Copyright is owned by the Author of the thesis. Permission is given for a copy to be downloaded by an individual for the purpose of research and private study only. The thesis may not be reproduced elsewhere without the permission of the Author. DUAL MUTUALISTIC ASSOCIATIONS IN SAINFOIN (Onobrychis viciifolia $cop.) A thesis presented in partial fulfilment of the requireme nts for the degree of Master of Agricultural Sci en ce in Agronomy at Massey Univers ity Kee Fui? Kon 1982 i i . ABSTRACT Recent studies established that many legumes , when in fected with the appropriate Rhizobiwn spp. and arbuscular fungi, nodul ated better and exhibited greater dinitrogen fi xa tion than plants infected with only the rhizobia. A similar study, therefore, was carried out in a glasshou se using sainfoin (Onobrychis viciifoZia Scop.J, a legume that is rapidly gaining recognition as a potential forag e crop in New Zealand and other parts of the world. Pre -germinated seeds (cv. Fakir) were planted in sterilized soils and incubated with an effective Rhizobiwn spp. (strain NZP 5301), a mixture of endophytes (Gigasp.o:t.'a magaPita Becker & Hal 1, G'lomus fasciculata (Thax. sensu Gerd.) Gerdemann & Trappe and G'lomus tenuis (Greenall) Hall), or both eht rhi zob ·ia and endophytes. The experiment also included a control, without any inocu lation. Endophyte in fection, nodulation and dinitrogen fixation, total nitrogen and phosphorus concentrations, and plant growth and development were determined on eleven sequential samplings over about twenty v,eeks, up to the stage of green inflorescence. Arbuscular mycorrhiza formation did not occur with the first endophyte inocul ation, containing GigaspoY'a magaY'ita Becker & Hall, even after 93 days of growth. This is probab ly because the inoculum used consisted of a low quantity of viable spores and mycelia. The second inoculation, containing the three endophyte species, produced only a low degree of infection between day 115 and 137, possibly because the extensive root lignification and relatively higher root phosphorus concentration (0.50%) restricted fungal invasion and establishn~nt within the root cortex. Mycorrhiza formation did not increase phosphate uptake, improve nodul ati on and dinitrogen fi xati on, or increase p 1 ant grov,th. This is due probably to the al ready \ve 11-deve loped root systems that were efficiently exploiting the small soil volume within the bags. Rhizobia-inoculated plants produced more nodules, l~rger nodules and consequently, a greater nodule dry weight than the uninocu­ lated plants. The nodules produced in the inoculated plants were red iii . instead of green as in the uninoculated plants, and exhibited a greater dinitrogen fixation. As a result, these inoculated plants contained a higher concentration of shoot~ root and nodule nitrogen, and a greater dry weight accumulation in the shoots and nodules. The shoot and nodule phosphorus concentrations, however, were lower in the rhizobia-inoculated than in the uninoculated plants due to the greater amount of shoot and nodule tissues which caused a dilution ·effect. These rhizobia effects on nodulation and dinitrogen fixation, nitrogen and phosphorus concentrations, and plant growth and development became more prominent with time. The relatively higher nodule phosphorus concentration when compared with the shoot and root phosphorus concentrations suggests that phosphorus was presumably required in large quantities by the dinitrogen-fixing system. iv. PREFACE of life. Coexistence of organisms has long been recognised as an axism In 1952, Paul R. Bulkholder formally and objectively inter- preted coexistence as different biological interactions. Based on his coaction theory, these interactions were classified into and named as ' nine separate categories of which the most studied in agricultural ecology are competition and mutualism. In this thesis, two examples, of mutualism, involving a forage legume (Onobrychis viciifolia Scop.), a nitrogen-fixing bacterium (Rhizobium spp.J and three species of arbuscular fungi (Gigaspora magari ta Becker & Hall, Glomus fasciculatus (Thax. sensu Gerd.) Gerdemann & Trappe and Glomus tenuis (Greenall) Hall), are examined. The intention of this study was to investigate the real value of coexistence of these organisms from an agricultural standpoint and, therefore, emphasis is placed on the effects of the bacterium and fungi on the nutrition, and growth and development of sainfoin. While the bulk of chapters 4, 5, 6 and 7 is devoted to these topics, the relevant background information of the research is also included in the first three chapters. Various persons were directly and indirectly involved in the completion of this work. I am deeply indebted to Mr Angus G. Robertson for his close supervision and unceasing availability in offering advice, suggestions and practical assistance during this entire masterate prog­ ramme, and his many criticisms and recommendations during editing of the manuscript. I must also ~cknowledge his foremost contribution to me as a research student in helping me to develop the skill of more effective thinking in scientifi~ research. Dr Conway Ll. Powell, of the MAF Ruakura Soil and Plant Research Station in Hamilton, was most generous in supplying a substantial quantity of fungal inocula as my initial cultures. Throughout the study, he, being an outstanding world authority on mycorrhiza research, showed a deep interest in the work and provided many prompt suggestions which were invaluable. Sainfoin seeds (cv. Fakir) were kindly supplied by Mr Jim A. Fortune, of the Agronomy Department. I am also grateful for the permission to sample some sainfoin plants from his experimental plots, and his many suggestions. v. The methodol ogy of acetyl ene reduction assay was kindly introduced and demonstrated by Dr Jim .A. Crush and Mr Paul Yarrell, of the DSIR Grass l ands Division 1 Palmerston North. Owing to certain unavailable glassware, the assaying procedure was s lightly modified, but the value of their contributions remains. I am thankful for the privil ege to use the Pye gas chromatograph and other facilities in the Botany Department as well as the technical assistance given by Dr David W. Fountain and his technician, Mr Chong Loong Kan. The colorimetric autoanalysis of both t ota l plant nitrogen and phosphorus was kindly conducted by Mr Russe ll W. Till man, of the Soil Science Department and , therefore, a considerab le amount of routine work was reduced, enabling me t o concentrate on other aspects of the study. Hi s instructions on the preparati on of the Kjeldahl digest reagent and Kje l dahl digestion are also fully apprec iated. I am exceedingly grateful to Mr Hugh Nielson, of the Horti­ culture and Plant Health Department , for the supp ly of some chemical reagents and classware, and hi s ass i stance i n compound-microscope photography . All the micrographs in this vol ume are hi s f ine work . Ap preciation i s expressed to Dr Murray J . Hill for the permi ss i on to use the wei gh ing faci li ties in the Seed Technology Centre and the assistance received f rom hi s technician, Mrs Karen Johnstone. I wi sh to th ank Dr Ian L. Gordon, of the Agronon~ Depa r tment and Mr Greg C. Arn old, of the Mathematics and Statistics Department, for advice in statistical methods . I am also exceedingly grateful to Dr Neil A. Macgregor for hi s general recommendations and the great interest he took in thi s research. To my typi st, Mrs Cecily Willbond, I wish to extend n~ si ncere appreciation fo r her efficient and excellent work . Very special appreciation must be made to my wife, Lih Ju, for her long-suffering~ fin ancial assistance and unsacrifi cial contri­ bution of her time and energy in hel ping me in the expe rimental work, while also fulfilling the role of a homemaker and breadwinner. Financial awards from Helen E. Akers (two years), John Alexander Hurley, William Hudson and the Christian Centre Palmerston North are also gratefully acknowledged. Finally, I wi sh to acknowl edge the inspiration from t he Holy vi. Spirit and God's gift of the ineffable awesome creation which I intim­ ately worked with for over five months. The opportunity is here for me to return the magnificent glory of His i ngeni'ous design whi eh aptly speaks of His omniscience. ABSTRACT PREFACE TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES LIST OF PLATES LIST OF APPE NDICES CHAPTER 1. INTRODU CTION 2. TERMINOLOGY TABLE OF CONTENTS 2.1 Introduction 2.2 Terms Used in Dinitrogen Fixation 2.21 Infection, invasi on and infectivity 2.22 Mutualism and symbiosis vi. i. Page ii . iv. vii. xi. xv. xvi i. xv·i ii . 1. 5. 5. 5. 5. 5. 2.23 Rhizobia, Rhizobiwn and Rhizobiwn spp. 6. 2.3 Terms Used in Mycorri zae 2. 31 Endophyte 2.32 Mycorrhiza types 3. REVIEW OF LITERATURE 3.1 Introduction 3.2 Sainfoin as a Crop 3.21 General distribution and hi story 3.22 Botanical description 3.23 Some agricultural characteristics 3.24 Suitability in New Zealand conditions 3.3 Dinitrogen-Fi xing System of Legumes 6. 6. 6. 7. 8. 8. 10. 12. 15. 17. 3.31 Development of the legume-Rhizobi wn system 17. 3.32 Dinitrogen fixation 20 . 3.33 Methods of improving dinitrogen fixation 25. CHAPTER 3.4 Mycorrhiza System of Legumes 3.41 Occurrence 3.42 Development of arbuscular mycorrhizae 3.43 Arbuscular mycorrhizae and phosphorus nutrition 3.44 Arbuscular mycorrhizae and nitrogen nutrition 3.5 Agricultural Significance of Dual Mutualistic viii. Page 28. 28. 29. 33. 40. Associations 42. 4. EXPERIMENTAL MATERIALS AND METHODS 4.1 Introduction 4.2 Preliminary Experiments 4.21 Glasshouse test 4.22 Field test 4.3 The Principal Experiment 4.31 Experimental design and layout 4.32 Schedule of destructive sampling 4.33 Experimental measurements 4.4 Preparation of Materials 4.41 Growth containers 4.42 Growth medium 4.43 Sainfoin seedlings 4.44 Rhizobia inoculum 4.45 Endophyte inoculum 4.46 Nutrient solutions 4.5 Experimental Procedures 4.51 Control of glasshouse conditions 4.52 Potting 4.53 Inoculation of endophyte 4.54 Planting 4.55 Inoculation of rhizobia 4.56 Thinning and replacement of seedlings 4.57 Watering 4.58 Nutrient application 4.6 Analytical Techniques 4.61 Sampling method 4.62 Removal and washing of plants 44. 44. 44. 44. 45. 45. 45. 47. 48. 48. 48. 48. 48. 50. 50. 50. 51. 51. 51. 51. 52. 52. 52. 53. 53. 53. 53. 54. CHAPTER 4.63 Determination of plant dry weight 4.64 Estimation of endophyte infection 4.65 Determination of dinitrogen fixation 4.66 Determination of total nitrogen and phosphorus . 4.7 Problem with Endophyte Infection 4.71 Reisolation of Fm 4.8 4.72 Reinoculation with MX 4.73 Detection of condensed tannins 4. 74 Determination of root morphology Statistical Methods 4.81 Data transformation 4.82 Analysis of variance 4.83 Coefficient of variation and standard error 4.84 Duncan's new multiple range test 4.85 Regression analys i s 5. EXPERIMENTAL RESULTS 5.1 Introduction 5.2 Preliminary Resu lts 5.21 Glasshouse test 5.22 Field test 5.3 Development of Arbuscular Mycorrhizae 5.31 First inoculation with Fm 5.4 5.5 5.6 5.32 Condensed tannins and root morphology of sain foin 5.33 Second inoculation with MX Development of Dinitrogen-Fi xing System 5. 41 Infection 5.42 Nodulation 5.43 5.44 Total 5.51 5.52 Dinitrogen fixation Regression analysis Nitrogen and Phosphorus in Plant Tissue Tota 1 nitrogen Total phosphorus 5.53 Correlation studies Growth and Deve lopment of Sainfoin i X. Page 54. 54. 55. 56 . 57. 57. 57. 57. 58. 58. 58. 59. 61. 61. 61. 64. 64 . 64. 64. 64. 66. 66. 66. 66. 69. 69. 69. 75. 80. 80 . 80. 90. 98. 101. x. CHAPTER Page 5.61 Plant dry weight 101. 5.62 Root-shoot ratio 107. 5.63 Secondary stem production 107. 5.64 Regression analysis 114. 6. DISCUSSION 115. 6.1 Introduction 115. 6.2 Development of Arbuscular Mycorrhizae 115. 6.21 First inoculation with Gm 115. 6.22 Second inoculation with MX 117. 6.3 Development of Dinitrogen-Fixing System 118. 6.31 Infection 118. 6.32 Nodulation 118. 6.33 Di nitrogen fixation 120. 6.4 Total Nitrogen and Phosphorus in Plant Tissue 122. 