best.” On the contrary, CIMMYT researchers have a critical role to play in investigating the very best new technology available anywhere in the world and working with local partners to determine the potential value of that technology for resource-poor farmers. To help create a policy environment that enables farmers to use improved soil fertility management technologies, policy research and advocacy will also be essential. The research in Africa, as well as the research for irrigated wheat systems, also points to developing countries’ great need for more integrated nutrient management strategies. In Africa, farmers’ resource constraints, the very seriousness of the soil fertility problem, and the extreme lack of technological options mean that no single technology—no chemical fertiliser, organic fertiliser, or new variety alone—can possibly solve all of farmers’ problems. The same is true for the intensive irrigated wheat systems that feed so many people. At first glance, wheat farmers in irrigated areas may appear to have many more resources than maize farmers of southern Africa, but they also confront enormous challenges in maintaining the productivity of their cropping systems without employing environmentally destructive practices. The design of more precise, economic, and environmentally friendly nutrient management practices is crucial for agriculture worldwide, and it is CIMMYT’s mission to ensure that developing country maize and wheat producers benefit as much as possible from these kinds of innovations. 3.8 References

Aldrich SR (1980) N in relation to food, environment and energy. Special Publication 61. Illinois Agricultural Experiment Station, United States of America.

Austin RB, Ford MA, Morgan CL, Yeoman Y (1993) Old and modern wheat cultivars compared to broadbalk wheat experiment. European Journal of Agronomy 2, 141-147.

Bänziger M, Cooper ME (2001) Breeding for low-input conditions and consequences for participatory plant breeding: Examples from tropical maize and wheat. Euphytica (in press).)

Bänziger M, Edmeades GO, Beck D, Bellon M (2000a) Breeding for Drought and Nitrogen Stress Tolerance in Maize: From Theory to Practice.( Mexico, D.F.: CIMMYT).

Bänziger M, Edmeades GO, Lafitte HR (1999) Selection for drought tolerance increases maize yields over a range of N levels. Crop Science 39, 1035-1040.

Bänziger M, Pixley KV, Vivek B, Zambezi BT (2000b) Characterisation of elite maize germplasm grown in eastern and southern Africa: Results of the 1999 regional trials conducted by CIMMYT and the Maize and Wheat Improvement Research Network for SADC (MWIRNET). (Harare, Zimbabwe: International Maize and Wheat Improvement Centre).

Bellon M, Karigwindi J, Waddington S (2000) Talk with Chiduku farmers conducting their own experiments with velvet bean as a green manure. Target (Newsletter of the Soil Fertility Research Network for Maize-Based Cropping Systems in Malawi and Zimbabwe) 24, 9.

Benson TD (1998) Developing flexible fertiliser recommendations for smallholder maize production in Malawi. In ‘Soil Fertility Research for Maize-Based Farming Systems in Malawi and Zimbabwe’. Proceedings of the Soil Fert Net Results and Planning Workshop held from 7 to 11 July 1997 at Africa University, Mutare, Zimbabwe. (Eds. SR Waddington, HK Murwira, JDT Kumwenda, D Hikwa and F Tagwira) pp. 275-285. (Harare, Zimbabwe: Soil Fert Net and International Maize and Wheat Improvement Centre).

Blackie MJ (1994) Maize productivity for the 21st century: The African challenge. Outlook on Agriculture 23, 189-195.

32

Blake RO (1995) The challenges of restoring African soil fertility. An open letter to the World Bank. Committee on Agricultural Sustainability for Developing Countries, Washington DC, United states of America.

Bolaños J, Edmeades GO (1993a) Eight cycles of selection for drought tolerance in lowland tropical maize. I. Responses in grain yield, biomass, and radiation utilisation. Field Crops Research 31, 233-252.

Bolaños J, Edmeades GO (1993b) Eight cycles of selection for drought tolerance in lowland tropical maize. II. Responses in reproductive behavior. Field Crops Research 31 253-268.

Bolaños J, Edmeades GO, Martinez L (1993) Eight cycles of selection for drought tolerance in lowland tropical maize. III. Responses in drought-adaptive physiological and morphological traits. Field Crops Research 31, 269-286

Byerlee D (1996) Modern varieties, productivity, and sustainability: Recent experience and emerging challenges. World Development 24, 697-718.

Byerlee D, Heisey PW (1997) Evolution of the African maize economy. In ‘Africa’s Emerging Maize Revolution’.(Eds. CK Eicher and D Byerlee) pp. 9-22. (Boulder, Colorado: Lynne Rienner, United states of America).

Byerlee D, Moya P (1993) Impacts of International Wheat Breeding Research in the Developing world, 1966-1990. (International Maize and Wheat Improvement Centre, Mexico, D.F).

Byerlee D, Siddiq A (1994) Has the green revolution been sustained? The quantitative impact of the seed-fertiliser revolution in Pakistan revisited. World Development 22, 1345-1361.

CAST (Council for Agricultural Science and Technology). 1985. Agricultural and Ground-water Quality. Report no. 103. Ames, Iowa: Council for Agricultural Science and Technology, United States of America.

Chameides WL, Kasibhatla PS, Yienger J, Levy H (1994) Growth of continental-scale metro-agro-plexes, regional ozone pollution, and world food production. Science 264, 74-77.

CIMMYT (International Maize and Wheat Improvement Centre) (1999) CIMMYT 1997/98 World Maize Facts and Trends. Maize Production in Drought-Stressed Environments: Technical Options and Research Resource Allocation. CIMMYT, Mexico, D.F.

Dhliwayo DKC, Mukurumbira L, Sithole T, Nemasasi H, Hikwa D, Gatsi T (1999) Liming in Zimbabwe: A Critical Look at the Potential Recovery from Acid Soil Infertility in the Communal Areas of Zimbabwe. Soil Fert Net Research Results Working Paper 5. (Harare, Zimbabwe: International Maize and Wheat Improvement Centre).

Dhugga KS, Waines JG (1989) Analysis of N accumulation and use in bread and durum wheat. Crop Science 29, 1232-1239.

Doerge TA, Roth RL, Gardner BR (1991) Nitrogen Fertiliser Management in Arizona. College of Agriculture, University of Arizona. Tucson, Arizona, United States of America.

Entz MH, Fowler DB (1989) Response of winter wheat to N and water: Growth, water use, yield and grain protein. Plant Science 69, 1135-1147.

FAO (Food and Agriculture Organization of the United Nations) (1990) Fertiliser Yearbook. FAO. Rome, Italy.

Fischer RA, Howe GN, Ibrahim Z (1993) Irrigated spring wheat and timing and amount of N fertilizer. I. Grain yield and protein content. Field Crops Research 33, 37-56.

Fujisaka S, Harrington LW, Hobbs PR (1994) Rice-wheat in South Asia: System and long-term priorities established through diagnostic research. Agricultural Systems 46, 169-187.

Galloway JN, Schlesinger WH, Levy H, Michaels A, Schnoor JL (1995) N fixation – anthropogenic enhancement-environmental response. Global Biogeochemical Cycles 9, 235-252.

Gambara P, Machemedze TE, Mwenye D (2000) Chihota Soil Fertility Project, Annual Report 1998 to 1999. Marondera, Zimbabwe: AGRITEX. Mimeo.

Ganunga R, Yerokun O, Kumwenda JDT (1998) Tithonia diversifolia: An organic source of nitrogen and phosphorus for maize in Malawi. In ‘Soil Fertility Research for Maize-Based Farming Systems

33

in Malawi and Zimbabwe’. Proceedings of the Soil Fert Net Results and Planning Workshop held from 7 to 11 July 1997 at Africa University, Mutare, Zimbabwe. (Eds. SR Waddington, HK Murwira, JDT Kumwenda, D Hikwa and F Tagwira) pp. 191-194. (Harare, Zimbabwe: Soil Fert Net and International Maize and Wheat Improvement Centre).

Granli T, Bockman OC (1994) Nitrous oxide from Agriculture. Norwegian Journal of Agricultural Sciences. Supplement No.12.

Heisey PW, Mwangi W (1996) Fertiliser Use and Maize Production in Sub-Saharan Africa. Economics Working Paper 96-01. (Mexico, D.F.: International Maize and Wheat Improvement Centre).

Hikwa, D, Murata M., Tagwira F, Chiduza C, Murwira H, Muza L, Waddington SR (1998). Performance of green manure legumes on exhausted soils in northern Zimbabwe: A Soil Fertility Network trial. In ‘Soil Fertility Research for Maize-Based Farming Systems in Malawi and Zimbabwe’. Proceedings of the Soil Fert Net Results and Planning Workshop held from 7 to 11 July 1997 at Africa University, Mutare, Zimbabwe. (Eds. SR Waddington, HK Murwira, JDT Kumwenda, D Hikwa and F Tagwira) pp. 81-84. (Harare, Zimbabwe: Soil Fert Net and International Maize and Wheat Improvement Centre).

Hobbs PR, Morris ML (1996) Meeting South Asia’s Future Food Requirements from Rice-Wheat Cropping Systems: Priority Issues Facing Researchers in the Post-Green Revolution Era. NRG Paper 96-01. Mexico D.F.: International Maize and Wheat Improvement Centre.

Hobbs PR, Sayre KD, Ortiz-Monasterio JI (1998) Increasing Wheat Yields Sustainably through Agronomic Means. NRG Paper 98-01. Mexico, D.F.: International Maize and Wheat Improvement Centre.

Howarth RW, Billen G, Swaney D, Townsend A, Jaworski N, Lajtha K, Downing JA, Elmgren R, Caraco N, Jordan T, Berendse D, Freney J, Kudeyarov V, Murdoch P, ZhaoLiang Z. (1996) Regional N budgets and riverine N and P fluxes for the drainages to the North Atlantic Ocean: natural and human influences. Biogeochemistry 35, 75-139

IAEA (International Atomic Energy Agency) (1999) Technical Document, in preparation. IAEA, Vienna, Austria.

Jain NK, Krantz BA, Bains KB, Bhardwaj RBL, Bhattacharya S, Gill GS, Gupta KP, Moolani MK, Morabad IR, Patel UJ, Reddy MK, Sandhu KS, Sharma AK, Sharma KC, Shekhawat GS, Singh M, Singh RR, Srivastava MMP, Wright BC (1975) Response of dwarf wheat varieties to N. In ‘Agronomy of Dwarf Wheats: Summary of Investigations of the All-India Co-ordinated Wheat Improvement Project’. (Eds. AM Wadhwani and R Singh) pp. 11-23. (Indian Council of Agricultural Research, New Delhi, India).

Jiri O, Waddington S (1998) Leaf prunings from two species of Tithonia raise maize grain yield in Zimbabwe, but take a lot of labour! Target (Newsletter of the Soil Fertility Research Network for Maize-Based Cropping Systems in Malawi and Zimbabwe) 16, 4-5.

Jones RB, Wendt JW (1995) Contribution of soil fertility research to improved maize production by smallholders in eastern and southern Africa. In ‘Maize Research for Stress Environments’. Proceedings of the 4th Eastern and Southern Africa Regional Maize Conference. (DC Jewell, SR Waddington, JK Ransom and KV Pixley) pp. 2-14. International Maize and Wheat Improvement Centre. Mexico D. F.

Keeney DR (1982) N management for maximum efficiency and minimum pollution. In ‘N in Agricultural Soils’. Agronomy Monograph 22. Madison, Wisconsin, USA. (Ed. FJ Stevenson) pp. 605-649. (American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America).

Kumwenda AS (1998) Observations on the Malawi maize productivity task force. In ‘Soil Fertility Research for Maize-Based Farming Systems in Malawi and Zimbabwe’. Proceedings of the Soil Fert Net Results and Planning Workshop held from 7 to 11 July 1997 at Africa University, Mutare, Zimbabwe. (Eds. SR Waddington, HK Murwira, JDT Kumwenda, D Hikwa and F Tagwira) pp. 263-269. (Harare, Zimbabwe: Soil Fert Net and International Maize and Wheat Improvement Centre).

34

Kumwenda JDT, Gilbert R (1998) Biomass production by legume green manures on exhausted soils in Malawi: A Soil Fertility Network trial. In ‘Soil Fertility Research for Maize-Based Farming Systems in Malawi and Zimbabwe’. Proceedings of the Soil Fert Net Results and Planning Workshop held from 7 to 11 July 1997 at Africa University, Mutare, Zimbabwe. (Eds. SR Waddington, HK Murwira, JDT Kumwenda, D Hikwa and F Tagwira) pp. 85-86. (Harare, Zimbabwe: Soil Fert Net and International Maize and Wheat Improvement Centre).

Kumwenda JDT, Waddington SR, Snapp SS, Jones RB, Blackie MJ (1996) Soil Fertility Management Research for the Maize Cropping Systems of Smallholders in Southern Africa: A Review. Natural Resources Group Paper 96-02. (Mexico, D.F.: International Maize and Wheat Improvement Centre).

Kumwenda JDT, Waddington SR, Snapp SS, Jones RB, Blackie MJ (1997) Soil fertility management in southern Africa. In ‘Africa’s Emerging Maize Revolution’. (Eds. CK Eicher and D Byerlee) pp. 157-172. (Boulder, Colorado: Lynne Rienner, United States of America).

Kwesiga F (1998) Improved fallows for sustainable food security in eastern Zambia. Transactions of the Zimbabwe Scientific Association 72 (Supplement), 72-83.

Kwesiga F, Coe RS (1994) The effect of short rotation Sesbania sesban planted fallows on maize yield. Forest Ecology and Management 64, 199-208.

Lafitte HR, Edmeades GO (1994a). Improvement for tolerance to low soil nitrogen in tropical maize. I. Selection criteria. Field Crops Research 39, 1-14.

Lafitte HR, Edmeades GO (1994b) Improvement for tolerance to low soil nitrogen in tropical maize. II. Grain yield, biomass production, and N accumulation. Field Crops Research 39, 15-25.

Lafitte HR, Edmeades GO (1994c) Improvement for tolerance to low soil nitrogen in tropical maize III. Variation in yield across environments. Field Crops Research 39, 27-38.

Large EC (1954) Growth scales in cereals. Illustrations of the Feekes Scale. Plant Pathology 3, 128-129.

Malama C (2000) The potential of Tithonia diversifolia, in combination with phosphate fertiliser and indigenous ground rock phosphate, to raise maize yield on the acid soils of Northern Zambia. Target (Newsletter of the Soil Fertility Research Network for Maize-Based Cropping Systems in Malawi and Zimbabwe) 24, 4-5.

Mann C (1998) Higher Yields for All Smallholders through “Best Bet” Technology: The Surest Way to Restart Economic Growth in Malawi. Soil Fert Net Research Results Working Paper 3. (Harare, Zimbabwe: International Maize and Wheat Improvement Centre).

Matson PA, Naylor R, Ortiz-Monasterio I (1998) Integration of environmental, agronomic and economic aspects of fertiliser Management. Science 280, 112-115.

Matson PA, Parton WJ, Power AG, Swift M (1997) Agricultural intensification and ecosystem properties. Science 277, 504-509.

McCalla AF (2000) Agriculture in the 21st Century. CIMMYT Economics Program 4th Distinguished Economist Lecture. (Mexico, D.F.: International Maize and Wheat Improvement Centre).

Meisner CA, Acevedo E, Flores D, Sayre KD, Ortiz-Monasterio I, Byerlee D (1992) Wheat Production and Growers’ Practices in the Yaqui Valley, Sonora, Mexico. Wheat Special Report No. 6. Mexico D.F., International Maize and Wheat Improvement Centre.

Moll RH, Kamprath EJ, Jackson WA (1982) Analysis and interpretation of factors which contribute to efficiency of N utilization. Agronomy Journal 74, 562-564.

Mushayi PT, Waddington SR, Chiduza C (1999) Low efficiency of N use by maize on smallholder farms in sub-humid Zimbabwe. In Maize Production Technology for the Future: Challenges and Opportunities. Proceedings of the 6th Eastern and Southern Africa Regional Maize Conference, 21-25 September, 1998. (Addis-Ababa, Ethiopia: International Maize and Wheat Improvement Centre and Ethiopian Agricultural Research Organization) pp. 278-281.

Ortiz-Monasterio JI, Matson PA, Panek J, Naylor RL (1996) N fertiliser management: Consequences for N2O and NO emissions in Mexican irrigated wheat. Transactions of the 9th N Workshop. Braunschweig, Germany. pp. 531-534.

35

Ortiz-Monasterio I, Sayre KD, Davila GF, Camacho M, Moreno O (1993) Algunas prácticas agronómicas relacionadas con la incidencia de carbon parcial (Neovossia indica). Memorias del Taller Sobre las Estrategias para el Control del Carbon Parcial del Trigo (Neovossia indica) en el Estado de Sonora. Mexico, D.F.: International Maize and Wheat Improvement Centre. pp.47-58.

Ortiz-Monasterio RJI, Sayre KD, Pena J, Fischer RA (1994) Improving the N-use efficiency of irrigated spring wheat in the Yaqui Valley of Mexico. 15th World Congress of Soil Science 5b, 348-349.

Ortiz-Monasterio RJI, Sayre KD, Rajaram S, McMahon M (1997) Genetic progress in wheat yield and N-use efficiency under four N rates. Crop Science 37, 898-904.

Pingali P, Heisey PW (1996) Cereal crop productivity in developing countries: Past trends and future prospects. In ‘Global Agricultural Science Policy for the Twenty-First Century’. (Natural Resources and Environment , Melbourne, Australia) pp. 51-94.

Pingali P, Rajaram S (1999) Global Wheat Research in a Changing World: Challenges and Achievements. Part 1 of CIMMYT 1998-99 World Wheat Facts and Trends. Mexico, D.F.: International Maize and Wheat Improvement Centre.

Pinstrup-Andersen P, Pandya-Lorch R, Rosegrant MW (1999) World Food Prospects: Critical Issues for the Early Twenty-first Century. International Food Policy Research Institute, Washington, DC, United States of America.

Pompi I, Mpepereki S, Gwata E (1998) Zimbabwe soybean promotion taskforce: Objectives, achievements, and agenda for the future. In ‘Soil Fertility Research for Maize-Based Farming Systems in Malawi and Zimbabwe’. Proceedings of the Soil Fert Net Results and Planning Workshop held from 7 to 11 July 1997 at Africa University, Mutare, Zimbabwe. (Eds. S.R. Waddington, H.K. Murwira, J.D.T. Kumwenda, D. Hikwa, and F. Tagwira) pp. 271-273. (Harare, Zimbabwe: Soil Fert Net and International Maize and Wheat Improvement Center).

Rajaram S, van Ginkel M, Fischer RA (1995) CIMMYT’s wheat breeding mega-environments (ME). In ‘Proceedings of the 8th International Wheat Genetics Symposium’, 19-24 July, Beijing, China. (Eds. ZS Li and ZY Xin) pp. 1101-1106. (China Agricultural Scientech., Beijing, China).

Reeves TG (1999) Intensification for the nine billion. Nature Biotechnology (supplement) 17, BV33−BV34.

Rosegrant MW, Sombilla MA, Gerpacio RV, Ringler C (1997) Global food markets and US exports in the twenty-first century. Paper presented at the Illinois World Food and Sustainable Agriculture Program Conference, Meeting the Demand for Food in the 21st Century: Challenges and Opportunities, 28 May, University of Illinois, Urbana-Champaign, United States of America.

Rusike J, Sukume C, Dorward A, Mpepereki S, Giller K (2000) The Economic Potential for Smallholder Soyabean Production in Zimbabwe. Soil Fert Net Special Publication. (Harare, Zimbabwe: International Maize and Wheat Improvement Centre).

Sanchez PA, Shepherd KD, Soule MJ, Place FM, Mokwunye AU, Buresh RJ, Kwesiga FR, Izac AMN, Ndiritu CG, Woomer PL (1997) Soil fertility replenishment in Africa: An investment in natural resource capital. In ‘Replenishing Soil Fertility in Africa’. Soil Science Society of America Special Publication 51. (Eds. RJ Buresh, PA Sanchez and F Calhoun) pp. 1-46. (Soil Science Society of America and American Society of Agronomy, Wisconsin, United states of America).

Sayre KD (1996) The role of crop management research at CIMMYT in addressing bread wheat yield potential issues. In ‘Increasing Yield Potential in Wheat in Wheat: Breaking the Barriers’. (Eds. M.P. Reynolds, S. Rajaram, and A. McNab) pp. 203-207. (Mexico, D.F.: International Maize and Wheat Improvement Centre).

Simmonds NW (1979) Principles of Crop Improvement. (Longmans, New York, United states of America).

Smaling E (1993) Soil nutrient depletion in sub-Saharan Africa. In ‘The Role of Plant Nutrients for Sustainable Food Crop Production in Sub-Saharan Africa’. (Eds. H van Reuler and WH Prins) pp. 53-67. (Leidschendam, The Netherlands)

36

Smil V (1999) Long-Range Perspectives on Inorganic Fertilisers in Global Agriculture. Travis P. Hignett Lecture. Muscle Shoals, Alabama: International Fertiliser Development Centre.

Stoorvogel JJ, Smaling EMA (1990) Assessment of Soil Nutrient Depletion in Sub-Saharan Africa, 1983-2000. Report 28. Wageningen, the Netherlands: Winard Staring Centre for Integrated Land, Soil and Water Research.

Stoorvogel JJ, Smaling EMA, Janssen BH (1993) Calculating soil nutrient balances in Africa at different scales. I. Supranational scale. Fertiliser Research 35, 227-235.

Swift MJ, Woomer P (1993) Organic matter and the sustainability of agricultural systems: Definition and measurement. In ‘Soil Organic Matter Dynamics and Sustainability of Tropical Agriculture’. (Eds. K Mulongoy and R Merckx) pp. 3-18. (Chichester, Wiley-Sayce, United Kingdom).

van Sanford DA, MacKown CT (1986) Variation in N-use efficiency among soft red winter wheat genotypes. Theoretical and Applied Genetics 72, 158-163.

Vitousek PM, Aber JD, Howarth RW, Likens GE, Matson PA, Schindler DW, Schlesinger WH, Tilman DG (1997) Human alteration of the global N cycle: Sources and consequences. Ecological Applications 7, 737-750.

Vitousek PM, Matson PA (1993) Agriculture, the global N cycle, and trace gas flux. In ‘The Biogeochemistry of Global Change: Radiatively Active Trace Gases’. (Ed. RS Oremland). (Champman and Hall, New York, United States of America).

Waddington SR, Gilbert R, Giller KE (1998) “Best Bet” technologies for increasing nutrient supply for maize on smallholder farms. In ‘Soil Fertility Research for Maize-Based Farming Systems in Malawi and Zimbabwe’. Proceedings of the Soil Fert Net Results and Planning Workshop held from 7 to 11 July 1997 at Africa University, Mutare, Zimbabwe. (Eds. SR Waddington, HK Murwira, JDT Kumwenda, D Hikwa and F Tagwira) pp. 245-250. (Harare, Zimbabwe: Soil Fert Net and International Maize and Wheat Improvement Centre)

Waddington SR, Heisey PW (1997) Meeting the N requirements of maize grown by resource-poor farmers in southern Africa by integrating varieties, fertiliser use, crop management, and policies. In ‘Developing Drought- and Low N-Tolerant Maize’. Proceedings of a Symposium, 25 to 29 March 1996, CIMMYT, El Batan, Texcoco, Mexico. (Eds. GO Edmeades, M Bänziger, HR Mickelson and CB Peña-Valdivia) pp. 44-57. (Mexico, D.F.: International Maize and Wheat Improvement Centre).

Wall PC, McMahon MA, Ransom, JK (1984) Do semidwarf wheats require more N than traditional tall varieties? Agronomy Abstracts. Madison, Wisconsin: American Society of Agronomy. P. 45

Woomer PL, Martin A, Albrecht A, Resck DVS, Scharpenseel HW (1994) The importance and management of soil organic matter in the tropics. In ‘The Biological Management’. (Eds. PL Woomer and MJ Swift).

World Bank (1997) World Development Indicators 1997. (Oxford University Press, New York, United States of America).

World Bank (2000) World Development Report 2000/2001: Attacking Poverty. Zadoks JC, Chang TT, Konzak CF (1974) A decimal code for growth stages of cereals. Weed

Research 14, 415-421. Zambezi, Mwambula (1997) The impact of drought and low soil nitrogen on maize production in the

SADC region. In ‘Developing Drought and Low N-Tolerant Maize’. Proceedings of a Symposium, March 25-29, 1996, CIMMYT. (Eds. GO Edmeades, M.Bänziger, HR Mickelson, and CB Peña-Valdivia) pp 29-34. (CIMMYT , El Batán, Mexico. Mexico D.F.).

37

4. The response of field-grown rice to inoculation with a multi-strain biofertiliser in the Hanoi district, Vietnam

Nguyen Thanh Hien, I. R. Kennedy and R. J. Roughley 4.1 Abstract

A multi-strain biofertiliser was found to provide statistically significant increases in rice yield in two out of three field trials in Vietnam. This biofertiliser contained three strains of bacteria selected from rice rhizospheres in paddies near Hanoi. The benefit possible for rice farmers from application of the inoculant biofertiliser was confirmed as a reliable effect by positive results in 65 farmer demonstrations over three seasons for both summer and winter rice crops, with the increases in grain yield compared to farm areas receiving urea alone usually much greater than 10 percent. Increases in the dose of biofertiliser organisms applied in the range 5.5 – 22.2 x 1012 cfu ha-1 had no significant effect suggesting that with suitable quality control to ensure its effectiveness, costs of application could be reduced. The three biofertiliser strains were selected respectively for their ability to reduce acetylene (N2 fixation), mobilise insoluble phosphates and to favour establishment of the other two under competition from other rhizosphere organisms. There is evidence of significant stimulation of early root and seedling growth and of panicle numbers and seeds per panicle as a result of applying biofertiliser but the precise mechanisms of increases in grain yield remains a topic for future research. Keywords: Rice, biofertiliser. 4.2 Introduction

Biofertilisers have been used to increase legume crop performance for centuries with the direct transfer of soil from areas where particular crops were growing well to those where the crop was to be introduced. Not till 100 years ago was the rationale behind the practice described and from which inoculation with pure cultures of root nodule bacteria followed (Fred et al. 1932). Almost concurrently, inoculation with other biofertilisers developed with the discovery of improved crop growth following inoculation with Azotobacter. Unpredictable results were common with both systems. The causes of those associated with root nodule bacteria were later attributed to poor strains and too small an inoculum potential. The causes of variable responses with other organisms remain poorly understood. The situation is complicated by the wide range of organisms involved, the mixed nature of cultures, a largely unknown mechanism for their stimulation of plant growth, and claims for them often based on unreplicated experiments. Recently interest in biofertilisers has been heightened by studies of the effect of asymbiotic nitrogen-fixing organisms and in particular Azospirillum on plant growth (Kennedy and Islam 2001). In Vietnam, experience is largely with biofertilisers based on studies by one of us (Nguyen Thanh Hien) from 1990. This work concentrated on making available to farmers a product which could both reduce their dependence on inorganic fertilisers and increase yields of rice. In the following years a considerable body of data was accumulated based largely on unreplicated trials established at many sites, which demonstrated a surprisingly consistent response. The substitution of biofertiliser for more expensive chemical fertilisers, using local labour inputs, could become a significant element in the alleviation of poverty in these poor farming communities.

38

The experiments described in this paper aimed to use the biofertiliser prepared at the Hanoi National University of Science, to quantify any responses obtained in replicated trials over three years. We took the opportunity to look in a preliminary way at the dose rates of biofertiliser required and their interaction with normal farmer inputs in the area of urea and farmyard manure. 4.3 Materials and Methods

Sites and experimental design A field trial was sown in the Hanoi area, Vietnam in July 1999, July 2000, and February 2001. All trials were established in alluvial soils of the Red River and those of 2000 and 2001 were in the same field at Dai Moi where the soil pH varied between 5.19 and 5.58 (CaCl2). Applications of farmyard manure (FYM) had no effect on soil pH. The trials were established in a split plot design replicated four times with rates of biofertiliser applied to the subplots. The trials in 1999 and 2000 had N levels as the main plots but in 2001, levels of FYM were allocated to the main plots. Plot sizes between replicates were not constant due to the shape of the fields but were at least 20 m2 up to a maximum of 40 m2. The amounts of fertiliser and biofertiliser applied to each plot were adjusted to allow for this variation. The plots were protected by well-prepared banks to minimise the movement of water between plots thereby reducing the possibility of uninoculated plots being contaminated with biofertiliser. A basal fertiliser of muriate of potash (MOP) at 55 kg ha–1 and triple superphosphate (TSP) at 417 kg ha–1 was applied in 1999 and 2000. In 2001 TSP was reduced to 208 kg ha–1. In addition in 1999, 11.1 t ha–1 wet weight of FYM was included but in 2000 this amount was halved. In 2001 urea was applied to all plots at 55 kg ha–1 replacing FYM which was a variable in this trial.

Planting geometry In 1999, 3-4 plants were transplanted in rows 17 cm apart with hills at 14-17 cm spacing providing 42–45 hills of rice m2. In 2000 the plantings were 14 cm apart in rows spaced at 21 cm. To simplify harvesting, in 2001 the hills were spaced at 45 hills per m2 and statistical analysis of the number of hills per m2 showed no significant difference between plots P<0.05.

Biofertiliser The biofertiliser contained three strains of bacteria all selected (by N. T. Hien) from rice rhizospheres in the Hanoi area of Vietnam. One (2N) was selected for its ability to reduce acetylene to ethylene as an indication of potential N2 fixation, a second, (4P) for its ability to solubilise PO4 in an agar medium and the other (3C) for its ability to produce toxic extra cellular compounds which inhibited 50% of a test group of 100 rhizosphere organisms but to which the inoculum strains were resistant. Its purpose was to aid the establishment of the inoculum in competition with other rhizosphere organisms. Each of the three bacteria were grown in separate broth cultures and added to separate bags of carrier formulated by mixing peat 50%, rice husks 25%, sugar 1%, plus water and inoculum 24%. These separate cultures were mixed in the field immediately before use in the ratio of 10 parts 2N: 10 parts of 4P : 1 part of 3C. Because strains 2N and 3C are difficult to count in the unsterile carrier, we were only able to make a direct count of 4P, which was 3 x 109 cfu g-1 carrier. We estimate that the numbers of 2N and 3C, based on counts of their broth cultures, to be approximately 1 x 108 and 1 x 107 g-1 carrier respectively. Biofertiliser was applied to the seedlings at sowing at 25% of the rate to be used in the plots to which they were to be transplanted. Biofertiliser was applied to the field plots by spreading the carrier evenly by hand directly to the soil. To plots which were to remain uninoculated, uninoculated carrier was added at a rate equivalent to 222 kg ha–1.

Composition of farmyard manure The composition of the farmyard manure used in the 2001 field trial is presented in Table 1. Farmyard manure is by its nature variable. The eight samples we analysed had a mean nitrogen value of 2.79% with a standard error of 0.27 or 10%.

39

Table 1. Composition of farmyard manure used in the field trial in 2001 Mean organic matter (%)

Total nitrogen (%) Total phosphorus (%) Total potassium (%)

33.66 ± 2.389 2.79 ± 0.271 0.72 ± 0.106 1.04 ± 0.105 Seedling production Seedlings of rice variety 9830 for the trials in 1999 and 2000 were grown in plastic trays of tapered hexagonal thimble shaped units 1.5 cm diam., filled with wet soil with seeds spread evenly over their surface. Seedlings were grown for 12 days before transplanting. In 2001 we used traditional field sown nursery bays to assist the farmers in planting a uniform 3 seedlings per hill. Separate seedling bays were used for each level of biofertiliser. Seedlings of rice variety Khang Dan, a shorter season variety than 9830 and better adapted to winter sowings, were transplanted 24 days after sowing rice sprouts.

Treatments 1999 Three levels of urea (0, 83 and 194 kg ha–1) were applied to main plots, and four levels of biofertiliser (0, 111, 222 and 444 kg ha–1) were applied to subplots.

Treatments 2000 Three levels of urea (0, 83 and 194 kg ha–1) were applied to main plots, and four levels of biofertiliser (0, 55, 111 and 222 kg ha–1) were applied to subplots.

Treatments 2001 Three levels of FYM (5,560; 11,120 and 22,240 kg ha–1) were applied to main plots, and four levels of biofertiliser (0, 111, 222 and 444 kg ha–1) were applied to subplots.

Sampling and harvest Plants were sampled in each trial six weeks after transplanting. In 1999 and 2000 the number of tillers, root and tiller dry weight and plant height at 12 sampling points selected at random were recorded and in 2001 plant height and tiller dry weight were recorded at 10 sampling points. At harvest, the number of panicles in three hills and the number of fertile and infertile seeds per panicle in five panicles were counted. Samples of 1000 seeds and grain yield from 5, 1 m2 quadrats in 1999 and 2000, and from 10 samples of 25 hills in 2001 were weighed and the results were expressed as kg ha–1. In 2000 and 2001, the grain was analysed for total N content (%), and nitrogen uptake (kg ha-1) by grain was calculated. The data were analysed using the Genstat statistical package.

Farmers’ tests of biofertiliser We collected data from demonstration trials sown by farmers of the effect of adding biofertiliser to rice. In July 1999 three communes were included and 25 farmers participated. In July 2000 four communes and 20 farmers, and in January 2001 two villages and 20 farmers participated. Each farmer was asked to divide his field into two plots one to receive biofertiliser and half the normal fertiliser inputs, and the other to include all normal inputs and no biofertiliser. Costs of the two systems were recorded.

40

4.4 Results

Field trial 1999 There was very little effect of treatment on the vegetative plant at 6 weeks. Biofertiliser had no significant effect on root dry weight, tiller dry weight but had a slight depressing effect (P = 0.05) on plant height. The difference between the two extremes was only 3 cm or 4.4%. Urea increased plant height from 68 to 70.8 cm. Application of urea reduced grain yield from 6933 kg ha–1 (mean yield in urea control plots) to 6623 and 6565 kg ha-1 when it was applied at 83 and 194 kg ha–1, respectively. There was a significant interaction between urea and biofertiliser (P<0.05). At urea levels of 0 and 83 kg ha–1, biofertiliser application at 111 kg ha-1 increased grain yield significantly although there was no response to increased doses of biofertiliser. When urea was applied at 194 kg ha–1 there was no effect of biofertiliser (Table 2).

Field trial 2000 Urea stimulated both early tiller and root growth; at the 6 week sampling without added urea, mean tiller dry wt was 5.54 g but increased to 7.96 and 10.00 g with urea applied at 83 and 194 kg ha–1, respectively. Root growth responded similarly with roots of 1.43 g without added urea and 2.09 and 2.71 g with added urea at 83 and 194 kg ha-1, respectively. At harvest neither the number of seeds per panicle or 1000 seed weight responded to treatment. Biofertiliser had a progressive negative effect on grain yield which declined by 413 kg ha–1 or 6.7 % with the highest application (Table 3). Grain yield increased significantly (P<0.05) due to application of each additional amount of urea. The increases were 1274 kg ha–1 (26.4%) and 717 kg ha–1 (11.8%). The response in grain yield was matched with a similar response in N uptake by grain, which also did not respond to the application of biofertiliser. Table 2. Effects of urea and biofertiliser on the grain yield of rice adjusted to 15% moisture in 1999

Urea (kg ha–1)

Biofertiliser (kg ha–1)

Mean* grain yield (kg ha–1)

0 0 6528 0 111 7254 0 222 7228 0 444 6702 83 0 5732 83 111 6730 83 222 6947 83 444 7081 194 0 6789 194 111 6466 194 222 6403 194 444 6603

*Means of 20 samples. LSD (0.05) = 542

41

Table 3. Effects of urea and biofertiliser on the grain yield of rice and N uptake by grain in 2000

Urea kg ha-1 Biofertiliser kg ha-1 Mean 0 55 111 222

Grain yield (kg ha-1) 0 5087 4985 4645 4599 4829 83 6288 6252 5988 5884 6103 194 7076 6787 6690 6727 6820 Mean 6150 6008 5774 5737

N uptake by grain (kg ha-1) 0 46 44 39 38 42 83 52 59 49 51 53 194 61 68 59 62 63 Mean 53 57 49 50

Grain yield: LSD (0.05) for urea means = 364.9, and for biofertiliser means = 272.0. N uptake by grain: LSD (0.05) for urea means = 4.5.

Field trial 2001 Applications of farmyard manure did not affect either vegetative growth or grain yield. Biofertiliser had an early effect on vegetative growth with both plant height and tiller dry weight responding positively (Table 4). At harvest the number of seeds per panicle was unaffected by biofertiliser but the weight of 1000 seeds responded directly to increasing the dose of biofertiliser. The difference between 0 and 444 kg ha –1 was significant (P<0.05). Grain yield increased significantly due to biofertiliser application at 111 kg ha-1 (Table 5), but the response did not increase further by applying more biofertiliser. Yield increased by 553 kg ha–1 or 9.9 % with biofertiliser applied at 111 kg ha–1. This increase was significant at the 0.1% probability level. Nitrogen uptake by grain (kg ha-1) increased significantly due to biofertiliser application at 111 kg ha-1, beyond this rate there was no further increase. This increase was significant at 1% level of probability. Farmyard manure (FYM) application at 11,120 kg ha-1 increased N uptake by grain significantly over the lowest rate of FYM, at the highest rate of FYM there was no further increase. The effect of FYM was significant at 5% level of probability.

Table 4. Effects of biofertiliser on tiller dry weight and plant height at 6 weeks, and number of panicles per hill, number of seeds per panicle and weight of 1000 seeds at harvest in 2001

Biofertilisier (kg ha-1)

Tiller dry wt. (g per hill)

Plant height (cm)

No. of panicles per hill

No. of seeds per panicle

Dry wt. of 1000 seeds (g)

0 10.83 62.5 4.4 194 18.411 111 12.07 66.7 4.6 197 18.624 222 13.52 65.8 5.1 200 18.726 444 12.97 66.4 4.8 201 18.983

LSD (0.05): tiller dry weight = 1.375, plant height = 1.98, number of panicles per hill = 0.42, number of seeds per panicle = 10.8, dry weight of 1000 seeds = 0.323.

42

Table 5. Effects of farmyard manure and biofertiliser on the grain yield of rice and N uptake by grain in 2001

FYM kg ha-1 Biofertiliser kg ha-1 Mean 0 111 222 444 Grain yield (kg ha-1) 5,560 5476 6170 5890 5801 5834 11,120 5443 6360 6111 5979 5973 22,240 5764 5813 6116 5854 5888 Mean 5561 b 6114 a 6039 a 5878 a N uptake by grain (kg ha-1) 5,560 50.40 55.89 53.59 51.14 52.76 B 11,120 51.41 59.28 57.09 54.69 55.62 A 22,240 50.67 54.62 57.29 55.78 54.59 A Mean 50.83 b 56.60 a 55.99 a 53.87 a

Grain yield: LSD (0.05) for biofertiliser means = 258.1. N uptake by grain: LSD (0.05) for FYM means = 1.669, and for biofertiliser means = 2.903. Means followed by a common small letter in a row and a common capital letter in a column for a parameter are not significantly different at 5% level by least significant difference (LSD).

Farmers’ tests of biofertiliser All the results of the 65 farmers that were averaged by commune or village and presented in Table 6 showed an increase in yield resulting from biofertiliser. The overall mean increase was 15% (728 kg ha–1) with extremes of 8.3-30.7%. The mean increase was valued at Aus$146. There was a supplementary saving in inputs in all but one commune and although its value was secondary to that from increased yield, averaged AUD15 per ha. The mean total economic benefit was AUD161 per ha which is a significant amount in this economy. The farmers with the highest rate of improvement increased their return by AUD274 which included AUD251 per ha from increased yield. None of the farmers appeared sufficiently confident to reduce fertiliser inputs to 50% of their normal practice. The mean reduction was only 7.6%.

43

Table 6. Effect of biofertiliser on cost of production and yield of rice in farmer’s demonstrations in 1999, 2000 and 2001

Year Commune (No. farmers)

Yield kg ha-1

% Increase yield

Value of increased grain AUD

Cost of fertiliser inputs AUD

Savings on inputs with biofert AUD

Total economic benefit AUD

- Biofert

+ Biofert

- Biofert

+ Biofert

1999 Van Hoa (8)

4809 5449 13.3 128 120 99 21 149

Tan Linh (8)

5365 6116 14 150 150 125 25 175

Ba Trai (9)

4921 5588 13.6 133 143 121 22 155

2000 Cam

Giang (10)

5699 6422 12.7 145 247 247 0 145

Tan Linh (3)

5304 6405 20.8 220 272 257 15 235

Ba Trai (4)

5154 5905 14.6 150 279 268 11 161

Vvan Hoa (3)

4083 5338 30.7 251 273 250 23 274

2001 Li Do

Village (10)

4337 4698 8.3 72 181 169 12 84

Bang Gia Village (10)

4859 5587 15 146 204 189 15 161

Mean 4859 5587 15 146 204 189 15 161 4.5 Discussion

Biofertiliser increased grain yield significantly in two of the three field trials and in all 65 farmer demonstrations over three seasons. This consistency obtained in both winter and summer crops, and in 66 different sites is particularly significant. In the 1999 and 2001 field trials where we obtained a positive response, the dose of biofertiliser was not a factor suggesting that even the low dose rate may well provide more than the minimum inoculum potential required. The amount applied was large in terms of that used with legume inoculants although in their case the organisms are usually strategically applied in the vicinity of the seed. The highest rate of biofertiliser we used, 222 kg ha-1, applies approximately 22.2 x 1012 cfu ha-1 and the lowest rate of 55 kg ha-1 applies 5.5 x 1012 cfu ha-1. Legume inoculants for soybean where the standards in Australia require 1 x 10 9 cfu g-1 when used at the recommended dose rate, add 5 x 1011 cfu ha–1. Even at the lowest rate, biofertiliser adds some 22 times the number of useful organisms per unit area. These high rates of application affect the economics of its use so that further experimentation is warranted to determine a minimum dose for general use. The interaction of other inputs with biofertiliser was not consistent. The best responses, those in 2001, were all obtained in conjunction with added urea. In 1999, the highest rate of urea (194 kg ha-

1), eliminated the biofertiliser response but at 83 kg ha-1 biofertiliser further increased yield.

44

Interestingly, added FYM did not affect yield in 2001 the only year in which it was a variable. FYM is nevertheless a key input in farmer sowings where it is applied in addition to 280 kg ha–1 urea. Using the N and P values form Table 1 as a guide, FYM contributed 30 kg ha–1 of N in 1999 trials and 15 kg ha–1 of N in 2000. In 2001 where FYM was a variable, approximately 15, 30 or 60 kg ha–1

of N was added depending on the amount applied to the main plots. The amount of P added in the FYM was approximately 25% of the amounts of N. These all represent high inputs against which to evaluate biofertiliser. In the farmer sowings, although they are advised to halve inputs, the data on input costs indicated they were reduced by only 7% so that the consistent responses to biofertiliser the 65farmers obtained were in the presence of relatively high levels of FYM, urea, TSP and MOP. It should be noted that in the 2000 season when we failed to obtain a response to biofertiliser, farmers using the same rice variety and biofertiliser in the same area all had a positive response. 4.6 Conclusion

Our trials and observed farmer demonstrations clearly indicated that biofertiliser, comprising strains 2N, 4P and 3C produced a consistent increase in rice yield. The factors, which caused the negative response in 2000, are not understood. This generally positive response, surprising in relation to the other trials and demonstrations, warrants further investigation together with detailed studies of the mode of action of each bacterial genus in the inoculum, the most economic rate at which to apply them and the factors which maximize their effect. 4.7 Acknowledgments

We wish to thank Australian Centre for International Agricultural Research (ACIAR) and Australian Agency for International Development (AusAID) for financial support, the Vietnamese farmers for their help in conducting the trials, Mrs Nhan of the Extension Services, Vietnam for direct help in conducting the trials and liaising with the farmers. 4.8 References

Fred EB, Baldwin IL, McCoy E (1932) Root Nodule Bacteria and Leguminous Plants. University of Wisconsin Studies in Science No. 5. (University of Wisconsin Press, Madison, Wisconsin, United States of America).

Kennedy IR, Islam N (2001) The current and potential contribution of asymbiotic nitrogen fixation to nitrogen requirements on farms: a review. Australian Journal of Experimental Agriculture 41, 447-457.

45

5. Azobiofer: A technology of production and use of Azolla as biofertiliser for irrigated rice and fish cultivation

M. H. Mian 5.1 Abstract

"Azobiofer" is a technology developed for production of Azolla in ponds and the utilisation of Azolla as biofertiliser for cultivation of irrigated rice. The guarantee of year round production of Azolla in ponds is the essential element for the sustenance of the technology using Azolla as biofertiliser. It has been shown that Azolla can be grown in ponds round the year and it is highly profitable. The cost-benefit (CB) ratio is 1:2.8. In fish nurseries, this ratio is even better. Azolla can be grown simultaneously with irrigated rice to produce 10-15 t ha-1

of biomass containing 20-30 kg N ha-1 in 3-4 weeks from initial inoculum of 0.2 kg m-2. In this way, the use of urea can be reduced by 30-40%. Azolla-based rice-fish farming has also been found possible and profitable. This can be a model for tropical and sub-tropical rice growing countries. Keywords: Azobiofer, azolla, rice, fish. 5.2 Introduction

The technology of producing Azolla year round and utilising it as biofertiliser for cultivation of irrigated rice has been named Azobiofer. The utilisation of Azolla as biofertiliser for rice and as feed for fish in Azolla-based rice-fish farming has also been found very promising. The idea of developing a technology based on Azolla on rice cultivation in Bangladesh originated from two needs - (i) to add organic matter into the soils which is very essential since the status is alarmingly low, there is little scope to grow green manures because of limitations of both extra land and time using intensive cultivation; and (ii) to reduce The dependence on chemical nitrogenous fertiliser such as urea, which accelerates the decomposition of inherent organic matter in soils. Azolla is a tiny free-floating fresh water fern of the tropical and sub-tropical Asia, Africa and America. There are now seven existing species of the family Azollaceae – Azolla caroliniana, A. maxicana, A. filiculoides, A. microphylla, A. rubra, A. nilotica and A. pinnata. Azolla is an ancient associate of green plants on earth, dating back to Cenozoic era (2-65 million years). There are fossil records of A. filiculoides and A. pinnata from Pleistocene deposits (West 1953; Moore 1969). Moore (1969) first reported the agronomic significance of Azolla. Literature on Azolla is now voluminous (Lumpkin and Plucknett 1982; Khan 1988; Kumarasinghe and Eskew 1993; Mian 1993). Azolla has the unique capacity to fix significant amounts of atmospheric dinitrogen (N2) through its phycsymobiont Anabaena azollae and thereby can act as the nitrogenous biofertiliser of the irrigated rice. Since Azolla could be grown simultaneously with irrigated rice needing no extra land nor extra water, its utility as biofertiliser had become a reality (Kikuchi et al. 1984; Singh and Singh 1990; Mian and Kashem 1995). Systematic research was necessary to find out the proper methods of culturing Azolla round the year to ensure the supply of inoculum and the adjustment of the growth of Azolla simultaneously with a rice crop. The technology Azobiofer is the product of our continuous target-oriented research on Azolla since 1978. This technology can be a model for most tropical and sub-tropical rice growing countries. This Azobiofer technology has been developed through five phases.

46

5.3 Phase I. Quantification of N transfer from Azolla to rice plants

The author has carried out numerous experiments for quantifying mineralisation of Azolla-N, transfer of Azolla-N to rice plants (pot study), denitrification losses of Azolla-N and finally making a balance sheet of Azolla-N. For definite quantification, 15N was used as the tracer. Azolla caroliniana was labelled with 15N by growing in 15N-enriched growth media. This programme was carried out by the author as a Commonwealth Scholar from 1978 to 1981 under the supervision of Professor W D P Stewart of the Department of Biological Sciences, University of Dundee, Scotland for PhD degree study for the thesis entitled “Biofertiliser and rice production – a 15N tracer study”. Total-N content in Azolla plant and in soil samples were determined by the microkjeldahl digestion and distillation method. The 15N contents were determined in VG Micromass 601 spectrometer (Winsford, Cheshire, England). Mian (1981) was possibly the first to prove definitely with 15N-tracer technique that rice plants received nitrogen from incorporated Azolla biomass. About 33% of the 15N from Azolla was found assimilated by the rice plants in 60 days (Mian 1984; Mian 1985a; Mian and Stewart 1984). A method was developed (Mian 1985b) for direct quantification of N2 by mass spectrometer determining the extent of dinitrification loss of nitrogen from incorporated Azolla biomass as N2. The 15N – labelled Azolla biomass and ammonium sulphate were used to prepare a nitrogen balance sheet (Mian and Stewart 1985a; 1985b). These were high precision pioneer works in this field establishing benefits from Azolla as biofertiliser (Table 1). 5.4 Phase II. Establishment of Azolla as the source of N at field

level

A fortunate ‘follow-up’ of the previous greenhouse studies made possible by a joint venture of The Food and Agricultural organisation (FAO) and International Atomic Energy Agency (IAEA). The title of this internationally co-ordinated project was “Isotopic Studies of Nitrogen Fixation and Nitrogen Cycling by Blue Green Algae”, continuing from 1984 to 1989. Field experiments were conducted to test the pattern of N availability from Azolla to rice plants. In such cases, Azolla fronds were labelled with 15N by growth in 15N-labelled urea. Urea was also used as the fertiliser for rice culture to compare the effects with that of Azolla. The 15N concentrations in Azolla, rice grain and straw, and soil samples were analyzed in the Seibersdorf Laboratory of the IAEA in Vienna.

Table 1. Transfer of 15N from Azolla to rice plants (variety IR8) during 60 days of growth in pots

15N applied 15N assimilated by rice plants

Source Rate (mg pot-

1) (mg pot-

1) % of total 15N applied

1.33 0.44 33 1.99 0.65 33

Azolla

3.98 1.37 34 1.85 1.11 60 2.77 1.68 61

Ammonium sulphate

5.54 3.35 61

Source: Mian and Stewart (1985b)

47

Table 2. Field verification of 15N transfer from Azolla to rice plants (1m x 1m field plots)

15N uptake (grain + straw) Treatments 15N applied

(mg m–2) mg m–2 % of total applied

Azolla 166.5 91.4 55 Azolla 88.6 30.4 34

Urea (2 splits) 64.8 17.2 27 Urea (3 splits) 24.2 7.7 32

Source: Mian (1990, 1993). These field trials with 15N-labelled Azolla helped establish that significant amounts of N from Azolla are taken up by the growing rice plants (Table 2). Field studies have proven clearly that the use of Azolla was comparable to the use of urea as the source of N, even better in some cases (Mian 1990; Mian 1991a; Mian 1992; Kumarasinghe and Eskew 1993). 5.5 Phase III. Field trials on farms to test Azolla as nitrogenous

biofertiliser for lowland rice

The findings of phases I and II provided the basis for the idea of applying Azolla at farm level as an alternative or a supplement to chemical nitrogenous fertiliser urea for cultivation of lowland rice. For this purpose, a short-term plan was made to test different ways of growing Azolla in the field plots, different methods of applying Azolla for cultivation of rice and the availability of other nutrients to rice plants from Azolla. Various methods of incorporating Azolla in dry and wet were also tested. Performance of Azolla in terms of increasing rice yields was compared with that of the recommended doses of urea. The results of these experiments clearly indicated the potential position of Azolla as biofertiliser for lowland rice cultivation on farms. These experiments tested: (i) the effects of incorporated or unincorporated Azolla biomass on the yield of rice; (ii) different methods of growing and incorporating Azolla; (iii) the differences between the effects of Azolla and urea on modern rice varieties; (iv) different timings of Azolla inoculation and biomass incorporation with respect to cultivation of irrigated rice; and (v) the uptake of P, S and Zn by rice plants in addition to N. Incorporation of two layers of Azolla produced either similar or better rice yields compared to applying 60 kg N ha-1 as urea (Table 3). These results have been reported elsewhere (Mian 1990, 1991b, 1993; Kumaransinghe and Eskew 1993). It is seen in Table 4 that higher amounts of Azolla biomass (18.2-22.9 t ha-1) can be produced in 21 days within 5-25 days after transplanting (DAT) containing 38.1-48.2 kg N ha-1. The second layer formation was found to be of poor effectiveness probably due to lack of space caused by closing-up of rice canopies.

48

Table 3. Effect of Azolla on rice grain yield in dry season

Year Variety Treatment Grain yield (t ha-

1)

% increase over control

Control 3.02 - Azolla 4.53 50

BR3

Urea 4.15 37 Control 4.02 - Azolla (2 layers) 5.30 32

1990

BR11

Urea (3 splits) 5.32 32 Control 3.03 - Azolla 5.75 90

BR3

Urea (60 kg N ha-1) 4.66 54 Control 4.00 - Azolla (2 layers) 6.33 58

1993

BR2

Urea (60 kg N ha-1) 5.67 42

Source: Mian (1990, 1993).

Table 4. Growth of Azolla simultaneously with irrigated rice plants in dry season, 1996

Inoculum size (kg m-2)

Azolla layers

Growth stage Fresh biomass (t ha-1)

Total N accumulation (kg ha-1)

0.1 1st 5-25 DAT (21 days)

22.9 48.2

2nd 26-59 DAT (25 days)

7.7 16.2

0.2 1st 5-25 DAT (21 days)

18.2 38.1

2nd 26-59 DAT (25 days)

7.3 15.3

DAT = Days after transplantation of rice plants. Dry season = December to to May.

Mian and Azmal (1989) also pioneered in proving that Azolla can supply significant amount of P (about 28% of Azolla P) for uptake of rice plants. Mian (1991) confirmed these results in a second study. Table 5 presents some of the results regarding the availability of S and Zn from incorporated Azolla biomass to the growing rice plants. It is seen that 21-27% of S and 17-21% of Zn from Azolla were available to rice crop (variety BR2).

49

Table 5. Availability of S & Zn from Azolla to rice (variety BR2)

Total S uptake, kg ha-1 Total Zn uptake, g ha–1

Observations 80 kg N ha–1

as urea

80 kg N ha–

1

as Azolla

80 kg N ha–

1

as urea

80 kg N ha–

1

as Azolla

Total amount applied 0 16.4 0 266

Total uptake (grain + straw)

7.4 10.8 325 381 Dry

season

1990 %Recovery

- 21 - 21

Total amount applied 0 12.4 0 335

Total uptake (grain + straw)

22.5 25.8 364 421 Dry

season

1993 %Recovery

- 27 - 17

(Uptake in Azolla-treated plants – uptake in urea-treated plants) x 100 %Recovery = Total amount applied 5.6 Phase IV. Modelling of the technology

A. Model tests for establishing the integration of Azolla as biofertiliser of irrigated rice. The following work was performed in this phase : (i) To test the rate of growth of Azolla cultivated simultaneously with rice during the first 30 days

of rice growth while rice fields were kept irrigated. Thus, neither extra land nor extra irrigation was necessary.

(ii) To test the effects of one or two layers of Azolla alone or in combination with different doses of urea-N on the yield of rice.

(iii) To identify the best combination of one layer of Azolla with remaining N applied as urea to achieve a satisfactory yield of rice.

(iv) To trace any possible changes in organic matter status of soils due to receiving Azolla biomass.

(v) Fine tuning to discover the most suitable Azolla and urea combination for adoption at farm level.

A critical review of all previous work was carried out to develop a possible model for the technology. The observations made were – (i) Biomass to be used as biofertiliser should be produced by growing Azolla simultaneously with the respective rice crop to avoid the need for extra land, extra irrigation and extra management; (ii) the size of initial Azolla inoculum should be 0.2 kg m-2; inoculation should be done within 5-10 days after transplanting (DAT) and incorporation of the layer should be done within 25-30 DAT; (iii) growing and incorporation of one layer of Azolla within 30 DAT should be the practice and the rest amount of N should be applied as urea; (iv) the organic matter

50

status of the soil under Azolla treatment was slightly higher in contrast to the decreasing trend in the urea-treated soils. This is an essential for tropical and subtropical rice soils where organic matter is very low and has gradually been decreasing. The model of one layer of Azolla incorporation within 30 DAT plus addition of different doses of 25-75% of urea-N has now been under trials for 4-5 years. Some of the results are shown in Table 6. It appeares that growing and incorporation of one layer of Azolla can reduce the need for applying urea by about 30-40%. Some positive trend in the status of soil organic matter has been observed (Table 7).

Round the year production of Azolla Various combinations of triple super phosphate (TSP), muriate of potash (MOP) and cow dung (CD) were tested for satisfactory growth of Azolla in ponds all the year round to ensure the supply of inoculum. Trials were also conducted to find out the economic ways of producing Azolla by the farmers in their household ponds. Cost-benefit ratios were calculated for large-scale production of Azolla in a bid to ensure benefits for the year round supply of Azolla inoculum. Different rates of TSP, MOP and cow dung; their combinations; and different frequencies of their applications have been tested to assess their effects on the growth of Azolla. In particular, the guarantee of the growth of Azolla year round is so essential that the existence of the technology on Azolla is directly dependent on it. Fortunately, the application of 10 kg TSP ha-1 day-1 + 50 kg cow dung ha-1 day-1 has been found the best treatment in producing Azolla profitably round the year (Table 8). If a farmer produces only Azolla in his pond, it will be highly profitable, the estimated cost-benefit ratio being 1: 2.8 (Table 9).

Integration of Azolla to rice-fish farming system An extra dimension to the utilisation of Azolla as biofertiliser is to fertilise both fish (as feed) and rice (as the source of N). This is now evidence that Azolla is a good fish food and that, fish be successfully cultivated in the rice fields. Few attempts have been made to integrate Azolla into a rice –fish farming system, where fish feed on Azolla, the resultant fish excreta released nutrients for both rice and phytoplankton, with greater economic return from the production of rice and fish together. Further more, it may be possible to further reduce the use of urea. This approach would emerge as a different technology altogether. Azolla is good feed for fish. It is so acceptable to some fish that Azolla cannot survive in ponds with fish of more than 3-months old. However, it has been proven that Azolla can be grown in the nursery ponds for rearing fish fries and fingerlings; the cost-benefit ratio is 1: 4-5 in most cases. The most interesting information is that it is possible to cultivate rice and fish together in a system where Azolla will act both as the feed of fish and as biofertiliser for rice whereby the use of urea can be reduced by about 50% at least (25% in many cases). In such a system, each rice field will have a ditch along one side covering 25% of the total field, which will act as the fish refuse tank. Azolla will grow both in the ditches and in the rice fields. Under this Azolla-based rice fish farming system the cost-benefit ratio was found to be 1: 4.7. Hopefully, a new technology will emerge from this endeavour. Integrating fish and Azolla into rice-duck farming in Asia has been reported by Cagauan et al. (2000). 5.7 Phase V. Finalization of the technology model

A preliminary model for year round production of Azolla and utilisation of Azolla as a supplement to urea fertiliser for cultivation of irrigated rice was developed on the basis of our assembled information. The package of technology has now been finalised for application after few years of field trials. This package has been developed using information refined through model tests.

51

A. Year round production of Azolla inoculum in ponds i. Select any pond of any size [preferably of 20-50 decimals (1 decimal = 40 square meter)];

remove all weeds ; remove all fishes by net (use 20-36 g rotnone per decimal per foot of water depth if necessary. In such case do not use water for household purposes for a week). [Alternatively the pond can be dried by removing water to make it weed and fish free].

ii. Dissolve lime in sufficient water, spread the dissolved lime evenly over the water surface; the rate will be 1kg lime per decimal.

iii. After three days of liming add decomposed cow dung as 4 kg decimal-1 and poultry/duck manure as 2 kg decimal-1 or 8 kg decomposed cow dung decimal-1.

iv. After three days of manuring irrigate the pond (if water was removed). Try to keep 1-1.5 m of water depth.

v. Apply 10 kg triple superphosphate (TSP) ha-1 and 50 kg cow dung ha-1. Dissolve TSP in water and spread over the water surface.

vi. Inoculate Azolla at 0.5 kg m-2. vii. Continue application of 10 kg TSP ha-1 and 50 kg cow dung ha-1 everyday. viii. Harvest Azolla at each alternate days keeping the initial inoculum of 0.5 kg m-2 for growing.