6.41 Total nitrogen 122. 6.42 Total phosphorus 124. 6.5 Growth and Development of Sainfoin 126. 6.51 Plant dry weight 126. 6.52 Root-shoot ratio 127. 6.53 Secondary stem production 128. 7. CONCLUSIONS 129. APPENDICES 131. REFERENCES CITED 175. Table xi. LIST OF TABLES Page 1.1 Concentration of sixteen elements in complex plants 2. 1.2 Efficiency and contribution from various dinitrogen-fi xing systems 2. 3.1 Carbon and nitrogen in the nodules of three legumes 24. 4.1 Schedule of destructive sampling 47. 4.2 Computed variables used in statisti ca l ana lyses 49. 4.3 Equations of transformation of various vari ab les 60. 5.1 Means of endophyte-infected root cortex of sa infoin (var. bifera Hort.) in the glasshouse 65. 5.2 Means of endophyte- in fected root cortex of sainfoi n (cvs . Fakir and Remont) in the field 65. 5.3 Means of infected root cortex of corn with Gigaspora magari t a Becker & Ha ll 67. 5.4 Means of conden sed t anni n score of different sainfoin plant components 67. 5.5 Mean root diameter, mean ex tension of the root hair cylinder from the epidermis and root hair abundance for tertiary rootlets of four legumes 68. 5.6 Means of infected root cortex of corn and sainfoin with a mixture of endophy tes (MX) 70. 5.7 Transformed treatment means of endophyte infection per field of view (x 10- 2 ) 70. 5.8 Trans formed treatment means of nodul e number per plant 73. 5.9 Transformed treatment means of nodul e dry weight per plant 74. 5 .10 Transformed treatment means of nodule dry weight per nodule (Xl0- 2 ) 76. 5.11 Transformed treatment means of acetylene reduced per hour per plant 77. Table xii • . Page 5.12 Original treatment means of acetylene reduced per hour per gram nodule dry weight 78. 5.13 T-test values for homogeneity of regression slopes between nodule dry weight and nodule number, and nodul e dry weight and nodule size 81. 5.14 Summary of stepwise multiple regression analysis for different treatments 82. 5.15 Transformed treatment means of total cotyledon nitrogen 84. 5.16 Transformed treatment means of total shoot nitrogen 85. 5.17 Transformed treatment means of total root nitrogen 86. 5.18 Transformed treatment means of total nodule nitrogen 87. 5.19 Original treatment means of total plant nitrogen 88. 5.20 Transformed treatment means of total cotyledon phosphorus 93. 5. 21 Original treatment means of total shoot phosphorus 94. 5.22 Original treatment means of total root phosphorus 95. 5. 23 iJ'rans formed trea triten t means of total rbd~le .~hosphorus 96. 5.24 Original treatment means of total plant phosphorus 97. 5.25 Correlations of tissue nitrogen and acetylene reduction 99. 5.26 Correlations of tissue phosphorus and acetylene reduction 100. 5.27 Original treatment means of cotyledon dry weight per plant (xl0- 3 ) 103. 5.28 Transformed treatment means of shoot dry weight per plant 104. 5.29 Transformed treatment means of root dry weight per plant 105. xiii. Table Page 5.30 Transformed treatment means of tota 1 plant dry weight per plant 106. 5.31 Trans formed treatment means of root-shoot ratio 109. 5.32 Transformed treatment means of secondary stern number per plant 110. 5.33 T-test values for homogeneity of regression slopes between shoot dry weight and root dry weight and between shoot dry weight and nodule dry weight 111. 5.34 Summary of stepwise multiple regression analysis for different treatments 113. ,. LIST OF FIGURES Figure 3.1 Factors affecting the energy supply for dinitrogen fixation in a legume nodule 3.:2 Model of soil phosphates and inorganic phosphorus in different pools 3.3 Root hairs and extramatrical mycelium in the soil 4.1 Experimental layout in the glasshouse 5.1 Mycorrhiza infection of sainfoin measured as score of infected cortex per microscope field of view over time (zero infection obtained for all treatments before day 115) 5.2 Nodule number per sainfoin plant over time (zero data set for all treatments before day 36) 5.3 Nodule dry weight per sainfoin plant over time (zero data set for all treatments before day 36) 5.4 Nodule dry weight per sainfoin nodule over time (zero data set for all treatments before day 36) 5.5 Acetylene reduced per hour per sainfoin plant over time (no assay made before day 77) 5.6 Acetylene reduction per hour per gram nodule dry weight of sainfoin over time (no assay made before day 77) 5.7 Simple regression plots of nodule dry \'Jeight against nodule number per sainfoin plant for treatments C, E, R and B 5.8 Simple regression plots of nodule dry weight per sainfoin plant against nodule dry weight per nodule for treatments Ci E, Rand B 5.9 Total cotyledon nitrogen concentration of sainfoin over time (inadequate tissue for analysis at day 93, while zero data sets after day 104) 5.10 Total shoot nitrogen concentration of sainfoin over time Xi. V. Page 26. 34. 34. 46. 72. 72. 72. 72. 79. 79. 79. 79. 83. 83. xv. Figure Page 5.11 Total root nitrogen concentration of sainfoin over time 83. 5.12 Total nodule nitrogen concentration of sainfoin over time (inadequate tissue for analysis at day 36, 45 and 61, while zero data set before day ~7) 83. 5.13 Total coty ledon phosphorus concentration of sainfoin over time (inadequate tissue for analysis at day 93, while xero data sets after day 104) 91. 5.14 Total shoot phosphorus concentration of sainfoin over time 91. 5.15 Total root phosphorus concentration of sainfoin over time 91. 5.16 Total nodule phosphorus concentration of sainfoin over time (inadequate tis sue for analysis at day 36, 45 and 61, while zero data set before day 27) 91. 5.17 Total plant ni ½rogen concentration of sainfoin over time 92. 5.18 Total plant phosphorus concentration of sainfoin over time 92. 5.19 Cotyledon dry weight per sa infoin plant over time (zero data sets after day 104) 102. 5.20 Shoot dry weight per sainfoin plant over time 102. 5.21 Root dry weight per sainfoin plant over time 102. 5.22 Total plant dry weight per sa infoin plant over time 102 . 5.23 Root-shoot ratio of sainfoin over time 108. 5.24 Secondary stem number per sai nfoin plant over time (zero data set before day 77) 108. 5.25 Regression plots between shoot dry weight and root dry weight of sainfoin (R 1 and 81 are plotted from day 18 to 77i while R2 and 82 are plotted from day 93 to 137) · 108. 5.26 Regression plots between sh oot dry weight and nodule dry weight of sainfoin (C 1 and E1 are p 1 otted from day 36 to 93; C2 and E2 are plotted from day 93 to 137; Rand Bare plotted from day 77 to 137) 108. Figure 5.27 Scatterplots of acetylene reduction against nodule dry weight from day 77 to 137 5.28 Scatterplots of acetylene reduction against shoot dry weight from day 77 to 137. xvi . . Page 112. 112. . I Plate 4.1 4.2 4.3 4.4 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 LIST OF PLATES Incubation and gas sampling in the first step of acetyl ene reduction assay Gas chromatography in the second step of acetylene reduction assay Auto-analyser used in the analysis of total nitrogen and phosphorus Auto-analyser recorders General view of infection with Gigaspora magarita Becker & Hall in corn (x 1 220) (HY, hypha ; AR, arbuscle; VE, ves i cle). Genera l view of in fect i on with MX in sainfoin at day 137 (X 380) (AP , appressorium; MY, myce lium; AR, arbuscle; RH , root hair) Endophyte entry points in sainfoin roots at day 137 (X 1 220) (AP , appressorium ; HY , hyph a) Arbuscles at differ i ng stages of de ve l opment in sainfoin at day 137 (X 1 220) (E , early; I, intermediate ; L, late) A sainfoin root hair infected with Rhizobiwn spp . (X 1 910) (RH , root hair ; IT , infection thread) General view of a region with many infected root hairs (X 1 910) (RH, root hair ; IT, infection thread) Sainfoin nodules of various s izes , shapes and colour Green (left) and red (right) nodules from the uninocul ated (C and E) and rhizobia-inocul ated (Rand B) treatments 5.9 Sainfoin plants at day 93 (C , control ; E, endophyte ; xvi i. Page 56a. 56a. 56b. 56n. 68a. 68a. 68b 68b. 71a. 71a . 71b. 71b. R, rhi zobi a; B, bot h inoculations) 101a. 5.10 Sainfoin plants at day 137 (C, control; E, endophyte; R, rhizobia; B, both inocu l ations) 5.11 Nitrogen defici ency symptoms of sainfoin plan t s in treatment C and Eat day 115 5.12 Secondary stem initiation at the end of rosett e leaf stage at day 77 101a. 101b. 101b. Appendix A B C D E F G H I J K L M LIST OF APPENDICES SI and SI-derived units and other common abbreviations Effects of endophytes on nodulation and dinitrogen fixation Programme summary Endophyte culture Soil - te s t results Yeast-mannitol medium Nutrient solutions Glasshouse temperatures Development of endophyte infection assessment key Gas chromatography Colorimetric automated analysis Pooled ANOVA for endophyte infection Pooled ANOVA for nodulation and dinitrogen fixation variables N Pooled ANOVA for total nitrogen and phosphorus concentrations 0 Pooled ANOVA for plant growth variables xvi ii . Page 131. 133. 135. 137. 139. 141. 142. 145. 146. 148. 151. 153. 154. 159. 169. CHAPTER 1 INTRODUCTION Nitrogen is an essential element for plant growth and repro­ duction as it is required in the synthesis of proteins, enzymes, deoxyribonucleic acids and many intermediate metabolic compounds. It is, therefore, a key constituent of all plant cells. Dry matter of vascular plants contains about 15 OOO µg N g- 1DM (1.5% nitrogen), making it the most abundant soil element in plant tissue (Table 1.1). The supply of nitrogen in current and future agricultural systems is a major determinent of ever-expanding human population. agricultural plants were selected adequate food production to meet the During the green revolution, many and bred for responsiveness to ferti- lizer supply, resulting in the necessity for intensive application of fertilizers particularly nitrogen (Cummings and Gleason, 1971; Engibous, 1975; Jackson et a~ 1975; Cox and Atkins, 1979). Scrimshaw and Taylor (1980) have worked out that the primary fa ctor responsible for increases in crop production between 1961 and 1976 was the increasing use of fertilizers. Between the same period, the world 1 s annual consumption of nitrogen fertilizers rose sharply from 11 to 45 Gkg, a drama tic 309% increase (Fertilizers - Annual Review, 1968; FAO Fertilizer Yearbook , 1980). In 1980, it reached a record of 57 Gkg year- 1 (FAO Fertilizer Yearbook, 1981). With this increasing t rend in nitrogen fertilizer usage, it is perhaps not an exaggeration to reaffirm the statement of Viets (1965) that more crops are deficient of nitrogen than of any other element. The primary supplement of nitrogen to crop plants is from indust­ rial nitrogen fertilizers. However, the existing method of industrial synthesis of nitrogen fertilizers requires a high input of expensive fossil energy. For instance, in the manufacture of ammonia, the pre­ cursor for various types of nitrogen fertilizers, a temperature of 400 to 500°C and a pressure of 15 to 35 MPa must be created to drive the Haber­ Bosch process in a modern plant, with a production capacity of about 900 Mkg day- 1 (Bridger et al, 1979). In this process alone, the natural gas feed and fuel cost contributes 25% of the total manufacturing cost (Finneran and Czuppon, 1979). With the addition of other fuel expend­ iture as in the conversion of ammonia into nitrogen fertilizers, in transport and in application, the final nitrogen fertilizer applied to TABLE L 1 CONCENTRATION OF SIXTEEN ELEMENTS IN COMPLEX PLANTS (AFTER STOUT, 1961) 2. Element Concentration (µg g- 1 OM) From the atmosphere and water, carbon oxygen hydrogen From the soil, nitrogen potassium calcium magnesium phosphorus sulphur chlorine iron manganese boron zinc copper molybdenum TABLE 1. 