Table 6. Integration of Azolla as biofertiliser into irrigated rice cultivation system

Year & Rice

Variety

Treatments Grain yield % increase over control

Az-1L (25 DAT)+50 % of urea-N 4.82 b 62 1996 BR2 100% of urea-N 5.28 a 77

Az-1L (25 DAT)+50 % of urea-N 4.19 a 51 1998 BR29 100% of urea-N 4.22 a 52

Az-1L (30 DAT)+50 % of urea-N 3.22 b 39 Az-1L (30 DAT)+75 % of urea-N 3.74 a 61

1999 BR21

100% of urea-N 3.26 b 41 Az-1L (30 DAT)+50 % of urea-N 4.10 b 46 Az-1L (30 DAT)+75 % of urea-N 5.50 a 96

2000 BR29

100% of urea-N 5.00 a 80 Az-1L (30 DAT)+50 % of urea-N 5.40 ab 82 Az-1L (30 DAT)+70 % of urea-N 6.00 a 102

2000 BR29

100% of urea-N 5.58 ab 88 The values have been extracted from different research reports of the author. The figures having a common letter do not differ significantly at 5% DMRT. Az-IL (25 DAT)=One layer of Azolla grown with rice crop and incorporated at 25 days after transplantation.

52

Table 7. Improvement of soil organic matter status due to addition of Azolla biomass

Treatments %

Organic matter

% Increase / decrease over initial status

Az-1L (25 DAT) + 50 % of urea-N 1.907 + 1.38 100 %of urea-N 1.823 - 3.08

1996 Initial status 1.881

Az-1L (25 DAT) + 50 % of urea-N 1.74 + 0.58 100 %of urea-N 1.71 - 1.16 199

8 Initial status 1.73 Az-1L (25 DAT) + 50 % of urea-N 3.19 +0.31 100 %of urea-N 3.09 -3.02 199

9 Initial status 3.18 Az-1L (25 DAT) + 50 % of urea-N 3.11 +0.65 100 %of urea-N 3.07 -0.65 200

0 Initial status 3.09 Table 8. Round the year production of Azolla in ponds (Data of June 1999 to May 2000)

Azolla biomass production (kg decimal-1) under two treatments (1 decimal = 40 square meter)

Month

Treatment 1: (10 kg TSP + 5 kg MP) ha-1 day-1

Treatment 2: (10 kg TSP + 5o kg cowdung) ha-1 day-1

January 120.2 123.9 February 178.5 181.0 March 181.9 182.1 April 143.9 150.1 May 138.3 134.8 June 149.3 144.4 July 153.9 181.2 August 201.8 206.0 September 139.5 173.2 October 115.0 126.0 November 169.5 206.0 December 100.2 110.8

53

Table 9. Cost benefit analysis for Azolla production in one hectare

Output Cost Amount (Taka) Azolla

biomass (kg)

Price (Taka kg-1)

Gross income (Taka)

Net income (Taka)

Cost benefit ratio

Labour 39,500.00 Materials 57,313.00 Overhead 30,070.00 Total 126,883.0

0

479,875.00

1.00 479,875.00

352,992.00

2.8

1 US$ = 56.00 Taka; material cost = costs for seed, fertilisers and pesticides; overhead cost = interests on input cost & land value, and miscellaneous costs.

B. Use as biofertiliser for irrigated rice i. Prepare the land for transplanting irrigated rice (in case of Bangladesh it is a cool and dry

season for cultivation of modern rice varieties with irrigation). Apply the basal doses of P, K, S, Zn and other nutrients.

ii. Divide the land into 3-5 decimal plots with temporary mud-bunds for easier management of water and Azolla growth.

iii. Transplant the rice seedlings as per standard spacing (25 cm from row to row and 15 cm from plant to plant) or any spacing normally followed. Irrigate the plots to a depth of 5-6 cm (no harm if more water is added).

iv. Inoculate Azolla as 0.2 kg m-2 (no harm if used more) within 5-10 days after transplanting (DAT).

v. Keep growing until a thick mat of Azolla is formed within 25-30 DAT; stop irrigation around 20-25 DAT so that water level is reduced and Azolla mat touches the soil; incorporate Azolla biomass by hand and feet or by a weeder.

vi. Irrigate the plots after 2-3 days if necessary or follow the normal practice. vii. Apply 60-70% of the urea-N in 3-splits i.e. at 15-20, 40-45 and 55-60 DAT. 5.8 Acknowledgements

The author is grateful to the Commonwealth Commission for awarding him Commonwealth Scholarships for PhD and Post-doctoral research on Azolla; to Professor W D P Stewart FRS, Department of Biological Sciences, Dundee University, Scotland, for guidance and support; to Professor Z H Bhuiya, Department of Soil Science, Bangladesh Agricultural University, Mymensingh for supports; to FAO/IAEA Division and the Ministry of Science and Technology, Peoples Republic of Bangladesh for funds. 5.9 References

Cagauan AG, Branckaert RD, Van Hove C (2000) Integrating fish and azolla into rice-duck farming in Asia. Naga, the ICLARM Quarterly 23, 4-10.

Khan MM (1988) A primer on azolla production and utilisation in agriculture. (IBS-UPLB & SEARCA, the Phillipines).

Kikuchi M, Watanabe I, Haws LD (1984) Economic evaluation of Azolla use in rice production. In ‘Organic Matter and Rice’. pp.569-592. (International Rice Research Institute, Los Bannos, Philippines).

54

Kumarasinghe KS, Eskew DL (1993) Isotopic studies of Azolla and nitrogen fertilization of rice. Developments in Plant and Soil Sciences 51, 1-145. (Kluwer Academic Publishers, Dordrecht, Netherlands).

Lumpkin TA, Pluckmett DL (1982) Azolla as a green manure : use and management in crop production. West View Tropical Agriculture. Series No. 5, Boulder Col, USA, p. 230.

Mian MH (1981) Biofertiliser and rice production - A 15N tracer study. PhD Thesis, Department of Biological Sciences, The Dundee University, Scotland, United Kingdom.

Mian MH (1984) A 15N tracer study to differentiate nitrogen supply to flooded rice plants by Azolla and Anabaena during their early and later stages of decomposition. Indian Journal of Agricultural Science 54, 733-738.

Mian MH (1985a) Relative nitrogen supply from 15N-labelled Azolla and blue-green algae to IR8 rice grown in pots under flooded conditions. Philippines Agriculturist 68, 415-423.

Mian MH (1985b) Detection of denitrification, by 15N tracer technique, of nitrogen released from Azolla and blue-green algae in flooded soils. Australian Journal of Soil Research 23, 245-252.

Mian MH (1990) Effects of 15N-labelled Azolla and urea on the yield of rice. Bangladesh Agricultural University Research Progress 4, 141-145.

Mian MH (1991a) Nitrogen, phosphorus and potassium uptake by flooded rice plants from incorporated Azolla. Progressive Agriculture 2, 75-80.

Mian MH (1991b) Response of BR11 rice to 15N-labelled Azolla and urea incorporated in Boro season. Bangladesh Agricultural University Research Progress 5, 240-245.

Mian MH (1992) Residual effects of Azolla and urea on the yield of rice. Bangladesh Agricultural University Research Progress 6, 133-139.

Mian MH (1993) Prospects of Azolla and blue-green algae as nitrogenous biofertiliser for rice production in Bangladesh. In Advances in Crop Science. Proceedings of The first Biennial Conferencer, Crop Science Society of Bangladesh. (Eds. L Rahman and MA Hashem). pp. 426-442.

Mian MH, Azmal AKM (1989) The response of Azolla pinnata R. Brown to the split application of phosphorus and the transfer of assimilated phosphorus to flooded rice plants. Plant and Soil 119, 211-216.

Mian MH, Kashem MA (1995) Comparative efficiency of some selected methods of applying Azolla for cultivation of irrigated rice. Bangladesh Journal of Crop Science 6, 29-36.

Mian MH, Stewart WDP (1984) A study on the availability of biologically fixed atmospheric dinitrogen by Azolla-Anabaena complex to the flooded rice crops. In Proceedings of The First International Workshop on “ Practical Application of Azolla for Rice Production’’. (Eds. W S Silver and EC Schroder) pp. 168-175. (Martinus Nijhoff Publishers, The Netherlands).

Mian MH, Stewart WDP (1985a) Fate of nitrogen applied as Azolla and blue-green algae (Cyanobacteria) in waterlogged rice soils – A 15N tracer study. Plant and Soil 83, 363-370.

Mian MH, Stewart WDP (1985b). A 15N tracer study to compare nitrogen supply by Azolla and ammonium sulphate to IR8 rice plants. Plant and Soil 83, 371-379.

Moore AW (1969) Azolla: biology and agronomic significance. Botanical Reviews 35, 17-30. Singh AL, Singh PK (1990) Intercropping of Azolla biofertiliser with rice at different crop geometry.

Tropical Agriculture 67, 350-354. West RG (1953) The occurrence of Azolla in the British interglacial deposits. New Phytoogy 52, 267-

271.

55

6. The spermosphere model to select for plant growth promoting rhizobacteria

J. Balandreau 6.1 Abstract

The spermosphere model allows the study of spontaneous plant-bacterial relationships under laboratory conditions. Axenic seeds are germinated in the dark and studied for a duration not exceeding the heterotrophic phase of the plant development. This device allows isolation ande enumeration of nitrogen fixing (acetylene-reducing) bacteria, instead of using regular synthetic nitrogen-free media. The model has selected unusual though efficient nitrogen-fixing bacterial strains, useful for inoculation of annual plants. The rationale for the model is a large diversity of diazotrophic bacteria compete for colonisation of roots of annual plants in the field; only after some time do the best adapted bacteria out-compete less efficient ones and colonise the roots. These best adapted candidate PGPRs are those which eventually reach large population densities on adult plants andexhibit highest efficiencies of N-fixation under xenobiotic conditions, such as in the spermosphere model. When correctly standardized this model allows the comparison of acetylene reduction rates and selection the best candidate PGPR (plant growth promoting rhizobacteria) strains can be selected using a mathematical procedure. The strategy is then to inoculate these strains at sowing to avoid the competition phase and establish an effective early association with a well adapted and efficient N-fixer. Ideally, candidate strains should be selected from the microflora of the very same soil in which these strain would have to be used. This strategy has been used on several occasions with success. For example, for rice in Egypt the most efficient diazotroph was a strain of Azospirillum brasilense now included in a commercial inoculant called Cerealin. In acid sulphate soils of South Vietnam, the same procedure was followed and lead to the selection of strain TVV75 of Burkholderia vietnamiensis sp. nov. In the case of maize, the spermosphere model selected a strain of Azospirillum lipoferum, now developed into the product named Azo Green. Keywords: Spermosphere model, rhizobacteria. 6.2 Introduction

Many bacteria have been tried as seed inoculants to improve plant growth, but published papers are mostly devoted to the results of field experiments and very few of them give a clear rational about the choice of a particular strain, leading to the "magical bug" concept: fashionable strains, i. e. strains giving rise to the larger number of publications were implicitly considered as good candidates for inoculation. Strain Sp7 of Azospirillum brasilense is a good instance of a strain used for reasons that have no ecological basis. The consequence is that this strain, originally isolated from a Digitaria decumbens lawn in Brazil has been used to inoculate plants that it had never met before. With such a low probability to be well adapted to these new host plants it is no surprise that these types of trial never gave good reproducible results. In the Laboratory of Microbial Ecology of the Rhizosphere (LEMIR) located in Nancy (France) we developed a rational approach based on experimental results to elaborate a strategy for choosing bacterial candidates for seed inoculation. The basic data, hypotheses and concepts leading to this strategy may be summarised as follows: • plant rhizospheres can be colonized by a vast diversity of N-fixing taxa (Young 1992) • these bacteria differ greatly in their N-fixing efficiencies: generally, enterobacteria are poorly

efficient, whereas bacteria such as Azospirillum, Bacillus and Burkholderia can be very good

56

fixers when associated with the plant (Heulin et al. 1989). The distinction roughly coincides with the distinction between R- and k-strategies (De Leij et al. 1993).

• hypothesis 1: during the phase of establishment of the rhizosphere microflora, these taxa enter a strong competition for root colonisation. Better adapted taxa (k-strategists) ultimately win this competition and become abundant.

• hypothesis 2: efficiency of N-fixation is a symptom of good interaction with the host plant.

A basic rationale ensues: 1. the main role of inoculation is to ensure an early colonisation by these bacteria which would

anyway ultimately win the competition for root colonisation. There is no need to introduce an alien bacterium.

2. these potential winners can be found on adult healthy N-fixing plants 3. seed inoculation by these, suppress the competition phase, inducing the earlier formation of the

best plant-bacterium combination. The role of inoculation merely speeds up a natural process. This approach was first attempted in Southern France, in Camargue where it produced strain 4B of Azospirillum lipoferum, a very efficient N-fixer which, unfortunately, was assayed only once in situ (Charyulu et al. 1985). This approach has been more successfully used in three other instances: in South Western France with the isolation of strain CRT1 of Azospirillum lipoferum, a good maize colonizer which is the basis of Azo-Green, produced by Liphatech; in Egypt where strain NO40 of Azospirillum brasilense is now included in a commercialised product called "Cerealin" used by rice growers and in Vietnam with the very promising strain TVV75 of Burkholderia vietnamiensis, which Agrium considered for a commercial product before Environmental Protection Agency (EPA) banned Burkholderia based products. These successes are a posteriori proof of the validity of this strategy. Another reason for this success could be the use of a peculiar isolation procedure to obtain representatives of the diazotrophic flora. Instead of using regular selective media, we decided to use an exudate-based medium to maintain the plant selective pressure as long as possible during the isolation procedure: we called it the spermosphere model. As we have not performed a comparative study of this method versus others, its benefit is difficult to assess. Nevertheless, as this method is a common starting point of all these success stories we shall present it first in this brief summary of the strategy we employed to select for candidate plant-growth promoting rhizobacteria (PGPR) strains. 6.3 The spermosphere model

The spermosphere model (Thomas-Bauzon et al. 1982) is a gnotobiotic experimental model allowing the study of plant-bacteria relationships under laboratory conditions. Axenic seeds are germinated in the dark and inoculated with the bacteria being studied. Inoculation must be delayed until the coleoptile is long enough (at least 1 cm) to avoid drowning the plant in the inoculum. Inoculated cell numbers must be low enough to avoid killing the seeds (probably through oxygen competition), i.e. in the range of 106 to 107 bacteria per ml, maximum. Hypochlorite surface sterilisation of seeds is very efficient but can cause toxicity and mutagenesis even after many rinses. Destroying chlorine remnants after sterilisation through 2% sodium thiosulfate rinses seems able to avoid these artefacts (unpublished data). The spermosphere model (Figure 1) is made of a test tube with a lateral arm reminiscent of the Pankhurst tube (Pankhurst 1967) used to grow anaerobic bacteria. The side arm contains 2ml of 1N sodium hydroxide (in the case of rice) to trap excess CO2 generated in large amounts by the germinating seed. During the germination process, the seed exudes large amounts of organic compounds. In the case of rice, at 30°C this amounts to approximately 10% of seed weight in 10 days. There cannot be any photosynthesis under these conditions so the tubes are kept in the dark. Exudates are due to the mobilisation of carbon storage products in the seed (Heulin et al. 1987). This means that the system cannot be used for long durations; in the case of rice 10 days is a maximum. The culture medium in

57

the main tube is an N-free medium to suit N-fixing bacteria. The medium is made semi-solid (0.5% agar) to allow good growth of roots and a convenient oxygen gradient for N-fixing bacteria. Six ml of medium per tube were enough in the case of rice. Several media designed for N-fixers have been used with success. Indeed they must be prepared devoid of any carbon source: the only carbon source for bacteria in the spermosphere model consists of the compounds exuded during germination. We commonly used the following medium modified after Weaver's medium for Rhodopseudomonas capsulata (Weaver et al. 1975): Solution A (g L-1): ZnSO4.7H2O, 0.42; MnSO4.H2O, 1.30; NaMoO4.2H2O, 0.75; H3BO3 , 2.8; CuSO4.7H2O, 0.026; CoSO4.7H2O, 0.07. Solution B (g L-1): MgSO4.7H2O, 2.00; CaCl2.2H2O, 2.00; FeSO4.7H2O, 0.44; EDTA O.40; solution A, 20 ml Solution C (g L-1): K2HPO4, 90; KH2PO4, 60. Final medium: solution B, 50 ml; solution C, 15 ml; distilled water 1 L. Tubes are cotton plugged and incubated in the dark until coleoptiles are approximately 1 cm high.

Figure 1. Spermosphere model. The medium is semi-solid and devoid of N and C. The side arm contains 2 ml of 1N NaOH to trap CO2. Tubes are kept in the dark. Rubber stoppers are used (Suba Seal type)

58

6.4 Isolation of bacteria

This device can be used first to enumerate and isolate nitrogen fixing (acetylene-reducing) bacteria, instead of a regular synthetic nitrogen-free medium. For most probable number (MPN) counting soil dilutions are inoculated in spermosphere models and 1% acetylene (C2H2) injected. This concentration is high enough to compete with nitrogen so that any nitrogenase activity is revealed by the formation of ethylene. Nevertheless this C2H2 concentration is low enough to allow the reduction of some nitrogen as well, providing combined nitrogen to ensure the growth of diazotrophs present in the inoculated soil dilution. As a consequence, after one week, tubes which had received a soil dilution containing N-fixing bacteria have evolved a measurable amount of ethylene. The numbers and distribution of ethylene-containig spermosphere models allow the calculation of most probable numbers of diazotrophs. Ethylene positive tubes can then be used for the following step. The contents of ethylene positive tubes are macerated and diluted in the same way as soil dilutions, and plated on a regular medium for diazotrophs. Subsequent purification comprises alternate streakings on N-free and complete media untill purity can be assessed (Omar et al. 1989). Somewhat unusual bacteria have been obtained using this procedure; for example, N-fixing Sphingomonas paucimobilis strains, a non-motile Azospirillum strain (Bally et al. 1983). 6.5 Comparison of strains

In the next step, strains are compared for their N-fixation efficiency when associated with the host plant. Spermosphere models are inoculated with a high density of cells of the strain under study to be sure that cell density is not the limiting factor of nitrogenase activity: final bacterial concentration after inoculation is 107 per ml. Beforehand, inoculated cells are grown overnight in a rich medium, centrifuged and washed twice in 0.85% KCl. The side arm receives 2 ml of 1N NaOH, and both arms are closed by rubber stoppers (Suba seal type). 10% acetylene is injected. This concentration is necessary to obtain a quantitative measurement of nitrogenase activity, whereas in the isolation procedure a qualitative estimate is sufficient. It is advisable to use an internal tracer gas to improve quantitation. For that purpose we used propane; injecting a precisely known quantity at the beginning of incubation allows us to relate the measured ethylene concentrations to this initial propane concentration, making unnecessary to correct for leaks or changes in the internal pressure. Incubation is conducted in the dark. Gas analyses are performed after 3, 5, and 7 days by gas chromatography. At the end of the incubation tubes are checked for purity by plating on a nutrient rich medium. Data from contaminated tubes are discarded. At time zero the amount of ethylene contained in acetylene is negligible and so is the trace amount of ethylene generated by the plant, as compared to what is evolved from nitrogenase activity. For each tube 4 ethylene concentrations are obtained, including 0 at time 0. Ethylene increases are calculated for the three time intervals. Figure 2 shows that replicate tubes greatly differ by the time lag before onset of ethylene evolution, whereas the rate of ethylene evolution is reproducible. It is constant for several days during which C supply is the limiting factor of nitrogenase activity (any injection of extra C supply during this period results in an increase in ethylene evolution). As a consequence, each tube is characterised by its maximum ethylene evolution rate. Depending on the lag it can be any of the three recorded values. Most often, but not always, it is the rate measured between day 5 and 7. Let us call this maximum rate of ethylene evolution (Rmax). Large numbers of replicates are advisable. As the limiting factor is the C supply through exudates, it is essential to use seeds of a very uniform size. It is preferable to do a preliminary study of seed weight distribution and use only seeds of the modal class.

59

Figure 2. Reproducibility of ethylene evolution in spermosphere model. The three replicate tubes differ by the time lag, but the maximum rates of ethylene evolution (Rmax) are roughly the same: 0.8, 0.9 and 1.1 micromoles C2H4 per tube per day 6.6 Mathematical procedure used to compare strains

We report here a comparison of 20 N-fixing strains isolated from the rhizosphere of rice growing on an Egyptian soil (Heulin et al. 1989). Due to the amount of work involved it was not easy to compare more than four strains at a time. We had to compare the 20 strains in nine series. To avoid series to series differences, a common reference strain (Enterobacter cloacae NO13) was included in all series, and (Rmax )s measured for each strain could be related to (Rmax )c, the value of the common reference strain in this particular series. We found a log transformation of data to be necessary: Xs = log (Rmax )s was thus obtained for each strain along with the corresponding Xc = log (Rmax )c obtained for the control in the same series. A global variance analysis showed that a significant strain effect was visible beyond a series effect. The procedure described by Dunnett (1964) was then followed. The comparison is obtained by the evaluation of

Z = (Xs - Xc) / s (1/ns + 1/nc)0.5

in which s is the residual standard deviation and ns and nc are the number of replicates for strain S and its control C. The value Z is then compared to the Λ value given by Dunnett's tables which take into account the number of replicates and the desired significance level. If IZI > Λ strain S is significantly less efficient (negative values of Z) or more efficient (positive values of Z) than strain C. If IZI < Λ the strains are not significantly different. Furthermore, using Dunnett's approach it is possible to calculate a confidence interval for Xs - Xc = log (Rmax )s - log (Rmax )c = log (Rmax )s / (Rmax )c . Its value is ci = Λ s(1/ns + 1/nc)0.5 Table 1 shows an instance of strain comparison in the case of rice rhizosphere isolates obtained from Moshtoor, an experimental agronomy station situated in the Nile delta, in Egypt. All isolates were Enterobacteriacae except NO40 (Azospirillum brasilense) and NO44 (Azospirillum sp.). In this study we also included strain 4B of Azospirillum lipoferum isolated in a Camargue rice field (South of France) strain FS (a non rhizosphere Azospirillum isolate from the Philippines) and Alcaligenes

0

1

2

3

4

5

0 1 2 3 4 5 6 7 8

Time (days)

60

faecalis A15. Most strains were not significantly different from the control. Three strains were significantly less efficient: NO23, A15 and NO44. Strain NO44 was a very poorly efficient Azospirillum strain which, unfortunately, was subsequently lost. The most active strains were 4B and NO40. NO40 has been used in the following years to do a series of field inoculations in the Nile delta. In Vietnam, the same procedure yielded strain TVV75 of Burkholderia vietnamiensis as a very efficient rice associated diazotroph (Tran Van et al. 1996). In the case of maize, this selection strategy has been followed by Fages and Mulard (1988), working for Pioneer-France-Maïs. From a mature maize plant in the field they obtained strain CRT1 of Azospirillum lipoferum. This procedure has been used to compare different plants associated with the same bacterial strain. Different cultivars or mutants of rice associated with A. lipoferum 4B showed significantly different levels of nitrogenase activity (Charyulu et al. 1985). 6.7 Fate of selected strains

Field experiment with rice Strain NO40 of Azospirillum brasilense has been used in six field inoculation trials in the Nile delta. One of these experiments is of special interest as it has been performed in Moshtoor, the very same place where this strain had been isolated. Two controls have been done: a non-inoculated control and a control in which rice was inoculated by a suspension of killed bacteria. In that place, inoculation had a significant effect on yield. Yield rate was 1.01 kg paddy m-2 in the inoculated treatment and 0.83 and 0.84 kg paddy m-2 in the two controls, respectively. This support the initial hypothesis that the role of inoculation was to speed up a natural process, not to introduce an alien strain. Field inoculation was done five times in the same plots in Sakha. In the experimental set up the effect of inoculation was assessed at three levels of nitrogen fertiliser (recommended rate, half this rate and a control without added nitrogen). Experiments gave increases in yield when nitrogen was the limiting factor, i. e. whenever the yield of the non-inoculated control was increasing with the rate of applied nitrogen. In two instances nitrogen had no effect on yield, in these two cases, inoculation had no effect either (Omar et al. 1989, 1992). Strain NO40 is now part of a commercial inoculant called Cerealin.

61

Table 1. comparison of rice rhizosphere isolates

Nc Strain ns s df Λ Z (Rmax )s /(Rmax )c

Conf. interv.

18 NO25 16 0.8 32 2.04 0.77 1.23 17 NO1 17 0.7 59 2.41 1.60 1.49 NO3 15 -0.73 .97 NO27 15 1.23 1.34 14 NO28 15 0.9 58 2.51 1.11 1.34 NO29 15 -0.63 .84 NO31 18 1.73 1.52 15 NO11 16 0.8 29 2.04 1.14 .88 18 NO7 19 0.8 63 2.41 0.39 1.07 NO36 17 -1.64 .60 NO37 13 -1.25 .87 15 NO23 17 0.7 44 2.29 -3.92 .09* 0.2-0.7 NO24 15 -1.47 .73 7 NO33 12 0.5 43 2.44 -0.94 .87 NO22 14 -0.64 .92 NO9 14 1.62 1.51 5 4B 11 0.4 28 2.32 4.34 2.65* 1.5-4.1 NO40 15 5.80 3.69* 2.1-5.5 11 NO42 12 0.8 39 2.54 0.86 1.35 NO44 12 -6.55 .12* 0.05-

0.25 FS 12 0.12 1.03 A15 8 -11.01 .04* 0.01-

0.06 The rice cultivar used was Giza 171. The common control was strain NO13 of Enterobacter cloacae. This strain gave an average value of 853 ± 22 nanomoles C2H4 per tube per day (± the standard deviation). df = degree of freedom. Significance (P = 0.05) is marked by *. In acid sulphate soils of South Vietnam, the same procedure was followed and led to the selection of strain TVV75 of Burkholderia vietnamiensis sp. nov. The strain has been used with success in 17 field experiments, in cooperation with AgBiologicals. Figure 3 shows an instance of field results. In this case, N was limiting the yield (yield increased with increasing N fertilisation in the control), and there was a significant effect of inoculation. The same yield (4.4 t ha-1) can be obtained either under regular practice, using 68 kg N ha-1 or using inoculation and only 43 kg N ha-1. The NFEI (nitrogen fertilizer equivalent of inoculation) is thus 25 kg N ha-1. Given the inoculum cost and the cost of N fertilisers, the NFEI value is an help to reach a decision about the feasibility of inoculation (Tran Van et al. 2000). Unfortunately this strain could not be registered given its relatedness to B. cepacia for which health hazards are known

Field experiment with maize Fages and Mulard (1988) used the spermosphere model to isolate dominant diazotrophs from the rhizosphere of an adult plant (flowering stage) and obtained strain CRT1 of Azospirillum lipoferum as a good candidate for inoculation. Large numbers of field tests followed, all over the world. Subsequently Pioneer France-Maïs stopped this research and sold it to Liphatech, who developed a CRT1 based inoculum named Azo Green.

62

Figure 3. Effect of bacterial inoculation on grain yield of rice under variable applied N levels. Bình Chanh field experiment, autumn season, rice cv. Nàng Thom inoculated by B. vietnamiensis TVV75 6.8 Conclusion

The approach developed in the course of this work was based on the ecology of rhizosphere dwellers. The strategy was not to introduce a bacterium absent from the soil, as is the case most often for legume inoculation. Instead it was chosen to hasten the colonisation of roots by a local well adapted strain making an efficient use of exudates in terms of N-fixation. Selecting a strain to inoculate amongst the local microflora was the first choice. Using the rate of acetylene reduction in spermosphere model as a criterion was another choice. At first, it was done considering that N-fixation was the mechanism to select for. Now it is rather considerd a high nitrogenase activity as a criterion of « good » plant bacterium interaction. Many parts of the rationale still await demonstration. For instance the population of introduced strains was not monitored, so that the claim that simply speeding up the natural colonisation process has not been substantiated. Nevertheless this strategy has been followed by success in at least three different instances, and using a pragmatic approach, this is enough to decide to base on it for the future selection of new strains to inoculate. 6.9 Acknowledgments

This work has been done in the Centre de Pédologie of CNRS in Nancy between 1980 and 1990, with contributions from R Bally, K Barbouche, O Berge, P B B N Charyulu, T Heulin, J L Hubert, A M N Omar, Z Rafidison, M Rahman, L Rondro-Harisoa, P Villecourt, D Thomas-Bauzon, V Trân Van, P Weinhard-Lebourg. Outside collaborators were J Fages, F Fourcassié, A Guckert, R Marie, J C Pierrat. The author is grateful to all collaborators for their contributions. 6.10 References

Bally R, Thomas-Bauzon D, Heulin T, Balandreau J (1983) Determination of the most frequent N2-fixing bacteria in a rice rhizosphere. Canadian Journal of Microbiology 29, 881-887.

Charyulu PBBN, Fourcassié F, Barbouche K, Rondro-Harisoa L, Omar AMN, Weinhard P, Marie R, Balandreau J (1985) Field inoculation of Rice using in vitro selected bacterial and plant genotypes. In ‘Azospirillum III. Genetics, Physiology, Ecology’. (Ed. Klingmuller) pp. 163-179. (Springer-Verlag Publishers, Berlin, Germany).

De Leij FAAM, Whipps JM, Lynch JM (1993) The use of colony development for the characterisation of bacterial communities in soil and on roots. Microbial Ecology 27, 81-97.

Dunnett CW (1964) New tables for multiple comparisons with a control. Biometrics 20, 482-491.