2 450 OOO 450 OOO 60 OOO 15 OOO 10 OOO 5 OOO 2 OOO 2 OOO 1 OOO 100 100 50 20 20 6 0.1 EFFICIENCY AND CONTRIBUTION FROM VARIOUS DINITROGEN-FIXING SYSTEMS (AFTER BURNS AND HARDY, 1975, WITH AUTERATI ONS IN PARENTHESIS FROM PAUL, 1978) Rate of Dinitrogen-fi xing Land use fixation system (Mha) (kg N2ha- 1 year- 1 ) Legume-Rhizobium legume 250 140 ( 80) Legume-Rhizobium permanent grassland 3 OOO 15 ( 8) Blue-green algae rice 135 30 Free-living and "loose" associations other crops 5 5 Total contribution (Gkg N2year- 1 ) 35 (20) 45 (24) 4 5 3. the field is an expensive item for many farme.rs·. In conjunction with th.is. h.i.gh cost, tracer studies reveal that the use of applied nitrogen fertilizers in the field by crop plants is inefficient. Depending on the type of crop, agricultural practices, fertilizer, climate and soils (Winteringham, 198Q), between 20 to 60% of the total applied nitrogen is absorbed by crop plants (Allison, 1965, 1966; Bartholomew, 1971; Gutschick, 1980; Hauck, 1981). The work of Myers and Paul (1971) showed that wheat (Tri t ioum aestivum L.) grown in a sandy loam and a clay soil, recovered about 25% and 50% of the applied ammonium nitrate respectively in the shoots. In two other studies, a first year maize (Zea mays L.) crop utilized only about 22% of the labelled urea, this being from grain and straw (Arora et al, 1980), while a first year dwarf bean (Phaseolus vulgaris L.) crop removed about 30% of the labelled ammonium sulphate (Cervellini et al, 1980). The unrecovered nitrogen is "lost" through immobilization, leaching, erosion, denitrification and volatilization of which leaching and erosion, if excessive, pose a serious threat to environmental pollution and public health (Mulder et al, 1977; Wild and Cameron, 1980 ). From the foregoing di scussion, it is apparent that the continual heavy reliance on nitrogen fertilizers in the future is becoming a ques­ tionable proposition. The emphasis of current nitrogen and crop research is, therefore, strongly orientated towards improv ing biological dinitrogen fixation (Evans, 1975 ; Hardy et al, 1975; Brill, 1980; Hardy, 1980a , b; Lambourg, 1980; Subba Roa, 1980). Several biological dinitrogen fixing systems are available for in corporation into agricultural production as shown in Table 1.2. The most important and efficient of which, in relative terms, is the legume-Rhizobiwn mutualistic association. The data in Table 1.2 indicate that this type of association fixes an average of between 80 to 88 kg N2 ha- 1year- 1 and, thus, contributes between 20 to 44 Gkg N2 year- 1 to cultivated land under legumes and permanent grassland. However, a fixation as high as 171 kg N2 ha- 1year- 1 has been obtained in the developed New Zealand pastures in which the principal legume compo­ nent is white clover (Trifoliwn r epens L.) (Hoglund et al, 1979). Although the legume-Rhizobiwn association is the most efficient by comparison, it is widely recognised that its dinitrogen-fixing activ­ ity seldom attains the optimal rate. For example, the clovers in the . New Zealand pastures are capabl e of fixing a potential of 215 to 336 kg 4. N2 ha- 1year- 1 (Sears et al,1965; Levy~ 1970). Improvement on the rate of dinitrogen fixation is, therefore, an imperative research endeavour in order to sustain the necessary agricultural production levels. The physical and biologica l factors that directly and indirectly influence the legume-Rhizobiwn relationship have been ident­ ified and comprehensively reviewed by various authors (Lie, 1974; Gibson, 1977; Munns, 1977; Pate, 1977; Parker et al, 1977; Dommergues, 1978; Vincent, 1980; Grandha 11 , 1981). One of these factors is soil phosphorus, an essential element for the growth and nodulation of legumes (van Schreven, 1958; Andrew, 1977; Andrew and Jones, 1978). Many legumes, \-.Jhen infected with arbuscular fungi, show an enhanced phosphate absorption and, subsequently, an associated increase in growth, nodul­ ation and dinitrogen fi xation (Crush, 1974; Daft and El-Giahmi, 1974, 1975,1976; Powell, 1976; Mosse et al, 1976; Mosse, 1977; Abbott and Robson, 1977; Smith and Daft, 1977; Carling et al, 1978; Azcon-G. de Aguilar et al, 1979; Smith et al, 1979). Similar studies on sainfoin (Onobrychis viciifolia Scop.) have yet to be carried out and since it is a legume which is gradually gaining world-wide recognition as a potential forage crop, the purpose of this study is to examine the endophyte­ phosphate interaction, and its effects on the nodulation and dinitrogen fixation in sainfoin. 5 • CHAPTER 2 TERMINOLOGY 2.1 INTRODUCTION Over the years, the interdisciplinary interest in plant science has resulted in some confusing terminology in the literature on soil-plant studies. This chapter, therefore, proposes to define the meanings of commonly used terms so as to ensure a greater precision of meaning. For convenience, each term is discussed in alphabetical order under two sections . Section 2.2 is concerned with the nomenclature used in dinitrogen fixation, while section 2.3 deals with the terms used in mycorrhizae. Similarly, to reduce ambiguity in units of expression (Incoll et al, 1981), it is proposed to use the Systeme International d'Unites (SI) endorsed by the International Organisation for Standardization in 1960. A list of the units (and other common abbreviations) used, with their symbols, names and definitions are presented in Appendix A. 2.21 2.2 TERMS USED IN DINITROGEN FIXATION Infection, Invasion and Infectivity (Fahraeus and Ljunggren, 1968) The term in6eetion is used to describe the establishment of a parasitic and not a pathogenic relationship within the host. The term invcv.,ion is sometimes used as a synonym of infection, irrespective of whether the process is pathogenic or non-pathogenic. Since virulence or pathogenicity expresses the capacity of an organism to produce a disease, it is substituted by another term, in6ec.,ti,vdy. 2.22 Mutualism and Symbiosis (Starr, 1975) Symbio-0-l6 is a general term which means the living together of two dissimilar organisms and, thus, includes all types of inter­ actions. However, mU,W.aiMm refers specifically to the association in which the two organisms involved, benefit from each other. Both terms as defined are also used to describe mycorrhizae. 2.23 Rhizobia, Rhizobium and Rhizobium s Jordan and Allen, 1974; Skerman et al, 1980) Rlu.zobiwn Frank 1889 is the generic name for a genus of bacteria that form a mutualism with legumes and fix dinitrogen. 6. Rlu.zobia is sometimes used as a common name when referring to this group of bacteria, but other terms such as "Rhizobia", and 11 rhi zobi al II are regarded as unsatisfactory. · 'Rhizobiwn spp. refers to certain unclass- ified dinitrogen-fixing bacteria. On the basis of cross-inoculation studies, these bacteria can be described either as "promiscuous", that is, capable of forming mutualism with legumes of a wide range of genera from different tribes, or as specific. 2.3 TERMS USED IN MYCORRHIZAE 2.31 Endophyte (Braun-Blanquet, 1928) The term endophyte (plural endophyte.1:i) means a plant that invades and lives in its host. It is commonly used to refer to fungi that form mycorrhizae (see Kelley, 1950). Fungi are not classified as plants of course, but the use of endophyte to define mycorrhiza­ forming fungi is continued here. However, it is not employed when referring to bacteria such as the Rhizobium spp. that reside in legumes. 2.32 Mycorrhiza Types (Lewis, 1975) The term myeoJzJLhiza (plural myeoMhizae in American usage and mycorrhizas in English usage) (Nicholson, 1967) means fungus-root and is strictly used to describe the mutualism between a fungus and plant root. Terms such as "mycorrhi za 1 fungus 11 ( fungus-root fungus) and 11 mycorrhi za 1 root" (fungus-root root) are, therefore, best avoided. If the fungus exists only in the intercellular spaces in the root and forms a fungal sheath around the root, the association is known as an edomyeoMhiza. However, if the fungus is both intercellular and intra­ cellular, and does not fonn a sheath, the association is called aJtbu..6c..ulaJt, o~ehidaeeou..6, or vueoid myeoMhiza depending on the host type. Orchid­ aceous and ericoid mycorrhiza occur specifically in the family Orchidaceae and the order Ericales respectively, while arbuscular mycorrhiza occurs generally in many families. A fifth group, named as aJtbuxoid myeoMhiza, is also specific to the order Ericales, but the fungus involved is intercellular and intracellular, as well as sheath-producing. 7. CHAPTER 3 REVIEW OF LITERATURE 3.1 INTRODUCTION The Fabaceae (or Leguminosae), comprising 650 genera and 18 OOO species, is the third largest family of flowering plants after the Asteraceae and Orchidaceae (Polhill et al, 1981). Within the Fabaceae, three subfamilies -- the Caesalpinioideae, the Mimosoideae and the Papilionoideae -- are generally recogni sed . In the Caesalpinioideae and the Mimosoideae, the species included are mainly trees and shrubs of the tropics and subtropics (Clapham et ai, 1952; Hutchinson, 1964). In contrast, the Papilionoideae consists mostly of annual and perennial herbs which are more widely distributed, extending from the tropical to the temperate regions (Hutchinson, 1969). Almost all the important agricultural legumes are also included in this subfamily (Okigbo, 1978). Economically, legumes are considered to be the second most important group of plants after grasses (Sinha,1977; National Academy of Sciences, 1979). The findings of Harlan (1976 ) suggest that legumes have been cultivated since the beginning of ancient civilization and agriculture . Throughout history, therefore, man appeared to have highly regarded the value of legumes . Today , legumes maintain their key role in modern agriculture, covering a wide range of purposes including agricultural uses (Skerman, 1977; Dobereiner and Campelo, 1977; Mulder et al, 1977), forestry (National Academy of Sciences, 1979), horticulture (Heywood, 1971; National Academy of Sciences, 1979), and others (Heywood, 1971; Willis, 1973). Although legumes are cultivated for various reasons, their primary economic importance is attributed to the ability of their root nodules to convert dinitrogen into ammonia which is readily utilized by the plants. This dinitrogen-fixing ability is the result of a successful mutualistic association between legume roots and a genus of soil bacteria, Rhizobiwn. Not all legumes exhibit this mutualistic relationship, but the majority of species in the Mimosoideae and the Papilionoideae are known to form root nodules (Allen and Allen, 1961), including all the agricultural legumes except for certain lines of red clover (Trifoliwn pratense L.) (Nutman, 1949) and soybean (Glycine max (L.) Merr.) (Williams and Lynch, 1954; Clark, 1957). 8. In recent years, the potential dinitrogen-fixing capacity of legumes is recognised to be of immense economic and ecological value. This is reflected in the release of several exclusive monographs by the National Academy of Sciences (1975, 1977, 1979), Skerman (1977), Duke (1980), and Summerfield and Bunting (1980) on the potential and the utilization of both food and forage legumes. In addition to the major agricultural legumes, the generated research interest is extended also to many under-utilized species with desirable characteristics. In this chapter, a review is presented of the botany and agronomy of one such under-utilized legume called sainfoin (Onobrychis viciifolia Scop.), with special emphasis on the supply of phosphates through mycorrhiza formation, and its effects on nodulation and dinitrogen fixation. In the section on dinitrogen fixation (section 3.3) and mycorrhizae (section 3.4), where information is scanty on sainfoin, an endeavour is made to broaden the review to include other legume species and, whenever necessary, other non-legume species. 3.2 SAINFOIN AS A FORAGE CROP 3.21 General Distribution and History 3.211 Distribution. The genus Onobrychis Mill., belonging to the tribe Hedysareae, is comprised of some 130 species (Clapham e t al, 1952; Polhill, 1981) herbaceous, or sometimes shrubby and spiny perennials (Bailey, 1944) and annuals (Ball, 1968). The origin of Onobrychis spp. is in the Near Eastern Centre which includes the regions of Turkey, Iraq, Iran, and the Caucasus and the east of the Caspian Sea, belonging to the Soviet Union (Vavilov, 1951). However, their present distribution is wide, covering the Mediterranean, central Europe, western Asia, and central Asia from the Caspian Sea to Lake Baykal (Kernick, 1978). A number of species such as o. arenaria (Kit.) D.C., 0. transcaucasica (Shain, 1959), 0. caput-galli (L.) Lam., 0. crista­ galli (L.) Lam., (Smith, 1976), O. vaginalis C.A. Mey., 0. tanaitica Sprengel, O. inermis Stev. and O. altissima Grossh. (Duke, 1980) have been cultivated in their native areas. However, the most important and widely grown species is o. viciifolia Scop. (sainfoin) which can be found commonly today in the Mediterranean nations, west Asian countries, 9. the United States and Canada (Kernick, 1978). 