1008060402003

4

5

6

Grain yield (t ha )

Inoculated

Control

Nitrogen (kg ha-1)

NFEI

-1

<

63

Fages J, Mulard D (1988) Isolement de bactéries rhizosphériques et effet de leur inoculation en pots chez Zea mays. Agronomie 8, 309-314.

Heulin T, Guckert A, Balandreau J (1987) Stimulation of root exudation of rice seedlings by Azospirillum strains: carbon budget under gnotobiotic conditions. Biology and Fertility of Soils 4, 9-14.

Heulin T, Rahman M, Omar AMN, Rafidison Z, Pierrat JC, Balandreau J (1989) Experimental and mathematical procedures for comparing efficiencies of rhizosphere N2- fixing bacteria. Journal of Microbiological Methods 9, 163-173.

Omar AMN, Heulin T, Weinhard P, Alaa-el-Din M, Balandreau J (1989) Field inoculation of rice by in vitro selected plant-growth-promoting-rhizobacteria. Agronomie 9, 803-808.

Omar AMN, Richard CL, Weinhard P, Balandreau J (1989) Using the spermosphere model technique to describe the dominant nitrogen-fixing microflora associated with wetland rice in two Egyptian soils. Biology and Fertility of Soils 7, 158-163.

Omar N, Berge O, Shalaan SN, Hubert JL, Heulin T, Balandreau J (1992) Inoculation of Rice with Azospirillum brasilense in Egypt: results of five different trials between 1985 and 1990. Symbiosis 13, 281-289.

Pankhurst ES (1967) A simple culture tube for anaerobic bacteria. Laboratory Practical 16, , 58-59. Thomas-Bauzon D, Weinhard P, Villecourt P, Balandreau J (1982) The spermosphere model: I. its’

use in growing counting and isolating N2-fixing bacteria from the rhizosphere of rice. Canadian Journal of Microbiology 28, 922-928.

Trân Van V, Berge O, Balandreau J, Ngo Kê S, Heulin T (1996) Isolement et activité nitrogénasique de Burkholderia vietnamiensis, bactérie fixatrice d'azote associée au riz (Oryza sativa L) cultivé sur un sol sulfaté acide du Viêt-nam. Agronomie 16, 479-491.

Trân Van V, Berge O, Ngô Kê S, Balandreau J, Heulin T (2000) Repeated beneficial effects of rice inoculation with a strain of Burkholderia vietnamiensis on early and late yield components in low fertility acid sulphate soils of Vietnam. Plant and Soil 218, 273-284.

Weaver PK, Wall JD, Gest H (1975) Characterisation of Rhodopseudomonas capsulata. Archieves of Microbiology 105, 207-216.

Young JPW (1992) Phylogenetic classification of nitrogen-fixing organisms. In ‘Biological Nitrogen Fixation’. (Eds. G Stacey, RH Burris, HJ Evans) pp. 43-86. (Chapman and Hall Publishers, London, United Kingdom).

64

7. Root stimulation and nutrient accumulation of hydroponically-grown tissue-cultured banana plantlets inoculated with rhizobacteria at lower level of nitrogen fertilisation M. A. B. Mia, Z. H. Shamsuddin, W. Zakaria and M. Marziah 7.1 Abstract

Banana (Musa spp.), the second most extensively grown fruit in Malaysia, which is usually grown by small land holders. Commercial cultivation requires a substantial amount of nitrogen (N) fertiliser, which is costly, and may cause environmental hazards. Biofertilisers could be an alternative source of nitrogen for sustainable banana production. A hydroponics experiment was conducted to observe the effects of plant growth-promoting rhizobacterial (PGPR) inoculation on root development and nutrient uptake of tissue-cultured banana plantlets at reduced level of N-fertilisation. The results showed that PGPR inoculation increased the cumulative primary root elongation and the secondary root initiation rate. Primary root length, root number, root volume, and root mass increased significantly in inoculated plants with and without N-fertilisation. Application of PGPR without N-fertilisation produced an equivalent root mass and volume as the 100% N-applied control plant. Concentrations of N, P, K, Ca and Mg were not influenced but total uptake of P, K, Ca and Mg increased significantly due to PGPR inoculation. Inoculation with the supplement of 33% N of full requirement could contribute an equivalent amount of N accumulation of 100% N-applied control plants. The results strongly indicated that PGPR could be used as a bio-enhancer and biofertiliser for banana seedlings production. Key words: Rhizobacteria, banana, hydroponics, roots, nutrient, growth. 7.2 Introduction

Banana is an important fruit crop, with domestic and export markets, contributing significantly to the national economy of many countries including Malaysia. In commercial cultivation it requires a substantial amount of nitrogen fertiliser, which is costly and may cause environmental hazards when leached into natural water bodies. Biofertiliser is globally accepted as an alternative source of nitrogen. It is claimed to be environmentally friendly and can ensure a sustainable banana production. Plant-growth promoting rhizobacteria (PGPR) (eg. Azospirillum spp., Bacillus spp.) are potential inoculant biofertilisers. The PGPR are rhizosphere bacteria those colonise plant roots and stimulate growth of host plants (Kloepper et al. 1980). They are also regarded as bio-enhancers due to their abilities to stimulate root development and to increase absorption of water and plant nutrients in bananas (Wange and Patil 1994; Shamsuddin et al. 1997). Azospirillum spp. have been widely reported to fix atmospheric nitrogen with grasses and cereals (Tarrand et al. 1978) and enhance plant nutrient uptake (Lin et al. 1983; Kapulnik et al. 1985; Murty and Ladha 1988; Bashan and Levanony 1990). A locally isolated strain, UPMB10 (Bacillus sp.), has been shown to produce beneficial effects on oil palm (Amir et al. 2001) and banana seedlings (Shamsuddin et al. 1997). PGPR inoculation along with 33% N of the full requirement could produce similar yield and biomass resulting the plants to less dependence on N fertiliser in sweet potatoes (Saad et al. 1999). Root development and nutrient uptake studies are the key factor to observe the beneficial effects of PGPR inoculation in bananas, which are easier in hydroponics system.

65

In the soil medium, extraction of undamaged roots, and ions remained attached to the soil colloids are major limitation for this kind of study. Hydroponics condition allows plants to grow more vigorously. In hydroponics, roots can develop uniformly, can be separated more easily with a minimum damage, and nutrients could be absorbed more easily. Therefore, this study was undertaken in hydroponics condition to investigate the PGPR inoculation on root stimulation and nutrient accumulation in bananas. 7.3 Materials and methods

The experiment was conducted in the shade house of Field 2, Universiti Putra Malaysia, Selangor, Malaysia under hydroponics condition, using the following nutrient solution (g L-1) modified from Clarkson et al. (1989): KNO3, 0.505; Ca (NO3)2.4H2O, 0.355; MgSO4.7H2O, 0.37; NaNO3, 0.17; KCl, 1.05x10-3; KH2PO4, 0.136; Fe-EDTA, 3.55x10-3; MnSO4.4H2O, 0.81x10-3; H3BO3, 0.57x10-3; ZnSO4.7H2O, 0.22x10-3; CuSO4.5H2O 0.04x10-3; (NH4)6 Mo7O24.4H2O , 0.02x10-3. The following treatments with five replications were used for the experiment: T1 (control: no N and no PGPR), T2 [no N + Azospirillum brasilense (strain Sp7))], T3 [no N + a locally isolated Bacillus sp. (strain UPMB10)], T4 (control, 100% N and no PGPR), T5 (33% N + Sp7) and T6 (33% N + UPMB10). PGPR strains Sp7 (Azospirillum brasilense) and UPMB10 were used in this experiment. PGPR strain Sp7 and UPMB10 were obtained from cultures maintained by Soil Microbiology Laboratory, Department of Land Management, Universiti Putra Malaysia. The strain Sp7 was originally provided by EMBRAPA, Brazil. Bacteria were cultured from a single colony into a stock culture. A 1 ml sample of bacterial suspension was then transferred into nutrient broth (Okon et al. 1977) in 250 ml Erlenmeyer flasks agitated to a rotary shaker (125 rpm, 24h, 30± 20C). Tissue-cultured banana plantlets cv. 'Berangan’ were used as the test plant, supplied by Professor Marziah Mahmood, Plant Tissue Culture Laboratory, Department of Biochemistry and Microbiology, Universiti Putra Malaysia. One tissue-cultured banana plantlet was transplanted into each plastic pot (4.0 L) containing nutrient solution. Forty ml broth cultures of PGPR strains Sp7 or UPMB10 with 107 cfu.ml-1 were applied to the respective pot prior to the transplanting process. The pots were wrapped with aluminium foil to prevent light effect from inhibiting root growth and aerated with air pumps at six hourly intervals to ensure an uninhibited root respiration and bacterial growth. The primary root elongation and secondary root initiation were measured one-day intervals until 15 d. Photographs of the whole plants were taken at six weeks after transplanting when the plants were terminally harvested. Roots parameters (primary root number, roots length, roots volume and roots dry weight) were measured. Plant parts were separated into roots, corms, stems and leaves. Separated plant samples were then oven-dried for 3 d at 700C and weighed. The dried samples were ground and digested with H2SO4 and H2O2 (Thomas et al. 1967). Nitrogen and P contents were determined using an auto-analyser while Ca and Mg contents were measured using an atomic absorption spectrophotometer. Total accumulation of nutrients were calculated according to the formula: accumulation = nutrient concentration x total dry matter of the plant. The data were analysed using the statistical analysis system (SAS Institute Inc. 1987). 7.4 Results

Banana roots Banana roots can be classified according to their origin viz. primary roots arise from the corm, secondary roots arise from primary roots and tertiary roots are those, which originates from the secondary roots. The primary roots can further be classified into two groups namely feeders and pioneers. Feeders are significantly longer but thinner than pioneers and have a higher density of secondary roots (Swennen et al. 1986).

66

Vertical bars denote LSD at 0.05 Figure 1. Effect of PGPR inoculation on root growth of tissue-cultured banana plantlets; A: primary root elongation; B: secondary root initiation

Primary root elongation PGPR inoculation positively influenced primary root elongation (Figure 1A). Plants inoculated without N application showed lower rate of root elongation compared to inoculated with N fertilisation. Inoculation with Sp7 plus 33% N showed higher root elongation compared to inoculation alone. Inoculation with UPMB10 plus 33% N showed the highest root elongation.

Secondary root initiation rate Secondary roots started to initiate at 3 DAI (days after inoculation). Inoculation with and without N fertilisation showed higher rate of root elongation compared to control (Figure 1B). Highest rate of secondary root initiation was observed in the treatment of Sp7 with 33% N application. Application of 100% N without inoculation showed the lowest rate.

Root number Total number of primary roots plant-1 was not significantly influenced by PGPR inoculation (Table 1). However inoculation with Sp7 and UPMB10 without N fertilisation gave 38%, and 17%, higher root number compared to control (without PGPR and without N fertilisation), respectively. Inoculation with 33% N gave slightly higher root numbers compared to un-inoculated with 100% N application.

Root length Total primary root (feeder and pioneer) length increased significantly due to inoculation with UPMB10 while Sp7 could not increase the total primary roots length significantly (Table 1). But feeder root length was positively influenced by PGPR inoculation both with UPMB10 and Sp7. Inoculation with UPMB10 also significantly increased feeder root length which are covered by secondary roots. There was no effect on pioneer root length due to PGPR inoculation. However inoculation with UMPB10 significantly increased the length of pioneer roots which are covered by secondary roots.

Figure 1A

0

5

10

15

20

25

0 1 2 3 6 8 10 12Days after inoculation

N0-PGPR N100%-PGPRN0+Sp7 N0+UPMB10N33%+Sp7 N33%+UPMB10

Figure 1B

-100

1020304050

0 1 3 6 8 10 12

Days after inoculation

N0-PGPR N0+Sp7N0+UPMB10 N33%+Sp7N33%+UPMB10 N100-PGPR

67

Table 1. Effect of PGPR inoculation on number and length of primary roots of hydroponically-grown tissue cultured banana plantlets

Primary root length (cm) Treatments Number of primary roots plant-1

Feeder roots

Feeder roots occupied by secondary roots

Pioneer roots

Pioneer roots occupied by secondary roots

Total primary root length

N0-PGPR 6.0b 170.7b 144.0b 65.5a 31.7b 236.2b N0+Sp7 8.3ab 262.7a 202.3ab 51.7a 22.0b 314.3ab N0+UPMB10 7.0ab 269.3a 269.3a 217.3

a 34.0a 341.8a

N100%-PGPR 8.6ab 269.0a 211.7a 103.1a

56.0a 372.1a

N33%+Sp7 9.6a 215.1ab 184.0ab 93.0a 57.0a 308.2ab N33%+UPMB10

9.6a 253.5a 213.8a 101.7a

55.3a 355.1a

Values having same letter(s) in a column do not differ significantly at 5% level by Duncan’s Multiple Range Test (DMRT).

Root base diameter This indicates the size and thickness of the roots. Inoculation with UPMB10 increased base diameter in feeder roots significantly whereas Sp7 could not contribute positively to the thickness of the roots (Table 2). Base diameter of pioneer root also increased due to inoculation with UPMB10 although the effect was not significant.

Root volume Remarkable increment in root volume was observed in inoculated plants compared to un-inoculated control (Table 2; Plate 1). Inoculation with UPMB10 and Sp7 without N application resulted in 142% and 59%, greater root volume, respectively. But application of N with PGPR inoculation could not show the similar trends over N application alone, although the treatments produced more root volume compared to 100% N without inoculation.

Root dry weight Root dry weights of inoculated plants increased significantly over control both with and without N application (Table 2). Inoculation with Sp7 along with N fertilisation gave the highest root dry weight.

Root to shoot ratio Nitrogen fertilisation showed lower root to shoot (R/S) ratio (Table 2). Inoculated plants without N fertilisation produced the lower R/S values compared to control although the effect was not significant.

Nutrient uptake Concentrations of N, P, K, Ca and Mg were not much influenced but total accumulation of different nutrients was significantly increased by PGPR inoculation [Figure 2 (A, B, C, D)]. Accumulation of N in inoculated plants without N application was slightly higher. Accumulation of P, K, Ca and Mg were increased significantly due to inoculation especially without N fertilisation. Inoculation with UPMB10 with 33% N showed significantly higher Ca and Mg uptake compared to 100% N-applied plant.

68

Inoculation of Sp7 with 33% N showed similar response to 100% N for Mg uptake. Among the elements, the highest uptake of K , followed by N, was noticed in all the treatments. Table 2. Effect of PGPR inoculation on base diameter, root volume, root dry weight and root to shoot ratio of hydroponically-grown tissue cultured banana plantlets

Base diameter (mm) Treatments Feeder Pioneer

Root volume (ml plant-

1)

Root weight (g plant-1)

Root to shoot ratio

N0-PGPR 0.13b 0.25c 11.3d 0.27c 0.54a N0+Sp7 0.17ab 0.27bc 18.0c 0.65b 0.33abc N0+UPMB10 0.23a 0.35abc 27.3b 0.61b 0.42ab N100%-PGPR 0.15ab 0.42a 50.0a 0.69b 0.13d N33%+Sp7 0.21ab 0.38b 55.0a 1.02a 0.22bdc N33%+UPMB10 0.19ab 0.42a 59.0a 0.99ab 0.16dc

Values having same letter(s) in a column do not differ significantly at 5% level by Duncan’s Multiple Range Test (DMRT).

Plate 1. Increased root growth of PGPR inoculated tissue-cultured banana plantlets 7.5 Discussion

The results and photograph (Plate 1) indicated that PGPR inoculation increased the production of primary and secondary roots. Inoculation showed higher primary root length, root volume and root dry weight compared to un-inoculated control. PGPR inoculation without N fertilisation produced the similar root dry weight of 100% N-treated plants, Sp7 with 33% N application showed highest root dry weight. Again UPMB10 with 33% N produced the highest root volume. Inoculation with UPMB10 resulted thicker and longer roots whereas Sp7 recorded more root mass, showing little increment in root length and diameter. This indicates that Sp7 produced more compact roots resulting more root mass. Our previous study on root colonisation by PGPR on banana roots indicated that Sp7 preferentially colonised the root hair zone with a few cells found on the root elongation and root tip zone whereas UPMB10 more efficiently colonized the entire roots.

69

Figure 2A

ba

a a

a

b bab a

a

bbc

ab a

a

a

cdb a

a

a

db a

a

a

cdb a

a

0

1

2

3

4

5

6

7

8

9

N P K Ca Mg

%N0-PGPR N0+Sp7 N0+UPMB10 N33%+Sp7 N33%+UPMB10 N100%-PGPR

Figure 2B

da a ab

a

c

a a a

a

da ab ab

a

b

a bc ab

a

a

ac a

a

a

a bc b

a

0

2

4

6

8

10

12

14

N P K Ca Mg

%

N0-PGPR N0+Sp7 N0+UPMB10 N33%+Sp7 N33%+UPMB10 N100%-PGPR

Figure 2. Effect of PGPR inoculation on nutrient concentration of bananas, A: leaf, B: stem The study indicated that PGPR inoculation affects the root function and an initial supplement of N could accelerate the root development, and using 33% N of the full requirement could contribute a similar or even higher root growth over the 100% N-supplied plants. It is in agreement with previous findings (Patriquin et al. 1983; Morgenstern and Okon 1987; Sumner 1990). They found that Azospirillum spp. inoculation affect the root function by stimulating the appearance of lateral roots and enhancement of the number of adventitious roots in different crop plants. Tien et al. (1979) also found that inoculation of PGPR in Setaria increased the number of lateral roots and root hairs without changing in root dry weight. Increase in diameter and length of the lateral roots in maize seedlings were observed by Hartmann et al. (1983). The possible explanations for the increased roots might be due to production of phytohormone or phytohormone-like substances, Bothe et al. (1992) concluded that Azospirillum enhanced the nitrite production in a nitrate base solution, which accelerate the lateral root formation in wheat. Without N, the PGPR were unable to establish them or might have taken much longer time to establish. Our study indicated that PGPR inoculation increased the total accumulation of P, K, Ca, and Mg. Higher uptake of N and K was due to the higher demand of those elements for banana. The enhanced mineral uptake of inoculated plants was due to more root growth. The accumulation of N was not much affected by PGPR inoculation, it might be due to lower amount of N2 fixation. Inoculation of Azospirillum often causes increases only in plant dry matter with decrease or no increases in concentrations of N (Bouton and Zuberer 1979; Kapulnik et al. 1981; Avivi and Feldman 1982; Smith et al. 1984). The ability of banana plantlets-adapted rhizobacteria to perform as biofertiliser inoculants that significantly improve the root growth and enhance nutrient uptake while reducing chemical N fertiliser inputs up to 67% under hydroponics condition is of major potential importance to sustainable banana production.

70

Figure 2C

cab

c a

b

cab

c ab

ab

bc

bc c

a

bc

ab aab

ab

b

abc

a

ab

a

ab bc

ab

0

2

4

6

8

10

12

N P K Ca Mg

N0-PGPR N0+Sp7N0+UPMB10 N33%+Sp7N33%+UPMB10 N100%-PGPR

Figure 2D

c dc

d dbc c

b

c cbc c

b

c c

ab

b

a

b b

a

a

a

a a bd

a

ab

a

0

50

100

150

200

250

300

350

400

N P K Ca Mg

N0-PGPR N0+Sp7 N0+UPMB10 N33%+Sp7 N33%+UPMB10 N100%-PGPR

Figure 2. Effect of PGPR inoculation on nutrient concentration of bananas, C: root, D: total accumulation of N, P, K, Ca and Mg ( root, stem, leaf) 7.6 Conclusion

The study concluded that PGPR inoculation significantly increased the root elongation and root growth. Application of PGPR without fertiliser N can produce an equivalent root mass of 100% N applied plant. Total accumulation of P, K, Ca and Mg was increased due to inoculation. Inoculation could reduce the dependency of N fertilisation at least 66% for banana plantlets growing for six weeks. The results strongly indicated that PGPR could be recommended as a bio-enhancer and biofertiliser for banana seedling production. 7.7 Acknowledgments

The authors are grateful to Universiti Putra Malaysia for technical assistance and facilities; to the Ministry of Science, Technology and Environment, Malaysia for the financial support under the project on Intensification of Research in Priority Areas (IRPA).

71

7.8 References

Amir HG, Shamsuddin ZH, Halimi MS, Ramlan MF, Marziah M (2001) Effects of Azospirillum inoculation on N2 fixation and growth of oil palm plantlets at nursery stage. Journal of Oil Palm Research 13, 42-49.

Avivi Y, Feldman M (1982) The response of wheat to bacteria of the genus Azospirillum . Isreal Journal of Botany 31, 237-245.

Bashan Y, Levanony H (1990) Current status of Azospirillum inoculation technology: Azospirillum as a challenge for agriculture. Canadian Journal of Microbiology 36, 591-608.

Bothe H, Korsgen H, Lehmacher T, Hundeshagen B (1992). Differential effects of Azospirillum, auxin and combined nitrogen on the growth of the roots of wheat. Symbiosis 13, 167-179.

Bouton JH, Zuberer DA (1979) Response of Panicum maximum Jcq. to inoculation with Azospirillum brasilense. Plant and Soil 52, 585-590.

Clarkson DT, Saker LR, Purves JV (1989) Depression of nitrate and ammonium transport in barley plants with diminished sulfate status. Evidence of coregulation of nitrogen and sulfate intake. Journal of Experimental Botany 40, 953-963.

Hartmann A, Singh M, Klingmuller W (1983) Isolation and characterization of Azospirillum mutants excreting high amounts of indoleacetic acid. Canadian Journal of Microbiology 29, 916-923.

Kapulnik Y, Gafny R, Okon Y (1985) Effect of Azospirillum spp. inoculation on root development and NO3

- uptake in wheat (Triticum aestivum cv. Miriam) in hydroponics systems. Canadian Journal of Botany 63, 627-631.

Kapulnik Y, Sarig S, Nur I, Okon Y, Henis Y (1981) The effect of Azospirillum inoculation on growth and yield of corn. Isreal Journal of Botany 31, 247-256.

Kloepper JW, Leong J, Teintze M, Schorth MN (1980) Enhanced plant growth by siderophores produced by plant growth promoting rhizobacteria. Nature 286, 885-886.

Lin W, Okon Y, Hardy RW(1983) Enhanced mineral uptake by Zea mays and Sorghum bicolor roots inoculated with Azospirillum brasilense. Applied Environmental Microbiology 45, 1775-1779.

Morgenstern E, OkonY (1987) Promotion of plant growth and NO3- and Rb+ uptake in Sorghum bicolor x Sorghum sudanense inoculated with A. brasilense-Cd. Arid Soil Research and Rehabilitation 1, 211-217.

Murty MG, Ladha, JK (1988) Influence of Azospirillum inoculation on the mineral uptake and growth of rice under hydroponics condition. Plant and Soil 108, 281-285.

Okon Y, Albrecht SL, Burris RH (1977) Methods for growing Spirillum lipoferum and for counting it in pure culture and different solution in association with plants. Applied Environmental Microbiology 33,85-88.

Patriquin DG, Döbereiner J, Jain DK (1983) Sites and processes of association between diazotrophs and grasses. Canadian Journal Microbiology 29, 900-915.

Saad MS, Sabuddin ASA, Yunus AG, Shamsuddin ZH (1999) Effects of Azospirillum inoculation on sweetpotato grown on sandy tin-tailing soil. Cummunications in Soil Science and Plant Analysis 30, 1583-1592.

SAS Institute Inc. (1987) SAS/STAT Guide for Personal Computers. (SAS Institute Inc., Carry, North Carolina, United States of America).

Shamsuddin ZH, Marziah M, Ismail MR, Yusof MK (1997) Beneficial effects of Azospirillum inoculation on growth of banana seedlings under different moisture regimes. Proceedings of the Fourth International Workshop on Plant Growth-Promoting Rhizobacteria: Present Status and Future Prospects. (Eds. A Ogoshi, K Kobayashi, Y Homma, F Kodam, N Kondo and S Akino) pp. 194-197. (Organisation for Economic Cooperation and Development, Sapporo, Japan).

Smith RL, Schank SC, Milam JR, Baltensperger AA (1984) Response of Sorghum and Pennisetum species to the N2 fixing bacterium Azospirillum brasilense. Applied Environmental Microbiology 47, 1331-1336.

72

Sumner ME (1990) Crop response to Azospirillum inoculation. Advances in Soil Sciences 12, 53-123. Swennen R, DeLanghe E, Janssen JD, Decoene D (1986) Study of the root development of some

Musa cultivars in hydroponics. Fruits 41, 515-524. Tarrand JJ, Krieg NR (1978) A taxonomic study of the Spirillum lipoferum group, with descriptions

of a new genus, Azospirillum gen. Nov. and two species, Azospirillum lipoferum (Beijerinck) comb.nov. and Azospirillum brasilense sp.nov. Canadian Journal of Microbiology 24, 967-980.

Thomas RL, Sheard RW, Moyer JR (1967) Comparison of conventional and automated procedures for nitrogen, phosphorus, and potassium analysis of plant materials using a single digest. Agronomy Journal 59, 240-243.

Tien TM, Gaskins MH, Hubbell DH (1979) Plant growth substances produced by Azospirillum brasilense and their effects on the growth of pearl millet (Pennisetumm americanum L.) Applied Environmental Microbiology 37,1016-1024.

Wange SS, Patil PI (1994) Effect of combined inoculation of Azotobacter and Azospirillum with chemical nitrogen on Basarai banana. Madras Agricultural Journal 81, 163-165.

73

8. The role of plant-associated beneficial bacteria in rice-wheat cropping system

K. A. Malik, M. S. Mirza, U. Hassan, S. Mehnaz, G. Rasul, J. Haurat, R. Bally and P. Normand 8.1 Abstract

This study indentifies, by 16S rRNA sequence analysis, two nitrogen-fixing, phytohormone producing bacterial isolates and evaluates the effects of the bacterial inoculants on growth of rice and wheat. The nitrogen-fixing isolates from rice (strain N4) and wheat (strain Wb3) were identified as Azospirillum lipoferum and Azospirillum brasilense, respectively. These isolates as well as other plant growth promoting rhizobacteria (PGPR) were used to inoculate rice and wheat plants grown in pots. In rice, among the six strains used as inoculants (Azospirillum lipoferum N4, Azospirillum brasilense Wb3, Enterobacter cloacae S1, Pseudomonas stutzeri K1, Pseudomonas 96-51, Zoogloea Ky1), significant increase in the biomass over non-inoculated controls was observed in plants inoculated with Azospirillum Wb3, Pseudomonas K1, Pseudomonas 96-51 and Zoogloea Ky1. Maximum increase in grain weight was recorded in plants inoculated with Zoogloea Ky1. As estimated by 15N isotopic dilution method, maximum nitrogen fixation (19.5% Ndfa i.e. nitrogen derived from atmosphere) was observed in plants inoculated with Pseudomonas K1. For wheat, four bacterial strains (Azospirillum N4, Azospirillum Wb3, Pseudomonas K1, Zoogloea Ky1) were used as inoculants. Maximum nitrogen fixation (Ndfa) and increase in biomass over control was observed in plants inoculated with Zoogloea Ky1. Unlike the case with rice the increase in grain weight of the inoculated wheat plants was not significantly different from that of controls. Keywords: biofertilizer, bacteria, rice, wheat. 8.2 Introduction

Beneficial bacteria comprise an important component of microbial community in the rhizosphere of plants, including rice and wheat. The mechanism involved in growth stimulation of plants was initially considered to be the nitrogen fixing activity of the associated bacteria (Lima et al. 1987; Malik et al. 1988, 1991; Urquiaga et al. 1992; Boddey et al. 1995). The emphasis has shifted from nitrogen fixation to growth-promoting effects of plant growth hormones produced by the bacteria present in the rhizosphere (Tien et al. 1979; Haahtela et al. 1990; Sarig et al. 1992; Kloepper 1994; Costacurta and Vanderleyden 1995; Rasul et al. 1998; Jacoud et al. 1999; Mehnaz et al. 2001). Production of antibiotics to reduce populations of minor plant pathogens, siderophores and other plant growth stimulatory compounds such as vitamins, are also implicated in the growth promoting effects of rhizosphere bacteria (Leong 1986; Rodelas et al. 1993). Studies on association of diazotrophic bacteria with the plants were initiated in the mid 80s in Pakistan when the rhizosphere of kallar grass (Leptochloa fusca (L.) kunth. was investigated for nitrogen fixing activity and the contribution of diazotrophic bacteria to nitrogen nutrition of plants was studied (Zafar 1985; Bilal 1988). This grass tolerates high salt concentrations and waterlogged conditions and grows luxuriantly in saline sodic soils. The soils that sustain such high yields of plant biomass are very low in fertility and thus it was worthwhile to investigate the rhizosphere of kallar grass for any biological nitrogen fixing activity. Nitrogen fixing activity (acetylene reduction activity) was detected in all samples of kallar grass roots and activity was higher during the active growth period (March-September) of grass (Malik et al. 1980; Zafar 1985). Estimation of nitrogen-fixing bacteria by ARA-based MPN technique showed that microbial population was higher on the root surface (rhizoplane) as compared to non-rhizosphere, rhizosphere and histoplane fractions (Bilal 1988).

74

Several nitrogen fixing bacterial strains including Klebsiella pneumoniae (NIAB-1), Beijerinckia sp. (Iso-2), Azotobacter sp., Azospirillum spp., Enterobacter spp., Zoogloea as well as some un-identified diazotrophs were obtained from kallar grass and other plants growing in saline sodic soils (Zafar et al. 1987; Bilal 1988; Bilal et al. 1990a,b; Malik et al. 1991). In addition to morphological characteristics and physiological/biochemical tests, detailed taxonomic studies were carried out in which DNA-based techniques (PCR, 16S rRNA sequence analysis, oligonucleotide probes) were used for identification promising isolates. Two diazotrophic bacterial isolates from kallar grass showing high nitrogen fixing activity (ARA) and phytohormone (indole acetic acid) production in pure culture (Rasul et al. 1998; Rasul 1999; Mehnaz 2000) were identified by sequence comparison of the PCR-amplified 16S rRNA gene with the database (Mehnaz 2000). The isolate K1 was identified as Pseudomonas stutzeri. This isolate (K1) had initially been identified as Azospirillum brasilense primarily on the basis of its nitrogen fixing ability and similar carbon source utilization pattern (Bilal et al. 1990b). On the basis of its morphological characters as well as its peculiar growth in shaken broth culture, where it aggregates to form macroscopic star-like flocks, the isolate Ky1 was originally identified as Zoogloea ramigera (Bilal and Malik 1987). The 16S rRNA sequence (1460 nucleotides) of Ky1 (Mehnaz, 2000) showed very high similarity (96%) to Pseudomonas sp. and relatively low similarity (84%) to Zoogloea ramigera sequence (Shin et al. 1993). By using The 15N isotopic dilution technique it was observed that a considerable amount of nitrogen in kallar grass was derived from the activity of the inoculated bacteria. By using three bacterial strains (Azospirillum brasilense, Klebsiella sp. (NIAB-I), Beijerinckia sp. (Iso-2) as inoculants for estimation of nitrogen fixation it was found that 50-70% N of the plants was derived from nitrogen fixing activity of the inoculated bacteria (Malik et al. 1988). In another study (Bilal 1988), in which bacterial isolates were used as inoculants for kallar grass, the highest 15N dilution was obtained in case of Enterobacter, followed by Azospirillum and then Azotobacter. This suggested considerable potential benefit from the N2-fixing bacteria-plant association. Encouraged by the results of the studies carried out on kallar grass-nitrogen-fixing bacterial associations, the emphasis of the research group naturally shifted to crops of economic importance like rice and wheat. A number of nitrogen-fixing, plant growth promoting rhizobacteria (PGPR) have been isolated and tested as inoculants for both these crops (Hassan et al. 1998; Malik et al. 1997; Rasul et al. 1998; Rasul 199; Mirza et al. 2000; Mehnaz, 2000; Mehnaz et al. 2001). In the present study, identification of the bacterial isolates from rice and wheat was carried out by 16S rRNA sequence analysis. Beneficial effects of these bacterial inoculants as well as other plant growth promoting rhizobacteria on rice and wheat and development of biofertilizers for these crops, are also discussed. 8.3 Materials and methods

PCR-amplification and 16S rRNA sequence analysis For extraction of template DNA for PCR, isolates N4 and Wb3 obtained from rice and wheat, respectively were grown in LB broth for 24 hr at 30oC. The cell pellets from 1.5 mL cultures were obtained by centrifugation at 13,000 rpm for five min and washed with TE buffer (10 mMTris.Cl; 1mM EDTA, pH 8). The cell pellets were then dissolved in 200µL of TE buffer. Cell lysis was obtained at 37oC for 30 min with lysozyme (2 mg mL-1; final concentration) and by using SDS (1%). The lysate was extracted twice with phenol/chloroform followed by two extractions with chloroform/isoamyl alcohol (24:1). After adding 1/10 volume of sodium acetate (3 M, pH 5.2) and 0.5 volume of isopropanol, the supernatant was incubated at –20 0C for 30 min. The nucleic acid pellets were then obtained by centrifugation at 13,000 rpm for 20 min, and washed with ethanol (70%) before drying under vacuum. The nucleic acid pellets were then dissolved in 100 µL TE buffer and used as template for PCR amplification of 16S rRNA gene. Each PCR reaction mixture (50 µL) contained 0.5 µL Taq DNA polymerase (5U µL-1; Gibco/BRL), 5 µL Taq buffer, 5 µL dNTPs (200 µM), 5 µL (100 ng µL-1) of each primer [Primer FGPS4-281 bis: AGA GTT TGA TCC TGG CTC AG; Primer FGPS1509’-153: AAG GAG GTG ATC CAG CCG CA; (Normand 1995)], 24.5 µL sterile water and 1 µL of template. After denaturation of the template at 94oC for 4 min, 30 rounds of

75

temperature cycling (94oC for 1 min, 55oC for 1 min and 72 oC for 1 min) were followed by incubation at 72oC for 7 min. The PCR products were gel purified (NuSieve, 1.2%) by using QIAquick spin (QIAGEN) kits and sequenced on Perkin-Elmer ABI PRISM Model 373. Amplification primers as well as internal primers (Normand 1995) were used for sequencing both strands of the PCR products. The sequences were deposited in the EMBL databank (strain N4, Accession no. AJ278445; strain Wb3, Accession no. AJ278446).

Inoculation of rice and wheat plants For The raising nursery of the rice variety Super Basmati, seeds were homogeneously spreaded on seed bed (100 g m-2) and kept covered with moist wheat straw till seed germination was completed. After removal of the plant cover used to retain moisture, sufficient canal water was added to keep lower part of the shoots of the seedlings submerged under water. After five weeks (first week of July), seedlings were removed from The nursery, roots were washed with canal water and transplanted after inoculations with bacterial strains. Seeds of wheat (Triticum aestivum, variey Inqalab) were directly sown in the pots (10 seeds pot-1). After gemination only five seedlings per pot were allowed to grow and additional seedlings were removed from the pots. Six bacterial strains (A. brasilense Wb3, A. lipoferum N4, Enterobacter S1, Pseudomonas K1, Pseudomonas 96-51 and Zoogloea Ky1) were grown in LB broth by incubation at 30oC on a shaker (100 rpm) for 24 h. Bacterial cells were harvested by centrifugation at 10,000 rpm for 5 min. Cell pellets were washed with saline (0.85% NaCl) and resuspended in saline to approximately 1 x 109 cells mL-1. Rice seedlings were inoculated by keeping the root system submerged in liquid bacterial cultures for 30 min and transplanted (five seedlings pot-1) in fiberglass pots (25 cm diameter) containing non-sterilized soil. Wheat seedlings were inoculated by adding one mL of the inoculum near the roots of each plant. To fill each pot, 15.5 kg air-dried marginally saline soil (Ec=4.87 mS /cm, pH=7.8, K=0.15 meq L-1, Na=66 meq L-1, Ca= 4.1 meq L-1, total nitrogen=0.04%) was used (Ali et al. 1998). After two weeks of transplantation, 15N labelled ammonium sulphate of 5 atom % excess (0.72 g pot-1) was added to all pots as a tracer to quantify nitrogen fixation. The pots were kept flooded with canal water until two weeks before rice harvest. For wheat, the same amount (0.72 g pot-1) of labeled ammonium sulphate was added to each pot after two weeks of seed germination. The plants were harvested at maturity and dried in an oven at 70 0C till no change in weight was noted. The dried plant samples were ground to a fine powder and total nitrogen in these samples was determined by using semi micro-kjeldahl method based on wet combustion in Rapid Kjeldahl System (Labconco, USA). The analysis for 15N excess was carried out on a double inlet Mass Spectrometer (MAT GD 150). Quantification of fixation based on isotope dilution was calculated by the formula of Fried and Middleboe (1977):

%Ndfa = 1- (15N atom % excess)fs . 100 (15N atom % excess)nfs

Plants inoculated with heat-killed bacterial cells were used as non-fixing reference (nfs) for the estimation of 15N dilution.

Acetylene reduction activity and enumeration of nitrogen fixing bacteria associated with roots For measuring acetylene reduction activity, roots as well as shoots of rice submerged under water (5 cm pieces of shoot near the base) were collected at panicle initiation and grain filling stages. The plant samples (approximately, 4 g fresh weight) were washed with sterile water and transferred to 16 mL capacity glass tubes with rubber stoppers. Acetylene (10%) was injected and the tubes were incubated at 30oC for 16 hr. Triplicate samples of roots and shoots, collected from 3 different plants, were used for ARA. The tubes with plant material (roots and shoots) but without C2H2 were used as control. Another set of tubes containing only 10% C2H2 and no plant material was also used as control. Ethylene production was measured on a gas chromatograph (Gasukuro kogyo model 370, Tokyo, Japan) using Porapak N column (Supelco Inc., Bellefonte, Pennsylvania). Quantitative

76

estimations of ethylene gas produced in the samples were made by measuring peak height relative to the standard (1% C2H4). Root and shoot samples were dried in an oven at 70oC to a constant weight. For estimation of the bacterial populations, root samples of rice and wheat were collected from each treatment at grain filling stage. Samples were washed with sterile water to remove the soil completely. One gram of each sample was homogenized in 9 mL sterile water and serial dilutions (10×) were prepared from this suspension. 100 µL of each dilution were used to inoculate 5 mL of semi-solid nitrogen-free Combined Carbon Medium (Renni 1981) in glass vials. Five vials were used for each dilution. After 24 h of bacterial growth at 30oC, vials were incubated with C2H2 (10% v/v) at 30oC. ARA was measured on a gas chromatograph (Gasokuro Kogyo, Model 370) after 48 h and ARA-positive vials were used for the estimation of MPN with the help of a probability table (Cochran 1950). 8.4 Results

Ribosomal rRNA sequence analysis Ribosomal RNA (16S rRNA) sequence analysis was used for identification of the two bacterial isolates N4 and Wb3 obtained from rice and wheat, respectively. By using conserved primers, PCR products of 16S rRNA gene were obtained from bacterial isolates and sequenced directly. The partial 16S rRNA sequence of the isolate N4 was obtained in two stretches (Figure 1). Comparison of this sequence information with the database showed highest similarity (98%) with the Azospirillum lipoferum sequences (Accession nos: X79729; Z29619.1). The partial 16S rRNA sequence of the isolate Wb3 showed highest sequence similarity (over 98% similarity) with the Azospirillum brasilense sequences (Accession nos: X79733; X79726.1).

Beneficial effect on rice To study the beneficial effects on plant growth, six bacterial isolates(Azospirillum Wb3, Azospirillum N4, Enterobacter S1, Pseudomonas K1, Pseudomonas 96-51 and Zoogloea Ky1 were used as inoculum for rice variety Super Basmati grown in non-sterilized saline soil. At panicle initiation stage, highest values of ARA were found in both the roots (74 nmol C2H4 g-1 root dry weight day-1) and basal parts of shoot (118 nmol C2H4 g-1 shoot dry weight day-1) of plants inoculated with Azospirillum Wb3 while at grain filling stage the ARA values for both plant tissues were highest (37 nmol C2H4 g-1 root dry weight day-1; 149 nmol C2H4 g-1 shoot dry weight day-1) in plants inoculated with Enterobacter S1 (Table 1). In most of the inoculation treatments, relatively low ARA was detected in roots at later growth stage (grain filling) as compared to panicle initiation stage. Like roots, ARA activity decreased in basal part of shoots at grain filling stage in majority of the treatments. However in plants inoculated with E. cloacae S1 high activity (149 nmol C2H4 g-1 shoot dry weight day-1) was detected at this stage. Increase in the grain yield and root and straw weight was recorded in all inoculated plants as compared to control (Table 2). As compared to control, maximum increase in grain yield was observed in plants inoculated with Zoogloea Ky1 whereas maximum increase in root and straw weight was observed in plants inoculated with Azospirillum Wb3. Nitrogen fixation, as determined by 15N isotopic dilution technique was highest in plants inoculated with Pseudomonas K1. Estimation of the nitrogen-fixing bacteria by ARA-based MPN method indicated presence of 107 cells g-1 dry root weight at grain filling stage of plant growth.

Beneficial effect on wheat Four bacterial inoculants (Azospirillum N4, Azospirillum Wb3, Pseudomonas K1 and Zoogloea Ky1) were used as inoculants to study the beneficial effects on growth of wheat plants grown in non-sterilized soil (Table 3). Zoogloea Ky1 strain proved to be the most effective strain as maximum increase in plant biomass, grain weight and %Ndfa over control was recorded in plants inoculated with this strain. However, the increase in grain weight of the plants inoculated with this strain as well as all other strains tested in the present study, was not significantly different from control. To determine the population of nitrogen-fixing bacteria associated with the roots of wheat, ARA-based MPN technique was used. The population of diazotrophs determined at grain filling stage indicated presence of diazotrophs in the range of 105 MPN g-1 dry weight of the roots.

77

Fragment A 1 70 N4 5’-GCATGCCTAACACATGCAAGTCGAACGAAGGCTTCGGCCTTAGTGGCGCACGGGTGAGTAACACGTGGGA Wb3 5’-GCATGCCTAACACATGCAAGTCGAACGAAGGCTTCGGCCTTAGTGGCGCACGGGTGAGTAACACGTGGGA 71 140 N4 ACCTGCCTTTCGGTTCGGAATAACGTCTGGAAACGGACGCTAACACCGGATACGCCCTACXGGGGAAAGT Wb3 ACCTGCCTTATGGTTCGGGATAACGTCTGGAAACGGACGCTAACACCGGATGTGCCCTTCGGGGGAAAGT 141 210 N4 TTACGCCGAGAGAGGGGCCCGCGTCGGATTAGGTAGTTGGTGTGGTAACGGCGCACCAAGCCGACGATCC Wb3 TTACGCCATGAGAGGGGCCCGCGTCCGATTAGGTAGTTGGTGGGGTAATGGCCCACCAAGCCGACGATCG 211 280 N4 GTAGCTGGTCTGAGAGGATGATCAGCCACACTGGGACTGAGACACGGCCCAGACTCCTACGGGAGGCAGC Wb3 GTAGCTGGTCTGAGAGGATGATCAGCCACACTGGGACTGAGACACGGCCCAGACTCCTACGGGAGGCAGC 281 350 N4 AGTGGGGAATATTGGACAATGGGCGCAAGCCTGATCCAGCAATGCCGCGTGAGTGATGAAGGCCTTAGCG Wb3 AGTGGGGAATATTGGACAATGGGGGCAACCCTGATCCAGCAATGCCGCGTGAGTGATGAAGGCCTTAGCG 351 420 N4 GTAGCGTGAGAAGAAGCCCCGGCTAACGGTTGTAAAGCTCTTTCGCACGCGACGATGATGATTCGTGCCA Wb3 GTAGCGTGAGAAGAAGCCCCGGCTAACGGTTGTAAAGCTCTTTCGCACGCXACGATGATGATTCGTGCCA 421 490 N4 GCAGCCGCGGTAATACGAAGGGGGCTAGCGTTGTTCGGAATTACTGGGCGTAAAGGGCGCGTAGGCGGCC Wb3 GCAGCCGCGGTAATACGAAGGGGGCGAGCGTTGTTCGGAATTACTGGGCGTAAAGGGCGCGTAGGCGGCC 491 560 N4 TGTTTAGTCAGAAGTGAAAGCCCCGGGCTCAACCTGGGAATAGCTTTTGATACTGGCAGGCTTGAGTTCC Wb3 TGTTTAGTCAGAAGTGAAAGCCCCGGGCTTAACCTGGGAACGGCTTTTGATACTGGCAGGCTTGAGTTCC 561 630 N4 GGAGAGGATGGTGGAATTCCCAGTGTAGAGGTGAAATTCGTAGATATTGGGAAGAACACCGGTGGCGAAG Wb3 GGAGAGGATGGTGGAATTCCCAGTGTAGAGGTGAAATTCGTAGATATTGGGAAGAACACCGGTGGCGAAG 631 700 N4 GCGGCCATCTGGACGGACACTGACGCTGAGGCGCGAAAGCGTGGGGAGCAAACAGGATTAGATACCCTGG Wb3 GCGGCCATCTGGACGGACACTGACGCTGAGGCGCGAAAGCGTGGGGAGCAAACAGGATTAGATACCCTGG

Fragment B 1 70 N4 5’- CCATCATTCAGTTGGGCACTCTGGTGGAACCGCCGGTGACAAGCCGGAGGAAGGCGGGGATGACGTCAAG Wb3 CCATCATTCAGTTGGGCACTCTGGTGGAACTGCCGGTGACAAGCCGGAGGAAGGCGGGGATGACGTCAAG 71 140 N4 TCCTCATGGCCCTTATGGGTTGGGCTACACACGTGCTACAATGGCGGTGACAGTGGGAGGCGAAGTCGCG Wb3 TCCTCATGGCCCTTATGGGTTGGGCTACACACGTGCTACAATGGCGGTGACAGTGGGATGCGAAGTCGCA 141 210 N4 AGATGGAGCAAATCCCCAAAAGCCGTCTCAGTTCGGATTGCACTCTGCAACTCGAGTGCATGAAGCTGGA Wb3 AGATGGAGCCAATCCCCAAAAGCCGTCTCAGTTCGGATTGCACTCTGCAACTCGGGTGCATGAAGTTGGA 211 280 N4 ATCGCTAGTAATCGCGGATCAGCACGCCGCGGTGAATACGTTCCCGGGCCTTGTACACACCGCCCGTCAC Wb3 ATCGCTAGTAATCGCGGATCAGCACGCCGCGGTGAATACGTTCCCGGGCCTTGTACACACCGCCCGTCAC Figure 1. Alignment of 16S rRNA gene sequence of Azospirillum lipoferum N4 and Azospirillum brasilense Wb3. The position 1 in the fragment A corresponds to E. coli 16S rRNA position 42; The position 1 in the fragment B corresponds to E.coli 16S rRNA position 1128

78

Table 1. Acetylene reduction activity (nmol C2H4 g-1 day-1) in roots and basal part of shoots of rice variety Super Basmati inoculated with plant growth promoting rhizobacteria Treatments Root Shoot

Panicle initiation

Grain filling

Panicle initiation

Grain filling

Control (heat killed mixed inoculum)

6 D 0.5 E 11 D 5 C

Azospirillum brasilense Wb3 74 A 3 DE 118 A 19 BC Azospirillum lipoferum N4 29 B 10 C 51 C 16 C Enterobacter cloacae S1 32 B 37 A 82 B 149 A Pseudomonas stutzeri K1 22 BC 8 CD 17 D 16 C Pseudomonas 96-51 13 CD 2 DE 13 D 42 BC Zoogloea Ky1 5 D 25 B 52 C 25 BC In a column, the values followed by the same letter do not differ significantly at 5% level of significance. Table 2. Effects of inoculation with plant growth promoting rhizobacteria on rice variety Super Basmati grown in pots

Treatments aARA-based MPN

(Mean±bSD)X 107

Root+Straw wt.

(g plant-1)

Grain wt. (g plant-1)

Ndfa (%)

Control (inoculated with heat killed bacteria)

0.02±0.01 14.8 C 5.0 C --

Azospirillum brasilense Wb3

0.37±0.03

22.5 A

7.8 AB

6.0 B

Azospirillum lipoferum N4

0.33±0.04 19.6 B

6.6 BC

2.5 B

Enterobacter cloacae S1 1.1±0.03

18.2 B

5.4 C

4.0 B

Pseudomonas stutzeri K1 2.1±0.06

22.2 A

7.9 AB

19.5 A

Pseudomonas 96-51 1.1±0.02

22.3 A

7.5 AB

3.0 B

Zoogloea Ky1 1.2±0.01

22.3 A

8.5 A

10.0 AB

In a column, the values followed by the same letter do not differ significantly at 5% level of significance. Data is based on six replicates. aAcetylene reduction assay-based most probable number, determined at grain filling stage. bSD, standard deviation. 8.5 Discussion

Ribosomal RNA sequence analysis has been extensively used to study phylogenetic relationships between micro-organisms as well as for taxon identification (Woese 1987; Woese et al. 1985, 1990). The availability and use of PCR-based amplification methods and sequencing of the PCR products on automated sequencers has dramatically expanded RNA databases during the past few years. Now sequences of over 16000 rRNA molecules from different organisms have been catalogued (Ludwig and Schleifer 1999; Normand 1999). This wealth of sequence information is now readily available in public databases for ever finer identification of new bacterial isolates by sequence comparisons. In the present study two nitrogen fixing bacterial isolates N4 and Wb3 from rice and wheat,

79

respectively, were identified by 16S rRNA sequence analysis. Both these strains were identified previously as Azospirillum strains on the basis of their typical spiral motility and carbon utilization pattern (Hassan et al. 1998). The 16S rRNA gene of the isolates N4 and Wb3 was amplified by using conserved primers and the PCR products were sequenced directly. Sequence comparison with the databases of the partial 16S rRNA sequence obtained in the present study confirmed identity of the isolate N4 to Azospirillum lipoferum showing 98% similarity. Similarly the isolate Wb3 was confirmed as A. brasilense strain on the basis of high 16S rRNA sequence similarities to those published for A. brasilense strains. The sequence information of the 16S rRNA gene of the Azospirillum strains N4 and Wb3 btained in the present study will be used to develop specific oligonucleotide probes and primers (Stahl and Amann 1991; Ward et al. 1992) for detection of these bacteria in soil or plant samples by dot blot hybridization or polymerase chain reaction (PCR). In the rice variety Super Basmati, nitrogen-fixing activity (ARA) was detected in roots and basal part of shoots of the inoculated as well as non-inoculated plants. Acetylene reduction activity detected in non-inoculated plants indicates presence of indigenous nitrogen fixing bacteria in the soil as plants were grown in non-sterilized soil. Detection of ARA in both the roots and shoots indicates colonization of not only roots but also basal parts of shoots by diazotrophic bacteria. In most of the treatments higher ARA was determined at panicle initiation stage than grain filling stage in the root samples. Variation in ARA with the plant growth stage has been reported by Watanabe et al. (1979) in two rice varieties IR36 and IR26 where maximum ARA was detected at heading stage. Rao and Rao (1984) determined maximum activity 60 days after transplantation while Barraquio et al. (1986) also detected maximum activity at The heading stage. Higher ARA detected during a particular growth stage may be due to reduction in the inhibitory nitrogen concentrations in the soil or overproduction of root exudates creating conducive conditions for growth and activity of diazotrophs (Dobereiner and De-Polli 1980; Jagnow 1983). Acetylene reduction activity detected in shoots and the presence of diazotrophs in high numbers may be of practical significance as the isolation and use of these bacteria as biofertilizers along with root colonizing bacteria may enhance efficiency of such inocula. In wetland rice, contribution of the basal portion of shoot to nitrogen fixation has been reported by Watanabe et al. (1981). Table 3. Effects of inoculations with plant growth promoting rhizobacteria on wheat Treatments

aARA-based MPN (Mean±bSD)X 105

Plant Biomass (g plant-

1)

Grain weight (g plant-

1)

%Ndfa

Control (inoculated with heat-killed cells)

4.6±1.6 5.6 B 2.0 A --

Azospirillum lipoferum N4 5.1±2.0 5.7 B 2.0 A 12.3 B Azospirillum brasilense Wb3 3.5±1.1 7.7 AB 2.6 A 7.1 C Pseudomonas K1 4.5±2.3 6.0 B 1.6 A 6.2 C Zoogloea Ky1 4.3±1.5 9.0 A 3.1 A 14.9 A

In a column, the values followed by the same letter do not differ significantly at 5% level of significance. Data is based on six replicates. aAcetylene reduction assay-based most probable number, determined at grain filling stage. bSD, standard deviation. The performance of four bacterial inoculants (Azospirillum N4, Pseudomonas K1, Pseudomonas 96-51 and Zoogloea Ky1) was comparable as reflected by the increase in plant biomass (root + straw weight) of the inoculated plants over control. However, maximum grain yield was observed in plants inoculated with Zoogloea Ky1 while maximum nitrogen fixation was recorded in plants inoculated with Pseudomonas K1. These results showed that the increase in plant biomass was not necessarily correlated with the amount of nitrogen fixed (%Ndfa) in inoculated plants. Therefore it may be

80

concluded that other mechanisms e.g. phytohormone produced by the inoculated bacteria may also be involved in the observed beneficial effects. All the bacterial strains used as inoculants have been reported (Rasul et al. 1998; Mehnaz 2000) to produce phytohormone (IAA) in pure culture. In wheat, maximum increase in plant biomass, grain weight and nitrogen fixation (%Ndfa) was recorded in plants inoculated with Zoogloea strain Ky1. The %Ndfa values determined in the present study for rice were much lower than those observed when the experiments were carried out under microbiologically controlled conditions (Mehnaz 2000). Similar low levels of nitrogen fixation in inoculated rice plants grown in non-sterilized soil as compared to plants grown under gnotobiotic conditions has been reported by Baldani et al. (2000). In the experiments carried out by Baldani et al. (2000), in which a number of strains belonging to genera Herbaspirillum and Burkholderia were used as inoculum for rice grown in greenhouse under natural soil conditions, the highest nitrogen contribution (Ndfa) from BNF was obtained with B. brasiliensis strain M130 (20%) followed by H. seropedicae strain ZAE67 (19%). In the rice plants grown under gnotobiotic conditions, Burkholderia strain M130 showed 28% Ndfa while Herbaspirillum strain ZAE67 fixed 32% of the total nitrogen in inoculated plants. In the same study strain ZAE94 of Herbaspirillum was found to fix 54% nitrogen in plants grown under gnotobiotic conditions while the same strain fixed only 17% nitrogen of the plants grown under natural soil conditions. In the present study beneficial effects (increase in plant biomass, %Ndfa) of the inoculations with plant growth promoting rhizobacteria on rice and wheat plants grown in non-sterilized saline soil were observed. The isolates from kallar grass (Pseudomonas K1 and Zoogloea Ky1) were also found very effective inoculants for both rice and wheat. These studies as well as several other investigations carried out in this laboratory (Zafar 1985; Bilal 1988; Hassan et al. 1998; Malik et al. 1997; Rasul et al. 1998; Rasul 1999; Mirza et al. 2000; Mehnaz 2000; Mehnaz et al. 2001) formed the basis for development of a biofertilizer for rice. This biofertilizer (commercial name “BioPower”) is based on a mixture of the bacterial strains isolated from rice as well as from other hosts like kallar grass. These bacterial strains were selected on the basis of their high nitrogen-fixing activity and phytohormone production in pure culture as well as their beneficial effects on the rice plants grown under controlled or field conditions. For production of the biofertilizer, the bacterial cells are grown and injected into polythene bags containing carrier material. This carrier material is a finely ground, dried and gamma-sterilized plant material that ensures survival during storage and transportation to farmer’s fields. Application of this biofertiliser in water-suspension form is recommended on the roots of rice plants at the time of transplantation from the nursery. Application of this biofertiliser during the past few years has increased to over 1000 ha per year. The rice farmers have reported saving of half the recommended doze of nitrogenous fertilisers while others have obtained nearly 20% yield increases over non-inoculated controls with full dose of fertiliser applications. The biofertiliser for wheat is in its final phase of testing before its large-scale production for farmers. The application of the biofertiliser developed for wheat has been recommended in a thick suspension form directly on the seeds just before sowing. 8.6 Acknowledgments

We are thankful to Dr. M. I. Sajjad, PINSTECH, Rawalpindi and Abdul Shakoor, NIBGE, for their help in 15N analysis of the samples. This work was supported in parts by a research grant (CRP/PAK 96-01) from International Centre for Genetic Engineering and Biotechnology (ICGEB), Trieste, Italy.

81

8.7 References

Ali S, Hamid N, Rasul G, Mehnaz S, Malik KA (1998) Contribution of non-leguminous biofertilizer to rice biomass, nitrogen fixation and fertilizer-N use efficiency under flooded soil conditions. In ‘Nitrogen Fixation with Non-legumes’. (Eds K A Malik, M S Mirza and J K Ladha) pp. 61-73. (Kluwer Academic Publishers, Dordrecht, The Netherlands).

Baldani VLD, Baldani JI, Döbereiner J (2000) Inoculation of rice plants with the endophytic diazotrophs Herbaspirillum seropedicae and Burkholderia spp. Biology and Fertility of Soils 30, 485-491.

Barraquio WL, Ladha JK, Yao HQ, Watanabe I (1986) Antigenic relationship of nitrogen-fixing Pseudomonas strain H8 to various known cultures and rice rhizosphere isolates studied by indirect enzyme-linked immunosorbent assay (ELISA). Canadian Journal of Microbiology 32, 402-408.

Bilal R (1988) Associative biological nitrogen fixation in plants in saline soils. PhD Thesis, Department of Botany, Punjab University, Lahore, Pakistan.

Bilal R, Malik KA (1987) Isolation and identification of a N2-fixing zoogloea-forming bacterium from kallar grass histoplane. Journal of Applied Bacteriology 62, 289-294.

Bilal R, Rasul G, Mahmood K, Malik KA (1990a) Nitrogenase activity and nitrogen-fixing bacteria associated with the roots of Atriplex spp. growing in saline sodic soils of Pakistan. Biology and Fertility of Soils 9, 315-320.

Bilal R, Rasul G, Qureshi JA, Malik KA (1990b) Characterization of Azospirillum and related diazotrophs associated with roots of plants growing in saline soils. World Journal Microbiology and Biotechnology 6, 46-52.

Boddey RM, de Oliveria OC, Urquiaga S, Reis VM, de Olivares FL, Baldani VLD, Dobereiner J (1995) Biological nitrogen fixation associated with sugarcane and rice: contributions and prospects for improvement. Plant and Soil 174, 195-209.

Cochran WG (1950) Estimation of microbial densities by means of the “Most Probable Number”. Biometrics 6,105-116.

Costacurta A and Vanderleyden J 1995 Synthesis of phytohormones by plant-associated bacteria. Critical Reviews Microbiology 21, 1-18.

Döbereiner J, De-Polli H (1980) Diazotrophic rhizocoenoses. In ‘Annual Proceedings of phytochemical Society of Nitrogen Fixation’. (Eds. W D P Stewart and WR Gallon) pp. 301-343. (Academic Press, New York, USA).

Fried M, Middleboe V (1977) Measurement of amount of nitrogen fixed by a legume crop. Plant and Soil 41, 713-715.

Haahtela K, Konkko R, Laakso T, Williams PH, Korhonen TK (1990) Root associated Enterobacter and Klebsiella in Poa pratensis: Characterization of an iron-scavenging system and a substance stimulating root hair production. Molecular Plant Microbe Interactions 3, 358-365.

Hassan U, Mirza MS, Mehnaz S, Rasul G, Malik KA (1998) Isolation and identification of diazotrophic bacteria from rice, wheat and kallar grass. In ‘Nitrogen fixation with nonlegumes’. (Eds. KA Malik, MS Mirza and JK Ladha) pp.197-205. (Kluwer Academic Publishers, Dordrecht, The Netherlands).