3.212 History. The writings of Vavilov (1951) suggest that sainfoin is indigenous to the mountains of the Near Eastern Centre, especially in the Caucasus, existing in many forms as land races and ecotypes. Sainfoin cultivation probably commenced in the tenth century in Azerbadjan, Armenia and Georgia, all of which are situated just south of the Caucasus (Shain, 1959). In the fifteenth century, its cultivation spread to France from which it was extended to Switzer­ land, Germany and England two centuries later (Stebler and Schroter, 1889). In the eighteenth century, some seeds were also carried to Italy (Stebler and Schroter, 1889) and Wales (Rees, 1928 ), and in the early nineteenth century to the United States (Clark and Malte, 1913). From the time of its introduction to the nineteenth century, sainfoin was highly esteemed as a forage or pasture plant in both conti­ nental Europe and England (Stebler and Schroter, 1889; . Whyte et al, 1953). This popularity was because of its successful adaptation to the then barren, dry calcareous soils, where it grew well for as long as 3 to 20 years without manuring (Stebler and Schroter, 1889; Davies, 1960). Owing to such extraordinary qualities, the French named it as sainfoin (formerly written incorrectly as Saint Foin), meaning wholesome hay (Clark and Malte, 1913). However, after the 1920's, the popularity of sainfoin steadily declined, with sainfoin acreages being considerably reduced in the continent (Smith, 1976) as well as in England (Hutchinson, 1965; Bland, 1971). This decline was due to the introduction of improved lucerne and clover cultivars that are higher-yielding and more adaptable to a wider range of soils and climate (Rogers, 1975; Smith, 1976). and due to the changing agricultural ststem that requires intensive application of nitrogen fertilizers instead of lower-yielding legumes (Hutchinson, 1965; Aldrich, 1974). In Europe today, therefore, the chief sainfoin- growing areas are restricted to France, Italy, the Soviet Union and Romania (Fortune and Withers, 1980). However, as a result of the escalating cost of nitrogen fertilizers in the last decade, interest in sainfoin has been renewed in recent years (Sheehy and Popple, 1981). In North America, sainfoin cultivation and evaluation were carried out on many farms and experimental stations in the nineteenth century, but it never attained agricultural importance (Piper, 1924; Jensen and Sharpe, 1968). In the 1960s, however, there was renewed 10. interest in this species in both the United States (Eslick, 1968) and Canada (Hanna, 1968). This is because of its potential in dry calcar­ eous soils (Dubbs, 1968; Roath, 1968; Ryerson and Taylor, 1968), its high palatability (Hanna and Smoliak, 1968; Jensen e t al, 1968), its non-bloating characteristics (Krall, 1968; Cooper e t al, 1968a), its resistance to alfalfa (lucerne) weevil (Hypera pos t i aa Gyll .) (Carleton et al, 1968b; Wallace, 1968) and most of all, the availability of genotypes well-adapted to the adverse winters (Eslick, 1965; Carleton, 1968; Hanna, 1968; Cooper et al, 1968b). With the release of improved cultivars such as Eski in 1964 (Eslick e t al, 1967), Melrose in 1969 (Hanna et a l, 1970), Remont in 1971 (Carleton and Delaney, 1972) and Nova in 1980 (Hanna, 1980), sainfoin cultivation is becoming profitable and important in North America. 3.22 Botanical Description 3.221 Morphology. Sainfoin is a deep-rooted perennial, with a soil-level crown which produces a number of hollow stems that grow up to about 80 cm high. The leaves are borne on long petioles and are imparipinnately compound, with 7 to 35 oblong-obvate and entire leaflets (Thomson, 1951a; Smith, 1972). The inflorescences are conspicuous, 5 to 13 cm long racemes (Piper, 1924), terminating on long, slender peduncles (Percival, 1921). Each raceme consists of 5 to 80 rose-pink florets (Carleton and Weisner, 1968). The pods formed are single­ seeded, indehiscent, slightly compressed, with reticulate ridges (Stebler and Schroter, 1889). Hulled seeds are kidney-shaped, yellowish­ green to brown or black, and weigh about 15g per 1 OOO seeds (Percival, 1921; Thomson, 1951b). The root system is extensive with a strong woody tap root which has several large branches and numerous fine laterals (Masaudilov, 1958). The tap root usually penetrates 1 to 2 m down the soil profile, but occasionally reaches to 10 m deep (Andreev, 1963). The nodules formed are white-orange and are found mainly on the fine lateral roots (Smith, 1972). They are large and characteristically branched, but when under-developed may be merely prolate, ovoid, elongate, or lobed, with apical meristems and, therefore, are classified as astragaloid nodules (Corby, 1981). 3.222 Growth ·pattern. Broadly, two types of sainfoin are 11. recognised (Rees, 1928; Thomson, 1938). One is a single-cut group, with greater stem production and l ate flowering, and with a relatively slow regrowth (e.g. variety aommunis Ahlef., cultivar Eski, Melrose and Nova) . The other is a double-cut (or multi-cut) group, with lower stem production and early flowering, and with a fairly quick regrowth (e.g. variety bifera Hort., cultivar Remont, Fakir and Othello). Both types of sainfoin exhibit three similar phases of growth and development in their seeding year. The first phase is germination which is described as phaneroepigeal (Duke and Polhill, 1981), with the thick, fleshy cotyledons turning green and opening above the soil surf­ ace (Thomson~ 1938). Delayed germination of unhulled sainfoin seeds has been indicated in several early te s ts (Rees, 1931, 1933; Thomson, 1938). In 1952, Thomson explained that it was probab ly due to mechan­ ical resistance of seed pod to radicle emergence and not due to a water­ soluble inhibitor in the seed pod. However, this exp l anation was refuted by Cavazza (1952), Carleton et al, (1968a) , and Smith (1979), who all found that at least one type of water-soluble inhibitor was . present. The second phase is associated with the formation of a rosette of about 6 leaves close to the ground (Percival, 1921; Thomson, 1938). The first leaf may be unifoliate, bifoliate, trifoliate, or even penta­ foliate, with the unifoliate form being predominant (Thomson, 1938) . Cooper (1974) has noted differences in leaf area among the various types of first leaf, but no associated long-term effect upon photosyn­ thesis and growth. The cotyledon may remain green to yellowish-green throughout this phase of growth. However, the major photosynthetic contribution from the cotyledons occurs during the unfolding and expansion of the first l eaf, with a 100% contribution when the cotyle­ dons are just unfolded, 54% when the first leaf is unfolded, 39% when the first leaf is fully expanded, and only 18% when the second leaf is unfolded (Cooper and Fransen, 1974). The same authors showed also that cotyledon food reserves do not contribute significantly to sainfoin growth during this phase of development as compared to germination. The positive correlation obtained, therefore, between seed size (in the range of 13.8 to 19.5 mg seed - 1 ) and seedling performance (Carleton and Cooper, 1972) is due possibly to the greater contribution of cotyle­ don food reserves from larger seeds during germination, and the greater photosynthesis from larger cotyledon area during the early second phase 12. of growth. In addition to seed size, this phase of growth depends also on the inherent vigour of the sainfoin seedling (Cooper, 1977). The third phase of growth begins soon after the formation of the sixth leaf, that is, when lateral buds start to develop (Thomson, 1938). These buds produce numerous leaves, with their internodes normally short initially, but which elongate later and climax in the development of inflorescences. According to Rees (1931), the amount of flowering in the seeding year is variable and depends on the type of sainfoin and time of sowing. Double-but sainfoin commonly pro­ duces elongated stems and flowers two to five times after each cutting in its first growth season. However, single-cut sainfoin, when sown in spring, does not produce elongated stems and inflorences until the second season (Thomson, 1951a; Badoux, 1965) when vernalization and photoperiodic requirements have been satisfied (Sheely, 1977). 3.23 Some Agricultural Characteristics 3.231 Adaptive characteristics. Sainfoin is well-known to perform favourably in marginal agricultural lands having a dry climate and poor, calcareous soils. Besides being tolerant to drought, lime­ stone and low fertility, it is reported also to exhibit tolerance to frost, steep country, salt, high pH and grazing (Duke, 1980). Space here permits only the discussion of its adaptation to drought and high pH including limestone. Although areas with an annual precipitation of 400 to 800 mm are recommended for sainfoin growing, it also shows tolerance to an annual rainfall of up to 1 200 mm (Duke, 1980), provided that the soil is well-drained (Mansfield, 1945). Two species, Onobryahis eahidna Lips. and o. aornita (L.) Desv. from central Asia, have been described as xerophytes (Kul~tiasov, 1961, 1962), but not sainfoin per se. However, it might well be classified as a xerophyte because of its semi-arid habitat, deep extensive root system, and typically small, thick and hairy leaflets. A cross-section of each leaflet reveals the characteristic thick article, thickened epidennal cell walls, and water-bearing cell layer under the epidermis (Dalenvoa, 1961). Such xeromorphic features intimate that sainfoin seems to possess both avoidance and tolerance mechanisms like many arid and semi-arid plants (Chabot and Bunce, 1979), but these mechanisms are yet to be elucidated. Avoidance is possibly· more important than tolerance because of its ability to maintain a relatively high and constant leaf water potential at greater than -400 kPa throughout its growth period (Sheehy et aZ, 1978; Sheehy and Popple, 1981). 13. Sainfoin is a calcicole, but it is tolerant to a relatively wide pH range of 5.3 to 8.2 (Duke, 1980). Information on pH effects and the physiology of this species is limited . Crop failure at pH below 5.3 is due probably to the high concentration of injurious hydrogen ions (Moore, 1974), toxic aluminium ions (Ronson, 1958, 1965) and toxic manganese ions (Murray, 1968) , low availability of calcium and magnesium ions, phosphates and molybdates (Donahue et aZ, 1977), or a combination of these factors. At pH greater than 8.2, other ion s such as those of iron, manganese, zinc, copper and boron exist in low concentrations in the soil solution and, thus, the related deficiencies may develop (Donahue et aZ, 1977). 