Jacoud C, Job D, Wadoux P, Bally R (1999) Initiation of root growth stimulation by Azospirillum lipoferum CRT1 during maize seed germination. Canadian Journal of Microbiology 45, 339-342.

Jagnow G (1983) Nitrogenase (C2H2) activity in non-cultivated and cereal plants: Influence of nitrogen fertilizer on population and activity of nitrogen fixing bacteria. Z. Pflanzenernaehr. Bodenkd. 146, 217-227.

Kloepper JW (1994) Plant growth promoting rhizobacteria (other systems). In ‘Azospirillum/Plant Associations’. (Ed Y Okon) pp. 137-166. (CRC Press, Florida, USA).

Leong J (1986) Siderophores: Their biochemistry and possible role in the biocontrol of plant pathogens. Annual Reviews Phytopathology 24, 187-209.

82

Lima E, Boddey RM, Döbereiner J (1987) Quantification of biological nitrogen fixation associated with sugarcane using 15N aided nitrogen balance. Soil Biology and Biochemistry 19, 165-170.

Ludwig W, Schleifer KH (1999) Phylogeny of bacteria beyond the 16S rRNA standard. ASM News 65, 752-757.

Malik KA, Bilal R, Azam F, Sajjad MI (1988) Quantification of N2-fixation and survival of inoculated diazotrophs associated with roots of kallar grass. Plant and Soil 108, 43-51.

Malik KA, Bilal R, Mehnaz S, Rasul G, Mirza MS, Ali S (1997) Association of nitrogen fixing and plant growth promoting rhizobacteria (PGPR) with kallar grass and rice. Plant and Soil 194, 37-44.

Malik KA, Bilal R, Rasul G, Mahmood K, Sajjad MI (1991) Associative N2-fixation in plants growing in saline sodic soils and its quantification based on 15N natural abundance. Plant and Soil 137, 67-74.

Malik KA, Zafar Y, Hussain A (1980) Nitrogenase activity in the rhizosphere of kallar grass (Diplachne fusca Linn Beauv). Biologia 26, 107-112.

Mehnaz S (2000) Molecular characterization of plant growth promoting rhizobacteria (PGPR) isolated from rice (Oryza sativa L.). PhD Thesis, Department of Biological Sciences, Quaid-i-Azam University, Islamabad, Pakistan.

Mehnaz S, Mirza MS, Huarat J, Bally R, Normand P, Malik KA (2001) Isolation and 16S rRNA sequence analysis of the beneficial bacteria from rice. Canadian Journal of Microbiology 47, 110-117

Mirza MS, Rasul G, Mehnaz S, Ladha JK, So RB, Ali S, Malik KA (2000) Beneficial effects of inoculated nitrogen-fixing bacteria on rice. In ‘The Quest for Nitrogen Fixation’. (Eds. JK Ladha and PM Reddy) pp. 191-204. (International Rice research Institute, Los Bannos, Philippines).

Normand P (1995) Utilisation des séquences 16S pour le positionnement phylétique d'un organisme inconnu. Oceanis 21, 31-56.

Normand P (1999) Molecular phylogeny of nitrogen-fixing bacteria in symbiosis with plant roots. In ‘Taxonomy,phylogeny and gnotobiotical studies of entomopathogenic nematode bacterium complexes’. (Eds. N Boemare, P Richardson and F Coudert) pp. 29-35. (Brussels, Belgium).

Rao VR, Rao JLN (1984) Nitrogen fixation (C2H2 reduction) in soil samples from rhizosphere of rice grown under alternate flooded and nonflooded conditions. Plant and Soil 81, 111-118.

Rasul G (1999) Production of growth hormones and nitrogenase by diazotrophic bacteria and their effect on plant growth. PhD Thesis, Institute of Biochemistry and Biotechnology, University of the Punjab, Lahore, Pakistan.

Rasul G, Mirza MS, Latif F, Malik KA (1998) Identification of plant growth hormones produced by bacterial isolates from rice, wheat and Kallar grass. In ‘Nitrogen fixation with nonlegumes’. (Eds. KA Malik, MS Mirza and JK Ladha) pp 25-37. (Kluwer Academic Publishers, Dordrecht, The Netherlands).

Rennie R J (1981) A single medium for the isolation of nitrogen fixing bacteria. Canadian Journal of Microbiology 27, 8-14.

Rodelas B, Salmeron V, Martinez-Toledo MV, Gonzalez-Lopez J (1993) Production of vitamins by Azospirillum brasilense in chemically defined media. Plant and Soil 153, 97-101.

Sarig S, Okon Y, Blum A (1992) Effect of Azospirillum brasilense inoculation on growth dynamics and hydraulic conductivity of Sorghum bicolor roots. Journal of Plant Nutrition 15, 805-819.

Shin YK, Hiraishi A, Sugiyama J (1993) Molecular systematics of the genus Zoogloea and emendation of the genus. International Journal of Systematic Bacteriology 43, 826-831.

Stahl DA, Amann R (1991) Development and application of nucleic acid probes. In ‘Nucleic acid techniques in bacterial systematics’. (Eds. E Stackebrandt and M Goodfellow) pp. 205-247. (John Willey and Sons, New York, USA).

Tien TM, Gaskins MH, Hubbel DH (1979) Plant growth substances produced by Azospirillum brasilense and their effect on growth of pearl millet (Pennisetum americanumL.). Applied and Environmental Microbiology 37, 1016-1024.

83

Urquiaga S, Cruz KHS, Boddey RM (1992) Contribution of nitrogen fixation to sugarcane:Nitrogen-15 and nitrogen-balance estimates. Soil Science Society American Journal 56,105-114.

Ward MW, Bateson MM, Weller R, Ruff-Roberts AL (1992). Ribosomal RNA analysis of micro-organisms as they occur in nature. In ‘Recent advances in microbiol ecology’. (Ed. KC Marshall) pp 219-286. (Plenum Press, New York).

Watanabe I, Barraquio WL, de Guzman MR, Cabera DA (1979) Nitrogen fixing (C2H2 reduction) activity and population of aerobic heterotrophic nitrogen fixing bacteria associated with wetland rice. Applied and Environmental Microbiology 37, 813-819.

Watanabe I, Cabrera D, Barraquio WL (1981) Contribution of basal portion of shoot to nitrogen fixation associated with wetland rice. Plant and Soil 59, 391-398.

Woese CR (1987) Bacterial evolution. Microbiology Reviews 51, 221-271. Woese CR, Stackebrandt E, Macke TJ, Fox GE (1985) A phylogenetic definition of the major eu-

bacterial taxa. Systematic and Applied Microbiology 6, 143-151. Zafar Y (1985) Some studies on the diazotrophic biocoeonosis in Kallar grass (Leptochloa fusca (L.)

Kunth). PhD Thesis, Department of Biological Sciences, Quaid-i-Azam University, Islamabad, Pakistan.

Zafar Y, Malik KA, Niemann EG (1987) Studies on N2 fixing bacteria associated with salt tolerant grass, Leptochloa fusca (L.) Kunth. MIRCEN Journal of Applied Bacteriology and Biotechnology 3, 45-56.

84

9. Facilitating a N2-fixing symbiosis between diazotrophs and wheat

N. Islam, C. V. S. Rao and I. R. Kennedy 9.1 Abstract

As part of a program to develop symbiosis between diazotrophic bacteria and wheat, three N2-fixing species [Azospirillum brasilense (HM53, HM53.1 Sp7-S), Herbaspirillum seropedicae and Citrobacter freundii] were used to inoculate wheat seedlings grown in controlled or greenhouse conditions. Improvement in the rate of plant growth due to inoculation with the bacterial strains using synthetic auxin to promote colonisation was assessed by shoot nitrogen content, shoot and spike dry weight and related to endophytic bacterial counts. In hydroponic culture, the total nitrogen content in the shoot increased significantly by co-inoculation of Azospirillum with Citrobacter freundii. An increased endorhizosphere population of the N-fixer was also observed by co-inoculation. In pot culture in soil, the effects of inoculation were not obvious at early stages of wheat growth, but at maturity, all treatments with these strains showed positive effects on shoot growth compared to the control. Increases in the grain yield of wheat were also associated with inoculation and supplementation of soil with malate to assist early colonization. Keywords: N2-fixation, symbiosis, plant growth, wheat. 9.2 Introduction

Nitrogen is a major macro-nutrient determining the productivity of cereals. The economic and environmental consequences of organic-N supply to plants by direct N2-fixation are well recognised. To reduce the cost of production and to minimise the risk of degrading the environment by applying chemical fertilisers, attention has been focused on the possibility of biological N2 fixation in cereals (Kennedy and Cocking 1997). A significant reduction in the use of inorganic-N with cereals would be possible if an effective and stable association between N2-fixing microbes and cereals were available (Kennedy and Islam 2000). A model of symbiosis between Azospirillum with wheat roots inducing with synthetic auxin to stimulate colonisation has been reported (e.g. Tchan et al. 1991; Kennedy and Tchan 1992; Kennedy et al. 1997). Estimating colonisation using reporter gene (lacZ) fusion techniques has further allowed the recognition of important features in endophytic colonisation in para-nodulated wheat roots (Vande Broek et al. 1993; Katupitiya et al. 1995). Using these techniques, additional desirable characteristics for colonisation diazotrophs have been identified (Katupitiya et al. 1995; Pereg-Gerk et al. 1998). Initial trials with an ammonia-excreting mutant of Azospirillum with wheat in McCartney bottles showed a significant transfer of newly fixed N after 72 h of exposure to 15N provided sufficient organic carbon was supplied (Wood 1999; Wood et al. 2001). A set of cooperative processes as a result of the plant genotype and bacterial interaction will be necessary to obtain eventual symbiosis between diazotrophs and cereals. Identification of these by processes of colonisation, N-fixation and other relevant action processes can indicate the procedures requiring for modification and selection has been described (Kennedy and Islam 2000). In this study, the practical effect of some of these processes, including enhanced colonisation and N-transfer are examined. In order to favour positive effects, measures were taken to promote microbial growth for colonisation by treatment with 2,4-D and by providing carbon substrate as malate for microbial growth.

85

9.2 Materials and methods

Bacteria and bacterial strains used in this experiment are listed in Table 1. Table 1. Bacteria and bacterial strains used in the study

Bacteria Strains Comments

Azospirillum brasilense

HM53 HM53.1 Sp7 Sp7-S

HM53 is ammonia-excreting mutant provided by Dr F Pedrosa (Pedrosa et al.1989). HM53.1 was a flcA- transposon-directed mutant of HM53 (see Pereg-Gerk et al. 1998) modified to favour endophytic colonisation, similar to Sp7-S. Sp7-S was isolated in this laboratory (Katupitiya et al. 1995).

Herbaspirillum seropedicae

(ATCC 35892) Endophytic, found effective for N-supply in sugarcane (Baldani et al. 1986), in sorghum (Döbereiner et al. 1993) and in wheat (Pereira et al. 1988).

Citrobacter freundii

C3 Provided by Professor Nguyen Thanh Hien, isolated from the rhizosphere of rice roots at the Hanoi University of Science and identified in this laboratory by 16S rRNA sequence analysis (unpublished).

Seed treatment and raising seedlings Seed of the Sunbri cultivar of wheat was rinsed with 90% ethanol and then surface-sterilised by treatment under vacuum with 0.5% HgCl2 for 75 s (Zeman et al. 1992). The seed was then transferred to plates of yeast-manitol agar (YMA) and incubated at 350C for germination. Two healthy seedlings were transferred to each sterile test-tube containing 15 ml of nitrogen-free hydroponic solution (Zeman et al. 1992). Filter paper was used to support the germinated seed. The seedlings were grown in controlled conditions in a growth chamber (Biotron-Japan) under constant light (200 µE m-2 sec-1), with alternate 12-h cycles at 180C and 250C. The hydroponics solution was treated with 0.2 ml of 50 ppm sterile 2,4-D stock solution, depending on treatment, giving a final concentration of 2,4-D in 15 ml solution of 0.67 ppm. On the following day, the seedlings were inoculated with bacteria/bacterial combinations and grown in the Biotron under the constant light (rate as mentioned above) for further six days before transfer to other growth media used as sand or soil.

Soil collection and preparation A red brown soil collected from a rice growing area (Yanco, NSW, Australia) was used for pot trials. The initial pH of the soil was 5.6 (H20 1:5). Malic acid (5g kg-1 soil) was added and the soil pH 4 was adjusted to 7 by adding lime (3 g kg-1 soil). To facilitate reaction, the soil was saturated with water, lime and malic acid (well distributed). After two days of incubation with mixing (2x), the soil was air-dried to field capacity, ground and sieved. The pH of the malate and lime treated soil (pH 7.8) was then mixed with untreated soil (1:1) yielding a final soil pH of 6.8. The soil was then potted for experimentation (1kg for the greenhouse and 100 g for the Biotron- growth cabinet). A set of plants with the same treatment was also grown in the Biotron using 100 g of sand per pot.

Transplanting and maintenance Two seedlings from hydroponics solution were transplanted to each pot and grown in the Biotron (250C/180C day/night 12h day). Plants grown in sand were supplied weekly with nitrogen-free hydroponics solution. A small amount of “starter-N” (0.0144 g L-1) was added to the hydroponic solution in the first and third month. In hydroponic culture, following 6 days of inoculation, 15 ml of

86

sterilised solution was added and the stem of the seedling was supported at the base with non-absorbent cotton to allow the tops of the seedlings to air to ensure normal transpiration.

Harvesting and data collection Plants were harvested at 30, 60, 90 days after transplanting (DAT). The final harvest for the glasshouse trial carried out in soil was at 120 DAT. Shoot dry weight was measured after oven drying at 700C for 3 days.

N- determination A portion of seedling grown in hydroponic solution was used for N determination, after drying of shoots at 700C for 3 days and grinding, using the Dumas total combustion method in a Leco CHN-1000 elemental analyser (Leco Inc., St Joseph, MI, USA).

Microbial counts Ten g of rhizosphere soil shaken from roots was suspended into 100 ml of sterile water and mixed thoroughly using a sterile magnetic stirrer. For the endorhizosphere samples, 1 g of the root per whole root sample was surface sterilised and ground using a mortar and pestle with 10 ml of sterile distilled water. Ten fold dilutions were prepared and 0.1 ml of the appropriate dilution was inoculated by streaking on both Congo red and nutrient agar media. The plates were incubated at 300C for 48 h and the number of colonies was counted.

Statistical analyses Statistical analyses were performed using JMP Software (SAS Institute Inc.). The standard error of means and least significant difference (LSD) at 5% was used to differentiate the treatments. 9.4 Results

Shoot growth and N uptake The shoot dry weights of plants co-inoculated HM53.1 and C3 were significantly higher than the control in hydroponically grown plants (Figure 1). By comparison HM53.1+C3 with 2,4-D application showed reduced shoot growth. Like the shoot dry-weight, the HM53.1+C3 co-inoculation showed increased total N uptake compared to the control (Figure 2). A 2,4-D application for growth in hydroponics solution, co-inoculated HM53.1 and C3, resulted in elevated N- concentration in the shoot but the effect was reduced when expressed as a total N, because of poor shoot growth in 2,4-D treated plants (Figure 2). Treatment effects on shoot dry weight of plants grown in the Biotron using sand and soil were similar both at 30 and 60 DAT (Table 2). At 90 DAT, a similar increase in shoot growth was observed in sand-grown plants inoculated with the C3, Sp7-S and Sp7-S + C3 when compared to the control, but this effect was not observed in soil. In the greenhouse trial, in unsterilised soil, differences in shoot-dry weight among treatments were not appreciable either at 30 or 60 DAT, however, significant differences were obvious when plants were harvested at 120 DAT (Figure 3). All treatments, including the C3 strain alone, showed better shoot growth compared to un-inoculated control. HM53.1, HM53.1+C3, Herbaspirillum seropedicae and Sp7-S showed the most pronounced effects on shoot production. Similar to the shoot weight, increased spike dry weight, indicating increased grain yield was also observed when seedlings were treated with HM53 plus C3, and Herbaspirillum seropedica.

87

05

101520253035404550

0

5

10

15

20

25

30ShootHeight

Figure 1. Dry shoot weight and plant height of wheat grown in hydroponics for 2 months. (LSD at 5% 8.0 for shoot and 2.2 for height, Bar represents standard error of mean for each treatment)

LSD for shoot dry weight at 5%

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

1

Control C3 HM53.1 HM53.1+C3 HM53.1+C3+2,4-Da

HM53.1+C3 +2,4-Db

Tota

l N u

ptak

e (m

g pl

ant-1

)

Figure 2. Total N uptake by shoot of wheat plants grown in hydroponics for two months

(Bar represents standard error of mean for each treatment. a100 ppm of 2,4-D applied as foliar spray 4 times, b 0.66 ppm of 2,4-D in hydroponic)

LSD for total N content at 5%

88

0.00

0.50

1.00

1.50

2.00

2.50Shoot 120DAT Grain 120DAT

Figure 3. Dry weight (g) of wheat plants at 120 days after transplanting (DAT)grown in glasshouse. (Bar represents standard error of mean for each treatment of three replicates)

Table 2. Dry shoot weight (g) of wheat at three stages of harvest grown in Biotron with sand and soil Treatments Sand cultured Soil cultured

30 DATa 60 DAT 90 DAT 30 DAT 60 DAT 90 DAT Control 0.036 0.125 0.208 0.053 0.156 0.282

C3 0.033 0.118 0.257 0.056 0.207 0.277

HM 53.1 0.043 0.136 0.224 0.048 0.204 0.279

HM 53.1 + C3 0.042 0.133 0.222 0.045 0.196 0.181

SP7-S 0.035 0.127 0.261 0.061 0.199 0.282

SP7-S + C3 0.037 0.123 0.249 0.058 0.185 0.267

Herbaspirillum 0.039 0.159 0.250 0.059 0.184 0.263

Herbaspirillum + C3 0.032 0.129 0.230 0.055 0.194 0.285

LSD (5%) 0.010 0.050 0.051 0.016 0.062 0.133

Seedlings were treated with 2,4-D. aDays After Transplanting (DAT)

Colonisation patterns From the dilution counts, the total number of root-colonising bacteria in wheat roots grown hydroponically for 2 months was greater in all cases where plants were co-inoculated. The greatest numbers (greater than 109 per g of dry weight) were observed in roots treated with HM53.1+C3 compared to the control (Figure 4). The bacterial population expressed per unit weight of fresh roots in the endorhizophere and rhizosphere declined sharply between 0 and 7 days after inoculation (DAI), indicating a initial high death rate and remained almost the same in Biotron-grown plants and or increased slightly between 7 and 30 DAI (Figures 5 and 6).

89

1.00E+001.00E+021.00E+041.00E+061.00E+081.00E+101.00E+12 Nutrient Agar

N-Free media

Figure 4. The number of endophytic bacteria in wheat roots grown hydroponically for two months. (Plants were grown in hydroponics for two months, roots treated with C3 and C3+HM53.1 were thick and curled suggesting a phytohormonal effect. Bar represents standard error of means for each treatment)

The population of C3 (companion) in rhizosphere and endorhizosphere after 7 DAI was higher compared to other microbes, although the difference is not very clear at 30 DAI in both the greenhouse and biotron grown conditions (Figures 5 and 6). In general, C3 showed better contribution in colonisation in both endo and exorhizosphere (Figures 5 and 6) 9.5 Discussion

Improved shoot growth due to inoculation with diazotrophs are in agreement with some previous findings (Smith et al. 1981; Bashan 1986; Baldani and Döbereinier 1986; Boddey 1995). But the exact mechanism of such increased shoot dry weight from bacterial inoculation, which could involve any of several agronomic traits such as nitrogen fixation, nutrient mobilisation or hormonal effects in plant growth, was not examined. Even though the bacterial strains used in this experiment have tested positive for their ability to fix N2, (except for the C3), no claim is possible that the effects observed are due to N2 fixation. In nonsymbiotic N2 fixation, the amount of fixed N supplied to the host plant by most of the N-fixers is not considered sufficient to support plant growth (Christiansen-Weniger and Van Veen 1991; Okon and Labandera 1994). Instead the effect of bacterial strains on plant growth parameters can be assigned to a number of factors such as hormonal effects, mobilisation of mineral nutrients from soil or decomposition of bacterial biomass. Poor initial survival of inoculated bacteria was observed in this study. Similar resulted were reported by Smith et al. (1981). The C3 strain (Citrobacter freundii) seemed to give enhance activity when co-inoculated with all other bacteria (Figures 3, 5 and 6). Higher colonization or inoculation with C3 both in the endo- and exo-rhizosphere was also affirmated by the better crop growth due to inoculation (Figures 1 and 2.). Xiaoji et. al. (1993) found better effect of Bradyrhizobium colonization by using pectin-lytic bacteria. Yanzchen and Fudi (1993) also reported higher nodular structure in several cereals and oil crops when treated with a helper bacteria. Although a trend of increased growth rate, with at least some of the treatments, compared to the control was observed in this experiment, further optimisation of such diazotrophs-crop plant associations has to be considered as essential for a sustainable technology under farming conditions.

90

In these experiments, measures were taken to ensure effective colonisation and to encourage bacterial growth. Because of limited resources, we could not include control treatments for the effect of 2,4-D and malate, which were needed to be done. Our previous experiments showed that the application of synthetic auxin (2,4-D) favoured root colonisation with Azospirillum ( Zeman et al. 1992) and the addition of malate should favour multiplication and should significantly increase the transfer of newly-fixed N2 from bacteria to the shoot tissue of wheat (Wood 1999; Wood et al. 2001). An extension of the research will allow us to examine if these treatments designed to favour a productive system are necessary to achieve increases in grain yield.

0

2

4

6

8

10

0 7 30

controlcompanionHM53.1HM53.1 + CompanionSp7-SSp7-S + companionHerbaspirillumHerbaspirillum + Com.

Sand cultured plants grown in Biotron

log

cfu

g-1 o

f fre

sh ro

ots

Days

0

2

4

6

8

10

0 7 30

controlcompanionHM53.1HM53.1 + CompanionSp7-SSp7-S + companionHerbaspirillumHerbaspirillum + Com.

Soil cultured plants grown in Biotron

log

cfu

g-1 o

f fre

sh ro

ots

Days

0

2

4

6

8

10

0 7 30

control

companion

HM53.1

HM53.1 + Companion

Sp7-S

Sp7-S + companion

Herbaspirillum

Herbaspirillum + Com.

HM53

HM53 + companion

Soil cultured plants grown in Glass House

log

cfu

g-1 f

resh

root

s

Days

Figure 5. Endophytic colonisation of bacteria at different days after inoculation The number of bacteria in soil at the time of inoculation-0 DAT and in endorizosphere at 7 and 30 DAT are shown Increases in root weight would be compensate the decline in the bacteria per unit fresh weight

91

0

2

4

6

8

10

0 7 30

control

companion

HM53.1

HM53.1 + Companion

Sp7-S

Sp7-S + companion

Herbaspirillum

Herbaspirillum + Com.

Sand cultured plants grown in Biotron

Days

log

cfu

g-1 s

oil

0

2

4

6

8

10

0 7 30

control

companion

HM53.1

HM53.1 + Companion

Sp7-S

Sp7-S + companion

Herbaspirillum

Herbaspirillum + Com.

Soil cultured plants grown in Biotron

Days

log

cfu

g-1 s

oil

0

2

4

6

8

10

0 7 30

controlcompanionHM53.1HM53.1 + CompanionSp7-SSp7-S + companionHM53HM53 + companionHerbaspirillumHerbaspirillum + Com.

Soil cultured plants grown in Glass house

Days

log

cfu

g-1 s

oil

Figure 6. Rhizosphere-soil colonisation of bacteria at different days after inoculation The number of bacteria in soil at the time of inoculation-0 DAT and in rhizosphere at 7 and 30 DAT are shown. It is likely that an initial decline in bacteria occurred soon after inoculation (not shown in data) 9.6 Acknowledgements

The authors are grateful to Grains Research and Development Corporation (GRDC) for their grant (US224), to the Australian Research Council (ARC) for their financial support, to Professor Nguyen Thanh Hien and Dr Fabio Pedrosa for supplying bacterial strains

92

9.7 References

Baldani JI, Baldani VLD, Seldin L, Döbereiner J (1986) Characterisation of Herbaspirillum seropedicae gen. nov., sp. nov., a root-associated nitrogen-fixing bacterium. Internation Journal of Systematic Bacteriology 36, 86-93.

Bashan Y (1986) Enhancement of wheat roots colonisation and plant development by Azospirillum brasilense Cd, following temporary depression of the rhizosphere microflora. Applied Environmental Microbiology 51, 1067-1071

Boddey RM, de Oliveira OC, Urquiga S, Reis VM, Olivares FL, Baldani VLD, Döbereiner J (1995) Biological nitrogen fixation associated with sugar cane and rice: contributions and prospects for improvement. Plant and Soil 174, 195-209.

Christiansen-Weniger C, Van Veen, J (1991) NH4+ -excreting Azospirillum brasilense mutants

enhance the nitrogen supply of a wheat host. Applied Environmental Microbiology 57, 3006-3012. Döbereinier J, Reis VM, Paula MA, Olivers F (1993). Endophytic diazotrophs in sugarcane, cereals

and tuber crops, In ‘New Horizon in Nitrogen Fixation. (Eds R Palacios, J Mora and WE Newton) pp 671-674. (Kluwer Academic Publishers, Dordrecht, Netherlands).

Katupitiya S, Millet J, Vesk M, Viccars L, Zeman A, Lidong Z, Elmerich, Kennedy IR (1995) A mutant of Azospirillum brasilense Sp7 impaired in flocculation with a modified colonization pattern and superior nitrogen fixation in association with wheat. Applied Environmental Microbiology 61, 1987-1995.

Kennedy IR, Cocking EC (1997) Biological Nitrogen Fixation: The Global Challenge and Future Needs, ISBN 1-86451-364-7, 83 pages, SUNFix Press, University of Sydney, Sydney, Australia.

Kennedy IR, Islam N (2000) The current and potential contribution of asymbiotic nitrogen fixation to nitrogen requirement on farms. Australian Journal of Experimental Agriculture 41, 447-457.

Kennedy IR, Pereg-Gerk L, Wood C, Deaker R, Gilchrist K, Katupitiya S (1997) Biological nitrogen fixation in non-leguminous field crops: Facilitating the evolution of an effective association between Azospirillum and wheat. Plant and Soil 194, 65-79.

Kennedy IR, Tchan YT (1992) Biological nitrogen fixation in non-leguminous field crops: Recent advances. Plant and Soil 141, 93-118.

Okon Y, Labandera-Gonzalez CA (1994) Agronomic applications of Azospirillum: an evaluation of 20 years worldwide field inoculation. Soil Biology and Biochemistry 26, 1591-1601.

Pedrosa FO, De Souza EM, Machado HB, Rigo LU, Funayama S (1989) Regulation of nif genes expression in Azospirillum brasilense and Herbaspirillum seropedicae. In ‘Nitrogen fixation with Non-Legumes’. Proceedings of the Fourth International Symposium on Nitrogen Fixation with Non-Legumes, Rio de Janeiro, Brazil. (Eds FA Skinner, RM Boddey and I Fendrik) pp. 155-163. (Kluwer Academic Publishers, Dordrecht, Nrtherlands).

Pereg-Gerk L, Paquelin A, Gounon P, Kennedy IR, Elmerich C (1998) A transcriptional regulator of the LuxR-UhpA family, FlcA, controls flocculation and wheat root surface colonisation by A. brasilense Sp7. Applied Environmental Microbiology 11, 177-187.

Pereira JAR, Cavalcante VA, Baldani JI, Döbereinier J (1988) Field inoculation of sorghum and rice with Azospirillum spp and Herbaspirillum seropedicae. Plant and Soil 110, 269-274.

Smith RS, Ellis MA, Smith RE (1981) Effect of Rhizobium japonicum inoculant rates on soybean nodulation in a tropical soil. Agronomy Journal 73, 505-508

Tchan YT, Zeman AMM, Kennedy IR (1991) Nitrogen fixation in para-nodules of wheat roots by introduced free-living diazotrophs. Plant and Soil 137, 43-47.

Vande Broek A, Michiels J, Van Gool A, Vanderleyden J (1993) Spatial-temporal colonisation patterns of Azospirillum brasilense on the wheat root surface and expression of the bacterial nifH gene during association. Molecular Plant-Microbe Interactions 6, 592-600.

Wood CC (1999) Ammonia fluxes across biological membranes: Towards and Azospirillum-wheat symbiosis. PhD thesis, University of Sydney, Australia.

93

Wood CC, Islam N, Ritchie RJ, Kennedy IR (2001) An experimental model for assessing critical parameters during associative 15N2 fixation between Azospirillum and wheat. Australian Journal of Plant Physiology 28, 969-974.

Xiaoji H, Xuejiang Z, Mulan, Qinjie H (1993) Inducing nodulation on oilseed rape roots inoculated with rhizobia with the help of pectin-lytic bacterium. In Proceedings of international Symposium on Nitrogen Fixation with Non-Leguminous Crops. (Eds N Yanfu, IR Kennedy and C Tingwei) pp. 66-71. (Qingdao Ocean University Press, Qingdao, China).

Yanzhen C, Fudi L (1993) Induction of para-nodules on non-leguminous crops by nitrogen fixing organisms with helper bacteria. In Proceedings of international Symposium on Nitrogen Fixation with Non-Leguminous Crops. (Eds N Yanfu, IR Kennedy and C Tingwei) pp. 72-76. (Qingdao Ocean University Press, Qingdao, China).

Zeman AMM, Tchan YT, Elmerich C, Kennedy I R (1992) Nitrogenase activity in wheat seedlings bearing para-nodules induced by 2,4-dichlorophenoxyacetic acid (2,4-D) and inoculated with Azospirillum. Research in Microbiology 143, 847-855.

94

10. Sesbania: A potential nitrogen source for sustainable rice production

A. T. M. A. Choudhury, S. K. Zaman and N. I. Bhuiyan 10.1 Abstract

The beneficial effect of Sesbania (Sesbania spp.), as a potential nitrogen (N) source, on rice production is well known. But it has not yet been adopted by the farmers in a large scale due to its competition with the main crop (rice) for land in intensive cultivation. Research works have been conducted at Bangladesh Rice Research Institute (BRRI) to find out the possible means to fit it in the rice cropping patterns without sacrificing rice crop. It has been found that Sesbania can be introduced during the fallow period of the dry season-fallow-rainy season rice-cropping pattern by different planting practices (seedling transplanting, seed broadcasting and planting stem cuttings). Different planting practices can accumulate 131.2-147.4 kg N ha-1 and give similar rice yield like 80 kg urea-N ha-1 in the succeeding rainy season rice. Among the planting practices stem cutting accumulated the highest amount of N (147.4 kg ha-1). The planting arrangement of 10 cm x 5 cm has been found superior to other planting arrangements (10 cm x 10 cm or 10 cm x 15 cm) of stem cutting with respect N accumulation, and grain production in the succeeding rainy season rice. Sesbania can also be introduced in summer season-rainy season rice cropping pattern by intercropping it with summer rice. This practice reduces the mean rice yield in the summer season by 0.9 t ha-1, but can completely substitutes the addition of urea-N in the following rainy season rice. These results clearly indicate that inclusion of Sesbania in both dry season-fallow-rainy season and summer season-rainy season rice cropping patterns is beneficial to sustain soil fertility and crop productivity in the long run. Keywords: Sesbania, nitrogen, rice. 10.2 Introduction

Nitrogen (N) is a primary macro nutrient element for all crops including rice. Rice plant requires large amount of mineral nutrients including N for its growth, development and grain production (De Datta 1981; Sahrawat 2000; Choudhury et al. 2001). Most of the rice soils of Asia are deficient in N. So fertiliser N application is essential to meet the crop requirement. Due to acute N deficiency in soils and higher N requirement for grain production, rice plant responses sharply to N fertilisation contributing higher yields (Choudhury and Bhuiyan 1991; Panda and Mohanty 1995; Shah et al. 1996; Choudhury et al. 1997). Generally urea is applied as N source for rice. But the efficiency of the added urea-N is very low, generally around 30-40%, and in many cases even lower (De Datta 1978; Choudhury and Bhuiyan 1994; Choudhury et al. 1994; Choudhury and Khanif 2001). This low N use efficiency is attributed to losses through denitrification, ammonia volatilisation, leaching and surface run-off (Ponnanperuma 1972; Fenn and Escarzuga 1976; Vleck and Craswell 1979; De Datta and Buresh 1989). Ammonia volatilisation and denitrification cause atmospheric pollution while leaching causes nitrate toxicity in the groundwater. These environmental problems are of great concern to the agronomists, soil and environmental scientists, and policy makers around the world. Alternate sources of N should be used in rice crop to minimise these problems. Green manuring can supplement the use of urea, and thereby can reduce these problems. Sesbania (Sesbania spp.) is an important green manuring crop with the potential to supply 100-150 kg N ha-1 within two months of sowing (Bhuiyan et al. 1988). It can assimilate both soil and atmospheric N (Dreyfus et al. 1983). In favourable environment, Sesbania can fix 70-458 kg N ha-1 from the atmosphere with 68-94% Ndfa (Nitrogen derived from atmosphere) within 45-65 days after sowing (Landa et al. 1992). It can be used as green manure to improve soil fertility and to increase nutrient supply for the crop. Its use increases soils capacity to absorb nutrients and improve soil structure and microbial activities (Zaman et al. 1994, 1997; BRRI 1996). Due to its extensive and deep root

95

systems, it can extract nutrients from deep soil layers, use insoluble or fixed forms of phosphorus, and make them available to the succeeding rice crop. Sesbania is salt-tolerant, and its incorporation improves the physico-chemical properties of saline-alkali soils, and thereby provides facilities for better rice growth and yields (Ladha and Kundu 1997). Literature on its N accumulation capacity and beneficial effects on rice cultivation is voluminous (Dargen et al. 1975; Bin 1983; Roger and Watanabe 1986; Ventura et al. 1987; Ghai et al. 1988; BRRI 1996). Despite available information on the beneficial effects of Sesbania, it is not yet adopted by the farmers in large-scale because of its competition with the rice crop for land. So it is necessary to find out the techniques to fit it in rice cropping patterns without sacrificing the rice crop. In this regard, research works have been conducted at Bangladesh Rice Research Institute (BRRI). Findings of these research works have been published elsewhere (Bhuiyan et al. 1988; Zaman et al. 1994, 1996; Bhuiyan and Zaman 1996; Choudhury et al. 1996). Some salient findings of these published works are reviewed here to accumulate the findings all together. 10.3 Potentials of Sesbania to supply plant nutrients

Although Sesbania is used as green manure to supply N, it has the potential to accumulate other nutrients, and supply those to the following crop (Bhuiyan and Zaman 1996). The data presented in Table 1 indicate that one ton dry weight of Sesbania aculeata can supply 33.2, 13.6, 14.0 and 16.2 kg N, K, Ca and Mg, respectively. In addition it can supply other elements in lesser extent. Sesbania can assimilate both soil and atmospheric N (Dreyfus et al. 1983). But the magnitude of N supply depends upon species (Bhuiyan and Zaman 1996). It has been found that Sesbania rostrata has the highest potential to supply N which is followed by Sesbania aculeata, and Sesbania sesban has the lowest N supply potential (Table 2). Table 1. Mineral nutrition of Sesbania aculeata and its nutrient supply capacity at eight weeks after sowing

Nutrients Content Supply (kg t-1 dry weight) Macronutrient (%) N 3.32 33.2 P 0.10 1.0 K 1.36 13.6 S 0.19 1.9 Ca 1.40 14.0 Mg 1.62 16.2 Micronutrient (mg kg-1) Zn 80.2 0.08 Cu 14.2 0.01 Mn 345 0.35 Fe 154 0.15

Source: Bhuiyan and Zaman (1996)

96

Table 2. Biomass production and nitrogen accumulation by three Sesbania species within 60 days of sowing

Sesbania species Shoot dry matter (t ha-

1) N content (%) N accumulation (kg ha-1)

Sesbania rostrata 7.4 3.4 252 Sesbania aculeata 6.7 2.7 181 Sesbania sesban 5.8 2.4 139

Source: Bhuiyan and Zaman (1996) 10.4 Establishment of Sesbania in the dry season-fallow-rainy

season cropping pattern

Despite its potential to supply N, Sesbania has not yet been adopted by the farmers in large-scale due to its competition with the main crop (rice) for land. Research works have been conducted at Bangladesh Rice Research Institute (BRRI) to find out the possible means to fit it in the the dry season (January-June)-fallow-rainy season (August-December) rice cropping pattern. Different planting practices and planting arrangements have been evaluated to find out the best mean of Sesbania establishment in this pattern.

Planting practices This work was conducted at BRRI during 1991 under an IFAD (International Fund for Agricultural Development) funded project in Bangladesh, co-ordinated by The International Rice Research Institute (IRRI). Details of the experimental procedures and findings are available in Zaman et al. (1994). Sesbania rostrata was planted in the fallow period of the dry season-fallow-rainy season cropping pattern. Three planting practices [seedling transplanting (20-day-old, spacing 20 cm x 10 cm), seed broadcasting (50 kg ha-1) and planting stem-cuttings (15 cm in length stem-cuttings, spacing 20 cm x 10 cm)] were evaluated. Sesbania rostrata was allowed to grow for 60 days. Then its biomass and the amount of N accumulation were recorded, and it was incorporated into the respective plots in-situ. After a week of Sesbania incorporation rice seedlings were transplanted. Data presented in Table 3 indicated that Sesbania rostrata accumulated 131.2-147.4 kg N ha-1. The performance of planting stem cutting was the best among the planting practices. It was also found that Sesbania substituted the use of urea-N completely in the succeeding rainy season rice crop by giving comparable grain yield (Table 4). Although there was no significant difference among the planting practices with respect to rice yield, planting stem cutting was identified as the best practice considering its highest amount of N accumulation (Table 3). Table 3. Effect of planting practices on nitrogen accumulation by Sesbania rostrata

Planting practice N accumulation (kg ha-1) by different plant parts Roo

t Stem Root nodule Stem nodule Tot

al Seedling transplanting

7.3 117.0 0.5 6.4 131.2

Seed broadcasting 6.8 102.6 0.8 3.6 113.8

Planting stem cutting 13.4

126.9 0.8 6.3 147.4

Source: Zaman et al. (1994)

97

Table 4. Effect of N fertilisation and Sesbania green manuring on yield and N uptake of rice

Treatment Grain yield (t ha-1)

Straw yield (t ha-1)

Total N uptake (kg ha-1)

Control (no nitrogen) 2.6 1.8 35 Urea-N (80 kg ha-1) 3.7 3.2 55 Sesbania seedling transplanting 3.8 3.5 60 Sesbania seed broadcasting 3.8 3.4 61 Planting Sesbania stem cutting 3.9 3.5 61 LSD (0.05) 0.45 0.57 -

Source: Zaman et al. (1994)

Planting arrangements The best performance of planting stem cuttings, as noticed in the previous study, led to conduct further research in order to find out the best planting arrangement of stem cuttings. In this regard a study was conducted at BRRI during 1993. Details of the experimental procedures and findings are available in Choudhury et al. (1996). Three planting arrangements (10 cm x 5 cm, 10 cm x 10 cm, and 10 cm x 15 cm) of Sesbania rostrata stem cuttings (25 cm in length) were evaluated. Sesbania rostrata stem cuttings were planted in the fallow period of dry season-fallow-rainy season rice cropping pattern using the three planting arrangements. Sesbania was allowed to grow for 60 days. Then its biomass and the amount of N accumulation were recorded. Sesbania was incorporated into the respective plots in-situ one week before planting rice. Two other treatments (urea-N at 80 kg ha-1 and control) were also included in the study to compare the effect of Sesbania on rice. Sesbania accumulated 61-140 kg N ha-1 (Table 5). The planting arrangement of 10 cm x 5 cm was identified as the best considering the biomass production and N accumulation. This planting arrangement gave significantly higher grain yield over 80 kg urea-N ha-1 in the succeeding rainy season rice (Table 6). Other planting practices were similar to urea-N in rice grain production. Table 5. Effect of planting arrangements of stem cutting Sesbania rostrata on biomass production and N accumulation

Planting arrangement Sesbania biomass (t ha-1) N content (%)

N accumulation (kg ha-1)

10 cm x 5 cm 4.5 3.1 140 10 cm x 10 cm 3.3 3.0 99 10 cm x 15 cm 1.9 3.2 61

Source: Choudhury et al. (1996)

98

Table 6. Effect of N fertilisation and planting arrangements of Sesbania rostrata green manuring (stem cutting) on yield and N uptake of rice

Treatment Grain yield (t ha-1) Straw yield (t ha-1) Total N uptake (kg ha-

1) Control (no nitrogen) 3.2 3.3 43 80 kg urea-N ha-1 4.0 5.1 65 Sesbania stem cutting at 10 cm x 5 cm spacing

4.5 4.8 76

Sesbania stem cutting at 10 cm x 10 cm spacing

4.0 5.2 64

Sesbania stem cutting at 10 cm x 15 cm spacing

3.9 4.6 60

LSD (0.05) 0.41 1.49 - Source: Choudhury et al. (1996) 10.5 Establishment of Sesbania in the summer season-rainy

season cropping pattern

Since there is no fallow period convenient for the growth of Sesbania between rice crops in the summer season (April-July)-rainy season (August-December) intensive rice cropping pattern, it is difficult to establish Sesbania in this pattern. Due to intensive cropping in this pattern soils would be depleted in N rapidly. To minimise this depletion problem there should be some means to include green manuring crop like Sesbania in this pattern, even by sacrificing rice yield to some extent, in order to maintain soil fertility status in the long run. In this regard research works have been also conducted at BRRI during 1993 by intercropping Sesbania with the summer rice. Details of the experimental procedures and findings are available in Zaman et al. (1996). Sesbania rostrata seedlings (30-day-old) were transplanted (spacing 5 cm x 20 cm) on the same day of rice transplanting (spacing 20 cm x 20 cm) between two adjacent rice rows. The experiment was conducted in a split-plot design using Sesbania intercropping (with and without) as the main plot and urea-N rate (0, 30, 60, 90 & 120 kg ha-1) as the sub-plots. Sesbania plants were clipped twice (at panicle initiation and flowering stages of rice) to avoid shading to rice crop. The clipped Sesbania biomass and the amount of its N accumulation were recorded before incorporating into the respective plots. Summer rice was harvested at maturity. Sesbania plants were allowed to grow for further 20 days. Then its biomass and the amount N accumulation were recorded before incorporating into the respective plots. Seven days after Sesbania incorporation, the rainy season rice seedlings were transplanted. Rice crop was harvested at maturity. Sesbania intercropping accumulated 138-209 kg N ha-1 (Table 7). The amount of Sesbania biomass and N accumulation decreased gradually with increasing N rates. Generally the activity of Rhizobium bacteria decreases with fertiliser N application. This may be the possible reason of decrease in Sesbania biomass with increasing rates of fertiliser N application. These results are supported by earlier findings (Gines et al. 1986).

99

Table 7. Effect of N fertilisation on biomass production and N accumulation by Sesbania rostrata intercropped with summer rice

N rate (kg ha-1)

Sesbania biomass (t ha-1)

Total N accumulation (kg ha-1)

Percent decrease in N accumulation over control

0 6.8 209 - 30 6.3 193 8 60 5.0 153 27 90 4.6 141 33 120 4.5 138 34

Source: Zaman et al. (1996) Sesbania intercropping decreased the mean grain yield of the summer season rice by 0.9 t ha-1 (Table 8). This was attributed to the competition between rice and Sesbania for light, space and nutrients. The beneficial effect of Sesbania intercropping was noticed in the succeeding rainy season rice (Table 9). Sesbania intercropping completely substituted the recommended rate of urea-N for rainy season rice (60 kg N ha-1). In control plots, Sesbania intercropping gave significantly higher grain yield over without Sesbania intercropped plots. The yield advantage was 0.8 t ha-1 in these plots. Although Sesbania intercropping decreased the mean rice yield by 0.9 t ha-1 in the summer season, it accumulated substantial amount of N which contributed in the complete substitution of urea-N in the following rainy season rice. This practice is beneficial for sustaining of soil fertility and productivity in the long run. Table 8. Effect of Sesbania rostrata intercropping and N fertilisation on grain yield of summer rice

Grain yield (t ha-1) N rate (kg ha-1) Without Sesbania With Sesbania Effect of N 0 2.5 1.4 2.0 b 30 2.7 1.8 2.3 ab 60 2.8 2.2 2.5 a 90 2.6 1.6 2.1 b 120 2.5 1.6 2.1 b Effect of Sesbania 2.6 A 1.7 B

Figures followed by different capital letters in a row and small letters in a column are significantly different at 5% level by Duncan's Multiple Range Test (DMRT). Source: Zaman et al. (1996) Table 9. Effect of Sesbania rostrata intercropping (with summer rice) and N fertilisation on the grain yield of the succeeding rainy season rice

Grain yield (t ha-1) N rate (kg ha-1) Without Sesbania With Sesbania 0 4.0 bB 4.8 aA 30 4.4 abA 4.6 aA 60 4.7 aA 4.6 aA 90 4.6 aA 4.0 bB 120 4.4 abA 3.5 cB

Figures followed by different capital letters in a row and small letters in a column are significantly different at 5% level by Duncan's Multiple Range Test (DMRT). Source: Zaman et al. (1996)

100

10.6 Conclusions

Sesbania can be established in both dry season-fallow-rainy season and summer season-rainy season rice cropping patterns. Planting of stem cutting with a spacing of 10 cm x 5 cm in the fallow period of the dry season-fallow-rainy season rice cropping pattern has been identified as the best technique to fit Sesbania in this pattern for complete substitution of urea-N for the rainy season rice. Intercropping Sesbania with summer rice in the summer season-rainy season rice cropping pattern can completely substitute urea-N for the rainy season rice. Although this practice reduces rice yield in the summer season, it is beneficial for sustaining soil fertility and productivity in the long run. 10.7 Acknowledgements

The authors are grateful to Bangladesh Rice Research Institute (BRRI) for providing facilities to conduct the research; to International Fund for Agricultural Development (IFAD) for funding part of the research works. 10.8 References

Bangladesh Rice Research Institute (BRRI) (1996) Annual Report for 1993. (Ed. HR Talukdar) pp. 45-67. (BRRI, Gazipur, Bangladesh).

Bhuiyan NI, Zaman SK (1996) Use of green manuring crops in rice fields for sustainable production in Bangladesh Agriculture. In 'Biological Nitrogen Fixation Associated with Rice Production'. (Eds. M rahman, AK Podder, CV Hove, ZNT Begum, T Heulin and A Hartmann) pp. 51-64. (Kluwer Academic Publishers, Dordrech, Netherlands).

Bhuiyan NI, Zaman SK , Panaullah GM (1988) Dhaincha green manure: A potential nitrogen source for rainfed lowland rice. In 'Proceedings of the Workshop on Experiences with Modern Rice Cultivation' pp. 108-125. (Bangladesh Rice Research Institute, Gazipur, Bangladesh).

Bin J (1983) Utilisation of green manure for raising soil fertility in China. Soil Science 135, 65-69. Choudhury ATMA, Bhuiyan NI (1991) Yield and nitrogen nutrition of modern rice as affected by

nitrogen fertilisation under irrigated culture. Bangladesh Rice Journal 2, 122-127. Choudhury ATMA, Bhuiyan NI (1994) Effect of rates and methods of nitrogen application on grain

yield and nitrogen uptake of wetland rice. Pakistan Journal of Scientific and Industrial Research 37, 104-107.

Choudhury ATMA, Bhuiyan NI, Hashem MA, Matin MA (1994) Nitrogen Fertiliser management in wetland rice culture. Thai Journal of Agricultural Science 27, 259-267.

Choudhury ATMA, Khanif YM (2001) Evaluation of the effects of nitrogen and magnesium fertilisation on rice yield and fertiliser nitrogen efficiency using 15N tracer technique. Journal of Plant Nutrition 24, 855-871.

Choudhury ATMA, Khanif YM, Aminuddin H, Zakaria W (2001) Potassium and magnesium adsorption properties of some Malaysian rice soils. In 'Proceedings of the International Conference on Agricultural Science and Technology, Session 2: Sustainable Agriculture'. (Eds. W Jingguo, H Bin, N Lihong and S Zhou) pp. 340-345. (Ministry of Science and Technology, Beijing, Peoples Republic of China).

Choudhury ATMA, Zaman SK, Bhuiyan NI (1996) Stem cutting of dhaincha (Sesbania rostrata) green manuring as substitute of urea-N for rainfed lowland rice. Bangladesh Journal of Agricultural Science 23, 101-104.

Choudhury ATMA, Zaman SK, Bhuiyan NI (1997) Nitrogen response behaviour of modern rice varieties with different agronomic traits. Thai Journal of Agricultural Science 30, 185-193.

Dargen SK, Chilli RK, Bhardwaj KKB (1975) Alkali soils: green manuring for more paddy. Indian Farming 25, 13-15.

De Datta SK (1978) Fertiliser management for efficient use in wetland rice soils. In 'Soils and Rice'. pp. 671-679. (International Rice Research Institute, Los Banos, Philippines).

101

De Datta SK (1981) Principles and Practices of Rice Production. pp. 348-419. (John Wiley and Sons Inc., New York, United States of America).

De Datta SK, Buresh RJ (1989) Integrated nitrogen management in irrigated rice. Advances in Soil Science 10, 143-169.

Drefus B, Elmerich C, Dommergues YR (1983) Free-living Rhizobium strain to grow under N2 as the sole nitrogen source. Applied Environmental Microbiology 45, 711-713.

Fenn LB, Escarzuga R (1976) Ammonia volatilisation from surface applications of ammonium compounds on calcareous soils. V. Soil water content and method of nitrogen application. Soil Science Society of America Journal 40, 537-541.

Ghai SK, Rao DLN, Batra L (1988) Nitrogen contribution to wetland rice by green manuring with Sesbania spp in an alkaline soil. Biology and Fertility of Soils 6, 22-25.

Gines HC, Furoc RE, Meelu OP, Dizon MA, Morris RA (1986) Studies on green manuring of rice in farmers' fields. Philippines Journal of Crop Science 7, 1-5.

Ladha JK, Kundu DK (1997) Legumes for sustaining soil fertility in lowland rice. In 'Extending Nitrogen Fixation Research', Proceedings of an International Workshop on Managing Legume Nitrogen Fixation in the Cropping Systems of Asia. (Eds. OP Rupela, C Johansen and DF Herridge) pp. 76-102. (International Crops Research Institute for the Semi-Arid Tropics, Patancheru, Andra Pradesh, India).

Ladha JK, Pareek RP, Becker M (1992) Stem nodulating legume-Rhizobium symbiosis and its agronomic use in lowland rice. Advances in Soil Science 20, 147-192.

Panda MM, Mohanty SK (1995) Time of application of low dose nitrogen to rainy season rice (Oryza sativa) for increasing N-use efficiency. Indian Journal of Agricultural Science 65, 283-285.

Ponnamperuma FN (1972) The chemistry of submerged soils. Advances in Agronomy 24, 29-96. Roger PA, Watanabe I (1986) Technology for utilising N2-fixation in wetland rice: potentials, current

usage and limiting factors. Fertiliser Research 9, 39-77. Sahrawat KL (2000) Macro and micronutrients removed by upland and lowland rice cultivars in

West Africa. Communications in Soil Science and Plant Analysis 31, 717-723. Shah AL, Choudhury ATMA, Rahman MS, Bhuiyan NI (1996) Nitrogen and sulphur interactions in

wetland rice. Bangladesh Journal of Scientific Research 14, 161-168. Ventura W, Mascarina GB, Furoc RE, Watanabe I (1987) Azolla and Sesbania as biofertilisers for

lowland rice. Philippine Journal of Crop Science 12, 61-69. Vleck PLG, Craswell ET (1979) Effect of nitrogen sources and management on NH3 volatilisation

losses from flooded soil. Soil Science Society of America Journal 42, 252-258. Zaman SK, Choudhury ATMA, Bhuiyan NI (1994) Stem cutting Sesbania rostrata: an approach of

green manure establishment for rainfed lowland rice. Thai Journal of Agricultural Science 27, 269-276.

Zaman SK, Choudhury ATMA, Bhuiyan NI (1996) Prospect of dhaincha (Sesbania rostrata) intercropping with T. Aus in a T. Aus-T. Aman cropping pattern. In 'Biological Nitrogen Fixation Associated with Rice Production'. (Eds. M rahman, AK Podder, CV Hove, ZNT Begum, T Heulin and A Hartmann) pp. 65-70. (Kluwer Academic Publishers, Dordrech, Netherlands).

Zaman SK, Parul SS, Ahmed HU, Bhuiyan NI (1997) Effect of ploughpan management on the performance of rice varieties under drought prone rainfed lowland environment. Thai Journal of Agricultural Science 30, 491-500.

102

11. An economic analysis of inoculant biofertiliser production and use in Vietnam

G. Barrett and S. Marsh 11.1 Abstract

Biofertiliser has the potential to increase rice yield and decrease the use of chemical fertilisers. Farmers in Ha Tay province reported yield increases of up to 20% in field trials using biofertiliser, as compared to conventional chemical fertiliser treatment. This practice increased average farm income substantially. The technology also offers economic and social benefits to communities engaged in biofertiliser production and use. Despite these benefits, this analysis identifies several constraints to the adoption of biofertiliser by the farmers. By focussing research efforts on achieving efficiencies in biofertiliser manufacture, combined with research and extension at farm level, there is further capacity for farmers to gain from the use of inoculant technology. Keywords: Economic analysis, biofertiliser, Vietnam.

11.2 Introduction

Biofertiliser inoculant technology offers potential economic and social benefits to individuals and rural communities at a number of different levels. Technical infrastructure associated with biofertiliser inoculant technology provides a new enterprise for communes involved in its production. At the farm level, the technology has economic advantages for farmers and communes, and also has potential long term ecological benefits associated with decreased use of chemical fertilisers and pesticides. The evidence presented in this paper suggests that there are significant economic benefits associated with income gains from production stemming from the adoption of biofertiliser by farmers in the communes under study. This paper outlines production costs, benefits and risks from an analysis of the Ba Vi biofertiliser factory’s operations. This is followed by an examination of expected benefits for farmers, including a hypothetical analysis of possible changes in income and costs, and discussion of benefits and risks for farmers. Finally, we make some suggestions on the focus of future work from a perspective emphasising farmer adoption of the technology. 11.3 Background

Biofertiliser can be used in the cultivation of rice and other crops, and it has the potential to increase yields and decrease the use of chemical fertilisers. In this paper, biofertiliser refers to the combination of starter (cultured microbial inoculants) with peat, rice husk, sugar and water to form an organic fertiliser. Biofertiliser production takes place in three factories located in the following provinces: Ha Tay, Hai Duong and Thai Nguyen. Starter culture for the factories is produced at a laboratory located at the Hanoi University of Science. Up to 200 kilograms of starter can be produced in the laboratory in each production period, and one kilogram of starter is sufficient to produce approximately 80 kilograms of biofertiliser. The Vietnam Women’s Union (VWU) and the Agricultural Extension Service of the Ministry of Agriculture and Rural Development, Vietnam, are actively involved in the management and production in the factories. In Ba Vi factory, located in Ha Tay province, four out of eight

103

memberson the Factory Management Board are from the VWU. Further, the VWU received the original grant for construction of the factory. The Agricultural Extension Service is responsible for the factory in Thai Nguyen province and is undertaking a marketing and extension role within other provinces to promote the use of biofertiliser. Biofertiliser is currently in its fourth season of use and there have been positive responses from those farmers who have adopted the new technology. Farmers from Tan Linh commune in Ba Vi district, who were recently surveyed, generally gave a positive indication to researchers that they were happy with results from the use of biofertiliser. From thirteen respondents who answered the questions, six responded that they strongly agreed with the statement ‘I will keep using biofertiliser’ and seven indicated that they agreed with the statement. Nine farmers indicated that they agreed or strongly agreed that crops grown with biofertiliser were healthier than crops grown with chemical fertiliser. A similar positive response was also found when farmers were asked whether they were satisfied with the results of the crop grown with biofertiliser, with eleven respondents saying they agreed or strongly agreed. We now turn our attention to an analysis of biofertiliser production at the factory level by examining input costs, expected benefits from production in Tan Linh commune and significant risks factory production may face. 11.4 The cost of biofertiliser production at Ba Vi biofertiliser

factory

Ba Vi factory, located in Tan Linh commune, commenced biofertiliser production in June 2000. Production takes place three times per year coinciding with the harvest period. At present total capacity for the factory is 600 t year-1 or 200 t crop-1. Current production is 45 t year-1 over the three production periods. It is assumed that if the factory can attain economies of scale in its production processes these gains will be passed on to farmers in the form of a reduced price for biofertiliser, a major input cost for farmers.

Ba Vi biofertiliser factory schedule of input costs Table 1 shows the breakdown of costs for production of 10 t of biofertiliser at Ba Vi factory. Analysis of the above schedule shows starter, transport and distribution are the major production costs of biofertiliser. Together these inputs account for over 50% of input costs. Ba Vi Factory Management has estimated that if production were to rise to 100 t costs would be reduced by approximately 5%. Lower prices for the starter culture may be possible for larger production volumes of biofertiliser as at current production levels there are assumed to be limited efficiencies in manufacture of the input. Benefits would also come from a reduction in fixed costs per kilo from inputs such as fuel, electricity, transport and distribution costs. Assuming that profit for the factory will be kept constant and efficiency gains from production will be transformed into a reduced cost of fertiliser, farmers can reduce fertiliser costs on average, by up to 5% or approximately 500 VND sao-1 at an application rate of 10 kg sao-1 [one Australian Dollar is equal to 7,500 VND (Vietnamese Dong), one sao is equal to 360 square meters].

104

Table 1. Breakdown of factory costs for the production of 10 t of biofertiliser (sold for 10,000,000 VND)

Input Cost (VND ‘000) Percentage of Cost (%) Starter (140 kg) 2,240 26 Peat (14 m3) 910 10 Sugar (140 kg) 840 10 Potassium carbonate (180 kg) 216 2 Petrol (fuel) 200 2 Electricity 72 1 Labour 1,000 12 Packaging (small) 450 5 Packaging (large) 240 3 Distribution to villages 1,300 15 Transport 1,000 12 Advocacy 200 2 TOTAL 8,668,000

Expected benefits for Ba Vi district Economic benefits accrue to the factory staff in the form of payments from the production of biofertiliser and subsequent flow on effects to the locality. For every tonne of biofertiliser produced, 100,000 VND is divided between the eight factory workers on the basis of the number of hours worked. This represents a transfer of funds from farmers in the district to factory workers who live locally, thereby spend wages locally, in contrast to chemical fertiliser companies which, can be assumed, do not inject that money back into the local economy. With an increase in factory production, there is the potential to employ more labour and hence provide greater flow on effects for the district economy. Although it is usual to exclude secondary costs and benefits from economic analyses, when non-competitive markets exist secondary outcomes should be identified and included (Sinden and Thampapillai 1995). In this case, the biofertiliser factory can increase production without imposing opportunity costs elsewhere in the economy, by using previously unemployed resources such as idle labour and surplus factory capacity. Even though the biofertiliser factory has considerable benefits for the district, there are still significant risks for factory production that need to be addressed.