3.232 Herbage quality. The nutritional qualities of sain- fain herbage include its non-bloating property, and relatively high palatability , digestibility, and protein and mineral content. The non-bloating characteristic of sainfoin forage when grazed by or fed to ruminants is an exceedingly important quality. Bloat is caused by the generation of a stable foam which traps the fermented gaseous products in t~e reticulo-rumen, leading to a rise in pressure in the rumen (Reid and Johns, 1957; Reid, 1960). Although a number of factors contribute to formation of this foam (Clarke and Reid, 1974), soluble plant protein appears to be a major foaming agent (Jones et aZ, 1970). The anti-bloating character­ istics of sainfoin herbage are attributed to the presence of certain compounds in the leaves (Cooper et aZ, 1966), later identified as condensed tannins ( Kendall, 1966; Jones et aZ, 1973) which inhibit the production of stable foam in the rumen (Reid et al, 1974). Gutek et aZ (1974) found large quantities of condensed tannins in sainfoin in all seasons and at all stages of growth and thus estab­ lished the persistence of this non-bloating property of sainfoin. In addition to the prevention of foam generation, condensed tannins in sainfoin can slow down deamination of proteins in the rumen, thereby increasi ng nitrogen assimilation (Reid et aZ, 1974). Plant tannins have been shown in many experiments to be toxic to 14. animals, but no evidence has been collected for · sainfoin tannins (Krall, 1968) or other forage tannins (McLeod, 1974). There is evidence that tannins are secondary metabolic compounds that are 11 distasteful 11 and do inhibit the action of proteases (Harborne, 1977). Such properties have been found to deter herbivores. For instance, the presence of condensed tannins, possibly procyanidin, {Sakar et aZ, 1976), in sericea (Lespedeza sericea Benth.) was suggested as the cause of its low palatability (Donelly and Anthony, 1969) and poor digestibility (Cope and Burns, 1971). However, such problems are not found in sainfoin due probably to the absence of procyanidin in the flavolan composition (Sarkar et aZ, 1976). The intake of sainfoin herbage by sheep (Raymond, 1966; Osbourn et aZ, 1966; Hanna and Smoliak, 1968) and cattle (Jensen et aZ, 1968; Ulyatt et aZ, 1977) respecitvely was higher or at least similar to that of lucerne (Medicago sativa L.), a recognised nutritious forage legume, indicating that the palatability of sainfoin herbage is better or equivalent to that of lucerne herbage. In another study involving Hereford cattle, the digestible dry matter percentage of sainfoin was as high as that of lucerne (Jensen et aZ, 1968). In addition, greater liveweight gains were obtained with steers grazed on sainfoin than on a mixture of v1hite clover (TrifoZiwn repens L.) and tall fescue (Festuca arundinacea Schreb.) or a mixture of bromegrass (Bromus in­ ermis Leyss.) and cocksfoot (pactyZis gZomerata L.) (Krall, 1968). 2.233 Herbage yield. Dry matter yield of sainfoin is highly variable, depending on the climate, soils, management, age of stand and cultivar. In its indigenous mountains of the Caucasus in Azer­ badjan, the Soviet Union, Mamedov and Aliev (1972) recorded an annual yield of 6 OOO kg OM ha- 1 on level plots, 4 050 kg OM ha- 1 on northern slopes, and 3 600 kg OM ha- 1 on southern slopes. On a sloping eroded land in Romania, sainfoin yielded 5 400 kg OM ha- 1 year- 1 (Popa, 1976). In Italy, a dry matter yield of 9 380 kg ha- 1 year- 1 was reported (Covarelli, 1975). In England, Cotswold Common yielded 11 OOO kg OM ha- 1 year -i in three cuttings in both the first and second harvest year (Smith, 1972). In a trial using monocultures or mixtures of Eski and Remont in the United States, Eski and Remont respectively contrib­ uted 9 450 and 9 590 kg OM ha- 1 year- 1 in two cuts in the second year (Cooper, 1972). In Canada, however, Nova, Melrose, Remont, and Eski respectively produced up to 14 600, 13 600, 13 500, and 13 500 kg OM 15. ha- 1 year- 1 in 3 cuts in the second s·eason (Hanna, 1980). When comparing with two other important forage legumes, lucerne and red clover, the herbage yield of sainfoin is generally lower (Cooper and Roath, 1965; Hanna and Smoliak, 1968; Smith, 1972; Rogers, 1975). However, in conditions suitable to sainfoin, it can outyield lucerne or red clover (Roath, 1968; Carleton e t al, 1968; Smoliak and Hanna, 1975). Since the improved cultivars such as Eski, Melrose, Remont and Nova have shown potential in dry matter yeild, the selection and breeding for suitable high-yielding cultivars is, therefore, a major priority in many countries including New Zealand. 3.24 Suitability in New Zealand Conditions 3.241 Early introduction and evaluation. In New Zealand, a series of studies on the causes and control of bloat in cattle that graze on clovers and lucerne (Johns, 1954; Reid and Johns, 1957; Mangan and Johns, 1957; Reid, 1958; Mangan et al , 1959; Reid e t al , 1961) has generated considerable interest in non-bloating legume species including sainfoin (Jones et al, 1970; Jones and Lyttleton, 1971; Jones et al , 1973; Ross and Jones, 1974; Wright and Reid, 1974). Introduced into New Zealand in the early 1970s, it was evalu- ated for its non-bloating property and the associated nutritional qualities (Special Correspondent, 1974). Excellent results were obtained with both sheep and cattle and, thus, it was recommended as a potential nutritive forage legume in New Zealand (Reid et al, 1974; Derrick, 1977; Ulyatt e t al, 1977). In the following years, the agronomic characteristics of sainfoin have been studied at the Ministry of Agriculture and Fisheries in Rotorua, and at the Grassland Division of the Department of Scien­ tific and Industrial Research (DSIR), Palmerston North, involving the Agronomy Department at Massey University. Currently in DSIR, a breeding programme, aiming to select a cultivar better suited to New Zealand is also underway ( W. Rumba 11, personal communication). 3.242 Diversity of genetic potential. The indigenous habit- ats of sainfoin are related to the semi-arid and sub-humid Mediterranean climatic zones (between 30 to 40° latitude), with an average rainfall of betwe01 400 and 600 mm, and 600 and 800 mm respectively ( Le Houeroa, 1977). The winter is wet, particularly in December and January, and 16. cool (-5 to 15°C), while the summer is dry and hot ( 30 to 40°C). In these climatic regions, sainfoin is known to exist in a range of habitats from the lowland to the substeppe zone, up to an elevation of about 2 OOO m (Kernick, 1978). It is dominant in calcareous pasture lands (Le Houeroa, 1977) and in marls (Kernick, 1978). Since its spread into the Soviet Union and Europe, sainfoin, like many other legumes, displays a wider adaptation to diversity in environmental conditions (Adams and Pipoly III, 1980). A number of different ecotypes of sainfoin have been collected from many geo­ graphical locations especially from its indigenous habitats (Llovet, 1963-1965; Sinskaya, 1958; Hanna, 1968, 1980). This wide biological variation suggests that if is possible to select and fit the specific requirements and desired performance of certain ecotypes to the specific resources of the New Zealand environment. 3.243 Sainfoin and the New Zealand environment. Much of the cultivated land in the North Island and east of the South Island lies between 37 and 45° latitudes, and experiences a Mediterranean-like climate, characterised by a wet and cool winter, and a dry and warm summer (Mitchell, 1963). The mean annual rainfall in these regions is generally between 380 to 1 520 mm with maximum precipitation occurring in the winter (New Zealand Meteorological Service, 1959). The average annual maximum and minimum temperature is -5 and 30°C respectively. These regions include areas in Gisborne, Hawke's Bay, Central Plateau, Manawatu, Wairarapa, Canterbury and Central Otago. The period which limits pasture production most in the above­ mentioned regions is during the summer and autumn, when a serious water deficit is compounded by the shallow, less vigorous root systems of pasture species and high surface soil temperatures (Mitchell, 1963). As much as 40% reduction in potential pasture production between October and April in Palmerston North was estimated (Brougham, 1966). Consequently, certain perennial Mediterranean plant species, having the characteristics of a vigorous, deep root system, low growing meristems, greater stem production and regrowth capacity, smaller leaves, tall vegetative structure of 1 to 2 m high, and good digest­ ability and nutritional value were proposed as desirable forage plants for these regions (Mitchell? 1960~1963,1966). Some cultivars and ecotypes of sainfoin seem to satisfy most of these requirements. Considering this infonnation~ sainfoin cultivation may be profitable in these regions especially the well-drained soils, but liming will be necessary also in many areas. At present, there is limited quantitative data available for critical assessment of the suitability of sainfoin in these regions. The only published research so far was conducted by Percival and McQueen (1980), between 1976 and 1980, on the pumice soils in the Central Plateau. The report reveals that in the overall 5 years, 17. the annual dry matter yield of the cultivar Melrose was 8 500 kg ha- 1 , similar to the dry matter yield of pasture, but much lower than the 12 700 kg ha- 1 of lucerne. Future research endeavour should be directed to investigating more cultivars and ecotypes at various places in both North and South Island, and the nature of their growth responses. At this moment, therefore, sainfoin appears to have a specific role in the New Zealand agriculture as a specialised forage for certain drier and well-drained areas due mainly to its drought resistance, its nutritive merits, and its resistance to pests that seriously affect lucerne such as the blue-green lucerne aphid (Acyrthosiphon kondoi Shinji), pea aphid (Acyrthosiphon piswn (Harris)), and spotted alfalfa aphid (Therioaphis trifolii fm maculata) (Lance, 1980). The spotted alfalfa aphid is recently reported to have arrived in New Zealand. Scott (1979) calculated that sainfoin growing for bloat control is less economical than lucerne growing plus the existing bloat control methods, but this financial analysis must be reserved for a later reconsideration, together with further yield information and other findings, of the economics of sainfoin cultivation. 3.3 DINITROGEN-FIXING SYSTEM OF LEGUMES 3.31 Development of the Legume-Rhizobiwn System A number of authors, including the more recent ones (Dart, 1975, 1977; Pate, 1977a; aergersen, 1978; Schmidt, 1978,1979; Dazzo, 1980; Vincent, 1980; Bauer, 1981), have reviewed comprehensively the sequence of interactive events from recognition to dinitrogen fixation between legumes and their respective Rhizobiwn spp. In these reviews, the information is derived from work done mainly on the major agricultural 18. legumes such as pea, soybean, dwarf bean, lucerne, clovers and lupins, but similar observations are expected to occur in other legumes. Here, a short discussion on the three generally recognised phases of inter­ action is presented in relation, as much as possible, to sainfoin. 3.311 Recognition. In this first phase of interaction, the appropriate strain or species of bacterium, specifically (Rovira, 1961; Robinson, 1967; Munns, 1968; van Egeraat, 1975) or non­ specifically (Peters and Alexander, 1966) stimulated by root exudates, attach to the surface of root hairs perpendicularly (Sahlman and Fahraeus, 1963; Marshall et ai, 1975; Dazzo and Hubbell, 1975). Using pea seedlings, van Egeraat (1978) detected a large quantity of a root-tip exudate, Y-L-glutamyl-D-alanine, \vhich inhibited the growth of Rhizobiwn legwninosanwn strain PRE before the lateral roots emerged. However, when the lateral roots were emerging, they exudated also substantial amounts of homose r ine which selectively stimulated rhizobia growth, while the exudation of another compound called 2-alanyl-isoxazolin- 5-one appeared to diminish the inhibiting action of y-L-glutamyl-D- alanine. The mechanism or mechanisms of recognition is still under much speculation although lately, certain legume proteins, called lectins (carbohydrate-binding glycoproteins) (H amblin and Kent, 1973; Boh]ool and Schmidt, 1974; Dazzo and Hubbell, 1975), are postulated to bind with the polysaccharide produced by the bateria (Wolpert and Albersheim, 1976; Dazzo and Brill, 1977; Calvert et al , 1978; Zevenhuizen et al, 1980). Firm adherence to the root hair surface at later stages of recognition is provided by the production of extracel­ lu,slar cellulose microfibrils (Napoli et a l, 1975). The first visible effect of recognition is root hair deform­ ation, characterized by swelling, curling and branching of the hairs often leading to the formation of shepherd's crooks (Fahraeus and Ljuggren, 1968). Root hair deformation is caused by a partially dialyzable, heat-stable, undefined substance (Li and Hubbell, 1969; Hubbell, 1970; Yao and Vincent, 1969,1976), containing a nucleic acid and a protein or polysaccharide (Solheim and Raa, 1973), produced by the bacteria. No study has been done on sainfoin in the recognition phase at the cellular level but sainfoin rhizobia have been found to secrete polysaccharides (Nalbandyan and Sayadyan, 1977), while a lectin has been isolated from sainfoin seeds and young roots (Hapner and Robbins, 19. 1979). Early inoculation studies suggested that sainfoin bacteria were strain-specific (Allen and Allen, 1981). However, in more recent experiments, bacterial isolates from sweetvetch (Hedyra:Pwn spp.) and crownvetch (Coronilla spp.) can also infect and nodulate sainfoin (Burton and Curley, 1968; van Schreven, 1972). 3.312 Infection. Infection begins when infection threads are initiated in deformed root hairs by way of invagination of the primary walls (Dart, 1977). Three mechanisms of threat initiation have been proposed. The first hypothesis suggests that thread initi- ation may be caused by inward cell wall synthesis induced by impediments to normal wall growth in curled root hairs (Nutman, 1956). Accordinq to the second hypothesis, bacterial penetration is probably enhanced by cell wall softening, induced by host polygalacturonase (Ljuggren and Fahraeus, 1961). The third hypothesis deduces that thread initiation may be due to direct infection by small, multi-flagellated coccoid rhizobia, called swarmers, through gaps in the cellulose microfibrils of the root hair cell wall (Dart and Mercer, 1964). The infection thread, once initiated, progresses towards the inner root cortex by a coordinated bacterial division and cell wall deposition along the thread (Dart, 1974,1977; Sprent, 1979). While passing through the cortex, it also branches and ramifies. The nuclei of cortical cells enlarge and cells adj acent to the threads may divide. Cortical cell division is probably stimulated by indole-acetic acid (IAA) and cytokinin, secreted by the bacteria enclosed in the threads (Libbenga e t al, 1973; Libbenga and Torrey, 1973; Liggenga and Bogers, 1974). Dangeard (1926) has observed some infection threads in sainfoin nodules, while Hume (1981) has related the greater number of nodules to the abundance of root hairs. These results indicate that, as in most legumes (Dart, 1974,1977), infection by rhizobia in sainfoin is probably via root hairs. 3.313 Nodulation. After the infection thread passes through the outer six layers of cortical cells, the inner cortical cells show increased RNA contents and subsequently divide to form a nodule primor­ dium which later become infected by the advancing infection thread (Sprent, 1979). This young primordium then loses its meristematic activity, but the adjacent cells become meristematic and form the nodule meristem which divides repeatedly and emerges from the root (Libbenga and Bogers, 1974). Cell division is probably stimulated by the cyto- 20. kinins in the meristems (Sy~no et aZ,· 1976). Meanwhile, back in the non-meristematic zone, bacteria are released from the infection thread tips as membrane droplets (Kijne, 1975; Newcomb, 1975; Bassett et aZ, 1977), later freed and dispersed into the host cells (Robertson et aZ, 1978). They multiply and enlarge progressively, and differentiate into dinitrogen-fixing bacteroids, enclosed in membrane envelopes of the host origin (Bergersen, 1974; Bassett et aZ, 1977). Nodulation is completed when leghaemoglobin, vascular tissues and nodule cortex are developed (Appleby, 1974; Bergersen, 1974,1978). With an effective inoculum, sainfoin nodule swellings can be detected between 14 to 30 days after inoculation (Dangeard, 1926; Karpov, 1957; Hume, 1981). At this stage, cell divisions and nodule primordia had occurred (Dangeard, 1926). Within 3 to 4 days, the swellings mature into nodules (Hume, 1981), containing bacteroids that are spherical, 1 to 5 µm in diameter, and which sometimes occur in pairs (Dangeard, 1926). Koter (1965) reported that a small amount of combined nitrogen stimulated nodulation and dinitrogen fi xation, but a large quantity decreased nodule number in sainfoin. Recently, Hume (1981) showed that in the presence of 210 µg N cm - 3 solution as sodium nitrate, the number and dry weight of nodules, as \-Jell as leghaemoglobin production were significantly suppressed. Other nutritional disorders such as massive accumulations of calcium carbonate in the roots (Ross and Delaney, 1977) may also depress nodulation and dinitrogen fi xation. In general, however, sainfoin (Burton and Curley, 1968) and other Onobrychis spp. (Hely and Ofer, 1972) are reasonably nodulated in the presence of an effective inoculum, with the nodules formed being larger than those in lucerne (Karpov, 1957). 3.32 Dinitrogen Fixation 3.321 Physiology and biochemistry. The present concept of the physiology and biochemistry of dinitrogen fixation in legume nodules is a synthesis of research results mainly from soybean, lupins, pea and seradella (Orni thopus sativus Brat.), but applicable to other legume species (Bergersen, 1980a). Dintrogen fixation in legume nodules is under rigid genetic control (Whiting and Dilworth, 1974; Page, 1978) and catalysed by the nitrogenase enzyme system (Kennedy e t aZ, 1966; Bergersen and Turner, 1967; Koch et aZ, 1967 in the bacteroids. The nitrogenase enzyme 21. system, consisting of two protein components, a heavier molybdenum-iron protein called component I and a lighter iron protein called component II (Brill, 1977; Eady et aZ, 1980), catalyses a series of complex reductive reactions in which dinitrogen is ultimately reduced to ammonium ions (Evans and Russell, 1971; Eady and Postgate, 1974; Stiefel, 1977; Chatt 1 1980) and protons are reduced to hydrogen gas (Bulen et aZ, 1965; Schubert and Evans, 1976,1977; Evans et aZ, 1980). Energy, required in large quantity in the fixation process, is derived from the oxidation of photosynthetic products, whose forms are yet to be established (Bergersen, 1978,1980a). The oxidation of these products generates reducing power which is used for the production of ATP in oxidative pathways (Appleby et aZ, 1975) and for the gener­ ation of reductants for the nitrogenase system (Evans and Philips, 1975). In the bacteroids of certain rhizobia strains, another enzyme called hydrogenase is present and oxidizes the generated hydrogen gas to protons and hence, recovering these reductants and possibly synthe­ sizing ATP as well (Dixon, 1972; Schubert and Evans, 1976,1977; Schubert et aZ, 1977; Schubert and Ryle, 1980). Oxygen, also required in the fixation process (Bergersen, 1974,1977), is transported to active sites through facilitated-diffusion by leghaemoglobin, thereby preventing the inactivation of the nitrogenase system (Bergersen, 1980b). The ammonium ions reduced are transported to and assimilated in the nodule host cells (Bergersen, 1980a). The assimilatory pathways and, thus, assimilatory products vary among species. In general, the final assimilatory products for xylem transport later in temperate legumes are primarily amides (Pate, 1977b, 1980), while in tropical legumes these are mainly ureides (Pate et aZ, 1980). Dinitrogen fixation in sainfoin, as in all other legumes, follows an exponential increase with time during the vegetative growth phase (Hume, 1981). There is as yet no published report on its rate of fixation during the reproductive growth phase, but dinitrogen fix­ ation is expected to decline rapidly from anthesis. At similar growth stages, the fixation in sainfoin is generally lower compared with that in the more improved legumes such as pea (Phillips et aZ, 1976), sub­ terranean clover (Eskart and Raguse, 1980), soybean (Nelson and Weaver, 1980; Cassman et aZ, 1980), lucerne (Sheehy et aZ, 1980; Craig et aZ, 1981; Hersman et aZ, 1981), red clover (Craig et aZ, 1981), and cowpea 22. (Zablotowicz et al, 1981). 3.322 Energetics. From various experiments involving legumes, the evidence obtained strongly indicates that the major limi­ tation in dinitrogen fixation of all the factors investigated is the energy source, photosynthate, available to the nodules (Hardy and Barelka, 1977; Hardy, 1977). However, other factors must also be considered because of their potential adverse effects on plant growth and photosynthesis and, thus, on dinitrogen fixation (Hardy, 1977, 1980a; Andersen and Shanmugam, 1977; Andersen et al, 1981). Measurements on the time course of dinitrogen fixation for field grown soybean (Hardy et al, 1968; Weber et al, 1971; Ham et al, 1976), pea (LaRue and Kurz, 1973), dwarf bean (Bethlenfalvay and Phillips, 1977; Graham and Halliday, 1977), cowpea (Vigna unguiculata (L.) Walp.) (Herridge and Pate, 1977) and white lupin (Lupinus albus L.) (Pate and Herridge, 1978) show a general exponential increase until pod-filling after which there is a rapid decrease. These results demonstrate that photosynthate is a limiting factor in dinitrogen fixation, when the demand for photosynthate by reproductive organ becomes large. Indirectly, enrichment of carbon dioxide to the canopy of field-grown legumes from pod-filling to senescence also increases dinitrogen fixation (Hardy and Havelka, 1975,1977; Havelka and Hardy, 1976) by reducing photorespiration and increasing photosynthate avail­ able to the nodules. Similarly, photosynthesis and dinitrogen fix­ ation are increased by supplementing light intensity (Lawn and Brun, 1974) or by grafting a second shoot on a rootstock (Streeter, 1974), while they are decreased by reducing light intensity or shading (Hardy et al, 1968; Chu and Robertson, 1974; Sloger et al, 1975) and defoli­ ation (Hardy et al, 1968; Moustafa et al, 1969; Chu and Robertson, 1974). In a more direct study involving the measurement of soluble carbohydrate, Graham and Halliday (1977) found a strong positive correlation between acetylene reduction and soluble nodule carbohydrate in Phaseolus spp. Phillips (1981), in a more recent experiment, showed that nitrogen was the chief limiting factor during the early growth stages in soybean. It appears, therefore, that carbon is a limiting factor at the later growth stages when the dinitrogen-fixing system has greatly developed and is more active in fixation. Various workers have estimated the energy cost of dinitrogen 23. fixation in whole plant studies using different techniques which pro­ duced striking differences in results (Phillips~ 1980; Schubert and Ryle, 1980). However, measurements based on intact root systems can be taken with reasonable confidence. Three such experiments involving intact root systems are summarised by PateandMinchin (1980) (Table 3.1). In the table, both the carbon consumed-nitrogen fixed ratio and percentage of net photosynthate utilized by nodules vary consid­ erably probably due mainly to differences in plant species, rhizobia strain and length of growth period. The net photosynthate utilized by nodules ranges from 10 to 32% and, thus, suggests that dinitrogen fixation is an energy-intensive process (Schubert and Ryle, 1980). Although the three species show large differences in the quantities of carbon used in the export of fixed nitrogen, in nodule respiration and in the accumulation of nodule dry matter, the proportion of consumed carbon used in these three activities are similar. Between 48 to 52%, 36 to 40% and 9 to 16% of the consumed carbon are used in the export of fixed nitrogen from the nodules, in nodule respiration lost and in accumulation of nodule dry matter respectively. These variations among species may be explained by the differences in nitrogen-metabolic pathways, and the complex relationships in dinitrogen fixation, hydrogen gas evolution and hydrogenase activity (Pate and Minchin, 1980). Cellular estimates of the energetics of the fixation process in vi va are based on the stoichiometry of ATP hydrolysis