Factory production risks Farmers place orders through the VWU up to a month prior to the biofertiliser becoming available. The factory then produces the ordered amount and payments for biofertiliser are made when it is delivered to the farmer. With this method of distribution there is a risk that farmers may not honour their agreement, made a month earlier, to purchase. This situation would leave the factory with a shortfall in funds needed to pay factory workers and other production costs. For example, in an earlier production period orders totalling 10 t were placed with the VWU however only 2 t was paid for, leaving a shortfall of 8 million VND for the Factory Management Board to cover. This risk could possibly be avoided by seeking deposits when orders for biofertiliser are made. However, in a survey conducted by the authors in September 2001, ten out of eleven farmers indicated they are strongly against paying a 20% deposit for orders. In addition, we presume asking for a deposit would make the biofertiliser option much less attractive for many farmers currently buying fertiliser from chemical companies offering attractive terms of payment. Price risks are also faced by the factory stemming from the need for a constant supply and price of inputs such as peat and sugar. Competition risk faced by the factory from chemical fertiliser companies is a major issue that the factory will have to face in the coming seasons. As trade barriers on chemical fertilisers are removed, in line with Vietnam’s entry into ASEAN (Association of South

105

East Asian Nations), the price of chemical fertilisers is likely to fall in the short to medium term, providing a greater incentive for farmers to use chemical fertilisers. To remain competitive and profitable, factories are going to have to look to research to find cheaper methods of production. 11.5 Expected economic benefits for farmers from the use of

biofertiliser technology

Farmers benefit from the use of biofertiliser in two ways: first, in the form of reduced input costs and second, in the form of increased income resulting from an increase in yield. In addition, flow on benefits resulting from an increase in the average level of income in the commune are also expected to occur in the long run if these benefits are sustained.

Economic benefits To investigate the economic benefits of biofertiliser use, data was collected from ten farmers in Cam Giang District, Hai Duong Province, comparing rice sown with and without biofertiliser. The field trials, supervised by scientists from Hanoi University of Science, were conducted by farmers who used biofertiliser on half of the plot and conventional fertiliser application on the other half of the plot. Results suggest that gains from increases in yield and hence income, are significant. Overall, nine farmers reduced chemical fertiliser inputs and experienced an increase in yield. Yield increases experienced by farmers in Cam Giang District were up to 20% (36 kg sao-1) higher with use of biofertiliser, with an average increase of 25 kg sao-1 (or the equivalent of 700 kg ha-1). Two farmers indicated negative cost savings, (i.e. an increase in fertiliser input costs) with the use of biofertiliser. We were informed that these farmers applied higher than the recommended amount of biofertiliser, or did not reduce the amount of chemical fertiliser application (Hien 2001, personal communications). Table 2 calculates the total benefit of using biofertiliser over a year using averages from eight farmers in Cam Giang District, Hai Duong Province (data adapted from Roughley 2000). For the farmers in this district, yield increases contributed up to an additional 55,000 VND sao-1 towards farm income, with an average increase in farmers’ income of 38,000 VND sao-1 (or AUD 142 ha-1). Results from these field trials indicated that there are substantial benefits to be gained by the communities involved in the use of biofertiliser technology.

These yield increases have been used to calculate potential income benefits for farmers in the Red River Delta. Based on average farm size and farm household income statistics, the use of biofertiliser has the potential to increase farmer income by 22% (Table 2). Despite these benefits the level of adoption of the technology was relatively low. Only 100 households out of 500 in the village were using biofertiliser in the fourth season of production.

106

Table 2. Actual and potential additional benefits from use of biofertiliser due to increase in yield VND AUD Calculations from field trials: Average savings in costs from use of biofertiliser

3,300 sao-1

12 ha-1

Average increase in income 38,000 sao-1 142 ha-1 Total increase in income attributable to biofertiliser use 41,300 sao-1 154 ha-1 Extrapolations based on average farm size in the Red River Delta: Average land holding per farmer (GSO Stats. 2000)

7 sao

2,520 sq. m

Potential increase in income (per farm per crop) 290,000 40 Potential increase in total income (assuming 2 crops per year) 580,000 80 Average rural income (GSO Stats. 2000) 2,700,000 360 Potential percentage increase in income from use of biofertiliser

22% 22%

Assumptions: AUD1 = VND7,500; 1ha = 28 sao; Constant rice price = VND1,500 kg-1; two rice seasons in a year. The above analysis illustrates that there are benefits from the use of biofertiliser over use of chemical fertilisers. Any incremental increase in income brings an additional capacity for the household to consume which, when expended locally, contributes to income generation for the district. Ultimately a sustained increase in the level of farmer income in the district would result in a rise in the general living standards of people in the area.

A hypothetical analysis of the economic benefits of fertiliser use Based on limited on-farm and experimental data, we have used a hypothetical analysis to look more closely at the potential economic benefits of using biofertiliser. We assume high and low fertiliser application rates, with accompanying average yield of 130 and 180 kg sao-1, respectively. Farmers at Ba Vi reported that the average district yield was 130 kg sao-1, and the fertiliser rates used were in the typical ranges (Hien 2001, personal communications). In the calculation of regimes using biofertiliser, the application rates of both urea and NPK (mixed nitrogen, phosphorus and potassium fertiliser) have been halved from their respective high and low rates. This was also suggested as appropriate practice. Table 3 shows the high and low fertiliser regimes, both with and without biofertiliser, and their respective costs. Under the low fertiliser regime, using biofertiliser results in a cost saving of 6,650 VND or approximately 17%. At the higher fertiliser application rate the savings are more substantial, 14,500 VND or approximately 24%. It is noticeable that these savings are substantially higher than those reported by the farmers in the field trials. Farmers may be either not cutting back on their chemical fertilisers as much as suggested, or alternatively, using more biofertiliser than suggested. However, for the purpose of this analysis, we assume that fertiliser cost savings are possible up to 25%, and that yield increases are possible up to 30% (increases up to 27% were reported by farmers taking part in the trials).

107

Table 3. Fertiliser regimes and costs used in the hypothetical analysis of economic benefits

Fertiliser applied (kg sao-1)* Fertiliser regime Urea NPK K Biofer

t.

Total cost** (VND sao-1)

Low – no biofertiliser 6 15 3 0 40,200 Low – with biofertiliser 3 7.5 3 10 33,550 High – no biofertiliser 10 20 5 0 60,500 High – with biofertiliser 5 10 5 10 46,000

* In each regime, farmyard manure is assumed to be applied at the same rate. **Fertiliser costs used are urea at 2300 VND kg-1, NPK at 1300 VND kg-1, K at 2300 VND kg-1 and biofertiliser at 1000 VND kg-1. Tables 4 and 5 show a range of possible economic outcomes from biofertiliser use at the two different fertiliser regimes. The low and high regimes are assumed to give standard yields of 130 and 180 kg sao-1, respectively. The benefits of a range of yield increases are then calculated with the price of rice at 1400 VND kg-1. These are combined with a range of possible cost savings on fertiliser inputs. The standard costs for the low and high fertiliser regimes are 40,000 and 60,000 VND sao-1, respectively (as calculated in Table 3). It can easily be seen that the benefits from potential yield increases outweigh the benefits from reduced input costs in an approximate ratio of 9:2. A 5% increase in yield is approximately as beneficial as a 25% reduction in costs. This provides some insight into the behaviour of some the farmers who conducted field trials. Farmers appear to be more interested in the possibility of obtaining higher yields, either by using high rates of chemical fertilisers with biofertiliser or by using higher than recommended rates of biofertiliser, than they are in obtaining savings from reduced input costs. Table 4. Potential economic benefits in total VND sao-1 from biofertiliser used in conjunction with a low fertiliser regime, assuming base rice yield of 130 kg sao-1 and rice price at 1400 VND kg-1

Yield increase % Reduction in fertiliser costs %

0 5 10 15 20 25 30

0 0 9100 18200 27300 36400 45500 54600 5 2000 11100 20200 29300 38400 47500 56600 10 4000 13100 22200 31300 40400 49500 58600 15 6000 15100 24200 33300 42400 51500 60600 20 8000 17100 26200 35300 44400 53500 62600 25 10000 19100 28200 37300 46400 55500 64600

At rice prices higher than 1400 VND kg-1, the potential benefits from yield increases (as opposed to saving on fertiliser costs) become even more attractive. At lower rice prices the benefits from yield increases are reduced. However, even at a rice price of 1000 VND kg-1, the benefits from yield increases still outweigh the benefits from reduced fertiliser costs in an approximate ratio of 3:1. Hence, even at low rice prices, the potential benefits from yield increases from a changed fertiliser regime using biofertiliser would still probably remain the major driving motivation for its adoption by farmers.

108

Table 5. Potential economic benefits in total VND sao-1 from biofertiliser used in conjunction with a high fertiliser regime, assuming base rice yield of 180 kg sao-1 and rice price at 1400 VND kg-1

Yield increase % Reduction in fertiliser costs %

0 5 10 15 20 25 30

0 0 12600 25200 37800 50400 63000 75600 5 3000 15600 28200 40800 53400 66000 78600 10 6000 18600 31200 43800 56400 69000 81600 15 9000 21600 34200 46800 59400 72000 84600 20 12000 24600 37200 49800 62400 75000 87600 25 15000 27600 40200 52800 65400 78000 90600

This analysis suggests that biofertiliser application should demonstrate yield increases for it to be widely adopted by the farmers. The savings in fertiliser costs are relatively small, and will potentially be smaller if the price of chemical fertilisers falls in the short to medium term in response to changed import regulations. This, combined with the attractive purchase conditions offered by chemical companies, make the adoption of biofertiliser on the basis of its cost savings alone of marginal benefit to farmers. Evidence suggests that farmers are generally slow to adopt technologies on the basis of environmental and social benefits, without accompanying individual economic benefits (Pannell 1999).

Impact of biofertiliser on farmers’ production and price risk There are two main risks that the rice farmer faces: crop failure (production risk) and income failure (price risk). The adoption of biofertiliser technology does not serve to reduce these farming risks, but it can help cushion the effects of a poor season or low prices by its potential to increase yields and hence allow the farmer an extra margin of income. Crop failure can be caused by the environmental factors that can potentially affect the crop, resulting in lower than expected yields. Because the nature of rice production in Vietnam is specific to geographic areas, environmental risks faced by farmers in each district varies. In the two communes observed in Ba Vi district, farms are irrigated so that the degree of water control is substantial. In other areas, droughts or floods or both can affect crops dependent on seasonal rainfall or natural flooding. Price risk relates to changes in the prices for farmers’ output during the growing period. It includes not only changes in the local price received by the farmer but also the stability of the world rice market which significantly affects local traders, particularly farmers growing for the export market. Higher yields from the same or lower input level allow the farmer a greater profit margin and therefore a greater capability to overcome seasonal or price variations. In the next section some of the environmental and social benefits, expected to be gained from the widespread use of biofertiliser, are outlined. 11.6 Perceptions of environmental and social benefits

Ecological and social benefits are expected to flow from the use of biofertiliser technology and the subsequent economic gains to be made if the practice is maintained. The extensive use of biofertiliser has the potential to better recycle the current nutrients contained in the soil and water of agricultural ecosystems and to reduce the negative impacts on ecosystems of chemical fertilisers (Kennedy and Hien 1999). Potential ecological benefits from the use of biofertiliser are: • reduction in the use of chemical fertilisers; • reduction in denitrification and other production of greenhouse gases;

109

• reduction in nitrate toxicity in the ground water by minimising leaching loss; and • reduction in pollution of waterways by reduction in the growth of algae and other harmful

microbes capable of toxin production. Widespread adoption of biofertiliser technology has the potential to alleviate problems associated with chemical fertiliser use, leading to benefits for the community of a cleaner agricultural environment. Social benefits include those stemming from the ecological benefits and, in addition, human capital benefits flowing from the education of villagers and farmers who will acquire training and production skills in new activities: the production of biofertiliser and the use of biofertiliser technology in their farming systems. Both of these activities have significant income generating potential. At present there remains scope for the introduction of biofertiliser technology because the current price and reported yield gains make it an attractive economic option for many farmers. Ecological and social benefits contribute to these direct economic benefits. In September 2001, fourteen farmers from Tan Linh commune in Ba Vi district were surveyed in relation to their opinions about biofertiliser use on their crops. Respondents were asked to write their opinions about advantages of using biofertiliser on their crops. Generally, respondents wrote two to three points and the most common responses are noted here. Twelve respondents indicated higher yields as the main advantage of using biofertiliser on their crops. Ten farmers indicated they experienced an improvement in soil quality (including softness and fertility). Up to half of the farmers surveyed wrote that a reduction in chemical fertiliser use was an advantage. Nearly all farmers indicated that they saw advantages in relation to the plant itself including greener leaves, longer stems and more seeds per panicle. It is interesting to note that the majority of farmers indicated changes to the soil quality and to the rice plant. These beneficial effects of biofertilisers are in agreement with some previous findings (Ladha and Kundu 1997; Kennedy and Islam 2001). Farmers were also asked to indicate some of the disadvantages stemming from the use of biofertiliser. Over half of the surveyed farmers indicated that they saw the short storage period of biofertiliser as a disadvantage of its use. Half of the farmers wrote that the ordering period (a requirement of fifteen days in advance) was also a disadvantage. A majority of farmers indicated that biofertiliser was difficult to handle (including mixing and spreading) when it was wet. Reflected in farmer opinion as advantages of using biofertiliser are the social and ecological benefits as outlined by scientists. The factors highlighted by farmers as disadvantages indicate areas for further research, and these are discussed below. 11.7 Areas for further study

The challenge for researchers is to discover a method of strain selection and production that further reduces the cost of biofertiliser. It should be highlighted that very few farmers noted the cost of biofertiliser as an advantage of adopting the new technology. If researchers are to increase the adoption rate of biofertiliser they may have to find more optimal combinations of other inputs with biofertiliser. The impact of a reduction in costs will be two fold. Lowered costs will mean that farmers will be more likely to use biofertiliser as a cheaper fertiliser alternative, and thereby reduce chemical fertiliser application as much as possible. Thus the ecological and social benefits discussed earlier can potentially be realised. Furthermore, comments from farmers indicated that other farmers may not be using biofertiliser because of the inconvenience of its preparation compared to chemical fertilisers. Farmers said that

110

the additional work required in preparing the biofertiliser for application, plus the need for forward ordering and lack of storability, makes biofertiliser a less attractive option. Addressing these issues at the factory and farm level and ensuring farmers are aware that the benefits of using biofertiliser, both in economic and environmental terms, outweigh the inconvenience will mean they are more receptive to changing their practices to biofertiliser use. In summary, research has a role to play in finding further means of costs reductions that will advantage farmers, by addressing strain selection procedures, manufacture and marketing of the biofertiliser, to ensure the enterprise remains viable in the face of changing relative prices. This is made more critical in light of the likelihood of falls in chemical fertiliser prices in the short-term, as domestic production protection is removed in line with requirements of the ASEAN (Association of South East Asian Nations) Free Trade Association, of which Vietnam is a member. 11.8 Conclusions

There remain areas in need of attention in the production of biofertiliser. These include the need to substantiate yield gains using biofertiliser, reduce the costs of biofertiliser, and address farmer concerns regarding biofertiliser use. Firstly, data from farmer field trials suggest that the gains from increases in yield are significant. Our hypothetical analysis shows that the economic benefits from yield increases far outweigh the economic benefits from cost savings in fertiliser application. There is a wealth of empirical evidence to show that farmers’ adoption behaviour is heavily influenced by self-interest, and that profitability is an important element of self-interest (Lindner 1987). A greater, proven yield increase with associated direct economic benefits to farmers will significantly improve farmer adoption of biofertiliser. Secondly, a further reduction in the cost of biofertiliser will also significantly increase the attractiveness of this organic substitute. It can be expected that competitive chemical fertiliser companies will actively try and ensure that their market share is not eroded by organic fertilisers. In addition, the price of chemical fertiliser is expected to fall in the short to medium term as a result of a Government decision to remove the required use of contracts when importing fertiliser. For biofertiliser producers this translates into another incentive to focus on lowering the price of the input for farmers. Lower application rates of biofertiliser for similar or the same results may also be an option. As mentioned previously, the contribution to profit from the use of biofertiliser made by the reduction in input costs is currently relatively small. Further gains in this area can only serve to make biofertiliser a more appealing option for many farmers. Finally, feedback from farmers in Ba Vi district has indicated that although there are many positive outcomes from the use of biofertiliser, the additional work required to prepare the biofertiliser for application, plus the need for forward ordering and lack of storability of biofertiliser erodes the attractiveness of the biofertiliser inoculant technology. In brief, giving attention to the above mentioned areas in research, extension, and the production and marketing of biofertiliser has the potential to increase the adoption of the product, thus giving many farmers the potential to increase their income through environmentally sound and economically viable practices.

111

11.9 Acknowledgements

The authors acknowledge information and assistance from members of the AusAID CARD Biofertiliser project, especially Professor Nguyen Thanh Hien, Hanoi University of Science. We also wish to acknowledge information and assistance offered by the Ba Vi District Biofertiliser Factory and farmers in Tan Linh commune, Ba Vi District, Ha Tay Province. Assistance given by AusAID Australian Youth Ambassadors for Development Program in the placement of one of the authors in Vietnam is also acknowledged, along with the support from colleagues at Hanoi Agricultural University No. 1, Hanoi, Vietnam. 11.10 References

GSO (General Statistical Office) (2000) Statistical Yearbook 2000. (Statistical Publishing House, Hanoi, Vietnam).

Hien, NT (2001). Hanoi University of Science, Vietnam (Personal Communications). Kennedy I, Hien NT (1999) Biofertiliser inoculant technology for the growth of rice in Vietnam:

Developing technical infrastructure for quality assurance and village production for farmers, Capacity Building for Agriculture and Rural Development (CARD) Program, Project Proposal. Australian Agency for International Development, Canberra, ACT, Australia.

Kennedy IR, Islam N (2001) The current and potential contribution of asymbiotic nitrogen fixation to nitrogen requirements on farms: a review. Australian Journal of Experimental Agriculture 41, 447-457.

Ladha JK, Kundu DK (1997) Legumes for sustaining soil fertility in lowland rice. In 'Extending Nitrogen Fixation Research', Proceedings of an International Workshop on Managing Legume Nitrogen Fixation in the Cropping Systems of Asia. (Eds. OP Rupela, C Johansen and DF Herridge) pp. 76-102. (International Crops Research Institute for the Semi-Arid Tropics, Patancheru, Andra Pradesh, India).

Lindner RK (1987) Adoption and diffusion of technology: An overview. In ‘Technological Change in Postharvest Handling and Transportation of Grains in the Humid Tropics’, Australian Agency for International Development Proceedings Number 19. (Eds. BR Champ, E Highley and JV Remenyi) pp. 144-51. (Australian Agency for International Development, Canberra, ACT, Australia).

Pannell DJ (1999) Social and economic challenges in the development of complex farming systems. Agroforestry Systems 45, 393-409.

Roughley RJ (2000) Report of the Fourth Visit to Vietnam during 18th October to 2nd November 2000. (Unpublished).

Sinden JA, Thampapillai DJ (1995) Introduction to Benefit-Cost Analysis. (Longman Publishing Pty Ltd, Melbourne, Australia).

112

12. A model for testing the effectiveness of biofertiliser for Australian rice production

R.L. Williams and I.R. Kennedy 12.1 Abstract

The success of inoculant biofertiliser in experimental and farmer trials in Vietnam encourages the application of similar technology in Australia. Benefits would arise from increased yield and in reduced input costs. In this short paper, we examine the nature of the current Australian rice production system to assess the likelihood of particular problems in applying biofertiliser and recommend a strategy to define and overcome these. It is recommended that trials begin in Australian rice fields as soon as possible because similar yield increases as in Vietnam would be welcomed by Australian rice farmers. 12.2 Introduction

Experimental and farmers’ field trials have shown the effectiveness of the multi-strain inoculant biofertiliser (BioGro) for rice in Vietnam. A biofertiliser containing three strains of bacteria has been shown to consistently produce increases in grain yield in the range 10-20% (Hien et al. 2002, this volume). These positive effects are obtained using inoculation of rice seedlings either before transplanting or at transplanting into flooded rice paddies. Similar positive effects on grain yield of around 20% with reduced chemical N-fertiliser input on farms for rice production with the application of a multi-strain product called BioPower are claimed in Pakistan (Malik et al. 2000). In this short paper we consider the specific challenges presented in attempting to apply such a biofertiliser to rice production in Australia, a much more capital-intensive system. 12.3 Methods

Current practices Compared to Vietnam where a significant proportion of the population of 75 million is engaged in rice production, each farmer harvesting from only a fraction of a hectare, the Australian rice industry involves much larger farms although it is quite small overall. The industry involves only 2,000 farmers irrigating 130,000 ha producing 1.1million tonnes of semi-dwarf Japonica varieties of paddy, however most of this is exported. The rice is grown in rotation with other crops and pasture, without government subsidies. The average yield has steadily increased from about 3 tonnes to 8 tonnes ha-1 in the last 70 years as a result of improved technology. Nearly all Australian rice is aerially seeded pre-germinated into pre-flooded rice bays, although 5% is combine-sowed, flushed three times and then flooded one month after planting. Currently, the only fertilizer additions are N as urea, P and zinc. In the past, additions of P and S to pastures in rotation have provided sufficient of these nutrients. Current inputs of N amount to 124 kg ha-1 with 93 exported in rice grain. Only K is in significant deficit. Nitrogen is applied prior to flooding and once mid-season added to flood water. The average rate of 120 kg ha-1 can increase yields from 6 to 12 tonnes ha-1 in deficient sites. The rice plant has a discontinuous root system ((Matsuo and Hoshikawa 1993), without a tap root. The first root systems, both the seminal and the coleoptilar, are temporary and poorly aerated. Nodal roots from the first stem node appear after about the third leaf stage of growth at around three weeks. The soil itself is strongly anaerobic during crop growth with a redox potential of –300 mV from early September, a month after planting.

113

Plate 1: Australian rice at harvest. Rice is grown in rotation with other crops and pasture for livestock production, as is evident in this photograph.

Application of inoculant biofertiliser to Australian rice The challenges to application of inoculant biofertiliser to rice involves uncertainty about the effects of the mode of discontinuous root development on microbial colonisation. Loss of the first root systems may reduce the degree of colonisation although if stems can also be infected this effect might be minor. This requires investigation and can be done beforehand in the laboratory. In the first instance, field trials similar to those described elsewhere in this volume (Hien et al. 2002) and in an ACIAR Final Report (Kennedy and Hien 2001) are recommended. The strategy would be designed to maximise the probability of obtaining positive effects on yield, in order to have a field system for subsequent optimisation. Permission from AQIS to employ the Vietnamese strains in field studies in Australia would be an advantage, although the final system adopted might well involve the use of new strains selected from the rhizospheres of Australian ricefields. New studies would be required to study the following factors likely to affect the degree of colonisation with microbial strains in rice paddies during aerial seeding:

• An inoculum of sufficient size could probably be applied to each pre-germinated plantlet used in aerial planting. Inocula can be prepared containing 109 bacterial cells per g and sufficient cells of the order of 104-106 per seed applied.

• Effective establishment and colonisation with microbes in rice paddies could require measures to ensure a low death rate of bacteria during seeding and subsequently. Negative factors are likely to be excessive drying or loss of inoculum from the seedling during planting, which may require adhesive additives to prevent.

• Evidence of continued significant colonisation during the period up to grain filling would be required. Some of the positive effects claimed for biofertiliser strains such as the PGPR effect probably depend on phytohormone production by relatively small numbers of microbial cells per rice plant (say 105 per fresh g). However other effects such as P-solubilisation and N-mobilisation or biological N2 fixation would require bacterial cell numbers of the order of 107 cells per g of fresh tissue in order to benefit the plant as improved grain yields.

• Any such study should involve quality control measures to ensure that benefits from biofertiliser applications are obtained and can be sustained.

114

12.4 Conclusion

Benefits from inoculation with biofertiliser strains shown to increase the yield of rice elsewhere are an exciting prospect for the Australian rice industry. A system that reduces input costs and also reduces the likelihood of release of nitrous oxide, a gas with a greenhouse effect 300 times greater than carbon dioxide, is very attractive to the industry. Most of all, however, consistent increases in grain yield would be the factor necessary to ensure adoption of the technology by farmers. Although there are scientific concerns about establishment of effective associations between the rice plants and microbes, trials to assess the performance of currently available products are recommended. 12.5 References

Kennedy IR, Hien NT (2001) Microbial Biofertilisers for Sustainable and Environmentally Sound Crop Production in Vietnam and Australia. A Final Project Report Submitted to Australian Centre for International Agricultural Research (ACIAR), Canberra, ACT, Australia.41pp.

Hien NT, Roughley RJ, Kennedy IR (2002) The response of field-grown rice to inoculation with a multi-strain biofertiliser in Hanoi district, Vietnam. This volume.

Malik KA, Miza S, Melnaz S, Rasul G (2000) The role of plant-associated beneficial bacteria in rice-wheat cropping system. Abstracts 8th International Symposium on Nitrogen Fixation with Non-legumes. p.46.

Matsuo T, Hoshikawa K (1993) Science of the Rice Plant, Morphology. (Food and Agriculture Policy Research Center, Japanese Ministry of Agriculture, Forestry and Fisheries, Tokyo, Japan).

115

13. Author Affiliations J. Balandreau Ecologie Microbienne, Bât. 741, Université Claude Bernard Lyon I 43 Boulevard du 11 novembre, 69622 Villeurbanne cedex, France. Email: balandreau@univ-lyon1.fr R. Bally Laboratoire d'Ecologie Microbienne du Sol, UMR CNRS 5557, UCB Lyon1, 69622 Villeurbanne Cedex, France M. Bänziger International Maize and Wheat Improvement Centre (CIMMYT) Apdo. Postal 6-641, 06600 Mexico, D.F., Mexico Posted in Zimbabwe G. Barrett Australian Youth Ambassador for Development, Intake 5, Vietnam Faculty of Economics and Rural Development Hanoi Agricultural University No. 1, Hanoi, Vietnam Corresponding author, email: galina_barret@hotmail.com N. I. Bhuiyan Director Research, Bangladesh Rice Research Institute Gazipur 1701, Bangladesh K. Cassaday International Maize and Wheat Improvement Centre (CIMMYT) Apdo. Postal 6-641, 06600 Mexico, D.F., Mexico A. T. M. A. Choudhury SUNFix Centre for Nitrogen Fixation; School of Land, Water and Crop Sciences, Ross Street Building A03, The University of Sydney, NSW 2006, Australia Corresponding author, email: a.choudhury@acss.usyd.edu.au E. C. Cocking Centre for Crop Nitrogen Fixation, University of Nottingham University Park, Nottingham NG7 2RD, Nottingham, United Kingdom Email: edward.cocking@nottingham.ac.uk U. Hassan Biofertilizer Division, National Institute for Biotechnology and genetic Engineering (NIBGE) P.O. Box 577, Jhang Road, Faisalabad, Pakistan J. Haurat Laboratoire d'Ecologie Microbienne du Sol, UMR CNRS 5557, UCB Lyon1, 69622 Villeurbanne Cedex, France Nguyen Thanh Hien Genetics Department, Faculty of Biology Hanoi University of Science 90 Nguyen Trai Road, Hanoi, Vietnam Corresponding author, email: hnuniversity.hien@hn.vnn.vn

116

N. Islam Division of Plant Genetic Engineering Kihara Institute for Biological Research Yokohama City University Maioka-cho*@641-122*Atotsuka-ku Yokohama 244-0813, Japan I. R. Kennedy SUNFix Centre for Nitrogen Fixation; School of Land, Water and Crop Sciences, Ross Street Building A03, The University of Sydney, NSW 2006, Australia Email: i.kennedy@acss.usyd.edu.au K. A. Malik Pakistan Atomic Energy Commission (PAEC), P.O.Box 1114, Islamabad Pakistan S. Marsh Department of Agricultural Economics University of Sydney, NSW 2006, Australia M. Marziah Department of Biochemistry and Microbiology Universiti Putra Malaysia, 43400 UPM, Serdang, Selangor Darul Ehsan, Malaysia S. Mehnaz Biofertilizer Division, National Institute for Biotechnology and genetic Engineering (NIBGE) P.O. Box 577, Jhang Road, Faisalabad, Pakistan M. A. B. Mia Departments of Land Management Universiti Putra Malaysia, 43400 UPM, Serdang, Selangor Darul Ehsan, Malaysia Corresponding author, E-mail: basetmia@hotmail.com M. H. Mian Vice Chancellor Hajee Mohammad Danesh Science and Technology University Bansherhat, Dinajpur, Bangladesh M.S. Mirza Biofertilizer Division, National Institute for Biotechnology and genetic Engineering (NIBGE) P.O. Box 577, Jhang Road, Faisalabad, Pakistan Corresponding author, email: sajjad_mirza@yahoo.com P. Normand Laboratoire d'Ecologie Microbienne du Sol, UMR CNRS 5557, UCB Lyon1, 69622 Villeurbanne Cedex, France I. Ortiz-Monasterio International Maize and Wheat Improvement Centre (CIMMYT) Apdo. Postal 6-641, 06600 Mexico, D.F., Mexico C. V. S. Rao SUNFix Centre for Nitrogen Fixation; School of Land, Water and Crop Sciences, Ross Street Building A03, The University of Sydney, NSW 2006, Australia

117

G. Rasul Biofertilizer Division, National Institute for Biotechnology and genetic Engineering (NIBGE) P.O. Box 577, Jhang Road, Faisalabad, Pakistan T. G. Reeves Director General International Maize and Wheat Improvement Centre (CIMMYT) Apdo. Postal 6-641, 06600 Mexico, D.F., Mexico Corresponding author, email: CIMMYT@cgiar.org R. J. Roughley SUNFix Centre for Nitrogen Fixation; School of Land, Water and Crop Sciences, Ross Street Building A03, The University of Sydney, NSW 2006, Australia Z. H. Shamsuddin Departments of Land Management Universiti Putra Malaysia, 43400 UPM, Serdang, Selangor Darul Ehsan, Malaysia S. R. Waddington International Maize and Wheat Improvement Centre (CIMMYT) Apdo. Postal 6-641, 06600 Mexico, D.F., Mexico Posted in Zimbabwe R. L. Williams Yanco Agricultural Institute PMB, Yanco NSW 2703, Australia Email: robert.williams@agric.nsw.gov.au W. Zakaria Departments of Crop Science Universiti Putra Malaysia, 43400 UPM, Serdang, Selangor Darul Ehsan, Malaysia S. K. Zaman Soil Science Division, Bangladesh Rice Research Institute Gazipur 1701, Bangladesh