using a reductant and purified nitrogenase system. The overall reaction (equation 3) is given by the reduction of dinitrogen (equation 1) and reduction of protons (equation 2) as shown below (Phillips, 1980):- 1 N2 + 12ATP + 6e + 8H+ -+ 2NH4+ + 12ADP + 12Pi 2 + 4ATP + 2e + 2H+ - H2 + 4ADP + 4Pi 3 N2 + 16ATP + Se + lOH+ - 2NH4 + + H2 + 16ADP + 16Pi Although 16 mole ATP mole -i N2 fixed is the total energy involved in the above stoichiometry, between 12 to 24 mole ATP mole- 1 N2 fixed is - the commonly cited range (Brill , 1977,1980), depending on the reaction conditions (Schubert and Ryle, 1980). The ATP-ADP ratio in nodules has been shown to relate directly to dinitrogen fixation (Appleby et aZ, 1975; Ching e t aZ, 1975). - TABLE 3.1 CARBON AND NITROGEN IN NODULES OF THREE LEGUMES (AFTER PATE AND MINCHIN; 1980) Cowpea White lupin Period of growth (days) 0 - 78 0 - 94 N2 fixed (mg plant- 1 ) 726 788 C used in export of fixed N from nodules (mg plant- 1 ) 969 (48%) 1789 (52%) C lost as C0 2 in nodule respiration (mg plant- 1 ) 789 (39%) 1372 (40%) C incorporated into nodule dry matter (mg plant- 1 ) 253 (13%) 298 (9%) Total C consumed by nodules (mg plant- 1 ) 2011 (100%) 3459 (100%) Plant net photosyn t hate 2.8 4.4 utilized by nodules (%) 10.2 13.0 24. Garden pea 21 - 30 27 53 (48%) 40 ( 36%) 18 (16%) 122 (100%) 4.1 32.0 25. Sainfoin, growing in a control led environment (Burton and Curley, 1968) and in the field (Sims et al, 1968), has been reported to develop symptoms of nitrogen deficiency although the plants we re well-nodulated. The deficiency was readily corrected with supplemen­ tary combined nitrogen (Sims et al, 1968). In a more recent study, Hume (1981) f ound that sa infoin plants depending solely on the dinitro­ gen fixed in the nodu les were capable of fixing on ly approximately 43% of the amount of nitrogen requi red. Wh en comparing these plants with those that were supp li ed with combined nitrogen , their relative growth rates were lower than the l atter. It is, therefore, eviden t that dinitrogen fi xation in sa infoin i s relatively inefficient due probably to inefficient strains of rhizob ia (Burton and Curley , 1968 ) and inadequa te supply of energy from photosynthesis (Hume , 1981). 3.33 Methods of Improving Dinitrogen Fi xation The many possibilities of enhancing dinitrogen fixation in legumes have been examined by Evans (1975) , Brown et at, (1975) , Postgate (1977 ,1978) , Hardy and Havelka (1977) , and Subba Rao (1980) . The principal, short- term method i s by ame li orating the efficiency of energy consumpti on by the nodu les. Theoreti cally and to a certain extent, pract ically, t his is achieved when the energy supply to the nodules is increased , and the energy wastage in hydrogen evo l ution is reduced (Figure 3.1). 3.331 Photosynthesis. As discussed earli er , dinitrogen fixati on is strongly dependen t on photosynthesis. This was undoubted ly demonstrated in experiments in which dinitrogen fixation was greatly enhanced by li ght supp lement, carbon dioxide enrichment and shoot grafting on a rootstock. Although these methods are not practical in extensive agricultural production systems, the underlying pri nci pl e i s the same, that is, to create an efficien t photosynthetic sys tem (Hardy, 1977) . This can be achieved through plant breeding involving three major approaches -(1) optimization of the canopy structure, primarily to attain an optimal l eaf area index and better canopy architecture to increase light intercepti on and absorption (Nasyrov , 1978; Gifford and Evans , 1981), (2) improvemen t of efficiency of the photosyntheti c process, with minimal photorespirati on and dark respir­ ation to reduce energy lost (Zelitch, 1971; Postgate, 1977; Tolbert , LIGHT 26. CANOPY STRUCTURE PHOTOSYNTHETIC EFFICIENCY H .. \2 SOURCE-SINK RELATIONSHIP ' 177177 7 JJ 7 7 7 7 7 ?J/7 ;7?7777/77717 I ------,,,-;,, E = ENERGY SUPPLY FIGURE 3.1 FACTORS AFFECTING THE ENERGY SUPPLY FOR DINITROGEN FIXATION IN A LEGUME NODULE 27. 1977), and (3) rational partition and usage of assimilates for maximum expression of yield (Nasyrov, 1978; Gifford and Evans, 1981) and, thus, in the case of forage legumes the shoot dry weight. 3.332 Phosphate supply. In many experiments, nodulation and dinitrogen fixation in legumes are closely related to phosphate fertil­ ization (van Schreven, 1958; Vincent, 1965; Gukova and Arbuzova, 1969; Gates, 1970,1974; Fillery, 1974; Andrew, 1977; Munns, 1977; Robson, 1978; Andrew and Jones, 1978; Hanson, 1978-1979). This is because of the common limiting phosphate supply from the soil solution in many soils (Dudal, 1976). Since phosphorus is a constituent of many metabolic compounds associated with dinitrogen fixation (Martin and Bushman, eds, 1978), it is very likely that phosphorus is inadequate for the synthesis of these metabolites especially ATP, the energy­ yielding compound required in large quantity (Mortenson, 1966; Bullen and LeComte, 1966). Moustafa et al, (1971) have positively correlated the nodule ATP, ADP and AMP with the foliar-applied labelled ortho­ phosphate in lotus major, but the relationships between inorganic soil phosphate supply and ATP and dinitrogen fixation are still to be i nves ti gated. 3.333 Hydrogenase activity. Part of the dinitrogen-fixing process in legumes is associated with the energy-wateful reduction of protons to hydrogen gas. In this reaction, up to 12 mole ATP mole- 1 N2 fixed or an equivalent of up to 50% of the total energy used in fixation is lost (Schubert and Evans, 1976,1977). However, the bac­ teroids of certain rhizobia strains possess another enzyme, hydrogenase, which is capable of oxidizing the hydrogen gas into protons, thereby also synthesizing ATP. Additional evidence also shows that different legumes inoculated with the same strain of bacteria produced nodules with different levels of hydrogenase activity, indicating that host plants play a role in the exppression of nitrogenase (Dixon, 1972; Schubert and Evans, 1976,1977). In the range of legumes surveyed, the hydrogenase activity in each legume species varies widely and, in general, this activity is greater in tropi ea 1 1 egumes such as soybean and cowpea than in temper- ate legumes (Evans et al, 1977). In certain tests, nodules of soybean and cowpea are also capable of taking up hydrogen gas from their external surroundings in the presence of oxygen (Schubert and Evans, 28. 1976, 1977). When determining the fate of recycled energy in the soybean cultivar Anoka-R. japonicwn strain USDA 110 system, part of the energy was channelled to dinitrogen fixation which showed a 7% increase, and the balance to dry matter accumulation which recorded a 24% increase (Schubert and Ryle, 1980). These results, therefore, show that more research ought to be conducted to identify and develop more energy-efficient systems of legume cultivars and rhizobia strains, capable of high hydrogenase activity. 3.334 Improving the dinitrogen-fixing efficiency of sainfoin. Sainfoin, being a relatively new forage crop, has not been well-selected for its dinitrogen-fixing capacity. Burton and Curley (1968) have studied the efficiency of the dinitrogen-fixing systems of various sainfoin cultivars using different rhizobia isolates, but no attempt was made to select or breed for efficient fixation. It is, therefore, apparent that such research endeavour is a requisite for further level of sainfoin yield. The prospective areas of research in plant breeding include the screening and selection for a lirger leaf area index, greater stem production, low hydrogen evolution, and more efficient rhizobia strains. Compared with lucerne, Sheehy and Popple (1981) have, in fact, demon­ strated a lower rate of canopy photosynthesis in sainfoin (var. bifera Hort.) because of its considerably lower leaf area index. This lower leaf index as compared to lucerne was also reported in the cultivars Remont, Melrose and Fakir (J.A. Fortune, unpublished data), and Hume (1981) has found that the leaf area of Fakir was positively correlated with dinitrogen fixation (R = 0.74, P <0.001). Currently at Massey University, an initial screening programme for cultivars with a larger leaf area index is in progress (N.J. Withers, personal communication). 3.4 MYCORRHIZA SYSTEM OF LEGUMES 3.41 Occurrence Some 200 families and 1 OOO genera of plants (Trappe and Fogel, 1977) have been identified as developing arbuscular, orchidaceous, ericoid and arbutoid rnycorrhizae (Lewis, 1975). This thesis concerns only with the arbuscular type. The phenomenon of arbuscular mycorrhizae is ubiquitous, occurr­ fog in numerous terrestrial species of the Bryophyta (liverworts, mosses and hornworts), Polypodiophyta (ferns), Pinophyta (gymnosperms), 29. and Magnoliophyta (angiosperms) (Gerd~mann, 196~,1974,1975). They are formed by some 90 distinct taxa of non-specialized fungi (Hall and Fish, 1979; Hall, in preparation) of the Endogonaceae, a family of the Phycomycetes (Mosse, 1956; Nicolson and Gerdemann, 1968; Gerdemann, 1971). Arbuscular mycorrhizae are of particular interest because of their widespread occurrence in numerous agricultural (Gerdemann, 1968; Mosse, 1973; Safi r, 1980) and horti cul tura 1 (Maronek et al, 1980; Hayman, 1981) crops. In the Fabaceae, they are found in most species of the two subfamilies, the Mimosoideae and the Papilionoideae (Malloch et al, 1980), including those in the agriculturally important genera Arachis, Glycine , Lens, Lotus, Medicago, Onobrychis , Phaseolus , Piswn, Trifoliwn, Trigonella and Vicia (Jones, 1923,1924; Peyronel, 1924; Samuel, 1926; Asai, 1944; Strzemska, 1969a,b, 1970,1973,1975b; Crush, 1975). Of the Onobrychis spp., sainfoin is the only one investigated (Strzemska, 1975b). Members of the third subfamily, the Caesalpinioideae, also develop arbuscular mycorrhizae except those in the tribes, Amherstieae and Detarieae; those regularly form ecto­ mycorrhizae (Malloch et al, 1980). 3.42 Development of Arbuscular Mycorrhizae On the basis of growth and development of the endophytes, arbuscular mycorrhiza formation is described in four phases -­ recognition, infection, arbuscle and vesicle production, and extra­ matrical development. All these four phases are found typically in different plant species. In the following account, therefore, data from non-legumes are also used wherever necessary. 3.421 Recognition. In the rhizosphere, spore germination and subsequent growth are stimulated by adjacent root exudates (Bevege and Bowen, 1975; Hepper and Mosse, 1975), decayed organic litter (Nicolson, 1958,1960,1963; Guttenbergt 1963; Went, 1974) particularly disintegrated roots, and the presence of other soil microorganisms (Mosse, 1956,1959,1962). Certain unknown compounds are particularly crucial as indicated in experiments involving artificial cultures, in which spore germination and subsequent fungal growth are extremely limited (Godfrey, 1957b; Mosse, 1963). The germ tubes produced from spores become the first hyphae which grow and branch, resulting in the 30. pre-infection mycelia (Bevege and Bowen, 1975; Hepper and Masse, 1975). The endophytes generally are known to infect a wide range of hosts (Gerdemann, 1955; Koch, 1961; Mosse, 1962,1963,1973,1975), but certain species are reported to exhibit some degree of specificity (Tolle, 1958). However, zero infection is less common than restricted infection, that is, "preferenti al" association (Hawker, 1962). Specificity in the recognition phase is the result of root­ endophyte interaction involving root exudates and poss ibly other physiological properties as shown in the study of Bevege and Bowen (1975). Using clover, onion (AUiwn aepa L.), common wall cress (Arabidopsis thaliana (L.) Heynh.), and pine (Pinus radiata Don.), inoculated with three arbuscular fungal species (white-reticulate spore type, Endogone arauaareae Bevege, GZomus mosseae (Nicol. & Gerd.) Gerdemann & Trappe), they observed that all these endophytes developed pre-infection growth and later infected the clover and onion roots in varying degrees, depending on the endophyte species. Ho1t1ever, these endophytes fail ed to infect the roots of common wall cress and pine although pre-infection myceli a were present except for the white­ reticul ate type which even failed to germinate in the rhizosphere of common wall cress . Despite the absence of infection in pine, E. arauaarea formed external vesicles besides hyphae, while G. mosseae also developed occasional appressoria in addition to hyphae and external vesicles around the roots. These cross-inoculation re sults suggest that the selection of highly susceptible hos t cultivars and highly infective fungal species is likely to effect a more intensive infection. 3.422 Infection. Infection begins when hyphae, produced from spores (Manjunath and Bagyaraj, 1981) or infected roots (Hirrel and Gerdemann, 1979; Heap and Newman, 1980; Manjunath and Bagyaraj, 1981), attach to the epidermal or root-hair surface (Nicolson, 1959; Davidson and Christensen, 1977) and produce appressoria, from which projections penetrate the wall and form the invading hyphae (Hepper and Masse, 1975) of 6 to 12 µm across (Jones, 1924). The penetration process is probably aided by enzymes and mechanical pressure (Kaspari, 1975). These aseptate, yellow hyphae become coarser, measuring 12 to 13 µm in diameter (Jones, 1924), and subsequently traverse and ramify the root cortex intercellularly and intracellularly (Jones, 1923, 1924; Samuel, 1925; Nicolson, 1959), developing an infection unit 31. (Cox and Sahders, 1974) often with coiled hyphae (Jones, 1924; Nicol­ son, 1959; Kinden and Brown, 1975a). The host cells react to invasion with an increase in cytoplasm and an accumulation of all organelles such as mitochondria, dictyosomes, and endoplasmic recticulum cisternae in the vicinity of the hyphae (Kaspari, 1975). Ultrastructural studies showed a continual digestion of the intracellular hyphae by the host cells (Scannerini et al , 1975; Kinden and Brown, 1975b), but the host plasmalemma remained morphologically and cytochemically unchanged in the interaction (Bonfante-Fasolo et al, 1981; Marx et ~l, 1982) in contrast to host-pathogen interactions (Littlefield and Bracker, 1972; Gil and Gay, 1977), indicating a mutualistic symbiosis instead of a true host-pathogen association. The degree of infection spread in the cortex depends on the interaction between the host species and fungal species. Bevege and Bowen (1975) observed a very rapid infection spread in cl over but a localised spread in onion for three different fungal species. Khan (1980) obtained a general infection spread in onions with Glomus fas aiaulatus (Thax. sensu Gerd.) Gerdemann & Trappe, Glomus mosseae , and Glomus maarocarpus Tul. & Tul. var . maarocarpus, but a localised one with Saleroaystia rubifosmis Gerdemann & Trappe. Generally, the spread is greater in the outer cortex than in the middle or inner cortex (Jones, 1924; Nicolson, 1959). 3.423 Arbuscle and vesicle production. In the middle and inner cortex, and to a certain extent in the outer cortex, the inter­ cellular hyphae parallel to the root axis produce lateral branches that penetrate further cortical cells and form arbuscles (Peyronel, 1923; Jones, 1924; Nicolson, 1959; Kinden and Brm\lTl, 1975a). Each arbuscle develops within a host cell by repeated dichotomous branching of the invading hypha to produce a cluster of fine filaments (Kinden and Brown, 1975c). The host nucleus is also significantly enlarged with a prominent nucleolus (Mosse, 1963). At a later stage, the arbuscles are digested within the host cells, forming a granular mass (Kaspari, 1975; Scannerini et al, 1975). Arbuscular digestion is initiated at the tips of the finest branches and progresses basipetally (Kinden and Bro¼f!, 1976). These greenish-yellow arbuscles (Jones, 1923) are structurally analogous to haustoria of the fungi in the family, Pero­ nosporaceae (Peyronel, 1923), and are probably the sites of photo­ synthate and nutrient exchange, and especially phosphate (Cox and Tinker, 1976; Tinker, 1978). 32. At the later stage of mycorrhiza formation, vesicles may develop in the middle or from the tips of hyphae, mainly in the intercellular spaces of the outer cortex where few arbuscles are present (Jones, 1924; Kinden and Brown, 1975a,b). They vary in shape but typically are ovate and vary in length from 25 to 65 µm (Jones, 1924). Their functions are associated with possibly the t emporary storage of certain compounds as indicated by the presence of oil droplets within (Kinden and Brown, 1975b), and in reproduction as shown by the germination and growth of single excised vesicles (Koch, 1961 ). Both the number of arbuscles and vesicles produced varies con­ siderably in different roots of the same plant (Jones , 1924) and also in different species (Jones, 1924; Samuel, 1926; Bevege and Bowen, 1975). Such variations are clearly due to host and endophyte inter­ actions (Bevege and Bowen, 1975; Khan, 1981). 3.424 Extramatrical development. While the internal fungal structures are being formed in the cortex, hyphae grow out into the soil to form a loose network of mycelium. These extramatrical hyphae are dimorphic with a system of coarser hyphae, 20 to 30 µm across, thick-walled and aseptate, and a system of finer hyphae, 2 to 7 µm across, thin-walled and septate with time (Masse, 1959; Nicolson, 1959, 1963). These hyphae extend several centimetres in to the soil and contribute to the entire nutrient absorptive system of the plant (Hattingh et al~ 1973; Pearson and Tinker, 1975; Rhodes and Gerdemann, 1975; Cooper and Tinker, 1978). The extent of extra­ matrical mycelium development differs enormously among different endophytes (Bevege and Bowen, 1975; Khan, 1981) . The coarser hyphae also produce vesi cles, single chlamydospores (resting spores) and sporocarps (fruiting bodies) (Nicolson, 1967). The sporocarps contain chlamydospores, zygospores or sporangia, but sometimes chlamydospores and zygospores together (Godfrey, 1957; Gerdemann, 1965). These chlamydospores, zygospores and sporangio­ spores of the sporacarps, and the single chlamydospores are able to produce new infections in adjacent roots (Gerdemann, 1955,1961; Gerdemann and Nicolson, 1963; Mosse and Bowen, 1968). The morphological features of these spores such as size, shape, wall structure, origin, as well as the appearance of the hyphae from which they arise have been employed to distinguish different spore groups (Gerdemann and Nicolson, 1962,1963; Mosse and Bowen, 1968) 33. which became the foundation for later endophyte classification into various genera and species (Gerdemann and Trappe, 1974,1975; Hall and Fish, 1979; Hall, 1982; Hall, in preparation). Besides these keys, a special slide collecti on has been prepared also for the identification and classification of arbuscular endophytes (Hal l and Abbott, in preparation). 3.43 Arbuscular Mycorrhizae and Phosphorus Nutrition 3.431 Soil phosphates and organic phosphorus . Soi l phos- phates (inorganic phosphorus compounds) are recognised to exist in three defined pools in equilibrium relationsh ips (Sutton and Gunary, 1969) as shown in Figure 3.2. The soil solut ion pool also links to the organic pool existing as organic matter (Dalal, 1977). The non-labil e pool is by far the larges t inorganic fraction (Larsen, 1967,1976), in which phosphates are he l d mai nly as calcium, fluorine, aluminium and iron compounds in mineral lattices such t hat they are not immediately exchangeable \'lith ions in the soil solution (Larsen, 1967). These compounds are only sparingly soluble and, thus, their solubility in the soi l solution is strongly influenced by the soil pH and the action of microorganisms (La rsen, 1967 ) . The fertilizer known as rock phosphate, being sparingly so lubl e, is also considered in this category (Tinker, 1975a). The labile pool describes the phosphate component that can be released into the soil solution (Sutton and Gunary, 1968). The size of this pool may be determined by isotopic dilution of label led phos­ phorus such as 32 P, with the pool sampled either phys ically via the soil solution, in which case the result is called E value, or biologi­ cally by growing a test plant in the soi l, in which case the result i s called L value (Larsen, 1967). If the E and L values are closely similar, then the test plant is absorbing only from the soil solution or the labile pool in equilibrium with it (Sanders and Tinker, 1971). The soil solution is the pool in which plants obtain their phosphate supply as orthophosphates mainly in the forms H2P04- and HP042- (Larsen, 1967). The concentration of total phosphorus in the soil solution is extremely low, usually between 0.1 and 10.0 µM, equivalent to less than 60 µg P kg- 1 moist soil or about 0.005% of the soil phosphorus (Pierre and Parker, 1927; Fitter and Hay, 1981). 34. 77T?J 77 I 7 JI TI 7 77777 7777 ORGANIC MATTER SOIL J SOLUTION ------~-; ~----~-----, NON-LABILE POOL FIGURE 3.2 LABILE POOL MODEL OF SOIL PHOSPHATES AND INORGANIC PHOSPHORUS IN DIFFERENT POOLS (THE THICKNESS OF THE ARROWS INDICATES THE BALANCE OF THE EQUILIBRIUM) (MODIFIED AFTER FITTER AND HAY, 1981) HYPHA 0 0.5 1.0 mm FIGURE 3.3 ROOT HAIRS AND EXTRAMATRICAL MYCELIUM IN THE SOIL (MODIFIED AFTER POWELL, 1975b) 35. Phosphates, the inorganic fraction, constitutes less than 0.1 µM (Pierre and Parker, 1927). As shown in Table 3.2, the availability of phos­ phates for plant uptake depends on the replenishment from the labile and, to a limited extent, the non-labile pool. The organic phosphorus pool may comprise between 20 and 80% of the total soil phosphorus (Wells and Saunders, 1960), and consists of organic as well as organic-inorganic fractions (Larsen, 1967; Dalal, 1976,1977). The compounds so far identified are inositol phosphates (0.4 to 83.0%), phospholipids (0.5 to 7.0% ), and nucleic acids (0.2 to 0.4%) (Andersen, 1967; Dalal, 1976,1977). About half the soil organic phosphorus is still in an unknown form (Andersen and Malcolm, 1974). 3.432 Absorption by uninfected roots . Phosphate ions and complexes, like all other mineral nutrients, move through t he soil to the root surface by convection (mass flow) and diffusion (Barber, 1962; Olsen and Kemper, 1968). Although the behaviour of t he soil-root interface is still ambiguous (Oades, 1978; Greenland, 1979) , actively absorbing plant roots are known t o quickly deplete phosphate in the solution in their immediate vicinity (Barber et al, 1963). For in­ stance, the amount of phosphorus needed by a crop of corn fo r a yield of 9 500 kg ha- 1 is 38 kg ha- 1 , but the approximate amount supplied by convection is only 2 kg ha- 1 (Barber and Olson, 1968) . The remain­ ing phosphate supply, therefore, depends on the release of phosphates from the labile pool and diffusion of these released phosphates to the root surface . However, because phosphates are extreme ly immobile , the rate of phosphate diffusion to the roots is the slowest (Nye, 1966) , with a diffusion coefficient of 0.3 to 3.3 x 10- 9 cm2 s- 1 (Rowell et al, 1967) compared to that of nitrate (1.0 x 10- 6 cm 2 s- 1 ) (Nye, 1966) and potassium (1 .0 to 28.0 x 10- 8 cm 2 s-1 ) (Drew et al, 1969). In such circumstances, root hairs play a supplementary role in phosphate absorption to meet the plant's demand (Bouldin, 1961; Nye, 1966,1969; Barley, 1970; Drew and Nye, 1970; Newman and Andrews, 1973, Phat and Nye, 1974a,b). The greater zone of phosphate depletion around root hairs as shown by autoradiographs (Lewis and Quirk, 1967a, b; Bhat and Nye, 1973) is due primarily to the greater length of root hairs that penetrate further into adjacent soils, generating a more effective soil-root interface, thereby decreasing the distance for 36. phosphate diffusion to the root surface (Nye, 1966; Brewster et aZ, 1976b). However, other reasons such as the larger total root absorbing surface area per unit radius (Bouldin, 1961; Bhat and Nye, 1974b; Brewster et al, 1976a). greater total volume of soil exploited (Nye, 1966; Bhat and Nye, 1974b) and possibly also greater surface root exudations that aid in solubilizin